SEARCH

SEARCH BY CITATION

Keywords:

  • T cell;
  • IL-17;
  • IFNγ;
  • FoxP3;
  • cancer;
  • inflammation

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Colorectal cancer is one of the five leading causes of cancer mortality worldwide. The mechanisms of pathogen clearance, inflammation and regulation by T cells in the healthy bowel are also important in controlling tumor growth. The majority of studies analyzing T cells and their relationship to colorectal tumor growth have focused on individual T cell markers or gene clusters and thus the complexity of the T cell response contributing to the growth of the tumor is not clear. We have studied the T cells in colorectal cancer patients and have defined a unique T cell signature for colorectal tumor tissue. Using a novel analytical flow cytometric approach in concert with confocal microscopy, we have shown that the tumor has a lower frequency of effector T cells (CD69+), but a higher frequency of both regulatory (CD25hi Foxp3+) and inflammatory T cells (IL-17+) compared with associated nontransformed bowel tissue. We have also identified minor populations of T cells expressing conventional markers of both inflammatory and regulatory T cells (CD4+IL-17+Foxp3+) in the tumor tissue. These cells may represent intermediate populations or they may dictate an inflammatory versus regulatory function in surrounding T cells. Together, these data describe an immune microenvironment in colorectal cancer unique to the tumor tissue and distinct from the surrounding healthy bowel tissue, and this distinct environment is reflected by a gradient of T cells expressing markers of multiple T cell populations. These findings may be used to improve diagnosis and prognosis of colorectal cancer patients.

T cells in the gut lymphoid tissue are subject to substantial regulation to ensure that pathogenic microorganisms are eliminated while commensal bacteria are well tolerated. This regulation involves several T cell populations with inflammatory or regulatory functions and the balance of these can determine the outcome of an immune response to an infection.1 Tumors arising in the gut may therefore be subject to different immune regulation than tumors in other sites since they have arisen in an immunomodulated environment. It is then likely that the gut-associated tumors, such as colorectal cancer, will have a unique immune microenvironment influencing tumor initiation and growth.

There is considerable evidence that T cells are important in destroying tumors, and colorectal cancer patients with a high T cell infiltrate into the tumor are more likely to have a positive outcome.2–8 Regulatory T cells act to down regulate effector immune responses, and high frequencies of these cells infiltrating tumor tissue have been associated with poor patient outcome in many cancers, including colorectal cancer and associated metastases.9–13 Paradoxically, in some studies of colorectal cancer, a high frequency of regulatory T cells has been associated with improved patient outcome.14 The high frequency of IL-17-producing pro-inflammatory T cells in the bowel also supports tumor growth.15–17

T cells are not easily divisible into distinct populations and there is considerable plasticity in the populations of T cells, particularly in inflammation-prone mucosal environments.18–20 The majority of studies analyzing T cells and their relationship to colorectal cancer growth have focused on individual T cell markers4, 21 and/or clearly defined “classical” T cell populations.17 The complexity of T cell interactions in the tumor environment has only recently begun to be studied—Svensson et al. showed a correlation between increasing frequencies of regulatory T cells and decreasing frequencies of effector T cells.12 The T cell infiltrate into tumor tissue in regard to those cells co-expressing markers of multiple T cell populations has not yet been well characterized.

We propose that the immunomodulated environment of the healthy bowel creates a situation whereby naturally arising cancers can grow with less immune control than tumors at other sites. Our hypothesis is that the tumor tissue consists of an immune microenvironment distinct, but deriving from, that of healthy associated bowel tissue, reflected by a gradient of T cells expressing markers of multiple T cell populations.

