Next-generation Sequencing Reveals Differentially Expressed Small Non-coding RNAs in Uterine Leiomyoma

Abstract

Objective

To determine the expression profile of small non-coding RNAs (sncRNAs) which have not been investigated to date in leiomyoma.

Design

Laboratory based investigation.

Setting

Academic center.

Patients

Women undergoing hysterectomy for benign indications.

Intervention

Next generation sequencing and screening of sncRNA database with confirmatory analysis by QRT-PCR.

Main Outcome Measure

Expression profile of sncRNAs in leiomyoma and matched myometrium.

Results

Screening our previously determined RNA sequencing data with sncRNAs database resulted in identification of 15 snRNAs, 284 snoRNAs, 98 piRNAs, 152 tRNAs and 45 rRNAs of which 15 snoRNAs, 24 piRNAs, 7 tRNAs and 6 rRNAs were differentially expressed at 1.5 fold change cutoff in leiomyoma as compared to myometrium. We selected 5 snoRNAs, 4 piRNAs, 1 tRNA, and 1 rRNA which were differentially expressed and confirmed their expression in paired tissues (N=20) from both phases of the menstrual cycle using QRT-PCR. The results indicated upregulation of snoRNAs (SNORD30, SNORD27, SNORA16A, SNORD46 and SNORD56) and down-regulation of piRNAs (piR-1311, piR-16677, piR-20365, piR-4153), tRNA (TRG-GCC5-1), and rRNA (RNA5SP202) expression in leiomyoma as compared to myometrium (P<0.05). The pattern of expression of these sncRNAs was similar to RNA sequencing analysis with no menstrual cycle-dependent differences detected except for SNORD30. Since Argonaute 2 (AGO2) is required for sncRNAs mediated gene silencing, we determined its expression and found greater abundance in leiomyoma.

Conclusion

Our results provide the first evidence for the differential expression of additional classes of sncRNAs and AGO2 in leiomyoma, and implicate their roles as a gene regulatory mechanism.

Introduction

Leiomyomas are benign uterine tumors with an unknown etiology affecting 40-70% of women during their reproductive years (1). Symptomatic tumors are a major cause of chronic pelvic pain, abnormal uterine bleeding and associated infertility, and account for over 30% of all hysterectomies performed in the United States annually (1). Although ovarian steroids are known to be key regulators of leiomyoma growth, altered expression of many protein-coding genes (1–3) as well as genetic heterogeneity associated with chromosomal re-arrangements and mutation in a number of genes have also been associated with their development and growth progression (1). In addition to protein-coding genes, conventional, microarray and recent high-throughput sequencing have provided further evidence for the expression of a large number of non-protein-coding RNAs (ncRNAs) in leiomyoma (4–6).

As of to date tens of thousands of non-protein coding RNAs (ncRNAs) derived from different genomic loci have been identified. These RNAs are classified as small (17–120 nucleotides long) and long (>200 nucleotides long) non-coding RNAs (sncRNAs and lncRNAs) (7). SncRNAs are transcribed as precursors of 60- to 300 nucleotides long and in their mature form are further classified as miRNAs (17–22 nucleotides), small nuclear/nucleolar RNAs (snRNA and snoRNAs)[70–120 nucleotides], PIWI-interacting RNAs (piRNAs)[26–33 nucleotides]. SnoRNAs are highly conserved and classified as C/D box (SNORDs) or H/ACA box (SNORAs) based on their sequence and structure (8). MiRNAs have been most intensively investigated in the past few years, and accumulated evidence supports their regulatory function on a vast number of protein-coding genes involved in various normal cellular activities. Aberrant expression of miRNAs is associated with a wide range of disorders, including tumorigenesis and tissue fibrosis (9, 10). In previous studies including our own high throughput sequencing identified the expression profile of a large number of miRNAs in leiomyoma and myometrium and provided support for altered expression and regulatory function of a number of them, including let7, miR-21, miR-29, miR-200 and miR-25/93/106 cluster in leiomyoma and leiomyoma smooth muscle cells (6, 11–15).

