f

.9. ‘3'
1,111 11nd éukh “1‘1"
6; F’L "" . |>

t

‘ Q~OA '
..- ~_- ..
. . air-‘N

.

    
   

J- N... ‘

fliIv.

 

.
.-‘..

5

pro} ~
“h

' .n...->

' “Writ?

1.

.

fl....

L..."\:-
.-.- -
‘4'.

)2.

“Yo~’

 

-v

< . -w—w.‘r.

 

 

:' ’~
3:1'5‘11 1111‘

fl...”
P...—
4....

."-l‘r<<

-40

..
a‘
I

-t
—.4. «-

l :i
v ' [t
.101 V‘ ~,
. l‘
21.: '-

{a }.'1J-.-\ mil”.
‘1"M3“1 ‘ - ,
“611111.? ii v L‘I‘. 11“}:41?

:11; \

1H1!” [Q Wu H ‘ 3 d1;
. , i‘i.‘ “gm.
M uéifi”“p‘“1“‘ Mil: 11:1 5‘

1x

1L~ KvI" "'13¥ "
m. r“. .. _ i..." ' ‘- 1,4“ 1H"“~w ‘,«~
... . - g. g2,
1111.11'H3! ‘~ , 11 1-111 9; .1 51.11.! H11 .1: W 113791

 

.
m‘ s

i
H

11W. W111?“ 1...“)

   

THEsm

2

 

HIGAN STATE UNNE SlTY LIBRAR

Ill/1°11 1111111111111111111111171

 

 

 

l I!

3 1293 01581 4423

LIBRARY
Mlchlgan State
University

 

 

 

 

This is to certify that the

dissertation entitled
Identification of _I_r_i Vivo Induced Genes

in Actinobacillus pleuropneumoniae

presented by
Troy Eugene Fuller

has been accepted towards fulfillment
of the requirements for

Ph 0 D 0 degree in MiCI‘Ob 10 logy

 

Major professor

Date 03‘3/‘77

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

PLACE ll RETURN BOX to remove We checkout from your record.
TO AVOID FINES return on or before dete due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

- l l

lCl—l
__Jl::l__l '
:JLJE

MSU leAn Atflrmetlve Action/Equel Opportunity Inetltulon
m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

uni-9.1

IDENTIFICATION OF IN VIVO INDUCED GENES IN ACTINOBACILLUS
PLEUROPNEUMONIAE

By

Troy Eugene Fuller

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Microbiology

1997

ABSTRACT

IDENTIFICATION OF IN VIVO INDUCED GENES IN ACTINOBACILLUS
PLEUROPNEUMONIAE

By

Troy Eugene Fuller

Actincbacillus pleumpneumoniae (APP) is the causative agent of porcine
pleuropneumonia. a severe and often fatal respiratory disease in swine. In order to
develop effective vaccines and antibiotics, there is a need for a better understanding of
the pathogenesis of APP. The goal of this research was to identify APP genes that are
specifically induced in vivo during infection.

The 4300 bp riboflavin operon (n'bGBAH) of APP serotype 5 was cloned and
sequenced. The biosynthetic functions of these genes were conflnned by sequence
homology with known riboflavin biosynthesis genes, by complementation of known
Escherichia coli riboflavin mutants, by analysis of the plasmid encoded proteins and by
analysis of the recombinant riboflavin product including spectral analysis, mass
spectroscopy and fluorescence. A deletion dismption riboflavin-requiring mutant of a
serotype 1 strain was made by targeted mutagenesis of the riboflavin operon. This
mutant (AP233) is avirulent at a dosage which is 500 times the established wild type
L050. AP233, as a live vaccine, provides a degree of protection against both serotype
1 and serotype 5 strains when administered intramuscularly and supplemented with
enough riboflavin to permit two to three generations of growth in viva.

An IVET (in vivo expression technology) system was constructed for APP that
is able to identify genes that are induced in viva, i.e., during infection, and not in vitro.
An IVET vector was assembled using the plasmid pGZRS-19 containing in sequence

the T4 terminator, a unique BamHl cloning site, promoterless quAB genes from an’o

harveyi and a pramaterless Bacillus subtilis n'bBAH operon. It was demonstrated that
wild type APP contains no native bioluminescence, that the B. subtilis n’bBAH genes
can complement the attenuating riboflavin mutation in AP233 thus restoring virulence
and that the T4 terminator effectively eliminates background expression from the
pGZRS-19 vector. This IVET system was used to identify five in viva induced genes in
APP, two of which were putatively identified as the mrp gene and the secE-nusG
operon. The further identification of in viva induced genes will help us better

understand and control the pathogenesis of APP.

Dedicated to:
Michelle Fuller

James Fuller and Karen Hanes

iv

 

Tl'

Dr

Jar

Mic:

ACKNOWLEDGMENTS

My years in graduate school could not have been accomplished without the support of

many. I wish to thank the following:

Dr. Martha H. Mulks, for all of your guidance, mentoring bath in and out of the
laboratory and far dedicating yourself to my professional development as a scientist

Dr. Brad Thacker, far all of your expertise and help with the animal experiments.

Drs. Pat Oriel, Michael Bagdasarian, Paul Caussens, Linda Mansfield and Ron
Patterson, members of my graduate committee for your advice, assistance and
excellent questions.

Drs. Sadhana (Chauhan) Matin, Vladimir Matin and Wilma Cruz, far patiently helping
me get started during my first year and for teaching me many valuable techniques.
The United States Department of Agriculture far 3 years of support through a National

Needs Graduate Fellowship in Animal Biotechnology

The DuVall family for their gracious financial support during my last year

Dr. David Kreps, my first mentor and inspiration for pursuing microbiology as a
profession

James and Sandra Fuller and Karen and Robert Hanes, my parents, for all at your love
and support and for allowing meta pursue my own interests and to make my own
decisions.

Michelle Fuller, my wife, without whom my accomplishments would not matter.

PREFACE

This thesis includes four chapters that are in manuscript format. The second
chapter entitled ”Characterization of Actinabacillus pleurapneumaniae Riboflavin
Biasynthesis Genes" is a full length version of a manuscript that was published in note
format in the Journal of Bacteriology Volume 177 (1995) pages 7265-7270, with Dr.
Martha H. Mulks as ca-author. Chapter three entitled ”A Riboflavin Auxatraph of
Actinabacillus pleurapneumaniae is Attenuated in Swine” was published in lnfecfian
and Immunity Volume 64 (1996) pages 4659-4664, with Dr. Martha H. Mulks and Dr.
Brad J. Thacker as ca-authars. Chapters four and five will be submitted to Infection
and Immunity with Dr. Martha H. Mulks and Dr. Brad J. Thacker as ca-authars. As a
ca-authar, Dr. Brad Thacker provided veterinary care and expertise, clinical evaluation

and pathological interpretation for the animal experiments.

TABLE OF CONTENTS

UST OF TABLES ...................................................................................................... ix
LIST OF FIGURES ..................................................................................................... x
CHAPTER 1
GENERAL INTRODUCTION ...................................................................................... 1
Actinabaciilus pleurapneumaniae ................................................................... 2
Epidemiology and Pathology ................................................................ 2
\frmlence Factors .................................................................................. 3
Prevention and Control ........................................................................ 9
Genetics ............................................................................................. 13
Genetic Technology for the Study of Pathogenesis ...................................... 15
In Viva Expression Technology .......................................................... 15
Signature-Tagged Mutagenesis ......................................................... 18
Resalvase-Based IVET ...................................................................... 18
Subtractive Hybridization .................................................................... 19
References .................................................................................................... 21
CHAPTER 2.
CHARACTERIZATION OF ACTINOBA CILLUS PLEUROPNEUMONIAE
RIBOFLAVIN BIOSYNTHESIS GENES .................................................................. 36
Abstract ......................................................................................................... 37
Introduction ................................................................................................... 38
Materials and Methods .................................................................................. 43
Results .......................................................................................................... 46
Discussion ..................................................................................................... 59
Acknowledgments ......................................................................................... 61
References .................................................................................................... 62
CHAPTER 3
A RIBOFLAVIN AUXOTROPH OF ACTINOBA CILLUS PLEUROPNEUMONIAE
IS ATTENUATED IN SWINE .................................................................................... 66
Abstract ......................................................................................................... 67
Introduction ................................................................................................... 68
Materials and Methods .................................................................................. 71
Results .......................................................................................................... 76
Discussion ..................................................................................................... 84
Acknowledgments ......................................................................................... 86
References .................................................................................................... 87

CHAPTER 4
A RIBOFLAVIN AUXOTROPH AS A LNE-ATTENUATED VACCINE

AGAINST ACTINOBA CILLUS PLEUROPNEUMONIAE ......................................... 91
Abstract ......................................................................................................... 92
Introduction ................................................................................................... 93
Materials and Methods .................................................................................. 96
Results ........................................................................................................ 100
Discussion ................................................................................................... 106
Acknowledgments ....................................................................................... 1 10
References .................................................................................................. 1 1 1

CHAPTER 5

IDENTIFICATION OF IN VIVO INDUCED GENES IN

ACTINOBA CILLUS PLEUROPNEUMONIAE ........................................................ 115
Abstract ....................................................................................................... 1 16
Introduction ................................................................................................. 1 17
Materials and Methods ................................................................................ 123
Results ........................................................................................................ 127
Discussion ................................................................................................... 138
Acknowledgments ....................................................................................... 141
References .................................................................................................. 142

CHAPTER 6

SUMMARY ............................................................................................................. 145

Table 2.1

Table 3.1
Table 3.2
Table 3.3
Table 4.1
Table 4.2
Table 4.3
Table 4.4

Table 4.5

LIST OF TABLES

Comparison of Amino Acid Sequence Similarity Between

Riboflavin Synthesis Proteins ............................................................ 54
Characteristics of Bacterial Strains and Plasmids ............................. 72
Mortality and Lung Scare Data .......................................................... 83
Clinical Scare Data ............................................................................ 83

Seralagic Analysis of Samples Collected at Challenge for Trial 1.... 103

Mortality and Lung Scare Data for Trial 1 ........................................ 104
Clinical Sign Data for Trial 1 ............................................................ 104
Mortality and Lung Scare Data for Trial 2 ........................................ 105
Clinical Sign Data for Trial 2 ............................................................ 105

 

 

Fl
Fl

Figure 2.1
Figure 2.2
Figure 2.3

Figure 2.4

Figure 2.5
Figure 2.6
Figure 3.1
Figure 3.2

Figure 3.3

Figure 3.4
Figure 5.1
Figure 5.2

Figure 5.3

Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7

LIST OF FIGURES

Proposed Bacterial Riboflavin Biasynthesis Pathway ................... 39-40
Physical Map of the Construct pTF10 ................................................ 47
Spectal Analysis of Riboflavin and Recombinant Riboflavin ............. 48
Complete Nucleotide Sequence of APP ribGBAH Genes

and Flanking Regions ................................................................... 49-53
Camplementatian of E. cali Mutants by Cloned APP rib Genes ........ 57
Minicell Analysis of pTF10 and Deletions .......................................... 58
Construction of pTF67a ..................................................................... 77
Southern Blots of Hindlll or EcaRl Digested DNA from Mutants

and Controls ...................................................................................... 78
Schematic Structure of the rib Locus of Parent and Mutant

Strains in Double and Single Crass-Over Events .............................. 79
Phenotypic Analysis of AP100, AP225 and AP233 ........................... 81
Construction of the IVET Vector pTF86 ........................................... 129
Identification of In Viva Induced Genes ............................................ 130
Primary Screening for Minimally Expressed IVET Clones

in AP233 ........................................................................................... 132
In Vitra Quantitative Lux Analysis of Selected IVET Clones ............. 133
Amino Acid Alignment of IVA and Mrp .............................................. 134
Amino Acid Alignment of IVF and SecE/NusG ................................. 136
In Viva Induction of IVET Clones ...................................................... 137

Chapter 1

GENERAL INTRODUCTION

 

Th
are
net
Vulr
stag

"eul

inflar.

fibfin

2
ACTINOBA CILLUS PLEUROPNEUMONIAE

1. Epidemiology and Pathology:

Actinabacillus pleurapneumaniae (APP) is a gram negative, encapsulated,
hemolytic, pleiamarphic rod belonging to the family Pasteurellaceae. APP strains can
be divided into two biatypes based on their nicotinamide adenine dicnucleatide
dependence (biatype 1) or independence (biatype 2). APP is the etiologic agent of
porcine contagious pleurapneumania, a severe and often fatal respiratory disease of
swine (38,112,141). The disease has become a major economic problem in swine-
raising areas throughout the world since its discovery and identification over 30 years
ago (38,67,102,112,119,123,141,142). Pleuropneumonia is typically characterized as
an acute, necrotizing, exudative bronchopneumonia with accompanying hemorrhage,
vascular injury and fibrinous inflammation of the pleural cavity (8,30,38,94,112,141,168).
Increased industrialization of swine farms has led to situations which are conducive to
transmission and infection. Disease outbreaks are frequently associated with stresses
which might depress normal respiratory defenses such as transportation, abrupt climatic
change, overcrowding or poor ventilation (38,112,139,141).

The disease may occur in a number of forms ranging from a peracute phase
which normally results in high morbidity and mortality, to a chronic carrier phase (112).
The disease most commonly effects 3 to 5 month-old pigs even though pigs of all ages
are susceptible. It has been proposed that the gradual loss of maternal toxin-
neutralizing antibodies is the reason that animals at the age of 12 weeks are particularly
vulnerable in an endemically infected herd (20,21). The pathology during the initial
stages of pleurapneumania includes congestion and platelet aggregation, edema, and
neutrophil infiltration, as well as development of multifocal petechial lesions (8,94,168).
Wlth progression, the alveoli, bronchioles, and bronchi become filled with an
inflammatory exudate that contains macrophages, dead and degenerate neutrophils,

fibrin, hemorrhage, and bacteria (112). It has also been suggested that the severity of

’6

"It

do

he;

"Wt

Ant)”.

3

clinical signs may be enhanced by a synergistic co-infection with viral agents, such as
pseudorabies virus (138). This synergism may be due to viral mediated hindrance of
normal physical respiratory defense mechanisms such as mucociliary clearance (111).

2. Virulence Factors:

Several proven and suspected virulence factors of APP include capsular
polysaccharide (CPS), lipopolysaccharide (LPS), extracellular toxins, and surface
proteins (OMPs). These factors may contribute to the ability to survive in viva by
escaping from or neutralizing host defenses.

There are currently 12 recognized serotypes based on CPS, the serotype-
specific antigen of APP. Seratypes 1 and 5 can both be divided further into subtypes A
and B. The composition and structure of the CPS of all 12 serotypes has been
determined (124). CPS does not cause pulmonary lesions when administered
endobronchially (41,68), does not demonstrate toxic activity (39,41,68) and does not
activate the complement cascade (68). The role of CPS is believed to be protection
against specific (bactericidal killing) and nonspecific (phagocytosis) respiratory defense
mechanisms because encapsulated APP strains are resistant to antibody and
complement mediated killing and to phagocytosis by PMNs (68,69,134). The
mechanism for this resistance appears to be inhibition of the binding of C9 and
membrane attack complex to the cells (163). Non-encapsulated mutants of APP
serotypes 1 and 5 have been shown to be at least 10 fold less virulent than their
respective wild type strains (71,130) although this work was done with chemically
mutagenlzed bacteria leading to the possibility of additional unknown attenuating
mutations. The gene cluster involved in the biosynthesis of CPS in APP-5 has been
cloned, and two genes homologous to polysaccharide export genes from other gram
negative bacteria have been identified (164). Further work with genetically defined
mutants has recently been reported to confirm the previous attenuation findings (163).
Anti-CPS antibodies provide serotype-specific protection against severe disease

 

fa

lik
LF
Dn
inn
res1

(4,5

”38,
expo

4

symptoms, but they do not prevent colonization or chronic infection (68,69,113,131,132)
suggesting the importance of other virulence factors.

LPS, or endotoxin, is produced by all gram-negative bacteria. As with CPS, the
composition and structure of the O-antigens (the outermost oligosaccharide chain of
LPS) has been determined for all 12 serotypes of APP (124). Structural similarities
between LPS O-side chains of some serotypes may explain many cross reactions in
serotyping. Although serotyping by antigenic differences in CPS is standard, it has been
proposed that APP strains be classified using both LPS (0) and capsule (K)
designations, as is done with other gram-negative bacteria (9,110,124). However, the
creation of additional subtypes using multiple antigens would likely lead to a decreased
reliability in the typing due to antisera variation.

Except for the typical hemorrhagic necrosis, LPS may cause many of the
pathologic lesions in the lungs of swine infected with APP, including inflammatory
edema, neutrophil infiltration, increased capillary permeability, hemorrhage into alveoli,
intravascular platelet aggregation and fibrinous thrombosis (41,154). Death during
acute infection may even be due to septic shock rather than to the pulmonary ventilation
failure associated with extensive lung damage (94). LPS is known to specifically induce
activated complement, production of fibrin and production of pro-inflammatory cytokines
like interieukin 1 and tumor necrosis factor (10,24,105). Administered intratracheally,
LPS can induce lesions ranging from mild interstitial pneumonia to consolidating lobular
pneumonia with inflammatory cell infiltration (41,154). In addition to its role in
inflammatory reactions, smoofl'l LPS plays a role in adherence of APP to porcine
respiratory tract cells, to mucus secretions and to porcine tracheal rings in organ culture
(4,5122). The lipid A-core region also has been shown to bind porcine hemoglobin (3).

Immunization of swine with a J5 strain of E. cali, which produces a non-toxic
LPS, provided a significant reduction in clinical signs and protection from mortality in

experimental infections and field trials; however, it did not prevent APP infection (37,40).

 

Serc

bind.

5

Possible reasons for this lack of protection from infection might be that antibodies
against LPS may be blocking the ability of complement components to bind to APP cells
(165), the LPS may be too dissimilar, or perhaps it is even partially hidden due to CPS.

Three extracellular RTX (Repeat-ln-Toxin) toxins have been described in APP
and are termed Apxl, II and Ill. An additional cryptic hemolysin, the source of some
confusion, has also been identified. All field isolates of APP produce one or more of
these toxins depending on the serotype (2,85). The genes encoding all three Apx toxins
have been sequenced by multiple laboratories (14-6,43,46,49,79-81,83). Apxl and
Apxlll are encoded by typical RTX toxin operons, consisting of four contiguous genes,
apxCABD (14,43,46,49,79,80,83), encoding respectively the toxin activator, the pro-
toxin and the two products necessary for a single step secretion pathway. The Apxll
operon is truncated, containing only the apxC and apr genes (15,81,96); flierefore,
secretion of Apxll is apparently mediated through Apxl or Apxlll B and D (81,96). RTX
toxins kill eukaryotic cells by forming pores in the cell membrane. Biophysical studies
have recently determined the channel size and channel-fanning activity of all three Apx
toxins (101), and me pore forming activity appears to directly correlate with the activity of
the toxins.

Apxl is a strongly hemolytic and cytolytic toxin produced by serotypes 1, 5a, 5b,
9, 10, and 11 (44,50-2,85). Like other RTX toxins, the Apxl toxin is a pore-fanning toxin
which binds calcium and requires calcium for activity, specifically for binding to target
cells such as erythrocytes and neutrophils (28,49,90,160). Expression of Apxl is
regulated at the transcriptional level by calcium through an unknown mechanism
(46,52,63). Apxll is a weakly hemolytic and weakly cytolytic toxin produced by all APP
serotypes except serotype 10 (45,52,54,85,86,144). The Apxll toxin requires calcium for
binding to target cells, but unlike Apxl its production is not induced by calcium (160).
The cytotoxic and hemolytic domains of this toxin are distinct and separable with the

hemolytic activity being more dependent on activation and secretion than cytotoxic

6

activity (103,152). Apxlll requires calcium for binding to porcine neutrophils and is
strongly cytotoxic for porcine alveolar macrophages and neutrophils; however, unlike
Apxl and Apxll it is not able to bind to erythrocytes and is therefore not hemolytic
(85,95,136,160). It is produced by serotypes 2, 3, 4. 6 and 8 (85,95).

In addition to the three Apx toxins, several laboratories have identified an APP
gene encoding a cryptic hemolysin, Hlyx, which is expressed in E. cali but not in APP
(53.76.93). HlyX is homologous to the FNR protein of Escherichia cali, which regulates
several genes involved in anaerobic respiration (60,97,145). It was originally thought
that the HlyX was responsible for the CAMP reaction hemolysis seen with APP, but now
this phenomena has been attributed to the Apx toxins (47,78). There is currently no
evidence that the thX gene is involved in vimlence in A. pieuropneumaniae, and me
reason for its hemolytic activity in E. coli remains to be determined. Perhaps it regulates
the expression of yet another cryptic hemolysin in E. coli.

The APX toxins are perhaps the most studied vlmlence factors in the
pathogenesis of APP. APP culture supematants are toxic in vitro for a wide variety of
porcine cells including erythrocytes. pulmonary macrophages and neutrophils
(7,85,129,160.161). The Apxl and Apxll toxins from serotype 1 have been shown to
stimulate the oxidative metabolism of porcine PMNs at low doses (153). kill PMNs at
higher doses (153), and damage porcine alveolar macrophages (18,147). Likewise, the
Apxll and Apxlll toxins from serotype 2 sfimulate neutrophils and porcine pulmonary
alveolar macrophages at low doses and kill at higher doses (31,33). The activities of
these toxins in viva probably interfere with pulmonary macrophage and neutrophil
function, thereby decreasing effective phagocytosis and killing. This lack of clearance
could then allow the rapid multiplication of bacterial cells leading to acute
pleurapneumania. In addition. damage to phagocytic cells will result in the release of
lysosomal contents and oxygen radicals contributing to massive inflammation and

necrosis.

 

(8
hr
9r;
ha
Sec
inflr

7

Toxin negative mutants of APP have been shown to be avirulent, with the degree
dependent on which toxins are knocked out. in vivo studies using isogenic toxin-
negative mutants conclusively prove that the RTX toxins are critical to the pathogenesis
of APP infection. Even though toxin is critical to the virulence of the organism, toxin
neutralizing antibodies are not sufficient to completely protect against pleurapneumania.
Mutants of APP-1 lacking Apxl and Apxll do not activate or kill porcine neutrophils (82)
and do not cause significant clinical signs or lesions in pigs at a dose that is normally
100% lethal with a wild type APP-1 strain (148). In contrast, APP-1 mutants that can still
produce Apxll but which lack Apxl are still pathogenic, although they are less vimlent
than wild type APP-1 (149). An Apxl-, Apxll- mutant of APP-5 caused neither mortality
nor lung lesions in either mice or pigs, even when tested at 10 times the L050 for the
wild type strain (70). An Apxll- Apxlll- mutant of APP-2 was also attenuated compared
to the wild type strain (23). However. an Apxll- Apxlll+ mutant of APP-2 was still able to
cause pleurapneumania and typical acute lesions in pigs, although the disease in
animals infected with this mutant was less severe than in those infected with the wild
type strain (137). An Apxll- mutant of APP-7 was also attenuated in pigs and could not
colonize the host as well as the wild type strain (58). These results suggest that the
Apxll toxin is critical when it is the only toxin produced.

Outer membrane protein (OMP) profiles of APP strains show three major
common proteins of 38-42 kDa, 29 kDa, and 16—17 kDa, as well as 10-20 minor proteins
(98.108.126.135). The 29 kDa OMP is a heat-modifiable protein which has not been
further characterized (126). The 16-17 kDa OMP was reported initially as a common
gram negative antigen (98), and the gene has recently been cloned, characterized and
named paIA (48). It is predicted to be a 16.2 kDa protein which is homologous in
sequence to the peptidoglycan-associated lipoprotein of E. cali and the Haemophilus
influenzae OMP P6 (48). Amino terminal sequencing of a 40 kDa major OMP from APP
suggests that it is related to a porin protein from Pasteurella mulfacida (64). Our lab has

tested purified OMPs as subunit vaccines; however, vaccination of pigs with a purified
38-42 kDa OMP or with a 48 kDa OMP (AopA) (22) did not provide protection against
APP challenge (106). A 42 kDa OMP expressed by all APP serotypes. which is
homologous to the P2 OMP of H. influenzae. does provide good protection against APP
challenge when used in combination with Apx toxins in a subunit vaccine (158-9).

A common host defense mechanism is the sequestration of iron from the
pathogen. APP responds to conditions of iron-limitation by inserting new OMPs into its
outer membrane including OMPs of 60-65. 76-79, and 99-105 kDa rather than by
production of iron scavenging siderophores. These iron regulated OMPs are expressed
in viva as detected by convalescent sera from infected pigs (26.57.117.128). The
sarkosyl-soluble 60-65 kDa OMP (T fbA) is a transferrin binding outer membrane
lipoprotein homologous to other prll type proteins (57.59). Similar be proteins are
made by all 12 APP serotypes. although there are three antigenically distinct families
(59). The beA protein from APP-7 has been tested in combination with Apxll from APP-
7 as a vaccine, providing significant protection against challenge with APP-7 but not with
APP-1 (133). The gene for the type one transferrin binding protein (tbpl) is located
downstream of fbpll (ffbA) and has recently been cloned and sequenced from serotype
1 (25).

Another sarkosyl-soluble outer membrane lipoprotein (OmlA) which is
recognized by convalescent serum has been cloned from serotypes 1 and 5 (11.56.75).
Soumem blot analysis demonstrated that all APP serotypes contain DNA homologous to
the amlA gene. and immunoblots detected homologous proteins to the serotype 1 OmlA
produced by APP 2. 8, 9, 11. and 12. and to me serotype 5 OmlA in serotype 10.
Vaccination with the recombinant protein provided protection against mortality when
challenged with the homologous serotype but still showed limited protection against
clinical signs and lung lesions. No data on protection against heterologous challenge

has been reported with this lipoprotein.

9

There are various other environmental conditions besides iron restriction that
stimulate the production of additional OMPs. For instance, the presence of maltose in
the medium (27) and variation of NAD concentration (120) can induce the production of
unique OMPs in APP. Several of these OMPs are expressed in vivo as detected by
convalescent pig sera. It is unknown whether they have any significant role in
virulence.

Adhesins are important for attaching and localizing the pathogen to invade or to
prevent physical clearance by the mucosal ciliary system. When the normal mucosal
ciliary system is inhibited by atropine or xylocaine. infection rapidly progresses seeming
to implicate adhesion as an important process. APP adheres to porcine tracheal rings,
tracheal epithelial cells, lung tissue, respiratory mucus, and hemoglobin (3,4,6,77,122)
and has been shown to closely associate with alveolar and bronchial epithelial cells in
viva during the initial stages of infection (32). The high molecular weight O antigens of
lipopolysaccharide have been identified as potential adhesins (4,122), while capsule
seems to inhibit adherence (6,77). APP cells freshly isolated from infected pigs have
also been shown to possess fimbriae that were rapidly lost upon subculture (155), but
their role in attachment has yet to be proven.

3. Prevention and Control:

Rapid progress has been made in the basic science of Actinobacillus
pleurapneumaniae (APP) in the past several years, yet APP continues to be a serious
worldwide economic problem in the swine industry. Current methods of treatment are
cosfly as well as fairly ineffective for preventing lung damage. Antibiotics such as
penicillin, ampicillin. chloramphenical and cotrimoxazole are fairIy effective against APP
(112) in the initial phase of the disease before significant tissue damage occurs.
Antibiotic therapy, however. will not eliminate infection in a herd as chronic infections
may persist in abscesses and tonsils of the swine or the onset of pleurapneumania may

be too rapid to treat effectively (112). Increasing consumer concerns for food safety

 

 

t1

(9
(1

0f

ho

10

also discourage the use of high levels of antibiotics for prophylactic feeding or even
treatment. In addition. the use of antibiotics encourages the development of antibiotic
resistant strains as seen with many different pathogens. Naturally occurring plasmids
have been identified in APP which encode resistance to many of the major antibiotics
currently in use. Although herd management techniques such as early segregated
weaning are gaining in popularity and effectiveness. the risk of a herd to potential
disaster is great because the animals' immune systems may not be primed specifically
or nonspecifically to prevent disease. There is also no difference in the number of days
to market or in the final weights between early segregated weaned animals and those
vaccinated and raised in the conventional manner (34).

It is more cost-effective and practical to focus on prevention rather than
treatment post-infection; however. APP prevenfion is still limited by the lack of effective
vaccines. Current commercial vaccines are killed whole cell vaccines. or bacterins,
which have been shown to induce mainly capsular polysaccharide (CPS) specific
immune responses (113). A serious limitation of these APP bacterins is their inability to
provide protection against heterologous serotypes. Unfortunately, there are no other
available types of vaccines, subunit or live, which are both economical and cross-
protective. Another limitation of the bacterin is its ability to induce mainly systemic
immunity rather than specific secretory mucosal immunity. There are many other
problems that result in vaccine failure. including production of ineffective bacterins
(150), blockage of immune response by passive immunity (151). and the existence of
many serotypes and subtypes of APP (84.116).

Many of the OMPs previously described appear to be common to all serotypes
(98.108.126.135), and several outer membrane vaccines have been tested against APP
(17.35.109.127,150). Several outer membrane vaccines prepared by sarkosyl extraction
of total APP cell membranes have provided significant if not complete protection against

homologous challenge (17,35,127). However, a vaccine which used APP-5 outer

 

d.

all

Vin

Vin
the

11

membranes prepared by sucrose density centrifugation as the antigen provided
complete protection against APP-5 challenge and also reasonable protection against a
heterologous APP-1 challenge (109). This suggests that sarkosyl soluble OMPs. such
as beA (57), OmlA (11.56.75) and our AopA (22), that are present in a sucrose density
gradient preparation, may contribute to cross-protective immunity.

Extracellular toxins should be included in a true cross-protective vaccine.
Several labs have demonstrated that a concentrated cell culture supernatant containing
hemolysin is an effective vaccine (1,35,36,87,118.150). The addition of this hemolysin-
containing supernatant to whole cell bacterins can also improve vaccine efficacy (151).
A purified cytotoxin preparation from APP-1, which likely contained both Apxl and Apxll,
conferred good protection against APP-1 challenge (29). A trivalent vaccine containing
Apxl, Apxlll, and a 42 kDa OMP conferred significant protection against challenge with
both APP-1 and APP-2 (89,158,159). In contrast, a recombinant Apxll from APP-7
provided protection against death following challenge with APP-7, but no cross-
protection against APP-1 (133), likely due to the absence of Apxl in the vaccine
preparation. A vaccine containing the Apxl protein conjugated to CPS and detoxified
LPS from APP-1 provided only minimal protection against APP-1 challenge (12).

Finally, several groups have begun testing avirulent or attenuated strains of APP
as live vaccines (71.121.130.156). While several of these have elicited good protection
and cross-protection (71,121,156), the use of live vaccines in the field is problematic,
particularly when the attenuating lesions in the vaccine strains have not been genetically
defined. Many live-attenuated vaccines have been selected by passage on laboratory
media or by chemical mutagenesis without fully understanding what causes the
attenuation. With undefined attenuating mutations the possibility of reversion to
virulence may be strong especially if the mutations are merely point mutations. The
virulent strain would also likely have a selective advantage over the attenuated strain in

the host. The protection data from these live vaccines do suggest, however, that

 

an
Sh.
me
8p;
env
Virm

have

12

rational development of attenuated strains with carefully defined deletion mutations that
limit growth in vivo and prevent reversion to wild type is a promising approach to
improved APP vaccines.

In contrast to bacterin vaccinated animals, which are protected against infection
with homologous serotypes but not against heterologous serotypes (39. 113, 115. 146).
animals infected either naturally or experimentally with a more virulent form of APP are
also strongly resistant to both homologous and heterologous serotype infection
(19.113.114.116). These reports suggest that there are common antigens expressed
during an infection that result in good cross-protective immunity but that these antigens
are not expressed, masked or are altered in killed whole cell vaccines. Work in our lab
has demonstrated that APP produces antigens which are expressed in viva but not in
vitra under standard culture conditions. It is possible that the absence of these antigens
in whole cell bacterins prepared in vifra results in reduced protection and that their
inclusion in a subunit vaccine would elicit cross-protective immunity against APP. It may
also be possible that a live strain produces immunomodulatory compounds which lead to
a different immune response and better protection.

Two methods currently exist for identifying in viva expressed antigens in APP:
immunological identification and mimicry of host conditions in laboratory media. The
immunological approach involves screening bacteria grown in vivo and in vifra by
Western blot with convalescent pig sera and then comparing the results to identify
antigens that are produced only in viva. Using immunological techniques, our lab has
shown that on SDS-PAGE gels of in viva gram APP whole cell Iysates and outer
membranes there are several bands between 106 kDa and 13 kDa which do not
appear in vitra (106). The other approach has been directed at duplicating host
environmental conditions on laboratory media to identify environmentally regulated
virulence factors that may be important in host-pathogen interaction. Both approaches

have limitations in that immunological screening mainly identifies surface-associated

 

 

Ce
int
Se

Inc,

Lair:
App

13

antigens and that media mimicry will perhaps never duplicate all of the possible host
conditions. Historically, research has focused on the studies of antigens expressed in
vitra; however, this at best only gives a fractional picture of virulence-associated
characteristics. Increasing attention is being given to the environmentally-stimulated
regulatory mechanisms that modulate bacterial attributes in the host (61,62,104).
Known environmental regulatory signals for bacterial virulence genes include
temperature, osmotic strength. pH, oxygen levels. availability of iron. calcium levels,
carbon sources, urea, growth phase. stress, starvation and availability of other nutrients
such as amino acids (61.62. 104).

