Differential expression of microRNAs in myometrium and leiomyomas and regulation by ovarian steroids.

Given the emerging roles of microRNAs (miRNAs) as key regulator of mRNA stability we assessed their expression profile in paired myometrium and leiomyoma, their isolated smooth muscle cells (MSMC and LSMC), a spontaneously transformed leiomyoma smooth muscle cells (T-LSMC) and SK-LMS-1, a leiomyosarcoma cell line using microarray and real time PCR. Based on global normalization of expression values of 385 miRNAs and statistical analysis (anova), 91 miRNAs were expressed above the threshold levels in myometrium, with a progressive decline in numbers in leiomyomas, MSMC, LSMC, T-LSMC and SK-LMS-1 (P<0.05). We selected and validated the expression of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142-5p and found their differential expression in tissue and cell-specific manners (P<0.05). Treatments of MSMC and LSMC with 17beta estradiol and medroxyprogesterone acetate (10(-8)M), or ICI-182780 and RU-486 (10(-6)M) resulted in differential regulation of these miRNAs (P<0.05). In conclusion, the expression of a number of miRNAs in myometrium and leiomyoma with their progressive aberrant from normal MSMC into LSMC, transformed and cancerous stage, suggests that miRNAs and their regulation by ovarian steroids play a key role in pathogenesis of leiomyoma through gene expression stability.

Introduction

Leiomyomas are benign uterine tumours that develop during the reproductive years and suppressed following menopause. Clinical observations and epidemiological studies have also estimated that a lifelong risk of developing leiomyomas is about 70%, with a higher risk among African-Americans compared to other ethnic groups [1]. Despite limited potential of becoming malignant, the presence of symptomatic leiomyomas account for over a one-third of all the hysterectomies performed annually in the United States alone [2].

Leiomyomas are considered to develop from cellular transformation of myometrial cells, however, the identity of the factor(s) and their mechanism of actions contributing toward this process is unknown. Genomic instability involving several genes, including estrogen and progesterone receptors, has also been associated with an increased risk of leiomyoma development, however in a majority of cases the evidence has been inconsistent [3-5].Conventional and large-scale gene expression studies have provided further evidence reflecting the molecular environments of leiomyoma and myometrium, and their possible involvement in transformation of myometrial cells into leiomyomas [6-15].However, the regulation and biological significance of many of these genes in the pathophysiology of leiomyoma remains to be established. Given that gene expression stability is regulated at multiple levels, recent studies have implicated microRNAs (miRNAs), a group of small non-coding RNA, as a key regulator of this process through mRNA degradation and repression [16-21] miRNAs are expressed as 70–90 base pair (bp) precursors that are processed and further cleaved forming a 17–24 bp mature miRNAs [16-21]. The mature miRNAs complex with their target gene sequences in a complementary manner, results in degradation or regression of gene expression [22].To date, several hundred miRNAs have been cloned and/or predicted, each with the ability to modulate the expression of many mRNA, although their target sequences could be only partially complementary to these miRNAs [22].The expression of many miRNAs has been identified in a number of cells and tissues and their altered and/or aberrant expression associated with various pathological disorders [23-25]. As such elucidating the biological function of miRNAs in normal and pathological conditions is a subject of intense investigations and the existing evidence suggests their importance in several developmental and tumourogenesis processes [26] such as cell growth, differentiation, apoptosis and malignant transformation [17, 19, 20]. In addition, the genes encoding a number of miRNAs have been located at chromosomal fragile sites and regions of cytogenetic abnormalities associated with cancer, and their altered expression has been associated with the stability of oncogenes involved in cellular transformation [16–21, 27].

Given the emerging roles of miRNAs in the above processes we hypothesized that the expression profile of miRNAs in leiomyomas differs from normal myometrium, their isolated smooth muscle cells and their transformation into cancerous phenotype.To test this hypothesis we profiled the expression of 385 miRNAs in paired leiomyomas and myometrium, their isolated smooth muscle cells, a spontaneously transformed leiomyoma smooth muscle cell and SK-LMS-1, a leiomyosarcoma cell line.We selected and validated the expression of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142–5p, based on their predicted target genes, which include TGF-βR2, ERα, ERβ and PR, respectively, in these tissues and cells using real time PCR. We also examined the influence of 17β estradiol and medroxyprogesterone acetate, as well as ICI-182780 and RU-486 a pure antiestrogen and anti progesterone, respectively on the expression of these miRNAs in MSMC and LSMC.

