SIDT2 mediates gymnosis, the uptake of naked single-stranded oligonucleotides into living cells.

Single-stranded oligonucleotides (ssOligos) are efficiently taken up by living cells without the use of transfection reagents. This phenomenon called 'gymnosis' enables the sequence-specific silencing of target genes in various types of cells. Several antisense ssOligos are used for the treatment of human diseases. However, the molecular mechanism underlying the uptake of naked ssOligos into cells remains to be elucidated. Here, we show that systemic RNA interference deficient-1 (SID-1) transmembrane family 2 (SIDT2), a mammalian ortholog of the Caenorhabditis elegans double-stranded RNA channel SID-1, mediates gymnosis. We show that the uptake of naked ssOligos into cells is significantly downregulated by knockdown of SIDT2. Furthermore, knockdown of SIDT2 inhibited the effect of antisense RNA mediated by gymnosis. Overexpression of SIDT2 enhanced the uptake of naked ssOligos into cells, while a single amino acid mutation in SIDT2 abolished this effect. Our findings highlight the mechanism of extra- and intracellular RNA transport and may contribute to the further development of nucleic acid-based therapies.

extracellular RNA

microRNA

miRNA–16

miRNA–19a

mouse embryonic fibroblasts

quantitative reverse transcription PCR

small guide RNA

systemic RNA interference defective-1

SID1 transmembrane family, member 1

SID1 transmembrane family, member 2

small interference RNA

single-stranded oligonucleotides

single-stranded RNA

Introduction

Nucleic acid based drugs, which could be applied to the treatment of various diseases, include antisense oligonucleotides, aptamers, small interference RNA (siRNA), antisense oligonucleotides targeting microRNAs (miRNAs), miRNA mimics, decoy oligonucleotides, and CpG oligonucleotides. Currently, antisense oligonucleotides, including RNA and/or DNA, unassisted by transfection reagents are clinically used for the treatment of homozygous familial hypercholesterolemia and cytomegalovirus retinitis.

Recently, Stein et al. reported that single-stranded oligonucleotides (ssOligos) can be efficiently taken up by living cells without the use of transfection reagents. The uptake of ssOligos without the use of a transfection reagent (naked ssOligos) is called ‘gymnosis’ (nakedness in Greek). These authors showed that naked antisense oligonucleotides can induce sequence-specific silencing of target genes in various types of cells. Small guide RNAs (sgRNAs) are a class of artificially designed single-stranded RNAs (ssRNAs) that can induce gene silencing by the tRNA maturase tRNase ZL-dependent mechanism. Multiple papers, including ours, reported that naked 2′-O-methylated sgRNAs targeting various miRNAs and mRNAs can be taken up by different types of living cells without the use of a transfection reagent and reduce target RNA levels. Iizuka et al. reported that the uptake of naked sgRNA is not mediated by clathrin or caveolin-dependent endocytosis; however, the molecular mechanism of naked ssRNA uptake into living cells remains unclear.

The Caenorhabditis elegans systemic RNA interference deficient-1 (SID-1) protein mediates the uptake of double-stranded RNA (dsRNA) into cells. SID-1 functions as a bidirectional dsRNA transporter, but does not transport ssRNA. SID-1 transmembrane family 1 and 2 (SIDT1 and SIDT2) are mammalian orthologs of SID-1. SIDT1 localizes to the plasma membrane and mediates the bidirectional transport of dsRNA in human cells. Whether SIDT1 transports ssRNA has not been elucidated. Considering that SIDT1 is expressed in restricted types of cells such as dendritic cells and lymphocytes, it is unlikely to account for the ubiquity of gymnosis. By contrast, SIDT2 is ubiquitously expressed in many types of cells. In addition, we recently found that SIDT2 on the lysosomal membrane mediates the uptake of ssRNA into lysosomes, indicating that SIDT2 can transport ssRNA. This led us to hypothesize that SIDT2 is involved in gymnosis. In the present study, we explored potential mechanisms underlying gymnosis and investigated whether the uptake of naked ssOligos is mediated by SIDT2.

