Sage Journals: Discover world-class research

Abstract

In recent years, it has been demonstrated that malignancy arises and advances through the molecular interplay between tumor cells and non-malignant elements of the tumor stroma, that is, fibroblasts and extracellular matrix. However, in contrast to the mounting evidence about the role of tumor stroma in the genesis and progression of the malignant disease, there are very few data regarding the uninvolved stromal tissue in the remote surrounding of the tumor. Using the objective morphometric approach in patients with adenocarcinoma, we demonstrate the remodeling of extracellular matrix of the lamina propria in the uninvolved rectal mucosa 10 and 20 cm away from the neoplasm. We show that the representation of basic extracellular matrix constituents (reticular and collagen fibers and ground substance) is decreased. Also, the diameter of empty spaces that appear within the extracellular matrix of the lamina propria is increased. These spaces do not represent the blood or lymphatic vessel elements. Very likely, they reflect the development of tissue edema in the remote, uninvolved lamina propria of the mucosa in patients with the malignant tumor of the rectum. We hypothesize that the remodeling of extracellular matrix in lamina propria of the rectal mucosa may increase its stiffness, modulating the mechano-signal transduction, and thus promote the progression of the malignant disease.

Keywords

Rectum cancer human extracellular matrix

Introduction

The surrounding tissue of colorectal cancers has attracted considerable interest in recent years for two main reasons. The first reason is that carcinogens act on an entire tissue exposed to this carcinogen, and consequently, genetic alterations are expected to occur at sites distant from the tumor itself. Therefore, several malignant or pre-malignant lesions may occur in the vicinity of an actual tumor. This concept of field carcinogenesis has originally been proposed for the oral mucosa but has been expanded to practically all tissues of the human body.^1–3 While the field carcinogenesis primarily targets the epithelial cells of the colorectum, the connective stroma has attracted considerable interest too. The masters of colorectal carcinogenesis Vogelstein and Kinzler⁴ stated recently that “Despite intensive efforts, no genetic alterations have been shown to be required to convert a malignant primary tumor into a metastatic lesion.” With reference to distant metastasis formation, the decreased presence of lymphatic vessels and reduced immune cytotoxicity have recently been highlighted as hallmarks for the occurrence of metastasis,⁵ thus putting the tumor stroma into the focus of metastasis research. Malignancies arise and advance through the interactions with the elements of the neighboring tissues, that is, the tumor microenvironment.^6,7 Special significance has been attributed to the molecular interplay between tumor cells and non-malignant elements of the tumor stroma, that is, fibroblasts and extracellular matrix (ECM).⁸ Tumor progression and aggressiveness are promoted by the surrounding tissue through cell-to-cell contacts or secreted molecules.^9,10 In return, tumor cells uphold the recruitment of fibroblasts into a tumor mass and their transdifferentiation into myofibroblasts, which produce factors stimulating cancer progression.^11–14 In this way, the interplay between the cancer cells and stromal cells is established, a positive feedback loop is closed and the advancement of malignant disease is promoted.

However, despite the mounting evidence about the role of tumor stroma in the genesis and progression of the malignant disease, there are very few data regarding the uninvolved stromal tissue in the remote surrounding of the tumor. Only very recently, we performed a morphometric study of the mucosa of rectum and revealed the alterations of Lieberkühn crypts, as well as reduced cellularity, 10 and 20 cm away from the malignant tumor. Our study also provided indications that ECM of the lamina propria could be affected.¹⁵

Therefore, the aim of this study was to investigate the organization of ECM components of the lamina propria of the rectal mucosa in the remote surrounding of the malignant tumor and quantitate them using an objective morphometric approach.

Our work is the first to reveal a profound remodeling of ECM of the lamina propria of the uninvolved human rectal mucosa in the remote surrounding of the malignant tumor. Thus, we document the interactions between the cancer and distant mucosal tissue of the affected organ.

Materials and methods

Tissue samples

This study has been approved by the Ethics Committee of Zvezdara Clinical Center, Belgrade, Serbia (27 November 2013) and performed in accordance with ethical standards laid down in the 1964 Declaration of Helsinki. All individuals involved received detailed verbal information and gave their informed consent prior to their inclusion into the study.

Only patients with newly discovered tumor were included in the study. Thus, they had not received any previous treatment for the malignant disease. Tissues were endoscopically sampled at the Center for Gastroenterology and Hepatology, Zvezdara Clinical Center, Belgrade, Serbia, from patients suspected on clinical grounds to suffer from a rectal cancer, which all were subsequently diagnosed as adenocarcinoma according to World Health Organization (WHO) histological classification of tumors. Patients with diverticular disease of the colon, previous infectious colitis, or inflammatory bowel disease were excluded from the study. Patients were divided into groups according to their gender. All patients were on standard mixed meals regimen and not on any particular form of diet.

