Activation of EGFL7 within High Grade Ovarian Serous Carcinoma is Associated with Lower Tumor Infiltrating CD4+ and CD8+ Lymphocytes

It has been suggested that Epidermal Growth Factor like domain 7 (EFFL7) promotes tumor cell escape from immunity by downregulating the activation of tumor blood vessels. Aim: to analyze mRNA expression of EGFL7 within tumor microenvironment of high-grade ovarian serous carcinoma and its association with number of intraepithelial CD4+/CD8 + lymphocytes and expression of molecule involved in diapedesis ICAM-1. Methods: qPCR analysis of EGFL7 mRNA in cancer cells and adjacent stromal endothelium microdissected from formalin-fixed paraffin-embedded tumors of 59 high-grade ovarian serous carcinoma patients, was performed. Infiltration of intraepithelial lymphocytes (CD4+/CD8+) and expression of ICAM-1 were evaluated by immunohistochemistry and compared between tumors with different status of EGFL7 expression. Results: EGFL7 was expressed in cancer cells (9/59, 15.25%), endothelium (8/59, 13.56%), or both cancer cells and adjacent endothelium (4/59, 6.78%). ICAM-1 was expressed on cancer cells (47/59, 79.66%), stromal endothelium (46/59, 77.97%), or both epithelium and endothelium (40 of 59, 67.8%). EGFL7 -positivity of cancer cells and endothelium was associated with lower intraepithelial inflow of CD4+ (p = 0.022 and p = 0.029, respectively) and CD8 + lymphocytes (p = 0.04 and p = 0.031, respectively) but impact neither epithelial nor endothelial ICAM-1 expression (p = 0.098 and p = 0.119, respectively). Patients’ median follow-up was 23.83 months (range 1.07–78.07). Lack of prognostic significance of EGFL7 -status and ICAM-1 expression was notified. Conclusion: EGFL7 was independently activated in the cancer epithelium


Abstract
It has been suggested that Epidermal Growth Factor like domain 7 (EFFL7) promotes tumor cell escape from immunity by downregulating the activation of tumor blood vessels. Aim: to analyze mRNA expression of EGFL7 within tumor microenvironment of high-grade ovarian serous carcinoma and its association with number of intraepithelial CD4+/CD8 + lymphocytes and expression of molecule involved in diapedesis ICAM-1.

Background
Ovarian cancer (OvCa) is the leading cause of deaths from gynecological malignancy in the developed world. The overall five-year survival rate for patients with advanced OvCa is only 30-40% [1].
Notwithstanding the good initial response to primary therapy, around 80% of the patients with advanced disease will develop recurrence and finally succumb to the illness [1]. Therefore, more insight in the interaction between OvCa and the immune system is needed to develop more effective anti-tumor immunotherapy to improve the clinical outcome.
Instead of being targeted for immune destruction, OvCa has ability to escape the immune surveillance by creating a highly suppressive environment [2]. Tumor escape from immunity is undoubtedly a promising phenomenon to consider while designing therapeutic tools aimed at preventing cancer progression and metastasis [3,4].
Blood vessels, which are a part of the tumor microenvironment, constitute a natural physical barrier which regulates the immune response via diapedesis [5]. Activated endothelial cells express high level of cell surface leukocyte adhesion molecules such as E-and P-selectin, intercellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1) which participate in the capture of circulating immune cells through the rolling, arrest, firm adhesion, and extravasation of leukocytes into stroma [5,6]. Thus, immune escape may be achieved by limiting the influx of immune effectors into cancer tissue [7,8] through down-regulation of endothelial adhesion molecules ICAM-1 and VCAM-1 [9][10][11][12].
Using experimental approaches, it has been demonstrated that, when expressed by cancer cells, EGFL7 promotes tumor escape from immunity by downregulating the activation of tumor blood vessels [17]. EGFL7 in mouse breast and lung carcinoma cells inhibit the endothelial expression of ICAM-1 and VCAM-1, preventing lymphocyte adhesion and transendothelial migration [17]. This observation was additionally confirmed by an analysis of tissue samples derived from large cohort of breast cancer patients, utilizing hybridization in situ as a method to localize a real source of EGFL7 secretion [18].
Since EGFL7 is plasma membrane-associated sialidase (lysosomal sialidase) which could accumulate at distant sites from the producing cells [15], it is impossible to answer if elevated expression of The patients were informed that surgical specimens would be used for conventional pathological diagnosis and that the remaining tissues would be used for research. No personal patient data was required for this study and the protocols carried no risk, so only verbal informed consent was required from the patients.

