Ultrasensitive loop mediated isothermal amplification (US-LAMP) to detect malaria for elimination

For Plasmodium genus-level detection, a set of genus-specific (Pan) primers was adopted from a previous study [13]. Modified P. falciparum-specific primers (Pf) were developed for this study (Table 1) [29]. Briefly, P. falciparum-specific primers were modified to amplify a specific region of the 18S rRNA gene located at chromosome 5 and 7, since these two copies are highly expressed at the blood stages of their life cycle [23, 30].

For whole blood specimens, a modification of the traditional Trizol Reagent (Invitrogen, Burlington, ON) based RNA extraction protocol was used. Briefly, 250 µL of whole blood was mixed with 50 µL of 5% saponin (Sigma-Aldrich, Oakville, ON) solution and kept at room temperature for 15 min. In this step, blood was mixed with saponin by shaking the tubes by hand, no vigorous mixing such as vortex mixing was applied. Then, saponin lysate was mixed and homogenized with 1500 µL of Trizol reagent (pH was adjusted to 7.2) and kept at room temperature for 10 min. Afterward, blood lysate was centrifuged at 4 °C and 10,000 rpm for 5 min and the supernatant decanted for downstream purification. The subsequent steps were performed according to the manufacturer’s instructions except for the final washing step (no 70% ethanol required) [31]. Two µL of the extract was used for each US-LAMP reaction. A detailed description of the protocol can be found in Additional file 3.

A modified total nucleic acid extraction protocol from DBS was used [24]. Fifty microliter of the whole blood sample was spotted onto Whatman 903 protein saver cards (GE Healthcare, Mississauga, ON) and allowed to air-dry overnight. A standard 6 mm-diameter hole puncher was used to cut spots into individual tubes. Dried blood containing filter paper pieces were mixed with lysis buffer [24] and incubated at 65 °C and 250 rpm shaking speed for 2.5 h in an orbital shaker. The 700 µL supernatant was transferred into HiBind^® DNA mini columns (Omega Bio-tek, Norcross, GA) and centrifuged at 2000 rpm for 2 min followed by another spinning step of 8000 rpm for 1 min. Column-bound nucleic acid was washed with 500 µL of the “wash buffer 1” by centrifuging the column at 8000 rpm for 1 min. Columns were washed with 500 µL of the “wash buffer 2” through centrifuging at 13,000 rpm for 3 min. An eluate of total nucleic acid with 50 µL of TE buffer (5 min wait after the addition of TE buffer) was obtained by centrifuging at 8000 rpm for 1 min. Ten microliters of the filter paper extract was used in a single US-LAMP reaction. A detailed description of the protocol can be found in Additional file 4.

Bst 2.0 WarmStart^® DNA polymerase was combined with WarmStart^® reverse transcriptase in 1X Isothermal Amplification Buffer (New England Biolabs, Whitby, ON) to perform the US-LAMP assay. In a 25-µL LAMP reaction mixture, 1.6 µM F1P and B1P, 0.8 µM LPF and LPB, 0.2 µM F3 and B3 primer concentrations, 8 mM MgSO₄, 1.4 mM dNTPs, 0.8 M Betaine (Sigma-Aldrich, Oakville, ON, Canada), 8 unit of Bst 2.0 WarmStart^® DNA Polymerase and 7.5 unit of Warm Start^® reverse transcriptase were used. The assay was optimized with pre-addition of 0.5 µL of 50X SYBR green (Invitrogen, Burlington, ON) in the reaction mixture. Amplification was measured based on increased relative fluorescence units (RFU) per minute in the CFX-96 Real-Time PCR detection system (Bio-Rad, Mississauga, ON). A threshold RFU value of 200 was chosen based on the background fluorescence levels. Optimization studies were performed to arrive at the ideal incubation of 63 °C for both Pan and Pf-LAMP assays. The final assay duration was fixed at 30 min for the Pan-LAMP assay and 60 min for the Pf-LAMP assay after optimizing the amplification curves using the CFX96 Real Time System. For visual detection of LAMP amplification, 1 µL of 0.35% (v/v) Gel green (Biotium, Freemont, CA), and 3 mM Hydroxynapthol blue (Sigma-aldrich, Oakville, ON), was added to the master mix [32]. Nuclease free water (VWR, Mississauga, ON) was used to constitute the final reaction volume. After amplification, the reaction tubes were exposed to blue LED light using a Blue Light Transilluminator (New England Biogroup, Atkinson, NH) to visualize the fluorescence due to the inter-chelation of gel green with the amplicons.

Uninfected blood, collected from a healthy donor, was spiked with in vitro culture of P. falciparum strain 3D7. A serial dilution of spiked blood with uninfected blood was made resulting in a parasite count range of 1 to 10,000 parasite/mL. Here, the parasite count was established by microscopy. Total nucleic acid was extracted from whole blood as well as from the DBS (50 µL) prepared from each dilution. Subsequently, the nucleic acid was amplified from the extracts by Plasmodium genus (Pan) and P. falciparum (Pf)-specific primers sets in the CFX-96 Real Time system. To detect real-time amplification, fluorescence measurement was taken every minute.

After initial assessment on culture-spiked blood, whole blood from one of each P. falciparum, Plasmodium vivax and Plasmodium ovale spp.-infected patients was diluted with healthy donor blood to obtain a parasite count ranging from 1 to 10,000 parasite/mL. Here, P. falciparum dilutions were tested in triplicate by both Pan and Pf-specific primers set while P. vivax and P. ovale spp. dilutions were tested in triplicate with only genus-specific primers. LODs were determined using the CFX-96 Real Time System subsequently confirmed through observing gel green fluorescence. If at least two out of three replicates of a certain dilution were tested positive by LAMP assays, the result was noted as positive.

