Metabolism of L-arabinose in Azospirillum brasilense

MISSING IMAGE

Material Information

Title:
Metabolism of L-arabinose in Azospirillum brasilense
Alternate title:
L-arabinose
Physical Description:
ix, 69 leaves : ill. ; 28 cm.
Language:
English
Creator:
Novick, Norman James, 1953-
Publication Date:

Subjects

Subjects / Keywords:
Nitrifying bacteria   ( lcsh )
Nitrifying bacteria   ( fast )
Azospirillum brasilense
Microbiology thesis Ph. D
Dissertations, Academic -- Microbiology -- UF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1980.
Bibliography:
Includes bibliographical references (leaves 65-68).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Norman James Novick.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
aleph - 020728868
oclc - 07216224
System ID:
AA00026371:00001


This item is only available as the following downloads:


Full Text











METABOLISM OF L-ARABINOSE IN AZOSPIRILLUM BRASILENSE


















BY

NORMAN JAMES NOVICK

















A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA

1980














ACKNOWLEDGEMENTS



The author wishes to thank Dr. Max E. Tyler for his advice and

guidance and wishes him the very fastest recovery from his present

illness. The author also wishes to thank Dr. Arnold S. Bleiweis for

his aid and advice under difficult circumstances and Dr. James F.

Preston for his technical assistance and suggestions in the preparation

of this manuscript. The author would also like to thank Dr. David H.

Hubbell and Dr. Stanley C. Schank for their suggestions in the

preparation of this manuscript.

My wife, Connie, has been a constant companion and advisor

during the time of my research and in the preparation of this manu-

script. I want to thank her for her love and guidance.



























ii















TABLE OF CONTENTS


ACKNOWLEDGEMENTS. . . . . . ... ii

LIST OF TABLES. . . . . . . v

LIST OF FIGURES . . . . . . vi

ABSTRACT. . . . . . . .. viii

SECTION

I IDENTIFICATION OF ENZYMES INVOLVED IN L-ARABINOSE
METABOLISM IN AZOSPIRILLUM BRASILENSE. ASSAY OF TCA
CYCLE DEHYDROGENASES AND IDENTIFICATION OF COMPONENTS
INVOLVED IN ELECTRON TRANSPORT . . . 1

Introduction . . . . . 1
Materials and Methods. . . . . 3
Bacterial Strain. . . . . 3
Media . . . . . . 3
Growth and Nitrogen Fixation. ... . . .. 3
-Acetylene Reduction Assay . . . 4
Determination of Arabinose Pathway Enzymes. . 4
Identification of Pathway End Product . . 5
Chromatography. . . . . 6
TCA Cycle Enzymes . . . . 7
Cytochrome Determinations . . . 7
Results. . . . . . . 8
Growth and Nitrogen Fixation. . . . 8
Arabinose Pathway Enzymes . . . 8
End Product Accumulation. . . . .. 14
Identification of Reaction Products . . 14
TCA Cycle Dehydrogenases and Cytochrome Content 18
Discussion . . . . . . 27

II PURIFICATION AND CHARACTERIZATION OF L-ARABINOSE
DEHYDROGENASE AND L-ARABONATE DEHYDRATASE FROM
L-ARABINOSE GROWN CELLS OF AZOSPIRILLUM BRASILENSE . 31

Introduction . . . . . 31
Materials and Methods. . . . . .. 32
Bacterial Strain. . . . . .. 32
Enzyme Purification . . . . 32
Protein Determination . . . . 34
Enzyme Assays . . . . . 34
Molecular Weight Determination. . . .. 34



iii










Results. . . . . . . 35
Characteristics of L-arabinose Dehydrogenase .. .. 38
Characteristics of L-arabonate Dehydratase. .. 44
Molecular Weight Determination. . . .. 44
Enzyme Product Identification . . . 44
Polyacrylamide Gel Electrophoresis. . .. 55
Discussion . . . . . . 55

APPENDIX. . . . . . . 61

LITERATURE CITED. . . . . . 65

BIOGRAPHICAL SKETCH . . . . . . 69












































iv















LIST OF TABLES


Table

1-1 L-arabinose pathway enzymes. . . . .. 13

1-2 Accumulation of alpha-ketoglutarate in crude
extracts of L-arabinose grown cells; with
L-arabonate as substrate . . . . 17

1-3 TCA cycle dehydrogenases . . . . 19

1-4 Cytochrome content of membrane and soluble fractions 26

2-1 Purification of L-arabinose dehydrogenase. . ... 36

2-2 Purification of L-arabonate dehydratase. . ... 37

2-3 Effect of divalent cations and reducing agents on
L-arabinose dehydrogenase activity . . . 43

2-4 Substrate specificity of L-arabinose dehydrogenase . 45

2-5 Effect of divalent cations and reducing agents on
L-arabonate dehydratase activity . . . 52



























V














LIST OF FIGURES


Figure

1 Growth and acetylene reduction on L-malate by
Azospirillum brasilense. . . . . 10

2 Growth and acetylene reduction on L-arabinose by
Azospiriltum brasilense. . . . ... 12

3 Reduction of NAD in crude extracts of L-arabinose
grown cells with I pmole potassium L-arabonate or
DL-KDA; as substrate . . . . 16

4 Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
105,000 xg supernatant. . . . . .. 21

5 Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
35,000xg pellet. . . . . .. 23

6 Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
105,000 xg pellet . . . . . 25

7 Elution from a Biogel A 1.5M (2 x 150 cm) column of
L-arabinose dehydrogenase and L-arabonate dehydratase. 40

8 Elution from a DEAE cellulose column (2 x 10 cm) of
L-arabinose dehydrogenase and L-arabonate dehydratase. 42

9 Lineweaver-Burk plot showing the effect of substrate
concentration on reaction velocity . . . 47

10 Effects of pH on reaction velocity of L-arabinose
dehydrogenase. . . . . . .. 49

11 Effect of pH on reaction velocity of L-arabonate
dehydratase. . . . . . 51

12 Molecular weight determination of L-arabinose
dehydrogenase and L-arabonate dehydratase by gel
filtration technique . . . . .. 54

13 Polyacrylamide slab gel of crude and partially purified
extracts of L-arabinose dehydrogenase. . . ... 57


vi










14 Polyacrylamide slab gel of partially purified
extracts of L-arabonate dehydratase. . . 59

15 Schematic of oxygen controller . . . 63





















































vii












Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


METABOLISM OF L-ARABINOSE IN AZOSPIRILLUM BRASILENSE

By

Norman James Novick

December 1980

Chairman: Arnold S. Bleiweis
Major Department: Microbiology and Cell Science


Under dinitrogen fixing conditions Azospirilium brasilense showed

comparable generation times using L-arabinose or malate as sole carbon

and energy sources. This root associated bacterium was found to

metabolize L-arabinose through an oxidative pathway that has also been

found in certain species of Pseudomonas and fast-growing Rhizobium.

L-Arabinose is converted to L-arabono-'-lactone by an NAD(P) dependent

dehydrogenase, hydrolyzed to L-arabonic acid by a lactonase, and

dehydrated to L-2-keto-3-deoxyarabonate (L-KDA) by dehydratase activity.

In crude extracts NAD is rapidly reduced if potassium L-arabonate or

DL-KDA is added as substrate. DL-KDA has been found to reduce NAD at

3 times the rate of L-arabonate. Alpha-ketoglutarate accumulated in

crude extracts to which L-arabonate and NAD were added. It is proposed

that L-KDA is dehydrated to alpha-ketoglutaric semialdehyde which is

then oxidized to alpha-ketoglutaric acid by an NAD dependent dehydrogena-

tion reaction. L-Arabinose dehydrogenase and L-arabonate dehydratase

have been partially purified and characterized. The NAD(P) dependent



viii










dehydrogenase, which has been found to be specific for the L-arabono-

configuration, has been purified 59 fold. No enhancement of activity

in the presence of any divalent cation or reducing agent tested has

been found. The K values of 75 pM and 140 pM were found with NADP

and NAD as cofactors, respectively. The enzyme has a pH optimum of

9.5 in glycine buffer and was stable when heated to 55C for 5 minutes.

The enzyme product has been identified as L-arabono-y-lactone. L-

Arabonate dehydratase has been purified 38 fold. The presence of

MgCl2 and MnCl2 has been found to significantly increase enzyme activity.

The pH optimum for the dehydratase has been found to be 7.8. The

enzyme product was identified as L-2-keto-3-deoxyarabonate, In

further studies TCA cycle dehydrogenases have been assayed and indicate

an active TCA cycle. The presence of b and c type cytochromes has been

confirmed and their distribution between membrane and soluble fractions

determined.



























ix














SECTION I
IDENTIFICATION OF ENZYMES INVOLVED IN L-ARABINOSE
METABOLISM IN AZOSPIRILLUM BRASILENSE. ASSAY OF TCA CYCLE
DEHYDROGENASES AND IDENTIFICATION OF COMPONENTS
INVOLVED IN ELECTRON TRANSPORT



Introduction


Azospirillum brasilense (ATCC 29145) grows and fixes molecular

nitrogen best on organic acids such as succinate, malate, and pyruvate

(13, 28, 29, 38). Variable results have been reported for A. brasilense

growth on sugars. Day and Dobereiner (13) reported that several sugars

including glucose could be used for growth under nitrogen fixing con-

ditions but only if "starter" organic acid was added to the medium.

Okon et al. (28) found growth and acetylene reduction rates were less

on galactose than on several organic acids tested and found little or no

growth under nitrogen fixing conditions with glucose as a carbon and

energy source. Tarrand and Krieg (38) have also found A. brasiZense

unable to use glucose as a sole carbon and energy source under nitrogen

fixing conditions. Okon et al. (29) found little or no increase in

oxygen uptake rates above endogenous levels in cell free extracts if

glucose, fructose, galactose, or several phosphorylated metabolic

intermediates were added as substrates.

Child and Kurz (6) reported that in certain cases nitrogenase

activity in A. brasilense could be enhanced by supplementing the

organic acid containing medium with arabinose. In examining A.

brasilense growth under nitrogen fixing conditions on a variety of






2



substrates, I found that L-arabinose served as an excellent sole carbon

and energy source. I therefore decided to investigate the mode of

L-arabinose catabolism in this root associated bacterium.

My initial failure to find a pathway involving phosphorylated

intermediates led me to look at other possible pathways. There are two

well documented pathways of L-arabinose metabolism not involving

phosphorylated intermediates. In the first pathway L-arabinose is

oxidized to L-arabono-y-lactone by an NAD(P) dependent dehydrogenase,

the lactone is cleaved by lactonase to L-arabonic acid, followed by two

successive dehydration reactions forming L-2-keto-3-deoxyarabonate and

alpha-ketoglutaric semialdehyde, respectively. The last step involves

the NAD dependent dehydrogenation of the semialdehyde to alpha-

ketoglutaric acid. The second pathway has the same initial steps but

the L-2-keto-3-deoxyarabonate is cleaved through an aldolase reaction

to glycolaldehyde and pyruvate. The first pathway has been demonstrated

in Pseudomonas saccharophila (42),Pseudomonas fragi (7, 41), and the

fast growing rhizobia (15, 16), R. metiZoti, R. trifolii, R. phaeseoZi,

and R. leguminosarum. The second pathway has been demonstrated in

Pseudomonas strain MSU-1 (8), and slow growing rhizobia; R. japonicum

(15, 31) and Rhizobium sp. 32HI (15). In addition to these L-arabinose

dehydrogenase has also been demonstrated in Aquaspirillum gracile (23)

although the subsequent pathway is unknown.

In this paper evidence is presented indicating the first pathway

(i.e. L-arabinose ---- alpha-ketoglutaric acid) is present in A.

brasilense. The presence of TCA cycle dehydrogenases and cytochromes

is also demonstrated, indicating that the alpha-ketoglutarate formed






3



provides energy to the cell through electron transport coupled to

TCA cycle dehydrogenases.


Materials and Methods


Bacterial Strain

Azospirillum brasilense (ATCC 29145) was obtained from the

American Type culture collection. For daily use cultures were maintained

on Trypticase soy agar (BBL) with biweekly transfer. Stock cultures

were frozen in liquid nitrogen.


Media

The growth medium was that of Nelson and Knowles (27) with twice

the concentration of potassium phosphate buffer (65 mM, pH 7.1).

Malate or L-arabinose (37 mM) was filter sterilized and added to the

autoclaved and cooled medium. Ammonium chloride, 0.1%, was added to

ammonia grown cells.


Growth and Nitrogen Fixation

Two and one-half liters of the nitrogen free medium were added to

a vessel consisting of a 2.8 L Fernbach flask tightly capped with a

rubber stopper through which had been placed a gas inlet port (ending

in a sparging stone), a gas outlet port, a sampling port, and an oxygen

probe (2). The oxygen probe and air line were connected to an oxygen

stat (see Appendix 1). The flask containing a stirring disc was

immersed ina 25C water bath set on a magnetic stirrer. Air and

nitrogen were mixed to maintain the 02 concentration at 0.5% + 0.05%.






4



For NH4CI grown cells, 1.5 L of medium were placed in a cotton stoppered

Fernbach flask and rapidly shaken at room temperature. In both N2 and

ammonia grown cells a 10% inoculum containing 0.05% NH 4 C was grown

aerobically.


Acetylene Reduction Assay

Six milliliters of cells were removed anaerobically from the growth

vessel and placed in a 125 ml flask that had previously been sparged

with argon and capped with a serum stopper. Oxygen, 0.5%,was added as

air back to the flask. The samples were shaken on a New Brunswick

rotary shaker for I h at 150 rpm and 25C. One-tenth milliliter gas

volume was removed and injected into a Varian 2400 series gas chro-

matograph with Poropak Q columns and flame ionization detectors.

Growth was followed by reading optical density at 560 nm and

protein determined by the method of Lowry (24).


Determination of Arabinose Pathway Enzymes

Both NH4Cl and N2 grown cells gave similar specific activities for

enzymes involved in arabinose metabolism. Due to this and the ease of

growing large batches of cells with a fixed nitrogen source, ammonium

chloride grown cells were used for the enzyme studies. Crude extracts

were prepared in the following manner. Cells, near the middle or end of

exponential growth, were collected by centrifugation and washed once in

O.IM sodium-potassium phosphate buffer, pH 7.5, resuspended to ca. 0.25

g per ml in the same buffer, passed twice through an Aminco french

pressure cell at 20,000 p.s.i., and centrifuged at 10,000 xg for 30 min.

The supernate, crude extract, could be stored for several months at






5



-80C without significant loss of enzymatic activity. L-arabinose

dehydrogenase was assayed in a reaction mixture containing 125 Jmoles

glycine/NaOH buffer, pH 9.0, 0.5 pmoles NAD, 0.5 pmoles L-arabinose,

1 pmole NaCN, and enzyme plus water to I ml. The change in optical

density at 340 nm was followed on a Zeiss spectrophotometer at 250C.

Lactonase activity was measured by the methodof Dahms and Anderson (10) using

alkaline hydroxylamine reagent (18) to quantitate the disappearance

of L-arabonolactone. Arabonate dehydratase assay was that of Pedrosa

and Zancan (31) with L-2-keto-3-deoxyarabonate (L-KDA) accumulation

being measured with semicarbazide reagent (25). KDA aldolase was

assayed in the reverse direction (31) with KDA formation being followed

by the thiobarbituric-periodate method (43). The following assay

mixture was used to determine NAD dependent KDA oxidation: 60 pmoles

potassium phosphate buffer pH 7.2, 10 pmoles NAD, 10 pmoles potassium

arabonate, 1 pmole NaCN, and water plus enzyme to 1 ml. In all cases

I unit of enzyme activity is that amount of enzyme which produces I

pmole of product per minute at 25C.


