Despite recent advances, non-small cell lung cancer (NSCLC) still leads to a great mortality in most of the continents1,2, reaching up to one third of deaths in certain countries1,3. Clinical factors such as chronic obstructive pulmonary disease (COPD) or airway obstruction underlie the pathophysiology of LC in many patients1,4–6. Several relevant investigations have clearly demonstrated that airway obstruction and emphysema render the patients more susceptible to the development of LC7–9. Despite this consolidated knowledge, full elucidation of the underlying biological features is still underway.

In the airways, lungs, and blood compartment of patients with LC and underlying COPD, mechanisms such as redox imbalance, inflammatory events, epigenetics, and immune alterations were shown to be disrupted compared to LC patients with no COPD10. As a result of the interaction of those biological events with key cellular processes, namely angiogenesis, cell death and repair, and the cell survival machinery, COPD patients are more prone to lung tumorigenesis11.

Stroma is defined as the part of a tissue or organ that confers mainly structure with no specific function, and is mainly composed by blood vessels, nerves, and connective tissue. In LC, several components such as extracellular matrix (ECM), endothelial cells, and cancer-associated fibroblasts (CAFs) play a significant role in tumorigenesis and cancer progression12. CAFs are a major component of the stroma in tumors. Growth factors, hormones, and cytokines mediate the tumor cell proliferation favored by CAFs. The most specific and widely used marker of CAFs is alpha-smooth muscle actin (SMA), which is indeed a specific marker of myofibroblasts12. The differentiation process of epithelial cells into mesenchymal cells is known as epithelial-mesenchymal transition (EMT), characterized by the appearance of mesenchymal properties13,14. Interestingly, CAFs may also regulate EMT 13.

Extracellular macromolecules such as collagen, enzymes, and glycoproteins conform a specific network of the ECM, which is also involved in tumor development and progression15. In cancer stroma, collagen was demonstrated to be the most abundant protein16. Importantly, type I collagen promotes growth of cancer cells, invasion, and distant metastasis, thus favoring tumor progression17, as well as resistance to therapy18. Whether a distinct expression of extracellular matrix markers or CAFs may take place in the stroma of lung tumor samples of patients with COPD remains to be answered.

The formation of new vessels in tumors can be identified using specific markers such as platelet endothelial cell adhesion molecule also known as cluster of differentiation (CD) 31. CD31 is involved in several physiological processes, namely maintenance of vascular endothelial and inflammatory cell functions and is also expressed in tumor cells19. In fact, the immunohistochemical measurement of CD31 expression can be reliably used as a marker of neoangiogenesis in tumors20. In mice with experimental airway inflammation mimicking COPD, an immunosuppressive microenvironment of the lung tumors was characterized by increased angiogenesis21. Whether differences in CD31 expression may exist in tumors of patients with COPD compared to non-COPD remain to be identified.

On this basis, we hypothesized that in LC patients with airway obstruction, cancer stroma as analyzed using specific markers may differ from tumors of patients with no underlying COPD. Accordingly, our objectives were to determine in lung tumors and non-tumor specimens (control samples) of LC patients with and without COPD the following parameters: 1) CAFs and non-tumor fibroblasts (cultured fibroblasts), 2) type I collagen as a marker of extracellular matrix, and 3) CD31 expression levels as a marker of endothelial cells and blood vessels.

Fresh human tumor and non-tumor lung samples were placed in Dulbecco's Modified Eagle Medium (DMEM) with 1% of penicillin, streptomycin, and fungiozone immediately after obtaining lung specimens and transported on ice to the molecular laboratory. Tumor and non-tumor specimens were minced finely and digested in 1% collagenase type I (Sigma-Aldrich, St. Louis, MO) at 37°C for two hours with occasional agitation. Then the digested tissue was centrifuged at 1,200rpm for two minutes. Cell suspensions were cultured on culture plates in proliferation medium consisting of the mixture of DMEM-medium, 10% fetal bovine serum, and 1% penicillin-streptomycin-fungizone solution at 37°C in a 5% CO2 atmosphere. The culture medium was changed after 48hours to remove unattached cells and debris in suspension. Cells were subcultured with 0.025% trypsin (Life Technologies, California, USA) and 0.01% EDTA when they reached 50-80% confluence for ten minutes. All the study experiments were performed on the cultured cells between passages 1 and 2 of the primary cultures to perform immunofluorescence as described below.

