Expression Pattern of Stemness-Related Genes in Human Endometrial and Endometriotic Tissues

Abstract

Endometriosis is a chronic disease characterized by the presence of ectopic endometrial tissue outside of the uterus and with mixed traits of benign and malignant pathology.

This study aimed at analysing in endometrial and endometriotic tissues the differential expression of a panel of genes involved in preservation of stemness status and consequently considered as markers of stem cell presence.

The expression profiles of a panel of 13 genes (SOX2, SOX15, ERAS, SALL4, OCT4, NANOG, UTF1, DPPA2, BMI1, GDF3, ZFP42, KLF4, TCL1) were analysed by RT-PCR in human endometriotic (n=12) and endometrial samples (n=14). The expression of SALL4 and OCT4 was further analysed by immunohistochemistry.

Genes UTF1, TCL1 and ZFP42 showed a trend for higher frequency of expression in endometriosis than in endometrium (p<0.05 for UTF1), while GDF3 showed an higher frequency of expression in endometrial samples. Immunohistochemical analysis revealed that SALL4 was expressed in endometriotic samples but not in endometrium, despite the expression of the corresponding mRNA in both the sample groups.

This study highlights a differential expression of stemness-related genes in ectopic and eutopic endometrium and suggests a possible role of SALL4-positive cells in the pathogenesis of endometriosis.

INTRODUCTION

Human endometrium undergoes cyclical processes of growth, differentiation, shedding and regeneration as part of the menstrual cycle during the reproductive life of women [1].

Endometriosis is a multifactorial oestrogen-dependent disease that affects 5 to 10% of women of reproductive age in the Western countries. Its defining feature is the presence of endometrium-like tissue in sites outside the uterine cavity, primarily on the pelvic peritoneum and ovaries [2].

Endometriosis can originate from anatomical or biochemical aberrations of uterine function. Theories on the histogenesis of endometriosis belong to five categories: coelomic metaplasia, retrograde menstruation, embryonic cell rest, induction and lymphatic and vascular dissemination [3].

Many studies focused so far on the biomolecular and cellular characteristics of endometriosis in comparison to endometrium and to the possible molecular mechanisms at the basis of the development of endometriotic lesions. Among them, of particular interest is a recent analysis revealing a list of 22 microRNAs (miRNAs) differentially expressed in paired ectopic and eutopic endometrial tissues, that could contribute to endometriosis progression through their cognate target mRNAs [4]. Other studies highlighted a differential expression in endometriotic tissue as compared with endometrium of genes SF1 and estrogen receptor beta, resulting to be primarily controlled by a methylation-dependent epigenetic mechanism [5–6]. Finally, different chromosomal aberrations have been reported in endometriotic samples and in ovarian carcinoma [7].

Differences in stromal cell migration, inflammatory markers and in other pathways between eutopic and ectopic endometrial tissues have been also highlighted [8].

It should also be mentioned that endometriosis can have a genetic basis, since its incidence in relatives of affected women is much more higher than the incidence in women without such familiar history [9].

Stem cells are increasingly becoming the focus of many areas of biomedical research. Stem cells are rare undifferentiated cells virtually present in all adult tissues and organs. These cells retain high proliferative, self-renewal and differentiation potential. The number of stem cells in adult tissues is actively regulated through a strict balance between cell proliferation, cell differentiation and cell death [10]. Recent studies revealed the presence of adult stem cells in endometrium. In particular, a work by Chan RW et al. [11] described clonogenic stromal and epithelial cells in human endometrium, possibly indicating the presence of stem cells.

Two studies [12,13] suggested the presence of stem cells in murine endometrium through the label-retaining cell (LRC) approach. Other studies, also conducted in murine models, demonstrated that stem cells in endometrium derive from bone marrow [14].

The presence of stem cells in endometrium has been demonstrated mainly through the analysis of their surface markers, their clonogenic properties and their differentiation ability. Endometriosis can evolve into ovarian cancer [3,15] and other malignant diseases in which stem cells could play a role, as recently demonstrated [16]. The relationship of endometriosis and ovarian cancer has been demonstrated both by epidemiological studies and by common genetic alterations [3].

Studies on transcriptional profiling of stem cells allowed a preliminary identification of stemness-related genes actively involved in the control of stem cell properties, such as self-renewal ability and retention of an uncommitted state. Initially, genes that control stemness were identified in embryonic stem cells [17, 18]. In adult stem cells, some embryonal “stemness genes” are not expressed.