Identification of Pathway End Product

The accumulation of alpha-ketoglutaric acid in crude extracts,

after the addition of potassium arabonate (10 imoles), was determined

as follows. The arabonate oxidation mix was the same as that used to

follow KDA oxidation. In some experiments water or 10 ymoles sodium

arsenite replaced NaCN. The reaction mix was incubated up to six hours

and alpha-ketoglutarate was enzymatically quantitated in a reaction

mixture containing potassium phosphate buffer, 80 pmoles, pH 7.2, 0.32

lmoles NADH, 40 pmoles NH4C1, 3.3 units glutamate dehydrogenase (Sigma)






6



and 0.05 ml of the above arabonate oxidation mixture in a total of I ml.

The amount of NADH oxidized was determined against controls without

NADH and without both glutamate dehydrogenase and NH4CI to correct for

NADH oxidase activity. Controls were also run using the arabonate

oxidation mixture to which arabonate had not been added. Pyruvate was

enzymatically determined in a reaction mixture containing potassium

phosphate buffer, 50 pmoles, pH 7.2, 1 unit lactic dehydrogenase

(Sigma), 0.32 pmoles NADH, 0.05 ml arabonate oxidation mix, and water

to 1 ml. Controls were run without lactic dehydrogenase to correct

for endogenous NADH oxidase activity and with arabonate oxidation

mixture to which arabonate had not been added. Glycolaldehyde was

quantitated with diphenylamine reagent (14).


Chromatography

Dehydrogenase and dehydratase reaction products, using partially

purified enzyme, as well as identification of the arabinose pathway

terminal product, were determined by thin layer and paper chromatography.

The samples were processed as follows. The reaction mixture was passed

through a small Dowex-X8 (H+ form) column, 0.5 ml bed volume, to remove

cations and precipitate proteins, then passed through a 0.45 p membrane

filter, and concentrated 20 fold by lyophilization. KDA was further

processed by the method of Weimberg (40) before spotting. Samples were

spotted on Whatman #1 paper and resolved in one of the following solvent

systems: pyridine: 1-butanol:water (6:4:3), l-propanol:formic acid:

water (6:3:1); or 1-butanol:l-propanol:water (10:7:5). Compounds were

detected with alkaline silver nitrate reagent (36) or aniline-xylose






7



reagent (36). Dinitrophenylhydrazones of alpha-keto acids, produced by the

methodof Cavallini et al. (5), were spotted on Silica G gel plates and

resolved in 1-butanol saturated with 3% NH4OH. Spots were accentuated

by spraying with 0.5N NaOH.


TCA Cycle Enzymes

Isocitrate and malate dehydrogenases were assayed by the method of

Reeves et al. (33). One unit of activity was the amount of enzyme which

reduced 1 pmole of NAD(P) per minute at 25C. Succinate dehydrogenase

was assayed by a modification of the method of Veeger et al. (39). The l ml

reaction mixture contained 50 imoles potassium phosphate buffer, pH

7.6, 40 pmoles sodium succinate, 1 pmole NaCN, I mg phenazine metho-

sulfate, and 0.25 pmoles 2,4 dichlorophenolindolphenol (DCPIP). One

unit of activitywas the amount of enzyme which reduced 1 lmole of DCPIP

per min. at 25C. The dye had an extinction coefficient at 600 nm of

19,100 (35). Alpha-ketoglutarate activity was assayed by following the

reduction of ferricyanide. The assay was the same as that for pyruvate

decarboxylase (32). One unit of activity was the amount of enzyme

necessary to reduce 2 pmoles of ferricyanide per hour.


Cytochrome Determinations

Crude extracts were separated into membrane and soluble fractions

by the method of Jones and Redfearn (19). Reduced minus oxidized spectra

were determined on a Beckman model 25 recording spectrophotometer at

room temperature. Samples were reduced by the addition of a few grains

of sodium dithionite to a I ml quartz cuvette containing the fraction

to be assayed. Samples were oxidized by rapidly shaking the cuvette





8



before insertion into the spectrophotometer. Cytochrome b was

quantitated using a difference spectrum of 560-538 nm and an estimated

extinction coefficient of 17,300. Cytochrome c was quantitated using

a difference spectrum of 549-575 and an estimated extinction coefficient

of 17,500. Flavoprotein was quantitated using a difference spectrum of

465-510 and an extinction coefficient of 11,000. All values for

difference spectra and extinction coefficients are those of Jones and

Redfearn (19).

Potassium arabonate was prepared by the hypoiodite oxidation of

L-arabinose (26). DL-2-keto-3-deoxyarabonate was chemically synthesized

by the method of Stoolmiller and Abeles (37) and purified by the method

of Weimberg (40). L-Arabonolactone was produced by boiling potassium

arabonate in 0.2N HCI for 5 min.


Results


Growth and Nitrogen Fixation

Azospirillum brasiZense showed a doubling time of 16-20 h when

grown on either malate (Fig. 1) or L-arabinose (Fig. 2) at 0.5% 02'

Malate showed a 40% higher specific activity of acetylene reduction

than did L-arabinose grown cells (average of 3 experiments). The pH

of the medium dropped only slightly after 72 h growth on L-arabinose

(7.1-7.05) while pH rose sharply in malate grown cells (7.1 to 8.6).


Arabinose Pathway Enzymes

Table I gives the specific activities of L-arabinose dehydrogenase,

L-arabonolactonase, L-arabonate dehydratase, and the DL-KDA reduction

































Figure 1. Growth and acetylene reduction on L-malate
by Azospirillum brasilense. Each acetylene
reduction point average of 3 samples. Standard
deviation given by width of bar through point.
Optical density, 0 acetylene reduction







10










3000-


-1.0

r-I
I

0 2500- 0.9



E -0.8



2000-
it -0.7 0



0) 0.6
.4-I

1500- \
0.5 "



0 o.4 *H

1000- 0

0 0.3




500- -0.2


-0.1




10 20 30 40 50 60 70


time (hours)
































Figure 2. Growth and acetylene reduction on L-arabinose by
Azospirillum brasiZense. Each acetylene reduction
point average of 3 samples. Standard deviation
given by width of bar through point.
a Optical density, 0 acetylene reduction







12










3000-


-1.0



H 2500- 0.9

0

.0.8


r-H 2000-
I U .7 0






1500-
0.5 r

4"-'

^ / -0-4 ,
a) H.

S1000- J/

S.0.3




500-


.0.1




10 20 30 4 0 50 60 70


time (hours)






13









Table 1-1. L-arabinose pathway enzymes. Average of
3 crude extract preparations.


Activity
(nmol/min/
Enzyme Cofactor mg protein)


L-arabinose NAD 70
dehydrogenase NADP 30



Arabonolactonase 13



L-Arabonate dehydratase 17



DL-KDA oxidation NAD 62



L-KDA aldolase 0






14



of NAD in crude extracts of L-arabinose grown cells. Figure 3 shows

the reduction of NAD in acrude extract to which L-arabonate or DL-KDA

was added as substrate. DL-KDA exhibited 3 times the rate of NAD

reduction than did potassium L-arabonate. No KDA aldolase activity

could be found in any crude extract of L-arabinose grown cells.


End Product Accumulation

Table 2 indicates that crude extracts of L-arabinose grown cells

accumulate alpha-ketoglutaric acid when L-arabonate is added as substrate.

The largest accumulation is seen if sodium arsenite, an inhibitor of

alpha-ketoglutarate and pyruvate dehydrogenase activity (21, 41), is

added to the reaction mixture. No accumulation of either pyruvate or

glycolaldehyde has been found in crude extracts to which L-arabonate

had been added as substrate. This lent further credence to the likely

absence of L-KDA aldolase in these cells.


Identification of Reaction Products

At pH 6.5 a major product of the L-arabinose dehydrogenase

reaction is arabono-y-lactone as identified on paper chromatograms run

against authentic lactone. The sample comigrates with standard in all

3 solvent systems given in Methods and Materials. Even at this low pH,

however, L-arabonic acid appears as a significant part of the product.

At pH 9.0 the sample comigrates with L-arabonic acid in all three

solvent systems. The product of the L-arabonate dehydratase reaction

comigrates with DL-KDA in all three solvent systems. In addition the

2,4 dinitrophenylhydrazone of the reaction product comigrated with the

derivatized standard DL-KDA on Silica G gel plates in 1-butanol































Figure 3. Reduction of NAD in crude extracts of
L-arabinose grown cells with I pmole
potassium L-arabonate o; or DL-KDAB;
as substrate.







16











1.2-







1-0-






0
_ 0.8-







0.6

r-4
C.-
0

P,
0

0.4-







0.2








0.5 1.0 1.5 2.0 2.5

time (minutes)






17

















Table 1-2. Accumulation of alpha-ketoglutarate in crude
extracts of L-arabinose grown cells with
L-arabonate as substrate.



Alphaketoglutarate
Rxn. Mix Accumulation
Addition (pmoles)

Sodium arsenite 7.6


Sodium cyanide 4.2


none 4.4


control* 0.49


No arabonate was added to arabonate oxidation mix.
See Materials and Methods.






18



saturated with 3% NH4OH. On spraying with 0.5N NaOH the spot gave the

reddish brown color reported by Weimberg (40). In addition to enzymatic

assay alpha-ketoglutaric acid was identified chromatographically in

crude extracts to which potassium L-arabonate had been added as substrate

in the presence of NAD. The reaction product comigrated with alpha-

ketoglutarate (Sigma) in all three solvents. The 2,4 dinitrophenyl-

hydrazone likewise comigrated with the derivitized alpha-ketoglutarate

in 1-butanol saturated with 3% NH4OH.


TCA Cycle Dehydrogenases and Cytochrome Content

Table 3 shows the specific activity of isocitrate, malate, alpha-

ketoglutarate, and succinate dehydrogenases in crude extracts of L-

arabinose grown cells. Clearly cells grown with L-arabinose as substrate

possess an active TCA cycle for the oxidation of alpha-ketoglutarate.

Membrane and soluble fractions clearly contain b and c type

cytochrome and a probable cytochrome oxidase (Fig. 4-6). Cytochrome b

shows an alpha band at 556 nm which appears as a shoulder on the c

cytochrome alpha band at 549. The c cytochrome shows a beta band at

520 nm. Soret bands at 429 (cytochrome b) and 416 (cytochrome c) are

also seen. The cytochrome oxidase shows a broad peak around 600 nm.

The 105,000 xg supernatant contains only c type cytochrome (Table 4),

but c cytochrome is also found in the 35,000 xg and 105,000 xg pellets.

All the b cytochrome is found in the 35,000 xg and 105,000 xg pellets.

The 105,000 xg supernate also exhibited a deep trough at 451 nm (Fig.

4) which is probably a flavoprotein. This trough only appears in the

soluble fraction. A broad peak seen at 600 nm in the soluble fraction

and in the 105,000 xg pellet probably represents a cytochrome oxidase.






19








Table 1-3. TCA cycle dehydrogenases.



Activity
Electron (nmoles/min/mg
Enzyme Acceptor protein)



Malate dehydrogenase NAD 2,080


Isocitrate dehydrogenase NADP 230


Succinate dehydrogenase PMS DCPIP 150


Alpha-ketoglutarate
dehydrogenase Ferricyanide 540*

nmoles/hr/mg protein

































Figure 4. Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
105,000 xg supernatant.







21
















0.1
-0.15






-0.10







-0.05
0



0









0.05







-0.10



I I I I
400 500 600 700

wavelength (nm)




































Figure 5. Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
35,000 xg pel let.







23



















0.25







0.20







0.15
0



0
0g

0.10 Ct







0.05












400 500 600 700

wavelength (nm)


































Figure 6. Dithionite reduced minus oxidized spectra of membrane
and soluble fractions from L-arabinose grown cells.
105,000 xg pel let.







25









-0.15



0.1



-0.10











o



0


-0.05








0.10








II
400 500 600 700


wavelength (nm)





26















Table 1-4. Cytochrome content of membr3ne
and soluble fractions. pmoles/g
protein.


Cytochrome Cytochrome
c b Flavoprotein

105,000 xg
supernatant 0.64 0 0.79


105,000 xg
pellet 0.56 0.83 0


35,000 xg
supernatant 0.20 0.31 0






27



Discussion


Azospiriltum brasilense grew at approximately the same rate on

both malate and L-arabinose. Malate grown cells have shown a consistently

higher specific activity of acetylene reduction than have L-arabinose

grown cells, although both rates are high and compare closely with the

rate reported by Nelson and Knowles (27) for malate grown cells at 0.5%

02. The doubling time of 16-20 h is much slower than the 5.5 to 7 h

reported by Okon et al. (30) at 0.5% 02 (with malate as substrate), but

approximately the same as the 20 h generation time reported for

stagnantly grown cultures (28). It should be noted in my study, that

cells grown in malate or L-arabinose were cultured under identical

conditions. The sharp rise in pH of malate grown cells may have

accounted for a similar growth rate compared to that of L-arabinose

grown cells despite the higher acetylene reduction rates shown by

malate grown cells.

Initial attempts in our laboratory to demonstrate an L-arabinose

catabolic pathway involving phosphorylated intermediates were unsuccess-

ful. No L-arabinose isomerase or phosphotransferase activity could be

found. I have since found transketolase and transaldolase activity, in

crude and soluble fractions of L-arabinose grown cells, with about one-tenth

the specific activity of L-arabinose dehydrogenase. The means by

which L-arabinose might enter an oxidative pentose cycle is still

unknown. Some ability to metabolize phosphorylated compounds has

been indicated in 02 uptake studies (29).

In examining the possibility of a pathway without phosphorylated

intermediates I found high NAD(P) dependent L-arabinose dehydrogenase






28



activity. In addition L-arabono-y-lactonase, L-arabonate dehydratase.

and NAD dependent 2-keto-3-deoxyarabonate oxidation activity were found.

Alpha-ketoglutarate semialdehyde has been shown to be the substrate of

the NAD dependent dehydrogenation reaction (37). I did not look for

this intermediate nor did .1 assay for the L-KDA dehydratase activity,

which produces the alpha-ketoglutarate semialdehyde. The rapid reduction

of NAD with DL-KDA as substrate, which has also been reported by Weimberg

(40) in Pseudomonas saccharophila, would seem to be evidence of

significant L-KDA dehydratase activity. Weimberg and Doudoroff (42)

reported a more rapid reduction of NAD if crude extracts of Ps.

saccharophila were first preincubated with L-arabonate before the

addition of NAD. This preincubation probably allowed time for the

accumulation of L-KDA and perhaps alpha-ketoglutarate semialdehyde in

the reaction mixture.

An average of 75% of the potassium arabonate added to the crude

extract, in the presence of NAD, accumulated as alpha-ketoglutaric

acid if sodium arsenite was added to the crude extract. Sodium arsenite

is an inhibitor of alpha-ketoglutarate (41) and pyruvate dehydrogenases

(21), probably by inhibiting the decarboxylation step. Significant

amounts of alpha-ketoglutarate still were found in the absence of sodium

arsenite. Sodium cyanide had no effect on alpha-ketoglutarate accumu-

lation.

The presence of L-arabinose dehydrogenase, L-arabono-y-lactonase,

and L-arabonate dehydratase activity, along with the rapid reduction

of NAD in crude extracts with DL-KDA as substrate, and the accumulation

of alpha-ketoglutaric acid in crude extracts, is clear indication of the

following pathway of L-arabinose metabolism in A. brasilense.





29


NAD(P) +H20
L-arabinose > L-arabino-y-lactone H L-arabonic

-H2O -H20
acid -2 L-2-keto-3-deoxyarabonic acid -.20 > alpha-keto-

NAD
glutaric semialdehyde NAD alpha-ketoglutaric acid

Oxygen uptake studies (29) have indicated a very active TCA cycle

in A. brasilense, although the specific enzymes have not been previously

assayed. My assays of TCA cycle dehydrogenases would seem to confirm

high TCA cycle activity and these activities were comparable to that

shown by Rhizobium metiZoti (16). The significance of the very high

activity of malate dehydrogenase relative to the other dehydrogenases

is unknown.