In order to determine whether a distinct immune microenvironment exists in colorectal cancer, we investigated colorectal tumor tissue and compared it with that of the normal bowel tissue from the same patients. We devised a novel strategy to investigate 64 distinct populations of T cells to take into account the extensive plasticity of T cell surface markers and functions. This has the advantage of revealing multiple populations, rather than previously published studies that analyzed individual markers of cells, or gene expression studies that cannot separate individual cells. Despite the large number of studies analyzing immune cells in colorectal cancer (reviewed in Ref.22), to our knowledge, this is the first study to undertake such an extensive analysis of multiple immune markers to define T cell populations infiltrating tumor tissue versus normal bowel tissue in colorectal cancer patients. Our research shows that the tumor has an immune microenvironment distinct from normal bowel tissue and that a population of T cells co-expressing inflammatory and regulatory markers may play a role in maintaining this difference.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Clinical samples

Patients undergoing elective surgery for colorectal cancer at Dunedin Hospital were invited to participate. The study was approved by the Lower South Regional Ethics Committee and all patients gave signed informed consent.

Resected specimens were dissected by a pathologist, and fresh samples were obtained from the tumor and uninvolved normal colon. Separate samples were obtained from the outer margin (S1) and central component (S3) of each tumor as determined by the pathologist. The normal bowel samples were obtained from macroscopically normal appearing colon more than 10 cm from the tumor. The samples were transported directly to the laboratory for immediate processing.

Patients were entered into a prospectively maintained database with detailed clinicopathological information and clinical follow-up (Supporting Information Table 1).

Sample preparation

Fresh tissue samples were weighed then washed with phosphate-buffered saline (PBS; 0.8% NaCl; ThermoFisher Scientific, Waltham MA, 0.114% Na2HPO4; VWR International, Radnor Penn; 0.02% KH2PO4, Merck, Whitehouse Station NJ; 0.02% KCl, VWR International and distilled water). Tissue was manually dissociated and suspended in 5 ml complete RPMI (supplemented with 100 U/ml Penicillin, 100 μg/ml Streptomycin, 55 μM 2-mercaptoethanol, all from Gibco; Invitrogen, Carlsbad CA) containing 0.5 mg/ml of collagenase (Sigma-Aldrich, St Louis, MO). Half of the sample was then removed for lymphocyte stimulation with 10 ng/ml phorbol 12-myristate 13-acetate (PMA) and 500 ng/ml ionomycin (both from Sigma-Aldrich). A 4-hr incubation period was allowed for digestion at 37°C for both stimulated and ex vivo fractions. Brefeldin A (BFA, Sigma-Aldrich) 1 μg/ml was added after 2 hr.

Flow cytometry

Cells from tissue sample suspensions were incubated in 1 ml PBS with 1 μl per sample of LIVE/DEAD Fixable Red Dead Stain (Invitrogen) for 30 min on ice in darkness. Cells were washed and re-suspended in 10 μl of fluoresence activated cell sorting (FACS) buffer ((PBS, 0.01% sodium azide (VWR International), 0.5% fetal calf serum (FCS; PAA Strasse, Pasching, Austria), 0.075% EDTA; VWR International) containing antibodies against surface markers CD3-APCCy7 (HIT3a, BioLegend, San Diego, CA), CD4-BDHorizonV450 (RPA-T4, Becton Dickinson, Franklin Lakes NJ), CD8-BDHorizonV500 (RPA-T8, Becton Dickinson), CD25-PECy7 (BC96, BioLegend) and CD69-APC (FN50, BioLegend). After 30 min on ice, the samples were washed and fixed in PBS containing 1% paraformaldehyde (PFA; Sigma-Aldrich). Cells were washed and resuspended with 1 ml permeabilization buffer (FACS buffer + 0.5% saponin (EMD Biosciences, San Diego CA)) for 1 hr, then washed and incubated for 30 min in permeabilization buffer containing intracellular antibodies against FOXP3-FITC (206D, BioLegend), IL17-PE (BL168, BioLegend) and IFNγ-PerCpCy5.5 (4S.B3, BioLegend). Flow cytometry was conducted using FACSAria IIu (Becton Dickinson). Single-stain controls were derived from a pooled suspension of samples. Anti-Mouse Compensation Beads (Becton Dickinson) and ArC Beads (for the LIVE/DEAD Fixable Red Dead Stain, Invitrogen) were used for compensation according to the manufacturer's instructions.

