mitochondria of tetrahymena pyriformis

J. Cell Sci. 61, 437-451 (1983)

437

Printed in Great Britain © The Company of Biologists Limited 1983

MITOCHONDRIA OF TETRAHYMENA PYRIFORMIS: ENUMERATION AND SIZING OF ISOLATED ORGANELLES USING A COULTER COUNTER AND PULSE-HEIGHT ANALYSER R. K. POOLE Department of Microbiology, Queen Elizabeth College (University of London), Campden Hill, London W8 7AH, England

SUMMARY An electronic particle counter (Coulter Counter ZBI) and pulse-height analyser (Channelyzer C1000) have been used to measure numbers and sizes of mitochondria isolated from the ciliated protozoon Tetrahymena pyriformis. Differential centrifugation of disrupted organisms, followed by single-step sub-fractionation of the mitochondrial fraction on sucrose gradients yielded a population of organelles extensively enriched in the activities of mitochondrial marker enzymes. Gradientpurified mitochondria (approx. 3 X 109 particles (mg protein)" 1 ) were stable in electrolyte, exhibited unimodal volume distributions and were somewhat larger (0-93 ± 0-13(S.D.) jwn3; 19 preparations) than organelles in a crude mitochondrial fraction. Glutaraldehyde fixation of mitochondria in sucrose gradients decreased the apparent volume to 0-6 ± 0-06jUm3 (6 preparations). Based on the recovery in the mitochondrial fraction of mitochondrial membrane-bound cytochromes from a suspension of intact cells, the number of mitochondria per cell was estimated to be approximately 1440, representing 15 % of the total cell volume. Isolated mitochondria were osmotically sensitive and exhibited an apparent marked contraction on adding Ca2+ (10/ZM— lOmM). Addition of chloramphenicol (500/lgml~') to exponentially growing Tetrahymena cultures resulted in an almost immediate cessation of cell division and a dramatic decrease in cell volume. Mitochondria purified from such cells were much smaller than control mitochondria (0-21 ± 0-02|Um3; 7 measurements); their population density was approximately 900 per cell, equivalent to 11 % of total cell volume. The measurements of mitochondrial populations using the Coulter Counter and electron microscopy are in good agreement.

INTRODUCTION

The sizes of individual mitochondria and of mitochondrial populations in cells are generally determined microscopically. The use of high-voltage electron microscopy and serial sectioning have led to reappraisals of earlier estimates of mitochondrial population densities, and shown that, in some (e.g., see Hoffmann & Avers, 1973) but not all (Heywood, 1977) cells, 'giant' branched mitochondria ramify throughout the cell (for further references, see Lloyd & Turner, 1980). Phase-contrast or interference microscopy, combined with vital staining, may also be valuable in elucidating gross changes in mitochondrial sizes or populations (Lloyd, 1974). Alternative approaches are used for isolated organelles: sizes may be deduced from light scattering (for references, see Lehninger, 1964), sedimentation velocities (e.g., see Poole et al. 19716), and the weighing of pelleted mitochondria (Glas & Bahr, 1966). A potentially powerful, but as yet little exploited, method of determining

438

R. K. Poole

organelle sizes and numbers is the use of an electronic (resistive flow) particle counter coupled with a pulse-height analyser. Modern instruments that are capable of routinely counting and sizing bacteria (Kubitschek, 1969) should also be suitable for organelles like mitochondria. Early studies on rat liver mitochondria (Gebicki & Hunter, 1964; Gear & Bednarek, 1972) were limited by the inability to detect particles smaller than 0-4-0-6^m3. Pulse-height analysers, such as the Coulter Channelyzer, extend the usefulness of the general method by allowing volume distributions to be recorded in seconds and organelle swelling and shrinking to be followed (Heath, Coulson & Chimiklis, 1973; Schmidt, Martin & Vorbeck, 1977). In this paper, Coulter measurements of mitochondria from control and chloramphenicol-supplemented cultures of the ciliated protozoon Tetrahymenapyriformis are shown to be in good agreement with estimates obtained by other methods.

MATERIALS

AND

METHODS

Organism, growth conditions and harvesting of cultures T. pyriformis strain ST (from Dr Y. Suyama) was grown as described previously (Lloyd et al. 1971). Where indicated, D-chloramphenicol (500/igmr 1 ) was added as a dry powder to cultures otherwise ready to harvest (in the mid-exponential phase of growth) and the cultures were harvested 45-48 h later (Turner & Lloyd, 1971).

