II': i This dissertation has been i microfilmed exactly as received 69-10,593
. ASATO, Robert Noriyoshi, 1942- MULTIPLE FORMS OF BACTERIAL HYDRO GENASES.
University of Hawaii, Ph.D., 1968 Biochemistry
University Microfilms, Inc., Ann Arbor, Michigan MULTIPLE FORMS OF BACTERIAL HYDROGENASES
A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF
THE UNIVERSITY OF HAWAII IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN BIOCHEMISTRY
SEPTEMBER 1968
by
Robert Noriyoshi Asato
Dissertation Committee:
Dr. Howard F. Mower Dr. Kerry T. Yasunobu Dr. Morton Mandel Dr. John B. Hall Dr. Hilmer A. Frank ii
ABSTRACT
Hydrogenase enzymes from many distinct bacterial species have been shown by disc electrophoresis on polyacrylamide gel to exist in multiple forms. The enzymes of the hydrogenase systems have different electrophoretic mobilities and produce a band pattern that is unique for each species. These studies have demonstrated the utility of this method for the taxonomic identification of hydrogenase containing bacteria. A tentative identification of an isolated bioluminescent bacteria as ~. phosphorium was made using this method. This multiplicity of hydrogenase forms was found both in bacteria which contain mostly soluble hydrogenases and in those where the hydrogenase is predominantly associated with particulate material. When solubilization of this particulate material could be effected, at least two solubilized hydrogenases were released, and, of these, one would have the same electrophoretic properties (RF) as one of the soluble hydrogenases already present in small amounts within the cell.
Different growth conditions for various types of bacteria, such as the nitrogen source, iron in the medium, the degree of aeration, and photosynthetic versus aerobic growth in the dark, as well as the conditions under which the cells were stored, markedly affected the hydrogenase activity of the cells, but not their hydrogenase band patterns.
The specificities of various isoenzymes of certain bacteria were tested with redox dyes of different potentials, and no difference among isoenzymes of a particular organism was observed. Three of the six forms of £. pasteurianum were found to be interrelated with the iii
~ and S forms always converting to the more stable y species. The molecular weights of these forms were approximately 50,000. Recently developed techniques such as sucrose density gradient centrifugation, and gel filtration allowed studies on hydrogenase in the presence of other proteins by a direct assay of its biological activity.
Comparative studies were done on the size and shape of hydrogenase enzymes of £. butyricum and £. butylicum, with the former possessing isoenzymes of the same weight as those of £. pasteurianum. Kinetic studies on purified preparations of £. pasteurianum and
C. butylicum hydrogenases demonstrated that molybdenum was involved in the one-electron transfer to methyl viologen. The participation of iron in the viologen assay of £. butylicum hydrogenase was also shown.
A purification scheme for the isolation of £. butylicum hydrogenase is presented. The final extract possessed a specific activity of approximately 4.Z liters HZ evolved per hour per mg protein.
This represented a 500 fold purification over the crude preparation and was stable when stored under HZ at 0° for up to several weeks. The estimation of the turnover number of this partially purified form of hydrogenase indicate that it is at least three times more active than catalase, the most active enzyme known.
It has been shown that for all hydrogenase enzymes examined there is an abrupt change in the Arrhenius activation energy at 17°. This was true for a variety of hydrogenase enzymes, catalyzing a variety of reactions. This transition temperature corresponds exactly with a similar transition in the electrophoretic mobilities of the various iv hydrogenases on polyacrylamide gel. It is believed that this phenomenon is the manifestation of a conformational change occurring at this temperature. v
TABLE OF CONTENTS
ABSTRACT •• •• ii
LIST OF TABLES vii
LIST OF ILLUSTRATIONS ix
INTRODUCTION •••••• 1
MATERIALS AND METHODS 23
Materials 23
Methods 23
Growth of Bacteria ..•• 23 Gel Electrophoresis . .••• 24 Sucrose Density Gradient Centrifugation • 25 Gel Filtration ••••.•••••.•• 26 Enzymatic Assays •.•.••.••.•• 27 Arrhenius Activation Energy Determinations 29
RESULTS 30
(A) Survey of Bacterial Hydrogenases • 30
Hydrogenase Band Patterns of Bacteria Having More Than 80% of Their Activity in a Soluble Form ...... 30
Hydrogenase Band Patterns of Bacteria Having More Than 50% of Their Activity in an Easily Sedimentable Form •.•• 47
Organisms Containing Little or No Hydrogenase Activity .••.•••• •• 59
(B) Redox Potentials and Specificities of Various Hydrogenases ••••.•••••. 64
Hydrogenases of High Specific Activities 64
Hydrogenases of Low Specific Activities 67
(C) Molecular Weight Determinations of Various Hydrogenases •••...•• 72
Sedimentation Analysis 73 vi
Gel Filtration •••• 78
(D) Study of the 3 Isoenzymes of C. Pasteurianum 93
Inter-relationship Studies on ~. pasteurianum 93
Iron and Hydrogenase Activity in ~. pasteurianum 94
Gel Concentrations and Electrophoretic Mobilities . .•••••••• 98
(E) Purification of C. Butylicum Hydrogenase •. 101
Purification Procedures • 101
Properties of Purified Hydrogenase • •• 109 (F) Kinetic Assays on the Co-factor Requirements of Hydrogenase •••.•.•••••••.•••••. 115
Effect of Co-factors on C. pasteurianum -Hydrogenase •.••••••••••.•• 115
Effect of Co-factors on~. butylicum Hydrogenase ••.• .•••...• 120
(G) Arrhenius Activati.on Energies of Hydrogenase Enzymes •.. • 128
Activation Energies of Various Hydrogenases ·• 129
DISCUSSION • 149
LITERATURE CITED · 164 vii LIST OF TABLES
Table
I SOURCE AND GROWTH CONDITIONS OF BACTERIA • 31
II BAND PATTERNS OF SOLUBLE HYDROGENASES 36
III BAND PATTERNS OF SOLUBLE AND SOLUBILIZED HYDROGENASES ••...•.•••. 48
IV BAND PATTERNS OF DIFFERENT STRAINS OF E. COLI HYDROGENASES •.••• 50
V EFFECT OF GROWTH AND STORAGE CONDITIONS ON R. RUBRUM HYDROGENASE ...•.. 53 VI REDOX POTENTIALS OF HYDROGENASESOF RELATIVELY HIGH SPECIFIC ACTIVITIES 65 VII REDOX POTENTIALS OF HYDROGENASES OF RELATIVELY LOW SPECIFIC ACTIVITIES • 69
VIII SPECIFIC ACTIVITIES OF VARIOUS HYDROGENASES CATALYZING THE EVOLUTION OF HYDROGEN FROM REDUCED METHYL VIOLOGEN ..••.•• · · · ·· 70 IX SEDIMENTATION CONSTANTS AND MOLECULAR WEIGHTS OF VARIOUS HYDROGENASES ··· · · · 74 X MOLECULAR WEIGHTS OF VARIOUS HYDROGENASES AS DETERMINED BY S VALUES .. ..•.•. · · · · · · 78 XI MOLECULAR WEIGHTS OF VARIOUS HYDROGENASES BY GEL FILTRATION •....•••..•. · · ·· ·· 79 XII MOLECULAR WEIGHTS OF VARIOUS HYDROGENASES FROM SEDIMENTATION AND GEL FILTRATION DATA · · ·· · · 92 XIII EFFECT OF IRON ON HYDROGENASE ACTIVITY ON £. PASTEURIANUM HYDROGENASE •..... 95
XIV PURIFICATION OF C. BUTYLICUM HYDROGENASE . 105 xv pH OPTIMUM OF CRUDE AND PURIFIED HYDROGENASE . 110
XVI REDOX SPECIFICITIES OF CRUDE AND PURIFIED HYDROGENASES ••.•...••.••••• 111 XVII EFFECT OF MOLYBDENUM AND IRON ON £. PASTEURIANUM HYDROGENASE AFTER GEL FILTRATION .•••.•.• 116 viii
Table
XVIII EFFECT OF MOLYBDENUM AND IRON ON THE METHYLENE BLUE ACTIVITY OF C. PASTEURIANUM HYDROGENASE AFTER GEL FILTRATION •••• 117 XIX VMAX DETERMINATION OF UNDIALYZED EXTRACT . 120 xx EFFECT OF CO-FACTORS ON C. BUTYLICUM HYDROGENASE .•.•••.•...... 123 XXI . EFFECT OF VARIOUS METALS ON THE HYDROGENASE OF .£. BUTYLICUM •. 127 XXII TEMPERATURE CHARACTERISTIC OF BACTERIAL SUBSTRATE COMBINATIONS •....•• 132 XXIII MOLECULAR WEIGHT DETERMINATIONS OF £. BUTYLICUM HYDROGENASE AT 5 AND 20 142 XXIV VARIATION OF PROTEIN MOBILITIES WITH TEMPERATURE ON POLYACRYLAMIDE GEL ..•.•.•. •. 143 ix LIST OF ILLUSTRATIONS
Figures
1 DESITOMETER TRACINGS OF HYDROGENASE BANDS ON POLYACRYLAMIDE GELS •••. 40
2 BAND PATTERNS OF VARIOUS HYDROGENASES 42
3 GROWTH CURVE OF .£. BUTYLICUM • 44
4 GROWTH CURVE OF A. VINELANDII 46
5 PROPORTION OF SOLUBLE HYDROGENASE IN A. VINELANDII WITH AGE •••. 56
6 HYDROGENASE BAND PATTERN OF H. EUTROPHA 63
7 SEDIMENTATION PATTERN OF THE Y ISOENZYME OF £. PASTEURIANUM ••..•..•.•. 77
8 MOLECULAR WEIGHTS OF HYDROGENASES AND VARIOUS PROTEIN STANDARDS FROM SEPHADEX G-75 •••.• 81
9 MOLECULAR WEIGHTS OF HYDROGENASES AND VARIOUS PROTEIN STANDARDS FROM SEPHADEX G-100 .•.. 83 10 ELUTION PATTERN OF VARIOUS HYDROGENASES FROM SEPHADEX G-100 .•.•..•.. 85
11 CORRELATION OF KAV WITH STOKES RADIUS ON SEPHADEX G-75 •...•••...• 89
12 CORRELATION OF KAV WITH STOKES RADIUS ON SEPHADEX G-100 •••. •••••• 91
13 EFFECT OF STORAGE ON THE HYDROGENASES OF C. PASTEURIANUM ...•.•••.. 97
14 RELATIVE MOBILITIES OF THE ISOENZYMES OF C. PASTEURIANUM IN VARIOUS GEL CONCENTRATIONS 100
15 POLYACRYLAMIDE ASSAY OF A PURIFIED EXTRACT OF C. BUTYLICUM HYDROGENASE .•••.•. 107 16 MAXIMUM VELOCITY AND TURNOVER NUMBER OF PURIFIED HYDROGENASE OF Q. BUTYLICUM •• 114 17 REACTIVATION OF POOLED ELUATES OF Q. PASTEURIANUM HYDROGENASE FROM SEPHADEX G-100 ••.•.•. 119 x
Figures
18 REACTIVATION OF C. PASTEURIANUM HYDROGENASE BY MOLYBDENUM ••• ...... • 122 19 EFFECT OF FAD ON A PURIFIED EXTRACT OF C. BUTYLICUM ••••.••••• . 126
20 ARRHENIUS PLOT FOR THE EVOLUTION OF HYDROGEN FROM REDUCED METHYL VIOLOGEN IN EXTRACTS OF .Q.. PASTEURIANUM .••••• .. •.••.•.• 131
21 VARIATION OF Rf OF THE ~ AND S BANDS OF .£. PASTEURIANUM WITH TEMPERATURE ••. 135
22 VARIATION OF Rf OF THE Y BAND OF C. PASTEURIANUM WITH TEMPERATURE 137
23 ARRHENIUS PLOT FOR THE EVOLUTION OF HYDROGEN FROM REDUCED METHYL VIOLOGEN IN EXTRACTS OF C. BUTYLICUM .•.•.•.•.•.• 139
24 VARIATION OF Rf OF THE HYDROGENASE BAND OF .£. BUTYLICUM WITH TEMPERATURE •.• 141
25 VARIATION OF Rf OF BOVINE SERUM ALBUMIN WITH TEMPERATURE .•..•..•... 145
26 VARIATIONS OF Rf'S OF VARIOUS PROTEINS WITH TEMPERATURE ...... ••••.. 147 INTRODUCTION
Bacteria have long been known which are capable of oxidizing hydrogen gas by molecular oxygen and living autotrophica11y on the energy thus liberated (1,Z,3). Stephenson and Stickland (4) while investigating the anaerobic fermentation of fatty acids to methane by a mixed culture from river mud, obtained a sample which reduced sulfate to sulfide, and decomposed formate quantitatively to methane, carbon dioxide, and water. The same culture also synthesized methane from a mixture of carbon dioxide and hydrogen and could simultaneously reduce sulfate to sulfide with the use of this gas. They concluded that carbon dioxide and sulfate were acting as hydrogen acceptors in a system where molecular hydrogen served as a donor and that the bacteria in the mixed
culture were somehow able to activate it. The activating process was
expressed as,
+
and the reversible enzyme system responsible was termed "hydrogenase" (4,5).
The reduction of methylene blue (Eo,+10mv) was used as a criterion
by Stephenson and Stickland for hydrogenase activity (4). Other
techniques have since been developed. They include the reduction or
evolution reactions with the natural acceptors such as ferredoxin (6);
the fixation of tritium (7); the conversion of para to ortho hydrogen
(8); hydrogen-deuterium exchange (8); the reduction of one and two
electron dyes (9); and the evolution of HZ from reduced one electron
dyes (9). All hydrogenase catalyzed reactions are reversible, and it Z is assumed that the activation of HZ is similar for all organisms
though the transfer of electrons after activation may occur via different pathways depending upon the final acceptor and organism
concerned (10).
The ability to produce or utilize HZ has been observed in a
relatively large number of micro-organisms since the initial
observation of Kaserer (Z). Hydrogenase activity has been shown to
exist in a wide variety of bacteria encompassing the strict aerobes,
facultative anaerobes, strict anaerobes, the photosynthetic and
chemo-autotrophic bacteria (10). To clarify the role that hydrogenase
plays in each group, the representative organisms can be separated into
several distinct categories based on their metabolic pathways and
prominent group characteristics.
The first category is comprised of the strict heterotrophic
anaerobes where growth is inhibited by molecular oxygen. As far as is
known, no microbes in this group possess electron carriers of the
cytochrome type. For most of the organisms, and in particular the
clostridia, pyruvate or reduced Z-carbon compounds, serve as electron
donors for hydrogen formation. In some species these or closely
related donors are generated by the anaerobic decomposition of amino
acids (10). Other anaerobic bacteria, such as the micrococci, can be
further distinguished by their ability to produce hydrogen from purines
and pyrimidines (10).
The metabolism of most bacteria of category I follow a similar
energy producing sequence. In the fermentation of the energy source,
pyruvate is generated as an important intermediate which is further 3 degraded by an energy yielding Iphosphorc1astic" reaction giving acetyl phosphate, carbon dioxide, and hydrogen as shown below:
The mechanism of this reaction is assumed to be similar to the oxidative decarboxylation of pyruvate in mammalian systems with the participation of coenzyme A thiamine pyrophosphate, lipoic acid and nicotinamide adenine dinucleotide. Free formate is not produced as an intermediate in this reaction and electrons are released during the oxidative decarboxylation of pyruvate as hydrogen gas (11,lZ,13,14).
The nature and sequence of catalysts constituting the terminal portion of the hydrogen evolving system was found to be similar for all organisms of category I. Early experimentors using clostridial hydrogenase observed many parallels between the ability of cell free extracts to produce hydrogen from pyruvate in one instance, and from sodium dithionite in the other; this led to the suggestion that a common intermediate was required for the shuttle of electrons to the hydrogenase (15). Methyl vio1ogen, a low redox potential one-electron dye, would effectively replace the unknown factor which was lost during the purification of the hydrogenase. Known electron-transfer cofactors, including flavin and pyridine nuc1eotides, were ineffective (15,16).
Subsequent investigators showed that the natural cofactor was an iron containing protein now named ferredoxin (17,18). A general scheme for the production of HZ by strict heterotrophic anaerobes can thus be presented as: 4
CH3CH20H Other electron donors J, (Na2S204, etc.) - ~ cH3co~::[CH3IHO_X] ~ (FdOXidized) \. hydrogenase • HZ Fdreduced
CH3COCoA CH3COOP04H2 + --~) + > H3P04 ADP
Where Fd represents ferredoxin which can be replaced by reduced methyl vio1ogen and_CH3CHO-X represents the enzyme bound complex containing several co-factors such as CoA and thiamine pyrophosphate. Artificial electron donors such as sodium dithionite can replace the natural substrates in cell free preparations.
Similar iron proteins have been isolated from various clostridia and other micro-organisms and show several similarities in amino acid composition and chemical and physical properties (19,20,21). In some instances the specific ferredoxins have been shown to be functionally interchangeable in cell-free preparations (22). It should be emphasized, however, that ferredoxins do not always act as electron carriers in processes resulting in hydrogen formation, but they also function in such reactions as the reduction of pyridine nuc1eotides by pyruvate (23), in N2 fixation (10); and the light-dependent reduction of NADP by photosynthetic bacteria (19). Also, ferredoxins are found in organisms which do not produce hydrogen (10).
The second group is comprised of the heterotrophic facultative anaerobes that contain cytochromes and evolve hydrogen from formate, when grown anaerobically. Escherichia coli and related bacteria are 5
the major representatives of this group. In the presence of oxygen,
these organisms obtain growth energy by the oxidation of pyruvate
through an aerobic citric acid cycle coupled to a primitive electron
transport system. As electrons from pyruvate travel through a sequence
of bacterial cytochromes and, finally to oxygen, ATP is formed by a mechanism assumed to be similar to the oxidative phosphorylative process
found in plants and animals. Under this condition, the cell economizes
by repressing the anaerobic components that are present. In the
absence of oxygen, as pyruvate is generated from glycolysis, it is
metabolized through various alternative routes, one of which includes
the type of "clastic cleavage" indicated below:
+ CH3COOP03H2 + HCOOH ;;P 4 ATP +
This reaction is similar to the clostridial phosphorclastic reaction
but differ in certain basic aspects. These differences were elucidated
by using the enzymes of E. coli as a model system (24,25,26).
In the strains of this species, formate produced by the
phosphorclastic reaction, is converted to carbon dioxide and hydrogen
by an additional sequence of reactions, called the "formate
dehydrogenlyase" system (25,26). This enzyme complex is comprised of
a soluble and a particulate dehydrogenase, a particulate hydrogenase
and 2 unknown electron carriers designated as Xl and X2 (25,26) as
shown below: 6
°2 (cytochromes) ---\-.~) H20 C02 i ATP -~ ~FDH HCOOH~ •P ~ FDHs Xl hydrogenase) H2 2H+7
Under aerobic conditions, the particulate (membrane bound) enzyme
(FDHp) oxidizes formate to carbon dioxide and water via a system of high potential cytochromes (26). ATP generated by the electron transport from formate to oxygen provides the growth energy. The activities of the soluble dehydrogenase and hydrogenase are low or nonexistent (27); Xl and X2 are not formed. Anaerobically, the synthesis of the particulate dehydrogenase and low potential cytochromes is derepressed and the soluble formate dehydrogenase (FDHs ) is synthesized. This enzyme is coupled to the particulate hydrogenase by carriers Xl and X2 oxidizing formate to carbon dioxide and H2' The particulate dehydrogenase activity still occurs but is greatly reduced as oxygen, its terminal acceptor, has been removed (27).
Recent studies with coliform variants have strongly indicated that
X2 is a c-type cytochrome of low redox potential (29). This low redox potential intermediate is produced only during growth under anaerobic conditions--conditions which are favorable for the development of the hydrogenlyase system (28). Xl may function as a cytochrome c-reductase or, as in the case of Desulfovibrio desulfuricans, a cytochrome c3-reductase (28).
The hydrogen-evolving systems of strict and facultative anaerobes differ in several other functional aspects. So far, no ferredoxin-like 7 compounds have been detected in facultative anaerobes (30). Components isolated from strict anaerobes have consistently failed to interact with those of the facultative anaerobes (25). As an example, the formate dehydrogenase of coliform organisms will not couple with the hydrogenase in crude extracts from clostridia to give a model hydrogenlyase system (31). These observations have led to the postulate that the bacteria of this group possess a formate dehydrogen1yase which is obligatorily linked to the low-redox potential cytochromes of the system and ferredoxin does not participate in the fermentative pathways of these bacteria. Evidence of a functional cytochrome c3 in the hydrogenlyase system of D. desulfuricans adds more weight to this conclusion (28,32).
Desu1fovibrio desu1furicans is an example of a heterotrophic strict anaerobe that contains cytochromes and carries out both a coli-type hydrogen1yase reaction and clostridial-type phosphorc1astic cleavage
(10). This bacteria is placed in a category by itself as it is thought to be a transitional form (between the strict anaerobes and the facultative anaerobes) which normally uses sulfate as the terminal oxidant for the energy yielding reactions. However, certain strains of this species can liberate hydrogen from pyruvate and formate in the absence of sulfate (10).
The next category consist of the obligate aerobes of which
Azotobacter vine1andii is an example. This organism obtains its growth energy through the operation of an aerobic tricarboxylic acid cycle and cytochrome system using oxygen as the terminal acceptor (33,34,35). It contains no ferredoxin or cytochrome-type carriers of low redox 8 potential (62). This organism, a strict aerobe possesses a hydrogenase with no obvious function as it does not produce or utilize hydrogen.
Recent work (66,71,77) indicates that the existence of hydrogenase
activity in these cells is related to its nitrogen fixing ability (67).
The ability to utilize hydrogen as the sole source of oxidizable
energy in an autotrophic environment characterizes the next group of
bacteria, the chemo-autotrophs. The micro-organisms of this group use
the hydrogenase catalyzed oxidation of hydrogen gas by molecular oxygen
(Knal1gas reaction) as their only source of growth energy (36). The
energy, liberated by the transport of reducing equivalents from an
activated hydrogen molecule to oxygen, is coupled to carbon dioxide
fixation and other metabolic reactions (37,38). These bacteria need
only a mixture of gases (H2, C02 and 02) and mineral salts for growth
(84). The fact that oxygen is a potent inhibi.tor of hydrogenase (85,
86), has prompted the current studies on the mode of coupling and the
role of hydrogenase in this system (39).
The final category consists of the photosynthetic bacteria in
which hydrogen can be produced in one of two ways. Cells grown in the
dark can produce this gas by a mechanism which is similar to that
occurring in categories I and II (10). When these cells are
illuminated, they continue to produce hydrogen but by a pathway
dependent upon light (40,41,42). Recent studies (43,44,45) indicate
the operation .of an anaerobic citric acid cycle which is coupled to
the photoproduction of hydrogen to effect the reoxidation of the
reduced pyridine nucleotides produced by enzymes of the cycle. This
can be represented schematically as, 9
PN+ Hydrogenase ( H2 ~(---~) Fd ( ~ ~ PNH < TCA CYCLE T/ ATP light where PN+ and PNH represent the oxidized and reduced forms of the pyridine nucleotides respectively. Light provides the energy for the generation of hydrogen and ATP. Ferredoxin (Fd) plays a central role in both photoproduction of hydrogen and photophosphorylation (46,47).
