Annexin A1 peptide and endothelial cell‐conditioned medium modulate cervical tumorigenesis

Cervical cancer is one of the leading causes of cancer death in women worldwide, and its tumorigenesis can be influenced by the microenvironment. The anti‐inflammatory protein annexin A1 (ANXA1) has been reported to be associated with cancer progression and metastasis, suggesting that it plays a role in regulating tumour cell proliferation. Here, we examined the effect of the N‐terminal peptide Ac2‐26 of ANXA1 on the HaCaT cell line (normal) and HeLa cell line (cervical cancer) co‐cultured with endothelium cell‐conditioned medium (HMC). Treatment with Ac2‐26 decreased proliferation and increased motility of cervical cancer cells, but did not affect cellular morphology or viability. Combined HMC stimulus and Ac2‐26 treatment resulted in an increase in apoptotic HeLa cells, upregulated expression of MMP2, and downregulated expression of COX2,EP3 and EP4. In conclusion, Ac2‐26 treatment may modulate cellular and molecular mechanisms underlying cervical carcinogenesis.

Cervical cancer is one of the leading causes of cancer death in women worldwide, and its tumorigenesis can be influenced by the microenvironment. The anti-inflammatory protein annexin A1 (ANXA1) has been reported to be associated with cancer progression and metastasis, suggesting that it plays a role in regulating tumour cell proliferation. Here, we examined the effect of the N-terminal peptide Ac2-26 of ANXA1 on the HaCaT cell line (normal) and HeLa cell line (cervical cancer) co-cultured with endothelium cell-conditioned medium (HMC). Treatment with Ac2-26 decreased proliferation and increased motility of cervical cancer cells, but did not affect cellular morphology or viability. Combined HMC stimulus and Ac2-26 treatment resulted in an increase in apoptotic HeLa cells, upregulated expression of MMP2, and downregulated expression of COX2, EP3 and EP4. In conclusion, Ac2-26 treatment may modulate cellular and molecular mechanisms underlying cervical carcinogenesis.
Cervical cancer is one of the main gynaecological cancers worldwide, and has been associated with the human papillomavirus (HPV) and/or infection with bacteria such as Chlamydia trachomatis and Neisseria gonorrhoeae [1,2]. Cervical neoplasm is associated with persistent HPV in more than 95% of cases, particularly in people with impaired immune system functions [2,3]. In overview, cervical cancer is the seventh most frequent type of tumour and fourth most common in women, disregarding non-melanoma skin cancer [2,4]. Diagnosis may be ascertained using the Papanicolaou test, which assists in the identification of persistent inflammation caused by the HPV, and in cases of positive results, treatment may be initiated [5].
Besides infections, the cancer microenvironment also influences the tumour development process because it is very different from that of corresponding healthy tissue [6]. The microenvironment has been characterised by its unsettled extracellular matrix composition, increased microvessel density, and abundance of inflammatory cells and fibroblasts with an activated phenotype [6][7][8]. There is evidence that the cancer progression and development does not rely only on genetic characteristics, but also on the interaction between the cancer and stromal cells, including endothelial cells, immune cells, fibroblasts, adipocytes and inflammatory cells [6]. The different tumour phases that are associated with inflammation are considered an imminent threat, directly connected to the development process [9]. In the inflammatory scenario, Annexin A1 (ANXA1) has anti-inflammatory effects and therefore plays a key role in the modulation of the inflammatory response.
Annexin A1, which is also expressed by the tumour cells, acts as a modulator of the inflammatory process and has been linked to tumourigenesis [10]. Studies involving ANXA1 and cancer are controversial, but nevertheless indicate that this protein may be a target for new therapeutic interventions and used as a potential biomarker [11]. Its functions are specific to each type of cancer and there is evidence to indicate that its regulation and subcellular localisation are linked to the development, invasion, metastasis, progression and treatment resistance of tumours [12][13][14][15].
Structurally, the protein has a C-terminal core, representing 80% of its composition; this portion is common to all members of the annexin superfamily, and has four repeated homologous sequences and the 'type 2' domain for calcium linkage [11]. The variable N-terminal core is unique in length and sequence to each member of the family, and includes potential phosphorylation, glycosylation and peptidase action sites [16,17].
The biological activity of ANXA1 can be reproduced by the first 26 amino acids of the N-terminal core [18] or by some smaller peptides [19]. Since this was discovered, it has become common practice to use these molecules in experimental models of acute [20][21][22][23], chronic [24] and systemic inflammation [25], and also in in vitro studies using different cancer cell lines [26][27][28][29]. Recently, the use of the peptide was evaluated in skin allograft [30] and in inflammatory ocular disease [21,31].
There is evidence of a relationship between ANXA1 expression and cervical tumourigenesis. To ascertain the upregulation of the phosphorylated protein according to disease progression, samples from dysplasia and cervical cancer stages I, II, and III have been used [32]. Other work showed that ANXA1 was downregulated in all stages of the disease [33], and another study, analysing healthy, stage I, II and III, and invasive cancer samples, demonstrated that the protein expression levels corresponded to the disease progression [34].
ANXA1's contributions to tumourigenesis are still not well known, and considering its role in inflammation, it is an important area of research. The available data also point to controversies in the expression of this protein in cervical carcinogenesis, indicating a possible research field.
Considering the important role of ANXA1 in the inflammatory response and in tumours, we analysed the activity of the synthetic peptide of the ANXA1 protein in a cervical carcinoma cell line, along with the conditioned medium of endothelial cells, to help elucidate the processes that occur in the tumour microenvironment and expand understanding of ANXA1 as a therapeutic alternative. The rationale for this co-treatment is that paracrine factors in the conditioned medium of human umbilical vein endothelial cells (HUVECs) simulate the cancer microenvironment, which influences the tumour development process, and is very different from that of corresponding healthy tissue.

