Histopathology 2009, 54, 550–560. DOI: 10.1111/j.1365-2559.2009.03273.x
Progressive cellular response in the lamina propria of the colorectal adenoma–carcinoma sequence Guanglin Cui, Aping Yuan, Barthold Vonen,1 & Jon Florholmen Laboratory of Gastroenterology, Institute of Clinical Medicine, Faculty of Medicine, University of Tromsø and Department of Gastroenterology, University Hospital of North Norway, and 1Division of Gastrointestinal Surgery, University Hospital of North Norway, Tromsø, Norway Date of submission 11 March 2008 Accepted for publication 26 October 2008
Cui G, Yuan A, Vonen B & Florholmen J (2009) Histopathology, 54, 550–560
Progressive cellular response in the lamina propria of the colorectal adenoma–carcinoma sequence Aims: The lamina propria is inevitably involved in epithelial transformation. The aim was to evaluate the dynamic cellular changes in the tumour lamina propria throughout the colorectal adenoma–carcinoma sequence. Methods and results: Using immunohistochemistry and double immunohistochemistry, we examined lamina propria cellular changes in 41 colorectal adenomas, 25 colorectal cancers and 15 control tissues. The results showed that the proliferation labelling index in lamina propria cells began to increase in the precancerous lesions (adenomas) and became even higher in the colorectal cancers; these proliferative cells were
primarily identified as myofibroblasts and lymphocytes. Phenotypic analysis revealed gradually increasing lymphocytic infiltration in both the lamina propria and adenomatous epithelium, as well as myofibroblasts in the lamina propria. However, the intraepithelial macrophage density also showed a tendency to increase gradually. Furthermore, cyclooxygenase-2-expressing cell density and microvessel density gradually increased in the tumour lamina propria throughout the adenoma–carcinoma sequence. Conclusions: Progressive cellular responses in the lamina propria could be involved in the adenoma–carcinoma transition.
Keywords: adenoma–carcinoma sequence, carcinogenesis, tumour–stromal interaction Abbreviations: COX-2, cyclooxygenase-2; HRP, Horseradish Peroxidase; LI, labelling index; LSAB, Labelled Streptavidin-Biotin; MVD, microvessel density; SMA, smooth muscle actin
Introduction Colorectal cancer is the fourth most frequent cancer and has a high mortality worldwide. Colorectal cancers, according to the adenoma–carcinoma sequence theory, arise primarily in a setting where a multistep carcinogenesis event progresses from colorectal adenoma with a low grade of dysplasia to high dysplasia and, finally, to carcinoma.1 Thus, colorectal adenoma has been recognized as the main precancerous lesion
Address for correspondence: Dr G Cui, Laboratory of Gastroenterology, Institute of Clinical Medicine, University of Tromsø, N-9037, Tromsø, Norway. e-mail:
[email protected] 2009 The Authors. Journal compilation 2009 Blackwell Publishing Limited.
for most colorectal cancers. Although many molecular and genetic alterations involved in this sequence have been identified,1 the exact mechanisms of the adenoma–carcinoma transition are still not fully understood. The lamina propria is the intraepithelial connective tissue and contains a complex arrangement of cells, blood vessels, immune cells and nerves plus extracellular matrix, growth factors, regulatory molecules and remodelling enzymes. These components act in a coordinated manner to play an essential regulatory role in organ ⁄ cell development, maintenance and function.2,3 Recent studies have provided considerable evidence that the dynamic microenvironment directly influences epithelial cell behaviour and
Tumour lamina propria and colorectal cancer
transformation.4 In response to carcinogenesis, the tumour lamina propria rapidly modulates to a reactive condition in which cells proliferate, phenotypes change and extracellular matrix and growth factors synthesize. All these cellular changes are likely to create a favourable microenvironment for cancer cell survival, proliferation and metastasis.5–9 Colorectal adenoma– carcinoma transition develops slowly over many years, and the lamina propria inevitably changes so as to establish a supportive environment. The altered tumour lamina propria cells, i.e. fibroblasts ⁄ myofibroblasts and inflammatory cells (lymphocytes, macrophages, mast cells and eosinophils) in colorectal cancers have been reported in many studies.5,10–14 The real significance of these lamina propria cells in human cancers has long been a matter of debate. They were once considered important components of anticancer responses, and the increased immune cells in the tumour microenvironment were thought to predict a good prognosis;8,15 however, it has now become more clear that those cells in the tumour lamina propria are not merely a secondary response but could be important participants in disease pathogenesis and could create a favourable microenvironment for human epithelial transformed cell (tumour cell) growth under appropriate conditions.13,16 In colorectal cancer, angiogenesis, or blood vessel formation, is another significant histological change in reactive tumour lamina propria;17 this neovascularization provides the necessary nutrients for cancer growth and invasion.18 Extensive evidence now suggests that the angiogenic switch is a critical step in the growth of adenomas as well as in the adenoma–carcinoma transition.19 This angiogenesis is regulated by many factors that can be released from both the tumour cells and host cells. Cyclooxygenase (COX)-2, which is produced by both tumour lamina propria cells and tumour cells, is one of the potential growth factors and pro-angiogenic factors linked to colorectal adenoma formation20,21 and colorectal cancer development.22,23 Therefore, up-regulated COX-2 expression and angiogenesis in the tumour lamina propria may represent the early conditional change for adenoma–carcinoma transition. However, since most previous studies have examined only tumour lamina propria changes in developed colorectal cancer,10,11 dynamic cellular changes throughout the adenoma–carcinoma sequence, particularly in precancerous lesions, are still incompletely characterized. Given the above background, we hypothesized that altered cellular components in the tumour lamina propria in the precancerous stage (adenoma) could be important factors in adenoma–carcinoma transition. In this study, therefore, we characterized the cellular
551
responses regarding the proliferation rate, phenotype, COX-2 expression and angiogenesis in the tumour lamina propria components throughout this sequence.