Modern molecular genetic tools are rapidly becoming available to identify APP
genes that are specifically expressed in viva during infection of the natural host Some
of mesa genes will likely encode infection-associated antigens. which can then be
cloned. purified, and tested as components of subunit vaccines against APP. In
addition. other in viva expressed genes are likely to encode enzymes for critical
biochemical pathways that could be used for the development of genetically defined,
stable deletion mutants for use as modified live vaccines, and potentially for the
development of antibiotics that target these specific biochemical pathways.

4. Genetics:

Significant advances in genetic studies on A. pleurapneumaniae have been
made within the last few years. While it does not seem to be possible to transform
certain serotypes of APP at high frequency (42,92,106), plasmids can be electroporated
into many serotypes of APP, although generally not into serotype 5 (42.91.92.167).
Several types of native plasmids of APP have been described and assigned to
incompatibility groups (73.74.140.169). Several shuttle vectors that replicate in both
APP and E. coli have been constructed from these naturally occurring plasmids.
Lalonde et al (92) developed several based on an APP Inc H3 plasmid, using the native

APP chloramphenical resistance (CmR) gene. One of the plasmids can be conjugally

 

bl.
us

Drc

her

14

transferred into APP. Frey (42) constructed a set of vectors based on the RSF1010
replicon. with a chloramphenical resistance (CmR) gene derived from Staphylococcus
aureus. West at al (167) have constructed several vectors based on an APP Inc H2
plasmid that contain either the B-lactamase gene from Tn3, under the control of a
putative APP promoter, or the kanamycin resistance gene from Tn903, as well as the
lacZ alpha-complementation region containing a multiple cloning site from pUC18/19.
Hodgson et al have developed vectors which are also able to function in E. coli.
Pasteurella haemolytica and APP (66). In addition. West has constructed a cosmid
vector that works well in APP (166).

Three different systems for construction of chromosomal mutants of APP have
been described. A system for targeted mutagenesis of APP, based on a conditionally
replicating plasmid vector and insertional mutagenesis by homologous recombinafion.
has been described and used to produce hemolysin-negative mutants of APP (82).
While the initial premise of using a temperature sensitive (Ts) replicon in this system
was promising. the Ts-vector does not appear to replicate in APP. and the efficiency of
this system is quite low. Because this system is based on elecfioporation. it will not
function well with APP-5 (82). Mulks and Buysee have recently developed a system for
targeted mutagenesis of APP that utilizes a broad-host range conjugative suicide vector.
pGP704, that functions in both APP-1 and APP-5 (107). This system has been used
successfully to construct knock-out mutants of the aopA gene in APP-1 and APP-5
(107). which fail to express the 48 kDa outer membrane protein. We have also
previously reported its use to construct riboflavin mutants of the ribBAH riboflavin
biosynthetic genes in serotypes 1 and 5 (55). A transposon mutagenesis procedure,
using a mini-Tn10 derivative delivered via a conjugative suicide plasmid, can be used to
produce random mutations in the chromosome of most serotypes of APP (148). This
system has been successfully used to produce both tryptophan auxotrophs and
hemolysin-deficient mutants of APP serotype 1 (82.148).

 

 

infe

15

GENETIC TECHNOLOGIES FOR THE STUDY OF BACTERIAL PATHOGENESIS

In the search for a more complete understanding of bacterial pathogenesis, new
genetic techniques are emerging which can be used to identify genes that are
expressed specifically during infection and not during growth on standard laboratory
media. These techniques are able to identify a wide variety of gene products that are
infection-associated. These genes may produce important antigens. enzymes that are
necessary for production of critical growth factors not available in viva (e.g., purines),
factors which allow survival within host cells (e.g., enzymes that inactivate phagocytic
killing mechanisms). or toxins that are not produced under standard in vitra conditions
(104). Identification of such in viva expressed genes should lead to further insights into
the pathogenesis of APP disease, to the identification of important antigens for inclusion
in subunit vaccines and to me development of defined avirulent mutants for use as live
vaccines.

Four genetic methods for the analysis of genes involved in the pathogenesis of
bacteria have been described. The first, designated in viva expression technology
(IVET) uses survival in the natural host to identify genes specifically tumed on in viva
during infection (99,143). The second, designated signature-tagged mutagenesis
(STM). uses negative selection with unique transposon mutants to identify mutations
that adversely affect survival in the natural host (65). The third relies on a heritable
genetic change promoted by the induction of me enzyme resolvase. and the fourth is
subtractive hybridization (SH) which identifies differentially expressed genes based on
RNA profiles.

1. In vivo expression technology:

An IVET (in viva expression technology) genetic system has been developed

which allows positive selection of bacterial genes that are specifically expressed during

infection and has been used to identify in viva induced genes in the gram negative

 

Qe

Clo
i"Itr

no

16

pathogen Salmonella typhimurium (99.143). The purpose of the system is to use
reporter genes to identify promoters that are tumed on specifically in the host Once
identified. the genes regulated by these promoters can be cloned. sequenced and
studied further. There are five basic requirements for constructing an iVET system.
First, it is necessary to have a biochemical mutant that attenuates growth of the
organism in viva resulting in lack of survival. The second requirement is to have a
promoterless copy of a wild type gene from a different organism that will complement
the attenuating biochemical lesion to restore growth capability when expressed in viva.
Third, it is necessary to have an appropriate vector allowing introduction of the DNA
into the organism. Fourth, a quantifiable promoterless lacZY or similar reporter gene is
needed to monitor in vitra expression. and finally one needs an animal model of
disease.

In Mahan's system. a purA auxotroph of S. typhimurium was used. A delivery
vector was constructed that contained a promoterless copy of a wild type purA gene
fused to a promoterless lacZY gene, which encodes the enzyme p-galactosidase, using
a broad host range suicide vector. Next, random fragments of the S. fyphimurium
genome were inserted upstream of the purA gene. creating transcriptional fusions that
would allow any properly positioned S. fyphimurium promoters to drive expression of
the purA:lacZY fusion. The entire library of transcriptional fusions was introduced into
the S. Iyphimurium purA auxotroph, and recombinants which contained the
transcriptional fusion integrated by a single crossover event into the bacterial
chromosome at the site homologous to the genomic fragment upstream of the purA
gene were selected. Thus, the integration event generated strains where the native
chromosomal promoter would drive expression of the purA-lacZY genes, while the
cloned promoter would drive expression of the chromosomal gene at the site of
integration. This assures that there would be no disruption of chromosomal genes and

no loss of expression of any potential virulence factors. The pool of S. fyphimurium

17

strains containing integrated purA-lacZY fusions was injected into mice, and surviving
bacteria recovered 3 days later from the mouse spleens. The environmental conditions
in the host induced transcription from infection-induced promoters, thus expressing the
wild type biochemical gene and relieving the attenuation. Only strains which contained
a transcriptional fusion that expressed the purA gene in viva should have survived. All
recovered strains isolated from the infected mice were screened in vitra for expression
of the lacZY gene. Strains which survived in mice, but which minimally expressed ji-
galactosidase activity in vitra. contained promoters that were turned on in viva and off in
vitra. From the strains identified, the in viva expressed (ivr) genes were cloned and
knock-out mutants of three different 8. Iyphimurium ivi genes (carAB, pheST-himA, rfb)
have been shown to be attenuated inimice (99).

A similar system to identify in viva expressed genes has been developed for use
in the intracellular pathogen Legionella pneumaphila. using an avirulent fhyA mutant, an
independently replicating plasmid. and infection of a macrophage cell culture rather than
an intact host (72,125). Other IVET selection methods for in viva expressed promoters,
in addition to complementation of defined biochemical mutants, have been developed
for use in other bacteria. One form of selection involves the use of a promoteriess
antibiotic-resistance cassettes, such as chloramphenical acetyltransferase (100).
Cloned promoters are selected for expression during infection by administering the
antibiotic and isolating surviving bacteria. These antibiotic selections appear to function
best in tissue culture or small animals rather than in whole animal models because of
the difficult phannacokinetics. Finally, a modified IVET system has been developed for
use with Pseudomonas aeruginosa consisting of an avirulent purEK mutant. plasmid
pBR322, a promoterless purEK operon and a mouse model of infection (162). One
major limitation to IVET type selections is the necessity for a high enough level of
expression during the entire infection process to reisolate the organism containing an in

viva expressed promoter.

 

re:
he:

fOr

18

2. Signature-tagged mutagenesis (STM):

This system overcomes the problem of an auxotrophic or antibiotic based IVET
system because genes do not have to be expressed during the entire infection process
to be reisolated. In this system, transposons carrying unique DNA sequence tags are
used to create a collection of mutants by insertional mutagenesis (65). The pool of
mutants is used to infect a host animal. and surviving bacteria are recovered from the
infected host after several days. Using the unique DNA sequence tags and appropriate
PCR primers. probes can be created for colony blot hybridization. It is then possible to
identify which specific members of the original mutant pool are absent upon recovery of
bacteria from the host. These mutants should carry mutations in genes necessary for
survival in the host When this system was tested with Salmonella typhimurium, almost
half of the mutations identified were in genes known to be involved in the virulence of
this organism. The STM system has the power to identify genes that may be turned on
only transiently. However, the STM system is at this time not applicable to A.
pleurapneumaniae, since the transposon (T n5) used as the basis for the system does
not function in APP. Additional problems may also exist with the use of this system in
larger animals. If the entire population of surviving clones is not reisolated due to the
large size of the lungs or other organs, then mutants will be falsely identified as not
being able to survive. A certain subset of vimlence genes (e.g. toxins) will likely not be
identified because other clones in the pool will express them and the transposon mutant
will be able to replicate normally even though it is not able to express these genes.

3. Resalvase based IVET:

An extremely promising selection system uses in viva induction of fnpR
resolvase. a site-specific recombinase of the transposable element 16 to induce a
heritable, detectable genetic change. In this construct, expression of resolvase. even

for a very brief period, results in excision of a tetracycline resistance cassette flanked by

 

 

le.

9e

the

19

direct repeats of the DNA sequence at which the resolvase functions (13). Any bacterial
cell in which the resolvase is expressed will give rise to progeny that are tetracycline-
sensitive. This selection system has the power to identify promoters that are only
transiently expressed in viva because all strains survive whether the cloned promoter is
not turned on, turned on transiently, or turned on constitutively during infection. It has
not yet been demonstrated whether this system would be functional in APP. but if so it
would potentially be the most sensitive and least technical method to identify in viva
expressed genes. One limitation, as the system is currently designed. is the reliance on
screening for the loss of antibiotic resistance after in viva passage. A second limitation
is the sensitivity of the system; it can be too sensitive yielding a high level of background
with the false identification of genes being tumed on. Finally, all promoters turned on in
viva will be identified, not just those that are in viva specific; thus, pro-screening of the
pools is required to eliminate in vitra-expressed clones. As a result, this system will
likely not identify genes that are minimally to moderately expressed in vitro and induced
in viva.
4. Subtractive hybridization:

Subtractive hybridization is based on identifying differences in RNA transcripts
(157) between bacteria isolated from an infected animal and bacteria grown in the
laboratory. One limitation on this type of procedure is the technical difficulty in isolation
and manipulation of bacterial RNA to obtain a cDNA clone. Another limitation is that
subtractive hybridization results may change dramatically depending on the stage of
infection at which bacteria are isolated therefore an impractical number of animals may
be needed in order to truly identify all genes involved in pathogenesis. Also, this
technique will tend to identify ”all or none” expressed genes rather than in viva induced
genes which have some level of constitutive expression.

The choice of these systems for applications in different organisms will vary with

the genetic tools that are functional in the organism and perhaps with the type of animal

20

used. Each system has advantages and disadvantages. but as a whole they provide
tools that will revolutionize the study of pathogenesis. In previous IVET studies in
Salmonella typhimurium (100,143). approximately 5% of the bacteria recovered from
infected animals had cloned promoters that were turned on in viva and turned off in
vitro, indicating that they were expressed specifically during infection. Many of the
genes identified as having promoters specifically turned on in vivo or as being required
for survival in viva (13.65.99.162) have been genes involved in the biosynthesis of
purines. pyrimidines, and essential amino acids, such as part), pyrE, and carAB, while
others have been genes known to encode virulence factors. such as invA. invG, sva.
fadB. and virB or even transcriptional regulators such as Fur. These infection
associated genes can be divided into four different classes of in viva expressed genes:
1) enzymes in critical biochemical pathways; 2) known virulence factors. or genes
involved in the synthesis of such factors, such as genes for LPS
biosynthesis/modification; 3) new genes, not identifiable by comparison with known
sequence data; and 4) genes involved in transcriptional regulation or signal
transduction. In viva-expressed genes that can be identified as steps in critical
biochemical pathways, such as purine biosynthesis. may be candidates for production of
knock-out mutants and testing as live avirulent vaccines. Any of the genes identified by
homology with virulence genes of other pathogens are likely to be of extreme interest for
antibiotic and vaccine development. and further work with the unidentifiable genes will

open up new horizons and lead to a better understanding of pathogenesis.

 

10.

11.

10.

11.

21

REFERENCES

Beaudet. R.. G. McSween. G. Boulay. P. Rousseau. J.G. Blsaillon. J.P.
Descoteaux. and R. Ruppanner. 1994. Protection of mice and swine against
infection with Acfinobacillus pleurapneumaniae by vaccination. Vet. Microbiol.
39:71-81.

Beck, M.. J. F. van den Bosch, l. M. Jongenelen, P. L. Loeffen, R. Nielsen. J.
Nicolet, and J. Frey. 1994. RTX toxin genotypes and phenotypes in
Actinabacillus pleurapneumaniae field strains. J. Clin. Microbiol. 32:2749-2754.

Belanger, M.. C. Begin. and M. Jacques. 1995. Lipopolysaccharides of
Actinobacillus pleurapneumaniae bind pig hemoglobin. Infect Immun. 63:656-
662.

Belanger, M.. D. Dubreull. J. Harel. C. Girard. and M.Jacques. 1990. Role of
lipopolysaccharides in adherence of Actinobacillus pleurapneumaniae to porcine
tracheal rings. Infect Immun. 58:3523-3530.

Belanger, M.. D. Debreuil, and M. Jacques. 1994. Proteins found within
porcine respiratory tract secretions bind lipopolysaccharides of Actinobacillus
pleurapneumaniae. Infect. Immun. 62:868-873.

Belanger, M.. S. Rioux. B. Fairy. and M. Jacques. 1992. Affinity for porcine
respiratory tract mucus is found in some isolates of Actinabacillus
pleurapneumaniae. FEMS Microbiol. Lett. 97:119-126.

Bendlxen. P.H., P.E. Shewen, and S. Rosendal. 1981. Toxicity of
Haemaphilus pleurapneumaniae for porcine lung macrophages, peripheral blood
monocytes, and testicular cells. Infect. Immun. 33:673-676.

Bertram, TA 1985. Quantitative morphology of peracute pulmonary lesions in
swine induced by Haemophilus pleurapneumaniae. Vet Pathol. 22:598-609.

Beynon. L. M.. D. W. Griffith, J. C. Richards. and M. B. Perry. 1992.
Characterization of the lipopolysaccharide O-antigens of Actinabacillus
pleurapneumaniae serotypes 9 and 11: Antigenic relationships among serotypes
9. 11. and 1. J. Bacteriol. 174:5324-5331.

Bradley, S. G. 1979. Cellular and molecular mechanisms of action of bacterial
endotoxins. Ann. Rev. Microbiol. 33: 67-94.

Bunka, S., C. Christensen, A. A. Potter, P. J. WllIson, and G. F. Gerlach.
1995. Cloning and characterization of a protective outer membrane lipoprotein
of Actinobacillus pleurapneumaniae serotype 5. Infect Immun. 63:2797-2800.

 

 

11

2C

21.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22

Byrd. W., 8.6. Harman and S. Kadis. 1992. Protective efficacy of conjugate
vaccines against experimental challenge with porcine Actinabacillus
pleurapneumaniae. Vet. Immunol. lmmunopathol. 34:307-324.

Camilli, A. and J. J. Mekalanos. 1995. Use of recombinase gene fusions to
identify Vibrio cholerae genes induced during infection. Mol. Microbiol. 18:671-
683.

Chang. Y.F.. J. Shi. D.P. Ma, S.J. Shin. and D.H. Lein. 1993. Molecular
analysis of the Actinabacillus pleurapneumaniae RTX toxin-Ill gene cluster. DNA
Cell. Biol. 12:351-362.

Chang, Y.F.. R.Y. Young, and D.K. Struck. 1991. The Actinobacillus
pleurapneumaniae hemolysin determinant: unlinked appCA and appBD loci
flanked by pseudogenes. J. Bacteriol. 173:5151-5158.

Chang, Y.F.. R. Young, and D.K. Struck. 1989. Cloning and characterization
of a hemolysin gene from Actinobacillus (Haemophilus) pleurapneumaniae. DNA
8:635—647.

Chiang, Y.W., T.F. Young. V.J. Rapp-Gabrielson, and R.F.Ross. 1991.
Improved protection of swine from pleurapneumania by vaccination with
proteinase K-treated outer membrane of Actinobacillus (Haemophilus)
pleurapneumaniae. Vet. Microbiol. 27:49-62.

Chung, W.B.. L. Backstrom. J. McDonald. and M.T. Collins. 1993.
Actinobacillus pleurapneumaniae culture supematants interfere with killing of
Pasteurella multocida by swine pulmonary alveolar macrophages. Can. J. Vet.
Res. 57:190-197.

Cruijsen, T.. L.A. van Leengoed. M. Ham-Raffles, and J.H. Verheijden. 1995.
Convalescent pigs are protected completely against infection with a homologous
Actinobacillus pleurapneumaniae strain but incompletely against a heterologou-
serotype strain. Infect Immun. 63:2341-2343.

Cruijsen, T.. L.A.M.G. vanLeengoed, E.M. Kemp, A. Bartelse. A. Korevaar
and J.H.M. Verheijden. 1995. Susceptibility to Actinobacillus
pleurapneumaniae infection in pigs fiom an endemically infected herd is related
to the presence of toxin neutralizing antibodies. Vet Microbiol. 47:219-228.

Cruijsen, T.. L.A.M.G. van Leengoed. E.M. Kemp. W.A. Hunneman. K.
Rlepema, A. Bartelse and J.H.M. Verheijden. 1995. Prevalence and
development of antibodies neutralizing the haemolysin and cytotoxin of
Actinobacillus pleurapneumaniae in three infected pig herds. Vet. Quarterly.
17:96-100.

 

 

 

31

32.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

23

Cruz. W.T.. Y. A. Nedlalkov, B.J. Thacker, and M.I-l. Mulks. 1996. Molecular
characterization of a common 48 kDa outer membrane protein of Actinobacillus
pleurapneumaniae. Infect. Immun. 64: 83-90.

Cullen. J. M.. and A. N. Rycroft. 1994. Phagocytosis by pig alveolar
macrophages of Actinobacillus pleurapneumaniae serotype 2 mutant strains
defective in haemolysin ll (Apxll) and pleurotoxin (Apxlll). Microbiology 140:237-
244.

Cybulsky, M. I.. M. K. W. Chen, and H. 2. Movat. 1988. Acute inflammation
and microthrombosis induced by endotoxin, interleukin-1. and tumor necrosis
factor and their implication in gram negative infections. Lab. Invest. 58:365-378.

Daban, M.. A. Medrano. and E. Querol. 1996. Cloning, sequencing and
expression of the transferrin-binding protein 1 gene from Actinobacillus
pleurapneumaniae. Biochem. J. 315:257-264.

Deneer. H.G.. and A.A. Potter. 1989. Effect of iron restriction on the outer
membrane proteins of Actinobacillus (Haemaphilus) pleurapneumaniae. Infect.
Immun. 57:798-804.

Deneer. H.G.. and A.A. Potter. 1989. Identification of a maltose-inducible major
outer membrane protein in Actinobacillus (Haemaphilus) pleurapneumaniae.
Microb. Pathog. 6:425-432.

Devenish. J.. and S. Rosendal. 1991. Calcium binds to and is required for
biological activity of the 104-kilodalton hemolysin produced by Actinobacillus
pleurapneumaniae serotype 1. Can. J. Microbiol. 37:317-321.

Devenish. J.. S. Rosendal. and J.T. Basso. 1990. Humoral antibody response
and protective immunity in swine following immunization with the 104-kilodalton
hemolysin of Ach’nobacillus pleurapneumaniae. Infect. Immun. 58:3829-3832.

Didier. P. J.. L. Perlno, and J. Urbance. 1984. Porcine Haemaphilus
pleurapneumania: Microbiologic and pathologic findings. J. Am. Vet. Med.
Assoc. 184:716-719.

Dom, P.. F. l-Iaesebrouck, and P. DeBaetseIier. 1992. Stimulation and
suppression of oxygenation activity of porcine pulmonary alveolar macrophages
by Actinobacillus pleurapneumaniae and its metabolites. Am. J. Vet. Res.
53:1113-1118.

Dom, P.. F. Haesebrouck, R. Ducatelle, and G. Charller. 1994. In viva
association of Actinobacillus pleurapneumaniae serotype 2 with the respiratory
epithelium of pigs. Infect. Immun. 62:1262-1267.

 

 

a)

 

(A

l.‘

(.9

(A,

42

43

33.

35.

37.

38.

39.

40.

41.

42.

43.

24

Dom. P.. F. Haesebrouck, E.M. Kemp. and M.A. Smite. 1992. Influence of
Actinobacillus pleurapneumaniae serotype 2 and its cytolysins on porcine
neutrophil chemiluminescence. Infect. Immun. 60:4328-4334.

Drum. 8., WE. Marsh, M.A. Mellencamp and RD. Walker. 1996. Effects of
segregated early weaning on growth performance of pigs from nursery to
slaughter. Abstracts, 77th Conference of Research Workers in Animal Diseases.
abstract #64.

Fedorka-Cray, P.J., M.J. Huether, D.L. Stine, and G.A. Anderson. 1990.
Efficacy of a cell extract from Actinobacillus (Haemophilus) pleurapneumaniae
serotype 1 against disease in swine. Infect. Immun. 58:358-365.

Fedorka-Cray. P.J., D.L.Stine. J.M. Greenwald, J.T. Gray. M.J. Heuther, and
G. A. Anderson. 1993. The importance of secreted vlmlence factors in
Actinobacillus pleurapneumaniae bacterin preparation: a comparison. Vet
Microbiol. 37285-100.

Fenwick. B. W., J. S. Cullor. B. I. Osbum, and H. J. Olander. 1986.
Mechanisms involved in protection provided by immunization against core
lipopolysaccharides of Escherichia cali J5 from lethal Haemaphilus
pleurapneumaniae infections in swine. Infect Immun. 53:298-304.

Fenwick. B.. and S. Henry. 1994. Porcine pleurapneumania. J. Am. Vet Med.
Assoc. 204:1334-1340.

Fenwick. B. W., and B. l. Osbum. 1986. Immune responses to the
lipopolysaccharides and capsular polysaccharides of Haemaphilus
pleurapneumaniae in convalescent and immunized pigs. Infect Immun. 54:575-
582.

Fenwick, B. W., B. l. Osbum, J. S. Cullor. S. C. Henry, and H. J. Olander.
1986. Mortality in swine herds endemically infected with Haemaphilus
pleurapneumaniae: Effect of immunization with cross-reacting lipopolysaccharide
core antigens of Escherichia cali. Am. J. Vet. Res. 47:1888-1891.

Fenwick. B. W., B. l. Osbum, and H. J. Olander. 1986. Isolation and biological
characterization of two lipopolysaccharides and a capsular-enriched
polysaccharide preparation from Haemaphilus pleurapneumaniae. Am. J. Vet.
Res. 47:1433-1441.

Frey. J. 1992. Construction of a broad host range shuttle vector for gene cloning
and expression in Actinobacillus pleurapneumaniae and other Pasteurellaceae.
Res. Microbial. 143: 263-269.

Frey, J.. M. Beck. U. Stucki, and J. NlcoIet. 1993. Analysis of hemolysin
operons in Actinobacillus pleurapneumaniae. Gene 123:51-58.

 

(h

55

45.

47.

48.

49.

50.

51.

52.

53.

55.

25

Frey, J.. J.T. Basso. Y.F. Chang, J.M. Cullen, B. Fenwick. G.F. Gerlach. D.
Gygl. F. Haesebrouck, T.J. Inzana. R. Jansen, et al. 1993. Actinobacillus
pleurapneumaniae RTX-toxins: uniform designation of haemolysins, cytolysins,
pleurtoxin, and their genes. J. Gen. Microbiol. 139:1723-1728.

Frey. J.. J.B. Deillon. D. Gygl. and J. Nicolet. 1991. Identification and partial
characterization of the hemolysin (Hlyll) of Actinobacillus pleurapneumaniae
serotype 2. Vet Microbiol. 28:303-312.

Frey. J.. A. Haldlmann. J. Nicolet, A. Boffinl. and P. Prentki. 1994. Sequence
analysis and transcription of the apxl operon (hemolysin I) from Actinobacillus
pleurapneumaniae. Gene 142:97-102.

Frey, J.. R. Kuhn, and J. Nicolet. 1994. Association of the CAMP phenomenon
in Actinobacillus pleurapneumaniae with the RTX toxins Apxl. Apxll and Apxlll.
FEMS Microbiol. Lett 124:245-251.

Frey, J.. P. Kuhnert, L Vllllger and J. Nicolet. 1996. Cloning and
characterization of an Actinobacillus pleurapneumaniae outer membrane protein
beloning to the family of PAL lipoproteins. Res. Microbial. 147:351-361.

Frey. J.. R. Meier, D. Gygl. and J. Nlcolet. 1991. Nucleotide sequence of the
hemolysin I gene from Actinobacillus pleurapneumaniae. Infect Immun.
59:3026-3032.

Frey, J.. and J. Nicolet. 1990. Hemolysin patterns of Actinobacillus
pleurapneumaniae. J. Clin. Microbiol. 28:232-236.

Frey, J.. and J. Nlcolet. 1988. Purification and partial characterization of a
hemolysin produced by Actinobacillus pleurapneumaniae type strain 4074.
FEMS Microbiol. Lett 55:41-46.

Frey. J.. and J. Nlcolet. 1988. Regulation of hemolysin expression in
Actinobacillus pleurapneumaniae serotype 1 by Ca2+. Infect. Immun. 56: 2570-
2575.

Frey, J.. J. Perrln, and J. Nlcolet. 1989. Cloning and expression of a
cohemolysin. the CAMP factor of Actinobacillus pleurapneumaniae. Infect
Immun. 57:2050-2056.

Frey, J.. H. van den Bosch. R. Segers, and J. Nicolet. 1992. Identification of
a second hemolysin (Hlyll) in Actinobacillus pleurapneumaniae serotype 1 and
expression of the gene in Escherichia cali. Infect Immun. 60:1671-1676.

Fuller. T.E.. B.J. Thacker and M.H. Mulks. 1996. A riboflavin auxotroph of
Actinobacillus pleurapneumaniae is attenuated in swine. Infect Immun.
64:4659-4664

 

 

 

l“

 

64

66

S7

56.

57.

58.

59.

60.

61.

62.

63.

65.

67.

26

Gerlach. G.F., C. Anderson, S. Klashinsky. A. Rossi-Campos. A.A. Potter,
and P.J. Willson. 1993. Molecular characterization of a protective outer
membrane lipoprotein (OmlA) from Actinobacillus pleurapneumaniae serotype 1.
Infect. Immun. 61:565-572.

Gerlach. G.F.. C. Anderson, A.A. Potter. S. Klashinsky. and P.J. Willson.
1992. Cloning and expression of a transferrin-binding protein from Actinobacillus
pleurapneumaniae. Infect. Immun. 60:892-898.

Gerlach. G.-F.. C. Anderson. A. RossI-Campos, and A.A.Potter. 1992. The
role of the 103 Kd Actinobacillus pleurapneumaniae cytolysin in virulence and the
association of the encoding gene with insertion sequence-like elements. Proc.
Int. Pig Vet. Soc. 12:199.

Gerlach. G.F., S. Klashinsky. C. Anderson. A.A. Potter, and P.J. Willson.
1992. Characterization of two genes encoding distinct transferrin-binding
proteins in different Actinobacillus pleurapneumaniae isolates. Infect Immun.
60:3253-3261.

Green, J.. A.D. Sharrocks, J.l. Maclnnes, and J.R. Guest. 1992. Purification
of HlyX. a potential regulator of haemolysin synthesis, and properties of
HlyX:FNR hybrids. Proc. R. Soc. Land. 8. Biol. Sci. 248:79-84.

Griffiths, E. 1991. Environmental Regulation of Bacterial Vimlence-
Implications for Vaccine Design and Production. Trends in Biotechnology
92309-315.

Gross. R. 1993. Signal Transduction and Virulence Regulation in Human and
Animal Pathogens. FEMS Micro. Rev. 104:301-326

Gygl. D.. J. Nicolet. C. Hughes. and J. Frey. 1992. Functional analysis of the
Ca(2+)-regulated hemolysin l operon of Actinobacillus pleurapneumaniae
serotype 1. Infect. Immun. 60:3059-3064.

Hartmann, L., W. Schroder, and A. Lubke. 1995. Isolation of the major outer-
membrane protein of Actinobacillus pleurapneumaniae and Haemaphilus
parasuis. J. Vet. Med. 8. 42:59-63.

Hansel, M. J. E. Shea, C. Gleason, M. D. Jones, E. Dalton, and D. W. Holden.
1995. Simultaneous identification of bacterial virulence genes by negative
selection. Science 269:400-403.

Hodgson. A.L.M., C. L. Wright, C.A. Podgson. R.A. Strugnell and C.T.
Prideaux. 1996. Genetic manipulation of Haemophilus. Actinobacillus and
Pasteurella (HAP). Proc. HAP-96, abstract #21.

Hunneman, W.A. 1986. Incidence. economic effects. and control of
Haemaphilus pleurapneumaniae infections in pigs. Vet. Q. 8:83-87.

 

 

 

 

7‘1

7e

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

27

Inzana, T. J. 1991. Vinilence properties of Actinobacillus pleurapneumaniae.
Microb. Path. 11: 305-316.

Inzana, T. J.. J. Ma. T. Workman, R. P. Gogolewski. and P. Anderson. 1988.
Virulence properties and protective efficacy of the capsular polymer of
Haemaphilus (Actinobacillus) pleurapneumaniae serotype 5. Infect. Immun. 56:
1880-1889.

Inzana, T. J.. J. Todd. J. Ma. and H. Veit. 1991. Characterization of a non-
hemolytic mutant of Actinobacillus pleurapneumaniae serotype 5: role of the 110
kilodalton hemolysin in virulence and immunoprotection. Microb. Path. 10:281-
296.

lnzana, T. J.. J. Todd, and H. P. Velt. 1993. Safety. stability, and efficacy of
nonencapsulated mutants of Actinobacillus pleurapneumaniae for use in live
vaccines. Infect Immun. 61:1682-1686.

lsberg. RR. 1994. Intracellular trafficking of Legionella pneumaphila within
phagocytic cells. pp.263-278. In V.L. Miller et al (ed). Molecular Genetics of
Bacterial Pathogenesis. ASM Press. Washington, DC.

lshli, H.. T. Fukuyasu, S. lyobe. and H. Hashimoto. 1993. Characterization of
newly isolated plasmids from Actinobacillus pleurapneumaniae. Am. J. Vet. Res.
54:701-708.

lshli, H.. F. Hayashl. S. lyobe. and H. Hashimoto. 1991. Characterization and
classification of Actinobacillus (Haemophilus) pleurapneumaniae plasmids. Am.
J. Vet Res. 52:1816-1820.

Ito, H.. l. Uchlda, T. Seklzakl, E. Ooishl, T. Kawai, T. Okabe. A. Taneno. and
N. Terakado. 1995. Molecular cloning of an Actinobacillus pleurapneumaniae
outer membrane lipoprotein (OmlA) from serotype 5a. Microb. Pathog. 18:29-
36.

lto. H.. l. Uchlda, T. Seklzakl, and N. Terakado. 1993. A cryptic DNA
sequence, isolated from Actinobacillus pleurapneumaniae. confers a hemolytic
activity upon Escherichia coli K12 strains. J. Vet Med. Sci. 55:173-175.

Jacques. M.. M. Belanger, G. Roy, and B. Fairy. 1991. Adherence of
Actinobacillus pleurapneumaniae to porcine tracheal epithelial cells and frozen
lung sections. Vet. Microbiol. 27:133-143.

Jansen, R.. J. Briaire, E. M. Kamp, A. L. Gielkins, and M. A. Smite. 1995. The
CAMP effect of Actinobacillus pleurapneumaniae is caused by Apx toxins.
FEMS Microbiol. Lett. 126:139-143.

 

 

88

89.

90.

79.

80.

81.

82.

83.

85.

87.

88.

89.

90.

28

Jansen. R.. J. Briaire, E.M. Kamp. A.L. Gielkens. and M.A. Smite. 1993.
Cloning and characterization of the Actinobacillus pleurapneumaniae-RTX-toxin
lll (Apxlll) gene. Infect. Immun. 61:947-54.

Jansen, R.. J. Briaire, E. M. Kamp. A.L. Gielkens. and M.A. Smite. 1993.
Structural analysis of the Actinobacillus plaurapneumaniae-RTX-toxin l (Apxl)
operon. Infect Immun. 61:3688-3695.

Jansen. R.. J. Briaire, E.M. Kamp. and M.A. Smite. 1992. Comparison of the
cytolysin ll genetic determinants of Actinobacillus pleurapneumaniae serotypes.
Infect. Immun. 60:630-636.

