Conceived and designed the experiments: MR IB DTT FN LR. Performed the experiments: MR SD LAR IB DTT VA FN. Analyzed the data: MR LAR IB DTT VA LR. Contributed reagents/materials/analysis tools: MR SD LAR VA FN. Wrote the paper: MR. Collected samples: MR LAR IB DTNT VA. Critically revised the article: IB DTNT VA FN. All authors approved the final version.
The authors have declared that no competing interests exist.
Tools for plague diagnosis and surveillance are not always available and affordable in most of the countries affected by the disease.
We developed two rapid immunochromatography-based tests for the detection of antibodies directed against F1 antigen of
The SIgT test is of importance for surveillance because it can detect Ig antibodies in a range of reservoir species. The HIgM test could facilitate the diagnosis of plague during outbreaks, particularly when only a single serum sample is available.
Plague is due to the bacterium
Plague, a bacterial infection caused by
Bubonic plague is the major form of the disease encountered in the modern plague outbreaks. It affects mainly rural people in developing countries, whose level of education is very low. In Madagascar, plague foci matches with the poorest rural area with the most vulnerable population
The control of plague involves diagnosis and recognition of the disease
The isolation of
A RDT for the detection of F1 antigen, a capsular antigen of
The serology for the detection of anti-F1 plague antibodies plays an important role in confirming plague diagnosis. Indeed, during the last outbreak of pneumonic plague in the Democratic Republic of Congo (DRC) in 2005, plague diagnosis was confirmed by evidence of seroconversion based on the anti-F1 IgG titre in paired serum samples
When neither the most appropriate specimen nor the pair of sera were available, the plague confirmation is “compromised”. A rapid test which can be performed with a unique serum collected during the early stage of the infection would be very helpful for the biological diagnosis.
For the surveys of plague infection foci, the sero-epidemiological or sero-epizootic investigations of anti-F1 antibody prevalence in human or animal population could be achieved. The species involved in the plague cycle have been identified in Madagascar and Viet Nam, but this is not necessarily the case in other countries, such as Algeria. ELISA is an efficient tool for the serology but it is difficult to carry out in field either for diagnosis or for surveillance. It requires specific equipment, expensive consumables and specific antibodies to each species for the revelation of anti-F1 to be detected. The available ELISA for human and rodent IgG detection were respectively 91.4% and 100% sensitive and 98.5% and 100% specific
There is still a need to develop a simple, rapid and cost-effective test for the detection of plague-specific antibodies. First, a test which can replace the ELISA and which can be used in a large-scale in the field would be beneficial. Second, another test which can be used for biological diagnosis of plague would be very helpful.
We aimed to develop two immunochromatography-based tests (dipstick). The SIgT test is able to detect total (T) anti-F1 immunoglobulin (Ig) in different species (S) (human and animal reservoirs). It was compared with ELISA. The SIgT test could have major applications in epidemiological investigations of plague and for surveillance. The HIgM test is able to detect human (H) anti-F1 IgM. Performance of the test with sera from individual with known status of plague was determined.
We developed two rapid tests based on a one-step, vertical-flow immunochromatography. The test consisted of a reaction pad, a conjugate pad and the absorbent pads. The reaction pad consists of a nitrocellulose membrane, with a pore size of 5 µm (Whatman International, Chateau Giron - France). The BioDot machine (BioDot, Irvine, California) was used to spray a line of Protein A (Pharmacia, Sweden) at a concentration of 2.5 mg/ml to capture Ig for SIgT test or a line of anti-human IgM (µ) at a concentration of 2 mg/ml (P.A.R.I.S, Compiègne – France) to capture human IgM for HIgM test. Protein A possesses two distinct Ig-binding domains, one that binds the Fc region of IgG and the other that binds the Fab region of Ig
The conjugate pad consists of a polyester Accuflow P (Whatman Schleicher & Schuell, England). The colloidal gold particles (40 nm diameter) were conjugated to F1 antigen by the British Biocell International (British Biocell International Cardiff, UK). The polyester was soaked in F1 conjugated at OD = 3 and was overnight lyophilised.
The absorbent pads (sample and wicking pads) consist of cellulose filter paper (Schleicher and Schuell, Ecquevelly - France). The sample pad was soaked in blocking buffer (0.1 M sodium borate supplemented with 1% Triton X-100) and overnight dried. The wicking pad was not processed prior to use.