In this study we aimed to detect the expression of a panel of 13 genes considered as stem cell markers in eutopic endometrium and in endometriotic tissue, through the analysis of the mRNA level for all the 13 genes, and verifying for two of them the data also at protein level. The 13 genes have been selected on the basis of the currently available literature data.

Among them, BMI1 plays a central role in the inheritance of stemness. It belongs to the polycomb group (PcG) genes and is involved in the maintenance of cellular memory through epigenetic chromatin modifications. Recent studies have implicated a role for PcG genes in the self-renewal of stem cells, a process in which cellular memory is maintained through cell division [19]. ERAS is a Ras membrane protein involved in proliferation and tumorigenicity of embryonic stem cells [20]. TCL1 is an oncogene involved in regulation of proliferation of embryonic stem cells and is a downstream gene of OCT4 [21]. UTF1 is a tightly DNA-associated protein with transcriptional repressor activity, and expressed in embryonic pluripotent stem cells [22]. All the other genes we analysed, including OCT4, SOX2, SOX15, NANOG, SALL4, DPPA2, GDF3, ZFP42 and KLF4, code for transcription factors for genes involved in the preservation of stem cells pluripotency (see also Supplementary file 1 for additional references specific for stemness-related genes).

Our results highlight the expression of stem cell markers both in endometrial and endometriotic tissues, suggesting that stem cells could play a role in disease progression.

METHODS

Patients and samples

Clinical samples of endometrial and endometriotic tissues were collected from 26 patients, (endometrial tissues from n=14 patients aged 29–58, mean 46.9; endometriosis samples from n=12 patients, aged 24–46, mean 34.4) at the Department of Gynaecology, Obstetrics and Reproductive Medicine of the Second University of Naples, undergoing hysterectomy, laparoscopy or laparotomy for benign pathologies. Informed written consent was obtained from each patient. Surgery was performed irrespective of the day of patient’s menstrual cycle. The patients had never received any hormonal treatment before surgery.

After surgery, endometrial biopsies and excised ovarian endometriotic lesions were formaldehyde-fixed and hematoxylin-stained cross-sections were analysed by experienced histopathologists for assessment of the grade of endometriosis (I–IV) and for determination of the stage of menstrual cycle (proliferative or secretory), referring to established histological criteria [23]. The clinical characteristics of the patients are shown in Table I.

Table I.

Patient clinical characteristics (PP: proliferative phase; SP: secretory phase; M: menopause).

Case No.	AGE (years)	PHASE OF MENSTRUAL CYCLE	GRADE OF ENDOMETRIOSIS	PATHOLOGY
Endometrium 1	58	M		Uterus fibromatosis
Endometrium 2	29	PP		Uterine myoma
Endometrium 3	53	SP		Endometrial polyp
Endometrium 4	54	SP		Uterus fibromatosis
Endometrium 5	46	SP		Uterus fibromatosis
Endometrium 6	58	M		Cystocele
Endometrium 7	52	SP		Uterus fibromatosis
Endometrium 8	43	PP		Uterine myoma
Endometrium 9	41	PP		Uterus fibromatosis
Endometrium 10	37	SP		Ovarian cyst
Endometrium 11	51	M		Endometrial polyp
Endometrium 12	41	SP		Uterine myoma
Endometrium 13	52	SP		Uterus fibromatosis
Endometrium 14	41	PP		Uterus fibromatosis
Endometriosis 1	38	SP	II
Endometriosis 2	39	SP	III
Endometriosis 3	44	SP	IV
Endometriosis 4	29	SP	II
Endometriosis 5	26	SP	III
Endometriosis 6	28	PP	III
Endometriosis 7	31	PP	III
Endometriosis 8	46	SP	IV
Endometriosis 9	24	PP	III
Endometriosis 10	42	PP	IV
Endometriosis 11	33	PP	III
Endometriosis 12	33	PP	I

The samples from each patient were either snap frozen and stored at – 80°C or fixed in buffered formaldelyde 4% (Sigma Aldrich) and embedded in paraffin using standard techniques for immunhistochemical analysis.

RNA extraction and RT-PCR

Total RNA was extracted from frozen tissue samples using the TRI REAGENT (Molecular Research Center Inc., OH, USA) and from paraffin-embedded tissues (RNeasy minikit, Qiagen) according to manufacturer’s instructions. RNA was treated with DNase I (Ambion) to remove DNA contamination. RNA concentration was measured using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies). RNA integrity was verified by electrophoresis on denaturing 1% agarose gel.

