Biosynthesis of phenylalanine and tyrosine in pseudomonas aeruginosa and zymomonas mobilis

MISSING IMAGE

Material Information

Title:
Biosynthesis of phenylalanine and tyrosine in pseudomonas aeruginosa and zymomonas mobilis molecular cloning of the genes encoding cyclohexadienyl dehydratase and cyclohexadienyl dehydrogenase, and characterization of the gene products
Physical Description:
vii, 120 leaves : ill., photos. ; 29 cm.
Language:
English
Creator:
Zhao, Genshi, 1961-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 111-119).
Statement of Responsibility:
by Genshi Zhao.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001750282
notis - AJG3187
oclc - 26486693
System ID:
AA00003315:00001

Full Text












BIOSYNTHESIS OF PHENYLALANINE AND TYROSINE IN PSEUDOMONAS
AERUGINOSA AND ZYMOMONAS MOBILIS:MOLECULAR CLONING OF THE
GENES ENCODING CYCLOHEXADIENYL DEHYDRATASE AND
CYCLOHEXADIENYL DEHYDROGENASE, AND
CHARACTERIZATION OF THE GENE PRODUCTS








BY



GENSHI ZHAO


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


UNIVERSITY OF FLORIDA


1991




























Dedicated to my father and my mother,
whose love and care gave me the encouragement
to complete this dissertation
















ACKNOWLEDGMENTS


I wish to express my deep and sincere gratitude to Dr.

Roy A. Jensen, chairman of my supervisory committee, whose

tremendous encouragement, invaluable guidance, endless ideas,

constant challenging, critical input, and financial support

made the fulfillment of this study all possible.

I would like to thank Dr. James F. Preston for his

sincere encouragement, critical support, invaluable help,

advice, guidance, and friendship during the course of this

study.

My sincere appreciation is extended to Dr. William B.

Gurley, Dr. Ernest Hiebert, and Dr. Edward M. Hoffmann for

their encouragement, help, advice, guidance, and critical

review of the dissertation.

My special thanks are also extended to Dr. Tianhui Xia,

a productive collaborator, and also a good friend, for his

tremendous help in all phases of this study and especially in

protein purification, and for sharing his knowledge of

enzymology and his philosophy during this study.

I am also thankful to Dr. Randy S. Fischer for his

initial help, Carol Bonner for her help in preparation of the

dissertation and in all phases of this study, and Premila Rao

iii









for providing arogenate, chorismate, and prephenate and for

creating an enjoyable working environment during this study,

Ronlou Doong and Prem Submanamium for sharing their knowledge

of chemistry, and Wei Gu for providing microbial cultures.

I am indebted to my family, especially to my father and

my mother, to whom this dissertation is dedicated. Without

their love, support, and understanding, this study could not

have been accomplished. I am also indebted to my sisters for

taking care of my parents.

Finally, but not least, I wish to express my sincere

appreciation to my wife, Ren Shiling, for her sincere

encouragement, understanding, patience and support during

these years of study, and to my three-month-old daughter,

Lily, for filling the family with all the joy and happiness.


















TABLE OF CONTENTS
page
ACKNOWLEDGMENTS...................................... iii

ABSTRACT........ ...... ................................ vi

CHAPTERS

1 LITERATURE REVIEW............................... 1

2 THE CYCLOHEXADIENYL DEHYDRATASE FROM
PSEUDOMONAS AERUGINOSA:MOLECULAR CLONING OF
THE GENE AND CHARACTERIZATION OF THE GENE
PRODUCT...... ........... ....... .......... ....... 12

Introduction...................................... 12
Materials and Methods............................ 16
Results..................... ............ ......... 23
Discussion....................................... 33

3 THE CYCLOHEXADIENYL DEHYDRATASE FROM
PSEUDOMONAS AERUGINOSA:EVIDENCE FOR A
PERIPLASMIC PROTEIN.............................. 45

Introduction ..................................... 45
Materials and Methods........................... 46
Results........................................... 51
Discussion...... .. .................. ............ 65

4 THE CYCLOHEXADIENYL DEHYDROGENASE FROM
ZYMOMONAS MOBILIS: MOLECULAR CLONING OF
THE GENE AND CHARACTERIZATION OF THE GENE
PRODUCT.................. ........................ 69

Introduction.................................... 69
Materials and Methods......................... 73
Results.............................. ............ 80
Discussion.. ...................................... 91

SUMMARY................... .... ................. 101

REFERENCES........................................... 110

BIOGRAPHICAL SKETCH.................................. 120
















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

BIOSYNTHESIS OF PHENYLALANINE AND TYROSINE IN PSEUDOMONAS
AERUGINOSA AND ZYMOMONAS MOBILIS: MOLECULAR CLONING OF
THE GENES ENCODING CYCLOHEXADIENYL DEHYDRATASE AND
DEHYDROGENASE, AND CHARACTERIZATION OF THE GENE PRODUCTS

BY

GENSHI ZHAO

December, 1991


Chairman: Roy A. Jensen
Major Department: Microbiology and Cell Science

The gene encoding cyclohexadienyl dehydratase (CDT),

denoted as pheC, was cloned from Pseudomonas aeruginosa. The

pheC product produced in Escherichia coli was purified and

showed identical physical and biochemical properties as those

obtained for CDT purified from P. aeruginosa. The CDT

exhibited K. values of 0.42 mM for prephenate and 0.22 mM for

arogenate, respectively. The DheC gene was 807 bp in length,

predicating a protein of 30,480 Mr. This compares with a Mr

value of 28,000 determined for the purified enzyme by SDS-

PAGE. Comparison of the CDT sequence of P. aeruginosa with

three P-proteins (chorismate mutase/prephenate dehydratase)

did not establish obvious homology. The N-terminal sequencing

revealed that the first 11 residues of the purified pheC gene

vi









product matched the residues 26 to 36 from the translation

start site deduced from the nucleotide sequence, indicating

that a 25 residue amino peptide was cleaved in E. coli. The

CDT from P. aeruginosa was then purified and exhibited an

identical molecular weight and sequence as did the pheC

product expressed in E. coli, thus showing that the cleavage

of the amino-peptide occurred in P. aeruginosa. Chloroform and

osmotic shock treatments released 95% of CDT activity from the

periplasmic spaces of P. aeruginosa and E. coli, respectively,

thus demonstrating that the CDT is a periplasmic protein.

The cyclohexadienyl dehydrogenase (CDH) gene, denoted as

tyrC, was cloned from Zvmomonas mobilis. The tyrC gene was 882

bp in length, encoding a protein of 32,086 Mr. The tyrC

product formed in E. coli was purified, and estimated to be

32,000 by SDS-PAGE. The activity ratios of arogenate

dehydrogenase (ADH) to prephenate dehydrogenase (PDH) (3:1)

remained constant throughout purification, indicating the two

activities were inseparable. K. values of 0.4 mM and 0.33 mM

were obtained for prephenate and arogenate, respectively. A k,

value of 0.11 mH for NAD+ was obtained when the enzyme was

assayed as ADH or PDH. Unlike the CDHs of E. coli and P.

aeruqinosa, that of Z. mobilis is not sensitive to tyrosine

inhibition. In pairwise alignments, the Z. mobilis CDH showed

a partial identity with the PDHs of Bacillus subtilis (32%),

Saccharomyces cerevisiae (19%), and the T-proteins (chorismate

mutase/CDH) of E. coli (21%) and Erwinia herbicola (23%).

vii

















CHAPTER 1
LITERATURE REVIEW


The Aromatic Pathway

Biosynthesis of the aromatic amino acids begins with the

condensation of erythrose-4-phosphate and phosphoenol pyruvate

by 3-deoxy-D-arabino-heptulosonate-7-phosphate (DAHP) synthase

to form chorismate, the last intermediate common to all

aromatic amino acids (Fig. 1-1). The pathway branches at

chorismate to L-tryptophan, and L-phenylalanine and L-

tyrosine. Chorismate, the major branchpoint metabolite, is

then converted by chorismate mutase to prephenate, forming a

unique precursor for L-phenylalanine and L-tyrosine. The

conversion of prephenate to phenylpyruvate via prephenate

dehydratase or to 4-hydroxyphenylpyruvate via prephenate

dehydrogenase was assumed to be the universal enzymatic steps

functioning in nature (Cotton and Gibson, 1965). In 1974, an

alternative route to L-tyrosine was discovered in

Cyanobacteria in which prephenate is transaminated to L-

arogenate (Stenmark et al., 1974). L-Arogenate is then

converted to L-tyrosine by arogenate dehydrogenase (Stenmark

et al., 1974).


-1-













1 C co >
m <, ea >



SI 4J I k
M>4 a4 0,
> ( '0 4 U




-M ,, ,U Er
* 0 10 4 -
t ,,- ,.Q.

5 9 (VC >
.0 00






*1 CA 0
* I0 0










0 -0>
r44







0 0

4. 44 "
,CH Ic e
0 '0






*O 0 > 0

>000
i 0 0 0


0 (o ,o
44 > 0 4J
. 0 0


4 0 a0
(0-3 0 J 0 C to

P ) ( *Hv




0 E
O 0 C
H4 aU)rIl U) ;.










.U) 43 4.J
0 .0 4J a)



*r4 .(0 4) 'U 0
S040 4J, g4














-3-


0
o

z-z





Z




0 Z



0
o o
O=U 00




*



0 ,


0L x
0= o


0
0


z-u











I O
Q0

o i

o=u 0








<- Ia)-


4J,










0
r0
S=u



a o o
- '- o-


II

0Z1
I')









-4-

L-Arogenate was initially identified as a unique

precursor for L-tyrosine biosynthesis (Stenmark et al., 1974)

in microorganisms, and now its role as a precursor for both L-

tyrosine and L-phenylalanine has been established not only in

microorganisms but also in higher plants (Byng et al., 1982).



The Dual Pathways to L-Phenylalanine and L-Tyrosine in
Pseudomonas aeruginosa and Zymomonas mobilis

The biosynthetic pathways to L-phenylalanine and L-

tyrosine have proven to be unexpectedly diverse in nature

(Byng et al., 1982). This diversity generates a range of

alternative character states that has greatly facilitated

tracing the evolutionary history of aromatic biosynthesis

within groups of phylogenetically related organisms (Jensen,

1985). P. aeruainosa is an opportunistic human pathogen which

belongs to Superfamily B, a lineage where evolutionary history

of aromatic biosynthesis has been studied in a greater detail

than in any other phylogenetic grouping (Jensen, 1985). The

disease of cystic fibrosis is the outcome of P. aeruinosa

infection and a genetic defect in human chromosome (Rommens et

al., 1989; Riordan et al., 1989; and Kerem et al., 1989).

Dual enzymatic routes to L-phenylalanine and L-tyrosine

were first described in P. aeruainosa (Patel et al., 1977). A

bifunctional P-protein leading to the biosynthesis of L-

phenylalanine consists of chorismate mutase-P and prephenate

dehydratase (Patel et al., 1977). The prephenate dehydratase









-5-

activity of the bifunctional P-protein is feedback inhibited

by L- phenylalanine, but activated by L-tyrosine (Patel et

al., 1977). In addition to the bifunctional P-protein, an

unregulated route to L-phenylalanine is also present, which

consists of monofunctional chorismate mutase-F, aromatic

aminotransferase, and cyclohexadienyl dehydratase, whose

catalytic reactions are often referred to as overflow pathway

(Fiske et al., 1983). The cyclohexadienyl dehydratase utilizes

prephenate and L-arogenate as alternative substrates (Patel et

al., 1977, Zhao et al., 1992a). Thus, L-phenylalanine can be

synthesized through the intermediate phenylpyruvate or L-

arogenate. The dual pathways to L-phenylalanine are apparently

due to substrate ambiguity of the cyclohexadienyl dehydratase.

Dual pathways to L-tyrosine biosynthesis also exist in P.

aeruginosa. These consist of the chorismate mutase-F, aromatic

aminotransferase, and cyclohexadienyl dehydrogenase (Patel et

al., 1977). As noted above, the chorismate mutase-F, and the

aminotransferase are also participating in L-phenylalanine

over-flow pathway (Fiske et al., 1983). The cyclohexadienyl

dehydrogenase catalyzes the conversion of prephenate to 4-

hydroxyphenylpyruvate and arogenate to L- tyrosine (Patel at

el., 1977). Thus, L-tyrosine can be synthesized via

intermediates of arogenate and 4-hydroxyphenylpyruvate.

Recently, the cyclohexadienyl dehydrogenase has been purified

from P. aeruinosa. An identical K, value was obtained

regardless whether the enzyme was assayed as prephenate









-6-

dehydrogenase or as arogenate dehydrogenase, showing that a

single NAD-linked cyclohexadienyl dehydrogenase exists which

can utilize either arogenate or prephenate as substrate (Xia

and Jensen, 1990). The enzyme is sensitive to feedback

inhibition by L-tyrosine (Patel et al., 1977; and Xia and

Jensen, 1990).

Zvmomonas mobilis is a gram-negative bacterium and also

a member of Superfamily C (Montenecourt, 1985). It exhibits

high ethanol tolerance and is capable of rapid ethanol

production, which makes it potentially useful for commercial

production of ethanol (Montenecourt, 1985). The enzymic

arrangement employed by this organism for the biosynthesis of

L-phenylalanine and L-tyrosine is essentially unknown. Some

studies (Bhatnagar and Jensen, unpublished) suggest that Z.

mobilis possesses a chorismate mutase-F, prephenate

dehydratase, arogenate dehydratase, and cyclohexadienyl

dehydrogenase. Thus, the dual pathways to L-phenylalanine and

L-tyrosine may also coexist. The presence of the

cyclohexadienyl dehydrogenase in this organism has recently

been established by Zhao et al. (1992b).



Evolutionary Scenario

The P-proteins from Escherichia coli (Davidson et al.,

1972; and Gething and Davidson, 1976), Alcaliaenes eutrophus

(Friedrich et al., 1976), and Acinetobacter calcoaceticus

(Ahmad et al., 1989) have been purified and characterized









-7-
thoroughly. The P-protein from E. coli has a native molecular

weight of 85,000, and is composed of two identical subunits of

40,000 (Davidson et al., 1972; and Duggleby et al., 1978). The

native form of enzyme tends to form tetramers at high protein

concentration or in the presence of feedback inhibitor of

phenylalanine (Dopheide et al., 1972). Evidence from chemical

modification (Gething and Davidson, 1977a; and 1977b) and

kinetic studies (Duggleby et al., 1978) strongly suggests that

the P-protein has two separate catalytic sites for chorismate

mutase and prephenate dehydratase. Both activities are

feedback inhibited by L-phenylalanine (Gething and Davidson,

1977a; and 1977b). Mutant strains of E. coli which were

defective in prephenate dehydratase, or chorismate mutase-P,

or both activities, have been isolated (Baldwin and Davidson,

1981). These genetic studies further support the existence of

independent catalytic sites for prephenate dehydratase and

chorismate mutase-P.