Materials and methods A total of 41 colorectal adenomas excised completely by endoscopic polypectomy (male ⁄ female ratio 24 ⁄ 17; age 43–92 years; histological types tubular ⁄ tubulovillous ⁄ villous 27 ⁄ 12 ⁄ 2; low dysplasia ⁄ moderate dysplasia ⁄ high dysplasia 24 ⁄ 14 ⁄ 3) and 25 colorectal cancers excised by surgery (male ⁄ female ratio 21 ⁄ 4; age 42–89 years; histological types adenocarcinoma ⁄ mucinous ⁄ signet ring 22 ⁄ 2 ⁄ 1; Duke’s stages A ⁄ B ⁄ C 4 ⁄ 10 ⁄ 11) between August 2003 and April 2005 were included in this study. Morphologically normal mucosa taken from patients undergoing colonoscopic examination at our department served as the control group, and included a total of 15 (male ⁄ female ratio 9 ⁄ 6; age 30–68 years) colorectal biopsy specimens with no pathological evidence on colonoscopic and microscopic examination. No patients or control subjects had a history of regular non-steroidal anti-inflammatory drug use, other immunomodulatory treatments or colorectal adenoma ⁄ cancer. All biopsy specimens were prepared and routinely embedded in paraffin. Sections (4 lm) were cut and then stained with haematoxylin and eosin. Histological diagnoses for all the specimens were reviewed at the Department of Pathology, University Hospital of North Norway. The study was approved by the Regional Ethics Committee of Northern Norway, and written informed consent was obtained from the patients. e xa m in at ion o f l ami na p ropri a ce ll phenotype, cox- 2 -expressing cells and an giogenic cha nges by immunoh istoch emist ry To examine the changes in density of lymphocytes, macrophages, myofibroblasts and COX-2-expressing cells and microvessels in the tumour lamina propria throughout the colorectal adenoma–carcinoma sequence, immunohistochemistry was performed on 4-lm paraffin sections from control tissues, adenomas and cancers with Labelled Streptavidin-Biotin (LSAB)-2 System, Horseradish Peroxidase (HRP) kits (Dako Corp., Carpinteria, CA, USA) according to the manufacturer’s instructions and our previous published method.24–26 The following primary antibodies were used: mouse anti-CD3 monoclonal antibody to label lymphocytes, mouse antihuman CD68 (clone KP-1) monoclonal antibody to label macrophages, mouse antihuman
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
552
G Cui et al.
smooth muscle actin (SMA)-a (clone 1A4) monoclonal antibody to label myofibroblasts, mouse antihuman CD34 monoclonal antibody to label microvessels and rabbit antihuman COX-2 polyclonal antibody to reveal COX-2-expressing cells (all antibodies purchased from Dako, except for the last antibody, purchased from Cayman Chemical, Ann Arbor, MI, USA). Antibodies were incubated at 4C overnight; 3-amino-9-ethylcarbazole (Vector Laboratories, Burlingame, CA, USA) was used as chromogen and slides were counterstained with Mayer’s haematoxylin. Negative control slides for all immunohistochemistry were prepared routinely: (i) primary antibodies were substituted with the isotypematched control antibodies; (ii) secondary antibody was substituted with phosphate-buffered saline. e va lu a t i o n of t h e p ro l i fe r a t i o n l a b e l li n g index i n the tumour lami na prop ria c ells To examine proliferation index changes in the tumour lamina propria cells, immunohistochemistry was performed using 4-lm paraffin sections from control tissues, adenomas and cancers. The proliferation labelling index (LI) was determined with a mouse anti human Ki67 monoclonal antibody (1:70; BD Biosciences Pharmingen, San Diego, CA, USA). Antigen retrieval was achieved by boiling sections for 20 min in 0.01 m citrate buffer, pH 6.0. After the primary antibody was incubated at 4C overnight, slides were further developed with LSAB-2 system HRP kits (Dako) as described above. To examine the proliferative activity in different types of cells in the tumour lamina propria, double immunohistochemistry with Ki67 ⁄ CD3, Ki67 ⁄ CD68 and Ki67 ⁄ SMA-a antibodies was performed using the EnVision Doublestain System kit (Dako) according to the manufacturer’s instructions in order to examine the activation of myofibroblasts in the tumour lamina propria. In brief, slides were incubated overnight at 4C with anti-Ki67 antibody after antigen retrieval and then incubated with labelled polymer-HRP-antimouse and antirabbit antibodies for 30 min at room temperature. Peroxidase activity was detected with the enzyme substrate 3,3¢-diaminobenzidine tetrachloride. After quenching the enzyme reaction, slides were incubated in Doublestain Block at room temperature for 5 min to block endogenous phosphatase. The slides were then incubated with anti-CD3, anti-CD68 and anti-SMA-a antibodies, respectively, for 2 h at room temperature. After washing, slides were incubated with labelled polymer-alkaline phosphatase antimouse and antirabbit antibody for 30 min at room temperature. Fast Red chromogen substrate solution was used to
