Download Immune response to tuberculosis infection

Tuberculosis detection in the Asian elephant
(Elephas Maximus) population of Thailand
 development of an IFNγ assay
 evaluation of a multiple antigen iELISA and a commercial rapid
test
Research paper
Ilona Bontekoning
Research project from September 2009 – December 2009
Kasetsart University, Bangkok, Thailand
Chiang Mai University, Chiang Mai, Thailand
Utrecht University, Utrecht, Holland
Supervised by prof. dr. V. Rutten
Table of contents
Abstract
3
Introduction
4
Background information
5
Materials and method
- Part I : expression of IFNγ
- Part II: iELISA and RT
11
11
13
Results
-
16
16
17
Part I : expression of IFNγ
Part II: iELISA and RT
Discussion
22
Acknowledgement
24
References
25
Appendix: iELISA procedure
28
2
Abstract
As tuberculosis might be a problem in the declining Asian elephant population but too
little is known about the exact prevalence of this disease, researchers in Bangkok,
Chiang Mai, Utrecht and Pretoria are currently busy developing blood tests for
detecting tuberculosis in elephants. During this research project, I participated in the
development of an IFNγ assay by working on the expression of glycosylated IFNγ
from insect cells, to be used as a positive control in the development of the IFNγ
assay. I also evaluated sensitivity, specificity, positive and negative predictive value
of indirect (i)ELISA tests employing ESAT6/CFP10, CFP10 & MBP83 antigens, and
the commercial TB STAT-PAK rapid test.
The expression of IFNγ has encountered quite some obstacles, but we succeeded in
constructing a pMT/BiP/V5-HisA vector carrying an IFNγ gene insertion (as assessed
by PCR product size). This plasmid has been transferred into Schneider 2 insect cells
(Drosophila melanogaster), using the Drosophila Expression System (DES®,
Invitrogen). A stable expression of IFNγ by these cells has not been established yet.
The comparison of the different diagnostic tests for detecting TB infection in
elephants provided some interesting results. Compared to the 100% sensitivity and
97% specificity claimed by the producer of the TB STAT-PAK test, in this project a
sensitivity of 80% and a specificity of 87.23% were found. The iELISA test
employing the immunodominant antigens ESAT6 & CFP10, showed a sensitivity of
75% and a specificity of 97.82%, and a positive predictive value of 88.24% opposed
to a positive predictive value of 57.14% of the TB STAT-PAK test. Negative
predictive values of these tests were about the same. This ESAT6/CFP10 iELISA test
might be an alternative test to the expensive TB STAT-PAK test or trunk wash culture
when screening for tuberculosis. However, limitations such as low sample size and –
variation, growing body of evidence for a low sensitivity of the trunk wash culture
(considered as the gold standard in this experiment), suboptimal negative controls
used in the iELISA procedure, and the use of near to expiry STAT-PAK tests, suggest
that more research has to be done regarding the use of this ESAT6/CFP10 iELISA
test.
3
Introduction
The number of Asian elephants currently decreases at an alarming rate. The
International Union for Conservation of Nature (IUCN) has estimated their decline at
50% in the past three generations and has put the Asian elephant on its Red List of
endangered animals [24]. Captive breeding of elephants is difficult and little is known
about the prevalence of diseases in elephants. That’s why Europe and Asia have set
up the EU-Asia Link Project on Management of Health and Reproduction of Elephant
Populations in Asia. In this three-year project, veterinary faculties of the EU and Asia
collaborate to develop research capacity at Asian institutes. By academic training of
staff and students in Asia, the knowledge to preserve the elephants of Asia, which are
considered both a unique natural heritage and a valuable economic resource to both
the agricultural and tourism sectors, can be disseminated [23].
As tuberculosis might be a problem in Asian elephants, but too little is known about
the exact prevalence of this disease and all of the current diagnostic tests show
limitations [6][11][13][19][20], researchers in Bangkok, Chiang Mai, Utrecht and
Pretoria are now working on the development of blood tests for detecting tuberculosis
in elephants. This will eventually enable effective disease management in the elephant
population. During my research internship I went to Kasetsart and Chiang Mai
Universities, Thailand, in order to help developing these blood tests. I will first
provide some background information about the Mycobacterium tuberculosis
complex, tuberculosis disease, the immune response to tuberculosis infection, and the
available diagnostic tests. Then, I will discuss my research project in more detail.
4
Background information
Tuberculosis: mycobacterium tuberculosis complex
Tuberculosis is a chronic necrotizing granulomatous disease caused by the pathogenic
mycobacterium tuberculosis complex (MTBC), which comprises the species
Mycobacterium tuberculosis, M. bovis, M. africanum, M. microti and M. canetti
[7][20]. Mycobacteria are obligate aeroob, non-spore forming, non-motile, rod
shaped, acid-fast micro-organisms. Their cell walls are rich in complex lipids and
waxes containing mycolic acid [16]. Mycobacteria are intracellular pathogens that
reside mainly within macrophages and grow best in tissues with high oxygen tension,
such as the apices of the lung or the renal cortex [10][7]. This way they are able to
survive many years in a slow-replicating or non-replicating state, induced by the host
immune response or fibrotic encapsulation [10]. M. tuberculosis and M. bovis are
most highly pathogenic. Humans are the natural and reservoir hosts for M.
tuberculosis, domestic cattle are the natural and reservoir hosts for M. Bovis [15].
Both pathogens have a wide host range and can be found in fish, reptiles, birds and
(marine) mammals worldwide [20]. The most common manifestation of tuberculosis
is pulmonary disease, but nearly all organ systems can be involved, such as the lymph
nodes, pleura, bones and joints [7].
Tuberculosis in elephants
Tuberculosis in captive elephants is most frequently caused by M. tuberculosis,
although infection with M. bovis also occurs. Davis et al (2001) speculate that humans
probably introduced tuberculosis into the elephant population centuries ago [4]. In
elephants infected with M. tuberculosis clinical signs are often absent or only shown
in the terminal stages of chronic disease [5][8]. Clinical signs are comparable to those
observed in humans and may include weight loss, anorexia, weakness, dyspneu,
coughing, and exercise intolerance. Transmission of tuberculosis occurs by
aerosolization of infected respiratory droplets when the animals cough, trumpet or
trunk spray. Transmission is influenced by the bacterial load, droplet size, length of
exposure, proximity to the infected animal and immune status of exposed individuals.
Tuberculosis can only be transmitted from elephants with active pulmonary disease
following primary infection or after reactivation of latent infection [5].
Tuberculosis: risk for humans and elephants
Transmission from elephants to other animals and even humans has been described.
Michalak et al. (1998) conducted research in an exotic animal farm in Illinois, USA,
in which three elephants died of tuberculosis and one was found culture-positive.
From the twenty-two animal handlers, who had moderate to frequent animal contact,
eleven showed positive result with the tuberculin skin test (discussed below).
