Here, we report that adenine nucleotides decreased, whereas carboxyatract-yloside increased, Ca2+ uptake capacity in mitochondria isolated from Artemia embryos.. Transmission electron mi
Trang 1from Artemia franciscana embryos is associated with
insensitivity to bongkrekate and atypical effects of
Csaba Konra`d1, Gergely Kiss1, Beata To¨ro¨csik1, Ja´nos L La´ba´r2, Akos A Gerencser3,
Miklo´s Ma´ndi1, Vera Adam-Vizi1and Christos Chinopoulos1
1 Department of Medical Biochemistry, Semmelweis University, Budapest, Hungary
2 Research Institute for Technical Physics and Materials Science, Budapest, Hungary
3 Buck Institute for Age Research, Novato, CA, USA
Introduction
Embryos of the brine shrimp Artemia franciscana
toler-ate anoxia at room temperature for several years [1,2],
by bringing their metabolism to a reversible standstill,
with no evidence of apoptotic or necrotic cell death
[3] While doing so, they maintain viability under con-ditions that are known to open the so-called mitochon-drial permeability transition pore (PTP) in mammalian species [4–6] This pore is of a sufficient size (cut-off of
Keywords
adenine nucleotide carrier; adenine
nucleotide translocator; bongkrekic acid;
diapause
Correspondence
C Chinopoulos, Department of Medical
Biochemistry, Semmelweis University,
Budapest, Hungary
Fax: +361 2670031
Tel: +361 4591500; ext 60024
E-mail: chinopoulos.christos@eok.sote.hu
(Received 22 October 2010, revised 30
November 2010, accepted 23 December
2010)
doi:10.1111/j.1742-4658.2010.08001.x
Mitochondria isolated from embryos of the crustacean Artemia franciscana lack the Ca2+-induced permeability transition pore Although the composi-tion of the pore described in mammalian mitochondria is unknown, the impacts of several effectors of the adenine nucleotide translocase (ANT) on pore opening are firmly established Notably, ADP, ATP and bongkrekate delay, whereas carboxyatractyloside hastens, Ca2+-induced pore opening Here, we report that adenine nucleotides decreased, whereas carboxyatract-yloside increased, Ca2+ uptake capacity in mitochondria isolated from Artemia embryos Bongkrekate had no effect on either Ca2+ uptake or ADP–ATP exchange rate Transmission electron microscopy imaging of
Ca2+-loaded Artemia mitochondria showed needle-like formations of elec-tron-dense material in the absence of adenine nucleotides, and dot-like for-mations in the presence of adenine nucleotides or Mg2+ Energy-filtered transmission electron microscopy showed the material to be rich in calcium and phosphorus Sequencing of the Artemia mRNA coding for ANT revealed that it transcribes a protein with a stretch of amino acids in the 198–225 region with 48–56% similarity to those from other species, includ-ing the deletion of three amino acids in positions 211, 212 and 219 Mito-chondria isolated from the liver of Xenopus laevis, in which the ANT shows similarity to that in Artemia except for the 198–225 amino acid region, demonstrated a Ca2+-induced bongkrekate-sensitive permeability transition pore, allowing the suggestion that this region of ANT may con-tain the binding site for bongkrekate
Abbreviations
ANT, adenine nucleotide translocase; BKA, bongkrekic acid; CaGr-5N, Calcium Green 5N hexapotassium salt; cATR, carboxyatractyloside; CypD, cyclophilin D; EFTEM, energy-filtered transmission electron microscopy; [Mg2+] f , free mitochondrial [Mg2+]; PTP, permeability transition pore; TEM, transmission electron microscopy; DWm, mitochondrial membrane potential.
