This study identifies and characterizes the antigen recognized by monoclonal antibody (mAb) 14C5. We compared the expression of antigen 14C5 with the expression of eight integrin subunits (α1, α2, α3, αv, β1, β2, β3, and β4) and three integrin heterodimers (αvβ3, αvβ5, and α5β1) by flow cytometry. Antigen 14C5 showed a similar expression to αvβ5 in eight different epithelial cancer cell lines (A549, A2058, C32, Capan-2, Colo16, HT-1080, HT-29, and SKBR-3). Specific binding of P1F6, an anti-αvβ5 specific antibody, was blocked by mAb 14C5. After transient expression of αvβ5 in 14C5-negative Colo16 cells, mAb 14C5 was able to bind a subpopulation of αvβ5-positive cells. We evaluated the tissue distribution of the 14C5 antigen in colon (n = 20) and lung (n = 16) cancer tissues. The colon carcinoma cells stained positive for 14C5 in 50% of tumors analyzed, whereas bronchoalveolar lung carcinoma and typical carcinoid were not positive for the antigen. More common types of non–small cell lung cancer, i.e., squamous (n = 5) and adenocarcinoma (n = 3), stained positive in 2 of 5 squamous carcinomas and in 1 of 3 investigated adenocarcinoma. Colon (95%) and lung (50%) carcinoma tissues showed extensive expression of antigen 14C5 in the stroma surrounding the tumor cells and on the membrane of the adjacent fibroblasts. We show for the first time that mAb 14C5 binds the vascular integrin αvβ5, suggesting that mAb 14C5 can be used as a screening agent to select colon and lung cancer patients that are eligible for anti-αvβ5–based therapies. [Mol Cancer Ther 2008;7(12):3771–9]

In search of new antibody therapeutics for therapy of metastatic breast cancer, several mouse monoclonal antibodies (mAb) were developed against epitopes on the extracellular membrane of SK-BR-3 human breast cancer cells (1). One of these mAbs is the IgG1 mAb 14C5, which recognizes an extracellular plasma membrane antigen expressed on SK-BR-3 and MCF-7 human breast cancer cells (1, 2). The antibody is capable of reversibly inhibiting the adhesion of SK-BR-3 cells to culture-treated plastic and pronectin-, fibronectin-, osteopontin-, and vibronectin-precoated culture plates.

Coene et al. (2) showed abundant expression of antigen 14C5 (Ag 14C5) on the tumor surface of in situ and invasive breast cancer tissue. The antigen was specifically overexpressed in 64% of invasive ductal adenocarcinomas of the breast (n = 33), in all investigated cases of invasive squamous cell carcinoma (n = 7), and in 40% of basocellular carcinomas of the skin (n = 5). Ag 14C5 is located on the cell membrane of the carcinoma cells. When the tumor is highly invasive, 65% of the cases also show an extensive stromal expression on the fibroblasts between the tumor cells (n = 71). In normal tissues and stroma surrounding in situ carcinomas of the breast (n = 15), no expression of the Ag 14C5 occurred (2). Normal epithelial, muscle, and connective tissues that did not show staining with mAb 14C5 were skin, thyroid, parathyroid, colon, stomach, lung, uterine tube, ovary, ureter, urethra, lymph node, nerve, chondroid tissue, skeletal muscle, esophagus, and artery (1, 2). Low-level staining of myoepithelial cells was sometimes observed in biopsies of breast tissue and in tubular cells of the kidney.

Furthermore, mAb 14C5 has been shown to prevent invasion and, subsequently, metastasis of SK-BR-3 and MCF-7 cells on host tissue in vitro (1). In addition, mAb 14C5 significantly inhibits tumor growth in a dose-dependent manner in Sprague-Dawley rats bearing HH-16 clone 1/2 adenocarcinomas or fibrosarcomas overexpressing the antigen 14C5 (3, 4). Therefore, antigen 14C5 offers an interesting target for anti-invasion, antiangiogenesis, or other antibody-based targeted strategies such as radioimmunotherapy. We have previously shown the biodistribution and stable antigen-binding properties of radio-iodinated mAb 14C5 (5, 6) and its Fab and F(ab′)2 fragments (7). mAb 14C5 readily internalizes in cancer cells, which makes it an attractive candidate for radioimmunotherapy (6).

Several attempts were made to identify the antigen 14C5. Immunoprecipitation of the putative antigen from MCF-7 cells was done using mAb 14C5 (1). Analysis by SDS-PAGE in the presence of a reducing agent showed that mAb 14C5 recognizes a protein fragment with a molecular size of 90,000 kDa (2). Based on the immunohistochemical staining of Ag 14C5 on the cell membrane extensions of highly invasive tumor cells, it had been suggested that Ag 14C5 could be related to the family of integrins (1, 2). The integrins, a family of related membrane receptors involved in cell-cell and cell-matrix interactions, are heterodimeric complexes of α- and β-subunits. The cell substrate adhesion inhibition after coating with osteopontin and vitronectin suggests that the antigen could be a receptor for these extracellular matrix proteins. Osteopontin is recognized by the αvβ3 vitronectin receptor. Integrins can also bind to different extracellular matrix proteins (8, 9).

The objective of this study was to identify the antigen 14C5 by investigating if Ag 14C5 could be related to the family of integrins and also to characterize the tissue distribution of antigen 14C5 in colon and lung cancer tissues to determine its usefulness for therapeutic development.

