Morphological alterations in the mitochondria of Amoeba Proteus induced by uncoupling agents

Isolated mitochondria show structural variations which are dependent on metabolic activity, and possibly related to respiratory cycles (Hackenbrock, 1966; Muscatello, Guarriera-Bobyleva & Buffa, 1972), differences in ion accumulation (Packer, Wriggles-worth, Fortes & Pressman, 1968) and/or to changes in nucleotide binding and translocation (Weber, 1972; Scherer & Klingenberg, 1974). Mitochondrial structural differences observed in situ have also been correlated with specific alterations in mitochondrial functioning (Hackenbrock, Rehn, Weinbach & Lemasters, 1971) and with overall changes in cell activity (Haydon, Smith & Seligamn, 1967; Innis, Beers & Craig, 1976; Rosano & Jones, 1976).

Different mitochondrial forms are preserved within healthy individual control Amoeba (Flickinger, 1968,1974; Ord, 1976; Smith, 1978) and in other protozoa such as Euplotes (Jurand & Lipps, 1973) following double aldehyde fixation. The co-existence of different types in close proximity within the same cell has been considered indicative of an actual chemical or physical difference between the mitochondria which may relate to functional or biogenetic events rather than an artefact of fixation. If so then these protozoa could prove valuable in attempts to link morphological changes with changes in mitochondrial functioning.

The present study was undertaken to determine whether treatments with chemicals which affected the functioning of the mitochondria in situ by uncoupling oxidation and phosphorylation induced recognizable changes in the mitochondria of Amoeba proteus, and if so whether these were similar for all chemicals which uncoupled these 2 activities. Uncoupling has previously been shown to cause mitochondrial structural changes both in in vitro preparations (Weinbach & Garbus, 1968; Blair & Munn, 1972; Muscatello, Guarriera-Bobyleva, Pasquali-Ronchetti & Ballotti-Ricci, 1975) and in whole cells (Buffa, Guarriera-Bobyleva, Muscatello & Pasquali-Ronchetti, 1970; Morisset, 1974). The 3 uncouplers chosen for use were: dinitrophenol (DNP), pentachlorophenol (PCP), and tn-chlorocarbonyl cyanide phenylhydrazone (CCCP), a particularly potent uncoupler in isolated mammalian systems (Heytler & Pritchard, 1962).

Amoeba proteus, strain PD_DAXJJ₆₉ were routinely cultured at 20 ± 1 °C and pH 5 9±o-i in a modified Chalkley’s medium (Chalkley, 1930; Ord, 1970). This is a weak salt solution and is henceforth referred to as the amoeba-medium. The amoebae were maintained using the Tetrahymena feeding regime of Griffin (1960).

Concentrated solutions of the 3 uncoupling agents were prepared and added to the amoebamedium to give the final concentrations required. Fresh solutions were always used as there was evidence of a reduction in potency (particularly evident for PCP) if aqueous solutions were left standing for any length of time. Treatments were carried out in watchglasses, the amoebae being allowed to attach to the substratum before the amoeba-medium was withdrawn and replaced by the uncoupler solution. At the end of the treatment periods the amoebae were returned to normal amoeba-medium and either cloned singly to assess cell activity and viability (Ord, 1978) or fixed for the electron microscope at defined periods following the removal of the agent.

Whole cell incubations in DNP, PCP and CCCP proved most effective when performed at pH 4, 4.5 and 6 respectively - values near the pK_a of each chemical. Since in vitro experiments with these uncouplers are successfully carried out at physiological pH’s, i.e. 7.2–7.4 (e.g. Blair & Munn, 1972; Muscatello et al. 1975), it was believed that the pH requirement for the amoeba was due to problems in penetration through the cell membrane as found with other whole-cell studies using uncoupling agents (Kahn, 1974). A means was, therefore, sought to enhance the membrane permeability of the amoebae so that the uncouplers could be tested at culturing pH. Treatments with dimethyl sulphoxide (Reiss, 1971) and sodium periodate (Sanders, 1969) were both investigated : although these reagents, particularly DM SO, did improve the performance of the uncouplers, they were not considered practical for general use as they introduced added cell changes which confused the interpretation of the uncoupler results. When direct injection by micropipette using DNP adjusted to pH 6.8 was investigated results similar to those for whole cell exposure to DNP at pH 4 were obtained. This confirmed that the crossing of the outer cell membrane by the uncoupler was creating the problem encountered at higher pHs. Since dose was more difficult to control for the injection technique, it was considered practical to use whole cell exposure at the pK_a of each chemical throughout the present investigation.