Results

Naked ssOligos can be taken up by living HeLa cells

We examined whether small ssOligos can be taken up by living HeLa cells without the use of a transfection reagent. Fully 2′-O-methylated and 5′-Alexa568-labeled 15-nt oligonucleotides containing 5′- and 3′-phosphates (Alexa568-ssOligos) were synthesized. HeLa cells were cultured in the presence or absence of 500 nM of naked Alexa568-ssOligos for 6 and 24 hours and analyzed by confocal fluorescence microscopy. The results showed that naked Alexa568-ssOligos were efficiently taken up by HeLa cells without the use of a transfection reagent (Fig. 1A). Furthermore, time-course experiments of the uptake of naked ssOligos into living HeLa cells performed by culturing cells in the presence of 500 nM of naked Alexa568-ssOligos for 0, 2, 4, 6 or 8 hours followed by confocal fluorescence microscope analysis showed that the oligonucleotides were taken up by cells in a time-dependent manner (Fig. 1B). We next investigated whether naked antisense oligonucleotide targeting microRNA–16 (miR–16) can inhibit the endogenous miR–16 expression levels in HeLa cells. Previously, we reported that naked antisense oligonucleotides targeting miR–16 can reduce miR–16 expression levels in 293 cells. Cells were incubated with or without 20 nM of naked antisense oligonucleotides targeting miR–16 or negative control oligonucleotide for 6 hours, and the endogenous miR–16 expression levels were analyzed by quantitative reverse transcription PCR (qPCR). We observed that naked antisense oligonucleotides targeting miR-16 reduced the endogenous miR–16 expression levels, while negative control oligonucleotides (poly-A) did not (Fig. 1C and D).

Figure 1.

Uptake of naked ssOligos by HeLa cells. (A) Confocal imaging of HeLa cells. HeLa cells were incubated with or without 500 nM of the naked 15-nt 5′-Alexa568-labeled oligonucleotides (Alexa568-ssOligos) for 6 and 24 hours. Cells were washed with PBS and stained with Hoechst 33258. (B) HeLa cells were incubated with 500 nM of Alexa568-ssOligos for 0, 2, 4, 6 and 8 hours, and analyzed by confocal microscopy. Arrows indicate the 5′-Alexa568-labeled oligonucleotides in cells. (C and D) HeLa cells were incubated with or without 20 nM of naked antisense oligonucleotides targeting miR-16 (C) or naked negative control oligonucleotides (D) for 6 hours, and analyzed by qPCR analysis, miR-16 levels were normalized against β-actin mRNA levels. Error bars indicate s.d. (n = 3). *, P<0.05., n.s., not significant.

SIDT2 partly localizes to the plasma membrane

In HeLa cells, SIDT1 mRNA expression was not detected, whereas expression of SIDT2 mRNA was detected (Fig. 2A). Similar results were obtained when we used 293FT (human embryonic kidney cells), HL60 (human promyelocytic leukemia cells), Neuro2a (mouse neuroblastoma cells), and mouse embryonic fibroblasts (data not shown), suggesting that SIDT1 does not account for ubiquitousness of gymnosis. SIDT2 mainly localizes to lysosomes. However, it is possible that a small portion of SIDT2 localizes to the plasma membrane. To test whether endogenous SIDT2 exists in the plasma membrane, we biotinylated cell surface proteins in HeLa cells, which express endogenous SIDT2 proteins, and performed a biotin-streptavidin pull-down assay followed by western blot analysis. β-actin, an established intracellular protein, and cathepsin D, an intralysosomal protein were not biotinylated, while N-cadherin, a plasma membrane-integrated protein was biotinylated (Fig. 2B), confirming that plasma membrane proteins were selectively biotinylated. Endogenous SIDT2 was detected in biotinylated cell surface proteins (Fig. 2B), indicating that a portion of endogenous SIDT2 localizes to the plasma membrane. Moreover, using a C-terminal EGFP-tag, we also observed that SIDT2 mainly localized to lysosomes and partly to the plasma membrane (Fig. 2C).

Figure 2.

Localization of SIDT2 to the plasma membrane. (A) mRNA levels in HeLa cells were analyzed by RT-PCR. Expression of SIDT2 mRNA was detected. In contrast, SIDT1 mRNA was not detected in Hela cells. (B) HeLa cells were incubated with (+) or without (−) biotin, and a biotin-streptavidin pull-down assay was performed to purify cell surface proteins as described in the Materials and Methods section. Proteins were analyzed by western blotting using anti-SIDT2 (Abnova), anti-β-actin (ACTB), anti-N-cadherin and anti-cathepsin D antibodies. (C) HeLa cells expressing EGFP-tagged SIDT2 were incubated with LysoTracker Red. Fluorescence images were visualized using a confocal laser-scanning microscope. Arrows indicate plasma membrane localization.