CAFs and NFs were identified by analyzing the fibroblast-specific protein vimentin and alpha-SMA (CAFs). Briefly, cells were fixed with acetone and methanol (1:1) on the slides at -20°C for ten minutes, and were then washed with PBS three times. Subsequently, slides were incubated with blocking solution (50mM Tris with PH=7.5, 150 Mm NaCl, 0.01% Triton, 1% bovine serum albumin and 1% skimmed milk powder) for one hour at room temperature in a humidifed chamber. Subsequently, primary antibodies incubation with anti-alpha SMA antibody (anti-alpha-SMA antibody, Santa Cruz) and anti-vimentin antibody (anti-vimentin antibody, Santa Cruz) was performed overnight at 4°C in the chamber. After washing with PBS three times, slides were incubated with corresponding secondary antibodies diluted in PBS for one hour: anti-mouse IgG FITC (Invitrogen, Thermo Fisher Scientific) and anti-rabbit IgG A647 (Invitrogen, Thermo Fisher Scientific) at room temperature. Finally, the sections were mounted using the fluorescent mounting medium 4’,6-diamidino-2-phenylindole (DAPI) G-Fluoromount medium (Southern Biotech, Birmingham, AL, USA), which specifically stained DNA (allowing identification of all nuclei) in the cell sections. A fluorescence microscope (x 40 objectives, Nikon Eclipse Ni, Nikon, Tokyo, Japan) coupled with a digitizing camera was used to identify and count the number of fibroblasts (30 fields) in each study sample. Results were expressed as the percentage of either both alpha-SMA and vimentin positively stained fibroblasts for identification of CAFs or vimentin-only positively stained for detection of NFs to the total number of counted fibroblasts in the 30 fields. Results are reported separately for both CAFs and NFs in each type of lung specimen and patient group.

Type I collagen and endothelial cells were identified on three-micrometer lung tumor and non-tumor cross-sections using immunohistochemical procedures as previously described10,29. Following deparaffinization, lung cross-sections were immersed in preheated antigen retrieval solution of ethylenediaminetetraacetic acid (EDTA, pH 9), incubated at 95°C for 40minutes to be then cooled down to room temperature. Slides were washed over the following steps with phosphate buffer saline (PBS). Endogenous peroxidase activity was blocked with 6% hydrogen peroxide for 15minutes. Primary antibody incubation with anti-collagen I antibody (anti-collagen I antibody, Abcam, Cambridge, UK) and anti-CD31antibody (anti-CD31 antibody, Abcam, Cambridge, UK) was performed for one hour. Slides were incubated with biotinylated universal secondary antibody for 30minutes followed by a 30-minute incubation with HRP-streptavidin and diaminobenzidine for five minutes (kit LSAB+HRP Dako Cytomation Inc., Carpinteria, CA, USA) as a substrate. Hematoxylin counterstaining was performed for two minutes and slides were dehydrated and mounted for conventional microscopy. Images of the stained lung sections (tumor and non-tumor) were captured with a light microscope (Olympus, Series BX50F3, Olympus Optical Co., Hamburg, Germany) coupled with an image-digitizing camera (Pixera Studio, version 1.0.4, Pixera Corporation, Los Gatos, CA, USA).

Expression of the markers collagen and CD31 was estimated as the percentage of type I collagen and CD31 using the semiquantitative immunohistochemical scoring system (Hscore) according to methodologies previously published30. Type I collagen and CD31 staining in the tumor and non-tumor specimens was established according to the following categories: Hscore 0 (indicated the absence of staining) and Hscore 1 (indicated the presence of staining). Data are shown as the percentage of both positively and negatively stained structures for all the histological sections in both tumor and non-tumor lung specimens.

In the current investigation, the main findings were that levels of the endothelial marker CD31 significantly decreased in tumors of LC patients, but not in tumors of patients with airway obstruction. In both groups of patients, a rise in the expression of CAFs was seen in lung tumors. Levels of type I collagen in tumor and non-tumor lungs did not differ between patient groups. The most relevant findings collected in the study are discussed below.

