• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Aug 9, 2011; 108(32): 13159–13164.
Published online Jul 25, 2011. doi:  10.1073/pnas.1108567108
PMCID: PMC3156200
Developmental Biology

Male germ cells express abundant endogenous siRNAs


In mammals, endogenous siRNAs (endo-siRNAs) have only been reported in murine oocytes and embryonic stem cells. Here, we show that murine spermatogenic cells express numerous endo-siRNAs, which are likely to be derived from naturally occurring double-stranded RNA (dsRNA) precursors. The biogenesis of these testicular endo-siRNAs is DROSHA independent, but DICER dependent. These male germ cell endo-siRNAs can potentially target hundreds of transcripts or thousands of DNA regions in the genome. Overall, our work has unveiled another hidden layer of regulation imposed by small noncoding RNAs during male germ cell development.

Keywords: RNA interference, spermatogenesis, testis, interferon response

RNA interference (RNAi) is a highly conserved gene silencing mechanism by which double-stranded RNAs (dsRNAs) are processed into single-stranded RNAs (ssRNAs) followed by loading onto effector complexes to modulate gene expression (1, 2). Small interfering RNAs (siRNAs) represent one of several distinct classes of small noncoding RNAs identified so far. Previously, siRNAs mainly referred to small ssRNAs processed in the host cells from exogenous dsRNAs (e.g., hairpin dsRNAs, synthetic short dsRNAs, etc.), and these artificial siRNAs have been widely used to suppress target gene expression both in vitro and in vivo (35). Endogenous siRNAs (endo-siRNAs) were initially identified in yeasts, plants, and Caenorhabditis elegans (69), and biogenesis of endo-siRNAs in these organisms depends on the activity of RNA-dependent RNA polymerase (RdRP), which catalyzes the replication of RNA from an RNA template (811). Double-stranded RNAs (dsRNAs) produced by RdRP are then cleaved by the RNase III DICER to generate single-stranded mature endo-siRNAs. By associating with Argonaute (AGO) proteins, siRNAs negatively regulate the expression of targeting genes at posttranscriptional levels by inducing mRNA degradation and/or translational suppression (1012). Alternatively, these endo-siRNAs can function as guidance molecules to direct associated protein factors to target specific genomic regions by DNA cytosine methylation or promoting the formation of heterochromatin (13, 14).

Although RdRP has not been found in flies or mice, certain cell types of these two species appear to be able to generate endo-siRNAs by processing the naturally occurring dsRNAs (1523). These dsRNA precursors include hairpin-dsRNAs, trans-natural antisense transcript-derived dsRNAs (trans-nat-dsRNAs), and cis-natural antisense transcript-derived dsRNAs (cis-nat-dsRNAs) (2426). Endo-siRNAs differ from microRNAs (miRNAs) in that the former are processed from long dsRNA precursors, whereas the later are mainly derived from precursors containing the short stem-loop structure, although production of both requires DICER activity in the cytoplasm. On the other hand, the microprocessor complex (DROSHA–DGCR8) is required for cleaving precursor miRNAs out of the primary miRNA transcripts in the nucleus (2729). miRNAs are thus dependent upon DROSHA activity in the nucleus, whereas endo-siRNAs do not need DROSHA in their biogenesis. In flies, endo-siRNAs have also been found to be involved in heterochromatin formation in addition to their roles in posttranscriptional regulation (30).

In mammals, endo-siRNAs have only been reported in murine oocytes and embryonic stem cells (2123). Two recent independent studies revealed that the function of miRNAs is virtually silenced during oocyte maturation and preimplantation embryonic development, whereas endo-siRNAs appear to be required for the cellular events occurring during the same time window (31, 32). Those studies suggest that endo-siRNAs play a critical role in normal development of oocytes and embryonic stem cells. Previous data appear to support the notion that both oocytes and ES cells, unlike most of the somatic cell types, lack or are insensitive to the IFN response triggered by dsRNAs because introduction of dsRNAs into these cells do not activate the OAS–RNaseL pathway, which usually leads to the cell's demise (3335). Given that the testis is an immune-privileged organ that has been shown to tolerate antigen introduction without eliciting immune responses (36, 37) and to respond poorly to RNA virus stimulation (38, 39), testicular cells, at least the developing male germ cells, may lack IFN response and thus can produce endo-siRNAs using naturally occurring long dsRNAs. Here, we show that the mouse testis, indeed, expresses numerous endo-siRNAs, which are mostly derived from trans-nat-dsRNAs and hairpin dsRNAs. These testicular endo-siRNAs could potentially target hundreds of transcripts expressed during all phases of spermatogenesis. Alternatively, they can recognize thousands of DNA sites across the genome, where they display complementary sequences to, and thus may have an epigenetic role as reported in yeasts, plants, C. elegans, and flies (13, 14, 30).