Data were acquired using FACSDiva (Version 6.1.3, Becton Dickinson) and analyzed using FlowJo (Version 9.2, Tree Star, Ashland OR). Live CD4+ and CD8+ T cells were gated as shown in Supporting Information Figure 1a. Population markers for CD25, CD69, IL-17, IFNγ and FOXP3 were gated against unstained tissue-matched controls and/or isotype controls (Supporting Information Fig. 1b). Bisecting gates established from population marker histograms were combined and compared to form a Boolean gating strategy. Markers evaluated as AND, OR, AND NOT and OR NOT implications of five populations led to 32 possible population combinations for both CD4+ and CD8+ T cell populations. These population values were reported as a frequency of the above-mentioned parents.

Microscopy

Confocal microscopy

About 10-μm serial sections of fresh-frozen normal bowel and tumor tissue were fixed with ice cold acetone then blocked with 0.25% casein in Tris-buffered saline (TBS; 0.88% NaCl, ThermoFisher, 0.24% tris(hydroxymethyl)aminomethane, (VWR International) and distilled water). Primary antibodies against human CD3 (mouse IgG2a, HIT3a, 1:50, Biolegend), CD4 (mouse IgG2b, OKT4, 1:50, Biolegend) and IL17 (mouse IgG1, ebio64DEC17, 1:50, eBioscience, San Diego CA), TBET (mouse IgG1, 4B10, 1:20, BioLegend) or FOXP3 (mouse IgG1, 206D, 1:50, BioLegend) were prepared in TBS containing 1% FCS and incubated on sections for 1 hr at room temperature. Secondary antibodies against mouse isotypes IgG2a DyLight 488 (poly24092, BioLegend), IgG2b AlexaFluor 555 and IgG1 AlexaFluor 647 (both from Molecular Probes) at 1:200 each with DAPI ((4′,6-diamidino-2-phenylindole) Molecular Probes) at 1:2000 were prepared and incubated on sections for 1 hr. Sections were visualized with Zeiss LSM 710 Upright confocal microscope and analyzed using ImageJ (version 1.43u, NIH, Bethesda MD) and Cytosketch (Cytocode, Auckland, New Zealand).

Immunohistochemistry

Five-micrometer formalin-fixed paraffin-embedded sections were antigen retrieved using KOS Microwave histoSTATION (Milestone, Sorisole, Italy) according to the manufacturer's instructions in citrate buffer pH 6.0. Ki67 primary antibody (Cell Marque, Rocklin CA) was stained with Envision Dual Link (Dako, Glostrup Denmark), secondary antibody followed by Diaminobenzidine (Cell Marque). Anti CD3 (Cell Marque) was stained with alkaline phosphatase (Abcam, Cambridge, UK) secondary antibody and AP Fast Red naphthol (Cell Marque). Hemotoxylin staining was included for connective tissue visualization. Tissue sections were viewed with an Olympus IX71 inverted microscope and analyzed by analySIS LS Research software (Version 3.1, Olympus Soft Imaging Solutions).

Statistical analysis

All data are expressed as mean and SEM of the frequency as a percentage of the parent population, for example, CD4+ T cells. Wilcoxon test was used for matched and unmatched samples; Tukey HSD was used for pairwise comparisons. These analyses were performed using the statistical software package R (version 2.12, Vienna Austria). Grouped analyses by two-way ANOVA were performed using Graphpad Prism (version 5.0a, La Jolla CA).

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Fifty-five patients were recruited to the study. The demographic and clinical details are summarized in Supporting Information Table 1. Thirty-nine patients had matching tumor and non-tumor bowel samples. Twenty-nine of the 39 patients had S3 and S1 tumor samples, the other 10 had S1 tumor samples only.

