THE PURIFICATION, CHARACTERIZATION,
AND MECHANISM OF ACTION OF
ORNITHINE CYCLASE (DEAMINATING)
FROM CLOSTRIDIUM SPOROGENES

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
WILLIAM LAWRENCE :MUTH

1972

 

 

This is to certify that the

thesis entitled

\

The Purification, Characterization, and Mechanism
of Action of Ornithine Cyclase (Deaminating)
from Clostridium sporogenes

presented by

William Lawrence Muth

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Microbiology and
Public Health

 

I Ma or professor

Date October 27, 1972

0-169

  
   

  

“ BIJIIK BINDERY mc. ‘
LIBRARY BINDERS 1 I
l senmsgoepflcmcgfl

ABSTRACT

THE PURIFICATION, CHARACTERIZATION, AND
MECHANISM OF ACTION OF ORNITHINE CYCLASE
(DEAMINATING) FROM CLOSTRIDIUM SPOROGENES

 

BY

William Lawrence Muth

Ornithine cyclase (deaminating) from Clostridium

 

sporogenes was purified to electrophoretic and ultracentri-

 

fugal homogeneity. The pure cyclase had a sedimentation
coefficient (£20,w) of 5.65 measured by high speed sedi-
mentation velocity centrifugation, and a molecular weight

of 81,000 calculated from high speed equilibrium sedimenta-
tion. Amino acid analysis calculations indicated a molecular
weight of 80,000 and a partial specific volume of 0.733 cm3
per gm for the protein. Sodium dodecyl sulfate poly-
acrylamide gel electrophoresis indicated the enzyme was
made up of two equal subunits with a molecular weight of
41,500. The ultraviolet absorption spectrum had a major
absorption peak at 267 nm which suggested the presence of
a nucleotide. Assay of the pure enzyme for bound NAD+

demonstrated the presence of approximately 1 mole of bound

NAD+ per mole of enzyme. Pure enzyme was stimulated by

William Lawrence Muth

NAD+ and either ADP or ATP. The enzyme was inhibited by
several sulfhydryl group inhibitors. Titration of the
enzyme with p-chloromercuribenzoate established the presence
of four sulfhydryl groups per mole. The conversion of
L-ornithine to L-proline by ornithine cyclase involves the
deamination of the a-NH3 group prior to cyclization.
Therefore, 2-keto-5-aminopentanoic acid and Al-pyrroline-
2-carboxylic acid are likely intermediates in the conversion.
Pure cyclase failed to incorporate any tritium from NADT
into proline when incubated in the presence of ornithine

alone or with Al-pyrroline-Z-carboxylic acid.

THE PURIFICATION, CHARACTERIZATION, AND
MECHANISM OF ACTION OF ORNITHINE CYCLASE

(DEAMINATING) FROM CLOSTRIDIUM SPOROGENES

 

BY

William Lawrence Muth

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Microbiology and Public Health

1972

[A DEDICATION

This thesis is dedicated to my wife and
parents, who through their encouragement
and support made my entire doctoral

program possible.

ii

ACKNOWLEDGMENTS

I would like to sincerely thank my major advisor,
R. N. Costilow, for his constant enthusiasm, encouragement
and discussion of the material presented in this thesis.
Thanks also go to my officemate, R. L. Somack, for many
long discussions concerning the thesis topic. A special
acknowledgment is due W. C. Deal, Jr., W. A. Wood, and
C. C. Sweeley, all of the Department of Biochemistry,
Michigan State University, for their help in the comple-
tion of the ultracentrifuge studies, the amino acid

analyses, and the mass spectrometry studies, respectively.

iii

LIST OF TABL

LIST OF FIGU

INTRODUCTION

ES 0 0

RES . .

LITERATURE REVIEW .

Part I:
Part II:
Part III
Part IV:
Part V:

Part VI:

REFERENCES

ARTICLE 1:

ARTICLE 2:

TABLE OF

CONTENTS

The Stickland Reaction

Energy Considerations

: Pathways for the Fermentation of

Ornithine
Conversion of Ornithine to Proline

Possible Roles for NAD+ in the

Enzyme Reactions .

Possible Mechanisms for the

Ornithine Cyclase Reaction

ORNITHINE CYCLASE (DEAMINATING):
PROPERTIES OF THE HOMOGENEOUS ENZYME
REMOVAL OF THE ALPHA-AMINO GROUP
DURING THE CONVERSION OF ORNITHINE

TO PROLINE

iv

Page

vi

10

12

13

18

22

60

LIST OF TABLES

TABLE Page
ARTICLE 1
I. Purification of Ornithine Cyclase . . . . . . 46
II. Amino Acid Analysis of Ornithine Cyclase . . 47
III. Estimation of Protein-Bound NAD+ . . . . . . 48
IV. Effect of Various Compounds on Enzyme

Activity . . . . . . . . . . . . . . . . . . 49
V. Effect of Sulfhydryl Group Inhibitors

on Enzyme Activity . . . . . . . . . . . . . 51

ARTICLE 2
I. Incorporation of 15N into Proline . . . . . . 81
II. Incorporation of 15N into Glutamate . . . . . 82

III. Incorporation of Tritium from NADT to
Proline with Ornithine Alone or with
Al-Pyrroline-Z-carboxy1ic Acid (2-PCA)

as substrate 0 O O O O C O O O O O C O O O C 83

FIGURE

LIST OF FIGURES

Page
LITERATURE REVIEW
Oxidative and reductive pathways for the
fermentation of ornithine . . . . . . . . . . 8

Proposed mechanism of action of UDP-D-
glucose 4'—epimerase (A) and TDP-D-

glucose oxidoreductase (B). . . . . . . . . . 16

ARTICLE 1
Preparative polyacrylamide gel electro-
phoresis of ornithine cyclase . . . . . . . . 53
Gel scan of the purified cyclase in a
7% analytical polyacrylamide gel taken
at a wavelength of 600 nm . . . . . . . . . . 55
Plot of fringe displacement (3 + log AY)
gs. the square of the distance from the
center of rotation (radiusz). . . . . . . . . 57

UV scan of homogeneous cyclase . . . . . . . 59

vi

FIGURE

Page
ARTICLE 2

Gas chromatography of the trimethylsilylated
amino acids separated from a reaction
mixture containing unlabelled ornithine,
proline, glutamate, and ornithine cyclase . . 85
Comparison of part of the mass spectrum
of (TMS)2-proline derived catalytically
from unlabelled ornithine (a) with that
of proline from [6-15N]-ornithine (b) . . . . 87
Comparison of part of the mass spectra of
(TMS)3-glutamate derived catalytically
from a-ketoglutarate and the ammonia
generated from either unlabelled (a) or
[6-15N1- (b) ornithine . . . . . . . . . . . 89
Proposed mechanism of action of ornithine

cyclase . . . . . . . . . . . . . . . . . . . 90

vii

INTRODUCTION

The putrefactive anaerobes of the genus Clostridium

 

ferment amino acid pairs as sole substrates for the pro-
duction of energy. Early work indicated that when such
amino acids as alanine, leucine, and valine were oxidized
there was a concomitant reduction of glycine or proline.
Ornithine has been found to function as a hydrogen acceptor

for Clostridium sporogenes and is reduced to O-aminovaleric

 

acid. However, in g. sticklandii it is primarily oxidized

 

to NH C02, and volatile fatty acids. 9. botulinum can

 

3'
ferment ornithine as a single substrate to C02, NH3, volatile

fatty acids, and O-aminovaleric acid utilizing both oxidative
and reductive pathways.

The oxidation of D-ornithine by g. sticklandii pro-

 

ceeds as follows:

CHZNHZCHZCHZCHNHZCOOH'——* CHBCHNHZCHZCHNHZCOOH (1)
CH3CHNH2CH2CHNH2COOH-—+ CH3COCH2CHNH2COOH + NH3 (2)
CH3COCH2CHNH2COOH-——+ CH3COOH + CH3CHNH2COOH (3)

Reaction (1) involves a cobamide and pyridoxal phosphate-

requiring mutase enzyme which transfers the O-amino group

to carbon 4 forming 2,4-diaminopentanoic acid. In reaction (2)
an NAD+-requiring dehydrogenase yields 2-amino-4-ketopentanoic
acid and ammonia. The enzymes catalyzing the conversion
indicated in (3) have not been studied. Presumably, there

is a CoA-dependent thiolytic cleavage of the keto-acid to

form acetyl-CoA and alanine followed by phosphorolysis of

the acetyl-GOA to acetyl phosphate with subsequent transfer

of the high energy phosphate to ADP to form ATP.

The reduction of L-ornithine by g. sporogenes involves

 

the following reactions:

2 2 2 ZCHNI-IZCOOH-"—'t NH3 + L-I: ]COOH (4)

CH NH CH CH

N
H
L- [ JCOOH —* D- [ ]COOH (5)
N N
H H
D- N COOH—+ CHZNH2 (CH2) 3COOH (6)
H

L-ornithine is converted in a complex NAD+-requiring
reaction to ammonia and L-proline (reaction 4) which is
racemized to the D-form (reaction 5). The final step
(reaction 6) involves a cleavage of the proline ring with
an NADH-dependent reductase to form O-aminovaleric acid.
The enzyme catalyzing the conversion of ornithine

to proline (ornithine cyclase) can be obtained from

either 9, botulinum or g. sporogenes in good quantity. It

 

 

has been purified about 2-fold by fractionation with
ammonium sulfate and dialysis against ornithine-phosphate
buffer with no loss of activity. Further purification
using ion exchange and molecular sieving columns has
resulted in considerable purification but extreme loss of
activity in spite of numerous attempts at stabilization of
the enzyme with substrate, cofactor, reducing agents, and
glycerol. The cyclase is stimulated by both NAD+ and ADP.
However, the partially purified enzyme appears to contain
tightly bound NAD+. No other cofactors have been identi-
fied.

Specific transaminase activity was not found in
partially purified preparations catalyzing the conversion.
Therefore, ornithine is probably deaminated and cyclized
by an oxidation-reduction mechanism. However, any semi-
aldehyde or a-keto acid intermediate formed in the course
of this reaction was probably bound to the enzyme, since
attempts to trap such intermediates were unsuccessful.

The conversion of ornithine to proline by a single enzyme
or enzyme complex appeared to be both unique and complex.
Therefore, we decided to purify the cyclase to homogeneity
and examine some of its more salient physical properties.
In addition, some aspects of the catalytic mechanism of

the cyclase were examined.

This thesis is organized into three sections.

The first section is a review of the pertinent literature
in which ornithine cyclase is discussed in relation to the
pathways involved in ornithine fermentation, energy produc-
tion, and possible mechanisms for the reaction. The second
section is a manuscript describing the purification and
properties of the enzyme which has been prepared for publi-
cation. The final section contains a manuscript describing
experiments to determine the mechanism of action of the

cyclase.

LITERATURE REVIEW

Part I
The Stickland Reaction

The non-saccharolytic clostridia are known to
ferment amino acid pairs as sole substrates for the pro-
duction of energy (Stickland, 1934). These fermentations
are collectively known as Stickland reactions, and involve
the oxidation of one of the amino acids in the pair while
the other is reduced (Woods, 1936). For each mole of
amino acid such as alanine, leucine, isoleucine, valine,
222! that is oxidized, two moles of glycine, proline or
hydroxyproline are reduced (Stickland, 1935). For instance:

-4H

CHBCHNHZCOOH + 2H20 > NH3 + CH3COOH + CO2

 

2CH2NH2COOH + 4H-——+ ZNH3 + 2CH3COOH

CHBCHNHZCOOH + ZCHZNHZCOOH + 2H20-——+ 3NH3 + 3CH3COOH + CO2

For a more complete review of Stickland reactions, see

recent reviews by Nisman (1954) and Barker (1961).

