Calpeptin

Calpeptin, not calpain, directly inhibits an ion channel of the inner mitochondrial membrane

Maria Derksen 1 & Christian Vorwerk 2 & Detlef Siemen 1

Received: 29 April 2015 /Accepted: 8 June 2015 # Springer-Verlag Wien 2015

Abstract The permeability transition pore (PTP) of inner mi- tochondrial membranes is a large conductance pathway for ions up to 1500 Da which opening is responsible for ion equilibration and loss of membrane potential in ap- optosis and thus in several neurodegenerative diseases. The PTP can be regulated by the Ca2+-activated mito- chondrial K channel (BK). Calpains are Ca2+-activated cystein proteases; calpeptin is an inhibitor of calpains. We wondered whether calpain or calpeptin can modulate activity of PTP or BK. Patch clamp experiments were performed on mitoplasts of rat liver (PTP) and of an astrocytoma cell line (BK). Channel-independent open probability (Po) was determined (PTP) and, taking into account the number of open levels, NPo by single chan- nel analysis (BK). We find that PTP in the presence of Ca2+ (200 μM) is uninfluenced by calpain (13 nM) and shows insignificant decrease by the calpain inhibitor calpeptin (1 μM). The NPo of the BK is insensitive to calpain (54 nM), too. However, it is significantly and reversibly inhibited by the calpain inhibitor calpeptin (IC50 =42 μM). The results agree with calpeptin- induced activation of the PTP via inhibition of the BK. Screening experiments with respirometry show
calpeptin effects, fitting to inhibition of the BK by calpeptin, and strong inhibition of state 3 respiration.

Keywords Calpain . Calpeptin . Calpain inhibitor . Mitochondria . Patch clamp . BK channel . Permeability transition pore

Introduction

m-calpain is the type protein of a calcium-dependent, non- lysosomal cysteine protease, which is ubiquitously found in all tissues (Reverter et al. 2001). μ-calpain (also calpain 1) is an isoform predominant in the central nervous system. Calpeptin (benzoyloxycarbonyldipeptidyl aldehyde) is a cell-permeable calpain inhibitor. Calpain seems to act, e.g., on long-term potentiation in neurons and on cell fusion in myoblasts. Either of these processes and also the onset of apoptosis cause a transient and localized influx of Ca2+ into the cell which then activates calpain to cleave its target pro- teins. It has been suggested that μ-calpain could cleave apoptosis-inducing factor (AIF) thereby inducing Ca2+ release from mitochondria (Polster et al. 2005). It is tempting to as- sume that this Ca2+ is released by a phenomenon called per- meability transition. It includes cytochrome c release from the outer membrane that is generally accepted as a sign of apo-

Handling Editor: Reimer Stick

* Detlef Siemen [email protected]
ptosis (Polster et al. 2005).
Another sign of the intrinsic signaling chain in apoptosis is opening of the permeability transition pore (PTP) in the inner mitochondrial membrane (IM). It seems to be an evolutionary old pathway as it is found in plants, too (Contran et al. 2007).

1

2
Department of Neurology, Otto-von-Guericke-Universität, Leipziger Str. 44, D-39120 Magdeburg, Germany
Ophthalmology, SRH Zentralklinikum Suhl, D-98527 Suhl, Germany
The PTP can be opened by elevated matrix Ca2+ or depolari- zation of the IM while it is inhibited by the antibiotic cyclo- sporine A (Loupatatzis et al. 2002; Sayeed et al. 2006). In single channel experiments by means of the patch clamp

technique, it shows up as a sequence of current events with extremely large amplitude and a large number of substates of different amplitude and different kinetics (Loupatatzis et al. 2002). Blockade of the PTP can be achieved by several phar- maceuticals all shown to act as neuroprotectants (Andrabi et al. 2004; Sayeed et al. 2006; Parvez et al. 2010). Though it had been assumed for years that the purine nucleotide antiporter would be the molecular basis of the PTP, recent papers claim that dimers of the F0/F1-ATPase are constituting the PTP with their conserved b-, e-, and g-F0 subunits (Bernardi et al. 2015). They are cooperating with other pro- teins like cyclophiline D containing the cyclosporine A bind- ing site (Giorgio et al. 2013; for review: Siemen and Ziemer 2013). However, the new model does not explain all observa- tions so that some points remain controversial. In our hands, the PTP was most frequently observed in mitoplasts obtained from liver mitochondria.
Next to cyclosporin A and hyperpolarization, opening of a mitochondrial Ca2+-activated K channel of the BK type (BK channel) can inhibit activity of the PTP (Cheng et al. 2008). This functional connection was shown by several groups in several preparations like guinea pig ven- tricular cells, isolated perfused rat heart, intact rat brain mitochondria (Xu et al. 2002; Cao et al. 2005; Gao et al. 2005; for review and their Fig. 3: Cheng et al. 2010). The BK channel exhibits a single channel conduc- tance (γ) of 290±10 pS and much more regular appear- ance of the square-shaped current events as compared with

