Citation
Regulation of an operon coding for the components of the molybdate-transport system (modABCD) by mode protein, a molybdate-dependent repressor in Escherichia coli

Material Information

Title:
Regulation of an operon coding for the components of the molybdate-transport system (modABCD) by mode protein, a molybdate-dependent repressor in Escherichia coli
Creator:
Grunden, Amy M., 1971-
Publication Date:
Language:
English
Physical Description:
xiv, 108 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Amino acids ( jstor )
DNA ( jstor )
Gels ( jstor )
Genetic mutation ( jstor )
Molybdates ( jstor )
Molybdenum ( jstor )
Operon ( jstor )
Plasmids ( jstor )
Proteins ( jstor )
Sodium ( jstor )
Dissertations, Academic -- Microbiology and Cell Science -- UF ( lcsh )
Microbiology and Cell Science thesis, Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 101-107).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Amy M. Grunden.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
026310551 ( ALEPH )
36659334 ( OCLC )

Downloads

This item has the following downloads:


Full Text











REGULATION OF AN OPERON CODING FOR THE COMPONENTS OF THE MOLYBDATE-TRANSPORT SYSTEM (mod4BCD) BY MODE PROTEIN. A
MOLYBDATE-DEPENDENT REPRESSOR IN ESCHERICHIA COLI












By

AMY M. GRUNDEN

















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 1996














ACKNOWLEDGMENTS


I would like to thank Dr. K. T. Shanmugam for his guidance and support. His insistence on scientific rigor and enthusiasm for research has certainly influenced my development as a scientist. I would also like to thank Drs. John Gander, Lonnie Ingram, Philip Laipis, and James Preston for their support and counsel while serving as members of my committee. My gratitude is also offered to all of my associates from Dr. Shanmugam's lab group for their encouragement and insights. Special thanks are extended to my husband for his tremendous support, encouragement and patience, as well as to my parents, in-laws and other family members and friends.























ii















TABLE OF CONTENTS





ACKNOWLEDGMENTS ............................................ ii

LIST OF FIGURES .................................................. v

LIST OF TABLES ................................................. vii

LIST OF ABBREVIATIONS ......................................... viii

LIST OF GENE SYMBOLS ................................................ xi

ABSTRACT ........................................ xiii

INTRODUCTION ........................................ 1

LITERATURE REVIEW ............................................... 4

Molybdate Transport .............................................. 8

Regulation of the modABCD Operon ...................... ........... 11

MATERIALS AND METHODS ........................................ 13

Materials ......................................................... 13

Bacterial Strains .......................... ........................... 13

Media ........................................... ...... ......... 13

Enzyme Activities and Culture Conditions ............................... 16

Genetic Experiments ...................................... ............ 16

Molecular Biology Experiments ......................................... 16


iii









Biochemical Characterization of ModE ................................. 23

RESULTS AND DISCUSSION ......................................... 30

Evidence for a regulator of the modABCD Operon .......................... 30

Analysis of modEF DNA ............................................ 35

Regulation ofmodA by ModE ....................................... 41

Characterization of the Interaction of ModE with modA-operator/promoter DNA .. 47 Molybdate-independent ModE Proteins .................................. 82

Regulation of modE .............................................. 88

Regulation of Other Genes by ModE ................................... 91

CONCLUSION ...................................................... 97

REFERENCES ........................................ 101

BIOGRAPHICAL SKETCH ........................................... 108


























iv









LIST OF FIGURES


Figure page

1 Steps involved in the biosynthesis of the organic portion of
the molybdenum cofactor in E. coli .................. 7

2 DNA sequence of the modEF operon. .. ............ 36

3 Amino acid sequences of homologs of E. coli ModE protein . 38

4 SDS-PAGE analysis of proteins from different stages of
ModE purification . . .. . . 43

5 SDS-PAGE analysis of proteins produced in an in vitro transcriptiontranslation experiment in which plasmid pRMI served as the template DNA. 45

6 Intergenic region ofmodEA DNA .................. 49

7 DNA-mobility shifts using plasmid pAM13-derived (panel I), plasmid
pAM 5-derived (panel II), and plasmid pAM16-derived DNA (panel III)
and M odE . . . . . . 54

8 DNA-mobility shift featuring 42-bp modA4-operator/promoter
DNA (-9 to +33) and ModE. ................. ...57

9 DNA-mobility shifts using a 43-bp oligomer spanning the -17 to +25
region of the modA-operator/promoter region and ModE. ........ 59

10 Percentage of ModE-43-bp-oligomer complex 'A' formed versus
concentration of ModE protein in the presence of 1 mM sodium molybdate. 61

11 Percentage of 43-bp-oligomer bound to ModE protein in relation
to the concentration of sodium molybdate present in the binding
reactions in a DNA-mobility shift experiment ..... ........ 62

12 DNA-mobility shift experiment in which four different oxyanions
were present in the binding reactions. . . . 68

13 Percent shifted DNA versus concentration of ModE resulting from DNAmobility shift binding reactions containing various oxyanions. ...... 69



V










14 Ferguson plot generated from the Rf values obtained after electrophoresis of the
non-denatured protein standards and ModE-43-bp-modA-operator/promoter
DNA complex in 5%, 6%, 7%, and 8% polyacrylamide-TBE gels. .... 72

15 DNase I-footprinting in which the binding reactions contained 1 mM sodium
molybdate, 0.1 pmol 32P-labeled-446-bp-modA-operator/promoter DNA
and various concentrations of ModE. ................ 73

16 DNase I-footprint of ModE in modA-operator/promoter DNA in which
sodium molybdate was excluded from the binding reaction mixtures. ... 76

17 DNA-mobility shift experiments using the modA-operator/promoter DNA
contained in plasmids pAM17, pAM18, and pAM19 and ModE. ..... 80

18 Plot of % shifted DNA versus concentration of ModE required for
the shift featuring the 36-bp modA DNA present in plasmid pAM17 ..... 81

19 Analysis of ModE mutants ............. ........ 83



























vi
















LIST OF TABLES


Table page

I Bacterial strains and phages used in this study. ............ .14

2 Plasmids used in this study...................... 17

3 Effect of mod gene mutation in mod operon transcription. ... .. 32

4 Complementation of the mutation in strain SE1811 for molybdatedependent repression of the modABCD operon. ....... . 34

5 Complementaion of the modE mutation in strain SE 1811 with H. influenzae ModE ..................... 40

6 Purification profile of ModE ................. 42

7 Titration of ModE protein in vivo by various DNA sequences in strain SE2069 (modA '- 'lacZ*)................... 50

8 Expression of modA '-'lacZ in strain SE2069 cultured in the presence of various oxyanions. ............... 65

9 Titration of ModE protein in vivo by various regions of the modA-operator/ promoter DNA in strain SE2069 (modA '- 'lacZ). ........ ... 78

10 List of proteins which are known to interact with molybdate and DNA and contain a possible molybdate-binding motif. ........... 86

11 Regulation of modE in different mod strains. .......... .. 89

12 Regulation of modE in various backgrounds. ...... ... ... 90

13 Regulation ofmodFby ModE ............ ....... 92

14 Regulation of narG expression by ModE ............. 95


vii













LIST OF ABBREVIATIONS


ABC....................................... ATP-binding cassette

ATP.............................. Adenosine triphosphate

A. vinelandii........................ Azotobacter vinelandii

AvModE.................................... Azotobacter vinelandii ModE protein

bp............................ ................ Base pair

C. pasteurianum............................ Clostridium pasteurianum

CpMopII.................................... Clostridium pasteurianum MopII protein

CTP................................ Cytosine triphosphate

Da................................... Dalton

DNase I.................................... Deoxyribonuclease I

DMSO.................................. .. Dimethyl sulfoxide

DTT................................... Dithiothreitol

EcModE.............................. ... Escherichia coli ModE protein

E. coli................................. Escherichia coli

EDTA............................... Ethylenediaminetetraacetic acid

FDH-H................................ Formate dehydrogenase linked to H-, evolution

FeMoCo............................... Iron-molybdenum cofactor

FHL................................ Formate hydrogenlyase

GTP.................................. Guanosine triphosphate


viii











H. influenzae ............................... Haemophilus influenzae

HiModE ....................... Haemophilus influenzae ModE protein

HYD.......................................... Hydrogenase

IPTG.......................................... Isopropyl-0-D-thiogalactopyranoside

KD................................................. Dissociation constant

kDa ............................................ Kilodalton

Kinase ................................ T4 polynuclotide kinase

Klenow.................................. DNA polymerase I- Klenow fragment

LB ........................................... Luria Broth

LBG................................ Luria Broth + glucose

Mo................................... Molybdenum

MoCo................................. Molybdenum cofactor

ModE................................. ModE protein

MGD.................................. Molybdopterin guanine dinucleotide

MPT..................................... Molybdopterin

OD .............................. Optical density

ONPG.................................... Ortho-nitrophenyl-3-D-galactopyranoside

ORF..................................... Open reading frame

PAGE......................................... Polyacrylamide gel electrophoresis

PCR..................................... Polymerase chain reaction

R. capsulatus........................... Rhodobacter capsulatus



ix









RcMopA....................................... Rhodobacter capsulatus MopA protein

SDS......... ............... Sodium dodecyl sulfate

TMAO.......................................... Trimethylamine-N-oxide

Tris............................... Tris-(hydroxymethyl)-aminomethane

TTP.................................. ..... Thymine triphosphate

UV ............................... Ultraviolet

X-gal................................ 5-Bromo-4-chloro-3-indo lyl-p-D-galactopyranoside





































X












LIST OF GENE SYMBOLS


All the genes listed below are from Escherichia coli unless otherwise indicated. Gene Alternate gene symbols: phenotype affected


arcA Aerobic respiratory control; putative DNA binding protein of Arc modulon arcB Aerobic respiratory control; histidine-protein-kinase of Arc modulon bla _-1actamase; ampicillin resistance chiD Chlorate resistance gene D [renamed as mod; molybdate transport] fdhF FDH-H; formate dehydrogenase (linked to H, evolution) fhlA Putative DNA binding protein necessary for transcriptional activation of
the fdhF and hyc operons

fnr 'Global' regulator of anaerobic respiration; similar to cAMP receptor
protein

galETK _Qactose operon

hyc Hydrogenase isoenzyme-3 1sg H. influenzae lipopolysaccharide synthesis gene moa Molybdopterin biosynthesis; as chlorate resistance gene A mob Molybdopterin biosynthesis; as chlorate resistance gene B mod Molybdate transport; as chlorate resistance gene D moe Molybdopterin biosynthesis; as chlorate resistance gene E mog Molybdopterin biosynthesis; as chlorate resistance gene G



xi









mop Molybdopterin binding protein narL Putative DNA binding protein of respiratory nitrate reductase operon narQ Putative nitrate sensor for activation of the NarL protein narX Putative nitrate sensor for activaton of the NarL protein nit-] mutation allele resulting in nitrate reductase deficiency in Neurospora
crassa





































xii














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


REGULATION OF AN OPERON CODING FOR THE COMPONENTS OF THE MOLYBDATE-TRANSPORT SYSTEM (modABCD) BY MODE PROTEIN, A
MOLYBDATE-DEPENDENT REPRESSOR IN ESCHERICHIA COLI By

Amy M. Grunden

December, 1996

Chairperson: Dr. K. T. Shanmugam
Major Department: Microbiology and Cell Science

The modABC genes encode a molybdate-specific transporter in Escherichia coli. When the intracellular molybdate concentration is sufficient to support the activity of molybdoenzymes, expression of the modABCD operon is very low and increases under molybdate-deficient conditions. Isolation and characterization of a mutant strain (strain SE 1811), which maintained the same level of expression of the modABCD operon in both the presence and absence of molybdate in the medium, resulted in the identification of a regulatory gene, designated modE. The modE gene constitutes the first gene in a twogene modEF operon which diverges from the modABCD operon and codes for a 262amino-acid protein (28 kDa). ModE, in the presence of molybdate, repressed the production of plasmid-encoded ModA and ModB' proteins in an in vitro transcriptionxiii









translation system. DNA-mobility shift experiments demonstrated that ModE is capable of binding as a dimer to DNA derived from the modA -operator/promoter region with an apparent KD of 0.22 ptM in the presence of 10 pLM molybdate or with an apparent KD of

1.3 jiM in the presence of 10 pM tungstate. DNA-mobility shift experiments also revealed that a 50% shift of DNA is achieved at a concentration of 6 j[M molybdate in the binding reaction. DNase I-footprinting experiments identified three regions of protection by ModE in the modA operator/promoter region, 'GTTATATTG' at position -15 to -7, 'CCTACAT' at position -3 to +4, and 'GTTACAT' at position +8 to +14. In vivo ModE titration experiments confirmed that the indicated protected regions are required for efficient binding of ModE to DNA. A highly conserved amino acid sequence TSARNQXXG (amino acids 125 to 133) was identified in ModE and homologs from Azotobacter vinelandii, Haemophilus influenzae, Rhodobacter capsulatus, and Clostridium pasteurianum. E. coli ModE mutant proteins with mutations in either the T or G repressed transcription of the modABCD operon in the absence of molybdate, suggesting that this stretch of amino acids is essential for binding molybdate by ModE protein. This study provides evidence that ModE serves as a repressor of the modABCD operon in response to intracellular molybdate levels.












xiv














INTRODUCTION

Molybdenum is required for the activity of several enzymes found in animals, plants, and bacteria such as sulfite oxidase, xanthine dehydrogenase, nitrate reductase, formate dehydrogenase, and nitrogenase (77). In humans, sulfite oxidase, xanthine dehydrogenase, and aldehyde oxidase have all been identified as molybdoenzymes, and the absence of these molybdoenzymes, particularly sulfite oxidase, results in severe pathological conditions (37, 57). This medical concern coupled with the fact that nitrogen cycling, which is principally mediated by nitrogen assimilation in plants, requires the participation of molybdoenzymes, has spurred interest in characterizing the steps involved in the successful incorporation of molybdenum into the appropriate apo-molybdoenzymes.

From structural studies, it has been determined that molybdenum exists in

molybdoenzymes in the form of a pterin-containing molybdenum cofactor (MoCo), with the exception of nitrogenase which has an iron-molybdenum cofactor (FeMoCo) (1, 31, 57). Other studies have described mutations which result in the pleiotropic loss of molybdoenzyme function (31). One set of these mutants, defined as chl mutants in the bacterium Escherichia coli, were isolated as chlorate resistant mutants and a major fraction of these were found to be defective in all molybdoenzymes, including nitrate reductase (18, 22, 23, 72).





I











A subset of these chl mutants, designated as chlD, exhibited a pleiotropic loss of molybdoenzymes that could be suppressed by molybdate supplementation of the growth medium (14, 22, 29, 48, 61, 67, 71). Further characterization of the chlD mutants indicated that the genes in this location comprise the molybdate-specific transporter (now called modABC; 69), with modA encoding a molybdate-specific periplasmic binding protein, modB encoding an integral membrane channel-forming protein, and modC encoding an ATP-binding energizer protein (33, 47, 58, 60, 75). This molybdate-specific transporter closely resembles the established ATP-binding cassette (ABC) transporter motif(3, 30). Moreover, similar high-affinity molybdate transporters have been described for Azotobacter vinelandii, Rhodobacter capsulatus, and Haemophilus influenzae (20, 43, 76).

The fact that molybdate supplementation of the growth medium restores

molybdoenzyme activity in E. coli mod mutant strains suggests that molybdate is capable of entering the cell through alternate routes, and preliminary experiments indicate that under high molybdate concentration conditions, molybdate can be transported by the sulfate-specific transporter as well as through other nonspecific anion transporters (41, 61). The molybdate uptake kinetics were also analyzed in mod mutant strains, and as expected, mod mutants exhibit much lower rates of molybdate transport as compared to mod' strains (14, 29).

Having shown that the modABCD genes code for a molybdate-specific transporter, studies concerned with the regulation of expression of the modABCD operon in E. coli were conducted (24, 48, 58, 61), and it was determined that expression of this operon is









3

regulated by the intracellular concentration ofmolybdate. Specifically, high levels of intracellular molybdate resulted in reduction of transcription of the modABCD operon, which implies that E. coli has a molybdate-dependent repressor which regulates expression of the modABCD operon.

Isolation of an E. coli strain (strain SE1811) that did not exhibit the classical repression of the modABCD operon upon molybdate-supplementation of the growth medium, afforded a starting point for the study of the molybdate-dependent repression of the modABCD operon. Determination of the location of the mutation in strain SE1811 led to identification and subsequent characterization of the putative molybdate-dependent repressor (ModE) of the modABCD operon. This study serves as an in-depth examination of the ModE and its interaction with the modABCD operon, which will ultimately further the understanding of the production of the molybdenum cofactor in E. coli and possibly other organisms as well.














LITERATURE REVIEW


One of the hallmark features of many bacteria is the metabolic diversity of their energy producing systems, and the bacterium Escherichia coli is no exception. E. coli, a facultative anaerobe, is capable of switching among its respiratory pathways depending upon the availability of terminal electron acceptors. A hierarchy of electron acceptors exists in E. coli as determined by their oxidation-reduction potential. Given its large positive redox potential (+ 0.82 V), dioxygen serves as the most efficient electron sink followed by a series of acceptors with incrementally smaller positive redox potentials. When E. coli is undergoing aerobic respiration, the proteins contributing to other less energetically favorable pathways of respiration are not synthesized. However, in the absence of dioxygen, the components of other respiratory pathways may be produced. Depending upon the availability of alternative electron acceptors, nitrate respiratory components may be synthesized followed by those necessary for the utilization of other alternative electron acceptors such as dimethyl sulfoxide, trimethylamine-N-oxide, and fumarate-utilizing respiratory pathways (25).

In E. coli, a number of proteins which transfer electrons from a donor to the final acceptor are molybdoenzymes. These include nitrate reductase, dimethly sulfoxide reductase, and trimethylamine-N-oxide reductase and are molybdoenzymes. All of these proteins contain the molybdenum at the site of electron transfer (77). Structural studies


4








5

have shown that the molybdenum in these enzymes is present as a molybdopterin guanine dinucleotide (MGD) moiety (4, 5, 35. 57).

The identification of the molybdenum cofactor structure as MGD and the

determination of the pathway by which it is synthesized resulted from the work of many individuals presented in a number of studies. One early revelation in this inquiry resulted from the identification of a series of pleiotropic mutations in Aspergillus nidulans each of which caused a deficiency in the molybdoenzymes xanthine dehydrogenase and nitrate reductase (53). This initial finding led investigators to believe that there is a molybdenum cofactor common to molybdoenzymes. It was subsequently demonstrated that the inactive apoprotein nitrate reductase produced by a pleiotropic nit-I mutant of Neurospora crassa could be reconstituted by the addition of denatured preparations of purified molybdoenzymes from animal, plant, fungal, or bacterial sources (34). This finding further indicated that molybdoenzymes have a dissociable molybdenum moiety and also showed that this cofactor was common to a variety of organisms.

Since the native molybdopterin cofactor proved extremely labile, initial attempts to determine the structure of the native molybdenum cofactor involved derivatization of the cofactor using vigorous oxidation at 1000 C in the presence of iodine gas and potassium iodide (KI), creating derivative Form A or KI oxidation in the presence of air resulting in derivative Form B (36). Analysis of these derivatives led to the identification of the molybdopterin (MPT) structure for the molybdenum cofactor. Later structural studies showed that dinucleotides such as guanine dinucleotide, adenine dinucleotide, cytosine dinucleotide, and hypoxanthine dinucleotide can also be associated with the cofactor in the











native molybdoenzymes (35). In E. coli, all of the molybdoenzymes contain the molybdopterin guanine dinucleotide form of the molybdenum cofactor (57).

Determination of the structure of the active form of the molybdenum cofactor facilitated studies of the biosynthetic pathway by which the molybdenum cofactor is produced. A great deal of this work was accomplished in E. coli since a number of pleiotropic molybdoenzyme mutants had been isolated as chlorate-resistant strains (18, 22, 23, 72). Chlorate, a structural analog of nitrate, can be reduced by nitrate reductase to chlorite, and in solution, chlorite exists in the form hypochlorite which is toxic to the cell. The chlorate-resistant mutants that have been isolated can be classified as those which are only deficient in nitrate reductase due to a mutation in the nitrate reductase structural genes (chiC renamed narGHJI; 4, 18, 31, 72, 73) and those which are pleiotropic (defined as chlA, renamed moa; chlB, renamed mob; chlD, renamed mod; chlE, renamed moe; chlG, renamed mog; 69). These pleiotropic mutants are deficient in nitrate reductase and presumably all of the other molybdoenzymes in E. coli (formate dehydrogenase-N, formate dehydrogenase-H, DMSO/TMAO reductase, and biotin-sulfoxide reductase) because of a failure to synthesize and incorporate the molybdenum cofactor into the appropriate apoprotein. However, it has been found that the mog mutants do contain functional molybdoenzyme formate dehydrogenase-H (73).

From the characterization of these various pleiotropic chlorate-resistant E. coli mutants, the biosynthetic pathway for MGD synthesis in E. coli was established and is illustrated in Fig. 1 (57). It was found that the biosynthesis of MGD begins with the conversion of guanosine or a guanosine derivative by the moa encoded proteins to a









7












Guanosine or Guanosine derivative

MoaA,B,C
Converting Factor Proteins
Large Subunit (MoaE protein)
Precursor Z

Active Converting Factor


Converting Factor Converting Factor-MPT Complex
Small Subunit (sulfo)
Intracellular MoO4 "Activated Mo" MoeB Protein (?) Mod, MoeA Proteins
Converting Factor Molybdenum Cofactor
Small Subunit (MoaD protein) (Molybdopterin Form)

SSource of GMP Mob protein r

Molybdenum Cofactor (Guanine Dinucleotide Form)










Fig. 1. Steps involved in the biosynthesis of the organic portion of the molybdenum cofactor in E. coli (modified from Rajagopalan and Johnson, Ref. 57)









8

precursor compound, termed precursor Z. Precursor Z joins with active converting factor to produce the converting factor-MPT complex. The active converting factor responsible for the conversion of precursor Z to the converting factor-MPT complex is capable of mediating the conversion only after the small unit of the converting factor (MoaD) protein is first activated by the MoeB protein so that it can then form a complex with the large subunit of the converting factor (MoaE protein), thereby producing the active converting factor. At this point, it is believed that activated molybdenum combines with the molybdopterin molecule. In the presence of guanosine triphosphate, the Mob protein then guanylylates the molybdopterin form of the molybdenum cofactor to the guanine dinucleotide form of the cofactor.

Molvbdate Transport

As has already been indicated, a number of pleiotropic E. coli chlorate-resistant mutants exist, and one type in particular, referred to as mod mutants, is deficient in the uptake of molybdate in the absence of molybdate supplementation of the growth medium (18, 22, 23, 71). Sequencing of the mod genes, located in the 17 minute region of the chromosome, resulted in the identification of a four gene operon designated modABCD. modA codes for a 257 amino acid molybdate-specific periplasmic binding protein; modB codes for a 229 amino acid integral membrane channel-forming protein; modC codes for a 352 amino acid transport energizer protein; and modD codes for a 231 amino acid putative outer membrane protein which may function as a porin (33, 47, 58, 60, 75). It should be noted that the molybdate-specific transporter encoded by the modABCD operon resembles the ATP-binding cassette transporter motif in which there is a substrate-specific








9

periplasmic-binding protein (3, 30), an integral membrane channel-forming protein(s), and an ATP-binding protein which is thought to couple the hydrolysis of ATP to transport.

Moreover, when molybdate uptake kinetics were investigated in an E. coli mod mutant strain, the rate of uptake was substantially lower than that seen in the wild type strain, and this transport was shown to be an energy requiring process (14, 29). Other presumptive mod mutant strains transported molybdate as efficiently as the wild type strain, but these strains were not capable of retaining the imported molybdate, as evidenced by a rapid loss of accumulated molybdate when chased with unlabeled molybdate (14). This result implies that the mutation in this second group of mutants is not located in one of the genes coding for the molybdate-specific transporter; it apparently is located in a gene necessary for the incorporation of the molybdenum into the cellular material.

Homologs of the E. coli molybdate-specific transport proteins have also been

identified in other facultative anaerobes such as H. Influenzae Rd as well as the nitrogenfixing bacteria A. vinelandii and R. capsulatus (20, 43, 76). In the case of the H. influenzae Rd Mod proteins, a 49% identity and 58% similarity for ModA was determined while a 65% identity and 73% similarity for ModB, and a 53 % identity and 66% similarity for ModC were found in relation to the respective E. coli proteins. The A. vinelandii homologs show 27% identity and 38% similarity compared to E. coli ModA, 31% identity and 47% similarity for ModB, and 44% identity and 58% similarity for ModC. The R. capsulatus molybdate transport protein homologs have 29% identity and 40% similarity








10

for ModA, 29% identity and 45% similarity for ModB, and 39% identity and 55% similarity for ModC compared to the E. coli proteins.

Aside from the sequence similarity implicating the A. vinelandii and R. capsulatus homologs' involvement in molybdate-specific uptake, there are also genetic studies which suggest that these proteins do indeed constitute the molybdate-specific transporter in the respective organism. One such study (50) demonstrated that mutations in the modB and modC genes of A. vinelandii allowed for expression of the alternate nitrogenase (heterometal-free nitrogenase) in the presence of 0.5 NtM sodium molybdate. Since it has previously been shown that expression of the alternate nitrogenase is repressed in the presence of molybdate in the wild type (43), it may be concluded that these mutations prevent molybdate-specific transport. Furthermore, mutations in modA, modB, or modC of R. capsulatus resulted in a loss of activity for the molybdenum-containing nitrogenase as well as derepression of the transcription of the genes encoding the heterometal-free alternative nitrogenase, which again indicates a loss of molybdate-specific transport by these mutants (76). In both of these organisms, increasing the concentration of the molybdenum in the medium resulted in repression of transcription of the alternate nitrogenase genes.

The fact that mutations in genes encoding the molybdate-specific transporter

proteins can be suppressed by the addition of sufficiently high concentrations of molybdate to the growth medium suggests that alternate routes for molybdate transport exist (41, 61). Furthermore, there is evidence that molybdate is capable of being transported through








11

the sulfate/thiosulfate and selenite transport systems (41, 61), which is not surprising given the similarity in charge and structure among molybdate, sulfate, and selenate.

Regulation of the modABCD Operon

Expression of the modABCD operon has been investigated in a variety of genetic backgrounds by means of monitoring P-galactosidase activities in strains harboring a(modA '- 'lacZ) (58, 60, 61). From these experiments, it was found that modA '- 'lacZ expression was high under conditions of low intracellular molybdate as in the case of mod mutants grown in medium lacking molybdate. However, when there is a high enough intracellular molybdate concentration to support nitrate reductase or formate hydrogenlyase activity as is the case in a molybdate transport competent strain or a mod strain grown in molybdate supplemented medium, modA '-'lacZ expression is very low. Since molybdate availability seemed to effect expression of the modABCD operon, the possible role of MGD synthesis in mod expression was also explored, and it was found that mutations in the moa, mob, or moe operons did not appreciably change expression of the modABCD operon. However, modA '-'lacZ expression was increased approximately twofold when the cells were grown under aerobic conditions or when cultured anaerobically in a strain containing anfnr mutation (60, 61). Inspection of the DNA sequence upstream of modA transcription start site did not reveal any putative FNR protein binding motif, consequently, this increase in expression of the modABCD operon in anfnr background is most likely a physiological effect of thefnr mutation on other operons like narGHJI.

A study by Rech et al. (58) examined the effects of molybdate, nitrate and oxygen on modABCD expression, and again it was found that molybdate limitation enhanced








12

modABCD expression, whereas nitrate and oxygen did not significantly modify expression levels. This study also evaluated the DNA sequence in the modA operator/promoter region which might be involved the molybdate-dependent regulation of the modABCD operon. It was determined by mutation analysis that a 'CATAA' sequence located at position +2 to +6 in relation to the modA transcription start site is involved in binding the molybdate-dependent regulator; however, a similar 'CATTAA' sequence located at position +12 to +17 was not required for regulation.

Further analysis of the modA operator/promoter DNA revealed that there is an eight base pair (bp) inverted repeat sequence ('TAACGTTA') spanning the +4 to +11 region. This inverted repeat may serve as the actual recognition sequence for the molybdate-dependent regulator of the modABCD operon since the inverted repeat sequence overlaps the +2 to +6 sequence that had been implicated in regulator protein binding. Inverted repeat sequences have been shown to constitute the binding sites for a number of repressor proteins such as the MetJ repressor protein and the Trp repressor (40, 54, 55).

Although the molybdate-dependent repression of modABCD expression has been reported by several investigators, the causative agent of this repression had not been determined. This study identifies this molybdate dependent repressor, designated ModE, and also locates the recognition sequences in the modA operator/promoter DNA required for binding the repressor. The interaction between the repressor and its target DNA is examined under a variety of conditions.















MATERIALS AND METHODS


Materials

Biochemicals were purchased from Sigma Chemical Co (St. Louis, MO).

Inorganic and organic chemicals were obtained from Fisher Scientific Co. and were analytical or molecular biological grade. Restriction endonucleases and DNA modifying enzymes were purchased from New England Biolabs, Inc. (Beverly, MA) or Promega (Madison, WI). S30 cell extract was purchased from Promega (Madison, WI). Sequenase 2.0 was obtained from United States Biochemical Corporation (Cleveland, OH).

Bacterial Strains

Bacterial strains used in this study are presented in Table 1. All bacterial strains used are derivatives of Escherichia coli K-12.

Media

Bacterial cultures were routinely grown in Luria Broth (LB) which contains 1% trypticase peptone (BBL, Cockeysville, MD), 0.5% yeast extract (BBL) and 0.5% NaC1. When necessary, the LB was supplemented with 0.3% glucose (LBG) and I mM sodium molybdate. Lactose-MacConkey agar was prepared by adding filter-sterilized lactose






13










14



Table I. Bacterial strains and phages used in this study.


Strain or Phage Geneotype Source and/or reference


Strains
BW545 A(lacU) 169 rpsL Laboratory collection MC4100 araD139 A(argF-lacU)205 rpsL 150 re/Al flbB5301 deoCI ptsF25 CGSC 6152 ECL618 arcA2 zjj::TnlO E. Lin RK4353 MC4100 gyrA219 non-9 V. Stewart RK5278 RK4353 narL215::TnlO10 V. Stewart VJS720 chlD247::Tn10 (modB247) V. Stewart VJS 1779 RK4353 moe-251::TnlOd(Tc) V. Stewart VJS1780 RK4353 mob-252::Tnl10d(Tc) V. Stewart VJS1782 RK4353 moa-254::TnlO10d(Tc) V. Stewart VJS 1784 RK4353 mog-256::Tn10d(Tc) V. Stewart BL21 ,DE3 hsdS gal [lacUV5-genel (T7)] F. Studier (74) SE1 325 cysC43 srl-300::Tn10 thr-1 leu-6 thi-1 lacY1 galK2 ara-14 xyl-5 ml-1 proA2 his-4 argE3 rpsL31 tsx-33 supE44 modBI38::Tn5 Laboratory collection SE1592 BW545 modA 115 Laboratory collection SE1595 BW545 modCl 18 Laboratory collection SE1602 BW545 modB125 Laboratory collection SE1811 SE2069 modEl This study SE1910 BW545 (AmodE-km)2 (24) SE 1934 RK4353 (AmodF-km)1 (24) SE1938 BW545 ,RM26 (24) SE 1940 BW545 (AmodF-km)1 RM26 P1 transduction (SE1938 x SE1934)
SE1942 BW545 (AmodE-km)2 ,RM26 PI transduction (SE1938 x SE1910)
SE1952 BW545 modBl38::Tn5 XRM26 P1 transduction (SE 1938 x SE1325)
SE2069 BW545 0(modA'-'lacZ) 102 Laboratory collection SE2105 BW545 XRM37 R. Ray SE2106 SE1910 ARM37 R. Ray SE2107 BW545 (AmodF-km)l XRM37 R. Ray SE2112 BW545 (AmodEF-km)l XRM37 R. Ray SE2114 BW545 modB::TnlO XRM37 R. Ray SE2116 SE2106 modB::TnlO R. Ray SE2118 SE2107 modB::TnlO R. Ray SE4110 MC4100 (AmodE-km)2 This study










15


Table 1. Cont.


Strain or Phage Geneotype Source and/or reference


Phages
XRZ5 'bla'lacZ lacr Laboratory collection XRM26 ?XRZ5 [QD(modE'-'lacZ)] (24) ASE 1 RZ5 [ D(modA '-'lacZ )] Laboratory collection XRM37 XRZ5(4P(modF'-'lacZ )] R. Ray ,PC50 XRS45(bla '-lacZ.) [ (narG'- 'lacZ)] R. P. Gunsalus









16

(final concentration 1%) to MacConkey agar base (Difco, Detroit, MI) after autoclaving. Ampicillin (100 ig/ml), kanamycin (50 pg/ml), tetracycline (15 pg/mL), chloramphenicol (15 pg/ml), and X-gal (40 ptg/ml) were included in the media as needed.

Enzyme Activities and Culture Conditions [-Galactosidase activities. Standing overnight LB cultures were used to inoculate (5% v/v) fresh medium in 13x100 mm screw cap tubes. The cultures in filled tubes were incubated at 370C until late-exponential or early stationary phase had been attained (4 to 6 h). The 3-galactosidase activity of Sodium-dodecyl-sulfate (SDS)-chloroform permeabilized cells was assayed by measuring the rate of hydrolysis of o-nitrophenyl-P-Dgalactopyranoside (ONPG) as per Miller (49). The activities are expressed as nanomoles of o-nitrophenol produced per minute per milligram of protein where a value of 350 ipg protein/ml was used as the protein content of I ml of cells at an optical density (O.D.) of

1.0 unit in a Spectronic 710 (Rochester, NY) spectrophotometer.

FHL activities. Formate hydrogen lyase activity was qualitatively determined as previously described (61).

Genetic Experiments

All genetic experiments were conducted as per standard protocols (49).

Molecular Biology Experiments

Plasmid constructions. The plasmids employed in this study are listed in Table 2. Standard published procedures (45) or DNA modifying enzyme suppliers' recommendations were used for plasmid constructions and manipulations.