Materials and methods

All the materials for isolation of myometrial (MSMC) and leiomyoma (LSMC) smooth muscle cells and culture media were purchased from commercial sources as previously described [7]. MirVana and miRNA Bioarrays were purchased from Ambion (Austin, TX) and Applied Biosystem (Foster City, CA), respectively. 17β estradiol (E2) medroxyprogesterone acetate (MPA) and RU486 (Mifepristone) were purchased from Sigma Aldrich (St. Louis, MO). ICI-182780 (Fulvestrant) was purchased from Tocris Cookson, Inc. (Ballwin, MO) and charcoal-stripped foetal calf serum was purchased from Hyclone (Logan, UT). Leiomyosarcoma cell line, SK-LMS-1 was purchased from American Type Culture Collection (ATCC;Manassas, VA).

Tissue collection

Portions of leiomyoma and matched myometrium were collected from pre-menopausal women (n= 7) who were scheduled to undergo hysterectomy for indications related to symptomatic leiomyomas. Four of the patients were Caucasians and three were African-Americans at ages ranging from 29 to 38 years. These women were not taking any medications, including hormonal therapy for the pervious 3 months prior to surgery and based on their last menstrual period and endometrial histology they were from early-mid secretory phase of the menstrual cycle. All the leiomyomas used in this study were 3–5 cm in diameter and were collected at the University of Florida affiliated Shands Hospital with prior approval from the Institutional Review Board.Immediately after collection, portions of the tissues were snapped frozen and kept in liquid nitrogen for further analysis, or used for smooth muscle cell isolation and culturing.

Leiomyoma and myometrial smooth muscle cell isolation and culture

A small portion of paired myometrium and leiomyoma was used for isolation of MSMC and LSMC as previously described [7].The isolated cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing antimycotic, antibiotics and 10% foetal bovine serum (FBS) and incubated at 37°C in a humidified 5% CO2 incubator until reaching confluence. Prior to use, the cell cultures were characterized using antibodies to α smooth muscle actin, desmin and vimentin based on immunofluoresence microscopy.

During the course of this study one of the LSMC originally isolated and cultured from leiomyoma of a 35-year-old woman undergoing hysterectomy for symptomatic leiomyoma was spontaneously transformed after the fifth passage. This cell culture was morphologically homogenous and at initial isolation and after first passage displayed 100% immunostaining for α smooth muscle actin and desmin. However, some of the cells adopted a different morphological appearance and progressively increased in numbers, and following two additional passages established into a uniform cell population with a rapid doubling time.We referred to this cell culture as transformed LSMC or T-LSMC. The T-LSMC were cultured in DMEM supplemented with 10% FBS, however they have not been further characterized. SK-LMS-1 was cultured in DMEM supplemented with 10% FBS.

miRNA isolation

Total RNA was isolated from myometrium, leiomyomas as well as LSMC, MSMC, T-LSMC and SK-LMS-1 using mirVana miRNA isolation kit according to manufacturer's instructions (Ambion).For isolation of miRNA fractions total RNAs were fractionated and cleaned up with flashPAGE Fractionator and reagents (Ambion). Briefly, 25 μg of total RNA was loaded onto the top of a column filled with a denaturing acrylamide gel matrix and the miRNA fraction was obtained by mixing RNA with 2X sample buffer and flashPAGE purification using flashPAGE precast gels and the flashPAGE fractionator system. A dye was loaded with the total RNA to track RNAs that are ∼40 nt in size. The quality, yield and size of miRNA fractions and total RNA were analysed using Agilent 2100 Bioanalyzer (Agilent Technologies, Foster City, CA).

miRNA labelling and hybridization

Purified miRNA isolated from the above cells and tissues were labelled at 3′ amine-modified tails using the mirVana miRNA array labelling kit (Ambion) and fluorescently coupled with Cy5 post-labelling reactive dye (Amersham, GE Healthcare Bio-Sciences Corp. Piscataway, NJ). Chemically synthesized oligoribonucleotides were used as positive control (Ambion) and labelled along with purified miRNAs. A 3x miRNA hybridization buffer (Ambion) was added to the fluo-rescently labelled miRNAs and the solution was heated at 95°C for 2 min. The Cy5 labelled samples were cleaned using post-labelling miRNA clean up kit, eluted, and stored at –70°C or analysed by hybridizing to miRNA arrays. The mirVana miRNA Bioarray slides (Ambion) were placed into hybridization chambers (Corning Incorporated Life Sciences, Acton, MA) and 10 _l of miRNA added to 20 _l of hybridization buffer mixture was applied under the Bioarray LifeSlip onto the array slide. The chambers were sealed and incubated at 42°C water bath for 12–16 hrs. Each probe on the Bioarray slide is printed in duplicate with 20 positive and 100 negative controls.