Results and Discussion

The Murine Testis Expresses Numerous Endo-siRNAs.

The size of endo-siRNAs is ~21 nt, which is distinct from that of piwi-interacting RNAs (piRNAs) at ~31 nt. The sequence of an endo-siRNA is completely complementary to its target transcripts, whereas a miRNA is usually partially complementary to their targets. Mature miRNAs are derived from precursor miRNAs, which possess the short stem-loop structures, whereas endo-siRNAs are processed from long dsRNAs without the short stem-loop structures (2426). These major characteristics distinguish endo-siRNAs from other two well-known small RNA species: piRNAs and miRNAs. Using these characteristics as criteria, we searched for endo-siRNAs in a small RNA library of the adult mouse testes sequenced using the 454 platforms (SI Appendix, Fig. S1). A total of 73 testicular endo-siRNAs were identified from approximately half a million reads, which were named endo-siRNA-T1 to -T73 (SI Appendix, Table S1). These murine testicular endo-siRNAs rarely displayed unique chromosome hits. Instead, the majority of endo-siRNAs were matched to hundreds of different sites on multiple chromosomes (Fig. 1 and SI Appendix, Table S2). For example, endo-siRNA-T19 was matched to every single chromosome, and on each chromosome except Y there were ~850 copies of this endo-siRNA (SI Appendix, Fig. S2 and Table S2). Therefore, endo-siRNAs are remarkably different from miRNAs in that miRNAs usually come from a unique locus or very few loci (40), whereas endo-siRNAs can have hundreds or even thousands of chromosome hits. A total of 57 out of the 73 testicular endo-siRNAs possessed fewer than 10 target transcripts predicted on the basis of sequence complementarity, whereas the remaining 16 displayed numerous target transcripts ranging from 10 up to 140 (Figs. 1 and and2A),2A), suggesting these endo-siRNAs can target a greater number of transcripts. Among all of the predicted target transcripts, the majority (~92%) were mRNAs and the remaining belonged to transcripts of pseudogenes (~3%), retrotransposons (~1%), and noncoding RNAs (~4%) (Fig. 2B). Among all of the mRNA targets, an average of 93% had endo-siRNA targeting sites in their 3′-UTRs, 4% in 5′-UTRs, and 3% in the coding regions (Fig. 2C). These results suggest that endo-siRNAs, like miRNAs, can target 3′-UTRs of mRNAs and thus may regulate gene expression at posttranscriptional levels.

Fig. 1.
Chromosome mapping of mouse testicular endo-siRNAs and their predicted targets. Each shade from solid blue to solid red in the spectrum represents each endo-siRNA from T1 to T73. The number and type of endo-siRNAs located in bins of ~140 kbp (corresponding ...
Fig. 2.
Transcripts predicted to be targeted by the 73 mouse testicular endo-siRNAs. (A) Number of predicted transcripts targeted by all 73 testicular endo-siRNAs. (B) Category and proportion of all predicted transcripts. Note that the majority of the predicted ...

It is noteworthy that each of these endo-siRNAs appears to have many more hits on DNA (thousands) than on RNA (hundreds) (Fig. 1 and SI Appendix, Tables S1 and S2). If these endo-siRNAs indeed interact with genomic DNA, their effects would be much greater at DNA levels than at RNA levels on the basis of the number of potential targets.

Mouse Testicular Endo-siRNAs Are Mainly Derived from Naturally Occurring dsRNAs.