Materials and methods

Solutions (concentrations in mM if not otherwise stated): Patch clamp (pH=7.2 with KOH)
Preparation solution: 250 sucrose, 5 HEPES Storage solution: 150 KCl, 10 HEPES Hypotonic solution: 5 HEPES
Hypertonic solution: 750 KCl, 30 HEPES Isotonic solution: 150 KCl, 10 HEPES
Bath solution: 2 ml hypotonic + 0.5 ml hypertonic solution If not otherwise stated, last four solutions contained
200 μM CaCl2
Respirometry (pH=7.4 with KOH)
MSE solution A: 225 mannitol, 75 sucrose, 20 MOPS, 1 EGTA
MSE solution B: 225 mannitol, 75 sucrose, 20 MOPS, 0.1 EGTA
KCl medium: 60 KCl, 120 mannitol, 40 MOPS, 5 KH2PO4, 5 MgCl2·6 H2O
Calpain was diluted in ethanol and then further diluted to experimental concentrations. Care was taken that the final ethanol concentration never exceeded 0.1 % (v/v). Even in
1% ethanol, single channel current–voltage curves measured for control did not show any deviation from the curves without ethanol, which is in agreement with a low ethanol sensitivity of the BK channel described earlier (Davies et al. 2003).

2+
the PTP. It is stimulated by Ca
and by depolarization
Chemicals

and can be blocked by iberiotoxin from the Indian red

scorpion (Buthus tamulus), charybdotoxin (from the scor- pion Leiurus quinquestriatus), and paxilline (Siemen et al. 1999; Gu et al. 2014). The cAMP-dependent kinase depo- larizes the mitochondrial membrane potential ΔΨ and re- duces the mitochondrial Ca2+ thereby modulating the BK channel, too (Sato et al. 2005). BK channels are most frequently found in mitoplasts from glioblastoma or astro- cytoma cell lines. We find them in mitochondria from other cells too, though at much lower density.
Openers of the BK channel were shown either to act cell protectively on cardiomyocytes or, in contrary, to induce cell death in the LN229 human glioma cell line (Xu et al. 2002; Debska-Vielhaber et al. 2009). Some openers increased the cytosolic Ca2+ concentration, possibly due to inhibition of the Ca2+-ATPase of the endoplasmic reticulum (ER) as a side effect. Because this Ca2+-increase activated calpain, it seems important to understand what calpain does to other proteins involved in cytoprotection or apoptosis. We therefore tried to apply both, calpain and one of its antagonists, calpeptin, to the PTP directly or to the BK channel and to look for modulatory effects on the single channel currents. It turned out that under our conditions, calpain has no significant effect on PTP or on BK while calpeptin effectively inhibits BK but not PTP.
Calpain 1 (from human erythrocytes), calpeptin, carboxyatractyloside, EGTA, sucrose, KCl, MgCl2·6 H2O, CaCl2·2 H2O, and DMSO were from Merck (Darmstadt, Ger- many). Dulbecco’s modified Eagle’s medium (DMEM with L-glutamin and sodium pyruvate), fetale calf serum (FCS), and Hank’s balanced salt solution (BSS) medium was from PAA Laboratories (Pasching, Austria). Penicillin and strepto- mycin were from Roche (Indianapolis, USA). All other chemicals were from Sigma (Taufkirchen, Germany).

Preparation of mitochondria from mouse and rat liver

C57BL/6SV129 wild-type mice and Wistar rats were kept and sacrificed according to the guidelines of the Animal Health and Care Committee of the state of Sachsen-Anhalt, Germany. Liver tissue was transferred into ice-cold solution B, washed, and minced with scissors. It was then homogenized in 50 ml of solution A and centrifuged at 800×g. The supernatant was filtered through sterile gauze and centrifuged 5100×g. The pellet was dispersed in solution B, homogenized again, and centrifuged at 12,300×g. This step was repeated, and the pellet finally dispersed in 2 ml solution B and stored on ice.

