Louisiana State University LSU Digital Commons
LSU Historical Dissertations and Theses Graduate School
1992 Edwardsiella Ictaluri: Pathogenic Mechanisms and Antigens Recognized by Infected Channel Catfish. Thomas James Baldwin Louisiana State University and Agricultural & Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses
Recommended Citation Baldwin, Thomas James, "Edwardsiella Ictaluri: Pathogenic Mechanisms and Antigens Recognized by Infected Channel Catfish." (1992). LSU Historical Dissertations and Theses. 5421. https://digitalcommons.lsu.edu/gradschool_disstheses/5421
This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 9316952
Edwardsie.Ua ictalurix Pathogenic mechanisms and antigens recognized by infected channel catfish
Baldwin, Thomas James, Ph.D.
The Louisiana State University and Agricultural and Mechanical Col., 1992
UMI 300 N. ZeebRd. Ann Arbor, MI 48106 EDWARDSIELLA ICTALURI: PATHOGENIC MECHANISMS AND ANTIGENS RECOGNIZED BY INFECTED CHANNEL CATFISH
A Dissertation
Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy
in
The School of Veterinary Medicine
by Thomas J. Baldwin B.S., Washington State University, 1985 D.V.M., Washington State University, 1988 December 1992 MANUSCRIPT THESES
Unpublished theses submitted for the Master’s and Doctor's
Degrees and deposited in the Louisiana State University Libraries are available for inspection. Use of any thesis is limited by the rights of the author. Bibliographical references may be noted, but passages may not be copied unless the author has given permission. Credit must be given in subsequent written or published work.
A library which borrows this thesis for use by its clientele is expected to make sure that the borrower is aware of the above restrictions.
LOUISIANA STATE UNIVERSITY LIBRARIES TABLE OF CONTENTS
LIST OF T ABLES ...... V
LIST OF FIGURES ...... vi
ABSTRACT ...... viii
INTRODUCTION ...... 1
CHAPTER I. LITERATURE REVIEW ...... 5 Enteric septicemia of catfish ...... 5 Geographic distribution ...... 5 Species affected ...... 5 Bacterial morphology and biochemical characteristics ...... 7 Isolate homogeneity ...... 7 Bacterial components ...... 13 C a p s u l e ...... 13 F i m b r i a ...... 13 F l a g e l l a ...... 14 Lipopolysaccharides ...... 15 Outer membrane proteins ...... 17 P l a s m i d s ...... 19 Occurrence and transmission ...... 22 Clinical signs ...... 2 3 Gross lesions ...... 24 Histopathology ...... 25 Pathogenesis ...... 27 D i a g n o s i s ...... 30 T r e a t m e n t ...... 31 Vaccination ...... 33 S a l m o n e l l a ...... 34 Mucosal invasion ...... 35 Cell culture pa s s ag e ...... 36 Chemotaxis and motility ...... 37 A t t a c h m e n t ...... 38 Penetration ...... 41 Transcytosis ...... 43 T o x i n s ...... 44 Serum and phagocyte resistance ...... 46 Iron siderophores...... 48 Genetic control of virulence ...... 48 S h i g e l l a ...... 50 A t t a c h m e n t ...... 52 Penetration ...... 54 Intracellular multiplication ...... 55 T o x i n s ...... 56 Iron siderophores ...... 58 Genetic control of virulence ...... 59 Regulation of virulence genes ...... 60
ii Y e r s i n i a ...... 61 Mucosal invasion ...... 62 Chemotaxis and motility ...... 64 Attachment and penetration ...... 64 Transcellular passage ...... 69 Serum and phagocyte resistance ...... 69 T o x i n s ...... 70 Iron siderophores...... 70 Genetic control of virulence ...... 71 Study objectives ...... 73
CHAPTER II. EARLY EVENTS IN THE PATHOGENESIS OF ENTERIC SEPTICEMIA OF CHANNEL CATFISH, CAUSED BY EDWARDSIELLA ICTALURI: LIGHT MICROSCOPIC AND BACTERIOLOGIC FINDINGS ...... 76 Introduction ...... 76 Materials and methods ...... 77 B a c t e r i a ...... 77 Maintenance of channel catfish fingerlings . 78 Experimental infection ...... 79 Tissue collection ...... 79 Tissue processing ...... 80 Quantitative cultures ...... 81 Rabbit anti-E. ictaluri sera ...... 81 Immunohistochemical detection of E. i c t a l u r i ...... 82 Statistical evaluation . 83 R e s u l t s ...... 84 Bacterial cultures ...... 84 Gross lesions ...... 86 Histologic findings ...... 86 I n t e s t i n e ...... 86 L i v e r ...... 90 Spleen, trunk kidney, and head kidney . 90 D i s c u s s i o n ...... 91
CHAPTER III. EARLY EVENTS IN THE PATHOGENESIS OF ENTERIC SEPTICEMIA OF CATFISH: ELECTRON MICROSCOPIC FINDINGS ...... 97 Introduction ...... 97 Materials and methods ...... 98 B a c t e r i a ...... 98 Rabbit anti-E. ictaluri sera ...... 98 Maintenance and experimental infection of c a t f i s h ...... 99 Tissue collection ...... 99 Bacterial culture ...... 100 Tissue processing and immunogold staining . . 100 Results ...... 101 Bacterial cultures ...... 101 Electron microscopic findings ...... 102 I n t e s t i n e ...... 102
iii L i v e r ...... 106 Spleen, trunk kidney, and head kidney . 106 D i s c u s s i o n ...... 109
CHAPTER IV. ANTIGENS OF EDWARDSIELLA ICTALURI RECOGNIZED IN NATURALLY INFECTED CHANNEL CATFISH...... 113 Introduction ...... 113 Materials and methods ...... 114 B a c t e r i a ...... 114 Enriched bacterial fractions ...... 115 Outer membrane proteins ...... 115 F l a g e l l a ...... 116 Lipopolysaccharide ...... 117 Whole cell lysates ...... 118 Quantification of LPS and p r o t e i n s ...... 119 Gel electrophoresis...... 119 Immunoblots ...... 120 Production of rabbit anti-catfish IgM . . . .122 R e s u l t s ...... 124 D i s c u s s i o n ...... 128
SUMMARY ...... 134
BIBLIOGRAPHY ...... 137
VITA ...... 157
iv LIST OF TABLES
Table 1: Biochemical characteristics of EL. ictaluri . . 8
Table 2: Sugar composition of EL. ictaluri ...... 18
Table 3: Antibiotic sensitivities of EL. ictaluri . . . 32
Table 4: Highlights of invasion mechanisms used by Salmonella, Shigella, and Yersinia species .... 74
v LIST OF FIGURES
Figure 1: Weight catfish processed, 1976-1991 ...... 2
Figure 2: Percentage of fish losses by cause, 1990 . . 2
Figure 3: Restriction endonuclease maps of plasmids pEIl and pEI2 from E^ ictaluri ...... 21
Figure 4: The percent of total cases received at the Mississippi Cooperative Extension for 1982 . . . . 24
Figures 5a and 5b: Logarithmic transformation of mean CFU/g stomach, intestine, liver, and trunk k i d n e y ...... 85
Figure 6: Photomicrograph of anterior intestine 0.5 hours P I ...... 88
Figure 7: Photomicrograph of anterior intestine 0.5 hours P I ...... 88
Figure 8: Photomicrograph of necrotic enterocytes at the tip of an intestinal fold 1 hour P I ...... 88
Figure 9: Photomicrograph of anterior intestine and immunostained serial section 72 hours PI .... . 88
Figure 10: Photomicrograph of liver 72 hours PI .. . 92
Figure 11: Photomicrograph of a portal vessel 72 hours PI ...... 92
Figure 12: Photomicrograph of spleen 72 hours PI .. . 92
Figure 13: Photomicrograph of trunk kidney 48 hours PI ...... 92
Figure 14: Percent positive liver and trunk kidney cultures for EU. ictaluri. 0.25 to 96 hours PI . . 102
Figure 15: Electron micrograph of immunogold labeled E. ictaluri adhered to the brush border of an enterocyte, 0.5 hours PI...... 104
Figure 16: Electron micrograph of a dilated basilar cell between enterocytes at 0.5 hours PI ...... 104
Figure 17: Electron micrograph of a necrotic enterocyte at the tip of an intestinal fold at 72 hours P I ...... 104
vi Figure 18: Electron micrograph of a leukocyte infiltrated between enterocytes at 48 hours PI . . 104
Figure 19: Electron micrograph of an enterocyte containing immunogold labeled EL ictaluri in the cytoplasm, 72 hours PI ...... 107
Figure 20: Electron micrograph of a hepatic vessel bordered by hepatocytes at 72 hours P I ...... 107
Figure 21: Electron micrograph of several hepatocytes at 72 hours PI ...... 107
Figure 22: Electron micrograph of head kidney at 72 hours P I ...... 107
Figure 23: Sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) profile of whole cell and enriched fractions of EL. ictaluri AL83189 . . 126
Figure 24: Transferred proteins from whole cell and enriched fractions of EL. ictaluri AL83189 stained with India ink...... 12 6
Figure 25: Immunoblot analysis of EL. ictaluri. AL83189 ...... 129
Figure 26: Immunoblot analysis of EL. ictaluri. AL83189 ...... 129
vii ABSTRACT
To clarify early events in the pathogenesis of enteric
septicemia of catfish, 140 channel catfish fingerlings
(Ictalurus punctatus) were each infected with approximately
1.0 X 109 colony forming units Edwardsiella ictaluri (E.
ictaluri^ by intragastric intubation. Fish were sacrificed
at 0, 0.25, 0.5, 1, 3, 6, 12, 24, 48, 72, 96, and 120 hours
post infection (PI). Multiple tissue samples were
evaluated by gross observation, light and electron
microscopy, and immunohistochemical methods. In addition,
stomach, intestine, trunk kidney, and liver were cultured
quantitatively for bacteria.
Trunk kidney cultures were positive by 0.25 hours PI,
indicating rapid transmucosal passage. Histologic lesions
were evident in the intestine at 0.5 hours PI and in other
tissues at 48 hours. In the intestine, E^_ ictaluri was
seen in contact with the brush border at 0.5 hours PI.
Also at 0.5 hours PI, dilated cells with large
intracytoplasmic inclusions were observed subadjacent to the basement membrane. From 1 to 3 hours PI, occasional
necrotic enterocytes were seen on tips of intestinal folds.
Proprial leukocytes were rare before 24 hours PI but common
thereafter, and many contained cytoplasmic bacteria.
In other tissues, earliest observed lesions (48 hours
PI) consisted of bacteria within vascular associated phagocytic cells. Electron microscopy showed bacteria were
viii in cytoplasmic vacuoles within phagocytes. Bacteria were
also seen within vacuoles in enterocytes and hepatocytes at
later time periods.
This study confirms EL. ictaluri can invade channel
catfish by crossing the intestinal mucosa and suggests E .
ictaluri may have a similar pathogenesis as other
enteroinvasive enterobacteriaceae.
In addition, antigens of EL. ictaluri recognized by
channel catfish infected naturally were determined by probing bacterial components with serum from diseased fish.
Whole cell preparations and enriched fractions of outer membrane proteins, flagella, and lipopolysaccharide (LPS) were separated by sodium dodecyl sulfate polyacrylamide gel
electrophoresis and transferred to nitrocellulose membranes. Transfer was verified by India ink staining.
Separated bacterial components affixed to nitrocellulose membranes were probed using catfish serum and catfish serum absorbed with purified LPS. Lipopolysaccharide and the major outer membrane protein (37 kD) were strongly
immunogenic. A flagellar antigen (36 kD) and five additional outer membrane proteins (51 kD, 3 9 kD, 3 3 kD,
23.5 kD and 21 kD) were weakly immunogenic.
ix INTRODUCTION
The farming of channel catfish (Ictalurus punctatus) in the southern United States has become a major agricultural industry. In 1991, surface hectares in the 16 states with catfish production totaled 65,182 (Catfish
Production 1992). States with the most hectares devoted to catfish farming are, in decreasing order, Mississippi
(38,475), Arkansas (8,302.5), Alabama (7,695), Louisiana
(4,050), and Texas (1,336.5). Catfish producers had an excess of 323 million dollars of sales in 1990 and over 304 million dollars in 1991 (Catfish Production 1991; Catfish
Production 1992). As illustrated in figure 1, weight processed has increased from 8600 metric tons in 1976 to
177,300 metric tons in 1991. Unfortunately, as production has increased, so have catfish losses.
In 1990, 288 million fish weighing a total of 11.7 metric tons were lost to producers (Catfish Production
1991). The vast majority (259 million) were fingerlings and fry. Percentage of fish losses by cause for 1990 is shown in Figure 2. The most significant source of loss was disease, accounting for 64% of the total or 184 million fish (7.5 metric tons). The average price-per-kilogram paid to producers in 1990 was $1.70 (Catfish Production
1991). Accounting only death losses, and not decreased production, which may be more significant but more
1 Metric tons 2 0 0
1 5 0-
100-
5 0-
1976 1978 1980 1982 1984 1986 1988 1989 1990 1991
WM Weight processed
Figure 1: Weight catfish processed, 1976-1991 (Catfish Production 1992).
% LOSS 60
5 0-
4 0-
3 0 -
2 0 -
10 -
W IHTER KILL FLOOD DISEASE LOW 0 THEFT CHEMICALS BIROS FREDATORS OTHER CAUSE Figure 2: Percentage of fish losses by cause, 1990 (Catfish Production 1991). 3
difficult to quantify, at least 12 million dollars were
lost to the catfish producer from disease in 1990.
The most significant disease of cultured channel
catfish is enteric septicemia of catfish (ESC) caused by
the bacterium Edwardsiella ictaluri (E. ictaluri)
(MacMillan 1985). Enteric septicemia of catfish currently
accounts for 34% of all cases submitted to the Louisiana
Aquatic Diagnostic Laboratory (J. Hawke, Louisiana State
University, personal communication), and 40% of all cases
received at the Mississippi Cooperative Extension Service
(M. Johnson, Mississippi State University, personal
communication). Freund et al. (1990) reports that in
Mississippi alone, ESC accounts for four to six million
dollars in losses each year.
The definitive pathogenesis of ESC is unknown.
Preliminary to studies at the molecular level, a descriptive sequence of events following exposure to E .
ictaluri is required. Thus, in this dissertation, early
events in the pathogenesis of ESC are reported. In addition, the immunogenicity of structural components of E.
ictaluri in naturally infected channel catfish is presented. Chapter I, summarizes the current literature available on E-. ictaluri, and since EL. ictaluri is a member of the bacterial family Enterobacteriaceae, invasion methods of enteroinvasive members of this family are also reviewed. The subsequent two chapters explore early pathogenic events at the light and electron microscopic level. The final chapter reports on the immunogenicity of
E. ictaluri to channel catfish. CHAPTER I. LITERATURE REVIEW
Enteric septicemia of catfish
Geographic distribution
Enteric septicemia of catfish was initially described in infected channel catfish from Alabama and Georgia (Hawke
1979). By 1982, ESC outbreaks had occurred in Mississippi,
Arkansas, Idaho, Colorado, Indiana, and Maryland, with suspected cases in Arizona and California (Plumb and
Schwedler 1982). Enteric septicemia has since been reported in every state where channel catfish are cultured
(Johnson 1989). Moreover, widespread intrastate dissemination of ESC has occurred in states where intensive channel catfish culture is practiced. In Mississippi, from
1980 through 1988, 427 producers in 56 counties had outbreaks of ESC (Freund et al. 1990). The distribution of
ESC in wild catfish populations is not known.
Species affected
Edwardsiella ictaluri. the causative agent of ESC, was initially considered only infectious for ictalurid species of fish, with the majority of cases reported in channel catfish. Solitary isolates from other ictalurid fish, the white catfish (Ictalurus catus) and the brown bullhead
(Ictalurus nebulosis), were subsequently reported (Plumb and Sanchez 1983).
5 Reports of ESC in non-ictalurid fish followed. Kent and Lyons (1982) reported an "ictaluri-like" bacterium in a green knife fish fEiqemannia virescens) with manifestations of ESC. Waltman et al. (1985) isolated E^_ ictaluri from diseased aquarium fish, Danio devario. in Florida. Tilapia
(Sarotherodon aureus) and European catfish (Silurus qlanis) are susceptible to infection by EL. ictaluri. but have considerably higher lethal dose requirements than channel catfish (Plumb and Sanchez 1983; Plumb and Hilge 1986). In
1985, EL. ictaluri was isolated in Australia from two separate consignments of Rosy barbs fPuntius conchonius), aquarium fish imported from Singapore (Humphrey et al.
1986). Edwardsiella ictaluri was also isolated from walking catfish (Clarias batrachus) in natural waters off
Thailand (Kasornchandra et al. 1987). Baxa and Hedrick
(1989) reported that rainbow trout (Oncorhvnchus mvkiss) and Chinook salmon (Oncorhvnchus tshawvtscha) are susceptible to experimental infections. In addition,
Blanch et al. (1990) reported a strain of salt water adapted EL. ictaluri which produced high mortalities in cultured sea bass, Dicentrarchus labrax. These reports show that although most infections caused by E^_ ictaluri are in channel catfish within the United States, there is potential for disease outbreaks in other species and geographical areas. 7
Bacterial morphology and biochemical characteristics
Edwardsiella ictaluri is a member of the family
Enterobacteriaceae (Farmer and McWhorter 1985). It is a
small, straight rod, measuring 1 /m in diameter and 2-3 /urn
in length (Hawke 1979). Peritrichous flagella are present and the organism is motile at 25 but not at 37 °C (Hawke et al. 1981). Optimum growth, determined by maximum motility
and fermentation, is between 25-30 °C.
Edwardsiella ictaluri is a facultative anaerobe with the ability to ferment glucose, fructose, galactose, and manno.se, and produces gas (Waltman et al. 1986) . The triple sugar iron reaction (TSI tubes) is alkaline over acid with little to no gas production and no H2S. Other
important differentiating characteristics include negative reactions for cytochrome oxidase and indole. A complete listing of biochemical parameters is given in table 1.
Growth in culture is slow, requiring 36 to 48 hours on enriched media (Farmer and McWhorter 1985). Brain heart infusion or blood agar plates are commonly used. Colonies at 3 6 hours post inoculation are approximately l mm in diameter, raised, white, and opaque.
Isolate homogeneity
Edwardsiella ictaluri isolates from different geographical locations and different species of fish have remarkable homogeneity (Plumb and Vinitnantharat 1989) .
Bertolini et al. (1990) examined 32 isolates of Table l: Biochemical characteristics of E. ictaluri
Characteristic or substrate tested Reaction
oc-Methyl-D-glucoside fermentation
B-Galactos idase
a-Nitrophenyl-JS-D-galactpyranoside (ONPG) +/-
Acetate utilization
Adonitol fermentation
Alginate
Amygdalin
Arabinose fermentation
Arabitol fermentation
Arbutin
Arginine dihydrolase
Bacteriolysis
Brilliant green agar
Casein
Catalase
Cellobiose fermentation
Cellulose
Cetrimide
Chitin
Chondroitin sulfate
Citrate (Simmon's) Characteristic or substrate tested Reaction
Collagen
Cytochrome oxidase
Deoxyribonuclease
Dextrin
Dulcitol fermentation
Elastin
Erythritol fermentation
Esculin hydrolysis
Fibrinogen
Flagellation +
Fromate (gas) +
Fructose fermentation +
Galactose fermentation +
Gas from D-glucose, 37 °C
Gas from D-glucose, 25 °C +
Gelatin hydrolysis (22 °C)
Gluconate
Glucose fermentation
Glycerol fermentation
Gram stain
Growth on KCN
Growth at 42 °c 10
Characteristic or substrate tested Reaction
H2S on triple sugar iron
Hemolysis +
Hyaluronic acid
Indole production
Inositol fermentation -
Inulin -
Lactose fermentation
Lecithin
Lipase
Lysine decarboxylase +
MacConkey agar +
Malonate utilization -
Maltose fermentation +
Mannitol fermentation
Mannose fermentation +
Melibiose fermentation
Melezitose fermentation
Methyl red +
Motility, 22 °C +
Motility, 37 °C
Mucate fermentation
NaCl, i.o% + 11
Characteristic or substrate tested Reaction
NaCl, 1.5%
NaCl, 2.0%
Nitrate reduction to nitrite
0/F test, 25 °C + / +
Ornithine decarboxylase +
Pectin hydrolysis -
Phenylalanine deaminase -
Pigment
Raffinose fermentation
Rhamnose fermentation -
Ribose fermentation +
Salmonella-shigella agar +
Sorbitol fermentation
Sorbose fermentation
Starch fermentation
Sucrose fermentation
Tartrate, Jordan's
Trehalose fermentation
Tween (20-80)
Urease -
Xylose fermentation -
(Hawke 1979; Hawke et al. 1981; Farmer et al. 1985; Waltman et al. 1986) 12
E. ictaluri, including bacteria cultured from Danio and green knife fish, for variations in proteins and LPS components using polyacrylamide gel electrophoresis andimmunoblots. No structural protein differences were observed and only slight out-of-phase variances were seen in LPS antigens.
In contrast, Plumb and Klesius (1988) used monoclonal antibodies to assess the antigenic homogeneity of E. ictaluri. and reported that antibody from one hybridoma reacted with 17 isolates tested, another with 16 isolates, and a third with 15. This indicates that serological differences between isolates do exist.
Plumb and Vinitnantharat (1989) compared 40 isolates of EL. ictaluri from different geographical areas and host species for biochemical, serological, and biophysical differences. Biochemical reactions were identical in 42 of
46 tests (91%). An isolate from Thailand failed to produce gas at 30 °C, and an isolate from a green knife fish was ornithine decarboxylase and D-mannose negative. All other reactions were homologous. All isolates grew optimally at
25-3 0 °C with a medium pH of 7.0-7.5. No serological differences were determined using polyclonal rabbit sera as the primary antibody.
Starliper et al. (1988) used multilocus isozyme electrophoresis to compare enzyme mobility patterns from 33 isolates of EL, ictaluri. A total of 23 different enzymes 13 were analyzed and genetic diversity estimates made.
Minimal genetic diversity was observed "indicating a high degree of homogeneity in this species". In addition to similar enzyme patterns, homogeneity is also observed in the structural and genetic components of EL. ictaluri.
Bacterial components
Capsule: A number of pathogenic bacteria have an extramembranous capsule or glycocalyx which acts as a virulence factor (Cross 1990). This extracellular matrix envelops an individual bacterium or groups of bacteria, and is composed of large chains of homo- or heteropolysac charides (Williams 1988). A glycocalyx acts as a virulence factor principally by inhibiting phagocytosis and protecting the bacteria from complement lysis (Williams
1988).
Efforts to visualize or extract capsular material from
E. ictaluri have so far failed (Wong et al. 1989; J. C.
Newton, Louisiana state University, personal communication), which implies that EL ictaluri is nonencapsulated. Further research is needed to firmly establish the presence or absence of a capsule and its role in virulence.
Fimbria: Fimbria are proteinaceous structures found on bacterial outer membranes which function in bacterial adhesion to eukaryotic cells (Williams 1988; Jann and
Hoschutzky 1990). A number of types are recognized based on size, binding sensitivity, and contractility (Jann and
Hoschutzky 1990). Within the enterobacteriaceae, type one
fimbria which bind to D-mannose residues on eukaryotic cell
membranes are most common (Finlay and Falkow 1989a). They
are composed of identical, 17 to 21 kilodalton (kD),
repeating subunits of protein assembled into a filament and
have associated minor proteins at the tips or spaced
periodically along the length of the filament. Binding
specificities are associated with the minor proteins
(Finlay and Falkow 1989a). Because of their role in
attachment, fimbria have long been considered virulence
factors (Reid and Sobel 1987).
Fimbria have not been definitively isolated from E.
ictaluri. No fimbria were visualized by electron microscopy by Wong et al. (1989), although hemagglutinating
studies revealed JL_ ictaluri has a mannose sensitive
adhesin. Mechanical shearing likewise failed to reveal
fimbria (Wong et al. 1989). Further research is required
to document the presence or absence of fimbria in E.
ictaluri and their role in the pathogenesis of ESC.
Flagella: Flagella, which provide bacterial motility
through rotation, are filamentous structures anchored in bacterial membranes. They have three structural domains: a basal body, hook, and filament (Aizawa et al. 1985; Macnab and DeRosier 1988). The basal body contains ring-like structures which provide required rotational energy (via a 15
proton gradient), transmembrane passage, and stability
(Aizawa et al. 1985; Macnab and DeRosier 1988). The hook
is external to the cell, is composed of a single molecular
species of protein, and connects the basal body to the
flagellar filament (Macnab and DeRosier 1988). Filaments
are typically 0.02 fivi in diameter and 10 to 20 jitm long.
Filaments are composed of polymers of a single molecular
weight protein known as flagellin (Macnab and DeRosier
1988).
Flagella act as virulence factors by providing
motility (Stocker and Makela 198 6). Not all pathogens
require motility to be virulent; Shigella flexneri. a
serious enteric pathogen of humans, is highly virulent and
non-motile (Hale and Formal 1987). However, motility has
noted effects on the virulence of Salmonella and Vibrio
species (Liu et al. 1988; Allweiss et al. 1977).
