 

.3ij «w.

I.

 

airﬂmww
JWV§.H(LL. n.8, .

; . .3513. x
Amwﬂr
.. l. .

am

u...
5.5...

 

.
a .u

5!. 1
5943.3.
3...

s
hwy! ﬁns. ;
11.5.1... (I.

3.1,].

331..

.ﬂsnbgmmuifuum
. d? if?
.mnﬁumwrwithi. IﬁHW
.~ 4 u "I 31".] .l
s W
"WU“.HPI. .4305:

g $.55
.II(

 

. ... Esau,» 1.." 5.

ii... S... :2

n 5
awznuﬁf
“Wﬂﬂhuﬂd

.3. w 15m... Mn...“

 

Wyn)
,.

d€03

65mm)“

This is to certify that the

dissertation entitled

Bioactivities of Bovine Anti-coliform

Antibodies Elicited by J5 Vaccinations

presented by

Anantachai Chaiyotwittayakun

has been accepted towards fulfillment
of the requirements for

Eh-“ - degree in _Laa:g.e_Anima1 Clinical

Sciences

 

 

DMe May 7, 2003

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c-JCIRC/Daleouepss-sz

BIOACTIVITIES OF BOVINE ANTI-COLIFORM
ANTIBODIES ELICITED BY J5 VACCINATIONS

By

Anantachai Chaiyotwittayakun

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
~ DOCTOR OF PHILOSOPHY

Department of Large Animal Clinical Sciences

2003

ABSTRACT

BIOACTIVITIES OF BOVINE AN TI—COLIFORM
ANTIBODIES ELICITED BY J5 VACCINATIONS

By

Anantachai Chaiyotwittayakun

Five healthy Holstein steers were immunized 12 times with a J 5 bacterin at time
0, 30 days later, and every 2 weeks for 10 subsequent immunizations. Sera were
collected from pre-immunization and on days 4, 6, 11, and 13 post immunizations, and
assayed for relative contents of anti-15 IgM, IgG1, and Ing antibodies by ELISA.
Selected sera from pre- and 6-day-post-3‘d, 6m, 9m, and 12th immunization were used to
determine isotype speciﬁc endpoint antibody titers and cross-reactivity against Gram-
negative bacteria isolated from clinical disease cases. Immunization number and day post
immunization signiﬁcantly (P < 0.0001) inﬂuenced anti-J5 E. coli antibody responses for
all 3 isotypes. Two immunizations were required to increase (P < 0.005) serum anti-J5 E.
coli IgM antibody levels above pre-immunization. However, 5 immunizations were
required to signiﬁcantly (P s 0.02) increase anti-J5 IgG1 and Ing antibody levels and
detect isotype switching. Anti-15 antibody titers for all isotypes were signiﬁcantly (P <
0.005) higher than background after 6 immunizations. Cross-reactivity of isotype
speciﬁc anti-15 antibodies to all Gram-negative bacteria in ELISA was signiﬁcantly (P <
0.05) increased with hyperimmunization. Anti-15 IgM and Ing antibodies were more

cross-reactive to

Serratia spp. , E. coli 487 and K. pneumoniae than other bacteria tested. Anti-15 IgG1
antibodies tended to cross-react with E. coli 487 and Pseudomonas Spp. more than other
bacteria tested. However, only low to undetectable cross-reactivity of isotype speciﬁc
anti-15 antibodies with puriﬁed LPS or Lipid A was observed, which was similar to the
result for S. aureus used as a negative control in ELISA.

Western blot analysis showed that the anti-15 IgG1 and Ing recognized at least
two predominant antigen clusters with molecular masses of 8-105 and 34-39 kDa An
identical pattern of 34-39 kDa antigen recognition by Ing antibody was observed in
commercial fetal bovine serum and in J5 antisera from the pre-, 3“, and 9‘11 immunization
collections. However, the 8-105 kDa proteins appeared to be the target antigens in the J5
bacterin because the speciﬁc recognition of these antigens by IgGl or Ing antibodies
was only observed when J5 hyperimmune sera were used in the assays. Western blotting
also revealed strong recognition of LPS by anti-J5 IgG1 antibodies. However, there was
little or no recognition of Lipid A or S. aureus proteins by any antiserum or isotype.
Protein A enriched Ing antibody from the hyperimmune serum signiﬁcantly (P<0.001)
enhanced neutrophil phagocytosis of GFP-J 5 E. coli compared to no antibody or enriched
IgG. antibody, especially with incubation times from 120 to 150 minutes. Thus, results
from this work indicated that J5 hyperimmunization elicited high levels of cross-reactive
serum Ing antibodies that increasingly recognized small (8-105 kDa) and- larger (34-39
kDa) antigens on most Gram-negative bacteria and promoted effective opsonization of E.
coli for enhanced neutrophil phagocytosis of the bacteria Increasing the number of
exposures of the bovine immune system to J5 bacterin elicits antibodies that should

protect animals from a variety of Gram-negative infectious diseases.

ACKNOWLEDGEMENTS

This study was greatly supported by the Michigan Agricultural Experiment
Station, Michigan State University-Research Excellence Funds, and the Michigan Animal
Initiative Coalition. I would like to thank Ofﬁce of the Civil Service Commission, Royal
Thai Government, Bangkok, Thailand for providing the ﬁnancial support for my PhD.
stipend, also the government officers and staffs over the Royal Thai Embassy,
Washington, DC for their kindness, and all convenience while I studied in the United
States.

I would like greatly acknowledge Dr. Ronald J. Erskine (Uncle Ron, my major
advisor), and my committee members, Dr. Jeanne L. Burton (also co-advisor), Dr. Patty
S. D. Weber, Dr. Paul C. Bartlett, Dr. Phillip M. Sears, Dr. James S. Cullor (an external
reviewer) for your great support and guidance to help me build research experience and
knowledge. Special thanks to Drs. Burton and Weber for guiding me through the world
of science. I want to thank all of you and your families for giving me opportunities to
leam a lot about American culture on many occasions, and for making me feel
comfortable while I have been thousands miles away from home.

I also thank Dr. James S. Cullor (University of California, Davis, CA) for
donating the J 5 E. coli used for ELISA work, Dr. Bruce A. Beachnau (Technical Service
Consultant, Phannacia & Upjohn Animal Health, Kalamazoo, MI) for donating the J5
Bacterin Vaccine, Mr. & Mrs. Powell (Powell Farm, Haslett, Michigan) for use of the
animal use in this study, and Ms. Karen L. Smith and Mrs. Christine R. Phipps for

assistance with the animal work. All my hard work cannot be complete without these

iv

places: the Immunogenetics Laboratory, the Mastitis Research Laboratory, the Meat
Animal Research Laboratory (Dr. Doumit), and the Molecular Pathogenesis Laboratory
(Dr. Paul Coussens).

I want to thanks many friends; Jennifer (Wells) Jacob, Rebecca Darch, Christine
R. Phipps, Sally Madsen, and Valencia Rilington for their love and support. All great
help from many students including Jason van Lente, Andy Smith, Michelle Taylor,
Amanda Goeman, Jenny Gardner, and Liz Konoski is acknowledged, as is their
assistance on my laboratory work.

At last, I would like to thank my grandfather (Prasit Phakam), my parents
(Pongpetch & Nurot), Anucha my youngest brother, and my beloved brother Ponganan
(who passed away during my PhD. studies) for your love and support And thanks for
always giving me a great inspiration to have a strong mind and to ﬁght any obstacle in

my life in every moment. Thanks to everybody for everything.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... v
LIST OF FIGURES ............................................................................................................ vi
KEY TO ABBREVIATIONS .......................................................................................... xiii
INTRODUCTION ............................................................................................................... 1
CHAPTER 1
REVIEWOF LITERATURE
15 Escherichia coli bacterin .................................................................................... 5
Basic structure of Gram-negative bacteria... . .. .. . .. 5
Core antigens elicit protective immunity against Gram-negatrve bacterial
mwwm ”m ...mnmnmumumumumumnmum9
Bovine antibody response and mastitis. ... ....10
Maturation ofbovine antibody... .. .16
Colifonn mastitis... .. .. ...20
Inﬂammatory and systemic response to 1ntramammary invading pathogens. ..22
'Coliform mastitis versus core antigen technology. .... .......24
CHAPTER 2

ISOTYPE SPECIFIC SERUM ANTI-COLIFORM ANTIBODIES CROSS-REACT
WITH GRAM-NEGATIVE BACTERIA
1. Introduction... .. 30
2. Materials andMethods 32
2.]
2.2.15 Immumzat1on schedule .32
2. 3 Blood collections... .....33
2. 4 Preparation of phenol- killed whole J5 Escherichia co'li for use in
ELISA... .. ...........34
2. 5 Isotype speciﬁc anti-J5 E. coli antibody response proﬁles .............. 35
2. 6 Isotype speciﬁc anti-J5 E. coli antibody titers in selected serum
samples... 37
2.7Antibody cross-reactwrty assay. 37
2.88tatisticalanalysis... 38

vi

3. Results... 39
3 1 Hyperimmune response proﬁles ..............39
3. 2 Isotype speciﬁc anti- J5 E. coli antibody endpomt titers ................. 40
3. 3 Isotype speciﬁc antibody cross-reactivity... ...40
4. Discussion... . 41
5. Conclusron45

CHAPTER3
GRAM-NEGATIVE BACTERIAL ANTIGEN RECOGNITION BY ISOTYPE
SPECIFIC ANTI-J5 E. coli ANTIBODY
Abstract... 56
1. Introduction... .. 57
2. Materials andMethods... 59
2. 1 Testand controlsamples... .............59
2. 2 Protein A puriﬁcation of serum lgG1 and Ing antlbodres .....59
2. 2.1 Ammonium sulfate precipitation .................................. 59
2. 2. 2 Dialysis... 61
2. 2. 3 Protein A afﬁnity chromatography 62
2.2.4 Quality test for puriﬁed antibodies. 6.........3
2.3 Preparation of bacterial antigens ............................................... 64
2.4 Assays for bacterial antigen recognition by anti-J5 antibodies. .. ...66
2.4.1 SDS-PAGE66
2.4.2 Silverstaining procedure......................................... 67
2.4.3Westemblotanalysesu 68
2 5 Opsonization assay. 69
2.5.1 Preparation of green fluorescent protern—transformed
JSE. coli... 69
2520psonization ......70
2. 6 Neutrophil isolation... .. . .71
2. 7 Phagocytosis assay and flow cytometric analysis of phagocytosrng
neutrophils... 72
2.8 Statistical analysrs73
3. Results73
4. Drscussron80
5.Conclusron89

CHAPTER4
SUMMARYANDCONCLUSIONS...... 90

CHAPTERS

CHAPTER 6
REFERENCES ....................................................................................................................... 97

LIST OF TABLES

Table 1 Molecular size characteristics of the major bovine antibody isotypes ......... 14

Table 2 Normal serum and milk (> 2 weeks postpartum) concentrations of the major
antibody classes (isotypes) in dairy cows 14

Table 3 Mean anti-15 E. coli antibody response elicited by J5 E. coli bacterin from
seracollectedfromﬁveHolsteinsteers...................................................... 94

Table 4 Mean anti-J5 E. coli antibody endpoint titers of selected serum samples
collected from ﬁve Holstein steers... 95

Table 5 Mean percentage of phagocytosrng bovine neutrophil of opsonized GFP-
transformedJSE.coli... 96

viii

LIST OF FIGURES
Images in this dissertation are presented in color.

Figure 1 General structure of the Lipopolysaccharide (LPS) structures of smooth
and rough Gram-negative bacteria... ....

Figure 2 Schematic drawings of Wild type Escherichia coli (leﬁ) showing 0-
polysaccharide chains, and J5 E. coli, Rc mutant (right) showing exposed core and
Lipid A regions... .... .. .... .... ..

Figure 3 Schematic model of the cell wall of Gram-negative bacteria (adapted from
Kremeretal. ,...1990) . . .

Figure 4 Monomeric antibody molecules are composed of two heavy (dark gray)
andtwolight(white)chains... . . . ..

Figure 5 Antibody molecules are synthesized as monomers, such as shown for IgG
(middle), but IgM and IgA molecules are usually found in serum as multimers
[either as pentameric IgM (left) or dimeric IgA (right)]. ...

Figure 6 Antibodies effectively block pathogens from interacting with host cells
and may have varying degrees of toxin neutralization capacity (a) or opsonization

capacity (b). ..

Figure 7 Repeated exposure of the immune system to the same antigens on bacteria
promotes isotype switching to occur inB cells...

Figure 8 B cell activation with subsequent irnmunoglobulin isotype switching
requires collaboration between B cells and helper T cells (TH) that recognize the
same pathogen. .. .. .... ....

Figure 9 Schedule of immunizations (white arrows) and blood sampling (blue and
back arrows). ..........................................................................................

Figure 10 Mean serum anti-J5 E. coli IgM antibody response proﬁle of ﬁve
Holstein steers immunized 12 times with a JS bacterin. .... ...

Figure 11 Mean serum anti-J 5 E. coli IgG1 antibody response proﬁle of ﬁve
Holstein steers immunized 12 times with a J5 bacterin... ..

Figure 12 Mean serum anti-J5 E. coli IgG; antibody response proﬁle of ﬁve
Holstein steers immunized 12 times with a J5

11

12

13

19

20

33

46

47

43

Figure 13 Mean (:1: SEM) isotype-speciﬁc anti-J5 E. coli antibody titers of selected
hyperimmune serum samples from ﬁve Holstein steers...

Figure 14 Mean (i SEM) cross-reactivity of anti-J5 E. coli IgM antibodies
contained 1n selected sera (at 1:50 dilution) from pre-, 3“, 6th, 9‘“, and 12th
immunizations to heterologous Gram-negative bacterial antigens, LPS and Lipid A. .

Figure 15 Mean (:h SEM) cross-reactivity of anti-J5 Eco/1'11 antibodies
contained 1n selected sera (at 1:200 dilution) from pre-, 3rd ,6 ,9‘”, and 12th
immunizations to heterologous Gram-negative bacterial antigens, LPS and Lipid A...

Figure 16 Mean (3: SEM) cross-reactivity of anti-15 E. coli IgG2 antibodies
contained 1n selected sera (at 1:50 dilution) from pre-, 3'd ,6 ,9 ,and 12th
immunizations to heterologous Gram-negative bacteria] antigens, LPS and Lipid A...

Figure 17 Protein A chromatography isolation of IgG1 and IgG2 from pooled serum
ofJSE. colihyperimmunizedsteers... . ... . .. . .

Figure 18 Elution pattern of affinity puriﬁed immunoglobulins from an Afﬁ-Prep
Protein A column using a pH gradient from 8.0 to 2.0 ...................................................

Figure 19 Scheme to illustrate the protocol used for extraction of bacterial proteins
for SDS- PAGE ...............................................................................................................

Figure 20 The ﬂow cytometric assay employed in this study to measure neutrophil
phagocytosis.......

Figure 21 Silver stain of Gram-negative (lane 1-7) and S. aureus (lane 8) bacterial
proteins prepared by sonication and proteins prepared by sonication and protein
precipitation, and of commercial LPS (lane 9) and Lipid A (lane 10) separated by
SDS-PAGE ...................................................................................................................

Figure 22 Western blot analysis of Gram-negative bacterial antigens in the
molecular mass ranges of 34-39 kDa, recognized by IgM, IgG1, and IgG2 antibodies
in various sera from non-immunized and J5 E. coli immunized cattle .........................

Figure 23 Western blot analysis of Gram-negative bacterial antigens in the
molecular mass ranges of 8-10.5 kDa, recognized by IgM, IgG1, and IgG2 antibodies
in various sera from non-immunized and J5 E. coli immunized cattle ..........................

Figure 24 Western blot analysis of Gram-negative bacteria] antigens in the
molecular mass ranges of 8 t010.5 kDa and 34 to 39 kDa, recognized by Protein A

enriched IgG1 and IgG2 antibodies from J5 E. coli hyperimmune serum .......................

49

50

52

54

6O

60

66

70

74

76

77

78

Figure 25 Westem blot analysis of IgG1 (peak 2) and IgG2 (peak 3) fractions
afﬁnity puriﬁed from J5 hyperimmune serum using a protein A column and a pH
gradient technique (as shown in Figure 17) .................................................................... 79

Figure 26 Flow cytometric analysis of bovine blood neutrophils showed that the
cells phagocytosed green ﬂuorescent protein (GFP)—transformed J5 E. coli better
when the bacteria were opsonized with enriched IgG2 versus IgG1. .. 81

Figure 27 Mean neutrophil phagocytosis of GFP-transformed J5 E. coli in the

presence of various opsonins; IgG1 (E ) and IgG2 (I ) (both puriﬁed from J5
hyperimmune serum using Protein A afﬁnity chromatography); bovine serum

albumin (I ); and no opsonin (El ). .... 83

KEY TO SYMBOLS OR ABBREVIATIONS

Ab(s) Antibody(-ies)

Ag(s) antigen(s)

ANOVA analysis of variance

APC antigen-presenting cell

CD cluster of differentiation

cfu colony-forming unit

C11 heavy chain constant

DC(s) Dendritic cell(s)

ELISA enzyme-linked immunosorbent assay

Fab fragment antigen-binding of immunoglobulin carrying one antigen-binding site
F c fragment crystallizable, heavy-chain constant regions of immunoglobulin
F I Fluorescent intensity

GFP green ﬂuorescent protein

Ig immunoglobulin

IL interleukin

INF-y Interferon-gamma

KDO 2-keto-3deoxyoctonate

LPS lipopolysaccharide

MAbs monoclonal antibodies

MHC major histocompatibility complex (i.e., class I and H)

xii

MLP murine lipoprotein

MWCO molecular weight cut-off

“O” oligosaccharide

OD optical density

OMPs outer membrane proteins

OmpA outer membrane protein A

PAL peptidoglycan-associated lipoprotein
PMN Polymorphonuclear neutrophil
PS(s) polysaccharide(s)

SASS saturated-ammonium sulfate sera
SAS-PAGE Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
TD thymus-dependent

TI thymus-independent

T111 T helper cell type 1

TNF tumor necrosis factor (i.e. TNF-a)
udg uridine diphosphate galactose

V11 heavy chain variable

V,1 7. light chain gene locus

xiii

INTRODUCTION

Core antigen technology has been demonstrated to be a potentially useful tool for
enhancing antibody protection against diseases caused by Gram-negative bacteria in
humans and animals. In cattle, commercial bacterins developed from the J5 mutant strain
of E. coli have been used extensively as part of routine coliform mastitis prevention
programs. Nonetheless, mastitis remains one of the most costly diseases in US. animal
agriculture for most dairy enterprises (DeGraves and F etrow, 1993, Reneau, 1993). In
well-managed dairy herds, clinical mastitis caused by pathogens of environmental origin,
such as coliforrns and streptococci, remains a signiﬁcant problem. In particular, colifonn
organisms are the most important cause of severe mastitis that can result in agalactiae,
premature culling, or death Although therapy, including use of antibacterials may
ameliorate the detrimental effects of coliform mastitis, many cases do not respond.
Additionally, therapy often requires milk to be discarded for variable periods following
treatment, and raises food safety and public health concerns regarding potential resistance
to antibacterials (White, 1999).

Therefore, preventive alternatives that can help to decrease the incidence of new
intramammary infections, reduce clinical severity, and reduce the need for therapy, are in
the best interests of dairy cattle, producers, processors, and consumers. However,
success of the J 5 vaccine varies, and cows in herds using this technology still suffer from
clinical coliform mastitis. Additionally, the immune mechanisms induced by J 5 bacterins
in dairy cattle in response to coliform infections are still unclear. The biological

relevance of antibody responses to this vaccine has not been well determined. Thus, it is

time to consider why the vaccines have limited success, and how to improve their
efﬁcacy from development of new applications or technology.

Therefore, we have developed unique resources in the form of anti-colifonn
antisera and enriched speciﬁc antibody from these sera to begin the study of the potential
immune basis of core-antigen vaccine protection against gram-negative bacteria,
including those that cause mastitis in dairy cows. Existing J 5 E. coli core antigen
technology was used to elicit these antibodies in hyperimmunized steers based on
literature that supports the idea that such antibodies may cross-react with a variety of
colifonn bacteria and their endotoxins (Tyler et al., 1992; Tomita et al., 1995).

We expect that results of this study should provide some answers regarding what
may constitute humoral immune protection against mastitis-causing coliforms in J5
vaccinated cattle. The goal of this work was to increase the knowledge of the diversity of
colifonn antigens recognized by isotype-speciﬁc anti-J5 E. coli antibodies, the biological
roles of these antibodies, and the potential to improve humoral immune responses against
coliform bacteria using an alternative vaccination protocol for an existing J5 bacterin
vaccine. With this knowledge, novel vaccination protocols and technologies aimed at

preventing colifonn infections could be developed and improved.

Overall goal

The overall goal of the study was to better understand humoral immune responses
elicited by available J 5 E. coli bacterin vaccine. This will promote future development of
more effective vaccination protocols and vaccine preparations for controlling mastitis-
causing coliforms and other diseases of cattle caused by Gram-negative bacteria. We
believe results of this study have: 1) addressed a signiﬁcant economic problem of the
dairy industry; 2) addressed a signiﬁcant gap in knowledge about humoral immunity
against mastitis-causing coliforms; 3) advanced available vaccination protocols aimed at
preventing coliform infections; and 4) opened the door of discovery to novel vaccine
technologies and other strategies for enhancing control of disease caused by coliforms

and other Gram-negative pathogens.

Main Hypothesis

The hypothesis of the current study was that repeated exposure of the bovine
immune system to Rc mutant J5 E. coli bacterin would result in highly mature antibodies
that cross-react in isotype-speciﬁc fashion with a variety of antigens on heterologous
Gram-negative bacteria for improved bacterial recognition and clearance by bovine

neutrophils.

Questions regarding the hypothesis
1. Does repeated exposure of the bovine immune system to a bacterin of the Rc
mutant J5 E. coli elicit a mature antibody response in hyperimmunized cattle?
Objectives in response to Question 1
1.1 To determine if a mature antibody response occurs during
hyperimmunization of cattle.
1.2 To determine isotypes of serum antibody responses and titers during
hyperimmunization of cattle.
2. Do maturing anti-J5 E. coli antibodies cross-react with a variety of Gram-negative
bacteria and antigens on the bacteria in an isotype-speciﬁc fashion?
Objectives in response to Question 2
2.1 To determine the cross-reactivity of bacterial recognition by maturing
isotype-speciﬁc anti-J5 E. coli antibodies.
2.2 To determine the diversity of bacterial antigen recognition by maturing
isotype-speciﬁc anti-J5 E. coli antibodies.
3. Do IgG2 antibodies from hyperimmune serum promote better phagocytosis of
mastitis-causing coliforms by bovine neutrophils?
Objectives in response to Question 3
3.1 To purify IgG2 antibodies from 15 bacterin-derived hyperimmune serum
for subsequent biological testing.
3.2 To test the effectiveness of the enriched IgG2 antibody to opsonize E. coli

for enhanced neutrophil phagocytosis.

CHAPTER 1

REVIEW OF LITERATURE

15 Escherichia coli bacteria

15 Escherichia coli bacterin is a core antigen technology developed from E. coli
011 1:84 (strain JS), which is a genetically stable rough mutant based on colony
morphology (Tyler et al., 1990b). This technology was convincing because, apparently,
Gram-negative core antigens posses marked chemical, structural, and immunological
homology across bacterial species, genera, and groups (Tyler et al., 1990b). Rough
mutant bacteria are designated by the capital letter R, followed by a subscript, a, b, c, d, c
(Figure 1, Tyler et al., 1990b). Like a variety. of mutant bacterial strains lacking speciﬁc
enzymes for complete somatic chain assembly, the J5 strain which is an Rc mutant, lacks
uridine diphosphate galactose (udg)-4-epimerase, which is required for complete
synthesis of the oligosaccharide side chain portion of the LPS molecule (Figures 1 and
2, Elbien and Heath, 1965; Tyler et al., 1990b). As a result, the bacteria lack O
polysaccharide chains and have exposed core protein and lipid A antigens. These also
make the J S E. coli cell wall structure different from wild type E. coli (smooth Gram-
negative bacteria) (Figures 1 and 2).
Basic structure of Gram-negative bacteria

In wild type Gram-negative bacteria, structures of the cell wall play important
roles as virulence factors, which subsequently cause pathophysiological changes in the
host. The cell wall components of Gram-negative bacteria consist of (1) an inner

cytoplasmic membrane; (2) a peptidoglycan layer; and (3) an outer membrane consisting

Smooth Gram-negative bacteria

 

 

 

 

 

 

 

 

f a
l O—speciﬁc side chain ) Outer R-core II Inner R-oore
f N r ‘r wr W

 

$ 5'? gig $ C.) raglan.
E—CD—C] Cl—E—D—I—
ln

Numbers of Repeating unit ' l
of sugar residues 1

 

 

KP}???