Transmission between elephants and human was strongly suggested by DNA finger
printing demonstrating the same pathogen strain in both an animal handler with
culture-positive active tuberculosis and the dead elephants [12]. Oh et al. (2002)
examined a tuberculosis infection in a Los Angeles Zoo in which DNA fingerprinting
suggested transmission between two Asian elephants, three Rocky Mountain goats
and one rhinoceros. Of the LA Zoo employees 18% showed a positive result on
tuberculin skin tests, but no clinical symptoms or radiographic findings [14]. Davis et
al. (2001) emphasize that people must be working in close proximity to, and have
more than incidental contact with an infected elephant for transmission of the disease
to occur. Activities such as blowing off or cleaning elephants, executing trunk
5
washes, other medical procedures, and attending necropsies, increase potential for
human exposure to tuberculosis, as mentioned by these authors [4]. They agree with
Michalak et al. (1998) who state that there has to be ‘prolonged and close contact’
between elephant and human for transmission to occur [12]. All together, these
publications confirm that there is a substantial risk for transmission of tuberculosis
from elephant to human, especially for people working in close contact with the
elephants.
And how about the risk for elephants? Tuberculosis is a major global health problem:
up to one third of the world’s population is infected with M. tuberculosis. The World
Health Organization (WHO) estimates that 8 million new cases and 1.9 million deaths
occurred in 2000 due to tuberculosis, making the disease the second leading cause of
death, exceeded only by HIV/AIDS. Tuberculosis occurs worldwide, but 95% of
cases and 98% of deaths occur in developing countries, especially in sub-Saharan
Africa and Asia [7]. As in Thailand the number of patients with tuberculosis is
increasing with the increased incidence of HIV, a further increase in the infection rate
among people and animals, especially elephants, is feared [20].
Immune response to tuberculosis infection
After inhalation, infectious droplet nuclei containing viable tubercle bacilli are
deposited in the lung. Here, they invade the alveolar epithelial cells or they are
phagocytosed by alveolar macrophages and dendritic cells. Macrophages and
dendritic cells present the mycobacterial antigens on their major histocompatibility
complex (MHC) proteins to T-lymphocytes with appropriate antigen-specific
receptors [7]. CD4 T-lymphocytes are the T-lymphocytes who play a key role in the
host immune response to tuberculosis. They regulate the acquired cellular immune
response which provides protection against mycobacteria, but they also participate in
the development and evolution of the infection, chronic reactivation and tissue
damage [9]. During the immune response, two separate waves of cytokine-producing
CD4 T-lymphocytes can be discriminated: T-helper 1 (Th1) and T-helper 2 cells
(Th2) [15][21].
The first wave consists of Th1 cells. Th1 cells promote cell mediated immunity (CMI)
by activating macrophages and by stimulating antigen presentation through regulation
of the MHC protein expression. They participate in delayed type hypersensitivity
(DTH), which may be measured by the tuberculin skin test (discussed below), and
they seem to induce a primary antibody response [15]. IFN-γ, the mayor cytokine
produced by Th1 cells, is considered critical for the control of mycobacterial infection
[21]. After in vitro infection of mice CD4 T-lymphocytes with mycobacterial
antigens, copious amounts of this IFN-γ cytokine are measured between day 10 and
30 post inoculation, with its decline starting after day 30 [15].
During progression of the infection, immune responsiveness tends to switch towards
the Th2 type [2]. A rise in the level of the major cytokine IL-4 can be measured
starting from day 30 after in vitro inoculation of CD4 T-lymphocytes with
mycobacterial antigens [15]. By antagonising the Th1 cell response, Th2 cells deviate
the initial CMI towards a humoral immune response [9]. They are important in
controlling isotype switching to secondary antibody responses [15].
6
In hosts with a competent immune system, the initial interactions between the bacillus
and the host cells lead to induction of specific acquired immunity against tuberculosis
[7]. Infection with mycobacteria can eventually result in the formation of granulomas.
The core of a granuloma exists of infected and killed macrophages, which are
surrounded and infiltrated by T-lymphocytes. Granulomas are formed to control and
to constrict the infection, but they may ultimately turn into necrotic or even calcified
lesions and cause significant tissue damage [21].
Diagnosing tuberculosis
Early in tuberculosis infection, when the mycobacteria reside mostly within
macrophages, little if any free antigen is able to evoke an antibody response.
Consequently, measurement of the CMI (IFNγ & DTH) rather than the humoral
response (antibodies) may indicate an early tuberculosis infection. However, if an
elephant develops active tuberculosis and the bacterial and antigenic load increase,
antibodies and free antigen can be detected in the blood, as depicted in figure 1 [2].
Figure 2 gives a good overview of the ability of some tests (which will be discussed
below) to detect infection of tuberculosis in cattle. This figure clearly illustrates the
decline of the CMI and the rise of the antibody level during progression of infection.
Definitive antemortem diagnostic techniques for tuberculosis in elephants have
limitations. No single diagnostic test provides optimal sensitivity, specificity, or
possibilities for use in the field; making them inadequate for effective disease
management. The major diagnostic tests for detecting tuberculosis will be discussed
here in three categories:
1) Direct identification of tubercle bacilli
2) Measurement of the CMI
3) Serology
1) Direct identification of M. tuberculosis/M. bovis
Identification of tuberculosis by culture of a trunk wash sample is the only officially
recognised diagnostic test, the gold standard. This test has only been validated in
domestic cattle, bison, and Cervidae [13]. There is a growing body of evidence that
this test has poor sensitivity, as it can only identify animals with extensive shedding
of mycobacteria which usually occurs in the late course of disease [11]. False negative
results may further be caused by insufficient mycobacterial numbers in trunk
secretions (<100 organisms/ml), inadequate collection procedures, contamination with
other nontuberculous mycobacteria, intermittent mycobacterial shedding, and
improper sample handling. Another drawback of this test is the fact that culturing of
the trunk wash may take up to eight weeks, creating an opportunity for an infected
elephant to disseminate mycobacteria in its environment and infect others [3].
Polymerase chain reaction (PCR) is a recently developed test for the direct detection
of mycobacterial DNA in samples such as blood, sputum, mucus or milk. It is based
on DNA amplification by DNA polymerase and specific primer sets. Unfortunately,
there is a high rate of false positives using this method [3].
7
Figure 1
Schematic representation of the immune response during the course of infection with increasing
bacterial load. The shade areas illustrate when it is possible to detect either antibodies, antigen or
IFN-γ. As early as two weeks after infection, CMI can be measured. This response is associated with
both delayed type hypersensitivity (DTH) responses, as measured by the tuberculin skin test, and
production of IFN-γ. This response is maintained throughout the course of infection, but may wane in
individuals who develop severe tuberculosis. Mycobacterial load remains low in this early phase of
infection and is therefore not detectable by antigen detection assays. However, in individuals with
acute disease, the mycobacterial load and soluble antigens increase and can be detected in the blood.