Array data processing and analysis

Following hybridization, the slides were washed, dried and scanned on a GenePix 4000B Array Scanner (Molecular Devices Corporation, Sunnyvale, CA). The miRNA spots and their intensity were determined using GenePix Pro 6.0 software as recommended by the manufacturer. The background-adjusted intensity was normalized for each miRNA using a global variance stabilization normalization procedure recommended by Ambion for data analysis [28, 29]. The analysis identified differentially expressed miRNAs with precision and quadratic relationship between the variance with lesser focus on absolute expression and fold-change difference [28].The results were subjected to unsupervised hierarchical clustering and Tree View analysis.

Treatments

To determine whether ovarian steroids regulate the expression of miRNAs in MSMC and LSMC 106 cells/well were seeded in 6-well culture dishes and incubated for 48 hrs. The cells were made quiescent under serum-free condition for 24 hrs, and then treated with E2 (10−8M), MPA (10−8 M), ICI-182780 (10−6M), RU486 (10−6 M), E2+ICI or MPA+RU486 for 6, 12 and 24 hrs added to phenol red-free medium containing 2% charcoal-stripped FBS as previously described [9].

Real time polymerase chain reaction

Total RNA isolated from the above tissues and cells was used to confirm the expression of mature miRNA identified by microarray using TaqMan® MicroRNA Assay for human mature miRNA (Applied Biosystems) listed in the Sanger miRBase database. Briefly, 10 ng of total RNA was reverse transcribed to cDNA with stem-loop primers for miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142–5p and Taqman microRNA reverse transcription kit. Quantitative real time PCR was carried out using an Applied Biosystems 7300 real time PCR System and a Taqman universal PCR master mix at 95°C for 10 min., 95°C 15 sec. and 60°C for 1 min. for 40 cycles. The results were analysed using the comparative method following normalization of expression values to U6 expression as recommended using Sequence Detection Software 2.2.1 (Applied Biosystems).

Statistical analysis

All the in vitro experiments were performed at least three times in duplicate using independent cell cultures. Where appropriate the results are expressed as mean ± standard error (S.E.M.) and statistically analysed using Student's t test for comparison of two groups and anova for multiple comparisons, with P<0.05 considered significant.

Results

Using microarray platform consisting of 385 miRNA probes (213 human miRNAs) we first assessed their expression profiles in paired myometrium and leiomyomas, myometrial and leiomyoma smooth muscle cells (MSMC and LSMC) isolated from these tissues, as well as in spontaneously transformed leiomyoma cells (T-LSMC) and SK-LMS-1. The observational assessment indicated that many of the miRNAs are expressed in these tissues and cells with a progressive reduction in their numbers in leiomyomas, LSMC, T-LSMC and SK-LMS-1 as compared to myometrium and MSMC (Figs not shown). Following global normalization, the mean expression value of each cohort was subjected to unsupervised hierarchical clustering and Tree-View analysis (Fig. 1A and B).These figures illustrate that these differentially expressed miRNAs in paired myometrium (MY1, MY2, MY3 and MY4) and leiomyomas (LY1, LY2, LY3 and LY4) all from Caucasians subjects, and LSMC, T-LSMC and SK-LMS-1 are separated into their respective subgroup some with overlapping relatedness. Of these miRNAs, the level of expression of 91 miRNAs was above the threshold levels set during the analysis in myometrium, of which 68 were similarly identified in leiomyomas (P<0.05;Table 1).The analysis also identified a substantially lower number of miRNAs expressed in MSMC, LSMC, T-LSMC and SK-LMS-1 as compared to myometrium and leiomyomas, with 48 miRNAs expressed above threshold levels in MSMC (P<0.05). The list of miRNAs expressed in these cells is presented in Table 2.

