Peritumor mucosa in advanced laryngeal carcinoma exhibits an aberrant proangiogenic signature distinctive from the expression pattern in adjacent tumor tissue

doi:10.21203/rs.3.rs-3453893/v1

Download PDF

Research Article

Peritumor mucosa in advanced laryngeal carcinoma exhibits an aberrant proangiogenic signature distinctive from the expression pattern in adjacent tumor tissue

https://doi.org/10.21203/rs.3.rs-3453893/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Field cancerization theory is an important paradigm in head and neck carcinoma as its oncological repercussions affect treatment outcomes in diverse ways. The aim of this study is to assess the possible interconnection between peritumor mucosa and the process of tumor neoangiogenesis. Sixty patients with advanced laryngeal carcinoma were enrolled in the study. The majority of patients express a canonical HIF-upregulated proangiogenic signature with almost complete predominancy of HIF-1α overexpression and normal expression levels of the HIF-2α isoform. Remarkably, more than 60% of the whole cohort exhibit a HIF-upregulated proangiogenic signature also in peritumoral benign mucosa. Additionally, the latter subgroup has a distinctly shifted phenotype towards HIF-2α upregulation compared to the one in tumor tissue, i.e., a tendency towards a HIF-switch is observed in contrast to the dominated by HIF-1α tumor phenotype. ETS-1 displays stable and identical significant overexpression in both proangiogenic phenotypes present in tumor and peritumoral mucosa. In the current study, we report for the first time the existence of an abnormal proangiogenic expression profile present in the peritumoral mucosa in advanced laryngeal carcinoma when compared to paired distant laryngeal mucosa. Moreover, we describe a specific phenotype of this proangiogenic signature that is significantly different from the one present in tumor tissue as we delineate both phenotypes, quantitively and qualitatively. This finding is per se cancer heterogeneity that extends beyond the “classical” borders of the malignancy and is proof of a strong interconnection between field cancerization and one of the classical hallmarks of cancer – the process of tumor neoangiogenesis.

field cancerization

laryngeal carcinoma

HIF

HIF switch

HIF-1α

HIF-2α

HIF-3α

VEGF-A

ETS-1

miR-210

Laryngeal carcinoma is a worldwide burden and despite advances in individualization of treatment for a number of other solitary malignancies, curative treatment options in this subdomain of the oncology remain limited to surgery and/or chemoradiation therapy based on anatomic location and TNM staging [1, 2]. One of the main theories for oncogenesis and recurrences in head and neck cancer is the field cancerization theory proposed by Slaughter [3]. According to it, on a molecular level, some processes of cell dysregulation characteristic of the malignancy are evident in clonal patches of the surrounding histologically normal mucosa around the tumor, and presumably, both have a common predecessor. Our team has already reported the existence of an aberrant microRNA signature in the peritumoral tissue surrounding the advanced laryngeal carcinoma that could be used as a biomarking tool for demarcating the field cancerization [4]. In the current study, we report for the first time in the literature the existence of an abnormal proangiogenic expression profile present in the peritumoral mucosa in advanced laryngeal carcinoma when compared to paired distant laryngeal mucosa. Moreover, we describe a specific phenotype of this proangiogenic signature that is significantly different from the one present in tumor tissue as we delineate both phenotypes, quantitively and qualitatively.

Study design and treatment protocols

A total of sixty patients (mean age at diagnosis 64.6 ± 8.7 years) with pathologically verified primary laryngeal carcinoma were recruited into the current prospective prognostic study. The diagnosis was made based on current standards. The patients were admitted to the Department of ENT, Head & Neck Surgery, Medical University – Sofia, during the years 2018–2019 when underwent primary laryngectomy. During surgery four samples from each patient were obtained: tumor site –surface, tumor site-depth, histologically healthy peritumor mucosa within 1 cm from the border of the tumor, and paired normal laryngeal mucosa distant to the tumor (contralateral, at least 3 cm distance). All samples were stored in RNAlater Solutions (Thermo Fisher Scientific, Massachusetts, USA) and frozen at -20°C for a short period until were transported to the Molecular Medicine Center, Department of Medical Chemistry and Biochemistry, Medical University – Sofia, and maintained at -80˚C until use.

