**3. Results**

#### *3.1. E*ff*ect of CsA on Extra-Matrix Free [Ca<sup>2</sup>*+*]*

To determine the e ffect of CsA on matrix Ca2+ uptake, we measured [Ca<sup>2</sup>+]e during repetitive additions of 20 μM CaCl2 boluses at 5 min (300 s) intervals to allow characterization of the detailed kinetics of steady-state Ca2+ dynamics (influx and bu ffering). Figure 2 shows the dynamics of [Ca<sup>2</sup>+]e during CaCl2 pulse challenges, with di fferent treatments. Panels A, B, C and panels D, E, F depict the Ca2+ dynamics profile using Protocols A and B, respectively. Each panel consists of five traces representing di fferent treatment groups (DMSO, ADP, OMN, OMN+ADP, and CsA) in the presence of approximately 40 μM EGTA (carried over from the isolation bu ffer). In response to each CaCl2 pulse, an increase in Fura-4F fluorescence intensity was observed, which then returned to a baseline; the

steady-state (ss) level is marked by the flat response as mitochondria take up and sequester the added Ca2+. The opening of mPTP is evident by cessation of mitochondrial Ca2+ uptake and a sharp rise in the extra-matrix dye fluorescent intensity. Ca2+ concentrations were determined from the fluorescence ratios using Equation (1).

**Figure 2.** Effect of CsA and AdN on extra-mitochondrial calcium ([Ca<sup>2</sup>+]e) dynamics. Mitochondrial Ca2+ uptake and buffering for each of the treatment groups: DMSO (control; black trace), CsA (red trace), ADP (brown trace), OMN (blue trace), or OMN+ADP (green trace) are shown using the protocols depicted in Figure 1. Mitochondrial suspension was exposed to 0.5 μM CsA, 250 μM ADP, 10 μM OMN, or OMN+ADP before adding boluses of 20 μM CaCl2 (Protocol A; left column). Mitochondrial suspension was exposed to added boluses of CaCl2 (20 μM) and rescued mitochondria from mPTP opening (Protocol B; right column) with similar treatments as in Protocol A, at a time point at which it would initiate pore opening. Representative traces show change in extra-matrix free Ca2+ ([Ca<sup>2</sup>+]e) over time (**A**), and rescue of mitochondria from mPTP opening (**D**). Insets (**A**,**D**) show Ca2+ uptake kinetics in detail. Steady-state [Ca<sup>2</sup>+]e (ss[Ca<sup>2</sup>+]e), 270 s after initiation of Ca2+ uptake, plotted as function of added Ca2+ (20 μM) every 300 s, in delay of mPTP opening (**B**), and rescue of mitochondria from mPTP opening (**E**). Insets (B,E) indicate the time points at which ss[Ca<sup>2</sup>+]e was calculated. Quantification of steady-state [Ca<sup>2</sup>+]e after a cumulative of 80, 140, and 180 μM CaCl2 during delay of pore opening (**C**) and cumulative of 100, 120, and 140 μM CaCl2 during rescue of mitochondria from mPTP opening (**F**). Error bars represent mean ± SEM (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.005). Arrowhead indicates time of addition of DMSO, ADP, OMN, OMN+ADP, or CsA during Protocol B.

