Calcium-induced precipitate formation in brain mitochondria: composition, calcium capacity, and retention

Abstract

Both isolated brain mitochondria and mitochondria in intact neurons are capable of accumulating large amounts of calcium, which leads to formation in the matrix of calcium- and phosphorus-rich precipitates, the chemical composition of which is largely unknown. Here, we have used inhibitors of the mitochondrial permeability transition (MPT) to determine how the amount and rate of mitochondrial calcium uptake relate to mitochondrial morphology, precipitate composition, and precipitate retention. Using isolated rat brain (RBM) or liver mitochondria (RLM) Ca(2+)-loaded by continuous cation infusion, precipitate composition was measured in situ in parallel with Ca(2+) uptake and mitochondrial swelling. In RBM, the endogenous MPT inhibitors adenosine 5'-diphosphate (ADP) and adenosine 5'-triphosphate (ATP) increased mitochondrial Ca(2+) loading capacity and facilitated formation of precipitates. In the presence of ADP, the Ca/P ratio approached 1.5, while ATP or reduced infusion rates decreased this ratio towards 1.0, indicating that precipitate chemical form varies with the conditions of loading. In both RBM and RLM, the presence of cyclosporine A in addition to ADP increased the Ca(2+) capacity and precipitate Ca/P ratio. Following MPT and/or depolarization, the release of accumulated Ca(2+) is rapid but incomplete; significant residual calcium in the form of precipitates is retained in damaged mitochondria for prolonged periods.

Fig. 1

Mitochondrial permeability transition (MPT) inhibitors increase calcium-loading capacity in brain mitochondria. Rat brain mitochondria in a suspension were infused with a buffered CaCl₂ solution at 100 nmol Ca²⁺/mg/min without MPT inhibitors or in the presence of 1 μmol/L cyclosporine A (CsA), 0.2 mmol/L adenosine 5′-diphosphate (ADP) plus 1 μg/mL oligomycin, ADP/oligomycin plus 1 μmol/L CsA, or 3 mmol/L adenosine 5′-triphosphate (ATP). Fura-6F (250 nmol/L) was present, as appropriate. (a) Traces of medium free Ca²⁺ concentration ([Ca²⁺]_e) as a function of time. Infusion begins at the arrow. The time to onset of MPT – taken as the sharp upward inflection of [Ca²⁺]_e – reveals a substantial protective effect for the endogenous MPT inhibitors ADP and ATP. CsA, which is ineffective when added alone, greatly amplifies the protective effect of ADP. (b) Inset: In the absence of mitochondria, the actual rate of [Ca²⁺]_e increase in the presence of ATP is ~70% of that for ADP because ATP has a significantly higher Ca²⁺ binding affinity. Main panel: Traces of light scattering intensities plotted as a function of the amount of available free Ca²⁺ by correcting for the effective slower infusion rate in the presence of ATP. Increased light scattering during Ca²⁺ infusion reflects mitochondrial Ca²⁺ uptake and precipitate formation. Maxima in these traces correlate well with the inflection points in panel (c) and coincide with MPT. (c) Traces of [Ca²⁺]_e as a function the amount of added free Ca²⁺. Ca²⁺ loading capacities, estimated as the cumulative amount of infused Ca²⁺ at the inflection point, are ~400 nmol/mg and ~600 nmol/mg in the presence of ATP and ADP, respectively. CsA strongly amplifies the protective effect of ADP, increasing the CLC to ~1500 nmol/mg.

Fig. 2

Mitochondrial calcium accumulation leads to precipitate formation. Digital transmission electron micrographs of high-pressure frozen rat brain mitochondria suspensions, freeze-substituted by means of a protocol that preserves calcium phosphate precipitates. (a) Control mitochondria are structurally normal. They are spherical and unswollen, with a homogeneous dense matrix and normal cristae; membranes are smooth and unbroken. (b) Mitochondria frozen after Ca²⁺ infusion for 200 s without mitochondrial permeability transition (MPT) inhibitors are swollen and damaged, but precipitates are rare. (c) In mitochondria frozen after 400 s Ca²⁺ infusion in the presence of 0.2 mmol/L adenosine 5′-diphosphate (ADP) with 1 μg/mL oligomycin, multiple precipitate foci are present in almost all mitochondria, which are also generally swollen. (d) Mitochondria frozen after 400 s Ca²⁺ infusion in the presence of 5 mmol/L adenosine 5′-triphosphate appear generally similar to those infused in an ADP medium. In all cases, mitochondria were frozen at times just after the presumed onset of MPT, as indicated by an increase in [Ca²⁺]_e (see Fig. 1). Bar (all panels), 1 μm.

