L-SORBOSE 1-PHOSPH’ATE REBUCTASE
FROM. AEROBACTER AEROGENES:
ETS PUREFECATION, CHARACTERIZMTGN, 7
AND ROLE EN THE METABOUSM OF
L-SORBGSE AND D-FRUCTDSE

Dissertation for the Degree of Ph. D.
MECHIGAN STATE UNIVERSITY
RONALD ALLEN SIMKINS
1975

 

This is to certify that the

thesis entitled

L-Sorbose l-Phosphate Reductase from Aero-
bacter aerogenes: Its Purification, Char-
acterization, and Role in the Metabolism
of L-Sorbose and D-Fructose.

presented by

 

Ronald Allen Simkins

has been accepted towards fulfillment
of the requirements for

 

 

Ph 0 Do degree in B10Chem18 try

ELEM

humupnmuxn

 

Date '7‘ Mag- ’91-;

0-7639

 

  

f mwue‘” ’- ~
*flnrlgflk
BUUK BINDERY INC.

LIBRARY BINDERS
D

    

 

:4?" -

 

ABSTRACT
L-SORBOSE l-PHOSPHATE REDUCTASE FROM AEROBACTER AEROGENES:

ITS PURIFICATION, CHARACTERIZATION, AND ROLE IN THE
METABOLISM OF L-SORBOSE AND D-FRUCTOSE

BY

Ronald Allen Simkins

In Aerobacter aerogenes L-sorbose is metabolized via the
pathway: L-sorbose +-L-sorbose l-phosphate *‘D-glucitol 6-phosphate
+ D-fructose 6-phosphate. The steps of the pathway are catalyzed
by three enzymes: a phosphoenolpyruvate-dependent L-sorbose l-
phosphotransferase system; L-sorbose-l-P reductase, and D-glucitol-
6-P dehydrogenase, respectively. A key reaction in this pathway
is the reduction of L-sorbose-l-P (L-xylo-Z-ketulose) with NADH or
NADPH to D-glucitol-é—P by the L-sorbose-l-P reductase (L-sorbose-
1-P:NAD(P) ketoreductase), which is inducible by growth on L-sorbose.
The work in this dissertation covers three facets of L—sorbose-l-P
reductase: (i) the isolation and identification of the product of
the phosphoenolpyruvate-dependent phosphorylation of L-sorbose to
demonstrate that L-sorbose-l—P is available as a substrate for the
reductase in the cell; (ii) the purification and characterization
of L-sorbose-l-P reductase, the first enzyme reported to catalyze

the reduction of a ketohexose l-phosphate; and (iii) the role of

Ronald Allen Simkins
this enzyme in the evolution of a novel pathway for the metabolism
of D-fructose in A. aerogenes.

The product of the phosphoenolpyruvate-dependent phosphoryla-
tion of L-sorbose was isolated from a reaction mixture containing
an extract of A. aerogenes supplemented with phosphoenolpyruvate
and L-sorbose. The phosphorylated product (which gave a positive
ketohexose test) was eluted from a Dowex-l bicarbonate column at a
concentration of KHCO3 which elutes sugar monophosphates. It was
identified as a ketohexose monophosphate on the basis of: (i) a
ketohexose to phosphate ratio of 1:1: (ii) a rate of hydrolysis in
IN HCl that was identical with that of authentic L-sorbose-l-P but
different from L-sorbose-G-P; and (iii) a mass spectrogram of the
trimethylsilyl 0-methyloxime derivative of the product which gave
a fragmentation pattern and molecular weight predicted for a 2-
ketohexose l—phosphate. The ketohexose moiety was identified as
L-sorbose by chemical, enzymatic, and chromatographic techniques.

L-Sorbose-l-P reductase was purified 41-fold by DEAR-cellulose
chromatography and by affinity chromatography on NADP-sepharose using
elution with various combinations of cofactor and D-fructose-l-P.
The final preparation contained no more than 1% contamination by
other proteins as determined by acrylamide gel electrophoresis.

The L-sorbose-l-P reductase was shown to have a molecular weight of
90,000 :_2,000 by gel filtration and ultracentrifugation and is
composed of two identical subunits of 45,000 :_l,000 molecular
weight as determined by acrylamide gel electrophoresis with and
without sodium dodecyl sulfate. The enzyme is an acidic protein

based on the following data: (i) amino acid analysis revealed

Ronald Allen Simkins
relatively low amounts of basic amino acids; (ii) the isoelectric
point is 5.0 by isoelectric focusing; (iii) the enzyme migrated
toward the anode at pH 6.2; and (iv) the enzyme bound to DEAE-
cellulose at pH 6.2.

Treatment of purified L-sorbose-l-P reductase with EDTA
abolished catalytic activity. The activity of EDTA-treated enzyme
could be restored by adding various divalent metal cations, but
only Mn+2 could reactivate the enzyme to levels shown before EDTA-
treatment. The enzyme was not capable of reducing L-sorbose, D-
fructose, D-fructose-6-P, or D—fructose-l,6-P2 an: concentrations
up to 10 mM. The enzyme could catalyze the reduction of either
L-sorbose-l-P or D-fructose-l-P with a Km of 0.65 mM and 0.38 mM,
respectively. The maximal velocity was three times as great for
L-sorbose-l-P as for D-fructose-l-P. The enzyme could use either
NADH or NADPH as a cofactor with a Km of about 20 uM for each.

The product of the reduction of D-fructose-l-P was identified
as D-mannitol-l-P by chemical, enzymatic, and chromatographic
techniques. The equilibrium constant of the reaction for the
reduction of L-sorbose-l-P and D-fructose-l-P at pH 6.15 was
determined to be 2.6 x 107 and 4.5 x 107 liters per mole, respec-
tively. The pH optimum of the reaction was 6.2 in three buffers
tested; the greatest activity was obtained in the presence of
2(N-morpholino)sulfonic acid buffer. The stoichiometry of the
reaction revealed that 1 mole of ketohexose-l-P was reduced by 1
mole of NAD(P)H for every mole of sugar alcohol phosphate and

NAD(P)+ produced.

Ronald Allen Simkins
The normal pathways of exogenous and endogenous D—fructose
catabolism (D-fructose + fructose-l-P + fructose-1,6-P2 and fructose

+ fructose-6-P + fructose-1,6-P , respectively) were blocked in A.

2
aerogenes by sequentially mutating the genes for D-fructose-l-P
kinase and fructokinase. This fructose-negative strain would grow
linearly on D-fructose if preinduced by growth on L-sorbose plus
D-mannitol. By selecting from this strain a mutant (SFl) which
was constitutive for L-sorbose-l-P reductase, the ability to
utilize D-fructose as a sole source of carbon and energy was con-
comitantly gained. Mutants of SP1 missing either L-sorbose-l-P
reductase or D-mannitol-l-P dehydrogenase were unable to grow on
D-fructose, whereas a mutant with increased levels of D-mannitol-
l-P dehydrogenase showed an increased growth rate on D-fructose.
When incubated with D-fructose, the L-sorbose-l-P reductase-
negative mutant accumulated D-fructose-l-P and no detectable levels
of D-mannitol-l-P, whereas the D-mannitol-l-P dehydrogenase-negative
mutant accumulated both D-fructose-l-P and D-mannitol-l-P at
greater levels than SFl. The mutant with increased levels of
D-mannitol-l-P dehydrogenase showed a proportional decrease in the
levels of D-mannitol-l-P accumulated. Thus the newly evolved path-
way uses the phosphoenolpyruvate:D-fructose l-phosphotransferase
system of normal exogenous D-fructose metabolism, the L-sorbose-l-P
reductase of L-sorbose metabolism, and the D-mannitol-l-P dehy-
drogenase of D-mannitol metabolism to convert D-fructose + D-
fructose-l-P + D-mannitol-l-P + D-fructose-6-P. This is a novel
pathway of D-fructose metabolism which has been previously unde-

tected in any organism.

L-SORBOSE l-PHOSPHATE REDUCTASE FROM AEROBACTER AEROGENES:
ITS PURIFICATION, CHARACTERIZATION, AND ROLE IN THE

METABOLISM OF L-SORBOSE AND D-FRUCTOSE

BY

Ronald Allen Simkins

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1975

ACKNOWLEDGEMENTS

I would like to thank Richard L. Anderson for his advice,
criticism, and patience: W. C. Deal, Ursula Koch, John Trujillo,
Andrew Mort, R. A. Laine, C. C. Sweeley, w. W. Wells, and Dave
Bishop for their valuable technical advice and assistance. I
would like to especially thank John Markwell and Don Bissett for
their suggestions, assistance, and wit which did much to maintain
a sense of proportion.

This work was supported by the National Institutes of Health

and the Michigan State University Employees Credit union.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS . . . . . . . . . . . . . .
LIST OF TABLES C O I O O O O O O C O O O O 0
LIST OF FIGURES O O C O C O C O C O O O O O 0
LIST OF SYMBOLS AND ABBREVIATIONS. . . . . .
IMRODUCTION O I O O O O O O 0 C O O O O O 0
LITERATURE REVIEW OF L-SORBOSE METABOLISM. .

PART I - ISOLATION AND IDENTIFICATION OF THE

PRODUCT OF

THE PHOSPHOENOLPYRUVATE:L-SORBOSE l-PHOSPHO-
TRANSFERASE SYSTEM OF AEROBACTER AEROGENES. . . . . .

IntrOduCtj-on O O O O O O O O O O O O O O O O O O O O O 0
Materials and Methods . . . . . . . . . . . . . . . . .
Chemicals 0 O O I O O O O O O O I O O O O O O O 0
Growth of Cells an Preparation of Cell Extracts
Analytical Methods . . . . . . . . . . . . . . .

Results 0 O O O I O O O O O O O O I O O 0 O O O O O O 0

Preparation of Product of the PEP:L-

sorbose PTS. . . . . . . . .
Isolation of Product . . . . .
Identification of Product. . .

Discussion. . . . . . . . . . . . . .

PART II - PURIFICATION AND PROPERTIES OF L-SORBOSE

l-PHOSPIi-ATE REDUCTASE o o o o o O 0
Introduction. . . . . . . . . . . . .
Materials and Methods . . . . . . . .

Chemi ca 1 S O O O O O O O I I O 0
Enzyme Assays. . . . . . . . .
L-Sorbose-l-P reductase

Other enzyme assays . .
Protein determinations.

iii

Page
ii
vi

vii
ix

10
10
10
10
11
12
12
12
13

15

18
19
l9

19
20
20
20
21

Page

Determination of Reaction Products . . . . . . . 21
Enzymatic determination . . . . . . . . . 21
Preparation of the hexaacetate derivatives

of the dephosphorylated product . . . . 22
Gas-liquid chromatography . . . . . . . . 22
Preparation of NADP-Substituted Sepharose
4B for Affinity Chromatography . . . . . . . . 23
Activation of Sepharose 4B and
coupling to adipic dihydrazide. . . . . 23
Binding of activated NADP to Sepharose-
adipic hydrazide. . . . . . . . . . . . 23

Polyacrylamide Gel Electrophoresis . . . . . . . 24
General procedures. . . . . . . . . . . . 24
Preparation of native gels for determina-

tion of homogeneity . . . . . . . . . . 26
Preparation of native gels for determina-

tion of molecular weight. . . . . . . . 27
Preparation of SDS gels for determina-

tion of subunit molecular weight. . . . 27

Other Methods for Determining Molecular Weight . 29
Sucrose density gradient centrifugation . 29
Sedimentation equilibrium ultracentri—

fugation. . . . . . . . . . . . . . . . 29
Amino acid analysis of L-sorbose-l-P

reductase . . . . . . . . . . . . . . . 30
Determination of pH . . . . . . . . . . . 30

Results 0 I O O O O O O O O O O O O O O O O O O O O O 0 3O

Purification . . . . . . . . . . . . . . . . . . 30
Preparation of crude extracts . . . . . . 31
Chromatography on DEAR-cellulose. . . . . 32
Affinity chromatography on NADP-

substituted Sepharose . . . . . . . . . 36

Properties . . . . . . . . . . . . . . . . . . . 48
Determination of molecular weight . . . . 48
Sedimentation equilibrium ultracentri-

fugation. . . . . . . . . . . . . . . . 56
Determination of Svedberg coefficient

from sucrose density gradient

centrifugation. . . . . . . . . . . . . 58
Summary of molecular weight determi-

nations . . . . . . . . . . . . . . . . 58
Catalytic properties. . . . . . . . . . . 62
Physical and chemical properties. . . . . 73

Characterization of the Reaction . . . . . . . . 76

Identification of the product of th

reduction of D-fructose-l-P . . . . . . 76
Equilibrium and stoichiometry of the

reaction. . . . . . . . . . . . . . . . 77

Discussion and Summary. . . . . . . . . . . . . . . . . 80

iv

Page

PART III - EVOLUTION OF A NOVEL PATHWAY FOR THE METABOLISM
OF D-FRUCTOSE BY AEROBACTER AEROGENES . . . . . . . 85

IntIOduction O O O O O O O O O O O O O O O O O O O O O O 86
Materials and Methods . . . . . . . . . . . . . . . . . 89

Isolation of Mutants-~Genera1 Procedures . . . . 89

Analytical Procedures. . . . . . . . . . . . . . 91

Enzyme assays . . . . . . . . . . . . . . 91
Determination of intracellular metabo-

lite levels in mutant strains . . . . . 92

IMViC tests . . . . . . . . . . . . . . . 93

Results . . . . . . . . . . . . . . . . . . . . . . . . 94

Linear Growth of Strain 14 on D-Fructose
after Induction with D-Mannitol and
L-Sorbose. . . . . . . . . . . . . . . . . . . 94
Selection of Mutants of L-Sorbose-l-P
Reductase and D-Mannitol-l-P Dehydrogenase . . 96
Mutants constitutive for L-sorbose-
1-P reductase . . . . . . . . . . . . . 96
Mutants negative for L-sorbose-l-P
reductase . . . . . . . . . . . . . . . 100
Mutants negative for D-mannitol-l-P
dehydrogenase . . . . . . . . . . . . . 101
Revertants of SFl-SS an SFl-MAZ. . . . . 102
Mutants with elevated levels of D-
mannitol-l-P dehydrogenase. . . . . . . 103
Diagnostic Tests to Demonstrate the Common
Ancestry of the Mutants. . . . . . . . . . . . 103
IMViC tests . . . . . . . . . . . . . . . 103
Colony characteristics on eosin-
methylene blue agar . . . . . . . . . . 105
Lysis by a bacteriophage specific for
A. aerogenes. . . . . . . . . . . . . . 105
Growth Response on Carbohydrates . . . . . . . . 107
Growth response on carbohydrates other
than D-fructose . . . . . . . . . . . . 107
Growth on D-fructose. . . . . . . . . . . 110
Enzyme Levels in Mutants of A. aerogenes . . . . 113
Intracellular Metabolite Levels in A.
aerogenes Mutants. . . . . . . . . . . . . . . 115

Discussion and Summary. . . . . . . . . . . . . . . . . 118

REE'ERENCES O O O O O O O O O O O O O O O O O O O O O O O O O O 1 2 2

LIST OF TABLES

Table Page
1 Summary of purification of L-sorbose-l-P reductase. . . 47
2 Summary of molecular weight determination of L-
sorbose-l-P reductase . . . . . . . . . . . . . . . . . 61
3 The effects of various divalent metal cations on the

activity of EDTA-treated L-sorbose-l-P reductase. . . . 64

4 Effects of varying concentrations of divalent metal
cations on the activity of L-sorbose-l-P reductase. . . 66

5 Amino acid analysis of L-sorbose-l-P reductase. . . . . 74

6 Gas-liquid chromatography of the hexaacetate deriva-
tive of the dephosphorylated reduction product of
D-fmCtOSG-I'P. o o o o o o o o o o o o o o o o o o o o 78

7 Genotype and phenotype of mutants used in this study. . 90

8 Results of IMViC tests on wild-type and mutant strains
of A. aerogenes and on an E. coli control . . . . . . . 104

9 Colony characteristics of wild-type and mutant strains
of A. aerogenes and an E. coli control on eosin-
methylene blue agar plates. . . . . . . . . . . . . . . 106

10 Effect of the presence (+) of D-fructose on the
growth rate of mutants of A. aerogenes on nutrient
broth 0 O O O C I O O O O O O O O O O O O C O I O O O O 112

11 Specific activities of L—sorbose-l-P reductase and
D-mannitol-l-P dehydrogenase in crude extracts of
mutants of A. aerogenes grown on various carbohydrates
in mineral broth. . . . . . . . . . . . . . . . . . . . 114

12 Intracellular metabolite levels in wild-type and
mutant strains of A. aerogenes induced for 5 hr on
0.5% D-fructose as compared to enzymes present and
growth rate on 0.3% D-fructose. . . . . . . . . . . . . 117

vi

Figure

10

11

12

13

14

LIST OF FIGURES

The pathway of L-sorbose metabolism in A. aerogenes

Two proposed pathways by which L-sorbose may be
modified in certain species of bacteria . . . . . .

Mass spectrum of the trimethylsilyl O-methyloxime

derivative of the product of the phosphoenolpyruvate-

dependent phosphorylation of L-sorbose. . . . . . .
Tracings of the gas-liquid chromatograms of the
trimethylsilyl derivatives of: A_: D-fructose—l-P;
§_: D-fructose-6-P; g_: L-sorbose-l-P; and 2.:
Product of the PEP:L-sorbose 1-phosphotransferase
system. . . . . . . . . . . . . . . . . . . . . . .

Intermediates and product of the NADP-substitution
of Sepharose 4B . . . . . . . . . . . . . . . . . .

Elution profile of first DEAE-cellulose column. . .
Elution profile of second DEAE-cellulose column . .
Affinity chromatography of L-sorbose-l-P reductase.

Affinity chromatography of L-sorbose-l-P reductase;
elut ion With NADH O O I O O O O O O O O O O O O O O

Affinity chromatography of L-sorbose-l-P reductase:
elution with NADH and D-fructose-l-P. . . . . . . .

Polyacrylamide gels of L-sorbose-l—P reductase at
various stages of purification. . . . . . . . . . .

Second affinity chromatography step . . . . . . . .

Determination of the molecular weight of L-sorbose-
1'? reductase by acrylamide gel electrophoresis . .

Determination of the subunit molecular weight of

L-sorbose-l-P reductase (SR) by SDS acrylamide gel
electrophoreSis O O O O O O O I O O O O O I O O O 0

vii

Page

14

16

25
33
35

37

4O

43

44

46

SO

53

Figure Page
15 Elution profile of Sephadex G-100 column. . . . . . . . 54

16 Determination of molecular weight of L-sorbose-l-P
reductase (0) using data obtained from Figure 15. . . . 55

17 Sedimentation equilibrium ultracentrifugation of
L-sorbose-l-P reductase to determine molecular weight . 57

18 Sucrose density gradient determination of Svedberg
coefficient of L-sorbose-l-P reductase. . . . . . . . . 60

19 Relative activity of L-sorbose-l-P reductase as a
function of pH and buffer . . . . . . . . . . . . . . . 63

20 Relationship between L-sorbose-l-P reductase concen-
tration and the velocity of the reaction. . . . . . . . 68

21 Double reciprocal plot showing the relationship
between D-fructose-l-P Or L-sorbose-l-P concentra:
tion and the reaction velocity using two concentra-
tions of enzyme . . . . . . . . . . . . . . . . . . . . 70

22 Double reciprocal plot showing the relationship
between NADH or NADPH concentration and reaction
veIOCity O O O O O O O O O O O I I O O O O O O O O O O O 72

23 Isoelectric focusing on an acrylamide gel from pH
3.0 to 6.0 of a partially purified preparation of
L‘SOIbOSG'I-P reductase o o o o o o o o o o o o o o o o 75

24 Growth of strain 14 on D-glucose (o) and D-fructose
(0) after preinduction by growth on D-mannitol and
L-sorbose . C I C O C O C O O C O C C O O C O O O O O I 95

25 Ability of mutants of A. aerogenes to utilize D-
fructose as a function of enrichment for constitu-
tivity of growth on L-sorbose . . . . . . . . . . . . . 99

26 Lysis of wild-type and mutant strains of A.
aerogenes by a bacteriophage. . . . . . . . . . . . . . 109

27 Growth curves of wild-type and mutant strains of
A. aerogenes on 0.3% D-fructose in mineral broth
at 30°. 0 O O C O O O I O O O O O I I O O O O O O O O O 111

28 Double reciprocal plot showing the relationship
between D-mannitol-l-P concentration and reaction
velocity for D-mannitol-l-P dehydrogenase from PRL-RB
and SFl-MCl . . . . . . . . . . . . . . . . . . . . . . 116

29 The metabolic pathways of D-fructose, L-sorbose,
and D-mannitol in A. aerogenes. . . . . . . . . . . . . 119

viii

ATP
DEAE-cellulose
EDTA

Fru
D-fructose-l-P
D-fructose-6-P
D-fructose-1,6-P2
D-glucitol-l-P
D-glucitol-6-P
D-glucose-6-P
D-gulose-6-P
D-iditol-l-P
D-mannitol-l-P

mercaptoethanol

MES

MES buffer

Mtl

NAD

NADH

LIST OF SYMBOLS AND ABBREVIATIONS

Adenosine 5'-triphosphate
Diethylaminoethyl cellulose
(Ethylenedinitrilo) tetraacetic acid
D~fructose

D-fructose l-phosphate
D-fructose 6-phosphate
D-fructose 1,6-biphosphate
D-glucitol 1-phosphate
D-glucitol 6-phosphate
D-glucose 6-phosphate
D-gulose 6-posphate
D-iditol l-phosphate
D-mannitol 1-phosphate

2-hydroxymethy1mercaptan or 2-
thioethanol

2[N-morpholino]ethanesulfonic acid

0.025 M MES (pH 6.15) containing 0.2%
mercaptoethanol and 2 mM EDTA

D-mannitol

oxidized nicotinamide adenine
dinucleotide

reduced nicotinamide adenine
dinucleotide

ix

NADP

NADPH

PEP:D-fructose PTS

PEP:L—sorbose PTS

Scs

SDS

Sor

D-sorbose—l-P

L-sorbose-l-P

TCA

TEMED

Tris

oxidized nicotinamide adenine
dinucleotide phosphate

reduced nicotinamide adenine
dinucleotide phosphate

phosphoeno1pyruvate-dependent D—
fructose l-phosphotransferase system

phosphoeno1pyruvate-dependent L—
sorbose l-phosphotransferase system

sucrose

sodium dodecyl sulfate

L-sorbose

D-sorbose 1-phosphate

L-sorbose l-phosphate
trichloroacetic acid
N,N,N',N'-tetramethylethylenediamine

tris(hydroxymethyl)aminomethane

INTRODUCTION

During on-going investigations in this laboratory on the
metabolism of L-sorbose (L—xglo-Z-ketulose) in Aerobacter aerogenes,
I became involved in the elucidation of the catabolic pathway in
collaboration with Norman E. Kelker and Richard L. Anderson. In
the initial stages, the elucidation of the pathway was facilitated
by examining mutants defective in hexose and hexitol metabolism.