All patients signed an individual consent form. The study was approved by the Ethics Committee of Medical University - Sofia. The enrolled cohort was a single-surgeon consecutive series and the inclusion criteria were as described in a previous study [4]. All patients underwent postoperative radiotherapy or combined chemoradiotherapy according to the protocol.

RNA extraction

Total RNA (including miRNAs) was isolated from 60 fresh-frozen tumor materials and adjacent normal tissue using the miRNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. The RNA concentration was determined with NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Massachusetts, USA). The Qubit™ RNA HS Assay Kit, Qubit™ 2.0 Fluorometer (Life Technologies, Thermo Fisher Scientific Inc.) was used for precise quantification of 50 ng/µl and 100 ng/µl RNA dilutions used further for miRNA array and RT-qPCR analysis, respectively. The RNA integrity numbers (RIN) were measured using the 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA) as described in a previous study [4].

Real-time quantitative polymerase chain reaction (RT-qPCR)

Quantitative real-time PCR was performed in a 7900HT Fast Real-Time PCR System machine (Applied Biosystems, California, USA) with primers for the selected mature RNAs HIF-1α, HIF-2α, HIF-3α VEGF-A, VEGFR1, VEGFR2, ETS-1 and miR-210, as QuantiTect Primer Assay and miScript Primer assays, respectively, designed by Qiagen (Qiagen, Hilden, Germany). Reactions were performed in triplicate with a total volume of 10 µl volume, and the mixture for mRNAs and miRNA included respective SYBR Green PCR Mix (Qiagen, Hilden, Germany), Primer Assay (Qiagen, Hilden, Germany) for the respective genes and miRNA and cDNA synthesized from 400 ng of total RNA using the miScript II RT Kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. Sample data were normalized to the β-actin for all mRNAs and to the U6 snRNA level for miR210 (as an internal controls). Negative and no-template controls were also evaluated. The relative quantification (RQ) of mRNAs as well as miRNAs in samples was analyzed by the 2-ΔΔCt method, as previously described [4].

Statistical analysis

SPSS software v.23.0 for Windows (IBM SPSS, NY, USA) and GraphPad Prism software were performed for data analysis. The Kolmogorov-Smirnov test for normality and the Wilcoxon test were used as appropriate. Spearman’s correlation test was used to analyze the relationship between two continuous variables. Statistical significance was defined as a two-tailed p-value < 0.05.

The study group included predominantly HPV-negative tumors (90.3%), validated with p16 immunohistochemistry, with all patients having a history of long-term smoking. In terms of tumor staging, the majority of the cases (87.1%) were classified as pT4a, one case was pT4b, and the rest were staged as pT3. Almost half of the group, 48.3% of the cases, had pathologically verified metastatic processes, and the distribution of N status was as follows: N1 (28.6%); N2a/N2b/N2c (57.1%), N3 (14.3%).

The majority of patients express a canonical HIF-upregulated proangiogenic signature with almost complete predominancy of HIF-1α overexpression and normal expression levels of the HIF-2α isoform, i.e., HIF-switch is not present in advanced laryngeal cancer.

More than half of patients (56%) with advanced laryngeal carcinoma presented with HIF-1α upregulation (RQ > 2) and in only 10% of the cases HIF-2α overexpression was evident [Figure 1]. Additionally, when a direct comparison of RQ expression levels of both isoforms was performed, unambiguously in 93,3% of all tumor samples HIF-1α had higher expression levels in comparison to its second isoform. This data was obtained with the Wilcoxon ranking test, which undeniably showed significantly higher expression levels of HIF-1α vs. HIF-2α (p < 0.001, z = 6.35). In summary, in advanced laryngeal cancer we observed a canonical proangiogenic phenotype with significant upregulation of HIF-1α but not HIF-2α, i.e., HIF-switch was not evident.