17

Table 2. Plasmids used in this study


Plasmid Relevent Genotype Source or reference


pACYC 184 Cam' Tet' (11) pBR322 Amp' Tet' (8) pTrc99A Ptrc lacP Amp' (2) pT7-7 pT7 lacZ' Amp' J. Maupin pUC19 Amp' (78) pUC4K lacZ' Kan' Amp Pharmacia pZ1918 'lacZ Amp' H. Schweizer (66) pMAK705 'lacZ rep(ts) cam' (26) pFGHI5 pACYC184 (modABCDEF) gal' (47) pSE1004 pBR322 (modABCDE)' (41) pSE1009 pUC 19 (modABC)' (41) pAG I pTrc99A (modE) This study pAG 1-A76V pTrc99A (modE76) This study pAGI-T1251 pTrc99A (modE125) This study pAG 1-G 33D pTrc99A (modE133) This study pAG 1-Q216* pTrc99A (modE216) This study pRM 1I pUC 19 (modA'B) (24) pRM9 pBR322 [modE-km (modABCD)'] (24) pRMIO pMAK705 (modE'-km) (24) pRM12 pUC 19 (modEF) 'galE' (24) pRM13 pBR322 (modEF) galG' (24) pRM14 pBR322 [4(modE'-'lacZ)] This study pRM 16 pBR322 (modE modF-km) (24) pRM 17 pMAK705 ['modE 0(modF-km)] (24) pRM22 pT7-7 (modEF) (24) pRM23 pT7-7 (modF) (24) pRM25 pBR322 (A HindIII-AvaI fragment) (24) (modEF) "galE'
pRM26 pBR322 ['D(modE'-lacZ) modF] (24) pAM4 pUC 19 (-247 to +25 modA operator/ This study promoter)
pAM5 pUC 19 (-5 to +25 modA operator/promoter) This study pAM6 pUC 19 (-7 to +25 modA operator/promoter) This study pAM7 pUC 19 (-55 to +25 modA operator/ This study promoter)
pAM8 pUC 19 (-30 to +25 modA operator/ This study promoter)
pAM9 pUC 19 (-247 to -55 modA operator/ This study promoter)
pAM 10 pUC 19 (-1 1 to +25 modA operator/promoter) This study










18




Table 2. Contd.


Plasmid Relevant Genotype Source or reference


pAM13 pUC19 ( mod DNA -15 to Accl in modA ) This study pAM15 pUC19 [-247 to +25, A(+4 to +11) modA This study operator/promoter]
pAM16 pUC 19 (-243 to +20 modA operator/ This study promoter
pAM17 pUC 19 (-17 to +6 modA operator/promoter) This study pAMI8 pUC 19 (-17 to -1, +5 to +15 modA operator/ This study promoter
pAM19 pUC 19 [-17 to +15, A(-6 to +9) modA This study operator/promoter
plsgG pGEM3ZF+ [(modE), Haemophilus Apicella influenzae]
plsgH pGEM3ZF+ [(modA*), Haemophilus Apicella influenzae]









19

A Q(modE '- 'lacZ) plasmid was constructed by first removing a 2.6-kb KpnIHindIII fragment from plasmid pFGHI 5 containing the mod'EF, galE' genes and ligating it into the KpnI-HindIII sites of plasmid pUC19 to yield plasmid pRM11. Since the 5' region of the modE gene was not present in plasmid pRMI11, a 0.6-kb KpnI fragment from plasmid pFGH15 containing this 5' region of the modE gene was inserted in frame to construct plasmid pRM12. The modEFgalE'DNA was then moved to plasmid vector, pBR322 by ligating a 3.2-kb EcoRI-HindIII fragment from plasmid pRM12 into the corresponding sites in plasmid pBR322, resulting in plasmid pRM13. Plasmid pRMI3 proved to be unstable; however, the removal of the tet gene in the vector DNA as a HindIII-AvaI fragment from plasmid pRM13 to give plasmid pRM25 resolved the instability problem. A 3.2-kb SmaI fragment encoding the lacZ gene from plasmid pZ 1918

(66) was then ligated into the AflII site (in modE gene) ofplasmid pRM13 which had previously been modified using the klenow fragment of DNA polymerase I (Klenow), thereby creating plasmid pRM26. Plasmid pRM26 produced (3-galactosidase activity from the modE promoter. Subsequent transfer of the QI(modE '- 'lacZ) into the chromosome of strain BW545 was accomplished by first recombining in vivo the bla and modE '- 'lacZ from plasmid pRM26 with 3RZ5 to produce XRM26. Strain BW545 was then transduced with phage ARM26.

In the course of this study, two different modE expression plasmids were

constructed. Plasmid pAG1, used for the production of ModE mutants, was created by ligating a Klenow-modified 870-bp EspI fragment containing the promoterless modE gene from plasmid pFGH 15 into the SmaI site of the ptac-based expression vector









20

pTrc99A. The second modE expression plasmid, pRM22, used for purification of ModE. was constructed by ligating an EcoRI-HindIII fragment which contained the modEF genes from plasmid pRM12 into the EcoRI-HindIII sites of plasmid pT7-7.

The modAB' containing plasmid, which was used in the in vitro coupled

transcription-translation experiments, resulted from the ligation of a 1.7-kb MscI fragment from plasmid pSE 1009 into the Smai site of plasmid pUC 19, yielding plasmid pRM1. Plasmid pRMI carries the entire modA gene but a truncated modB gene (modB' lacks the C-terminal ten amino acids).

A battery of modA-operator/promoter DNA-containing plasmids were constructed for the purpose of in vivo ModE titration experiments. One such plasmid, pAM4, was produced by ligating a KpnI-BstEII fragment from plasmid pSE 1009 (the BstEII end was modified using Klenow enzyme) which covered the -247 to +25 sequence of the modAoperator/promoter region into the KpnI-HinclI sites of plasmid pUC19. The plasmid pAM4 deletion derivative plasmids pAM5 and pAM6 were obtained by nuclease Bal-31 treatment of KpnI-digested plasmid pAM4 followed by religation. The extent of the deletions in plasmids pAM5 and pAM6 was determined by sequencing the DNA. Another derivative of plasmid pAM4, plasmid pAM7, was created by removal of an EcoNI-EcoRI fragment (-247 to -56 of modA -operator/promoter region) from plasmid pAM4 and subsequent religation after the ends had been filled in with Klenow. Plasmid pAM8 was constructed by removal of a 217-bp ApoI fragment (-247 to -31 of modA-operator/ promoter region) from plasmid pAM4, and plasmid pAM9 was achieved by ligating a KpnI-Apol fragment (-247 to -55 of modA-operator/promoter region) from plasmid









21

pAM4 into the KpnI-Apol sites of plasmid pUC19. Plasmid pAMIO was produced by digesting plasmid pAM8 with Apol followed by nuclease Bal-31 treatment and religation. Again, the extent of the deletion was determined through sequencing the DNA. Construction of plasmid pAM13 relied on polymerase chain reaction (PCR) generation of a 648-bp fragment using the following primers synthesized by National Biosciences, Inc. (NBI, Plymouth, MN): Primer #1, 5' AAGGATCCGTTATATTGTCGCCTAC 3', which contains an engineered BamHI site and is complementary to the -16 to +2 sequence of the modA-operator/promoter and Primer #2, 5' GCTGGCAACTGCGTC 3', which is complementary to sequence located in the 3' end of modA. The 648-bp fragment was digested with BamHI and AccI to give a 634-bp fragment which was ligated into the BamHI/AccI sites of plasmid pUC19, thereby yielding plasmid pAM13. Plasmid pAMI5 was obtained as a result of mung bean nuclease treatment and religation of Pspl406Idigested plasmid pAM4. Plasmid pAM16 resulted from mung bean nuclease treatment of the BstEII end of a KpnI-BstEII fragment from plasmid pSE 1009 which was then ligated into the KpnI-HincII sites of plasmid pUC19. Plasmids pAM17, pAM18, and pAM19 were constructed by ligating the oligomers which contained various portions of modAoperator/promoter sequence into vector plasmid pUC 19. These DNA oligomers were synthesized by NBI. Complementary oligomers were annealed, polynucleotide kinasetreated and digested with BamHI prior to ligating into the BamHI-SmaI sites of plasmid pUC19. The mod-insert DNA in plasmid pAM17 has the following 36-bp sequence 5' CGCAATCGTTATAT TGTCGCCTACATAAGATCCCG 3', while plasmid pAMI8's insert has a 45-bp sequence, 5' CGGAATCGTTATATTGTCGCCTTGGAACGTTAC











ATTGGATCCCG 3', and plasmid pAM19 contains the following 3 l-bp sequence 5 'CGGAATCGTTATATTGTTACATTGGATCCCG 3' insert.

Hydroxylamine mutagenesis of plasmid DNA. Hydroxylamine mutagenesis of plasmid DNA was carried out essentially as described by Davis et al. (15) with minor modifications. A reaction mixture containing 7.5 pg plasmid pAGI DNA, 40 pl of phosphate-EDTA buffer (0.5 M K-PO4, pH 6.5; 5mM EDTA), 80 Pl of freshly prepared hydroxylamine-hydrochloride solution. pH 6.0 (0.35 g of NHOH-HCI, 0.56 ml of 4M NaOH, 4.44 ml distilled HO0), and 50 tl deionized H,O was incubated at 370C for 18 h. Following the incubation period, hydroxylamine was removed by three successive 1:1 phenol extractions and a 1:1 chloroform extraction. At this point, the plasmid DNA was ethanol precipitated, and the dissolved DNA was used to transform competent cells of strain RK4353. Total plasmid DNA isolated from these transformants was then transformed into competent cells of strain SE 1811 for identification of mutant ModE containing plasmids. The resultant transformants were plated on lactose-MacConkey agar with and without ImM sodium molybdate. White colonies were picked from media lacking molybdate, and red colonies were picked from the media supplemented with molybdate. The levels of P-galactosidase activity were determined for the SE 1811 cells containing the mutant plasmids to confirm the presumptive phenotypes. The entire modE sequence was then determined for each selected mutant.

DNA Sequencing experiments. All DNA sequences generated in the course of this study were obtained using the Sanger dideoxy method and appropriate primers and plasmids which had been isolated by alkaline lysis procedure and purified by cesium









23

chloride density-gradient centrifugation (28, 62). For sequencing mutant modE genes, the primers used were synthesized by the DNA Synthesis Core Facility of the Interdisciplinary Center for Biotechnology Research at the University of Florida based on the sequence of the wild type modE DNA. The DNA sequences were analyzed by using computer software programs GCG (6, 13, 16) and Genepro (Riverside Scientific, Seattle, WA).

Biochemical Characterization of ModE

Expression and purification of ModE. For isolation of ModE protein, a lL LB culture of strain BL21,DE3(pRM22) was incubated at 37C with vigorous shaking (250 rpm) until an optical density of 1.0 was reached. Isopropyl-j-D-thiogalacto-pyranoside (IPTG) was then added to a final concentration of 0.5 mM to induce phage T7-RNA polymerase and high level expression of modE. After incubation for an additional two h., the cells were harvested by centrifugation at 8,300 x g for 10 min. and were resuspended in 8 ml of extraction buffer [50 mM Tris-HC1, pH 7.5; 0.5 mM EDTA; 0.5 % glycerol; ImM dithiothreitol (DTT)]. This cell suspension was passed twice through a French pressure cell (20,000 lb/in2). The broken cell suspension was centrifuged at 100,000 x g for 60 min. to remove cellular debris. The resultant supernatant was loaded onto a 10 ml Q Sepharose fast-flow column (Pharmacia) that had been equilibrated with 50 mM TrisHCl, pH 8.0. Progress of the protein sample through the column was monitored using a chart recorder connected to a UV light (280 nm) absorbance detector. After application of the protein sample to the Q sepharose column, the column was washed with 50 mM TrisHC1, pH 8.0 until a baseline on the UV absorbance trace had been established. The column was subsequently washed with Tris-HC1, pH 8.0 buffer containing 0.1 M NaCl. The









24

column was washed with this buffer until protein no longer eluted from the column as indicated by a return to the baseline on the UV absorbance trace. The column was then washed successively with Tris-HC1, pH 8.0 containing 0.2 M, 0.3 M, 0.4 M. 0.5 M. and

1.0 M NaCI. Fractions containing the ModE protein eluted with 0.2 M NaC1. The buffer of the ModE-containing protein solution was exchanged to 10 mM phosphate buffer, pH 7, by using an EconoPac-10 DG desalting column (BioRad Laboratories, Hercules, CA) prior to loading the protein on to a 5-ml heparin column (HiTrap; Pharmacia) that had been equilibrated with 10 mM phosphate buffer, pH 7. Bound proteins were stripped from the heparin column by successive washes with 10 mM phosphate buffer, pH 7.0 containing 0.1 M, 0.2 M, 0.3 M, 0.4 M, 0.5 M, and 1.0 M NaCI. The elution procedure proceeded as indicated above for the Q sepharose chromatography. ModE-containing protein fractions eluted at a salt concentration of 0.2 M. At this point, the ModEcontaining protein sample was concentrated using a Centricell-20 spin cartridge (Polysciences, Warrington, PA). The buffer of the ModE solution was exchanged with 25 mM histidine-HCI buffer, pH 6.2 and the protein was applied to a 10-ml polybuffer exchanger 94 column (Pharmacia) equilibrated with 25 mM histidine-HCI buffer, pH 6.2. The column was washed with polybuffer 74, pH 4.0, and the ModE protein was eluted when the pH of the effluent reached approximately 4.5. The ModE protein was stored in 10 mM phosphate buffer-10% glycerol, pH 7.0 at a final protein concentration of 0.42 mg/ml. The entire purification process was conducted at 40C. Since the intrinsic properties of the ModE protein did not lend itself to conventional detection by UV light at 280 nm or detection by the Bradford assay (9), ModE protein was monitored by the









25

bicinchoninic acid protein determination assay (70) as well as by electrophoresis in (SDS)polyacrylamide gels, N-terminal amino acid sequencing (performed at the Interdisciplinary Center for Biotechnology Research Protein Chemistry Core Laboratory at the University of Florida) was used to confirm the identity of the purified protein as ModE. Matrixassisted-desorption time-of-flight spectrometric determination of the molecular weight of purified ModE protein was kindly performed by Dr. Preston.

In vitro coupled transcription-translation experiment. In vitro transcriptiontranslation experiments using S30 extract and L-[35S]methionine to monitor the expression of modA and modB' from plasmid pRMI in the presence and absence of ModE and 1 mM sodium molybdate were conducted as suggested by Promega. Three gjg of plasmid pRM1, and when appropriate, 40 pmol of ModE and 1 mM sodium molybdate were included in the S30 reaction mixtures. The 35S-labeled proteins were separated by electrophoresis through a SDS-15% polyacrylamide gel. Following completion of the electrophoresis, the resulting gel was transferred to Whatman 3mm paper and was dried under vacuum. X-ray film (Fuji RX) was then applied to the gel for autoradiographic visualization of the labeled proteins. Quantification of labeled proteins was accomplished by using a phosphorimager (Molecular Dynamics, Sunnyvale, CA).

Mobility shift experiments. DNA-mobility shift experiments were performed as

described by Fried and Crothers (21) with modification. These experiments used a binding reaction buffer consisting of 10 mM Tris-HC1, pH 7.9, 10 mM MgC,, 50 mM NaCl, I mM DTT, and 5% glycerol. The reaction mixture contained the binding reaction buffer, DNA, and protein in a final volume of 10 jl. The reaction mixtures were incubated at









26

37oC for 30 min. prior to loading onto 5%-polyacrylamide gels in Tris-borate-EDTA buffer that had been prerun at 100 V at room temperature for approximately 60 min (51). During the electrophoresis, tank buffer was continuously circulated using a Minipuls II pump (Gilson). The migration patterns of the binding reactions in the resulting gels were visualized by autoradiography, and various bands in the gels were also quantified using a phosphorimager.

Experiments designed to ascertain the apparent dissociation constant (KD) for the binding of ModE to its target DNA featured a 43-bp oligonucleotide synthesized by NBI which spans the -18 to +25 region of the modA-operator/promoter sequence. For the DNA-mobility shift experiments, the 43-bp oligonucleotide was end labeled using kinase to phosphorylate the 5' ends with y-[2p] from y-[32P]-deoxyadenosine triphosphate (dATP). Unincorporated label was removed by passing the labeled oligonucleotide through a G-25 microspin column (Pharmacia). The reaction mixtures prepared for these experiments contained 0.1 pmol of the 43-bp labeled oligonucleotide as well as 0, 1, 5, 10, 25, 50, 100, 250, 500, and 1000-fold excess of ModE. When appropriate, sodium molybdate at a final concentration of 10 jM or 1 mM was also included in the reaction mixtures. Experiments which had sodium molybdate in the reaction mixtures also contained sodium molybdate in the gels and electrophoresis buffers at the same concentration as that present in the binding reactions. In some experiments, 10 PM sodium molybdate was replaced with sodium tungstate, sodium sulfate or sodium orthovanadate at the same concentration as in the reaction mixtures, gels and tank buffers.









27

For the DNA-mobility shift experiments designed to determine the concentration of molybdate required for half-maximal binding of ModE, the binding reactions included different concentrations of sodium molybdate ranging from 1 pM to 1 mM as well as 0.1 pmol 43-bp oligomer and 50-fold excess of ModE. Two polyacrylamide gels were run for this experiment, one with 1 pM sodium molybdate present in the gel and tank buffer and a second gel having 1 pM molybdate present. Binding reactions with 1 pM to 1 jPM molybdate were electrophoresed through the 1 pM molybdate gel, and binding reactions having 1 pM to 1 mM molybdate were subjected to electrphoresis through the gel containing 1 M molybdate.

In this study, the measure of the affinity of ModE for its target DNA is reported as an apparent dissociation constant value (KD value). The dissociation constant is defined as the concentration of protein required for 50% binding to the target DNA. Apparent KD values describing binding of ModE to DNA were determined by plotting the concentration of ModE present in the binding reaction versus the resultant percent of shifted DNA.

In vitro DNase I footprinting experiments. The DNase I protection experiments were carried out essentially as described previously (42, 56, 65). For these experiments, a 446-bp FspI-HindIII fragment from plasmid pAM4 which carries modA-operator/ promoter DNA spanning from -247 to +25 was labeled using a-32P-dATP, dCTP, dGTP, and dTTP (NEN, Boston, MA) using Klenow to fill in the HindIII end. Binding reactions having a total volume of 20 pl were prepared which contained 0.1 pmol of the labeled DNA, 0, 1, 10, 100, or 1000-fold excess ModE, 1 mM sodium molybdate, and the binding reaction buffer used in the DNA-mobility shift experiments. The binding reactions were









28

incubated for 30 min. at 37oC prior to adding 1.5 ng of DNase I (Sigma). The DNase I treatment continued for two minutes before adding 15 pl stop solution (34 mM EDTA. 6.5 M ammonium acetate). Ethanol (110 pl) was then added to precipitate the cleaved DNA fragments. The precipitated DNA was resuspended in 6 pl deionized HO and 4 pl Sequenase stop solution (USB), and the samples were electrophoresed through an 8% polyacrylamide-TBE-denaturing gel. After drying the gel, autoradiography was used to determine the areas of DNA protected by ModE from DNase I cleavage.

Determination of the molecular weight of the ModE-43-bp-DNA complex using native gel electrophoresis. In an effort to determine the stoichiometry of the ModE-DNA association, binding reactions containing 0.1 pmol of a 43-bp oligomer (-18 to +25 region) and 250-fold excess ModE were prepared as described for the DNA-mobility shift experiments. The binding reaction samples were then electrophoresed through a 5.0%, 6.0%, 7.0% or 8.0%-polyacrylamide-TBE-nondenaturing gel as described under DNAmobility shift experiments. The differences in the Rf values [migration of the sample (cm)/ migration of the dye front (cm)] for the ModE-DNA complexes after electrophoresis through the different concentration polyacrylamide gels was then compared to values obtained for protein standards. For the determination of the Rf values of the protein standards, 15 pig of lactalbumin (14,200 Da), 20 [lg of carbonic anhydrase (29,000), 20 Plg of chicken egg albumin (45,000 Da), 15 jig of bovine serum albumin (66,000 Da monomer, 132,000 Da dimer) and 6 [ig of urease (240,000 Da dimer and 480,000 Da tetramer) were electrophoresed along with the ModE-DNA complex in various polyacrylamide-TBE-nondenaturing gels. A plot of the [log(Rf x 100)] versus % gel









29

concentration for the molecular weight standards was then generated, and the slopes derived for each of the standards from this plot were then used to produce a Ferguson plot

(19), which relates the negative slope to molecular weight of the protein. The molecular weight of the ModE-43-bp-oligomer complex was determined by extrapolation from the Ferguson plot.














RESULTS AND DISCUSSION


Evidence for a Regulator of the modABCD Operon

Previous work demonstrated that a mutation resulting in loss of function in any one of the proteins which form the molybdate-specific transporter, ModA, ModB, or ModC will cause a loss of activity for the cell's complement of molybdoenzymes (58, 61). This phenotype, which can be suppressed by molybdate supplementation of the growth medium, suggests that the cells containing these mutations are unable to transport molybdate under conditions of low molybdate concentration in the growth medium (22, 29, 48, 61, 67, 71). Restoration of the activities of molybdoenzymes with the addition of

1 mM molybdate to the growth medium also indicates that a sufficient intracellular concentration of molybdate is generated by alternate routes of molybdate transport other than the molybdate-specific transport system in E. coli (41, 61). Although a molybdatespecific transport system has been defined in various organisms, the mechanism for the regulation of the genes coding for the components of the transport machinery is not known.

A general approach for initial investigation of the regulation of genes of interest in Escherichia coli is to construct lacZ fusions in the appropriate genes and then monitor the resulting P-galactosidase activities under a variety of experimental conditions and in a number of genetic backgrounds. In this case, the genes of interest comprise the 30









31

modABCD operon, and E. coli strain SE2069, which contains a chromosomal D(modA ''lacZ) fusion was used. Since strain SE2069 has its modA gene interrupted by the lacZ gene fusion, it is functionally a mod mutant strain. Using strain SE2069, it was found that a high level of modA '-'lacZ expression, monitored as -galactosidase activity, was only achieved in cells grown in LBG, a medium which contains trace amounts of contaminating molybdate (Table 3). Therefore, the fact that high level expression of modA- 'lacZ is only seen in the absence of added molybdate intimates that the modABCD operon is subject to repression only when sufficient intracellular levels ofmolybdate are present. If this is indeed the case, a mod' E. coli (strain BW545) should transcribe the modABCD operon at very low levels. In order to test this possibility, the modA '- 'lacZ fusion from strain SE2069 was transferred to phage (XSE 1) and incorporated into the ;1 att site of the chromosome.

Low levels of expression of modA '- 'lacZ in the wild type mod" strain, strain

BW545(XSE 1), regardless of supplementation of molybdate to the growth medium (Table 3), implies that the functional molybdate-specific transporter brings in sufficient levels of molybdate from the LBG medium to repress the modABCD operon. In the presence of a mutation in any of the first three mod genes (undefined point mutation in a given mod gene), derepressed levels of modA '- 'lacZ from XSE 1 result only in the absence of molybdate supplementation of the growth medium.

Further information concerning the nature of the molybdate-dependent repression of the modABCD operon was afforded by complementation studies of a spontaneous









32

Table 3. Effect of mod gene mutation in mod operon transcription


Strain mod P-galactosidase Activity Genotype -Mo +Mo


SE2069 4(modA '- 'lacZ) 2,400 250 BW545(XSE 1) mod' 20 <10 SE1592(.SE1) modA 2,330 165 SE1595(XSE1) modC 1,980 150 SE1602(XSEI) modB 1,860 60 aCells were grown in LBG medium supplemented with 1 mM sodium molybdate when specified. 3-galactosidase activities are expressed as nmoles-minute'-milligram of cell protein-'.









33

mutant of strain of SE2069, designated as strain SE1811. The level of modA '- 'lacZ expression by strain SE 1811 remained high even in the presence of 1 mM molybdate, suggesting that this mutation inactivated the gene responsible for the molybdate-dependent repression of the modABCD operon (Table 4).

By phage P -mediated transduction, the mutation which derepressed modA

expression in strain SE1811 was mapped close to modA. The identification of the gene altered in strain SE 1811 was accomplished through the introduction of plasmids containing different regions of the mod DNA into strain SE1811. As presented in Table 4, the introduction of plasmid pSE 1004 which harbors modABCD DNA as well as an open reading frame diverging from this operon into strain SE 1811 resulted in both the restoration of molybdate transport as indicated by the return of FHL activity in the absence of added molybdate in strain SE1811 as well as the restoration of molybdatedependent repression of the modABCD operon in this strain. Introduction of plasmid pSE 1009 which has modABC DNA only restored the molybdate-specific transport in strain SE1811 as seen by the presence of FHL activity but did not restore the molybdatedependent repression of modA '- 'lacZ expression in this strain. As expected both plasmids complemented the modA mutation in parent strain SE2069.

It should be noted that there is a disparity in the [-galactosidase activity values produced by strain SE2069 (Tables 3 and 4) which is most likely the result of allelic variations in strain SE2069. The higher value from Table 3 was obtained approximately 2 years prior to the value shown in Table 4, while the value from Table 4 is characteristic of values which are now routinely obtained for strain SE2069.









34

Table 4. Complementation of the mutation in strain SE1811 for molybdate-dependent
repression of the modABCD operon


Strain Relevant Genotype P-galactosidase Activitya FHLb
-Mo +Mo


SE1811 'D(modA-lacZ), modE 1,500 1,500 SE 1811(pSE1004) Q(modA-lacZ), modE, (modABCDE) 290 260 + SE1811 (pSE1009) #D(modA-lacZ), modE, (modABC)' 1,500 1,500 + SE181 1(pAG) D(modA-lacZ), modE, (modE)' 600 100 SE2069 Q(modA-lacZ) 1,300 200 SE2069(pSE 1004) 'D(modA-lacZ), (modABCDE)' 300 300 + SE2069(pSE 1009) #(modA-lacZ), (modABC)' 300 300 +


aCells were grown in LBG medium. Sodium molybdate concentration was 1 mM. 3galactosidase activities are expressed as nanomoles, minutel- milligram of cell proteinb-, absent; +, present.








35

The DNA present in plasmid pSE1004 but not in plasmid pSE 1009 was subcloned into plasmid vector pTrc99A to yield plasmid pAGI. The introduction of this plasmid into strain SE1811 resulted in full restoration of the molybdate-dependent repressor activity as shown by the low level of modA '-'lacZ expression for strain SE 1811 (pAG1) with the addition of 1 mM sodium molybdate to the growth medium. The lower level of modA ''lacZ expression for SE 1811 (pAG 1) in the absence of added molybdate, (600 units of 13galactosidase activity, as opposed to 1,500 units for strain SE1811), is most likely due to the presence of multiple copies of the plasmid in each cell. Presumably, when there is this higher dosage of repressor in strain SE 1811 cells as a result of the presence of plasmid pAG 1, any trace amount of molybdate present in the cell would serve to enhance the partial repression of the modA '-'lacZ seen under these conditions.

Based on the results presented in Tables 3 and 4, it can be concluded that there is a molybdate-dependent repressor which acts to limit expression of the modABCD operon when a sufficient intracellular concentration ofmolybdate is sensed. The gene that presumably codes for this repressor is designated modE, and the modE gene is located adjacent to the modABCD operon but is transcribed in the opposite direction compared to the modABCD operon.

Analysis of modEF DNA

modE gene. The region of DNA which complemented the mutation in strain SE 1811 was sequenced, and a two-gene operon starting 444 bp upstream of the modA translation start site was identified (Fig. 2; modified from 60). These two open reading














36


-nocA

CGAGCCATTGGTAACCCCTTAATGTAACGTTATGTAGGCGACAATATAACGACTAACTCGGAATTTTCCCAGCAGTTATTGCTAACCTTT 90
RAM
AGGTAGATAAGAAAAAATCGGCA AGAAAAGCAGGAAGTTGAGAAAAAGAAAATGCCCOACTAAGCGGGCATTCAGGGAATCAAT

ATTTTGTCCGGGCTGGTCTTTTTTACCAACACCAGAAAAGATGTTGAATACTTCACCAAGACCGTAAATCAGACCCAGGATGATGGCCAT 270
CACGACAGGTACCATGATTACGGCGAATACCAGACTTTTTAATAACTCTAACATGGTCCAAC ATATAGTCATGAGAC TT AAC 360 CGCTAAGCACAGAAAGCACTCCCCTT TTTTCCTCACAATAAGACTTTTGCCAACATTG 450
modE (lacZ in pRM14 and pRM26)
TTATGCAGGCCGAAATCCTTCTCACCCTTAAGCTCCAACAAAAATTATTCGCCGACCCGCOCCGCATTTCGCTACTAAAACACATTGCGC 540
MQA E I LLTLKLQQK LFA DPRR SLL K H A L
TTTCCGGTTCCATTAGCCAGGGAGCGAAAGATGCCOGTATTAGCTATAAAAGCGCCTOGGATGCCATTAACGAGATGAATCAGTTAAGTG 60
SGS ISQGA KDA G I SYKSAW DA I NEMNQ LS E
AGCATATTCTGGTCGAGCGCGCAACAGGCGGTAAAGGTGGCGGCGGGCAGTACTGACCCGCTATGGTCAGCGACTGATTCAGCTCTATG 720
H I L V E RATGG KGGGG A V L T R Y Q R L I Q LY
ACTTACTGGCGCAAATCCAGCAAAAAGCCTTTGATGTGTTAAGTGACGATGACGCCCTGCCGCTGAACAGCCTGCTGGCCGCGATCTCAC 910
L L A Q I Q 0 K A F D V L S D D D A L P L N S L L A A I S R
GTTTTTCACTGCAAACCAGTGGTTCGGTACCATCACCGCCCGCGATCATGGACGTTCAACAGCATGTTGATGTCT 900
F S L Q T S A R N Q W F G T I T A R D H D D V Q Q H V D V L
TACTGGCTGACGGAAAAACACGCCTGAAAGTCGCAATTACCGCACAACGCGGCCGCGTCTGGGGCTGGATGAAGGCAAAGAAGTGTTGA .90
LA DG KTRLKVA TAQS A R G LDEG E KEV L I
TATTGCTAAAAGCGCCGTGGGTAGGTATTACTCAGGACGAGGCGGTCGCGCAAAACGCTGACAACCAATTACCGGGTATTATTAGTCATA 1080
LLK A P W V I TQ DEAVAQNA DNQ L P I I SH
TTGAGCGCGGCGCAGAGCAGTGCSAAGTATTAATGGCGCTACCCGACGGGCAAACACTGTGCGCCACAGTGCCGGTAAATGAACGACTT 1
ER G A EQ CE V L M A L P D G Q T L C A TV P V N E A T S
CTCTTCAGCAAGGACAGAATGTCACGGCCTACTTTAATGCCGACAGCGTGATTATCGCCACGCTGTGCTAAGCGTGTTGACAATTTGTTA 1260
LQQGQN V TAY FNA DSV I ATLC
moaF
TGAAACACGTATCCCTGTCAGTAATCGCTGCACAAAGTGGGTATAAAATGTCATCGTTGCAAATTTTGCAAGGCACGTTTCGTCTTAGC 1350 M S S L Q I L Q G T F R L S
GACACAAAAACGCTGCAATTGCCTCAGCTAACGTTAAACSCGGGTGATAGTTGGGCGTTTGTCGGTTCGAATGGAAGCGGGAAATCGGCC 1440
DT K T L Q L P Q L T L N A G D S WA F V G S N G S GK S A
CTGGCCCGCGCGCTGGGGGGGC AAAGCCAGTTTTCCCACATCACTCGTCTCTCCTTCGAGCAA 1530
L A R A L A G EL PLL K G ER Q S Q F S H I T R L S FE Q
TTGCAAAAACTCGTCAGCGACGAATGGCAGCGGAATAACACCGATATGCTCGGCCCTGGCGAAGATGACACCGGACGCACTACGGCTGAG 1620
L Q K LV S D E W Q R N N TD M L G P G E D D T G R T TA E
ATCATTCAGGATGAAGTAAAGGATGCACCGCGTTGCATGCAACTGGCGCAGCAGTTCGGTATTACCCGCCTCCTCGACCGACGCTTTAAA 1710 I I Q D E V K D A PR C MQ L A Q F G I TA L L D R R F K
TACCTTTCCACTGGCGAGACGCGAAAAACCCTGCTGTGTCAGGCGCTGATGTCGGAGCCTGACTTGTTGATTCTTGATGAGCCGTTCGAT 1800
YLSTGETRKT LLCQA LMS E PDLLI LDE P FD
GGCC TGATTTGCCTCACGT TGGCTGAGCGACTCGCCTCGTTACA TCCGGTATTACTCTGGTACTGGTGCTCAATCGC 1890
G L D V A S R Q Q L A E R L A S L H Q S G I T L V L V L N R
TTCGATGAGATCCCGGAGTTTGTCCAGTTTGCTGGCGTGCTGGCGGATTGCACGTTAGCGGAAACTGGCGCTAAAGAGGAACTGCTCCAA 1980
F D E I P E F V Q F A G V L A D C T L A E T G A K E E L L 0
CAAGCACTCGTCGCGCAACTGGCGCATAGTGAACAGCTTGAAGGTGTGCAACTGCCGGAGCCGGATGAACCCAACCTGTA 2070 QA L V A Q LA H S E Q L E G V Q L PE P D E PS A R HA L CCCGCCAACGAACCGCGCATTGTGCTGAACAATGGCGTGGTTTTTATAAATATTCTTATAACGATCGCCCCATTCTTAATAACCTTAGCAGGTGAAT 2160 PANE PR IVLNNGVV SYN DR PI LNN LSWOV N
CCAGGCGAACACTGGCAAATTGTCGGGCCAAATGGTGCAGGAAAATCGACGTTATTAAGCCTGGTTACTGGCGATCATCCGCAAGGTTAC 2250 P G E H W Q I V GPNGA GKS T L L S L VT G D H P Q G Y
AGCAACGATTTGACGCTTTTCGGACGACGTCGCGGCAGCGGCGAAACCATCTGGGATATCAAAAAGCATATCGGTTACGTCAGCAGTAGT 2340 S N D L T L F R R R G S G E T I W D K K H I G Y V S SS
TTGCATCTGGATTACCGGGTCAGCACTACCGTGCGTAATGTGATTCTTTCTGGCTATTTTGATTCGATTGGCATTTATCAGGCCGTTTCG 2430 L H L D Y R VS TT V R N V I L S G Y F D S I G I Y Q A V S
GATCGCCAGCAAAAACTGGTGCAGCAGTGGCTGGATATTCTCGGCATTGATAAACGCACGGCTGACGCTCCGTTCCATAGTCTTTCCTGG 2520 D R Q Q K L V Q Q W L D I L G I D KR TA D A P F H S L S W
GGACAGCAGCGTCTGGCGCTGATTGTCCGCGCACTGGTGAAACATCCGACGTTGCTTATTCTCGATGAACCACTACAGGGGCTTGATCCG 2610 G Q Q R L A L I V R A L V K H P T L L I L D E P L Q G L D P CTGAATCCAGCTTATCCCCGTTTTGTTGATGTGCTGATTAGCGAAGGTGAAACGCAATTGTT TCGCACCACGCTGAAGAT 2700 L N R Q L I R R FV DV LI S E G E T LL F V S H H A E D
GCGCCTGCCTGTATTACCCATCGTCTTGAGTTCGTGCCGGACGGTGGACTCTATCGCTATGTGCTGACAAAAATATATTGAGTCGGTAGT 2790 A PAC IT H R L E FV P D G G L YR Y V L T K I Y GCTGACCTTGCCGGAGGCGGCCTTAGCACCCTCTCCGGCCAACGGTTCGACGCATGCAGGCATGAAACCGCGTCTTTTTTCAGATAAAAA 2880
GCGCAATCATTCATAAACCCTCTGTTTTATAATCACTTAATCGCGCATAAAAAACGGCTAAATTCTTGTGTAAACGATTCCACTAATTTA 2970
galE
TTCCATGTCACACTTTTCGCATCTTTGTTATGCTATGGTTATTTCATACCATAAGCCTAATG 3033


Fig. 2. DNA sequence of the modEFoperon. Amino acid sequences are listed below the DNA sequence.
Presumptive -10 and -35 sequences are underlined and ribosomal binding sites are double underlined.
Asterisks denote translation terminations. The location of the lacZ insertion in the modE gene in plasmid
pRM26 is indicated by the down arrowhead. The stem-loop structure is shown by the two opposing arrows. ATP-GTP binding motifs in ModF are in boldface type.