The extent of hydrogen production in all organisms is strongly influenced by the competition for electrons among alternative acceptors.
In general, the yield of hydrogen is directly related to the state of reduction of the fermented energy source, and tends to be inversely related to the total quantity of reduced end products which accumulates in the medium (10). It has been suggested that the level of hydrogenase activity is an important control point regulating the flow of electrons into various pathways (63,30). This hypothesis is supported by the results of several experiments with clostridial and coli-aerogenes bacteria in which there was a shift in fermentative pathways whenever
the hydrogen evolving system was inhibited. For example, when the hydrogen-evolving complex was inhibited by carbon monoxide, fermentation
of glucose by Clostridium butyricum was characterized by the production
of lactate rather than H2' C02, and fatty acids (48). Similar shifts
occur if iron, which is thought to be essential for hydrogenase
activity, is withheld from the media of various bacteria (64,65). 10
In certain clostridial species, the control systems which regulate
anaerobic electron flow, seem to reutilize the evolved hydrogen. This
is suggested by the observation that preparations of clostridial hydrogenase can reduce pyridine nucleotides by a ferredoxin dependent
reaction (23). Since the reduced pyridine nuc1eotides playa crucial
role in many physiological reactions, a reversible hydrogenase system
serves as a delicate regulator of electron flow (23). The tight control
of the flow of electrons in the anaerobic mode of growth is important as
the mechanism for ATP formation under these conditions are rather
inefficient and requires the disposal of a comparatively large number
of electrons (10).
In photosynthetic bacteria, maximum photoproduction of H2 occurs
under conditions in which the cell might be expected to produce ATP, or
its precursors, and reduced pyridine nucleotides in excess of its
requirements (43,45). Accordingly, the photoproduction of hydrogen can
be viewed as a regulatory mechanism which aids in the balancing of the
availability of the stored energy forms with the demands imposed on the
cell by its metabolic activity (43,44,46). This interpretation is
supported by the action of inhibitors of photophosphorylation such as
antimycin A and certain redox dyes. These compounds, while abolishing
the formation of H2 by interfering with the light-induced phosphorylation,
effects a shift in the fermentative pathways causing the cells to behave
as though in darkness where the fermentation of endogenous fatty acids
occur (43). In addition, uncouplers of oxidative phosphorylation in
mitochondria are also potent inhibitors of the photoproduction of H2
(46). 11
In conformity with the foregoing, hydrogenase activity has been
found to be linked with nitrogen metabolism (49,50,52,53). The studies on several nitrogen fixing bacteria which possess hydrogenase can be summarized in the following way. Lee and Wilson (54) working with
Azotobacter vinelandii demonstrated that the level of hydrogen uptake
activity depended to a large extent upon the nitrogen fixing system of
this obligate aerobe. Cells grown on a fixed nitrogen source had
considerably less hydrogenase activity than those grown under molecular nitrogen. Clostridium pasteurianum reacted in a similar manner (52), but in Rhodospirillum rubrum, the hydrogenase activity was inhibited by both fixed nitrogen and molecular nitrogen (49,50).
Subsequent investigation with A. vinelandii (66,67), with £.
pasteurianum (52,55,57,58), and R. rubrum (50,59,68) uncovered an
ATP-dependent hydrogen evolution activity which was situated within the
N2-fixing complex called nitrogenase. It is not known whether this
enzyme is a different protein or the same compound as the normal or
ATP-independenl hydrogenase but altered by a conformational change
(58,59,60). The ATP dependent enzyme can only work in one direction,
towards the evolution of HZ, in all species studied. On the other
hand, the normal hydrogenases of A. vinelandii and R. rubrum exhibit
a very strong reduction reaction (uptake of H2) and a weak or
nonexisting hydrogen evolving capacity (77,78). Extracts of C.
pasteurianum catalyze both reactions equally (9).
The correlation between the ATP-dependent hydrogen evolution
activity in all three species and growth conditions can be explained
if it is assumed that this enzyme is situated within the nitrogen 1Z fixing complex. In the presence of a fixed nitrogen source, there would be no need for nitrogenase, and the enzymes of this system are repressed
(66), and the activities will be low or nonexistent. Molecular nitrogen stimulates both activities by de-repressing the synthesis of the nitrogen-fixing complex.
The correlation between the normal hydrogenase activities and growth conditions is still obscure. Nitrogen fixation in cell free extracts of A. vinelandii, which does not contain ferredoxin, can occur only if an artificial electron donor (clostridial hydrogenase and ferredoxin, or sodium dithionite) is present (59,67). Whether this is because the Azotobacter hydrogenase itself does not participate in NZ reduction or whether it needs a specific carrier similar to ferredoxin is not known. Recently, there has been a report (6Z) of the isolation of a nonheme iron proten in A. vinelandii which may couple with the
ATP-independent enzyme to furnish electrons for the fixing of NZ. If this were so, though the results are still fragmentary, the system of
A. vindelandii could be explained by the scheme proposed for the hydrogenase-nitrogenase complex of I. pasteurianum and R. rubrum (next page) • /" Mo+6 - hydrogenase W) Cdred) (2 ox. t (TP HZ Fdox. DZ red.)\ ADP 't Mo+5 NaZSZ04 \ ATP - dependent hydrogenase NITROGENASE-HYDROGENASE SYSTEM
ATP Dependent Hydrogenase W / Mo +6 Hydrogenase - - -..... N D lilt I ( Fe +2 ~ I / Z Z ( ATP ( D3 ) ADP Fe +z ) , \ I -¥ NH3 HZ .. Fd ' -'D ) Z (ox. ) D (red. ) \ 1 /. r Mo+ ...... " Sodium 5 - dithionite HZ
I--' W 14
According to this interpretation, hydrogenase functions as a donor or acceptor of electrons supplied to, or furnished by ferredoxin (Fd).
Ferredoxin can be replaced by artificial electron donors such as sodium dithionite or reduced methyl viologen (58). D1 and DZ are two protein components, recently isolated by Mortenson (58), containing molybdenum and devoid of molybdenum, respectively. A third protein component has been proposed (D3)' These proteins are synthesized only under nitrogen fixing conditions (55,57,58). Ferredoxin, or in the case of A. vine1andii, a nonheme iron protein, serves as a common donor of electrons to the nitrogenase system and the hydrogenase enzyme. Whether the activities of the normal hydrogenase enzymes are stimulated or inhibited depends on the extent of competition between the hydrogenase and nitrogenase for electrons, and the availability of electrons produced by the fermentative pathways of the organism. Hence in some cases, molecular NZ inhibits and in others it stimulates Hz evolution. A
complete explanation of the results awaits further experimentation. It has been suggested that the protein that catalyzes the ATP-dependent
evolution of HZ can be controlled by an allosteric transition caused by
the nitrogen molecule. According to this idea (45), the appearance of
a utilizable nitrogen source which satisfies the requirements for the
synthesis of amino acids and proteins would signal the cessation of HZ
evolution from organic compounds.
The existence of hydrogenase activity in many diverse organisms
with significantly different metabolic pathways raises the question on
the mechanism of action of this enzyme. Though the enzyme has not yet
been purified, hydrogenase has been subject to many investigations 15
about its mode of operation. Its properties vary from one bacteria to
another, but there are several points of general agreement (69,73).
Krasna and Rittenberg (69), working with the hydrogenase system of
Proteus vulgaris, discovered that both whole cells and cell-free
extracts of this facultative anaerobe catalyzed the exchange reaction
between heavy water (DZO) and HZ and the conversion of para to ortho hydrogen. While extracts, suspended in HZO, catalyzed the conversion
of pure para to normal hydrogen (ratio of 3 ortho to 1 para), no
conversion occurred if DZO was substituted for HZO. In accord with
these experiments, the following mechanism was proposed:
HZ + H:E + H+ slow exchange rapid with water exchange with water
where E represented the enzyme, H:E the hydride of the enzyme and the
effective reducing agent.
This representation explained the para to ortho conversion in HZO,
for the back reaction (kZ) will produce normal hydrogen (HZn) (70). In pure DZO, however, H+ is diluted with the large amount of D+ andHD is
formed, not HZn • The above scheme will also explain the exchange reaction, since the H+ ion is in rapid equilibrium with water.
According to this mechanism, hydrogenase action consists of Z steps; .
the activation of HZ, and its transfer to various substrates. The
transfer is effected by the hydride of the enzyme, H:E-, as indicated
in the reduction of fumarate.
+ + fumarate + succinate + E 16
In a system containing E. coli, HZO, DZ gas, and fumarate, it was found that the hydrogen added to the carbon-carbon double bond contained ZO% deuterium (74). It also indicated that the activated hydrogen was not in immediate equilibrium with water.
The mechanism of the enzymatic reaction is entirely different from that involved in the platinum catalyzed reaction. In experiments in which_~Z and pure DZO were the reactants, DZ was the most abundant molecule formed. The splitting of the hydrogen molecule by platinum appeared to be a homolytic one where both hydrogen atoms attached to the metal could easily exchange with the water, as shown in the
following scheme:
k1 D+ D+ HZ + Pt c t HPtH ~HPtD , ) DptD kZ k3 k3 k' Z 1 1"kZ HD + Pt DZ +Pt , " , This suggested that k3 & k3 were much greater than k2, k2 , kZ' This
explains the large amount of DZ formed and a small amount of HD, and further the fact that the catalyst does not affect the para-ortho
conversion (69,70).
These results are just the opposite of that predicted by the mechanism for hydrogenase and that obtained from the hydrogenase-
catalyzed reaction in K. vulgaris. For if H:E- is not immediately
exchangeable with water, the HZ molecule interacts with the enzyme and
DZO and at most, in one reaction, returns to the gas phase as HD, i.e.
only one-half of the molecule will come to equilibrium with water. 17
Therefore, each molecule most react at least twice with the enzyme before it can yield D2. The experiments with~. vulgaris demonstrated that the rate of the conversion was actually three times the rate of
the exchange reaction (69,75).
Later investigations (8,64,77,78,79), indicated the participation of ferrous iron in hydrogenase action. In addition, it was found that
complexes of metal sulfides could catalyze the reduction of redox dyes
such as benzyl viologen (a common substrate of hydrogenase) and carry out a heterolytic activation of the H2 molecule (70). Incorporating these results in the earlier mechanism, Rittenberg proposed a more
detailed model involving a metal to combine with the hydride ion and a base to react with the proton (70). The active site of hydrogenase would then have the structure shown below:
I I Fe++ :B where B represents the base donated by the protein with its free pair
of electrons. A H2 molecule would then fit in to yield:
I ~(---7) H+(Water) H:B
The hydrogen atom attached to B must be able to exchange with the
hydrogen proton of water. At low pH, the Fe++ ion is present as such
and the base is neutralized as H:B; activity will be low. At a very
high pH, the base will free as :B but the iron will take up any
available OH- to form FeOH+ making it inaccessible to the H- ion. At
neutral pH, both sites will be relatively free to yield an active 18 hydrogenase. This offers one explanation why the optimal activity of hydrogenase is between pH 6.4 to 7.0 (5).
Krasna and Rittenberg did not account for the participation of co-factors and substrates such as the redox dyes, as they were concerned only with the activation of HZ. However, for the purposes of purification and for the study of the properties of hydrogenase as they occur in the organism, where the transfer of HZ to substrates is
an important reaction, dyes such as methylene blue are ordinarily
employed as electron acceptors. The question of how the transfer of
the activated hydrogen occurred was proposed by Peck, San pietro, and
Gest (71).
The methylene blue or HZ uptake reaction is represented as,
HZ + methylene blue ~ Methylene b1ue:ZH
The opposite or reverse reaction as developed by Peck and Gest (9), is
based on the evolution of HZ from reduced methyl vio1ogen a one
electron dye:
Z methyl vio1ogen-e- + ZH+ ~E--~) HZ + Zmethy1vio1ogen
A survey of hydrogenase preparations from a variety of organisms using
the foregoing assays disclosed that both activities were usually
present (9). Of particular significance, however, was the observation
that certain bacteria which catalyzed only the HZ uptake reaction also
exhibited a slow or nonexistent exchange reaction (77,64). These
results suggested that the hydrogenase may be a double-headed enzyme
with Z prosthetic groups; one mediating the electron transport to dyes 19 or physiological acceptors of relatively high redox potential, while the other might be concerned with reduction of "one-electron" acceptors of low potential involved in the formation of H2.
In addition to the requirement for iron, Shug et a1. (80) demonstrated that Mo03 and FAD were needed for the reduction of cytochrome c by a purified hydrogenase from £. pasteurianum. On the basis of the foregoing results, Peck et a1. (71) formulated the following scheme:
H2 + E(Fe++) +-i E(Fe++) :2H ~(-----"t') E(Fe++)-f1avin:2H L flavin methylene blue 2-vio1ogen-e + 2H+ where molecular H2 is activated by a ferrous enzyme (E), to form a reduced enzyme complex. The latter can be oxidized via a flavin complex by methylene blue and acceptors or similar redox potentials.
Alternatively, if molybdenum is properly combined with the enzyme, the
flavin:2H can be reoxidized by one electron (e-) acceptors such as the viologen dyes.
The reduction of the viologen dyes by H2 and the evolution of H2
from reduced substrates would require the complete system. Likewise,
the isotope exchange activity would be catalyzed by only the complete
system, since all systems, with a few exceptions as mentioned
previously, which produce H2 from physiological intermediates, also
show hydrogen evolution activity from reduced dyes, it would appear
that a metal component permitting one electron transfers is essential
in normal hydrogen formation by bacteria, and it is possible that 20 molybdenum is of significance in this regard.
All the enzymes from bacteria studied to date (81,82,83,84,73)
adhere to the overall scheme of Peck et a1. and the mechanism of
activation of Krasna and Rittenberg. More detailed work on this enzyme
has been limited as the enzyme is very labile to oxygen (5,85,86).
Further work had to await the development of more sensitive and milder
methods of experimentation.
The development of new procedures in examining the physical and
chemical properties of this enzyme in the past few years have rekindled
the interest in hydrogenase. Bone et a1. (87) found that the properties
of Hydrogenomonas ruh1andii hydrogenase can be changed by the addition
of Mn++ ions which cause a rearrangement or aggregation of the enzyme.
Also, Sadana and Rittenberg (88), working with D. desu1furicans,
discovered that the hydrogenase of this bacteria possessed different
and distinct pH optima depending upon the chemical reagents used to
activate it. This suggested the possibility of the existence of
multiple forms of the enzyme hydrogenase. Work in this laboratory
(unpublished results) has shown that the Arrhenius plot for the
evolution of hydrogen from reduced methyl vio1ogen by several
hydrogenase enzymes has a characteristic break with a change in slope
occurring at about 15°. This nonclassical Arrhenius plot could be the
result of an enzyme which exists in more than one form with different
activation energies (89).
Disc electrophoresis on polyacrylamide gel has been used to show
that the hydrogenase activity of £. pasteurianum is the result of six
active and distinct enzyme species (90). These hydrogenase species 21 were shown to have different electrophoretic mobilities on the gel column and were identified by their ability to reduce methyl vio1ogen in the presence of hydrogen gas. The reduced dye bands were correlated with protein bands after staining the gel with amido black in acetic acid.
The application of the gel technique towards the study of the hydrogenase system of ~. pasteurianum exhibited many advantages, and it was decided to extend this method to the study of other bacteria to see if the observations from this species were unique or whether they were common to other hydrogenase containing micro-organisms. To test the utility and reproducibility of the polyacrylamide technique, the hydrogenase isoenzyme band patterns from various bacteria were examined under a variety of conditions; different growth conditions, various methods of cell disruption, and different degrees of purity. The studies were also extended to particulate species whenever it was possible to gain some information on the relationship between the soluble and membrane bound hydrogenases and between bacterial species.
As a check on the polyacrylamide technique, starch gel electrophoresis was also performed.
The total hydrogenase activities of representative bacteria with respect to dyes of various redox potentials were tested to determine whether some enzymes possessed unusual properties or whether these reactions were common to all h~drogenases. These specificity studies were also extended to the hydrogenase band patterns on the gel to learn if all isoenzymes carried out the same dye reactions or whether certain electrophoretic species were different. 22
The molecular weights and shapes of three hydrogenase enzyme
systems were determined using the sedimentation method of Martin and
Ames (91), and the gel filtration technique of Andrews (92). These
experiments were performed to add to the present knowledge of shape and weight parameters of various representative hydrogenases (88,97,98),
and to aid in the taxonomy of these bacteria.
Kinetic assays were done on purified preparations of ~.
pasteurianum and~. butylicum in the presence of various components
thought to be necessary for hydrogenase action (71,80). These
"co-factor" requirements were tested with hydrogenases at various
stages of purity to see if the properties of the enzyme were altered
by purification procedures. In addition attempts were made, using a
variety of recently developed techniques, to isolate this enzyme in
pure form and to compare the properties of this highly purified enzyme
with that of crude hydrogenases.
To obtain some insight into the physical properties of the
hydrogenases, the Arrhenius activation energies of several hydrogenase
systems were determined. Polyacrylamide gel electrophoresis was
performed on these systems at the various temperatures used in the
determination of the activation energies to detect any change in
properties of the enzymes. MATERIALS AND METHODS
Materials
The compressed gases used in these studies were obtained from
Gaspro, Ltd., Honolulu, Hawaii. Bovine thyroglobulin, bovine fibrinogen, beef liver catalase, yeast alcohol dehydrogenase, bovine serum albumin, and horse cytochrome c were supplied by Sigma Chemical
Company, St. Louis, Missouri. Mann Research Laboratories, New York, supplied apoferretin, human gamma globulin, myog1obu1in, ovalbumin, and beef pancreas chymotrypsinogen.
The redox dyes benzyl vio1ogen and methyl vio1ogen were supplied also by the Mann Research Laboratories. Methylene blue was obtained from Conway Products Co., New York, and bromophenol blue from the
Allied Research Corp., New York. Blue dextran and all types of Sephadex were obtained from Pharmacia Fine Chemicals, Inc., Piscataway, N. J.
The redox dyes 1,1'trimethy1ene, 2,2'dipyridi1ium di-iodide, Eo = -560 mv, and 1,1'tetramethy1ene 2,2'dipyridi1ium di-iodide, Eo = -640 mv were obtained through the courtesy.of Dr. E. A. Ca1derbank of I.C.I.,
Jea1ott's Hill Research Station, Brackne11 Berks, England, and Dr. R.
C. Valentine of the University of California at Berkeley, California.
Methods
Growth of Bacteria and Preparation of Extracts
All bacterial cultures were harvested by centrifugation at middle
to late log phase, and the cell paste was washed in cold O.lM P04
buffer (pH 6.8) prior to preparation of extracts. Table I gives 24 reference to the source of the bacteria, the growth medium, and the method of preparing extracts containing hydrogenases. The methods used to obtain extracts were, (A) autolysis of 1.0 g of dried cells in 10 m1 of 0.1 M phosphate buffer (pH 6.8) according to Mortensen (1), (B) rupture of cells in a suspension (lg wet weight of cell paste in 5 m1 of buffer) by a French pressure cell (American Instrument Co., Inc.;
Silver Spring, Md.) or sonic treatment (by a branson Ultrasonic, model
L5-75 sonicator) for 1 to 2 minutes, and (C) disruption of cells by method (B) followed by solubilization of the hydrogenase associated with the particulate fraction of cells by a modification of the method
of Kondo et a1. (32). The material sedimented after centrifugation at
144,000 x g for 1/2 hour was washed in 0.01 M phosphate buffer (pH 7.5)
and suspended in the same buffer (2 m1s per gram of material). NaOH was added to adjust the pH to 8.0 and trypsin (0.1 mg per 10 mg protein
content) was added. The suspension was gassed with H2 stoppered, and
incubated at 37° for 1 hour. The mixture was then centrifuged at
32,000 x g for 15 minutes and the supernatant was designated as the
solubilized extract. All suspensions and extracts, except where
otherwise stated were maintained in 0.1 M phosphate buffer (pH 6.8),
frozen, and under H2. The criterion for solubility was non
sedimentation at 144,000 x g for 30 minutes.
Gel Electrophoresis
Polyacrylamide gels (7.5%) were prepared and run in a Tris
(hydroxymethy1)-aminomethane-g1ycine buffer system (pH 8.3) according
to Ornstein and Davis (3). The marker dye was bromophenol blue. 25
Electrophoresis of samples of crude extracts (500-1000 ~g) was carried out on gel columns (6 x 60 rom) at 2.5 milliamperes per column at room temperature. When the marker dye had traversed approximately three fourths of the gel column, the gel was removed from its glass tube and cut transversely at the position of the marker dye band. The ratio of the distance travelled by a hydrogenase band to that travelled by the marker dye is recorded as an Rf value.
The vertical slab gel technique of Raymond (4) was also used for the separation of single hydrogenase species except that the Ornstein and Davis formulation for small pore gels was used as well as the
Tris-g1ycine buffer. In experiments comparing the molecular weight of the enzymes, the concentration of acry1amide was varied from 6% to 14%.
To confirm the identify of each species of isoenzymes, appropriate sections containing the hydrogenase, were cut from the gel slab and placed in 0.1 M phosphate buffer (pH 6.8) with 0.1% sodium dithionite and a drop of methyl vio1ogen (14 mg/m1). The sections were homogenized in a Potter E1vehjem tissue homogenizer and rerun on a polyacrylamide disc gel. The hydrogenase band was then excised from the glass tube, developed and identified.
Starch gel electrophoresis was carried out according to the method
of Asuton and Braden (135). The assay for hydrogenase was done in the
same manner as in the polyacrylamide technique.
Sucrose Density Gradient Centrifugation
The method of Martin and Ames (91) was employed. A Buchler
Corporation gradient maker was used to prepare linear gradients of 5% 26 to 20% sucrose in 0.1 M phosphate buffer (pH 7.0). Sodium dithionite
0.05% and a drop of methyl viologen (14 mg/ml) were added to insure that the solutions were free of oxygen. Reduced methyl viologen exhibits a blue color and this color persisted throughout the experiment. Samples
(0.1 to 0.2 ml) were layered on top of the gradient at a protein concentration not exceeding 3%. Tubes were centrifuged at 35,000 to
38,000 rev/min in a Beckman SW 39 rotor using a Spinco Model L or
Beckman L2-65 ultracentrifuge from 15 to 38 hours at 4°, then punctured and 0.4 ml fractions collected for the polyacrylamide assay or 0.8 ml for Warburg manometric assay.
Beef heart cytochrome c with an S20w value of 1.9s (Altman and
Dittmer (7) and molecular weight of 12,500; rabbit hemoglobin with an
S20w value of 4.5s (Hasserodt and Vinograd (8) and molecular weight of
66,500; yeast alcohol dehyrogenase with anS20w value of 7.4 (6) and molecular weight of 150,000; were used as standards in separate
centrifuge tubes. Their position in the-gradient was determined by
optical density readings at 280 m~.