Antibodies

A monoclonal antibody was purified from 14C5 antibody-producing hybridoma cells as described previously (6). The 9E10 mAb (anti–c-myc, mouse IgG1, Becton Dickinson, BD) was used as a control for nonspecific Fc-receptor binding during flow cytometric analysis. Alexa Fluor 488–conjugated goat anti-mouse IgG antibody (Invitrogen) was used during flow cytometry. Integrin antibodies (Millipore) used in flow cytometric analysis were anti-α1 (FB12, mouse IgG1), anti-α2 (P1E6, mouse IgG1), anti-α3 (ASC-1, mouse IgG1), anti-αv (P3G8, mouse IgG1), anti-β1 (P4G11, mouse IgG1), anti-β2 (P4H9-A11, mouse IgG3), anti-β3 (25E3, mouse IgG2a), anti-β4 (ASC-3, mouse IgG1), anti-α5β1 (HA5, mouse IgG2b), anti-αvβ3 (LM609, mouse IgG1), and anti-αvβ5 (P1F6, mouse IgG1). Anti-β5 (B5-IVF2, mouse IgG1) was obtained from Sigma.

Cell Lines

HT-29 (colon), Capan-2 (pancreas), and C32 (melanoma) cancer cells were a kind gift of J&J Pharmaceutical Research & Development. HT-1080 (fibrosarcoma), A2058 (melanoma), and A549 (lung) cancer cells were obtained from the Laboratory of Tumor and Developmental Biology (University of Liège, Liege, Belgium). SK-BR-3 (breast), Colo16 (squamous carcinoma), and MOLT-4 (lymphoblastic leukemia) cancer cells were obtained from the N. Goormaghtigh Institute of Pathology, University of Ghent, Ghent, Belgium. All cell lines were cultured in standard medium with supplements according to the American Type Culture Collection recommendations, except for Colo16, and MOLT-4 was grown in RPMI 1640 medium containing 10 mmol/L HEPES and 10% fetal bovine serum (Cambrex).

Flow Cytometry

Confluent cells were harvested using the nonenzymatic Cell Dissociation Buffer (Invitrogen). Aliquots of 2 × 105 cells were incubated with mAb 14C5 or integrin antibodies under saturating conditions (67 nmol/L) in PBS-0.5% bovine serum albumin-0.02% (w/v) sodium azide (Sigma; w/v; PBSA) on ice for 2 h. After washing the cells with PBSA, cells were incubated with Alexa Fluor 488–conjugated goat anti-mouse on ice for 1 h. Again, cells were washed with PBSA and suspended in a final volume of 300 μL of PBSA. In control samples, the primary antibody was replaced by the mAb 9E10 (67 nmol/L) isotype control or omitted.

Flow cytometric analysis was done using a FACScan flow cytometer (Becton Dickinson). Tumor cell populations were gated based on forward- and side-scatter variables. Data analysis was done using the cytometry software WinMDI (Joseph Trotter).

αv and β5 cDNA Transfection of Colo16 Cells

Full-length cDNA encoding the human β5 integrin (RZPD) was amplified using BETA5 F (sense, ATGAATTCGCCACCATGCCGCGG) and BETA5 B (antisense, ATGGATCCTCAGTCCACAGTGCC) primers and was ligated into the mammalian expression vector pES31, a derivative of pCAGGS (10), containing a zeocin selection site. The full-length cDNA encoding the human αv integrin (OriGene Technologies) was amplified by PCR using ALPHAV F (sense, ATGAATTCGCCACCATGGCTTTT) and ALPHAV B (antisense, ATGGATCCTTAAGTTTCTGAGTTT) primers and was ligated into pES31. In this vector, the zeocin-selectable marker was replaced by a neomycin selection gene derived from the pCDNA3 expression vector by BamHI/PvuI and XhoI restriction digest. The sense primers were designed to contain an EcoRI site and Kozak consensus; the antisense primers contained a BamHI restriction site. These EcoRI and BamHI sites were used for integrating the amplicon in the vector. The complete sequence of both pES31αvneo and pES31β5zeo were confirmed by sequencing using the dideoxynucleotide chain termination method with the Big Dye Terminator version 3.1 cycle sequencing kit (Applied Biosystems). All nucleotide analyses were done using Clone Manager 8.0 (Scientific & Educational Software).

Transfection into antigen 14C5–negative Colo16 was done using the Saint-Mix Lipofectamine reagent (Synvolux Therapeutics) according to the manufacturer's instructions. Transfections were done in 6-well dishes with 1 μg DNA per well (1 μg vector containing αv, 1 μg vector containing β5, or 0.5 μg of both αv and β5 containing vectors). After 48 h, flow cytometry analysis was done using anti-αv (P3G8), anti-β5 (B5-IVF2), and anti-αvβ5 (P1F6) specific antibodies.

Blocking Experiment

Blocking studies were done with mAb 14C5, anti-αvβ5 (P1F6), anti-αv (P3G8), anti-β5 (B5-IVF2), and anti-αvβ3 (LM609) antibodies. Antibodies were radio-iodinated with 123I by the Iodo-Gen method as previously described (6). Unbound radio-iodine was removed by gel filtration over a PD-10 desalting column equilibrated with 0.5% BSA in PBS. Washed target cells (SKBR-3, 0.5 × 106) were incubated with 300 nmol/L of the unlabeled mAb 14C5 to block antigen 14C5. Test solutions containing increasing amounts of 123I-labeled mAb 14C5 or 123I-labeled anti-αvβ5 in a total volume of 1 mL cell medium were incubated at 4°C for 2 h. Then, the supernatant was removed by centrifugation (15 s, 8,000 rpm, 4°C) and cells were washed twice with 1 mL ice-cold PBS and peletted for 15 s at 8,000 rpm. For each sample, total binding was determined in the absence of unlabeled mAb 14C5. Radioactivity was counted by a gamma counter (Cobra II, Perkin-Elmer). Incubations were done in triplicate.