In contrast to the wild-type strain, mutants missing either enzyme
I of the phosphoenolpyruvate-dependent phosphotransferase system,
D-glucitol-G-P dehydrogenase, or D-fructose-6-P kinase were unable
to utilize L-sorbose as a sole source of carbon and energy. These
findings implicated D-glucitol-G-P, D-fructose-6-P, and some
phosphorylated product of a phosphoenolpyruvate-dependent L-sorbose
phosphotransferase system (PEP:L-sorbose PTS) in the pathway.

There was no evidence for ATP-dependent phosphorylation of L-
sorbose. These data suggested that the product of the PEP:L-
sorbose PTS might be L-sorbose-l-P and, consequently, that there
should be a reductase capable of catalyzing the reduction of L-
sorbose-l-P to D-glucitol-6-P. By using chemically synthesized
substrate, a pyridine nucleotide-dependent L-sorbose-l-P reductase
was in fact detected in crude extracts of cells grown on Lésorbose,

but not in extracts of cells grown on D-glucose, D—glucitol, or

2
glycerol. The L-sorbose inducibility of the reductase was con-
sistent with its participation in L-sorbose metabolism. This
supposition was confirmed by the isolation of an L-sorbose-l-P
reductase-negative mutant which simultaneously lost the ability to
grow on L-sorbose, and by the isolation of a revertant of this
mutant which concomitantly regained the reductase activity and the
ability to utilize L-sorbose.

In crude extracts of L-sorbose-grown cells supplemented with
phosphoenolpyruvate and L-sorbose, it was possible to demonstrate
a precursor-product relationship between L-sorbose-l-P and
D-glucitol-6-P that would be predicted by the catabolic pathway
depicted in Figure 1. The first step of this pathway involves
the phosphorylation of L-sorbose at carbon atom 1 by a PEP:L-
sorbose PTS. This system was found to be present in crude extracts
of cells grown on a variety of carbohydrates, suggesting that the
system is constitutive. The second step is the reduction of
L-sorbose-l-P to D-glucitol-6-P by the inducible L-sorbose-l-P
reductase. The third step is the oxidation of D-glucitol-6-P to
D—fructose-G-P by an inducible D—glucitol-6-P dehydrogenase. This
is a unique pathway that has not been previously described in any
organism.

This dissertation concerns three topics that deal directly
with or were derived from the elucidation of the pathway shown in
Figure 1. The first part reports the isolation and identification
of the product of the PEP:L-sorbose PTS. To prove that the pro-
posed pathway was correct it was essential to show unequivocally

that L-sorbose-l-P is the first intermediate. The second part

.ouucomouomnon ouunouosmum Houaosgonn " m.
omuuosoom ouosomosmua oncouomuu ” m.
Eoummm omuuommcouuosomosmna amonuomuaummm " «. I.ouu ucoauomou
.mocwmouoo nouoooouo< a“ aofifiooouoa oooAMOunA mo hasnuom u fl.mmmwmm.

 

may wcfinzfiouoo mmazucm

 

 

 

$3863: $2230.: 1-3233 manta-I
IoaIw 5qu 8qu IONIw
o- -w - I- w o.I - I- we o..I - IJIUI
55:10 Io-..oI1 . Io..-QI . Ice-I
I- -T__. .I +2 I- ino- -OI seas I- -9" I e. III-OI
I- o- I I- o -OI o- o o-w
”822.. ”Baa ”0%qu Io~In_.

4
describes the purification and properties of the enzyme catalyzing
the second step in the pathway: L-sorbose-l-P:NAD(P) ketoreductase
(or trivialy, L-sorbose-l-P reductase). The third part deals with
a project that arose from the determination of the substrate speci-
ficity of the reductase: the evolution of a novel pathway for the
metabolism of D-fructose by conscripting enzymes normally associated
with D-fructose, L-sorbose, and D-mannitol metabolism. The data
from the first part of my dissertation have been published in a
paper dealing with the elucidation of the pathway of L-sorbose

metabolism in A. aerogenes (1).

LITERATURE REVIEW OF L-SORBOSE METABOLISM

Investigations of metabolic reactions of L-sorbose have been
concerned primarily with those found in microorganisms. There has
been little reported work with higher organisms. Although rat liver
fructokinase can catalyze the phosphorylation of L-sorbose with ATP,
the reaction is considered not to be of physiological significance
(2) .

In addition to the metabolic pathway of L-sorbose in A.
aerogenes which has been discussed in the Introduction to this
dissertation (Figure 1), two other series of reactions in which
L-sorbose is modified have been proposed to occur in microorganisms
(Figure 2). Early studies demonstrated that 2-keto-L-gulonate
accumulated in the extracellular medium when species of Acetobacter,
Alcaligenes, Aerobacter, Bacillus, Escherichia, Gluconobacter,
Klebsiella, Micrococcus, Pseudomonas, Serratia, or xanthomonas
were grown in the presence of L-sorbose (3,4). Later studies by
Perlman and co-workers demonstrated that cell-free extracts of
Gluconobacter melanogenus were capable of converting L-sorbose to
2-keto-L-gulonate (5,6). A subsequent abstract suggested that the
initial step in this conversion was the oxidation of L-sorbose to
L-sorbosone (7). Furthermore, Makover and co-workers reported in

abstract form that extracts of A. aerogenes and Pseudomonas putida

.u«uouomo
mo momooom owwuuoo om oommmoos on use mooouooua nouns mo o>o35uod oooooouo 039 u M.muomfim

 

uncuooumun occuosumunuOuoxuw

 

 

 

 

 

mommo mommo
ouwzmuwmwmmoz n- menfinm A on+
m moxmm mo 1w 1m mo 1% Im
m-w do m-w do
one o
mommw mono-M
m.MdeH«M_ coonuooua
mommo
o-w-om
Io - -m
m- «o:
L.
« 331193 momow
mummmmmum. oumcoHnquIOuoxum ocoooou0mLH
mommo momma mommo
mane do .1- ------------ m-nuo do 11 IA... do
mOIwum meswum notwum
so do m-o do Id do
so -w I oLo ouw
ooow moonw omnw

7
contained an L-sorbose oxidase which converted L-sorbose to L-
sorbosone (8), and that extracts of P. putida and G. melanogenus
contained an L-sorbosone oxidase which converted L-sorbosone to
2-keto-L-gulonate (9). These findings led to the proposal of
pathway A in Figure 2. The latter group also reported the presence
of a 2-keto-L-gulonate reductase in extracts of P. putida and G.
melanogenus which converted 2-keto-L-gulonate to L-idonate (10).
Since this reaction could not be reversed, an earlier proposal that
L-idonate was an intermediate in 2-keto-L-gulonate formation (11)
may not be correct. Due to the fact that the majority of these
findings were reported only in short abstracts of verbal presenta-
tions, it is not possible to evaluate the evidence for this pro-
posed pathway. Regardless of its possible existence, the signifi-
cance and function of the pathway also remain to be established.
Tsukada and Perlman (6) have reported that only 2% of the metabolized
L-sorbose is converted to 2-keto-L—gulonate in G. melanogenus, thus
leaving the fate of the remaining 98% of the metabolized L-sorbose
apparently in need of further investigation.

Reports of an L-sorbose oxidase in Trametes sanguinea (12)
and a 2,6-dichlorophenol indophenol-linked L-sorbose dehydrogenase
in Gluconobacter suboxydans (13), both of which oxidize L-sorbose
to S-keto-D-fructose, in addition to reports on the purification and
characterization of NADP-linked 5-keto-D-fructose reductases from
G. cerinus (14) and G. albidus (15), led some researchers to propose
that L-sorbose may be metabolized in some microorganisms by pathway

B in Figure 2 (16). In order for this pathway to be accepted as an

8
established route for the metabolism of L-sorbose, however, it would
be necessary to (i) determine the physiological electron acceptor
for the L-sorbose dehydrogenase and (ii) demonstrate the presence of
both the L-sorbose dehydrogenase and the S-keto-D—fructose reductase

in the same organism.

PART I

ISOLATION AND IDENTIFICATION OF THE PRODUCT OF THE

PHOSPHOENOLPYRUVATE:L-SORBOSE l-PHOSPHOTRANSFERASE

SYSTEM OF AEROBACTER AEROGENES

10

Introduction

 

In the course of elucidating the metabolism of L-sorbose in
A. aerogenes, L-sorbose-l-P was proposed to be the product of the
phosphoenolpyruvate-dependent phosphorylation of L-sorbose in the
first step of the pathway. Most of the evidence that we had accumu-
lated for this was based on work with mutants missing various enzymes
involved in the metabolism of hexoses and hexitols. The discovery
of the L-sorbose-l-P reductase and the demonstration of the ordered
appearance of L-sorbose-l-P and D-glucitol-6-P in crude extracts of
L-sorbose-grown cells gave additional substance to our proposal that
the phosphorylation was on carbon atom 1 of L-sorbose. To prove
conclusively that L-sorbose-l-P is indeed the first intermediate in
the proposed pathway it was necessary to isolate and characterize
this compound by physical and chemical methods. In this part of my
dissertation I will show that the product of the phosphoenolpyruvate-
dependent L-sorbose phosphotransferase system in A. aerogenes is

L-sorbose-l-P.

Materials and Methods

Chemicals
Authentic L-sorbose-l-P was prepared by the chemical procedure
described by Mann and Lardy (17). Other chemicals were obtained

from commercial sources.

Growth of Cells and Preparation of Cell Extracts
Aerobacter aerogenes strain PRL-R3 was grown at 30° overnight

in 100 ml of mineral broth (63) supplemented with 0.5% D-glucose.

11
Cells were harvested by centrifugation at 12,000 x g_for 10 min.
The supernatant was discarded, the sedimented cells were suspended
in 50 ml of 0.85% NaCl, and were centrifuged at 12,000 x g_for 10
min. After discarding the supernatant fluid, the pellet was
suspended in 3.0 ml of 0.04 M Tris-HCl buffer, pH 7.5, and the
suspension was sonicated for 20 min in a Ratheon 250 W, 10 kHz sonic
oscillator cooled with circulating ice water. Unbroken cells and
debris were removed by centrifugation at 48,000 x g_for 20 min and
the supernatant, containing 6 mg of protein per ml, was removed and

used for the preparation of the product of the PEP:L-sorbose PTS.

Analytical Methods

Protein was determined by the method of Lowry et a1. (18).
Total phosphate and inorganic phosphate were determined by the
modified method (18a) of Fiske and SubbaRow (19). Ketohexose was
determined by a modification of Roe's method as discussed by Putman
(20). Qualitative determination of ketohexoses was done by the
cysteine-H2804 method (21,22). Trimethylsilyl derivatives were
prepared by dissolving the sodium salts of sugar phosphates in
trimethylsilylating reagent as described by Sweeley, Wells, and
Bentley (23): the final concentration was one mg of sugar phosphate
per m1. Gas-liquid chromatography was performed on a Hewlett-Packard
gas chromatograph model 402, employing a 1.2 m column of 3% OV-l
(methyl silicone) at a temperature of 190°. Trimethylsilyl 0-
methyloxime derivatives of authentic L-sorbose-l-P and the product

were prepared and subjected to combined gas chromatography and mass

spectroscopy as described by Laine and Sweeley (24).

12
Results

Preparation of Product of the
PEP:L—sorbose PTS

The reaction mixture (10 ml final volume) contained 0.04 M
Tris-HCl buffer (pH 7.5), 15 mM sodium fluoride, 0.5 mM magnesium
chloride, 10 mM phosphoenolpyruvate (tricyclohexylammonium salt),
20 mM L-sorbose, and 2.0 m1 of cell extract (24 mg of protein).
The reaction was allowed to proceed for 3 hr at 28° and was then
terminated by heating at 100° for seven minutes. Precipitated
protein was removed by centrifugation for 10 min at 48,000 xlg

after cooling the suspension on ice.

Isolation of Product

 

The deproteinized reaction mixture was then applied to a
1 x 7 cm column of Dowex-l, 200-400 mesh, equilibrated in the
bicarbonate form. The column was washed with distilled water until
no more ketohexose was detected in the effluent. The unbound keto-
hexose was nonphosphorylated L-sorbose. The column was then eluted
with a stepwise gradient consisting of 100 ml each of 0.15 M, 0.30 M,

and 0.45 M KHCO3. In this system the monophosphates were eluted

from the column with 0.15 M KHCO and the diphosphates were eluted

3

with 0.30 M KHC03. A substance that gave a positive reaction with

the Roe test was eluted only with 0.15 M KHCO , suggesting that the

3

product was a ketohexose monophosphate. The peak fractions of this
0.15 M KHCO3 wash were combined and the pH of the solution lowered
to 2 by treatment with Dowex-SO (hydrogen form). The solution was

then quickly frozen in a dry ice-acetone bath and lyophilized to

13
dryness. The resulting white powder was dissolved in 5 m1 of
distilled water and the pH adjusted to 7.0 with NaOH. This solu—

tion contained 24 umoles of ketohexose phosphate.

Identification of Product

The determination of total phosphate in the product indi-
cated that there was 0.99 umole of phosphate for every umole of
ketohexose. This result substantiates the conclusion that the
product is a monophosphate, as was suggested by the elution pattern
from the Dowex-l bicarbonate column.

The rate of hydrolysis of the product and authentic L-
sorbose-l-P in l M HCl at 100° exhibited first-order kinetics
with a 50% hydrolysis time of 16 min for each. This time is
distinct from that reported for L-sorbose-6-P, for which 50%
hydrolysis occurs at 60 min (25).

When the trimethylsilyl O-methyloxime derivatives of the
product and L-sorbose-l-P were subjected to combined gas-liquid
chromatography and mass spectrometry, identical fragmentation
patterns were obtained. The fragmentation pattern for the product
is shown in Figure 3. The ions occurring at m£g_103, 205, 307, 414,
and 516 originate from the cleavage pattern shown on the right of
the figure and are those predicted for a 2-ketohexose l-phosphate
without regard for stereochemical configuration (24). The frag—
ments at gig 299, 314, 315, and 387 are found in most mass spectra
of sugar phosphates and are attributable to various ions of

methylsilyl-substituted phosphate (26). The molecular weight of

14

 

.mwm.mg communes onu was muuoeouca

opwuufiou om ousomouo 0:8 .comumusmamooo goowaosooououm cu ouowou uncouaa cumsooonoua moox05000xum m
«o o>wuu>fiuoo orb ouoooouaou :30:o ouauoouuo any .oooouooug we coauufimuooouoso ucoocoooououo>oumdfioco
rosewood era «0 uoooouo 0:0 mo o>guo>auoo camxoamnuoauo ngmoamsuuawuu osu mo asuuooou man: u W.MMMMMM.

 

 

 

 

 

 

com-10mm 0.2 0...:- omz. owm 0mm 0mm 0......- 0mow 0m: 0? 00m 0.5 00m 0mm 0- 02 010%: 00
J -410L .3.». 1.43.0...» trflr-mwlbbqlhipr1tprqfl) 1 1 44).-
4 2n «a 41 H 4114
02 u nth-0.99. 8.. o \
1 . 4. 2.: I 39 n2 ..o~
now. 2.6-w: 2m. < Zn. 2.
So < 2.6.9. 2n a \/
- » mx n; c; 10:
w A
”Ea-w! c? I
_ 1
- 9.8.3 "To: . 2: do
2: _ p8
- Sued-um: E.» a: do
2.. 35:3... _. 3:8
as
2

 

 

 

 

oo_

15
721 is that predicted for the trimethylsilyl O-methyloxime deriva-
tive of 2-ketohexose-l—P.
The cysteine-H2804 method for the determination of ketose
provides a means for distinguishing sorbose from fructose, psicose,
and tagatose (21,22). After 15 hr in cysteine-H2804 the ratio of

absorbance at 412 nm to the absorbance at 604 nm (A /

412 ) 15

A604
0.4 for L—sorbose and greater than 3.0 for the other ketohexoses
(21). When samples of the product, authentic L-sorbose—l—P, and
D-fructose-l-P were treated according to this method the A412/A604
after 15 hr were 0.41, 0.43, and 3.3, respectively. These results
show that the ketohexose moiety is sorbose.

The results of the gas-liquid chromatography of the trimethyl-
silyl derivatives of the product and sugar phosphate standards are
shown in Figure 4. The retention times and peak patterns are the
same for the product (D) and authentic L-sorbose-l-P (C); there are
two minor peaks at 9.5 and 11.0 min and a major peak at 13.7 min.
These retention times and peak patterns are distinct from those of

D-fructose-l-P (A) with major peaks at 9.5, 11.0, and 11.5 min,

or of D-fructose-6-P (B) with one major peak at 10.7 min.

Discussion

 

These results establish unequivocally that the product of
the PEP:L-sorbose PTS is L-sorbose-l-P. The elution pattern from
the Dowex-l-bicarbonate column and the phosphate to ketohexose ratio
show that the product is a ketohexose monophosphate. The hydrolysis

rate in l M HCl at 100° and the mass spectrum of the trimethylsilyl

16

 

 

 

 

 

 

RELATIVE INTENS ITY
o m >

 

 

RETENTION TIME (min)

Figure A - Tracings of the gas-liquid chromatograms of the trimethyl-
silyl derivatives of: A : D-fructose-l-P; _B : D-fructose-6-P;

_c_ : L-sorbose-l-P; and 2 : Product of the PEP:L-sorbose l-phospho-
transferase system.

17
O-methyloxime derivative of the product show that the phosphate is
attached to carbon atom l of the ketohexose moiety.
The ketohexose moiety was identified as sorbose on the basis

of its characteristic A / value after 15 hr reaction with

412 A604

cysteine-H2504. Both the product and authentic L-sorbose-l—P gave
identical peak patterns and retention times when trimethylsilylated
and gas chromatographed.

The product was identified as the L isomer of sorbose-l-P
on the basis of several facts: (i) the mutant studies indicated
that D-glucitol-6-P and D-fructose-6-P were intermediates in the
pathway; (ii) there was a precursor-product relationship between
L-sorbose-l-P and D-glucitol-6-P; (iii) in vitro the reaction path-
way of L-sorbose could be demonstrated with L-sorbose-l-P reductase,
D-glucit01-6-P dehydrogenase, phosphoglucoisomerase, and D-glucose-
6-P dehydrogenase to be L-sorbose-l-P -———+ D-glucit01—6-P -———+
D-fructose-6-P --+-D-glucose-6-P -—-—+ D-gluconate-6-P; and (iv)
if the product were D-sorbose-l-P the reaction for its formation
from L-sorbose would require three epimerizations (at C-3, C-4, and
C-5) and the product of its reduction by a D-sorbose-l-P reductase
would be D-iditol-l-P, which, if oxidized by D-glucitol-G-P
dehydrogenase, would form D-sorbose-G-P, which, if isomerized by
phosphoglucoisomerase, would form D-gulose-6—P. It is highly
unlikely that any of the four enzymes involved--the reductase, the
two dehydrogenases, and the isomerase--would fortuitously use the
four intermediates described. By this reasoning, and with the data
previously discussed, it is established that the product of the

PEP:L-sorbose 1-phosphotransferase system is L-sorbose-l-P.

PART II

PURIFICATION AND PROPERTIES OF L-SORBOSE

1-PHOSPHATE REDUCTASE

18

19

Introduction

 

The second reaction in the pathway of L-sorbose metabolism
in A. aerogenes was demonstrated to be the reduction of L-sorbose-l-P
to D-glucitol~6-P, catalyzed by an inducible L-sorbose-l-P:NAD(P)
ketoreductase (1; see Figure 1). Since this was the first demonstra-
tion of an enzyme that catalyzes the reduction of a 2-ketohexose—l-P,
a study of its properties was undertaken. In this section of my
dissertation I will describe (1) the purification of this enzyme
essentially to electrophoretic homogeneity, (ii) the physical,
chemical, and catalytic properties of the enzyme, and (iii) the
reactions catalyzed by the enzyme, particularly with respect to

equilibrium and identity of products.

Materials and Methods

 

Chemicals

L-Sorbose-l-P was prepared as described in part I of this
dissertation. D-Mannitol-l-P and D-glucitol-6-P were prepared by
the procedure of Wolff and Kaplan (27), i.e., the sodium borohydride
reduction of D-mannose-6-P and D-glucose-G-P, respectively. Adipic
dihydrazide for affinity chromatography was prepared by refluxing
diethyl adipate, hydrazine hydrate (98%), and ethanol (1:2:2,
v/v/v) for 3 hr. The resulting flaky precipitate was crystallized
twice from an ethanol-water mixture and dried in vacuo. Other
chemicals were obtained from commercial sources.

All enzymes and proteins used as markers in the determina-

tions of molecular weight were obtained from Sigma except catalase

20
(W0rthington), D-glucose-6-P dehydrogenase (Calbiochem), and 2-keto-

3-deoxy-6-phosphogluconate aldolase (a gift from W. A. Wood).