1

Hierarchical cluster and tree-view analysis of differentially expressed miRNAs profile in paired myometrium (MY1, MY2, MY3 and MY4) and leiomyomas (LY1, LY2, LY3 and LY4) (Fig. 1A) and in paired myometrial smooth muscle cells (MSMC1 and MSMC2) and leiomyoma smooth muscle cells (LSMC1 and LSMC2) isolated and cultured from the above tissues, transformed leiomyoma cells (TLSMC-1 and T-LSMC-2) and leiomyosarcoma cell line (SLKMS1 and SKLMS2). Each column represents data from a single cohort with shades of red and green indicating up- or down-regulated miRNA according to the colour scheme shown below.miRNAs represented by rows were clustered according to their similarities in pattern of expression in each tissue. The dendrogram at the top of the image displays similarity of expression among these cohorts (see Table 1 and Table 2 for the list of differentially expressed miRNAs).

1

The list of miRNAs identified in paired myometrium and leiomyomas

Name	Myometrium (Mean)	Leiomyoma (Mean)	P-values
hsa_let_7a	7.83	6.40	0.003
hsa_let_7b	8.89	6.90	0.001
hsa_let_7c	8.95	6.95	0.0001
hsa_let_7d	6.63	4.73	0.001
hsa_let_7e	6.64	3.78	0.009
hsa_let_7i	5.46	4.36	0.018
hsa_miR_100	6.81	5.03	0.0002
hsa_miR_103	6.82	4.94	0.0003
hsa_miR_106b	4.59	3.64	0.0106
hsa_miR_107	6.59	5.02	0.0009
hsa_miR_10b	7.15	6.70	0.00025
hsa_miR_125a	9.08	7.02	0.00005
hsa_miR_125b	9.31	8.23	0.0015
hsa_miR_126	6.09	6.49	0.004
hsa_miR_132	4.72	2.77	0.0025
hsa_miR_133a	7.56	5.39	0.0006
hsa_miR_133b	7.27	4.62	0.001
hsa_miR_140	2.61	3.46	0.019
hsa_miR_143	8.79	7.54	0.007
hsa_miR_145	10.26	9.36	0.0005
hsa_miR_146b	5.65	2.88	0.008
hsa_miR_149	3.70	2.04	0.03
hsa_miR_150	2.48	3.63	0.007
hsa_miR_151	4.79	3.25	0.01
hsa_miR_152	5.21	3.44	0.006
hsa_miR_15a	4.69	3.72	0.014
hsa_miR_15b	5.45	3.30	0.03
hsa_miR_16	7.57	6.74	0.003
hsa_miR_181a	5.26	3.37	0.002
hsa_miR_181b	4.53	2.64	0.01
hsa_miR_191	6.52	5.32	0.005
hsa_miR_193b	6.87	3.65	0.002
hsa_miR_195	2.75	7.10	0.0003
hsa_miR_197	4.65	2.98	0.004
hsa_miR_199a	6.33	3.74	0.00003
hsa_miR_199a_AS	6.27	5.60	0.002
hsa_miR_20a	2.98	3.10	0.06
hsa_miR_21	7.53	7.31	0.02
hsa_miR_214	7.39	5.63	0.00002
hsa_miR_22	6.56	4.00	0.01
hsa_miR_221	6.62	6.17	0.002
hsa_miR_222	6.83	5.69	0.004
hsa_miR_224	3.49	2.37	0.045
hsa_miR_23a	8.17	6.63	0.002
hsa_miR_23b	9.09	7.22	0.002
hsa_miR_24	8.32	7.57	0.003
hsa_miR_25	4.99	3.51	0.02
hsa_miR_26a	8.50	7.50	0.001
hsa_miR_27a	5.81	4.47	0.05
hsa_miR_27b	7.53	6.50	0.003
hsa_miR_28	4.76	3.16	0.002
hsa_miR_29c	2.24	3.93	0.0004
hsa_miR_30a_5p	6.64	5.70	0.0009
hsa_miR_30c	6.76	4.98	0.001
hsa_miR_30d	7.28	5.75	0.003
hsa_miR_30e_5p	5.35	4.26	0.04
hsa_miR_320	7.51	5.68	0.002
hsa_miR_324_3p	5.50	3.07	0.0005
hsa_miR_331	5.46	3.17	0.0008
hsa_miR_339	4.78	3.10	0.01
hsa_miR_342	7.29	5.36	0.0002
hsa_miR_345	3.47	2.15	0.03
hsa_miR_34a	6.67	3.28	0.00004
hsa_miR_361	5.72	2.93	0.006
hsa_miR_378	4.14	2.54	0.004
hsa_miR_422a	4.81	2.57	0.01
hsa_miR_422b	6.33	3.64	0.0008
hsa_miR_423	5.92	4.60	0.008
hsa_miR_432	3.74	2.49	0.01
hsa_miR_487b	3.65	2.23	0.03
hsa_miR_494	5.56	6.10	0.006
hsa_miR_497	2.39	3.94	0.005
hsa_miR_511	4.85	2.15	0.009
hsa_miR_92	5.92	4.76	0.02
hsa_miR_93	5.22	3.11	0.007
hsa_miR_99a	8.25	6.01	0.0001
hsa_miR_99b	7.05	4.90	0.005
hsa_miR_20a	2.98	3.10	0.05
hsa_miR_20b	2.39	2.64	0.05