SIDT2 knockdown reduces the uptake of naked ssOligos by cells

To investigate whether the uptake of naked ssOligos is mediated by SIDT2, we assessed the effect of SIDT2 knockdown on gymnosis. HeLa cells were transfected with siRNA targeting SIDT2 or control siRNA, and incubated for 72 hours (Fig. 3A–C). Then, cells were cultured in the presence or absence of 1 μM or 500 nM of Alexa568-ssOligos for 6 hours and analyzed by confocal microscopy. When 1 μM and 500 nM of naked Alexa568-ssOligos were added to the culture media, the Alexa568 intensity was decreased to 60% and 64% in SIDT2 knockdown cells, respectively, compared with control cells (Fig. 3D and E). Similar results were obtained when we used another siRNA against SIDT2 (Fig. 3F), or with another cell line, wild-type (WT) mouse embryonic fibroblasts (MEFs) (Fig. S1). These results indicated that SIDT2 mediates uptake of ssOligos by cells.

Figure 3.

Knockdown of SIDT2 reduces the uptake of naked ssOligos. (A) Experimental paradigm for the confocal microscopic analysis. (B and C) HeLa cells were transfected with 10 nM of control siRNA (siControl) or siRNA against SIDT2 (siSIDT2#1 or #2). At 72 hours after transfection, SIDT2 mRNA levels (B) and SIDT2 protein levels (C) were analyzed. In the qRT-PCR analysis, SIDT2 mRNA levels were normalized against β-actin mRNA levels (B). In the western blot analysis, anti-SIDT2 (custom antibody) and anti-β-actin antibodies were used (C). Error bars indicate s.d. (n = 3). *, P < 0.05. (D–F) HeLa cells were transfected with siControl or siSIDT2#1 (D and E) and siControl or siSIDT2#2 (F). Seventy-two hours after transfection, cells were washed with PBS, and incubated with or without 1000 nM (D) or 500 nM (E and F) of the naked Alexa568-ssOligos for 6 hours. Cells were washed with PBS again, and analyzed by confocal microscopy. Quantification of fluorescence intensity was performed with ImageJ. Error bars indicate s.d. (n = 3). *, P < 0.05.

Knockdown of SIDT2 inhibits the antisense effect induced by gymnosis

We next investigated whether the antisense effect induced by gymnosis is mediated by SIDT2. Therefore, we examined the effect of SIDT2 knockdown on the antisense effect induced by naked antisense oligonucleotides targeting miR–16. Cells were transfected with siRNA against SIDT2 or control siRNA, and cultured for 72 hours (Fig. 4A). Then, cells were incubated with or without 20 nM of naked antisense oligonucleotides targeting miR–16 for 6 hours, and the endogenous miR–16 expression levels were analyzed by qPCR. The results showed that HeLa cells transfected with siRNA targeting SIDT2 (#1 and #2) inhibited the antisense effect induced by gymnosis, compared with negative control siRNA (Fig. 4B and C). These results indicate that SIDT2 mediates the antisense effect induced by gymnosis.

Figure 4.

Knockdown of SIDT2 inhibits the antisense effect induced by gymnosis.(A) Experimental paradigm for the qRT-PCR analysis. (B and C) HeLa cells were transfected with siControl or siSIDT2#1 (B) and siControl or siSIDT2#2 (C). Seventy-two hours after transfection, cells were washed with PBS, incubated with (oligo+) or without (oligo–) 20 nM of the naked antisense oligonucleotides targeting miR–16 for 6 hours, and RNA levels were analyzed by qPCR. The miR-16 levels were normalized against β-actin mRNA levels. The miR-16 levels were expressed as percentages compared to the levels of oligo– samples (expressed as 100%). Error bars indicate s.d. (n = 6). *, P < 0.05.