Trang 21.5 kDa) to allow the passage of solutes and water,
resulting in the swelling and ultimate rupture of the
outer membrane Almost all studies on the mammalian
PTP concur on the conditions that open or inhibit the
pore [7,8] However, PTP characteristics in
nonmam-malian species show significant deviations from the
mammalian consensus For example, mitochondria
from the yeast species Saccharomyces cerevisiae have a
PTP that is inhibited by ADP and has comparable size
exclusion properties to the homologous structure in
mammalian mitochondria, but these mitochondria are
cyclosporin A-insensitive [9–11] Mitochondria isolated
from pea stems (Pisum sativum L.) and potatoes
(Sola-num tuberosumL.) require dithioerythritol for the
cyclosporin A to inhibit the PTP [12,13] In contrast,
cyclosporin A failed to afford protection from the PTP
in wheat (Triticum aestivum L.) mitochondria, even in
the presence of dithioerythritol [14] Furthermore, no
Ca2+-induced PTP could be found in mitochondria
from the yeast Endomyces magnusii [15–17] Likewise,
no Ca2+-induced PTP could be found in mitochondria
from embryos of the crustacean A franciscana [18]
The lack of a Ca2+-inducible PTP in embryos of
A franciscanamarks a cornerstone in our
understand-ing of the long-term tolerance, extendunderstand-ing for years,
to anoxia and diapause, conditions that are
invaria-bly accompanied by large increases in intracellular
Ca2+[3]
Despite intense research on the mammalian PTP
since its characterization by Hunter and Haworth in
1979 [19–21], the identity of the proteins comprising it
is debated; the voltage-dependent anion channel,
hexo-kinase, creatine hexo-kinase, the mitochondrial peripheral
benzodiazepine receptor, adenine nucleotide
translo-case (ANT), cyclophilin D (CypD) and the phosphate
carrier have all been proposed to participate in the
for-mation of the pore [7,22] Recent findings excluded the
voltage-dependent anion channel (all isoforms) [23],
CypD [24–27] and ANT (isoforms 1 and 2) [28] as
being the constituents of the pore itself, although
CypD and ANT have gained support as playing a
modulatory rather than structural role in pore
forma-tion [28–30] The modulatory role of ANT has been
firmly established by extensive literature on the effects
of its ligands on mitochondrial Ca2+ uptake capacity
Mitochondrial Ca2+ uptake capacity is defined as the
amount of Ca2+that mitochondria sequester, prior to
opening of the PTP PTP inhibitors increase, and
acti-vators decrease, this cumulative bioenergetic
parame-ter Regarding ANT, three endogenous ligands – ADP,
ATP (both inhibiting the PTP) [4,31], and acyl-CoA
and its esters (opening the PTP) [32,33] – plus four
poisons – atractyloside, carboxyatractyloside (cATR)
(both favoring pore opening), bongkrekic acid (BKA) and isobongkrekic acid (both promoting pore closure) [34–36] – have been identified Other, less well-charac-terized, inhibitors of ANT have also been reported [37] Mindful of (a) the well-established ligand profile
of ANT, (b) the modulatory role of ANT in the mam-malian PTP, and (c) the absence of a Ca2+-induced PTP in mitochondria from the embryos of A francis-cana, we investigated the effect of ANT ligands on
Ca2+ uptake capacity in mitochondria isolated from brine shrimp embryos We also showed that the matrix
Ca2+ precipitates show needle-like morphology in the absence of adenine nucleotides or Mg2+ but dot-like structures in their presence, unlike the ring-like struc-tures observed in mammalian mitochondria [38–40]
By sequencing of the mRNA coding for ANT in this organism, we show that the complete coding sequence
is dissimilar to those from human, mouse, Xenopus, Drosophila, and many other species, which are them-selves similar to each other Specifically, protein sequence comparison revealed a 28 amino acid region comprising positions 198–225 in Artemia ANT that shows only 48–56% similarity to those from other spe-cies, including the deletions of four amino acids Finally, we show that the ADP–ATP exchange rate mediated by ANT expressed in mitochondria of
A franciscana and Ca2+ uptake capacity are insensi-tive to BKA Resistance to BKA may be a direct con-sequence of the unique con-sequence of the Artemia ANT
Results
Effect of adenine nucleotides on Ca2+
sequestration of Artemia mitochondria
