Skip to main content

Full text of "Partial purification and characterization of DCCD-sensitive ATPase from pea cotyledon mitochondria"

See other formats

For Reference 


0X 11BB15 


Digitized by the Internet Archive 
in 2020 with funding from 
University of Alberta Libraries 




TITLE OF THESIS Partial Purification and Characterization of DCCD- 

-Sensitive ATPase from Pea Cotyledon Mitochondria 

Permission is hereby granted to THE UNIVERSITY OF ALBERTA 
LIBRARY to reproduce single copies of this thesis and to lend or sell 
such copies for private, scholarly or scientific research purposes only. 

The author reserves other publication rights, and neither the thesis 
nor extensive extracts from it may be printed or otherwise reproduced 
without the author's written permission. , 


Partial Purification and Characterization of DCCD-Sensitive ATPase from Pea 

Cotyledon Mitochondria 

Mark Burnett WHISSON 



OF Master of Science 

Plant Biochemistry 

Department of Plant Science 

Spring, 1981 




The undersigned certify that they have read, and recommend to 
the Faculty of Graduate Studies and Research, for acceptance, a thesis entitled 
Partial Purification and Characterization of DCCD-Sensitive ATPase from Pea 
Cotyledon Mitochondria submitted by Mark Burnett WHISSON in partial fulfilment 
of the requirements for the degree of Master of Science in Plant Biochemistry. 


The DCCD-sensitive ATPase of pea (Pisum sativum L.) cotyledon 
mitochondria was solubilized from submitochondrial particle (SMP) membranes. The 
solubilized ATPase (FoFI) was partially purified and characterized. 

To release FoFI from SMP membranes, the detergents sodium cholate, 
Triton X-100, and sodium deoxycholate were investigated. The latter two 
detergents were found to either interfere with the ATPase assay, or deactivate 
the pea FoFI. Sodium cholate was thus chosen for the solubilization of the 
enzyme. The parameters of FoFI solubilization were varied, and it was 
concluded that 0.6 mg cholate/mg protein in 10% saturated ammonium sulphate 
resulted in solubilization of 85% of the total ATPase activity. Sensitivity of the 
enzyme to DCCD was maintained. Incubation of SMP in the detergent medium 
decreased the sensitivity of the solubilized enzyme to DCCD, and its specific 

Purification of the crude enzyme preparation (CE) by affinity 
chromatography with Affigel Blue was attempted, but this procedure caused a 
50% inactivation of the FoFI, and no active FoFI could be released from the 
Affigel Blue column. Ammonium sulphate precipitation did result in an increase in 
specific activity. At between 38% and 45% saturated ammonium sulphate, 20% 
of the ATPase activity was precipitated, with a specific activity 4 to 5 times 
higher than that of the CE. The precipitate, designated NSE ATPase, was highly 
sensitive to DCCD, and had a specific activity 12 to 13 times higher than the 
specific activity of pea SMP. 

Methods for further purification of NSE were investigated with CE. 
Ammonium sulphate precipitation chromatography of CE was attempted, and 
although a separation of FoFI from other proteins was observed, over 90% of 
the ATPase activity was lost. A separation of CE from other proteins was also 
obtained by sucrose density gradient centrifugation, but over 80% of the ATPase 
activity was lost. It was thought that the deactivation may have been a result of 
the long centrifugation times that are required for sucrose density gradients 
because of their high viscosity. Accordingly, CE was separated by centrifugation 


through Percoll density gradients, which have low viscosities even at high 

densities. It was found that the Percoll density gradients that resulted in poorer 
separations maintained much more ATPase activity than Percoll gradients that 
gave better separations. 

When the Percoll density gradient technique was used on NSE, or when 

the ATPase peak from a Percoll density gradient was fractionated with 
ammonium sulphate, all of the ATPase activity was destroyed. It was suggested 
that the deactivation may be caused by removal of one or more factors that 
are necessary to either the function or the stability of the pea FoFI, such as 

phospholipids. Crude soybean phospholipids stimulated NSE only slightly however. 

The properties of the NSE preparation were investigated. Preliminary 
experiments revealed that NSE contained levels of cytochrome and NADH 

dehydrogenase contamination comparable to those of the highly purified FoFI 
preparations in the literature. The pea FoFI was found to break down during 

polyacrylamide gel isoelectric focusing. Preliminary examination by SDS 
electrophoresis also suggested that the NSE may contain as few impurities as 

the highly purified mitochondrial FoFI preparations in the literature. 

It was found that in most respects, the pea FoFI had properties similar 
to pea SMP, rather than pea FI. These included specificity for nucleotide 

triphosphates, anion effects, stimulation of the ATPase activity by aging at 25 
C, inhibition of ATPase activity by DCCD, and biphasic kinetic properies. It was 
suggested that the biphasic kinetics displayed by pea FoFI may have 

physiological significance. It seemed that Fo modified many properties of FI by 
binding to it. 

In other respects (cold lability and cation specificity), the FoFI was 
intermediate between pea SMP and purified pea FI. It was suggested that the 
pea FoFI preparation may be in equilibrium with free FI and Fo. 




Thirty years ago Lehninger and coworkers proved that the chemical 

energy of Krebs cycle intermediates is conserved by the coupling of electron 
flow along the respiratory chain to oxygen, with oxidative phosphorylation. The 
subsequent development of three competing hypotheses (Racker, 1977) in the 

following 15 years spurred an ever increasing volume of research into the 

mechanism by which the energy of electron flow through the electron transport 

chain is coupled to the endothermic phosphorylation of ADP. In the early 1960s 
Racker and his collegues discovered a coupling factor (FI), that when released 
from mitochondrial membranes displayed ATPase activity. Coupling factor 
preparations were able to restore oxidative phosphorylation capabilities to 

coupling factor-depleted membranes and were able to hydrolyse ATP, but could 
not catalyse ATP synthesis or any of the exchange reactions that are associated 
with ATP synthesis in mitochondria (Pedersen, 1975). During the late 1960s 
other workers isolated coupling factors remarkably similar to FI from a 
tremendous variety of sources, including rat liver mitochondria, yeast 

mitochondria, bacteria, and even chloroplasts (Kagawa et a!, 1979). 

According to Mitchell's chemiosmotic hypothesis (Mitchell, 1961, 1966) the 
membrane protein (Fo) that anchors FI to the membrane also functions as a 
proton pore and perhaps gate. The membrane itself is a barrier that allows the 
buildup of a proton gradient that the hypothesis states drives ATP synthesis on 

FI during proton leakage back through the ATPase (FoFI). It is now generally 
accepted (Boyer et at, 1977) that FoFI embedded in a closed membrane that 
enables the formation of a proton gradient (FI side alkaline) is the minimal 
requirement to observe ATP synthesis and many of the associated exchange 
reactions. Consequently submitochondrial particles (SMP), subbacterial particles 
(SBP) and subchloroplast particles (SCP) have become the standard tools for 

investigating ATP synthesis and exchange reactions, while FI, which has been 
highly purified from a variety of sources in the last 10 years, has become the 
most common tool for investigating the ATP hydrolysis function (Pedersen, 



However, many properties of FI (kinetics, specificity, nucleotide binding) 

are dissimilar to the membrane bound enzyme (Nelson, 1976, Pedersen, 1975), 

and it has even been suggested that the ATPase reaction may not in fact be 

the reverse of the ATP synthetase reaction (Penefsky, 1974b). On the other 
hand, data obtained from SMP, SBP, or SCP are rarely conclusive because of 
the presence of unknown lipids, proteins, and carbohydrates. Consequently, many 
investigators have attempted to purify the FoFI complex in a soluble form (see 
Chapter 1). However, since "mitochondrial ATPase is perhaps the most complex 
enzyme system known to man" (Pedersen, 1975), this approach has met with 

limited success until recently, because it is a very labile system. 

Although chloroplasts are a popular source of FI and more recently of 
FoFI, plant mitochondria have received very little attention in this regard, 
perhaps because of the low yields of FI obtained. To the author's knowledge, 
only two groups have purified FI from plant mitochondria (Yoshida and 
Takeuchi, 1970 and Malhotra and Spencer, 1974, Grubmeyer and Spencer, 

1979), and no one has yet investigated a solubilization of FoFI from this 
ubiquitous and important source. 

Since FI from peas has been shown to have properties that differ in 

important points to those of animal mitochondrial FI (Grubmeyer and Spencer, 
1978), it was concluded that a FoFI preparation from plant mitochondria may 
eventually yield valuable insights into the mechanism of FoFI function. Thus the 
aim of this research project was to solubilize the FoFI of pea mitochondria. 
The FoFI extract was then to be purified if possible, and at least partially 




I wish to thank my supervisor. Dr. Mary Spencer, for her encouragement 
and helpful criticism throughout this work, and also for her help in obtaining 

the financial assistance of the Department of Plant Science. 

Special thanks are due to Dr. Dara Melanson and Dr. Arnost Horak for 

their discussions and interest during this research. Dr. Melanson also assisted in 

experimentation with the sucrose step gradients. I would also like to thank Anne 
Johnson for help in developing SMP2, Dr. Ken Eastwell for his guidance in the 

fine art of polyacrylamide gel isoelectric focusing, and Valeta Gregg for her 

invaluable enthusiasm and assistance while trying to obtain data from SDS 

polyacrylamide gel electrophoresis. 

I am also grateful to Dara for her help in typing this thesis into the 

computer, and to Dave Liverman for the use of his graphing program. 


Table of Contents 

Chapter Page 


1.1 Reviews .1 

1.2 FoFI Solubilization .2 

1.3 Properties of the Isolated FoFI .7 

1.3.1 Purity .7 

1.3.2 Structure-Function Relationships of FoFI .8 

1.3.3 Catalytic Properties of FoFI: ATPase Activity .9 

1.3.4 Catalytic Properties of FoFI: Energy Linked Reactions .11 


2.1 General .13 

2.2 Tissue .13 

2.3 Preparation of Mitochondria .13 

2.4 Preparation of SMP .14 

2.5 Preparation of Cholate Extract (CE) .15 

2.6 Column Chromatography with Affigel Blue .15 

2.7 Ammonium Sulphate Precipitation .16 

2.8 Ammonium Sulphate Precipitation Chromatography .16 

2.9 Sucrose Density Gradient Centrifugation .16 

2.10 Percoll Density Gradient Centrifugation .17 

2.11 Polyacrylamide Gel Isoelectric Focusing .17 

2.12 SDS Polyacryamide Gel Electrophoresis .18 

2.13 ATPase Assays .19 

2.14 ATPase Assay with Regeneration of ATP .20 

2.15 Protein Assay .21 

2.16 Cytochrome Determination .21 

2.17 Assay for NADH dehydrogenase . 22 


3.1 Preliminary Testing of Detergents .23 

3.1.1 Effects of Detergents on the ATPase Assay .23 



3.1.2 Effects of Detergents on the ATPase 

3.2 Preliminary Detergent Extractions .29 

3.3 Optimization of the Cholate Extraction Procedure .32 

3.4 Chromatography of CE on Affigel-Blue .35 

3.5 Ammonium Sulphate Precipitation .38 

3.6 Ammonium Sulphate Precipitation Gel Chromatography .40 

3.7 Sucrose Density Gradient Centrifugation .43 

3.8 Percoll Density Gradient Centrifugation .46 

3.9 Further Purification .48 


4.1 Purity of NSE .53 

4.1.1 Enzymatic Assays for Impurities .53 

4.1.2 Gel Separations .55 

4.2 Specificity .60 

4.2.1 Cations .60 

4.2.2 Nucleotide Triphosphates .60 

4.3 pH Optimum .62 

4.4 Kinetic Properties of FoFI .62 

4.5 Anion Effects .71 

4.6 Effects of Temperature on the Stability of CE and NSE .73 

4.6.1 CE .73 

4.6.2 NSE .78 


5.1 Similarities to Pea SMP and Pea FI .82 

5.2 Similarities to Other FoFI Preparations .83 

5.3 Future Work .84 



5.4 Development of New Pi Assay .94 


5.5 Development of Tissue Growth and Organelle Isolation Procedures .99 



List of Tables 

Table Page 

3.1 Affigel Blue Chromatography of CE .37 

3.2 Ammonium Sulphate Precipitation of CE .39 

3.3 Percoll Density Gradient Centrifugation of CE .47 

3.4 Further Purification of NSE .49 

4.1 Recovery of Respiratory Components during Isolation of NSE .54 

4.2 Specificity of NSE and CE .61 

4.3 Anion Effects on ATPase activity of NSE .72 

5.1 Development of Tissue Growth and Organelle Isolation Procedures .100 


List of Figures 

Figure Page 

3.1 Effects of Detergents on the ATPase Assay .25 

3.2 Effects of Detergents on the Apparent Activity of the ATPase .27 

3.3 Rate of Precipitation of SMP during Centrifugation in an Airfuge 

A100 Rotor .31 

3.4 Optimization of the Cholate Extraction Procedure .34 

3.5 Ammonium Sulphate Precipitation Chromatography . 42 

3.6 Sucrose Density Gradient Centrifugation .45 

3.7 Representative Examples of Percoll Density Gradients .51 

4.1 Polyacrylamide Gel Isoelectric Focusing of Ammonium Sulphate 

Fractions of CE .57 

4.2 pH Profile of NSE ATPase .64 

4.3 ATP Kinetics of CE .66 

4.4 ATP Kinetics of NSE .68 

4.5 Storage Stability of CE .75 

4.6 Aging of CE at 30 C .77 

4.7 Aging of NSE .80 

5.1 Colour Development by the Reduced phosphomolybdate Complex .96 

5.2 Standard Curve for the New Pi Assay .98 


List of Abbreviations 








FoF 1 

ATPase unit 



adenosine triphosphatase 

carbonyl cyanide m-chlorophenyl hydrazine 

cholate extract of SMP 

N,N’-dicyclohexyl carbodiimide 

membrane portion of proton-ATPase of SBP,SCP or 

proton-ATPase of SBP,SCP or SMP 

soluble portion of proton-ATPase of SBP,SCP or SMP 


Michaelis constant 

2(N-morpholino)ethane sulphonic acid 

reduced nicotinamide adenine dinucleotide 

ammonium sulphate enzyme 


optical density 


proton motive force 

inorganic phosphate 

subbacterial particles 

subchloroplast particles 

sodium dodecyl sulphate 

submitochondrial particles 

N,N,N',N'-tetramethylethylene diamine 

N-tris (hydroxymethyl}-methyl-2-amino ethane sulphonic 


1 umole of Pi released/min 



1.1 Reviews 

There is today a very large body of literature on the coupling ATPase, 
derived in the main from work with SMP, SBP, SCP, and FI preparations from 
mitochondria, bacteria, and chloroplasts. Instead of repeating material in the many 
excellent reviews that are available on FI and membrane-ATPase preparations, it 
is the intention of the author to refer the reader to a few of the better ones 
and then summarize the literature, first, on the solubilization of FoFI, and 

second, on the enzymatic properties of FoFI. 

For FI from chloroplasts, the reader is referred to Baird and Hammes 

(1979) and Nelson (1976, 1977). For FI from mitochondria, or all sources, one 
should read Senior (1973) and Pedersen (1975), both of which are very 
comprehensive. Kozlov and Skulachev (1977) have also published an excellent 
review, in which they attempt to reconstruct the mechanism of ATPase action. 
The review of Harris (1978) summarizes the hundreds of nucleotide binding 

experiments that have been reported, and their possible significance to ATPase 
mechanism models. Many of these reviews contain sections on the membrane 
bound ATPase, i.e. of SMP, SBP, and SCP, since their properties are often quite 

different from those of FI. Membrane bound ATPases are reviewed by Baird 

and Hammes (1979), Harris (1978), Kozlov and Skulachev (1977), and Nelson 


There are available two comprehensive reviews (Kagawa, 1978 and 
Kagawa et ah 1979) on the structural and functional aspects of the FoFI 

complex. Fillingame's (1979) review contains a large section on the subunits of 
FoFI and the function of Fo. Pedersen (1975) includes small sections on the 

properties of the isolated FoFI. 



1.2 FoFI Solubilization 

In 1966, a preparation designated Fo was isolated by sonication and 
enzyme digestion from beef heart SMP (Kagawa and Racker). Fo could be 
combined with FI preparations to confer sensitivity to oligomycin on them, as 
well as other properties previously only found in SMP, SBP, or SCP. However, 
the specific activity of the combined preparations was low and the purity 

The first recorded detergent extraction of an FoFI preparation is that of 
Tzagoloff et at (1968). The procedure was fairly lengthy but produced a 
preparation free of all respiratory components aside from some cytochromes 
and flavins. The preparation had a high specific activity of 5-8 umoles Pi 
released/min/mg protein and was 90% inhibited by oligomycin. 

It was not until 1971, however, that two groups (Kagawa and Racker, 
and Tzagoloff and Meagher) isolated FoFI preparations (from SMP), that could 
be reconstituted with phospholipids into vesicles. These vesicles could then 
catalyze ATP-Pi exchange, an important reaction that SMP, SBP, and SCP but no 
previous enzyme preparations were capable of catalyzing. Both methods required 
only two or three steps although the Tzagoloff and Meagher preparation was 
not highly sensitive to energy transfer inhibitors and the Kagawa and Racker 
preparation was greatly stimulated by the addition of other factors, such as FI. 
The preparation from Tzagoloff and Meagher (1971) was subsequently modified 
to purify a very active FoFI from yeast SMP (Ryrie, 1975a). The principle 
modifications consisted of the addition of ATP during isolation to stabilize the 
complex, and the addition of a molecular sieving step that removed high 
molecular weight contaminants. 

Swanljung and Frigeri (1972) published an isolation technique that used as 
its major purification step affinity chromatography. SMP were incubated in 0.3% 
Triton X-100 and centrifuged at high speed to remove remaining vesicles. The 
supernatant layer was passed through a column to which purified ATPase 
inhibitor had been covalently bound. The subsequently eluted enzyme complex 
had a very high specific activity, but because of problems in binding the 
enzyme to the column and later releasing it, the yield was only 20% or so. In 


view of suggestions that the binding of the ATPase inhibitor to the enzyme 
may be controlled by membrane energization (Van de Stadt et at, 1973) this 
result is not surprising. Furthermore, the purified enzyme complex was not 
inhibited by oligomycin, except at very high concentrations. 

Later workers have also used affinity chromatography, but with ADP 
rather than the ATPase inhibitor bound to the support (Brodie et al, 1979 and 
Higashi et al, 1975). Although the binding and release of the enzyme to the 
ADP-column can be easily manipulated, resulting in higher yield, there are other 
membrane proteins that also bind ADP and may therefore bind to the column, 
eg. the adenine nucleotide antiporter. A further purification step in the form of 
a sucrose density gradient, perhaps modified from the glycerol gradient of 
Tzagoloff and Meagher (1971), was thus added. 