Saturation Binding Study

Saturation binding studies were done with 123I-labeled mAb 14C5 and 123I-labeled P1F6 using SKBR-3 cells. Twelve test solutions (in triplicate) containing increasing amounts of radiolabeled antibody and 0.5 × 106 cells in a total volume of 1 mL cell medium were incubated at 4°C for 2 h. Then, the supernatant was removed by centrifugation (15 s, 8,000 rpm, 4°C) and cells were washed twice with 1 mL ice-cold PBS. For each sample, nonspecific binding was determined in the presence of 300 nmol/L of the unlabeled mAb 14C5. Radioactivity was counted by a gamma counter (Cobra II, Perkin-Elmer). Kd values were determined with GraphPad Prism 3.1 software.

Immunohistochemistry

Biopsy specimens from human colon and lung epithelial tumors and normal colon and lung tissues were examined for the expression of the antigen 14C5. Sections (5 μm) from fresh frozen tissues were cut and fixed in acetone for 1 min at 4°C. For immunohistochemical detection, an avidin-biotin system (DakoCytomation) was applied using a 1 mg/mL dilution of mAb 14C5 as previously described (2).

Statistical Analysis

A nonparametric Mann-Whitney test was used for comparisons.

Comparative Flow Cytometry Analysis with Integrin Antibodies and mAb 14C5

A comparison of mAb 14C5 and 11 integrin antibodies (α1, α2, α3, αv, β2, β3, β4, α5β1, αvβ3, and αvβ5) was done by flow cytometry analysis under saturating conditions. Figure 1A shows the reactivity of mAb 14C5 and anti-integrin antibodies with the positive control, the SK-BR-3 breast cancer cell line. Figure 1B shows the reactivity with the negative control, the Colo16 squamous carcinoma cell line. Expression patterns of six other cell lines (A2058 skin melanoma, A549 lung carcinoma, C32 melanoma, Capan-2 pancreatic carcinoma, HT1080 fibrosarcoma, and HT29 colon carcinoma) were obtained (data not shown).

Figure 1.

Flow cytometry analysis showing binding of irrelevant anti–c-myc IgG1 (open histograms, thin line), mAb 14C5 (open histograms, thick line), or integrin antibodies (gray-filled histograms) under saturating conditions. Different anti-integrin antibodies were compared with binding of mAb 14C5 to (A) SKBR-3 breast cancer cells and (B) Colo16 squamous carcinoma cells. Binding to different cell lines of (C) anti-αv and (D) anti-αvβ5 was compared with mAb14C5.

Figure 1.

Flow cytometry analysis showing binding of irrelevant anti–c-myc IgG1 (open histograms, thin line), mAb 14C5 (open histograms, thick line), or integrin antibodies (gray-filled histograms) under saturating conditions. Different anti-integrin antibodies were compared with binding of mAb 14C5 to (A) SKBR-3 breast cancer cells and (B) Colo16 squamous carcinoma cells. Binding to different cell lines of (C) anti-αv and (D) anti-αvβ5 was compared with mAb14C5.

Close modal

Two antibodies showed similar reactivity as mAb 14C5, namely anti-αv and anti-αvβ5 (Fig. 1C-D). However, differences in detection levels of anti-αv and mAb 14C5 were observed in A2058 and C32 cells (Fig. 1C), whereas anti-αvβ5 showed equal reactivity in all tested cell lines (Fig. 1D). All other integrin antibodies showed differences in detection levels in more than two cell lines.

mAb 14C5 Binds Transiently Expressed αvβ5 Integrin in Colo16 Cells

To determine whether antigen 14C5 is the integrin αvβ5 receptor, Colo16 cells were cotransfected with a vector containing a neomycin resistance cassette and the human αv integrin subunit and a second expression vector containing a zeocin resistance cassette and the β5 integrin subunit. After transfection of Colo16 cells with αv or β5 or cotransfection of αv and β5, ∼30% of transfected cells showed expression of the αv subunit, the β5 subunit, or αvβ5, respectively (Fig. 2A). Some overlap is observed between mAb 14C5 binding and the anti-αv subunit or anti-β5 subunit alone. Only when both integrin subunits were cotransfected, the expression of mAb 14C5 matched completely with the anti-αvβ5 expression levels. After cotransfection of αv and β5 genes, the positive transfectant subpopulation binds anti-αvβ5–specific antibody and mAb 14C5 at equivalent levels. These results suggest that integrin αvβ5 might be the antigen for mAb 14C5 and that both subunits are involved in the antigen-antibody binding.

Figure 2.

Identification of antigen 14C5. A, flow cytometry analysis of Colo16 transfected cells with human αvv+/β5−), β5v−/β5+), or αvβ5v+/β5+). Cell surface expression was analyzed using anti-αv (P3G8, dotted line), anti-β5 (B5-IVF2, dashed line), and anti-αvβ5 (P1F6, full line) specific antibodies and compared with mAb 14C5 (gray histograms) binding. B, blocking of integrin αvβ5 by mAb 14C5. Total binding (filled triangles) and nonspecific binding (open triangles) to SKBR-3 cells of 123I-mAb 14C5, 123I-labeled anti-αvβ5, and 123I-labeled anti-αv are shown. Bars, SD (n = 3).

Figure 2.

