Field of the invention

The present invention relates to the field of detection of tumour neoangiogenesis, in particular associated with colorectal cancer.

Background

Neoangiogenesis is needed for the growth of solid tumours. Accumulating evidence indicates that, in addition to the sprouting, tumour angiogenesis is supported by the mobilization and incorporation of bone marrow (BM)-derived progenitor cells. The study of endothelial progenitor cells (EPCs) has generated considerable interest and controversy, as the contribution of EPCs to tumour vasculature varies considerably among mouse models and the type of tumour. So, significant but variable contribution of EPCs to tumour vasculature, ranging from about 90% in lymphomas (1) to 10% in Ewing's sarcoma has been reported (2). However, in other models, EPCs did not contribute to tumour growth (3), they were only required in late-stage tumours (4) or indirectly promoted angiogenesis (5-7). The involvement of EPCs in the human tumour vascularization has been also investigated. So, by performing fluorescence in situ hybridization with sex chromosome-specific probes, the contribution of EPCs to tumour endothelium ranged from 1% to 12% (8). On the other hand, EPCs have been reported to be elevated in some cancers (9), and even have been proposed as biomarkers for antiangiogenic therapy and prognosis (10-13).

Recently, increased attention has been directed to BM-derived hematopoietic cells recruited to the tumour microenvironment by angiogenic factors. These cells remain in proximity to the new blood vessels, contributing to the angiogenic process (14). Moreover, tumour associated macrophages appear to be infiltrated into the tumours (15,16), and in mice, Tie-2-expressing monocytes home to tumours, where they are required for angiogenesis (17,18). Additionally, it has been reported that recruitment of hematopoietic cells is necessary and sufficient to restore tumour angiogenesis (1,19). Although substantial evidence indicates that tumour angiogenesis is supported by hematopoietic cells, few reports exist supporting the role of these cells in human solid tumours.

Colorectal cancer (CRC), which includes colon cancer and rectal cancer, is responsible for half a million deaths worldwide every year. There are also about one million new cases diagnosed annually, making it the third most common cancer in the world. Early detection, accurate diagnosis and intensive surveillance are important for best improving a patient's prognosis and response to therapy.

There are a number of methods that doctors use to detect CRC including sigmoidoscopy, colonoscopy and barium enema. Some newer techniques being investigated for effective detection and monitoring of CRC are CT colonography (computed tomographic colonography) and molecular biomarkers.

Tumour markers are used during the treatment of cancer in order to monitor the effectiveness of a therapy and how the patient may be responding to the treatment. If levels of a tumour biomarker decrease it may mean that the cancer is responding to treatment. If levels remain the same or increase after treatment it may be an indication that the therapy is not working. Continued monitoring of tumour biomarker levels following treatment can be used to check for recurrence of the cancer.

In the early detection or screening of patients for CRC, stool based markers are widely used. Tissue -based markers have been studied as possible prognostic or predictive markers of disease, while CRC biomarkers obtained from serum (blood) are primarily used for the postoperative surveillance of patients.

Tissue based markers

Tissue based markers have been investigated as possible prognostic markers and predictors of response to treatment. Thymidilate synthase (TS) is an enzyme required for DNA synthesis and has been studied as a marker that can predict how well a patient may respond to treatment with drugs such as 5-fluorouracil (5-FU) and 5- fluorodeoxyuridine (for a further review about TS, see Popat S, Matakidou A, Houlston RS et al. J Clin Oncol 2004; 22:529-536.; Lecomte T, Ferraz JM, Zinzindohoue F et al. Clin Cancer Res 2004; 10:5880-5888; Shirota Y, Stoehlmacher J, et al. J Clin Oncol 2001; 19:4298-4304). The transcription factor p53 has also been widely investigated as a biomarker that may predict the severity of cancer how it may respond to particular anticancer drug (for a further review about p53, see Tejpar S, Bertagnolli M, Bosman F et al. The Oncologist 2010; 15:390-404.; Braun MS, Richman SD, Quirke P et al. J Clin Oncol 2008; 26:2690-2698.).

The K-ras oncogene is often associated with cancer as abnormalities in this gene have been found in many tumours. K-ras is involved in sending signals that can regulate how much cells grow or multiply. K-ras mutations are linked to approximately half of all CRCs and have been found to be important in the early stages of the disease. Studies have identified an association between K-ras mutations and poor disease outcome in patients with CRC. For further details about K-ras mutations, see Lievre A, Bachet JB, Le Corre D et al. Cancer Res 2006; 66:3992-3995; Karapetis CS, Khambata-Ford S, Jonker DJ et al. N Engl J Med 2008; 359: 1757-1765.

Stool based markers

Faecal occult blood testing (FOBT) is the most commonly used screening test for CRC. There are two main types of FOBT, the guaiac test and the immunochemical test. Both tests detect proteins that may be indicators of CRC.

The advantages of FOBT for CRC are that the tests are simple and affordable, non-invasive, require very little patient preparation and have the capability of examining the entire colorectal tract. They do however have relatively low specificity and sensitivity for both benign (or precancerous adenomas) and malignant CRC.

Faecal DNA tests are used in screening for CRC on the basis that abnormal DNA is excreted in cells shed from cancerous colorectal lesions. Tests usually use a panel of DNA markers in order to identify mutant genes. DNA markers can provide a more accurate test than FOBT and there are no restrictions on diet or medication. The test is however quite laborious, expensive, and also lacks specificity. Examples of stool based DNA markers include K-ras, APC (adenomatous polyposis coli) and p53. Serum based markers

Serum-based markers of CRC are mainly used for monitoring patients following the surgical removal of malignant tumours. Patients are monitored regularly following surgery in order to detect any cancer recurrences or metastases. As up to

50% of patients develop recurrent disease or metastases following surgery, this is an important part of CRC management. CEA (carcinoembyonic antigen) was the first serum marker used in patients with CRC, and although it is the oldest, it still remains the most widely used. CEA is mainly used to monitor patients following surgery for primary CRC. A number of studies have shown that intensive monitoring after cancer surgery is associated with an improved outcome if regular CEA measurements were taken. Other serum-based tumour biomarkers used for CRC include CA-19-9, TPA (Tissue polypeptide antigen), TPS (tissue polypeptide specific antigen) and TIMP-1 (tissue inhibitors of metalloproteinases) (Cancer Res 2004; 64:952-961. Rhee JS, Diaz R, Korets L, Hodgson JG, Cousens LM).

In the state of art, there are several documents which use biomarkers related to

CRC.

For example, in the document by Hua Xue et al., Identification of Serum biomarkers for Colorectal Cancer Metastasis Using a Differential Secretome Approach, Journal of Proteome Research 2010, 9, 545-555 there is disclosed the application of an LC-MS (liquid chromatography-mass espectrometry)/MS-based label-free quantitative proteomics approach to compare the differential secretome of a primary cell line SW480 and its lymph node metastatic cell line SW620 from the same CRC patient.

In the document by Sudhir Srivastava et al., Biomarkers for early detection of colon cancer, Clinical Cancer Research, Vil. 7, 118-1126, May 2001, there is disclosed that the major advances in understanding CRC include the identification and involvement of APC, p53 and Ki-ras in the development and progression of the disease, the identification of the aberrant crypt foci as an early preinvasive lesion and its relation to the development of cancer.