visualize anti-CD3, anti-CD68 and anti-SMA-a antibodies. The sections were lightly counterstained with Mayer’s haematoxylin. morphometric analysis Since macrophages and lymphocytes have been reported to infiltrate into both the lamina propria and epithelium,27,28 their numbers were evaluated in the lamina propria and epithelium, respectively. The semiquantified density grading of macrophages and lymphocytes according to the method described in our previous publication26 was employed. In brief, semiquantified scoring was done in at least five welloriented fields with abundant positive cell distribution from each slide under 400· high-power magnification and scored as nil (0), 1–19 cells ⁄ field (1+), 20–49 cells ⁄ field (2+) and >50 cells ⁄ field (3+). Myofibroblasts (labelled by SMA-a) were found at very high density in the stroma and were evaluated with light microscopy using the following criteria described by Adegboyega et al.21: negative immunoreactivity ()), 1–25% positive cells (+), 25–50% positive cells (++) and >50% positive cells (+++). The numbers of Ki67+ (nucleus) cells, COX-2-expressing cells and CD34+ microvessels located in the (tumour) lamina propria were counted in at least five well-oriented fields with abundant positive cell distribution on each slide at 400· highpower magnification. The average values were used for the statistical analysis. statistical a nalysis The results were expressed as mean ± SEM unless otherwise stated. Statistical significance was evaluated by the Mann–Whitney test and the Kruskal–Wallis test. Values of P < 0.05 were considered to be significant.
Results i n c r e a s e d pr o l i f e r a t i o n la b e l l i n g i n d e x i n t h e tu m o ur la m i n a pr o p r i a ce l l s o f c o l o r e c t a l a d e n o ma s a n d c a n c e r s The proliferation LI in tumour lamina propria cells was evaluated with Ki67 immunohistochemistry (Figure 1). The results showed that the LIs in the tumour lamina propria cells were increased in both the adenomas (Figure 1B) and cancers (Figure 1C) compared with the controls (Figure 1A). This observation was confirmed by Ki67+ cell quantification (Figure 1M), which showed a gradual increase of proliferative cell numbers in the tumour lamina propria
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
Tumour lamina propria and colorectal cancer
throughout the adenoma–carcinoma sequence relative to the controls (both P < 0.0001, Figure 1M), whereas a non-significant increase was observed from adenomas to cancers (Figure 1M, P > 0.05). Double immunohistochemistry demonstrated that most Ki67+ lamina propria cells in the adenomas and cancers were lymphocytes and myofibroblasts: the proliferation LI (labelled by Ki67; brown) in lymphocytes (labelled by CD3; red) was rarely observed in the controls (Figure 1D), but was frequently present in the tumour lamina propria of adenomas (arrows in Figure 1E) and cancers (arrows in Figure 1F). The proliferation LI (brown) in macrophages (labelled by CD68; red) was not observed in any of the three groups (Figure 1G–I). The increased colocalization of proliferation LI (labelled by Ki67, brown) with myofibroblasts (labelled by SMA-a; red) could be found largely in the tumour lamina propria of the adenomas (arrows in Figure 1K) and cancers (arrows in Figure 1L) relative to the controls (Figure 1J). These results indicated that increased proliferative activity was specifically higher in lymphocytes and myofibroblasts than in other types of cells. t u mo u r l a m i n a p r op ri a ce l l d en si t i e s wer e a l t e r e d i n co l o re c t a l a denoma s an d ca nc ers Lymphocytes Lymphocytes were observed in all the cases of controls, adenomas and carcinomas in this study. In the control mucosa, lymphocytes were mainly present in the lamina propria (arrow in Figure 2A), although some infiltrated the epithelium (arrowheads in Figure 2A). In the adenomas, high-density lymphocytes were present in the region of the tumour lamina propria close to the epithelium (the subepithelial region) (Figure 2B). In the cancers, lymphocytes in the tumour lamina propria were distributed diffusely (Figure 2C). Relative to controls (Figure 3A, white bar), the density of lymphocytes in the whole tumour lamina propria began to increase at the adenoma stage (Figure 3A, grey bar) and became even higher at the cancer stage (Figure 3A, black bar). Intraepithelial lymphocytes were frequently observed in both the adenomatous epithelium (arrowheads in Figure 2E) and cancerous epithelium (arrowheads in Figure 2F) compared with the control epithelium (arrowheads in Figure 2D); their density was slightly increased in the adenomatous epithelium (Figure 3B, grey bar) and was significantly higher in the cancerous epithelium (Figure 3B, black bar) than in the control epithelium (Figure 3B, white bar).