Far less is known about the function and expression profile of other members of sncRNAs. Functionally snoRNAs play a central role in modifying and processing sncRNAs, ribosomal (rRNA) and transfer RNA (tRNA) (16). PiRNAs are the largest class of sncRNA family and functionally regulate epigenetic, post-transcriptional gene expression and degradation (17). Transfer RNAs (tRNAs) are derived from pre-tRNA by endonuclease RNase Z and in their mature form play a well-defined role in protein translation (18). However, shorter tRNA fragments (tRFs) of 30–35 nucleotides long are also generated through cleavage of mature tRNAs at both 5′ and 3′ end at the anticodon loop by angiogenin (19). In addition, snRNAs, snoRNAs, tRNAs and rRNAs can be further processed into miRNAs or siRNAs, and incorporated into Argonaute proteins to form RNA-induced silencing complex (RISC) for cleavage, degradation or inhibition of translation of downstream target mRNA (20).

Argonaute proteins are the active part of RISC and play a central role in RNA silencing processes, resulting in RNA degradation or translation inhibition (21). In humans, there are eight Argonaute family members and are separated into two subfamilies: 1). four members of Argonaute-like subfamily interact with miRNAs and siRNAs, and 2). four members of PIWI-like subfamily bind to piRNAs (20). The Argonaute family members are composed of four characteristic domains: the N- terminal domain, the PIWI-Argonaute-Zwille (PAZ) domain, the middle (MID) domain and the C-terminal PIWI domain (20). Among them, the PIWI domain, structurally resembles RNase H, is essential for target RNA cleavage through an active catalytic motif “Asp-Asp-Asp/His/Glu/Lys” that harbors a divalent metal ion, and is necessary for catalysis. The PIWI domain also mediates protein-protein interaction with Dicer at one of its RNase III domain (20). Although four Argonaute-like proteins are capable of loading miRNAs or siRNAs, only Argonaute 2 (AGO2) has endonuclease activity in human, thus playing a key role in gene silencing (22).

The identity and pattern of expression of snRNAs, snoRNAs, piRNAs, tRNAs and rRNAs as well as AGO2 in leiomyoma and myometrium has not been evaluated to date. As such in the present study using our existing high-throughput sequencing data set generated from leiomyoma and paired myometrium we carried out an in-depth data analysis to identify the expression profile of snRNAs, snoRNAs, piRNAs, tRNAs and rRNAs. Using QRT-PCR we selected and confirmed the expression of a number of differentially expressed sncRNAs along with detection of AGO2 protein in twenty pairs of leiomyoma and matched myometrium tissues from both phases of the menstrual cycle.

Materials and Methods

Myometrium and Leiomyoma Tissues Collection

Leiomyomas (3–5 cm in diameters) and paired myometrium (N=20) were collected at Harbor-UCLA Medical Center with prior approval obtained from Institutional Review Board at LA BioMed Research Institute at Harbor-UCLA Medical Center. The paired tissues were from white Hispanic (N=15) and African-American (N=5) women ages ranging from 30 to 49 years (mean=42±5.6 years). This sample set included 3 paired tissues previously used for high throughput sequencing (4). Based on endometrial histology the paired leiomyoma and myometrium were from follicular (N=10) and luteal (N=10) phase of the menstrual cycle. The tissues were snap-frozen and stored in liquid nitrogen for further analysis.

Total RNA isolation and RNA sequencing

Total RNA was isolated from leiomyoma and matched myometrium using miRNeasy Mini Kit (Qiagen, Germantown, MD) and RNA concentration and integrity were determined as previously described (4). Samples with RNA integrity numbers (RIN) greater than or equal to 9 were used for library preparation. The RNA sequencing was carried out at UCLA Clinical Microarray Core Facility (http://pathology.ucla.edu/tcgb) as previously described (4). Briefly, for the production of small RNA sequencing libraries 500 ng of total RNA from each tissue was used and libraries were prepared according to manufacturer’s instructions of the Truseq small-RNA kit (Illumina, San Diego, CA). Each library was then pooled and sequenced on a MiSeq single-ended 35-bp run (Illumina) to 10 million reads with 80-90% alignment per library. Our sequencing GEO accession number is GSE100338.