We plotted steady-state [Ca<sup>2</sup>+]e (ss[Ca<sup>2</sup>+]e) as a function of cumulative added CaCl2 at each pulse (Figure 2B). The detailed dynamics of ss[Ca<sup>2</sup>+]e for the initial four CaCl2 pulses, in each group, are illustrated in the enlarged scale inset (Figure 2A). The exposure of mitochondria to DMSO and OMN, followed by repeated boluses of CaCl2, resulted in a gradual increase in ss[Ca<sup>2</sup>+]e with less mitochondrial Ca2+ uptake and rapid Ca2+ release by the third or fourth CaCl2 pulse. The total CRC for the DMSO and OMN were comparable (i.e., 133.3 ± 13.3 nmol Ca<sup>2</sup>+/mg protein and 146.6 ± 13.3 nmol Ca<sup>2</sup>+/mg protein, respectively) (Figure S1). In the presence of ADP, mitochondria took up more Ca2+ before pore opening and the CRC was further augmented with OMN+ADP; the CRC value increased from 213.3 ± 13.3 nmol Ca<sup>2</sup>+/mg protein for ADP alone to 373.3 ± 35.3 nmol Ca<sup>2</sup>+/mg protein for OMN+ADP (Figure S1). Mitochondria treated with CsA before the addition of CaCl2 boluses displayed a more robust Ca2+ uptake, with a significantly higher CRC value, 573.3 ± 26.6 nmol Ca<sup>2</sup>+/mg protein, compared with all other groups (Figure 2A and Figure S1). Importantly, in CsA-treated mitochondria, the addition of CaCl2 pulses (20 μM) did not significantly increase the ss[Ca<sup>2</sup>+]e until the sixth to seventh pulse (Figure 2B), suggesting enhanced Ca2+ uptake. Figure 2C, summarizes the effects of the different treatments on the ss[Ca<sup>2</sup>+]e for cumulative additions of 80, 140, and 180 μM of Ca2+, which corresponds to the fourth, seventh, and ninth CaCl2 pulses, respectively. The addition of CsA strongly blunted the Ca2+-induced increase in ss[Ca<sup>2</sup>+]e by stimulating faster and more Ca2+ uptake. We observed that the ss[Ca<sup>2</sup>+]e was significantly lower for the CsA-treated mitochondria than for OMN+ADP-treated mitochondria after the cumulative addition of CaCl2 of 80 μM (0.28 ± 0.0 μM vs. 0.50 ± 0.02 μM), 140 μM (0.57 ± 0.06 μM vs. 0.77 ± 0.06 μM), and 180 μM (0.71 ± 0.02 μM vs. 0.95 ± 0.1 μM) CaCl2, respectively (Figure 2B,C). The sustained low ss[Ca<sup>2</sup>+]e for an extended period of CaCl2 additions in the CsA-treated group indicates a maintained ΔΨm for Ca2+ uptake, resulting in enhanced Ca2+ loading capacity because of improved buffering.

We next examined [Ca<sup>2</sup>+]e dynamics in the situation in which mitochondrial matrix Ca2+ nearly reached threshold, as determined by the predicted opening of mPTP; in this case we added CsA just before the anticipated mPTP opening. Using Protocol B (Figure 1), [Ca<sup>2</sup>+]e was measured and the kinetics were compared in response to adding either DMSO, ADP, OMN, OMN+ADP, or CsA just before the onset of the mPTP opening (Figure 2D). The dynamic changes in ss[Ca<sup>2</sup>+]e during the addition of different treatments are illustrated in more detail in the inset of Figure 2D. In the DMSO-treated mitochondria, three to four Ca2+ pulses (cumulative addition of 68 ± 4.9 μM CaCl2) were sufficient to induce the release of matrix Ca2+. Addition of ADP or OMN reversed the initial pore opening and delayed matrix Ca2+ release by one to two pulses compared to DMSO. Addition of OMN+ADP also showed a significant reversal of Ca2+ release with reduction in the ss[Ca<sup>2</sup>+]e (0.74 ± 0.18 μM). Thus, there was a considerable increase in the CRC by OMN+ADP compared to DMSO (Figure 2D and Figure S1) and a further delay in mPTP opening by one additional CaCl2 bolus compared to OMN or ADP alone. More impressively, the addition of CsA not only reversed the increasing trend of ss[Ca<sup>2</sup>+]e to the baseline levels (0.42 ± 0.06 μM) (Figure 2D,E), it further maintained the ss[Ca<sup>2</sup>+]e at a constant low value for an additional twelve to thirteen Ca2+ pulses. This resulted in a four-fold and a two-fold increase in the CRC, compared to DMSO and OMN+ADP, respectively (Figure 2E,F and Figure S1).