Enzyme Assays

 

All assays were done in 150 pl total volume. The rates of
absorbance change (at 340 nm unless otherwise noted) were measured
with a Gilford multiple-sample absorbance recorder thermostated at

30°.

L-Sorbose-l-P reductase. The standard assay for L-sorbose-

 

l-P reductase contained 0.033 M MES buffer (2[N-morpholino]ethane-
sulfonic acid), adjusted to pH 6.15 with NaOH; 0.33 mM NADH (unless

specified as NADPH); 1.0 mM MnCl and 3.3 mM D-fructose-l-P (unless

2;
specified as L-sorbose-l-P). One unit of reductase is defined as

the amount that catalyzes the oxidation of one umole NADH per min

in the standard assay.

Other enzyme assay_. In the case of those assays coupled to
reactions involving reduction or oxidation of nicotinamide adenine
dinucleotides, the reactions were monitored for increase or decrease
in absorbance at 340 nm. Escherichia coli alkaline phosphatase was
assayed by the procedure of Garen and Levinthal (28); the assay
mixture contained 1.0 mM pfnitrophenyl phosphate in 1.0 M Tris-RC1
buffer, pH 8.0. The reaction was monitored by increase in absorbance
at 410 nm. Lactate dehydrogenase from beef heart was assayed in a
reaction mixture containing 0.067 M glycylglycine buffer (pH 7.5)-
3.3 mM sodium pyruvate, and 0.33 mM NADH. Glucose-6-P dehydrogenase

from yeast was assayed in a reaction mixture containing 0.067 M

21
glycylglycine buffer (pH 7.5), 3.3 mM D-glucose-6-P, and 0.33 mM
NADP+ (adjusted to pH 7.0). Yeast hexokinase was assayed in a reac-
tion mixture containing 0.067 M glycylglycine buffer (pH 7.5), 0.33
mM NADP+ (adjusted to pH 7.0), 3.3 mM D-glucose, and nonlimiting

amounts of yeast D-glucose-6-P dehydrogenase.

Protein determinations. Protein was determined by the method

 

of Lowry et a1. (18) except in those cases where protein concentra-
tion made it more desirable to determine concentration by measuring
the absorbance at 220 nm (29). Hemoglobin from sheep erythrocytes

was determined by the absorbance at 410 nm.

Determination of Reaction Products

 

Enzymatic determination. D-Fructose-l-P was determined by

 

a coupled end-point assay containing 0.067 M glycylglycine buffer

(pH 7.5), 3.3 mM ATP, 6.6 mM MgCl 0.33 mM NADH, nonlimiting

2-
amounts of rabbit muscle aldolase, triosephosphate isomerase, o-
glycerophosphate dehydrogenase, and partially purified D-fructose—
l-P kinase. D-Mannitol-l-P and D-glucitol-6-P were determined under
identical assay conditions; the reaction mixture contained 0.067 M
Tris-HCl buffer (pH 9.0), 0.33 mM NAD+, 0.33 mM NADP+, and nonlimit-
ing amounts of phosphoglucoisomerase and D-glucose-G-P dehydrogenase.
The end—point assay for D-mannitol-l-P also contained nonlimiting
amounts of partially purified D-mannitol-l—P dehydrogenase (30);
the end-point assay for D-glucitol-G-P also contained nonlimiting

amounts of partially purified D-glucitol-6-P dehydrogenase (30).

All assays were performed in a total volume of 150 ul.

22

Preparation of the hexaacetate derivatives of the dephos-
phorylated product. Approximately 50 pg of the enzymatically
dephosphorylated product of the reductase-reduced D-fructose—l-P
was dried by passing a stream of dry nitrogen gas over the solution
containing the product. The residue was dried by suspending it in
absolute ethanol and evaporating with nitrogen. The sample was
then placed in a vacuum desiccator at 50° for 3 hr. After drying,
the residue was dissolved in 50 ul of acetic anhydride, the con-

tainer sealed, and heated in a 120° heating block for 1 hr.

Gas-liquid chromatography. Gas-liquid chromatography of the

 

hexaacetate derivative of the dephosphorylated reaction product was
performed on a Perkin Elmer 900 gas chromatograph as described by
Jones and Alberscheim (31). The manifold temperature was 250°,

the injection port temperature was 265°, and the column temperature
was programmed from 130° to 180° at a rate of 1 degree per min.

The column (1.83 m x 2 mm) consisted of a solid phase of Gas-Chrom
P, 100-200 mesh with a liquid phase consisting of 0.2% ethylene
glycol adipate, 0.2% ethylene glycol succinate, and 0.4% XF-llS
silicone oil. The data were processed by an Autolab System 4
computer. Standards were run of the hexaacetate derivatives of
rhamnitol, fucitol, arabitol, xylitol, mannitol, galactitol,

grlucitol, and inositol. The standards were obtained from Dr.

Derek Lamport.

23

Preparation of NADP-Substituted Sepharose
4B for Affinity Chromatography

 

 

Activation of Sepharose 4B and coupling to adipic

dihydrazide. Sepharose 4B was activated by the method of March

 

et a1. (32). Sepharose 4B was washed with 100 vol of distilled
water on a coarse sintered-glass funnel and suspended in two

volumes of l M sodium carbonate by slow stirring on a magnetic
stirrer. The rate of stirring was increased and 0.05 vol of a 66%
(w/v) solution of cyanogen bromide in acetonitrile was added and
stirred vigorously for 2 min. The slurry was then poured onto a
sintered-glass funnel and washed with 5 to 10 vol each of 0.1 M
sodium bicarbonate (pH 9.5), distilled water, and 0.2 M sodium
bicarbonate (pH 9.5). The slurry was filtered under vacuum to a
moist cake and the cake transferred to a plastic bottle containing
one volume of a cold (4°) saturated solution of adipic acid dihydra-
zide in 0.2 M sodium bicarbonate (pH 9.5). The coupling was allowed
to continue for 20 hr at 4° with gentle stirring. The coupled
Sepharose-adipic acid dihydrazide was filtered on a sintered-

glass funnel and alternately washed with distilled water and 0.1 M
sodium chloride until the washings gave only a slight color test
with 2,4,6-trinitrobenzene sulfonate (33). Samples of substituted
sepharose, when subjected to this test, became deep red in color,
indicating the presence of hydrazide. The gel could be stored for

several days at 4° in 0.02% sodium azide.

Binding of activated NADP to Sepharose-adipic hydrazide (34).

 

+
NADP was activated by adding a cold (0°) solution of 0.02 M sodium

24
metaperiodate to an equal volume of cold (0°) 0.02 M NADP+ adjusted
to pH 7.5 with NaOH. The mixture was incubated on ice in the dark
for 60 min. Sepharose-adipic acid dihydrazide (compound I in
Figure 5), washed free of sodium azide and washed with 10 vol of
0.1 M sodium acetate buffer, pH 5.0, was suspended in 3.5 vol of
the same buffer and 0.5 vol of the periodate-activated NADP+
(Figure 5, II) added. The reaction was allowed to proceed for 3
hr with gentle stirring at 4°, at which time 7.5 vol of 2 M NaCl
were added to remove uncoupled nucleotide from the gel. The mixture
was stirred for an additional hour and then filtered on a coarse
sintered-glass funnel. The filtrate was saved to determine the
extent of coupling to the gel. The cake was washed with 5 vol of
0.02% sodium azide and stored in the same solution. Generally,
there were 1.7 to 2.5 umoles of NADP+ coupled per m1 of gel

(Figure 5, III).

Polyacrylamide Gel Electrophoresis

 

General procedures. Gel tubes 0.5 cm x 10 cm were prepared

 

by cleaning with chromic acid solution, neutralizing with sodium
thiosulfate, and rinsing with distilled water. The cleaned tubes

were immersed in a solution of 0.5% Photoflo for 30 min and the

tubes air dried. Components of the gel mixture were degassed

prior to mixing. The bottoms of the gel tubes were sealed with
paraffin film. Electrophoresis was performed with a Spinco Duostat
regulated D.C. power supply at constant current. After electrophoresis

the gels were removed from the tubes by directing a stream of water

25

 

 

HI
o o
III. I_
Io\ /IU
III/Io\\\mww-w
ooIEuquoon OMMuOu
o
_
ouI-o
mooo Io
o
I.\I I/II
III/onxxmmo-o

ochNo<

.m: omouocoom mo oowuouwumosmumamz ecu mo uoooouo ocu mouowooauoucH u w.ou=mmm

woouocoomuunamzuo Nmo Nmo Nmo N:ouu §\\wm m%/

_/o\/IIHI

ooIaaquooIz

III

ono-o

_
LI

mooo m

s.\/... I/

 

\IIM

$0

 

ochoo<

_m

mmz-oz-m-mmo-mIo-mIo-mIo-m-mz--mmooueomm

26
delivered with a 22 gauge hypodermic needle attached to a 50 m1
syringe. Gels to be stained for protein were immersed in 18 ml of
12.5% trichloroacetic acid (TCA) for 20-30 min prior to the addi-

tion of the staining solution.

Preparation of native_gels for determination of homogeneity.

 

Gel tubes were filled to a height of 9 cm with a solution containing
7.5% acrylamide, 0.18% bisacrylamide (N,N'-methy1enebisacrylamide)-
1.25% glycerol, 0.028% TEMED (N,N,N',N'-tetramethylethylenediamine),
0.09% ammonium persulfate (prepared fresh each time), and 0.375 M
Tris-RC1 buffer (pH 9.0). The solution was overlaid with 50 ul of
the electrophoresis buffer with the mercaptoethanol deleted and
allowed to polymerize for 60 min at room temperature. The tubes
were pre-electrophoresed at a constant current of 2 mA per tube at
4° for 60 min. The electrophoresis buffer contained 0.2 M glycine,
0.02 M Tris, 0.2% mercaptoethanol with a pH of 8.8. Prior to
application of the protein sample, 0.05% bromphenol blue dye in

10% glycerol in a volume of 2.5 ul was applied to the gels and

the gels electrophoresed for approximately 5 min. The sample was
then applied in a total volume of 10 to 20 ul and the gels were
electrophoresed at 2 mA per tube for 2.5 to 3 hr, at which time the
tracking dye was within 1 cm of the bottom of the tube. Gels were
stained for protein by adding 2 ml of an aqueous solution of 1%
Coomassie Brilliant Blue R to 18 ml of 12.5% TCA and allowing to
stain overnight. Destaining was achieved by immersion in 12.5% TCA
with frequent changing of solution until the background was a faint

blue.

27

To determine the location of L-sorbose-l-P reductase activity
a separate series of gels was run using 0.1 M sodium phosphate
buffer, pH 6.2, instead of the Tris buffer. Gels were run for
6 to 8 hr at 4 mA per tube. After gels were removed from the
tubes, one set was stained for protein and the other was placed
in a 1 x 10 cm glass tube, sealed at the bottom, containing 0.05
M MES buffer (pH 6.1), 3.3 mM D-fructose-l-P, 0.33 mM NADH, 1.0
mM MnClZ, and 0.2% mercaptoethanol. A control tube was run with
the D-fructose-l-P omitted. Gels were incubated at room temperature
for 30 min and scanned at 340 nm in a 1 cm x 12 cm quartz gel-
scanning cuvette on a Gilford 240 spectrophotometer with a model

2410-5 linear transport attachment.

Preparation of native gels for determination of molecular
weight. Molecular weight in native gels was determined by the method
of Hedrick and Smith (35). Gels were polymerized and submitted to
electrophoresis in the same manner as described above except that
the concentration of acrylamide was varied from 6 to 10% and the
bisacrylamide to acrylamide ratio was maintained at 1:30. The gel
runs varied due to the change in gel concentration with each run
from 1.5 hr for 6% to almost 3 hr for 10% gels. Gels were removed

and stained for protein as previously described.

Preparation of SDS gels for determination of subunit
molecular weight. Gels were prepared by cleaning the same glass
tubes used in the native gel electrophoresis studies and soaking

the tubes in 1% SDS (sodium dodecyl sulfate) for 1 hr, rinsing with

28

distilled water, and placing in a drying oven until dry. The gels
were prepared according to the procedure of Fairbanks et a1. (36).
The gel solution contained 5.6% acrylamide, 0.21% bisacrylamide,
1% SDS, 0.15% ammonium persulfate, 0.025% TEMED, and 0.04 M Tris-
0.02 M sodium acetate-0.002 M EDTA, pH 7.4. Gels were allowed to
polymerize for 40 min at room temperature with an overlaid solution
containing 0.15% ammonium persulfate, 0.05% TEMED, and 0.1% SDS.
After polymerization the overlay solution was replaced with the
electrophoresis buffer: 0.04 M Tris-0.02 M sodium acetate-0.002 M
EDTA (pH 7.4) containing 1% SDS and allowed to stand at room
temperature overnight. Prior to application of the sample the
gels were submitted to electrophoresis for 30 min at 4 mA per tube.

Samples were prepared by dissolving 1 mg of standards or
the enzyme preparation in a l-ml solution containing 1% SDS, 10%
sucrose, 10 mM Tris, 1 mM EDTA, at pH 8.0. Immediately prior to
electrophoresis, mercaptoethanol was added to a final concentration
of 2% and the sample heated to 100° for 15 min. Two-hundredths
vol of a 10% solution of pyronin B in 5% sucrose was added to the
sample and 20 ul of sample was added to each gel. Gels were sub-
jected to electrophoresis for 4 hr at 4 mA per tube, the gels were
removed from the tubes, and the dye fronts were marked by inserting
a small sliver of bamboo into the center of the dye band.

Gels were stained for protein in a solution containing 10%
TCA, 33% methanol, and 0.4% Coomassie brilliant blue R overnight.

Destaining was performed by immersing the gels in a solution of

29
10% TCA and 33% methanol overnight and completing the destaining

with a solution of 10% TCA.

Other Methods for Determining Molecular Weight

Sucrose density gradient centrifugation. Sucrose density

 

gradient centrifugation was performed using 1.2 x 7.5 cm cellulose
nitrate centrifuge tubes containing 4.6 m1 of 5 to 20% sucrose in
0.05 M MES buffer (pH 6.1) with 0.2% mercaptoethanol. The linear
gradients were poured at room temperature and then the tubes were
cooled for several hours to 4°. The sample, containing partially
purified L-sorbose-l-P reductase and standards of sheep erythrocyte
hemoglobin and E. coli alkaline phosphatase, was applied in a volume
of 0.1 ml in buffer containing 0.05 M MES (pH 6.1), 0.2% mercapto-
ethanol, and 5% sucrose. The centrifugation was performed in a
model L-2 preparative Beckman centrifuge with a SW 50.1 rotor for
16 hr at 35,000 rpm and 0°. After centrifugation the tubes were

punctured and 5-drop fractions collected and assayed.

Sedimentation equilibrium ultracentrifugation. Sedimentation
equilibrium centrifugation of L-sorbose-l-P reductase was done by
standard procedures (37,38) on a Beckman Model E ultracentrifuge
equipped with Rayleigh optics. Three samples of 0.5, 1.0, and 2.0
mg protein per ml were used in standard cells oriented for 6°
above, 6° below, and 0° with respect to the reference lines. The

centrifugation was done for 24.5 hr at 15° at a speed of 12,573 rpm.

30

Amino acid analysis of L-sorbose-l-P reductase. L-Sorbose-

 

1-P reductase which had been purified 99% (as determined by acrylamide
gel electrophoresis) was carboxymethylated by the method of Crest-
field as modified by Robertson et a1. (39) to protect cysteine
residues. Amino acid analysis was performed on a noncommercial

amino acid analyzer (39). Tryptophan was determined by the method

of Edelhoch (40).

Determination ofng. All pH measurements of buffers were

 

determined prior to the addition of mercaptoethanol on a Beckman

Zeromatic pH meter with a Sargent 30070-10 combination electrode.

Results

Purification

 

L-Sorbose-l-P reductase appeared at first to be relatively
unstable, since most of my attempts to purify it by conventional
means resulted in low recovery and little purification. Various
steps which were attempted but which failed to give the desired
purification and recovery were: precipitation by ammonium sulfate
or by lowering the pH; ion exchange chromatography on phosphocellu-
lose or carboxymethylcellulose; gel filtration on various commercial
gels; and adsorption to calcium phosphate or bentonite. Hydroxy-
apatite chromatography was partially successful; however, the
phosphate buffer used in the elution inhibited the enzyme activity,
and the elution pattern was not reproducible.

The enzyme's stability could be increased by the addition of

0.2% mercaptoethanol and 2 mM EDTA (ethylenediaminetetraacetic

31
acid) to the buffer. Buffer containing 0.025 M MES (pH 6.1) and
mercaptoethanol and EDTA was used in all subsequent purification
steps and will be referred to as MES buffer; in cases where the MES
concentration was increased to 0.05 M MES, this will be noted. The
enzyme could be stored for several weeks with a minimal loss of
activity in 0.05 M MES buffer containing 20% glycerol at -20°.
With the exception of dialysis, any manipulative steps resulted in
a loss of greater than 10% of the activity. Such steps include
ultrafiltration, lyophilization, filtration on Sephadex G-25, and

ammonium sulfate and pH precipitation.

Preparation of crude extracts. Cells grown on 0.5% L-sorbose

 

in mineral broth (63) and harvested as previously described (1)
were suspended in 50 m1 of 0.05 M MES buffer. The cells were dis-
rupted and the crude extract prepared as described in Section I of
this dissertation. A sample was removed and the L-sorbose-l-P
reductase activity determined by the standard assay described in
methods. To get an accurate determination of reductase activity in
the crude extract it was necessary to run a control with D-fructose-
l-P omitted. This control measured the endogenous NADH oxidase
activity present; in crude extracts of A. aerogenes this activity
was about equal to that of the reductase. The specific activity of
the L-sorbose-l-P reductase, as determined by subtracting the
endogenous rate of NADH oxidation from the rate in the presence of
D-fructose-l-P, varied from 0.040 to 0.075 units per mg protein in

different preparations.

32

Chromatography on DEAE-cellulose

 

First DEAE-cellulose column. The 50 ml of crude

 

extract was applied to a 1.6 x 8 cm column of diethylaminoethyl
cellulose (DEAR—cellulose from Sigma, Lot No. 16C-l940). After

the sample was applied, the column was washed with one column

volume (16 ml) of MES buffer. An assay of the column wash-through
indicated that all of the reductase activity was retained on the
column but that all of the NADH oxidase activity was present in the
effluent. Figure 6 is an elution profile of the first DEAE-
cellulose column; the reductase activity elutes at a sodium chloride
concentration of approximately 0.15 M. The total activity that
could be recovered from this column was 67%; fractions 31 through
46, representing 54% of the activity applied to the column, were
combined and dialyzed against MES buffer containing 20% glycerol for
18 hr at 4°. The peak fraction, number 36, represented a purifica-
tion of 7.5-fold, whereas the combined fractions (31-46) represented

a 5.1-f01d-purification.

Second DEAF-cellulose column. The dialyzed fractions
from the first DEAE-cellulose purification step were applied to a
1.2 x 8 cm column of DEAE-cellulose and washed with one volume (9
ml) of MES buffer containing 0.05 M NaCl. There was no detectable
reductase activity in the wash-through.‘ Bound proteins were eluted
with a 0.05 M to 0.25 M NaCl gradient (90 ml total volume) in MES
buffer. Fractions of 1.4 ml were collected and assayed for protein

and L-sorbose-l-P reductase activity (Figure 7). The activity peak

33

(o------o)

mo ucowoouw Heuoma 0 new: mo3 cOwuon

.Imooso mm: II Iooz z muo as o

“Is my momzoz onaoamo

 

oo om o: on om oI o
0. q. . .. .. I q 0 0.0
\
INFJK. JV
fix
H. m .
. I m.o
I
_
_
..
d
H m. ..
.
I _..
\I _
I b ._
d —
a _
1 —
..
m... .. I m4
I .e
I
I
I
I
s
I
II I
m Ob [IOoN

 

 

 

 

)

(1m 13d SJIUH) ALIAIIOV HSVIOHGHH d'I'HSOflHOS'T

(

.caoaoo omoHoHHoonmmmn umpfim mo sawmouo :owuoam : m ouswfih

34

.cofiumofimfiusm uwnuusw How UwCHnEoo mumz Nv zmsounu mm mcoauomuw can

.vmuomaaoo mum: IHE v.av mcoauomum .nfidaou umuaw any you “mumsn wEMm may

2H z m~.o ou mo.o Eoum Homz mo ucmflcmum Hauom m zuaz amusam mu: cEdHoo
mph .sfidaoo mmoHsHHmunm4ma ocoomm mo mawmoum coHusHm .5 wusmfim

.-5.0

35

(1m 19d 3m) ulamoua
(cw-mo)

O
m

- h.0

 

 

2.5'

T

 

200’P

m
o
H

1.0

U\
C)

(0--—-—0)
(1m 18d 331““) ALIAILDV zsvmonaau a-I-zsoquos-1

  
  
  

FRACTION NUMBER (1.u m1)

F1 ure 1

36
eluted from the column at a concentration of approximately 0.15 M
NaCl and represented an overall purification of 13.7-fold. Frac-
tions 25 through 38, representing 74% of the activity applied to
the column, were combined and dialyzed overnight against MES buffer
containing 20% glyercol. Only 5% of the activity was lost in the

dialysis step.

Affinity chromatography on NADP—substituted Sepharose

 

Previous studies with NAD or NADP covalently attached to
immobilized supports such as Sepharose have revealed that many
dehydrogenases exhibit a certain specificity in binding towards
the cofactor (41—43). Those enzymes which require NADP(H) seem
able to bind to either NAD- or NADP-substituted sepharose. This
specificity suggested that an NADP-sepharose column should bind
L-sorbose-l-P reductase (since it could use either NADH or NADPH
as a cofactor) while allowing all enzymes which do not bind NADP
to wash through. Elution with NAD+ or NADH should then separate
those enzymes binding more tightly to NADP from those which might

use either cofactor.