It has been well established that in mammalian mito-chondria, adenine nucleotides increase Ca2+ uptake capacity [4,38,39] In order to investigate whether this also applies to Artemia mitochondria, we tested the effect of ADP and ATP in the presence and absence of the ANT inhibitory ligand cATR, and of the F0F1 -ATP synthase inhibitor oligomycin The results are shown in Fig 1 In Fig 1A, ADP was present prior to the addition of mitochondria in all traces In the pres-ence of ADP (Fig 1A, trace a) when neither cATR nor oligomycin was present, a clamped [Ca2+] is diffi-cult to achieve, owing to the interconversions of ADP
to ATP by mitochondria, as these two nucleotides show different Kdvalues for Ca2+ When cATR or oli-gomycin was present, the amount of ADP was assumed to be static (see below), and therefore the esti-mations of free extramitochondrial Ca2+ were reliable
In the presence of ATP (Fig 1B), as the mitochondrial
Trang 3membrane potential (DWm) did not exceed the reversal
potential of ANT (see Fig 3A), the amount of ATP
added was assumed to be static, assisting the reliable
calculations of the total amount of CaCl2added What
is apparent from Fig 1A,B is that both ADP and
ATP significantly decreased Ca2+uptake rates as
com-pared with the condition in which adenine nucleotides
were absent (Fig 1C), and thereby Ca2+ uptake
capacity The effect of ADP was considerably
miti-gated by cATR and oligomycin (Fig 1A), implying
that ADP mediated its effect after being taken up by
mitochondria, most likely through ANT Inhibition of
F0F1-ATP synthase by oligomycin also lead to
cessa-tion of the funccessa-tion of ANT [41] It is apparent from
Fig 1C that even cATR alone slightly accelerated mitochondrial Ca2+ uptake, in the absence of nucleo-tides We conservatively attributed this to the inhibi-tion of mitochondrial ADP or ATP uptake (depending
on the prevalent DWm) by cATR, thereby eliminating any effect of nucleotides released from broken mito-chondria in the suspension The effect of oligomycin alone is hard to predict, because this inhibitor blocks both ATP formation by polarized mitochondria and ATP hydrolysis by depolarized mitochondria found in the same suspension It is of note that BKA had no effect as compared with its vehicle (5 mm ammonium hydroxide; not shown), but it also failed to inhibit the ADP–ATP exchange rate of Artemia mitochondria (see below)
In summary, Fig 1 shows that exogenously added adenine nucleotides decrease Ca2+ uptake rate and capacity in mitochondria isolated from embryos of
A franciscana, a phenomenon that is apparently at odds with the mammalian consensus
Fig 1 Effect of ANT ligands on Ca2+ uptake capacity in Artemia
mitochondria (A) Reconstructed time courses of extramitochondrial
[Ca 2+ ] calculated from CaGr-5N fluorescence Mitochondria were
added at 50 s, and this was followed by the addition of 2 m M ADP;
200 l M CaCl2(free) was added where indicated by the arrows For
trace b (blue), 4 l M cATR was added, and for trace c (green),
10 l M oligomycin was added, followed by 2 m M ADP prior to
addi-tion of mitochondria In trace a, no inhibitors were present (B) As
for (A), but ATP was added instead of ADP (C) As for (A) and (B),
but no nucleotides were present Results shown in all panels are
representative of at least four independent experiments.
Fig 2 Absence of the PTP evaluated by 660 ⁄ 660-nm excita-tion ⁄ emission in Artemia mitochondria (A) ADP (1 m M ), oligomycin (olgm, 10 l M ), CaCl 2 (0.1 m M , free), n-butyl-malonate (nBM, 50 l M ), N-ethylmaleimide (NEM, 0.5 m M ), SF 6847 (250 n M ) and alamethi-cin (ALM, 80 lg) were added where indicated (B) CaCl2(0.2 m M ) was added as indicated by the arrows In the upper trace (black),
1 m M ADP was added prior to addition of mitochondria Alamethicin (80 lg) was added where indicated as a calibration standard of maximum swelling Results shown in both panels are representa-tive of at least four independent experiments.
Trang 4Artemia mitochondria lack the PTP
In order to confirm that mitochondria obtained from
the embryos of this crustacean lack the PTP as
origi-nally shown by Menze et al [18], we adapted our
scheme [42] demonstrating a cyclosporin A-refractory
PTP In this scheme, addition of an uncoupler in the
presence of phosphate carrier blockers to Ca2+-loaded
rat liver mitochondria previously treated with
oligomy-cin causes an immediate and precipitous opening of