37




frames, designated modE and modF, are transcribed in the opposite direction as compared to modA and encode a 262-amino-acid protein (28,200 Da) and a 490-amino-acid protein (53,900 Da), respectively. An eight-bp-stem six-bp-loop can be identified in the DNA between the operator/promoter regions of modA and modE and is indicated by the two opposing arrows in Fig. 2. This DNA is located between gal and modABCD operons in the E. coli chromosome, and the galETK operon is transcribed in the same direction as the modEF operon.

A search for proteins that bear similarity to ModE was conducted resulting in the detection of four homologs (Fig. 3) The homologs were identified as ModA from A. vinelandii (35% identity, 45% similarity) (43) (renamed ModE [50]), ModE from H. influenzae Rd (48 % identity, 58% similarity) (20), MopA from R. Capsulatus (34% identity, 46% similarity) (76), and MoplI from C. pasteurianum (29% identity, 46% similarity) (32). A possible helix-turn-helix motif which is a known prokaryotic DNAbinding protein structure motif (52) has been identified in the N-terminal part of the E. coli ModE as indicated by the underlined sequence in Fig. 3.

Since H. influenzae ModE is almost 60% similar to E. coli ModE and both are

facultative anaerobes, it was thought that the H. influenzae ModE homolog may be able to substitute for the E. coli ModE protein in the ModE deficient E. coli strain SE 1811. This possibility was tested by monitoring the modA '- 'lacZ expression in strain SE 1811 which had been transformed with a plasmid containing the H. influenzae modE gene












38


AvModE MTATRFLARMSLDTDVG--TALSDTRIRLLEAIEREGSINRAAKVVPLSYKAAWDAIDTMNN 60 EcModE MQAEILLTLKLQQK--LFADPRRISLLKHIALSGSISQGAKDAGISYKSAWDAINEMNQ 57 HiModE MKNTEILLTIKLQQA--LFIDPKRVRLLKEIQQCGSINQAAKNAKVSYKSAWDHLEAMNK 58 RcMopA MNEQPLIAALSLQRAGAPRVGGDRIRLLEAIARHGTIAGAAREVGLSYKTAWDAVGTLNN 60


AvModE LAPEPLVVRVAGGRQGGGTQLTDYGRRIVAMYRALEIEYQSALDRLSERLNEVTGGDIQA 120 EcModE LSEHILVERATGGKGGGGAVLTRYGQRLIQLYDLLAQIQQKAFDVLSDDDALPLNSLLAA 117 HiModE ISPRPLLERNTGGKNGGGTALTTYAERLLQLYDLLERTQEHAFHILQDE-SVPLDSLLTA 117 RcMopA LFEQPLVEAAPGGRTGGNARVTEAGQALIAGFGLLEGALTKALGVLEGGVSAPEKALNTL 120


AvModE FQRLMHSMSMKTSARNQFAGIVTGLRVGGVDYEVRIRL-DAENEIAAVITKASAENLELA 179 EcModE ISRF---- SLQTSARNQWFGTITARDHDDVQQHVDVLLADGKTRLKVAITAQSGARLGLD 173 HiModE TARF----SLQSSARNQFFGRVAQQRIIDSRCVVDVNVQGLPTPLQVSITTKSSARLKLI 173 RcMopA WSL----- TMRTSNRNTLRCTVTRVTLGAVNAEVELALTDGHS-LTAVITERSATEMGLA 174 CpMopII MSISARNQLKGKVVGLKKGVVTAEVVLEIAGGN-KITSIISLDSVEELGVK 50


AvModE IGKEVFALVKSSSVMLTTEPSLKL-TARNQLWGEVIDIHEGPVNNEVTLALPSGRSVTCV 238 EcModE EGKEVLILLKAPWVGITQDEAVAQ-NADNQLPGIISHIERGAEQCEVLMALPDGQTLCAT 232 HiModE TEKEVMLMFKAPWVKISEQPLA---NQPNQFPVNIKSLN---EEEAILQFAESNIEFCAT 219 RcMopA PGVEVFALIKASFVMLAAGGDPGRISACNRLTGIVAARTDGPVNTEIILDLGNCKSITAV 234 CpMopII EGAELTAVVKSTDVMILA 68 AvModE VTADSCKALGLAPGVAACAFFKSSSVILAVYG 270 EcModE VPVN--EATSLQQGQNVTAYFNADSVIIATLC 262 HiModE V-H---QPNQWQIEQQVWIHIDQEQIILATLG 255 RcMopA ITHTSADALGLAPGVPATALFKASHVILAMP 265






Fig. 3. Amino acid sequences of the homologs of E. coli ModE protein. AvModE, A. vinelandii ModE (43); EcModE, E. coli ModE; HiModE, H. influenzae Rd (20); RcMopA, R. capsulatus ModE (76); CpMopII, C. pasteurianum MoplI (31). Double dots represent identity, and single dots represent conservative substitutions. Shadow print indicates the conserved SARNQ sequence. All identity and similarity designations are in relation to E. coli ModE. A presumptive helix-turn-helix DNA binding region in E. coli ModE is underlined.









39

(plasmid plsgG). A low level of modA '-'lacZ expression was seen for strain SE 1811 (plsgG) that was cultured in LBG supplemented with 1 mM sodium molybdate (Table 5) verifying that the H influenzae ModE protein can functionally substitute for E. coli ModE.

modF gene. Expression of the modF gene from a T7-based expression vector,

plasmid pRM23, yielded a 54 kDa protein when visualized after SDS-PAGE. This size is in agreement with the expected size based on the DNA-sequence derived amino acid sequence. Database searches revealed no homologs of E. coli ModF protein, but further examination of the amino acid sequence identified two ATP-GTP binding site motifs (Fig. 2).

As of yet, the function of modF has not been determined, since deletions of the modF gene did not result in a detectable phenotype for FHL production or modA gene expression or activity under the conditions used in this study. A study by Dorrel et al.

(17) has described and presented the sequence of a gene involved in photoreactivation (phrA). This DNA sequence is situated within the modF gene presented in Fig. 2. Based on their studies, Dorrel et al. proposed that the 38 kDa PhrA protein is necessary for photorepair. However, the findings that a 54 kDa protein and not a 38 kDa product was observed when modF was expressed in a T7 RNA polymerase-driven expression system and that the expression of <)(modF'-'lacZ) responds to regulation by ModE (data presented later) suggest that the DNA in question codes for modF.









40





Table 5. Complemetation of the modE mutation in strain SE1811 with H. influenzae
modE.


Strain Relevant genotype P-galactosidase Activity'
-Mo +Mo


SE1811 4(modA '-'lacZ), modE 1,300 1,300 SE 1811(plsgG) 4(modA '- 'lacZ), modE, 1,500 < 10 (H.i.- modE')


aCells were grown in LBG medium. Sodium molybdate concentration was 1 mM. 3galactosidase activities are expressed as nanomoles, minute-'- milligram of cell protein'.








41

Regulation of modA by ModE

In order for the interaction of ModE with its target operon modABCD to be

investigated, the ModE protein first had to be purified. ModE protein was overexpressed and purified as indicated in "Materials and Methods" section. The protein yield after each purification step is shown in Table 6, while the purity of ModE after each purification procedure is presented in Figure 4. A final yield of purified ModE protein of 5% of total protein in the extract was obtained, and this protein was judged to be pure by native polyacrlyamide electrophoresis (not shown), SDS-PAGE (Fig. 4), and N-terminal amino acid sequencing of the protein. The first twelve amino acids are MQAEILLTLKLQ and these correspond to the predicted amino acid sequence of the protein (Fig. 3). The molecular mass of the purified protein was determined to be 28,271 Da using matrixassisted laser desorption ionization time-of-flight mass spectrophotometry. This mass value is in agreement with the predicted value of 28,200 Da based on the DNA sequencederived amino acid sequence as well as a value of 29,000 Da obtained after SDS-PAGE.

Coupled in vitro transcription-translation experiments. Experiments in which the P-galactosidase actvity of modA '-'lacZ was monitored in two isogenic strains, one having a functional modE gene (strain SE2069) and the other carrying an inactivating mutation in modE (strain SE 1811) suggested that ModE represses transcription of modA '-'lacZ and presumably the modABCD operon in a molybdate-dependent fashion (Table 4). A direct approach to investigating the interaction of ModE with its target DNA is to determine the









42

Table 6. Purification profile of ModE

Fraction Total Purification protein (fold)
(mg)


Extract 238 1.0 Q Sepharose fast flow 57 4.2 HiTrap Heparin 16.5 14.4 Chromatofocusing 12 19.8









43















-I-w
-5O



1 2 3 4 5










Fig. 4. SDS-PAGE analysis of proteins from different stages of ModE purification. Lane 1, molecular weight markers (from top: phosphorylase b, 97,400; bovine serum albumin, 66,200; ovalbumin, 45,000; carbonic anhydrase, 31,000; trypsin inhibitor, 21,500; lysozyme, 14,400); lane 2, extract; lane 3, proteins after Q Sepharose fast-flow purification; lane 4, proteins after HiTrap Heparin purification; lane 5, proteins after chromatofocusing. A 25-pg amount of protein was loaded into each of lanes 2-4. A 10jIg amount of protein was loaded into lane 5.









44

level of expression of target genes (modAB ') with and without the effector (ModE) in an in vitro coupled transcription-translation experiment with S30 extract.

For these coupled transcription-translation experiments, plasmid pRMI DNA, containing modAB' was incubated with E. coli S30 extract with and without 1 mM sodium molybdate and 40 pmol of ModE. In these experiments, the molybdate concentration was maintained at 1 mM since the affinity between the protein and molybdate is not known. Also, it is not known whether the Mo species bound to ModE is molybdate or some other derivative of molybdate. It should be noted that the S30 extract provides all the constituents necessary for production of proteins from a DNA template and that addition of 3S-methionine allows for labeling of proteins produced in the reaction mixture. The resultant autoradiogram (Fig. 5) and derivative Phosphorimager data indicate that the presence of ModE in the reaction mixture reduced production of the 26kDa ModA protein and the 24-kDa ModB' protein by 48 and 65%, respectively (lanes 1 and 2). In Fig. 5, two bands are labeled as ModA. ModA, which has been shown to function as a molybdate-binding protein (59), has a putative 24 amino acid leader signal peptide thought to be responsible for its localization in the cell's periplasmic space (60). Given that ModA has a signal peptide, the upper band would be the unprocessed modA while the lower band would be ModA without its leader peptide. With the addition of sodium molybdate, repression of modAB' expression was even greater with over 90% decrease in production of ModA and ModB' (Fig. 5, lanes 3 and 4). The decrease in production of ModA and ModB' even in the absence of added sodium molybdate is most likely due to the presence of contaminating molybdate in the S30 extract. It should also









45












Bla -- ~

ModA { ...
ModB




















Fig. 5. SDS-PAGE analysis of proteins produced in an in vitro transcription-translation experiment in which plasmid pRM1 served as the template DNA. Sodium molybdate and ModE concentrations were 1 mM and 40 pmol per reaction mixture, respectively. Lane 1, without ModE; lane 2, with ModE; lane 3, without ModE, with sodium molybdate; lane 4, with ModE and sodium molybdate. Bla, p-lactamase. The upper ModA protein band represents the precursor protein bearing the signal peptide, while the lower ModA protein band represents the protein with the signal peptide removed. ModB, truncated ModB protein.









46

be pointed out that, although equal amounts of radioactivity were loaded in each lane, the samples in lanes 2 and 4 appear to contain a total of less labeled protein than is present in the other lanes. This disparity in the amounts of labeled protein is due to the presence in these samples of low molecular weight (partially degraded?) protein products which migrate faster than the dye front during electrophoresis. Addition of molybdate alone did not significantly reduce the amount of ModA and ModB' produced in the reaction mixture. Thus, the reduction in the production of ModA and ModB' proteins with ModE in the reaction mixtures confirms that ModE acts as a repressor of the modABCD operon and that this repression is enhanced by molybdate.

In an effort to further show that ModE was binding to the modA operator/

promoter DNA and thereby reducing expression of the operon, a 42-bp DNA fragment, believed to encompass the ModE target binding region, was added to the transcriptiontranslation reactions along with the ModE protein in order to titrate out ModE and relieve the repression of modAB '. The added DNA is a 42-bp DNA derived from the operator/promoter region of modA spanning -9 to +33. With the addition of 50-fold more 42-bp oligomer as compared to plasmid pRM1 in the reaction mixtures, production of ModA was restored to 66% of the levels seen without ModE, and inclusion of 100-fold excess of 42-bp oligonucleotide fully restored production of ModA and ModB' to the levels seen in the absence of ModE (data not shown). Clearly, even though the DNA used to compete for the binding of ModE carried only a part of the operator/promoter region of the modABCD operon, the 50-fold and 100-fold excess of this DNA compared to the









47

template DNA present in the reactions was sufficient to eliminate efficient binding of ModE to plasmid-borne target DNA.

Characterization of the Interaction of ModE with modA Operator/promoter DNA

Determination of ModE's target binding site sequences using in vivo titration

experiments. The rationale for this series of experiments aimed at identifying the region of DNA in the modA operator/promoter region required for successful binding of ModE protein is based on the idea that the presence of a large number of copies of DNA, in the form of plasmid DNA, containing the repressor's binding site(s) should result in the binding of a certain fraction of the repressor molecules produced from the chromosomal copy of modE. With the repressor molecules bound to the plasmid DNA, and thus not available as free protein, repressor binding at its target site on the chromosome should be reduced. This reduction of binding to the chromosomal ModE-binding site can be monitored using an appropriate modA '-'lacZ fusion in the chromosome. More specifically, various lengths of DNA ranging from -247 to +25 of the modAoperator/promoter region were cloned into the multiple cloning site of the high copy plasmid vector pUC19, as described in "Materials and Methods" section. The resultant plasmids were subsequently transformed into strain SE2069, which contains a chromosomal 4'(modA 'lacZ).

An initial inspection of the modA operator/promoter DNA suggested that perhaps an eight-bp inverted repeat (TAACGTTA) spanning the +4 to +11 region might be involved since inverted repeats had been shown to function as the binding sites for the trp and met repressors (40, 54, 55). However, a closer inspection of the modA operator/









48

promoter DNA revealed that there is a preponderance of GTTA sequences in this DNA (Fig. 6). Specifically, there are 8 GTTA tetramers between the translation start sites of modA and modE and all are located near the modA end. It is especially notable that two of the GTTA tetramers straddle the -10 region, and that there is a GTTA tetramer inverted repeat (the eight-bp inverted repeat referred to above) located 4-bp downstream of the transcription start site. Thus, the number and location of these GTTA tetramers provided compelling reasons for investigating their potential role in the binding of ModE.

For determination of the degree of titration of ModE by the various plasmids, 3galactosidase assays were performed as per standard protocol, and ampicillin was added to the culture medium to ensure the maintenance of the plasmids in the cells. Supplementation of 1 mM molybdate to the culture medium was also required since strain SE2069 cells are functionally Mod; thus, in the absence of added molybdate, ModE fails to serve as a repressor of modA '- 'lacZ expression (Table 3). The resulting 1galactosidase activities are presented in Table 7. The mod DNA sequence that is present in each of the plasmids is also indicated in Table 7. It should be noted that the data is presented in a manner between plasmids pAM7 to pAM5 that the first 1-galactosidase activity corresponds to the construct bearing all of the GTTA tetramers (plasmid pAM7), while each following plasmid has successive GTTA sequences removed from left to right, respectively, until plasmid pAM5 is reached.

The level of modA '- 'IacZ expression for strain SE2069 with plasmid pAM4, which contains the entire intergenic DNA sequence shown in Fig. 6, is 510 1-galactosidase activity units while expression for SE2069 cells alone or for SE2069 harboring plasmid









49










TA
T G C T G -C C -G C -G C -G modE G -C T -A A-T E 276nt GATTCCCTGA -TTTCTTTTTCTCAACTTCCTGC


TTTTCCTGCCGATATTTTTTCTTATCTACCTCACAAAG CAA

-35 -10
rTGC n AATTCCGA IEGTC ::CGC

modA

CT 5C ATTAAGG Z CA K



Fig. 6. Intergenic region of modEA DNA. Boxes outline the modA -10 and -35 sequences. The modA transcription start site is indicated by the boxed 'A' with the arrow, and the translation start site appears as the boxed 'ATG' or 'CAT'. 'GTTA' sequences are shown in reverse print.










50





Table 7. Titration of ModE protein in vivo by various DNA sequences in
strain SE2069 (modA '-'lacZ+)



Plasmid DNA Sequence P-Galactosidase Activity

No Plasmid 150
1 2
pAM4 GGTACC- 186nt CAAAG CA TGCTGGGAAAATTCC-CC-TA c-TTAA 5 C 510
3 4 5 6 7 8

pAM9 GGTACC- 186nt 140 pAM7 CAAAGMLCA MGCTGGGAAAATTCC
GAWTCIL CTCTT~~~A 510 pAM8 AAATTCC~g~CT CGCCTACAC TTAAGG 710 pAN13 iU CGCCTACAATATAAGGI 750 pAM10 IU CGCCTAC AC TTAA 150
pAM6 IrCGCCTACcAU ATTAAGGGi 120 pAMS CGCCTAC A ATTAAGG 60 pAMi5 GGTACC- 186nt CAAAG CAAMc TGGGAAAATTCC~CGAo rcGL CTCA------CATTAAGGG 90 pAM16 GGTACC- 186nt CAAAG cAK i CTGGGAAAATTCCGGT*T. C T -C-GG-- C 410


Cells were grown in LBG medium with molybdate (1 mM). -galactosidase activities are expressed as nanomoles.minute-'.milligram of cell protein'. The putative ModE protein binding sites are indicated by inverse print and are numbered (pAM4).









51

pAM9 which contains only the 5' 192-bp of the intergenic region is 150 and 140 units. respectively. These data indicate that the modEA intergenic region is capable of titrating a fraction of ModE, thereby relieving repression of the chromosomal modA '- 'lacZ and that the first 192-bp of the insert in plasmid pAM4 is not necessary for this titration. Based on the results presented in Table 4, a P-galactosidase activity of 1,300 units is expected for strain SE2069 under conditions of derepression; thus, the relief of repression of modA '- 'lacZ by plasmid pAM4 seen in the titration experiments is not total, but is of sufficient level to discriminate the sequence required for the binding of ModE. Removal of the first two GTTA tetramers resulted in an increased titration of ModE by the plasmid DNA as shown by the increase in P-galactosidase activity from 510 units to 710 units when plasmid pAM8 is present rather than plasmid pAM4 or pAM7. Likewise, removal of the next GTTA tetramer (plasmid pAMI 3) yields an even higher level of titration, 750 P-galactosidase units as opposed to 710. The higher P-galactosidase activities obtained using plasmids pAM8 and pAM13 suggest that the first three GTTA tetramers may interact with ModE forming unstable associations, which results in an overall less efficient binding of ModE to its true target binding regions. The fact that these first three GTTA tetramers do naturally exist in the modEA intergenic region implies that perhaps these GTTA tetramers help to modulate the repression of the modABCD operon by ModE under suboptimal intracellular molybdate concentrations.

Removal of the first four as well as the first four and part of the fifth GTTA tetramers (plasmids pAM10 and pAM6, respectively) results in the complete loss of titration of ModE, which demonstrates that the fourth and fifth GTTA tetramers are








52

necessary for successful binding of ModE. Moreover, these two GTTA tetramers are the ones that are situated in the modA -10 sequence, and would be expected to participate in binding of the repressor given that binding of a repressor at this point would serve to obstruct the binding of RNA polymerase and thereby limit transcription of the ensuing gene. Furthermore, removal of only the sixth and seventh GTTA tetramers (pAM15) also results in complete loss of titration as indicated by the drop in P-galactosidase activity to 90 units. This finding establishes the requirement for the sixth and seventh GTTA tetramers for ModE binding, and again the location of these two GTTA sequences 4-bp downstream of the transcription start site, supports the assertion that these sites are integral to the binding of repressor. Additionally, the fact that the binding of ModE to the DNA requires the presence of the fourth and fifth as well as the sixth and seventh GTTA tetramers intimates that the binding of ModE to one set of sites necessitates the initial binding of ModE to the other site. This potential binding order in turn evokes the possibility of cooperative binding occurring; however, none of the methods used in this study to characterize the binding of ModE to its target DNA is capable of evaluating cooperativity. Introduction of plasmid pAM16, which only lacks the eighth GTTA sequence, into strain SE2069 results in a partial loss of titration of ModE, thereby implicating the involvement of this eighth tetramer in stabilizing ModE once it is bound to the DNA or possibly aiding in the initial recruitment of ModE.

Based on these in vivo titration experiments, it has been determined that the

binding of ModE minimally requires the presence of the fourth, fifth, sixth, and seventh GTTA tetramers, while the presence of the eighth GTTA tetramer increases the efficiency









53

of binding. Furthermore. the arrangement of these GTTA tetramers required for binding (the fourth and fifth GTTA tetramers [GTTATATTG] seem to constitute one binding site, while the sixth and seventh GTTA tetramers [TAACGTTA] could form a second site) and the size of ModE protein (28,271 Da), predicts that ModE likely binds at each of the two sites as a dimer. The differences in the sequences also suggests different binding efficiencies.

In vitro studies characterizing the binding of ModE to its target DNA. Having

established putative binding sites for ModE using the in vivo titration experiments, in vitro techniques such as DNA-mobility shift assays and DNase I-footprinting experiments were undertaken in an effort to confirm the in vivo findings as well as to provide further information on the kinetics of the interaction. One set of experiments designed specifically to determine whether the in vivo titration results would be mirrored by DNA-mobility shift assay results, involved preparing binding reactions containing various amounts of ModE and 32P-labeled DNA-oligomers derived from the mod DNA in plasmids pAM13, pAMI5 and pAM16. In these experiments, a 211-bp DraI/EcoRI fragment derived from plasmid pAM13 (40 bp of modA-operator/promoter DNA from -15 to +25 and 171 bp of plasmid pUC 19) and a 313-bp EcoRI/HindIII fragment from plasmid pAM15 (272 bp of modAoperator/promoter DNA from -247 to +25 and 41 bp plasmid pUC 19 DNA) and a 305bp-EcoRI/HindIII fragment from plasmid pAM16 (267 bp modA-operator/promoter DNA from -247 to +20 and 38 bp plasmid pUC19 DNA) was used as the source of mod operator/promoter DNA. Standard DNA-mobility shift assay protocol outlined in "Materials and Methods" section was followed, and the results are presented in Fig. 7.








54







V V





II










Ill









1 2 3 4 5 6 7 8 9 10



Fig. 7. DNA-mobility shifts using plasmid pAM13-derived (panel I), plasmid pAMI5derived (panel II), and plasmid pAMI6-derived DNA (panel lII) and ModE. A 0.1 pmol amount of labeled DNA and 1 mM sodium molybdate were present in each binding reaction. The following amounts of ModE had been added to the reactions as indicated: lane 1, without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6, 50-fold excess; lane 7, 100fold excess; lane 8, 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess. The arrow indicates a 'biologically significant" shifted-band.









55

In the DNA-mobility shift for the plasmid pAM13-derived DNA (Fig. 7, panel I), a detectable shift to position 'A' (complex A) occurred at a protein to DNA ratio of 1.0 (25.7% DNA shifted to form 'A'). This 'shift form' predominated until a ModE excess of 25-fold was reached, at which point, a second band begins to emerge. Other ModE-DNA complexes also emerged at 250-fold, 500-fold, and 1000-fold ModE excess. These different complexes most likely represent additional binding of ModE to the DNA in a nonspecific manner with little biological significance. When DNA is included in the binding reactions which is incapable of titrating ModE (plasmid pAMI 5-derived DNA: Table 7), shift-form 'A' ModE-DNA complex does not arise (Fig. 7, panel II), but the higher-mass shift complexes did arise at ModE excesses of 250, 500, and 1000-fold. Thus, the DNA-mobility results, combined with the inability of plasmid pAM15 to titrate ModE in the in vivo experiments (Table 7), indicates that the 'A' complex form is the DNA-target-specific form. Based on the ModE titration data (Table 7), plasmid pAM16derived DNA should prove capable of forming the shift-form 'A' ModE-DNA complex, albeit at higher ModE/DNA ratios than required for plasmid pAM13-derived DNA-ModE complex formation. This expectation was realized in the DNA-mobility shift experiments, as complex 'A' did form but was only first detectable at a ModE to DNA ratio of 5.0, and over 50% shift of DNA did not occur until 100-fold excess ModE to DNA (Fig. 7, panel III). Again, at higher ModE concentrations, 250, 500, and 1000-fold excesses, the larger ModE-DNA complexes arose with this DNA also.

In all three of the DNA-mobility shifts presented in Fig. 7, multiple ModE-DNA complexes other than the complex identified as shift-form 'A' are observed. Since these









56

ModE-DNA complexes arise regardless of whether the DNA is capable of titrating ModE, it is thought that these complexes result from non-specific binding of ModE to non-mod sequence or to mod sequence located upstream of the modA-operator/promoter when there is greater than 250-fold excess ModE to DNA in the binding reaction. In order to determine whether the non-mod DNA contributes to formation non-specific ModE-DNA associations, a DNA-mobility shift experiment was conducted in which a 42-bp oligomer (modA-operator/promoter DNA from -9 to +33) that is essentially the sequence present as the insert in plasmid pAM10 (modA-operator/promoter DNA from -10 to +25) was used. As indicated in Table 7, this particular DNA does not titrate ModE and, therefore, should not support formation of complex 'A'. Results of this DNA-mobility shift experiment are presented in Fig. 8, and as can be seen, DNA was not shifted at any ModE concentration used except at 1000-fold excess ModE compared to 42-bp oligomer. Furthermore, this shifted DNA-protein complex migrated at a much slower rate than the shift-form 'A' observed with DNA from plasmid pAM13 (Fig. 7, panel I). Inclusion of a smaller DNA fragment (42 bp, as opposed to 211 or 305 or 313 bp), which does not contain non-mod DNA, eliminated formation of the larger ModE-DNA complexes at all ModE concentrations except at the very high 1000-fold excess ModE. This result supports the assertion that the interaction of ModE with non-mod DNA at high ratios of ModE to DNA leads to the formation of non-specific ModE-DNA associations.

The DNA-mobility shift experiments (Figs. 7 and 8) have demonstrated that the formation of the ModE-DNA complex (complex 'A') in vitro requires the presence of the fourth, fifth, sixth, and seventh GTTA tetramers, and that the eighth tetramer is necessary









57


























1 2 3 4 5 6 7 8 9 10




Fig. 8.- DNA-mobility shift featuring 42-bp modA-operator/promoter DNA (-9 to +33) and ModE. A 0.1 pmol amount of DNA and 1 mM sodium molybdate and the following amounts of ModE were included in each binding reaction. Lane, 1 without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6, 50-fold excess; lane 7, 100-fold excess; lane 8, 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess.









58

for efficient binding (Table 7: Fig. 7). The mobility shift experiments fully support the results from the in vivo ModE titration experiments (Table 7), further lending credence to the use of the titration experiments as a means of probing the association of protein with its target DNA sequence.

It has already been shown both in vivo and in vitro that repression of the

modABCD operon by ModE requires the presence of intracellular molybdate (Tables 3 and 4; Fig. 5). Therefore, DNA-mobility shift experiments were conducted for the purpose of establishing whether molybdate is required for the binding of ModE to its target DNA. For these DNA-mobility shift experiments, a 43-bp oligonucleotide (-18 to +25), was used in the binding reaction along with the appropriate amount of ModE protein. For one set of shifts, sodium molybdate was added to the binding reactions, to the gel the binding reaction samples were electrophoresed in, and to the electrophoresis tank buffer at a final concentration of 1 mM. For the second set of mobility shifts, molybdate was excluded. These two mobility shifts, presented in Fig. 9, indicate that molybdate is necessary for efficient binding of ModE, as the 'A' complex is first detectable at a ModE/DNA ratio of 1.0 in the presence of 1 mM sodium molybdate, whereas in the absence of molybdate, the ModE-DNA complex was not detectable until a 100-fold excess ModE was included in the binding reaction. Furthermore, binding of 50% or greater of DNA with ModE was never achieved in the binding reactions lacking molybdate, while in the presence of 1 mM sodium molybdate, greater than 50% shift of the DNA to the ModE-DNA complex 'A' was obtained at a ModE/DNA ratio of 50.









59



I























1 2 3 4 5 6 7 8 9 10


Fig. 9. DNA-mobility shifts using a 43-bp oligomer spanning the -17 to +25 region of the modA-operator/promotor region and ModE. Panel I- sodium molybdate was not added to the binding reaction, gel or running buffer. Panel II- 1 mM sodium molybdate was present in the binding reaction, gel, and running buffer. A 0.1 pmol amount of DNA was present in the binding reaction. The following amounts of ModE were present in the binding reaction samples as indicated. Lane, 1 without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6. 50-fold excess; lane 7, 100-fold excess; lane 8, 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess.









60

The apparent dissociation constant (KD) for the binding of ModE to the 43-bp oligomer in the presence of ImM molybdate was determined after quantitation of the percent of shifted DNA using a Phosphorimager. This correlation is presented in Fig. 10 and an apparent KD of 0.33 pLM was calculated for the ModE-DNA complex 'A'. An apparent KD for the binding of ModE to the 43-bp DNA in the absence of 1 mM sodium molybdate could not be calculated since even at 1,000 fold excess of protein, the amount of shifted DNA was less than 50% of the total DNA.

Interaction of molybdate with ModE. Given that the presence of molybdate is

required for efficient binding of ModE to its target DNA, the concentration of molybdate needed for a 50% shift of the 43-bp oligomer to the form 'A' ModE-DNA complex was determined. A ModE concentration of 0.5 pM (protein to DNA ratio of 50) was chosen for these shift experiments because, at this ModE concentration in the binding reaction, a 60% shift of the DNA was obtained in the presence 1 mM sodium molybdate, and at this ratio, the contribution of sodium molybdate to the ModE-DNA interaction would be readily discernable.

Based on these experiments, a 50% shift of the 43-bp oligomer was accomplished at a molybdate concentration of 6 tM (Fig. 11). It is not unreasonable to consider that an artificially high apparent K, value for the association of molybdate with ModE could have been obtained given the inherent limitations of the experiment used to derive the apparent KD. Specifically, for the molybdate concentration DNA-mobility shift experiments, once the binding reaction samples were loaded onto the gel, the effective concentration of sodium molybdate present in those samples would decrease as the sodium molybdate








61








100


80 60


40


20



0.01 0.1 1.0 10.0

[ModE] (M)


Fig. 10. Percentage of ModE-43-bp-oligomer complex 'A' formed versus concentration of ModE protein in the presence of 1 mM sodium molybdate. These values are from the DNA-mobility shift experiment presented in Fig. 8, panel II.








62









100



80



S60

-I
40



20




0.0001 0.001 0.01 0.1 1.0 10.0 100.0 1000.0 [Molybdate] (gM) Fig. 11. Percentage of 43-bp-oligomer bound to ModE protein in relation to the concentration of sodium molybdate present in the binding reactions in a DNA-mobility shift experiment.









63

concentration for the entire system undergoes equilibration, and secondly, electrophoresis of the binding reaction complexes may perturb the ModE-molybdate association.

If the actual binding of ModE to its target DNA constitutes the rate-limiting step of the association of ModE with the DNA rather than the interaction of molybdate with ModE, then this concentration of molybdate (6 pM) required for half-maximal binding most likely represents an upper limit value since it has previously been shown that the apparent dissociation constant for the interaction between ModE-molybdate and modoperator/promoter DNA is 0.3 pM.

If however, it is the association of molybdate with ModE and not the binding of ModE to the DNA which serves as the rate limiting step, then the 6 pM apparent KD value for the dissociation of molybdate from ModE may indeed prove valid. This possibility is especially compelling in light of the KD value of 3 tpM that was determined for the affinity of molybdate for the periplasmic molybdate-specific-binding protein, ModA, using fluorescence studies (59). Furthermore, an apparent K of 6 pM for the binding of molybdate to ModE seems physiologically acceptable since molybdate entering the E. coli cell would preferentially be activated and ultimately incorporated into an appropriate apomolybdoenzyme. Under these physiological conditions, the apparent K, value for molybdate in ModE-system would be expected to be higher than the Km value for molybdate for the activation enzyme. Only when the internal molybdate saturates the biosynthetic pathway, would molybdate be available for association with ModE resulting in the repression of the modABCD operon and cessation of further production of the components of the molybdate-specific-transporter. If the apparent Kn for the dissociation









64

of molybdate from ModE were much lower, then it would be likely that molybdate's interaction with ModE would be favored above its activation and incorporation into molybdoenzymes which would prove physiologically untenable for the cell. Given the uncertainty of whether it is the interaction of molybdate with ModE or it is the binding of activated ModE to DNA which acts as the limiting step for the repression, it seems that the only way to establish an accurate description of the kinetics governing the association of molybdate with ModE, and in turn, the binding of the ModE-molybdate complex to its target DNA is to determine the interaction in solution and compare to the appropriate enzymes competing with ModE for molybdate.

Oxyanion specificity of ModE. Having established that molybdate is required for the repression of the modABCD operon in vivo and the efficient binding of ModE to its target DNA in vitro, the question of whether structural analogs of molybdate could substitute for molybdate was addressed. This question was answered in two ways: (1) modA '-'lacZ expression in strain SE2069 cells that were cultured in LBG medium supplemented with 1 mM sodium molybdate, 1 mM sodium tungstate, I mM sodium sulfate, or 1 mM sodium ortho-vanadate, and (2) DNA-mobility shift experiments in which the binding reactions, gels, and tank buffers contained 10 gM sodium molybdate, 10 [IM sodium tungstate, 10 jM sodium sulfate, or 10 gIM sodium ortho-vanadate.