Part II
Energy Considerations

It is possible for cells to derive energy in two
ways from Sticland reactions. The first mechanism has not
been extensively studied but it probably involves substrate
level phosphorylation of one or more of the products of an
oxidative pathway. For example, when alanine is oxidized
and deaminated, pyruvate is formed. Pyruvate is presumably
thiolytically cleaved by pyruvate dehydrogenase forming
CH3COOSCoA and CO2 since CoASH is required for activity of
the enzyme. Inorganic phosphate could then be incorporated
as CH3COO<:)by phosphate acetyltransferase with the loss of
CoASH. The high energy phosphate may be transferred to
ADP to form ATP and CH3COOH by acetate kinase.

The second mechanism for the production of energy
occurs with the reduction of one of the amino acids in the
pair (Stadtman, 1967). The reduction of glycine has been
carefully shown to be linked to the esterification of ortho-
phosphate into ATP probably through an anaerobic electron
transport system containing ferredoxin and an acidic low
molecular weight protein (Stadtman et al., 1958; and Stadt-
man, 1966). Early papers describing the proline reductase
step either did not discuss or disclaimed any ATP formation

associated with this step (Stadtman, 1956; Stadtman and

Elliot, 1957). However, more recent data indicate that ATP

is formed during reduction of proline to 6-aminovaleric
acid (Stadtman, 1962; Stadtman, 1967). The system appears
to be so labile that the stoichiometry is poor and the
mechanism is not understood. Hydroxyproline or tryptOphan
reduction have not been examined as sources of orthophosphate
esterification.
Part III

Pathways for the Fermentation of Ornithine

Ackermann (1907) showed that ornithine was converted
to O-aminovaleric acid when added to mixed cultures of putre-
factive bacteria. Woods (1936) clearly demonstrated ornithine
to be a hydrogen acceptor when fermented with alanine by

C. sporogenes and that O-aminovaleric acid and ammonia were

 

the end products. In contrast, ornithine was fermented

primarily as an electron donor by g. sticklandii in the

 

presence of lysine (Stadtman: 1954; Stadtman and White,

1954). In whole cells and extracts of E. botulinum, ornithine

 

was fermented as a single substrate with the production of
C02, NH3, acetate, propionate, butyrate, valerate, and
O-aminovaleric acid (Mitruka and Costilow, 1967).

The oxidative pathway for ornithine utilization has

been partially elucidated in E. sticklandii (Figure l).

 

The first reaction in the sequence is a cobamide-requiring
mutase step which transfers the O-amino group to carbon 4

forming 2,4-diaminopentanoic acid (Dyer and Costilow, 1970;

D,L-CH NH CH CH CHNH COOH

2 2 2 2 2

OXIDATIVE REDUCTIVE

I I

D-CH3CHNH2CH2CHNH2COOH

 

L—
COOH + NH
-2H+ 3
H
I
D-CHBCOCHZCHNHZCOOH + NH3 D_
COOH
—2H+ N
H
CH3COOH + CH3CHNH2COOH +2H+

CHZNHZCHZCHZCHZCOOH

FIGURE 1. Oxidative and reductive pathways for the

fermentation of ornithine

Tsuda and Friedmann, 1970; Somack et 31., 1971; Somack and
Costilow, submitted for publication). The second step is
an NAD+-requiring oxidative deamination yielding 2-amino-
4-ketopentanoic acid and ammonia (Somack and Costilow,
1972). Previously, a pyridoxal phosphate stimulated
epimerization was postulated to occur prior to the C4
dehydrogenase step (Tsuda and Friedmann, 1970), but more
recent evidence indicates that the pyridoxal phosphate is
a cofactor in the mutase reaction (Somack and Costilow,
submitted for publication). The final step is the CoA-
requiring thiolytic cleavage of 2-amino-4-ketopentanoic
acid by several enzymes to form the primary products,
acetate, alanine, and ATP (Dyer and Costilow, 1968).

The reductive pathway of ornithine utilization has

been elucidated in extracts of g, botulinum, C. sporogene§_

 

 

and g. sticklandii cells (Figure l). The first step of the

 

sequence involves a complex NAD+-requiring deamination and
cyclization to form L-proline (Costilow and Laycock, 1968,
1969, 1971). More will be said concerning this reaction
later (see Part IV below). The second step is the racemi-
zation of L-proline to D—proline (Stadtman and Elliott,
1957). The final step involves the cleavage of the proline
ring and reduction to O-aminovaleric acid (Stadtman, 1956;
Stadtman and Elliott, 1957). This final reaction is com—
plex and requires a number of cofactors for activity

(Stadtman, 1962; Hodgins and Abeles, 1967).

10

Part IV
Conversion of Ornithine to Proline

The conversion of ornithine to proline in fresh

extracts of Q, botulinum cells was first reported by Costilow

 

and Laycock (1968). Considerable evidence was presented
that proline was an intermediate in the conversion of orni—
thine to O-aminovaleric acid. Thus, (a) proline accumulated
in reactions containing excess ornithine but limited reducing
power, but did not accumulate with limited substrate and
excess reducing power; (b) the rate of accumulation of pro-
line plus O-aminovaleric acid in a system with limited
reducing power was the same as the accumulation of 6-amino-
valeric acid when excess reducing capability was present;
(c) when proline reductase was inhibited, proline accumulated
at the same rate as 6-aminovaleric acid with no inhibitor
present; (d) when unlabelled proline and [14C]-U-ornithine
were added to reaction mixtures, radioactive proline
accumulated until the exogenous proline supply was exhausted;
and (e) when small amounts of extract and limited reducing
supply were incubated for short periods of time (2 min)
with high specific activity [14C1-ornithine, a small but
significant amount of radioactive proline accumulated,
whereas essentially no O-aminovaleric acid was formed.
Compounds which stimulated the conversion of orni-

thine to proline included NAD+ and ADP (Costilow and

11

Laycock, 1968), although ADP stimulation could no longer
be demonstrated after a 2-fold purification. No other
cofactor was found, but FAD and NADP appeared to partially
substitute for NAD+ in stimulating the activity in more
purified enzyme preparations (Costilow and Laycock, 1971).
The conversion of ornithine to proline in a number
of other microorganisms and animal tissues proceeds via
ornithine-O-transaminase to glutamic-Y-semialdehyde, which
is in equilibrium with Al-pyrroline-5-carboxylic acid
(5-PCA), and thence, through NADH or NADPH-dependent reduc-
tion to proline (Stetten, 1955; Scher and Vogel, 1957).

Evidence for a similar pathway was sought in Q. botulinum

 

and Q, sporogenes without success. Although 5-PCA was

 

produced from proline in reaction mixtures using crude
extracts, and 5-PCA could be rapidly reduced by these
extracts when NADH was present, further purification of the
extract demonstrated that 5-PCA reductase was not involved
in the conversion of ornithine to proline (Costilow and
Laycock, 1969, 1971). The formation of proline from orni-
thine was shown to be carried out by a single enzyme or
enzyme complex, and the enzyme was given the trivial name
of ornithine cyclase (deaminating). The catalytic protein
was unstable after purification through DEAE cellulose
column chromatography. The enzyme at this stage of purity

was characterized as being specific for L-ornithine and

12

producing L-proline in an essentially irreversible reaction.
The apparent Km for ornithine was 11.1 mM which indicated a
rather low affinity of the enzyme for substrate. On the
other hand, the apparent Km for NAD+ was 6.1 x 10-3 mM
indicating a strong affinity. The enzyme was most active
at pH 8.0 and at 44°. Increasing levels of D-ornithine
appeared to inhibit the enzyme but L-lysine, L-alanine,
L-leucine or isoleucine and L-proline did not significantly
affect the activity. No tritium exchange took place when

the reaction was run in the presence of T O. No intermediate

2
of a pyrroline nature could be trapped in the course of the
reaction. The presence of hydrazine or semicarbazide did
not inhibit the reaction, thereby indicating a lack of
either pyridoxal phosphate or bound pyruvate which could
possibly be implicated in Schiff base formation.

Part V

Possible Roles for NAD+ in
Enzyme Reactions

NAD+ is involved in reactions catalyzed by a
number of different kinds of enzymes including dehydro-
genases, epimerases, and oxidoreductases. Its action may
be catalytic in nature as it is in the vast majority of
enzymes in which NAD+ is found. However, NAD+ may also
function as an activator for an enzyme and not be directly

involved with catalysis. For instance, nicotinamide

l3

adenine dinucleotide phosphate-ferredoxin reductase from

g. kluyveri which catalyzes the reduction of ferredoxin by
NADPH is controlled by the oxidation state of the NAD+—NADH
couple (Thauer 33 31., 1971). NAD+ serves as an obligatory
activator for this enzyme. As expected, NADH is an inhibitor,
but is competitive with NAD+ rather than NADPH.

Most NAD+-dependent enzymes have an apparent Km for
the cofactor in the range of 0.01 to 1.0 mM; and thus, have
a relatively low affinity for it. On the other hand, there
are several enzymes which display very low apparent Km's
for their cofactors (0.1 uM to 1 HM), and have NAD+ tightly
bound to the enzyme. Velick et_gl. (1953) have shown that
NAD+ is tightly bound to glyceraldehyde-3-phosphate dehydro-
genase. Other enzymes known to contain tightly bound NAD+
include UDP-D-glucose 4'—epimerase from either E. 321i
(Wilson and Hogness, 1964, 1969) or yeast (Darrow and Rod-
strom, 1968) and TDP-D-glucose oxidoreductase (Wang and
Gabriel, 1969); Zarkowsky and Glaser, 1969). In these cases
in which NAD+ is tightly bound to the enzyme, a catalytic
role has been ascribed to the cofactor.

Part VI
Possible Mechanisms for the
Ornithine Cyclase Reaction
The conversion of ornithine to proline must proceed

via either a specific transaminase and a reductase or an

l4

oxidative deamination and a reduction (Meister, 1965;
Costilow and Laycock, 1969). Costilow and Laycock (1969,
1971) eliminated the possibility of a specific transaminase,
since they were unable to identify either pyridoxal phosphate
or any a-keto acid involvement in the conversion. The pos-
sibility that NAD+ participates in the reaction as an
essential activator controlling the reaction is remote,

since the apparent Km for NAD+ indicates a strong interaction
between enzyme and cofactor (Costilow and Laycock, 1971),
whereas a cofactor which appears to have an activator function
usually has a much larger apparent Km (Thauer §E_§l3, 1971).
In addition, the cyclase displayed normal saturation kinetics
with NAD+, and thus it is not likely that NAD+ functions

as an allosteric effector. Costilow and Laycock (1971) have
shown that the conversion of ornithine to proline proceeds
via a single enzyme or enzyme complex. Since a transamina-
tion has been ruled out, and the status of NAD+ involvement
as an activator is highly questionable, the most logical
mechanism for the conversion involves an oxidative deamina-
tion and reduction in which NAD+ participates as a coenzyme.
The oxidative deamination may occur in a fashion similar to
that which occurs in glutamate dehydrogenase (Meister, 1965)

which may be outlined as follows:

15

 

 

coon FCOOH ‘T TOOH

IHZ THZ H20 THZ

CH2 + NADP : THZ + NADPH -—:>>+ THZ + NH3 + NADPH
CHNHZ C=NH C=O

~COOH LCOOH _ COOH

Glutamate is first oxidized by NADP to an imino acid. The

imino group is replaced by oxygen originating from H O,

2
and ammonia is released.