Jansen, R.. J. Briaire, H.E. Smith. P. Dom. F. Haesebrouck, E.M. Kamp.
A.L.J. Gielkens. and M.A. Smite. 1995. Knockout mutants of Actinobacillus
pleurapneumaniae serotype 1 that are devoid of RTX toxins do not activate or kill
porcine neutrophils. Infect Immun. 63:27-37.

Janeen, R.. J. Briaire, A. B. van Geel, E. M. Kamp. A. L. Gielkins. and M. A.
Smite. 1994. Genetic map of the Actinobacillus pleurapneumaniae RTX-toxin
(Apx) operons: characterization of the Apxlll operons. Infect. Immun. 62:4411-
4418.

Jolie. R.A.V.. M. H. Mulks, and B. J. Thacker. 1994. Antigenic differences
within Actinobacillus pleurapneumaniae serotype 1. Vet. Microbiol. 38:329-349.

Kamp. E.M., J. K. Popma, J. Anakotta, and M. A. Smite. 1991. Identification of
hemolytic and cytotoxic proteins of Actinobacillus pleurapneumaniae by use of
monoclonal antibodies. Infect. Immun. 59:3079-3085.

Kamp. E.M., and L. A. van Leengoed. 1989. Serotype-related differences in
production and type of heat-labile hemolysin and heat-labile cytotoxin of
Actinobacillus (Haemaphilus) pleurapneumaniae. J. Clin. Microbiol. 27:1187-
1191.

Kamp. E.M., T.M.M. Venneulen, T. Cruyeen. J.E. van Dijk. K. Riepema, M.A.
Smite. and J.M. van Leeuwen. 1992. Actinobacillus pleurapneumaniae
cytotoxins: Cross protective effect of a crude cytolysin vaccine. Proc. Int. Pig
Vet. Soc. 12:209.

Kamp. E.M., T.M.M.Venneulen. M.A.Smlte, and J. Haagsma. 1994.
Production of Apx toxins by field strains of Actinobacillus pleurapneumaniae and
Actinobacillus suis. Infect. Immun. 62: 4063-4065.

Koblsch, M.. and J. F. van den Bosch. 1992. Efficacy of an Actinobacillus
pleurapneumaniae subunit vaccine. Proc. Int. Pig. Vet. Soc. 12:216.

Lalonde, 6., TN. McDonald, P. Gardner, and PD. O'Hanley. 1989.
Identification of a hemolysin from Actinobacillus pleurapneumaniae and

 

 

 

 

(O

(n

11

1C

91.

92.

93.

95.

97.

98.

99.

100.

101.

29

characterization of its channel properties in planar phospholipid bilayers. J. Biol.
Chem. 264:13559-13564.

Lalonde. G.. J. F. Miller, L. S. Tompkins, and P. D. O'Hanley. 1989.
Transformation of Actinobacillus pleurapneumaniae and analysis of R factors by
electroporation. Am. J. Vet. Res. 50:1957-1960.

Lalonde. G.. and P. D. O'Hanley. 1989. Development of a shuttle vector and a
conjugative transfer system for Actinobacillus pleurapneumaniae. Gene
85:243-246.

Lian, C.J.. S. Rosendal. and J. l. Maclnnes. 1989. Molecular cloning and
characterization of a hemolysin gene from Actinobacillus (Haemaphilus)
pleurapneumaniae. Infect Immun. 57:3377-3382.

nggett. A. D., L. R. Harrison. and R. L. Farrell. 1987. Sequential study of
lesion development in experimental haemophilus pleurapneumania. Res. Vet
Sci. 42:204-221.

MacDonald, J.. and A. N. Rycroft. 1993. Actinobacillus pleurapneumaniae
haemolysin II is secreted from Escherichia coli by A. pleurapneumaniae
pleurotoxin secretion gene products. FEMS Microbiol. Lett 109:317-322.

MacDonald, J.. and AN. Rycroft. 1992. Molecular cloning and expression of
ptxA, the gene encoding the 120-kilodalton cytotoxin of Actinobacillus
pleurapneumaniae serotype 2. Infect Immun. 60:2726-2732.

Maclnnes. J.I.. J. E. Kim. C.-J. Lian. and G.A. Soltes. 1990. Actinobacillus
pleurapneumaniae hlyX gene homology with the fnr gene of Escherichia coli. J.
Bacteriol. 172:4587-4592.

Maclnnes. J.I.. and S. Rosendal. 1987. Analysis of major antigens of
Haemophilus (Actinobacillus) pleurapneumaniae and related organisms. Infect
Immun. 55:1626-1634.

Mahan, M.J.. J.M. Slauch. and J.J.Mekalanos. 1993. Selection of bacterial
virulence genes that are specifically induced in host tissues. Science 259:686-
688.

Mahan, M.J.. J.W. Tobias. J.M. Slauch, P. C. Hanna. R. J. Collier. and J. J.
Mekalanos. 1995. Antibiotic-based selection for bacterial genes that are
specifically induced during infection of a host. PNAS 92:669-673.

Meier, E.. N. Reinhard, R. Benz and J. Frey. 1996. Channel-forming activity
and channel size of the RTX toxins Apxl. Apxll. and Apxlll of Actinobacillus
pleurapneumaniae. Infect Immun. 64:4415-4423.

 

 

 

 

 

 

1(

1C

11

11

11:

115

114

115

102.
103.

104.
105.

106.

107.

108.
109.
110.
111.
112.
113.

114.

115.

30

Matthews. P. R. J.. and I. H Pattlson. 1961. The identification of a
Haemaphilus-like organism associated with pneumonia and pleurisy in the pig.
J. Comp. Pathol. 71:44-52.

McWhinney. D.R., Y-F. Chang. R. Young, and D.K.Struck. 1992. Separable
domains define target cell specificities of an RTX hemolysin from Actinobacillus
pleurapneumaniae. J. Bacteriol. 174:291-297.

Mekalanos. J.J. 1992. Environmental signals controlling expression of
virulence determinants in bacteria. J. Bacteriol. 174:1-7.

Morrison, D. C.. and J. L. Ryan. 1987. Endotoxins and disease mechanisms.
Ann. Rev. Med. 38:417-432.

Mulks. M.H. Unpublished data.

Mulks. M.H.. and J. M. Buysse. 1995. A targeted mutagenesis system for
Actinobacillus pleurapneumaniae. Gene 165:61-66.

Mulks. M.H., D.Y. Hunter-Simon. J.W.Sprague. 8. B.J.Thacker. 1986.
Identification of immunogenic outer membrane proteins of Haemaphilus
pleurapneumaniae in infected swine. Proc. Int Pig Vet. Soc. 9:266.

Mulks. M.H.. and B.J.Thacker. 1988. Efficacy of an Haemaphilus
pleurapneumaniae outer membrane subunit vaccine in swine. Proc. Int. Pig Vet
Soc. 10:81.

Nakai, T.. E. Ono, K. Ike. and K. Kume. 1992. Serotyping of Actinobacillus
pleurapneumaniae strains by use of purified capsular polysaccharide or
lipopolysaccharide. Proc. Int. Pig Vet. Soc. 12:186.

Narita. M.. K. Kawashlma. T. Morozuml and H. Takashima. 1995. Effect of
physical defenses of the respiratory tract on the development of pneumonia in
pigs inoculated endobronchially with Actinobacillus pleurapneumaniae. J. Vet
Med. Sci. 57:839-844.

Nicolet. J. 1992. Actinobacillus pleurapneumaniae. pp. 401-408. In A. D.
Leman et al, eds. Diseases of swine. Iowa State University Press, Ames, Iowa.

Nielsen, R. 1984. Haemaphilus pleurapneumaniae serotypes - Cross protection
experiments. Nord. Vet. Med. 36:221-234.

Nielsen, R. 1979. Haemaphilus parahaemolyticus serotypes: pathogenicity and
cross immunity. Nord. Vet. Med. 31:407-413.

Nielsen. R. 1976. Pleuropneumonia of swine caused by Haemaphilus
parahaemolyticus. Studies on the protection obtained by vaccination. Nord. Vet.
Med. 28:337-338.

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

31

Nielsen. R. 1974. Serological and immunological studies of pleurapneumania of
swine caused by Haemaphilus parahaemolyticus. Acta Vet. Scand. 15:80-89.

Niven, D.F.. J. Donga. and F.S. Archibald. 1989. Responses of Haemaphilus
pleurapneumaniae to iron restriction: changes in the outer membrane protein
profile and the removal of iron from porcine transferrin. Mol. Microbiol. 3:1083-
1089.

Oishi. E.. T. Kitajima, Y. Koyama. T. Ohgitanl. S. Katayama. and T. Okabe.
1995. Protective effect of the combined vaccine prepared from cell-free-antigen
of Actinobacillus pleurapneumaniae serotypes 1, 2 and 5 in pigs. J. Vet. Med.
Sci. 57:1125-1128.

Olander. H.J. 1963. A septicaemic disease of swine and its causative agent
Haemaphilus parahaemalyticus. Ph.D. thesis, U. California.

O'Rellly, T.. D.F. Niven. and MR. Brown. 1991. Phenotypic variation in the
outer membrane protein composition of Actinobacillus (Haemophilus)
pleurapneumaniae: non-specific effect of exogenous pyridine nucleotide supply.
Vet. Microbiol. 29:159-172.

Paltlneanu. D., R. Pambucol. E. Tirzlu. and I. Scobercea. 1992. Swine
infectious pleurapneumania: Aerosol vaccination with a live attenuated vaccine.
Proc. Int Pig. Vet Soc. 12:214.

Paradis, S.E., D. Debreuil, S. Rioux. M. Gottschalk. and M. Jacques. 1994.
High-molecular-mass lipopolysaccharides are involved in Actinobacillus
pleurapneumaniae adherence to porcine respiratory tract cells. infect Immun.
62:331 1-3319.

Pattison. l. H.. D. G. Howell. and J. Elliot. 1957. A Haemaphilus-like organism
isolated from pig lung and the associated pneumonic lesions. J. Comp. Path.
67:320-329.

Perry, M.B., E. Altman, J.-R. Brisson. L M. Beynon. and J. C. Richards.
1990. Structural characteristics of the antigenic capsular polysaccharides and
lipopolysaccharides involved in the serological classification of Actinobacillus
pleurapneumaniae strains. Serodiag. Immunother. Infect. Dis. 42299-308.

Rankin. S., and R. R. lsberg. 1993. Identification of Legionella pneumaphila
loci expressed in response to an intracellular environment. Infect Agents Dis.
2:269-271.

Rapp. V.J.. R.S.Munson, 8. R.F.Ross. 1986. Outer membrane profiles of
Haemaphilus pleurapneumaniae. Infect. Immun. 52:414-420.

127.

128.

129.

130.

131.

132.

133.

134.

135.

136.

137.

32

Rapp, V.J.. and RF. Ross. 1988. lmmunogenicity of outer membrane
components of Haemophilus (Actinobacillus) pleurapneumaniae. Can. Vet. J.
29:585-587.

Ricard, M.A., F.S. Archibald. and D.F. Niven. 1991. Isolation and identification
of a putative porcine transferrin receptor from Actinobacillus pleurapneumaniae
biotype 1. J. Gen. Microbiol. 137:2733-2740.

Rosendal. S., 0.8. Carpenter. W.R. Mitchell, and MR. Wilson. 1981 .
Vaccination against pleurapneumania in pigs caused by Haemaphilus
pleurapneumaniae. Can. Vet. J. 22:34-35.

Rosendal. S.. J. Devenish. J.l. Maclnnes. J.H. Lumsden, S. Watson. and H.
Xun. 1988. Evaluation of heat-sensitive, neutrophil-toxic, and hemolytic activity
of Haemophilus (Actinobacillus) pleurapneumaniae. Am. J. Vet Res. 49: 1053-
1058.

Rosendal, S.. and J. l. Maclnnes. 1990. Characterization of an attenuated
strain of Actinobacillus pleurapneumaniae, serotype 1. Am. J. Vet Res. 51:711-
717.

Rosendal. S.. O. P. Miniate. and P. Sinclair. 1986. Protective efficacy of
capsular extracts of Haemophilus pleurapneumaniae in pigs and mice. Vet
Microbiol. 12:229-240.

Rossi-Campos. A.. C. Anderson. G.F. Gerlach. S. Klashinsky. A.A. Potter,
and P.J. Willson. 1992. Immunization of pigs against Actinobacillus
pleurapneumaniae wiflr two recombinant protein preparations. Vaccine.
10:512-518.

Rycroft. A. N.. and J. M. Cullen. 1990. Complement resistance in
Actinobacillus (Haemaphilus) pleurapneumaniae infection of swine. Am. J. Vet
Res. 51:1449-1453.

Rycroft. A.N.. and D.J. Taylor. 1987. Preparation and characterization of
envelope proteins from Haemaphilus pleurapneumaniae. Vet Microbiol. 1 5:303-
314.

Rycroft. A.N.. D. Williams. J.M. Cullen, and J. Macdonald. 1991. The cytotoxin
of Actinobacillus pleurapneumaniae (pleurotoxin) is distinct from the haemolysin
and is associated with a 120 kDa polypeptide. J. Gen. Microbiol. 137:561-568.

Rycroft. A.N.. D. Williams. I.A. McCandlish, and D.J. Taylor. 1991.
Experimental reproduction of acute lesions of porcine pleurapneumania with a
haemolysin-deficient mutant of Actinobacillus pleurapneumaniae. Vet Rec.
129:441-443.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

33

Sakano, T.. l. Shibata. Y. Samegai, A. Taneda, M. Okada. T. lrlsawa, and S.
Sato. 1993. Experimental pneumonia of pigs infected with Aujesky's disease
virus and Actinobacillus pleurapneumaniae. J. Vet. Med. Sci. 55:575-579.

Sanford. S. E.. and G. K. A. Josephson. 1981. Porcine Haemaphilus
pleurapneumaniae epizootic in Southwestem Ontario: Clinical, microbiological,
pathological, and some epidemiologic findings. Can. J. Comp. Med. 45:2-7.

Scholz. P.. V. Haring, B. WIttrnann-Liebold, K. Aehman, M. Bagdasarian, and
E. Scherzinger. 1989. Complete nucleotide sequence and gene organization of
the broad host range plasmid RSF1010. Gene 75:271-288.

Sebunya, T.N.K. and Saunders, JR 1983. Haemaphilus pleurapneumaniae
Infection in Swine: A Review. J. Am. Vet Med. Assoc. 182:1331-1337.

Shape. RE. 1964. Porcine contagious pleurapneumania. l. Experimental
transmission. etiology, and pathology. J. Exp. Med. 119:357-368.

Slauch, J. M. M. J. Mahan, and J. J. Mekalanos.1994.ln viva expression
technology for selection of bacterial genes specifically induced in host tissues.
Moth. Enzymol. 235: 481-492.

Smite. M.A.. J. Briaire, R. Jansen. H.E. Smith. E. M. Kemp. and A.L.
Gielkens. 1991. Cytolysins of Actinobacillus pleurapneumaniae serotype 9.
Infect. Immun. 59:4497-4504.

Soltes, G.. and J.I. Maclnnes. 1994. Regulation of gene expression by the
Hin protein of Actinobacillus pleurapneumaniae. Microbiology 140:839-845.

Stine. D. L. P. J. Fedorka-Cray. M. J. Huether, M. J. Gentry, and G. A.
Anderson. 1994. Comparison of semm responses in swine after vaccination
and challenge exposure with Actinobacillus pleurapneumaniae. Am. J. Vet Res.
55:1238-1243.

Tarigan, S.. R. F. Slocombe. G. F. Browning. and W. Kimpton. 1994.
Functional and structural changes of porcine alveolar macrophages induced by
sublytic doses of a heat-labile. hemolytic, cytotoxic substance produced by
Actinobacillus pleurapneumaniae. Am. J. Vet Res. 55: 1 548-1 557.

Tascon, R. l.. E. F. Rodriguez-Fern. C. B. Gutierrez-Martin, I. Rodriguez-
Barbosa, P. Berche. and J. A. Vazquez-Boland. 1993. Transposon
mutagenesis in Actinobacillus pleurapneumaniae with a Tn 10 derivative. J.
Bacteriol. 175:5717-5722.

Tascon, R.I.. J.A. Vazquez-Boland. C.B. Gutierrez-Martin. l. Rodriguez-
Barboea, and E.F. Rodriguez-Ferd. 1994. The RTX haemolysins Apxl and
Apxll are major virulence factors of the swine pathogen Actinobacillus

150.

151.

152.

153.

154.

155.

156.

157.

158.

1 59.

160.

161.

34

pleurapneumaniae: evidence from mutational analysis. Mol. Microbiol. 14:207-
216.

Thacker, B.J.. M.H. Mulks, and R.A.V Jolie. 1992. Development of improved
Actinobacillus pleurapneumaniae vaccines. Proc. Int. Pig Vet Soc. 12:193.

Thacker, B.J.. M.H. Mulks. F. Udeze, and RR. Simonson. 1994.
Development of disrupted whole-cell Actinobacillus pleurapneumonaie (APP)
vaccines. HAP-94 Proc., abstract if P35.

Tu, A.H.T., C. Hausler, R. Young. and D.K. Struck. 1994. Differential
expression of the cytotoxic and hemolytic activities of the ApxllA toxin from
Actinobacillus pleurapneumaniae. Infect. Immun. 62:2119-2121.

Udeze, F.A.. and S. Kadis. 1992. Effects of Actinobacillus pleurapneumaniae
hemolysin on porcine neutrophil function. Infect. Immun. 60:1558-1567.

Udeze, F.A.. K. S. Latimer, and S. Kadis. 1987. Role of Haemaphilus
pleurapneumaniae lipopolysaccharide endotoxin in the pathogenesis of porcine
Haemaphilus pleurapneumania. Am. J. Vet Res. 48:768-773.

Utrera, V.. and C. Pijoan. 1991. Fimbriae in A. pleurapneumaniae strains
isolated from pig respiratory tracts. Vet. Rec. 128:357-358.

Utrera. V.. C. Pijoan. and T. Molitor. 1992. Evaluation of the immunity induced
in pigs after infection with a low vimlence strain of A. pleurapneumaniae
serotype 1. Proc. Int. Pig. Vet Soc. 12:213.

Utt, E.A.. J.P. Irousal. L.C. KIkuta-Oshima and FD. Quinn. 1995. The
identification of bacterial gene expression differences using mRNA-based
isothermal subhactive hybridisation. Can. J. Microbiol. 41:152-156.

van den Bosch. J.F., I.M.C.A. Jongenelen, A.N.B. Pubben. F.G.A. van Vugt,
and R.P.A.M. Segers. 1992. Protection induced by a trivalent A.
pleurapneumaniae subunit vaccine. Proc. Int. Pig. Vet. Soc. 12:194.

van den Bosch. J.F., A.N.B. Pubben. F.G.A. van Vugt, and R.P.A.M. Segers.
1994. Synergistic effect of a P2-homologous outer membrane protein and Apxl
toxin in protection against Actinobacillus pleurapneumaniae. HA P-94
Proceedings. abstract # T12.

van Leengoed. LA, and H.W. Dickerson. 1992. Influence of calcium on
secretion and activity of the cytolysins of Actinobacillus pleurapneumaniae.
Infect. Immun. 60:353-359.

van Leengoed. L.A., E.M. Kamp. and J.M. Pol. 1989. Toxicity of Haemaphilus
pleurapneumaniae to porcine lung macrophages. Vet Microbiol. 19:337-349.

162.

163.

164.

165.

166.

167.

168.

169.

35

Wang, J.. A. Musheglan. S. Lory and S. Jin. 1996. Large-scale isolation of
candidate virulence genes of Pseudomonas aeruginosa by in viva selection.
PNAS 93:10434-10439.

Ward, C.K.. and T.J. Inzana. 1996. Mutations in capsular polysaccharide
biosynthesis render Actinobacillus pleurapneumaniae avirulent. HAP-96 Proc.,
abstract #25.

Ward. C.K.. and T.J.lnzana. 1994. Cloning of a DNA region putatively involved
in the encapsulation of Actinobacillus pleurapneumaniae. HAP-94 Proc.,
abstract if T25.

Ward. C.K.. and T.J. Inzana. 1994. Resistance of Actinobacillus
pleurapneumaniae to bactericidal antibody and complement is mediated by
capsular polysaccharide and blocking antibody specific for lipopolysaccharide.
J. Immunol. 153:2110-2121.

West. S.E.H., and M.J.M. Romero. 1995. Construction of an Actinobacillus
pleurapneumaniae-Escherichia cali cosmid cloning vector. Gene 160287-88.

West, S.E.H., M.J.M. Romero. L.B. Regassa. N.A. Zielinski, and RA. Welch.
1995. Construction of Actinobacillus pleurapneumaniae-Escherichia cali shuttle
vectors: Expression of antibiotic resistance genes. Gene 160:81-86.

Westfall. J. A., A. Chattergee, and N. V. Anderson. 1994. Actinobacillus
pleurapneumaniae in gnotobiotic piglets: ultrastructural changes in the
pulmonary alveoli with dose and time. J. Submicrosc. Cytol. Pathol. 26:41-50.

Willson. P.J., H.G. Deneer. A. Potter. and W. Albritton. 1989.
Characterization of a streptomycin- sulfonamide resistance plasmid from
Actinobacillus pleurapneumaniae. Antimicr. Agents Chemother. 332235-238.

Chapter 2

CHARACTERIZATION OF ACTINOBACILLUS PLEUROPNEUMONIAE RIBOFLAVIN
BIOSYNTHESIS GENES

37

ABSTRACT

Four genes encoding enzymes involved in riboflavin biosynthesis have been
cloned from Actinobacillus pleurapneumaniae (APP) serotype 5. These genes, which
are encoded on a single 5.0 kb Hinalll DNA fragment and are expressed from the
endogenous APP promoter in E. cali DH5a, specify the production of large amounts of
a bright yellow, water-soluble, fluorescent compound identified as riboflavin by mass
spectroscopy and characteristic absorbance spectra. Nucleotide sequence analysis of
the cloned DNA demonstrated four open reading flames of 1232, 647. 1205. and 461
bps. Based on homology with the riboflavin biosynthetic operon of Bacillus subtilis.
these ORFs were designated ribG. ribB, ribA, and rin. Predicted amino acid
sequences showed strong homology with related enzymes from B. subtilis and E. cali,
and suggested that the APP ribGBAH genes encode, respectively. a rib-specific
deaminase, the a-subunit of riboflavin synthase, a bifunctional enzyme containing
GTP cyclohydrase and 3,4-dihydroxy 2-butanone 4-phosphate synthase (DHBP)
activities, and the p-subunit of riboflavin synthase. Minicell analysis and SDS-PAGE
demonstrated bands of comparable molecular weights to those predicted from the
sequence. The cloned APP genes were able to complement mutants of E. cali
deficient in GTP cyclohydrase ll (RibA), 3,4-DHBP synthase (RibB) and riboflavin
synthase (RibC) activities. The APP genes were expressed canstitutively in E. coli.
presumably due to the lack of a cloned trans-acting repressor such as a homologue to
the B. subtilis ribC gene product. Further analysis of the APP rib operon and its
regulation should yield interesting information on APP genetics and on the role of

riboflavin biosynthesis in the pathogenesis of infections caused by this organism.

38

INTRODUCTION

Actinobacillus pleurapneumaniae (APP) is the causative agent of porcine
pleurapneumania (9.23.39). The disease is characteristically an acute necrotizing
hemorrhagic bronchopneumonia. with accompanying fibrinous pleuritis (9,39).
Pleuropneumonia is an economically devastating, severe and often fatal disease with
clinical courses ranging from peracute to chronic infection (9.14). The existence of at
least twelve antigenically distinct capsular serotypes (31) has made development of a
cross-protective vaccine difficult. Killed whole cell bacterins provide at best serotype-
specific protection (25.26.35.43). In contrast, natural or experimental infection with
virulent APP frequently elicits protection against reinfection with any serotype
(24.25.27). Avirulent strains of APP have been tested as live vaccines and have
elicited cross-protective immunity against subsequent challenge (15.28.44). However,
the use of live vaccines in the field is problematic, particularly when the attenuating
lesions in the vaccine strain have not been genetically defined. Development of
attenuated strains with defined biochemical mutations that limit growth in viva and
prevent reversion to wild type is a promising approach to improved vaccines against
APP infection.

Riboflavin (vitamin 82). a precursor of the coenzymes flavin adenine
dinucleotide (FAD) and flavin mononucleotide (FMN), is essential for basic metabolism.
It Is synthesized by plants and by most microorganisms but not by higher animals (1).
Many pathogenic bacteria are apparently unable to utilize flavins from their environment
and are entirely dependent an endogenous production of riboflavin (38). Therefore.
riboflavin biosynthesis may be essential for survival of pathogens in vivo. and mutations
in the riboflavin biosynthetic pathway may be attenuating.

The proposed metabolic pathway for bacterial riboflavin synthesis shown in

Figure 2.1 begins with guanosine triphosphate (GTP ) as the precursor (for a review

39

Figure 2.1 - Proposed Bacterial Riboflavin Biasynthesis Pathway. Proposed gene
functions are as indicated although the functions of ribG and ribT have not been
determined conclusively. Structures correspond to the following: I, GTP; II, 2.5-
diamino-6-(ribosylamino)-4(3H)-pyrimidinone 5'-phosphate; Ill. 5-amino-6-
(ribosylamino)-2,4(1H,3H)-pyrimidinedione 5'-phosphate; IV. 5-amino-6-(ribitylamino)-
2.4(1H,3H)-pyrimidinedione 5‘-phosphate; V, 5-amino-6-(ribitylamino)-2.4(1H,3H)—
pyrimidinedione; Vl. ribulose 5'-phosphate; VII, 3.4-dihydroxy-2-butanone 4-
phosphate; Vlll, 6.7-dimethyl-8-ribityllumazine; IX. riboflavin. Structures are adapted
from Bacher (1).

 

41

see reference 1). The most extensively studied system for bacterial riboflavin synthesis
is Bacillus subtilis (for a review see reference 29). The B. subtilis riboflavin synthesis
genes are located and co-regulated in an operon structure (12) that consists of five
open reading frames designated as ribG, rib 8. 11b A, rin and n'bT (19,29). The
ribGBAHT genes encode. respectively, a rib-specific deaminase; the a-subunit of
riboflavin synthase (lumazine synthase); a bifunctional enzyme containing GTP
cyclohydrase and 3.4-dihydroxy 2-butanone 4-phosphate synthase (DHBP) activities;
the p-subunit of riboflavin synthase; and a rib-specific reductase (29). The complete
sequence of the B. subtilis riboflavin operon has been determined in two individual
laboratories (19.30). The B. subtilis structural ribGBAHT genes code for predicted
proteins of 361 (MW 39.700), 215 (MW 23.600), 398 (MW 43,800). 154 (MW 16,900),
and 124 (MW 13.600) amino acids in length (19. 29). Two functional promoters have
been identified in the B. subtilis n’b operon. The main promoter, P1. for which a
transcriptional start site has been determined 294 bps upstream of ribG (12,30). is
responsible for transcription of all five structural genes (12). Another promoter, P2,
produces a secondary transcript encoding the last three genes of the operon, n'bAHT
(12). A possible third promoter has been postulated that would express rin (7). In
addition, the operon has been shown to be transcriptionally coregulated (30) by a trans-
acting repressor, RibC (3,6). which acts at a regulatory site, n’bO (3.20), upstream of
ribG, apparently by a transcription tennination-antitermination mechanism (29). The
RibC repressor appears to respond to FMN and FAD. as well as to riboflavin and
several of its biosynthetic intermediates, and regulates expression from both ribP1 and
ribP2 (4.20.29).

E. coli is the second most chemically characterized system for riboflavin
synthesis. In contrast to B. subtilis, the rib genes of E. cali are scattered around the
chromosome and are expressed constitutively (2,46). Rather than having a bifunctional

ribA, E. cali has two separate genes, ribB and n’bA. that encode the functions of 3.4-

42

DHBP synthase (34) and GTP cyclohydrase II (33), respectively. ribB is homologous to
the 5' end of B. subfilis ribA while n‘bA is homologous to the 3' and (33.34). E. cali
genes with sequence homology to the B. subtilis ribG (42), rin (42), and ribB genes
have also been identified.

In this study, we have identified a fragment of APP serotype 5 chromosomal
DNA that triggers overproduction of riboflavin when cloned in E. coli. Nucleotide
sequence analysis demonstrated four open reading frames with significant identity and

a similar operon arrangement to the ribGBAH genes from Bacillus subfilis.

43

MATERIALS AND METHODS

Bacterial strains and media. A. pleurapneumaniae lSU178, a serotype 5
strain. was cultured at 37°C in brain heart infusion broth or agar (Difco Laboratories,
Detroit, MI) containing 10 [la/"II B-NAD (Sigma Chemical Company. St Louis, MO). E.
cali DH5—a (supE44, AlacU169, (e80lacZAM15), hst17, recAf. endA 1. gyrA96, thi-1,
reIA 1) was used for construction of the APP genomic library. E. cali strain DS410 (azi-
8, tonA2, minA 1, minB2, rpsL135. xyl-7, mil-2. thi-1, t-) was used for minicell isolation
and protein labeling experiments. E. coli ribA:Tn5 (BSV18), n'szTn5 (BSV11) and
ribC:Tn5 (BSV13) mutants used for complementation studies were described by
Bandrin et al (2) and were kindly provided by Barbara Bachmann (E. cali Genetic Stock
Center, Yale University). E. cali strains were cultured in Luria-Bertani medium or in
M9 (36) supplemented with 15 gIL NZ amine (Sigma) and with riboflavin at 200 uglmL
when necessary. Ampicillin was added to 100 pg/ml for plasmid selection.

DNA manipulations. DNA modifying enzymes were supplied by Boehringer-
Mannheim Biochemicals (lndianapolis, IN) and used according to manufacturer's
specifications. Genomic and plasmid DNA preparations, gel electrophoresis, and E.
coli transformation were all performed by conventional methods (36).

Cloning and sequencing. APP serotype 5 genomic DNA was digested with
Hindlll and fragments ranging in size from 4 to 7 kb were ligated into the Hindlll site in
the polyllnker of the plasmid vector pUC19 (45). A recombinant plasmid, designated
pTF10, which overproduced riboflavin was isolated from this library. Unidirectional
nested deletions were constructed with exonuclease Ill and S1 nuclease digestion,
using the Erase-a base system (Promega Corp., Madison. WI). Nucleotide sequencing
was performed on alkali-denatured double-stranded DNA by the dideoxy chain-
tennination method of Sanger et al. (37) using the Sequenase 2.0 kit (U.S.
Biochemical, Cleveland, OH) and [35S]dATP (Amersham Corp., Arlington Heights. IL).

Sequencing primers used included universal forward and reverse primers for pUC

44

sequencing (U.S. Biochemicals), as well as several oligonucleotide primers designed
from previously obtained sequence data. These internal primers were synthesized by
the Michigan State University Macromolecular Structure Facility and included MM4 (5'-
AAT-CCG-GCA-AAA-ATT-GAA-GGC-a'). MM5 (5'-GCA-CCG-TGA-CGC-ACT-AAC-G-
3'). MM6 (5'-GCG-CCA-ATA-C‘l'l'-GCT-CAC-CG-3'). MM9 (5'-GGT-T'l'C-TTT-A‘IT-
CGT-ATG-CGG-3'). MM10 (5'-TGA-AGA-AGT-CGG-CAA—A'l'T-GCT-C-3‘). MM11 (5'-
CGG-ATT-GGG-A‘I'T-CGT-CCA-GCC-S'). MM13 (5'-GGC-GAC-ACG-ATl’-GCG-GTG-
3'), MM14 (5'-GCC-AGT-TAG-TGC-AGA-CAG-CG-3'). and MM38 (5'-CTC-ACC-GGT-
TCC-TGC-CAA-ACC-3').

DNA sequences were analyzed using the GCG sequence analysis programs
(1 1).

Mass spectroscopy. Positive and Negative Ion Fast Atom Bombardment (FAB)
mass spectroscopy was performed at the Michigan State University Mass Spectroscopy
Facility.

Quantification of riboflavin. Bacterial cells were pelleted in a microfuge, and
the absorbance at 445 nm of the culture supernatant was measured using a Beckman
DU-7 spectrophotometer (Beckman Instruments, Fullerton, CA). The absorbance at
445 nm was multiplied by a factor of 31.3 to yield the riboflavin concentration in mglliter
(1a).

Minicell Analysis. The minicell-producing E. coli strain DS410 (32) was
transformed by calcium chloride/heat shock treatment with pUC19 or pTF rib clones.
Transforrnant colonies which produced a large number of minicells were selected by
microscopy. Cultures were grown overnight at 37°C in 500 ml LB broth, and minicells
were isolated by differential centrifugation followed by glass fiber filtration as described
by Christen et al (8). Minicells were resuspended to an 00590 of 0.5-1.0 in 200 pl
labeling mix (22.0 ml M9 media. 20.0 ml 50 mM HEPES pH 7.5, 2.5 ml of 20% glucose,
0.05 ml of 10 mg/ml adenine. 0.05 ml of 10 mg/ml pyridoxine, 5.0 ml of NEDA amino

45

acid stock (21) lacking methionine and cysteine, and 0.2 ml of 10 mg/ml cycloserine-D)
and incubated at 37°C for 30 minutes. Trans-label ([3581methionine plus
[35S]cysteine, ICN Biomedicals, Irvine, CA) was added to a final concentration of 22 u.
Ci per reaction and cells were incubated at 37°C for1 hour. Total and TCA precipitable
counts were measured by liquid scintillation counting to determine amount of
incorporation. Cells were pelleted in a microfuge and washed with cold HEPES (50
mM, pH7.5) plus 10 mM methionine plus 10 mM cysteine. Labeled proteins (50.000
cpm/lane) were separated by discontinuous SDS-PAGE on a 12% polyacrylamide gel
and were visualized by autoradiography on Kodak XAR-5 film.