Carmeli and Racker (1973) reported the first solubilization of FoFI from 
chloroplasts. This was acheived by incubating chloroplasts with 2% cholate and 
0.4 M ammonium sulphate and spinning down the remaining vesicles and SCP at 
high speed. When dialysed, this crude extract formed vesicles that catalysed an 
uncoupler- and DCCD- sensitive ATP-Pi exchange When the crude extract 
was fractionated by ammonium sulphate precipitation, it gave an ATPase with an 
apparent molecular weight lower than that of the ATPase of the crude extract. 
However, the ATP-Pi exchange catalysed' by the vesicles after removal of 
ammonium sulphate by dialysis was not sensitive to uncouplers or DCCD. They 
also experimented with molecular sieving but although an ATPase peak was 
found, no purification resulted. Later authors (Clarke and Morris, 1976) also tried 
to purify detergent FoF 1-extracts by molecular sieving with only slightly more 
success (4 x purification). 

Bragg and Hou (1976) were the first workers to report a large 
purification by the use of molecular sieves. After chromatography of a cholate 
extract of E. coli membranes on Sepharose 6B followed by chromatography on 
Sepharose 4B, a 16 fold purification was obtained. However, Hare (1975) (see 
below) obtained a higher specific activity of the FoFI preparation from the 
same source and a higher purification (20-25 fold) with one spin through a 
sucrose density gradient. The former preparation contained 4 more polypeptide 



bands (as shown by SDS electrophoresis) than the latter, but was stimulated 

more by added phospoiipids. It was stated that the former FoFI preparation 

contained no respiratory chain components although this was not shown. 

An "ATP-Pi Exchangease" was prepared (Sadler et at, 1974) from SMP 

by disintegration of the particles with lysolecithin and removal of large 

membrane fragments by a short high speed centrifugation. Because the exchange 

reaction is believed to require closed vesicles (Boyer et at, 1977) and because 

the "exchangease" itself was spun down at 100,000 g (the speed usually used 
to rid preparations of vesicles and large membrane fragments), the lysolecithin 
treatment probably did not even solubilize the enzyme although it did reduce 
contamination by many electron transport components. The procedure resulted in 
a four fold increase in ATPase activity. 

Hatefi et al (1974) reported a preliminary preparation, derived from that 
of Tzagoloff et a / (1968), of complex V from beef heart SMP, Complex V has 
very similar properties to FoFI and is most probably the same enzyme (see 

Stiggall et at, 1978). Hatefi's preparation had a lower specific activity, but was 
able to catalyse an ATP specific ATP-Pi exchange without addition of or 
reconstitution with any other factors. This method was later made very complex 
by the addition of even more detergent extraction and salting in and out steps 
(Stiggall et at, 1978). The resulting complex V had a high specific activity (8-10 

umoles Pi released/min/mg protein), low contamination by respiratory components 
and a variable high ATP-Pi exchange activity that was inversely related to the 
concentration of cholate used in the second detergent extraction. The cholate 
concentration also caused the amount of phospholipid to vary from 5-13 ugm 
phosphorous/ mg protein. If it was less than 7 ugm/mg, the ATP-Pi exchange 
had an absolute requirement for added phospholipid. If it was greater than 7 
ugm/mg, the preparation catalysed ATP-Pi exchange, although the exchange was 
always stimulated by phospholipid added up to concentrations of 100 ugm/mg 

Jackl and Sebald (1975) attempted to use immunoprecipitation as the 
principle purification step after solubilization of the membrane with Triton 
X-100. Although the yield and the specific activity were not reported, they did 



find 9 proteins other than the 5 normally attributed to FI. This is 5-7 more 
than are found by SDS electrophoresis of the highly purified preparations. 

Probably the most important contribution to the field of FoFI research 
to date was made by Sone et at, (1975). After much searching, they chose as 
their source material a thermophilic bacterium, PS3. The FI and FoFI ATPases 
from PS3 are highly stable, even under conditions that cause rapid dissociation 
or denaturation of ATPases from other sources, e.g. high temperature or 4 M 
urea. Consequently, Kagawa and his group were able to develop for the PS3 
FoFI a fairly long procedure that utilized principally ion exchange 

chromatography, to which many other FoFIs appear to be sensitive. This gave a 
15 fold purification (18 umoles Pi released/min/mg protein) with a 40% yield 
and 75% inhibition by DCCD. SDS electrophoresis revealed only 8 bands, i.e. 5 
from FI and 3 from Fo. The fact that the source was available in large 
quantities, coupled with the high yield procedure and the stability of the 

enzyme, has allowed the completion of many elegant experiments by Kagawa's 
team since 1975, including reconstitution of active FoFI entirely from individual 
subunits, electron density mapping of crystallized FI and so on. Their 

discoveries will be discussed later in this chapter. 

Soper and Pedersen (1976) investigated solubilization of FoFI from rat 
liver SMP. They concluded that deoxycholate was superior to Triton X-100 and 
other detergents in terms of specific activity and sensitivity to DCCD. They did 
not attempt to purify the deoxycholate extract any further. 

The procedure of Carmeli and Packer (1973) was further refined by 
Serrano et al (1976), principally by careful optimization of the cholate extraction 
conditions and by addition of a sucrose density gradient centrifugation. The 

resulting mitochondrial FoFI preparation had a specific activity of 15 umoles Pi 
released/min/mg protein (which was dependent on added phospholipids), 0-3% 
respiratory components, and some adenine nucleotide transporter contamination. It 
could be reconstituted with phospholipid into vesicles with a very high ATP-Pi 
exchange and the ability to function as reversible ATP driven proton pumps. 
This was the first purified FoFI from mitochondria that was shown to have 
this capability. It showed 1 1 bands on SDS gel electrophoresis, 3 of which 


were very faint and which were concluded to be trace contaminants. This 
procedure still produces perhaps the best FoFI available from mitochondria. 

Recently Berden and Voorn-Brouwer (1978) have reported the isolation 
of an FoFI from beef heart SMP with properties very similar to those of the 
FoFI of Serrano et a! (1976). It had a similar specific activity, even though it 
showed 4 more bands on SDS gels than the FoFI of Serrano et al (1976). 
When reconstituted into vesicles, the preparation also catalyzed ATP driven 
proton translocation. The notable point about the method of Berden and 
Voorn-Brouwer is that it is appreciably shorter and gives a very high yield 
(88% compared to 30% for that of Serrano et a / (1976), which itself is high in 
comparison to that obtained by many other methods). After removal of Triton 

X-100 by dialysis the procedure consisted merely of fractionation with 

ammonium sulphate in cholate. 

With the exception of the two reports mentioned above, since 1976 
investigators appear to have been concentrating on purifying FoFI preparations 
from different sources by slightly modifying fractionation and purification 
techniques that had been previously published. 

The technique of Carmeli and Racker (1973) was barely altered by 

Winget et at (1977) to extract and partially purify an FoFI from chloroplasts 
that could be put into synthetic phospholipid vesicles with bacteriorhodopsin. 
These vesicles were able to catalyse ATP-Pi exchange and light driven ATP 
synthesis. Pick and Racker (1979) modified the procedure of Serrano et a / 

(1976) to obtain a highly purified FoFI from chloroplasts. 

The Tzagoloff et a / (1968) procedure was somewhat streamlined and 

adapted to extraction and partial purification of FoFI from E. coli membranes 
by Freidl et al (1977). In 1979, Freidl et al used a novel detergent (Aminoxid 
WS35) and a procedure based on that of Sone et al (1975) to highly purify 
the FoFI of E. co/i. The ATPase activity was increased 21 fold and the 
preparation showed only 8 bands after SDS gel electrophoresis. 

Another FoFI was purified from E. co/i membranes by Foster and 
Fillingame (1979), who used a partially modified method of Hare (1975) with 
only the addition of an ammonium sulphate precipitation step. This precipitate 



also showed 8 subunits after SDS gel electrophoresis and had a specific 

activity of 16-18 umoles Pi released/min/mg protein. 

Oren and Gromet-Elhanan (1977) used a simple Triton X-100 extraction 

coupled with a glycerol gradient to solubilize a very low activity FoFI from the 
nonsulphur purple bacterium Rhodospi ri / / um rubrum. Schneider et a / (1980) 

purified a considerably higher specific activity FoFI from the same bacterium by 
adapting the method of Freidl et a / (1979). Even so, it still showed 12 or so 
bands after SDS electrophoresis and quite low activity (2 umoles Pi released/- 

min/mg protein) when compared with other purified preparations. 

The mitochondria of the mould, AspergiIlus nidulans, were used as a 

source by Marahiei et at, (1977). The purification procedure was based on that 
of Tzagoloff and Meagher (1971) and resulted in a 6 fold purification. 

1.3 Properties of the Isolated FoFI 

1.3.1 Purity 

The first and most important question that should be asked of a new 
enzyme purification is, of course, how pure is the enzyme preparation that it 
produces 7 This question has normally been addressed with direct assays of 
FoFI preparations for specific impurities (eg. spectophotometric assays for 
cytochromes or enzymatic assays for NADH dehydrogenase) and/or SDS 
polyacrylamide gel electrophoresis of FoF 1 preparations. 

Kagawa and Racker (1971) found in their FoFI preparation no 
cytochrome a + a3, 79% of the mitochondrial cytochrome b in SMP, 37% of 
the cytochrome c + cl in SMP, and 20% of the mitochondrial phospholipid in 
SMP (i.e. 0.64 umol Pi/mg protein). The preparation of Tzagoloff et at (1968) 
had a considerably higher specific activity than that of Kagawa and Racker 
(1971) (10 fold higher) and not surprisingly, lower impurity levels. It also had no 
cytochrome a, 8% of the mitochondrial amount of cytochrome b, and only 4% 
of the cytochrome c + cl. 

The more recent highly purified preparations have specific activities over 
2 times higher than that of Tzagoloff and Meagher (1971). Analyses commonly 



reveal 0 or less than 0.5% cytochrome a (3% in the case of Sone et a / 

(1975)), 2% NADH dehydrogenase, 2% cytochrome b, 1-3% cytochrome c + cl, 
and 5% or less phospholipid, relative to the source membranes (Sone et a / 
1975, Serrano et al 1976, Stiggall et al 1978, Brodie et a! 1979). 

Most workers, however, have judged the purity of their preparation by 
the specific activity and by the number of bands it shows after 

SDS-electrophoresis. Early preparations often showed 12 or more bands, e.g. 
Sadler et al (1974). In fact, some later preparations have shown even more 
bands, e.g. 16 (Marahiel et al 1977, Bragg and Hou 1976), even some of the 

highly purified FoFIs (Stiggall et al , 1978). 

However, Capaldi (1973) compared FoFI made by the methods of 
Tzagoloff et a! (1968) and Kagawa and Racker (1966) with SDS-gel 

elecrophoresis. He also fractionated the preparations by repeated NaBr 
precipitation and ran the fractions and the FoFI reconstituted from these 

fractions on SDS gels. He concluded that as well as the 5 FI subunits, there 

were 4 subunits in mitochondrial FoFI. This conclusion has been strengthened by 
the results from recent highly purified FoFIs (Serrano et al, 1976, Pick and 
Racker, 1979), although other workers (Sone et al, 1975, Freidl et al, 1979, 

Schneider and Altendorf, 1980) have purified fully functional FoFI from bacteria 
with only 8 subunits. There is even at present some controversy over whether 
or not the third Fo subunit is necessary for bacterial Fo function (Sone et al, 
1978, Fillingame et al, 1980). 

1.3.2 Structure-Function Relationships of FoFI 

The FoFI complex catalyses the last step in the processes of oxidative 
and photosynthetic phosphorylation. Therefore its function is to provide a 

channel that protons are able to travel through as they move down the 

protonmotive gradient, and to use the energy released by these protons to 

drive the endergonic synthesis of ATP. In prokaryotes it probably also uses the 
energy of ATP hydrolysis to form a protonmotive gradient, that is then used 

used for active uptake of substrates (Kaback, 1976, Boyer et al, 1977). 


Reconstitution has been a powerful tool in the investigation of 
structure-function relationships. Some of the reviews mentioned at the beginning 
of this chapter detail the use of reconstitution to elucidate the function of 
each subunit of FI. The reviews of Kagawa (1978), Kagawa et al (1979), and 
Fillingame (1980), detail the same for the FoFI, principally of the thermophilic 
bacterium PS3. They concluded that the D and E subunits bind the rest of FI 

to Fo, which itself is composed probably of 3 different subunits, although only 

2 appear to be necessary to reconstitute ATP-Pi exchange and proton 

translocation (Sone et al, 1978). The number (eg. Foster and Fillingame, 1979, 
Freidl et at, 1979) and even the molecular weight of the subunits of Fo 

(Schneider and Altendorf, 1980) have been confirmed from eukaryotic sources, 
but it should be noted that other workers have shown that Fo from higher 

sources, beef heart SMP (Alfonzo and Racker, 1979) and chloroplasts (Pick and 
Racker, 1979), may well contain 4 different types of subunits. Although it 

seems unlikely from an evolutionary point of view (Hasan and Rosen, 1979), the 
Fo from higher eukaryotic sources may differ from that of PS3, from which 
the vast majority of our knowledge to date has come. 

1.3.3 Catalytic Properties of FoFI: ATPase Activity 

ATPase activity is the most convenient way to monitor purification of the 
enzyme, and when used in conjunction with inhibition by DCCD or similar 

inhibitor, is very specific for FoFI. Aside from the maximum rate of ATP 

hydrolysis measurable, however, few workers have investigated this activity in 
FoFI preparations, which is rather surprising in view of the amount of literature 
on the same activity in SMP and purified FI. 

Soper and Pedersen (1976) found that deoxycholate-solubilized FoFI from 
beef heart showed biphasic ATP kinetics, with Kms of 29 and 313 uM. Other 
workers have found only linear Michaelis-Menten kinetic plots with Kms from 
13 uM (Schneider et al, 1980) to 300 uM (similar to the membrane bound 
enzymeKOren and Gromet-Elhanan, 1977) for FoFI from a nonsulphur purple 
bacterium, and from 140 uM (Swanljung et al, 1973) to 160 uM (Stiggall et al, 
1978) for beef heart FoFI. At present, therefore, the general consensus would 


appear to indicate a Km of 10-300 uM, although one should keep in mind that 
many variables can alter the Km drastically. For example, the use of Ca++ in 
place of Mg++ as the cation raised the Km by 400% (Schneider et a /, 1980) 
and activation with cardiolipin instead of lysolecithin increased the Km by 500% 
(Swanljung et al, 1973). It may also vary with the type of detergent used, the 
amount of membrane still present, and so on. However, the Km of FoFI is 
roughly one order of magnitude lower than that of purified FI, and almost the 
same as that for membrane bound ATPase (i.e. SMP, SBP, etc.) (Pedersen, 1975). 

Carmeli and Racker (1973) and Swanljung et al (1973) have both done 
pH profiles of ATPase activity of chloroplast and mitochondrial FoFI respectively 
, and found the optimum pH to be 8.0, with or without added phospholipids, 

the same as purified FI and membrane bound ATPase. 

Only three groups have so far investigated the nucleotide specificity for 

hydrolysis by FoFI. Ryrie (1975) reported that yeast FoFI hydrolysed ITP, ATP, 
and GTP at almost the same rates. UTP was hydrolysed at 20% of the previous 
rates, and other nucleotides (TTP, CTP, and ADP) were hydrolysed at less than 
5% of this rate. Stiggall et a / (1978) found that GTP and ITP were hydrolysed 

at 66% and 44% respectively, of the rate of ATP hydrolysis. UTP was not 
hydrolysed. Tzagoloff et al (1968) found that with beef heart FoFI the rate of 
hydrolysis of GTP and ITP was only 30% of the rate of ATP hydrolysis. Thus, 
SMP from yeast and beef heart show specificities similar to the FoFI 

preparation from beef heart by Tzagoloff et al (1968) while purified FI 

characteristically shows specificities similar to those found by Ryrie (1975). The 
most highly purified FoFI preparation (Stiggall et al, 1978) falls in between the 


With both beef heart FoFI (Tzagoloff et at, 1968) and yeast FoFI (Ryrie, 
1975), Ca++ has been found to be relatively ineffective in combining with ATP 
to form a substrate complex, compared to Mg++ (and Mn++). It is similar in 
this respect to the membrane bound and F1 forms of the enzyme (Pedersen, 

1975). Pick and Racker (1979) found that Ca++ was almost as effective as 
Mg++ in forming a substrate complex for chloroplast FoFI. Chloroplast FoFI 
would thus appear to be intermediate between membrane bound and soluble F1 


(Nelson, 1976). 

1.3.4 Catalytic Properties of FoFI: Energy Linked Reactions 

The presence of, or the potential for reconstitution of ATP-Pi exchange 
has been used as a marker to indicate the capability for energy-linked 

reactions, which of course, are essential features of the in vivo enzyme that 
the well studied FI apparently lacks. Early workers (eg. Sadler et at, 1974) 
purified "exchangease particles” that were able to catalyse ATP-Pi exchange 
without the addition of extra factors or phospholipids, it was soon realized, 
however, that if extracted and purified gently enough, purified FoFI could 
catalyse ATP-Pi exchange merely by reconstitution with phospholipids into 

vesicles in accordance with the predictions of Mitchell (1961, 1966). Since then, 
exchange rates from 2.3 nmol Pi/min/mg protein (Schneider et at, 1980) to 410 
nmol Pi/min/mg protein (Stiggall et at , 1978) for FoFI reconstituted into vesicles 
have been recorded. Most exchange rates fell between 25 nmol Pi/min/mg 
protein and 200 nmol Pi/min/mg protein, with the higher rates catalysed by the 
more highly purified preparations (e.g. Ryrie, 1975), irrespective of the source 
or detergents used. Carmeli and Racker (1973) found that the optimum pH of 
the exchange was 8.0 or 8.5, much the same as the optimum for ATPase 
activity. Stiggall et a / (1978), Pick and Racker (1979), Hatefi et a! (1974), and 
Carmeli and Racker (1973) have investigated the nucleotide specificity of the 
exchange and found it to be highly specific for ATP. Other nucleotide 
exchanges were less than 10% of the rate of ATP exchange, with the 
exception of the chloropiast FoFI of Pick and Racker (1979) that could 
catalyse a GTP-Pi exchange about 30% as fast as the ATP-Pi exchange. This is 
much the same as the exchange and hydrolytic specificity of the membrane 

bound enzyme. Pick and Racker (1979) also observed that the exchange was 

dependent on Mg++ and that Ca++ was ineffective. Stiggall et a / (1978) found 
that the Km (ATP) of the exchange was approximately 2.5 mM, 250 times 
higher than the Km (ATP) for the hydrolysis reaction. 