The resulting levels of modA '- 'lacZ expression in strain SE2069 cultured with the various oxyanions are presented in Table 8. Only in those cultures of strain SE2069 grown in medium supplemented with 1 mM sodium molybdate or 1 mM sodium tungstate did repression of the modABCD operon occur, as indicated by their low P-galactosidase








65

Table 8. Expression of modA '- 'lacZ in strain SE2069 cultured in the presence of various
oxyanions.


Strain Relevant Genotype Oxyanion 3-galactosidase Activitya Added


SE2069 4(modA '- 'lacZ) None 1.300 MoO;2 30 WO4-2 40 SO4-2 1,400 VO-3 1,900 aCells were grown in LBG with the appropriate oxyanion at a final concentration of 1 mM. Enzyme activities are expressed in nanomoles- minute--milligram of cell protein'.








66

activities. Therefore. it appears that tungstate is the only oxyanion tested which is capable of functionally substituting for molybdate in the molybdate-dependent activation of ModE in vivo. Furthermore, the fact that higher modA '- 'lacZ expression levels were observed when the cells were grown in LBG medium containing sodium sulfate and especially sodium vanadate compared to cells grown in medium without any oxyanion supplementation (1,400 and 1,900 P-galactosidase units, respectively, as opposed to 1,300 units) implies that perhaps the presence of these two oxyanions inhibits the interaction of ModE with any of the trace amount of molybdate that may have been present in the cells, thereby reducing binding of ModE-molybdate complex to the modA-operator/promoter region.

The results of the DNA-mobility shift experiments (Fig. 12) which contained 10 gM amounts of the previously indicated oxyanions in the binding reactions, were consistent with the in vivo modA '- 'lacZ expression data (Table 8). More specifically, high affinity binding of ModE to the 43-bp DNA and shift to complex 'A' only occurred in the presence of sodium molybdate (Fig. 12, panel I) or sodium tungstate (Fig. 12, panel II), while a low level of binding was seen for the shifts with sodium sulfate (Fig. 12, panel III). Sodium vanadate was fifty times less effective in supporting DNA-ModE protein interaction (Fig. 12, panel IV). The apparent KD values for the ModE-DNA complex obtained from the correlation of percent of shifted DNA (ModE-DNA complex species 'A') versus the concentration of ModE for the shifts with the different oxyanions were as follows: 0.22 LM with 10 giM sodium molybdate supplementation, 1.3 jiM with sodium tungstate, and 4.7 pM with sodium sulfate (Fig. 13). An apparent KD value for the


















Fig. 12. DNA-mobility shift experiment in which four different oxyanions were present in the binding reactions. The binding reactions contained 10 pM sodium molybdate (panel I), 10 pM sodium tungstate (panel II), 10 pM sodium sulfate (panel III), or 10 piM sodium ortho-vanadate (panel IV) along with 0.1 pmol of the 43-bp modAoperator/promoter DNA. The following concentrations of ModE were present in the binding reactions per lane as indicated: lane 1, no ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6, 50-fold excess; lane 7, 100-fold excess; lane 8, 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess.






















B




A













1 2 3 4 5 6 7 8 910 1 2 3 4 5 6 7 8 9 10
i iiii i iii~ iiii~iiiii~ ii iii .............................. ...i ii








69








100


80


60
O II .SO .i

40 MoO42 W04


20


0

0.01 0.1 1.0 10.0 [ModE] (M)




Fig. 13. Percent shifted DNA versus concentration of ModE resulting from DNAmobility shift binding reactions containing various oxyanions. Bindng reactions included 10 CpM sodium molybdate (filled squares), 10 pLM sodium tungstate (open squares), 10 jiM sodium sulfate (open circles), or 10 tM sodium ortho-vanadate (closed triangles).








70

binding of ModE in the presence of 10 tM sodium vanadate could not be calculated since 50% shift of the DNA was not achieved with the ModE concentrations used. It is not surprising that tungstate proved to be the only oxyanion of those tested which was capable of substituting for molybdate in activating ModE to any significant degree, since it has previously been shown that activated tungstate can be incorporated successfully into some E. coli apo-molybdoenzymes in place of molybdate (57). However, these tungstatesubstituted molybdoenzymes are not functional probably because of a failure of electron transfer across the tungsten center. Yet, the stable association of tungstate with ModE would result in the promotion of the activation of ModE, since presumably, the activation of ModE relies only on the binding of an appropriately sized and shaped moiety to affect the change in conformation of the protein which allows for the efficient binding of DNA, and does not require any further chemical process such as the transfer of electrons.

Stoichiometry of ModE-DNA complex. Since a stable ModE-43-bp DNA shift complex (complex form 'A') has been reproducibly obtained under a variety of conditions, this general method was used to determine the molecular weight of the ModEDNA complex which would then indicate the number of ModE proteins which are associated with the DNA in complex 'A'. For the molecular weight determination of the ModE-DNA complex, standard binding reaction samples supplemented with 1 mM sodium molybdate were electrophoresed in 5%, 6%, 7%, and 8% nondenaturingpolyacrylamide-TBE gels. The resulting Rf values for ModE-DNA complex and the protein standards after migration through the four different gels were used to prepare a Ferguson plot (19) from which the molecular weight of the ModE-DNA complex could be








71

extrapolated (Fig. 14). Based on the results of these experiments, a value of 81.247 Da was obtained for the size of this complex. This molecular weight compares favorably with a molecular weight of 83,144 expected for the association of a ModE-dimer (56. 424 Da) with the 43-bp DNA (26,602 Da), as there is only a 2.3% difference in the expected value versus the observed molecular weight. Thus, the ModE-DNA complex form 'A' is minimally a ModE dimer bound to the 43-bp modA-operator/promoter DNA. In these experiments, the axial ratio and electrophoretic mobility of the protein-DNA complex was considered to be similar to that characteristic of the protein itself. This assumption is based on the small size of the DNA used in this experiment.

DNase I-footprinting studies for the identification of the ModE binding sites in the DNA. Although, the ModE titration experiments (Table 7) and mobility shift experiments (Fig. 7) implicated the GTTA tetramers found in the modA operator/promoter DNA in the binding of ModE, direct determination of the DNA sequence bound by ModE is required to confirm the involvement of the GTTA tetramers. To this end, DNase I-footprinting experiments were carried out. For one set of DNase I-footprinting experiments, 1 mM sodium molybdate was added to the binding reactions which also included 0.1 pmol of a 446-bp 32P-labeled DNA fragment containing the entire modA-operator/promoter region as well as 1-fold, 10-fold. or 100-fold excess ModE compared to DNA. Analysis of the protection patterns resulting from the DNase I cleavage of the binding reaction mixtures (Fig. 15) reveals that there are three areas of protection. These protected regions are detectable at 10 and 100-fold excess ModE. It is notable that the first protected region








72









40




30




0 20




10
ModE-DNA


0 "
0 100 200 300 400 500 600

MOLECULAR WEIGHT (10"3)




Fig. 14. Ferguson plot generated from the Rf values obtained after electrophoresis of the non-denatured protein standards and ModE-43-bp-modA-operator/promoter DNA complex in 5%, 6%, 7% and 8% polyacrylamide-TBE gels. The dotted lines indicate the extrapolation for the determination of the molecular weight of the ModE-DNA complex form 'A'.










73



























2.




















1 234




Fig. 15. DNase I-footprinting experiment in which the binding reactions contained 1 mM sodium molybdate,
0.1 pmol 32P-labeled-446-bp-modA-operator/promoter DNA and various concentrations of ModE. Lane 1, DNA alone; lane 2, 1-fold excess ModE compared to DNA; lane 3, 10-fold excess ModE; lane 4, 100-fold excess ModE. The boxed sequence represents the nucleotides located in the protected regions. The indicates bases that are subject to hypersensitivity of DNase I cleavage.









74

(from top of Fig. 15) (GTTATATTG; -15 to -7) encompasses the fourth and fifth GTTA tetramers located in the -10 region of the modA-operator/promoter region (Table 7), which would confirm the importance of GTTA sequences in the binding of ModE. Also there appears to be a 'G' located three bases upstream of this first protected region which is hypersensitive to DNase I-cleavage. However, the second (CCTACAT; -3 to +4) and third (GTTACAT; +8 to +14) protected areas, respectively, do not uphold the prediction of the involvement of the GTTA tetramers located in these regions in the binding of ModE since only part of the sixth and seventh GTTA tetramers (Table 7) which comprise the 8bp inverted repeat (+4 to + 11) are located in the protected area. A second DNase I hypersensitive nucleotide ('A') is located at the lower boundary of the second protected region. The location of these hypersensitive cleavage spots are not unexpected since it has previously been shown that hypersensitive cleavage sites often flank areas of protection

(10).

In light of the findings from the DNase I-footprinting experiments, a reevaluation of the sequence involved in binding ModE was undertaken which resulted in the assignment of the sequences TACAT and TATAT as those necessary for the binding of ModE, since these sequences are common to all three protected regions. The fact that the TATAT and TACAT sequences found in the three protected regions overlap the fourth and fifth GTTA tetramers (first protected region) and sixth and seventh GTTA tetramers (second and third protected regions), respectively (Fig. 15 and Table 7), supports the interpretation that removal of the above indicated GTTA sequences from certain plasmids used in the ModE titration experiments resulted in the loss of binding by ModE.








75

A DNase I-footprinting experiment was also conducted in which molybdate was withheld from the binding reactions in order to determine whether ModE is capable of binding the DNA, thereby protecting it from cleavage by DNase I, even in the absence of molybdate, and secondly, to determine, if protection does occur, whether the protected regions are identical to those obtained from the reactions supplemented with sodium molybdate. As can be seen from the results of these experiments (Fig. 16), protection of the DNA from DNase I cleavage did occur in the presence of 100-fold and 1000-fold excess ModE, which is ten times more ModE required for the protection in this experiment than was needed when the reactions included sodium molybdate. It was expected that protection of the DNA from DNase I cleavage in the absence of molybdate would be compromised since it had previously been shown in mobility shift experiments (Fig. 9) that the binding of ModE to the 43-bp modA-operator/ promoter fragment was severely reduced without the inclusion of sodium molybdate in the binding reactions. The protection pattern that did result from the binding of ModE to DNA in the absence of added molybdate was identical to that obtained from the experiment conducted with sodium molybdate supplementation. However, the 'G' at -18 is hypersensitive to DNase I cleavage only in the presence of molybdate. It is possible that in this experiment, a small fraction of ModE is bound to trace amounts of molybdate present in the solution and that only this ModE-molybdate complex binds to the DNA. Alternatively, molybdate-freeModE can bind to DNA but at a very low efficiency. The differential hypersensitivity of 'G' at position -18 suggests that molybdate-free-ModE is capable of binding to the DNA.










76






























c12345 E-4



















F 1. s f pn f i dr o d f teilanE t; l oe n ois l d i nc td wv to D c i~a ~ E-4 sd~ ~ C) taOr ~ ~n *~ 0





1 2 3 4 5r















nucleotides which are hypersensitive to DNase I cleavage.








77

The protection of the modA-operator/promoter DNA by ModE in three distinct regions is not consistent with the determination of dimeric binding of ModE to DNA in the ModEDNA complex (Fig. 14). These results taken together suggest that, the association of ModE with the modA-operator/promoter DNA is not simply binding of ModE to each site. Rather, these data suggest a possible scenario in which a monomeric ModE protein may initially bind loosely to one of the three protected regions, and this initial binding then allows for the rapid recruitment of additional ModE-molybdate to the region for stable binding of ModE protein as a dimer to the other two sites. This monomeric association with the DNA is probably unstable and is rapidly removed by electrophoresis in the DNAmobility shift experiment. Yet, at this point, any further discussion of the course of the binding of ModE to the DNA must await binding experiments carried out and analyzed in solution.

In order to evaluate the role of the three ModE-protected regions in vivo, another set of plasmids was constructed for ModE-titration experiments similar to the ones previously described (Table 7). Plasmid pAM17 contains protected sites I and 2, while plasmid pAM18 has protected sites 1 and 3 with the spacing between the sites preserved (Table 9). Plasmid pAM19 contains only the protected regions 1 and 3 separated by only two bp. The modA '- 'lacZ expression levels, measured as P-galactosidase activity, for strain SE2069 containing the above-mentioned plasmids, indicated that only plasmid pAM17 is capable of partially titrating ModE (Table 9). Strain SE2069 harboring plasmid pAM17 produced only 220 P-galactosidase units, while only 70 P-galactosidase units were obtained for strain SE2069 lacking any plasmid. The level of titration of ModE in









78




Table 9. Titration of ModE protein in vivo by various regions of the modA-operator
/promoter DNA in strain SE2069 (modA '-'lacZ+)


Plasmid DNA Sequence 4-Galactosidase Activity

No Plasmid 70 pAM4 GGTACC- 186nt CAAAGGTTAGCAATAACTGCTGGGAAAATTCCGAGTTAGTCGTI GTCGCC AACGU AAGGGGTTAC 600
1 2 3

pAM17 TCGT rGTCGCC AA 220
1 2

pANS CGCCAACGTG AAGGGGTTAC 60
2 3

pAM18 TCG~[ir GTCGCCTgAACGTr ZT 60
1 3

pAM19 TCGrI- --------------GT so
1 3


Cells were grown in LBG medium with molybdate (1 mM). 0-galactosidase activities are expressed as nanomoles.minute-'.milligram of cell protein-'. The ModE protein binding sites are indicated by inverse print and the numbers designate the particular binding site (see Text for details). The underlined sequence "CTTGG" replaces the sequence in region 2 preserving the spacing between regions 1 and 3.









79

the cells containing plasmid pAM17 is not the maximal level possible, since 600 Pgalactosidase units of activity was produced by strain SE2069 with plasmid pAM4 which contains all three protected regions. Thus, it appears that only protected regions I and 2 are absolutely required for the binding of ModE, but that in addition, region 3 is necessary for maximal binding.

DNA-mobility shift experiments were also conducted using the modAoperator/promoter DNA contained in plasmids pAM17, pAM18. and pAM19. As would be expected, based on the in vivo titration experiments, only the modA-operator/promoter DNA present in plasmid pAM17 was capable of forming the 'A' complex in the presence of ModE protein as presented in Fig. 17 (Panel I). Moreover, as is also suggested by the partial titration seen in the ModE titration experiments, the half-maximal binding of ModE to the 36-bp modA-operator/promoter fragment occurs at a ModE concentration of 1.8 gIM (Fig. 18) which is considerably higher than the 0.33 glM apparent KD value obtained for the binding of ModE to DNA containing all three sites (Fig. 10). The results of the in vivo ModE-titration experiments coupled with the in vitro DNA-mobility shift experiments indicate that the binding of ModE only requires sites 1 and 2, but site 3 is necessary for efficient binding at sites 1 and 2. These data lend credence to the premise that site 3 may serve as a location to which ModE initially binds, thereby promoting the stable binding of ModE to sites 1 and 2.









80






I























1 2 3 4 5 6 7 8 9 10




Fig. 17. DNA-mobility shift experiments using the modA-operator/promoter DNA contained in plasmids pAMI7, pAM18, and pAMI9 and ModE. For the shift with DNA present in plasmid pAMI7 (panel I), a 32P-labeled-36-bp modA DNA fragment was included in the binding reactions along with 1 mM sodium molybdate and the appropriate amount of ModE (lane 1, without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess ModE; lane 4, 10-fold excess ModE; lane 5, 25-fold excess ModE; lane 6, 50-fold excess ModE; lane 7, 100-fold excess ModE; lane 8, 250-fold excess ModE; lane 9, 500-fold excess ModE; lane 10, 1000-fold excess ModE). The mobility shifts featuring the 45-bp modA DNA present in plasmid pAMI8 (panel II) and 31-bp modA DNA present in plasmid pAM19 (panel III) also contained 1 mM sodium molybdate and the amount of ModE indicated for the lanes in panel I.








81







100



80



Z 60



r0 40



20



0
0.01 0.1 1.0 10.0 [ModE] (gM)



Fig. 18. Plot of% shifted DNA versus concentration of ModE required for the shift featuring the 36-bp modA DNA present in plasmid pAM17. The data were from the experiment presented in Fig. 17, Panel I.








82

Molybdate-independent ModE Proteins

Mutation of DNA which codes for proteins of interest is often employed as an approach to study structure/function characteristics of the protein. In this study hydroxylamine, which causes G-to-A transitions, was used to mutate the DNA coding for ModE (plasmid pAG1) in order to identify critical amino acids responsible for the association of molybdate with ModE and/or the binding of ModE to its target DNA. Following the hydroxylamine treatment, two different mutant phenotypes were observed: partial loss of repression of the modABCD operon in the presence of molybdate and complete repression of the modABCD operon in the absence of molybdate (superrepressor). The phenotypes observed were confirmed by assaying strain SE 1811 harboring plasmids with modE mutations for 3-galactosidase activity, and the results of these experiments are presented in Fig. 19.

The partially derepressed (A76V) ModE mutant has a substitution of valine for

alanine at amino acid position 76. Three "superrepressor" mutants which were capable of repressing the modABCD operon in the absence of molybdate were also isolated, and two of these mutants harbor the following mutations: threonine at amino acid position 125 replaced with isoleucine (T1251) and glycine at position 133 changed to aspartic acid (G133D). For the third "superrepressor" mutant (Q216*), the C-terminal 47 amino acids starting from the amino acid at position 216 were deleted.








83



Plasmid Mutation -galactosidase Activity
-Mo +Mo


262
pAG1 600 100

pAG 1(A76V) 1 1,300 600 A76V
pAGI(T125I) 125 110 T1251
pAG1(G133D) 40 40 G133D
pAGl(Q216*) <10 <10 Q216*



Fig. 19. Analysis of ModE mutants. 3-galactosidase activity produced by strain SE1811 carrying the indicated mutation was determined and expressed as nanomoles-min1
-milligram of cell protein'. Cells were grown in LBG with (+) or without (-) sodium molybdate (Mo) at a final concentration of 1 mM. represents translation termination.








84

The mutation in the partially derepressed mutant (A76V) lies immediately after a stretch of three glycines which are identical in three of the four ModE homologs presented in Fig. 3, the fourth (RcMopA) having two of the three glycines. Two of the four homologs (EcModE and RcMopA) have alanine at this position while the other two (AvModE and HiModE) have threonine at this position. Furthermore, this mutation is 23 amino acids away (towards the C-terminal end) from the putative helix-turn-helix structure, so it is possible that this mutation may have disrupted the helix-turn-helix structure resulting in a partial loss of DNA binding by the mutant protein.

Two of the "superrepressor mutants", T125I and G133D, have mutations situated near a five amino-acid long sequence SARNQ (amino acids 126 to 130), which is conserved among the five homologs, with the exception ofR. capsulatus MopA which has an asparagine in place of alanine and a threonine in place of glutamine. In fact, the T125I mutation directly precedes the SARNQ sequence and is itself conserved among three of the homologs (AvModE, EcModE, and RcMopA) with serine substituting for threonine in H. influenzae ModE. The G133D mutation lies two amino acids further on the C-terminal side of the SARNQ sequence, and the glycine is also conserved in four of the homologs (AvModE, EcModE, HiModE, and CpMopII). The glutamine which is replaced by a stop codon in the third superrepressor mutant (Q216*) is unique to E. coli ModE at this position; however, there are regions of conserved amino acids present in the missing Cterminal end of ModE whose loss undoubtedly contributes to the phenotype of this mutant.








85

It is also interesting to note that the amino acid sequence deleted in this mutant (Q261*) contains the only three cysteines present in the wild type E. coli ModE protein, and it is possible that two of these three cysteines may be involved in forming a disulfide bond. Moreover, it is tempting to speculate that ModE adopts an open conformation when molybdate is not bound to the protein, and that upon binding of molybdate, a change in the structural conformation ensues which involves formation of a disulfide bond. This formation of the disulfide bond could stabilize a more compact structure which promotes the efficient binding of ModE to the appropriate DNA sequences. Furthermore, it is apparent that the deletion of the region of the protein which contains these cysteine residues in the (Q216*) mutant ModE protein resulted in the adoption of the "activated" or molybdate-bound conformation even in the absence of bound molybdate.

At the current level of information available concerning the structure of the ModE protein, a more defensible assertion at this point in time is that the SARNQ sequence may be integral to the binding of molybdate by ModE given that two of the three isolated "superrepressor" mutations abut this SARNQ region and that all of the ModE homologs as well as other proteins that are either known to have or are suspected to have interactions with molybdate contain variations of this sequence (Table 10). It should also be mentioned that it would be expected that binding of the negatively charged molybdate moiety by ModE would most likely require positively charged amino acids such as those present in the SARNQ region. Furthermore, if the nature of the G133D "superrepressor" mutation is considered, it is possible that the substitution of aspartic acid in place of










86

Table 10. List of proteins which are known to interact with molybdate and DNA and
contain a possible molybdate-binding motif.



Protein Molybdate-binding motif sequence Reference




ModE from E. coli SARNQ (24,47,60) ModE from H. influenzae SARNQ (20) ModE from A. vinelandii SARNQ (43) MopII from C. pasteurianum SARNQ (32) MopA from R. capsulatus SnRNt (76) FhlA from E. coli SGRNN (47,64) NarX from E. coli SGRNe (38) NarQ from E. coli SIRmQ (12) Estrogen receptor from Rat LdRNQ (44)
pAtNQ

Estrogen receptor from Human LARgQ (39)



Lower case letters represent non-identical or non-similar amino acids. NarX and NarQ are membrane-bound components of the nar regulatory system but do not bind DNA themselves.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E93X8NMCX_A008SW INGEST_TIME 2014-12-23T19:17:23Z PACKAGE AA00026531_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

REGULATION OF AN OPERON CODING FOR THE COMPONENTS OF THE MOLYBDATE-TRANSPORT SYSTEM {modABCD) BY MODE PROTEIN, A MOLYBDATE-DEPENDENT REPRESSOR IN ESCHERICHIA COLI AMY M. GRUNDEN 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 1996

PAGE 2

ACKNOWLEDGMENTS I would like to thank Dr. K. T. Shanmugam for his guidance and support. His insistence on scientific rigor and enthusiasm for research has certainly influenced my development as a scientist. I would also like to thank Drs. John Gander, Lonnie Ingram Philip Laipis, and James Preston for their support and counsel while serving as members of my committee. My gratitude is also offered to all of my associates from Dr. Shanmugam' s lab group for their encouragement and insights. Special thanks are extended to my husband for his tremendous support, encouragement and patience, as well as to my parents, in-laws and other family members and friends. ii

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF FIGURES LIST OF TABLES LIST OF ABBREVIATIONS LIST OF GENE SYMBOLS ABSTRACT INTRODUCTION LITERATURE REVIEW Molybdate Transport Regulation of the modi BCD Operon MATERIALS AND METHODS Materials Bacterial Strains Media Enzyme Activities and Culture Conditions Genetic Experiments Molecular Biology Experiments iii

PAGE 4

23 Biochemical Characterization of ModE RESULTS AND DISCUSSION 30 Evidence for a regulator of the modABCD Operon 35 Analysis of modEF DN A 41 Regulation of modA by ModE Characterization of the Interaction of ModE with modA -operator/promoter DNA 47 82 88 Molybdate-independent ModE Proteins Regulation of modE Regulation of Other Genes by ModE 91 CONCLUSION REFERENCES BIOGRAPHICAL SKETCH 97 101 108 iv

PAGE 5

LIST OF FIGURES Figure page 1 Steps involved in the biosynthesis of the organic portion of the molybdenum cofactor in E. coli 7 2 DNA sequence of the modEF operon 36 3 Amino acid sequences of homologs of E. coli ModE protein 38 4 SDS-PAGE analysis of proteins from different stages of ModE purification 43 5 SDS-PAGE analysis of proteins produced in an in vitro transcriptiontranslation experiment in which plasmid pRMl served as the template DNA. 45 6 Intergenic region of modEA DNA 49 7 DNA-mobility shifts using plasmid pAM 13 -derived (panel I), plasmid pAM15-derived (panel II), and plasmid pAM16-derived DNA (panel III) and ModE 54 8 DNA-mobility shift featuring 42-bp modA -operator/promoter DNA (-9 to +33) and ModE 57 9 DNA-mobility shifts using a 43-bp oligomer spanning the 1 7 to +25 region of the /woc£4-operator/promoter region and ModE 59 10 Percentage of ModE-43-bp-oligomer complex 'A' formed versus concentration of ModE protein in the presence of 1 mM sodium molybdate. 61 1 1 Percentage of 43-bp-oligomer bound to ModE protein in relation to the concentration of sodium molybdate present in the binding reactions in a DNA-mobility shift experiment 62 12 DNA-mobility shift experiment in which four different oxyanions were present in the binding reactions 68 1 3 Percent shifted DNA versus concentration of ModE resulting from DNAmobility shift binding reactions containing various oxyanions 69 v

PAGE 6

14 Ferguson plot generated from the Rf values obtained after electrophoresis of the non-denatured protein standards and ModE-43 -bp-modA -operator/promoter DNA complex in 5%, 6%, 7%, and 8% polyacrylamide-TBE gels 72 15 DNase I-footprinting in which the binding reactions contained 1 mM sodium molybdate, 0.1 pmol 32 P-labeled-446-bp-woc//l -operator/promoter DNA and various concentrations of ModE 73 1 6 DNase I-footprint of ModE in modA -operator/promoter DNA in which sodium molybdate was excluded from the binding reaction mixtures 76 1 7 DNA-mobility shift experiments using the modA -operator/promoter DNA contained in plasmids pAM17, pAM18, and pAM19 and ModE 80 1 8 Plot of % shifted DNA versus concentration of ModE required for the shift featuring the 36-bp modA DNA present in plasmid pAM 17 81 19 Analysis of ModE mutants 83 vi

PAGE 7

LIST OF TABLES Table page 1 Bacterial strains and phages used in this study 14 2 Plasmids used in this study 17 3 Effect of mod gene mutation in mod operon transcription 32 4 Complementation of the mutation in strain SE 1 8 1 1 for molybdatedependent repression of the modABCD operon 34 5 Complementaion of the modE mutation in strain SE 1 8 1 1 with H. influenzae ModE 40 6 Purification profile of ModE 42 7 Titration of ModE protein in vivo by various DNA sequences in strain SE2069 (modA ''lac?) 50 8 Expression of modA ''lacZ in strain SE2069 cultured in the presence of various oxyanions 65 9 Titration of ModE protein in vivo by various regions of the modA -operator/ promoter DNA in strain SE2069 (modA ''lac?) 78 10 List of proteins which are known to interact with molybdate and DNA and contain a possible molybdate-binding motif. 86 1 1 Regulation of modE in different mod strains 89 12 Regulation of modE in various backgrounds 90 1 3 Regulation of modF by ModE 92 14 Regulation of narG expression by ModE 95 vii

PAGE 8

LIST OF ABBREVIATIONS ABC ATP-binding cassette ATP Adenosine triphosphate A. vinelandii Azotobacter vinelandii AvModE Azotobacter vinelandii ModE protein bp Base pair C. pasteurianum Clostridium pasteurianum CpMopII Clostridium pasteurianum MopII protein CTP Cytosine triphosphate Da Dalton DNase I Deoxyribonuclease I DMSO Dimethyl sulfoxide DTT Dithiothreitol EcModE Escherichia coli ModE protein E. coli Escherichia coli EDTA Ethylenediaminetetraacetic acid FDH-H Formate dehydrogenase linked to H, evolution FeMoCo Iron-molybdenum cofactor FHL Formate hydro genlyase GTP Guanosine triphosphate viii

PAGE 9

H influenzae Haemophilus influenzae HiModE Haemophilus influenzae ModE protein HYD Hydrogenase IPTG Isopropyl-p-D-thiogalactopyranoside Kq Dissociation constant kDa Kilodalton Kinase T4 polynuclotide kinase Klenow DNA polymerase IKlenow fragment LB Luria Broth LBG Luria Broth + glucose Mo Molybdenum MoCo Molybdenum cofactor ModE ModE protein MGD Molybdopterin guanine dinucleotide MPT Molybdopterin OD Optical density ONPG Ortho-nitrophenyl-P-D-galactopyranoside ORE Open reading frame PAGE Polyacrylamide gel electrophoresis PCR Polymerase chain reaction R. capsulatus Rhodobacter capsulatus ix

PAGE 10

RcMopA Rhodobacter capsulatus MopA protein SDS Sodium dodecyl sulfate TMAO Trimethylamine-N-oxide Tris Tris-(hydroxymethyl)-aminomethane TTP Thymine triphosphate UV Ultraviolet X-gal 5-Bromo-4-chloro-3-indolyl-P-D-galactopyranoside x

PAGE 11

LIST OF GENE SYMBOLS All the genes listed below are from Escherichia coli unless otherwise indicated. Gene Alternate gene symbols ; phenotvpe affected Symbol arc A Aerobic respiratory control; putative DNA binding protein of Arc modulon arcB Aerobic respiratory control; histidine-protein-kinase of Arc modulon bla p-lactamase; ampicillin resistance chlD Chlorate resistance gene D [renamed as mod; molybdate transport] fdhF FDH-H; formate dehydrogenase (linked to H, evolution) fhlA Putative DNA binding protein necessary for transcriptional activation of the fdhF and hyc operons f nr 'Global' regulator of anaerobic respiration; similar to cAMP receptor protein galETK Galactose operon hyc Hydrogenase isoenzyme-3 Isg H. influenzae lipopolysaccharide synthesis gene moa Molybdopterin biosynthesis; as chlorate resistance gene A mob Molybdopterin biosynthesis; as chlorate resistance gene B mod Molybdate transport; as chlorate resistance gene D moe Molybdopterin biosynthesis; as chlorate resistance gene E mog Molybdopterin biosynthesis; as chlorate resistance gene G xi

PAGE 12

mop Molybdopterin binding protein narL Putative DNA binding protein of respiratory nitrate reductase operon narO Putative nitrate sensor for activation of the NarL protein narX Putative nitrate sensor for activaton of the NarL protein nit-1 mutation allele resulting in nitrate reductase deficiency in Neurospora crassa xii

PAGE 13

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 REGULATION OF AN OPERON CODING FOR THE COMPONENTS OF THE MOLYBDATE-TRANSPORT SYSTEM (modABCD) BY MODE PROTEIN, A MOLYBDATE-DEPENDENT REPRESSOR IN ESCHERICHIA COLI By Amy M. Grunden December. 1996 Chairperson: Dr. K. T. Shanmugam Major Department: Microbiology and Cell Science The modABC genes encode a molybdate-specific transporter in Escherichia coli. When the intracellular molybdate concentration is sufficient to support the activity of molybdoenzymes, expression of the modABCD operon is very low and increases under molybdate-deficient conditions. Isolation and characterization of a mutant strain (strain SE181 1), which maintained the same level of expression of the modABCD operon in both the presence and absence of molybdate in the medium, resulted in the identification of a regulatory gene, designated modE. The modE gene constitutes the first gene in a twogene modEF operon which diverges from the modABCD operon and codes for a 262amino-acid protein (28 kDa). ModE, in the presence of molybdate, repressed the production of plasmid-encoded ModA and ModB' proteins in an in vitro transcriptionxiii

PAGE 14

translation system. DNA-mobility shift experiments demonstrated that ModE is capable of binding as a dimer to DNA derived from the modA -operator/promoter region with an apparent K D of 0.22 uM in the presence of 10 uM molybdate or with an apparent K D of 1.3 fiM in the presence of 10 uM tungstate. DNA-mobility shift experiments also revealed that a 50% shift of DNA is achieved at a concentration of 6 uM molybdate in the binding reaction. DNase I-footprinting experiments identified three regions of protection by ModE in the modA operator/promoter region, 'GTTATATTG at position -15 to -7. 'CCTACAT' at position -3 to +4, and 'GTTACAT' at position +8 to +14. In vivo ModE titration experiments confirmed that the indicated protected regions are required for efficient binding of ModE to DNA. A highly conserved amino acid sequence TSARNQXXG (amino acids 125 to 133) was identified in ModE and homologs from Azotobacter vinelandii, Haemophilus influenzae, Rhodobacter capsulatus, and Clostridium pasteurianum. E. coli ModE mutant proteins with mutations in either the T or G repressed transcription of the modABCD operon in the absence of molybdate, suggesting that this stretch of amino acids is essential for binding molybdate by ModE protein. This study provides evidence that ModE serves as a repressor of the modABCD operon in response to intracellular molybdate levels. xiv

PAGE 15

INTRODUCTION Molybdenum is required for the activity of several enzymes found in animals, plants, and bacteria such as sulfite oxidase, xanthine dehydrogenase, nitrate reductase, formate dehydrogenase, and nitrogenase (77). In humans, sulfite oxidase, xanthine dehydrogenase, and aldehyde oxidase have all been identified as molybdoenzymes. and the absence of these molybdoenzymes, particularly sulfite oxidase, results in severe pathological conditions (37, 57). This medical concern coupled with the fact that nitrogen cycling, which is principally mediated by nitrogen assimilation in plants, requires the participation of molybdoenzymes, has spurred interest in characterizing the steps involved in the successful incorporation of molybdenum into the appropriate apo-molybdoenzymes. From structural studies, it has been determined that molybdenum exists in molybdoenzymes in the form of a pterin-containing molybdenum co factor (MoCo), with the exception of nitrogenase which has an iron-molybdenum cofactor (FeMoCo) (1,31, 57). Other studies have described mutations which result in the pleiotropic loss of molybdoenzyme function (31). One set of these mutants, defined as chl mutants in the bacterium Escherichia coli, were isolated as chlorate resistant mutants and a major fraction of these were found to be defective in all molybdoenzymes, including nitrate reductase (18, 22, 23, 72). 1

PAGE 16

A subset of these chl mutants, designated as chlD, exhibited a pleiotropic loss of molybdoenzymes that could be suppressed by molybdate supplementation of the growth medium (14, 22, 29, 48, 61, 67, 71). Further characterization of the chlD mutants indicated that the genes in this location comprise the molybdate-specific transporter (now called modABC; 69), with modA encoding a molybdate-specific periplasmic binding protein, modB encoding an integral membrane channel-forming protein, and modC encoding an ATP-binding energizer protein (33, 47, 58, 60. 75). This molybdate-specific transporter closely resembles the established ATP-binding cassette (ABC) transporter motif (3, 30). Moreover, similar high-affinity molybdate transporters have been described for Azotobacter vinelandii, Rhodobacter capsulatus, and Haemophilus influenzae (20, 43, 76). The fact that molybdate supplementation of the growth medium restores molybdoenzyme activity in E. coli mod mutant strains suggests that molybdate is capable of entering the cell through alternate routes, and preliminary experiments indicate that under high molybdate concentration conditions, molybdate can be transported by the sulfate-specific transporter as well as through other nonspecific anion transporters (41, 61). The molybdate uptake kinetics were also analyzed in mod mutant strains, and as expected, mod mutants exhibit much lower rates of molybdate transport as compared to mod* strains (14, 29). Having shown that the modABCD genes code for a molybdate-specific transporter, studies concerned with the regulation of expression of the modABCD operon in E. coli were conducted (24, 48, 58, 61), and it was determined that expression of this operon is

PAGE 17

3 regulated by the intracellular concentration of molybdate. Specifically, high levels of intracellular molybdate resulted in reduction of transcription of the modABCD operon, which implies that E. coli has a molybdate-dependent repressor which regulates expression of the modABCD operon. Isolation of an E. coli strain (strain SE181 1) that did not exhibit the classical repression of the modABCD operon upon molybdate-supplementation of the growth medium, afforded a starting point for the study of the molybdate-dependent repression of the modABCD operon. Determination of the location of the mutation in strain SE181 1 led to identification and subsequent characterization of the putative molybdate-dependent repressor (ModE) of the modABCD operon. This study serves as an in-depth examination of the ModE and its interaction with the modABCD operon, which will ultimately further the understanding of the production of the molybdenum cofactor in E. coli and possibly other organisms as well.