Nucleotide sequence accession number. The nucleotide sequence of the A.
pleurapneumaniae ribGBAH genes has been submitted to GenBank and assigned an

accession number of: U27202.

46

RESULTS

Identification of a riboflavin producing alone. A genomic library of A.
pleurapneumaniae serotype 5 DNA was constructed in pUC19 and transformed into E.
coil DH5-a. A single clone, designated pTF10 (Figure 2.2), containing a 5.0 kb insert,
was identified that produced a bright yellow extracellular, water-soluble compound that
fluoresced under ultraviolet light. The compound was crudely purified by filtration
through a 3000 Da cut off membrane filter (Amicon Corporation. Bedford, MA).
Absorbance spectra of this compound in aqueous solution under neutral conditions
showed absorbance peaks at 373 and 443 nm, which coalesced to a single peak at
388 nm under acidic conditions; these results compared well to a riboflavin standard
(Figure 2.3). Positive and negative ion fast atom bombardment mass spectroscopy
indicated that the compound was a flavin (data not shown). Culture of E. cali DH5-a/
pTF10 in M9 medium plus NZ-amine plus 0.6% glucose yielded 10 mg riboflavin per
liter in 24 hours.

Sequence of APP rib genes. The nucleotide sequence and corresponding
predicted amino acid sequence of a 4312 bp region of the insert in pTF10 is shown in
Figure 2.4. Four open reading frames of 1232. 647, 1205. and 461 bp were observed
that encoded proteins with predicted molecular masses of 45,438 Da. 23,403 Da,
44,739 Da and 16,042 Da. respectively. Based on homology with the riboflavin
biosynthetic genes of B. subtilis (T able 2.1). these ORFs were designated ribG, ribB,
n’bA. and rin, respectively. All four ORFs were preceded by potential ribosome
binding sites (RBS), although the RBS upstream of ribG is not as strong as the other
three. Production of riboflavin by this clone is not dependent on its orientation in
pUC19 or on induction by IPTG, indicating that it is produced under the control of a
native promoter included in the cloned DNA fragment A consensus promoter
sequence of the -35I-10 type (12) was identified within the sequenced region 224 bp

upstream from the n'bG start codon. A second potential consensus promoter was

47

Hindlll
Sphl Ndel
Pstl
Sail
Xbal
BamHl
Smal
Kpnl
Sstl
EcoRl

 

g Hindlll

Figure 2.2 - Physical Map of the Construct pTF10.

48

Riboflavin and Recombinant Riboflavin

 

 

0.4
3 0.3 -———Riboflavin
E
g 0-2 —Recombinant
4: 0,1 Riboflavin
<
0.0 . i r
300 400 500
Wavelength (nm)
Riboflavin and Recombinant Riboflavin
with Acidification
0.4
3 0 3 -———Riboflavin
C
(U
E 0 2 —Recombinant
§ 01 Riboflavin
0.0 T = l
300 400 500

. Wavelength (nm)

Figure 2.3 - Spectral Analysis of Riboflavin and Recombinant Riboflavin.

49

Figure 2.4 - Complete Nucleotide Sequence of APP ribGBAH Genes and Flanking
Regions. The amino acid translations are shown for ribG, ribB. ribA, and rin and
correspond to base pairs 330-1560, 1685-2330. 2393-3596 and 3709-4168. Putative
ribosome binding sites are underlined. Potential promoters for the operon and for rin
are double-underlined. An inverted repeat that may function as a transcription
terminator is overlined.

61

121

181

241

301

361

421

481

541

601

661

721

781

841

901

961

1021

1081

1141

AATTCGGTCGGACGTACTTT

TCAGAGTGTGAAACCTGTAA
-10

CCGATTACTTTATTATCAAT

CAAAAGTCGCAAATTTGCAA

TCGGCGGTAAAGTCCGCGAG

TACCGACAGTATAGTCTGGA

S F L Q A. L
TATCTTTTTTACAAGCCTTG

M P V’ M C F
ACATGCCTGTAATGTGTTTT

Y M R R .A I
ACTATATGCGCCGTGCCATT

L V G C 'V I
CGCTTGTCGGTTGTGTAATT

I G G W H A
AGATTGGTGGATGGCATGCG

.A T .A Y V' T
GGGCGACTGCTTATGTAACG

D L L I E R
CGGATTTATTAATTGAACGA

L V .A G R G
CTTTAGTAGCAGGGCGGGGA

L L K E E C
GTTTACTCAAAGAAGAATGT

R P Y V' L M
AACGTCCGTATGTGCTAATG

G E S K W I
GCGGCGAATCCAAATGGATT

Q Y S .A I M
ATCAATATAGTGCGATTATG

S R M P N .A
ATAGCCGAATGCCGAATGCG

P L D C Q L
CACCGTTAGATTGCCAGTTA

50

ATTTGAGCATATCAATGAAG

ATGGCAGATTGATATGTCGA

GGCATTGGATAAACGCTAAT

CAATTTTTTAATAATCTTCA

CCGAACGAAAAAGGTTTGGC

M K L P
TGGAAGAAGATGAAATTACC

R S K D F K A
AGATCGAAAGATTTCAAGGC

P L P S N S F
CCTCTGCCCTCAAATAGTTT

.A L .A K Q G L
GCACTGGCAAAACAAGGTTT

V K N G E I V
GTCAAAAACGGTGAAATCGT

E R N A.'V L H
GAACGTAATGCCGTTTTACA

L E P C C H H
CTTGAGCCTTGTTGTCATCA

G I K K V' F I
GGCATTAAAAAAGTATTTAT

A. N Q L R Q .A
GCAAATCAGCTACGCCAAGC

D A. L N P I F
GATGCGTTAAACCCGATTTT

K Y A. M T A. D
AAATATGCCATGACGGCAGA

T G E S A. R .A
ACCGGTGAATCGGCAAGAGC

V G V D T V’ L
GTCGGTGTAGATACGGTACT

K Q P V R I V
AAACAACCGGTCCGGATTGT

V Q T A. K E Y
GTGCAGACAGCGAAAGAATA

GAGGTTTTGATTATGTGATT
-35

GCAATGTGACTTGTTTACAT

TCTTGCTTGACTTTGACAAT

GGGCAGGGTGAAATTCCCGA

AGGAACCGGTGAGATTCCGG

C K R W F F L
GTGTAAGCGGTGGTTTTTCC

F F I I R V N
TTTTTTCATCATTAGGGTAA

K T M T D L D
CAAAACAATGACGGATTTAG

G W T N P N P
AGGCTGGACGAATCCCAATC

.A E G Y H E K
TGCCGAAGGTTACCATGAAA

C K E D L S G
TTGTAAGGAAGATCTTTCCG

G R T P P C S
CGGCCGCACGCCGCCTTGTT

G S S D P N P
CGGTTCGAGCGATCCGAATC

G V' E V ‘V E G
CGGCGTGGAAGTGGTGGAAG

F H Y I Q T K
TTTCGACTATATTCAAACTA

G K I .A T G S
CGGCAAAATTGCAACCGGTA

R 'V Q Q T R H
AAGAGTGCAGCAAACACGTC

A. D N P M L N
TGCCGATAACCCGATGTTAA

C D S Q L R T
CTGCGATAGCCAATTACGTA

R T V I .A T V
TCGCACCGTAATTGCAACCG

1201

1261

1321

1381

1441

1501

1561

1621

1681

1741

1801

1861

1921

1981

2041

2101

2161

2221

2281

2341

S D D L Q K
TTAGTGACGATTTGCAAAAA

K A R N K R
GTAAAGCACGAAACAAGCGG

I D S L L L
AGATCGACAGCCTCTTATTG

I V N R V’ R
GTATCGTGAATCGAGTACAT

T P I G G E
AAACCCCAATCGGCGGTGAG

S T E L I G
AATCGACCGAACTCATCGGC

AGCAAAGAGGGGTCGGGGGA

GGCGTTGTTAAATCTCCCCT

M F T G I I
TTATATGTTCACAGGTATTA

E F .A V V T I
CGAATTTGCGGTAGTCACAA

T I A. V N G V
CACGATTGCGGTGAACGGCG

A. D V M S E T
CGCCGATGTAATGTCGGAAA

P V N L E R .A
TCCGGTTAATTTAGAACGCG

G H I D G T G
GGGGCATATTGACGGCACCG

Y R I K T S P
GTATCGCATTAAAACCTCTC

I D G I S L T
CATTGACGGTATTAGCCTGA

I P H T I K E
TATTCCGCATACGATTAAAG

L E N D I V G
TTTAGAAAATGATATTGTCG

E P K S N L S
TGAGCCGAAAAGTAATCTTA

GGACACACTGAGTGTATCCT

51

I E Q F R P L
ATTGAACAATTTAGACCGCT

V D L Q D L L
GTAGATTTGCAAGATCTTTT

E G G S S L N
GAAGGCGGTTCAAGTTTGAA

C Y I .A P K L
TGTTATATTGCGCCTAAATT

G I Q Q I D Q
GGAATTCAACAAATCGACCA

E D I L L D Y
GAAGATATTTTGTTGGATTA

GATTTGAGATAATGTTGAAA

AACCCCTCTTTACAAAAGAG

E E V G K I
TTGAAGAAGTCGGCAAAATT

N A. T K V L
TTAATGCGACCAAAGTATTA

C L T V T S
TATGTTTAACCGTAACTTCT

L K R T S L
CGTTAAAACGTACTTCATTA

M .A .A N G R
CGATGGCGGCAAACGGACGT

E I A. E I T
GCGAAATTGCGGAAATCACA

K L M R Y I
CAAAATTAATGCGTTATATT

V ‘V D T D D
CCGTAGTCGATACCGATGAT

T N L G S K
AAACCAATTTAGGTTCGAAA

K Y I E Q F
GTAAATATATCGAACAGTTT

L D F L K Q
GTTTAGACTTTTTAAAGCAG

ACCGACAAAAATATATATTT

G V' D V L V C
TGGCGTAGATGTATTAGTGT

Q K L G E M Q
GCAAAAGCTCGGTGAAATGC

F S .A L E S G
TTTCAGTGCGTTAGAAAGCG

V G G K Q .A K
AGTCGGTGGTAAACAAGCGA

.A V K L K L K
AGCGGTTAAATTAAAATTGA

V 'V I S P L *
TGTAGTCATCTCCCCTCTTT

TTTACACCGCCTTTCACTTT

AGGGATCAATAATGAGGAAA

A. Q I H K Q G
GCTCAAATTCATAAGCAAGG

Q D V H L G D
CAAGACGTTCATTTAGGCGA

F S S N Q F T
TTTTCGAGTAATCAGTTTAC

G E L K S N S
GGCGAATTAAAGTCGAATAG

F G G H I ‘V S
TTCGGCGGACACATCGTTTC

P .A H N S T W
CCGGCACATAATTCGACATG

I E K G S I T
ATTGAGAAAGGTTCGATCAC

E S F R V S I
GAAAGTTTCCGTGTATCGAT

K I G S I V' N
AAAATCGGCAGTATTGTCAA

L L K K P .A D
TTACTGAAAAAGCCGGCGGA

A G F *
GCGGGATTTTAAGATTTGTA

M T D
TAGGAAAAGAAGATGACAGA

2401

2461

2521

2581

2641

2701

2761

2821

2881

2941

3001

3061

3121

3181

3241

3301

3361

3421

3481

F Q F S K V E
TTTCCAATTTTCAAAAGTAG

V T D D E D R
AGTGACTGACGATGAAGATC

T P E N I N F
CACACCGGAAAATATCAATT

S T E I A K K
TTCAACCGAAATCGCTAAAA

H E T A. F T V
TCATGAAACGGCGTTTACCG

F E R S I T A
TTTTGAGCGTTCGATTACCG

R R P G H M F
CCGCCGCCCGGGGCATATGT

G H T E A. T V
CGGTCATACCGAAGCAACAG

C C E I M .A D
ATGTTGTGAAATTATGGCGG

.A V E H N M P
TGCGGTAGAACACAATATGC

H D S L V K Q
GCATGACAGCTTGGTGAAAC

M .A H S F V E
TATGGCACATAGCTTTGTTG

D L T D G E Q
CGATTTAACCGACGGTGAGC

A F G S Q R C
CGCTTTCGGTTCTCAACGTT

E Q E G R G V
TGAGCAAGAGGGCAGAGGTG

I N K L R A Y
AATCAATAAGCTACGTGCTT

V .A L G F K E
CGTCGCTTTAGGATTTAAAG

Q L G ‘V K S I
GCAGTTAGGCGTAAAATCGA

K E Q G L N I
AAAAGAGCAAGGATTAAATA

52

D A I E A I
AAGATGCGATCGAAGCGATT

E N E G D F
GCGAAAACGAAGGCGATTTT

M A T Y G K
TTATGGCAACTTACGGCAAA

L N F H P M
AATTAAATTTCCATCCGATG

S V D H I D
TATCGGTGGATCATATTGAT

M K I V D D
CAATGAAAATTGTCGATGAT

P L I A. K E
TTCCGTTAATCGCTAAAGAA

D L .A R L .A
TGGATTTAGCTCGTTTAGCC

D G T M M T
ATGACGGCACGATGATGACT

F I T I Q Q
CGTTTATCACGATTCAACAA

I S V"V K M
AAATTTCTGTGGTAAAAATG

V I S G K E
AAGTGATTTCAGGTAAAGAA

V L .A R I H
AAGTATTGGCGCGTATCCAT

D C G Q Q F
GTGATTGCGGTCAGCAATTT

I L Y L R Q
TGATTCTGTATTTACGCCAA

E L Q D K G
ACGAACTACAAGATAAAGGG

D E R E Y Y
AAGACGAACGTGAGTACTAT

R L L T N N
TCCGTTTATTAACCAATAAT

V .A R E P I
TCGTTGCACGTGAGCCGATT

R Q G K I I L
CGACAAGGCAAAATCATTTT

I C A..A E F .A
ATCTGTGCGGCGGAATTTGC

G L I C T P I
GGTTTGATTTGTACGCCGNT

V .A V' N Q D N
GTTGCGGTCAATCAAGATAA

T G T G I S .A
ACGGGAACGGGTATCTCAGC

N A. K .A T D F
AATGCTAAAGCAACGGATTT

G G V L V' R N
GGTGGAGTGTTAGTGCGTAA

G L K H .A G L
GGTTTAAAACACGCCGGTTT

M P D L Q K F
ATGCCGGATCTACAAAAATT

L Q E Y R R K
TTACAAGAATATCGCCGTAA

P T K Y G E F
CCGACAAAATACGGTGAGTT

H V .A L V K G
CACGTTGCGTTAGTCAAAGG

S E C L T G D
TCGGAATGTTTAACCGGTGA

A. A. A. M T Q I
GCCGCAGCAATGACCCAAAT

E G R G I G L
GAAGGTCGTGGTATCGGTTT

M D T V E .A N
ATGGATACCGTTGAAGCGAA

I G .A Q M F Q
ATCGGTGCACAAATGTTCCA

P A. K I E G L
CCGGCAAAAATTGAAGGCTT

I V E P N K N
ATTGTAGAACCGAACAAAAA

3541

3601

3661

3721

3781

3841

3901

3961

4021

4081

4141

4201

4261

D I D Y L K V
TGACATTGATTACCTAAAAG

TTAACAACCGTATGTAGTAT

-10

53

K Q I K M G
TCAAACAGATAAAAATGGGG

TAGGGAAGCAAGCGTTGCGT

H M F N F *
CATATGTTTAACTTCTAACT

-35
CCCTACTATAGAATGATACA

M .A K I

AGCGGTCACTTTTTTATAAA.ATTTTGCATATTTCGAGAGG ACAAAAAAATGGCAAAGATT

T G N L V .A T
ACAGGTAACTTAGTTGCGAC

F I N D K L L
TTTATCAACGATAAATTATT

E N D I D T A
GAAAACGATATTGATACGGC

K M A. N S G K
AAAATGGCAAACAGCGGTAA

S T T H Y D Y
TCGACAACTCACTATGATTA

L E T G V P V
TTAGAAACCGGCGTACCGGT

I E R .A G T K

G L K F G I V
AGGTTTAAAATTCGGTATTG

S G A. I D T L
AAGCGGTGCAATTGATACGT

W V P G .A F E
ATGGGTTCCGGGTGCATTTG

Y D A. V I C L
ATATGATGCGGTAATCTGTT

V C N E A. A. K
CGTATGTAATGAAGCGGCAA

I F G V L T T
AATTTTCGGTGTATTAACCA

A G N K G S E

T A. R F N D
TAACCGCACGTTTCAACGAT

V R H G .A Y
TAGTGCGTCACGGTGCGTAT

I P L V'.A K
AGATTCCATTAGTTGCGAAA

G T V I R G
TAGGTACGGTAATTCGCGGT

G I G .A V .A
AAGGTATCGGTGCGGTAGCA

E N I E Q A
CAGAAAATATTGAACAGGCG

C A. L G .A I

ATTGAACGCGCGGGTACTAA.AGCAGGTAATAAAGGTTCAG AATGTGCATTAGGCGCAATC

E I V N ‘V L K

A I *

GAAATAGTAAACGTATTAAA.AGCGATCTAATTTTCGTTTG.ACGTGCTAAAAACAAGCGGT

 

CGTTTTTGACTGGAATTTTG

.AGTAAAGACCTTCTTTCTCG

 

CAAATTTCCCGTTAAAAACG

TACCAGATTTTGTTGATATA

ACCGCTTATATTTTATGTCT

TAGCAAGCTTGG 4312

Table 2.1 - Comparison of Amino Acid Sequence Similarity Between Riboflavin
Synthesis Proteins. '

 

% Similarity with A. pleurapneumaniae

 

 

 

RibG RibB RibA Rin
Bacterium Compared % Compared % Compared % Compared %
with with: with: with:
Bacillus subtilis RibG 63 R1138 69 RibA 73 Rin 83
Escherichia coli b mm 62 RibC 58 RibB 63 11in 74
RibA 73
Haemaphilus influenzaec RibG 58 RibC 60 R1133 65 11le 75
. . RibA 71
Photobacterium leiognathi NA RibI 64 Rin 61 RibIlI 69
Photobacterium phosphoreum d NA mm 63 RibII s9 Rihlll 73
RibIV 63
Vibrio harveyi NA NA Until 59 NA

 

3 Identity is expressed in percent similarity as calculated by the GCG Needleman—thsch algorithm (22). Proteins
withpartial identitywereeomparedtotheentimappropfiateAPPRibprotein.

t’I:'.coli RibB is homologous to the 5' end ofAPP RibA. Ecoli RihA is homologous to the 3‘ end ofAPP mm.

c Hinfluenzae RibB is homologous to the 5' end of APP RibA. fun/lame RihA is homologous to the 3' end of APP
RibA.

d P.phosphorcwn RibIV is homologous to the 3' end ofAPP RibA.

55

identified between the genes n‘bA and n'bH. However, no consensus promoter was
identified between n‘bB and ribA, as is found in B. subtilis. The ORF encoding n‘bH is
followed by an inverted repeat stem-loop structure with a AG = -56.0 that may function
as a rho-dependent transcriptional terminator (13).

Homology of APP rib genes. Predicted amino acid sequences of the APP
RibGBAH proteins were compared with B. subfilis RibGBAH (19); E. coli RibA, RibB,
RibC, RibG, and Rin (33.34.42); Photobacten‘um Ieiognathi Ribl-lll (17),
Photobacten‘um phosphoreum Ribl-IV (16), and Vibn‘o harveyi LuxH (41) proteins, using
the SOS Gap program (Table 2.1). APP RibG showed 62-63% similarity to the RibG
proteins from B. subtilis and E. coli. APP RibB showed 58—69% and APP Rin showed
69-8396 similarity to homologous genes from B. subtilis, E. coli, and Photobacten'um
species. APP RibA showed 73% similarity to the entire RibA protein of B. subfilis and
61% to the Ribll protein of P. Ieiognathi, both of which encode a bifunctional enzyme
catalyzing two distinct steps in the riboflavin pathway. In addition, the carboxy terminal
half of APP RibA, encompassing ~200 amino acids, shows 59—63% similarity to E. coli
RibB, and V. harveyi LuxH, which encode 3,4-DHBP synthase. The N-terminal region
of APP RibA, encompassing the remaining ~200 amino acids, shows 63-73% similarity
to E. coli RibA and P. phosphomum RiblV, which encode GTP cyclohydrase ll.

Complementation of E. coll mutants. The original pTF10 clone and several
deletion derivatives were tested for their abilities to complement ribA (GTP
cyclohydrase II), ribB (3,4-DHBP synthase), and n‘bC ([3- subunit of riboflavin synthase)
mutations in E. coli (2) (Figure 2.5). Complementation of the E. coli mutation was
determined by restoration of the ability to grow on M9 minimal medium in the absence
of riboflavin. Plasmids containing a complete copy of the APP n’bB gene
complemented the E. coli n'bC mutation. Plasmids containing the 5' end of APP n’bA
complemented the E. coli ribB mutation. Plasmids containing a complete copy of APP

n'bA complemented both E. coli n‘bB and mm mutations.

56

Minicell analysis. Plasmid pTF1O and its deletion derivatives were transformed
into the minicell—producing E. coli strain DS410, and the proteins encoded by these
plasmids were radioactively labeled, separated by SDS-PAGE, and visualized by
autoradiography. Compared with the pUC19 vector, plasmid pTF1O shows four unique
proteins with apparent molecular masses of 45 kDa, 27.7 kDa, 43.7 kDa, and 14.8 kDa
(Figure 2.6), which correspond well with the sizes predicted for the RibG, RibB, RibA,
and Rin proteins by amino acid sequence data. The RibG protein did not appear to
be as strongly expressed as RibB, RibA, and Rin. Analysis of proteins encoded by
plasmid pTF 19 (Figure 2.5), which contains no rin and a slightly truncated n‘bA gene,
eliminates the 14.8 kDa protein (Rin) and truncates the 43.7 kDa protein (RibA) to
42.5 kDa (Figure 2.6). Plasmid pTF12 (Figure 2.5), which does not contain ribB, ribA,
or n’bH genes, does not express the 27.7, 43.7, or 14.8 kDa proteins (data not shown).

57

EcoflEcollEcol
rectum

 

 

 

 

 

 

 

 

Cornplernentadon rle ribB ribA rle
ofE.coli mm mm * IIIIIIIIIIIIIIIWW _
it K E C All I ll
ribA ' use ' 7le '
"d - - pm: }
' - + pTF26 }
' ' "d pTF23 }
' + M pTFZZ :
- 4» nd pTF19 }
'l' + nd er16 i
+ 'I' + pTFio [L
Farr—l

Figure 2.5 - Complementation of E. coli Mutants by Cloned APP rib Genes. A
physical map for the APP ribGBAH genes is shown as well as several deletions that
were made from the 3' end of the APP n’b clone. The E. coli gene designations are
indicated above their APP homologues. A ”-0-” indicates complementation of the
indicated E. coli mutation by the recombinant plasmid. nd = not done. Enzyme
abbreviations: A, Aval; C, Clal; E,EcoRl; H,Hindlll; K, Kpnl; N, Ndel

Figure 2.6 - Minicell Analysis of pTF10 and Deletions. Minicells contained: Lane 1,
pUC19; Lane 2, pTF10; Lane 3, pTF19. Molecular weight standards are indicated on
the left. Proteins encoded by the APP genes are indicated by the arrows on the right.
Apparent molecular weights for the APP Rib enzymes are: RibG, 45 kDa; RibA, 43.7
kDa; RibB, 27.7 kDa, and Rin, 14.8 kDa.

59

DISCUSSION

In this paper, we report the identification, cloning and complete nucleotide
sequence of four genes from Actinobacillus pleurapneumaniae that are involved in
riboflavin biosynthesis. The cloned genes can specify production of large amounts of
riboflavin in E. coli, can complement several defined genetic mutations in riboflavin
biosynthesis in E. coli, and are homologous to riboflavin biosynthetic genes from both
E. coli and Bacillus subtilis. The genes have been designated APP ribGBAH due to
their similarity in both sequence and arrangement to the B. subtilis ribGBAH operon.

The DNA sequence data, complementation, and minicell analysis strongly
suggest that the four at: genes are transcribed from a single APP promoter upstream of
the n'bG gene. This promoter, among the first described for housekeeping genes in
APP, is a good match for an E. coli consensus -35I-1O promoter. There is a 4 of 6 bp
match at the -35 region, a 17 bp interval, a 4 of 6 bp match at the -10 region, an 8 bp
interval, and a CAT box at the -1/+1 site. There is also a second potential promoter
located between n’bA and n'bH, although it is not clear whether this promoter is
functional.

Biosynthesis of riboflavin by APP appears to be more similar to that in the gram-
positive bacterium B. subtilis than in the gram-negative bacterium E. coli. First, APP n'b
genes are arranged in an operon similar to that seen in B. subtilis, rather than scattered
throughout the chromosome as is found in E. coli. However, the B. subtilis rib operon
contains a fifth gene, n‘bT, that is proposed to mediate the third step in riboflavin
biosynthesis; it is unlikely that a n’bT homologue is present as part of the operon in
APP because of the presence of a strong inverted repeat following rin and the lack of
a likely reading frame downstream. Second, APP contains a n’bA gene that encodes a
bifunctional enzyme with both GTP cyclohydrase II and DHPB synthase activities, as is
found in B. subtilis; E. coli has two genes, n‘bA and 11728, that encode these two

enzymes separately. Finally, the APP riboflavin biosynthetic enzymes are more similar

60

at the amino acid level to the enzymes of B. subtilis than to those of E. coli, although
alignment of the proteins from all three sources shows highly conserved sequences
(data not shown).

It should be noted that in three bioluminescent species from the family
Vibn'onaceae, Vibn'o harveyi, Photobacten’um Ieiognathi, and P. phosphoreum,
riboflavin biosynthesis genes have been shown to be closely linked to the qu operon
(10,11,41). FMNHZ is the substrate for the light-emitting reaction, and therefore an
increase in bioluminescence requires an increased supply of the cofactor. Since
riboflavin is the precursor for FMN, linkage and possibly coordinate regulation of M
and rib genes may facilitate the expression of bioluminescence in these bacteria.

The recombinant E. coli DH5—a containing plasmid pTF10 showed a marked
increase in extracellular riboflavin production, most likely due to the lack of regulation
(40) and high copy number of the cloned synthetic genes (45). Although the APP n’b
operon is similar in structure to that of B. subtilis, it is not yet known whether the genes
are tightly regulated in APP by a repressor similar to B. subtilis RibC, or whether they
are constitutively expressed as appears to be true in E. coli (33). We hypothesize that
APP must synthesize riboflavin to meet its own metabolic demands during infection,
since riboflavin is not synthesized by mammals and therefore is not likely to be freely
available to APP within its porcine host. Further study of this model operon should
reveal interesting information on regulation and promoter/operon structure in APP, as

well as information on the role of riboflavin biosynthesis in APP infection.

61

ACKNOWLEDGMENTS
Mass spectral data were obtained at the Michigan State University Mass
Spectrometry Facility which is supported, in part, by a grant (ORR-00480) from the
Biotechnology Research Technology Program, National Center for Research
Resources, National Institutes of Health.
We also thank Barbara Bachmann for supplying the E.coli mutant strains,
Vladimir Motin for interpretation of Russian papers, and Robert Hausinger, Pahick Oriel

and Douglas Gage for helpful discussion.

10.

11.

12.

62

REFERENCES

Bacher, A. 1991. Biosynthesis of flavins, p. 215-259. In F. Muller (ed.),
Chemistry and Biochemistry of Flavins, vol. 1. Chemical Rubber 00., Boca
Raton, FL.

Bandrin, S.V., P.M. Rabinovich, and AI. Stepanov. 1983. Three linkage
groups of genes involved in riboflavin biosynthesis in Escherichia coli. Sov.
Genet. 19:1103-1109.

Bresler, S.E., E.l. Cherepenko, T.P. Chemik. V.L. Kalinin, and D.A.
Perumov. 1970. Investigation of the operon of riboflavin synthesis in Bacillus
subtilis. l. Genetic mapping of the linkage group. Genetika 6:116—124.

Bresler, S.E., E.A. Glazunov, G.l. Chernik, T.N. Shevchenko, and D.A.
Perumov. 1973. Investigation of the operon of riboflavin synthesis in Bacillus
subtilis. V. Flavin mononucleotide and flavin adenine dinucleotide as effectors
in the operon of riboflavin biosynthesis. Genetika 9:84—91.

Bresler, S.E., V.L. Kalinin, A.S. Kriviskii. D.A. Perumov. and T.P. Chemik.
1969. Mutant of Bacillus subtilis synthesizing notable amounts of riboflavin.
Genetika 5:133-138.

Bresler, S.E., D.A. Perumov. G.I. Chemik, A.P. Skvortsova. 1976.
Investigation of the operon of riboflavin synthesis in Bacillus subtilis. XI.
Determination of the type of regulation by a test for dominance of operator-
constitutive and regulator-constitutive mutations. Genetika 12:124-130.

Chikindas, M.L., G.I. Morozov, V.N. Mironov, E.V. Luk'yanov, V.V.
Emel'yanov, and Al. Stepanov. 1988. Regulatory regions of the riboflavin
operon in Bacillus subtilis. Dokl. Akad. Nauk SSSR 298:997-1000.

Christen, A.A., M.L. Pall, T. Manzara and P.F. Lurquin. 1983. Rapid isolation
of Escherichia coli minicells by glass-fiber filtration: study of plasmid-coded
polypeptides. Gene 23:195-198.

Fenwick. B. 1994. Porcine pleuropneumonia. J. Am. Vet Med. Assoc.
204:1334-1340.

Foster, E.w., D.C. Gyure, D.L. Heefner, C.A. Weaver, M.J. Yarus, L.A.
Burdzinski. June 1992. US Patent 5,120,655.

Genetics Computer Group. Sept. 1994. Program Manual for the Wisconsin
Package, Ver. 8. Genetics Computer Group. Madison, Wisconsin.

Hawley, D.K. and W.R. McClure. 1983. Compilation and analysis of
Escherichia coli promoter DNA sequences. Nucleic Acids. Res. 11 :2237-2255.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24

25.

63

Holmes, W.M., T. Platt, and M. Rosenber. 1983. Termination of transcription
in E. coli. Cell 32:1029-1032.

Hunneman, W.A. 1986. Incidence, economic effects, and control of
Haemophilus pleurapneumaniae infections in pigs. Vet. Quarterly 8:83-87.

Inzana, T.J., J. Todd, and HP. Veit. 1993. Safety, stability and efficacy of
nonencapsulated mutants of Actinobacillus pleurapneumaniae for use in live
vaccines. Infect. Immun. 61:1682-1686.

Lee, C.Y., D.J. O'Kane, and EA. Meighen. 1994. Riboflavin synthesis genes
are linked with the qu operon of Photobacten‘um phosphoreum. J. Bacteriol.
176:2100-2104.

Lee, CY. and EA Meighen. 1992. The lux genes in Photobacterium
leiognathi are closely linked with genes corresponding in sequence to riboflavin
synthesis genes. Biochem. Biophys. Res. Commun. 186:690-697.

Mironov, V.N., M.L. Chlkindas. A.S. Kraev, A.l. Stepanov. and K.G. Skryabin.
1990. Operon organization of genes of riboflavin biosynthesis in Bacillus
subtilis. Dok. Akad. Nauk SSSR 312:237-240.

Mironov, V.N., A.S. Kraev, B.K. Chernov, A.V. Ul’yanov. Y.B. Golva, G.E.
Pozmogova, M.L. Simonova, V.K. Gordeev, A.I. Stepanov, and K.G.
Skryabin. 1989. Genes of riboflavin biosynthesis of Bacillus subtilis-complete
primary structure and model of organization. Dokl. Akad. Nauk SSSR 3052482-
487.

eronov, V.N., D.A. Perumov. A.S. Kraev, A.I. Stepanov, K.G. Skryabin.
1990. Unusual structure in the regulation region of the Bacillus subtilis riboflavin
biosynthesis operon. Molekulyamaya Biologiya 24:256-261.

Morse, S.A. and L. Bartenstein. 1980. Purine metabolism in Neisseria
gononhoeae: the requirement for hypoxanthine. Can. J. Microbiol. 26:13-20.

Needleman, 8.8. and CD. Wunsch. 1970. A general method applicable to
the search for similarities in the amino acid sequence of two proteins. J. Molec.
Biol. 48:443-453.

Nicolet, J. 1992. Actinobacillus pleurapneumaniae, p. 401-408. In A.D.
Leman et al (eds), Diseases of Swine, 7th edition. Iowa State University Press,
Ames, Iowa.

Nielsen, R. 1979. Haemaphilus parahaemolyticus serotypes: pathogenicity
and cross immunity. Nord. Vet. Med. 31:407-413.

Nielsen, R. 1984. Haemaphilus pleurapneumaniae serotypes - Cross
protection experiments. Nord. Vet. Med. 36:221-234.

26.

27.

28.

29.

30.

31.

32.

33.

35.

64

Nielsen, R. 1976. Pleuropneumonia of swine caused by Haemaphilus
pleurapneumaniae. Studies on the protection obtained by vaccination. Nord.
Vet. Med. 28:337-338.