Overexpression of SIDT2 enhances the uptake of naked ssRNA into cells

The effect of SIDT2 overexpression on the uptake of naked ssOligos into living cells was examined next. Cells were transfected with empty vector, mouse SIDT2 plasmid or human SIDT2 plasmid, and cultured for 48 hours (Fig. 5A and B). Then, cells were incubated with or without 500 nM of naked Alexa568-ssOligos for 24 hours and analyzed by confocal fluorescence microscopy (Fig. 5A). The Alexa568 intensities in cells overexpressing mouse or human SIDT2 were increased to 186% and 175%, respectively, compared with control cells with empty vector (Fig. 5C), showing that overexpression of SIDT2 enhances the uptake of ssRNA into cells.

Figure 5.

Effect of overexpression of SIDT2 on the uptake of naked ssRNA. (A) Experimental paradigm for the confocal microscope analysis. (B) HeLa cells were transfected with empty vector, pCI-neo/mouse SIDT2 or pCI-neo/human SIDT2 plasmids. β-actin was used as a loading control. Overexpression of SIDT2 was confirmed by western blot analysis using an anti-SIDT2 antibody (Abnova). (C) HeLa cells were transfected as indicated and cultured for 48 hours. Cells were washed with PBS, incubated with or without 500 nM of the naked Alexa568-ssOligos, washed with PBS again, and analyzed by confocal microscopy. Quantification of fluorescence intensity was performed. Error bars indicate s.d. (n = 3). *, P < 0.05.

Mutant SIDT2 does not enhance the uptake of naked ssOligos into cells

Overexpression of missense mutant SID-1 (A173T or S536I), which is found in a systemic RNAi-defective mutant C. elegans, cannot enhance the uptake of dsRNA into cells. The Ala-173 and Ser-536 of SID-1 correspond to Phe-154 and Ser-564 of mouse SIDT2, respectively. Therefore, we hypothesized that mutant SIDT2 containing a mutation on Phe-154 or Ser-536 could not mediate gymnosis. To test this hypothesis, cells were transfected with empty vector, WT SIDT2 plasmid or mutant SIDT2 (F154T or S564A) plasmids (Fig. 6A), and the effect of the mutations on the uptake of naked Alexa568-ssOligos was assessed. The Alexa568 intensities in cells overexpressing SIDT2 mutants were not significantly increased compared with that of control cells (Fig. 6B), indicating that both Phe-154 and Ser-564 of SIDT2 are essential for SIDT2-mediated gymnosis. The subcellular localization of both mutants was shown to be comparable with that of WT SIDT2, namely WT and mutant SIDT2 mainly localized to lysosomes and partly to the plasma membrane (Fig. 1B and Fig. 6C).

Figure 6.

Effect of mutations of SIDT2 on the uptake of naked ssOligos. (A) HeLa cells were transfected with empty vector, pCI-neo/WT SIDT2, pCI-neo/F154T SIDT2, or pCI-neo/S564A SIDT2 plasmids. Overexpression of SIDT2 variants was confirmed by western blot analysis using an anti-SIDT2 antibody (Abnova). β-actin was used as a loading control. (B) HeLa cells were transfected as indicated and cultured for 48 hours. Cells were washed with PBS, incubated with or without 500 nM of the naked Alexa568-ssOligos, washed with PBS again, and analyzed by confocal microscopy. Quantification of fluorescence intensity was performed. Error bars indicate s.d. (C: n = 3, D: n = 6). *, P < 0.05. n.s., not significant. (C) Hela cells expressing EGFP-tagged SIDT2 (S564A or F154T) were incubated with LysoTracker Red. Fluorescence images were acquired using a confocal laser-scanning microscope. Arrows indicate plasma membrane localization.

SIDT2 does not enhance nonselective type endocytosis

As a control experiment, we investigated whether overexpression of SIDT2 enhances endocytosis using Rhodamine B and Rhodamine B-dextran as endocytotic markers. Cells were transfected with empty vector, mouse SIDT2 plasmid or human SIDT2 plasmid, and cultured for 48 hours. Then, cells were incubated with or without 500 nM of Rhodamine B or Rhodamine B-dextran for 24 hours, and analyzed by confocal fluorescence microscopy. The Rhodamine B and Rhodamine B-dextran intensities in cells transfected with mouse SIDT2 plasmid or human SIDT2 plasmid were not changed compared with that in control cells with empty vector (Fig. 7A–C). These results indicated that SIDT2 does not enhance nonselective type endocytosis.

Figure 7.

