To identify the availability of stem-loop RT-PCR technology for cloning conserved miRNAs from non-model insects, the sequences of miR-14, miR-2a and bantam in model insects were searched from miRBase (Table S1). Two groups of primers for amplification were designed based on the sequences of mature miRNAs in D. melanogaster (shown in Experimental Section).

The putative homologue of the three miRNAs in S. litura, namely sli-miR-14, sli-miR-2a and sli-bantam were cloned by a stem-loop RT-PCR technique. Ten positive clones of each sample were sequenced. When Group 1 was used for amplification of miR-14, miR-2a and bantam, PCR products were 86 bp, 89 bp, 89 bp, respectively (Figure 1a); when Group 2 was used, PCR products were 76 bp, 79 bp and 79 bp in length, respectively (Figure 1b). Sequence analysis of the single bands showed that specific amplifications consisted of all bases of putative miRNA, PCR forward primer and reverse primer. Sequences of these cloned miRNAs in S. litura showed 100% match to those in D. melanogaster.

To verify the present results, fragments appeared in blank control group were also purified from 2.5% agarose gels, ligated into the T-vector and then sequenced. These non-specific fragments amplified with Group 1 primers were diverse according to agarose gel electrophoresis analysis, sequence analysis showed that all fragments just consisted of PCR forward primer and reverse primer, rather than the target miRNAs (Figure 1a). By using primers of Group 2, the fragments in blank control group of each miRNA displayed a single band by agarose gel electrophoresis (Figure 1b). Similarly, sequence analysis showed these fragments in each experimental group just included the forward primer, reverse primer, and/or other non-specific bases.

Performances of the two groups of primers were also different in the experimental group. Some non-specific amplifications appeared in experimental group with primers of Group 1, but the majority of sequenced colonies contained all bases of putative miRNA. By using primers of Group 2, agarose gel electrophoresis showed amplified fragment was a single band when cDNA was added into the reaction volume, sequence analysis also confirmed it. All indicated that primers bound to cDNA much more tightly than themselves in experimental group. Although two groups of primers could be used for cloning conserved miRNAs from S. litura, Group 2 performed better and was more stable; same results were shown in electrophoretogram (Figure 1). Based on the electrophoresis and sequence analysis of RT-PCR, Group 2 primers were more suitable for a quantitative PCR (qPCR) assay.

To verify the present results, homology of pre-miR-14 and pre-miR-2a were cloned from S. litura. According to the two ESTs of Spodoptera frugiperda which are highly homologous to putative sli-miR-14 and four ESTs of Heliothis virescens, a 322 bp fragment was amplified using degenerate primers Pre-14F and Pre-14R from genomic DNA of adult of S.litura. This fragment showed 97.5% identity to the ESTs of S. frugiperda and 82 bp putative pre-sli-miR-14 was contained in it (Figure 2a). Blast procedure had been performed to analyze the homology among the putative pre-miR-14 in S.litura and that in B. mori and D. melanogaster. Results showed the putative pre-sli-miR-14 shared 83.8% and 73% homology to pre-bmo-miR-14 and pre-dme-miR-14 respectively (Figure 2b).

A cluster of miR-2a-1, miR-2a-2, miR-2b, miR-13a and miR-13b is localized on chromosome 1 of B. mori, indicating there are two precursors of miR-2a, and miR-2a-2 is in the interior of this cluster. Therefore, Primers Pre-2aF and Pre-2aR were designed according to the conserved miRNAs for amplifying pre-miR-2a-2, and rising nealing temperature to 65 °C to inhibit non-specific amplification. A 552bp fragment containing putative pre-sli-miR-2a-2 was amplified from genomic DNA of adult of S. litura. Sequence analysis showed that this fragment was 68.9% identity to the homologue in B. mori, there are three highly conserved regions besides two primers’ sites, where highly conserved miRNAs including miR-2a-2, miR-13a and miR-13b exist (Figure 3a). The putative pre-sli-miR-2a showed 81.1% and 70.4% similarities with pre-bmo-miR-2a and pre-dme-miR-2a respectively (Figure 3b).