Ac2-26 peptide response
Proliferation, motility and cytotoxicity of the human immortalised keratinocyte (HaCaT) cell line and the HeLa cell line (human cervical adenocarcinoma cells infected with HPV18) in response to Ac2-26 peptide treatment were studied. The HaCaT cell line showed an increase in proliferation after 72 h (Fig. 1A), and motility after 24 h, closing the experimental wound, and for this reason the cells detached from the well plate, after 24 h ( Fig. 1B and C). In the HeLa cell line, proliferation was decreased after 2, 24, 48 and 120 h (Fig. 1A), while motility was increased after 24 and 48 h (Fig. 1B). Cytotoxicity was not observed in either cell line at any of the experimental times (Fig. 1D). Late apoptosis was decreased in both cell lines after the treatment ( Fig. 2A). Gene expression showed an upregulation of all six genes analysed in the HaCaT cell line, and of prostaglandin E receptor 4 (EP4), matrix metallopeptidase 2 (MMP2) and matrix metallopeptidase 9 (MMP9) in the HeLa cell line (Fig. 2B).

Conditioned medium of endothelial cells (HMC) and Ac2-26 peptide response
In the HaCaT cell line, secreted factors from endothelial cells (HUVECs) without Ac2-26 peptide treatment (HMCS) increased proliferation after 24 h (Fig. 3A). With the combination of secreted factors of endothelial cells and Ac2-26 treatment (HMCT), it was possible to observe an increase of the proliferation at 48 and 120 h, but a decrease at 72 h (Fig. 3B). Motility decreased after 24 h in the HaCaT cells (Fig. 3C,D) after induction with the conditioned medium without (HMCS) and with (HMCT) Ac2-26 peptide treatment.
Moreover, both conditions showed cytotoxicity to these cells only at 48 h (Fig. 3E,F).
In the HeLa cell line, the secreted factors of endothelial cells without treatment (HMCS) led to a decrease in proliferation after 24 h (Fig. 4A), while with induction with HMC and the peptide treatment there was a decrease in proliferation at 72 h, but an increase at 24 and 48 h (Fig. 4B). In HeLa cells, motility had increased at 4 h after the induction with the conditioned medium without and with the treatment (HMCS and HMCT), but at 24 h there was a decrease after HMCT induction that only became statistically significant in 120 h (Fig. 4C,D). As with the HaCaT cell line, HeLa cells showed cytotoxicity only at 48 h (Fig. 4E,F).
Late apoptotic cells were increased in the HaCaT cell line (Fig. 5A,B) after induction with the conditioned medium without and with the treatment (HMCS and HMCT), but decreased in the HeLa cells ( Fig. 5C) after induction with the conditioned medium without the treatment (HMCS). In the HeLa cells after the induction with the conditioned medium with the treatment (HMCT), it was possible to observe an increase in apoptotic cells, but still showing more viable cells (Fig. 5C,D).
The peptide treatment upregulated COX2, EP3, EP4 and MMP9 gene expression in the HaCat cell line and also upregulated EP4 and MMP9 gene expression in (D) cytotoxicity. P < 0.05 was considered significant; one symbol, P < 0.05; two symbols, P < 0.01; three symbols, P < 0.001: * vs HaCaT, # vs HeLa; ANOVA followed by Bonferroni's test. Assays were performed with three independent experiments. Error bars indicate SD. Scale bars: 500 lm.
the HeLa cell line. The combination of the secreted factors from endothelial cells with Ac2-26 treatment upregulated MMP2 gene expression in HaCat cells, and downregulated COX2, EP3 and EP4 gene expression, but upregulated MMP2 gene expression in HeLa cells (Fig. 6B).

