• No results found

Molecular analysis of the structure and expression of the Aeromonas Salmonicida surface layer protein gene vapA

N/A
N/A
Protected

Academic year: 2021

Share "Molecular analysis of the structure and expression of the Aeromonas Salmonicida surface layer protein gene vapA"

Copied!
191
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

MOLECULAR ANALYSIS OF THE STRUCTURE AND EXPRESSION OF THE AEROMONAS SALMON1CIDA SURFACE LAYER PROTEIN GENE v a p A

A C C E P T E D ^

ACULiTY OF GRADUATE d T S h i j i a n Chu

B. Med., Shanghai Medical University, 1982 oAf|3issertation Subm itted in Partial Fulfillment of j the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

in the D epartm ent of Biochemistry a n d Microbiology

We accept this thesis as conforming to the required standard

Dr. T. T./Trust, Supervisor

Dr. W. W. Kay, Departm ental Member

Dr, Ry^V. ^ a ^ f v Dcpaftm entj.l Member

Dr. P. j. RomaniukJ jDe^artrrfeptaf Member

Dr. M. J. Ashwood-Sm ith,'Outside M ember

Di .TV. J. F(a^e,' E te rn a l Exi miner ©SHIJIAN CHU, 1993 University of Victoria.

All rights reserved. Dissertation may not be reproduced in whole or in part, by photocopying or other means, w ithout the permission of the author.

(2)

Supervisor: Dr. Trevor J. Trust

ABSTRACT

Aeromonas calmonicida is a Gram negative rod shaped bacterium capable

of causing furunculosis in salmonid fish and other chronic and inflam m atory diseases in goldfish, carp and also salm onids. The surface layer of A ,

salmonicida, the A-layer, has been dem onstrated to be a major virulence factor

for the organism , and its su b u n it A -protein has b een p u rifie d an d its Structural gene vapA lias been cloned.

The vapAx gene from A, salmonicida strain A450 w as subcloned (pSC150) and expressed in Escherichia coli. Its DNA sequence w as then determ ined to consist of 1,506 bp encoding a 502-amino acid residue protein, containing a 21- residue signal peptide and a m ature protein of 50,778 Dalton. The A-protein assembled on the cell surface in the form of an 5-layer w as refractile to trypsin cleavage while trypsin digestion of the purified m ature protein revealed a highly resistant 39,400 D alton N -term inal fragm ent and a 16,700 D alton C- term in al frag m en t w ith m o d erate resistance, These try p sin -re sis ta n t fragm ents m ay form distinct stru ctu ral dom ains, consistent w ith three- dim ensional ultraStructural observations.

The plasm id pSC150 contained 62 bp of Aeromonas DNA in front of the

vapA structural gene. A prom oter (P2) w as predicted in this region w hich

show ed sequence hom ology to the E. c o lic70 prom oter. H ow ever, prim er extension in the w ild type strain A44S> show ed a transcriptional start site 181 bp upstream from the gene, and thus, another prom oter (PI) w as show n to be the m ajor prom oter. The DNA sequence coding for the untranslated leader mRNA contained two stem -loop structures, a putative sm all open reading frame spanning the stem-loop structures, and a palindrom ic sequence which

(3)

overlaps the predicted ribosome binding site. Northern analyses of A449 vapA mRNA show ed that incubation at 15°C produced the highest level of the transcript, and the transcript half-life w as 22 m in in cells grow n at 15°C com pared to 11 min in cells grow n at 20°C. DNA gyrase inhibitors nalidixic acid a n d novobiocin significantly reduced the vapA transcript level.

A . salmcnicida 30°C. m u tan ts w ere1 found to produce significantly

reduced levels of A-protein an d some of them were shown t;o have the native insertion elements, ISAl and ISA2, inserted in the vapA area. These insertion elem ents have been cloned a sequenced, and also identified in the w ild type strains A449 and A450. ISA2 w as show n to have sequence sim ilarity to other bacterial insertion elements.

Plasm id encoded vapA expression in E, coli was also affected by a dow nstream gene abcA, w hich, w hen deleted from the clone, significantly reduced vapA expression. This reduction could be complemented by the abcA gene carried on a second plasmid. In addition, the lipopolysaceharide (LPS) O- chain deficient phenotype of A449 m u tan t strain TM4, w hich has the abcA gene in terru p ted by ISA l, w as also com plem ented by abcA. DNA sequence analysis show ed that the abcA gene coded for a 308 am ino acid residue protein, w hich w as Confirmed by in vivo and in vitro expression and gene fusion w ith lacZ, and w as localized in the inner m em brane fraction of E. coli. At the N -term inal p a rt of th e protein, the predicted sequence of AbcA d isp lay ed high hom ology w ith a bacterial tran sp o rt p rotein su p er family/ in clu d in g a w ell conserved nucleotide b in d in g sequence. This binding sequence was show n b y site-directed m utagenesis to be required for LPS O- chain com plem entation in TM4. ATP binding activity w as confirm ed in the purified AbcA-LacZ fusion protein. A leucine zipper-basic region sequence w ith p re d o m in a n tly a -h e lic e l c o n fo rm a tio n w as p re d ic te d fu rth e r

(4)

dow nstream , w ith leucine residues in four of the five h eptad repeats and valine residue in the rem aining heptad repeat,

(5)

Examiners:

Dr. T. J. Trust, Supervisor

Dr. W. W. Kay, Departmeijj&l Member

Dr. P. J. Romahitjjfk, Departmenia^^Hember

Dr, M. J. Ashwood-Smith, Outsia5"Member

Dr. W. J. Page, fe?)terr^J Examiner

(6)

TABLE OF CONTENTS

ABSTRACT...ii

TABLE OF CONTENTS...v i LIST OF TABLES...ix

LIST OF FIGURES...,... x

LIST OF ABREVIATIONS... xiii

ACKNOWLEDGEMENTS ... xv

INTRODUCTION... 1

Bacterial S-layers... 1

Biological functions of S-layers... ...1

S-layer proteins... 4

S-layer protein genes...5

S-layer protein gene p ro m o ters... 7

Aeromonas salmonicida... ...8

The genus Aeromonas... 8

The species... 9

Pathogenesis ... 10

A -layer ... 13

Morphology ... 13

S u b u n its...15

A-layer form ation and anchoring... 16

V iru len ce ... .17

Other biological functions ... 18

O ther Aeromonas S-layers... ... ... 20

Lipopolysaccharides (LPS)... 21

Purpose of this dissertation... ... ... ... ... ,22

MATERIALS AND METHODS... ... 25

Bacterial strains, vectors and m ed ia... 25

DNA techniques ... .,...25

DNA preparation... ... 25

Aeromonas chrom osomal D N A ,... 25

Plasmid DNA preparation ... 28

M13 DNA preparation.. ... 29

Oligonucleotide preparation ... 30

Molecular cloning... ...30

Preparation of E, coli com petent cells ...30

DNA restriction digestion... 31

DNA lig atio n ... 31

E lectroporation. ... 32

M13 cloning... ...32

Southern an aly sis... ,.33

Non-radioactive labeled DNA....!... 33

DNA fragments radiolabeled b y nick translation ...34

End-labeled oligonucleotides ... 34

Polymerase chain reaction (PCR)... 35

DNA sequencing.,., ... 35

(7)

Autom ated sequencing,., ...36

Site-directed m utagenesis... 37

C om puter analysis... ,...38

TINA preparation and detection... 38

RNA p rep aratio n ... 38

N orthern hybridization ... 39

RNA stability assay ... 39

Prim er extension experim ents... 39

Detection and purification of proteins... 40

Sodium dodecyl sulfate-poiyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie Blue stain in g 40 Im rnunogold electron microscopy assay ,... ,..,.41