Similarly, antibodies are not detectable in the early phase of disease, but only when the infection
progresses to active tuberculosis [2].
Figure 2
Schematic representation of the spectrum of responses of the bovine immune system to various tests for
tuberculosis [17]
8
2) Measurement of the CMI
The tuberculin skin test, which measures a delayed-type hypersensitivity (DTH)
response based on immunological recognition of mycobacterial antigens in exposed
animals [2], shows poor sensitivity (16.7%) and specificity (74.2%) in elephants [5].
Similar to the Mantoux test applied in humans, a small volume of purified protein
derivative (PPD), prepared by precipitation of proteins of heat-killed cultures of M.
tuberculosis, is injected into the skin and skin reactivity is measured 72 hours later.
As most of the protein components of the PPD are shared between mycobacterial
species, elephants sensitized by prior exposure to non-tuberculosis mycobacteria may
respond to PPD the same way as elephants that have actually been infected with M.
tuberculosis [3]. Another drawback of this test is that intradermal application is not
possible in elephants. The test has to be executed on the conjunctiva, which is very
unpleasant for elephants, causing a lot of stress and restraint. A limitation of this test
is also the lack of objectivity when reading the test result, and in case of uncertainty it
cannot be repeated for 2 months due to desentization [1]. In addition elephants have to
be kept under observation for 72 hours, which may be hard to realise when screening
elephants on large scale or in the wild.
A good alternative is the IFNγ release assay. This test detects the release of IFNγ in
whole blood cultures, which have been stimulated overnight with mycobacterial
antigen, using an enzyme-linked immunosorbent assay. The IFNγ release assay is able
to detect infected animals early in disease, but the sample should be processed within
6-8 hours. This is often impossible in field conditions. Wood and Jones (2001)
evaluated several studies on sensitivity and specificity of conventional tests in cattle
in different countries, and reported a median sensitivity of 87.6% and a median
specificity of 96.6% of this IFNγ release assay [17].
3) Serology
As illustrated in figure 1 and 2, antibodies can be detected in the blood during
progression of the infection. Alternatives to the tests mentioned above, are rapid blood
tests such as the Elephant TB STAT-PAK test and, as yet experimental, the Dual
pathway platform (DPP) test, both using lateral flow technology [11]. These tests
show high sensitivity and specificity, for example the 100% sensitivity and 97%
specificity claimed by the producer of the TB-STAT-PAK test [22]. Rapid tests are
quick and handy to use, but they are too expensive to be used on large scale in
Thailand and other developing countries. The latter is also true for the MAPIA (multiantigen print immuno assay), that defines specificity of antibodies in more detail [3].
Another, cheaper possibility to detect antibodies upon progression of infection is the
multiple-antigen ELISA test. Larsen et al. (2000) found a 100% and a 95% specificity
of a multiple-antigen ELISA test employing CF, PPD, MPB70, ERD, RA and
AVPPD antigens [10]. The ESAT-6 and CFP10 antigens, which can be found in M.
tuberculosis and M. Bovis, may also be used in this ELISA test. Greenwald et al.
(2009) found that ESAT-6 and CFP10 are the immunodominant antigens recognized
by elephant antibodies during disease [8]. In addition, Aagaard et al. (2006) tested
several antigens on their potential as diagnostic markers for tuberculosis in cattle,
using MAPIA, and found ESAT6 and CFP10 to be the superior diagnostic antigens.
They also found a synergistic effect on sensitivity when combining these antigens [1].
The genes encoding for ESAT6 and CFP10 antigens are absent in most nontuberculous mycobacteria and deleted in all BCG vaccine strains, minimizing the risk
9
of false positive results [2]. The ELISA blood test operating ESAT6 and CFP10
antigens should thus be able to diagnose animals infected with TB, although only
upon progression of infection.
The IFNγ release assay and the multiple-antigen ELISA test leave behind a lot of
limitations of the tests described above. Both tests require only one blood sample,
show a prompt result, and are cheaper than the commercial rapid tests. The IFNγ
assay detects animals early in disease but is impractible for use in the field as the
sample has to be processed within 6-8 hours; the multiple-antigen ELISA test detects
animals during the more progressive phase of disease and is more suitable for
population screening as the samples can be frozen and tested later. It is important to
develop these tests and asses their reliability in order to map the prevalence of
tuberculosis in domesticated elephants in Thailand, asses the risks for both animals
and humans, and establish an effective disease management.
10
Materials and method
Set up of the research project
The aim of this research project was to contribute to the development of the IFNγ
capture assay and the multiple-antigen indirect ELISA test, which both have not been
developed and evaluated for use in Asian elephants yet. In Kasetsart University,
Bangkok, I contributed to the development of the IFNγ capture assay, as depicted in
figure 3. I worked on the expression of IFNγ by insect cells, to be used as a positive
control during development of the assay; potentially for immunisation of mice to raise
monoclonal antibodies for capture and detection of IFNγ in a capture ELISA. In
Chiang Mai University, I run three different iELISA tests, employing the recombinant
antigens ESAT6, CFP10, and MPB83, on 144 elephant sera and compared the result
with the TB STAT-PAK rapid test outcome and TB status (confirmed by trunk wash
culture or necropsy) of the elephants.
 Part I: Expression of glycosylated IFNγ from insect cells
In order to express IFNγ from insect cells, a plasmid containing an IFNγ gene had to
be constructed. After synthesizing this plasmid, the next step was to transfect the
plasmid into insect cells in order to produce glycosylated IFNγ.
IFNγ Capture ELISA
Color reaction: read
optical density
Add substrate
Rabbit polyclonal antibodies to
chicken anti-IFNγ
Mouse/chicken anti IFNγ
polyclonal antibodies
Serum containing IFNγ
Plate coated with monoclonal
antibodies specific to recombinant IFNγ
Figure 3
Schematic presentation of the IFNγ capture ELISA (modification of design of prof. dr. V. Rutten)
11
Construction of a T&A vector
We started with dead BL21 cells, stored in glycerol at -80°C, containing a pET15
vector with IFNγ insertion. After taking some colonies into 20 μl of TE, boiling them
at 10 minutes at 100°C, and centrifuging for 5 minutes at 13.000 rpm; we run PCR on
the supernatant with 52-IFN-1W and 52-IFN-2W primers (see table 1 for sequence)
with annealing at 53°C, 30 cycles. 5 μl of the PCR product was loaded into a 1.5%
agarose gel and checked with UV-light camera. The bands were cut from the gel and
purified with GeneJETTM Gel Extraction Kit (Fermentas). The product was ligated into
a T&A vector (T&A Cloning Vector Kit (RBC Bioscience) as indicated on the kit
instructions and incubated for 15 hours at 15°C. DH5α E.coli cells were prepared
following instructions of TransformAidTM Bacterial Transformation Kit (Fermentas)
and incubated overnight. The T&A vector was then transformed into the DH5α cells
on an agar plate containing 100μg/ml ampicilline. We incubated the plate for 16 hours
in a water-jacketed incubator at 37°C. The next day, 20 colonies were picked from the
plate and checked with PCR using 52-IFN-1W and 52-IFN-2W primers (annealing at
53°C, 30 cycles). We cut the band from the gel, purified the DNA from the gel using
the GeneJETTM Gel Extraction Kit (Fermentas) and send it for sequencing. In order
not to lose time, we did not wait for the sequencing result but continued the work. We
grew some colonies, which showed IFNγ insertion (as judged by size of the PCR
product) into 20 ml LB broth containing 100µg/ml ampicilline, 18 hours in a shaking
incubator at 37°C, 250 rpm. The plasmids were extracted from the LB broth using the
Nucleospin® Plasmid (Macherey-Nagel) kit and extraction was checked with gel
electroforesis.