High-resolution respirometry

The protein content of the mitochondrial suspension was de- termined by photometry using the biuret or the bicichoninic acid method (Wiechelman et al. 1988). Protein (1 mg/ml) in
2ml KCl medium was used for the experiments. Measure- ment of respiration was performed with an Oroboros oxygraph (Bioenergetics and Biomedical Instruments, Inns- bruck, Austria). Complex I of the respiratory chain was acti- vated routinely by 5 mM glutamate and 5 mM malate follow- ed by 100 μM ADP and 1 mM NADH. Before NADH, 100 μM Ca2+ was added for opening of the PTP and thus allowing access of the NADH to the respiratory chain. Calpeptin was added either 10 min in advance or after glutamate/malate application. O2 content of the chamber as well as its first derivation, the change in O2 concentration, were recorded.

Cell culture and preparation of astrocyte mitochondria The human astrocytoma cell line U-87MG was cultured in
DMEM supplemented by 10 % FCS, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10 μg/ml tylosin. Cells were incubated in a humidified atmosphere with 5 % CO2 at 37 °C. They were fed and reseeded every third day. Harvesting was done in Hank’s BSS solution by means of a scraper and subsequent centrifugation at 800×g (10 min), resuspension in preparation solution, and homogenizing in a tight glass potter. One fast centrifugation at 9200×g and one slow at 770×g, both for 10 min, isolated the purified mito- chondria. Finally, sucrose was removed by two additional fast

capillary system. The patch clamp amplifier was an EPC 7 (HEKA, Lambrecht, Germany). Patches were inside-out or mitoplast-attached. All potentials are given as at the inner side of the membrane. Filtering of the analog data was performed at a corner frequency of 0.5 KHz. Data were digitized at a sampling rate of 250 μs by a Digidata 1322A (Molecular Devices, Sunnyvale, USA). Stimulation and recording were done by the pClamp software (Molecular Devices, Sunnyvale, USA) running on a Windows PC. In the figures, outward currents are generally directing upward. All experiments were performed at room temperature (24±1 °C).

Analysis

Analysis of the experimental data was done by means of the pClamp software, too. The probability (Po) that a channel occupies the open state can be determined in two ways. We used the all-point analysis for the PTP cutting the current traces into continuous segments of 1 min duration and analyz- ing its amplitude point to point. Showing these amplitudes as histogram, the areas (An) under the different Gaussian curves represent the time the channel spends at this level. A0 repre- sents the time in the closed state. Due to the characteristic large number of substates each with distinct amplitude, we did not use the natural number of the different current levels to account for a different number of open channels but used instead the different current amplitudes (Bn), i.e., the mean value of a Gaussian for relating Po to the amount of charges transferred. Bmax is the largest current amplitude observed in that experiment (Eq. 1, also used in (Loupatatzis et al. 2002)).

centrifugation steps at 9200×g for 10 min in storage solution and kept on ice.
Po ¼ ðA1 *B1

Þ þ ðA2 *B2 ð

ð
n *Bn
n Þ

Þ

ð1Þ

Patch clamp experiments

A small sample of either type of mitochondria was transferred into a chamber with 2 ml of hypotonic solution in which they
As the BK channel shows only a small number of clearly distinguishable levels, NPo can be easily determined by the single channel search mode of the pClamp software (Eq. 2):

rapidly swelled to mitoplasts. After about 45 s, 0.5 ml of hypertonic solution was added in order to restore isotonicity. In phase contrast microscopy, mitoplasts could be easily rec-
NPo ¼
T o
T o þ T c
ð2Þ

ognized as absolutely round fragile spheres with one (mostly in liver) or more (in astrocytoma) dark spots. Staining with a mouse anti-inner mitochondrial membrane antibody, staining and mitochondrial localization with an anti-BK channel anti- body, and occasional coactivity of the mtBK channel with the mitochondrial PTP further confirmed the identity of these ves- icles as consisting out of inner membrane material (Siemen
To represents the total duration in the open state of the channel, Tc the total time in the closed state.
The concentration–response curve was fitted with a modified Hill equation (Eq. 3). c is the concentration of calpeptin used, and IC50 is the concentration of calpeptin that blocks the BK channel by 50 %. n is the Hill coefficient.

et al. 1999; Xu et al. 2002; Bednarczyk et al. 2013).
The measuring pipette was pulled from borosilicate glass (Harvard Apparatus Ltd. Edenbridge, UK) with a pipette re-

NPo ¼
cn
cn þ IC50n

ð3Þ

sistance of about 17 MΩ and filled with isotonic solution. Test solutions were applied through a peristaltic pump-driven
For statistics, we always calculated the standard deviation (SD). Significance was determined by a two-sided Student’s t

test. As levels of significance, we used p ≤ 0.05 (*), p ≤ 0.01 (**), and p ≤ 0001 (***).