Rough, mutant Gram-negative bacteria

ﬂ D-Galactose O Phosphate
egg N-acetyl-D-glucosamine 0 2-keto-3-deoxy-D-manno-octulosonate
E] D-Glucose V Ethanolamine

L-glycero-D-manno—heptose

Figure 1 General structure of the Lipopolysaccharide (LPS) structures of smooth
and rough strains of Gram-negative bacteria. In smooth Gram-negative bacteria,
LPS consists of the “0” antigen side chain, R-core polysaccharide and Lipid A.
“11” represents number of repeating units of sugar residues in the O—speciﬁc side
chain Rough Gram-negative bacteria lack “O” antigens and consist of Lipid A
and some R-core antigens. These are termed “R” mutants, with the various
mutations given as subscripts (a, b, c, d, or e). Ra indicates the complete core, Rb
to Re indicate the incomplete core structures. (Adapted from Appelmelk et al.,
1993; Tyler et al., 1990)

of two layers: (a) a phospholipid protein layer, and (b) an outer lipopolysaccharide (LPS)
layer, which is incomplete in mutant Gram-negative bacteria including J5 E. coli (Figure
3, Kremer et al., 1990; Tyler et al., 1990b). The inner cytoplasmic membrane contains
enzymes for synthesis of basic chemical units of the peptidoglycan, the phospholipid
protein and the LPS layers (Kremer et al., 1990). The peptidoglycan layer maintains the
bacterial shape and is relatively thinner in Gram-negative bacteria (5-20% of the cell
wall) than in Gram-positive bacteria (as much as 90% of the cell wall) (Bergquist and

Pogosian, 2000; Kremer et al., 1990).
O antigens

 
  
  
 

-07 Core & Lipid A regions

Figure 2 Schematic drawings of Wild type Escherichia coli (leﬂ) showing 0-
polysaccharide chains, and J 5 E. coli, Rc mutant (right) showing exposed core and
Lipid A regions.

 

\4 P111
Capsule { Outer membrane protein A
O-polysaccharide LPS
COIC antigen } (3b)
(3) Outer L1p1d A
Membrane ‘— Phospholipid 3a)

Lipoprotein

 

 

(l) Cytoplasmic { ......................................

Membrane ........
Figure 3 Schematic model of the cell wall of Gram-negative bacteria (adapted from
Kremer et al., 1990). The cell wall components consist of 3 main layers, including
(1) an inner cytoplasmic membrane; (2) a peptidoglycan layer; and (3) an outer
membrane consisting of two layers: a) a phospholipid-protein layer and b) an LPS
layer. Some bacteria also have an outer membrane protein A layer, that readily binds
the Fc region of IgG antibodies, and pili that extend away from the surface of the
bacteria and facilitate adhesion to host cells and tissues.

Bacterial outer membranes are in direct contact with the environment unless the
cells process an additional capsular polysaccharide, or slime layer, which is found in a
number of E. coli strains (Bergquist and Pogosian, 2000; Kremer et al., 1990). However,
the biochemical structure of capsular antigens and their role in the pathogenesis in bovine
mastitis are not clear (Sanchez-Carlos et al., 1984). Capsules may inhibit activation of
the alternative pathway of the complement system (without involvement of
immunoglobulins) (Stevens et al., 1978), thus decreasing susceptibility to lysis and
phagocytosis. It is well established that strains with a heat-labile capsule resist serum
bactericidal activity more than strains without a capsule (Hill, 1981). Opsonization of
encapsulated strains of E. coli with antibodies is also more difﬁcult (Hill et al., 1983b).
Protein ﬁlaments termed ﬁmbriae or pili, protrude from the outer membrane and through
the capsule, which generally confers adhesive properties of bacteria (reviewed by Kremer
et al., 1990). However, E. coli pili do not seem to play a signiﬁcant role in the
pathogenesis of colifonn mastitis, as opposed to mastitis caused by Staphylococcus
aureus and S. agalactiae (reviewed by Kremer et al., 1990). There is another surface
structure of E. coli called curli, which may be involved in adherence during wound
colonization or colonization of ﬁbronectin—coated surfaces (Olsen et al., 1989), and this
may be problematic in cows with damaged teats.

Lipopolysaccharide, also termed endotoxin, is composed of a variable
oligosaccharide region (0 or somatic antigens) covalently linked to a highly conserved
core polysaccharide [N-acetylglucosamine, 2-keto-3deoxyoctonate (KDO), heptose and
glucose residues] and lipid A (LA) regions (Figures 2 and 3, Elbien and Heath 1965 ;

Jann and Jann, 1987; Raetz, 1993). With combinations of hexose molecule numbers,

composition, and linkage, O-speciﬁc antigens can vary widely and speciﬁcally for each
speciﬁc Gram-negative organism (Elbien and Heath 1965; Raetz, 1993). LPS is a key
virulence factor released from the Gram-negative bacteria upon cell growth and death.
Toxicity of LPS varies, depending on Lipid A~associated proteins and the organism, and
is dose-dependent (Frost, 1984; Giri, et al., 1984; Lohuis et al., 1988b). Generally, as
dosage increases, latency time decreases, the peak effect becomes more pronounced, and
the duration of the effect protracted (Lohuis et al., 1988b).

Natural antibodies elicited against 0 polysaccharides do well to protect animals
against homologous Gram-negative bacterial strains but are not useful in protection
against heterologous strains. This is a problem for protection against coliform mastitis
because the causative pathogens represent a diverse population of species and antigens.
Because the lipid A component of LPS is so well conserved across Gram-negative
bacteria, antibodies elicited against this core antigen logically should confer cross-
reactive protection against most Gram-negative bacteria. Indeed, this is the basis for
vaccinating cows with mutant forms of Gram-negative coliforms for protection against
mastitis. One such vaccine strain commonly employed is J 5 E. coli.

Core antigens elicit protective immunity against Gram-negative bacterm1 diseases

Core antigens (i.e. J 5 bacterin) have been used as immunogens (active
immunization) in both humans and animals to enhance immune resistance against Gram-
negative bacteria that are responsible for many diseases, including neonatal colifonn
septicemia, salmonellosis, and colifonn mastitis. Evidence suggests that this bacterin
also elicits cross-reactive serum antibodies that protect against sepsis caused by

heterologous Gram-negative bacterial infections, for example E. coli, Klebsiella, spp.,

and Pseudomonas aeruginosa, (Bhattacharjee et al., 1994; Ziegler et al., 1973, 1975,
1982; Baumgartner et al., 1985; Calandra et al., 1988;). The relative risk of developing
septic shock was about 6 times greater in untreated surgical patients who became infected
with E. coli, Klebsiella, P. aeruginosa, Enterobacter sp. or Proteus sp., than in surgical
patients who received plasma taken from donors immunized with the endotoxin
glycolipid core of J5 E. coli (Baumgartner et al., 1985). As in the Gram-negative
infection model for humans, antibodies elicited by heat-killed J5 E. coli in experimentally
Haemophilus-infected mice provided partial protection, with 80% survival in treated
mice as compared to 28% survival in untreated mice (Marks et al., 1982). IgG and J5
LPS-speciﬁc IgG antibodies afﬁnity puriﬁed from hyperimmune serum of rabbits
immunized with heat-killed J5 E. coli signiﬁcantly protected neutropenic rats against
lethal consequences of Pseudomonas aeruginosa infection as compared to rats not
administered antibody treatment (Bhattacharjee et al., 1994). While these antibodies
elicited by J5 E. coli vaccination provided some cross-protective activity against existing
Gram-negative infections, they did not prevent new infections with Gram-negative
bacteria (Baumgartner et al., 1985). This suggests that anti-J5 antibodies have a short
half-life in vivo. In cattle, success of the J5 vaccine varies, and many cows in herds using
this technology still suffer from clinical diseases (Hill, 1991; Tomita, et al., 2000).
Additionally, the immune mechanisms and the biological relevance of the antibody
response induced by 15 vaccination in dairy cattle are still unclear.
Bovine antibody response and mastitis

As shown in Figure 4, bovine antibodies (Abs) are composed of two heavy (dark

grey) and two light (white) chains. These chains combine to form two antigen binding

10

sites (Fab) and one Fc region of the full monomeric antibody molecule (Figure 4).
Antibodies, which are the speciﬁc product of the adaptive immune response (humoral
immunity), are found in the ﬂuid compartment of blood and inextracellular ﬂuid
(Janeway et al., 2001). Bovine antibodies exist as six isotypes called immunoglobulin
(lg) M (IgM), IgG1, IgG2, IgG3, IgA, and IgE, of which relatively little is known of IgG3
and IgE (reviewed by Burton et al., 2002, Figure 4). Unlike humans, cattle lack IgD
(Nassens, 1997). Molecular weight characteristics of bovine IgM, IgG1, IgG2, and IgA

are given in Table 1.

Figure 4 Monomeric antibody
molecules are composed of two heavy
(dark gray) and two light (white)
‘/regions\ chains. These chains combine to form
two Fab and one Fc region of the full
antibody molecule. The Fab are
encoded by VDJ gene segments in
immunoglobulin genes and undergo
editing during immune responses for
"3M (1‘) improved afﬁnity of binding with
Fc region 189101) antigen. The F c region determines an
(biological < ‘8G2 (72) antibody isotype and is responsible for
function) IgG-103) the biological functions of antibodies.
18“ (a) Isotype switching from IgM to IgG1
L13": (8) and IgG2 occurs in response to repeated
antigen exposure and increases the
biological repertoire of antibodies for
more effective clearance of pathogens
(ﬁgure from Burton et al., 2002).

   
   

 

Bovine IgA antibodies are usually dimerized (Figure 5) and are present in very
low concentrations in bovine serum and milk (Table 2; reviewed by Butler, 1981; 1983;
Tizard 1996). This is in contrast to non-ruminant species for which IgA constitutes the

major 11qu immunoglobulin, and reﬂects the fact that the bovine mammary gland is part

11

of the peripheral immune system rather than the mucosal immune system (Harp et al.,
1988; Kehrli and Harp, 2001, reviewed by Burton et al., 2002).

Bovine IgM occurs predominantly as a pentamer (Figure 5) and in relatively
modest concentrations in serum and milk (Table 2; reviewed by Butler, 1981; 1983;
Tizard, 1996). However, because of its ability to bind with 5 to 10 antigens and to ﬁx
complement, IgM is thought to be a highly effective antibody for neutralizing toxins
(Figure 6a), complement ﬁxation, and also may opsonize bacteria for efﬁcient clearance
via tissue macrophages and blood neutrophils (reviewed by Kremer et al., 1990).
However, given that concentrations of IgM in milk are low, the contribution of this
antibody class to clearance of mastitis pathogens may be relatively small. Instead,
antibody protection against inﬂammatory pathogens may depend on IgG1 and IgG2

(reviewed by Burton et al., 2002; Burton and Erskine, 2003).

save

IgM IgG IgA
Figure 5 Antibody molecules are synthesized as monomers, such as is shown for

IgG (middle), but IgM and IgA molecules are usually found in serum as multimers
[either as pentameric IgM (left) or dimeric IgA (right)].

Monomeric IgG1 and IgG2 antibodies (Figure 5) are much smaller than IgA or
IgM antibodies (Table 1) and are found in relatively high and almost equivalent
concentrations in bovine serum (Table 2; Tizard 1996; Butler, 1981; 1983). These, along

with IgM, are the main classes of antibodies of the peripheral immune system. IgG

antibodies predominate over all other isotypes in serum, accounting for approximately
90% of the total serum immunoglobulin (Table 2, McGuire and Musoke, 1981).
However, the IgG1 to IgG2 ratio in non-mastitic bovine milk is 7:1 as compared with an
IgG1 to IgG2 ratio of 0.7:] in serum (Guidry et al., 1980b). This is due to selective and
active transport of IgG1 into milk of healthy mammary glands (Butler, 1998). However,
after intramammary bacterial challenge, the IgG1:IgG2 ratio in milk reverses to 1:16,
leading some to hypothesize that IgG2 must be the most important antibody isotype in

defense against intramammary infections in dairy cows (Burton and Erskine, 2003).

(b) Opsonizing antbody

 

 

   

 

' MN

Figure 6 Antibodies effectively block pathogens from interacting with host cells and may
have varying degrees of toxin neutralizauiapacity(a) or opsonizatiorcapacity(b) In

cows, bacteriaopsonizedwith IgG antibody leads to rapid receptor-mediaphagocytosis

by neutrophilsresulting in pathogen degradation through activation of the respiratory
burst (generating toxic ROS) and throughagosomelysosomefusion(c). (adapted from
Burton et al., 2002).

13

Table 1 Molecular size characteristics of the major bovine antibody isotypes. N/A = not
available. (adapted from Butler, 1981)

 

 

 

 

 

 

 

 

 

Molecular Weight. (kDa)

Antibody isotype Intact molecule Heavy chain Light chain
IgM 1,000 61.8-76 22.5-26
IgG1 146-163 56-59 22.5-26
IgG2 146-150 54-59 22.5-26
IgA: 22.5-26

in colostrum 385-430 61 N/A
in respiratory tract 430 & 642 62.7 N/A

 

 

 

 

Table 2 Normal serum and milk (> 2 weeks postpartum) concentrations of the major
antibody classes (isotypes) in dairy cows (from Burton et al., 2002). -

 

 

 

 

 

 

 

 

 

 

Antibody Serum levels Milk levels Main biological
isotype (mg/ml) (mg/ml) functions (via Fc region)

IgM 3.05 0.86 Agglutination,
(6.0 — 4.30) (0.37 — 1.5) complement ﬁxation,

opsonization
blocking/neutralization
IgG1 11.20 0.58 Complement ﬁxation,
(6.00 .— 15.10) (0.33 - 1.20) ' blocking/neutralization

IgG2 9.20 0.055 Opsonization,
(5.00 — 13.50) (0.037 — 0.06) blocking/neutralization

IgA 0.37 0.08 Agglutination,
(0.06 — 1.00) (0.05 — 0.110) blocking/neutralization

 

l4

 

Substantiating this hypothesis, IgG2 antibodies are well known opsonins for
bovine neutrophils, the key cellular defense against intramammary bacteria (Burton and
Erskine, 2003; Figure 6b, 6c). Freshly migrated neutrophils express high levels of
surface Fc receptors that are speciﬁc for pathogen-bound IgG2 antibodies (reviewed by
Burton and Erskine, 2003). Thus, rapid leakage of blood-derived IgG2 antibodies into
milk shortly after infection must play a signiﬁcant role in subsequent neutrophil clearance
of the intramammary pathogens (reviewed by Burton et al., 2002; Burton and Erskine,
2003 ).

Passive transudation of IgG2 is highly associated with neutrophil recruitment into
the gland and signiﬁcantly increases the opsonic activity of milk (reviewed by Burton et
al., 2002; Burton and Erskine, 2003). Interestingly, the selective transfer of blood IgG1
into milk may be suppressed during the acute phase of the mammary inﬂammatory
response (Lascelles, 1979) possibly explaining why researchers observe relatively greater
increases in milk IgG2 compared to IgG1 in 4 to 12 hours post infection, to the extent that
the normal IgG1/IgG2 ratio can become reversed (reviewed by Burton and Erskine, 2003).
It has also been shown that return to active IgG1 transport post acute inﬂammation is
accompanied by the formation of immune complexes in milk, which appear to inhibit
neutrophil phagocytosis (Targowski and Clucinski, 1985), possibly by blocking IgG2 Fc
receptors and competing with IgG2 for binding'sites on pathogens (Targowski and
Klucinski, 1985; Tao et al., 1995). Virtually nothing is known about if or how anti-J5 E.
coli IgG2 antibodies mediate defense against mastitis-causing coliforms (reviewed by

Burton et al., 2002; Burton and Erskine, 2003)

15

IgG1 is able to ﬁx complement by the classical pathway (C3b molecules) through
its F c portion (McGuire and Musoke, 1981). Together with serum-derived complement,
IgG1 impedes the grth of serum-sensitive colifonn bacteria (Carroll, 1974; Carroll et
al., 1969). AlthoughO IgG2 had been thought as a non-complement—ﬁxing IgG subclass,
studies demonstrated bovine IgG2 ﬁxed bovine complement in vitro, but was much less
effective than IgG1 (reviewed by Burton and Erskine, 2003). The ability of IgG1 to ﬁx
complement (Butler, 1981; 1983) may be important to neutrophil phagocytosis in non-
mammary tissues but its role in mammary defense against mastitis-causing coliforms is
not at all clear. Despite the fact that mammary neutrophils have been demonstrated to
bind more complement than blood neutrophils (DiCarlo et al., 1996), complement
concentrations in normal milk are extremely low (reviewed by Burton et al., 2002;
Burton and Erskine, 2003). Endotoxin-induced mammary inﬂammation has been shown
to increase the levels of complement detected in bovine milk (Rainard et al., 1998), but
milk itself has also been shown to mask the inﬂammatory activity of activated
complement (Colditz and Maas, 1987). Thus, enhanced IgG1 levels in milk, subsequent
to the active inﬂammatory response to infection, have the potential to actively reduce the
effectiveness of diffused IgG2 antibodies in neutrophil-mediated bacterial clearance.
Both IgG1 and I gG2 may be effective blocking or neutralizing antibodies (Table 2).
Maturation of bovine antibody

Generally, maturation of the antibody response is the process of development of
highly speciﬁc antibodies by T-cell-dependent antigen-activated B cells. This process
includes afﬁnity maturation of the Fab regions by editing of the variable (VDJ) region of

the heavy chain gene segments, and switching of gene segments in the constant (Fe)

16

region (Figure 7, Bonhomme et al., 2000; Butler, 1998; Kaattari et al., 2002). This
process is considered an integral part of adaptive immune responses that usually occurs in
the germinal centers of secondary lymphoid organs including the spleen and lymph nodes
(Bonhomme et al., 2000; Butler, 1998; Kaattari et al., 2002; Liu et al., 1992). The
maturation process ultimately improves afﬁnity of antibodies produced by some B cell
clones for their corresponding antigens (Bonhomme et al., 2000; Butler, 1998; Kaattari et
al., 2002). Memory B cells stimulated by the same antigens (Figure 8) produce mature
antibodies with higher afﬁnity as compared to nai've B cells. This continued or repeated
stimulation of memory B cells also leads to isotype switching, mediated in large part by
T—cell derived cytokines (e.g. IL-4 for IgG1 and IgE, IFN-( ﬁr IgG2) present in the
responding germinal center of lymph nodes draining the affected tissue.

Studies in mice demonstrated that antibody maturation occurs following multiple
exposures with the same antigen (F urukawa et al., 1999). In human antibodies, affinity
maturation rapidly generates multiple variants to a broader range of epitopes that react
with tumor associated antigens, and that display greater than 100-fold higher afﬁnity and
distinct speciﬁcities than the ﬁrst antibodies produced (Pancook et al., 2001). It was also
determined that trout are capable of generating new, higher aﬂ'mity antibodies relatively
late in the antibody response, which then become dominant (Kaattari et al, 2002).

In cattle, maturation of the antibody response is not completely understood but
appears to share marked similarity to that of sheep (Aitken et al., 1999; Lei Parng et al.,
1996). For example, antigen-stimulated B cells rely on somatic hypermutation and gene
conversion to generate diversity in the Fab-encoding VDJ (Heavy chain) and VJ (light

chain) regions of antibody genes in ruminants (Aitken et al., 1999; Butler, 1998; Lei

l7

Pamg et al., 1996). Somatic hypermutation occurs as point mutations accumulate when
cells rapidly proliferate. Gene conversion is a nonreciprocal transfer of genetic
information from one locus to another. Indeed, Lei Pamg et al., (1996) have proposed
that the pseudogenes occurring in the 7» light chain gene locus (VQ, which are the
predominant light chains expressed in bovine antibodies, may serve to diversify the light
chain repertoire through gene conversion. Because bovine immunoglobulin genes
possess an extremely limited number of heavy chain variable (V3) region gene segments
and gene families, the generation of diversity in heavy chain gene segments relies mostly
on somatic rearrangement. Consequently, cattle are unable to establish a diverse primary
antibody repertoire from the process of heavy chain gene rearrangements, a paradigm that
departs signiﬁcantly from the murine/human system (Aitken et al., 1999; Butler, 1998).
Thus, naive B cell clones in cattle may express very similar or identical VDJ gene
combinations (and thus Fab regions in antibody molecules). Subsequently, antigen-
activated and proliferating B cells undergo variable-region somatic hypermutation and
gene conversion in the spleen and lymph nodes under the inﬂuence of cognate B cell-T
helper cell recognition of speciﬁc antigens. In this case, the antigen-speciﬁc T cell-
derived cytokines present in the germinal center will determine if and what isotype
switching occurs (Estes, 1996; Butler, 1998).

In summary, maturation of the antibody response involves editing of the variable
(VDJ) region of heavy chain gene segments, and constant region gene switching (isotype
switching, Figure 7). VDJ region editing modiﬁes the sequences of Fab regions of
antibody molecules, ultimately improving antibody affinity produced by some B cell

clones for speciﬁc antigens (Figure 7, Butler, 1998). Isotype switching enables

18

antibodies with the same or afﬁnity matured Fab regions to use different heavy chain
constant region genes. This permits alteration of the biological functions (F c region,
Figure 4) of the antibodies to given antigens, thus expanding the capabilities of the
humoral immune response (Estes, 1996). Thus, maturation of the antibody response
should improve its overall effectiveness in clearing infections and toxins that cause
clinical disease, including coliform mastitis. It is important to understand if the antibody
response to J5 E. coli vaccination is capable of maturation in vaccinated dairy cattle.
This knowledge would have signiﬁcant implications for the efficacy of this vaccine to
protect cows against a large variety of gram-negative bacteria that cause clinical mastitis

as well as other diseases.

(a) Immunoglobulin gene
Fab region Fc region DNA segments

55

Heavy
did:

ll!

     

 

(b) Antibody response
Serum ‘ w-
Antibody
Level /
l°'IgM / 2"'lgGl > 2°"IgG2

 

A '4 A
Figure 7 Repeated exposure of the immune system to the same antigens on bacteria promotes
isotype switching to occur in B cells. Panel (a) shows switching of Fc region gene segments
in bovine immunoglobulin genes after the immune sytem has been exposed to one (1°), two
(2°), or more (> 2°), doses of bacterin (arrows). Panel (b) shows the corresponding changes
in serum isotype-speciﬁc antibody levels following the ﬁrst (1°), second (2°), or more (> 2°)
doses of bacterin (arrows).

Figure 8 B cell activation with subsequent immunoglobulin isotype switching
requires collaboration between B cells and helper T cells (TH) that recognize the
same pathogen. In panel (a), membrane-bound antibody on the surface of a speciﬁc
B cell recognizes a pathogen and binds to it. The B cell then intemalizes the
pathogen and processes it into small pieces, some of which are peptides that bind to
the antigen binding groove of MHC II molecules for surface presentation to nearby
CD4+ TH cells (a and b). When the TH cell recognizes the peptide presented in MHC
II molecules by the B cell, the TH cell responds by secreting a variety of cytokines
(a) that cause clonal expansion of the B cell (c), isotype switching from IgM to IgG
antibody subclasses (d), and secretion of large amounts of antigen and isotype
speciﬁc antibodies from differentiated plasma cells (e). Some B cells so activated
will also differentiate into memory cells that can respond rapidly to the same antigen
upon subsequent exposures (t).

 

 

 

 

@ Plum-cell(e) Memorynun (r)
wt‘t IaG

 

Coliform mastitis

Colifonn mastitis is one of the most costly diseases in animal agriculture in the
US, with costs from decreased milk production, discarded milk, therapeutic costs,
mortality, and culling for dairy enterprises (DeGraves and F etrow, 1993; Reneau, 1993).
Studies in California, Ohio, and Pennsylvania have determined that the incidence of

clinical mastitis in herds with low somatic cell counts averages 22 to 50 cases per 100

20

cows milking per year (Erskine et al., 1988; Hogan et al., 1989a; Gonzalez et al., 1990).
In the Pennsylvania study, the proportion of clinical mastitis cases attributable to
coliform bacteria (lactose-fennenting, Gram-negative bacilli of the family
Enterobacteriaceae, including the genera Escherichia, Klebsiella and Enterobacter) was
signiﬁcantly higher in low somatic cell count (LSCC, 5 150,000 cells/ml) herds (43.5 2:
3.5%, n = 12) than in high somatic cell count (HSCC, 2 700,000 cells/ml) herds (8.0 2
3.4%, n = 6) (Erskine et al., 1988). In the other studies, coliforms were isolated from 30
to 46% of the clinical cases of which E. coli, Klebsiella spp., and Enterobacter spp.
accounted for 14.6%, 2.6%, and 2.8% of the cases, respectively (Bartlett et. al., 1992;
Gonzales et al., 1990; Hogan et al., 1989a; Smith et al., 1985). '

Furthermore, coliforms are the predominant cause of both clinical mastitis and
severe, life threatening, mastitis in well-managed dairy herds (Anderson et al., 1982;
Erskine et al., 1988; Hogan et al., 1989a; Gonzalez et al., 1990; Erskine et al., 2002).
Studies have isolated coliforms as the causative agent in 58.9% (56/95) and 55% (56/ 104)
of severe clinical cases (Erskine et al., 2002). In addition, almost 30% of clinical
coliform cases result in severe mastitis with abnormal milk, swelling of quarter, and
systemic signs (review by Eberhart et al., 1979; Hogan et al., 1989a). However, it is
suspected that when no bacteria are isolated on culture from milk samples collected from
clinical mastitis cases, that the causative agent is often a colifonn infection already
eliminated by the cow’s immune defenses (Bartlett et al., 1993).

Severe colifonn mastitis may result in a total cessation of the lactation or in death.
In an Illinois study, approximately 14% (6/42) of acute colifonn cases resulted in death

of the cow, as opposed to no death among 59 acute cases caused by Gram-positive

2]

bacteria (Anderson et al., 1982). Similarly, 25% (14/56) of severe coliform mastitis cases
died or were culled as compared to no loss cows in severe clinical mastitis caused by
other pathogens (Erskine, 2002). Unlike S. aureus, coliforms do not colonize the teat
canal and are minimally invasive of the mammary tissue (Frost and Wanasinghe, 1976).
Thus, chronic intramammary coliform infections are uncommon (Eberhart, 1977; Jain,
1979). Typically, only one-fourth of colifonn infections exceeded 30 days in duration
(Smith, 1983).