Remarkably, the majority of patients (60%) exhibit a HIF-upregulated proangiogenic signature also in peritumoral benign mucosa. Additionally, the latter has a distinctly shifted phenotype towards HIF-2α upregulation compared to the one in tumor tissue, i.e., a tendency towards a HIF-switch is observed.

The analysis of the samples from the histologically normal mucosa adjacent to the tumor tissue revealed that in 60% of all cases, there was an aberrant proangiogenic overexpression of either or both HIF-1α/ HIF-2α when compared to paired distant laryngeal mucosa along with other classical molecules from the canonical proangiogenic cascade. Moreover, when analyzed, the expression pattern of those molecules differentiated significantly from the pattern evident in paired tumor samples. First, despite still having upregulation of HIF-1α in 40% of cases (versus 56% in tumor samples), its RQ expression levels were significantly lower in peritumoral tissue when compared to tumor tissue (Wilcoxon test, p = 0.03, z = 2.16) [Figure 1]. Even more pronounced was the change in HIF-2α expression – we detected significantly higher levels of HIF-2α expression in peritumoral mucosa when compared to paired tumor samples (Wilcoxon test, p < 0.001, z = 5.44). Thus, we saw two distinct abnormal proangiogenic signatures in the tumor and the paired adjacent normal peritumoral mucosa – in tumor samples we recognize the leading role of the overexpressed HIF-1α, whereas in peritumor mucosa there is an evident shift towards HIF-2α upregulation and lowering of the expression of the first isoform. Ergo, there is a shift towards a “HIF switch” in peritumoral mucosa in contrast to tumor tissue, where HIF-1α holds the predominant role [Figure 2]. Additionally, the other major molecules from this cascade had significantly higher expression in peritumoral mucosa when compared to paired distant control laryngeal mucosa: VEGF-A overexpression was displayed in 40% of patients, ETS-1 in 42%, VEGFR1 in 36.7% and VEGFR2 in 41.7%. Interestingly, VEGF-A expression levels in peritumor tissue were significantly lower than those in tumor tissue (p < 0.001, z = 3.931) while VEGFR2 levels were significantly higher in peritumoral mucosa in comparison to tumor tissue (p < 0.001, z = 4.56) [Figure 1].

HIF-3α expression levels are significantly higher in peritumoral mucosa as part of the HIF-switch but no definite association can be established with the other proangiogenic molecules.

HIF-3α is a molecule that was massively silenced in 78.3% of the tumor samples (RQ < 0.5). In peritumoral mucosa samples, we saw a statistically significant rise in expression levels of HIF-3α when compared to paired tumor samples (Wilcoxon test, p = 0.006, z = 2.75), despite being still largely downregulated (55%) in relation to the paired control samples (RQ < 0.5) [Figure 1]. In both tumor and peritumoral mucosa when we analyze correlations with the other molecules we obtain p-values beneath 0,05 (Pearson’s correlation) but visual evaluation of the scatterplot does not validate those results since the relationship is not monotonic [Figure 3, last column].

Mir-210 is overexpressed in tumor tissue but does not show a well-established association with any of the main proangiogenic genes including HIF-1α

Among the listed differences in proangiogenic expression pattern between tumor and peritumoral mucosa aligns also the expression of angiomir miR-210. In tumor tissues, we recorded that almost half of the patients displayed overexpression of miR-210 (48.3%) that was in contrast to only 16.6% overexpression rate of this molecule in peritumoral mucosa. Additionally, pairwise comparison with the Wilcoxon rank test revealed significantly higher levels of expression of miR-210 in tumor tissue compared to peritumoral mucosa (p < 0.001, z = 5.03) [Figure 1]. Contrary to expectations and literature data, when correlation tests were run between miR-210 and HIF-1α, HIF-2α, VEGF-A, VEGFR1, and VEGFR2, scatterplots visual evaluation concluded that there was no clear monotonic relationship between the expression levels of this microRNA and the other proangiogenic molecules, despite obtaining certain p-values beneath 0.05 [Figure 3, first row]

ETS-1 displays stable and identical significant overexpression in both proangiogenic phenotypes present in tumor and peritumoral mucosa. Correlations between proangiogenic molecules in both sample groups are matching but tend to differ in terms of strength.