Immunohistochemistry
The immunohistochemical stainings were performed on whole sections of representative FFPE blocks.
The staining was performed according to the following protocol (antibody details, suppliers and dilutions are listed in Table 1). Four µm-thick serial sections were cut, deparaffinized and subjected to a heat-induced epitope retrieval step before being incubated with the primary antibodies. Sections were immersed in Target Retrieval Solution (pH 6,0; Dako Cytomation, Denmark) and heated in a pressure cooker. The slides were incubated for 90 minutes with the primary antibodies. The reaction was visualized by the Novolink polymer detection system (Novocastra Laboratories). Appropriate positive (tonsil for both: CD4 and CD8, placenta for ICAM) and negative controls (the primary antibody was replaced with normal mouse IgG at an appropriate dilution) were included for each case. The results of immunohistochemistry were evaluated by two independent pathologists who were blinded for the clinical data.

Statistical analysis
To determine statistically significant differences between the variables, the Mann-Whitney U test was used. The statistical analysis was performed using chi-square test or Fisher's exact probability test.
Correlations and differences between variables were assessed using the Spearman's rank correlation coefficient. Overall survival curves were estimated by the Kaplan-Meier method and compared with use of the two-sided log-rank test. P-values of < 0.05 were regarded as significant in each of the analysis. All analyses were performed with the statistical software Statistica 13 (Stat Soft Inc.).

Patient population
The median age of the patients was 59 years (range 39-86), the median duration of follow-up was 23.83 months (range 1.07-78.07). The 5-year disease free survival (DFS) rate was 20.35%. The clinicopathological HGOSC data are presented in Table 2. We noted the increase in the PCR product only in a low proportion of samples, therefore, the initially planned quantitative analysis was not possible and the qualitative scoring had to be applied. In Correlation of clinicopathological HGOSC population data and EGFL7 expression of cancer cells and endothelium are presented in Table 3 and Table 4, respectively. Table 3 The clinicopathological characteristics of HGOSCs population related to Egfl7 and ICAM-

ICAM-1 expression
We observed that ICAM-1 was expressed by both cancer and endothelial cells (Fig. 1C). ICAM-1 expression in cancer cells was noted in 47 cases (79.66%) and in endothelial cells in 46 cases (77.97%); in 40 cases (67.8%) the expression was observed in both compartments. ICAM-1 expression in cancer cells was score with High-score method [21] with the median expression level of 20 (range 0 to 200). Endothelial ICAM-1 expression was calculated as the proportion of positive vessels (median 18.75%, range 0%-94.12%) and then scored semi-quantitatively with a 4-tier system using the thresholds of 1%, 30% and 60% [18].  Table 2 and Table 3, respectively. We observed no differences in OS or PFS between high and low ICAM-1 expression in either compartment (p > 0.05).
We observed no significant association between EGFL7 and ICAM-1 expression in either compartment were also similar in both groups (Fig. 2).