Results

Calpeptin and calpain action on the PTP

Single channel current was recorded from mitoplasts prepared from mitochondria of rat hepatocytes. Channel events showed large amplitude with many substates several of them with distinct kinetics (Fig. 1a). Thirty-four pA at 20 mV corre- sponds to a single channel conductance (γ) of 1.7 nS, which is about the maximum conductance of a PTP. Channel activity was blocked when the measuring pipette was transferred to a capillary of the flow system that contained isotonic solution either without Ca2+ or with 1 μM cyclosporine A. All togeth- er, these features are known characteristics of the PTP from the inner mitochondrial membrane (Loupatatzis et al. 2002).
In order to test for a PTP-blocking effect, 1 μM of the calpain antagonist calpeptin was added to the isotonic solution

within the flow system. At a holding potential (Eh) of +20 mV, the current amplitude of the PTP was reduced from 34 to 5.5 pA (Fig. 1b). This effect was almost irreversible (Fig. 1c). For a quantification of the effect on the mean, open probability (Po) was determined from 3 to 11 1-min segments of the current traces of n =5 experiments by means of Eq. 1 (Fig. 2). It turned out that in five independent experiments, 1 μM calpeptin showed a tendency to irreversibly reduce Po. The effect was insignificant, however. Calpain (13 nM) did not exhibit any effect on Po (Fig. 2).

Calpeptin but not calpain causes blockade of the mitochondrial BK channel

Recording single channels from astrocytoma mitoplasts, the mitochondrial BK channel was easier to recognize than the PTP (Fig. 3). It exhibited a mean γ of 290±10 pS (n =120) and a much smaller number of substates with less different kinetics than the PTP (Bednarczyk et al. 2013; Siemen et al. 1999). It was activated by Ca2+ ions (Fig. 3) and could be blocked by the selective BK channel inhibitor iberiotoxin (not shown).
Calpain (54 nM) had no significant effect on γ. The open probability NPo (Eq. 2) showed an insignificant tendency to- ward lower values (Fig. 4). When the single channel events were studied in more detail, it turned out that under our exper- imental conditions (filter frequency and sample rate), the closed times and open times could be described by two expo- nentials: a fast and a slow one (Fig. 5). The closed times of the BK channel changed in two ways: (i) The fast component increased slightly but significantly and (ii) the slow compo- nent, which is hardly detectable under control conditions be- fore, increased enormously. For the open times, (i) a similar but insignificant increase was observed for the fast compo- nent, while (ii) the slow component was highly significantly increased. Obviously, slowing of both together, the slow

Fig. 1 Single channel current traces of the PTP from rat liver mitoplasts were recorded before, during, and after application of 1 μM calpeptin to the isotonic solution. Activity of the PTP with a conductance of up to 1.7 nS was considerably reduced by calpeptin, however, not significantly in the mean of five 1-min segments of five experiments (together 25-min recording time compared with 25-min control records before calpeptin). The calpeptin effect was irreversible after washing with isotonic solution without calpeptin. Dashed line gives zero current. Note large number of substates particularly in the record before calpeptin

Fig. 2 Neither 1 μM calpeptin (left) nor 13 nM calpain (right) showed significant effects on the mean channel open probability Po of the PTP at rat liver mitoplasts (left; n =5). Po was determined by all point analysis of the PTP by means of Eq. 1

Fig. 3 Original current traces of the BK channel from astrocytoma cells
2+
(a) demonstrate Ca2+ dependence (b), independence of the Ca dependence from the presence of 54 nM calpain (c), increasing slow components of particularly of the closed times in 54 nM calpain (d), irreversibility of this calpain effect (e), blockade of the BK channel by 100 μM calpeptin (f), and partial reversibility of the calpeptin effect (g). Records from a single experiment

closed time and the slow open time resulted in an almost unchanged NPo.

Fig. 4 One hundred micromolar calpeptin (n =3) but not 54 nM calpain (n =3) cause a significant and reversible reduction of the open probability (NPo) as determined by single channel analysis from recordings of the BK-Kanal

Fig. 5 Reduced activity of the BK channel by calpain can be explained by a significant change of the according time constants (τ). Fifty-four nanomolar calpain raised highly significant and irreversibly the slow open time constant. Additionally, it irreversibly induced the slow time constant of the closed state