dsRNAs formed by two complementary transcripts derived from different loci are called intermolecular dsRNAs. These include trans-natural antisense transcript-dsRNAs (trans-nat-dsRNAs) in which two transcripts come from different loci and cis-natural antisense transcript-dsRNAs (cis-nat-dsRNAs) in which two transcripts come from the bidirectional transcription of the same chromosome locus. Hairpin-dsRNAs, which result from complementary sequences within a single transcript, are called intramolecular dsRNAs (2426). All three types of long dsRNAs could serve as precursors for endo-siRNA production. For example, Tmod1 and Tstd2 are two neighboring genes on chromosome (Chr.) 4 with the opposite orientations, which partially overlap in their last exons. This overlapping region can be transcribed in a bidirectional manner, which can result in the potential formation of a cis-nat-dsRNA. This cis-nat-dsRNA is formed by the partial sequences of the 3′-UTRs of these two mRNAs and seeds three endo-siRNAs (endo-siRNAs-T26, -T32, and -T71) (Fig. 3A). Hsd3b2 locates on Chr. 3 and Zfp488 locates on Chr. 14. The partial 3′-UTR sequences of these two mRNAs are complementary to each other; these 3′-UTRs could, therefore, form a trans-nat-dsRNA, which contains two endo-siRNAs (endo-siRNAs-T27 and -T45) (Fig. 3B). The Zfp353 transcript possesses extensive 5′-UTR and 3′-UTR, and their sequences are partially complementary to each other. Therefore, a large hairpin-dsRNA can be formed by the noncoding sequences of this mRNA, which seeds one endo-siRNA, endo-siRNA-T50 (Fig. 3C).

Fig. 3.
Double-stranded RNA (dsRNA) precursors for mouse testicular endo-siRNAs. (A) Representative cis-natural antisense transcript-derived dsRNA (cis-nat-dsRNA) precursor for three testicular endo-siRNAs (T26, T32, and T71). (B) Representative trans-natural ...

Although our endo-siRNAs could be mapped to multiple chromosome sites, we found dsRNA precursors for only 42 of the 73 endo-siRNAs. This was likely due to the incomplete collection of transcripts in the currently available databases. A total of 60 intermolecular dsRNA precursors were studied here, which could potentially generate 42 unique endo-siRNAs. We further analyzed these precursor dsRNAs for length, percentage of complementarity, and transcript sources. The length of these dsRNAs ranged from 118 to 1,400 bp, with an average of 337 bp. The percentage of complementarity between the two strands ranged from 80 to 100%, with an average of 90% (Fig. 3D and SI Appendix, Dataset 1A). Among all 60 intermolecular dsRNAs, 58 were trans-nat-dsRNAs whereas only two appeared to be cis-nat-dsRNAs (Fig. 3D and SI Appendix, Dataset 1A). By studying the transcript sources for dsRNA precursors, we found that 40 out of the 60 predicted intermolecular dsRNA precursors formed by paring between two 3′-UTRs (Fig. 3E and SI Appendix, Dataset 1A). In addition, we found 17 intramolecular dsRNA precursors that could potentially produce 12 endo-siRNAs. The average length of these hairpin-dsRNAs was 176 bp, and the average percentage of complementarity between the two strands was 83% (Fig. 3F and SI Appendix, Dataset 1B). The 5′-UTR and the 3′-UTR of the Zfp353 transcript could form a hairpin dsRNA, whereas the other 16 hairpin-dsRNAs were derived from hairpin folding of single 3′-UTRs (SI Appendix, Dataset 1B).

It is quite remarkable that almost all of the dsRNA precursors are formed between noncoding regions of the mRNAs (either 5′-UTRs or 3′-UTRs) and/or pseudogene or transposon transcripts. This finding suggests that endo-siRNAs possess a lower degree of sequence specificity toward their target transcripts because they appear to mainly come from 3′-UTRs and predominantly target 3′-UTRs of transcripts. This may allow endo-siRNAs to target a greater number of genes in the developing male germ cells. Because the formation of intermolecular dsRNA precursors requires concurrent expression of two transcripts, we examined the spatiotemporal expression profiles of these predicted precursor transcripts during the testicular development using semiquantitative PCR and/or by analyzing the Gene Expression Omnibus (GEO) database (SI Appendix, Fig. S3). Transcript pairs predicted to form intermolecular dsRNA precursors appeared to display similar or at least partially overlapping spatiotemporal expression patterns during testicular development, suggesting that these predicted dsRNAs indeed could be formed in vivo. The same analyses were also performed for the transcripts that were predicted to form intramolecular dsRNA precursors, and results confirmed that they did express during testicular development.

Stage-Specific Expression of Testicular Endo-siRNAs in Developing Male Germ Cells.