553
Macrophages The infiltration pattern of macrophages was similar to that of lymphocytes. In control mucosa, lymphocytes were primarily present in the lamina propria (Figure 2G) and occasionally observed in the epithelium (inset in Figure 2G). Macrophages in the tumour lamina propria of adenomas were present primarily in the subepithelial regions (Figure 2H), but were also distributed diffusely in the tumour lamina propria of cancers (Figure 2I). The scores for macrophages in the tumour lamina propria were similar in adenomas (Figure 3C, grey bar), cancers (Figure 3C, black bar) and controls (Figure 3C, white bar). Intraepithelial infiltration of macrophages was observed in the adenomas (inset in Figure 2H) and in the cancers (arrowheads in Figure 2I). The density scores of intraepithelial macrophages also showed a tendency to increase gradually throughout the adenoma–carcinoma sequence, although statistical significance was not reached (see Figure 3D). Myofibroblasts Myofibroblasts were observed in both the lamina propria and muscular layers in the tissues of all three groups. Here, we only describe their presence in the lamina propria. Myofibroblasts were observed in the whole lamina propria in the three groups within both pericryptal and non-pericryptal regions (Figure 2J–L). In the adenomas, dense myofibroblasts were observed in the whole lamina propria, but were particularly dense in the pericryptal region (Figure 2K) compared with the controls (Figure 2J). In the cancers, myo- fibroblasts were highest in the invading edge of the tumour lamina propria (stroma) (Figure 2L). When the density of myofibroblasts was graded, the grading scores were greatly increased in the adenomas (Figure 3E, grey bar) relative to the controls (Figure 3E, white bar) and increased continually towards the cancer stage (Figure 3E, black bar) to a higher level than that at the adenoma stage. gradually i ncreasing c ox - 2 -expressing cell d e n s i t y an d a n g i o ge n e s i s i n t h e tu m o u r l a m i n a p r o p r i a pa r a l l e l e d th e c o l o r e c t a l adenoma carcinoma s equence In the lamina propria of the controls, low COX-2expressing cell density was observed (Figure 2M), whereas it was increased in the tumour lamina propria of adenomas (Figure 2N), where these cells primarily aggregated in the subepithelial region of the lamina propria in the adenomas (Figure 2N) and in both the tumour lamina propria and the cancerous epithelium
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
554
G Cui et al.
in the cancers (Figure 2O). When the COX-2-expressing cells in the tumour lamina propria were quantified under the microscope, they were slightly increased at Control
the adenoma stage and significantly increased at the cancer stage (Figure 3F). Furthermore, COX-2 expression was particularly high in the cancerous epithelium
Adenoma B
A
E
Ki67(Brown) + CD3 (Red) x400
F
Ki67 (Brown) + CD68 (Red) x400 J
I
Ki67 (Brown) + CD68 (Red) x400
Ki67 (Brown) + CD68 (Red) x400 L
K
Ki67 (Brown) + SMA-alpha (Red) x400
Ki67 (Brown) + CD3 (Red) x400
Ki67(Brown) + CD3 (Red) x400 H
G
Ki67 (Red) x200
Ki67 (Red) x200
Ki67 (Red) x200 D
Carcinoma C
Ki67 (Brown) + SMA-alpha (Red) x400
Ki67 (Brown) + SMA-alpha (Red) x400
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
Tumour lamina propria and colorectal cancer
50 Proliferative index labelled by Ki67 in the lamina propria of colorecta ladenoma and carcinoma (per field)
ated with either the grade of dysplasia in the adenomas or with Duke’s stage in the cancers (data not shown).
P < 0.0001
M
555
P > 0.05
Discussion
40 P < 0.0001 30
20
10
0 Control
CRA
CRC
Figure 1. (Continued).
(inset in Figure 2O), but was rarely seen in the control and adenomatous epithelium. Angiogenesis in the lamina propria was determined by the measurement of CD34 immunoreactivitylabelled microvessel density (MVD, Figure 3F). In the adenomas, CD34-labelled microvessels were distributed in the tumour lamina propria between adenomatous epithelium (Figure 2Q). Whereas in the cancers, CD34labelled microvessels were more abundant in the tumour lamina propria (Figure 2R) than in controls (Figure 2P), MVD was particularly high in the invading edges. When CD34-labelled MVD was measured in the stained sections, a gradual increasing trend was demonstrated through the adenoma–carcinoma sequence (Figure 3F): MVD was slightly increased at the adenoma stage compared with the controls, and was significantly higher at the cancer stage than in the controls or adenomas (control versus adenoma, P > 0.05; control versus cancer, P < 0.0001; cancer versus adenoma, P < 0.001; Mann–Whitney test). The changes in lamina propria cell proliferative index, populations and angiogenesis were not associ-
The current study has demonstrated that reactive proliferation and progressive phenotypic changes in the tumour lamina propria begin at the precancerous stage (adenoma) and continue throughout the whole sequence. These findings suggest that dynamic proliferative and cellular changes in the tumour lamina propria could be involved in the transition from adenoma to carcinoma. Because the cells have a low proliferation activity and secrete only those factors necessary to maintain normal tissue homeostasis and function under physiological conditions,29 it is not surprising that the number of Ki67+ cells in the lamina propria was low in the control tissues. However, throughout the adenoma–carcinoma sequence the proliferative capacity in the tumour lamina propria was significantly increased. This proliferative change began at the early stage (adenomas) of the colorectal adenoma–carcinoma sequence; cells with high proliferative capacity were mostly identified as lymphocytes and myofibroblasts. To our knowledge, this is the first study to examine the proliferative capacity of tumour lamina propria cells. The exact mechanisms of elevated proliferation in the tumour lamina propria are thus far undetermined; many factors may be related to the initiation of tumour lamina propria activation during adenoma–carcinoma transition. Chromosomal aberrations in the transformed colorectal mucosal cells are frequently observed and may modulate the stromal response.30 Cytokines, chemokines and growth factors derived from colorectal cancer cells may also participate in the modulation of cell proliferative activity.7,31–33 Phenotypic changes are a key feature of the tumour lamina propria reaction to carcinogenesis. In the current study, progressive cell phenotype changes in the tumour lamina propria were observed throughout the adenoma–carcinoma sequence. The lymphocyte