Cutting the gene from the vector
The next step was to cut the gene from the plasmid to enable transfection into a pMT
vector. We used the FastDigestTM EcoRI kit (Fermentas), employing two partial cuts.
As depicted in figure 4, our plasmid contained three different cutting sites for BglII.
By employing a first partial cut with BglII and a second cut with KpnI, we would be
sure that the fragment obtained, would contain our gene in the right alignment and not
reversed (when employing a double digestion, you need sequencing to be sure that
your gene is not reversed). We tried this digestion procedure several times, using
different concentrations of plasmid (up to 10 μl), different incubating times (up to 3
hours), and new buffer. Unfortunately, we only succeeded the first cut. Sequencing
showed us a normal alignment so the reason for this failed cutting was not
understood. That’s why we constructed a new T&A vector with different cutting sites,
as depicted in figure 5. We run PCR using 52-IFN-2W & 52-IFN-3W primers
(annealing at 53°C, 30 cycles) on the colonies containing the former T&A vector, and
ligated the PCR product into a new T&A vector using the same cloning kit. We
transformed, checked the colonies with PCR, and grew the colonies in LB broth as
described above. After extraction of the plasmid and checking the extraction with gel
electrophoresis to assess PCR product sizes, we tried to cut the plasmid again. We
tried a first partial cut with KpnI and a second cut with BamHI. Both procedures were
executed with 10 μl of plasmid, incubating for 3 hours at 37ºC, and a final enzyme
inactivation step for 10 minutes at 80ºC. Again only the first cut succeeded. We
rechecked our plasmids again with different primersets, but they all showed IFNγ
insertion. We reconsidered the cutting process. As the gel showed us only one band
after the first cut, we were sure that the gene alignment was correct, so we could do a
double digestion. We added both enzymes at the same time and after 3 hours at 37ºC
and 10 min at 80 ºC, we finally succeeded to cut the gene from the plasmid.
12
Transformation of the gene into pMT vector and transfection into insect cells
As we also succeeded the digestion of our BiP/V5-HisA pMT vector, we transformed
our gene into the pMT vector using the pMT/BiP/V5-HisA, B, C kit (Invitrogen). The
product was transformed into DH5α cells and checked with PCR using MTfw and 52IFN-3W primers. Some clones were grown into LB broth, and their plasmids were
extracted with the PureYieldtm Plasmid Miniprep System (Promega), containing a
Endotoxin Removal Wash which removes proteins, RNA and endotoxines from the
plasmid. The plasmid was then transfected into Schneider 2 cells (Drosophila
melanogaster), using the Drosophila Expression System (DES®, Invitrogen).
Expression of IFNγ by these cells is still experiencing some problems.
Figure 4
Schematic representation of the first T&A vector with gene
insertion (green box) and its cutting sites using 52-IFN-1W
& 52-IFN-2W primers. The arrows indicate the preferred
cutting sites.
KpnI
BglII
BglII BglII
Figure 5
Schematic representation of the second T&A vector with
gene insertion (green box) and its cutting sites using 52-IFN2W & 52-IFN-3W primers. The arrows indicate the
preferred cutting sites.
KpnI KpnI
Primer name
52-IFN-1W
52-IFN-2W
52-IFN-3W
BamHI BamHI
Sequence (5’ to 3’)
GAAGATCTTTTTTKARAGARATAVMRAWC
GGGGTACCDYTGATGCTCTCYGGCCT
ggATCCTTTTTKARAgARATAVMRAWC
Company Name
Bio Basic Inc.
Bio Basic Inc.
Bio Basic Inc.
Table 1
Sequences of primers used in this experiment
13
 Part II: evaluation of multiple-antigen iELISA blood test & rapid test
Elephants
144 Blood- and trunk wash samples were taken from 25 elephants from the National
Elephant Institute hospital, Lampang, Thailand, during a period of five years. The
elephants differ in age, sex, origin and background. Some elephants had already died.
Samples
Trunk wash samples were collected by instilling 150 ml of saline into the trunk and
recollecting the liquid into a plastic bag. If the elephant was properly trained, this
procedure was repeated two more times. The samples were stored at -20°C and
processed within a week. Blood samples were collected from the ear vein, processed
to sera, and stored at -20°C.
Tests
A. Trunk wash culture
Trunk wash culture samples were cultured following the standard operating procedure
for trunk wash culture from dr. J.B. Payeur, United states Department of Agriculture,
National Veterinary Services Lab. Only the L-J slant & L-J gruft slant media have
been used in this experiment.
B. Commercial test kit (Rapid Test, RT)
The Elephant TB STAT-PAK® Assay (Chembio) was used to test the elephant sera,
as indicated on the test kit’s instructions. The test results were interpreted by eye.
Unclear results were considered negative.
C. iELISA
An indirect iELISA procedure employing three different antigens was evaluated for
its ability to detect tuberculosis infection in elephants. As ESAT6 and CFP10 antigens
have been proved to be immunodominant antigens [1][8], a CFP10/ESAT6
combination and CFP10 only were used as antigens to coat the plates. Unfortunately,
there was no purified ESAT6 available during this project. MPB83, an antigen
unspecific for the MTB complex but included in the commercial rapid test, was also
used as iELISA antigen in order to help identify false positive results with the rapid
test. It is beyond the scope of this project to address this subject in more detail. For a
schematic representation of the iELISA test, more details on the antigens used and the
exact iELISA procedure, see appendix 1. On each iELISA plate, negative and positive
controls were added, as depicted in table 2. Our positive control was an Asian
elephant from Thailand, confirmed positive for tuberculosis by necropsy and trunk
wash. The negative controls were an Asian elephant living in Thailand and an African
elephant living in the Netherlands. As the incidence of tuberculosis in the Netherlands
in both the animal and the human population is almost zero, this serum is considered
to be a real negative control.