Edwardsiella ictaluri has peritrichous flagella and is
motile at 25 °C (Hawke et al. 1981). Newton and Triche (in
press, a) have examined flagellin profiles from 10 isolates
of EL. ictaluri. On polyacrylamide gel electrophoresis, all
isolates showed two flagellin bands with apparent molecular
masses of 38 kD and 44 kD. The role of flagella in the
pathogenesis of EL. ictaluri is unknown.
Lipopolvsaccharides: Lipopolysaccharides (LPS) are
complex molecules found in the outer membranes of most
Gram-negative bacteria (Doyle and Sonnenfeld 1989). They are composed of three elements: lipid A, a core oligosaccharide (KDO), and O-side chain polysaccharides
(Luderitz et al. 1982). The O-side chains are composed of repeating units of oligosaccharides and vary both in structure and quantity between bacteria (Doyle and
Sonnenfeld 1989). These changes are partly responsible for the serological and structural differences between bacteria
(Orskov et al. 1977). Bacteria with complete LPS O-side chains are referred to as smooth strains due to their appearance on culture plates. Bacteria lacking significant amounts of O-side chain polysaccharide units are referred to as rough strains (Doyle and Sonnenfeld 1989).
The LPS of Ei ictaluri has been isolated and examined by polyacrylamide gel electrophoresis, gas chromatography, spectrophotometry, and immunoblotting techniques (Nomura and Akoki 1985; Weete et al. 1988, Bertolini et al. 1990;
Newton and Triche in press, b). Nomura, Bertolini, and
Newton, in separate experiments, found the LPS of 63 different isolates of JL ictaluri to be of the smooth type.
Thirty of these isolates were examined by Newton and Triche by SDS-PAGE, and a complete ladder-like array of low, medium, and high molecular weight bands were observed.
Little variation was noted between isolates. This corresponds with observations by Bertolini et al. (1990) on
3 2 isolates. Weete et al. (1988) examined an isolate originally obtained in an outbreak of ESC in Georgia and found it to be of the rough type. Analysis of this strain showed the lipid A portion was similar to that found in Salmonella tvphimurium and Escherichia coli. but the core oligosaccharide contained twice the hexose. The O-antigen sugar content was also different, having high quantities of glucose and galactose but no mannose or dideoxy sugars.
The percent sugar composition of this rough strain is given in table 2. Whether this analysis is representative of smooth strains is unknown. The importance of LPS in the virulence of EL. ictaluri is also unknown.
Outer membrane proteins: In addition to LPS and fimbria, a number of proteins are in the outer membranes of bacteria (Doyle and Sonnenfeld 1989). These usually number less then 10 and are structural proteins only, possessing no enzymatic activity. Examples include the lipoprotein which binds to intramembranous peptidoglycan, and a number of porins (Williams 1988).
Outer membrane hemolysins are exceptions and may act as enzymatic cytotoxic virulence factors (Reitmeyer et al.
1986). Shigella and enteroinvasive Escherichia utilize a contact-hemolysin to escape from endocytotic vacuoles following cellular uptake, gaining access to host cytoplasm and avoiding lysosomal degradation (Hale and Formal 1987).
In addition, outer membrane proteins may serve as virulence Table 2: Sugar composition of E^. ictaluri
Sugar Percent within O-antigen side chains
E. ictaluri S. tvphimurium E. coli
Rhamnose 9.3 25.1 2.8
Arabinose 0 1.5 0
Mannose 0 16.5 0
Glucose 40.9 8.0 16.2
Galactose 29.2 20. 0 16.5
Mannohepto s e 10.1 1.1 0.6
Unknown 10.5 0.9 4.3
(Weete et al. 1988)
factors by contributing to attachment and penetration
(Hoffman 1992).
Recently, Newton et al. (1990) examined protein profiles from 33 E^. ictaluri isolates by N-lauryl- sarcocinate extraction and polyacrylamide gel electrophoresis. One major protein, with apparent molecular mass of 35 kD, and nine minor proteins, with apparent molecular masses of 71, 51, 46, 43.5, 38.5, 37.5,
31.5, 28.5, and 19.5 kD, were found on 27 channel catfish isolates. Identical gel patterns were found for Bengal danio and walking catfish isolates. A green knife fish isolate had a major protein of 37 kD, a moderate increase in the amount of the 38.5 kD protein, lacked the 46 kD 19 band, and had a unique 45 kD protein. A white catfish isolate had two major proteins, 43 and 35 kD, but was otherwise similar to the channel catfish isolates. These studies further indicate the common theme of homogeneity demonstrated by EL. ictaluri. The role of these proteins as virulence factors is unknown.
Plasmids: Extrachromosomal segments of circular DNA known as plasmids carry genes encoding a variety of virulence factors in pathogenic bacteria (Elsell and
Shipley 1980). Factors such as resistance to complement lysis, antibiotic and heavy metal resistance, and production of outer membrane proteins, LPS, fimbria, and toxins are all known to be plasma encoded (Elwell and
Shipley 1980; Foster 1983; Kopecki and Formal 1984). The importance of plasmids in the virulence of EL. ictaluri is unknown although isolate analyses has demonstrated remarkable strain homogeneity suggesting an important function.
Speyerer and Boyle (1987) investigated plasmid contents from 18 isolates of EL. ictaluri by agarose gel electrophoresis. Two small plasmids of 5200 and 4400 base pairs were isolated. Plasmids from all 18 isolates hybridized with radiolabeled portions of the 52 00 base pair plasmid indicating homologous regions. This work was corroborated by Lobb and Rhodes (1987) who isolated two plasmids, 5700 and 4900 base pairs, from 18 different 20
strains of E-. ictaluri. Moreover, restriction enzyme
analysis showed that each strain had identical restriction
patterns.
Minor variations in plasmid profiles from isolates in
non-ictalurid fish were reported by Newton et al. (1988)
who examined 55 isolates of E^. ictaluri. All 49 isolates
from channel catfish and one from a Bengal danio had
plasmid profiles similar to those reported by Lobb and
Rhodes (1987), and were designated as pEIl and pEI2.
Restriction endonuclease maps of the plasmids were
performed and are shown in figure 3. Isolates from walking
catfish from Thailand had three plasmids with molecular masses of 2.5, 3.7 and 45 megadaltons (MD). A green knife
fish isolate also had three plasmids with masses of 2.1,
2.5, and 3.7 MD. A white catfish isolate contained one
plasmid of 3.2 MD. Reid and Boyle (1989) further
characterized the plasmid profiles of E^_ ictaluri isolated
from ictalurid and non-ictalurid fish. Homologous plasmids, 5600 base pairs and 4000 or 4700 base pairs, were
found in every isolate except the white catfish, which
contained only a solitary plasmid. Purified regions of the
5600 base pair plasmid was used as a hybridization probe
and every 5600 base pair plasmid from all isolates of E.
ictaluri were complementary. None of the 4000 or 4700 base pair plasmids hybridized as reported earlier by Speyerer and Boyle (1987). It was postulated contamination of the 21
PUC 19 PUC 19 — Eco R 1 — Eco R 1 Bgl I — — Pvu II Eco R V — Bgl I — — Pst I — Eco R V
Eco R V — — Bgl II — Bgl II Pst I — — Nde I Pvu II — — Pst I Bam H I — Ava I — — Sma I — Pvu II Bgl I — — Sst I Pvu I —
Ava I — — Sma I — Bgl 1 — Eco R V
Bgl I — Bgl I —
— Bgl I
— Pvu I
Nde I —
pEI 1 pEI2
— Eco R I — Eco R I
PUC 19 PUC 19
Figure 3: Restriction endonuclease maps of plasmids pEIl and pEI2 from JL. ictaluri (Newton et al. 1988) . 22 earlier probe produced the observed cross-reactions. Reid and Boyle (1989) concluded that "all isolates, regardless of source, contained classes of homologous plasmids with similar but not identical sizes". Although the role of plasmids in virulence of E^. ictaluri infections is not known, their uniform presence in most isolates suggests they encode important products.
Occurrence and transmission
Outbreaks of ESC are highly correlated with water temperature (Hawke 1979; MacMillan 1985; Johnson 1989). In the southern United States, peak incidence-of disease occurs in June and September when water temperatures are between 22 and 28 °C (MacMillan 1985). This infection temperature phenomenon is observed yearly. Francis-Floyd et al. (1987) studied the effect of temperature on the clinical manifestations of ESC on catfish experimentally infected with EU. ictaluri by intraperitoneal injection.
The percent mortality was highest in fish kept in water temperatures from 23 to 25 °C. They concluded that fish in water temperatures from 22 to 28 °C were most susceptible to infection.
In an epidemiological study on the occurrence of ESC in Mississippi from 1980 to 1988, Freund et al. (1990) found that 90.7% of all cases occurred in the spring (May and June) and fall (September and October) when air temperatures were within 16.1 to 28.3 °C. This correlation 23 is illustrated in Figure 4 which shows the percent of total cases received at the Mississippi Cooperative Extension
Service and average monthly temperatures for 1982
(MacMillan 1985).
Transmission methods of EU. ictaluri from pond to pond are unknown but dissemination may partly be due to inadvertent transport of infected fingerlings
(MacMillan 1985). A possible carrier state, in which E. ictaluri can be cultured from the intestine in the absence of clinical disease, has been reported (Johnson 1989), and may be a source of new pond infections. Transport of bacteria laden water on machines and equipment should also be considered. Edwardsiella ictaluri can survive in pond water for eight days (Hawke 1979) and in pond mud for greater than 95 days (Plumb and Quinlan 1986). Pond mud may be the source of repeat infections in the same pond in successive years.
Clinical signs
Enteric septicemia of catfish can occur in both acute and chronic forms (Johnson 1989). The clinical signs of acute ESC are similar to those of other systemic bacterial infections of channel catfish. Infected fish are lethargic, weak, often hang vertically at the water surface, and swim erratically or in spiraling circles when disturbed (Plumb and Schwedler 1982; Blazer et al. 1985). 24 Water temperature 3 5
3 0 -
2 5-
2 0 -
1 5-
10 -
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Month
M % MORTALITY — WATER TEMPERATURE
Figure 4: The percent of total cases received at the Mississippi Cooperative Extension for 1982 (MacMillan 1985).
Chronically infected fish may have no clinical signs or may show evidence of neurological dysfunction.
Gross lesions
In naturally acquired acute infections, fingerlings are pale, exophthalmic, have a distended abdomen (ascites), and frequently show numerous petechiae and ecchymoses at the fin bases, opercular flaps, and skin over the ventral abdomen and jaw (Jarboe et al. 1984; Miyazaki and Plumb
1985; Miyazaki and Kaige 1985; Blazer et al. 1985).
Scattered, raised, 1-5 mm, white plaques and reddened epithelial ulcers are common on the head and body skin
(Jarboe et al. 1984; Miyazaki and Kaige 1985). Erosion and 25
ulceration of the olfactory sacs are seen {Miyazaki and
Plumb 1985). Common visceral lesions include a swollen,
mottled, pale to deep red liver with multifocal areas of
necrosis and hemorrhage. The head kidney, trunk kidney,
and spleen are edematous and may have areas of necrosis and
hemorrhage. The anterior portion of the intestine is
mildly dilated and contains clear yellow to blood tinged
fluid.
Chronically infected fish frequently have a swelling
or prominent ulcer on the dorsum of the head caudomedial to
the eyes (Blazer et al. 1985; Miyazaki and Plumb 1985).
Other organs may or may not have areas of necrosis.
In experimentally infected fish, lesions are similar
(Areechon and Plumb 1983; Shotts et al. 1985; and Newton et
al. 1989). Some variation may occur depending on the method of exposure. For example, Areechon and Plumb
indicated an absence of intestinal lesions when the
bacteria were given by intraperitoneal injection.
Experimental infections have provided information on
the time of lesion occurrence and progression. The acute
lesions are reported to occur between 4 and 12 days post
exposure and the chronic lesions 3 to 4 weeks post exposure
(Newton et al. 1989).
Histooatholoav
Lesions observed grossly are further clarified histologically. In acute infections, the lamina propria and submucosa of the small intestine are multifocally infiltrated with macrophages, lymphocytes, and fewer neutrophils (Blazer et al. 1985; Newton et al. 1989).
Mucosal ulceration is rare. Histologic lesions frequently precede gross lesions. For example, intestinal lesions were seen by Shotts et al. (1985) in one of three fish, infected by gastric intubation, and examined at 1 day post infection. Newton et al. (1989) reported intestinal lesions by 2 days post infection in 28 out of 37 fingerlings immersion exposed to EL. ictaluri. Lesions were most severe on days 5 through 8 post infection.
Multifocal necrotic foci infiltrated with inflammatory cells are observed in the liver, trunk kidney, head kidney, and spleen (Jarboe et al. 1984; Miyazaki and Plumb 1985;
Shotts et al. 1985; and Newton et al. 1989). Hemorrhage, leukocyte infiltrates without necrosis, and edema are also reported in these organs. Multifocal areas of epidermal necrosis with associated leukocyte infiltrates are in the skin (Jarboe et al. 1984; Newton et al. 1989).
Mononuclear cell infiltration into the olfactory sacs and occasional meningoencephalitis are also observed.
In chronically infected fish, accumulations of macrophages within and around the olfactory nerves and in the olfactory bulbs of the brain are present (Miyazaki and
Plumb 1985; Miyazaki and Kaige 1985; Blazer et al. 1985; and Newton et al. 1989). Extension into the meninges and 27 adjacent fat is common. Inflammation also extends through, the sutra fontanels of the skull and into the subcutaneous tissues. The overlying skin ulcerates producing the grossly observed head ulcer.
Few studies on clinical pathologic changes observed in fish with ESC have been done. Areechon and Plumb (1983) reported a normocytic normochromic anemia, leukopenia, hypoproteinemia, and hypoglycemia from fish injected intraperitoneally with EU. ictaluri. No alterations in the leukocyte differential counts were noted.
Pathogenesis
A descriptive sequence of pathogenic events following exposure toL. ictaluri is not available. However, pathologic studies have provided some insights. Shotts et al. (1985) successfully infected fish by gastric intubation and indicated the intestine as a point of bacterial access.
Miyazake and Plumb (1985) suggested the olfactory sac as a site of pathogen passage. Shotts et al. (1985) and Newton et al. (1989) both proposed intestinal invasion by the bacteria was responsible for the acute syndrome, while passage through the olfactory sac and subsequent ascending olfactory neuritis and perineuritis produced the chronic syndrome. Bacterial passage across either mucosal barrier has not been observed.
Molecular mechanisms of pathogenesis have also not been defined. In other Enterobacteriaceae, outer membrane 28 structures such as lipopolysaccharides, proteinaceous adhesins, and fimbria are considered virulence factors
(Williams 1988). In addition, the production of toxins, motility due to flagella, and plasmid and chromosomally encoded serum and antibiotic resistance are important for pathogenesis (Finlay and Falkow 1989a). These structures aid in chemotaxis, bacterial attachment and penetration of host cells, resistance to complement lysis, allow for bacterial multiplication, and provide resistance to phagocytic destruction.
A mannose sensitive adhesin has been described for E. ictaluri (Wong et al. 1989). Whether this is an outer membrane protein or important for attachment or internalization is unknown. A cell associated hemolysin which may be important for cellular penetration, replication, or cellular escape has also been described
(Waltman et al. 1986). Edwardsiella ictaluri does possess chondroitin sulfatase and this may be partially responsible for the observed erosion through the sutra fontanels in chronic infections (Waltmann et al. 1986).
As can best be determined, only one cell culture invasion study has been done with EL-. ictaluri (Janda et al.
1991). Edwardsiella ictaluri was unable to penetrate monolayers of HEp-2 cells (human larynx origin) whereas E. tarda and some strains of E^. hoshinae were. The reasons underlying this difference are unknown. 29
A number of researchers have noted large numbers of E. ictaluri within macrophages in various tissues (Blazer et al. 1985; Miyazaki and Kaige 1985; Shotts et al. 1985;
Miyazaki and Plumb 1985; and Newton et al. 1989). Shotts et al. (1985) suggested the bacteria within macrophages were viable and "may actually proliferate within the macrophage".
In electron microscopic studies of catfish macrophages infected with E^. ictaluri. Stanley (1991) found numerous virulent bacteria attached to and within macrophages.
Avirulent bacteria had little association with macrophages.
Moreover, virulent bacteria were resistant to phagolysosomal degradation but avirulent strains were not.
The molecular mechanisms responsible are unknown.
Regardless of the mechanisms used, Eh, ictaluri can replicate to high numbers in channel catfish. Areechon and
Plumb (1983) injected ictaluri into the coelomic cavity and quantified bacterial numbers in various tissues.
Highest bacterial numbers, exceeding 107 bacteria per gram tissue, were in the spleen, trunk kidney, and liver 4 days post injection and declined thereafter.
Mechanisms underlying observed clinicopathologic changes are also unknown. Areechon and Plumb (1983) proffered a number of possible explanations for the observed anemia, hypoproteinemia, and leukopenia in experimentally infected fish. Hemorrhage, bacteria-mediated 30 hemolysis and hypoglycemia, destruction of hematopoietic cells, and protein loss through damaged renal tubules were considered. Further research in this area will clarify responsible mechanisms.
Diagnosis
In a typical clinical situation, ESC is diagnosed by gross lesion observation and bacterial cultures of internal organs, a procedure requiring several days (Ainsworth et al
1986) . Blood or brain heart infusion agar are commonly used (Hawke 1979). Monitoring environmental water levels as a means of detecting exposure has been difficult due to plate contamination by faster growing bacteria. Recently, a selective medium for the isolation of EU. ictaluri has been developed (Shotts and Waltmann 1990) which should aid in water detection.
Serological methods of detecting E^. ictaluri have also been developed. Serum agglutination tests and enzyme linked immunosorbant assays for detecting antibody against
E. ictaluri have been developed which determine previous exposure (Saeed and Plumb 1987; Waterstrat et al. 1989).
In addition, polyclonal and monoclonal antibodies directed against E^. ictaluri and coupled with florescent compounds or color producing enzyme reactions have been developed
(Rogers 1981; Ainsworth et al. 1986). The monoclonal indirect florescent antibody technique by Ainsworth et al.
(1986) claims a 90.3% correlation with culture results and 31 requires 3 to 4 hours. A rapid diagnosis is advantageous since prompt treatment can be initiated.
Treatment
Outbreaks of ESC are currently controlled using antibiotic medicated feeds. Edwardsiella ictaluri is sensitive to a number of antibiotics but resistant to others (Hawke 1979; Hawke et al. 1981; Waltman and Shotts
1986; Plumb et al. 1987); these are listed in table 3.
Plumb et al. (1987) and Johnson (1989) report that only two antibiotics are registered by the Food and Drug
Administration for treatment of catfish raised for food: ormetoprim-sulfadimethoxine (Romet-3 0*) and oxytetracycline
(Terramycin*) . Ormetoprim-sulfadimethoxine is fed at 50 milograms (mg) active ingredient per kilogram (kg) of fish for 5 days. Oxytetracycline is fed at 55 to 82 mg active ingredient per kg fish for 10 days (Johnson 1989). An advantage of the potentiated sulfa is a shorter withdrawal time prior to marketing - 3 days verses 21 days for oxytetracycline. Moreover, Romet-30* can be incorporated into floating feeds whereas Terramycin* is destroyed by heat during pelleting (Johnson 1989). Currently, there are no published reports of resistance to these therapeutics, but resistant strains have been encountered clinically (M.
Johnson, Mississippi State University, personal communication). 32
Table 3: Antibiotic sensitivities of EL. ictaluri
Sensitive Resistant
A u r e o m y c m Bacitracin
Chiortetracyc1ine Clindamycin
Cinoxacin Cloxacillin
Dihydrostreptomycin Colistin
Furacin Erythromyc in
Kanamycin Lincomycin
Moxalactam Sulfadiazine
Nalidixic acid Sulfamerazine
Neomycin Triple sulfa
Nitrofurantoin
Ormetoprim-sulfadimethoxine (Romet-30*)
Oxolinic acid
Oxytetracycline (Terramycin*)
Ticarcillion
Trimethoprim
(Hawke 1979; Hawke et al. 1981; Waltman and Shotts 1986; Plumb et al. 1987)
In addition to medicated feeds, high levels of dietary vitamin C have been advanced as an aid in controlling and preventing outbreaks of ESC (Liu et al. 1989). The amount of vitamin C needed by growing catfish is between 30 and 90 mg/kg diet (Li and Lovell 1985). Fish receiving 1000 mg/kg 33 feed had a 100-fold increase in the exposure dose required to kill 50% of the fish (LD50) following vaccination with a whole cell bacterin (Lin et al. 1989). The significance of this report is uncertain since all fish had pre-trial titers to ictaluri. indicating previous exposure.
Vaccination
Most vaccines developed against ESC have been whole cell bacterins or sonicated cells treated with formalin
(Plumb 1988). Results have been equivocal. Saeed and
Plumb (1986) used purified lipopolysaccharide arid formalin- killed whole cells injected intraperitoneally to vaccinate fish, and subsequently challenged by intraperitoneal injection. Mortality rates were significantly reduced in vaccinated fish. In contrast, Plumb et al. (1986) compared intraperitoneal vaccination with whole cell preparations to immersion vaccination. The injected fish had higher antibody titers, as measured by agglutination, and significantly higher mortalities, implying vaccinated fish were more susceptible to fatal disease.
Vaccination against spring viremia of carp, cold water vibriosis, and enteric red mouth disease have been successful (Ellis 1988). These successes and the annual losses to ESC are providing impetus for further research in this area.
Edwardsiella ictaluri is a member of the
Enterobacteriaceae (Farmer and McWhorter 1985). Other 34 members of this family are known enteric pathogens, namely
Salmonella, Shigella, Yersinia, and Escherichia genera
(Finlay and Falkow 1989a). An overview of invasive strategies utilized by these bacteria is provided since E. ictaluri may use similar mechanisms in accessing fish. In each instance, membrane passage is described followed by a discussion of possible virulence mechanisms and known genetic controls.
Salmonella
Salmonella are responsible for a variety of diseases in humans and animals (Groisman and Saier 1990). Most
Salmonella infections occur following oral ingestion of bacteria and can be roughly categorized as enteric or systemic infections. Enteric pathogens, enteritis in man resulting from ingestion of food contaminated with
Salmonella tvphimurium for example, pass through the stomach and are taken up by intestinal epithelial cells.
Bacteria replicate in enterocytes causing cytotoxic damage and subsequent inflammation (Groisman and Saier, 1990).
Enteric species rarely penetrate deeper than the lamina propria and the target cell of infection are enterocytes or modified enterocytes (M cells) overlying Peyer's patches
(Finlay and Falkow, 1988a). In contrast, Salmonella species producing systemic infections, typhoid fever caused by Salmonella tvphi for example, penetrate the intestinal 35 epithelium and enter cells of the reticuloendothelial system, and subsequently disseminate to the spleen, lymph nodes, liver, and other organs (Edelman and Levine 1986).
Mechanisms for mucosal invasion for both categories appear similar with minor interspecies variation (Finlay and
Falkow 1988a) and are discussed below.
Mucosal invasion
Following oral ingestion, Salmonella are carried to the distal small intestine where they attack two cell types: principally the modified epithelial cells (M-cells) overlying Peyer/s patches, but also columnar epithelial cells (Kohbata et al. 1986). Takeuchi (1967), using gastric intubation in guinea pigs, described the invasion of enterocytes by Salmonella tvphimurium. Bacteria were seen adjacent to the brush border, and the associated microvilli and terminal web of the enterocyte were degenerate. A cavity or indentation was formed in the apical cytoplasm underlying the degenerate microvilli. As bacteria advanced deeper into the cell, the top of the cavity fused, enclosing the microbe in a membrane bound vacuole. Vacuoles containing bacteria then separated from the apical plasma membrane.
Rarely, some bacteria penetrated intercellular junctions and accessed the lamina propria by passing between enterocytes. 36
Only villus tip enterocytes were attacked; crypt
epithelium was not involved even though luminal bacteria
were observed in these regions. Most bacteria were seen in
epithelial cells between 8 and 12 hours post infection. By
24 hours, penetration was infrequent and most organisms
were in proprial phagocytes.
Cell culture passage
Further insight on mucosal passage has been obtained
using cell culture infection studies. Cell types commonly used include HeLa (human cervical cells), HEp-2 (human
larynx epithelial cells), CHO (Chinese hamster ovary
fibroblast cells), CAC02 (human intestinal epithelial
cells), and MDCK (Madin Darby canine kidney epithelial
cells) (Finlay et al. 1989b).
Madin Darby canine kidney epithelial cells are
especially valuable since polarized epithelial monolayers with distinct apical and basilar orientations can be made with this cell line (Simons and Fuller 1985). Polarized monolayers, maintained by tight junctions, can be produced
if cells are cultured on permeable supports, usually
filters with 3 fim. pores (Simons and Fuller 1985) .
Utilizing this model, Finlay et al. (1988b) investigated the invasiveness of Salmonella choleraesuis. a species which produces fatal bacteremias in humans. The bacteria began to adhere and penetrate the monolayer in 2 hours, and 37 the numbers adhering and penetrating increased over the next 4 hours.
Morphologic changes in infected cell monolayers parallel those described by Takeuchi (1967) in enterocytes.
The bacteria contact the polarized monolayer and brush border degeneration follows. At contact points, a cavity is formed in the apical membrane and the bacteria are internalized in membrane bound vacuoles. These indentations are often, but not always, associated with coated pits. This suggests Salmonella may use normal cellular transport systems (receptor mediated endocytosis) to gain access into the cell. Bacteria were motile within vacuoles and underwent replication with a generation time of approximately 50 minutes. No bacteria were seen exiting the basolateral membrane but culture of the media underlying the permeable filter confirmed passage.