Nuclear transfers were performed with a Fonbrune micromanipulator at magnification 200 × using the oil chamber technique of Comandon & deFonbrune (1939). Microinjections were carried out using an oil-filled Alga microsyringe outfit (Wellcome Reagents Ltd, England), injecting volumes of approximately 60 μm’ at pH 6.8.

Amoebae were fixed for 1 h at 4 °C in freshly prepared 5% glutaraldehyde/4% formaldehyde buffered with 01 M cacodylate at pH 7.1 (Karnovsky, 1965). After a 5-min rinse in buffer, the cells were postfixed in 1% osmium tetroxide for i h, washed in buffer, then blocked in 2% agar (this increased the ease of handling during dehydration, embedding and sectioning, while avoiding centrifugation at any stage). Agar blocks of cells were dehydrated in a graded series of ethanol, stained for i h with 2% uranyl acetate at 50% ethanol, and embedded in Spurr epoxy resin (Taab Laboratories, Reading, England). Sections of 60–80 nm thickness were examined in a Philips 300 EM.

The effectiveness of each of the uncoupling agents was found to be dependent on 4 main factors: (i) The concentration to which the amoeba were exposed (Fig. 1); (ii) The length of the exposure period (Fig. 2); (iii) The pH of the treatment solution: each uncoupler was more potent at its pK_a where cell membrane penetration was improved (see methods); and (iv) The feeding condition of the treated cells: starved cells were more sensitive than well fed cells treated with DNP (Table 1).

Continual exposure of amoebae to any one of the uncouplers produced a similar pattern of behavioural changes in the form and activity of the cell. The following sequence of events resulted from the immersion of amoebae in an effective uncoupler (i.e. 5 ×10⁻⁵M DNP at pH 4; 5 × 10⁻⁶ PCP at pH 4.5 ; or 8 × 10⁻⁶M CCCP at pH 6). (a) Cell adhesion decreased, the cell losing its attachment to the substratum within 10 min of being placed in the uncoupler. (b) Pseudopod formation was significantly diminished by 20-30 min, and the functioning of the contractile vacuole gradually inhibited, (c) The cells generally assumed a spherical rosette configuration after approximately 1 h of exposure. (d) This rosette form was replaced within 3–4 h by a smooth membrane form, such membrane change frequently being accompanied by central aggregation of all large cytoplasmic organelles and a loss of any visible signs of organelle movement, (e) These changes became irreversible in the majority of amoebae by 6–7 h, the cell either cytolysing in the uncoupler or within 5 h following its removal. Control amoebae incubated at the lowered pHs but in the absence of the uncouplers did not display this series of activity changes.

The damage sustained by amoebae while exposed to uncouplers was reversible providing the cells were transferred to control culturing medium before the extreme stage was reached. Recovery began with reattachment of cells to the substratum in a rosette form. With short uncoupler incubations the return to normal was rapid with the extension of pseudopods from the rosette amoebae leading to normal locomotion within 1–2 h and the cell showing no significant change in cell cycle. As the length of the exposure to an uncoupler increased, the interval necessary for any surviving cell to recover normal activity increased correspondingly.

When the site of DNP damage in the living cell was investigated by means of reciprocal nuclear transfers between cells treated with 1 × 10⁻⁴M DNP at pH 4 for 6 h and untreated controls, cell damage was shown to be mainly cytoplasmic. Of 60 hybrid cells containing a ‘DNP nucleus + control cytoplasm’ 90% survived and divided to form clones as compared with a 33% survival for ‘DNP nucleus+DNP cytoplasm’ amoebae. The reciprocal transfer of a control nucleus into DNP-treated cytoplasm, on the other hand, produced no survivors. This inability of ‘control nucleus + DNP-treated’ cytoplasm amoebae to recover from the operation and divide appeared to be due to operational damage superimposed upon cytoplasm already changed by DNP treatment. DNP cytoplasm generally proved incapable of closing the membrane break made by either the outgoing treated nucleus or the incoming control nucleus. The membrane hole allowed a slow leakage of cytoplasm. Once approximately 30% of the cytoplasm had escaped in such cells, cytolysis was inevitable. This type of membrane leakage is probably induced indirectly by the mitochondrial lesion, since energy is required by the microfibres responsible for closing membrane tears (Jeon & Jeon, 1975).