Moreover, both mature sequences of miR-14 and miR-2a in their precursors shared 100% identity to the results shown by stem-loop RT-PCR. In order to verify the result, secondary structure of two putative pre-miRNAs were predicted by mfold program. Results showed that putative pre-miR-14 and pre-miR-2a of S. litura can form stable stem-loop structures (initial ΔG = −41.80 and ΔG = −41.20 respectively) and are highly homologous to those of B. mori and D. melanogaster, indicating they are the precursors of sli-miR-14 and sli-miR-2a (Figure 4). Above all suggested stem-loop RT-PCR could be used in identification of highly conserved miRNAs in non-model insects.

The expression patterns of the putative miRNAs, sli-miR-14, sli-miR-2a and sli-bantam were also studied to make a further validation of this method. Group 2 primers were used for a qPCR assay. Although putative sli-bantam could be cloned by stem-loop RT-PCR, no stable results of qPCR could be obtained here (data not shown). However, the expression patterns of sli-miR-14 and sli-miR-2a, two highly conserved miRNAs in insect, could be obtained in this assay. PCR efficiency was determined on the miRNAs as well as 5S rRNA. The efficiency of miR-14, miR-2a and 5S rRNA was close to the ideal value of 2 (Table 1), therefore relative quantification (RQ) of miRNA expression was calculated with 2^{−Δ ΔCt} method [28,29].

Sli-miR-14 was expressed strongly in all developmental stages of S. litura, but there were some differences among developmental stages. The relative expression levels of sli-miR-14 were 1.47-, 4.22-, 3.78-, 2.26-, 1.87-, 0.37- and 9.07-fold higher in the first instar, third instar, fourth instar, six instar, pre-pupa, pupa and adult than in eggs, respectively (Figure 5a), which indicated that sli-miR-14 expressed in pupal stage lower than in other stages, and especially strongly in adult stage.

Development-specific expression patterns for sli-miR-2a were also detected by qPCR. There were significant differences among various developmental stages, the relative expression levels of sli-miR-2a were 2.89-, 15.65-, 24.55-, 4.88-, 2.23-, 38.68- and 1.81-fold higher in the first instar, third instar, fourth instar, six instar, prepupa, pupa and adult than in egg, respectively (Figure 5b). Sli-miR-2a expressed at low level in egg, prepupa and adult stages, instead expressed most strongly in pupa stage, which accord with that in model insects, D. melanogaster [30] and B. mori [24].

As we know, stem-loop RT-PCR is a very sensitive method for quantitative determination of known miRNAs, even one base difference between two miRNAs in the same family could be detected and distinguished by this method [31]. Whether stem-loop RT-PCR could be used for identifying conserved miRNAs in non-model insects was studied in this paper. Two groups of primers were designed to access the availability of this method. Six nucleotides of 3′ end of stem-loop RT primer were complementary with 3′ end of mature miRNAs, other nucleotides of RT primer were partially complementary with themselves to form a stem, and the remaining nucleotides form a loop. The PCR reverse primer was universal, while the forward primer was designed by mature miRNAs (Figure 6a). In order to make a distinction between specific and non-specific amplifications, only partial bases of miRNA were contained in forward primer. Two bases, three bases and four bases were not included in primers of miR-14, miR-2a and bantam, respectively (Figure 6b). The designing of stem-loop RT primers often presents learners with difficulties in some way; the primers designed in this assay also provide references for later research on other miRNAs.

MiR-14 and miR-2a are more conserved members of miRNAs family than bantam in model insects including D. melanogaster and B. mori (Table S1). Sequence analysis indicated that homologue of the three miRNAs could be cloned from S. litura by stem-loop RT-PCR. In order to confirm whether the PCR products are endogenous miRNAs, putative precursors of miR-14 and miR-2a as representatives were also cloned from S. litura, results showed that both their sequences and secondary structures were highly conserved with model insects. The mature sli-miR-14 and sli-miR-2a in their precursors shared 100% identity to the results shown by stem-loop RT-PCR. Then, sli-miR-14 and sli-miR-2a were successfully detected by a qPCR assay and their expression patterns showed similar characteristics with those in model insects. Instead, abnormal amplification appeared in qPCR of sli-bantam, there is no significant difference in Ct values in the presence or absence of cDNA added to the RT reactions (data not shown). As mentioned above, only six bases of 3′ end of putative miRNA were complementary with stem-loop RT primer, but PCR forward primer was restricted by the 5′ end of miRNAs, suggesting this strategy can be used for identification of highly conserved miRNAs and distinguished variation appeared in 3′ region of miRNAs. However, this method is difficult to differentiate variation appeared in 5′ region of miRNAs. Some miRNAs with evolutional variation in their 5′ region, such as bantam, cannot show accurate sequences in non-model insects and toned to be identified together by other methods.