Increased incidence of colifonn mastitis is associated with the early postpartum
period, bedding materials, warm humid weather, and increasing parity (Bramley, 1985;
Erskine et al., 1988; Hogan et al., 1989a; 1989b; 'Smith et al., 1985). Slow migration of
blood polymorphonuclear neutrophil (PMN) into the mammary gland and impaired
oxidative burst activity are considered the important immune risk factors for severe
colifonn mastitis during the early postpartum period (reviewed by Dosogne et. al., 2002).
The susceptibility of individual cows to severe coliform mastitis appears to be associated
with the degree of neutropenia or impairment of blood PMN function (Dosogne et. al.,
2002). Lymphocyte function including proliferation and antibody response has also been
found to be impaired during early lactation, but a relation with the severity of coliform
mastitis has not been clearly established (reviewed by Dosogne et. al., 2002).
Inﬂammatory and systemic response to intramammary invading pathogens

The teat streak canal is a physical barrier that normally prevents intramammary
infection in cows. However, when breached, bacteria gain easy entry into the duct
system and mammary epithelium. This initiates a nonspeciﬁc innate immune response

resulting in inﬂammation. Inﬂammation is the leakage of serum proteins (including

22

antibodies) from blood into the infected tissue, with subsequent massive recruitment of
blood-derived neutrophils. Rapid inﬂammatory responses must occur if the neutrophils
are to successfully clear the infecting pathogen before it proliferates and causes tissue
damage and clinical disease. However, the inﬂammatory response must also be
controlled to avoid excess tissue damage from reactive oxygen species and enzymes
elaborated by inﬁltrating neutrophils. Antibodies are key in facilitating rapid and
controlled inﬂammation.

Inﬂammation at the site of infection is initiated by the response of resident
macrophages to common components of bacterial cell walls, such as LPS or lipoproteins
of coliforms. This induces rapid production and secretion of various macrophage
cytokines and tissue chemokines that cause effusion of blood components through the
capillaries, and recruit large number of neutrophils to the site. In dairy cattle, neutrophils
play a pivotal role as the ﬁrst line of cellular immune defense against intramammary
pathogens. Blood-derived neutrophils gain access to the mammary parenchyma and milk
through the process of migration and chemotaxis. The destruction of infecting bacteria
by phagocytosis uses both oxygen-dependent and oxygen-independent killing
mechanisms (Figure 6c; Burton and Erskine, 2003; Burvenich et al., 1994; Hill et al.,
1978). This mammary inﬂammatory response, like all inﬂammatory responses, results in
ﬁve hallmark clinical signs: swelling, redness, pain, heat, and abnormal milk (Burton and
Erskine, 2003; Lohuis et al., 1988b; Tizard, 1996).

Systemic responses to intramammary infection are also noted, especially during
severe colifonn mastitis. These include fever, increased concentrations of serum cortisol,

ﬁbrinogen, complement, haptoglobin, and ceruloplasmin, transient decreases in serum Fe

23

and Zn, and mobilization of leukocytes out of the blood (neutrophils) and lymph nodes
(mononuclear leukocytes) into the mammary gland. These systemic changes are in large
part caused by cytokines, especially IL-1, IL-6, and TNF-a, that escape from the
inﬂamed mammary gland (Bishop et al., 1976), and are observed with both natural
coliform infections and when animals are treated with intramammary doses of LPS
(Conner et al., 1986; Erskine et al., 1989; Erskine et al., 1993; Jackson et al., 1990;
Lohuis et al., 1990; Shuster et al., 1992). Released eicosanoids following LPS exposure
are also responsible for many pathophysiologic changes, including shock, pulmonary
hypertension, abortion, and diarrhea (Tyler and Cullor, 2002).
Coliform mastitis versus core antigen technology

The dairy industry has adopted core-antigen vaccine technology to promote the
prevention of colifonn mastitis. Currently, commercially available core-antigen vaccines
for mastitis include J5 BACTERINTM (Pharmacia Animal Health), Mastiguardm, J -
VACTM (Merial Ltd), Endovac-Bovi® (Re-mutant Salmonella typhimurium bacterin-
toxoid. IMMVAC Inc), and Master Guard® J5 (Agri Laboratories). Typically, these
vaccines are administered subcutaneously into dairy cows one to three times prior to the
early lactation period, the time of the lactation cycle that is believed to have the greatest
risk for new intramammary infections (Clark, & VanRoekel, 1994; Gonzalez et al., 1989;
Hogan et al., 1992a; 1992b; 1995, 1999). However, disparity exists for both the
recommended vaccination protocols and the antibody responses elicited by various
vaccines and vaccination protocols.

Historically, the ﬁrst 60 days of lactation has been the period of peak occurrence

of clinical coliform mastitis. Evidence from early research trials indicated that a three-

24

immunization protocol can signiﬁcantly reduce the incidence, severity and duration of
clinical colifonn mastitis (Clark & van Roekel, 1994; Gonzalez et a1, 1989; Hogan et
al,1992a; 1992b; 1995). Studies in two commercial California dairy herds reported that
the incidence rate of Gram-negative mastitis was lower in cows vaccinated by 3 doses of
whole cell bacterin of J 5 E. coli plus F reund's incomplete adjuvant (2.6%) than in
unvaccinated cows (12.8%) during the ﬁrst three months of lactation (Gonzalez et al.,
1989). Only 20% of cows vaccinated with this J 5 bacterin (prepared from the heat-killed
Escherichia coli J5 Rc mutant at a 5 ml dose emulsiﬁed with 1 ml of Freund's incomplete
adjuvant) had IMI caused by Gram-negative bacteria as compared to % of non-
vaccinated controls (Hogan, et al., 1992a). As compared to placebo, cows vaccinated by
commercial J 5 bacterin at drying off, 30 days later, and 48 hours after calving and
challenged by intramammary infusion of E. coli 727, had reduced bacterial numbers in
milk from infected quarters and reduced local signs and shortened duratiOn of mild
clinical mastitis compared to non-vaccinated controls (Hogan et al., 1995). However,
there was no effect of vaccination on systemic signs, dry matter intake, and somatic cell
count in this study (Hogan et al., 1995). Two commercially available vaccines 15
Bacterin and J-VAC were compared in another study (Tomita et al., 2000). Here, serum
IgG, IgG1 and IgG; antibody titers to E. coli J5 whole cell antigens were higher in
vaccinated cows around 2-6 weeks after calving than in unvaccinated control cows. Also,
both serum and milk IgG1 and IgG; antibody titers elicited by JS Bacterin were
consistently higher than those elicited by J -VAC (Tomita et al., 2000). Nonetheless,
neither vaccine prevented colifonn mastitis, though all vaccinated cows tended to have a

reduced severity of disease compared to unvaccinated cows (Tomita et al., 2000). It is

25

possible that the increased antibody titers achieved by vaccination were not high enough
or mature enough to provide effective humoral immune protection against the infecting
bacteria.

Although much of the attention of J 5 vaccination protocols has focused on the
prevention of colifonn mastitis in the early part of lactation, a ﬁeld trial of six Michigan
dairies reported that the highest incidence of severe clinical coliform mastitis was
between the 3rd and 5th months of lactation, even when herd J5 vaccination programs
were employed (Erskine, 2002). In other studies, cows immunized with J5 according to
these protocols still had clinical coliform mastitis and systemic disease, especially in the
period between calving and 180 days in milk (Hogan et al., 1992a; 1999). As was
observed in the study of Tomita, et a1. (2000), this suggests that the anti-colifonn
antibody response elicited by dry period and peripartum J5 vaccinations may beshort-
lived and (or) ineffective in today’s high producing cows. A potential practical limitation
of active immunization using J5 bacterin may be the shorter half-life of core-antigen
speciﬁc IgG antibodies (7.6 days) in the serum of cattle as compared to total serum IgG
(22.7 days; Douglas et al., 1989).

Low titers (< 1:240) among cows of naturally occurring serum anti-J5 E. coli
IgG1 antibodies have been shown to be negatively correlated with mastitis incidence,
resulting in 5.33 times higher risk of disease than cows with titers > 1:240 (Tyler et al.,
1988). Similarly, titers in response to J-5 immunization were negatively correlated with
severity of clinical outcome following experimental challenge using heterologous E. coli
(Hogan et al., 1992a). In another experimental challenge trial, vaccination using J5

bacterin elaborated signiﬁcantly higher antibody responses in serum and milk, higher

26

serum IgG titers, and less depression of milk production and dry matter intake than was
observed in non-vaccinated cows (Hogan et al., 1992b). However, differences in
antibody titers between vaccinated and control cows were not impressive in that study
despite the statistically signiﬁcant differences highlighted. This could be why J5
vaccines are not effective in preventing colifonn mastitis in some cows.

In an attempt to improve antibody responses to J5 vaccination, researchers have
considered different routes of administration. For example, Hogan et a1. (1997)
demonstrated that intramammary infusion of J5 enhanced serum and milk IgG and IgM
antibody titers as compared to the systemic route of administration. Total milk IgG titers
in cows administered intramammary vaccination were higher (13.1) than in cows
receiving subcutaneously administered bacterin (10.1) at calving. This may have been
due to the combination of active transfer of IgG1 from serum and local production of
IgG; from B cells present in the mammary gland parenchyma. This was substantiated by
the ﬁnding that serum IgG titers were higher in intramammarily vaccinated cows than in
cows vaccinated subcutaneously (Hogan et al., 1997). However, antibodies elicited by J5
intramammary infusion may not have been high enough or of the protective isotypes
because they did not prevent infection or reduce the severity of coliform mastitis (Smith
et al., 1999). Also, rectal temperatures were increased in the intramammarily vaccinated
cows but not in the subcutaneously vaccinated cows (Hogan et al., 1997). Therefore, the
intramammary route of JS vaccination is not more efficacious in preventing or reducing
signs of clinical colifonn mastitis than the subcutaneous route.

The induction of cross-reactive antibodies is most likely responsible for

successful applications of J 5 vaccine programs when these do work well. However, there

27

is a paucity of research available that describes the biological activities of anti-J5 E. coli
antibodies, including isotype-speciﬁcity of pathogen and antigen recognition, and
biological functions of each responding antibody isotype. The purpose of the current
study was to begin to characterize the bioactivities of bovine ant-colifonn antibodies
elicited by J5 vaccination. The hypothesis of the study was that repeated exposure of the
bovine immune system to a bacterin of Re mutant J5 E. coli would result in highly
mature serum antibody responses, producing in particular a strong IgG2 response that
would cross react with increasingly diverse antigens on a variety of heterologous Gram-
negative bacteria. This would result in enhanced bacterial opsonization for effective

phagocytic uptake by bovine neutrophils.

28

CHAPTER 2

ISOTYPE SPECIFIC SERUM ANTI-COLIFORM ANTIBODIES
CROSS-REACT WITH GRAM-NEGATIVE. BACTERIA

Anantachai Chaiyotwittayakun‘, Jeanne L. Burtonb, .
Kadia Kizilkayab, Fernando F. Cardosob, and Ronald J. Erskinea'

Departments of “Large Animal Clinical Sciences and ”Animal Sciences,
Michigan State University, East Lansing 48824 USA

Abstract

This study was conducted to determine response proﬁles, titers, and cross
reactivity for Gram-negative bacteria of isotype speciﬁc anti-.15 E. coli antibodies elicited
by J5 hyperimmunization in healthy Holstein cattle. Five steers were used.
Immunizations were administered at time 0, 30 days later, and every 2 weeks for 10
subsequent immunizations. Blood samples were collected pre-immunization and on days
4, 6, 11, and 13 post each immunization, harvested for sera, and assayed for relative
contents of anti-J5 E. coli IgM, IgG1, and IgG2 antibodies by ELISA. Anti-J 5 E. coli
antibody response proﬁles were recorded as optical densities of appropriately diluted
sera. Selected sera from pre-immunization and 6 days post-3“, 6‘“, 9th, and 12th
immunization were also assayed for isotype speciﬁc end-point titers. Those selected sera
were then used to determine cross-reactivity against a variety of Gram-negative bacteria,
most of these isolated from clinical mastitis cases. Repeated measurement analysis
showed that immunization number and day post immunization signiﬁcantly (P < 0.0001)

inﬂuenced anti-15 E. coli antibody responses for all three isotypes. Two immunizations

 

Corresponding author. Dr. Ronald J. Erskine, Large Animal Clinical Science, Michigan State University,
East Lansing, MI 48824-1314, Tel. 517-332-2484, Fax. 517-432-1042, e-mail address:
erskine@cvm.msu.edu

29

were required to increase (P s 0.05) mean serum anti-J5 E. coli IgM and IgGlantibody
above pre-immunization levels. However, ﬁve immunizations were required to detect
signiﬁcant IgG1 to IgG2 isotype switching, and to increase anti-J5 E. coli IgG; antibody
above pre-immunization levels (P s 0.02). Anti-J 5 antibody titers of all isotypes tended
to increase through 12 immunizations, and were signiﬁcantly higher than pre-
immunization levels after 6 immunizations (P < 0.005). Cross-reactivity of isotype
speciﬁc anti-J5 antibodies to all Gram-negative bacteria was increased (P < 0.05 ) in
hyperimmunized steers. Anti-J5 IgM and IgG2 antibodies tended to cross-react with
Serratia spp., E. coli 487 and K. pneumoniae more than with other bacteria Anti-J5 IgG1
antibodies tended to cross-react with E. coli 487 and Pseudomonas spp. more than with
other bacteria. However, no cross-reactivity of anti-J5 antibodies with either LPS or
Lipid A was observed, which was similar to the result for S. aureus. Results of our study
suggest that multiple dose immunizations with J5 E. coli bacterin may be required to
elicit high levels of serum IgG and IgG2 antibodies that strongly cross-react with mastitis
causing Gram-negative bacteria. Our results also suggest that LPS and Lipid A core
antigens in the J5 vaccine might not be the key antigens stimulating high IgG1 and IgG;

antibody titers in vaccinated cattle.

Key words: humoral immunity, dairy cattle, core antigen, cross-reactivity

1. Introduction
Active immunization of dairy cattle, with Escherichia coli J5 bacterin, has been

used by dairy producers as a method to enhance immune resistance against Gram-

3O

negative bacteria, including mastitis-causing coliforms. The J5 bacterin is derived from a
mutant stain of E. coli (0111284; Re mutant), which lacks the “0” antigen capsular
portion of the cell wall but retains core lipopolysaccharide (LPS) and Lipid A antigens
that are highly conserved across all Gram-negative bacteria (Cullor, 1991b; Gonzalez et
al., 1989; Tyler et al., 1990b). Immunization of cattle with bacterins of J5 E. coli was
previously shown to elicit serum antibodies with cross-reactivity to a wide variety of
Gram-negative bacteria (Tyler et al., 1990a; 1992). However, some researchers have
claimed that J5 bacterin is a poor immunogen, eliciting poor serum and milk antibody
responses in vaccinated cattle (reviewed by Kehrli and Harp, 2001). Others have
reported relatively good IgM and IgG antibody responses in serum and milk of
vaccinated cows (Tyler et al., 1991, 1992; Tomita et al., 1995). Nevertheless, the
biological relevance of these antibody responses has not been well determined, leaving a
paucity of data on the mechanism(s) by which J5 vaccination protects dairy cattle from
disease caused by Gram-negative bacteria, especially coliforms. This is apparent by the
lack of knowledge of the magnitude of isotype-speciﬁc antibody responses, diversity of
antigens recognized, and the biological roles of isotype-speciﬁc serum antibodies elicited
by J5 vaccination. There is also little data available exploring the potential of improving
humoral immune responses against colifonn bacteria using vaccination protocols or
vaccine preparations different from those currently recommended and available.

We hypothesized in the current study that cattle hyperimmunized with J5 E. coli
bacterin would mount strong humoral immune responses, inducing IgG1 and IgG2
antibodies that are cross-reactive with a wide variety of Gram—negative bacteria The

objectives were to determine response proﬁles for isotype speciﬁc anti-J5 E. coli

31

antibodies in steers immunized 12 times with J5 E. coli bacterin; to determine titers of
isotype speciﬁc anti-J5 E. coli antibody in sera collected pre—immunization and on 6 days
post-3m, 6*, 9m, and 12th immunization, and to determine the cross reactivity of

antibodies contained in those selected sera with a variety of Gram-negative bacteria.

2. Materials and Methods
2. I Animals

Five healthy, 5-month old Holstein donor steers were used for this study. Steers
were owned by a private dairy in Michigan, and housed and cared for at the farm for the
duration of the study. Animals were group-housed with free access to water and feed. In
preparation for entry into our study, steers were immunized twice for BVD, IBR, P13
BRSV and 5 serovars of Leptospira (Triangle 9; Fort Dodge Animal Health, Fort Dodge,
IA), and dewonned (Ivermec, Merk & Co., Inc, Rahway, NJ) at 3 to 4 mOnths of age
(before the trial). All steers were deemed healthy at the start of the study, and remained
healthy throughout the study.
2. 2 J5 Immunization schedule

All steers were started on the J5 hyperimmunization protocol, bled and
immunized on the same days (see below). A total of 12 immunizations were
administered to each steer by subcutaneous injection of 5 ml of a commercially available
15 Escherichia coli bacterin (Pharmacia Animal Health, MI). The ﬁrst immunization was
administered at study time 0 (immunization number 1, Figure 9). The second

immunization was administered 30 days later. The subsequent 10 immunizations

32

(immunization numbers 3 through 12) were administered every two weeks, with an 8-

week rest period between immunization numbers 6 and 7 (Figure 9).

Figure 9 Schedule of immunizations (white arrows) and blood sampling (blue and
black arrows). Blood was collected from each steer immediately before the ﬁrst and
seventh immunizations (preimmunization samples), on days 4, 6, and 11 post
immunization for immunization numbers 1 through 6, and on days 4, 6, 11, and 13 post
immunization for immunization numbers 7 through 12.

Blood sampling ‘ (day relative to immunizations)

Preimmunizationamples *

11411 11 11 13

'3‘ 331113361113 $31613
#ﬁ I I ,1
etc. 7 8 9 etc. 12

Immunization numberﬂ

46 6 46 6
it! H {i *

r'r—T—r‘r
l 2 3

 

aid-:5

2.3 Blood collections

Blood was collected from each steer immediately before the ﬁrst immunization
(pm-immunization), on days 4, 6, and 11 post immunization for immunization numbers 1
through 6, and on days 4, 6, 11, and 13 post immunization for immunization numbers 7
through 12 (Figure 9). One-hundred and ﬁfty milliliters of blood were collected from
jugular veins into 50-ml sterile conical tubes using sterilized 14G, 3.75-cm needles.
Samples were allowed to clot overnight at 4°C and were then centrifuged at 2400 x g

[(GS-6R BECKMAN Centrifuge (Beckman Instrument, Inc., Schaumburg, IL)] for 30

33

min at 4°C. Sera were harvested and stored in sterile 50-ml tubes at -20°C until assayed
by ELISA.
2.4 Preparation of phenol-killed whole J5 Escherichia coli for use in ELISA

J5 E. coli Rc mutant, donated by Dr. James Cullor (University of California,
Davis, CA), was plated on 5% sheep blood agar and incubated overnight at 37°C. The
plate was then checked for culture purity. Single colonies from a pure culture plate were
selected and inoculated in 15 ml of Trypticase Soy Broth (TSB) in an orbital shaker [120
rotations per minute (rpm); J .T Baker, Phillipsburg, NJ] for 18 hr at 37°C. In ordewr to
establish purity of the inoculum, a 0.01 ml aliquot was collected from the inoculum and
plated on 5% sheep blood agar for isolation of individual colonies, and incubated
overnight at 37°C. The inoculum was stored at 4°C during this incubation, and if the
culture provided a pure culture, the inoculum was transferred into 1 liter-TSB and
incubated with shaking @ 120 rpm for 18 hr at 37°C. This preparation was then stored at
4°C while purity was assured by culture, and then bacteria were killed by adding 15 ml of
a 99% liquid phenol solution (Sigma; St. Louis, MO) per liter of the bacterial
preparation. This solution was incubated with shaking for 1 hr at 37°C and 120 rpm. The
phenol-killed whole cell pellets were recruited by centrifugation at 1000 x g for 12 min at
4°C. The pellets were washed twice by suspension in sterile 0.9% NaCl solution and
centrifugation (1000 x g for 12 min at 4°C). The pellets were suspended in sterile 0.9%
NaCl solution to reach 13% optical transmission (determined by Beckman DU 650
Spectrophotometer, BECKMAN COULTER, Inc., Fullerton, CA), which represented
approximately 1 x 109 cfu of killed bacteria/ml. This was the antigen used to coat wells

of ELISA plates (see below) ands was stored in 12 ml aliquots at -20°C.

34

2. 5 Isotype specific anti-J5 E. coli antibody response proﬁles

Serum anti-J5 E. coli IgM, IgG1, and IgG2 antibody responses were determined
using a J5 E. coli ELISA protocol modiﬁed from Tyler et al. (1990a) according to the
methods of Burton et al. (1993). Brieﬂy, one hundred microliters of J5 E. coli antigen
per well were used to coat wells of 96-well ELISA plates (ProBind Flat Bottom plates,
Fisher Scientiﬁc). The plates were covered and left to incubate on a ﬂat surface at room
temperature for 15 hours, after which unbound antigen was removed by six washes using
200 ul/well of biowash solution (0.9% saline containing 0.05% Tween 20). Test sera
were prepared for ELISA using separate 96-well, U-bottomed, polystyrene microtiter
plates (Becton Dickinson and Company, F lankin Lakes, NJ). Individual samples were
placed in the top well of each column and doubly diluted down the 8 rows of the column,
starting with a 1:50 dilution in row A and ending with a dilution of 1:6400 in row H.
Serum dilutions were made using a diluent, which was PBS, pH 7.3 with 0.5% Tween 20.
A positive control serum consisting of pooled steer sera from blood samples collected
following the 5th immunization was similarly diluted down rows of one column of each
plate. The negative control in this assay was fetal bovine serum (heat-inactivated,
endotoxin s 10 EU/ml, Life Technologies, Rockville, MD) and was diluted in the same
fashion in one column of each plate. Samples contained in each well of the U-bottomed
dilution plates were then transferred (100 til/well) to washed ELISA plates using a multi-
channel pipetter (Costar Corporation, Cambridge, MA). Additional controls in this assay

included quadruplicate wells of 1:400 dilutions of the positive and negative control sera,

35

duplicate blank wells, duplicate wells that received only initial J5 coating, and duplicate
wells that received all reagents except test sample.

Once all samples were delivered to appropriate wells, ELISA plates were sealed
and incubated at 37 °C for 45 min, washed 6 times (as above) to remove unbound
antibodies, and 100 ul/well of detection antibodies (Horseradish peroxidase-conjugated
sheep anti-bovine IgM, IgG], or IgG2, BETHYL Laboratories, Inc., Montgomery, TX)
diluted to 125,000 with the diluent were added to wells of appropriate plates. Plates
were again sealed and incubated for 30 min at 37°C, washed 6 times to remove unbound
detection antibody, and 125 til/well of substrate (hydrogen peroxide-azino-bis-3-
ethylbenzthiazoline sulfonic acid, Sigma Chemical Co., St. Louis, MO) added to all wells
except the duplicate blank wells and duplicate J5 antigen only well for 45 min at 37°C.
Relative antibody concentrations in samples were recorded as optical density (OD)
following spectrometric analysis at a dual wavelength of 405—450 nm using an ELISA
plate reader (Benchmark, Bio-RAD Laboratories, Hercules, CA). The plate reader was
blanked against the 2 blank wells of each plate. In our ELISA assays, the greatest and
most repeatable differences in IgG; ODs between the positive and negative control sera
occurred at the 1:400 dilution. All samples in a plate were repeated if OD values of the
positive and negative control sera were not within pre-established ranges (0.99 5. OD _<_
1.10 and OD _<_ 0.09, respectively). Mean antibody response was reported from statistical
analysis of test sample OD at serum dilutions of 1:100 for IgM and of 1:400 for IgG and
IgG2. Isotype speciﬁc anti-J 5 E. coli antibody responses based on ODs of these selected

dilutions of test sera were parallel to the response proﬁles obtained from OD of one

36

doubling dilution above and below the selected dilution (correlation coefﬁcients between
sample dilutions within isotype > 0.90; P < 0.0001).
2.6 Isotype speciﬁc anti-J5 E. coli antibody titers in selected serum samples

The previously described ELISA was used to determine isotype speciﬁc anti-J5 E.
coli antibody endpoint titers in selected sera (pre—immunization, and 6 days post 3“, 6‘“,
9m, and 12'h immunizations). These sera were doubly diluted down columns of each plate
in duplicate, including 1:4, 1:8, 1: 16,... , and 1: 131,072 dilutions. Pooled hyperimmune
steer sera (7‘h immunization) and fetal bovine serum (heat-inactivated, endotoxin s 10
EU/ml, Life Technologies, Rockville, MD) served as positive and negative controls,
respectively, and were doubly diluted down 2 columns in the same fashion as test sera.
End-point titers of the test and control sera were recorded as log; of the last dilution to
yield an OD < 0.10, which we considered as zero titer because the FBS negative control
always had an OD < 0.10.