Assembly of snRNA, snoRNA, piRNA, tRNA and rRNA annotations

For data quality control, FASTQC was used to check the raw fastq data quality and Trimmomatic was used to remove adaptors and to trim quality bases. After adapter clipping, we removed leading and trailing ambiguous or low quality bases (below Phred quality scores of 33). Trimmomatic works with a user-defined window spanning the read from 5′ to 3′ and removes bases only at the 3′-end; we set up a window length of 4 and a quality threshold Q of 20. When the average quality droped below 20, the 3′-end was clipped. The low-quality reads and reads shorter than 15nt were trimmed by Cutadapt (23). The filtered reads were mapped to known human sncRNA reference index (24), including mature miRNA (miRBase, release 21), precursor miRNA (miRBase, release 21), tRNA (Genomic tRNA Database), piRNA (piRBase), rRNA (ENSEMBL, release 76), snRNA, snoRNA and mitochondrial RNA (ENSEMBL, release 76). Read count outputs from AASRA pipeline (An Anchor Alignment-Based Small RNA Annotation Pipeline) (25) were statistically normalized by DESeq2 (26). Analyzed read counts for snRNA, snoRNA, piRNA, tRNA and rRNA were further sorted based on 1.5 fold change cutoff and p<0.05.

Quantitative RT-PCR Analysis

Total RNA was extracted from leiomyoma and matched myometrium (N=20) using TRIzol (Thermo Fisher Scientific, Carlsbad, CA) and their quantity and quality was determined (ND-1000 Spectrophotometer, NanoDrop Technologies, Wilmington, DE) as previously described (27–29). One μg of RNA was reverse transcribed using random primers for snoRNAs detection. Quantitative RT-PCR was carried out using SYBR gene expression master mix (Applied Biosystems). Reactions were incubated for 10 min at 95°C followed by 40 cycles for 15 seconds at 95°C and 1 min at 60°C. For piRNAs, tRNA and rRNA detection, primer design and PCR conditions were followed as described previously (30, 31). The level of sncRNAs was determined using Invitrogen StepOne System with RNU6B used for normalization. All reactions were run in triplicate and relative expression was analyzed with the comparative cycle threshold method (2^−ΔΔCT) according to the manufacturer (Applied Biosystems). Values were expressed as fold change compared to the control group. The primer sequences used are listed in supplementary Table 1.

Immunoblotting

Total protein was isolated from leiomyoma and paired myometrium (N=20) and subjected to immunoblotting as previously described (32, 33). AGO2 antibody (Cell Signaling Technology, Inc., Danvers, MA) was used for AGO2 detection. The band densities were determined using image J program (http://imagej.nih.gov/ij/), normalized to a band obtained from staining the membrane with Ponceau S. Results were expressed as a ratio relative to the control group designated as 1.

Statistical analysis

Throughout the text, results are expressed as mean ± SEM and analyzed by PRISM software (Graph-Pad, San Diego, CA). Comparisons involving two groups were analyzed using Student’s t-test. Statistical significance was established at P<0.05.

Results

Analysis of snRNAs sequencing in Leiomyoma and Matched Myometrium

Using high throughput RNA sequencing we have recently reported the expression profile of lncRNAs, miRNAs (a member of sncRNAs) and mRNAs in leiomyoma and paired myometrium (4). Using the existing sequencing data set we remapped and aligned our data with sncRNA sequenced database (miRbase, piRBase, Genomic tRNA database, and ENSEMBL annotations) which contains the curated sncRNAs, with specific emphasis placed on snoRNAs, snRNAs, piRNAs, tRNAs and rRNAs. Screening of 437 snoRNAs, 1828 snRNAs, 23439 piRNAs, 610 tRNAs and 522 rRNAs in the database resulted in identification of 31 snRNAs, 284 snoRNAs, 180 piRNAs, 152 tRNAs and 63 rRNAs in leiomyoma and paired myometrium (Supplementary Table 2). Despite considerable variation, further analysis after normalization by DESeq2 (26) revealed that 15 snoRNAs, 24 piRNAs, 7 tRNAs, and 6 rRNAs were differentially expressed at 1.5 fold change cutoff with no significant difference in snRNA expression in leiomyoma as compared to myometrium (Table 1).

