Localization of phosphofructokinase on the mitochondria of Tetrahymena pyriformis.

Abstract At least three-fourths of the phosphofructokinase activity in homogenates of Tetrahymena pyriformis is localized on the mitochondria. The mitochondrial phosphofructokinase activity is stabilized by ATP and by fructose 6-phosphate and is inhibited by ATP and by citrate.

natant was decanted, the cells were resuspended in about 5 ml of Buffer A containing 0.5 mM ATP and 0.5 mM dithiothreitol. A cell count was taken and the remaining cells were then broken by mechanical homogenization or by sonication. Mechanical homogenization was accomplished in a Potter-Elvehjem grinder with a Teflon pestle rotating at about 100 rpm; over 907, of the cells were broken after 100 strokes. Sonication treatment consisted of two 30-s exposures in a model  at a setting of 7. The disrupted cell preparations were centrifuged at 12,000 X g for 20 min and the pellet was resuspended in a small volume of the Buffer A with ATP and dithiothreitol.
The volumes of the supernatant and of the resuspended pellet were measured.
Zonal centrifugation was carried out in a TiXIV rotor equipped with a 29 liner essentially as described by Porter et al. (7). Four hundred milliliters of a continuous 10 to 50% (w/w) sucrose gradient in Buffer A plus 0.5 mM ATP and 0.5 mM dithiothreitol were loaded in from the edge, and rested on a cushion of 55y0 sucrose in Buffer A. The sample volume was 15 ml and was overlaid with about 90 ml of a 1: 1 dilution of Buffer A with water.
Centrifugation was at 5000 rpm for 16 min. Fractions of about 11 ml were collected from the outer edge.
Phosphofructokinase was assayed essentially as described by Ho and Anderson (8). The assay mixture contained, in 1 .O ml: 0.5 pmole of ATP, 0.25 pmole of NADH, 0.5 pmole of WIgSOr, 1 pmole of fructose 6-phosphate, 20 pmoles of glycylglycine (pH 8.0), 0.2 unit of aldolase, 12.3 units of triosephosphate isomerase, and 1.4 units of a-glycerophosphate dehydrogenase. The reaction was started by adding the solution to be assayed for phosphofructokinase.
For each assay a control with fructose 6-phosphate omitted was performed.
The coupling enzymes were dialyzed overnight against 20 mM glycylglycine at pH 8.0 prior to use. It was ascertained at the end of each day that an excess of each coupling enzyme was present.
The reaction rate was computed from the change in absorbance at 340 nm, using cells with a l-cm light path in a thermostatted cell compartment maintained at 26". One unit of phosphofructokinase activity is equal to 1 kmole of fructose 1 ,6-diphosphate produced per hour. Activity was expressed as units per 10G ceils or units per ml of fraction.
Hexokinase was assayed according to the method of Risse and Blum (4) except that glycylglycine buffer was used instead 7445 7446 of Tris-HCl.
Glucose 6-phosphate isomerase activity was meas-  Table I I  I  I  I  I  I  I  ,  I  I  homogenizer. Most of the phosphofructokinase activity was I 2 3 4 5 HOURS "soluble" (i.e. stayed in solution after centrifugation for 20 min at 12,000 x g) in cells disrupted by ultrasound, but three-fourths of the phosphofructokinase activity was localized in the crude mitochondrial fraction.
Washing the crude mitochondrial fraction by centrifugation and resuspension in fresh buffer with or without 0.04% Triton X-100 did not solubilize an appreciable amount of the act,ivity.
Centrifugation of the supernatant of the sonicated aliquot at 100,000 X g for 1 hour in a Beckman model L ultracentrifuge did not lead to the sedimentation of any activity; by this criterion, sonication leads to a true solubilization of the enzyme without any loss of activity.
The particulate phosphofructokinase activity decayed rapidly after homogenization.
Since it is well known (9) that this enzyme from other species is stabilized by its substrates, ATP and fructose 6-phosphate were added to the homogenate.
As shown in Fig. 1, 1 mM fructose B-phosphate reduced the rate of loss of activity while 0.5 mM ATE' almost prevented loss of activity of particles kept for over 5 hours at pH 8.0. In all subsequent esperimeuts ATl' was added to the washed cell suspension just prior to homogenization.
It should be noted, however, that even ill the presence of 0.5 mM ATP some activity is frequently lost. It was also observed that the greater the dilution of the crude mitorhondrial preparation, the greater the tendency for loss of activity.
I,ocalizafion oJ Phosphofructokinase Activity-To det.ermine the nature of the particles containing the phosphofructokinase activity, particles obtained from a homogenate were resolved by zonal gradient centrifugation as described in Fig. 2. Isocitrate lyase alId cntalase served as perosisomal markers; lactate and glutamate dehydrogeuases served as mitochondrial markers (7). It can be seen that an almost complete separation between mitochontlria and perosisomes was achieved, and that the phosphofructokinasc activity was localized on the mitochondria. The It should also be noted that only half of the total phosphofructokinase activity put onto the gradient was recovered, presumably due to inactivation resulting from dilution as mentioned above. Some Properties of Xitochondrial Phosphojructokinase-In a freshly prepared crude mitochondrial pellet phosphofructokinase activity is low at pH about 6 and increases to a fairly flat maximum as the pH is increased to pH 8 (Fig. 3). Although the presence of 0.5 m&t ATP in the homogenization buffer appears to protect the activity measured at pH 8, this may not be true of the activity measured below pH 7. Two different kinds of results were obtained with time in ice after homogenization. In Fig. 3A the shape of the activity versus pII curve appears to change so that with time after homogenization the activity decreases more at low pIl than at high PH.