2. 7 Antibody cross-reactivity assay

The cross-reactivity of isotype-speciﬁc antibodies from the titered sera was
determined by ELISA, basically as described for the proﬁling of isotype speciﬁc anti-15
antibodies. However multiple bacterial solutions (approximately 1 x 109 cfu/ml each) in
addition to the J5 E. coli (Rc mutant) were prepared for well coating of the ELISA plates.
Escherichia coli McDonald 48 7, Pseudomonas spp., Serratia spp., Klebsiella '
pneumoniae, Salmonella newport, Salmonella typhimurium, J5 E. coli and
Staphylococcus aureus were cultured and killed with phenol to make antigen solutions as
previously described. S. newport and S. typhimurium were donated by the Animal Health

Diagnostic Laboratory, East Lansing MI. The other bacteria originated from milk

37

cultures of cows with clinical mastitis (bacterial stocks available through the Mastitis
Research Laboratory, Michigan State University; Department of Large Animal Clinical
Sciences. The ELISA plates were coated using 100 til/well of each bacterial solution.
Similar preparations of J 5 E. coli and S. aureus were used as antigen positive and
negative controls, respectively. LPS (Sigma-Aldrich Co., St. Louis, MO, 10 ug/ml) and
lipid A [monophosphoryl from S. minnesota (Re 595), Sigma-Aldrich Co., St Louis,
MO, 10 ug/ml] were also used as antigens in some wells and were prepared as described
in Freudenburg (1989). The sera from pre-immunization, and 6 days post-3“, 6th, 9th, and
12th immunizations were then diluted at 1:50 (for IgG; and IgM) or 1:200 (for IgG1) for
determining isotypespeciﬁc cross-reactivity of the antibodies contained in the various
sera and to quantify relative concentrations (as OD) of the isotype speciﬁc cross-reactive
antibodies.
2.8 Statistical analysis

For statistical analysis of the antibody response proﬁles, immunization number
and sample day nested within immunization number were included in a mixed model as
ﬁxed effects along with assay day, assay plate nested within assay day, and animal as
random effects. In addition, autoregressive order one covariance structure based on
Schwarz' Bayesian model-ﬁt criterion was speciﬁed in the REPEATED statement of SAS
to model covariance structure within subject. Statistical differences in mean OD over
days following each immunization, and speciﬁc differences between the mean OD for
immunizations 1, 2, and 3 versus the pre-immunization mean OD or the ODs of all other
immunizations were determined at a signiﬁcant level P < 0.05 using LSMEAN and

ESTIMATE statements of SAS with the Tukey-Kramer adjustment factor.

38

Repeated measures analysis of variance was used to analyze anti-J5 E. coli IgM,
IgG1, and IgG; antibody end-point titer data and isotype speciﬁc cross-reactivity data
sets. The analyses were carried out utilizing the MIXED procedure of SAS software
(SAS Institute, Cary, NC). Overall signiﬁcance of effects in the model included number
of immunizations for the isotype speciﬁc anti-J 5 E. coli antibody titer data, the
interaction of “BACTERIA” and immunization number, and BACTERIA for cross-
reactivity OD data, and was assessed by the Type III F -test. Statistical signiﬁcance of
adjusted mean comparisons of effects declared relevant by the t-test was determined by
the Tukey-Krarner multiple comparison procedure. The signiﬁcance level was set at _ =

0.05.

3. Results
3.1 Hyperimmune response proﬁles

The Least Squares Means (LSMean :1: SEM) reported in Figures 10, 11, and 12
were derived from the mean ODs of samples collected aﬁer each immunization. Steer
did not contribute signiﬁcantly to variation in mean post immunization antibody
responses for any isotype studied (P > 0.10). However, immunization number and day
post immunization signiﬁcantly inﬂuenced all antibody response curves (P < 0.0001).
The following trends were observed for the day post immunization effect (data not
shown). There was no difference between sampling days following the ﬁrst 3
immunizations and for the 8th to 12th immunizations. For all other immunizations,
highest ODs were observed at 11 days post immunization ( 4th, 5th and 6"), or at 6 and 11
days post 7th immunization. Adjusted differences in mean OD within immunization over

days post immunization showed that only two doses of J 5 bacterin were required to

39

signiﬁcantly (P s 0.05) increase serum anti-J5 E. coli IgM antibody above pre-
immunization levels (Figure 10). Similarly, it took two J5 doses to elicit increases in
anti-J5 E. coli IgG1 antibodies pre-immunization levels (P = 0.004), and this isotype
continued to increase in serum through the 8'h immunization, after which the IgG
attained a plateau (P s 0.02, Figure 11). For IgG2, ﬁve doses of J5 vaccine were required
to achieve isotype switching, and increase serum anti-J 5 E. coli IgG2 antibodies above
pre-immunization levels (P < 0.02; Figure 12).
3.2 Isotype speciﬁc anti-J5 E. coli antibody endpoint titers

Pre-immunization sera contained varying amounts of naturally occurring anti-J5
antibodies for all three isotypes (Figure 13). These anti-J5 antibody endpoint titers
increased through 12 immunizations for all three isotypes, and were signiﬁcantly higher
(P < 0.005) than pre-immunization following administration of the 6th dose of J5 bacterin
(Figure 13). It is noteworthy that three doses of vaccine were not sufﬁcient to increase
anti-J5 E. coli IgM, IgG], or IgG; titers because this is the number of doses recommended
by the vaccine manufacturer.
3.3 Isotype specific antibody cross-reactivity

As evidenced by increased OD values, cross-reactive IgM antibodies in sera of
the test steers were high after 6 immunizations for Klebsiella app. , Pseudomonas, and
Serratia app. (P < 0.01), after 9 immunizations for S. typhimurium and E. coli 487, (P <
0.01), and after 12 immunizations for S. Newport (P < 0.01) and Lipid A (P < 0.05) as
compared to corresponding pre-immunization titers (Figure 14). Anti-J5 IgG1
signiﬁcantly increased after the 6th immunization for all bacterial antigens (P < 0.01)

except for E. coli 487, which only required 3 immunizations (P < 0.001), and S. neWport,

40

which required 9 immunizations (P < 0.05, Figure 15). IgG; antibody titer increased
signiﬁcantly after the 6th immunization for all bacterial antigens compared to pre-
immunization titers.(P < 0.01). Also, the IgG1 and IgG2 antibodies in test sera were more
likely to cross-react with E. coli 48 7 and Pseudomonas spp. than Salmonella bacteria (P
< 0.05, Figure 15 & 16). Low cross-reactivity for LPS, Lipid A, and S. aureus was
observed for all three isotypes, especially as compared with cross-reactivity to Gram-

negative bacteria (P < 0.001, Figure 14-16).

4. Discussion

In the current study, J5 hyperimmunization of dairy cattle achieved increased
isotype speciﬁc immune responses, titers, and cross-reactivity of anti-J5 antibodies with
multi-heterologous Gram-negative bacteria compared to no immunization and the
manufacturer’s recommended three immunizations. Therefore, results of this work
suggest that more doses of J5 bacterin than currently recommended may be required to
induce strong cross-reactive anti-J5 IgM, IgG1, and especially, IgG2 serum antibody
responses in dairy cattle.

Generally, antibodies of the IgM isotype are the ﬁrst to be secreted from B cells in
response to primary exposure to thymus-dependent (TD) antigens (Janeway et al., 2001).
With successive exposures to the same antigen, IgM-to-IgGl isotype switching occurs in
the responsive B cell clones, effectively shutting down their IgM secretion. It is now
known that this classic humoral immune response occurs via a speciﬁc class of B cells
that are characterized by lack of surface CD5 expression (called B-2 cells). However, a

second class of CD5-expressing B cells (CD5+, B-l cells) has been identiﬁed and secretes

41

predominantly IgM antibodies in response to repeated stimulation by thymus-
independent (TI) antigens (Haas and Estes, 2000; Hayakawa et al., 1984; Kantor 1991;
Mond et al., 1995). TI antigens characteristically possess highly repetitive epitopes, such
as those that occur in capsular polysaccharides of Gram-negative bacteria (Mond et al.,
1995)

Because core lipopolysaccharide antigens are exposed in the J5 strain of E. coli, it
is very possible that the sustained IgM responses to 15 hyperimmunization observed in
the current study resulted from preferential activation of CD5+ B cells. CD5"° B cells
exist in cattle and have even been shown to account for up to 30% of adult peripheral
blood lymphocytes during certain infections (Buza et al., 1997; Haas and Estes, 2000;
Meirom et al., 1993; Nassens and Williams, 1992). Antigen-activated bovine CD5+ B
cells do not undergo signiﬁcant IgM-to-IgGl isotype switching, eventually leading to
very high serum IgM antibody titers (Hayakawa et al., 1984; Janeway et al., 2001;
Nassens and Williams, 1992). So, while some B cell clones obviously switched from
IgM to IgG1 production in our J5 hyperimmunized steers, other B cell clones that may
have been CD5+ also continued to produce increasing amounts of IgM with each
successive J5 immunization.

In the current study, evidence exists for isotype switching (from IgM to IgG1 and
IgG2) after repeated exposure of the bovine immune system to Re mutant J5 E. coli
bacterin. Theoretically, strong IgG antibody responses result when CD5' B cells are
continuously exposed to TD antigens in the context of major histocompatibility complex
(MHC) class II antigens on antigen presenting cells (Haas and Estes, 2000; Janeway et

al., 2001). Speciﬁc interaction between the antigen-presenting cells (APCs) and CD4+ T-

42

helper cells (TH2) via CD402CD40L ligation causes the TH2 cells to secrete potent
stimulatory cytokines (IL-4, IL-5, and IL-6) that promote isotype switching in CD5' B
cells from IgM to IgG1 (Haas and Estes, 2000; Janeway et al.,2001; McGuire et al.,
1987). This humoral response may have occurred in our J5 hyperimmunized steers
because we observed signiﬁcant serum IgG1 antibody responses after the 5th
immunization, even as the IgM response continued to increase. In addition,
CD402CD40L ligation must have occurred in some APC and T31 CD4+ cells, with
subsequent release of proinﬂammatory IFN-y from the T cells, because there was also
isotype switching to IgG2 after the 5th exposure to J5 bacterin. IFN-y has been shown to
be the major cytokine for IgM to IgG2 switching in bovine CD5' B cells (Estes, 1996).

However, peak anti-J 5 E. coli IgG; antibody responses were not achieved until the
11th exposure to .15 bacterin, at which time the IgG1 response had plateaued.
Interestingly, the IgM response continued to increase beyond immunization number 11.
While it is unknown what components of the J5 bacterin are recognized best or most
quickly by responsive B and T cells, the early and continued increase in the serum IgM
response may suggest that the TI-LPS antigens may have rapidly stimulated CD5+ B cells
production of IgM. In contrast, bacterial proteins (TD antigens) may have required more
exposures and time to be properly processed and presented in MHC II molecules of APC
for TH cell-induced production of IgG1 and IgG2 by CDS' B cells. °

If the scenario just described was true, we would have expected to observe
signiﬁcant cross-reactivity of the serum IgM antibodies with LPS and (or) Lipid A during
ELISA. However, this was not the case. In fact, antibodies of all 3 isotypes reacted

weakly only with LPS or Lipid A, and only after 12 J 5 immunizations. Instead, the

43

antibodies contained in the test sera demonstrated increasing cross-reactivity with whole
(killed) Gram-negative bacteria (but not with S. aureus). We also observed considerable
variation in binding preference of the three antibody isotypes across bacteria. For
example, IgG1 and IgG2 antibodies recognized E. coli 487 better than IgM antibodies,
which preferentially cross-reacted with Serratia spp. However, there were some pre-
existing antibodies against Gram-negative bacteria, which was previously reported
(Gonzalez et al., 1989). This might be explained through the natural exposure of the
animal to Gram-negative bacteria in its environments, and especially in the digestive
tract. Also, in an agreement with a previous study (Tyler et al., 1991), the polyclonal
antibodies in hyperimmune serum did not recognize a Gram-positive isolate, S. aureus.
There is controversy over how LPS inﬂuences the humoral immune response.
Most J5 vaccine studies-have focused attention on anti-LPS and anti-Lipid A IgG
antibodies in immune serum suggesting that the levels of these are relatively high
following immunization and thus that these antibodies are responsible for cross-
protective against heterologous Gram-negative pathogens (Aydintug et al., 2001;
Baumgartner et al., 1987; Pollack et al., 1989; Tyler et al., 1991; 1992). Our IgM, IgG1
and IgG2 cross-reactive antibody data do not support this notion because we found little
cross-reactivity of antibodies, even in hyperimmunized serum, with LPS or Lipid A.
Hellman and colleagues (1997) suggested that solid-phase assays, such as the ELISA,
may not be the best approach to study anti-LPS antibody. However, our results on
antibody cross-reactivity with LPS is in agreement with results of previous ELISA

studies that IgG elicited from E. coli J5 bacterin binds only weakly to LPS from

heterologous, smooth, Gram-negative bacteria in solid-phase assays (Siber et al., 1985;

Warren et al., 1987).

5. Conclusion

Mean anti-J 5 E. coli IgG1 and IgG2 serum antibody responses elicited by three J5
immunizations were not impressive. In contrast, J5 hyperimmunization (ﬁve or more
vaccinations) elicited strong and sustained serum IgG; and IgG; antibody responses. In
this steer model of hyperimmunization, at least 5 injections of the manufacturer’s
recommended dose of J5 bacterin were required to successfully elevate anti-J 5 E. coli
IgG1 and IgG; antibody levels above background (pm-immunization) levels and to
initiate IgM to IgG; or IgG2 isotype switching. Even so, it took 8 to 10 immunizations to
elicit maximum anti-J5 E. coli IgG1 and IgG; antibody responses, the latter which could
be sustained during the subsequent 11th and 12th immunization periods. Anti-J 5 E. coli
IgM antibody responses rose sharply from background after only 2 immunizations and,
quite unexpectedly, continued to increase over the remaining 10 immunization periods.
Importantly, the levels of cross-reactivity for heterologous Gram-negative bacteria also
increased for each antibody isotype with increasing immunizations, but responses to LPS
and Lipid A were always low.

Therefore, results of our-study suggest that repeated exposure of the bovine
immune system to Re mutant J5 E. coli bacterin results in antibodies of all isotypes that
preferentially recognize non-LPS antigens shared by a diverse group of Gram-negative
bacteria. The slow, but continuous increase in cross-reactive serum IgG; antibodies with

increasing doses of administered JS bacterin would suggest that J5 vaccine use protocols

45

be modiﬁed to obtain this response in target animals because IgG; is the main opsonic

and cytophilic isotype for bovine neutrophils recruited to sites of Gram-negative

infections.

Optical Density (1:100 dilution of sera)

0.9 r
0.8 -
0.7 -
0.6 -
0.5 a
0.4 -
0.3 -
0.2 ~
0.1 a

 

 

 

T

re-immunization

I
0
P

I I I I I I I I I

1 2 3 4 5 6 7 8 9
Immunization Number

I I I

I
10 ll 12 13

Figure 10 Mean serum anti-J5 E. coli IgM antibody response proﬁle of ﬁve
Holstein steers immunized 12 times with a J5 bacterin. Each data point is an
immunization number LSMean from sera collected either on days 4, 6, and 11 post
immunization (ﬁrst six immunizations) or on days 4, 6, 11, and 13 post
immunization (last six immunizations), expressed as optical density (:1: SEM) when
sera were diluted 1:100. Immunization number (P < 0.0001) and day within
immunization number (P = 0.018) signiﬁcantly inﬂuenced the mean anti-J 5 E. coli
IgM antibody response. [* Indicates the point at which mean optical density was
signiﬁcantly higher than the pre-immunization mean (P s 0.05). it? Indicates the
mean optical density was signiﬁcantly different than that at the 3rd immunization (P

s 0.03)].

46

1°- ‘X- . * ‘* *
.. . i iii ..
,~"°.°°'.°‘° WA.
0.8 " T ..O‘ l 1 1° 03’]-
* A..... .‘
0.7- 3.» l 1
l... l
0.6- ot~ .4

0.4 -

l
0.3- 1

Optical Density (1:400 dilution of sera)

 

 

0.2-
0.14
0 I l I I I I I l I I I I I I
-1012345678910111213
T Immunization Number '
Pre-immunization

Figure 11 Mean serum anti-.15 E. coli IgG] antibody response proﬁle of ﬁve Holstein
steers immunized 12 times with a 15 bacterin. Each data point is an immunization
number LSMean from sera collected either on days 4, 6, and 11 post immunization (ﬁrst
six immunizations) or on days 4, 6, 11, and 13 post immunization (last six
immunizations), expressed as optical density (:1: SEM) when sera were diluted 1:400.
Immunization number (P < 0.0001) and day within immunization number (P = 0.011)

signiﬁcantly inﬂuenced the mean anti-J 5 E. coli IgG1 antibody response. [* Indicates
. the point at which mean optical density was signiﬁcantly higher than the pre-

immunization mean (P = 0.004). 4* Indicates mean optical densities that were

signiﬁcantly higher than optical density at the 3rd immunization (P s 0.02)].

47

0.4 -

 

 

--0-- IgG2
at ale *
gos- * I T l I
a * I ..0"°--0
° 1 o°"° l l
s * x1 1 l
g 0.21 at» TAT
is T ,0’
'§' T To'.0 1
g 0.1" T i LOO-’6 .l.
0% 0"]. J- J- J-
“ 1
e 01
‘3.
o
'0.1 I I I I l I I I I I l I I I
-0 2345678910111213
T Immunization Number

Pre-immunization

Figure 12 Mean serum anti-J 5 E. coli IgG; antibody response proﬁle of ﬁve Holstein
steers immunized 12 times with a J5 bacterin. Each data point is an immunization
number LSMean from sera collected either on days 4, 6, and 11 post immunization
(ﬁrst six immunizations) or on days 4, 6, 11, and 13 post immunization (last six
immunizations), expressed as optical density (: SEM) when sera were diluted 1:400.
Immunization number (P < 0.0001) and day within immunization number (P <

0. 0001) signiﬁcantly inﬂuenced the mean anti-J 5 E. coli IgG; antibody response.

[it Indicates mean optical densities that were signiﬁcantly higher than the optical
density at 3“i immunization (P s 0. 001)].

48

(a) 16

12
10
Log2 Titer 8

l l l l l l l l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0))
IgG,
Log, Titer
0 3 6 9
(e) 15 1
1:0. 14 - * a.
12 T _J_ . L—_I-_.
10; .I. t-:-: t:-:-
Log, Titer 8 .2" FL". ..‘_'
6 ~ *-:-: -:-:-: -:-:
4 ‘ Z-Z- I-Z-I- I: .
2 a .-_- -_-_' _-.-
0 r P - - r - - ° 1 - - I l
0 3 6 9 12

Immunization number

Figure 13 Mean (:9: SEM) isotype-speciﬁc anti-J 5 E. coli antibody titers of
selected hyperimmune serum samples from ﬁve Holstein steers. Titers are
expressed as log 2 of endpoint titer. (a) IgM; (b) IgG1; (c) IgG2. * Mean titer is
different from the pre-immunization titer at P < 0.005.

49

Figure 14 Mean (.4: SEM) cross-reactivity of anti-15 E. coli IgM antibodies contained in
selected sera (at 1:50 dilution) from pre-, 3‘“, 6‘“, 9‘“, and 12‘h immunizations to
heterologous Gram—negative bacterial antigens, LPS and Lipid A. Data were presented as
OD. Each panel represents anti-J5 IgM antibody against test Gram-negative bacterial
antigens compared to .15 E. coli as a positive control and S. aureus as a negative control.
Test Gram-negative bacteria include: panel (a) S. typhimurium, Serratia spp., and S.
newport; panel (b) E. coli 487, Pseudomonas spp., and Klebsiella spp.; and panel (c) LPS
(from E. coli 111:B4) and Lipid A (from S. minnesota, Re mutant). Mean OD
signiﬁcantly higher following immunizations; a (P < 0.01), b (P < 0.05). Mean OD
signiﬁcantly higher than LPS, Lipid A and S. aureus, * (P < 0.001).

IgM

ES. typhimurium E Klebsiella spp

lSerratia spp. B LPS (E. coli01112B4)

ISIS. newport S Lipid A (S. minnesota)

EE. coli487 IJ5 E. coli (positive control)
EPseudomonas spp. C] S. aureus (negative control)

    

 

 

 

 

 

 

 

 

0 3 6 9 12
Serum sample (1:50 dilution)

51

Figure 15 Mean (i SEM) cross-reactivity of anti-.15 E. coli IgG1 antibodies contained in
selected sera (at 1:200 dilution) from pre-, 3'“, 6‘“, 9‘“, and 12th immunizations to
heterologous Gram—negative bacterial antigens, LPS and Lipid A. Data were presented as
OD. Each panel represents anti-15 IgGl antibody against test Gram-negative bacterial
antigens compared to JS E. coli as a positive control and S. aureus as a negative control.
Test Gram-negative bacteria include: panel (a) S. typhimurium, Serratia spp., and S.
neWport; panel (b) E. coli 48’7, Pseudomonas spp., and Klebsiella spp.; and panel (c) LPS
(from E. coli 111:B4) and Lipid A (from S. minnesota, Re mutant). Mean OD
signiﬁcantly higher following immunizations; a (P < 0.01), b (P < 0.05). Mean OD
signiﬁcantly higher than LPS, Lipid A and S. aureus, * (P < 0.001).

52

OD

OD

1.40 s "

1.20 i

0.20 "

 

1.40 r
1.20 n
1.00 ~
0.80 “
0.60 -*
0.40 ~
0.20 *

IS whimurium
DISerratia spp.
EDS. newport
EIE. coli 487

(a)

 

IIIIIIIIIIIlI-l

 

 

 

IgG1

IPseudomonas mo.

Klebsiella spp
BLPS (E. coli01112B4)
D Lipid A (S. minnesota)

IJ5 E. coli (positive control)
[:1 S. aureus (negative control)

.,
'9: .
1:2;

i125-._ I

 

  

321% U

lllIlIIIllllllllllIl 9

 

 

 

 

 

 

C. I I.

O I... I I I I I O O O I O. C. D I I O
I. C I O O I O O I. C I D O I. I. O I
A‘A‘A‘AAAAAA‘AAAAAAAAJ

   

"iiiiii'"

 

 

 

 

0.00

 

 

 

 

Serum sample (1:200 dilution)

53

 

Figure 16 Mean (i- SEM) crossoreactivity of anti-15 E. coli IgG2 antibodies contained in
selected sera (at 1:50 dilution) from pre-, 3'“, 6m, 9‘“, and 12ul immunizations to
heterologous Gram-negative bacterial antigens, LPS and Lipid A. Data were presented as
OD. Each panel represents anti-.15 IgG2 antibody against test Gram-negative bacterial
antigens compared to J5 E. coli as a positive control and S. aureus as a negative control.
Test Gram-negative bacteria include: panel (a) S. typhimurium, Serratia spp., and S.
newport; panel (b) E. coli 487, Pseudomonas spp., and Klebsiella spp.; and panel (c) LPS
(from E. coli 111:B4) and Lipid A (from S. minnesota, Re mutant). Mean OD
signiﬁcantly higher following immunizations; a (P < 0.01), b (P < 0.05). Mean OD
signiﬁcantly higher than LPS, Lipid A and S. aureus, * (P < 0.001).

OD

OD

OD

0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00

0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00

0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00

I S. typhimurium
ZSerratia spp.

BS. newport

53E. coli 487
EPseudomonas spp.

 

 

IgG2

E Klebsiella spp I

ELPS (E. coli01112B4)

S Lipid A (S. minnesota)
.J5 E. coli (positive control)
[:1 S. aureus (negative control)

      
    
 

  

i

   

’JJ.IJJIJ-

/
/
i

I
__I

 

 

 

  

9
Serum sample (1:50 dilution)

3 6

55

CHAPTER 3

GRAM-NEGATIVE BACTERIAL ANTIGEN RECOGNITION
BY ISOTYPE SPECIFIC ANTI-J5 E. coli ANTIBODY

Anantachai Chaiyotwittayakun", Ronald J. Erskine”, ‘
Rebecca A. Darch, Patty S.D. Weberb, Jeanne L. Burtonb’

Departments of “Large Animal Clinical Sciences and bAnimal Sciences,
Michigan State University, East Lansing 48824, USA
Abstract
We conducted this study to determine the cross-reactivity and diversity of
bacterial antigen recognition by serum antibodies following multiple-dose administration
of J5 bacterin in steers, and determine if enriched anti-J 5 E. coli IgG1 and IgG2 antibodies
opsonized E. coli for improved neutrophil phagocytosis. Lysates were prepared from
heterologous Gram-negative bacteria by sonication and used to determine isotype speciﬁc
antibody recognition of various proteins and LPS. Anti-J5 E. coli IgG1 and IgG2 bound
distinctively to Gram-negative bacterial protein antigens of a few key sizes, the most
predominant groups being of molecular masses 8-10.5 and 34-39 kDa. IgG; recognized
these proteins more strongly than IgG1, whereas there was little recognition of them by
IgM antibodies. Therefore, the 8-10.5 kDa antigens may be the targets of T and B cell
recognition during immunization of cattle with J5 vaccines. we also observed modest
interaction between IgG1 in hyperimmune serum and puriﬁed LPS. However, there was
little or no cross-reactivity of our antisera or enriched IgG1 and IgG2 antibodies to Lipid

A or S. aureus. Also only enriched IgG2 antibodies enhanced phagocytosis of GFP-

 

Correaponding author: Dr. Jeanne L. Burton, Department fo Animal Science, Michigan State University,
East Lansing, MI 48824-1314, Tel. 517-3 53-9702, Fax. 517-353-1699, e-mail address: burtonj@msu.edu

56

transformed J5 E. coli by bovine neutrophils compared to no antibody controls. Thus,
multiple J5 immunizations appear to enhance IgG2 antibodies that increasingly recognize
small proteins in heterologous Gram-negative bacteria and enhance neutrophil

phagocytosis of E. coli in vitro.