Absence of residual genomic DNA was verified by PCR on total RNA without reverse transcription. Genomic human DNA was used as positive control of PCR reactions.

cDNA was generated from 200 ng of each RNA sample. Reverse transcription was done at 42°C for 1 h in presence of random examers and Moloney-Murine Leukemia Virus (M-MULV) reverse transcriptase (Finnzymes). GeneBank sequences for human mRNAs SOX2, SOX15, ERAS, SALL4, OCT4, NANOG, UTF1, DPPA2, BMI1, GDF3, ZFP42, KLF4, TCL1 and the Primer Express software (Applied Biosystem) were used to design primer pairs for the genes and the house keeping gene GAPDH. Primer sequences are listed in Table II. They were chosen to yield 100–150 bp Each PCR was repeated for 35 cycles. PCR products and were validated running the PCR products on agarose gel to confirm a single band.

Table II.

Summary of RT-PCR primer sequences, position, annealing temperature and chromosome mapping position of the stemness-related target genes.

GENE	PRIMER POSITION	PRIMER SEQUENCE	ANNEALING TEMPERATURE (C°)	PCR PRODUCT (bp)	CHROMOSOME MAPPING OF THE GENE
GAPDH	472 799	5′-GCATCCTGCACCACCACCTG-3′ 5′-GCCTGGTTCACGACGTTCTT-3′	55	347	12p13
SOX2	1563 1701	5′-CCATCCACACTCACGCAAAA-3′ 5′-TATACAAGGTCCATTCCCCCG-3′	59	139	3q27
OCT4	1121 1223	5′-TCCCATGCATTCAAACTGAGG -3′ 5′-CCAAAAACCCTGGCACAAACT -3′	60	103	6p21,31
NANOG	1169 1310	5′-TGGACACTGGCTGAATCCTTC -3′ 5′-CGTTGATTAGGCTCCAACCAT -3′	59	142	12p13,31
KLF4	1508 1689	5′-CTGCGGCAAAACCTACACAA -3′ 5′-GGTCGCATTTTTGGCACTG -3′	60	182	9q31
ERAS	969 1103	5′-AATGTAGACCTTTCCCCAGGC -3′ 5′-AAAGCCCCTCACCAAGTGAA -3′	58	135	Xp11,23
GDF3	778 887	5′-AAAAGGAAGAGCAGCCATCCCT -3′ 5′-GCAATGATCCACTTGTGCCAA -3′	60	110	12p13.1
SOX15	315 441	5′-GAACAGGTTGGAAGCAAAGGC -3′ 5′-GCGTCGATCCTGAAAATGGA -3′	59	127	17p13
DPPA2	798 905	5′-AGCCATGTTGGCATCATGG -3′ 5′-GAGGCTTGCAGCAAAAAGGC -3′	58	108	3q13,13
SALL4	2394 250	5′-GCCCAGATATCCTGGAAACCA -3′ 5′-TTCTCGGAGCTCTCTGCTTTG -3′	60	115	20q13,13/13,2
TCL1	667 793	5′-CTCGGCTTTTTCTCAGCTGGAT -3′ 5′-GGTGAATCGGCTGTGTTCTCA -3′	59	127	14q32,1
ZFP42	953 1085	5′-ATGACAGTCTGAGCGCAATCG -3′ 5′-AACGCTTTCCCACATTCCG -3′	60	133	4q35,2
UTF1	876 992	5′-CGACATCGCGAACATCCTG -3′ 5′-AGAATGAAGCCCACGGCCA -3′	64	117	10q26
BMI1	437 575	5′-AATGTCTTTTCCGCCCGCT -3′ 5′-ACCCTCCACAAAGCACACACAT -3′	59	139	10p11,23

Each RT-PCR reaction was repeated at least three times. A semi-quantitative analysis of mRNA levels was performed by the GEL DOC UV system (Bio-Rad, Hercules, CA) on agarose gels containing the GelStar Nucleic Acid Gel Stain (Lonza), an highly sensitive fluorescent stain able to detect as little as 20 picograms of DNA, with a 4–16 fold increase of sensitivity compared to Ethidium Bromide.

In order to determine the lowest number of molecules of a given mRNA in a pool that can be detected by RT-PCR, it is warranted to know the percentage of that mRNA in the pool. In many cases, it is not possible to determine this percentage. Consequently we previously set up in our laboratory an alternative method based on serial dilutions of total RNA, ranging from 1000 ng to 1 ng, used to carry out RT-PCR to detect high- (GAPDH), medium- (HPRT) and low-expressed (E2F2) mRNAs after 35 cycles.