Virulence factors allowing Salmonella species to locate, attach, penetrate, replicate, and escape from epithelial cells, as well as resist serum killing and survive and replicate within macrophages, are only partially characterized. Summaries of possible mechanisms for each step are provided.
Chemotaxis and motility
Chemotaxis, the ability of bacteria to move toward or away from chemical stimuli, is a virulence factor as pathogens can be directed toward host surfaces (Finlay and Falkow 1989a). Uhlman and Jones (1982) showed a diffusible agent secreted by HeLa cells was chemotactic for Salmonella tvphimurium and increased infectivity. Motility is a natural part of chemotaxis and is achieved in Salmonella by flagella. Liu et al. (1988) identified 63 Salmonella tvphi mutants that were either non-motile, non-flagellated, or non-chemotactic, and all were non-invasive for HeLa cell monolayers. Spontaneous revertants were as invasive as wild types.. Carsiotis et al. (1984), however, tested non- chemotactic, non-motile, and non-flagellated Salmonella tvphimurium mutants and concluded only the flagella were virulence factors. Flagella were not required for the bacteria to penetrate the intestinal tract, but for replication in the liver and spleen. Apparently flagella impart increased resistance to phagocytic killing or allow for improved intracellular replication. Considered together, chemotaxis and motility appear to be virulence factors for Salmonella but their exact roles need further clarification.
Attachment
Bacterial attachment to host cells is commonly mediated by outer membrane proteinaceous adhesins which bind to cellular receptors (Finlay and Falkow, 1989a).
Jones and Richardson (1981) identified two adherence factors in Salmonella tvphimurium: type one fimbria and a mannose resistant hemagglutinin (MRHA). Type one fimbria 39 were determined to "participate little, if at all, in the attachment of salmonellae to HeLa cells". Oral challenges with piliated and non-piliated strains have been performed with minimal differences in virulence observed (Duguid et
al. 1976).
Mannose resistant hemagglutinin is expressed by those
Salmonella species primarily involved in enteric infections and not from species producing systemic disease. It has not been characterized on a molecular basis and its role in pathogenicity is largely undetermined.
Adherence of Salmonella to epithelial cells requires de novo bacterial protein and ribonucleic acid syntheses, but not DNA replication (Finlay et al. 1988c). Inhibitors of protein (chloramphenicol) or RNA (rifampin) syntheses completely block invasiveness. When protein profiles from bound bacteria were examined by two-dimensional electrophoresis, six new protein bands were observed and seven previously observed proteins were no longer produced
(Finlay et al. 1988c). one newly synthesized protein (40 kD) was localized on the bacterial outer membrane by
[123]iodine labeling. These proteins were shown to be different from heat shock proteins (Finlay et al. 1988c).
Monolayers pre-treated with trypsin (a serine protease), periodic acid (a sugar oxidant), and neuraminidase (an enzyme which disrupts sialic acid membrane structures), do not induce bacterial de novo 40
protein synthesis and are refractory to adherence and
invasion by Salmonella (Finlay et al. 1989b). This implies
glycoprotein structures on the cell surface stimulate
production of specific proteins in Salmonella, and that
these proteins are required for attachment and penetration.
The cellular structures stimulating bacterial protein
synthesis may be membrane receptors (Finlay and Falkow
1988d). Drugs inhibiting receptor mediated endocytosis
prevent invasion by Salmonella. Dansylcadaverine, such a
reagent, is reported to prevent receptor-ligand clustering
in coated pits by inhibiting receptor cycling (Davies et al
1984) . It is postulated dansylcadaverine inhibits entry of
Salmonella by reducing the number of surface membrane
receptors available for bacterial adhesion or penetration
(Finlay and Falkow 1988d). Required receptors may remain
internalized and unavailable for binding with the surface
proteins of Salmonella.
Contact with eukaryotic cells may not be the only
factor inducing de novo protein synthesis in Salmonella.
Recently, low oxygen tension in broth cultures induced
synthesis of two of the six proteins (Lee and Falkow 1990).
Bacteria grown in 1% oxygen were able to adhere 20- to 100-
fold better than those grown in standard aerobic cultures.
It was hypothesized low oxygen tension in the brush border region may act as an environmental stimulus to induce
Salmonella protein synthesis. The LPS component of the outer membrane of Salmonella has been shown to aid in attachment to eukaryotic cells
(Mroczenski-Wildey et al 1989). Most Salmonella producing systemic disease require complete LPS O-side chains for invasiveness (Finlay et al. 1988a). Rough mutants are unable to penetrate the intestinal epithelium and gain access to the target reticuloendothelial cells. In contrast, enteroinvasive species may not require complete
LPS for attachment. Tannock et al. (1975) showed that rough mutants of Salmonella tvnhimurium could enter eukaryotic cells and replicate.
Penetration
Following adhesion, Salmonella are internalized in epithelial cells. This process has been shown to require functional host microfilaments and host energy expenditure
(glycolysis) in HEp-2 cells and MDCK cells (Finlay and
Falkow 1988e). Cytochalasins B and D are powerful inhibitors of microfilament polymerization and cell lines pre-treated with these agents are refractory to invasion
(Finlay and Falkow I988d; Finlay and Falkow 1988e).
Attachment is not inhibited but penetration does not occur.
Polymerized F actin, but not G actin monomers, can be visualized using the fluorescent stain phallicidin (Finlay et al. 1989b). When MDCK monolayers were stained with phallicidin following infection with Salmonella choleraesuis. "small, bright staining balls were apparent 42 in the infected monolayers" (Finlay et al. 1989b).
Bacteria were subsequently found inside these structures.
Microtubule inhibitors such as colchicine, vincristine, and vinblastine have no effect on invasion of polarized cell cultures by Salmonella, indicating microtubules are not involved in this process (Finlay and
Falkow 1989c).
A key aspect of receptor mediated endocytosis, which the entry of Salmonella appears to share some components, is acidification of the internalized endosome prior to phagolysosomal fusion (Dautry-Varsat and Lodish 1989).
Proton pumps (ATPases) in the endosome transport hydrogen ions into the vacuole lowering the pH, often causing the uncoupling of receptor-ligand complexes (Mellman et al.
1986). Is this portion of the endocytotic process required for the penetration of Salmonella? Monensin blocks endosome acidification by inducing the exchange of protons for potassium ions (Finlay and Falkow 1988e). Treatment of monolayers with monensin failed to inhibit Salmonella invasion or intracellular replication indicating endosome acidification is not required (Finlay and Falkow 1988d;
Finlay and Falkow 1988e).
Actin polymerization is known to occur in phagocytosis but not in receptor mediated endocytosis (Dautry-Varsat and
Lodish 1984). Thus, the internalization of Salmonella seems to include aspects of both phagocytosis and receptor 43 mediated endocytosis, and has been termed parasite directed or parasite mediated endocytosis (Moulder 1985).
Transcvtos is
The passage of Salmonella through polarized cell monolayers is considered to occur by a form of transcytosis. Transcytosis describes the transport of bound receptors and ligands from the endosome to the surface of the cell opposite from the membrane of entry
(Mostov and Simister 1985). The best characterized examples of transcytosis are the transport of immunoglobulins across epithelial surfaces. Tears, bile, milk, and saliva are all high in immunoglobulins, for example. Immunoglobulins bind to surface receptors, are internalized by receptor mediated endocytosis, and transported across the cell (Mostov and Simister 1985).
Similar mechanisms may be involved in the transport of
Salmonella, although no bacteria have been seen crossing the basilar membranes of polarized monolayers (Finlay et al. 1989b) or in ultrastructural intestinal studies
(Takeuchi 1966; Kohbata et al. 1986). However, cultures of basilar cell suspension media are positive for Salmonella approximately 4 hours post infection of the monolayer
(Finlay and Falkow 1988b), and morphologic intestinal studies have described bacteria in the lamina propria, indicating transepithelial passage. 44
Transepithelial passage of Salmonella through monolayers produces no grossly observable cell pathology until approximately 8 hours post infection (Finlay and
Falkow 1988b). Trypan blue staining of monolayers indicates most cells are viable at this time even though large numbers of organisms have passed through. Non- invasive Escherichia coli. added to the apical surface prior to infection with Salmonella, are not cultured from the basilar media until after 8 hours - further indicating viability of the monolayer.
Measurements of electrical resistance and
[35S]methionine uptake by cells are more sensitive indicators of cell viability. By 4 hours post infection, electrical resistance is lost, even though tight junctions required to maintain polarity are morphologically normal
(Finlay and Falkow 1988b; Finlay and Falkow 1989c).
Methionine uptake remains normal up to 8 hours indicating the cells are still viable in spite of the loss of polarity. Mechanisms responsible for monolayer depolarization are unknown but it is postulated rearrangements of intracellular actin, occurring during bacterial penetration, disrupt tight junction integrity
(Finlay and Falkow 1989c).
Toxins
Considering Salmonella can pass through epithelial cell barriers and leave the cells viable, the role of 45
cellular toxins in virulence has been questioned.
Salmonella possess three known toxins: an enterotoxin
morphologically and genetically similar to cholera toxin, a
non-LPS outer membrane component that inhibits eukaryotic
cellular protein synthesis, and endotoxin or the lipid A
portion of outer membrane LPS (Reitmeyer et al. 1986;
Finlay and Falkow 1988a; Groisman and Saier 1990).
The enterotoxin functions similar to cholera toxin
produced by Vibrio cholera: toxin binds to a Gni!
ganglioside and activates adenylate cyclase. Subsequent
elevations in cAMP disrupt normal cellular metabolism
resulting in excessive ion loss and passive afflux of
water.
The outer membrane component which disrupts cellular
protein synthesis has not been definitively identified;
further work should clarify the importance of this product
as a virulence mechanism.
The endotoxic effects of lipid A are well established;
the molecule exerts its effect primarily on macrophages and
lymphocytes (Cybulsky et al. 1988). These leukocytes
subsequently release a number of cytokines (interleukins,
tumor necrosis factor) which have broad physiologic effects
such as pyrogenicity, leukocytosis, increased hepatic acute phase protein synthesis, increased plasma proteinase
inhibitors, increased plasma transport proteins, and 46
potentially life-threatening hypotension (Cybulsky et al.
1988).
Serum and phagocyte resistance
Complement mediated lysis of bacteria is a major
defense mechanism impeding bacterial invasion (Doyle and
Sonnenfeld 1989). The end product of the complement
cascade is a transmembrane pore, inserted in the bacterial
membrane, which disrupts osmotic balance killing the
bacteria. The principle protective component of Salmonella
appears to be the LPS in the outer membrane.
Lipopolysaccharide prevents both the activation and
deposition of complement on the bacterial surface (Groisman
and Saier 1990). Rough strains lacking 0-side chains are
more susceptible to complement lysis (Finlay and Falkow
1989a).
In addition, complement coated bacteria are
preferentially taken up and destroyed by macrophages. The
number and arrangement of sugar residues composing the
repeating sugar units of the 0-side chains vary in
efficiency of protection. For example, modified Salmonella
tvphimurium strains with 0-6,7 side chains activate the
complement cascade and are efficiently phagocytosed and
destroyed by macrophages. Strains with 0-4,12 side chains
do not activate complement and are less efficiently phagocytosed (Finlay and Falkow 1988a). Virulent Salmonella producing systemic disease are able to survive within macrophages. Salmonella do not inhibit phagolysosomal fusion and consequently must be resistant to oxygen dependent and independent mechanisms of cellular destruction (Groisman and Saier 1990). Oxygen dependent mechanisms include the free radicals superoxide and hydrogen peroxide produced in the respiratory burst.
Oxygen independent mechanisms include hydrolytic enzymes, acidification of the phagolysosome, and lyses by cationic peptides (defensins). How Salmonella defends against these factors has only begun to be established.
Fields et al. (1986) identified 83 ayirulent mutants of Salmonella tvphimurium which no longer could survive in target macrophages. These fell into several classes: one group had increased sensitivity to lysosomal oxidation; another had minimal resistance to cationic proteins in phagocytic granules; and, others were auxotrophs and required nutrients not available intracellularly. Pathways commonly altered included purine, pyrimidine, aromatic amino acid, methionine, and histidine biosyntheses.
Other important phenotypes for intracellular survival included motility, intact O-side chains, and an intact virulence plasmid. Molecular mechanisms by which these characteristics confer virulence have yet to be established. 48
Iron siderophores
Iron is an essential nutrient for bacterial development and growth (Martinez et al. 1990). Free iron is rare in mammalian systems since most is complexed by iron-binding proteins such as transferrin and lactoferrin.
Bacteria counter this by producing iron-solubilizing and iron-transporting molecules called siderophores (Yancy et al. 1979). Salmonella tvphimurium mutants deficient in siderophore production have reduced virulence when injected intraperitoneally in mice (Yancy et al. 1979). In contrast, Benjamin et al. (1985) determined siderophore production was necessary only for growth in mouse serum and had no effect on virulence when bacteria were injected into mice. Stocker and Makela (1986) suggest iron is readily available once Salmonella are intracellular and thus siderophore production is not essential.
Genetic control of virulence
Bacterial genes necessary for the entry of Salmonella into cells are poorly characterized. Salmonella typhimurium has a genome of approximately 4.5 million base pairs which corresponds to about 3 000 genes (Groisman and
Saier 1990). Transposon induced mutants tested for virulence in mice indicate that four percent of the genome, or roughly 120 genes, code for virulence associated structures (Fields et al. 1986) . Mutants have demonstrated some phenotypes associated with virulence but little of the 49 underlying genetic control is known (Finlay and Falkow
1989c; Groisman and Saier 1990; Fields et al. 198 6).
A chromosomal regulatory locus known as phop, which encodes a transcriptional regulator protein, has been described (Fields et al. 1989). The phoP product has homology with other regulator proteins including PhoB and
OmpR which are two-compartment regulatory proteins (Miller et al. 1989a). A second gene downstream from phoP is phoQ which encodes a protein with kinase activity. It is thought PhoQ is the kinase which phosphorylates PhoP and that phosphorylated PhoP exerts transcriptional activity over several genes.
PhoP is known to activate at least 5 genes: phoN, which codes for an acid phosphatase, psiD, a phosphate starvation inducible gene, and pagA, pagB, and page, which code for proteins of unknown function (Groisman and Saier
1990). Mutations in the phoP activator result in mutants with increased susceptibility to defensins (Fields et al.
1989). These mutants enter and replicate normally in epithelial cells but are destroyed by macrophages. This suggests resistance to defensin lysis is at least one phenotype regulated by phoP. Other factors under phop regulation are unknown.
Besides phop, a chromosomal locus termed inV has recently been described (Finlay and Falkow 1989c).
Insertional inactivation of this locus produces mutants 50 which attach to but cannot penetrate epithelial cells. The molecular mechanisms responsible are unknown.
Most Salmonella species have a large (100 kD) plasmid which is essential for successful survival within phagocytes (Finlay and Falkow 1988a). It is not required for attachment and penetration of epithelial cells or colonization of Peyer's patches (Gulig and Curtis 1987).
Plasmid-cured mutants have no variations in LPS from wild- types so other factors must be involved.
Shigella
Bacillary dysentery in humans is caused by enteric infectious with Shigella species and dysenteric strains of
Escherichia coli (Maurelli and Sansonetti 1988a).
Bacillary dysentery is endemic throughout the world but poses serious health concerns only in developing countries.
It affects principally children and is usually self- limiting, but can be life-threatening in infants, the elderly, or in malnourished individuals (Hale and Formal
1987) . Shigella are also important pathogens in subhuman primates (Mulder 1971)
The critical event in the pathogenesis of bacillary dysentery is invasion of the colonic mucosa by the bacteria. Shigella species and enteroinvasive Escherichia coli utilize the same mechanisms for invasion, even at the genetic level (Finlay and Falkow 1988e; 1989a), and in this 51 discussion will be considered together. A general description of mucosal invasion followed by a discussion of possible virulence factors is provided.
Infection is by oral ingestion of bacteria and the infection dose can be alarmingly small, 10-100 bacteria
(Clerc et al. 1988). Transmission is usually by the fecal to oral route. Target cells for penetration and replication are the large intestinal columnar epithelium.
Electron microscopic studies have shown Shigella access epithelial cells in a manner which is morphologically similar to that described for Salmonella
(Finlay and Falkow 1989a). Bacteria contact the epithelium and are enclosed in a vacuole as they advance through the apical membrane. Unlike Salmonella, Shigella rapidly lyse the vacuolar wall and enter the cytoplasm. In virulent strains, lysis of the vacuole occurs in 15 to 3 0 minutes post cellular contact (Sansonetti et al. 1986). Once within the cytoplasm, bacteria replicate to high numbers with a generation time of approximately 40 minutes (Clerc et al. 1988). Within 4 to 6 hours post contact, cells containing bacteria lyse and neighboring cells are infected
(Hale and Formal 1987). Cell lysis produces the colonic ulcers in clinical disease and the epithelial plaques in monolayer cell cultures. In the colon, infection is limited to the superficial epithelium and lamina propria except in immune-suppressed individuals who may become 52 bacteremic (Struelens et al. 1985). As in Salmonella, virulence factors involved in invasion are only partly character i z ed .
The colonic mucosal epithelial cell is the site of bacterial invasion which indicates the bacteria can survive the low stomach pH and small intestinal pancreatic enzymes and bile salts. The LPS component of the outer membrane is partially responsible for protecting the bacteria from bile salt solubilization (Sansonetti et al. 1983a); rough colonies lacking LPS are more sensitive to bile salt degradation.
Pancreatic enzymes may have a role in colonic epithelial trophism. Virulent Shigella, treated with trypsin or chymotrypsin, are non-invasive to HeLa cell monolayers suggesting that outer membrane proteins are involved in attachment (Hale and Formal 1987).
Invasiveness returns shortly after removal of the enzymes.
High levels of pancreatic enzymes in the small intestine may prevent attachment until new membrane proteins can be synthesized (Hale and Formal 1987).
Attachment
Shigella lack flagella, are not motile, and are incapable of chemotaxis (Hale and Formal 1987). How do these bacteria penetrate the intestinal mucus layer and attach to epithelial cells? Shigella have a mucinase and an a-galactosidase (Formal and Lowenthal 1956; Prizont 1982), but alone these do not allow for mucus penetration
(Hale and Formal 1987). Prizont and Reed (1980) showed that Shigella flexneri is incapable of attaching to intestinal sections in gnotobiotic mice but will attach if these sections are exposed to normal fecal material.
Similarly, in the absence of normal flora, monkey colon explants are refractory to Shigella infection (Hale and
Formal 1987). These findings imply that enzymes produced from normal bacterial flora are required for Shigella to penetrate the mucus barrier and contact epithelial cells.
The physiologic uptake of water by. the large intestine may also assist Shigella in epithelial contact (Hale and
Formal 1987). Bacteria become more concentrated and have greater opportunity for epithelial contact as water is absorbed.
The nature of initial binding to epithelial cells is unknown. No evidence of fimbrial attachment has been seen
(Finlay and Falkow 1988d) and it is not known which, or if, host receptors are involved (Finlay and Falkow 1988e). A plasmid encoding for afimbrial attachment proteins has been described in a pyelonephritic strain of Escherichia coli, and when the plasmid is transferred to Shigella, make the
Shigella highly adherent to HeLa cells (Labigne-Roussel
1984) . The significance of this in colonic infections is unknown. 54
Inhibitors of receptor cycling such as dansylcadaverine, which reduce invasion by Salmonella in cell cultures, have no effect on invasion by Shigella
(Finlay and Falkow 1988d). This suggests Shigella may use receptors with slower turnover rates. The trypsin induced loss of invasiveness indicates outer membrane proteins are likely involved (Hale and Formal 1987), but definite interactions are not known.
Penetration
Like Salmonella, the internalization of Shigella in eukaryotic cells is thought to occur by parasite mediated or directed endocytosis. Host microfilaments and cellular energy expenditure are required for internalization (Hale and Formal 1987; Maurelli and Sansonetti 1988a; Finlay and
Falkow I988d; Finlay and Falkow 1988e; Finlay and Falkow
1989a). Following bacterial attachment, polymerized actin appears within 6 minutes beneath the apical plasma membrane
(Clerc et al. 1988). Like Salmonella, penetration can be inhibited by cytochalasins B and D which block actin polymerization. Microtubules and intermediate filaments are not involved and endosome acidification is not required.
An alternative to parasite binding of surface receptors to initiate internalization has been proposed
(Sansonetti et al. 1986). Shigella possess a cell associated contact hemolysin which may focally damage the 55 apical plasma membrane. The damaged membrane, with attached bacteria, may be internalized by "a natural plasma membrane repair mechanism" (Hale and Formal 1987). This cell associated hemolysin may also be responsible for the rapid lysis of the vacuole following penetration.
Intracellular multiplication
To replicate intracellularly, Shigella must escape the endocytotic vacuole. Strains lacking this property do not replicate and do not invade adjacent epithelial cells (Hale and Formal 1987; Finlay and Falkow 1989a). Lysis is attributed to the contact hemolysin even though it is not a phospholipase (Hale and Formal 1987). Escape from the vacuole is critical because it blocks phagolysosomal fusion and places the bacteria in the nutrient rich cytoplasm
(Clerc et al. 1988).
Disruption of cellular functions occurs rapidly once
Shigella access the cytoplasm. Clerc et al. (1987) reported altered mitochondria adjacent to bacteria within
15 minutes post exposure. This correlates with observed decreases in cellular ATP, a 60% decrease by 30 minutes post infection (Sansonetti and Mounier 1987). This loss of
ATP is accompanied by complete cessation of cellular protein synthesis (Finlay and Falkow 1989a) and fermentation (Sansonetti and Mounier 1987). In addition, cAMP levels are elevated in infected cells severely disrupting cellular metabolism. The net effect of 56
bacterial cytoplasmic access is paralysis of cellular
metabolism which frees cytosolic nutrients for use by the
bacteria.
Within 4 to 6 hours, large numbers of Shigella are
within the cytoplasm and infected cells begin to lyse. By
48 hours post infection, grossly visible plagues are seen
in monolayer cell cultures (Oakes et al. 1985). The mechanism of cell lysis has not been defined but the
contact hemolysin may again be involved (Hale and Formal
1987). In addition, paralysis of cellular metabolism and
the production of bacterial toxins may also play a role in
cell lysis leading to bacterial escape.
Toxins
Toxins produced by Shigella include shiga toxin, produced by Shigella dvsenteriae, and shiga-like toxin, produced by all other species (Hale and Formal 1987).
These toxins vary in one amino acid due to 3 base pair differences in the transcribed chromosomal DNA (Maurelli
and Sansonetti 1988a). They cross react serologically and
enzymatically (O'Brien et al. 1977). Shiga toxin is composed of six subunits: an enzymatically active, 32 kD, A subunit, and five identical, 7.7 kD, B subunits which are responsible for cellular binding (Maurilli and Sansonetti
1988a). Shiga toxin is a highly specific N-glycosidase which depurinates adenine 4324 near the 3' end of the cellular 28 S RNA, one of the 60 S ribosomal subunits 57
(Maurelli and Sansonetti 1988a). This disables ribosomal
translation of mRNA halting cellular protein synthesis.
The importance of shiga toxin to bacterial virulence
is less clear. Shigella dvsenteriae produces the highest
amounts of shiga toxin and is associated with the most
severe disease, implying shiga toxin is a virulence factor.
Moreover, an early event in bacillary dysentery is small
intestinal secretory diarrhea. If the small intestine is not exposed to bacteria (inoculum injected straight into the colon) secretory diarrhea is not observed (Rout et al.
1975). This suggests shiga toxin is responsible since
Shigella do not adhere to or invade the small intestine.
However, shiga toxin alone cannot produce disease since noninvasive but highly toxigenic strains fail to induce diarrhea when given orally (Gemski et al. 1972a).
Shiga toxin does not appear to be a factor in attachment and penetration since non-toxigenic strains invade cell cultures as efficiently as do wild-type strains (Clerc et al. 1987; Sansonetti and Mounier 1987).
Maurelli and Sansonetti (1988a) propose the toxin acts in three ways: as an enterotoxin, blocking cellular protein synthesis; as a neurotoxin, producing the observed megacolon in clinical infections by damaging the vas nervosum in the colon; and, as a vascular toxin, causing ischemia of the intestine by local or systemic diffusion into small vessels disrupting endothelial integrity. 58
Taken together, the role of shiga toxin in invasion of colonic epithelium is unclear. However, it is an established virulence factor since it contributes significantly to the severity of disease.
The role of endotoxin (LPS) in the pathogenesis of
Shigella infections is also uncertain, but LPS appears necessary for intracellular replication. By constructing intergeneric hybrids, Gemski et al. (1972b) showed that
Shigella flexneri expressing one type of O-antigen could produce keratoconjunctivitis in guinea pigs (positive
Sereny test), but bacteria expressing a different O-antigen type could not. Effects on attachment and penetration are probably not responsible for observed differences in virulence since rough mutants can invade HeLa cells
(Okamura et al. 1983). This suggests intracellular replication may be responsible; the ability to replicate intracellularly may vary with differing O-antigens.