Table 1.

Summarized Results of RNA Sequencing Data in leiomyoma as compared to matched myometrium (N=3; ≥ 1.5 or ≤ 0.75 fold change) and the background information of the differential expressed snoRNAs, piRNAs, tRNAs and rRNAs.

	Transcripts Total	Expressed in all Paired tissues	Differentially Expressed	Gene symbol	Gene description	Aliases	Regulation	Fold Change	p Value	Chromosome Location
piRNA	180	85	24	hsa_piR_000753	Piwi-interacting RNA 753	piR-31052	Down	0.5263	0.0177	chr17
				hsa_piR_000794	Piwi-interacting RNA 794	piR-31104	Down	0.5347	0.0136	chr6
				hsa_piR_001311	Piwi-interacting RNA 1311	piR-31924	Down	0.6059	0.0194	chr1 and 3
				hsa_piR_004153	Piwi-interacting RNA 4153	piR-43772	Down	0.5468	0.0009	chr3
				hsa_piR_004307	Piwi-interacting RNA 4307	piR-43993	Down	0.6105	0.1194	chr3, 5 and 6
				hsa_piR_006426	Piwi-interacting RNA 6426	piR-46895	Down	0.6222	0.0626	chr17
				hsa_piR_009228	Piwi-interacting RNA 9228	piR-32608	Down	0.6283	0.1112	chr17
				hsa_piR_009294	Piwi-interacting RNA 9294	piR-32678	Down	0.6953	0.1063	chrY
				hsa_piR_009295	Piwi-interacting RNA 9295	piR-32679	Down	0.6867	0.2222	chr17
				hsa_piR_015026	Piwi-interacting RNA 15026	piR-57660	Down	0.5742	0.0610	chr1, 3 and 4
				hsa_piR_016677	Piwi-interacting RNA 16677	piR-33065	Down	0.6695	0.0412	chr2
				hsa_piR_016742	Piwi-interacting RNA 16742	piR-33161	Down	0.4766	0.0123	chr1, 5 and 16
				hsa_piR_016945	Piwi-interacting RNA 16945	piR-33468	Down	0.5214	0.0031	chr1, 5, 6 and 16
				hsa_piR_017723	Piwi-interacting RNA 17723	piR-60576	Down	0.6963	0.2539	chr2, 3, 5 and 6
				hsa_piR_017724	Piwi-interacting RNA 17724	piR-60577	Down	0.6487	0.1722	chr1, 5, 6 and 11
				hsa_piR_018569	Piwi-interacting RNA 18569	piR-61645	Down	0.7030	0.2534	chr1, 3 and 4
				hsa_piR_018570	Piwi-interacting RNA 18570	piR-61648	Down	0.6259	0.0271	chr1, 2, 5, 6, 16 and 17
				hsa_piR_019825	Piwi-interacting RNA 19825	piR-35284	Down	0.4448	0.0023	chr1
				hsa_piR_019914	Piwi-interacting RNA 19914	piR-35413	Down	0.5331	0.0102	chr1, 5, 6 and 16
				hsa_piR_020326	Piwi-interacting RNA 20326	piR-35982	Down	0.5240	0.0319	chr1, 3, 6 and 16
				hsa_piR_020365	Piwi-interacting RNA 20365	piR-36041	Down	0.6139	0.0467	chr11
				hsa_piR_000765	Piwi-interacting RNA 765	piR-31068	Down	0.7164	0.2774	chr1, 2, 5 and 16
				hsa_piR_001312	Piwi-interacting RNA 1312	piR-31925	Down	0.7116	0.1113	chr1, 2, 5, 6, 16 and 17