ETS-1 molecule was overexpressed to a similar extent in both tumor and peritumoral mucosa (45% vs. 42%) and no statistically significant difference was reported. The investigated classical proangiogenic molecules in tumor tissue displayed strong correlations with each other that can be seen in detail in the scatter dot matrix in Fig. 3. We additionally analyzed the correlations of those molecules from their expression profile in peritumor tissue and in terms of significance they did not differ widely from those in the tumor samples. Nevertheless, when comparing the strength of association, one could undoubtedly distinguish the difference. To compare the strength of association among these molecules, we created two heatmaps with the correlation coefficients (Spearman's ρ (rho) or r_s) given in Fig. 4. Firstly, we identified that in contrast to tumor tissue, where we found a moderate correlation between HIF-1α and HIF-2α, such an association was not valid in peritumoral tissue. ETS-1 had a far stronger association with VEGF-A in peritumoral tissue compared to tumor tissue (r_s= 0.32 vs 0.63). HIF-1α/ HIF-2α correlated at a far greater extent with the two VEGF receptors and ETS-1 in tumor tissue compared to peritumor tissue but kept their level of association with VEGF-A identical in both sample groups. Additionally, the association between VEGF-A and VEGFR1 displayed a higher correlation coefficient than VEGF-A and VEGFR2 in tumor tissue, and this relation was inverted in peritumoral mucosa [Figure 4].

Several studies investigated and displayed the molecular dysregulation on a cellular level of the histologically “normal” peritumoral tissue bearing identical genetic and epigenetic traits with its adjacent malignancy due to their probable common monoclonal origin [5]. The current study is the first to our knowledge to address the possibility that peritumoral mucosa could express an aberrant proangiogenic signal. The results from the mRNA expression analysis of the main proangiogenic genes undoubtedly reveal the existence of such a pathological activity [Figure 1].

A well-known fact is that in normoxia HIFs are being degraded via the pVHL-E3 ubiquitin ligase complex whereas in a hypoxic environment, HIF-1α/ HIF-2α stabilize and translocate to the nucleus where they induce their effect by binding to the hypoxia responsive elements (HREs) [6]. As peritumoral mucosa by axiom has a normal supply of oxygen, we could hypothesize two possible explanations for these findings – either there is some unrecognized hypoxia in the peritumor microenvironment or we have intracellular processes of normoxic upregulation of the HIF molecules. The first alternative could be rationalized with the existence of a “steal phenomenon” whereby tumor blood flow could decrease peritumoral perfusion locally which would lead to a certain level of hypoxic microenvironment surrounding the tumor. The second alternative involves assuming the existence of an intracellular non-canonical proangiogenic phenotype. It has been observed that aggressive cancers constitutively express hypoxia-related proteins (HRPs) even in the presence of oxygen, a condition known as pseudohypoxia [7, 8]. Such phenotype is closely related to the so-called “acid-mediated invasion hypothesis” which states that expression of glycolytic proteins at the edge of tumors creates an acidic microenvironment, which could induce apoptosis in stromal cells, resulting in space being created into which tumor cells can proliferate [9]. We could speculate that such changes in the peritumor environment could lead to phenotype changes in the surrounding tissues and result in aberrant phenotypes. Such hypothesis is supported by the fact that we see a strong correlation between expression levels of MMP2 in tumor and peritumor mucosa (p < 0,001, z = 0.528, unpublished data) and acid has been shown to induce the release of matrix metalloproteinases, resulting in degradation of the matrix [8]. Moreover, a number of studies have reported on pseudohypoxic tumor phenotypes that occur due to specific mutations that upregulate HIF-1α and prevent its degradation – e.g., mutations in isocitrate dehydrogenase (IDH) [10], (PTEN), von Hippel–Lindau protein (pVHL), p53, epidermal growth factor (EGF), and mutant Ras and Src [11]. In the light of field cancerization theory, we could speculate that a possible explanation for this aberrant proangiogenic activity in peritumoral mucosa could be the presence of some mutations that are found in both tumor and peritumoral cells due to their common monoclonal ancestry. All those possible explanations should find their answer in future studies in that subfield of oncological research.