In all cases tested, the exchange was sensitive to uncouplers and DCCD. 
Workers have also demonstrated ATP driven proton translocation into vesicles 


reconstituted with purified FoFI from mitochondria, E. coli, and PS3 (Schneider 
and Altendorf, 1980, Ryrie and Blackmore, 1976, Sone et at, 1975), either by 

direct pH measurement (eg. Kagawa et at, 1973), fluorescent dye quenching (e.g. 
Freidl et at, 1979), or fluorescent dye enhancement (Sone et at, 1975). Kagawa 
et a / (1973) found that the phospholipids used in the reconstitution with 

mitochondrial FoFI must contain unsaturated fatty acid side chains. Kagawa and 
coworkers have used the more stable FoFI to investigate in detail the 

electrochemical gradient that is formed during ATP hydrolysis (Kagawa et at, 

1979), and have concluded that the gradient is larger than predicted by 
Mitchell's (1966a) postulate that 2 protons are pumped/ATP. As was the case 

with ATP-Pi exchange, ATP driven proton exchange was found to be inhibited 
by low concentrations of uncouplers, DCCD or oligomycin in all instances. 

ATP synthesis by FoFI preparations (eg. Racker and Stoeckenius, 1974) 

and by purified FoFI (see below) in vesicles has been demonstrated and shown 

to be dependent upon a correctly oriented electrochemical gradient. A sufficient 
gradient was formed by: acid to base transition (Pick and Racker, 1979), 

illumination of purified bacteriorhodopsin that was reconstituted into the 

membrane with FoFI (Winget et at, 1977), or by a chemical redox potential 

across the membrane (external ascorbate and internal ferricyanide) coupled to a 

permeable proton carrier (PMS) (Ryrie and Blackmore, 1976). Synthesis of ATP 
was insensitive to all specific electron transport inhibitors, but sensitive to 
uncouplers or the absence of any part of the gradient production system. 
Although none of these experiments utilized purified lipids for reconstitution, 
they did provide strong support for Mitchell's (1966) chemiosomotic hypothesis. 
Sone et a / (1977) did substantiate the primary role of proton translocation in 
phosphorylation by using the highly purified FoFI of Sone et a / (1975) and 
reconstituting it with a defined mixture of phospholipids from PS3. The 

phosphorylation was driven by acid-base transition and was concluded to be 40 
times faster/mg protein than phosphorylation by mitochondria, SMP, or SBP 
(Kagawa et at, 1979). 



2.1 General 

ATP was obtained from Terochem Inc., Edmonton. Sephadex G~25 and 
Percoll were from Pharmacia, Uppsala, Affigel Blue from Bio-Rad, ammonium 
sulphate (Enzyme Grade) from Serva, Heidlberg, and Ampholines from LKB, 
Bromma, Sweden. All other chemicals and biochemicals were obtained from 
Sigma or Fisher and were of the highest purity available. pH was measured at 
25 C. All other manipulations were done at 0-2 C except where stated 

2.2 Tissue 

Pea seeds (Pisum sativum L. cv Homesteader) were soaked in tap water 
for 6 hours and then planted in trays of vermiculite (horticultural grade). The 

trays were incubated at 27 C (high humidity) in the dark for 4 days. The plants 
were harvested, rinsed in tap water and the cotyledons removed by hand and 
kept on ice until use (within 2 hours). 

In early experiments various growing methods that could have decreased 

the time required for the removal of the cotyledons were tried. For example, 
peas were grown for one or two days on trays between sheets of moistened 
blotting paper at 27 C (high humidity), or suspended in constantly agitated and 
aerated vats of water at 25 C. However, all of these methods resulted in 

lower yields and/or specific activities, than did the vermiculite procedure (see 

Appendix 2). 

2.3 Preparation of Mitochondria 

Mitochondria were prepared essentially by the method of Solomos et at, 

(1972) as follows. One and a half litres of freshly prepared pea cotyledons 

(see Appendix 2) was divided into 3 lots of 500 ml, and ground with 3 

volumes of 0.5 M mannitol, 5 mM EDTA, 0.5% BSA, 0.05% cyseine, and 50 
mM TES (pH 7.4) in a large mortar and pestle. The homogenate was strained 




through 2 layers of Miracloth (Calbiochem) and centrifuged at 900 g for 8 min. 
The supernatant layer was removed and centrifuged at 19,000 g for 20 min. 
The pellet was suspended in 50 ml of 0.3 M mannitol, 0.3% BSA, and 25 mM 
TES (pH 7.2), and centrifuged at 23,000 g for 15 min. The pellet thus obtained 
was resuspended in 250 mM sucrose and 50 mM TES (pH 7.0) 

2.4 Preparation of SMP 

One batch of mitochondria, preferably fresh (see Appendix 2), was 

divided into two lots and each was diluted to 60 ml (120 ml total) with 250 

mM sucrose and 50 mM TES (pH 7.0). The diluted mitochondria were sonicated 

on ice at 90% full power with an Artek Sonic Dismembrator (model 300 large 
tip) for 2 one min bursts separated by a one minute cooling period. The 

temperature of the sonicate was thus kept below 5 C. The same beaker was 
always used to contain preparations during sonication. The sonicate was 

centrifuged at 23,000 g for 15 min. to remove unbroken mitochondria. The 
supernatant layer was centrifuged for one hour at 100,000 g and the pellet 
was resuspended in 2-4 ml of 250 mM sucrose and 50 mM TES (pH 7.0) and 
frozen at -20 C in 1 ml containers sealed with rubber bungs. For storage for 

longer than two weeks, the frozen SMP was stored in liquid nitrogen in 

unsealed tubes. 

Some batches of mitochondria were resuspended in 10 mM magnesium 
chloride, 30 mM mercaptoethanol, and 20 mM TES (pH 6.9). These mitochondria 
were sonicated for four one minute bursts with a one minute cooling period 
between each sonication. The SMP were resuspended in 0.5 mM EDTA, 1 mM 
magnesium sulphate, 0.5 mM dithiothreitol, and 10 mM TES (pH 7.5) and stored 
as above. The resulting SMP (SMP2) had a specific activity 5 to 10 times 
higher (0.3-1.0 umoles Pi released/min/mg protein) than ordinary SMP (0.05-0.1 
umoles Pi released/min/mg protein) but the total protein content was lower. This 
will be discussed further in a later chapter. 

' . 


2.5 Preparation of Cholate Extract (CE) 

ATPase was solubilized from SMP by the method of Serrano et al 

(1976), slightly modified as follows. Five ml of SMP or SMP2 (protein content 

15-25 mg/ml) were thawed and brought to 10% saturation of ammonium 
sulphate by the addition of solid ammonium sulphate. Sodium cholate (25% w/v) 

was added to give a concentration of 0.6 mg cholate /mg protein and gently 

stirred until the ammonium sulphate had dissolved and the cholate dispersed. The 
resulting dark orange brown mixture was immediately centrifuged at either 
100,000 g for one hr or 235,000 g for 35 min. The supernatant layer (CE) 

was divided into 1 ml aquilots and frozen at -20 C. If it was to be stored 

for more one week the frozen aquilots were stored in liquid nitrogen as 

described above. 

Initial experiments to determine the extraction conditions were performed 
under the conditions described in Chapter 3 in a Beckman Airfuge with an 

A100 rotor spun at 145,000 g at 25 C. 

2.6 Column Chromatography with Affigel Blue 

Affigel Blue was rinsed and degassed at room temperature in double 
distilled water. The slurry was packed into a 1 cm by 10 cm column and 
washed with five volumes of Eluting Buffer (250 mM sucrose, 1 mM 
magnesium sulphate, 1.0% sodium cholate, and 50 mM TES (pH 7.0 or pH 8.0)), 
at 0.5 ml/min. At either 2 C or 25 C, 0.35-0.70 ml of CE were loaded onto 
the column. One volume of Eluting Buffer was then pumped through, followed 
by three volumes of Eluting Buffer whose pH increased linearly from 7.0 (or 
8.0) to 11.0 (or 11.3) and one volume of pH 11 (or 11.3) buffer (the high pH 
wash). The pH 11.0 (or 11.3) buffer was identical to the Eluting Buffer except 
that it was buffered with 50 mM glycine instead of TES. The OD of the eluant 
was monitored at 280 nm with a Pharmacia Duo Optical Monitor and collected 
in one ml fractions by an LKB Ultrorac Fraction Collecter. Fractions were 
assayed for ATPase activity and protein content by the methods described 



2.7 Ammonium Sulphate Precipitation 

This procedure was modified slightly from that of Serrano et at (1976). 
Cholate extract (0.5 or 1.0 ml) was brought to 38% saturation by the addition 
of 0.45 ml of saturated ammonium sulphate/ ml CE while stirring constantly. The 
solution was immediately spun at 23,000 g for 20 min. The supernatant layer 
was carefully decanted and brought to 45% saturation by the slow addition of 
0.127 ml of saturated ammonium sulphate/ ml of beginning CE. The solution 
was stored on ice for 20-30 min with brief mixing every five min and 
subsequently spun at 23,000 g for 20 min. The supernatant layer was discarded 
and the small pellet resuspended in 50 ul of 50 mM sucrose, 0.5 mM EDTA, 
1.0 mM magnesium sulphate,0.5 mM dithiothreitol and 10 mM TES (pH 7.5). 

2.8 Ammonium Sulphate Precipitation Chromatography 

Expanded and degassed Sephadex G-25 (48 ml) was packed into a 1.5 
cm by 30 cm column and equilibrated with two volumes of 50 mM sucrose, 
0.5 mM EDTA, 1.0 mM magnesium sulphate, 0.5 mM dithiothreitol, 1.0% sodium 
cholate, 10 mM TES (pH 7.5), and 50% saturated ammonium sulphate. One and a 
half ml of CE that had been centrifuged in 38% ammonium sulphate was loaded 
onto the column, and was subsequently eluted with 50 ml of a 50% to 38% 
saturation linear ammonium sulphate gradient in equilibration buffer at a flow 
rate of 0.5 ml/min. Fractions were collected and analysed as described above. 

2.9 Sucrose Density Gradient Centrifugation 

The procedure of Serrano et al (1976) was modified as follows. CE (0.8 
ml) was layered onto 4.7 ml linear sucrose gradient, from 0.35 M to 0.70 M, 
containing 0.5 mM EDTA, 1 mM magnesium sulphate, 0.5 mM dithiothreitol, 1.0% 
sodium cholate, and 10 mM TES (pH 7.5). The gradient was centrifuged for 6 
hr at 235000 g and collected by siphoning off measured fractions. The 
fractions were assayed as described below. 




2.10 Percoll Density Gradient Centrifugation 

Two different types of Percoll density gradients were used: preformed 
linear gradients and self-generating gradients. 

Ten ml preformed gradients from 0-95% Percoll, 25-75% Percoll, and 
0-50% Percoll were made with a Buchler gradient maker and contained 0.5 mM 
EDTA, 1 mM magnesium sulphate, 0.5 mM dithiothreitol, from 0 to 3% sodium 
cholate, and 10 mM TES (pH 7.5). Half a ml of CE was layered onto the 
gradient and centrifuged at various speeds for various lengths of time (see 
Chapter 3). 

For experiments using self-generating gradients 7.5 ml of medium A (0.5 

mM EDTA, 1 mM magnesium sulphate, 0.5 mM dithiothreitol, and 10 mM TES 

(pH 7.5)) in 95% Percoll, 2.1 ml of medium A, 0.4 ml of 25% sodium cholate, 

and 0.5 ml of CE were mixed together in a tube and centrifuged at various 

speeds for various times (see Chapter 3). 

The gradients were removed by siphoning and fractioned into one ml 
volumes. ATPase activity and protein were assayed as described below. 

2.11 Polyacrylamide Gel Isoelectric Focusing 

The method of Eastwell (1980) was used as follows. Gels (7.5% 

acrylamide/ 2.5% BIS- acrylamide) containing 0.19% N,N'-methylene bis-acrylamide 
(Bio-Rad, Richmond, Calif.), 7.31% acrylamide, 5% glycerol (v/v), 1.0% (v/v) 
Ampholine pH 5/7, 1.0% (v/v) Ampholine pH 7/9, 0.1% (v/v) TEMED, and 0.5 

mg/ml ammonium persulphate were mixed and cast in 3.5 mm ID glass tubes at 
5 C after 20 min degassing. The gels were 10 cm long. 

After curing the gels for 16 hr the sample was applied as a 20% (v/v) 
glycerol solution, and overlaid with 25 ul of 1.0% (v/v) Ampholine pH 7/9, and 
5% (v/v) glycerol. The anode electrolyte was 40 mM glutamate, and the cathode 
electrolyte was 100 mM ethanolamine. The gels were run at 18.8 V/cm for 17 

hrs, then 75 V/cm for 30 min, and finally at 100 V/cm for one hr with a 

Buchler model 3-1014 power supply. Immediately after focusing, the gels were 
placed on ice to reduce diffusion. 


The gels were stained overnight for protein with Reisner's stain (Reisner 
et at, 1975), which contains 0.04% (w/v) Coomassie Brilliant Blue G-250 in 3.5% 
perchloric acid. The gels were then soaked in 7% (v/v) acetic acid for 2 hr 

and then transferred to 7% (v/v) acetic acid and 5% (v/v) methanol at 

approximately 45 C. The solution was stirred continuously for two days in a 
diffusive destainer (model 172A, Bio-Rad Laboratories, Richmond, California. The 
gels were scanned at 490 nm in a 10 cm gel boat with a Cary 219 


Gels were stained for ATPase activity for two days at room temerature 

with Horak's stain (Horak, 1972) (300 mM sucrose, 50 mM calcium chloride,5 
mM ATP (added separately), and 25 mM TES (pH 8.0)). The gels were scanned 
at 578 nm. ATPase activity was revealed by a white band of calcium phosphate 

2.12 SDS Polyacryamide Gel Electrophoresis 

Slab gels were made by the method of Laemmii (1970), as follows. A 

10 cm long separation gel was cast in a Bio-Rad model 220 Dual Vertical Slab 
Gel Electrophoresis Cell. The gel consisted of 12.5% acrylamide, 0.33% 

N, N'-methylene bis-acrylamide, 0.1% SDS, 0.025% (v/v) TEMED, 0.025% 

ammonium persulphate, and 0.375 M Tris-HCI (pH 8.8). A stacking gel 1 cm 
long was cast on top of the separation gel and consisted of 3% acrylamide, 

O. 08% N,N’-methylene bis-acrylamide, 0.1% SDS, 0.025% (v/v) TEMED, 0.025% 

ammonium persulphate, and 0.125 M Tris-HCI (pH 6.8). The electrolyte contained 
0.1% SDS, 0.192 M glycine, and 0.025 M Tris-HCI (pH 8.3). The sample, in 2% 

SDS, 10% (v/v) glycerol, 5% (v/v) 2-mercaptoethanol, 0.0001% Bromophenol blue, 

and 0.0625 M Tris-HCI (pH 6.8), was incubated at 100 C for 90 sec before 

application onto the gel. 

The gels were run at room temperature with a Bio-Rad model 500 
power supply at 20 mV for 1 hr and at 70 mV for 1.5 hr. Upon completion 
of the run, the slabs were immediately stained with Reisner's stain as described 




2.13 ATPase Assays 

The ATPase assay that was used in the initial stages (up to the studies 

on the effects of detergents on ATPase) of this research was that of Malhotra 
and Spencer (1974). The assay medium contained 300 mM sucrose, 3 mM 
magnesium sulphate, and 25 mM TES (pH 8.0). The reaction was started by the 

addition of SMP or enzyme, or ATP to a concentration of 3 mM and a final 

volume of 2 ml. The reaction ran for 10 min at 30 C and was quenched by 

2 ml of ice cold 120 mM glycine, 1.8 M sodium perchlorate and 0.3 N HCI. 
After incubation on ice for 6 min the tubes were centrifuged to remove 
protein, and 2 ml were removed to 2 ml of Mozersky’s (1966) molybdate 

reagent (2.1 N sulphuric acid, 600 mM sodium perchlorate, and 12.5 mM 

ammonium molybdate). Four ml of isobutanofbenzene (1:1) were added to the 

mixture, which was then stirred vigorously for 10 sec. The two phases were 

separated by centrifugation, and the presence of oxidized phosphomolybdate 
complex in the hydrocarbon phase was quantified by measurement of the OD at 

313 nm in a quartz cuvette. 

This method was subsequently modified. The reaction volume was reduced 
to 1 ml, as was the volume of quench medium. The protein precipitation step 
was omitted, and the entire 2 ml reaction/quench mixture treated as before. 

The removal of the protein precipitation step had two advantages. Firstly, 
because the volume was halved, the amounts of enzyme and substrate required 

for the same reading were halved. Secondly, the time taken to perform an 
assay was reduced as a result of removal of a pipetting step and a 

centrifugation step, and because the capacity of the centrifuge had previously 
limited the assay to only twelve tubes at a time. 

The protein (if any) that was previously precipitated, banded at the 

hydrocarbon/water interface in the last step of the assay. When using SMP or 
higher specific activity preparations, the assay was found to give identical 
results to the unmodified method of Malhotra and Spencer (1974). 

Later in the project a new assay was developed to even further reduce 
the length of the time required (see Appendix 1). The reaction conditions were 
left unchanged and the Pi detection method of Serrano et a / (1976) was used, 


modified as follows. One ml of reaction medium was quenched with 2 ml of 
0.72 N sulphuric acid and 0.7% ammonium molybdate Any resulting turbidity was 
cleared by addition of 100 ul of 10% (w/v) SDS. To develop colour, 50 ul of 
freshly prepared 1%(w/v) sodium ascorbate was added. After 10 min, the 
reduced phosphomolybdate complex was measured by recording the OD at 750 

This assay was considerably faster than either of the previous assays. 
Turbidity from protein was rarely a problem, so that after quenching the 
reaction, only one step i.e. adding and mixing in ascorbate, was required before 
reading the amount of Pi released. The assay was thus found to be more 
reproducible, and was linear to over 300 nmoles of Pi (see Appendix 1). The 
sensitivity of this assay (100 nmoles Pi=0.14 OD at 750 nm) was approximately 
one fifth of that of the previous assay (100 nmoles Pi=0.70 OD at 313 nm) 
and one third of the first assay (100 nmoles Pi=0.40 OD at 313 nm). 
However, because of its better reproducibility, and speed, it was considered 
much more useful. It was used in all of the experiments done after the cholate 
extraction procedure was optimized. 

One unit of ATPase was defined as 1 umole of Pi released/min under 
the assay conditions described in this chapter. 

2.14 ATPase Assay with Regeneration of ATP 

During kinetic experiments an enzyme trap was employed to remove ADP 
and maintain the initial ATP concentration. One ml of the assay medium (2 mM 
magnesium sulphate, 50 mM KCI, 2 mM PEP, and 25 mM TES (pH 8.0)) 
contained 50 ugm of pyruvate kinase. The enzyme was added and the mixture 
incubated at 30 C for 10 min. Then the reaction was started by the addition 
of equal volumes of 100 mM sodium ATP and 100 mM magnesium sulphate. 
After 10 or occasionally 15 min the reaction was quenched and assayed for 
released Pi with the reduced molybdate assay. 