Samples of rectal mucosa were obtained from 32 patients older than 60 years (18 males and 14 females; Table 1). Cancers were located in the sigmoid colon 25−30 cm from the external anal verge and the biopsies were taken 10 and 20 cm away from the malignant tumor in the caudal direction. The samples of rectal mucosa collected at the same institution from 30 healthy persons of the corresponding age (16 males and 14 females; Table 1) during the active endoscopy screening of individuals with a family history of intestinal malignancy, possibly suffering from an as yet unidentified, asymptomatic cancer in which no disease involving the rectal mucosa was found, were used as control. Samples were collected from the upper third of rectum.

Table 1.

Demographic characteristics of patients included in the study.

Patients	Number	Age (years)	Gender
Patients	Number	Age (years)	Male	Female
Cancer	32	72.12	18	14
Healthy	30	70.90	16	14

Tissue preparation, staining, and morphometric measurements

The biopsies of the rectal mucosa were fixed in 10% neutral buffered formalin, processed to Paraplast. Tissue sections (3–5 µm thick) were routinely stained with hematoxylin–eosin, while Gomori’s silver impregnation and Masson trichrome staining were used for demonstration of reticular and collagen fibers, respectively.¹⁶

Microphotographs were acquired with a digital camera Olympus C3030-Z (Olympus Deutschland GmbH, Hamburg, Germany) connected to a light microscope Opton Photomicroscope III (Carl Zeiss AG, Oberkochen, Germany) or digital camera Leica DFC295 (Leica, Heerbrugg, Switzerland) connected to a light microscope Leica DM4000 B LED (Leica, Wetzlar, Germany).

Gomori’s silver impregnation technique was used to identify the reticular fibers which stained black. A random selection of three fields per slide from the subepithelial region, between the crypts, and above the lamina muscularis mucosae was assessed for further analysis.

The diameters of spaces between reticular fibers were measured at the magnification 630×. In order to avoid the inclusion of crypts in the measurement of the spaces between reticular fibers, a rectangle was drawn that always occupied the same area of the lamina propria (3600 µm²) excluding the crypts, and our measurements were performed within this rectangle. The plugin BoneJ within open-source software Fiji¹⁷ was used to estimate the diameter of spaces between reticular fibers in the lamina propria of the rectal mucosa of healthy individuals (n = 30), as well as 10 cm (n = 27) and 20 cm (n = 22) away from the malignant tumor. BoneJ determines the diameter (µm) of the largest sphere that can be inserted into the spaces between the reticular fibers and assigns different colors to these spaces according to their diameter.

The presence of reticular fibers in colonic lamina propria was determined in the following manner. A random selection of 15 fields per slide (5 fields in subepithelial region, 5 fields between the crypts of lamina propria, and 5 fields above lamina muscularis mucosae) was assessed at the magnification of 400× (n = as above). Reticular fibers were extracted using Color Picker Threshold Plugin within open community platform for bioimage informatics Icy.¹⁸ For each slide, 10 positive and 10 negative colors were selected as recognition patterns of stained and unstained tissue elements. The resulting images were systematically compared with the corresponding originals, and in cases in which the detection of reticular fibers was not accurate, the threshold was fixed manually. The presence of reticular fibers in rectal lamina propria was expressed as the relative percentage of the area occupied by the reticular fibers divided by the area of the lamina propria selected with an imaging processor.

Masson’s trichrome staining identified collagen fibers, which were green. The presence of collagen fibers of healthy individuals (n = 30), as well as 10 cm (n = 32) and 20 cm (n = 32) away from the malignant tumor, was determined in the same manner as described above for reticular fibers.

Also, for the identification of collagen fibers, the unfixed and label-free colon tissue samples were imaged on an original labframe nonlinear laser scanning microscope (NLM). The microscope is described elsewhere.¹⁹ Experimental setup for NLM was used for second harmonic generation (SHG) imaging of 3D collagen distribution from the samples. The incoming infrared femtosecond pulses from the tunable mode-locked Ti:sapphire laser (Coherent, Mira 900) were directed onto the sample by a dichroic mirror through the Zeiss EC Plan-Neofluar 40×/1.3 NA Oil objective. The laser wavelength was 840 nm. The SHG was selected by narrow bandpass filter at 420 nm (Thorlabs FB420-10, FWHM 10 nm). The average laser power on the sample was 30 mW and the peak laser power was 2.5 kW.

Immunohistochemical analysis was performed on formalin-fixed, paraffin-embedded sections using following antibodies and dilution ratios: monoclonal mouse anti-human CD34 (M7165, dilution 1:100; Dako, Carpinteria, CA, USA), CD105 (clone SN6h, MA5-11854, dilution 1:100; Thermo Fisher Scientific, Rockford, IL, USA), podoplanin (clone D2-40, M3619, dilution 1:100; Dako), as well as hyaluronic acid binding protein (HABP, biotinylated, 385911-50 UG, dilution 1:75; Calbiochem, Spring Valley, CA, USA).