Figure 1. Examination of proliferation labelling index (LI) in the cells in the tumour lamina propria throughout the colorectal adenoma– carcinoma sequence. The proliferation LI labelled by Ki67 is increased in the tumour lamina propria of colorectal adenomas (arrows in B) and cancers (arrows in C) compared with that in controls (arrow in A). Double IHC (D-IHC) with Ki67 ⁄ CD3, Ki67 ⁄ CD68 and Ki67 ⁄ smooth muscle actin (SMA)-a antibodies further shows that increased proliferation LI [Ki67 visualized by diaminobenzidine (DAB), brown colour] frequently involves lymphocytes (CD3 visualized by Fast Red, red colour in E for adenoma and F for cancer), and myofibroblasts (SMA-a visualized by Fast Red, red colour in K for adenoma and L for cancer), except in the controls (D for Ki67 ⁄ CD3, J for Ki67 ⁄ SMA-a). Proliferation LI (Ki67 visualized by DAB, brown colour) is rarely found in macrophages (CD68 visualized by Fast Red, red colour) in all three groups (G for control, H for adenoma and J for cancer). Counting Ki67+ cells in whole lamina propria (M) confirmed a gradually increased trend for proliferation LI thoughout the adenoma–carcinoma sequence (adenoma versus control and cancer versus control: both P < 0.0001; adenoma versus cancer: P > 0.05; Mann–Whitney test) (A–C, immunohistochemistry; D–L, double immunohistochemistry; all counterstained with haematoxylin). 2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
556
G Cui et al.
B
Adenoma CD3-IR in lamina propria C
Adenocarcima CD3-IR in lamina propria
D
CD3-IR in epithelium E
CD3-IR in epithelium F
CD3-IR in epithelium
G
CD68-IR in lamina propria H
A
Control CD3-IR in lamina propria
SMA-alpha-IR
I
CD68-IR in lamina propria
epithelium
epithelium
J
CD68-IR in lamina propria
K
SMA-alpha-IR
L
SMA-alpha-IR
M
COX-2-IR in lamina propria N
COX-2-IR in lamina propria O
COX-2-IR in lamina propria
P
CD34-IR in lamina propria Q
CD34-IR in lamina propria R
CD34-IR in lamina propria
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
Tumour lamina propria and colorectal cancer
number increased in both the tumour lamina propria and adenomatous ⁄ cancerous epithelium, beginning in the precancerous lesions (adenoma stage) and becoming significant at the cancerous stage. In addition, macrophage numbers increased non-significantly in adenomatous ⁄ cancerous epithelium. Initially, these cells were thought to be passive participants in anticancer immunity.8,15 However, recent evidence suggests that the immune cells may act as a doubleedged sword in response to tumorigenesis; they attack transformed cells when the reaction occurs in a prompt and efficient manner but, alternatively, facilitate tumorigenesis under certain conditions, such as chronic inflammation, by releasing many growth factors, cytokines and carcinogens that participate in the epithelial–stromal interaction.8,15,34–36 Whether their role is to promote or prevent cancer development probably depends on their function, which is more important than their density in the microenvironment. Several studies have shown that a functional switch can occur in the ‘education’ of tumour cells.14,16 We have previously shown that macrophages and lymphocytes are the main cellular sources of cytokines in colorectal adenomas and cancers, but that Th1 cytokine characterization at the adenoma stage is distinctly different from that at the cancer stage (increased Th1 type in adenoma and decreased in cancers).26 Thus, a functional switch of immune cells in the tumour microenvironment might contribute to adenoma–carcinoma transition.15,16,26 Myofibroblasts are the main cells in the lamina propria, and the most common marker of reactive tumour lamina propria may be the activation of fibroblasts into myofibroblasts.11,37,38 It is widely accepted that myofibroblasts are capable of remodelling tumour lamina propria tissue and play a key role in controlling tumour cell behaviour.39 Ngan et al. have previously demonstrated increased fibroblast ⁄ myofibroblast densities in the tumour stroma of colorectal cancers, an observation that is related to clinical
557
outcome.10,11 In the current study, gradually increased proliferation capacity and myofibroblast numbers were observed in the lamina propria as the sequence progressed. In addition to the changes in proliferation and density, the location of myofibroblasts was also modulated: in the adenomas, dense myofibroblasts were observed surrounding the adenomatous crypts (named pericryptal myofibroblasts). This localization pattern may provide close, direct contact with adenomatous epithelium and promote tumorigenesis by releasing numerous factors.21 In the colorectal cancers, abundant myofibroblasts were scattered in the invading front (stroma); this location has been linked to the stimulating effect on cancer cell growth and invasion via release of numerous growth factors.11,38 As previously described, angiogenesis is required for cancer growth, invasion and metastasis.2 COX-2 is a potential proangiogenic factor in many human cancers, including colorectal cancer.20 Previously, high expression of COX-2 mRNA has been demonstrated in colorectal adenomas and cancers by our own and other groups,25,40 and an angiogenic switch along the adenoma–carcinoma sequence has also