Similar frequency of T cells infiltrating tumor and non-tumor bowel tissue of colorectal cancer patients

We investigated the immune microenvironment of tumor versus non-tumor bowel tissue of colorectal cancer patients. The frequency of CD3+ T cells was calculated on live lymphocytes defined by size (Supporting Information Fig. 1a). A similar frequency of CD3+ (T) cells was present in tumor and non-tumor bowel tissue from the same colorectal cancer patients (Table 1). We included the use of a dye to discriminate live and dead cells since tumor infiltrating cells can have low viability—only live cells were included in our analyses (in our laboratory, the average frequency of live CD3+ T cells is 71% recovered from either tumor tissue or healthy bowel tissue). Recovered T cells were not cultured in IL-2 to enhance survival since this may have altered the phenotype of the cells. Both CD4+ and CD8+ T cell populations were also present in similar frequencies in each tissue (Table 1). We were therefore able to accurately compare T cell populations from paired tissues.

Table 1. T cell infiltrates into tissue
inline image

Activated effector T cells, regulatory T cells and inflammatory T cells infiltrate both tumor and non-tumor bowel tissue

Figure 1a shows that proliferating T cells (Ki67+CD3+) were present in both tumor tissue and healthy bowel tissue of colorectal cancer patients.

thumbnail image

Figure 1. Activated effector, regulatory and inflammatory T cells infiltrate tumor tissue and healthy bowel tissue of colorectal cancer patients. (a) Tumor tissue (top) and healthy (normal) bowel tissue (bottom) incubated with anti-CD3 (red) and Ki67 (brown). Arrows show co-localized staining. (bd) Confocal microscopy of tumor tissue (top panels) or healthy bowel tissue (bottom panels) incubated with DAPI (grey) CD3 (green) and CD4 (orange), and IL-17 (red; b), T-bet (red; c) or FoxP3 (red; d). Arrows show co-localized staining. Images are from matched tumor tissue and healthy gut tissue from one representative colorectal cancer patient. Similar data are seen from 10 other patients.

Download figure to PowerPoint

Activated T cells with many different phenotypes or functions have been shown to be present in tumors, including IFNγ+ effector T cells23, FoxP3+ regulatory T cells14 and IL-17+ αβ24 or γδ T cells.25, 26 To investigate the composition of the proliferating infiltrating T cells, we used confocal microscopy of tumor and healthy tissue. T cells were identified throughout both tumor and non-tumor bowel tissue (Fig. 1). At least three distinct populations of T cells were identified infiltrating both tumor and non-tumor bowel tissue; inflammatory (IL-17+; Fig. 1b), effector (T-bet+; Fig. 1c) and regulatory (FoxP3+; Fig. 1d). These data show that T cells identified throughout the rest of this study are bona fide infiltrating T cells, and not T cells present only in blood vessels within the tissue. In order to compare T cell infiltration into the tumor microenvironment versus the normal non-tumor bowel tissue microenvironment, we devised a flow cytometric approach that allowed us to quantify multiple T cell populations simultaneously.

Activated T cells and regulatory T cells exist in a reciprocal relationship in tumor and non-tumor bowel tissue

To examine the environments of tumor versus healthy tissue in colorectal cancer patients, while taking into account the issue of T cell plasticity and the lack of defined markers, we devised a novel analytical strategy based on flow cytometry. Matched samples of tumor tissue and healthy bowel tissue from the same patient were incubated with antibodies to CD3, CD4 and CD8 (T cells) as well as CD69 (activated T cells), CD25, FoxP3 (as a transcriptional marker of regulatory function in T cells), IFNγ (as an anti-tumor effector cytokine27) and IL-17 (as an inflammatory cytokine) directly ex vivo, or after a 4-hr stimulation to induce detectable cytokine production (but not de novo protein synthesis). Attempts to measure IFNγ or IL-17 production following culture with tumor cells only were not successful-expression of cytokines was too low to be detected. Cells were gated using a Boolean approach that allowed analysis of each marker alone or in every possible combination with every other marker, providing 64 distinct T cell populations (Supporting Information Fig. 1).