Altogether, these results demonstrate that CsA enhances mitochondrial Ca2+ uptake, thereby inhibiting a consequent increase in free [Ca<sup>2</sup>+]e during CaCl2 pulse challenges, leading to an increase in the CRC of the mitochondria. This sustained low ss[Ca<sup>2</sup>+]e and concomitant increase in Ca2+ uptake are likely explained by enhanced [Ca<sup>2</sup>+]m buffering to maintain basal [Ca<sup>2</sup>+]m, which results in a preserved ΔΨm for Ca2+ uptake and greater CRC. To investigate further the potential for CsA on mediating Ca2+ buffering, it was necessary to examine the effects of CsA on matrix [Ca<sup>2</sup>+]m dynamics in the next set of experiments.

#### *3.2. E*ff*ect of CsA on Matrix Free [Ca<sup>2</sup>*+*] Handling*

Matrix Ca2+ was assessed with Fura-4 AM as described in Materials and Methods. We explored the effect of CsA on [Ca<sup>2</sup>+]m, under identical conditions and protocols as shown in Figure 1 (Protocols A,B). Mitochondrial Ca2+ buffering was measured as a function of a decrease in Ca2+ fluorescence, reaching a steady-state at approximately 270 s after each bolus of CaCl2 added. The magnitude of mitochondrial Ca2+ uptake for the first CaCl2 pulse (20 μM) was similar in all groups; however, on subsequent additions of CaCl2, the ADP- and/or OMN-treated groups showed faster declines in [Ca<sup>2</sup>+]m with lower mitochondrial steady-state [Ca<sup>2</sup>+]m (ss[Ca<sup>2</sup>+]m) and delayed mPTP opening compared to DMSO (Figure 3A,B). Interestingly, the CsA-treated group showed a small increase in ss[Ca<sup>2</sup>+]m with each CaCl2 pulse, but a gradual decline in ss[Ca<sup>2</sup>+]m was observed after [Ca<sup>2</sup>+]m exceeded 3 ± 0.10 μM with the cumulative addition of 100–150 μM CaCl2 (Figure 3A,B) and a significant increase in CRC up to fifteen to sixteen pulses. This suggested that the buffering effect of CsA on matrix Ca2+ is triggered when [Ca<sup>2</sup>+]m reaches a certain value.

**Figure 3.** Effect of CsA and AdN on intra-matrix free Ca2+ ([Ca<sup>2</sup>+]m) dynamics. Mitochondrial Ca2+ uptake and buffering for each treatment groups, DMSO (control; black trace), CsA (red trace), ADP (brown trace), oligomycin (OMN, blue trace), or combination of OMN+ADP (green trace) are shown using the protocols depicted in Figure 1. Mitochondrial suspension was exposed to 0.5 μM CsA, 250 μM ADP, 10 μM OMN, or OMN+ADP before adding boluses of 20 μM CaCl2 (Protocol A; left column). Mitochondrial suspension was exposed to added boluses of CaCl2 (20 μM) and rescued from mPTP opening (Protocol B; right column) with similar interventions as in Protocol A, at a time point at which it would initiate mPTP opening. Representative traces show changes in [Ca<sup>2</sup>+]m over time in delay of mPTP opening (**A**) and rescue of mitochondria from mPTP opening (**D**). Insets (A,D) show Ca2+ uptake kinetics in detail. Steady-state [Ca<sup>2</sup>+]m (ss[Ca<sup>2</sup>+]m), 270 s after initiation of Ca2+ uptake, plotted as function of added Ca2+ (20 μM) every 300 s in delay of mPTP opening (**B**) and rescue of mitochondria from mPTP from opening (**E**). Insets (B,E) indicate the time points at which ss[Ca<sup>2</sup>+]m was calculated. Change in matrix-bound Ca2+:free Ca2+ over time in delay of mPTP opening (**C**) and rescue of mitochondria from mPTP opening (**F**). Arrowhead indicates time of addition of DMSO, ADP, OMN, OMN+ADP, or CsA during Protocol B.

To estimate the mitochondrial Ca2+ buffering capacity, the ratio of bound Ca2+:free Ca2+ was calculated from the change in [Ca<sup>2</sup>+]m (Figure 3A) to the total amount of Ca2+ taken up from the extra-matrix medium (ΣCa2+uptake): (ΣCa2+uptake-[Ca<sup>2</sup><sup>+</sup>]m)/[Ca<sup>2</sup><sup>+</sup>]m, as described previously [45]. Although the extent of bound Ca2+:free Ca2+ at each Ca2+ pulse was comparable in all the treated groups (Figure 3C), the addition of CsA maintained the buffering capacity, with a gradual increase in the capacity to bind Ca2+ up to fifteen or sixteen Ca2+ pulses (Figure 3C).