PAGE 18

LITERATURE REVIEW One of the hallmark features of many bacteria is the metabolic diversity of their energy producing systems, and the bacterium Escherichia coli is no exception. E. coli, a facultative anaerobe, is capable of switching among its respiratory pathways depending upon the availability of terminal electron acceptors. A hierarchy of electron acceptors exists in E. coli as determined by their oxidation-reduction potential. Given its large positive redox potential (+ 0.82 V), dioxygen serves as the most efficient electron sink followed by a series of acceptors with incrementally smaller positive redox potentials. When E. coli is undergoing aerobic respiration, the proteins contributing to other less energetically favorable pathways of respiration are not synthesized. However, in the absence of dioxygen, the components of other respiratory pathways may be produced. Depending upon the availability of alternative electron acceptors, nitrate respiratory components may be synthesized followed by those necessary for the utilization of other alternative electron acceptors such as dimethyl sulfoxide, trimethylamine-N-oxide, and fumarate-utilizing respiratory pathways (25). In E. coli, a number of proteins which transfer electrons from a donor to the final acceptor are molybdoenzymes. These include nitrate reductase, dimethly sulfoxide reductase, and trimethylamine-N-oxide reductase and are molybdoenzymes. All of these proteins contain the molybdenum at the site of electron transfer (77). Structural studies 4

PAGE 19

5 have shown that the molybdenum in these enzymes is present as a molybdopterin guanine dinucleotide (MGD) moiety (4, 5, 35, 57). The identification of the molybdenum co factor structure as MGD and the determination of the pathway by which it is synthesized resulted from the work of many individuals presented in a number of studies. One early revelation in this inquiry resulted from the identification of a series of pleiotropic mutations in Aspergillus nidulans each of which caused a deficiency in the molybdoenzymes xanthine dehydrogenase and nitrate reductase (53). This initial finding led investigators to believe that there is a molybdenum cofactor common to molybdoenzymes. It was subsequently demonstrated that the inactive apoprotein nitrate reductase produced by a pleiotropic nit1 mutant of Neurospora crassa could be reconstituted by the addition of denatured preparations of purified molybdoenzymes from animal, plant, fungal, or bacterial sources (34). This finding further indicated that molybdoenzymes have a dissociable molybdenum moiety and also showed that this cofactor was common to a variety of organisms. Since the native molybdopterin cofactor proved extremely labile, initial attempts to determine the structure of the native molybdenum cofactor involved derivatization of the cofactor using vigorous oxidation at 100 C in the presence of iodine gas and potassium iodide (KI), creating derivative Form A or KI oxidation in the presence of air resulting in derivative Form B (36). Analysis of these derivatives led to the identification of the molybdopterin (MPT) structure for the molybdenum cofactor. Later structural studies showed that dinucleotides such as guanine dinucleotide, adenine dinucleotide, cytosine dinucleotide, and hypoxanthine dinucleotide can also be associated with the cofactor in the

PAGE 20

6 native molybdoenzymes (35). In E. coli, all of the molybdoenzymes contain the molybdopterin guanine dinucleotide form of the molybdenum cofactor (57). Determination of the structure of the active form of the molybdenum cofactor facilitated studies of the biosynthetic pathway by which the molybdenum cofactor is produced. A great deal of this work was accomplished in E. coli since a number of pleiotropic molybdoenzyme mutants had been isolated as chlorate-resistant strains (18, 22, 23, 72). Chlorate, a structural analog of nitrate, can be reduced by nitrate reductase to chlorite, and in solution, chlorite exists in the form hypochlorite which is toxic to the cell. The chlorate-resistant mutants that have been isolated can be classified as those which are only deficient in nitrate reductase due to a mutation in the nitrate reductase structural genes {chlC renamed narGHJI; 4, 18, 31, 72, 73) and those which are pleiotropic (denned as chlA, renamed moa; chlB, renamed mob; chlD, renamed mod; chlE, renamed moe; chlG, renamed mog; 69). These pleiotropic mutants are deficient in nitrate reductase and presumably all of the other molybdoenzymes in E. coli (formate dehydrogenase-N, formate dehydrogenase-H, DMSO/TMAO reductase, and biotin-sulfoxide reductase) because of a failure to synthesize and incorporate the molybdenum cofactor into the appropriate apoprotein. However, it has been found that the mog mutants do contain functional molybdoenzyme formate dehydrogenase-H (73). From the characterization of these various pleiotropic chlorate-resistant E. coli mutants, the biosynthetic pathway for MGD synthesis in E. coli was established and is illustrated in Fig. 1 (57). It was found that the biosynthesis of MGD begins with the conversion of guanosine or a guanosine derivative by the moa encoded proteins to a

PAGE 21

Guanosine or Guanosine derivative Converting Factor Large Subunit (MoaE protein) MoaA.B.C Proteins Precursor Z Active Converting Factor Converting Factor Small Subunit (sulfo) MoeB Protein Converting Factor Small Subunit (MoaD protein) 1 Converting Factor-MPT Complex (?) r Intracellular "Activated Mo" MoO< Mod, MoeA Proteins Molybdenum Cofactor (Molybdopterin Form) Mob protein Source of GMP Molybdenum Cofactor (Guanine Dinucleotide Form) Fig. 1. Steps involved in the biosynthesis of the organic portion of the molybdenum cofactor in E. coli (modified from Rajagopalan and Johnson, Ref. 57)

PAGE 22

8 precursor compound, termed precursor Z. Precursor Z joins with active converting factor to produce the converting factor-MPT complex. The active converting factor responsible for the conversion of precursor Z to the converting factor-MPT complex is capable of mediating the conversion only after the small unit of the converting factor (MoaD) protein is first activated by the MoeB protein so that it can then form a complex with the large subunit of the converting factor (MoaE protein), thereby producing the active converting factor. At this point, it is believed that activated molybdenum combines with the molybdopterin molecule. In the presence of guanosine triphosphate, the Mob protein then guanylylates the molybdopterin form of the molybdenum cofactor to the guanine dinucleotide form of the cofactor. Molybdate Transport As has already been indicated, a number of pleiotropic E. coli chlorate-resistant mutants exist, and one type in particular, referred to as mod mutants, is deficient in the uptake of molybdate in the absence of molybdate supplementation of the growth medium (18, 22, 23, 71). Sequencing of the mod genes, located in the 17 minute region of the chromosome, resulted in the identification of a four gene operon designated modABCD. modA codes for a 257 amino acid molybdate-specific periplasmic binding protein; modB codes for a 229 amino acid integral membrane channel-forming protein; modC codes for a 352 amino acid transport energizer protein; and modD codes for a 23 1 amino acid putative outer membrane protein which may function as a porin (33, 47, 58, 60, 75). It should be noted that the molybdate-specific transporter encoded by the modABCD operon resembles the ATP-binding cassette transporter motif in which there is a substrate-specific

PAGE 23

9 periplasmic-binding protein (3, 30), an integral membrane channel-forming protein(s). and an ATP-binding protein which is thought to couple the hydrolysis of ATP to transport. Moreover, when molybdate uptake kinetics were investigated in an E. coli mod mutant strain, the rate of uptake was substantially lower than that seen in the wild type strain, and this transport was shown to be an energy requiring process (14, 29). Other presumptive mod mutant strains transported molybdate as efficiently as the wild type strain, but these strains were not capable of retaining the imported molybdate, as evidenced by a rapid loss of accumulated molybdate when chased with unlabeled molybdate (14). This result implies that the mutation in this second group of mutants is not located in one of the genes coding for the molybdate-specific transporter; it apparently is located in a gene necessary for the incorporation of the molybdenum into the cellular material. Homologs of the E. coli molybdate-specific transport proteins have also been identified in other facultative anaerobes such as H. Influenzae Rd as well as the nitrogenfixing bacteria A. vinelandii and R. capsulatus (20, 43, 76). In the case of the H. influenzae Rd Mod proteins, a 49% identity and 58% similarity for ModA was determined while a 65% identity and 73% similarity for ModB, and a 53 % identity and 66% similarity for ModC were found in relation to the respective E. coli proteins. The A. vinelandii homologs show 27% identity and 38% similarity compared to E. coli ModA, 31% identity and 47% similarity for ModB, and 44% identity and 58% similarity for ModC. The R. capsulatus molybdate transport protein homologs have 29% identity and 40% similarity

PAGE 24

10 for ModA, 29% identity and 45% similarity for ModB, and 39% identity and 55% similarity for ModC compared to the E. coli proteins. Aside from the sequence similarity implicating the A. vinelandii and R. capsulatus homologs' involvement in molybdate-specific uptake, there are also genetic studies which suggest that these proteins do indeed constitute the molybdate-specific transporter in the respective organism. One such study (50) demonstrated that mutations in the modB and modC genes of A. vinelandii allowed for expression of the alternate nitrogenase (heterometal-free nitrogenase) in the presence of 0.5 uM sodium molybdate. Since it has previously been shown that expression of the alternate nitrogenase is repressed in the presence of molybdate in the wild type (43), it may be concluded that these mutations prevent molybdate-specific transport. Furthermore, mutations in modA, modB, or modC ofR. capsulatus resulted in a loss of activity for the molybdenum-containing nitrogenase as well as derepression of the transcription of the genes encoding the heterometal-free alternative nitrogenase, which again indicates a loss of molybdate-specific transport by these mutants (76). In both of these organisms, increasing the concentration of the molybdenum in the medium resulted in repression of transcription of the alternate nitrogenase genes. The fact that mutations in genes encoding the molybdate-specific transporter proteins can be suppressed by the addition of sufficiently high concentrations of molybdate to the growth medium suggests that alternate routes for molybdate transport exist (41, 61). Furthermore, there is evidence that molybdate is capable of being transported through

PAGE 25

11 the sulfate/thiosulfate and selenite transport systems (41,61), which is not surprising given the similarity in charge and structure among molybdate, sulfate, and selenate. Regulation of the modABCD Operon Expression of the modABCD operon has been investigated in a variety of genetic backgrounds by means of monitoring p-galactosidase activities in strains harboring §(modA ''lacZ) (58, 60, 61). From these experiments, it was found that modA 'lacZ expression was high under conditions of low intracellular molybdate as in the case of mod mutants grown in medium lacking molybdate. However, when there is a high enough intracellular molybdate concentration to support nitrate reductase or formate hydrogenlyase activity as is the case in a molybdate transport competent strain or a mod strain grown in molybdate supplemented medium, modA ''lacZ expression is very low. Since molybdate availability seemed to effect expression of the modABCD operon, the possible role of MGD synthesis in mod expression was also explored, and it was found that mutations in the moa, mob, or moe operons did not appreciably change expression of the modABCD operon. However, modA ''lacZ expression was increased approximately twofold when the cells were grown under aerobic conditions or when cultured anaerobically in a strain containing an/w mutation (60, 61). Inspection of the DNA sequence upstream of modA transcription start site did not reveal any putative FNR protein binding motif; consequently, this increase in expression of the modABCD operon in an/w background is most likely a physiological effect of the fnr mutation on other operons like narGHJI. A study by Rech et al. (58) examined the effects of molybdate, nitrate and oxygen on modABCD expression, and again it was found that molybdate limitation enhanced

PAGE 26

modABCD expression, whereas nitrate and oxygen did not significantly modify expression levels. This study also evaluated the DNA sequence in the modA operator/promoter region which might be involved the molybdate-dependent regulation of the modABCD operon. It was determined by mutation analysis that a 'CATAA' sequence located at position +2 to +6 in relation to the modA transcription start site is involved in binding the molybdate-dependent regulator; however, a similar 'CATTAA' sequence located at position +12 to +17 was not required for regulation. Further analysis of the modA operator/promoter DNA revealed that there is an eight base pair (bp) inverted repeat sequence ('TAACGTTA') spanning the +4 to +1 1 region. This inverted repeat may serve as the actual recognition sequence for the molybdate-dependent regulator of the modABCD operon since the inverted repeat sequence overlaps the +2 to +6 sequence that had been implicated in regulator protein binding. Inverted repeat sequences have been shown to constitute the binding sites for a number of repressor proteins such as the MetJ repressor protein and the Trp repressor (40, 54, 55). Although the molybdate-dependent repression of modABCD expression has been reported by several investigators, the causative agent of this repression had not been determined. This study identifies this molybdate dependent repressor, designated ModE. and also locates the recognition sequences in the modA operator/promoter DNA required for binding the repressor. The interaction between the repressor and its target DNA is examined under a variety of conditions.

PAGE 27

MATERIALS AND METHODS Materials Biochemicals were purchased from Sigma Chemical Co (St. Louis, MO). Inorganic and organic chemicals were obtained from Fisher Scientific Co. and were analytical or molecular biological grade. Restriction endonucleases and DNA modifying enzymes were purchased from New England Biolabs, Inc. (Beverly, MA) or Promega (Madison, WI). S30 cell extract was purchased from Promega (Madison. WI). Sequenase 2.0 was obtained from United States Biochemical Corporation (Cleveland, OH). Bacterial Strains Bacterial strains used in this study are presented in Table 1 All bacterial strains used are derivatives of Escherichia coli K-12. Media Bacterial cultures were routinely grown in Luria Broth (LB) which contains 1% trypticase peptone (BBL, Cockeysville, MD), 0.5% yeast extract (BBL) and 0.5% NaCl. When necessary, the LB was supplemented with 0.3% glucose (LBG) and 1 mM sodium molybdate. Lactose-MacConkey agar was prepared by adding filter-sterilized lactose 13

PAGE 28

14 Table 1 Bacterial strains and phages used in this study. Strain or Phage Geneotype Source and/or reference Strains BW545 MC4100 ECL618 RK4353 RK5278 VJS720 VJS1779 VJS1780 VJS1782 VJS1784 BL2UDE3 SE1325 SE1592 SE1595 SE1602 SE1811 SE1910 SE1934 SE1938 SE1940 SE1942 SE1952 SE2069 SE2105 SE2106 SE2107 SE2112 SE2114 SE2116 SE2118 SE4110 A(lacU)J69rpsL araDl39 A(argF-lacU)205 rpsL150 relAl flbB5301 deoCl ptsF25 arcA2 zjjv.TnlO MC4\00gyrA2 19 non-9 RK4353 narL215::TnlO chlD247::Tn!0 (modB247) RK4353 moe-251 ::Tn/0d(Tc) RK4353 mob-252::Tnl0d(Tc) RK4353 moa-254::Tnl0d(Tc) RK4353 mog-2 56::Tnl0d(7c) hsdSgal X[lacUV5-gQnel {11)} cysC43 srl-300::TnlO thr-1 leu-6 thi-1 lacYl galK2 ara-14 xyl-5 mtl-1 proA2 his-4 argE3 rpsL31 tsx-33 supE44 modB138::ln5 BW '545 modA\\5 BW545mo{modA'-'lacT)102 BW545 A.RM37 SE1910 A.RM37 BW545 (AmodF-km)l ARM37 BW545 (AmodEF-km)\ ARM37 BW545 modByJnlO A.RM37 SE2106 modBy.TnlO SE2107 modB-.-.TnlO MC4100 (AmodE-km)2 Laboratory collection CGSC6152 E. Lin V. Stewart V. Stewart V. Stewart V. Stewart V. Stewart V. Stewart V. Stewart F. Studier (74) Laboratory collection Laboratory collection Laboratory collection Laboratory collection This study (24) (24) (24) PI transduction (SE1938 xSE1934) PI transduction (SE1938 xSE1910) PI transduction (SE1938 xSE1325) Laboratory collection R. Ray R. Ray R. Ray R. Ray R. Ray R. Ray R. Ray This study

PAGE 29

Table 1 Cont. Strain or Phage Geneotype Source and/or reference Phages A.RZ5 A.RM26 A.SE1 A.RM37 A.PC50 'bla'lacZ lacT XRZ5[$(modE'-'lacr)] XRZ5[$(modA'-'lacZ + )] XRZ5[(modF'-'lacZ)] XRS45(bla '-lacZJ [0>{narG '-lacT)\ Laboratory collection (24) Laboratory collection R. Ray R. P. Gunsalus

PAGE 30

16 (final concentration 1%) to MacConkey agar base (Difco, Detroit, MI) after autoclaving. Ampicillin (100 ^g/ml), kanamycin (50 ug/ml), tetracycline (15 ug/mL), chloramphenicol (15 ug/ml), and X-gal (40 ug/ml) were included in the media as needed. Enzyme Activities and Culture Conditions P-Galactosidase activities Standing overnight LB cultures were used to inoculate (5% v/v) fresh medium in 13x100 mm screw cap tubes. The cultures in filled tubes were incubated at 37C until late-exponential or early stationary phase had been attained (4 to 6 h). The p-galactosidase activity of Sodium-dodecyl-sulfate (SDS)-chloroform permeabilized cells was assayed by measuring the rate of hydrolysis of o-nitrophenyl-P-Dgalactopyranoside (ONPG) as per Miller (49). The activities are expressed as nanomoles of o-nitrophenol produced per minute per milligram of protein where a value of 350 [xg protein/ml was used as the protein content of 1 ml of cells at an optical density (O.D.) of 1.0 unit in a Spectronic 710 (Rochester, NY) spectrophotometer. FHL activities Formate hydrogen lyase activity was qualitatively determined as previously described (61). Genetic Experiments All genetic experiments were conducted as per standard protocols (49). Molecular Biology Experiments Plasmid c onstructions The plasmids employed in this study are listed in Table 2. Standard published procedures (45) or DNA modifying enzyme suppliers' recommendations were used for plasmid constructions and manipulations.

PAGE 31

Table 2. Plasmids used in this study Plasmid Relevent Genotype Source or reference pACYC184 Cam r Tet r (11) pBR322 Amp r Tet r (8) pTrc99A ?trc lacF Amp' (2) pT7-7 pT7 lacT Amp' J. Maupin pUC19 Amp r (78) pUC4K lacZ' Kan r Amp r Pharmacia pZ1918 'lacZ Amp r H. Schweizer (66) pMAK705 'lacZrep(ts) cam' (26) pFGH15 pAC YC 1 84 {modABCDEFf gal (47) pSE1004 pBR322 (modABCDE)* (41) pSE1009 pUC19 {modABCy (41) pAGl pTrc99A (modET) This study pAGl-A76V pTrc99A (modE76) This study pAGl-T125I pTrc99A (modEl25) This study pAGl-G133D pTrc99A (modE133) This study pAGl-Q216* pTrc99A (modE216) This study pRMl pUC19 (modA + ff) (24) pRM9 pBR322 [modE-km (modABCDy] (24) pRMlO pMAK705 (modE'-km) (24) pRM12 pUC19 (modEFygalE' (24) pRM13 pBR322 {modEFf galC (24) pRM14 pBR322 [§{modE'-'lacZ)] This studv pRM16 pBR322 [modE* modF-km) (24) pRM17 pMAK705 ['modE (modF-km)] (24) pRM22 pT7-7 (modEFY (24) pRM23 pT7-7 (modF) (24) pRM25 pBR322 (A HindlU-Aval fragment) (24) (modEF) 'galE pRM26 pBR322 [Q(modE'-lacZ) modF\ (24) pAM4 pUC19 (-247 to +25 modA operator/ This study promoter) pAM5 pUC19 (-5 to +25 modA operator/promoter) This study pAM6 pUC19 (-7 to +25 modA operator/promoter) This study pAM7 pUC19 (-55 to +25 modA operator/ This study promoter) pAM8 pUC19 (-30 to +25 modA operator/ This study promoter) pAM9 pUC19 (-247 to -55 modA operator/ This study promoter) pAMlO pUC19 (-11 to +25 modA operator/promoter) This study

PAGE 32

Table 2. Contd. Plasmid Relevant Genotype Source or reference pAM13 pUC 1 9 ( mod DNA 1 5 to Accl in modA ) This study pAM15 pUCl 9 [-247 to +25, A(+4 to +1 1 ) modA This study operator/ promoter] pAM16 pUC19 (-243 to +20 modA operator/ This study promoter pAM17 pUC 1 9 (-1 7 to +6 modA operator/promoter) This study pAM18 pUC 1 9 (1 7 to 1 +5 to +1 5 modA operator/ This study promoter pAM19 pUC 1 9 [1 7 to +1 5, A(-6 to +9) modA This study operator/promoter plsgG pGEM3ZF+ [(modF), Haemophilus Apicella influenzae] plsgH pGEM3ZF+ [(modA + ), Haemophilus Apicella influenzae] /

PAGE 33

A (modE ''lacZ) plasmid was constructed by first removing a 2.6-kb KpnlHindlll fragment from plasmid pFGH15 containing the mod'EF,galE' genes and ligating it into the Kpnl-Hindlll sites of plasmid pUC19 to yield plasmid pRMl 1 Since the 5' region of the modE gene was not present in plasmid pRMl 1, a 0.6-kb Kpnl fragment from plasmid pFGH15 containing this 5' region of the modE gene was inserted in frame to construct plasmid pRMl 2. The modEF,galE DNA was then moved to plasmid vector, pBR322 by ligating a 3.2-kb £coRI-//wdIII fragment from plasmid pRMl 2 into the corresponding sites in plasmid pBR322, resulting in plasmid pRM13. Plasmid pRM13 proved to be unstable; however, the removal of the tet gene in the vector DNA as a Hin&lll-Aval fragment from plasmid pRMl 3 to give plasmid pRM25 resolved the instability problem A 3.2-kb Smal fragment encoding the lacZ gene from plasmid pZ1918 (66) was then ligated into the AjTll site (in modE gene) of plasmid pRMl 3 which had previously been modified using the klenow fragment of DNA polymerase I (Klenow), thereby creating plasmid pRM26. Plasmid pRM26 produced p-galactosidase activity from the modE promoter. Subsequent transfer of the <&(modE'lacZ) into the chromosome of strain BW545 was accomplished by first recombining in vivo the bla and modE 'lacZ from plasmid pRM26 with XRZ5 to produce A.RM26. Strain BW545 was then transduced with phage ARM26. In the course of this study, two different modE expression plasmids were constructed. Plasmid pAGl, used for the production of ModE mutants, was created by ligating a Klenow-modified 870-bp Espl fragment containing the promoterless modE gene from plasmid pFGH15 into the Smal site of the ptac-based expression vector

PAGE 34

pTrc99A. The second modE expression plasmid. pRM22, used for purification of ModE. was constructed by ligating an EcoRl-Hindlll fragment which contained the modEF genes from plasmid pRM12 into the ^coRI-Z/mdlll sites of plasmid pT7-7. The modAB containing plasmid, which was used in the in vitro coupled transcription-translation experiments, resulted from the ligation of a 1.7-kb Mscl fragment from plasmid pSE1009 into the Smal site of plasmid pUC19, yielding plasmid pRMl. Plasmid pRMl carries the entire modA gene but a truncated modB gene (modET lacks the C-terminal ten amino acids). A battery of /woc£4-operator/promoter DNA-containing plasmids were constructed for the purpose of in vivo ModE titration experiments. One such plasmid, pAM4, was produced by ligating a Kpnl-BstEll fragment from plasmid pSE1009 (the BstEll end was modified using Klenow enzyme) which covered the -247 to +25 sequence of the modAoperator/promoter region into the KpnI-Hincll sites of plasmid pUC19. The plasmid pAM4 deletion derivative plasmids pAM5 and pAM6 were obtained by nuclease Bal-3 1 treatment of AT/wI-digested plasmid pAM4 followed by religation. The extent of the deletions in plasmids pAM5 and pAM6 was determined by sequencing the DNA. Another derivative of plasmid pAM4, plasmid pAM7, was created by removal of an EcoM-EcoRI fragment (-247 to -56 of modA -operator/promoter region) from plasmid pAM4 and subsequent religation after the ends had been filled in with Klenow. Plasmid pAM8 was constructed by removal of a 217-bp Apol fragment (-247 to -31 of modA -operator/ promoter region) from plasmid pAM4, and plasmid pAM9 was achieved by ligating a Kpnl-Apol fragment (-247 to -55 of modA -operator/promoter region) from plasmid

PAGE 35

pAM4 into the Kpnl-Apol sites of plasmid pUC19. Plasmid pAMlO was produced by digesting plasmid pAM8 with Apol followed by nuclease Bal-3 1 treatment and religation. Again, the extent of the deletion was determined through sequencing the DNA. Construction of plasmid pAM13 relied on polymerase chain reaction (PCR) generation of a 648-bp fragment using the following primers synthesized by National Biosciences, Inc. (NBI, Plymouth, MN): Primer #1, 5* AAGGATCCGTTATATTGTCGCCTAC 3', which contains an engineered BamHl site and is complementary to the -16 to +2 sequence of the modA -operator/promoter and Primer #2, 5' GCTGGCAACTGCGTC 3', which is complementary to sequence located in the 3' end of modA. The 648-bp fragment was digested with BamHl and Accl to give a 634-bp fragment which was ligated into the BamHl/ Accl sites of plasmid pUC19, thereby yielding plasmid pAM13. Plasmid pAM15 was obtained as a result of mung bean nuclease treatment and religation of Psp\406ldigested plasmid pAM4. Plasmid pAM16 resulted from mung bean nuclease treatment of the BstEU end of a Kpnl-BstEU fragment from plasmid pSE1009 which was then ligated into the Kpnl-Hincll sites of plasmid pUC19. Plasmids pAM17, pAM18, and pAM19 were constructed by ligating the oligomers which contained various portions of modAoperator/promoter sequence into vector plasmid pUC19. These DNA oligomers were synthesized by NBI. Complementary oligomers were annealed, polynucleotide kinasetreated and digested with BamHl prior to ligating into the BamHl-Smal sites of plasmid pUC19. The mod-insert DNA in plasmid pAM17 has the following 36-bp sequence 5' CGCAATCGTTATAT TGTCGCCTACATAAGATCCCG 3', while plasmid pAM18's insert has a 45-bp sequence, 5' CGGAATCGTTATATTGTCGCCTTGGAACGTTAC

PAGE 36

22 ATTGGATCCCG 3'. and plasmid pAM 19 contains the following 3 1 -bp sequence 5 'CGGAATCGTTATATTGTTACATTGGATCCCG 3' insert. Hydro xylamine mutagenesis of plasmid DNA. Hydroxylamine mutagenesis of plasmid DNA was carried out essentially as described by Davis et al. (15) with minor modifications. A reaction mixture containing 7.5 ug plasmid pAGl DNA, 40 ul of phosphate-EDTA buffer (0.5 M K-P0 4 pH 6.5; 5mM EDTA), 80 ul of freshly prepared hydroxylamine-hydrochloride solution. pH 6.0 (0.35 g of NH 2 OH-HCl, 0.56 ml of 4M NaOH, 4.44 ml distilled H 2 0), and 50 ul deionized H,0 was incubated at 37C for 18 h. Following the incubation period, hydroxylamine was removed by three successive 1:1 phenol extractions and a 1 : 1 chloroform extraction. At this point, the plasmid DNA was ethanol precipitated, and the dissolved DNA was used to transform competent cells of strain RK4353. Total plasmid DNA isolated from these transformants was then transformed into competent cells of strain SE181 1 for identification of mutant ModE containing plasmids. The resultant transformants were plated on lactose-MacConkey agar with and without ImM sodium molybdate. White colonies were picked from media lacking molybdate, and red colonies were picked from the media supplemented with molybdate. The levels of P-galactosidase activity were determined for the SE181 1 cells containing the mutant plasmids to confirm the presumptive phenotypes. The entire modE sequence was then determined for each selected mutant. DNA Sequencing experiments. All DNA sequences generated in the course of this study were obtained using the Sanger dideoxy method and appropriate primers and plasmids which had been isolated by alkaline lysis procedure and purified by cesium

PAGE 37

23 chloride density-gradient centrifugation (28, 62). For sequencing mutant modE genes, the primers used were synthesized by the DNA Synthesis Core Facility of the Interdisciplinary Center for Biotechnology Research at the University of Florida based on the sequence of the wild type modE DNA. The DNA sequences were analyzed by using computer software programs GCG (6, 13, 16) and Genepro (Riverside Scientific, Seattle, WA). Biochemical Characterization of ModE Expression and purification of ModE. For isolation of ModE protein, a 1L LB culture of strain BL2UDE3(pRM22) was incubated at 37 C with vigorous shaking (250 rpm) until an optical density of 1 .0 was reached. Isopropyl-P-D-thiogalacto-pyranoside (IPTG) was then added to a final concentration of 0.5 mM to induce phage T7-RNA polymerase and high level expression of modE. After incubation for an additional two h., the cells were harvested by centrifugation at 8,300 x g for 10 min. and were resuspended in 8 ml of extraction buffer [50 mM Tris-HCl, pH 7.5; 0.5 mM EDTA; 0.5 % glycerol; ImM dithiothreitol (DTT)]. This cell suspension was passed twice through a French pressure cell (20,000 lb/in 2 ). The broken cell suspension was centrifuged at 100,000 x g for 60 min. to remove cellular debris. The resultant supernatant was loaded onto a 10 ml Q Sepharose fast-flow column (Pharmacia) that had been equilibrated with 50 mM TrisHCl, pH 8.0. Progress of the protein sample through the column was monitored using a chart recorder connected to a UV light (280 nm) absorbance detector. After application of the protein sample to the Q sepharose column, the column was washed with 50 mM TrisHCl, pH 8.0 until a baseline on the UV absorbance trace had been established. The column was subsequently washed with Tris-HCl, pH 8.0 buffer containing 0. 1 M NaCl. The

PAGE 38

column was washed with this buffer until protein no longer eluted from the column as indicated by a return to the baseline on the UV absorbance trace. The column was then washed successively with Tris-HCl, pH 8.0 containing 0.2 M, 0.3 M, 0.4 M, 0.5 M. and 1 .0 M NaCl. Fractions containing the ModE protein eluted with 0.2 M NaCl. The buffer of the ModE-containing protein solution was exchanged to 10 mM phosphate buffer. pH 7, by using an EconoPac-10 DG desalting column (BioRad Laboratories, Hercules, CA) prior to loading the protein on to a 5-ml heparin column (HiTrap; Pharmacia) that had been equilibrated with 10 mM phosphate buffer, pH 7. Bound proteins were stripped from the heparin column by successive washes with 10 mM phosphate buffer, pH 7.0 containing 0.1 M, 0.2 M, 0.3 M, 0.4 M, 0.5 M, and 1.0 M NaCl. The elution procedure proceeded as indicated above for the Q sepharose chromatography. ModE-containing protein fractions eluted at a salt concentration of 0.2 M. At this point, the ModEcontaining protein sample was concentrated using a Centricell-20 spin cartridge (Polysciences, Warrington, PA). The buffer of the ModE solution was exchanged with 25 mM histidine-HCl buffer, pH 6.2 and the protein was applied to a 10-ml polybuffer exchanger 94 column (Pharmacia) equilibrated with 25 mM histidine-HCl buffer, pH 6.2. The column was washed with polybuffer 74, pH 4.0, and the ModE protein was eluted when the pH of the effluent reached approximately 4.5. The ModE protein was stored in 10 mM phosphate buffer10% glycerol, pH 7.0 at a final protein concentration of 0.42 mg/ml. The entire purification process was conducted at 4 C. Since the intrinsic properties of the ModE protein did not lend itself to conventional detection by UV light at 280 nm or detection by the Bradford assay (9), ModE protein was monitored by the

PAGE 39

bicinchoninic acid protein determination assay (70) as well as by electrophoresis in (SDS)polyacrylamide gels, N-terminal amino acid sequencing (performed at the Interdisciplinary Center for Biotechnology Research Protein Chemistry Core Laboratory at the University of Florida) was used to confirm the identity of the purified protein as ModE. Matrixassisted-desorption time-of-flight spectrometric determination of the molecular weight of purified ModE protein was kindly performed by Dr. Preston. In vitro coupled transcription-translation experiment. In vitro transcriptiontranslation experiments using S30 extract and L-[ 35 S]methionine to monitor the expression ofmodA and modB from plasmid pRMl in the presence and absence of ModE and 1 mM sodium molybdate were conducted as suggested by Promega. Three p.g of plasmid pRMl, and when appropriate. 40 pmol of ModE and 1 mM sodium molybdate were included in the S30 reaction mixtures. The 35 S-labeled proteins were separated by electrophoresis through a SDS1 5% polyacrylamide gel. Following completion of the electrophoresis, the resulting gel was transferred to Whatman 3 mm paper and was dried under vacuum. X-ray film (Fuji RX) was then applied to the gel for autoradiographic visualization of the labeled proteins. Quantification of labeled proteins was accomplished by using a phosphorimager (Molecular Dynamics, Sunnyvale, CA). Mobility shift experiments. DNA-mobility shift experiments were performed as described by Fried and Crothers (21) with modification. These experiments used a binding reaction buffer consisting of 10 mM Tris-HCl, pH 7.9, 10 mM MgCl 2 50 mM NaCl, 1 mM DTT, and 5% glycerol. The reaction mixture contained the binding reaction buffer, DNA, and protein in a final volume of 10 pi. The reaction mixtures were incubated at

PAGE 40

37C for 30 min. prior to loading onto 5%-polyacrylamide gels in Tris-borate-EDTA buffer that had been prerun at 100 V at room temperature for approximately 60 min (51). During the electrophoresis, tank buffer was continuously circulated using a Minipuls II pump (Gilson). The migration patterns of the binding reactions in the resulting gels were visualized by autoradiography, and various bands in the gels were also quantified using a phosphorimager. Experiments designed to ascertain the apparent dissociation constant (Kq) for the binding of ModE to its target DNA featured a 43-bp oligonucleotide synthesized by NBI which spans the -18 to +25 region of the modA -operator/promoter sequence. For the DNA-mobility shift experiments, the 43-bp oligonucleotide was end labeled using kinase to phosphorylate the 5* ends with y[ 32 P] from y-[ 32 P]-deoxyadenosine triphosphate (dATP). Unincorporated label was removed by passing the labeled oligonucleotide through a G-25 microspin column (Pharmacia). The reaction mixtures prepared for these experiments contained 0.1 pmol of the 43-bp labeled oligonucleotide as well as 0, 1, 5, 10, 25, 50, 100, 250, 500, and 1000-fold excess of ModE. When appropriate, sodium molybdate at a final concentration of 10 uM or 1 mM was also included in the reaction mixtures. Experiments which had sodium molybdate in the reaction mixtures also contained sodium molybdate in the gels and electrophoresis buffers at the same concentration as that present in the binding reactions. In some experiments, 10 uM sodium molybdate was replaced with sodium tungstate, sodium sulfate or sodium orthovanadate at the same concentration as in the reaction mixtures, gels and tank buffers.