Identification of antigen 14C5. A, flow cytometry analysis of Colo16 transfected cells with human αvv+/β5−), β5v−/β5+), or αvβ5v+/β5+). Cell surface expression was analyzed using anti-αv (P3G8, dotted line), anti-β5 (B5-IVF2, dashed line), and anti-αvβ5 (P1F6, full line) specific antibodies and compared with mAb 14C5 (gray histograms) binding. B, blocking of integrin αvβ5 by mAb 14C5. Total binding (filled triangles) and nonspecific binding (open triangles) to SKBR-3 cells of 123I-mAb 14C5, 123I-labeled anti-αvβ5, and 123I-labeled anti-αv are shown. Bars, SD (n = 3).

Close modal

mAb 14C5 Blocks Specific Binding of 123I-Labeled Anti-αvβ5 Antibody (P1F6)

To reinforce the identity of antigen 14C5 as integrin αvβ5, a blocking experiment was done with 123I-labeled mAb 14C5 and 123I-labeled anti-αvβ5–specific antibody. The blocking assay was done by monitoring loss of the ability of 123I-labeled antibodies to bind to target cells in the presence of excess of unlabeled mAb 14C5. The blocking agent used in the study represents a 10-fold excess of saturating levels obtained in saturating binding analysis with 123I-labeled mAb 14C5 (300 nmol/L; Fig. 2B). Figure 2B shows that the total binding of 123I-labeled anti-αvβ5 is similar to the total binding of 123I-labeled mAb 14C5. In the presence of 300 nmol/L unlabeled mAb 14C5, total binding was reduced to nonspecific binding levels both for mAb 14C5 and anti-αvβ5. On the contrary, the binding of anti-αv was not blocked by mAb 14C5. As seen in the flow cytometry studies, there was no specific binding of 123I-labeled β5 and 123I-labeled αvβ3 to SKBR-3 cells (data not shown). These results show that mAb 14C5 and anti-αvβ5 bind the same antigen and confirm previous tentative results where transient expression of the αvβ5 integrin leads to specific binding of mAb 14C5 to a subpopulation of αvβ5-expressing transfectants.

mAb 14C5 Binds to Integrin αvβ5 with Higher Affinity than P1F6

Supplementary Fig. S14

4

Supplementary material for this article is available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

shows the saturation plots of 123I-labeled mAb 14C5 to SKBR-3 cells (Supplementary Fig. S1A)4 and 123I-labeled P1F6 to SKBR-3 cells (Supplementary Fig. S1B).4 mAb 14C5 bound to SKBR-3 with a 10-fold higher affinity (Kd 0.16 ± 0.01 nmol/L) than P1F6 (Kd 1.40 ± 0.11 nmol/L). The Kd found with 123I-mAb 14C5 and SKBR-3 cells was similar to what has been reported on the binding affinity of 125I-mAb 14C5 to A549 lung cancer cells (Kd 0.19 ± 0.07 nmol/L) or LoVo colon cancer cells (0.20 ± 0.05 nmol/L; ref. 6).

mAb 14C5 Targets Integrin αvβ5 in Lung and Colon Cancer

The analysis of human colon cancer tissues (n = 20) by immunohistochemical staining is shown in Tables 1 and 2. The carcinoma cells stained positive in 50% of tumor samples analyzed. This staining was generally weak; however, there was intermittent strong focal staining of cancer cells (Fig. 3A). Sometimes, only apical staining of tumor cell membrane was observed (Fig. 3B). However, the fibroblasts in the stroma surrounding the tumor cells showed extensive positive staining for Ag 14C5.

Table 1.

Antigen 14C5 expression on human colon cancer tissues

Patient
Ag 14C5 expression*
Tumor
Normal
No.DiagnosisStromaCells
Rectum CA +/− − NA 
Colon CA − NA 
Colon CA +++ − NA 
Rectum CA ++ − NA 
Rectum CA ++ +/− NA 
Rectum CA − NA 
Colon CA +++ − NA 
Rectum CA NA 
Colon CA − NA 
10 Rectum CA +++ NA 
11 Colon CA +++ — 
12 Rectum CA ++ NA 
13 Rectum CA (Fig. 3A) +++ ++ NA 
14 Colon CA − — 
15 Rectum CA − NA 
16 Colon CA NA 
17 Colon CA +++ +/++ NA 
18 Colon CA (Fig. 3B) +++ +/++ NA 
19 Recto-sigmoid CA NA 
20 Colon CA − NA 
Patient
Ag 14C5 expression*
Tumor
Normal
No.DiagnosisStromaCells
Rectum CA +/− − NA 
Colon CA − NA 
Colon CA +++ − NA 
Rectum CA ++ − NA 
Rectum CA ++ +/− NA 
Rectum CA − NA 
Colon CA +++ − NA 
Rectum CA NA 
Colon CA − NA 
10 Rectum CA +++ NA 
11 Colon CA +++ — 
12 Rectum CA ++ NA 
13 Rectum CA (Fig. 3A) +++ ++ NA 
14 Colon CA − — 
15 Rectum CA − NA 
16 Colon CA NA 
17 Colon CA +++ +/++ NA 
18 Colon CA (Fig. 3B) +++ +/++ NA 
19 Recto-sigmoid CA NA 
20 Colon CA − NA 

NOTE: +/−, weak positive; +, positive; ++, strong positive; +++, very strong positive; −, negative.

Abbreviations: NA, not available; CA, carcinoma.

*

Results show scoring of cryosections evaluated both with mAb 14C5 and P1F6.

Table 2.