Twozgood examples of enzymes that carry out a sequen-
tial oxidation and reduction are the nucleotide-sugar .
epimerase, UDP-D-glucose 4'—epimerase (Glaser, 1963), and
the nucleotide-sugar oxidoreductase, TDP—D-glucose oxido-
reductase (Zarkowsky and Glaser, 1969). The postulated
mechanisms for these reactions are shown in Figure 2-A and
2-B, and are the net result of data gathered in several
laboratories (see review by Glaser, 1972, for the epimerase;
and review by Glaser and Zarkowsky, 1972, for the oxido-
reductase).

The similarity between what is currently known about
the cyclase and several facts which are consistent with the
mechanisms proposed for the epimerase and the oxidoreductase
may be valuable to this discussion of possible mechanisms
for the cyclase. For instance, neither the epimerase nor

the cyclase incorporate tritium into the product when the

reactions are carried out in tritiated water. Also, one

16

can A<v ommnoEHQOI.v omoozamlalmao mo cowuom mo Sufismzooe comomoum

.Amv ommuosponoowxo omoosamlalmaa

mmOUDAOln—lwxomalmtgmvutfilmQB

 

+

 

 

.N mMDUHh

 

 

+9§TH
o_m om
moeo mo mono moeo moeo
0
mac EON mo EON mo
omznm mo¢2nm moezum +o¢z m L
mmoonqononmoe
+
taaTm
ommuosoouonflxo Omoosaunoumoa Amy
mo fl so I
WOOD 20 V Ago +glfl + mg
+ +g1m Al m0 0 mat-W
o O: o
N I a
no no no mo .1 so «so

mmmuOEHmmu.v mucosaonoumoo .ev

17

mole of NAD+ is tightly bound to both the epimerase and the
oxidoreductase and Costilow and Laycock (1968, 1971) have
indicated that NAD+ may also be bound tightly to the cyclase.
In addition, both the cyclase and the oxidoreductase appear
to catalyze irreversible rearrangements (Costilow and Lay-
cock, 1971; Gabriel and Lindquist, 1968). Finally, attempts
to trap or isolate any of the proposed intermediates or
demonstrate the formation of free NADH during reactions

with the cyclase, epimerase, or the oxidoreductase have

been unsuccessful (Costilow and Laycock, 1968, 1971; Glaser,
1972; Glaser and Zarkowsky, 1972). The inability to trap
intermediates has been partially explained for UDP—D—glucose
4'-epimerase through changes in the conformation in the
enzyme by Bertland 25 El! (1966), Bertland and Kalckar
(1968), and Kalckar et_§1, (1970). These authors have
demonstrated that the enzymatic reduction of enzyme bound
NAD+ to bound NADH occurs with a gross change in the tertiary
structure of the enzyme as measured by circular dichroism,
optical rotary dispersion, or the rate of exchange of hydro-
gens in the epimerase backbone with those in the medium.
This conformational change in structure is probably
responsible for the increased binding of the substrate
analogue, 5'-UMP, to the reduced enzyme. Similarly, con-
formational changes may occur in the cases of the oxido-

reductase and the cyclase.

10.

REFERENCES

Ackermann, D. (1907-1908), Hoppe-Seyl. Z. 54, l.

 

Barker, H. A. (1961), in The Bacteria, Vol. II,

 

Gunsalus, I. C., and Stanier, R. Y., Eds. Academic
Press, p. 151.
Bertland, A. U., Bugge, B., and Kalckar, H. M. (1966),

Arch. Biochem. Biophys. 116, 280.

 

Bertland, A. U., and Kalckar, H. M. (1968), Proc.
Nat'l. Acad. Sci., U.S.A. 61, 629.

Costilow, R. N., and Laycock, L. (1968), J. Bacteriol.

 

96, 1011.

Costilow, R. N., and Laycock, L. (1969), J. Bacteriol.

 

100, 662.

Costilow, R. N., and Laycock, L. (1971), J. Biol. Chem.

 

246, 6655.

Darrow, R. A., and Rodstrom, R. (1968), Biochemistry

 

7, 1645.

Dyer, J. K., and Costilow, R. N. (1968), J. Bacteriol.

 

96, 1617.

Dyer, J. K., and Costilow, R. N. (1970), J. Bacteriol.

 

101, 77.

18

11.

12.

13.

14.

15.

16.

17.

18.

19.
20.

21.

22.

23.

24.

19

Gabriel, 0., and Lindquist, L. C. (1968), J. Biol.
Chem. 243, 1479.

Glaser, L. (1963), Physiol. Rev. 43, 215.

 

Glaser, L. (1972), in The Enzymes, Vol. VI, 3rd Ed.,

 

Boyer, P. D., Ed., Academic Press, New York, p. 355.

Glaser, L., and Zarkowsky, H. (1972), in The Enzymes,

 

Vol. VI, 3rd Ed., Boyer, P. D., Ed., Academic Press,
New York, p. 465.

Hodgins, D., and Abeles, R. H. (1967), J. Biol. Chem.

 

242, 5158.
Kalckar, H. M., Bertland, A. U., and Bugge, B. (1970),
Proc. Nat'l. Acad. Sci., U.S.A. 65, 1113.

Meister, A. (1965) in Biochemistry_of the Amino Acids,

 

Vol. I, 2nd Ed., Academic Press, New York, p. 269.

Mitruka, B. M., and Costilow, R. N. (1967), J. Bacteriol.

 

93, 295.

Nisman, B. (1954), Bact. Rev. 18, 16.

 

Sher, W. I., and Vogel, H. J. (1957), Proc. Nat'l.

 

Acad. Sci., U.S.A. 43, 796.

 

Somack, R. L., Bing, D. H., and Costilow, R. N. (1971),

Anal. Biochem. 41, 132.

 

Somack, R. L., and Costilow, R. N. (1972), J. Biol.
Chem. in press.

Stadtman, T. C. (1954), J. Bacteriol. 67, 314.

 

Stadtman, T. C. (1956), Biochem. J. 62, 614.

 

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

20

Stadtman, T. C. (1962), Arch. Biochem. Biophys. 99, 36.

 

Stadtman, T. C. (1966), Arch. Biochem. Biophys. 113, 9.

 

Stadtman, T. C. (1967), in Proc. Int. Workshop, 5th,

 

Fredette, V., Ed., Montreal Univ. Press, Laval-des-
Rapides, Can., p. 25.
Stadtman, T. C., Elliott, P., and Tiemann, L. (1958),

J. Biol. Chem. 231, 961.

 

Stadtman, T. C., and Elliott, P. (1957), J. Biol.
Chem. 228, 983.
Stadtman, T. C., and White, F. H., Jr. (1954), J.

Bacteriol. 67, 651.

 

Stetten, M. R. (1955), in A Symposium on Amino Acid

 

Metabolism, McElroy, W. D., and Glass, H. B., Eds.,
The Johns Hopkins Press, Baltimore, p. 277.

Stickland, L. H. (1934), Biochem. J. 28, 1756.

 

Stickland, L. H. (1935), Biochem. J. 29, 288.

 

Thauer, R. K., Rupprecht, E., Ohrloff, C., Jungermann,

K., and Decker, K. (1971), J. Biol. Chem. 246, 954.

 

Tsuda, Y., and Friedmann, H. C. (1970), J. Biol.
Chem. 245, 5914.
Velick, S. F., Hayes, J. B., Jr., and Harting, J.

(1953), J. Biol. Chem. 203, 527.

 

Wang, S. F., and Gabriel, 0. (1969), J. Biol. Chem.

 

244, 3430.

38.

39.

40.

41.

21

Wilson, D. B., and Hogness, D.
9222: 239, 2469.
Wilson, D. B., and Hogness, D.
£23m, 244, 2132.

Woods, D. D. (1936), Biochem.

S. (1964), J. Biol.

S. (1969), J. Biol.

J. 30, 1934.

 

Zarkowsky, H., and Glaser, L.

244, 4750.

(1969), J. Biol. Chem.

 

ARTICLE 1

ORNITHINE CYCLASE (DEAMINATING):

PROPERTIES Cfi‘ THE HOMOGENEOUS ENZYMEl

BY

W. L. Muth1L and R. N. Costilow

Manuscript to be submitted to

Biochemistry

 

FOOTNOTES
1‘Graduate Fellow under Title IV, National Defense
Education Act. This work was submitted in partial fulfill-
ment of the requirements for the degree of Doctor of

Philosophy.

1The abbreviations used are: DTT, dithiothreitol;

SDA, sodium dodecyl sulfate; p-CMB, parachloromercuriben-

zoate.

23

ABSTRACT

Ornithine cyclase from Clostridium sporogenes was

 

purified to electrophoretic and ultracentrifugal homo-
geneity. The pure cyclase had a sedimentation coefficient
(E20,w) of 5.68 measured by high speed sedimentation
velocity centrifugation, and a molecular weight of 81,000
calculated from high speed equilibrium sedimentation.
Amino acid analysis calculations indicated a molecular
weight of 80,000 and a partial specific volume for the
protein of 0.733 cm3 per gm. Sodium dodecyl sulfate poly-
acrylamide gel electrophoresis indicated the enzyme was
made up of two equal subunits with a molecular weight of
41,500. The ultraviolet absorption spectrum had a major
absorption peak at 267 nm which suggested the presence of
a nucleotide. Assay of the pure enzyme for bound NAD+
demonstrated the presence of approximately one mole of
bound NAD+ per mole of enzyme. Pure enzyme was stimulated
by NAD+ and either ADP or ATP. The enzyme was inhibited
by several sulfhydryl group inhibitors. Titration of the
enzyme with p-chloromercuribenzoate established the

presence of four sulfhydryl groups per mole.

24

The clostridia are known to ferment a variety of
amino acid pairs as sole substrates for the production of
energy (Stickland, 1934; Woods, 1936). Woods (1936)

demonstrated that Clostridium sporogenes used ornithine

 

as a hydrogen acceptor and reduced it to O-aminovaleric

acid. In contrast, ornithine is used primarily as an

 

electron donor by Q, sticklandii (Stadtman, 1954; Stadt—
man and White, 1954). Recently, ornithine was found to

be fermented as a single substrate by whole cells and

2, NH3, volatile fatty acids,

and O-aminovaleric acid (Mitruka and Costilow, 1967).

extracts of g. botulinum to CO

 

Previous studies in this laboratory demonstrated
that proline is an intermediate in the fermentative path-
way which reductively converts ornithine to O—aminovaleric

acid in both E. botulinum and C, sporogenes (Costilow and

 

 

Laycock, 1968; 1969). More recent data indicate the

direct conversion of ornithine to L-proline and ammonia

by a single protein or enzyme complex which was partially
purified and characterized and given the trivial name,
ornithine cyclase (deaminating) (Costilow and Laycock,
1971). The present report describes procedures for obtain-
ing the cyclase in a homogeneous state and identifies some

of the more salient properties of the pure enzyme.

25

EXPERIMENTAL SECTION

Culture and Cultural Conditions. A putrefactive

 

anaerobe, Q. sporogenes (National Canners Association,
PA3679, American Type Culture Collection, ATCC 7955) was
the source of all enzyme used in this study. The cultural
conditions have been previously described (Costilow and

Laycock, 1971).