Highly expressed mRNA was detected in all experimental conditions we used in the presence of GelStar, whereas 10 ng of total RNA was the lowest quantity to detect medium and low expressed mRNAs.

In the RT-PCR analysis in this study we used 200 ng of total RNA and 35 cycles for amplification, far above the limit of detection of low-expressed mRNAs.

When minimal differences in gene expression were detected by PCR, experiments were repeated using the real-time PCR assays, run on an Opticon 4 machine (Bio-Rad). Reactions were performed according to the manufacturer’s instructions using the SYBR Green PCR master mix (Stratagene). Relative quantitative RT-PCR was used to determine the fold difference for genes. Melting curves (65–94°C) were also generated to determine whether there were any spurious amplification products. The real time PCR efficiency was calculated for each primer pair using a dilution series and the MJ Opticon II analysis software.

Immunohistochemical analysis

Tissue samples from patients were fixed in 4% buffered formaldehyde, dehydrated and embedded in paraffin. Consecutive 5 μm cross-sections were placed on coated slides, deparaffinized through a series of xylene and ethanol washes and used for immunohistochemical analysis of SALL4 and OCT4 expression. We verified the immunohistochemical signal for SALL4 using as positive and negative controls sections of mouse adult testis and heart, respectively (supplemental file 1). We verified the immunohistochemical signal for OCT4 using as positive and negative controls sections of mouse embryo testis (E13.5) and mouse adult heart, respectively (supplemental file 2).

Antigen retrieval was obtained through incubation in citrate buffer at pH 6.0 for 10 minutes followed by gradual cooling at room temperature for 20 minutes. After 1hr incubation in blocking solution (5% BSA and 1% donkey serum), slides were incubated overnight at 4°C with SALL4 mouse monoclonal antibody (1:100, Abnova, Walnut, CA, USA) or OCT4 rabbit polyclonal antibody (1:250, Abcam, Cambridge, UK) diluted in blocking solution, according to manufacturer’s instructions. In negative controls the primary antibodies were omitted.

After washing, slides were incubated with biotinylated anti-mouse or anti-rabbit secondary antibodies for 30 minutes at R.T. After washing, slides were incubated with streptavidin-peroxidase (HRP) (Vector Laboratories, Burlingame, CA) for 30 minutes at R.T. Finally, specific hybridisation of antibodies was highlighted through incubation with diaminobenzidine (DAB) and HRP substrate buffer (Vector). The DAB substrate solution gives a brown precipitate at the site of the target antigen recognized by the primary antibody. Nuclei were counterstained blue with Mayer’s hematoxylin (Merck, Darmstadt, Germany). Dried slides were immersed in xylene solution and cover-slipped using ultramount.

Image screening and photography of serial cross-sections were performed using a Leica IM 1000 System. Slides were analysed by two blinded independent observers.

Statistical analysis

The Multivariate Statistical Package (Kovach Computing Service, Isle of Anglesey, UK) was used for Ward’s Minimum variance clustering method to evaluate gene expression variability among different samples.

Statistical analyses (Fisher’s exact test; Student’s t and Bonferroni’s tests) were evaluated using the GraphPad Software (Prism 4.0).

RESULTS

RT-PCR analysis of stemness-related genes

We analysed by RT-PCR the expression of a set of 13 stemness-related genes (Table I) in endometrial (n=14) and endometriotic (n=12) biopsies. Overall results are shown in Table III, while histogram in Fig. 1A shows the percentage of expression of each gene in the endometrium and endometriotic sample groups and histogram in Fig. 1B reports the number of expressed stemness-related genes in endometrial and endometriotic samples.

Table III.

Qualitative RT-PCR analysis of stemness-related genes in 14 endometrial tissues and in 12 endometriotic samples. The table summarizes the results on the presence (+) or absence (−) of gene expression for each patient (PP: proliferative phase; SP: secretory phase; M: menopause).