Bifunctional T-proteins for L-tyrosine biosynthesis

consist of chorismate mutase-T and cyclohexadienyl

dehydrogenase. Bifunctional P-proteins are of ancient origin,

and are apparently distributed throughout all of Superfamily

B and the connecting Superfamily A (Ahmad and Jensen, 1986).

On the other hand, bifunctional T-proteins are of more recent

origin, and are distributed only in some species of

Superfamily B (Ahmad and Jensen, 1986; and Jensen, 1985).









-8-

The T-proteins from E. coli (Hudson et al., 1984; and

Koch, 1971) and Aerobacter arogenes (Koch, 1972) have been

purified, and they consist of two identical subunits each with

an estimated molecular weight of 39,000. Subsequently,

kinetic, chemical modification, and mutational studies have

suggested that the two reactions of the T-proteins occur at

one active site or overlapping active sites (Hudson et al.,

1984; Heyde, 1979; and Rood et al., 1982). Recently, in vitro

separation of the two activities of T-protein from E. coli has

been achieved genetically by Maruya et al. (1987), an

indication that the T-protein also has two separate catalytic

sites.

No P-proteins have been reported to utilize arogenate.

However, the dehydrogenase component of T-proteins in enteric

bacteria has been found to utilize both prephenate and

arogenate (Ahmad and Jensen, 1987). The failure of the P-

protein dehydratase to utilize arogenate might suggest that

the dehydratase component of the P-proteins has lost the

substrate ambiguity of the ancestral cyclohexadienyl

dehydratase during evolution. The utilization of arogenate by

T-proteins, on the other hand, may indicate that T-proteins

did not have enough evolutionary time for the loss of

recognition of arogenate as substrate since, compared with P-

protein, T-protein is of relatively more recent origin.

The P-protein and T-protein genes from E. coli have

been cloned and sequenced, and the two genes are contiguous









-9-

(Hudson and Davidson, 1984). Considerable similarity was

observed between N-terminal portions of the two proteins,

indicating that the components of chorismate mutase are

located on their N-termini. Recently, the prephenate

dehydratases from Corynebacterium alutamicum and Bacillus

subtilis were found to share certain similarities with the C'-

terminal portion of the P-proteins from E. coli and P

stutzeri (Fischer et al., 1992; Follettie and Sinskey, 1986;

Hudson and Davidson, 1984; and Trach and Hoch, 1989),

suggesting that the prephenate dehydratase component of P-

protein resides on its C-terminus, and they may share a common

evolutionary ancestor. More recently, it has shown that no

sequence homology has been found between the chorismate

mutase-F from Bacillus subtilis (Gray et al., 1991) and either

P-protein or T-protein from E. coli (Hudson and Davidson,

1984). It would be interesting to see how the B. subtilis

bifunctional chorismate mutase is related to the P-protein and

T-protein from E. coli.

In view of all the evidence, it is reasonable to assume

that T-protein evolved via the fusion of the gene encoding

cyclohexadienyl dehydrogenase with one of two duplicates of

the gene specifying chorismate mutase-F. Firstly, the

dehydrogenase component of the T-proteins studied so far

utilizes both prephenate and arogenate, functioning as a

cyclohexadienyl dehydrogenase. Secondly, the presence of T-

proteins in enteric bacteria is correlated with the absence of









-10-

cyclohexadienyl dehydrogenases. The presence of

cyclohexadienyl dehydrogenases in Superfamilies B and C is

also correlated with the absence of T-proteins. Finally, the

dehydrogenase domain of T-proteins and many cyclohexadienyl

dehydrogenases are feedback inhibited by L-tyrosine.

The scenario for the P-proteins is more complex. The P-

proteins could have originated from chorismate mutase-F and

cyclohexadienyl dehydratase. If so, the presence of

cyclohexadienyl dehydratase in P. aeruainosa and many other

members of Superfamily B suggests that the gene encoding this

enzyme must have duplicated before gene fusion. However, the

absence of cyclohexadienyl dehydratase in some members of

Superfamily B suggests another possibility. Perhaps this gene

was duplicated before fusion but then lost due to mutation. It

is also very possible that P-proteins evolved from the fusion

of the genes coding for prephenate dehydratase and chorismate

mutase-F. Firstly, the presence of P-proteins in Superfamilies

A and B is correlated with the absence of prephenate

dehydratases. Secondly, the dehydratase component of P-

proteins and many monofunctional prephenate dehydratases, but

not cyclohexadienyl dehydratases, are feedback inhibited by L-

phenylalanine, and some of them are allosterically activated

by L-tyrosine. Thirdly, P-proteins examined so far possess

prephenate dehydratase activity but not arogenate dehydratase

activity.









-11-

One of the objectives of this study was to clone and

characterize the genes coding for cyclohexadienyl dehydratase

from P. aeruginosa and cyclohexadienyl dehydrogenase from Z.

mobilis and products. Another objective was to test the

proposed evolutionary hypotheses, one being that P-proteins

evolved from the gene fusion of monofunctional chorismate

mutase and cyclohexadienyl dehydratase, and another being that

T-proteins evolved from the gene fusion of monofunctional

chorismate mutase and cyclohexadienyl dehydrogenase.
















CHAPTER 2
CYCLOHEXADIENYL DEHYDRATASE FROM PSEUDOMONAS AERUGINOSA:
MOLECULAR CLONING OF THE GENE AND CHARACTERIZATION OF
THE GENE PRODUCT


Introduction

Biosynthesis of the aromatic amino acids begins with the

condensation of erythrose-4-phosphate and phosphoenolpyruvate

by DAHP synthase to form chorismate, the last intermediate

common to all aromatic amino acids (Weiss and Edwards, 1980).

Chorismate is then converted to prephenate via chorismate

mutase, forming an initial precursor that is unique for the

biosynthesis of L-phenylalanine and L-tyrosine in

microorganisms (Weiss and Edwards, 1980). The conversion of

prephenate to phenylpyruvate via prephenate dehydratase or to

4-hydroxyphenylpyruvate via prephenate dehydrogenase was well

established first in microorganisms and thought to be the

universal enzyme steps functioning in nature (Cotton and

Gibson, 1965). In 1974, an alternative route to L-tyrosine

biosynthesis was established in cyanobacteria in which

prephenate was converted to arogenate via prephenate

aminotransferase rather than to 4-hydroxyphenylpyruvate via

prephenate dehydrogenase (Stenmark et al., 1974). Arogenate

was initially identified as a precursor of L-tyrosine, and now

its function as a precursor of L-phenylalanine has become

-12-









-13-

apparent in both microorganisms and higher plants (Byng et

al., 1982).

The pathway to phenylalanine biosynthesis in

microorganisms has been demonstrated to be complex and diverse

not only because of the presence of the alternative pathway

but also because of the presence of the different combinations

of enzymes that catalyze the overall reactions. For example,

Escherichia coli and other Gram-negative bacteria synthesize

L-phenylalanine via phenylpyruvate utilizing a bifunctional P-

protein (Weiss and Edwards, 1980). The bifunctional P-protein

catalyzes the overall conversion of chorismate to prephenate

and prephenate to phenylpyruvate. On the other hand, other

microorganisms such as Pseudomonas diminuta synthesize L-

phenylalanine via arogenate using the alternative route (Zamir

et al., 1985). In these organisms, three monofunctional

enzymes, chorismate mutase-F, prephenate aminotransferase, and

arogenate dehydratase, are utilized to carry out the

conversion of chorismate to prephenate, prephenate to

arogenate, and arogenate to L-phenylalanine. P. aeruginosa,

however, was the first example of microorganisms possessing

the dual pathways to L-phenylalanine biosynthesis (Patel et

al., 1977) (Fig. 2-1). In addition to the bifunctional P-

protein, P. aeruginosa possesses chorismate mutase-F, and

cyclohexadienyl dehydratase which consist of the overflow

pathway (Fiske et al., 1983). The cyclohexadienyl dehydratase

of P. aeruginosa has a broad substrate specificity that











0 C M- IC O I
E- (0 0 -4 M
V rC. A u


C a .0 i c
is ku --.i 0









?O o 0" .
0A 0 ) H

0 .. 4C
4 J rf M.



3 > 0 0
0 1 C r -. >




























034 t f-4
goZto to


































0 0 *
HM u 0 4) 10

































I 4. tNi 0
a* f 0s V> VH 0a
















t> a > S 3O
S0 0 E
1 0 ) 4-) 43 0- 1



rz U 3j V







C-H (0 EA frl S
0 a U ( 4 )


5 rito 06



:3 V a -oq 0 V
$4 aPC Ng I-u ..



4-) f ) U >il C-l '00







0 Q ( o to 9 Vt













-15-


=

I I
z-o


8
I
00O


I


z-o
1


T_
n









-16-

accommodates prephenate or arogenate as substrate for L-

phenylalanine biosynthesis (Patel et al., 1977; and Zhao et

al., 1992). The overflow pathway to L-phenylalanine

biosynthesis is widely spread in Superfamily B microorganisms

(Ahmad and Jensen, 1988a).

The physiological role of the overflow pathway in nature

is essentially unknown, and the evolutionary relationship of

the cyclohexadienyl dehydratase to the monofunctional

prephenate dehydratases and the dehydratase domain of the

bifunctional P-proteins is unclear. In this chapter, I

describe the molecular cloning and nucleotide sequence of the

cyclohexadienyl dehydratase gene from P. aeruainosa, and the

purification, characterization, and expression of its gene

product.



Materials and Methods

Strains. Plasmids. and Media

All bacterial strains and plasmids used in this study are

listed in Table 2-1.

LB (Maniatis et al., 1982) was used as a rich growth

medium, and M9 (Maniatis et al., 1982) was used as a minimal

medium for coli and Pseudomonas aeruainosa. Ampicillin (50

ug/ml), tyrosine (50 ug/ml), and thiamine (17 ug/ml) were

supplemented when appropriate. Agar was added at 15 g/liter

for plates.










-17-


Table 2-1. Bacterial strains and plasmids

Strain/plasmid Genotype or description Source


E. coli K-12
JM83

JP2255


KA197


P. aeruainosa
PAO1

Plasmids
pUC18
pUC19
pJZIa


pJZla



pJZlb



pJZlc



pJZld-S


pJZld-O


pJZle


pJZlf


pJZIg


ara (proAB-lac) rpsL
thi80 dlac ZM15
aroF363 p eA361 phe0352
tyrA382 thi strR712
lacYl xy15

thil pheA97 relAl sDoT1


Prototroph


Apr lacZ
AP- lacZ
Original clone carrying
pheC gene isolated from
PAO1 library
2.0 kb derivative of pJZl
generated by removal of a
3.8 kb EcoRI fragment

4.0 kb derivative of pJZl
generated by removal of a
1.8 kb KpnI fragment

4.5 kb derivative of pJZl
generated by removal of a
1259 bp SmaI-SmaI fragment

1.8 kb KpnI fragment of pJZl
cloned into pUC18 with the
same orientation as pJZl

1.8 kb KpnI fragment of pJZl
cloned into pUC18 with the
opposite orientation of pJZl

741 bp SmaI fragment of
pJZld-S cloned into pUC18

515 bp SmaI fragment of
pJZld-S cloned into pUC18

1110 bp SphI-SmaI fragment
of pJZld-S cloned into pUC18


Yanisch-
Perron

Baldwin &
Davidson

CGSC 5243


Holloway


BRL


This study


This study



This study



This study



This study



This study


This study


This study


This study









-18-


Isolation of P. aeruqinosa (PAO1) Chromosomal DNA and
Construction of a Gene Library

Chromosomal DNA was isolated as described previously

(Marmur, 1961). This DNA was partially digested with Sau3A,

and fragments of 5 to 10 kilobases (kb) were isolated from

agarose gels after electrophoresis. A library was constructed

by ligation of these fragments into the dephosphorylated BamHI

site of pUC18 (Yanisch-Perron et al., 1985). The ligation

mixtures were transformed into E. coli JM83 (Yanisch-Perron et

al., 1985). The transformants obtained on LB plates

(supplemented with 50 ug/ ml ampicillin) were collected and

stored in 50% glycerol at -750C. Recombinant plasmids were

purified from this library as described (Davis et al., 1980).

The recombinant plasmids were used to transform E. coli heA

mutant of JP2255 (Baldwin and Davidson, 1981) in order to

select for clones carrying the pheC gene.

DNA Manipulations

All restriction enzymes, T4 DNA ligase, and calf

intestine phosphatase were obtained from Bethesda Research

Laboratories and Promega, and were used according to the

manufacturer instructions. Analyses of restriction sites and

subcloning were conducted by standard methods (Maniatis et

al., 1982). Southern blot hybridization, using biotinylated

probes, was conducted following the instructions of Promega.









-19-

DNA Sequencing and Data Analysis

The subclones of pJZle, pJZlf, and pJZlg were first

purified in a large scale by CsCl gradient centrifugation

(Humphreys et al., 1975), and then sequenced in both

directions as described (Prober et al., 1987) at the DNA Core

Facility of the Univ. of Florida. Nucleotide sequence of the

gene along with the deduced amino acid sequence was analyzed

by using the Univ. of Wisconsin Genetics Computer Group (GCG)

package (Devereux et al., 1984).

Crude Extract Preparation and Enzyme Assay

Cultures of E. coli JP2255 transformed by various clones

were grown at 37C in 450 ml LB supplemented with ampicillin,

and harvested by centrifugation during late exponential phase

of growth. The cells were suspended in 3 ml of 50 mM potassium

phosphate buffer, pH 7.5, and disrupted by sonication. The

resulting suspension was centrifuged at 150,000 g for 60 min

at 4C. The supernatant fraction collected was passed through

a PD-10 Sephadex column to remove small molecules, and used

for enzyme assay.

Prephenate dehydratase was assayed as described (Cotton

and Gibson, 1965). A reaction mixture (200 ul) contained 1.0

mM prephenate, enzyme and 50 mM potassium phosphate (pH 7.5).

After incubation at 370C for 20 min, 0.8 ml of 2.5 N NaOH was

added, and phenylpyruvate was measured spectrophotometrically

at 320 nm. An extinction coefficient of 17,500 was used for

calculation of phenylpyruvate formation (Cotton and Gibson,









-20-
1965). Arogenate dehydratase was assayed as described

previously for the formation of phenylalanine using HPLC (high

performance liquid chromatography) (Zamir et al., 1980). A

reaction mixture of 200 ul contained 1.0 mM arogenate, enzyme

and 50 mM potassium phosphate (pH 7.5). After incubation at

37C for 20 min, the phenylalanine formed was measured using

authentic phenylalanine as a standard. Protein was measured as

described by Bradford (1976). One unit of enzyme activity was

defined as the formation of 1 nanomole of phenylpyruvate or

phenylalanine per min at 370C.