Standard Buffer Preparation. All buffers used in

 

this study were prepared in the following manner. Various
levels of Tris plus 10% (w/w) glycerol were added to dis-
tilled water to which HCl was added to give the appropriate

pH. After the buffer was briefly boiled to remove dissolved

oxygen and cooled under argon, 10-2 M ornithine, 10—3 M

1 4

DTT and 10- M NAD+ were added as stabilizers. Wherever

possible, oxygen was excluded from the buffer.

Enzyme Assay. Ornithine cyclase was assayed in
14

 

a radiochemical assay utilizing L-[U- C]-ornithine as
previously described (Costilow and Laycock, 1971). One
unit of enzyme activity equals one umole proline formed

per minute. Al-Pyrroline-S-carboxylic acid reductase was

assayed according to Strecker (1965) as amended by Costilow

26

27

and Laycock (1971). Protein was measured by the method of
Lowry EE.EL° (1951) using crystalline bovine serum albumin

as the standard.

Polyacrylamide Gel Electrmphoresis. Analytical
disc gel electrophoresis was carried out essentially by the
technique of Davis (1964) using 7% resolving gels and 3.5%
stacking gels. The lower buffer reservoir and gels were
cooled with tap H20 (14-15°) during electrophoresis. The
gels were stained in 1% amido Schwartz in 7% (v/v) acetic
acid. They were destained either by diffusion or in an
electrophoretic destainer with 7% acetic acid.

Preparative polyacrylamide gel electrophoresis was
carried out in either a Buchler "PolyPrep-lOO" unit or a
Canalco "Prep-Disc“ apparatus with the PD-2/150 lower column
insert. The gel system was that of Davis (1964) as modified
by Gilpin and Sadoff (1971) except that (i) the gel contained
10% (w/w) glycerol, (ii) the upper buffer reservoir buffer

3 M DTT and (iii) the elution buffer contained

both 10% (w/w) glycerol and 10-3 M DTT. Each collection

contained 10-

tube contained 0.2 ml of 0.1 M ornithine. The flow rate
was 0.9 ml/min and 1.8 m1 fractions were collected. The
7% resolve gel was 40 mm in length whereas the 3.5% stacking
gel was 10 mm in length. A constant current of 20 or 25 mA

(Buchler system) or 10 mA (Canalco system) was applied.

28

The gel was cooled during electrophoresis by means of a
circulating ice water bath.

Subunit analysis of the homogeneous enzyme was
accomplished with SDS polyacrylamide gel electrophoresis
by a modification of the technique of Weber and Osborn
(1969) using 5% gels, 10 cm in length. Ovalbumin (43,000),
D-amino acid oxidase (37,000), chymotrypsin A (25,000),
yeast alcohol dehydrogenase (37,000) and glutamate dehydro-
genase (53,000) were employed as standard polypeptide markers
(polypeptide chain molecular weights in parentheses).
Approximately 5 pg of each marker and ornithine cyclase were
used in each gel. The gels,were stained with 0.4% coomassie
blue in 10% trichloroacetic acid-33% methanol and were

destained in the same solvent mixture.

Ultracentrifugation Studies. All measurements were

 

carried out in a Spinco model E analytical ultracentrifuge
equipped with an RTIC temperature control unit and either
Rayleigh interference optics and 12 mm double sectored cells
or Schlieren optics and 30 mm double sectored cells. High
speed sedimentation equilibrium analysis was carried out at
4.7° at a rotor speed of 25,978 rpm. The molecular weight
calculations according to the meniscus depletion technique
of thantis (1964) were based on the fringe displacement
after equilibrium was achieved. Sedimentation velocity

analysis was conducted at 5.4° at a rotor speed of

29

47,544 rpm. The sedimentation coefficient was calculated

by the method of Schachman (1957) and corrected to the
viscosity and density of water at 20°. Enzyme samples

used in these studies were dialyzed against 1 liter standard
0.1 M Tris chloride buffer (pH 8.0) prior to use. Dialysate
was used in the reference cells and in making all necessary

dilutions.

Amino Acid Analysis. The pure enzyme was exhaustively

 

dialyzed against three 1 liter changes of distilled-deionized
water prior to hydrolysis of 500 pg samples of protein in
constantly boiling HCl for 24 hours. The analyses of ali-
quots of sample equivalent to 25 or 50 pg of protein were
conducted with an ultrasensitive amino acid analyzer accord-
ing to the procedure outlined by Robertson EE.E£° (1971).
Tryptophan content was measured by the method of Beaven and
Holiday (1952). Approximate corrections for decomposition
of serine and threonine were made as outlined by Moore and
Stein (1963). The half-cystine content was below the limit
of resolution of the analyzer. Therefore, these residues
were estimated by titration of the sulfhydryl groups con-
tained in the dialyzed native enzyme with 0.665 mM p—CMB
according to the procedure of Finlay and Adams (1970) except
that 0.1 M Tris chloride buffer (pH 7.0) was substituted

for 50 mM potassium phosphate buffer. The absorbance

30

difference at 250 nm for the free reagent and its

3 -l

mercaptide was taken as 7.6 x 10 M-lcm (Boyer, 1954).

Absorption Spectrum and Determination of Protein-

Bound NAD+. One mg pure cyclase (in 1 ml) was dialyzed

 

against three 500 ml changes of standard 0.1 M Tris chloride
buffer minus NAD+. Dilution of the NAD+ present in the
original sample was sufficient to bring the concentration

to 8.0 x 10-13 M. The spectrum was then taken from 240 to
700 nm on a Cary model 15 recording spectrophometer using
the last dialysate as blank. To determine bound NAD+, an
aliquot of enzyme was acidified with HCl to approximately

pH 1.0 and incubated at 37° for 15 min, and then neutralized
with NaOH. Standard NAD+ solutions were treated identically.
The NAD+ released by this technique was assayed by the
procedure described by Maxwell and deRobichon-Szulmajster
(1960) in which NAD+ is detected by the disappearance of
pyruvate in a cyclic coupled enzymatic system containing
pyruvate and an excess of lactic dehydrogenase, alcohol
and alcohol dehydrogenase. The remaining pyruvate was

determined spectrophotometrically with fresh lactic

dehydrogenase and NADH.

RESULTS

Enzyme Purification. Ornithine cyclase was
purified in the presence of standard buffer unless other-
wise noted. All operations were carried out in a cold room
(5°) or in an ice bath (0°). Wherever feasible, columns
and buffer reservoirs were maintained under argon. Frac-
tions of eluate from each column containing cyclase activity
were purged of air with a stream of argon and stoppered as
soon as possible after collection. Preparation of the crude
extract, treatment with streptomycin sulfate and ammonium
sulfate fractionation were conducted as described by Costilow
and Laycock (1971). A summary of the overall purification

is presented in Table I.

DEAE-Cellulose Chromatography. Ammonium sulfate treated
extract (19.2 ml) was placed on a large DEAE-cellulose
(DE-52) column (3.5 x 15 cm) which had been equilibrated
with standard 0.15 M Tris chloride buffer (pH 7.5). The
column was washed with the equilibrium buffer until the
A280 was less than 0.3. The column was then washed with
standard 0.25 M Tris chloride buffer (pH 7.5) to elute the

cyclase rapidly. Fractions containing the cyclase activity

31

32

were pooled and concentrated by ultrafiltration through a
Diaflo UM-lO membrane (Amicon Corporation, Cambridge, Massa-
chusetts) using argon as the pressurizing gas. The con-
centrated preparation was then diluted with distilled water
containing the standard stabilizers to a Tris chloride con-
centration of 0.15 M and placed on a smaller DEAE-cellulose
column (1.0 x 25 cm) which had been equilibrated with
standard 0.15 M Tris chloride buffer (pH 7.5). The column
was washed with the equilibration buffer until the A280
reached 0.15. Then a 500-ml linear gradient from 0.15 M
to 0.25 M Tris chloride (pH 7.5) (250 ml of each) was used
to elute the enzyme from the column. Fractions of 7 ml
were collected at a flow rate of approximately 1.7 ml per
min. Fractions were assayed and the cyclase was found to
elute at approximately 0.21 M Tris chloride concentration.

The active fractions were pooled and concentrated as

described above.

Sephadex Gel Filtration. The 3.0 ml concentrated sample
from the second DEAE-cellulose column was applied to the
top of a 2.5 x 100 cm Pharmacia column which was packed
with Sephadex G-200 supported on siliconized glass beads
according to the technique of Sachs and Painter (1971) and
equilibrated with standard 0.1 M Tris chloride buffer

(pH 8.0). Fractions of 3.8 ml were collected at a flow

33

rate of 40 ml per hr. A large protein peak eluted with the
void volume (fractions 16-18). The enzyme was eluted in a
second peak (fractions 25-30). Much contaminating protein
was removed by this step although there was but little
increase in the specific activity of the cyclase. One con-
taminating protein which migrated on disc gel electrophoresis
with an RF almost coincident with the cyclase was totally
removed. The active fractions were pooled and concentrated

to 3 ml by ultrafiltration as before.

Preparative Polyacrylamide Gel Electrgphoresis. Two m1 of
the concentrated cyclase activity from the gel filtration
step were applied to a preparative gel column described in
the Experimental Section along with 0.2 ml glycerol and
0.05 ml 0.1% bromphenol blue. The enzyme was eluted 6 hours
after application to the column.

The elution profile of the cyclase from preparative
polyacrylamide gel electrophoresis is illustrated in
Figure 1. The first peak corresponds to the tracking dye
and the second was primarily NAD+ which was present in the
sample when applied to the column. The fractions corre-
sponding to the shoulder on the trailing edge of the NAD+
peak contained the cyclase free of other protein. The
specific activity of the final preparation was about 36-fold
that of the crude extract and about 3.7% of the total activity

in the crude extract was recovered.

34

Homogeneity. The ornithine cyclase prepared as
described above was homogeneous by several criteria. A
single protein band was observed following analytical disc
gel electrophoresis (Figure 2). A single symmetrical
Schlieren peak was obtained from high speed sedimentation
velocity centrifugation. One subunit species was obtained
from SDS polyacrylamide gel electrophoresis. A plot of the
fringe displacement upon high speed equilibrium sedimenta—
tion ya. the square of the radius was linear (see below).
In addition, no Al-pyrroline-S-carboxylic acid reductase
activity (an enzyme earlier thought to be involved in the
conversion of ornithine to proline) was present. Also,
neither glutamate dehydrogenase nor NADH oxidase activities
(enzymes present in large quantities in extracts of this
Clostridium) were found. Cyclase of this purity is stable
for at least 6 months when stored in sealed ampoules under
argon in liquid nitrogen. Cyclase stored at 4° under argon
generally lost at least 50% of the original activity within
3 days. Freezing the enzyme (-l4°) did not appreciably

increase the stability.

Molecular weight and Subunit Analysis. The
equilibrium sedimentation analysis for determination of the
molecular weight was carried out on two samples, one of

0.9 mg/ml and the other of 0.45 mg/ml. A plot of the

35

fringe displacement (3 + log AY) XE: the square of the
distance from the center of rotation (radiusz) was linear
(Figure 3). The weight-average molecular weight averaged
from the two runs was 81,040.

Analysis of the plates taken during sedimentation
velocity centrifugation indicated a single symmetrical peak.
Using a partial specific volume calculated from amino acid
analysis data of 0.733 cm3/gm, a corrected Svedburg
constant (gzo'w) of 5.68 was calculated for the enzyme,
which corresponds roughly to a molecular weight of 83,000.