Preparation of mitochondrial

fractions

Homogenates were prepared from suspensions of washed cells (Lloyd et al. 1971) but using 60 strokes of a loosely fitting homogenizer (no. B 22087, Thomas, Philadelphia, U.S.A.). In some experiments, the pellet containing unbroken organisms and pellicles was resuspended in a small volume of buffer and re-homogenized to yield a second homogenate that was pooled with the first. The whole homogenate was centrifuged to yield a post-mitochondrial supernatant, which was decanted, together with most of a white loosely packed layer (Turner, Lloyd & Chance, 1971). The remaining, cream-coloured pellet (Pi) was further purified either by resuspending in disruption buffer and then repeating the above procedure twice (Turner et al. 1971) or by centrifuging through a sucrose gradient. Gradients (10 ml) were formed by applying the following layers (1-67 ml each) in 14ml centrifuge tubes: 2 0 , 1-076, 0-887, 0-698, 0-509 and 0-32M-sucrose, each buffered with lOmM-Tris-HCl to pH7-2. These gradients are similar in profile and composition to those used previously for satisfactory rate separation of mitochondria in a zonal rotor (Poole etal. 19716). Pellet Pi was resuspended gently in a minimum volume of disruption buffer; 0-5 ml samples were layered onto gradients and centrifuged for the times indicated in Results at 13090 g (r av , 11 • 1 cm) in a swingout rotor. The mitochondrial band was removed using a hypodermic syringe and needle, and diluted with disruption buffer. After centrifugation at 7000 g (at r av ), the mitochondrial pellet was finally resuspended in buffer. Slight departures from this procedure are described in Results.

Counting and sizing ofparticles in mitochondrial fractions using the Coulter Counter and Channelyzer A Coulter Counter model ZBI fitted with a probe having an aperture diameter of 30fim and a Channelyzer C1000 was used. Interference from mains 'noise' was reduced by a Radio Frequency Interference Filter L1829 (Belling & Lee Ltd, Enfield, Middlesex, England) and by housing the entire apparatus in a walk-in screened Faraday cage (Pickett & Lester, 1979). Volume distributions displayed on the Channelyzer oscilloscope were transformed into sets of 100 data points on punch tape using a Coulter Channelyzer Teleprinter Interface and a Data Dynamics or Westrex Teletype. Particle volumes and statistical analyses were calculated using a CDC 6600 Computer (Bazin, Richards & Saunders, 1975). The electrolyte contained 160mM-NaCl and lOmM-Tris, brought to

Quantification ofmicrobial mitochondria

439

pH 7-2 with 1 M-HCI. Before use, it was filtered under reduced pressure three times through a single membrane filter (0-2/zm diameter pores, type SM 11307, Sartorius) and then once through a layer of two such filters. The absence of a preservative such as azide required the filtration to be done immediately before counting. All vessels used for filtering and storing the electrolyte, as well as the sample containers (Accuvettes; Coulter Electronics Ltd, Luton, Beds), were rinsed at least twice with filtered electrolyte before use. Counts presented are the means of at least four counts of the same sample (two at each polarity) and are corrected for the background count of the electrolyte alone and for coincidence errors. Instrument settings were as follows: manometer volume, 0-05 ml; lower threshold, 2; upper threshold, off; edit switch, on; matching switch, 40K; count range, 1000; 1/current = 1; 1/amplification = 1/4 (particles from normal cells) or 1/8 (particles from chloramphenicol-treated cells). Base-channel threshold and window-width settings were varied as appropriate. Background counts were normally 300 to 1000 (1/amplification = 1/4) or 6000 to 7000 (l/amplification = 1/8). The instrument was calibrated using latex spheres of 2-03(im diameter (Coulter Electronics Ltd, Luton, Beds.), suspended in electrolyte.

Counting and sizing of cells The above apparatus was used, except that the probe had an aperture diameter of 140/im. Cells were counted in the growth medium, diluted where necessary with fresh medium, so that the coincidence correction did not exceed 10 % of the cells counted (approx. 14 500) with a manometer volume of 0-5 ml. Settings were as above, except that the matching constant was 20 K and 1 /amplification was 4. Background counts of unfiltered medium were usually about 200. Calibration was with latex spheres of 19'0/im diameter (Coulter Electronics Ltd), suspended in growth medium. Cell counts performed microscopically in a Rafter cell (Gallenkamp, London) were in good agreement with Coulter counts.

Enzyme assays These were performed as described by Poole et al. (19716) using a Pye-Unicam SP 1700 spectrophotometer. Antimycin A was dissolved in methanol at lmgml" 1 and added to the test cuvette in the assay of NADH-cytochrome c oxidoreductase (EC 1.6.99.3), where indicated, to a final concentration of 5jUgml~'. Appropriate controls showed that methanol alone was without effect.