Gel Fitration
Cell-free extracts of various bacteria were passed through columns
of Sephadex G-lOO (1.6 x 113 cm) and G-150 and G-75 (1.6 x 100 cm)
prepared and eluted according to the method of Andrews (92) modified by
an increase in the molarity of the Tris-HCL buffer to 0.2 M and the
addition of 0.05% sodium dithionite and methyl viologen. The columns
were standardized for molecular weight estimations by determining the
elution pattern of blue dextran, beef heart lactic dehyrogenase, bovine 27 serum albumin, and bovine heart cytochrome c. These compounds were detected by their absorption of light at 280 m~.
A small amount, 1 to 2 mg of sodium dithionite, was added to each collection tube before elution to maintain anaerobic conditions after elution.
Enzymatic Assays
Except where otherwise stated in the text the hyrogenase bands on
the gel were located by immersing the gel after electrophoresis and its
removal from the glass tube, in a 0.25% solution of methyl viologen in
0.1 M phosphate buffer (pH 6.8) contained in a 12.5 x 1.5 cm test tube
fitted with a serum stopper (#1083, Bitner Corp.). The air was removed
and replaced by H2 by repeated evacuation and filling with H2 through a
22 gauge needle. If immediate reduction (the production of a blue band
at the site of hydrogenase) did not occur, a solution of dithionite
(5% in 0.1 M phosphate buffer) was titrated into the test tube until a
faint blue color persisted iri the methyl viologen-buffer solution. This
insured that traces of oxygen trapped within the gel had been removed
and the enzyme was in its reduced and therefore most active state (15).
Bands of reduced methyl viologen typically appeared within 20 minutes
after introduction of H2 or trace amounts of dithionite. Continued
reduction often caused bands to fuse and darken the entire gel column.
The presence of hydrogenase activity in different bacterial
fractions was checked manometrically by either the evolution of
hydrogen assay according to Peck and Gest (15) or by the reduction of
a suitable redox dye in the presence of H2' 28
The hydrogenase preparation, together with buffer of appropriate pH
(total volume of 3.0 m1), was placed in the main compartment of a
Warburg vessel of approximately 20 m1 capacity. A solution of 14%
NaOH (0.1 m1) was placed in the center well (with filter paper) to absorb any C02 given off in the reaction and 0.1 m1 redox dye (55 ~m/m1) in the side arm bearing the solid plug. The vessel was then attached to the manometer and flushed with the appropriate gas (H2 or He) for 10 to 15 minutes. Samples of cell-free extracts were then added to the gassing side arm, and the vessel was placed in the water bath (30°) and allowed to equilibrate while shaking. In the case of the assay measuring evolution of H2 0.1 m1 of sodium dithionite (250 mg/5 m1 0.2
M phosphate buffer pH 6.8) was added to the dye compartment before the
flask was placed in the water. After several minutes of gassing the vessels were closed-off to the atmosphere and equilibrated. At zero
time the contents of both side arms were tipped into the main chamber.
Assays of highly purified samples of hydrogenase were done in the
presence of 0.1% bovine serum albumin to offset any denaturation due to
dilution of the enzyme.
In the experiments involving the determination of the redox
potentials and specificities of various hydrogenases, all manometric
assays were run in 0.1 M phosphate buffer at pH 7.0. The pH was
checked after the completion of each experiment to determine whether
there was a change in the hydrogen ion concentration during the run.
All preparations that were to be used in several hours or days
were stored under H2 at 0°. Sodium dithionite and methyl vio1ogen, as
an indicator, were added to these preparations. The presence of reduced 29 methyl vio10gen, cysteine or mercaptoethano1 did not affect the hydrogenase band patterns, and a1iquots stored in this manner were stable up to a period of a week without any app~pciab1e loss in activity.
Arrhenius Activation Energy Determinations
The Arrhenius activation energy determinations for each of the bacteria studied were done in the following manner. Bacterial extracts were prepared just prior to the experiments and stored in separate vials under H2 at 0°. Enough material was prepared to supply the Lineweaver-
periodic checks on the activity at a reference temperature at 30°. This was done to insure that the enzyme activity was constant for each set of experiments (5°-35°) and that no change in enzyme activity occurred during the storage period. These experiments (for each bacteria) took
3-4 days to complete. If any change did occur a correction was made.
All experiments for each bacteria were repeated at least three times by
Warburg manometry. Substrate concentrations were varied over a range
of 5 to 55 ~moles for each assay at each temperature.
Protein was estimated by the method of Lowry et a1. (129). RESULTS
To test the utility of polyacrylamide gel electrophoresis in the study of hydrogenases, a survey of various representative bacteria was conducted. The effect of growth conditions, age of the cells, extent of purification, different methods of cell disruption and dyes of various redox potentials on the soluble and solubilized enzymes were studied to check the constancy of the hydrogenase band patterns, and its usefulness in solving taxonomic problems. Different strains of the same species were, also, tested as a further check on this technique.
A. Survey of Bacterial Hydrogenases
For the purposes of discussion it is convenient to group the hydrogenase containing bacteria according to the distribution of hydrogenase activity among soluble and particulate forms rather than by genus.
Hydrogenase band patterns of bacteria having more than 80% of their hydrogenase activity in a soluble form. Soluble hydrogenases were obtained from Vei11one11a a1ca1escens, Micrococcus 1acty1iticus,
Bacillus po1ymyxa, and 11 species of clostridia. Of these only ~. po1ymyxa, Clostridium tetanomorphum, Clostridium thermosaccharo1yticum,
Clostridium aurantibutyricum, and Clostridium f1avum possessed hydrogenase activity that could not be resolved into more than one band by disc electrophoresis. The Rf values of the hydrogenase bands which
constitute the band patterns of the various species are given in Table
II. Underscored Rf values refer to the bands that gave the largest TABLE I. SOURCE AND GROWTH CONDITIONS OF BACTERIA
Method of Reference to cell Species Source and strain growth conditions disruptiona
l. Azotobacter vine1andii ATCC 12518 (114) B,C
2. Bacillus brevis ATCC 999 (115) B
3. Clostridium acidi-urici Rabinowitzb (116) A,B
4. Clostridium buty1icum ATCC 14823 (104)C A,B
5. Clostridium butyricum ATCC 8260 (104) A,B
6. Clostridium fe1sineum ATCC 13160 (104) B
7. Clostridium fe1sineum McClung 541d (104) B
8. Clostridium fe1sineum McClung 638 (104) B
9. Clostridium fe1sineum McClung 644 (104) B
10. Clostridium fe1sineum McClung 539 (104) B
1I. Clostridium fe1sineum McClung 639 (104) B
12. Clostridium fe1sineum McClung 2822 A (104) B (var. vinaceum)
13. Clostridium histo1yticum ) Lyophilized cells Not known A obtained from ) Not known A w 14. Clostridium k1uyveri Worthington Corp. t-' TABLE 1. (Continued) SOURCE AND GROWTH CONDITIONS OF BACTERIA
Method of Reference to cell Species Source and strain growth conditions disruptiona
15. Clostridium pasteurianum ATCC 6013 (104) A,B
16. C10s tridium perfri~eIls Worthington Corp. Not known B
17. Clostridium sporogenes ATCC 7905 (104)e B
18. Clostridium tetanomorphum ATCC 3606 (117) B
19. Unclassified Clostridium FA 3679hf (104)e B
20. Desu1fovibrio desu1furicans ATCC 7757 (126) B,C
21. Escherichia coli ATCC 13706 (118) B,C
22. Escherichia coli ATCC 9637 (118) B,C
23. Escherichia S2!i ML 3g (118) B,C
24. Escherichia coli ML 30g (118) B,C
25. Escherichia coli Bg (118) B,C
26. Escherichia coli ATCC 4157 (118) B,C h 27. Escherichia S2!i K-12 (118) B,C
28. Micrococcus 1ysodei~ticus Worthington Corp. Not known A w N TABLE I. (Continued) SOURCE AND GROWTH CONDITIONS OF BACTERIA
Method of Reference to cell Species Source and strain growth conditions disruptiona
29. Micrococcus 1acti1yticus ATCC 14894 (127) B
30. Rhodomicrobium vannie1ii Bergerg (93) A,B,C
3l. Rhodospiri11um rubrum Bergerg (119) A,B,C
32. Vei11one11a a1ca1escens ATCC 12641 (128) B
33. Vei11one11a a1ca1escens ATCC 12642 (128) B
34. Clostridium haumanii ATCC 17791 (104) B
35. Clostridium roseum ATCC 17797 (104) B
36. Clostridium roseum McClung 653 (104) B
37. Clostridium f1avum ATCC 17790 (104) B
38. Clostridium aurantibutyricum ATCC 17777 (104) B
39. Rhodospiri11ium mo1ischianum Bergerg (119) B
40. Rhodospiri11ium spheroides Bergerg (119) B
4l. Rhodopseudomonas_ pa1ustris Bergerg (119) B
42. Chromatium warmingii Krasnaj (122) B k w 43. HydrQg~nomollas eutropha Repaske (120) B w TABLE I. (Continued) SOURCE AND GROWTH CONDITIONS OF BACTERIA
Method of Reference to cell Species Source and strain growth ccmditions dis~j.lQtiona
44. Hydrogenomonas faci1is Burris1 (121) B
45. Hydrogenomonas faci1is ATCC 11228 (121) B
46. Bacillus po1ymyxa ATCC 842 (124) B
47. Pseudomonas saccarophi1a Barberm (123) B
48. Photobacterium phosphorium ATCC 11040 (125) B
49. Photobacterium fischeri ATCC 7749 (125) B
50. Photobacterium harveyii ATCC 14126 (125) B
51. Mycoplasma Fo1someh Not known B
52. Clostridium rubrum ATCC 14949 (104) B
53. Unclassified Bioluminescent Octopus (123) B Marine bacteria
54. Unclassified Bio1ouminescent Squid (123) B Marine bacteria
55. Unclassified Bioluminescent Fish Stomach (123) B Marine bacteria
w .po TABLE 1. (Continued) SOURCE AND GROWTH CONDITIONS OF BACTERIA
aA, autolysis; B, sonication or French Pressure Cell; C, method B followed by the solubilization of particulate material according to Kondo et al. (2). bCulture supplied by Dr. Rabinowitz, University of California. cAdded to each liter of the usual medium of Carnahan and Castle (10) were tryptone 19 and yeast extract 0.5g. dCulture supplied by Dr. McClung, University of Indiana. eTryptone medium: Triptone 109, yeast extract 5g, K2HP04 1.25g, CaC03 19, glucose 109 per liter H20. FeS04 O.lg added aseptically. fCulture supplied by Dr. Frank, University of Hawaii. gCulture supplied by Dr. Berger, University of Hawaii. hCulture supplied by Dr. Folsome, University of Hawaii. iLactic acid substituted for ethanol. jCulture supplied by Dr. A. I. Krasna,. College of Physicians and Surgeons, Columbia University. kCulture supplied by Dr. R. Repaske, National Institutes of Health, Bethesda, Md. lCulture supplied by Dr. Burris; University of Wisconsin, Madison, Wis.
W \Jl TABLE II. HYDROGENASE BAND PATTERNS OF BACTERIA WHICH CONTAIN SOLUBLE HYDROGENASES+
Source and strain
I. C. buty1icum ATCC 14823 0.30 0.87
2. f.. buty1icum ATCC 14823 0.30 0.82
3. f.. butyricum ATCC 8260 0.83 0.94
4. f.. fe1sineum ATCC 13160 0.60 0.70
5. C. fe1sineum 638 0.60 0.72
6. C. fe1sineum 541 0.60 0.72
7. .£. kJ.uyveri Worthington Corp. 0.62 0.76
8. f.. .Easteurianum ATCC 6013 0.48 0.52, 0.57 0.68 0.77
9. f.. .Easteurianum ATCC 6013 0.48 0.52, 0.57 0.69 0.78
10. f.. tetanomorphum ATCC 3606 0.76
1I. M. 1acti1yticus ATCC 14894 0.92 0.96
12. V. a1ca1escens ATCC 12641 0.90 0.97
13. V. a1ca1escens ATCC 12642 0.86 0.92
+Solub1e hydrogenases accounted for at least 80% of the total hydrogenase actiyity of the cells.
w *underscored Rf value refers to most prominent reduction band. C' TABLE II. (Continued) HYDROGENASE BAND PATTERNS OF BACTERIA WHICH CONTAIN SOLUBLE HYDROGENASES+
Species Source and strain Rf va1ues+ -- 14. C. haumanii ATCC 17791 0.56 0.63
15. C. roseum ATCC 17797 0.52 0.67
16. C. roseum McClung 653 0.53 0.67
17. C. aurantibuytricum ATCC 17777 0.52
18. C. f1avum ATCC 17790 0.63
19. C. fe1sineum McClung 644 0.56 0.70
20. C. fe1sineum McClung 539 0.56 0.70
21. C. fe1sineum McClung 639 0.56 0.70
22. C. fe1sineum (var. vinaceum) McClung 2822A 0.46 0.59
23. C. rubrum ATCC 14949 0.55 0.69
24. B. po1ymyxa ATCC 842 0.71
w ...... 38 peaks when the reduced dye bands on the gel were scanned by a recording electrophoresis densitometer (Densicord model, Photovolt Corp.).
Examples of such tracings for Clostridium felsineum and Clostridium btityricum are shown in Figure 1. Band patterns for bacterial species used in this investigation are also represented schematically in Figure
2.
In order to ascertain the effect of different methods of extract preparation on the band patterns, cells of Clostridium butylicum and
Clostridium pasteurianum were extracted by several techniques; e.g.
French press, sonication, liquid nitrogen treatment (26), autolysis of dried cells (25), and acetone drying (16). All preparations maintained their respective bands regardless of the method of disruption.
In addition the hydrogenase of ~. butylicum was subjected to several fractionation schemes (15). The resulting preparations always had an increase in specific activity and at least one species, at
Rf = 0.70, appeared not to have been altered by the heat treatment, and protamine sulfate, alcohol, and ammonium sulfate fractionations.
The extracts of ~. pasteurianum and~. butylicum grown on a fixed nitrogen source possessed identical band patterns to those of cells grown under nitrogen fixing conditions (Table II).
No variation of hydrogenase band patterns could be observed during the growth of ~. butylicum and A. vinelandii. Identical band patterns of hydrogenase were observed in cells harvested from cultures only a few hours old and in cells harvested from cultures past peak log phase.
The total hydrogenase activity, however, reached a maximum in middle log phase and decreased towards late log phase (Figures 3 & 4). This 39
Figure 1
The relative activities of the hydrogenase forms that
comprise the band patterns of (a) £. felsineum and (b) £.
butyricum. Gels containing the hydrogenases were incubated
with methyl viologen in O.1M P04 buffer under an atmosphere of HZ. After development of reduced dye bands the gels were
scanned by a recording electrophoresis densitometer. 40
100 ...I I&J \!» C. felslnlurn 90 (!)z MIGRATION ) 80 ~ 0 en I&J 70 II: ... 60 0
... I&J 50 0: >- c( 0 ...en 40 II: I&J :II' 30 1 II: c( :2 l 20 J. 10 z 0 i= Q. II: 100 0 en m c( 90 C. butyrlcurn
80 ...I I&J (!) MIGRATION 70 ~ 60 :>g I&J 50 II! II: >- ... 0 40 0 II: ... I&J II: :II' II: 30 ~ c( en ::I 20 1 J, 10 41
Figure 2
Schematic representation of the band patterns exhibited
by the multiple forms of some bacteria. Bands were developed
on gel column as described in text. If methyl vio10gen
(colorless oxidized form) was used to locate the position
of the hydrogenase enzymes, colored (blue) bands of reduced
methy1vio10gen against a white background were obtained,
but if methylene blue (blue) was employed, colorless bands
of 1euco-methy1ene blue appeared against a blue background. 42
~.' vlnelandli Q; d8sulfurlcans
Soluble Solubilized Soluble Solubilized Soluble Solubilized
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
E vulgaris ~ rubrum ~ perfringens
Soluble Solubilized Soluble Soluble
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Figure 3
The relationship between growth and hydrogenase activity
of a batch culture of £. buty1icum. Samples of culture were withdrawn from the culture vessel, centrifuged at 1Z,000 x g
for 10 min. and the cells washed and resuspended at a final
concentration of 10 g cell paste per ZO m1 O.lM P04 buffer
pH 6.8. This suspension was sonicated for 1 min. and
centrifuged at 144,000 x g for 30 min. and the supernatant
tested for hydrogenase activity by the evolution assay of
Peck and Gest (9). 0---0, 10g10 density of culture;
specific activity of supernatant (~1 HZ evo1ved/hr/mg N). 44 4·5
Figure 4
The relationship between growth and hydrogenase activity
of a culture of A. vine1andii. Samples of culture were
withdrawn from the culture vessel, centrifuged at 12,000 x g
for 10 min. and the cells washed and resuspended at a final
concentration of 10 g cell paste per 20 m1. O.lM P04 buffer.
Hydrogenase activity was assayed for by the reduction of
methylene blue. 0---0, bacterial count per m1; x x, total
activity of hydrogenase per m1 of whole cells. 46
'IW I 0IJ'I:»Dg .... , _.,.- CD CD CD CD 52 2 52 2 .c M .c 'It It)" N 0 0 N
0 ~O CD
/0 0 ~ 0 / )( J ~ 0 N c( 0\
0 )(! ~ en 0:: :J 0 ::I: 0 ~ CD ~ 0 ~'"0,,, U) ~ 0v 0, 0 \0)("- N r 0, 0 N 0 en CD ,... CD It) 'It It) N
(~Ol x 'IW/,J4/~H salll If) A.II\I.:»\t eaouaDoJpAH ID.OJ. 47 type of variation in hydrogenase activity during the growth of the culture was similar to that found with A. agi1is (17) and H. faci1is'
(36).
Hydrogenase band patterns of bacteria having more than 50% of their activity in an easily sedimentab1e form. Bacterial extracts containing a high proportion of particulate hydrogenase were obtained from Escherichia coli, Desu1fovibro desu1furicans, Azotobacter vine1andii, Rhodospiri11um rubrum, Clostridium perfringens, Clostridium sporogenes, Clostridium histo1yticum, Chromatium warmingii,
HIdrogenomonas eutropha, Rhodopseudomonas pa1ustris, Rhodospiri11um rubrum and the photobacteria species. In addition to the particulate hydrogenases, extracts of each of these bacteria also contained a small amount of soluble hydrogenases which could easily be detected on the gel column. The band patterns of the soluble hydrogenases of these extracts are listed in Tables III and IV, together with the distribution of hydrogenase activity among soluble and particulate forms as determined in separate manometric experiments. Included also, where
the modified Kondo method (20) of solubilization proved effective, are
the band patterns of the solubilized particulate forms.
The efficiency of solubilizing the hydrogenase enzyme in the particulate fraction of cells by the solubilization method varied
greatly being higher than 90% for A. agi1is, ~. coli, and D.
desu1furicans and to almost zero for those of H. eutropha, £.
sporogenes, and a related but unclassified clostridium, PA 3679h. The
latter 3 species thus resembled the thermophile Clostridium
nigrifican~ (133) in this respect. No attempts were made to TABLE III. HYDROGENASE BAND PATTERNS OF SOME BACTERIA THAT CONTAIN BOTH SOLUBLE AND PARTICULATE HYDROGENASEa
Rf values of soluble Speciesb Cell fraction % total hyde and activity solubilized hyde forms
L A. vinelandii soluble 17 0.53 0.73
particulate 83 0.52 0.65
2. C. perfringens soluble 50 0.78 0.87
3. .£. histolyticum soluble 40 0.56 0.80
4. D. desulfuricans soluble 20 0.32 0.44
particulate 80 0.42 0.51
5. E. coli (ML-3) soluble 10 0.47 0.63
particulate 90 0.33 0.61 0.79
6. R. rubrumc soluble 9 0.45 0.58 0.65 0.73
7. P. vulgaris soluble 40 0.23 0.37 0.41 0.53
particulate 60 0.49 0.53 aparticulate hydrogenases solubilized by method of Kondo et ale (32). bAll species grown on a fixed nitrogen source.
~ CR. rubrum grown on the modified medium of Duchow et ale (93). (Xl
50
TABLE IV. Rf VALUES OF THE SOLUBLE AND SOLUBILIZED HYDROGENASE FORMS OF SOME STRAINS OF ,§,. COLI
a Strain Hydrogenase forms Rf values Grown semi-aerobica11yb
ML-3 soluble 0.47 0.63 solubilized 0.33 0.61 0.79
ML-30 soluble 0.47 0.63 solubilized 0.33 0.65 0.81
K-12 soluble 0.46 0.61 solubilized 0.33 0.60 0.80
ATCC 9637 soluble 0.50 0.62 solubilized 0.31 0.61 0.78
'B' soluble 0.50 0.58 solubilized 0.32 0.59 0.80
ATCC 13706 soluble 0.47 0.63 solubilized 0.32 0.59 0.80
ATCC 4157 soluble 0.52 0.6 0.80 solubilized 0.33 0.61
Grown aerobica11yc
K-12 soluble 0.49 0.62 solubilized 0.32 0.62 0.83
B soluble 0.45 0.59 solubilized 0.35 0.61 0.84
Cells of strain K-12 grown semi-aerobically and aerobically were disrupted by sonication (1 min) and centrifuged at 12,000 x g for 15 min. The evolution activities of the supernatants were 625 and 188 ~1 H2/hr/mg protein respectively. aA11 strains grown on a NH4+ nitrogen source. bCu1tures grown without aeration or agitation in 20 liter bottles. cCu1tures grown on Gyrotary shaker in 200 m1 lots in 1 liter flasks. 51 solubilize the particulate fractions of £. perfringens, £. histo1yticum and the photosynthetic bacteria. In the latter case, pre'vious investigators have also reported their failure to solubilize the particulate fraction (109,127).
In those instances where comparisons between the Rf values of the soluble and solubilized forms could be made, the bacteria always possessed one solubilized form with an Rf value identical to one of the soluble forms (Table IV). In addition other hydrogenase species, unlike the soluble forms, were also produced. In separate experiments the soluble hydrogenases of these bacteria were not affected by the solubilization treatment. The solubilized forms which do not have soluble counterparts may represent different isoenzymes of the particulate fraction or, they may be intermediates in the degradation process of the particulate material which results in the formation of the solubilized hydrogenase.
The soluble hydrogenase in crude extracts of £. perfringens were very susceptible to inhibition during electrophoresis. Oxidation of hydrogenase was indicated because the passage of 0.1 m1 dithionite solution (5%) through the gel by electrophoresis before migration of the extract prevented complete inhibition. Prominence and rate of band pa~ern appearance on the gel was enhanced by the addition of ferrous ions to the assay mixture of methyl vio1ogen. Loss of ferrous ions and corresponding drop in activity is not uncommon in hydrogenase preparations (34,35,30,1).
The hydrogenases of A. agi1is and R. rubrum do not couple with reduced methyl vio1ogen (9,29). These enzymes were detected on the gel 52 column by their ability to reduce methylene blue. After electrophoresis and removal of the gel from the glass tube, the oxidized form of this dye was first allowed to diffuse into the gel to stain it blue. In the presence ofH2' colorless bands appear at the site of reduction when the rate at which the dye is reduced to the 1euco-form exceeds the rate at which the oxidized form of the dye diffuses into the gel. The reduction bands for A. agi1is and R. rubrum are depicted in Figure 2.