				hsa_piR_017716	Piwi-interacting RNA 17716	piR-60565	Down	0.7487	0.3280	chr1, 3 and 6
tRNA	152	72	7	tRNA-Glu-CTC-1-3	Transfer RNA-Glu (CTC) 1-3	TRNAE22	Down	0.6755	0.2088	chr1
				tRNA-Glu-CTC-1-5	Transfer RNA-Glu (CTC) 1-5	TRNAE5	Down	0.6395	0.1482	chr1
				tRNA-Gly-GCC-5-1	Transfer RNA-Gly (GCC) 5-1	TRNAG20	Down	0.5275	0.0059	chr16
				tRNA-Lys-CTT-4-1	Transfer RNA-Lys (CTT) 4-1	TRNAK9	Down	0.6500	0.1105	chr16
				tRNA-Glu-CTC-1-2	Transfer RNA-Glu (CTC) 1-2	TRNAE13	Down	0.7134	0.2638	chr1
				tRNA-Glu-CTC-1-7	Transfer RNA-Glu (CTC) 1-7	TRNAE19	Down	0.7554	0.3511	chr6
				tRNA-His-GTG-1-9	Transfer RNA-His (GTG) 1-6	TRNAH5; TRH-GTG1-9	Down	0.7128	0.2502	chr15
rRNA	63	29	6	RNA5SP202	RNA, 5S ribosomal pseudogene 202	RN5S202	Down	0.5882	0.0259	chr6
				RNA5SP150	RNA, 5S ribosomal pseudogene 150	RN5S150	Down	0.6798	0.1125	chr3
				RNA5SP355	RNA, 5S ribosomal pseudogene 355	RN5S355	Down	0.6894	0.2412	chr12
				RNA5SP199	RNA, 5S ribosomal pseudogene 199	RN5S199	Down	0.7243	0.2294	chr5
				RNA5SP103	RNA, 5S ribosomal pseudogene 103	RN5S103	Down	0.7558	0.2727	chr2
				RNA5SP450	RNA, 5S ribosomal pseudogene 450	RN5S450	Down	0.7270	0.3132	chr18
snoRNA	284	243	15	SNORD36A	Small nucleolar RNA, C/D box 36A	RNU36A; U36a	Up	1.5998	0.0228	chr9
				SNORA68	Small nucleolar RNA, H/ACA box 68	RNU68A; U68	Up	1.5299	0.0928	chr19
				SNORD113	Small nucleolar RNA, C/D box 113	14q(I)	Up	1.7936	0.0201	chr14
				SNORD56	Small nucleolar RNA, C/D box 56	RNU56; U54	Up	1.5767	0.0272	chr20
				SNORD54	Small nucleolar RNA, C/D box 54	RNU54; U54	Up	1.6456	0.0626	chr8
				SNORD52	Small nucleolar RNA, C/D box 52	RNU52; U52	Up	1.5259	0.0202	chr6
				SNORA77	Small nucleolar RNA, H/ACA box 77	ACA63; SNORA77A	Up	1.5513	0.1650	chr1
				SNORD114-17	Small nucleolar RNA, C/D box 114-17	14q(II-17)	Up	1.9226	0.0101	chr14
				SNORA16A	Small nucleolar RNA, H/ACA box 16A	ACA16	Up	1.6404	0.0310	chr1
				SNORD96A	Small nucleolar RNA, C/D box 96A	U96A	Up	1.5495	0.0210	chr5
				SNORD73A	Small nucleolar RNA, C/D box 73A	RNU73A; U73a	Up	1.5996	0.0377	chr4
				SNORD94	Small nucleolar RNA, C/D box 94	U94	Up	1.4956	0.0495	chr2
				SNORD49B	Small nucleolar RNA, C/D box 49B	U49B	Up	1.6227	0.0227	chr17
				SNORD16	Small nucleolar RNA, C/D box 16	U16	Up	1.6850	0.0095	chr15
				SNORD113-5	Small nucleolar RNA, C/D box 113-5	14q(I-5)	Up	1.8938	0.0057	chr14
snRNA	31	15	0