Cell fractionation... *..,.41

C ulture su p ern atan t... 41

P erip lasm ... ,...42

C ytoplasm ... *42

Inner m em brane an d o uter m em brane ...43

W estern blot ... *... ...43

Silver stain for proteins... .,.,,..44

Protein preparation for N -term inal sequencing ... 44

T7 polymerase in vivo transcription... 45

In vitro gene expression ... 46

Purification of the AbcA-LacZ fusion p rotein... ,.47

P-galactosidase assay ,,... 48

N ucleotide affinity chrom atography... 48

LFS detection ... 49

Silver stain for LPS... 49

RESULTS... 51

I. Structure of vapA gene and its product, A-protein.. ... ,.51

Subcloning and sequencing of the A. salmonicida A449 A-protein gene (vapA)... ... »... 51

Subcloning cf the vapA gene from A. salmonicida A450... 54

DNA sequencing of the A450 vapA gene... 55

Conservation of the vapA g e n e ... 62

A-protein structure... 63

Cellular localization of A-protein in E, coli...67

A -protein conform ation in p erip lasm ...67

II. Characterization of the vapA gene prom oter a rea... 73

Subcloning and sequencing of the flanking region in front of the vapA gene... 73

Prim er extension analysis ... 77

vapA transcript siz e... 79

G row th phase dependence of vapA transcription... 79

Effects of grow th tem perature on vapA transcription ... ....82

vapA transcript stability... i,... 82

Variation of vapA transcript1: levels in other grow th co n d itio n s... ,... ... ,... 85

Iffects of nalidixic acid and novobiocin on vapA transcription... ....,... 85

(8)

III. The vapA gene in some A, salmonicida 30°C m utants is

interrupted by insertion elem ents.,... 89

DNA rearrangem ents in 30°C m utants... 89

Identification of two insertions in the v'ipA. area ill A, salmonicida 30°C m utan ts... 93

Cloning and sequencing of ISA2... ...95

Effect of insertion elements in the vapA prom oter region on A-protein production... 102

IV, Characterization of a gene dow nstream from oapA ...103

A-protein expression in E. coli affected by a dow nstream Aeromonas DNA sequence... >103 vapA expression in E. coli can be complem ented by the downstream sequence... 103

DNA sequence analysis of the dow nstream reg io n ... 105

Predicted AbcA protein structure... 113

Identification and localization of AbcA p ro tein ... ,,.,...115

Analysis of abcA expression... ...118

Purification of the AbcA-p-galactosidase fusion protein and identification of ATP binding activity ...122

abcA pene complements LPS O-chain expression in TM4... 124

Site-directed mutagenesis in the P-loop of A bcA ... ...126

DISCUSSION ,... ,,... 129

I. Structure of vapA and analysis of its prom oters...129

II. Aeromonas insertion elem ents...140

III. The abcA gene ... ...143

IV. S um m ary ... 149

REFERENCES... 151

APPENDIXES ... 172

I. Entire DNA sequence determ ined in this study...172

(9)

ix

l i s t o f t a b i

.

e s

Table 1. Aeromonas strains vised in this study ... ... ... 26

Table 2. Vectors used in this study... ,... ,... 27

Table 3. Codon usage for the A. salmonicida A450 vapA gene, and A. hydrophila genes coding for exported p ro te in s ... 60

Table 4. vapA transcript level affected by p H ... 86

Table 5. vapA transcript level affected by Ca++ addition and depletion...86

Table 6. vapA transcript level affected by Fe+++ depletion...87

Table 7. vapA transcript level affected by M g++ addition... 87

Table 8. vapA transcript level affected by oxygen... 87

Table 9. vapA transcript level affected by nalidixic acid ... 88

Table 10. vapA transcript le vel affected by novobiocin... 88

Table 11. Results of DNA hybridization of A. salmonicida 30°C m utants 96 Table 12. Codon usage for the A. salmonicida A450 abcA gene and vapA gene

(10)

Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11.

LIST OF FIGURES

Schematic illustrations of the bacterial envelopes containing S-layers ... ... ... ... 2

Electron micrographs of Aeromonas A 4 a y e r... ...14 Physical m ap o f the cloned A. salmonicida chromosomal DNAs containing the vapA g e n e ... 52 W estern blot of A-protein subclones in pTZ18R ... 56 Nucleotide sequence of the vapA gene and flanking DNA from A.

salmonicida strains A449 and A450, an d the translated am ino acid

sequence ... 58

Southern blot analysis of BamHI digested chrom osomal DNAs from A. salmonicida strains probed w ith pSC650 ... 64 Kyte-Doolittle hydrophilicity plot of the A -p ro tein ... 64 SDS-PAGE analysis of A. salmonicida A450 A-protein structure by TPCK-trypsin cleavage and CNBr h y d ro ly sis...66

Schematic m aps of A. salmonicida A450 A-protein secondary structure, CNBr hydrolysis fragm ents and TPCK-trypsin

fragments ... 68

W estern im m unoblot analysis of 7.5% SDS-PAGE show ing the cellular localization of A-protein coded by pSC150 in E. coli

D H 5 a ... 69 Western blot analysis o f 12% SDS-PAGE of TPCK-trypsin partially digested periplasmic A-protein from A. salmonicida and E. coli

stra in s... 71

Figure 12. DNA sequence of the 5* region of the vapA gene of A450, the im m ediate upstream flanking DNA, and the predicted

(11)

Figure, 13. Figure 14, Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26.

amir?o acid seq u en ce .... ... ... ... ... 74 Prim er extension of the vapA gene in A449 and A450 ... 78 N orthern hybridization of w ild type and m utant Aeromonas

strains probed w ith TJ41... ... ... ... 80

vapA transcript dynamics during cell grow th at 1 5 ° C ... ,, 81 vapA transcript level at various tem peratures m easured by

N orthern h y b rid izatio n ... ... 83

vapA transcript stability a ssa y ... ... ... 84

Electron m icrographs of im m unogold labeling of A450 and

A450-3 c e lls... 90 Southern hybridization of A450 an d A450-3 chromosomal

DNAs probed w ith oligonucleotide TJ44... ... 91 Com parison of physical m aps of A450 and A450-3 chromosomes ... ... 92 W estern blot of whole cel] lysate of A. salmonicida 30°C m utants w ith rabbit anti-A-protein a n tise ru m ... ...94

A. salmonicida 1SA2 sequence ... ... . 97

Amino acid sequence alignm ent of the predicted 1SA2 encoded protein and other proteins coded by prokaryotic insertion

e lem en ts... ... . 100

Restriction m aps of inserts in each subclone used for abcA

subcloning and com plem entation ... ... 104 W estern blot detection of A-protein and densitom etrie analysis of the blot show ing the compleinentation in B, coli by abcA,,, 106 A-protein production from E. coli strains w ith and w ithout abcA gene detected by in vitro transcription-translation... . . 107

(12)

Figure 27. Nucleotide sequence of the abcA gene and flanking DNA from A.

salmonicida strain A450, and translated aminp acid se q u en c e 108

Figure 38. Schematic m ap of deduced sequence of A b cA ... 114 Figure 29. a-helix structure of the leucine zipper region betw een L211 to

L239 viewing from one end of the h e lix ... 116 Figure 30. Alignment of the AbcA protein of A. salmonicida w ith other

ABC-transporter proteins ... ... 117 Figure 31. In vivo plasm id expression and localization of AbcA protein

from pSC161 using the T7 polymerase expression system ... 119 Figure 32. Expression, localization and purification of AbcA-p--galactosidase

fusion protein as show n by SDS-PAGE and W estern blot analysis using monoclonal antibody apainst E. coli p-galactosidase 121 Figure 33. Nucleotide binding ability of the purified AbcA-P-galactosidase

fusion protein .,... 123

Figure 34. Silver stained SDS-PAGE gel of LPS samples ... 125 Figure 3F Sequences indicating m utation sites in the abcA gene P-loop

constructed by site-directed m utagenesis... 127 Figure 36. Phylogenetic tree of putative transposases from ISA2 and other

(13)

LIST OF ABREVIATIONS

A600 Absorbance at 600 nm ATP Adenosh.-;: triphosphate

1 i '