The list of miRNAs identified in paired myometrium and leiomyomas with their mean expression values determined following global normalization and statistical analysis using paired student's t-test as described in the materials and methods.

2

The list of miRNAs identified in isolated myometrial and leiomyoma cells

Name	MSMC	LSMC	T-LSMC)	SK-LMS-1	P-value
hsa_let_7a	7.99	6.64	5.72	5.47	0.04
hsa_let_7b	7.86	6.23	5.46	4.53	0.009
hsa_let_7c	7.69	5.90	5.06	4.04	0.01
hsa_let_7i	6.17	5.37	4.60	4.77	0.002
hsa_miR_10b	5.69	5.07	4.13	3.57	0.02
hsa_miR_125b	6.68	6.14	4.11	5.76	0.03
hsa_miR_143	6.68	5.82	3.70	4.20	0.01
hsa_miR_145	7.75	6.00	3.09	3.46	0.003
hsa_miR_181a	6.06	4.39	3.24	4.15	0.003
hsa_miR_199a_AS	6.53	3.83	3.68	3.72	0.04
hsa_miR_20a	4.25	4.41	5.02	4.10	0.06
hsa_miR_21	9.34	7.74	7.30	7.58	0.06
hsa_miR_22	6.76	5.52	3.90	4.54	0.01
hsa_miR_222	6.39	5.05	4.55	3.45	0.0003
hsa_miR_23a	7.70	6.19	5.82	6.15	0.05
hsa_miR_26a	6.77	6.13	4.80	4.45	0.01
hsa_miR_27b	6.49	5.27	5.61	4.99	0.02
hsa_miR_29a	7.87	6.25	4.54	6.44	0.01
hsa_miR_30a_5p	5.14	4.02	5.55	4.34	0.05
hsa_miR_30e_5p	3.91	2.65	5.44	3.72	0.02

The list of miRNAs identified in paired isolated myometrial (MSMC) and leiomyoma (LSMC) smooth muscle cells, spontaneously transformed LSMC (t-LSMC) and leiomyosarcoma cell line (SKLMS-1) with their mean expression values determined following global normalization and statistical analysis using anova as described in the materials and methods.

Using real time PCR we validated the expression of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a, and miR-142–5p in three-paired leiomyoma and myometrium as well as their isolated MSMC and LSMC. As illustrated in Figure 2, the level of expression of these miRNAs is significantly different in myometrium as compared to leiomyomas and different in the tissues of Caucasians compared to African-Americans (P<0.05).The level of expression of these miRNAs was also different among MSMC, LSMC, T-LSMC and SKLM-S1 with LSMC expressing lower levels of miR-20a, miR-21 and miR-26a as compared to MSMC and T-LSMC and SKLM-S1 (Fig. 3).However, these cells expressed very low levels of miR-18a and miR-181a, while miR-142–5p and miR-206 were expressed below the CT threshold levels (Fig. not shown).

2

Bar graphs show real time PCR expression of miR-20a, miR-21 and miR-26a, miR-18a, miR-206, miR-181a and miR-142-5p in paired myometrium (Myo) and leiomyomas (LYOM) form African-Americans (AA) and Caucasians (C). The data is presented as relative expression following normalization and setting the expression of each miRNA independently in Caucasian myometrium arbitrarily as 1.Data represent mean ±standard error of three paired tissues from each ethnic group with asterisks *, ** and *** significantly different from C-Myo. P<0.05 was considered significant. Arrows indicates significant difference between the expression of these miRNAs in AA-Myo and AA-LYMO.