Iron siderophores
Two different siderophores are reported in Shigella species: the phenolate, enterobactin, which is common in many enteric bacteria, and the hydroxamate, aerobactin
(Lawlor and Payne 1984). Lawlor et al. (1987) tested aerobactin negative transposon mutants for invasiveness and concluded siderophore production has no effect on invasiveness or intracellular replication. It is possible siderophores have a role in extracellular growth and hence 59 are virulence factors. This is supported by ileal loop infection studies using aerobactin negative mutants. These mutants produced significantly fewer mucosal lesions then wild-type bacteria. Since conflicting data exist, the role of siderophores in Shigella virulence is questionable.
Genetic control of virulence
All virulent Shigella and enteroinvasive Escherichia coli contain a 180-220 kilobase plasmid which contains all genes necessary for bacterial invasion of eukaryotic cells
(Sansonetti et al. 1981; 1982; 1983b). Hybridization studies on plasmids from various strains show high levels of homology (Sansonetti et al. 1983b). Two dimensional gel electrophoresis has shown eight unique polypeptides: designated IpaA-G, for invasion plasmid antigens, and pl40
(Hale et al. 1985; Hale and Formal 1986). The proteins vary from 140 kD (P140) to 20 kD (IpaG). Peptides A-D are immunogenic indicating their location on the outer membrane
(Hale et al. 1985). Baudry et al. (1987) demonstrated that peptides B - D are determinants of virulence using insertion mutants. These plasmid encoded antigens may constitute portions of the bacterial ligand responsible for cell attachment since solubilized preparations of these peptides adhere to the surface of HeLa cells (Hale and
Formal 1987).
In addition to the ipa genes, two other plasmid loci, termed virF and virG have been identified (Makino et al. 60
1986; Sakai et al. 1986a). The virF region codes for three polypeptides whose quaternary structure forms 6-sheets.
This gene confers the congo red binding phenotype on
Shigella and also affects virulence (Sakai et al. 1986b).
The products of the virG region have not been determined; virG mutants can invade and multiply within epithelial cells but are unable to infect neighboring cells
(Makino et al. 1986). It is postulated bacteria defective in virG are unable to lyse the cell plasma membrane permitting escape.
Three chromosomal regions have also been shown to affect virulence. DNA located between the xyl and rha loci encode for aerobactin in Shigella flexneri (Griffiths et al. 1985), and for shiga toxin in Shigella dvsenteriae
(Timmis et al. 1985). Both are potential virulence factors. The second region is near the his locus and is linked to the rfb gene cluster (Strum et al. 1986).
Expression of this cluster is necessary for complete 0- antigen biosynthesis on the outer membrane LPS. The third area is the kcp locus named for its ability to cause keratoconjunctivitis provocation in guinea pigs (Formal et al. 1971). The molecular basis for expression of this phenotype is not known.
Regulation of virulence genes
A virulence regulon, a number of diverse, unlinked genes which have a common regulatory signal, has been defined for Shigella (Maurelli and Sansonetti 1988a), and the regulatory signal is environmental temperature.
Strains grown at 37 °C are virulent and invade eukaryotic cells. The same strains grown at 30 °C are avirulent
(Maurelli et al. 1984). Virulence is restored if the temperature is returned to 37 °C. Insertional mutagenesis studies by Maurelli and Sansonetti (1988b) identified a virulent mutant that was not temperature sensitive. The dysfunctional gene was called virR and is thought to encode a repressor protein. When low temperatures are present, transcription is active, the repressor protein is produced, and virulence genes are not expressed. At body temperature, the repressor is not produced and virulent phenotypes result. This allows Shigella to economize on protein synthesis by repressing expression of virulence proteins until needed (Maurelli and Sansonetti 1988a).
Yersinia
The Yersinia species are pathogenic to a variety of animals and man (Ewing et al. 1978; Cornells et al. 1987).
Species affecting man are best characterized and include
Yersinia pestisf etiological agent of bubonic plague, and
Yersinia pseudotuberculosis and Yersinia enterocolitica which cause gastroenteritis and mesenteric lymphadenitis but rarely systemic infections (Cover and Aber 1989).
Infections by enteroinvasive Yersinia result from oral ingestion of bacteria, usually with contaminated food and water (Brubaker 1972). Bacteria pass through the stomach and adhere to and penetrate the distal small intestinal epithelium, including the epithelium covering Peyer's patches (Une 1977a). Eventually, bacteria escape the epithelium, enter the lamina propria, and are taken up by resident and recruited inflammatory cells in which they multiply (Une 1977a; Miller et al. 1988), Infected leukocytes localize in the mesenteric lymph nodes; rarely the spleen and liver are subsequently infected.
Mucosal invasion
Like Salmonella and Shigella, the first and critical step in the pathogenesis of Yersinia caused enteric infections is invasion of intestinal epithelial cells
(Isberg 1990). Attachment and penetration of Yersinia in intact intestines and cell culture monolayers has been described using transmission and scanning electron microscopy (Bovallius and Nilsson 1975; Lee et al. 1977;
Une 1977a; Une 1977b).
Following contact with Yersinia, eukaryotic cells extend protoplasmic projections encircling the bacteria.
The apical cytoplasm then indents and the bacteria is internalized within a vacuole, a process morphologically similar to that described for Salmonella and Shigella.
Coated pits are occasionally observed with the bacteria but vacuoles free of coated pits are common. Since portions of 63 the apical membrane form the phagocytic vacuole, the association with coated pits may be by chance alone {Miller et al. 1988).
The bacteria reside in vacuoles with no detectable increase in numbers for at least 6 hours (Small et al.
1987), and then slowly begin to replicate (Finlay and
Falkow 1988d). The vacuoles gradually increase in size but no vacuolar fusion occurs (Finlay and Falkow 1989a).
Finlay and Falkow (1988d) studied the infection of cell lines with Yersinia enterocolitica. They were unable to observe bacteria exiting host cells, although culture results confirmed passage. In experimental intestinal infections, bacteria pass through the intestinal epithelium and enter the lamina propria where they are phagocytized by reticuloendothelial cells (Une 1977a).
Uptake in resident macrophages or recruited neutrophils appears to occur by classic phagocytosis
(Miller et al. 1988). Like Salmonella, internalized bacteria remain within cytoplasmic membrane-bound vesicles and replicate (Une 1977a; Une 1977c). Phagolysosomal fusion and leukocyte degranulation occur normally in macrophages containing Yersinia oestis. but the oxidative burst is reduced (Charnetzky and Shuford 1985). Une
(1977b) studied the interaction of Yersinia enterocolitica with mouse peritoneal macrophages and determined numbers of viable phagocytosed bacteria initially decreased, but after 64
3 hours steadily increased. Responsible mechanisms are
unclear; it is believed the synthesis of products allowing
for intracellular survival are induced upon phagocytosis
(Cornelis et al. 1987).
Mechanisms of attachment and penetration of the
mucosal membrane, intracellular replication, resistance to
serum killing, and intra-leukocyte survival are only
partially characterized. Summaries of possible virulence
factors for each step follow.
Chemotaxis and motility
Yersinia species are motile at 28 °C by peritrichous
flagella, but not at 37 °C (Kihlstrom and Magnusson 1983).
Motility may aid the bacteria, in cell contact and
consequently act as a virulence factor. However, since
motility is lost at body temperature, the importance is
questionable. Centrifugation of Yersinia onto cell
monolayers does not increase the total number of bacteria
passing through the monolayer (Vesikari et al. 1982).
Moreover, no chemotactic attraction to host cells (directed
motility) has been demonstrated (Miller et al. 1988).
Attachment and penetration
The attachment and penetration of Yersinia in cell
monolayers occurs in minutes (Brunius and Bolin 1983).
Attachment and penetration have common or interrelated
components and are difficult to separate experimentally
(Isberg and Falkow 1985). A large complement of cell lines 65 are invaded by Yersinia, and in some cell lines nearly 100 percent of applied bacteria invade (Miller and Falkow
1988). This suggests factors necessary for attachment and penetration are common among many cell types, or that multiple mechanisms for entry are employed by Yersinia. In fact, common cell receptors and multiple bacterial entrance mechanisms are involved (Isberg 1990).
In 1987, Isberg et al. (1987) isolated a 103 kD, outer membrane, surface-exposed protein (invasin) that allows
Yersinia to attach and enter eukaryotic cells. Latex beads coated with invasin efficiently enter cells, indicating invasin binding alone stimulates cellular uptake (Isberg
1990). The DNA segment encoding invasin has been cloned and transforms non-invasive strains of Escherichia coli into invasive strains (Isberg and Leong, 1988). Monoclonal antibodies against invasin neutralize this effect. Invasin has subsequently been shown to mediate entry into large numbers of mammalian cell types, indicating its non specificity in promoting cellular uptake (Isberg 1990) .
What cellular receptors are on multiple cell lines which interact with invasin to promote cellular uptake? A clue was provided by Finlay and Falkow (1988e) who showed microfilament polymerization was required for bacterial uptake by eukaryotic cells. This implied the cell receptor recognizing invasin was linked to the cytoskeleton. To identify cell receptors, Isberg (1990) bound invasin to 66 agarose beads and exposed these to cell surface extracts.
Bound proteins were eluted and examined by reducing and non-reducing polyacrylamide gel electrophoresis. Bound proteins were heterodimeric glycoproteins composed of large and small subunits (140-160 kD and 115 kD respectively).
Distinctive migration patterns between reducing and non reducing gels suggested the proteins were members of the integrin family of receptors.
Integrins are cell surface proteins which connect the cytoskeleton with the extracellular matrix (Hynes 1987).
Different integrins bind to collagen, fibronectin, laminin, vitronectin, fibrinogen, and certain cell adhesion molecules (Hynes 1987). Each integrin is a a/B dimer and is grouped into subfamilies based on differences in the B subunits. Immunoprobing with anti-integrin antibodies showed invasin binds four members of the Bx subfamily, namely a^fl,, a6Blr cis&i, which is the integrin that binds fibronectin, and a4B1, a receptor on T-cell lymphocytes
(Isberg 1990).
Regulation of invasin expression is temperature controlled (Isberg et al. 1988). The amount of invasin on the bacterial surface decreases with increasing temperature. Thus, bacteria cultured at 28 °C enter cells
10 times more efficiently than those cultured at 3 7 °C
(Isberg et al. 1988). At environmental temperatures,
Yersinia produce abundant invasin and are primed for 67 cellular uptake. At body temperatures, invasin synthesis is inhibited. This implies rapid cellular uptake, in a variety of cell types, is important early in infection, but may be a disadvantage later. Interestingly, binding of the fibronectin integrin a 5/3! on macrophages results in phagocytic activation. Continued expression of invasin could activate phagocytic cells resulting in bacterial destruction.
Recently, another outer membrane protein has been shown to mediate binding and internalization of Yersinia in eukaryotic cells (Miller and Falkow 1988). Like invasin, the locus encoding this protein, termed ail for attachment invasion locus, has been cloned and transforms previously innocuous Escherichia coli into a strongly adherent and invasive strain (Miller and Falkow 1988). Ail is surface exposed and has a molecular weight of 15 kD.
Unlike invasin, Ail binds selectively to cell lines, promoting entry in some, but only attachment in others
(Isberg 1990). Moreover, serum factors are required for invasion in certain cell lines. It is suggested Ail mediated entry requires binding of both bacterial and serum components to cell receptors (Isberg 1990). Cell lines lacking the serum receptor are unable to internalize the bacteria. However, certain cell lines, CHO for example, can rapidly bind and internalize bacteria in the absence of serum. 68
Like invasin, the synthesis of Ail is temperature dependent (Isberg 1990), but is expressed in an opposite manner. The synthesis of Ail is induced at 37 °C and inhibited at lower temperatures. This implies a sequential expression of proteins on the bacterial surface: invasin in the cool extracellular environment, and Ail when bacteria are internalized. Currently, invasin is thought to mediate uptake of Yersinia across epithelial surfaces (oral, intestinal) and Ail is believed to mediate spread of the bacteria in host tissues (Isberg 1990).
Isberg (1990) indicated a third pathway for entry into eukaryotic cells. Bacteria with defective invasin synthesis can be induced to invade monolayers if a virulence associated plasmid is introduced into the mutants. This pathway is not temperature controlled, in contrast to invasin and Ail mediated entry. It is not known what plasmid encoded products are involved. Isberg
(1989) suggested the plasmid encodes for a cytotoxin and that cell cytotoxicity induces bacterial uptake.
In addition to the above mentioned plasmid encoded product, a number of plasmid encoded outer membrane proteins are implicated in attachment and intracellular survival (Martinez 1983). Five outer membrane proteins, known as plasmid outer membrane proteins (POMPs), are expressed in Yersinia pseudotuberculosis and Yersinia enterocolitica (Portnoy et al. 1981; Portnoy et al. 1984). 69
A number of virulence phenotypes are attributed to these outer membrane proteins, although underlying mechanisms are unknown. These phenotypes include autoagglutination, hydrophobicity, serum resistance, and resistance to phagolysosomal destruction (Cornelius et al. 1987).
One of the outer membrane proteins, Pt (160 kD) is implicated in attachment. This protein forms a fibrillar matrix around the bacteria that is thought to interact with the cell surface. Hydrophobicity may also be involved, since this affects bacteria cell surface interactions
(Cornelius et al. 1987).
Transcellular passage
Mechanisms by which Yersinia exit epithelial cells are unknown, but like Salmonella it is thought to be a transcytotic event (Finlay and Falkow 1988d). Yersinia can pass through polarized monolayers and leave the monolayer intact. Yersinia exiting the basilar surface have not been observed; passage has been confirmed by culture (Finlay and
Falkow 1988d). Intestinal infections have also failed to show exiting cells; bacteria are seen in the lamina propria indicating successful escape. These bacteria are rapidly engulfed by phagocytic cells (Une 1977a).
Serum and phagocyte resistance
Resistance to serum lysis by Yersinia enterocolitica has been correlated with plasmid encoded outer membrane proteins, and treatment with trypsin abolishes this 70 protection (Martinez 1983). This has not been shown for
Yersinia pseudotuberculosis. In Yersinia enterocolitica. mutants lacking the P, fibrillar matrix are serum sensitive implying this protein may play a major role (Balligand
1985). Other POMPs are also involved since mutants expressing only ?! are serum sensitive (Balligand et al.
1985).
Plasmid encoded products are also responsible for resistance to phagocytic destruction (Lian and Pai 1985).
Plasmid bearing bacteria were estimated to be four- to six fold more resistant to phagocytic destruction then non plasmid bearing strains. This effect was partially blocked by protease treatment indicating the role of surface proteins.
Toxins
Yersinia enterocolitica produces an enterotoxin which is heat stable and has been compared to the ST toxin of
Escherichia coli (Pai and Mors 1978; Boyce et al. 1979).
It is unknown what, if any role, this toxin has in pathogenesis since non-pathogenic, environmental strains also express it (Pai et al. 1978).
Iron siderophores
Iron availability influences the outcome of infections by several serotypes of Yersinia enterocolitica. Serotypes
0:3 and 0:9 have a 100,000-fold reduction in 50% lethal dose in mice if the animals are pre-treated with 71 desferroxamine B, an iron supplement (Robins-Browne and
Prpic 1985) . By contrast, serotype 0:8 has no additional iron requirement for virulence (Carniel et al. 1987). The
0:8 serotype has two high molecular weight outer membrane proteins believed to be involved in iron sequestration which are lacking in the 0:3 and 0:9 serotypes (Carniel et al. 1987). Both aerobactin and enterochelin are used by yersinia to obtain iron (Perry and Brubaker 1979).
Genetic control of virulence
The genes encoding iron related proteins, and the attachment proteins invasin and Ail, are chromosomal
(Carniel et al. 1987; Isberg 1990). The invasin locus
(inv) is 3.2 kilobases and codes for only one protein, invasin. The ail locus is only 650 base pairs and again encodes for a single outer membrane protein. Are these loci associated primarily with virulent bacteria or do non pathogens also possess them?
Hybridization studies using restriction fragments from both genes have been performed (Miller et al. 1989b). For the inv locus, five patterns of probe hybridization were observed but only two patterns (TI and TI/II) were seen in virulent strains (Miller et al. 1989b). Of tested strains showing one of these two patterns, 90% were virulent. The ail locus is even a better prognostic measure of virulence.
A 98% correlation between ail hybridization and invasive 72 phenotypes is seen (Miller et al 1989b). Invasiveness is clearly correlated with possession of these loci.
In addition to chromosomal loci, a 70 kilobase plasmid is required for virulence (Portnoy and Martinez 1985). The plasmid is not required for invasion but is necessary for intracellular replication and subsequent infection of other tissues (Cornelis et al 1987). As previously mentioned, phenotypes associated with the virulence plasmid include autoagglutination, hydrophobicity, serum resistance, resistance to phagolysosomal destruction, and additionally, congo red binding, mannose resistant hemagglutination, and a nutritional requirement for calcium at 37 °C (Martinez
1983; Kapperud et al. 1985). Expression of these phenotypes are temperature controlled.
The regulatory protein or encoding locus responsible for temperature sensitivity in Yersinia is undetermined, but it is apparent control is well coordinated. As stated earlier, invasin is maximally produced at temperatures near
28 °C and minimally produced at 37 °C (Isberg et al 1988).
Ail has exactly the opposite expression pattern (Isberg
1990). Plasmid encoded phenotypes are also optimally expressed at 37 °C verses 28 °C (Martinez 1983; Bolin et al. 1985).
The low intra-cellular calcium concentration in eukaryotic cells may also be an important regulatory signal. Plasmid encoded protein synthesis is sensitive to 73
calcium levels and is inhibited at extracellular
concentrations (Portnoy and Martinez 1985).
Taken together, bacterial passage from the
extracellular to the intracellular environment results in
the repression of some loci (invasin) and activation of
others (ail and plasmid associated). Like Shigella,
yersinia appears to control these changes by environmental
temperature induced mechanisms. Invasion mechanisms
utilized by Salmonella, Shigella, and Yersinia to invade
intestinal epithelium, and which may be used by
Edwardsiella ictaluri to access fish are highlighted in
table 4.
Study objectives
As previously mentioned, the definitive pathogenesis
of ESC is unknown. The acute form is thought to result
from intestinal invasion, while passage across the
olfactory sac is believed to produce chronic ESC. Initial
goals of this study were to verify the intestine as a route
of infection, to provide a descriptive sequence of events
following enteric exposure, and to clarify mechanisms of
transmucosal passage and dissemination.
In addition, a number of bacterial components in other
enterobacteriaceae are known to be immunogenic (LPS, outer membrane proteins, flagella). An additional goal was to 74
Table 4: Highlights of invasion mechanisms used by Salmonella, Shigella, and Yersinia species.
Genera Salmonella Shigella Yersinia
Cell type Epithelial Colonic Epithelial
entered cells and M epithelial cells and
cells cells M cells
Microfilaments Yes Yes Yes
required
Endosome No No No
acidification
required
Intracellular Vacuole Cytoplasm Vacuole
location
Multiplication Yes Yes Slow
within entero-
cytes
Bacterial metab Yes Yes No
olism required
for entry
Plasmid neces No Yes No
sary for
invasion
Motile Yes No Yes
Chemotactic Yes No Unknown 75
Genera Salmonella Shigella Yersinia
Temperature reg No Yes Yes
ulated invasion
Identified cel No No Integrins
lular receptors
Toxins 1) Cholera 1) Shiga Entero-
like toxin; and shiga- toxin
2) Surface like toxins
toxin; 3) 2) endo
endotoxin toxin
establish the immunogenicity of structural components of E. ictaluri in naturally infected channel catfish.
Specific objectives were: 1) To successfully infect channel catfish by gastric intubation and establish the time required for bacterial passage; 2) To quantify bacterial numbers reaching internal organs; 3) To clarify possible mechanisms of mucosal attachment and invasion; 4)
To determine early pathologic changes associated with bacterial invasion; and, 5) To establish the immunogenicity of whole cell preparations and enriched fractions of E. ictaluri LPS, outer membrane proteins, and flagella in naturally infected channel catfish. CHAPTER II. EARLY EVENTS IN THE PATHOGENESIS OF ENTERIC
SEPTICEMIA OF CHANNEL CATFISH, . CAUSED BY EDWARDSIELLA
ICTALURI: LIGHT MICROSCOPIC AND BACTERIOLOGIC FINDINGS
Introduction
Channel catfish farming has become a major
agricultural industry in the southern United States. As
catfish production has increased, so have catfish losses,
particularly due to disease (Catfish Production 1991).
Enteric septicemia of catfish (ESC), caused by the
bacterium EL_ ictaluri, is the most significant disease of
cultured channel catfish (Macmillan 1985), currently
accounting for 34% of all disease cases submitted to the
Louisiana Aquatic Diagnostic Laboratory (J. Hawke,
Louisiana State University, personal communication), and
40% of cases seen at the Mississippi Cooperative Extension
Service (M. Johnson, Mississippi State University, personal
communication).
Enteric septicemia of catfish occurs in acute and
chronic forms (Johnson 1989). The acute form is a
bacterial septicemia characterized by hemorrhage and
necrosis in multiple organs. The chronic form is a meningoencephalitis with dorsal extension through the sutra
fontanel of the skull. The pathogenic mechanisms underlying the development of these forms are unknown.
76 77
Shotts et al. (1985) and Newton et al. (1989) both proposed
that intestinal penetration by EL. ictaluri results in the
acute form of ESC, and that passage through the olfactory
sac and the subsequent ascending olfactory neuritis and
perineuritis produces chronic ESC. However, bacterial
passage across either mucosal barrier has not been
documented.
The objectives of this study were: 1) to confirm the
intestine as a site of bacterial entry and establish
bacterial access times; 2) to identify intestinal
morphologic changes associated with bacterial penetration;
3) to identify early morphologic changes in other tissues;
and, 4) to quantify bacterial numbers accessing internal
organs.
Materials and methods
Bacteria
Edwardsiella ictaluri. strain LA 89-9, a virulent
isolate obtained from a natural epizootic of ESC, was used
for experimental infection. Virulence was ensured by four
serial passages of EL. ictaluri, strain LA 89-9, in channel
catfish fingerlings using an intragastric route of
infection (1.0 X 109 colony forming units [CFU]/fish).
Twenty-four hours PI, fingerlings were sacrificed and the trunk kidneys cultured. Resulting bacterial colonies were used to reinfect new fingerlings. Bacteria were then 78
stored at -70 °C in brain heart infusion broth (BHIB, Difco
Laboratories, Detroit, MI) containing 20% glycerol until needed for infection trials. Culture identity was routinely verified using standard biochemical tests (Farmer
& McWhorter 1985).
Maintenance of channel catfish fingerlings
Enteric septicemia free fingerling channel catfish
(10-12 cm, 9-13 g ) , hatched from disinfected eggs and raised in ESC-free recirculating systems were used.
Fingerlings were maintained in 115 L glass aquaria under
slow, flow-through conditions (two water changes/day), using dechlorinated Baton Rouge city water with supplemental filtration provided by in-tank, air-lift sponge filters. Water temperature was maintained at 26 ± 2
°C with in-tank, submersible, aquarium heaters. Water quality (ammonia, nitrites, pH) was monitored biweekly using a commercially available kit (Hach Company, Loveland,
CO). Nitrite and ammonia levels were consistently below 1 part per million; water pH was 8.0. Fish were fed a commercially prepared ration (Zeigler Brothers,
Incorporated, Gardners, PA) three times a week.
Infection trials were done under the conditions described above but with flow through rates increased to four water changes per day and the in-tank sponge filters removed (to avoid filter contamination). Water quality was the same as in maintenance tanks. 79
Experimental infection
Frozen virulent bacteria were cultured on brain heart infusion agar (BHIA, Difco Laboratories) for 24 hours at 26
°C. Single colonies were transferred to BHIB and incubated for 18 hours at 26 °C with moderate shaking (100 revolutions/minute) in a water bath. Bacteria were harvested by centrifugation and resuspended in 1% sterile peptone broth (Difco Laboratories), pH 7.2, to a concentration of approximately 5.0 X 109 bacteria/mL diluent, as measured by spectrophotometry. Prior plate counts were used to established the correct spectrophotometric density.
All fingerlings were inoculated by intragastric intubation using an 18 gauge, stainless steel, animal feeding needle (Popper and Sons Incorporated, New Hyde
Park, NY). Each fingerling, except control fish, received
200 fill of the bacterial suspension (approximately l.o X 109 bacteria). Control fish received peptone broth only. Fish were not anesthetized prior to intubation. Following intubation, fish were returned to tanks and watched closely for regurgitation. Only rarely was regurgitation observed and these fish were reintubated.
Tissue collection
Ten fish were sacrificed at time 0 to serve as controls. Infected fingerlings were sacrificed at 0.25,
0.5, 1, 3, 6, 12, 24, 48, 72, 96, and 120 hours PI. From 80
0.25 to 3 hours PI, twenty fish per time period were sampled (individual fish were infected and sacrificed before the next fish was infected to allow for strict adherence to sampling time frames). After 3 hours, ten fish per time period were sampled. Fish were euthanized by exposure to 0.5% 3-aminobenzoic acid (Methanesulfonate salt, Sigma Chemical Company, St, Louis, MO) for 5 minutes
(time included in total infection time) and necropsied.
Samples of liver, trunk kidney, stomach, and intestine were sterilely collected for bacterial quantitative culture. Remaining portions of intestine (from throughout the tract), liver, trunk kidney, spleen, and head kidney were fixed in 1.25% glutaraldehyde and 2% paraformaldehyde in 0.1 M sodium cacodylate buffer, pH 7.2, at room temperature. Tissues were gently agitated in fixative for
5 hours, rinsed twice for 3 0 minutes in cacodylate buffer, and stored in buffer at room temperature for several days prior to processing.