Nielsen, R. 1974. Serological and immunological studies of pleuropneumonia
of swine caused by Haemaphilus parahaemolyticus. Acta Vet. Scand. 15:80-
89.

Paltineanu, D., R. Pambucol. E. Tirziu, and I. Scobercea. 1992. Swine
infectious pleurapneumania: Aerosol vaccination with a live attenuated vaccine.
Proc. Int Pig. Vet. Soc. 12:214

Perkins, J.B., and J.G. Pero. 1993. Biosynthesis of riboflavin, biotin, folic
acid, and cobalamin, p. 319-334. In A. Sonenshein (ed.), Bacillus subtilis and
other gram-positive bacteria: biochemistry, physiology, and molecular genetics.
American Society for Microbiology. Washington, DC.

Perkins, J.B., J.G. Pero, and A. Sloma. January 1991. Riboflavin
overproducing strains of bacteria. European patent application 0405370.

Perry, M.B., E. Altman, J.-R. Brlsson, LM. Beynon, and J.C. Richards.
1990. Structural characteristics of the antigenic capsular polysaccharides and
lipopolysaccharides involved in the serological classification of Actinobacillus
pleurapneumaniae strains. Serodiag. Immunother. Infect. Dis. 4:299-308.

Reeve, J. 1977. Bacteriophage infection of minicells: a general method for
identification of in vivo bacteriophage directed polypeptide biosynthesis. Molec.
Gen. Genet. 158:73-79.

Richter, G., H. Ritz, G. Katzenmeier, R. Volk, A. Kohnle, F. Lottspeich, D.
Allendorf, and A. Bacher. 1993. Biosynthesis of riboflavin: Cloning,
sequencing, mapping and expression of the gene coding for GTP
cyclohydrolase II in Escherichia coli. J. Bacteriol. 175:4045-4051.

Richter, G., R. Volk, C. Krleger, H.W. Lahm, U. Rothllsberger, and A.
Bacher. 1992. Biosynthesis of riboflavin: cloning, sequencing, and expression
of the gene coding 3,4-dihydroxy-2-butanone 4-phosphate synthase of
Escherichia coli. J. Bacteriol. 174:4050-4056.

Rosendal. S., 0.8. Carpenter, W.R. Mitchell, and MR. Wilson. 1981.
Vaccination against pleurapneumania in pigs caused by Haemaphilus
pleurapneumaniae. Can. Vet. J. 22:34-35.

Sambrook, J.. E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A
Laboratory Manual, 2nd. ed. Cold Spring Harbor Press, Cold Spring Harbor,
New York.

37.

38.

39.

40.

41.

42.

43.

45.

65

Sanger, F., S. Milken, and AR. Coulson. 1977. DNA sequencing with chain-
tenninating inhibitors. Proc. Natl. Acad. Sci. 74:5463-5467.

Schott, K., J. Kellermann, F. Lottspeich, and A. Bacher. 1990. Riboflavin
synthases of Bacillus subtilis: purification and amino acid sequence of the a.-
subunit. J. Biol. Chem. 265:4204-4209.

Sebunya, T.N.K., and JR. Saunders. 1983. Haemaphilus pleuropneumoniae
infection in swine: A review. J. Amer. Vet. Med. Asscoc. 182:1331-1337.

Shavlovskli, G.M., and EM. Logvinenko. 1988. Flavin oversynthesis in
microorganisms and its molecular mechanisms (review). Prikladnaya
Biokhimiya i Mikrobiologiya 24:435-447.

Swartzman, E., C. Mlyamoto, A. Graham, and EA. Meighen. 1990.
Delineation of the transcriptional boundaries of the qu operon of .Vibrio harveyi
demonstrates the presence of two new qu genes. J. Biol. Chem. 265:3513-
3517.

Tetsuya, T.. C. Ueguchi, K. Shiba., and K. Ito. 1992. lnsertional disruption of
the nusB (ssyB) gene leads to cold-sensitive growth of Escherichia coli and
suppression of the secY24 mutation. Mol. Gen. Genet 234:429-432.

Thacker, B.J., and M.H. Mulks. 1988. Evaluation of commercial Haemaphilus
pleurapneumaniae vaccines. Proc. Int Pig Vet Soc. 10:87.

Utrera, V.. C. Pijoan, and T. Molitor. 1992. Evaluation of the immunity
induced in pigs after infection with a low virulence strain of A.
pleurapneumaniae serotype 1. Proc. Int Pig Vet Soc. 12:213.

Vielra, J. and J. Messing. 1982. The pUC plasmids, an M13mp7-derived
system for insertion mutagenesis and sequencing with synthetic universal
primers. Gene 19:259-268.

Wilson, AC. and AB. Pardee. 1962. Regulation of flavin synthesis by
Escherichia coli. J. Gen. Microbiol. 28:283-303.

 

 

Chapter 3

A RIBOFLAVIN AUXOTROPI-I OF ACTINOBA CILLUS PLEUROPNEUMONIAE IS
ATTENUATED IN SWINE

67

ABSTRACT

Actinobacillus pleurapneumaniae is the etiological agent of a highly contagious
and often fatal pleuropneumonia in swine. A riboflavin-requiring mutant of A.
pleurapneumaniae serotype 1, designated AP233, was constructed by deleting a
portion of the riboflavin biosynthetic operon (ribGBAH) and replacing it with a gene
cassette encoding kanamycin resistance. The genes affected included both the a and
B-subunits of riboflavin synthase as well as a bifunctional enzyme containing GTP
cyclohydrase and 3,4-dihydroxy-2-butanone-4-phosphate synthase (DHBP) activities.
AP233 was unable to grow in the absence of exogenous riboflavin, but othenvise was
phenotypically identical to the parent wild type strain. Experimental infection studies in
pigs demonstrated that the riboflavin-requiring mutant was unable to cause disease,
based on mortality, lung pathology and clinical signs, at dosages as high as 500 times
the normal L050 of the wild type parent. This is the first demonstration of the
attenuation of A. pleumpneumoniae by introduction of a defined mutation in a
metabolic gene and the first demonstration that mutations in the genes required for

riboflavin biosynthesis can lead to attenuation in a bacterial pathogen.

68

INTRODUCTION

Actinobacillus pleurapneumaniae, a gram negative coccobacillus belonging to
the family Pasteurellaceae, is the causative agent of porcine pleuropneumonia
(4.20.32). This disease is characteristically an acute necrotizing hemorrhagic
bronchopneumonia. with accompanying fibrinous pleuritis (4,32). Pleuropneumonia is
an economically devastating. severe and often fatal disease with clinical courses
ranging from hyperacute to chronic infection (4,9). The existence of at least twelve
antigenically distinct capsular serotypes (28) has made development of a cross-
protective vaccine difficult Killed whole cell bacterins provide at best serotype-specific
protection (22.23.29.37). In contrast, natural or experimental infection with a highly
virulent serotype of A. pleurapneumaniae elicits protection against reinfection with any
serotype (21.22.24). In several recent studies. attenuated strains of A.
pleuropneumoniae produced by chemical mutagenesis, serial passage. or other
undefined spontaneous mutation have been tested as live vaccines, with promising
results (10.27.38). However, the use of live vaccines in the field is problematic,
particularly when the attenuating lesion in the vaccine strain has not been genetically
defined. A well-defined mutation that prevents reversion to wild-type would be
extremely desirable for the development of a live attenuated vaccine against
pleuropneumonia. We hypothesize that a mutation in a critical biosynthetic pathway
which limits growth in vivo but does not othenrvise alter expression of important antigens
such as capsular polysaccharide, lipopolysaccharide and extracellular toxins, could
produce an attenuated vaccine strain capable of inducing cross-protective immunity
against A. pleurapneumaniae.

A variety of mutations in biosynthetic pathways are known to be attenuating in
other organisms. Lesions in are (6-8, 13, 17, 26, 39), pur (26,33), and thy (1) loci,
which affect the biosynthesis of aromatic amino acids, purines and thymine,

respectively, are attenuating because they eliminate the ability of the bacterium to

69

synthesize critical compounds that are not readily available within mammalian hosts.
For example, are mutants of Salmonella and Shigella species have been shown to be
attenuated in their natural hosts (6-8, 13. 17, 26). Lesions that affect the biosynthesis
of LPS (3,25) and of cyclic AMP (14, 36) have also been shown to be attenuating in
Salmonella species. It is important to note that not all attenuating mutations are good
vaccine candidates in different organisms because some attenuating mutations result
in poor persistence and immunogenicity (26,33).

Riboflavin (vitamin 82). a precursor of the coenzymes flavin adenine
dinucleotide (FAD) and flavin mononucleotide (FMN), is essential for basic metabolism.
It is synthesized by plants and by most microorganisms but not by higher animals (2).
Many pathogenic bacteria are apparently unable to utilize flavins from their environment
and are entirely dependent on endogenous production of riboflavin (31). Even with the
ability to utilize exogenous riboflavin, there may not be enough of the vitamin present in
mammalian host tissues to permit growth, particulariy not in sites devoid of normal
bacterial flora. Therefore. we hypothesized that riboflavin biosynthesis would be
essential for survival of A. pleurapneumaniae in vivo, and that mutations in the
riboflavin biosynthetic pathway would be attenuating due to the scarcity of riboflavin
present on the mucosal surfaces of the respiratory tract.

Previously we reported the identification of a fragment of A. pleurapneumaniae
serotype 5 chromosomal DNA that triggers overproduction of riboflavin when cloned in
E. coli. Nucleotide sequence analysis demonstrated four open reading frames with
significant identity and a similar operon arrangement to the ribGBAH genes from
Bacillus subtilis (5). The ribGBAH genes encode, respecfively. a rib-specific
deaminase, the a-subunit of riboflavin synthase, a bifunctional enzyme containing GTP
cyclohydrase and 3.4-dihydroxy-Z-butanone-4-phosphate synthase activities, and the )3-
subunit of riboflavin synthase (5). In this study we report the construction of a deletion-

dismption riboflavin mutant of A. pleuropneumoniae serotype 1 and show that this

70

mutation attenuates A. pleurapneumaniae in viva. This is the first published report
demonstrating that riboflavin auxotrophy can lead to attenuation of a bacterial pathogen

in its natural host.

71

MATERIALS AND METHODS

Bacterial strains and media. The bacterial strains and plasmids used in this
study are listed in Table 3.1. A. pleurapneumaniae strains were cultured at 37 °C in
either brain heart infusion (BHI), heart infusion (HI), or tryptic soy agar (T SA) (Difco
Laboratories, Detroit, MI) containing 10 pg/ml p-NAD (V factor) (Sigma Chemical
Company. St. Louis, MO). Riboflavin (Sigma) was added to a final concentration of 200
uglml when needed. E. cali strains were cultured in Luria-Bertani medium. Ampicillin
was added to 100 pglml and kanamycin to 50 rig/ml for plasmid selection in E. coli
strains. For A. pleurapneumaniae strains, 50 [Lg/"II kanamycin sulfate and 25 pglml
nalidixic acid were added as required, except for selection after matings which were
performed with 100 pglml kanamycin sulfate and 50 [IQ/rill nalidixic acid.

DNA manipulations. DNA modifying enzymes were supplied by Boehringer-
Mannheim Biochemicals (Indianapolis, IN) and used according to the manufacturer's
specifications. Genomic DNA was prepared according to the lysislproteinase K method
of the Gene Fusion Manual (34). Plasmid DNA preparations. agarose gel
electrophoresis, and E. coli transformation were all performed by conventional
methods (30).

Filter mating targeted mutagenesis. Filter mating between E. coli 817-1 (Apir)
lpTF67A and AP225 was performed according to the protocol of Mulks and Buysse
(19). Briefly. bacterial cultures were grown overnight at 37°C. Equal cell numbers of
donor and recipient cultures, as determined by optical density at 520 nm. were added
to 5 ml 10 mM M9804 and the bacteria pelleted by centrifugation. The pellet
containing the cell mating mixture, resuspended in 100 pl of 10 mM MgSO4, was plated
on a sterile filter on BHIV + riboflavin agar and incubated for 3 h at 37°C. Cells were
washed from the filter in sterile phosphate buffered saline (pH 7.4), centrifuged,
resuspended in 400 pl BHIV broth and plated in 100 pl aliquots on BHIV containing

riboflavin, kanamycin, and nalidixic acid. Kanamycin and nalidixic acid resistant

72

Table 3.1 - Characteristics of Bacterial Strains and Plasmids.

 

 

Strain I Plasmid Characteristics Source/Reference
Strain
E. coli DH5-a supE44, Alch169. (¢80IacZAM15), BRL (USA)
hst17, recA1. endAl, gyrA96, (hi-1, reIA1
E. coli OHS-a (kpir) kpir. supE44, AlacU169, (tSOIacZAM15), Mulks 8 Buysse
hst17. recA1, endAl, gyrA96, thi-1, reIA1
E. coliS17-1 (kpir) 2. plr, recA. thi, pm, hsd, (r-m+), RP4-2, Simon et al.
(T czzMu), (Km::Tn7), [TmpR]. [SmR]
AP100 A. pleurapneumaniae ATCC 27088, ATCC
serotype 1. passaged through pigs
AP106 A. pleurapneumaniae lSU178, a serotype 5 Iowa State University
field isolate, passaged through pigs
AP225 A spontaneous nalidixic acid resistant This work
mutant of AP100
AP233 A double cross-over riboflavin auxotroph of This work
AP225
AP234 A single cross-over riboflavin auxotroph of This work
AP225
Plasmid
pUC19 ApR Rcloning vector Vielra 8. Messing
pUC4K ApR, vector, source of the ken Phannacia (USA)
pGP704 Apfebroad host range suicide vector Miller 8 Mekalanos
pGZRS19 ApR APP-E. cali shuttle vector West at al.
pTF10 AP106 ribGBAH genes cloned into pUC19 Fuller 8 Mulks
pTF66 A 2.9 kb fragment containing AP106 This work
ribBAH in pGP704
pTF67a pTF66 with all of ribA and part of n‘bB This work
deleted and replamd with the ken cassette
from pUC4K
pTF76 5.2 Kb insert from pTF10 cloned into This work

pGZRS1 9

73

colonies were selected from filter mating plates and screened for riboflavin auxotrophy
by replica plating onto TSAV, observing for inability to grow in the absence of added
riboflavin.

Southern Analysis of Transconjugants. Chromosomal DNA and plasmid
controls were digested with the appropriate restriction enzymes and the DNA fragments
were separated on an 0.7% ultrapure agarose gel in TAE buffer. Southern blots were
performed as described by Sambrook et al (30). DNA probes were labeled with
digoxygenin by random priming using the Genius V. 3.0 kit from Boehringer Mannheim.
Probes included the 5.2 Kb insert from pTF10 containing the intact riboflavin operon
from AP106 (Rib), the 1.4 Kb Clal/Ndel fragment deleted from the riboflavin operon in
the construction of pTF67a (R.Del.), the 1.2 Kb kanamycin cassette from pUC4K (Km)
and the intact plasmid pGP704 (pGP704). Hybridization was carried out in 50%
fonnamide at 42°C for 16h. Blots were washed twice in 2 X SSCIO.1% SDS for 15 min
at room temperature, than twice in 0.1x SSCIO.1% SDS for 30 min at 65°C. Blots were
developed with alkaline phosphatase-conjugated anti-digoxygenin and colorimetric
substrate (Boehringer Mannheim) according to the manufacturer's instructions.

Phenotypic analysis of mutant strains. Whole cell Iysates and supematants
of AP100, AP225 (NalR), and AP233 (KmR, NaIR. Rib-) were prepared from ovemight
cultures grown in HIV + 5 mM CaCI2 + appropriate antibiotics. Cells were separated by
microcentrifugation and resuspended in SOS-PAGE sample buffer (16). The culture
supernatant was precipitated with an equal volume of 20% trichloroaceu'c acid 0' CA)
and resuspended in SOS-PAGE sample buffer. Cellular polysaccharides, including
lipopolysaccharide (LPS) and capsular polysaccharide, were prepared according to the
cell lysislproteinase K method of Kimura et al (15). All samples were analyzed on an
0.125% SOS-12% acrylamide gel using a discontinuous buffer system (16). Samples
were transferred to nitrocellulose according to standard protocols (30) and probed with

convalescent serum from a pig infected with A. pleurapneumaniae serotype 1.

74

Antigen-antibody complexes were detected with horseradish peroxidase-conjugated
protein A (Boehringer Mannheim) and the colorimetric substrate 4-chloro-naphthol
(BioRad. Hercules, CA).

Production of serotype-specific capsular polysaccharide was measured by
caagglutination assay using hyperimmune rabbit anti-sera complexed to
Staphylococcus aureus whole cells (11).

Electroporation of A. pleurapneumaniae. AP233 was grown in 100 ml BHIV
with riboflavin at 37°C, with shaking at 150 RPM, to an OD520 of 0.7. Cells were
chilled on ice and centrifuged at 5,000 X g at 4°C for 10 min. Cells were washed twice
in ice cold sterile 15% glycerol. Cells were resuspended in 2 ml 15% glycerol and
frozen in 50 pl aliquots using a dry ice-ethanol bath. Plasmid DNA was added to an
aliquot of competent cells thawed on ice and then transferred to a 0.1 cm gap
electroporation cuvette (BioRad). Cells were electroporated using a Gene Pulsar ll
(BioRad) with the following settings: voltage, 1.8 kV; resistance, 200 a; capacitance,
25 de.

Experimental infections. Eight-week-old specific-pathogen-free castrated
male pigs (Whiteshire Hamroc, lnc.. Albion, IN) were allotted to six challenge groups by
a stratified random sampling procedure, balancing each group for body weight. Each
challenge group was housed in a separate ESL-2 isolation room at the Michigan State
University Research Containment Facility. All experimental protocols for animal
experiments were reviewed by the Michigan State University All University Committee
on Animal Use and Care. and all procedures conformed to university and USDA
regulations and guidelines.

For preparation of challenge inocula, bacteria were grown in 30 ml HIV + 5 mM
CaClz + riboflavin and antibiotics as needed, in 300 ml baffled side-arm flasks, at 37°C
with shaking at 160 RPM, to an OD520 of 0.8. Ten ml of each culture was harvested

by centrifugation at room temperature and washed once with sterile 0.9 % saline. The

75

cell pellet was resuspended in 10 ml of saline and diluted in saline to obtain the desired
cfu/ml. The actual inoculating doses were retrospectively calculated by viable cell
counts on agar plates.

For the challenge procedure. pigs were anesthetized by intravenous injection
with ketamine (4.4 mglkg) and xylazine (1.65 mglkg) and inoculated by percutaneous
intratracheal injection with the appropriate dose of bacteria suspended in 10 ml saline.
Clinical signs of pleurapneumania. including respiration rate, temperature, dyspnea.
appetite and activity/attitude (depression), were monitored and scored as previously
described (12). Seriously ill animals, as determined by severe dyspnea and/or
depression, were euthanized immediately. Survivors were euthanized three days post-
challenge. All animals were necropsied, and lungs were examined macroscopically for
A. pleurapneumaniae lesions, including edema, congestion, hemorrhage, necrosis,
abscessation, fibrosis, and pleuritis. The percentage of lung tissue and pleural surface
area affected was estimated for each of the seven lung lobes. and the total %
pneumonia and % pleuritis calculated using a formula that weights the contribution of
each lung lobe to the total lung volume (12). Representative lung samples were

collected for histopathology and for bacterial culture.

76

RESULTS

Construction of A. pleurapneumaniae rib mutants. To construct riboflavin-
requiring auxotrophic mutants of A. pleurapneumaniae, a suicide plasmid with part of
the riboflavin operon deleted and replaced with a kanamycin-resistance (KmR)
cassette was designed (Figure 3.1). A 2.9 kb EcoRI fragment from pTF10 (5)
containing the A. pleurapneumaniae ribBAH genes was cloned into the EcoRI site of
the conjugative suicide vector pGP704 (18) to create plasmid pTF66. pTF66 was
digested with Clal and Ndel to excise the 3' end of ribB and the entire ribA gene. After
Klenow treatment of the DNA, the 1.2 kb KmR cassette, excised with EcoRI from
pUC4K , was blunt-end ligated into the rib deletion site to create pTF67a.

pTF67a was transformed into E. coli S17-1 (xpir) and mobilized into AP225
(NalR ) to produce >100 transconjugant colonies demonstrating resistance to both
nalidixic acid and kanamycin. Transconjugants were replica plated onto TSAV and
TSAV + riboflavin to assess the requirement for riboflavin and the stability of the
riboflavin auxotrophy. Two classes of transconjugants were found. The majority of the
transconjugants, e.g. AP234. were unstable and produced revertants capable of growth
without supplemental riboflavin in the absence of kanamycin selection. One
transconjugant, AP233. was very stable, maintaining kanamycin resistance as well as
the inability to grow without exogenous riboflavin. All transconjugants were confirmed
as A. pleurapneumaniae by gram stain, colonial morphology, and requirement for V
factor (Ii-NAD).

Southern blot analysis of transconjugants. Two transconjugants were
selected for further analysis based on their phenotypes as potential single (AP234)
and double cross-over mutants (AP233). Southern blot analysis of transconjugant
genomic DNA from the two mutants indicated that AP233 and AP234 were indeed
double and single cross-over insertion mutants respectively (Figure 3.2). Predicted

band sizes for single and double cross-over events are shown in Figure 3.3.

  

gritgrtzgg

regirisgrzgrssgrtzg;

g

5 igrtgrsirgzr

 

git:

EcoRI, Purlfy

Cl ”Ndel, mu-
ltler.iow fill-In Klenow ""4"

 

Izggzstgttgtiigirzt:

Figure 3.1 - Construction of pTF67a.

78

 

Rib
-' <5.2
.. U I <30
m... «2.2
-. «1.3
- «0.7
R.Del
-' <5_2
- <30
pGP704
- <36
Km
- «2.0
-. 41.3

123456

Figure 3.2 - Southern blots of Hindlll or EcoRl Digested DNA from Mutants and
Controls. Blots were prepared in quadnrplicate and hybridized at high stringency with
one of four probes: Rib, the entire ribGBAH operon from pTF10; R. Del., the deleted
portion (Clal/Ndel fragments) of the ribGBAH operon; pGP704, the entire plasmid; Km,
the kanamycin cassette from pUC4K. Lanes: 1, pTF10 digested with Hindlll; 2, AP106
+ Hindlll; 3, AP100 + Hindlll; 4, pTF67a + EcoRl; 5, AP233 + Hindlll; and 6, AP234 +
Hindlll.

79

H‘— 2.2” -—.fl‘—— 30“ ———..||

 

«— 21» —~—1ka -><-‘|,3kh -o-
u H .1

I‘ll“

 

 

 

92.2“ —>‘-1.7kb “1.3": 3.8” >= 3.0”: ——>
ll

x. L-F'g “3' 2.11, ..... Em

Apt um «rink

 

 

 

 

 

 

 

¢——2.2b -—>¢— ”lb :u: 3.6“: ———T-1.7kb *13» ->
_. .___""" _in “"" 'P
M ~a rum «Hurt

 

Figure 3.3 - Schematic Structure of the rib Locus of Parent and Mutant Strains in
Double and Single Cross-Over Events. The predicted sizes of Hindlll genomic
fragments are shown for two possible single cross-over events and for a double cross-
over event. The results show that for AP233 the chromosomal rib operon has been
replaced with the cloned riboflavin operon containing the Km cassette by a double
cross-over event, while AP234 is the result of a single cross-over event either upstream
or downstream of the kanamycin cassette. Restricb'on enzymes used: E = EcoRl; H =

Hindlll.

 

 

80

Genomic DNA from AP233 contained a 2.2 Kb Hindlll fragment that hybridized with the
riboflavin operon (Rib) probe. as well as 1.7 and 1.3 Kb fragments that hybridized with
both the Rib and Km probes; however, there was no reaction with either pGP704 nor
the deleted portion of the riboflavin operon (Figure 3.2). This is the pattern of
hybridization predicted in transconjugants that replaced the wild type riboflavin operon
with the mutated ribzszR locus by a double-crossover event (Figure 3.3). In contrast,
genomic DNA from AP234 shows the presence of DNA homologous to the fragment
deleted from the riboflavin operon (R. del), pGP704, and the kanamycin cassette
(Figure 3.3). This is the pattern of hybridization predicted in transconjugants that
inserted the entire pTF67a plasmid into the wild type rib operon by a single crossover
event (Figure 3.2).

Phenotypic analysis of the A. pleurapneumaniae rib mutant. Whole cell
Iysates, TCA-precipitated culture supematants. and polysaccharide preparations were
analyzed on silver stained SOS-PAGE and on immunoblots developed with
convalescent swine sera. No differences in protein, LPS. extracellular toxin. or
capsular polysaccharide profiles were detected between wild type AP100, its NalR
derivative AP225, and the riboflavin mutant AP233 (Figure 3.4). There was no
difference in reactivity with serotype-specific antisera as determined by caagglutination
assay (data not shown).

Complementation of the rib mutation with a cloned wild type rib operon.
The 5.2 Kb insert from pTF10, containing the wild-type A. pleurapneumaniae riboflavin
operon, was cloned into pGZRS19, an E. cali-A. pleurapneumaniae shuttle vector (41),
to form pTF76. pTF76 was transformed into AP233 by electroporation. restoring the
ability of AP233 to grow in the absence of exogenous riboflavin and restoring the

virulence of the mutant (Table 3.2 and Table 3.3).

81

123456789123456789

5--

 

66.2— F 53?: -
i: :2- —--fl
42.7- 93 3g" .-
e- "' "'..._
.22
eta-ff? 99.—'3’
2t5-
14.4- ”5'
.-.
-
..
A B

Figure 3.4 - Phenotypic Analysis of AP100, AP225 and AP233.

Panel A: All samples were boiled for 5 minutes before SDS-PAGE on a 12% gel.
Panel B: Westem blot analysis was performed using convalescent a-APP-1 swine
serum and detected colorimetrically with a protein A-HRP system.

Lanes 1-3 are whole cell Iysates of AP100, AP225 and AP233.
Lanes 4-6 are TCA precipitated supematants of AP100, AP225 and AP233
Lanes 7-9 are LPS preparations of AP100, AP225 and AP233.

82

Attenuation of virulence of the rib mutant in swine. Six groups of three pigs
each were infected with: group 1, 1 L050 (5 X 106 cfu) of AP225; groups 2-5, AP233
at doses equivalent to 4. 20, 100, and 500 times the wild-type LDso; and group 6,
AP233/pTF76 at a dose equivalent to 1 L050 for the wild-type. Mortality. lung score,
and clinical score data. shown in Tables 3.2 and 3.3, all indicate that the riboflavin
auxotroph is avinrlent in pigs at doses as high as 500 times the wild-type L050. The
pigs infected with the rib mutant AP233 displayed no dyspnea. elevated respiration
rate, depression, or loss of appetite, and had no typical pleurapneumania pamology at
necropsy at even the highest dose tested. In contrast. 1 of 3 pigs infected with the wild-
type AP225 strain died, and all three exhibited significant clinical signs of infection,
including elevated respiration rates, dyspnea, depression, loss of appetite, and fever,
and severe pneumonia and pleuritis was evident at necropsy. Pigs infected with
AP233 containing the riboflavin genes in trans (pTF76) also exhibited obvious clinical
signs and significant pneumonia and pleuritis, although somewhat less severe than the
wild-type strain. These results indicate that restoration of the ability to synthesize
riboflavin does restore virulence.

Bacteria were readily reisolated at necropsy from the lungs of pigs receiving
AP225 and AP233 IpTF76. All reisolated organisms were characterized by gram stain.
colonial morphology, requirement for V factor (ti-NAD). antibiotic sensitivity, and
serotyping by caagglutination. Reisolated organisms showed no differences from the
initial inocula, including the presence of plasmid pTF76 in bacteria reisolated from pigs
infected with AP233/pTF76. In contrast, we were unable to recover organisms from
the lungs of animals infected with AP233 and euthanized 48 hours post infection.

83

Table 3.2 - Mortality and Lung Score Data.

 

 

 

Group Strain Dose W. Mortality % Pneumonia” % Pleuritis°
1 AP225 (WT) 1 1/3 66.7 71.7
2 AP233 (Rib-) 4 0/3 0 0
3 AP233 (Rib-) 20 on 0 0
4 AP233 (Rib-) 100 0/3 0 0
5 AP233 (Rib-) 500 on 0 0
6 AP233+ pTF76 1 0/3 27.6 20.2

 

' Doses are multiples of the established wild-type L050 of 5.0 x 10° cfu (12).
% lung tissue exhibiting A. pleurapneumaniae lesions
° % pleural surface area exhibiting pleuritis

C

Table 3.3 - Clinical Score Data

 

Dose T63?
Group Strain 0339' RR Max” M Oyspnead Depression' Appetite'

 

 

1 AP225 1 20 105.7 5.5 6.7 4.2
2 AP233 4 8 102.5 0 0 0
3 AP233 20 8 103.3 0 0 0
4 AP233 100 8 103.5 0 0 0
5 AP233 500 8 102.8 0 0 0
6 AP233 + 1 19.3 105.4 4.5 4.7 3.7

pTF76

Normal 8.0 <103.0 0 0 0

Maximum 25 15 15 5

 

Doses are multiples of the established wild-type L050 of 5.0 x 106 cfu (12).
Maximum respiratory rate observed after challenge. Respiratory rate recorded as number of breaths
per 15 sec observation period.

c Maximum rectal temperature after challenge, in degrees Fahrenheit.

‘

Oyspnea score measures degree of respiratory distress and labored breathing. Scored as 0 = normal;
= slight; 2 = moderate; 3 = severe. Total score = sum of scores taken at 12 hour intervals
after challenge.
Depression scare evaluates attitude and activity. Scored as 0 = normal; 1= slight inactivity; 2 =
moderate; 3 = severe. Total score = sum of scores taken at 12 hour intervals after challenge.
Appetite was scored as 0 = did eat; 1 = did not eat. Total score = number of 12 hour periods not
eating over 60 hour observation period.

84

DISCUSSION

In this paper. we report the construction of a serotype 1 Actinobacillus
pleurapneumaniae deletion-dismption riboflavin mutant that is attenuated in vivo. The
A. pleurapneumaniae ribGBAH operon was disrupted by deleting an internal segment
of the operon (ribBA) and replacing it with a KmR cassette using a targeted
mutagenesis technique (19). A stable riboflavin-requiring. KmR mutant, AP233, was
phenotypically identical to its wild-type parent based on analysis of proteins.
extracellular toxin, LPS. and capsular polysaccharide by SDS-PAGE, immunoblot, and
caagglutination.

A riboflavin mutant of A. pleurapneumaniae serotype 5 was also constructed
and was also found to be attenuated in a preliminary animal challenge experiment
However, further studies were conducted in serotype 1 because serotype 5 seemed to
be very resistant to transformation by standard heat shock or electroporation
procedures. In order to complement the rib mutation in trans, and for ease of future
genetic manipulations, it was necessary to use a serotype 1 strain for these studies.

Experimental infection of pigs, the only natural host for A. pleurapneumaniae,
demonstrated that the riboflavin-requiring mutant was unable to cause disease at
dosages as high as 500 times the LD5o for the wild-type parent. In the four groups of
pigs infected with AP233 by intratracheal inoculation. there was no mortality, no
significant clinical signs were observed, and no typical pleurapneumanic lesions were
discovered upon necropsy. Complementation of AP233 in trans with the wild-type A.
pleurapneumaniae riboflavin operon restored both the ability to grow without
exogenous riboflavin and virulence, demonstrating that the riboflavin mutation itself is
responsible for the attenuation in viva.

It is important to note that the riboflavin-requiring mutant used in these studies is
a deletion mutant, with ~1.4 Kb of the riboflavin operon removed from the chromosome

and replaced with an antibiotic resistance marker. We observed neither reversion to

85

prototrophy nor loss of kanamycin resistance in this mutant in the laboratory. In the
preliminary experiment with a serotype 5 riboflavin mutant, we were able to reisolate
the mutant from the pig lungs at 16 hours post-infection. All colonies isolated in this
experiment were kanamycin-resistant, nalidixic acid-resistant, and riboflavin requiring,
suggesting that reversion to prototrophy and thus virulence will not occur in vivo.

In the dosage trial experiments, AP233 was not recovered from the lungs of
infected swine at 48 hours post-infection. These results may indicate poor persistence
of the organism in viva, a potential problem for its use as a live-attenuated vaccine. If
necessary, sufficient exogenous riboflavin could be added to the vaccine to allow the
organism to replicate minimally and therefore persist long enough to induce a protective
immune response.

This is the first report that a mutation in riboflavin biosynthesis in a pathogenic
bacterium is attenuating. This finding represents a new addition to the group of
biosynthetic mutations which can be used to construct attenuated strains of bacteria.
This is also the first report of a genetically defined attenuated mutant of A.
pleurapneumaniae that is still capable of production of all of the major known virulence
factors of this organism. including extracellular toxins and capsular polysaccharide.
We intend to continue studies on the use of riboflavin mutants of A. pleurapneumaniae
as candidates for a new generation of live attenuated vaccines against this disease.
In addition. the fact that riboflavin biosynthesis is essential for this pathogen and is not
synthesized by higher eukaryotes could potentially lead to discovery of a new

generation of antibiotics which inhibit riboflavin biosynthesis.

ACKNOWLEDGMENTS
This research was supported by grants from the State of Michigan Research
Excellence Fund Center for Animal Production Enhancement and from the Michigan
State University Office of Intellectual Property. T.E. Fuller is supported by a U.S.D.A.
National Needs Graduate Fellowship in Animal Biotechnology.