Summary of the immunohistochemical investigation of human lung and colon carcinomas and their surrounding stroma evaluated with mAb 14C5 and P1F6

Tumor typenStroma
Tumor cells
++
Lung      
    Squamous cell carcinoma 5 (100%) 0 (0%) 2 (40%) 3 (60%) 
    Adenocarcinoma 3 (100%) 0 (0%) 1 (33%) 2 (67%) 
    BAC 1 (25%) 3 (75%) 0 (0%) 4 (100%) 
    Large cell cancer 1 (100%) 0 (0%) 0 (0%) 1 (100%) 
    Carcinoid 0 (0%) 1 (100%) 0 (0%) 1 (100%) 
Colon      
    Adenocarcinoma 20 19 (95%) 1 (5%) 10 (50%) 10 (50%) 
Tumor typenStroma
Tumor cells
++
Lung      
    Squamous cell carcinoma 5 (100%) 0 (0%) 2 (40%) 3 (60%) 
    Adenocarcinoma 3 (100%) 0 (0%) 1 (33%) 2 (67%) 
    BAC 1 (25%) 3 (75%) 0 (0%) 4 (100%) 
    Large cell cancer 1 (100%) 0 (0%) 0 (0%) 1 (100%) 
    Carcinoid 0 (0%) 1 (100%) 0 (0%) 1 (100%) 
Colon      
    Adenocarcinoma 20 19 (95%) 1 (5%) 10 (50%) 10 (50%) 

NOTE: BAC, bronchoalveolar adenocarcinoma.

Figure 3.

Immunohistochemistry with mAb 14C5 on frozen tissue of human colon and non–small cell lung carcinomas. A, adenocarcinoma of the rectum. There is a distinct expression on the cytoplasmic membrane of the carcinoma cells. The surrounding fibrous tissue and the fibroblasts express the antigen 14C5 at high level (×400). B, colon adenocarcinoma with apical expression on the cytoplasmic membrane of the carcinoma cells. There is also a distinct expression of the antigen 14C5 on the surrounding fibrous tissue and the fibroblasts (×200). C, moderately differentiated squamous lung carcinoma. Focal positive staining of cell membranes (×400). D, moderately differentiated squamous lung carcinoma. No staining of tumor cell membranes was observed. The surrounding fibrous tissue and the fibroblasts express the antigen 14C5 at high level (×200). E, lung adenocarcinoma. Strong to very strong expression of the antigen 14C5 on the surrounding fibrous tissue and the fibroblasts (×400). F, moderately differentiated squamous lung carcinoma. Focal positive staining of tumor cell membranes with high expression in surrounding stroma (×400). Filled arrows, tumor cell membrane staining; Asterisk, stroma.

Figure 3.

Immunohistochemistry with mAb 14C5 on frozen tissue of human colon and non–small cell lung carcinomas. A, adenocarcinoma of the rectum. There is a distinct expression on the cytoplasmic membrane of the carcinoma cells. The surrounding fibrous tissue and the fibroblasts express the antigen 14C5 at high level (×400). B, colon adenocarcinoma with apical expression on the cytoplasmic membrane of the carcinoma cells. There is also a distinct expression of the antigen 14C5 on the surrounding fibrous tissue and the fibroblasts (×200). C, moderately differentiated squamous lung carcinoma. Focal positive staining of cell membranes (×400). D, moderately differentiated squamous lung carcinoma. No staining of tumor cell membranes was observed. The surrounding fibrous tissue and the fibroblasts express the antigen 14C5 at high level (×200). E, lung adenocarcinoma. Strong to very strong expression of the antigen 14C5 on the surrounding fibrous tissue and the fibroblasts (×400). F, moderately differentiated squamous lung carcinoma. Focal positive staining of tumor cell membranes with high expression in surrounding stroma (×400). Filled arrows, tumor cell membrane staining; Asterisk, stroma.

Close modal

The analysis of human non–small cell lung cancer (NSCLC) tissues (n = 16) by immunohistochemistical staining for Ag 14C5 are shown in Tables 2 and 3. In bronchoalveolar carcinoma, which tends to be organ confined and is less likely to metastasize, the carcinoma cells stained negative and stroma surrounding bronchoalveolar carcinoma stained only weakly positive in one case (n = 4). Typical carcinoid represents another uncommon group of pulmonary neoplasms. They are well differentiated and represent the least biologically aggressive type of pulmonary neuroendocrine tumor. These tumors characteristically grow slowly and tend to metastasize infrequently. The investigated carcinoid did not show staining with mAb 14C5. The more common types of NSCLC, i.e., squamous (n = 5) and adenocarcinoma (n = 3), stained focal positive in 2 of 5 squamous carcinomas (Fig. 3C and F) and in 1 of 3 investigated adenocarcinoma (Fig. 3E). Fibroblasts in the stroma surrounding the tumor cells showed positive to very strong positive staining in all squamous and adenocarcinoma samples (n = 8; Fig. 3D-F).

Table 3.

Antigen 14C5 expression on human lung cancer tissue

Patient
Ag 14C5 expression*
Tumor
Normal tissue
No.DiagnosisStromaCells
Md squamous − NA 
Md adeno ++ − NA 
Pd squamous +++ − NA 
BAC − − NA 
Md squamous (Fig. 3C) +++ NA 
BAC − − NA 
Md-wd adeno +++ − NA 
Pd giant cell adeno − NA 
BAC − − NA 
10 Md squamous (Fig. 3D) +++ NA 
11 Carcinoid − − NA 
12 BAC +/− − NA 
13 Md squamous (Fig. 3E) +++ − — 
15 Adeno (Fig. 3F) +++ NA 
16 LCUC ++ − NA 
Patient
Ag 14C5 expression*
Tumor
Normal tissue
No.DiagnosisStromaCells
Md squamous − NA 
Md adeno ++ − NA 
Pd squamous +++ − NA 
BAC − − NA 
Md squamous (Fig. 3C) +++ NA 
BAC − − NA 
Md-wd adeno +++ − NA 
Pd giant cell adeno − NA 
BAC − − NA 
10 Md squamous (Fig. 3D) +++ NA 
11 Carcinoid − − NA 
12 BAC +/− − NA 
13 Md squamous (Fig. 3E) +++ − — 
15 Adeno (Fig. 3F) +++ NA 
16 LCUC ++ − NA 

NOTE: +/−, weak positive; +, positive; ++, strong positive; +++, very strong positive; −, negative.