For anti-CD34, -CD105, and -podoplanin antibodies, the following procedure was used. Briefly, after dewaxing and rehydration, a heat-inducing antigen retrieval procedure using Tris-EDTA at pH 9.0 for 30 min was performed on all tissue sections, with subsequent washing in Tris-buffered saline (TBS) with Tween and incubation with primary antibodies for 60 min. The sections were treated by applying the commercial Ultra Vision/3,3′-diaminobenzidine (DAB) staining kit (Thermo Scientific Lab Vision TL-060-HD, Rockford, IL, USA). Immunoreactions were developed by DAB substrate.

For HABP antibody, Dako retrieval solution (S1699 at pH = 6) was used. After washing in TBS, protein blocking was performed with 1% bovine serum albumin (BSA) in TBS. After incubation with primary biotinylated antibodies for 60 min and with VECTASTATIN ABC-AP kit (Vector Laboratories, Burlingame, CA, USA) for 30 min, staining with Permanent Red (Dako K0640, Carpinteria, CA, USA) was used for visualization. The sections were counter-stained with Mayer’s hematoxylin. Negative controls were performed on colon sections using the same methodology, but with the omission of primary antibody. For CD105 antibody, sections of human colon cancers were used as a positive control and for HABP antibody, sections of human nasal mucosa.

The number of blood vessels demonstrated with CD34 or CD105 in rectal lamina propria (per 0.1 mm² of the tissue) was determined at 5 fields per slide at 200× magnification using the open-source software Fiji. A region of interest was drawn on each field, which excluded the crypts and included only the lamina propria. Counting of blood vessels was performed on healthy tissue (n = 25), 10 and 20 cm away from tumor (n = 10), using the Multi-point tool.

The diameter of spaces within ECM and the representation of hyaluronan, as demonstrated with HABP staining, in healthy tissues (n = 4) 10 cm (n = 10) and 20 cm (n = 12) away from the malignant tumor, were determined as described for reticular fibers.

Statistical analysis

The statistical package SPSS for Windows 12.0 (SPSS Inc., Chicago, IL, USA) was used to calculate the means and standard deviations, as well as to indicate significant differences (Student’s t-test and one-way analysis of variance (ANOVA) at p < 0.05).

Results

Morphological alterations of the uninvolved rectal mucosa in cancer patients

The mucosa of normal, healthy persons exhibited an orderly organization. The crypts were well developed, generally regular and round in shape, with luscious epithelium consisting mostly of goblet cells and fewer enterocytes (Figure 1(a)). The connective tissue of lamina propria also displayed a tidy appearance with very little free space within ECM (Figure 1(a)).

Figure 1.

Representative photomicrographs of rectal mucosa in healthy persons (a, c, e, and g) and cancer patients 10 cm away from the tumor (b, d, f, and h). (a) On hematoxylin–eosin staining, the connective tissue of lamina propria of healthy person shows orderly organization with very little free space within extracellular matrix, (b) while at the distance of 10 cm away from the tumor, lamina propria is disorganized with prominent spaces notable within extracellular matrix. Gomori’s silver impregnation technique identified reticular fibers which were (c) thick and closely appositioned in normal mucosa and (d) feeble and loosely arranged at the distance 10 cm away from the tumor. Measurement of the spaces between reticular fibers in (e) a healthy person and (f) 10 cm away from tumor was performed within the rectangle using plugin BoneJ (Fiji). BoneJ determines the diameter of the largest sphere that can be inserted into the spaces between reticular fibers and assigns different colors to these spaces according to their diameter. Brighter spheres have larger diameter. Graphical output from (g) BoneJ of healthy tissue and (h) 10 cm away from tumor.

At the distance of 10 cm away from the tumor, the number of crypts appeared decreased, with diminished diameter, slightly irregular in shape, and with frail epithelium (Figure 1(b)). The connective tissue of lamina propria showed a disorderly organization, whereby the prominent spaces were notable within the ECM (Figure 1(b)). Similar alterations of lamina propria of the rectal mucosa, but less prominent in comparison with the normal tissue, were observed in cancer patients 20 cm away from the tumor (not shown).

Therefore, we wished to objectively investigate the extent of disorderly organization of the mucosal lamina propria 10 and 20 cm away from the adenocarcinoma of the rectum and quantify the basic components of ECM (namely, reticular and collagen fibers, as well as ground substance). To this end, we used the computer-aided image analysis of the affected tissue to estimate the diameter of spaces and the representation of the components of ECM in the lamina propria of the rectal mucosa 10 and 20 cm away from the malignant tumor.