The presence of b and c type cytochromes and a cytochrome oxidase

in deoxycholate extracts of Azospirillum brasilense has previously been

shown (4). For the first time the distribution of the cytochromes

between particulate and soluble fractions has been demonstrated. The

distribution is similar to that found in A. vinetandii (19). There also

appears to be a large flavoprotein factor. The possible pathway of

electron transport in A. brasilense may be the following. The X represents

a possible ubiquinone type electron carrier.

NADH Flavoprotein cyt.> X cyt

cyt cyt
c oxidase 2

It has been suggested by Eskew et al. (17) that the pink pigment

found in some strains of Azospirillum is due to a c type cytochrome.

The reduced minus oxidized spectrum published by Eskew et al. (17) of

the concentrated pigment is virtually identical to the spectrum I found

in the soluble fraction (105,000 xg supernatant) of the L-arabinose






30



grown cells. This, along with the significant amount of c cytochrome

I found in the soluble fraction, is further evidence that the pink

pigment produced by this strain of Azospirillum is in fact a c cytochrome.

We have demonstrated a totally oxidative pathway by which L-arabinose

is metabolized in Azospirillum brasilense. This ability may offer some

distinct advantages to this soil bacterium. Plant cell walls are rich

in compounds with the L-arabono-configuration (i.e. L-arabinose, D-

Fucose, D-galactose) (1). L-Arabinose is oneof the few isomers found in plant

cell walls. The cellulose fibers in the walls are cemented together by xylo-

glucans,arabinogalactans,and rhamnogalacturonan (1). It is possible that

root cell death and the sloughing off of the cells provides root associated

bacteria, likeAzospirillumwhich is able to metabolize these sugars, a

plentiful carbon substrate. The advantage provided to these bacteria in

the highly competitive soil environment is obvious.

Another benefit of this pathway is that there is a ready pool of alpha-

ketoglutarate which can act as substrate for glutamate dehydrogenase and

glutamate synthase activity (29). Both enzymes are important in nitrogen

assimilation under fixed nitrogen and dinitrogen fixing conditions,

respectively. Finally there is the increasing similarities demonstrated

between Azospirillum brasilense and some Rhizobium species. Included in

these is the oxygen sensitivity of their nitrogenase (20,30), high dissimi-

latory nitrate reductase activity (34, 38, accumulation of PHB granules

in cells (3, 29), the enhancementof nitrogenaseactivityby inclusionof a

pentosewith the organic acid substrate (6), and similar GC content (3, 38).

The L-arabinose pathway described in this paper is one that had previously

been found only in some species of Pseudomonas (7, 41, 42) and fast growing

rhizobia (15, 10. The presence of this pathway is another indication of

a close relationship between Azospirillum and Rhizobiumn species.














SECTION II
PURIFICATION AND CHARACTERIZATION OF L-ARABINOSE
DEHYDROGENASE AND L-ARABONATE DEHYDRATASE FROM
L-ARABINOSE GROWN CELLS OF AZOSPIRILLUM BRASILENSE



Introduction


Azospirillum brasilense (ATCC 29145) can metabolize L-arabinose

by the following series of reactions

NAD(P) +H20
L-arabinose -NAD(P) L-arabono-y-lactone -+H2 L-arabonic
-H2 -H00
acid L-2-keto-3 deoxyarabonic acid -2 alpha-keto-

NAD
glutaric semialdehyde -NAD alpha-ketoglutaric acid

This pathway has also been demonstrated in Pseudomonas saccharophila

(42), Pseudomonas fragi (7, 41), and the fast growing rhizobia (15, 16)

R. meliloti, R. trifolii, R. phaeseoli, and R. leguminosarum. A second

pathway which carries through the first 3 steps of the alpha-keto-

glutarate pathway but cleaves L-2-keto-3-deoxyarabonate (L-KDA) by an

aldolase reaction to glycolaldehyde and pyruvate has been found in

Pseudomonas strain MSU-1 (8) and the slow growing rhizobia; R. japonicum

(15, 31) and Rhizobium sp. 32H] (15).

The enzymes of these pathways have not been extensively studied.

L-arabinose dehydrogenase and the L-arabonate oxidation system (L-

arabonate alpha-ketoglutaric acid) have been partially purified

in Pseudomonas saccharophila (42). L-2-keto-3-deoxyarabonate dehydratase

has been purified and characterized by StoolmillerandAbeles (37). Dahms

and Anderson have partially purified L-KDA aldolase from Ps. MSU-I (8).


31






32



D-Fucose metabolism in Ps. MSU-I is very similar to that found for

L-arabinose. In the D-fucose pathway, D-2-keto-3-deoxyfuconate is

cleaved to pyruvate and lactaldehyde. All the enzymes in this pathway

have been purified and characterized by Dahms and Anderson (9-12).

I have partially purified and characterized L-arabinose dehy-

drogenase and L-arabonate dehydratase from L-arabinose grown cells

of Azospirillum brasiZense.


Materials and Methods

Bacterial Strain

Azospirillum brasilense (ATCC 29145) was obtained from the American

Type Culture Collection. For daily use cultures were maintained on

Tryticase soy agar (BBL) slants with biweekly transfer. Stock cultures

were frozen in liquid nitrogen.

Enzyme Purification

Crude extracts were prepared as stated in Section 1. All steps in

the purification of enzymes were carried out at 40C unless otherwise

stated.

Nucleic acid precipitation

A 2% solution of protamine sulfate (Sigma, Grade II) in O.IM

sodium-potassium phosphate buffer, pH 7.5 was slowly added to the crude

extract to give a final concentration of 0.33%, stirred in the cold for

30 min., and centrifuged at 20,000 xg for I hr. The pellet was dis-

carded.






33



Ammonium sulfate fractionation

Solid ammonium sulfate (Sigma, Grade III) was slowly added to the

protamine sulfate supernatant while maintaining the pH at 7.5 with 0.1N

NaOH. The sample was stirred in the cold for I hr then centrifuged at

20,000 xg for I hr. The pellets were resuspended in 0.1M sodium-

potassium phosphate buffer pH 7.5. Ninety-two percent of the dehydratase

activity was found in the 30% and 40% ammonium sulfate precipitate

fraction. These fractions were pooled before being placed on Biogel A

column. Eighty-nine percent of the dehydrogenase was found in the 50%

ammonium sulfate precipitate fraction. This fraction was carried through

a heat treatment step before being loaded onto the Biogel column.

Heat treatment Oehydrogenase only)

The resuspended 50% ammonium sulfate fraction was heated to 55C

for 5 min, immediately cooled in ice to VC, and centrifuged at 20,000

xg for I hr.

Gel filtration

Sample was added to a 2 x 150 cm Biogel A 1.5M column equilibrated

with 25 mM sodium-potassium phosphate buffer pH 7.4. Flow rate was

25 ml per hr and 5 to 20 ml fractions were collected.

Ion exchange chromatography

A 2 x 10 cm DEAE cellulose (fine) column was equilibrated with

3 column volumes of 25 mM sodium-potassium phosphate buffer, pH 7.4.

Samples were either placed directly on the column from active, pooled

Biogel A fractions or Biogel fractions were concentrated first in an






34



Amicon pressure cell with UM30 filter. Enzyme was eluted in a 0 to 0.2M

(dehydrogenase) or 0 to 0.3M (dehydratase) 300 ml linear NaCl gradient.

All NaCl solutions were prepared in 25 mM sodium-potassium phosphate

buffer, pH 7.4. Active DEAE fractions were immediately desalted on a

G-25 column, equilibrated with 25 mM sodium-potassium phosphate buffer,

pH 7.4. Flow rate was 50 ml per hr and 20 ml fractions were collected.

Active fractions were concentrated by lyophilization.


Protein Determination

Protein was determined by the method of Lowry et al. (24) with

BSA as a standard.


Enzyme Assays

L-arabinose dehydrogenase was assayed in 125 pmoles glycine/NaOH

buffer, pH 9.0, 0.5 pmoles NAD, 0.5 pmoles L-arabinose, and enzyme plus

water to I ml. In crude extracts 1 pmole NaCN was added to the reaction

mix. The change in optical density at 340 nm was followed on a Zeiss

spectrophotometer at 25C. One unit of L-arabinose dehydrogenase activity

was the amount of enzyme which reduced I pmole NAD(P) per min. Arabonate

dehydratase activity was assayed by the method of Pedrosa and Zancan (31)

with 2-keto-3-deoxyarabonate accumulation being measured with semi-

carbazide reagent (25). One unit of activity was the amount of enzyme

which produces 1 pmole of L-KDA per min.


Molecular Weight Determination

The 2 x 150 cm Biogel A column was equilibrated with

25mMsodium-potassium phosphate buffer, pH 7.4, and 2 mg of

each standard (Boehringer Calibration Proteins 11), including






35



catalase (240,000 daltons), aldolase (158,000), albumin (68,000), and

albumin (45,000), were dissolved in 20 ml of buffer and placed on the

column. The column was run at 25 ml per hour and 7.5 ml fractions were

collected. Standard peaks were located by A280. L-Arabinose dehy-

drogenase and L-arabonate dehydratase were located by enzyme activity.

A plot of the molecular weight of the standards versus the logarithm

of the volume of the half-height of the leading edge of each compound

was made. The standards generated a straight line from which unknown

molecular weights were determined.


Product Identification

Products of L-arabinose dehydrogenase and L-arabonate dehydratase

were identified as previously stated in Section I.


Polyacrylamide Gel Electrophoresis

Slab gels were prepared by the method of Laemmli (22). A 5-15%

exponential gel was cooled to 4C and run at 32 ma. Enzyme activity was

located by slicing the gel in 8 mm sections, placing the sections in

13 x 100 mm tubes, and eluting the enzyme in 1 ml of 25 mM potassium

phosphate buffer, pH 7.2. Enzyme activity was assayed in the standard

manner.


Results


The purification of L-arabinose dehydrogenase and L-arabonate

dehydratase are outlined in Tables I and 2, respectively. The dehy-

drogenase was purified 59 fold with 1.2% yield and L-arabonate dehydra-

tase was purified 38 fold with 9% yield. Attempts were made to improve






36







Table 2-1. Purification of L-arabinose dehydrogenase.


Sp. Act. Total
Protein Units/mg Activity Purifi- Recovery
Fraction (mg) Protein Units cation %

Crude 1484 0.07 103.8

Protamine
sulfate
supernatant 1280 0.07 89.6 0 86

Ammonium
sulfate
(50% prec.) 733 0.08 58.6 1.1 56

Heat
treatment 288 0.20 57.6 2.9 55

Biogel A
1.5M
(fractions
pooled) 20.6 1.43 29.4 20.4 28

DEAE
G-25 washed
conc. O1X 0.3 4.12 1.2 58.9 1.2






37









Table 2-2. Purification of L-arabonate dehydratase.


Sp. Act. Total
Protein Units/mg Activity Purifi- Recovery
Fraction (mg) Protein Units cation %

Crude 2497 0.011 27.5

Protami ne
sulfate
supernatant 1783 0.015 26.7 1.4 97

Ammonium
sulfate prec.
(30 & 40%
fractions) 635 0.007 4.4 16

Biogel A
1.5M
(pooled frac-
tions) 19 0.19 3.6 17.2 13

DEAE
G-25 washed 6 0.42 2.5 38.2 9






38



the purification and yield of L-arabinose dehydrogenase. Calcium

phosphate gels, affinity chromatography (Affi-gel Blue), and DEAE

Sephadex ion exchange columns failed to improve either of the above

factors. The elution profiles of L-arabinose dehydrogenase and L-

arabonate dehydratase activities from a Biogel A 1.5M column are shown

in Fig. 7. It is clear that only partial resolution of these two

enzymes was achieved by gel filtration. Therefore it was necessary

to further chromatograph the Biogel fractions on DEAE cellulose eluted

with a linear NaCl gradient. Figure 8 demonstrates preliminary results

obtained from crude 30% and 40% ammonium sulfate fractions. Excellent

resolution of the two enzymes was achieved.

Dehydrogenase activity started to elute from the DEAE column at

0.08M NaCl while the dehydratase did not start to elute from the column

until 0.19M NaCl was applied. The dehydrogenase came off in unstable

condition and it was necessary to desalt the eluate immediately on a

G-25 column. Thus ion exchange chromatography allowed excellent

separation of enzymes from each other.


Characteristics of L-Arabinose Dehydrogenase

L-Arabinose dehydrogenase showed no enhancement of activity in the

presence of any of the divalent cations listed in Table 3. The presence

of CaCl2, FeSO4, MnCl2 and CoCl2 severely inhibited enzyme activity. The

presence of EDTA has a slight negative effect on activity and the presence of

the reducing agents 2-mercaptoethanol, glutathione (reduced), and dithio-

threitol also had a small negative effect on activity.

Only those substrates with the L-arabono configuration (i.e. D-

galactose, D-fucose) were good substrates for the L-arabinose































Figure 7. Elution from a Biogel A 1.5M (2 x 150 cm) column
of L-arabinose dehydrogenase and L-arabonate de-
hydratase. The resuspended 30% and 40% ammonium
sulfate precipitate fractions were pooled and
placed directly on the column. Assays are as
stated in Materials and Methods. Fractions (20 ml)
were collected and protein determined by A280.












absorbance (o dehydratase, 250 nm)

0 t I I I











C+
o


















absorbance (e dehydrogenase, 340 nm)
(a protein, 280 nm)































Figure 8. Elution from a DEAE cellulose column (2 x 10 cm) of
L-arabinose dehydrogenase and L-arabonate dehydratase.
The resuspended 30% and 40% ammonium sulfate pre-
cipitate fractions were pooled and placed on the
column. Enzyme was eluted from the column with a
300 ml,0 to 0.3M NaCl linear gradient. Fractions
(5 ml) were collected and protein determined by A280.
Enzyme assays were as in Materials and Methods.







42











r 1.6-
0
N -0.3


1.4-





1.2-

O



0 /

- 0.8- 0.2


*-Hl 0.8-
+> H r-t
- 0o


0.6-


S I -0.1


,B 0.4-
0



0.2-






2 .4 6 8 10 12 14 16 18 20 22 24 26 28 30


fraction number






43





Table 2-3. Effect of divalent cations and reducing
agents on L-arabinose dehydrogenase
activity.


Relative
Activity
Reagent Concentration %

None 100

MgSO4 10 mM 100

NH4SO4 10 mM 100

CaCl2 10 mM 70

FeSO4 10 mM 0

MnCl2 10 mM 55

CoCI2 10 mM 46

EDTA 2.5 mM 92

2-mercapto-
ethanol I mM 95

Glutathione
(reduced) 1 mM 93

Dithiothreitol I mM 90






44



dehydrogenase (Table 4). Lineweaver-Burk plots (Fig. 9) show a K value

of 140 yM with 0.5 pmole NAD as cofactor and 72 pM with 0.5 pmole NADP

as cofactor. The heat stable dehydrogenase (no loss of activity when

heated to 55C for 5 min) had a pH optimum of 9.5 in glycine/NaOH

buffer (Fig. 10)

Characteristics of L-arabonate Dehydratase

L-Arabonate showed a pH optimum of 7.8-8.0 in Hepes/NaOH buffer

(Fig. 11). Magnesium sulfate and manganese chloride (Table 5) enhanced

enzyme activity 35% and 56%, respectively. The presence of ZnCl2, CaCl2,

FeSO 4, and CoCl2 severely inhibited activity. The addition of EDTA

(2.5 mM) resulted in a 95% inhibition of activity. Reducing agents 2-

mercaptoethanol and glutathione (reduced) had a small negative effect

on activity while dithiothreitol greatly reduced activity.


Molecular Weight Determination

L-Arabinose dehydrogenase had a molecular weight of 175,000 and

L-arabonate dehydratase had a molecular weight of 225,000 according to

gel filtration determinations (Fig. 12).


Enzyme Product Identification

At pH 6.6 the product of the L-arabinose dehydrogenase reaction

was found to be L-arabono-y-lactone, although, L-arabonic acid was also

found in the reaction mixture. At pH 9.0 L-arabonic acid was the only

product found. The product of the dehydratase reaction was identified

as L-2-keto-3-deoxyarabonate. See Section I for a complete description

of enzyme product identification.