Greater uptake of Ca2+ from the extra-matrix space (indicated by lower ss[Ca<sup>2</sup>+]e), combined with lower ss[Ca<sup>2</sup>+]m, indicated a greater Ca2+ buffering capacity of mitochondria in the presence of CsA. Consistent with this notion, the calculated matrix Ca2+ buffering capacity (mβCa) in CsA-treated mitochondria was about ten-fold higher compared to DMSO and two-fold higher than with OMN+ADP (Figure 4). This CsA-mediated increase in <sup>m</sup>βCa is possibly due to an effect of CsA in triggering the matrix physiological buffers to enhance sequestration of Ca2+.

**Figure 4.** Effect of CsA and AdN on mitochondrial Ca2+ buffering capacity. Mitochondrial Ca2+ buffering capacity calculated from trend fits of [Ca<sup>2</sup>+]e and [Ca<sup>2</sup>+]m for DMSO-(control), CsA-, and OMN+ADP-treated mitochondria as described by Equations (2)–(4) in Materials and Methods. Buffering capacity for each treatment was calculated from three-five experiments each for [Ca<sup>2</sup>+]e and [Ca<sup>2</sup>+]m and averaged. Error bars represent mean ± SEM (\* *p* < 0.01 compared with DMSO).

After observing the high buffering capacity of mitochondria pre-treated with CsA before the CaCl2 bolus challenges, we next examined the effect of CsA on the rescue of mitochondria from Ca2+ release when the matrix Ca2+ buffering system (MCBS) becomes overwhelmed by the added boluses of CaCl2 (Figure 1B). As shown in Figure 3D,E, OMN and ADP, each failed to reverse the mitochondrial Ca2+ efflux with added boluses; however, adding CsA or OMN+ADP at similar time points significantly reduced ss[Ca<sup>2</sup>+]m by reinstating Ca2+ sequestration. This reversal was more effective and sustained in the presence of CsA than with OMN+ADP. This observation is consistent with the calculated values of bound Ca2+: free Ca2+ , which increased two-fold for CsA compared to OMN+ADP (Figure 3F). Taken together, these data demonstrate that CsA increases the mitochondrial Ca2+ threshold for mPTP opening by activating [Ca<sup>2</sup>+]m buffering that results in maintenance of a low ss[Ca<sup>2</sup>+]m.

#### *3.3. E*ff*ect of CsA on Ca2*+*-Mediated Changes in* ΔΨ*m, NADH, and Matrix pH*

A major driving force for Ca2+ uptake, in addition to the chemical gradient, is a high IMM potential gradient (ΔΨm); but increased Ca2+ uptake without efflux or sequestration can decrease ΔΨm by flooding the matrix with positive charges. To strengthen the thesis that CsA increases the capacity of mitochondria to sequester Ca2+, we next investigated the effect of CsA on mitochondrial bioenergetics. ΔΨm, NADH, and pHm were assessed using the same protocols as described in Figure 1 for CRC to correlate changes in [Ca<sup>2</sup>+]m to changes in bioenergetics over time. mPTP opening was marked by a sudden rise in the TMRM signal, indicating maximal depolarization of Ψm. Correspondingly, the oxidation of NADH was marked by a decrease in matrix NADH signal intensity when mPTP opens. Figure 5 shows representative traces of ΔΨm, NADH, and pHm for each experimental condition. The rate of ΔΨm depolarization and NADH oxidation correlated well with the induction of mPTP, as seen in the CRC data. The loss of CRC coincided with total ΔΨm