Cultures of R. rubrum were grown under photosynthetic or aerobic conditions to examine the influence of different culture methods on the hydrogenase level and band pattern of this organism. The harvested cells were resuspended in 0.1 M phosphate buffer (pH 6.8) at a concentration of 6 g cell paste per 15 m1 buffer and the hydrogenase activity of whole cells and the soluble fractions from sonicated cells was determined by the reduction of methylene blue. Those cells grown under photosynthetic conditions possessed eight times as much hydrogenase activity as those grown chemo-synthetica11y in the dark
(Table V). Despite the difference in growth conditions and the levels of hydrogenase activity in the soluble extracts, all soluble extracts had identical band patterns.
R. rubrum was also grown semi-anaerobically in daylight on the medium of Duchow and Douglas (14) modified by substituting lactic acid for ethanol. A1iquots of the harvested cells in 0.025 M phosphate buffer (pH 6.8) were stored for 16 hours at 22° under helium or hydrogen in either the dark or light. The suspensions of whole cells and the soluble fractions of these cells were tested for hydrogenase activity at the beginning (to·give a control value) and end of this TABLE V. AFFECT OF GROWIH AND STORAGE CONDITIONS UPON HYDROGENASE LEVEL IN R. RUBRUM
Growth conditionsa Storage conditionsb
Total activityC Specificc Total activityC Specificc of whole cell activity of whole cell activity suspension of soluble suspension of soluble fraction fraction
Aerobic, light 1,670 1.86 Control 216,000 60.8
Aerobic, dark 1,800 2.82 Helium, dark 216,000 65.0
Photosynthetic 16,950 4.70 Helium, light 96,000 72.0
Hydrogen, dark 144,000 92.0
Hydrogen, light 96,000 27.5 aCu1ture grown on medium of Cohen-Bazire et a1. (119). bCu1ture grown on modified medium of Duchow and Douglas (93). Harvested cells stored in 0.025 M P04 buffer (pH 6.8) at 22° for 16 hours under conditions given in table. Light was supplied by 150 watt photof1ood units at a distance of 4' according to Koh1mi11er and Gest (127). CHydrogenase activity as estimated by the reduction of MeB is given as ~1 H2 absorbed/hr by extract and the specific activity as ~1 H2 absorbed/hr/mg protein.
IJ1 W 54 storage period,'the results being compared in Table V. It can be seen that the loss of activity was greatest in cells stored in the light.
However comparison of the activities of whole cells and specific activities of these soluble extracts did show that soluble hydrogenase were being retained at the expense of other proteins. Again, and despite the different storage conditions and losses in activity, the band patterns of the soluble extracts remained identical.
In experiments with A. vinelandii to assess the effect of nitrogen-fixing conditions on the level of hydrogenase in the cell, it was found that cells grown on a NH4+ nitrogen source had much less hydrogenase than those grown under nitrogen-fixing conditions (the specific activities of the soluble extracts were 48 and 560 ~liters of
H2 per hr per mg protein, respectively). The type of band patterns exhibited by the soluble and solubilized hydrogenase forms did not depend on the nitrogen source for growth. However, the proportion of soluble enzyme increased with total cell count (Figure 5) going from
4% to 23% in 80 hours. This increase may be due to a structural alteration of aged cells resulting in a greater release of soluble enzymes upon lysis. This change in structure may be an indirect result of lower oxygen tension within the medium. The rate of aeration was constant and as the cells increased in number, the amount of oxygen available to each cell was lowered, thereby decreasing the need for protection of the hydrogenase as a membrane bound form which is very stable to oxygen (109).
Similarly, vigorous agitation of~. coli cultures growing aerobically caused a four fold decrease in overall level of hydrogenase 55
Figure 5
The proportion of soluble hydrogenase against
A. vinelandii culture growth. Bacteria were counted during
growth using a Petroff Hausser counter (curve A). The activity
contribution by the soluble fraction expressed as a percentage'
of the total activity was also determined as a function of
growth time (curve B). ~he manometric reduction assay of
methylene blue was used (Materials and Methods). 56
'1W / Vlij3.LOVS
en en en en 0 Q 0 0
)( )( )( )( ~ rt) N 0 0 N
0 0 O~ en 0 0/ to
)( 0 / ~ )( 0 0 N « m
(I) \ )(I 0 et: 0 0 :::> 0 / ::J: 0 en b')( 0 ,,\~, to o )( 0 ", ~ o~~ 0 N 0 " )( 0 10 0 U') 0 N N 10 A.L IAI.L O\' 3190108 1 'flO! % 57 activity than usually obtained with cells grown semi-anaerobically.
However, there was an increase in the percentage of membrane bound activity. The band patterns of the soluble and solubilized enzymes were not affected (Table III).
To test the accuracy and consistency of the polyacrylamide technique towards the study of these isoenzymes, 7 strains of E. coli,
7 strains of £. felsineum, and 2 strains of V. alcalescens were tested for their band patterns (Tables II and V). No difference in hydrogenase band patterns among different strains of the same species were observed except for the variant strain of C. felsineum, 2822 A
(Table II). M. lactyliti~s possessed identical hydrogenase band patterns as that of the 2 strains of V. alcalescens, and this would support the reclassification of M. lactyliticus as a Veillonella species as suggested by Rogosa (108). This premise was extended to certain biolumi~escent marine bacteria which possessed hydrogenase activity when grown semi-anaerobically. Three unclassified species obtained from octopus, squid, and fish stomach, respectively possessed identical soluble hydrogenase band patterns. Their solubilized hydrogenases were similar except that the bacteria isolated from octopus possessed one more solubilized form than the other two (Table III). No gross differences were observed among these three species with respect to cultural characteristics, morphology, and physiology. Classical microbiological methods indicate that these organisms belong to the genus Photobacterium, but positive'identification could not be obtained by this method. A more positive identification of these species as
Photobacterium phosphorium was possible by examination of the 58 hydrogenase band patterns of the bacteria. Table III shows the similarity of P. phosphorium hydrogenase band patterns with that of the organism isolated from octopus and a tentative classification of these bacteria as ~. phosphorium is suggested. The other two micro-organisms may be a variant strain of this species since they differ from pO. phosphorium by not containing the 0.62 band upon solubilization of their particulate hydrogenases.
Hydrogenomonas eutropha yielded hydrogenase activity only when grown on an autotrophic medium. Upon gel electrophoresis soluble fractions yielded two hydrogenase bands (Figure 6). The two bands were identical with respect to redox dyes of various potentials (Table
VII). However, the slower moving band (Rf = 0.31) reduced NAD and its acetyl pyridine analog. The reduction of these compounds were determined in two ways; by excising the band after electrophoresis, and adding it to a cuvette with NAD or acetyl pyridine (0.1%). The cuvette was stoppered and gassed with hydrogen, and the reduction of the pyridine nuc1eotides was followed at 340 m~ on a Beckman DB spectrophotometer. The other method consisted of locating the NAD specific band by UV light; after electrophoresis of crude extracts, the gels were inserted into test tubes with buffer and the acetyl analog (0.1%), the tubes were stoppered, gassed with hydrogen, and incubated for 1/2 hour at room. temperature. The gel was then removed from the test tube and scanned with UV light; the band at Rf = 0.31 was located by its fluorescence. Protein stains of the gels revealed only 1 protein band at this RF. 59
Of all the bacterial hydrogenases investigated, the system of
H. eutropha is the only instance where there is a specificity
difference among isoenzymes of the same strain. This may be due to an
enzyme having dual specificities. Previous investigators have reported
the reduction of pyridine nuc1eotides by other hydrogenase enzymes
(130,131,132), but this occurred only in the presence of ferredoxin.
Bone (86), working with highly purified extracts of Hydrogenomonas
ruh1andii, observed the reduction of NAn by a hydrogen activating enzyme which he named hydrogen dehydrogenase. It was not determined whether
this reaction was catalyzed by a single protein or mediated by a
coupling factor, though it was known that ferredoxin did not participate
in the reduction process. It appears that the NAD reducing enzyme of
H. eutropha is similar to the hydrogen dehydrogenase of H. ruh1andii.
Cells of H. eutropha grown on a heterotrophic medium, using
various carbon sources, were devoid of hydrogenase activity. No
hydrogen uptake or hydrogen evolving capacities were detected in the
2 strains of Hydrogenomonas faci1is tested, grown autotrophica11y and
heterotrophically. This is in contrast to the findings of Schatz and
Bove11 (121) who demonstrated a very active hydrogenase in cells of
this species whe~ grown autotrophica11y and a less active hydrogenase
in cells grown in a heterotrophic medium. The cause of this
discrepancy in results is not known.
Organisms containing little or no hydrogenase activity. Certain
organisms known not to possess hydrogenase activity were tested for
band patterns as negative controls. They were the aerobes Bacillus
brevis, Micrococcus 1ysodeikticus, and the photosynthetic heterotroph 60
Rhodomicrobium vannie1ii grown on the medium of Duchow and Douglas
(93) modified by substituting lactic acid for ethanol. In addition,
two species of marine bacteria, Photobacterium harveyii, and
Photobacterium fischeri, and a species of pleuropneumonia like organism,
Mycoplasma, were tested. In each case no detectable hydrogenase bands
were observed with either methyl vio1ogen or methylene blue.
The metabolism of ~' acidi-urici would seem not to warrant a
hydrogenase (117) but Valentine (112) has reported trace activity in
cell extracts. In our experiments, no hydrogenase activity could be
detected by the reduction of methylene blue or the evolution assays.
"~~n addition no activity could be detected on a gel after the soluble ,j fraction of C. acidi-urici had been subjected to electrophoresis even
when procedures such as those employed with ~' perfringens were used,
i.e. passing dithionite through the gel prior to the C. acidi-urici
extract and adding Fe++ to the assay mixture. Disc electrophoresis of
extracts was also carried out at 5° as well as room temperature in the
case this particular hydrogenase was heat labile. The sensitivity of
the disc electrophoresis technique on polyacrylamide gels is such that
0.5 ~g of a purified preparation of ~' buty1icum with a specific
activity of 1 x 106 ~1 H2 per hr per mg nitrogen may be detected.
Using manometric techniques, 2.5 ~g of this protein appeared to be the
lower limit of detection, even when serum albumin was used to offset
any effects due to dilution.
Certain precautions had to be taken when the gel contained weak
hydrogenase activity and dithionite was used to ensure reducing
conditions prior to the development of band patterns (see Materials 61 and Methods). Under these circumstances prolonged exposure of the gel column to the colored reduced form of the vio1ogen and tetrazo1ium dyes sometimes resulted in the appearance of faintly colored bands apparently caused by non-specific staining of protein bands by the reduced dye. Braun (140) has shown that dyes with sulfhydryl groups of proteins and Miller and Miller (141) have found very strong bonds produced between N-methy1 amino dyes and proteins or nuc1eoproteins but not nucleic acids.
The weakly colored band resulting from protein 'tagging' by reduced dye and a band of reduced dye produced by hydrogenase action could be easily be distinguished since the former necessarily required the presence of reduced dye, was sharply defined and did not expand, and occurred under an atmosphere of hydrogen and helium. Hydrogenase band development on the other hand was complete within 15-20 minutes, the bands of reduced dye rapidly expanded and an atmosphere of hydrogen was a prerequisite.
As a check on the polyacrylamide technique, crude extracts of
C. Easteurianum and £. buty1icum were subjected to starch gel electrophoresis and the reduced dye bands were developed under hydrogen.
In each instance the same number of isoenzymes appeared on the starch gel as on the polyacrylamide disc gels with identical Rf's. 62
Figure 6
Hydrogenase band pattern of H. eutropha. extracts
obtained by sonic disruption of cells were run on
polyacrylamide gel columns as described in the Methods. Assays
for hydrogenase activity were done as des.cribed in Materials
and Methods. The reduction of NAD and its acetyl derivative was
followed spectrophotometrically or by the fluorescence of the
reduced forms on the gel when irradiated by UV light. 63
Hydrogenase Band Pattern of -_H. eutropha...... -
0.1 0.2 methylene blue, benzyl vlologen, methyl 0.3 vlologen, pyocyanloe perchlorate, 0.4 phenoslne meth08ulfate, phenosoffranine, 0.:5 0.6 tetrozollum re d, tetrazollum blue 0.7 0.8 0.9 1.0
O. I 0.2 0.3 NAD, Acetyl pyridine 0.4 0.5 0.6 0.7 0.8 O.g 1.0 64
B. Redox Potentials and Specificities of Various Hydrogenases
The total hydrogenase activities of various bacteria were examined to determine the relative efficiencies of their respective hydrogenases with respect to redox dyes of different potentials; to determine whether a common pattern of reaction was displayed by their hydrogenases, or, whether some enzymes had unusual dye specificities.
This investigation was also extended to the band patterns on polyacrylamide gel to learn if all isoenzymes had the same dye reactions or if there was any specificity difference with different compounds.
The bacteria were categorized by their reactivities with respect to various dyes rather than by genus or by the distribution of hydrogenase activity among soluble and particulate forms.
Hydrogenases of high specific activities. Relatively high hydrogen uptake activity and evolution of hydrogen gas from reduced methyl viologen were exhibited by cell free extracts of the faculatative and strict anaerobes (Table VI). Reduction activity was highest with methylene blue (MeB), Eo = + 10 mv, and decreased with decreasing redox potentials of benzyl viologen (BV), Eo = - 315 mv; methyl viologen (MV), EO = - 447 mv and - 511 mv; 1,1' trimethylene,
2,2' dipyridilium di-iodide, Eo = -560 mv; and to zero with 1,1'
tetramethylene, 2,2' dipyridilium di-iodide, Eo = - 640 mv. The hydrogenase of £. butylicum is the only exception with no activity with methylene blue. This variation of activity in this species is in
agreement with the observation of Peck and Gest (15). TABLE VI. SPECIFIC ACTIVITIESa OF VARIOUS HYDROGENASES WITH REDOX DYES (1 H2/hr/mg PROTEIN)
Species + 10 mv - 315 mv - 447 mv - 511 mv - 560 mv - 640 mv
l. C. roseum 3000 2100 1150 1100 15 0
2. C. rubrum 5200 2100 1530 1100 40 0
3. C. haumanii 9000 4420 1530 1100 40 0
4. C. f1avum 6750 6000 3700 3600 265 0
5. ~. aurantibutyricum 8000 6750 3800 3700 200 0
6. ~. pasteurianum 9000 9000 4900 4500 15 0
7. C. buty1icum 0-5 9000 5000 4800 268 0
8. ~. fe1sineum 538 7000 4500 4000 4100 60 0
9. C. fe1sineum 638 6750 4500 4000 4000 45 0
10. C. fe1sineum 644 7560 4900 4000 4000 50 0
1I. C. fe1sineum 2822A 6700 4500 4100 4000 40 0
12. ~. coli 3700 2840 1440 1400 10 0 (soluble)
13. E. coli 2400 2100 665 640 5 0 (particulate) VI'" TABLE VI. (Continued) SPECIFIC ACTIVITIESa OF VARIOUS HYDROGENASES WITH REDOX DYES (1 H2/hr/mg PROTEIN)
Species + 10 mv - 315 mv - 447 mv - 511 mv - 560 mv - 640 mv
14. P. vulgaris 4400 2100 1250 1250 10 0 (soluble)
15. P. vulgaris 4500 2300 1400 1300 10 0 (particulate)
16. B. polymyxa 4770 3000 1340 1000 5 0
17. D. desulfuricans 9000 8450 3800 3000 252 0 (soluble)
18. D. desulfuricans 12000 10000 6000 5500 300 0 (particulate)
aActivities were determined by Warburg manometry using the uptake of H2 gas reaction. Appropriate enzyme concentrations were used to give a rate of about 5 to 10 microliters per minute. Dye concentration was constant at 6.4 micromoles. Total volume of 0.1 M phosphate buffer, pH 6.8, was 3.2 mls. Gas phase was hydrogen.
C1\ C1\ 67
There was an abrupt drop (10-100 fold) in activity at Eo = - 560 mv. Little or no hydrogen uptake activity was observed with dyes of redox potentials equal to, or lower than - 560 mv. Hydrogen evolution activity from reduced dyes was observed only with methyl vio1ogen and its - 511 mv derivative (Table VIII). The hydrogen evolution reaction with the reduced form of methyl vio1ogen, - 447 mv was always greater or equal to that of the - 511 mv derivative. The evolution of hydrogen from reduced methyl vio1ogen was always higher than the reverse reaction (Hz-uptake) involving this dye, and this is in agreement with earlier observations (9). In the case of E. coli,
~. vulgaris and D. desu1furicans, there was no significant differences in the reactivities of the soluble and particulate hydrogenases with
these dyes.
Hydrogenases of low specific activities. The hydrogenases of the obligate aerobe, A. vine1andii, the chemoautotroph, H. eutropha, and
5 species of photosynthetic bacteria behave similarly to the anaerobes
in the HZ-uptake reactions except that the activities are much lower.
The drop in hydrogenase activity seems to occur at a higher redox potential (- 511 mv vs •. - 560 mv) than in the anaerobes. With the
exception of R. rubrum , H. eutropha, and chromatium warmingii, no hydrogen evolution from reduced methyl vio1ogen was observed with cell
free extracts of these organisms. Peck and Gest (15) reported no
evolution activity from reduced methyl vio1ogen in extracts of R.
rubrum (Table VIII); however, the cells used in our experiments were
grown in a modified medium of Duchow and Douglas (93), and possessed
10 times the activity with respect to methylene blue than previously 68 reported by these investigators. Hence the discrepancy in the results may be due to their inability to detect the low level of the hydrogenase reaction.
The hydrogenase band patterns on polyacrylamide gel of all species listed in Tables VI and VII were constant irregardless of the dyes used. The isoenzymes of each organism either appeared together or not at all. The reactivities of the various hydrogenases on the gel corresponded to the Warburg manometric results, and no specificity differences among isoenzymes of a particular strain were indicated. TABLE VII. SPECIFIC ACTIVITIES OF HYDROGENASES OF RELATIVELY LOW ACTIVITIESa (LITERS HZ/hr/mg PROTEIN)
Species + 10 mv - 315 mv - 447 mv - 511 mv - 560 mv
L H. eutropha 960 480 100 0 0 z. A. vine1andii 400 ZOO 50 0 0 (soluq1e)
3. A. vine1andii 500 Z10 50 0 0 (particulate)
4. ~. rubrum 1Z0 60 ZO 10 0 (soluble)
5. ~. speroides 40 15 5 0 0 (soluble)
6. R. mo1ischianum 40 15 10 5 0 (soluble)
7. R. pa1ustris 60 ZO 10 5 0 (soluble)
8. Ch. warmingii 150 50 30 20 0
aAssay conditions were the same as those in Table VIII.
0' \0 70
TABLE VIII. EVOLUTION OF HYDROGEN FROM REDUCED METHYL VIOLOGEN
Specific activitya Species (~1 H2/hr/mg protein)
1. C. roseum 1880
2. f.. rubrum 1800
3. C. haumanii 1900
4. f.. f1avum 4500
5. C. aurantibutyricurn 5000
6. f.. pasteruianum 4000
7. C. buty1icum 5200
8. f.. fe1sineum 538 4800
9. f.. fe1sineum 638 4500
10. C. fe1sineum 2822 A 4500
11. E. coli (soluble) 1500
12. E. coli (particulate) 480
13. P. vulgaris (soluble) 1900
14. P. vulgaris (particulate) 1500
15. B. po1ymyxa 2000
16. D. desu1furicans (soluble) 5500
17. D. desu1furicans (particulate) 6000
18. H. eutropha (soluble) 250
19. A. vine1andii (soluble) o
20. A. vine1andii (particulate) o
21. R. rubrum (soluble) 40
22. R. spheroides (soluble) o 71
TABLE VIII. (Continued) EVOLUTION OF HYDROGEN FROM REDUCED METHYL VIOLOGEN
Specific activitya Species (~l H2/hr/mg protein)
23. R. molischianu~ (soluble) o 24. R. palustris (soluble) o
25. Ch. warmingii (soluble) 50
26. C. felsineum 644 4500
aActivities determined manometrically by the evolution of hydrogen gas from methyl viologen reduced by sodium dithionite. Dye concentrations were standardized at 6.4 micromoles; sodium dithionite at 3.0 micromoles; total volume of 0.1 M phosphate buffer, pH 6.8 was 3.2 mls. Enzyme concentrations were adjusted to give a rate of 10 microliters of hydrogen evolved per minute. The gas phase was helium. Temperature was constant at 30°. 72
C. Molecular Weight Determinations of Various Hydrogenases
The molecular weights of three hydrogenase enzyme systems were determined by the technique of column chromatography and density gradient centrifugation. These experiments were done to add to the knowledge of the shape and weight parameters of various hydrogenases
determined by A. D. kidman (98). The determinations were done on
crude extracts of all three bacteria and on a purified extract of one
of the organisl~ to see if any change in properties occurred during purification of this enzyme.
It was necessary to maintain anaerobic conditions during gel
filtration and density gradient centrifugation. In the absence of
such precautions no hydrogenase activity could be detected. In each
of the two procedures sodium dithionite (0.05%), was used to eliminate
oxygen with methyl vio10gen as an indicator. To prepare linear
gradients of 5% to 20%, the appropriate amounts of sucrose were
dissolved in 0.1 M phosphate buffer (pH 7.0) with dithionite and
methyl vio10gen; the gradient acquired a blue color which persisted
throughout the experiment. The molarity of the buffer was of
sufficient strength to stabilize the pH at this dithionite
concentration. In gel filtration however, the molarity of the
Tris-HCL buffer had to be raised to 0.2 M.
In spite of these precautions much of the hydrogenase activity
was lost, 65-95% after gel filtration and 95-99% after sucrose
gradient centrifugation. The addition of ferrous sulfate or
mercaptoethano1 to the buffers did not increase the level of activity. 73
Sedimentation analysis. Sucrose density gradient centrifugation of crude preparations of £. pasteurianum, £. buty1icum, and £. butyricum was performed in a 5-20% linear sucrose gradient. Anaerobic conditions were ensured by addition of mineral oil to the top of each tube at the start of each run. Fractions eluted from the gradient tubes centrifuged in a SW 39 and SW 41 rotors were assayed by both
Warburg manometry and polyacrylamide disc gel electrophoresis. More than 95% of the total activity by methyl vio1ogen assay was lost routinely in the gradient experiments. Whenever the gel assays were used, the presence of activity was determined by analyzing each fraction and recording each isoenzyme as they appeared. The peak of activity was assumed to occur in the middle of a consecutive series of fractions, all of which contained on disc gel electrophoresis the same hydrogenase isoenzyme (Table X). Only one activity peak was observed with each of the three bacteria studied.