Purification of The pheC Product from E. coli JP2255 and
The Cyclohexadienyl Dehvdratase from P. aeruginosa

The E. coli JP2255 carrying pJZlg was grown in 1000 ml LB

supplemented with 50 ug/ml of ampicillin at 37C, and

harvested by centrifugation during late exponential phase of

growth. The cells were washed once with 20 mi potassium

phosphate, 1 mH DTT, pH 8.5 (Buffer A), resuspended in the

same buffer, and disrupted by sonication. The resulting

suspension was centrifuged at 150,000 g for 60 min. The

supernatant fraction collected was applied to a DEAE-cellulose

column (2.5 x 30 cm) that was pre-equilibrated with Buffer A.

The column was first washed with 100 ml of Buffer A, and then

eluted with a 1,000 ml linear KC1 gradient from 0 mM to 300 mM

in Buffer A. The fractions of 2.8 ml were collected. The

fractions showing high cyclohexadienyl dehydratase activity

were pooled, and concentrated by using an Amicon-PM 10









-21-
membrane. The concentrated preparation was washed twice with

20 mM potassium phosphate, 1 mM DTT, pH 7.2 (Buffer B),

and then applied to a hydroxylapatite column (1.5 x 15 cm)

which was equilibrated with Buffer B. The column was eluted

with a 600 ml linear gradient of phosphate from 20 mM to 300

mM. Fractions of 2.8 ml were collected, and those showing high

cyclohexadienyl dehydratase activity were pooled. The pooled

fractions were concentrated as described before, and then

applied to a Sephadex G-200 column (2.5 x 98 cm) previously

equilibrated with Buffer B. The column was eluted with

Buffer B, and the fractions exhibiting cyclohexadienyl

dehydratase activity were collected for further study.

P. aeruainosa PAO1 (Holloway, 1955) was grown in a M9

minimal medium (3,000 ml). The culture was harvested by

centrifugation during late exponential phase of growth. The

crude extract was prepared as described before. The

supernatant fraction collected after ultra-centrifugation was

adjusted to 40% ammonium sulfate saturation with solid salt.

After stirring for 10 min, the pellet was removed by

centrifugation. The supernatant fraction was collected and

brought to 60% saturation. The pellet collected was dissolved

in 20 ml of Buffer A, and dialyzed against several changes of

20 volumes of the same buffer. The dialyzed enzyme preparation

was applied to a DEAE-cellulose column and purified as

previously described under identical conditions.









-22-


Characterization of The pheC Product from E. coli and the
Cyclohexadienvl Dehydratase from P. aeruginosa

The subunit molecular weight of the purified protein was

determined by sodium dodecyl sulfate- polyacrylamide gel

electrophoresis (SDS-PAGE) (Laemmli, 1970). The molecular

weight of the native proteins was determined by gel filtration

(Sephadex G-200) as described above for enzyme purification.

Biochemicals and Chemicals

Prephenate was prepared from a tyrosine auxotroph of

Salmonella typhimurium (Dayan and Sprinson, 1970), and

arogenate was prepared from a triple auxotroph of Neurospora

crassa (Zamir et al., 1980). Sephadex G-200 was obtained from

Sigma Chemical Company, DEAE-cellulose was obtained from

Whatman, and hydroxylapatite was obtained from Bio-Rad.

Molecular weight standards for SDS-PAGE (alpha-lactalbumin,

14,400; soybean trypsin inhibitor, 20,100; carbonic anhydrase,

30,000; ovalbumin, 43,000; bovine serum albumin, 67,000; and

phosphorylase, 94,000), and for gel filtration (carbonic

anhydrase, 30,000; bovine serum albumin, 67,000; alcohol

dehydrogenase, 150,000; and b-amylase, 200,000) were obtained

from Pharmicia Fine Chemical and Sigma Chemical Company,

respectively. All the chemicals were purchased from Sigma.









-23-


Results
Cloning of The Gene Encoding Cyclohexadienvl Dehydratase

Approximately 4000 recombinants were obtained after

transformation of E. coli JM83. Purified plasmids from these

recombinants were used to transform E. coli JP2255. The

transformants of JP2255 were first allowed to grow in LB

overnight at 37C, washed twice with liquid M9 medium, and

then plated on M9 plates which were supplemented with L-

tyrosine and ampicillin. The plates were examined for the

presence of colonies 4 days after incubation at 370C. Two

colonies were observed. Plasmids were purified from cultures

derived from each of these transformants. The two plasmids

isolated, designated as pJZl and pJZ2, respectively, Were

found to be able to transform JP2255 to L-phenylalanine

prototropy. The transformants of JP2255 carrying the clones of

pJZl and pJZ2 grew slowly on M9 plates. However, the

transformation of pJZl or pJZ2 into KA197, a pheA mutant,

resulted in a relatively faster growing phenotype, apparently

due to the presence of chorismate mutase encoded by tyrA.

The presence of prephenate dehydratase and arogenate

dehydratase activities was examined in the crude extracts of

E. coli JP2255 carrying pJZl and pJZ2. A high level of both

activities was observed in the crude extract of E. coli

JP2255 carrying pJZl and pJZ2 while no enzyme activity was









-24-

detected in the crude extract of E. coli JP2255 carrying

pUC18 (Table 2-2).

Digestion of pJZ1 and pJZ2 with HindIII, PstI, and KpnI

showed that they carried two identical DNA fragments estimated

to be 5.7 kb in length. The plasmid designated as pJZl was

used for further study.

Southern blot hybridization analysis showed that, when

the SmaI-EcoRI fragment of pJZl was biotinylated, it

hybridized with a 4 kb fragment of P. aeruainosa chromosomal

DNA which was completely digested with SmaI and EcoRI;

however, it did not hybridize with E. coli chromosomal DNA

which was also digested with SmaI and EcoRI (data not shown).


Localization of the P. aeruginosa pheC Gene and Expression
of the pheC Gene in E. coli

Digestion and religation of pJZl with EcoRI, KpnI, and

SmaI yielded subclones of pJZla, pJZlb, and pJZlc,

respectively (Fig. 2-2a). These subclones were unable to

complement E. coli JP2255 and E coli KA197, and

cyclohexadienyl dehydratase activity was not detected in crude

extracts of the transformants carrying the subclones (Table 2

-2), indicating that the gene may be located within the KpnI

fragment. The KpnI fragment (Fig. 2-2a) was cloned into pUC18

at the KpnI site in both orientations. When the KpnI fragment

was cloned in the same orientation as the original clone

(designated as pJZld-S), the same high levels of enzyme

activities conferred by pJZl were observed (Table 2-2).










-25-


Table 2-2. Expression of the P. aeruainosa ph gene in E
coli JP2255


Specific activity+
Plasmids
in strain
JP2255 Prephenate dehydratase Arogenate dehydratase


pJZl 621 191

pJZ2 700 211

pJZla 0.0 0.0

pJZlb 0.0 0.0

pJZlc 0.0 0.0

pJZld-S 628 186

pJZld-O 5.6 1.7

pJZle 0.0 0.0

pJZlf 0.0 0.0

pJZlg 1314 387


+Specific activity is defined as nmole phenylpyruvate or
phenylalanine formed per min per mg of protein. All the clones
or subclones listed were first transformed into L. coli
JP2255, and the crude extracts of the transformed E. coli
JP2255 were used for enzyme assay.









-26-

However, when this fragment was cloned in the opposite

orientation (designated as pJZld-O), relatively lower levels

of the activities were observed (Table 2-2). These results

showed that the over-expression of the enzyme activity of

these clones was largely dependent on the LacZ promoter of the

plasmid, and also indicated that the promoter of the cloned

gene was functioning in E. coli although not efficiently.

Further localization of the gene was carried out by

cloning the two Smal fragments (released upon digestion of

pJZld-S with Smal) into pUC18 at SmaI site. The two resulting

subclones, designated as pJZle, and pJZlf (Fig. 2-2a), were

unable to complement the defects of E. coli strains JP2255 and

KA197, suggesting the gene was localized within the two SmaI

fragments. In order to obtain an intact SphI-SmaI-SmaI

fragment, the original clone was first completely digested

with SphI, and then partially digested with SmaI. The SphI-

SmaI-SmaI fragment was isolated, and cloned into pUC19,

yielding a subclone denoted pJZlg. The subclone pJZlg

complemented E coli strains JP2255 and KA197, and exhibited

a 2-fold increase in enzyme activities when compared with

pJZld-S (Table 2-2). The increased enzyme activity was

probably due to the decrease of the distance between lacZ

promoter and the transcriptional start site of the gene.













-27-


COMPLEMENTATION
A OFphe A
rC Co 0 Co

pJZ1 LL I I I I I YES
1 2 3 4 5 Kb

pJZ1a NO

pJZ1b NO

pJZlc NO
ORF
pJZld YES

pJZle NO

pJZ1f NO

pJZ1g YES

500 bp



B

___ E
S ---0----- E
0U) ORF U)0




c- < ---
,R I C








100 bp







Fig. 2-2. A. Physical localization of the P. aeruqinosa phC
gene. Linear maps of the clone and subclones are shown, and
their ability to complement the pheA mutants of E. coli JP2255
and KA197 is also indicated. Restriction sites are labeled in
the cloned P. aeruainosa DNA and in the flanking pUC18
multiple cloning site (depicted by heavy lines).
Abbreviations: ORF, open reading frame. B. The sequencing
strategy used for the pheC gene.









-28-


DAN Seauence of the P. aerucinosa pheC Gene and Its Flanking
Regions

The complete nucleotide sequence of the 1.3 kb SmaI

fragment (Fig. 2-2b) is presented in Fig. 2-3. The structural

gene coding for cyclohexadienyl dehydratase was located within

a single open reading frame, 807 bp in length. The deduced

amino acid sequence presented in Fig. 2-3 indicates that the

P. aeruginosa cyclohexadienyl dehydratase contained 268

residues with a molecular weight of 30,480. This compares with

a value of 29,500 determined for the purified enzyme by SDS-

PAGE (Fig. 2-4). The sequence AGGAG, located 6 bp upstream of

the start cordon is presumably the ribosome binding site

(Shine and Dalgarno, 1974). The open reading frame was

terminated by a TAA codon.

The G+C content of the pheC gene was 65.6% which falls

within the 60.6%-66.3% range for P. aeruginosa genome (West

and Iglewski, 1988). The codon usage of the gene was typical

of P. aeruginosa (West and Iglewski, 1988), exhibiting a

striking preference for G or C in the third base position in

91.1% of the codons. As is the case for most P. aeruinosa

genes, C (52.4%) was utilized more frequently than G (38.7%)

in the third position.

A portion of an unidentified open reading frame encoding

a truncated peptide of 74 residues (Fig. 2-3) was found

upstream of the pheC gene. A search of Genbank sequences did

not reveal obvious homology with any known protein sequences.














-29-


10 30 50
GGATCAGCTTCCCGGCCTACCGGGCACGGCCTGGAGATGCTGCTGCGACCACC
G A 8 R PTR 8 TA R C CC A T T R
70 90 110
AATGGCTGCAGGGOGTACCGCTOTCGATOGCGGTGTAGGTCGTCAGCCGTTTCOCGCAC
N C 0 RCR WRC V V 8 A
130 150 170
TTTTTTCCCTCCTTCCTOCCGCACTCOOCCCGCOCCCOCGOTCATCOGCOGTTCCCC
S LL L P A R P A P R R R A P
190 210 230
TGCATTCCGOGATTTOGCCGCGOCTOCCOACTTGCOTATCTCTTGCTCCCCATCC
CI r R GC
250 270 290
CGA OTCGCCATGCCGAAGTCATTCCCCCATCTCGTCCAGGCCCTGGCCTCCTTOC
S K r R L V Q A L A C L A
310 330 350
CTOCTGGCCACGOCCAGCCTCCAOGCGCAGGAGAGCCCCTCACCOCATCCTCAAAGC
L L A 8 A S L Q A Q B 8 R L D I L
370 390 410
O0COTOCTOCOCOTCACCACCACTGGCGACTACAAOCCCTTCACTACCOCAC0GAAGA
SVL R V T T G D Y P F 8 Y R
430 450 470
OOCOQTTACOCCGOTTTCOACGCOGTCATGCOCCAGOCCTOCOCCOAGACCTGOCC
S0 TA D V D D A Q R L A R L A
490 510 530
AAOCTOOTAOTGOTOCCGACC AQTTGGCCGAACCTGATOGCOATTTCOCCOCACCC
SL V V VP W P N L RD ADD
550 570 590
TTCGACATCOCCATGAOCCATCTCOATCAACCTOACGCCAGCOCACOCATTTC
F D I A M G 8S I N L E R Q R QA 8 F
610 630 650
TCGATTCCCTACCTGCGCAACAGCAAGACGCCGATCACCCTCTGTAGCGAAGAAGCGT
S I P Y L R N S K T P I T L C S E A R
670 690 710
TTCCAGACCCTGOAGCAGATCGACCAGCCGGGCTGACGGCCATCGTCAACCCCGGCGC
F Q T L I D Q P GV T A I VN P G
730 aIa 750 770
ACCAACGAGAAGTTCO=CCG CGAACCTGAAGAAGGCCCGGATCCTOGTGCATCCGGAC
TN K ARA L K A R I L V H P D
790 810 830
AACTGACGATCTTCCAGCAGATCGTCGACGGCAAGGCCGACCTGATGATGACCGACGCC
NVT I QQ V D G K A D L TDA
850 870 890
ATCGAGGCCCGCCTGCAGTCGCGTCTGCACCCGGAACTCTGCGCCGTGCATCCGCAGCAA
I EAR L Q S R L H P EL C A V H P
910 930 950
CCCTTCOACTTCGCCGAGAAGGCCTACCTGCTGCCGCGACGAGGCCTTCAAGCGCTAC
P F D A E AY L L P RD EA R Y
970 990 1010
GTCGACCAGTGGCTGCACATCGCCGAGCAGAGCGGCTTGTTGCGCCAGCGCATGGAGCAC
V D Q WLB I A Q S G L L R Q R E
1030 1050 1070
TGGCTCGAATACCGCTGGCCCACCGCGCACGGCAAGTAATACAGGGGCGGCGAGGGTGGC
N L Y R PTA HG K *
1090 1110 1130
CCGCGGCCCGCGCGGCCTTCCTTGGCGGCGGCAAAAACGTTATGGTCGGCGCCCCATCCT
1150 1170 1190
GGTGCCTGGTCCATGCGTTATCTACTGTTCGTCACCGTCCTCTGGGCGTTCTCCTTCAAC
1210 1230 1250
CTGATCGGCGAGTACCTCGCCGGCCAGGTCGCACTACTTCGCCGTGCTTACCCGGGG






Fig. 2-3. Nucleotide sequence of the P. aeruginosa phg gene
and of its flanking regions. The deduced amino acid sequence
of the gene along with its upstream flanking region is shown
beneath the corresponding codons. The Shine-Dalgarno (1974)
region is underlined and labeled (RBS). The Smal site
localized within the ORF of the phgC gene is also shown.