PAGE 41

27 For the DNA-mobility shift experiments designed to determine the concentration of molybdate required for half-maximal binding of ModE, the binding reactions included different concentrations of sodium molybdate ranging from 1 pM to 1 raM as well as 0. 1 pmol 43-bp oligomer and 50-fold excess of ModE. Two polyacrylamide gels were run for this experiment, one with 1 pM sodium molybdate present in the gel and tank buffer and a second gel having 1 uM molybdate present. Binding reactions with 1 pM to 1 uM molybdate were electrophoresed through the 1 pM molybdate gel, and binding reactions having 1 uM to 1 mM molybdate were subjected to electrphoresis through the gel containing 1 uM molybdate. In this study, the measure of the affinity of ModE for its target DNA is reported as an apparent dissociation constant value (K D value). The dissociation constant is defined as the concentration of protein required for 50% binding to the target DNA. Apparent K D values describing binding of ModE to DNA were determined by plotting the concentration of ModE present in the binding reaction versus the resultant percent of shifted DNA. In v itro DNase I footnrinting experiments. The DNase I protection experiments were carried out essentially as described previously (42, 56, 65). For these experiments, a 446-bp Fspl-Hindlll fragment from plasmid pAM4 which carries mo<&4 -operator/ promoter DNA spanning from -247 to +25 was labeled using a32 P-dATP, dCTP, dGTP, and dTTP (NEN, Boston, MA) using Klenow to fill in the Hindlll end. Binding reactions having a total volume of 20 ul were prepared which contained 0.1 pmol of the labeled DNA, 0, 1, 10, 100, or 1000-fold excess ModE, 1 mM sodium molybdate, and the binding reaction buffer used in the DNA-mobility shift experiments. The binding reactions were

PAGE 42

incubated for 30 min. at 37C prior to adding 1.5 ng of DNase I (Sigma). The DNase I treatment continued for two minutes before adding 15 ul stop solution (34 mM EDTA. 6.5 M ammonium acetate). Ethanol (1 10 ul) was then added to precipitate the cleaved DNA fragments. The precipitated DNA was resuspended in 6 ul deionized H 3 0 and 4 ul Sequenase stop solution (USB), and the samples were electrophoresed through an 8% polyacrylamide-TBE-denaturing gel. After drying the gel, autoradiography was used to determine the areas of DNA protected by ModE from DNase I cleavage. Determination of the molecular weight of the ModE-43-bp-DNA complex using native gel electrophoresis. In an effort to determine the stoichiometry of the ModE-DNA association, binding reactions containing 0.1 pmol of a 43-bp oligomer (-18 to +25 region) and 250-fold excess ModE were prepared as described for the DNA-mobility shift experiments. The binding reaction samples were then electrophoresed through a 5.0%, 6.0%, 7.0% or 8.0%-polyacrylamide-TBE-nondenaturing gel as described under DNAmobility shift experiments. The differences in the Rf values [migration of the sample (cm)/ migration of the dye front (cm)] for the ModE-DNA complexes after electrophoresis through the different concentration polyacrylamide gels was then compared to values obtained for protein standards. For the determination of the Rf values of the protein standards, 15 ug of lactalbumin (14,200 Da), 20 ug of carbonic anhydrase (29,000), 20 ug of chicken egg albumin (45,000 Da), 15 ug of bovine serum albumin (66.000 Da monomer, 132,000 Da dimer) and 6 ug of urease (240,000 Da dimer and 480,000 Da tetramer) were electrophoresed along with the ModE-DNA complex in various polyacrylamide-TBE-nondenaturing gels. A plot of the [log(R f x 100)] versus % gel

PAGE 43

29 concentration for the molecular weight standards was then generated, and the slopes derived for each of the standards from this plot were then used to produce a Ferguson plot (19), which relates the negative slope to molecular weight of the protein. The molecular weight of the ModE-43-bp-oligomer complex was determined by extrapolation from the Ferguson plot.

PAGE 44

RESULTS AND DISCUSSION Evidence for a Regulator nf the modARCD Qp ernn Previous work demonstrated that a mutation resulting in loss of function in any one of the proteins which form the molybdate-specific transporter, ModA, ModB, or ModC will cause a loss of activity for the cell's complement of molybdoenzymes (58, 61). This phenotype, which can be suppressed by molybdate supplementation of the growth medium, suggests that the cells containing these mutations are unable to transport molybdate under conditions of low molybdate concentration in the growth medium (22, 29, 48, 61, 67, 71). Restoration of the activities of molybdoenzymes with the addition of 1 mM molybdate to the growth medium also indicates that a sufficient intracellular concentration of molybdate is generated by alternate routes of molybdate transport other than the molybdate-specific transport system in E coli (41, 61). Although a molybdatespecific transport system has been defined in various organisms, the mechanism for the regulation of the genes coding for the components of the transport machinery is not known. A general approach for initial investigation of the regulation of genes of interest in Escherichia coli is to construct lacZ fusions in the appropriate genes and then monitor the resulting p-galactosidase activities under a variety of experimental conditions and number of genetic backgrounds. In this case, the genes of interest comprise the in a 30

PAGE 45

modABCD operon. and E. coli strain SE2069, which contains a chromosomal <$>(modA ''lacZ) fusion was used. Since strain SE2069 has its modA gene interrupted by the lacZ gene fusion, it is functionally a mod mutant strain. Using strain SE2069. it was found that a high level of modA ''lacZ expression, monitored as p-galactosidase activity, was only achieved in cells grown in LBG, a medium which contains trace amounts of contaminating molybdate (Table 3). Therefore, the fact that high level expression of modA 'lacZ'xs only seen in the absence of added molybdate intimates that the modABCD operon is subject to repression only when sufficient intracellular levels of molybdate are present. If this is indeed the case, a mo(T E. coli (strain BW545) should transcribe the modABCD operon at very low levels. In order to test this possibility, the modA 'lacZ fusion from strain SE2069 was transferred to phage (ASE1) and incorporated into the X att site of the chromosome. Low levels of expression of modA 'lacZ in the wild type mod" strain, strain BW545(ASE1), regardless of supplementation of molybdate to the growth medium (Table 3), implies that the functional molybdate-specific transporter brings in sufficient levels of molybdate from the LBG medium to repress the modABCD operon. In the presence of a mutation in any of the first three mod genes (undefined point mutation in a given mod gene), derepressed levels of modA ''lacZ from A.SE1 result only in the absence of molybdate supplementation of the growth medium. Further information concerning the nature of the molybdate-dependent repression of the modABCD operon was afforded by complementation studies of a spontaneous

PAGE 46

32 Table 3. Effect of mod gene mutation in mod operon transcription Strain mod P-galactosidase Activity 3 Genotype -Mo +Mo SE2069 tyimodA 'lacZ) 2,400 250 BW545(ASE1) mod' 20 <10 SE1592(ASE1) modA 2,330 165 SE1595(ASE1) modC 1,980 150 SE1602(ASE1) modB 1,860 60 "Cells were grown in LBG medium supplemented with 1 mM sodium molybdate when specified, p-galactosidase activities are expressed as nmolesminute" 1 milligram of cell protein" 1

PAGE 47

mutant of strain of SE2069, designated as strain SE 1 8 1 1 The level of modA 'lacZ expression by strain SE181 1 remained high even in the presence of 1 mM molybdate, suggesting that this mutation inactivated the gene responsible for the molybdate-dependent repression of the modABCD operon (Table 4). By phage PI -mediated transduction, the mutation which derepressed modA expression in strain SE181 1 was mapped close to modA. The identification of the gene altered in strain SE181 1 was accomplished through the introduction of plasmids containing different regions of the mod DNA into strain SE 1 8 1 1 As presented in Table 4, the introduction of plasmid pSE1004 which harbors modABCD DNA as well as an open reading frame diverging from this operon into strain SE181 1 resulted in both the restoration of molybdate transport as indicated by the return of FHL activity in the absence of added molybdate in strain SE181 1 as well as the restoration of molybdatedependent repression of the modABCD operon in this strain. Introduction of plasmid pSE1009 which has modABC DNA only restored the molybdate-specific transport in strain SE181 1 as seen by the presence of FHL activity but did not restore the molybdatedependent repression of modA ''lacZ expression in this strain. As expected both plasmids complemented the modA mutation in parent strain SE2069. It should be noted that there is a disparity in the P-galactosidase activity values produced by strain SE2069 (Tables 3 and 4) which is most likely the result of allelic variations in strain SE2069. The higher value from Table 3 was obtained approximately 2 years prior to the value shown in Table 4, while the value from Table 4 is characteristic of values which are now routinely obtained for strain SE2069.

PAGE 48

34 Table 4. Complementation of the mutation in strain SE 1 8 1 1 for molybdate-dependent repression of the modABCD operon Strain Relevant Genotype P-galactosidase Activity 2 FHL b -Mo +Mo SE1811 SE1811(pSE1004) SE1811(pSE1009) SE1811(pAGl) SE2069 SE2069(pSE1004) SE2069(pSE1009) (modA-lacZ), modE Q{modA-lacZ), modE. (modABCDEY <5>{modA-lacZ), modE, {modABCf ^(modA-lacZ), modE, {modEY $(modA-lacZ) Q(modA-lacZ), {modABCDEY Q(modA-lacZ), (modABQ* 1,500 1,500 290 600 1,300 300 300 260 1,500 1,500 100 200 300 300 + + "Cells were grown in LBG medium. Sodium molybdate concentration was 1 mM. 0galactosidase activities are expressed as nano molesminute" 1 milligram of cell protein" 1 b -, absent; +, present.

PAGE 49

The DNA present in plasmid pSE1004 but not in plasmid pSE1009 was subcloned into plasmid vector pTrc99A to yield plasmid pAGl. The introduction of this plasmid into strain SE181 1 resulted in full restoration of the molybdate-dependent repressor activity as shown by the low level of modA ''lacZ expression for strain SE181 l(pAGl) with the addition of 1 mM sodium molybdate to the growth medium. The lower level of modA 'lacZ expression for SE181 l(pAGl) in the absence of added molybdate, (600 units of pgalactosidase activity, as opposed to 1,500 units for strain SE181 1), is most likely due to the presence of multiple copies of the plasmid in each cell. Presumably, when there is this higher dosage of repressor in strain SE181 1 cells as a result of the presence of plasmid pAGl, any trace amount of molybdate present in the cell would serve to enhance the partial repression of the modA 'lacZ seen under these conditions. Based on the results presented in Tables 3 and 4, it can be concluded that there is a molybdate-dependent repressor which acts to limit expression of the modABCD operon when a sufficient intracellular concentration of molybdate is sensed. The gene that presumably codes for this repressor is designated modE, and the modE gene is located adjacent to the modABCD operon but is transcribed in the opposite direction compared to the modABCD operon. Analysis of modEF DNA modE gene. The region of DNA which complemented the mutation in strain SE181 1 was sequenced, and a two-gene operon starting 444 bp upstream of the modA translation start site was identified (Fig. 2; modified from 60). These two open reading

PAGE 50

modA CGAGCCATTGGTAACCCCTTAATGTAACGTTATGTAGGCGACAATATAACGACTAACTCGGAATTTTCCCAGCAGTTATTGCTAACCTTT 90 RAM GTGAGGTAGATAAGAAAAAATATCGGCAGGAAAAGCAGGAAGTTGAGAAAAAGAAAATGCCCGACTAAGCGGGCATTCAGGGAATCAATG 180 > < ATTTTGTCCGGGCTGGTCTTTTTTACCAACACCAGAAAAGATGTTGAATACTTCACCAAGACCGTAAATCAGACCCAGGATGATGGCCAT 270 CArGACAGGTACCATGATTACGGCGAATACCAGACTTTTTAATAACTCTAACATGGTCAAC TCCAGA TATAGTCATGAGAC TATTCTA AC 360 CGCTAAGCACAGAAAAGCACTCCCCTTTTGTGCGGTCAGCTTTGCGTGGCGTTCGTTTTCCGTCACAATAAGACTTTTGCC AGGAC ATTG 450 mod£ • UacZ in pRM14 and pRM26) TTATGCAGGCCGAAATCCTTCTCACCCTTAAGCTCCAACAAAAATTATTCGCCGACCCGCGCCGCATTTCGCTACTAAAACACATTGCGC 54 0 MQAEILLTLKLQQKLFADPRRISLLKHIAL TTTCCGGTTCCATTAGCCAGGGAGCGAAAGATGCCGGTATTAGCTATAAAAGCGCCTGGGATGCCATTAACGAGATGAATCAGTTAAGTG 630 SGS I SQGAKDAGI SYKSAWDAINEMNQLSE AGCATATTCTGGTCGAGCGCGCAACAGGCGGTAAAGGTGGCGGCGGCGCAGTACTGACCCGCTATGGTCAGCGACTGATTCAGCTCTATG 720 HI LVERATGGKGGGGAVLTRYGQRLIQLYD ACTTACTGGCGCAAATCCAGCAAAAAGCCTTTGATGTGTTAAGTGACGATGACGCCCTGCCGCTGAACAGCCTGCTGGCCGCGATCTCAC 810 LLAQIQQKAFDVLSDDDALPLNSLLAAISR GTTTTTCACTGCAAACCAGCGCCCGTAACCAGTGGTTCGGTACCATCACCGCCCGCGATCATGATGACGTTCAACAGCATGTTGATGTCT 900 FSLQTSARNQWFGTITARDHDDVQQHVDVL TACTGGCTGACGGAAAAACACGCCTGAAAGTCGCAATTACCGCACAAAGCGGCGCGCGTCTGGGGCTGGATGAAGGCAAAGAAGTGTTGA 990 LADGKTRLKVAITAQSGARLGLDEGKEVLI TATTGCTAAAAGCGCCGTGGGTAGGTATTACTCAGGACGAGGCGGTCGCGCAAAACGCTGACAACCAATTACCGGGTATTATTAGTCATA 1080 LLKAPWVG I TQDEAVAQNADNQLPG I ISHI TTGAGCGCGGCGCAGAGCAGTGCGAAGTATTAATGGCGCTACCCGACGGGCAAACACTGTGCGCCACAGTGCCGGTAAATGAAGCGACTT 1170 ERGAEQCEVLMALPDGQTLCATVPVNEATS CTCTTCAGCAAGGACAGAATGTCACGGCCTACTTTAATGCCGACAGCGTGATTATCGCCACGCTGTGCTAAGCGTGTTGACAATTTGTTA 1260 LQQGQNVTAY FNADSVI I A T L C modF TGAAACACGTATCCCTGTCAGTAATCGCTGCACAAAGT GGGA TATAAAATGTCATCGTTGCAAATTTTGCAAGGCACGTTTCGTCTTAGC 1350 MSSLQILQGTFRLS GACACAAAAACGCTGCAATTGCCTCAGCTAACGTTAAACGCGGGTGATAGTTGGGCGTTTGTCGGTTCGAATGGAAGCGGGAAATCGGCC 1440 DTKTLQLPQLTLNAGDSWAFVGSNGSGKSA CTGGCCCGCGCGCTGGCGGGGGAACTTCCGCTTTTGAAAGGTGAACGGCAAAGCCAGTTTTCCCACATCACTCGTCTCTCCTTCGAGCAA 1530 LARALAGELPLLKGERQSQFSHITRLSFEQ TTGCAAAAACTCGTCAGCGACGAATGGCAGCGGAATAACACCGATATGCTCGGCCCTGGCGAAGATGACACCGGACGCACTACGGCTGAG 1620 LQKLVS DEWQRNNTDMLG PGE DDTGRTTAE ATCATTCAGGATGAAGTAAAGGATGCACCGCGTTGCATGCAACTGGCGCAGCAGTTCGGTATTACCGCCCTCCTCGACCGACGCTTTAAA 1710 I I QDEVKDAPRCMQLAQQ FG I TALLDRR FK TACCTTTCCACTGGCGAGACGCGAAAAACCCTGCTGTGTCAGGCGCTGATGTCGGAGCCTGACTTGTTGATTCTTGATGAGCCGTTCGAT 1800 YLSTGETRKTLLCQALMSEPDLLILDEPFD GGCCTGGATGTTGCCTCACGTCAGCAGCTGGCTGAGCGACTCGCCTCGTTACATCAGTCCGGTATTACTCTGGTACTGGTGCTCAATCGC 1890 GLDVASRQQLAERLASLHQSGITLVLVLNR TTCGATGAGATCCCGGAGTTTGTCCAGTTTGCTGGCGTGCTGGCGGATTGCACGTTAGCGGAAACTGGCGCTAAAGAGGAACTGCTCCAA 1980 FDEI PEFVQFAGVLADCTLAETGAKEELLQ CAAGCACTCGTCGCGCAACTGGCGCATAGTGAACAGCTTGAAGGTGTGCAACTGCCGGAGCCGGATGAACCTTCAGCACGTCACGCCTTA 2070 QALVAQLAHSEQLEGVQLPEPDEPSARHAL CCCGCCAACGAACCGCGCATTGTGCTGAACAATGGCGTGGTTTCTTATAACGATCGCCCCATTCTTAATAACCTTAGCTGGCAGGTGAAT 2160 PANE PR IVLNNGVVSYNDRPI LNNLSWQVN CCAGGCGAACACTGGCAAATTGTCGGGCCAAATGGTGCAGGAAAATCGACGTTATTAAGCCTGGTTACTGGCGATCATCCGCAAGGTTAC 2250 PGEHWQIVGPNGAGKSTLLSLVTGDHPQGY AGCAACGATTTGACGCTTTTCGGACGACGTCGCGGCAGCGGCGAAACCATCTGGGATATCAAAAAGCATATCGGTTACGTCAGCAGTAGT 2340 SNDLTLFGRRRGSGETIWDIKKHIGYVSSS TTGCATCTGGATTACCGGGTCAGCACTACCGTGCGTAATGTGATTCTTTCTGGCTATTTTGATTCGATTGGCATTTATCAGGCCGTTTCG 2430 LHLDYRVSTTVRNVILSGYFDSIGIYQAVS GATCGCCAGCAAAAACTGGTGCAGCAGTGGCTGGATATTCTCGGCATTGATAAACGCACGGCTGACGCTCCGTTCCATAGTCTTTCCTGG 2520 DRQQKLVQQWLDILGIDKRTADAPFHSLSW GGACAGCAGCGTCTGGCGCTGATTGTCCGCGCACTGGTGAAACATCCGACGTTGCTTATTCTCGATGAACCACTACAGGGGCTTGATCCG 2610 GQQRLALIVRALVKHPTLLILDEPLQGLDP CTGAATCGCCAGCTTATCCGCCGTTTTGTTGATGTGCTGATTAGCGAAGGTGAAACGCAATTGTTGTTTGTTTCGCACCACGCTGAAGAT 2700 LNRQLI RRFVDVLISEGETQLLFVSHHAED GCGCCTGCCTGTATTACCCATCGTCTTGAGTTCGTGCCGGACGGTGGACTCTATCGCTATGTGCTGACAAAAATATATTGAGTCGGTAGT 2790 APACITHRLEFVPDGGLYRYVLTKIY* GCTGACCTTGCCGGAGGCGGCCTTAGCACCCTCTCCGGCCAACGGTTCGACGCATGCAGGCATGAAACCGCGTCTTTTTTCAGATAAAAA 2880 GCGCAATCATTCATAAACCCTCTGTTTTATAATCACTTAATCGCGCATAAAAAACGGCTAAATTCTTGTGTAAACGATTCCACTAATTTA 2970 galE TTCCATGTCACACTTTTCGCATCTTTGTTATGCTATGGTTATTTCATACCATAAGCCTAATG 3033 Fig. 2. DNA sequence of the modEF operon. Amino acid sequences are listed below the DNA sequence. Presumptive -10 and -35 sequences are underlined and ribosomal binding sites are double underlined. Asterisks denote translation terminations. The location of the hcZ insertion in the modE gene in plasmid pRM26 is indicated by the down arrowhead. The stem-loop structure is shown by the two opposing arrows. ATP-GTP binding motifs in ModF are in boldface type.

PAGE 51

37 frames, designated modE and modF, are transcribed in the opposite direction as compared to modA and encode a 262-amino-acid protein (28,200 Da) and a 490-amino-acid protein (53,900 Da), respectively. An eight-bp-stem six-bploop can be identified in the DNA between the operator/promoter regions of modA and modE and is indicated by the two opposing arrows in Fig. 2. This DNA is located between gal and modABCD operons in the E. coli chromosome, and the galETK operon is transcribed in the same direction as the modEF operon. A search for proteins that bear similarity to ModE was conducted resulting in the detection of four homologs (Fig. 3) The homologs were identified as ModA from A vinelandii (35% identity, 45% similarity) (43) (renamed ModE [50]), ModE from H. influenzae Rd (48 % identity, 58% similarity) (20), MopA from R. Capsulatus (34% identity, 46% similarity) (76), and MopII from C. pasteurianum (29% identity, 46% similarity) (32). A possible helix-turn-helix motif which is a known prokaryotic DNAbinding protein structure motif (52) has been identified in the N-terminal part of the E. coli ModE as indicated by the underlined sequence in Fig. 3. Since H. influenzae ModE is almost 60% similar to E. coli ModE and both are facultative anaerobes, it was thought that the H. influenzae ModE homolog may be able to substitute for the E. coli ModE protein in the ModE deficient E. coli strain SE181 1. This possibility was tested by monitoring the modA ''lacZ expression in strain SE181 1 which had been transformed with a plasmid containing the H. influenzae modE gene

PAGE 52

38 AvModE MT AT R FLARMS L DT D VG — TALSDTRIRLLEAIEREGSINRAAKWPLSYKAAWDAIDTMNN 6C EcModE MQAEILLTLKLQQK — LFADPRRISLLKHIALSGSISOGAKDAGISYKSAWDAINEMNO 57 HiModE MKNTEILLTIKLQQA— LFIDPKRVRLLKEIQQCGSINQAAKNAKVSYKSAWDHLEAMNK 56 RcMopA 60 AvModE LAPEPLWRVAGGRQGGGTQLTDYGRRIVAMYRALEIEYQSALDRLSERLNEVTGGDIQA 120 EcModE LSEHILVERATGGKGGGGAVLTRYGQRLIQLYDLLAQIQQKAFDVLSDDDALPLNSLLAA 117 HiModE ISPRPLLERNTGGKNGGGTALTTYAERLLQLYDLLERTQEHAFHILQDE-SVPLDSLLTA 117 RcMopA LFEQPLVEAAPGGRTGGNARVTEAGQALIAGFGLLEGALTKALGVLEGGVSAPEKALNTL 120 AvModE FQRLMHSMSMKTSARNQFAGIVTGLRVGGVDYEVRIRL-DAENEIAAVITKASAENLELA 179 EcModE ISRF SLQTSARNQWFGTITARDHDDVQQHVDVLLADGKTRLKVAITAQSGARLGLD 173 HiModE TARF SLQSSARNQFFGRVAQQRIIDSRCWDVNVQGLPTPLQVSITTKSSARLKLI 173 RcMopA WSL TMRTSNRNTLRCTVTRVTLGAVNAEVELALTDGHS-LTAVITERSATEMGLA 174 CpMopII MS I SARNQLKGKWGLKKGWTAE WLE I AGGN-KITS 1 1 SLDSVEELGVK 50 AvModE IGKEVFALVKSSSVMLTTEPSLKL-TARNQLWGEVIDIHEGPVNNEVTLALPSGRSVTCV 238 EcModE EGKEVLILLKAPWVGITQDEAVAQ-NADNQLPGIISHIERGAEQCEVLMALPDGQTLCAT 232 HiModE TEKEVMLMFKAPWVKISEQPLA NQPNQFPVNIKSLN EEEAILQFAESNIEFCAT 219 RcMopA PGVEVFALIKASFVMLAAGGDPGRISACNRLTGIVAARTDGPVNTEIILDLGNCKSITAV 234 CpMopII EGAELTAWKSTDVMILA 68 AvModE VTADSCKALGLAPGVAACAFFKSSSVILAVYG 270 EcModE VPVN — EATSLQQGQNVTAYFNADSVIIATLC 262 HiModE V-H QPNQWQIEQQVWIHIDQEQIILATLG 255 RcMopA I THTS ADALGLAPGVPATAL FKASHVI LAMP 265 Fig. 3. Amino acid sequences of the homo logs of E. coli ModE protein. AvModE, A. vinelandii ModE (43); EcModE, E. coli ModE; HiModE, H. influenzae Rd (20); RcMopA, R. capsulatus ModE (76); CpMopII, C. pasteurianum MopII (31). Double dots represent identity, and single dots represent conservative substitutions. Shadow print indicates the conserved SARNQ sequence. All identity and similarity designations are in relation to E. coli ModE. A presumptive helix-turn-helix DNA binding region in E. coli ModE is underlined.

PAGE 53

39 (plasmid plsgG). A low level of modA ''lacZ expression was seen for strain SE181 1 (plsgG) that was cultured in LBG supplemented with 1 mM sodium molybdate (Table 5) verifying that the H. influenzae ModE protein can functionally substitute for E. coli ModE. modF gene. Expression of the modF gene from a T7-based expression vector, plasmid pRM23, yielded a 54 kDa protein when visualized after SDS-PAGE. This size is in agreement with the expected size based on the DNA-sequence derived amino acid sequence. Database searches revealed no homologs of E. coli ModF protein, but further examination of the amino acid sequence identified two ATP-GTP binding site motifs (Fig. 2). As of yet, the function of modF has not been determined, since deletions of the modF gene did not result in a detectable phenotype for FHL production or modA gene expression or activity under the conditions used in this study. A study by Dorrel et al. (17) has described and presented the sequence of a gene involved in photoreactivation iphrA). This DNA sequence is situated within the modF gene presented in Fig. 2. Based on their studies, Dorrel et al. proposed that the 38 kDa PhrA protein is necessary for photorepair. However, the findings that a 54 kDa protein and not a 38 kDa product was observed when modF was expressed in a T7 RNA polymerase-driven expression system and that the expression of <$>(modF'lacZ) responds to regulation by ModE (data presented later) suggest that the DNA in question codes for modF.

PAGE 54

40 Table 5. Complemetation of the modE mutation in strain SE181 1 with H. influenzae modE. Strain Relevant genotype P-galactosidase Activity 3 -Mo +Mo SE1811
PAGE 55

Regulation of modA by ModE In order for the interaction of ModE with its target operon modABCD to be investigated, the ModE protein first had to be purified. ModE protein was overexpressed and purified as indicated in "Materials and Methods" section. The protein yield after each purification step is shown in Table 6, while the purity of ModE after each purification procedure is presented in Figure 4. A final yield of purified ModE protein of 5% of total protein in the extract was obtained, and this protein was judged to be pure by native polyacrlyamide electrophoresis (not shown), SDS-PAGE (Fig. 4), and N-terminal amino acid sequencing of the protein. The first twelve amino acids are MQAEILLTLKLQ and these correspond to the predicted amino acid sequence of the protein (Fig. 3). The molecular mass of the purified protein was determined to be 28,271 Da using matrixassisted laser desorption ionization time-of-flight mass spectrophotometry. This mass value is in agreement with the predicted value of 28,200 Da based on the DNA sequencederived amino acid sequence as well as a value of 29,000 Da obtained after SDS-PAGE. Coupled in vitro transcription-translation experiments. Experiments in which the P-galactosidase actvity of modA ''lacZ was monitored in two isogenic strains, one having a functional modE gene (strain SE2069) and the other carrying an inactivating mutation in modE (strain SE181 1) suggested that ModE represses transcription of modA ''lacZ and presumably the modABCD operon in a molybdate-dependent fashion (Table 4). A direct approach to investigating the interaction of ModE with its target DNA is to determine the

PAGE 56

Table 6. Purification profile of ModE Fraction Total Purification protein (fold) (mg) Extract 238 1.0 Q Sepharose fast flow 57 4.2 HiTrap Heparin 16.5 14.4 Chromato focusing 12 19.8

PAGE 57

43 1 2 3 4 5 Fig. 4. SDS-PAGE analysis of proteins from different stages of ModE purification. Lane 1, molecular weight markers (from top: phosphorylase b, 97,400; bovine serum albumin. 66,200; ovalbumin, 45,000; carbonic anhydrase, 31,000; trypsin inhibitor, 21,500; lysozyme, 14,400); lane 2, extract; lane 3, proteins after Q Sepharose fast-flow purification; lane 4, proteins after HiTrap Heparin purification; lane 5, proteins after chromatofocusing. A 25-ug amount of protein was loaded into each of lanes 2-4. A 10ug amount of protein was loaded into lane 5.

PAGE 58

level of expression of target genes (modAB ') with and without the effector (ModE) in an in vitro coupled transcription-translation experiment with S30 extract. For these coupled transcription-translation experiments, plasmid pRMl DNA, containing modAB was incubated with E. coli S30 extract with and without 1 mM sodium molybdate and 40 pmol of ModE. In these experiments, the molybdate concentration was maintained at 1 mM since the affinity between the protein and molybdate is not known. Also, it is not known whether the Mo species bound to ModE is molybdate or some other derivative of molybdate. It should be noted that the S30 extract provides all the constituents necessary for production of proteins from a DNA template and that addition of 35 S-methionine allows for labeling of proteins produced in the reaction mixture. The resultant autoradiogram (Fig. 5) and derivative Phosphorimager data indicate that the presence of ModE in the reaction mixture reduced production of the 26kDa ModA protein and the 24-kDa ModB' protein by 48 and 65%, respectively (lanes 1 and 2). In Fig. 5, two bands are labeled as ModA. ModA, which has been shown to function as a molybdate-binding protein (59), has a putative 24 amino acid leader signal peptide thought to be responsible for its localization in the cell's periplasmic space (60). Given that ModA has a signal peptide, the upper band would be the unprocessed modA while the lower band would be ModA without its leader peptide. With the addition of sodium molybdate, repression of modAB expression was even greater with over 90% decrease in production of ModA and ModB' (Fig. 5, lanes 3 and 4). The decrease in production of ModA and ModB' even in the absence of added sodium molybdate is most likely due to the presence of contaminating molybdate in the S30 extract. It should also

PAGE 59

45 Bla Mod A { ModB 12 3 4 Fig. 5. SDS-PAGE analysis of proteins produced in an in vitro transcription-translation experiment in which plasmid pRMl served as the template DNA. Sodium molybdate and ModE concentrations were 1 mM and 40 pmol per reaction mixture, respectively. Lane 1 without ModE; lane 2, with ModE; lane 3, without ModE, with sodium molybdate; lane 4, with ModE and sodium molybdate. Bla, P-lactamase. The upper ModA protein band represents the precursor protein bearing the signal peptide, while the lower ModA protein band represents the protein with the signal peptide removed. ModB, truncated ModB protein.

PAGE 60

46 be pointed out that, although equal amounts of radioactivity were loaded in each lane, the samples in lanes 2 and 4 appear to contain a total of less labeled protein than is present in the other lanes. This disparity in the amounts of labeled protein is due to the presence in these samples of low molecular weight (partially degraded?) protein products which migrate faster than the dye front during electrophoresis. Addition of molybdate alone did not significantly reduce the amount of ModA and ModB' produced in the reaction mixture. Thus, the reduction in the production of ModA and ModB' proteins with ModE in the reaction mixtures confirms that ModE acts as a repressor of the modABCD operon and that this repression is enhanced by molybdate. In an effort to further show that ModE was binding to the modA operator/ promoter DNA and thereby reducing expression of the operon, a 42-bp DNA fragment, believed to encompass the ModE target binding region, was added to the transcriptiontranslation reactions along with the ModE protein in order to titrate out ModE and relieve the repression of modAB '. The added DNA is a 42-bp DNA derived from the operator/promoter region of modA spanning -9 to +33. With the addition of 50-fold more 42-bp oligomer as compared to plasmid pRMl in the reaction mixtures, production of ModA was restored to 66% of the levels seen without ModE, and inclusion of 100-fold excess of 42-bp oligonucleotide fully restored production of ModA and ModB' to the levels seen in the absence of ModE (data not shown). Clearly, even though the DNA used to compete for the binding of ModE carried only a part of the operator/promoter region of the modABCD operon, the 50-fold and 100-fold excess of this DNA compared to the

PAGE 61

template DNA present in the reactions was sufficient to eliminate efficient binding of ModE to plasmid-borne target DNA. Characterization of the Interaction of ModE with modA Operator/promoter DNA Determination of ModE's target binding site sequences using in vivo titration experiments. The rationale for this series of experiments aimed at identifying the region of DNA in the modA operator/promoter region required for successful binding of ModE protein is based on the idea that the presence of a large number of copies of DNA, in the form of plasmid DNA. containing the repressor's binding site(s) should result in the binding of a certain fraction of the repressor molecules produced from the chromosomal copy ofmodE. With the repressor molecules bound to the plasmid DNA, and thus not available as free protein, repressor binding at its target site on the chromosome should be reduced. This reduction of binding to the chromosomal ModE-binding site can be monitored using an appropriate modA ''lacZ fusion in the chromosome. More specifically, various lengths of DNA ranging from -247 to +25 of the modAoperator/promoter region were cloned into the multiple cloning site of the high copy plasmid vector pUC19, as described in "Materials and Methods" section. The resultant plasmids were subsequently transformed into strain SE2069, which contains a chromosomal <$>(modA 'lacZ). An initial inspection of the modA operator/promoter DNA suggested that perhaps an eight-bp inverted repeat (TAACGTTA) spanning the +4 to +1 1 region might be involved since inverted repeats had been shown to function as the binding sites for the trp and met repressors (40, 54, 55). However, a closer inspection of the modA operator/

PAGE 62

promoter DNA revealed that there is a preponderance of GTTA sequences in this DNA (Fig. 6). Specifically, there are 8 GTTA tetramers between the translation start sites of modA and modE and all are located near the modA end. It is especially notable that two of the GTTA tetramers straddle the -10 region, and that there is a GTTA tetramer inverted repeat (the eight-bp inverted repeat referred to above) located 4-bp downstream of the transcription start site. Thus, the number and location of these GTTA tetramers provided compelling reasons for investigating their potential role in the binding of ModE. For determination of the degree of titration of ModE by the various plasmids, 0galactosidase assays were performed as per standard protocol, and ampicillin was added to the culture medium to ensure the maintenance of the plasmids in the cells. Supplementation of 1 mM molybdate to the culture medium was also required since strain SE2069 cells are functionally Mod"; thus, in the absence of added molybdate, ModE fails to serve as a repressor of modA 'lacZ expression (Table 3). The resulting 0galactosidase activities are presented in Table 7. The mod DNA sequence that is present in each of the plasmids is also indicated in Table 7. It should be noted that the data is presented in a manner between plasmids pAM7 to pAM5 that the first P-galactosidase activity corresponds to the construct bearing all of the GTTA tetramers (plasmid pAM7), while each following plasmid has successive GTTA sequences removed from left to right, respectively, until plasmid pAM5 is reached. The level of modA ''lacZ expression for strain SE2069 with plasmid pAM4, which contains the entire intergenic DNA sequence shown in Fig. 6, is 510 p-galactosidase activity units while expression for SE2069 cells alone or for SE2069 harboring plasmid

PAGE 63

49 T A T G C T G-C C-G C-G C-G modE ^ ~jj Elll276nt GATTCCCTGA -TTTCTTTTTCTCAACTTCCTGC TTTTCCTGCCGATATTTTTTCTTATCTACCTCACAAAGgJgGCAA -35 -10 g^TGC iTflftftAAl AATTCCGAgJgGTCgJj^2iiB rCGC modA CT liTcfjgggggCATTAAGGGgjgCCA fAT^ Fig. 6. Intergenic region of modEA DNA. Boxes outline the modA -1 0 and -35 sequences. The modA transcription start site is indicated by the boxed 'A' with the arrow, and the translation start site appears as the boxed 'ATG' or 'CAT'. 'GTTA' sequences are shown in reverse print.

