1
APPENDIX ONE
Characterization of vaccine candidate strains
for
E. coli
vaccine 2
1. Serogroup, fimbriae and enterotoxins possessed by the candidate strains

The virulence characteristics (OK-antigen serogroup, fimbriae and enterotoxins) of the three strains
selected for vaccine production were independently confirmed by The Pig Health and Research Unit
(PHRU), Victorian Department of Primary Industry (Table 1). These strains have been stored as
freeze dried specimens in three separate laboratories (NIVR, UQ and PHRU).
Table 1
:
E. coli
strains used for the preparation of vaccine
Virulence Characteristics
Designation of
E. coli

vaccine strains
O-serogroup Fimbriae Enterotoxin(s)
NVP613
(CARD-VN1)
O8 5F-* STa/STb/LT
NVP1402
(CARD-VN2)

E. coli
grown at 18-20
o
C will not produce fimbriae. Therefore, to prepare sera that is specific
for the possible new fimbrial type, the antisera from rabbits is mixed with inactivated NVP613 cells
harvested from an 18-20
o
C culture. Non-specific antigen:antibody complexes are then removed by
centrifugation and the remaining antisera should only agglutinate in the presence of cells expressing
the new fimbrial type. However, it must be remembered that this is an inexpensive but crude
technique and all cross-reactive antibodies may not be removed during cross-absorption.

3
Method:
- Grow isolate (NVP613 5F-) at 18-20
o
C (to prevent production of fimbriae) for 6-8 hours
- Inactivate with formalin (0.3% v/v) overnight
- Wash and centrifuge 3 times in PBS/saline
Absorption:
- Make a 20% cell suspension with the inactivated cells grown at 18-20
o
C in the rabbit antisera
- Incubate 1 hour at 37
o
C, then refrigerate overnight
- High spin to pellet cells
- Remove antisera
- Coagglutinate antisera
- Test against cells grown at 37C

NVP625 37
o
C O8 5F-/STa/STb/LT +++
NVP1372* 37
o
C O64 F5/STa -
NVP1392 37
o
C O149:K91 F4/STa/STb/LT -
NVP1402* 37
o
C O149:91 F4/STa/STb/LT -
NVP613* 37
o
C O8 5F-/STa/STb/LT +++

20
o
C

+++
NVP1272 37
o
C O8:G7 F4/STa/STb -

- no agglutination observed
+ Weak agglutination (25%)
++ Moderate agglutination (50%)
+++ Strong agglutination (100%)
* vaccine strains

(CARD-VN1)
O8 5F
-
, F? STa, STb, LT
EC-VN8 O8 5F
-
, F? STa, STb, LT

B.

Results of mannose-resistant haemagglutination:
The two ETEC strains were examined for mannose-resistant haemagglutinating activity using Sheep
Red Blood Cells. Both strains were tested using overnight cultures grown at 18
o
C and 37
o
C. The
density of bacteria in each culture was adjusted to OD=1 (
A
=660
i
m) in NaCl 0.85% using a PYE
Unicam PU-8600 UV/VIS spectrophotometer (Philips), prior to mixing with red blood cells. The
results of haemagglutination of cultures of each strain are presented in Table 1. Mannose-resistant
haemagglutination was observed at 37
o
C, but not at 18
o
C for both strains, confirming the production
of adhesins (ie fimbriae) at 37

Transmission electron microscopy photographs taken at low and high magnification showed the
presence of hair-like structures on the surface of the bacteria cells (an example is shown in Figure 1). 31
Scale bar = 2
t
m

Scale bar = 1
t
m

Scale bar = 200
i
m

5F- ETEC strain (grown
overnight in BHI broth) under
transmission electron
microscopy

6
D. Preliminary results on purification of 5F- fimbriae:
The purification of the 5F- fimbriae was performed using the OIE
E. coli
Laboratory protocol
EcL1000 (Production of Fimbriae) with some modifications. All steps were exactly the same as
described in the protocol, except that:
- In the Precipitation step, the concentration of sulfate ammonium was increased to 30% and 40%