10.

11.

12.

87

REFERENCES

Ahmed, Z.U., M.R. Sarker, and D.A. Sack. 1990. Protection of adult rabbits
and monkeys from lethal shigellosis by oral immunization with a thymine-
requiring and temperature-sensitive mutant of Shigella flexneri Y. Vaccine.
8:153-158.

Bacher, A. 1991. Biosynthesis of flavins. p. 215-59. In F. Muller (ed.).
Chemistry and Biochemistry of Flavins, Vol. 1. Chemical Rubber Company,
Boca Raton, Fla.

Collins, L.V., S. Atti'idge, and J. Hackett. 1991. Mutations at rfc or pmi
attenuate Salmonella typhimurium virulence for mice. Infect. Immun. 5921079-
1085.

Fenwick, B. and S. Henry. 1994. Porcine pleurapneumania. J. Am. Vet Med.
Assoc. 204:1334-1340.

Fuller, T.E. and M.H. Mulks. 1995. Characterization of Actinobacillus
pleurapneumaniae riboflavin biosynthesis genes. J. Bacteriol. 177:7265-7270.

Hoiseth S. K. and B.A.D. Stacker. 1981. Aromatic-dependent Salmonella
typhimurium are non-virulent and effective as live vaccines. Nature (Iondon).
291 :238-239.

Homchampa, P., R. A. Strugnell and B. Adler. 1992. Molecular analysis of
the aroA gene of Pasteurella multocida and vaccine potential of a constructed
araA mutant Mol. Microbiol. 623585-3593.

I-Iomchampa, P., R. A. Strugnell and B. Adler. 1994. Construction and
vaccine potential of an aroA mutant of Pasteurella heemolyb‘ca. Vet Microbiol.
42:35-44.

Hunneman, W.A. 1986. Incidence, economic effects, and control of
Haemaphilus pleurapneumaniae infections in pigs. Vet Quarterly 8:83-87.

Inzana, T.J., J. Todd, and l-I.P. Veit. 1993. Safety. stability and efficacy of
nonencapsulated mutants of Actinobacillus pleurapneumaniae for use in live
vaccines. Infect Immun. 61:1682-1686.

Jolie, R. A. V.. M. H. Mulks, and B. J. Thacker. 1994. Antigenic differences
within Actinobacillus pleurapneumaniae serotype 1. Vet Microbiol. 38:329-349.

Jolie, R. A. V.. M. H. Mulks, and B. J. Thacker. 1995. Cross-protection
experiments in pigs vaccinated with Actinobacillus pleurapneumaniae subtypes
1A and 1B. Vet Microbiol. 45:383-391.

13.

14.

15.

16.

17.

18.

19.

20.

21.

23.

24.

25.

Kamell, A., P.D. Cam, N. Verrna and A.A. Lindberg. 1993. Arab deleteion
attenuates Shigella flexneri strain 2457T and makes it a safe and efficacious
oral vaccine in monkeys. Vaccine 8:830-836.

Kelly, S.M., B.A. Bosecker and R. Curtiss Ill. 1992. Characterization and
protective properties of attenuated mutants of Salmonella cholerasuis. Infect.
Immun. 60:4881-4890.

Kimura, A. and E.J. Hansen. 1986. Antigenic and phenotypic variations of
Haemophilus influenzae type B lipopolysaccharide and their relationship to
virulence. Infect Immun. 51 :69-79.

Laemmli, UK 1970. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature 227:680-685.

Lindberg. A.A., A. Kamell, B.A.D. Stacker, S. Katakura, H. Sweiha and ER
Reinholt. 1988. Development of an auxotrophic oral live Shigella flexneri
vaccine. Vaccine 6:146-150.

Miller, V.L. and J.J. Mekalanos. 1988. A novel suicide vector and its use in
construction of insertion mutants: osmoregulation of outer membrane proteins
and virulence determinants in Vibrio cholerae requires toxR. J. Bacteriol.
170:2575-2583.

Mulks, M.H. and J.M. Buysse. 1995. A targeted mutagenesis system for
Actinobacillus pleurapneumaniae. Gene 165:61-66.

Nicolet, J. 1992. Actinobacillus pleurapneumaniae, p. 401-408. In A.D.
Laman et al (eds), Diseases of Swine. 781 edition Iowa State University Press,
Ames, Iowa.

Nielsen, R. 1979. Haemaphilus paraheemolyticus serotypes: pathogenicity
and cross immunity. Nord. Vet Med. 31:407-413.

Nielsen, R. 1984. Haemaphilus pleurapneumaniae serotypes - Cross
protection experiments. Nord. Vet. Med. 36:221-234.

Nielsen, R. 1976. Pleuropneumonia of swine caused by Haemaphilus
pleurapneumaniae. Studies on the protect'on obtained by vaccinafion. Nord.
Vet. Med. 28:337-338.

Nielsen, R. 1974. Serological and immunological studies of pleurapneumania
of swine caused by Haemaphilus peraheemalyticus. Acta Vet Scand. 15:80-
89.

Nnalue, N.A., and B.A.D. Stacker. 1987. Tests of the virulence and live-
vaccine efficacy of auxotrophic and gaIE derivatives of Salmonella cholerasuis.
Infect. Immun. 55:955-962.

26.

27.

28.

29.

30.

31.

32.

33.

35.

36.

89

O'Callaghan, D. D. Maskell, F.Y. Lieu, C.S.F. Easmon and G. Dougan. 1988.
Characterization of aromatic and purine dependent Salmonella typhimurium:
attenuation, persistence and ability to induce protective immunity in BALB/c
mice. Infect Immun. 56:419-423.

Paltlneanu. D., R. Pambucol, E. Tirziu, and l. Scobercea. 1992. Swine
infectious pleurapneumania: Aerosol vaccination with a live attenuated vaccine.
Proc. Int. Pig. Vet Soc. 12:214.

Perry, M.B., E. Altman, J.-R. Brisson, L.M. Beynon, and J.C. Richards.
1990. Structural characteristics of the antigenic capsular polysaccharides and
lipopolysaccharides involved in the serological classification of Actinobacillus
pleurapneumaniae strains. Serodiag. Immunother. Infect Dis. 4:299-308.

Rosendal. S., 0.8. Carpenter, W.R. Mitchell, and MR. Wilson. 1981.
Vaccination against pleurapneumania in pigs caused by Haemaphilus
pleurapneumaniae. Can. Vet. J. 22:34-35.

Sambrook, J.. E.F. Fritsch and T. Maniatis. 1989. Molecular Cloning: A
Laboratory Manual, 2nd. ed. Cold Spring Harbor Press, Cold Spring Harbor.
New York.

Schott, K., J. Kellerman, F. Lottspeich and A. Bacher. 1990. Riboflavin
syntheses of Bacillus subtilis: purification and amino acid sequence of the a-
subunit. J. BioI.Chem. 265:4204-4209.

Sebunya, T.N.K. and JR. Saunders. 1983. Haemaphilus pleurapneumaniae
infection in swine: a review. J. Am. Vet. Med. Assoc. 182:1331-1337.

Sigwart, D.F., B.A.D. Stacker, and JD. Clements. 1989. Effect of a purA
mutation on the efficacy of Salmonella live vaccine vectors. Infect. Immun.
57:1858-1861.

Silhavy, T.J. 1984. DNA extraction from bacterial cells. p. 137-139. In
Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring
Harbor. NY.

Simon, R., U. Priefer, and A. Puhler. 1983. A broad host range mobilization -
system for in viva genetic engineering: transposon mutagenesis in Gram
negative bacteria. Bio/Technol. 1:784-791.

Tacket, C.l., D.M. Hone, R. Curtiss III, S.M. Kelly, G. Losonsky, L. Guers.
A.M. Harris, R. Edalman. M.M. Levine. 1992. Comparison of the safety and
immunogenicity of AaroC AaroD and AcyeAcrp Salmonella typhi strains in adult
volunteers. 60:536-541.

37.

38.

39.

40.

41.

90

Thacker, B.J., and M.H. Mulks. 1988. Evaluation of commercial Haemaphilus
pleurapneumaniae vaccines. Proc. Int. Pig Vet. Soc. 10:87.

Utrera, V.. C. Pijoan. and T. Molitor. 1992. Evaluation of the immunity
induced in pigs after infection with a low virulence strain of A.
pleurapneumaniae serotype 1. Proc. Int. Pig Vet. Soc. 12:213.

Vaughan, L.M., P.R. Smith, and T.J. Foster. 1993. An aromatic-dependent
mutant of the fish pathogen Aeromonas salmonicide is attenuated in fish and is
effective as a live vaccine against the Salmonid disease furunculosis. Infect.
Immun. 61:2172-2181.

Vieira, J. and J. Messing. 1982. The pUC plasmids, an M13mp7-derived
system for insertion mutagenesis and sequencing with synthetic universal
primers. Gene192259-268.

West, S.E.H., M.J.M. Romero, L.B. Regassa. N.A. Zielinski, and RA Welch.
1995. Construction of Actinobacillus pleurapneumaniae-Escherichia cali shuttle
vectors: expression of antibiotic resistance genes. Gene 160: 81-86.

Chapter 4

A RIBOFLAVIN AUXOTROPH OF ACTINOBA CILLUS PLEUROPNEUMONIAE AS A
LNE-ATTENUATED VACCINE

91

92

ABSTRACT

Actinobacillus pleurapneumaniae (APP) is a gram negative pleiamarphic rod
that is the causative agent of a severe, highly infectious and often fatal
pleurapneumania in swine. We have previously reported the construction of a
genetically defined stable riboflavin auxotroph (AP233) by replacing a portion of the
APP serotype 1 strain (ATCC 27088) riboflavin biosynthetic operon (ribGBAH) with an
antibiotic cassette encoding resistance to kanamycin. The riboflavin mutant was
previously shown to be avinilent at a dosage equivalent to 500 times the established
L050 for the wild type parent. We have tested AP233 in two studies for its ability to
stimulate protective immunity against pleurapneumania. In the first study, four groups of
six pigs each were inoculated with 5 x 108 cfu of AP233 either by intratracheal or
intramuscular injection, with or without the addition of 5 jig/ml of riboflavin to the live
vaccine inoculum. Two additional groups of six pigs each were either vaccinated with a
serotype 1A bacterin or kept as unvaccinated controls. The pigs received a second
vaccination at three weeks followed by experimental challenge with 1 L050 of serotype
1A two weeks later. Results of the challenge experiment indicate that intramuscular
vaccination with the live attenuated rib mutant. in a vaccine formulation that included a
limiting amount of exogenous riboflavin, provided significant protection against
homologous challenge with virulent A. pleurapneumaniae. In a second study, two
groups of six pigs each were vaccinated twice with 5 x 109 cfu of AP233 in P88 + 5 p
glml riboflavin by intramuscular injection. Two additional groups of six pigs each were
sham vaccinated with P88 + 5 [Ag/MI riboflavin. Challenge with both a homologous
(serotype 1) strain and a heterologous (serotype 5) strain demonstrated limited
protection against pleurapneumania. It is inconclusive at this time as to the degree of

protection obtained by vaccination with AP233.

93

INTRODUCTION

Actinobacillus pleurapneumaniae (APP) is the causative agent of porcine
pleurapneumania. a severe and often fatal respiratory disease of swine (5.24.37) first
described over 30 years ago (38). While our understanding of the pathogenesis of
APP infections and the virulence factors involved has increased dramatically over the
past decade, there is still no safe effective vaccine against APP that provides
protection against all the major serotypes (26.27.33.40). This is in part due to the fact
that there are twelve antigenically distinct capsular serotypes (31) as well as antigenic
subtypes of some serotypes (17, 25), and vaccination with a killed whole cell vaccine
prepared from one serotype does not generally confer protection against other
serotypes (26.27.33.40), or even subtypes of the same serotype (16). Current
commercial vaccines are still primarily killed whole cell bacterins. which generally
reduce mortality from APP infection but frequently fail to prevent severe morbidity and
economic loss due to chronic effects of the disease on growth rate and feed efficiency
(5,13).

In contrast to bacterin vaccines, natural or experimental infection with a virulent
strain of APP generally elicits protection against reinfection with any serotype
(26.27.33.40), implying that there are protective antigens or immunomodulatory
compounds produced by the organism in viva which are not produced in vitro. This
suggests that a live avirulent vaccine. with limited replication ability, would have the
potential to produce these factors and elicit broad cross-protective immunity against all
serotypes of APP. Avinilent strains of APP, including capsule-deficient strains and
other strains for which the attenuating lesion is completely undefined. have been
tested as live vaccines and have elicited cross-protective immunity against subsequent
challenge (14.30.34.42). However, the use of live vaccines in the field is problematic,
particularly when the attenuating lesions in the vaccine strain are not genetically

defined. The development of attenuated strains of APP with genetically defined

94

biochemical mutations that limit growth in vivo and prevent reversion to wild type for
use as live avirulent vaccines has. until recently, been impossible due to the lack of
genetic tools for manipulation of APP chromosomal genes. We have developed a
targeted mutagenesis system for APP (23) and are now using this methodology to
constructgenetically defined deletion mutants of APP.

Many live-attenuated strains of organisms such as Shigella fiexnen‘ (1,
18.20.28.35), Salmonella typhi (8), Salmonella typhimurium (4,10), Salmonella
choleraesuis (19). Aeramanas salmonicide (44), Bacillus anthracis (15), Bordeteila
pertussis (32), Yersinia enterocolitica (3). Pasteurella heemolytica (11) and Pasteurella
multocida (12) provide moderate-to-excallent protection against infection, varying with
the attenuating lesion and the dosage regimen. Many of these live-attenuated vaccine
strains have been constructed by disnipting biochemical pathways or critical virulence
genes. For example. live attenuated vaccines have been constructed using mutations
in em (3.10.11.12,15,18,20,32,44), icsA (35), iuc (35), guaB-A-virG (28). cya-crp (19),
thy (1), rfc, pmi (4) and gel (8), which respectively encode aromatic amino acid
biosynthetic enzymes. intracellular spreading factors. siderophores. guanine nucleotide
biosynthetic enzymes, adenylate cyclase and cAMP, thymine biosynthetic enzymes,
and enzymes involved in lipopolysaccharide biosynthesis. Live vaccines appear to hold
much promise for the future of vaccination, due to the ability to vaccinate with a small
amount of organism, the production and more natural presentation of protective
antigens, and the ability to stimulate the production of cytokines influencing the immune
system (21).

The goal of this research was to test a genetically defined attenuated
biochemical mutant of APP as a live-attenuated vaccine (LAV). Previously we reported
the cloning and characterization of the operon encoding riboflavin biosynthesis from
APP. Riboflavin (vitamin 82), a precursor of the coenzymes flavin adenine dinucleotide

(FAD) and flavin mononucleotide (FMN). is essential for basic metabolism. It is

95

synthesized by plants and by most microorganisms but not by higher animals (2).
Therefore. exogenous riboflavin is not likely to be freely available on the mucosal
surfaces of the respiratory tract. In addition. many pathogenic bacteria are apparently
unable to utilize flavins from their environment and are entirely dependent an
endogenous production of riboflavin (36). We have constructed genetically defined
mutants of APP that lack part of the riboflavin biosynthetic operon (6) and require
exogenous riboflavin for growth. We have confirrnad that riboflavin biosynthesis is
essential for survival of APP in vivo and that mutations in the riboflavin biosynthetic
pathway are highly attenuating (7).

In this study we have conducted two vaccine trials to test the potential efficacy
of this riboflavin-requiring mutant of APP as a live-attenuated vaccine. The results
presented here demonstrate that intramuscular vaccination with the live riboflavin
auxotroph can elicit some protective immunity against challenge with two different

serotypes of vinilent wild type A. pleurapneumaniae.

96

MATERIALS AND METHODS

Animals. Eight-week-old crossbred (Yorlrshire/Landrace) castrated male pigs
from a herd known to be free of A. pleurapneumaniae and related respiratory
pathogens (Whiteshire Hamroc, lnc.. Albion. IN) were allotted to challenge groups by a
stratified random sampling procedure, balancing each group for body weight. Each
challenge group was housed in a separate BSL-2 isolation room at the Michigan State
University Research Containment Facility and fed a standard antibiotic-free diet
provided by the Michigan State University Swine Research and Teaching Center. All
experimental protocols for animal experiments were reviewed by the Michigan State
University All University Committee on Animal Use and Care. and all procedures
conformed to university and USDA regulations and guidelines.

Bacterial Strains. A. pleuropneumoniae ATCC 27088 (APP-1A), reisolated
from experimentally infected pigs, was used as the serotype 1 challenge strain. A.
pleurapneumaniae lSU178 (APP-5). also reisolated from infected pigs, was used as the
serotype 5 challenge strain. AP233 (7), a Km', Nal'. riboflavin-requiring mutant of APP-
1A, was used as the live attenuated vaccine strain.

MediaICulture Conditions. All bacteria were grown in heart infusion broth (HI)
(Difco, Detroit. MI) with 10 pglml B-nicotinamide adenine dinucleotide (V factor) (Sigma.
St. Louis, MO). Riboflavin (Sigma) was added at 200 uglml for growth of strain AP233.
Kanamycin was used at 100 jig/ml and nalidixic acid was used at 50 jig/ml in agar
plates.

Preparation of Vaccines. AP233 for the live vaccine was grown in 30 ml HIV
broth + 5 mM CaClz + 200 jig/ml riboflavin, in 300 ml baffled side-am flasks at 37°C
with shaking at 160 RPM. to an optical density at 520 nm of 0.8. Bacteria were
harvested by centrifugation. washed once in phosphate buffered saline (PBS). pH 7.0,
diluted in PBS to the appropriate cell density, and used immediately as vaccine. For

vaccine groups containing riboflavin, 5 jig/ml riboflavin was added to the PBS (PBS-

97

Rib) used for washing and resuspension. The actual inoculating doses were
retrospectively calculated by viable cell counts on agar plates.

To prepare the fonnalinized whole cell bacterin, virulent APP-1A bacteria were
similarly grown in HIV broth + 5 mM CaClz to an optical density at 520 nm of 0.8.
Bacteria were harvested by centrifugation and washed once with Tris-acetate-EDTA-
DT'I' buffer. Bacteria were resuspended in buffer containing 0.2% formalin to a
concentration of 5 X 10° cfulml, and kept at room temperature for 1 hour. then stored at
4°C. Each vaccine dose contained 1 ml fonnalinized cells, 0.5 ml saline. and 0.5 ml
Emulsigen adjuvant (MVP Laboratories, Ralstan, Nebraska).

Vaccination. In trial one. six groups (six pigs per group) of six- to eight-week-
old APP-free pigs were vaccinated twice with a 3 week interval. Group 1 was
vaccinated intratracheally (IT) with 5 X 108 cfu (100 X the established LDso for wild type
APP-1A) of live AP233 in 10 ml of sterile PBS by percutaneous transtracheal
inoculation, after anesthesia with xylazine (1.65 mglkg) and telazol (4.4 mglkg)
administered intramuscularly. Group 3 received 5 X 10'3 cfu of live AP233
intramuscularly (IM) in 2 ml PBS. Groups 2 and 4 received the same treatment as
Groups 1 and 3, respectively, except that the bacteria were suspended in PBS
containing 5 uglml riboflavin. Group 5 received a fonnalinized whole cell bacterin
prepared from APP-1A containing the equivalent of 5 X 109 cfu per dose in 2 ml of 25%
Emulsigen adjuvant (MVP Laboratories, Ralstan, Nebraska). Group 6 was
unvaccinated controls.

In trial two, four groups (six pigs per group) of six- to eight- week- old APP-free
pigs were vaccinated twice with a 3 week interval. Groups 1 and 2 were vaccinated
with 2 X 109 cfu (400 X the wild type L050) of live AP233 intramuscularly in 2 ml PBS-
Rib. Groups 3 and 4 were sham vaccinated with 2 ml PBS-Rib.

Pigs were monitored for 24 hours post-vaccination for any clinical signs of APP

infection, as described below.

98

Experimental Challenge. Two weeks after the second vaccination, all pigs
were challenged with either 1 LDso (5 X 106 cfu) of virulent wild type APP-1A or 1 LDso
(2 X 107 cfu) of virulent wild type APP-5A, by percutaneous intratracheal inoculation
(41). For preparation of the challenge inoculum. bacteria were grown to an 00520 of
0.8 in HIV broth containing 5 mM CaClz. washed once in sterile saline, and diluted in
saline to the appropriate cell density. Pigs were anesthetized by intravenous injection
with ketamine (4.4 mglkg) and xylazine (1.65 mglkg) and inoculated by percutaneous
intratracheal injection with the appropriate dose of bacteria suspended in 10 ml saline.
Clinical signs of pleurapneumania. including increased respiration rate. elevated rectal
temperature, dyspnea, decreased appetite and activity/attitude (depression), were
monitored and scored as previously described (16). Seriously ill animals, as
determined by severe dyspnea and/or depression, were euthanized immediately.
Survivors were euthanized three days post-challenge in trial 1 and four days post-
challenge in trial 2. All animals were necropsied. and lungs were examined
macroscopically for A. pleurapneumaniae lesions, including edema, congestion,
hemorrhage, necrosis. abscess, fibrosis, and pleuritis. The percentage of lung tissue
and pleural surface area affected was estimated for each of the seven lung lobes, and
the total % pneumonia and % pleuritis calculated using a formula that weights the
contribution of each lung lobe to the total lung volume (16). Representative lung
samples were collected for histopathology and for bacterial culture. The neck muscles
of all animals were also examined for any macroscopic signs of injection site reactions
such as granuloma formation. Protection against challenge was measured as a
reduction in mortality, in the severity of lung lesions, and in the severity and duration of

clinical signs as compared to the unvaccinated control animals.

Seralagic analysis of immune responses. Sera were collected from all pigs

on days 0. 21, and 35, Le. prior to the first vaccination, at the second vaccination, and

99

at challenge. Antibody titers against APP were quantitated by indirect ELISA against

APP outer membranes (16), hemolysin neutralization (22), and complement fixation

(9)-

Statistical analysis. Statistical analysis of the data was conducted using the
Statistix microcomputer program (Analytical Software, St. Paul, MN) for analysis of
variance (ANOVA) and Epistat (T .L. Gustafson. Round Rock, TX) for nonparametric

analyses.

100
RESULTS

Addition of riboflavin to the vaccine inoculum. In preliminary studies, it was
found that riboflavin-requiring strains of APP failed to persist in the porcine respiratory
tract for more than 16-24 hours. To permit expression of infection-associated antigens
by the live attenuated vaccine strain of bacteria after immunization of pigs, we needed
to ensure that the bacteria had sufficient available riboflavin to permit 2-3 generations
of growth in vivo. We determined that addition of 510 pg of riboflavin per ml of the
vaccine inoculum was sufficient to permit this amount of growth, and no more, in vitro.
Therefore, as part of the first vaccine trial, we compared intratracheal (IT) and
intramuscular (IM) administration of the live attenuated vaccine, with and without the
addition of 5 jig/ml exogenous riboflavin.

Safety. Pigs were monitored post-vaccination for any clinical signs of APP
disease, such as fever. dyspnea, and increased respiratory rate, and for injection site
reactions in animals that received IM vaccines. The intramuscular administration of
AP233 as a live vaccine caused no significant clinical signs other than slight depression
and decreased appetite for less than eight hours post vaccination and caused no
injection site reaction. In contrast, bacterin-vaccinated animals showed mild fever.
depression and decrease in appetite for eight to sixteen hours post vaccination. and
injection site granulomas in several cases. lntratracheal immunization with the LAV did
cause slight clinical signs. including increased respiratory rates, fever, decreased
appetite, and mild depression for 8-16 hours post-immunization. These results
demonstrate that the live intramuscular vaccine is at least as safe as, if not safer than,
a fonnalinized whole cell bacterin of the type currently used commercially.

lmmunogenicity. The immune responses of the pigs to vaccination in the first
vaccine trial, were evaluated by ELISA against APP outer membranes (16), hemolysin
neutralization titer (22). and complement fixation (9)(T able 4.1). At challenge, the

bacterin-vaccinated animals showed significant ELISA and complement fixation titers,

101

but low or negative hemolysin neutralization titers. The four groups receiving live
vaccines showed low or negative ELISA and CF titers. However. the animals that were
vaccinated intramusculariy did show significant hemolysin neutralization titers,
indicating that hemolysin was expressed in vivo by the LAV bacteria.

Protection against experimental challenge. In the first vaccine trial, all
vaccine groups were challenged with virulent APP-1A, the homologous serotype to the
vaccine strain. The mortality and clinical data from this homologous challenge are
reported in Table 4.2 and Table 4.3. APP was cultured from the lungs of all the pigs
except for 1 animal in Group 4. All cultures were confirmed as APP-1A by gram stain,
requirement for NAD, and caagglutination. In this experiment, the LAV prepared with
exogenous riboflavin and delivered intramuscularly (Group 4) provided complete
protection against mortality (0/5 animals died) and significant reductions in lung
damage and some clinical signs of pleurapneumania. In contrast, 6/6 unvaccinated
control animals died from overwhelming pleurapneumania as a result of this
experimental challenge. Other LAV formulations, as well as the fonnalinized bacterin,
did not afford significant protection.

Note that the experimental challenge in this case was overwhelming. Normally,
in our hands. the killed whole cell bacterin provides 100% protection against mortality
and some reduction in morbidity; whereas in this experiment, half (3/6) of the bacterin-
vaccinated animals died. In addition, all (6/6) unvaccinated control animals died,
indicating that the dose administered was much higher than the intended wild type
LDso. However. despite the enormous challenge, the group that received the LAV IM
with riboflavin added had significant reductions in mortality and in % pneumonia as
determined by Least Significant Difference (LSD) analysis (p<0.05). No other groups
were significantly different.

The second vaccine trial was conducted to determine whether intramuscular

administration of the LAV, with limiting amounts of riboflavin added to the inoculum,

102

could elicit cross-protective immunity against challenge with a heterologous serotype of
APP (APP-5). In this experiment, the concentration of bacteria in the vaccine dosage
was increased four-fold from the dosage previously used; however, the amount of
riboflavin in the dosage was kept constant at 5 jig/ml. Mortality, lung lesion scores.
and clinical scores from trial 2 are shown in Table 4.4 and Table 4.5. It is clear that. as
in trial 1, the challenge doses were still overwhelming. even though flie challenge dose
of APP-1A was decreased to half that administered in trial 1. All sham vaccinated
controls. whether challenged with APP-1A or APP-5, died from severe
pleurapneumania in less than 48 hours. The vaccinated group challenged with the
homologous serotype (Group 1) did show a reduction in mortality and a significant
reduction in clinical signs of APP infection, but no statistically significant reduction in
lung pathology. The vaccinated group challenged with the heterologous serotype
(Group 3) showed slightly, but not significantly, reduced mortality, death time, and
clinical scores. APP of the appropriate challenge serotype was cultured from the lungs
of all experimentally infected pigs.

103

Table 4.1 - Seralagic Analysis of Samples Collected at Challenge for Trial 1.

 

 

 

 

 

 

 

 

Group Vaccine HNT‘ ELISA-APP12 or (L 2 3
1 Live, IT, PBS 3129 :l: 14785 227 i 90‘ 1.7 :1: 2.8
2 Live, IT, PBS + riboflavin 2520 :l: 741“ 164 i 73” 1.6 a 3.1”
3 Live, IM, PBS 10760 a 6245' 120 :1; 32” 0.0 :I: 0.0”
4 Live, IM, PBS + riboflavin 6293 :I: 2662"-b 236 a; 173” 2.0 :l: 4.0”
5 APP-1A bacterin 3035 :I: 265” 1119 a 170‘ 24.3 a; 7.4'
6 Unvaccinated control 2240 :1: 243° 67 :l: 21° 0.0 :l: 0.0ID

 

1 Hemolysin neutralization titer, <3000 = negative; 3000-6000 = suspect; >6000 = positive. Assays were
performed in the laboratory of Dr. Brad Fenwick, Kansas State University.

2 ELISA vs APP-1 outer membranes; <200 = negative, 200-300 = suspect, >300 = positive. Assays were
performed in the laboratory of Dr. Martha H. Mulks, Michigan State University.

3 Complement fixation test; 0 = negative; >0 = positive; data presented as geometric mean titer. Assays
were performed at the Veterinary Diagnostic Laboratory, Iowa State University.

104

Table 4.2 - Mortality and Lung Score Data for Trial 1.

 

 

 

 

 

 

 

 

Group Vaccine1 Mortality % Pneumonia2 % Pleuritis3
1 Live. IT. PBS 3/5 56.6 :l: 235““ 73.3 a: 39.3I
2 Live, IT, PBS + riboflavin 6/6 63.2 a 6.2“ 66.7 a 51 6I
3 Live, IM, PBS 4/6 57.7 i 23.2” 73.3 a 42.5'I
4 Live, IM, PBS + riboflavin 0/5 24.5 a: 150° 21.5 a: 20.7'
5 APP-1A bacterin 3/6 54.1 a 24.6b 73.9 :I: 41 .2'
6 Unvaccinated control 616 80.9 a 13.2‘ 63.3 :I: 40.6'

 

1 IT: live vaccine administered by intratracheal inoculation; IM: live vaccine administered by intramuscular

injection. Inoculating dose is 5x10a cfu.
Percentage of lung tissue exhibiting A. pleurapneumaniae lesions; results presented as mean :I:

standard deviation.

 

3 Percentage of pleural surface area exhibiting pleuritis; results presented as mean :I: standard deviation.
°Values with different superscripts among the six vaccine groups were significantly different (p<0. 05) by
Least Significant Difference (LSD) analysis.

Table 4.3 - Clinical Score Data for Trial 1.

 

 

 

 

 

 

 

 

 

Group1 RR Maxz Oyspnea3 Depression4 Appetite"5
1 22.0 a; 53' 1.60 i .45“ 1.40 a .55T 2.00 a 71"—
2 19.7 :I: 5.7' 2.17 1: .41' 1.67 :1: .62' 0.75 :l: .96”
3 19.2 a 1.2' 1.63 1 .41“ 1.27 :l: .75' 2.33 :l: 1.21'I
4 16.2 :I: 3.4'I 1.20 1 .75” 0.40 :r .69‘ 0.20 1: .45°
5 23.3 a: 1.6' 1.63 a: .52“ 1.83 a: .75' 1.67 a 1.03“
6 23.0 a: 5.6' 2.33 a .52' 1.83 a: .75' 1.83 :t . 90"
Norm. 6.0 0 0 0
Max. 25 3 3 3

 

; See Table 4.2 for vaccination types.
Maximum respiratory rate observed after challenge. Respiratory rate recorded as number of breaths per
15 sec observation period.
Maximum dyspnea score observed alter challenge. Oyspnea score measures degree of respiratory
distress and labored breathing. Scored as 0 = normal; 1 = slight, 2 = moderate; 3 = severe.
Maximum depression score observed after challenge. Depression score evaluates attitude and activity.
Scored as 0 = normal; 1: slight inactivitjn 2 = moderate; 3 = severe.
Appetite was scored as 0: did eat; 1- - did not eat. Total score = number of 12 hour periods not eating
over 36 hour observation period.
°Values with different superscripts among the six vaccine groups were significantly different (p<0. 05)
by Least Significant Difference (LSD) analysis.

3

5

105

Table 4.4 - Mortality and Lung Score Data for Trial 2.

 

 

 

 

 

 

 

Death Time % 3 % 4
Group Vaccine1 Challefle Mortality (hrs)2 Pneumonia Pleuritis
1 APP-1A. APP-1A we 74. 0 1 16 53.4 1 26.5'I 51.9 1 343‘-
PBS-Rib
2 PBS-Rib APP-1A 6/6 26.0 1 13.1"° 60.3 1 4.6' 66.7 1 51 .6'
3 APPI-A. APP-5 4/5 40.0 1 26.1b 54.6 1 10.2' 49.0 1 50.1'
PBS-Rib
4 PBS-Rib APP-5 6l6 16.7 1 33° 61.0 1 5.6' 65.6 1 51 .0‘

 

 

 

 

1 Live vaccine administered by intramuscular injection. Inoculating dose is 2x10" cfu.
Results presented as mean :I: standard deviation.
Percentage of lung tissue exhibiting A. pleurapneumaniae lesions; results presented as mean :1:
standard deviation.

:“Percentage of pleural surface area exhibiting pleuritis; results presented as mean :t standard deviation.
°Values with different superscripts among the six vaccine groups were significantly different (p<0. 05) by
Least Significant Difference (LSD) analysis.

Table 4.5 Clinical Score Data for Trial 2.

 

 

 

 

 

 

 

 

 

Group1 RR Max: Dyspnea3 Depression“ Appetites
1 13.7 1 3.015 1.17 1 0.96I 1.17 1 0.96' 0.5 1 055"—
2 19.0 1 3.6' 2.33 1 0.52b 2.17 1 0.75” 2.0 1 0 b
3 15.6 1 1.5" 2.20 1 0.45” 1.60 1 0.45'“b 1.6 1 0.69”
4 14.0 1 2.6b 2.60 1 0.45b 2.20 1 0.45b 2.0 1 0 b
Norm. 6.0 0 0 0
Max. 25 3 3 3

 

; See Table 4.4 for vaccination types.
Maximum respiratory rate observed after challenge. Respiratory rate recorded as number of breaths per
3 15 sec observation period.
Maximum dyspnea score observed after challenge. Dyspnea score measures degree of respiratory
4 distress and labored breathing. Scored as 0 = normal; 1 = slight; 2 = moderate; 3 = severe.
Maximum depression scare observed after challenge. Depression score evaluates attitude and activity.
Scored as 0 = normal; 1= slight inactivity; 2 = moderate; 3 = severe.
Appetite was scored as 0= did eat; 1" - did not eat. Total score = number of 4 hour periods not eating
over 12 hour observation period
6Values with different superscripts among the six vaccine groups were significantly different (p<0. 05)
by Least Significant Difference (LSD) analysis.