PAGE 64

50 Table 7. Titration of ModE protein in vivo by various DNA sequences in strain SE2069 (modA'-'lacZ + ) Plasmid DNA Sequence /S-Galactosidase Activity No Plasmid pAM4 pAM9 pAM7 pAM8 pAM13 pAMlO pAM6 pAM5 PAM15 pAM16 GGTACC186nt CAAAGgjjgGCAA GAgJgGTC2i2 T ISSiE TCGCCTAC ~ 3 4 5 GGTACC186nt CAAAGgJgGCAA GAgj^GTCggJrgJj^CGCCTAC. ~ GAG gGT CGCCTACA CGCCTACA TggjgTCGCCTACA gTCGCCTACA CGCCTACA TAACGTT TAACGTT TAACGTT TAACGTT TAACGTT GCTGGGAAAATTCC ATTAAGGGjjggC 8 GCTGGGAAAATTCC ATTAAGGGgJgC AAATTCCJCATTAAGGGgJgc JCATTAAGGGgJgC gCATTAAGGG^^C gCATTAAGGGgjJgC gCATTAAGGGgggC G GTACC 1 86nt CA AA gaSSJJgtcS EtEEHEtc G GTACC 1 86nt CAAA GAgJgGTCgJgTgggTC' CAfT GCTGGGAAAATTCC CATTAAGGGgggC GCTGGGAAAATTCC ATTAAGGG C 150 510 140 510 710 750 150 120 60 90 410 Cells were grown in LBG medium with molybdate (1 mM). /3-galactosidase activities are expressed as nanomoles.minute '.milligram of cell protein 1 The putative ModE protein binding sites are indicated by inverse print and are numbered (pAM4).

PAGE 65

pAM9 which contains only the 5' 192-bp of the intergenic region is 150 and 140 units, respectively. These data indicate that the modEA intergenic region is capable of titrating a fraction of ModE. thereby relieving repression of the chromosomal modA 'lacZ and that the first 192-bp of the insert in plasmid pAM4 is not necessary for this titration. Based on the results presented in Table 4, a P-galactosidase activity of 1,300 units is expected for strain SE2069 under conditions of derepression; thus, the relief of repression of modA ''lacZ by plasmid pAM4 seen in the titration experiments is not total, but is of sufficient level to discriminate the sequence required for the binding of ModE. Removal of the first two GTTA tetramers resulted in an increased titration of ModE by the plasmid DNA as shown by the increase in P-galactosidase activity from 510 units to 710 units when plasmid pAM8 is present rather than plasmid pAM4 or pAM7. Likewise, removal of the next GTTA tetramer (plasmid pAM13) yields an even higher level of titration, 750 P-galactosidase units as opposed to 710. The higher P-galactosidase activities obtained using plasmids pAM8 and pAM13 suggest that the first three GTTA tetramers may interact with ModE forming unstable associations, which results in an overall less efficient binding of ModE to its true target binding regions. The fact that these first three GTTA tetramers do naturally exist in the modEA intergenic region implies that perhaps these GTTA tetramers help to modulate the repression of the modABCD operon by ModE under suboptimal intracellular molybdate concentrations. Removal of the first four as well as the first four and part of the fifth GTTA tetramers (plasmids pAMlO and pAM6, respectively) results in the complete loss of titration of ModE, which demonstrates that the fourth and fifth GTTA tetramers are

PAGE 66

necessary for successful binding of ModE. Moreover, these two GTTA tetramers are the ones that are situated in the modA -10 sequence, and would be expected to participate in binding of the repressor given that binding of a repressor at this point would serve to obstruct the binding of RNA polymerase and thereby limit transcription of the ensuing gene. Furthermore, removal of only the sixth and seventh GTTA tetramers (pAM15) also results in complete loss of titration as indicated by the drop in p-galactosidase activity to 90 units. This finding establishes the requirement for the sixth and seventh GTTA tetramers for ModE binding, and again the location of these two GTTA sequences 4-bp downstream of the transcription start site, supports the assertion that these sites are integral to the binding of repressor. Additionally, the fact that the binding of ModE to the DNA requires the presence of the fourth and fifth as well as the sixth and seventh GTTA tetramers intimates that the binding of ModE to one set of sites necessitates the initial binding of ModE to the other site. This potential binding order in turn evokes the possibility of cooperative binding occurring; however, none of the methods used in this study to characterize the binding of ModE to its target DNA is capable of evaluating cooperativity. Introduction of plasmid pAM16, which only lacks the eighth GTTA sequence, into strain SE2069 results in a partial loss of titration of ModE, thereby implicating the involvement of this eighth tetramer in stabilizing ModE once it is bound to the DNA or possibly aiding in the initial recruitment of ModE. Based on these in vivo titration experiments, it has been determined that the binding of ModE minimally requires the presence of the fourth, fifth, sixth, and seventh GTTA tetramers, while the presence of the eighth GTTA tetramer increases the efficiency

PAGE 67

53 of binding. Furthermore, the arrangement of these GTTA tetramers required for binding (the fourth and fifth GTTA tetramers [GTTATATTG] seem to constitute one binding site, while the sixth and seventh GTTA tetramers [TAACGTTA] could form a second site) and the size of ModE protein (28.271 Da), predicts that ModE likely binds at each of the two sites as a dimer. The differences in the sequences also suggests different binding efficiencies. In vitro studies characterizing the binding of ModE to its target DNA. Having established putative binding sites for ModE using the in vivo titration experiments, in vitro techniques such as DNA-mobility shift assays and DNase I-footprinting experiments were undertaken in an effort to confirm the in vivo findings as well as to provide further information on the kinetics of the interaction. One set of experiments designed specifically to determine whether the in vivo titration results would be mirrored by DNA-mobility shift assay results, involved preparing binding reactions containing various amounts of ModE and 32 P-labeled DNA-oligomers derived from the mod DNA in plasmids pAM13, pAM15 and pAM16. In these experiments, a 211 -bp Dral/EcoJH. fragment derived from plasmid pAM13 (40 bp of /noc£4-operator/promoter DNA from -15 to +25 and 171 bp of plasmid pUC19) and a 313-bp EcoRl/Hindlll fragment from plasmid pAM15 (272 bp oimodAoperator/promoter DNA from -247 to +25 and 41 bp plasmid pUC19 DNA) and a 305bp-iscoRI/Z/wdlll fragment from plasmid pAM16 (267 bp modA -operator/promoter DNA from -247 to +20 and 38 bp plasmid pUC19 DNA) was used as the source of mod operator/promoter DNA. Standard DNA-mobility shift assay protocol outlined in "Materials and Methods" section was followed, and the results are presented in Fig. 7.

PAGE 68

54 II III 1 23 456789 10 Fig. 7. DNA-mobility shifts using plasmid pAM 13 -derived (panel I), plasmid pAM15derived (panel II), and plasmid pAM16-derived DNA (panel III) and ModE. A 0.1 pmol amount of labeled DNA and 1 mM sodium molybdate were present in each binding reaction. The following amounts of ModE had been added to the reactions as indicated: lane 1, without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6, 50-fold excess; lane 7, 100fold excess; lane 8, 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess. The arrow indicates a 'biologically significant" shifted-band.

PAGE 69

55 In the DNA-mobility shift for the plasmid pAM 13 -derived DNA (Fig. 7. panel I), a detectable shift to position 'A' (complex A) occurred at a protein to DNA ratio of 1 .0 (25.7% DNA shifted to form 'A'). This "shift form' predominated until a ModE excess of 25-fold was reached, at which point, a second band begins to emerge. Other ModE-DNA complexes also emerged at 250-fold, 500-fold, and 1000-fold ModE excess. These different complexes most likely represent additional binding of ModE to the DNA in a nonspecific manner with little biological significance. When DNA is included in the binding reactions which is incapable of titrating ModE (plasmid pAM15-derived DNA; Table 7), shift-form 'A' ModE-DNA complex does not arise (Fig. 7, panel II). but the higher-mass shift complexes did arise at ModE excesses of 250, 500, and 1000-fold. Thus, the DNA-mobility results, combined with the inability of plasmid pAM15 to titrate ModE in the in vivo experiments (Table 7), indicates that the 'A' complex form is the DNA-target-specific form. Based on the ModE titration data (Table 7), plasmid pAM16derived DNA should prove capable of forming the shift-form 'A' ModE-DNA complex, albeit at higher ModE/DNA ratios than required for plasmid pAM13-derived DNA-ModE complex formation. This expectation was realized in the DNA-mobility shift experiments, as complex 'A' did form but was only first detectable at a ModE to DNA ratio of 5.0, and over 50% shift of DNA did not occur until 100-fold excess ModE to DNA (Fig. 7, panel III). Again, at higher ModE concentrations, 250, 500, and 1000-fold excesses, the larger ModE-DNA complexes arose with this DNA also. In all three of the DNA-mobility shifts presented in Fig. 7, multiple ModE-DNA complexes other than the complex identified as shift-form 'A' are observed. Since these

PAGE 70

ModE-DNA complexes arise regardless of whether the DNA is capable of titrating ModE, it is thought that these complexes result from non-specific binding of ModE to non-mod sequence or to mod sequence located upstream of the modA -operator/promoter when there is greater than 250-fold excess ModE to DNA in the binding reaction. In order to determine whether the non-mod DNA contributes to formation non-specific ModE-DNA associations, a DNA-mobility shift experiment was conducted in which a 42-bp oligomer (modA -operator/promoter DNA from -9 to +33) that is essentially the sequence present as the insert in plasmid pAMlO (modA -operator/promoter DNA from -10 to +25) was used. As indicated in Table 7, this particular DNA does not titrate ModE and, therefore, should not support formation of complex 'A'. Results of this DNA-mobility shift experiment are presented in Fig. 8, and as can be seen, DNA was not shifted at any ModE concentration used except at 1000-fold excess ModE compared to 42-bp oligomer. Furthermore, this shifted DNA-protein complex migrated at a much slower rate than the shift-form 'A' observed with DNA from plasmid pAM13 (Fig. 7, panel I). Inclusion of a smaller DNA fragment (42 bp, as opposed to 21 1 or 305 or 313 bp), which does not contain non-mod DNA, eliminated formation of the larger ModE-DNA complexes at all ModE concentrations except at the very high 1000-fold excess ModE. This result supports the assertion that the interaction of ModE with non-mod DNA at high ratios of ModE to DNA leads to the formation of non-specific ModE-DNA associations. The DNA-mobility shift experiments (Figs. 7 and 8) have demonstrated that the formation of the ModE-DNA complex (complex 'A') in vitro requires the presence of the fourth, fifth, sixth, and seventh GTTA tetramers, and that the eighth tetramer is necessary

PAGE 71

57 1 23456789 10 Fig. 8. DNA-mobility shift featuring 42-bp modA -operator/promoter DNA (-9 to +33) and ModE. A 0. 1 pmol amount of DNA and 1 mM sodium molybdate and the following amounts of ModE were included in each binding reaction. Lane, 1 without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6, 50-fold excess; lane 7, 100-fold excess; lane 8. 250-fold excess; lane 9, 500-fold excess; lane 10, 1000-fold excess.

PAGE 72

58 for efficient binding (Table 7; Fig. 7). The mobility shift experiments fully support the results from the in vivo ModE titration experiments (Table 7), further lending credence to the use of the titration experiments as a means of probing the association of protein with its target DNA sequence. It has already been shown both in vivo and in vitro that repression of the modABCD operon by ModE requires the presence of intracellular molybdate (Tables 3 and 4; Fig. 5). Therefore. DNA-mobility shift experiments were conducted for the purpose of establishing whether molybdate is required for the binding of ModE to its target DNA. For these DNA-mobility shift experiments, a 43-bp oligonucleotide (-18 to +25), was used in the binding reaction along with the appropriate amount of ModE protein. For one set of shifts, sodium molybdate was added to the binding reactions, to the gel the binding reaction samples were electrophoresed in, and to the electrophoresis tank buffer at a final concentration of 1 mM. For the second set of mobility shifts, molybdate was excluded. These two mobility shifts, presented in Fig. 9, indicate that molybdate is necessary for efficient binding of ModE, as the 'A' complex is first detectable at a ModE/DNA ratio of 1 .0 in the presence of 1 mM sodium molybdate, whereas in the absence of molybdate, the ModE-DNA complex was not detectable until a 100-fold excess ModE was included in the binding reaction. Furthermore, binding of 50% or greater of DNA with ModE was never achieved in the binding reactions lacking molybdate, while in the presence of 1 mM sodium molybdate, greater than 50% shift of the DNA to the ModE-DNA complex 'A' was obtained at a ModE/DNA ratio of 50.

PAGE 73

59 B— A— Fig. 9. DNA-mobility shifts using a 43-bp oligomer spanning the -17 to +25 region of the modA -operator/pro motor region and ModE. Panel Isodium molybdate was not added to the binding reaction, gel or running buffer. Panel II1 mM sodium molybdate was present in the binding reaction, gel, and running buffer. A 0.1 pmol amount of DNA was present in the binding reaction. The following amounts of ModE were present in the binding reaction samples as indicated. Lane, 1 without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess; lane 4, 10-fold excess; lane 5, 25-fold excess; lane 6. 50-fold excess; lane 7, 100-fold excess; lane 8, 250-fold excess; lane 9, 500fold excess; lane 10, 1000-fold excess.

PAGE 74

60 The apparent dissociation constant (K D ) for the binding of ModE to the 43-bp oligomer in the presence of ImM molybdate was determined after quantitation of the percent of shifted DNA using a Phosphorimager. This correlation is presented in Fig. 10 and an apparent K D of 0.33 uM was calculated for the ModE-DNA complex 'A'. An apparent K D for the binding of ModE to the 43-bp DNA in the absence of 1 mM sodium molybdate could not be calculated since even at 1 ,000 fold excess of protein, the amount of shifted DNA was less than 50% of the total DNA. Interaction of molybdate with ModE. Given that the presence of molybdate is required for efficient binding of ModE to its target DNA, the concentration of molybdate needed for a 50% shift of the 43-bp oligomer to the form 'A' ModE-DNA complex was determined. A ModE concentration of 0.5 uM (protein to DNA ratio of 50) was chosen for these shift experiments because, at this ModE concentration in the binding reaction, a 60% shift of the DNA was obtained in the presence 1 mM sodium molybdate, and at this ratio, the contribution of sodium molybdate to the ModE-DNA interaction would be readily discernable. Based on these experiments, a 50% shift of the 43-bp oligomer was accomplished at a molybdate concentration of 6 uM (Fig. 11). It is not unreasonable to consider that an artificially high apparent K D value for the association of molybdate with ModE could have been obtained given the inherent limitations of the experiment used to derive the apparent K D Specifically, for the molybdate concentration DNA-mobility shift experiments, once the binding reaction samples were loaded onto the gel, the effective concentration of sodium molybdate present in those samples would decrease as the sodium molybdate

PAGE 75

Fig. 10. Percentage of ModE-43-bp-oligomer complex 'A' formed versus concentration of ModE protein in the presence of 1 mM sodium molybdate. These values are from the DNA-mobility shift experiment presented in Fig. 8, panel II.

PAGE 76

62 100 80 < 60 40 20 0 0.0001 0.001 0.01 0.1 1.0 10.0 100.0 1000.0 [Molybdate] (jiM) Fig. 1 1 Percentage of 43-bp-oligomer bound to ModE protein in relation to the concentration of sodium molybdate present in the binding reactions in a DNA-mobility shift experiment.

PAGE 77

concentration for the entire system undergoes equilibration, and secondly, electrophoresis of the binding reaction complexes may perturb the ModE-molybdate association. If the actual binding of ModE to its target DNA constitutes the rate-limiting step of the association of ModE with the DNA rather than the interaction of molybdate with ModE, then this concentration of molybdate (6 uM) required for half-maximal binding most likely represents an upper limit value since it has previously been shown that the apparent dissociation constant for the interaction between ModE-molybdate and modoperator/promoter DNA is 0.3 uM. If however, it is the association of molybdate with ModE and not the binding of ModE to the DNA which serves as the rate limiting step, then the 6 uM apparent K D value for the dissociation of molybdate from ModE may indeed prove valid. This possibility is especially compelling in light of the K D value of 3 uM that was determined for the affinity of molybdate for the periplasmic molybdate-specific-binding protein, ModA, using fluorescence studies (59). Furthermore, an apparent K D of 6 uM for the binding of molybdate to ModE seems physiologically acceptable since molybdate entering the E. coli cell would preferentially be activated and ultimately incorporated into an appropriate apomolybdoenzyme. Under these physiological conditions, the apparent K D value for molybdate in ModE-system would be expected to be higher than the K M value for molybdate for the activation enzyme. Only when the internal molybdate saturates the biosynthetic pathway, would molybdate be available for association with ModE resulting in the repression of the modABCD operon and cessation of further production of the components of the molybdate-specific-transporter. If the apparent K D for the dissociation

PAGE 78

of molybdate from ModE were much lower, then it would be likely that molybdate's interaction with ModE would be favored above its activation and incorporation into molybdoenzymes which would prove physiologically untenable for the cell. Given the uncertainty of whether it is the interaction of molybdate with ModE or it is the binding of activated ModE to DNA which acts as the limiting step for the repression, it seems that the only way to establish an accurate description of the kinetics governing the association of molybdate with ModE, and in turn, the binding of the ModE-molybdate complex to its target DNA is to determine the interaction in solution and compare to the appropriate enzymes competing with ModE for molybdate. Oxyanion specificity of ModE. Having established that molybdate is required for the repression of the modABCD operon in vivo and the efficient binding of ModE to its target DNA in vitro, the question of whether structural analogs of molybdate could substitute for molybdate was addressed. This question was answered in two ways: (1) modA ''lacZ expression in strain SE2069 cells that were cultured in LBG medium supplemented with 1 raM sodium molybdate, 1 mM sodium tungstate, 1 mM sodium sulfate, or 1 mM sodium ortho-vanadate, and (2) DNA-mobility shift experiments in which the binding reactions, gels, and tank buffers contained 10 uM sodium molybdate, 10 uM sodium tungstate, 10 uM sodium sulfate, or 10 uM sodium ortho-vanadate. The resulting levels of modA ''lacZ expression in strain SE2069 cultured with the various oxyanions are presented in Table 8. Only in those cultures of strain SE2069 grown in medium supplemented with 1 mM sodium molybdate or 1 mM sodium tungstate did repression of the modABCD operon occur, as indicated by their low P-galactosidase

PAGE 79

65 Table 8. Expression of modi ''lacZ in strain SE2069 cultured in the presence of various oxyanions. Strain Relevant Genotype Oxyanion P-galactosidase Activity 2 Added SE2069 ${modA'-'lacZ) None 1,300 MoOy 2 30 WOy 2 40 SOy 2 1,400 VO/ 3 1,900 "Cells were grown in LBG with the appropriate oxyanion at a final concentration of 1 mM. Enzyme activities are expressed in nanomolesminute" 'milligram of cell protein'.

PAGE 80

66 activities. Therefore, it appears that tungstate is the only oxyanion tested which is capable of functionally substituting for molybdate in the molybdate-dependent activation of ModE in vivo. Furthermore, the fact that higher modA 'lacZ expression levels were observed when the cells were grown in LBG medium containing sodium sulfate and especially sodium vanadate compared to cells grown in medium without any oxyanion supplementation (1,400 and 1,900 p-galactosidase units, respectively, as opposed to 1,300 units) implies that perhaps the presence of these two oxyanions inhibits the interaction of ModE with any of the trace amount of molybdate that may have been present in the cells, thereby reducing binding of ModE-molybdate complex to the modA -operator/promoter region. The results of the DNA-mobility shift experiments (Fig. 12) which contained 10 uM amounts of the previously indicated oxyanions in the binding reactions, were consistent with the in vivo modA ''lacZ expression data (Table 8). More specifically, high affinity binding of ModE to the 43-bp DNA and shift to complex 'A' only occurred in the presence of sodium molybdate (Fig. 12, panel I) or sodium tungstate (Fig. 12, panel II), while a low level of binding was seen for the shifts with sodium sulfate (Fig. 12, panel III). Sodium vanadate was fifty times less effective in supporting DNA-ModE protein interaction (Fig. 12, panel IV). The apparent K D values for the ModE-DNA complex obtained from the correlation of percent of shifted DNA (ModE-DNA complex species 'A') versus the concentration of ModE for the shifts with the different oxyanions were as follows: 0.22 uM with 10 uM sodium molybdate supplementation, 1.3 uM with sodium tungstate, and 4.7 uM with sodium sulfate (Fig. 13). An apparent Kq value for the

PAGE 81

2 -o s O co 1 2 CO i CU i >o at) -h •S ^ T3 p • S 5 X) CU •5 a, .£ H s cu co co &b CU c Oh 2 S £ > -T3 co O C c IS S3 3. 2 o CO CO CO* X cu w CU T) X £ S Cy i "55 — O cS co ^ O Q. CS CO _0 CD CO M CO Si -* J s H-> .3 Q Ch X W Os u C CO CO CU X CU c — cu — — CU sS o -o & b cu cS 1 1 I • s I O -3 8 o CU 2 Oh =!. 1 5 — CU .§ 3 o .5 o o u £ c < -B Z a Q | H J> • *o .£P S o -~ £ c ^ 0V 5 a h. C id • cu — O co • g Q MOO O -a a £ "2 o o o\ 00" cu e jg co' CO cu cu X CU Oh d ^ o C O O o o C co cu cu ?5 o O CU h3 H— C3 2 1 O cu 2 0 TD g|2 E O „ 3 = (N T3 c2 U O ... C, co J> -3 — -O — a. g 0 ^ nS S Q 9 ^ S3 a cu C J2 CO CO cu U X cu O sC CU O co CG co CU O a o = £ CU o cu S .3 CO Q. .. w o w 23 9? ;3 3 On-g co o .a cu W c2 2 6 cS o 1 o >T) — ^ o in — h cu cu c S 03 S3

PAGE 82

68

PAGE 83

69 Fig. 13. Percent shifted DNA versus concentration of ModE resulting from DNAmobility shift binding reactions containing various oxyanions. Bindng reactions included 10 uM sodium molybdate (filled squares), 10 uM sodium tungstate (open squares), 10 uM sodium sulfate (open circles), or 10 uM sodium ortho-vanadate (closed triangles).

PAGE 84

70 binding of ModE in the presence of 10 uM sodium vanadate could not be calculated since 50% shift of the DNA was not achieved with the ModE concentrations used. It is not surprising that tungstate proved to be the only oxyanion of those tested which was capable of substituting for molybdate in activating ModE to any significant degree, since it has previously been shown that activated tungstate can be incorporated successfully into some E. coli apo-molybdoenzymes in place of molybdate (57). However, these tungstatesubstituted molybdoenzymes are not functional probably because of a failure of electron transfer across the tungsten center. Yet. the stable association of tungstate with ModE would result in the promotion of the activation of ModE, since presumably, the activation of ModE relies only on the binding of an appropriately sized and shaped moiety to affect the change in conformation of the protein which allows for the efficient binding of DNA, and does not require any further chemical process such as the transfer of electrons. Stoichiometrv of ModE-DNA complex. Since a stable ModE-43-bp DNA shift complex (complex form 'A') has been reproducibly obtained under a variety of conditions, this general method was used to determine the molecular weight of the ModEDNA complex which would then indicate the number of ModE proteins which are associated with the DNA in complex 'A'. For the molecular weight determination of the ModE-DNA complex, standard binding reaction samples supplemented with 1 raM sodium molybdate were electrophoresed in 5%, 6%, 7%, and 8% nondenaturingpolyacrylamide-TBE gels. The resulting R,values for ModE-DNA complex and the protein standards after migration through the four different gels were used to prepare a Ferguson plot (19) from which the molecular weight of the ModE-DNA complex could be

PAGE 85

71 extrapolated (Fig. 14). Based on the results of these experiments, a value of 81,247 Da was obtained for the size of this complex. This molecular weight compares favorably with a molecular weight of 83,144 expected for the association of a ModE-dimer (56. 424 Da) with the 43-bp DNA (26,602 Da), as there is only a 2.3% difference in the expected value versus the observed molecular weight. Thus, the ModE-DNA complex form 'A' is minimally a ModE dimer bound to the 43-bp wo
PAGE 86

Fig. 14. Ferguson plot generated from the Rf values obtained after electrophoresis of the non-denatured protein standards and ModE-43-bp-modA -operator/promoter DNA complex in 5%, 6%, 7% and 8% polyacrylamide-TBE gels. The dotted lines indicate the extrapolation for the determination of the molecular weight of the ModE-DNA complex form 'A'.

PAGE 87

73 rttt 1 2 3 4 Fig. 15. DNase I-footprinting experiment in which the binding reactions contained 1 mM sodium molybdate. 0.1 pmol 32 P-labeled-446-bp-moc£4-operator/promoter DNA and various concentrations of ModE. Lane 1, DNA alone; lane 2, 1-fold excess ModE compared to DNA; lane 3, 10-fold excess ModE; lane 4, 100-fold excess ModE. The boxed sequence represents the nucleotides located in the protected regions. The indicates bases that are subject to hypersensitivity of DNase 1 cleavage.

PAGE 88

(from top of Fig. 15) (GTTATATTG; -15 to -7) encompasses the fourth and fifth GTTA tetramers located in the -10 region of the modi -operator/promoter region (Table 7), which would confirm the importance of GTTA sequences in the binding of ModE. Also there appears to be a "G' located three bases upstream of this first protected region which is hypersensitive to DNase I-cleavage. However, the second (CCTACAT; -3 to +4) and third (GTTACAT; +8 to +14) protected areas, respectively, do not uphold the prediction of the involvement of the GTTA tetramers located in these regions in the binding of ModE since only part of the sixth and seventh GTTA tetramers (Table 7) which comprise the 8bp inverted repeat (+4 to +1 1) are located in the protected area. A second DNase I hypersensitive nucleotide ('A') is located at the lower boundary of the second protected region. The location of these hypersensitive cleavage spots are not unexpected since it has previously been shown that hypersensitive cleavage sites often flank areas of protection (10). In light of the findings from the DNase I-footprinting experiments, a reevaluation of the sequence involved in binding ModE was undertaken which resulted in the assignment of the sequences TACAT and TAT AT as those necessary for the binding of ModE, since these sequences are common to all three protected regions. The fact that the TAT AT and TACAT sequences found in the three protected regions overlap the fourth and fifth GTTA tetramers (first protected region) and sixth and seventh GTTA tetramers (second and third protected regions), respectively (Fig. 1 5 and Table 7), supports the interpretation that removal of the above indicated GTTA sequences from certain plasmids used in the ModE titration experiments resulted in the loss of binding by ModE.

PAGE 89

A DNase I-footprinting experiment was also conducted in which molybdate was withheld from the binding reactions in order to determine whether ModE is capable of binding the DNA, thereby protecting it from cleavage by DNase I. even in the absence of molybdate, and secondly, to determine, if protection does occur, whether the protected regions are identical to those obtained from the reactions supplemented with sodium molybdate. As can be seen from the results of these experiments (Fig. 16), protection of the DNA from DNase I cleavage did occur in the presence of 100-fold and 1000-fold excess ModE, which is ten times more ModE required for the protection in this experiment than was needed when the reactions included sodium molybdate. It was expected that protection of the DNA from DNase I cleavage in the absence of molybdate would be compromised since it had previously been shown in mobility shift experiments (Fig. 9) that the binding of ModE to the 43-bp modA -operator/ promoter fragment was severely reduced without the inclusion of sodium molybdate in the binding reactions. The protection pattern that did result from the binding of ModE to DNA in the absence of added molybdate was identical to that obtained from the experiment conducted with sodium molybdate supplementation. However, the 'G' at -18 is hypersensitive to DNase I cleavage only in the presence of molybdate. It is possible that in this experiment, a small fraction of ModE is bound to trace amounts of molybdate present in the solution and that only this ModE-molybdate complex binds to the DNA. Alternatively, molybdate-freeModE can bind to DNA but at a very low efficiency. The differential hypersensitivity of 'G' at position -18 suggests that molybdate-free-ModE is capable of binding to the DNA.

PAGE 90

76 1 2 3 4 5 Fig. 16. DNase I -footprint of ModE in /noc£4-operator/promoter DNA in which sodium molybdate was excluded from the reaction mixtures. The sample in lane 1 contains no ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 10-fold excess ModE; lane 4, 100-fold excess ModE; lane 5, 1000-fold excess ModE. The boxed sequences represent nucleotides located in the protected regions. The indicates nucleotides which are hypersensitive to DNase I cleavage.

PAGE 91

77 The protection of the motZ^-operator/promoter DNA by ModE in three distinct regions is not consistent with the determination of dimeric binding of ModE to DNA in the ModEDNA complex (Fig. 14). These results taken together suggest that, the association of ModE with the modA -operator/promoter DNA is not simply binding of ModE to each site. Rather, these data suggest a possible scenario in which a monomelic ModE protein may initially bind loosely to one of the three protected regions, and this initial binding then allows for the rapid recruitment of additional ModE-molybdate to the region for stable binding of ModE protein as a dimer to the other two sites. This monomeric association with the DNA is probably unstable and is rapidly removed by electrophoresis in the DNAmobility shift experiment. Yet, at this point, any further discussion of the course of the binding of ModE to the DNA must await binding experiments carried out and analyzed in solution. In order to evaluate the role of the three ModE-protected regions in vivo, another set of plasmids was constructed for ModE-titration experiments similar to the ones previously described (Table 7). Plasmid pAM17 contains protected sites 1 and 2, while plasmid pAM18 has protected sites 1 and 3 with the spacing between the sites preserved (Table 9). Plasmid pAM19 contains only the protected regions 1 and 3 separated by only two bp. The modA ''lacZ expression levels, measured as P-galactosidase activity, for strain SE2069 containing the above-mentioned plasmids, indicated that only plasmid pAM17 is capable of partially titrating ModE (Table 9). Strain SE2069 harboring plasmid pAM17 produced only 220 P-galactosidase units, while only 70 p-galactosidase units were obtained for strain SE2069 lacking any plasmid. The level of titration of ModE in

PAGE 92

78 Table 9. Titration of ModE protein in vivo by various regions of the modA -operator /promoter DNA in strain SE2069 (modA'-'lacZ + ) Plasmid DNA Sequence /S-Galactosidase Activity No Plasmid 70 pAM4 GGTACC186nt CAAAGGTTAGCAATAACTGCTGGGAAAATTCCGAGTTAGTCGT^^TGTCGCC^^^AACGTgJ^jTAAGGGGTTAC 600 1 2 3 pAM17 TCGTgggjTGTCGCCg^^jAA 22 0 1 2 pAM5 CGCCjyjjJjJjgAACGlQ^ggTAAGGGGTTAC 60 2 3 PAM18 TCGT^ggjTGTCGCCCTTGGAACGTg^ggT 60 1 3 PAM19 TCGT^JJggjT GTSjgggjT 50 1 3 Cells were grown in LBG medium with molybdate (1 mM). 0-gaIactosidase activities are expressed as nanomoles.minute'.milligram of cell protein 1 The ModE protein binding sites are indicated by inverse print and the numbers designate the particular binding site (see Text for details). The underlined sequence "CTTGG" replaces the sequence in region 2 preserving the spacing between regions 1 and 3.

PAGE 93

79 the cells containing plasmid pAM17 is not the maximal level possible, since 600 0galactosidase units of activity was produced by strain SE2069 with plasmid pAM4 which contains all three protected regions. Thus, it appears that only protected regions 1 and 2 are absolutely required for the binding of ModE, but that in addition, region 3 is necessary for maximal binding. DNA-mobility shift experiments were also conducted using the modAoperator/promoter DNA contained in plasmids pAM17, pAM18. and pAM19. As would be expected, based on the in vivo titration experiments, only the modA -operator/promoter DNA present in plasmid pAM17 was capable of forming the *A* complex in the presence of ModE protein as presented in Fig. 17 (Panel I). Moreover, as is also suggested by the partial titration seen in the ModE titration experiments, the half-maximal binding of ModE to the 36-bp modA -operator/promoter fragment occurs at a ModE concentration of 1.8 uM (Fig. 18) which is considerably higher than the 0.33 uM apparent K D value obtained for the binding of ModE to DNA containing all three sites (Fig. 10). The results of the in vivo ModE -titration experiments coupled with the in vitro DNA-mobility shift experiments indicate that the binding of ModE only requires sites 1 and 2, but site 3 is necessary for efficient binding at sites 1 and 2. These data lend credence to the premise that site 3 may serve as a location to which ModE initially binds, thereby promoting the stable binding of ModE to sites 1 and 2.

PAGE 94

I 1 23456789 10 Fig. 17. DNA-mobility shift experiments using the modA -operator/promoter DNA contained in plasmids pAM17, pAM18, and pAM19 and ModE. For the shift with DNA present in plasmid pAM17 (panel I), a 32 P-labeled-36-bp modA DNA fragment was included in the binding reactions along with 1 mM sodium molybdate and the appropriate amount of ModE (lane 1, without ModE; lane 2, 1-fold excess ModE compared to DNA; lane 3, 5-fold excess ModE; lane 4, 10-fold excess ModE; lane 5, 25-fold excess ModE; lane 6, 50-fold excess ModE; lane 7, 100-fold excess ModE; lane 8, 250-fold excess ModE; lane 9, 500-fold excess ModE; lane 10, 1000-fold excess ModE). The mobility shifts featuring the 45-bp modA DNA present in plasmid pAM18 (panel II) and 31 -bp modA DNA present in plasmid pAM19 (panel III) also contained 1 mM sodium molybdate and the amount of ModE indicated for the lanes in panel I.

PAGE 95

100 80 0.01 0.1 1.0 10.0 [ModE] (pM) Fig. 18. Plot of % shifted DNA versus concentration of ModE required for the shift featuring the 36-bp modA DNA present in plasmid pAMl 7. The data were from the experiment presented in Fig. 17, Panel I.