First, we analyzed effector T cell populations and observed a statistically significant decrease in the frequency of activated effector CD4+ T cells (cells expressing CD69 but not CD25 or Foxp3) in tumor tissue compared with non-tumor bowel tissue of patients (Fig. 2a). We were unable to find a difference in CD4+ or CD8+ T cells co-expressing CD69 and IFNγ (Figs. 2b and 2c).

thumbnail image

Figure 2. Tumor tissue consists of a lower frequency of effector T cells than matched healthy bowel tissue. Boolean gating of flow cytometry data from tissue sample analysis revealed distinct T cell populations. Dot plots show individual patient means. (a) Frequency of ex vivo isolated CD4+ CD69+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue; inset represents paired individual frequencies (n = 39). (b) Frequency of stimulated CD4+ CD69+IFNγ+ T cells in tumor tissue compared with healthy bowel tissue. (c) Frequency of stimulated CD8+ CD69+IFNγ+ T cells in tumor tissue compared with healthy bowel tissue. Unmatched Wilcoxon test, ‡p < 0.05, ‡‡p < 0.01, ‡‡‡p < 0.001. Matched Wilcoxon test, §p < 0.05, §§p < 0.01, §§§p < 0.001, ns: not significant.

Download figure to PowerPoint

However, regulatory CD4+ T cell populations (CD25hi or CD25hiFoxP3+) were increased in frequency in tumor tissue compared with healthy bowel tissue (Figs. 3a and 3b). This was not true of CD8+FoxP3+ regulatory T cells (Fig. 3c). When tissue from the outer margin of the tumor (designated S1) was compared with tissue from the central area (S3), we found that the frequency of regulatory CD4+ T cell populations increased (Fig. 4b) and that of effector CD4+ (CD69+) T cells decreased (Fig. 4a) with increasing tumor depth. Together, these data imply a reciprocal relationship between effector and regulatory T cell populations in the tumor and healthy bowel tissue and demonstrate distinct immune microenvironments between two closely associated tissues in colorectal cancer patients.

thumbnail image

Figure 3. Tumor tissue consists of a higher frequency of regulatory T cells than matched healthy bowel tissue. Boolean gating of flow cytometry data from tissue sample analysis revealed distinct T cell populations. Dot plots show individual patient means. (a) Frequency of ex vivo isolated CD4+CD25hi or CD4+CD25hiFoxP3+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue. (b) Paired individual frequencies of CD4+ CD25hi or CD4+CD25hiFoxP3+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue (n = 39). (c) Frequency of ex vivo isolated CD8+CD25hiFoxP3+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue. Unmatched Wilcoxon test, ‡‡‡p < 0.001. Matched Wilcoxon test, §§§p < 0.001. ns: not significant.

Download figure to PowerPoint

thumbnail image

Figure 4. Tumor tissue is an immune microenvironment distinct from matched healthy bowel tissue, consisting of a higher frequency of regulatory T cells and a lower frequency of effector T cells. Boolean gating of flow cytometry data from tissue sample analysis revealed distinct T cell populations. Dot plots show individual patient means. (a) Frequency of CD4+CD69+ T cells in tumor (outer margin (S1) and central area (S3)) tissue and healthy bowel tissue from the same patients. (b) Frequency of CD4+CD25hi or CD4+CD25hiFoxP3+ T cells in tumor (outer margin (S1) and central area (S3)) tissue and healthy bowel tissue from the same patients. (c) Frequency of CD4+69+IL-17+ T cells in tumor (outer margin (S1) and central area (S3)) tissue and healthy bowel tissue from the same patients. Tukey HSD test, *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant.

Download figure to PowerPoint

The immune microenvironment of the tumor is enriched for inflammatory T cells producing IL-17

IL-17-producing T cells are reported to be over-represented in the gut tissue28 and have been shown to have cancer-promoting functions.29 We found that Th17 cells (CD4+CD69+IL-17+ or CD4+IL-17+) were not present at higher frequencies in tumor tissue compared with non-tumor bowel tissue (Fig. 5a); however, when paired samples from individual patients were analyzed, there was a statistically significant increase in CD4+IL-17+ T cells in the tumor tissue compared with that of the healthy bowel. There was no increase in frequency of Th17 cells with increasing tumor depth (Fig. 4c). Interestingly, CD8+CD69+IL-17+ T cells were also increased in tumor tissue compared with healthy bowel tissue (Fig. 5b). These data indicate the presence of an immune environment of the tumor higher in both inflammatory and regulatory T cells than in normal bowel tissue.