5

106

DISCUSSION

Current commercial bacterin vaccines do not provide effective cross-protective
immunity against all strains of APP. However, infection with a virulent strain of APP
does elicit cross-protection against other serotypes. This suggests that there are
common antigens expressed during infection that elicit cross-protective immunity and
that these antigens may not be present in killed vaccines prepared from bacteria grown
in the laboratory. This was certainly true in the past when APP bacterins were routinely
prepared in media that did not elicit expression of Apx toxins. It remains to be
determined what other in viva specific antigens and virulence factors of APP are
produced as well as the specific environmental stimuli, such as temperature, lack of
available iron, pH, or osmotic conditions, that trigger their production. One way to
produce an APP vaccine that contains these cross-protective antigens, without
identifying them or the growth conditions which induce their production, is to use a
biochemically attenuated strain of APP that is still capable of synthesis of these
antigens and virulence factors as a live vaccine.

The genetically defined riboflavin auxotroph of APP serotype 1 used in these
studies as a live attenuated vaccine is phenotypically identical to its wild type parent
and is fully capable of synthesis of known virulence factors such as extracellular Apx
toxins, outer membrane proteins, lipopolysaccharide, and capsular polysaccharide (7).

In previous attenuation studies (7), we found that AP233, the riboflavin-requiring
vaccine strain, failed to persist in the porcine respiratory tract for more than 16 hours.
Poor persistence of live vaccine strains in viva can lead to a failure to elicit a protective
immune response (29,39). In order to assure that infection-associated antigens would
be expressed by this live attenuated vaccine strain after immunization of pigs, we
needed to ensure that the bacteria had sufficient available riboflavin to permit 2-3

generations of growth in vivo. We determined in vitro that addition of 5 pg of riboflavin

107

per ml to our vaccine inoculum containing 5 X 10° cfu was sufficient to permit this
amount of growth.

In trial 1, we compared intratracheal (IT) and intramuscular (IM) administration of
the live attenuated vaccine, with and without the addition of 5 pg/ml exogenous
riboflavin. Intramuscular vaccination with the live attenuated riboflavin-requiring
APP mutant, with the addition of a limited amount of exogenous riboflavin. led to
complete protection against mortality and to significant reductions in lung damage and
clinical signs of pleurapneumania. lntratracheal immunization with the same dosage of
live vaccine did not elicit the same degree of protective immunity. As predicted, the
addition of a limited amount of riboflavin to the live vaccine inoculum did improve the
efficacy of the vaccine, at least with lM immunization, most probably due to improved
persistence of the vaccine strain in vivo. We did not predict that IM immunization
would be more effective than the intratracheal route. This could be due to dispersal of
the IT dosage throughout the lung, and subsequent lack of containment of enough
riboflavin in close proximity to the bacterial cells to permit sufficient in viva replication
and persistence. It is also possible that intratracheal inoculation circumvented
important antigen sampling sites in the nasal passages that are implicated in the
generation of a protective immune response (43).

The live vaccine administered intramuscularly did not provide complete
protection against clinical signs of pleurapneumania in this experiment. We suggest
two possible reasons for this result. First the challenge administered, although
designed to be an LDso dose, was overwhelming, with 6/6 control animals killed in less
than 24 hours and 3/6 animals vaccinated with a bacterin that usually affords significant
protection also killed. Second, the dose of live vaccine used was daterrnined based on
the dosage that could be safely administered intratracheally. It should be possible to

use a higher dose for intramuscular immunization without adverse effects.

108

Seralagic analysis indicates that both forms of IM vaccination elicited high toxin
neutralization titers, but not significant ELISA or complement fixation (CF) titers.
However, only the IM vaccination with riboflavin elicited good protection. In contrast,
the whole cell bacterin elicited strong ELISA and CF titers, but elicited no hemolysin
neutralizing antibody and did not confer adequate protection against severe challenge.
This strongly suggests that the current serologic tests are not truly indicative of
protective immune responses. Certainly the antigen preparation used for this ELISA,
which is comprised of APP outer membranes prepared from bacteria grown in brain
heart infusion, does not contain APP antigens that are only expressed in viva.
Therefore, if antibodies against in vivo expressed antigens are critical to protection. an
accurate measure of the protective immune response would not possible with this
ELISA.

In the second vaccine trial, we attempted to duplicate, and perhaps improve, the
homologous protection seen in trial 1 and to extend mesa results to include
heterologous protection. However, there was a significant loss in homologous
protection in trial 2 as compared to trial 1, and we failed to achieve significant
heterologous protection. We believe this is the result of increasing the dosage of
AP233 fourfold without correspondingly increasing the amount of riboflavin in the
dosage. In the second trial, there was one-quarter the amount of riboflavin available to
each organism and thus not enough to allow sufficient growth in viva. Therefore this
vaccination was analogous to a bacterin vaccine. which we know is not entirely
protective against pleurapneumania in our challenge model. Because of this lack of
consistency in our two trials, we are unable to prove conclusively the degree to which
AP233 is protective against both homologous and heterologous challenge.

It is also interesting to note that we recently began acquiring our specific
pathogen free pigs from a high health status herd and using the new Biosafety Level 2

Michigan State University Research Containment Facility. We have determined that

109

the L050 for these animals kept under these very clean conditions is at least two fold
and likely a full log lower than had been previously established for a typical hard in a
more typical environment. It is evident that we will need to re-establish an LDso dosage
for future experiments in order to obtain accurate protection data. This may also have
implications for many of the high health status herds which are being managed to
prevent diseases like pleurapneumania as these herds may be extremely sensitive to
potential outbreaks of disease.

In summary, a deletion-disruption riboflavin mutant (AP233) of APP serotype 1
(ATCC 27088) which requires riboflavin for growth is highly attenuated in the porcine
host. causing no clinical signs, no typical pleurapneumanic lung pathology. and no
mortality. AP233 was tested as a live attenuated vaccine, prepared either with or
without the addition of riboflavin to the vaccine inoculum and delivered either
intratracheally or intramuscularly. Intramuscular vaccination with 5 X 108 cfu of live
AP233 with 5 pglml of exogenous riboflavin provided complete protection against
mortality and significant reductions in lung damage and clinical signs of
pleurapneumania. However, intramuscular vaccination with 2 x 109 cfu of live AP233
and 5 pg/ml of exogenous riboflavin provided only moderate protection against
homologous challenge and minimal protection against heterologous challenge. We will
need to repeat these vaccination experiments, altering the concentration of bacteria
and/or the amount of riboflavin in the vaccination inoculum and lowering the challenge

dose before conclusively evaluating the protective effects of vaccination with AP233.

110

ACKNOWLEDGMENTS
This research was supported by grants from Pfizer Animal Health and the
National Park Producers' Council.
T. E. Fuller is supported by a U.S.D.A. National Needs Graduate Fellowship in
Animal Biotechnology.

10.

11.

111

REFERENCES

Ahmed, Z.U., M.R. Sarker, and D.A. Sack. 1990. Protection of adult rabbits
and monkeys from lethal shigellosis by oral immunization with a thymine-
requiring and temperature-sensitive mutant of Shigella flexneri Y. Vaccine
82153-158.

Bacher, A. 1991. Biosynthesis of flavins. p. 215-59. In F. Muller (ed.).
Chemistry and Biochemistry of Flavins, Vol. 1. Chemical Rubber Company.
Boca Raton, Fla.

Bowe, F., P. O'Gaora, D. Maskell, M. Cafferkey and G. Dougan. 1989.
Virulence, persistence and immunogenicity of Yersinia enterocolitica 0:8 aroA
mutants. Infect. Immun. 57:3234-3236.

Collins, L.V., S. Attridga, and J. Hackett. 1991. Mutations at rfc or pmi
attenuate Salmonella typhimurium virulence for mice. Infect. Immun. 59:1079-
1085.

Fenwick, B. and S. Henry. 1994. Porcine pleurapneumania. J. Am. Vet. Med.
Assoc. 204:1334-1340.

Fuller, T.E. and M.H. Mulks. 1995. Characterization of Actinobacillus
pleurapneumaniae riboflavin biosynthesis genes. J. Bacteriol. 177:7265-7270.

Fuller, T.E., B.J. Thacker, and M.H. Mulks. 1996. A riboflavin auxotroph of
Actinobacillus pleurapneumaniae is attenuated in swine. Infect. Immun.
64:4659-4664.

Gennanier, R. and E. Furer. 1975. Isolation and characterisation of gaIE
mutant Ty21a of Salmonella tyth a candidate strain for a live oral typhoid
vaccine. J. Infect. Dis. 131:553-558.

Hoffman, LJ. 1989. Actinobacillus (Haemophilus) pleurapneumaniae: Use of
caagglutination and complement fixation to determine the relationship between
presence of organisms and antibody titer in slaughterhouse pigs. J. Vet Diagn.
Invest. 1:12-15.

Hoiseth S. K. and B.A.D. Stacker. 1981. Aromatic-dependent Salmonella
typhimurium are non-virulent and effective as live vaccines. Nature (Iondon).
291:238-239.

Homchampa, P., R. A. Stnignell and B. Adler. 1994. Construction and
vaccine potential of an aroA mutant of Pasteurella heemolytica. Vet. Microbiol.
42:35-44.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

112

Homchampa, P., R. A. Strugnell and B. Adler. 1992. Molecular analysis of
the aroA gene of Pasteurella multocida and vaccine potential of a constructed
aroA mutant. Mol. Microbiol. 623585-3593.

Hunneman, W.A. 1986. Incidence, economic effects, and control of
Haemaphilus pleurapneumaniae infections in pigs. Vet. Quarterly 8:83-87.

Inzana, T.J., J. Todd, and HP. Veit. 1993. Safety, stability and efficacy of
nonencapsulated mutants of Actinobacillus pleurapneumaniae for use in live
vaccines. Infect. Immun. 61:1682-1686.

lvins, B.E., S.L. Welkos. G.B. Knudsen and SF. Little. 1990. Immunization
against anthrax with aromatic compound-dependent (Aro-) mutants of Bacillus
anthracis and with recombinant strains of Bacillus subtilis that produce anthrax
protective antigen. Infect. Immun. 58:303-308.

Jolie, R.A.V., M.H. Mulks, and B.J. Thacker. 1995. Cross-protection
experiments in pigs vaccinated with Actinobacillus pleurapneumaniae subtypes
1A and 1B. Vet Microbiol. 45:383-391.

Jolie, R.A.V., M. H. Mulks, and B. J. Thacker. 1994. Antigenic differences
within Actinobacillus pleurapneumaniae serotype 1. Vet. Microbiol. 38:329-
349.

Kamell, A., P.D. Cain, N. Vanna and A.A. Lindberg. 1993. Arab deleteion
attenuates Shigella flexneri strain 2457T and makes it a safe and efficacious
oral vaccine in monkeys. Vaccine 8:830-836.

Kelly, S.M., B.A. Bosecker and R. Curtiss III. 1992. Characterization and
protective properties of attenuated mutants of Salmonella cholerasuis. Infect
Immun. 60:4881-4890.

Lindberg. A.A., A. Kamell, B.A.D. Stacker, S. Katakura, H. Sweiha and F.P.
Reinholt. 1988. Development of an auxotrophic oral live Shigella flexneri
vaccine. Vaccine 6:146-150.

Mekalanos, J.J. 1994. Live attenuted vaccine vectors. lntl. J. Tech. Assess.
Hlth Care 10:131-142.

Montaraz, J.A., B. Fenwick, H. Hill and M. Rider. 1996. Evaluating antibody
isotype-specific ELISA, complement fixation, and Apxl hemolysin neutralization
tests to detect serum antibodies in pigs infected with Actinobacillus
pleurapneumaniae serotype 1. Swine Health and Production 4:79-83.

Mulks, M.H. and J.M. Buysse. 1995. A targeted mutagenesis system for
Actinobacillus pleurapneumaniae. Gene 165261-66.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

113

Nicolet, J. 1992. Actinobacillus pleurapneumaniae, p. 401-408. In A.D.
Laman et al (eds). Diseases of Swine, 7th edition Iowa State University Press,
Ames, Iowa.

Nielsen, R. 1986. Serology of Haemophilus (Actinobacillus) pleurapneumaniae
serotype 5 strains: establishment of subtypes A and B. Acta Vet. Scand. 27:
49-58.

Nielsen, R. 1984. Haemaphilus pleurapneumaniae serotypes - Cross
protection experiments. Nord. Vet. Med. 36:221-234.

Nielsen, R. 1974. Serological and immunological studies of pleurapneumania
of swine caused by Haemaphilus parahaemolyticus. Acta Vet. Scand. 15:80-
89.

Noriega, F.R., G. Losonsky, C. Lauderbaugh, F.M. Liao, J.Y. Wang and M.M.
Levine. 1996. Engineered AguaB-A AvirG Shigella flexneri 2a strain CVD
1205: construction, safety, immunogenicity and potential efficacy as a mucosal
vaccine. Infect. Immun. 64:3055—3061.

O'Callaghan, D. D. Maskell, F.Y. Lieu, C.S.F. Easmon and G. Dougan. 1988.
Characterization of aromatic and purine dependent Salmonella typhimurium:
attenuation, persistence and ability to induce protective immunity in BAme
mice. Infect. Immun. 56:419-423.

Paltlneanu. D., R. Pambucol, E. Tirziu, and I. Scobercea. 1992. Swine
infectious pleurapneumania: Aerosol vaccination with a live attenuated vaccine.
Proc. Int. Pig. Vet Soc. 12:214

Perry, M.B., E. Altman, J.-R. Brisson, LM. Beynon, and J.C. Richards.
1990. Structural characteristics of the antigenic capsular polysaccharides and
lipopolysaccharides involved in the serological classification of Actinobacillus
pleurapneumaniae strains. Serodiag. Immunother. Infect Dis. 4:299-308.

Roberts, M.. D. Maskell, P. Novotny and G. Dougan. 1990. Construction and
characterization in viva of Bordetella pertussis aroA mutants. Infect. Immun.
58:732-739.

Rosendal. S., 0.8. Carpenter, W.R. Mitchell, and MR. Wilson. 1981.
Vaccination against pleurapneumania in pigs caused by Haemaphilus
pleurapneumaniae. Can. Vet. J. 22:34-35.

Rosendal. S.. and J. l. Maclnnes. 1990. Characterization of an attenuated
strain of Actinobacillus pleurapneumaniae, serotype 1. Am. J. Vet. Res. 51:
711-717.

35.

37.

38.

39.

40.

41.

42.

43.

114

Sansonetti P.J. and J. Arondel. 1989. Construction and evaluation of a
double mutant of Shigella flexneri as a candidate for oral vaccination against
shigellosis. Vaccine 72443-450.

Schott, K., J. Kellerman, F. Lottspeich and A. Bacher. 1990. Riboflavin
synthases of Bacillus subtilis: purification and amino acid sequence of the a-
subunit. J. Biol. Chem. 265:4204-4209.

Sebunya, T.N.K. and JR. Saunders. 1983. Haemaphilus pleurapneumaniae
infection in swine: a review. J. Am. Vet. Med. Assoc. 182:1331-1337.

Shape, R. E. 1964. Porcine contagious pleurapneumania. 1. Experimental
transmission, etiology, and pathology. J. Exp. Med. 119:357-368.

Sigwart, D.F., B.A.D. Stacker, and JD. Clements. 1989. Effect of a purA
mutation on the efficacy of Salmonella live vaccine vectors. Infect. Immun.
57:1858-1861.

Thacker, B.J., and M.H. Mulks. 1988. Evaluation of commercial Haemaphilus
pleurapneumaniae vaccines. Proc. Int. Pig Vet. Soc. 10:87.

Thacker, B. J.. M. H. Mulks, B. Yamini, 8 J. Krehbiel. 1988. Clinical,
immunological, hematological, microbiological, and pathological evaluation of a
percutaneous intratracheal injection Haemaphilus pleurapneumaniae challenge
model. Proc. Int. Pig Vet Soc. 10:69.

Utrera, V., C. Pijoan. and T. Molitor. 1992. Evaluation of the immunity
induced in pigs after infection with a low virulence strain of A.
pleurapneumaniae serotype 1. Proc. Int. Pig Vet. Soc. 12:213.

van Ginkel, F.D., J.L. VanCott, S. Yamamoto, M. Marinara, K. Fujihashi, M.
Yamamoto, M.N. Kwaon, H. Kiyono and JR. McGhee. 1996. Current
vaccination approaches with emphasis on mucosal pathogens and immunity in
the respiratory tract. Abstracts, HAP 96. #27.

Vaughan, L.M., P.R. Smith, and T.J. Foster. 1993. An aromatic-dependent
mutant of the fish pathogen Aeramanas salmonicide is attenuated in fish and is
effective as a live vaccine against the Salmonid disease furunculosis. Infect.
Immun. 61:2172—2181.

Chapter 5

DEVELOPMENT OF AN IVET (IN VIVO EXPRESSION TECHNOLOGY) SYSTEM FOR
ACTINOBA CILLUS PLEUROPNEUMONIAE

115

116

ABSTRACT

To aid in the identification of genes involved in the pathogenesis of APP
infection. we have developed an in viva expression technology (IVET) system to
identify APP gene promoters that are specifically induced in viva during infection. A
functional APP IVET system requires a defined biochemically attenuated APP mutant.
the wild type biochemical gene(s) from a different source that can complement the
mutation, a readily quantifiable reporter gene with no background in APP and a delivery
vector. We have constructed an APP mutant that lacks most of the operon required for
synthesis of riboflavin and that fails to survive in viva. When the ribBAH genes from
Bacillus subtilis were cloned into this Rib- mutant under the control of a functional APP
promoter, both ability to grow without riboflavin and virulence were restored. We have
shown that the quAB genes from Vibria harveyi can be used to monitor gene
expression in APP and that APP is not natively bioluminescent. Our initial APP-IVET
promoter-trap vector (pTF86) contains, in sequence, the T4 terminator, a unique
BamHI site, a promoterless copy of the B. subtilis ribBAH genes, and a promoterless
copy of the V. harveyi quAB genes in the E. cali-APP shuttle vector pGZRS-19.
Plasmid pools were constructed by cloning Sau3A fragments from APP into the BamHI
site in pTF86 and fiansforming into the APP Rib- mutant Pigs were infected with pools
of 400 transforrnants by endobronchial inoculation and surviving bacteria isolated from
the pigs' lungs at 12 to 24 hours post-infecfion. Only those transformants containing
cloned promoters active in viva should have survived due to expression of the cloned
ribBAH genes. Strains that survived in viva, but which minimally expressed Lux in vitro.
should contain cloned promoters that are specifically induced in vivo. We report the
identification of 5 clones containing promoters that are induced in viva (during
infection). Two of these clones were putatively identified by amino acid sequence
similarity as the secE-nusG operon and the mrp gene. This is the first report of an

IVET system for use in the family Pasteurellaceae.

117

INTRODUCTION

Actinobacillus pleurapneumaniae (APP) is the causative agent of an
economically devastating. acute necrotizing hemorrhagic bronchopneumonia with
accompanying fibrinous pleuritis (4,12,19,29). We have chosen to use APP as a model
system to study the pathogenesis of respiratory disease in order to identify potenfial
targets for the development of new live vaccines and antibiotics. There are many
characteristics which APP shares with other members of the family Pasteurellaceae.
For instance, RTX type toxins are also produced by Actinobacillus suis and Pasteurella
heemolytica (30). Also, as with Pasteurella multocida and Pasteurella heemolytica,
there are multiple antigenic serotypes (24,30), and killed whole cell bacterins provide at
best serotype-specific protection or no signifcant protection (2022.26.30.33) while
natural or experimental infection elicits good cross-protective immunity (20.21.23.30).
This suggests that there are cross-protective antigens expressed in viva which are not
expressed during in vitro growth for bacterin production.

Historically. research on this and other pathogens has focused on the studies of
antigens expressed in vitro; however, this at best only gives a fractional picture of
virulence-associated characteristics. Increasing attention is being given to the
environmentally-stimulated regulatory mechanisms that modulate bacterial attributes in
the host (9.10.16). Known environmental regulatory signals for bacterial virulence
genes include temperature, osmotic strength, pH, oxygen levels. availability of iron,
calcium levels, carbon sources. urea, growth phase, stress, starvation and availability of
other nutrients such as amino acids (9.10.16). Attempts to mimic host environmental
conditions during in vitro growth have demonstrated that there are many differences in
gene expression in response to various environmental signals. A serious limitation of
this technique is the multitude of environmental factors that can induce gene
expression. One or more of these factors, individually or in combination, could produce

a wide variety of responses in the pathogen. Which combination of known

118

environmental stimuli are important at any given stage of infection is virtually impossible
to determine in vitro. In addition, there may be many environmental conditions currently
undiscovered in the host which have a significant effect on gene expression.

Another method for studying these antigens is to use convalescent antisera to
directly compare the antigenic profiles of bacteria isolated directly from the site of
infection with bacteria that have been cultured in the laboratory. We have used
bronchoalveolar lavage (BAL) to recover bacteria directly from the lungs of infected
pigs and have demonstrated that there are indeed differing immunogenic bands on
Western blot with convalescent pig sera as compared to bacteria cultured in the lab
(18). This technique is also restrictive because it will only identify antigenic differences
and will not identify regulatory or non-immunogenic factors. Also, further analysis of
these factors is limited by the ability to identify, clone and express these antigens in
vitro.

In view of these limitations, genetic technology has emerged to the forefront as a
means to study bacterial pathogenesis. Genetic techniques have been recently
described that can be used to identify genes that are specifically induced only during
infection and not during growth on standard laboratory media. The power in these
techniques is that they can identify genes required for survival in the host such as
enzymes necessary for production of critical growth factors not available in viva (e.g.,
purines and aromatic amino acids), factors which allow survival within host cells (e.g.,
enzymes that inactivate phagocytic killing mechanisms), or even toxins that have not yet
been identified because they are not produced under standard in vitro culture conditions
(16). Identification of such in viva expressed genes should lead to further insights into
the pathogenesis of APP disease as well as to identification of important antigens for
inclusion in subunit vaccines and to development of defined avirulent mutants for use as

live vaccines.

119

Four genetic methods for the study of bacterial pathogenesis have been
described to date: subtractive hybridization, in viva expression technology (IVET).
resolvase based-IVET and signature-tagged mutagenesis (STM). Subtractive
hybridization is based on identifying differences in RNA transcripts (34) between
bacteria isolated from an infected animal and bacteria grown in the laboratory.
Limitations may be the technical difficulty in isolation and manipulation of bacterial RNA
to obtain a cDNA clone, and the results may change dramatically depending on the
stage of infection at which bacteria are isolated necessitating a large number of
animals in order to truly perform a saturation study of all the genes involved in
pathogenesis. Also, this technique will tend to identify "all or none” expressed genes
rather than in viva induced genes which have some level of constitutive expression.

A second method termed IVET has been developed which allows positive
selection of bacterial genes that are specifically expressed during infection and has
been used to identify in viva induced genes in the pathogen Salmonella typhimurium
(15,32). The basis of this system is to use a tandem set of in vivo and in vitro
promoterless "reporter” genes to identify promoters and genes in random genomic
fragments which are turned on specifically in the host. The host organism is a
biochemically attenuated mutant which is unable to survive unless complemented by
expression of the wild type biochemical genes driven by a cloned promoter (in viva
"reporter“ gene). Bacterial strains that survive in vivo are isolated and expression of the
in vitro reporter gene is measured. The reisolated strains which do not express the
quantifiable reporter gene contain ivi (in viva induced) promoters that are turned on in
viva and off in vitro. Knock-out mutants of three different S. typhimurium ivi genes
(carAB, pheST-himA, rib) have been shown to be attenuated in mice (15).

Modifications of this system using different promoterless genes for the selection
process, such as chloramphenicol acetyltransferase (16), have been developed.

Cloned promoters are selected for expression during infection by treating with the

120

antibiotic and isolating surviving bacteria. These antibiotic selections would appear to
function best in tissue culture or in small animal models rather than in large animal
models because of phannacokinetics. Another system to identify in viva expressed
genes has been developed for use in the intracellular pathogen Legioneiia
pneumaphila. using an avinilent thyA mutant. an independently replicating plasmid, and
infection of a macrophage cell culture rather than an intact host (13,25). Finally, a
modified IVET system has been developed for use with Pseudomonas aeruginosa
consisting of an avinilent purEK mutant, plasmid pBR322, a promoterless purEK
operon and a mouse model of infection (35). The major limitation with the IVET
technology is its ability to recover only promoters which are expressed throughout the
infection process. Promoters which are expressed only transiently or during latter
stages of infection will not be recovered because of fl'IS lack of complementation and
thus the lack of survival of the biochemical mutant.

The remaining two technologies are designed to overcome the limitations
of IVET and subtractive hybridization and can identify genes that are expressed only
transiently or at low levels. Signature-tagged mutagenesis (STM) overcomes the
problem of an IVET type system in that genes do not have to be expressed during the
entire infection process to be reisolated. In this system, transposons carrying unique
DNA sequence tags are used to construct a collection of mutants by insertional
mutagenesis (11). The pool of mutants is used to infect a host animal, and surviving
bacteria are recovered from the infected host after several days. Using the unique DNA
sequence tags to make probes, it is possible to identify which specific members of the
original mutant pool are absent upon recovery of bacteria from the host. These mutants
are predicted to carry mutations in genes necessary for survival in the host. The STM
system is not yet applicable to A. pleurapneumaniae, since the transposon's (T n5)
antibiotic marker does not appear to function in APP. The system also requires the

generation of individually tagged transposon mutants of each new organism to be

121

studied, and in large animal models the entire surviving pool will be difficult to reisolate,
possibly leading to the false identification of mutants as unable to survive. A certain
subset of virulence genes (e.g. toxins) will likely not be identified because other clones
in the pool will express them and the transposon mutant will be able to replicate normally
even though it is not able to express these genes.

The fourth and perhaps most promising selection system based on resolvase
activity uses in viva induction of tnpR resolvase, a site-specific recombinase of the
transposable element ya, to induce a heritable, detectable genetic change. In this
system, even brief expression of the promoterless resolvase by a cloned promoter
results in excision of a tetracycline resistance cassette flanked by direct repeats of the
DNA sequence at which the resolvase functions (1). Any bacterial cell in which the
resolvase is expressed will give rise to progeny that are tetracycline-sensitive. This
selection system has the power to identify promoters that are only transiently expressed
in viva because all strains survive whether or not the promoter is inactive, turned on
transiently, or turned on constitutively during infection. A second limitation as it currently
exists is the reliance on a negative selection requiring replica plating of colonies to
determine whether they have lost antibiotic resistance. Another limitation is the
sensitivity of the system; it can be too sensitive yielding a high level of background with
the false identification of genes being turned on. Another limitation is that all promoters
turned on in viva will be identified. not just those that are in viva specific; thus, pre-
screening of the pools is required to eliminate in vitro-expressed clones. As a result, this
system will likely not identify genes that are minimally to moderately expressed in vitro
and induced in viva. It is not yet clear whether this system could be made to function in
APP.

Many of the genes identified as having promoters specifically turned on in vivo
or as being required for survival in viva (1,11,15,35) have been genes involved in the

biosynthesis of purines, pyrimidines, and essential amino acids, such as purD, pyrE,

122

and carAB, while others have been genes known to encode vimlence factors. such as
invA. invG, sva, fedB. and virB or even transcriptional regulators such as Fur.
Previously identified genes have been divided into four different classes of ivi
expressed genes: 1) enzymes in critical biochemical pathways; 2) known virulence
factors, or genes involved in the synthesis of such factors. such as genes for LPS
biosynthesis/modification; 3) new genes, not identifiable by comparison with known
sequence data; and 4) genes involved in transcriptional regulation or signal
transduction. In vivo expressed genes that can be identified as steps in critical
biochemical pathways, such as purine biosynthesis, may be targets for the production
of knock-out mutants to be used as live avirulent vaccines or may be targets for a new
generation of antibiotics.

In this study, we report the construction and utilization of a modified IVET
system for APP based on an attenuated riboflavin mutant of APP, the shuttle vector
pGZRS-19 (36), the T4 terminator, the Bacillus subtilis ribBAH genes and the quAB
genes from Wbrio harveyi. Using this system initially with a limited number of clones.
representing approximately 10% of the number required to be truly representative of
the entire genome. we have identified five clones that are specifically induced in viva.
Two of these clones have been putatively identified as clones of the secE-nusG operon
and the mrp gene. The identity of the other clones remains undetermined. but they will
perhaps be the most interesting clones as no homologous genes have previously been

sequenced.

123

MATERIALS AND METHODS

Bacterial strains and media. A. pleurapneumaniae strains were cultured at 37
°C in either brain heart infusion (BHI) or heart infusion (HI) (Difco Laboratories, Detroit,
MI). containing 10 pglml B-NAD (V factor) (Sigma Chemical Company. St. Louis, MO).
Riboflavin (Sigma) was added to a final concentration of 200 pglml as necessary. E.
coli strains were cultured in Luria-Bertani medium. Ampicillin was added to 100 pglml
for plasmid selection in E. coli strains. For A. pleurapneumaniae strains, 50 pglml
ampicillin was added except for selection after electroporation which was done with 20
pglml ampicillin.

DNA manipulations. DNA modifying enzymes were supplied by Boehringer-
Mannheim Biochemicals (Indianapolis, IN) and used according to the manufacturer‘s
specifications. Genomic DNA was prepared according to the lysislproteinase K method
of the Gene Fusion Manual (31). Plasmid DNA preparations, agarose gel
electrophoresis. and E. coli transformation were all performed by conventional methods
(27).

PCR. PCR was performed to eliminate Xbal and BamHI restriction sites and to
introduce Sphl sites in the 5' and 3' ends of the quAB genes. Primers were MM43 (5'-
GCA-GCA-TGC-ACT-AGA-GGA-ACC-CCA-TG-3')for the forward and MM44 (5'-GCA-
GCA-TGC-G‘IT-AAA-CGT—TAC-GAG-TG—3') for the reverse direction. The PCR
reaction was carried out using Vent polymerase (New England Biolabs, Beverly, MA),
450 pmol of each primer. 100 ng of the pRL10626 template, 2 mM MgSO4 with 2mM
dNTPs in a reaction volume of 100 ul. Reaction conditions were repeated for 30 cycles
with 1 minute at 95°C for denaturing, 1 minute at 48°C for annealing and 2-1/2 minutes
at 72°C for polymerization. The desired 2.2 kb PCR product was purified by phenol-
chloroform extraction, restriction digested with Sphl and isolated by purification from an
agarose gel with the QIAEX gel extraction kit (QIAGEN, Chatsworth. CA).

124

Electroporation of A. pleurapneumaniae. AP233 was grown in 100 ml BHIV
+ riboflavin at 37°C, with shaking at 150 RPM, to an 00520 of 0.7. Cells were chilled
on ice and centrifuged at 5,000 x g at 4°C for 10 min. Cells were washed twice in 10.0
ml ice cold sterile 15% glycerol. Cells were resuspended in 2.0 ml 15% glycerol and
frozen in 50 pl aliquots using a dry ice-ethanol bath. Plasmid DNA was added to an
aliquot of competent cells thawed on ice and then transferred to a 0.1 cm gap
electroporation cuvette (BioRad. Hercules, CA). Cells were electroporated using a
Gene Pulsar ll (BioRad) with the following settings: voltage. 1.8 kV; resistance, 200 a;
capacitance, 25 de. Cells were allowed to recover in BHIV + riboflavin at 37°C with
slow shaking for approximately 3 to 4 hours before plating on selective media.

Experimental infections. Eight-week-old specific-pathogen-free castrated
male pigs (Whiteshire Hamroc, lnc., Albion, IN) were housed in separate cage units in
BSL-2 isolation rooms at the Michigan State University Research Containment Facility.
All experimental protocols for animal experiments were reviewed by the Michigan State
University All University Committee on Animal Use and Care, and all procedures
conformed to university and USDA regulations and guidelines.

For preparation of challenge inocula, each pool of approximately 100
transfonnants was added to 3 ml HIV + 5 mM CaCIz + riboflavin + ampicillin to produce
a starting 00520 of 0.2. Pools were grown in 20x100 mm culture tubes, at 37°C with
shaking at 160 RPM, to an 00520 of 0.8. Appropriate amounts of each pool were
combined to produce 10.0 ml culture containing ~ 2 X 106 cfu each of 300-400
individual strains. The combined culture was harvested by centrifugation at 2100 x g at
room temperature and washed once with sterile PBS pH 7.0 with 5 pglml riboflavin
(PBS-Rib). The cell pellet was resuspended in 10 ml of PBS-Rib and diluted to obtain
the desired 8 x 10° cfu dose.