-bp Base pair

cAMP Cyclic adenom onophosphate cfu Colony form ing u n it

CRP cAMP receptor protein DEPC D iethylpyrocarbonate DTT D ith io th reito l

EDTA Ethylenediam inetetraacetic acid

EGTA Ethylene glycol-bis(P-aminoethyl ether)N,N ,N ',N'-tetraacetic acid h H o u r kb Kilobase Kdi K ilodalton j LB Luria-Bertani broth i

L-agar Luria-Bertani agar LPS Lipopolysaccharide M M olar m g m illig ram ! i pg m icrogram m in M in u te rr 1 m illiliter (xl m icroliter m M m illim o lar M r M olecular w eight

(14)

ng n an o g ram

ORF O pen reading frame PBS Phosphate buffered saline Pfc polyethylene glycol

pg pieogram

pi Isoelectric point pm ol picom olar

PSI Pounds p er Square inch rbs ribosome binding sequence

sec Second

SD Shine Dalgarno ribosomal binding sequence SDS Sodium dodecyl sulfate

SDS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis SPK buffer The 10 X SPK buffer contains 200 mM Tris Cl (pH 8.0), 50 mM

MgCl2, 5 mM DTT, 1 mM EDTA, 500 mM KC1, and 50% glycerol. TE buffer 10 mM of Tris and 1 mM of EDTA, pH 8.0

Tris Tris (hydroxyn\ethyl) am inom ethane

(15)

ACKNOWLEDGEMENTS

I am g ratefu l to my su p erv iso r Dr, T. J. T ru st for his co n stan t inspiration, guidance and encouragem ent. I th an k all the m em bers in Dr. T rust's lab, past a n d present, for their valuable help and su p p o rt in my experim ents. I also th an k m y family for their su p p o rt and understanding during m y program .

This stu d y w as su p p o rted by U niversity of Victoria Fellowship and research grants to Dr. T. J. Trust.

(16)

INTRODUCTION

Bacterial S-iayers

S-layers, or paracrystalline surface protein arrays, are reg u lar two-dim ensional assemblies of protein o r glycoprotein m onom ers th at constitute the o uter layer of the cell envelope of a wide variety of bacteria (143,191) (Fig. 1). The subunits in these S-layer lattices are held together an d onto the u nderlying surface b y oion-covalent forces (142). S-layers w ere relatively

; ! , , : ;

unknow n three decades ago. The first S-layer to be described w as by H ouw ink in a study of a Spirillum species by electron microscopy in 1953 (94). Today, S-layers have been identified on hundreds of bacteria and archaea (143,192, 193) and studies on S-layers have expanded from sim ple morphological analysis to considerations of their biochem istry, im m unochem istry, m olecular biology and their role in pathogenesis.

For com parative an d classification p u rp o ses, S-layers have been classified according to their space groups, unit cell size, ahd the position of their protom ers and pores relative to the sym m etry elements (186). In theory, six distinct arrangem ents can be found for a P6 crystal, three for a P4, a n d one only for P3, P2 and PI crystals. So far, P6, P4 a n d P2 sym m etry lattices have been identified. Although P I and P3 lattices are theoretically possible, they are relatively rare if they exist at all.

Biological functions o f S-layers

I

A lth o u g h a w id e ra n g e of S -lay ers h a v e b e e n id e n tifie d m orphologically, further detailed studies on the biological functions of S- layers are still at an early stage. Up to now, bacterial S-layers have been

(17)

©

Thi n s e c l i o n l l M o l e c u l a r a r c h i t e c t u r e ( f r p u r e f r a c t u r e I J , m u n m mmmmsmk

n

urn

Pro te in or g ly c o p r o t e i n C 2 > M e m b r a n e p r o te i n P e p t i d o g l y c a n ( o t h e t p b l y m e r s ) It? L ip o p ro te in Lipopoly s a c c h a r i d e P h o s p h o lip id P o r e p r o t e i n

Fig. 1. Schem atic illu stra tio n s of the b acterial envelopes containing S-layers. (a) Archaea. (b) Gram positive bacteria rand some archaea possess a pseudom urein layer, (c) Gram negative bacteria. CM: cytoplasm ic membrane; CW: rigid cell wall layer m ad e u p of p e p tid o g ly c a n ; OM: o u te r m em b ran e; PG: peptidoglycan layer; S: the surface layer. SA: additional S-layers. From Sleytr and M essner (192).

(18)

shown to function in m aintaining cell shape, blocking entrance of large size m olecules, p rev en tin g b in d in g of certain harm ful m olecules to the cell membranes, binding bacterial phages, an d m ediating cell adhesion, In some cases S-layers determ ine and m aintain cell shape or contribute to envelope rigidity (192). This w as best illustrated in H alobacterium strain s w here glycosylated S-layers are the exclusive cell w all com ponent (141). U pon blocking the glycosylation by exposing grow ing cells to bacitracin, the cell shape changed from rods to spheres (232).

The physical characteristics of the protein m eshw ork determ ines the perm eability of a particular S-layer. U ltrastructural studies have suggested the existence of channels of 2 to 6 nm diam eter (143). This pore size w o u ld not significantly affect the perm eation of sm aller nutrients and degradation products, but could block lytic enzym es and bacteriophage particles from the environm ent w hich could be harm ful to the cells (191, 192). As a protective coat, for example, the Campylobacter fetus S-layer prevents C3b binding to the

I

bacterial cells (27) and renders the cells serum and phagocytosis resistant (26). S-layers have also been show n to protect cells from bacteriophages (105) and from predation, by Bdellovibrio bacteriovorus (120), a n d also function as bacteriophage receptors (50, 95). From available data, it also appears th at S- layers can provide organism s w ith advantages in cell adhesion a n d surface recognition (192). S-layers m ediating cell adhesion have been dem onstrated am ong bacterial cells (auto-agglutination) an d betw een bacterial cells and epiderm al surfaces (55),

The bacterial S-layer is also a p ro m isin g tool for a v a rie ty of biotechnological applications. For exam ple, w ith w ell defined m olecular w eight cut-off v alu es (10,000-15,000 (143)), S-layers can be u se d as ultrafiltration m em branes. In contrast to the am o rp h o u s stru ctu re an d a

(19)

pore-size distribution of regular ultrafiltration m em branes, S-layers provide a crystalline structure and uniform pore sizes. In addition, functional groups on the Surface of the S-layers such as carboxyl, amino and hydroxyl groups can be m odified to suit various applications. These structural features could be also u seful for the im m obilization of m acrom olecules, a process which

I

traditionally em ploys am o rp h o u s polym ers w ith a random structure and binding sites. In addition to these potential in d u strial applications, S-layers also have p o ten tial in biom edical research. Because of th e ir a b u n d an t production and their outside location, S-layers could become ideal carriers for other proteins. Indeed as early as 1989, Yamagata, Udaka and their group reported that the hum an epiderm al grovyth factor (hEGF) gene could be fused to B. brevis 47 "middle wall" S-layer protein gene and successfully expressed in B. brevis 47 (235), The protein w as produced efficiently in a large am ount (0.24 g /lite r of culture), and the m ature hEGF protein w as correctly cleaved from the bacterial m iddle wall protein leader sequence and secreted.

S-layer p ro tein s

S-layers are usually composed of a single protein species and are among the m ost ab u n d an t of all prokaryotic cellular proteins. Thus far, a num ber of

! I

bacterial S-layer proteins have been purified an d biochemically characterized. The m olecular w eights of these proteins are in the range of 40 to 220 Kd (194). Some of the S-layer proteins are glycosylated, w ith a rem arkable structural diversity of the carbohydrate chains even am ong strains of the same species (142). S-layers com posed of two protein subunits, although rare, have been rep o rted (86). S-layer proteins share a num ber of com m on features. For exam ple, they contain 40 - 50% of hydrophobic am ino acids, m ore acidic amino acid residues rather than basic residues (193), an d a Very low content of

(20)

Cys a n d Met residues. In term s of am ino acid sequence, sim ilar prim ary sequences can be seen am ong strains w ith in the sam e g en u s such as

Methanothermus (M- fervidus and M. sociability (30), Cauldbacter strains (show n

by Southern hybridization) (195), Rickettsia (R, prowazekii and R. typhi) (33), and

Bacillus brevis (B. brevis HPD31 hexagonal array protein HW P an d B. brevis 47

m iddle wall protein MWP) (48). However, S-layer proteins from unrelated bacteria usually do not display sequence homology, except for a few examples such as the hom o lo g y seen b etw een the N -term in al reg io n s of the

Acetogetiium kivui S-layer protein and B. brevis 47 m iddle wall protein (165).