45











Table 2-4. Substrate specificity of L-arabinose
dehydrogenase. The standard assay was
used except substrate was varied.




Relative
Substrate Velocity
10 mM %

L-arabinose 100


D-ribose 0


D-galactose 63


D-xylose 4


D-fucose 100


L-fucose 0


L-rhamnose 0


D-mannose 0


D-glucose 0
































Figure 9. Lineweaver-Burk plot showing the effect of substrate
concentration on reaction velocity. NAD o and NADPm
(both 10 mM) were used as cofactors. The standard
assay was used except substrate concentration was
varied.







47






















8

100-




80





.* 60-
0




40-





20-






2 4 6 8 10 12 14 16 18 20

substrate conc. (mM)-

































Figure 10. Effect of pH on reaction velocity of L-arabinose
dehydrogenase. Standard assay was used except pH
of the glycine/NaOH buffer was varied.







49




















100-





80-





* -1 60-
.r-I
4--l
0


0 40-





20






7 8 9 10 11

pH































Figure II. Effect of pH on reaction velocity of L-arabonate
dehydratase. Standard assay was used except pH
of the HEPES/NaOH buffer was varied.







51






100 -










80-










60-



*Hp

cd

4 -

0)
p, 40-










20











6 7 8 9


pH






52





Table 2-5. Effect of divalent cations and reducing
agents on L-arabonate dehydratase
activity.


Relative
Concentration Activity
Reagent mM %

None 100

ZnCl2 1 34

MgSO4 10 135

(NH )2SO4 10 100

CaCl2 10 55

FeSO4 10 55

MnCl2 10 156

CoCl2 10 11

EDTA 2.5 5

2-mercapto-
ethanol I 93

Glutathione
(reduced) 1 92

Dithiothreitol 1 41


































Figure 12. Molecular weight determination of L-arabinose
dehydrogenase and L-arabonate dehydratase by gel
filtration technique (see Materials and Methods).






54










1000-





500-




catalase (240,000)
0 dehydratase

0
r- dehydrogenase
x aldolaseo
S(158,000)

100-


C3d

C (68,000)

0 50 albumin
(45,000)











10-
0.1 0.2 0.3 0.4 0.5 0.6 0.7
kav






55



Polyacrylamide Gel Electrophoresis

Gels of purified dehydrogenase and dehydratase still showed several

bands indicating that the enzymes were not totally pure. Assay of

dehydrogenase activity in the gel showed the enzyme activity to be

located 60 mm from the top of the gel (Fig. 13). Two dark bands, 41 mm

and 53.5 mm from the top of the gel (Fig. 14), appeared in the gels run

with partially purified dehydratase. Dehydratase activity could not be

located within the gel. It is possible that the dehydratase dis-

associated into two subunits which could not individually catalyze the

dehydration reaction or that the enzyme was inactivated in the gel with-

out disassociation.



Discussion


L-Arabinose dehydrogenase from Azospirillum brasilense apparently

metabolizes only substrates having the L-arabono configuration. Other

aldose dehydrogenases are generally not specific for this configuration

(9). According to K values the dehydrogenase had a greater affinity

for substrate with NADP as cofactor than with NAD, although, V was
max
twice as great with NAD as cofactor than with NADP. The enzyme had no

requirements for any of the divalent cations tested nor was there enhance-

ment of activity in the presence of reducing agent. The enzyme had a

high pH optimum of 9.5.

L-Arabinose dehydrogenase from A. brasilense shows several simi-

larities todehydrogenases from other bacteria metabolizing L-Arabinose

through similar pathways. L-Arabinose dehydrogenase from R. japonicum




































Figure 13. Polyacrylamide slab gel of crude and partially
purified extracts of L-arabinose dehydrogenase.
Lane A and B; protaminesulfate supernatant
fraction (50 pg); Lanes C and D; DEAE-G-25 washed
fraction (50 pg). Arrow points to area of
enzyme activity.














r



1
Lr%


















n---------



N& W


































Figure 14. Polyacryamide slab gel of partially purified extracts
of L-arabonate dehydratase. Lanes A and B contain
DEAE-G-25 washed fraction (50 1g).















































-J






















U






60



(31) and D-fucose dehydrogenase from Ps. MSU-l (9) are both capable of

reducing NAD and NADP. L-arabinose dehydrogenase from Ps. saccharophila

is reported only capable of reducing NAD (42). L-arabinose dehydrogenase

from Ps. saccharophila (42); R. japonicum (31), and Aquaspirillum

graciZe (23) are all assayed at high pH, although, optimum pH is not

reported. D-Fucose dehydrogenase from Pseudomonas MSU-1 is similar to

L-arabinose dehydrogenase from A. brasilense in its lack of divalent

cation requirement, lack of reducing agent requirement, heat stability,

and pH optimum as well as substrate specificity (9).

The immediate product of L-arabinose dehydrogenasewas L-arabono-y-

lactone, whichwas cleaved to L-arabonic acid by a lactonase. Even at

pH of 6.6, though, there was considerable spontaneous breakdown of the

lactone.

L-arabonate dehydratase showed greatly enhanced activity in the

presence of MgSO4 or MnCl2. Its requirement for cations was clearly

demonstrated by the near total loss of activity in the presence of the

chelating agent EDTA. This was similar to that found for D-fuconate

dehydratase (11). Unlike D-fuconate dehydratase this enzyme showed no

enhancement of activity in the presence of any of the reducing agents

tested. In fact reducing agents had a negative effect on activity. The

semicarbazide positive product of the dehydratase reaction has been

identified as L-2-keto-3-deoxyarabonate.














APPENDIX
DESIGN OF OXYGEN CONTROLLER



A simple oxygen controller has been built for use in growth and

nitrogen fixation studies. The sensing part of the instrument consists

of a Borkowski and Johnson (2) oxygen probe. The probe acts as a tiny

battery with electrons moving from the pure lead cathode to a pure silver

anode covered by a teflon membrane. The amount of current running be-

tween cathode and anode is proportional to the rate at which oxygen

is reduced at the anode. A resistor placed between the anode and

cathode produces a potential drop which drives the sensing system. The

whole system is outlined in Figure 1. A 741 operational amplifier

allows the probe signal to be amplified up to 100 times. This signal

is then fed to a 50 mv potentiometer (as part of the LFE compact con-

troller). The system is calibrated by bubbling pure nitrogen (Matheson)

through the sterile medium and setting the mv meter to zero with the

741 zero adjust. Two percent oxygen (Matheson) is then bubbled through

the sterile medium and full scale is set with the amplification adjust.

The heart of the controller part of the instrument is an LFE

Compact Controller with high and low relays. If the mv indicator drops be-

low the low relay setting, the relay closes; likewise, if it rises above

the high relay setting, the high relay closes. The motor (115 Vac, I

rpm) which drives the micrometering valve has a common, clockwise, and

counterclockwise control wire. If the common and clockwise wires are


61






































Figure 15. Schematic of oxygen controller.






63


1 Meg

+12
Oxygen + 10 K 2 Amplification
Probe 7 Adjust
741 6 LFE
10 K Compact
5 Controller
3+
To Growth c
Chamber 50 mV
Chambe -12 Potentiometer
9.9 K 10 K
Nitrogen
Zero
Adjust
-12


Air



Low High LFE
Relay Relay Compact
Controller
4 Relays
Motor

|I rpm
115 Vac
Nupro
Micrometering
Valve


+ 9V



RI
IM 4 14 10 R3
1 75K

R2
2 12 115 Vac
6 556 13 C2+tD I
+ C2+ DI
C" 5 9 33 F IN914
100uF 8 11.
_C C3 D2 4 '



Radio Shock
275-004






64



connected to 115 Vac the motor turns clockwise, if the common and counter-

clockwise wires are connected to 115 Vac the motor runs counterclockwise.

The common wire runs from the motor through a relay (connected to a

556 timer) to line voltage. The clockwise wire runs through the con-

troller high relay and the counterclockwise wire runs through the compact

controller low relay. The high and low relay wires come together to

form the other line for 115 Vac connection.

In use the 556 timer is set for I second on time and I minute delay

time (time between on periods). The common line therefore closes for

I second every minute. The high relay is set at approximately 14.5

and the low relay at 10.5. The mv indicator starts out at 12.5 on

the 50 my scale (0.5% 02). Every minute the instrument "checks" to

see if either the high relay or low relay is closed. If, for instance,

the high relay is closed, the motor will move approximately one-sixtieth

of a turn (1 second on a 1 rpmmotor) counterclockwise closing the valve

a fraction of a turn and decreasing 02 concentration in solution. This

method manages to keep the 02 concentration at 0.5% + 0.05%. Typically

with A. brasiZense 02 levels being pumped into the growth chamber will

rise from 0.5% at the beginning of growth to 5.0% 02 by late exponential

phase (in order to keep dissolved 02 at 0.5%).














LITERATURE CITED



1. Albersheim, P. 1975. The walls of growing plant cells. Sci.
Am. 232: 81-94.

2. Borkowski, J. D., and M. J. Johnson. 1967. Long-lived steam-
sterilizable membrane probes for dissolved oxygen measurement.
Biotech. and Bioeng. 9: 635-639.

3. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey's Manual
of Determinative Bacteriology, 8th ed. The Williams and Wilkins
Co., Baltimore.

4. Burris, R. H., Y. Okon, and S. L. Albrecht. 1978. Properties and
reactions of Spirillum lipoferum. Environmental role of nitrogen-
fixing blue-green algae and asymbiotic bacteria. Ecol. Bull.
(Stockholm) 26: 353-363.

5. Cavallini, D., N. Frontali, and G. Toschi. 1949. Keto acid content
of human blood and urine. Nature 164: 792.

6. Child, J. J. and W. G. W. Kurz. 1978. Inducing effect of plant
cells on nitrogenase activity of Spirillum and Rhizobiwn in vitro.
Can. J. Microbiol. 24: 143-148.

7. Dagley, S., and P. W. Trudgill. 1965. The metabolism of
galactarate, D-glucarate and various pentoses by species of
Pseudomonas. Biochem. J. 95: 48-58.

8. Dahms, A. S., and R. L. Anderson. 1969. 2-keto-3-deoxy-L-arabonate
aldolase and its role in a new pathway of L-arabinose degradation.
Biochem. Biophys. Res. Commun. 36: 809-814.

9. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a
pseudomonad. II. Oxidation of D-fucose to D-fucono-y-lacone by an
L-arabino-aldose dehydrogenase and the hydrolysis of the lactone by
a lactonase. J. Biol. Chem. 247: 2228-2232.

10. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a
pseudomonad. I. Oxidation of D-fucose to D-fucono-y-lacone by a
D-aldohexose dehydrogenase. J. Biol. Chem. 247: 2222-2227.

11. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a
pseudomonad. III. Conversion of D-fuconate to 2-keto-3-deoxy-D-
fuconate by a dehydratase. J. Biol. Chem. 247: 2233-2237.

65





66



12. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a
pseudomonad. IV. Cleavage of 2-keto-3-deoxy-D-fuconate to pyruvate
and lactaldehyde by a 2-keto-3-deoxy-L-arabonate aldolase. J. Biol.
Chem. 247: 2238-2241.

13. Day, J. M., and J. Dobereiner. 1976. Physiological aspects of
N2-fixation by a Spirillum from Digitaria roots. Soil Biol.
Biochem. 8: 45-50.

14. Dische, Z., and E. Borenfreund. 1949. A specific color reaction
of glycolic aldehyde. J. Biol. Chem. 180: 1297-1300.

15. Duncan, M. J. 1979. L-Arabinose metabolism in rhizobia. J. Gen.
Microbiol. 113: 177-179.

16. Duncan, M. J., and D. G. Fraenkel. 1979. Alpha-ketoglutarate
dehydrogenase mutant of Rhizobium meliZotti. J. Bacteriol. 37:
415-419.

17. Eskew, D. L., D. D. Focht, and I. P. Ting. 1977. Nitrogen fixation,
denitrification, and pleomorphic growth of a highly pigmented
Spirillum lipoferum. Appl. Environ. Microbiol. 34: 582-585.

18. Hestrin, S. 1949. The reaction of acetylcholine and other car-
boxylic acid derivatives with hydroxylamine and its analytical
application. J. Biol. Chem. 180: 249-261.

19. Jones, C. W., and E. R. Redfearn. 1966. Electron transport in
Azotobacter vinelandii. Biochem. Biophys. Acta. 113: 467-481.

20. Keister, D. L., and W. R. Evans. 1976. Oxygen requirements for
acetylene reduction in pure cultures of rhizobia. J. Bacteriol.
129: 149-153.

21. Kersters, K., and J. De Ley. 1971. Tests with cell free extracts.
p. 46-51. In J. R. Norris and D. W. Ribbons (ed.) Methods in
Microbiology. v. 6A. Academic Press, New York.

22. Laemmli, U. K. 1970. Cleavage of structural proteins during the
assembly of the head of Bacteriophage T4. Nature 222: 680-685.

23. Laughon, B. E., and N. R. Krieg. 1974. Sugar catabolism in
Aquaspirillum gracile. J. Bacteriol. 119: 691-697.

24. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.
1951. Protein measurement with the folin phenol reagent. J.
Biol. Chem. 193: 265-275.

25. MacGee, J., and M. Doudoroff. 1954. A new phosphorylated inter-
mediate in glucose oxidation. J. Biol. Chem. 210: 617-626.






67



26. Moore, S., and K. P. Link. 1940. Carbohydrate characterization.
I. The oxidation of aldoses by hypoiodite in methanol. J. Biol.
Chem. 133: 293-311.

27. Nelson, L. M., and R. Knowles. 1978. Effect of oxygen and nitrate
on nitrogen fixation and denitrification by Azospirillum brasiZense
grown in continuous culture. Can. J. Microbiol. 24: 1395-1403.

28. Okon, Y., S. L. Albrecht, and R. H. Burris. 1976. Factors affect-
ing the growth and nitrogen fixation of Spirillum lipoferum. J.
Bacteriol. 127: 1248-1254.

29. Okon, Y., S. L. Albrecht, and R. H. Burris. 1976. Carbon and
ammonia metabolism in Spirillum lipoferum. J. Bacteriol. 128:
592-597.

30. Okon, Y., J. P. Houchlins, S. L. Albrecht, and R. H. Burris. 1977.
Growth of Spirillum lipoferum at constant partial pressures of oxygen
and the properties of nitrogenase in cell free extracts. J. Gen.
Microbiol. 98: 87-93.

31. Pedrosa, F. 0., and G. T. Zancan. 1974. L-Arabinose metabolism in
Rhizobium japonicum. J. Bacteriol. 119: 336-338.

32. Reed, L. J. and C. R. Williams. 1966. Purification and resolution
of the pyruvate dehydrogenase complex. p. 258-259.
In W. A. Wood (ed.) Methods in Enzymology. V. 9. Academic Press,
New York.

33. Reeves, H. C., R. Rabin, W. S. Wegener, and S. J. Ajl. 1971.
Assays of enzymes of the tricarboxylic acid and glyoxylate cycles.
p. 425-462. In J. R. Norris and D. W. Ribbons (ed.) Methods in
Microbiology. V.6A. Academic Press, New York.

34. Rigaud, J., F. J. Bergersen, G. L. Turner, and R. M. Daniel. 1973.
Nitrate dependent anaerobic acetylene reduction and nitrogen fix-
ation by Soybean bacteroids. J. Gen. Microbiol. 77: 137-144.

35. Singer, T. P., and T. Cremona. 1966. D-(-)-Lactate cytochrome c
reductase. p. 303-305. In W. A. Wood (ed.), Methods in Enzymology
V. 9. Academic Press, New York.

36. Smith, I (Ed.) 1960. Chromatographic and Electrophoretic Techniques.
I. Chromatography. Interscience Publishers, Inc., New York.

