Mechanism of LAMP reaction and hybridization of probe
The primers used for LAMP reaction is schematically depicted in enclosure of Figure 1. Forward Inner Primer (FIP) consists of F2 and the complementary sequence of F1, and Backward Inner Primer (BIP) contains of B2 and the complementary sequence of B1 when each sequences (F1, F2, B1, and B2) are defined on the template sequence as shown in Figure 1. In some references such as Notomi et al. [1] or Parida et al. [5], a spacer of few thymidines was inserted between F1c or B1c and F2 or B2 in the inner primer (FIP or BIP) so that one and two thymidine spacers were inserted in FIP and BIP of HBV, respectively. However, the spacer was not used in this study because the LAMP reaction can progress with the use of inner primers without the spacer as shown by Hong et al. [4], Hirayama et al. [6], and Iwasaki et al. [7]. The LAMP reaction takes place isothermally in the three steps shown in Figure 1, i.e., starting material production step, cycling amplification step, and elongation and recycling step, by using of polymerase with strand displacement activity. First, the starting structure (structure 6) is generated from the template nucleic acid in the starting material production step. Next, the starting structure becomes structure 7 with a stem-loop structure by self-primed DNA synthesis. When the forward inner primer is hybridized to the loop segment and strand displacement synthesis takes place, structure 11, which is a complementary strand of structure 6, is generated. This means that an auto cycle reaction was established between structure 6 and structure 11. In addition, products bound by an inverted repeat with two amplified regions, like structures 8 and 13, are generated in association with the auto cycle reaction (cycling amplification step). Then, with these structures as starting points, products elongated to a length of several kbp and products with complex structures with cauliflower-like structures (14, 15) are ultimately generated.
Since the LAMP products have a loop structure, oligo DNA probes (green arcs in the figure) in the reaction solution can sequentially hybridize to products as the LAMP reaction proceeds. The cauliflower structures, in particular, contain two or more loops to which the probes can hybridize. This characteristic of the LAMP reaction plays an important role in the detection method described here.
Sequence-specific visualization of LAMP amplicons and it's estimated mechanism
Plasmid DNA cloned with HBV or HCV sequence was added to a LAMP reaction solution containing both HBV primers and the probe (FITC-labeled) and HCV primers and the probe (ROX-labeled) and amplified, after which low molecular weight PEI (Mw = 600; 0.2 μmol as a monomer) was added. As shown in Figure 2A, precipitate emitting green fluorescence characteristic of FITC was obtained in LAMP reaction solution containing the HBV template, precipitate emitting red fluorescence characteristic of ROX was obtained in LAMP reaction solution containing the HCV template, and precipitate emitting a color (orange) that was a combination of FITC green and ROX red fluorescence was obtained in LAMP reaction solution containing both templates. They could be observed using an ordinary UV illuminator or UV-LED (light emitted diode). When LAMP products (lambda DNA) not related to HBV and HCV were present, precipitate with no fluorescence was obtained, and visible precipitate was not generated in a sample not containing a template (LAMP reaction negative). This means that it was possible to assess whether the HBV template nucleic acid was in the reaction solution, HCV template nucleic acid was in the reaction solution, or both were in the reaction solution by visualizing the fluorescent color of the precipitate. The LAMP method is a nucleic acid amplification method that is so sensitive that it is possible to create an amplification reaction from only six copies of template nucleic acid [1]. Therefore, a combination of LAMP amplification and this detection method makes possible sequence-specific visual presentation of the presence of trace amounts of nucleic acid, i.e., only several copies, found in a sample.
A model that represents the principle of this method is shown in Figure 2B. Oligo DNA probes labeled with fluorescent dye are hybridized to a specific LAMP product. When an optimized amount of low-molecular-weight PEI (Mw = 600) is then added, the positive charge of PEI neutralizes the negative charge of the DNA, which results in formation of an insoluble DNA-PEI complex. When this solution is left to stand for a few minutes or it is centrifuged with a small, desk-top centrifuge for a few seconds, the complex is deposited at the bottom of the tube. When the precipitate is observed on an illuminator (365 nm), the fluorescence of the dye from the probe taken up by the precipitate is visualized. On the other hand, probes unrelated to the sequence of the amplified LAMP product are not taken up by the LAMP-PEI complex because they are not hybridized to the LAMP product. When the molecular weight of the PEI used is small, oligo DNA probes and PEI cannot interact sufficiently to form an insoluble complex. Therefore, unrelated probes remain in the supernatant. Since the fluorescent probes in the supernatant are dispersed, the fluorescence cannot be visualized. Consequentially, bound/free separation of labeled oligo DNA is achieved as a result of insolubilization by PEI of LAMP products.