To validate the expression of these cloned endo-siRNAs in vivo, we examined their expression levels in developing testes (Fig. 4) and purified spermatogenic cells (Fig. 5 and SI Appendix, Fig. S7). We determined the expression profiles for 58 of the 73 testicular endo-siRNAs because the PCR method used requires the use of sequences of small RNAs as the forward primers (41), whereas the sequences of some endo-siRNAs were not qualified as workable primers. Four general expression patterns were observed among the 58 endo-siRNAs examined (Fig. 4): (i) Eleven endo-siRNAs displayed an onset of expression at approximately postnatal day 14 (P14), followed by increasing levels from P17 to P28 and decreasing levels from P35 to adulthood. These endo-siRNAs are likely to be expressed in pachytene spermatocytes and/or round spermatids because the developmental timing coincides with the first appearance and accumulation of these spermatogenic cells in the testis (i.e., pachytene spermatocytes first appear at P14 and round spermatids first appear at P20). (ii) Fifteen endo-siRNAs were first detected at ~P14 and their levels kept increasing thereafter until adulthood, suggesting these endo-siRNAs are mainly expressed in pachytene spermatocytes, round and elongated spermatids. (iii) Four endo-siRNAs (T10, T9, T64, and T12) showed the highest levels of expression between birth and P14, and levels diminished drastically thereafter, suggesting that these endo-siRNAs are either expressed in the somatic cell types (Sertoli or Leydig cells) or spermatogonia because these cells are proportionally diluted significantly by the increasing number of meiotic and haploid germ cells. (iv) The remaining 28 endo-siRNAs displayed relatively constant levels in the developing testes, suggesting these endo-siRNAs may represent those housekeeping ones. These dynamic expression patterns imply that these endo-siRNAs are expressed in specific stages during male germ cell development. Moreover, these endo-siRNAs are mainly expressed in developing germ cells because levels should have decreased if they are solely expressed in somatic cell components of the testis due to the dilution effects of the increasing number of spermatogenic cells during testicular development (42). Indeed, all except four endo-siRNAs with early expressions (pattern 3 described above) were detected in purified pachytene spermatocytes (see below). Moreover, their levels were all significantly down-regulated in the Dicer-null, but not in Drosha-null spermatogenic cells (see below), supporting the notion that these endo-siRNAs are expressed in germ cells rather than somatic cell types in the testis.

Fig. 4.
Expression profiles of testicular endo-siRNAs during development. Semiquantitative PCR was performed to determine levels of 58 testicular endo-siRNAs in the testes at the ages of newborn, postnatal day 7 (P7), P10, P14, P17, P21, P28, P35, and adult. ...
Fig. 5.
Levels of 54 endo-siRNAs in Drosha- or Dicer-null pachytene spermatocytes. TaqMan-based quantitative real-time PCR was used to quantify levels of endo-siRNAs in pachytene spermatocytes purified from the testes of the control (Stra8-iCre-Rosa26mTmG+/tg ...

Biogenesis of Testicular Endo-siRNAs Is DROSHA Independent, but DICER Dependent.

Given that endo-siRNAs are mainly derived from DICER-mediated processing of dsRNAs formed in the cytoplasm, the true endo-siRNAs should be independent of the microprocessor (DROSHA–DGCR8 complex) activity, which is required for precursor miRNA production in the nucleus. To test whether the endo-siRNAs that we identified were affected by either Dicer or Drosha inactivation, we generated two conditional knockout mouse lines with postnatal germ cell-specific inactivation of Dicer or Drosha, respectively, by crossing a postnatal male germ cell-specific Cre line, Stra8-iCre, with Dicer-loxp or Drosha-loxp line. The Stra8-iCre male mice start to express Cre exclusively in spermatogonia starting at postnatal day 3 (43). To visualize the Cre-expressing cells, we further crossed the Stra8-iCre-Dicer+/lox mice with Rosa26mTmGtg/tg-Dicerlox/lox mice and generated compound, male germ cell-specific Dicer knockout mice (Stra8-iCre-Dicerlox/lox-Rosa26mTmG+/tg), in which all Cre-expressing (true knockout) cells were green due to Cre-mediated activation of eGFP expression (44). Similarly, Stra8-iCre-Droshalox/lox-Rosa26mTmg+/tg male mice were generated for the purification of Drosha-null pachytene spermatocytes. Both Stra8-iCre-Dicerlox/lox and Stra8-iCre-Droshalox/lox males were infertile due to oligozoospermia or azoospermia caused by constant depletion of pachytene spermatocytes and spermatids in the adult mouse testes. However, the germ cell depletion was progressive with age, and at P25, a portion of Dicer-null or Drosha-null spermatocytes and round spermatids were being depleted, but the majority of spermatogenic cells remained normal looking within the seminiferous tubules (SI Appendix, Fig. S4A), which allowed us to purify pachytene spermatocytes using the STA-PUT method (45) for molecular analyses reported below.