CASE No. (cycle phase- endometriosis grade)	GENE
	SOX2	DPPA2	GDF3	TCL1	ZFP42	UTF1	ERAS	SALL4	NANOG	SOX15	OCT4	KFL4	BMI1
Endometrium 1 (M)	−	−	−	−	−	−	−	−	+	+	+	+	+
Endometrium 2 (PP)	−	−	−	−	−	−	−	+	+	+	+	+	+
Endometrium 3 (SP)	−	−	−	+	−	−	−	+	+	+	+	+	+
Endometrium 4 (SP)	−	−	−	−	−	+	+	+	+	+	+	+	+
Endometrium 5 (SP)	−	−	+	+	−	+	+	+	+	+	+	+	+
Endometrium 6 (M)	−	−	−	−	−	−	−	+	+	+	+	+	+
Endometrium 7 (SP)	−	−	+	−	−	+	+	+	+	+	+	+	+
Endometrium 8 (PP)	−	−	−	+	−	−	−	+	+	+	+	+	+
Endometrium 9 (PP)	−	−	+	−	−	+	+	+	+	+	+	+	+
Endometrium 10 (SP)	−	−	+	+	−	−	+	−	+	+	+	+	+
Endometrium 11 (M)	−	−	+	+	−	−	+	+	+	+	+	+	+
Endometrium 12 (SP)	−	−	−	−	−	−	+	−	−	−	+	+	+
Endometrium 13 (SP)	−	+	+	+	−	+	+	+	+	+	+	+	+
Endometrium 14 (PP)	−	+	+	+	−	+	+	+	+	+	+	+	+
Endometriosis 1 (II )	−	−	+	+	−	+	+	+	+	+	+	+	+
Endometriosis 2 (III)	−	−	+	+	−	+	+	+	+	+	+	+	+
Endometriosis 3 (IV)	−	−	−	+	+	+	+	+	+	+	+	+	+
Endometriosis 4 (II)	−	−	−	−	−	+	+	−	+	+	+	+	+
Endometriosis 5 (III)	−	−	−	+	+	−	−	+	+	+	+	+	+
Endometriosis 6 (III)	−	−	−	+	+	+	+	+	+	+	+	+	+
Endometriosis 7 (III)	−	−	−	+	−	+	−	+	−	+	+	+	+
Endometriosis 8 (IV)	−	−	+	+	−	+	−	+	−	+	+	+	+
Endometriosis 9 (III)	−	−	−	+	−	−	−	+	−	+	+	+	+
Endometriosis 10 (IV)	−	−	−	+	−	+	−	+	+	+	+	+	+
Endometriosis 11 (III)	−	+	−	−	−	+	+	+	+	−	+	+	+
Endometriosis 12 (I)	−	+	−	+	−	+	−	+	+	+	+	+	+

Figure 1.

A) The histogram shows the frequency of expression of stemness-related genes in endometrial tissues (white columns) and in endometriotic samples (grey columns). B) The histogram shows the number of expressed stemness-related genes in endometrial tissues (white columns) and in endometriotic samples (grey columns).

SOX2 mRNA resulted to be not expressed in any of the samples we analysed (Table III, Fig. 1A). Conversely, OCT4, KFL4 and BMI1 mRNAs were expressed in all the endometrium and endometriotic samples we examined (Table III, Fig. 1A).

Other genes, such as DPPA2 and SOX15 resulted to be expressed in the same percentage of patients in endometrial and endometriotic sample groups (Fig. 1A).

ERAS, NANOG and GDF3 showed a slightly higher (but not statistically significant) frequency of expression in endometrial than in endometriotic samples (Fig. 1A).

The remaining genes we analysed (SALL4, UTF1, TCL1) showed a different percentage of expression in endometrium and endometriotic sample groups, with a trend for higher percentage of expression in endometriotic samples than in endometrium samples. In more detail, UTF1 (also known as undifferentiated embryonic cell transcription factor 1) showed a significantly higher frequency of expression in endometriotic samples than in endometrium (83% vs. 43%, p<0.05). Also TCL1 showed a remarkable difference in the percentage of expression between endometrial and endometriotic samples (50% vs. 83%), even if not statistically significant. Of note, ZFP42 was expressed only in 25% of endometriotic tissues (classified as III and IV grade) and in none of the endometrial biopsies.

The 12 endometriotic samples we analysed co-expressed a minimum of 6 to a maximum of 10 stemness-related genes, (Fig. 1B). Conversely, the 14 endometrial samples we analysed co-expressed a minimum of 4 to a maximum of 11 stemness-related genes (Fig. 1B). No significant differences were observed in the number of expressed genes between the two groups of samples.

In this study we report only qualitative RT-PCR data about the expression of a panel of 13 stemness-related genes, since the endometrial and endometriotic biopsies were harvested during the last decade and in some cases the quality of RNA extracted from frozen or paraffin-embedded tissues did not allow to obtain fully reliable quantitative RT-PCR data. Nevertheless, in some patients we found a correlation between the expression level of stemness-related genes and the grade of endometriosis, as well as a trend (not statistically significant) for an higher expression level of some genes (e.g. SALL4) in endometriotic tissues rather than in endometrium samples (data not shown).