In the document by DG Ward et al., Identification of serum biomarkers for colon cancer by proteomic analysis, British Journal of Cancer (2006) 94, 1898-1905, there are disclosed new biomarkers for early detection of CRC. In particular, they used surface-enhanced laser desorbtion/ionisation (SELDI) to investigate the serum proteome of 62 CRC patients and identified proteins (complement C3a des-arg, ccl- antitrypsin and transferring) with diagnostic potential.

In the document by Mariaelena Mata et al., Circulating tumour cells: utility for predicting response to anti-EGFR therapies, Expert Rev. Mol. Diagn. 9(2), 115-119 (2009) there is disclosed the development of technology platforms for capturing circulating tumour cells in the clinical setting providing thereby the possibility of characterising the tumour using non-invasive biopsies from peripheral blood.

A possible benefit of molecular characterization of circulating tumour cells is that it can be used to design a personalized medicine as many primary tumours negative for some markers become positive at the time of metastasis. Enumeration of circulating tumour cells has proven valuable as a prognostic marker of breast, colorectal and prostate cancers. However, it has been reported that circulating tumour cells were detected in only 40% of patients with CRC (Annals of Oncology 2008; 19:935-938. Circulating tumour cells in colorectal cancer: correlation with clinical and pathological variables. Sastre J et al.)

In EP 05772471, there is disclosed a method for detecting colorectal adenoma and/or colorectal carcinoma based on the determination of the level of C3a, being used then as a biomarker. It is also disclosed a method for discriminating between colorectal adenoma and colorectal carcinoma as well as to a method for monitoring the course of colorectal adenoma and/or colorectal carcinoma and/or the treatment of colorectal adenoma and/or colorectal carcinoma. A test system and an array for use in these methods are also disclosed.

It should also be mentioned the use of serum microRNAs which are currently used as potential biomarkers for CRC. For example, A. Chajeu et al. in the 2009 Gastrointestinal Cancers Symposium (category of Colon and Rectum and subcategory of Prevention, diagnosis and screening) propose microRNAs, a familiy of small non- coding regulatory RNAs involved in human development and pathology, as an emerging class of effective serum markers. The authors demonstrate that certain microRNAs are found in different amounts in sera of CRC patients compared with healthy controls.

According to previous paragraphs, a first object of the present invention is to provide a new method for detecting tumour neoangiogenesis, in particular associated with CRC providing the following list of advantages:

a) It is a non-invasive method.

b) Only up to 8 ml of blood sample are needed.

c) No risk associated with morbidity/mortality (events shown in for example colonoscopies).

d) It would be useful to evaluate the hematopoietic system during chemotherapy and/or radiotherapy.

e) No need to determine the level or concentration of molecules or cytokines in serum or plasma, neither the level of other proteins, although if considered appropriate can be determined, since plasma obtained can be stored frozen.

f) No need to determine the amount of circulating tumour cells which renders the present method easier and available in all hospitals.

g) Easy way of detection by flow cytometry available in all hospitals.

A second object of the present invention is to provide an immunoassay to be used in said method.

A third object of the present invention is to provide a kit for detecting tumour neoangiogenesis, in particular associated with CRC.

A fourth object of the present invention is the use of several parameters as biomarkers for detecting tumour neoangiogenesis, in particular associated with CRC. Summary of the invention The present invention relates to a method for detecting tumour neoangiogenesis, in particular associated with CRC, comprising the steps of:

a) Providing an isolated blood sample from an individual;

b) Obtaining nucleated cells from the isolated blood sample;

c) Determining the level of circulating CD 14+ monocytes, the level of circulating CD34 ⁺ hematopoietic progenitor cells and the percentage of circulating CD34 ⁺ hematopoietic progenitor cells expressing ACE/BB9 (CD 143) antigen in the nucleated cells from the isolated blood sample;

d) Comparing the ratio between the level of circulating CD14 ⁺ monocytes divided by the level of circulating CD34 ⁺ hematopoietic progenitor cells in respect of the percentage of circulating CD34 ⁺ hematopoietic progenitor cells expressing ACE/BB9 (CD 143) antigen determined in c) with values obtained for control individuals or healthy individuals.

The present invention also relates to an immunoassay used in step c) of the method for detecting tumour neoangiogenesis, in particular associated with CRC, in an isolated blood sample from an individual, comprising:

a) binding an anti-human CD45 monoclonal antibody to CD45 antigen present in leukocytes (white blood cells) from the isolated blood sample (CD45 ⁺ hematopoietic cells);

b) binding at least one anti-human monoclonal antibody to an antigen from circulating CD14 ⁺ monocytes, circulating CD34 ⁺ hematopoietic progenitor cells and circulating CD34 ⁺ cells expressing ACE/BB9 (CD 143) antigen, preferably said at least one anti-human monoclonal antibody being selected from the group of anti-human monoclonal antibodies consisting of anti-CD34, anti-CD14, and anti-ACE/CD143 (BB9);

c) mixing reactants and cells in a container;

d) detecting said binding between antibodies and antigens by means of at least one reactant.

The present invention further relates to a kit for detecting tumour neoangiogenesis, in particular associated with CRC.

The present invention additionally relates to the use of level of circulating CD14 ⁺ monocytes, the level of circulating CD34 ⁺ hematopoietic progenitor cells and the percentage of circulating CD34 ⁺ hematopoietic progenitor cells expressing ACE/BB9 (CD 143) antigen in an isolated blood sample as a biomarker for detecting tumour neoangiogenesis, in particular associated with CRC. Brief description of the drawings

Figure 1. Schematic overview of the flow cytometric protocol for enumeration of CPCs (circulating progenitor cells). The gates used to select and analyze CD34 ⁺ cells are shown. Representative FACS plots that show the expression of CD31, CD133, CD184/CXCR4 and BB9/CD143 within the CD34 ⁺ cell gate and their corresponding isotype-matched controls are presented. Cells in the R3 gate were used to analyze the two subsets of CD34 ⁺ cells, CD34 ⁺ CD45 ^~ and CD34 ⁺ CD45 ^dim . Rl, R2, R3 and R4 indicate the gates used in the analysis. 7-Amino-actinomycin D (Sigma) was used to detect and exclude non-viable cells in the analysis.

Figure 2. Identification of human monocytes and circulating endothelial cells in peripheral blood. MNCs (mononucleated cells) were gated on to identify the CD45 ^neg and CD45 cell subsets. CD146 and CD31 expression on CD45neg cells was used to identify CECs (circulating endothelial cells). CD45 ⁺ cells were gated on to enumerate and analyze CD14 ⁺ monocytes. The expression of CD 16 and Tie-2 on CD14 ⁺ cells was determined. The corresponding isotype-matched controls are shown.