14
Data analysis
Tuberculosis (TB) status (infected-not infected), confirmed by trunk wash culture
and/or necropsy, was used as a gold standard to evaluate the iELISA and TB STATPAK test results. As TB status was only confirmed for 21 elephants, 118 out of the
144 samples were used for data analysis. From the raw iELISA data, the mean OD
from each sample (added in triplicate), 30 minutes after adding ABTS, was calculated
by Excel. Next, the S/P ratio, which corrects for the differences in between plates, was
calculated by hand for each sample using the formula
S/P:
mean OD sample - mean OD negative control
x 100%
mean OD positive control - mean OD negative control
Computer program STATA/SE 9.2™ was used to analyze these data. iELISA results
with a negative S/P ratio were introduced into the program as negative values. A twosample t-test with 166 degrees of freedom and α=0.05 was used to compare the mean
test outcomes of samples from both the noninfected and infected elephants, testing
H0: μ+=μ- and HA: μ+≠μ-. The SEM was calculated for each mean S/P ratio. For each
iELISA test, the program was instructed to show a detailed sensitivity and specificity
report and receiver operating characteristic (ROC) curve, which shows the sensitivity
and 1-specificity of each cut off point in more detail. As the iELISA test will be used
as a screening test, a high sensitivity is more important than a high specificity. The
following criteria were taken into account when determining the cut off points: 1. high
sensitivity 2. high percentages of correctly classified values 3. high area under ROC
curve 4. high as possible specificity 5. high positive and negative predictive value
(calculated with Win Episcope 2.0). When comparing the ROC areas of different cut
off points, the chi-square test was used to determine if the difference in the area under
ROC curve was significant. Values with the most optimal criteria were established as
cut off points. After determining these cut off points, STATA™ was used to design a
2x2 table and the Win Episcope 2.0 program was used to determine sensitivity,
specificity, positive and negative predictive values and 95% confidence intervals of
each test in more detail. As we had only some trunk wash results and only one true
positive animal, this diagnostic test was not evaluated but only used to confirm the TB
status of the elephants.
1
5
9
13
17
21
25
29
1
5
9
13
17
21
25
29
1
5
9
13
17
21
25
29
2
6
10
14
18
22
26
Th +
2
6
10
14
18
22
26
Th+
2
6
10
14
18
22
26
Th +
3
7
11
15
19
23
27
Th -
3
7
11
15
19
23
27
Th -
3
7
11
15
19
23
27
Th -
4
8
12
16
20
24
28
NL -
4
8
12
16
20
24
28
NL -
4
8
12
16
20
24
28
NL -
Table 2
Schematic representation of the ELISA plate alignment. On each plate, 29 sera in triplicate, and three
controls, were added. Th+ is a Thai Asian elephant confirmed positive for TB, Th- is a Thai Asian
elephant confirmed negative for TB, and NL- is a Dutch African elephant confirmed negative for TB.
15
Results
 Part I: Expression of glycosylated IFNg from insect cells
Prof. dr. W. Wajjwalku and his team of the faculty of veterinary medicine, Kasetsart
University, Bangkok, were still working on the expression in IFNγ by the time this
paper was written. In the pictures below, some intermediate results are shown.
Figure 6
Dead BL21 cells showing IFNγ
gene insertion using 52-IFN-1W
& 52-IFN-2W primers.
Figure 7
T&A clones no.1,2,4,5,8-11 showing IFNγ gene insertion using
IFN-1W & M13re primers. NC=negative control.
Figure 8
First partial cut of the second T&A vector
with KpnI enzyme(succeeded).
MK=marker, C=T&A vector,
cut=T&A vector digestion product.
Figure 9
Second cut of pMT and second T&A vector using
BamHI enzyme (FAILED: only one band!).
MK=marker, PC/IC=intact pMT/T&A vector,
Pcut/Icut=pMT/T&A vector digestion product.
Figure 10
Double digestion of second T&A vector
with KpnI & BamHI enzymes (succeeded).
Only the T&A vector shows two bands as
the fragment cut from the pMT vector is
too small to be shown in the gel.
PMT=pMT vector digestion product,
E2=T&A vector digestion product.
16
 Part II: evaluation of multiple-antigen iELISA blood test & rapid test
TB STAT-PAK test: association with TB status
Pearson Chi-square testing indicated that there is an association between the rapid test
result and elephant TB status. Further test characteristics are discussed below.
iELISA: mean S/P ratio and hypothesis testing
Two sample t-tests determined significant differences in mean S/P ratios between the
infected and noninfected groups with ESAT6/CFP10, CFP10 and MPB83 iELISA
testing (see table 3). This means that the null hypothesis H0: μ+=μ- is to be rejected.
When calculating the standard error of the mean (SEM), it shows us that the given
mean S/P ratios are quite adequate estimates of the population mean, especially for
the ESAT6/CFP10 iELISA test. This also proves the null hypothesis can be rejected.
iELISA
Infected (n=20)
Noninfected (n=98)
ESAT6/CFP10
75.70 (± 7.75)
17.15 (± 1.32)
CFP10
58.80 (± 9.12)
23.13 (± 1.33)
MPB83
41.94 (± 12.07)
26.95 (±1.70)
Table 3
Mean S/P ratio and standard error of the mean (SEM) of different iELISA tests, calculated for samples
from infected and noninfected elephants. The null hypothesis H0: μ+=μ- can be rejected. The SEM
shows that the given S/P ratios are quite adequate estimates of the population mean.
iELISA: cut off points
The ROC curve of the ESAT6/CFP10, CFP10 and MPB83 iELISAs are shown in
figure 11, 12 and 13. The ESAT6/CFP10 iELISA shows the most optimal curve with
cut off points with a high sensitivity and specificity; the curves of CFP10 and MPB83
iELISA are less optimal. The red arrows in each curve indicate the cut off points
which have been compared in the figures 14-16 and tables 4-6. For the ESAT6/CFP10
iELISA, nwec2 with a value of 55.14 was chosen as the cut off point. Although
nwec1 showed the highest sensitivity, nwec2 showed a significantly higher area under
the ROC curve (chi-square tested), higher correctly classified percentage, a much
higher positive predictive value (88.24% for nwec2 versus 64% for nwec1) and a
higher specificity (see figure 14 and table 4). Choosing a cut off point for the CFP10
iELISA was somewhat easier. Although its somewhat lower percentage of correctly
classified values, ncfp1 with a value of 32.25 shows the highest sensitivity and a
significantly higher area under the ROC curve (see figure 15 and table 5). Finally, for
the MPB83 test, determining the cut off point was a real challenge. Nmpb3 with a
value of 28.56 seems to be the best pick because of its much higher percentage of
correctly classified values and a significantly higher are under the ROC curve (see
figure 16 and table 6).