A membrane-based immunochromatographic card was prepared by fixing the different pads onto a plastic card with double-sided sticky tape (Adhesives Research Ltd, Melville House United Kingdom): the reaction pad in the middle, the conjugate pad under the sample pad at the bottom and the wicking pad at the top of the card (
The dipstick test was performed on sera or seropads (blood on blotting paper). Sera were previously diluted ten times in distilled water. For seropads, one pastille was soaked in 200 µl of distilled water and incubated for one hour at 37°C or at room temperature. This step allows the elution of the serum from the filter. The test was carried out in a plastic tube containing 150 µl of diluted sample. Diluted serum samples containing antibody absorbed from the bottom of the dipstick bound to the conjugated antigen, and the antigen-antibody binding formed complexes moved by capillary action into the nitrocellulose membrane. The complexes reacted with the immobilised protein A, generating a signal. Excess of the conjugate moved on the nitrocellulose and bound to the anti-F1 Mab, giving a second signal corresponding to the control line (
A: SIgT test, B: HIgM test, C: Negative test, D: Positive test for Ig anti-F1, E: Positive test for IgM anti-F1.
The shelf life of the strip for long-term storage was assessed by testing serial dilution of the positive control sera with the dipsticks stored at 60°C twice per week for three weeks. Storage at 60°C for three weeks is equivalent of two years at room temperature
The tests were evaluated on human and canine stored sera at Institut Pasteur de Madagascar and on small mammal samples freshly collected by the Pasteur Institutes of Algeria, Madagascar and Nha-Trang or from their frozen collection. The human samples tested were anonymous frozen sera from the collection of Institut Pasteur of Madagascar. Samples without turbidity, with sufficient quantity and with clearly identification number were selected and blind tests were performed. A total of 288 sera from the following groups were included in the study:
Human samples from plague suspected cases (n = 131) were taken from sera collected during sero-epidemiological investigation in 1999 at CHU Androva-Mahajanga, the only coast plague focus in Madagascar and in 2002 at Ankazobe, one of the active foci in Antananarivo province. These sera were sampled within the framework of the routine plague diagnosis during outbreaks. Suspect plague was defined as compatible clinical and epidemiological features
Frozen sera collected from individuals in plague-free area were used as negative controls (n = 71): 24 sera from Fort-Dauphin in Madagascar collected in 2000 during malaria surveillance, 23 sera from Comoros collected in 1999 during cholera outbreak and 24 sera from Seychelles collected in 1997 during arbovirosis surveillance.
To assess the cross-reaction with other bacterial or parasitic infections prevalent in Madagascar, samples (n = 86) collected from patients with schistosomiasis, cysticercosis, toxoplasmosis, hepatitis B, hepatitis C, streptococcosis and with unknown diseases were tested. These sera were sampled within the framework of the evaluation of a serodiagnostic test in 1997.
Evaluation on specimens from dogs (n = 63) and from rodent or small-mammals (n = 352) specimens was performed by testing sera or seropads from animals in plague foci or in areas non endemic for plague in Madagascar, Viet Nam and Algeria. Canine samples were collected in 1999 within the framework of diagnosis tool development, while rodent samples were collected during plague monitoring in rodent populations. For small mammal sera collection, we adhered to our institutional guidelines for animal husbandry and experiments.
For HIgM test evaluation, sensitivity was studied with 23 acute-phase sera specimen from bacteriologically-confirmed plague cases. Only the 26 negative controls from Madagascar and the 86 sera from patients with others infectious diseases were used for studying the specificity of this HIgM test.
For SIgT test, as the purpose of this test is to be an alternative method to the ELISA in the field, ELISA was used as reference methods. Both methods were performed with each sample. Human sera were tested by human IgG anti-F1 ELISA. The sensitivity of this ELISA was 91.4%, and its specificity was 98.5%
ELISA for rodent and insectivore was standardized by testing sera from experimentally infected animals and from animals caught in plague-free areas. For rodent ELISA, its' sensitivity and specificity were 100%
For HIgM test, the evaluation was carried out with sera from individual of known plague status according to the biological diagnosis by bacteriology (culture)
The data, such identification number of the sample and the results of the two compared tests, were entered in Microsoft Excel sheet and analysed by Pivot Table reports. Sensitivity (Se) and specificity (Sp) estimates are given as percentages with 95% confidence interval.
The use of the human sera from the collection of Institut Pasteur de Madagascar, including the sera from plague patients, the sera from individual in plague–free area and the sera from patients affected by others diseases, was approved by the ethics committee of the Madagascar Ministry of Health.