-30-


Characterization of the Cloned pheC Product Purified from E.
coli JP2255 and the One Purified from P. aeruginosa

Purification of the cloned pheC product expressed in L.

9oli is summarized in Table 2-3. The ratio of 3:1 obtained for

the activity of prephenate dehydratase compared to that of

arogenate dehydratase remained constant throughout the process

of purification. Only one major band was resolved by SDS-PAGE

after the Sephadex G-200 column (Fig. 2-4). The subunit

molecular weight of the cloned pheC product was 29,500 as

determined by SDS-PAGE, and the molecular weight of the native

product was 72,000 as determined by gel filtration on Sephadex

G-200. Therefore, the native enzyme is made up of 2 identical

subunits.

Purification of the cyclohexadienyl dehydratase from P.

aeruainosa was essentially carried out under the identical

conditions as those used for the cloned gene product. The

cyclohexadienyl dehydratase from P. aeruainosa and the ph

product isolated from coli failed to bind to the DEAE-

cellulose column at the pH value lower than 7.4, and they were

found to be eluted into the equivalent fractions throughout

the purification process. A specific activity ratio of

prephenate dehydratase to that of arogenate dehydratase of 3:1

was also obtained and remained unaltered during the

purification process (see Chapter 3). The native molecular

weight of this enzyme was shown to be 72,000, a value

identical to that obtained from the pheC product in E.










-31-


Table 2-3. Purification of the cloned
in EL coli JP2255(pJZ1g)


Rhre product expressed


Total Specific Activity Ratio Purification
Protein (nmole/min/mg) of Factor
(mg) PDT/ADT

PDT ADT

Crude
Extract 1118 1319 381 3.46 1

DEAE-
Cellulose 83.5 10616 3518 3.02 8.1

Hydroxyl
apatite 4.83 87750 26340 3.32 67

Sephadex 3.25 158242 48352 3.27 120
G-200

Abbreviations: PDT, prephenate dehydratase; ADT, arogenate
dehydratase.









-32-


1 2345
o 94,000
S67,000
43,000
S -- 30,000

-. 20,100

S* 14,400












Fig. 2-4. SDS-polyacrylamide gel electrophoresis of the cloned
.heQ product expressed in E. coli JP2255. The protein samples
were run on a 15% gel and stained with coomassie blue. From
left to right: Lane 1, crude extract of E. coli JP2255
carrying the pheg gene; Lane 2, the collected fractions after
DEAE-cellulose column; Lane 3, the collected fractions after
Hydroxylapatite column; Lane 4, the collected fractions after
Sephadex G-200 column; Lane 5, molecular weight standards.









-33-

coli. A subunit molecular weight of 28,000 was also obtained

for the cyclohexadienyl dehydratase from P. aeruinosa

(Chapter 3).

The K. values of 0.42 mH for prephenate, and of 0.22 mM

for L-arogenate were obtained for the cloned enzyme (Fig. 2-

5). The corresponding values obtained for the cyclohexadienyl

dehydratase directly isolated from P. aeruginosa were 0.40 mH

and 0.19 mi, respectively. Vmax values of 307.7 umole/min/mg

for prephenate and 102.8 umole/min/mg for L-arogenate were

obtained for the cloned enzyme. Since the cyclohexadienyl

dehydratase isolated from P. aeruainosa was not homogeneous,

the Vmax values were not determined. Prephenate dehydratase

activity of both preparations was competitively inhibited by

L-arogenate with a Ki value of 0.20 mM, whereas arogenate

dehydratase activity was competitively inhibited by prephenate

with a Ki value of 0.40 mM. The P. aeruginosa cyclohexadienyl

dehydratase was not subjected to allosteric regulation by L-

phenylalanine, L-tyrosine and L-tryptophan.



Discussion

The Identity of the Cloned Gene and Its Product

This study has clearly demonstrated that the cloned gene

coding for cyclohexadienyl dehydratase was obtained from P.

aeruginosa. Southern blotting analysis showed that the cloned

DNA fragment only hybridized with the Pl aeruainosa

chromosomal DNA. Analysis of the cloned DNA sequence by











-34-


3.0- A B

2.5
+ 0.4 mM AGN
2.0- 2.0-
V-1 .1. +0.4 mM PPA
1.5- 1.5-
-1
1.0 v' 1.0

0.5 t 0.5

I I I I I I I I I I


3.0 C 6.0 D

2.5- 5.0
+ 0.4 mM AGN +0.4 mM PPA
S2.0 4.0 -
1.5 3.0

1.0- 2.0

0.5 1.0

I IA -I I 1 1 I 1 I
-5 0 5 10 15 20 25 -5 0 5 10 15 20 25
rmM PREPHENATEl-1 [mM AROGENATE]-1




Fig. 2-5. Double reciprocal plots of the P. aeruginosa pheC
product expressed in E. coli (panels A and B) and the
cyclohexadienyl dehydratase purified from P. aeruginosa
(panels C and D). When assayed as prephenate (PPA) dehydratase
(panels A and C), v is expressed as nmoles of phenylpyruvate
formed per min at 370C in presence of () or absence () of 0.4
mM L-arogenate (AGN). When assayed as arogenate dehydratase
(panels B and D), v was expressed as nmoles of L-phenylalanine
formed per min at 370C in the presence () or absence () of
0.4mM PPA.









-35-

Testcode has shown that the DNA sequence from 200 bp to 1000

bp corresponds to a coding region. An open reading frame

(ORF), well situated within this region, codes for a protein

with calculated molecular weight of 30,480. This is compatible

with the value determined for the purified enzyme by SDS-PAGE.

The SmaI restriction site was found to be localized within the

ORF (Fig. 2-2). The interruption of this ORF by digestion with

SmaI caused the simultaneous loss of both prephenate

dehydratase and arogenate dehydratase activities. Subcloning

analysis has shown that as long as the clones retained the

SphI-SmaI-SmaI fragment, they complemented the phA mutants of

E. coli, JP2255 and KA197. The disruption of the integrity of

the SphI-SmaI-SmaI fragment resulted in the loss of

complementation of E. coli JP2255 and KA197.

Cyclohexadienyl dehydratase was first described in P.

aeruginosa by Petal et al. (1977). In present study, a more

extensive investigation has been carried out. The

cyclohexadienyl dehydratase isolated from P. aeruainosa and

the cloned gene product expressed in E. coli JP2255 have been

purified to electrophoretic homogeneity. The extensive

purification process was unable to separate the prephenate

dehydratase activity from the arogenate dehydratase activity.

The two activities were not regulated by either L-

phenylalanine or L-tyrosine. The physical parameters obtained

for the two purified preparations were essentially identical.









-36-
Therefore, the cloned gene product is a cyclohexadienyl

dehydratase.

Unlike the bifunctional P-protein from P aeruPinosa, the

cyclohexadienyl dehydratase is a monofunctional enzyme

although possessing both prephenate dehydratase and arogenate

dehydratase activities. The kinetic study has shown that

prephenate competitively inhibited the arogenate dehydratase

activity, and arogenate competitively inhibited prephenate

dehydratase activity, an indication that the enzyme possessed

only one single substrate binding site. Recently, the similar

results have been obtained for the cyclohexadienyl dehydratase

from Erwinia herbicola (Xia et al., 1991). The bifunctional P-

protein from P. aeruainosa and those from many other Gram-

negative bacteria possess two substrate binding sites, one for

chorismate, and one for prephenate (Ahmad and Jensen, 1986;

and Baldwin and Davidson, 1981).

The Role of the Cvclohexadienyl Dehydratase of P. aeruginosa

Auxotrophs of P. aeruginosa for L-phenylalanine and L-

tyrosine have not been reported. The reluctant auxotrophy of

E. aeruainosa for L-phenylalanine was explained as the

consequence of independent dual pathways to L-phenylalanine

(Patel et al., 1978). A mutant lacking the bifunctional P-

protein has been identified (Berry et al., 1987), and the

mutant exhibited a leaky requirement for L-phenylalanine,

showing that the exclusive biosynthesis of L-phenylalanine via

cyclohexadienyl dehydratase is rate-limiting to growth.









-37-

Another mutant of P. aeruinosa possessing a L-tyrosine

insensitive DAHP synthase was found to excrete L-

phenylalanine, presumably through the overflow pathway (Fiske

et al., 1983).

L-Arogenate is generated by transamination of prephenate.

Five aromatic aminotransferases capable of transamination of

prephenate have been isolated from P. aeruginosa (Whitaker et

al., 1982), and shown to have a relatively poor affinity for

prephenate compared to the prephenate dehydratase component of

the bifunctional P-protein. A possible candidate that may

generate prephenate molecules for transamination is the

monofunctional chorismate mutase. However, it has a poor

affinity for chorismate compared to the chorismate mutase

component of the bifunctional P-protein (unpublished results).

As discussed earlier, the cyclohexadienyl dehydratase has

relatively poor affinity for both L-arogenate and prephenate

although slightly favoring L-arogenate over prephenate.

Therefore, under normal growth conditions, most if not all L-

phenylalanine is synthesized via the bifunctional P-protein

rather than via the cyclohexadienyl dehydratase, and the L-

phenylalanine synthesized is probably all derived from

phenylpyruvate but not from L-arogenate. Thus, under

ordinary growth conditions, the cyclohexadienyl dehydratase

along with the monofunctional chorismate mutase probably does

not play a significant role in terms of the contribution to L-

phenylalanine biosynthesis. However, the overflow pathway









-38-

(the monofunctional chorismate mutase and the cyclohexadienyl

dehydratase) could be well suited as a backup system for L-

phenylalanine biosynthesis. It is interesting that P.

stutzeri, a very close relative of aeruainosa, lacks the

overflow pathway (Byng et al., 1983). The loss of the

prephenate dehydratase activity of the bifunctional P-protein

has yielded a tightly blocked L-phenylalanine auxotroph

(Carlson et al., 1984), in contrast to the bradytrophy of the

corresponding mutant of P. aeruainosa (Berry et al., 1987).

The physiological role of cyclohexadienyl dehydratase as

discussed above is not well understood and is currently under

investigation in our laboratory. To have a better

understanding of the role of the cyclohexadienyl dehydratase

in vivo, mutants lacking this enzyme activity would be

desirable and such mutants now can be obtained by using the

cloned cyclohexadienyl dehydratase gene to target the

corresponding region of chromosome in P. aeruFinosa through

gene-scrambling mutagenesis (Mohr and Deretic, 1990).


The Basis for the pheC Gene to Complement the E. coli pheA
Mutants of JP2255 and KA197

E. coli JP2255 was initially employed to select for the

clones carrying the bifunctional P-protein gene, and KA197 was

to be used for selection of the clones carrying the

cyclohexadienyl dehydratase gene. Although a low leaky

chorismate mutase activity was observed in E. coli JP2255

(Baldwin and Davidson, 1981), such outcome was not anticipated









-39-

because it was thought that such low activity was impossible

to generate enough prephenate molecules for L-phenylalanine

biosynthesis. Based on the slowness of growth, it was

initially thought that the clones could carry a truncated

bifunctional P-protein gene. This possibility has, however,

been ruled out. First, no chorismate mutase activity was found

to be associated with any clones. Second, extensive studies

have shown that the bifunctional P-proteins are competitively

inhibited by L-phenylalanine (Ahmad et al., 1989; Calhoun et

al., 1973; Cotton and Gibson, 1965; and Friedrich et al.,

1976). In case of the bifunctional P-protein of P. aeruinosa,

the prephenate dehydratase component is inhibited by L-

phenylalanine but activated by L-tyrosine (Calhoun et al.,

1973; and Patel et al., 1977). Third, the P-proteins examined

so far do not utilize L-arogenate as substrate (Ahmad et al.,

1988). Finally, when the antibody prepared against the P-

protein from Acinetobacter calcoaceticus (Ahmad et al., 1989)

which cross-reacted with the P-protein from P. aeruinosa was

added to and mixed with the crude extract of E. coli JP2255

carrying the clones, no immuno-inhibition of cyclohexadienyl

dehydratase activity was observed (data not shown). Therefore,

the complementation of E. coli JP2255 for L-phenylalanine

requirement was due to the catalytic activities of the leaky

chorismate mutase of the host and the cyclohexadienyl

dehydratase of the clones. The slow growing phenotype of E

coli JP2255 transformants indicated that in vivo L-









-40-

phenylalanine was still limiting which was directly imposed by

the low activity of the chorismate mutase.

L. coli KA197 was chosen to select for the

cyclohexadienyl dehydratase gene. This was based on an

assumption that the chorismate mutase component of the

bifunctional T-protein would divert some prephenate for L-

phenylalanine biosynthesis. Enzymological studies of the T-

protein has suggested that chorismate molecules are tightly

subjected to sequential enzymatic action (Heyde, 1979; Heyde

and Morrison, 1978; and Rood et al., 1982). When the original

clones and the resulting subclones were transformed into E

coli KA197, a relatively faster growing phenotype resulted, an

indication that the limitation of L-phenylalanine in vivo was

relieved due to the elevation of chorismate mutase activity.

Since E coli KA197, a pheA mutant lacking a bifunctional P-

protein, still possesses an intact bifunctional T-protein

(chorismate mutase/prephenate dehydrogenase), the prephenate

molecules generated in vivo were probably derived from the

catalytic activity of the chorismate mutase component of the

T-protein. This study clearly indicates that the chorismate

mutase component of the T-protein had the potential to

participate in L-phenylalanine biosynthesis. Our recent study

showed that the bifunctional P-protein was also participating

in L-tyrosine biosynthesis (Zhao et al., 1992b).