Figure 3. Correlations between hematopoietic cells and angiogenic parameters. The ratio CD14 /CD34 ⁺ was calculated by dividing the number of circulating CD14+ cells/mL by the number of circulating CD34 ⁺ cells/mL. Plasma values of VEGF, P1GF and IL-6, expressed in fg/mL, were divided by the number of circulating CD34+ cells/mL. The ratio CD 167CD34 ⁺ was calculated by dividing the number of CD14 CD16 ⁺ cells/mL by the number of circulating CD34 ⁺ cells/mL. Pearson or Spearman correlation coefficient and P value are given. (●) control group, n: 25 to 32; (o) cancer group, n: 54. In panel F, (●, Α) correspond to CD14 ⁺ /CD34 ⁺ values and CD16 ⁺ /CD34 ⁺ values from control group, respectively, and (o, Δ) correspond to CD14 ⁺ /CD34 ⁺ values and CD16 ⁺ /CD34 ⁺ values from cancer group, respectively.

Figure 4. The level of circulating CD34 ⁺ CXCR4 ⁺ HPCs correlate to the plasma concentrations of angiogenic factors per CD34 ⁺ cell. The values corresponding to ratios VEGF/CD34 ⁺ and P1GF/CD34 ⁺ were calculated as described in Figure 3. Plasma values of MMP-9 were expressed in (ng/mL) x 100, and divided by the number of circulating CD34 ⁺ cells/mL. Pearson correlation coefficient and P value are given. (●) control group, n: 15 to 18; (o) cancer group, n: 34 to 37.

Figure 5. The ability of CD34 ⁺ HPCs to differentiate into monocytes correlates to the percentage of circulating CD34 ⁺ BB9 ⁺ HPCs in patients with colorectal cancer. The values corresponding to ratios VEGF/CD34 ⁺ and P1GF/CD34 ⁺ were calculated as described in Figure 3. The %CD34 BB9 cells represent the percentage of CD34+ cells expressing BB9 antigen. Pearson or Spearman correlation coefficient and P value are given. (●) control group, n: 15 to 18; (o) cancer group, n: 34 to 37.

Figure 6. Diagnostic value of hematopoietic parameters. ROC curves were drawn with CD14/CD34 ratio values (o-o) and with CD14/CD34 ratio values divided by the percentage of CD34 ⁺ cells expressing BB9 (●-●).

Detailed description of the invention

1.- Definitions

The term "Colorectal cancer" or "colon cancer" or "large bowel cancer" or

"CRC", as used herein, includes cancerous growths in the colon, rectum and appendix.

The term "biomarker", as used herein, follows the definition provided by NIH (National Institutes of Health) i.e. a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention. In the present invention, said biomarkers are the level of circulating CD14 ⁺ monocytes, the level of circulating CD34 ⁺ hematopoietic progenitor cells and the percentage of circulating CD34 ⁺ hematopoietic progenitor cells expressing antigen ACE/BB9 (CD 143) in an isolated blood sample.

The term "invididual", as used herein, is taken to mean a human being.

The term "circulating CD14 ⁺ monocytes", as used herein is taken to mean the number of hematopoietic cells expressing CD45 and CD 14 antigens, also named monocytes, in peripheral blood, and it can be expressed as number of CD14 ⁺ monocytes per ml of blood or number of CD14 ⁺ monocytes per ml of nucleated cells, obtained after gradient centrifugation of blood, or as percentage of nucleated cells positive for CD 14 antigen.

The term "circulating CD34 ⁺ hematopoietic progenitor cells", as used herein, is taken to mean the number of hematopoietic cells with dim or negative expression of CD45 antigen and expressing CD34 antigen, a marker of hematopoietic progenitor cells in blood, expressed as number of CD34 ⁺ hematopoietic progenitor cells per ml of blood or number of CD34 ⁺ hematopoietic progenitor cells per ml of nucleated cells, obtained after gradient centrifugation of blood, or as percentage of nucleated cells positive for CD34 antigen with a dim or negative expression of CD45 antigen.

The term "circulating CD34 ⁺ hematopoietic progenitor cells expressing ACE/BB9 (CD 143) antigen" as used herein, is taken to mean the number of CD34 ⁺ hematopoietic progenitor cells positive for CD 143 (BB9) antigen per ml of blood or the number of CD34 ⁺ hematopoietic progenitor cells positive for CD 143 (BB9) antigen per ml of nucleated cells, obtained after gradient centrifugation of blood, or as percentage of nucleated cells positive for CD34 ⁺ hematopoietic progenitor cells expressing CD 143 (BB9) antigen.

Unless otherwise stated, all circulating cells referred to in the present invention are human cells.

2 - Development of the invention

The present invention relates to a method for detecting tumour neoangiogenesis, preferably associated with CRC, comprising the steps of: a) Providing an isolated blood sample from an individual;

b) Obtaining nucleated cells from the isolated blood sample;

c) Determining the level of circulating CD14 ⁺ monocytes, the level of circulating CD34 ⁺ hematopoietic progenitor cells and the percentage of circulating CD34 ⁺ cells expressing ACE/BB9 (CD 143) antigen in the nucleated cells from the isolated blood sample;

d) Comparing the ratio between the level of circulating CD14 ⁺ monocytes divided by the level of circulating CD34 ⁺ hematopoietic progenitor cells in respect of the percentage of circulating CD34 ⁺ hematopoietic progenitor cells expressing ACE/BB9 (CD 143) antigen determined in c) with values obtained for control individuals or healthy individuals.

Individuals with CRC show a strong correlation between the ratio CD14 ⁺ /CD34 ⁺ and the percentage of CD34 ⁺ progenitor cells expressing BB9, whereas in control individuals or healthy individuals both parameters (CD14 :CD34 ⁺ and the percentage of CD34 ⁺ progenitor cells expressing BB9) are independent. For more details see example section and figure 5D.

Said determination can be carried out by flow cytometry.

In another aspect, the present invention further relates to an assay system for detecting tumour neoangiogenesis, preferably associated with CRC, in an isolated blood sample from an individual. The assay system is based on the specificity of an antibody to specifically bind an epitope or structural element from the antigen or receptor. In particular said assay system is an immunoassay which is used in the step c) of the method for detecting tumour neoangiogenesis disclosed herein, preferably associated with CRC, in an isolated blood sample from an individual, comprising: a) binding an anti-human CD45 monoclonal antibody to CD45 antigen present in leukocytes (white blood cells) from the isolated blood sample (CD45 ⁺ hematopoietic cells);

b) binding at least one anti-human monoclonal antibody to an antigen from circulating CD14 ⁺ monocytes, circulating CD34 ⁺ hematopoietic progenitor cells and circulating CD34 ⁺ cells expressing ACE/BB9 (CD 143) antigen, preferably said at least one anti-human monoclonal antibody being selected from the group of anti-human monoclonal antibodies consisting of anti- CD34, anti-CD 14, and anti-ACE/CD143 (BB9);

c) mixing reactants and cells in a container, preferably said container is a tube;

d) detecting said binding between antibodies and antigens by means of at least one reactant, preferably being said reactant for detection selected from the group consisting of fluorescein-isothiocyanate (FITC), phycoerythrin (PE), peridin chlorophyll (PerCP), PE-cyanin 7 (PE-Cy7), allophycocyanin (APC), APC-cyanin 7 (APC-Cy7) and APC-H7. The list of antibodies and reactants used in the immunoassay above is not limited thereto, since last generation of cytometers with 3 lasers can also be used allowing to visualize up to eight colours and , therefore, it is possible to carry out the present immunoassay of the invention with other antibody conjugated to suitable fluorochromes. Nevertheless, the list shown as preferable makes reference to antibodies and reactants experimentally used in the present invention with a cytometer which allows to visualize six colours.