2.15 Protein Assay 

Protein was assayed by the method of Sedmak and Grossberg (1977) as 
follows. A sample containing from 10 to 50 ugm of protein was diluted to 2 
ml with double distilled water. To this, 2 mis of protein dye containing 0.06% 
Coomassie Brilliant Blue G-250 and 3.0% (w/v) perchloric acid were added and 
after a timed wait of 5 to 10 min, the OD at 625 nm was read. 

2.16 Cytochrome Determination 

Cytochromes a, b, cl, and c were determined by the method of 

Williams (1964). To each of two 1 cm light path cuvettes were added 100 to 

200 ul of sample, 200 ul of 120 mM potassium phosphate (pH 7.4) and 200 

ul 10% sodium deoxycholate. To one cuvette 100 ul of 50 mM potassium 
ferricyanide was also added, and to the other was added 100 ul of 50 mM 

sodium ascorbate and a few grains of sodium hydrosulphite. The contents of 

both cells were mixed by inversion and a difference spectrum recorded from 

500 to 630 nm. From the curve, the optical density differences of the 

following wavelength pairs were calculated: Aa=550-535 nm, Ab=554-540 nm, 
Ac=563-577 nm, and Ad=605-630 nm. From these values the following 

mathematical expressions were calculated; 

E=(Ad+0.22B+0.482C-0.076D)/1 2.0 

From these expressions the concentration of each of the cytochromes was 
calculated as follows; 



cyt. a conc.(umol/mi)=Xd=E 

cyt. b conc.(umol/ml)=Xc=D+0.0263Xd 

cyt cl conc.(umol/ml)=Xb=C-0.0484Xd-0.225Xc 

cyt. c conc.(umol/ml)=Xa=B-0.03Xd+0.149Xc-0.491 Xb 

2.17 Assay for NADH dehydrogenase 

The NADH dehydrogenase assay that was used was modified slightly from 
that of King and Howard (1967) by reducing the reaction volume to 2.88 ml 
and increasing the concentration of the reactants accordingly. The OD of the 
reaction media (1.74 or 0.87 mM potassium ferricyanide, 41.7 mM potassium 
phosphate (pH 7.4), and 0.16 mM NADH at 30 C) was recorded at 420 nm 
against a water blank. After the non-enzymatic decrease in OD became linear 
with time, 10 to 50 ugm of enzyme was added and the absorbance decrease 
recorded for 1 min. To convert the decrease in OD 420 nm to mM NADH a 
factor of 0.5 was used. 


3.1 Preliminary Testing of Detergents 

In late 1978 when this project was begun, the three best FoFI 
solubilizations and purifications that were available were those of Sone et at 
(1975), Serrano et a / (1976), and Stiggall et a / (1978). To solubilize FoFI, the 
procedures used pre-extraction of the membranes with cholate and then 
solubilization with 2% Triton X-100, solubilization with 1.5% cholate, and 
solubilization with 0.3 mg deoxycholate/g protein respectively. Therefore, cholate, 
deoxycholate, and Triton X-100 were investigated as suitable agents for the 
solubilization of FoFI. 

3.1.1 Effects of Detergents on the ATPase Assay 

Sodium cholate was found to have little effect on the ATPase assay 
when used at concentrations of 1.0% or less (Fig 1). The detergent Triton 
X-100 caused cloudiness of the isobutanol-benzene mixture at concentrations 
greater than 0.1%, and thus increased the OD at 313 nm. Deoxycholate also 

increased the OD at 313 nm, but at all concentrations. The increases in OD 
may have been caused by a reaction of the detergent with the molybdate or 
phosphomolybdate complex. Triton X-100 may even react with 
isobutanolbenzene, or perhaps the Kraft point (Helenius and Simons, 1975) of 
Triton X-100 in isobutanol-benzene is above 25 C and thus the cloudiness was 
a crystalline suspension of Triton X-100. 

3.1.2 Effects of Detergents on the ATPase 

Increasing the concentration of cholate in the presence of a fixed 

amount of SMP resulted in a 200% increase in the OD at 313 nm (Fig. 2). 
Since, as shown in Figure 1, cholate itself does not affect the assay except at 
the higher concentrations, cholate must have stimulated the ATPase. This could 

have been achieved through breakdown of the membrane structure and thus the 
PMF, or by release of the ATPase inhibitor protein (Serrano et at, 1976). 

Previous experiments in this laboratory showed that the uncoupler CCCP, did not 



Figure 1. Effects of Detergents on the ATPase Assay 

The standard assay mixture (in 1 ml final volume) was 300 mM 
sucrose, 3 mM magnesium sulphate, and 25 mM TES (pH 8.0). One 
hundred nmoles of Pi and appropriate amounts of detergent (25% 
w/v) were added to the final concentrations shown. The assay was 
performed as described in Chapter 2. (□): sodium cholate; (o): Triton 
X-100; (a): sodium deoxycholate. 

OD at 313 nm 




Figure 2. Effects of Detergents on the Apparent Activty of the ATPase 

Assay conditions as in Figure 3 except that 100 nmoles of Pi was 

replaced by 19 ugm of protein. After incubation of the assay mixture 
at 30 C for 10 min each reaction was started by addition of ATP 
to a final concentration of 3 mM. The Pi detection assay was 

performed as described in Chapter 2. (□): sodium cholate; (o): Triton 
X-100; (a): sodium deoxycholate. 

OD at 313 nm 



stimulate the ATPase of SMP in this system (Grubmeyer, 1978). Subsequent 
experiments (see Chapter 4) demonstrated that aging caused only a 40% 
stimulation of activity in CE, whereas aging of SMP stimulated the ATPase 

activity 10 fold. Aging is believed to cause release of the ATPase inhibitor 
(Grubmeyer, 1978). This suggests that the activation of ATPase by cholate was 
a result of dissociation of the inhibitor polypeptide. As would be expected if 
this was true, the stimulation did not continue indefinitely as the concentration 

Between 0 and 0.1%, Triton X-100 caused a rapid increase in the 
apparent activity of the ATPase and from 0.1% onwards, a steady increase in 
apparent activity at about the same rate as Triton X-100 affected the assay 
itself (Fig. 2). Therefore, low concentrations of Triton X-100 also caused a 

large increase in the apparent activity of the ATPase, perhaps by the same 
mechanism as did cholate. 

In order to reconstitute energy-transducing activities, it is necessary to 
remove most of the detergent used for solubilization. It has been found that 

non-ionic detergents such as Triton X-100 are more difficult to remove by 
mild methods, eg. dialysis, than cholate and deoxycholate (Helenius and Simons, 

1975). Although reconstitution of these activities was not attempted in this 
study, it was concluded that Triton X-100 would be unsuitable for FoFI 
solubilization since it also affected the ATPase assay itself. (See below also). 

Increasing concentrations of deoxycholate progressively deactivated the 
enzyme (Fig. 2). Deactivation by deoxycholate has been reported by other 
authors (eg. Soper and Pedersen, 1976) and although the mechanism is unknown, 
it may possibly be caused by an unfavourable interaction between Fo and the 
hydrophobic environment created by deoxycholate. However, this inhibition may 
be released by removal of the deoxycholate (Tzagoloff et a!, 1968). 


3.2 Preliminary Detergent Extractions 

Detergent extraction of FoFI from membranes usually requires high 
protein concentrations (from 10-30 mg/ml). Therefore, in order to reduce the 
amount of SMP used in the experiments to determine the optimum detergent 
extraction conditions, these experiments were performed in a Beckman Airfuge 
as described in Materials and Methods. Each tube in the Airfuge rotor has a 
capacity of 150 ul and can be spun at speeds of up to 145,000 g. In control 
spins it was found that almost 90% of activity sedimented in less than 30 min 
and that a further 5% could be sedimented by 90 min of centrifugation (Fig 3). 
It was suggested (personal communication with Beckman, Inc.) that this latter 
finding may be caused by "wall effects” resulting from the small size of the 


A scaled down cholate extraction by the method of Serrano et at (1976) 
resulted in 82% of the total protein, and 93% of the total ATPase activity 
remaining in the supernatant layer after 30 min at 145,000 g. Since the activity 
was still in the supernatant layer, it was judged to be soluble (Serrano et at, 

1976). Upon addition of cholate, the ATPase activity became insensitive to 
oligomycin although sensitivity to DCCD was not affected (data not shown). 
There was no activation of the enzyme by cholate. Deoxycholate extraction of 
SMP under the conditions described by Stiggall et a / (1978) resulted in 

deactivation of 80% of the ATPase (data not shown). Furthermore, although 89% 
of the protein was solubilized, there was no detectable ATPase activity in the 
supernatant layer. It would seem, therefore, that the latter method was too 
harsh for this enzyme, while the former gave good results. 

The procedure of Sone et a / (1975), which produced the most highly 
purified FoFI preparation, involved pre-extraction of the membranes with 1% 
cholate and 4.5% saturated ammonium sulphate and subsequent solubilization of 
the ATPase with Triton X-100. Since the preextraction conditions are almost the 
same as those used successfully in the paragraph above, and because Triton 

X-100 affected the assay itself (Fig. 1) and is difficult to remove (Helenius and 
Simons, 1975), solubilization with FoFI by Triton X-100 was not pursued 





Figure 3. Rate of Precipitation of SMP during Centrifugation in Airfuge A100 

One hundred and fifty ul of SMP (23 mg/ml) were centifuged at 
145,000 g for the time indicated at 25 C. Protein and ATPase 
activity were determined as described in Chapter 2. (□): % of total 
ATPase activity present after centrifugation that was in the pellet; (O): 

% of total protein in the pellet. 

in Pellet 



The cholate extraction procedure was thus the most promising technique. 

3.3 Optimization of the Cholate Extraction Procedure 

Since the FoFI of peas may have properties different to those of the 
FoF 1 of beef heart, the parameters of cholate extraction were varied to 

determine the optimum extraction conditions for pea SMP. Serrano et a! (1976) 
incubated beef heart SMP in cholate and 10% saturated ammonium sulphate for 
7 min before centrifugation down of insoluble protein. It was found that 

incubation of pea SMP in the same solution reduced the specific activity of the 

ATPase, the yield of ATPase in the supernatant layer, and especially the 

sensitivity of the enzyme to DCCD (Fig. 4). Incubation thus appeared to allow 

the detergent to damage the ATPase. 

Since the cholate molecules must bind to the protein and lipid molecules 
to "cover" the hydrophobic regions, it is not the concentration of cholate that 

is important, but the ratio of cholate to protein. When less than 0.6mg of 

cholate/mg of protein was used, the percentage and the specific activity of 

FoFI solubilized were significantly reduced (Fig. 4). Higher ratios of 

cholate:protein lower the specific activity and DCCD sensitivity of FoFI (Fig. 4), 
both of which are obviously undesirable. At these higher levels of cholate, the 

detergent molecules may displace too many phospholipids from around Fo and 
thereby interfere with the functioning of FoFI, or the detergent molecules may 
even be so numerous as to bind in positions that effectively block proton 
movement through the pore. In any case, more cholate did not solubilize more 
ATPase, and 0.6g cholate/g protein was thus concluded to be the optimal ratio. 
This was the same ratio as used by Serrano et a / (1976). 

Ionic strength is known to greatly affect the effectiveness of the bile 
salts (Tzagoloff and Penefsky, 1971) probably by changing the CMC 1 and 
micellar properties (Helenius and Simons, 1975). Serrano et at (1976) noted that 
the use of 2% saturated ammonium sulphate resulted in the solubilization of only 

20% of the ATPase. In our system, it was found that a lower concentration of 

1 critical micellar concentration 


Figure 4. Optimization of the Cholate Extraction Procedure 

Solid ammonium sulphate and sodium cholate (25% w/v) were added 
to final concentrations of 10% saturated and 0.6 mg/mg protein (or 
as otherwise specified), to 150 ul of SMP (23 mg/ml) in 250 mM 
sucrose and 50 mM TES (pH 7.0). The ammonium sulphate and cholate 
were quickly dissolved by pipetting, and the resulting translucent 
mixture was either immediately centrifuged, or incubated at 0 C for 
the indicated time interval. The mixture was centrifuged in a Beckman 
Airfuge at 145,000 g for 30 min at 25 C. Protein and ATPase 
activity were determined as described in Chapter 2. In appropriate 
assays, DCCD was added to a concentration of 0.1 mM ???? (o): % 
of total ATPase activity after centrifugation, that was in the 
supernatant layer; (□): % of total protein that was in the supernatant 
layer; ( a ); % inhibition by DCCD of the ATPase activity of the 
supernatant layer; (+): specific activity of the supernatant layer (umoles 
Pi released/min/mg protein). 


091 SET 021 SOI 06 SL 09 S V OE SI 0 

oueieujadns uj ase dld % 

3.uej.BUJ3dns uj ujapJd % 

0330 uoiiiqTMUI % 

(Sui/ujuj/paseapj id sajouin) JijpaiS 

osr set ozv sor oeo- sto* oso* s^o* oeo* siq* 

.0 15.0 30.0.00 0.30 0.60 0-94.0 9.0 

Incubation Tine (min) mg chol/mg prot % Sat NH 4 SO 


ammonium sulphate also solubilized much less ATPase, although the solubilized 
FoF 1 did have a higher sensitivity to DCCD than FoF 1 released with 10% 
ammonium sulphate (Fig. 4). When the concentration of ammonium sulphate was 
raised above 10% saturation, no release of extra FoFI was observed, and the 
sensitivity of the enzyme to DCCD was reduced. Thus the optimal ammonium 
sulphate concentration was 10% saturated, the same as used by Serrano et a / 

Interestingly, when scaled up to a 5 ml volume (see Chapter 2), the 

optimized procedure increased the total ATPase activity 2-3 fold. A similar 

effect was noted earlier (see Testing of Detergents) and by Serrano et al 
(1976) , who suggested that it may be caused by dislocation of the inhibitor 
protein. Three results supported this sugestion. First, as discussed earlier (see 
Testing of Detergents), the cholate stimulation was probably not caused by 
release of inhibition by proton motive back pressure. Secondly, the ATPase 
activity of CE was increased by only 40% (see Chapter 4) by aging, which 
dissociates the inhibitor protein. Thirdly, SMP2, which were made in the 

presence of dithiothreitol, did not show an increase in total ATPase activity 

after extraction with cholate. Dithiothreitol is commonly used to release the 

inhibitor protein of chloroplast ATPase (Nelson, 1976) and is probably 
responsible for the 5-10 fold higher specific activity of SMP2 than is found in 

3.4 Chromatography of CE on Affigel-Blue 

Probably no other technique can purify a protein as much in a single 
step, as affinity chromatography is potentially able to do. Failure of the 

technique is most often caused by poor binding of the protein to the 
immobilized ligand, or by poor release of the protein from the immobilized 
ligand. Often the ligand is inactivated during or by its immobilization. 

In order to avoid at least this last pitfall, affinity chromatography with 
Affigel-Blue was attempted. Affigel-Blue consists of agarose beads to which 
Cibacron Blue F3GA is covalently bound. Cibacron Blue is a dye whose structure 



is similar to that of adenosine, so Affigel-Blue will bind to all enzymes that 
have a dinucleotide fold, ie. most dehydrogenases and kinases. 

High substrate/product concentrations, high pH, or high ionic strengths 

could be used to elute FoFI from the dye, once bound. ADP is a potent 
inhibitor of the ATPase (Lowe et at, 1979), and ATP and Pi would require 
removal before fractions could be assayed. At high ionic strengths, the enzyme 
appeared to be unstable (see below). Therefore, a pH gradient was used to 
release the FoFI from the dye. 

When the column was not pre-equilibrated with cholate, the enzyme was 
inactivated (Table 1) probably because the agarose support acted as a molecular 
sieve that separated the enzyme from the cholate, thus causing aggregation and 
deactivation. Previous workers have noted an increase in apparent molecular 
weight, and partial deactivation upon detergent removal without its replacement 
by phospholipids or other detergents (eg. Carmeli and Racker, 1973; Pick and 

Racker, 1979). The total deactivation that ocurred suggests that pea FoFI may 
be more fragile than FoFI from other sources. 

Pre-equilibration of the column with 1% cholate resulted in elution of 3 
major protein peaks, one of which, the peak in the void volume, was 

considerably larger than either of the remaining peaks (Table 1). All of the 
ATPase activity (50% of the amount applied) coincided with the large protein 
peak in the void volume. Since the pH optimum of FoFI is 8.0 and the activity 
drops sharply as the pH rises (see Chapter 4), it was concluded that FoF 1 
should be eluted between pH 8.0 and 1 1.3. The pH at which CE was loaded 
onto the column was raised to 8 to promote better binding of the enzyme. 
The protein peak eluted between pH 8.0 and 11.3 was considerably larger than 
that eluted with a pH 7-11 gradient, but the ATPase activity eluted was still 
found only in the void volume. To improve binding to the column even further, 
after the CE was loaded the column was incubated at 25 C for 30 min. 

Although the protein peak was even larger and squewed, no activity at all was 
eluted from the column (Table 1). The void volume was always cloudy, as if the 
previously solubilized intrinsic membrane proteins had for some reason come out 
of solution (Swanljung et at, 1973). It seemed unlikely that much FoFI had 



Table 1. Affigel Blue Chromatography of CE 

The 1 cm by 10 cm column was prepared and run as described in Chapter 2. 
All manipulations were performed at 0 to 2 C except that in run 4 0.35 ml of 
CE was loaded onto the column and incubated at 25 C for 30 min. The eluant 
was fractionated and assayed as described in Chapter 2. In all runs, 0.7 ml 
of CE containing 1.550 units of ATPase activity, was loaded onto the column, 
except in run 4, which had only 0.35 ml of CE loaded. 

Conditions ATPase activity OD^ P ea k 









pH wash 



pH wash 

1 . 

no cholate 








]% cholate, 









1% cholate. 









1% cholate, 
incubate at 

25 C 







> < 


bound to the dye but if it did it was either denatured by the high pH before 
it was released, or some factor that is necessary to maintain the active state 
of FoFI may have been removed by the column. 

3.5 Ammonium Sulphate Precipitation 

Ammonium sulphate precipitation is a gentle separation technique that has 
been used for many FoFI purifications. Most FoFIs are precipitated between 25 
and 50% saturation ammonium sulphate. Serrano et at (1976) found that 
precipitation of CE between 38 and 45% saturation ammonium sulphate gave a 
200% increase in specific activity over CE, with 43% yield. 

When this ammonium sulphate precipitation protocol was followed with 
pea CE, only 8% of the activity was recovered in the 38-45% precipitate 
fraction, but the specific activity was increased almost 5 fold (Table 2). Two 
different approaches were tried to increase the yield of FoFI without 
significantly reducing the specific activity: incubation of the solution at 45% 
saturation ammonium sulphate before precipitation, and an increase in ammonium 
sulphate concentration for the second precipitation. A wait of 45 min increased 
the yield to 20% (Table 2), which is lower than that of Serrano et a / (1976), 
but the specific activity was increased over CE by 350%, twice the increase 
obtained by Serrano et a! (1976). Further work with ammonium sulphate 
precipitations showed that after approximately 20 min the yield did not continue 
to rise, while the specific activity declined. 