-41-


Evolutionary Implications

The deduced sequence of the P. aeruainosa heC product

was pairwise aligned with those of the Corynebacterium

glutamicum and Bacillus subtilis prephenate dehydratases

(Follettie and Sinskey, 1986; and Trach and Hoch, 1989), and

E. coli and P. stutzeri P-proteins (Fischer et al., 1992; and

Hudson and Davidson, 1984). The FP aeruainosa enzyme was found

to be marginally similar to these four proteins with an

averaged 17.5% identity. The monofunctional prephenate

dehydratases of C_. lutamicum and B. subtilis have shown

significant similarity to the two bifunctional P-proteins of

E. coli and E. stutzeri (Fischer et al., 1992; Follettie and

Sinskey, 1986; Hudson and Davidson, 1984; and Trach and Hoch,

1989 ), indicating that the prephenate dehydratase might share

a common evolutionary origin. The marginal similarity of the

P. aeruginosa cyclohexadienyl dehydratase to the prephenate

dehydratases as well as the bifunctional P-proteins might

suggest that the cyclohexadienyl dehydratase and the

prephenate dehydratases evolved independently. However, a more

detailed analysis, focusing on short, highly conserved

sequence segments rather than the total peptide, revealed a

conserved motif which includes the essential threonine residue

demonstrated by Hudson and Davidson (1984) and a number of

flanking residues. This motif, shown in Fig. 2-6, suggests

residues (within boxes) that may prove to be common to all of

dehydratases. It is interesting that all of the prephenate










-42-

dehydratases share the TRF sequence, whereas the

cyclohexadienyl dehydratase sequence is TIF. It remains to be

seen whether other cyclohexadienyl dehydratases will also

possess TIF sequences. Note that in this alignment there was

high degree of conservation of amino acid sequence between the

peptides of the P. aeruginosa cyclohexadienyl dehydratase and

the P. stutzeri P-protein. These organisms are very closely

related.











4 I MaI r-1
10 .r-- C r

00 -r 0)
S +-1 01


A' to ca
a) 4N qj 44



OA t o



0c ., a-1
J o. Q

0V~ 0


o o 0 Q)


r) V r) 44
MMg J^MOM'
4.) -I 0

34 0 ) 4 J
< 0 W 10
E1 -6 X





4J 0d 0>- 0 V


s( U P.. (14 .0

SE-4 P4 -H
, C -1 ^



ac 0 g
W00 ) V 0 0
0 ) o to En .


: se,~U
-' 0 :

r.C 4 -Q1 4
0 *I tl a y


*-l r 4 ) 0 0




-,.I :,. .
- ,- 0 0 H
WC -4 r 4J0 9
-r- j O k1 f -A
d 0 a4) *0 p
V 9 ^- M 10 (d
* Cw S d l t4 >











-44-


Z miiH > 41i".

CI 0 Ol<





He'<>HOI
H O 00I
* E< E3 c.- E








tioclog 0 >iiilli
J0 4> h Mi4
H 1S > H Y0






= 4 pw t >






In Q a >
cl< oc w 10







00000
E1 SSSO SE


CPUW (0
n u ft
















CHAPTER 3
THE CYCLOHEXADIENYL DEHYDRATASE FROM PSEUDOMONAS AERUGINOSA:
EVIDENCE FOR A PERIPLASMIC PROTEIN


Introduction

Cyclohexadienyl dehydratase catalyzes either the

conversion of prephenate to phenylpyruvate or the conversion

of L-arogenate to L-phenylalanine. The enzyme was first

described in Pseudomonas aeruainosa, an organism which

possesses dual pathways to L-phenylalanine biosynthesis (Patel

et al., 1977). The gene encoding cyclohexadienyl dehydratase,

denoted as pheC, has been cloned and sequenced recently from

P. aeruainosa (Zhao et al., 1992a). The cloning and sequencing

analyses have shown that the two catalytic activities,

prephenate dehydratase and arogenate dehydratase, were

apparently encoded by the single gene, pihe. The pheC gene

product expressed in Escherichia coli was purified to

electrophoretic homogeneity. Results of kinetic studies are

consistent with this enzyme possessing one single substrate

binding site. Thus, the dual pathways to L-phenylalanine in P.

aeruainosa were partly due to the substrate ambiguity of the

enzyme.

Cyclohexadienyl dehydratase is widely distributed among

Superfamily B microorganisms (Ahmad et al., 1988a). However,

-45-









-46-

its physiological role in nature remains to be elucidated. The

pheC gene product of P. aeruginosa was expressed to a large

quantity in E. coli. The sequence of the protein in comparison

with the deduced sequence has predicated a processing event

and a possible location in the periplasmic space of P.

aeruainosa. In this chapter, I report the results of the

biochemical and immunological studies.



Materials and Methods

Bacterial Strains. Plasmids. and Media

Bacterial strains and plasmids used in this study are

listed in Table 3-1.

LB and M9 (Maniatis et al., 1982) were used for E. coli

and P. aeruainosa strains. Ampicillin (50 mg/ml), L-tyrosine

(50 ug/ml), and thiamine (17 mg/ml) were supplemented when

appropriate.

Biochemicals

Restriction enzymes and T4 DNA ligase were obtained from

BRL. Prephenate was prepared from Salmonella typhimurium

(Dayan and Sprinson, 1970), and L-arogenate was prepared from

Neurospora crassa (Zamir et al., 1980). Sephadex G-200, DEAE-

cellulose, and hydroxylapatite were obtained from Sigma

Chemical Company, Whatman, and Bio-Rad, respectively.

Molecular weight standards for SDS-polyacrylamide gel

electrophoresis (a-lactalbumin, 14,400; soybean trypsin

inhibitor, 20,100; carbonic anhydrase, 30,000; ovalbumin,










-47-


Table 3-1. Bacterial strains and plasmids

Strain or plasmid Genotype or description Source


EL .cli K-12
JM83


JP2255


E. aeruainosa
PAO1

PAT1051


Plasmids
pUC19


pJZlg


aEr (lac-proAB) r.sL 80
lacZ M15

aroF363 pheA361 DRhe352
tyrA382 thi strR712
lacYl xy"5

Prototroph

RheA


Apr lacZ


1110-bp SphI-SmaI fragment
containing the intact phe
gene cloned into pUC19


BRL



1*


BRL



4*


pJZpp 1477-bp SphI-PstI fragment
containing the intact pheA
gene cloned into pUC19


This study


(1955); 3,


': 1, Baldwin and Davidson (1981); 2, Holloway
Berry et al. (1987); 4, Zhao et al. (1992a).









-48-

43,000; bovine serum albumin, 67,000; and phosphorylase,

94,000) and for gel filtration (carbonic anhydrase, 30,000;

bovine serum albumin, 67,000; alcohol dehydrogenase, 150,000;

and b-amylase, 200,000) were obtained from Pharmacia Fine

Chemicals and Sigma Chemical Company, respectively. Other

biochemicals were obtained from Sigma Chemical Company.

Expression Construct of The P. stutzeri pheA

The intact pheA gene coding for bifunctional P-protein

from P. stutzeri (Fischer et al., 1991) was cloned from

pJF1954 into pUC19 at SphI/PstI sites, and the resulting

construct, designated as pJZpp was found to over-express the

P-protein in E. coli.


Purification of The Cloned pheC Product from E. coli and
The Cyclohexadienyl Dehydratase from P. aeruginosa

The cloned pheC product expressed in E. coli was purified

as previously described (Zhao et al., 1992a). The

cyclohexadienyl dehydratase from P. aeruginosa was purified as

described before (Zhao et al., 1992a) with modifications.

Ammonium sulfate precipitation was omitted. After the Sephadex

G-200 column, the fractions showing cyclohexadienyl

dehydratase activity were pooled and applied to a small

hydroxylapatite column (1.5 cm x 30 cm) and then to a small

DEAE-cellulose column (1.5 cm x 30 cm) as described before

(Zhao et al., 1992a).









-49-

N-terminal Amino Acid Sequencina

The N-terminal sequences of the purified proteins were

determined by using an Applied Biosystems 470A Protein

Sequencer with On-Line 120A PTH-Analyzer at the Protein Core

Facility of the University of Florida.

Molecular Weight Determinations

The native molecular weight of the purified proteins was

estimated by gel filtration as described above for enzyme

purification, and the subunit molecular weight of the proteins

was determined by SDS-polyacrylamide gel electrophoresis (SDS-

PAGE) (Laemmli, 1970).

Antibody Preparation and Immunoloaical Analysis

The P. aeruginosa pheC gene product purified from E. coli

(0.6 mg) was injected into a New Zealand white rabbit (Hiebert

et al., 1984), and the antiserum was collected one week after

the third injection. The Western blot was performed as

described by Hiebter et al. (1984).

Immunological localization of the the pheC gene product

in E coli was carried out as described below. E. coli JP2255

carrying pJZlg (DheC gene) was grown in LB supplemented with

50 ug/ml of ampicillin at 37*C, harvested during the late

exponential phase of growth, fixed in cacodylate-buffered (pH

7.2) 2.5% formaldehyde/0.5% glutaraldehyde for 10 min on ice,

dehydrated, and embedded in Lowicryl K4M (Carlemalm and

Villiger, 1989). The resin was polymerized with UV at -20C

for 48 hr. Sections were allowed to react with the antibody









-50-

prepared against the purified phge gene product, and then

lebeled with gold particles (Carlemalm and Villiger, 1989).

Osmotic Shock and Chloroform Treatment

E. coli JP2255 carrying pJZlg and pJZpp was grown in

either M9 medium (1.2 liters) supplemented with ampicillin (50

ug/ml), L-tyrosine (50 ug/ml), and thiamine (17 ug/ml), or in

LB medium (1.2 liters) supplemented with ampicillin (50

ug/ml). The cells were harvested by centrifugation during the

late exponential phase of growth. Each culture was divided

equally into three parts (400 ml each). One part of the

culture (400 ml, 1.5 g) was subjected to osmotic shock (Neu

and Heppel, 1965). After each centrifugation, supernatant

fractions were collected and referred to as periplasmic

fractions. For preparation of the cytoplasmic fractions, the

cell pellet recovered after osmotic shock was sonicated. This

sonicated preparation was centrifuged at 150,000 g for 60 min,

and the resulting supernatant fraction was collected and then

passed through a PD-10 Sephadex column. This preparation is

referred to as cytoplasmic fraction. The second part of the

culture was subjected to chloroform treatment as described by

Ames et al. (1984) except that the culture mixed with

chloroform was incubated for 45 min. The supernatant fraction

collected after incubation is referred to as periplasmic

fraction. The third part of the culture was treated as

described above for the preparation of cytoplasmic fraction,

and this preparation is referred to as crude extract.









-51-

P- aeruainosa PAOI (Holloway, 1955), and PAT1051 (Berry

et al., 1987) were grown in M9 medium (1.2 liters) and

harvested during late exponential phase of growth. The

cultures were also divided equally into three parts (400 ml

each, 1.5 g). One part of the culture was treated with

chloroform (Ames et al., 1984) as described above, and one

part was suspended in Tris buffer (20 mM, pH 8.0) and shaken

at 40C for 45 min (Beacham, 1979). The pellets recovered after

Tris treatment and the third part of the culture were treated

as described above for the preparation of cytoplasmic

fraction, and the resulting preparations were referred as to

cytoplasmic fraction and crude extract, respectively.

Enzyme Assay and Protein Determination

Prephenate dehydratase was assayed as described by Cotton

and Gibson (1965). Arogenate dehydratase was assayed as

described by Zamir et al. (1985). Alkaline phosphatase was

assayed by the procedure of Brickman and Beckwith (1975).

Specific activity is defined as nmoles of phenylpyruvate, L-

phenylalanine, or nitrophenyl formed per mg protein per min.

Protein was measured according to Bradford (1976).



Results
Purification of The Cloned DheC Product from E. coli and
The Cyclohexadienvl Dehydratase from P. aeruainosa and
The N-terminal Amino Acid Sequencina

Purification of the cloned pheC was carried out under the

identical conditions as previously described (Zhao et al.,









-52-

1992a). After the Sephadex G-200 column chromatography, only

one single protein band was resolved by SDS-PAGE. The subunit

molecular weight of the p~he product was determined to be

28,000 (Fig. 3-1) in comparison with a value of 29,500

obtained earlier (Zhao et al., 1992a). The native molecular

weight of the purified pheC product was estimated to be

72,000, identical to the value determined previously (Zhao et

al., 1992a).

The N-terminal sequence of the purified pheC gene product

was determined to be Gln-Glu-Ser-Arg-Leu-Asp-Arg-Ile-Leu-Glu-

Ser. The sequence determined herein is found to match with the

deduced amino acid sequence starting from the residues 26 to

36 (Zhao et al., 1992a). This indicates that an amino-terminal

peptide (25 residues in length), a possible signal sequence

was cleaved in E. coli (Fig. 3-2). In order to establish if

this cleavage also occurred in P. aeruginosa, the

cyclohexadienyl dehydratase from this organism was also

purified.

Purification of the cyclohexadienyl dehydratase from P.

aeruginosa is summarized in Table 3-2. The activities of

prephenate dehydratase and arogenate dehydratase were found to

coelute into the same fractions, and the ratio of the two

activities was constant throughout purification. After the

five steps of purification, the enzyme was homogenous as

judged by SDS-PAGE and found to have the same molecular weight

(28,000) (Fig. 3-1) as the cloned pheC gene product. The








-53-


1234

94000
0 67000
43000


.- .*' 30000


20100

14400





Fig. 3-1. SDS-PAGE analysis of the cloned DheC gene product
purified from E. coli and the cyclohexadienyl dehydratase
purified from P. aeruainosa. The samples were run on a 15%
polyacrylamide gel. From left to right: Lane 1 and 3, the
purified cyclohexadienyl dehydratase from P. aeruiosa; Lane
2, the purified pheC gene product from E. coli; and Lane 4,
the molecular weight standards.









-54-


N'-Gln-Glu-Ser-Arg-Leu-Asp-Arg-Ile-Leu-Glu-Ser ... C'


N'-Gln-Glu-Ser-Arg-Leu-Asp-Arg-Ile-Leu-Glu-Ser ....C'


5'-ATG CCG AAG TCA TTC CGC CAT CTC GTC CAG GCC CTG GCC
Met Pro Lys Ser Phe Arg His Leu Val Gin Ala Leu Ala

TGC CTT GCG CTG CTG GCC AGC GCC AGC CTC CAG GCG CAG
Cys Leu Ala Leu Leu Ala Ser Ala Ser Leu Gin Ala Gin

GAG AGC CGC CTC GAC CGC ATC CTC GAA AGC GGC GTG. .-3
Glu Ser Arg Leu Asp Ara Ile Leu Glu Ser Gly Val..


Fig. 3-2. A and B, The N-terminal amino acid sequences of the
pheC product purified from E. coli and the cyclohexadienyl
dehydratase purified from .. aeruinosa, respectively; and C,
the deduced N-terminal amino acid sequence for the cloned pheC
gene product. The underlined sequence starting from the
positions of 26 to 36 was identical to those of A and B.









-55-

native molecular weight of the purified enzyme was determined

to be 72,000 by gel filtration, a value identical to the one

previously determined for both the cloned phe- gene product

and the cyclohexadienyl dehydratase directly isolated from E.

aeruainosa (Zhao et al., 1992a).