S-layer p ro tein genes

In contrast to their relatively simplicity w ith which S-layer proteins Can be isolated, cloning and expression of S-layer protein genes has often proven to be difficult. In som e cases, S-layer genes have h a d to he cloned on overlapping fragm ents (165). Partly due to these difficulties, an d also due to the late start of genetic research on S-layers, studies of S-layer structural genes are still at a quite prelim inary stage. Udaka an d their coworkers cloned the first S-layer protein gene from Bacillus brevis 47 in 1984 (217), How ever they were not able to clone the prom oter region into E, coli, an d were subsequently forced to use Bacillus subtilis as the host cell (234). Since this first cloning of an S p ro tein gene, a num ber of S-protein genes have been cloned and sequenced. SO far, the S-layer protein genes being sequenced include Bacillus brevis 47 (48, 214, 216), Rickettsia rickettsii (76), Bacillus sphctericus (29), Acetogentum kivui (166),

Bacillus brevis HPD31 (48), Halobacterium halobium (128), Haloferax volcanii (201), M ethanothermus fervidus and Methanothermus lociabilis (30), Deinococcus radiodutans (164, 166), Campylobacter fetus (25), Rickettsia prowazekii (33), Thermus thermophilus (56), Lactobacillus brevis (222) and Caulobacter crescentus (75). From

(21)

these studies it appears that S-layer proteins are usually encoded by single copy genes (195) and transcribed as monocistronic units (59, 222). In the case of C. fetu s, w here S-layer protein antigenic variability is displayed, there appear to be several copies of the gene, only one of w h k h is expressed at a time (M. Blaser, personal com m unication). B. brevis is another exception. This organism has a polycistronic operon (cwp operon) since two different S- layer p ro tein genes coding for o uter and m iddle w all proteins (OWP and MWP) are expressed in this species (236). A lthough OWP and MWP are c o tra n sc rib e d from the sam e o p ero n , tra n sc rip tio n a l term in ato r-lik e sequences w ere found follow ing each of these genes (216). The monocistronic nature o f other S-layer genes is also supported by the fact that transcriptional term inator-like sequences have been found im m ediately follow ing the genes (25,29/ 48, 75,128,166, 222). For example, TV thermophilus HB8 S-layer gene has a n 11-base-long inverted repeat w ith a predicted free energy value of -24.2 k C a l/m o l follow ed by a T-rich stretch 15 nucleotides dow nstream from the translational stop codon (56).

In m ost of the S-layer genes, a region coding for a protein lead er sequence has been identified at the 5' end. In the case of B. brevis 47 m iddle w all p ro tein gene, tw o tandem ly located translation initiation sites were found (3). T ran slatio n in itiatio n from these tw o sites resu lted in tw o different leader sequences w ith the length of 54 residues and 23 residues respectively. H ow ever, the, sequence of the m ature protein rem ained the same since the precursors w ere cleaved at the same site. In the longer precursor, the initiation codon TTG w as found to be efficiently used to code for a M et residue. The tandem translation initiation Sites were also found in other B. brevis strains (48). Several S-layer proteins do not have a cleaved leader sequence. Examples include Campylobacter fetus strain 84-32 (23D) (25),

(22)

Caulobacter crescentus (59, 75), and Rickettsia prowazekii (33). In the latter case,

the C -term inal region of the p ro tein w as p ro p o sed to h av e a potential transm em brane sequence which w as cleaved after translocation.

S-layer p ro tein gene prom oters

As S-layer proteins are a b u n d an tly produced, S-layer p ro tein gene prom oters are generally considered! to be efficient prom oters. H ow ever, only a few S-layer protein gene prom oters have been identified. In m any cases,

j'

only a single prom oter appears to be present (56, 59, 218). In other cases, S- layer protein genes have been Shown to have more than one prom oter (2,

222)., In the case of B. brevis 47 w hich contains tw o S-layer protein genes coding for OWP and MWP) Adachi, U daka and their coworkers found several tandem ly arranged prom oters in the 5' region of the cotranscriptional u n it in the cwp operon (2). Two of the m ost active of these prom oters, P2 a n d P3, displayed different transcriptional activities at different stages of cell growth. This prom oter containing sequence also contains a 12-nucleotide palindrom ic sequjenee dow nstream from an infrequently used prom oter P I (234). This complex structure in the prom oter region w as found to be highly conserved am ong HW P (S-layer protein) genes from B. brevis strains HPD31, HPD52 and HP033, suggesting th at the synthesis of the S-layer proteins is intricately regulated (48). In the case of Lactobacillus brevis, tw o p ro m o ters direct transcription of the S-layer protein gene (222). These tw o prom oters function w ith roughly equal efficiency in exponential grow th phase. In the case of C.

crescentus w hich undergoes cell differentiation, S-layer gene transcription

appeared not to be subject to developm ental regulation (59). In the case of cloned S-layer protein gene prom oters, the T. thermophilus HB8 S-layer gene p ro m o ter w as recognized by E. coli RNA polym erase (56), w hile th e C.

(23)

crescentus prom oter was not recognized in E. cdli although the sequence of the

p rom oter w as reasonably sim ilar to the E, coli a70 p ro m o te r consensus sequence (59). Fisher et al. proposed that transcription of the adjacent region or auxiliary transcriptional factors m ay have some effects on the transcription of the Caulobacter S-layer gene (59).

A erom onas salm onicida

The genus A ero m o n a s

A erom onads are Gram negative rod shaped bacteria that are native to aquatic an d soil environm ents w orld w ide (32). M any of them also cause diseases am ong both w arm - an d cold-blooded animals (106, 209). The genus

Aeromonas w as classified in the fam ily Vibriondceae in the latest Bergey's

M anual (172), how ever a new family Aeromonadacete has also been proposed (39). U sing D N A -D N A reassociation kinetics a n d m ultilocus enzym e electrophoresis analysis, Janda et al. have identified at least 12 genospecies or hybridization g roups (HGl-12) in this family: HG1 - Aeromonas hydrophila, HG2 - unnam ed, HG3 - Aeromonas salmonicida, HG4 - Aeromonas caviae, HG5 -

Aeromonas media, HG6 - Aeromonas eucrenophila, HG7 - Aeromonas sobria, H G8 -

Aeromonas veronii biotype sobria, HG9 - Aeromonas jandaei, HG10 - Aeromonas veronii, HG11 - unnam ed, and HG12 - Aeromonas schubertii (107).

W ith the exception of HG3 - A . salmonicida, the aerom onads are m esophilic and motile. A num ber of the motile aerom onads are associated w ith h u m a n in fectio n s such as septicem ia, m eningitis, osteom yelitis, gastroenteritis, a n d w o u n d infections am ong pediatric and adult populations, both im m unocom prom ised and otherwise healthy individuals (61, 224). The five im plicated in hum an diseases are: A, hydrophila, A, veronii biotype sobria,

(24)

well (149). The psychrophilic non-motile group HG3 - A. salmonicida has also been show n to cause fish diseases (105, 209). A com m on feature shared by disease isolates in both mesophilic and psychrophilic groups is the presence of an S-layer (108, 111, 116, 219). In the case of A. salmonicida the S-layer has been dem onstrated to be an im portant virulence factor (105).