These experiments suggested that the pea FoFI was not highly stable at 
0 C and high ionic strengths. This conclusion was supported by precipitation of 
the enzyme from higher ammonium sulphate concentrations. Both the 38-50% 
precipitate and the 38-55% precipitate fractions had reduced specific activities 
and reduced sensitivities to DCCD, although the yields were not increased (Table 
2). Furthermore, the 70% or so of FoFI activity that previously remained in 
solution after the 45% ammonium sulphate precipitation, was apparently destroyed 
by the higher ionic strengths. The beef heart FoFI is more stable at ammonium 
sulphate concentrations of over 55% saturation (Serrano et a!, 1976). Pick and 



Table 2. Ammonium Sulphate Precipitation of CE 

Precipitations were performed as described in Chapter 2, except for the 
following details. In experiment 1, the mixture was centrifuged immediately 
after bringing to 45% saturation of ammonium sulphate. In experiments 3 and 
4, the concentration of ammonium sulphate was brought to 50% and 55% satura¬ 
tion, respectively, instead of 45% saturation, and the mixtures were centri¬ 
fuged immediately. The specific activity of the CE was 0.105 units/mg. 


]. Serrano et al 

% sens, to DCCD 

2. wait 45 min (control) 

% sens, to DCCD 

3. to 50% sat. 

% sens, to DCCD 

4. to 55% sat. 

% sens, to DCCD 

Specific activity 

% saturation 



























% of total 
Original ATPase 

% saturation 

10-38 38-45 >45 

12 8 80 

10 20 34 

10 22 0 

% ATPase 




8 22 




Racker (1979) have, however, noted that although fractions from over 45% 
saturated ammonium sulphate had ATPase activity, they had low energy 
transducing capabilities. 

Therefore, ammonium sulphate precipitations were done by the method of 
Serrano et a / (1976) except for an incubation of 20 min in 45% saturated 
ammonium sulphate before the second spin. The 38-45% precipitate fraction 
was designated NSE (ammonium sulphate enzyme). 

3.6 Ammonium Sulphate Precipitation Gel Chromatography 

Since the ammonium sulphate precipitation step was successful, a 
modification of it was attempted to increase the purification even further. CE in 
50% ammonium sulphate was washed through a molecular sieve column that 
excluded the FoFI, but not the ammonium sulphate, with a decreasing 
ammonium sulphate gradient. At 50% ammonium sulphate, FoFI precipitates (Table 
2) and thus should deposit on the column. As the ammonium sulphate gradient 
passes by the FoFI, a lower ammonium sulphate concentration at which FoFI is 
soluble should be reached. When it redissolves, FoFI would move into the void 
volume and should repass the critical ammonium sulphate concentration at which 
it becomes soluble (since ammonium sulphate would not also move in the void 
volume) and thus reprecipitate. Other proteins in CE should have different critical 
ammonium sulphate concentrations and should therefore be separated from FoFI 
(see also Chapter 2). 

Upon elution of the ammonium sulphate gradient, two protein peaks were 
observed, one of which overlapped a peak of ATPase activity (Fig. 5). However, 
the specific activity of the peak was 0.1 umoles Pi released/min/mg protein, 
which was 10 times lower than that of the CE loaded onto the column. This 
inactivation may have been caused by many factors, such as removal of a 
factor required for stability of the enzyme, or by the high ionic strength. The 
latter, however, should not by itself cause a 10 fold inactivation (see Table 2), 
so this technique was discontinued in favour of previously published procedures, 
such as sucrose density gradient centrifugation. 



Figure 5.Ammonium Sulphate Precipitation Chromatography 

The Sephadex G-25 column was prepared, equlibrated, loaded, and 
run as described in Chapter 2. The specific activity of the CE used 
was 0.68 umoles Pi released/min/mg protein. The volume of each 
fraction was 2 ml. (□): mg protein/fraction; (o): ATPase activity per 
fraction (umoles of Pi released); ( a ); specific activity (umoles Pi 
released/min/mg protein). 

ATPase activity Curtails Pi released/fraction) 





ml eluant 

Specific Rctivity (erodes Pi released/ndn/mg 


3.7 Sucrose Density Gradient Centrifugation 

The method of Serrano et al (1976) was modified as described in 
Materials and Methods. The modification consisted of omitting lysolecithin, 
methanol and deoxycholate from the gradient. Serrano et al (1976) included 

these compounds in the gradient to prevent inactivation of the enzyme, but the 
compounds also destroyed the sensitivity of the enzyme to DCCD (Serrano et 
al, 1976). However, sensitivity to DCCD was necessary because it provided the 
only specific assay for FoFI, since no energy transformation reactions were 
measured during this research project. 

Two well separated protein peaks were obtained, the smaller of which 
had high ATPase activity (Fig. 6). The second larger protein peak probably 
contained adenine nucleotide transporter (Serrano et at, 1976). Floating on the 
meniscus of the tube was a smaller peak of ATPase activity, which possibly 

contained membrane fragments. 

Although the density gradient provided very good separation, it also 
inactivated over 80% of the ATPase that was added (data not shown). The 
specific activity of the ATPase peak was 0.48 umoles Pi released/min/mg 
protein (compared to the CE, which was 0.68 umoles Pi released/min/mg 
protein), and contained 3-4% of the activity that was loaded onto the gradient. 
The deactivation may have been caused by removal of some protective factor(s), 
as suggested before, or the enzyme may be unstable and thus be dissociated 
during the long spin. The long spin time is required because dense sucrose 
solutions are also quite viscous, and molecular movement is thus slowed by 

high sucrose concentrations. 

In a recent paper by Pick and Racker (1979), deactivation of chloroplast 

FoFI during sucrose density gradient centrifugation was eliminated by addition of 
phospholipids to the gradient. The phospholipids apparently did not alter the 
position of FoF 1 on the gradient. For future work with this enzyme, it might 
be fruitful to try this approach with sucrose density gradients, and perhaps with 
Percoll density gradients also (see below). 




Figure 6.Sucrose Density Gradient Centrifugation 

Sucrose density gradient centrifugation was performed and the 
fractions assayed as described in Chapter 2. The specific activity of 
the CE used was 0.68 umoles Pi released/min/mg protein. (□): mg 
protein/ml; (o): ATPase activity (umoles Pi released/ml); (a): specific 
activity (umoles Pi released/min/mg protein). 

Specific Activity (umoles Pi released/min/mg protein) 


bottom Vol (ml) top 

RTPase activity (umoles Pi released/ 5 mi) 


3.8 Percoll Density Gradient Centrifugation 

Percoll is a colloidal silica sol coated with polyvinylpyrrolidone to reduce 

its toxicity to biological systems. At 100% concentration (as it is supplied from 

Pharmacia), it has a density of 1.13 gm/ml, a viscosity of only 10 cP, and a 

low osmolality (less than 20 mOs/kg water). It therefore overcomes many of 
the problems normally associated with sucrose gradients by allowing separation 
of organelles by density gradients in as little as 20 min at as low as 10,000 

g. It will even spontaneously form gradients at moderate g forces and spin 

times. Although successful separation by Percoll of particles smaller than large 

viruses has not been reported, purification of FoFI on Percoll gradients was 
attempted because it was hoped that the short centrifugation times that would 
be necessary would minimize FoFI inactivation. 

It was expected that self generating Percoll gradients would give the 
best separation of FoFI from contaminants , since such a gradient should range 
from 0-100% Percoll. Table 3 shows that no activity was recovered from self 
generated Percoll gradients. FoFI may have been inactivated by the dilution, or 
as suggested above, by removal of some factor(s) that are neccessary for 
maintenance of an active conformation. The former is unlikely in view of the 
observation that when CE was applied into preformed Percoll gradients by 

pipetting through a bent pasteur pipette at the density at which the CE banded, 
the specific activity of the peak was no higher than when CE was layered on 

top of the gradient (data not shown). 

It was found that if cholate was absent from preformed gradients, no 

separation was observed (Table 3). This is in agreement with previous authors 
(Carmeli and Racker, 1973, and Helenius and Simons, 1975) who have shown 
that removal of detergent from solutions of detergent solubilized proteins, eg. 

by dialysis, results in aggregation. Cholate concentrations of 0.5 and 1.0% 

resulted in good separations with the former giving a peak of slightly higher 
specific activity. At higher concentrations, cholate deactivated the enzyme. In 
most of the successful published procedures the detergent concentration is 
reduced, typically to 0.5%, after extraction of FoFI from the membranes (eg. 
Sone et at, 1975, Foster and Fillingame, 1979). Higher detergent concentrations 



Table 3. Peroolt Density Gradient Centrifugation 

Percoll gradients were made and assayed as described in Chapter 2. The 
gradients were centrifuged at the g force indicated for the time indicated. 

Conditions are expressed as 

undi1uted Percol1. 

the concentration 

range of 

the gradient 

in % 







(g x 1000) 

% Choi ate Separation 




% of CE) 

self formed 






self formed 






self formed 






25 to 75 






25 to 75 






25 to 75 






25 to 75 






25 to 75 






25 to 75 






25 to 75 






0 to 95 






25 to 75 






0 to 50 







may alter the hydrophobic environment of Fo, or interact more directly with the 

enzyme (eg. by blockage of the proton pore, or by disruption of hydrophobic 
bonds within FoFI. 

A centrifugation step 40 min long at 30,000 g was observed to be 

sufficient to cause the FoFI to move down into the gradient (see Table 3 and 
Fig. 7). Longer and/or faster centrifugations did not improve separation and 
usually precipitated a portion of the silica sol. A 20 min centrifugation gave a 
higher specific activity peak but did not result in a separation equivalent to a 
40 min centrifugation. The 20 min centrifugation moved The ATPase only one 
cm or so into the gradient while other proteins moved no more than 4 cm 
into the gradient (Table 3). Once again it appeared that the pea FoFI was 
inactivated by either dilution, or by removal of necessary stabilization factors. 

This problem was encountered again while experimenting with different 
Percoll density gradient ranges. FoFI banded at approximately 30% Percoll on 

25-75% gradients (Fig. 7). This meant that on 0-95% and 0-50% gradients, the 

enzyme moved all the way down through a large portion of the gradient (Fig. 
7), and apparently this caused deactivation. The ATPase peaks of both of these 
gradients had a lower specific activity than either the peak on 25-75% 
gradients or the CE. 

3.9 Further Purification 

Deactivation was observed again when the 38-45% saturated ammonium 
sulphate fraction (NSE) was centrifuged through a 25-75% Percoll density 
gradient, or on fractionation with 38-45% saturated ammonium sulphate of the 
ATPase peak of a 25-75% Percoll gradient (Table 4). 

Therefore, before further work on the purification of this FoFI is 
attempted, the nature of this deactivation must be elucidated. Is it caused simply 
by dilution or by removal of stabilization factor(s), as suggested above? Recent 
authors have shown that the presence of phospholipid is essential to ATPase 
activity (Berden and Voorn-Brouwer, 1978). Asolectin (an undescribed partially 
purified phospholipid preparation from soybean) has been shown to be effective 


Table 4. Further Purification of NSE 

The ATPase activity peak (1 ml) from a 25 to 75% Percoll density gradient 
was brought to 10% saturated ammonium sulphate and centrifuged at 23,000 g 
for 40 min. The supernatant layer was fractionated by ammonium sulphate 
precipitation as described in Chapter 2. In the second treatment, 100 yl of 
NSE was layered on top of a 25 to 75% Percoll density gradient (Chapter 2) 
and centrifuged at 30,000 g for 40 min. Assays were performed as described 
in Chapter 2. 

Conditions Specific activity Specific activity % inactivation of 

before (units/ of target fraction total activity in 


after second step second step 

Percoll gradient 

then ammonium 

sulphate pre- 





Ammonium sulphate 


then Percoll 






Figure 7. Representative Examples of Percoll Density Gradients 

Percoll density gradients of the ranges specified 
centrifuged as described in Chapter 2, at 30,000 
Assays were performed as described in Chapter 2. (i 
(o): ATPase activity (nmoles Pi released/50 ul); (a): 
(umoles Pi released/min/mg protein). 

were made and 
g for 40 min. 
]): mg protein/ml; 
specific activity 

Specific flctivty tirades Pi rdeased/rairs/mg) 


RTPase Rctivity (redes Pi released/50 ul) 


in restoration of ATPase activity of phospholipid depleted preparations (Serrano 
et at, 1976). Since the pea FoFI may be quite similar in such properties as 
phospholipid content to the FoFI prepared by Serrano et at (1976), it is 

possible that the deactivation is caused by phospholipid depletion. (The FoFI 
preparation of Serrano et at (1976) contains only 0.08 umole phospholipid/mg 
protein and is stimulated 10 fold by asolectin). However, the NSE was found to 
be stimulated by only 20% by addition of 100 ugm of soybean phospholipids 
(Sigma, 40% lecithin)/mg protein. 

In any case, an increase in specific activity from 0.07 umoles Pi 

released/min/mg protein (SMP) to 0.93 umoles Pi released/min/mg protein (NSE) 
represents a 13 fold purification, barring stimulation and deactivation , which is 
a high purification compared to many other preparations. Highly purified FoFI 
from mitochondria usually is purified 5-10 fold (eg. Serrano et a!, 1976, 

Stiggall et at, 1978). Preparations of Fofl from bacteria, gg. E. co/i, when 
highly purified show a 15-20 fold purification (eg. Foster and Fillingame, 1979, 
Schneider et a/, 1980). These figures presumably reflect the greater number of 
transport and other proteins in the latter source materials. 

These results led us to forgoe further purification attempts, and to 

instead characterize the NSE preparation. 




4.1 Purity of NSE 

4.1.1 Enzymatic Assays for Impurities 

In an effort to determine the purity of the NSE, the level of NADH 
dehydrogenase activity and the amount of cytochromes present were quantified. 
However, in all but a few experiments, the assays were not sufficiently 
sensitive to give accurate data with the amount of sample that was available. 

This was particularly true of the cytochrome assay. The phospholipid assay of 
Dittmer and Wells (1969) was also experimented with, but was found to require 
too much sample to obtain readings. 

NADH dehydrogenase activity, although lower in activity in pea SMP than 
in other SMP (Serrano et at, 1976), was reduced 80% by extraction with 
cholate and fractionation with ammonium sulphate (Table 5). This compares 
favourably with the results of Serrano et a! (1976) who reported an 87% 

reduction of NADH dehydrogenase activity after detergent extraction and 
ammonium sulphate fractionation. Sone et a / (1975) showed only 0.05 umol of 

NADH were reduced per minute by one mg of their highly purified PS3 FoFI. 

Negligible amounts of cytochrome a were detected in NSE (Table 5), 

comparable to the FoFI preparations of Serrano et at (1976) and Stiggall et at 
(1978). Sone et at (1975) reported 0.02 nmoles cyt. a/mg FoFI preparation. The 
content of cytochrome b of NSE was reduced approximately 90% from that of 
SMP (Table 5). Serrano et at (1976) obtained a similar purification. However, the 
highly purified preparations of Sone et at (1975) and Stiggall et at (1978) both 
had significantly higher cytochrome b contents (0.06 nmol/mg and 0.15 nmol/mg 

respectively). The measured levels of cytochromes may be lower than the levels 
actually present because of the difficulties that were mentioned above of 

obtaining data. Cytochromes cl+c were reduced approximately 80% by the 

solubilization and ammonium sulphate fractionation procedure detailed in this 

thesis. Serrano et at (1976) obtained identical results with their procedure. The 
FoFI purified by the procedure of Stiggall et at (1978) contained approximately 
twice as much cytochromes cl+c as the pea NSE, while FoFI from PS3 (Sone 



Table 5. Recovery of Respiratory Components during Isolation of NSE 

The assays were performed as described in Chapter 2 on between 14 to 80 ygm 
of protein. (-) indicates that data were not measurable. 

Sample NADH dehydrogenase Cyt a Cyt b Cyt e Cyt cl 

(ymoles/min/mg) (n mol/mg) (n mol/mg) (n mol/mg) (n mol/mg) 

Mitochondria 0.66 
SMP 1.71 
CE 0.79 
10 to 38% sat. 

ammonium sulphate 
NSE 0.38 

0.13 0.46 

0.14 0.09 

0.00 0.03 

0.11 0.09 

0.20 0.08 

0.02 0.02 


et at, 1975) contained less than half of the contamination shown by NSE. 

In summary, the reduction of cytochrome content of NSE, by our 
procedure appeared to be very similar to that obtained by some of the best 
currently available FoFI purification procedures (Serrano et at, 1976, Sone et at, 
1975, and Stiggall et at, 1978), although the absolute cytochrome contents were 
often lower than expected. It would seem that a large proportion of the 
cytochromes were precipitated between 10 and 38% saturated ammonium 
sulphate (Table 5), with the exception of cytochrome b. This is in contrast the 
results of Kagawa and Racker (1971), which indicated that beef heart 
cytochromes c, b, and cl precipitated above 33% saturated ammonium sulphate. 
Serrano et a / (1976) found that most cytochromes were not precipitated by 
45% saturated ammonium sulphate. The contrast in results may reflect intrinsic 
differences in the properties of the cytochromes from pea and from beef 
heart, or the previously mentioned difficulties with the cytochrome assay. 

4.1.2 Gel Separations 

It was originally thought that gel isoelectric focusing would be a good 

indicator of how pure NSE was, since each contaminating protein should have a 
discrete different isoelectric point. Furthermore, the FoFI should not be entirely 
inactivated during the focusing thereby allowing staining for ATPase activity to 
be done to pinpoint FoFI. 

Approximately 16 protein peaks were observed on the gels (Fig. 8), 14 
of which appeared in the 38-45 precipitate fraction. However, when the gel 
was stained for ATPase activity, one sharp peak with two prominent shoulders 
was observed (Fig. 8). Hence, the FoFI was present on the gels in three 

slightly different active forms, and after consideration of its evident fragility 

(see Chapter 3), most probably it was also present in one or many more 

inactive forms, eg. dissociated subunits. In fact, gel isoelectric focusing over a 
different pH range produced 6 peaks in 3 major groups, that reacted with the 
ATPase stain (data not shown), one of which was insensitive to DCCD. We 
therefore concluded that gel isoelectric focusing was unsuitable for our 
purposes because we could draw no conclusions from the data that it yielded. 



Figure 8. Polyacrylamide Gel Isoelectric Focusing of Ammonium Sulphate 
Fractions of CE 

Three hundred and fifty ul of 10 to 38% saturated ammonium 
sulphate fraction, 100 ugm of NSE (38 to 45% saturated ammonium 
sulphate), or 300 ugm of protein not precipitable by 45% saturated 
ammonium sulphate were loaded onto polyacrylamide gels prepared as 
described in Chapter 2. The latter fraction was centrifuged through a 
1 ml Biogel-P column preequilibrated with 0.25 M sucrose, and 50 
mM TES (pH 7.0), to remove ammonium sulphate and cholate. The gels 
were stained for either protein or ATPase activity as described in 
Chapter 2. 