The purified cyclohexadienyl dehydratase was also

subjected to N-terminal sequencing. The 11 residues

sequenced were found to be identical to those determined for

the cloned pheC gene product purified from E. coli and to

match with the deduced sequence starting from residues 26 to

36, thus showing the cleavage of the amino-terminal peptide

also occurred in P. aeruinosa.

The amino peptide sequence of the cyclohexadienyl

dehydratase from P. aeruginosa is shown in Fig. 3-2 and found

to share several features with the known signal sequences for

periplasmic proteins (Oliver, 1986). For example, the P.

aeruainosa amino peptide contained a basic amino terminus

(Lys, Arg, and His), a hydrophobic core region (Ala, Leu, Val,

and Ser), and a putative processing site (Leu-Gln-Ala). The

length of this amino peptide also falls within the range for

the periplasmic signal sequences (Oliver, 1986).

Osmotic Shock and Chloroform Treatment

The results of osmotic shock and chloroform treatment of

E. coli carrying pJZlg and pJZpp are summarized in Table 3-3.

Over 90% of cyclohexadienyl dehydratase and alkaline

phosphatase activities was recovered in the periplasmic









-56-

Table 3-2. Purification of the cyclohexadienyl dehydratase of
E. aeruginosa

Total Specific activity Ratio Purifi-
protein (nmol/mg/min) of cation
(mg) PDT/ADT ratio
PDT ADT

Crude extract 2205 15+ 3.7 4.1

DEAE-cellulose 406 48.5 12.5 3.9 3.2

Hydroxylapatite 34.4 403 107 3.8 26

Sephadex G-200 6.0 1996 537 3.7 130

Hydroxylapatite 0.95 10301 2918 3.5 672

DEAE-cellulose 0.32 19882 5598 3.5 1298

+ prephenate dehydratase was assayed in presence of 4.0 mM
phenylalanine. Abbreviations: PDT, prephenate dehydratase;
ADT, arogenate dehydratase.









-57-

fractions when E. coli JP2255(pJZlg) was subjected to osmotic

shock.

The pheA construct of P. stutzeri was chosen for this

study because this gene was over-expressed in E coli and the

gene product is a cytoplasmic protein. Thus, it was served as

a negative control in this study.

Less than 1% of the bifunctional P-protein activity, but

over 90% of the alkaline phosphatase activity was found in the

periplasmic fractions when coli JP2255(pJZpp) was subjected

to osmotic shock. However, over 95% of the P-protein activity

was detected in the cytoplasmic fraction.

When E. coli carrying pJZlg and pJZpp were treated with

chloroform, similar results were obtained (Table 3-3) except

that the chloroform treatment was only 30-50% as efficient as

osmotic shock based on the amount of activity released.

Similar results were also obtained when EL. coli JP2255

carrying pJZlg or pJZpp was grown in minimal medium (Table 3-

4). Since the alkaline phosphatase activity was repressed to

a very low level by phosphate present in the medium, its

activity was not determined.

The results of chloroform and Tris treatments of L.

aeruainosa strains are summarized in Table 3-5. Approximately,

50% of cyclohexadienyl dehydratase activity was recovered in

periplasmic fraction of the strain PAT1051, but less than 10%

of the activity was recovered in the periplasmic fraction of

the strain PAO1 when the cells of these strains were subjected










-58-


Table 3-3. Release of the cyclohexadienyl dehydratase
expressed in E. coli JP2255 by osmotic shock and chloroform
treatment when the cells were grown in rich medium


Total AP PDT
protein (OD/min) --(umol/min)--
(mg) + Phe Phe


ADT


pJZlg (CDT)


Periplasmic fraction
(Osmotic shock)

Periplasmic fraction
(Chloroform treatment)

Cytoplasmic fraction
(Osmotic shock)

Crude extract
(Intact cells)


63 90


1168


30 46 480


250


316


93 1231


pJZpp (P-protein)


Periplasmic fraction
(Osmotic shock)

Periplasmic fraction
(chloroform treatment)

Cytoplasmic fraction
(Osmotic shock)

Crude extract
(Intact cells)


35 63


12 26


196


231


1174


486


26


1257


318


318


7


354


0.0


0.0


1.9


2


1.1


0.09


162


172


0.0


0.0


0.0


0.0


Abbreviations: AP, alkaline phosphatase; PDT, prephenate
dehydratase; ADT, arogenate dehydratase; CDT, cyclohexadienyl
dehydratase; Phe, phenylalanine.










-59-


Table 3-4. Release of the cyclohexadienyl dehydratase
expressed in E. coli JP2255 by osmotic shock and chloroform
treatment when the cells were grown in minimal medium

Total PDT ADT
protein ------(umol/min)-------
(mg) + Phe Phe
pJZlg (CDT)

Periplasmic fraction 33 444 444 120
(Osmotic shock)

Periplasmic fraction 10 219 219 58
(Chloroform treatment)

Cytoplasmic fraction 79 4 4 1
(Osmotic shock)

Crude extract 116 462 462 122
(Intact cells)

pJZpp (P-protein)

Periplasmic fraction 37 0.0 0.4 0.0
(Osmotic shock)

Periplasmic fraction 11 0.03 19 0.0
(Chloroform treatment)

Cytoplasmic fraction 70 0.11 40 0.0
(Osmotic shock)

Crude extract 108 0.14 42 0.0
(Intact cells)

Abbreviations: PDT, prephenate dehydratase; ADT, arogenate
dehydratase; CDT, cyclohexadienyl dehydratase; Phe,
phenylalanine










-60-


Table 3-5. Release of cyclohexadienyl dehydratase of P.
aeruainosa PAO1 and PAT1051 by chloroform and Tris treatments
when cells were grown in minimal medium

Total PDT ADT
protein -----(nmole/min)------
(mg) + Phe Phe


PAO1

Periplasmic fraction 7.5 198.7 218.2 56.7
(Tris treatment)

Periplasmic fraction 11.4 2349.4 2475.3 647.1
(Chloroform treatment)

Cytoplasmic fraction 155.8 2153.3 5754.3 581.9
(Tris treatment)

Crude extract 166.7 2435.9 6489.0 667.4
(Intact cells)

PAT1051

Periplasmic fraction 4.06 317.2 306.2 88.1
(Tris treatment)

Periplasmic fraction 17.8 150.3 148.7 40.1
(Chloroform treatment)

Cytoplasmic fraction 170.4 267.9 255.6 78.5
(Tris treatment)

Crude extract 181.3 608.3 580.9 164.3
(Intact cells)


Abbreviations: PDT, prephenate
dehydratase.


dehydratase; ADT, arogenate









-61-
to Tris treatment. The fact that the cyclohexadienyl

dehydratase rather than the bifunctional P-protein was present

in the periplasmic fractions was established by lack of

feedback inhibition by L-phenylalanine and capability to

utilize L-arogenate, which are all the characteristics of the

cyclohexadienyl dehydratase but not of the bifunctional P-

protein (Patel et al., 1977, Zhao et al., 1992a; and Berry et

al., 1987). Furthermore, E. aeruainosa PAT1051 does not

possess the P-protein activities (Berry et al., 1987).

The chloroform treatment yielded a better result for the

strain PAO1 since over 85% of the activity was released from

periplasmic space. On the other hand, chloroform treatment

released less than 40% of the activity from the periplasmic

space of the strain PAT1051. Since cells of P, aeruginosa are

known to undergo massive lysis in the presence of EDTA,

osmotic shock could not be used for this study.


Immunological Localization of The Cyclohexadienyl Dehydratase
in E. coli and P. aeruginosa

All the periplasmic fractions collected were first

concentrated using a PM-10 membrane and then subjected to SDS-

PAGE along with the crude extracts. The proteins were then

transferred to a nitrocellulose membrane and probed with a

preimmune serum and the antibody prepared against the purified

pheC product. The presence of the cyclohexadienyl dehydratase

was detected in the periplasmic fractions of JP2255(pJZlg),

PAO1, and PAT1051 (Fig. 3-3). The antibody prepared against













0 a C
,- o> .1 .







0 04 4
(A r. 4 44 + 4 >4 r.




Me ## OC UM '-'O1-'
+ r..q 1 {/)4-1>> G


to (0 t Qto m rn




( ,) 0-l I O 4,-J + 4J ) 0 -
) 3; 44 W










.,. 01 t .qA0 .

o. .o .no I0








Sw t -) (0 ,,'4 r-
r( 3 Oat C t
V W ON r. :4) k T -

0 r + r. +,UV)rll
51, 4 .2 2 ) .- '




4 4) g-:3 0 0








C0D 0 ( C4 >












.. 0r).-.0 ,0 0,
S4 )ll H WV) 0





















*4 Wr4040r4 Nr
4' C 0 P > 0 k 04





i -H0 0 to0 0-l40

r0 O, 44 QH 0 to P4 .I
r4 k o -+j tz t +4
44 04J r d- J k o v
0o V 'o ..


.rd~d PI,) 0.H No 4 -1 e
I0 :3 v 0 : ,+ (l) r
k -0 t-3 1 'r.
c'0 0-4 00 U 44
[,, +a V 4 a PW aM











-63-


CO

r
0)


'I
0


LuII










-64-


hr

~~ 34 .J.


-. 08


I~


.. 4


Fig. 3-4. Immunological localization of the pheC gene product
formed in E. coli. Cells of E. coli JP2255 (pJZlg) were fixed
by 2.5% formaldehyde/0.5%glutaraldehyde, and embedded in
Lowicryl K4M. Thin sections of these cells were incubated with
the antibody prepared against the purified pheC gene product
(upper panel) or preimmune serum (lower panel), labeled with
gold-particles, and examined with a JEOL 100-CX electron
microscope (52,000 x).


it .. ;S









-65-

the pheC product did not react with the bifunctional P-protein

apparently because of lack of a significant identity.

The localization of the cyclohexadienyl dehydratase in

the periplasmic space of E coli was clearly established by

electron microscopy (Fig. 3-4). Interestingly, the

cyclohexadienyl dehydratase expressed in E. coli was only

found in the periplasmic space of the two polar regions (Fig.

3-4).




Discussion

Cyclohexadienyl Dehydratase from P. aeruainosa: A
PeriDlasmic Enzyme

Several lines of evidence in this study have led to a

conclusion that the cyclohexadienyl dehydratase from P.

aeruainosa is a periplasmic protein. The sequencing of the

pheC product expressed in E. coli indicated that an amino

peptide, 25 residues in length was cleaved. This could be due

to some proteases which acted on the expressed phgC product in

E. gcli. However, this possibility was ruled out since the

enzyme purified directly from P. aeruginosa exhibited the

identical molecular weight and sequence, and was shown to

possess no methionine residue at its N-terminus. A substantial

amount of the cyclohexadienyl dehydratase activity was

released from the periplasmic spaces when the cells were

subjected to either osmotic shock or chloroform treatment, an

indication of a periplasmic protein. This contention was









-66-

confirmed by Western blotting analysis of the periplasmic

fractions of P. aeruginosa and E coli and by immunological

localization study.

The cloned pheC product was recovered in the periplasmic

fractions when expressed in E. coli, indicating that the E.

coli machinery was able to recognize the signal of the

protein, to process and translocate it to its correct destiny

within the cell. Similar results have often been obtained not

only for microorganisms and but also for eukaryotes (Benson et

al., 1985; and Talmadge et al., 1980).



Cyclohexadienyl Dehydratase from P. aeruQinosa: Its
In Vivo Function

In P. aeruinosa, L-phenylalanine can be synthesized

either through the bifunctional P-protein or through the

monofunctional chorismate mutase and the cyclohexadienyl

dehydratase. The latter two enzymes are referred to as

overflow pathway (Fiske et al., 1983). The presence of the

overflow pathway in microorganisms may confer some

physiological advantages. A mutant of P. aeruinosa lacking

the bifunctional P-protein exhibited a leaky requirement for

L-phenylalanine, showing the exclusive biosynthesis of

phenylalanine through the overflow pathway is rate-limiting to

growth (Berry et al., 1987). On the other hand, a mutant of P.

stutzeri, a close relative of aeruginosa, defective in the

bifunctional P-protein activity, yielded a tight phenylalanine

auxotroph (Carlson et al., 1984). Therefore, the overflow









-67-

pathway may be used as a backup system for phenylalanine

biosynthesis (Zhao et al., 1992a).

This study has shown that the cyclohexadienyl dehydratase

of P. aeruginosa is a periplasmic enzyme rather than a

cytoplasmic one as previously assumed. Thus, its function in

vivo may be also different. Most of bacterial proteins

localized in the periplasm are either binding proteins or

scavenging enzymes (Oliver, 1986). It is possible that the

cyclohexadienyl dehydratase of P. aeruainosa is also a

phenylalaninee) scavenging enzyme. Many bacteria such as P.

aeruginosa, P. sringa, Xanthomonas campestris, and Erwinia

herbicola that possess cyclohexadienyl dehydratases are plant

pathogens. It is known that higher plants synthesize L-

phenylalanine through an intermediate, L-arogenate (Jensen,

1986). Cyclohexadienyl dehydratase in microorganisms may be

involved in scavenging L-arogenate of plants. Upon binding

with L-arogenate, the enzyme converts L-arogenate to L-

phenylalanine that is then transported into bacterial cells.

Two lines of evidence to some extent support this hypothesis.

The two cyclohexadienyl dehydratases, one from P. aeruainosa

(Zhao et al., 1992a), and one from E. herbicola (Xia et al.,

1991) were characterized and found to have a higher affinity

for L-arogenate than for prephenate. A search of Genbank has

revealed that the cyclohexadienyl dehydratase from P.

aeruginosa shared some sequence similarity with the products

of hisJ, argT and gInQ which are known to code for the amino









-68-

acid binding proteins and permease (Higgins et al., 1982,

Nonet et al., 1987). In order to study the function of the

cyclohexadienyl dehydratase, the pheL mutants of P.aeruainosa

will be needed. Experiments designed to selectively inactivate

the P. aeruinosa pheC gene in vivo (Mohr and Deretic, 1991)

are currently in progress.
















CHAPTER 4
CYCLOHEXADIENYL DEHYDROGENASE FROM ZYMOMONAS
MOBILIS:MOLECULAR CLONING OF THE GENE
AND CHARACTERIZATION OF THE GENE PRODUCT


Introduction

Cyclohexadienyl dehydrogenase catalyzes either the

conversion of prephenate to 4-hydroxyphenylpyruvate or the

conversion of L-arogenate to L-tyrosine (Patel et al., 1977;

and Xia and Jensen, 1990) (Fig. 4-1). Thus, this enzyme may

function as either prephenate dehydrogenase or as arogenate

dehydrogenase (Fig. 4-1). Cyclohexadienyl dehydrogenase was

first described in Pseudomonas aeruainosa, an organism that

possesses dual pathways to L-tyrosine (Patel et al., 1977).