A further aspect of the present invention relates to a kit for detecting tumour neoangiogenesis, preferably associated with CRC, comprising: a) at least one receptacle means for receiving a test sample of blood from the individual, preferably tubes containing citrate and Ficoll;

b) an antibody selected from anti-human CD45, anti-human CD34, anti-human CD 14 and anti- human ACE/CD 143 (BB9);

c) a reactant for detection selected from the group consisting of fluorescein- isothiocyanate (FITC), phycoerythrin (PE), peridin chlorophyll (PerCP), PE- cyanin 7 (PE-Cy7), allophycocyanin (APC), APC-cyanin 7 (APC-Cy7) and APC-H7;

d) isotype-matched controls to be used as negative controls; said isotype controls matched to the specific primary antibodies being used in order to accurately determine the level of specific staining by the primary antibody;

e) instructions for using the kit.

A further aspect of the present invention relates to the use of the level of circulating CD14 ⁺ monocytes, the level of circulating CD34 ⁺ hematopoietic progenitor cells and the percentage of circulating CD34 ⁺ cells expressing ACE/BB9 (CD 143) antigen in an isolated blood sample as a biomarker for detecting tumour neoangiogenesis, preferably associated with CRC.

Examples

The following Examples are provided to explain and illustrate the present invention and are not intended to be limiting of the invention.

Materials and Methods

Study Subjects

Fresh blood was obtained from 65 newly diagnosed CRC patients before tumour resection. Of these, 39 were male and 26 female, with a mean age of 67 years (range: 25 to 84 years). 9% of patients were in stage I, 31% in stage II, 40% in stage III and 20% in stage IV. 63% of tumours were localized in colon and 37% in rectum. All tumours were pathologically classified as adenocarcinoma. Neoadjuvant treatment was given to 52%> of rectal cancer patients before surgery, consisting of an oral dose of 825 mg/m of capecitabine and concomitant radiotherapy (11 out 13 patients), or only radiotherapy (2 out 13 patients). The control group comprised 32 healthy donors (18 male and 14 female), with a median age of 52 years (range: 25 to 72 years). The study was approved by the Bioethical Committee of Valencia University General Hospital, Valencia, Spain. The written informed consent was obtained from each patient.

Quantification of Soluble Factors

Peripheral blood was obtained previous signed informed consent from patients before surgery. Blood samples were collected in BD Vacutainer Cell Preparation Tubes (BD Diagnostics) containing citrate and Ficoll. After centrifugation, mononuclear cells (MNCs) and plasma were aseptically obtained according to manufacturer instructions. Enzyme-linked immunosorbent assay was used to determine plasma levels of vascular endothelial growth factor (VEGF), placental derived growth factor (P1GF), matrix metalloproteinase-9 (MMP-9) (DRG Instruments, GmbH, Marburg, Germany), soluble VEGF receptor 2 (VEGFR-2/KDR), soluble VEGF receptor 1 (VEGFR-l/Fltl), Interleukin-6 (IL-6) and angiotensin- converting enzyme (ACE) (R&D Systems Europe Ltd. Abingdon, UK).

Flow Cytometry

A total of 1 x 10 ⁶ cells were incubated for 30 minutes at 4°C with the following monoclonal antibodies (mAbs): CD45 (2D1, BD Biosciences), CD34 (8G12, BD Biosciences), CD14 (ΜΦΡ9, BD Biosciences) and CD16 (3G8, BD Biosciences), CD31 (WM59, BD Pharmingen), CD146 (P1H12, BD Pharmingen), CXCR4/CD184 (12G5, BD Pharmingen), ACE/CD 143 (BB9, BD Biosciences); CD133/2 (293C3, Miltenyi Biotec), angiopoietin receptor Tie-2 (83715, R&D Systems) and VEGFR- 2/KDR (89106, R&D systems, and 2, Milteny Biotec). Antibodies were labeled with fluorescein-isothiocyanate (FITC), phycoerythrin (PE), peridin chlorophyll (PerCP), PE-cyanin7 (PE-Cy7), allophycocyanin (APC), APC-cyanin7 (APC-Cy7) or APC-H7. As controls, the matched labeled isotypes were used. 7-aminoactinomycin D was used to discard non-viable cells. To perform enumeration of HPCs and EPCs a minimum of 200 CD34 ⁺ cells were acquired. Monocyte subsets were analyzed after at least 200,000 events were acquired. Flow cytometry analyses were performed using a six-color FACSCanto flow cytometer (Becton Dickinson) and the FACSDiva software.

Statistics

Statistical analyses were performed with GraphPad PRISM version 5.00 for Windows (GraphPad Software). Data are expressed as mean ± SD and 95% confidence intervals (CI). The normal distribution of data was tested by Kolmogorov-Smirnov or Shapiro-Wilk tests. Comparisons of variables between the two groups were analysed with a Student's t-test or Mann- Whitney test where applicable. Spearman or Pearson correlation coefficient was used to analyze correlation between variables. The power of the suggested biomarkers was evaluated with the areas under the ROC curve with the 95% CI. The optimal cutoff point was given by the maximum of the Youden's index. Differences were significant if p value was less than 0.05.

Results

Circulating HPCs are decreased in CRC patients

CD34 ⁺ progenitor cells from human peripheral blood were analyzed according to international procedure for enumeration of human HPCs (Fig. 1). It is known that more than 95% of CD34 ⁺ progenitor cells coexpress CD31, a common marker of HPCs and EPCs, and so the entire CD34 ⁺ CD31 ⁺ cell population was defined as CPCs. Moreover, most of these CD34 cells exhibits a low expression of the pan-leukocyte antigen CD45 and are hematopoietic in origin, whilst the remaining CD34 ⁺ cells do not express CD45 and are enriched in EPCs and HPCs. To analyze the different progenitor cell subsets, we took advantage of the expression of CD133, an antigen restricted to hematopoietic lineage, that down-regulates as HPCs proliferate (15). Accordingly, HPCs were phenotyped as CD34 ⁺ CD31 ⁺ CD133 ⁺ with dim or negative expression of CD45; committed or proliferating HPCs as CD34 ⁺ CD31 ⁺ CD133 ^" CD45 ^dim , and EPCs as CD34 ⁺ CD31 ⁺ CD133 ^" CD45 ^" .

Flow cytometric analysis evidenced a dramatic twofold decrease of CPCs content in CRC patients (Table 1).