Salmonella
. As I am not fully convinced, we will run another SDS-PAGE and
this time, we will cut out any bands we observe around 18 to 22 kDa that seem to be less expressed
at 18°C, and send them off for MS. It costs around $60 per reaction. We are doing this at the
moment.
As the DNA sequence of the gene in
Enterobacter
is known, we could also test by PCR or probe (see
pdf attached) phenotypically positive and negative isolates to see if there is any correlation.
I have tested all of the 140 isolates in our collection from Thuy by immunofluorescence using the
adsorbed anti-F19 serum. There is a good correlation with the virotype Paa,STa,STb,LT,East1. All of
the isolates have been virotyped for 20 virulence genes. Also, the postweaning isolates (about 40)
have been tested by MIC, for the paper with Thuy. We did not do the preweaning isolates for MIC,
as you had said that these isolates had already been tested and will be in another paper. Several of the
PWD isolates are enroflox resistant, hence my interest to test for the qnr genes. I am compiling all
these results at the moment.
I know that the F19 characterisation has been taking a lot of time to get results. It is a question of
man power. When my technician has had spare time, she has worked on it. However, she went off on
pregnancy leave! Now that the summer holidays are finished, I have put another technician on the
project. We will run more gels this week or next and send off bands for MS. Our problem has always
been that the adhesin band is not produced in great quantities as we observed with F4, F165 etc, so it
is not easy to identify the correct band.

7
Do you think that at this point you will have enough for AUSAID? I think that if
we want to go further, it would be a great project for a student. Another approach if
we still have problems could be DIGE, the proteomic approach. We could send the
18C and 37C preps off for 2-D gel electrophoresis and MS.

Let me know what you think

An inactivated whole cell multivalent vaccine was prepared according to the protocol of the
Bacteriology Laboratory, NIVR. The vaccine contained three
E. coli
strains with the O-antigen and
virulence characteristics of each listed in Table 1.
Table 1
:
E. coli
strains used for the preparation of vaccine

Characteristics
Designation of
E. coli

vaccine strains
O-serogroup Fimbriae Enterotoxin(s)
CARD-VN1 O8 F? STa/STb/LT
CARD-VN2 O149: K91 F4 STa/STb/LT
CARD-VN3 O64 F5 STa

2.
Specialised culture media were prepared in order to provide favourable growth conditions for the
production of fimbriae. For efficient expression of F4, strain CARD-VN2 was grown on Buffered
Glucose Nutrient Agar as described by Jones & Rutter (1972), whereas for the production of F5
fimbriae on strain CARD-VN3, Minca agar as described by Guinee
et al.
(1977a) was used. For the
strain with currently uncharacterized fimbriae (CARD-VN-1), it was shown in Appendix One that
Buffered Glucose Nutrient Agar was suitable for the production of a possible new fimbrial type.
The procedure used to prepare the vaccine is summarised in Figure 1. In brief, cultures of each strain

Preparation of
E. coli
multivalent vaccine (1 ml of vaccine contains approximately 10
10

bacteria)
10% (v/v) bufferred
formaldehyde to a final
concentration of 0.3%
Mix with equal colume of
each bacterin

Add 2% (v/v) aluminum
hydroxide to a final
concentration of 20%
Freeze-dried
cultures
2 ml TSB (37
o
C,
overnight)
SBA (37
o
C,
overnight)

NIVR
E. coli

vaccine
References:
Jones, G. W. & Rutter, J. M. (1972).
Role of the K88 antigen in the
pathogenesis of neonatal diarrhoea caused by
Escherichia coli
in piglets.
Infection
and Immunity
6
, 918-927.
Guinee, P. A. M., Veltkamp, J. & Jansen, W. H. (1977a).
Improved Minca
medium for the detection of K99 antigen in calf enterotoxigenic strains of
Escherichia coli
.
Infection and Immunity1
APPENDIX THREE
Results of safety and efficacy studies conducted on NIVR’s
E. coli
vaccine in
comparison to commercially available vaccines (Pfizer Litterguard and
Intervet EcoVac)

C, the growth was harvested into sterile PBS and the
suspension was adjusted to a density of approximately 10
10
bacteria/ml.
Following birth, piglets were allowed access to colostrum and nursed a further 24 h, after which
they were intragastrically inoculated with 1 ml of bacterial suspension (10
10
bacteria/ml) via a
stomach tube.
Piglets born to both vaccinated gilts and non-vaccinated gilts were challenged with a homologous
bacterial suspension prepared from one of the five vaccine strains (one strain per two litters). In
each litter, five piglets were challenged with a single
E. coli
strain and the remaining piglets were
left as non-inoculated in-contact control piglets. Only two strains (NVP613 and NVP1402) were
used to challenge the piglets from the control gilts.
None of the vaccinated or control gilts showed abnormal clinical signs in the 24 h period
immediately following vaccination or administration of the placebo. From each gilt, 10-12 piglets
were born alive. All piglets were healthy and had an average weight of 0.9-1.2 kg/piglet.