Effect of overexpression of SIDT2 on endocytosis. (A) Experimental paradigm for the confocal microscope analysis. (B and C) HeLa cells were transfected with empty vector, pCI-neo/mouse SIDT2 or pCI-neo/human SIDT2 plasmids, and cultured for 48 hours. Cells were washed with PBS, and incubated with or without 500 nM of Rhodamine B (B) or Rhodamine B isothiocyanate-dextran (C). Cells were washed with PBS again, and analyzed by confocal microscopy. Quantification of fluorescence intensity was performed. Error bars indicate s.d. (n = 3). *, n.s., not significant.

Discussion

In the present study, we showed that the uptake of naked ssOligos into living cells is decreased in SIDT2 knockdown cells (Fig. 3D–F), and knockdown of SIDT2 inhibited the antisense effect induced by naked oligonucleotides (Fig. 4B and C). These results indicate that gymnosis is partly mediated by SIDT2.

In contrast to the ubiquitous expression of SIDT2, SIDT1 is not expressed in the cell lines tested. Overexpression of SIDT1 has been reported to enhance the uptake of dsRNA into cells. Although the ability of SIDT1 to transport ssRNA remains unknown, it is possible that SIDT1 also mediates gymnosis in SIDT1-expressing cells.

Knockdown of SIDT2 did not completely inhibit the uptake of Alexa568-ssOligos (Fig. 3), indicating the possible existence of SIDT2-independent pathway(s) for the uptake of extracellular oligonucleotides. Oligonucleotides can be taken up via clathrin-, caveolin- or dynamin-dependent endocytotic pathways, and these pathways could be the SIDT2-independent pathways.

We previously reported that SIDT2 primarily localizes to lysosomes and mediates RNautophagy/DNautophagy, an autophagic pathway in which RNA and DNA are directly taken up by lysosomes. In the present study, we showed that a portion of SIDT2 localizes to the plasma membrane, suggesting that plasma membrane SIDT2 mediates gymnosis. The localization of SIDT2 seems reasonable, considering that even a lysosomal marker protein, LAMP2, partly localizes to plasma membranes. In general, the extracellular domains of plasma membrane proteins topologically correspond to luminal regions of lysosomal membrane proteins. Therefore, from the point of view of topology of SIDT2, RNA translocation occurs in opposite directions during RNautophagy and gymnosis, and our results indicate that SIDT2 can bidirectionally transport RNA. This notion is consistent with other studies showing that SID-1 and SIDT1 mediate the bidirectional translocation of RNA on the plasma membrane. Whether SIDT2 also mediates the export of RNA from the cell is currently unknown.

The mechanism by which SIDT2 translocates RNA into living cells is currently unclear. Given that SID-1 is a putative bidirectional RNA channel, it is likely that SIDT2 also function as an RNA channel during gymnosis. Our results showing that SIDT2 can bidirectionally transport RNA support this idea. However, the possibility that SIDT2 mediates an endocytosis-like pathway cannot be completely excluded. Although we showed that overexpression of SIDT2 does not enhance the uptake of Rhodamine B or Rhodamine B dextran into living cells (Fig. 7), gymnosis could be explained by SIDT2-mediated RNA-specific endocytosis. In this scenario, a portion of RNAs taken up by cells would be released from endosomes to the cytosol. Further exploration of this issue would be of interest.

SIDT2 is a member of the putative transmembrane hydrolase superfamily. Ser-564 of SIDT2 is a conserved residue in the superfamily and is thought to be important for the putative hydrolase activity of SIDT2. We observed that overexpression of SIDT2 enhanced the uptake of naked ssOligos into living cells (Fig. 5C), whereas overexpression of the S564A mutant of SIDT2 did not (Fig. 6B). This indicated that S564 is essential for RNA translocation by SIDT2 during gymnosis. We previously showed that S564 is also indispensable for RNA translocation by SIDT2 in the process of RNautophagy. Identification of the hydrolase substrate of SIDT2 would contribute to clarify the mechanism of RNA translocation via SIDT2.

We also showed that overexpression of the F154T mutant of SIDT2 does not enhance the uptake of naked ssOligos into cells (Fig. 6B). A recent paper reported that the N-terminal extracellular domain of SIDT2 (22-292 amino acids of mouse SIDT2) shows affinity for RNA and the F154T mutation attenuates RNA binding, indicating that the Phe154 residue mediates RNA binding. Taken together with these observations, our results suggest that the RNA binding activity of the extracellular domain of SIDT2 is required for SIDT2-mediated gymnosis.