In the case of £. pasteurianum, only 3 of the 6 isoenzymes appeared in all fractions tested. These 3 forms possessed Rf values of 0.74, 0.64, 0.56, and were designated as ~, S, and y, respectively
(95). The predominant form, the y species, had an S20w value of 4.0s using hemoglobin as a standard (Table IX). The ~ and S species were not always present because of lower specific activities and the conversion of the ~ and S forms to the y species. To obtain the ~ and
S forms in these centrifugation studies, cells grown on a high iron medium possessing twice Ehe activity of the normally grown'ce11s were used. Even then, these 3 species never appeared toegether in the sedimentation and gel filtration studies; either the ~ and y bands or TABLE IX. S VALUES OF VARIOUS HYDROGENASE ISOENZYMES
Rf's bands in Rf's of bands under Average S20 ~ Speciel:L .._ crude ~~ej)aratj.ons_ .a.cJ:iyit.Y_..I>eaksa value f.. butylicum 0.30 0.70 0.70 5.8 f.. butylicum 0.30 0.70 0.70 5.8 (purified 450 fold)
C. pasteurianum 0.32 0.44 0.52 0.56 0.65 0.74 3.95 0.57 0.65 0.74
C. butyricum 0.58 0.65 0.65 0.58 4.0
aAssays for hydrogenase activity were done by both gel electrophoresis and Warburg manometry as described in Materials and Methods. bS values calculated by the ratio of the distance travelled by the hydrogenase to that of the standard multiplied by the S value of the standard.
-...J .po. 75 the Sand y appeared but never ~, S, and y, together. However, when the position of the ~ and S species could be located, they always had an S value that was within the 90% confidence range of the y form
(Table X).
A plot of the ratio of the distance moved during centrifugation by the unknown to that of the standard vs. rev./min.-time indicated that the S value of the y hydrogenase as referred to hemoglobin is constant and independent of the length of centrifugation. This treatment of the data according to Bergi and Hershey (94) suggested that the y species is the same shape as hemoglobin (Fig. 7). With cytochrome c a nonlinear plot was obtained at low rev./min for short runs, however, longer runs resulted in linearity.
In the case of the C. butyricum hydrogenase system, the two bands of the crude preparations appeared under one peak after density gradient centrifugation. These two forms possessed a similar S value to that of the hydrogenases of £. pasteurianum (Table X). The two isoenzymes of £. felsineum has also been reported to have similar S value and molecular weight.
The hydrogenase system of £. butylicum appeared to be different in this regard, as one of its isoenzymes had an S value of 5.8s and a molecular weight value of about twice that of the other clostridial enzymes (Table X). A highly purified preparation of this enzyme
(approximately 450 fold with respect to specific activity) yielded an identical S value and molecular weight (Table X). The size of the other isoenzyme of £. butylicum could not be determined. 76
Figure 7
Relative sedimentation pattern of the y isoenzyme.
The ratio of the distance moved during sucrose density
gradient centrifugation by the y hydrogenase during to
that against centrifugal force by time of run expressed
as rev./min-hr. X X hemoglobin standard,
0 0 cytochrome c standard...... -0- -ON a:0 2.0 ~ 0 ~ 1.8 o 0 ~ (/) ~O 0 o 1.6 o ~ cr '1 HYDROGENASE/CYTOCHROME-C l"2 1.4 >- Nz 1.2 l" >- CD 1.0 0 x XXXX l" X X x X >0.8 -x 0 '¥ HYDROGENASE IHEMOGLOBIN ~ l" 0.6 0 z ~0.4 en Q 0.2
1L. 0 - I , I I , 0 2 3 4 5 4 -~ ° CENTRIFUGAL FORCE (Rev./Min.)· Hr.' 10 ~ -I a:
---.J ---.J 78
An estimate of the molecular weight of small enzymes can be made using the following relation (96):
2/3 = Mol. Wt'l (1) Mol. Wt'2 provided the molecule is assumed to be spherical and of partial specific volume of 0.725. The results are shown in the following table.
TABLE X. MOLECULAR WEIGHTS OF HYDROGENASE ENZYMES DET£RMINED BY SEDIMENTATION DATA
Rf of Species hydrogenase Ave. S value Ave. MW £. buty1icum 0.70 5.8 100 000 £. buty1icum 0.70 5.8 100 000 (purified) £. butyricum 0.56 0.65 3.9 53 000 £. pasteurianum 0.57 4.0 56 000 0.64 3.8 51 000 0.74 4.1 59 000
Gel filtration. Samples of crude extracts of £. buty1icum,
~. butyricum, and ~. pasteurianum were added to Sephadex G-75 and
G-100 columns. The elution volume of the activity peaks were
independent of the protein concentrations used in these experiments.
The results are summarized in the following table. 79
TABLE XI. MOLECULAR WEIGHTS OF VARIOUS HYDROGENASES DETERMINED BY SEPHADEX GEL FILTRATION
Rf of Species hydrogenase Ave. MoL Wt. Gel size
C. buty1icum 0.70 125 000 G-100 0.70 150 000 G-75 .£. pasteurianum 0.56 53 000 G-75 0.64 54 000 and 0.74 50 000 G-100
C. butyricum 0.56 0.66 60 000 G-75 0.56 0.66 56 000 G-100
C. buty1icum 0.70 120 000 G-100
All fractions were assayed as in the centrifugation studies and as
expected, the analysis of these fractions by gel electrophoresis
confirmed the presence of each species at an elution volume
corresponding to the molecular weights determined by the sucrose
density gradient centrifugation experiments (Figures 8 and 9).
As in the case of the sedimentation studies, the 0.74 band of
.£. pasteurianum appeared only in high iron extracts, and the 0.74 and
0.64 bands never appeared together. Each of these forms, however,
appeared in conjunction with the 0.56 band. The hydrogenase activities
in the three species studied were found under only one peak as in the
sedimentation studies (Figure 10). This observation agrees with the
results of Tamiya et a1. who have shown that .£. pasteurianum
hydrogenase was eluted as a single peak; but gave no estimate of the
molecular size (97). 80
Figure 8
Plot of elution volume against log molecular weight., The
column (1.6 x 97.5) contained Sephadex G-75 equilibrated with
0.2 M Tris-Hel buffer (pH 7.5). Flow rate was 9 ml/h. Standards
were eluted separately; cyto. c and myoglobin were determined
by optical density measurements at 408 m~; the remainder were
detected by absorbance at 280 m~. Hydrogenase activity was
detected by the manometric technique measuring the evolution
of hydrogen gas from reduced methyl viologen. Methyl viologen
concentration was standardized at 5.5 micromoles and reduced
with 3 micromoles of sodium dithionite. A volume of 0.4 mls
was used for these assays. Total volume of the assay mixture
was 3.2 mls. The gas phase was helium. 170 Cytochrome C (12,500) 160
150
140
130 Chymotrypsinogen (25,000) 120
-10 E 110
100 ." Ovalbumin (45,000) ell E hydr~enase~ ::s .£.. butxricum x 90 ( 66,500) >0 C. pa5teurianum hydrogenase ----+ x 80 c 0 ";:i ::s 70 Blue ~ 60 Dextran
50 • 2 3 4 :5 (,7 .. I 10 20 30 40 :50 60 100 200 1000 Molecular Weight X 10-3
co I-' 82
Figure 9
Plot of elution volume against log molecular weight.
The column (1.6 x 111.5) contained Sephadex G-IOO equilibrated
with 0.2 M Tris-Hel buffer (pH 7.5). Flow rate was 8 ml/h.
Standards were eluted and detected as described in Fig. 8. 160
150 ",yog'ob;. ( 17, 800) 140 "'-0 ~ .. 130 Chymotrypsinogen (25,000) E 120
110 ~o I Oval'um'. (45, 000 ) =' ~ 100 c Co ~ 0 ._,,".um ""ag,.a., 90 -.= .£:. butyricum hydrog.nose----t x O~BoVine Serum Albumin (66,500) IIJ 80 C. butylicum hydrogenase - ...... x Beef Heart Lactic Acid o /0 Yeast Alcohol Dehydrogenase 0...... Dehydrogenase ( 135,000 ) ( 150,000) ) 60 Blue Dextran 50 o 2 3 4 5 6 7 8 9 10 20 30 40 60 80 100 200 1000 Molecular Weight X 10-3
co w 160
150 Myoglobi' ( 17, 800] 140 "'0 • 130 ""0 Chymot,yp,;,oge, (25.000) E 120
~ 110 ~o (45. ::J Ovolhomin 000 I "0 > 100 c C. p""..,'o,om hyd,ogeno,e ~ ~ 90 .=- .£:. butyricum hydrogena,se--l x O~BoVine Serum Albumin (66,500) LLI 80 C. butylicum hydrogenase~ - x Beef Heart Lactic Acid o 70 Yeast Alcohol Dehydrogenas,: 0...... Dehydrogenase ( 135.000 ) ( 150,000) ) 60
50 o 2 3 4 5 6 7 8 9 10 20 30 40 60 80 100 200 1000 Molecular Weight X 10-3
00 w 84
Figure 10
Elution pattern of the hydrogenases from~. buty1icum,
C. pasteurianum, and~. butyricum from Sephadex G-100. The
relative activities from the column are shown. The elution
volumes correspond to a molecular weight of 125 000, 60 000,
and 64 000 respectively. Hydrogenase activity was detected
as described in Fig. 8. 85
2000
x
c. post euri Q num hydrogenase
1600
- E ...... : 1200 .c ...... : x )( ~
C. butyllcum > -u 800 nO··aI~ oCt h'd,o••
.,Q) c r:: CI co o ~ 'C x >0 X 400
\Jx / j\ >~"","m
,!-'L:."_-L"";o:::-.J...'_-L_---1.1-:..:,)(---II--!.=.·:...... ::0:...... J1c·=-=..::...·11--_.1-_.1-'::""..1.-_..1.-__ 60 70 80 90 100 110 120
Elution Volume ( mls.) 86
The two isoenzymes of C. butyricum were also eluted as a single peak and at an elution volume corresponding to a molecular weight value of 56 000 to 60 000 (Table XI). The hydrogenase isoenzyme of C. buty1icum obtained from crude and purified extracts was also eluted as a single peak at a molecular weight value of 125 000.
Siegel and Monty (99) reported the calculation of molecular weights and frictional ratios of proteins in impure systems from data obtained by Sephadex gel filtration and density gradient centrifugation. Their work confirmed the mathematical correlations by Ackers (100) and Laurent et a1. (101) relating the elution position of a macromolecule with its Stokes radius. This is the radius of the equivalent sphere calculated from the diffusion constant using Stoke's formula (102) to obtain a general measure of the size of each protein.
The molecular radius of any protein can be determined by calibrating a gel column with an excluded macromolecule, proteins of known Stokes radii and by knowledge of the amount of Sephadex used. This can be expressed mathematically as,
1/2 (2) (-log Kav) = 0: (8 + a)
a = Stokes radius, 0: & 8 are constants
(3) Kav = Ve Vo Vt Vo
where Ve = the elution volume corresponding to the peak concentration of a solute, Vo = void volume of the column, and Vt = the total volume of the gel bed.
Plots of (-log Kav)1/2 against Stokes radius for various protein standards from Sephadex G-75 and G-100 columns have been done by 87
A. D. Kidman (98) and are shown in Figures 10 and 11. With the Stokes radius calculated in this manner and the sedimentation coefficients determined in the foregoing experiments, an accurate estimate of the molecular weight and of the shape of a protein in impure fractions can be obtained by the following equations.
1/3 (4) M = 67TnN a s f/fo = a/(3vM/47TN) (1 - vq) where n is the viscosity of the medium, N is Avogadro's number, a is the Stokes radius, s is the sedimentation constant, v is the partial specific volume, and q is the density of the medium, and f/fo is the frictional ratio. The diffusion constant (D) is also related to the elution position of macromolecules from a Sephadex gel column (103), and is related to the Stokes radius by the equation:
kT (5 ) D = 67Tna where k = Boltzmann's constant; T = the absolute temperature; n = viscosity of the medium.
To check the molecular weights obtained by the sucrose gradient centrifugation and those obtained by gel filtration, the Stokes radius and diffusion coefficients of the various hydrogenases were calculated by A. D. Kidman (98). The molecular weights of the hydrogenases estimated by use of the Stokes radii and sedimentation constants are shown in the following table and agree reasonably well with those obtained through gel filtration or sedimentation alone. 88
Figure 11
Correlation of Kav with Stokes radius. The gel
filtration data of standard proteins shown from the Sephadex
G-75 column were plotted according to the correlation of
Laurent and Killander, Equation (2) in the text. 89
1.0
0.9 Bovine Serum Albumin • C. bllly,lcum hYd,ogen~7
0.8 C. pasteurianum hydroQenase x
Ovalbumin ./ 0.7 ::::-~- ClI-> ~ 0.6 Chymot,ypslnogen / CIt 0 ..J I /T'YPSln - 0.5 ./ ~YOglObln /
0.3
4,1-1 -I-__L--_--a..__.....L-__.L--_---L.__...J.-_ 5 10 15 20 25 30 35
Stokes Radius )( 108 em. 90
Figure 12
Correlation of Kav with Stokes radius. The gel
filtration data of standard protein shown from the Sephadex
G-100 column were plotted according to the correlation of
Laurent and Killander, Equation (2) in the text. 91
1.1 Yeast Alcohol Dehydrogenase • c. butylicum hydrogenase
1.0
0.9 Bovine Serum Albumin
;::.N c. butyricum hydrogenase --; c 0.8 x/Xc. pasteurianum hydrogenase ~
CI / . 0 OvalbumIn ..J /0 - 0.7
/. Chymotrypsinogen 0.6
0.5
0.4
~ ,1-'--1-_-J.-_--1-_...... L-_...I.-_.l-.--_ 20 25 30 35 40 45 50
Stokes Radius x 10 8 TABLE XII. MOLECULAR WEIGHTS OF HYDROGENASES FRO}! GEL FILTRATION AND SEDIMENTATION DATA
Species MW Stokes radius (10-8) fifo D2Ow(10-7) S20 w
C. pasteurianum 50 000 30 1.2 7.1 4.0
C. buty1icum 100 000 44 1.4 4.8 5.8
C. butyricum 53 000 33 1.3 6.5 3.9
""N 93
D. Study of the 3 Isoenzymes of C. Pasteurianum
To gain further information on these hydrogenase systems, Q. pasteurianum was chosen as it possessed a number of well-defined isoenzymes which appeared regularly in crude preparations. The effect of growth conditions, different gel concentrations and storage methods, were investigated in hopes of adding to the present knowledge of hydrogenases.
Inter-relationship studies-on C. pasteurianum. The three most active hydrogenase species of Q. pasteurianum were the ~, S, and y bands of Rf's on 7.5% gel of 0.74, 0.64 and 0.55, respectively. Their relative activities were determined by measuring their relative intensities on the gel of reduced methyl viologen by a recording electrophoresis densitometer. Their activities were found in the ratio of 1.5: 0.3 : 5.0. This ratio could be the relative proportions of the enzyme forms if theiLaffinities are identical and no substrate inhibition occurred. The y species exhibited the highest activity as it was the only form that could be measured manometrically and on the gel'it appeared first and spread rapidly.
Each species was obtained by use of the Raymond gel slab after excising, homogenizing and extracting with 0.1 M phosphate buffer pH
7.0. The identity of each form was checked by immediately rerunning the individual extracts on a 7.5% polyacrylamide disc gel.
In order to show that the isoenzymes were not an artifact due to the polyacrylamide gel, a preparation was subjected to electrophoresis on starch gel as described in the Methods. The three species were also obtained on this medium. 94
Storage of the 3 forms at 4°C under hydrogen overnightat pH 7 and pH 10 resulted in the complete conversion of the ~ and S forms to the y species. Storage at pH 4 resulted in loss of all activity as determined by the gel assay which is about 10 times more sensitive than the manometric technique in detecting the presence of hydrogenase activity (Figure 13). B-mercaptoethano1 (0.7 - 1.5 M) and 0.5% dithiothreito1 did not affect the conversion and 8 M urea, 1 M guanidine hydrochloride, 1 M sodium chloride eliminated all activity.
Th~ y~species appeared to be the most stable form to which the ~ and S species converted. This predominant form could not be converted to the ~ and S species. It should be emphasized at this point that the assays employed in these studies were mainly qualitative and depended upon the presence of active hydrogenase forms.
Iron and hydrogenase activity in C. pasteurianum. The
concentration of added iron (FeS04) in the growth medium and its
effect on the specific activity of hydrogenase is shown in the
following table: 95
TABLE XIII. EFFECT OF IRON ON HYDROGENASE ACTIVITY OF C. PASTEURIANUM
FeS04 (mg/m1) Specific activity of hydrogenase added to growth medium (1l1 H2 evo1ved/hr/mg protein)
o 1.1 x 10 3
25 2.1 x 10 3
3.3 x 10 3
125 3.5 x 10 3
250 7.0 x 10 3
500 4.2 x 10 3
2500 3.0 x 10 3
aConcentration recommended by Carnahan and Castle (104). Manometric assays were performed as described in Table VIII. 96
Figure 13
Effect of storage on the isoenzyme. The section with the
appropriate Rf was excised from the gel slab. Each section
containing an individual species was placed in 0.1 M phosphate
buffer (pH 7.0) containing 0.1% sodium dithionite and a drop
of methyl vio1ogen and homogenized. The homogenized gel
particles were removed by centrifugation and the supernatant
containing the appropriate hydrogenase isoenzyme was rerun on
a polyacrylamide disc gel immediately to confirm its identity.
Then the supernatant was stored under hydrogen for 12 h
at 4° at the indicated pH and reexamined by the gel assay. 97 Storage. 4· Section Immediate 12 hours at Rf Rerun pH 7 pH 10
0.1 0.2 0.3 0.4 0.5- 0.6 0.5 r ,. 0.6 0.7 0.8 0.9 1.0
0.1 0.2 0.3 0.4 no activity 0.6 - 0.7 0.5 r 0.6 8 0.7 0.8 0.9 1.0
0.1 0.2 0.3 0.4 0.7 - 0.85 0.5 r 0.6 0.7 0.8 0.9 1.0 98
The optimum concentration of iron appeared to be 5 times the iron concentration recommended by Carnahan and Castle (104). The band patterns of these "high" iron extracts were identical with that of the preparations obtained from normal cells. However, the ~ and y forms appeared with greater intensity in high iron preparations, and the y species spread so rapidly that it obscured the S band on the gel. As in extracts obtained from cultures grown at normal iron concentrations, the y form seemed to be the most stable and predominant and changing the iron concentration did not alter the conversion of ~ and S to y.
Gel concentrations and electrophoretic mobilities. The individual species (~, S, y) obtained from the Raymond gel slab procedure as described previously, were then subjected to disc electrophoresis at polyacrylamide concentrations varying between 6 and 14% on small gels.
Components of the same size but different charges should migrate proportionally giving a straight-line relationship with changing gel concentrations. Components of different sizes will be affected by increasing amounts as the polyacrylamide concentration increases; these components falling on a curved line. The ratios of the Rf's of the ~ and S bands to that of the y band remained constant over the whole range of gel concentrations used. A plot of the logarithm of the Rf
~ gel concentration with these forms is shown in Figure 14, and parallel lines were observed. This suggested that the 3 species possessed the same molecular size but differed in resultant charge since a log plot of the Rf's of components of different sizes versus gel concentrations would exhibit convergence of the lines at some point (105). 99
Figure 14
Relative mobility of the individual species is plotted
against gel concentration. The individual species were excised
from a gel slab, placed in 0.1 M phospha~e buffer (pH 7.0)·
containing 0.1% sodium dithionite, homogenized and centrifuged.
The supernatant was immediately rerun on 7.5% polyacrylamide
disc gels to confirm the identity of each species. Then the
individual species were run on polyacrylamide disc gels varying
in concentration from 6% to 14%. The log of the Rf ~ gel
concentration was plotted for the individual ~, 8, and y
isoenzymes. 100
.)( :!
o • )( )(
-~ Q - c 0 0 • ; ..0 en c -cu CJ c 0 (.) co
~ 0 • (!) £ .....
CD 0 )(
~ ~ To 0 0 0 0 0 0 0 C\I ", V II) 0 0 0 0 0 0 I I I I~ D01 101
E. Purification of C. Buty1icum Hydrogenase
Purification procedures. As hydrogenase has never been successfully purified by previous investigators (15,88), it was decided to attempt a fractionation of this enzyme using recently developed techniques including polyacrylamide gel electrophoresis. As most of the work done on hydrogenases were performed on crude or partially purified preparations, it is important to determine whether there is any change in properties of a highly purified hydrogenase enzyme. £. buty1icum was chosen for this work as it possessed a very active enzyme form and Peck and Gest (15) had previously obtained about a 300-fo1d purification of this enzyme and reported that it was stable.
The hydrogenase of C. buty1icum was purified by following a modification of various procedures used by several investigators for the partial purification of hydrogenases (15,88). A flow sheet summarizes this method. The crude preparation was obtained by either the autolysis of dried cells (1 g dried cells per 5 m1 of 0.01 M phosphate buffer pH 6.8) under H2 or by the sonication of cells (2 g/5 m1 buffer) for 1.5 minutes and by centrifugation at 32,000 x g for 15 minutes. The supernatant, designated as the crude extract, contained
50-60 mg protein per m1 and exhibited a deep brownish-yellow color.
DEAE-ce11u1ose prepared according to Sober (110) and washed with
0.01 M phosphate buffer pH 7.0 was added (5 g of powder/50 m1 extract) to the crude extract under H2 . The suspension was stirred for 15 minutes and centrifuged at 15,000 x g for 10 minutes. The precipitate was discarded and the pH of the supernatant was adjusted to pH 5.6 with acetic acid (7.5%) before heating in a water bath at 60° under H2 for 102
PURIFICATION SCHEME FOR Q. BUTYLICUM HYDROGENASE
DISCARD Crude prep. ~ I _. DEAE-ce11u1ose, pH 7.0, 0.01 M phosphate buffer ~ 15,000 x g 10 minutes ppt. supernatant Adjust to pH 5.6 Heat 60° under H2 for 10 minutes, cool. 15,000 x g 10 minutes ppt. supernatant Adjust pH 5.0, add 0.2 vol. absolute ethanol cool to 0° ~ 15,000 x g 10 minutes supernatant ppt. add 1/3 vol. original extract volume of 0.05 M phosphate buffer pH 6.4; add 1.2 volumes ~ saturated (NH4)2S04 ; 15,000 x g 10 min. ppt. supernatant add 0.8 vol. saturated (NH4)2S04 15,000 x g 10 minutes
dialyze in 50 volumes 0.05 M acetate pH 5.0 for 6 hours; 15,000 x g 10 minutes ppt. supernatant adjust pH to 6.0 add 5% total concentration of protamine sulfate 15,000 x g 10 minutes
PPt.~upernatant add 10% total concentration protamine sulfate 10,000 x g 10 minutes supernatant ppt. dissolve in 1/10 vol. 0.05 M Tris-HC1 pH 8.0 add 1 mg/m1 phosphocellulose powder stir under H2 1/2 hour 32,000 x g 15 minutes ppt. supernatant Stable under hydrogen at 0° in presence of 0.02% dithionite. W3
10 minutes. Failure to lower the pH resulted in a loss of activity.