Although the identification of miRNAs in non-model insects with stem-loop RT-PCR is limited by 5′ region of putative miRNA, large amounts of miRNAs that are highly conserved, such as miR-14 and miR-2a, can be simply cloned by this method. It has been clearly stated that cell autonomous anti-apoptotic activity mediated by miR-14 and miR-2a [12,27], miR-14 also plays critical role in molting process [25], indicating miRNAs are involved in strict developmental regulation in insects. Studies on miRNAs in various insects, especially agricultural pests, may provide new ideas for pest control. However, so many difficulties stand in our way for identifying miRNAs in non-model insects, a simple and efficient method is urgently necessary.

Applying stem-loop RT-PCR in identification of highly conserved miRNAs from S. litura provides a new strategy for studying miRNAs in a more extensive field. As mentioned above, direct cloning is more appropriate for the cloning of massive miRNAs and the discovery of novel miRNAs, this powerful but blind task produces a huge workload. With the help of computer programs, researchers can make an easy prediction on miRNAs in the genomic database of model insects, and then make a direct cloning according to the predicted results. Unfortunately, this has no advantages for non-model insects. Currently, most of the studies on miRNAs of non-model insects have focused on the direct cloning of miRNAs, however, the cloned miRNAs mostly are conserved; many even have identical sequences with those in model insects. Compared with direct cloning, stem-loop RT-PCR has a clearly defined aim for cloning selected conserved miRNAs, which simplifies operation and saves cost. Meanwhile, the discovery of massive miRNAs in model insects gives us lots of resources for further study in more extensive field. When one or a few important and highly conserved miRNAs were expected to research in non-model insects, this strategy greatly improves efficiency.

A laboratory strain of S. litura was purchased from the Entomology Institute of SUN YAT-SEN University, Guangzhou, China. Insects were routinely reared at 25 ± 1 °C and 70 ± 10% relative humidity under a 16:8 h (light:dark).

S. litura ovary cell lines SL-1 were routinely maintained at 27 °C at the laboratory. SL-1 cells were cultured in insects’ Grace medium, supplemented with 7% fetal bovine serum (FBS) (Gibco, USA), 0.3% yeast extract (Oxoid, UK) and 0.3% lactalbumin hydrolysate (Oxoid, UK). Culture medium was changed every 3 days, and subcultured SL-1 cells every 6 days.

Eight samples including eggs, larvae (selected first, third, forth, sixth instar as representatives), pre-pupae, pupae and adults were collected and stored at −80 °C for total RNA extraction. The third-instar larvae and SL-1 cells were prepared for stem-loop RT-PCR, and samples of eight stages were prepared for real-time qPCR.

Two groups of primers including stem-loop primer and corresponding PCR primers were designed based on the mature sequences of the three miRNAs in D. melanogaster (Table S1). The Group 1 primers were designed by previous reports [15], Group 2 primers haven’t reported before. The partial bases of stem-loop RT primer of three miRNAs were universal, Group 1 was: 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGAC-3′ (Table 2), Group 2 was: 5′-GCACTTCAGTGTCGTGGTCAGTGACGGCAATTTGAAGTGC-3′ (Table 3). Six bases of 3′ end of stem-loop RT primer were complementary with 3′ end of mature miRNA, and the PCR forward primer was paired with 5′ end of mature miRNA (principle in Figure 3).

The extractions of total RNA from all stages of S. litura and SL-1 cells were performed by using TRIzol kit (Invitrogen) according to the manufacturer’s instructions. cDNA was synthesized from total RNA with miRNA-specific stem-loop RT primer. Reverse transcriptase reactions contained 2 μg of RNA samples, 125 nM stem-loop RT primer, 1× RT buffer, 2.5 mM each of dNTPs (TaKaRa), 200 U·μL⁻¹ PrimerScript Reverse Transcriptase (TaKaRa) and 40 U·μL⁻¹ RNase Inhibitor (TaKaRa). The 11 μL reactions consisting of total RNA, stem-loop RT primer and ddH₂O was first incubated for 10 min at 70 °C, 2 min for ice-bath, and other reagents were mixed together after a transient centrifugation. These 25 μL reactions were incubated in a MyCycler Thermal Cycler (Bio-Rad, Hercules, CA) for 60 min at 42 °C, 15 min at 72 °C, and then held at 4 °C.