Disorganization of reticular fibers in the lamina propria of the uninvolved rectal mucosa in cancer patients

In normal tissue, the reticular fibers were orderly organized and showed a tidy appearance. They were thick and intimately appositioned to the epithelium of the crypts. The basal lamina of the crypts was surrounded by three to four layers of reticular fibers. A dense meshwork of reticular fibers leaving little space for non-fibrous ground substance connected the crypts with each other (Figure 1(c)).

At the distance of 10 and 20 cm away from the tumor, the reticular fibers around the crypts consisted only of one thin layer, and those fibers spanning the ECM were also thin, thus leaving the prominent empty spaces between the reticular fibers (Figure 1(d)). We revealed that the diameter of spaces between the reticular fibers was significantly increased at the distance of 10 cm away from the tumor lesion (5.66 ± 2.21 µm), in comparison with both healthy controls (3.32 ± 0.81 µm; p < 0.01; Figures 1(e)–(h)) and tissue samples taken 20 cm away from the tumor (3.72 ± 1.27 µm; p < 0.01). There was no statistically significant difference between samples taken 20 cm away from the tumor and control (Table 2). These results clearly documented the disturbed organization of the lamina propria in the remote, uninvolved rectal mucosa 10 cm away from the malignant lesion.

Table 2.

Morphometric measurements of the rectal mucosa in healthy subjects and cancer patients 10 and 20 cm away from tumor.

	Healthy controls	10 cm away from tumor	20 cm away from tumor
Diameter of spaces between reticular fibers (µm)	3.32 ± 0.81	5.66 ± 2.21*	3.72 ± 1.27⁺
Representation of reticular fibers (%)	42.72 ± 11.33	27.99 ± 7.86*	38.76 ± 10.01⁺
Representation of collagen fibers (%)	48.05 ± 8.92	26.43 ± 6.22*	35.15 ± 8.34**⁺
Diameter of spaces between HABP (µm)	2.00 ± 0.40	4.27 ± 1.58*	3.78 ± 1.45^§
Representation of HABP (%)	60.41 ± 9.31	42.71 ± 11.81*	45.42 ± 13.35**

HABP: hyaluronic acid binding protein.

10 cm away from tumor versus healthy controls (p < 0.01).

10 cm away from tumor versus 20 cm away from tumor (p < 0.01).

20 cm away from tumor versus healthy controls (p < 0.01).

20 cm away from tumor versus healthy controls (p < 0.05).

Furthermore, we wished to explore whether the increased diameter of spaces was accompanied by a decreased representation of reticular fibers in the lamina propria of the remote rectal mucosa 10 and 20 cm away from the neoplasia. We revealed that the representation of reticular fibers was significantly decreased in the lamina propria of rectal mucosa 10 cm (27.99 ± 7.86; p < 0.01) but not 20 cm (38.76 ± 10.01) away from the tumor in comparison with the tissue of healthy individuals (42.72 ± 11.33; Table 2). The representation of reticular fibers was significantly lower (p < 0.01) in the lamina propria at the distance of 10 cm away from the tumor compared with the representation at the distance of 20 cm (Table 2).

Disorganization of collagen fibers in the lamina propria of the uninvolved rectal mucosa in cancer patients

In normal tissue, the collagen fibers showed a notably tidy appearance: the fibers were massive, intimately appositioned, and orderly organized (Figure 2(a)).

Figure 2.

Representative photomicrographs of rectal mucosa in healthy persons (a, c, and e) and cancer patients 10 cm away from the tumor (b, d, and f). Masson’s trichrome staining identified collagen fibers, which were massive, intimately appositioned, and orderly organized in (a) normal rectal tissue and (b) frail and loosely arranged 10 cm away from the tumor. NLM was used for SHG imaging of collagen fibers in (c) unfixed and unstained healthy tissue and (d) tissue 10 cm away from the tumor. SHG images confirmed the findings described above. (e) Immunohistochemical staining for hyaluronan using HABP in healthy person shows hyaluronan (red) organized in confluent bands between the crypts with small spaces within it. (f) At the distance 10 cm away from the tumor, hyaluronan formed thin bands with prominent spaces notable within it.

At the distance of 10 and 20 cm away from the tumor, the collagen fibers were frail and loosely arranged. Thus, the prominent spaces were notable between the collagen fibers (Figure 2(b)).

We further wished to exclude the possibility that the observed changes of the collagen fibers in the mucosal lamina propria were artificially produced (i.e. due to fixation or staining procedure). Therefore, NLM was used for SHG imaging of collagen distribution. In this method, the fresh, unfixed, and unstained tissue is used for demonstration of collagen fibers. SHG images wholly confirmed the above-described findings: in healthy individuals, the collagen fibers were massive, closely appositioned, and orderly organized around the crypts and throughout the lamina propria (Figure 2(c) and Online Resource 1). In the lamina propria 10 cm away from tumor, the collagen fibers were thinner, disordered, and loosely arranged with noticeable spaces between them (Figure 2(d) and Online Resource 2).