37. Stoolmiller, A. C., and R. H. Abeles. 1966. Formation of alpha-
ketoglutaric semialdehyde from L-2-keto-3-deoxy-arabonic acid and
the isolation of L-2-keto-3-deoxy-arabonate dehydratase from
Pseudomonas saccharophila. J. Biol. Chem. 241: 5764-5771.

38. Tarrand, J. J., and N. R. Krieg. 1978. A taxonomic study of Spirillum
lipoferum group,with descriptions of a new genus, Azospirillum gen.
nov. and two species, Azospirillum lipoferum (Beijerinck) comb. nov.
and Azospirillum brasilensesp. nov. Can. J. Microbiol. 24: 967-980.






68



39. Veeger, C., D. V. DerVartanian, and W. P. Zeylemaker. 1969.
Succinate dehydrogenase. p. 81-90. In J. M. Lowenstein (ed.),
Methods in Enzymology. V. 13. Academic Press, New York.

40. Weimberg, R. 1959. L-2-keto-4,5 dihydroxyvaleric acid: An
intermediate in the oxidation of L-arabinose by Pseudomonas
saccharophita. J. Biol. Chem. 234: 5764-5771.

41. Weimberg, R. 1961. Pentose oxidation by Pseudomonas fragi.
J. Biol. Chem. 236: 629-634.

42. Weimberg, R., and M. Doudoroff. 1955. The oxidation of L-
arabinose by Pseudomonas saccharophila. J. Biol. Chem. 217:
607-624.

43. Weissbach, A., and J. Hurvitz. 1959. The formation of 2-keto-
3-deoxyheptonic acid in extracts of Escherichia coli B. J.
Biol. Chem. 234: 705-709.














BIOGRAPHICAL SKETCH



Norman James Novick was born in Portland, Maine,on March 12,

1953. He was graduated from Portland High School in June, 1971, and

attended Rensselaer Polytechnic Institute from September, 1971, to

June, 1975. He received the degree of Bachelor of Science from RPI

with a major in biology. After working for the Rensselaer Fresh Water

Institute from June to August, 1975, he entered graduate school at the

University of Florida in September, 1975. Norman received the degree

of Master of Science from the Department of Microbiology and Cell

Science in March, 1977. Since that date he has been pursuing a study

of the nitrogen-fixing bacterium Azospirillum brasilense and is a

candidate for the degree of Doctor of Philosophy in December, 1980.


























69










I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.




Arnold S. Bleiweis, Chairman
Professor of Microbiology and Cell
Science



I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.




/dames F. Preston
Associate Professor of Microbiology
and Cell Science


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.




David H. #bbefl \"
Professor of Soil Science



I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.




Stanley C. Schank
Professor of Agronomy










This dissertation was submitted to the Graduate Faculty of the College
of Agriculture and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.

December, 1980




Dean/ ol lege of AgaiRul ture



Dean for Graduate Studies and Research




Full Text

PAGE 1

METABOLISM OF L-ARABINOSE IN AZOSPIRILLUM BRASILENSE BY NORMAN JAMES NOVICK A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1980

PAGE 2

ACKNOWLEDGEMENTS The author wishes to thank Dr. Max E. Tyler for his advice and guidance and wishes him the very fastest recovery from his present illness. The author also wishes to thank Dr. Arnold S. Bleiweis for his aid and advice under difficult circumstances and Dr. James F. Preston for his technical assistance and suggestions in the preparation of this manuscript. The author would also like to thank Dr. David H. Hubbell and Dr. Stanley C. Schank for their suggestions in the preparation of this manuscript. My wife, Connie, has been a constant companion and advisor during the time of my research and in the preparation of this manuscript. I want to thank her for her love and guidance. i i

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGEMENTS LIST OF TABLES. LIST OF FIGURES ABSTRACT. SECTION I IDENTIFICATION OF ENZYMES INVOLVED IN L-ARABINOSE METABOLISM IN AZOSPIHILLUM BEASILENSE. ASSAY OF TCA CYCLE DEHYDROGENASES AND IDENTIFICATION OF COMPONENTS INVOLVED IN ELECTRON TRANSPORT 1 Introduction ] Materials and Methods 3 Bacterial Strain 3 Media I 3 Growth and Nitrogen Fixation. 3 Acetylene Reduction Assay k Determination of Arabinose Pathway Enzymes k Identification of Pathway End Product 5 Chromatography 6 TCA Cycle Enzymes 7 Cytochrome Determinations 7 Results 8 Growth and Nitrogen Fixation 8 Arabinose Pathway Enzymes 8 End Product Accumulation ]li Identification of Reaction Products ]k TCA Cycle Dehydrogenases and Cytochrome Content ... 18 Discussion 27 It PURIFICATION AND CHARACTERIZATION OF L-ARABINOSE DEHYDROGENASE AND L-ARABONATE DEHYDRATASE FROM L-ARABINOSE GROWN CELLS OF AZOSPIHILLUM BRASILENSE .... 31 Introduction 3j Materials and Methods 32 Bacterial Strain 32 Enzyme Purification 32 Protein Determination 3i4 Enzyme Assays 3/j Molecular Weight Determination 3Zf i i i

PAGE 4

Results 35 Characteristics of L-arabinose Dehydrogenase 38 Characteristics of L-arabonate Dehydratase hk Molecular Weight Determination kh Enzyme Product Identification 44 Polyacrylamide Gel Electrophoresis 55 Discussion 55 APPENDIX 61 LITERATURE CITED 65 BIOGRAPHICAL SKETCH 69 i V

PAGE 5

LIST OF TABLES Table 1-1 L-arabinose pathway enzymes 13 1-2 Accumulation of al pha-ketogl utarate in crude extracts of L-arabinose grown cells; with L-arabonate as substrate 17 1-3 TCA cycle dehydrogenases 19 1h Cytochrome content of membrane and soluble fractions ... 26 21 Purification of L-arabinose dehydrogenase 36 2-2 Purification of L-arabonate dehydratase 37 2-3 Effect of divalent cations and reducing agents on L-arabinose dehydrogenase activity 43 2-4 Substrate specificity of L-arabinose dehydrogenase .... 45 2-5 Effect of divalent cations and reducing agents on L-arabonate dehydratase activity 52 V

PAGE 6

LIST OF FIGURES Figure 1 Growth and acetylene reduction on L-malate by Azospivillion brasilense 10 2 Growth and acetylene reduction on L-arabinose by Azospirillum brasilense 12 3 Reduction of NAD in crude extracts of L-arabinose grown cells with 1 ymole potassium L-arabonate or DL-KDA; as substrate 16 k Dithionite reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 105,000 xg supernatant 21 5 Dithionite reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 35,000 xg pel let 23 6 Dithionite reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 105,000 xg pel let 25 7 Elution from a Biogel A 1 5M (2 x 150 cm) column of L-arabinose dehydrogenase and L-arabonate dehydratase. kO 8 Elution from a DEAE cellulose column (2 x 10 cm) of L-arabinose dehydrogenase and L-arabonate dehydratase. hi 9 Lineweaver-Burk plot showing the effect of substrate concentration on reaction velocity hi 10 Effects of pH on reaction velocity of L-arabinose dehydrogenase i^g 11 Effect of pH on reaction velocity of L-arabonate dehydratase 5) 12 Molecular weight determination of L-arabinose dehydrogenase and L-arabonate dehydratase by gel filtration technique 5if 13 Polyacrylamide slab gel of crude and partially purified extracts of L-arabinose dehydrogenase 57 vi

PAGE 7

14 Polyacrylamide slab gel of partially purified extracts of L-arabonate dehydratase 59 15 Schematic of oxygen controller 63 vi i

PAGE 8

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy METABOLISM OF L-ARABINOSE \n AZOSPIEILLUM BRASILENSE By Norman James Novick December I98O Chairman: Arnold S. Bleiweis Major Department: Microbiology and Cell Science Under dinitrogen fixing conditions Azospivittwn bvasilense showed comparable generation times using L-arabinose or malate as sole carbon and energy sources. This root associated bacterium was found to metabolize L-arabinose through an oxidative pathway that has also been found in certain species of Pseudomonas and fast-growing Rhisobium. L-Arabinose is converted to L-arabono-^lactone by an NAD(P) dependent dehydrogenase, hydrolyzed to L-arabonic acid by a lactonase, and dehydrated to L-2-keto-3-deoxyarabonate (L-KDA) by dehydratase activity. In crude extracts NAD is rapidly reduced if potassium L-arabonate or DL-KDA is added as substrate. DL-KDA has been found to reduce NAD at 3 times the rate of L-arabonate. Al pha-ketogl utarate accumulated in crude extracts to which L-arabonate and NAD were added. It is proposed that L-KDA is dehydrated to alpha-ketogl utar ic semialdehyde which is then oxidized to a 1 pha-ketogl utar i c acid by an NAD dependent dehydrogenation reaction. L-Arabinose dehydrogenase and L-arabonate dehydratase have been partially purified and characterized. The NAD(P) dependent vi i i

PAGE 9

dehydrogenase, which has been found to be specific for the L-arabonoconf i gurat ion, has been purified 59 fold. No enhancement of activity in the presence of any divalent cation or reducing agent tested has been found. The values of 75 yM and 1^0 yM were found with NADP and NAD as cofactors, respectively. The enzyme has a pH optimum of 9.5 in glycine buffer and was stable when heated to 55C for 5 minutes. The enzyme product has been identified as L-arabono-ylactone. LArabonate dehydratase has been purified 38 fold. The presence of MgCl2 and MnCl2 has been found to significantly increase enzyme activity. The pH optimum for the dehydratase has been found to be 7.8. The enzyme product was identified as L-2-keto-3-deoxyarabonate. In further studies TCA cycle dehydrogenases have been assayed and indicate an active TCA cycle. The presence of b and c type cytochromes has been confirmed and their distribution between membrane and soluble fractions determined. i X

PAGE 10

SECTION I IDENTIFICATION OF ENZYMES INVOLVED IN L-ARABINOSE METABOLISM IN AZOSPIHILLUM BRASILENSE. ASSAY OF TCA CYCLE DEHYDROGENASES AND IDENTIFICATION OF COMPONENTS INVOLVED IN ELECTRON TRANSPORT I ntroduct ion Azospirillum brasilense (ATCC 29145) grows and fixes molecular nitrogen best on organic acids such as succinate, malate, and pyruvate (13, 28, 29, 38). Variable results have been reported for A. brasilense growth on sugars. Day and Dobereiner (13) reported that several sugars including glucose could be used for growth under nitrogen fixing conditions but only if "starter" organic acid was added to the medium. Okon et al. (28) found growth and acetylene reduction rates were less on galactose than on several organic acids tested and found little or no growth under nitrogen fixing conditions with glucose as a carbon and energy source. Tarrand and Krieg (38) have also found A. brasilense unable to use glucose as a sole carbon and energy source under nitrogen fixing conditions. Okon et al. (29) found little or no increase in oxygen uptake rates above endogenous levels in cell free extracts if glucose, fructose, galactose, or several phosphory 1 ated metabolic intermediates were added as substrates. Child and Kurz (6) reported that in certain cases nitrogenase activity in A. brasilense could be enhanced by supplementing the organic acid containing medium with arabinose. In examining A. brasilense growth under nitrogen fixing conditions on a variety of 1

PAGE 11

2 substrates, I found that L-arabinose served as an excellent sole carbon and energy source. I therefore decided to investigate the mode of L-arabinose catabolism in this root associated bacterium. My initial failure to find a pathway involving phosphorylated Intermediates led me to look at other possible pathways. There are two well documented pathways of L-arabinose metabolism not involving phosphorylated intermediates. In the first pathway L-arabinose is oxidized to L-arabono-y1 actone by an NAD(P) dependent dehydrogenase, the lactone is cleaved by lactonase to L-arabonic acid, followed by two successive dehydration reactions forming L-2-keto-3-deoxyarabonate and a Ipha-ketogl utari c semialdehyde, respectively. The last step involves the NAD dependent dehydrogenat ion of the semialdehyde to alphaketoglutaric acid. The second pathway has the same initial steps but the L-2-keto-3~deoxyarabonate is cleaved through an aldolase reaction to glycolal dehyde and pyruvate. The first pathway has been demonstrated in Pseudomonas saccharophila (hi), Pseudomonas fragi (7, h\) and the fast growing rhizobia (15, 16), R. meliloti, R. tri-foli-i-, R. phaeseoti^ and R. leguminosarum. The second pathway has been demonstrated in Pseudomonas strain MSU-1 (8), and slow growing rhizobia; R. japonicwn (15, 31) and Rhizobium sp. 32H1 (15). In addition to these L-arabinose dehydrogenase has also been demonstrated in Aquaspirillion graaile (23) although the subsequent pathway is unknown. in this paper evidence is presented indicating the first pathway (i.e. L-arabinose > a 1 pha-ketogl utar i c acid) is present in A. brasilense The presence of TCA cycle dehydrogenases and cytochromes is also demonstrated, indicating that the a 1 pha-ketogl utarate formed

PAGE 12

3 provides energy to the cell through electron transport coupled to TCA cycle dehydrogenases. Materials and Methods Bacterial Strain Azospirillim hrasilense (ATCC 29145) was obtained from the American Type culture collection. For daily use cultures were maintained on Trypticase soy agar (BBL) with biweekly transfer. Stock cultures were frozen in liquid nitrogen. Med i a The growth medium was that of Nelson and Knowles (27) with twice the concentration of potassium phosphate buffer (65 mM, pH 7.1). Malate or L-arabinose (37 mM) was filter sterilized and added to the autoclaved and cooled medium. Ammonium chloride, was added to ammonia grown cells. Growth and Nitrogen Fixation Two and one-half liters of the nitrogen free medi urn were added to a vessel consisting of a 2.8 L Fernbach flask tightly capped with a rubber stopper through which had been placed a gas inlet port (ending in a sparging stone), a gas outlet port, a sampling port, and an oxygen probe (2). The oxygen probe and air line were connected to an oxygen Stat (see Appendix 1). The flask containing a stirring disc was immersed ina25C water bath set on a magnetic stirrer. Air and nitrogen were mixed to maintain the 0„ concentration at O.S% + 0.05%.