The results of two separate SDS polyacrylamide gel
electrophoresis experiments (see Experimental Section for
details) indicated that the cyclase consists of two identi-
cal subunit species. The average molecular weight of the
subunits estimated by this procedure was 41,500. No minor

or additional bands were observed.

Amino Acid Analysis. The analysis of the average

 

of 3 replicate samples of cyclase is shown in Table II.

The calculations for each amino acid are based on the
number of residues per 32 glycine residues. Calculations,
based on residues per 12 tyrosine, 24 alanine or 24 isoleu-
cine residues yielded similar results. A molecular weight
of 80,220 is indicated. The partial specific volume calcu-
1ated from the amino acid data by the procedure of McMeekin

and Marshall (1952) was 0.733 cm3/gm.

36

Absorption Spectrum and Determination of Protein-

Bound NAD+. Previous studies (Costilow and Laycock, 1969;

 

1971) indicated that the conversion of ornithine to proline
was NAD+ dependent. However, the highly purified cyclase
was not completely dependent on added NAD+. Thus, it was
of interest to look for enzyme-bound NAD+. The spectrum
from 340 to 700 nm was not distinctive and decreased slowly
from an absorbance of mo.1 at 340 nm to 0 at 700 nm. Addi-
tion of ornithine to the buffer had no effect on the absor-
bance at 340 nm. However, in the ultraviolet range, the
standard protein peak at 280 nm is little more than a
shoulder on a peak that has a maximum at 267 nm (Figure 4).
The peak is broad and exhibits a minimum at 245 nm.

Three aliquots of a solution containing cyclase
denatured with HCl to release bound NAD+ were assayed.
The results indicate that the purified enzyme does indeed
have tightly boundhqu+(Table III). The average number of
moles of bound NAD+ per mole of enzyme was calculated to

be 1.10.

Effect of Various Compounds on Enzyme Activity.

 

The data presented in Table IV indicate that NAD+, ADP,
and ATP are stimulatory to the pure enzyme. AMP and Pi
had no effect. Several divalent and trivalent cations

+

(Mg+ , Mn++ and Fe+++) were inhibitory at the concentra-

tions employed (10 mM to 10 pM final concentration. The

37

apparent Km for L-ornithine was not changed significantly
by the presence of 5 mM ADP in the reaction mixture. The
values with and without ADP were 6.2 mM and 5.5 mM respec-

tively.

Effect of Sulfhydryl Gromp Inhibitors on Enzyme
Activity. Table V illustrates the effect of p-CMB, iodo-
acetate, n-ethylmaleimide or arsenite on cyclase activity.
p-CMB and n-ethylmaleimide inhibited activity 100% at the
highest levels used. Iodoacetate inhibited to a lesser
extent and arsenite had but little effect on enzyme

activity.

DISCUSSION

These results establish with a high degree of
certainty that the conversion of ornithine to proline is
catalyzed by a single enzyme. The fact that a single sub-
unit species of one-half the molecular weight of the holo-
enzyme was observed in SDS polyacrylamide gels essentially
eliminates the possibility of an enzyme aggregate. Previous
data (Costilow and Laycock, 1971) indicated that this was
probably true, but failed to eliminate the possibility of
a complex.

It is apparent that the purified protein obtained
in this study did not retain maximum specific activity.

At least one preparation of the partially pure enzyme
studied earlier by Costilow and Laycock (1971) had a
specific activity of 4.8 units per mg of protein while

that of the homogeneous preparation obtained in this study
was only 1.8. However, the previous report also notes that
there was a great loss in specific activity during further
purification of their enzyme preparation. The reason for
this loss of activity is unknown. we have attempted to
reactivate the pure enzyme by adding heat inactivated crude

or partially purified extracts, ashed residues of crude

38

39

extracts, or various mineral ions without success. It
appears probable that much of the protein is denatured
during the final phases of purification.

Previous papers (Costilow and Laycock, 1968; 1969;
1971) have indicated that NAD+ is a required cofactor for
this enzyme. The apparent Km for NAD+ obtained with a
partially purified preparation was 6.1 x 10-3 mM. However,
no absolute requirement for NAD+ was demonstrable when more
highly purified preparations were used. We also failed to
find a complete dependence on added NAD+ by the homogeneous
enzyme. However, an analysis of an acid-denatured enzyme
preparation using a dehydrogenase assay demonstrated the
presence of approximately 1 mole of NAD+ per mole of enzyme.
Also, the absorption spectrum of an exhaustively dialyzed
preparation indicated the presence of a purine and/or a
pyrimidine base. The major absorption peak was at 267 nm
with only a shoulder at 280 nm. One would not expect this
much of a shift in the absorption maximum of an average
protein by the presence of 1 mole of NAD+ per mole enzyme.
The A280:A260 for the cyclase was only 0.92, as compared
to an A280:A260 of 1.19 for uridine diphosphogalactose
4'-epimerase which has 1 mole NAD+ per 72,000 gm enzyme
(Wilson and Hogness, 1964). Thus, it appears that our

estimate of bound NAD+ may be low or that there are other

nucleotides bound to the cyclase.

40

We failed to detect any evidence of reduced pyridine
nucleotide by measuring the absorbance of the enzyme at
340 nm in the presence of substrate. Such an absorption
peak has been observed with other enzymes containing bound
NAD+ (Taylor SE §£., 1948; Wilson and Hogness, 1964; Bert-
land and Kalckar, 1968; Zarkowsky and Glaser, 1969). Our
failure to detect a 340 nm absorbance may have resulted
from the low specific activity of the dialyzed preparation
in the absence of added NAD+ (No.6 units/mg). Also, it is
possible that NADH is oxidized as rapidly as it is formed
with this enzyme, since the cyclase reaction is either
irreversible or proceeds very far in the direction of pro-
line (Costilow and Laycock, 1971). It is obvious that
more extensive efforts will be required to conclusively
establish a catalytic role of NAD+ in this reaction.

Strong inhibition of enzyme activity by three sulf-
hydryl-binding reagents suggests the participation of
sulfhydryl groups during catalysis. Titration of the sulf-
hydryl groups present in the enzyme by p-CMB indicates the
presence of 4.1 such groups per mole. However, the low
level of inhibition of activity by the arsenite anion
indicates a lack of vicinal sulfhydryl groups in the active
site.

Studies are currently in progress to elucidate the

mechanism of the ornithine cyclase reaction and to

41

establish with confidence the cofactor role of NAD+ in the

reaction.

ACKNOWLEDGMENT

We wish to acknowledge the assistance of W. C. Deal,
Jr., and W. A. Wood at the Department of Biochemistry,
Michigan State University, for the completion of the ultra-

centrifuge studies and the amino acid analysis, respectively.

42

REFERENCES

Beaven, G. H., and Holiday, E. R. (1952), Advan. Protein

 

Chem. 7, 319.

 

Bertland, A. U., and Kalckar, H. M. (1968), Proc. Natl.

 

Acad. Sci. U.S.A. 61, 629.

 

Boyer, P. D. (1954), J. Amer. Chem. Soc. 76, 4331.

 

Chirpich, T. P., Zappia, V., Costilow, R. N., and Barker,

H. A. (1970), J. Biol. Chem. 245, 1778.

 

Costilow, R. N., and Laycock, L. (1968), J. Bacteriol.

 

96, 1011.

Costilow, R. N., and Laycock, L. (1969), J. Bacteriol.

 

100, 662.

Costilow, R. N., and Laycock, L. (1971), J. Biol. Chem.

 

246, 6655.

Davis, B. J. (1964), Ann. N. Y. Acad. Sci. 121, 404.

 

Finlay, T. H., and Adams, E. (1970), J. Biol. Chem. 245,

 

5248.

Gilpin, R. W., and Sadoff, H. L. (1971), J. Biol. Chem.

 

246, 1475.
Lowry, O. H., Rosenbrough, N. J., Farr, A. L., and Randall,

R. J. (1951), J. Biol. Chem. 193, 265.

43

44

Maxwell, E. S., and De Robichon-Szulmajster, H. (1960),
J. Biol. Chem. 235, 308.

McMeekin, T. L., and Marshall, K. (1952), Science 116,
142.

Mitruka, B. M., and Costilow, R. N. (1967), J. Bacteriol.

 

93, 295.

Moore, 5., and Stein, W. H. (1963), i2 Methods in Enzy-
mology, Vol. VI, Colowick, S. P., and Kaplan, N. 0.,
Eds., Academic Press Inc., New York, p. 819.

Morley, C. G. 0., and Stadtman, T. C. (1970), Biochemistry

 

9, 4890.
Robertson, D. C., Hammerstedt, R. H., and WOod, W. A.

(1971), J. Biol. Chem. 246, 2075.

 

Sacks, D. H., and Painter, E. (1971), Science 175, 781.

Schachman, H. K. (1957), i2 Methods in Enzymology, Vol. IV,
Colowick, S. P., and Kaplan, N. 0., Eds., Academic
Press Inc., New York, p. 32.

Stadtman, T. C. (1954), J. Bacteriol. 67, 314.

Stadtman, T. C., and White, F. H., Jr. (1954), J. Bacteriol.

 

67, 651.

Stickland, L. H. (1934), Biochem. J. 28, 1746.

 

Strecker, H. J. (1965), J. Biol. Chem. 241, 1225.
Taylor, J. F., Velick, S. F., Cori, G. T., Cori, C. F.,
and Slein, M. W. (1948), J. Biol. Chem. 173, 619.

Weber, K., and Osborn, M. (1969), J. Biol. Chem. 244, 4406.

 

45

Wilson, D. B., and Hogness, D. S. (1964), J. Biol. Chem.

 

239, 2469.

WOodS, D. D. (1936), Biochem. J. 30, 1934.

 

thantis, D. A. (1964), Biochemistry 3, 297.

 

Zarkowsky, H., and Glaser, L. (1969), J. Biol. Chem. 244,

 

4750.

46

 

 

 

 

 

 

 

 

 

 

n.m om.a v.~ mm.H o.~ Hem weesmaauom
Imaom m>wumhmmonm h
m.- omv.a n.¢H m.m o.m oomuo xmcmsmom m
¢.mm mH~.H n.mm ma.am n.v HH mdma m
m.m> mmm. v.Hm o.mmH o.om H m<ma v
o.~m Nmo. o.om m.mmm «.ma Auso womlomv
sommlsmzs m
m.vHH mmo. m.wh omma m.mm usefiumouu
:womEoumOHuw mm.~ m
ooa omo. m.mm oama ~.ma womnuxo Opsuo H
mE\muacsv Amuassv
huo>ooom A . . AmEv AHEV
>ua>wu0¢ hua>wuom coaumummoum moum
unwoumm owaaoomm Hmuoa swououm OESHO>
.ommaomo ocflnuwsuo mo coaumowmausm .H wands

47

TABLE II. Amino Acid Analysis of Ornithine Cyclase.a

 

 

 

Amino Raises?
Asp 38.68 39
Thrb 20.88 21
Serb 25.76 26
Glu 42.56 43
Pro 14.56 15
Gly 32.00 32
Ala 23.84 24
Val 21.60 22
Met 18.56 19
Ile 24.24 24
Leu 29.44 29
Tyr 11.56 12
Phe 10.28 10
Lys 25.28 25
His 5.80 6
Arg 9.44 9
Cysc 2
Tryd 3

 

 

 

aThe data were obtained from the average of trip-
licate analyses of 25 or 50 pg cyclase samples which had
been hydrolyzed for 24 hours in constantly boiling HCl in

bApproximate corrections

sealed, evacuated tubes at 110°.
for decomposition of these amino acids were made according
to Moore and Stein (1963). cTitrated as free sulfhydryl

groups with p-CMB. dEstimated by spectral means described

in the Experimental Section (Beaven and Holiday, 1952).