We also wished to explore whether the presence of free spaces between the collagen fibers was accompanied by a decreased representation of collagen in the lamina propria of the uninvolved rectal mucosa 10 and 20 cm away from the neoplasia. We revealed that the representation of collagen fibers in the lamina propria in the remote rectal mucosa 10 and 20 cm away from the cancer was significantly lower (26.43 ± 6.22 and 35.15 ± 8.34, respectively; p < 0.01) in comparison with the control, healthy individuals (48.05 ± 8.92; Table 2). Notably, the representation of collagen fibers was significantly lower (p < 0.01) in the lamina propria at the distance of 10 cm away from the tumor compared with that at the distance of 20 cm (Table 2).

Disorganization of ground substance in the lamina propria of the uninvolved rectal mucosa in cancer patients

The ECM ground substance was visualized by demonstration of HABP binding to hyaluronan in affinity histochemically stained tissue sections of healthy individuals and cancer patients and analyzed using computer-aided morphometric approach.

In normal tissue, the ECM ground substance was organized in confluent, uniform bands between the crypts with very few small spaces within it (Figure 2(e)).

At the distance of 10 and 20 cm away from the tumor, the ECM ground substance recapitulated the structural organization similar to that observed for reticular and collagen fibers: namely, the prominent spaces were notable within it (Figure 2(f)). The morphometric analysis revealed that the diameter of spaces within the ground substance was significantly increased at the distance of 10 and 20 cm away from the tumor lesion (4.27 ± 1.58 µm, p < 0.01 and 3.78 ± 1.45 µm, p < 0.05, respectively), in comparison with healthy controls (2.00 ± 0.40 µm; Table 2). Furthermore, we revealed that the representation of ECM ground substance in the lamina propria of the remote rectal mucosa 10 and 20 cm away from the cancer was significantly lower (42.71 ± 11.81 and 45.42 ± 13.35, respectively; p < 0.01) in comparison with the control, healthy individuals (60.41 ± 9.31; Table 2).

Edema of the lamina propria of the uninvolved rectal mucosa in cancer patients

Finally, we wanted to elucidate whether the enlarged spaces within the ECM in the lamina propria of the remote, unaffected rectal mucosa 10 and 20 cm away from the tumor could represent the blood or lymphatic vessel elements.

On the routinely stained sections of the healthy tissue, the small blood vessels and capillaries were readily discernible, but infrequently seen. On the contrary, in lamina propria of the rectal mucosa 10 and 20 cm away from the tumor, the small blood vessels and capillaries were abundant and much more frequent than in the healthy tissue. To more specifically uphold our observation, we stained the tissue samples immunohistochemically with CD34 and CD105 antibodies for total and newly formed blood vessels, respectively. We counted the number of blood vessel profiles per unit area of tissue (0.1 mm² of the lamina propria).

The CD34-positive blood vessels were abundant in healthy tissue (Figure 3(a)), as well as 10 and 20 cm away from the malignant tumor (Figure 3(b)). However, their number per 0.1 mm² of the lamina propria 10 cm away (65.15 ± 18.94, p < 0.01) and 20 cm away from the tumor (45.77 ± 16.48; p < 0.01) was significantly higher compared with the lamina propria of healthy tissue (36.57 ± 7.59). The number of CD34-positive blood vessels 10 cm away from the tumor was significantly increased compared with lamina propria 20 cm away from the tumor (p < 0.01; Table 3).

Figure 3.

Immunohistochemical staining for CD34 and CD105 of rectal mucosa in healthy persons (a and c) and cancer patients 10 cm away from the tumor (b and d). (a) The CD34-positive blood vessels were abundant in healthy tissue. (b) Their number was significantly increased 10 cm away from the tumor. (c) The blood vessels in healthy lamina propria were either minimally positive or totally CD105-negative. (d) The number of CD105-positive blood vessels was significantly increased 10 cm away from the rectal adenocarcinoma (some blood vessels are indicated by arrows).

Table 3.

Number of CD34- and CD105-positive blood vessels per 0.1 mm² of the rectal mucosa in healthy subjects and cancer patients 10 and 20 cm away from tumor.

	Healthy controls	10 cm away from tumor	20 cm away from tumor
Number of CD34-positive blood vessels (per 0.1 mm² of lamina propria)	36.57 ± 7.59	65.15 ± 18.94*	45.77 ± 16.48** +
Number of CD105-positive blood vessels (per 0.1 mm² of lamina propria)	5.8 ± 2.52	14.93 ± 5.65*	9.53 ± 4.97** +

10 cm away from tumor versus healthy controls (p < 0.01).