thumbnail image

Figure 5. Presence of unconventional regulatory and inflammatory T cell populations at higher frequencies in tumor tissue than healthy bowel tissue. Boolean gating of flow cytometry data from tissue sample analysis revealed distinct T cell populations. Dot plots show individual patient means. (a) Frequency of stimulated CD4+CD69+IL-17+ or CD4+IL-17+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue; inset represents paired individual frequencies (CD4+IL-17+). (b) Frequency of stimulated CD8+CD69+IL-17+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue. (c) Frequency of stimulated CD4+ IL-17+FoxP3+ or CD4+IL-17+FoxP3+CD25hi T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue. (d) Paired individual frequencies of CD4+IL-17+FoxP3+, CD4+IL-17+FoxP3+CD25hi or CD4+CD69+IL-17+FoxP3+ T cells in tumor tissue in colorectal cancer patients and healthy bowel tissue (n = 19). Unmatched Wilcoxon test, ‡p < 0.05, ‡‡p < 0.01, ‡‡‡p < 0.001. Matched Wilcoxon test, §p < 0.05, §§p < 0.01, §§§p < 0.001, ns: not significant.

Download figure to PowerPoint

Tumor tissue contains higher frequencies than normal bowel tissue of T cells co-expressing markers and cytokines typical of both regulatory and inflammatory T cells

IL-17 has been associated with several T cell phenotypes, including “intermediate” populations of cells co-expressing IFNγ (Th1/Th17 cells) or FoxP3 (Th17/Treg cells). Our unique approach to T cell analysis in the tumor tissue allowed us to describe the distribution of these minor T cell populations in relation to tumor growth. Whereas the frequency of stimulated CD8+IL-17+IFNγ+ or CD4+IL-17+IFNγ+ T cells was not different between tumor tissue and non-tumor bowel tissue (data not shown), the frequency of CD4+IL-17+FoxP3+ or CD4+IL-17+CD25hiFoxP3+T cells was significantly increased in the tumor tissue compared with non-tumor bowel tissue (Figs. 5c and 5d).

Together, these data demonstrate that the tumor immune microenvironment differs considerably from associated non-tumor bowel tissue and is characterized by an increased infiltrate of regulatory and inflammatory T cells, as well as a population of T cells expressing both inflammatory and regulatory markers.

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Several recent studies have shown an important role for T cells in controlling colorectal cancer.4, 21, 30 These excellent studies have highlighted a role for effector and memory CD4+ and CD8+ T cells, and production of IFNγ, in agreement with data from animal models.23, 27 Analysis of regulatory T cells in colorectal cancer patients has shown both positive and negative correlations with patient outcomes,31 and an inverse association with activated conventional T cells.12 Our hypothesis was that the tumor tissue immune microenvironment would be reflected by a gradient of T cells expressing markers of multiple T cell populations, thus demonstrating the complexity and local influence of the immunomodulated site of the bowel.

In order to understand the role of T cell populations in the context of colorectal cancer, we had to devise a novel analytical strategy that allowed us to simultaneously study 64 distinct T cell populations, and incorporate both surface expression and function (cytokine production). This meant that the complexity of T cell responses could be analyzed, and the plasticity of T cell populations could also be addressed. Furthermore, by using a flow cytometry approach, we could quantify and distinguish individual T cells infiltrating the tissue. We were thus able to include minor T cell populations that might have otherwise been missed and whose contribution to the anti-cancer response might have been discounted.

We found that the tumor microenvironment of colorectal cancer patients is distinct from that of the normal bowel in the same patient, with tumors containing higher frequencies of regulatory and inflammatory T cells and lower frequencies of effector T cells, in agreement with Svensson et al.12 However, we were also able to show that as tumor depth increased, effector T cells and regulatory T cells decreased and increased, respectively (Fig. 4). This supports our hypothesis that a gradient of changing T cell responses exists as the tumor develops within bowel tissue. In addition, we identified a population of FoxP3+IL-17+ T cells that were present at a higher frequency in the tumor tissue compared with normal bowel. It is possible that these non-conventional T cell populations may be involved in tumor initiation and immune evasion.