The heated extract was then cooled to 4° and centrifuged at
15,000 x g for 10 minutes. The precipitate was discarded and the pH of the supernatant was lowered to pH 5 with 0.25 M acetic acid and
0.25 M acetate buffer (pH 5.0) was added to restore the preparation to
its original volume. The suspension was then cooled to 0° in an NaC1
ice bath and 0.2 volumes of absolute ethanol was added. The mixture was allowed to stand for 10 minutes and centrifuged for 10 minutes at
15,000 x g. The supernatant was discarded and the precipitate was
brought up to one-third the original extract volume with 0.05 M
phosphate buffer (pH 6.4).
Saturated ammonium sulfate (1.2 volumes) was added and the
suspension was allowed to stand for 20 minutes at 4°. The mixture was
spun at 15,000 x g for 10 minutes and the precipitate discarded. The
hydrogenase band of Rf 0.30 was removed at this stage of purification.
Addition of ammonium sulfate seemed to stabilize the hydrogenase as
preparations at this stage of fractionation could be stored up to 24
hours without appreciable loss of activity. To the supernatant of the
first ammonium sulfate addition was added another 0.8 volumes of the
saturated solution, and mixture was allowed to stand for 15 minutes at
0° and centrifuged at 15,000 x g for 10 minutes. The supernatant was
discarded and the precipitate was dissolved in one-fifth original
extract volume and dialyzed against 0.05 M acetate buffer pH 5.1
containing 0.01% dithionite under H2' After 4-6 hours a precipitate
was formed and this was centrifuged and discarded. 104
The supernatant was then adjusted to pH 6.0 and treated with protamine sulfate to a final concentration of 5% and allowed to stand for 15 minutes. The precipitate was discarded after centrifugation at
15,000 x g and the supernatant was further treated with protamine sulfate to a concentration of 10%, allowed to stand for ZO minutes, and centrifuged at 10,000 x g for 10 minutes. The final precipitate was dissolved in one-tenth original volume of 0.05 M Tris-He1 buffer pH
8.0. Phosphoce11u10se powder (1 mg/m1) was then added to the suspension
and stirred under HZ for l/Z hour in the presence of sodium dithionite
(0.1%) and centrifuged for 15 minutes at 3Z,000 x g. The final
supernatant was stored under HZ at 0° in the presence of a drop of
dithionite (0.1%). The hydrogenase preparations at this stage
represented a 400 to 600 fold purification (Table XIV) and were stable
for up to a week without any appreciable loss (less than 5%) in
activity.
-When disc electrophoresis on polyacrylamide gel were performed on
these purified extracts a single, very active band appeared at Rf 0.70.
The gels were then stained for protein with 0.1% amido black in 7.5%
acetic acid and at least 10 protein bands were observed; the weakest of
which was that of the hydrogenase at Rf 0.70 (Figure 15).
Attempts to further purify this species of hydrogenase have so
far failed. DEAE-ce11u10se column chromatography as employed by Sadana
and Rittenberg, who have isolated a highly purified form of hydrogenase
from D. desu1furicans, was used in the attempt to fractionate this
hydrogenase starting with crude and highly purified extracts. A pH
range of 4.5 to 8.5 and a molarity range of 0.005 to 0.3 was tested TABLE XIV. PURIFICATION OF C. BUTYLICUM HYDROGENASE
Protein Volume Specific activity Total activity % activity Fraction (mg/m1) (m1) (v1/hr/mg protein) (J11/hr/ml) recovered
Crude 60 40 10 000 600 000
Heat 20 30 25 000 500 000 60
Alcohol ppt. 5 13 100 000 500 000 27
Ammonium sulfate 1.5 10 300 000 450 000 17
Final extract 0.1 4 6 200 000 620 000 10
Hydrogenase activity was determined as described in Table VIII.
I-'o VI 106
Figure 15
Polyacrylamide gel assay on the final extract of the
purification scheme for C. buty1icum. A total of 0.2 m1
of final extract was added to the gel columns. Electrophoresis
was carried out as desc~ibed in the Materials and Methods. Only
one band appeared in the hydrogenase assay. When the gel was
stained by 1% amido black in 7 1/2% acetic acid, at least 9
protein bands were observed. 107
f Hydrogenase Band. Rf • O. 68 108 in an attempt to adsorb the hydrogenase onto the column, but routinely
95% was eluted in the first few fractions with little or no purification. The remaining 5% was never eluted in active form even up to salt concentrations as high as 0.8 M. Calcium phosphate gels and carboxymethyl-cellulose chromatography at various pH's and molarities also gave negative results.
Preparative gel electrophoresis of crude and partially purified extracts on medium size polyacrylamide gels (lcm x 10cm) was also tried. The gel columns were subjected to electrophoresis for Z hours prior to addition of samples (3-4mg of crude extract or 0.5-1.0 mg purified preparation), for one hour with the addition of sodium dithionite to eliminate any excess ammonium persu1fate, a strong oxidant. The strongest hydrogenase species, the 0.70 band, was excised after electrophoresis; homogenized under reduced conditions and extracted for 3 hours in 0.05 M phosphate buffer pH 7.0 under HZ at 4°.
The suspension was then centrifuged at 3Z,000 x g for 15 minutes, and the supernatant tested for activity by the manometric measurement of evolved HZ from reduced methyl vio1ogen. The identity of this species was tested by rerunning the extract on 7.5% polyacrylamide disc gels.
Protein stains of the small gels revealed the presence of not more than
3 protein bands. The activity calculated from preparations obtained by
this method was usually between 50,000-90,000 ~1 HZ evolved per hour
per'm1 of sample. Protein determinations were difficult to obtain as
the polyacrylamide interfered with the methods employed. However
repeated dialyses against several changes of 0.05 M phosphate buffer
eliminated most of the polyacrylamide and by using the buffer as a 109 blank, the protein content was estimated to be always less than 10 ~g per m1. These preparations exhibited an increase of 500-900 fold in specific activity as assayed by the evolution of hydrogen gas. These extracts when stored under H2 at 0° were stable up to a week.
An apparatus for preparative temperature-regulated polyacrylamide gel electrophoresis (106), in discontinuous buffer systems was employed in attempts to purify extracts of this hydrogenase. A new formulation of polyacrylamide gels in which the N,N,N 1 ,N'-tetramethy1ethy1enediamine
(Temed) and ammonium persu1fate were increased 2 and 4 fold, respectively was used for electrophoresis at low temperatures, and relatively high voltages (500-800v). Electrophoresis of 40 mg samples were carried out at 50 milliamps under reduced conditions (0.05% sodium dithionite). This technique allowed a continuous elution of protein bands with minimal dilution; however in 20 separate experiments no hydrogenase activity was detected.
Properties of purified hydrogenase. The properties of a hydro genase preparation of £. buty1icum purified 600 fold were compared with those of crude preparations from this bacteria. There was no change in pH optimum of this purified extract (Table XV). Both preparations
appeared to exhibit the same optimal activity at about pH 6.8. This
agrees with previous observations (69).
The similarities of the two extracts were also exhibited when
tested with redox dyes of various potentials (Table XVI). The highly purified hydrogenase, like the crude enzyme, catalyzed the evolution
of H2 gas only from the reduced forms of methyl vio10gen (- 447 and
- 511 mv). 110
TABLE XV. OPTll1UM pH OF CRUDE AND PURIFIED HYDROGENASE AS DETERMINED BY THE REDUCTION OF METHYL VIOLOGEN
Specific activity of crude Specific activity pH (~1 H2/hr/mg protein) of purified prep.
6.0 3000 4.8 x 10 5
6.5 3600 5.0 x 10 5
6.7 3800 5.5 x 10 5
6.8 4000 5.8 x 10 5
7.0 3600 5.2 x 10 5
7.4 3200 4.0 x 10 5 111
TABLE XVI. REDOX SPECIFICITIES OF THE CRUDE AND PURIFIED HYDROGENASESRECORDED AS SPECIFIC ACTIVITIES (~1 H2/hr/mg PROTEIN)
Crude extract Purified extract Dyes (reduction assay) (reduction assay) Methylene blue o o
Benzyl vio1ogen 9 x 10 3 5.5 x 10 5
Methyl vio1ogen 4.8 x 10 3 5.0 x 10 5
Methyl vio1ogen 4.5 x 10 3 4.8 x 10 5 (- 511 mv)
1,1'trimethy1ene 15 100 2, 2'dipyridi1ium di-iodide (- 560 mv)
1,1'tetramethy1ene o o 2,2' dipyridi1ium di-iodide (- 640 mv) 112
Both crude and purified preparations were stable at 60° under hydrogen for 15 minutes. Sedimentation and gel filtration experiments
gave identical molecular weight values and shape properties for the
unfractionated and purified extracts (Tables IX, X, XI). The only
difference observed between these two preparations was the substrate
inhibition that occurred during the kinetic experiments. Inhibition of
the purified enzyme occurred at lower concentrations of substrate than
with the crude extract.
It appears therefore, that the enzyme purified to this extent
(600 fold) is not significantly different to the enzyme in crude
preparations.
The maximum velocity of a highly purified preparation of £.
buty1icum hydrogenase was determined (Figure 15) in order to estimate
its turnover number. The protein content was estimated at 5 x 10-5 ~
moles assuming a mo1ec~lar weight of 100 000 for hydrogenase. This
preparation was run on polyacrylamide gel and stained for protein.
This hydrogenase band was one of nine protein bands exhibited on the
gel (Figure 15). If each band contributed equally to the protein
estimation, then the actual enzyme content is only 1/9 of the total
protein concentration. The specific activity of the hydrogenase has
been calculated as 2 x 106 ~ moles H2 evo1ved/min./~ mole protein
(Figure 14). By multiplying by a factor of 9, the turnover number
(~mo1e substrate/min./~mo1e enzyme) can be calculated and this is
18 x 106 ~mo1e H2 evo1ved/min./~mo1e enzyme. This value is more than
3 times that of catalase (107) which is the most active enzyme yet
studied. 113
Figure 16
Vmax determination of £. buty1icum hydrogenase.
Manometric assays employing the method Lineweaver and Burk (89)
were used to determine the maximum velocity of the hydrogenase
reaction from reduced methyl viologen. Enzyme concentrations,
about 0.01 m1 in volume, were adjusted to give rates of 10
~liters per minute. Methyl vio1ogen concentrations were varied
over a range of 5.5 ~mo1es to 55 ~mo1es. Sodium dithionite
concentration was set at 3 ~mo1es. Total volume of the reaction
mixture was 3.2 m1s (phosphate buffer pH 6.8). The gas phase
was helium. The slope of the line and the intercepts were
determined by the "least squares" method. Temperature was
constant at 30°. Substrate inhibition at high methyl vio1ogen
concentrations is indicated by the curved line. Molecular
weight of 100,000 assumed for hydrogenase. 5
·N 0 4
JC C E "-..• 3 / '0 \ / Protein Cone. • 5 lC 10-5 Jlmole• E 'e /e ~ 7- Vrnax • 100 Jlmol.. ~ I min. - 2 6 > Specific Activity • 2 lC 10 ...... H2/min./jImOle "- / protein
/ Turnover Nurrber = 18 It I06J1mOle. ~/minJSlmole enzyme / /
-I o 2 3 4 5 G liS (Jlmoles)
I-' I-' ~ 115
F. Kinetic Assays on the Co-Factor Requirements of Hydrogenase.
The mechanism of hydrogenase action proposed by Peck and Gest (71) implicated the participation of at least 3 co-factors. Kidman (98) has observed, with extracts of £. pasteurianum hydrogenase after
Sephadex gel filtration, that while 70% of the methylene blue activity
(2-e1ectron acceptor) was retained, only 35% of the methyl vio10gen activity (1-e1ectron dye) was present. This cou1dhave been due to the loss of some co-factor essential to one-electron transfer but unnecessary for the 2-e1ectron reaction. It has been demonstrated that molybdenum, when incubated with the fractions of C. pasteurianum hydrogenase eluted from a Sephadex column, restored part of the methyl vio10gen activity (98). The role of iron has not been clearly established; however, crude and partially purified extracts of hyd~ogenase, when dialyzed in the presence of ferrous sulfate under reducing conditions, retained significantly more activity than those preparations without iron. Also, the results of the growth experiments indicate the importance of iron for hydrogenase activity.
Effects of co-factors on C. pasteurianum hydrogenase. To further study the effects of co-factors, a dinetic analysis was undertaken with Sephadex treated £. pasteurianum hydrogenase and because of its highly purified state (2 x 106 ~1 H2 evo1ved/hr/mg protein) the hydrogenase of £. buty1icum. Enzyme activity was tested over a range of substrate concentrations using the double reciprocal plots of
Lineweaver and Burk (89) to determine the kinetic parameters Vmax which is the maximum velocity as substrate concentration (S) approaches 116
infinity, and Km a constant which can be defined as,
Km = S at Vo = Vrnax/2 where Vo is the initial velocity.
A1iquots (1 m1) were obtained from the pooled eluates containing
C. pasteurianum hydrogenase after gel filtration from a Sephadex G-100
column. Sodium molybdate (1 mg) and ferrous sulfate (2 mg) were added
to separate vials and incubated under hydrogen at room temperature.
The control consisted of an equal volume of eluant stored under
hydrogen. At various times after incubation the Vmax and Km values
were determined for the control and co-factor preparations (Table XVII).
TABLE XVII. EFFECT OF MOLYBDENUM AND IRON ON HYDROGENASE AFTER GEL FILTRATIONa
Time Vmax Km (hrs) ()ll/hr/m1) (mole/liter x 10-3) Co-factor
0.0 440 3.6 None
1.5 360 3.3 Fe
3.0 1000 7.0 Mo
4.0 800 4.5 Mo
5.0 240 1.4 Fe
6.5 420 5.0 None
aThe manometric assays were the same as in Figure 8 except that 0.5 m1s of enzyme was used. 117
The incubation with molybdenum results in more than a 100% stimulation of activity over the controls, and iron seemed to inhibit rather than activate. In a separate experiment, it was found that the methylene blue activity was not stimulated by the addition of molybdenum and iron as shown by the following table.
TABLE XVI II. EFFECT OF MOLYBDENUM AND IRON ON THE METHYLENE BLUE ACTIVITY OF C. PASTEURIANUM AFTER GEL FILTRATIONa
Time Vmax Km (hrs) (]J1/hr/m1) (mole/liter x 10-3) Co-factor
a 21 000 3.0 None 2 18 000 2.9 Fe
4 12 000 2.5 Mo
5 11 000 2.5 Mo
6 11 500 2.5 Fe aDetermined by manometric assays in which reaction vessels contained 0.4 m1s of enzyme preparation, 6.4 ~mo1es methyl vio1ogen, 2.7 m1s of 0.1 M phosphate buffer pH 6.8. The gas phase was hydrogen.
In a separate experiment the pooled eluates were incubated under
hydrogen; nitrogen; air; Mo and hydrogen; Mo, FAD and hydrogen
respectively, for up to a period of 10 hours. The results are
summarized in Figure 17. The reactivation by Mo is shown and FAD seems
to inhibit the enzyme. Hydrogen and nitrogen effect a slight
stimulation of the activity while air inhibits it completely after
9.5 hours. 118
Figure 17
Reactivation of pooled eluate with time. Fractions
containing £. pasteurianum hydrogenase from a Sephadex G-100
column were pooled and 1.0 m1 a1iquots were incubated under
nitrogen, nitrogen and air, 1 mg of sodium molybdate and
sodium molybdate plus FAD (1 mg each). The hydrogenase
preparations contained approximately 0.1 grams protein per
m1. The manometric assays were the same as described in
Figure 16 except that 0.2 m1 of enzyme was used. 119
900 j\0 800 \ 700 \ / 0 \ 600
- j \ Mo, H2. E 0 ...... ~ .s::. 500 ...... (\J o Mo, FAD, H , ::I: / 2 >. -> 300 u j «-
200 / o 2 100 V 0 . 0 N 0 O"C::g Air o 2 3 4 5------6 7 8 9 10 HOU RS 120
The effect of Mo on the activity peak eluted from a Sephadex G-100 column is shown in Figure 18. Effect of co-factors on C. buty1icum hydrogenase. £. buty1icum hydrogenase does not possess methylene blue activity (Table VII) hence the stimulation or inhibition by added co-factors should be directly related to the one-electron transfer reaction of methyl vio1ogen. Highly purified preparations of hydrogenase from £. buty1icum of specific activity of 3 x 106 ~1 Hi evo1ved/hr/mg protein were divided into two fractions. One fraction was dialyzed in phosphate buffer pH 6.9 under hydrogen with no metals for 3 hours. The other was stored under hydrogen and tested at various times by the Lineweaver Burk method. The activity of the undia1yzed control over a period of 10 hours is shown below.
TABLE XIX. VMAX DETERMINATION OF UNDIALYZED EXTRACT
Time Vmax Km 1 (Hrs) (v1/hr/m1 x 105 ) (mole/liter x 10- )
o 4.0 2.3 4 4.2 2.1
10 4.1 2.2
aThe manometric assays were the same as described in Figure 15. 121
Figure 18
Reactivation by molybdate after five hours incubation.
Fractions (1 ml) containing hydrogenase from a G-100 Sephadex
column were assayed manometrically as described in Figure 16,
and then incubated with 1 mg sodium molybdate for five hours
under hydrogen and assayed. The volume of enzyme preparation
was 0.4 mls containing 40 ~g of protein. , 122
800 A - Control
B- Mo. H2
700
~ 600 -e ...... : ...... c N :J: 500 0 f eu AB -::3,
>. 400 '>- x :0: «u xl
300
x 0 200 0-0\. 100 It \\0-0, x-x-x0 80 90 100 110 120 130 140
Elution Volume ( mls ) 123
To one m1 a1iquots obtained from the dialyzed extract was added sodium molybdate (] mg/m1); ferrous sulfate (2 mg/m1) and ferrous sulfate + molybdate (1 mg of each per m1). The control consisted of the extract stored under hydrogen. The Vmax and KID values are shown in the following table.
TABLE XX. EFFECT OF CO-FACTORS ON HYDROGENASE ACTIVITya
Time Vmax KID (Hrs) Metal (}l1/hr/m1 x 105) (mole/liter x 10-2)
0.0 None 2.0 3.02
2.5 Mo 2.3 3.10
4.5 Fe 2.3 3.20
5.0 Mo 2.8 4.50
5.5 Mo + Fe 2.9 4.60 6.5 Mo + Fe 3.0 7.00 7.5 Fe 2.5 4.00
8.0 None 2.1 3.10
~anometric assays were performed in the same way as described in Figure 15.
The addition of iron stimulated the hydrogenase activity .up to
60% of the undia1yzed extract, while molybdenum gave a value of 70%.
The combined metal .additions gave values up to 75% to that of the undia1yzed extract. No stimulation was observed with undia1yzed extracts. lZ4
The effect of FAD incubation was tested in the same manner in a separate experiment. A concentration range of 5 to 110 ~g of FAD/ml extract was used as recommended by Peck and Gest (9). Samples were tested l/Z hour after incubation and every hour thereafter up to six hours. No stimulation was observed and a pronounced inhibition occurred at concentrations equal to or greater than 50 ~g FAD/ml of preparation (Figure 19).
To insure that the increase in activity was uniquely caused by iron and molybdenum, several other metals were tested. Magnesium sulfate, sodium chloride, manganous sulfate, and cupric sulfate were added (1 mg of each) to separate tubes containing 1 ml of dialyzed extract. The test tubes were stoppered, gassed with hydrogen, and incubated up to 6 hours. The control consisted of an undialyzed and dialyzed extract incubated under hydrogen. The fractions were tested at Z hour intervals (Table XXI).
As mentioned previously (in the introduction) metal sulfides may in some cases effect the reduction of viologen dyes and the evolution of HZ from the reduced forms of these dyes. As an added precaution iron and molybdenum were incubated under the exact conditions as were the enzyme mixtures except that bovine serum albumin was added in place of the hydrogenase and, in a separate experiment no protein was included. Both controls had no HZ uptake or evolution activity up to
8 hours of indubation. 125
Figure 19
Effect of FAD on hydrogenase. A1iquots consisting of
1 m1 of purified extract (specific activity of 2 x 106 ~1
H2/hr/mg protein) were incubated for 1/2 and 2 hours and assayed by the method described in Figure 16. 126
.: .s: 0 f .: en .s: C\I .s: ...... C\I C\I. C\I 0 0 :::::: c( c( 0 II. II. l"-
e /O/K C -0 u /~ @ ....: E ...... 0 0;::' E I/ ... o ~ 0 " U') '"::a. I I o ~ c I I .2 OK .. 0v .."... C GI 1/ U 0 OM C II u0 OM 0 0 If) c( II. OM1/
o OKrt 0 C\I I J OM II OM Q
OMII //
0 en en ,... CD U') V If) C\I ('IWI 'JlIl ~ H SJ8111 r1' ) SOl )( 1\ 127
TABLE XXI. EFFECT OF VARIOUS METALS ON A DIALYZED EXTRACT OF £. BUTYLICUM PURIFIED 400 FOLD
Vmax Incubation time Metal (v1 H2/ml/hr x 1Q5)a (hours)
Na 2.0 2
Na 2.1 4
Na 2.0 6
Mg 1.9 2
Mg 1.9 4
Mg 1.9 6
Mn 2.0 2
Mn 2.1 4
Mn 2.1 6
Cu 2.1 2
Cu 2.0 4
Cu 1.8 6
Control 3.5 2,6 (undialyzed)
Control 2.0 a (dialyzed)
Control 2.1 2 (dialyzed)
Control 2.2 4,6 (dialyzed)
aTo 1 m1 a1iquots of purified hydrogenase (specific activity of 4 x 106 Vl H2/hr/mg protein were added sodium chloride, magnesium sulfate, manganous sulfate, cupric sulfate (1 mg of each to separate vials). Manometric assays were carried out according to the method described in Figure 15. 128
G. Arrhenius Activation Energies of Hydrogenase Enzymes
It has been customary for many years to express the variation of biological rates with temperature in terms of the temperature characteristic, u, according to the Arrhenius formulation (133),
u = 2.303 R
(R is the gas constant, T the absolute temperature, and v the velocity
of the reaction, u is customarily expressed as calories/mole and is a measure of the activation energy of an enzyme-catalyzed reaction).
This energy of activation is a fundamental physical property which can
be studied with nonhomogeneous preparations or even in the intact cell
(89). The variation of activation energy with temperature of many
enzymes has been determined and often, deviations from the expected
linear behavior have been observed (89). These deviations could be
the result of inhibitors acting on the enzyme at certain temperatures,--
or a change in conformation of the enzyme with the variation in
temperature (89).
We have determined the Arrhenius activation energies of
hydrogenases from several sources to gain some insight into the
physical properties of these enzymes from the kinetic data.
All data was treated by the "least square" method to obtain the
best rate value and duplicate runs gave values ± 2% of each other at
high substrate concentration and within 10% at low substrate
concentrations. The temperature dependence of each hydrogenase 129 extract was observed by a plot of log Vmax, obtained by the Lineweaver
Burk method, or log v obtained at a given substrate concentration against l/T, the absolute temperature.
