1. Development of RT-LAMP method for HPeVs
The 5′ UTR is 700 bp in length. When we attempted to find combinations of primers using whole 700 bp bases of the 5′ UTR by using PrimerExplorer V5 (Eiken Chemical Co., Ltd.), we could not find any appropriate combinations of primers. Therefore, while shifting the range (e.g., 1–400, 11–410, 21–340, …, 301–700 (i.e., the last 5′ UTR), we comprehensively extracted all combinations of primers (i.e., FIP, BIP, F3, and B3). The results showed eight candidate combinations (Supplemental Table 2). From these combinations, we chose three combinations (Set Aloop-, Bloop-, and Cloop-) because they were most frequently duplicated (Figure 1-a and Table). Sets Aloop-, Bloop-, and Cloop- were extracted 12 times in our experiments, but the other five combinations of the primers were extracted only once or twice.
When we used HPeV3 viral RNAs for RT-LAMP, Sets Aloop-, Bloop-, and Cloop- showed positive results. However, we tested multiple times and these sets had unstable or unreproducible rates (positive rate, %; [i.e., number of positive samples/number of tested samples]), as follows: Set Aloop-, 33.3% (6/18); Set Bloop-, 44.4% (8/18), and Set Cloop-, 86.0% (6/7).
Therefore, we added two more primers, loop-F and loop-R, to increase the sensitivity of the RT-LAMP method. We automatically designed loop-R by using PrimerExplorer V5 (Eiken Chemical Co., Ltd.), but we manually designed loop-F because PrimerExplorer V5 could not answer the loop-F candidate (Figure 1-a and Table). As a result, we could detect HPeV3 by using primer Set A (Set Aloop- + loop-F/R), Set B (Set Bloop- + loop-F/R), and Set C (Set Cloop- + loop-F/R) (Figure 1-b). The RT-LAMP products that included the target sequence and the reverse complementary sequence were confirmed by direct sequencing (data not shown).
2. Minimum reaction time and amount of viral RNAs
We tested RT-LAMP by changing the reaction times to 20 minutes, 30 minutes, 40 minutes, …, 110 minutes. We also tested RT-LAMP by applying 1.18×102 ng, 1.18×10 ng, 1.18 ng, …., 1.18×10-5 ng of total viral RNAs. The results are shown in Supplemental Table 3. When we applied more than 1.18×10-1 ng of the total viral RNAs, positive results were visible at 20 minutes in Set A and Set B. Set C showed optically positive results at 30 minutes. However, when we applied less than 1.18×10-1 ng of the total viral RNAs, the reaction became unstable. Only Set C showed positive results at 50 minutes for all amounts of viral RNAs. Therefore, we confirmed that our RT-LAMP technique could detect at least 118 pg of the total viral RNAs in stool samples and could show positive results in 30 minutes.
We also changed the reaction temperature, using 59°C, 61°C, 63°C, 65°C, and 67°C. The most appropriate and stable temperature was 63°C (data not shown).
3. Specificity of the RT-LAMP method
We conducted RT-LAMP using samples that were already known the etiology (two more HPeV3, one HPeV1, one Coxsackievirus B5, one Echovirus 11, one Enterovirus D68, one Enterovirus A71, three Noro, two Adeno. HPeV3 and HPeV1 showed positive results in our RT-LAMP assay. However, all other samples were negative (Figure 2 and Supplemental Table 4).
4. RT-LAMP analysis using clinical samples
The stools of six anonymous febrile infants were obtained at the time of suspected HPeV infection. From these stools, we extracted viral RNAs. The concentration of viral RNA solutions was 84.8±25.2 μg/mL (expressed as the mean±standard deviation).
We applied 2 μL of viral RNA solution, which was 169.7±50.4 ng of viral RNAs, for the RT-LAMP experiments. The LAMP products were visible in the stool samples of four patients (Patients #3–#6) and the positive control (Figure 3). However, no RT-LAMP products were visible in the negative control and in Patients #1 and #2. RT-LAMP experiments were conducted using both Sets A, B, and C, and all showed the same results (Supplemental Figure).
In the six stools, we simultaneously tried to detect HPeVs genomes by using the nested PCR technique. We detected HPeV3 in four samples and did not detect HPeVs in two samples. This finding corresponded to the RT-LAMP results. An important finding was that we could know the results in 2 hours after collecting stool samples when using the RT-LAMP technique, but in 2 days when using the nested PCR technique. In addition, all RT-LAMP products were confirmed by direct sequencing (data not shown).