3

Bar graphs show real time PCR expression of miR-20a, miR-21 and miR-26a in myometrial and leiomyoma smooth muscle cells (MSMC and LSMC), spontaneously transformed LSMC (tLSMC) and leiomyosarcoma cell line, SKLM-1. The cells (1_106/well in 6-well plates) were cultured as described in materials and methods and their total RNA was isolated and subjected to real time PCR. The data is presented as relative expression following normalization and setting their expression level in MSMC arbitrarily as 1. Data represent mean ± standard error from three separate experiments, with asterisks ** and *** displaying significant difference with * with P<0.05 considered significant. Arrow indicates significant difference in the expression of these miRNAs between the cell types.

We next examined the effect of ovarian steroids on the expression of miR-20a, miR-21 and miR-26a in MSMC and LSMC, and found that E2 inhibited the expression of miR-21 and miR-26a in MSMC and LSMC respectively, while enhancing miR-26a expression in MSMC as compared to controls (Fig. 4; P<0.05). In contrast, ICI-182780 increased the expression of miR-20a and miR-21 in MSMC and LSMC, and miR-26a in MSMC, while inhibiting the expression of miR-26a in LSMC (Fig. 4; P<0.05).Co-treatment with E2+ICI-182780 resulted in similar changes in the expression of these miRNAs as seen with ICI-182780 alone, with the exception of miR-21 expression that returned to levels induced by E2 (Fig. 4). Unlike E2, MPA-treatment increased the expression of miR-21 and miR-26a in LSMC and MSMC respectively, and inhibited miR-26a in LSMC, without significantly effecting miR-20a expression as compared to untreated controls (Fig. 4; P<0.05). RU-486 increased the expression of miR-20a and miR-21 in MSMC and LSMC respectively, and inhibited miR-26a expression in LSMC (Fig. 4; P<0.05). Co-treatment with MPA+RU-486 resulted in a similar effect as seen with MPA-treated cells with the exception of miR-20a expression that was reduced as compared to other treatments (Fig. 4).There was no significant difference in the expression of these miRNAs following 6, 12 or 24 hrs of treatments and the results presented here are for the 24 hrs time course.

4

Bar graphs show the expression of miR-20a, miR-21 and miR-26a in myometrial and leiomyoma smooth muscle cells (MSMC and LSMC).The cells (1_106/well in 6-well plates) were cultured as described in materials and methods and following 24 hrs of treatment with 17b estradiol (E2), ICI-182780 (ICI), E2+ICI, medroxyprog-esterone acetate (MPA), RU-486 (RU) and MPA+RU, their total RNA was isolated and subjected to real time PCR. The data is presented as relative expression following normalization and setting their expression level in untreated MSMC arbitrarily as 1. Data represent mean ± standard error from three separate experiments. The asterisks * indicate statistical difference between the expression of these miRNAs in treated as compared to untreated controls (Ctrl), with arrows pointing out the difference in their expression between MSMC and LSMC. A probability level of P<0.05 was considered significant.

Discussion

In the present study we profiled the expression of a number of miRNAs in paired myometrium and leiomyomas, their isolated smooth muscle cells, a spontaneously transformed LSMC (t-LSMC) and SKLM, a leiomyosarcoma cell line. Of the 213 human miRNAs on the array the expression value of 91 miRNAs was detected above the threshold levels in the myometrium. There was a significant lower number of miRNAs expressed in leiomyomas and those commonly expressed with myometrium displayed an altered level of expression. The number of miRNAs expressed in isolated MSMC and LSMC was even lower as compared to their corresponding tissues. Consistent with the current view that cellular transformation is associated with altered expression of several miRNAs [17, 19, 20, 27], we also found a further reduction in the number of miRNAs expressed in t-LSMC and SKLMS-1, as compared to MSMC and LSMC. In addition to altered level of their expression, a decrease in the number of miRNAs has been reported in several normal cells and their benign and malignant cells and tissue counterparts [20, 24, 30–32].Consistent with these observations we found a progressive loss of miRNA expression from myometrium and MSMC into leiomyomas, LSMC, T-LSMC and SKLM-1, implying the biological significance of these miRNAs and their target genes in pathogenesis of leiomyoma and possibly leiomyosar-coma. While commencing our study, a recent microarray study reported the expression profile of miRNAs in paired leiomyomas and myometrium demonstrating some overlapping similarities with our observations [33].Of the 206 miRNAs the expression of 45 was altered in leiomyomas, which included let-7 family, miR-21, miR-23b, miR-29b and miR-197 [33].Additionally this study identified a significant difference in the expression of a selective number of miRNAs in leiomyomas based on the phases of the menstrual cycle, ethnicity and size of the tumours [33].We recognize the low number of tissues representing the secretory phase of the menstrual cycle as a limitation of our study;however, these and several other miRNAs were also identified among the differentially expressed miRNAs in our study.