48

 

 

 

TABLE III. Estimation of Protein-bound NAD+.a
., Amer
None 0.0000 -
0.167 nmole NAD+ 0.056 -
0.333 nmole NAD+ 0.112 -
0.500 nmole NAD+ 0.167 -
0.677 nmole NAD+ 0.222 -
Acid treated enzyme: '
0.101 nmoles 0.041 1.21
0.203 nmoles 0.073 1.08
0.305 nmoles 0.104 1.02

 

 

 

aPreincubation mixture contained 100 p1 of 0.02M
potassium phosphate buffer (pH 7.4), 20 pl of 95% ethanol,
10 pl each of 0.025 M sodium pyruvate, lactic and alcohol
dehydrogenases and either standard samples of NAD+ or acid
treated enzyme solutions and sufficient water to bring the
total volume to 0.2 ml. The mixtures were incubated at
25° for 10 min and then plunged into boiling H20 for 2 min

to inactivate. Residual pyruvate was determined spectro-

photometrically by addition of fresh LDH and NADH.

49

TABLE IV. Effect of Various Compounds on Enzyme Activity.a

 

 

 

Ad d i t ion (:1 rficttsi vxi tly0 3 ) PeCrOcnetnrt; 10 f
Experiment 1:
None 18.4 100
2.0 pM NAD+ 21.2 115.2
20.0 pM NAO+ 24.5 133.0
200.0 pM NAO+ 25.3 137.5
Experiment 2:
None 11.3 100
1.0 mM AMP 11.0 97.2
5.0 mM AMP 11.8 105.0
1.0 mM ADP 13.6 120.7
5.0 mM ADP 15.4 137.1
10.0 mM ADP 14.7 130.1
Experiment 3:
None 8.1 100
5.0 mM ADP 13.4 164.6
1.0 mM Pib 8.6 106.9
5.0 mM Pi 9.2 113.9
1.0 mM Pi + 1.0 mM ADP 10.6 130.0
5.0 mM Pi + 5.0 mM ADP 11.5 141.3
1.0 mM ATP 11.7 143.8
5.0 mM ATP 12.2 149.3

 

 

 

aThe standard radiochemical assay was used in each
experiment except that no NAD+ was added to the reaction
mixtures in Experiment 1 unless indicated. In each case the
added compound was pre-incubated with enzyme for 10 min at
39° before adding substrate. In Experiment 1, 29.5 pg

50

TABLE IV.--Continued.

protein (specific activity = 0.857) was used which had been
exhaustively dialyzed to removeruuffprior to assay. In

Experiments 2 and 3, 20 and 15 pg enzyme (specific activity
= 0.565) were used. Reaction mixtures in Experiments 2 and

4

3 contained 1 x 10- M NAD+. bInorganic phosphate was

potassium salt at the pH of the reaction mixture.

51

TABLE V. Effect of Sulfhydryl Group Inhibitors on Enzyme

 

 

 

Activity.a
. . Activity Percent
Add1t1°n (units x 103) Inhibition

None 13.0 0
p-CMB:

0.01 mM 11.7 10

0.1 mM 6.1 53.1

1.0 mM 0.0 100
Iodoacetate:

0.02 mM 13.8 ' 0

0.2 mM 10.5 19.2

2.0 mM 3.9 70.0
n-Ethylmaleimide:

0.02 mM 12.0 7.7

0.2 mM 5.6 56.9

2.0 mM 0.0 100
Arsenite:

0.02 mM 10.8 16.9

0.2 mM 9.6 26.2

2.0 mM 12.7 2.3

 

 

 

aPure cyclase (1 mg in 1 ml) was dialyzed against
2 liters of standard Tris chloride buffer (pH 8.0) minus
DTIPfor 12 hours. The various inhibitors were added to
the standard radiochemical assay minus mercaptoethanol.
Fifteen pg protein with 1.3 x 104-2 units of activity were
used in each assay.

 

52

.mufl>fluom ommaomo smoamcmfluu “Aomm

«v mafimoum coauzao
may .mmHOHHU .uxou on» CH posflauso one mHHmuop Hmucoafluomxo one

.ommaomo Osaguflsuo mo memouonmouuooaw How opflemamuomwaom O>Humnmmmum

.H mmDUHh

53

con: Emma 90 51mm

 

_ manor...

mums—DZ 29.—.041...
on O? on ON 0.

 

 

 

 

 

 

 

 

 

QTQR‘QQ‘Im‘“

BONVSHOSBV

54

FIGURE 2. Gel scan of the purified cyclase in a 7%
analytical polyacrylamide gel taken at a wavelength of
600 nm. For details, see the Experimental Section.
Thirty-six pg protein (specific activity = 1.80) were

applied to the gel.

 

RELATIVE ABSORBANCE

55

 

 

k

'

 

I. I I I
DYE MARKER PIN

 

 

J

l l l l

 

 

 

 

U

 

 

TOP

FIGURE 2

2 3 4
GEL LENGTH (cm)

BOTTOM

 

56

.ommaowo poamausm
mE mv.o pmcflmucoo camp omozu How poms HHOO one .mOHSOOOOHQ poaamuoo Mom
sowuoom Hmucoeflummxm map mom .Ammsflpmuv GOHDMDOH mo Housoo osu Eoum

mocoumflp may mo OHMOOm on» .MN Aw< mod + mv ucoEoomammfip mmcflum mo uon

.m MMDUHh

57

0.00

assumes:
0.00

m 0130.“.

one.

 

 

 

_ _ 0 _

I N._
c..
0..
0..
0.N
N.N
¢.N
0.N
0N
0.0
«.0

 

 

(AV) 50w;

 

 

58

.Amsmmsnse noses «No.0

“mammamwv ouowon om.H umpfi>wuom

oemwoommv HE\mE o.H mo3 sowumuucoosoo aflououm .coapoom amasoEwuomxm

on“ OH UOGHHDOO ma ouspoooum one

.ommaoxo msoocomoaos 00 snow >9

.v HMDme

 

 

59

0V0

AE:VIPOZMJM><3
0N0 000 00m

.v 0130...;

00m

0¢N

 

 

_

 

q

_ s _ _ _ .

_

l

 

 

OQQE‘QQVZR“!

BONVBHOSBV

ARTICLE 2

REMOVAL OF THE ALPHA-AMINO GROUP DURING THE

CONVERSION OF ORNITHINE TO PROLINE*

BY

W. L. Muth+ and R. N. Costilow

Manuscript to be submitted to

J. Biological Chemistry

 

 

*

1

FOOTNOTES

From the Department of Microbiology and Public Health,
Michigan State University, East Lansing, Michigan 48823.

Received . This investigation has been sup-

 

ported by Research Grant No. from the National

 

Institutes of Health. Journal Article No.

 

from the Michigan Agricultural Experiment Station.

Graduate Fellow under Title IV, National Defense Educa-
tion Act. This work was submitted in partial fulfillment

of the requirements for the degree of Doctor of Philoso—

PhY .

Abbreviations used in this paper include: MET, mercapto-
ethanol; 2-PCA, Al—pyrroline-2-carboxylic acid; and TMS,

trimethysilyl.

61

 

Ornithine cyclase catalyzes the transformation of
ornithine to proline in a unique enzymatic reaction (Costi-

low and Laycock, 1971) and is outlined as follows:

 

Cyclase H2? ?H2
CHZNHZCHZCHZCHNHZCOOH + > HZC\\ //CHCOOH + NH3
NAD ADP N
H

The cyclase, which is a single protein, appears to catalyze
a balanced oxidation-reduction reaction. It is stimulated
by catalytic amounts of NAD+ and the pure enzyme contains
approximately 1 mole of bound NAD+ per mole of enzyme (Muth
and Costilow, submitted for publication). Some stimulation
of the cyclase was apparent with ADP or ATP, though no
evidence of allosterism was observed and the apparent Km
for ornithine was unchanged by the addition of ADP. The
enzyme contains 4 labile sulfhydryl groups and is sensitive
to oxygen. No other cofactors have been identified.

The present study was conducted to determine which
amino group is removed from ornithine during the cyclase
reaction. An attempt was also made to incorporate tritium
from NADT into proline during the conversion of ornithine

to proline.

62

EXPERIMENTAL SECTION

Synthesis of [O-lSN]-Ornithine. [O-lSN]-Ornithine was

 

synthesized from a-aminoadipic acid and potassium azide-
15N in a Schmidt degradation reaction by a procedure
adapted from two sources (Adamson, 1939; Wolff, 1946).

Nine ml anhydrous CHCl 3 ml concentrated H 804, and

3' 2

1.5 mmoles (241 mg) a-aminoadipic acid were added to an
acid-cleaned 16 x 150 mm tube containing a small stirring
bar. The reaction was started by slowly adding 1.5 mmoles
(98 mg) KNNlSN (96.4 atom % enriched, Mallinkrodt Nuclear,
St. Louis, Mo.) to the stirred mixture maintained at 43-
45°. The reaction proceeded until gas evolution had

ceased (5 hr). The sulfuric acid containing the newly
synthesized amino acid was diluted in 297 m1 cold distilled
H20 and was then applied to a small (0.9 x 6.0 cm) column
containing Dowex-SO H+—form resin (400 mesh). vThe unreacted
a-aminoadipic acid was eluted from the column with 1N HCl
and the newly synthesized ornithine was then eluted with

3N HCl. The fractions containing ornithine were pooled

and dried under a vacuum at 45° in a rotary evaporator.

The amino acid was dissolved in water and dried several

times to remove excess HCl. The sample was taken up in

63

64

2 ml of distilled H20 and compared with an authentic
ornithine standard by (a) ionophoresis in 0.125N sodium
acetate (pH 5.2), 0.5N acetic acid (pH 2.6), and 0.2M
formate (pH 2.0); (b) descending chromatography on Whatman
3 mm paper developed in butanol: acetic acid: water (120:
30:50); and (c) by ascending chromatography on thin layer
silica gel developed with chloroform: methanol: 15% ammonium
hydroxide (36:46:20). The relative migrations of the
synthesized product and the standard were identical in all
cases. Assay of the product by the reduced ninhydrin method
of Moore and Stein (1954) indicated that the yield was

250 pmoles or 16.6%. A small ninhydrin-positive impurity
was present, but was neither identified nor removed.
Samples of the synthesized product and the standard were
examined in the combined gas chromatograph-mass spectro-

graph after preparing trimethylsilyl derivatives of each.

Conversion of Ornithine to Proline. Two reactions were

 

carried out in this experiment, one with [O-lsNJ-labelled
and the other with unlabelled ornithine. Both reactions
were carried out in 6 x 50 mm tubes inserted in 13 x 100 mm
stoppered tubes which had been flushed with argon. The
reaction mixtures were set up in a manner similar to the
radiochemical assay previously described (Costilow and

Laycock, 1971). Each reaction tube contained: 0.015 ml

65

l

I

each of 0.25M Tris chloride buffer (pH 8.0), 0.1M MET

3

10' M NAD+, and 6.6 x 10'2M CuCl The Cu++ was added to

2.
inhibit the reduction of some of the proline formed to
O-aminovaleric acid by the enzyme preparation used (Costi-
low and Laycock, 1968). Each reaction also received 0.075 ml
(0.2 mg protein) of an ammonium sulfate-treated and dialyzed
extract of Clostridium sporogenes cells (Muth and Costilow,
submitted for publication). The reactions were started by
adding 0.035 ml of either 0.2M [O-lsN]-1abelled or unlabelled
ornithine to individual reaction mixtures. The tubes were
incubated at 37° for 2 hr and the reactions stopped by
placing each tube in an 80° water bath for 5 min. A radio-
chemical assay (Costilow and Laycock, 1971) run concurrently
indicated that about 3.3 pmoles of proline were formed by
the enzyme.