the PTP As shown in Fig 2A, this was not observed
in mitochondria isolated from embryos of A francis-cana It is of note that, in the presence of oligomycin and ADP, addition of Ca2+ failed to induce an increase in light scattering (Fig 2A), consistent with the notion that ADP entering mitochondria is required for Ca2+–Pi complexation [39] Addition of the pore-forming peptide alamethicin induced mitochondrial swelling, manifested as an abrupt decrease in light scattering (Fig 2A,B) However, in accordance with the mammalian consensus, addition of ADP in the absence of oligomycin to the suspension caused Artemia mitochondria to show ‘shrinkage’ upon addi-tion of CaCl2 (Fig 2B), which is known to occur because of complexation of matrix Ca2+with Pi affect-ing light scatteraffect-ing [38,39] From Fig 2B, it is notable that addition of Ca2+ even in the absence of ADP caused a considerable increase in light scattering, although to a lesser extent than in the presence of the nucleotide This is at odds with the finding that mito-chondrial Ca2+ capacity is decreased in the presence
of adenine nucleotides, and even the volume fraction
of the calcium-rich and phosphorus-rich electron-dense material is smaller in the latter case (see Fig 5B); how-ever, the effect of alamethicin in mitochondria treated with ADP was not as great as the effect of the peptide
in the absence of the nucleotide, and therefore a reli-able comparison cannot be made
Demonstration of the function of ANT in
A franciscana by the ADP–ATP exchange rate–DWmprofile
As adenine nucleotides produced unusual effects on the Ca2+ uptake characteristics in Artemia mitochon-dria, it was important to evaluate the functional status
of ANT in these mitochondria For this, a recently described method was used [41], in which the ADP– ATP exchange rate mediated by ANT is measured as a function of DWm Such an experiment is shown in Fig 3A The ADP–ATP exchange rate mediated by ANT (in the presence of diadenosine pentaphosphate,
a blocker of adenylate kinase) was measured by exploiting the differential affinity of ADP and ATP for
Mg2+ The rate of ATP appearance in the medium fol-lowing addition of ADP to energized mitochondria was calculated from the measured rate of change in free extramitochondrial [Mg2+] by the use of standard binding equations [41] During the course of this experiment, ADP–ATP exchange rates were gradually altered by stepwise additions of an uncoupler (10 nm
SF 6847) until complete collapse of DWm In parallel experiments, DWm was measured by safranine O
Fig 3 ADP–ATP exchange rate and DWmprofile of Artemia
mito-chondria The effect of ANT ligands on the ADP–ATP exchange
rate (A) Reconstructed time courses of DWm, calculated from
safra-nine O fluorescence, and extramitochondrial [ATP] appearing in the
medium upon addition of ADP (at 150 s) calculated from
Magne-sium Green fluorescence as described in Experimental procedures.
For both traces, small arrows indicate the addition of 10 n M
SF 6847 (B) Reconstructed time courses of extramitochondrial
[ATP] appearing in the medium upon addition of ADP (where
indi-cated) in Artemia mitochondria, and effect of mitochondrial
inhibi-tors cATR (trace a), oligomycin (olgm, trace b), vehicle (5 m M
NH 4 OH, trace c) or BKA (50 l M , trace d) was added where
indi-cated (C) Reconstructed time courses of extramitochondrial [ATP]
appearing in the medium upon addition of ADP (where indicated) in
rat liver mitochondria, and effect of mitochondrial inhibitors cATR
(trace a), oligomycin (olgm, trace b), vehicle (5 m M NH4OH,
tra-ce c), BKA (50 l M , in buffer at pH 7.25, trace d), or BKA (50 l M , in
buffer at pH 7.5, trace e) was added where indicated Results
shown in all three panels are representative of at least four
independent experiments.
Trang 5Fig 4 TEM and EFTEM images of Ca 2+ -loaded Artemia mitochondria (A, B) TEM images of Artemia mitochondria loaded with
Ca 2+ , incubated in the absence (A) or presence (B) of ADP (C) TEM images of Artemia mitochondria loaded with Ca2+in the presence of 2 m M MgCl2incubated in the absence of ADP The 1-lm bar applies
to all images in (A–C) (D) Calcium map obtained from EFTEM imaging (E) Phospho-rus map obtained from EFTEM imaging (F) Pseudocolor image of (D) (G) Pseudocolor image of (E) The scale bars of (D) and (E) also apply to (F) and (G), respectively.