In Fig. 3B, however, the activity decreased with time more or less uniformly in the pII range studied. Increasing the ATP concentration in the assay medium to 10 mi\z caused a marked inhibition of activity, but the highest activity still occurred at about pH 8. Mitochondrial phosphofructokinase was inhibited by citrate (Fig. 4). The enzyme was much more sensitive to inhibition by citrate at ~1-1 6 than at pH 8. In experiments not shown, it was established that both orthophosphate and ammonium sulfate enhance the activity of this enzyme.
There was no effect of cyclic 3': 5'-AMI' on the activity of the mitochondrial phosphofructokinasc whether it was added to the homogenization medium or to the assay medium.
Several experiments were done in which the sensitivity of the supernatant phosphofructokinase to pH, citrate, and ATP was compared to the sensitivity of the mitochondrially localized enzyme. These very preliminary experiments did not indicate any major differences between the two forms of phosphofructokinase, and the subject was not pursued further.
It was found that in homogenates of Tetrahymena prepared in a Potter-Elvehjem grinder at pH 7.9 most of the hexokinase ac- KHzPO~ (20 mM) was-added to all assay mixtures and the pH adjusted to 6.0 or 8.0. One hundred ner cent activitv d (measured immediately after homogenization) corresponds to 1.40 and 1.10 PFK units per lo6 cells for the studies at pH 8.0 and pH 6.0, respectively. At several times throughout each experiment control assays (at pH 8.0) were performed. All assays were corrected for the loss of activity with time at pH 8.0, on the assumption that the loss of activity at pH G.0 was proportional to the loss at pH 8.0. tivity is localized on the mitochondria, whereas in homogenates prepared at pH 6.9 most of the activity was soluble (4). The effect of pH of the homogenization buffer on localization of phosphofructokinase was therefore studied.
No change in the distribution of phosphofructokinase was observed when a homogenate made at pH 7.0 or 7.5 was compared with one made at pH 8.0. The possibility that phosphohesose isomerase was present on the mitochondria was tested for by adding glucose and some more ATP to the assay mixture for phosphofructokinase, but leaving out the fructose B-phosphate.
It was ascertained that the hesokinase was active under these conditions and that if fructose 6-phosphate was added, the phosphofructokinase was active.
In the absence of added fructose B-phosphate, however, 7445 no oxidation of NADH was observed, indicating that phosphohexose isomerase was not present. Several experiments were done in which the cells were grown in medium supplemented with 6 mM acetate or 12 mM glucose, or were grown under conditions of partial anaerobiosis, or were grown to stationary phase. In no case did we find a large and reproducible change in total amount of phosphofructokinase activity nor was there any reproducible change in per cent of distribution between the mitochondrial pellet and the supernatant.
The amount of activity varied from 10 to 25 nmoles per min per mg of cell protein, comparable to the values obtained by Warnock and Van Eys (6). DISCUSSION It is generally found that phosphofructokinase in many species is a soluble enzyme (9), although Mansour et al. (10) noted that this enzyme was localized on an insoluble fraction of sheep heart homogenates.
In so far as we have been able to ascertain, however, Tetrahymena is the first cell for which a mitochondrial localization of this glycolytic enzyme has been established. Hexokinase is relatively loosely bound to the mitochondria of Tetrahymenu; changing the pH from 7.9 to 6.9 leads to an almost complete solubilization of this enzyme. Lactate dehydrogenase, also localized on the mitochondria of this cell, is at the opposite extreme, and remains particulate even after sonication.
Phosphofructokinase appears to be intermediate. It is entirely solubilized by sonication but remains on mitochondria prepared by mechanical homogenization even after washing with 0.04% Triton X-100.
It is therefore very unlikely that the phosphofructokinase was localized in the cytosol and was adsorbed onto the mitochondria because of changes brought about by the mechanical homogenization procedure.
We have also noticed that the distribution is very sensitive to the particular set of Potter-Elvehjem homogenizers used. The maximum yield we have obtained for the mitochondrial form is about 80% of the total, but, in view of the sensitivity to mechanical disruption, it is possible that almost all of the phosphofructokinase of Tetruhymenu is localized on the mitochondria in viva. This possibility is further strengthened by our failure to note any differences in kinetic behavior between the enzyme in the supernatant of homogenates as compared to the pellet.
Regardless of the nature of the non-mitochondrially bound enzyme, the fact that at least 75% of this enzyme, most of the hexokinase and lactate dehydrogenase (4, 7) and probably over half of the glyceralde-hyde 3-phosphate dehydrogenase (5) activities are on the mitochondria raise important questions as to the structure of gluconeogenesis and glycolysis in this cell. It is possible that with gentler methods of disruption, other glycolytic enzymes will also be found to be mitochondrial in this cell and, indeed, available evidence suggests that Tetruhymenu is not unique in this respect (1).
The properties of the mitochondrial phosphofructokinase of Tetruhymenu in general conform to those of this enzyme from other sources (9) : (a) it is inhibited at pH 8.0 and more strongly at pH 6.0 by citrate and by ATP; (b) it is relatively unstable but can be stabilized by the presence of fructose g-phosphate and, to a greater extent, by ATP; (c) activity is enhanced by (NH&S04 and by orthophosphate; (d) its stability decreases with increasing dilution.
Unlike the enzyme from liver flukes, however (9), no effect of cyclic 3':5'-AMP on activity was observed. The similarity of the effect of pH, citrate, ATP, (NH&SO+ and orthophosphate on the activity of the mitochondrial phosphofructokinase of Tetruhymenu to the effect of these modifiers on this enzyme from other sources suggests that the enzyme is comparable to the other enzymes in most of its kinetic properties and differs primarily in its localization.
Further kinetic studies on this enzyme as well as localization studies on other cells are indicated.
Aclinowledgments-We are grateful to Mrs. Carolyn Edwards and Mr. Alvernon Hayes for technical help in some of these experiments.