We purified pachytene spermatocytes from the control (Stra8-iCre-Rosa26mTmG+/tg), Drosha KO (Stra8-iCre-Droshalox/lox-Rosa26mTmg+/tg), and Dicer KO (Stra8-iCre-Dicerlox/lox-Rosa26mTmG+/tg) testes and the purified pachytene spermatocytes displayed a purity ranging from 85 to 95% on the basis of microscopic examination (green vs. red cells) (SI Appendix, Fig. S4B) and quantitative real-time PCR analyses of marker genes for different types of testicular cells (SI Appendix, Fig. S5). Levels of Dicer and Drosha mRNAs, compared with those of the controls, were drastically reduced in pachytene spermatocytes purified from Dicer KO (Stra8-iCre-Dicerlox/lox-Rosa26mTmG+/tg) and Drosha KO (Stra8-iCre-Droshalox/lox-Rosa26mTmg+/tg) testes, respectively (SI Appendix, Fig. S6 A and B). Furthermore, levels of 12 miRNAs known to be expressed in pachytene spermatocytes (46) were all significantly lowered in both purified Dicer-null and Drosha-null pachytene spermatocytes than in the controls (SI Appendix, Fig. S6C). Together, these data confirmed the effective inactivation of Dicer and Drosha in the pachytene spermatocytes of Stra8-iCre-Dicerlox/lox-Rosa26mTmG+/tg and Stra8-iCre-Droshalox/lox-Rosa26mTmg+/tg mice, respectively.

In purified pachytene spermatocytes, 54 out of the 58 endo-siRNAs examined showed a similar pattern of changes with significantly lowered levels in Dicer-null cells (P < 0.001, n = 54), but not in Drosha-null cells (Fig. 5 and SI Appendix, Fig. S7). Average levels of these endo-siRNAs in Dicer-null pachytene spermatocytes were reduced by ~75% compared with those in the controls (Fig. 5). In contrast, these endo-siRNAs displayed an average of ~27% decrease in levels in the Drosha-null pachytene spermatocytes compared with the controls (Fig. 5). Given that the depletion of those Drosha-null pachytene spermatocytes was ongoing at P25 (SI Appendix, Fig. S4A), this degree of decrease was likely caused by down-regulation of global transcription or transcript stability in cells that were undergoing apoptosis. Overall, these data demonstrate that biogenesis of these testicular endo-siRNAs requires DICER, but not DROSHA. This is consistent with the findings on endo-siRNAs identified in mouse oocytes and embryonic stem cells (23, 31). The expression pattern of these endo-siRNAs (SI Appendix, Fig. S7) was different from that of miRNAs (SI Appendix, Fig. S6C) in the same sets of samples, further demonstrating that miRNA production is dependent upon both DROSHA and DICER activities, whereas biogenesis of endo-siRNAs requires DICER, but not DROSHA. Expression profiles of four endo-siRNAs (T10, T9, T64, and T12) during testicular development (Fig. 4) suggested that they were expressed mainly in spermatogonia or in testicular somatic cell types (Sertoli and Leydig cells). No significant changes in levels of these four endo-siRNAs in the control, Dicer-null, or Drosha-null pachytene spermatocytes further support that these endo-siRNAs are mainly expressed in spermatogonia and/or somatic cells within the testis.

We also examined the expressions of these 58 endo-siRNAs in human testes. Although primers for these endo-siRNAs were designed on the basis of the mouse sequences, PCR products with the expected sizes were observed in both mouse and human testes samples (SI Appendix, Fig. S8). We then sequenced the PCR products and the sequencing results confirmed that sequences of these testicular endo-siRNAs were exactly the same between mice and humans (sequencing data available upon request). These data suggest that endo-siRNAs are conserved in male germ cells between mice and humans.