Moreover, a more exhaustive analysis demonstrated that within CPCs, only the percentage corresponding to HPCs was significantly decreased, while that corresponding to the most proliferating CD34 ⁺ CD31 ⁺ CD133 ^~ CD45 ^dim HPCs was increased (Table 1), suggesting a higher rate of entry into the cell cycle of HPCs in patients with cancer. The proportion of CPCs corresponding to EPCs was similar in control and cancer groups (Table 2). Additionally, we determined the expression of VEGFR2/KDR, a marker originally shown to be restricted to vascular cells and used to enumerate EPCs, although it is also expressed by some HPCs subsets. Our results show that no differences were found in the percentage of CPCs positive for KDR between healthy donors [1.3% ± 1.9% (95% CI=0.47% to 2.19%)] and CRC patients [0.6% ± 1.0% (95% CI=0.18% to 1.04%)]. Furthermore, we also determined the frequency of mature circulating endothelial cells (CECs) whose phenotype correspond to CD45 ^" CD31 CD146 ⁺ (Fig. 2). The average of these cells in peripheral blood was 0.33% (95% CI=0.19% to 0.47%) and 0.41% (95% CI=0.29% to 0.52%) of MNCs cells in control and cancer groups, respectively. Therefore, no differences were found between healthy donors and cancer patients.

HPCs from CRC patients exhibit a higher ability to differentiate into monocytes with a marked proinflammatory signature

Our results show that the proportion of MNCs corresponding to CD14 ⁺ monocytes was significantly increased in CRC patients (Table 2). To assess whether circulating levels of these myeloid cells were associated with those of CD34 ⁺ HPCs, all values were plotted in a graph. In fact, a good correlation between these two cell populations was obtained in both study groups, but at single cell level a higher number of CD 14 monocytes was obtained per CD34 ⁺ cell in CRC patients (Fig. 3A), indicating a greater efficiency of HPCs to differentiate into monocytes. The expression of Tie-2 and CD16 on CD14 ⁺ monocytes was also analyzed (Fig. 2). Unexpectedly, the proportion of pro-angiogenic CD14 ⁺ Tie-2 ⁺ monocytes within the CD14 ⁺ cell subset was similar in both groups, 0.77% ± 0.42% (95% CI=0.56% to 0.98%) in controls and 1.41% ± 1.36% (95% CI=0.79% to 2.04%) in CRC patients. However, pro- inflammatory CD14 ⁺ CD16 ⁺ monocytes were dramatically increased from 5.2% ± 1.9%

(95% CI=4.27% to 6.2%) in controls to 10.5% ± 4.8% (95% CI=8.7% to 12.3%) in patients (P<0.0001).

Plasma levels of angiogenic factors and related proinflammatory molecules are increased in CRC patients It is widely known that circulating levels of angiogenic factors are elevated in CRC. Accordingly, plasma levels of both VEGF and PIGF were significantly increased in CRC patients when compared to control group, but plasma levels of their soluble receptors VEGFR2/KDR and VEGFRl/Fltl were similar between the two groups, suggesting a greater bioavailability of these angiogenic factors in these patients. Additionally, IL-6 and MMP-9, two molecules involved in inflammatory and angiogenic processes, were seven- and two-fold increased, respectively in CRC patients (Table 3).

Angiogenic factors modulate hematopoiesis

To evaluate the impact of angiogenic factors concentration on circulating levels of HPCs, we expressed plasma values of selected molecules per CD34 ⁺ cell. In doing so, a scenario of strong relationships among these factors appeared (Fig. 3B and 3C). But importantly, at single cell level, plasmatic concentrations of VEGF, P1GF and IL-6 per CD34 ⁺ cell were directly correlated to the CD14 :CD34 ⁺ ratio, a measure of the ability of CD34 ⁺ cells to differentiate into monocytes (Fig. 3D, 3E, 3F). Of note IL-6 concentration per CD34 ⁺ cell was also correlated to the CD14 CD16 :CD34 ⁺ ratio (Fig. 3F).

CRC induces mobilization of CD34 ⁺ cells

As circulating angiogenic factors were increased in CRC patients, we hypothesized that this increase could affect the degree of mobilization of CD34 ⁺ HPCs. The egress of CD34 ⁺ cells from the BM is facilitated by the interaction between stromal-derived factor- 1 (SDF-1) and its unique receptor CXCR4 on CD34 ⁺ cells. Moreover, it has been demonstrated that HPCs mobilization is mediated by degradation of SDF-1 within the BM (16, 17) and as consequence, mobilized CD34 ⁺ cells have a lower proportion and density of CXCR4 (18). Therefore, we determined the expression of CXCR4 on CD34 ⁺ progenitor cells. In fact, a higher number of CD34 ⁺ HPCs were mobilized in patients as demonstrated by the lower proportion of

CD34 ⁺ cells expressing CXCR4 [12.1% ± 5.0% (95% CI=10.37% to 13.9%) versus 18.0% ± 7.3% (95% CI=13.5% to 22.6%) in controls, P = 0.02], as well as by the lower membrane expression of CXCR4, determined as mean fluorescence intensity (414 ± 95 AU versus 557 ± 209 in controls, P =0.014).

As depicted in Figure 4A and 4B, circulating levels of CD34 ⁺ CXCR4 ⁺ cells were dependent of the amount of both VEGF and P1GF per CD34 ⁺ cell, but a more efficient mobilization was observed in CRC patients. It has been proposed that SDF-1 - mediated migration and recruitment of CPCs depends on MMP-9 (19). Our results show that indeed the number of circulating CD34 CXCR4 ⁺ cells were well correlated with plasma concentration of MMP-9 per CD34 ⁺ cell in healthy donors, but independent in the CRC patients (Fig. 4C), despite the observed good correlation between circulating VEGF and MMP-9 per CD34 ⁺ cell (Fig. 4D). Additionally, we observed that the number of circulating monocytes was well correlated with that of CD34 ⁺ CXCR4 ⁺ cells in healthy individuals, but not in CRC patients (Fig. 4E). The CD14 ⁺ :CD34 ⁺ ratio in CRC patients is related to the percentage of the circulating CD34 ⁺ HPCs expressing angiotensin-converting enzyme.

Finally, we tested the possibility that the increased plasma levels of angiogenic factors could reduce the reservoir of the most pluripotent CD34 ⁺ cells and thus satisfying the myeloid requirement necessary to promote tumour angiogenesis. To this end, we took advantage of the recent description of CD143/BB9 (angiotensin- converting enzyme, ACE), which defines CD34 ⁺ HPCs with the ability to engraft and repopulate irradiated BM (20) and identifies hemangioblasts differentiating from human embryonic stem cells (21). Our results show that 19.9% ± 8.1% (95%> CI=15.1% to 24.8%) of the CD34 ⁺ cells expressed BB9 in the control group, but only 10.8% ± 4.4% (95% CI=9.0% to 12.7%) in the CRC group expressed it (P = 0.001). Moreover, no differences in fluorescence intensity were observed.