PCR was performed in a 50 μL reaction volume containing 1 μL cDNA, 10 μM of each primer, 10× PCR buffer, 2.5 mM each of dNTPs (TaKaRa) and 5 U·μL⁻¹ Taq DNA polymerase (TaKaRa). The condition for the PCR was 94 °C for 3 min, followed by 35 cycles of 94 °C for 30 s, 55~60 °C for 30 s, 72 °C for 30 s, and additional polymerization step at 72 °C for 10 min. Target fragments were purified from 2.5% agarose gels and ligated into the T-vector (TaKaRa) and then sequenced.

Two gene fragments of S. frugiperda (FP355748.1 and FP352735.1) were found in EST Database (NCBI) by homologous searching according to the sequence of pupative sli-miR-14. Partial sequence of the two ESTs showed high similarities with pre-bmo-miR-14 and pre-dme-miR-14. Then four gene fragments of Heliothis virescens (GT205268.1, GT212851.1, GT204572.1 and GT204571.1) was found according to the two ESTs of S. frugiperda. Based on these homologous sequences, degenerate primers Pre-14F and Pre-14R were designed for cloning pre-miR-14 from S. litura. Pre-14F: TTTGTC ATGTGACTCGCGCC; Pre-14R: CGCAGCCGAATCAAATAACA.

By searching bmo-miR-2a in Silkworm Genome Database, it is noticed that there are two precursors of pre-bmo-miR-2a which localized on chromosome 1 and organized as a cluster with pre-miR-2b, pre-miR-13a and pre-miR-13b (Figure 5). Based on the conserved character of miRNA, Pre-2aF and Pre-2aR for cloning pre-miR-2a-2 were designed according to the mature sequences of miR-2b and miR-2a-1 which located at both ends of the cluster. Pre-2aF: ACTCAACAAAGCTGGCTGTGA; Pre-2aR: CACAGCCAGCTTTGATGAGC.

Total genomic DNA was isolated from the adult of S. litura according to the instruction of E.Z.N.A. Insect DNA Kit (OMEGA, USA). The PCR reaction was performed with the following protocol: 94 °C for 3 min; followed by 35 cycles of 94 °C for 30 s, 55~65 °C for 30 s, 72 °C for 45 s, and additional polymerization step at 72 °C for 10 min. Amplified DNA was purified by 2% agarose gels, and ligated into T-vector (TaKaRa), then positive clones were sequenced.

The cDNAs were diluted 10 times to perform PCR for expression confirmation or qPCR for expression patterns analysis. The 25 μL reaction volume consisting of 2 μL cDNA, 12.5 μL SYBR Green (TaKaRa), 9.5 μL ddH₂O, 0.5 μL of forward primer (10 μM) and 0.5 μL of reverse primer (10 μM) for each miRNAs. The reactions were performed on BIO-RAD CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA) following the manufacturer’s recommendations. The optimized real-time PCR program was 94 °C for 20 s, followed by 40 cycles of 95 °C for 10 s, 60 °C for 20 s, 72 °C for 10 s. After the cycling protocol, melting curves were obtained by increasing the temperature from 70 °C to 95 °C (0.4 °C/s) to denature the double-stranded DNA. PCR products were detected by electrophoresis with 3% agarose gel containing ethidium and photographed under UV light. There has not been a standard control for expression normalization for miRNAs, thus 5S rRNA as an internal control was used in this assay [19]. The forward primer was: CCAACGTCCATACCAT GTTGA, the reverse primer was: GCGGTCACCCATCCAAGTA.

Relative quantification (RQ) of miRNA expression was calculated with 2^{−Δ ΔCt} method [29]. To experimentally determine PCR efficiency, diluted cDNAs (10⁰, 10⁻¹, 10⁻² and 10⁻³) were amplified by real-time PCR. Plots were made of the log of template concentration versus the CT and the PCR efficiency was calculated from the slope of the line using the equation: E = 10^−1/slope.

Sequence analysis was performed by DNASTAR 7.0. Sequence similarity for conserved domains of miRNAs was analyzed by MIRBASE, Silkworm Genome Database and BLAST programs on the National Center for Biotechnology Information (NCBI) Multiple sequence alignments of partial pri-miRNA and entire pre-miRNA among related species were performed by ClustalW2. The pre-miRNA sequence was subjected to an RNA secondary structure check using mfold program.