Keywords: antigens, humoral immunity, vaccine, phagocytosis

1. Introduction

Cell wall components of Gram-negative bacteria are considered to play an
important role as virulence factors, which subsequently cause pathophysiological changes
in the host (Kremer et al., 1990; Tyler et al., 1990b). Most researchers have been
focusing on two cellwall components, including LPS and Lipid A regarding speciﬁc host
immune response following immunization with J5 E. coli bacterins. Studies reported
relatively high antibody responses (both monoclonal and polyclonal antibody) against
LPS and Lipid A antigens in the vaccine, which is thought to be the key response that
provides protection against a variety of Gram-negative pathogens in immunized animals
(Aydintug et al., 2001; Baumgartner et al., 1987; Pollack et al., 1989; Tyler et al., 1991;
1992). However, other evidence suggests that antisera containing IgG antibodies bind
only weakly to LPS from multiple heterologous Gram-negative bacteria (Siber et al.,
1985; Warren et al., 1987). In addition, our recent work has shown that serum from J5
hyperimmunized steers contains only low or undetectable levels of anti-LPS and anti-

Lipid A IgM, IgG1 and IgG2 antibodies (Chapter 2).

57

It is known that LPS and Lipid A are not the only core antigens contained in the
cell walls of Gram-negative bacteria, but that other components including lipoprotein,
peptidoglycan, and outer membrane proteins are also shared across bacterial isolates
(OMPs). Currently, these bacterial antigens are believed to play a role in the
pathogenesis of Gram-negative sepsis (Hellman et al., 1997; 2000a). Recent studies in
humans and mice reported conserved gram-negative bacterial OMPs including outer
membrane protein A (OmpA), and peptidoglycan-associated lipoprotein (PAL) (Hellman
et al., 1997; 2000a). These OMPs are released into human blood circulation during
sepsis as a complex with LPS, which is most likely to occur during rapid bacteria growth
and may play a role in the pathogenesis of Gram-negative sepsis (Hellman et al., 1997;
2000a). OMPs are also recognized by anti-15 IgG polyclonal antibody elicited by heat-
killed E. coli J5 (Hellman et al., 1997; 2000a; 2000b; 2001). OmpA (i.e., Klebsiella
pneumoniae) is also used as a carrier protein for some polysaccharides (PS) to increase
the immunogenicity of PS and enhance protective antibodies against lethal bacterial
infection (Libon et al., 2002; Rauly et a, 1999). Thus, use of OMPs in vaccines might
contribute toward improved efficacy for host immunity against Gram-negative bacteria.
If true, it may be possible to improve J5 bacterins currently used to actively immunize
dairy cows against mastitis-causing coliforms.

However, studies in-cattle regarding isotype-speciﬁc anti-J5 antibodies and their
recognition of speciﬁc Gram-negative antigens have not been conducted. For our study,
we have used serum antibodies from J5 hyperimmunized steers to examine which protein
and polysaccharide antigens from Gram-negative bacteria are recognized. Based on work

of Chapter 2, our hypothesis was that repeated exposure of the bovine immune system to

58

Rc mutant J 5 E. coli bacterin would result in highly mature IgG2 antibodies that cross-
react with a variety of protein antigens on heterologous Gram-negative bacteria, and that
these antibodies would enhance neutrophil recognition and phagocytosis of Gram-
negative bacteria. Therefore, the current study was conducted: 1) to determine the cross-
reactivity and diversity of bacterial antigen recognition by maturing isotype speciﬁc
antibodies in J5 hyperimmune serum; and 2) to determine if the IgG; fraction of
hyperimmune serum opsonizes E. coli better than the IgG fraction for improved

neutrophil phagocytosis of coliform bacteria.

2. Materials and Methods
2. 1 Test and control samples

Test samples included various pre-immune and immune test sera, IgG1 and IgG2
antibodies afﬁnity puriﬁed from immune test sera, and fetal bovine senim as a negative
control (GibcoBRL, Life-technologies). Previously described in Chapter 2, test sera were
harvested from blood samples collected from ﬁve healthy Holstein steers that were
subcutaneously administered multiple doses of J 5 bacterin over time (Pharmacia-Animal
Health). The sera were pooled by immunization number (pre-immunization, 3”, 6th, and
9th immunization) and stored at -20 °C until use.
2.2 Protein A purification of serum IgG; and IgG; antibodies
2. 2. I Ammonium sulfate precipitation

Hyperimmune sera collected from the ﬁve Holstein steers vaccinated 9 times
using J 5 E. coli bacterin were prepared for Protein A column chromatography by a

modiﬁcation of the method of Harlow and Lane (1999) (Figure 17). Brieﬂy, one liter of

59

Figure 17 Protein A chromatography isolation of IgG1 and IgG; from pooled serum of J5
E. coli hyperimmunized steers. An Afﬁ-Prep Protein A column was used in conjunction
with a pH gradient elutant to produce two eluted fractions of IgG1 and IgG2 (see Figure 18).

 

Hyperimmune
serum

pH gradient
technique

is eiﬁ

g Eluted Eluted 2 Wash Efﬂuent

5 IgG IgG : -
...-IIIII:II.IIIII.II..2.I.; ﬂuld N IgG! & IgM

   
     

 

95%;) protein A
5:}:{5 Column

   

Figure 18 Elution pattern of afﬁnity puriﬁed immunoglobulins from an Afﬁ-Prep
Protein A column using a pH gradient from 8.0 to 2.0. One-milliliter fractions were
collected, neutralized immediately by adding 1 M Tris pH 9.0, and protein
concentrations (rig/ml) determined a optical density at )t = 280 nm. Peak 1 was the
efﬂuent. Peak 2 eluted between pH 8.0 and 7.5 and peak 3 at pH 5.0, and were shown
by subsequent Western blot analysis (see Figure 25) to be IgG and IgG2, respectively.

—a— Optical Density [3 pH Gradient
4 10

 

Peak 1 Peak 2 Peak 3

 

 

 

Fraction Number

the pooled hyperimmune serum was centrifuged at 3210 x g (GS-6R BECKMAN
Centrifuge, Beckman Instruments, Inc., Schaumburg, IL) at 4 °C for 15 minutes to
remove debris. Protein concentration of the supernatant was determined using the
Warburg-Christian program by spectrophotometry (Model DU° 650, Beckman Coulter,
Inc., Fullerton, CA). Seven hundred milliliters of 4.1 M saturated ammonium sulfate, pH
7.0 was gradually added to the serum and the proteins were allowed to precipitate
overnight at 4°C. The suspension was then gently stirred for 5 minutes and centrifuged at
3210 x g at 4 °C for 25 minutes. After discarding the supernatant, the pellet was
reconstituted with sterile Milli-Q water to a ﬁnal volume of 30 to 50% of the starting
serum, and termed saturated-ammonium sulfate serum (SASS). Subsequently, the SASS
was dialyzed for 6 changes every 6 to 8 hours against pre-cooled 0.9 % NaCl at 4 °C
using 50,000 molecular weight cut-off (MWCO) dialysis membranes (Spectra/Pro° 6
Membrane, Spectrum Laboratories, Inc.).
2.2.2 Dialysis

Two l-liter beakers containing 500 ml of Milli-Q H20 (~18 mQCm, Milli-Q
water system) and 500 ml of 0.9% NaCl were prepared and pre-cooled at 4 °C. Dialysis
membrane (Molecular weight cut off (MW CO) = 50,000, Spectra/Pro” 6 Membrane,
Spectrum Laboratories, Inc.) was cut to approximately 45.8 to 50.8 cm long for 6 - 8
pieces or as needed, and placed into the beaker containing 500 ml °of Milli-Q H20 at 4 °C
for 15 minutes. The membranes were transferred to another beaker containing 0.9%
NaCl, allowed to soak for another 15 minutes, and rinsed with 20 ml of Milli-Q H20.
One end of each dialysis membrane was double-clamped with dialysis clamps. The

membrane was ﬁlled by the sample (i.e., SASS) approximately half full and secured by a

61

dialysis clamps at the other end. The membranes containing samples were placed into a
24-liter carboy containing 0.9% NaCl, which had been pre-cooled at 4 °C and stirred with
a magnetic stir bar. SASS was dialyzed for 6 changes every 6 to 8 hours and transferred
to a graduated cylinder to record volume and determine the amount of protein at 280 nm
using Nucleic Acid Module in the Spectrophotometer (Model DU° 650, Beckman
Coulter, Inc., Fullerton, CA). The samples were then stored in 50-ml sterile conical tubes
at -20 °C for further analysis.
2.2.3 Protein A aﬁinity chromatography

Afﬁnity chromatography was performed at room temperature by using a 25-ml-
Afﬁ-Prep Protein A support/Econo column attached to an Econo-Column Flow Adapter
and equipped with a Biologic LP system Model 2128 Fraction collector (BioRad
Laboratories, Hercules, CA). Prior to loading sample, the column was equilibrated with
200 ml of Protein A MAPSD II binding buffer, pH 9.0 (BioRad Laboratbries, Hercules,
CA) and then 150-170 mg of total protein from dialyzed SASS (in 2.5 ml) containing
0.725 g of Protein A MAPSO II binding buffer solid [calculated by Equation (a), Bio-
Rad, Hercules, CA] was loading in the column. The pH of SASS was 9.0 and the ionic
strength of the SASS approaches that of the MAPS Binding Buffer to ensure optimal
immunoglobulin binding.

gggm of Protein M_aps 11 solid" ' x volume of SASS ..(a)

100 ml of SASS

* This number is based upon using SASS dialyzed into 0.9% NaCl.

62

Non-speciﬁcally bound components were removed by pumping 600-ml of Protein
A MAPS° II binding buffer through the column until the absorbance (OD) of the efﬂuent
was less than 0.001. After washing the column, a pH gradient of citrate-phosphate buffer
system (0.1 M Citric acid and 0.2 M NazHPO4) from pH 8.0 through pH 5.5 was used to
elute the speciﬁcally bound IgG1 and IgG2 antibodies (modiﬁed from Schmerr et al.,
1985, Figure 18). A pH gradient was generated by applying sequentially; 40 m1 of pH
8.0, 20 ml each of pH 7.5 and 7.0, 10 ml each opr 6.5, 6.0, and 5.5, and 75 ml opr
5.5 citrate-phosphate buffer. The eluted samples were collected in aliquots of 1 ml each,
neutralized immediately with 1 M Tris, pH 9.0, and absorbance determined at 280 nm.
The highest OD fractions were°pooled, concentrated, desalted, and stored at -20 °C for
further analysis.
2. 2. 4 Quality test for purified antibodies

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE,
modiﬁed from Laemmli, 1970) and Western blot analyses (modiﬁed from Lee and
Shewen, 1996) were used to test the quality of puriﬁed IgG1 and IgG2 antibodies. A
molecular weight marker (Precision Plus All Blue; 10, 15, 20, 25, 37, 50, 75, 100, 150
and 250 kDa; BioRad) was loaded on every gel as a size standard. One microgram each
of pooled fraction (2“d or 3rd peak, Figure 18) was prepared by adding loading buffer at
1:2 ratio [95% (v/v) of Laemnli sample buffer (6.25 mM Tris-HCl, pH 6.8, 25.0%
glycerol, 2.0% SDS and 0.01% Bromophenol Blue, BioRad) containing 5% (v/v) B-
mercaptoethanol, (Sigma Chemical Co., St Louis, MO)]. The mixed sample was heat-
denatured at 95 °C for 4 min and immediately placed on ice prior to sample loading on

pre-casted mini gels (12.0% Tris-HCl, 10-well, 30 ul, 0.57-mm Ready Gel, BioRad,

63

Hercules, CA). Electrophoresis was conducted at constant 200 volts for 35 min using
Mini PROTEAN 11 cell electrophoresis units, Power/Pac 300 power supplies (both from
BioRad), and 10% Tris/Glycine/SDS buffer in milli-Q water, pH 8.3 (Bio-Rad) as
electrode buffer.

After electrophoresis was complete, the gel was transferred electrophoretically
onto a nitrocellulose membrane (0.45 pm; Pierce Chemical Co., Rockford, IL) at a
constant voltage (100 volts, PowerMac 300, Bio-Rad) for an hour using a Mini Transblot
apparatus (Bio-Rad Laboratories, Hercules, CA) containing transfer buffer (192 mM
Glycine and 25 mM Tris HCl (Bio-Rad Laboratories, Hercules, CA), 20.0% (v/v)
Methanol). Non-speciﬁc binding sites on the membranes were then blocked with a
SuperBlock° blocking buffer (Pierce, Rockford) for 1 hr at room temperature with 0.1%
added Polyoxythelene Sorbitan Monolaurate (Tween 20). The blocked membranes were
probed with the sheep anti-bovine HRPO-conj ugated isotype-speciﬁc detection
antibodies (IgG1 and IgG2, heavy chain speciﬁc, BETHYL Laboratories, Inc,
Montgomery, TX) for 60 min at room temperature. The membranes were then washed 6
times and substrate [SuperSignal West Pico Chemiluminescent Substrate system (Pierce)]
added for 5 min before being exposed to x-ray ﬁlms (Kodak BioMax MS; Fisher
Scientiﬁc) for 10 seconds. Films were developed (Futura 2000 E Automatic X-Ray ﬁlm
processor; Fischer Industries Inc, Geneva, IL).
2.3 Preparation of bacterial antigens

Lipopolysaccharide (E. coli serotype 0111:B4, Sigma-Aldrich Co., St. Louis,
MO) and Lipid A (monophosphoryl, S. minnesota Re 595, Sigma-Aldrich Co., St. Louis,

M0) were both dissolved by vigorous vortexing in physiological saline solution, and

0.05% Triethylamine (TEA) to reach a concentration of 1 mg/ml. Bacteria selected for
lysates preparation included Escherichia coli McDonald 487, Pseudomonas spp. , Serratia
spp., Klebsiella pneumoniae, Salmonella newport, S. typhimurium, J5 E. coli and
Staphylococcus aureus. J 5 E. coli and Staphylococcus aureus represented positive and
negative controls, respectively. S. newport and S. typhimurium were kindly donated by
Animal Health Diagnostic Laboratory, East Lansing MI. All other bacteria, except the J5
E. coli, were isolated from milk of clinical mastitis cases. As shown in Figure 19,
bacteria were grown on 5% sheep blood agar overnight at 37°C. Single colonies were
inoculated in 15 ml of Trypticase soy broth (J .T. Baker, Phillipsburg, NJ) and incubated
‘while shaking (225 rpm) for 18 hours at 37°C. Purity of culture was again determined by
culture on 5% sheep blood agar overnight at 37°C (National Mastitis Council, 1987). The
inoculums were centrifuged (8000 x g for 15 min at 4°C) to collect bacterial cell pellets,
which were then resuspended in 3.0 ml of lysing buffer containing 50 mM Tris-HCl, pH
7.5, 50 mM NaCl, 1 mM dithiothreitol (DTT), and 5% glycerol. Using a probe sonicator
(Soniﬁer), the bacterial cells were disrupted with multiple short bursts of 15 second
pulses with 15 second rests for 1-2 min and kept on ice at all times (Harlow and Lane,
1999; Promega, 1996). Resulting crude lysates were cleared of cell debris by
centrifugation (10,000 x g for 15 min at 4°C). Trichloro acetic acid (TCA) was added to
reconstitute the pellet to a ﬁnal concentration of 10% (w/v) and placed on ice for 5 min to
precipitate proteins (Promega, 1996). The TCA-treated mixture was centrifuged (12,000
x g for 2 min at 4°C) to collected the protein pellets, which were washed twice (12,000 x
g for 2 min at 4°C) and resuspended in ice cold 0.9% sterile NaCl. Protein concentrations

were determined spectrophotometrically by the speciﬁc 0D. (A. = 260, 280 and 320 nm)

65

reading and using the Warburg-Christian Concentration Assay [Model DU" 650,

BECKMAN COULTER, Inc, Fullerton, CA, Equation (b)], and stored at -—20°C until

use.
[O.D. (260.0 nm) x (757290)] + [O.D. (280.0 nm) x 1552.00] ...(b)
8000 x g
% m ‘
4°C, 15 min
Blood agar TSB
37°C, 18 hr 37°C, 18 hr Lysmg ”“9"

  

+4.

g g “ g 10,000xg, g . , §§
4°02, 2 min 4°C, 15 min Sonication

0'9% NaCl Wash 2x, Precipitation 151°C; cycle,
0.9% NaCl TC A, 5 min - mm

 

Figure 19 Scheme to illustrate the protocol used for extraction of bacterial proteins
for SDS-PAGE. Aﬁer the various bacterial species were grown on blood agar and
propagated in TBS, the cells were pelleted by centrifugation, then lysed and sonicated,
and the resulting fragments pelleted and proteins precipitated from them using TCA.
Following two centrifugation steps to pellet the precipitated proteins, proteins were
washed in normal saline twice and then suspended in the same for use in SDS-PAGE.

2.4 Assays for bacterial antigen recognition by anti-15 antibodies
2. 4. 1 SDS-PA GE

SDS-PAGE was used to separate the prepared bacterial protein, LPS, and lipid A
antigens by molecular weight Molecular weight markers (Precision Plus All Blue; 10,

15, 20, 25, 37, 50, 75, 100, 150 and 250 kDa; BioRad) were loaded on every gel as size

66

standards. Test and control antigens (5 ug each) were prepared by adding loading buffer
at a 1:2 ratio [95% (v/v) of Laemnli sample buffer (BioRad) containing 5% (v/v) [l-
mercaptoethanol, Sigma Chemical Co., St Louis, MO]. Themixed sample was heat-
denatured at 95°C for 4 min and placed on ice prior to sample loading on pre-casted Tris-
HCl polyacrylamide gels (12.0%, 160x160x0.75 cm). Electrophoresis was conducted at
a constant 15 and 30 mA on stacking gel and separating gel, respectively, for 3__ hr at
4°C (PROTEAN 11 xi cell electrophoresis units, Gibco BRL; (model 4001P
programmable power supply, Life TechnologiesTM ), and 1M Tris/Glycine/SDS buffer,
pH 8.3 as an electrode buffer. Two gels were prepared per sets of samples. One gel was
reserved for silver staining (Silver Stain Plus, BioRad) to visualize location and variety of
antigens in the samples. The other gel was used for antigen transfer on to nitrocellulose
membrane.
2. 4. 2 Silver staining procedure

After gel electrophoresis, the gel was submerged in 100 ml of Fixative Enhancer
Solution [50% methanol (J .T. Baker), 10% acetic acid, and 10% Fixative Enhancer
Concentrate (Bio-Rad, Hercules, CA)], and gently agitated (an orbital shaker, Bellco
Glass, Inc, Vineland, NJ) for 20 min at room temperature. The Fixative Enhancer
Solution was decanted from the staining vessel, which was replaced by 200 ml of 18-
megohm-cm resistivity (Milli-Q) water twice for 10 min each with gentle agitation.
About 5 min before the staining step, 5 ml each of Silver Complex Solution (NHaNO3
and AgNOg), Reduction Moderator Solution (tungstosilicic acid), and Image
Development Reagent (formaldehyde) were orderly added and stirred in 35 ml of Milli-Q

water, and followed by 50 ml of Development Accelerator Solution (Na2CO3)

67

immediately before use. The gel was stained with gentle agitation for approximately 5-
15 minutes to reach desired staining intensity. The stained gel was then gently agitated in
5% acetic acid to stop the reaction for 15 min and rinsed using 200 ml of Milli-Q water
for 5 minutes. The silver stained gel was photographed (Gel Documentation
System/Fluoro—S image Analysis System; BioRad) for data presentation.
2.4.3 Western blot analyses

Three gels were used for antigen transfer (Trans-Blot Cell with plate electrode,
BioRad) to nitrocellulose blotting membranes (0.45 am; Pierce, Rockford) at 25 constant
volts (PowerMac 300, BioRad) for 6 hr at 4°C, to visualize isotype-speciﬁc antibody
recognition of antigens by Western blot analyses. The membranes were incubated in a
blocking buffer 1 hr at room temperature, with 0.1% added Tween 20. Blocked
membranes were probed with test sera (1 mg/ml). diluted 1:1000 in sterile milli-Q water
for 60 min at room temperature, and washed six times [BupH Tris Buffer Saline (Pierce)
containing 0.05% Tween 20]. The membranes were further probed with anti-bovine
HRPO-conjugated isotype-speciﬁc detection antibodies (IgM, IgG, or IgG2, heavy chain
speciﬁc, BETHYL Laboratories, Inc, Montgomery, TX) for 60 min at room temperature.
After six more washes, the substrate was added [Super Signal West Pico
Chemiluminescent Substrate system (Pierce)] for 5 min before the membranes were
exposed to x-ray ﬁlms (Kodak BioMax MS; Fisher Scientiﬁc) for 2 min; F ilrns were
developed (F utura 2000 E Automatic X-Ray ﬁlm processor; Fischer Industries Inc,
Geneva, IL) for data presentation. This step was repeated using fetal bovine serum, J5
antisera from pre-, 3“, and 9th immunizations, and Protein A-enriched IgG1 and IgG;

antibody fractions for antigen recognition comparisons.

68

2.5 Opsonization assay
2. 5. 1 Preparation of green fluorescent protein-transformed .15 E. coli

Green ﬂuorescent protein (GFP)—transformed J5 E. coli was prepared using
electroporation as modiﬁed from Ausubel et al. (1995 ). Brieﬂy, J 5 E. coli were grown on
MacConkey agar at 37 °C overnight. Single colonies were picked and inoculated shaking
(250 rpm) in 20-ml Luria-Bertani (LB) broth at 37 °C overnight. Two milliliters of the
inoculum were added in 400-ml LB broth and incubated shaking at 37 °C until the O.D.,
= 600 m, was between 0.5 and 1.0. The bacterial pellet was recruited by centrifugation
(3,500 x g, for 20 min) and then washed using 200-ml sterile water. The pellet was
resuspended in 20 ml of 10% glycerol, centrifuged at 3500 rpm for 15 min, and the
supernatant discarded. The remaining pellet was resuspended to a ﬁnal volume of 2 ml
using 10% glycerol, which was equivalent to approximately 1-3 x 1010 bacteria/ml. Four
hundred microliters of the cell suspension was put into a 1.5-ml ice cold polypropylene
tube, and 1 u] of pGFPuv plasmid DNA added (3.3 kb, Clontech Laboratories, Inc,
PaoAlto, CA). The suspension was mixed well, and left on ice for 1 min. The
suspension was transferred to a 0.2-cm electroporation cuvette, which was placed in a
chilled safety chamber and securely sealed by the chamber slide.

The transformation was performed under condition set for 100 ohms and 500
uFD, which produced a 4-5 msec. constant pulse with a 12.5-kV/cm ﬁeld strength The
cuvette containing transformed bacteria was immediately removed from the chamber and
1 ml of SOC medium added Transformed J5 E. coli (10 ul) were inoculated shaking
overnight (~200 rpm, 37 °C) in 15 ml of LB broth containing ampicillin (100 rig/ml).

The inocultun was then grown on MacConkey agar at 37 °C overnight and stored in

69

glycerol at -20 °C until use. The transformed J5 E. coli were checked for green

ﬂorescence of colonies on UV light source (<30 seconds) and visualized using a UV light

    

microscope (Figure 20).
GFP-J55. coli as viewed under a
ﬂuorescence microscope
Phagocytosis ‘—
assay

-’ + Y4l+ @= Ire:

Antibodies
GFP JSE coli in test m
Opsonized bacteria Neutrophils Phagocytosis

Figure 20 The ﬂow cytometric assay employed in this study to measure neutrophil
phagocytosis. Brieﬂy, the J5 E. coli were transformed using a plasmid containing the green
ﬂuorescent gene (GFP-J 5 E. coli) and were used for IgG; and IgG; opsonizations prior to
incubating with neutrophils enriched from blood of three cows for various times.
Phagocytising neutrophils ﬂuoresced green upon ﬂow cytometric analysis, and thus could
be enumerated.