Calpeptin (100 μM), however, reduced γ from 270 to 133 pS (Fig. 6) and NPo of the BK channel from 0.80±0.19 to 0.21±0.05 (n =3), even in the presence of 200 μM Ca2+ (Figs. 3 and 4). Both effects were reversible. Blockade of the BK channel by 100 μM calpeptin could be partially washed out in the presence of 50 μM calpain, too (not shown). The inhibition by calpeptin was concentration dependent with an IC50 of 42 μM (Fig. 7). A detailed analysis of five 1-min segments under each condition demonstrated the contribution of each of the kinetic parameters to the dramatic change of NPo (Fig. 8). Again, the slow component of the closed time was so small that it could be hardly detected, even in 1 μM calpeptin. Only when the calpeptin concentration was raised to 30 μM the slow component of the closed times showed up considerably while the fast component of the closed times that was unaffected by 1 μM calpeptin, too, became shorter. Con- trary, the fast and slow components of the open times

Fig. 6 Calpeptin reduces the single channel current irreversibly. Current– voltage relation in bath solution (diamond), 10 μM calpeptin in isotonic solution (square), and isotonic solution without calpeptin (triangle). i: single channel current in pA, Eh: holding potential in mV. Application of calpeptin and isotonic solution as a control by the flow system. The straight line gives a single channel conductance (γ) of 270 pS, the dashed line a γ of 133 pS

Fig. 7 Calpeptin inhibits the BK channel concentration dependently. Curve calculated by a modified Michaelis–Menten equation accounting for the decline by inhibition (Eq. 3, IC50 =42 μM)

responded already to 1 μM calpeptin by an irreversible de- crease of about 50 %. At 30 μM calpeptin, this decrease was extended and continued during the control afterwards in a way that the fast component gained on the cost of the slow so that the slow disappeared completely. In summary, while the effect of calpain is more a slowing of the closed times, calpeptin causes an enormous reduction open times.

Respirometry demonstrates involvement of the BK channel

In order to find out if the strong calpeptin effect and the ab- sence of a calpain effect found by single channel experiments at mitoplasts could be detected on intact mitochondria as well, we performed high-resolution respirometry on mouse liver

Fig. 8 The change of NPo of the BK channel can be explained by significant changes of the relevant time constants. In single channel experiments, 1 μM calpeptin reduced extremely significant and irreversibly the slow time constant of the open state, the effect increasing at higher concentration (30 μM). The slow time constant of the closed state is induced only at a calpeptin concentration of 30 μM, while the fast time constant is irreversibly reduced at this concentration

mitochondria. In two screening experiments, 50 μM calpeptin induced inhibition of the respiratory chain (RC) demonstrated by a reduced state 3 respiration after application of 100 μM ADP as compared with respiration before ADP. As the medium contained high Ca2+ concentrations (100 μM), we tested additionally media without Ca2+. They demonstrated that 50 μM calpeptin inhibited the RC by 11.4 pmol/(s ml) (n =2), however, less pronounced than in the presence of 100 μM Ca2+ 54.7 pmol/(s ml). Even 100 μM Ca2+ alone was able to inhibit the RC by 44.0 pmol/(s ml). Ca2+ acts concentration dependently as 10 μM Ca2+ inhibited only by 12.4 pmol/(s ml). The IM is normally impermeable to NADH, but after opening of the PTP by Ca2+, NADH gets access to complex I thus fueling the respiratory chain (Fontaine et al. 1998). Testing for PTP opening with 50 μM calpeptin, we found that respiration in the absence of Ca2+ did not show a clear increase after application of 1 mM NADH (n =2). The latter are screening experiments in numbers too small to pro- vide safe statistics. We mention them, but we do not show figures, therefore.
In contrast to calpeptin, 1.25 or 0.2 μM calpain did not affect state 3 respiration significantly (n =3) in the presence of 10 μM Ca2+ (not shown). Control experiments without Ca2+ showed the same (n =2) leaving us with the result that calpain was completely ineffective.

Discussion

In glioma cells, a BK channel opener not only stimulates the mitochondrial BK channel but also raises the cytosolic as well as the mitochondrial Ca2+ concentration (Debska-Vielhaber et al. 2009). This Ca2+ originates from the ER by an unknown mechanism. The increased cytosolic Ca2+ enters the matrix via the so-called Ca2+ uniport, a voltage-dependent Ca2+ channel (for review: Szabo and Zoratti 2014), then opens the PTP and activates calpains, both together finally leading to death of the glioma cell (Debska-Vielhaber et al. 2009). The precise mech- anism by which calpain is acting at the mitochondria was unknown. It was not even clear whether it was a direct or an indirect effect.
Similarly, a rise of intracellular Ca2+ due to glutamate excitotoxicity may raise the matrix Ca2+ as a conse- quence of acute rodent brain injury (Polster et al. 2005). The authors suggested that cleavage of the AIF by endogeneous μ-calpain could mediate its release from the mitochondria via the activated PTP. Inhibition of the AIF release by CsA supported this idea. Such a proapoptotic mechanism would differ from Bax activa- tion of the PTP via inhibition of the BK channel (Cheng et al. 2011). It was thus tempting to test for a direct effect of μ-calpain on the open probability of the

PTP by patch clamping inner mitochondrial membrane in the single channel mode.