been reported.18,19 Although the cancer cells can produce significant levels of COX-2, it is likely that surrounding lamina propria cells are also involved in angiogenesis.21,41 Consistent with this idea, the current study has demonstrated that COX-2-expressing cell density gradually increased in parallel with the increased MVD in the tumour lamina propria along the sequence. Since COX-2 is the most potent proangiogenic factor in colorectal cancer, the current findings suggest that the COX-2 cell density change in the tumour lamina propria is an important signal for the angiogenic switch during colorectal adenoma–carcinoma transition. The tumour lamina propria may also be influenced by factors from the intestinal wall and contents. It has been demonstrated epidemiologically that patients with chronic constipation have an increased risk for colorectal cancer, probably through over-absorption of
Figure 2. Examination of cellular phenotypic changes, cyclooxygenase (COX)-2 expressing cells and microvessels in the tumour lamina propria in colorectal adenomas and cancers with immunohistochemistry. In controls, the lymphocytes labelled by CD3 immunoreactivity are mostly located at the tumour lamina propria (A, pointed arrows), but are occasionally found in the epithelium (D, pointed arrowheads). In adenoma tissues, increased infiltrating lymphocytes are mostly distributed in the upper region of the lamina propria towards the epithelium (B, pointed arrows), with some in adenomatous epithelium (E, pointed arrowheads). In cancer tissues, increased infiltrating lymphocytes are mainly located in the tumour lamina propria (stroma) in the invading edge (C, pointed arrows) and intraepithelial lymphocytes are also seen (F, pointed arrowheads). The distribution patterns of macrophages labelled by CD68 immunoreactivity are similar to those of lymphocytes in all three groups (G–I): macrophages can be observed in both the tumour lamina propria (H,I, pointed arrows) and intraepithelial tissue of adenoma and cancer (insets in G and I, respectively, pointed arrowheads). Myofibroblasts are detected in the lamina propria (J–L for myofibroblasts labelled with smooth muscle actin-a immunoreactivity). In the tumour lamina propria of adenomas, abundant myofibroblasts are frequently observed in both the pericryptal and non-pericryptal regions (K), whereas in the tumour lamina propria of cancers, myofibroblasts are diffusely distributed in the tumour lamina propria and highly scattered in the invading edges (L). Increased density of COX-2-expressing cells is seen in the lamina propria of adenoma (N) and becomes more significant in cancers (O) compared with normal controls (M). Microvessel density in the stroma also shows a similar trend (P for control, Q for adenoma and R for cancer). 2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
G Cui et al.
2
1
0
Control Adenoma Cancer
Grades of intraepithelial macrophages in the colorectal adenoma and cancer
D
C
1.5
P > 0.05
1.0
0.5
0.0
20
15
Control Adenoma Cancer
P < 0.05 P > 0.05
2.0 1.5
P > 0.05
1.0 0.5 0.0
E
G CD34 (+) microvessel density in the lamina propria of the colorectal adenoma and cancer (per field)
B
3
Control Adenoma Cancer P < 0.0001 P < 0.01 P < 0.0002
2
1
0
Grades of macrophages in the lamina propria of the colorectal adenoma and cancer
P < 0.05 P < 0.05 P > 0.05
3
P > 0.05
2
1
0
F COX-2 expressing cells in the lamina propria of the colorectal adenoma and cancer (per field)
3
Grades of intraepithelial lymphocytes in the colorectal adenoma and cancer
Grades of lymphocytes in the lamina propria of the colorectal adenoma and cancer
A
Grades of myofibroblasts in the lamina propria of the colorectal adenoma and cancer
558
12.5
Control Adenoma Cancer P < 0.01 P < 0.05
10.0
Control Adenoma Cancer
P > 0.05 7.5 5.0 2.5 0.0
Control Adenoma Cancer
P < 0.0001 P < 0.0001 P > 0.05
10
5
0
Control Adenoma Cancer
Figure 3. Graphic analysis of the cellular phenotypic changes, cyclooxygenase (COX)-2-expressing and microvessel (MVD) density changes in the tumour lamina propria of colorectal adenomas and cancers. The scores for infiltrating lymphocyte grading in either the lamina propria (A) or epithelium (B) of both adenomas (grey bar) and cancers (black bar) are significantly different from controls (white bar) (both P < 0.05, non-parametric Kruskal–Wallis test), whereas the score of infiltrating macrophage grading in the tumour lamina propria of adenoma (C, grey bar) is on the same scale as in controls (white bar) (P > 0.05, non-parametric Kruskal–Wallis test). The scores for infiltrating macrophage grading (D) in the adenomatous epithelium and cancerous epithelium gradually increase, although not significantly (P > 0.05, non-parametric Kruskal–Wallis test). The scores for myofibroblast grading in the lamina propria (E) increase remarkably (adenoma versus control: P < 0.0002; cancer versus control: P < 0.0001; and adenoma versus cancer: P < 0.01, Mann–Whitney test). The density of COX-2-expressing cells (F) and CD34-labelled MVD (G) in the tumour lamina propria both show a gradually increasing trend thoughtout the adenoma–carcinoma sequence (COX-2: P < 0.01; MVD: P < 0.0001; both non-parametric Kruskal–Wallis test).
carcinogens from stool.42 Since an existing adenoma has long-term contact with the stool, it will be important to evaluate whether the chemical components of the stool accelerate lamina propria changes during adenoma–carcinoma transition; notably, the colonic luminal contents have been found to be altered in patients with colorectal adenomas and have an impact on human colonic cancer cell lines.43–45 Finally, hyperplastic polyps are also linked to colonic cancer that appears via the recently characterized
sessile serrated adenomas,46,47 so comparison of the lamina propria alterations between hyperplastic polyps and adenomatous polyps would further aid the understanding of the developmental mechanisms of colonic cancer.