Confirmation of the mechanism of the sequence-specific visualization method
The following experiments were conducted to confirm this principle. First, we investigated the effect of the added amount of PEI on this precipitation titration (Figure 3A). Oligo DNA probes for lambda DNA were captured in the precipitate when 0.2 to 1 μmol of PEI was added to 25 μL of LAMP reaction solution for lambda DNA. If the amount of PEI added is too high or too low relative to the optimal range, the PEI-DNA complex precipitate is not formed, resulting in oligo DNA probes remaining in the solution. This phenomenon is characteristic in ionic interaction between cationic polymers and anionic polymers [14]. Namely, when the amount of the cationic polymer PEI is too low, the PEI-DNA complex becomes anionic. In contrast, when the amount of PEI is too high, the PEI-DNA complex becomes cationic. In both cases, the PEI-DNA complex is solubilized as a result. This characteristic, which is shown in Figure 3A, indicates that this precipitation titration is based on neutralization of the negative electric charge of the DNA by the cationic polymer PEI. In the case of LAMP negative, a small amount of free probe was deposited under certain conditions when 0.4 μmol to 0.8 μmol of PEI was added as a monomer (< 20%). However, the amount of precipitate in this case was so small that it was impossible to confirm it visually. On the other hand, when LAMP products (PSA) unrelated to the probe sequence were present, no free probe at all was deposited. This is because almost all of the PEI molecules added were consumed in the precipitation of unrelated LAMP products, an excess of which was present relative to the amount of probe. In other words, more reliable detection is achieved by first confirming whether the LAMP reaction has occurred by checking whether white DNA-PEI complex precipitate is generated as a result of addition of PEI and then determining for which nucleic acid template the LAMP reaction occurred based on the fluorescent color of the precipitate. Because of the above results, the amount of PEI added was fixed at 0.2 μmol for the following experiments in order to avoid generation of free probe precipitate as much as possible.
We investigated the effect of the Mw of PEI on this detection system (Figure 3B). When PEI with different Mw was added to the LAMP solution so that the amount per monomer of each was the same, we found that sufficient BF separation occurred if LAMP amplicons were present, even if the Mw of PEI increased up to 10,000. That is, the fluorescence intensity of the supernatant of LAMP solution with specific amplification (lambda DNA) is lower than that with unrelated amplification (PSA). As the Mw of PEI increased, however, almost all probes in LAMP reaction negative solution formed an insoluble PEI-oligo DNA complex. This result means that the LAMP reaction negative solution cannot be distinguished from the LAMP reaction positive solution, which successfully amplifies the targeting sequence if PEI with a high molecular weight is used. As was also observed in Figure 3A, when LAMP amplicons are present, almost all of the added PEI reacts with LAMP amplicons, so there is little opportunity for interaction with free oligo DNA probes, but when LAMP amplicons are not present, all of the added PEI interacts with oligo DNA probes. Under conditions where a large amount of PEI with a high molecular weight can strongly interact with oligo DNA, an undesirable insoluble PEI-oligo DNA probe complex forms because the molecular weight of PEI is high. Therefore, the average molecular weight of PEI in this detection system should be about 600. We conducted a similar experiment using spermine, which is a polyamine with a lower molecular weight. In that experiment, an adequate amount of insoluble complex was not generated compared with PEI of an average molecular weight of 600 under conditions used for the present research (data not shown). It is well known that spermine can also make DNA insoluble, as indicated in many other reports [10]. Therefore, optimal conditions in the case of spermine as a precipitant might exist, but since further investigation would have gone beyond the scope of this paper, no further investigation was carried out.
Insolubilization of LAMP products by PEI was inhibited by addition of an excessive amount of KCl to the LAMP reaction solution after amplification (Figure 3C). This was because excessive amounts of potassium ions and chloride ions inhibited the electrostatic interaction between DNA and PEI. This finding is further indication that this detection method is based on neutralization of the negative electric charge of DNA by the positive electric charge of PEI. The effect of ionic strength on the LAMP reaction has been investigated and found that the presence of 200 mM or more of KCl markedly delayed the LAMP reaction (data not shown). Therefore, it can be said that this detection method will function without trouble if the solution used has a composition that is optimized for the LAMP reaction.
We investigated the sensitivity of this detection method (Figure 3D). Almost all probes were taken up by the precipitate in the case of up to 1 μg of LAMP product. Moreover, we were able to visualize the fluorescence in the precipitate even if the specific LAMP product was 0.2 μg. We found that this detection system was sensitive enough as a detection system for visual assessment to be used in simple g-POCT devices. We can see from the results shown in Figure 3C that all of the 1 pmol of probe added hybridized to 1 μg (= 3 nmol nucleotide) of LAMP product. In other words, one molecule of probe bound to every 1,500 base pairs of LAMP products. The LAMP product is a mixture of products of several different sizes, with an average molecular size of 2 kbp [1]. This finding that one molecule of probe binds to every 1,500 base pairs reflects well the fundamental characteristic of this LAMP reaction.