Fig. 3

Calcium capacity and accumulation rates influence precipitate composition. (a) Representative energy-dispersive X-ray spectra recorded from mitochondrial precipitates formed in the presence of adenosine 5′-triphosphate (ATP) (solid line) or adenosine 5′-diphos-phate (ADP) (dashed line). Elements corresponding to the major characteristic K-line X-ray peaks are identified. Calcium and phosphorus are the major inorganic elements in precipitates, but the molar Ca/P ratio (approximately reflected by the ratio of areas under their characteristic X-ray peaks) is higher in the presence of ADP than with ATP. For clarity, the x-axis of the ADP spectrum is offset by +500 counts. (b) In rat brain mitochondria (RBM), stronger mitochondrial permeability transition (MPT) inhibition, e.g. ADP or ADP + cyclosporine A (CsA) and faster Ca²⁺ infusion favors relatively high (~1.5) Ca/P precipitate ratios, while slower infusion, e.g. ATP or ADP at 70 nmol/mg/min, leads to formation of precipitates with lower (~1.2) ratios. See also Table 1. Rat liver mitochondria (RLM) fail immediately at standard infusion rates (not shown), but at slow infusion rates in the presence of ADP they form precipitates that have relatively low Ca/P ratios. Ratios are higher if MPT is strongly inhibited by CsA. Symbols indicate a significant decrease relative to fast Ca²⁺ infusion of an RBM suspension in the presence of ADP (leftmost bar; **p < 0.001) or an increase in the presence of CsA and ADP, relative to ADP only, for RLM (#p < 0.05). (c) Traces of extramitochondrial free calcium ([Ca²⁺]_e) as a function of the amount of free Ca²⁺ added at 50 nmol/mg/min in the presence of ADP show that the Ca²⁺ capacity of RLM is much less than RBM under comparable conditions. Similar to RBM, however, the higher Ca²⁺ capacity when CsA is added correlates with a higher Ca/P ratio (panel b).

Fig. 4

Precipitate formation is partially supported by adenosine 5′-triphosphate (ATP) hydrolysis. (a) Records of extramitochondrial free calcium ([Ca²⁺]_e) versus time for rat brain mitochondria (RBM) suspended in a K⁺-based buffer containing ATP but no added phosphate reveal that the Ca²⁺ loading capacity (CLC) of RBM is reduced from ~400 to ~250 nmol Ca²⁺/mg protein in the presence of oligomycin (compare solid curves, ±oligomycin). This indicates that inorganic phosphate (P_i) derived from ATP hydrolysis supports Ca²⁺ sequestration. Further reduction of phosphate availability by depleting endogenous P_i (cf. text for rationale and Materials and methods for procedure) further reduces RBM Ca²⁺ capacity (compare dotted curves). The effects are additive, so that Ca²⁺ uptake and sequestration are essentially eliminated when both P_i sources are removed (leftmost curve). (b) Electron micrographs confirm that precipitates are evident when ATP hydrolysis alone supplies phosphate (P_i-depleted, oligo (−); top panel), but are not formed in the complete absence of phosphate (P_i-depleted, oligo (+); bottom panel). Electron micrographs are from conventional preparations. Bar, 1 μm.

Fig. 5

Only part of the mitochondrial calcium load is rapidly released upon pore opening and depolarization. (a) Trace of extramitochondrial free calcium ([Ca²⁺]_e) versus time for rat brain mitochondria (RBM) Ca²⁺-loaded in the presence of adenosine 5′-triphosphate (ATP). Infusion was terminated at the onset of mitochondrial permeability transition (MPT) (stop). Depolarization with 1 μmol/L carbonyl cyanide 4-(trifluoro-methoxy)phenylhydrazone (FCCP), added ~250 s later, elicited a large but incomplete burst of Ca²⁺ release, which was unaffected by the addition of 5 μg/mL oligomycin or 2 μmol/L rotenone at the points indicated. Subsequent application of alamethicin, 40 μg/mg protein, released a significant amount of residual Ca²⁺. (b) Digital transmission electron micrograph of high-pressure frozen, freeze-substituted RBM incubated as in panel (a) and frozen 200 s after application of FCCP (no other drugs) confirms that residual precipitates are retained in both swollen and non-swollen mitochondria. Bar, 1 μm.