Using real time PCR we validated the expression of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142–5p in paired myometrium and leiomyomas as well as MSMC, LSMC, tLSMC and SKLM-S1. Although miR-206 was not on the array, the level of expression of other miRNAs differed considerably as compared to their expression values detected by microarray analysis, which could be due to the detection of both precursor and mature miRNAs by microarray and the mature forms by real time PCR. Detection of a low level of miR-18a and miR-181a, and a non-detectable level of miR-206 and miR-142–5p by real time PCR and low levels by microarray imply that the genes targeted by these miRNAs are protected from degradation and/or regression in these cells, although other miRNAs can target the stability of these genes.

It is estimated that approximately 30% of genes are the potential target of miRNAs regulatory functions [26, 32].Since each miRNA is predicted to have a broad range of target mRNA, even an alteration in the expression of a single miRNA could have a significant impact on the outcome of diverse biological functions regulated by the product of these genes [17, 19, 20, 22, 34]. As such the absence or altered expression of these and other miRNAs could result in expression re-programming of many of their target genes in leiomyomas as compared to myometrium. Thus, it is necessary to identify and correlate the expression of some of these miRNAs and their target genes, to define their biological relevance in myometrium, pathogenesis of leiomyomas as well as during cellular transformation from MSMC into LSMC, t-LSMC an SKLM-S1. Altered expression of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142-5p in leiomyomas of African-Americans, which develop more frequently, grow rapidly and are more symptomatic as compared to other ethnic group, is also of interest. Such an ethnic association and correlation of miRNA expression with leiomyomas' size, stage of the menstrual cycle and age of the patients [33] imply further regulation of miRNAs expression in these tissues. Although our study involved leiomyomas of one size from the secretory phase of the menstrual cycle, the identification of similar miRNAs suggest that these miRNAs play a significant regulatory function in pathogenesis of leiomyomas. Considering the current view that miRNAs act as negative regulators of gene expression, their altered expression in leiomyomas of this sub-population may further enhance and/or reduce the expression of target genes resulting in their rapid growth and associated symptoms.

Among the genes predicted as the target of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142-5p (Table 3), the expression of a considerable numbers have been identified in myometrium, leiomyomas and their isolated smooth cells and to a certain degree are regulated by the ovarian steroids [9, 10, 14, 34, 35].We found that E2, MPA, ICI-18727 and RU-486 differentially regulate the expression of miR-20a, miR-21 and miR-26a in MSMC and LSMC, suggesting that ovarian steroid actions on the expression of these genes are mediated in part through the regulation of miRNAs that target these genes. Such a regulatory mechanism may control the expression of target genes necessary for leiomyoma growth and regression. Furthermore, differential expression of ERα, ERβ, and PR in leiomyomas and myometrium is considered to reflect their differential response to ovarian steroid actions [36, 37] and that could be in part regulated by changes in expression of miR-18a, miR-181, miR-206 and miR-142-5p which are predicted to target ER and PR, respectively. Interestingly, RU-486 and ICI-182780 which are found effective in regressing leiomyoma growth and LSMC proliferation in vitro, respectively [7, 38–40], also altered the expression of miRNAs in MSMC and LSMC. Analysis of miR-20a, miR-21, miR-26a, miR-18a, miR-206, miR-181a and miR-142–5p expression in other cell types indicated their functions in gene regulation involved in cellular apoptosis, differentiation, cell–cell communication, transformation and tumourogenesis [41-43]. Overexpression of miR-20a, a member of miR-17-92 cluster [42, 44, 45], and miR-21 [43, 46] in several cancer cells prevented apoptosis and malignant cellular transformation, and analysis of many cancer-derived cell lines for sequence variations in 15 miRNAs for tumour-associated mutations implicated miR-21 and miR-26a in tumourogenesis [18, 46, 47]. Elevated expression of miR-206 has been found in ER-α-negative MB-MDA-231 cells, but not ERα-positive MCF-7 cells [48] and miR-181 up-regulation has been associated with cellular differentiation and establishment of muscle phenotype [49]. These examples imply that in leiomyomas, the expression and differential regulation of these miRNAs resulting in re-programming of their target mRNAs expression influence the outcome of their growth and regression.