Cytochromes Difference spectra (dithionite-reduced minus H^C^-oxidized) were recorded with a Pye-Unicam SP 1700 spectrophotometer, fitted with an attachment allowing measurements to be made at 77 K (Salmon & Poole, 1980). The band width was l-0nm and the scanning speed 1 nms~'. The extinction coefficients used for the a-absorption bands of cytochromes b and c were those measured using cytochromes isolated from this organism (Yamanaka, Nagata & Okunuki, 1968). The value of 16mM~'cm~' (Chance & Williams, 1956) for mammalian cytochrome oxidase (605-630 nm) was used to measure the Tetrahymena cytochrome oxidase at 617-650 nm. The peak heights of the a-absorption bands were intensified approximately 10-fold at 77 K relative to room temperature.

Protein This was measured using the method of Lowry, Rosebrough, Farr & Randall (1951).

Light scattering Mitochondrial fractions were diluted into the electrolyte used for Coulter counting to an £520 nm (lcm light-path) of about 0-5 (= 0-2-0-3 mg protein ml" 1 ). Turbidity (measured against an electrolyte blank) was continuously monitored using a Pye-Unicam SP 1700 spectrophotometer. Additions of salts were as described in Results.

Electron microscopy Mitochondrial preparations were fixed in 1-5% (w/v) glutaraldehyde in 0-35in-sucrose for 60 min, then rinsed in similar buffer and pre-fixed in 1 % (w/v) OsCU , also made up in this buffer.

440

R. K. Poole

Dehydration by passage through graded water/ethanol mixtures was followed by embedding in Spurr's resin. Sections (50-60 nm) were examined in an AEI EM6B instrument at 60 kV.

Chemicals Oligomycin, cytochrome c (type III, horse heart), jS-NADH (yeast, grade III, disodium salt), 104 phosphatase substrate tablets, cw-oxalacetic acid (grade I) and Ficoll (type 400) were all from Sigma. Antimycin A was from Boehringer and Triton X-100 from Koch-Light. Glutaraldehyde was from EMscope (13 Bedford Road, London SW4 7SH). All other reagents were of the highest purity available and were from Fisons Scientific Apparatus or B.D.H.

RESULTS

Performance of the apparatus The addition to electrolyte of increasing volumes of a suspension of mitochondria (prepared by differential centrifugation) resulted in increased particle counts, which, after coincidence corrections, were linearly proportional to the volume of suspension added. Increasing volumes of suspension also resulted in a slight increase in the apparent mean volume of the particles (coincidence error) (Fig. 1). Dilutions of all fractions used subsequently were such that the number of particles counted were in the range 7-8 (X104) (equivalent to 0-12-0-6/ig protein ml" 1 ; see later). At this level, coincidence corrections were about 5 % of the observed count and mean particle volume fluctuated by only 10 % for a 50 % change in the number of particles counted. Background counts were less than 1 -5 % of the counts of mitochondria from normal cells, but rose to 10% for the smaller mitochondria from chloramphenicol-treated cells (see below). The background signals corresponded to particles of volume 0-06-0-08/im3; further filtration of the electrolyte was without effect. Thus, the limits of detection were at least as good as those of Gear & Bednarek (1972; 0-06 ;Um3), and significantly better than those of Gebicki & Hunter (1964; 0-4-0-6;Um3) or Glas & Bahr (1966; 0-2-0-3 ^im3). Resolution of the instrument was assessed by using monodisperse latex spheres (2-0 /zm diam.). Measured volume distributions were unimodal, almost symmetrical about the mode, and the half-width of the distribution at half the peak height (Kubitschek, 1969) was 10-4% of the distribution mean (results not shown). Purification of mitochondria by density gradient centrifugation It was important to define conditions for the preparation of mitochondrial fractions as free as possible from contaminating organelles. In isopycnic subcellular fractionations of Tetrahymena homogenates, mitochondria are heavily contaminated with acid phosphatase-containing particles and band at densities very close to peroxisomes (Lloyd et al. 1971). In rate separations, mitochondria sediment faster than the activities of catalase and acid phosphatase (Poole et al. 19716). Table 1 shows that rate separation of a Pi fraction on sucrose gradients achieves a considerable enhancement of specific activity of malate dehydrogenase, which is 85 % mitochondrial in location in T. pyriformis (Miiller, Hogg & de Duve, 1968). The increases in specific activities of catalase and, especially, acid phosphatase were much

Quantification of microbial mitochondria

441

18

15

IS

12 |

3

CO

a "o 6

E n 1-1 "«

CO

3

E

o

"a. - 0-9 I o

rtici

1 t

9