Preliminary studies; elution with NaCl. Initial experi-
ments involving a NADP-substituted sepharose column were directed
towards determining whether L-sorbose-l-P reductase would bind to
such a column, whether the enzyme could be recovered quantitatively
from the column, and whether this type of purification step would
enable the reductase to be greatly purified. Figure 8A shows the
results of one of these early studies. The column was a 0.8 x 8

cm column of NADP-sepharose and the gradient was 0.0 to 0.5 M NaCl

37

 

0.0 to 0.5 M NaCl

AHa you mav szaomm
35.3

-¢ 0.2

 

 

 

 

 

 

30

o
FRACTION NUMBER (1.0 ml)

2

 

 

 

200-

 

\1-../ 55055
O
z; m nonualncnc
1 I.
1 AB
(
R
E
mm
N
0
m
Wm W
I
O I
.II I
I
I
I
I
I
I
I
I
I
I
I
. . .
o. s o a. a a
.1 no _ 1r 1. mm
Iosllla

AIS Moo muwasv WHH>HHU< mm<HUDQNM mquMmOmMOmnA

‘.
6
O

.
h.
0

002-

-P (F-l-P). Elutions were performed in MES buffer.

lution with NaCl; B : Elution with various concentrations of

E

'+

Figure'§ - Affinity chromatography of L-aorbose-l-P reductase.
NAD and D-fructose-l

A o

38
in MES buffer with EDTA omitted (30 m1 volume). The sample (6
ml of a preparation from the first DEAR-cellulose step) was applied
to the column and washed with 6 ml of buffer. In this case 25% of
the reductase activity failed to bind to the column and 81% of
the protein did not bind. Elution with NaCl gave a recovery of
64.5% of the activity applied to the column or 85% of the reductase
that bound. The protein eluted at the same concentration of NaCl
as the reductase and represented 17% of the protein applied to the
column. The purification for the combined activities (fractions
13-19) was 4.5-fold or a 13.4-fold purification overall. This
purification was not as great as the purification from later
columns where the elution was made more specific by eluting with

4.
NAD or NADH instead of NaCl.

+ .
Elution with NAD and D-fructose-l-P. Previous studies

 

had shown that some enzymes could be eluted from affinity columns
by the presence of substrate, cofactor, or both in the elution
buffer (41-43). It was reasoned that since L-sorbose-l-P reductase
could use either NADH or NADPH, elution in the presence of NADH or
NAD+ would give additional specificity in elution since this should
cause only those enzymes which have an affinity for both NADP and
NAD+ to elute. Attempts to elute with NAD+ alone gave no better
purification than using NaCl and, in fact, gave multiple bands when
examined on acrylamide gels. Elution was tried in the presence of
D-fructose-l-P but this resulted in a very slow leak of reductase
from the column at such low levels that it was deemed impractical

for the purposes of purification. Ohlsson et a1. (41) showed that

39
lactate dehydrogenase could be selectively eluted from an AMP-
sepharose column by the formation of a ternary complex of the enzyme
with pyruvate and the oxidized form of its cofactor, NAD+. This
approach was attempted with L-sorbose-l-P reductase by elution with
NAD+ and D-fructose-l-P. Figure 8B shows the elution pattern from
such an experiment. A preparation from a first DEAE-cellulose
column (10 ml) was applied to a l x 6.4 cm column of NADP-sepharose;
25% of the activity and 54% of the protein did not bind to the
column. Elution with 0.5 mM NAD+ resulted in a recovery of 23%
of the reductase activity and 5% of the protein with the peak puri-
fication of 5.5-fold in fraction number 24. Elution with 0.5 mM
NAD+ and 1.0 mM D—fructose-l—P gave a recovery of an additional 18%
of the reductase activity and 3% of the protein, whereas when 2.5
mM NAD+ and 2.5 mM D-fructose-l-P was used to elute the column,
22% of the reductase activity and 2.5% of the protein were recovered.
The remaining protein could be eluted with a 0.5 M solution of NaCl
in buffer; there was no detectable reductase activity. The peak
fraction, 38, represented an 8.3-fold purification or a 30-fold
purification overall. Acrylamide gel electrophoresis of the frac-
tions that eluted with 2.5 mM NAD+ and 2.5 mM D-fructose-l-P
revealed a major band (which activity gels revealed as possessing
the reductase activity) and three contaminants. Since this pro-
cedure resulted in the loss of 66% of the activity which bound to
the column, further studies were made to determine if a better

procedure could be found.

4O

Elution with NADH. Attempts were made to determine if

 

NADH would prove to be more specific than NAD+ in eluting the reduc-
tase from the sepharose. A preparation from a first DEAE-cellulose
column was applied to a l x 6.4 cm column of NADP-sepharose; 33%

of the activity applied did not bind and 73% of the protein remained
unbound. A 40-ml gradient of 0 to 50 uM NADH was applied to the
column. The results of the elution are shown in Figure 9. The
initial peak in the figure represents part of the reductase activity
from the buffer wash representing an additional 13.6% of the
activity applied, and 1.8% of the protein. The gradient started

at fraction 10 and the majority of the activity and protein binding
to the column was eluted with 10 to 20 uM NADH. The peak fraction,
number 14, represented a purification of 11.7-fold or 33-fold
overall. The total activity recovered from the column by elution
with NADH was 45.3% or 85% of the activity which bound to the column.
Attempts to increase purification by the presence of D—fructose-l-P
revealed that no activity was eluted from the column in the presence
of both NADH and D-fructose-l-P although some protein was. These
results suggested a modification of procedure to take advantage of

this property.

Elution with NADH and D-fructose-l-P. The final pro-

 

cedure for the affinity chromatographic purification of L-sorbose-
l-P reductase rested on a series of observations: (i) L-sorbose-
l-P reductase actually seemed to bind tighter to NADP-sepharose in
the presence of both D-fructose-l-P and NADH; (ii) metal ion such

+2 . . .
as Mn was necessary for the enzyme's catalytic actiV1ty when it

41

 

 

 

 

 

 

C: 08 - r 01+
B
L0
0)
D.
U)
u
-.-4
S
V {0.3
E
H
>/\
H
BI
<C
at, - 0.2
B
E
”f
.7 - 0.1
[21
U)
E”: .
O O
U.) \
r-1 \‘
O r 9 A O
10 20 50

FRACTION NUMBER (1.1 1111)

Figure 2 - Affinity chromatography of L-sorbos'e-l-P reductase:
elution with NADH. Column was eluted with a SO-ml gradient of
0 to 50 pH NADH.

(Om-0)

PROTEIN (mg per m1)

42
had been treated with EDTA; (iii) Mn+2 was not necessary for the
binding of the enzyme to NADP since EDTA-treated enzyme would
bind to NADP-sepharose; and (iv) with EDTA-treated enzyme in the
absence of Mn+2 the D—fructose-l-P had no effect on the binding of
the enzyme to NADP-sepharose. These properties of the enzyme
suggested that elution of the column in the presence of D-fructose-
l—P, MnClZ, and NADH would remove those enzymes binding to NADP-
sepharose which have no affinity for D-fructose-l-P, while leaving
the reductase tightly bound. Elution with NADH alone should then
allow the specific elution of the reductase. Figure 10 shows the
results of such a procedure. The dialyzed combined fractions from
the second DEAR-cellulose step were applied to a 1 x 6.3 cm column
of NADP-sepharose in a buffer containing 0.025 M MES (pH 6.1),
0.2% mercaptoethanol, 2 mM EDTA, and 20% glycerol. The column was
then washed with the same buffer with EDTA and glycerol omitted.
The wash-through contained 58% of the protein and less than 1% of
the reductase activity. When the column was eluted with the same

buffer used for the wash containing 1 mM MnCl 2 mM D-fructose-l-P,

2:
and 20 mM NADH, 3.2% of the reductase activity and 5.9% of the pro-
tein were eluted from the column. The column was then eluted with
the wash buffer containing only 2 mM EDTA and 20 mM NADH, giving a
recovery of 75.5% of the activity and 23% of the protein. When
these fractions were combined, they represented a 32-fold purifi-
cation overall. Acrylamide gel electrophoresis of a lyophilized

preparation of these fractions revealed two minor contaminants

comprising less than 5% of the total protein (Figure 11). This

43

.mcanwmouosuwno paw
Iwmmuozpwu mnmnmmonuomnq mo xnmmumoumeouno muwcwmw< n mm.wummHm

is 7: $5.52 22.85

momz :uwz cofiusfiw

 

 

 

    
 

 

 

   

  

om om OI on om OH
0 1r! 0 c c o a I O
Oar \ .Aamm-Imwwcw. .m .IN an
a I... . a ... ..
My I w m
p. u I
m.o 6 II . _ I
I II _ II
Why no W
I m.o
WW I I m
mm o.H o m I
numw .
m/I I
I
3 Am II
.7 I. ..
J . _ I
m. m I I
(x 04
MO
ma<z :1 om _ I
mcIcImucoo nowmsb am: I U m
I
o.m mauazrza H I I
can «moInomouosumnn I m
:5 m £32 :1 om fit. w.
Iwmuuaao «somI ummmsa mm: I m

Avenuqao «ynmI hummus mm: “ < . m.~

  

)
VIOOOEN a-I-zsosuos-I

(
(1m lad 931”") ALIAILDV as

.HHH .. $4 mm 3953 can .. uuumsb ouunnuono Havoc

5 «6 um us can How u m. 8339:. 3393 m: 03 9:33 minimum vacuum Scum 035m .. HHH
goofing—u cgououn m1 0N: 55500 huucumwm unuaw aouw 0.35m u HH IavoIHn—mu caououn w: CNS.
55.30 23:33:39 cocoon scum 0.35m u H 33.35 unashawunaua 3 0.0 mm u.» can 3qu n <
.aoauuuawausm mo mom-um $5.32, an umuuusvou mnI-Ononuouua mo quw ogacahuoubom .. m mam.

 

 

45
combined eluent was used for most of the characterization studies.
Where a purer preparation was required the combined fractions from
this step were applied to a second affinity column as described

below.

Second affinity purification step. A second affinity

 

purification was performed by applying the 32-fold purified prepa-
ration from the first affinity step to the same NADP-sepharose
column and eluting with NAD+ and D-fructose-l-P as described on
pages 38 and 39. Figure 12 shows the elution profile of this
column. Since the preparation had been lyophilized, some of the
reductase was inactivated (the specific activity had decreased to
1.21) and so 68% of the protein applied did not bind to the column
and 18% of the reductase activity did not bind. In the presence

of 1 mM NAD+ 12.5% of the reductase activity and 14% of the protein
eluted from the column. Increasing the NAD+ concentration to 2.5
mM and adding 2.5 mM D-fructose-l-F to the MES buffer (with 1 mM
MnClz) allowed a recovery of 49% of the reductase activity and 20%
of the protein applied to the column, and gave an overall purifica-
tion of 4l-fold. This preparation was lyophilized and applied to
acrylamide gels (Figure 11). One minor contaminant was present
representing less than 1% of the total protein. This preparation
was used in the amino acid analysis. A summary of the purification

data is presented in Table l.

L-SORBOSE-l-P REDUCTASE ACTIVITY (units per m1)
(H)

2.0

1.6

1.2

0
CD

.0
4:—

46

- -0.5

— 0.1+

O
\JJ

‘

 

(0--o)

PROTEIN (mg per m1)

9
m

 

.LO.1

 

 

 

 

ERAcrION NIMBER (1.0 m1)

Figure _1_2_ - Second affinity chromatography step. Column was eluted

with:

A-
2
.9

MES buffer

: MES buffer plus 1 mM NAD+
: MES buffer (EDTA omitted) plus 2 5 mM NAD+ , 2 5 mM

D-fructose-l- P, and 1 NM MnCla.

.mmum casaoo auflsflumu maaz umufim
on» How mm pwumsnom mum mOsHm> .muaummouoauMIo 28 m.~ can +o¢z Es m.N spas oouzao :EsHoUm

.moum was» 0» wouOOnnsm coon moum msow>oum on» scum mafia: 059 Ham on: vocamuno on vaso3
and? $5.35 :33 pcommmuuoo on pennants can omuuomou mosaflr £954 2: cm :33 noun—IO EHOUN

.oom .H.o mm .Ouscws mom omosku mualmmouosumuo mo moaofi mm wouuommu undead

 

47

.usao +o¢z a muanm

 

m.av 4H owm.~ m.m ~.HH o.o~ m .H00 auflcfimum maaz
~.usao madz z: om

m.mm mm vom.~ m.m v.mm o.o~ .Hoo auficammm modz
mnumw .o: .uouum

m.m mm Hwo.o vv m.mm m.mH .cesaoo mdmo cam
oqum .oc .uomum

H.m 4m mom.o «Ha m.a¢ m.hH .aasHOO mama umH

a ooa ano.o mmoa m.m> o.mm H.m mm .pumuuxo mosuo

vmwmwuzm >uo>oowu I.uoum ma\muflcdv Imsv IGHE\mOHOE:I IHEI moum coaumowmwnsm

oaom unmoumm >ufl>wuom cfiououm Hmuoa muwcd Hmuoa QEUHO>
ownaoodw H

 

mmmuosomu mIHIOmOQHOmIA mo coaumuamauom Mo huusssm .H manna

48

Properties

Determination of molecular weight

Native acrylamide gel electrophoresis. When a graph of
the log of protein mobility (relative to a marker dye at a constant
pH) versus the percentage of acrylamide in the gels is plotted, a
series of lines is obtained. The slope of each of these lines is
directly proportional to the molecular weight of the protein pro-
ducing the line (35). This type of molecular weight determination
was performed for L-sorbose-l-P reductase using three marker pro-
teins: bovine erythrocyte carbonic anhydrase, MW 29,000 (44);
bovine serum albumin, MW 65,000 (45); and P. putida 2-keto-3-
deoxy-G-phosphogluconate aldolase, MW 73,000 (46). Figure 13 shows
the relationships obtained from such a molecular weight determina-
tion. Part A of Figure 13 is a graph of the log of relative mobility
versus the percentage of acrylamide in the gel at pH 8.8. The
slopes of the lines were determined by the method of the least
squares and these slopes plotted against the molecular weight in
Part B of Figure 13. The accuracy of the slope for the carbonic
anhydrase (Figure 13A) could be questioned due to the scatter in
the points; however, Figure 138 shows that the slope determined by
statistical methods must be correct or the line would not have
intersected at the origin, as it must in this type of graph. When
the slope of L-sorbose-l-P reductase is plotted on the standard
curve in Part B of the figure, a molecular weight of 44,000 :_2,000
is obtained. As will be seen below, this value actually represents

the subunit molecular weight.

49

.IoI Ommuospmu msalmmonuomlq can I¢0I upwusm .m

Scum OmeOpHm wumcoosamonmmonmuouhxoooumlouoxlm IIOI cassnau adumm

ch>on IIOI Ommuomncu owsonudo ”ouo3 poms acauuoum .mmuuospou

mIHIOmOQHOmIA can moumccmum mounu mo unmaoz unapomaos can pcu.m

unmm Eoum mocaa Onu mo Immumsvm unwed mo ponuwfi on» ma pocafiumump mmI

macaw on» coosumn mfinmcofiucaou on» msonm.m .Hom on» ca opaeuamuou

mo coHucuucmocoo on» can muaaanoe 0>Huuamu asp cowauon manmcoau

Imamu on» m3onm.m .mawouonmouuooao How moasuahuon an Omuuosuou mud
ImmobuOmtq mo usmwoz unasooaoa 0:» mo cofiumcfleumuma .MH ouamam

Hm. enough

3838 no: 6853 «5850: $3 zofiéazmuzoo gnu
om om o: om o 2 m w. I. m

50

HdOTS HAILVSHN

00.0

I0.0

No.0

no.0

:0.0

mod

00.0

No.0

wo.0

mod

 

I

 

 

0’

 

 

 

mo-

0.0-

Iko-

0.0-

m.o..

4.0-

no-

N.0n

H.0n

0.0

LLITIHON HAILV138 901

51

Subunit molecular weight determined by SDS acrylamide
gel electrophoresis. Electrophoresis in the presence of 1% SDS was
performed as described in methods for L-sorbose-l-P reductase and
the following standards: bovine serum albumin, MW 65,000 (45);
beef liver catalase, MW 60,000 (47); ovalbumin, MW 45,000 (48);
yeast alcohol dehydrogenase, MW 37,000 (49); and bovine erythrocyte
carbonic anhydrase, MW 29,000 (44). The data obtained by this
method are presented as a graph (Figure 14) of the MW on a log
scale versus the relative mobility (expressed as the distance
traveled from the origin divided by the length of the gel). Each
point was determined by averaging three values for each protein.
Using the standard deviation of the relative mobility of L-sorbose-
l-P reductase as a source of error, the molecular weight of the
subunit of L-sorbose-l-P reductase was determined to be 46,000

_+_ 1,000.

Gel filtration on Sephadex G-100. Gel filtration was
performed on a 1.5 x 57 cm column of Sephadex G-100. The sample
in a volume of 2 ml contained L-sorbose-l-P reductase and the
following standards: beef heart lactate dehydrogenase, MW 131,000
(50); yeast D-glucose-G-P dehydrogenase, MW 12,000 (51); bovine
serum albumin, MW 65,000 (45); yeast hexokinase, MW 50,000 (52);
and ovalbumin, MW 45,000 (48). The column was eluted with MES
buffer with EDTA omitted and fractions of 1.1 ml collected and
assayed for enzyme activities and protein. The elution pattern
from the column is presented in Figure 15. From this pattern the

standard curve (Figure 16) of the molecular weight (plotted on a

52

Figure 14. Determination of the subunit molecular
weight of L-sorbose-l-P reductase (SR) by SDS acrylamide
gel electrophoresis. Standards used were: bovine serum
albumin (BSA); beef liver catalase (CAT); ovalbumin (0A);
yeast alcohol dehydrogenase (ADH); and bovine erythro-
cyte carbonic anhydrase (CA).

53

 

 

DH

SR
0A

 

 

I)
‘1

Amcouamn voIv HmUHm3 M<ADUMA02

0.7

0.6

0.5
RELATIVE MOBILITY

0.1+

0.3

Figure 1.}:

S4

 

 

 

‘5 .§
p.\ ’P ‘0‘
100 y f E) I \ f l 0
\ ’ ‘
l ’ \5 ‘
f \ I o.
\ o No
" é I, \
' I \
l \‘ I \\
I, \ I 0‘
80 b A ‘ . \ u 0.8
I . \
I k ‘b
{ \ \
A ' \ \
\
:5 : \ b
I I |
a I \ I x
E: 60 f Y } . 0.6
E : I
< I o’ A ‘:
I l \ |
I I \ 1
E: I f \ i
E ’ " ‘ ' '.
ho - f ’ \ 3,0 O.h
g! f f) A
E : " \
\
j I, ’ld ‘\
3’ , . “\
I . \
20 - A ,or *3 L 0.2
d'
O I l i _‘I ‘1 o . O
55 no h5 50 55

FRACTION NO. (1.1 m1)

Figure 15 - Elution profile of Sephadex G-100 column. Marker proteins
used were: lactate dehydrogenase (A); Dbglucose-6-P dehydrogenase (5);
hexokinase (I); bovine serum.a1bumdn (peakwé) and ovalbumin (peak.§)
as determined by protein absorbance at 220 nm (0). L-sorbose-l-P re-
ductase is shown by O

PROTEINzABSORBANCE AT 220 nm

55

 

 

 

 

2o
15 .
LDH
fl GPDH

10 T

9 d
A 8 q
2 7 .
£3 BSA
3 6 -
'1'
O
:1 5 ‘ W

L 0v

5 u a
g
5
<9 3 1
{‘3
E

2 q

1 s A

55 [*0 1+5 50 55

FRACTION NUMBER

Fi ure lé - Determination of molecular weight of L-sorbose-l-P reductase
(0; using data obtained from Figure 12. Abbreviations used are:

LDH : lactate dehydrogenase; GDPH : D-glucose-6-P dehydrogenase; BSA :
bovine serum albumin; HK : hexokinase; and 0V : ovalbumin.

56
log scale) versus the fraction number was obtained. Allowing that
the activity peak of the L-sorbose-l-P reductase may be off by one
half of a fraction either way, the molecular weight was determined

as 92,000 1 2,000.

Sedimentation equilibrium ultracentrifugation. Ultracentri-
fugation was performed as described in methods. Three concentrations
of L-sorbose-l-P reductase were used: 0.5 mg/ml, 1.0 mg/ml, and
2.0 mg/ml. After 24.5 hr only the 0.5 mg/ml sample appeared to
have reached equilibrium and all measurements were obtained from
the Rayleigh interferogram of this concentration. The interfero-
gram of this sample appears as the upper image in the photograph
appearing to the left in Figure 17. Measurements were made by
standard methods (37,38) and plotted (Figure 17) in a graph of
three plus the logarithm of the vertical displacement from the
meniscus versus the square of the distance from the center of
rotation. The slope of the resulting straight line was substituted

in the equation.

M = 2 R T (2'303) x slope from Figure 17,

(l - yo) ”2

where M2 is the z-average molecular weight, 5.18 the gas constant
(8.314 x 107 erg°K-1mol-1), $_is the absolute temperature of the
centrifuge run (288°). 0 is the density of the solvent at the

given temperature (0.999), m is the speed of rotation of the run

in radians per sec, and G is the partial specific volume of the
protein, which was assumed to be 0.733 g/ml3 (that for an "average"

protein). Most of the error in this method comes from the value

57

AV 801 + Q

.chuouou «0 houses 0:» scum consumav ago no chance onu was uses

noouaduav guofluuo> onu mo coauocsm wofi u cooauob adachHuuch osu wowsonm «.cw nauuwouowuouca scum
occasuno cusp mo uoam " m..Aomoaq ucucoov Ha\sauuoua we o.m can “Amuse“ umzoav Haxcaououa wa_o.H
"Remus“ bouncy Ha\ca0uoud we m.o "can; con: acoausuuccucoo .uouou uuz< cc mw a: m.:m nouns Odfi

us Asap Arm.mfiv can owsmauucoo scum msmuwoucmuouca nwwoahum mo undnquuonm “ < .uawaoa magnuoaoa
ocaaucuov ou ousuosvou muaaomonuomuq mo coauomsmauucooouuas suaubflaasuc coaucuccsacom a Hm.cummwm

 

 

ANEOV NM
mm :m mm
o; _ .
m.:
O.N. a v -
wood a and n «so: . . . . 0
mm. . . .