In general, nucleic acid medicines have high specificity. Some nucleic acid-based drugs may become therapeutic agents for a variety of disorders. However, Sledz et al. reported that dsRNAs such as siRNAs induce an interferon response. In many cases, the interferon response is associated with cytotoxicity. On the other hand, naked ssRNAs, such as the 7-nt sgRNA, do not induce interferon-β production and can be taken up by living cells without a transfection reagent. Therefore, our study will encourage further development of ssOligos without transfection reagents.

Our findings may contribute to the further development of nucleic acid based drugs such as ssOligos. Naked ssOligos targeting miRNAs specific for the SIDT2 mRNA or the mRNAs of other gymnosis-related genes might be able to reduce the expression of these miRNAs. Consequently, it is expected that gymnosis will be enhanced. However, because SIDT2 mainly localizes to lysosomes, the upregulation of SIDT2 expression should enhance RNautophagy and lead to RNA degradation. It is possible that ssOligos are taken up by lysosomes after the uptake into the cytoplasm. Therefore, the application of SIDT2 to enhance the functional uptake of naked ssOligos is currently difficult.

Recent papers reported that some RNAs are exported from cells and play a role in biological processes, such as cell regulation and cell-to-cell communication. These RNAs are called extracellular RNAs (exRNAs), and some exRNAs have been identified as biomarker candidates in cancer. At least a portion of exRNAs is found within vesicles such as exosomes. However, we showed that naked ssRNA is taken up by living cells via the putative RNA transporter SIDT2. Whether exRNAs exist outside of extracellular vesicles and whether gymnosis occurs during normal biological processes are interesting issues to be resolved.

Materials and methods

RNA/DNA preparation

The following fully 2′-O-methylated, 5′-Alexa568-labeled, 5′- and 3′-phosphate oligonucleotides containing phosphodiester backbones were chemically synthesized and subsequently purified by high-performance liquid chromatography as a custom service by Nippon Bioservice (Saitama, Japan): 15-nt 5′-Alexa568-labeled oligonucleotide, 5′-AAAAAAAAAAAAAAA-3′; antisense oligonucleotide targeting miR–16, 5′-CGCCAAUAUUUACGUGCUGCUA-3′; and negative control oligonucleotide, 5′-AAAAAAAAAAAAAAAAAAAAAA-3′. The following primers used for real-time reverse-transcription polymerase chain reaction (RT-PCR) were obtained from Fasmac (Kanagawa, Japan) or Hokkaido System Science (Hokkaido, Japan): SIDT2 forward primer, 5′-ATGAGTTCCCTGAAGGCGTG-3′; SIDT2 reverse primer, 5′-TCTTTGCGCTGTACGGTGAT-3′; SIDT1 forward primer A, 5′-AGATTTGGGGAATTTTCCGGCG-3′; SIDT1 reverse primer A, 5′-AGAGGGCGAAGGACCAGTTA-3′; SIDT1 forward primer B, 5′-GTGGCATCTCATCCCATTGC-3′; SIDT1 reverse primer B, 5′-GGACATCTGCCTTCCAGGAC-3′; β-actin forward primer, 5′-ACAATGTGGCCGAGGACTTT-3′; β-actin reverse primer, 5′-TGTGTGGACTTGGGAGAGGA-3′; miR–16 reverse transcription primer, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACCGCCAA-3′; miR–16 forward primer, 5′-CGCGCTAGCAGCACGTAAAT-3′; miR–16 reverse primer, 5′-GTGCAGGGTCCGAGGT-3′.

Plasmids

The plasmids pCI-neo-mSIDT2 [WT or S564A] and pCI-neo-hSIDT2 (WT) for the expression of SIDT2 variants without a tag, and pEGFP-mSIDT2 (WT or S564A) for the expression of SIDT2 with an EGFP tag at the C-terminus were prepared as described previously. Plasmids for the expression of the F154T mutant SIDT2 were constructed using the QuikChange Site-Directed Mutagenesis Kit. All resulting constructs were confirmed by sequencing.

RNA interference

For siRNA-mediated knockdown of SIDT2, cells were transfected with 10 nM SIDT2-targeting or control siRNAs. The target sequences were as follows: 5′-GAGGAUGACUACGACACAU-3′ [human SIDT2 siRNA#1], 5′-CCGCUCCUUUGAACCUGUA-3′ [human SIDT2 siRNA#2], and 5′-CAGCACGACUUCUUCAAGU-3′ [EGFP siRNA]. EGFP siRNA was used as a control.