CAFs play a crucial role in cancer cell invasion through several mechanisms31. Vimentin, which is expressed in normal mesenchymal cells, maintains cellular integrity and provides resistance against stress. Its function has also been proposed in different cancer cell types including LC32. In the present investigation, the expression of CAFs was significantly greater in the tumor specimens in both groups of LC patients with and without COPD. No significant differences in the levels of cultured CAFs in tumor lungs were seen between the study groups of patients. These findings suggest that CAFs are similarly expressed in lung tumors regardless of underlying airway obstruction. They also imply that fibroblasts are not likely to be involved in an accelerated process of cancer invasion and progression in patients with airway obstruction. Conversely, the percentage of fibroblasts-expressing vimentin-only was significantly reduced in the tumors of both groups of patients. These results also reinforce the concept that CAFs are likely to be a predominant feature of the stroma in lung tumor progression in the patients regardless of the presence of airway obstruction.

Whether a similar profile of CAFs expression can be detected in lung tumors of patients with other underlying respiratory diseases remains to be elucidated. In idiopathic pulmonary fibrosis, myofibroblasts are persistently activated, which secrete collagen type I, and express alpha-SMA fibers, thus they may favor lung tumorigenesis35. Conversely, in patients with non-cystic fibrosis bronchiectasis, a lower or no risk of LC was demonstrated33,34.

Activated myofibroblasts synthesize extracellular components that contribute to the remodeling of the ECM taking place during carcinogenesis. As such CAFs secrete type I collagen, which plays an important role in tumor development, growth, and epithelial-mesenchymal transition36. Moreover, overall survival correlated with low levels of expression of type I collagen and cancer cell differentiation36. In the present study, expression levels of collagen did not significantly differ between tumor and non-tumor samples or between the study groups. These findings suggest that collagen was not a major driver in lung tumor development in these patients, probably because well-differentiated tumor types were analyzed in the study.

CD31 is a glycoprotein expressed in endothelial cells, leukocytes, T cells, and platelets20. CD31 is also expressed in lung tumors37. In the current investigation, a significant decline in CD31 expression levels (Hscore 1) was detected in the tumor specimens of patients with LC, while in patients with underlying airway obstruction no significant differences were seen between lung tumor and non-tumor samples. These findings imply that the vascular endothelial component of stroma was probably involved in the prognosis of LC in patients with and without COPD. In fact, 47% of LC-COPD and 80% of LC patients are still alive in this series (10-year follow-up, data not shown). In keeping with, CD31 has proven to be a useful marker to evaluate angiogenesis in lung tumors38 as well as to monitor the response to specific anti-angiogenic molecules such as vascular endothelial growth factor (VEGF) in clinical settings38,39. In this regard, several investigations have demonstrated that VEGF inhibitors, through reduced angiogenesis (CD31 marker among others), are currently prescribed as single agents in the third-line treatment of patients with NSCLC38,39.

A limitation in the study was the relatively reduced number of analyzed patients. Nonetheless, calculations of sample size estimated 12 patients in each group (24 in total), thus the number of patients included was sufficient to detect statistically significant differences in the study. The degree of airway obstruction might have influenced the study results. However, as most of the patients were in GOLD stages I and particularly II, COPD severity did not exert any significant impact on the results. Almost half of the patients were non-smokers, thus cigarette smoking might have influenced the study results. Nevertheless, a subanalysis in which non-smokers and ex-smokers were included revealed identical results to those obtained with the entire population.

If non-tumor samples had been obtained from a closer distance from the tumors, the profile of biological events might have differed as shown previously for other components of the extracellular matrix (integrins) that probably play a significant role in recurrence40. Nonetheless, this was not explored in the present study, and warrants further attention.

Within the stroma, the expression of the vascular endothelial marker CD31 was reduced in tumors of patients without airway obstruction, while expression levels of the ECM component type I collagen did not differ between patient groups. A rise in the levels of CAFs was detected in the lung tumors of patients irrespective of underlying airway obstruction.

Low levels of CD31 may have implications in the overall survival of LC patients, especially in those without underlying airway obstruction. Investigations aiming to decipher the specific role of CD31 as a predictor of survival and as a biomarker to monitor anti-angiogenic agents in lung tumors of patients with underlying respiratory diseases are warranted.

Study conception and design: EB, VC; Patient assessment and recruitment: JT, VC, DRC, MMJ, ARF, RA, LP; Molecular biology analyses: JT, DRC, KA; Statistical analyses and data interpretation: XD, JT, DRC, EB; Manuscript drafting and intellectual input: EB, JT; Manuscript writing final version: EB.

This study has been supported by FIS 18/00075 (FEDER, ISC-III) & CIBERES (ISC-III), SEPAR 2018 and SEPAR 2020, and an unrestricted research grant from Menarini SA 2018 (Spain).

None.