QRT-PCR Expression of sncRNAs in Leiomyoma and Matched Myometrium

To provide support and validate RNA sequencing data analysis we selected a number of differentially expressed snoRNAs (SNORD30, SNORD27, SNORA16A, SNORD46 and SNORD56), piRNAs (piR-1311, piR-16677, piR-20365, piR-4153), tRNA (TRG-GCC5-1) and rRNA (RNA5SP202) and confirmed their expression by QRT-PCR in leiomyoma and paired myometrium (N=20) including the same tissues used in RNA sequencing. The selection of these sncRNAs was based on their differential expression obtained from RNA sequencing in leiomyoma as compared to myometrium (N=3) as well as prior reports demonstrating their altered expression in various disorders, including tumorigenesis and tumor progression (34–40). QRT-PCR analysis demonstrated that all the above snoRNAs, piRNAs, tRNA, and rRNA are expressed in leiomyomas and paired myometrium with up-regulated expression of snoRNAs and down-regulated expression of piRNAs, tRNA, rRNA in leiomyomas as compared to matched myometrium (P<0.05; Fig. 1A–C). Expression of these sncRNAs as determined by QRT-PCR displayed a similar pattern as RNA sequencing analysis. Further analysis of QRT-PCR results was also carried out in paired tissues based on the phase of the menstrual cycle. This analysis indicated that there were no statistically significant differences in expression of selected piRNAs, tRNA, rRNA and snoRNAs between follicular (N=10) and luteal (N=10) phase, with the exception of SNORD30 which was down-regulated in myometrium from luteal phase as compared to follicular phase (P<0.05; Fig. 2A–C).

Figure 1.

The expression of selected (A) snoRNAs (SNORD30, SNORD27, SNORA16A, SNORD46 and SNORD56), (B) piRNA (piR-1311, piR-16677, piR-20365 and piR-4153) and (C) tRNA (TRG-GCC5-1) and rRNA (RNA5SP202) in twenty paired leiomyoma tissues by QRT-PCR. The results are presented as mean ± SEM with P values (*P<.05) indicated by corresponding lines.

Figure 2.

The expression of selected (A) snoRNAs (SNORD30, SNORD27, SNORA16A, SNORD46 and SNORD56), (B) piRNA (piR-1311, piR-16677, piR-20365 and piR-4153) and (C) tRNA (TRG-GCC5-1) and rRNA (RNA5SP202) in twenty paired leiomyoma tissues separated by follicular (N=10) or luteal (N=10) phase of menstrual cycle by QRT-PCR. The results are presented as mean ± SEM with P values (*P<.05) indicated by corresponding lines. #: P<.05 as compared to Myo-Follicular group.

AGO2 expression in Leiomyoma and Matched Myometrium

Since AGO2 plays an essential role in sncRNAs mediated gene silencing (41), we assessed the expression of AGO2 in paired leiomyoma and myometrium (N=20) using western blot analysis. The results indicated that AGO2 was expressed in greater abundance in leiomyomas as compared to myometrium (P<0.05; Fig. 3A–B) with no significant difference in AGO2 expression between follicular as compared to luteal phase of the menstrual cycle (Fig. 3C).

Figure 3.

Representative protein levels of AGO2 (A) in twenty paired leiomyoma tissues. The mean expression of AGO2 is shown in (B) (N=20) and separated according to menstrual cycle phase: Follicular (F, N=10) and Luteal (L, N=10) phase is shown in (C). The results are presented as mean ± SEM. *P<.05.