Cell culture and transfection

WT MEFs were kind gifts from Dr. Noboru Mizushima (The University of Tokyo, Tokyo, Japan). WT MEFs and HeLa cells were cultured in Dulbecco's modified Eagle's medium (Thermo Fisher Scientific, C11995500BT) supplemented with 10% fetal bovine serum (Sigma-Aldrich, 172012) at 37°C in a 5% CO2 humidified incubator. Cells were transfected with each siRNA or each plasmid vector using Lipofectamine 3000 reagent (Thermo Fisher Scientific, L3000015) according to the manufacturer's protocol, and cultured for 72 or 48 hours, respectively. Then, transfected cells were washed with phosphate-buffered saline (PBS), and treated with naked oligonucleotides in culture medium.

Biotin-streptavidin pull-down assay

HeLa cell surface proteins were biotinylated using EZ-Link™ Sulfo-NHS-Biotin (Thermo Fisher Scientific, 21217) according to the manufacturer's protocol. Then, HeLa cell lysates were prepared in a lysis buffer (50 mM Tris-HCl pH7.5, 1% TritonX-100, 150 mM NaCl, 5 mM EDTA), and biotinylated proteins were pulled down using Streptavidin Sepharose™ High Performance (GE Healthcare, 17-5113-01) according to the manufacturer's protocol. After the beads were washed three times with lysis buffer, proteins were eluted with sodium dodecyl sulfate (SDS) sample buffer (10 mM Tris, pH 7.8, 3% SDS, 5% glycerol and 0.02% bromophenol blue, 14% 2-mercaptoethanol), and analyzed by western blotting.

Real-Time RT-PCR

Total RNA was extracted from HeLa cells using TRIzol (Thermo Fisher Scientific, 15596018), and cDNA was synthesized using a PrimeScript™ RT reagent Kit with gDNA Eraser (Perfect Real Time) (Takara, RR047A) according to the manufacturer's protocols. The SIDT2 mRNA, β-actin mRNA, and miR-16 were quantified using a CFX96™ Real-Time System (Bio-Rad) with SYBR Premix Ex Taq™ II (Tli RNaseH Plus) (Takara, RR0820A).

Western blotting

SDS polyacrylamide gel electrophoresis and western blotting were performed according to standard procedures. The antibodies used were polyclonal antibody against SIDT2 (Abnova, PAB27211), polyclonal anti-SIDT2 antibody raised in a rabbit against synthetic peptides (C+DLDTVQRDKIYVF) containing an amino acid sequence corresponding to the C-terminal region of SIDT2 (custom antibody), monoclonal antibody against β-actin (Sigma-Aldrich, A-4700), monoclonal antibody against N-cadherin (BD Biosciences, 610921), and monoclonal antibody against cathepsin D (BD Biosciences, 610800).

Confocal fluorescence microscopic analysis

Plasmid or siRNA transfected HeLa cells or untransfected cells were prepared in 35 mm glass bottom dishes (IWAKI, 3910-035 or ibidi, 81158). After cells were washed with PBS, cells were incubated with culture medium in the absence or presence of 500 nM or 1 μM of Alexa568-labeled polyA 15mer, 500 nM of Rhodamine B (Sigma-Aldrich, R6626) or 500 nM of Rhodamine B isothiocyanate-dextran (Sigma-Aldrich, R7623). Then, cells were washed with PBS unless otherwise indicated. Hoechst 33258 (Dojindo, 343–07961) was used to stain DNA and the cells were analyzed with a FLUOVIEW FV10i or FV1000 confocal microscope (Olympus). Quantification of fluorescence intensity was performed with ImageJ (NIH).

For analysis of intracellular localization, cells were overexpressed with EGFP-tagged SIDT2 (WT or mutants), and incubated with 100 nM LysoTracker Red DND 99 (Thermo Fisher Scientific, L7528) in growth medium for 1 h. Then, fluorescence image acquisition was performed with a FLUOVIEW FV10i or FV1000 confocal microscope.

Statistical analysis

For comparison of two groups, statistical differences were evaluated by Student's t-test. We used Tukey's multiple comparison test for the comparison of three groups.