 

on.

 

 

 

58
used for the partial specific volume; an error of it in G will
result in a 3% error in the molecular weight (37). From the
equation the molecular weight of L-sorbose-l-P reductase was calcu-

lated to be 87,900 :_2,600.

Determination of Svedberg coefficient from sucrose density
gradient centrifugation. Centrifugation was performed as described
in methods using a 0.1-ml sample containing L-sorbose-l-P reductase
and two marker proteins: sheep hemoglobin, 4.13 S (53) and E. coli
alkaline phosphatase, 6.10 S (28). A profile of the protein peaks
obtained from the centrifugation is shown in Figure 18; the dotted
lines are extensions to determine the distance that the markers
and the reductase traveled from the origin. Since this distance
is inversely proportional to the Savalue, then the average calcu-
lated from each marker for L-sorbose-l-P reductase is 5.54 :_0.04

x 1013 sec.

Summary of molecular weight determinations. The results of
the various methods by which the molecular weight of L-sorbose-l-P
reductase was determined is summarized in Table 2. The molecular
weight of the native enzyme as determined by gel filtration and
ultracentrifugation correlated well with the molecular weight of
the enzyme calculated from the values of the subunit molecular
weight. The average molecular weight determined by all of the
methods is calculated to be 90,000 :_2,000. The average molecular
weight of the subunit molecular weight is calculated to be 45,000

i_1,000, indicating that the enzyme is a dimer in the native state.

59

.Aollllsv HE mom ou9c«6 mom powwowxo modz moaofin no c303» ma smcuosnou
mnHIomOQHOmnq .AOIIIIOV He Mom ouacafi mom oav< ca unsound“ an
omuommoe mm Ao.m may Hcmusn Humlnwue z o.H ca oumnmmozmaaconmou0chm
mo mammaoucmn no ousu an owcfifiumuwo Am oa.mv smoumnmmonm mcaaoxao
wwoo .m can “ADIIIIDV Es oav us moccQHOmno ha pmcHEumumu Am ma.vv
cHnonoamn mownm “ouoz poms muoxuoz .uouou uoxosn ocamcfism 0.0m 3m

0 2H Emu ooo.mm up nu: ma non oo us can oumz A80 m.> x so ~.Hv moose

.Hocsaumoummoums-m w~.o can AH.o may ummuan mm: 2 omo.o as mmouosm sow o»

wo.m no as >.v no: ucoapnuo .mmouonvmu mlaswmonuOmlq mo unmaoammmoo
oumnpo>m mo cowumcflaumuwo unmaooum huwmcwp smouosm .ma madman

 

 

 

 

 

 

#3 U‘ o
'1' O. 3 8
o o (5 d
- 'l I
INHIGVUD JO HOLLOEI
_L
)-
JNHIGVHD JO c101. A
I 4r I
<2 u: 0. “3 <2
01 .—I F. o O
(o—o)
‘I F I I i
:1” M (\J H O

(0 c1)

 

3-5

5.0

2.5

0.5

DISTANCE TRAVELED ( cm)

Fi ure 18

61

Table 2. Summary of molecular weight determination of L-sorbose-l-P

 

 

 

reductase
Sedimentation

Method of Molecular weight coefficient
determination Subunit Native (Svedbergs)
"Native" acrylamide 44,000 : 2,000 88,000 :_4,000* ---
gel electrophoresis
SDS acrylamide gel 46,000 :_l,000 92,000 :_2,000* -—-
electrophoresis
Gel filtration on --- 92,000 :_2,000 ---
Sephadex G-100
Sedimentation equili- --- 87,900 :_2,600 —--
brium centrifugation
Sucrose density gra- —-- —-- 5.54 i 0.04

dient centrifugation

 

*
Calculated by assuming two subunits per native enzyme.

62

Catalyticlproperties

 

Effect of pH. The effect of pH on the relative activity

 

of L-sorbose-l-P reductase in MES, maleate-NaOH, and citrate-
phosphate buffers is shown in Figure 19. The pH optimum in all

three buffers is 6.2, but the velocity is greatest in MES.

Metal ion requirements. Monovalent metal cations such

 

as Na+, K+, MHZ, and Li+ have an inhibitory effect on the activity
of L-sorbose-l-P reductase; in concentrations above 10 mM, the
velocity is inhibited approximately 50%. In the presence of EDTA
all L-sorbose-l-P reductase activity is abolished, indicating the
divalent metal cations are necessary for the enzyme's activity. In
order to identify the divalent metal ion requirements, EDTA-treated
L-sorbose-l-P reductase [enzyme which had been purified in 0.025 M
MES buffer (pH 6.15) containing 0.2% mercaptoethanol and 2 mM

EDTA] was passed over a Biogel P-6 column equilibrated with 0.025
M MES buffer (pH 6.15) containing 0.2% mercaptoethanol. Fractions
containing protein were combined and assayed; they exhibited no
L-sorbose-l-P reductase activity in the absence of divalent metal
cation.

Assay mixtures (with various divalent metal cations) were
prepared containing all components of the standard mixture but one
and incubated for 5 min at 30°. The reaction was initiated by
adding the missing component to the mixture. The relative activity
of L—sorbose-l-P reductase in the presence of different metal ions

is shown in Table 3; the four columns represent the effect of

 

 

 

 

100
80~
>4
f: 60-
E.
8
<3
E
E-I
3 ho
k]
or.
O D/J/KD
20 “/0
D
O I l I I I J l L
5.2 5.1; 5.6 5.8 6.0 6.2 6.1+ 6.6 6.8 7.0

Figure _12 - Relative activity of L-sorbose-l-P reductase as a function
of pH and buffer. Relative rates in 0.067 M MES buffer (0) represent

the average of four determinations at each pH. The pH was determined

for each complete assay mixture after the rates were determined: The

relative rates for 0.067 Mmaleate-NaOH (D) and 0.067 M citrate-phos-

phate (0) represent one determination at each pH value.

Table

3. The effects of various divalent metal cations on the
activity of EDTA-treated L-sorbose-l-P reductase.
assay conditions and methodology are described in the

 

 

 

 

 

text.

Metal Activity expressed asgpercent activity with MnC12
added Component of assay_mixture used to initiate rxn
(1.0 mM) D-fru-l-Pf Metal ion NADH Enzyme
BaCl2 3 2 3 8
CaCl2 30 38 23 41
CoCl2 225 ++ 294 if
FeSO4 94 95 86 83
MnCl2 100a 100b 100C 100d
MgCl2 73 4 3 77
NiCl2 63 54 55 80
ZnSO4 4O 37 35 36

1-

D-fructose-l-P
++

Rates were nonlinear

a0.511 nmoles NADH oxidized per min.

b

0.445 nmoles NADH oxidized per min.

C0.445 nmoles NADH oxidized per min.

d

0.507 nmoles NADH oxidized per min.

65
initiating the reaction with various components of the assay mixture.
+2 ,
The addition of Mn reactivated the enzyme to the same rates
obtained before the EDTA-treatment. Ferrous ion produced rates
, , +2 , +2 +2 +2
closest to those obtained with Mn . Neither Ba , Ca , nor Zn
produced rates greater than one half those obtained in the presence
+2 .+2 . .
of the same amount of Mn . Ni produced intermediate rates of
. . . +2 . . .
activation (i.e., between 50 and 100%). Mg produced Significant
activation only in those cases where D-fructose-l—P and the enzyme
. . . . +2
were not present together in the preincubation mixture. Co was
a better activator (i.e., gave greater initial velocities) of
+2 . .
reductase than was Mn . However, the reaction rates were nonlinear
+
when either the enzyme or Co 2 were used to initiate the reaction.
Attempts at determining the optimal concentration of divalent
metal cation necessary for optimal activation of the reductase pro-
. +2 +2
duced variable results (Table 4). Ba and Ca were unable to
. . +2 .
activate the enzyme at concentrations below 0.1 mM. Zn acti-
vated the reductase optimally at a concentration of 0.01 mM and the
. . . . +2 .
actiVity was comparable to that obtained with Mn at this concen-
tration. All other divalent metals produced optimal activity at
. + . . .
a concentration of 0.1 mM except Mg 2 (which activated maximally

at 10 mM).

Substrate specificity. Preliminary investigations of

 

the substrate specificity of less purified preparations of L-
sorbose-l-P reductase revealed that D-fructose-l-P would serve as

a substrate for the enzyme nearly as well as L—sorbose-l-P.

66

Table 4. Effects of varying concentrations of divalent metal cations
on the activity of L-sorbose—l-P reductase. The reaction
conditions were the same as in Table 3 except that the
reaction was initiated by the addition of enzyme to the
assay mixture.

 

Activity (nmoles NADH oxidized per min per ml)
Concentration (mM)

 

 

 

Metal ion 10 1.0 0.1 0.01

8ac12 0.074 0.018 0.000 0.000
Ca012 0.160 0.064 0.000 0.000
MnC12 0.288 0.322 0.346 0.346
Mgc12 0.105 0.090 0.090 0.035
Nic12 0.136 0.234 0.293 0.254
Peso4 ppt* 0.244 0.351 0.293
Znso4 0 0.156 0.302 0.351

 

*

Formed a precipitate at this concentration.

67

A graph of reaction velocity versus amount of purified
enzyme added to the assay revealed that at concentrations above
1.2 ug (0.008 mg/ml in the assay mixture), the reaction velocity
did not increase at the same rate as below this concentration
(Figure 20). A double reciprocal plot (Figure 21) shows that at
higher concentrations of enzyme the Km for D—fructose-l-P and
L-sorbose-l-P increases from 0.38 mM to 1.2 mM and from 0.65 mM
to 1.73 mM, respectively. The Vmax for D-fructose-l-P and L-
sorbose-l-P increases from 0.61 nmole/min to 2.04 nmole/min and
from 1.83 nmole/min to 6.08 nmole/min, respectively. Since all
assays and other kinetic determinations in this dissertation were
done at enzyme concentrations below 1.2 ug, this phenomenon was
not studied further.

A study of the Km and Vmax determined for L-sorbose-l—P
and D-fructose-l-P from Figure 21 reveals that although the Km
of the reductase for L-sorbose-l-P is 1.7 times greater than that
for D-fructose-l-P, the Vmax of the enzyme is three times greater
with L-sorbose-l-P than with D-fructose-l-P. The specificity of
the enzyme was limited to these two ketohexose l-phosphates;
L-sorbose, D-fructose, D-fructose-6-P, and D-fructose-l,6-P2
would not serve as substrates for the enzyme at concentrations
of up to 10 mM.

Determination of the Km and Vmax of the enzyme for both
NADH and NADPH (Figure 22) indicated that the values were nearly
identical for both and were comparable to the levels of NADH

needed to elute the reductase from an NADP-sepharose column (20 uM).

68

.606: was doom coHuuoHMfiuao oasaouwounaouno zuacqmmo uuufim cnu Boom coauouuooua < .coauouou
may no huwooao> «so was cofiuouucoocoo ououusvou muHaouOAHOwIA cmosuon afinmcoHusHom I mM.mummHm

 

17m Tm m; m; m; mo v.0 no
in 1 a fl - i I 0.0
“‘IL
““|“““““““ o
. A. O.H
\\#
O s \\\\\

\\\\\ .. O.N
\\\\\\ \mllm
. \\\ o m
\\ .I- I
\\ a o
\\\ 8 N

\\\ i 0.“ d

\\\ 3 M
\\\ J m
0\ m 3
mm.

4 Of:

1. Com

6 on muaumuonuouuq_ze.n.n % 0.0

ADV muguomouosumun 23.n.n

 

 

 

69

.owms oum3 samuoum me mom mamas mom.o mo huw>fiuom camaowmm o sua3
mmmuospmu mnauomonuOmIA mo cofiumummoum o no Amn<z suw3 muauomouosumuo

mo COwuosowu wnu an wousmmwa may Adxuv mafia: mloa x mo.a Ho “0.0v
muflcs muoa x mo.o nocufim .mexucm mo macaumuucoocou ozu moans huaoonb
cofiuowmu on» was cowumuucmocoo mIHIwmonuomIA no mnalmmouosumuo cmoBDmn

mfismcowumawu on» mcfl3o:m uon Hmooumfioou cannon .HN ousmam

70

.d'

flM.ouoth

$.23 EA -mmoxmaoamsu:

N H

d *

 

(310mu/uim) A/I

 

a 0.: cHB\oHoac wo.m
cfia\ofioac n®.~
:«a\osoac do.m
owe a can .

ro.m \H Go

 

xoa>

:8 2.4
as 6.0
28 om;
:8 and

Acv muauomOAuomuq
Asv mufiuomOQuomnq
Adv muauomouosumnn
on muauomouosumua
oucuumosm

 

71

.cwououm as sea mafia: mm.a mo

>ua>auom oamfioomm m cues coduoummmum 6 mo Amadz nuwz mIHImmouoduMIa

mo cofiuoopou mnu >n pmnsmmwfi mmv >ua>auom wmouodpmu mIHImmonuom

Ia mo was: MIOH x vhv.o was mIHImmouosumua z: m.m pwcamucoo humus

comm .xufiooao> coauoomu one cofiucuucoocoo mmadz no 3042 cmosuon
mflnmcoflumamu may mafizonm uon Hmooumflomu cannon .NN madman

72

MM ouswwm

$.25 “£563:

 

 

 

03.: OS OOH ow om 0+. ON 0 ON: 0.:- om:
. . . . . . . . e
o.m
I
. .0... u
a \mI
I.
o N
. m
0
I
16.6 P
o
1 O.
o w
582686 No.0 28 nmod A3 $3.42
0
mas—":08: 5.1.0 Zammod AOV madz
.235 mm... 33033

73

Physical and chemical properties

 

Amino acid composition. Determination of the amino

 

acid composition of L-sorbose-l-P reductase purified from a second
affinity step (99% pure by electrophoretic criteria) was performed
as described in methods. The results of this determination are
summarized in Table 5. L-Sorbose-l-P reductase is rather low in

the basic amino acids-~histidine, arginine, and lysine--and contains
relatively high amounts of aspartate-asparagine (Asx) and glutamate-
glutamine (Glx). Such a distribution of amino acids suggests that
L-sorbose-l-P reductase is an acidic enzyme at neutral pH and

has an acidic isoelectric point.

Isoelectric point. In order to determine the isoelectric

 

point of L—sorbose-l-P reductase, duplicate gels, composed as
described by Wrigley (53a), were subjected to 4 hr of focusing

at 300 V constant voltage. For these studies a preparation from

a first affinity step was used since purity was not being determined.
The activity peak focused at a pH of 5.0 in the gradient and was

the same place as the protein peak (Figure 23). This isoelectric
point is in agreement with the prediction that the L-sorbose-l-P
reductase is acidic because it contains few basic amino acid residues,
migrates towards the anode when subjected to electrophoresis at pH
6.2, and binds to DEAE-cellulose at pH 6.15. Early studies with
acid precipitation showed that the L-sorbose-l-P reductase precipi-
tated only at pH values below 5.0, which is also in agreement with

the other data.

74

Table 5. Amino acid analysis of L—sorbose-l-P reductase

——-—- _-
-.-—-

 

No. residues No. residues
per subunit per subunit
Amino acid of enzyme Amino acid of enzyme

Ala 26.2 Lys 11.0

Arg 7.4 Met 3.2

Asx 35.0 Phe 7.6

1

Cys 3.2 Pro 15.2

Glx 63.7 Ser 58.1

Gly 104.6 Thr 18.9

His 12.2 Trp2 26.2

Ile 9.8 Tyr 13.1

Leu 17.0 Val 13.1

 

lesteine determined as carboxymethyl derivative.

2Tryptophan determined spectrophotometrically as described
in text.

75

(o—o) 311319313 Hd

.cfiououa pow cocoons no «mo amuw>uuos ofihuco you command was: maow noaga scams an :
you A> oonv owuu~o> ucoumcoo us can whoa maow muoofiadsn .mmouospou muauomonuomua mo cowuoucocua
powmwuso mafiufluuua a mo o.m cu o.m ma Scum flow cowasfimuoo co co wcamsoOM cauuooHoOuH I mm ouowwh

Aauv mfizfi one

 

 

 

 

 

 

   

 

 

_ o m 1. m m s n m a o - o
. VIIIII w .. y w . a . .
"" ’
(pl!
I [I’ll u- H HI.
/I ".6
II J 0
ll H v.0
I/ m m
III \\\\III s} an S
III\ I a I 3
I In ‘ ‘ N N p
. s . ...l
0 ~
— |\P\ .d
. xx x.
— \\I\ a n
. x- _ m. .u
s _
. . .. _ - .6 >m
. A a.
a a . a s V.
o s - J S
. . Q. on
. 8 J \l
a . O B J
- I A: - 1 d a
- (0.. Us .I.
— a (9
. u 3
. a I.
. 3 A
a
. n a
u. I m N
._ w I.
x o m
.-
.. m .M
.8- (
O I:
. w
.
.

 

 

76

Characterization of the Reaction

 

Identification of the product of the reduction of D-

fructose-l-P. Reduction of D-fructose-l-P at the keto group can

 

produce either D-mannitol-l-P or D-glucitol-l-P. By analogy with
the reduction of L-sorbose—l-P by L-sorbose—l-P reductase to form
D-glucitol-G-P (1), D-mannitol-l-P would be the expected product of
the reduction of D-fructose-l-P by L-sorbose-l-P reductase. To
identify the product of the reduction a scaled up reaction mixture
was prepared containing 0.067 M MES buffer (pH 6.1), 0.2% mercapto-
ethanol and 10% glycerol (to stabilize the enzyme during the extended
incubation period), 2 mM D-fructose-l-P, 2 mM NADH, and 1 mM

MnCl2 in a total volume of 2 ml. The reaction was started by
adding 0.1 ml of partially purified L-sorbose-l-P reductase (0.75
total units of activity and 0.48 mg total protein). The reaction
was allowed to proceed for 3 hr at 30° at which time it was
terminated by making the mixture 2.5% (w/v) trichloroacetic acid.
The acidified solution was clarified by centrifugation for 20 min
at 48,000 x g, extracted three times with 1 volume of ethyl ether
(to remove the TCA), and the pH adjusted to 8.0 with 10% ammonium
hydroxide. The mixture was placed on a Dowex-l column (bicarbonate)
as described in part I of this dissertation (p. 12). The column
was eluted with 0.5 M KHCO3 so that all phosphorylated products
could be analyzed. The product isolated in this manner served as

a substrate for D-mannitol-l-P dehydrogenase but not for D-glucitol-

6-P dehydrogenase when added to a coupled assay containing phospho-

glucoisomerase and D-glucose-6-P dehydrogenase. The same assay

77
mixture with phosphoglucoisomerase omitted produced only half the
amount of reduced nicotinamide adenine dinucleotide cofactor.
Dephosphosphorylation of the product was performed as described
in methods and the dephosphorylated mixture was assayed for inorganic
phosphate after 30 min. There was 1 umole of inorganic phosphate
present for every umole of D-mannitol-l-P present in the initial
mixture prior to dephosphorylation (as determined by the D-mannitol-
l-P dehydrogenase assay) and there were no detectable levels of
product capable of being used as a substrate for the coupled D-
mannitol-l-P dehydrogenase assay. When the dephosphorylated product
was acetylated as described in methods and subjected to gas-liquid
chromatography there was one major peak (representing 99% of the
detectable peaks) with the same retention time (40.7 min) as the
hexaacetate derivative of authentic D-mannitol in a mixture of sugar
alcohol standards (Table 6). The results of these experiments
indicate that the product of the reduction of D-fructose-l-P by
L-sorbose-l-P reductase is D-mannitol-l-P as shown chemically,

enzymatically and chromatographically.

Equilibrium and stoichiometry of the reaction. The equilibrium
of the reduction reaction was determined under standard assay condi-
tions. The concentration of NADH present initially was determined
by measuring the absorbance at 340 nm, the amount of NADH oxidized
was determined from the decrease in absorbance at 340 nm when the
mixture had reached equilibrium (i.e., when there was no longer a
decrease in absorbance at 340 nm). Controls were run with no Mn+2

present. After equilibrium was attained the reaction was terminated

 

78

Table 6. Gas-liquid chromatography of the hexaaacetate derivative of
the dephosphorylated reduction product of D-fructose-l-P*

 

Dephosphorylated D-fructose-

 

 

Hexaacetate 1-P reductiog_product
standards RT (min) % peak height RT (min)
Rhamnitol 16.4

Fucitol 18.0

Arabitol 24.3

Xylitol 29.9

Mannitol 41.1 99 40.7
Galactitol 44.1

Glucitol 45.9

Inositol 50.5

 

*Product was prepared as described in methods. Retention
times (RT) are compared to hexaacetate derivatives of sugar alcohol
standards. Chromatography was performed on a 0.002 x 1.83 m column
of 0.2% polyethyleneglycol adipate, 0.2% polyethyleneglycol suc-
cinate, and 0.4% XF-115 silicone oil with a Perkin Elmer 900 gas
chromatograph programmed to run from 130° to 180° at a rate of 1°
per min.