A combination of whole blood specimens (symptomatic returning travellers to Calgary, Canada) and DBS (asymptomatic individuals from Ethiopia and Bangladesh) were used for validation of the assay. For symptomatic travellers, 41 malaria positive specimens and 72 negative specimens based on microscopy were obtained between September 2017 to May 2018 were used. Of the 41 positives, 24 were positive for P. falciparum, 12 were P. vivax positive and five were positive for P. ovale spp. Total nucleic acid was extracted from the fresh blood samples within 24 h of collection by the modified Trizol extraction mentioned above. Additionally, to validate the utility of US-LAMP for detection of low-level infections, DBS samples collected previously from asymptomatic individuals were obtained from a relatively high transmission area (Gondar, Ethiopia, n = 308) and a low transmission area (Bandarban, Bangladesh, n = 186). Whole blood samples were collected in EDTA tubes through venipuncture, and subsequently, 50 µL of the EDTA mixed blood specimens were spotted onto Whatman 903 protein saver card immediately after collection. Then, the blood spots were air-dried and stored at the room temperature in Calgary which is around 20–25 °C. Additionally, thick and thin blood smears were prepared for microscopy prior to the mixing with EDTA. Total nucleic acid was extracted from the DBS by the aforementioned protocol, and LAMP assays were conducted on the extracts with the pre-addition of gel green and hydroxynaphthol blue in the reaction mixture. Ethical approval were obtained from the corresponding Ethical Review board of University of Gondar (CMHS08/28/2013), International Center for Diarrheal Disease Research, Bangladesh (icddr,b:PR-15021), and University of Calgary Conjoint Health Research Ethics Board (REB17-2220). For sensitivity and specificity analysis, qRT-PCR [24] targeting 18S rRNA was used as the gold standard.

A standard curve was obtained by plotting time to amplification against the logarithms of initial parasite count per mL. The goodness of fit to the straight line (R²) values were 0.813 and 0.641 respectively from whole blood and DBS. The data confirm that the Plasmodium genus-level US-LAMP assay consistently detects the presence of P. falciparum as low as 10 parasite/mL of culture-spiked whole blood, whereas the detection limit was 25 parasite/mL from DBS (Additional file 1: Figure S1).

Standard curves were plotted using threshold time against the logarithms of initial parasite count per mL for P. falciparum using genus- and species-specific primers to determine the assay dynamic range (Figs. 1 and 2). The LOD of 50 parasite/mL and 50–100 parasite/mL from whole blood and DBS in case of the P. falciparum-specific LAMP assay was achieved, with a goodness of fit to the straight line of 0.884 and 0.927, respectively (Fig. 3). Standard curves were plotted separately for whole blood and DBS using time to amplification (threshold time) against the logarithms of initial parasite count per mL for P. falciparum, P. vivax, and P. ovale spp. serial dilutions prepared from patient specimens. In the case of P. falciparum, LODs of 25 parasite/mL and 50–100 parasite/mL from whole blood and DBS were obtained, respectively (Fig. 1). Corresponding goodness of fit to the straight line values were 0.931 (whole blood) and 0.800 (DBS). LODs of 5–10 parasites/mL and 25–50 parasite/mL for P. vivax were obtained using whole blood and DBS (Fig. 2). LODs for P. ovale spp. were 25 parasites/mL from whole blood and 25–50 parasite/mL from DBS. The goodness of fit to the straight line was 0.896 and 0.827 for corresponding whole blood and DBS extracts of P. ovale. Through pre-addition of gel green, visual detection of amplification was also possible (Fig. 4). Additionally, data demonstrate that a maximum of four spots can be batched together without compromising the limit of detection (Additional file 1: Figure S2).

From P. falciparum strain 3D7 culture-spiked blood, DNA and RNA are stable for at least 5 months on Whatman 903 protein saver card. However, best results were obtained within 1 month. After 1 month, inconsistent amplification from lower dilutions at 25 and 50 parasites/mL occurred (Table 2).

An initial verification study was performed on a set of clinical specimens from symptomatic returning travellers. Pan-LAMP and Pf-LAMP assays were 100% (95% CI 82.8–100) sensitive for the detection of symptomatic malaria, whereas specificity was 98.6% (95% CI 91.5–99.9) and 97.8% (95% CI 91.4–99.6) for pan-LAMP and Pf-LAMP assay, respectively (Table 3). Subsequently, DBS samples from asymptomatic individuals in Gondar (Ethiopia) and Bandarban (Bangladesh) were used to validate the assay on low-level infections. Overall, Pan-LAMP was 97% sensitive (95% CI 82.5–99.8) and 99.1% (95% CI 97.6–99.7) specific for identifying asymptomatic Plasmodium infection while Pf-LAMP was concluded to be 100% sensitive and 99.8% specific for asymptomatic P. falciparum cases (Table 3). QRT-PCR and Pan-LAMP detected 29 and 32 positives, respectively, from the 308 microscopy negative DBS samples obtained from Gondar. The additional infections detected by US-LAMP but not microscopy comprised 10 P. falciparum, 16 P. vivax, and 3 P. falciparum and P. vivax mixed infections. US-LAMP detected one additional P. falciparum asymptomatic carrier not identified by microscopy from the Bandarban region (Additional file 2: Table S1). Detailed sensitivity and specificity calculations can be found in the Additional file 2: Tables S2–S5.