Trang 6fluorescence, and calibrated to millivolts as detailed in
Experimental procedures As shown in Fig 3A, ATP
appeared in the medium after ADP addition, and at
the same time there was a depolarization by 25 mV
Subsequent stepwise additions of the uncoupler
SF 6847 led to a stepwise decrease in DWm
accompa-nied by a decrease in the ADP–ATP exchange rate
This culminated at approximately ) 90 to 100 mV,
and thereafter ANT was gradually reversed The ATP
influx rate (reverse mode of ANT) was much slower
than the ADP influx rate, i.e the forward mode of
ANT From these experiments, we concluded that the
ANT of our mitochondrial preparations of A
francis-canaembryos is fully functional
ANT of A franciscana is refractory to inhibition
by BKA
As mentioned above, BKA was without an effect on
Ca2+uptake rate and capacity in Artemia mitochondria
Here, we tested whether BKA (three different LOT stocks were tested) was able to act on the fully func-tional ANT As shown in Fig 3B, addition of either cATR (trace a) or oligomycin (trace b) immediately stopped further ATP appearance in the medium, implying a cessation of ANT operation In contrast, addition of BKA (50 lm, trace d) failed to inhibit ANT operation as compared with the control (5 mm
NH4OH, which is the vehicle of BKA, trace c) With the same BKA stocks, this poison fully inhibited ANT operation in rat liver mitochondria (Fig 3C) and also induced state 4 from state 3 respiration (not shown) BKA was also tested at pH 7.5, the pH of the buffer used for experiments with Artemia mitochondria; this
is important, because BKA needs to be protonated in order to exert its action [43], and at pH 7.5 it will be less efficient Still, as shown in Fig 3C, 50 lm BKA inhibited the ADP–ATP exchange rate in rat liver mitochondria (trace e), although with a delay, as explained in [44–46], as compared with its vehicle
(tra-ce c) NH4OH at 5 mm reduced the ADP–ATP exchange rate, probably because of matrix alkaliniza-tion, in accordance with our findings reported earlier [41], however, this was not observed in Artemia mito-chondria It is also notable that at pH 7.5 (traces c and e of Fig 3C), ADP–ATP exchange rates are smaller than those obtained in buffer at pH 7.25, in
Fig 6 Effect of Ca 2+ uptake on light scattering in mitochondria iso-lated from the liver of X laevis (A) Time courses of light scattering
of X laevis liver mitochondria followed by 660 ⁄ 660-nm excita-tion ⁄ emission CaCl 2 (20 l M ) was added where indicated by the arrows Trace a, only Ca 2+ addition; trace b, Ca 2+ addition plus BKA (50 l M ); trace c, no Ca2+addition; trace d, Ca2+addition plus 1 l M cyclosporin A Cyclosporin A or BKA was present in the medium prior to the addition of mitochondria (B) Reconstructed time course
of extramitochondrial [Ca2+] obtained from CaGr-5N fluorescence Mitochondria were added at 50 s, and this was followed by addi-tion of 20 l M CaCl2, where indicated by the arrows Results shown
in both panels are representative of at least four independent experiments.
Fig 5 Quantification of the Ca 2+ –P i -rich areas of Ca 2+ -loaded
Art-emia mitochondria by adaptive thresholding (A) Images of ArtArt-emia
mitochondria loaded with Ca 2+ : (i) incubated in the absence of
MgCl 2 or adenine nucleotides; (ii) same image with adaptive
thres-holding (red); (iii) incubated in the presence of ADP; (iv) same
image as in (iii), with adaptive thresholding (red) (B) Volume
frac-tions of the electron-dense material in the mitochondria loaded with
Ca2+with or without ADP, in the absence of MgCl 2 , as calculated
by the fractional area of positive pixels [red in (A)] of the
mitochon-drion (P = 0.031 by Mann–Whitney rank-sum test; 29 TEM images
in total).
Trang 7line with the results obtained in [41] Furthermore, as
shown below, the same BKA inhibited Ca2+-induced
swelling in Xenopus liver mitochondria From the
results shown in Fig 3B,C, we postulated that the
ANT isoform(s) of A franciscana may lack a
BKA-binding site
Ca2+–Pimatrix complexation in Artemia
mitochondria shows a unique morphology
As shown above, mitochondria from the embryos of
A franciscanasequester Ca2+, although adenine
nucle-otides decrease uptake rates and capacity The effect of
ADP probably took place at the matrix side, as cATR
mitigated its action However, the effect of ATP also
seems to be mediated by a cATR⁄