(38-45% sat.) 







10-38% sat. 


(38-45% sat.) 

>45% sat. 



There are no reports in the literature of FoFI being subjected to this 


Many authors have, however, used SDS polyacrylamide gel electrophoresis 
to determine FoFI purity (eg. Alfonzo and Racker, 1980, Clarke and Morris, 
1976, Kagawa et ai, 1978). It is now well known that FI has 5 types of 

subunits (Pedersen, 1975 and Nelson, 1976) that will be referred to here as A, 
B, C, D, and E. Subunits A and B have molecular weights in the range of 
50,000 to 60,000, depending on the source, while C, D, and E are smaller and 
constitute a small proportion of FI. In reconstitution experiments, Sone et a! 
(1978) showed that a fully functional FoFI from PS3 could be made from only 
7 types of subunits, although the conclusion that FoFI consists of only 7 types 

of subunits in vivo, has been challenged on genetic grounds by Fillingame et a / 
(1980), who worked with E. co/i. The most highly purified bacterial preparations 
of FoFI all contain 8 types of subunits (eg. Sone et ai , 1975). FoFI from 

higher sources, that is, chloroplasts (Pick and Racker, 1979) and mitochondria 
(Serrano et at, 1976, and Capaidi, 1973) appears to consist of 9 different 
subunits (therefore, 4 in Fo), although the best preparations still contain more 
than this (Serrano et a!, 1976; 12 subunits, Stiggali et ai , 1978; 13 subunits). 

Preliminary SDS gel electrophoresis of NSE and the ATPase peak from 

Percoll gradients suggested that there are 9 different subunits in NSE, and 8 in 
the gradient purified enzyme (data not shown). Their approximate molecular 
weights were as follows; 1 (72,000), 2 (58,000), 4 (44,000), 5 (36,000), 6 

(28,000), 7 (19,000), 8 (12,000), and 9 (5,000). Grubmeyer (1978) showed that 
the A and B subunits of pea FI ran together in a very prominent band at 

58,000 molecular weight on SDS electrophoresis. Band 2 thus probably contains 

the A and B subunits of FI. Grubmeyer (1978) also showed that subunits C 

and D of pea FI had apparent molecular weights of 36,000 and 22,000. Thus 
bands 5 and 7 probably correspond to subunits C and D respectively. Band 9 

was observed to be diffuse and probably contains both subunit E of the FI, 
and the smallest of the Fo subunits (the DCCD binding protein), which in 
bacteria has a molecular weight of 5,000 to 6,000 (Sone et ai, 1975). 

Alternatively, the DCCD binding protein (DBP) may be in band 8, which was also 


observed to be diffuse. Workers have found that mammalian mitochondrial DBP 
has a molecular weight of 11,000 to 12,500 (Serrano et at, 1976 and Stiggall 

et al, 1978). Serrano et a / (1976) suggested that bands of higher molecular 
weight than A and B were respiratory components, as most probably is band 1 
in our gels and the unexplained 65,000 band that is observed in SDS 
electrophoresis of pea FI (Grubmeyer, 1978). Band 4 was absent from density 

gradient purified ATPase and is, therefore, probably a contaminant. Serrano et at 
(1976) found that purification of FoFI by density gradient centrifugation reduced 
contamination by the adenine nucleotide transporter. Thus band 4 may be the 
nucleotide transporter, although its molecular weight in animal mitochondria is 

believed to be approximately 30,000 to 32,000 (Kiehl, 1980 and Serrano et al 
(1976)). Kiehl (1980) also showed that incubation of FoFI preparations with 
carboxyatractylate removed nucleotide transporter contamination. It could be 
fruitful to try this technique in future work. 

Capaldi (1973), after comparison of two beef heart FoFI preparations, 
concluded that Fo contained a 28,000 subunit, which indicates that band 6 
(28,000) may be part of FoFI. Alfonzo and Racker (1979) have since shown 

that the 28,000 subunit of beef heart Fo is necessary for Fo functioning. 
Capaldi (1973) also concluded that the oligomycin sensitivity conferring protein 
(OSCP) had a molecular weight of 19,000. If the pea mitochondrial OSCP has 
the same molecular weight, it may be in band 7 along with D. Pick and Racker 
(1979) found that a subunit that was attributed to Fo had the same apparent 
molecular weight as subunit D of chloroplast FI (ie. 17,500). 

In summary, preliminary SDS gel electrophoresis suggested that there are 
at least 9 different proteins in NSE. Two of these appeared to be contaminents 
(bands 1 and 4) and five of the rest were attributed to FI (bands 2, 3, 5, 7, 
and 9). The remaining two proteins (band 6, molecular weight=28,000 and band 
8, molecular weight= 12,000) were thought to be components of Fo. It was 
also thought that band 7 (molecular weight= 19,000) and band 9 (molecular 
weight= 5,000) may contain more than one protein each. Since bands 1 and 4 
were small (data not shown) the NSE preparation may be as pure or purer than 
the FoFI preparations of Serrano et a / (1976) and Stiggall et a / (1978). 



4.2 Specificity 

4.2.1 Cations 

Pea FoFI was able to catalyse ATP hydrolysis in the presence of either 

Mg++ or Ca++, the latter suporting slightly higher rates (Table 6). This is in 
contrast to the work of some previous authors, who have found that Ca++ is 
relatively ineffective as a cosubstrate for ATP for animal and yeast 
mitochondrial FoFI (Tzagoloff et at, 1968, Ryrie, 1975), as it is for membrane 

bound mitochondrial ATPase and purified mitochondrial FI (Pedersen, 1975). Pea 
SMP also have low Ca++-ATPase activity, but pea FI shows rates of Ca++ATP 
activity 3 fold higher than Mg++ATP activity. FoFI was thus intermediate in 
cation specificity between SMP and FI. Pick and Racker (1979) found that 
chloroplast FoFI was able to catalyse Ca++-ATP hydrolysis almost as well as 
Mg++-ATP hydrolysis, and was therefore also intermediate between thylakoid 
bound ATPase and chloroplast FI (chloroplast FI is specific for Ca++) (Nelson, 
1976). Schneider et at (1980) concluded that the cation requirements of purified 
FoFI were similar to those of membrane bound ATPase in photosynthetic 

Therefore, although Fo from animal and yeast mitochondria determines the 
cation specificity of these FIs, it would seem that the cation specificity of the 
pea FoFI is only partly determined by the Fo portion. The remainder is 
presumably determined by the membrane and/or a membrane protein(s). 

4.2.2 Nucleotide Triphosphates 

NSE was slighty more specific in its nucleotide requirements than CE 

(Table 6). Both however, had similar specificities to pea SMP, rather than pea 
FI (Grubmeyer, 1978). Other workers have found that the specificity of 
mitochondrial FoFI is either the same as that of FI (Ryrie, 1975a), intermediate 
between FI and SMP (Stiggall et al, 1978), or the same as SMP (Tzagoloff et 
at, 1968). Pedersen (1975) has advised cautious interpretation of specificity data 
reported at one NTP concentration and at one point in time. Of the above 
studies including this one, only Stiggall et al, (1978) have reported Vmax, Km 

and so forth. It is probable therefore, that the specificity of FoFI is 



Table 6 . Specificity of NSE and CE 

The standard assay mixture (in 1 ml final volume) was 300 mM 
sucrose, 3 mM magnesium sulphate or calcium sulphate, 25 mM TES 
(pH 8.0) and 3 mM NTP. From 5 to 18 ygm of protein were added 
to start the reaction. After 10 min the reaction was stopped 
and free Pi measured as described in Chapter 2. 

Conditions NTPase activity 


% of control Specific activity % of control 


Mg ++ , ATP 0.35 100 

Ca ++ , ATP 0.40 114 

Mg + \ 




Mg , 



Mg , 



Mg , 



Mg +T , 



Mg ++ , 



Mg ++ , 











intermediate between that of SMP and FI. The pea enzyme may be different to 
the beef heart FoFI, ie. be similar to the pea SMP enzyme, or it may only 
appear to be so, since Grubmeyer (1978) found that activation of pea SMP (by 
removal of the inhibitor protein) reduced the GTPase activity of the SMP. If the 
stimulation of the ATPase by cholate (Chapter 3) is caused by dissociation of 
the inhibitor protein, then one might expect a higher specificity for ATP from 
an FoFI that otherwise might be more like FI (ie. have relatively low specificity 
for ATP). It should be noted though, that dissociation of sufficient inhibitor to 
decrease GTPase by 40% resulted in a 10 fold increase in ATPase activity, and 
that stimulation by cholate typically resulted in a twofold increase in specific 

4.3 pH Optimum 

In common with pea SMP and pea FI (Grubmeyer, 1978) pea NSE 

displayed a sharp ATPase activity peak at pH 8.0 (Fig. 9). Most other ATPases 
from mitochondria, bacteria, and chloroplasts are also assayed at pH 8.0 (eg. 
Stiggall, et a / 1978, Sone et at, 1975, Carmeli and Racker, 1973), although 
Ryrie (1975a) has reported that the partially purified FoFI of yeast mitochondria 
had a pH optimum of 9.5. Previous workers with FoFI had shown pH optima 

of 8.0 (Carmeli and Racker, 1973) and 8.0-9.0 (Tzagoloff et at, 1968). It is 

therefore probable that the pH optimum of purified FoFI is the same as that 
of the other forms of the enzyme, ie. 8.0, although it may be more 

pronounced in some cases, eg. this one. 

4.4 Kinetic Properties of FoFI 

Pea CE and pea NSE were found to have biphasic Michaelis-Menten 
kinetics (Fig. 10 and 11). CE had a Vmax of 0.58 umoles Pi released/min/mg 
protein, and Kms of 41 uM and 430 uM (Fig. 10). NSE had a Vmax of 1.33 
umoles Pi released/min/mg protein and Kms of 76 uM and 480 uM (Fig. 11). 

On closer examination one may observe that each biphasic line could be 
more closely approximated by a curve. Christensen (1975) has pointed out that 

* <. ,3' .Hq i* beya>%6 vek *\* fc*a uk<i ~0 ‘ r * ’ * *' 


Figure 9. pH Profile of NSE ATPase 

The reactions were performed as described in Chapter 2 with either 
3.5 ugm or 7.0 ugm protein, except that the pH was adjusted as 
shown. At pH 6.5, 50 mM TES was replaced with 50 mM MES. At 
pH 8.5 and 9.0, 50 mM TES was replaced with 50 mM glycine. 
Separate blanks were provided at each pH. 



Figure 10. ATP Kinetics of CE 

The standard ATPase assay with regeneration, was used as described 

in Chapter 2. Seventy five ugm of CE was incubated in the medium 

at 30 C for 10 min. The reactions were begun by addition of 

MgATP to the concentrations indicated. After 10 min the reactions 

were terminated and free Pi measured as described in Chapter 2. 
Blanks were provided for each MgATP concentration. Because of the 
large range of MgATP concentrations required, duplicates of each 
concentration could not be performed. Instead, duplicates at 0.01 mM 
MgATP were done to ensure that reproducibility was within 10%. CE 

was stored at -20 C between experiments. (□): CE at day 1 plus 

values from day 4 and day 5 CE corrected to day 1 by 

proportionally increasing the specific activity at each MgATP 

concentration so that the activity at 1 mM MgATP was always the 
same; (o); Ct at day 4, uncorrected; (a): CE at day 5, uncorrected. 





S£ oe 32 02 

(UTUi/Sui/sajouin/T) a/[ 









1/RTP (1/mfl) 


Figure 11. ATP Kinetics of NSE 

Assay as Figure 10 except that the reactions were run for 20 min 
and each contained 20 ugm of protein. (□): NSE aged 1.5 hr at 0 C 
plus data from NSE aged 3.5 hr at 0 C, corrected as in Figure 10; (o 
): NSE aged 3.5 hr at 0 C, uncorrected. 


(ujui/Buj/saxoiiin/i) A /i 

1/RTP (1/mM) 


if there are two catalytically active sites on an enzyme, Lineweaver-Burke plots 
should show curves such as those in Figures b and c, and that each point on 

the curve would be a product, to varying degrees, of both sites. Therefore, in 
order to determine the Km of each site, the velocity at each substrate 

concentration must be split into its two components. This technique 

(computerized) was used by Melanson and Spencer (1981) and Grubmeyer et a / 
(1979) for analysis of kinetic data of ATP synthesis by pea SMP. The 

calculations raised the higher Km considerably, from 30 uM to 160uM, but did 

not greatly affect the lower Km, which was lowered from 9 uM to 4.8 uM. 

This treatment was not applied to the data shown in Fig. 10 and 11, partly 

because no one else has applied it to ATPase data and it thus could not be 

compared to previous work. That the Kms shown in this thesis may not be real 
should, therefore, be kept in mind by the reader. The data was also plotted in 

the form of Eadie-Hofstee plots (v/s versus v) (plots not shown) but graphical 
analysis was even more difficult. NSE from SMP2 also displayed curvilinear 
kinetics (data not shown). 

Such curvilinear plots may be caused by any one of a number of 

factors, such as negative cooperativity (Cleland, 1967), two kinetically different 

ATPase sites (Pedersen, 1976, Soper and Pedersen, 1976), or a subpopulation 

of damaged enzyme (Melanson and Spencer, 1981) caused by detergent or 
some other treatment. Since the pea FI showed curvilinear ATP kinetics when 

purified (Grubmeyer, 1978) and curvilinear ADP and ATP kinetics when membrane 
bound (Melanson and Spencer, 1981, Grubmeyer, 1978), it is unlikely that the 
curvilinear kinetics are caused by damage to some ATPases by detergent 
although Kms, etc. were different (see below). Pedersen (1976) found biphasic 
kinetics in SMP isolated by detergent fractionation, in which case damage could 

not have been caused by sonication (since there was none). Pedersen (1976) 
thus concluded "that at least two kinetically distinct classes of ATP binding sites 
must be present on (rat liver) mitochondrial ATPase...". It seems most probable 
that the same is true of pea FoFI. 

Although many workers report linear ATP kinetics for other 

proton-ATPases, curvilinear kinetics are not unknown. As mentioned above. 


Melanson and coworkers (1981 and 1979) have shown that pea membrane 
bound ATPase shows curvilinear kinetics with respect to ADP esterification, 
although the two Kms were approximately one order of magnitude lower than 

the Kms reported here. It should be noted that Swanljung et at, (1973) 
concluded that stimulatory phospholipids lowered the Km (ATP) of FoFI by over 
80%. The Kms reported here thus may be variable Grubmeyer (1978) has 
shown that pea FI displays "negative cooperativity", ie., curvilinear kinetics. The 
high Km was calculated graphically to be 170 uM (40% of that of CE and 
NSE), but the low Km was not calculated. Pea SMP also displayed curvilinear 
kinetics (Grubmeyer, 1978), although the range of ATP concentrations was 

insufficient to allow calculation of the higher Km. The lower Km was 
approximately 70 uM, which is quite similar to that of CE and NSE (Fig. 10 
and 11). This is interesting in view of the data of Kayalar et at, (1976) who 

also showed that addition of an uncoupler increased the Km of SMP with 
respect to ADP and Pi. Since the FoFI should be entirely uncoupled, it is 

evident that uncoupling does not also increase the Km (ATP). 

Soper and Pedersen (1976) also found little difference between the Kms 
(ATP) of SMP and solubilized (but not purified) FoFI. Furthermore, the Kms (29 
uM and 310 uM for the FoFI) were quite close to those of pea NSE and CE 

(Fig. 10 and 11). The remaining publications on FoFI kinetics all show linear 

ATP kinetics and considerably lower ATP affinities (22 mM, Stiggall, et a / 1978; 
300 uM, Oren and Gromet-Elhanan, 1977; and 140 - 9,000 uM, Swanljung et 
at, 1973). The absence of two ATP sites in the latter two preparations may be 
traced to the TX-100 that was used for solubilization of these complexes, 
since Soper and Pedersen (1976) found that TX-100 apparently destroyed one 
site and thus gave linear kinetics for rat liver FoFI. The FoFI of Stiggall et a / 

(1978) was only assayed at very high ATP concentrations (the lowest was 2 
mM), which accounts for the lack of biphasic kinetics (see Fig. 10 and c). 

Aging of CE at subzero temperatures for up to 5 days and aging of 
NSE on ice for up to 4 hr slowly deactivated the enzyme (Fig. 10 and 11), 

although in each case the Kms were unchanged. Lower enzyme activity indicates 
that either all enzyme molecules have been partially deactivated, or that some 



molecules have been entirely deactivated while the rest are unchanged. A 
decrease in Vmax with no concommitant decrease of the Kms would tend to 
support the latter conclusion. Some molecules may be more sensitive to such 
deactivation than others within the same population, because they may be closely 
associated with different amounts and compositions of phospholipids and other 

4.5 Anion Effects 

Previous work from this laboratory (Grubmeyer, 1978) has shown that 
the ATPase of pea mitochondria possess a unique intrinsic stimulation by 
chloride ions, in addition to the bicarbonate stimulation that both membrane 
bound ATPases and purified FI from other sources normally possess. 

Preliminary results (Table 7) indicated that NSE was stimulated by both 40 
mM NaCI (17%) and 40 mM sodium bicarbonate (43%), but much less than FI 
was (125% and 220% respectively), and slightly less than were SMP (73% by 
100 mM NaCI and 64% by 20 mM sodium bicarbonate). This provides further 
evidence that the chloride ion stimulation of pea FI may be of physiological 
significance, and in addition suggests that the Fo portion somehow exerts a 
modifying effect on these anion stimulations, possibly by restricting 
conformational changes. 

Although pea FI is stimulated 210% by 250 mM NaCI (Grubmeyer, 1978), 
pea FoFI was not (Table 7). This lack of stimulation may reflect the instability 
of the pea FoFI at higher ionic strengths (see Chapter 3). Instability at higher 
ionic strengths may also have been the reason for the lower stimulations under 
the other treatments, as compared to SMP, although one would not expect 10 
mM NaCI to accelerate enzyme dissociation. 

Calcium ATPase activity was inhibited by chloride ions as was pea FI 

(Grubmeyer, 1978). 



lable 7. Anion Effects on ATPase Activity of NSE 

Assay conditions as in Chapter 2. Each tube 
contained 10 ygm of protein, which was incubated in 
the assay medium for 10 min before initiation of the 
reaction by addition of ATP to a concentration of 
3 mM. Sodium bicarbonate or sodium chloride was 
added as shown. In experiments 6 and 7, 3 mM 
magnesium sulphate was replaced with 3 mM calcium 



Specific activity 

%o of control 



.. ++ 




Mg , 



NaHC0 3 




.. ++ 
Mg , 



NaC 1 




Mg , 







Mg , 

250 mM 





r 4"f 





Ca , 






(91% of Ca 4+ ) 



4.6 Effects of Temperature on the Stability of CE and NSE 

4.6.1 CE 

When stored in liquid nitrogen for less than a week, CE lost 

approximately 10% of its original activity (Fig. 12). This 10% loss of activity 

appeared to be intrinsic to the liquid nitrogen since repeated freezing and 
thawing at each assay did not decrease the activity any more than one freeze 

and thaw (Fig. 12). If stored for longer than 7 days in liquid nitrogen, the CE 
loses further activity. 