Discussion
EGFL7 expression has been previously analyzed in one study on human ovarian cancer tissue samples derived from 177 patients [22]. In that study immunohistochemical staining for EGFL7 was performed using formalin fixed paraffin-embedded tissue microarrays. 72 of 177 analyzed cases (40.7%) had tumors with serous histology, but information about grading was not specified [22]. High levels of EGFL7 expression were noted in 23 of 72 (31.9%) serous cancers. Survival analysis performed for the entire cohort showed that the epithelial ovarian cancer patients having tumors with high EGFL7 expression had a poorer DFS but similar OS to those with low EGFL7 expression [22].
Here, we analyzed EGFL7 mRNA transcripts in HGOSC using different methodology, thus, the results of both studies are difficult to compare.
Interestingly, our study supports the suggestion that EGFL7 expression is frequent in epithelial ovarian cancers as we detected EGFL7 mRNA transcripts in 13 of 59 (22.03%) HGOSC cases.
Unfortunately, we did not manage to perform quantitative analyses of mRNA in microdissected tissue samples, thus, the impact of intensity of expression of EGFL7 mRNA on OS and PFS was not assessable. However, EGFL7-status of tumor microenvironment (cancer epithelium and adjacent endothelium) had no predictive and prognostic value in analyzed cohort of HGOSC patients.
EGFL7-expression levels were found to correlate with a higher tumor grade in gliomas [23] and colon cancer [24], not influencing the prognosis. However, it conferred poorer prognosis and higher metastatic score in hepatocellular carcinoma [25], but better prognosis and absence of lymph node invasion in human breast cancer [26]. These ambiguous results indicate that EGFL7 expression in human cancer needs to be carefully analyzed as EGFL7 may play a complex role in cancer biology depending on cancer origin [23][24][25][26] and the source of EGFL7 secretion: cancer cells, endothelium or both.
Since it had been proven that EGFL7 is plasma membrane-associated sialidase, which could accumulate in distant tissues from the producing cells [15], it is impossible to answer if expression of EGFL7 observed in OvCas [22] comes from activated cancer epithelium or up-regulated endothelial cells in the surrounding stroma. Therefore, we had decided to analyze the impact of EGFL7 on diapedesis using LCM technology, which enables to harvest the epithelial and endothelial cells directly to give histologically enriched cell populations for separate qPCR analyses [19].
We expected to detect EGFL7 mRNA at least in endothelial cells as these reflect upregulation of EGFL7 gene in activated cancer vessels due to neoangiogenesis and/or inflammatory process.
Presence of EGFL7 mRNA transcripts found in microdissected cancer cells indicates the endogenic activation of EGFL7 gene in human HGOSCs. To the best of our knowledge this is the first study documenting activation of EGFL7 in human HGOSC. Interestingly the frequency of epithelial and endothelial activation of EGFL7 was comparable. Further, ICAM-1 was also found to be expressed both on cancer epithelium and endothelium of adjacent stromal vessels thus we decided to assess the correlation between EGFL7 and ICAM-1 matched according the source of origin.
Although it was suggested that EGFL7 negatively regulates the expression of ICAM-1 in endothelial cells [27] we did not find the difference in the intensity of ICAM-1 expression between endothelium with different EGFL7-status. Similarly lack of difference of ICAM-1 expression was observed for cancer ICAM-1 expressing cells with different Egfl7-statuses.
Our understanding of the role of ICAM-1 in ovarian cancer development remains limited. The elevated expression of ICAM-1 in freshly isolated ovarian cancer cells [28] suggest that ICAM-1 expression may promote the malignant progression of ovarian cancer. Findings of studies on other malignant tumors suggest that ICAM-1 contributes to carcinogenesis by at least two mechanisms. It mediates the accumulation of inflammatory cells, which facilitates the instability of the tumor environment, triggers tumorigenesis, and maintains the release of trophic factors to enhance cancer cell survival [29,30].
These observations were not confirmed for our cohort of HGOSC patients as we found that cancer-ICAM-1 expression was not correlated with FIGO stage and lacked prognostic significance.
To assess the potential impact of EGFL7 on diapedesis we decided to compare intraepithelial immune infiltrates between tumors having EGFL7 activated in cancer nests and/or adjacent vessels and those without EGFL7 activation.
Presence of EGFL7 mRNA transcripts in epithelium and/or endothelium of tumor microenvironment was associated with a lower influx of adaptive immune effectors (CD4 + and CD8 + lymphocytes) into cancer nests. This provides direct information that in some part, immune escape is achieved by limiting the influx of immune effectors into cancer tissue by activation of EGFL7. The role of epithelial and endothelial EGFL7 should be further evaluated in HGOSCs.
The weakness of the current study is the lack of quantitative assessment of EGFL7, the retrospective

Consent for publication: not applicable
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Competing interested: The authors declare that they have no competing interests.

Figure 2
Categories of endothelial-ICAM-1 expression in relation to Egfl7 status of endothelial cells

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. CoverLetterEgfl7JOCR.docx