PAGE 13

4 For NH2^C1 grown cells, 1.5 L of med i urn were p 1 aced in a cotton stoppered Fernbach flask and rapidly shaken at room temperature. in both N2 and ammonia grown cells a ]0% inoculum containing 0.05^ NH^Cl was grown aerobical ly. Acetylene Reduction Assay Six milliliters of cells were removed anaerob i ca 1 1 y from the growth vessel and placed in a 125 ml flask that had previously been sparged with argon and capped with a serum stopper. Oxygen, 0,5%, was added as air back to the flask. The samples were shaken on a New Brunswick rotary shaker for 1 h at 150 rpm and 25C. One-tenth milliliter gas volume was removed and injected into a Varian mOO series gas chromatograph with Poropak Q columns and flame ionization detectors. Growth was followed by reading optical density at 560 nm and protein determined by the method of Lowry (ih) Determination of Arabinose Pathway Enzymes Both NH^Cl and grown cells gave similar specific activities for enzymes involved in arabinose metabolism. Due to this and the ease of growing large batches of cells with a fixed nitrogen source, ammonium chloride grown cells were used for the enzyme studies. Crude extracts were prepared in the following manner. Cells, near the middle or end of exponential growth, were collected by cen t r i f ugat ion and washed once in O.IM sodium-potassium phosphate buffer, pH 7-5, resuspended to ca. 0.25 g per ml in the same buffer, passed twice through an Aminco french pressure cell at 20,000 p.s.i., and centrifuged at 10,000 xg for 30 min. The supernate, crude extract, could be stored for several months at

PAGE 14

5 -80C without significant loss of enzymatic activity. L-arabinose dehydrogenase was assayed in a reaction mixture containing 125 ymoles glycine/NaOH buffer, pH 9-0, 0.5 ymoles NAD, 0.5 pmoles L-arabinose, 1 ymole NaCN, and enzyme plus water to 1 ml The change in optical density at 3^0 nm was followed on a Zeiss spectrophotometer at 25C. Lactonase activity was measured by the method of Dahms and Anderson (10) using alkaline hydroxyl ami ne reagent (18) to quantitate the disappearance of L-arabonolactone. Arabonate dehydratase assay was that of Pedrosa and Zancan (31) with L-2-keto-3"deoxyarabonate (L-KDA) accumulation being measured with semicarbazide reagent (25). KDA aldolase was assayed in the reverse direction (31) with KDA formation being followed by the th iobarb i tur i c-per iodate method (^43). The following assay mixture was used to determine NAD dependent KDA oxidation: 60 ymoles potassium phosphate buffer pH 7.2, 10 ymoles NAD, 10 ymoles potassium arabonate, 1 ymole NaCN, and water plus enzyme to 1 ml In all cases 1 unit of enzyme activity is that amount of enzyme which produces 1 ymole of product per minute at 25C. Identification of Pathway End Product The accumulation of a 1 pha-ketogl uta r i c acid in crude extracts, after the addition of potassium arabonate (10 ymoles), was determined as follows. The arabonate oxidation mix was the same as that used to follow KDA oxidation. In some experiments water or 10 ymoles sodium arsenite replaced NaCN. The reaction mix was incubated up to six hours and a 1 pha-ketogl utarate was enzymat i ca 1 1 y quantitated in a reaction mixture containing potassium phosphate buffer, 80 ymoles, pH 7.2, 0.32 ymoles NADH, 40 ymoles NH^Cl 3-3 units glutamate dehydrogenase (Sigma)

PAGE 15

6 and 0.05 ml of the above arabonate oxidation mixture in a total of 1 ml. The amount of NADH oxidized was determined against controls without NADH and without both glutamate dehydrogenase and NH^Cl to correct for NADH oxidase activity. Controls were also run using the arabonate oxidation mixture to which arabonate had not been added. Pyruvate was enzymat i cal 1 y determined in a reaction mixture containing potassium phosphate buffer, 50 ymoles, pH 1.1, 1 unit lactic dehydrogenase (Sigma), 0.32 ymoles NADH, 0.05 ml arabonate oxidation mix, and water to 1 ml Controls were run without lactic dehydrogenase to correct for endogenous NADH oxidase activity and with arabonate oxidation mixture to which arabonate had not been added. Glycolaldehyde was quantitated with d i phenyl ami ne reagent (14). Chromatography Dehydrogenase and dehydratase reaction products, using partially purified enzyme, as well as identification of the arabinose pathway terminal product, were determined by thin layer and paper chromatography. The samples were processed as follows. The reaction mixture was passed through a small Dowex-X8 (H+ form) column, 0.5 ml bed volume, to remove cations and precipitate proteins, then passed through a 0.45 y membrane filter, and concentrated 20 fold by 1 yophi 1 i zat ion KDA was further processed by the method of Weimberg (40) before spotting. Samples were spotted on Whatman #1 paper and resolved in one of the following solvent systems: pyridine: 1-butanol :water (6:4:3), 1 -propanol : formi c ac i d : water (6:3:1); or 1 -butanol : 1 -propanol :water (10:7:5). Compounds were detected with alkaline silver nitrate reagent (36) or aniline-xylose

PAGE 16

7 reagent (36). D i n i t ropheny 1 hydrazones of alpha-keto acids, produced by the method of Caval I i n i et a 1. (5), were spotted on Silica G gel plates and resolved in 1-butanol saturated with 3% NH^^OH. Spots were accentuated by spraying with 0.5N NaOH. TCA Cycle Enzymes Isocitrate and malate dehydrogenases were assayed by the method of Reeves et al. (33). One unit of act i vi ty was the amount of enzyme which reduced 1 ymole of NAD(P) per minute at 25C. Succinate dehydrogenase was assayed by a modification of the method of Veeger et al. (39). The 1 ml reaction mixture contained 50 ymoles potassium phosphate buffer, pH 7.6, 40 ymoles sodium succinate, 1 ymole NaCN, 1 mg phenazine methosulfate, and 0.25 ymoles 2,^ d i chlorophenol i ndol phenol (DCPIP). One unit of act i vi ty Was the amount of enzyme which reduced 1 ymole of DCPIP per min. at 25C. The dye had an extinction coefficient at 600 nm of 19,100 (35). Al pha-ketogl utarate activity was assayed by following the reduction of ferr i cyan i de The assay was the same as that for pyruvate decarboxylase (32). One unit of act i vi ty was the amount of enzyme necessary to reduce 2 ymoles of ferricyanide per hour. Cytochrome Determinations Crude extracts were separated into membrane and soluble fractions by the method of Jones and Redfearn (19). Reduced minus oxidized spectra were determined on a Beckman model 25 recording spectrophotometer at room temperature. Samples were reduced by the addition of a few grains of sodium dithionite to a 1 ml quartz cuvette containing the fraction to be assayed. Samples were oxidized by rapidly shaking the cuvette

PAGE 17

8 before insertion into the spectrophotometer. Cytochrome b was quantitated using a difference spect rum of 560-538 nm and an estimated extinction coefficient of 17,300. Cytochrome c was quantitated using a difference spectrum of 5^9-575 and an estimated extinction coefficient of 17,500. Flavoprotein was quantitated using a difference spectrum of ^165-510 and an extinction coefficient of 11,000. All values for difference spectra and extinction coefficients are those of Jones and Red f earn (19). Potassium arabonate was prepared by the hypoiodite oxidation of L-arabinose (26). DL-2-keto-3"deoxyarabonate was chemically synthesized by the method of Stoolmiller and Abeles (37) and purified by the method of Weimberg (4o) L-Arabonol actone was produced by boiling potassium arabonate in 0.2N HCl for 5 min. Resul ts Growth and Nitrogen Fixation Azo spirillum brasilense showed a doubling time of 16-20 h when grown on either malate (Fig. 1) or L-arabinose (Fig. 2) at 0.5^ O2. Malate showed a kO% higher specific activity of acetylene reduction than did L-arabinose grown cells (average of 3 experiments). The pH of the medium dropped only slightly after 72 h growth on L-arabinose (7.1-7.05) while pH rose sharply in malate grown cells (7-1 to 8.6). Arabinose Pathway Enzymes Table 1 gives the specific act i vi ties of L-arab i nose dehydrogenase, L-arabonolactonase, L-arabonate dehydratase, and the DL-KDA reduction

PAGE 18

Figure 1. Growth and acetylene reduction on L-malate by Azospirillum brasilense. Each acetylene reduction point average of 3 samples. Standard deviation given by width of bar through point. Optical density, 0 acetylene reduction

PAGE 19

10 3000time (hours)

PAGE 20

Figure 2. Growth and acetylene reduction on L-arabinose by Azospir-itlum bvasilense. Each acetylene reduction point average of 3 samples. Standard deviation given by width of bar through point, a Optical density, 0 acetylene reduction

PAGE 21

12

PAGE 22

Table 1-1, L-arabinose pathway enzymes. Average of 3 crude extract preparations. Act i vi ty (nmol/min/ Enzyme Cofactor mg protein) IkNAD 70 L-arabinose dehydrogenase ^^^^ Arabonolactonase 13 L-Arabonate dehydratase 17 DL-KDA oxidation NAD 62 L-KDA aldolase 0

PAGE 23

14 of NAD in crude extracts of L-arabinose grown cells. Figure 3 shows the reduction of NAD in a crude extract to which L-arabonate or DL-KDA was added as substrate. DL-KDA exiiibited 3 times the rate of NAD reduction than did potassium L-arabonate. No KDA aldolase activity could be found in any crude extract of L-arabinose grown cells. End Product Accumulation Table 2 indicates that crude extracts of L-arabinose grown cells accumulate a1 pha-ketogl utar i c acid when L-arabonate is added as substrate. The largest accumulation is seen if sodium arsenite, an inhibitor of a Ipha-ketogl utarate and pyruvate dehydrogenase activity (21, ^1), is added to the reaction mixture. No accumulation of either pyruvate or glycol aldehyde has been found in crude extracts to which L-arabonate had been added as substrate. This lent further credence to the likely absence of L-KDA aldolase in these cells. Identification of Reaction Products At pH 6.5 a major product of the L-arabinose dehydrogenase reaction is arabono-y1 actone as identified on paper chromatograms run against authent ic lactone. The sample comigrates with standard in all 3 solvent systems given in Methods and Materials. Even at this low pH, however, L-arabonic acid appears as a significant part of the product. At pH 9.0 the sample comigrates with L-arabonic acid in all three solvent systems. The product of the L-arabonate dehydratase reaction comigrates with DL-KDA in all three solvent systems. In addition the 2,^* dinitrophenylhydrazone of the reaction product comigrated with the derivatized standard DL-KDA on Silica G gel plates in 1-butanoI

PAGE 24

Figure 3. Reduction of NAD in crude extracts of L-arabinose grown cells with 1 ymole potassium L-arabonate o; or DL-KDA>; as substrate.

PAGE 26

Table 1-2. Accumulation of al pha-ketogl utarate in crude extracts of L-arabinose grown cells with L-arabonate as substrate. Al phaketogi utarate Rxn. Mix Accumulation Addition (ymoles) Sodium arsenite 7-6 Sodium cyanide A. 2 none i( 4 control" 0.^9 No arabonate was added to arabonate oxidation mix. See Materials and Methods.

PAGE 27

]8 saturated with }>% NH^OH. On spraying with 0.5N NaOH the spot gave the reddish brown color reported by Weimberg (AO). In addition to enzymatic assay al pha-ketoglutar ic acid was identified chromatograph i ca 1 1 y in crude extracts to which potassium L-arabonate had been added as substrate in the presence of NAD. The reaction product comigrated v-/ith alphai
PAGE 28

19 Table I-3. TCA cycle dehydrogenases, Enzyme Electron Acceptor Act i vi ty (nmoles/mi n/mg protein) Malate dehydrogenase Isocitrate dehydrogenase Succinate dehydrogenase NAD NADP PMS DCPIP 2,080 230 150 Al pha-ke tog 1 uta rate dehydrogenase Ferr i cyan i de nmoles/hr/mg protein

PAGE 29

Figure h. Dithionlte reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 105,000 xg supernatant.

PAGE 30

wavelength (nm)

PAGE 31

Figure 5. Dithionite reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 35,000 xg pel let.

PAGE 32

—J— I I J— 400 500 600 700 wavelength (run)

PAGE 33

Figure 6. Dithionite reduced minus oxidized spectra of membrane and soluble fractions from L-arabinose grown cells. 105,000 xg pel let.

PAGE 34

25 wavelength (nm)

PAGE 35

26 Table Cytochrome content of membrane and soluble fractions, ymoles/g protein Cytochrome Cytochrome c b Flavoprotein 105,000 xg supernatant 0.6^ 0 0.79 105,000 xg pellet 0.56 0.83 35,000 xg supernatant 0.20 0.31

PAGE 36

27 Pi SCU5S i on Azospirillum brasilense grew at approximately the same rate on both malate and L-arabinose. Malate grown cells have shown a consistently higher specific activity of acetylene reduction than have L-arabinose grown cells, although both rates are high and compare closely with the rate reported by Nelson and Knowles (27) for malate grown cells at 0.5% The doubling time of 16-20 h is much slower than the 5.5 to 7 h reported by Okon et al. (30) at 0.5^ O2 (with malate as substrate), but approximately the same as the 20 h generation time reported for stagnantly grown cultures (28). It should be noted in my study, that cells grown in malate or L-arabinose were cultured under identical conditions. The sharp rise in pH of malate grown cells may have accounted for a similar growth rate compared to that of L-arabinose grown cells despite the higher acetylene reduction rates shown by malate grown cells. Initial attempts in our laboratory to demonstrate an L-arabinose catabolic pathway involving phosphory 1 ated intermediates were unsuccessful. No L-arabinose isomerase or phosphotransferase activity could be found. I have since found transketolase and t ransa 1 dol ase activity, in crude and soluble fractions of L-arabinose grown cells, with about one-tenth the specific activity of L-arabinose dehydrogenase. The means by which L-arabinose might enter an oxidative pentose cycle is still unknown. Some ability to metabolize phosphory 1 ated compounds has been indicated in 0^ uptake studies (29). In examining the possibility of a pathway without phosphory 1 ated intermediates I found high NAD(P) dependent L-arabinose dehydrogenase

PAGE 37

28 activity. In addition L-arabono-y 1 actonase L-arabonate dehydratase, and NAD dependent 2-keto-3~deoxyarabonate oxidation activity were found. Al pha-ketogl utarate semi aldehyde has been shown to be the substrate of the NAD dependent dehydrogenat ion reaction (37). I did not look for this intermediate nor did I assay for the L-KDA dehydratase activity, which produces the a 1 pha-ketogi utarate semialdehyde. The rapid reduction of NAD with DL-KDA as substrate, which has also been reported by Weimberg (^O) in Pseudomonas saaaharophila, would seem to be evidence of significant L-KDA dehydratase activity. Weimberg and Doudoroff (42) reported a more rapid reduction of NAD if crude extracts of Ps. saoaharoph-ila were first preincubated with L-arabonate before the addition of NAD. This preincubation probably allowed time for the accumulation of L-KDA and perhaps a 1 pha-ketog 1 utarate semialdehyde in the reaction mixture. An average of 75% of the potassium arabonate added to the crude extract, in the presence of NAD, accumulated as al pha-ketogl utari c acid if sodium arsenite was added to the crude extract. Sodium arsenite is an inhibitor of al pha-ketogl utarate (41) and pyruvate dehydrogenases (21), probably by inhibiting the decarboxylation step. Significant amounts of al pha-ketogl utarate still were found in the absence of sodium arsenite. Sodium cyanide had no effect on al pha-ketogl utarate accumulation. The presence of L-arabinose dehydrogenase, L-a rabono-y1 actonase and L-arabonate dehydratase activity, along with the rapid reduction of NAD in crude extracts with DL-KDA as substrate, and the accumulation of alpha-ketoglutaric acid in crude extracts, is clear indication of the following pathway of L-arabinose metabolism in A. brasilense.

PAGE 38

29 NAD(P) '^^2^ L-arabinose > L-arab ino-y lactone > L-arabonic -H2O -H2O acid — V L-2-keto-3-deoxyarabonic acid y alpha-ketoglutaric semi a 1 dehyde >. a 1 phaketog 1 utar i c acid Oxygen uptake studies (29) have indicated a very active TCA cycle in A. brasilense although the specific enzymes have not been previously assayed. My assays of TCA cycle dehydrogenases would seem to confirm high TCA cycle activity and these activities were comparable to that shown by Rhizobiim meliloti (16). The significance of the very high activity of malate dehydrogenase relative to the other dehydrogenases is unknown. The presence of b and c type cytochromes and a cytochrome oxidase in deoxycholate extracts of Azospirillwn brasilense has previously been shown (if). For the first time the distribution of the cytochromes between particulate and soluble fractions has been demonstrated. The distribution is similar to that found in A. vinelandii (19). There also appears to be a large flavoprotein factor. The possible pathway of electron transport in A. brasilense may be the following. The X represents a possible ubiquinone type electron carrier. NADH -y Flavoprotein y X > "^^^ v b cyt ^ cyt c oxidase ^ ^2' It has been suggested by Eskew et al. (1?) that the pink pigment found in some strains of Asospirillum is due to a c type cytochrome. The reduced minus oxidized spectrum published by Eskew et al. (1?) of the concentrated pigment is virtually identical to the spectrum 1 found in the soluble fraction (105,000 xg supernatant) of the L-arabinose

PAGE 39

30 grown cells. This, along with the significant amount of c cytochrome I found in the soluble fraction, is further evidence that the pink pigment produced by this strain of Azospirn-ltum i s in fact a c cytochrome. We have demonstrated a totally oxidative pathway by which L-arabinose is metabolized in Azosp'ir'illwn hvasilense. This ability may offer some distinct advantages to this soil bacterium. Plant cell walls are rich in compounds with the L-arabono-conf i gurat i on (i.e. L-arabinose, DFucose D-ga lactose) { 1 ) L-Arab i nose i s one of the few i somers found in plant cell walls. The cellulose fibers inthe wa 1 1 s are cemented together by xyloglucans, arabinogalactans, and rhamnogalacturonan (l). It is possible that root cell death and the s 1 ough i ng of f of the eel 1 s provides root associated bacteria, ] \ke Azospirillum wh't ch is able to metabolize these sugars, a plentiful carbon substrate. The advantage provided to these bacteria in the highly competitive soil environment is obvious. Another benefit of this pathway is that there i s a ready pool of al phaketogl utarate which can act as substrate for glutamate dehydrogenase and glutamate synthase activity (29). Both enzymes are important in nitrogen assimilation under fixed nitrogen and dinitrogen fixing conditions, respectively. Finally the re i s the i nc reas i ng s imi 1 a r i t i es demonstrated betvieen Azospirzllion brasilense and some Rhizobiim spec] es. Included in these is the oxygen sensitivity of their nitrogenase (20,30), high dissimilatory nitrate reductase activity (3^, 38), accumulation of PHB granules in eel 1 s (3, 29), the enhancement of n i t rogenase activity by inclusionofa pentose wi th the organic acid substrate (6), and similar GC content (3, 38), The L-arabinose pathway described in this paper is one that had prev i ous 1 y been found only in some species of Pseudomonas (7, Al, 42) and fast growing rhizobia (15, iQ. The presence of this pathway is another i nd i cat i on of a close relationship between Azospirillum and Rhizobium species.