CE was also stored at -20 C. There was an initial 15-25% loss of 
activity (Fig. 12), but further storage of up to a week at -20 C did not 

increase the loss. As with storage in liquid nitrogen, there was no apparent 
difference in stability between samples that were thawed and refrozen between 
assays and those that were not. 

At 0 C it was observed that CE lost 86% of its original activity within 
4 days (Fig. 12). This loss was considerably more rapid than in either of the 

two treatments mentioned above, but much slower than normally observed with 
the cold labile FI. It is, therefore, probably reasonable to conclude that the CE 

is not cold labile. 

CE is more unstable than pea SMP, which can be stored for up to 2 
weeks at -20 C without loss of its most labile property- the capacity for 
oxidative phosphorylation (Grubmeyer, 1978). Other workers routinely store 

membrane extracts at -70 C (Stiggall et at, 1978, Serrano et at, 1976). 

The sensitivity of pea FoF 1 to DCCD was found to be sustained for 

more than 4 days only if stored in liquid nitrogen (data not shown). This was 
taken as an indication that the native conformation of FoFI was maintained for 
longer periods when kept at liquid nitrogen temperatures than at -20 C. 

Therefore, in this study CE was divided into 1 ml volumes and frozen in 
stoppered cellulose nitrate test tubes at -20 C. When the CE was solid, the 
stoppers were removed and the tubes and the frozen CE were stored in liquid 

During incubation at 30 C, CE showed a stimulation of ATPase activity 
(Fig 13) similar to that of SMP during aging (Grubmeyer, 1978), although the 


Figure 12. Storage Stability of CE 

CE was frozen at -20 C and stored for the indicated length of time 
in unstoppered testtubes in liquid nitrogen. Tubes containing 'bulk' CE 
were thawed, assayed, and refrozen (at -20 C and then placed in 
liquid nitrogen), whereas 'individual' CE tubes contained sufficient CE 
for one ATPase assay only. CE was stored at -20 C in stoppered 
tubes for the indicated lengths of time. 'Bulk' CE was thawed and 
refrozen for each assay, while ’individual' CE was only frozen and 
thawed once. CE was also stored in stoppered tube on ice for the 
indicated time periods. 

Assays were as described in Chapter 2. (□): 'bulk' CE in liquid 
nitrogen; (o): 'individual' CE in liquid nitrogen; (a): 'bulk' CE at -20 C; 
(+): 'individual' CE at -20 C; (x): CE at 0 C. 

fll Rase (%) 



Storage (days) 


Figure 13. Aging of CE at 30 C 

Assays as in Chapter 2. Undiluted CE was incubated at 30 C for the 
time indicated. In appropriate reactions, DCCD was added to a 
concentration of 0.1 mM. (□): ATPase activity of CE; (o): sensitivity to 
DCCD of the ATPase activity of CE. 




latter is stimulated considerably more (10 fold compared to 43% for CE). The 
stimulation of pea ATPase has been suggested to be caused by dissociation of 
an inhibitor protein (Grubmeyer, 1978). Thus, the solubilized FoFI may be 

depleted in inhibitor protein. This could be caused by the high ionic strength 
that was used during detergent solubilization, as suggested in Chapter 3. 

Alternatively, the Fo portion may be quickly dissociated at 30 C (the sensitivity 
to DCCD decreased rapidly - Fig. 13) leaving a population that was 
predominately FI. Grubmeyer (1978) showed that pea FI is not stimulated by 

4.6.2 NSE 

Some previous workers (eg. Stiggall et a!, 1978) have routinely stored 

purified FoFI at -70 C for short periods before use. One must assume that 
many other workers store only membranes or membrane extracts, and purify 

fresh FoFI for every experiment, since storage conditions for membrane 

extracts but not purified FoFI are specified (eg. Serrano et a! (1976), Foster 
and Filligame, 1979). Occasionally, left over NSE was frozen at -20 C at the 
end of an experiment. Usually 3-5 days storage resulted in a 40% reduction in 
ATPase activity (data not shown), and this, coupled with the fact that most 

experiments used almost all the NSE precipitated for that experiment, led us to 

precipitate NSE fresh from stored CE for every experiment. 

The partially purified FoFI preparation was found to be slightly stimulated 

by more than 30 min incubation at 25 C (Fig. 14). This stimulation was most 
probably also caused by protein inhibitor dissociation. In any case, numerous 

NSE preparations were assayed for detectable stimulation by incubation at 25 C, 
over the time span of the ATPase assay but stimulation was never found. Ryrie 
(1975a) also noted a stimulation of FoFI ATPase by aging, and likewise 
attributed it to release of the inhibitor protein. In contrast to its evident stability 
at room temperature, NSE was steadily deactivated at 0 C (Fig. 14). Although 
more rapid than the deactivation of CE at 0 C, it was still much slower than 
that of pea and other purified FIs, which may lose up to 90% activity in 20 
min at 0 C (Grubmeyer, 1978, Penefsky, 1974). It is possible that this gradual 



Figure 14. Aging of NSE 

Assays as in Chapter 2. NSE was aged undiluted. (□): ATPase activity 
during incubation at 25 C; (O): ATPase activity during incubation at 0 




decrease in activity of NSE is caused by a slow interchange between the FoFI 
form and the FI and Fo forms. The decrease may therefore be prevented by 
addition of p-amino benzamidine, since Freidl et at (1979) have suggested that 
this compound inhibits release of FI from E. coli membranes. However, Ryrie 
(1975a) found that yeast FoFI showed only one band during analytical 
ultracentrifugation after cold inactivation. Approximately 50% of FoFI ATPase 
activity was lost within 5-10 hrs but none was lost thereafter. 



5.1 Similarities to Pea SMP and Pea FI 

Previous workers have found that purified FI and membrane bound 

ATPase differ significantly in many properties, eg. cold lability, Km(ATP), 
nucleotide specificity (Pedersen, 1975). Consequently, some workers (eg. Soper 
and Pedersen, 1976) have used only solubilization and purification procedures 
that resulted in FoFI preparations with similar properties to the membrane 
bound ATPase. Although that approach to solubilization and purification was not 
adopted in this study, most NSE properties were similar to, or the same as 

those of pea SMP rather than pea FI (Chapters 3 and 4). These included 

specificity for nucleotide triphosphates, anion effects, stimulation of ATPase 
activity by aging at 25 C, sensitivity of ATP hydrolysis to DCCD, and kinetic 
properties. Some properties however (ie. cation specificity and cold lability), 
were concluded to be intermediate between pea SMP and pea F1 (Chapter 4). It 
was suggested that this may have been caused by spontaneous dissociation of 
some FoFI into FI and Fo. If the dissociation was reversible with the 
equilibrium lying to the FoFI side, one could expect FoFI to be intermediate in 

cold lability, but in all other properties be similar to SMP, since under the cold 

lability experimental conditions the dissociation to FI and Fo would be 

irreversible. (At 0 C, FI dissociates rapidly and thus would displace the 
equilibrium to the FI side.) This hypothesis could be tested in at least two 

ways. The cold deactivation of FoFI could be monitored until it ceased. If there 
normally was an equilibrium between FoFI and free FI and Fo, all of the FoFI 
ATPase activity would eventually be lost, although FoFI itself may be slightly 

unstable at 0 C. Alternatively, FoFI could be incubated with p-amino 

benzamidine, which is thought to stop dissociation of FoFI (Freidl et al, 1979). 

If p-amino benzamidine caused FoFI to become stable at 0 C, one could 

conclude that the usual cold deactivation was caused by spontaneous dissociation 
into FI and Fo. It is unlikely that NSE is contaminated by significant amounts of 

free FI, because all steps in the solubilization and purification of NSE were 

performed at 0-2 C (as described in Chapter 3). 




It is unclear however why the cation specificity of FoFI is intermediate 
to that of FI and SMP. Perhaps the specificity in SMP is in part determined 
by one or more membrane lipids or proteins (by restricting conformational 
changes, Grubmeyer, 1378) as well as by Fo. 

In contrast to both pea FI and pea SMP, the ATPase activity of FoFI 
has been shown to be quite labile (Chapter 3). The lability may be caused by 
removal during purification of one or more factors required for stability of the 
enzyme or maintenance of ATPase activity, such as phospholipids, as as was 
suggested in Chapter 3. Alternatively, it may be caused by dissociation of FoFI 
to Fo and F1 as discussed above. 

5.2 Similarities to Other FoFI Preparations 

Pea FoFI appeared to be more easily deactivated than FoFI from other 

sources. It was found to be partially cold labile and may have been sensitive to 

removal of one or more factors during purification that were essential to 
ATPase activity. 

In spite of this, solubilization and fractionation with ammonium sulphate 
increased the specific activity of FoFI 12 to 13 fold. DCCD was shown to be 

a potent inhibitor of NSE. Thus under the assay conditions used, free FI was 

not a significant impurity of the NSE. Very clean preparations of bacterial FoFI 

usually have a specific activity 15 to 20 fold higher than the membranes from 

which they are purified (eg. Schneider et at, 1980, Foster and Fillingame, 1979, 
Sone et al , 1975). The most pure FoFI preparations from mitochondria (Stiggall 
et al, 1978, and Serrano et al, 1976) have specific activities that are 5 to 10 

fold higher than the SMP from which the FoFI was purified. Such differences 

in degree of increase of specific activity probably reflect differences in the 
number of different proteins in vivo in each source and consequently, the 

proportion of FoFI to total protein. Plant mitochondria are very similar in most 
properties to animal mitochondria (Wiskich, 1977), so the 12 to 13 fold 

increase in specific activity observed in pea mitochondrial FoFI preparations 

probably indicates that NSE is also highly purified. 


NSE was also shown to contain low levels of NADH dehydrogenase, and 
although lack of sensitivity of the cytochrome assays precluded conclusive data, 
low levels of cytochromes a, b and cl+c were found that were comparable to 
the highly purified FoFIs of Serrano et a / (1976), Sone et a/ (1975), and 
Stiggall et a / (1978). The % reduction of cytochrome content by purification of 
pea SMP was also comparable to that of the above-mentioned purifications. 

Polyacrylamide gel electrofocusing was found to give inconclusive data 
for analysis of NSE, because of enzyme dissociation. However, preliminary 
results with SDS polyacrylamide disc electrophoresis (by the method of Laemmli, 
1970) indicated that there may be as few as 10 or 11 different proteins in 9 
bands in NSE. Two minor bands were deduced to be contaminants (bands 1 and 
4). Heavier gel loading or two dimensional gel techniques, such as were used 
by Pick and Racker (1979), may well show more contaminant bands, but should 
also allow separation of the small molecular weight proteins (discussed in 
Chapter 4). In any case, these data indicated that our FoFI preparation could 
possibly have as little contamination as the highly purified FoFI preparations in 
the literature. The preparation of Serrano et al (1976) showed 12 bands on 
SDS electrophoresis, and that of Stiggall et a / (1978) showed 13 bands. Pea 
mitochondrial Fo may even be shown to contain less than the four types of 
subunits found in animal mitochondrial Fo (Capaldi, 1973, Alfonzo and Racker, 

FoFI preparations from other sources (Stiggall et at, 1978, Tzagoloff et 
al, 1968, Ryrie, 1975 and 1975a, etc.) show properties that are much the 
same as those of membrane bound ATPase, but not FI. This indicates that Fo 
is able to modify some of the catalytic properties of FI, perhaps by restricting 
conformational changes. 

5.3 Future Work 

The first priority of future work on pea FoFI should be to elucidate the 
cause of the ATPase inactivation that has hampered further purification of the 
enzyme in the present study. As suggested earlier in this study, experiments 



with the antidissociation compound p-amino benzamidine (Freidl et at, 1979) may 

prove fruitful in this regard. The NSE may also be inhibited by a lack of 
certain phospholipids. To this end, a study on the stimulatory effects of a 
variety of types of phospholipids should be done. 

If stabilization of the FoFI could be accomplished, the enzyme could be 

further purified. The NSE contained measurable amounts of NADH dehydrogenase 
and cytochromes, and it is unlikely that one ammonium sulphate precipitation 
step could remove all other contaminations, especially since almost identical 
procedures have not done so for detergent extracts from chloroplasts (Pick and 
Racker, 1979) and animal mitochondria (Serrano et at, 1976). If stabilizing 

procedures were developed, eg. addition of phospholipids to density gradients, 
Percoll density gradient centrifugation could be most useful as a suitable final 
purification step for the pea FoFI. 

A major reason for purifying FoFI instead of just FI, is to obtain a 

tool for investigations of energy transducing reactions, ie. phosphorylation of 
ADP, ATP-Pi exchange and so on. Accordingly, NSE, whether it is purified 

further or not, should be incorporated into vesicles. There are many new 

systems available that when incorporated into vesicles, are able to produce a 
PMF suitable for driving ADP phosphorylation by FoFI, such as the purified 
Photosystem 1 reaction centre (Bengis and Nelson, 1975). 

There is a great deal still to be iearned about how the FoFI catalyses 
ATP synthesis and hydrolysis, and the FoFI incorporated into vesicles is an ideal 
tool for such studies. For example, Choate et a / (1979) have used two 
different FoFI preparations to invesigate isotopic oxygen exchanges in FoFI. 

Baird et a / (1979) experimented with singlet-singlet resonances of FoFI 
reconstituted into vesicles and concluded that FoFI protrudes 3 nm from the 
membrane surface in vivo . There are many possibilities for research with such 
a system. 

The pea FoFI has some properties that are apparently unique, such as 

the stimulation of ATP hydrolysis by chloride ions. Grubmeyer (1978) has 
hypothesized that chloride ion stimulation may indicate a type of enzyme control 
different to that of other FoFIs. Elucidation of chloride ion stimulation of 


ATP-Pi exchange or ATP synthesis, if any, may yield information on the 
mechanism of this stimulation, and thus on the mechanism of action of the 
enzyme itself. 



Alfonzo M, E Racker 1979 Components and mechanism of action of ATP-driven 
proton pumps. Can J Biochem 57: 1351-1358 

Baird BA, GG Hammes 1979 Structure of oxidative and photophosphorylation 
coupling factor complexes. Biochem Biophys Acta 549: 31-53 

Baird BA, U Pick, GG Hammes 1979 Structural investigation of reconstituted 
chloroplast ATPase with fluorescence measurements. J Biol Chem 254: 

Bengis C, N Nelson 1975 Purification and properties of the photosystem I 
reaction center from the chloroplasts. J Biol Chem 250: 2783-2788 

Berden JA, MM Voorn-Brouwer 1978 Studies on the ATPase complex from 
beef heart mitochondria. Biochem Biophys Acta 501: 424-439 

Boyer PD, B Chance, L Ernster, P Mitchell, E Racker, EC Slater 1977 Oxidative 
phosphorylation and photophosphorylation. Ann Rev Biochem 46: 955-1026 

Bragg PD, C Hou 1976 Solubilization of a phospholipid stimulated ATPase 
complex from membranes of E. co/i . Archiv Biochem Biophys 174: 

Brodie A, VK Kalra, SM Lee, NS Cohen 1979 Properties of energy transducing 
systems in different types of membrane preparations from Mycobacterium 
ph/ei - preparation, resolution and reconstitution. Methods in Enzymology LV: 

Capaldi RA 1973 On the subunit structure of oligomycin sensitive ATPase. 
Biochem Biophys Res Commun 53: 1331-1337 

Carmeli C, E Racker 1973 Partial resolution of the enzymes catalysing 
photophosphorylation. J Biol Chem 248: 8281-8287 

Choate GL, RL Hutton, PD Boyer 1979 Occurrence and significance of oxygen 
exchange reactions catalysed by mitochondrial ATPase preparations. J Biol 
Chem 254: 286-290 

Christensen HN 1975 Biological Transport 2 nd Edition. WA Benjaminlnc. Don 
Mills 514 pp 

Clarke DJ, JG Morris 1976 Partial purification of a DCC sensitive membrane 
ATPase complex from the obligately anaerobic bacterium Clostridium 
pasteurianum. Biochem J 154: 725-729 



Cleiand WW 1970 Steady state kinetics. The Enzymes II: 1-65 

Coleman JOD, JM Palmer 1972 The oxidation of malate by isolated plant 
mitochondria. Euro J Biochem 26: 499-509 

Dittmer JC, MA Wells 1969 Quantitative and qualitative analysis of lipids and 
lipid components. Methods in Enzymology 14: 482-530 

Eastwell KC 1980 The effects of ethylene on the gibberellic acid-enhanced 
synthesis and release of amylase by aleurone layers isolated from Hordeum 
vu/gare cv Himalaya. Ph D Thesis. Univ of Alberta 

Fillingame RH 1980 The proton translocating pumps of oxidative phosphorylation. 
Ann Rev Biochem 49: 1079-1113 

Fillingame RH, DL Foster, RS Negrin, ME Mosher 1980 Proton-conducting Fo 
sector of the proton-ATPase of E. co/i. J Supramol Struct Supp 4: 86 

Foster DL, RH Fillingame 1979 Energy-transducing proton-ATPase of E. coll'. 
purification, reconstitution and subunit composition. J Biol Chem 254: 

Freidl P, Bl Schmid, HU Schairer 1977 A mutant ATP synthetase of E. coli 
with an altered sensitivity to DCCD: characterization in native membranes 
and reconstituted liposomes. Euro J Biochem 73: 461-468 

Freidl P, C Freidl, HU Schairer 1979 The ATP synthetase of E. col i K12: 
purification of the enzyme and reconstitution of energy-transducing 
activities. Euro J Biochem 100: 175-180 

Grubmeyer C 1978 The energy coupling ATPase complex of pea cotyledon 
mitochondria. Ph D Thesis, Univ of Alberta 

Grubmeyer C, M Spencer 1978 Pea cotyledon mitochondrial ATPase. in Plant 
Mitochondria, Ducet G, C Lance eds. Elsevier/North-Holland Biomedical Press 

Grubmeyer C, DL Melanson, I Duncan, MS Spencer 1979 Oxidative 
phosphorylation in pea cotyledon submitochondrial particles. Plant Physiol 64: 


Grubmeyer C, M Spencer 1979 Effects of anions on a soluble ATPase from 
mitochondria of pea cotyledons. Plant & Cell Physiol 20: 83-91 

Hare JF 1975 Purification and characterization of a DCCD sensitive ATPase 
complex from membranes of E. coli. Biochem Biophys Res Comm 66: 

Harris DA 1978 The interactions of coupling ATPases with nucleotides. Biochim 
Biophys Acta 463: 245-273 



Hasan SM, BP Rosen 1979 Properties and function of the proton-translocating 
ATPase of Clostridium periringens. J Bacteriol 140: 745-747 