Activation energies of various hydrogenases. The temperature
dependence of hydrogen evolution from methyl vio1ogen reduced by
sodium dithionite by a hydrogenase extract of £. pasteurianum is shown
in Figure 20. This Arrhenius curve has two major slopes; one in the
temperature range 5° to 15° with a temperature characteristic, u, of
13,800 calories/mole, and one above 20° with a u value of 7000
calories/mole. A similar type of curve was obtained by Ackre11 (134)
in experiments where ferredoxin was used in place of methyl vio1ogen
as the substrate. The u values for this reaction above and below 17° were 13,000 and 23,000 calories/mole, respectively (Table XXII).
Ackre11 (134) also observed an inflection of the Arrhenius plot around
17° with pyruvate serving as the electron source, in place of dithionite,
for evolved hydrogen. In this instance the phosphorc1astic reaction
was followed in ferredoxin-free extracts of £. pasteurianum supplemented with methyl vio1ogen or ferredoxin. The terminal step in the oxidative
decarboxylation of pyruvate is the transfer of electrons from reduced
co-factor to hydrogenase and this step appears to be rate limiting
since high levels of reduced methyl vio1ogen were observed under steady
state conditions. The evolution of hydrogen by hydrogenase is thus
common and rate limiting in both evolution and phosphorc1astic
experiments.
Extracts of £. fe1sineum, D. desulfovibrio, E. coli, A. vinelandii and £. butylicum (Table XXII) also exhibited a break in the Arrhenius 130
Figure 20
Arrhenius plot for the evolution of hydrogen in
C. pasteurianum extracts. Manometric assays were run as
described in Figure 16. Specific activity of the extract
was 4 x 103 ~l H2/hr/mg protein and contained all three
forms of the hydrogenase. The Vmax values are listed
below.
Temperature °c Vmax (~l H2 evolved/min.) 5 2.47 ± 0.083
10 3.92 0.030
15 6.57 0.42
20 15.00 4.10
25 8.30 0.73
30 9.31 0.64
35 31.30 10.00 131
oI
o
......
o
/o
o
,... CD It) o CD ..: o TABLE XXII. TEMPERATURE CHARACTERISTIC (U) OF BACTERIAL-SUBSTRATE COMBINATIONS
Bacterium Substrate Reaction u 5-15° u 20-30° ca1/mo1e ca1/mo1e
C. pasteurianum Fd Evolution 23,000 13,000
" HV " 13,000 8,200 C. fe1sineum MV " 22,700 8,000 D. desu1furicans MV " 12,600 6,000 E. coli MV " 13,000 6,000 C. pasteurianum Fd Phosphorc1astic 9,400 13,300
" MV " 11,700 13,700
" MV Reduction 9,600 12,800 A. vine1andii MeB " 6,500 13,700
~. buty1icum MV Evolution ----- 8,700 aThe evolution reaction was carried out as described in Figure 20. bThe reduction reaction was carried out as described in Table XVIII except that 0.01 m1 of enzyme preparation was used. cphosphorc1astic reaction was done according to the method of Wolfe and O'Kane (142).
I-' Vol N 133 curve at around 17°. This break occurred in both the hydrogen uptake and hydrogen evolution reactions though the break in the hydrogen
"uptake" reaction was not as pronounced as that in the evolution of hydrogen.
A correlation of electrophoretic mobilities of the hydrogenases of two bacteria with the Arrhenius activation energy was also observed.
Crude extracts of C. pasteurianum were run on 7.5% polyacrylamide gel at temperatures ranging from 5° to 35° (Figures 21, 22), and the variation in Rf's of the ~, Sand y forms were recorded. A pronounced break in the curve occurred at about 17° to 18° with the ~ species and a slight, but disproportionate increase occurred at about the same temperature with the y band (Figures 21, 22). The slope of the S band paralleled that of the ~ form at temperatures above 20°; however, this band could not be detected at temperatures below 18°. This may be due to the inactivation of this form or to the reduced mobility of this form so
that it migrated at the same rate as the y band. This behavior may be a manifestation of a slight and different conformational change for
each isoenzyme above and below 17°.
Similar results were also obtained with £. buty1icum hydrogenase.
The variation of Rf and the energy of activation with temperature of
a purified extract possessing a single hydrogenase species is shown in
Figures 23 and 24. It is significant that this deviation from
linearity also occurred at about 17°. 134
Figure 21
Variation of Rf with temperature. Crude extracts of
£. pasteurianum hydrogenase were subjected to polyacrylamide
gel electrophoresis as described in the Materials and Methods
section. The gels were polymerized at the respective temperature
of each electrophoretic experiment from 5° to 35°. 135
0.65
0.64
0.63
0.62
0.61 ~- hydrogenase
0.60
-0::
'( - hydrogenase
0,52
5 10 15 20 25 30
Tomperature - 0 C 136
Figure 22
Variation of the ~ band of £. pasteurianum with
temperature. Experiments were performed as described in
Figure 21. 137
0.75
0.74
0.73
0.72
0.71 -a: 0.70
0.69 Cl(, - Hydrogenase
5 10 15 20 25 30 35 Temperature - °C 138
Figure 23
Arrhenius plot of the evolution of hydrogen reaction of a
highly purified extract (450 fold) of £. buty1icum. The
preparation consisted of only one hydrogenase species.
Manometric assays were performed according to the method described
in Figure 16. Specific activity of the preparation was 2 x 106
~l/hr/mg protein. A volume of 0.005 m1s of enzyme was used.
Temperature DC Vmax (~1 H2/minute)
5 24.2 ± 0.24
10 32.7 0.27
15 38.5 0.89
20 48.0 0.90
25 47.6 1.67
30 43.0 0.50
35 40.7 1.60 139
CD 10
II) oI 10
o
... o "
o
o
V 10 (\J - - ~ - ('UIW/ ~H SJ91IPf) XDwA 001 140
Figure 24
Rf of 450 fold purified C. buty1icum hydrogenase vs
temperature. A highly purified extract of Q. buty1icum
containing only one species of hydrogenase was run on
polyacrylamide gel as described in Figure 21. 141
II') N
0 0 0 • N .a".. ~ a.. 0 Do !a "E I- \0 "
\ 0
\0 \ 0
U) It) ..,. It) N o C7l CD ,... U) It) ..,. U) ~ U) U) U) U) U) It) It) It) It) It) It) 000 000 000 o 0 o 0 142
A change in the molecular weight of the enzyme at the transition temperature could account for these results. However, Sephadex chromatography and sucrose density gradient determinations of molecular weight at 20° and 5° show the same molecular weight for the hydrogenase of C. buty1icum.
TABLE XXIII. MOLECULAR WEIGHT DETERMINATIONSa OF C. BUTYLICUM HYDROGENASE AT 5° AND AT 20° -
Method Molecular weight Temperature
Gel filtration 120,000
Gel filtration 120,000
Sedimentation 105,000
Sedimentation 105,000
aMo1ecu1ar weight determinations were performed as described in Materia1S-and Methods. Manometric assays were carried out according to the method described in Figure 8.
Proteins of .various sizes and shapes were run on the polyacrylamide gel at various temperatures to serve as controls to indicate the unique behavior of hydrogenases in these experiments.
All proteins examined yielded straight lines when their Rf's were plotted against the change in temperatures although the rate of change of Rf's with temperatures was greatly different with various proteins tested (Table XXIV, Figures 25, 26). TABLE XXIV. VARIATION OF PROTEIN MOBILITIES loJITH TEMPERATURE ON POLYACRYLAMIDE GEL a
Change in Rf Protein Source Molecular weight f/fo__ _ _ witl1.....temperature
YADH Yeast 150,000 1.28 1.8, 0.6b b BSA Bovine 66,500 1.30 5.0, 3.6 Catalase Beef liver 250,000 1.25 ° G1ucose-6-P04 Yeast 120,000 1.40 3.4 dehydrogenase
aE1ectrophoresis on polyacrylamide gel was performed as described in Figure 15, except that 0.1 mg of protein was used. bIn Yeast alcohol dehydrogenase (YADH) and Bovine serum albumin (BSA) two protein bands appeared after staining. These have been designated as number 2 in Figures 25 and 26.
I-' ~ (.oJ 144
Figure 25
Plot of Rf of Bovine serum albumin (MW 66,500) vs
temperature. Electrophoresis carried out as described in
Figure 15, except that 0.1 mg of protein was used. BSA #1
and BSA #2 denotes the two bands observed after staining. 0.43 0.31 0.30 0.42
0.41 0.29
0.40 0.28
0.39 BSA # 1 0.21 0.26 0.38 0/ /' N ~5 • 0.31 c:( 24 enCD 0 •36 /./' * c:( ~3 en .... 0.35 CD o .... 0.34 0~'~U2 ~2 0 ...... a:: 0.33 Q21 a::
0.32 ~O 0.31 of' 0.19 5 10 15 20 25 30 35 I-' ~ Temperature _ °C l.n 146
Figure 26
Plot of Rf's of various proteins vs temperature.
Electrophoresis was performed as described in Figure 15,
except that 0.1 mg of protein was used. Yeast alcohol
dehydrogenase (MW 150,000) possessed two bands upon
staining. G1ucose-6-phosphate (MW 120,000, axial ratio of
11:1; and Catalase (MW 250,000 and axial ratio of 5:1)
possessed only one band. 147
0.35 0.34 /0 0.33 YADH • ~2 ---0-7~ 0.32 _ x __-- x ------x. _. x- 0.31
0.30
0.29 Glucose _ 6 _ Phosphate 0 / Dehydrogenase ~ 0 0.28 0 0~0/ -0:: 0.27. 0.26 /0/0~0// 0.25 yO YADH" I 0.24 /0 0.23
0.22
0.21 Catalase 0 0 C 0 0 0 0
5 \0 15 20 25 30 35
Temperature -•C 148
This change of activation energy and mobilities on the gel appears to reflect changes in the physical and chemical properties of the enzyme itself. The effects of enzyme-substrate interaction do not appear to be as important since these changes occur irrespective - . of the electron source or substrate and appear in both the evolution and reduction assay. This type of behavior has been observed with several other enzymes and several possible causes have been suggested to explain this phenomenon (see discussion). DISCUSSION
It is possible to demonstrate that the hydrogenase enzyme of many diverse types of bacteria exists in several different molecular forms.
This has been accomplished by subjecting crude extracts of bacteria to disc electrophoresis in polyacrylamide gel and then detecting these enzymes with a hydrogen-dependent reduction of a suitable redox dye.
These multiplicity of forms also extend to the examples of the particulate (membrane-bound) hydrogenases studied; as treatment of these particulate fractions with deoxycholate and trypsin yielded at least 2 different soluble forms. Efforts to convert the solubilized
forms into the other have so far failed, and, one might conclude that different particulate hydrogenases may exist within the same bacteria.
Of more than 45 bacterial species examined in these studies, only
extracts of ~. tetanomorphum, ~. thermosaccharo1yticum, ~.
aurantibutyricum, f. f1avum, R. pa1ustris and ~. po1ymyxa exhibited a single hydrogenase band when run on polyacrylamide gel at pH 8.3. The band pattern of each of the other species was unique, differing from one species to another, and thus might prove useful in taxonomic considerations. For example, the hydrogenase band patterns of
V. a1ca1escens and M. 1acty1iticus are very similar and would support
the reclassification of ~ 1a£ty1iticus as a Vei110ne11a species as suggested by Rogosa (108). No differences in band patterns were found among different strains of the same species. This premise was tested with 2 strains of V. a1ca1escens (ATCC 12641, and ATCC 12642); 7
strains of E. coli (ATCC 4157, 4637, 13706, ML 3, ML 30 B, and K12)
and in 7 strains of C. fe1sineum (ATCC 8260, 13160, and McClung 541, 150
644, 539, 639, and 2822 A var. vinaceum). The one exception was McClung
2822 A a variant strain of £. felsineum which possessed the same number of bands as the rest of the other felsineum strains but at different
Rf values (Table II).
The results from the bioluminescent marine bacteria (Table III) serves to further emphasize the use of polyacrylamide technique in categorizing hydrogenase containing bacteria. A tentative identification of an uncharacterized marine micro-organism as a ~-train of Photobacterium phosphorium has been made using this method. The classical microbiological means of classifying bacteria were employed but because of their complicated and ambiguous nature no definite conclusions could be drawn. The sensitivity and consistency of the gel technique may prove helpful in these cases.
Experimental evidence has been obtained which indicate that the existence of these enzyme species is not dependent upon the method used to disrupt the cell. Further, it appears that they are not influenced by growth conditions even though the total hydrogenase
activity level may vary over a 10-fold range with different culture methods. A fixed nitrogen source in the case of A. vinelandii and
R. rubrum markedly lowered the total hydrogenase activity but the
isoenzymes of each species retained their identities. Likewise, in
£. pasteurian~m high concentrations of iron increased the activity
several fold but did not affect the band patterns of this bacterium.
In R. rubrum different storage methods resulted in large decreases in
the total hydrogenase activity, however the same band pattern was
observed throughout the course of the experiment. This may be the 151 result of the various hydrogenase species reacting at the same rate in the denaturation reactions or it may be a manifestation of an equilibrium among enzyme species.
The isoenzymes of each of the bacteria studied were identical with respect to substrates of different redox potentials (Tables VI and
VII). There was a quantitative difference in the activities of the hydrogenases obtained from the anaerobes when compared to the activities of that obtained from the photosynthetic bacteria, the obligate aerobe A. vine1andii, and the chemo-autotroph
Hydrogenomonas eutropha (Table VI and VII). It is not possible to determine at this time the redox potentials of the various enzymes but several trends were noticed. The activities of all hydrogenases varied in the pame way with redox dyes of various potentials. As the redox potentials of substrates increased, the activities of all the hydrogenases decreased. The isoenzymes of each bacteria reacted in a similar fashion with no difference in specificities of isoenzymes of a particular strain. The only exception was the hydrogenase system of
H. eutropha (Figure 6) in which one enzyme reacted with NAD and its acetyl derivative. This may be due to an enzyme with dual specificities as in the case of the hydrogen dehydrogenase described by Bone (86).
It appears that the oxidation-reduction potentials of the hydrogenases from the strict and facultative anaerobes is between
- 447 mv and - 560 mv as the greatest drop in activity occurred a point between these two potentials. An accurate determination awaits the purification of the enzyme. The potential of the hydrogenases from 152
the photosynthetic bacteria appear to be similar to that of the
anaerobes (Table VII) while that of the obligate aerobe and the
chemoautotroph appear to possess a hydrogenase of higher potential
(- 447 mv to - 511 mv, Table VII).
Clostridium pasteurianum possessed a maximum of six hydrogenase
species and three, designated as the ~, S, and y forms were chosen for
study, as they were the most active, appeared regularly, and were
routinely reproducible. They were interrelated by the conversion of
the ~ and S species to the more stable y form. Their respective sizes
were similar, with molecular weight values of 50,000 to 60,000 as
determined by sucrose gradient centifugation, gel filtration and by
their relative movements in various concentrations of polyacrylamide
gels (Tables IX, X, Figure 14). The mechanism of conversion appears
not to include the possibility that these forms were produced by the
oxidation of a single species, as all experiments were performed in
a reducing environment. Storage of the ~ and S forms, or the dilution
that occurred during partial purification, invariably brought about
the conversion of ~ and S species to the y species (Figure 13). An
explanation of the multiplicity of forms may be that the loss of some
co-factor on storage or purification produces a change in
configuration or a change in the resultant charge;- or both, causing a
separation of these species on the gel.
Sedimentation constants from centrifugation and the Stokes radii
determined from Sephadex gel filtration (Table XII, Figures 11, 12)
were used to estimate the frictional ratios, diffusion constants, and
-molecular weights of protein standards and of these enzymes. The 153
frictional ratio, fifo, is always greater than unity and is the measure of the deviation of a shape from that of a perfect sphere. The
diffusion constant is defined as a net flow of molecules across a
concentration gradient. Both these parameters are related to the
Stokes radius of a protein or macromolecule (equations 3, 4, 5) and all
three are involved in the determination of the elution position of a molecule from a Sephadex gel column (92,99,100,103). The combination
of the gel filtration and sucrose gradient centrifugation technique
enables one to determine the size and shape of macromolecules in
impure preparations and this was done to check the values obtained by
each method alone (Table XII). The results agreed reasonably well.
Except for f. butylicum, the strict anaerobes possessed hydrogenases of similar molecular sizes. The isoenzymes of f. butyricum, like £. Easteurianum, had similar Dolecular weight values
of 50,000 to 60,000. This value was also reported for the isoenzymes
of f. felsineum (98). The hydrogenase of £. butylicum, on the other hand, was twice as large (Table XII). Attempts to dissociate this
enzyme into smaller active species have failed and it is not known why
the enzyme of this particular species is of a different size.
Kidman (98) has demonstrated that the soluble hydrogenases
obtained from the facultative anaerobes ~' vulgaris and E. coli); an
obligate aerobe ~' vinelandii); and D. desulfuricans to be significant
ly different from the strict anaerobes. The nonsedimentable activities
of these bacteria could be resolved into two peaks by gel filtration;
a relatively low molecular weight species (MW 104-105) and large molecular complexes (MW -- 106). 154
The existence of two hydrogenases of different sizes is difficult to explain. It has been observed that the amount of soluble hydrogenase produced by A. vine1andii increases with growth time and it may be that the smaller soluble enzyme is the end product of a sequence of solubilization reactions as the cell ages and breaks apart with time. The larger species may then be an intermediate between the completely soluble hydrogenase and the membrane-bound complex (which is always present in all bacteria tested except the strict anaerobes).
The existence of only soluble hydrogenase activity in the strict anaerobes and the presence of both soluble and membrane-bound
(particulate) enzymes in all other bacteria may be a manifestation of the type of electron transport in which hydrogenase participates and of a protective mechanism.o..f the bacteria.
D. desu1furicans serves as an example of the correlation of the two types of hydrogenases with the electron transport systems within an-organism. This bacteria possesses both a cytochrome-linked hydrogen1yase and clostridial "clastic" reaction; it may be possible that the soluble hydrogenase functions in the cytoplasmic phosphorc1astic reaction (a ferredoxin-dependent reaction) while the particulate hydrogenase participates in the membrane-bound hydrogen1yase sequence. This is supported by the demonstration of the presence of ferredoxin, in small amounts, in this organism by
Valentine (112). A comparison of the specificities of these forms of hydrogenases with respect to their natural substrates (ferredoxin and the cytochromes within the hydrogen1yase complex) could clarify the situation. 155
It has been shown that the particulate hydrogenase was more stable to oxygen than its soluble counterpart (109). In addition our work with the facultative organism, E. ~oli, demonstrated that the level of membrane bound enzyme depended to a certain extent on the oxygen tension during growth. Therefore, in the strict anaerobes, which grow in the absence of oxygen, there is no need to protect the enzyme against this inhibitor. Thus only soluble enzymes are present.
A fractionation scheme for the purification of £. buty1icum hydrogenase has been presented. It was necessary to maintain anaerobic conditions throughout the procedure; however, ammonium sulfate treatment appeared to lessen the inactivation of the methyl vio1ogen (one-electron transfer activity of the enzyme even in the presence of oxygen, this has also been observed by other investigators (9,81). Enzyme preparations with specific activities as high as 6.Z x 106 ~1 HZ evo1ved/hour/mg protein and representing a
600 fold purification have been routinely obtained. The protein estimation by running the sample on polyacrylamide gel and then staining with amido black in acetic acid (Figure 15) has shown that the contamination by protein other than hydrogenase represented the majority of the actual protein content in these extracts. It is estimated that a further purification of 10 fold or a total purification of 6000 fold over the crude extract is required to obtain this enzyme in pure form.
By use of this method (Figure 16) the turnover number of the enzyme was calculated and found to be at least 3 times as active as catalase the most active enzyme known to date. Preparative gel 156 electrophoresis on medium size columns resulted in higher purification
(900 fold) but ~he amount of enzyme obtained was very small. A feasible method of obtaining a reasonable amount of purified enzyme remains a problem. It appears that any type of method which involves an effective dialysis step, such as DEAE-ce11u1ose column chromatography and Sephadex gel filtration results in an alteration of the enzyme with a.corresponding loss of activity.
For the first time evidence has been obtained that hydrogenase is a very active enzyme contained in small amounts within the cell.
It may be that hydrogenase functions as a control factor in the anaerobic metabolism of micro-organisms as suggested by certain investigators (10). It possesses a high turnover number and the ability to direct electrons to various metabolic pathways (10,132) or to dispose of the excess reducing power as hydrogen gas.
Properties of a highly purified preparation of hydrogenase of f. buty1icum were compared to that of a crude extract and no differences were observed (Tables XV, XVI) in molecular size, pH optima, dye specificity and heat stability. It appears therefore, that the enzyme is not drastically altered when subjected to fractionation procedures. These results support the validity of the molecular size determinations of other hydrogenases in impure preparations. This is in agreement with the premise of various other investigators (92,99,
100,101,103).
Iron, molybdenum, and flavin adenine dinucleotide (FAD) were associated with hydrogenase according to various investigators (71,80) and a suggestion as to their role in enzyme activity has implicated 157 molybdenum in the reaction with the one-electron vio10gen dyes but not in the two electron transfer to methylene blue.
In our gel filtration experiments with the hydrogenase system of
C. pasteurianum, at least 65% of the methyl vio10geli activity was lost while only 25% of the activity could not be recovered in the methylene blue assay. The mechanism would account for this if the enzyme had an easily dissociable molybdenum moeity and more tightly bound iron and
FAD components, and consequently contained much less molybdenum than iron and FAD after elution. This would result in reduced methyl vio10gen activity. In support of this, iron was the only metal found by X-ray f10urescence spectroscopy in fractions of Q. desu1furicans obtained from gel filtration (73).
The activation and kinetic studies on~. pasteurianum hydrogenase supported the role of molybdenum in the one electron transfer activity
(Table XVII, Figure 17) but not in the two electron process (Table
XVIII). The role of iron and FAD was not clearly established.
Evidence for the participation of iron comes from the growth experiments (Table XIII) and by the inhibition of hydrogenase activity by CN, CO, and iron che1ating agents (15,69,71,112).
As C. buty1icum hydrogenase does not possess methylene blue activity, the stimulation of hydrogenase activity by iron and molybdenum can be construed as direct evidence for the participation of these metals in the one-electron transfer step to methyl vio1ogen.
The effects of dialysis on purified extracts of £. buty1icum may be also explained by the loss of molybdenum as incubation of the dialyzed enzyme with this metal resulted in a significant stimulation of 158 activity. The addition of iron also resulted a marked increase over the dialyzed control. No stimulation was observed by the incubation with FAD and an explanation may be that the flavin is tightly bound to the enzyme and could not be removed by a simple dialysis procedure; or the flavin component may be bound irreversibly to the enzyme and once removed it cannot be restored. This may explain the results of
Sephadex and DEAE column chromatography in which there was a 95-100% loss in activity. Also, the lack of stimulation by these components in the undia1yzed control may be explained by the loss of an irrep1acib1e FAD moeity during fractionation procedures, and little or no loss of metals until dialysis.
There has never been any direct evidence for the participation of iron, molybdenum, and FAD in the hydrogenase reaction, but the experiments using other metals as controls (Table XXI) indicate that the reactivation of hydrogenase activity is dependent on iron and molybdenum.