2.5.2 Opsonization

Regular and GFP-transformed J5 E. coli stored in glycerol at -20 °C were grown
on Luria agar at 37 °C overnight after conﬁrming culture purity. A single colony from
each culture was selected and inoculated separately in 15-ml of Luria broth containing 20

ul of ampicillin (100 ug/ml) at 37 °C and 125 rpm overnight. The inoculum was then

70

centrifuged at 700 x g, 4 °C for 10 minutes and the supernatant discarded. The bacterial
cell pellet was resuspended in 30 ml of sterile PBS, cultured, and washed two more times
in 10 ml of sterile PBS, pH 7.2. The ﬁnal cell suspension was brought to 13.0%
transmission (approximately 1 x 109 cfu/ml) with sterile PBS as determined by
spectrophotometry (Beckman DU 650 BECKMAN COULTER, Inc, Fullerton, CA) and
placed on ice until used. Opsonization treatments included no antibody [GFP-J5 E. coli
(2.35 x105 cells) only] and GFP- J5 E. coli opsonized with 370 pg each of Protein A-
enriched serum IgG1 or IgG;, antibody (Figure 21) or bovine serum albumin as a protein
control [determined by the Warburg-Christian Concentration Assay
spectrophotometrically (Model DU° 650, BECKMAN COULTER, Inc, Fullerton, CA)].
Opsonization was performed in 5-ml round-bottomed tubes (Beckton Dickinson) by
mixing on an orbital shaker 125 rpm; (Bellco Glass, Inc, Vineland, NJ) at room
temperature for 20 min. °
2. 6 Neutrophil isolation

Neutrophils were harvested from 20—ml blood samples collected from tail veins of
three healthy multiparous cows (mid-lactation Holsteins) into vacutainers containing
Acid Citrate Dextrose (ACD, Becton & Dickinson). Blood samples were immediately
placed on ice and transported to our laboratory. After centrifugation (approximately 700
x g for 20 min at 4 °C), the plasma and top 1/3 of the red cell pack was aspirated and
discarded. The remaining red cell pack (containing neutrophils) in each tube was
reconstituted in 15 ml of sterile PBS, pH 7.2 and centrifuged at 700 x g, 4 °C for 10 min.
Supematants were aspirated and 12 ml of ice cold hypotonic lysing solution was added

for every 3-4 ml of cells to for 1__ min to lyse red blood cells. Immediately after lysis, 6

71

ml of ice cold Hypertonic restoring solution was added to each tube. The tubes were
swirled gently and centrifuged at 500 x g, 4 °C for 10 min to pellet remaining leukocytes
(2 85% neutrophils, determined ﬂow cytometrically). After aspiration of supematants,
cell pellets were resuspended in 2 ml of sterile PBS/tube, the number of cells per ml
determined using an electronic cell counter (Z1 counter, Particle counter, Beckman
Coulter), cells resuspended to 2 x 107 cells/ml in sterile PBS, and kept on ice for
immediate use in the phagocytosis assay.
2. 7 Phagocytosis assay and ﬂow cytometric analysis of phagocytosing neutrophils

The various opsonized bacteria (total volume of 100 ill/well with 2.35 x 107
bacteria/well) were added at 8:1 ratio to prepared bovine neutrophils (2.9 x106 cells/well)
and allowed to undergo phagocytosis for varying hours in on 96-well ﬂat bottom tissue
culture plates (Costar, Corning Incorporated, Corning, NY). Incubations proceeded in an
orbital shaker (Forma Scientiﬁc) at 120 rpm, 39 °C to standardize bacteria-neutrophil
contact time across wells for 60, 90 120, and 150 min. Aﬂer incubations, the contents of
each well were treated with 20 ul of Trypsin EDTA (~10%, GibcoBRL, Life
Technologies) for 10 min at room temperature to disengage surface-bound bacteria from
neutrophils before ﬂow cytometric data acquisition. The contents of each well were then
analyzed in duplicate using ﬂuorescence activated ﬂow cytometry (F ACSCalibur ﬂow
cytometer; Becton Dickinson). For ﬂow cytometric data acquisition, the green
ﬂuorescence detector (FL-1) and the side scatter (or granularity) detector (SSC) were set
against normal neutrophils (i.e., ones that were not phagocytosing and thus did not
ﬂuoresce green). In this situation, all cells fell in the upper leﬂ quadrant of SSC-FL-l

density dot plots. Background ﬂuorescence markers were placed on dot plots based on

72

ﬂuorescence of neutrophils exposed to only regular J 5 E. coli (with no opsonization).
Therefore, any neutrophils containing GFP E. coli and ﬂuorescing brighter green than
background were considered as phagocytosing cells, and were observed to lie in the
upper right quadrant on SSC-FL-l density dot plots (Figure 20). Thus, phagocytosis
data were recorded as % phagocytosing neutrophils based on their counts out of a total of

5,000 neutrophils per sample counted.

2.8 Statistical analysis

Repeated measures analysis of variance (AN OVA) was used to analyze the
percentage of phagocytosing neutrophil data set. The analysis was carried out utilizing
the MIXED procedure of SAS software (SAS Institute, Cary, NC). Overall signiﬁcance
of effects included in the model (immunization numbers, opsonins, and the interaction of
immunization number and opsonins) was assessed by the Type III F-test. Statistical
signiﬁcance of adjusted mean comparisons of phagocytosing neutrophil (%) over the
length of incubation by each opsonin declared relevant by the t-test was determined by
the Tukey-Kramer multiple comparison procedure. Also, the differences of LSMEAN
phagocytosing neutrophil (%) between each opsonin and negative control in each
incubation time were determined with the Dunnett adjustment factor. The signiﬁcance

level was set at P = 0.05.
3. Results

Using SAS-PAGE followed by silver staining, we were able to visualize multiple

complex protein, LPS, and Lipid A antigens from heterologous Gram-negative bacteria

73

(Figure 21). Our Western blot analyses demonstrated Gram-negative antigen recognition

by isotype speciﬁc serum antibodies. At least two groups of bacterial antigens of

approximawa 8.5-10.5 and 34-39 kDa were recognized, predominantly by IgG1 and IgG2

antibodies (Figure 22-24).

M12345678910

37 kDa—b

lOkDa —>

M marker 5 Serratia spp.
1 J55. coli 6 K pneumoniae
2 S. neuport 7 Pseudomonasspp.

3 S. ophimurium 8 S. aureus
4 E. coli 487 9 LPS
10 Lipid A

 

Figure 21 Silver stain of Gram-
negative (lanes 1-7) and S. aureus
(lane 8) bacterial proteins prepared
by sonication and protein
precipitation, and of commercial
LPS (lane 9) and Lipid A (lane 10)
separated by SDS-PAGE. A
molecular weight marker (kDa) is
in lane M. Bacterial proteins (5
uglane) were separated by SDS-
PAGE (12% gel). The bacterial
antigens included J5 E. coli (lane
1), S. newport (lane 2), S.
whimurium (lane 3), E. coli 487
(lane 4), Serratia spp. (lane 5),
Klebsiella spp. (lane 6),
Pseudomonas spp. (lane 7), S.
aureus (lane 8), LPS (011:B4, lane
9), and Lipid A (S. minnestoa Re
mutant, lane 10). The highlighted
boxes around bands in the 10- and
37-kDa regions (red arrows) were
the groups of bacterial antigens
recognized by serum antibodies
from J5 vaccinated steers in
subsequent Western blot analyses
(Figure 22-24).

In the 34-39 kDa molecular mass range, the pattern of antigen recognition by
antibodies was almost identical across the three antibody isotypes, except that the
recognition by IgG2 antibodies was more intense than for IgM and IgG1 (Figure 22).
Similar results were observed when different sources of serum antibodies were used,
including those in fetal bovine serum and in J5 antisera from pre-, 3rd and 9th
immunizations. For most bacteria, only one molecular mass size of this large antigen was
recognized by the antibodies for each bacterial species. In contrast, two antigens from
Salmonella bacteria appeared to be recognized by the serum antibodies, with one protein
of ~36 kDa being the same antigen of this size class recognized in Klebsiella. The
antigen recognized by serum antibodies in E. coli 487 and Pseudomonas seemed to be the
same size as that recognized in J5 E. coli (~35 kDa). There was apparent isotype speciﬁc
recognition of antigens in Serratia spp. , where IgG1 bound with a 34 kDa protein and
IgG2 bound with a 39 kDa protein (Figure 22 and 24). Surprisingly, we observed the
same binding pattern of IgG1 and IgG2 antibodies from fetal bovine serum to these 34-39
kDa proteins of gram-negative bacteria as from the preimmune, immune, and
hyperimmune sera from the 15 vaccinated steers. Therefore, the antibodies recognizing
these proteins may be “natural” antibodies.

In contrast to the 34-39 kDa proteins, the intensity of antibody binding to the 8-
' 10.5 kDa proteins increased with increasing number of 15 immunizations and. was
stronger for IgG1 than IgG2 antibodies (Figure 23). IgM recognition of this group of
small proteins was low. The IgG1 (and to a lesser degree IgG2) antibodies of
hyperimmune serum also recognized LPS in pure form and in the bacterial lysates

(Figures 23 and 24). Protein A-enriched G IgG1 than IgG2 antibodies had the same

75

12345678 2345678

 

 

(:1) Fetal bovine serum (b) Pro-immune serum

2345678

_s-. n_ [36311325 2391;,- .5 k133i.“ ﬁx: 2M3: ..bﬂn..§'u( 1.1-3.423%
.v.-. . >115. 41,"

1 . . . . . ...-“.f'ini’ _ _5.
'jj— 5275' 4‘6me %°”.""°°' ' ' ‘-, ‘(4o 9 . ‘3‘}; IgM
. _ 2:42.53 ' 'iLstiremfgj-gtge 7:33;; «1‘; ‘4”st “,4;ng “mg
-. : 2.4;: ... eIgG
° " It! ,‘I. -' .-' v°°°~::.'°.-' _ .°. 1°; fir-J . ° 9.3:;ng l
"|..

    

t . .J—P‘i ‘1".‘3 “E“ 7": V7 7""- “$705? f4". “35°" m#,;xr7"wrvswre “I:

(c) 3“1 immune serum ((1) 99' immune serum

- 37 kDa Molecular weight marker

Figure 22 Western blot analyses of Gram-negative bacterial antigens in the
molecular mass ranges of 34-39 kDa, recognized by IgM, IgG1 and IgG2
antibodies in various sera from non-immunized and 15 E. coli immunized cattle.
Gram-negative bacterial proteins (5 ug/ lane) were separated by SDS-PAGE (12%
gel), and transferred to nitrocellulose membranes. The membranes were ﬁrst
probed with fetal bovine serum (a) or antisera from pre-immunization (b) or
following the 3rd immunization (c), and 9th immunization (d) of steers with 15
bacterin. The membranes were further probed with horseradish peroxidase-
conjugated sheep anti-bovine IgM, IgG1, and IgG2, heavy chain speciﬁc detection
antibodies (shown in each panel). Test antigens included 15 E. coli (lane 1), S.
newport (lane 2), S. whimurium (lane 3), E. coli 487 (lane 4), Serratia spp. (lane
5), K. pneumoniae, (lane 6), Pseudomonas spp. (lane7), and S. aureus (lane 8).
The black arrows indicate the position of a 37-kDa molecular weight marker
included in each membrane but not shown in the ﬁgure.

76

12345678910 12345678910

   

" no. :—':M05
' and, -—Iecz
(a) Fetal bovine serum (b) Pre-immune serum

12345678910 12345678910

      
 

‘r :v”.-,’-;:.
I?" ‘95};1‘”, gir’kgnivges".1‘t‘i_nsﬂ*ﬁ4iﬂ

'. i, ([1302 .

mars eve-ainltvxvt is} M? sax;- ) ctr-3- 9‘?) ru.‘ ‘:-'::'r

 

(c) 3rd immune serum ((1) 9“I immune serum

- 10 kDa Molecular weight marker

Figure 23 Western blot analyses of Gram-negative bacteria] antigens in the
molecular mass ranges of 8 to 10.5 kDa, recognized by IgM, IgG1 and IgG2
antibodies in various sera from non-immunized and 15 E. coli immunized
cattle. Gram-negative bacterial proteins, LPS, and Lipid A (5 rig/lane) were
separated by SDS- PAGE (12% gel), and transferred to nitrocellulose
membranes. The membranes were ﬁrst probed with fetal bovine serum (a)
or antisera from pre-immunization (b) or following the 3rd immunization (c),
and 9th immunization (d) of steers with 15 E. coli bacterin The membranes

were further probed with horseradish peroxidase-conj ugated sheep anti-
bovine IgM, IgG1, and IgG2, heavy chain speciﬁc detection antibodies
(shown in each panel) . Test antigens included 15 E. coli (lane 1), S. newport
(lane 2), S. typhimurium (lane 3), E. coli 487 (lane 4), Serratia spp. (lane 5),
K. pneumoniae, (lane 6), Pseudomonas spp. (lane7), S. aureus (lane 8), LPS
(0112B4, lane 9), and Lipid A (S. minnestoa Re mutant, lane 10). The black
arrows indicate the position of a 37-kDa molecular weight marker.

77

12345678910 12345678910

37 -0
kDa

 

12345678910 12345678910
18—
kDa
10-
kDa

IgGI from hyperimmune serum IgG2 from hyperimmune serum

   

Figure 24 Western blot analyses of Gram-negative bacterial antigens in the molecular
mass ranges of 8 to 10.5 kDa and 34 to 39 kDa, recognized by Protein A enriched
IgG1 and IgG2 antibodies from 15 E. coli hyperimmune serum. Gram-negative
bacterial antigens (5 rig/lane) were separated by SDS-PAGE (12% gel), and
transferred to nitrocellulose membranes. The membranes were ﬁrst probed with
enriched IgG1 (a) or IgG; (b) antibodies afﬁnity puriﬁed from 15 hyperimmune serum
(9th immunization) from ﬁve Holstein steers. The membranes were further probed
with horseradish peroxidase-conjugated sheep anti-bovine IgG1 (a) or IgG; (b) heavy
chain speciﬁc detection antibodies. Test antigens included 15 E. coli (lane 1), S.
newport (lane 2), S. whimurium (lane 3), E. coli 487 (lane 4), Serratia spp. (lane 5),
K. pneumoniae, (lane 6), Pseudomonas spp. (lane7), S. aureus (lane 8), LPS (0111:B4,
lane 9), and Lipid A (S. minnestoa Re mutant, lane 10). The black arrows indicate the
position of the 10- and 37-kDa molecular weight markers.

78

antigen recognition patterns as antibodies in sera, but recognized antigens in Serratia and
Klebsiella more weakly than antigens in other Gram-negative bacteria Neither puriﬁed
IgG; or IgG; in any of the test anti-sera recognized antigens in S. aureus (the assay
negative control).

Protein A afﬁnity puriﬁcation using the pH gradient technique was successful in
isolating nearly pure IgG1 and IgG2 antibodies from 15 hyperimmune serum (Figure 18).
Western blot analysis demonstrated single 50 kDa heavy chains of the antibody
molecules from each immunoglobulin fraction collected from the protein A column
(Figure 25). Minor contamination of IgG1 antibody was found in the IgG2 fraction based
on the Western blot analysis; however, no IgG2 antibodies were detected in the IgG1

fraction (Figure 25).

Pooled elution fractions from
peaks 2 and 3 in Figure 17.
Peak Peak Peak Peak
2 3 2 3

#m1:.wLNZI-'LH;HEG,. ~ .-‘~1‘. 5.-. 4.th Limtﬁmﬁﬁlﬂ
r. I 4 r . 1 . ° ‘ a a ,a'. J' 3° '
,, .-. . . s e ..«154 .e n;s~vr.... I

 
    

.. .I-A.'°ly"’ l'
" ' ‘s ,1: °
,, .

’RVW ° '9'“ °.° °;°‘

IgG1 detection IgG2 detection

Figure 25 Western blot analysis of IgG1 (peak 2) and IgG2 (peak 3) fractions affmity
puriﬁed from 15 hyperimmune serum using a Protein A column and a pH gradient
technique (as shown in Figure 17). One microgram each of pooled fraction was
resolved by SDS-PAGE, transferred onto nitrocellulose membrane, and detected with
heavy chain speciﬁc Horseradish peroxidase-conjugated sheep anti-bovine IgG1 (leﬁ
panel), and 1ng (right panel) antibodies. Arrow indicates a 50-kDa molecular weight
protein, which is the heavy chain of IgG1 and IgG2.

The enriched IgG1 and IgG; antibodies were compared against each other and
against BSA and no opsonization as opsonins for enhancing neutrophil phagocytosis of

E. coli. Percentage of neutrophils phagocytosing E. coli opsonized with IgG2 antibodies

79

was higher (P < 0.01) than IgG.-BSA-, and no treatment E. col at all incubation times
studied. In fact, IgG1 antibodies were ineffective in this assay (Figures 26 and 27). The
untreated, BSA-, and IgG1- treated cultures had 9% phagocytosing neutrophils while the
IgG; cultures had 24% phagocytosing neutrophils by 150 minutes of incubation (Figure
26 a,b). There was no incubation time effect on percentage of phagocytosing neutrophils
for the no-, BSA-, or IgG1-opsonin treatments. However, phagocytosis was signiﬁcantly
higher for IgG2-opsonized bacteria at 90, 120, and 150 minutes of incubation compared

to 60 minutes (P < 0.05; Figure 27).

4. Discussion

In this study, we demonstrated antibody maturation in response to repeated
exposure of the bovine immune system to 15 E. coli bacterin. Maturation was observed as
increased recognition through Fab regions of small (8-10.5 kda) proteins in a variety of
Gram-negative bacteria, increased isotype switching from IgM to IgG1 or IgG;
antibodies, and improved functional activities of the antibodies for neutrophil
phagocytosis as part of the serum response to hyperimmunization-induced mature IgG2
antibodies.

Generally, maturation of the antibody response is the process of development of
highly antigen-speciﬁc antibodies by T-cell-dependent-antigen-activated B cells. This
development includes afﬁnity maturation of the Fab regions of antibodies achieved by
hypermutaton and gene conversion of the variable regions of heavy and light chain gene
segments, as well as maturation of the Fc regions of the molecule through constant region

gene segment switching (isotype switching) (Bonhomme et al., 2000; Butler, 1998;

80

Figure 26 Flow cytometric analysis of bovine blood neutrophils showed that the
cells phagocytosed green fluorescent protein (GFP)—transfromed 15 E. coli better
when the bacteria were opsonized with enriched IgG2 versus IgG1. The enriched
immunoglobulins were from pooled sera of 5 Holstein steers hyperimmunized with
a commercial 15 E. coli bacterin. Neutrophils were from blood collected from the
tail veins of 3 mid-lactation Holstein cows. GFP-transformed 15 E. coli were
incubated with identical amounts (370 11g) of IgG1 or IgG2 for 20 minutes at room
temperature in an orbital shaker (120 rpm) prior to being added to duplicate
cultures of blood neutrophils at 8:1 (bacteriazneutrophils) for 150 minutes at 39°C
and 120 rpm. Panel (a) shows representative ﬂow cytometric density dot plots for
neutrophils of one cow after 150 minutes of incubation with IgG1-opsonized (left
side of panel) and I ng-opsonized (right side of panel) GFP-15 E. coli. The Yeaxes
are side scatter (granularity) of the neutrophils while the X-axes are intensity of
green ﬂuorescence associated with the neutrophils. Colored dots lying in the upper
right quadrants of these plots are individual neutrophils that have phagocytosed
GFP-15 E. coli and thus fluoresced bright green. These represented 9% of all
neutrophils in the IgGl plot and 24% of all neutrophils in the IgG2 plot. Data in
panel (b) show mean (: SEM) phagocytising neutrophils for duplicate neutrophil
cultures from the 3 cows at 150 minutes of incubation with bacteria the were
opsonized with bovine serum albumin (BSA; white bar), IgGl (light gray bar),
IgG2 (dark gray bar), or nothing (stippled bar). *The mean of phagocytising
neutrophils was signiﬁcantly higher for IgG2 Opsonization than for all other
opsonization treatments at P < 0.01. ‘

81

 

(a) 150 minute incubation

 

 

  
   
 

 
   
 

IgGl opsonization IgG2 opsonization
i Shaft? i
: ,_ _- .‘ ,--‘.-; J; ‘7.
C611 . ' ° 3 -° 9% green .'°.-°'T51°3 '°°°°’i°~'i;' 24% green

. . : ' .23.» . '
_ _‘ -.° 3;:- ﬂuorescent

 
    

 

 

 

 

 

 

 

 

 

Count _ . :- .
‘ 2" neutrophils " ‘ is}, -neutr0phlls
Green Fluorescence Intensity
(b) 150 Minute Incubation
30 _ D BSA I IgG1 I IgG2 None
*
25 -
20 -
% Phago-
cytosis 15 °
10 5
5 .1
0 - ' I ‘° °' 1 '
BSA IgGl IgG2 None

Type of Opsonin

82

25

20
Percent
Phagocytosing 15
Neutrophils

10

 

60 90 120 150

Time of incubation (min)

Figure 27 Mean neutrophil phagocytosis of GFP-transformed J5 E. coli in the
presence of various opsonins; IgG1 (E) and IgG2 ( ) (both puriﬁed from 15
hyperimmune serum using Protein A afﬁnity chromatography); bovine serum
albumin ( E; ); and no opsonin (D ). Data represent the percentage of
neutrophils ﬂuorescing green at each incubation time, determined from density
dot plots as shown in Figure 26a. * Indicates signiﬁcant differences within
incubation time (P < 0.01). 1: Indicates signiﬁcant difference from 60 minutes
(P < 0.05).

83

Kaattari et al., 2002). This process is an integral part of effective adaptive immune
responses and occurs in the germinal centers of secondary lymphoid organs, including the
spleen and lymph nodes (Bonhomme et al., 2000; Butler, 1998; Kaattari et al., 2002; Liu
etaL,1992)

Isotype switching is largely regulated by T-cell-driven cytokines (i.e. IL-4 for
IgG1 & IgE; IF N-y for IgG; in cattle) and occurs when helper T cells ligate their plasma
membrane bound CD4OL molecules with CD40 on the B cells (Butler, 1998; Estes,
1996). CD40L-Cd40 binding sends signals to the B cells that they are to acquire a CDS'
phenotype capable of isotype switching from IgM to IgG] or IgG2 (depending on the
cytokines produced by the T cells). The maturation process ultimately improves the
afﬁnity and functional diversity of antibodies for bacterial antigens resulting in their
clearance (e. g. by neutrophil phagocytosis). .F or example, VDJ region editing modiﬁes
the amino acid sequences of the hypervariable regions of Fab, ultimately improving the
afﬁnity of the antibody for its cognate antigen (Butler, 1998). Isotype switching enables
antibodies with the same or afﬁnity matured Fab regions to carry out avariety of
biological functions, thus expanding the capabilities of the humoral immune response
(Estes, 1996). For example, IgM is a strong ﬁxer of complement, IgG1 is an effective
neutralizing antibody, and, in cattle, Ing is a key opsonin for enhanced neutrophil
phagocytosis of the infecting bacteria (Figure 6; Table 2). Hence, maturation of the
antibody response improves its overall effectiveness in clearing infections and toxins that
cause clinical disease.

In this study, numerous antigens were extracted from a variety of Gram-negative

bacteria with SDS-PAGE and silver stained. We also separated species of LPS on

84

acrylamide gels and visualized the ladders by silver staining, as previously reported
(Cross et al., 1989; Tsai and Frasch, 1982). We had shown in Chapter 2, using ELISA,
that antibodies in J5 hyperimmune serum were cross reactiVe with heterologous Gram-
negative bacteria, but had low cross-reactivity with LPS and non-reactive with Lipid A.
We expanded results of Chapter 2 in this study by showing which bacterial antigens were
bound by the cross-reactive antibodies and by showing that such antigen recognition by
IgG2 antibodies facilitated enhanced E. coli phagocytosis by bovine blood neutrophils.
Results of our Western blotting experiments suggested that bacterial antigen
recognition by serum antibodies may possess isotype speciﬁcity. IgG1 and IgG2 bound
distinctively to differently sized Gram-negative protein antigens. At least two
predominant groups of molecular mass proteins, ranging from of 8-10.5 and 34-39 kDa,
were recognized by the IgG1 and IgG2 antibodies. Based on the molecular masses of
these antibody-recognized antigens, we suspect that they might be outer membrane
proteins of the Gram-negative bacterial cell wall, including OmpA, murine lipoprotein
(MLP), and perhaps PAL. Evidence from studies employing human and mouse J5
antisera suggested that at least three conserved Gram-negative OMPs may be detected by
the serum antibodies. These included OmpA (35 or 37 kDa), PAL (18 or 24kDa), and
MLP (5-9 kDa) (Helhnan et al., 1997; 1999, 2000a; 2001). OMPs are thought to
contribute to the structural integrity of the outer membrane layer of Gram-negative
bacteria (Braun and Bosch, 1972; Lazzaroni and Portalier, 1992), and would be clearly
accessible for immune system recognition during immunization with J5 E. coli.
Interestingly, OmpA from Klebsiella pneumoniae is used as a carrier protein for some

polysaccharide antigens because these are typically poor immunogens. The OmpA

85

dramatically increases immunogenicity of the polysaccharide PS and induces IgG
antibodies during vaccination, which protects mice from lethal bacterial challenges
(Libon et al., 2002; Rauly et a., 1999). Therefore, if the prOteins recognized by the IgG
and IgG2 antibodies in our J5 hyperimmune serum during Western blotting were such
OMPs, these antibodies may be protective against a wide variety of gram-negative
bacteria, including those that cause mastitis. However, the ELISA data in Chapter 2
indicated that multiple doses of J5 vaccine are required to elicit high titers of these
putatively protective IgG1 and IgG2 anti-colifonn antibodies.

In addition, the 34-39 kDa antigens identiﬁed on our Western blots may not have
been the target antigens that increased serum anti-J 5 IgM, IgG; or IgG; antibodies in our
hyperimmunized steers (Chapter 2) because an identical pattern of recognition of these
proteins was observed for the IgG; antibody present in very different sources of test
samples, including fetal bovine serum and J5 antisera raised from, pre-, 3“, and 9th
immunizations. These observations were confusing because the fetal bovine serum used
in Western blotting was the same as that used for the negative control (no antibody) for
ELISAs in Chapter 2. It is possible that our ELISA was not sensitive enough to detect
these antibodies recognizing the 34-39kDa antigens, or large antigens that are “hidden”
from antibody recognition in phenol-killed whole bacteria, but are exposed in the lysates
of these same bacteria, were used for the Western blot analysis. Still, it is difficult to
reconcile that such high levels of antigen speciﬁc IgG2 antibodies are present in fetal
bovine serum. The dogma is that maternal IgG does not cross the placenta in ruminants
because this tissue is syndesmochorial with the uterine epithelium being maintained

throughout the entire pregnancy. As a result, it is believed that no prenatal transfer of

86

maternal antibodies to the fetus occurs. Indeed, using the relatively insensitive RID
assay, calves have been shown to be born agammaglobulinaemic or signiﬁcantly
gammaglobulinaemic (Watson et al., 1994). However, our sensitive Western blot
analysis suggests that this is not true and that the bovine fetus has serum IgG1, and
especially IgG2 that cross-reacts with large proteins in heterologous Gram-negative
bacteria. Unfortunately, we cannot ﬁnd any solid evidence to explain our data, which
raises the possibility that in cattle, IgG2 antibodies recognizing some proteins of Gram-
negative bacteria are transferred through the placenta. Further investigation of this
possibility is required.