The species

The first definitive isolation of Aeromonas salmonicida w as from brow n trout in G erm any by Em m erich an d W eibel in 1894 (51). Initially, it was thought to be the causative agent for fish furunculosis in hatcheries only. But studies by Plehn show ed that the disease w as w idely prevalent in Bavarian trout strains (171). At the same time, a num ber of investigations show ed that the disease was w idely spread out in m any European countries (63, 144, 170). The first description of its occurrence in the Am ericas w as b y M arsh in M ichigan hatcheries in 1902 (134). It was later found in w ild salm onids in B.C. (47) and also other regions in Americas. N ow , the occurrence of A.

salmonicida has been reported virtually all over the w orld, including Australia

and Asia.

W hen A. salmonicida w as first isolated, it w as called B a cteriu m

sali^ionicida (51). In 1953, Griffin et al (82) proposed th at the species B. salmonicida be in clu d ed in the Genus Aeromonas as Aeromonas salmonicida.

Since then, although argum ents have persisted on the classification of the species, the organism has rem ained in G enus Aeromonas. Bergey's M anual describes three subspecies of A, salmonicida (172). Subspecies salmonicida p ro d u ce s a b ro w n p ig m e n t on m edia c o n ta in in g 0.1% ty ro sin e or phenylalanine, does n o t produce indole, an d does hydrolyze esculin and ferm ent mannlioL Subspecies achfomogenes does not produce the brow n

(25)

pigm ent, may produce indole, and does not hydrolyze esculin and feunenr m annitol. Subspecies masoudda does not produce the brow n pigment, does produce indole, hydrolyze esculin, and ferm ent m annitol (172), For w orking purposes, the subspeciesi salmonicida and achromogenes are term ed "typical" A,

salmonicida strains a n d cause furunculosis in salm onids. H ow ever, there

have b een a n u m b er of other reports concerning "atypical" A , salmonicida strains causing diseases in other fishes ouch as goldfish, m innow s, carp, as well as salm onids. This g ro u p of stra in s u su ally exhibits fastidious nutritional requirem ents, slow grow th and distinctive biochemical properties (103). The diseases are often chronic and inflam m atory, and can involve surface ulceration and erythroderm atitis (28, 89, 160, 190, 197), In the late 70's, M cCarthy undertook a comprehensive analysis in w hich 29 "atypical" strains were co m p ared to 145 o th er bacteria, p rim arily "typical" strain s of A ,

salmonicida a n d v ario u s representatives of the m otile A erom onads (138),

From h is phene :ypic an d genotypic analysis, he proposed that the species A,

salmonicida include three subspecies. The first subspecies com prised the

"b pical" strains of A salmonicida isolated from sahnonid fish species. The second subspecies w as also restricted to isolates from salm onid fish and retained the nam e achromogenes by historical precedent. The third subspecies proposed w as subspecies nova, which was retained for A, salmonicida isolates from non-salm onid sources. A DNA-DNA hybridization study by Belland and T rust (22) confirm ed McCarthy's proposal.

P a th o g e n e sis

A . salmonicida produces fish diseases in freshw ater, estuarine, and

m arine environm ents (209). The diseases Caused by this organism in the sea are often related to the establishes >nt of an asym ptom atic carrier state during

(26)

the fishes' early grow th in fresh w ater. The laten t infection may n o t be detectable by culture m ethods, so corticosteroid injections and heat stress are com m only em ployed to prom ote acute infection (31). Studies using this technique have show n th at laten t infection w ith A . salmonicida can be common in salmonids. Furtherm ore, strains isolated from carrier fish show no reduced virulence. The organism appears to be carried in the kidney and later becomes active w hen fish go to sea or are stressed in some fashion (209). A lthough the latent state of the disease is not clearly understood, McIntosh and A ustin suggested th at L-forms m ay play a role (139). A fter Atlantic salmon were injected w ith a suspension of A. salmonicida L-forms, no clinical signs o f disease were observed, and parental type or L-form colonies were not recovered. However, microscopic exam ination revealed possible L-form cells rem aining w ithin the iish tissues, particularly in the kidney (139).

There h av e been a n u m b er o f virulence factors re p o rte d fo r A .

salmonicida. Am ong them , the best described is the surface protein array of A. salmonicida know n as A-layer, which will be discussed in detail later in this

dissertation. O ther virulence factors include proteases, glycerophospholipid- cholesterol acyltrartsferase (GCAT), as well as leukocytolytic and hemolytic activities. The proteinases an d hemolysins identified include a 70 Kd serine proteinase, a 56 Kd hem olysin (66), a 25.9 Kd H -lysin (broad-spectrum hemolysis) (206), a T-lysin (lysis of tro u t erythrocytes) (205), an d a 200 Kd salmolysin (lysis of salmonid erythrocytes) (154)

W hen injected intram uscularly, the 70 Kd serine proteinase produced a lesion histologically sim ilar to n atu ral infections (65). The gene (aspA) has been cloned and sequenced. The predicted am ino acid sequence indicated that the protein is of 64,173 Dalton, and has a 24-amino acid residue signal peptide and the NGTS consensus sequence of serine proteinases (231). Further

(27)

evidence for a role for protease as a virulence factor in the pathogenesis of fish furunculosis has been provided by Sakai (181). Sakai found that although the protease-deficient m utant NTG-1, w hich was induced w ith

N-methyl-N'-! : i

nitro -N -n itro so g u an id in e, rem aine au to ag g lu tin atiy e, hem agglutinative, serum resistant, adhesiye, hem olysin positive, an d leukocytolysin positive, its LD50 in sockeye salmon a n d rainbow trout w as increased to more than 10® as opposed to lO^-lO^ of its virulent parent strain A-7301. NTG-1 strains were also elim inated from rainbow trout in a Short time. In other studies, Lee and Ellis rep o rted th at purified A. salmonicida extracellular protease w as lethal to Atlantic salm on (Salmo salar L.) w hen intraperitoneally injected (129). The LD50 w as 2,400 n g /g fish.

In the case of hem olytic activities, Titball and M unn purified the H- lysin a n d show ed th at it h a d horse erythrocyte lysis activity an d GCAT activity. It also exhibited cytotoxicity w ith rainbow tro u t gonad cells and rain b o w tro u t leukocytes. H ow ever, w h en the H -lysin w as injected intravenously into rainbow trout, no pathological effect w as observed (206), Fyfe et al. show ed that the 56 Kd hemolysin had hemolytic activity w ith trout erythrocytes, although fhey d id not test for lysis of horse erythrocytes (66). In other studies, N om ura et al. purified the heat labile salm olysin from the culture su p ern atan t an d show ed that it w as a glycoprotein w hich lacked p ro te a se a ctiv ity , b u t w as leth al to rain b o w tro u t w h e n in jected intram uscularly (154).

A GCAT/LPS complex w as isolated and purified from A. salmonicida extracellular p ro d u cts (130). This com plex w as lethal for A tlantic salm on

(Salmo salar L.) by injection of 0.045 m g p er g of body weight. The virulence

m echanism s of th is 2,000 K d com plex m ay in clu d e its hem olytic, leukocytolytic an d other cytotoxic activities. The LPS w as found to play an

(28)

active role in the complex. The specific hemolytic activity a n d lethal toxicity of thei complex w as about eight tim es higher than free GCAT. Also, the complex w as m ore resistant to proteolytic and heat inactivation th an free GCAT (130). Lee and Ellis also found ah additive relationship betw een purified protease and the hem olysin com plex w h en injected in A tlantic salmon by the intraperitoneal route (129).

Thornton et al reported tw o attenuated m utants that possessed A-layer and w ere avirulent to salm onid fish (2P4). These m utants w ere deficient in aerobic m etabolism . In addition, although the cell surface com position of these m u ta n ts w ere v e ry sim ilar to w ild ty p e strain s, th eir A -layer architecture w as drastically altered. Electromicroscopic exam ination indicated that the A-layer in these strains m ight be m ultilayered or aggregated rather than a m onom olecular layer. This may indicate that the native organization of the A-layer is essential for virulence.