20 cm away from tumor versus healthy controls (p < 0.01).

10 cm away from tumor versus 20 cm away from tumor (p < 0.01).

The microvessels in rectal lamina propria from healthy controls were either minimally positive or totally CD105-negative. Rare CD105-positive blood vessels were preferentially located either superficially immediately below the basement membrane or at the cryptal base (Figure 3(c)). The number of CD105-positive blood vessels was significantly increased 10 and 20 cm away from the rectal adenocarcinoma (Figure 3(d), 14.93 ± 5.65 and 9.53 ± 4.97, respectively; p < 0.01) in comparison with the control, healthy tissue (5.8 ± 2.52). Also, the number of CD105-positive blood vessels 10 cm away from the tumor was significantly increased compared with lamina propria 20 cm away from the tumor (p < 0.01; Table 3).

So, we demonstrate that both the number of total and newly formed blood vessels in the lamina propria 10 and 20 cm away from the tumor was increased in comparison with the lamina propria of healthy tissue. However, as the large spaces within the ECM of diseased rectal lamina propria remained CD34- and CD105-negative, we concluded that they do not represent the blood vessel elements. Thus, the question of their nature remained unanswered.

Therefore, we decided to investigate whether these spaces could represent newly developed lymphatic vessels. To clarify this issue, we immunostained the tissue samples with anti-podoplanin that labels lymphatic endothelium, whereas it is unreactive with vascular endothelium. We did not reveal any lymphatic vessels in the lamina propria of healthy colon and they were present only at the level of lamina muscularis mucosae where an abundant array could be seen (not shown). On the contrary, tiny profiles of lymphatic vessels were sometimes identifiable within the lamina propria 10 and 20 cm away from the tumor (not shown). These results neatly corresponded to those of Kenney and Jain.²⁰ However, the lymphatic vessels in such cases were very few and could not account for numerous spaces in the lamina propria, which were podoplanin-negative. Therefore, we concluded that these spaces do not represent the newly developed lymphatic vessels.

It should be notified that no differences in structure of the mucosal elements were observed between male and female healthy persons. Also, no gender difference for any structural element of the rectal mucosa was observed neither at 10 cm nor at 20 cm distance away from the tumor.

Taking our results together, we concluded that they reflect the development of a veritable tissue edema in the remote, uninvolved lamina propria of the mucosa in patients with the neoplastic tumor of the rectum.

Discussion

Our present work, to the best of our knowledge, is the first to document the remodeling of lamina propria in the uninvolved rectal mucosa remote from the malignant lesion using an exact morphometric analysis. We demonstrate the decreased representation of basic ECM constituents (reticular and collagen fibers, and ground substance), as well as the increased diameter of free spaces within the ECM in the lamina propria of the distant, uninvolved rectal mucosa.

These findings are in line with the finding that collagen is reduced in tumor-associated lamina propria.^21,22 Recent studies show that modifications of collagen fibers (e.g. cross-linking), associated with changes of pore sizes, strongly affect the ECM stiffness.^7,23–25 The cells grown in fibrillar collagen gels, which are nowadays extensively used to study tumor–microenvironment interactions in vitro, adopted a motile phenotype in gels with larger pores and a less motile phenotype in gels with small pores.²⁶ The increased rigidity of the ECM modulates the mechano-signal transduction and promotes the aggressiveness of neoplastic cells.^27,28 One is tempted to speculate that the remodeling of the lamina propria in the remote rectal mucosa that we registered in vivo corresponds to these in vitro modifications of ECM, which sustain the aggressive behavior of tumor cells.

Two main questions arise from our study. First, what do the free spaces observed within the ECM represent? Second, does the remodeling of the remote rectal lamina propria reflect the influence of the malignant tissue, or conversely, denotes the intrinsic modifications of the stromal tissue which provide a suitable ground for tumor development?

Regarding the first question, and having in mind that enlarged spaces in the remote rectal lamina propria of cancer patients are CD34-, CD105-, and podoplanin-negative, we favor the possibility that these changes disclose a tissue edema at the greater distance from the tumor. This observation extends the finding that edema is observed in the immediate vicinity of tumors.²⁹

Regarding the second question, the former hypothesis is strengthened by the fact that neoplastic tissues can affect the structure and function of very distant organs.^30,31 It is also supported by our finding that the changes of lamina propria are somewhat more prominent at the distance of 10 cm than at 20 cm away from the tumor. The latter notion is underpinned by our finding that the remodeling of the rectal lamina propria still remains evident at a greater distance (20 cm) away from the tumor. This is in line with the concept of “field carcinogenesis”: it postulates that genetic and environmental risk factors induce large areas of tissue injury suitable for cancer development.^32,33 So, it remains possible that the remodeling of lamina propria that we observed in cancer patients reflects the overall alteration of larger fields of the colorectal mucosa and its connective tissue. At this moment, based on our data, it cannot be concluded with certainty which hypothesis holds true. However, studies currently in progress in our laboratories—in which the samples of mucosal lamina propria are taken from the part of the large bowel the most distant from the neoplastic lesion—are expected to greatly clarify the situation.