Calpain and calpeptin do not directly alter activity of the PTP

As a first guess, we expected that, due to its well-known proapoptotic effect, calpain would activate the PTP (Altznauer et al. 2004). As a fast and reliable measure of PTP activity, we chose all-point analysis of the open probability Po. μ-calpain requires 5 to 50 μM Ca2+ for activation (Kd) (Reverter et al. 2001; Dutt et al. 2002), so that the 200 μM Ca2+ used here which are useful for getting reliable seals between mitoplast membrane and patch pipette is well in the activating range. It turned out that calpain had no effect on Po although openings of the very small Ca2+-conducting substate of the PTP may have escaped detection (Ichas and Mazat 1998). At least, the total Po of about 0.45 (45 % of the maximum conductance of a single PTP) did not significantly increase anymore. Thus, our hypothesis of a calpain-induced opening of the PTP that had been suggested as the reason for cell death in glioma cells seems wrong (Debska-Vielhaber et al. 2009). The calpain an- tagonist calpeptin tended to reduce the Po values, though insignificantly.
We cannot exclude that a very slow effect of calpain or calpeptin may not have been recorded because for unknown reasons the majority of the patches of liver mitoplasts did not last longer than 25 min. This left about 7 min for each, the control before, the test phase, and the control after which was sufficient to observe direct effects on PTP by Ca2+, dopamine agonists, and others (Cheng et al. 2010; Sayeed et al. 2006; Parvez et al. 2010). Taken together, a direct effect of calpain on the PTP could not be observed.

Calpeptin but not calpain inhibits the BK channel significantly

From earlier work, it was clear that the PTP may be modulated by the activity of K channels and thus of the BK channel, too (Foster et al. 2012; Cheng et al. 2010; Gulbins et al. 2010). Therefore, we tested whether calpain would be able to reduce BK channel activity, thereby having the potential to open the PTP indirectly. We did not expect a very strong effect because only a few mV depolarization of ΔΨ was estimated upon opening of the K(ATP) channel in mitochondria of cardiomyocytes (Garlid 2000). An alternative, and more likely in this case, would be a scenario in which calpain would close an open BK channel thereby briefly hyperpolarizing ΔΨ by a few mV. Subsequently, the PTP would be activated by the closed BK channel (details unknown), and ΔΨ would be strongly depolarized. Such a mechanism was proposed for the Kv1.3 in lymphocyte mitochondria (Gulbins et al. 2010). Though we observed in our experiments a calpain-induced

tendency toward a reduction of the Po of the BK channel and an increased Po of the PTP, none of the observations was significant.
However, when examining the open and closed times of the BK channel in more detail, it turned out that 54 nM calpain increased both components of the closed times, the slow one dramatically because it was practically not existing before calpain. However, the open times increased too, thus probably explaining why the calpain effect on the open probability of the BK channel was not significant. As the number of ana- lyzed events was never below 400 (most about 2000) for a 1- min segment and five segments were analyzed under each condition, statistics should be sound. It cannot be excluded that the picture would differ slightly if by a faster sampling rate shorter events would have been included.
Single channel analysis of the calpeptin effect shows that inhibition at low concentrations was due to increased open time constants while at higher concentration, the closed times were concerned, as they were more and more redistributed from the fast component to the slow. Altogether, these calpeptin effects explain the significant and reversible inhibi- tion of the NPo of the BK channel, an effect that was dose- dependent and at lower concentrations partly reversibly.

Calpeptin effect measured by respirometry

The strong effect of the calpain inhibitor calpeptin on the NPo of the BK channel should be visible in respirometry measure- ments as well. Some screening experiments were performed in order to test whether they are in agreement with the patch clamp data. By respirometry, we saw that calpeptin reduces state 3 respiration after application of ADP. The missing ADP peak could either be due to a calpeptin effect on the purine nucleotide exchanger or due to a direct effect of blocking the BK channel on the respiratory chain like the one recently shown in reverse direction, i.e., RC affecting the BK channel (Bednarczyk et al. 2013). Both, the BK channel and the PTP are Ca2+- and ΔΨ-dependent in their activity (depolarization opens). Without Ca2+ in the medium, it is unlikely that either of them will open (for review: Cheng et al. 2010). Therefore, we believe that the effects we saw with respirometry are caused by direct action of calpeptin on the RC.
Opening of the PTP by Ca2+ in a Ca2+-containing medium might lead not only to a decline of ΔΨ, which could cause a positive feedback, as it would result in a depolarization. Open- ing could additionally induce swelling leading to membrane rupture, which would facilitate influx of RC substrates (for review: Cheng et al. 2010). As a consequence, respiration should be increased, as observed in our experiments after NADH application (Fontaine et al. 1998; Parvez et al. 2010), or light absorbence should be decreased. Thus, instead of con- clusively proving the patch clamp experiments, respirometry introduces new questions. A thorough answer of them would

require substantial additional work being beyond the scope of this paper.