Tissue processing
Fixed tissues were dehydrated by immersion in graded ethanol (3 0% to 100%), infiltrated and embedded in LR-white resin (London Resin Company, Basingstoke, Hampshire, UK), and sectioned at 0.5 /mt on a Sorval MT 5000 ultra microtome. Sections were stained with 1% toluidine blue in
1% sodium borate for 2 minutes, washed in deionized water, and counter stained with 1% basic fuchsin for 3 0 seconds. 81
Quantitative cultures
Quantitative cultures of liver and trunk kidney were done on brain heart infusion agar (BHIA, Difco
Laboratories) at 26 °C. Quantitative cultures of stomach and intestine, including contents, were done on a media selective for Eh. ictaluri (Shotts and Waltman 1990) at 26
°C. Tissues were aseptically weighed and then homogenized in 7 mL glass tissue grinders (Kimax, Toledo, OH) with 1 mL added sterile peptone broth. Serial dilutions were immediately made in peptone broth and 40 /iL, from each dilution plated. Culture plates were incubated for 3 days and CFU were counted. Bacteria per gram of tissue were calculated.
Rabbit anti-E. ictaluri sera
To identify Eh ictaluri by immunostaining in tissue sections, polyclonal antisera specific for Eh. ictaluri was produced. Virulent Eh. ictaluri were cultured in BHIB and harvested by centrifugation. The bacterial pellet was suspended at a concentration of approximately 1% in 3.7% neutral buffered formaldehyde in 0.85% NaCl, and incubated with gentle rocking at 4 °C for 12 hours. Fixed bacteria were pelleted by centrifugation, washed three times by suspension in 10 mM Hepes buffered saline, pH 7.4, for 5 minutes, and recentrifuged. The bacterial mass was dried at 32 °C, finely ground with a mortar and pestle, and 82 resuspended in phosphate buffered saline (PBS), pH 7.2, at
1 mg/mL.
One mL of the bacterin was emulsified with 1 mL complete Freund's adjuvant (Difco Laboratories) and administered subcutaneously, at multiple sites, to a New
Zealand white rabbit. The first booster, of similar volume but in incomplete Freund's adjuvant (Difco Laboratories), followed in 3 weeks. Three subsequent boosters were given at 3 week intervals. Twenty mL of blood were collected from the auricular artery 3 weeks following the last booster. The serum was harvested and stored at -70 °C. A dilution titer of 1:10,000 was obtained by enzyme linked immunosorbant assay (ELISA) with sonicated, formalized E. ictaluri as antigen. Antibody specificity was checked by
ELISA using sonicated formalized Escherichia coli.
Aeromonas hydrophila. and IL_ tarda as antigens.
Escherichia coli and Aeromonas hydrophila coated plates had no cross reactivity. Edwardsiella tarda coated plates had cross reactivity for dilutions of 1:100 and less, but were negative for dilutions greater than 1:100 indicating good antibody specificity for EL. ictaluri. Biotinylated, goat anti-rabbit IgG (Vectastain ABC kit, Vector Laboratories,
Burlingame, CA) was purchased.
Immunohistochemical detection of E. ictaluri
A modified immunoperoxidase staining technique was used to identify E^. ictaluri in tissue sections (Hsu et al. 1981) . Immunoperoxidase staining procedures followed standard protocols with all steps performed at room temperature. Endogenous peroxidase activity was eliminated by immersion in 0.3% hydrogen peroxide for 15 minutes.
Sections were washed and nonspecific binding sites blocked by immersion for 1 hour in 1:100 goat serum diluted in 0.1
M Tris buffer, pH 7.2. Primary rabbit anti-E. ictaluri serum at 1:1000 dilution in 0.1 M Tris buffer, pH 7.2, was applied for 1 hour. After washing, biotinylated, goat, anti-rabbit IgG (Vectastain ABC kit) diluted to 1:1000 in
Tris buffer was applied for 1 hour. Following washing, an avidin-biotin-peroxidase complex (Vectastain ABC kit) was applied for 1 hour. Color development was done using 3-3'- diaminobenzidine tetrahydrochloride (DAB, Sigma Chemical
Company, St. Louis, MO) solubilized in 0.1 M Tris buffer, pH 7.2, containing 0.02% hydrogen peroxide and 0.04% nickel chloride. Color development was monitored visually and stopped by immersion in deionized water. Slides were counter stained in 1% basic fuchsin.
Statistical evaluation
Logarithmically transformed bacterial counts per time period were analyzed using the general linear model for an unbalanced analysis of variance. Logarithmically transformed means were used due to the wide distribution of means. Comparison of means was by Tukey's studentized 84
range test (SAS Institute 1988). Results were considered
significant if P<0.05.
Results
Bacterial cultures
No fish cultured positive for E^ ictaluri in control
fish. At 0.25 h PI, trunk kidneys were positive for E.
ictaluri in six of twenty fish. Positive liver cultures were seen in six of twenty fish at 0.5 h PI, while eleven
of twenty kidney cultures were positive at this time
period. All liver and kidney cultures were positive by 72
hours PI. Logarithmic transformations of mean CFU per g
liver, trunk kidney, stomach, and intestine, for time
periods 0.5 to 96 hours PI, with standard deviations, are
represented graphically in figures 5a and 5b.
Significant increases from all earlier means were seen
in the liver from 48 to 96 hours PI and in the trunk kidney
from 24 to 96 hours. Bacterial numbers in trunk kidney
cultures were higher than those in liver cultures at every
time period. Colony forming units were to numerous to
count in the liver and kidney at 120 h PI. Counts
exceeding 106 bacteria per gram tissue were common in trunk kidney cultures from 48 days PI. In the intestine, CFU at
96 hours PI were significantly higher than those at 6 and
12 hours. No means were significantly different at any time period in stomach cultures. 85
A 9 -O* Stomach 8 Intestine
7
2
1 0 0.51 3 6 12 24 48 96
Time (h)
B 9 -D- Liver 8 Kidney 7
6
u 5 c « 4 a ST3 J 2 1 0 -1 0.5 1 3 6 12 24 48 72 96 Tine (h)
Figures 5a and 5b: Logarithmic transformation of mean CFU/g stomach, intestine, liver, and trunk kidney, with standard deviations. Means significantly different are marked (*). 86
Gross lesions
The earliest gross lesions, seen at 48 hours PI, were petechiae in the skin at the base of the fins and in the liver in three of ten fish. At 72 hours p i , similar changes were observed in six of ten fish. In addition, the intestine was edematous in four of the fish. At 96 hours
PI, five of ten fish had multifocal petechiae in the skin, intestine, liver, and head kidney. In these fish, the liver, trunk kidney, spleen, and intestine were severely edematous and soft. Ascites was common and the blood was thin. Multifocal areas of epidermal hemorrhage and necrosis were present. One fish had a intussusception in the anterior intestine. The intussusceptum was deep red but retractable. By 120 hours PI, nine of ten fish had gross lesions similar to those described above.
Histologic findings
Intestine: All sections from fish at time 0 and 0.25 h
PI were essentially normal. At 0.5 hours PI, bacteria were seen along the brush border of the anterior and middle portions of the intestine (figure 6). These were identified as E^_ ictaluri by immunostains (figure 6, insert). In addition, individual cells of unknown identity, located near or on the basement membrane, were dilated, vacuolated, and contained large intracytoplasmic inclusions (figure 7). These dilated basilar cells were most numerous at tips of intestinal folds, decreased in 87 number along sides of folds, and were rare in fold crypts.
These cells increased in number from 0.5 hours to 24 hours
PI.
In some intestinal sections, between 1 and 3 hours PI, individual enterocytes located on fold tips were necrotic
(figure 8). Adjacent cells were morphologically normal.
Bacteria were often, but not always, associated with necrotic cells. The bacteria were identified as E. ictaluri by immunostains (figure 8, .insert). Prominent dilated basilar cells were always subadjacent to these necrotic cells.
Proprial leukocyte infiltrates were observed rarely within the first 24 hours of infection (5 of 110 fish).
Thereafter, proprial infiltrates were common, thickened the lamina propria, and extended between enterocytes (figure
9). Leukocyte margination in large vessels, mucosal edema, and rarely, proprial hemorrhage were also seen. Numerous bacteria, identified as ictaluri by immunostains, were within leukocytes (figure 9, insert).
At 24 hours PI, most leukocytes containing bacteria had only one or two intracytoplasmic organisms. Within some leukocytes, bacteria were apposed end-to-end and incompletely separated. Few degenerate leukocytes were seen. By 72 to 120 hours PI, leukocytes contained numerous intracytoplasmic bacteria. Numerous inflammatory cells were dilated, had fragmented cytoplasm, and pyknotic or 88
Figure 6: Photomicrograph of anterior intestine 0.5 hours PI with EL. ictaluri and immunostained serial section (insert). Numerous luminal bacteria (arrows) line the brush border (LR-white sections X 720).
Figure 7: Photomicrograph of anterior intestine 0.5 hours PI with EL, ictaluri (LR-white section X 720) . Numerous dilated cells (arrow) subadjacent to the basement membrane are seen near the tips of intestinal folds. Prominent cytoplasmic inclusions are in these cells. No bacteria are seen in the lumen (n).
Figure 8: Photomicrograph of necrotic enterocytes (e) at the tip of an intestinal fold 1 hour PI with EL. ictaluri (LR-white sections X 720). Dilated basilar cells (arrow) are seen. Bacteria are within the lumen (n). The insert is a photomicrograph of an immunostained section which identifies the bacteria (arrow) as EL, ictaluri.
Figure 9: Photomicrograph of anterior intestine and immunostained serial section (insert) 72 hours PI (LR-white sections X 720). Numerous leukocytes (c) have infiltrated the lamina propria and are among enterocytes (e). Leukocytes contain few to numerous EL. ictaluri (arrow). 89
i n 90 absent nuclei. Inflammatory cells extended into the mucosa which was thin but not ulcerated (figure 9).
Bacteria were not observed within or between viable enterocytes in any sections examined. Free bacteria in the lamina propria were rare and always associated with degenerate leukocytes.
Liver: Sinusoid associated leukocytes containing intracytoplasmic bacteria were first observed at 48 hours
PI, and the presence of these cells containing bacteria was the earliest morphologic change observed in the liver
(figure 10) . The bacteria were identified as ictaluri by immunostains (data not shown) and bacteriologic culture.
The number of liver sections containing observable bacteria increased over time; by 120 hours PI, most sections had identifiable bacteria. Leukocytes containing bacteria were seen in larger vessels at 72 and 96 hours PI (figure 11).
Bacteria were not seen in cells that definitively could be identified by light microscopy as hepatocytes.
Soleen. trunk kidney, and head kidnev: In all tissues, multifocal interstitial leukocyte aggregates were first observed 48 hours PI and increased in number and size with time. In the spleen, infiltrates surrounded central arterioles (figure 12). In the trunk kidney, bacteria- laden leukocytes were between renal tubules (figure 13).
With time, aggregates increased in size and leukocytes extended between tubular epithelial cells. At 12 0 hours 91
PI, leukocytes containing bacteria were seen in glomerular capillaries and larger blood vessels. In the head kidney,
leukocytes containing bacteria separated and replaced interrenal and hematopoietic cells. No bacteria were observed in interrenal cells in the head kidney, renal tubular epithelium, or in hematopoietic cells in all organs. In all tissues, the bacteria seen were identified as Ei ictaluri by immunostains (data not shown) and bacteriologic culture.
Discussion
Results of this research indicate ictaluri may have an invasion strategy similar to other enteroinvasive enterobacteriaceae. Both Salmonella and Yersinia species, which cause systemic infections, penetrate the intestinal epithelium of the host and parasitize underlying resident macrophages (Takeuchi 1967; Finlay and Falkow 1988a).
These pathogens disseminate to other tissues within or attached to phagocytes, and replicate within these cells
(Finlay and Falkow 1989a), Salmonella and Yersinia species which are systemic pathogens utilize existing cellular pathways to penetrate the intestine but produce little epithelial necrosis (Finlay and Falkow 1988a).
A similar pathogenesis for infection by E^. ictaluri is supported by the following observations: l) ictaluri crossed the intestinal epithelium and accessed internal -92
Figure 10: Photomicrograph of liver 72 hours PI (LR-white section X 720). Sinusoid associated cells morphologically similar to mononuclear leukocytes contain bacteria (arrows). Bacteria were not definitively identified in hepatocytes (h).
Figure 11: Photomicrograph of a portal vessel 72 hours PI (LR-white section X 720). Within the vessel are leukocytes with cytoplasmic bacteria (arrows). Adjacent pancreatic cells (p) and hepatocytes (h) do not contain bacteria.
Figure 12: Photomicrograph of spleen 72 hours PI (LR-white section X 720). Surrounding the central arteriole (a) are numerous pale staining leukocytes (c). Some leukocytes (arrow) contain cytoplasmic bacteria.
Figure 13: Photomicrograph of trunk kidney 48 hours PI (LR- white section X 720). Moderate numbers of leukocytes (c) have infiltrated between tubules (t). Occasional intracytoplasmic bacteria are seen (arrow) within leukocytes. organs rapidly following exposure by intragastric intubation; 2) Widespread necrosis of the intestinal mucosa was not seen in fish culturing positive for E. ictaluri. Early lesions consisted of dilated basilar located cells and occasional necrotic enterocytes on tips of intestinal folds; 3) In all tissues, bacteria were consistently seen within leukocytes and not other cells; and, 4) bacterial counts were highest in tissues possessing higher numbers of resident macrophages.
By 0.25 hours PI, 3 0% of trunk kidney cultures were positive for E^. ictaluri indicating rapid bacterial passage through the intestinal epithelium. Yersinia species also attach to and penetrate cell membranes in minutes (Brunius and Bolin 1983). Yersinia do not have to be metabolically active to access eukaryotic cells since formalized bacteria are also internalized (Finlay and Falkow 1989a). In contrast, Salmonella do not penetrate cell membranes for several hours PI, and bacterial de novo protein synthesis is required for this organism to invade cells (Finlay et al. 1988b). In this respect, EU. ictaluri is most like
Yersinia, successfully crossing the intestinal epithelium in minutes. This rapid passage suggests E^_ ictaluri may utilize normal cellular transport systems to cross epithelium and does not require de novo bacterial protein synthesis to access fish. In fish culturing positive for E^_ ictaluri. widespread necrosis of the intestinal mucosa was not seen. Earliest histologic lesions were the rapid appearance of numerous dilated basilar located cells with prominent cytoplasmic inclusions, and occasional necrotic enterocytes on villus tips. The identity of the basilar reactive cells is unknown. Perhaps they are resident macrophages, and the cytoplasmic inclusions are phagocytosed cellular debris.
Alternatively, they could be necrotic enterocytes although this is considered unlikely; their consistent basilar location and morphologic characteristics are not typical of enterocytes.
The significance of the necrotic enterocytes observed on some tips of intestinal folds is unknown. They were an irregular finding on sections from 1 and 3 hours PI and were not observed at other time periods. Some, but not all, had associated bacteria. Perhaps the passage of E. ictaluri through some enterocytes results in cellular death. Salmonella and Yersinia species have been shown to replicate in and destroy non-phagocytic eukaryotic cells, although phagocytic cells are principally targeted for infection (Finlay and Falkow 1988c; 1989a).
In the intestine, liver, spleen, trunk kidney, and head kidney, bacteria were consistently seen in leukocytes and the bacteria were commonly apposed end-to-end and incompletely separated. Intracellular replication must be confirmed, however, since the bacteria may have been replicating prior to phagocytosis. Interestingly, it was vascular associated phagocytes which were first seen containing bacteria, and in large vessels, leukocytes with intracytoplasmic bacteria were seen. This suggests that dissemination may be within phagocytes. Rare bacteria were free in the interstitial tissues, but these were usually associated with degenerate leukocytes, suggesting they were formerly intracellular.
Why trunk kidney cultures are more sensitive than liver cultures and consistently have higher bacterial counts is unknown. This difference may be due to increased numbers of resident macrophages in the trunk kidney compared with the liver (Grizzle and Rogers 1976).
In conclusion, this study confirms the intestine as one site of bacterial access in ESC, and establishes that intestinal invasion occurs as early as 0.25 hours PI.
Furthermore, the rapid mucosal passage, the absence of widespread intestinal epithelial necrosis, and the consistent location of E^_ ictaluri within leukocytes are characteristics of other enteropathogenic enterobacteriaceae that may be shared by ictaluri. CHAPTER III. EARLY EVENTS IN THE PATHOGENESIS OF ENTERIC
SEPTICEMIA OF CATFISH: ELECTRON MICROSCOPIC FINDINGS
Introduction
Enteric septicemia of catfish (ESC), caused by the
Gram-negative bacterium Eh. ictaluri. is the most significant disease of the channel catfish farming industry
(MacMillan 1985) . The definitive pathogenesis of ESC is unknown. The work described in chapter II showes E. ictaluri rapidly crosses the intestinal epithelium, and morphologic changes at the light microscopic level associated with bacterial invasion were described. At 30 minutes PI, bacteria lined intestinal epithelial cells and large dilated cells subadjacent to the basement membrane were seen. Between 1 and 3 hours PI occasional necrotic enterocytes were on tips of intestinal folds. Proprial leukocyte infiltrates were common after 24 hours PI and bacteria were consistently seen within leukocytes.
Vascular associated leukocytes containing bacteria were the first lesions seen in other organs. Ultrastructural studies were undertaken to better define these light microscopic findings: 1) To clarify bacterial attachment and penetration; 2) To better characterize early morphologic changes in the intestine and other organs; and,
3) To clarify leukocyte-bacterial interactions.
97 98
Materials and methods
Bacteria
E. ictaluri. strain LA89-9, originally isolated from a
natural epizootic of ESC, was used in infection trials, and
as an antigen source for the production of specific anti-E.
ictaluri sera. Bacteria were cultured and stored as
described in the previous chapter. Bacterial identity was
verified periodically utilizing standard biochemical and
growth parameters (Hawke et al. 1981; Farmer and McWhorter
1985). Virulence was ensured by serial passage through
fingerlings as described in the previous chapter.
Rabbit anti-E. ictaluri sera
To identify bacteria in electron micrographs as E.
ictaluri. polyclonal rabbit antisera specific for E.
ictaluri was produced as described in the previous chapter.
Dilution titers were checked by enzyme linked immunosorbant
assays (ELISA) with sonicated formalized E^_ ictaluri as
antigen. A dilution titer of 1:10,000 was obtained
following the fourth booster. The rabbit was bled and the
serum harvested and stored at -70 °C. Antibody specificity was checked by ELISA using sonicated formalized Escherichia
coli. Aeromonas hvdrophila. and tarda as antigens.
Escherichia coli and Aeromonas hvdrophila coated plates had no cross reactivity. Edwardsiella tarda coated plates had cross reactivity for dilutions of 1:100 and less, but were negative for dilutions greater than 1:100. Commercially 99 available, 5 and 15 nm, gold-tagged, goat anti-rabbit IgG
(Bio Cell, Cardiff, Uk) was used as secondary antibody for identification of E-_ ictaluri on electron microscopy.
Maintenance and experimental infection of catfish
Enteric septicemia free fish, hatched and raised in
ESC free recirculating systems were used. Channel catfish fingerlings, 10-12 cm long and weighing 9-13 g, were maintained in 115 L glass aquaria as described in the previous chapter.
Virulent bacteria were cultured and fingerlings inoculated by gastric intubation as described previously.
Each fingerling, except control fish, received approximately 1.0 X 109 colony forming units E^. ictaluri.
Control fish received diluent only. To avoid regurgitation, fish were not anesthetized prior to intubation. Following intubation, fingerlings were monitored for regurgitation. The rare fish that regurgitated were re-intubated.
Tissue collection
Ten fish were sacrificed at time 0 and served as controls. Twenty fish were sacrificed per time period at
0.25, 0.5, 1, and 3 hours PI (PI). To strictly adhere to early time periods, a fish was infected, transferred to a new tank, and sacrificed before the next fish was infected.
Ten fish were sacrificed per time period at 6, 12, 24, 48,
72, 96, and 120 hours PI. Individual fingerlings were 100 euthanized by exposure in 0.5% 3-aminobenzoic acid
(Methane-sulfonate salt, Sigma Chemical Company, St. Louis,
MO) for 5 minutes (included in total exposure times) and necropsied.
Trunk kidney and liver samples were sterilely collected for bacteriologic culture, and samples of liver, trunk kidney, spleen, head kidney, and multiple intestinal samples from all areas were collected for electron microscopy (EM). Samples were cut into 1 mm square pieces with a razor blade and immersed in 1.25% glutaraldehyde and
2% paraformaldehyde, in 0.1 M sodium cacodylate buffer, pH
7.2. Tissues were fixed for 5 hours at room temperature on a laboratory rocker. Tissues were rinsed for 3 0 minutes in
0.1 M sodium cacodylate buffer, pH 7.2, two times. Tissues were stored at room temperature in cacodylate buffer for several days prior to processing.
Bacterial culture
To assess bacterial access to internal organs, trunk kidney and liver samples were plated directly onto brain heart infusion agar. Plates were cultured at 26 °C for 3 days and percent positive cultures noted.
Tissue processing and immunoqold staining
Tissues stored in cacodylate buffer were dehydrated in graded alcohols (30 to 100%) and infiltrated and embedded in LR-white resin (London Resin Company, Basingstoke,
Hampshire, Uk). Thick sections were cut and stained as 101 described in the previous chapter. Thin sections were cut
at 60 run on a Sorvall MT 5000 ultra-microtome. Sections were placed on 200 mesh nickel grids (Ted Pella, Inc.,
Redding, CA).
Immunogold staining of EL. ictaluri was done at 37 °C.
Nonspecific binding sites were blocked by immersing grids
in 1% normal goat serum dissolved in 0.1 M Tris buffer, pH
7.2 (TB), for 20 minutes. A 1:1000 dilution of rabbit anti-E. ictaluri in TB was subsequently applied for 1 hour.
Following washing in TB, 5 or 15 nm, gold-tagged, goat anti-rabbit IgG, at a dilution of 1:100 in TB was applied for 1 hour. Grids were washed in deionized water and stained for 10 minutes each in uranyl acetate and lead citrate. Sections were viewed with a Zeiss EM109 transmission electron microscope.
Results
Bacterial cultures
Percent positive cultures in liver and trunk kidney per time period are given in figure 14 (all tissues at time
0 had no growth). At 0.25 hours PI, 30% of trunk kidney cultures were positive for EL. ictaluri. At 0.5 hours PI,
26% of liver cultures and 52% of kidney cultures were positive for EL. ictaluri. By 24 hours PI, virtually all liver and trunk kidney cultures were positive for E. ictaluri. 102
POSITIVE FISH (%) 100
80 -
60 -
40 -
20 -
1 i I 0.25 0.5 1 3 6 12 24 48 72 TIME (HOURS)
LIVER KIDNEY
Figure 14: Percent positive liver and trunk kidney cultures for E_j_ ictaluri, 0.25 to 96 hours PI.
Electron microscopic findings
Intestine: All sections from fish sampled at 0 and
0.25 hours PI were essentially normal. At 0.5 hours PI, bacteria identified as E^. ictaluri by immunogold staining lined the brush border in sections from the anterior and middle portions of the intestine (figure 15).
Proteinaceous fimbrial attachments and brush border degeneration were not observed. Also at 0.5 hours PI, dilated cells containing large cytoplasmic inclusions were seen between enterocytes subadjacent to the basement membrane (Figure 16). These cells were most numerous on 103 tips of intestinal folds, decreased in number along sides of folds, and were rare in intestinal troughs. The dilated, basilar located cells were large, roughly oval, had round nuclei with small amounts of scattered heterochromatin, and multiple, cytoplasmic inclusions.
Inclusions consisted of either solid, fairly homogeneous, proteinaceous aggregates, or accumulations of degenerate membranous material (myelin figures). Nuclei were occasionally pyknotic and the cellular organelles degenerate. Dilated cells never extended to the intestinal lumen. Bacteria were not sequestered in these cells.
Numbers of these cells increased up to 24 hours PI and were found in all later time periods.
From 1 to 3 hours PI, occasional enterocytes on the tips of intestinal folds were necrotic (figure 17). Some, but not all, were associated with bacteria. Adjacent enterocytes were usually normal. At 6 hours PI, no necrotic enterocytes were found. Proprial leukocyte infiltrates and leukocytes extending between enterocytes were rarely observed in the first 24 hours PI, but commonly thereafter. Leukocytes were primarily agranular
(macrophages) although some granular leukocytes
(neutrophils) were seen. Bacteria, identified as E. ictaluri by immunogold labeling, were in vacuoles within leukocytes (figure 18). At 24 hours PI, most leukocytes were nondegenerate and contained low numbers of bacteria. 104
Figure 15: Electron micrograph of immunogold labeled E. ictaluri (i) adhered to the brush border (e) of an enterocyte, 0.5 hours PI.
Figure 16: Electron micrograph of a dilated basilar cell (d) between enterocytes (e) at 0.5 hours PI. Two types of cytoplasmic inclusions are present: one containing membranous figures (m) and another composed of proteinaceous material (p).
Figure 17: Electron micrograph of a necrotic enterocyte at the tip of an intestinal fold at 72 hours PI. Immunogold labeled E^. ictaluri (arrow) border the necrotic cell (n) . Adjacent enterocytes (e) are relatively unaffected.