Discussion
Cervical cancer is related to HPV infection, and microbial infection is associated with 20-25% of the cancer cases worldwide [35]. It has been observed that the start of the carcinogenesis is more associated with persistent inflammation, as a result of infections or autoimmunity [36]. Besides the infections, the tumour microenvironment also plays a role in the tumourigenesis, due to its complex structure, in which different factors may influence and modify the cancer [37][38][39], and our experiments demonstrate this heterogeneity.
Our results demonstrate that cervical cancer cells exhibit a decrease in growth when treated with the Ac2-26 peptide, and are stimulated by the conditioned medium of endothelial cells. This is important for our understanding of the role of this protein in cancer. These conditions also result in an increase in motility after 120 h, and the maintenance of viability, the stimulation becoming cytotoxic only at 48 h. Regarding apoptosis and gene expression, an increase in apoptotic cells, the downregulation of COX2, EP3 and EP4, and the upregulation of MMP2 were observed. The pro-and anti-inflammatory effects of ANXA1, a protein known to be an endogenous glucocorticoid, depend on the type of ligand, and have been described previously [12]. ANXA1 protein was demonstrated to initiate the p38 mitogen-activated protein kinase (MAPK) pathway and the Ac2-26 peptide promotes the initiation of the c-Jun N-terminal protein kinase (JNK) pathway [40]. The MAPK family pathway is one of the most important cascades responsible for reading an extracellular stimulus and unleashing a response in the cells, and one of these pathways is the JNK pathway, a signalling cascade known to be activated by stress [41]. The JNK cascade may act by phosphorylating different molecules, among them p53 [42]. Through this action, we could suggest that the proliferation decrease after the Ac2-26 treatment and HMC stimulation may be through activation of the MAPK family and targeting of p53, resulting in cell cycle arrest. The proliferation decrease found in this study is corroborated by refs [26,27], which also observed this phenomenon in cancer cell lines after treatment with ANXA1. One of the authors indicated that ANXA1 may act as a tumour suppressor, possibly acting on the extracellular signal-regulated kinases 1 and 2 pathway, also a signalling cascade from the MAPK family [27]. This response may result in overexpression of this pathway, which modifies the expression of the molecules involved in the cell cycle, resulting in arrest in the G1/S phase [43]. This adds to the suggestion that the ANXA1 protein may have an anti-proliferative role in the cervical cancer cells through the MAPK family pathway, as observed after the peptide treatment and stimulation with HMC.
The cytotoxicity assay demonstrated that the cells were as viable as the healthy control cells. Therefore, although the proliferation decreased after stimulation with the HMC and peptide treatment, the cells remained viable, making it possible to proceed with the cellular processes. There is an association between nuclear factor-jB and ANXA1, suggesting that Annexin A1 protein activates this pathway promoting metastasis [44], which could be a result of the maintenance of motility.
The apoptosis analysis showed that the majority of the cancer cells were viable, but some were in apoptosis and late apoptosis. The cervical cancer apoptosis may be altered because of HPV18 infection. This virus has E6 and E7 domains, which express oncoproteins, and they are involved with p53 and retinoblastoma degradation [45][46][47]. The E6 protein, specifically, can form a complex that causes p53 polyubiquitination and its degradation by the proteasome [45,48]. The p53 gene is also associated with the control of apoptosis by regulating the expression of two genes, B-cell lymphoma 2 (BCL2; anti-apoptotic) and BCL2-associated X apoptosis regulator (BAX; pro-apoptotic) [49]. ANXA1 has a role relating to apoptosis regulation according to [11], and some research groups suggest the involvement of this protein as a pro-apoptotic factor, whereas other groups indicate its role as an anti-apoptotic factor. The activation of the MAPK pathways could also have an important role in the apoptotic process, and a range of anti-cancer substances activate the MAPK cascade with the aim to promote apoptosis of the cancer cells [41]. Although there were a greater number of viable cells, we could observe an increase in apoptotic cells after the peptide treatment and stimulation with HMC.