Some of the NH3 liberated by the cyclase reaction
was trapped in glutamate by the use of glutamate dehydro-
genase. The following reagents for this reaction were
added to each of the two tubes: 0.030 ml each of 1.5M Tris
chloride buffer (pH 8.2), 5.0 x lO-ZM a-ketoglutarate,
2.0 x 10-2 ADP, and 4.0 x lO-ZM NADH. The reaction was
started by adding 8.5 units (0.010 ml) of glutamate dehydro-
genase. The tubes were incubated at 30° for 1 hr. Again,

the protein was precipitated by heating each tube to 80°

for 5 min and the tubes were centrifuged at 3,000 RPM for

66

15 min in a clinical centrifuge to remove the denatured
protein. The supernatent solutions were removed and the
pellets washed by resuspending in 0.3 ml distilled H20.
After centrifuging again, the supernatent solutions from
the first and second centrifugations were mixed. The
entire solution from each reaction was applied to a small
(0.9 x 6.0 cm) column containing Dowex-50 H+-form resin
(400 mesh) and washed with 12 m1 H O. The amino acids were

2

eluted with 10 m1 1M NH4OH. The eluate from each column
was dried under reduced pressure at 45° in a rotary
evaporator. The dried amino acids were dissolved in

0.1 ml distilled water.

Preparation of Trimethylsilyl Derivatives. Samples con-

 

taining 100 pg each of [O-lSN]-labelled or unlabelled
ornithine and samples of the resuspended enzymatic products
estimated to contain approximately 100 pg of proline and

35 pg of glutamate were dried at 50° in small (6 x 50 mm)
tubes inside 13 x 100 mm screw-capped tubes. After drying,
the sample tubes were maintained under dry nitrogen or
argon. The dried samples were then solublized in 0.050 ml
of acetonitrile and 0.050 ml bis-(trimethylsilyl)trifluoro-
acetamide containing 1% (v/v) trimethylchlorosilane
(Andrews SE g;., 1971; Gehrke, 1971). Each sample was
heated to 77° for at least 30 min to facilitate the tri—

methylsilylation reaction.

67

Isotope Analysis. Aliquots (0.002-0.004 m1) of the

 

trimethylsilyl derivatives were analyzed in an LKB-9000
combined gas chromatograph-mass spectrograph equipped with
a 3.0 mm x 1.83 m silanized glass column packed with 1%
(w/v) SE-30 on silanized Supelcoport (100-120 mesh,
Supelco, Inc.). The column temperature was maintained at
100° until proline was eluted and then the temperature was
increased at a rate of 7.3°/min to 140° to facilitate the
elution of glutamate. Peaks containing ornithine deriva-
tives eluted at 140° and at 160°. The helium carrier gas
flow rate was 25 cc/min. The ion source temperature of
the mass spectrograph was 290°, and the ionizing voltage
was 70 eV. All determinations of intensities of the various
ions were made from normalized bargraphs (Sweeley 23.21:!
1970). Isotope incorporation was measured by determining
the ratio Ip+1/Ip + Ip+1, where Ip is the intensity of the

ion at m/e = p and I is the intensity of the ion at m/e =

p+l
p+1.

Chemical Reduction of NAD+ in T20. The synthesis of NADT
labelled with tritium in the 4-position of the pyridine ring
was similar to the procedures used by Ohlmeyer (1938) and
Fisher 25 El: (1953). One hundred mg NAD+ (89% pure) were

dissolved in 2 ml of tritiated water (total activity

250 mCi) contained in an acid-cleaned 16 x 150 mm stoppered

68

tube flushed with argon. Fifty mg of sodium hydrosulfite

(dithionite) and 104 mg NaHCO were then added and dissolved

3
as a mixture. The tube was flushed with argon again and
incubated at 30° for 3-5 hr. The reduced NAD+ was crystal-
lized by addition of 25 ml of absolute ethanol at -20°. A
second crop of crystals was obtained by adding 11 ml of
absolute ethanol to the supernatent solution. The crystals
of NADT were dissolved in a small volume of H20 and
recrystallized to remove excess tritium from the compound.
To remove any exchangeable tritium left, the recrystallized
NADT was dissolved in 8 ml of H20 and concentrated in a
Diaflo ultrafilter using a UM-2 membrane. The concentrated
compound was diluted to 10 ml and concentrated again to a
volume of 2.0 m1. A small sample was counted for tritium
content, and an additional sample was assayed for reduced
NAD+ content by measuring the absorbance of a diluted sample
at 340 nm. The yield was 49% or 70 pmoles NADT with

15,000 cpm/pmole. Approximately 30% of the reduced com-

pound carried the tritium label in the A form and 70% in

the B form (Colowick and Kaplan, 1957).

Production of Al-Pyrroline-Z-carboxylic Acid. Al-Pyrroline-

 

2-carboxylate was produced by the oxidation of D-proline by
D-amino acid oxidase in the presence of excess catalase in
a Warburg respirometer (Strecker, 1960; Costilow and Laycock,

1969). The protein was removed by treatment of the reaction

69

with 5% trichloroacetic acid followed by extraction of the
trichloroacetic acid with ether. Ether solublized in the
water was removed by bubbling argon through the solution.
Oxygen consumption data indicated that approximately

300 pmoles of 2-PCA were formed from 500 pmoles of D-proline.

The yield was 60%.

Incorporation of NADT into Proline. Reactions to determine

 

the extent of incorporation of tritium from NADT into proline
by the enzyme in the presence of ornithine alone or with
2-PCA were similar to the standard radiochemical assay for
the cyclase (Costilow and Laycock, 1971). Each reaction
contained: 19.2 mM Tris chloride (pH 8.0), 7.7 mM mercapto-
ethanol, 1.5 mM ADP, and 0.025 ml pure cyclase (specific
activity 0.86) prepared as described by Muth and Costilow
(submitted for publication). Where indicated, 2 pmoles of
ornithine, 1.2 pmoles 2-PCA, 5 nmoles NAD+, and 0.35 pmoles
NADT were added alone or in various combinations. The total
volume of each reaction was adjusted to 0.065 ml with dis-
tilled water. Each reaction was incubated for 2 hr at 37°.
Reactions were stopped and protein precipitated by addition
of 0.025 ml of 0.5M formate. An aliquot (0.040 ml) of each
reaction mixture was spotted on Whatman 3 mm paper and
developed in descending fashion with butanol: acetic acid:

water (120:30:50). The chromatogram was dried and sprayed

70

with 0.01% ninhydrin. The proline spots were cut out and
the radioactivity determined in a liquid scintillation

spectrometer.

RESULTS

Analysis of [0-15

N]-Ornithine. Gas chromatographic analyses
of TMS derivatives prepared from either [O-ISNJ-labelled or
unlabelled ornithine each yielded two peaks. Mass scans
taken as each peak eluted from the column were consistent
with a tri-TMS and a tetra-TMS derivative of ornithine
corresponding to peaks B and D of Figure 1, respectively.
Analysis of the parent ion and four other prominent peaks
obtained from the tetra—TMS—ornithine peak indicated the
labelled compound was 28.0 atom % enriched for 15N. The

method of synthesis insured that all label was directed to

the 6-position.

Incorporation of 15N into the Products of the Ornithine
Cyclase Reaction. TMS derivatives of proline and glutamate
formed from either labelled or unlabelled ornithine were
readily separated from each other and from residual orni-
thine and several other compounds present in the derivatized
sample by gas chromatography (Figure 1). Four ion peaks
from the mass spectrum of proline were examined (Table I).
The parent ion [(TMS)2-proline (M)], was not included since

it was either of very low intensity or absent in every mass

71

72

spectrum taken. The largest molecular ion consistently
present was that at m/e 244 (Figure 2) which corresponds
to M-lS (the loss of one methyl group). Other prominent
ions used were those at m/e 216, 142, and 70. Each of
these contain the nitrogen from the proline ring. The ion
at m/e 216 may be regarded as TMS-N=CH-COO-TMS. The base
peak at m/e 142 (M-117) represents one of the expected

products of a-fission (BergstrOm, 1970):

The final ion considered is that at m/e 70 which may be

regarded simply as the pyrroline ring from proline:

I—I“2
C CH

2 \/

H2
H

It is evident from a comparison of the mass spectra
(Figures 2A and 2B) and the ion ratio intensities (Table I)

of proline produced from unlabelled and from [6-15N]-orni-

15

thine that the N and thus the 5-NH group of ornithine is

3

conserved in proline. The calculated atom-percent 15N in

the proline formed was the same as that determined in the

73

[6-15N]-ornithine used as the substrate for the cyclase

reaction.
Four ion peaks from the mass spectrum of glutamate
were also examined (Table II and Figure 3). A small but
consistent parent ion of m/e 363 was present which sug-
gested a tri-TMS derivative. The readily identifiable
second ion of m/e 348 (M-15) results from the loss of one
methyl group from the parent ion. The third useful ion
was located at m/e 246 (M-117) and corresponded to an
CH

a-fission product: di-TMS-(OOC-CH CHNH). The last

2 2

molecular ion examined was m/e 218 (M-145). The ion was
regarded as a B-fission product containing both the amino
group and the a-carboxyl group (TMS-+NH=CHCOOTMS). Thus,

all four ions contained the NH3 group of glutamate which

was generated from the ammonia liberated during the con-

version of ornithine to proline by the cyclasea Compari-
son of the mass spectra (Figure 3A and 3B) and of the ion
intensities (Table II) of glutamate from reaction mixtures
in which unlabelled and [6-15NJ-ornithine were substrates

for the cyclase reaction show that there was no significant

enrichment of 15N. Therefore, the a—NH group of glutamate

3

must be derived from the a-NH group released from ornithine

3
by the cyclase.

Incorporation of NADT into Proline. There was no apparent

 

incorporation of tritium from NADT into proline during the

74

cyclase reaction (Table III). It also appeared that the
enzyme failed to incorporate tritium from NADT into proline
when 2-PCA was added. Based on the control reaction, the
maximum radioactivity which could have been observed in the
samples examined would have been about 450 cpm above back-
ground if the A form of NADT was used and about 1000 cpm
above background if the B form was used. Incorporation of
very low percentages of either form could have occurred

without detection in this experiment.

 

DISCUSSION

Previous data (Costilow and Laycock, 1969; 1971)
indicated that the conversion of ornithine to proline by
ornithine cyclase might proceed in one of two possible
directions (Figure 4). The direction taken would obviously
depend on which amino group was involved in the deamina-
tion. Loss of the G-amino group would lead to the forma-
tion of glutamic-Y-semialdehyde which would undergo ring
closure to form Al-pyrroline-5-carboxylic acid (5-PCA).
Deamination in the a-position would form 2-oxo-amino-

15

pentanoic acid and 2-PCA. The conservation of N from

15N]-ornithine in proline during the conversion of

[6—
ornithine to proline by the cyclase enzyme establishes with
certainty that it is the a-amino group of ornithine which
is removed. Thus, it is highly probable that the latter
compounds are intermediates in the reaction.