The mitochondria of Amoeba proteus are preserved as 2 morphologically distinct forms which are present under normal conditions in an approximately 1:1 ratio (Fig. 3). These forms were initially called ‘dark’ and ‘light’ mitochondria, the difference in matrix density after aldehyde fixation being used primarily as the distinguishing feature (Flickinger, 1968). They have since been referred to as Type 1 (the dark matrix form) and Type II (the light matrix form) mitochondria by Ord (1976) and Smith (1978). This later nomenclature is used in the present communication, with mitochondrial types characterized by profile shape, cristal dimension and matrix density (Table 2).

Exposure of Amoeba proteus to sublethal doses of DNP, PCP or CCCP, resulted in a rapid loss of the control mitochondrial forms with changes involving all 3 distinguishing characteristics (Table 2). These distortions in mitochondrial structure, however, were reversible except when lethal to supralethal uncoupler doses were used. A gradual return of mitochondria of the 2 control types occurred when cells were transferred from the uncoupler to control culturing conditions. Ultrastructural changes were detected in the mitochondria before the living cell showed any reaction to the uncoupler, i.e. alterations in mitochondrial shape occurred with only a 2-min exposure to CCCP, but the first change in the living cell (loss of adhesion) was seldom visible before 10—20 min. The reverse was found on termination of treatment. Normal whole cell appearance and behaviour were re-established long before the mitochondria returned to normal, e.g. following a i-h exposure to either 8 × 10⁻⁵MCCCP or 1 × 10⁻⁴M DNP (pH 4) the amoebae required only i h in control amoeba-medium to resume normal locomotion but the mitochondria still retained an aberrant form for up to 4I1 and an abnormal ratio of Type I: Type II mitochondria for up to 20 h after removal from the uncoupler.

Although the general reaction of the mitochondria was similar for the 3 uncouplers, different timings for the lesions to occur and subtle differences in form were noted for each reagent

Incubation with 1 × 10⁻⁴M DNP at pH 4 induced recognizable changes in all mitochondria (Fig. 4). The mitochondria of control amoebae in amoeba-medium maintained at pH 4 were of normal appearance. The DNP treatment generated forms with a matrix of intermediate density, and cristae of an intermediate width, within 5 min of the beginning of treatment. Some profiles were 2–3 times normal length and often bent into an L-shape. If the DNP concentration was decreased to 1 × 10⁻⁵M, however, the incubation period had to be extended to induce mitochondrial change. Removal from DNP with a return to amoebae-medium proved the lesion resulting from exposures to 5 ×10⁻⁵ to 1 × 10⁻⁴M was reversible. The time needed for a return of normal mitochondrial forms was much longer than the time required to induce the lesion (Fig. 5).

PCP produced a more varied effect on the mitochondria than that found with exposure to DNP. Usually 2–-30% of a cell’s mitochondria retained control configurations while the remainder assumed an intermediate, often distorted, form. The matrix density, cristae and overall shape were all affected and in some profiles matrical inclusions were present. Abnormal forms were evident within 30 min on exposure to the sublethal dose of 5 × 10⁻⁸M at pH 4-5. With exposure to 1 x IO^-BM PCP at pH 4.5 for i h (a lethal dose) many mitochondria were broken with an apparent loss of organelle integrity.