The plots of circulating CD34 BB9 ⁺ cells against plasma concentrations of angiogenic factors per CD34 ⁺ cell lead us to observe that these primitive cells were well correlated to VEGF and P1GF levels in CRC group, but only with VEGF levels in control group (Fig. 5A and 5B). But importantly, we observed that the level of circulating monocytes was dependent of the level of the most primitive circulating CD34 ⁺ BB9 HPCs in both control and CRC groups (Fig. 5C), indicating a intimate association between these two cell populations. However, at single cell level, a higher number of CD14 ⁺ cells per CD34 BB9 ⁺ cell was obtained in the cancer group, corroborating the higher efficiency of HPCs towards differentiation into monocytes, previously observed. Additionally, whilst in the control group the CD14 :CD34 ⁺ ratio was independent of the percentage of CD34 ⁺ progenitor cells expressing BB9 a strong correlation between these two parameters was obtained in the CRC group (Fig. 5D). For this reason we assessed for diagnostic value of these hematopoietic parameters. So, the area under the ROC curve for CD14/CD34 ratio values was 0.885 (95% CI, 0.819 to 0.952: p<0.0001), and the optimal cutoff point of >221.5 had a 79.7% of sensitivity (95% CI, 68.3% to 88.4%), and 89.3% of specificity (95% CI, 71.8% to 97.7%). Next, we calculated the quotient of CD14/CD34 ratio values per the percentage of BB9 expressing HPCs, and in this case the area under the ROC curve reached a value of

0.929 (95% CI, 0.840 to 1.019; pO.0001) with an optimal cutoff of >13.1 that had 97.8% of sensitivity (95% CI, 88.5% to 99.9%) and 87.5 of specificity (95% CI, 61.6% to 98.5%) (Fig 6).

Interestingly, plasma levels of ACE were slightly but significantly decreased in CRC [(134 ± 44 ng/niL) (95% CI=122 ng/niL to 1474 ng/niL)] versus the control group [(181 ± 59 ng/mL) (95% CI=157 ng/mL to 204 ng/mL)], P =0.0008, but when expressed per CD34 ⁺ cell the relative concentrations of ACE were well correlated to those of VEGF and of P1GF (data not shown). These data strongly suggest that in addition to the classic angiogenic factors, ACE must also play an important role in the modulation of angiogenesis, as well as myeloid differentiation, particularly in CRC. Discussion

It is now believed, in light of recent advances in progenitor cell biology, that neovascularization occurs via angiogenesis and/or vasculogenesis, and thus the contribution of endothelial cells in the new blood vessels has been the subject of much study. So, circulating endothelial cells have been reported to be elevated in some human cancers (7), and interestingly, a decrease in CD31 ^bright CD45 ^" cells was observed in patients with rectal cancer after treatment with bevacizumab, a VEGF-specific antibody, (22). But, it is important to mention that CD 146, a marker of endothelial cells, could not be detected on CD31 ^bnght CD45 ^" viable cells from rectal cancer patients (23). Accordingly, we found that the frequency of these cell subsets in peripheral blood was extremely low in both healthy donors and CRC patients. Moreover, using a panel of specific markers, we demonstrate that percentages of MNCs corresponding to EPCs and CECs were similar in both CRC and control groups, suggesting a lack of mobilization and probably of contribution of these endothelial cell subsets to tumour angiogenesis. These results question the validity of using the enumeration of endothelial cells in peripheral blood as biomarkers of anti-angiogenic therapy, at least in CRC patients.

Here, we use the term CPCs to include HPCs and EPCs, and we show that the percentage of MNCs corresponding to CPCs was severely reduced in CRC patients, but only CPCs with phenotypic characteristics of HPCs were decreased. Remarkably,

HPCs from CRC patients exhibited an increased ability to differentiate into monocytes, and the resultant increase in circulating levels of these myeloid cells can be of great importance, given the implications of this cell subset in tumour angiogenesis. So, myeloid cells not only exhibit the ability to promote sprouting angiogenesis ex vivo, but their perivascular position contributes and seems to be necessary to tumour angiogenesis in experimental animal models (24, 25). It has been proposed that Tie-2 expressing monocytes are the most proangiogenic subsets of these myeloid cells (24, 26), and circulating Tie-2-monocytes are increased in some human cancers (27). In contrast, our results demonstrate that within CD14 ⁺ cells, the proportion of Tie-2- expressing monocytes was similar between healthy donors and CRC patients. However, circulating CD16-expressing monocytes, the most proinflammatory myeloid cells (28) were significantly increased, emphasizing the strong association between inflammation and angiogenesis in CRC (29). As inflammatory processes are related to increased angiogenesis (30) monocytes must contribute to the creation of a proangiogenic environment by delivering proinflammatory mediators and angiogenic factors. Accordingly, the single infusion of bevacizumab was able to decrease microvascular density and increase the pericyte coverage of a fraction of vessels (22). Additionally, targeted deletion of Vegfa in cells of the myeloid lineage not only resulted in a less tortuous tumour vasculature with increased pericyte-coverage and decreased vessel length, but contributed to the chemotherapeutic efficacy of cytotoxic agents (31), highlighting the important role of myeloid cells in tumour angiogenesis.

It is known that cytokines secreted by hematopoietic cells induce angiogenesis whilst angiogenic factors, including VEGF, play an essential role in hematopoiesis. Here we demonstrate that an intimate association exists between circulating levels of angiogenic factors and hematopoietic cells in both healthy state and CRC, and so indeed the plasma elevations of VEGF and P1GF can aid to explain the increase ability of HPCs to differentiate into monocytes in CRC patients. Moreover, and supporting the inducer effect of angiogenic factors on HPCs mobilization (32, 33) we demonstrate, on the basis of CXCR4 expression, that CRC induces a higher degree of HPCs mobilization.

These findings can have a major impact on tumour development, as recruitment of BM progenitor cells has been shown to be necessary and sufficient for tumour angiogenesis (1, 25, 34), and CD34 ⁺ HPCs have been observed to home into the tumour periphery, adhered or near to the blood vessel wall within tumour parenchyma in mice with breast carcinomas (35). Additionally, the sustained elevation of VEGFi ₆ 5 in plasma, target organs or in tumour has been shown to result in remodeling of BM vascular architecture and mobilization of HPCs to the bloodstream and to extrameduUary organs with concomitant splenomegaly and hepatomegaly (33, 36). On the other hand, clinical studies have provided evidence that elevation of VEGF plasma levels is associated with BM failure (37) and hemangiomas (38) among others.

Therefore it is possible that angiogenic factors induce the recruitment of HPCs into the tumour. However, although we have no evidence of the induction of extrameduUary hematopoiesis in these patients, this possibility can not be disregarded.

Additionally, and in the light of recent reports, the observed hematopoietic imbalance may also be of relevance in the outcome of CRC patients. So, clinical studies have reported that after treatment with bevacizumab, circulating levels of VEGF and P1GF were significantly increased in patients with rectal cancer (39), and after bevacizumab monotherapy, patients with rectal carcinoma exhibited higher circulating levels of SDF-1, and gene expression of SDF-1, CXCR4 and CXCL6 in rectal cancer cells was up-regulated (40). Taken into account that trafficking and recruitment of hematopoietic cells occurs after perivascular expression of SDF-1, that functions to capture and retain incoming CXCR4 ⁺ cells (17, 25, 41), it is plausible that anti-VEGF therapy should induce more than prevent the recruitment of hematopoietic cells, to tumour sites. Furthermore, the demonstration that recruitment and migration of hematopoietic cells into the tumour occurs prior to the initiation of angiogenesis in the tumour mass (34), lead us to hypothesize that the modest benefit seen on overall survival after anti-angiogenic therapies could be linked to the hematopoietic imbalance exhibited by these patients.