dissipation and NADH oxidation. DMSO-treated mitochondria (control) exhibited rapid Ca2+-induced ΔΨm depolarization and NADH oxidation (black trace) after only a few CaCl2 pulses. Addition of OMN+ADP significantly delayed the Ca2+ induced ΔΨm depolarization and NADH oxidation when compared to DMSO, with 533.3 ± 26.7 nmol Ca<sup>2</sup>+/mg protein vs. 200 ± 23 nmol Ca<sup>2</sup>+/mg protein and 546.7 ± 18.9 nmol Ca<sup>2</sup>+/mg protein vs.173.3 ± 13.3 nmol Ca<sup>2</sup>+/mg protein Ca2+ capacity, respectively (Figure 5A,B). Mitochondria treated with CsA maintained ΔΨm and NADH for a higher number of CaCl2 pulses than with OMN+ADP (666.7 ± 13.3 nmol Ca<sup>2</sup>+/mg protein, and 626.6 ± 13.3 nmol Ca2+ /mg protein, respectively) (Figure 5A,B). Mitochondrial matrix pH (pHm) is known to modulate mitochondrial Pi concentration and thus influence the matrix Ca2+ buffering [14]. The presence of CsA maintained pHm at a basal level until mPTP opened (Figure 5C).

**Figure 5.** Effect of CsA and AdN on mitochondrial bioenergetics. The bioenergetic responses during Protocols A (left column) and B (right column) were monitored using the ΔΨm sensitive dye TMRM (tetramethylrhodamine methyl ester perchlorate) (**A**,**D**), NADH autofluorescence (**B**,**E**), and pHm-sensitive dye BCECFAM (2,7-Bis-(2-Carboxyethyl)-5-(and-6)-carboxyfluorescein, acetoxymethyl ester AM) (**C**,**F**). Purple arrowhead indicates time of addition of DMSO (1 μM), ADP (250 μM), OMN (10 μM), OMN+ADP, or CsA (0.5 μM) during Protocol B.

In Protocol B, intervention with OMN + ADP or CsA maintained ΔΨm, mitochondrial NADH, and pHm (Figure 5D–F), and contributed to the improved capacity of mitochondria to take up and sequester additional Ca2+ after CaCl2 pulses. However, OMN+ADP was less effective in preserving ΔΨm, NADH, and pHm compared to CsA. This incapacity to sustain the bioenergetic status in the OMN+ADP- vs. CsA-treated mitochondria during CaCl2 challenges reflects a lower capacity to sequester Ca2+ in the matrix for a protracted time.

In summary, maintenance of ΔΨm, NADH, and pHm in the presence of CsA is consistent with changes in [Ca<sup>2</sup>+]e and [Ca<sup>2</sup>+]m that reflect greater Ca2+ sequestration (Figure S2) and uptake. Collectively, these results indicate that CsA reduced the accumulation of [Ca<sup>2</sup>+]m, by potentiating matrix Ca2+ buffering, which in turn, maintained ΔΨm, NADH, and pHm necessary for normal mitochondrial function. Together, these mitochondrial variables preserve mitochondria and protect against mPTP opening.

#### *3.4. Time Dependent E*ff*ect of CsA Addition on Rescue of Mitochondria from Imminent Ca2*+*-Induced mPTP Opening*

After demonstrating that CsA can reverse the induction of mPTP opening (Figures 2, 3 and 5), we next investigated the dynamics of [Ca<sup>2</sup>+]e, [Ca<sup>2</sup>+]m and ΔΨm, by adding CsA at three different time points, before the onset of mPTP opening. This approach allowed us to determine the threshold at which CsA can effectively restore the mitochondrial sequestration system that will protect mitochondria from Ca2+ overload-mediated pore opening. Figure 6, panels A-C, show changes in [Ca<sup>2</sup>+]e, [Ca<sup>2</sup>+]m, and ΔΨm depolarization, induced by adding CsA at 1, 2, and 3 min after the last CaCl2 bolus in which mitochondrial Ca2+ uptake was observed before pore opened. Right panels D-F show detailed (close up) comparison of kinetics of [Ca<sup>2</sup>+]e, [Ca<sup>2</sup>+]m, and ΔΨm after adding CsA at different time points. Adding CsA at all three tested time points, markedly delayed the large increase in [Ca<sup>2</sup>+]e due to mitochondrial Ca2+ release. However, the effect of CsA to prolong Ca2+ uptake, which eventually maintains ss[Ca<sup>2</sup>+]e at baseline, diminished as the interval before CsA addition and [Ca<sup>2</sup>+]e accumulation was lengthened (Figure 6A). Adding CsA at 1 min caused a decline in [Ca<sup>2</sup>+]e, with a marked decrease in ss[Ca<sup>2</sup>+]e (0.39 ± 0.07 μM) of the succeeding Ca2+ pulses, compared to adding CsA at 2 min (0.67 ± 0.03 μM) and 3 min (0.84 ± 0.05 μM) (Figure 6D; inset). In addition, we examined for changes in kinetics of [Ca<sup>2</sup>+]m with CsA added at the same time points (Figure 6B). The rate of maximal Ca2+ buffering (i.e., the time to reach steady-state [Ca<sup>2</sup>+]m) and the Ca2+ threshold for pore opening was significantly higher when CsA was added at the early time points (i.e., 1 and 2 min) compared to the late time point of 3 min (Figure 6E, inset).