Abbreviations: Adeno, adenocarcinoma; LCUC, large-cell undifferentiated carcinoma; Md, moderately differentiated; Pd, poorly differentiated; RCC, renal cell carcinoma; wd, well differentiated.

*

Results show scoring of cryosections evaluated both with mAb 14C5 and P1F6.

Table 2 provides a summary of the immunohistochemical investigation of human lung and colon carcinoma tissues and their surrounding stroma. Normal tissues of lung (n = 5) and colon (n = 5) were analyzed for expression of the antigen 14C5. No expression was detected on normal lung tissues (data not shown). On colon tissues, small blood vessels and mucous-secreting cells showed staining with mAb 14C5.

This study was also performed with the αvβ5 specific antibody (P1F6). The location of the staining on the tissues was identical. The semiquantitative scoring of cryosections with mAb 14C5 and P1F6 was the same for these two antibodies (Tables 13). Parameters used were intensity of the staining and number of tumor cells that stained positive.

This work has identified the antigen 14C5 as the integrin αvβ5. In the past, many studies failed to purify the antigen and consequently could not deduce the identity of Ag 14C5 by sequence analysis (1, 2).

It had been suggested that Ag 14C5 could be related to the family of integrins (1, 2). Therefore, we compared the expression of the antigen 14C5 with the expression of eight different integrin subunits (α1, α2, α3, αv, β1, β2, β3, and β4) and three different integrins (αvβ3, αvβ5, and α5β1). We selected these integrins based on their reported role in cancer pathology and blockade of cell-substrate adhesion (1115). At saturating conditions, anti-αvβ5 specific antibody showed equivalent antigen expression levels as mAb 14C5. We successfully blocked the αvβ5 receptors with mAb 14C5, indicating that antigen 14C5 is, or is extremely similar to, the αvβ5 receptor. Flow cytometry analysis of Colo16 cells transfected with human αv and β5 cDNA suggested that both subunits are involved in the mAb 14C5 binding.

The finding that Ag 14C5 is integrin αvβ5 is very important for further drug development using mAb 14C5. The role of integrins αvβ3 and its close relative αvβ5 in cancer growth, angiogenesis, and metastasis has been widely discussed (11, 1518) and remains uncertain and complex. αvβ3 was the first vascular integrin targeted to suppress tumor-induced angiogenesis. The evidence indicates that both αvβ3 and αvβ5 promote angiogenesis via distinct pathways; αvβ3 is involved in the response to basic fibroblast growth factor and tumor necrosis factor α; and αvβ5 is involved in the response to vascular endothelial growth factor and transforming growth factor α (19). Others hypothesized that these integrins could be negative regulators of angiogenesis and that drugs targeting them may act as agonists rather than antagonists (20). This hypothesis is based on the fact that although pharmacologic agents directed against integrins αvβ3 and αvβ5 have been reported to block angiogenesis, genetic ablations of the genes encoding these integrins fail to block angiogenesis. In a later study, both αvβ3 and αvβ5 were not essential for tumor growth and progression, although they might play some role in mammary gland development (18, 21). Moreover, αvβ5 integrins can mediate early steps of metastasis formation (22).

The identification of antigen 14C5 being αvβ5 offers the opportunity to assess the benefit of combining the favorable radioimmunotargeting properties of mAb 14C5 with its other possible mechanisms of action (e.g., blocking angiogenesis, blocking metastasis). These properties of mAb 14C5 should be further assessed and evaluated in comparative therapeutic studies using agents that are already available. RGD peptide inhibitors binding to αvβ3 and αvβ5 have been tested in preclinical (2326) and clinical studies (27). Monoclonal antibodies specific for αv (CNTO 95, fully human; refs. 2831) and αvβ3 (LM609, murine; ref. 32; Vitaxin, MEDI-522, humanized; ref. 33) have been investigated as therapeutic agents. Cilengitide, a cyclic RGD peptide in clinical trials for metastatic cancer, has been tested in an aggressive breast cancer model in combination with 90Y-labeled chimeric anti-L6 radioimmunotherapy, which remarkably increased efficacy and increased apoptosis, compared with single-modality therapy with either agent without additional toxicity (34). Noninvasive positron emission tomography and single-photon emission computed tomography imaging of integrin expression have been done with monomeric, dimeric, and tetrameric RGD peptides and peptidomimetics (3538). We have shown here that mAb 14C5 has a 10-fold higher affinity than P1F6. The higher affinity of mAb 14C5 compared with P1F6 could be beneficial to the delivery of payloads, such as radioisotopes, to the tumor where higher binding affinity could retain the antibody better within the tumor tissue. Also, although both antibodies scored equal in intensity and localization in the immunohistochemistry studies, the nonspecific background obtained with the P1F6 antibody was higher than mAb 14C5, making the use of mAb 14C5 for staining of crysections more favorable. Further studies are necessary to investigate whether the two antibodies have differences in their working mechanism that may lead to a different use of these antibodies in targeting αvβ5–expressing cancer cells.