Conclusion

1.The patch clamp experiments with calpain are neither demonstrating a significant activating effect of calpain on the PTP as suggested by (Altznauer et al. 2004) nor on the BK channel. Inhibition of these channels could not be seen, too.
2.The calpain antagonist calpeptin was not able to activate the PTP. In contrast, the experiments hint to an insignifi- cant inhibiting effect of calpeptin on the PTP.
3.The calpain antagonist calpeptin is a strong inhibitor of the BK channel by itself.
4.Earlier observations that a closed BK channel tends to open the PTP prompted us to expect increased opening of the PTP. That this did not apply might be due to the Ca2+ required for patch clamping.
5.Respirometry on intact mouse liver mitochondria sup- ports the assumption that calpeptin additionally inhibits the RC.

Acknowledgments We are indebted to F. N. Gellerich and Z. Gizatullina for advice and helpful discussions throughout the respirome- try experiments. Mrs. K. Kaiser, C. Höhne, and J. Witzke gave us tech- nical support, and Mrs. D. Koch from Leibniz Institute of Neurobiology, Magdeburg, kindly provided us with rat livers. Financial support by Deutsches Zentrum für Neurodegenerative Erkrankungen (DZNE) is gratefully acknowledged.

Conflicts of interest None of the authors has any conflicts of interest. (Detlef Siemen on behalf of the authors)

References

Altznauer F, Conus S, Cavalli A, Folkers G, Simon HU (2004) Calpain-1 regulates Bax and subsequent Smac-dependent caspase-3 activation in neutrophil apoptosis. J Biol Chem 279:5947–5957
Andrabi SA, Sayeed I, Siemen D, Wolf G, Horn TF (2004) Direct inhi- bition of the mitochondrial permeability transition pore: a possible mechanism responsible for anti-apoptotic effects of melatonin. FASEB J 18:869–871
Bednarczyk P, Wieckowski MR, Broszkiewicz M, Skowronek K, Siemen D, Szewczyk A (2013) Putative structural and functional coupling of the mitochondrial BKCa channel to the respiratory chain. PLoS One 8(6):e68125
Bernardi P, Di Lisa F, Fogolari F, Lippe G (2015) From ATP to PTP and back: a dual function for the mitochondrial ATP synthase. Circ Res 116:1850–1862
Cao CM, Xia Q, Gao Q, Chen M, Wong TM (2005) Calcium-activated potassium channel triggers cardioprotection of ischemic precondi- tioning. J Pharmacol Exp Ther 312:644–650
Cheng Y, Gu XQ, Bednarczyk P, Wiedemann FR, Haddad GG, Siemen D
(2008)Hypoxia increases activity of the BK-channel in the inner