Acknowledgements This work was financially supported by grants from Medical Research Program, Northern Norway Regional
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
Tumour lamina propria and colorectal cancer
Health Authority to G.C. (SFP-44-04) and Erna & Olav Aakres Foundation to G.C. (A4783). We express our sincere gratitude to Ms Dana Frederick (Department of Cell Biology, University of Massachusetts Medical School) for manuscript proofreading.
References 1. Khosraviani K. Colorectal adenoma–carcinoma sequence. Gut 1996; 39; 342. 2. Tuxhorn JA, Ayala GE, Rowley DR. Reactive stroma in prostate cancer progression. J. Urol. 2001; 166; 2472–2483. 3. Wernert N, Locherbach C, Wellmann A, Behrens P, Hugel A. Presence of genetic alterations in microdissected stroma of human colon and breast cancers. Anticancer Res. 2001; 21; 2259–2264. 4. Sugiyama Y, Farrow B, Murillo C et al. Analysis of differential gene expression patterns in colon cancer and cancer stroma using microdissected tissues. Gastroenterology 2005; 128; 480– 486. 5. Banner BF, Sonmez-Alpan E, Yousem SA. An immunophenotypic study of the inflammatory cell populations in colon adenomas and carcinomas. Mod. Pathol. 1993; 6; 295–301. 6. Bhowmick NA, Neilson EG, Moses HL. Stromal fibroblasts in cancer initiation and progression. Nature 2004; 432; 332– 337. 7. De Wever O, Mareel M. Role of tissue stroma in cancer cell invasion. J. Pathol. 2003; 200; 429–447. 8. Whiteside TL. The role of immune cells in the tumor microenvironment. Cancer Treat. Res. 2006; 130; 103–124. 9. Almholt K, Johnsen M. Stromal cell involvement in cancer. Recent Results Cancer Res. 2003; 162; 31–42. 10. Ngan CY, Yamamoto H, Seshimo I et al. Quantitative evaluation of vimentin expression in tumour stroma of colorectal cancer. Br. J. Cancer 2007; 96; 986–992. 11. Tsujino T, Seshimo I, Yamamoto H et al. Stromal myofibroblasts predict disease recurrence for colorectal cancer. Clin. Cancer Res. 2007; 13; 2082–2090. 12. Banner BF, Savas L, Baker S, Woda BA. Characterization of the inflammatory cell populations in normal colon and colonic carcinomas. Virchows Arch. B, Cell Pathol. Incl. Mol. Pathol. 1993; 64; 213–220. 13. Lieubeau B, Heymann MF, Henry F, Barbieux I, Meflah K, Gregoire M. Immunomodulatory effects of tumor-associated fibroblasts in colorectal-tumor development. Int. J. Cancer 1999; 81; 629–636. 14. Pollard JW. Tumour-educated macrophages promote tumour progression and metastasis. Nat. Rev. Cancer 2004; 4; 71–78. 15. Brigati C, Noonan DM, Albini A, Benelli R. Tumors and inflammatory infiltrates: friends or foes? Clin. Exp. Metastasis 2002; 19; 247–258. 16. Lewis CE, Pollard JW. Distinct role of macrophages in different tumor microenvironments. Cancer Res. 2006; 66; 605–612. 17. Hanrahan V, Currie MJ, Gunningham SP et al. The angiogenic switch for vascular endothelial growth factor (VEGF)-A, VEGF-B, VEGF-C, and VEGF-D in the adenoma–carcinoma sequence during colorectal cancer progression. J. Pathol. 2003; 200; 183–194. 18. Shpitz B, Gochberg S, Neufeld D et al. Angiogenic switch in earliest stages of human colonic tumorigenesis. Anticancer Res. 2003; 23; 5153–5157.