Figure 18: Electron micrograph of a leukocyte (1) infiltrated.between enterocytes (e) at 48 hours PI. Immunogold labeled EL. ictaluri (arrow) are within cytoplasmic vacuoles in the leukocyte.
106
By 72 hours PI, numerous bacteria were within phagocytes
and phagocytes were degenerate.
Large numbers of leukocytes extended between and
replaced enterocytes by 4 days PI. On many sections, only
a thin rim of enterocytes bordering the lumen remained. No
areas of ulceration were seen.
Bacteria were not seen within or between enterocytes
in early time periods. However, by 72 hours PI, occasional
enterocytes contained bacteria within the cytoplasm (figure
19). Bacteria not associated with cells but free in the
connective tissues were rarely observed. These were always
adjacent to degenerate leukocytes.
Liver: Macrophages, containing bacteria in cytoplasmic vacuoles, were within and adjacent to small vessels and sinusoids at 48 hours PI (figure 20); this was the first observable lesion in liver sections. Numbers of bacteria-laden macrophages among hepatocytes increased in later time periods.. Phagocytic cells containing large numbers of bacteria were degenerate. At 72 hours PI, bacteria were seen in the cytoplasm of hepatocytes (figure
21) .
Spleen, trunk kidney, and head kidnev: Changes in these tissues are considered together since lesions were similar. Sinus associated macrophages containing bacteria were seen at 48 hours PI (figure 22). At later time periods, macrophages infiltrated between and replaced 107
Figure 19: Electron micrograph of an enterocyte containing immunogold labeled E^. ictaluri (i) in the cytoplasm, 72 hours P I .
Figure 20: Electron micrograph of a hepatic vessel bordered by hepatocytes (h) at 72 hours PI. Several erythrocytes (r) and a leukocyte (1) containing immunogold labeled E. ictaluri (arrow) are in the vessel.
Figure 21: Electron micrograph of several hepatocytes (h) at 72 hours PI. Immunogold labeled E^ ictaluri (arrow) is in the cytoplasm of a hepatocyte.
Figure 22: Electron micrograph of head kidney at 72 hours PI. Adjacent to an erythrocyte (r) are leukocytes (1) containing E^. ictaluri within cytoplasmic vacuoles. The insert (bottom left) shows immunolabeling.
109
normal cells. Many contained bacteria within cytoplasmic vacuoles. Macrophages were often degenerate in later time periods and occasional bacteria were free in the supporting
stroma adjacent to necrotic leukocytes.
Discussion
Results of this research support earlier observations that Ej. ictaluri may have a similar pathogenesis as other enteroinvasive enterobacteriaceae, particularly Salmonella
and Yersinia species. Isolates of Salmonella and Yersinia which are systemic pathogens, penetrate the intestinal epithelium and target resident macrophages where they replicate (Finlay and Falkow 1989a). Dissemination to regional lymph nodes and other tissues occurs within phagocytes (Edelman and Levine 1986; Finlay and Falkow
1989a).
Like Yersinia, which can attach to and penetrate cellular membranes in minutes (Brunius and Bolin 1983), E.
ictaluri passed through the intestinal epithelium by 0.25 hours PI. Preliminary intestinal ligation studies currently in progress suggest bacterial passage may occur as early as 5 minutes post luminal inoculation.
Unfortunately, bacteria crossing the intestinal mucosa at early time periods were not observed. In investigating other enteroinvasive enterobacteriaceae, intestinal ligation and cell culture models have been used to clarify 110 mechanisms of epithelial passage (Takeuchi, 1967; Finlay et al. 1988a). These systems allow bacteria to be applied on a limited surface area and maximize the possibility of observing transepithelial passage. Bacteria were seen within enterocytes and hepatocytes in later stages of infection, indicating that eukaryotic cell invasion does occur.
Salmonella and Yersinia species causing systemic disease utilize existing cellular transport pathways to invade eukaryotic cells (Finlay and Falkow 1989a) in a process known as parasite directed or mediated endocytosis
(Moulder 1985) . The rapid mucosal passage of E_j_ ictaluri suggests a similar mechanism. Widespread, bacteria- induced, enterocytic necrosis with subsequent access to internal tissues would likely require more time.
Like Yersinia and Salmonella, no proteinaceous fimbrial structures were seen in E^. ictaluri closely apposed to enterocytes. Fimbria have a minimal role in invasion by Salmonella and Yersinia species (Jones and
Richardson 1981; Isberg et al. 1987). ictaluri has not definitively been shown to have fimbria. If fimbria are present, their role in pathogenesis is unknown.
The presence of dilated basilar cells and occasional necrotic enterocytes on the tips of intestinal folds may be indicators of mucosal passage by JL ictaluri. The identity of the dilated cells is unknown; their nuclear and Ill cytoplasmic ultrastructure are similar to macrophages.
This suggests they may be resident macrophages and the prominent cytoplasmic inclusions are accumulated debris from adjacent enterocytes. Bacteria, however, were not sequestered within these cells.
The significance of observed necrotic enterocytes is also unknown. They were sporadically seen in sections from
1 and 3 hours PI and at no other time frames. Some necrotic cells had associated bacteria, but in others no bacteria were observed. Perhaps E^. ictaluri injures cells through which it passes. Salmonella have been shown to replicate within eukaryotic cells although the bacteria can pass through epithelial monolayers without causing cellular death (Finlay et al. 1988b; Finlay and Falkow 1988c).
In all tissues examined, E^_ ictaluri was almost exclusively found within cytoplasmic vacuoles in phagocytic cells. The absence of cytoplasmic granules suggests these cells are macrophages. In earlier time periods, phagocytes contained single bacteria, while in later time periods phagocytes contained multiple bacteria. Moreover, occasional bacteria were apposed at the ends and incompletely separated, suggesting bacterial replication.
Further research is needed since the bacteria may have been replicating when phagocytized.
Perhaps Ej. ictaluri is similar to Salmonella and
Yersinia which target host phagocytes. Vascular associated 112 macrophages consistently contained intracytoplasmic bacteria. Only in later time periods were bacteria seen in other cell types: enterocytes and hepatocytes. On light microscopy, enterocytes and hepatocytes containing bacteria could not definitively be identified. The increased resolution provided by electron microscopy was necessary to positively identify these infected cells. Interestingly, both Salmonella and Yersinia, though preferentially targeting phagocytes, can slowly replicate within other eukaryotic cells {Finlay and Falkow 1988c).
In conclusion, E^_ ictaluri has been shown to access fish via the intestinal tract in this study. Passage occurs as early as 0.25 hours PI suggesting pre-existing cellular transport pathways are utilized and de novo bacterial protein synthesis is not required. No fimbria or other proteinaceous attachments were observed in any sections. Leukocytes appear important in the pathogenesis of Ej. ictaluri infections; perhaps EL. ictaluri is similar in this regard to other enteroinvasive enterobacteriaceae. CHAPTER IV. ANTIGENS OF EDWARDSIELLA ICTALURI RECOGNIZED
IN NATURALLY INFECTED CHANNEL CATFISH.
Introduction
Edwardsie11a ictaluri. the cause of enteric septicemia of catfish (ESC) and a member of the enterobacteriaeeae, is a short, Gram-negative, motile rod (Farmer and McWhorter
1985) . In immunologic studies of other pathogenic Gram- negative bacteria, outer membrane structures such as lipopolysaccharide (LPS), outer membrane proteins (OMP), flagella, fimbria, and extra-membranous polysaccharide capsules have been shown to be antigenic in infected hosts
(Broes et al. 1989; Puohiniemi et al. 1990; Aleksic et al.
1991; Brussow and Sidoti 1992) . Edwardsiella ictaluri has not definitively been shown to possess fimbria or an extramembranous polysaccharide capsule, but flagella, OMP, and LPS have been partially characterized, structurally and immunologically.
Newton et al. (1990) evaluated OMP from 33 isolates of
E. ictaluri by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE). One major protein with apparent molecular mass of 35 kD and nine minor proteins with apparent molecular masses of 71, 51, 46, 43.5, 38.5,
37.5, 31.5, 28.5, and 19.5 kD were found. Flagella profiles were similarly examined from 10 isolates of Ei.
113 114
ictaluri and two flagellin bands with apparent molecular
masses of 38 and 44 kD were seen (Newton and Triche in
press, a). Newton and Triche (in press, b) also extracted
and characterized LPS from 40 isolates of EL. ictaluri by
SDS-PAGE. All isolates had complete ladder-like arrays of
low, medium, and high molecular weight LPS.
It is unknown if flagella, LPS, or OMP are immunogenic
in naturally infected channel catfish. Saeed and Plumb
(1986) demonstrated that purified LPS injected into channel
catfish produced high antibody titers, and immunized fish
were resistance to infection. Plumb and Klesius (1988)
immunoblotted whole cell lysates (WC) using pooled serum
from infected catfish, and identified two immunogenic
proteins with apparent molecular masses of 3 4 and 60 kD.
The location of the proteins in intact bacteria was not
determined. The purpose of this study was to clarify the
immunogenicity of EL. ictaluri in naturally infected channel
catfish. Specifically, WC preparations and enriched
fractions of OMP, flagella, and LPS were examined for
immunogenicity by immunoblot techniques.
Materials and methods
Bacteria
Edwardsiella ictaluri. strain AL83189, originally
isolated from diseased channel catfish in Auburn, Alabama, was used in the preparation of enriched bacterial 115 components and as whole cell antigen. Bacteria were cultured on brain heart infusion agar (BHIA, Difco
Laboratories, Detroit, MI) at 2 6 "C or in brain heart infusion broth (BHIB, Difco Laboratories) at 26 °C with moderate shaking (100 revolutions/minute). culture identity was routinely verified using standard biochemical tests (Farmer and McWhorter 1985). Bacteria were stored at
-70 °C in BHIB containing 20% glycerol.
Enriched bacterial fractions
Outer membrane proteins: Outer membrane proteins were prepared using a modification of the sarcosinate extraction procedure of Barenkamp, et al. (1981). Bacteria were cultured in BHIB for 18 hours and harvested by centrifugation at 4,000 X g for 10 minutes at 4 °C.
Pelleted bacteria were washed by suspension in 10 mL of 10 mM HEPES buffer, pH 7.4, and recentrifuged. Bacteria were resuspended in 15 mL HEPES buffer, placed in an ice water bath, and sonicated five times for 1 minute with 3 0 seconds between sonications (Model W-375 Heat systems -
Ultrasonics, Inc., Plainview, NY). Cells not lysed were removed by centrifugation at 12,000 X g for 15 minutes at 4
°C. The supernatant was collected and a total membrane preparation obtained by centrifugation at 100,000 X g for 1 hour at 4 °C. The resulting pellet was resuspended in 15 mL of a 1% sodium lauryl sarcosinate solution and incubated for 3 0 minutes at room temperature on a laboratory rocker. 116
The detergent insoluble fraction was collected by
centrifugation at 100,000 X g for 60 minutes at 4 °C and
resuspended in 250 /xL distilled water. Outer membrane
enriched fractions were stored at -20 °C.
Flagella: Flagella (provided by Joseph C. Newton,
Louisiana State University) were obtained using a modification of the procedure of Ibrahim, et al. (1985).
To enhance production of flagella, pure cultures of E.
ictaluri were serially passed through semi-solid agar composed of BHIB with 0.25% bacteriologic agar (Difco
Laboratories). A single colony was stab inoculated into the center of the agar plate and incubated for 24 hours at
26 °C. Peripheral growth was reinoculated into the centers of subsequent plates; the procedure was repeated for 8 to
10 passages. The highly motile isolates obtained were cultured in BHIB as described above.
Bacteria were harvested by centrifugation at 4 000 X g for 15 minutes at 10 °C. The resulting pellet was resuspended in 0.85% NaCl to form a thick solution. The pH was decreased to 2.0 using IN HCL and the solution incubated at room temperature for 3 0 minutes with constant stirring. Bacteria, now devoid of flagella, were pelleted by centrifugation at 4 000 X g for 15 minutes at 10 °C. The supernatant, containing flagellin monomers, was further centrifuged at 100,000 X g for 1 hour at 10 °C to sediment all nonsolubilized material. The supernatant was collected and adjusted to pH 7.2 using 1 N NaOH. At this pH, flagellin monomers reassociate forming filaments of varying lengths. Filaments were precipitated by slow addition of ammonium sulfate to a final concentration of 2.67 M and the mixture slowly stirred for 16 hours at 4 °C. The precipitate was collected by centrifugation at 15,000 X g for 15 minutes at 4 °C. The supernatant was discarded and the precipitate solubilized in 2 mL distilled water.
Ammonium sulfate was removed by dialysis at 4 "C for 24 hours (tubing with 50,000 molecular-weight cut off) against distilled water containing 0.5% activated charcoal.
Filaments were collected by centrifugation for 15 minutes at 15,000 X g at 4 °C. The pellet was reconstituted in l mL distilled water and stored at -20 °C.
Lipooolvsaccharide: Purified LPS (provided by Joseph
C. Newton) was obtained using Johnson and Perry's (1976) modification of the hot aqueous phenol extraction method of
Westphal and Jann (1965). Bacteria harvested from 10 L of
BHIB by centrifugation at 4000 X g for 15 minutes at 4 °C were resuspended in 50 mL of 0.01 M PBS, pH 7.0, containing
5 mM ethylenediamine-tetracetic acid (tetrasodium salt,
Sigma Chemical Company, St. Louis, MO). One hundred mg of egg white lysozyme (American Research Company, Solon, OH) was added and the suspension stirred at 4 °C for 16 hours.
After this incubation, the suspension was placed in a water bath at 37 °C for 3 0 minutes. The volume was adjusted to 118
100 mL with 20 mM MgCl2 and bovine pancreas ribonuclease and deoxyribonuclease (Sigma Chemical Company) were added at final concentrations of l nq/mL. The suspension was incubated 10 minutes at 37 °C and for an additional 10 minutes at 60 °C in a water bath.
An equal volume (100 mL) of preheated (70 °C) 90%
(V/V) aqueous phenol was added to the suspension and the resulting mixture was stirred for 15 minutes. The suspension was then rapidly cooled to 15 °C by stirring in an ice water bath. Centrifugation at 18,000 X g for 15 minutes followed, separating the solution into aqueous and phenol layers. The aqueous phase was retained and dialyzed against running tap water for 2 hours at room temperature and against distilled water until no phenol odor could be detected. The LPS was pelleted by centrifugation at
100,000 X g for 2 hours at 4 °C. The pellet was reconstituted in distilled water, lyophilized, and stored at -20 °C.
Whole cell lvsates
Single colonies of EL. ictaluri were inoculated into 10 mL BHIB and cultured as described above for 18 hours.
Bacteria were pelleted by centrifugation at 4000 X g for 15 minutes at 4 °C and washed once by suspension in 10 mL of
10 mM HEPES buffer, pH 7.4. One mL aliquots were microcentrifuged, the supernatants discarded, and the bacterial pellet resuspended in l mL sample buffer (10% 119 glycerol, 2% SDS, 5% 2-mercaptoethanol, 0.125 M Tris-HCl
[pH 6.8], and 0.01% bromphenol blue). Whole cell preparations were stored at -20 °C.
Quantification of LPS and proteins
Total protein in flagella and OMP preparations was determined using the procedure of Markwell et al. (1978).
The amount of LPS in samples was determined by the method of Karkanis et al. (1978) using 2-keto-3-deoxyoctonate
(KDO) as standard.
Gel electrophoresis
Gel electrophoresis was performed using a 4.5% stacking gel, pH 6.8, and a 12,5% separating gel, pH 8.8, with a discontinuous buffer system as described by Laemmli
(1970). Gels were cast in 8 X 10 cm by 0.75 mm vertical slab cells (Mighty Small™ SE 250, Hoefer Scientific
Instruments, San Francisco, CA). Flagella, OMP, and LPS preparations were solubilized in sample buffer and all samples were heated to 85 °C for 10 minutes prior to loading. Five jxl WC (amount to load empirically determined), 40 protein/lane flagella, 160 ng protein/lane OMP, and 35 ^g KDO/lane LPS were loaded in duplicate on each gel. Antigens were separated by electrophoresis at a constant current of 10 mA per gel.
Electrophoresis was stopped when the tracking dye was approximately 0.2 cm from the gel bottom. Proteins were stained with Coomassie brilliant blue. Lipopolysaccharide 120
fractions were stained with silver according to the
procedure of Hitchcock and Brown (1983).
Immunoblots
Bacterial fractions separated by SDS-PAGE were
transferred onto 0.2 jLtm nitrocellulose paper (Schleicher
and Schuell, Keane, WV) from unstained gels. Gels were
equilibrated in transfer buffer (25 mM Tris buffer, pH 8.3,
193 mM glycine, 1.3 mM SDS, and 20% methanol) for 15
minutes and placed in a semi-dry transfer unit (Hoefer TE
77 SemiPhor™) . Proteins and LPS were transferred from
PAGE gels to nitrocellulose at a constant current of 24 mA
(0.5 mA/cm2 gel) for 35 minutes.
Following transfer, nitrocellulose membranes were cut
in half. One half, containing molecular weight standards,
WC, OMP, and flagellar proteins was stained with India ink
according to standard protocols (Harlow and Lane 1988),
India ink staining was done to verify transfer of
individual bands. The remaining half, containing WC, OMP,
and flagellar proteins and LPS, was immunostained.
Nonspecific binding sites were blocked by incubating
nitrocellulose membranes containing separated bacterial
components in phosphate buffered saline, pH 7.2, (PBS) containing 5% nonfat dry milk (Carnation Company, Los
Angeles, CA), 0.2% Tween 20 (American Research Company,
Solon, OH), and 0.02% sodium azide (Sigma Chemical Company)
for 2 hours at room temperature with gentle agitation. 121
Membranes were subsequently washed with PBS containing
0.05% Tween 20 (washing buffer) two times for 10 minutes each.
Catfish sera obtained from channel catfish infected naturally with ictaluri was used to probe the separated bacterial components affixed to the nitrocellulose membranes. Untreated serum and serum absorbed with purified LPS to remove the anti-LPS antibodies were used.
Antibodies against LPS were removed by mixing the serum with an equal volume of washing buffer containing 1 mg/mL solubilized LPS and incubated for 2 hours at room temperature on a laboratory rocker. Antibody-LPS complexes were pelleted by microcentrifugation for 5 minutes at room temperature. The supernatant was harvested and the procedure repeated. Absorbed serum was diluted to a final concentration of 1:10 in washing buffer. Nonabsorbed serum was diluted to 1:50. Absorbed and nonabsorbed serum were applied to separate nitrocellulose membranes for 1 hour at room temperature with gentle agitation.
Following incubation in diluted catfish serum, membranes were washed four times with washing buffer for 5 minutes at room temperature with gentle agitation. Rabbit anti-catfish IgM sera (production described below), diluted to 1:1000 in washing buffer, was subsequently applied for 1 hour at room temperature with gentle agitation.
Horseradish peroxidase conjugated, goat anti-rabbit IgG 122
(Sigma Chemical Company) was applied for 1 hour at room temperature. Following washing, color development was done using 3-3'-diaminobenzidine tetrahydrochloride (DAB, Sigma
Chemical Company) solubilized in 0.1 M Tris buffer, pH 7.0, containing 0.02% hydrogen peroxide and 0.04% nickel chloride. Color development was monitored visually and stopped by washing in deionized water.
Production of rabbit anti-catfish IoM
To allow for detection of bound catfish IgM to separated bacterial components, rabbit antisera to catfish
IgM was produced. Catfish IgM was obtained from channel catfish, 15 to 20 cm in length, which were maintained in
115 L glass aquaria under slow, flow-through conditions
(two water changes/day), using dechlorinated Baton Rouge city water with supplemental filtration provided by in-tank airlift sponge filters. Temperature was maintained at 26 ±
2 °C using in-tank glass aquarium heaters. Fish were fed a commercially prepared ration (Zeigler Brothers,
Incorporated, Gardners, PA) three times a week. Water ammonia, nitrite, and pH were monitored biweekly using a commercially available kit (Hach Company, Loveland, CO).
Nitrite and ammonia were consistently below 1 part per million; pH was 8.0.
Bovine serum albumin (Sigma Chemical Company) was solubilized in PBS at 1 mg/ml and emulsified with a similar volume of complete Freund's adjuvant (CFA, Difco Laboratories). One mL of the emulsion was administered
intraperitoneally to each of eight catfish. Three boosters
of similar volume but in incomplete Freund's adjuvant
(ICFA, Difco Laboratories) followed at 2 week intervals.
Two weeks following the fourth injection, catfish were bled through the tail vein and the serum harvested. Catfish anti-BSA IgM was concentrated using a commercially available affinity chromatography kit according to the manufacturers directions (ImmunoPure Ag/Ab Immobilization
Kit 1, Pierce, Rockford, IL). The total protein of the elutions were between 450 and 480 /xg/ml whole catfish serum as determined by the procedure of Markwell et al. (1978).
Purity of the IgM sample was examined by SDS-PAGE followed by Coomassie brilliant blue staining. Two protein bands with apparent molecular masses of 70 (heavy chain) and 24
(light chain) kD were seen.
One mL of PBS containing 450 to 480 fig protein
(catfish IgM) was emulsified with 1 mL CFA and administered subcutaneously, at multiple sites, to a New Zealand white rabbit. Three boosters of similar volume but in ICFA followed at 3 week intervals. Three weeks following the last booster, 2 0 mL of blood was collected from the auricular artery and the serum harvested. A dilution titer of 1:5000 was obtained by enzyme link immunosorbant assay
(ELISA) . Polystyrene plates coated with 2 /xg/well catfish
IgM were used. Serum was stored at -20 °C. 124
To ensure that the rabbit sera did not cross react with components of EL. ictaluri. WC immunoblots were
performed omitting the catfish serum. No staining was
observed.
Results
Proteins from EL. ictaluri OMP, flagella, and WC
preparations and purified LPS were examined by SDS-PAGE
(figure 23). In whole cell preparations, 38 protein bands,
ranging in apparent molecular mass from 108 to 13
kilodaltons (kD), were observed in Coomassie brilliant blue
stained gels, of these 38 bands, six, having apparent molecular masses of 77, 72, 66, 55, 44, and 37 kD were major bands (as evidenced by staining intensity), while the
remaining 32 were minor bands.
Sixteen proteins with molecular masses of 82, 77.5,
73, 70, 65, 59, 57, 54.5, 51, 39, 37, 33, 28, 23.5, 21, and
19 kD were observed in OMP preparations. One major band with an apparent molecular mass of 37 kD contained over 50%
of the protein in the OMP preparation (subjective
evaluation based on total staining intensity).
In flagella preparations, two major proteins with apparent molecular masses of 40 and 3 6 kD were seen on
Coomassie brilliant blue stained gels. Additionally, on overloaded flagellar gels, lightly staining protein bands with molecular masses between 90 and 13 kD were seen. The 125 exact origin of these minor protein bands is unknown but they probably represent contaminant proteins co-purified with flagellar proteins.
Silver stained LPS preparations show medium and low molecular mass bands. Lower molecular mass bands, as subjectively measured by staining intensity, contain over
80% of the total LPS.
India ink stains of the OMP, flagella, and WC preparations are shown in figure 24. Twenty four of the 38 protein bands seen in WC preparations stained by India ink.
Proteins with molecular weights greater than 70 and less than 20 were not seen. In OMP preparations, the 57, 54.5,
51, 39, 37, 33, and 28 kd bands stained with India ink. In addition, a 26 kD band not seen on Coomassie brilliant blue stained gels was observed. As in whole cell preparations, the high and low molecular mass OMP were not seen. In flagella enriched fractions, proteins with apparent molecular masses of 45, 43, 36, 33, and 27 kD stained faintly with India ink. The 40 kD flagellar protein was not seen.
Immunostains of flagella, WC, OMP, and LPS preparations using non-absorbed catfish serum are shown in figure 25. The 37 kD major OMP, in both WC and OMP preparations, was strongly immunogenic as evidenced by intense staining. Other protein bands in WC and OMP lanes could not be distinguished from LPS which also stained 126
Figure 23: Sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) profile of whole cell and enriched fractions of E_;_ ictaluri AL83189. Lanes a through d are stained with Commassie Brillant Blue; lanes e and f are stained with sliver. Lane a, molecular weight standards; lane b, outer membrane enriched fraction; lane c, whole cell lysate; lane d, flagella preparation; lane e, lipopolysaccharide (LPS) from EL. ictaluri AL83189; lane f, LPS from Escherichia coli 0111:B4 (Sigma Chemical Company).
Figure 24: Transferred proteins from whole cell and enriched fractions of EL. ictaluri AL83189 stained with India ink. Lane a, molecular weight standards; lane b, flagella enriched fraction; lane c, whole cell lysate; lane d, outer membrane protein preparation. 127
CO CM
r i
I I I o i ,1 CO o o a t <0 CO CM 128 in these preparations. The purified LPS also stained; bands were seen in high/ medium, and low molecular mass ranges. In flagella preparations, the 36 kD band stained lightly, but no other bands were visualized.
Immunostains using catfish serum absorbed with purified LPS are shown in figure 26. In whole cell preparations, proteins with apparent molecular masses of
51.5, 49.5, 46, 43.5, 40, 37, 35, 33, 30, 28.5, 24, 22, 19,
16.5, and 15 kD were seen. Only the 36.5 and 35 kD bands stained strongly. In the outer membrane protein lane, the
51, 39, 37, 33, 23.5, and 21 kD proteins were immunogenic.