Hatefi Y, DL Stiggall, Y Galante, WG Hanstein 1974 Mitochondrial ATP-Pi 
exchange complex. Biochem Biophys Res Comm 61: 313-321 

Helenius A, K Simons 1975 Solubilization of membranes by detergents. Biochim 
Biophys Acta 415: 29-79 

Higashi T, VK Kalra, SH Lee, E Bogin, AF Brodie 1975 Energy-transducing 
membrane-bound coupling factor ATPase from Mycobacterium phiei. J Biol 
Chem 250: 6541-6548 

Horak A 1972 A coupling factor for photosynthetic phosphorylation in pea and 
bean plastids. Ph D Thesis, Univ of Manitoba 

Jackl G, W Sebald 1975 Identification of 2 products of mitochondrial protein 
synthesis associated with mitochondrial ATPase from Neurospora crassa. 
Euro J Biochem 54: 97-106 

Kaback HR 1976 Molecular biology and energetics of membrane transport. J Cell 
Physiol 89: 575-594 

Kagawa Y, E Racker 1966 Partial resolution of the enzymes catalysing oxidative 
phosphorylation. J Biol Chem 241: 2461-2466 

Kagawa Y, E Racker 1971 Partial resolution of the enzymes catalysing oxidative 
phosphorylation. J Biol Chem 246: 5477-5487 

Kagawa Y, A Kandrach, E Racker 1973 Partial resolution of the enzymes 
catalysing oxidative phosphorylation. J Biol Chem 248: 676-684 

Kagawa Y 1978 Reconstitution of the energy transformer, gate, and channel. 
Biochim Biophys Acta 505: 45-93 

Kagawa Y, N Sone, H Hirata, M Yoshida 1979 Structure and function of 
proton-ATPase. J Bioenerg Biomem 1 1: 39-78 

Kayalar C, J Rosing, PD Boyer 1976 2,4-DNP causes a marked increase in the 
apparent Km of Pi and of ADP for oxidative phosphorylation. Biochem 
Biophys Res Comm 72: 1153-1 159 

King TE, RL Howard 1967 Preparations and properties of soluble NADH 
dehydrogenases from cardiac muscle. Methods in Enzymology 10: 275-294 

Kozlov A, VP Skulachev 1977 Proton-adenosine triphosphatase and membrane 
energy coupling. Biochim Biophys Acta 463: 29-89 


Leammli UK 1970 Cleavage of structural proteins during assembly of the head 
of bacteriophage T4. Nature 227: 68-685 

Lowe PN, PE Linnett, H Baum, RB Beechey 1979 Mg-ATP induced inhibition of 
the enzymatic activity of chloroform released ox heart mitochondrial 
ATPase. Biochem Biophys Res Comm 91: 599-605 

Malhotra SS, M Spencer 1974 Preparation and properties of purified (Na & 
K)-stimulated mitochondrial ATPase from germinating pea seeds. Can J 
Biochem 52: 491 

Malhotra SS, M Spencer 1973 Structural development during germination of 
different populations of mitochondria from pea cotyledons. Plant Physiol 52: 

Marahiel MA, G Imam, P Nelson, NJ Pieniazek, PP Stepien, H Kuntzel 1977 
Identification of an intramitochondrially synthesized proteolipid associated 
with the mitochondrial ATPase complex as the product of a mitochondrial 
gene determining oligomycin resistance in Aspergi //us nidu/ans. Euro J 
Biochem 76: 345-354 

Melanson DL, MS Spencer 1981 Plant Physiol In press 

Mitchell P 1961 Coupling of phosphorylation to electron and hydrogen transfer 
by a chemiosmotic type of mechanism. Nature 191: 144-148 

Mitchell P 1966 Chemiosmotic coupling in oxidative and photosynthetic 
phosphorylation. Glynn Res Ltd, Bodmin, Cornwall, England 

Mitchell P 1966a Chemiosmotic coupling in oxidative and photosynthetic 
phosphorylation. Biol Rev 41: 445-502 

Mozersky SM, JD Pettinati, SD Kolman 1966 An improved method for the 
determination of orthophosphate suitable for the assay of adenosine 
triphosphatase activity. Anal Chem 38: 1182-1187 

Nelson N 1976 Structure and function of chloroplast ATPase Biochim Biophys 
Acta 456: 314-338 

Nelson N 1977 Chloroplast coupling factor. Encyc Plant Physiol 5: 393-404 

Oren R, Z Gromet-Elhanan 1977 Coupling factor adenosine 
triphosphatase-complex of Rhodosp/ri/ium rubrum - Isolation of an 
oligomycin-sensitive Ca, Mg-ATPase. FEBS Lett 79: 147-150 

Pedersen PL 1975 Mitochondrial adenosine triphosphatase. Bioenerg 6: 243-275 

Pedersen P 1976 ATP-dependent reactions catalysed by inner membrane vesicles 
of rat liver mitochondria. J Biol Chem 251: 934-940 


Penefsky HS 1974 Mitochondrial and chloroplast ATPases. The Enzymes 3 rd 
edition: 375-394 

Penefsky HS 1974b Differential effects of adenyl imidodiphosphate on ATP 
synthesis and the partial reactions of oxidative phosphorylation. J Biol Chem 
249: 3579-3585 

Pick U, E Racker 1979 Purification and reconstitution of the DCCD sensitive 
ATPase complex from spinach chloroplasts. J Biol Chem 254: 2793-2799 

Racker E, W Stoeckenius 1974 Reconstitution of purple membrane vesicles 
catalysing light driven proton uptake and adenosine triphosphate formation. J 
Biol Chem 249: 662-663 

Racker E 1977 Mechanisms of energy transformations. Ann Rev Biochem 46: 

Reisner AH, P Nemes, C Bucholtz 1975 The use of Coomassie Brilliant Blue 
G-250 perchloric acid solution for staining in electrophoresis and 
isoelectric focusing on polyacrylamide gels. Anal Biochem 64: 509-516 

Ryrie IJ 1975 Studies of the oligomycin sensitive ATPase from yeast 
mitochondria: Reconstitution of ATP-Pi exchange in the presence of 
phospholipids. J Supramol Structure 3: 242-247 

Ryrie IJ 1975a Purification and partial characterization of the oligomycinsensitive 
ATPase from yeast mitochondria. Arch Biochem Biophys 168: 712-719 

Ryrie IJ, PF 

Blackmore 1976 Energy linked activities in reconstituted 
triphosphatase proteoliposome. Arch Biochem Biophys 



Sadler MH, DR Hunter, RA Haworth 1974 Isolation of an ATP-Pi exchangease 
from lysolecithin-treated electron transport particles. Biochem Biophys Res 
Comm 59: 804-812 

Schneider E, K Altendorf 1980 Reconstitution of the purified proton conductor 
(Fo) of the ATPase complex from E. coli. FEBS Lett 116: 173-176 

Schneider E, P Friedl, U Schwulera, K Dose 1980 Energy-linked reactions 
catalysed by the purified ATPase complex (FoF 1) from Rhodospi ri / turn 
rubrum chromatophores. Euro J Biochem 108: 331-336 

Sedmak JJ, SE Grossberg 1977 A rapid, sensitive, and versatile assay for 
protein using Coomassie Brilliant Blue G-250. Anal Biochem 79: 544-552 

Senior AE 1973 The structure of mitochondrial ATPase. Biochim Biophys Acta 
301: 249-277 


Serrano R, Bl Kanner, E Racker 1976 Purification and properties of the 
protontranslocating ATPase complex of beef heart mitochondria. J Biol 
Chem 251: 2453-2461 

Solomos T, SS Malhotra, S Prasad, SK Malhotra, M Spencer 1972 Biochemical 
and structural changes in mitochondria and other cellular components of 
pea cotyledons during germination. Can J Biochem 50: 725-737 

Sone N, M Yoshida, H Hirata, Y Kagawa 1975 Purification and properties of a 
DCCD sensitive ATPase from a thermophilic bacterium. J Biol Chem 250: 

Sone N, M Yoshida, H Hirata, Y Kagawa 1977 ATP synthesis by electrochemical 
proton gradient in vesicles reconstituted from purified ATPase and 

phospholipids of thermophilic bacterium. J Biol Chem 252: 2956-2960 

Sone N, IVI Yoshida, H Hirata, Y Kagawa 1978 Resolution of the membrane 
moiety of the proton ATPase complex into two kinds of subunits. Proc 
Nat Acad Sci USA 75: 4219-4223 

Soper JW, PL Pedersen 1976 ATPase of rat liver mitochondria: Detergent 
solubilization of an oligomycin and DCCD sensitive form of the enzyme. 
Biochem 15: 2682-2690 

Stiggall PL, YM Galante, Y Hatefi 1978 Preparation and properties of an ATP-Pi 
exchange complex (Complex V) from beef heart mitochondria. J Biol Chem 
253: 956-964 

Swanljung P, L Frigeri 1972 Affinity chromatography of mitochondrial ATPase 
dispersed with TX-100. Biochim Biophys Acta 283: 391-394 

Swanljung P, L Frigeri, K Ohlson, L Ernster 1973 Studies on the activation of 
purified mitochondrial ATPase by phospholipids. Biochim Biophys Acta 305: 

Tzagoloff A, KM Byington, DH Maclennan 1968 Studies on the mitochondrial 
ATPase system. J Biol Chem 243: 2405-2412 

Tzagoloff A, P Meagher 1971 Assembly of the mitochondrial membrane system. 
J Biol Chem 246: 7328-7336 

Tzagoloff A, H Penefsky 1971 Extraction and purification of lipoprotein 
complexes from membranes. Methods in Enzymology XXII: 219-230 

Van de Stadt RJ, BL De Boer, F Van Dam 1973 The interaction between the 
mitochondrial ATPase (FI) and the ATPase inhibitor. Biochim Biophys Acta 
292: 338-349 

Williams JN 1964 A method for the simultaneous quantitative estimation of 
cytochromes a, b, cl, and c in mitochondria. Arch Biochem Biophys 107: 




Winget GD, N Kanner, E Racker 1977 Formation of ATP by the adenosine 
triphosphatase complex from spinach chloroplasts reconstituted together 
with bacteriorhodopsin. Biochim Biophys Acta 460: 490-499 

Wiskich JT 1977 Mitochondrial metabolite transport. Ann Rev Plant Physiol 28: 

Yoshida K, Y Takeuchi 1970 Properties of a soluble ATPase from castor bean 
endosperm mitochondria. Plant Cell Physiol 1 1: 403-409 



5.4 Development of New Pi Assay 

The third ATPase assay that was developed in this study consisted of 
the ATPase reaction conditions used in the second assay, coupled with a 
modified form of the Pi assay of Serrano et a! (1976). As explained in 
Materials and Methods, colour is developed by reduction of the 
phosphomolybdate complex by ascorbate, and the OD is read at 750 nm. 

The OD at 750 nm increases with time (Fig. 15). At low molybdate 
concentrations (0.1% w/v) the OD increases steadily for over an hour. At higher 
molybdate concentrations (0.35% and 0.7% w/v) the OD reached a plateau within 
5 minutes and only increased slowly from there on. Samples left overnight 
always turned blue/black by morning. The time interval that elapsed between 
ascorbate addition and OD reading therefore gave rise to a further source of 
variability and so was strictly limited to a convenient period, 10 minutes. The 
highest (0.70% w/v) concentration of molybdate apparently resulted in a more 
level OD plateau and was therefore used in subsequent assays. 

Construction of a standard curve showed the assay to be linear to 
approximately 400 nmoles (Fig. 16). Figure 16 also shows that the components 
of the assay are stable and they will give reproducible results over a period of 
at least 3 months. 

In any case, to eliminate any variability caused by the change in OD with 
time of reduction, or by aging of assay solutions, Pi standards (usually 100 
nmoles) accompanied every ATPase assay. 




Figure 15. Colour Development by the Reduced Phosphomolybdate Complex 

The assay mixture contained 125 nmoles of Pi in 1 ml (final volume) 
of water. Two ml of quench medium (0.72 N sulphuric acid and 
appropriate concentrations (w/v) of ammonium molybdate) were added, 
along with 50 ul of 1% (w/v) sodium ascorbate, and mixed. The 
optical density at 750 nm was measured immediately and as indicated. 
The mixtures were stored at 25 C between readings. (□): 0.1% (w/v) 
ammonium molybdate; (o): 0.35% (w/v) ammonium molybdate; (a): 0.7% 
ammonium molybdate. 

OD at 750 nm 

.00 0.05 0.10 0.15 0.20 0-25 0.30 

Reaction Time (min) 

Figure 16. Standard Curve for the New Pi Assay 

One ml of assay mixture (300 mM sucrose, 3 mM magnesium 
sulphate, 25 mM TES (pH 8.0) and an appropriate concentration of Pi) 
was mixed with 2 ml of quench medium (0.72 N sulphuric acid, and 
0.7% (w/v) ammonium molybdate). To develope colour 50 ul of 1% 
(w/v) sodium ascorbate was added. The solution was incubated at 25 
C for 10 min before measurement of the optical density at 750 nm. 

OD at 750 nrc 



nmoles Pi 



5.5 Development of Tissue Growth and Organelle Isolation Procedures 

The first goal of the project was recognized to be an increase in the 
scale of production of mitochondria. Most previous projects in this lab have 
required small amounts of freshly prepared mitochondria. The established 
procedure was, therefore, most suitable for the growth and grinding of 100 ml 
of expanded pea cotyledons (or up to 300 ml for SMP production). The most 
time consuming part of this procedure was the separation of the cotyledons 
from the shoot, root, and testa. In the present project, to eliminate this 
procedure mitochondria were isolated from peas that were imbibed for only 
24-48 hours (instead of 4 days). After 24-48 hr, the root and shoot were 
2-4 mm long and the testa usually unbroken. Whole peas had the disadvantages 
of not being a homogeneous tissue, and of the ATPase perhaps not being fully 
functional (Solomos et a/,1972). However, the great savings in time in harvesting 
and peeling peas were considered to outweigh the disadvantages, at least for 
the preparation of tissue for preliminary experiments. Most of these preliminary 
experiments, eg. optimization of cholate extraction procedures, and Affigel Blue 
chromatography, required large amounts of SMP. SMP used in later experiments 
(from ammonium sulphate precipitation chromatography on) were always 4 day 
old greenhouse peas (see Chapter 2). 

Peas used in technique 1 (Table 8) were imbibed for 24 hr, while 
suspended in aerated tap water in a large round bottom flask (the bubbler 
technique). Various methods were tried to keep the peas suspended and 
therefore sufficiently aerated (magnetic stirrer, mechanical stirrer, air jets). As the 
amount of peas in the flask was scaled up it was progressively more difficult 
to keep all of the peas suspended without incurring mechanical damage to the 
peas. Furthermore, the specific activity of the SMP, although variable, was low 
(Table 8). After experimentation with different grinding methods, the amount of 
grinding was concluded to be the major factor in maintenance of the yield of 
protein (data not shown). Centrifuge tube and bottle shapes were also varied to 
reduce contamination of the mitochondrial and SMP pellets by the lipid cake 


Table 8. Development of Growth and Organelle Isolation Froaedures 

Peas were imbibed and grown and mitochondria were isolated as described 
in Chapter 2, except for treatment 3. For this procedure, the wash step 
was omitted. The mitochondria were instead loaded immediately onto a step 
gradient containing 40 ml of 0.6 M sucrose with 50 mM TES (pH 7.0) and 
20 ml of 1.6 M sucrose with 50 mM TES (pH 7.0). The gradient was centri¬ 
fuged at 30,000 g for 35 min in a swingout head. Approximately 60% of the 
ATPase activity was recovered at the 0.6 M to 1.6 M interface. SMP2 were 
made as described in Chapter 2. Protein and ATPase activity were measured 
as described in Chapter 2. The data in this table were from representa¬ 
tive experiments. —=not determined. 


Dry peas (grn) 





activity of SMP 


1. bubbler 




2. tray 




3. tray, step 





4. tray, roots 




& testa removed 

5. greenhouse 




6. greenhouse 

1 ,000 





which normally floated on top of the supernatant layer after the second spin. 
However, this problem eventually appeared (data not shown) to be correlated to 
over long grinding of tissues, which possibly caused organelle and membrane 
destruction. The best results were obtained when the tissue was pounded in a 
mortar and pestle until on 90% of strokes the pestle did not hit any unground 
cotyledons. Neither of these findings, however, enabled the specific activity of 
the SMP to be raised any further, and so the bubbler technique was abandoned 
in favour of the tray method that is described below. 

For the tray technique, peas were soaked for 6 hrs in tap water and 
then sandwiched in a single layer between 2 layers of paper towels soaked in 
deionized water. The paper towels and peas were laid in fiberglass trays that 
were covered with plastic mesh and filled to the bottom layer of paper with 
deionized water. The trays were kept in a high humidity environment at 27 C 
until the peas were removed for isolation of mitochondria. 

This technique was easily scaled up to as much as 1700 ml of peas, 
and gave SMP with only slightly below average specific activity (see Table 8). A 






Spencer, 1973) 



to try 




the mitochondria 

before sonication, however. 









8). This may 





inactivation by high sucrose concentrations, or by the undeveloped mitochondria 
having a different density to that of the mature mitochondria used by Malhotra 
and Spencer (1973). Removal of the roots, shoots, and testa of tray grown 

peas was also experimented with, but this resulted in a slight decrease in 

specific activity (Table 8). 

Gradient "purified", tray SMP (48 hr) and 24 hr bubbled, "unpurified" SMP 
were both used for preliminary detergent extractions. The former type of SMP 

was also used for the optimization of cholate extraction, Affigel Blue 

chromatography, and ammonium sulphate precipitation. The susceptibility of both 

types of preparations to detergents, etc., was found to be identical. 

In a direct comparison, with the same dried peas, media, grinding 
methods, etc., 4 day old greenhouse SMP (Chapter 2) had a 10% higher 
specific activity than SMP from 48 hr tray peas (preparations 2 and 5, Table 



8), therefore, SMP2 (see Chapter 2) were made from 4 day old greenhouse 
peas by the method of Coleman and Palmer (1972), with modifications as 
described in Materials and Methods. SMP2 had a specific activity up to 10 
times higher than that of SMP, although the protein yield was less than half of 
the usual yield (Table 8). It was found that four 1 min bursts of sonication 

raised the yield 20% (data not shown) and that sonication of fresh mitchondria 

(rather than previously frozen mitochondria) raised the specific activity even 
more (60%) (data not shown). Furthermore, the extraction of SMP2 with cholate 
resulted in CE with a slightly lower specific activity than SMP2, whereas cholate 
extraction of SMP normally resulted in CE with a 2-3 fold higher specific 

activity than SMP. (This was discussed in the section on detergent extraction. 
Chapter 3.) Nevertheless, SMP2 was used for the remaining experiments detailed 
in chapter 3 while other possible ways of increasing the yield were being 
explored. Four day old greenhouse SMP were exclusively used for the 

experiments in Chapter 4.