Development and Clinical Evaluation of sdaA Loop-Mediated Isothermal Amplification Assay for Detection of Mycobacterium tuberculosis with an Approach To Prevent Carryover Contamination

Abstract

A rapid and sensitive loop-mediated isothermal amplification assay for the sdaA gene of Mycobacterium tuberculosis was developed using a dUTP-uracil-N-glycosylase (dUTP-UNG) strategy to prevent carryover contamination. Evaluation of the assay using clinical specimens (n = 648) showed high specificity (97.2%) and sensitivity (100%), demonstrating its potential as a diagnostic test for tuberculosis, especially in resource-limited settings.

TEXT

The major challenge in combating tuberculosis (TB) is the lack of a rapid, reliable, and inexpensive diagnostic test for detection of Mycobacterium tuberculosis (1). The loop-mediated isothermal amplification (LAMP) assay is a diagnostic technique which can aid in the fight against TB in resource-poor countries. The LAMP technology amplifies DNA with high sensitivity and specificity, relying on an enzyme with strand displacement activity under isothermal conditions in autocycling reactions leading to accumulation of a large amount of the target DNA (2). LAMP-based assays targeting gyrB, rrs, IS6110, rimM, and hspX have been described for the detection of M. tuberculosis (3–7). However, there is a scope of improvement of this diagnostic method in terms of the target gene, reaction time, detection limit, and prevention of contamination. In this study, we investigated a promising sdaA-based LAMP diagnostic assay for M. tuberculosis which is highly sensitive and specific, generates visual results rapidly, and is robust to carryover contamination.

LAMP primers were designed for the 190-bp region of the sdaA gene (GenBank accession number NC_000962.3) of M. tuberculosis using the online Primer Explorer v. 4 software. This region offers excellent diagnostic sensitivity as a PCR target based on previous evaluation (8). The LAMP primer sequences, with the italicized letters indicating poly(T) linkers in the inner primers, are F3 (5′-TAGGGAAGGGCAACTGAGCA-3′), B3 (5′-AGCGTGATATCGACCTGCAT-3′), FIP (5′-CACGGAACAGACCAGCGG TTTT GGATGTTGGCCGCTGTTG-3′), and BIP (5′-CCGCGGCAGTGAACGTC TTTT GCCAACGCATCCCAACG-3′) (D. Saluja, M. Nimesh, and D. Joon, 10 January 2013, Indian Patent Office, patent application no. IN77/DEL2013).

The optimized reaction mixture (25 μl) contained 20 mM Tris-HCl (pH 8.8), 50 mM KCl, 10 mM (NH₄)₂SO₄, 0.1% Tween 20, 8 mM MgSO₄, 0.8 M betaine, 1.0 mM each deoxynucleoside triphosphate (dNTP) (dATP, dGTP, dCTP, and dUTP), 0.2 μM (each) F3 and B3 primers, 0.8 μM (each) FIP and BIP, and 8 U Bst 2.0 DNA polymerase (New England BioLabs, USA) and 3 U uracil-N-glycosylase (UNG). After a 10-min incubation at 37°C, amplification was carried out at 65°C for 45 min and inactivation at 80°C for 5 min. The products were visualized with the naked eye after addition of 2 μl of 1,000× SYBR green I (Invitrogen, USA) to the reaction mixture after amplification and agarose gel electrophoresis.

Serial dilutions of the genomic DNA (1 ng to 5 fg) of the M. tuberculosis H37Rv strain were tested in triplicate to check the analytical sensitivity of the primers. The sdaA LAMP assay detected up to 5 fg of purified M. tuberculosis DNA, which theoretically corresponds to 1.3 copies of the M. tuberculosis genome, which is same as the detection limit of the IS6110 LAMP assay (5). This detection sensitivity is greater than that of the other targets, gyrB, rrs, rim, and hspX, which showed detection limits of 50 copies, 20 copies, 200 copies, and 10 copies of the genome, respectively (3, 4, 6, 7).

The specificity of LAMP primers was tested by carrying out amplification with 1 ng of purified genomic DNA from 14 different mycobacterial species (M. avium, M. bovis [AN-5], M. chelonae, M. fortuitum, M. gordonae, M. kansasii, M. marinum, M. scrofulaceum, M. simiae, M. smegmatis, M. szulgai, M. terrae, M. tuberculosis H37Rv, and M. xenopi). Positive amplification was observed with M. bovis (AN-5) and M. tuberculosis H37Rv, whereas no amplification was obtained with nontuberculous mycobacteria.

The efficacy of the dUTP-UNG approach in preventing contamination in the LAMP assay procedure was investigated by complete replacement of dTTP with dUTP. Dilutions of LAMP products were added as the template and reaction mixture was treated with UNG, which could not be amplified again, whereas amplification was observed in the reaction mixture containing LAMP products as the template in the absence of UNG (Fig. 1).

FIG 1.