A -layer

A . salmonicida has the capacity to produce a paracrystalline surface-

protein array (Fig. 2) (111, 219), know n as the A-layer, w hich appears to be essential to virulence (105). Its location external to the outer m em brane w as dem onstrated b y thin section electrom icroscopy (105, 210) an d cell surface labeling using 125ID ISA (diazotized (125I) iodosulfanilic acid) (111).

M o rp h o lo g y

The ultrastructure of the A-layer is best observed w h en the layer has sloughed off the cell surface d u rin g p rep aratio n of sam ples for negative staining and electron microscopy. Early studies show ed that A-layer displayed P4 sym m etry (105, 210). Two tetragonal patterns/ type I a n d type II, w ere s u b s e q u e n tly o b s e r v e d b y im a g e e n h a n c e m e n t te c h n iq u e s in

(29)

Fig. 2. A: Negatively stained electron m icrograph of A450 cell an d sloughed A-layer sheets (X 150,000). B: N egatively stained electron m icrograph of a sloughed A -layer sheet from A450 (X337,000) (Courtesy of R. Garduno).

(30)

two-dim ensional (2-D) m ass distribution projections of negatively stained A- layer preparations (198). Dooley et al. later show ed that the type I pattern was the resu lt of tw o layers in a back to back arrangem ent, w hile projection th ro u g h the single layer gave the type II p attern (44). U sing com puter- sim ulated superim position, G arduno arid Kay recently confirm ed th at the type I p attern s were indeed form ed from superim posed A-layers d u rin g preparation for negative staining (70). Three dim ensional reconstruction of the A-layer (44) show ed a lattice constant of 12.5 rim, an d a major tetragon and a m inor tetragon w ere observed at a resolution of 1.6 nm . The major tetragon contains the four major dom ains of four subunits, an d forms a large depression tow ards the inside of the layer. The m inor tetragon is located tow ard the outside of the layer and has been proposed to provide connectivity w ithin the layer. This structure provides the surface of the layer w ith a certain am ount of three dim ensional architecture. N orm al organization of the A -layer seem s to require the presence of Ca++, since C a++ lim itation resulted in a sequence of structural rearrangem ents in the A-layer (72).

Subunits

The A-protein has been identified as the protein subunit of the A-layer and has been isolated, purified, and characterized biochemically (111, 169). It is the m ajor p ro tein com ponent in o uter m em brane fractions of A -layer producing strains. The apparent molecular weight of A450 A-protein is 48,000 to 53/000 as determ ined by SDS-PAGE. Early studies show ed that A -protein had several isoelectric forms w ith pis of 4.8-5.3, however, later studies show ed that the native protein on the surface of m id-exponential phase cells h ad a single p i corresponding to the m ost basic value in the previously observed pi range (167). Am ino acid com position stu d ies sh o w ed th a t A -protein contained 45% non polar residues (Val, Met, lie, Leu, Ala, Phe, Trp, and Pro)

(31)

and the N -term inal sequence w as show n to have very few charged residues. O ther studies show ed th at A-proteins from various A, salmonicida strains had conserved biochem ical p ro p erties such as m olecular w eight, am ino acid composition, N-term inal sequence, pi, chym otrypsin digestion sensitivity and im m unological reactivity (167).

A-laver formation and anchoring

In a transposon m utagenesis study using A. salmonicida A449, Belland an d T rust d em o n strated th at tw o Tn5 insertion m utants, TM1 a n d TM2, accum ulated the m ajority of their A-protein in the periplasm (20), suggesting that A -protein w as translocated through the periplasm ic space, a n d that at least one gene product w as required for A-protein export. Prelim inary studies indicate th a t the Tn5 insertions in m utants TM1 and TM2 are in different loci. The location of the Tn5 insertion in TM2 has been recently m apped relative to the A -protein gene (155), an d the DNA sequence of the region revealed a n open read in g fram e coding for a putative protein w ith high sequence sim ilarity to another bacterial transport protein (Noonan, personal com m unication).

A nother Tn5 insertion m u tan t A . salmonicida A449-TM5 excreted A- p ro tein in to the cu ltu re su p ern atan t in quantities approaching 1 mg A- protein per m l (20). W hen the LPS of this m u tan t was exam ined b y silver staining of SDS-PAGE, O -polysaccharide chains w ere show n to be absent. Electron m icroscopy show ed how ever that A -protein h ad assem bled into sheets. These findings suggest th at the anchoring of A-layer to the cell surface requires intact LPS O-chains. This hypothesis was su p p o rted by the evidence th at m any O-polysaccharide chains penetrated the A-layer an d were exposed on the cell surface (36) and th at A-protein could be reattached to the

(32)

V iru len ce

Among all of the bacterial S-iayers studied to date, the A , salmonicida A- layer is the best described in term s of biological activities. E arly studies by Ishiguro and colleagues dem onstrated the im portance of the A-layer in the virulence of A. salmonicida (105). Most cells in a culture grow n at higher than optim um tem peratures (>26°C) w ould die, and the rem aining cells w ere A- and avirulent. The LD50 for coho salm on by the intraperitoneal route was increased frorh 8 X 10? cfu per 7 g fish for the w ild type strain to 1 X 10? cfu per 7 g fish for the attenuated strains. In another experim ent the virulence of the single-site Tn5 insertion periplasm ic accum ulating m u ta n t TM4, i.e., an isogenic A-layer m utant (20), w as com pared w ith its parent strain. W hile the LD50 for the parent A449 strain was <10? per fish, the m utant w as avirulent (209).

The role of A-layer in virulence w as further illustrated b y following the fate of injected A + ah d A" cells in fish tissues (209). In tram u scu lar injection pf 10^ v iru len t A+ cells into fish w as follow ed by a very early bacterem ia an d extrem ely rap id accum ulation of organism s in the spleen, kidney/ and liver. The viable A + cells in each organ m ultiplied to very high levels, the tissues became necrotic after 72 h, an d all fish died 72-96 h post­ infection. Septicem ia w as first seen 24 h post-infection. Injection of an identical dose of high-grow th tem perature derived A" cells was also followed by art initial accum ulation in the spleen, kidney, liver, and heart. By 48 h post-infection, however, the fish had cleared itself of viable A" cells, a n d all fish survived. W hen fish were injected w ith as m any as 10^ A" cells, the cells persisted in the organs in high num bers for several days, b u t w ere rapidly cleared after 96 h; all fish survived. In neither experim ent w ith A ' cells were bacteria recovered from the blood (209).

(33)

The role of the A-layer in pathogenesis wap also investigated in serum resistance experim ents (150). W ith both non im m une an d im m une fish and rabbit serum , Munri an d coworkers compared w ild type A. salmonicida strains w ith h ig h tem perature m utants and spontaneous m utants that lack A-layer or bo th A -layer arid LPS O-chains, W hile A + w ild type strains show ed resistance to all sera applied, A" strains, regardless of their LPS, dem onstrated significantly less resistance. This indicated that A-layer contributed protection against the bactericidal activities of both im m une and non im m une sera.

O th e r biological functions

The A -layer has been d em o n strated to b in d to a w ide v ariety of substrates including m acrophages (211) and a num ber of molecules including Congo red (104), p o rp h y rin s (113), im m unoglobulin (168), fibronectin and laminin (41), an d collagen type IV (212).

Interaction betw een Aeromonas cells and m acrophages has been Well docum ented w ith a num ber of cell lines including resident peritoneal marine m acro p h ag es, rain b o w tro u t h e ad k id n ey tissue m acro p h ag es (211), m am m alian cell line BHK-21, salm onid cell line RTG-2 (157, 158), peritoneal elicited-exudate cells from rainbow tro u t and coho or m asu salm on (180), elicited o r activated peritoneal m acrophages from brook trout (156), rainbow trout blood leukocytes (110), salm onid cell lines (182), m urine m acrophage cell line P388D1 (71) and rainbow trout m acrophages (69). Recent studies by G a rd u n o a n d K ay u s in g m u rin e m ac ro p h ag e s a n d rain b o w tro u t m acrophages show ed th at only A+ bacterial cells were readily internalized in phosphate-buffered saline in the absence of opsonins (69, 71). In addition, A~ layer coated latex beads w ere m ore efficiently phagocytosed than A -protein coated beads, w hile uncoated bead s w ere n o t taken u p by m acrophages,

(34)

indicating that A-layer m ediated the phagocytosis of both A+ cells and A-layer coated beads.