Another important aspect of the findings of our study is the difference in the proangiogenic phenotypes between tumor tissue and peritumoral mucosa. Due to the inversion of the HIF-1α/ HIF-2α expression ratio in peritumor samples, we could categorize this phenotype as more shifted towards a “HIF switch” in contrast to the dominated by HIF-1α tumor phenotype. Firstly, our data once again confirms a finding from a previous study of our team that excludes the existence of such a HIF switch in advanced laryngeal carcinoma, a phenotype that is typical in many other malignancies and is characteristic of advanced cancer [13]. This difference in phenotypes is illustrated in Fig. 3 that is an adapted graphic from the magnificent review on HIF switch regulation published by the Rafał Bartoszewski’s team (with permission of the authors) [12]. This variance of the expression levels is per se cancer heterogeneity that extends beyond the “classical” borders of the malignancy. As a consequence, this would inevitably be affecting tumor resistance in terms of treatment. Beneath the shift of the HIF isoform’s expression levels, one could also identify a rearrangement of the association’s strength between the major proangiogenic genes [Figure 4] that could be a consequence of a change in the regulatory framework. An important fact that should be considered is the stability of the expression levels of ETS-1 does not change significantly in contrast to the other major proangiogenic molecules. These findings could position ETS-1 as a more independent gene in terms of regulation of the canonical cascade of neoangiogenesis.

To our knowledge, this is the first comprehensive study that uncovers the existence of a pathophysiological proangiogenic signature in peritumor mucosa in head and neck cancer. Additionally, its pattern of expression of the major proangiogenic molecules is distinctly different in comparison to the tumor proangiogenic signal. This finding is per se cancer heterogeneity that extends beyond the “classical” borders of the malignancy and is proof of a strong interconnection between field cancerization and one of the classical hallmarks of cancer – the process of tumor neoangiogenesis.

Competing interests

The authors declare no competing interests.

Acknowledgments

All authors provided substantial input into the conceptualization, drafting, and editing of this report. Each has given approval for the publication of the article.

Funding: This study is financed by the European Union-Next Generation EU, through the National Recovery and Resilience Plan of the Republic of Bulgaria, project № BG-RRP-2.004-0004-C01, group 3.1.5.

Ethics approval: The Ethical Committee of Medical University—Sofia approved the study and every patient signed written informed consent. The study was performed in accordance with the Declaration of Helsinki.

Authors' contributions:

Popov TM: study design, surgical treatment, sample collection, data analysis, survival analysis, manuscript drafting
Hadzhiev Y, Dobriyanova V: sample collection
Stancheva G: study design, RT-PCR, data analysis, manuscript drafting
Kyurkchiyan SG: RT-PCR, microarray, manuscript drafting
Petkova V: RT-PCR, microarray
Panova S: RT-PCR, microarray
Kaneva RP, Hadzhiev Y: data analysis, draft review
Stancheva I, Dobriyanova V: study design, data analysis, draft review

Data availability statement: Raw data were generated at Medical University - Sofia. Derived data supporting the findings of this study are available from the corresponding author [TP] on request.