PAGE 40

SECTION I I PURIFICATION AND CHARACTERIZATION OF L-ARABINOSE DEHYDROGENASE AND L-ARABONATE DEHYDRATASE FROM L-ARABINOSE GROWN CELLS OF AZOSPIEILLUM BRASILENSE I ntroduct ion Azospirillum brasilense (ATCC 291^5) can metabolize L-arabinose by the following series of reactions NAD(P) "^^2 L-arabinose > L-arabono-7lactone > L-arabonic -H„0 -H„0 acid ^ >L-2-keto-3 deoxyarabon i c acid ^ alpha-ketoglutaric semialdehyde — > al pha-ketogl utar i c acid This pathway has also been denranstrated in Pseudomonas saooharophila {k2) Pseudomonas fragi (7, h]) and the fast growing rhizobia (15, 16) E. meliloti, R. trifolii, R. phaeseoli, and R. leguminosamm. A second pathway which carries through the first 3 steps of the alpha-ketoglutarate pathway but cleaves L-2-keto-3-deoxyarabonate (L-KDA) by an aldolase reaction to gl ycol a 1 dehyde and pyruvate has been found in Pseudomonas strain MSU-1 (8) and the slow growing rhizobia; R. {japonicum (15, 31) and Rhizobiujv sp. 32H1 (15). The enzymes of these pathways have not been extensively studied. L-arabinose dehydrogenase and the L-arabonate oxidation system (Larabonate > al pha-ketogl utar i c acid) have been partially purified in Pseudomonas saooharophila {hi). L-2-keto-3-deoxyarabonate dehydratase has been purified and characterized by Stoolmi 1 ler and Abeles (37). Dahms and Anderson have partially purified L-KDA aldolase from Ps. MSU-1 (8). 31

PAGE 41

32 D-Fucose metabolism in Ps. MSU-I is very similar to that found for L-arabinose. In the D-fucose pathway, D-2-keto-3"deoxyf uconate is cleaved to pyruvate and 1 acta 1 dehyde All the enzymes in this pathway have been purified and characterized by Dahms and Anderson (9-12). I have partially purified and characterized L-arabinose dehydrogenase and L-arabonate dehydratase from L-arabinose grown cells of Azospirillum brasilense. Materials and Methods Bacterial Strain Azospirillian hrasilense (ATCC 23]kS) was obtained from the American Type Culture Collection. For daily use cultures were maintained on Tryticase soy agar (BBL) slants with biweekly transfer. Stock cultures were frozen in liquid nitrogen. Enzyme Purification Crude extracts were prepared as stated in Section I. All steps in the purification of enzymes were carried out at ^"C unless otherwise stated. Nucleic acid precipitation A 2% solution of protamine sulfate (Sigma, Grade l|) in O.IM sodium-potassium phosphate buffer, pH 7-5 was slowly added to the crude extract to give a final concentration of 0.33%, stirred in the cold for 30 min., and centrifuged at 20,000 xg for 1 hr. The pellet was discarded.

PAGE 42

33 Ammonium sulfate fractionation Solid ammonium sulfate (Sigma, Grade III) was slowly added to the protamine sulfate supernatant while maintaining the pH at 7-5 with O.IN NaOH. The sample was stirred in the cold for 1 hr then centrifuged at 20,000 xg for 1 hr. The pellets were resuspended in O.IM sodiumpotassium phosphate buffer pH 7-5. Ninety-two percent of the dehydratase activity was found in the 30% and kOZ ammonium sulfate precipitate fraction. These fractions were pooled before being placed on Biogel A column. Eighty-nine percent of the dehydrogenase was found in the 50% ammonium sulfate precipitate fraction. This fraction was carried through a heat treatment step before being loaded onto the Biogel column. Heat treatment (dehydrogenase only) The resuspended S0% ammonium sulfate fraction was heated to 55C for 5 min, immediately cooled in ice to 4C, and centrifuged at 20,000 xg for 1 hr. Gel f i 1 trat ion Sample was added to a 2 x 150 cm Biogel A 1.5M column equilibrated with 25 mM sodium-potassium phosphate buffer pH 7.4. Flow rate was 25 ml per hr and 5 to 20 ml fractions were collected. Ion exchange chromatography A 2 x 10 cm DEAE cellulose (fine) column was equilibrated with 3 column volumes of 25 mM sodium-potassium phosphate buffer, pH 7.4. Samples were either placed directly on the column from active, pooled Biogel A fractions or Biogel fractions were concentrated first in an

PAGE 43

3^ Amicon pressure cell with UMBO filter. Enzyme was eluted in a 0 to 0.2M (dehydrogenase) or 0 to 0.3M (dehydratase) 300 ml linear NaCl gradient. All NaCl solutions were prepared in 25 mM sodium-potassium phosphate buffer, pH 7'^Active DEAE fractions were immediately desalted on a G-25 column, equilibrated with 25 mM sodium-potassium phosphate buffer, pH 7.^. Flow rate was 50 ml per hr and 20 ml fractions were collected. Active fractions were concentrated by lyophi 1 ization. Protein Determination Protein was determined by the method of Lowry et a1 (2k) with BSA as a standard. Enzyme Assays L-arabinose dehydrogenase was assayed in 125 ymoles glycine/NaOH buffer, pH 9-0, 0.5 ymoles NAD, 0.5 ymoles L-arabinose, and enzyme plus water to 1 ml In crude extracts 1 ymole NaCN was added to the reaction mix. The change in optical density at 3^*0 nm was followed on a Zeiss spectrophotometer at 25C. One unit of L-arabinose dehydrogenase activity was the amount of enzyme which reduced I ymole NAD(P) per min. Arabonate dehydratase activity was assayed by the method of Pedrosa and Zancan (31) with 2-keto-3-deoxyarabonate accumulation being measured with semicarbazide reagent (25). One unit of activity was the amount of enzyme which produces 1 ymole of L-KDA per min. Molecular Weight Determination The 2 X 150 cm Biogel A column was equilibrated with 25mM sod i um-potass i um phosphate buffer, pH 7,^4, and 2 mg of each standard (Boehringer Calibration Proteins II), including

PAGE 44

35 catalase (2^0,000 daltons), aldolase (158,000), albumin (68,000), and albumin (45,000), were dissolved in 20 ml of buffer and placed on the column. The column was run at 25 ml per hour and 7-5 ml fractions were collected. Standard peaks were located by A^gQ. L-Arabinose dehydrogenase and L-arabonate dehydratase were located by enzyme activity. A plot of the molecular weight of the standards versus the logarithm of the volume of the half-height of the leading edge of each compound was made. The standards generated a straight line from which unknown molecular weights were determined. Product Identification Products of L-arabinose dehydrogenase and L-arabonate dehydratase were identified as previously stated in Section I. Pol yacry 1 ami de Gel Electrophoresis Slab gels were prepared by the method of Laemml i (22). A 5-15^ exponential gel was cooled to kC and run at 32 ma. Enzyme activity was located by slicing the gel in 8 mm sections, placing the sections in 13 X 100 mm tubes, and eluting the enzyme in 1 ml of 25 mM potassium phosphate buffer, pH 7-2. Enzyme activity was assayed in the standard manne r. Resul ts The purification of L-arabinose dehydrogenase and L-arabonate dehydratase are outlined in Tables 1 and 2, respectively. The dehydrogenase was purified 59 fold with 1.2^ yield and L-arabonate dehydratase was purified 38 fold with SZ yield. Attempts were made to improve

PAGE 45

36 Table 2-1. Purification of L-arabinose dehydrogenase. Fract ion Protein (mg) Sp. Act. Uni ts/mg Protei n Total Act i vi ty Uni ts Puri f icat i on Recovery Crude US'* 0.0? 103.8 Protami ne sul fate supernatant 1280 0.07 89.6 Ammon i urn sul fate (50^ prec.) 733 0.08 58.6 Heat treatment 288 0.20 57.6 Biogel A 1.5M (fractions pooled) 20.6 l.i3 29.^ DEAE G-25 washed cone. lOX 0.3 ^.12 1.2 1.1 2.9 20.^ 58.9 86 56 55 28 1.2

PAGE 46

37 Table 2-2. Purification of L-arabonate dehydratase. Fract ion Protein (mg) Sp. Act. Un i ts/mg Protein Total Act i vi ty Units Purification Recovery Crude 2^97 0.011 Protami ne sul fate supernatant 1783 0.015 Ammon i um sulfate prec. (30 S kQ% fractions) 635 0.007 Biogel A 1 .5M (pooled fractions) 19 0.19 DEAE G-25 washed 6 O.kl 27.5 26.7 3.6 2.5 1.4 17.2 38.2 97 16 13 9

PAGE 47

38 the purification and yield of L-arabinose dehydrogenase. Calcium phosphate gels, affinity chromatography (Affi-gel Blue), and DEAE Sephadex ion exchange columns failed to improve either of the above factors. The elution profiles of L-arabinose dehydrogenase and Larabonate dehydratase activities from a Biogel A 1.5M column are shown in Fig. 7. It is clear that only partial resolution of these two enzymes was achieved by gel filtration. Therefore it was necessary to further chromatograph the Biogel fractions on DEAE cellulose eluted with a linear NaCl gradient. Figure 8 demonstrates preliminary results obtained from crude 30% and 'O^ ammonium sulfate fractions. Excellent resolution of the two enzymes was achieved. Dehydrogenase activity started to elute from the DEAE column at O.OBM NaCl while the dehydratase did not start to elute from the column until 0.19M NaCl was applied. The dehydrogenase came off in unstable condition and it was necessary to desalt the eluate immediately on a G-25 col umn Thus ion exchange chromatography allowed excellent separation of enzymes from each other. Characteristics of L-Arabinose Dehydrogenase L-Arabinose dehydrogenase showed no enhancement of activity in the presence of any of the divalent cations listed in Table 3. The presence of CaCl^, FeSO|j, MnCl^ and CoCl^ severely inhibited enzyme activity. The presenceof EDTA has a si ight negative effect on activity and the presence of the reducing agents 2-mercaptoethanol glutathione (reduced), and dithiothreitol also had a small negative effect on activity. Only those substrates with the L-arabono configuration (i.e. Dgalactose, D-fucose) were good substrates for the L-arabinose

PAGE 48

Figure 7Elution from a Biogel A 1.5M (2 x 150 cm) column of L-arabinose dehydrogenase and L-arabonate dehydratase. The resuspended 30% and 40^ ammonium sulfate precipitate fractions were pooled and placed directly on the column. Assays are as stated in Materials and Methods Fractions (20 ml) were collected and protein determined by A„o-.

PAGE 49

fraction number

PAGE 50

Figure 8. Elution from a DEAE cellulose column (2 x 10 cm) of L-arabinose dehydrogenase and L-arabonate dehydratase. The resuspended 30% and ^0^ ammonium sulfate precipitate fractions were pooled and placed on the column. Enzyme was eluted from the column with a 300 ml, 0 to 0.3M NaCl linear gradient. Fractions (5 ml) were collected and protein determined by f^2S0' Enzyme assays were as in Materials and Methods.

PAGE 51

2 .4 6 8 10 12 14 16 18 20 22 24 26 28 30 fraction number

PAGE 52

Table 2-3. Effect of divalent cations and reducing agents on L-arabinose dehydrogenase activity. Relative Act i vi ty Rea aen t Concentration % None 100 MgSO^ 10 mM 100 10 mM 100 CaCl2 10 mM 70 FeSO|j 10 mM 0 MnCl2 10 mM 55 C0CI2 10 mM 46 EDTA 2.5 mM 92 2-mercaptoethanol 1 mM 95 Gl utathione (reduced) 1 mM 93 Di thiothrei tol 1 mM 90

PAGE 53

dehydrogenase (Table k) Lineweaver-Burk plots (Fig. 9) show a value of ]hO yM with 0.5 ymole NAD as cofactor and 72 yM with 0.5 ymole NADP as cofactor. The heat stable dehydrogenase (no loss of activity when heated to 55C for 5 min) had a pH optimum of 9.5 in glycine/NaOH buffer (Fig. 10) Characteristics of L-arabonate Dehydratase L-Arabonate showed a pH optimum of 7-8-8.0 in Hepes/NaOH buffer (Fig. 11). Magnesium sulfate and manganese chloride (Table 5) enhanced enzyme activity 3S% and 5(>t, respectively. The presence of ZnCl2, CaCl2 FeSO^^, and C0CI2 severely inhibited activity. The addition of EDTA (2.5 mM) resulted in a 35% inhibition of activity. Reducing agents 2mercaptoethanol and glutathione (reduced) had a small negative effect on activity while d i th iothrei tol greatly reduced activity. Molecular Weight Determination L-Arabinose dehydrogenase had a molecular weight of 175,000 and L-arabonate dehydratase had a molecular weight of 225,000 according to gel filtration determinations (Fig. 12). Enzyme Product Identification At pH 6.6 the product of the L-arabinose dehydrogenase reaction was found to be L-arabono-ylactone although, L-arabonic acid was also found in the reaction mixture. At pH 9.0 L-arabonic acid was the only product found. The product of the dehydratase reaction was identified as L-2-keto-3-deoxyarabonate. See Section I for a complete description of enzyme product identification.

PAGE 54

Table 2-k. Substrate specificity of L-arabinose dehydrogenase. The standard assay was used except substrate was varied. Relat i ve Substrate Velocity 10 mM % L-arabinose 100 D-ribose 0 D-galactose 63 D-xylose 1| D-fucose 100 L-fucose 0 L-rhamnose 0 D-mannose 0 D-glucose 0

PAGE 55

Figure 9. Lineweaver-Burk plot showing the effect of substrate concentration on reaction velocity. NAD o and NADPb (both 10 mM) were used as cofactors. The standard assay was used except substrate concentration was varied.

PAGE 56

hi

PAGE 57

Figure 10. Effect of pH on reaction velocity of L-arabinose dehydrogenase. Standard assay was used except pH of the glycine/NaOH buffer was varied.

PAGE 58

^3 PH

PAGE 59

Figure 11. Effect of pH on reaction velocity of L-arabonate dehydratase. Standard assay was used except pH of the HEPES/NaOH buffer was varied.