oligomycin-insensi-tive mechanism Adenine nucleotide-sensioligomycin-insensi-tive site(s)
that alter maximum Ca2+ uptake capacity other than
ANT have been reported in a variety of mitochondria
[40], although their identity is still unknown The
com-plexation⁄ precipitation of Ca2+ with Pi in the
mito-chondrial matrix and the involvement of matrix
adenine nucleotides as phosphate donors have been
firmly established in mammalian mitochondria
[38,39,42] We were therefore interested in the nature of
this phenomenon in Artemia mitochondria, as the
func-tional data deviated so significantly from the
mamma-lian consensus As shown in Fig 4A, mitochondria
from the crustacean incubated in the absence of
adenine nucleotides and MgCl2 showed needle-like
electron-dense structures If ADP (Fig 4B) or MgCl2
(Fig 4C) was present during Ca2+ loading, dot-like
electron-dense structures were observed instead In
order to confirm that the electron-dense structures were
indeed Ca2+–Pi precipitates, we performed
energy-filtered transmission electron microscopy (EFTEM) of
Ca2+-loaded mitochondria in the absence of adenine
nucleotides and MgCl2, as detailed under Experimental
procedures Spatial maps of calcium and phosphorus
were recorded (Fig 4D,E), and confirmed a high
degree of colocalization (Fig 4F,G) Image stability
was insufficient (owing to very long exposure times –
10 min each – under high magnification, bar 50 nm) to
allow the same experiments to be performed in
mito-chondria loaded with Ca2+ in the presence of adenine
nucleotides or MgCl2, during which dot-like electron dense structures are observed Quantification by adap-tive thresholding (Fig 5B) revealed that the volume fraction of the electron-dense material in the volume bounded by the inner boundary membrane was signifi-cantly higher in mitochondria untreated with ADP than in those treated with the nucleotide This is in line with the experimental findings on Ca2+uptake capacity
in the presence and absence of adenine nucleotides
Mitochondria isolated from Xenopus liver reveal
a classical Ca2+-induced PTP that is sensitive to cyclosporin A and BKA
By alignment of the ANT sequences from various organisms, we deduced that the two closest homologs
of A franciscana ANT were those expressed in Drosophila melanogaster and Xenopus laevis, both of which are similar to each other but not to A francis-cana ANT regarding the 198–225 amino acid region (see below) D melanogaster may show a Ca2+ -regu-lated permeability pathway with features intermediate between the PTP of yeast and that of vertebrates (S von Stockum, personal communication) [11], but the PTP in X laevis has not been yet studied We were therefore interested in whether mitochondria isolated from tissues from X laevis show the Ca2+-induced PTP As shown in Fig 6A, when 20 lm CaCl2 was added to Xenopus liver mitochondria, a decrease in light scatter was observed (trace a) as compared with
no addition of CaCl2 (trace d) that was completely sensitive to cyclosporin A (trace c) and partially sensi-tive to BKA (trace b) From this experiment, we con-cluded that Xenopus liver mitochondria have a classical PTP that is induced by Ca2+ and is sensitive to cyclo-sporin A and BKA
ANT of A franciscana shows low similarity to ANTs from other species
The results obtained above prompted us to clone and sequence ANT of A franciscana In the literature, an incomplete 834-bp sequence has been reported (EF660895.1) Gene-specific primers for RACE PCR were designed on the basis of highly conserved regions
Fig 7 Multiple sequence alignment of primary amino acid sequences (in single-letter code) of ANT from Artemia cysts and other organisms (lower panel) and superimposed three-dimensional reconstruction of the known bovine ANT and the predicted conformation of the Artemia ANT (upper panel) In the lower panel, every 10 amino acids are marked by a dot above the sequence box; a dot within the sequence box indicates a deletion Conserved regions are highlighted in red In the upper panel, the three-dimensional reconstructions of bovine ANT (iso-form 1) and Artemia ANT are shown in red and blue, respectively Protein structures differ in the designated areas a, b, and c Yellow repre-sents the part of the bovine ANT that is different from the Artemia ANT; the latter is depicted in magenta Regions a, b and c are marked
on the aligned sequence in the lower panel.