The dual pathways to L-tyrosine biosynthesis appear to result

from substrate ambiguity of the enzyme (Patel et al., 1977;

1978). The prephenate dehydrogenase and arogenate

dehydrogenase activities of the purified cyclohexadienyl

dehydrogenase from P. aeruginosa were inseparable, and the

ratio of the two activities remained constant throughout

purification (Xia and Jensen, 1990). A single protein band

was resolved by sodium dodecyl sulfate (SDS)-polyacrylamide

electrophoresis (Xia and Jensen, 1990). Evidence was

obtained in support of the conclusion that both

cyclohexadienyl substrates bind at a common catalytic site,

-69-












** 4
00
C4-




to

0 .0


-H 0
.0 H




-H-E-'- 0
0 0
H 0 0*
4) 1



0- 0




V -H
0






4 -
o 9;



4Q) 9i














U4*14
,CO0








0 *4
0 y O)












I AS
0 wI *



r4 0 93
-00




1 s H >i
r u to u










-71-


z
0
0

-C
'IC


w

= 0
.a
;!r


0
L0





=
0.


z


0O
ir'


0



0
0 8
1^ ..-o


z
o /
O /
C.A









-72-

and therefore, the enzyme is a monofunctional protein having

multiple activities.

The dehydrogenase component of bifunctional T-proteins

(named as chorismate mutase/prephenate dehydrogenase) present

in enteric bacteria (Ahmad and Jensen, 1986) has recently been

shown to utilize L-arogenate in addition to prephenate as

substrate (Ahmad and Jensen, 1987). Therefore, the named

prephenate dehydrogenase component of all T-proteins is in

fact a cyclohexadienyl dehydrogenase (Ahmad and Jensen, 1987).

Cyclohexadienyl dehydrogenase is ubiquitous throughout

Superfamilies A, B, and C of gram-negative bacteria (Ahmad and

Jensen, 1986; 1988b; and Jensen, 1985). An evolutionary

relationship between cyclohexadienyl dehydrogenase and the T-

protein has been proposed (Ahmad and Jensen, 1986; 1988; and

Jensen, 1985).

Z. mobilis is a gram-negative member of rRNA Superfamily

C. Genes and enzymes of the glycolytic pathway in this

organism have been studied extensively (Montenecourt, 1985).

However, the pathway arrangement for tyrosine and

phenylalanine biosynthesis is essentially unknown in this

organism and indeed within the entire Superfamily to which it

belongs. In this chapter, I describe the molecular cloning

and sequencing of tyrC, the gene coding for cyclohexadienyl

dehydrogenase from Z. mobilis CP4, as well as the purification

and characterization of the gene product.









-73-


Materials and Methods

Bacterial Strains, Plasmids and Media

Bacterial strains and plasmids used in this study are

described in Table 4-1. LB medium was used as an enriched

medium, and the M9 recipe (Maniatis et al., 1982) was used to

prepare minimal medium for E. coli strains. Ampicillin was

supplemented to media where indicated at 50 Mg/ml, and

thiamine was added at 17 jg/ml. A complex medium was used for

Z. mobilis (Osman et al., 1987). Agar was added to media at 15

g/liter for solid medium.


Isolation of Z. mobilis Chromosomal DNA and Construction of
A Gene Library

Chromosomal DNA was isolated from Z. mobilis CP4 as

described previously (Clark-Curtiss et al., 1985). The DNA was

partially digested with Sau3A, and fragments of 5 to 7

kilobases (kb) were isolated by using a sucrose gradient

(Clark-Curtiss et al., 1985). Libraries of these fragments

were constructed by ligation into the dephosphorylated BamHI

site of pUC18 (Yanisch-Perron et al., 1985). The ligation

mixtures were transformed into E. coli DH5a. Transformants

were collected from LB plates, and recombinant plasmids were

purified. The purified recombinant plasmids were then used to

transform E coli AT2471 to independence of the tyrosine

requirement.










-74-


Table 4-1. Bacterial strains and plasmids

Strain/plasmid
Genotype or description Source


E. coli K-12
JM83


DH5a



AT2471

JP2255


araA (proAB-lac)rpsL p80
lacZAM15

480dacAM15A(lacZ15acZYA-argF)
U169recAl hsdR17 (r-^,mk)
SupE44A-thi-l gyrA relAl

thil tyrA

aroF363 pheA361 phe0352
tyrA382 thi strR712
lacYl xyl5


Gibco-BRL


BRL

CGSC

CGSC


Z. mobilis
(CP4)

Plasmids
pUC18
pUC19
pGEM-5Zf(+)


Prototroph


lacZ Apr
lacZ Apr
lacZ Apr


Original clone of cyclohex-
xadienyl dehydrogenase gene
isolated from CP4 library

A derivative of pJZ5
generated by removal of a
1-kb EcoRI-EcoRI fragment

A derivative of pJZ5
generated by removal of a
3.2-kb HindIII fragment

A derivative of pJZ5
generated by removal of
a 4.2-kb SphI fragment

A derivative of pJZ5
generated by removal of
a 5.3-kb EcoRV-PstI
fragment

A derivative of pJZ5
generated by removal of a
6.7-kb SstII-PstI fragment


This study



This study



This study



This study




This study


This study


Ingram


BRL
Promega


pJZ5



pJZ5a


pJZ5b



pJZ5c


pJZ5d




pJZ5e










-75-


Table 4-1 (continued)


pJZ5f


2.5-kb NcoI fragment of
pJZ5 subcloned into
pGEM-5Zf(+) at NcoI site

1.1-kb NsiI-(SstII)
HindIII fragment of
pJZ5f subcloned into
pUC18 at PstI-HindIII sites

1.1-kb NsiI-(SstII)
HindIII fragment of pJZ5f
subcloned into pUC19 at
PstI-HindIII sites

0.6-kb StuI fragment of
pJZ5f subcloned into
pUC18 at SmaI site

0.8-kb StuI-PstI fragment
of pJZ5f subcloned into
pUC18 at SmaI-PstI sites

1.4-kb StuI-StuI-(SstII)
PstI fragment of pJZ5f
subcloned into pUC18 at
SmaI-PstI sites


This study


This study




This study




This study



This study



This study


pJZ5g-1




pJZ5g-2




pJZ5h



pJZ5i


pJZ5j









-76-


DNA Manipulations

All restriction endonuclease enzymes, T4 DNA ligase, and

calf intestine phosphatase were obtained from Gibco-BRL or

Promega and were used according to manufacturer instructions.

Subcloning was conducted by standard methods (Maniatis et al.,

1982). Southern blot hybridization, using a biotinylated

probe, was carried out under stringent conditions according to

the instructions of Promega.

DNA Sequencing and Data Analysis

Subclones pJZ5g, pJZ5h, and pJZ5i were purified by use of

a CsCl gradient (Humphreys et al., 1975), and sequenced in

both directions (Prober et al., 1987) at the DNA Core Facility

of the University of Florida. The nucleotide sequence and the

deduced amino acid sequence were analyzed by using the

University of Wisconsin Genetics Computer Group (GCG) package

(Devereux et al., 1984).

Crude Extract Preparation and Enzyme Assay

Cell cultures were grown at 37C in 450 ml of LB broth

containing ampicillin at 50 Mg/ml. The cells were harvested by

centrifugation during the late exponential phase of growth,

suspended in 3 ml of 50 mM potassium phosphate buffer (pH

7.5), and sonicated for 30 sec using a Lab-Line Ultratip

Labsonic System (Lab-Line Instruments, Inc., Melrose Park,

IL). The resulting suspension was centrifuged at 150,000 g for

60 min at 4C. The supernatant fraction was collected and

passed through a PD-10 Sephadex column to remove small









-77-

molecules. This preparation was used to assay for enzyme

activity.

Arogenate dehydrogenase and prephenate dehydrogenase

activities were assayed by following the appearance of NADH on

a spectrophotofluorometer (excitation at 340 nm and emission

at 460 nm) (Byng et al., 1982). For kinetic studies, arogenate

dehydrogenase was assayed by following L-tyrosine formation

using HPLC (Zamir et al., 1985). One unit of enzyme activity

was defined as formation of 1 nanomole of L-tyrosine or NADH

per min at 37C. Protein concentration was determined by the

method of Bradford (1976).


Purification of The Cloned Cyclohexadienyl Dehydrogenase
from E. coli AT2471

E. coli AT2471 carrying the subclone pJZ5f was grown at

37C in 3 liters of minimal medium supplemented with 50 Ag/ml

of ampicillin and harvested during the late exponential phase

of growth. The cells were washed with 50 ml of 20 mH potassium

phosphate containing 1 mM DTT (pH 7.4), resuspended in the

same buffer, and disrupted by sonication. After centrifugation

(150,000 g for 60 min at 40C), the supernatant was applied to

a DEAE-cellulose column (2.5 x 30 cm) equilibrated with the

same buffer. The column was first washed with 100 ml of the

buffer and then eluted with 800 ml of buffer containing a

linear gradient from 0-to-400 mH KC1. Fractions of 3 ml were

collected, and those showing high cyclohexadienyl

dehydrogenase activity were pooled and concentrated by use of









-78-

an Amicon PM-30 membrane. The concentrated preparation was

washed twice with the phosphate buffer, and then applied to a

hydroxylapatite column (2.5 x 30 cm) which was equilibrated

with the phosphate buffer. The column was first washed with

100 ml of the buffer and then eluted with a 700-ml linear

gradient of phosphate from 20-to-400 mM. Fractions of 3.0 ml

were collected, and those showing high cyclohexadienyl

dehydrogenase activity were pooled. The pooled fractions were

concentrated as described before and then applied to a

Sephadex G-200 column (2.5 x 98 cm) previously equilibrated

with the phosphate buffer. The column was eluted with the

buffer, and fractions exhibiting cyclohexadienyl dehydrogenase

activity were combined. The preparation was then loaded on a

hydroxylapatite column (1 x 30 cm), and the column was eluted

as described before. Fractions containing cyclohexadienyl

dehydrogenase were collected and used for further study.

Amino Acid Sequencing of The Cloned Cyclohexadienyl
Dehydrogenase

The purified enzyme preparation was denatured by SDS and

then subjected to polyacrylamide gel electrophoresis (Laemmli,

1970). The protein was then transferred to a polyvinylidene

difluoride membrane and sequenced using an Applied Biosystems

470A Protein Sequencer with On-line 120A PTH-Analyzer at the

Protein Core Facility of the University of Florida.

Molecular Weight Determinations

The molecular weight of the native enzyme was estimated

by gel filtration (Sephadex G-200) as described above for









-79-

enzyme purification. The subunit molecular weight of the

cloned cyclohexadienyl dehydrogenase was determined by SDS-

PAGE (Laemmli, 1970). In order to confirm further the identity

of the protein band resolved by SDS-PAGE, the purified enzyme

preparation was first subjected to electrophoresis on a native

gel which was performed as described for SDS-PAGE (Laemmli,

1970) except that SDS was omitted. After electrophoresis, a

part of the gel was stained in 1 mM tetrazolium blue, 1 mM

prephenate, and 1 mM NAD*, a part of the gel was stained in 1

mM tetrazolium blue, and 1 mM NAD*, and a part of the gel was

stained in Coomassie blue.

Biochemicals and Chemicals

Prephenate was prepared from Salmonella typhimurium

(Dayan and Sprinson, 1970), and L-arogenate was prepared from

Neurospora crassa (Zamir et al., 1980). DEAE-cellulose was

obtained from Whatman; Sephadex G-200, ampicillin, thiamine

and amino acids were obtained from Sigma Chemical Company; and

hydroxylapatite was obtained from Bio-Rad. Molecular weight

standards for SDS-PAGE (a-lactabumin, 14,400; soybean trypsin

inhibitor, 20,100; carbonic anhydrase, 30,000; ovalbumin,

43,000; bovine serum albumin, 67,000; and phosphorylase,

94,000), and for gel filtration (carbonic anhydrase, 29,000;

bovine serum albumin, 67,000; alcohol dehydrogenase, 150,000;

and B-amylase, 200,000) were obtained from Pharmacia Fine

Chemicals and Sigma Chemical Company, respectively. LB medium

and agar were purchased from Difco.









-80-


Cloning of The Gene Encoding Z. mobilis Cyclohexadienyl
Dehvdrogenase

E. coli AT2471 was employed to select clones carrying the

cyclohexadienyl dehydrogenase gene. E. coli AT2471 was first

transformed with the recombinant plasmids purified from gene

libraries prepared from Z. mobilis, and the ampicillin-

resistant transformants were then allowed to grow in LB medium

for 3 hr, 6 hr, and 16 hr. The transformants were washed twice

with growth-volume amounts of liquid minimal medium, and

plated out on agar plates of minimal medium supplemented with

ampicillin. No colonies were observed 2 days after incubation

at 379C when the transformants were first grown in LB medium

for 3 hr. However, 4 and 16 colonies were observed when the

transformants were grown in LB medium for 6 hr, and 16 hr,

respectively. Plasmids were purified from the 20 colonies

obtained, and all were able to transform E. coli AT2471 to

tyrosine independence.

All 12 of the 20 clones which were examined showed

cyclohexadienyl dehydrogenase activity in crude extracts (data

not shown). Restriction endonuclease cleavage analysis showed

that the 12 clones shared a 2.5-kb NcoI fragment. One of the

12 clones, designated as pJZ5, was chosen for further study

(Table 4-2). This clone carried a DNA fragment from Z.

mobilis estimated to be 10 kb in size (Fig. 4-2).









-81-

Southern blot hybridization showed that a labeled 6-kb

EcoRI-HindIII fragment of pJZ5 (Fig. 4-2) hybridized with a 6-

kb fragment of Z. mobilis chromosomal DNA digested with EcoRI

and HindIII, but did not hybridize with E. coli chromosomal

DNA digested with PstI (data not shown).


Subclonina and Expression of The Cyclohexadienyl
Dehvdroaenase Gene in E. coli

The subcloning strategy used for the clone pJZ5 is shown

in Fig. 4-2. Subclones denoted pJZ5a, pJZ5b, and pJZ5c, pJZ5d,

and pJZ5e expressed cyclohexadienyl dehydrogenase activity at

a similar level (Table 4-2). The gene was therefore localized

within the NcoI fragment. When this fragment was cloned into

pGEM-5Zf(+) at the NcoI site to give pJZ5f, a relatively high

level of cyclohexadienyl dehydrogenase activity was observed

(Table 4-2). The increase of the activity may be due to the

decrease of the distance between the vector promoter and the

transcriptional start site of the gene.