iELISA and TB STAT-PAK: test characteristics
Table 7 shows the sensitivity, specificity, positive predictive value, negative
predictive value and the confidence intervals for all iELISA tests and the TB STATPAK test, as calculated by Win Episcope 2.0. The rapid test’s sensitivity of 80% and
specificity of 87.23% do not agree with the 100% sensitivity and 97% specificity
claimed by the producer [22]. In this experiment, the rapid test and the ESAT6/CFP10
iELISA seem to have the most optimal test characteristics. The rapid test shows a 5%
higher sensitivity, but the ESAT6/CFP10 iELISA test shows a higher specificity and
much higher positive predictive value. As the positive predictive value stands for the
17
0.50
0.00
0.25
Sensitivity
0.75
1.00
probability that an elephant with positive test result is correctly diagnosed, this is also
an important measure for diagnostic methods. And when we consider the sensitivity
and specificity calculated by STATA™, the ESAT6/CFP10 iELISA test seems be
superior to the rapid test as the sensitivities of both tests are equal but the specificity
and positive predictive value of the ESAT6/CFP10 test are higher.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.8923
0.50
0.25
0.00
Sensitivity
0.75
1.00
Figure 11
ROC curve of all possible cut off values of ESAT6/CFP10 iELISA . The red arrows indicate possible cut
off points.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.7997
Figure 12
18
0.00
0.25
0.50
Sensitivity
0.75
1.00
ROC curve of all possible cut off values of CFP10 iELISA. The red arrows indicate possible cut off
points.
0.00
0.25
0.50
1 - Specificity
0.75
1.00
Area under ROC curve = 0.5171
0.50
0.25
0.00
Sensitivity
0.75
1.00
Figure 13
ROC curve of all possible cut off values of MPB83 iELISA. The red arrows indicate possible cut off
points.
0.00
0.25
0.50
1-Specificity
nwec1 ROC area: 0.8541
nwec3 ROC area: 0.85
0.75
1.00
nwec2 ROC area: 0.8648
Reference
Figure 14
ROC curve of nwec1, nwec2 and nwec3 cut off points of the ESAT6/CFP10 iELISA test.
Cut off point
Value
Sens.
Spec.
Correctly classified
ROC area
Nwec1
32.85
85%
90.82%
89.83%
0.8541
Nwec2
55.14
80%
97.82%
94.92%
0.8648
Nwec3
69.81
75%
100%
95.76%
0.85
Table 4
Characteristics of three possible cut off points for ESAT6/CFP10 iELISA as calculated by STATA™.
Nwec2 was chosen as a cut off point because of its high sensitivity, high percentage of correctly
classified values, a significantly higher area under the ROC curve and its high predictive value.
19
1.00
0.75
0.50
0.00
0.25
Sensitivity
0.00
0.25
0.50
1-Specificity
nwcfp1 ROC area: 0.7429
nwcfp3 ROC area: 0.7138
0.75
1.00
nwcfp2 ROC area: 0.7337
Reference
Figure 15
ROC curve of nwcfp1, nwcfp2 and nwcfp3 cut off points of the CFP10 iELISA test.
0.50
0.25
0.00
Sensitivity
0.75
1.00
Cut off point
Value
Sens.
Spec.
Correctly classified
ROC area
Ncfp1
32.25
70%
78.57%
77.12%
0.7532
Ncfp2
35.41
65%
86.73%
83.05%
0.7355
Ncfp3
38.16
60%
87.76%
83.05%
0.7185
Table 5
Characteristics of three possible cut off points for CFP10 iELISA as calculated by STATA™. Ncfp1
was chosen as the cut off point as this value shows the highest sensitivity and a significantly higher
area under the ROC curve.
0.00
0.25
0.50
1-Specificity
nwmpb1 ROC area: 0.4888
nwmpb3 ROC area: 0.5357
0.75
1.00
nwmpb2 ROC area: 0.5194
Reference
Figure 16
ROC curve of nmpb1, nmpb2 and nmpb3 cut off points of the MPB83 iELISA test.
Cut off point
Value
Sens.
Spec.
Correctly classified
ROC area
Nmpb1
19.64
70%
37.76%
43.22%
0.5138
Nmpb2
22.54
60%
43.88%
46.61%
0.5194
Nmpb3
28.56
55%
57.14%
56.78%
0.5357
Table 6
Characteristics of three suitable cut off points for MPB83 iELISA as calculated by STATA™. Nmpb3
was chosen as the cut off point as much higher percentage of correctly classified values and a
significantly higher area under the ROC curve.
20
Test
Sensitivity
Specificity
+ Pred. value
- Pred. value
TB STAT-PAK
80%
(62.47-97.53%)
75%
(56.02-93.98%)
70%
(49.92-90.08%)
50%
(28.09-71.91%)
87.23%
(80.49-93.98%)
97.82%
(95.16-100%)
78.57%
(70.45-86.67%)
57.14%
(47.35-66.94%)
57.14%
(38.81-75.47%)
88.24%
(72.92-100%)
40%
(23.77-56.23%)
19.23%
(8.52-29.94%)
95.35%
(90.90-99.80%)
95.05%
(90.82-99.28%)
92.77%
(87.02-98.34%)
84.85%
(76.2-93.50%)
ESAT6/CFP10
iELISA
CFP10 iELISA
MBP83 iELISA
Table 7
Overview of test characteristics of the tests used in this experiment using a 95% confidence interval as
calculated by Win Episcope 2.0.
Test
Sensitivity
Specificity
TB STAT-PAK
80%
84.69%
ESAT6/CFP10 iELISA
80%
97.82%
CFP10 iELISA
70%
78.57%
MBP83 iELISA
55%
57.14%
Table 8
Overview of test characteristics of the tests used in this experiment, as calculated by STATA™.
21
Discussion
The aim of this research project was to contribute to the development of an IFNγ
assay by working on the expression of IFNγ by insect cells, and to evaluate iELISA
tests employing ESAT6/CFP10, CFP10 & MBP83 antigens and the commercial TB
STAT-PAK rapid test.
As concerned the IFNγ expression, this process is not finished yet. We experienced a
lot of problems with the digestion of the T&A vector. This may be caused by a low
accuracy of the commercial kit that was used. This kit worked with the volume of
DNA and enzymes instead of concentration, which is less specific. It would have
been better to determine the concentration of the DNA and adjust the amount of
enzyme to this concentration, in order to optimize the digestion process. Eventually
the investigators succeeded in constructing a pMT/BiP/V5-HisA vector carrying an
IFNγ gene insertion (as assessed by PCR product size). This plasmid has been
transferred into Schneider 2 insect cells (Drosophila melanogaster), using the
Drosophila Expression System (DES®, Invitrogen). A stable expression of IFNγ by
these cells has not been established yet
Compared to the 100% sensitivity and 97% specificity claimed by the producer of the
TB STAT-PAK test, in this project a sensitivity of 80% and a specificity of 87.23%
were found (using Win Episcope 2.0 software). The iELISA test employing the
immunodominant antigens ESAT6 & CFP10, showed a sensitivity of 75% and a
specificity of 97.82%, and a positive predictive value of 88.24% opposed to the
positive predictive value of 57.14% of the TB STAT-PAK test. Calculation of
sensitivity and specificity with the STATA™ program showed an equal sensitivity of
80% for both the ESAT6/CFP10 iELISA and the rapid test and a higher specificity
(97.82%) of the iELISA test compared to the 84.69% sensitivity of the TB STATPAK test. This means that the ESAT6/CFP10 iELISA could be a good alternative to
the expensive TB STAT-PAK test when screening for TB infection in Asian
elephants.