559
19. Staton CA, Chetwood AS, Cameron IC, Cross SS, Brown NJ, Reed MW. The angiogenic switch occurs at the adenoma stage of the adenoma carcinoma sequence in colorectal cancer. Gut 2007; 56; 1426–1432. 20. Khan KN, Masferrer JL, Woerner BM, Soslow R, Koki AT. Enhanced cyclooxygenase-2 expression in sporadic and familial adenomatous polyposis of the human colon. Scand. J. Gastroenterol. 2001; 36; 865–869. 21. Adegboyega PA, Ololade O, Saada J, Mifflin R, Di MariJF, Powell DW. Subepithelial myofibroblasts express cyclooxygenase-2 in colorectal tubular adenomas. Clin. Cancer Res. 2004; 10; 5870– 5879. 22. Wendum D, Masliah J, Trugnan G, Flejou JF. Cyclooxygenase-2 and its role in colorectal cancer development. Virchows Arch. 2004; 445; 327–333. 23. DuBois RN, Smalley WE. Cyclooxygenase, NSAIDs, and colorectal cancer. J. Gastroenterol. 1996; 31; 898–906. 24. Cui G, Koh TJ, Chen D et al. Overexpression of glycine-extended gastrin inhibits parietal cell loss and atrophy in the mouse stomach. Cancer Res. 2004; 64; 8160–8166. 25. Yuan A, Steigen S, Goll R et al. Dendritic cell infiltration pattern along the colorectal adenoma–carcinoma sequence. APMIS 2008; 116; 445–456. 26. Cui G, Goll R, Olsen T et al. Reduced expression of microenvironmental Th1 cytokines accompanies adenomas-carcinomas sequence of colorectum. Cancer Immunol. Immunother. 2007; 56; 985–995. 27. Naito Y, Saito K, Shiiba K et al. CD8+ T cells infiltrated within cancer cell nests as a prognostic factor in human colorectal cancer. Cancer Res. 1998; 58; 3491–3494. 28. Sickert D, Aust DE, Langer S, Haupt I, Baretton GB, Dieter P. Characterization of macrophage subpopulations in colon cancer using tissue microarrays. Histopathology 2005; 46; 515–521. 29. Owens GK. Regulation of differentiation of vascular smooth muscle cells. Physiol. Rev. 1995; 75; 487–517. 30. Fijneman RJ, Carvalho B, Postma C, Mongera S, van Hinsbergh VW, Meijer GA. Loss of 1p36, gain of 8q24, and loss of 9q34 are associated with stroma percentage of colorectal cancer. Cancer Lett. 2007; 258; 223–229. 31. Hankemeier S, Keus M, Zeichen J et al. Modulation of proliferation and differentiation of human bone marrow stromal cells by fibroblast growth factor 2: potential implications for tissue engineering of tendons and ligaments. Tissue Eng. 2005; 11; 41–49. 32. Salgado R, Benoy I, Vermeulen P, van Dam P, Van Marck E, Dirix L. Circulating basic fibroblast growth factor is partly derived from the tumour in patients with colon, cervical and ovarian cancer. Angiogenesis 2004; 7; 29–32. 33. Li F, Cao Y, Townsend CM Jr, Ko TC. TGF-beta signaling in colon cancer cells. World J. Surg. 2005; 29; 306–311. 34. Bronte V, Cingarlini S, Marigo I et al. Leukocyte infiltration in cancer creates an unfavorable environment for antitumor immune responses: a novel target for therapeutic intervention. Immunol. Invest. 2006; 35; 327–357. 35. de Visser KE, Eichten A, Coussens LM. Paradoxical roles of the immune system during cancer development. Nat. Rev. Cancer 2006; 6; 24–37. 36. Mueller MM, Fusenig NE. Friends or foes – bipolar effects of the tumour stroma in cancer. Nat. Rev. Cancer 2004; 4; 839–849. 37. De Wever O, Mareel M. Role of myofibroblasts at the invasion front. Biol. Chem. 2002; 383; 55–67.
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
560
G Cui et al.
38. Martin M, Pujuguet P, Martin F. Role of stromal myofibroblasts infiltrating colon cancer in tumor invasion. Pathol. Res. Pract. 1996; 192; 712–717. 39. Desmouliere A, Guyot C, Gabbiani G. The stroma reaction myofibroblast: a key player in the control of tumor cell behavior. Int. J. Dev. Biol. 2004; 48; 509–517. 40. Yuri M, Sasahira T, Nakai K, Ishimaru S, Ohmori H, Kuniyasu H. Reversal of expression of 15-lipoxygenase-1 to cyclooxygenase-2 is associated with development of colonic cancer. Histopathology 2007; 51; 520–527. 41. Bamba H, Ota S, Kato A, Adachi A, Itoyama S, Matsuzaki F. High expression of cyclooxygenase-2 in macrophages of human colonic adenoma. Int. J. Cancer 1999; 83; 470–475. 42. Watanabe T, Nakaya N, Kurashima K, Kuriyama S, Tsubono Y, Tsuji I. Constipation, laxative use and risk of colorectal cancer: The Miyagi Cohort Study. Eur. J. Cancer 2004; 40; 2109–2115.
43. Glinghammar B, Rafter J. Colonic luminal contents induce cyclooxygenase 2 transcription in human colon carcinoma cells. Gastroenterology 2001; 120; 401–410. 44. Geltner-Allinger U, Branting C, Rafter JJ. Effect of lipid extracts of fecal water from colorectal tumor patients on DNA synthesis in Swiss 3T3 cells. Scand. J. Gastroenterol. 1994; 29; 736–739. 45. Haza AI, Glinghammar B, Grandien A, Rafter J. Effect of colonic luminal components on induction of apoptosis in human colonic cell lines. Nutr. Cancer 2000; 36; 79–89. 46. Higuchi T, Sugihara K, Jass JR. Demographic and pathological characteristics of serrated polyps of colorectum. Histopathology 2005; 47; 32–40. 47. Jass JR, Baker K, Zlobec I et al. Advanced colorectal polyps with the molecular and morphological features of serrated polyps and adenomas: concept of a ‘fusion’ pathway to colorectal cancer. Histopathology 2006; 49; 121–131.
2009 The Authors. Journal compilation 2009 Blackwell Publishing Ltd, Histopathology, 54, 550–560.