Discussion

Through high-throughput RNA sequencing we have recently reported the expression profile of a large number of lncRNAs with concurrent expression of miRNAs and protein-coding RNAs profiles in leiomyoma and paired myometrium (4). Using the existing RNA sequencing data in the present study through annotation with AASRA pipeline and miRbase, piRBase, Genomic tRNA database, and ENSEMBL annotations, which provides comprehensive up-to-date database information on human sncRNAs (437 snoRNAs, 1828 snRNAs, 23439 piRNAs, 610 tRNAs and 522 rRNAs), enabled us to further characterize the expression profile of other classes of sncRNAs (snRNAs, snoRNAs, piRNAs, tRNAs and rRNAs) in leiomyoma and paired myometrium. The annotation analysis resulted in identification of 284 snoRNAs, 31 snRNAs, 180 piRNAs, 152 tRNAs and 63 rRNAs transcripts in leiomyoma and myometrium of which 15 snoRNAs, 24 piRNAs, 7 tRNAs, and 6 rRNAs were differentially expressed in leiomyomas, and confirmed by QRT-PCR. Although we expected to detect a substantially larger number of sncRNAs transcripts in these paired tissues, we consider possible optimal library preparation utilized in our initial study and/or degradation or chemical modification of transcripts may have accounted for detection of fewer numbers of sncRNAs. Despite this limitation to the best of our knowledge these results are the first to identify differential expression of these sncRNAs in leiomyoma.

Leiomyomas used in our RNA sequencing study were from the luteal phase while the QRT-PCR analysis included paired tissues from both phases of the menstrual cycle in leiomyomas that were 3–5 cm in diameters in size. Under these conditions, QRT-PCR analysis displayed similar pattern of expression as RNA sequencing data. Several studies have previously reported the influence of menstrual cycle phase, tumor size and race on expression of mRNAs, lncRNAs as well as miRNAs, a member of sncRNA family (5, 42). Additional studies are warranted to determine the influence of race, tumor size and menstrual cycle phase on expression profile of sncRNAs in leiomyomas and the influence of ovarian steroids in isolated leiomyoma smooth muscle cells.

Currently with the exception of a number of studies that addressed the regulatory function of few miRNAs (6, 11–15), the regulatory function of other classes of sncRNAs in leiomyoma remains unknown. Accumulated evidence suggests that snoRNAs play key regulatory functions in various normal cellular processes and their altered expression has been associated with a wide range of disorders, including chronic lymphocytic leukemia, hepatocellular carcinoma and colorectal, prostate, endometrial and lung cancers. As such snoRNAs might be useful as diagnostic and prognostic biomarkers in colorectal and lung cancers (43–48). In addition, snoRNAs have a housekeeping role and are involved in biogenesis and modification of rRNAs, tRNAs and snRNAs (49). Other functions of snoRNAs include alternative splicing of the trans gene transcript as exemplified by SNORD115 (50), and may function as miRNAs after further processing into 20–24 nucleotides long small RNAs (51). Interestingly, there is overexpression of SNORD30 and SNORD27 in smoldering myeloma and this increased expression correlates with a higher incidence of tumor progression (34). SNORD27 modifies rRNA methylation by targeting fibrillarin, a methyltransferase, and through direct RNA-RNA interaction regulates alternative pre-mRNA splicing of several genes, including E2F7, MAP4K3, ZBTB37, FER, and ABCA8 (35). Additionally, SNORD27 has been shown to associate with ribosomes from the early pre-ribosomes to the later maturation stage, whereas SNORD56 is only recruited at the stage of later pre-ribosomal complexes, mediated by DDX21 RNA helicase (36). Moreover, SNORD46 which is embedded within the intronic regions of RPS8 was down-regulated in breast tumor (37). Furthermore, elevated expression of SNORA16A correlated with a higher risk of relapse in squamous cell carcinomas (38). Our results demonstrated that the expression levels of SNORD30, SNORD27 and SNORA16A were all increased in leiomyomas as in other tumors (34, 38); however, leiomyomas expressed elevated levels of SNORD46 in contrast to repressed SNORD46 expression in breast tumors (37), indicating the tumor-specificity of this snoRNA expression. The function and the regulatory mechanisms of differentially expressed snoRNAs in leiomyoma progression remain to be elucidated.