SIgT test was able to detect total anti-F1 antibodies in both humans and animal reservoirs species. The lower detection limit of SIgT test was 20 ng/ml of anti-F1 MAb (G6-18). The shelf-life study indicated that SIgT test was stable after storage for 14 days at 60°C, and its threshold increased to 40 ng/ml after 21 days at 60°C. The application of SIgT test to serum samples from several species was summarized in
SIgT test | ELISA | |||||
Human | Small mammal | Dog | ||||
N (n = 223) | P (n = 65) | N (n = 278) | P (n = 74) | N (n = 49) | P (n = 14) | |
219 (98.2%) | 10 (15.4%) | 251 (90.3%) | 9 (12.2%) | 48 (98%) | 1 (7.2%) | |
4 (1.8%) | 55 (84.6%) | 27 (9.7%) | 65 (87.8%) | 1 (2%) | 13 (92.8%) |
N : Negative, P: Positive, n: sample number.
We tested a total of 131 sera from human clinically suspected plague from which 65 were positive by ELISA. Compared to ELISA, the SIgT test gave 55/65 (84.6%) positive results. Among ten SIgT-negative and ELISA-positive samples, six had a low IgG titre by ELISA and four were sera collected early. Among the 66 negative sera by ELISA, 63 were also negative with SIgT test and 3 were positive (
SIgT test was evaluated with 352 sera from animal reservoirs of plague: 88 small mammals (
Dogs respond to
We evaluated HIgM test with 23 acute-phase sera from bacteriologically-confirmed plague patient and 110 plague-negative (24 negative control from Madagascar and 86 with others infectious disease). All these sera were from the collection of Institut Pasteur de Madagascar. Results of HIgM tests are presented according to the patient status for plague (
HIgM test | Patients status | |
Plague negative (n = 110) | Plague confirmed (n = 23) | |
110 (100%) | 4 (17%) | |
0 (0%) | 19 (83%) |
We developed and evaluated two tests for the qualitative detection of plague anti-F1 antibodies in sera: the SIgT test for total Ig anti-F1 antibodies during and after plague infection in humans, rodents and other animals and the HIgM test for anti-F1 IgM in humans.
Using ELISA as reference method, SIgT test detected plague antibodies in human with a sensitivity of 84.6% and a specificity of 98% according to the reference test ELISA which sensitivity and specificity were respectively 91.4% and 98.5%
In addition, since SIgT test detects plague-specific antibodies in many species of animal reservoirs, it is suitable for large scale serological survey of reservoirs in remote and impoverished areas endemic for plague. This could help to determine the risk of plague in a given zone, leading to a progress in disease prevention. SIgT test, used for canine sera proved sensitive and specific enough for this purpose, since it provided evidence of plague antibodies production in 93% of the samples collected from area of endemic plague, whereas over 98% of the samples from areas considered free of plague tested negative. Indeed, dogs are useful sentinels of plague prevalence, since animals living in or in adjacent to areas endemic for plague may be in contact with
HIgM test was developed for the detection of anti-F1 IgM in humans to provide an alternative diagnostic method for plague, particularly when serum is the only clinical specimen available. HIgM test in plague confirmed cases gave a specificity of 100% and a sensitivity of 83%. This low sensitivity will generate some false negative results. However, of the 4 “false negative” samples; 2 were taken early (within 2 days after onset of the disease) before IgM was likely to be detectable in blood and 2 were collected 1 week after the onset of the disease. Owing to its high specificity, HIgM test could be used with significant advantage on serum samples collected during the acute phase as early as three days after onset of the disease. It could be performed with only a single serum sample while plague diagnosis by ELISA usually need a pair of sera (early and late sera collected at 4–6 weeks interval)
Our aim was to develop some simple, rapid and affordable tools for a large scale use in plague monitoring (seroepidemiological investigations) and as an alternative test to ELISA.
In the majority of endemic area, particularly in Madagascar, the poor sanitary and economic situation renders difficult the control and surveillance of plague. Bacteriology techniques including culture-isolation and mouse infection require appropriate laboratory. In developing countries, at the district level, simple tests like the dipstick assay can be implemented in the health centres. Most of the suspected cases of plague are detected in remote villages in rural area. As soon as transport of specimen from these places to a central laboratory is long and difficult, it is essential to possess an alternative tool for plague diagnosis and surveillance on site.
A pilot assessment of the new tests by users at the periphery level could be considered to define the utility and performance of these tools in field conditions.
In conclusion, the rapid serodiagnostic tests developed in this study are promising, not only for epidemiological studies, but also for the surveillance of reservoirs and active foci and for plague diagnosis. Application in case of bioterrorism attack can also be considered as
Translation of the Abstract into French by Lila Rahalison
(0.03 MB DOC)
CONSORT Checklist
(0.08 MB PDF)
We wish to thank Ratsimba Mamy and Dr Beguin Pierre for their contributions in data collection and English revision, respectively.