Our previous results showed the potential role of the radiolabeled anti-αvβ5 specific antibody mAb 14C5 in radioimmunodetection and radioimmunotherapy (57). In these studies, we used lung and colon cancer xenograft models in mice to investigate the in vivo properties of radioiodinated mAb 14C5. Although the expression of Ag 14C5 had been extensively investigated in breast cancer tissues, its expression had not been investigated in human colon and lung cancer tissues. In these tumor types, Ag 14C5 expression was detected predominantly in the stroma surrounding the tumor cells and on the stromal fibroblasts (squamous cell carcinoma, 5 of 5; adenocarcinoma of lung, 3 of 3; large cell carcinoma, 1 of 1; colon adenocarcinoma, 19 of 20) and only in some cases was localized at the tumor cells themselves (squamous cell carcinoma, 2 of 5; adenocarcinoma of lung, 1 of 3; large cell carcinoma, 0 of 1; colon adenocarcinoma 10 of 20). The mechanisms that regulate fibroblast activation and their accumulation in cancer are not fully understood. Cancer-associated fibroblasts might serve as novel therapeutic targets (39). Such therapies can be given alone or in combination with chemotherapy, radiation, or surgery. Fibroblast-directed therapy can be used to ablate or to normalize the cancer-associated fibroblasts. One such target that has already been associated with cancer fibroblasts is fibroblast-activating protein (40). A phase I dose escalation study with sibrotuzumab (anti–fibroblast-activating protein IgG) in patients with advanced colorectal carcinoma or non–small-cell lung carcinoma showed that sibrotuzumab specifically binds to the tumor sites with no apparent side effects (41). Further investigation on the expression of antigen 14C5/αvβ5 and its possible role as a marker for cancer-associated fibroblast is necessary to fully understand the therapeutic and diagnostic potential of mAb 14C5.