PAGE 61

Table 2-5. Effect of divalent cations and reducing agents on L-arabonate dehydratase activity. Re 1 a t i ve Concentration Activity Reagent mM % None 100 ZnCl^ 1 34 MgSO^ 10 135 (nh, )^S0, 10 1 nn CaCl2 10 55 FeSO^ 10 55 MnCl2 10 156 C0CI2 10 11 EDTA 2.5 5 2-mercaptoethanol 1 93 Glutathione (reduced) 1 92 Di thiothrei tol 1 41

PAGE 62

Figure 12. Molecular weight determination of L-arabinose dehydrogenase and L-arabonate dehydratase by gel filtration technique (see Materials and Methods).

PAGE 64

55 Pol yacrylami de Gel Electrophoresis Gels of purified dehydrogenase and dehydratase still showed several bands indicating that the enzymes were not totally pure. Assay of dehydrogenase activity in the gel showed the enzyme activity to be located 60 mm from the top of the gel (Fig. 13). Two dark bands, ^1 mm and 53.5 rm from the top of the gel (Fig. I't), appeared in the gels run with partially purified dehydratase. Dehydratase activity could not be located within the gel. It is possible that the dehydratase disassociated into two subunits which could not individually catalyze the dehydration reaction or that the enzyme was inactivated in the gel without disassociation. Discussion L-Arabinose dehydrogenase from ABospirilZum brasilense apparently metabolizes only substrates having the L-arabono configuration. Other aldose dehydrogenases are generally not specific for this configuration (9).. According to values the dehydrogenase had a greater affinity for substrate with NADP as cofactor than with NAD, although, V was ^ max twice as great with NAD as cofactor than with NADP. The enzyme had no requirements for any of the divalent cations tested norwas there enhancement of activity in the presence of reducing agent. The enzyme had a high pH optimum of 9-5. L-Arabinose dehydrogenase from A. brasitense shows several similarities to dehydrogenases from other bacteria metabolizing L-Arabinose through similar pathways. L-Arabinose dehydrogenase from i?. japonicion

PAGE 65

Figure 13. Polyacryl amide slab gel of crude and partially purified extracts of L-arabinose dehydrogenase. Lane A and B; protaminesul fate supernatant fraction (50 yg) ; Lanes C and 0, DEAE-G-25 washed fraction (50 yg) Arrow points to area of enzyme activity.

PAGE 66

57 A B CD

PAGE 67

Figure 1^. Polyacryamide slab gel of partially purified extracts of L-arabonate dehydratase. Lanes A and B contain DEAE-G-25 washed fraction (50 yg)

PAGE 68

59 A B

PAGE 69

60 (31) and D-fucose dehydrogenase from Ps. MSU-1 (9) are both capable of reducing NAD and NADP. L-arabinose dehydrogenase from Ps. saoaharophila is reported only capable of reducing NAD (^2). L-arabinose dehydrogenase from Ps. saccharophila (42); R. japonioum (31), and Aquaspirillwn gvaaile (23) are all assayed at high pH, although, optimum pH is not reported. D-Fucose dehydrogenase from Pseudomonas MSU-1 is similar to L-arabinose dehydrogenase from ^4. hrasilense in its lack of divalent cation requirement, lack of reducing agent requirement, heat stability, and pH optimum as well as substrate specificity (9). The immediate product of L-arabinose dehydrogenase was L-arabono-y lactone, whi ch was cleaved to L-arabonic acid by a lactonase. Even at pH of 6.6, though, there was considerable spontaneous breakdown of the 1 actone. L-arabonate dehydratase showed greatly enhanced activity in the presence of MgSO^ or MnCl^. its requirement for cations was clearly demonstrated by the near total loss of activity in the presence of the chelating agent EDTA. This was similar to that found for D-fuconate dehydratase (II). Unlike D-fuconate dehydratase this enzyme showed no enhancement of activity in the presence of any of the reducing agents tested. in fact reducing agents had a negative effect on activity. The semicarbazide positive product of the dehydratase reaction has been identified as L-2-keto-3-deoxyarabonate.

PAGE 70

APPENDIX DESIGN OF OXYGEN CONTROLLER A simple oxygen controller has been built for use in growth and nitrogen fixation studies. The sensing part of the instrument consists of a Borkowski and Johnson (2) oxygen probe. The probe acts as a tiny battery with electrons moving from the pure lead cathode to a pure silver anode covered by a teflon membrane. The amount of current running between cathode and anode is proportional to the rate at which oxygen is reduced at the anode. A resistor placed between the anode and cathode produces a potential drop which drives the sensing system. The whole system is outlined in Figure 1. A 7^1 operational amplifier allows the probe signal to be amplified up to 100 times. This signal is then fed to a 50 mv potentiometer (as part of the LFE compact controller). The system is calibrated by bubbling pure nitrogen (Matheson) through the sterile medium and setting the mv meter to zero with the 7^1 zero adjust. Two percent oxygen (Matheson) is then bubbled through the sterile medium and full scale is set with the amplification adjust. The heart of the control ler part of the instrument is an LFE Compact Controller with high and low relays. If the mv indicator drops below the low rel ay sett i ng, the relay closes; likewise, if it rises above the high relay setting, the high relay closes. The motor (115 Vac, 1 rpm) which drives the micrometering valve has a common, clockwise, and counterclockwise control wire. If the common and clockwise wires are 61

PAGE 71

Figure 15Schematic of oxygen controller.

PAGE 72

63 1 M eg + 12 Kl Radio Shock 275-004

PAGE 73

64 connected to 115 Vac the nx3tor turns clockwise, if the common and counterclockwise wires are connected to 115 Vac the motor runs counterclockwise. The common wire runs from the motor through a relay (connected to a 556 timer) to line voltage. The clockwise wire runs through the controller high relay and the counterclockwise wire runs through the compact controller low relay. The high and low relay wires come together to form the other Hne for 115 Vac connection. In use the 556 timer is set for 1 second on time and 1 minute delay time (time between on periods). The common line therefore closes for I second every minute. The high relay is set at approximately 14.5 and the low relay at 10.5. The mv indicator starts out at 12.5 on the 50 mv scale {0.5% O2). Every minute the instrument "checks" to see if either the high relay or low relay is closed. If, for instance, the high relay is closed, the motor will move approximately one-sixtieth of a turn (1 second on a 1 rpmmotor) counterclockwise closing the valve a fraction of a turn and decreasing O2 concentration in solution. This method manages to keep the O2 concentration at 0.5% +0.05%. Typically with A. brasilense O2 levels being pumped into the growth chamber will rise from 0.5% at the beginning of growth to 5.0% 0^ by late exponential phase (in order to keep dissolved 0„ at 0.5%).

PAGE 74

LITERATURE CITED 1. Albersheim, P. 1975. The walls of growing plant cells. Sci. Am. 232: 81-94. 2. Borkowski, J. D., and M. J. Johnson. 1967. Long-lived steamsterilizable membrane probes for dissolved oxygen measurement. Biotech, and Bioeng. 9: 635-639. 3. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Sergey's Manual of Determinative Bacteriology, 8th ed. The Williams and Wilkins Co. Ba 1 1 imore. 4. Burris, R. H., Y. Okon, and S, L. Albrecht. 1978. Properties and reactions of Spirillum lipoferwn. Environmental role of nitrogenfixing blue-green algae and asymbiotic bacteria. Ecol. Bull. (Stockholm) 26: 353-363. 5. Cavallini, D. N. Frontal i, and G. Toschi. 1949. Keto acid content of human blood and urine. Nature I64: 792. 6. Child, J. J. and W. G. W. Kurz. 1978. Inducing effect of plant cells on nitrogenase activity of Spirillum and Rhizobium in vitro. Can. J. Microbiol. 24: 143-148. 7. Dagley, S., and P. W. Trudgill. 1965The metabolism of galactarate, D-glucarate and various pentoses by species of Pseudomonas. Biochem. J. 95: 48-58. 8. Dahms, A. S., and R. L. Anderson. I969. 2-keto-3-deoxy-L-arabonate aldolase and its role in a new pathway of L-arabinose degradation. Biochem. Biophys. Res. Commun. 36: 809-814. 9. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a pseudomonad. II. Oxidation of D-fucose to Df ucono-y1 acone by an L-arabi no-al dose dehydrogenase and the hydrolysis of the lactone by a lactonase. J. Biol. Chem. 247: 2228-2232. 10. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a pseudomonad. I. Oxidation of D-fucose to D-f ucono-y1 acone by a D-aldohexose dehydrogenase. J. Biol. Chem. 247: 2222-2227. 11. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a pseudomonad. III. Conversion of D-fuconate to 2-keto-3-deoxy-Dfuconate by a dehydratase. J. Biol. Chem. 247: 2233-2237. 65

PAGE 75

66 12. Dahms, A. S., and R. L. Anderson. 1972. D-fucose metabolism in a pseudomonad. IV. Cleavage of 2-keto-3"deoxy-D-fuconate to pyruvate and lactaldehyde by a 2-keto-3-deoxy-L-arabonate aldolase. J, Biol. Chem. Ikl: 2238-2241 13. Day, J. M., and J. Dobereiner. 1976. Physiological aspects of N2-fixation by a Spirillum from Digitavia roots. Soil Biol. Biochem. 8: 45-50. 14. Dische, Z., and E. Borenfreund. 1949. A specific color reaction of glycol ic aldehyde. J. Biol. Chem. 180: 1297-1300. 15. Duncan, M. J. 1979. L-Arabinose metabolism in rhizobia. J. Gen. Microbiol. 113: 177-179. 16. Duncan, M. J., and D. G. Fraenkel. 1979. Al pha-ketogl utarate dehydrogenase mutant of Rhizohimi melilotti. J. Bacterid. 37: 415-419. 17. Eskew, D. L., D. D. Focht, and 1. P. Ting. 1977. Nitrogen fixation, den i t r i f i cat ion and pleomorphic growth of a highly pigmented Spirillum lipoferum. Appl. Environ. Microbiol. 34: 582-585. 18. Hestrin, S. 1949. The reaction of acetylcholine and other carboxylic acid derivatives with hyd roxy 1 ami ne and its analytical application. J, Biol. Chem. I8O: 249-261. 19. Jones, C. W. and E. R. Redfearn. I966. Electron transport in Azotobaoter vinelandii Biochem. Biophys. Acta. 113: 467-481. 20. Keister, D. L., and W. R. Evans. 1976. Oxygen requirements for acetylene reduction in pure cultures of rhizobia. J. Bacteriol 129: 149-153. 21. Kersters, K. and J. De Ley. 1971. Tests with cell free extracts, p. 46-51. In J. R. Norris and D. W. Ribbons (ed.) Methods in Microbiology, v. 6A. Academic Press, New York. 22. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of Bacteriophage T4, Nature 222: 68O-685. 23. Laughon, B. E. and N. R. Krieg. 1974. Sugar catabolism in Aquaspirillwn graaile. J. Bacteriol. 119: 69I-697. 24. Lowry, 0. H. N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the folin phenol reagent. J Biol. Chem. 193: 265-275. 25. MacGee, J., and M. Doudoroff. 1954. A new phosphory 1 ated intermediate in glucose oxidation. J. Biol. Chem. 210: 6I7-626.

PAGE 76

67 26. Moore, S., and K. P. Link. ]SkQ. Carbohydrate characterization. I. The oxidation of aldoses by hypoiodite in methanol. J. Biol. Chem. 133: 293-311. 27. Nelson, L. M. and R. Knowles. 1978. Effect of oxygen and nitrate on nitrogen fixation and den i tr i f i cat ion by Azospi-rillion hvasilense grown in continuous culture. Can. J. Microbiol. lh\ 1395-1403. 28. Okon, Y., S. L. Albrecht, and R. H. Burris. 1976. Factors affecting the growth and nitrogen fixation of Sp'Lvillum lipoferum. J. Bacterid. 127: 12i8-1254. 29. Okon, Y., S. L. Albrecht, and R. H. Burris. 1976. Carbon and amnxDnia metabolism in Spirillum lipoferum. J. Bacterid. 128: 592-597. 30. Okon, Y., J. P. Houchlins, S. L. Albrecht, and R. H. Burris. 1977. Growth of Spirillum lipoferum at constant partial pressures of oxygen and the properties of nitrogenase in cell free extracts. J. Gen. Microbiol. 98: 87-93. 31. Pedrosa, F. 0., and G. T. Zancan. 197^. L-Arabinose metabolism in Rhizobium japonicum. J. Bacteriol 119: 336-338. 32. Reed, L. J. and C. R. Williams. 1966. Purification and resolution of the pyruvate dehydrogenase complex, p. 258-259. In W. A. Wood (ed.) Methods in Enzymology. V. 9. Academic Press, New York. 33. Reeves, H. C, R. Rabin, W. S. Wegener, and S. J. Aj 1 1971. Assays of enzymes of the tricarboxylic acid and glyoxylate cycles, p. hlS-kil. In J. R. Norris and D. W. Ribbons (ed.) Methods in Microbiology. V. 6A. Academic Press, New York. 34. Rigaud, J., F. J. Bergersen, G. L. Turner, and R. M. Daniel. 1973. Nitrate dependent anaerobic acetylene reduction and nitrogen fixation by Soybean bacteroids. J. Gen. Microbiol. 77: 137-144. 35. Singer, T. P., and T. Cremona. I966. D(-)Lactate cytochrome c reductase, p. 303-305. In W. A. Wood (ed.). Methods in Enzymology V. 9. Academic Press, New York. 36. Smith, I. (Ed.) I96O. Chromatographic and Electrophoretic Techniques. I. Chromatography. Interscience Publishers, Inc., New York. 37. Stoolmiller, A. C, and R. H. Abeles. I966. Formation of alphaketoglutaric semialdehyde from L-2-keto-3-deoxy-arabon i c acid and the isolation of L-2-keto-3-deoxy-a rabonate dehydratase from Pseudomonas saaaharophila. J. Biol. Chem. 241: 5764-5771. 38. Tarrand, J. J., and N. R. Krieg. I978. A taxonomic study of Spirillum lipoferum group, with descriptions of a new genus, Azospirillum gen. nov. and two species, Azospirillum lipoferum (Beijerinck) comb. nov. aud Azospirillum brasilense sp nov. Can. J. Microbiol 24: 967-98O.

PAGE 77

68 39. Veeger, C. D. V. DerVartan i an and W. P. Zey 1 emaker. 1969. Succinate dehydrogenase, p. 81-90. In J. M. Lowenstein (ed.), Methods in Enzymology. V. 13. Academic Press, New York. AO. Weinberg, R. 1959L-2-keto-4,5 d i hydroxyval er i c acid: An intermediate in the oxidation of L-arabinose by Pseudomonas saooharophila. J. Biol. Chem. 2J,k: 5764-5771. Al. Weimberg, R. 1961. Pentose oxidation by Pseudomonas fragi. J. Biol. Chem. 236: 629-63A. kl. Weimberg, R. and M. Doudoroff. 1955The oxidation of Larabinose by Pseudomonas sacaharophita J. Biol. Chem. 217: 607-62A. 43. Weissbach, A., and J. Hurvitz. 1959. The formation of 2-keto3-deoxyhepton i c acid in extracts of Escherichia coli B. J. Biol. Chem. 23^: 705-709.

PAGE 78

BIOGRAPHICAL SKETCH Norman James Novick was born in Portland, Maine, on March 12, 1953. He was graduated from Portland High School in June, 1971, and attended Rensselaer Polytechnic Institute from September, 1971, to June, 1975He received the degree of Bachelor of Science from RPI with a major in biology. After working for the Rensselaer Fresh Water Institute from June to August, 1975, he entered graduate school at the University of Florida in September, 1975Norman received the degree of Master of Science from the Department of Microbiology and Cell Science in March, 1977Since that date he has been pursuing a study of the nitrogen-fixing bacterium Azospiriltum brasilense and is a candidate for the degree of Doctor of Philosophy in December, I98O. 69

PAGE 79

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Arnold S. Bleiweis, Chairman Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
PAGE 80

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1980 Dean ^y^ol 1 ege of Ag^i-culture Dean for Graduate Studies and Research