There are studies with ANXA1 showing its pro-and anti-inflammatory responses and also associating its expression with metastasis [50]. During the invasion process, it is suggested that a cell population, called 'leader' cells, invade the adjacent matrix, opening space for following cells [51]. It is necessary that the 'leader' cells secrete factors in order to initiate the process, which will provide a path for the other cells; one of these factors is matrix metalloprotease [52]. Boudhraa et al. [53] have observed that ANXA1 shows expression patterns related to specific tumour types, and they have associated the protein cleavage with melanoma cell aggressiveness. The same group showed that in melanoma cells, the administration of Ac2-26 peptide activates formyl peptide receptors and MMP2 gene expression [54]. In our current work, the level of MMP9 gene expression equalled the control, after HMC stimulation and peptide treatment, and the upregulation of MMP2 was also observed in the healthy HaCaT cells; therefore this pattern could be due to the secreted paracrine factors of the endothelium, possible as an attempt to restore a healthy environment.
Pro-carcinogenic functions have been attributed to the COX2 gene, supporting the strategy of gene suppression; some of the processes related to COX2 are apoptosis inhibition, proliferation increase, and induction of angiogenesis [55][56][57]. The available literature shows that this gene may be triggered by different factors, for instance, bacterial lipopolysaccharide, interleukin (IL)-1b, IL-2, tumour necrosis factor a, epidermal growth factor and transforming growth factor b; and it may be blocked by anti-inflammatory factors, such as corticosteroids, IL-13, IL-10 and IL-4 [55,58,59]. In this work, it was possible to observe that the peptide treatment and HMC stimulation downregulated COX2 gene expression in the cervical cancer cells. Prostaglandin E2 (PGE2) is one of the most abundant COX2 products due to the fact that PGE2 exists in all cell types [55]. Besides stimulation of pain and inflammation, PGE2 participates in proliferation, apoptosis, and metastasis mechanisms [60,61]. PGE2 initiates its actions through specific G-protein-coupled membrane receptors, namely EP1, EP2, EP3 and EP4, with each of them initiating a specific cellular pathway [62]. The EP2 and EP4 receptors stimulate adenylate cyclase to increase cyclic adenosine monophosphate (cAMP) production, a glycogen synthase kinase-3 activator involved in the cell cycle [55]. cAMP promotes b-catenin phosphorylation and degradation by the proteasome 26S, stopping the cell cycle [55]. The EP3 receptor has the opposite action to that described above; its interaction with PGE2 leads to the downregulation of adenylate cyclase and cAMP decrease [60].
The culture of HeLa cells with HMC stimulation showed that the treatment did not alter cell motility, but it did result in a decrease in cell proliferation, and in COX2, EP3 and EP4 gene expression. The remaining production of PGE2 could be acting through the EP4 receptor, culminating in a decrease of cellular growth after treatment, which could be interfering with apoptosis, since a lower number of cells was observed in this cellular process.
Our research studied for the first time the interaction of ANXA1 with the conditioned medium of endothelial cells, an attempt to simulate the tumour microenvironment. It has been possible to observe the likely existence of a complex interaction, and the diverse array of secreted factors may influence cellular and molecular mechanisms in different ways. The diverse types of tumours exhibit different biological roles and signalling pathways when ANXA1 is involved, and this contradiction is explained by the differential expression patterns and biological behaviours of this protein [11]. In addition, calcium binding to ANXA1's C-terminal core, which binds to the phospholipids, may determine the expression specificity and promote its functional diversity [16].
We suggest that along with the tumour cell secreted paracrine factors, ANXA1 diminishes proliferation and does not alter the cellular cytotoxicity. It was possible to observe an increase in apoptotic cells and a downregulation of COX2, EP3 and EP4. Altogether, these data provide further insight into the protective effect of ANXA1 and its mimetic peptides in cervical tumourigenesis, and more research needs to be carried out to confirm its use.
We did all the experiments with another cervix cell line (SiHa -ATCC, Manassas, VA, USA) treated with the peptide, but without the co-treatment with conditioned medium of endothelial cells (HUVECs), and the results were the same with the HeLa; for this reason we did not continued with the co-treatment.