The RT-PCR data concerning the presence or absence of gene expression in the 26 samples under analysis were used to carry out a Minimum Variance test to evaluate gene expression variability among different patients. Our goal was to obtain a minimum variance clustering based on a matrix constructed with the presence/absence of gene expression points, such that patients having similar patterns of expressed/not expressed genes fall in the same cluster and have a more “genetic homogeneity” compared to those showing different expression patterns, which are then classified in distinct clusters. We did not find any correlation between the phase of the menstrual cycle and the number of expressed stemness-related genes (Table III). Similarly, no significant evidence was detected for a correlation between the grade of endometriosis and the number of expressed stemness-related genes (Table III).

Immunohistochemical detection of SALL4 and OCT4 proteins in endometrial and endometriotic samples

Endometrial (n=14) and endometriotic (n=12) samples embedded in paraffin were submitted to IHC-mediated analysis of the expression of SALL4 and OCT4. We selected these two proteins for IHC analysis since they play an important role in stemness preservation [24], to clarify possible misleading results deriving from RT-PCR analysis of OCT4 expression and finally because quantitative RT-PCR data indicated a trend for an higher expression level for their mRNAs in endometriosis samples rather than in endometrium, even if not statistically significant.

We analysed at least five consecutive cross-sections for each tissue sample. Only cross-sections of endometrial and endometriotic tissues with markedly brown stained cells, showing a clear structure, were scored positive for SALL4 and OCT4 protein expression.

Positive cells for SALL4 and OCT4 were detectable in different consecutive cross-sections of the tissue samples we analysed (Fig. 2 and 3). The staining for both SALL4 and OCT4 showed a nuclear localisation.

Figure 2.

Representative immunohistochemical staining of SALL4 in human endometrium and in endometriotic tissue. Hematoxilin counterstaining. Endometriotic tissue (A, B) is compared to endometrial tissue (E, F). Immunohistochemical staining of serial sections of the tissue used in A without Ab’ was done as negative control of the reaction (C, D). Black arrow in B indicates a representative SALL4 IHC-positive cell. Subparts (B, D, F) represent 100x magnification of the area enclosed in the black perimeter in A, C, E (40x magnification).

Figure 3.

Representative immunohistochemical staining of OCT4 in human endometrium and in endometriotic tissue. Hematoxilin counterstaining. Endometriotic tissue (A, B) is compared to endometrial tissue (E, F). Immunohistochemical staining of serial sections of the tissue used in A without Ab’ was done as negative control of the reaction (C, D). Immunohistochemical staining of serial sections of the tissue used in E without Ab’ was done as negative control of the reaction (H, G). Black arrows in B and F indicate representative OCT4 IHC-positive cells. Subparts (B, D, F, H) represent 100x magnification of the area enclosed in the black perimeter in A, C, E, G (40x magnification).

Cells positive for SALL4 were found in all the endometriotic tissues we analysed (Fig. 2). None of the endometrial samples revealed cells positive for SALL4. To further confirm this data, IHC detection of SALL4 was also conducted on paired ectopic and eutopic endometrium from the same patient (sample endometriosis 8, Tables I and III), revealing SALL4-positive cells only in endometriotic tissue.

Cells positive for OCT4 were found in the stroma of all the endometriotic tissues we analysed. Stromal cells positive for OCT4 were detected also in the endometrial samples (Fig. 3).

We observed only single stromal cells positive for OCT4 immunostaining both in endometrium and in endometriotic samples. Conversely, SALL4-positive cells in endometriotic tissues were located also in a periglandular position and in the stromal vasculature. Control immunohistochemical reaction for SALL4 was positive on mouse adult testis and negative on mouse adult heart (Supplemental Fig. 1).

Control immunohistochemical reaction for OCT4 was positive on mouse embryo testis (E13.5) and negative on mouse adult heart (Supplemental Fig. 2).

DISCUSSION

In this study, we have characterized at mRNA level the expression of a panel of 13 embryonic stemness-related genes in two sets of human endometrium and endometriotic samples, together with the immunohistochemical verification for a subgroup of two factors, to evaluate which of them were present in endometrial and endometriotic tissues.