In very recent time, a view has emerged that the role of stroma is not secondary to that of epithelium in the phenomenon of “field carcinogenesis.”^34,35 Our results strongly support this opinion and suggest that the concept of “field carcinogenesis” should incorporate not only the changes in epithelial component, but the stroma of the malignantly affected organ as well.²

Our findings are in fine agreement with the results of genetic studies which showed that the significant gene expression alterations exist in the unaffected colon mucosa from patients with colon tumor.^36–38 In addition to changes in ECM components, we also found an increased number of CD34- and CD105-positive capillaries and small blood vessels in the remote rectal mucosa. This was also detected in endoscopically normal rectal mucosa in patients with multiple adenomas anywhere in the colon and was completely absent in patients with benign colonic diseases.³³ Together, these data show that the use of remote uninvolved mucosa from patients with tumors of the gut as control tissue³⁹ should be regarded as unjustified. Despite its unremarkable endoscopic appearance, this tissue is not normal and only truly healthy tissue should be used as controls.

In conclusion, our study reveals profound remodeling of the ECM of lamina propria in the rectal mucosa 10 and 20 cm away from the malignant lesion. It documents the complex interplay between the tumor and stromal tissues not only of the neoplasm itself, but of the distant, uninvolved rectal lamina propria, as well. The search for these changes may be used as a diagnostic tool and a valuable indicator of occult tumors of the large bowel.

Footnotes

Acknowledgements

The authors are grateful to Luca Vannucci for critically reading the manuscript and Jovanka Ognjanović for excellent technical assistance.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Ministry of Education, Science and Technological Developments of the Republic of Serbia (grant nos 175005, and 45016).

References

Hawthorn

Lan

Mojica

Evidence for field effect cancerization in colorectal cancer. Genomics 2014; 103(2–3): 211–221.

Lochhead

Chan

Nishihara

. Etiologic field effect: reappraisal of the field effect concept in cancer predisposition and progression. Mod Pathol 2015; 28(1): 14–29.

Patel

Tripathi

Gopalakrishnan

. Field cancerisation in colorectal cancer: a new frontier or pastures past? World J Gastroenterol 2015; 21(13): 3763–3772.

Vogelstein

Kinzler

KW.

The path to cancer—three strikes and you’re out. N Engl J Med 2015; 373(20): 1895–1898.

Mlecnik

Bindea

Kirilovsky

. The tumor microenvironment and immunoscore are critical determinants of dissemination to distant metastasis. Sci Transl Med 2016; 8(327): 327ra26.

Goubran

Kotb

Stakiw

. Regulation of tumor growth and metastasis: the role of tumor microenvironment. Cancer Growth Metastasis 2014; 7: 9–18.

Vannucci

Stroma as an active player in the development of the tumor microenvironment. Cancer Microenviron 2015; 8(3): 159–166.

Hemmings

Is carcinoma a mesenchymal disease? The role of the stromal microenvironment in carcinogenesis. Pathology 2013; 45(4): 371–381.

Hale

Hayden

Grabsch

HI.

Tumour-microenvironment interactions: role of tumour stroma and proteins produced by cancer-associated fibroblasts in chemotherapy response. Cell Oncol 2013; 36(2): 95–112.

10.

Valcz

Sipos

Tulassay

. Importance of carcinoma-associated fibroblast-derived proteins in clinical oncology. J Clin Pathol 2014; 67(12): 1026–1031.

11.

Fuyuhiro

Yashiro

Noda

. Upregulation of cancer-associated myofibroblasts by TGF-β from scirrhous gastric carcinoma cells. Br J Cancer 2011; 105: 996–1001.

12.

Mazzocca

Dituri

Lupo

. Tumor-secreted lysophostatidic acid accelerates hepatocellular carcinoma progression by promoting differentiation of peritumoral fibroblasts in myofibroblasts. Hepatology 2011; 54: 920–930.

13.

Polanska

Orimo

Carcinoma-associated fibroblasts: non-neoplastic tumour-promoting mesenchymal cells. J Cell Physiol 2013; 228(8): 1651–1657.

14.

Martinez-Outschoorn

Curry

. Oncogenes and inflammation rewire host energy metabolism in the tumor microenvironment: RAS and NFκB target stromal MCT4. Cell Cycle 2013; 12(16): 2580–2597.