79
by adding trichloroacetic acid so that the final concentration in
the mixture was 0.2 M and the acidified solution was placed on ice
for 30 min at which time the acid was neutralized by adding an
appropriate amount of NaOH. Samples were removed and D-fructose-
l-P and D-mannitol-l-P were determined by coupled assays as
described in methods. The results of these determinations indi~
cated that the D-fructose-l-P levels at equilibrium were too low
to permit a very accurate determination: however, since 0.96 nmoles
of D-mannitol-l-P was produced for every umole of D-fructose-l-P
present initially (as determined from the control mixture) the
difference between these values (D-fructose-l-P minus D-mannitol-
l-P) was taken to represent the lvel of D-fructose-l-P present at
equilibrium. By using these values and the initial and final
determinations of NADH concentration, it was determined that the
ratios of NADH oxidized to D-fructose-l-P initially present to
D-mannitol-l-P formed was l.2:1.0:0.96. Any errors in these values
may be the results of the manipulations performed on the sample
after equilibrium was reached. The equilibrium constant (Keq) for

the reaction at pH 6.15 was determined from the equation:

K _ [D-mannitol-l-P] [mfi
eq [D—fructose-l-P] [NADH] [8+]

 

where the bracketed values are the concentration of the various
species present at equilibrium. The Keq calculated in this manner
is 4.5 x 107 liters per mole.

The equilibrium for L-sorbose-l-P was determined in the same

manner as for D-fructose-l-P except that the amount of L-sorbose-l-P

80
present at equilibrium could not be determined. The amount of L-
sorbose-l-P added initially was determined as organically bound
phosphate (using alkaline phosphatase as described for dephosphoryla-
tion of the reduction product of D-fructose-l-P). The ratio of NADH
oxidized to L-sorbose-l-P initially present to the amount of D-
glucitol-6-P formed (as determined in a coupled assay with D-
glucitol-6-P dehydrogenase) was l.2:1.0:0.93. The equilibrium
constant at pH 6.15 was determined by substituting the appropriate

equilibrium concentrations into the equation

K = [D-glucitol-6-P] [NAD+]
eq [L-sorbose-l-P] [NADH] [8‘17

 

The Keq calculated in this manner is 2.6 x 107 .liters per
mole. The equilibrium constants for the reduction of D~fructose~l-P
and L-sorbose-l-P indicate that the reaction catalyzed by the
reductase under the conditions used lies approximately 95% in the

direction of the formation of the sugar alcohol phosphate products.

Discussion and Summary

 

L-Sorbose-l-P reductase was purified 4l-fold by DEAE-
cellulose chromatography and by affinity chromatography using
selective elution with combinations of cofactor and D-fructose-l-P.
Acrylamide gel electrophoresis indicated that the purified enzyme
contained no more than 1% contamination by other proteins.

An examination of the physical properties of the purified
enzyme revealed that the native enzyme has a molecular weight of
90,000 (by ultracentrifugation and gel filtration) and consists of

two subunits of identical molecular weight, i.e., 45,000 (by

81
electrophoresis on native and SDS acrylamide gels). Although native
gels are normally used to determine the molecular weight of native
enzymes, the conditions under which my determination was performed
(pH 8.8 in Tris-glycine buffer) appeared to cause the reductase to
dissociate into its subunits. This dissociation allowed the conclu-
sion that the subunits were not only molecular weight—isomers but
also charge-isomers (35), thereby providing further evidence that
the subunits are identical monomers. The value of the Svedberg
coefficient (determined by sucrose gradient centrifugation) was
5.54 x 1013 sec and is similar to those values obtained for other
Proteins possessing a molecular weight of around 90,000 (54).

Several sets of data indicate that the L-sorbose-l-P reductase
is an acidic protein: (i) the amino acid analysis reveals that the
enzyme has relatively low amounts of basic amino acids (Table 5);
(ii) isoelectric focusing shows that the isoelectric point is 5.0
(Figure 23); (iii) the enzyme precipitates in acidic solutions
only when the pH is lowered to below 5.0; (iv) the reductase migrates
towards the anode during electrophoresis at pH 6.2 (Figure 12); and
(v) the enzyme will bind to DEAR-cellulose at pH 6.15.

The pH optimum is 6.2 in the three buffers tested, and the
highest activity was obtained in MES buffer. The enzyme requires a
divalent metal cation for catalytic activity. Treating L-sorbose-
l-P reductase with EDTA abolishes the ability of the enzyme to
reduce L-sorbose-l-P and D-fructose-l-P, but the activity can be
restored by the addition of an appropriate divalent metal ion.

EDTA-treated enzyme will bind to NADP-sepharose but divalent metal

82

ion (Mn+2) is required for the enzyme to bind more tightly in the
presence of D-fructose-l-P. Manganese appears to be the metal
normally associated with the enzyme since the presence of this metal
restores the enzyme to the same levels of activity possessed prior
to treatment with EDTA. The best reactivation of enzyme occurs when
the metal ion is preincubated in the presence of either D-fructose-
l-P without enzyme present, or enzyme without D-fructose-l-P
present (Table 3). In the instances in which the sugar phosphate
and enzyme are allowed to form complexes either before or at the
same time as the metal ions are added, the activation of the enzyme
is not as high. The aforementioned results and observations are
consistent with a model in which the metal ion binds to the enzyme
to facilitate the binding of the sugar phosphate in a complex with
the correct conformation to allow catalysis to occur. The exact
nature of the interactions involved remains to be determined but
several types of enzyme--metal ion--substrate interactions have been
discussed by Dwyer (55).

L—Sorbose-l-P reductase is capable of catalyzing the reduc-
tion of either L-sorbose-l-P or D—fructose-l-P with either NADH
or NADPH to form D-glucitol-6-P or D-mannitol-l-P, respectively.
The Michaelis constants for either ketohexose-l-P are similar
(Figure 21) but the maximal velocity for the reduction of L-sorbose-
l-P is three times that for the reduction of D-fructose-l-P. The
Michaelis constant and maximal velocity of the reaction are nearly
identical for either NADH or NADPH and agree with the concentra-

tion of NADH required to elute the enzyme from an NADP-sepharose

83

affinity column. The product of the reduction of D-fructose-l-P
was identified as D-mannitol-l-P on the basis of chromatographic
and enzymatic analysis. The equilibrium of the reduction of L-
sorbose-l-P and D-fructose-l-P lies far in the direction of the
formation of the sugar alcohol phosphate products (Keq = 2.6 x 107
and 4.5 x 107 liters/mole, respectively), and the stoichiometry of
the reaction shows that one molecule of ketohexose-l-P is reduced
by one molecule of NAD(P)H for every molecule of hexitol phosphate
and NAD(P)+ formed.

Studies of the substrate specificity reveal that L—sorbose,
D-fructose, D-fructose-6-P, and D-fructose-l,6-P2 will not serve as
substrates for the enzyme. This specificity plus the fact that the

only two reactions that I was able to demonstrate to be catalyzed

by L-sorbose-l-P reductase were

THZOP THZOP THZOP THZOP
|=O HO-T-H T=O HO-T-H
HO-C-H HO-C-H HO-C-H HO-C-H
| + l l + l
H-T-OH H-T-OH and H-T-OH H-T-OH
HO-f—H HO-T-H H-f-OH H-T-OH
CHZOH CHZOH CflzOH CHZOH
L-sorbose-l-P D-glucitol-6-P D-fructose-l-P D-mannitol-l-P

suggests that the specificity of the enzyme is for those ketohexose
l-phosphates which possess structures which are the same as L-
sorbose-l-P at C-1 through C-4. The reduction involves the forma-

tion of a hydroxyl group at the second carbon atom to make an

84
erythro structure for the hexitol phosphate at the carbon atoms which
correspond to C2 and C3 of the ketohexose-l-P.

L-Sorbose-l-P reductase from A. aerogenes is the first keto—
hexose-l-P reductase to be reported for any organism. Since the
enzyme is induced in this organism only by growth on L-sorbose,
its main role is in the metabolism of L-sorbose (1). Although the
enzyme is capable of reducing D-fructose-l-P to D-mannitol-l-P, it
normally serves no function in the metabolism of D-fructose: the
enzyme is not present when the wild-type strain is grown on D-
fructose and D-fructose has been shown to be metabolized almost
exclusively via two alternative pathways in the wild-type organism
(76). However, the presence of D-mannitol-l-P dehydrogenase in
cells uninduced by growth on D-mannitol coupled with the ability
of L-sorbose-l-P reductase to convert D-fructose-l-P to D-mannitol-
1-P suggested that under certain conditions, D-fructose might be
metabolized in A. aerogenes by a novel pathway in which L-sorbose-
l-P would play a key role. The evolution of such a pathway for
the metabolism of D-fructose in A. aerogenes will be demonstrated

in Part III of this dissertation.

PART III

EVOLUTION OF A NOVEL PATHWAY FOR THE METABOLISM

OF D-FRUCTOSE BY AEROBACTER AEROGENES

85

86

Introduction

 

There are several routes by which D-fructose is known to be
metabolized in various organisms. The classical studies were done
with yeast, for which it was shown that D-fructose is phosphorylated
with ATP by hexokinase to yield D—fructose-6-P (56).

In mammals the main site of D-fructose metabolism is the
liver, wherein: (i) D-fructose is phosphorylated to D-fructose-l-P
by an ATP-dependent l-fructokinase; (ii) D-fructose-l-P is cleaved
by an aldolase into dihydroxyacetone phosphate and D-glyceraldehyde;
and (iii) D-glyceraldehyde is phosphorylated to D-glyceraldehyde-
3-P by an ATP—dependent triokinase (57).

In certain pseudomonads, such as Gluconobacter cerinus, carbo-
hydrates are metabolized almost exclusively via the hexose mono-
phosphate pathway (58). It has been established that D-fructose
also plays a key role in the regeneration of the NADP+ that is
reduced during the oxidation of D-glucose-6-P when G. cerinus is
grown on D-fructose (59,60). D-Fructose is oxidized by a particu-
late, respiratory-chain-linked fructose 5-dehydrogenase to yield
S-keto-Defructose (61), which is then reduced by an NADPH-dependent
S-keto-D-fructose reductase (14), thereby regenerating both D—
fructose and NADP+.

Another major pathway for the metabolism of exogenously
supplied D-fructose in microorganisms was first elucidated in A.
aerogenes by Anderson and co-workers (62-65). It was shown that:
(i) D-fructose is simultaneously transported and phosphorylated

by an inducible P-enolpyruvate—dependent phosphotransferase system;

 

87
(ii) the phosphorylation is at carbon atom l of D-fructose; and (iii)
D-fructose-l-P is then phosphorylated with ATP by a specific,

inducible kinase to form D-fructose-l,6-P This pathway was

2.
subsequently found in Escherichia coli (66-68), Clostridium
pasteuranium (69-71), C. roseum (71), C. rubrum (71), C. butyricum
(71), C. thermocellum (72), Arthrobacter pyridinolis (73), and
Rhodopseudomonas capsulata and other species of phototrophic bacteria
(74). One of the enzymes, D-fructose-l-P kinase, has been reported
to occur in Bacteroides symbiosis (75) and in Clostridium
formicoaceticum (71).

Endogenous D-fructose is utilized in A. aerogenes by a dif-
ferent route. Sucrose is transported into the cell as the free
sugar, where a hydrolase cleaves the disaccharide into its component
monosaccharides, D-fructose and D-glucose. This intracellular
D-fructose undergoes an ATP-dependent phosphorylation at carbon
atom 6 by means of a sucrose-induced, specific D-fructokinase (76).
D—Fructokinase activity is also present in very low amounts in
cells grown on carbohydrates other than sucrose, but even these
low amounts are sufficient for mutants missing D-fructose-l-P
kinase to grow slowly on D-fructose.

This section of my dissertation will describe the evolution
of a novel pathway for the metabolism of D—fructose. Previous work
on the evolution of new pathways in microorganisms has been con-
cerned almost exclusively with the metabolism of "xenobiotics",
i.e., compounds normally not found in the natural environment of

the cell (77,78). Acquisition of the ability to grow on xenobiotics

88
has been found to occur by mutation in one of three ways: (1) con-
stitutive production of enzymes for which the novel compounds
normally serve as fortuitous substrates, but which are normally
induced only by natural substrates; (ii) mutation in a structural
gene which allows the new enzyme to also use the novel compound as
a substrate; or (iii) formation of a new enzyme which has no known
function other than to facilitate the metabolism of the novel
compound. The discovery (part II of this dissertation) that L-
sorbose-l-P reductase would also reduce D-fructose-l-P to D-
mannitol-l-P suggested the possibility of genetically manipulating
A. aerogenes to allow it to utilize D-fructose for growth by a
totally novel pathway. This would be the first report of the
evolution of a new metabolic pathway for a compound which is
commonly found in the milieu of the cell.

It was known that D-mannitol-l-P dehydrogenase is present
in relatively high levels even in cells grown on substrates other
than D-mannitol. By using a double mutant (strain 14) missing
both D-fructose-l-P kinase and D-fructokinase, and which is conse-
quently unable to grow on D-fructose, it was possible to select for
a mutant which was constitutive for L-sorbose-l-P reductase. This
mutant is able to utilize D-fructose as a sole source of carbon
and energy by conscripting enzymes from three normally separate
metabolic pathways. D-Fructose is transported and converted to
D-fructose-l-P by the PEP:D-fructose phosphotransferase system of
the normal D-fructose pathway. The D-fructose-l-P is then reduced

at the keto group by L-sorbose-l-P reductase. The product,

89
D-mannitol-l-P, is then oxidized at carbon atom 5 by D-mannitol—l-P
dehydrogenase from the D-mannitol pathway. The bacterium is then
able to metabolize the resulting D-fructose-G-P via the same path-

ways as in the wild-type strain.

Materials and Methods

 

Isolation of Mutants--General Procedures

 

The phenotype and genotype of the parental and mutant
strains used or developed in this study are shown in Table 7.
Parental strains were inoculated into culture tubes containing 7
ml of nutrient broth and grown overnight at 30°. The fully grown
cultures were transferred to a sterile 12.5-m1 centrifuge tube and
the cells were washed by alternatively centrifuging for 5 min at
12,000 x q_and suspending the pelleted cells in 7 m1 of sterile
mineral broth. After three such washes a 2-ml sample of suspended
cells was transferred to a sterile 5-ml test tube and 30 ul of
ethyl methane sulfonate was added. The tube was incubated with
shaking for 2 hr at 30°. The mutagenized sample was transferred
to a sterile 12.5-ml centrifuge tube, 7 ml of sterile mineral broth
was added, and the tube was centrifuged for 5 min at 12,000 x g,
The cells were washed as described above and the final suspension
in mineral broth was transferred to a sterile culture tube contain-
ing 0.2 m1 of a 17.5% sterile carbohydrate solution. After the
cells were allowed to grow for four generations (one generation
was considered to be one doubling in absorbance at 520 nm), a l-ml

portion was transferred to a fresh tube of 0.5% carbohydrate in

9O

 

 

 

 

 

 

 

 

 

 

 

Table 7. Genotype and phenotype of mutants used in this study
. i if
Strain Parent Genotype Phenotype Reference
+ + +
PRL-R3 --- Wild-type Fru+, Mtl , Sor ,
Scs
s + +
DD31 PRL-R3 fru-31 Fru , Mtl , Sor , 76
Scs (FlPK-negative)
- + +
14 DD31 fru—3l, scs-14 Fru , Mtl , Sor , 76
Scs (F1PK-neg.;
FK-neg.)
+ + + , .
SFl 14 fru-3l, scs-l4, Fru+, Mtl , Sor , this dis-
sor-l Scs (FlPK-neg.; sertation
FK—neg.; SR-
constructive)
- + ..
SFl-SS SFl fru-3l, scs-l4, Fru , Mtl , Sor , this dis-
sor-l, sor-S ScsS (F1PK-neg.; sertation
FK-neg.; SR—neg.)
+ + + , .
SFl-SSR SFl-SS Revertant to Fru+, Mtl , Sor , this dis-
SFl Scs (same as SFl) sertation
- - +
SFl-MA2 SFl fru-31, scs-l4, Fru , Mtl , Sor , this dis-
sor-l, mtl-2 ScsS (FlPK-neg.; sertation
FK-neg.; SR-const.;
M1PDH-neg.)
+ + + , ,
SFl-MAZR SFl-MAZ Revertant to Fru+, Mtl , Sor , this dis-
SFl Scs (same as SFl) sertation
+ + + . .
SFl-MCl SFl fru-3l, scs-14, Fru , Mtl , Sor , this dis-
sor-l, mtl-l Scs (FlPK-neg.; sertation
FK—neg.; SR—const.;
MlPDH-Z to 3 times
higher)
+Fru = D-fructose; Mtl = D-mannitol; Sor = L-sorbose; Scs =
sucrose.
T?

+ = able to grow; - =

unable to grow; 5 =

able to grow

slowly; F1PK = D-fructose-l-P kinase; FK = D-fructokinase; SR =
L-sorbose-l-P reductase; MlPDH =

D-mannitol-l-P dehydrogenase.

91
mineral broth and allowed to grow for three generations. This
transfer was repeated once more, the culture was allowed to grow
fully, and was selected for the apprOpriate phenotype as described

in the Results section.

Analytical Procedures

 

Enzyme assays. All assays were performed in a total volume

 

of 0.150 ml and the rates monitored by change in absorbance at 340
nm on a Gilford multiple-sample absorbance-recording spectropho-
tometer thermostated at 30°. The assay for L-sorbose-l-P reductase
has been described in part II of this dissertation. The assay for
D-mannitol-l-P dehydrogenase contained 0.067 M Tris-HCl buffer

(pl-I 9.0) , 3.3 mM D-mannitol-l-P, 0.33 mM NAD+ (adjusted to pH 7.0
with NaOH), 0.33 mM NADP+ (adjusted to pH 7.0 with NaOH), and non-
limiting amounts of phosphoglucoisomerase and D-glucose-G-P dehy-
drogenase. Controls were run with D-mannitol-l-P omitted. Rates
were constant for the initial 2 min of the assay. The assay for
D-fructose-l-P kinase contained 0.067 M glycylglycine buffer (pH
7.5), 3.3 mM D-fructose-l-P (sodium salt), 3.3 mM ATP, 6.6 mM
MgC12, 67 mM KCl, 0.33 mM NADH, and nonlimiting amounts of rabbit
muscle aldolase and a-glycerophosphate dehydrogenase-triose phosphate
isomerase. Controls were run with ATP omitted. The assay for
D-fructokinase contained 0.067 M glycylglycine buffer (pH 7.5),

+
3.3 mM D-fructose, 3.3 mM ATP, 6.6 mM MgCl , 0.33 mM NADP (adjusted

2

to pH 7.0 with NaOH), and nonlimiting amounts of phosphoglucoisomerase

and D—glucose-S-P dehydrogenase. Controls were run with ATP omitted.

92

Determination of intracellular metabolite levels in mutant
strains. Each strain was grown overnight in 100 ml of nutrient
broth at 30°. Cells were washed by alternately centrifuging and
suspending in 14 ml of sterile mineral broth a total of three times.
Cells were suspended finally in 50 m1 of sterile mineral broth and
this was added to 75 ml of mineral broth and D-fructose so that
there was a final concentration of 0.5% D-fructose in 100 ml of
mineral broth. Cells were incubated for 3 hr at 30° after which
they were washed as before with mineral broth. The pellet contain-
ing the cells from the final centrifugation was weighed and then
suspended in 1 ml of mineral broth. D-Fructose (80 nmoles) was
added to this suspension and the cells incubated for 10 min at
room temperature. The incubation was terminated by the addition
of 0.1 ml of 70% perchloric acid and mixing periodically for 5
min at room temperature. The tubes were placed on ice for 20 min
and the solution then neutralized with approximately 0.5 ml of
2.5 M potassium bicarbonate. Precipitated protein and cellular
debris were removed by centrifugation at 12,000 x g for 10 min.
The supernatant was removed and the pH adjusted with KHCO3 to
neutrality if needed. A control to determine the extent of leakage
of metabolites from the cells during incubation and subsequent
manipulations was performed. A sample of cells incubated for three
hr in the presence of 0.5% D-fructose was separated into two por-
tions and incubated with 80 nmoles of D-fructose, as described
above, for 5 min at room temperature. The two portions were centri-

fuged at 12,000 x g_at 22°. The samples were then divided as

 

93
follows: one sample was divided into supernatant and pellet; the
other sample was treated in toto. The three resulting fractions
were treated with perchloric acid and potassium bicarbonate as
described above. After assaying for the metabolite levels as
described below, there were no detectable metabolites in the super-
natant fraction and the metabolites present in the cells or the

cells plus the supernatant were the same.

IMViC tests. The IMViC tests were performed according to

 

standard procedures (80,81) and modified as described below.

For the indole test, cultures of E. coli and A. aerogenes
strains were grown at 37 and 30 degrees, respectively, in culture
tubes containing 7 ml of 1% tryptone broth. Inocula were allowed
to grow for 24 hr without shaking at which time 0.2-0.3 ml of
reagent were added. The reagent was composed of 5 gm 2:
dimethylaminobenzaldehyde in 75% 27amyl alcohol-25% concentrated
HCl. The test was considered to be positive if a deep purple color
formed at the interface between the medium and the air.

The methyl red and nges—Proskauer tests were performed on
cultures grown on mineral broth supplemented with 0.5% D-glucose.
After 24 and 48 hr l-ml samples were removed and tested for the
accumulation of acetoin (3-hydroxy-2-butanone) in the medium.

The Voges-Proskauer test for acetoin consists of adding to the 1-m1
sample, 0.6 ml of 5% a-naphthol in absolute ethanol followed by

0.2 m1 of 0.3% creatine in 40% KOH. After mixing a red color
developing within a few minutes constitutes a positive test. The

same cultures, allowed to incubate without shaking for 5 days, were

94
used for the methyl red test for acid production. To the cultures
5 drops of 0.3% methyl red in 50% ethanol were added. A red color

persisting for several hours was considered a positive test.