Trang 9of the known ANT nucleotide sequences from other
species and the partial A franciscana ANT sequence
RACE PCR products were sequenced, and the final
assembled 1213-bp nucleotide sequence was submitted
to GenBank (accession number: HQ228154)
Align-ment revealed 99% similarity to the partial A
francis-cana ANT sequence (EF660895.1) and significant
similarity (69–76%) to the sequences of human,
bovine, rat, mouse, Xenopus and Drosophila isoforms
(see below, and Fig 7) The deduced amino acid
sequence of the ORF comprises 301 amino acids and
includes the signature of nucleotide carriers
(RRRMMM) as well as 77–79% similarity to other
species [47,48] However, the region between amino
acids 198 and 225 showed a low degree of similarity
with the other ANT sequences, and harbored amino
acid deletions in positions 211, 212, and 219 (see
below)
Comparison of the primary sequence of Artemia
ANT with that of other species
Multiple alignment of the Artemia ANT protein
sequence with that of other species (Xenopus, Drosophila,
mouse isoforms 1, 2, and 4, rat isoforms 1 and 2,
bovine isoforms 1, 2, 3, and 4, and human isoforms 1,
2, 3, and 4) is shown in Fig 7 (lower panel) It is
evi-dent that region 198–225 of Artemia ANT shows low
similarity to that from other species, and there are,
overall, four amino acid deletions, at positions 46, 211,
212, and 219 The deletions that correspond to
posi-tions 46 and 219 are of highly conserved amino acids
(lysine and glutamine, respectively) However, as seen
below, only the deletions at positions 211 and 212
affect the predicted three-dimensional structure of
Artemia ANT, as compared with the known structure
of bovine ANT
Comparison of the predicted three-dimensional
structure of Artemia ANT with that of bovine
ANT
The structure of bovine ANT (isoform 1) is known
(structure: pdb1okc) [47], and we were therefore able
to compare it with the predicted structure of Artemia
ANT, on the basis of its amino acid sequence The
two proteins are superimposed in Fig 7 (upper panel)
Bovine ANT is shown in red, and Artemia ANT in
blue It becomes immediately apparent that the two
proteins are very similar, except for the three
desig-nated areas (a, b, and c) The part of bovine ANT that
is different from Artemia ANT is colored yellow, and
the corresponding part of Artemia ANT is colored
magenta In region a, this corresponds to His209– Gln218 in bovine ANT, which corresponds to Phe212– Ala218 in Artemia ANT In region b, this corresponds
to Leu41–Ser46 in bovine ANT, which corresponds
to Val45–Ala49 in Artemia ANT In region c, this corresponds to Asp3–Leu6 in bovine ANT, which corresponds to Leu10 and Ser11 in Artemia ANT
Discussion
The identity of the mitochondrial PTP remains unknown after 30 years However, its involvement in a variety of currently untreatable diseases has been repeatedly demonstrated [49–51] Therefore, the need
to discover the protein(s) of which it is composed is pressing Studies focusing on functional evidence for the pore and its modulation are numerous [29,52–56]; without pinpointing the identity of the PTP, they have provided considerable support for the role of two mitochondrial proteins, CypD and ANT However, experiments with genetically modified mice that lack either of these proteins have still demonstrated the PTP [24–28] It was therefore inferred that CypD and ANT do not form the pore, but rather modulate it The latter notion is firmly supported by a wealth of data showing the impact of all ANT ligands (without
a single exception) on the probability of pore opening [4,31–37,57] Therefore, seeking interactions of CypD and⁄ or ANT with other proteins may provide new candidates regarding the identity of the pore Indeed,
it was shown recently that the phosphate carrier – by means of interaction with the ANT – may be a critical component of the PTP [58], and also that ablation of CypD or treatment with cyclosporin A does not directly cause PTP inhibition, but rather unmasks an inhibitory site for Pi [29] Most recently, it has also been shown that CypD not only interacts with F0F1 -ATP synthase, but it also modulates its activity [59] Hereby, we present additional data linking the lack
of a Ca2+-induced PTP to the ANT and the Ca2+-Pi precipitation mechanism Specifically: (a) adenine nucleo-tides decreased Ca2+ uptake rate and capacity – the effect of cATR was conservatively attributed to the inhibition of mitochondrial ADP or ATP uptake (depending on the prevalent membrane potential), thereby eliminating any effect of nucleotides released from broken mitochondria in the suspension; and (b)
Ca2+–Pi precipitates appeared as needles in the absence of exogenously added adenine nucleotides or
Mg2+, a phenomenon that has also been observed in mitochondria isolated from rabbit heart [60], and dots when adenine nucleotides or Mg2+ were present This
is in stark contrast to the ring-like structures observed
Trang 10in Ca2+-loaded mammalian mitochondria [39,40] At
present, there is no explanation for the formation or
usefulness of the formation of such precipitates, and
neither has a connection – if any – to the type of ANT
expressed in A franciscana been established
In the present study, the most important finding is