, Although the A-layer appears to m ediate phagocytosis, it is probably not cytotoxic. W hile A + cells w ere show n to b e m ore cytotoxic to m acrophages than A" cells, A-protein and A-layer coated latex beads failed to cause m orphological changes in macrophages w hen phagocytosed (71).

The A -protein subunit has been show n to bind Congo red (104). On

1

Congo red containing solid culture media, red-colored A-layer* colonies can be readily distinguished from w hite or light orange-colored A-layer- colonies. Two classes of binding have been observed. W hen high concentrations of Congo red were used, a nonspecific hydrophobic interaction betw een Congo red an d A-protein was dem onstrated at a dye-to-protein m olar ratio of about 8 - 30. This nonspecific b in d in g co u ld b e en h an ced b y h ig h e r salt concentrations (104, 113). Specific binding w as observed at low Congo red c o n c e n tra tio n s a n d th is b in d in g w as c o m p e titiv e ly in h ib ite d b y protoporphyrin IX and hem in and vice versa (113). Based on this competitive inhibition and the sim ilar structures of Congo red a n d porphyrins, it w as hypothesized that a hydrophobic porphyrin binding dom ain exists in the A- protein which could act as a nucleation site for fu rth er b in d in g a t higher Congo red concentrations. The porphyrin binding m ay represent a n initial stage in iron transport.

A + cells of A. salmonicida also b in d rab b it IgG a n d h u m an IgM specifically (168). In contrast to Staphylococcus aureus protein A, w hich binds im m u n o g lo b u lin Fc fra g m e n ts , b in d in g o f n a tiv e A -la y e r to im m unoglobulins req u ired structurally intact IgG. In addition, a special m olecular arrangem ent o f A -protein m onom ers in a native A -layer w as required for the binding since purified A-protein b ound IgG only w eakly and

(35)

reconstituted array on A-layer negative cell surfaces d id not bind IgG. The im m unoglobulin b in d in g by A-layer w as show n to be saturable, b u t there was no single easily definable binding site on the A-protein m onom ers. Indeed one IgG m olecule w as estim ated to bind four or five A -protein monomers, leading to the speculation th at a four fold sym m etric p it in the A-layer forrried by four m onom ers m ay act as a specific trap for im m unoglobulin m olecules. This im m u n o g lo b u lin b in d in g a ctiv ity m ay serve as a m echanism to shield the bacterium from host im m une defenses (168).

W hile im m u noglobulin b in d in g req u ired an intact supram olecular array an d could not be inhibited by A-protein, the binding of the extracellular m atrix proteins lam inin and fibronectin w as significantly inhibited by A- p ro te in an d its 37.6K N -term inal m ajor try p sin -resistan t dom ain. This suggests that the b inding sites for these m atrix proteins are localized in this segm ent of A -protein (41). Since fibronectin is p resen t in body fluids including blood plasm a, it w as suggested th at binding of this protein to the surface of the bacterium via the A-layer could block the host's im m une response to the bacterium by sterically m asking im m unogenic epitopes. The b in d in g m ight also facilitate adhesion of the bacterium to host cells such as m acrophages. In ad d itio n , lam inin is a m ajor com ponent on basem ent m ^mbranes, an d the binding of this glycoprotein m ight facilitate binding of A.

salmonicida to basem ent m em branes in ulcerated regions. A nother basem ent

m em brane protein, collagen type IV, has also been reported to bind to the A- layer, a n d the b in d in g of this pro tein w as rapid, specific, saturable, and essentially irreversible. This m ight ad d to the ability of the bacterium to bind the basem ent m em branes of ulcerated regions (212).

(36)

The S-layers of mesophilic aerom pnads have been show n to have very sim ilar m orphology to the A-layer of A . salmonicida (116, 151). The best described is th at of the A. hydrophila strain TF7, This S-layer show s a lattice constant of 12.2 nm and two structural units, a major tetragon and a lesser tetragon (151, Dooley, 1989 #91). It has an ap p aren t m olecular w eight of 52,0C£, and a m easured p i of 4.6 (43), sim ilar to A . salmonicida A -protein. However; differences exist in the prim ary structures and antigenicity of these S-layer proteins (119). Also, in contrast to A-protein w hich is antigenically conserved, th e S-layer pro tein s from m esophilic aerom onads d isp lay ed considerable antigenic variability (119). Furtherm ore, unlike A-layer, the A .

hydrophila TF7 S-layer apparently does not possess a w ide range of binding

activities, and its role in fish pathogenesis rem ains undefined. Recently, however, a Tn5 transposon insertion m u tan t w as isolated in A. hydrophila TF7 w hich p roduced a truncated S-layer protein, which d id not form an S- layer (202). This m u tan t w as show n to be m ore sensitive to im m une trout Serum and have a 5-fold higher LDgo in rainbow trout com pared to the w ild type strain, indicating a possible role of the S-layer in pathogenesis (Thomas, personal com m unication).

L ipopolysaccharides (LPS)

| Early reports suggested that A, salmonicida LPS had no toxic effects on salm onid fish th ro u g h injection of u p to 0.714 g /k g of fish (161, 230). H ow ever, serum resistance experim ents have show n th at the LPS is a virulence factor (150). W hen A“ 0 + (A -layer negative, O -polysaccharide chains positive) a n d A" O - m u ta n ts w ere exam ined in se ru m k illing experiments, O-chains were show n to contribute to the serum resistance of A .

(37)

salmonicida, especially in norm al sera. Com parison w ith A+ and A" strains

suggested that O-chains were the major contributors to serum resistance. The s tru c tu ra l m o rp h o lo g y of LPS from v a rio u s strain s of A .

salmonicida has been analyzed by intrinsic ^^P -rad io lab elin g an d silver

' |

staining by C hart et al. (36). The high-m olecular-weight fraction of LPS was resolved into a small num ber of distinct bands, indicating a very high degree

1

.

of hom ogeneity in O-polysaccharide chain length. These LPS O-chains from

L.

diverse strain s are antigenically cross-r re active. Their m onosaccharide com position is also v ery sim ilar (36). W herj co m p ared w ith S-layer p ro d u c in g m eso p h ilic a ero m o n ad s, D ooley et al. o b se rv ed th a t O- polysaccharide chains w ere longer in A. salmonicida, and there were common epitopes and species-specific epitopes in the tw o species (45). Chart et al. also show ed that se me LPS O-polysaccharide penetrated to the outer surface of the |

i

A-layer (36). These exposed O-polysaccharide still functioned as phage 55R-1 j

receptors, although the phage adsorption to an A+ strain was slower than that to an A“ strain. This w as also confirm ed b y positive reaction of anti-O- p o ly sa c c h a rid e rn cn o c lo n al a n tib o d y w ith A + cells (36). A sim ilar j

arrangem ent of LPS an d S-layer w as also observed in A. hydrophila (45). Because of th eir surface exposure, the O-chains were strongly im m unogenic and ap p eared to be antigenically conserved (36). However, w hile both LPS and A-protein prom oted antibody response in fish (35, 161), neither served as a protective antigen (209).

Purpose of this dissertation

Infectious diseases of fish are a w orld wide problem . Today, as the

I

w orld population increases rapidly and people are m ore and more concerned about the preservation of o u r lim ited natural resources, more fish farm s are

(38)

required to provide daily source of fish for the ever increasing w orld dem and. As a result, infectious diseases in fish farm s become a major problem in this intensive culture practice, and often lead to serious economic consequences. Because of this economic significance, the control of fish disease is urgently needed.