Next, in a parallel study, we monitored the corresponding changes in ΔΨm profile at the same rescue time points (1, 2, or 3 min). Adding CsA reversed the Ca2+-induced ΔΨm depolarization even after a large depolarization (i.e., at 3 min; Figure 6C,F). Similar to its effect on [Ca<sup>2</sup>+]e and [Ca<sup>2</sup>+]m, CsA restored and maintained ΔΨm for a longer period at rescue points of 1 min vs. 2 and 3 min. Thus, at these points of intervention, CsA suppressed mPTP opening by increasing matrix Ca2+ buffering capacity, which maintained ΔΨm and the driving force for further Ca2+ uptake (Figure 6C). Together, these results demonstrate that the magnitude of CsA-mediated increase in Ca2+ threshold for mPTP opening and maintenance of mitochondrial integrity is dependent on the [Ca<sup>2</sup>+]m level before CsA intervention.

**Figure 6.** Time-dependent effects of CsA on mitochondrial Ca2+ dynamics and bioenergetics during rescue of mitochondria from mPTP opening. Changes in (**A**) extra-matrix free Ca2+ ([Ca<sup>2</sup>+]e), (**B**) intra-matrix free Ca2+ ([Ca<sup>2</sup>+]m) and (**C**) ΔΨm, when CsA was added at 1 min (blue trace), 2 min (green trace), and 3 min (red trace) after the last Ca2+ bolus before another Ca2+ bolus would have caused mPTP opening. Right panels show the effect of CsA on (**D**) [Ca<sup>2</sup>+]e, (**E**) [Ca<sup>2</sup>+]m, and (**F**) ΔΨm dynamics during rescue of mitochondria from mPTP opening in greater detail. Insets (**D**,**E**) show relative ss[Ca<sup>2</sup>+]e and decay time constant (ms) at specified time points (black dotted box), respectively. Arrows indicate time of addition of CsA (0.5 μM). Error bars represent mean ± SEM (\* *p* < 0.05, \*\* *p* < 0.01 vs. 3 min and # *p* < 0.05 vs. 2 min).

#### *3.5. Role of Inorganic Phosphate in CsA-Induced [Ca<sup>2</sup>*+*]m Regulation.*

Inorganic phosphate (Pi) is a required component for mitochondrial matrix Ca2+ buffering [14,29]. To gain insight into the mechanism that underlies CsA-mediated activation of the MCBS, we monitored mitochondrial Ca2+ handling and ΔΨm during repeated boluses of 20 μM CaCl2 every 5 min, as described in Materials and Methods, but now in the absence of Pi. With mitochondria depleted of Pi, and in Pi-free media, the CRC of mitochondria treated with CsA before the CaCl2 pulses was not different from DMSO (control). In addition, these mitochondria showed a gradual increase in ss[Ca<sup>2</sup>+]e and interestingly, after cumulative additions of CaCl2 to 80 ± 15 μM, there was a significant decrease in mitochondrial Ca2+ uptake during additional CaCl2 pulses (Figure 7A). These results implicated a Pi-dependent mechanism in the CsA-mediated delay in mPTP opening. In contrast, ADP and OMN+ADP, but not OMN alone, caused a significant delay in mPTP opening (Figure 7A) in the absence of Pi.