For the challenge procedure, pigs were anesthetized by intravenous injection

with xylazine (1.65 mglkg) and telazol (4.4 mglkg) and inoculated by shallow

125

endobronchial injection with the 8 x 103 cfu suspended in 10 ml saline. Clinical signs of
pleurapneumania. including increased respiration rate, elevated temperature, dyspnea,
decreased appetite and depression were monitored and scored as previously described
(14). Animals were euthanized at approximately 12 hours post infection. All animals
were necropsied. and lungs were examined macroscopically for A. pleurapneumaniae
lesions, edema, congestion, hemorrhage, necrosis, abscessation, fibrosis and pleuritis.
APP strains were recultured by one of two methods. Bronchoalveolar lavage (BAL)
(18) was performed by insertion of a tube into the lungs followed by the instillation and
recovery of PBS-Rib. The recovered BAL fluid was first centrifuged at 450 x g to pellet
eukaryotic cells or extraneous tissue and then the supematant was centrifuged at 2100
x g to pellet any bacterial cells. The pellet was resuspended in 1 ml of PBS-Rib and
either used directly for Lux quantitation or diluted appropriately for plate counts and
recovery of single colonies. The second method consisted of excising the lungs and
using sterile scalpel blades to make incisions every 1-2 cm which were swabbed with
sterile cotton-tipped applicators and cultured onto BHIVR2°°A5° agar plates.

Lux Analysis: Qualitative screening of strains for expression of the quAB
genes in vitro was performed on a Hamamatsu C1966 Photonic Microscope System.
Briefly, colonies on agar plates were exposed to 50 pl of N-decyl aldehyde distributed
evenly on a glass petri dish lid. The plate was analyzed using appropriate sensitivity
settings, which varied with the amount of expression by the colonies on each plate.
Final screening to select clones for quantitation after eliminating strong Lux-i- colonies '
was done with an aperture setting of 11, digital gain of 2 and photon counting for 59
seconds. Quantitative analysis of Lux expression was performed using a Turner Model
20a Luminometer. N-decyl aldehyde (Sigma) substrate was made by dissolving 20
mglml Essentially Fatty Acid Free BSA (Sigma) in 1.0 ml of H20 with 1 pl/ml N-decyl
aldehyde. This mixture was sonicated in a glass screw cap test tube for approximately

1 hour to disperse the N-decyl aldehyde into micelles. For analysis, 20 pl of the culture

126

was mixed with 20 pl of substrate in polypr0pylene Iuminometer cuvettes (Turner
Designs, Sunnyvale, CA) and mixed for 10 seconds. The mixture was then read in full
integral, autoranging mode with a pre-delay of 0 seconds, delay of 10 seconds. and
integration time of 30 seconds. Luminometer readings were normalized to micro-
relative light units per colony forming unit (pRLUICFU) with the number of bacteria in
the sample as determined by plate counts on selective media. or by normalizing to the
00520 when in pure culture.

DNA/Protein sequence analysis. Nucleotide and amino acid sequences were
searched against the GenBank database using the FASTA and TFASTA functions of
the GCG Programs (8). All alignments, calculated percent identities. and similarities
reported were made using the GAP function of the 606 Programs (8).

127
RESULTS

In order to construct an IVET system we need an attenuated biochemical
mutant and a vector which contains the following: random chromosomal DNA
fragments from the organism, promoteriess wild type biochemical genes from a
different organism to complement the attenuating mutation, and a
quantitative/qualitative reporter gene. We have previously reported the cloning of the
riboflavin operon from APP (6) and the construction of a defined deletion-disruption
riboflavin mutant of APP that is attenuated in swine (7). In this paper, we report the
construction of an APP IVET vector and its use in our riboflavin auxotroph to identify in
viva induced genes in Actinobacillus pleurapneumaniae.

Verification of complementation by Bacillus subtilis ribBAH genes. A 2.7
kb EcoRV fragment containing a promoterless copy of the ribBAH genes from Bacillus
subtilis strain 168 was excised from plasmid pRF2 (provided by Dr. John Perkins,
OmniGene Bioproducts Inc.. Cambridge, MA) and ligated into the Smal site of pGZRS-
19 (36) to construct plasmid pTF74. This plasmid was electroporated into AP233. a
riboflavin-requiring, nalidixic acid and kanamycin resistant attenuated mutant of APP
serotype 1. The restoration of virulence was determined as described previously (7). A
dosage of AP233 + pTF74 equivalent to the wild type L050 was administered to 3 six-
to eight-week-old pigs, resulting in significant signs of pleurapneumania including
increased respiration rate, dyspnea, depression. loss of appetite and pneumonia and
pleuritis as determined by necropsy. These results indicate that the B. subtilis ribBAH
genes are able to complement the AP233 riboflavin mutation in trans as AP233 is
normally unable to cause disease at a dosage which is 500 times the established wild
type L050.

Verification of Lux Activity in Actinobacillus pleurapneumaniae. A
promoterless quAB operon from Vibrio harveyi was cloned in both orientations into the

APP-E. cali shuttle vectors pGZRS-18 and pGZRS-19. When the resulting four

128

plasmids were transformed into AP233 and screened for Lux activity. all four conferred
detectable Lux activity, indicating background expression from unidentified promoters in
both vectors. In contrast, neither AP233 alone nor AP233 containing either pGZRS-18
or pGZRS-19 showed detectable Lux activity, indicating that there is no native
bioluminescence in APP.

Construction of the IVET Vector: pTF86. Figure 5.1 outlines the construction
of the IVET vector pTF86. A 2.3 kb fragment containing the promoterless quAB genes
was cloned from pRL10626 (provided by Dr. Peter Walk, MSU-PRL, East Lansing, MI)
by PCR, replacing the Xbal/BamHl sites with Sphl sites for insertion into pGZRS-19 to
make pTF77b. The 2.7 kb EcoRV fragment containing the promoterless B. subtilis
ribBAH genes from pRF2 was blunt-end ligated into the Klenow-filled Hindlll site of
pTF77b to make pTF85. The T4 terminator from pJFF224-NX (5) was inserted into
pTF85 on a 0.3 kb Kpnl-Xbal fragment to make pTF86. There is a unique BamHI site
upstream of the promoterless quABzribBAH fusion into which Sau3A fragments can be
cloned. AP233 containing pTF85. which does not contain the T4 tanninator, is able to
grow without riboflavin and is weakly Lux positive. pTF86 is unable to grow without
riboflavin and is Lux negative. indicating that the T4 terminator does indeed prevent
background expression from the pGZRS-19 vector.

Construction of an APP promoter library in pTF86. Chromosomal DNA
fragments generated by digestion with Sau3A, 0.4 to 1.0 kb in size. were purified from
an agarose gel and ligated into the BamHI site in pTF86. The ligation mix was
electroporated into AP233. Single colonies were selected, and sub-cultured onto
BHIVR2°°A5° plates in groups of 100. Colonies were washed off each plate and frozen
at -70°C in HIVR broth + 5 mM CaCI2 + 20% glycerol.

In vivo screening of IVET clones. An initial library of 1800 clones was
screened for the identification of chromosomal fragments containing in viva induced

promoters (Figure 5.2). Groups of 300 to 400 clones were used to infect pigs by

129

“mewicfl

  

locZ' ‘

  

 

 
 

 

    
 

”a: “T 5""1
., pCZRS-I9 4 ,
‘9 4.3 kb quA luxB

    
 
  

PCR from pRL10620
- eliminate Xbol-Bomlll siles
- inlroduce Sphl sites

 

 

 

 

sun
iii
no
5:. turn in our th. lb. rear
.3... g l i II I
as

an ribA rin

EcoRV fragment
from pRF2

 

    
 
  

ribB I018

   

quA

   
 
   

. pTF85
rin
L 9.3 kb locZ' .

mirror

Kpnl-Xbal fragment

Irorn pir790
lieu—Xbal

luxB

    
 
   

ribB

   

“a:
luxA
ribI-l pTF86 as 3."
9.6 kb 14° "" f-‘m
\ locZ"

    
 

Figure 5.1 - Construction of the IVET Vector pTF86.

1 30
Plasmid pTF86 APP-1A chromosomal DNA

  
 

Digest
Bani HI - CIP

Digestwlth Sau3A
Purify 0.4 - 1.0 kb fragnnnts

Ugate and elecboporde Into AP233 (RIM

Select Ap resistant colonies on media containing rlboflavin

I

Poolingroupsoffllllandfroeaeat-GIIc

I

Crow 4 pools Indlvldually to an 00520 of 0.8

I

Combine cultures, centrifuge aid wash with smile PBS, dilute appropriately
lnoeulate 8 x io'cru endobronchially

jam.

 

Euthailze at III - 48 hours p.l.

I

Culture strains from lungs aseptically by swab or bronchoalveolar lavage
onto BHIV + riboflavin «i Ap

I

ScreenforLirxexpressloiilnviiro
aridselectLux-colonles

I

Isolate plasrnld DNA and sequence Into chromosomal Insert

Figure 5.2 - Identification of In Vivo Induced Genes.

131

endobronchial inoculation. At either 10, 24 or 48 hours post infection, the pigs were
euthanized and necropsied. and surviving bacteria were isolated from the lungs.
Screening of this pool of clones in vitro before in viva passage demonstrated that
approximately 90% to 95% of the pool was composed of strains which had minimal or
no Lux expression, presumably containing no functional promoters or promoters which
were not expressed in vitro, and that 5% to 10% were Lux+. After reisolation from
infected pigs, the entire pool was approximately 1% Lux- and 99% Lux+, indicating a
strong selection in viva for those clones which had promoters expressed in viva.

Quantitation of In vivo induced (ivi) clones. A total of 750 individual colonies
were isolated from four pigs infected with the IVET pools. An example of the range of
promoter activity observed during primary qualitative screening is shown in Figure 5.3.
Clones with expression between the negative controls (pTF86 and pGZRS19) and a
minimally expressed clone (IVF) are selected for further analysis. Clones containing
strong in vitro expression (e.g. IVH) are likely to not be induced in viva. Forty-six
reisolated clones were selected during a primary screen on agar plates as being Lux
negative or as having relatively minimal Lux expression. Sixteen additional clones were
selected that contained strong in vitro promoters for future promoter studies as little is
known about APP promoter structure. All 62 clones were assayed for Lux activity in
vitro (Figure 5.4) and those with an activity of less than 200 RLU/OD520 were partially
sequenced with a primer reading upstream of the quA gene in plasmid pTF86. The 20
selected clones that were sequenced represented 5 unique clones.

Recovery of pTF85. As a positive control in a preliminary experiment, pTF85
(Figure 5.1) was included in an equimolar amount with 300 other clones. This clone
was reisolated after 24 hours and was confirmed by sequencing to demonstrate the
lack of a Sau3A insert and T4 terminator. The reisolation of this clone, with Lux activity
of only 75.9 RLUs per ODS-4, indicates that this system is extremely sensitive to

expression by a cloned promoter.

132

GUIbUM-P

 

Figure 5.3 - Primary Screening for Minimally Expressed IVET Clones in AP233.

Panel A is a incidental light photograph of triplicate colonies on an agarose plate.

Panel

Row 1
Row 2
Row 3
Row 4
Row 5
Row 6

B is the corresponding photograph of the bioluminescence

pGZRS—19

pTF86

IVA (mrp promoter)

IVF (secE-nusG promoter)

|V42 (moderately expressed promoter)
IVH (strong promoter)

133

Lux Activity of Selected NET Clones

16m «-

1400

 

1200

 

1M

 

 

 

§

 

 

04 Illlllllllllllllllll'lll'l
2 223333333

"82‘28R22 85333883

11
91785
i3

.170

Clonal

Figure 5.4 - In Vitro Quantitative Lux Analysis of Selected IVET Clones.

134

Sequence comparison. TFASTA and FASTA searches were performed
against the GenBank database with either the nucleotide sequence or predicted amino
acid sequences of the IVET clones. and two of the clones were putatively identified.
The predicted amino acid sequence from IVA had 54.8% similarity and 45.2% identity
with the Mrp (methionyl t-R NA synthetase related protein) from Haemophilus influenzae
and 61.3% similarity and 45.2% identity with E. coli Mrp protein (Figure 5.5). The 31
amino acid overlap shown in Figure 5.5 occurs 50 amino acids from the amino terminus

of E.cali Mrp (a 369 amino acid polypeptide) .

Amino Acid Alignment of APP IVA with E. coli Mrp

APP MPFAWNSGFEALKADTEVKLKQVTGANEVKW
lll.|:|:|l.l|.:....| .:l|l..:.l
E.c. ...TLHVELVMPFVWHSAFEELKBQCSAELLRITGAKAIDWKLSHNI...

Amino Acid Alignment of APP IVA with H. Influenzae Mrp

APP MPFAWNSGFEALKADTEVKLKQVTGANEVKW

=|||Il|| |-l|---- l--|=-- =Il
8.1. ...TLRIELQLPFAWNSGAEQLKQAVSDALLKATDCKLIKWAVAYQI...

Figure 5.5 - Amino Acid Alignment of NA and Mrp. Identical residues are indicated
with " | " and conserved residues are indicated with "z " or " . "

135

The two predicted ORFs within the sequence for IVF were homologous to the
carboxy terminus of SecE (a 127 amino acid polypeptide) and the amino terminus of
NusG (a 181 amino acid polypeptide) which are separated by two nucleotides in E. coli
and co-transcribed. ORF1 (25 amino acids) was 88% similar and 60% identical with the
carboxy terminus of Haemophilus influenzae SecE and 88% similar and 64% identical
with the carboxy terminus of Escherichia coli SecE. ORF2 (40 amino acids) was 63.3%
similar and 55.3% identical with NusG from Haemophilus influenzae and 60% identical
and 54% similar with NusG from Escherichia coli. (Figure 5.6).

The remaining three ivi clones could not be assigned a putative identity based
on their sequence.

Measurement of In vivo induction. To confirm that the reisolated clones had
indeed been induced in viva, each clone was quantitated for Lux expression in viva.
One clone which contained a good constitutive promoter (IVH) was used as a control
for the regeneration of virulence of AP233. Another previously identified clone
containing a weak in vitro and in viva promoter and partial ORF putatively identified as
a thioredoxin (T rxA) homologue was also included in these experiments. A dosage of 5
x 103 cfu of each IVET clone was administered to six- to eight-week-old pigs by
endobronchial instillation. After eight hours. animals were euthanized and
bronchoalveolar lavage (BAL) was performed to reisolate surviving bacteria. Animals
infected with all IVET clones except the pig containing the TrxA clone showed
significant clinical signs and evidence of acute pleurapneumania at necropsy. Bacteria
obtained by BAL after differential centrifugation to remove eukaryotic contaminants
were immediately assayed for Lux activity and plate counts of viable bacteria were
determined to standardize the readings. Each clone was induced in viva ranging from
8 to 88 fold induction (Figure 5.7) except the trxA clone which was unable to be

accurately quantitated due to the lack of survival within the host.

136

Amino Acid Alignment of APP IVF with H. Influenzae SecE

APP WSLVLWGIDSIIVTLVTFLTNLRF
:.Il.:l::llllll::.lll:lll
3.1. . . .TMIASLE‘FWAVDSIIV‘I‘VINFLTDLRF

Amino Acid Alignment of APP IVF with H. Influenzae NusG

APP MSEVENTEATKMRWYVLQAFSGFENRVAVTLRDISNYTKW

-|:l...|-llllllllllll.|ll=|||=- .
8.1. MTEETTVSKKRWYVLQAFSGFESRVALTLREYIKQQQMEDQFG...

Amino Acid Alignment of APP IVF with E. coil SecE

APP WSLVLWGIDSIIVTLVTFLTNLRF
I:Il:lll:l:l:l ll.l:l.lll
rs. c. . . .TAVMSLILWGLDGILVRLVSFITGLRF

Amino Acid Alignment of APP IVF with E. coil NusG

APP MSEVENTEATKMRWYVLQAFSGFENRVAVTLRDISNYTKW
-I|-|-l|ll=|l|||I|-|l|--Il= -
8.6. MSEAPKKRWYVVQAFSGFEGRVATSLREHIKLHNMEDLFG...

Figure 5.6 - Amino Acid Alignment of IVF and SecEINusG. Identical residues are
indicated with " | " and conserved residues are indicated with ": " or " . "

137

1000-

100 1

uRLUlcfu 10 q

 

 

Clone
El in vitro Iin viva

Figure 5.7 - In Vivo Induction of IVET Clones. Fold induction is indicated above each
in viva bar.

138
DISCUSSION

In this study we report the construction of an IVET system for use in
Actinobacillus pleurapneumaniae. To our knowledge, this is the only IVET system
which has been made specifically for use in an infection model of the pathogen's
natural host rather than a laboratory animal or tissue culture model. We have
demonstrated that a functional IVET vector can be constructed with the E. cali-APP
shuttle vector pGZRS-19 (36), the T4 tanninator, the promoterfess wild type riboflavin
biosynthesis genes (ribBAH) from Bacillus subtilis and the promoterless luxAB reporter
genes from Vibrio harveyi. We have shown that the Bacillus subtilis genes can
complement the riboflavin mutation in AP233 restoring the ability to grow without
riboflavin and restoring virulence. We have shown that there is no background
bioluminescence in APP and that the luxAB genes can be used as a simple and
quantifiable reporter of gene expression in APP. Finally, our IVET system has been
used to identify 5 specifically in viva induced genes which are likely to be important for
survival in vivo or virulence.

Two of the in viva induced clones have been putatively identified based upon
their sequence homology as clones of the secE-nusG (3. 28) operon and the mrp gene
(2). SecE is an essential integral cytoplasmic membrane protein necessary for the
signal sequence dependent export of proteins (3.28). It is likely that the increase in in
viva expressed proteins (e.g. transferrin binding protein) which contain a signal
sequence for export may require a concomitant increase in the secretion machinery
necessary for its export. It is unknown what role the NusG protein. known to be
involved in lambda antitennination. plays as no bacteriophage have been previously
identified in APP; however, in E. coli NusG has been shown to be essential for viability
and perhaps this protein also may play a role in the antiterrnination regulation of
virulence genes. The Mrp protein has been putatively identified as a potential ATPase

(2) which may be involved in the salvage of nucleosides in viva for production of

139

recycled nucleotides or for obtaining carbon. nitrogen or energy. One other possibility
is that mrp gene expression by virtue of its overlap in the antisense orientation with a
metG promoter, can regulate the expression or activity of methionyl t-RNA synthetase
(2). It remains to be determined what the exact role for these clones and the
unidentified clones is in the pathogenesis of APP.

We have had some problems identifying significant in viva induction and
obtaining exact quantitative numbers for our IVET clones in viva. Part of this problem is
that the riboflavin mutant AP233 is unable to survive well when diluted and plated for
colony counts. Another potential problem is that a clone containing the IVET vector
with a strong constitutive promoter (IVH) also appeared to be induced in viva. It is
unknown whether this induction is due to experimental artifact or whether it is indeed
induced 9 fold in viva. If this alone is not induced in vivo and represents the error in our
measurement, then perhaps only clones IVB, IVE and IVF (secE-nusG) are truly
induced in viva. However, the vimlence of AP233 containing these IVET plasmids can
also yield important information. For example, the weak trxA promoter in the IVET
plasmid in AP233 did not restore virulence even at 100 times the normal wild type L050
dosage, indicating there was not enough expression of the wild type rib genes to
overcome the attenuating mutation. We also recovered approximately 1000 fold fewer
organisms from the animal containing this clone. In contrast, the other IVET clones
restored the virulence of AP233 to a level equal or surpassing that of IVH which we
know contains a strong promoter. This would suggest that even though all 5 clones
have similar minimal expression to the trxA clone in vitro, they are induced to a level at
least equal to the expression of IVH, and thus all 5 clones could qualitatively be said to
be induced in viva.

It will likely be necessary in the future to alter our IVET procedure to provide a
more effective screen for in viva induced promoters. It is important that the in viva

selection process is allowed to continue for an appropriate amount of time such that

140

week promoters are not carried through and reisolated. Also, it may be necessary to
pre-screen the IVET pools and eliminate strong constitutive promoter clones. This
would allow the use of a greater number of colonies in each animal as well as perhaps
extending the time available for selection because the animal would not become
seriously ill as fast. To overcome the problem of quantitation of Lux activity from IVET
clones in viva. it may be necessary to re-transfonn any IVET plasmids containing
potential in viva induced promoters into the wild type strain of APP serotype 1, which
appears to be less fragile and able to provide more accurate plate counts. This will
provide a more accurate quantitative measurement but will not reconfirm the ability of
the promoter to restore virulence by expression of the wild type riboflavin genes. The
further identification of genes specifically induced in viva will likely lead to new insights
about the pathogenesis of this organism and may identify targets for the production of

new antibiotics or live attenuated vaccines.

141

ACKNOWLEDGMENTS
This research was supported by a grant from the USDA T. E. Fuller is
supported by a U.S.D.A. National Needs Graduate Fellowship in Animal Biotechnology.

10.

11.

12.

13.

142

REFERENCES

Camilli, A., and J. J. Mekalanos. 1995. Use of recombinase gene fusions to
identify Vibrio cholerae genes induced during infection. Mol. Microbiol. 18:671-
683.

Dardel, F., M. Panvert and G. Fayat. 1990. Transcription and regulation of
expression of the Escherichia coli methionyl t-RNA synthetase gene. Mol. Gen.
Genet. 223:121-133.

Downing, W.L., S.L. Sullivan, M.E. Gottesman and P.P. Dennis. 1990.
Sequence and transcriptional pattern of the essential Escherichia coli secE-
nusG operon. J. Bacteriol. 172:1621-1627.

Fenwick, B. and S. Henry. 1994. Porcine pleurapneumania. J. Am. Vet. Med.
Assoc. 204:1334-1340.

Frey, J. 1992. Construction of a broad host range shuttle vector for gene cloning
and expression in Actinobacillus pleurapneumaniae and other Pasteurellaceae.
Res. Microbial. 143:263-269.

Fuller, T.E. and M.H. Mulks. 1995. Characterization of Actinobacillus
pleurapneumaniae Riboflavin Biasynthesis Genes. J. Bacteriol. 177:7265-
7270.

Fuller, T.E., B.J. Thacker and M.H. Mulks. 1996. A riboflavin auxotroph of
Actinobacillus pleurapneumaniae is attenuated in swine. Infect. Immun.
64:4659-4664.

Genetics Computer Group. Sept. 1994. Program Manual for the Wisconsin
Package, Ver. 8. Genetics Computer Group. Madison, Wisconsin.

Griffiths, E. 1991. Environmental Regulation of Bacterial Virulence- Implications
for Vaccine Design and Production. Trends in Biotechnology 9:309-315.

Gross, R. 1993. Signal Transduction and Vinilence Regulation in Human and
Animal Pathogens. FEMS Micro. Rev. 104:301-326.

Hansel, M. J. E. Shea, C. Gleason, M. D. Jones, E. Dalton, and D. W. Holden.
1995. Simultaneous identification of bacterial virulence genes by negative
selection. Science 269:400-403.

Hunneman, W.A. 1986. Incidence, economic effects, and control of
Haemophilus pleumpneumoniae infections in pigs. Vet. Quarterly 8:83-87.

lsberg, RR. 1994. Intracellular trafficking of Legionella pneumophila within
phagocytic cells. pp.263-278. In V.L. Miller et al (ed). Molecular Genetics of
Bacterial Pathogenesis. ASM Press. Washington. DC.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

143

Jolie, R. A. V.. M. H. Mulks, and B. J. Thacker. 1995. Cross-protection
experiments in pigs vaccinated with Actinobacillus pleurapneumaniae subtypes
1A and 1B. Vet. Microbiol. 45:383-391.

Mahan, M.J., J.M. Slauch, and J.J.Mekalanos. 1993. Selection of bacterial
virulence genes that are specifically induced in host tissues. Science 2592686-
688.

Mahan, M.J., J.W. Tobias, J.M. Slauch, P. C. Hanna, R. J. Collier, and J. J.
Mekalanos. 1995. Antibiotic-based selection for bacterial genes that are
specifically induced during infection of a host PNAS 92: 669-673.

Mekalanos, J.J. 1992. Environmental signals controlling expression of
virulence determinants in bacteria. J. Bacteriol. 17421-7.

Mulks, M.H. Unpublished data.

Nicolet, J. 1992. Actinobacillus pleurapneumaniae, p. 401-408. In A.D.
Laman et al (eds), Diseases of Swine. 7th edition Iowa State University Press,
Ames, Iowa.

Nielsen, R. 1984. Haemaphilus pleurapneumaniae serotypes - Cross
protection experiments. Nord. Vet. Med. 362221-234.

Nielsen, R. 1979. Haemaphilus parahaemolyticus serotypes: pathogenicity
and cross immunity. Nord. Vet. Med. 312407-413.

Nielsen, R. 1976. Pleuropneumonia of swine caused by Haemaphilus
pleurapneumaniae. Studies on the protection obtained by vaccination. Nord.
Vet. Med. 28:337-338.

Nielsen, R. 1974. Serological and immunological studies of pleurapneumania
of swine caused by Haemaphilus parahaemolyticus. Acta Vet. Scand. 15:80-
89.

Perry, M.B., E. Altman, J.-R. Brisson, L.M. Beynon, and J.C. Richards.
1990. Structural characteristics of the antigenic capsular polysaccharides and
lipopolysaccharides involved in the serological classification of Actinobacillus
pleurapneumaniae strains. Serodiag. Immunother. Infect. Dis. 42299-308.

Rankin, S., and R. R. lsberg. 1993. ldenfification of Legionella pneumophila
loci expressed in response to an intracellular environment. Infect. Agents Dis.
22269-271.

Rosendal. S., D.S. Carpenter, W.R. Mitchell, and MR. Wilson. 1981.
Vaccination against pleurapneumania in pigs caused by Haemaphilus
pleurapneumaniae. Can. Vet. J. 22234-35.

27.

28.

29.

30.

31.

32.

33.

35.

36.

144

Sambrook, J.. E.F. Fritsch and T. Maniatis. 1989. Molecular Cloning: A
Laboratory Manual. 2nd. ed. Cold Spring Harbor Press. Cold Spring Harbor,
New York.

Schatz, P.J., P.D. Riggs, A. Jacq, M.J. Path and J. Beckwith. 1989. The
secE gene encodes an integral membrane protein required for protein export in
Escherichia coli. Genes Dev. 3:1035-1044.

Sebunya, T.N.K. and JR. Saunders. 1983. Haemaphilus pleurapneumaniae
infection in swine: a review. J. Am. Vet. Med. Assoc. 182:1331-1337.

Shewen, RE. and J.A. Rice Conlon. 1993. 18IPasteureIIa, p. 216-225. In
C.L. Gyles and CO. Thoen (eds), Pathogenesis of Bacterial Infections in
Animals, 2nd edition Iowa State University Press, Ames, Iowa.

Silhavy, T.J. 1984. DNA extraction from bacterial cells. p. 137-139. In
Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring
Harbor, NY.

Slauch, J.M., M.J. Mahan, and J.J. Mekalanos. 1994. In vivo expression
technology for selection of bacterial genes specifically induced in host tissues.
Math. Enzymol. 235:481-492.

Thacker, B.J., and M.H. Mulks. 1988. Evaluation of commercial Haemaphilus
pleurapneumaniae vaccines. Proc. Int. Pig Vet. Soc. 10:87.

Utt, E.A., J.P. Brousal, L.C. Kikuta-Oshima and FD. Quinn. 1995. The
identification of bacterial gene expression differences using mRNA-based
isothermal subtractive hybridisation. Can. J. Microbiol. 412152-156.

Wang, J.. A. Musheglan. S. Lory and S. Jin. 1996. Large-scale isolation of
candidate virulence genes of Pseudomonas aeruginosa by in vivo selection.
PNAS 93:10434-10439.

West, S.E.H., Romero, M.J.M., Regassa. L.B., Zielinski, NA and RA. Welch.
1995. Construction of Actinobacillus pleurapneumaniae-Escherichia cali shuttle
vectors: expression of antibiotic resistance genes. Gene 160:81-86.

Chapter 6

SUMMARY

145

146

Our long term objectives were to better understand the pathogenesis of
Actinobacillus pleurapneumaniae infections as a model of respiratory disease and to
use this information to generate better vaccines and new antibiotics. To contribute to
these objectives, the goal of this thesis research was to design and utilize a genetic
system to identify specifically in viva induced (e.g., infection associated) genes in this
organism. At the time this project was started little was known about the genetics of the
swine pathogen Actinobacillus pleurapneumaniae. Only one unstable shuttle vector
and one transposon mutagenesis system were available to us, and there were no
defined biochemical mutants.

We have identified, cloned and completely sequenced four genes from
Actinobacillus pleurapneumaniae that are involved in riboflavin biosynthesis. The
cloned genes can specify production of large amounts of riboflavin in E. coli, can
complement several defined genetic mutations in riboflavin biosynthesis in E. coli, and
are homologous to riboflavin biosynthetic genes from both E. coli and Bacillus subtilis.
The genes have been designated APP ribGBAH due to their similarity in both sequence
and arrangement to the B. subtilis ribGBAH operon. The DNA sequence data,
complementation, and minicell analysis strongly suggest that the four rib genes are
transcribed from a single APP promoter upstream of the ribG gene. Biosynthesis of
riboflavin by APP appears to be more similar to that in the gram-positive bacterium B.
subtilis than in the gram-negatives E. coli or Haemaphilus influenzae. Further study of
this model operon should reveal interesting information on regulation and
promoter/operon structure in APP, as well as information on the role of riboflavin
biosynthesis in APP infection.

We hypothesized that APP must synthesize riboflavin to meet its own metabolic
demands during infection, since riboflavin is not synthesized by mammals and therefore
is not likely to be freely available to APP within its porcine host, especially within the

relatively sterile lungs. We constructed a serotype 1 Actinobacillus pleurapneumaniae

147

deletion-disruption riboflavin mutant that is attenuated in vivo. The A.
pleurapneumaniae ribGBAH operon was disrupted by deleting an internal segment of

R cassette using a targeted mutagenesis

the operon (ribBA) and replacing it with 6 Km
technique. A stable riboflavin-requiring, genetically-defined, KmR mutant, AP233, was
phenotypically identical to its wild type parent based on analysis of proteins,
extracellular toxin, LPS, and capsular polysaccharide by SOS-PAGE, immunoblot, and
caagglutination.

Experimental infection of pigs, the only natural host for A. pleurapneumaniae,
demonstrated that the riboflavin-requiring mutant was unable to cause disease
including lung lesions or clinical signs at dosages as high as 500 times the LDso for the
wild-type parent. It is important to note that the riboflavin-requiring mutant used in
these studies is a deletion mutant, with ~1.4 kb of the riboflavin operon removed from
the chromosome and replaced with an antibiotic resistance marker. We observed
neither reversion to prototrophy nor loss of kanamycin resistance in this mutant in the
laboratory suggesting that reversion to prototrophy and thus virulence will not occur in
viva.

This is the first report that a mutation in riboflavin biosynthesis in a pathogenic
bacterium is attenuating. This finding represents a novel addition to the group of
biosynthetic mutations which can be used to construct attenuated strains of bacteria.
This is also the first report of a genetically-defined attenuated mutant of A.
pleurapneumaniae that is still capable of production of all of the major known virulence
factors of this organism. including extracellular toxins and capsular polysaccharide.
The fact that riboflavin biosynthesis is essential for this pathogen and that riboflavin is
not synthesized by higher eukaryotes could potentially lead to discovery of a new
generation of antibiotics which inhibit riboflavin biosynthesis.

Live-attenuated vaccine strains of numerous pathogenic organisms are being

developed and appear to hold the greatest promise for the future of vaccination. Many

 

148

of these strains contain mutations in biochemical pathways which are necessary for
survival inside the host. We have tested our riboflavin-requiring mutant of APP as a
live-attenuated vaccine. Poor persistence of live vaccine strains in viva can lead to a
failure to elicit a protective immune response. In order to assure that infection-
associated antigens would be expressed by the live-attenuated vaccine strain of
bacteria after immunization of pigs, we needed to ensure that the bacteria had
sufficient available riboflavin to permit 2-3 generations of growth. An IM immunization
with a dose containing sufficient riboflavin gave very good protection against APP in
the face of an ovenivhelming challenge. Further experimentation with varying dosages
and riboflavin concentrations is required in order to determine the optimal vaccine
regimen to elicit good protection.

We have constructed an IVET system for use in our riboflavin-requiring strain of
Actinobacillus pleurapneumaniae, the first of its kind for use in the family
Pasteurellaceae. We have demonstrated that a functional IVET vector can be
constructed with the E. cali-APP shuttle vector pGZRS-19, the T4 terminator, the
promoterless wild type riboflavin biosynthesis genes (ribBAH) from Bacillus subtilis and
the promoterless luxAB reporter genes from Vibrio hanreyi. We have shown that the
Bacillus subtilis genes can complement the riboflavin mutation in AP233, restoring the
ability to grow without riboflavin and restoring virulence. We have shown that there is
no background bioluminescence in APP and that the luxAB genes can be used as a
simple and quantifiable reporter of gene expression in APP. Finally. our IVET system
has been used to identify 5 specifically in viva induced genes which are likely to be
important for growth or virulence of APP in its natural host. Based upon their sequence
similarity, two of the identified clones have been predicted to encode polypeptides
homologous to E. coli SecE/NusG and E. coli Mrp. It remains to be determined what
the exact role for these clones and the unidentified clones is in the pathogenesis of

APP.

149

With a rapid mutation rate and the potential for horizontal gene transfer,
pathogens seem to have the advantage in the continual struggle against their hosts.
We now realize that antibiotics and vaccines are not the magical answers they were
originally thought to be. Therefore, it is imperative that we continue to use modern
genetic technology to study the pathogenesis of organisms which are presently causing
disease and anticipate how to overcome the next resurgence of diseases that seem to
be irrelevant at this time. The further identification of genes specifically induced in viva
will likely lead to new universal insights about the pathogenesis of these organisms and
may identify targets for the production of a new generation of antibiotics or new live

attenuated vaccines.