The hydrogenase enzymes examined in the study of the Arrhenius law behavior varied greatly in physical properties; some were particulate or mixtures of particulate and soluble enzymes, such as the hydrogenases of E. coli, D. desu1furicans, and A. vine1andii, while some are soluble such as the enzymes of~. fe1sineum, ~. buty1icum, and ~. pasteurianum. The organisms themselves also differed widely in their optimum growth conditions, the clostridia and D. desu1furicans being strict anaerobes; E. coli a facultative anaerobe and A. vine1andii an aerobe. Despite the diverse nature of the source, all
the hydrogenases studied exhibited the same marked discrepancies in 159 their Arrhenius plots, i.e. a change in the apparent activation energies occurred at a temper~ture of 15-20°.
This change in activation energy appears to reflect changes in the physical and chemical properties of the enzyme itself. The effects of the enzyme-substrate interaction do not appear to be as important since ,these changes occur irrespect'ive of the electron source or substrate and appear in both the evolution and reduction assay (Table XXII). This type of behavior has been observed with
several other enzymes and several possible causes have been suggested'
to explain this phenomenon.
The enzyme urease displays classical Arrhenius plot behavior
except in the presence of 803 when a marked deviation from linearity
occurs below 22°. Kisteakowsky and Lumry (136) have shown that
sulfite reversibly inhibits the enzyme, the inhibition being greater
at the lower temperatures. Thus the observed Arrhenius plot in the
presence of 803 represents a gradual decrease in sulfite inhibition
with increasing temperature until at 22° inhibition is negligible.
Any sharp break in the curve would thus be an artifact of kinetic
measurements. It is difficult to imagine the presence of an inhibitor
of this type in the extract of each of the bacteria examined in this
study.
Kavanau (137) and later Maier et a1. (138) have investigated
some hydrolytic and oxidizing enzymes and have shown marked deviation
of the Arrhenius plot at low temperatures. These authors interpret
their data as representing confirmationa1 changes occurring at the
lower temperatures. These changes are postulated to take place as a 160 result of~±ncreased hydrogen bond formation which results in the formation of an altered enzyme with a higher activation energy. At high temperatures thermal alteration occurs with minimum intra hydrogen bond formation, .whi1e at lower temperatures increasing hydrogen bond formation occurs resulting in altered enzyme activity.
Sheraga et a1. (139) points out that hydrophobic bonds have the opposite thermal stability pattern being less stable at lower temperatures. Therefore, the conformational state of an enzyme reflects a balance between both types of temperature sensitive bonds.
It is thus postulated over a range of temperatures, enzymes can exist in a continuum of conformational states.
The hydrogenase enzymes appear to represent an example of this type of behavior. The conformational changes in hydrogenases appear to have a sharp temperature transition resulting in Arrhenius plots which show marked deviations from straight lines (Figures 20, 23).
The problem is further complicated by the appearance of multiple forms of hydrogenase in crude extracts of these organisms and the conversion of some of these forms into others (Figure 13). However, experiments reported in thi~ thesis with extracts of £. buty1icum in which only one hydrogenase species was present indicate that each isoenzyme by itself exhibits nonclassical Arrhenius behavior. Further evidence that the abrupt change in activation energy at 17° for hydrogenase enzymes is a manifestation of a general physical change in the enzyme
structure was obtained by observations of the electrophoretic mobility
on polyacrylamide gel electrophoresis.
In £. pasteurianum, 2 of the three isoenzymes (Figures 21, 22) 161 exhibited a disproportionate increase in Rf value above 17°. These two forms, the ~ and y bands could be distinguished from each other by the type curve they produced. The S band could not be detected below temperatures of 18°. This could have been due to the complete inactivation of this form or the conversion to the more stable y species below these temperatures. It appears then that each form is slightly different from each other.
This physical change does not appear to include a gross alteration in molecular size as indicated by the molecular weight determinations of~. buty1icum hydrogenase at SO and 20° (Table XXIII).
As the migration of proteins in polyacrylamide gel electrophoresis depends on their size and charge, the discontinuous behavior of the isoenzymes may be explained by an alteration in their charge at the transition temperature, and this change may be due to a conformational modification of the protein. This behavior seems to be a unique property of the hydrogenases as other proteins examined yielded straight lines when their Rf's were plotted against the variation in
temperature (Tables XXV, XXVI). Further studies are currently under way. The following scheme is a summary of the physical properties of hydrogenase isoenzymes within bacteria of dissimilar fermentative pathways. Certain aspects are hypothetical. These include the in vivo reversible solubilization of the particulate hydrogenase into two different solubilized forms and the interconversion of the isoenzymes
of the strict anaerobes. Several common characteristics of hydrogenases are listed such as the similar trends in specificities of
the enzymes to dyes of different redox potentials and nonclassical Organisms containing only soluble hydrogenases strict anaerobe s
similar MW I behavior above ~ and below I 7 0 C ,..,...... l. I enzyme forms LL possible Intercanverslon
Organisms containill9 soluble and particulate hydrogenases
facultative anaerobes, obligate aerobes, .n. desulfurlcans Redox spe.clficities similar MW I behavior soluble, common SolUbilized) dlscontinuos Arrhenius 1 enzymes above a below I 7° C behQvior- 5°- 35° I l --..-.-- uncommon solubilized <. 17° 717° enzymes i I trypsin I --./ in vivo I LL particulate enzymes !-' C'\ N 163
Arrhenius behavior from 5° to 35°. Similarities within certain categories of bacteria are also noted. Organisms containing only soluble hydrogenases possess similar sized enzymes, the same trend is noted for the organisms containing soluble and particulate hydrogenases. However, the sizes of the soluble enzymes from the two different categories of bacteria are different.
A conformation change in the enzymes from both soluble and particulate forms is postulated at about 17°.
The ultimate solution to these problems, the mechanism of hydrogenase action and its physical and chemical properties, must await the development of the complete purification of this enzyme.
The technique of polyacrylamide gel electrophoresis promises to offer continuing rewards. BIBLIOGRAPHY
1. Kaserer, R., Biochem. J. 25, 205 (1931).
2. Grohman, N. , Biochem. J. 12.., 205 (1931) . 3. Nik1ewski, B. , Biochem. J. 12.., 205 (1931) . 4. Stephenson, M. , and Stickland, L. H. , Biochem. J. 12.., 205 (1931). 5. Stephenson, M. , and Stickland, L. H. , Biochem. J. 1.§., 712 (1932) .
6. Tamiya, N. Y., Yamaguchi, Y., Honya, M., and Yagi, T., Biochem. Biophys. Res. Commun. 1£, 43 (1966)~
7. Smith, G. N., and Marshall, R. G., Arch. Biochem. Biophys. ~, 395 (1952).
8. Krasna, A. I., and Rittenberg, D. J., Am. Chem. Soc. lls 3015 (1954) .
9. Peck, H. D., and Gest, H., J. BacterioL- ll, 70 (1956).
10. Gray, C. T., and Gest, H., Science 148, 186 (1965).
11. Koepsell, H. J., and Johnson, M. J., J. BioI. Chem. 145, 379 (1942) .
12. Wilson, J., Krampitz, L. 0., and Werkman, G. H., Biochem. J. 42, 598 (1948).
13. Krampitz, L. 0., Suzuki, I., and Greu11, R., Federation Proc. 20, 970 (1961).
14. Whitely, H. R., and McCormick, N. G., J. Bacterio1. 85, 382 (1963).
15. Peck, H. D., and Gest, H., J~ Bacterio1. 11, 569 (1957).
16. Peck, H. D., and Gest, H., J. Bacterio1. 11, 70 (1956).
17. Mortensen, L. E., Valentine, R. C., and Carnahan, J. E., Biochem. Biophys. Res. Commun. 2, 448 (1962).
18. Ibid., J. BioI. Chem. 238, 794 (1963).
19. San Pietro, A., 1961, "A Symposium on Light and Life," p. 631, John Hopkins Press, Baltimore, Md.
20. Tagawa, K., Arnon, D. I., Nature 195, 537 (1962). 21. Buchanan, B. B., Lovenberg, W., and Rabinowitz, J. C., Proc. Nat1. Acad. Sci. U.S. 49, 345 (1963).
22. Valentine, R. C., Jackson, R. L., and Wolfe, R. S., Biochem. Biophys. Res. Commun. 7, 453 (1962).
23. Valentine, R. C., Brill, W. J., and Wolfe, R. S., Proc. Nat1. Acad. Sci. U.S. 48, 1856 (1962).
24. Strecker, H. J., J. BioI. Chem. 189, 815 (1951).
25. Gest, H., and Peck, H. D., J. Bacterio1. ]1, 326 (1957).
26. Ibid., J. Bacterio1. ]1, 706 (1957). ,.
27. Linnane, A. W., and Wrigley, C. W., Biochim. Biophys. Acta 12, 408 (1963).
28. Wimpenny, J. W. T., Ran1ett, M., and Gray, C. T., Biochim. Biophys. Acta 11, 170 (1963). 29. Gray, C. T., Hughes, D. E., and Ran1ett, M., Biochim. Biophys. Acta &1, 157 (1963). 30. Mortensen, L. E. , Ann. Rev. Microbiol. g, 115 (1963) .
31. Valentine, R. C. , and Wolfe, R. S. , J. Bacterio1. .§1, 1114 (1963). 32. Ishimoto, M. , Kondo, Y. , Kameyama, T. , Yagi, T. , and Shiraki, M. , in "Proc. Intern. Symp. Enzy. Chem. " (Tokyo) , 1958, p. 229.
33. Lindstron, E. S., Lewis, S. M., and Pinsky, M., J. Bacterio1. &1, 481 (1951).
34. Wilson, P. W., and Burris, R. H., Bacteriol. Rev. 11, 41 (1947).
35. Stone, R. W., and Wilson, P. W., J. Bacterio1. 63, 605 (1961).
36. Repaske, R., and Wittenberger, C. L., Biochim. Biophys. Acta iL, 542 (1961).
37. Packer, L., and Vishniac, W., Bacterio1. Proc., p. 116 (1955).
38. Ibid., Biochim. Biophys. Acta 17, 153 (1955).
39. Repaske, R., Biochem. Biophys. Res. Commun. 30, 136 (1968).
40. Omerod, J. G., and Omerod, K. S., Arch. Biochem. Biophys. 2i, 449 (1961).
41. Gest, H., and Kamen, M. D., Science 109,558 (1949). 166
42. Bregoff, H. M., J. Bio1. Chem. 182, 153 (1950).
43. Gest, H., Omerod, J. G., and Omerod, K. S., Arch. Biochem. Biophys. 22, 21 (1962).
44. Omerod, J. G., and Gest, H., Bacterio1. Rev. 26, 51 (1962).
45. Gest, H., in "Bacterial Photosynthesis," Gest, H., San Pietro, A., and Vernon, L. P. (ed.), Antioch Press, Yellow Springs, Ohio, 1963, p. 129.
46. Bose, S. K., and Gest, H., in "Energy Linked Functions of Mitochondria," Chance, B. (ed.), Academic Press, New York, 1963, p. 209.
47. Arnon, D. 1., in "Nonheme Iron Proteins," San Pietro, A. (ed.), Antioch Press, Yellow Springs, Ohio, 1965, p. 137.
48. Kubowitz, F., Biochem. Zeit. 274, 285 (1934).
49. Kamen, M. D., and Gest, H., Science 109,560 (1949).
50. Ibid., Arch. Biochern. Biophys. ~, 202 (1952).
51. Shethna, Y. L., Wilson, R. W., Hansen, R. E., and Beinert, H., Proc. Nat1. Acad. Sci. U.S. 52, 1263 (1964).
52. Mortensen, L. E., in "Nonheme Iron Proteins," San Pietro, A. (ed.), Antioch Press, Yellow Springs, Ohio, 1965, p. 243.
53. Burris, R. H., Winter, H. C,) Munson, T. 0., and Garcia-Riveria, J., Ibid., p. 315. - 54. Lee, S. B., and Wilson, P. W., J. BioI. .Chern. 151, 377 (1943).
55. Gest, H., Bacterio1. Rev. 15, 183 (1951).
56. Bulen, W. A., LeComte, J. R., Burns, R. C., and Hinkson, J., in "Nonherne Iron Proteins," San Pietro, A. (ed.), Antioch Press, Yellow Springs, Ohio, 1965, p. 261.
57. Hardy, R. W. F., Knight, E., D'Eustachio, A. J., and McDonald, C. C., Ibid., p. 275.
58. Mortenson, L. E., Biochim. Biophys. Acta 127, 18 (1966).
59. Burns, R. C., and Bu1ern, W. A., Proc. Nat1. Acad. Sci. U.S. ~, 979 (1966).
60. Burris, R. H., and Winter, H. C., J. Bio1. Chern. 243, 940 (1968). 61. Linstrom, E. S., Burris, R. H., and Winter, H. C., J. Bacterio1. 58, 313 (1949).
62. Cole, J., and Wimpenny, J. W. T., Biochim. Biophys. Acta 128, 419 (1966) •
63. Gest, H., Bacterio1. Rev. 18, 43 (1954).
64. Waring, W. S., and Werkman, C. H., Arch. Biochem. Biophys. i, 75 (1944) .
65. ~ap~enheimer, A. M., and Shaskan, E., J. Bio1. Chern. 155, 265 (1944).
66. Bulen, W. A., Burns, R. C., and LeComte, J. R., Biochem. Biophys. Res. Commun. 17, 265 (1964).
67. Bulen, W. k., in "Nonheme Iron Proteins," San Pietro, A. (ed.), Antioch Press, Yellow Springs, Ohio, 1965, p. 261.
68. Burns, R. C., and Bulen, W. A., Arch. Biochem. Biophys. 113, 461 (1962) .
69. Krasna, A. I., and Rittenberg, D., J. Am. Chern. Soc. 1£, 3015 (1954).
70. Rittenberg, D., in "International Symposium on Enzyme Chemistry," (Tokyo), 1958, p. 256.
71. Peck, H., Gest, H., and San Pietro, A., Proc. Nat1. Acad. Sci. U.S. 42, 13 (1956).
72. Krasna, A. I., and Aramed, S. R., Biochem. J. i, (1965).
73. Krasna, A. I., Rik1is, E., and Rittenberg, D., J. Bio1. Chern. 235, 2717 (1960).
74. Farkas, L., and Schneidmesser, B., J. Bio1. Chern. 167, 807 (1947).
75. Krasna, A. I., and Rittenberg, D., Proc. Nat1. Acad. Sci. U.S. 40, 225 (1954).
76. Hoberman, H., and Rittenberg, D., J. Bio1. Chern. 147, (1943).
77. Hyndman, L. A., Burris, R. H., and Wilson, P. W., J. Bacterio1. 65, 522 (1953).
78. Gest, H., and Kamen, M. D., J. Bio1. Chern. 182, 153 (1950).
79. Warburg, 0., in "Heavy Metal Prosthetic Groups and Enzyme A~tion," Oxford University Press, Oxford, England, 1949, Chap. 18. 168
80. Shug, A. L., Wilson, P. W., Green, D. E., and Mahler, H. R., J. Am. Chern. Soc. 12.., 3355 (1954).
81. Riklis, E., and Rittenberg, D., J. Biol. Chern. 236, 2526 (1961).
82. Tamiya, W., Yamaguchi, Y., Honya, M., and Yagi, T., Biochem. Biophys. Res. Commun. ~, 43 (1966).
83. Krasna, A. I., Riklis, E., and Rittenberg, D., J. Biol. Chern. 235, 2717 (1960).
84. Atkinson, D. E., and McFadden, B. A., J. Biol. Chern. 210, 885 (1954).
85. Fischer, H. F., Krasna, A. I., and Rittenberg, D., J. Bio1. Chern. 209, 569 (1954).
86. Bone, D. H., Bernstein, S., and Vishniac, W., Biochim. Biophys. Acta &2, 581 (1963). 87. Bone, D. H., Biochim. Biophys. Acta &2, 589 (1963). 88. Sadana, J. C., and Rittenberg, D., Proc. Nat1. Acad. Sci. U.S. 1Q., 900 (1963).
89. Dixon, M., and Webb, E. C., in "Enzymes," Academic Press, Inc., New York, 1964, p. 150.
90. Ackrel1, B. A. C., and Mower, H. F., in "Nonheme Iron Proteins," San Pietro, A. (ed.), Antioch Press, Yellow Springs, Ohio, 1965, p. 283.
91. Martin, R. G., and Ames, B. N., J. Bio1. Chern. 236, 1372 (1961).
92. Andrews, P., Biochem. J. 91, 222 (1964).
93. Duchow, E:-, and Douglas, H. I., J. Bacterio1. 58, 409 (1944).
94. Bergi, E., and Hershey, A. D., Biophys. J. ~, 309 (1963).
95. Kidman, A. D., Ackre11, B. A. C., and Asato, R. N., Biochim. Biophys. Acta, in press, June (1968).
96. Ha1sa11 , H. B., Nature 215, 880 (1967).
97. Kidman, A. D., Ackre11, B. A. C., and Asato, R. N., Biochim. Biophys. Acta 159, 185 (1968).
98. Kidman, A. D., Thesis, University of Hawaii, 1968.
99. Monty, K. J., and Seigel, L. M., Biochim. Biophys. Acta 112, 346 (1966). 100. Ackers, G. K., Biochemistry .1, 723 (1964).
101. Laurent, T. C., and Ki11ander, J., J. Chromatog: 14, 317 (1964).
102. Cohn, E. J., and Edsall, J. R., in "Proteins, Amino Acids and Peptides," Reinhold, New York, 1943, p. 402.
103. Andrews, P., Biochem. J. ~, 595 (1965).
104. Carnahan, J. E., and Castle, J. E., J. Bacterio1. 75, 121 (1958).
105. Hedrick, J. L., and Smith, A. J., "Abstracts of the Pacific Slope Biochemical Conference," Davis, California, p. 56 (1967).
106. Jovin, T., Chrambach, A., and Naughton, M., Ana1yt. Biochem. 2,., 351 (1964).
107. West, E. S., Todd, W. R., Mason, H. S., Van Bruggen, J. T., in "Textbook of Biochemistry," Macmillan Co., New York, 1966, p. 468.
108. Rogosa, M., J. Bacterio1. 87, 162 (1964).
109. Robrish, S. A., and Marr, A. G., J. Bacterio1. 83, 158 (1962).
110. Sober, H. A., Guther, F. G., Wyckoff, M. M., and Petersen, E. A., J. Am. Chem. Soc. ~, 756 (1956).
111. Green, M., and Wi1s9n, P. W., J. Bacteriol. ~, 511 (1953).
112. Valentine, R. C., Bacterio1. Rev. 28,497 (1964).
113. Spruit, C. J. P., in "Physiology and Biochemistry of Algae," R. A. Lewin (ed.), Academic Press, New York, 1962, p. 48.
114. Bulen, W. A., J. Bacterio1. ~, 130 (1962).
115. Okuda, K., Edwards, G. C., and Winnick, T., J. Bacterio1. 85, 239 (1963).
116. Rabinowitz, J., "Methods in Enzymology," &., 703, Academic Press, New York, 1963.
117. Barker, H. A., "Bacterial Fermentations," John Wiley and Sons, Inc., New York, 1957, p. 1.
118. Clark, J. M., "Experimental Biochemistry," W. H. Freeman and Co., San Francisco, 1964, p. 101.
119. Cohen-Bazire, G., Sistrom, W. R., and Stanier, R. Y., J. Cell. Compo Physio1. 49, 25 (1957). 170
120. Cohen, J. S., and Burris, R. H., J. Bacteriol. ~, 316 (1955).
121. Shatz, A., and Bove11, C., J. Bacterio1. 63, 87 (1952).
122. Smillie, P., Naturwissenschaften 48,136 (1961).
123. Doudoroff, D., Enzymologia 1, 59 (1940).
124. Winogradsky, P., J. Bacteriol. ~, 160 (1953).
125. Harvey, E. N., "Bioluminescence,1I Academic Press, Inc., New York, 1952.
126. Baker, F. D., Papiska, H. R., and Campbell, L. L., J. Bacterio1. . 84, 973 (1962).
127. Ko1hmi11er, E. F., and Gest, H., J. Bacterio1. ~, 269 (1951).
128. Whitely, H. R., and Douglas, H. C., J. Bacterio1. 61, 605 (1951).
129. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J. BioI. Chern. 193, 265 (1951).
130. Korkes, S., J. BioI. Chern. 216, 737 (1955).
131. Arnon, D. I., and Tagawa, K. I., Nature 195, 400 (1962).
132. Akagi, J. M., and Campbell, L. L., J. Bacterio1. ~, 927 (1961).
133. Webb, J. L., "Enzymes and Metabolic Inhibitors, II Academic Press, New York, 1963, Vol., p. 792.
134. Ackre11, B. A. C. (Unpublished results, this laboratory).
135. Ashton, G. C., and Braden, A. W. H., Aust. J. BioI. Sci. 14, 248 (1961) .
136. Kisteakowsky, G. B., and Lumry, R., J. Amer. Chern. Soc. 11, 2006 (1949).
137. Kavanau, J. L., J. Gen. Physio1. 34, 193 (1950).
138. Maier, V. P., Tappe1, A. L., and Vo1nen, D. H., J. Amer. Chern. Soc. 12, 1278 (1955). 139. Sheraga, H. A., Nemethy, G., Schrier, E. G., Schneider, H., and Krischeck, G. C., Fed. Proc. 23, 687 (1964).
140. Braun, A. D., and Feldman, N. L., Compt. Rend. Acad. Sci. U.R.S.S. ~, 757 (1949). 171
141. Miller, E. C., and Miller, J. A., Cancer Res. 12, 547 (1952).
142. Wolfe, R. S., and OlKane, D. J., J. Biol. Chern. 205, 755 (1953). BIOGRAPHICAL
1942 --Born in Lahaina, Hawaii. 1964 --B.S., Lewis and Clark College, Portland, Oregon. 1964-1966 --Graduate Assistant, Department of Biochemistry, University of Hawaii. 1966-1968 --National Institutes of Health Trainee, Department of Biochemistry, University of Hawaii.
GRADUATE STUDIES
Graduate Field: Biochemistry. University of Hawaii, Honolulu, Hawaii. General Biochemistry. Professor Theodore Winnick. Topics in Biophysics. Professor Lawrence H. Piette. Advanced Enzymology. Associate Professor George A. Barber. Survey of Intermediary Metabolism. Associate Professor Howard F. )19~~~. Nucleic Acids. Asd~ate Professor John B. Hall. Other Studies. Ultrastructure Microbiology. Associate Professor Hans Hohl., PUBLICATIONS
"Multiple Forms of Bacterial Hydrogenases." (With B. A. C. Ackre11 and H. F. Mower). d. Bacterio1. ~:828. 1966. "Properties of 'L'hree Isoenzymes of Q. Easteurianum Hydrogenase." (With A. D. Kidman and B. A. C. Ackrell). Biochim. Biophys. ~. In press.
"Properties of Bacterial Hydrogenases. II (VT1:th A. D. Kidman and B. A. C. Ackre11). ~.~. 1l:589. 1968. "Comparative Studies of Hydrogenase from Various Bacteria." (With A. D. Kidman and R. Ya~gihara). In process of publication.