3

miRNAs and gene targets

miRNA	Chromosomal location	Predicted Target Genes
hsa-miR20a	13q31.3	TGFBR2, RUNX1, SMAD7, ADAM9, CCND1, CCND2, E2F1, EGR3, HMGB3, IL17E, MAPK4, WEE1, BRCA1, CASP6, CASP8AP2, CCL25, CCL5, CSF2RA, CYB5D2, F2RL3, FGFR2, FLT3, FNDC3A, HIF1A, HIG2, HSD17B1, HSD17B6, HSPA2, IFNA21, IL-10, IL-25, IL-27Ra.
hsa-miR-21	17q23.2	TGFBR2, BCL2, FASLG, SMAD7, TGFBI, TIMP3, CXCL5, E2F3, SMARCD1, ADAM8, CCL18, CCL20, CCNA1, CCR7, COL4A4, FBLN4, FOSL1, HSD17B4, IGFBP7, IL12A, IL13RA2, IL-15, IL-7, IL-9, JAK3, MAP3K8, MAPK10, MMP12, MMP7, NFIB, PSG4, PSRC1, THBS1, TNFRSF10B, TNFRSF11B, VIM
hsa-miR26a	3P22.3	SMAD4, CYP26A1, CYP2E1, EGF, FGF21, HSD17B 13, HSD17B6, IFNG, IL18R1, IL20, IL22, IL5RA, LEFTY1, MAPK6, MMP12, MMP7, OGN, PCNA, TCF12, ADAMT- SL1, BCAR3, CCL13, CDC27, CDC6, COL4A2, CREBL2, CSF2RA. CXCR6, CYC1, PRR11, PRR8, S100A16, S100B, SOX2, CDK6, CCND2.
hsa-miR18a	13q31.3	ESR1, IGF1, IGF2BP2, IRF2, NCOA1, PTGFRN, RUNX1, SMAD2, SMAD6, SOX21, TEX2, THRAP1, CYB5R2, CYB5R2, CYHR1, CYP39A1, CYP3A7, CYP4A11, FLT1, FMOD, IL10RB, IL11RA, IL17B, IL9R, TFDP1, VEGFB.
hsa-miR206	6p12.2	CALM1, CALM2, CITED2, CTNND2, EDN1, EID1, FGF14, FN1, GAS2L1, IGF1, IGFBP5, ITGB3, MTSS1, NCOA1, NCOA3, NGFR, PAFAH1B1, PDCD10, PDCD4, PDGFA, RASA1, SDCBP, SOX6, STC2, TAGLN2, TGFBR3, TGIF1, TGIF2, THRB
hsa-miR181a	9q33.3	ESR2, ABI1, ABI3BP, ADAM11, BMP3, BMPER, BMPR2, EGR1, EGR3, FGFR3, MMP14, MPP5, NCOA2, NEGR1, PAK4, PAK7, RUNX1, S100PBP, SMAD7, SOX5, SOX6, STC1, TGFBI, TGFBR1, TIMP3, TSC22D2.
hsa-miR142-5p	17q23.2	PGR, HSD17B3, CYB5, CYP26A1, E2F7, EGR3, GAS7, IGF1, IGF2BP3, RUNX1T1, SMAD2, SOX11, SOX5, STC1, TAF10, TCF12, TGFBR2, TGIF2, VCAM1, STAT5B, PLAUR, PDGFRL, IL13RA1, IL17F, IL18RAP, IL1RAPL2, IL1RL1, IL22RA2.

Note:See citations in the text and http://microrna.sanger.ac.uk/sequences regarding most updated predicted target genes.

In conclusion, the results provided evidence that a number of miRNAs are expressed and regulated by the ovarian steroids in myometrium and leiomyomas and their profile is altered in myometrial cellular transformation into leiomyoma and progression from benign into cancerous stage. Because of the key regulatory function of miRNAs in gene expression detailed investigation is necessary to identify their target genes in leiomyomas and that may serve as a useful strategy in regulating their growth.