Results

Linear Growth of Strain 14 on D—Fructose after
Induction with D-Mannitol and L-Sorbose

 

 

The feasibility of genetically producing an organism with
the proposed pathway for D-fructose metabolism (see Introduction
to part III of this dissertation) was examined in a mutant of A.
aerogenes in which both the normal pathways for the metabolism of
D-fructose were genetically blocked. The levels of L-sorbose-l—P
reductase and/or D-mannitol-l-P dehydrogenase were temporarily
increased in this mutant (strain 14) by growth on L-sorbose, D-
mannitol, or a mixture of both. Figure 24A shows that merely
increasing the level of D-mannitol-l-P dehydrogenase by growth on
D-mannitol alone is insufficient to allow strain 14 to grow on
0.3% D—fructose in mineral broth. When strain 14 is grown on L-
sorbose, the L—sorbose-l-P reductase which is induced is sufficient
to allow limited linear (not logarithmic) growth on D-fructose
(Figure 248). Cells in which both the level of L-sorbose-l-P
reductase and D-mannitol-l—P dehydrogenase have been temporarily
increased by growth on L-sorbose and D-mannitol (Figure 24C) grow
7.5 times as fast as those with elevated levels of L-sorbose-leP
reductase alone. Such results suggest in these studies that the
level of D-mannitol-l-P dehydrogenase is the limiting factor in

the metabolism of D-fructose by the new pathway. The linear growth

a“.

I
R
I

95

.omOAuOuuA *ma.o moan gouaccuaun $ma.0 :0 song

nosocmoum " o Homonuomua fin.o so coauusvcwoum u m "gouaccsaun $n.0 co coauosocwoum ” <
.uswsu room we fin.o co oscwauouoo was nusouu .omoouOunA can gouaccuann so nusouw mo
cowuosocwoua nouns AOV oaouoauunn can ADV owoosHmun so :a campus no nusouu ulmw.mmmmwm.

 

 

A53 Baa
oom ems ooH Om 0 One ooH cm 0

V
H
m
m
D
:m.0 3
o m
C4
3
O
m

213.0

0
1m.o
mu 41

 

 

 

 

 

 

96
rates evinced in Figure 24 are those expected in the case of a
constant amount of enzyme being present in the whole inoculum of
cells. The logarithmic growth curves shown by the cells grown on
D-glucose are those expected when each cell synthesizes new enzyme
as it grows.

Selection of Mutants of L-Sorbose-l-P Reductase
and D-Mannitol-l-P Dehydrogenase

 

 

To develop and elucidate the new pathway of D-fructose
metabolism, mutants with altered levels and/or regulation of L-
sorbose-l-P reductase and D-mannitol-l-P dehydrogenase were selected
as described below. It was unnecessary in this study to select
mutants in the PEP:D-fructose l-phosphotransferase system since
its role in the transport and phosphorylation of D-fructose in A.
aerogenes had already been established (63,65). The genealogy and
description of the mutants used in this study was summarized in

Table 7.

Mutants constitutive for L-sorbose-l-P reductase. Strain l4,

 

a double mutant missing both D-fructose-l-P kinase and D-fructokinase
(Table 7), was mutagenized as described in methods and grown on 0.5%
L—sorbose as the carbohydrate. One milliliter of the fully grown
culture was transferred to a culture tube containing 7 ml of mineral
broth with 16.7 mM (0.3%) L-sorbose and allowed to grow for three
generations. A l-ml sample was inoculated into a culture tube con-
taining 7 ml of mineral broth with 16.7 mM (0.3%) D-glucose. After
growth proceeded for 3 to 4 generations, a 0.2-ml sample was trans-

ferred to a culture tube containing 16.7 mM L-sorbose in mineral

97
broth and was incubated at 30° for 3 to 4 generations. Sterile D-
glucose was added to a final concentration of 27 mM (0.5%) and the
cells were allowed to grow at 30° overnight. This alternate transfer
between 16.7 mM L-sorbose and 27 mM D-glucose was considered to be
one cycle of enrichment. After 31 cycles, samples were removed,
serially diluted, and plated on MacConkey agar supplemented with
1.0% D-fructose. Mutants were selected as red colonies appearing
after incubation of the plates at 30° for 24 to 36 hr. Individual
colonies were selected and transferred to a culture tube containing
7 m1 of nutrient broth. After cultures were fully grown, crude
extracts were prepared as described in part II, methods, and
assayed for L-sorbose-l-P reductase and D-fructose-l-P kinase as
described in methods. Mutants obtained in this manner either
possessed very low constitutive levels of the reductase (0.007
to 0.010 units per mg protein in crude extracts) or were revertants
possessing D-fructose-l-P kinase. Unexpectedly, the mutants with
low constitutive levels of reductase were found not to grow on
mineral broth supplemented with 0.3% D-fructose. At this point I
modified both the enrichment and selection procedures. Mutagenized
cells of strain 14 were cycled between 0.1 mM (0.018%) L-sorbose
and the same concentration (16.7 mM) of D-glucose. By recycling
between 0.1 mM L-sorbose and 16.7 mM D-glucose I reasoned that
there would be a doubly selective advantage. As in the former
enrichment procedure, cycling between L-sorbose and D-glucose would
enrich for those mutants with a constitutive reductase because they

would not require the normal two hour lag period to induce the

"IT

98
reductase and, hence, would start growing before the wild-type cells.
By lowering the L-sorbose concentration I thought that those mutants
with higher constitutive levels of the reductase would have a
selective advantage because they would grow at the normal induced
rate initially and thereby deplete the medium of the limited amount
of L-sorbose before mutants with lower levels of constitutive
reductase could be induced to the normal levels of the enzyme for
faster growth. This method would allow the enrichment of the higher .
constitutive mutants at a faster rate than the old method. The
selection procedure was now modified because of the necessity to
distinguish between those mutants able to metabolize D-fructose
sufficiently to appear as red colonies on MacConkey agar and those
mutants truly able to utilize D-fructose as a sole source of carbon
and energy. To select for the latter mutants, samples from various
cycles of enrichment were plated onto mineral agar supplemented with
1.0% D-fructose. With the modified procedure one would predict
that the number of mutants with increased levels of reductase,
appearing as red colonies on MacConkey agar supplemented with 1.0%
D-fructose, would increase with the number of cycles. At the same
time those mutants with the higher levels of constitutive reductase
(and according to my theory able to utilize D-fructose as a sole
source of carbon and energy) should also increase, but at a rate
greater than the increase in constitutive mutants in general. This
prediction was substantiated as shown in Figure 25. Note that the
total number of constitutive reductase mutants is initially greater

than those with constitutive levels high enough to permit growth on

99

 

   

 

 

 

100+ F100
80- .80
60- p60
40- )hn
c:
a:
[.4
20! -20 j
a.
m M
5‘: 55
>4 10‘ --10 m <
8 _8 U I-‘l
E l o a
0 6d '6 SEE
U H
0" Ah z
§ 4- .4 O m
(I)
g Eio
° 8
”3 (9::
av 2d '2 z N
z VH It:
5 go
8 1- -1 ”‘5
Q [a _I
E E z
3% 0.61 -006 go
0.44 -O.h cf
01
2:
()02'I -002
0.1‘ ‘ 5 t :— 9 —0.1
5 10 15 20 25 30

NUMBER OF TIMES CYCLED

Figure g2 - Ability of mutants of A. aerogenes to utilize D-fructose
as a function of enrichment for constitutivity of growth on L-sorbose.
Mutants were cycled on 0.1 mM L-sorbose and 16.7 mM D-glucose as de-
scribed in the text.

100
D-fructose alone. By the end of 30 cycles 90% of the constitutive
mutants are those which can utilize D-fructose as a sole source of
carbon and energy. Controls which were recycled on 0.1 mM D-
mannitol and 16.7 mM D-glucose showed no increase in the number of
mutants able to grow on D-fructose and those that didlat no time
exceeded 0.8%. SFl was selected as a mutant growing on mineral

agar supplemented with 1.0% D-fructose as was used in further char-

acterization studies.

Mutants negative for L-sorbose-l-P reductase. SFl was muta-
genized and grown on D-glucose as the carbohydrate as described in
methods. Mutants missing either D—mannitol-l-P dehydrogenase or
Enzyme I or HPr of the P-enolpyruvate-dependent phosphotransferase
system were removed by growth overnight on 0.5% D-mannitol in
mineral broth. Cells fully grown on D—mannitol were washed as
described in methods and suspended finally in 2 ml of mineral
broth. Samples of this suspension were added to a culture tube
containing mineral broth so that the initial absorbance at 520 nm
was 0.075 when determined in a 1.8-cm culture tube with a Model 6A
Coleman Jr. spectrophotometer. A sterile solution of 17.5%
L-sorbose was added to this culture and the cells grown at 30°
until the absorbance at 520 nm had reached 0.30, at which point
8 mg of penicillin G (benzylpenicillin) was added. The tube was
vortexed and the incubation continued for 4 hr, at which time the
absorbance at 520 nm was 0.075. Cells were washed by alternate
centrifugation and suspension in mineral broth three times. A

sterile 17.5% solution (0.2 ml) of D-mannitol was added to the

 

101
final mineral broth suspension and the cells allowed to grow at
30° overnight. The fully grown culture was serially diluted and
102 cells plated on MacConkey agar supplemented with 1.0% L-sorbose.
After incubation for 24 hr at 30°, 28 pale colonies were selected,
inoculated into culture tubes containing 7 ml of nutrient broth,
grown overnight at 30°, and streaked on MacConkey agar plates
containing 1.0% D-glucose, D-mannitol, L-sorbose, or D-fructose.
Eighteen of these mutants grew as pale colonies on L-sorbose and
D-fructose and as red colonies on D-mannitol and D-glucose. Four
of these, SFl-SS, SFl-Slo, SFl-SlS, and SFl-SZO, were negative for
L-sorbose-l-P reductase when crude extracts were tested. SFl-SS

was selected for further use in the study.

Mutants negative for D-mannitol-l-P dehydrogenase. Selection
procedures paralleled those of the previous section except that
D-mannitol and L-sorbose in the media were interchanged. However,
mutants which were unable to metabolize D-fructose and D-mannitol
were also unable to use L-sorbose or D-glucose, appearing as pale
colonies on MacConkey agar containing any of these four sugars. The
procedure was then modified in the following manner: instead of
just penicillin G, penicillin G plus 0.1 M cycloserine (79) was
added to growing cultures on 0.5% D-mannitol and the incubation
continued for 6 hr, at which point the final absorbance at 520 nm
was 0.070. Cells were washed as described in methods, and colonies
appearing pale on MacConkey agar supplemented with 1.0% D-mannitol
were selected. All such mutants appeared as pale colonies on

.MacConkey agar supplemented with 1.0% of either D-fructose,

102
D-glucose, or L-sorbose, indicating that the mutants possessed a
mutation or mutations affecting the general growth characteristics
of the cell. Since it was possible that D-mannitol might cause
growth inhibition of cells unable to utilize D—mannitol, mutants
from the penicillin-cycloserine procedure were plated onto mineral
agar containing 0.5% D-mannitol and 0.005% D-glucose. After incu-
bation for 24 hr at 30° pinpoint-size colonies were selected,
grown on nutrient broth, and plated on MacConkey agar containing
1.0% D-glucose, L-sorbose, D—fructose, or D-mannitol. Five mutants
appeared as pale colonies on D-mannitol or D—fructose and red on
D-glucose or L—sorbose. These mutants were grown on nutrient broth
and crude extracts were prepared and assayed for the presence of
D-mannitol-l-P dehydrogenase. SFl-MAZ has the lowest level of

dehydrogenase and was selected for further characterization.

Revertants of SFl-SS and SFl-MAZ. Revertants were selected
to show that SFl-SS and SFl-MA2 were the result of single point
mutations. After several days of incubation in 0.5% L-sorbose
mineral broth and 0.5% D-mannitol mineral broth, respectively,
growth appeared in the cultures. Samples were taken, serially
diluted, and plated on MacConkey agar supplemented with 1.0% L—
sorbose for SFl-SS revertants or 1.0% D-mannitol for SFl-MAZ
revertants. Colonies appearing red on these plates after 24 hr
incubation at 30° were selected, grown on nutrient broth, and
streaked on MacConkey plates containing D-fructose, L-sorbose,

D-mannitol, or D-glucose. Revertants were selected as those that

9.4..

103
gave red colonies on all four sugars. One mutant each (SFl-SSR

and SFl-MAZR) was selected for further characterization.

Mutants with elevated levels of D-mannitol-l-P dehydrogenase.
Mutants of SFl were mutagenized and treated as described for the
selection of L-sorbose constitutive mutants except that the cells
were cycled with 0.1 mM D-mannitol substituting for 0.1 mM L-
sorbose. After 30 cycles the culture was serially diluted and
plated on mineral agar containing 1.0% D-fructose. After incuba-
tion for 24 hr at 30° the four largest colonies were selected for
assay of D-mannitol-l-P dehydrogenase. SFl-MCl had the highest
levels of dehydrogenase present in crude extracts of mineral
broth-grown cells.

Diagnostic Tests to Demonstrate the Common
Ancestry of the Mutants
In all of these tests the Escherichia call control was the

strain B/r ara-2.

IMViC tests. The IMViC tests, which are used to differen-
tiate between Aerobacter aerogenes and Escherichia coli (80,81),
were deemed useful markers for confirming that the mutants were
derivatives of A. aerogenes and not contaminants. When mutants of
A. aerogenes and an E. coli control were subjected to these tests
the data summarized in Table 8 were obtained. All of the A.
aerogenes strains produced indole when grown on 1% tryptone broth,
gave a negative methyl red test, and could utilize citrate as a

sole source of carbon and energy. The Voges-Proskauer test for

 

104

Table 8. Results of IMViC tests on wild-type and mutant strains of
A. aerogenes and on an E. coli control

 

 

Indol Methyl red Voges-Proskauer Growth on

Strain test1 test2 test3 citrate4
E. coli

B/r 2£§:2_ pos. pos. + neg.
A. aerogenes

PRL—R3 pos. “neg. +++++ pos.

14 pos. neg. +++ pos.

SFl pos. neg. ++ pos.

SFl-SS pos. neg. ++ pos.

SFl-MAZ pos. neg. ++ pos.

SFl-SSR pos. neg. + pos.

SFl-MA2R pos. neg. + pos.

SFl-MCl pos. ' neg. ++++ pos.

 

1pos. = deep purple color forming at interface of medium and

air.

red color persisting for several hours.

pos.

3+ = relative intensity of red color developing after
several minutes.

4pos. = extensive growth on 0.5% sodium citratedmineral
broth within 24 hr.
neg. = no detectable growth after 24 hr.

105
acetoin production was variable. The reason for the variability
was not pursued, but it could be related to the rate of utilization
of acetoin as well as to the rate of its formation. In light of
the other data this was not considered to militate against the common

ancestry of the strains.

Colony characteristics on eosin-methylene blue agar. When

 

appropriate dilutions of A. aerogenes mutants and the E. coli con-
trol were plated on eosin-methylene blue agar, the colony character-
istics described in Table 9 resulted. The E. coli colonies were
distinct from those of the A. aerogenes strains. Although all of
the colonies of the A. aerogenes mutants were large and slightly
irregular in shape, the colonies formed by strains 14, SFl-SS,

and SFl-MAZ lacked the dark center which was apparent in the other
mutants and the wild-type strain. This may be explained by the
fact that current commercial preparations of eosin-methylene blue
agar (Difco) contain 0.5% sucrose. The three strains of different
morphology grew slowly on sucrose (Table 7), probably due to inhi-
bition of growth by an accumultion of intracellular D-fructose.
This same inhibition could have affected the acid production neces-

sary for the formation of the metallic green product (80).

Lysis by a bacteriophage specific for A. aerogenes. The

 

various mutant strains of A. aerogenes and an E. coli control were
exposed to infection by a bacteriophage discovered in our labora-
tory (Hanson, T. E., Ph.D. dissertation, 1969, Michigan State

University, p. 327). This phage lyses A. aerogenes PRL-RB but will

Table 9.

106

Colony characteristics of wild-type and mutant strains of

A. aerogenes and an E. coli control on eosin-methylene blue

agar plates

 

 

Strain Colony characteristics Appearance
E. coli
B/r ara-2 Small dark colonies appearing with a .

A. aerogenes

PRL-R3

14

SFl
SFl-SS
SFl-MAZ
SFl-SSR
SFl-MAZR

SFl-MCl

green metallic sheen when viewed by
reflected light

Large pale colonies, irregular with
small dark centers appearing with a
green metallic sheen when viewed by
reflected light

Large pale colonies, irregular with
light concave center and no metallic
green sheen

Same

Same

Same

Same

Same

Same

as

as

as

as

as

as

PRL-R3

14

14

PRL-R3

PRL-R3

PRL-R3

Same

Same

Same

Same

Same

Same

as

as

as

as

PRL-R3

14

14

PRL-RB

PRL-RB

PRL-R3

 

107

not lyse various strains of E. coli, Pseudomonas, or other common
bacteria which might be contaminants in our laboratory. Figure
26 shows that all mutant strains used in this study were lysed
when the bacteriophage was added to rapidly growing cultures in
nutrient broth. The lysis was similar to that observed with the
wild-type parent, PRL-R3, whereas E. coli control was unaffected
by addition of the phage.

Additional evidence that all mutants developed for this
study had a common ancestry is provided by growth patterns on

various carbohydrates discussed below.

Growth Response on Carbohydrates

Growth response on carbohydrates other than D-fructose. All
the mutant strains of A. aerogenes grow on mineral broth supplemented
with 0.5% each of D-glucose, D-galactose, D-mannose, L-rhamnose,
D—ribose, L-arabinose, D-xylose, maltose, melibiose, cellobiose,
or glycerol. All strains grow slowly on lactose. Growth response
on D-mannitol and L-sorbose is summarized in Table 7. Only the
mutants missing enzymes in the pathways for D-mannitol or Lesorbose
metabolism fail to grow on D—mannitol or L-sorbose: SFl-SS, the
L—sorbose-l-P reductase-negative mutant, is unable to utilize L-
sorbose for growth and SFl-MAZ, the mutant with very low levels of
D—mannitol—l-P dehydrogenase, will not grow on mineral broth sup-

plemented with D-mannitol.

 

108

.oom um s3oum oumz unwound nonuo Has .onm us nacho no: w~oo

owcowhmcomw

.nuoun ucoauusc ca wwoo .m one monomeuoo .w mo mousuaso

maesouo >Ho>fluoo on see o no poops no: sound .omonmoauouocn a an
mocmmoumm .w mo mcamuum accuse one cumuloawz mo mfimhq .mN ousoam

109

 

Wm... opal—u:

 

   

 

 

“53 and.
8m 8m 62 6 com 02 o
a n b D n 1 a

mo 6 - amassed . «unwise mo 6

. o «Edam / 4 madam

.. 4 x.

.1 .
2.6.. d 2.6

C \

8.9 8.6
on 8 /</4\4 .on o
3.9 +3.0
no.6- 08:18 0 1: no.6

. o Em

u o
2.6. /\
ON.O._ \

o
\ \

on o. o
3.6. o

8.0.

 

 

 

 

 

““1 085 IV ”MOSHV

110

Growth on D-fructose. Although strain 14 is unable to
utilize D-fructose as a sole source of carbon and energy, the L-
sorbose-l-P reductase-constitutive mutant, SFl, is able to grow on
0.3% D-fructose in mineral broth at a rate which is 38% that of the
wild-type, PRL-R3 (Figure 27). The earlier studies on the effect
of preinducing strain 14 on L-sorbose and D-mannitol (Figure 24)
suggested that the D-mannitol-l-P dehydrogenase might be the
limiting factor for optimal utilization of D-fructose by the new
pathway. This speculation is supported by the fact that a mutant
with increased levels of D-mannitol-l-P dehydrogenase, SFl-MCl,
is able to grow on D-fructose at a rate which is 78% that of PRL-R3
and twice that of SFl. Thus the levels of D-mannitol-l-P dehy-
drogenase (Table 11) correlate with the rate of growth on
D-fructose.

Net only is strain 14 unable to grow on D-fructose, but its
growth rate on nutrient broth is inhibited 83% by the presence of
0.5% D-fructose in the medium (Table 10). In contrast to strain
14, SFl is inhibited.only 13% when grown on nutrient broth in the
presence of D-fructose. SFl-MCl, which uses the new pathway to
metabolize D-fructose at a faster rate than SFl, not only is unin-
hibited but grows slightly faster when nutrient broth is supplemented
with 0.3% D-fructose. Mutants missing either of two enzymes of
the new pathway, L-sorbose-l-P reductase (SFl-SS) or D-mannitol-l-P
dehydrogenase (SFl-MAZ), are inhibited by D-fructose to nearly the

same extent as is strain 14.

111

1.00”
0.80

 

I J. I l

0.60
0.50
0.h0'«

PRL- R5

j

0.50 --

0.20 -

   
 

SFl-MCl

,A
0.06

o—o—o A.
0.05 \o y 14
0.04

0.05 «(I /

‘
0.02 A/

0.10 0
0.08?

 

ABSORBANCE AT 520 nm

 

 

 

0.01 T f I 'I t
0 100 200 300 400

mm (min)

Figure g1 - Growth curves of wild-type and mutant stragns of
‘5. aerogenes on 0.5% Defructose in mineral broth at 30 .

112

Table 10. Effect of the presence (+) of D-fructose on the growth
rate of mutants of A. aerogenes on nutrient broth

 

Percent inhibition
of growth rate on

 

Growth rate nutrient broth by

Strain D-fructose (generations/hr) D-fructose

PRL-R3 - 1.5 0
+ 1.5

14 - 1.1 83
+ 0.18

SFl - 0.92 13
+ 0.80

SFl-SS - 0.97 71
+ 0.27

SFl-MAZ - 0.97 71
+ 0.27

SFl-MCl - 1.0 -10
1 1

 

113

Enzyme Levels in Mutants of A. aerogenes

 

Table 11 summarizes the specific activities of L-sorbose-l-P
reductase and D-mannitol-l-P dehydrogenase found in crude extracts
of various mutants of A. aerogenes grown on L-sorbose, D-mannitol,
D—fructose, or glycerol. There were no detectable levels of D-
fructose-l-P kinase in any of the strains tested. D-Fructokinase
activity was below 0.008 units per mg protein and was apparently
too low to allow utilization of D-fructose for growth (76).