Aeromonas salmonicida is a fish pathogen of m ajor im portance. Its

"typical" and "atypical" strains cause diseases in various fishes, and have been isolated w orld w ide. A lthough a num ber of virulence factors have been suggested for A. salmonicida, the best described virulence factor is the surface protein layer A-layer. The A-layer of A . salmonicida has been show n to be involved in fish pathogenesis using an isogenic m utant in serum resistance an d fish injection experim ents. W hile detailed structures and biological functions of m ost other A. salmonicida virulence factors are still undefined, studies of the A. salmonicida A-layer have revealed a num ber of biological

i

activities w hich are relevant to fish pathogenesis. In addition, this surface layer has provided a good exam ple for the study of bacterial surface layers which have been found in h u ndreds of bacteria and archaea. The S-layers com prise up to 15% of the cellular protein an d can be efficiently exported through cell m em branes and assem bled to cover the entire cell surface. The S-layers usually constitute the outerm ost layer of the cell envelope an d directly interact w ith various molecules, cells, and environm ental conditions, as well as the underneath cell surface. Indeed, A, salmonicida A-layer is one of th e b e st stu d ie d b a cterial S-layers w ith k n o w n th ree d im e n sio n al ultfastructure an d a num ber of defined biological functions. Therefore, the im pact of the stu d y o n A. salmonicida A -layer is tw o-fold, it involves econom ically im p o rtan t fish pathogenesis a n d the biologically im p o rtan t bacterial S-layers.

(39)

M olecular biological characterization of the gene for the A -layer su b u n it A -p ro tein is essential fo r u n d e rsta n d in g the gene stru ctu re, expression an d regulation. It will also provide inform ation on the protein's p rim ary and tertiary structure, biological activity, a n d location of their corresponding dom ains. How ever, prior to this study, m ost of the w ork on A-protein dealt w ith biochemical, structural and functional characterizations, A few pioneer genetic studies by Bellanc f20, 21) and Phipps (167) not only refle cted th e needs, b u t also p ro v id e d a solid g ro u n d , for fu rth e r investigations. Following the cloning of the A. salmonicida A-protein gene, it became possible to carry on the detailed characterization and analysis of the gene. This stu d y w as therefore aim ed at u n d erstan d in g the structure, expression a n d regulation of the A -protein and its gene at the m olecular genetic level. The results described in this dissertation are organized into

I

four m ajor groups, i.e., stru ctu ral characterization of th e A -protein gene, transcriptional studies on the A -protein gene expression, native insertion elem ents w hich affect the A -protein gene expression in A, salmonicida, and a structural and functional description of a dow nstream gene w hich may play a role in A p ro tein gene expression.

(40)

MATERIALS AND METHODS

Bacterial strains, vectors and m edia

A, salmonicida strains from the culture collection of T. J. T rust are listed

in Table 1. E. coli strains used were D H 5a for recom binant plasm ids (87) and JM109 for MI3 recom binants (237). All strains were grow n in L-broth or on L- agar (145). Wild type A. salmonicida strains were grown at 20°C. Strain A450-3 and other 30°C m utant A. salmonicida strains w ere grow n at 20°C or 30°C. E.

coli strains w ith plasm ids w ere grow n at 37°C w ith antibiotics. Antibiotics

used included ampicillin (50 m g/L ), kanam ycin (50 m g /L ) and streptom ycin (100 m g/L ). W hen required, CaCl2 or MgCl2 (1 - 30 mM) in the presence and absence of ethylene glycol-bis (p-aminoethyl ether) N -N -N ’-N'-tetraacetic acid (EGTA, 5 and 10 mM, Sigma Chemical Co., St. Louis, MO), or FeCl3 (0.01 -1.00 inM) in the presence an d absence of 2, 2'-dipyridyl (0.01 an d 0.10 mM, Sigma), w as added. Vectors used including plasm id and phage vectors are listed in Table 2. W hen anaerobic incubation w as required, a BBL GasPak w as used in a BBL G asPak jar (BBL M icrobiology Systems, Cockeysville, MD) w ith shaking.

DN A tech n iq u es D N A p reparation

A erom onas chrom osom al DNA

A 50 ml-Culture of Aeromonas cells w as centrifuged at 5,000 X g for 10 min an d the cell pellet was w ashed once in 20 ml of saline-EDTA (0.75% NaCl (v /v ) an d 5 mM EDTA). The cells w ere resuspended in 4.25 ml o f saline- EDTA a n d SDS was ad d ed to a final concentration of 3% (w /v ). This cell su sp en sio n w a s th en h e a te d to 50° C fo r 25 m in w ith occasional

(41)

Table 1. Aeromonas strains used in this study

Species Strain Sc urce S-layer

A., salmonicida i

typicals A185 Canada ii 4 .

A202 Japan

-A203 Japan

-A251 Atlantic salmon, U.K.

-A362 Atlantic salmon, U.K. ; +

A395 U.K.

-A440 Brook trout, U.S.A. _

A447 U.K. ~

A449 Brown trout, France +

A450 Brown trout, France +

A451 Rainbow trout, France +

A488 Brook trout, U.S.A. +

A500 Atlantic salmon, U.K. +

A591 Ganada ; +

atypicals A206 Japan

-A400 Goldfish, A ustralia +

A401 Goldfish, A ustralia

-A419 Goldfish, U.S.A. + 1

A438 Coho salmon, Ganada

-A460 Atlantic salmon +

A461 Atlantic salmon, C anada : j - .

A477 Carp, The N etherlands +

A480 Carp, The N etherlands i +

m utants TM1 Tn5 insertion m utant of A449 i L TM4 Tn5 insertion m utant of A449 f *

GEG4 30°C m utant of A449 i

-GEG6 30°C m utant of A450 km

GEG7 30°C m utant of A450 f

-CEG22 30°C m utant of A449 L

CEG24 30°C m utant of A449 i u i

i r i

CEG27 30°C m utant of A450 r !

GEG28 30°C m utant of A450 ! 1

-CEG29 30°C m utant of A450 r

A450-2 30°C m utant of A450 1 - ;

A450-3 30°C m utant of A450 L

i i

A. hydrophila i

! 1

TF7 Trout, Canada

i t !

A274 Sloth, Australia +

A, veronii

biotype sobria

A700 H um an diarrhea +

(42)

Table 2. Vectors used in this study i N am e Type Reference pUC18/19 Plasm id (237) pTZ18R/19R Plasm id (140) pMC1871 Plasm id (188) pMMB67EH / HE Plasm id (64) pDSK19 Plasm id , (114) pLG338 Plasm id (199) pSU2718 Plasm id (135) M 13m pl8/19 Phage (220) Xgtll Phage (238)

Referenties

GERELATEERDE DOCUMENTEN

This paper aims to investigate the lead-lag relationship between the stock index and stock index futures, and find out whether the spot market leads the futures market, whether the

Fig. Very high-energy γ-ray spectrum of the region centred on the position of HESS J1746−285, fitted with the sum of two power laws. The GC ridge contribution is fixed and the

Onze ervaring leert ons dat velen dat niet weten en juist daarom is onderzoek zoals dat onder andere door SWOV wordt uitgevoerd naar deze kosten en de oorzaken daar- van zo

De mens wordt niet meer bepaald door waar hij geboren is, maar ontwikkelt zich in vrijheid tot wat hij zelf wenst te worden.. De menselijke identiteit is geen werk van God, maar

This thesis demonstrates two key patterns of four presidential administrations (Bush I, Bush II, Obama I, Obama II), with a focus on the political discourse that constructed US Arctic

Songs, group narrative, and sign play can quickly be ruled out as potential categories where visual vernacular could reside, because visual vernacular does not have the

Montenegro is a typical case since it chose cooperation by joining NATO in 2017 (North Atlantic Treaty Organization, 2017), being small state in Europe that joined NATO after

In this paper, we select Monte Carlo method to forecast the VaR of log return series over the equally weighted European stock portfolio consisting of the CAC 40 stock index and FTSE