The major outer membrane protein (37 kD) is the only band which stained strongly. The 36 kD flagellar band stained weakly. No staining was seen in the LPS lane.
Discussion
Whole cell lysates of Ej. ictaluri and enriched fractions of outer membrane proteins, flagella, and LPS were separated by SDS-PAGE, blotted onto nitrocellulose membranes, and probed for immunogenicity using sera obtained from channel catfish naturally infected with E. ictaluri. To distinguish immunogenic proteins from LPS, serum used to stain protein preparations was absorbed with purified LPS prior to immunostaining.
Results of this research indicate that LPS and the major outer membrane protein (37 kD) of EL. ictaluri were 129
Figure 25: Immunoblot analysis of EU, ictaluri. AL83189. All lanes were incubated with 1:50 dilution of catfish serum from naturally infected fish. Lane a, flagella enriched fraction; lane b, whole cell lysate; lane c, outer membrane protein preparation; lane d, purified LPS.
Figure 26: Immunoblot analysis of E.._ ictaluri. AL83189. All lanes were incubated with 1:10 dilution of catfish serum absorbed with purified LPS. Lane a, flagella enriched fraction; lane b, whole cell lysate; lane c, outer membrane protein preparation; lane d, purified LPS. kD kD
94-
67-
43 -
30 -
20-
•it
abed d 25
26 130 131 strongly immunogenic in this group of naturally infected catfish. The 36 kD flagella band, and outer membrane proteins with apparent molecular masses of 51, 39, 33,
23.5, and 21 kD were weakly immunogenic. Proteins with molecular weights above 70 kD and below 20 kD are of unknown immunogenicity since bands were not visualized on
India ink stains (antigen may have been limiting).
Why were the high and low molecular mass proteins not visualized on India ink stains? Proteins with molecular masses greater than 70 kD were retained on the polyacrylamide gels as demonstrated by Coomassie brilliant blue stained post-blotted gels. The small molecular weight proteins passed through the nitrocellulose membranes at the current and transfer times used (confirmed by stacking membranes). The current and transfer times used were chosen to provide a spectrum of molecular weight proteins on a single membrane for immunoscreening. Further studies using separate membranes and different transfer times are in progress.
Outer membrane preparations stained with Coomassie brilliant blue showed six additional minor protein bands than previous reports (Newton et al. 1990). This may be due to sample overloading in attempts to ensure sufficient antigen transfer to nitrocellulose membranes. In addition, observed bands differed in apparent molecular masses. 132
This may be due to greater inherent measuring errors in mini verses full size gels.
The enriched bacterial fractions used in conjunction with WC preparations assisted in the identification of some of the immunogenic WC antigens. For example, of the fifteen protein bands seen in WC preparations exposed to
LPS-absorbed catfish serum, six (molecular masses of 51.5,
40, 37, 33, 24, and 22 kD) are outer membrane proteins, and one may be the 38 kD flagellar protein. The identity of the remaining nine proteins is unknown. The immunogenic 60 kD protein seen in whole cell preparations by Plumb and
Klesius (1988) was not seen in whole cell lysate immunoblots exposed to LPS extracted serum. Protein bands with apparant molecular mass near the reported 34 kD band
(Plumb and Klesius 1988) were seen, but due to strain and preparation differences, it is unknown if any of the proteins are similar.
The lipopolysaccharide in whole cell, outer membrane, and extracted preparations was strongly immunogenic. High molecular weight LPS bands were not seen with silver stains, but were visualized by immunostains using non-LPS absorbed catfish serum. The complete ladder-like array of
B. ictaluri LPS was clearly visible on immunostains.
Protection afforded the host by antibodies against these immunogenic proteins or LPS is unknown. Saeed and
Plumb (1986) showed that phenol extracted LPS from EL. 133 ictaluri is immunogenic and somewhat protective when injected into channel catfish. Moreover, LPS are important antigens in other fish diseases, and are the protective antigens in the Vibrio and Yersinia ruckeri vaccines used in salmonids (Amend et al. 1983; Ellis 1988). It is unknown if antibodies against the LPS of ictaluri are protective in naturally exposed fish.
Little work has been done on the role of outer membrane proteins in fish diseases. In other enterobacteriaceae, antibodies against major outer membrane proteins have been demonstrated and may provide protective immunity (Matsui and Arai 1989; Pal et al. 1989). It is unknown if antibody directed against the outer membrane proteins of EL. ictaluri are protective. The strong reaction of the major outer membrane protein invites further investigation.
The 36 kD flagella band was weakly immunogenic. Since the 44 kD band was not seen on India ink stains, its immunogenicity is uncertain. Further investigation varying blotting conditions is required.
In conclusion, LPS and the 37 kD major outer membrane protein of E^. ictaluri have been shown to be strongly immunogenic in these naturally infected channel catfish.
Five additional outer membrane proteins and one flagellar antigen are weakly immunogenic. SUMMARY
To clarify early events in the pathogenesis of ESC,
140 channel catfish fingerlings were each given approximately 1.0 X 109 CFU E^. ictaluri by intragastric intubation. Fingerlings were sampled at 0.25, 0.5, 1, 3,
6, 12, 24, 48, 72, 96, and 120 hours post infection (PI).
Edwardsiella ictaluri from a natural outbreak of ESC was used as inoculum, and for the production of rabbit anti-E. ictaluri antisera. Bacterial virulence was ensured by serial passage through catfish fingerlings.
To identify E^_ ictaluri by specific immunostaining in tissue sections, polyclonal antisera (rabbit) was produced.
Formalized, ground, whole cell ictaluri was administered to a New Zealand white rabbit. Titers and specificity were monitored by enzyme linked immunosorbant assays. Antisera was used at 1:1000 dilution.
At each time frame, fingerlings were sacrificed and samples of liver, kidney, stomach, and intestines were quantitatively cultured. Moreover, tissues were collected, fixed, processed, and sectioned for light and electron microscopy. Immunohistochemical detection of E_j. ictaluri was accomplished using immunoperoxidase and immunogold staining.
Bacterial cultures indicated E^. ictaluri rapidly crossed the gastrointestinal tract. Thirty percent of kidney cultures were positive at 15 minutes PI.
134 135
Earliest gross lesions, consisting of skin and hepatic petechiae were seen at 2 days PI. Subsequently observed lesions were ascites, widespread petechiation in multiple tissues, visceral edema, and necrosis in multiple organs.
Histologic lesions were first seen at 0.5 hours PI and consisted of IL. ictaluri aligned along the intestinal brush border, and the appearance of dilated cells subadjacent to the basement membrane. No proteinaceous attachments were seen on electron microscopy. From 1 to 3 hours PI, occasional enterocytes on the tips of intestinal folds were necrotic. Adjacent cells were essentially normal.
Leukocytes, primarily macrophages, were observed in the lamina propria at 24 hours PI. Leukocytes subsequently infiltrated the mucosa which became thin but did not ulcerate. Bacteria were commonly seen within vacuoles in the cytoplasm of leukocytes.
Leukocytes containing bacteria were seen in the liver, kidney, head kidney, and spleen by 48 hours PI. Leukocytes accumulated within the interstitial tissues adjacent to and within vascular spaces. Degenerate leukocytes were commonly seen by 72 hours PI. Bacteria were seen in enterocytes and hepatocytes in addition to leukocytes at 72 hours PI.
Results of this research suggest Eh. ictaluri may have an invasion strategy similar to other enteroinvasive enterobacteriaceae, principally some yersinia species. 136
Like Yersinia, E*. ictaluri appears to cross the intestinal mucosa in a matter of minutes- Replication and dissemination to various tissues may occur within leukocytes.
In a separate study, immunogenic antigens of E. ictaluri were determined by probing bacterial components with serum from naturally infected channel catfish. Whole cell preparations and enriched fractions of outer membrane proteins, flagella, and LPS were separated by SDS-PAGE and transferred to nitrocellulose membranes. Bacterial components were probed for antigenicity using catfish serum and catfish serum absorbed with purified LPS. Outer membrane LPS and the major outer membrane protein were strongly immunogenic. Flagellar antigens and five additional outer membrane proteins were weakly immunogenic. BIBLIOGRAPHY
Ainsworth, A. J., Capley, G., Waterstreet, P., and Munson, D. i986. Use of monoclonal antibodies in the indirect fluorescent antibody technique (IFA) for the diagnosis of Edwardsiella ictaluri. Journal of Fish Diseases 9:439-434.
Aizawa, S., Dean, G. E., Jones, C. J . , Macnab, R. M . , and Yamaguchi, S. 1985. Purification and characterization of the flagellar hook-basal body complex of Salmonella tvnhimurium. Journal of Bacteriology 161:83 6-849.
Aleksic, S., Burwitz, K., Bockemuhl, J., and Aleksic, V. 1991. Immune response to flagellar antigens in human infections with Yersinia enterocolitica. Contributions to Microbiology and Immunology 12:110- 116.
Allweiss, B., Dostal, J., Carey, K. E., Edwards, T. F., and Freter, R. 1977. The role of chemotaxis in the ecology of bacterial pathogens of mucosal surfaces. Nature 266:448-450.
Amend, D. F., Johnson, K. A., Croy, T. R., and McCarthy, D. H. 1983. Some factors affecting the potency of Yersinia ruckeri bacterins. Journal of Fish Diseases 6:337-344.
Areechon, N. and Plumb, J. A. 1983. Pathogenesis of Edwardsiella ictaluri in channel catfish, Ictalurus punctatus. Journal of the World Mariculture Society 14:249-260.
Balligand, G., Laroche, Y., and Cornelis, G. 1985. Genetic analysis of virulence plasmid from a serogroup 9 Yersinia enterocolitica strain: role of outer membrane protein PI in resistance to human serum and autoagglutination. Infection and Immunity 48:782-786.
Barenkamp, S. J., Munson, R. S. Jr., and Granoff, D. M. 1981. Subtyping isolates of Haemophilus influenzae type b by outer-membrane protein profiles. The Journal of Infectious Diseases 143:668-676.
Baudry, B., Maurelli, A. T., Clerc, P., Sadoff, J. c., and Sansonetti, P. J. 1987. Localization of plasmid loci necessary for Shigella flexneri entry into HeLa cells, and genetic organization of one locus encoding four immunogenic polypeptides. Journal of General Microbiology 133:3403-3413.
137 138
Baxa, D. V., and Hedrick, R. P. 1989. Two more species are susceptible to experimental infections with Edwardsiella ictaluri. FHS/AFS Newsletter 17:4.
Benjamin, W. H., Turnbough, C. L., Posey, B. S., and Briles, D. E. 1985. The ability of Salmonella tvphimurium to produce the siderophore enterobactin is not a virulence factor in mouse typhoid. Infection and Immunity 50:392-397.
Bertolini, J. M., cipriano, R. c., Pyle, s. w., and McLaughlin, J. J. A. 1990. Serological investigation of the fish pathogen Edwardsiella ictaluri. cause of enteric septicemia of catfish. Journal of Wildlife Diseases 26:246-252.
Blanch, A. R . , Pinto, R. M . , Jofre, J. T. 1990. Isolation and characterization of an Edwardsiella sp. strain, causative agent of mortalities in sea bass (Dicentrarchus labrax). Aquaculture 88:213-222.
Blazer, V. S., Shotts, E. B., Waltman, W. D. 1985. Pathology associated with Edwardsiella ictaluri in catfish, Ictalurus punctatus Rafinesque, and Danio devario (Hamilton-Buchanan, 1822). Fish Biology 27:167-175.
Bolin, I., Portnoy, D. A., and Wolf-Watz, H. 1985. Expression of the temperature-inducible outer membrane proteins of Yersinia. Infection and Immunity 48:234- 40.
Bovallius, A., and Nilsson, G. 1975. Ingestion and survival of Y. pseudotuberculosis in HeLa cells. Canadian Journal of Microbiology 21:1997-2007.
Boyce, J. M . , Evans, D. J. Jr., Evans, D. G., and DuPont, H. L. 1979. Production for heat-stable, methanol- soluble, enterotoxin by Yersinia enterocolitica. Infection and Immunity 25:532-537.
Broes, A., Fairbrother, J. M., Jacques, M., and Lariviere, S. 1989. Requirement for capsular antigen KX105 and fimbrial antigen CS1541 in the pathogenicity of porcine enterotoxigenic Escherichia coli 08:KX105 strains. Canadian Journal of Veterinary Research 53:43-47.
Brubaker, R. R. 1972. The genus Yersinia: biochemistry and genetics of virulence. Current Topics in Microbiology and Immunology 57:111-158. 139
Brunius, G. and Bolin, I. 1983. Interaction between Yersinia pseudotuberculosis and the HeLa cell surface. Journal of Medical Microbiology 16:245-261.
Brussow, H., and Sidoti, J. 1992. Reactivity of human serum antibody with lipopolysaccharide 0-78 antigen from enterotoxigenic Escherichia coli. Epidemiology and Infection 108:315-322.
Carniel, E., Mazigh, D., and Mollaret, H. H. 1987. Expression of iron-regulated proteins in Yersinia species and their relation to virulence. Infection and Immunity 55:227-280.
Carsiotis, M., Weinstein, D. L., Helge, K., Holder, I. A., and O'Brien, A. D. Flagella of Salmonella tvphimurium are a virulence factor in infected C57BL/6J mice. Infection and Immunity 46:814-818.
Catfish Production, February 1, 1991. National Agricultural Statistics Service, United States Department of Agriculture.
Catfish Production, February 3, 1992. National Agricultural Statistics Service, United States Department of Agriculture.
Charnetzky, W. T., and Shuford, W. W. 1985. Survival and growth of Yersinia pestis within macrophages and an effect of the loss of the 47-megadalton plasmid on growth in macrophages. Infection and Immunity 47:234- 241.
Clerc, P. L., Ryter, A., Mounier, J., and Sansonetti, P. J. 1987. Plasmid-mediated early killing of eukaryotic cells by Shigella flexneri as studied by infection of J774 macrophages. Infection and Immunity 55:521-527.
Clerc, P., Baudry, B., and Sansonetti, P. J. 1988. Molecular mechanisms of entry, intracellular multiplication and killing of host cells by Shigella. Current Topics in Microbiology and Immunology 138:3- 13.
Cornells, G., Larroche, Y. Ballignad, G., Sory, M. P., and Wauters, G. 1987. Yersinia Enterocolitica. a primary model for bacterial invasiveness. Reviews of Infectious Diseases 9:64-87.
Cover, T. L. and Aber, R. C. 1989. Yersinia enterocolitica. New England Journal of Medicine 321:16-24. 140
Cross, A. S. 1990. The biologic significance of bacterial encapsulation. Current Topics in Microbiology and Immunology 150:87-95.
Cybulsky, M. I., Chan, W . , Movat, H. Z. 1988. Acute inflammation and microthrombosis induced by endotoxin, interleukin-1, and tumor necrosis factor and their implication in Gram-negative infection. Laboratory Investigation 58: 365-378.
Dautry-Varsat, A., and Lodish, H. F. 1984. How receptors bring proteins and particles into cells. Scientific American 250:52-57.
Davies, P. J. A., Cornwell, M. M., Johnson, J. D., Reggianni, A., Myers, M., and Murtaugh, M. P. 1984. Studies on the effects of dansylcadaverine and related compounds on receptor-mediated endocytosis in cultured cells. Diabetes Care Supplement 1 7:35.
Doyle, R. J., and Sonnenfeld, E. M. 1989. Properties of the cell surfaces of pathogenic bacteria. International Review of Cytology 118:33-92.
Duguid J. P., Darekar, M. R., and Wheater, D. W. 1976. Fimbriae and infectivity in Salmonella tvphimurium. Journal of Medical Microbiology 9:459-473.
Edelman, R., and Levine, M. M. 1986. Summary of an International Workshop on Typhoid Fever. Reviews of Infectious Diseases. 8:329-349.
Ellis, A. E. 1988. Current aspects of fish vaccination. Diseases of Aquatic Organisms 4:159-164.
Elwell, L., and Shipley, P. 1980. Plasmid-mediated factors associated with virulence of bacteria to animals. Annual Review of Microbiology 34:465-496.
Ewing, W. H., Ross, A. J., Brenner D. J., and Fanning, G. R. 1978. Yersinia ruckeri sp. nov., the redmouth (RM) bacterium. International Journal of Systematic Bacteriology 28:37:44.
Farmer, J. J., McWhorter, A. C. 1985. Genus X. Edwardsiella Ewing and McWhorter 1965, 37AL. Pages 486-491 in N. R. Krieg, and J. G. Holt, editors. Bergey's Manual of Systemic Bacteriology, ninth edition. The Williams and Wilkin Company, Baltimore, Maryland. 141
Farmer, J. J., Davis, B. R., Hickman-Brenner, F. W., McWhorter, A., Huntley-Carter, G. P., Asbury, M. A., Riddle, C., Wathen-Grady, H. G., Elias, C., Fanning, G. R., Steigerwalt, A. G., O'Hara, C. M., Morris, G. K., Smith, P. B., and Brenner, D. J. 1985. Biochemical identification of new species and biogroups of Enterobacteriaceae isolated from clinical specimens. Journal of Clinical Microbiology 21:46-76.
Fields, P. I., Swanson, R. V., Haidaris, C. G., and Heffron, F. 1986. Mutants of Salmonella tvphimurium that cannot survive within the macrophage are avirulent. Proceedings of the National Academy of Sciences of the United States of America 83:5189-5193.
Fields, P. I., Groisman, A. Heffron, F. 1989. A Salmonella locus that controls resistance to microbicidal proteins from phagocytic cells. Science 243:1059-1062.
Finlay, B. B., and Falkow, S. 1988a. Virulence factors associated with Salmonella species. Microbiological Sciences 5: 324-328.
Finlay, B. B., Gumbiner, B., and Falkow, S. 1988b. Penetration of Salmonella through a polarized Madin- Darby Canine Kidney epithelial cell monolayer. The Journal of Cell Biology 107:221-230.
Finlay, B. B., Heffron, F., Falkow, S. 1988c. Epithelial cell surfaces induce Salmonella proteins required for bacterial adherence and invasion. Science 243:940- 943.
Finlay, B. B . , and Falkow, S. 1988d. A comparison of microbial invasion strategies of Salmonella, Shigella, and Yersinia species. Pages 227-243 in M A. Horwitz, editor. Bacteria-Host Cell Interaction. Alan R. Liss, Incorporated, New York.
Finlay, B. B., and Falkow, S. 1988e. Comparison of the invasion strategies used by Salmonella choleraesuis. Shiaella flexneri. and Yersinia enterocolitica to enter cultured animal cells: endosome acidification is not required for bacterial invasion or intracellular replication. Biochimie 70:1089-1099.
Finlay, B. B., and Falkow, S. 1989a. Common themes in microbial pathogenicity. Microbiological Reviews 53:210—230. 142
Finlay, B. B., Fry, J., Rock, E. P., and Falkow, S. 1989b. Passage of Salmonella through polarized epithelial cells: role of the host and bacterium. Journal of Cell Science. Supplement 11:99-107.
Finlay, B. B., and Falkow, S. 1989c. Salmonella as an intracellular parasite. Molecular Microbiology 3:1833-1841.
Formal, S. B., and Lowenthal, J. P. 1956. Distribution of mucinolytic activity in strains of Shigella. Proceedings of the Society for Experimental Biology and Medicine 92:10-12.
Formal, S. B., Gemski, P. Jr., Baron, L. S., and LaBrec, E. H. 1971. A chromosomal locus which controls the ability of Shigella flexneri to evoke keratoconjunctivitis. Infection and Immunity 3:73-79.
Foster, T. 1983. Plasmid-determined resistance to antimicrobial drugs and toxic metal ions in bacteria. Microbiological Reviews 47:361-409.
Francis-Floyd, R., Beleau, M. H., Waterstrat, P. R., and Bowser, P. R. 1987. Effect of water temperature on the clinical outcome of infection with Edwardsiella ictaluri in channel catfish. Journal of the American Veterinary Medical Association 191:1413-1416.
Freund, J. D., Durborow, R. M., MacMillan, J. R., Crosby, M. D., Wellborn, T. L., Taylor, P. W., and Schwedler, T. L. 1990. Using diagnostic laboratory records to monitor occurrence of enteric septicemia of channel catfish in Mississippi. Journal of Aquatic Animal Health 2:207-211.
Gemski, P. Jr., Takeuchi, A., Washington, 0., and Formal, S. B. 1972a. Shigellosis due to Shigella dvsenteriae 1: relative importance of mucosal invasion versus toxin production in pathogenesis. Journal of Infectious Diseases 126:523-530.
Gemski, P. Jr., Sheahan, D. G . , Washington, 0., and Formal, S. B. 1972b. Virulence of Shigella flexneri hybrids expressing Escherichia coli somatic antigens. Infection and Immunity 6:104-111. 143
Griffiths, E., Stevenson, P., Hale, T. L., and Formal, S. B. 1985. Synthesis of aerobactin and a 76,000-dalton iron-regulated outer membrane protein by Escherichia coli K-12 - Shiaella flexneri hybrids and by enteroinvasive strains of Escherichia coli. Infection and Immunity 49:67-71.
Grizzle, J. M., and Rogers, W. A. 1976. Page 17 in Anatomy and Histology of the Channel Catfish. Auburn University Agricultural Experiment Station. Auburn, Alabama.
Groisman, E. A., and Saier, M. H. 1990. Salmonella virulence: new clues to intramacrophage survival. Trends in Biochemical Sciences 15:30-33.
Gulig, P. A., and Curtiss, R. 1987. Plasmid-associated virulence of Salmonella tvphimurium. Infection and Immunity 55:2891-2901.
Hale, T. L. , Oaks, E. V., and Formal, S. B. 1985. Identification and antigenic characterization of virulence-associated, plasmid-coded, proteins of Shigella spp. and enteroinvasive Escherichia coli. Infection and Immunity 50:620-629.
Hale, T. L., and Formal, S. B. 1986. Genetics of virulence in Shigella. Microbial Pathogenesis 1:511-518.
Hale, T. L., and Formal, S. B. 1987. Pathogenesis of Shigella infections. Pathology and Immunopathology Research 6:117-127.
Harlow, E. and Lane, D. 1988. Antibodies: A Laboratory Manual. Cold Springs Harbor Laboratory, Cold Spring Harbor, New York.
Hawke, J. P. 1979. A bacterium associated with disease of pond cultured channel catfish, Ictalurus punctatus. Journal of the Fisheries Research Board of Canada 36:1508-1512.
Hawke, J. P., McWhorter, A. C., Steigerwalt, A. G., and Brenner, D. J. 1981. Edwardsiella ictaluri sp. nov., the causative agent of enteric septicemia of catfish. International Journal of Systematic Bacteriology 31:396-400.
Hitchcock, P. J., and Brown, T. M. 1983. Morphological heterogeneity among Salmonella lipopolysaccharide chemotypes in silver-stained polyacrylamide gels. Journal of Bacteriology 154:269-277. 144
Hoffman, M. 1992. New clues to how bacteria get into cells. Science 255:35.
Hsu, S., Raine, L . , and Fanger, H. 1981. Use of avidin- biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. The Journal of Histochemistry and Cytochemistry 29:577-580.
Humphrey, J. D., Lancaster, C., Gudkovs, N., and McDonald, W. 1986. Exotic bacterial pathogens Edwardsie11a tarda and Edwardsiella ictaluri from imported ornamental fish Betta splendens and Puntius conchonius. respectively: Isolation and quarantine significance. Australian Veterinary Journal 63:369-371.
Hynes, R. 0. 1987. Integrins: a family of cell surface receptors. Cell 48:549-554.
Ibrahim, G. F., Fleet, G. H., Lyons, M. J., and Walker, R. A. 1985. Method for the isolation of highly purified Salmonella flagellins. Journal of Clinical Microbiology 22:1040-1044.
Isberg, R. R., and Falkow, S. 1985. A single genetic locus encoded by Yersinia pseudotuberculosis permits invasion of cultured animal cells by Escherichia coli K—12. Nature 317:262-264.
Isberg, R. R., Voorhis, D. L., and Falkow, S. 1987. Identification of invasin: a protein that allows enteric bacteria to penetrate cultured mammalian cells. Cell 50:769-778.
Isberg, R. R., and Leong, J. L. 1988. Cultured mammalian cells attach to the invasin protein of Yersinia pseudotuberculosis. Proceedings of the National Academy of Sciences of the United States of America 85:6682-6686.
Isberg, R. R., Swain, A., and Falkow, S. 1988. Analysis of expression and thermoregulation of the Yersinia pseudotuberculosis inv gene with hybrid proteins. Infection and Immunity 56:2133-2138.
Isberg, R. R. 1989. Two determinants for Yersinia Pseudotuberculosis cell binding and penetration detected in the absence of invasin. Infection and Immunity 57:1998-2005. 145
Isberg, R. R. 1990. Pathways for the penetration of Enteroinvasive Yersinia into mammalian cells. Molecular Biology and Medicine 7:73-82.
Janda, M. J., Abbott, S. L., and Oshiro, L. S. 1991. Penetration and Replication of Edwardsiella spp. in HEp-2 cells. Infection and Immunity 59:154-161.
Jann, K . , and Hoschutzky, H. 1990. Nature and organization of adhesins. Current Topics in Microbiology and Immunology 151:55-70.
Jarboe, H. H., Bowser, P. R., and Robinette, H. R. 1984. Pathology associated with a natural Edwardsiella ictaluri infection in channel catfish (Ictalurus nunctatus Rafinesque). Journal of Wildlife Diseases 20:353-354.