15.

Despotović

Milićević

Milosević

. Morphometric study of uninvolved rectal mucosa 10 cm and 20 cm away from the malignant tumor. Histol Histopathol 2014; 29(2): 229–234.

16.

Bradbury

Gordon

. Connective tissues and stains. In: Bancroft

Stevans

(eds) Theory and practice of histological techniques. Edinburgh: Churchill Livingstone, 1982, pp. 121–144.

17.

Doube

Kłosowski

Arganda-Carreras

. BoneJ: free and extensible bone image analysis in ImageJ. Bone 2010; 47: 1076–1079.

18.

de Chaumont

Dallonqeville

Chenouard

. Icy: an open bioimage informatics platform for extended reproducible research. Nat Methods 2012; 9(7): 690–696.

19.

Rabasović

Pantelić

Jelenković

. Nonlinear microscopy of chitin and chitinous structures: a case study of two cave-dwelling insects. J Biomed Opt 2015; 20(1): 016010.

20.

Kenney

Jain

Identification of lymphatics within the colonic lamina propria in inflammation and neoplasia using the monoclonal antibody D2–40. Yale J Biol Med 2008; 81(3): 103–113.

21.

Tarpey

White

FH.

Ultrastructural morphometry of collagen from lamina propria during experimental oral carcinogenesis and chronic inflammation. J Cancer Res Clin Oncol 1984; 107(3): 183–194.

22.

Song

. Cancer-associated fibroblasts from invasive breast cancer have an attenuated capacity to secrete collagens. Int J Oncol 2014; 45(4): 1479–1488.

23.

Brugge

JS.

A stiff blow from the stroma: collagen crosslinking drives tumor progression. Cancer Cell 2009; 16(6): 455–457.

24.

Radosevich

Stypula-Cyrus

. Spatially resolved optical and ultrastructural properties of colorectal and pancreatic field carcinogenesis observed by inverse spectroscopic optical coherence tomography. J Biomed Opt 2014; 19(3): 36013.

25.

Sapudom

Rubner

Martin

. The phenotype of cancer cell invasion controlled by fibril diameter and pore size of 3D collagen networks. Biomaterials 2015; 52: 367–375.

26.

Carey

Kraning-Rush

Williams

. Biophysical control of invasive tumor cell behavior by extracellular matrix microarchitecture. Biomaterials 2012; 33(16): 4157–4165.

27.

Parekh

Weaver

AM.

Regulation of cancer invasiveness by the physical extracellular matrix environment. Cell Adh Migr 2009; 3(3): 288–292.

28.

Seewaldt

ECM stiffness paves the way for tumor cells. Nat Med 2014; 20(4): 332–333.

29.

White

Wunder

Bell

. Histologic assessment of peritumoral edema in soft tissue sarcoma. Int J Radiat Oncol Biol Phys 2005; 61(5): 1439–1445.

30.

Lopez

Charyulu

Adkins

Influence of breast cancer on thymic function in mice. J Mammary Gland Biol Neoplasia 2002; 7(2): 191–199.

31.

Carrio

Lopez

DM.

Insights into thymic involution in tumor-bearing mice. Immunol Res 2013; 57(1–3): 106–114.

32.

Vannucci

Fiserova

Horvath

. Cancer evolution and immunity in a rat colorectal carcinogenesis model. Int J Oncol 2004; 25(4): 973–981.

33.

Backman

Roy

HK.

Advances in biophotonic detection of field carcinogenesis for colon cancer risk stratification. J Cancer 2013; 4: 251–261.

34.

Dotto

GP.

Multifocal epithelial tumors and field cancerization: stroma as a primary determinant. J Clin Invest 2014; 124(4): 1446–1453.

35.

Meng

Zhou

. Could stroma contribute to field cancerization? Med Hypotheses 2010; 75(1): 26–31.

36.

Chen

Hao

Chiu

. Alteration of gene expression in normal-appearing colon mucosa of APC(min) mice and human cancer patients. Cancer Res 2004; 64(10): 3694–3700.

37.

Belshaw

Pal

Tapp

. Patterns of DNA methylation in individual colonic crypts reveal aging and cancer-related field defects in the morphologically normal mucosa. Carcinogenesis 2010; 31(6): 1158–1163.

38.

Lian

Yang

. Aberrant gene expression profile of unaffected colon mucosa from patients with unifocal colon polyp. Med Sci Monit 2015; 21: 3935–3940.

39.

Genovese

Zawada

Tosoni

. Cellular localization, invasion, and turnover are differently influenced by healthy and tumor-derived extracellular matrix. Tissue Eng Part A 2014; 20(13–14): 2005–2018.

Remodeling of extracellular matrix of the lamina propria in the uninvolved human rectal mucosa 10 and 20 cm away from the malignant tumor