In order to localize the gene precisely, the NsiI-

(SstII)HindIII fragment was cloned into pUC18 and pUC19 by

replacement of the PstI/HindIII fragment, and the resulting

subclones, designated as pJZ5g-l, and pJZ5g-2, respectively,

were transformed into E. coli AT2471. Only strain

AT2471(pJZ5g-1) exhibited cyclohexadienyl dehydrogenase

activity in crude extract, whereas strain AT2471(pJZ5g-2) did

not yield detectable enzyme activity, indicating that the










Q) -- a) -4J
4J'-4,l
4 4.) V

S0 0H
9j5 D4.o



0



CV
0 .- 0-


ron
C,0



*0 I4)




0 0 -I
,1 r-4 -H



,-I 0.0







0 to GO
a) q-H-
r-e
U
k ,t 0r






ssS0
0 V
2 o4)0






*) 0 V
4 )$4







0 0 4* d


SJ0 r
,H 0

0. c +) V)
-H1 I r-





H o rq 0q-4










-83-


D- + + + + + + + + I I +




ISd I


IHpu!H-



I4ds-



lAUoo33-


IOON_
mrlss-
In0s_

I!SN-
InlS-
I !sN -



IOON-

I U003 -


I U0 3
CO .Q 0 0 ) 0) J
in in iO iLn n go uo to in in an
N N N N N N N N N N N
3 -3 n n -7 -7
CLa a a a a a a









-84-


Table 4-2. Expression of the Z. mobilis cyclohexadienyl
dehydrogenase gene in E. coli


Specific activity"


Plasmid ADH PDH


pJZ5 15.2 4.8

pJZ5a 16.5 5.9

pJZ5b 15.7 5.1

pJZ5c 16.0 5.1

pJZ5d 14.7 4.9

pJZ5e 17.3 6.1

pJZ5f 37.8 12.9

pJZ5g-1 40.4 14.6

pJZ5g-2 -0- -0-

pJZ5h -0- -0-

pJZ5i -0- -0-

pJZ5j 35.5 12.6

pUC18 -0- -0-

pGEM-5Zf(+) -0- -0-


aSpecific activity is defined as nmole NADH formed per min per
mg of protein.
Abbreviations: ADH, arogenate dehydrogenase; PDH, prephenate
dehydrogenase.
The plasmids were first transformed into E. coli AT2471, and
the crude extracts of the transformed AT2471 were used for
enzyme activity assay.









-85-
transcription of the gene proceeded from the same direction as

lacZ of pUC18. The promoter of this Z. mobilis gene is

apparently destroyed or removed by digestion with NsiI. Thus,

expression of this Z. mobilis gene in pJZ5g-l depended upon

the lacZ promoter of the plasmid. Consistent with the

enzymological results, the subclone pJZ5g-1 was able to

complement E coli AT2471, while pJZ5g-2 failed to do so.

When the StuI-(SstII)PstI fragment was cloned into pUC18

at SmaI-PstI sites, the resulting subclone, designated as

pJZ5i, was unable to complement E. coli AT2471. The

transformants of E. coli AT2471 carrying pJZ5i did not show

detectable activity of either arogenate dehydrogenase or

prephenate dehydrogenase. This loss of cyclohexadienyl

dehydrogenase activity suggested that the cleavage of pJZ5g by

StuI disrupted the integrity of the structural gene, a

supposition which was later confirmed by the sequencing data.


Nucleotide Sequence of The Cvclohexadienvl Dehvdrogenase
Gene

The complete nucleotide sequence of the cyclohexadienyl

dehydrogenase gene along with its flanking regions is shown in

Fig. 4-3. The structural gene was located within a single open

reading frame (ORF), 882 bp in length. The deduced amino acid

sequence yields a protein containing 293 residues with a

molecular weight of 32,086. This agreed well with the value of

32,000 determined for the purified enzyme by SDS-PAGE (Fig.

4-4). The ORF was started at codon GTG, and terminated at









-86-

codon TAA. The sequence AGGCAGG, located 8 bp upstream of the

start codon may serve as the ribosome binding site (Pond et

al., 1989; and Shine and Dalgarno, 1974).

The G+C content of the gene was 44.44% which falls into

the range for the Z. mobilis genome (Montenecourt, 1985). The

codon usage of the gene was typical of Z. mobilis (Pond et

al., 1989), showing a preference for A+T. However, the codon

usage of this gene (data not shown) was less biased than that

of highly expressed genes (Pond et al., 1989).



Purification and Identification of The Cloned Dehvdrogenase

A ratio of 3:1 was obtained in comparison of the activity

of arogenate dehydrogenase to that of prephenate dehydrogenase

(Table 4-3). This ratio remained constant throughout

purification, and the two activities of the enzyme co-eluted

from chromatography columns with identical profiles. A single

protein band with an estimated molecular weight of 32,000 was

visualized by SDS-PAGE of the purified enzyme (Fig. 4-4).

Thus, the product of the cloned gene is a cyclohexadienyl

dehydrogenase.

When the purified enzyme preparation was run on a native

gel, and then stained in Coomassie blue or in tetrazolium

blue, prephenate and NAD', a single blue band or a single

purple band was observed, respectively. The blue and purple

bands were located at the exactly the same position on the

gel. When the gel was stained in tetrazolium blue and NAD*, no










-87-


10 30 50
cctctgccaattttactctactgctgtttgaagaggcagctgacggctaaaaccgcctata
70 90 110 usix
aagccttgatggatcacggctataccacccgttggttgccgggacagcgccttcctcatg
130 150 170
cattacgtatcactatcggcagtgaaaaacatatgcaggatgtcgctggtattttaactt
190 210 230
ccttggttaggcagcgctctaagtgaccgtctttaagcatattgccattatcggattag
ss M T V F K H I A I I G L G
250 270 stul 790
gactgatcggttcctctgcggcacgggcaacaaaqggqtattgtcctgatgtaacggtca
L I G S S A A R AT K A Y C P D V T V S
310 33T 350
gtctctatgacaaaagcgaatttgtctgcgacgagactagagcgctcaatctcggcgaca
L Y D K S E F V C D R A R A L N L G D N
70 390 410
atgtcaccgatgatattcaagatgcggttcgtgaggctgatctggtgctattatgqgtgc
V T D D I Q D A V R E A D L V L L C V P
430 450 470
cagtcagggcaatgggtatcgtcgcggcagcgatggcaccggcgctgaaaaagacgtta
V R A M G I V A A A M A P A L K K D V I
490 510 530
ttatctgcgatacaggttcggtaaaagtcagcgttataaaaacgctgcaagacaatttac
I C D T G S V K V S V I K T L Q D N L P
550 570 590
ccaatcacattattgttcctagcccatctttggctgggactgaaaataacggacccgacg
N H I I V P S H P L A G T E N N G P D A
610 630 650
ccggttttgctgaattattccaagaccatcctgttattttgacccccgatgcccatacac
G F A E L F Q D H P V I L T P D A H T P
670 690 710
cggcacaggctatcgcctatatcgccgattattgggaagaaattggtgggcgtatcaatc
A Q A I A Y I A D Y W E E I G G R I N L
730 750 770
tgatgtcggcggaacatcacgatcacgttttagcgcttaccagccatttgcctcatgtca
M S A E H H D H V L A L T S H L P H V I
790 810 830
ttgcataccaacttatagggatggtatcgggttatgagaaaaaaagccggacacccatca
A Y Q L I G M V S G Y E K K S R T P I M
850 870 890
tgcgttattcggcaggcagctttcgggatgcgacgcgggtagcggcttcggaacctcgtc
R Y S A G S F R D A T R V A A S E P R L
910 930 950
tctggcaagatattatgctggaaaatgcgcctgctcttttaccagtgctggatcatttta
W Q D I M L E N A P A L L P V L D H F I
970 990 1010
tcgcagatctcaaaaaattgcggacagctattgcttcgcaagatgaggattatcttcttg
A D L K K L R T A I A S Q D E D Y L L E
1030 1050 1070
agcatttcaaagaatcgcagaaagcgcgtttagccttaaaaacagaccacgatattcacc
H F K E S Q K A R L A L K T D H D I H P
1090 1110 1130
cttaaaatttgccagtattattcagccgtcagatatttttctgacggcttttttactgat
1150 1170 satzz
taaagcgattgatcgcagcatgaacaagcgcttcggcttccttccgcgg






FIG. 4-3. Nucleotide sequence of the cyclohexadienyl
dehydrogenase gene along with its flanking regions. The
deduced amino acid sequence of the gene is shown beneath the
corresponding codons. The Shine-Dalgarno (S-D) region and the
restriction sites are underlined and labeled. The residues
which correspond to the NAD binding domain are also
underlined.










-88-


Table 4-3. Purification of the cloned cyclohexadienyl
dehydrogenase from E. coli tyrA mutant of AT2471


Total SAa Purification
protein (nmol/mg/min)
(mg) ADH/PDH factor
ADH PDH Ratio


Crude
extract 908 42 14 3 1

DEAE-
cellulose 216 150 53 2.9 3.7

Hydroxyl
apatite 19 1188 409 2.9 29

Sephadex 3.6 2357 3929 3.2 277
G-200

Hydroxyl
apatite 1.2 18825 6618 3.0 466


aSpecific activity is defined as nmol NADH formed per min per
mg of protein.
Abbreviations: SA, specific activity; ADH, arogenate
dehydrogenase; PDH, prephenate dehydrogenase.










-89-

detectable band was observed in the absence of prephenate.

When the region corresponding to the purple band was excised

(Hiebert et al., 1984) from the unstained gel and run on SDS-

PAGE, the protein migrated to a molecular weight position of

32,000. Thus, the subunit molecular weight of the cloned

cyclohexadienyl dehydrogenase was 32,000. The molecular weight

of the native enzyme was 76,000 as determined by gel

filtration on Sephadex G-200. Therefore, the native enzyme is

probably a homodimer.


N-terminal Amino Acid Sequence of The Cyclohexadienyl
Dehvdrogenase

In order to confirm that the ORF was started at the GTG

codon, the N-terminal portion of the cyclohexadienyl

dehydrogenase was sequenced. The 11 residues determined were

shown to be Met-Thr-Val-Phe-Lys-His-Ile-Ala-Ile-Ile-Gly, a

sequence which was identical to that deduced from the

nucleotide sequence (Fig. 4-3).


Kinetic and Regulatorv Properties of The Cyclohexadienyl
Dehydroaenase

Both arogenate dehydrogenase and prephenate dehydrogenase

activities of the purified enzyme required NAD' as a cofactor

and failed to utilize NADP*. Uncomplicated, first-order

substrate saturation curves were obtained. A K, value of 0.11

mM was obtained for NAD, regardless of whether the enzyme was

assayed as arogenate dehydrogenase (Fig. 4-5A) or as

prephenate dehydrogenase (Fig. 4-5B). The cyclohexadienyl









-90-


123


94,000 -
67,000 m

43,000 ,


30,000


20,100 ,










FIG. 4-4. The analysis of the cloned cyclohexadienyl
dehydrogenase purified from E. coli AT2471 by SDS-
polyacrylamide gel electrophoresis. The protein samples were
run on a 12% gel and stained with Coomassie blue. From left to
right: Lane 1, the protein markers; Lane 2, the purified
cyclohexadienyl dehydrogenase preparation after the final step
of hydroxyapatite chromatography; Lane 3, a different fraction
of the purified enzyme preparation after the final step of
hydroxyapatite chromatography.









-91-

dehydrogenase had a K, value of 0.40 mM for L-arogenate (Fig.

4-5A) and a K. value of 0.33 am for prephenate (Fig. 4-5B).

Cyclohexadienyl dehydrogenase was not sensitive to

feedback inhibition by 1 mM L-tyrosine. The enzyme was assayed

at various substrate concentrations of prephenate or of L-

arogenate, as low as 0.04 mA, in the presence of 1 mM L-

tyrosine in order to detect any possible weak competitive

inhibition. L-Phenylalanine and L-tryptophan were also tested

as possible allosteric agents, but no effects upon activity

were found.



Discussion

Identity of The Cloned Gene and Its Gene Product

The conclusion that we have cloned and sequenced the

structural gene coding for cyclohexadienyl dehydrogenase from

Z mobilis is amply supported. This gene complemented the tyrA

defect of E. coli AT2471. E. coli transformants produced

cyclohexadienyl dehydrogenase having the properties of the Z.

mobilis (unpublished, Jensen) enzyme rather than those of the

cyclohexadienyl dehydrogenase component of the E. coli T-

protein. Southern blotting analysis has established that the

tyrC gene hybridized to Z. mobilis DNA but not E. coli one.

Complementation of the tyrA mutant of coli required a NsiI-

SstII insert. The disruption of the physical integrity of this

fragment in subclone derivatives abolished functional

complementation, and resulted in the simultaneous loss of









-92-


FIG. 4-5. Double-reciprocal plots of purified cyclohexadienyl
dehydrogenase assayed as L-arogenate dehydrogenase (A), or as
prephenate dehydrogenase (B). A. Velocities on the ordinate
scale are expressed as nmoles min-' of L-tyrosine formed at
37C. A 1.0 mM concentration of NAD* (left side) and a 1.6 mM
concentration of L-arogenate (right side) was used as the
fixed substrate. B. Velocities on the ordinate scale are
expressed as nmoles min-' of NADH formed at 37C. A 1.0 mM
concentration of NAD* (left side) and a 1.0 mM concentration
of prephenate (right side) was used as the fixed substrate.









-93-

arogenate and prephenate dehydrogenase activities in crude

extracts of the transformants.

Only one ORF was localized within the NsiI-SstII

fragment. This ORF would encode a protein with a calculated

molecular weight of 32,086, a value which agrees well with the

subunit size determined for the purified enzyme by SDS-PAGE.

A StuI restriction site was located in the 5'-terminus of the

ORF in a region coding for the NAD* binding site. A subclone

retaining the StuI-SstII fragment was unable to complement the

tyrosine auxotroph of E. coli since most of the coding region

for NAD* binding domain was cleaved upon digestion by StuI.


Synonymy of Prephenate Dehvdrogenase and Arocenate
Dehvdrogenase

Biochemical studies with cyclohexadienyl dehydrogenase

from P. aeruainosa (Xia and Jensen, 1990) have indicated

that prephenate and L-arogenate are cyclohexadienyl substrates

utilized at a common catalytic site. Similarly, the Z. mobilis

enzyme cloned in E. coli exhibited a constant ratio of

prephenate dehydrogenase:arogenate dehydrogenase during

purification to electrophoretic homogeneity. Strictly

competitive inhibition produced by each cyclohexadienyl

substrate in the presence of the other, as well as an

identical Km value for NAD* regardless of which cosubstrate

was present indicated a monofunctional protein exhibiting

substrate ambiguity. The genetic evidence presented here

definitively establishes that prephenate dehydrogenase and