that A franciscana ANT has a stretch of amino acids
in the 198–225 region that is significantly different from
that in mammalian homologs, including the deletion of
three amino acids at positions 211, 212, and 219
Fur-thermore, BKA did not alter the activity of the ANT
synthesized in this crustacean Currently, experiments
are under way in which the Artemia ANT coding
mRNA sequence will be introduced into ANT-less cells
to determine whether the particular effects of adenine
nucleotides or the lack of effect of BKA can be
repro-duced Nonetheless, the present findings, together with
the previous report that mitochondria isolated from the
embryos of A franciscana lack a Ca2+-induced PTP
[18], strongly reaffirm the implication of ANT in
modu-lation of the PTP However, even though we propose
that the altered amino acid sequence of Artemia ANT
that has been deduced here from the coding mRNA
may be associated with the insensitivity to BKA and
the particular effects of adenine nucleotides on
maxi-mum Ca2+ uptake capacity, it is still possible that an
as yet unfound Artemia ANT isoform is responsible for
some of these findings A diminished effect of BKA has
been demonstrated in yeast mutants [61,62], but the
site(s) of the mutation(s) have never been identified,
although in another study mutations in transmembrane
segments I, II, III and VI were reported to confer
par-tial resistance to BKA [63] So far, the exact binding
site of BKA on ANT has remained unknown [64,65],
but from the present study it seems possible that the
part of ANT in mammalian mitochondria that exhibits
low similarity to that found in A franciscana,
specifi-cally the C2 loop-H5 transmembrane domain region,
interacts with genuine components of the pore and may
also harbor the binding site for BKA
Experimental procedures
Isolation of mitochondria
Mitochondria from embryos of A franciscana were
pre-pared as described elsewhere, with minor modifications [2]
Dehydrated, encysted gastrulae of A franciscana were
obtained from Salt Lake, Utah, through Global Aquafeeds
(Salt Lake City, UT, USA) or Artemia International LLC
(Fairview, TX, USA) and stored at 4C until use Embryos
(15 g) were hydrated in 0.25 m NaCl at room temperature
for at least 24 h After this developmental incubation, the
embryos were dechorionated in modified antiformin solu-tion (1% hypochlorite from bleach, 60 mm NaCO3, 0.4 m NaOH) for 30 min, and this was followed by a rinse in 1% sodum thiosulfate (5 min) and multiple washings in ice-cold 0.25 m NaCl as previously described [66] After the embryos had been filtered through filter paper, 10 g was homogenized in ice-cold isolation buffer consisting of 0.5 m sucrose, 150 mm KCl, 1 mm EGTA, 0.5% (w⁄ v) fatty acid-free BSA, and 20 mm K+-Hepes (pH 7.5) with a glass–Tef-lon homogenizer at 850 r.p.m for 10 passages The homog-enate was centrifuged for 10 min at 300 g and 4C, the upper fatty layer of the supernatant was aspirated, and the remaining supernatant was centrifuged at 11 300 g for
10 min The resulting pellet was gently resuspended in the same buffer, but without resuspending the green core This green core was discarded, and the resuspended pellet was centrifuged again at 11 300 g for 10 min The final pellet was resuspended in 0.4 mL of ice-cold isolation buffer con-sisting of 0.5 m sucrose, 150 mm KCl, 0.025 mm EGTA, 0.5% (w⁄ v) fatty acid-free BSA, and 20 mm K+-Hepes (pH 7.5), and contained 80 mg proteinÆmL)1 (wet weight) Mitochondria from the livers of Xenopus were iso-lated in a similar manner as for rat liver mitochondria, as described elsewhere [41] Male Sprague-Dawley rats weigh-ing 300–350 g were used All animal procedures were per-formed according to the local animal care and use committee (Egyetemi Allatkiserleti Bizottsag) guidelines The X laevis liver is a melanin-containing organ, owing to the presence of melanomacrophage centers [67]; the pres-ence of melanin in the mitochondrial pellet precluded the reliable calibration of the Calcium Green 5N
hexapotassi-um salt (CaGr-5N) fluorescence signals (see below)
DWmdetermination
DWm was estimated by fluorescence quenching of the cat-ionic dye safranine O owing to its accumulation inside ener-gized mitochondria [68] Mitochondria (5 mg for Artemia mitochondria) were added to 2 mL of the incubation med-ium containing 500 mm sucrose, 150 mm KCl, 20 mm Hepes (acid), 10 mm potassium phosphate, 5 mm potassium glutamate, 5 mm potassium malate, 5 mm potassium succi-nate, 1 mm MgCl2 (where indicated), 5 mgÆmL)1 BSA (fatty-acid free), and 5 lm safranine O (pH 7.5) Fluores-cence was recorded in a Hitachi F-4500 spectrofluorimeter (Hitachi High Technologies, Maidenhead, UK) at a 2-Hz acquisition rate, with 495- and 585-nm excitation and emis-sion wavelengths, respectively Experiments were performed
at 27C To convert safranine O fluorescence into milli-volts, a voltage–fluorescence calibration curve was con-structed To this end, safranine O fluorescence was recorded in the presence of 5 nm valinomycin and stepwise increasing [K+] (in the 0.2–120 mm range), which allowed calculation of DWm from the Nernst equation, assuming a matrix [K+] of 120 mm [68] Pilot experiments with various