piRNAs are derived from long RNA precursors through a dicer-independent mechanism, and are also transcribed from the 3′ untranslated regions of intergenic non-protein coding and protein-coding genes (52). Although piRNAs transcripts were initially identified in gonadal tissues they are also expressed in somatic tissues (53), circulating or cancer stem cells (54). Functionally piRNAs through interaction with PIWI proteins regulate expression of genes through translational control, silencing transposons or epigenetically through control of telomeres via particular methyltransferases or histone modifications (55–58). Aberrant PIWI protein expression has been reported in various cancers including breast, esophageal, pancreas, gastric and endometrial carcinoma (59–67). Several studies reported elevated expression of piR-20365 in breast tumors and lymph node metastasis (39), while no differences in piR-4153 expression was found in synovial fibroblasts from rheumatoid arthritis and osteoarthritis (40). Despite the existence of a large number of piRNAs in the database our analysis only detected 180 piRNAs of which the expression of 24 piRNAs including piR-1311, piR-16677, piR-20365 and piR-4153 were significantly decreased in leiomyoma as compared to myometrium. The role of these piRNAs as potential silencers of protein coding genes in leiomyomas remains to be determined.

Transfer RNAs (tRNAs) are derived from pre-tRNA by endonuclease RNase Z and in their mature form play a well-defined role in protein translation (18). However, tRNAs derived small RNAs (tsRNAs) or tRNA fragments (tRFs) of 30–35 nucleotides long are also generated through cleavage of mature tRNAs at both 5′ and 3′ end at the anticodon loop during the maturation process of tRNAs (18). tRFs similar to piRNAs have been shown to associate with PIWIL2, a protein involved in the silencing of transposons (68). Altered expression of tRFs has been reported in several tumors including colon, breast and ovarian cancers upon oncogene activation and during cancer staging (68). Under in vitro condition, knock-out of tRFs-101 and tRFs-46 in HEK-293 cells resulted in significant alteration of gene-expression patterns, with activation of genes involved in cell survival with concurrent down-regulation of genes involved in apoptosis and chromatin structure. Overexpression of tRFs-46 and tRFs-47 in lung cancer cell lines resulted in inhibition of colony formation and cell proliferation, implying tsRNAs play a key role in cell growth and survival (68). Although our sequencing resulted in identification of 152 tRFs transcripts only 7 tRFs were differentially expressed in leiomyoma. It remains to be determined if these tRFs have a role in leiomyoma cell proliferation and or apoptosis. Our results also demonstrated that leiomyoma expressed significantly more AGO2 protein as compared to matched myometrium, possibly suggesting greater levels of post-transcriptional RNA processing in leiomyomas. However, the impact of altered AGO2 expression in leiomyoma tumorigenesis and progression remain further elucidation.

A dynamic regulatory interaction among the noncoding (sncRNA and lncRNAs) and protein-coding genes has emerged to influence the outcome of various normal cellular activities while their aberrant expression has been associated with a wide range of disorders. In the present study and our previous report (4) through next generation RNA sequencing we provided evidence that leiomyoma also display an altered expression of a number of sncRNAs and lncRNAs as compared to myometrium. Although the result presented here provided the necessary initial step in sncRNAs identification in leiomyoma, further studies are required to address their expression, regulation and molecular mechanism of actions in leiomyoma pathogenesis and progression.

Supplementary Material

1

Supplementary Table 1: Primers sequences used for QRT-PCR.

Supplementary Table 2: Excel spread sheets showing the raw data of expression profile of all sncRNAs (snoRNAs, snRNAs, piRNAs, tRNAs and rRNAs) in leiomyoma and paired myometrium.

Capsule.

The expression profile of sncRNAs was determined by next generation sequencing and confirmed by QRT-PCR, indicating differential expression of selected snoRNAs, piRNAs, tRNA and rRNA in leiomyoma.