**Figure 7.** Effect of Pi on CsA-induced mitochondrial Ca2+ handling and bioenergetics. Time course of [Ca<sup>2</sup>+]e (**A**), [Ca<sup>2</sup>+]m (**B**), ΔΨm (**C**), and matrix-bound Ca2+:free Ca2+ (**D**) during consecutive additions of 20 μM CaCl2 to a suspension of Pi-depleted mitochondria, pre-exposed to DMSO (control), CsA, ADP, OMN, or OMN+ADP.

Along with observing the Pi-mediated effect of CsA on [Ca<sup>2</sup>+]e dynamics, we also measured [Ca<sup>2</sup>+]m under identical conditions. In the absence of Pi, mitochondria showed a gradual increase in ss[Ca<sup>2</sup>+]m; matrix Ca2+ sequestration was strongly blunted in both DMSO-and CsA-treated groups. This reflected diminished buffering capacity with the increase in [Ca<sup>2</sup>+]m (Figure 7B). However, in the presence of ADP and OMN+ADP in the Pi-depleted condition, mitochondria displayed robust CRC and enhanced Ca2+ buffering and thus decreased [Ca<sup>2</sup>+]m (Figure 7B). Intriguingly, this effect was stronger than in the Pi replete condition (Figure 3). Mitochondria also showed an increased ratio of bound Ca2+:free Ca2+ in the OMN+ADP-treated group, but not in the DMSO and CsA groups (Figure 7D). These data further support the premise that Pi is crucial in CsA-induced matrix Ca2+ buffering and Pi is a requisite component of matrix calcium sequestration.

Since we observed significant attenuation of Ca2+ uptake and buffering by CsA in the absence of Pi, we addressed how the altered mitochondrial Ca2+ dynamics impacted ΔΨm. Analysis of ΔΨm in mitochondria depleted of Pi during CaCl2 bolus challenges revealed a gradual depolarization with each Ca2+ pulse over time in the DMSO-, OMN-, and CsA-treated groups (Figure 7C); this was consistent with the low CRC in these three groups due to the poor buffering after additional CaCl2 pulses. In contrast, mitochondria exposed to ADP or OMN+ADP in the Pi-depleted state exhibited restored and sustained ΔΨm, which supported a robust CRC (Figure 7C).

To further confirm the requisite role of Pi in mediating CsA-induced activation of the MCBS, a rescue experiment with 5 mM Pi was performed with DMSO-and CsA-treated groups in Pi-depleted condition. With addition of deionized H2O (vehicle), pore opening was not prevented in either group (data not shown). The addition of exogenous Pi to the buffer triggered a rapid reversal of Ca2+ release (decrease in ([Ca<sup>2</sup>+]e) in parallel with complete restoration of ΔΨm (Figure 8). In contrast, additional Ca2+ pulses in the Pi free DMSO-treated group failed to maintain ss[Ca<sup>2</sup>+]e and basal Ψm, and induced rapid Ca2+ efflux (Figure 8). However, the CsA-treated mitochondria showed a robust uptake of [Ca<sup>2</sup>+]e with low ss[Ca<sup>2</sup>+]e and sustained ΔΨm maintenance with additional CaCl2 boluses (Figure 8). Taken together, these results establish that Pi is required for CsA-mediated mitochondrial Ca2+ buffering that maintains low [Ca<sup>2</sup>+]m and preserves ΔΨm; this in turn contributes to the capacity for more Ca2+ uptake and thus increases the Ca2+ threshold for mPTP opening.

**Figure 8.** Mitochondrial Ca2+ modulation by CsA is phosphate (Pi)-dependent. Representative traces show change in extra-matrix Ca2+ fluorescence (Fura-4F Ratio) and ΔΨm during consecutive 20 μM CaCl2 boluses to induce mPTP opening in Pi-depleted mitochondria. Pi was added (purple arrowhead) at threshold point when mitochondria exhibited limited uptake of Ca2+ from the buffer.