Different studies highlighted so far the presence of stem cells in endometrium. In particular, Du et al. [14] demonstrated that lethally irradiated female mice receiving bone marrow transplantation from male donors show male-derived cells incorporated into the endometrium. The presence of stem cells has been demonstrated also in women submitted to bone marrow transplantation from mismatched donors [25]. The bone marrow compartment can be subdivided into 2 interdependent spaces: the hematopoietic cell compartment and the stroma. The stroma is composed of mesenchymal stem cells (MSCs), fibroblasts, adipocytes, nerves, and the bone marrow’s vascular system. MSCs are quite rare, as they represent between 0.01% and 0.001% of nucleated cells in adult human bone marrow, depending on the age of individuals [26]. Nonhemapoietic stem cells from bone marrow can potentially contribute to the preservation of multiple tissues. Some studies indicate that stem cells in endometrium are of bone marrow origin and that share many characteristics with mesenchymal stem cells (MSCs), as they are able to differentiate into condrocytes, osteocytes and adipocytes and express peculiar antigens [27].

Other recent studies highlighted the presence of stem cells also in the menstrual blood, characterized by an high proliferative rate in vitro, high differentiation ability, expression of a number of stemness-related non-hematopoietic markers (including OCT4), and production of matrix metalloproteases (MMPs), cytokine growth factors and angiogenic factors [28, 29]. Nevertheless, the presence of hematopoietic stem cells (HSCs) has been also demonstrated immunologically in endometrium [30]. The endometrial stem cells, both of hematopoietic or non-hematopoietic nature, probably contribute to the de novo formation of stroma, glands and vasculature in the reproductive cycle.

In this study, we highlighted the possible presence of stem cells in all the endometrium and endometriotic samples through the expression of 13 stemness-related genes.

Our RT-PCR data highlight a significantly higher number of endometriotic samples expressing UTF1 mRNA in comparison to endometrial biopsies (p<0.05). UTF1 is highly and almost exclusively expressed during embryogenesis [31]. In more detail, it is specifically expressed in the inner cell mass and primitive ectoderm and is down-regulated at early primitive streak stages [32]. Of interest, it has been reported that UTF1 expression is maintained in the primordial germ cells in developing embryos and in the gonads in adult animals [33].

ZFP42 (also known as REX-1) is expressed only in about 25% of endometriotic samples, classified as III and IV grade (Table 3). A recent study by Kristensen DM et al. [34] showed that ZFP42 and UTF1 are expressed throughout human testes development and in testicular germ cell tumours and in testicular carcinoma, showing similarities with pluripotent embryonic stem cells.

Promoter analysis indicated that the murine UTF1 gene is transcriptionally regulated by OCT4 and SOX2 [35]. Finally, a recent study indicated that UTF1 is a stably chromatin-associated transcriptional repressor protein involved in the initiation of embryonic stem (ES) cell differentiation, but not in ES cell self-renewal [36].

RT-PCR data indicate a trend for a higher frequency of expression also of TCL1 in endometriotic samples. TCL1 (also known as T cell leukemia 1) is a proto-oncogene highly activated in various human neoplastic diseases, while its physiological expression is tightly limited to early developmental cells as well as various developmental stages of immune cells [37].

One of the analyzed genes (SOX2) was not detected neither in endometrial or endometriotic tissue, while DPPA2 was expressed only in two patients for each group. This result is not surprising, since embryonic stem cells have broader stemness properties (self-renewal, pluripotency) compared with adult stem cells.

The analysis of Minimum Variance did not reveal any homogeneous clusters of samples on the basis of gene expression data, possibly because of the relatively low number of samples we analysed or because of the heterogeneity of samples in relation to the number and type of cells they contain.

Recently, it has been discovered that rare cells in endometrial stroma of about 44% of women are positive for OCT4 (also known as OCT3/4,OCT3 and POU5f1) [38], a protein member of the POU transcription factor family. OCT4 is expressed in pluripotent cells, and its down-regulation is associated with loss of pluripotency. The results of the mentioned study are in agreement with our RT-PCR and IHC data, since we highlighted the expression of OCT4 mRNA and protein in all the eutopic endometrium samples we analysed.

The latest results about OCT4 isoforms reveal the presence of three alternative splice variants (OCT4-A, OCT4-B, OCT4-B1) [39].