Finally, the most important finding was the close relationship between the angiogenic factors in plasma and circulating levels of hematopoietic progenitor cells expressing ACE/BB9, suggesting that angiogenic factors must play a key role in the maintenance of these primitive cells, and therefore be responsible for their decline in CRC patients. Moreover, as circulating levels of both CD34 BB9 ⁺ cells and CD14 ⁺ cells were well correlated, our results indicate that the elevated concentration of angiogenic factors must contribute to reduce the reservoir of the primitive HPCs to satisfy the myeloid requirement necessary to promote tumour angiogenesis in these patients. Importantly, we demonstrate, for the first time that the CD14 :CD34 ⁺ ratio in CRC patients was dependent of the proportion of CD34 ⁺ cells expressing ACE, whilst in healthy steady state conditions no correlation was observed. Collectively, these data strongly suggest that the shift of the dynamic balance between HPCs and myeloid cells to a greater generation of monocytes should occur at the expense of increased consumption of the CD34 BB9 ⁺ cells.

The involvement of these primitive progenitor cells raises the possibility that the rennin-angiotensin system (RAS) plays an unrecognized role in hematopoietic imbalance present in CRC patients. To note, a review of the literature suggests that

ACE must be involved in enhancing the recruitment of primitive HPCs into S-phase in BM (42, 43), and ACE inhibitors, as well as antagonists of angiotensin (Ang) II receptors, AT1/AT2, have been shown to affect human hemangioblasts colony expansion, HPCs proliferation and differentiation (21). On the other hand, the RAS has been involved in angiogenic processes, but also in tumour angiogenesis. For instance, the overexpression of Hu-Angiotensinogen in different animal models inhibited tumour growth and metastasis (44, 45), and the expression of ATI in stromal cells promoted tumour angiogenesis by inducing VEGF synthesis (46). It is important to mention that tumour-associated macrophages overexpress ATI (47), and a recent report has just shown that carcinoma-associated fibroblasts represents a transitional phenotype from the monocytic lineage to the fibroblast lineage (48). Therefore, myeloid cells by signaling through the RAS could contribute to human tumour angiogenesis.

Collectively, our data provide compelling evidence that angiogenic factors acting in concert with proinflammatory molecules and probably with the RAS, induce HPCs to differentiate into monocytes, leading to a hematopoietic imbalance, which in turn should contribute to the early stages of tumour growth as well as in the angiogenic phase, since it was observed in patients at all stages. Additionally, it is of great relevance that hematopoietic values obtained allowed discriminate between healthy donors and CRC patients, suggesting that enumeration of circulating CD14 ⁺ monocytes and CD34 ⁺ cells expressing BB9 may have diagnostic potential for CRC. In fact, when values from peripheral blood HPCs and monocytes were analyzed using ROC curve, CRC patients could be distinguished from unaffected controls with 89.3% specificity and 79.7% sensitivity; however, sensitivity increased up to 97.8%, whilst specificity was not hardly affected when percentage values of CD34 ⁺ BB9 ⁺ progenitor cells were considered, implying that these hematopoietic cell subsets may be sensitive markers to detect CRC, even at early-stage.

Mortality resulting from CRC is largely preventable if the disease is detected early. Although colonoscopy remains the most efficient screening test, the degree of compliance and the acceptance of this technique in the general public are low; therefore, sensitive stool- or blood-based tests are preferred.

Our assays rely on enumeration of specific hematopoietic cell subsets by flow cytometry, and therefore practical for clinical use. In summary, we have identified HPCs and CD14 ⁺ monocytes as accurate biomarkers for the early detection of CRC in small volumes of blood. Moreover, HPCs hold the potential to be used in combination with other well-known angiogenic or inflammatory markers. Although these results need adequate validation in a larger set of samples, they are consistent with the contribution of myeloid cells to tumor angiogenesis. Additionally, if as in experimental tumor models, myeloid cells confer resistance to anti-angiogenic treatment; it would be of great interest to validate these proposed biomarkers to monitor efficacy of VEGF- targeted therapy.

References

1. Lyden D, Hattori K, Dias S, et al. Impaired recruitment of bone marrow-derived endothelial and hematopoietic precursor cells blocks tumour angiogenesis and growth. Nat Med 2001 ; 7 : 1194-201.

2. Yu L, Su B, Hollomon M, Deng Y, Facchinetti V, Kleinerman ES.

Vasculogenesis driven by bone marrow-derived cells is essential for growth of Ewing's sarcomas. Cancer Res 2010; 70: 1334-43.

3. Purhonen S, Palm J, Rossi D, et al. Bone marrow-derived circulating endothelial precursors do not contribute to vascular endothelium and are not needed for tumour growth. Proc Natl Acad Sci USA 2008; 105:6620-5.

4. Spring H, Schiller T, Arnold B, Hammerling GJ, Ganss R. Chemokines direct endothelial progenitors into tumour neovessels. Proc Natl Acad Sci USA 2005;

102: 18111-6.

5. Ziegelhoeffer T, Fernandez B, Kostin S, et al. Bone marrow-derived cells do not incorporate into the adult growing vasculature Circ Res 2004: 94:230-8.

6. Peters BA, Diaz LA, Polyak K, et al. Contribution of bone marrow-derived endothelial cells to human tumour vasculature. Nat Med 2005; 11 :261-2.

7. Mancuso P, Burlini A, Pruneri G, Goldhirsch A, Martinelli G, Bertolini F.

Resting and activated endothelial cells are increased in the peripheral blood of cancer patients. Blood 2001; 97:3658-61.

8. Beaudry P, Force J, Naumov GN, et al. Differential effects of vascular endothelial growth factor receptor-2 inhibitor ZD6474 on circulating endothelial progenitors and mature circulating endothelial cells: implications for use as a surrogate marker of antiangiogenic activity. Clin Cancer Res 2005; 11 :3514-22.

9. Sessa C, Guibal A, Del Conte G, Ruegg C. Biomarkers of angiogenesis for the development of antiangiogenic therapies in oncology: tools or decorations?. Nat Clin Prac Oncology 2008; 5:378-91.

10. Rafii S, Lyden D, Benezra R, Hattori K, Heissig B. Vascular and hematopoietic stem cells: novel targets for anti-angiogenesis therapy? Nat Rev Cancer 2002; 2:826-35.

11. Takakura N, Watanabe T, Suenobu S et al. A role for hematopoietic stem cells in promoting angiogenesis. Cell 2000; 102: 199-209. 12. Pollard JW. Tumour-educated macrophages promote tumour progression and metastasis. Nat Rev Cancer 2004; 4:71-8.

13. Dirkx AE, Oude Egbrink MG, Wagstaff J, Griffioen AW. Monocyte/macrophage infiltration in tumours: modulators of angiogenesis. J Leukoc Biol 2006; 80: 1183-96.

14. De Palma M, Venneri MA, Galli R, et al. Tie2 identifies a hematopoietic lineage of proangiogenic monocytes required for tumour vessel formation and a mesenchymal population of pericyte progenitors. Cancer Cell 2005; 8:211-26

15. Encabo A, Mateu E, Carbonell-Uberos F, Minana MD. IL-6 precludes the differentiation induced by IL-3 on expansion of CD34+ cells from cord blood.