Prevention of contamination in sdaA LAMP assay by uracil-N-glycosylase (UNG). Incorporated with dUTP, different dilutions of sdaA LAMP products (10³, 10⁵, and 10⁷ dilutions), with and without UNG treatment, were used as the templates for sdaA LAMP amplification to test the anticontamination ability of UNG. Tubes/lanes 1 and 3, LAMP reaction with 10³-times diluted products; tubes/lanes 2 and 4, LAMP reaction with 10⁵-times diluted products; tubes/lanes 3 and 6, LAMP reaction with 10⁷-times diluted products. Tubes/lanes 1 to 3 were treated with UNG prior to amplification; tubes/lanes 4 to 6 were not treated with UNG. (A) Visual inspection of color change in ambient light by adding SYBR green I dye (orange indicates a negative result; green indicates a positive result). (B) Electrophoretic analysis of LAMP products on 2% agarose gel stained with ethidium bromide. Lane M, 100-bp DNA ladder.

Two groups of clinical samples were used for the present study per the Institute's ethics guidelines (ethical clearance no. F50–2/Eth.Com/ACBR/11/2108). For group I, 236 sputum specimens were collected from patients visiting the Vallabhbhai Patel Chest Institute (VPCI), New Delhi, during the period from July 2013 to August 2013. For group II, stored total DNA samples from the sputum specimens collected from 412 patients visiting the registered DOTS Center from East Delhi during the period from June 2009 to May 2012 were retrospectively analyzed with the LAMP assay. Subjects already on antitubercular therapy (ATT) at the time of specimen collection were excluded from the study.

The sputum specimens were processed by the universal sample processing (USP) method for the DNA extraction as described by Chakravorty et al. (9). The isolated DNA was stored at −20°C and used for LAMP and PCR assays. sdaA PCR was carried out according to a published protocol (8).

The results of the sdaA LAMP assay were compared with the results of smear microscopy and culture along with available clinical data. The culture method was considered the gold standard. Statistical analysis was done using the SPSS program (v. 16.0) (Chicago, USA) and MedCalc. Among 236 sputum specimens collected from the same number of patients, 14 and 18 specimens were positive by smear microscopy for acid-fast bacilli (AFB) and by culture, respectively (Table 1). The sensitivity and specificity of the sdaA LAMP test in comparison with culture as a gold standard were 100% and 97.24%, respectively, values is comparable to those of previous studies (10).

TABLE 1.

Comparison of sdaA LAMP assay, smear microscopy, and culture results in clinical specimens

Assay resulta	Culture or smear result (no.)		No. of specimens	Sensitivity (%) (95% CI)b	Specificity (%) (95% CI)	PPVc (%) (95% CI)	NPVd (%) (95% CI)	Positive likelihood ratio (95% CI)	Negative likelihood ratio (95% CI)	κ value (95% CI)
	Culture+	Culture−
LAMP+	18	6	236	100 (81.3–100)	97.2 (94.1–99.0)	75 (53.3–90.2)	100 (98.3–100)	36.3 (16.5–80.0)	0 (NA)e	0.8 (0.7–0.9)
LAMP−	0	212
	Culture+	Culture−
Smear+	10	4	236	55.6 (30.8–78.4)	98.1 (95.4–99.5)	71.4 (41.9–91.4)	96.3 (93–98.4)	30.2 (10.5–87.0)	0.4 (0.2–0.7)	0.6 (0.3–0.8)
Smear−	8	214
	Smear+	Smear−
LAMP+	14	10	236	100 (76.7–100)	95.5 (91.9–97.8)	58.3 (36.7–77.9)	100 (98.3–100)	22.2 (12.1–40.7)	0 (NA)	0.7 (0.5–0.8)
LAMP−	0	212
	Culture+/smear+	Culture−/smear−
LAMP+	10	2	224	100 (69–100)	99 (96.7–99.9)	83.3 (51.6–97.4)	100 (98.3–100)	107 (26.9–425.1)	0 (NA)	0.9 (0.7–1)
LAMP−	0	212

^a+, positive result; −, negative result.

^bCI, confidence interval.

^cPPV, positive predictive value.

^dNPV, negative predictive value.

^eNA, not applicable.

Another set of 412 clinical samples (group II) were also evaluated with the LAMP assay and the results were compared with those for the in-house sdaA PCR. The PCR results were in high concordance with the LAMP assay results for all the samples (Table 2).

TABLE 2.

Concordance between the results of sdaA LAMP and sdaA PCR assays to analyze clinical specimens (group I and group II) for the detection of M. tuberculosis

sdaA PCR result by group	sdaA LAMP result		Concordance between results
	No. positive	No. negative	%	κ value (95% CI)a
Group I (n = 236)			99.6	0.976 (0.93–1)
No. positive	23	0
No. negative	1	212
Group II (n = 412)			99.3	0.984 (0.966–1)
No. positive	151	1
No. negative	2	258

^aCI, confidence interval.

LAMP is a versatile technique offering many convenient read-out options for results (11). In this study, easy and sensitive visual detection using SYBR green I was carried out. A major problem of carryover contamination leading to false positives has been reported in various previous studies (5–7, 12). This challenge was successfully addressed by using the dUTP-UNG method, which is widely used for PCR. Previous studies applied the dUTP-UNG method to prevent carryover contamination in LAMP assays by five-eighths replacement of dTTP with dUTP using Bst polymerase enzyme, but it adversely affected the reaction efficiency (13, 14). In this study, for the first time, contamination risk was avoided in LAMP assay for the detection of M. tuberculosis by complete replacement of dTTP with dUTP in the reaction mix without any adverse effects on amplification efficiency.

In conclusion, the sdaA LAMP assay has the potential to be developed further and applied as a point-of-care test for the diagnosis of tuberculosis in peripheral laboratories and resource-limited settings.