mitochondrial membrane and reduces activity of the permeability transition pore. Cell Physiol Biochem 22:127–136
Cheng Y, Debska-Vielhaber G, Siemen D (2010) Interaction of mitochon- drial potassium channels with the permeability transition pore. FEBS Lett 584:2005–2012
Cheng Y, Gulbins E, Siemen D (2011) Activation of the permeability transition pore by Bax via inhibition of the mitochondrial BK chan- nel. Cell Physiol Biochem 27:191–200
Contran N, Cerana R, Crosti P, Malerba M (2007) Cyclosporin A inhibits programmed cell death and cytochrome c release induced by fusicoccin in sycamore cells. Protoplasma 231:193–199
Davies AG, Pierce-Shimomura JT, Kim H, VanHoven MK, Thiele TR, Bonci A, Bargmann CI, McIntire SL (2003) A central role of the BK potassium channel in behavioral responses to ethanol in C. elegans. Cell 115:655–666
Debska-Vielhaber G, Godlewski MM, Kicinska A, Skalska J, Kulawiak B, Piwonska M, Zablocki K, Kunz WS, Szewczyk A, Motyl T
(2009)Large-conductance K+ channel openers induce death of hu- man glioma cells. J Physiol Pharmacol 60:27–36
Dutt P, Spriggs CN, Davies PL, Jia Z, Elce JS (2002) Origins of the difference in Ca2+ requirement for activation of mu- and m-calpain. Biochem J 367:263–269
Fontaine E, Eriksson O, Ichas F, Bernardi P (1998) Regulation of the permeability transition pore in skeletal muscle mitochondria. Modulation By electron flow through the respiratory chain complex i. J Biol Chem 273:12662–12668
Foster DB, Ho AS, Rucker J, Garlid AO, Chen L, Sidor A, Garlid KD, O’Rourke B (2012) Mitochondrial ROMK channel is a molecular component of mitoK(ATP). Circ Res 111:446–454
Gao Q, Zhang SZ, Cao CM, Bruce IC, Xia Q (2005) The mitochondrial permeability transition pore and the Ca2+-activated K+ channel contribute to the cardioprotection conferred by tumor necrosis fac- tor-alpha. Cytokine 32:199–205
Garlid KD (2000) Opening mitochondrial K(ATP) in the heart—what happens, and what does not happen. Basic Res Cardiol 95:275–279
Giorgio V, von Stockum S, Antoniel M, Fabbro A, Fogolari F, Forte M, Glick GD, Petronilli V, Zoratti M, Szabó I, Lippe G, Bernardi P (2013) Dimers of mitochondrial ATP synthase form the permeability transition pore. Proc Natl Acad Sci U S A 110:5887–5892
Gu XQ, Pamenter ME, Siemen D, Sun X, Haddad GG (2014) Mitochondrial but not plasmalemmal BK channels are hypoxia- sensitive in human glioma. Glia 62:504–513
Gulbins E, Sassi N, Grassmè H, Zoratti M, Szabò I (2010) Role of Kv1.3 mitochondrial potassium channel in apoptotic signalling in lympho- cytes. Biochim Biophys Acta 1797:1251–1259
Ichas F, Mazat JP (1998) From calcium signaling to cell death: two con- formations for the mitochondrial permeability transition pore. Switching from low- to high-conductance state. Biochim Biophys Acta 1366:33–50
Loupatatzis C, Seitz G, Schönfeld P, Lang F, Siemen D (2002) Single- channel currents of the permeability transition pore from the inner mitochondrial membrane of rat liver and of a human hepatoma cell line. Cell Physiol Biochem 12(5–6):269–278
Parvez S, Winkler-Stuck K, Hertel S, Schönfeld P, Siemen D (2010) The dopamine-D2-receptor agonist ropinirole dose-dependently blocks the Ca2+-triggered permeability transition of mitochondria. Biochim Biophys Acta 1797:1245–1250
Polster BM, Basañez G, Etxebarria A, Hardwick JM, Nicholls DG (2005) Calpain I induces cleavage and release of apoptosis-inducing factor from isolated mitochondria. J Biol Chem 280:6447–6454
Reverter D, Strobl S, Fernandez-Catalan C, Sorimachi H, Suzuki K, Bode W (2001) Structural basis for possible calcium-induced activation mechanisms of calpains. Biol Chem 382:753–766
Sato T, Saito T, Saegusa N, Nakaya H (2005) Mitochondrial Ca2+-acti- vated K+ channels in cardiac myocytes: a mechanism of the

cardioprotective effect and modulation by protein kinase A. Circulation 111:198–203
Sayeed I, Parvez S, Winkler-Stuck K, Seitz G, Trieu I, Wallesch CW, Schönfeld P, Siemen D (2006) Patch clamp reveals powerful block- ade of the mitochondrial permeability transition pore by the D2- receptor agonist pramipexole. FASEB J 20:556–558
Siemen D, Ziemer M (2013) What is the nature of the mitochondrial permeability transition pore and what is it not? IUBMB Life 65: 255–262
Siemen D, Loupatatzis C, Borecky J, Gulbins E, Lang F (1999) Ca2+- activated K channel of the BK-type in the inner mitochondrial

membrane of a human glioma cell line. Biochem Biophys Res Commun 257:549–554
Szabo I, Zoratti M (2014) Mitochondrial channels: ion fluxes and more. Physiol Rev 94:519–608
Wiechelman KJ, Braun RD, Fitzpatrick JD (1988) Investigation of the bicinchoninic acid protein assay: identification of the groups respon- sible for color formation. Anal Biochem 175:231–237
Xu W, Liu Y, Wang S, McDonald T, Van Eyk JE, Sidor A, O’Rourke B (2002) Cytoprotective role of Ca2+-activated K+ channels in the cardiac inner mitochondrial membrane. Science 298:1029–1033