We also examined the tissue distribution of these testicular endo-siRNAs by PCR using multiple mouse organs including brain, heart, liver, spleen, lung, kidney, small intestine, stomach, ovary, and uterus. We detected PCR products of slightly larger sizes, and sequencing analyses revealed that these small RNAs amplified from tissues other than the testis were ~31 nt long and their sequences overlapped with those of the testicular endo-siRNAs, which are 21 nt long. Because partial sequences of these longer small RNAs are the same as those testicular endo-siRNAs, the PCR primers designed on the basis of endo-siRNA sequences could not distinguish these two types of small RNAs, resulting in amplification of endo-siRNAs in the testis and those larger small RNAs in other tissues. The larger small RNAs appear to be similar to those piRNA-like RNAs identified in tissues other than the testis previously (47). Further investigation on this small RNA species is underway. These findings suggest that these testicular endo-siRNAs are exclusively expressed in the testis, more specifically in the testicular germ cells. None of the 73 testicular endo-siRNAs was found among the mouse oocyte- or ES cell-expressed endo-siRNAs previously reported (2123). Although there might be more endo-siRNAs that are yet to be identified in both male and female germ cells, two unique sets of endo-siRNAs identified so far in the male and female gametes, respectively, suggest that male germ cell-expressed endo-siRNAs have functions that are specific to the male germ cell developmental program and vice versa.

Endo-siRNAs Effectively Induce Target mRNA Degradation in Vitro.

Endo-siRNA-T6 and its two predicted targets—Frmpd1 and Lrrc2—displayed similar expression patterns during testicular development (Fig. 4 and SI Appendix, Fig. S9A). Similar expression patterns were also observed between endo-siRNA-T40 and its predicted target Kif17 and between endo-siRNA-T19 and its predicted targets Spata1 and Syt11 (Fig. 4 and SI Appendix, Fig. S9A). Similar spatiotemporal expression patterns between endo-siRNAs and their predicted targets support a true targeting relationship in vivo To further test the target relationship between endo-siRNAs and predicted target transcripts, we performed in vitro luciferase assays using National Institutes of Health (NIH) 3T3 cells. Three endo-siRNAs and five predicted target transcripts were not expressed in NIH 3T3 cells (SI Appendix, Fig. S9 B and C), which eliminated the endogenous influences on the results.

The native firefly luciferase plasmid and the plasmid bearing Klhl10 3′-UTR were used as controls because neither contained any targeting sites for the three endo-siRNAs tested. No significant changes in levels of firefly mRNAs were detected in the control groups, whereas highly reduced mRNA levels were consistently observed in experimental groups (Fig. 6). Moreover, the efficiency for each of the three endo-siRNAs tested to degrade the target mRNAs was >50% (Fig. 6). Accordingly, protein production was significantly suppressed in all three sets of experiments (SI Appendix, Fig. S9D). These data demonstrated that endo-siRNAs could effectively induce degradation of their target mRNAs in vitro. It is conceivable that these testicular endo-siRNAs may exert similar effects in vivo, causing degradation of their target mRNAs.

Fig. 6.
Endo-siRNAs can induce degradation of their target mRNAs in vitro. In a luciferase-based reporter assay, NIH 3T3 cells were cotransfected with firefly luciferase expression plasmids bearing the 3′-UTR of Frmpd1, Lrrc2, Kif17, Spata1, or Syt11 ...

In addition to their role as a posttranscriptional regulator, endo-siRNAs of plants and yeasts have been found to function as an epigenetic regulator and thus affect transcriptional activities of genes through either RNA-directed DNA methylation (RdDM) mechanism or RNA-mediated heterochromatin formation (13, 14). In both cases, endo-siRNAs may act as guide molecules to direct their associated epigenetic factors [e.g., site-specific DNA methyltransferase (MET1) or histone methyltransferases (HMTs)] to target specific genomic regions through DNA cytosine methylation or promoting the formation of heterochromatin, respectively. Interestingly, like yeast and plants, endo-siRNAs appeared to be involved in chromatin modifications in flies (30). Given that the testicular endo-siRNAs display numerous hits on multiple chromosomes (Fig. 1), it is intriguing to further explore whether these murine endo-siRNAs have any nuclear effects on methylation and/or chromatin modification in addition to their well-established cytoplasmic roles as posttranscriptional regulators.

Materials and Methods

Quantitative analyses of endo-siRNAs, miRNAs, and other transcripts were conducted as described in refs. 41 and 46. Spermatogenic cells were purified as described in refs. 42 and 46. Bioinformatic analyses were performed using an in-house computer program. For an extensive description of the materials and methods, see SI Appendix, Materials and Methods.

Supplementary Material

Corrected Supporting Information:


The Drosha-loxp mouse line was provided by Dr. Dan R. Littman, Skirball Institute of Biomolecular Medicine, New York University School of Medicine. This work was supported by Grants HD050281 and HD060858 from the National Institutes of Health (NIH) (to W.Y.). The software for bioinformatic analyses was developed in the Imaging Core (Core D), with support by Centers of Biomedical Research Excellence Grant P20-RR18751 from the NIH.