Previous studies have shown a positive association between T cell infiltration into tumor tissue and colorectal cancer patient outcome21, 32; however, the vast majority of these studies used immunohistochemistry, tissue gene arrays or flow cytometry to quantify T cell populations based on limited parameters (reviewed in Ref.22). Our study has examined 64 distinct T cell populations (32 CD4+, 32 CD8+) by analyzing multiple parameters simultaneously. This approach has confirmed some earlier findings, including an infiltrate of regulatory T cell populations into tumor tissue of colorectal cancer patients, but has also identified minor populations of T cells that may have important roles in disease progression. Interestingly, we found a large and diverse infiltrate of both CD4+ and CD8+ T cells, in the tumor tissue compared with normal bowel. This is in agreement with the findings of Pages et al., who showed a high correlation between metastatic invasion and the expression of CD8 mRNA in tumor tissue.33 The experimental design of this study took advantage of flow cytometry to study freshly acquired patient tissue over the course of 2 years and so currently we do not have clinical outcome data available. As the sample size increases and follow-up data become available, we will correlate different patterns of immune infiltrate with disease characteristics and clinical outcome to determine the clinical significance.

We were able to identify distinct types of regulatory CD4+ T cells using a combination of these markers and to show that all were present at a higher frequency in the tumor tissue compared with the normal bowel. These cells may create a permissive environment for tumor cells to arise by inhibiting the action of anti-tumor effector T cells. Interestingly, we were unable to identify differences in any population of CD8+ regulatory T cells as described by others.34 Our data showing a decreased frequency of effector T cells in tumor tissue compared with normal bowel support this conclusion and is in agreement with others.12

Inflammatory T cells (IL-17+) can enhance tumor development by promoting a tumor microenvironment enriched in regulatory cells but depleted of anti-tumor cells.29 In addition, Th17 cells in the tumor have been postulated to offset the anti-tumor response of IFNγ+ effector T cells.1 IL-17+ T cells are often present alongside regulatory T cells within tumor tissue or in cancer patients.35–37 It has been proposed that the paradoxical link between the high infiltrate of regulatory T cells and positive patient outcome may be due to the effect of regulatory T cells on inflammatory (pro-tumor) T cells38–40 rather than on anti-tumor effector T cells.41 Our finding that both regulatory T cells and inflammatory T cells are increased in frequency in tumor tissue compared with normal bowel supports this hypothesis.

Finally, we identified a population of T cells expressing both IL-17 and FoxP3 (Fig. 5) whose frequency was increased in tumor tissue compared with normal bowel. Our data support that of Kryczek et al., who showed an increase in these double positive cells in the bowel cancers of patients with underlying inflammatory bowel disease, specifically ulcerative colitis and associated colon carcinoma.42 A recent study by Ma and Dong identified this population in cancer tissue from colorectal cancer patients and showed that these cells had the ability to suppress tumor-specific CD8+ T cells.43

In conclusion, we have developed a novel analytical technique to study the complexity of T cell responses in colorectal cancer. Using this methodology, we have found that the immune microenvironments of tumor and non-tumor colon from the same individuals are immunologically distinct and consist of a gradient of T cells expressing combinations of surface and functional markers of conventional T cell populations. The drivers of this distinction between tumor and normal bowel will be important in determining outcome and identifying mechanisms for enhancing anti-tumor immune reactivity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

The authors thank Mandy Fisher for technical support and guidance and Rod Dunbar for help with confocal microscopy. This work was funded by the New Zealand Health Research Council (09/267).

REFERENCES

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
IJC_27855_sm_SuppFig1.ppt323KSupporting Information Figure 1.
IJC_27855_sm_SuppTab1.rtf97KSupporting Information Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.