PAGE 96

82 Molybdate-independent ModE Proteins Mutation of DNA which codes for proteins of interest is often employed as an approach to study structure/function characteristics of the protein. In this study hydroxylamine, which causes G-to-A transitions, was used to mutate the DNA coding for ModE (plasmid pAGl) in order to identify critical amino acids responsible for the association of molybdate with ModE and/or the binding of ModE to its target DNA. Following the hydroxylamine treatment, two different mutant phenotypes were observed: partial loss of repression of the modABCD operon in the presence of molybdate and complete repression of the modABCD operon in the absence of molybdate (superrepressor). The phenotypes observed were confirmed by assaying strain SE181 1 harboring plasmids with modE mutations for P-galactosidase activity, and the results of these experiments are presented in Fig. 19. The partially derepressed (A76V) ModE mutant has a substitution of valine for alanine at amino acid position 76. Three "superrepressor" mutants which were capable of repressing the modABCD operon in the absence of molybdate were also isolated, and two of these mutants harbor the following mutations: threonine at amino acid position 125 replaced with isoleucine (T125I) and glycine at position 133 changed to aspartic acid (G133D). For the third "superrepressor" mutant (Q216*), the C-terminal 47 amino acids starting from the amino acid at position 216 were deleted.

PAGE 97

83 Plasmid pAGl pAGl(A76V) pAGl(T125I) pAGl(G133D) pAGl(Q216*) Mutation P-galactosidase Activity 2b2 A76V T125I G133D Q216" -Mo +Mo 600 100 1,300 600 125 110 40 40 <10 <10 Fig. 19. Analysis of ModE mutants. P-galactosidase activity produced by strain SE181 1 carrying the indicated mutation was determined and expressed as nanomoles-min" 1 •milligram of cell protein 1 Cells were grown in LBG with (+) or without (-) sodium molybdate (Mo) at a final concentration of 1 mM. represents translation termination.

PAGE 98

84 The mutation in the partially derepressed mutant (A76V) lies immediately after a stretch of three glycines which are identical in three of the four ModE homo logs presented in Fig. 3, the fourth (RcMopA) having two of the three glycines. Two of the four homologs (EcModE and RcMopA) have alanine at this position while the other two (AvModE and HiModE) have threonine at this position. Furthermore, this mutation is 23 amino acids away (towards the C-terminal end) from the putative helix-turn-helix structure, so it is possible that this mutation may have disrupted the helix-turn-helix structure resulting in a partial loss of DNA binding by the mutant protein. Two of the "superrepressor mutants", T125I and G133D, have mutations situated near a five amino-acid long sequence SARNQ (amino acids 126 to 130), which is conserved among the five homologs, with the exception of R. capsulatus MopA which has an asparagine in place of alanine and a threonine in place of glutamine. In fact, the T125I mutation directly precedes the SARNQ sequence and is itself conserved among three of the homologs (AvModE, EcModE, and RcMopA) with serine substituting for threonine in H. influenzae ModE. The G133D mutation lies two amino acids further on the C-terminal side of the SARNQ sequence, and the glycine is also conserved in four of the homologs (AvModE, EcModE, HiModE, and CpMopII). The glutamine which is replaced by a stop codon in the third superrepressor mutant (Q216*) is unique to E. coli ModE at this position; however, there are regions of conserved amino acids present in the missing Cterminal end of ModE whose loss undoubtedly contributes to the phenotype of this mutant.

PAGE 99

85 It is also interesting to note that the amino acid sequence deleted in this mutant (Q261*) contains the only three cysteines present in the wild type K coli ModE protein, and it is possible that two of these three cysteines may be involved in forming a disulfide bond. Moreover, it is tempting to speculate that ModE adopts an open conformation when molybdate is not bound to the protein, and that upon binding of molybdate, a change in the structural conformation ensues which involves formation of a disulfide bond. This formation of the disulfide bond could stabilize a more compact structure which promotes the efficient binding of ModE to the appropriate DNA sequences. Furthermore, it is apparent that the deletion of the region of the protein which contains these cysteine residues in the (Q216*) mutant ModE protein resulted in the adoption of the "activated" or molybdate-bound conformation even in the absence of bound molybdate. At the current level of information available concerning the structure of the ModE protein, a more defensible assertion at this point in time is that the SARNQ sequence may be integral to the binding of molybdate by ModE given that two of the three isolated "superrepressor" mutations abut this SARNQ region and that all of the ModE homologs as well as other proteins that are either known to have or are suspected to have interactions with molybdate contain variations of this sequence (Table 10). It should also be mentioned that it would be expected that binding of the negatively charged molybdate moiety by ModE would most likely require positively charged amino acids such as those present in the SARNQ region. Furthermore, if the nature of the G133D "superrepressor" mutation is considered, it is possible that the substitution of aspartic acid in place of

PAGE 100

86 Table 10. List of proteins which are known to interact with molybdate and DNA and contain a possible molybdate-binding motif. Protein Molybdate-binding motif sequence Reference ModE from E. coli SARNQ (24,47,60) ModE from H. influenzae SARNQ (20) ModE from A. vinelandii SARNQ (43) MopII from C. pas teurianum SARNQ (32) MopA from R. capsulatus SnRNt (76) FhlA from E coli 9Hnn (4 / b4) NarX from E. coli SGRNe (38) NarQ from E. coli SIRmQ (12) Estrogen receptor from Rat LdRNQ pAtNQ (44) Estrogen receptor from Human LARgQ (39) Lower case letters represent non-identical or non-similar amino acids. NarX and NarQ are membrane-bound components of the nar regulatory system but do not bind DNA themselves.

PAGE 101

8" glycine mimics the binding of molybdate due to the presence of the negatively-charged carboxyl group in the amino acid side chain. A listing of regulatory proteins which interact with molybdate and corresponding SARNQ-like sequences found in these proteins are presented in Table 10. As has already been indicated, four of the five ModE homologs contain the SARNQ sequence, while R. capsulatus ModE homolog, MopA, has an asparagine in place of alanine and threonine in place of glutamine. FhlA which serves as a molybdate-responsive activator of the hyc operon which codes for hydrogenase-3 (47, 63, 64, 68) has a sequence SGRNN (amino acid positions 426 to 430) which has three of five conserved amino acids while the remaining two are similar. NarX protein, which functions to activate the regulator protein, NarL, in response to the presence of nitrate, contains the sequence SGRNE (amino acid positions 204 to 208; 38). Interestingly, specific mutations in the SGRNE region result in the molybdate-independent activation of NarL (38). A second nitrate sensor protein, NarQ, has the sequence SLRMQ (amino acids positions 48 to 52; 12), which bears similarity to the general motif. Furthermore, the rat estrogen receptor-hsp90 complex which is stabilized by molybdate (44) contains two sequences which share similarity with SARNQ, one defined as PATNQ (amino acid positions 227 to 23 1) and a second LDRNQ which is located at amino acid positions 415 to 419 (44). The human estrogen receptor also contains a sequence similar to SARNQ, LARGQ, located at amino acid positions 190 to 194 (39). However, this SARNQ sequence or any similar sequence is not found in any of the periplasmic molybdate binding proteins of the transport system.

PAGE 102

The fact that certain mutations affecting amino acids located near the SARNQ sequence result in the molybdate-independent phenotype for ModE suggests that these amino acids are important for the proper functioning of the ModE protein. Secondly, the fact that all of the ModE homologs and a number of proteins which are thought to interact with molybdate contain sequences which are either completely conserved or bear striking similarity to the SARNQ amino acid sequence further implies that the amino acids in this region may be involved in activation of the ModE aporepressor by means of binding molybdate. Regulation of modE The regulation of ModE was investigated by introducing §(modE ''lacZ) borne on ARM26 into various backgrounds and assaying for P-galactosidase activity. The results presented in Table 1 1 indicate that there is little difference in the levels of P-galactosidase obtained from a wild type strain, SE1938, a modB mutant strain, SE1952; a modE mutant strain, SE1942; or a mo dF mutant strain, SE1940. These results show that modE is transcribed constitutively at low levels, which is consistent with ModE's proposed role as a repressor. A plasmid-based
PAGE 103

89 Table 1 1 Regulation of modE in different mod strains Strain Relevant Genotype P-galactosidase activity" -Mo +Mo SE1938 §(modE'-'lacZ) 324 297 SE1952 (b(modE'lacZ) modB::ln5 316 249 SE1942
PAGE 104

90 Table 12. Regulation of modE in various backgrounds Strain Relevant Genotype p-galactosidase activity 2 -Mo +Mo RK4353(pRM14)
PAGE 105

91 observed in any of these backgrounds. The higher P-galactosidase activity in these plasmid containing modE ''lacZ strains is a consequence of multiple copies of the modE ''lacZ fusion. Regulation of Other Genes by ModE ModE has already been shown to exert a negative effect on the transcription of the modABCD operon when sufficiently high amounts of molybdate are present in the cell. Investigation of the effect of ModE on the transcription of other genes has led to findings that ModE also regulates transcription of modF as well as narG. The P-galactosidase activities produced by a modF''lacZ operon fusion (ARM37) in the X att site, which demonstrate the regulation of modF by ModE, are presented in Table 13. Since modE and modF appear to comprise a two-gene operon transcribing from a single promoter through both genes, the study of the regulation of modF by ModE required the uncoupling of the genes so that expression of modF can be established without interference by ModE produced from the same operon. This uncoupling of the genes was accomplished by removal of a segment of the modE gene from plasmid DNA containing the §(modF'lacZ) before recombining (inframe deletion) with phage XRZ5 to produce ARM37, thereby preventing production of functional ModE protein, but leaving the entire modF gene in frame with respect to the modEF promoter. Production of P-galactosidase activity by a strain carrying modF''lacZ and a functional chromosomally encoded ModE protein (Strain SE2105) is very low (70 pgalactosidase activity units) as compared to 1,050 P-galactosidase activity units obtained for the modE mutant strain (strain SE2106). Introduction of a plasmid containing wild

PAGE 106

Table 13. Regulation of modF by ModE Strain Relevant Opnotvne IvWlW V CU.1L V.J V_ 1 1V_> L V [ P-galactosidase activity 3 -Mo +Mo SE2105 70 70 SE2106 (Ti( Wtf\i1h iflf*'7\ A YYic\ti /^-Irm 1,050 1,180 SE2106 (pRM20) \yyffl\Jl41 ILiLz^ f LA frlUCil—, ivl 1 L frlUtilZ, <10 <10 SE2107 <\>(modF'lacZ) AmodF-km 180 170 SE2112 (modF'lacZ) AmodF-km, modB::Tn\0 1,030 31 a Cells were grown in LBG medium with or without sodium molybdate at a final concentration of 1 mM. Enzyme activities are expressed as nanomoles-mm" 1 milligram of cell protein" 1

PAGE 107

93 type modE (plasmid pRM20) into the modE mutant, strain SE2106 restores the repression of modF''lacZ expression as evidenced by the very low P-galactosidase activities (< 10) obtained for strain SE2106(pRM20). Expression of modF''lacZ was approximately 2 times greater in a modF mutant strain (strain SE2 1 07) than in the wild type strain. This higher level of expression was also observed for the strain deficient in both ModE and ModF (strain SE21 12) compared to the strain deficient only in ModE, suggesting that modF expression may be subject to autoregulation as well as repression by ModE. As would be expected, given that in vivo as well as in vitro experiments have indicated that the DNA binding capacity of ModE depends on the presence of molybdate, repression of modF''lacZ expression was alleviated in molybdate transport mutant strains (strains containing modBv.TnlO). Supplementation of the growth medium with 1 mM sodium molybdate, restored repression of modF'lacZ expression in all the molybdate transport mutant strains except for the modE, modB::Tn\0 double mutant strain (strain SE2116). Based on the data presented in Table 13, it appears that ModE represses the transcription of modF when there is a sufficient intracellular concentration of molybdate. Moreover, the regulation of modF by ModE is similar to the regulation of the modABCD operon by ModE (Tables 3 and 4). This similarity in the regulation of both modF and modABCD by ModE is not surprising, since it is likely that the functions of all of the effected genes are in some way directed towards the successful transport of molybdate and incorporation of activated molybdenum into the appropriate apo-molybdoenzymes, and as such would most likely be coordinately regulated.

PAGE 108

94 The regulation of narG by ModE on the other hand was somewhat unexpected since it has already been shown that molybdate exerts influence over the transcription of narG presumably through the NarX and NarQ proteins, and an additional level of molybdate-dependent control through ModE seems superfluous (12, 38). However, as can be seen by the levels of narG 'lacZ expression presented in Table 14, ModE clearly has an effect on transcription through narG. More specifically, the absence of ModE results in reduction of transcription through narG ''lacZ, as evidenced by the drop in Pgalactosidase activity from 28,280 units in the wild type strain to 1 1,010 units in the ModE strain. As would be expected, complementation of the modE mutation by introduction of a modE containing plasmid, pAGl, into the modE mutant strain resulted in an increase in the p-galactosidase activity to a value comparable to wild type levels. The nature of the regulation of narG exerted by ModE is opposite of that seen for either the modABCD operon or for modF, as it appears that ModE is functioning as an enhancer of activation of narGHJI operon expression. This shows the complexity of the regulation of anaerobic respiratory pathways, in this case, the regulation of narG expression has already been shown to be positively activated by NarL as well as FNR. Apparently the additional level of regulation afforded by ModE helps optimize the expression of narG operon in response to molybdate availability. The possibility of ModE serving as an activator for certain genes is currently being explored for the hyc and moeAB operons (27, 68). Yet, at this juncture it is premature to state that the ModE functions as both a repressor of the modABCD operon and modF as well as an activator of other genes until studies providing direct proof of ModE's

PAGE 109

95 Table 14. Regulation of narG expression by ModE Strain Relevant Genotype P-galactosidase activity 3 MC4100(APC50) §{narG ''lacZ) 28,280 SE4110(A.PC50) §{narG ''lacZ) AmodE-km 11,010 SE4110(APC50) (pAGl) §{narG ''lacZ) AmodE-km, modE* 33.460 a Cells were grown in LBG medium supplemented with 40 mM potassium nitrate and 1 mM sodium molybdate. Enzyme activities are expressed as nanomoles min" 1 milligram of cell protein" 1

PAGE 110

interaction with these other genes (binding studies, footprinting experiments, etc.) have been completed.

PAGE 111

CONCLUSION Proper function of the complement of molybdoenzymes present in E. coli requires the synthesis of the molybdopterin guanine dinucleotide form of the molybdenum cofactor. The successful production of the molybdenum cofactor is predicated on the presence of an intracellular pool of molybdate. In the typical E. coli habitat, only trace amounts of molybdate are likely to be found. This low level of molybdate in E. coifs environment necessitates the existence of a high-affinity molybdate-specific transport system in E. coli. Such a transport system has been identified and is coded by the modABC genes. However, a number of considerations such as the fact that E. coli possesses only a few enzymes requiring molybdenum and that the majority of those enzymes are membrane-associated and are, therefore, sparingly produced in the cell, and the fact that, at high intracellular concentrations, molybdate is toxic to the cell would suggest that the molybdate transport system be regulated in response to molybdate availability. Indeed, expression of the molybdate-specific transporter genes was found to be negatively regulated under conditions of intracellular molybdate saturation. Isolation of an E. coli strain (strain SE181 1) which did not exhibit molybdate-induced repression of the modABC D operon led to the identification of a protein, designated ModE, which is responsible for the molybdate-dependent repression of transcription of the transport genes. 97

PAGE 112

Subsequent studies confirmed the initial assignment of repressor function to ModE by demonstrating that ModE. in the presence of 1 mM molybdate, substantially reduced the production of plasmid-encoded ModA and ModB' proteins in a coupled transcriptiontranslation system. Further supporting evidence was provided by the successful binding of ModE to DNA derived from the operator/promoter region of the modA sequence in DNA-mobility shift experiments. Also, the DNA-mobility shift experiments showed that the binding of ModE to its target DNA required molybdate since an apparent Kq of 0.22 uM was calculated for the binding of ModE to DNA in the presence of 10 uM molybdate, whereas binding in the absence of molybdate did not occur to an extent which would allow calculation of a K D value. It was also shown that half-maximal binding of ModE to modA operator/promoter DNA resulted when 6 uM molybdate was present in the binding reaction. The difference in the dissociation constants for the binding of ModE to the DNA (0.22 uM) and the apparent dissociation constant of molybdate to ModE (6 uM) suggests that the binding of molybdate to the ModE protein serves as the limiting step in the interaction of ModE with the modA operator/promoter DNA. DNase I-footprinting experiments revealed that ModE binds to the modA operator/promoter DNA at three locations, and that the sequences 'TATAT' and 'TACAT', present in these DNase I-cleavage-protected regions, constitute the ModE recognition sequences. The fact that the ModE-DNA complex formed in the DNAmobility shift experiments consists of only a ModE dimer bound to the 43-bp modA DNA implies that ModE is forming only a loose association with one of the three DNase Icleavage-protected regions, and that this loose association promotes the subsequent stable

PAGE 113

99 dimeric binding of ModE to the other two binding sites. This possible binding scenario was supported by ModE titration experiments and the corollary DNA-mobility shift experiments (Table 9 and Fig. 17) which indicate that the first and second binding sites are minimally required for ModE binding, but that efficient binding requires the presence of all three sites; however, further studies which can discriminate the order of binding need to be conducted to confirm this particular scenario for the binding of ModE to the DNA. Isolation of the T125I and G133D mutant ModE proteins which are capable of repressing the modABCD operon in the absence of molybdate, was crucial to the identification of the TSARNQXXG amino acid sequence (or variations thereof) that is present in the H. influenzae, A.vinelandii, R. capsulatus, and C. pasteuhanum ModE homologs. Similar amino acid sequences were also found in other proteins that are known to interact with molybdate suggesting that this stretch of amino acids is most likely involved in the binding of molybdate by the ModE protein as well as in the other proteins. Although, the mutation analysis has provided some information in terms of structure/function, confirmation of the interpretations of these results requires X-ray crystallographic or NMR-based structural studies. Lastly, modE expression was shown to occur constitutively at low levels which is characteristic of repressor proteins that require modification for function. However, ModE, itself, was also shown to regulate additional genes other than the modABCD operon when it was demonstrated that ModE negatively regulates modF in a molybdatedependent fashion but enhances the expression of narG. Preliminary studies also indicate that ModE enhances the transcription of moeA and hycA genes as well (27, 68). Thus, it

PAGE 114

100 appears that ModE functions primarily to repress the expression of the modABCD genes in a molybdate-dependent manner, but additionally may enhance the transcription of select genes which are involved in molybdate metabolism in order to optimize E. coifs biosynthetic and energy-producing systems under varying conditions.

PAGE 115

REFERENCES 1. Allen, R. M., R. Chatterjee, M. S. Madden, P. W. Ludden and V. K. Shah. 1994. Biosynthesis of the iron-molybdenum cofactor of nitrogenase. Crit. Rev. Biotechnol. 14:225-249. 2. Amann, E. B. Ochs, and K. -J. Abel. 1988. Tightly regulated tac promoter vectors useful for the expression of unfused and fused proteins in Escherichia coli. Gene 69. 301-315. 3. Ames, G. F. -L. 1986. Bacterial periplasmic transport systems: structure, mechanism, and evolution. Ann. Rev. Biochem. 55:397-425. 4. Amy, N. K. 1 98 1 Identification of the molybdenum cofactor in chlorate-resistant mutants of Escherichia coli. J. Bacteriol. 148:274-282. 5. Amy, N. K., and K. V. Rajagopalan. 1979. Characterization of molybdenum cofactor in Escherichia coli. J. Bacteriol. 140:1 14-124. 6. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tools. J. Mol. Biol. 215:403-410. 7. Bohm, R., M. Sauter and A. Bock. 1990. Nucleotide sequence and expression of an operon in Escherichia coli coding for formate hydrogenlyase components. Mol. Microbiol. 4:231-243. 8. Bolivar, F., R. L. Rodriguez, P. J. Greene, M. C. Betlach, H. L. Heyneker, and H. W. Boyer. 1977. Construction and characterization of new cloning vehicles II. A multipurpose cloning system. Gene 2:95-113. 9. Bradford, M. M. 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. 10. Brenowitz, M., A. Pickane, and E. Jamison. 1991. Stability of lac repressor mediated "loop complex". Biochemistry. 30:5986-5998. 101

PAGE 116

102 11. Chang, A. C. Y. and S. N. Cohen. 1978. Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol. 134:1141-1156. 12. Chiang, R. C, R. Cawicchioli. and R. P. Gunsalus. 1992. Identification and characterization of NarQ, a second nitrate sensor for nitrate-dependent gene regulation in Escherichia coli. Mol. Microbiol. 6:1913-1923. 13. Chou, P. Y. and G. D. Fasman. 1978. Prediction of the secondary structure of proteins from their amino acid sequence. Adv. Enzymol. 47:45-147. 14. Corcuera, G. L., M. Bastidas, and M. Dubourdieu. 1993. Molybdenum uptake in Escherichia coli K12. J. Gen. Microbiol. 139:1869-1875. 15. Davis, R. W., D. Botstein, and J. R. Roth. 1980. A manual for genetic engineering advanced bacterial genetics. Cold Spring Harbor Laboratory. Cold Spring Harbor. NY. 16. Devereux. J., P. Haeberli, and O. Smithies. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395. 17. Dorrell, N, A H. Ahmed, and S. H. Moss. 1993. Photoreactivation in a phrB mutant of Eschericia coli K-12: evidence for the role of a second protein in photorepair. Photochem. and Photobiol. 58:831-935. 18. Dudourdieu, M., E. Andrade, and J. Puig. 1976. Molybdenum and chlorate resistant mutants of Escherichia coli K12. Biochem. Biophys. Res. Commun. 70:766-773. 19. Ferguson, K. A. 1964. Starch gel electrophoresis-application to classification of pituitary proteins and polypeptides. Metabolism. 13:985-1002. 20. Fleischmann, R. D., M. D. Adams, O. White, R. A. Clayton, E. F. Kirkness, A. R. Kerlavage, C. J. Bult, J. F. Tomb, B. A. Dougherty, J. M. Merrick, K. McKenney, G. Sutton, W. FitzHugh, C. A. Fields, J. D. Gocayne, J. D. Scott, R. Shirley, L. I. Liu, A. Glodeck, J. M. Kelley, J. F. Weidman, C. A. Phillips, T. Spriggs, E. Hedblom, M. D. Cotton, T. R. Utterback, M. C. Hanna, D. T. Nguyen, D. M. Saudek, R. C. Brandon, L. D. Fine, J. L. Fritchman, J. L. Fuhrman, N. S. M. Geoghagen, C. L. Gnehm, L. A. McDonald, K. V. Small, C. M. Fraser, H. O. Smith, and J. C. Venter. 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269:496-512.

PAGE 117

103 21. Fried, M., and D. M. Crothers. 1981. Equilibria and kinetics of lac repressoroperator interactions by polyacrylamide gel electrophoresis. Nucleic Acids Res. 9:6505-6525. 22. Glaser, J. H., and J. A. DeMoss. 1971. Phenotypic restoration by molybdate of nitrate reductase activity in chlD mutants of Escherichia coli. J. Bacteriol. 108:854-860. 23. Graham. A., H. E. Jenkins. N. H. Smith, M. A. Mandrand-Bertholet, B. A. Haddock, and D. H. Boxer. 1 980. The synthesis of formate dehydrogenase and nitrate reductase proteins in various fdh and chl mutants of Escherichia coli. FEMS Microbiol. Lett. 7:145-151. 24. Grunden, A. M., R. M. Ray, J. K. Rosentel, F. G. Healy, and K. T. Shanmugam. 1996. Repression of the Escherichia coli modABCD (molybdate transport) operon by ModE. J. Bacteriol. 178:735-744. 25. Gunsalus. R. P. 1992. Control of electron flow in Escherichia coli: coordinated transcription of respiratory pathway genes. J. Bacteriol. 174:7069-7074. 26. Hamilton, C. M, M. Aldea, B. K. Washburn, P. Babitzke, and S. R. Kushner. 1989. New method for generating deletions and gene replacements in Escherichia coli. J. Bacteriol. 171:4617-4622. 27. Hasona, A. 1996. Personal communication. 28. Hattori, M., and Y. Sakaki. 1986. Dideoxy sequencing method using denatured plasmid templates. Anal. Biochem. 152:232-238. 29. Hemschemeier, S., M. Grund, B. Keuntje, and R. Eichenlaub. 1991. Isolation of Escherichia coli mutants defective in uptake of molybdate. J. Bacteriol. 173:64996506. 30. Higgins, C. F. 1992. ABC transporters: From microorganisms to man. Ann. Rev. Cell Biol. 8:67-113. 3 1 Hinton, S. M., and D. Dean. 1990. Biogenesis of molybdenum cofactors. Crit. Rev. Microbiol. 17:169-188. 32. Hinton, S. A., C. Slaughter, W. Eisner, and T. Fisher. 1987. The molybdenumpterin binding protein is encoded by a multigene family in Clostridium pasteurianum. Gene. 54:211-219.

PAGE 118

104 33. Johann, S. and S. M. Hinton. 1987. Cloning and nucleotide sequence of the chlD locus. J. Bacteriol. 169:1911-1916. 34. Johnson, J. L. 1980. in Molybdenum and molybdenum containing enzymes. M. P. Coughlan, ed. Pergamon Press, Oxford, pp. 345-383. 35. Johnson, J. L., N. R. Bastian, and K. V. Rajagopalan. 1990. Characterization of the molybdenum cofactor of sulfite oxidase, xanthine dehydrogenase, and nitrate reductase. Proc. Natl. Acad. Sci. USA. 87:3190-3194. 36. Johnson, J. L., B. E. Hainlain. K. V. Rajagopalan. and B. H. Arison. 1984. The pterin component of molybdenum cofactor. Structural characterization of two fluorescent derivatives. J. Biol. Chem. 259:5414-5422. 37. Johnson, J. L. and S. K. Wadman. 1989. Molybdenum cofactor deficiency. In: The metabolic basis of inherited disease. C. R. Scriver, et al. eds. McGraw-Hill, New york. pp.1463-1475. 38. Kalman. L. V. and R. P. Gunsalus. 1990. Nitrateand molybdenum-independent signal transduction mutations in narX that alter regulation of anaerobic respiratory genes in Eschericia coli. J. Bacteriol. 172:7049-7056. 39. Khavari, P. A., C. L. Peterson, J. W. Tamkun, D. B. Mendel, and G. R. Crabtree. 1993. BRG1 contains a conserved domain of the SW12/SNF2 family necessary for normal mitotic growth and transcription. Nature. 366:170-174. 40. Lawson, C. L., and J. Carey. 1993. Tandem binding in crystals of a trp repressor/ operator half-site complex. Nature 366:178-182. 41 Lee, J. H., J. C. Wendt and K. T. Shanmugam. 1990. Identification and characterization of a new gene, molR, essential for utilization of molybdate by Escherichia coli. J. Bacteriol. 172:2079-2087. 42. Li, J., S. Kustu, and V. Stewart. 1994. In vitro interaction of nitrate-responsive regulatory protein NarL with DNA target sequences in the fdnG, narG, narK, and frdA operon control regions of Escherichia coli K12. J. Mol. Biol. 241:150-165. 43. Luque, F., L. A. Mitchenall, M. Chapman, R. Christine, and R. N. Pau. 1993. Characterization of genes involved in molybdenum transport in Azotohacter vinelandii. Mol. Microbiol. 7:447-459. Maggi, A. M. A. 1991. Unpublished. Genbank Accession number X61098.

PAGE 119

105 45. Maniatis. T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 46. Maupin-Furlow J. A., J. K. Rosentel, J. H. Lee, U. Deppenmeier, R. P. Gunsalus, and K. T. Shanmugam. 1995. Genetic analysis of the modABCD operon (molybdate transport) of Escherichia coli. J. Bacteriol. 177:4851-4856. 47. Maupin, J. A. and K. T. Shanmugam. 1990. Genetic regulation of formate hydrogenlyase of Escherichia coli: role of fhlA gene product as a transcriptional activator for a new regulatory gene,/fc/5. J. Bacteriol. 176:4416-4423. 48. Miller, J. B., D. J. Scott, and N. K. Amy. 1987. Molybdenum-sensitive transcriptional regulation of the chlD locus of Escherichia coli. J. Bacteriol. 169:1853-1960. 59. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor. NY. 50. Mouncey, N. J., L. A. Mitchenall, and R. N. Pau. 1995. Mutational analysis of genes of the mod locus involved in molybdenum transport, homeostasis, and processing in Azotobacter vinelandii. J. Bacteriol. 177:5294-5302. 51. Ogden, R. C. and D. A. Adams. 1987. Electrophoresis in agarose and acrylamide gels. Methods in Enzymology. 152:61-87. 52. Pabo, C. O. and R. T. Sauer. 1992. Transcription factors: structural families and principles of DNA recognition. Ann. Rev. Biochem. 61:1053-95. 53. Pateman, J. A., D. J. Cove, B. M. Rever, and D. B. Roberts. 1964. A common cofactor for nitrate reductase and xanthine dehydrogenase which also regulates synthesis of nitrate reductase. Nature. 201:58-60. 54. Phillips, S. E. V., I. Manfield, I. Parsons, B. E. Davidson, J. B. Rafferty, W. S. Somers, D. Margarita, G. N. Cohen, I. Saint-Girons, and P. G. Stockley. 1989. Cooperative tandem binding of met repressor of Escherichia coli. Nature 341 71 1715. 55. Phillips, S. E. V., and P. G. Stockley. 1994. Similarity of met and trp repressor. Nature 368:106. 56. Rabin, R. S., L. A. Collins, and V. Stewart. 1992. In vivo requirement of integration host factor for nar (nitrate reductase) operon expression in Escherichia coli K12. Proc. Natl. Acad. Sci. USA. 89:8701-8705.

PAGE 120

106 57. Rajagopalan. K. V., and J. L. Johnson. 1992. The pterin molvbdenum cofactors. J. Biol. Chem. 267:10199-10202. 58. Rech, S., U. Deppenmeier. and R. C. Gunsalus. 1995. Regulation of the molybdate transport operon, modABCD, of Escherichia coli in response to molybdate availability. J. Bacteriol. 177:1023-1029. 59. Rech, S., C. Wolin and R.P. Gunsalus. 1996. Properties of periplasmic ModA molybdate-binding protein of Escherichia coli. J. Biol. Chem. 271:2557-2562. 60. Rosentel. J. K. 1995. The molybdate transport system of Escherichia coli. Master thesis, University of Florida, Gainesville, Florida. 61 Rosentel, J. K., F. Healy, J. A. Maupin-Furlow, J. H. Lee. and K. T. Shanmugam. 1995. Role of molybdate transport system(s) in the regulation of formate hydrogenlyase synthesis in Escherichia coli. J. Bacteriol. 177:4857-4864. 62. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chainterminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. 63. Schlensog, V.,A. Birkmann. and A. Bock. 1989. Mutations in trans which affect the anaerobic expression of formate dehydrogenase (fdhF) structural gene. Arch. Microbiol. 152:83-89. 64. Schlensog, V, A. and A. Bock. 1990. Identification and sequence analysis of the gene encoding the transcriptional activator of the formate hydrogenlyase system of Escherichia coli. Mol. Microbiol. 4:1319-1327. 65. Schroeder, I., S. Darie, and R. P. Gunsalus. 1993. Activation of the Escherichia coli nitrate reductase (na/GHJI) requires integration host factor. J. Biol Chem 269:771-774 66. Schweizer, H. P. 1 993 Two plasmids, X 1 9 1 8 and Z 1 9 1 8, for easy recovery of the xylE and lacZ reporter genes. Gene. 134:89-91. 67. Scott, D. J., and N. K. Amy. 1989. Molybdenum accumulation in chlD mutants of Escherichia coli. J. Bacteriol. 171:1284-1287. 68. Self, W. T., 1996. Personal communication.

PAGE 121

107 69. Shanmugam. K. T., V. Stewart, R. P. Gunsalus. D. H. Boxer, J. A. Cole, M. Chippaux. J. A. DeMoss, G. Giordano, E. C. C. Lin, and K. V. Rajagopalan. 1992. Proposed nomenclature for the genes involved in molybdenum metabolism in Escherichia coli and Salmonella typhimurium. Mol. Microbiol. 6:3452-3454. 70. Smith. P. K., R. I. Krohn, G. T. Heranson, A. K. Mallia, F.H. Ganther, M. D. Provanzano, E. K. Fujimoto, N. M. Goeke, B. J. Olson, and D. C. Klenk. 1985. Measurement of protein using bicinchoninic acid. Anal. Biochem. 150:76-85. 71 Sperl, G. T. and J. A. DeMoss. 1975. chlD Gene function in molybdate activation of nitrate reductase. J. Bacteriol. 122:1230-1238. 72. Stewart, V. 1988. Nitrate respiration in relation to facultative metabolism in enteric bacteria. Microbiol. Rev. 52:190-232. 73. Stewart, V., and C. H. MacGregor. 1982. Nitrate reductase in Escherichia coli K12: involvement of chlC, chlE, and chlF loci. J. Bacteriol. 151 :788-799. 74. Studier, W. F. and B. A. Moffatt. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol 1 89* 113130. 75. Walkenhorst, H., S. Hemschemeier and R. Eichenlaub. 1995. Molecular analysis of the molybdate uptake operon, modABCD, of Escherichia coli and modR a regulatory gene. Microbiol. Res. 50:347-361. 76. Wang, G., S. Angermuller and W. Klipp. 1993. Characterization of Rhodobacter capsulatus genes encoding a molybdenum transport system and putative molybdenum-pterin-binding proteins. J. Bacteriol. 175:3031-3042. 77. Wootton J. C, R. E. Nicolson, J. M. Cock, D. E. Walters, J. F. Burke, W. A. Doyle and R. C. Cray. 1991. Enzymes depending on the pterin molybdenum cofactor: sequence families, spectroscopic properties of molybdenum and possible cofactor-binding domains. Biochim. Biophys. Acta. 1057:157-185. 78. Yanisch-Perron, C, J. Vieira, and J. Messing. 1985. Improved Ml 3 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

PAGE 122

BIOGRAPHICAL SKETCH Amy M. Grunden was born on January 5, 1971, in Lakeland. Florida. In May, 1993, she graduated with a Bachelor of Science degree in microbiology and cell science from the University of Florida. During her senior year of undergraduate studies, she worked in the laboratory of Dr. K. T. Shanmugam. After graduating with a B.S., she decided to continue her studies with Dr. Shanmugam in pursuance of an advanced degree and since Fall of 1994 has been enrolled in the Ph.D. program in the Department of Microbiology and Cell Science. 108

PAGE 123

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. IcTT. K. T. Shanmugam. Chairman Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. /ohn E. Gander ^Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is folly adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Lonnie O. Ingram A Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosoi tit James F. Preston III 'rofessor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy / / Philip J. Laipis/ [/ Professor of Biochemistry and Molecular Biology

PAGE 124

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December. 1996 ^ Dean. College of Agriculture Dean. Graduate School