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1108567108/-/DCSupplemental.


1. Hannon GJ. RNA interference. Nature. 2002;418:244–251. [PubMed]
2. Tomari Y, Zamore PD. Perspective: Machines for RNAi. Genes Dev. 2005;19:517–529. [PubMed]
3. Elbashir SM, et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature. 2001;411:494–498. [PubMed]
4. Paddison PJ, Caudy AA, Hannon GJ. Stable suppression of gene expression by RNAi in mammalian cells. Proc Natl Acad Sci USA. 2002;99:1443–1448. [PMC free article] [PubMed]
5. Lu PY, Xie F, Woodle MC. In vivo application of RNA interference: From functional genomics to therapeutics. Adv Genet. 2005;54:117–142. [PubMed]
6. Cogoni C, Macino G. Post-transcriptional gene silencing across kingdoms. Curr Opin Genet Dev. 2000;10:638–643. [PubMed]
7. Ruby JG, et al. Large-scale sequencing reveals 21U-RNAs and additional microRNAs and endogenous siRNAs in C. elegans. Cell. 2006;127:1193–1207. [PubMed]
8. Pak J, Fire A. Distinct populations of primary and secondary effectors during RNAi in C. elegans. Science. 2007;315:241–244. [PubMed]
9. Sijen T, Steiner FA, Thijssen KL, Plasterk RH. Secondary siRNAs result from unprimed RNA synthesis and form a distinct class. Science. 2007;315:244–247. [PubMed]
10. Vaucheret H. Post-transcriptional small RNA pathways in plants: Mechanisms and regulations. Genes Dev. 2006;20:759–771. [PubMed]
11. Martienssen RA, Zaratiegui M, Goto DB. RNA interference and heterochromatin in the fission yeast Schizosaccharomyces pombe. Trends Genet. 2005;21:450–456. [PubMed]
12. Yigit E, et al. Analysis of the C. elegans Argonaute family reveals that distinct Argonautes act sequentially during RNAi. Cell. 2006;127:747–757. [PubMed]
13. Verdel A, Vavasseur A, Le Gorrec M, Touat-Todeschini L. Common themes in siRNA-mediated epigenetic silencing pathways. Int J Dev Biol. 2009;53:245–257. [PubMed]
14. Lejeune E, Allshire RC. Common ground: Small RNA programming and chromatin modifications. Curr Opin Cell Biol. 2011;23:258–265. [PubMed]
15. Ghildiyal M, et al. Endogenous siRNAs derived from transposons and mRNAs in Drosophila somatic cells. Science. 2008;320:1077–1081. [PMC free article] [PubMed]
16. Okamura K, et al. The Drosophila hairpin RNA pathway generates endogenous short interfering RNAs. Nature. 2008;453:803–806. [PMC free article] [PubMed]
17. Czech B, et al. An endogenous small interfering RNA pathway in Drosophila. Nature. 2008;453:798–802. [PMC free article] [PubMed]
18. Kawamura Y, et al. Drosophila endogenous small RNAs bind to Argonaute 2 in somatic cells. Nature. 2008;453:793–797. [PubMed]
19. Okamura K, Balla S, Martin R, Liu N, Lai EC. Two distinct mechanisms generate endogenous siRNAs from bidirectional transcription in Drosophila melanogaster. Nat Struct Mol Biol. 2008;15:581–590. [PMC free article] [PubMed]
20. Chung WJ, Okamura K, Martin R, Lai EC. Endogenous RNA interference provides a somatic defense against Drosophila transposons. Curr Biol. 2008;18:795–802. [PMC free article] [PubMed]
21. Watanabe T, et al. Endogenous siRNAs from naturally formed dsRNAs regulate transcripts in mouse oocytes. Nature. 2008;453:539–543. [PubMed]
22. Tam OH, et al. Pseudogene-derived small interfering RNAs regulate gene expression in mouse oocytes. Nature. 2008;453:534–538. [PMC free article] [PubMed]
23. Babiarz JE, Ruby JG, Wang Y, Bartel DP, Blelloch R. Mouse ES cells express endogenous shRNAs, siRNAs, and other Microprocessor-independent, Dicer-dependent small RNAs. Genes Dev. 2008;22:2773–2785. [PMC free article] [PubMed]