The PCR primers we used for OCT4 mRNA analysis (Table 2) are both enclosed within the exon 5 sequence and cannot distinguish among the variants OCT4-A, OCT4-B and OCT-4B1 and the RNA transcribed by the two pseudogenes identified by the GeneBank numbers NG_005793 and NG_006104. For this reason, together with the observation that the OCT4 RT-PCR signal resulted to be higher in endometriosis samples than in the endometrium group, we decided to further analyse the OCT4 expression in the two sets of human endometrial and endometriotic samples at protein level. The antibody for OCT4 we used was obtained using as immunogen a synthetic peptide derived from within residues 300 to the C-terminus of human OCT4. The OCT4 spliced variants OCT4-A and OCT4-B share an identical C-terminal domain, while the recently discovered OCT4-B1 lacks the C-terminal domain because of a stop codon in the criptic exon 2b, and consequently the immunohistochemical data we obtained potentially related to the OCT4-A and -B isoforms. Nonetheless, since we obtained a clear nuclear localisation of the immunohistochemical signal for OCT4 (Fig. 3), we can argue that it corresponds to the OCT4-A variant, as it has been reported that the OCT-4B variant is localised in the cytoplasm [40,41].

The variant OCT4-B1 has been discovered very recently [39], and consequently all the currently available literature data concerning the expression of OCT4 protein concern the isoforms A and B, as the antibody specific for the putative truncated protein translated by the OCT4-B1 splice variant is not available. The translation of the OCT4-B1 mRNA variant identified by Atlasi et al. has not yet been demonstrated, and its putative role in stemness and in carcinogenesis has been only suggested, but not demonstrated experimentally (e.g. through RNA interference assays). Moreover, nothing is known about the cellular localization (at nuclear or cytoplasmatic level) of the protein possibly expressed by the novel OCT4 mRNA splicing variant.

It should be underlined that OCT4 has been considered for a long time a reliable marker for stemness, but a recent study demonstrated the expression of OCT4 also in normal differentiated adult cells from human peripheral blood, thus suggesting that the presence of OCT4 alone is no more sufficient to define a cell as pluripotent [42]. Nevertheless, in our experiments we supported the presence of OCT4 as a marker of stemness with the expression data of adjunctive 12 stemness-related genes.

Parallel experiments revealed the presence of SALL4 mRNA both in eutopic and ectopic endometrium samples, but revealed the presence of SALL4 protein only in endometriotic samples.

It should be underlined that we were also able to analyse the SALL4 expression in paired ectopic and eutopic endometrial tissue from a same patient (sample endometriosis 8 in Table I), identifying SALL4-positive cells only in ectopic endometrium. The direct comparison between autologous ectopic and eutopic endometrium can exclude variables related to individual genetic variability and to different effect of hormonal stimulation during menstrual cycle, and thus it can further clarify the contribution of stem cells to the pathogenesis of endometriosis.

Nevertheless, it should be considered that this differential expression of SALL4 protein between endometrial and endometriotic tissues could be related not necessarily to a translational mechanism of regulation of SALL4 expression, but could be related to the very low expression of SALL4 protein in endometrium.

The presence of OCT4- and SALL4-positive cells mainly in the stroma of endometrial and endometriotic samples is in agreement with other studies based on stem cell detection through the analysis of “stemness” markers [38, 16]. Nevertheless, we found some SALL4-positive cells also in the vasculature and in periglandular position.

SALL4 and OCT4 work as essential stemness factors. Our choice to analyze at protein level both SALL4 and OCT4 relies also on experimental evidence that SALL4 forms a crucial interconnected autoregulatory network with OCT4 in embryonic stem cells [43]. It has also been demonstrated in mouse embryonic stem cells that SALL4 is a transcriptional regulator of OCT4 and has a critical role in the maintenance of stem cell pluripotency by modulating OCT4 expression [44].

CONCLUSIONS

Our data indicating an increased presence of stem cell markers in endometriotic samples are in agreement with the recent studies revealing an increased expression of the adult stem cell marker Musashi-1 in endometriosis and endometrial carcinoma [16]. Our preliminary results indicate that the percentages of single cells positive for SALL4 and OCT4 we detected in the stroma of endometriotic tissues are comparable to those found by Gotte M et al. for Musashi-1 positive cells (data not shown). The contribution of stem cells to endometriosis has been hypothesized by many studies and reviews [45, 14].

If further verified, the presence of stem cells in ectopic and eutopic endometrium can provide new insights in the mechanisms at the basis of gynaecological diseases related to cell proliferation, including endometrium carcinoma.

To our knowledge, this is the first study highlighting the expression of a panel of stemness-related genes in human endometrial and endometriotic samples, with a particular relevance for UTF1 and TCL1. Moreover, we report for the first time the expression of SALL4 and OCT4 proteins in endometriotic samples. Overall data obtained in this study suggest a possible role for stem cells in the pathogenesis of endometriosis, even if further data are warranted to support this hypothesis.