Strain 14, the mutant blocked in the normal pathways for
D-fructose metabolism, has detectable levels of L-sorbose-l-P
reductase only when grown on L-sorbose. All mutants which can
utilize D-fructose as a sole source of carbon and energy for
growth (i.e., SFl, SFl-SSR, SFl-MAZR, and SFl-MCl) possess reduc-
tase activity regardless of the carbon source on which they are
grown. A mutant missing the reductase completely, SFl-SS, also
loses the ability to grow on D-fructose. Thus the presence of
constitutive levels of L-sorbose-l-P reductase correlates with the
ability to grow on D-fructose.

D—Mannitol-l-P dehydrogenase is present in all strains
except SFl-MAZ, which possesses only 10-20% the normal uninduced
levels of this enzyme. Growth on D-mannitol increases the levels
of D-mannitol-l-P dehydrogenase dramatically in all mutants except
SFl-MAZ. SFl-MCl, the mutant selected by cycling an 0.1 mM
D-mannitol and 16.7 mM D-glucose, has at least twice the uninduced
levels of dehydrogenase found in other strains and at least one

and one-half times the levels of dehydrogenase when grown on

114

.muoouuxo moans cw cwououm ma mom :43 mom monsoon +o¢z modes: m00.0

mo muw>auom owmwoomm o o>om Houwccdauo om.0 no 00090:“ was spawn sameness so czoum maaou I ozv

.uuoouuxo moons cw onauosoou mnauononuomlq

mo namboa manouoouoo on on: omonuounq wn.0 co pounced odd spawn baseman: co ssouo naaoo I ozm

.ua v~ nouns spaoum senescence on u 02

.+04z sues mnanaouwccmeIa mo sofluoowxo on» ma oozwauouoo ouocooouoznoo mlauaouuccdalo I saw

.momz sues muauomouosuuno mo noduosoou on» ma vocaauouoo omouosoou mIHIomonuomIA I am

 

 

 

 

 

 

a
vm~.0 n~0.0 mav.a ~H0.0 00~.0 Hm0.0 mom.0 HH0.0 Hozlamm
«00.0 0v0.0 omw.0 om0.0 H50.0 0m0.0 ~00.0 0v0.0 awmzuamm
oz oz voz oz NH0.0 MN0.0 mH0.0 m~0.0 Nexunmm
00H.0 mm0.0 000.0 m~0.0 050.0 060.0 m~H.0 000.0 zmmIHmm
oz oz mao.0 000.0 oz moz ¢m0.0 000.0 mmuamm
000.0 ~m0.0 «No.0 0N0.0 hm0.0 000.0 m0a.0 0v0.0 How
oz oz 505.0 000.0 «no.0 m>0.0 00H.0 000.0 «H
(was Hmm Non Hum .mmo Hum .Nmo .Hmm aaauum
snouosuuuo HouascsSIo omOAHOmIA Houoomam

 

Acwououm ma mom owe you couscoum mooz no 002 moaoanv
.Nua>nuoa oamaodmm +

 

 

nuoun assures ca monouoznonuoo macauob so czoum nocomouoa .< mo mundane mo uuosuuxo
moons ca omssooouomroo mIHIHouaccosIo one omcuosoou muauomonuomuq no moeua>auos oamauomm .HH canoe

115

D-mannitol. These increased levels of dehydrogenase are not due to
a mutation affecting the affinity of the enzyme for D-mannitol-l-P
(Figure 28) but are due to changes in the level of activity of the
enzyme. The data from Table 11 indicate two conditions which are
necessary for a mutant blocked in the normal pathways of D-fructose
metabolism to utilize D-fructose as a sole source of carbon and
energy for growth. The first condition is that there be consti-
tutive levels of D-mannitol-l-P dehydrogenase present in the cell.
The second condition is that constitutive levels of L-sorbose-l-P
reductase must approach those found in cells induced by growth on

L-sorbose.

Intracellular Metabolite Levels in A. aerogenes Mutants

The intracellular accumulation of D-fructose-l-P and D-
mannitol-l-P in mutants of A. aerogenes correlates in the predicted
manner with the levels of L-sorbose-l-P reductase and D-mannitol-l-P
dehydrogenase and with the rate of growth on D-fructose (Table 12).
PRL-R3, which has D-fructose-l-P kinase when induced on D-fructose,
has on the order of 1% the level of D-fructose-l-P found in the
mutants which are missing this enzyme. Strains PAL-RB and 14
possess L-sorbose-l-P reductase only when grown on L-sorbose
(Table 11) and hence lack the ability to convert D-fructose-l-P
to D-mannitol-l-P. Those mutants missing enzymes in the new
D-fructose pathway (strains 14, SFl-SS, and SFl-MAZ) accumulate
greater amounts of D—fructose-l-P than those mutants with a func-
tional pathway. Strains l4 and SFl-SS, both lacking the reductase,

accumulate no detectable levels of D-mannitol-l-P as would be

..II I ll" ‘1' III

116

 

3.01
2.5 -
200 "

’3

'3

E

E 105 "'

....

Ji

3’

\

l-.
1.0 d
O 4

 

fi ‘— r j r v

4

m4

-5 -’+ -3 -2 -1 0 1 2 5 h 5
l/[D-mannitol-l-P] (nail)

Figgre g8 - Double reciprocal plot showing the relationship between
D-mannitol-l-P concentration and reaction velocity for D-mannitol-l-P
dehydrogenase from PRL-R} and SFl-MCl. Assays were performed with
crude extracts of nutrient broth-grown cells of _A_. aerogenes strains
PRL-R5 (o) and SFl-MCl (o).

d

117

 

 

 

Table 12. Intracellular metabolite levels in wild-type and mutant
strains of A. aerogenes induced for 5 hr on 0.5% D-
fructose as compared to enzymes present and growth rate
on 0.3% D-fructose
Enzymes present after 2
5 hr incubation in Metabolite levels Growth rate on
presence of 0.5% (nmoles/gram cells) 0.3% D-fructose

Strain D-fructosel F-l-P3 Mtl-l-P4 (generations/hr)

- +

PRL-R3 FK , FZl-PK , + 0.024 0.005 0.77
S-l-PR , M-l-PDH

14 8x7} 521-pr) + 4.1 0.005 0.00
S-l-PR , M-l-PDH

SFl 5x7} F-i-pr} 3.4 6.9 0.29
S-l-PR , M-l-PDH

sri-ss at?) le-PKTZ + 4.4 0.005 0.00
S-l-PR , M-l-PDH

SFl-SSR Same as SFl 2.1 6.6 0.21

SFl-MAZ 8x7} F;1-px7} __ 5.1 12.6 0.00
S-l-PR , M-l-PDH

SFl-MAZR Same as SFl 3.0 5.5 0.23

SFl-MCl 1m", F—i—px", 3.1 2.3 0.55

s-i-PRT, M-l-PDH

 

lEnzymes present (+), absent (-), or present at increased
levels (++).

2There were less than 0.005 nmoles per gram of cells (wet
weight) of D-fructose-G-P and D-fructose-l,6-P2.

PRL-R3 contained
0.083 nmoles of D-fructose-l,6-P2 per gram of cells.

3D-Fructose-l-P determined with D-fructose-l-P kinase (see
methods for details).

4D-Mannitol-l-P determined with D-mannitol-l-P dehydrogenase

(see methods for details).

118
predicted according to the order of steps in the new pathway.
SFl-MA2, which has greatly reduced levels of D-mannitol-l-P
dehydrogenase, accumulates twice the amount of D-mannitol-l-P
and, as a consequence, increased amounts of D-fructose-l-P.
SFl-MCl, with twice the level of dehydrogenase, accumulates half
the level of D-mannitol-l-P, and grows at twice the rate of

mutants with normal levels of dehydrogenase on D—fructose.

Discussion and Summagy,

 

This section of my dissertation has demonstrated the evolu-
tion of a novel pathway for the metabolism of D-fructose in A.
aerogenes. By using a mutant (strain 14) blocked in the two path-
ways by which D-fructose is normally metabolized, I derived a mutant
capable of metabolizing D-fructose via the pathway shown in Figure
29.

The studies presented in Figure 24 indicated that the presence
of L-sorbose-l-P reductase within the bacterium is sufficient to
allow the bacterium to grow on D-fructose when the normal pathways
of D-fructose are blocked. These same studies also showed that
D-mannitol-l—P dehydrogenase is the limiting enzyme for growth on
D—fructose when the reductase is also present. Since the P-enol-
pyruvate-dependent phosphotransferase system for L-sorbose appears
to be constitutive (l) and D-glucitol-l-P dehydrogenase is induced
by D-glucitol-G-P (1), then enriching for mutants constitutive for
growth on L—sorbose should consequently select mutants constitutive
for L-sorbose-l-P reductase. With the modified procedure described

in Results, those mutants with the higher levels of constitutive

119

.onouosu " won one “Houaosawun

" Hum "Heuoccssin ” Hus “encasemun " use monouoonmuo u sum "sounzu uuuuomscouuonaacao ucov
Iconocummm " mam .oszuco mo uHo>oa sabouoouoc on no concouoov now assauuusa one wummm a.o.a
announce omuo nosed oocaauovcz "masses we uHo>o~ vouuuuuco no o>eu=uauwcoo pom ocoausuaa

ous Tues .64 .9533 some nosed mucous—Sung mo 6930039: new zuanusa soz Alv
“nzssnuoo ofifionuuma Haauoz A v “cucuosumun ucfisuauucuuca mo monsoon caboouom
.mmmMMMHMN .fi ca Heuficcqauo was auncouoana «uncuosuuun we aheanusa ufiHonsusa 05H 1

 

 

 

 

 

 

 

 

   

 

 

cocoa—om...— sec—c0210
_ bOJfiWPL _uSHWFL
auwflom earn—a!
6.618 as...
7:8
:Hamum
$303“. 0111.11-11! E... J7 m o 3....—
.4
xx N anusuw
\. flmroozoilll $0.72... All. 13.3".
m \\ a: ,

 

 

120
reductase should have a selective advantage over those mutants with
lower constitutive levels of reductase. The results shown in
Figure 25 confirmed this prediction. One of the mutants selected
in this manner, SFl, possesses constitutive levels of reductase
comparable to the levels of reductase induced in the parental
mutant, strain 14, when grown on L—sorbose (Table 11).

Further evidence that the new pathway is as depicted in
Figure 29 was provided by the isolation of two D-fructose-negative
mutants derived from SFl. These mutants, missing either L-sorbose-
l-P reductase (SFl-SS) or D-mannitol-l-P dehydrogenase (SFl-MAZ),
lost the ability to utilize D-fructose as a sole source of carbon
and energy, whereas their revertants, SFl-SSR and SFl-MAZR, con-
comitantly regained the missing enzyme activities and the ability
to grow on D-fructose at the same rate as SFl.

The accumulation of intracellular metabolite levels of
D-fructose-l-P and D-mannitol-l-P correlates in a predictable
fashion with the enzymes present or absent in the various mutants
(Table 12). The data for SFl-MCl show that this mutant has about
2.5 times the level of dehydrogenase present in SFl, SFl-SSR, and
SFl-MAZR and also grows on D-fructose at a rate which is 2.3 times
the rate of these three mutants. The levels of D-mannitol-l-P
accumulated in SFl, SFl-SSR, SFl-MAZ, SFl-MAZR, and SFl-MCl are
inversely proportional to the levels of dehydrogenase present in
crude extracts of these bacteria.

By these results I have demonstrated that it is possible

to genetically manipulate A. aerogenes strains lacking the ability

121
to utilize D-fructose via the normal metabolic pathways, to produce
a mutant possessing a novel pathway for the metabolism of D-fructose.
This pathway is: D-fructose +D-fructose-l-P +'D-mannitol-l—P‘+
D-fructose-G-P. The genetic manipulations resulted in a mutant
which possesses this pathway by conscripting enzymes normally
associated with three separate metabolic pathways: (1) the PEP:D-
fructose l-phosphotransferase systemIOf D-fructose metabolism;
(ii) the L-sorbose-l-P reductase of L-sorbose metabolism; and (iii)
the D-mannitol-l-P dehydrogenase of D-mannitol metabolism. This is
a unique pathway previously unreported in any other organism and
is the first new pathway evolved in the laboratory for the metabolism

of a normal substrate.

REFERENCES

10.

11.

12.

13.

14.

REFERENCES

Kelker, N. E., Simkins, R. A., and Anderson, R. L. J. Biol.
Chem., 247, 1479 (1972).

Adelman, R. C., Ballard, F. J., and weinhouse, S., J. Biol.
Chem., 242, 3360 (1967).

Tengerdly, R. P., J. Biochem. Microbial. Technal. Eng., 3,
241 (1961).

Isono, M., Nakanishi, I., Sasajima, K., Motizuki, K., Kanzaki,
T., Okazaki, H., and Yoshino, M., Agr. Biol. Chem., 35, 424
(1968).

Tsukada, Y., and Perlman, D., Biatechnal. Bioengineer., 12!
799 (1972).

Tsukada, Y., and Perlman, D., Biatechnal. Bioengineer.,.lg,
811 (1972).

Kitamura, 1., and Perlman, D., Amer. Sac. Microbial. (Abstr.),
E14, 3 (1974).

Makover, 8., Ramsey, G. B., Witt, C. G., and Vane, F. M.,
Amer. Sac. Microbial. (Abstr.), P64, 155 (1974).

Makover, S., and Ramsey, G. 8., Amer. Sac. Microbial. (Abstr.),
P287, 188 (1973).

Ramsey, G. B., Makover, S., and Vane, F. M., Amer. Sac. Micro-
bial. (Abstr.), P65, 155 (1974).

Kanzaki, T., and Okazaki, E., Agr. Biol. Chem., 32, 432 (1970).

Yamada, Y., Iizuka, K., Aida, K., and Uemura, T., J. Biachem.,
62, 223 (1967).

Sato, K., Yamada, Y., Aida, K., and Uemura, T., J. Biochem.,
66, 521 (1969).

Avigad, G., Englard, 5., and Pifko, 3., J. Biol. Chem., 241,
373 (1966).

122

15.

16.

17.

18.

18a.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

123

Yamada, Y., Aida, K., and Uemura, T., J. Biachem., 641, 803
(1967).

Sato, K., Yamada, Y., Aida, K., and Uemura, T., Agr. Biol.
Chem., 33, 1612 (1969).

Mann, K. M., and Lardy, H. A., Methods Enzymal., 3, 177 (1957).

Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall,
R. J., J. Biol. Chem., 193, 265 (1951).

Clark, Jr., J. 0., Experimental Biochemistry, p. 32, W. H.
Freeman and Co., San Francisco, California.

Fiske, C. H., and SubbaRow, Y., J. Biol. Chem., 33, 375 (1925).
Putman, E. W., Methods Enzymal., 3, 62 (1957).

Dische, 2., and Devi, A., Biachim. Biophys. Acta, 32, 140
(1960).

Bissett, D. L., Hanson, T. E., and Anderson, R. L., Microchem.
J., 23_(1974).

Sweely, C. C., Wells, W. W., and Bentley, R., Methods Ehzyma1.,
3, 9S (1966).

Laine, R. A., and Sweely, C. C., Anal. Biachem., 33, 533 (1971).
Mann, K. M., and Lardy, H. A., Methods Enzymal., 3, 179 (1957).

Zinbo, M., and Sherman, W. R., J. Amer. Chem. Sac., 33, 2105
(1970).

Wolff, J. 8., and Kaplan, N. O., J. Bacterial., 23, 557 (1956).

Garen, A., and Levinthal, C., Biachim. Biophys. Acta, 33,
470 (1960).

Tombs, M. P., Souter, F., and MacLagan, N. F., Biochem. J.,
23, 167 (1959).

Lisa, M., Horowitz, S. 8., and Kaplan, N. O., J. Biol. Chem.,
237, 1342 (1962).

Jones, T. M., and Albersheim, P., Plant Physiol., 32, 926
(1972).

March, S. C., Parikh, I., and Cuatrecasas, P., Anal. Biachem.,
32, 149 (1974).

Cuatrecasas, P., J. Biol. Chem., 245, 3059 (1970).

 

III I!

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

SO.

51.

124

Lamed, R., Levin, Y., and Wilchek, M., Biachim. Biophys. Acta,
304, 231 (1973).

Hedrick, J. L., and Smith, A. J., Arch. Biochem. Biaphys., 126
155 (1968).

Fairbanks, G., Steak, T. L., and Wellach, D. F. H., Biochemistry,
39, 2606 (1971).

Chervenka, C. H., Manual af'Methads for the Analytical Ultra-
centrifuge. Beckman Instruments, Palo Alto, California (1970).

Van Holde, K. E., Sedimentation Equilibrium, 9 pp. Beckman
Instruments, Palo Alto, California (1967).

Robertson, D. C., Hammerstedt, R. H., and waod, W. A., J.
Biol. Chem., 246, 2075 (1971).

Edelhoch, H., Biochemistry, 3, 1948 (1967).

Ohlsson, R., Brodelius, P., and Mosbach, R., F.E.B.S. Letters,
33, 234 (1972).

Lowe, C. R., Mosbach, R., and Dean, P. D. G., Biochem. Biophys.
Res. Cbmmun., 33, 1004 (1972).

Mosbach, R., Guilford, H., Ohlsson, R., and Scot, M., Biochem.
J., 127, 625 (1972).

Reyneud, J., Luccione, F., Bouthier, 0., Savay, J., and
Derrien, Y., Biachim. Biophys. Acta, 221, 367 (1970).

Phelps, R. A., and Putnam, F. W., The Plasma Proteins, 3, 435
(1960) .

Hammerstedt, R. H., Mahler, H., Decker, K. A., and waod, W. A.,
J. Biol. Chem., 246, 2069 (1971).

Sund, H., Weber, R., and Molbert, E., Eur. J. Biachem., 3,
40 (1967).

Castellino, F. J., and Barker, R., Biochemistry, Z, 2207
(1968).

Sund, H., Biachem. 2., 333, 205 (1960).

Peace, A., McKay, R. H., Stolzenbach, F., Cahn, R. D., and
Kaplan, N. 0., J. Biol. Chem., 239, 1753 (1964).

Andrews, P., Biachem. J°'.21' 222 (1964).

52.

53.

53a.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

125

Agren, G., Engstrom, L., and Eklund, 8., Biochim. Biophys.
Acta, 32, 143 (1963).

Chiancone, E., Vecchini, P., Forlani, L., Antonini, E., and
Wyman, J., Biochim. Biophys. Acta, 127, 549 (1966).

Wrigley, C. W., Hethads Enzymal., 33, 559 (1971).

Smith, M. H., "Molecular weights of Proteins and Some Other
Materials Including Sedimentation, Diffusion, and Frictional
Coefficients and Partial Specific Volumes", pp. C-3 to C-35,
Handbook of Biochemistry, H. A. Sober, Ed. The Chemical
Rubber Company, Cleveland, Ohio (1970).

Dwyer, F. P., "Enzyme-Metal Ion Activation and Catalytic
Phenomena with Metal Complexes", p. 355, Chelating Agents
and Metal Chelates, F. W. Dwyer, Ed. Academic Press, New
York (1964).

Sumner, J. B., and Soames, G. F., The Chemistry and Methods
of Enzymes, p. 343. Academic Press, Inc., New York (1947).

Hers, H. C., Biochim. Biophys. Acta, 33, 427 (1953).
DeLey, J., J. Gen. Microbial., 33, 31 (1961).

Mowshowitz, S., Avigad, G., and Englard, S., J. Bacterial.,
118, 1051 (1974).

Mowshowitz, S., Avigad, G., and Englard, S., J. Bacterial.,
119, 363 (1974).

Englard, 5., and Avigad, G., J. Biol. Chem., 240, 2297 (1965).

Hanson, T. E., and Anderson, R. L., J. Biol. Chem., 240,
1644 (1966).

Hanson, T. E., and Anderson, R. L., Prac. Natl. Acad. Sci.
U.S.A., 33, 269 (1968).

Sapico, V., Hanson, T. E., welter, R. W., and Anderson, R. L.,
J. Bacterial., 33, 51 (1968).

Kelker, N. E., and Anderson, R. L., J. Bacterial., 112, 1441
(1972).

Morrissey, A. T. E., and Fraenkel, D. G., Biochem. Biophys.
Res. Commun., 33, 467 (1968).

Ferenci, T., and Kornberg, H. L., F.E.B.S. Letters, 33, 127
(1971).

68.
69.

70.
71.
72.
73.
74.
75.
76.
77.
78.
79.

80.

81.

126
Ferenci, T., and Kornberg, H. L., Biochem. J., 132, 341 (1973).
Kotze, J. P., S. Afr. J. Agri. Sci., 33, 349 (1968).

Hugo, H. v., and Gottschalk, G., Eur. J. Biachem., 33, 455
(1974). .

Hugo, a. v., and Gottschalk, G., F.E.B.S. Letters, 39, 106
(1974).

Patni, N. J., and Alexander, J. R., J. Bacterial., 105, 226
(1971).

Sobel, M. E., and Krulwich, T. A., J. Bacterial., 113, 907
(1973). “"

Conrad, R., and Schlegel, H. G., Biochim. Biophys. Acta, 358,
221 (1974).

Reeves, R. E., warren, L. G., and Hsu, D. S., J. Biol. Chem.,
241, 1257 (1966).

Kelker, N. E., Hanson, T. E., and Anderson, R. L., J. Biol.
Chem., 245, 2060 (1970).

Mayo, J. W., and Anderson, R. L., J. Bacterial., 100, 948
(1969). ""‘

Hegeman, G. D., and Rosenberg, S. L., Annu. Rev. Microbial.,
33, 429 (1970).

Ornston, L. H., Ornston, M. R., and Chou, G., Biochem. Biophys.
Res. Cammun., 33, 179 (1969).

DifCa Manual of Dehydrated Culture Media and Reagents for
Microbiological and Clinical Laboratory Procedure, 9th Ed.
Difco Laboratories, Inc., Detroit, Michigan, 350 pp. (1953).

Manual af'Methads for Pure Culture Study of Bacteria, Leaflet V,
12th ed., H. J. Conn, Ed. Biotech Publications, Geneva, New
York (1951).

HIWHIIUIHWI

9

l

174 8845

03

TI

3 1