Kemmer JD, Johnson DE, Grandis JR. Leveraging genomics for head and neck cancer treatment. J. Dent. Res. 2018;97(6):603–13. doi: 10.1177/0022034518756352.
Valenciano A, Henríquez-Hernández LA, Lloret M, Pinar B, Lara PC. New biological markers in the decision of treatment of head and neck cancer patients. Clin. Transl. Oncol. 2014;16(10):849–58. doi: 10.1007/s12094-014-1193-7.
Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer. 1953;6(5):963–968. doi: 10.1002/1097-0142(195309)6:5<963::AID-CNCR2820060515>3.0.CO;2-Q.
Popov TM, Stancheva G, Kyurkchiyan SG, Petkova V, Panova S, Kaneva RP, Popova DP. Global microRNA expression profile in laryngeal carcinoma unveils new prognostic biomarkers and novel insights into field cancerization. Sci Rep. 2022 Oct 12;12(1):17051. doi: 10.1038/s41598-022-20338-w. PMID: 36224266; PMCID: PMC9556831.
Curtius K, Wright NA, Graham TA. An evolutionary perspective on field cancerization. Nat. Rev. Cancer. 2018;18(1):19–32. doi: 10.1038/nrc.2017.102.
Bruick RK, McKnight SL. A conserved family of prolyl-4-hydroxylases that modify HIF. Science. 2001 Nov 9;294(5545):1337-40. doi: 10.1126/science.1066373. Epub 2001 Oct 11. PMID: 11598268.
Pressley, M., Gallaher, J.A., Brown, J.S. et al. Cycling hypoxia selects for constitutive HIF stabilization. Sci Rep 11, 5777 (2021). https://doi.org/10.1038/s41598-021-85184-8
Russell, S., Gatenby, R. A., Gillies, R. J. & Ibrahim-Hashim, A. Pseudohypoxia: life at the edge. In Ecology and evolution of cancer (eds Ujvari, B. et al.) 57–68 (Academic Press, London, 2017).
Martin NK, Gaffney EA, Gatenby RA, Maini PK. Tumour-stromal interactions in acid-mediated invasion: a mathematical model. J Theor Biol. 2010 Dec 7;267(3):461-70. doi: 10.1016/j.jtbi.2010.08.028. Epub 2010 Sep 8. PMID: 20816684; PMCID: PMC3005191.
Fu Y, Zheng S, Zheng Y, Huang R, An N, Liang A, Hu C. Glioma derived isocitrate dehydrogenase-2 mutations induced up-regulation of HIF-1α and β-catenin signaling: possible impact on glioma cell metastasis and chemo-resistance. Int J Biochem Cell Biol. 2012 May;44(5):770-5. doi: 10.1016/j.biocel.2012.01.017. Epub 2012 Jan 31. PMID: 22309944.
Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 2003 Jun;9(6):677-84. doi: 10.1038/nm0603-677. PMID: 12778166.
Serocki M, Bartoszewska S, Janaszak-Jasiecka A, Ochocka RJ, Collawn JF, Bartoszewski R. miRNAs regulate the HIF switch during hypoxia: a novel therapeutic target. Angiogenesis. 2018 May;21(2):183-202. doi: 10.1007/s10456-018-9600-2. Epub 2018 Jan 27. PMID: 29383635; PMCID: PMC5878208.
Popov TM, Goranova T, Stancheva G, Kaneva R, Dikov T, Chalakov I, Rangachev J, Konov D, Todorov S, Stoyanov O, Mitev V. Relative quantitative expression of hypoxia-inducible factor-1α, -2α and -3α, and vascular endothelial growth factor A in laryngeal carcinoma. Oncol Lett. 2015 Jun;9(6):2879-2885. doi: 10.3892/ol.2015.3070. Epub 2015 Mar 24. PMID: 26137164; PMCID: PMC4473705.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Peritumor mucosa in advanced laryngeal carcinoma exhibits an aberrant proangiogenic signature distinctive from the expression pattern in adjacent tumor tissue

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and methods

3. Results

4. Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1