The results of this study also suggest that the speciﬁc antigen recognition of IgG1
and IgG2 antibodies contained in serum of J5 vaccinated steers improved with increasing
. number of doses. For example, IgG; antibody gradually recognized antigens in the 8-

. 1 0.5 kDa range across Gram-negative bacterial species. Little or no recognition of these
antigens was observed when fetal bovine serum and sera from pre-immune, and 3rd
immunization were used. This suggested prolonged antigen exposure may be required
for bovine T cells and B cells to speciﬁcally recognize these particular antigens in the J5
E. coli bacterin. If so, it may be possible to improve the J5 vaccine by including an
excess of these 8-lO.5 kDa Gram-negative bacterial proteins, which may be OMPs. Thius
possibility certainly warrants further investigation. It is also possible that addition of
detoxiﬁed LPS may improve the efﬁcacy of J5 vaccines to prevent coliform diseases
including septic shock, because the IgG1 antibodies in our hyperimmune steer serum

speciﬁcally recognized LPS during Western blotting.

87

The patterns of antigen recognition by IgG1 and Ing serum antibodies observed
in this study might be explained through the types of antigen presentation of Gram-
negative bacteria by B cells and other antigen presenting cells present in lymph nodes
that drained the vaccination site. In the case of I ng, it is most likely that the peptide
antigens were processed and presented on surface MHC class II molecules (reviewed by
Burton and Erskine, 2003). If true, MHC H presentation of the antigens to cognitive TH]
cells could have activated them to produce IFN-(, which would cause responsive B cells
to switch production of IgM or IgG1 antibodies to IgG2 (Figure 7). As exposure to the
same antigen increased, afﬁnity maturation also might have occurred in the IgG;

antibodies, explaining the increasing recognition of 8—10.5 kDa bacterial proteins by
these molecules with increasing number of J 5 immunizations. Thus, our
hyperimmunization protocol may have shiﬁed the antibody response to a predominately
TH] response with steadily increasing levels of serum IgG2 that had increasing afﬁnity for
the small bacterial proteins. 1

In contrast, speciﬁc LPS recognition by serum IgG antibodies might be explained
through LPS antigen-presentation to TH2 cells via CD1 molecules on responding B cells.
(Howard et al., 1993a; 1993b). Recent studies in human and mice reported that CD1
molecules act as a separate lineage of antigen-presenting molecules and present
predominantly non-peptide antigens, including microbial carbohydrate fractions and
lipid-containing antigens, especially oligosaccharide, glycolipid, and glycopeptide
antigens from a variety of pathogens (Maher and Kronenberg, 1997; reviewed by Porcelli
et al., 1996). Cognate THZ cells responding to CD1-presented non-speciﬁc peptide

antigens induce primarily IgM and IgG1 antibody responses from the antigen-sensitized B

88

cells (reviewed by Burton and Erskine, 2003). Thus, LPS present in the J5 vaccine may
have induced strong serum IgG1 responses in the hyperimmunized steers of this study
through CD1 presentation to THZ cells. If true, the J5 vaccine may promote a strong anti-
LPS THZ response in addition to a THI response.

Finally, we were curious to know if the THI-dependent IgG2 antibodies or the
THZ-dependent IgG1 antibodies in our bovine hyperimmune serum were capable of
opsonizing J5 E. coli for enhanced phagocytosis by neutrophils. With Protein A-puriﬁed
IgG1 and IgG; from hyperimmune serum we clearly demonstrated that only the IgG;
fraction possessed direct opsonizing activity for bovine neutrophils. This is in agreement
with observations of others (Miller, et al., 1988; Butler, 1998), and is logical because
bovine neutrophils express receptors that speciﬁcally bind the F c region of pathogen-

bound IgG2, but not IgG1 (reviewed by Burton and Erskine, 2003).

5. Conclusion

The bovine antibody response elicited by repeated exposure to J5 E. coli bacterins
can mature into highly effective THI-dependent IgG; antibody responses when given
suﬁicient antigen stimulation and time to support afﬁnity maturation and isotype
switching. As a result, matured IgG2 antibodies increase their biological activities by
developing higher afﬁnity antigen recognition and promoting bacterial clearance through

neutrophil phagocytosis.

89

 

CHAPTER 4

SUMMARY AND CONCLUSIONS

The current study elucidated new information about the basic biology of serum
antibodies induced by vaccinating cattle with J5 bacterin. A three-dose protocol of J5
immunization, recommended by the manufacturer, yielded unimpressive serum IgG; and
IgG; antibody responses, as detected in ELISA using whole cell (killed) J5 E. coli as
antigen. We determined that anti-J5 antibody titers and cross-reactivity with other Gram-
negative bacteria for three key antibody isotypes (IgM, IgG], and IgG2), increased
signiﬁcantly only after 6 doses of the vaccine were administered. In our protocol, J5
bacterin was administered every two weeks. Therefore, our study suggests that frequent
and continuing J5 immunization in dairy cattle may be required to sustain high levels of
serum IgM, IgG1, and IgG2 antibodies that strongly cross-react against heterologous
Gram-negative bacteria, including those that cause colifonn mastitis.

We also determined that the IgG; and IgG; antibodies of immune serum bind
distinctively to different sizes of Gram-negative bacterial protein antigens and LPS. At
least two predominant groups of proteins, of molecular masses 8-10.5 and 34-39 kDa,
were recognized by the IgG1 and IgG2 antibodies. IgG2 predominately recognized the
larger proteins. According to basic immunology textbooks, we expected to see a trend of
enhanced bacterial antigen recognition by maturing IgG2 antibodies as additional vaccine
doses were administered. However, this was not true for our serum I ng antibodies that

recognized the larger 34-39 kDa antigens, because the banding patterns on Western blots

were the same in this protein size category regardless if the IgG; was from fetal bovine
serum, pre-immune, 3rd immunization, or 9th immunization serum from J5 vaccinated
steers. In contrast, the 8- 10.5 kDa protein antigens were observed as these increasingly
immune sera were applied to the Western blots. Therefore, these small antigens may be
excellent candidates for improving serum IgG; antibody responses in vaccinated cattle
because they indicated that the IgG; response of J 5 immunized steers improved with
increasing doses of vaccine. If true, the 8—10.5 kDa proteins in heterologous Gram-
negative bacteria increasingly recognized by IgG2 antibodies in J5 immune serum could
be used to improve coliform mastitis vaccines.

This would seem highly relevant because only the Protein A-enriched IgG2
fraction of our hyperimmune steer serum enhanced neutrophil phagocytosis of opsonized
E. coli. IgG1 opsonization was totally ineffective, inducing no greater neutrophil
phagocytosis than was achieved using no opsonization or BSA opsonization of the E.
coli. Our phagocytosis results were not unexpected because Miller et al. (1988) showed
that IgG2 antibodies were the key opsonizing antibodies for bovine neutrophil
phagocytosis of S. aureus, and bovine neutrophils express F c receptors for IgG;
(reviewed by Burton and Erskine, 2003). Therefore, if the increasingly mature IgG;
antibodies in J5 vaccinated steers recognized the same 8-10.5 kDa proteins on E. coli
during opsonization for enhanced neutrophil phagocytosis as they did on the Western
blots, these proteins should make ideal candidate antigens for improvements to the J5
vaccine. While our Protein A puriﬁed IgG1 antibodies did not affect neutrophil
phagocytosis of E. coli, these antibodies did recognize LPS on Western blots. If the same

occurs in vivo, the current J5 vaccine may be effective in eliciting protection against the

91

systemic shock that accompanies severe coliform mastitis, as long as enough doses (at
least 6) of the vaccine are administered. Because Ig01 is produced as a result of TH2-
dominated immune responses, the exposed LPS in the J5 bacterin may have been
recognized by surface IgM on naive B cells. Subsequent processing and presentation in
CD1 molecules to cognate THZ cells, and ensuing cytokine responses may have triggered
IgM to IgG1 isotype switching in the original B cells. If so, it may be beneﬁcial to
improve the current J5 vaccine by incorporating additional, detoxiﬁed LPS to stimulate
strong anti-LPS IgG1 responses. The modest IgG1 responses elicited by three doses of
the current J 5 vaccine may be responsible for the reduced severity of colifonn mastitis
observed in some immunized cows. However, as for the IgG2 response, the vaccine
requires improvement to elicit strong serum IgG responses in fewer administered doses.
In conclusion, we caused an increasingly mature serum antibody response
in Holstein steers by repeated exposure of their immune systems to antigens in the J5
bacterin. Resulting antibodies increasingly recognized a variety of whole Gram-negative
bacteria in ELISA assays, 8;10.5 kDa protein antigens (IgG2) and LPS (IgG1) from these
heterologous Gram-negative bacteria in Western blot analyses, and effectively opsonized
J5 E. coli for enhanced clearance of the bacteria by bovine neutrophils (IgG2 only).
However, at least 6 doses of the currently available J5 bacterin were required to mature
these IgGl and IgG2 antibodies and increase their blood levels above'pre-immunization
levels. Therefore, if the beneﬁts of these serum antibodies are to be realized in the ﬁeld
(protection against coliform mastitis for example), the current J 5 vaccine needs to be
improved such that one to two doses elicits the same Ig61 and Ing responses we

observed after six to eight doses of the current vaccine. We have identiﬁed two classes

92

of antigens worthy of pursuit in vaccine improvement, LPS (for IgG) and a cluster of
small (8 — 10.5 kDa) proteins present in lysates of all Gram-negative bacteria we tested
(for IgG2). Isolation, identiﬁcation, and immunogenicity testing of these antigens in
lactating dairy cows will be the next major steps of our continuing research on vaccines

to prevent colifonn mastitis.

93

DATA FOR FIGURES IN CHAPTERS 2 AND 3

CHAPTER 5

APPENDIX

Table 3 Mean anti-15 E. coli antibody response elicited by J5 E. coli bacterin from sera

collected from ﬁve Holstein steers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mean optical densitykmgso ..., :1: S.E.M.

Serum samples IgM IgG1 IgG2

Pre-immunized 0.137 1 0.021 0.337 1 0.100 0.045 1 0.051
1st 0.157 1 0.022 0.588 1 0.113 0.083 1 0.056
2“d 0.229 1 0.022 0.612 1 0.111 0.092 1 0053
3rd 0.220 1 0.015 0.573 1 0.110 0.100 1 0.051
4th 0.225 1 0.019 0.552 1 0.109 0.108 1 0.051 '.
5* 0.235 1 0.027 0.727 1 0.108 0.151 1 0.051
6'1‘ 0.259 1 0.029 0.340 1 0.106 0.177 1 0.050
7‘h 0.239 1 0.024 0.882 1 0.103 0.239 1 0.052
8‘11 0.315 1 0.023 0.975 1 0.102 0.270 1 0.051
9th 0.340 1 0.035 0.947 1 0.101 0.262 1 0.051
10th 0.399 1 0.041 0.982 1 0.100 0.294 1 0.051
11* 0.389 1 0.040 0.964 1 0.101 0.335 1 0.053
12?h 0.457 1 0.046 0.811 1 0.113 0.282 1 0.073

 

 

 

 

94

 

Table 4 Mean anti-J5 E. coli antibody endpoint titers of selected serum samples collected

from ﬁve Holstein steers

 

 

 

 

 

 

 

 

Log; anti-J5 antibody endpoint titers :1: S.E.M.
Serum sampr IgM IgG1 IgG2
Pre-immunized 5.4 t 0.458 11 :1: 0.513 7.6 :1: 0.696
3rd 6.4 1 0.458 12 1 0.513 9.4 1 0.696
6H! 8.2 1 0.458 14 1 0.513 11.6 1 0.696
9th 9.2 1 0.458 15 1 0.513 12.2 1 0.696
12“ 10.0 1 0.453 15 1 0.513 12.8 1 0.696

 

 

 

 

 

95

Table 5 Mean percentage of phagocytosing bovine neutrophil of opsonized GFP-
transformed J5 E. coli

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Incubation Opsonin-Control % Phagocytosing neutrophil
(min) Mean :1: S.E.M.
60 BSA
10.48 0.907
60 IgG1
9.11 0.907
60 IgG;
18.61 0.907
60 None
10.64 0.907
90 BSA
8.20 0.907
90 IgG1
12.58 0.907
9O 1ng '
23.01 0.907
90 None
7.86 0.907
120 BSA
10.20 0.907
120 Ig61
10.16 0.907
120 IgG2
21.98 0.907
120 None
10.39 0.907
.150 BSA
11.60 0.907
150 IgG1
10.66 0.907
150 IgG;
25.46 0.907
150 None
11.72 0.907

 

 

 

96

 

CHAPTER 6
REFERENCES

Appelmelk, B.J., Maaskant, J .J ., van der Meer, N.M., VerWeij-van Vught, A.M.J.J.,
Thijs, B.G., MacLaren, D.M., 1993. Antigenic and immunogenic differences in
lipopolysaccharides of Escherichia coli J5 vaccine strains of different origins. J.
Gen. Microbiol. 139, 2641-2647.

Aitken, R., Hosseini, A., MacDuff, R., 1999. Structure and diversiﬁcation of the bovine
immunoglobulin repertoire. Vet. Immunol. Immunopathol. 71(1), 21-29.

Anderson, K. L., Smith, A.R., Gustaﬁson, B.K., Spahr, S.L., Whitmore, H.L., 1982.
Diagnosis and treatment of acute mastitis in a large dairy herd. J. Am. Vet Med.
Assoc. 181, 690-696.

Anderson, K.L., Smith, A.R., Shanks, R.D., Whitmore, H.L., Davis, L.E., Gustafsson,
B.K., 1986. Endotoxin-induced bovine mastitis: immunoglobulins, phagocytosis,
and effect of ﬂunixin meglumine. Am. J. Vet. Res. 47:2405-2410.

Ausubel, F.M., Brent, R., Kingston, RE, Moore, D.D., Seidman, J .G., Smith, J.A., Struhl,
K., High efﬁciency transformation by Electroporation. In: Current Protocols.
John Wiley & Sons, Inc.

Aydintug, M.K., Inzana, T.J., Letonja, T., Davis, W.C., Corbeil, LB, 2001. Cross-
reactivity of monoclonal antibodies to Escherichia coli J5 with heterologous gram-
negative bacteria and extracted lipopolysaccharides. J. Infect. Dis. 160(5), 846-57.

Barbior, BM, 1978. Oxygen-dependent microbial killing by phagocytes. II. N. Eng. J.
Med. 298, 721-725.

Bartlett, P.C., Miller, G.Y., 1993. Mastitis microbiology: What is considered
normal? Agri-Practice. 14(6), 12-14.

Bartlett, P.C., Miller, G.Y., Lance, S.E., Heider, LE, 1992. Clinical mastitis and
intramammary infections on Ohio dairy farms. Prev. Vet. Med. 12,59-71.

Bartlett, P.C., VanWijk, J., Wilson, D.J., 1991. Temporal patterns of lost milk production
following clinical mastitis in a large Michigan Holstein herd J. Dairy Sci. 74, 1561-
1572.

Baumgartner, J .D., O’Brien, T.X., Kirkland, T.N., Glauser, M.P., Ziegler, B.J., 1987.

Demonstration of cross-reactive antibodies to smooth gram-negative bacteria in
antiserum to Escherichia coli J5. J. Infect. Dis. 156(6), 136-43.

97

Baumgartner, J ., McCutchan, J ., van Melle, G., Vogt, M., Luethy, E., Glauser, M.P., Ziegler,
B.J., Klauber, M.R., Muehlen, E., Chiolero, R., Geroulanos, S., 1985. Prevention of
Gram-negative shock and death in surgical patients by antibody to endotoxin core
glycolipid. Lancet. July 13“: 59-63.

Bergquist, L.M., Pogosian, B., 2000. Structure of microbial cells. In: Microbiology:
Principle 5 and health science applications. W.B. Saunders Company, Philadelphia.
USA. 60-81.

Berning, L.M., Paape, M.J., Peters, RR, 1993. Functional variation in endogenous and
exogenous immunoglobulin binding to bovine neutrophils relative to parturition. Am.
J. Vet. Res. 54, 1145-1152.

Bhattacharjee, A.K., Opal, S.M., Palardy, J E Drabick, J.J., Collins, H., Taylor, R.,
Cotton, A., Cross, AS, 1994. Afﬁnity-puriﬁed Escherichia coli J 5
lipopolysaccharide-speciﬁc IgG protects neutropenic rats against gram-negative
bacterial sepsis. J. Infect. Dis. 170, 622-629.

Bishop, J.G., Schanbacher, F .L., Ferguson, L.C., Smith, KL, 1976. In vitro grth
inhibition of mastitis-causing coliform bacteria by bovine apo-Iactoferrin and

reversal of inhibition by citrate and high concentrations of apo-lactoferrin. Infect.
Immun. 14, 911.

Bonhomme, D., Hammarstrm, L., Webster, D., Chapel, H., Herrnine, 0., Le Deist, F .,
Lepage, E., Romeo, P.H., Levy, Y., 2000. Impaired antibody afﬁnity maturation
process characterizes a subset of patients with common variable
immunodeﬁciency. J Immunol. 165, 4725-4730.

Bramley, A. J., 1985. The control of coliform mastitis. Proc. (4m)Ann. Meeting Natl.
Mastitis Council. 4.

Braun, V., Bosch, V., 1972. Sequence of the murein and the attachment site of the lipid.
Eur. J. Biochem. 28, 51-69.

Burton, J .L., Erskine, R. J ., 2003. Immunity and mastitis: some new ideas for an old
disease. 2003. Vet. Clin. North Am. (Food Ani. Pract.) 19, 1-45.

Burton, J .L., Mallard, B.A., Mowat, D.N., 1993. Effects of supplemental chromium on
immune responses of periparturient and early lactation dairy cows. J. Anim. Sci.
71, 1532-1539.

Burton, J.L., Chaiyotwittayakun, A., Smith, K.L., Darch, R.A., Erskine, R.J., 2002.

Novel applications for colifonn vaccine programs. In Proceedings of the 41st
Annual Meeting of the National Mastitis Council. Feb 3-6, Orlando, PL. 89-110.

98

Burvenich, C., M.J., Paape, M.J., Hill, A.W., Guidry, A.J., Miller, R.H., Heyneman, R.,
Kremer, W.D.J., Brand, A., 1994. Role of the neutrophil leukocyte in the local
and systemic reactions during experimenatally induced E.coli mastitis in cows
immediately after calving. The Veterinary Quarterly. 16(1), 45-50.

Butler, J E 1981. A concept of humoral immunity among ruminants and an approach
to its investigation. InzThe Rurninant Immune System, J .E. Butler, J .R. Duncan,
and K. Nielson (eds). Adv. Exp. Med. Biol. (Plenum Press, New York, NY)
137, 3-55.

Butler, J.E., 1983. Bovine immunoglobulins: an augmented review. Vet. Immunol.
Immunopathol. 4, 43.

Butler, J .E., 1998. Immunoglobulin diversity, B-cell and antibody repertoire
development in large farm animals. Rev. Sci. Tech. Off. Epiz. 17, 43-70.

Buza J, Sileghem M, Gwasika P, Naessens, J ., 1997. CD5+ B lymphocytes are the main
source of antibodies reactive with non-parasite antigens in T typanosoma
congolense-infected cattle. Immunol. 92, 226-33.

Caffrn, J .P., Poutrel, B., 1988. Physiological and pathological factors inﬂuencing bovine
immunoglobulin G; concentration in milk. J. Dairy Sci. 71, 2035-2043.

Calandra, T., Glauser, M.P., Schellekens, J., Verhoef, J ., 1988. Swiss-Dutch J5
Immunoglobulin Study Group. Treatment of gram-negative-septic shock with

human IgG antibody to Escherichia coli J5: a prospective, double blind,
randomized trial. J. Infect. Dis. 158, 312.

Carroll, B.J., 1974. In vitro bactericidal reactions of serums and milks obtained from
cows inoculated with selected serum-resistant serum-sensitive colifonn bacteria.
Am. J. Vet. Res. 35, 175-211.

Carroll, B.J., Schalm, O.W., Lasmanis, J ., 1969. Experimental colifonn (Aerobacter
aerogens) mastitis. Bacteria and host factors in virulence and resistance. Am. J.
Vet, Res. 30, 1795-1804.

Chew, BR, 1996. Importance of antioxidant vitamins in immunity and health animals.
Animal Feed Science Technology. 59, 103-114.

Clark, P., van Roekel, DE, 1994. Efficacy of an Escherichai coli bacterin for the control
of colifonn mastitis in dairy cows. Agri. Pract. 15, 19-25.

Colditz, I.G., Maas, PI, 1987. The inﬂammatory activity of activated complement
in ovine and bovine mammary glands. Immunol. Cell. Biol. 65, 433-436.

99

Conner, J .G., Eckersall, PD, 1986. Acute phase response and mastitis in the cow.
R. Vet. Sci. 41, 126-128.

Cullor, J .S., 1991a. The Escherichia coli J5 vaccine: investigating a new tool to combat
coliform mastitis. Vet. Med. 86, 836. '

Cullor, J .S., 1991b. The role of vaccines in the prevention and moderation of clinical
mastitis, in Proceedings. 30th Annl Mtng Natl Mast Council. 68.

DeChatelet, LR. 1978. Initiation of the respiratory burst in human polymorphonuclear
neutrophils: A critical review. J. Reticuloendothelial Soc. 24, 73-91.

DeGraves, F .J ., Fetrow, J ., 1993. Economics of mastitis control. Vet. Clin. N. Am. 9:421-
434.

DiCarlo, A.L., Paape, M.J., Miller, RH, 1996. Reactivity of puriﬁed complement
component 3b with bovine neutrophils and modulation of complement receptor 1.
Am. J. Vet. Res. 57, 151-156.

Dosogne, H., Vangroenweghe, F., Burvenich, C., 2002. Potential mechanism of action of J5
vacccie in protection against severe bovine colifonn mastitis. Vet. Res. 33, 1-12.

Douglas, V.L., Cullor, J .S., Tyler, J.W., Thurmond, M.C., Bushness, RB, 1989. Rapid
decay of serum IgG recognizing gram-negative cell wall core antigens in neonatal
calves. Am. J. Vet. Res. 50, 1138-1140. ~

Eberhart, R. J ., 1977. Colifonn mastitis. J. Am. Vet. Med. Assoc. 170, 1160-1163.
Eberhart, R., Buckalew, J .M., 1977. Intramammary infection in a dairy herd with a low
incidence of Streptococcus agalactiae and Staphylococcus aureus. J. Am. Vet.

Med. Asooc. 171, 630-634.

Eberhart, R.J., Natzke, R.P., Newbould, F.H.S., Nonnecke, 3., Thompson, P., 1979.
Colifonn mastitis—a review. J. Dairy Sci. 62, 1-22.

Elbien A.D., Heath EC. 1965. The biosynthesis of cell wall lipopolysaccharide in
Escherichia coli: Properties of uridine diphosphate galactose 4-epimerase1eSs
mutant. J. Biol. Chem. 5, 1919-1925.

Erskine, RJ., Bartlett, PC, 1993. Serum concentrations of copper, iron, and zinc
during Escherichia coli mastitis. J. Dairy Sci. 76, 408-413.

Erskine, R.J., Bartlett, PC, 1995 . Assessing the success of clinical mastitis
treatment: What are production medicine goals? Agri-Practice 16, 6-10.

100

Erskine, R.J., Bartlett, P.C., van Lente, J .L., Phipps, CR, 2002. Efﬁcacy of systemic
ceﬁiofur as a therapy for severe clinical mastitis in dairy cattle. J. Dairy. Sci. 85,
2571-75.

Erskine, R.J., Eberhart, R.J., Hutchinson, L.J., Spencer, S‘.B., Campbell, M.A., 1988.
Incidence and types of clinical mastitis in dairy herds with high and low somatic cell
counts. J. Am. Vet. Med. Assoc. 192, 761-768.

Erskine, R.J., Eberhart, R.J., Grasso, P.J., Scholz, R.W., 1989. Induction of
Escherichia coli mastitis in cows fed selenium-deﬁcient or selenium-
supplemented diets. Am. J. Vet. Res. 50, 2093-2100.

Estes, MD. 1996. Differentiation of B cells in the bovine. Role of cytokines in
immunoglobulin isotype expression. Vet. Immunol. Immunopathol. 54, 61-67.

Frank, M.M., Joiner, K., Hammer, C., 1987. The function of antibody and complement
in the lysis of bacteria. Rev. Infect. Dis. 9(supp1. 5), 3537-45.

F reudenburg, M.A., Fomsguard, A., Mitov, I., Galanos, C. 1989. ELISA for lipid'A,
lipopolysaccharide and other hydrophobic antigens. Infection. 17, 322.

Frost, A. J ., Wanasinghe, DD, 1976. The pathogenesis of bovine mastitis with special
reference to the role of selective tissue adhesion. Proc. (53“) Ann. Meeting
Australian Vet. Assoc. 101-102.

Frost, A., Brooker, J .B.E., Hill, A.W., 1984. The Effect of Escherichia coli endotoxin
and culture ﬁltrate on the lactating bovine mammary gland. Aust. Vet. J. 61, 77.

Furukawa, K., Akasako-Furukawa, A., Shirai, 11., Nakamura, I-I., Azuma, T., 1999.
J unctional amino acids determine the maturation pathway of an antibody.
Immunity. 11(3), 329-338.

Furukawa, K., Shirai, H., Azuma, T., Nakamura, H., 2001. A role of the third
complementarity-determining region in the afﬁnity maturation of antibody. J Biol
Chem. 276(29), 27622-27628.

Giri, S., Chen, N.Z., Caroll, B.J., Mueller, R., Schiedt, M.J., Panico, L., 1984. Role of
prostaglandins in pathogenesis of bovine mastitis induced by Escherichia coli
endotoxin. Am. J. Vet. Res. 45(3), 586-591.