The failure of our efforts to incorporate tritium
from NADT into the proline formed during the reaction was
not unexpected. Some of the sugar epimerases which contain
tightly bound NAD+ and probably operate by an oxidation and

reduction mechanism also fail to exchange tritium from

reduced NAD+ or from solvent during substrate catalysis

75

76

(Maxwell, 1957; Glaser, 1972). Two explanations of these
results are evident. First, the enzyme-NADH—intermediate
complex may be of such short duration or of such an undis-
sociable nature that exchange simply cannot take place.
Such is the case with UDP-D-glucose 4'-epimerase (Bertland
and Kalckar, 1968). In the case of the cyclase, this
possibility is reinforced by the fact that numerous
attempts to trap an intermediate of a pyrroline nature with
o-amino-benzaldehyde were unsuccessful (Costilow and Lay-
cock, 1968). The second explanation involves isotope
selection. Using UDP-D-glucose 4'-epimerase with UDP-D-
glucose-4-T as substrate Beville 31; 31.. (1965) noted a
large negative isotope effect. Thus, it is possible that
NADT fails to exchange with enzyme bound NADH simply because
of total isotope selection.

The possible role of NAD+ in the cyclase reaction
(see Figure 4) is still unresolved. Data with partially
purified enzymes (Costilow and Laycock, 1969; 1971) show
that the conversion of ornithine to proline is almost com-
pletely dependent on added NAD+. However, enzyme prepara-
tions of high purity (Costilow and Laycock, 1971; Muth and
Costilow, unpublished data) were only stimulated by added
NAD+. There is evidence that this cofactor is tightly
bound to the pure enzyme.

The fact that no tritium from the medium was incor-

porated into proline when the reaction was carried out in

77

tritiated water (Costilow and Laycock, 1971) is consistent
with an enzyme-bound oxidation-reduction mechanism. All
known dehydrogenases add or remove hydrogens from non-
exchangeable positions, while the other involved hydrogen
is in a freely exchangeable position on the substrate
molecule (Vennesland and Westheimer, 1954). The nucleotide-
sugar epimerase also has a postulated internal oxidation-
reduction mechanism and fails to exchange tritium from
T20 into the product (Kalckar and Maxwell, 1956).

Further elucidation of the mechanism of this unique

enzyme may be forthcoming from experiments currently being

planned.

REFERENCES

Adamson, D. W. (1939), J. Chem. Soc., 1954.

 

Andrews, T. J., Lorimer, G. H., and Tolbert, N. E. (1971),

Biochemistry 10, 4777.

 

BergstrOm, K., Gfirtler, J., and Blomstrand, R. (1970),

Anal. Biochem. 34, 74.

 

Bertland, A. U., and Kalckar, H. M. (1968), Proc. Nat'l.

 

Acad. Sci., U.S.A. 61, 629.

 

Beville, R. D., Hill, E. A., Smith, F., and Kirkwood, S.

(1965), Can. J. Chem. 43, 1577.

 

Colowick, S. P., and Kaplan, N. O. (1957), im Methods in
Enzymology, Vol. IV, Colowick, S. P., and Kaplan,
N. 0., Eds., Academic Press Inc., New York, p. 840.

Costilow, R. N., and Laycock, L. (1968), J. Bacteriol.

 

96, 1011.

Costilow, R. N., and Laycock, L. (1969), J. Bacteriol.

 

100, 662.

Costilow, R. N., and Laycock, L. (1971), J. Biol. Chem.

 

246, 6655.
Fisher, H. F., Conn, E. E., Vennesland, B., and westheimer,

F. H. (1953), J. Biol. Chem. 202, 687.

 

Gehrke, C. W., and Leimer, K. (1971), J. Chromatogr. 57, 219.

 

78

79

Glaser, L. (1972), im_The Enzymes, Vol. VI, 3rd Ed.,
Boyer, P. D., Ed., Academic Press Inc., New York,
p. 355.

Kalckar, H. M., and Maxwell, E. S. (1956), Biochem. Biophys.

 

Acta. 22, 588.

Maxwell, E. S. (1957), J. Biol. Chem. 229, 139.

 

Moore, 8., and Stein, W. H. (1954), J. Biol. Chem. 211,

 

407.

Ohlmeyer, P. (1938), Biochem. Z. 297, 66.

 

Strecker, H. J. (1960), J. Biol. Chem. 235, 2045.

 

Sweeley, C. C., Ray, B. D., Wood, W. 1., Holland, J. F.,

and Krichevsky, M. I. (1970), Anal. Chem. 42,

 

1505.

Vennesland, B., and Westheimer, F. H. (1954), im_The
Mechamism of Enzyme Action, McElroy, W. D., and
Glass, B., Eds., Johns Hopkins Press, Baltimore,
p. 357.

Wolff, H. (1946), im_0rganic Reactions, Vol. 3, Adams, R.,
Bachman, W. E., Feiser, L. F., Johnson, J. R., and
Snyder, H. R., Eds., John Wiley and Sons, Inc.,

New York, p. 307.

ACKNOWLEDGMENTS

We are greatly indebted to C. C. Sweeley, Depart-
ment of Biochemistry, Michigan State University, for the
useiof the combined gas chromatograph-mass spectrograph.
Our: thanks also to Jack Harten and Norman Young for

excellent technical assistance.

80

81

.mump omen» Eoum pocweuoump on OHDOO
osam> mamsflm O hasom .mcoflumcwfiuouoc mounu mo sowucw>oc cumpcmum can on 000050
Houuo OER

.omoHO>O poemwnsm madmauumm an mcwnuwsHOIHZ 1mg Eoum couscoum Oswaoumo

 

 

 

 

 

n ma
.omoao>o pennansm adamfluumm an ocwzuwcuo poaaonoacs Eouu Umospoum OGHHOHmQ .sowuoom
Housoefiuomxm mom .ansom cOHuoOHMRHOQ was cowuomon some 00 HGORUAOGOO Homo
m.~ H m.w~ u .o>m

H.m~ m.s a «.ma s.o n H.ns ma~\ssm
m.mm m.~ H m.mv mm.o a o.mH bam\mam
m.a~ 6N.os ma.as msaxaaa
m.Hm H.m a m.mv pm.m H o.ma Hh\oh

z I z H+a a H+a H+m a H+d
ea ma HA H + H0\ SuzmH "A H + Hex HLQZsH m\s

As sound udmusoo zmH lav mmausmsoudH mo mosses

 

 

 

 

m.osfinoum Open 2 mo coflumuomuoosH .H mamma

0H

82

.mCOHUMGflEHOUOU Own—”flu. HO COfluMflkrmU UHMUflMuh 0S“. mH

oouosv Houuo one
m

.omoHo>o ooHMAHsm aaaofiuuom an ocwsuwcHOIHz IQH Eoum oomooaou

0 0H
mz onu nufl3 pououosom ouoEOHSHoo .omoaomo cognauam haaowuuom an ocflnuflcno
mmz onu suw3 pouoposom ouofiousao

ooHHonoHcs Eonm oomooaou .sowuoom Housofiauomxm

 

 

 

 

 

n
oom .ofionom GORHMOAMHHSQ can sORuoooH sumo Cw pom: md0fiufi©soo Homo.
a.m n H.H+ u m><
s.HI s.m s n.sa m.a a H.o~ ssm\msm
n.o+ H.H n n.a~ m.o n.o.s~ asm\mam
H.Hn o.n a o.o~ ~.s a ~.HN nsmxssm
H.s+ n.~ a o.s~ em.m n m.aa mamxaam
z u z H+a a H+a H+a a H+d
as ma Hi H + H0\ HLOZmH H1 H + H0\ dazeH m\s
Am Eouov usousoo 20H Amy mowuwmsousH mo mowuom

 

 

 

 

.
A

o.ouofiousH0 OUGH Z No soauohomuoocH .HH wands

0H

83

TABLE III. Incorporation of Tritium from NADT into
Proline with Ornithine Alone or with
Al-Pyrroline-Z-carboxylic Acid (2-PCA) as

Substrates.a

 

 

Addition (pmoles)

 

+ Proline (c.p.m.)
Ornithine Z-PCA NAD Cyclase

 

1. 2.0 - - - 320
2. - 1.2 - + 316
3. - - - + 327
4. - 1.2 0.005 + 301
5. 1.0 1.2 - + 293

 

 

 

aEach reaction contained 0.005 ml each of 0.25 M
Tris chloride (pH 8.0), 0.1 M mercaptoethanol, 0.02 M ADP,
and 0.35 pmoles NADT (specific activity = 15,000 c.p.m./
pmole). Cyclase (29.5 pg, specific activity = 0.86) added
to each reaction contained 0.5 pmoles ornithine and
2.5 nmoles NAD+. In a control reaction, 0.24 pmoles of
proline were formed when 2.0 pmoles of l4C-ornithine and
0.35 pmoles NADH were added with the cyclase. For other
aspects of procedure, consult the Experimental Section.

84

FIGURE 1. Gas chromatography of the trimethylsilylated
amino acids separated from a reaction mixture
containing unlabelled ornithine, proline,
glutamate, and ornithine cyclase. The tracing
is the total ion current detector response
and the dotted line is the column temperature.
For other components of the reaction mixture
and isolation and preparation of samples, con-
sult Experimental Section. Mass spectra
indicated that peak A was (TMS)2—proline,
peak B was (TMS)3-ornithine, peak C was (TMS)3-

glutamate, and peak D was (TMS)4-ornithine.

85

O
ale

0
4

0
3

wmahdmmaiwh

o
w.

 

 

- I60

 

L

-

_

q

—

0
_

D

_

mmZOdmmm Mahdi—mm

0
0
H

- -------qbd

 

20

IO

 

  

MINUTES

FIGURE I

86

.lsv mansnasnOIHZ -a_ soap

ma
ocflaoum 00 away nuwz Amy ocflnuflcuo poaaonoass Eoum MHHMOHDNHMDMO

oo>fluop ocwHonmlmAszv mo Eduuoomm mmoE onu mo unom mo comwuomfiou

.N mMDme

87

QE

010m 0mm 00m omfi Om:

 

 

N meoE

om

I : _ ION

 

[Os
.8
-om

 

 

 

 

0X

 

OOH

ION
10¢
tom
tom
OOH

 

 

(O/o) AlISNEIlNI BAIIVTEIEI

88

.mssssssno Inc .12 -sL no Ass edssmnnass nmsssm

ma
Eoum pououosom oHcoEEo oau can ououousamouoxlo Eonm MHHEOHumHouoo

po>wnop ouoEousHm mAmzev mo ouuoomm nose on» 00 unom mo COmHHoQEOU

.m mMDUHm

89

omm

L s

_

OJm

psbp

0mm

r _

98

com omm omm oqm

—-_»_EL.hp_

n 0130....

0mm

 

 

NJ

 

<AI<<

id

 

 

13 :_

 

_

.: ..:_..

 

TON
roq
r00
[om

 

OOH

 

 

 

 

 

10m
roq
low
100
OOH

 

 

(%) AllSNEIlNI BAllV—IEIH

90

CH CH CH CHCOOH

 

 

 

 

 

 

I: 2 1.
2 2
{K’7‘NAD+ NAD+-\E}
NADH
NH3+ HCPCHZCHZCHOOH CHZCHZCHZC-COOH + NH3
I '32 I. II
H20) K31120 HZO/ KAHZO
V 0
H2? CH2 HZT-——-TH2
HC HC-COOH H c C-COOH

 

\N/ 2\./
\R\\\{E:NAD+;>

2| TH

H2C HC-COOH

FIGURE 4. Proposed mechanism of action of ornithine

cyclase.