No potential conflicts of interest were disclosed.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
De Potter CR, Schelfhout AM, De Smet FH, et al. A monoclonal antibody directed against a human cell membrane antigen prevents cell substrate adhesion and tumor invasion.
Am J Pathol
1994
;
144
:
95
–103.
2
Coene E, Schelfhout AM, De Ridder L, De Potter CR. Generation of a monoclonal antibody directed against a human cell substrate adhesion molecule and the expression of the antigen in human tissues.
Hybridoma
1997
;
16
:
77
–83.
3
Giffels P, Kohler S, De Potter C, et al. MAb 14C5 against a human cell substrate adhesion molecule for inhibition of tumor growth in vivo.
Eur J Cancer
1997
;
33
:
41
–2.
4
Wagner U, Kohler S, Prietl G, et al. Monoclonal anti-idiotype antibodies in immunotherapy of ovarian carcinoma (MAb ACA125) and breast carcinoma (MAb ACA14C5).
Zentralbl Gynakol
1999
;
121
:
190
–5.
5
Lahorte CM, Bacher K, Burvenich I, et al. Radiolabeling, biodistribution, and dosimetry of (123)I-mAb 14C5: a new mAb for radioimmunodetection of tumor growth and metastasis in vivo.
J Nucl Med
2004
;
45
:
1065
–73.
6
Burvenich I, Schoonooghe S, Cornelissen B, et al. In vitro and in vivo targeting properties of iodine-123- or iodine-131-labeled monoclonal antibody 14C5 in a non-small cell lung cancer and colon carcinoma model.
Clin Cancer Res
2005
;
11
:
7288
–96.
7
Burvenich IJ, Schoonooghe S, Blanckaert P, et al. Biodistribution and planar gamma camera imaging of (123)I- and (131)I-labeled F(ab′)(2) and Fab fragments of monoclonal antibody 14C5 in nude mice bearing an A549 lung tumor.
Nucl Med Biol
2007
;
34
:
257
–65.
8
Reinholt FP, Hultenby K, Oldberg A, Heinegard D. Osteopontin—a possible anchor of osteoclasts to bone.
Proc Natl Acad Sci U S A
1990
;
87
:
4473
–5.
9
Miyauchi A, Alvarez J, Greenfield EM, et al. Recognition of osteopontin and related peptides by an αvβ3 integrin stimulates immediate cell signals in osteoclasts.
J Biol Chem
1991
;
266
:
20369
–74.
10
Willems A, Leoen J, Schoonooghe S, Grooten J, Mertens N. Optimizing expression and purification from cell culture medium of trispecific recombinant antibody derivatives.
J Chromatogr B Analyt Technol Biomed Life Sci
2003
;
786
:
161
–76.
11
Mizejewski GJ. Role of integrins in cancer: survey of expression patterns.
Proc Soc Exp Biol Med
1999
;
222
:
124
–38.
12
Wehrle-Haller B, Imhof BA. Integrin-dependent pathologies.
J Pathol
2003
;
200
:
481
–7.
13
Imhof BA, Vadas MA. New therapeutic targets in vascular biology.
Thromb Haemost
2003
;
90
:
557
–9.
14
Jin H, Varner J. Integrins: roles in cancer development and as treatment targets.
Br J Cancer
2004
;
90
:
561
–5.
15
Ruegg C, Dormond O, Mariotti A. Endothelial cell integrins and COX-2: mediators and therapeutic targets of tumor angiogenesis.
Biochim Biophys Acta
2004
;
1654
:
51
–67.
16
Leavesley DI, Ferguson GD, Wayner EA, Cheresh DA. Requirement of the integrin β3 subunit for carcinoma cell spreading or migration on vitronectin and fibrinogen.
J Cell Biol
1992
;
117
:
1101
–7.
17
Kerr JS, Slee AM, Mousa SA. The α v integrin antagonists as novel anticancer agents: an update.
Expert Opin Investig Drugs
2002
;
11
:
1765
–74.
18
Taverna D, Crowley D, Connolly M, Bronson RT, Hynes RO. A direct test of potential roles for β3 and β5 integrins in growth and metastasis of murine mammary carcinomas.
Cancer Res
2005
;
65
:
10324
–9.
19
Friedlander M, Brooks PC, Shaffer RW, Kincaid CM, Varner JA, Cheresh DA. Definition of two angiogenic pathways by distinct α v integrins.
Science
1995
;
270
:
1500
–2.
20
Hynes RO. A reevaluation of integrins as regulators of angiogenesis.
Nat Med
2002
;
8
:
918
–21.
21
Mitjans F, Meyer T, Fittschen C, et al. In vivo therapy of malignant melanoma by means of antagonists of αv integrins.
Int J Cancer
2000
;
87
:
716
–23.
22
Enns A, Korb T, Schluter K, et al. αvβ5-integrins mediate early steps of metastasis formation.
Eur J Cancer
2005
;
41
:
1065
–72.
23
Dechantsreiter MA, Planker E, Matha B, et al. N-methylated cyclic RGD peptides as highly active and selective α(V)β(3) integrin antagonists.
J Med Chem
1999
;
42
:
3033
–40.
24
Li LS, Rader C, Matsushita M, et al. Chemical adaptor immunotherapy: design, synthesis, and evaluation of novel integrin-targeting devices.
J Med Chem
2004
;
47
:
5630
–40.
25
Dayam R, Aiello F, Deng J, et al. Discovery of small molecule integrin αvβ3 antagonists as novel anticancer agents.
J Med Chem
2006
;
49
:
4526
–34.
26
Dijkgraaf I, Kruijtzer JA, Frielink C, et al. αvβ3 integrin-targeting of intraperitoneally growing tumors with a radiolabeled RGD peptide.
Int J Cancer
2007
;
120
:
605
–10.
27
Eskens FA, Dumez H, Hoekstra R, et al. Phase I and pharmacokinetic study of continuous twice weekly intravenous administration of Cilengitide (EMD 121974), a novel inhibitor of the integrins αvβ3 and αvβ5 in patients with advanced solid tumours.
Eur J Cancer
2003
;
39
:
917
–26.
28
Davis H, Prabhakar U, Jang H. A rational approach for a phase I clinical study design to evaluate CNTO 95, a novel, fully human anti-a(v) monoclonal antibody (MAb), in patients with solid tumors.
J Clin Oncol
2004
;
22
:
190S
.
29
Jayson G, Mullamitha S, Ton C. Phase I study of CNTO 95, a fully human monoclonal antibody (mAb) to αv integrins, in patients with solid tumors.
J Clin Oncol
2004
;
22
:
224S
.
30
Trikha M, Zhou Z, Nemeth JA, et al. CNTO 95, a fully human monoclonal antibody that inhibits αv integrins, has antitumor and antiangiogenic activity in vivo.
Int J Cancer
2004
;
110
:
326
–35.
31
Martin PL, Jiao Q, Cornacoff J, et al. Absence of adverse effects in cynomolgus macaques treated with CNTO 95, a fully human anti-αv integrin monoclonal antibody, despite widespread tissue binding.
Clin Cancer Res
2005
;
11
:
6959
–65.
32
Brooks PC, Montgomery AM, Rosenfeld M, et al. Integrin αvβ3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels.
Cell
1994
;
79
:
1157
–64.
33
Gutheil JC, Campbell TN, Pierce PR, et al. Targeted antiangiogenic therapy for cancer using Vitaxin: a humanized monoclonal antibody to the integrin αvβ3.
Clin Cancer Res
2000
;
6
:
3056
–61.
34
Burke PA, DeNardo SJ, Miers LA, Lamborn KR, Matzku S, DeNardo GL. Cilengitide targeting of α(v)β(3) integrin receptor synergizes with radioimmunotherapy to increase efficacy and apoptosis in breast cancer xenografts.
Cancer Res
2002
;
62
:
4263
–72.
35
Lewis MR. Radiolabeled RGD peptides move beyond cancer: PET imaging of delayed-type hypersensitivity reaction.
J Nucl Med
2005
;
46
:
2
–4.
36
Dijkgraaf I, Kruijtzer JA, Liu S, et al. Improved targeting of the α(v)β(3) integrin by multimerisation of RGD peptides.
Eur J Nucl Med Mol Imaging
2007
;
34
:
267
–73.
37
Haubner R. αvβ3-integrin imaging: a new approach to characterise angiogenesis?
Eur J Nucl Med Mol Imaging
2006
;
33
Suppl 1:
54
–63.
38
Zhang X, Xiong Z, Wu Y, et al. Quantitative PET imaging of tumor integrin αvβ3 expression with 18F-FRGD2.
J Nucl Med
2006
;
47
:
113
–21.
39
Kalluri R, Zeisberg M. Fibroblasts in cancer.
Nat Rev Cancer
2006
;
6
:
392
–401.
40
Garin-Chesa P, Old LJ, Rettig WJ. Cell surface glycoprotein of reactive stromal fibroblasts as a potential antibody target in human epithelial cancers.
Proc Natl Acad Sci U S A
1990
;
87
:
7235
–9.
41
Scott AM, Wiseman G, Welt S, et al. A Phase I dose-escalation study of sibrotuzumab in patients with advanced or metastatic fibroblast activation protein-positive cancer.
Clin Cancer Res
2003
;
9
:
1639
–47.

Supplementary data