As in every investigation, there are quite some remarks about the statement above. As
the TB status was only confirmed for 21 of the 25 elephants, only 118 out of 144
samples were used in the research. In the near future, the trunk wash culture result
will be known of the additional four elephants, extending the sample group to 144
samples from 25 different elephants. Still, these samples originate from only a few
animals, resulting in dependent data. In this data analysis, the data have been
considered to be independent which may have resulted in an inaccurate calculation of
test characteristics. Additionally, elephants were considered to be infected when
necropsy and/or trunk wash culture gave a positive result. Trunk wash culture is
considered the gold standard, although there is growing body of evidence that this test
has a low sensitivity [10][11]. Thus, this test might have missed out on some positive
elephants, affecting the test characteristics of the iELISAs and the rapid test, although
the elephants have been under close health supervision and no clinical signs of the
negative animals have been reported yet. Necropsy and culture would be the ideal
gold standard in this investigation, but considering the life span of the elephant this
was not possible here.
22
Other factors that could have influenced the results are the difference in sera quality
(as they have been collected under different circumstances by different veterinarians,
during five years) and the occurence of quite some negative OD values and thus
negative S/P ratios in the data (due to the use of a negative control which wasn’t
really negative?). It would be useful to develop a method to account for this negative
data and recalculate S/P ratios and test characteristics, or run the iELISA tests again
with a different negative control. During this improved analysis, I would recommend
to use only one software program instead of the two different software programs used
in my analysis.
Finally, the rapid test outcome in this research might be distorted as the test kits were
expired two months prior to testing. But as each testing device showed a clear control
line, I suspect minimal effect of this expiry on the test outcomes. In this research a
first generation TB STAT-PAK test has been used, which employs less antigens and
probably has a lower sensitivity than the second and third generation tests. It would
have been be better to use a second or third generation test for our samples, but
unfortunately, money affected our choice of the test.
23
Acknowledgement
Great thanks to prof. dr. Victor Rutten, who made it possible for me to come to
Thailand and join this research about elephant tuberculosis. Thank you for your
commitment and supervision of the process, although you were many miles away.
I would like to thank prof. dr. Worawidh Wajjwalku for his dedication to my research
project and for taking care of me during my stay at his faculty. By taking me on his
trips to the Wildlife Sanctuaries, he showed me the importance of the preservation of
the Thai forests and their wildlife. Thank you, dr. Anucha Sirimalaisuwan, for your
support during my stay in Chiang Mai and the help with my accommodation and data
analysis.
I would like to thank Taweepoke Angkawanish for letting me join his research about
tuberculosis in elephants. He showed me the meaning of hard working and taught me
a lot of understanding and respect for elephants in Thailand and the National Elephant
Institute hospital.
And finally I would like to thank the Thai master/PHD students Manakorn Sukmak,
who teached me about the lab work and who got never tired of my never ending
questions, and Suthathip Dejchaisri, who always helped me with my questions and
issues besides the research. And thanks to all my Thai friends, who made my stay in
Thailand a time to never forget!
24
References
1] Aagaard, C., Govaerts, M., Meikle, V., Vallecillo, A.J., Gutierrez-Pabello, J.A.,
Suarez-Guemes, F., McNair, J., Cataldi, A., Espitia, C., Andersen, P. & Pollock, J.M.
Optimizing antigen cocktails for detection of Mycobacterium bovis in herds with
different prevalences of bovine tuberculosis: ESAT6-CFP10 mixture shows optimal
sensitivity and specificity, 2006. Journal of clinical microbiology, vol. 44, no. 12, pp.
4326-4335.
2] Andersen, P., Doherty, T.M., Pai, M. & Weldingh, K. The prognosis of latent
tuberculosis: can disease be predicted? 2007. Trends in Molecular Medicine, vol. 13,
no. 5, pp.175-183.
3] Andersen, P., Munk, M.E., Pollock, J.M. & Doherty, T.M. Specific immune-based
diagnosis of tuberculosis, 2000. Lancet, vol. 356, pp 1099-1104.
4] Davis, M. Mycobacterium tuberculosis risk for elephant handlers and
veterinarians, 2001. Applied occupational and environmental hygiene, vol 16, no. 3,
pp. 350-353.
5] Fowler, M.E. & Mikota, S.K. Biology, medicine and surgery of elephants, 2006.
Blackwell publishing, 1st edition, USA.
6] Furley, C.W. Tuberculosis in elephants, 1997. Lancet, vol. 350, p. 224.
7] Guerrant, R.L., Walker, D.H. & Weller, P.F. Tropical infectious diseases –
Principles, Pathogens, & Practice, 2006. Elsevier Churchill Livingstone, 2nd edition,
USA, pp. 394-410.
8] Greenwald,R., Lyashchenko, O., Esfandiari, J., Miller, M., Mikota, S., Olsen, J.H.,
Ball, R., Dumonce & Lyashchenko, K.P. Highly accurate antibody assays for early
and rapid detection of tuberculosis in African and Asian elephants, 2009. Clinical and
vaccine immunology, vol. 16, no. 5, pp. 605-612.
9] Hernandez-Pando, R., Orozcoe, H., Sampieri, A., Pavon, L., Velasquillo, C.,
Larriva-Sahd, J., Alcocer, J.M. & Madrid, M.V. Correlation between the kinetics of
Th1/Th2 cells and pathology in a murine model of experimental pulmonary
tuberculosis, 1996. Immunology, vol. 89, no. 1, pp. 26-33.
10] Larsen, R.S., Salman, M.D., Mikota, S.K., Isaza, R., Montali, R.J. & Triantis, J.
Evaluation of a multiple-antigen enzyme-linked immunosorbent assay for detection of
Mycobacterium tuberculosis infection in captive elephants, 2000. Journal of Zoo and
Wildlife Medicine, vol. 31, no. 3, pp. 291-302.
11] Lyashchenko, K.P., Greenwald, R., Esfandiari, J., Olson, J.H., Ball, R.,
Dumonceaux, G., Dunker, F., Buckley, C., Richard, M., Murray, S., Payeur, J.B.,
Andersen, P., Pollock, J.M., Mikota, S., Miller, M., Sofranko, D. & Waters, W.R.
Tuberculosis in elephants: antibody responses to defined antigens of Mycobacterium
tuberculosis, potential for early diagnosis, and monitoring of treatment, 2006.
Clinical and Vaccine Immunology, vol. 13, no. 7, pp. 722-732.
25
12] Michalak, K. Austin, C., Diesel, S., Bacon, J.M., Zimmerman, P. & Maslow, J.N.