Cellular co-culture
The co-culture was performed with the conditioned medium of the HUVEC (HMC) cell line, which was used in the cocultures with the other two cell lines. The HMC was obtained after HUVEC culture when the cells had acquired 80% confluence. The medium (HMC) was collected after 24 h of culture, from the HUVEC cell line medium without treatment (HMCS) and with treatment (HMCT). The treatment was performed with the N-terminal Ac2-26 peptide of ANXA1 (Ac-MVSEFLKQAWFIENEEQEYVQTVK) [63], at a concentration of 10 lgÁmL À1 [27]. The co-culture was performed at a 1 : 1 dilution, according to Rodrigues-Lisoni et al. [64].

Pharmacological treatment
The cell lines were cultured in complete medium, as described above, and subsequently, submitted for co-culture with HMC and treated with the N-terminal Ac2-26 peptide of ANXA1 (Ac-MVSEFLKQAWFIENEEQEYVQTVK) [63], at a concentration of 10 lgÁmL À1 for six different time points (2,4,24,48,72 and 120 h) to perform the proliferation, wound healing, and cytotoxicity assays. After the analysis of these experiments one time point was chosen to continue the other assays. The experimental groups are described below: HaCaT HeLa HaCaT + Ac2-26 HeLa + Ac2-26 HaCaT + HMCS HeLa + HMCS HaCaT + Ac2-26 + HMCT HeLa + Ac2-26 + HMCT This study was performed to analyse the response of Ac2-26 peptide treatment in cancer cell line, and in that way we developed the experiments only with the cells and the peptide. After analysing the results and in an attempt to mimic the tumour microenvironment we added the conditioned medium of endothelial cells (HMCS and HMCT), along with the peptide treatment, and observed if in these conditions the peptide would exhibit different results.

Proliferation assay and cellular morphology analysis
To analyse cell proliferation in the HaCaT and HeLa cells, a growth curve was performed. To count the number of cultured cells, they were seeded at a concentration of 3 9 10 4 in 1 mL of complete medium for 24 h. After this period, the medium was replaced with a serum-free medium, with the purpose of maintaining the same cellular phase. After a further 24 h, this medium was replaced again with complete medium, with the addition of specific HMC and Ac2-26 peptide, according to the experimental groups. The cells were analysed and counted at six different time points (2,4,24,48,72 and 120 h). The cellular morphology was evaluated with inverted microscopy, using an Olympus CKX41 (Olympus, Tokyo, Japan).

Wound healing assay
The HaCaT and HeLa cells were distributed in 12-well plates; after reaching adherence and confluence, a wound was made in the centre of the well. Subsequently, the cells were subdivided according to the experimental groups and analysed at 0, 2, 4, 24, 48, 72, and 120 h. The cellular motility was monitored with images obtained using a photographic camera coupled to the microscope. The wound areas were determined in six different microscopic fields and quantified using 'IMAGEJ software (National Institute of Health -NIH, Bethesda, MD, USA).

Flow cytometry
The cells were subdivided into the experimental groups and handled according to the manufacturer's protocol for the FITC Annexin V Apoptosis Detection Kit I (BD Pharmingen TM , São Paulo, SP, Brazil), to investigate viability, apoptosis, late apoptosis and necrosis, and utilised Guava Easy Cyte (Millipore, Barueri, SP, Brazil) equipment to perform the analysis.

Selection of the genes
The genes were previously selected from potential markers in the tumour microenvironment, evaluated in the studies of [64,65]. The metabolic pathways from these potential markers were evaluated and some related genes were also chosen. The gene selections took into consideration those that were potentially involved in the tumour invasion and inflammatory processes.

Quantitative real time PCR
The reactions were performed in the 7500 Fast Real-Time PCR system thermocycler (Applied Biosystems, Foster City, CA, USA); all the reactions were carried out in a final volume of 20 lL with 100 ng of cDNA, SYBR Ò Green PCR Master Mix, and 100 nM of each primer (F and R). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was the best endogenous control gene when tested alongside actin beta (ACTB). The analyses were performed from the cycle threshold (C t ) of each sample according to [66].

Statistical analysis
All assays were performed with three independent experiments. PRISM version 6.0 (GraphPad Software Inc., San Diego, CA, USA) was used to perform all statistics. The Kolmogorov-Smirnov normality test was used to analyse distribution. Variance analysis (ANOVA) was used for between group comparisons, followed by the appropriate parametric or non-parametric post-hoc test. Statistical significance was set at a probability value of less than 0.05. Table 1. Primers for the studied genes.