24. Nilsen TW. Endo-siRNAs: Yet another layer of complexity in RNA silencing. Nat Struct Mol Biol. 2008;15:546–548. [PubMed]
25. Golden DE, Gerbasi VR, Sontheimer EJ. An inside job for siRNAs. Mol Cell. 2008;31:309–312. [PMC free article] [PubMed]
26. Okamura K, Lai EC. Endogenous small interfering RNAs in animals. Nat Rev Mol Cell Biol. 2008;9:673–678. [PMC free article] [PubMed]
27. Denli AM, Tops BB, Plasterk RH, Ketting RF, Hannon GJ. Processing of primary microRNAs by the Microprocessor complex. Nature. 2004;432:231–235. [PubMed]
28. Gregory RI, et al. The Microprocessor complex mediates the genesis of microRNAs. Nature. 2004;432:235–240. [PubMed]
29. Han J, et al. The Drosha-DGCR8 complex in primary microRNA processing. Genes Dev. 2004;18:3016–3027. [PMC free article] [PubMed]
30. Fagegaltier D, et al. The endogenous siRNA pathway is involved in heterochromatin formation in Drosophila. Proc Natl Acad Sci USA. 2009;106:21258–21263. [PMC free article] [PubMed]
31. Suh N, et al. MicroRNA function is globally suppressed in mouse oocytes and early embryos. Curr Biol. 2010;20:271–277. [PMC free article] [PubMed]
32. Ma J, et al. MicroRNA activity is suppressed in mouse oocytes. Curr Biol. 2010;20:265–270. [PMC free article] [PubMed]
33. Wianny F, Zernicka-Goetz M. Specific interference with gene function by double-stranded RNA in early mouse development. Nat Cell Biol. 2000;2:70–75. [PubMed]
34. Yang S, Tutton S, Pierce E, Yoon K. Specific double-stranded RNA interference in undifferentiated mouse embryonic stem cells. Mol Cell Biol. 2001;21:7807–7816. [PMC free article] [PubMed]
35. Stein P, Zeng F, Pan H, Schultz RM. Absence of non-specific effects of RNA interference triggered by long double-stranded RNA in mouse oocytes. Dev Biol. 2005;286:464–471. [PubMed]
36. Fijak M, Bhushan S, Meinhardt A. Immunoprivileged sites: The testis. Methods Mol Biol. 2011;677:459–470. [PubMed]
37. Fijak M, Meinhardt A. The testis in immune privilege. Immunol Rev. 2006;213:66–81. [PubMed]
38. Dejucq N, Liénard MO, Jégou B. Interferons and interferon-induced antiviral proteins in the testis. J Reprod Immunol. 1998;41:291–300. [PubMed]
39. Melaine N, et al. Production of the antiviral proteins 2’5'oligoadenylate synthetase, PKR and Mx in interstitial cells and spermatogonia. J Reprod Immunol. 2003;59:53–60. [PubMed]
40. He L, Hannon GJ. MicroRNAs: Small RNAs with a big role in gene regulation. Nat Rev Genet. 2004;5:522–531. [PubMed]
41. Ro S, Park C, Jin J, Sanders KM, Yan W. A PCR-based method for detection and quantification of small RNAs. Biochem Biophys Res Commun. 2006;351:756–763. [PMC free article] [PubMed]
42. Bellvé AR, et al. Spermatogenic cells of the prepuberal mouse. Isolation and morphological characterization. J Cell Biol. 1977;74:68–85. [PMC free article] [PubMed]
43. Sadate-Ngatchou PI, Payne CJ, Dearth AT, Braun RE. Cre recombinase activity specific to postnatal, premeiotic male germ cells in transgenic mice. Genesis. 2008;46:738–742. [PMC free article] [PubMed]
44. Muzumdar MD, Tasic B, Miyamichi K, Li L, Luo L. A global double-fluorescent Cre reporter mouse. Genesis. 2007;45:593–605. [PubMed]
45. Bellvé AR. Purification, culture, and fractionation of spermatogenic cells. Methods Enzymol. 1993;225:84–113. [PubMed]
46. Song R, et al. Many X-linked microRNAs escape meiotic sex chromosome inactivation. Nat Genet. 2009;41:488–493. [PMC free article] [PubMed]
47. Ro S, et al. Cloning and expression profiling of testis-expressed piRNA-like RNAs. RNA. 2007;13:1693–1702. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...