CCCP induced the most striking alterations in the mitochondria compared with the other 2 uncouplers. Changes included gross alterations in profile shape, a matrix of intermediate density and cristae which were often almost indistinguishable. With 8 × 10⁻⁵M CCCP at pH 6, these changes were already present in 95-100% of the mitochondria with only a 2-min exposure, and after 1 h the morphology was greatly affected (Fig. 6). With CCCP, form changes were generated by doses (e.g. 1×10⁻⁵M for i h at pH 6) which had no observable effect on whole cell activity or viability either during the treatment time or on subsequent cloning (Fig. 7). Both the change in living cell activities and in the mitochondrial morphology were reversible if CCCP exposure was terminated at sublethal dose levels. Upon return to control culture conditions, however, the amoebae resumed normal behaviour and appearance well before control mitochondrial types were generated.

In the present study the exposure of Amoeba proteus to the 3 uncouplers of oxidative phosphorylation resulted in ultrastructural changes to 80–100% of the mitochondria with the loss of normal control forms. Mitochondria of control Amoeba proteus are normally present in 2 distinct forms: a condensed form with a dark matrix and wide cristae (Type 1), and a round, lighter form with narrower cristae (Type II). In active fed cells these are found in an approximate 1:1 ratio (though this varies if cell activity is interfered with (Smith, 1978)). Mitochondrial change by uncouplers was evidenced by replacement of the control Type I and Type II forms with elongated mitochondria of intermediate matrix density and cristae size. This was produced with rapidity and was always present before any significant changes could be recognised in the behaviour of the living cell. The change was dose dependent, with either an increase in concentration, or in exposure time, contributing to the alterations in form. As dose increased the mitochondrial changes became more extreme. At lethal levels the cristae were vacuolated and inclusions were present in the matrix. Though individual variations occurred, particularly in the timing and degree of change, it was seen that the major alterations to a form with a matrix intermediate between Type I and Type II forms was common with all 3 uncouplers.

The generation of the abnormalities with the elimination of the control forms occurred within 2 min from the beginning of an exposure period and suggested that the in vivo effects of the uncouplers were extremely rapid. This timing for treated whole cells compares well with in vitro timings of 30 s (Blair & Munn, 1972; Mus-catello et al. 1975), as in the latter no cell membrane barrier intervenes between the uncoupler and mitochondria. It is this cell membrane barrier which is believed responsible for the pH considerations encountered in whole-cell exposure studies. In the present work when the activity of each uncoupler was examined over the range pH 3’5–6, maximal effect was observed at a pH close to the pK_a (i.e. DNP at pH 4’0, PCP at 4’5, and CCCP at pH 6.0). When the penetration of the outer membrane was bypassed by directly injecting into the amoebae, then a pH common to in vitro studies could be employed. These results bear out an early theory of uncoupler action which proposed that only undissociated species of the weak acids penetrated intact cells, and that for each reagent concentrations existed which would cause complete inhibition of function at one pH (that near its pK_a) while having no effect at other pHs (Simon, 1953). Data to support this theory have been reported for other in vivo studies including those on the single-celled Euglena (Khan, 1974) and freshwater fish (Crandall & Goodnight, 1959). An inability to penetrate the membrane other than near its pK_a would explain the conflicting reports for DNP exposures of Amoeba proteus: i.e. toxicity limits of 4×10⁻⁵M when used at pH 4 (Chapman-Andresen, 1967), and the absence of effects even after 2 days with 5 ×10⁻⁴M but at pH 6.2 to 6.8 (Schumaker, 1958; Flickinger, 1972).

Ultrastructural changes in mitochondria were reversible even for the extreme variations seen in Fig. 6. However, the time required to reverse these form changes was of hours rather than the minutes taken to induce them. The return of a mitochondrial population to its ratio of Type I:Type II forms could require up to 24 b. Similar results have been found with subsequent aeration following anaerobic culturing and and exposure of amoebae to metabolic inhibitors (Smith, Bell & Ord, 1979).

Since (i) the major changes in mitochondrial form were common to all 3 uncouplers (ii) Type i and Type II forms were both affected, (iii) all changes were reversible providing the cells were removed from the uncouplers before they had received a dose, and (iv) the changes in mitochondrial morphology related well to those taking place in the activity of the cell, it is concluded that the mitochondrial changes observed in this study are reflecting a change in their functioning.

RAS was supported by an MRC studentship

MJ O is a member of the MRC Toxicology Unit, Carshalton, Surrey.

The authors wish to thank Dr E. A. Munn for his helpful suggestions and discussion of the work.