Haematologica 2003; 88:388-95.

16. Hattori K, Heissig B, Tashiro K, et al. Plasma elevation of stromal cell-derived factor- 1 induces mobilization o mature and immature hematopoietic progenitor and stem cells. Blood 2001; 97:3354-60.

17. Ceradini DJ, Kulkarni AR, Callaghan MJ, et al. Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med 2004; 10:858-64.

18. Dlubek D, Drabczak-Skrzypek D, Lange A. Low CXCR4 membrane expression on CD34+ cells characterizes cells mobilized to blood. Bone Marrow Transplant 2006; 37: 19-23.

19. Kollet O, Shivtiel S, Chen YQ, et al. HGF, SDF-1, and MMP-9 are involved in stress-induced human CD34 ⁺ stem cell recruitment to the liver. J Clin Invest. 2003; 112: 160-9.

20. Jokubaitis VJ, Sinka L, Driessen R, et al. Angiotensin-converting enzyme (CD 143) marks hematopoietic stem cells in human embryonic, fetal, and adult hematopoietic tissues. Blood 2008; 111 :4055-63.

21. Zambidis ET, Park TS, Yu W, et al. Expression of ACE (CD143) identifies and regulates primitive hemangioblasts derived from human pluripotent stem cells. Blood 2008; 112:3601-14.

22. Willet CG, Boucher Y, di Tomaso E, et al. Direct evidence that the VEGF- specific antibody bevacizumab has antivascular effects in human rectal cancer. Nat Med 2004; 10: 145-7.

23. Duda DG, Cohen KS, di Tomaso E, et al. Differential CD146 expression on circulating versus tissue endothelial cells in rectal cancer patients: implications for circulating endothelial and progenitor cells as biomarkers for antiangiogenic therapy. J Clin Oncol 2006; 24: 1449-53.

24. De Palma M, Venneri MA, Roca C, Naldini L. Targeting exogenous genes to tumour angiogenesis by transplantation of genetically modified hematopoietic stem cells. Nat Med 2003; 9:789-95.

25. Grunewald M, Avraham I, Dor Y, et al. VEGF-induced adult neovascularization: recruitment, retention, and role of accessory cells. Cell 2006; 124: 175-89.

26. Coffelt SB, Tal AO, Scholz A, et al. Angiopoietin-2 regulates gene expression in TIE2-expressing monocytes and augments their inherent proangiogenic functions. Cancer Res; 70:5270-80.

27. Venneri MA, De Palma M, Ponzoni M, et al. Identification of proangiogenic TIE2-expressing monocytes (TEMs) in human peripheral blood and cancer. Blood 2007; 109:5276-85.

28. Ziegler-Heitbrock L. The CD14+CD16+ blood monocytes: their role in infection and inflammation. J Leukoc Biol. 2007; 81 :584-92.

29. Terzic J, Grivennikov S, Karin E, Karin M. Inflammation and colon cancer.

Gastroenterology. 2010; 138:2101-14.e5.

30. Monaco C, Andreakos E, Kiriakidis S, Feldmann M, Paleolog E. T-cell-mediated signalling in immune, inflammatory and angiogenic processes: the cascade of events leading to inflammatory diseases. Curr Drug Targets Inflamm Allergy.

2004 ;3:35-42.

31. Stockmann C, Doedens A, Weidemann A, et al. Deletion of vascular endothelial growth factor in myeloid cells accelerates tumourigenesis. Nature 2008; 456:814- 8.

32. Carlo-Stella C, Di Nicola M, Longoni P, et al. Placental growth factor- 1 potentiates hematopoietic progenitor cell mobilization induced by granulocyte colony-stimulating factor in mice and nonhuman primates. Stem Cells 2007;25:252-61.

33. Hattori K, Dias S, Heissig B, et al. Vascular endothelial growth factor and angiopoietin-1 stimulate postnatal hematopoiesis by recruitment of vasculogenic and hematopoietic stem cells. J Exp Med 2001; 193: 1005-14.

34. Okamoto R, Ueno M, Yamada Y, et al. Hematopoietic cells regulate the angiogenic switch during tumourigenesis. Blood 2005;105:2757-63.

35. Jin H, Aiyer A, Su J, et al. A homing mechanism for bone marrow-derived progenitor cell recruitment to the neovasculature. J Clin Invest 2006;116:652-62. Kitajima S, Liu E, Morimoto M, et al. Transgenic rabbits with increased VEGF expression develop hemangiomas in the liver: a new model for Kasabach-Merritt syndrome. Lab Invest 2005; 85: 1517-27.

Aguayo A, Estey E, Kantarjian H, et al. Cellular vascular endothelial growth factor is a predictor of outcome in patients with acute myeloid leukemia. Blood 1999;94:3717-21.

Ohnishi S, Miyagishima T, Nakagawa M, et al. Diffuse neonatal hemangiomatosis without cutaneous lesions in an adult— a case report. Angiology 2002;53:235-7.

Willett CG, Duda DG, di Tomaso E, et al. Efficacy, Safety, and Biomarkers of Neoadjuvant Bevacizumab, Radiation Therapy, and Fluorouracil in Rectal Cancer: A Multidisciplinary Phase II Study. J Clin Oncol 2009; 27:3020-6.

Xu L, Duda DG, di Tomaso E, et al. Direct evidence that bevacizumab, an anti- VEGF antibody, up-regulates SDF1α, CXCR4, CXCL6, and neuropilin 1 in tumours from patients with rectal cancer. Cancer Res 2009; 69:7905-10.

Jin DK, Shido K, Kopp HG, et al. Cytokine-mediated deployment of SDF-1 induces revascularization through recruitment of CXCR4+ hemangiocytes. Nat Med 2006; 12 :557-67.

Park TS, Zambidis ET. A role for the rennin-angiotensin system in hematopoiesis. Haematologica 2009 94:745-7.

Haznedaroglu IC, Oztiirk MA. Towards the understanding of the local hematopoietic bone marrow rennin-angiotensin system. Int J Biochem Cell Biol 2003; 35 :867-80.

Bouquet C, Lamande N, Brand M, et al. Suppression of angiogenesis, tumour growth, and metastasis by adenovirus-mediated gene transfer of human angiotensinogen. Mol Ther 2006; 14: 175-82.

Vincent F, Bonnin P, Clemessy M, et al. Angiotensinogen delays angiogenesis and tumour growth of hepatocarcinoma in transgenic mice. Cancer Res 2009; 69:2853-60.

Deshayes F, Nahmias C. Angiotensin receptors: a new role in cancer?. Trends Endocrinol Metab 2005; 16:293-9.

Egami K, Murohara T, Shimada T, et al. Role of host angiotensin II type 1 receptor in tumour angiogenesis and growth. J Clin Invest 2003; 112 : 67-75. Abangan Jr RS, Williams CR, Mehrotra M, Duncan JD, LaRue AC. MCPl directs trafficking of hematopoietic stem cell-derived fibroblast precursors in solid tumour. Am J Pathol 2010; 176: 1914-26.