Johnson, K. G., and Perry, M. B. 1975. Improved techniques for the preparation of bacterial lipopolysaccharides. Canadian Journal of Microbiology 22:29-34.
Johnson, M. R. 1989. Culture methods and diseases of channel catfish Ictalurus punctatus. Proceedings— Annual Meeting of the United States Animal Health Association 93:86-100.
Jones, G. W., and Richardson, L. A. 1981. The attachment to, and invasion of HeLa cells by Salmonella typhimurium: the contribution of mannose-sensitive and mannose-resistant hemagglutinating activities. Journal of General Microbiology 127:361-370.
Kapperud, G., Skarpeid, H. J., Solberg, R. and Bergan, T. 1985. Outer membrane proteins and plasmids in different Yersinia enterocolitica serogroups isolated form man and animals. Acta Pathologica Microbiologica, Et Immunologica Scandinavica. Section B Microbiology 93:27—35.
Karkhanis, Y. D., Zettner, J. A., Jackson, J.J., and Carlo, D. J. 1978. A new and improved microassay to determine 2-keto-3-deoxyoctanate in lipopolysaccharide of Gram- negative bacteria. Analytical Biochemistry 85:595- 601.
Kasornchandra, J., Rogers, W. A., and Plumb, J. A. 1987. Edwardsiella ictaluri from walking catfish, Clarias batrachus L., in Thailand. Journal of Fish Diseases 10:137-138. 146
Kawaoka, Y., Otsuki, K., and Tsubokura, M. 1983. Growth temperature dependent variation in the bacteriophage- activated lipopolysaccharide. Journal of General Microbiology 129:2739-2747.
Kent, M. L., and Lyons, J. M. 1982. Edwardsiella ictaluri in the green knife fish, Eiaenmannia virescens. Fish Health News, 2:ii.
Kihlstrom, E., and Magnusson, K. E. 1983, Hemagglutinating, adhesive and physicochemical surface properties of different Yersinia enterocolitica and Yersinia enterocolitica-like bacteria. Acta Pathologica Microbiologica, Et Immunologica Scandinavica. Section B Microbiology 91:113-119.
Kohbata, S., Yokoyama, H., and Yabuuchi, E. 1986. Cytopathogenic effect of Salmonella tvohi GIFU 10007 on M cells of murine ileal Peyer's patches in ligated ileal loops: an ultrastructural study. Microbiology and Immunology 30:1225-1237.
Kopecki, D., and Formal, S. 1984. Plasmids and the virulence of enteric and other bacterial pathogens. Annals of Internal Medicine 101:260-262.
Labigne-Roussel, A. F., Lark, D., Schoolnik, G., and Falkow, S. 1984. Cloning and expression of an afimbrial adhesin (AFA-I) responsible for P blood group-independent, mannose-resistant hemagglutination from a pyelonephritic Escherichia coli strain. Infection and Immunity 46:251-259.
Laemrali, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685.
Lawlor, K. M . , and Payne, S. M. 1984. Aerobactin genes in Shigella spp. Journal of Bacteriology 160:226-272.
Lawlor, K. M., Daskaleros, P. A., Robinson, R. E., and Payne, S. M. 1987. Virulence of iron transport mutants of Shigella flexneri and utilization of host iron compounds. Infection and Immunity 55:594-599.
Lee, C. A., and Falkow, S. 1990. The ability of Salmonella to enter mammalian cells is affected by bacterial growth state. Proceedings of the National Academy of Sciences of the United States of America 87:43 04-4308. 147
Lee, W. H., McGrath, P. H., Carter, P. H., and Eidie, E. L. 1977. The ability of some Yersinia enterocolitica strains to invade HeLa cells. Canadian Journal of Microbiology 23:1714-1722.
Li, Y. and Lovell, R. T. 1985. Elevated levels of dietary ascorbic acid increase immune responses in channel catfish. Journal of Nutrition 115:123-131
Lian, C. J., and Pai, C. H. Inhibition of human neutrophil chemiluminescence by plasmid-mediated outer membrane proteins of Yersinia enterocolitica. Infection and Immunity 49:145-51.
Liu, S. L., Ezaki, T., Miura, H., Matsui, K., and Yabuuchi, E. 1988. Intact motility as a Salmonella tvphi invasion-related factor. Infection and Immunity 56:1967-1973.
Liu, P. R., Plumb, J. A., Guerin, M., and Lovell, R. T. 1989. Effect of megalevels of dietary vitamin C on the immune response of channel catfish Ictalurus punctatus in ponds. Diseases of Aquatic Organisms 7:191-194.
Lobb, C. J., and Rhoades, M. 1987. Rapid plasmid analysis for identification of Edwardsiella ictaluri from infected channel catfish (Ictalurus punctatus^. Applied and Environmental Microbiology 53:1267-1272.
Luderitz, 0., Freudenberg, M. A., Galanos, C., Lehmann, V., Rietchel, E. T . , and Shaw, D. H. 1982. Lipopolysaccharides of Gram-negative bacteria. Current Topics in Membrane Transport 17:79-151.
MacMillan, J. R. 1985. Infectious diseases. Pages 405-441 in C. S. Tucker, editor. Channel Catfish Culture. Elsevier, Amsterdam— Oxford— New York— Tokyo.
Macnab, R. M., and DeRosier, D. J. 1988. Bacterial flagellar structure and function. Canadian Journal of Microbiology 34:442-451.
Makino, S., Sasakawa, C., Kamata, K., Kurata, T., and Yoshikawa, M. 1986. A genetic determinant required for the continuous reinfection of adjacent cells on large plasmid in S. flexneri 2a. Cell 46:551-555.
Markwell, M. K., Haas, S. M., Bieber, L. L., and Tolbert, N. E. 1978. A modification of the Lowery procedure to simplify protein determination in membrane and lipoprotein samples. Analytical Biochemistry 87:206- 210. 148
Martinez, R. J. 1983. Plasmid-mediated and temperature- regulated surface properties of Yersinia enterocolitica. Infection and Immunity 41:921-930.
Martinez, J. L., Delgado-Iribarren, A., and Baquero, F. 1990. Mechanisms of iron acquisition and bacterial virulence. Federation of European Microbiological Societies Microbiological Reviews 75:45-56.
Matsui, K., and Arai, T. 1989. Specificity of Salmonella porin as an eliciting antigen for cell-mediated immunity (CMI) reaction in murine salmonellosis. Microbiology and Immunology 33:1063-1067.
Maurelli, A. T., Blackmon, B . , and Curtiss, R. III. 1984. Temperature dependent expression of virulence genes in Shigella species. Infection and Immunity 43:195-201.
Maurelli, A. T., and Sansonetti, P. J. 1988a. Genetic determinants of Shigella pathogenicity. Annual Review of Microbiology 42:127-150.
Maurelli, A. T., and Sansonetti, P. J. 1988b. Identification of a chromosomal gene controlling temperature regulated expression of Shigella virulence. Proceedings of the National Academy of Sciences of the United States of America 85:2820-2824.
Mellman, I., Fucho, R., and Helenius, A. 1986. Acidification of the endocytic and exocytic pathways. Annual Review of Biochemistry. 55:663-700.
Miller, V. L., Finlay, B. B., and Falkow, S. 1988. Factors essential for penetration of mammalian cells by Yersinia. Current Topics in Microbiology and Immunology 138:15-39.
Miller, V. L., and Falkow, S. 1988. Evidence for two genetic loci from Yersinia enterocolitica that can promote invasion of epithelial cells. Infection and Immunity 56:1242-1248.
Miller, J. F., Mekalanos, J. J., and Falkow, S. 1989a. Coordinate regulation and sensory transduction in the control of bacterial virulence. Science 243:916-922.
Miller, V. L., Farmer, J. T., Hill, W. E., and Falkow, S. 1989b. The ail locus is found uniquely in Yersinia enterocolitica serotypes commonly associated with disease. Infection and Immunity 57:121-131. 149
Miyazaki, T., and Kaige, N. 1985. Comparative Histopathology of Edwardsiellosis in Fishes. Fish Pathology 20:219-227.
Miyazaki, T., and Plumb, J. A. 1985. Histopathology of Edwardsiella ictaluri in channel catfish, Ictalurus punctatus (Rafinesque). Journal of Fish Diseases 8:389-392.
Mostov, K. E., and Simister, N. E. 1985. Transcytosis. Cell 43:389-390.
Moulder, J. W. 1985. Comparative biology of intracellular parasitism. Microbiological Reviews 49:298-3 37.
Mroczenski-Wildey, M. J., Di Fabio, J. L., and Cabello, F. C. 1989. Invasion and lysis of HeLa cell monolayers by Salmonella tvohi: the role of lipopolysaccharide. Microbial Pathogenesis 6:143-152.
Mulder, J. B. 1971. Shigellosis in nonhuman primates. A review. Laboratory Animal Science 21:734-738.
Newton, J. c., Bird, R. c., Blevins, W. T., wilt, G. R., and Wolfe, L. G. 1988. Isolation, characterization, and molecular cloning of cryptic plasmids isolated from Edwardsiella ictaluri. American Journal of Veterinary Research 49:1856-1860.
Newton, J. C., Wolfe, L. G., Grizzle, J. M., and Plumb, J. A. 1989. Pathology of experimental enteric septicaemia in channel catfish, Ictalurus punctatus (Rafinesque), following immersion-exposure to Edwardsiella ictaluri. Journal of Fish Diseases 12:335-347.
Newton, J. C., Blevins, W. T., Wilt, G. R., and Wolfe, L. G. 1990. Outer membrane protein profiles of Edwardsiella ictaluri from fish. American Journal of Veterinary Research 51:211-215.
Newton, J. C., and Triche, P. L. Isolation and characterization of flagella from Edwardsiella ictaluri. Journal of Aquatic Animal Health (In press, a) .
Newton, J. C., and Triche, P. L. Electrophoretic and immunochemical characterization of lipopolysaccharide of Edwardsiella ictaluri from channel catfish. Journal of Aquatic Animal Health (In press, b). 150
Nomura, J., and Takashi, A. 1985. Morphological analysis of lipopolysaccharide from Gram-negative fish pathogenic bacteria. Fish Pathology 20:193-197.
O'Brien, A. D., Thompson, M. R., Gemski, P., Doctor, B. P., and Formal, S. B. 1977. Biological properties of Shiaella flexneri 2a toxin and its serological relationship to Shigella dvsenteriae 1 toxin. Infection and Immunity 15:796-798.
Oaks, E. V., Wingfield, M. E., and Formal, S. B. 1985. Plaque formation by virulent Shigella flexneri. Infection and Immunity 48:124-129.
Okamura, N., Nagai, T., Nakaya, R., Kondo, S., Murakami, M., and Hisatsune, L. 1983. HeLa cell invasiveness and 0-antigen of Shigella flexneri as separate and prerequisite attributes of virulence to evoke keratoconjunctivitis in guinea pigs. Infection and Immunity 39:505-513.
Orskov, I. F., Orskov, F., Jann, B., and Jann, K. 1977. Serology, chemistry, and genetics of 0 and K antigens of Escherichia coli. Bacteriological Review 41:667- 710.
Pai, C. H., and Mors, V. 1978. Production of enterotoxin by Yersinia enterocolitica. Infection and Immunity 19:908-911.
Pal, T., Formal, S. B., and Hale, T. L. 1989. Characterization of virulence marker antigen of Shigella spp. and enteroinvasive Escherichia coli. Journal of Clinical Microbiology 27:561-563.
Perry, R. D. , and Brubaker, R. R. 1979. Accumulation of iron by Yersinia. Journal of Bacteriology 137:1290- 1298.
Plumb, J. A., and Schwedler, T. E. 1982. Enteric septicemia of catfish (ESC): a new bacterial problem surfaces. Aquaculture Magazine (May-June):26-27.
Plumb, J. A., Sanchez, D. J. 1983. Susceptibility of five species of fish to Edwardsiella ictaluri. Journal of Fish Diseases 6:261-266.
Plumb, J. A., and Hilge, V. 1986. Susceptibility of Europe an catfish (Silurus glanis) to Edwardsiella ictaluri. Journal of Applied Ichthyology 3:45-48. 151
Plumb, J. A., and Quinlan, E. E. 1986. Survival of Edwardsiella ictaluri in pond water and bottom mud. The Progressive Fish-Culturist 48:212-214.
Plumb, J. A., Wise, M. L., and Rogers, W. A. 1986. Modulary effects of temperature on antibody response and specific resistance to challenge of channel catfish, Ictalurus punctatus, immunized against Edwardsiella ictaluri. Veterinary Immunology and Immunopathology 12:297-304.
Plumb, J. A., Maestrone, G . , and Quinlan, E. 1987. Use of a potentiated sulfonamide to control Edwardsiella ictaluri infection in channel catfish (Ictalurus punctatus). Aquaculture, 62:187-194.
Plumb, J. A. 1988. Vaccination against Edwardsiella ictaluri. Pages 152-161 in A. E. Ellis, editor. Fish Vaccination. Academic Press Limited, London, United Kingdom.
Plumb, J. A., and Klesius, P. 1988. An assessment of the antigenic homogeneity of Edwardsiella ictaluri using monoclonal antibody. Journal of Fish Diseases 11:499- 509.
Plumb, J. A., and Vinitnantharat, S. 1989. Biochemical, biophysical, and serological homogeneity of Edwardsiella ictaluri. Journal.of Aquatic Animal Health 1:51-56.
Portnoy, D. A., Moseley, S. L., and Falkow, S. 1981. Characterization of plasmids and plasmid-associated determinants of Yersinia enterocolitica pathogenesis. Infection and Immunity 31:775-782.
Portnoy, D. A., Wolf-Watz, H., Bolin, I., Breeder, A. B., and Falkow, S. 1984. Characterization of common virulence plasmids in Yersinia species and their role in the expression of outer membrane proteins. Infection and Immunity 43:108-114.
Portnoy, D. A., and Martinez, R. J. 1985. Role of a plasmid in the pathogenicity of Yersinia species. Current Topics in Microbiology and Immunology 118:29-51.
Prizont, R., and Reed, W. P. 1980. Possible role of colonic content in the mucosal association of pathogenic Shigella. Infection and Immunity 29:1198-1199. 152
Prizont, R. 1982. Degradation of intestinal glycoproteins by pathogenic Shigella flexneri. Infection and Immunity 36:615-620.
Puohiniemi, R. , Karvonen, M., Vuopio-Varkila, J., Muotiala, A., Helander, I. M., and Sarvas, M. 1990. A strong antibody response to the periplasmic C-terminal domain of the OrapA protein of Escherichia coli is produced by immunization with purified OmpA or with whole E^_ coli or Salmonella tvphimurium bacteria. Infection and Immunity 58:1691-1696.
Reid, G., and Sobel, J. D. 1987. Bacterial adherence in the pathogenesis of urinary tract infection: a review. Reviews of Infectious Diseases 9:470-487.
Reid, W. S., and Boyle, J. A. 1989. Plasmid homologies in Edwardsiella ictaluri. Applied and Environmental Microbiology 55:3253-3255.
Reitmeyer, J. C., Peterson, J. W., and Wilson, K. J. 1986. Salmonella cytotoxin: a component of the bacterial outer membrane. Microbial Pathogenesis 1:503-510.
Robins-Browne, R. M., and Prpic, J. K. 1985. Effects of iron and desferrioxamine on infections with Yersinia enterocolitica. Infection and Immunity 47:774-779.
Rogers, W. A. 1981. Serological detection of two species of Edwardsiella infecting catfish. Developments in Biological Standardization 49:169-172.
Rout, W. R., Formal, S. B., Giannalla, R. A., and Dammin, G. J. 1975. The pathophysiology of Shigella diarrhea in the rhesus monkey: intestinal transport, morphology, and bacteriological studies. Gastroenterology 68:270-278.
Saeed, M. O., and Plumb, J. A. 1986. Immune response of channel catfish to lipopolysaccharide and whole cell Edwardsiella ictaluri vaccines. Diseases of Aquatic Organisms 2:21-25.
Saeed, M. 0., and Plumb, J. A. 1987. Serological detection of Edwardsiella ictaluri Hawke lipopolysaccharide antibody in serum of channel catfish, Ictalurus punctatus Rafinesque. Journal of Fish Diseases 10:2 05- 209. 153
Sakai, T., Sasakawa, C., Makino, S., and Yoshikawa, M. 1986a DNA sequence and product analysis of the virF locus responsible for congo red binding and cell invasion in Shigella flexneri 2a. Infection and Immunity 54:395-402.
Sakai, T., Sasakawa, C. , Makino, S., Kamata, K., and Yoshikawa, M. 1986b. Molecular cloning of a genetic determinant for congo red binding ability which is essential for the virulence of Shigella flexneri. Infection and Immunity 51:476-482.
Sansonetti, P. J., Kopecko, D. J., and Formal, S. B. 1981. Shigella sonnei plasmids: evidence that a large plasmid is necessary for virulence. Infection and Immunity 34:75-83.
Sansonetti, P. J., Kopecko, K. J., and Formal, S. B. 1982 Involvement of a plasmid in the invasive ability of Shigella flexneri. Infection and Immunity 35:852-860.
Sansonetti, P. J., Hale, T.L., Dammin, G. J., Kapfer, C., Collins, H. H., and Formal, S. B. 1983a. Alterations in the pathogenicity of Escherichia coli K-12 after transfer of plasmid and chromosomal genes from Shigella flexneri. Infection and Immunity 39:1392- 1402.
Sansonetti, P. J., d/Hauteville, H., Ecobichon, C., and Pourcel. C, 1983b. Molecular comparison of virulence plasmid in Shigella and enteroinvasive Escherichia coli. Annales De L Institut Pasteur. Microbiologie 134A:295-318.
Sansonetti, P. J., Ryter, A., Clerc, P., Maurelli, A. T., and Mounier, J. 1986. Multiplication of Shigella flexneri within HeLa cells: lysis of the phagocytic vacuole and plasmid-mediated contact hemolysis. Infection and Immunity 51:461-469.
Sansonetti, P. J., and Mounier, J. 1987. Metabolic events mediating early killing of host cells infected by Shigella flexneri. Microbial Pathogenesis 3:53-61.
SAS Institute Inc, 1988. Pages 593-600. SAS Institute Inc. SAS/STAT™ User's Guide, Release 6.03 Edition. Cary, NC: SAS Institute Inc. 1988. Pages 593-600. 154
Shotts, E. B., Blazer, V. s., and Waltman, W. D. 1985. Pathogenesis of experimental Edwardsiella ictaluri infections in channel catfish fIctalurus punctatus^. Canadian Journal of Fisheries and Aquatic Sciences 43:36-42.
Shotts, E. B., and Waltman, W. D. 1990. A medium for the selective isolation of Edwardsiella ictaluri. Journal of Wildlife Diseases 26:214-218.
Simons, K., and Fuller, S. D. 1985. Cell surface polarity in epithelia. Annual Review of Cell Biology 1:243-288.
Small, P. L. C., Isberg, R. R . , and Falkow, S. 1987. Comparison of the ability of enteroinvasive Escherichia coli. Salmonella tvphimurium. Yersinia pseudotuberculosis. and Yersinia enterocolitica to enter and replicate within HEp-2 cells. Infection and Immunity 55:1674-1679.
Speyerer, P. D., and Boyle, J. A. 1987. The plasmid profile of Edwardsiella ictaluri. Journal of Fish Diseases 10:461-469.
Stanley, L. A. 1991. Electron microscopic observations of channel catfish macrophages infected with Edwardsiella ictaluri. Pages 84-85 in G. W. Bailey, editor. Proceedings of the 49th Annual Meeting of the Electron Microscopy Society of America. San Francisco Press, Incorporated, San Francisco, California.
Starliper, C. E., Schill, W. B., Shotts, E. B., and Waltman, W. D. 1988. Isozyme analysis of Edwardsiella ictaluri. Microbios Letters 37:81-87.
Stocker, B. A., and Makela, P. H. 1986. Genetic determination of bacterial virulence, with special reference to Salmonella. Current Topics in Microbiology and Immunology 124:149-172.
Struelens, M. J., Patte, D., Kabir, I., Salam, A., Nath, S. K., and Butler, T. 1985. Shigella septicemia: prevalence, presentation, risk factors, and outcome. Journal of Infectious Diseases 152:784-790.
Sturm, S., Jann, B., Jann, K., Fortnagel, P., and Timmis, K. N. 1986. Genetic and biochemical analysis of Shigella dvsenteriae 1 O-antigen polysaccharide biosynthesis in Escherichia coli K-12: structure and functions of the rfb gene cluster. Microbial Pathogenesis 1:307-320. 155
Takeuchi, A. 1967. Electron microscopic studies of experimental Salmonella infection. American Journal of Pathology 50:109-13 6.
Tannock, G. W. Blumershine, R. V. H., and Savage, D. C. 1975. Association of Salmonella typhimurium with, and its invasion of, the ileal mucosa in mice. Infection and Immunity 11:3 65-370.
Timmis, K. N., Clayton, C. L., and Sekizaki, T. 1985. Localization of shiga toxin gene in the region of Shigella dvsenteriae 1 specifying virulence functions. Federation of European Microbiological Societies Microbiology Letters 30:301-305.
Uhlman, D. L . , and Jones, G. W. 1982. Chemotaxis as a factor in interactions between HeLa cells and Salmonella typhimurium. Journal of General Microbiology 128:415-418.
Une, T. 1977a. Studies on the pathogenicity of Yersinia enterocolitica. I. Experimental infection in rabbits. Microbiology and Immunology 21:341-363.
Une T. 1977b. Studies on the pathogenicity of Yersinia pseudotuberculosis. II. Interaction with cultured cells in vitro. Microbiology and Immunology 21:365- 377.
Une, T. 1977c. Studies on the pathogenicity of Yersinia enterocolitica. III. Comparative studies between Y. enterocolitica and Y. pseudotuberculosis. Microbiology and Immunology 21:505-506.
Vesikari, T., Bromirska, J . , and Maki, M. 1982. Enhancement of ivasiveness of Yersinia enterocolitica and Escherichia coli in HEp-2 cells by centrifugation. Infection and Immunity 36:834-836.
Waltman, W. D., Shotts, E. B., and Blazer, V. S. 1985. Recovery of Edwardsiella ictaluri from Danio (Danig devajrig) . Aquaculture 46:63-66.
Waltman, W. D., and Shotts, E. B. 1986. Antimicrobial susceptibility of Edwardsiella ictaluri. Journal of Wildlife Diseases 22:173-177.
Waltman, W. D., Shotts, E. B., and Hsu, T. C. 1986. Biochemical characteristics of Edwardsiella ictaluri. Applied and Environmental Microbiology 51:101-104. 156
Waterstrat, P., Ainsworth, J., and Capley, G. 1989. Use of an indirect enzyme-linked immunosorbent assay (ELISA) in the detection of channel catfish, Ictalurus punctatus (Rafinesque), antibodies to Edwardsiella ictaluri. Journal of Fish Diseases 12:87-94.
Weete, J. D., Blevins, W. T., and Chitrakorn, S. 1988. Chemical characterization of lipopolysaccharide from Edwardsiella ictaluri. a fish pathogen. Canadian Journal of Microbiology 34:1224-1229.
Westphal, 0., and Jann, K. 1965. Bacterial lipopolysaccharide extraction with phenol-water and further applications of the procedure. Pages 83-96 in R. L. Whistler, editor. Methods in Carbohydrate Chemistry, Volume 5. Academic Press, Incorporated, New York, New York.
Williams, P. 1988. Role of the cell envelope in bacterial adaptation to growth in vivo in infections. Biochimie 70:987-1011.
Wong, J. D., Miller, M. A., and Janda, J. M. 1989. Surface properties and ultrastructure of Edwardsiella species. Journal of Clinical Microbiology 27: 1797-1801.
Yancy, R. J., Breeding, S. A., and Lankford, C. E. 1979. Enterochelin (enterobactin): virulence factor for Salmonella typhimurium. Infection and Immunity 24:174- 180. VITA
Thomas J. Baldwin was born to LaGrand and Joan
(Reichert) Baldwin on March 20, 1960 in Salt Lake City,
Utah, as the last of six children. Tom graduated from
Skyline High School in Holladay, Utah in 1978. Following high school, he enrolled first at the University of Utah and subsequently at Brigham Young University, where he completed the required pre-veterinary curriculum and was admitted to the College of Veterinary Medicine at
Washington State University. He received his Bachelor of
Science degree in Veterinary Science from Washington State
University in December 1985 and his Doctor of Veterinary
Medicine (DVM) 3 years later, in May of 1988.
After receiving his DVM, Tom accepted a graduate assistant position in the Department of Veterinary
Pathology where he has pursued his Doctor of Philosophy degree. Upon completion of his degree, Tom will accept an assistant professor position at Washington State
University. He is married to Sandra Jean (Harris) Baldwin and has four children, Jessica, age 7 years, Matthew, age 5 years, Sarah, age 3 years, and Jason, age 1 year.
157 DOCTORAL EXAMINATION AND DISSERTATION REPORT
Candidate: Thomas James Baldwin
Major Field: Veterinary Medical Sciences (Vet. Pathology)
Title of Dissertation: Edwardsiella ictaluri: Pathogenic Mechanisms and Antigens Recognized by Infected Channel Catfish
Approved:
.(ZW a c a Jv /J Major Professor and Chairman
^ J? Dean of the Gradddte School
EXAMINING COMMITTEE:
D
Date of Examination:
September 1, 1992