2
Table 2:
Designation of groups of piglets used for challenge
No. of piglets
Gilt
Group (litter
number)
Challenge In-contact
Strain
challenged

Challenge In-contact

Challenge In-contact

1 (I, II) NVP613
3/10
a

(30.0)
1/11
b

(9.1)
4/10
a

(40.0)
3/11
b

(27.3)
2 (III, IV) NVP1402
3/10
a

(30.0)
1/12
b

(8.3)

b

(0.0)
3/10
a

(30.0)
2/11
b

(18.2)
Vaccinated
gilts
5 (IX, X) NVP1372
3/10
a

(30.0)
1/13
b

(7.7)
5/10

(50.0)
4/13
b

(30.8)
6 (XI) NVP613


P< 0.001 (vaccinated groups vs. control group) (two-tailed Fisher’s exact test)

Table 4:
Duration of diarrhoea and excretion of challenge strains in piglets born from vaccinated
ilts and unvaccinated controls
Diarrhoea
(average days)
Excretion of
challenge strain
(average days)
Gilt
Group (litter
number)
Strain
challenged
Challenge In-contact Challenge In-contact
1 (I, II) NVP613 2.0
a
1 2.5
a
1.3
2 (III, IV) NVP1402 2.0
a
1 2.3
a
1.7
3 (V, VI) NVP1271 1.7 1 1.8 2.0
4 (VII, VIII) NVP2081 1.5 - 2.0 1.5
Vaccinated

Table 5:
Morbidity requiring euthanasia in piglets born from vaccinated gilts and unvaccinated
controls
Gilts
Group
(Litter
number
Strain
challenged
No. of piglets
euthanased/inoculated
(%)
No. of piglets
euthanased /in-contact
(%)
1 (I, II) NVP613 0/10 (0)
a
0/11 (0)
2 (III, IV) NVP1402 0/10 (0)
a
0/12 (0)
3 (V, VI) NVP1271 0/10 (0)
a
0/12 (0)
4 (VII, VIII) NVP2081 0/10 (0)
a
0/11 (0)
Vaccinated
gilts
5 (IX, X) NVP1372 0/10 (0)

Piglets born alive 80 82
Stillbirths 3 4
Mummies or (died before parturition) 2 2
Deformities eg splay legs 4 5
Abortion 0 0

There were no significant differences between two groups of sows (all P values >0.05).

3. Efficacy study:
For the efficacy study, 4 groups of pigs (12 weeks of age) were divided as follows:
Control group n = 5 pigs
Litterguard group n = 10 pigs
Intervet vaccine n = 10 pigs
NIVR vaccine n = 10 pigs
The procedure for bleeding and vaccination as follows:
Age of pigs (weeks)

12 Bleeding and 1
st
vaccination
16 2
nd
vaccination
18 Bleeding

All pigs showed excellent health during and after the experiment, except that 1 pig from the
control group developed shock and died after the first bleeding. A total of 34 sera samples were
examined for antibody level against F4 antigens by ELISA test (according to the protocol from
E.
coli

uard 0.981 1.842 0.861
litterguard 1.176 1.843 0.667
litter
g
uard 1.376 2.182 0.806
litterguard 0.914 2.026 1.112
ecovac 1.11 2.162 1.052
ecovac 1.164 2.173 1.009
ecovac 1.954 2.252 0.298
ecovac 0.906 2.088 1.182
ecovac 0.76 2.299 1.539
ecovac 0.432 1.961 1.529
ecovac 0.83 1.701 0.871
ecovac 1.062 1.826 0.764
ecovac 1.236 1.699 0.463
ecovac 0.944 1.643 0.699
NIVR 0.562 1.871 1.309
NIVR 1.192 1.715 0.523
NIVR 1.071 2.122 1.051
NIVR 1.124 1.943 0.819
NIVR 0.34 1.599 1.259
NIVR 1.028 1.893 0.865
NIVR 1.135 1.798 0.663
NIVR 1.41 2.023 0.613
NIVR 1.014 2.048 1.034
NIVR 1.121 1.68 0.559
control 1.447 1.664 0.217
control 0.8 0.91 0.11
control 0.593 0.729 0.136
control 0.611 1.052 0.441

NIVR 0.8695 0.0814

Statistical analysis
An analysis of the difference between the Pre and post vaccination OD values was done using
linear regression with pairwise differences. Statistical Analysis was preformed by Genstat 8
th

Edt, Laws Agricultural Trust, Rothamsted Experimental Station.
Results

There was a significant difference between the OD levels for the treatment groups (p<0.003).
There was no significant difference between the antibody response elicited (as demonstrated by
OD values) by Litterguard, EcoVac or NIVR vaccines (p>0.1). All three vaccines were
significantly different from the control group (p< 0.005) (Table 1)
Table 1
Treatment group Mean
Control 0.2260
a

Litter
g
uard 0.8129
b

EcoVac 0.9406
b

NIVR 0.8695
b