Gonzalez, R.N., Cullor, J .S., Jasper, D.E., Farver, T.B., Bushness, RB., Oliver M.N.,

1989. Prevention of clinical colifonn mastitis in dairy cows by a mutant
Escherichia coli vaccine. Can J. Vet. Res. 53, 301-305.

101

Gonzalez, R.N., Jasper, D.E., Kronlund, N.C., Farcer, T.B., Cullor, J .S., Bushness, R.B.,
Dellinger, J .D., 1990. Clinical mastitis in two California dairy herds
participating in contagious mastitis control programs. J. Dairy Sci. 73, 648-654.

Guidry, A.J., Paape, M.J., Pearson, R.E., 1980a. Effect of udder inﬂammation on
milk immunoglobulins and phagocytosis. Am J Vet Res 41, 751-753.

Guidry, A.J., Ost, M., Mather, I.H., Shainline, W.E., Weinland, ET, 1983. Sequential
response of milk leukocytes, albumin, immunoglobulins, monovalent ions, citrate,

and lactose in cows given infusions of Escherichia coli endotoxin into the
mammary gland. Am. J. Vet. Res. 44, 2262-2267.

Guidry, A.J., Pearson, R.E., Paape, M.J., Williams, W.F., 1980b. Relationship among
leukocyte phagocytosis, milk immunoglobulins, and susceptibility to
intramammary infection. Am. J. Vet. Res. 41, 997-1001.

Haas, K.M., Estes, D.M., 2000. Activation of bovine B cells via surface
immunoglobulin M cross-linking or CD40 ligation results in different B-cell
phenotypes. Immunol. 99, 272-278.

Haig, D.M., Hopkins, J., Miller, H.R.P., 1999. Local immune responses in afferent and
efferent lymph. Immnol. 96, 155-163.

Harlow, E. ,,Lane D. 1999. Using antibodies, a laboratory manual. Cold Spring Harbor
Laboratory Press New York. 234- 5, 313.

Harp J .A., Runnels, P.L., Pesch, BA, 1988. Lymphocyte recirculation in cattle:
Patterns of localization by mammary and mesenteric lymph node lymphocytes.
Vet. Immunol. Immunpathol. 20, 31-39.

Hayakawa, K., Hardy, R.R., Honda, M., Herzenberg, L.A., Steinberg, A.D., Herzenberg,
LA, 1984. Ly-l B cells: functionally distinct lymphocytes that secrete IgM
autoantibodies, Proc. Natl. Acad. Sci. U S A 81, 2494-8.

Hellman, J., Loiselle, P.M., Tehan, M.M., Allaire, J.E., Boyle, L.A., Kumick, J.T.,
Andrews, D.M., Sik Kim, K., Warren. H.S., 2000a. Outer membrane protein A,
peptidoglycan-associated lipoprotein, and murein lipoprotein are released by
Escherichia coli bacteria into serum. Infect. Immunol. 2566-25 72.

Hellman, J ., Loiselle, P.M., Zanzot, E.M., Allaire, J.E., Tehan, M.M., Boyle, L.A.,
Kumick, J .T., Warren, H,S., 2000b. Release of Gram-Negative Outer-Membrane
Proteins into Human Serum and Septic Rat Blood and Their Interactions with
Imunoglobulin in Antiserum to Escherichia coli J5. J. Infect. Dis. 181, 1034-43.

102

Hellman, J ., Roberts, JD. Jr., Tehan, M.M., Allaire, J .E., Warren, HS, 2002. Bacterial
peptidoglycan-associated lipoprotein is released into the bloodstream in gram-

negative sepsis and causes inﬂammation and death in mice. J. Biol. Chem.
277(16), 14274-14280.

Hellman, J. Warren, HS, 2001. Outer membrane protein A (OmpA), peptidoglycan-
associated lipoprotein (PAL), and murein lipoprotein (MLP) are released in
experimental Gram-negative sepsis. J. Endo. Res. 7(1), 69-72.

Hellman, J., Zanzot E.M., Loiselle P.M., Arnato, S.F., Black, K.M., Ge, Y., Kumick, J.T.,
Warren, HS, 1997. Antiserum against Escherichia coli J5 contains antibodies
reactive with outer membrane proteins of heterologous gram-negative bacteria. J.
Infect. Dis. 176, 1260-8.

Hill, A.W., 1981. Factors inﬂuencing the outcome of Escherichia coli mastitis in the
dairy cow. Res Vet Sci. 31, 107-112.

Hill, A.W., Heneghan, D.J., Field, T.R., Williams, M.R., 1983a. Increase in speciﬁc
opsonic activity in bovine milk following experimental Escherichia coli mastitis.
Res. Vet. Sci. 35, 222-226.

Hill, A.W., Heneghan, D.J., Williams, M.R., 1983b. The opsonic activity of bovine milk
whey for the phagocytosis and killing by neutrophils of encapsulated and non-
encapsulated Escherichia coli. Vet. Microbiol. 8, 293-300.

Hill-AW, Shears AL, Hibbitt KG. 1978. The elimination of serum-resistant Escherichia
coli from experimentally infected single mammary glands of healthy cows. Res.
Vet. Sci. 25(1), 89-93.

Hoblet, K.H., Schnitkey, G.D., Arbaugh, D., Hogan, J .S., Smith, K.L., Schoenberger,
P.S., Todhunter, D.A., Pritchard, D.E., Bowman, G.L., 1991. Costs associated
with selective preventive practices and episodes of clinical mastitis in nine herds
with low SCC. J. Am. Vet. Med Assoc. 199, 190-196.

Hogan, J .S., Bogacz, V.L., Aslam, M., Smith, KL, 1999. Eﬁicacy of an Escherichia
coli 15 bacterin administered to primigravid heifers. J. Dairy Sci. 82, 939-943.

Hogan, J .S., Smith, K.L., Hoblet, K.H., Schoenberger, P.S., Todhunter, D.A.,
Hueston, W.D., Pritchard, D.E., Bowman, G.L., Heider, L.E., Brockett, B.L.,
1989a. Field survey of clinical mastitis in low somatic cell count dairy herds. J.
Dairy. Sci. 72, 1547-1553.

Hogan, J. S., K. L. Smith, K. H. Hoblet, D. A. Todhunter, P. S. Schoenberger, W. D.
Hueston, D. E. Pritchard, G. L. Bowman, L. E. Heider, B. L. Brockett, and H. R.
Conrad. 1989b. Bacterial counts in bedding materials used on nine Ohio
commercial dairies J. Dairy Sci. 72:250-258.

103

Hogan, J.S., Smith, K.L., Schoenberger, P.S., Romig, S., Thompson, L., 1997.
Responses of antibody titers to intramammary immunization with Escherichia
coli .15 bacterin. J. Dairy Sci. 80, 2398-2402.

Hogan, J.S., Smith, K.L., Todhunter, D.A., Schoenberger, P.S., 1992a Field trial to
determine efficacy of an Escherichia coli J5 mastitis vaccine. J. Dairy Sci. 75(1),
78-84.

Hogan, J.S., Weiss, W.P., Todhunter, D.A., Smith, K.L., Schoenberger, P.S., 1992b.
Efﬁcacy of an Escherichia coli J5 mastitis vaccine in an experimental challenge
trial. J. Dairy Sci. 75, 415-422.

Hogan, J .S., Smith, K.L., Todhunter, D.A., Schoenberger, P.S., Shuster, D., 1995.
Efﬁcacy of recombinant bovine interleukin-2 as an adjunct to dry cow therapy. J.
Dairy Sci. 78, 1062-1067.

Howard, C.J., Brooke, G.P., Werling, D., Sopp, P., Hope, J .C., Parsons, K.R., Collins,,
RA, 1999. Decdritic cells in cattle: phenotype and function. Vet. Immunol.
Immunopathol. 72, 119-124.

Howard, C.J., Brownlie, J ., 1979. Interaction of certain bovine immunoglobulins with
complement in a single radial haemolysis system. Res. Vet. Sci. 27(3), 388-9.

Howard, C.J., Sopp, P., Bembridge, G., Young, J ., Parsons, K.R., 1993a. Comparison of
CD1 monoclonal antibodies on bovine cells and tissues. Vet. Immunol.
Immunopathol. 39(1-3), 77-83.

Howard, C.J., Sopp, P., Parsons,. K.R., Bembridge, G.P., Hall, G., 1993b. A new bovine
leukocyte antigen cluster comprising two monoclonal antibodies, CC43 and
CC118, possibly related to CD1. Vet Immunol. Immunopathol. 39(1-3), 69-76.

Howard, C.J., Taylor, G., Brownlie, J ., 1980. Surface receptors for immunoglobulin

on bovine polymorphonuclear neutrophils and macrophages. Res. Vet. Sci.
29, 128-134.

Jain, NC, 1979. Common mammary pathogens and factors in infection and mastitis.
J. Dairy Sci. 62, 128-134. ' ‘

Jain, N.C., Lasmanis, J ., 1978. Phagocytosis of serum-resistant and serum-sensitive
colifonn bacteria (Klebsiella) by bovine neutrophils from blood and mastitis milk.
Am. J. Vet. Res. 39, 425-427.

Janeway, C.A., Travers, P., Walport, M., Shlomchik, M., 2001. The humoral immune

response. In: Immunobiology (5th ed.): The immune system of health and disease.
Garland Publishing, New York. 341-380.

104

 

Jann K., Jann, B., 1987. Polysaccharide antigens of Escherichia coli. Rev, Infect, Dis.
9 (Suppl 5), 8517-26

Janzen, J .J., Bishop, J.R., Bodine, A.B., Caldwell, C.A., Johnson, D.W., 1982.
Conposted dairy waste solids and crushed limestone as bedding in free stalls.
J. Dairy Sci. 65, 1025.

Kaattari S.L., Zhang, H.L., Khor, I.W., Kaattari, I.M., Shapiro, DA, 2002. Afﬁnity
maturation in trout: clonal dominance of high afﬁnity antibodies late in the
immune response. Dev. Comp. Immunol. 26(2), 191-200.

Kantor, A., 1991. A new nomenclature of B cells. Immunol. Today. 12, 388.

Kacskovics, 1., Wu, Z., Simister, N.E., Frenyo, L.V., Hammarstrdm, L., 2000. Cloning

and characterization of the bovine MHC class I-like F 0 receptor. J. Immunol.
164, 1889-1897.

Kehrli, M.E., Jr, Harp, J .A., 2001. Immunity in the mammary gland. Vet. Clin.
North. Am. Food Anim. Pract. 17, 495-516.

Kehrli, M.E., Shuster, DE, 1994. Factors affecting milk somatic cells and their
role in health of the bovine mammary gland. J. Dairy Sci. 77, 619-627.

Kremer, W.D.J., Noordhuizen-Stassen E.N., Lohuis J.A.C.M.. 1990. Host defense and
bovine colifonn mastitis. Vet. Q. 12(2), 103-113.

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

Lascelles, AK, 1979. The immune system of the ruminant mammary gland and its role
in the control of mastitis. J. Dairy Sci. 62, 154-167.

Lei Pamg, C., Hansal, S., Goldsby, R.A., Osborne, BA, 1996. Gene conversion
contributes to lg light chain diversity in cattle. J. Immunol. 157(12), 5478-86.

Libon, C., Haeuw, J .F ., Crouzet, F., Chantal Mugnier, C., J. Y., Bonnefoy, Beck,
' A., Corvaria, N:, 2002. Streptococcus pneumoniae polysaccharides conjugated
to the outer membrane protein A from Klebsiella pneumoniae elicit protective
antibodies. Vaccine. 20(17-18), 2174-80.

Liu, Y.J., Johnson, G.D., Gordon, J ., MacLennan Ian, C.M., 1992. Germinal centers in
T-cell-dependent antibody responses. Immunol Today. 13(1), 17-21.

Lohuis, J.A.C.M., Van Leeuwen, W., Verheijden, J .H.M., Van Miert, A.S.J.P.A.M.,

Brand, A., 1988a. Effect of dexarnethasone on experiment Escherichia coil
mastitis in the cow. J. Dairy Sci. 71, 2782-2789.

105

Lohuis, J. A. C. M., Y. H. Schukken, J. H. M. Verheijden, A. Brand, and A. S. J. P. A. M.
Van Miert. 1990. Effect of severity of systemic signs during the acute phase of
experimentally induced Escherichia coli mastitis on milk production losses. J.
Dairy Sci. 73:333-341.

Lohuis, J.A.C.M., Verheijden, J.H.M., Bervenich, C., Van Miert, A.S.J.P.M., 1988b.
Pathophysiological effects of endotoxins in ruminants. 1. Changes in body
temperature and reticulo-rumen motility, and the effect of repeated
administration. The Veterinary Quarterly. 10(2), 109-125.

Mackie, D.P., Logan, ER, 1986. Changes in immunoglobulin levels in whey during
experimental Streptococcus agalactiae mastitis. Res. Vet. Sci. 40, 183-188.

MacLennan, I.C.M., Liu, Young-Jun, Johnson, GD, 1992. Maturation and dispersal of
B cell clones during T cell-dependent antibody responses. Immunol. Rev. 129,
143-161.

Maher J .K. and Kronenberg M. 1997. The role of CD1 molecules in immune responses
' to infection. Curr. Opin. Immunol. 9, 456-461.

Marks, M.I., Ziegler, E.J., Douglas, H., Corbeil, L.B., Braude, AL, 1982. Induction of
immunity against lethal Haemophilus inﬂuenzae type b infection by Escherichia
coli core Lipopolysaccharide. J. Cli. Invest. 69, 742-749.

McGuire, T.C, Musoke, A.J., 1981. Biologic activities of bovine IgG subclasses. Adv.
Exp. Med. Biol. 137, 359-366.

McGuire, T.C., Musoke, A.J., Kurtti, T., 1987. Functional properties of bovine IgG1
and IgG2: interaction with complement, macrophage, neutrophils and skin.
Immunol. 38, 249-256.

Meirom, R., Brenner, J ., Trainin, Z., 1993. BLV—infected lymphocytes exhibit two
patterns of expression as determined by Ig and CD5 markers. Vet. Immunol.
Immunopathol. 36, 179-86.

Miller, R.H., Guidry, A.J., Paape, M.J., Dulin, A.M., Fulton, LA, 1988. Relationship
between immunoglobulin concentrations in milk and phagocytosis by bovine
neutrophils. Am. J. Res. Res. 49, 42-45.

Mond, J .J ., Lee, A., Snapper, CA, 1995. T cell-independent antigen type 2. Ann. Rev.
Immunol. 13,655.

Mueller, R., Carroll, B.J., Panico, L., 1983. Hemolytic complement titers and

complement C3 levels in endotoxin-induced mastitis. Am. J. Vet. Res. 44, 1442-
1445.

106

 

Natzke, R.P., LeClair, B.J., 1976. Coliform contaminated bedding and new
infections. J. Dairy Sci. 59, 2152.

Naessens, J ., 1997. Surface Ig on B lymphocytes from cattle and sheep. Intl Immunol.
9(3), 349-54. ~

Nassens, J ., Williams, D.J.L., 1992. Characterization and measurement of CD5+B cells
in normal and T rypanosoma congolense-infected cattle. Eur. J. Immunol. 22,
1713-171 8.

Olsen, A., Jonsson, A., Norrnark, S., 1989. Fibronectin binding medicated by a novel
class of surface organelles on Escherichia coli. 338, 652-5.

Paape, M.J., Shafer-Weaver, K., Capuco, A.V., Van oostveldt, K., Burvenich, C., 2000.
Immune surveillance of mammary tissue by phagocytic cells. Adv. Exp.
Med. Biol. 480, 259-277.

Pancock, J .D., Beuerlein, G., Pecht, G., Tang, Y., Nie, Y., Wu, H., Huse, W.D., Watkins,
J .D., 2001. In vitro affinity maturation of human IgM antibodies reactive with
tumor-associated antigens Hybridoma and hybridomics. 20(6), 383-396.

Pollack, N., Chia, J .K., Koles, N.L., Miller, M., Guelde, G., 1989. Speciﬁcity and cross-
reactivity of monoclonal antibodies reactive with the core and lipid A regions of
bacterial lipopolysaccharide. J. Infect. Dis. 159(2), 168-88.

Porcelli SA. 1995. The CD1 family: a third lineage of antigen-presenting molecules.
Adv. Immunol. 59, 1-98.

Porcelli, S.A., Modlin, R.L., 1999. The Cdl system: antigen-presenting molecules for T
cell recognition of lipids and glycolipids. Annu. Rev. Immunol. 17, 297-329.

Porcelli, S.A., Morita, C. T., Modlin RL., 1996. T-cell recognition of non-peptide
antigens. Curr. Opin. Immunol. 8, 510-516.

Promega, 1996. Protocols and Application Guide, thrid edition. Promega Corporation
(Catalogue # P1610). 255 and 281.

Raetz, C.R.H., 1993. Bacterial endotoxin: extraordinary lipids that activate eukaryotic
signal transduction. J. Bact. 175, 5745-53.

Rainard, P. 1983. Experimental mastitis with Escherichia coli: sequential response of
leukocytes and opsonic activity in milk of immunised and unimmunised cows.
Ann Rech. Vet. 14, 281-286.

Rainard, P., Poutrel, B., 2000. Generation of complement fragment C5a in milk is
variable among cows. J. Dairy Sci. 83, 945-951.

107

 

Rainard, P., Caﬁin, J .P., 1983. Sequential changes in serum albumin,
immunogloblin (IgG1, IgG2, IgM) and lactoferrin concentrations in milk
following infusion of Escherichia coli into the udder of immunised and
unimmunised cows. Ann. Rech. Vet. 14, 271-279.

Rainard, P., Sarradin, P., Paape, M.J., Poutrel, B., 1998. Quantiﬁcation of
C5a/C5a(des Arg) in bovine plasma, serum and milk. Vet. Res. 29, 73-88.

Rauly, 1., Goetsch, L., Haeuw, J .F ., Tardieux, C., Baussant, T., Bonnefoy, J .Y., Corvaia,
N., 1999. Carrier properties of a protein derived from outer membrane protein A
of Klebsiella pneumoniae. Infect. Immun. 67(11), 5547-51.

Reneau, J .K. 1993. Clinical mastitis records in production medicine programs. Comp.
and Food Anim. Pract. 15, 497-503.

Riollet, C., Rainard, P., Poutrel, B., 2000. Differential induction of complement
fragment C5a and inﬂammatory cytokines during intramammary infections with
Escherichia coli and Staphylococcus aureus. Clin. Diag. Lab. Immunol. 7, 161-
167.

Sanchez-Carlos, V., Wilson, R.A., McDonald, J .S., Packer, RA, 1984. Biochemical
and serologic properties of E. coli isolated from cows with acute mastitis. Am J.
Vet. Res. 45, 1771-4.

Schmerr, M.J.F., Patterson, J.M., Van Der Maaten, M.J., Miller, J.M., 1985.

Conditions for binding bovine IgG1 to protein A-sapharose. Mol. Immunol.
22, 613-616.

Shuster, D.E., Harmon, R.J., 1992. High cortisol concentration and mediation of
the hypogalactia during endotoxin-induced mastitis. J. Dairy Sci. 75, 739-746.

Siber, G.R., Kania, S.A., Warren, HS, 1985. Cross-reactivity of rabbit antibodies to
lipopolysaccharides of Escherichia coli J5 and other gram-negative bacteria. J.
Infect. Dis. 152, 954-64.

Smith, J .L., Hogan, J .S., Smith, K.L., 1999. Efﬁcacy of intramammary immunization
with an Escherichia coli J5 bacterin. J. Dairy Sci. 82, 2582-2588.

Smith, K.L., 1983. Mastitis control: A discussion. J. Dairy Sci. 66, 1790-1794.

Smith, K.L., Todhunter, D.A., Schoenberger, PS, 1985. Symposium: environmental
effect on cow health and performance. J. Dairy Sci. 65, 1531-53.

Steinman, RM, 1991. The dendritic cell system and its role in immunogenicity. Ann.
Rev. Immunol. 9, 271-296.

108

 

Stevens, R, Hung, SN-Y., Welch, W.D., Young, LS, 1978. Restricted complement
activation by Escherichia coli with the K-1 capsular serotype: a possible role in
pathogenicity. J. Immunol. 121, 2174-81.

Tao, W., Corbett, MJ., Pickett, W., 1995. Monomeric bovine IgG; is a potent stimulus
for bovine neutrophils. J. Leuk. Biol. 58, 203-210.

Targowski, SP, 1983. Role of immune factors in protection of mammary gland J.
Dairy Sci. 66(8), 1781-9.

Targowski, S.P., Niemialtowski, M., 1986. Inhibition of lacteal leukocyte

phagocytosis by colostrum, nonlactating secretion, and mastitic milk Am. J. Vet.
Res. 47, 1940-1945.

Targowski, S.P., Klucinski, W., 1985. Effect of immune complexes from mastitic
milk on blocking of F c receptors and phagocytosis. Infect Immun. 47, 484-488.

Tomita, GM, Ray, C.H., Nickerson, S.C., Owebs, W.B., Gallo, G.F., 2000. A
comparison of two commercially available Escherichia coli vaccines against E.
coli intramammary challenge. J. Dairy Sci. 83, 2276-2281.

Tomita, G.M., Todhunter, D.A., Hogan, J.S., Smith, K.L., 1995. Antigenic
crossreactivity and lipopolysaccharide neutralization properties of bovine
immunoglobulin G. J. Dairy Sci. 78, 2745-2752.

Tizard, IR, 1996. Inﬂammation. In: Veterinary immunology, An introduction, ﬁfth
edition. W.B. Saunders Company, Philadelphia. 43-54.

Tsai, Chao-Ming, Frasch, CE. 1982. A sensitive silver stain for detecting
lipopolysaccharides in polyacrylamide gels. 119, 115-119.

Tyler, J.W., Cullor, J.S., 2002. Mammary gland health and disorders. In: Smith, B.P.,
(ed). Large animal internal medicine: disease of horses, cattle, sheep and goats
3rd edition. Mosby, Inc. St Louis, M0. 1019-1038.

Tyler, J.W., Cullor, J.S., Dellinger, J.D., 1990a Cross reactive affinity puriﬁcation of

immunoglobulin recognizing common gram-negative bacteria] core antigens. J.
Immunol. Methods. 129, 221-226.

Tyler, J.W., Cullor, J.S., Osburn, B.I, Bushnell, RB, Fenwick, B.W., 1988.
Relationship between serologic recognition of Escherichia coli 0111:B4 (J 5) and
clinical coliform mastitis in cattle. Am J Vet Res 49, 1950-1954.

Tyler, J.W., Cullor, J .S., Spier, S.J., Smith, B.P., 1990b. Immunity targeting common
core antigens of gram-negative bacteria J. Internal Med 4, 17-25.

109

Tyler, J ., Spears, H., Cullor, J ., Smith, W., Nelson, R., Martin, J ., 1991. Antigenic
homology among gram-negative organisms isolated from cattle with clinical
mastitis. J. Dairy Sci. 74, 1235-1242.

Tyler, J .W., Spears, H., Nelson, R. 1992. Antigenic homology of endotoxin with a
colifonn mastitis vaccine strain, Escherichia coli 0111:B4 (J5). J Dairy Sci;
75, 1821-1825.

Van Steenhouse, J .L., 1987. Free radicals: Relation to tissue damage—a review. Vet.
Cli. Pathol. 116, 29-35.

Verbeet, M.P., Vermeer, H., Warmerdam, G.C., de Boer, H.A., Lee, SH, 1995.
Cloning and characterization of the bovine polymeric immunoglobulin receptor-
encoding cDNA. Gene 164, 329-333.

Warren, H.S., Norvitsky, T.J., Bucklin, A., Kania, S.A., Siber, GR, 1987. Endotoxin
neutralization with rabbit antisera to Escherichia coli J 5 and other gram-negative
bacteria. 55, 1668-73.

Watson, D.L., Golditz, I.G., Gill, HS, 1994. Immunological effector mechanisms in
ruminants. In: Wood, P.R., Willadsen, P, Vercoe, J E Hoskinson, R.M.,
Demeyer, D., (eds). Vaccine in agriculture: Immunological applications to animal
health and production. CSIRO, Australia. 21-36.

White, D.G., 1999. Antimicrobial resistance, Proceedings of the National Mastitis
Council.

Worku, M., Paape, M.J., Filep, R., Miller, RH, 1994. Effect of in vitro and in vivo
migration of bovine neutrophils on binding and expression of Fc receptors
for IgG; and IgM. Am. J. Vet. Res. 55, 221-226.

Zhang, G., Young, J .R., Tregaskes, C.A., Sopp, P, Howard, C.J., 1995. Identiﬁcation of
a novel class of mammalian F cy Receptor. J. Immunol. 155, 1534-1541.

Ziegler, B.J., Douglas, H., Sherman, J .E., Davis, GB, Braude, A., 1973. Treatment of
E. coli and Klebsiella bacteremia in granulocytic animals with antiserum to a
UDP-gal epirnerase deﬁcient mutant. J. Immunol. 111, 433.

Ziegler, B.J., McCutchan, J .A., Douglas, H., Braude, A.I., 1975. Prevention of lethal
Pseudomonas bacteremia with epirnerase deﬁcient E. coli antiserum. Trans.
Assoc. Am. Physicians. 88, 101.

Ziegler, B.J., McCutchan, J .A., Fierer, J ., Glauser, M.P., Sadoff, J .C., Douglas, H.,

Braude, A.I., 1982. Treatment of gram-negative bacteremia and shock with
human antiserum to a mutant Escherichia coli. N. Engl. J. Med. 307, 1225.

110

1:

u"111111111131111111131

   

...........