Mycobacterium tuberculosis infection as a zoonotic disease: transmission between
humans and elephants, 1998. Emerging infectious diseases, vol. 4, no. 2, pp. 283-287.
13] Mikota, S.K., Peddie, L., Peddie, J., Isaza, R., Dunker, F., West, G., Lindsay, W.,
Larsen, R.S., Salman, M.D., Chatterjee, D., Payeur, J., Whipple, D., Thoen, C., Davis,
D.S., Sedgwick, C., Montali, R.J., Ziccardi, M. & Maslow, J. Epidemiology and
diagnosis of Mycobacterium tuberculosis in captive Asian elephants (elephas
maximus), 2001. Journal of zoo and wildlife medicine, vol. 31, no.1, pp. 1-16.
14] Oh, P., Granich, R., Scott, J., Sun, B., Joseph, M., Stringfield, C., Thisdell, S.,
Staley, J., Workman-Malcolm, D., Borenstein, L., Lehnkering, E., Ryan, P., Soukup,
J., Nitta, A. & Flood, J. Human exposure following Mycobacterium tuberculosis
infection of multiple animal species in a metropolitan zoo, 2002. Emerging infectious
diseases, vol. 8, no. 11, pp. 1290-1293.
15] Orme, I.M., Roberts, A.D., Griffin, J.P. & Abrams, J.S. Cytokine secretion of
CD4 T lymphocytes acquired in response to Mycobacterium tuberculosis infection,
1993. Journal of immunology, vol. 151, pp. 518-525.
16] Quinn, P.J., Markey, B.K., Carter, M.E., Donnelly, W.J. & Leonard, F.C.
Veterinary microbiology and microbial disease, 2002. Blackwell Publishing, 1st
edition, United Kingdom, pp. 97-102.
17] de la Rua-Domenech, R., Goodchild, A.T., Vordermeier, H.M., Hewinson, R.G.,
Christiansen, K.H., Clifton-Hadley, R.S. Ante mortem diagnosis of tuberculosis in
cattle: a review of the tuberculin skin tests, γ-interferon assay and other ancillary
diagnostic techniques, 2006. Research in veterinary science, vol. 81, pp. 190-210.
18] Sarma, K.K., Bhawal, A., Yadav, V.K., Saikia, G. & Das, J. Investigation of
tuberculosis in captive Asian elephants of Assam vis-à-vis its cross infection with the
handlers, 2006. Intas polivet, vol. 7, no. 2, pp. 269-274.
19] Sreekumar, E., Janki, M.B.V., Arathy, D.S., Hariharan, R., Avinash Premraj, C. &
Rasool, T.J. Molecular characterization and expression of interferon-γ of Asian
elephant (Elephas maximus), 2007. Veterinary Immunology and Immunopathology,
vol. 118, pp. 75-83.
20] Une, Y. & Mori, T. Tuberculosis as a zoonosis from a veterinary perspective,
2006. Comparative immunology, microbiology & infectious diseases, vol. 30, 415425.
21] Widdison, S., Schreuder, L.J., Villarreal-Ramos, B., Howard, C.J., Watson, M. &
Coffey, T.J. Cytokine expression profiles of bovine lymph nodes: effects of
Mycobacterium bovis infection and bacille Calmette-Guerin vaccination, 2006.
Clinical and experimental immunology, vol. 144, no. 2, pp. 281-289.
22] Website Chembio Diagnostic Sytems: http://www.chembio.com
26
23] Website EU-Asia link project: http://www.vet.cmu.ac.th/elephant/index.php
24] Website International Union for Conservation of Nature:
http://www.iucnredlist.org/apps/redlist/details/7140
27
Appendix 1
Multiple antigen iELISA
Color reaction: read
optical density
Add substrate: ABTS
Goat polyclonal antibodies to
rabbit polyclonal antibodies
Rabbit polyclonal antibodies to
serum antibodies
Serum containing antibodies
to ESAT-6 and CPF-10
Plate coated with ESAT-6 and
CPF10 antigens
Schematic presentation of the multiple antigen iELISA (modification of design of prof. dr. V.
Rutten)
28
iELISA procedure
1. NUNC MaxiSorp ELISA plates were coated with either an ESAT6/CFP10
antigen combination 1μ/ml (liquid solution produced at Statens Serum
Institute, Copenhagen, Denmark), CFP10 antigen 1μ/ml, (Statens Serum
Institute), or MPB83 antigen 1μ/ml, (powder dissolved in distilled water,
origin Ireland), 50 μl per well. Plates were incubated on a shaking plate for 60
minutes, and washed one time with 0.1% TWEEN 20 solution (PBS).
2. The wells were blocked with 1x blocking reagent, 200 μl per well. The plates
incubated on a shaking plate for 15 minutes and washed one time with 0.1%
TWEEN 20 solution (PBS).
3. Then, the elephant sera were added, as indicated in table 2. On each plate, 29
different elephant sera, diluted 1:800 into PBS, were added in triplicate, 60 μl
per well. Positive and negative controls were added on each plate. The plates
were incubated for 120 minutes on a shaking plate, and then washed 3 times
with 0.1% TWEEN 20 solution (PBS).
4. Monoclonal anti-elephant rabbit antibodies (obtained after six weeks of
immunization of rabbits with 500µg elephant IgG injected three times every
two weeks, and stored in sodium azide) were then added, diluted 1:8000 into
PBS, 50 μl per well. The plates were incubated for 30 minutes on a shaking
plate and then washed 3 times with 0.1% TWEEN 20 solution (PBS).
5. Purified peroxidase labeled anti-rabbit goat antibodies (KPL) dilution 1:2000
(PBS) were added 50 μl per well. The plates were incubated for 30 minutes on
a shaking plate, and then washed 5 times with 0.1% TWEEN 20 solution
(PBS).
6. Finally, ABTS substrate (1mg/ml, Roche) was added, 50 μl per well. The
optical density of the sample wells was measured with TECANtm
spectrofotometer, 405 nm, at both 15 and 30 minutes after adding the ABTS
substrate.
1
5
9
13
17
21
25
29
1
5
9
13
17
21
25
29
1
5
9
13
17
21
25
29
2
6
10
14
18
22
26
Th +
2
6
10
14
18
22
26
Th+
2
6
10
14
18
22
26
Th +
3
7
11
15
19
23
27
Th -
3
7
11
15
19
23
27
Th -
3
7
11
15
19
23
27
Th -
4
8
12
16
20
24
28
NL -
4
8
12
16
20
24
28
NL -
4
8
12
16
20
24
28
NL -
Table 2
Schematic representation of the ELISA plate alignment. On each plate, 29 sera in triplicate, and three
controls, were added. Th+ is a Thai Asian elephant confirmed positive for TB, Th- is a Thai Asian
elephant confirmed negative for TB, and NL- is a Dutch African elephant confirmed negative for TB.
29