Internalizing cancer antibodies from phage libraries selected on tumor cells and yeast-displayed tumor antigens.
Journal: 2010/December - Journal of Molecular Biology
ISSN: 1089-8638
Abstract:
A number of approaches have been utilized to generate antibodies to cancer cell surface receptors that can be used as potential therapeutics. A number of these therapeutic approaches, including antibody-drug conjugates, immunotoxins, and targeted nucleic acid delivery, require antibodies that not only bind receptor but also undergo internalization into the cell upon binding. We previously reported on the ability to generate cancer cell binding and internalizing antibodies directly from human phage antibody libraries selected for internalization into cancer cell lines. While a number of useful antibodies have been generated using this approach, limitations include the inability to direct the selections to specific antigens and to identify the antigen bound by the antibodies. Here we show that these limitations can be overcome by using yeast-displayed antigens known to be associated with a cell type to select the phage antibody output after several rounds of selection on a mammalian cell line. We used this approach to generate several human phage antibodies to yeast-displayed EphA2 and CD44. The antibodies bound both yeast-displayed and mammalian cell surface antigens, and were endocytosed upon binding to mammalian cells. This approach is generalizable to many mammalian cell surface proteins, results in the generation of functional internalizing antibodies, and does not require antigen expression and purification for antibody generation.
Relations:
Content
Citations
(14)
References
(52)
Grants
(46)
Chemicals
(5)
Organisms
(2)
Processes
(3)
Anatomy
(1)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
J Mol Biol 404(1): 88-99

Internalizing cancer antibodies from phage libraries selected on tumor cells and yeast displayed tumor antigens

Introduction

Antibodies specific to cell surface receptors overexpressed on a number of cancers have been utilized for development of targeted immunotherapeutics. For example, HER2, CD20, and EGFR are overexpressed on a number of tumors and antibodies recognizing these receptors have been developed to treat metastatic breast cancer (Trastuzamab) 1, lymphoma (Rituximab) 2, and colorectal cancer (Cetuximab) 3. As such, identification of overexpressed cell surface receptors and antibodies which bind them provide a route to the development of cancer therapies, especially for those cancer subtypes with poor prognosis and resistance to traditional therapies.

Extensive studies of cancer transcriptional patterns have led to the discovery of many potential molecular targets to distinguish the malignant from the benign and the most aggressive cancers from those that are less aggressive 4; 5; 6 However, multiple steps are required, from the identification of genes by transcriptional pattern to the development of antibodies that can be used to treat humans. These steps include validation of protein expression, tumor associated antigen (TAA) expression and purification, generation of antibodies specific to the TAA, characterization of antibodies to identify therapeutic candidates, and further modification of antibodies which are therapeutic candidates including humanization. As an alternative, human antibodies binding TAA have been generated by direct selection of phage antibody libraries on live cancer cell lines 7; 8. In a modification of this approach, it has proven possible to select not only for cell binding, but also for binding and internalization into cancer cell lines 9; 10; 11; 12; 13. Such antibodies may prove especially useful for the development of antibody-drug or antibody-toxin conjugates, where the therapeutic molecule must enter the cell for efficacy 14; 15; 16; 17. Identification of the TAA bound by phage antibodies resulting from cell selections is a challenge which requires multiple steps, including the identification of specific phage antibodies, immunoprecipitation of the TAA using the phage antibody, and indentifying the TAA, for example by using mass spectrometry 11; 13; 18. Since this approach requires the isolation of the TAA by immunoprecipitation, it is not successful when the TAA is not abundant or easily immunoprecipiated by the antibody. Moreover, this approach is applied one antibody at a time to identify the antigen bound.

For this work, we sought to develop a general method to direct tumor cell selection of internalizing phage antibodies to a single tumor antigen. In this approach, the phage antibody library is subjected to several rounds of selection for internalization on a cancer cell line to enrich for internalizing phage antibodies binding native cell surface receptors. Next, the polyclonal phage output is subjected to several rounds of selection on yeast displaying a specific TAA thought to be associated with that cancer. Using this approach, we generated several internalizing human single chain Fv (scFv) phage antibodies to two TAA (CD44 and EphA2) known from transcriptional profiling and proteomic analysis to be overexpressed in basal breast cancers 19; 20; 21; 22; 23; 24.

Results

Display of tumor associated antigens on the surface of yeast

Two TAA (CD44 and EphA2) known to be overexpressed in basal breast cancers 19; 20; 21; 22; 23; 24 were selected for display on the surface of Sachromyces cerevisiae. For yeast surface display, the full-length extracellular domain (ECD) of EphA2 (aa 1-510) and the link domain of CD44 (aa 1-149) (domain 1) were cloned into the yeast display vector pYD2 for (C-terminal) fusion to Aga2 25. Vector DNA was used to transform EBY100, and cell surface display was induced. Both extracellular domains of CD44 and EphA2 were well displayed on the yeast surface as quantitated by a monoclonal antibody to a C-terminal epitope tag (Figure 1). Specific binding to yeast displayed EphA2 and CD44 extracellular domains of the EphA2 natural ligand Ephrin A1, and antibodies to EphA2 and CD44 suggests that the domains are not only displayed but displayed in a form that can be specifically recognized by antibodies (Figure 1).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f1.jpg
Display of antigen domains on the surface of yeast

(a) The extracellular domain (ECD) of receptor EphA2 was displayed on yeast surface and recognized by anti-EphA2 antibody and recombinant mouse Ephrin A1 (R&D) as determined by flow cytometry analysis. (b) The link domain of CD44 (domain 1, or D1) was displayed on the yeast surface and recognized by anti-CD44 rabbit monoclonal antibody as determined by flow cytometry analysis. Both anti-EphA2 and anti-CD44 antibodies did not recognize an irrelevant protein displayed on the yeast surface.

Efficient recovery of antigen specific phage antibodies from yeast cell surface displayed antigen

A scFv phage antibody that specifically bound to EphA2 was used to study the ability to select phage antibodies on yeast displayed antigen. Approximately 10 phage particles displaying anti-EphA2 human scFv 2D6 (YZ and JDM, unpublished) were incubated with 10 yeast cells displaying the target antigen EphA2 ECD (Y-EphA2). As a control, an identical number of anti-EphA2 phage antibodies were incubated with 10 yeast cells displaying an irrelevant scFv (Y-CON). The recovery of anti-EphA2 phage antibody from yeast displaying the EphA2 ECD was more than 10 fold higher than the recovery of anti-EphA2 phage antibody from yeast displaying the scFv (Figure 2a). To determine the optimal buffer to elute phage antibodies from the yeast surface, different buffers (PBS, 2-MEA, DTT, triethylamine (TEA) and glycine) were evaluated. Although yeast surface display of TAA results from the disulfide linkage between the Aga2 and Aga1 proteins on the yeast surface, reducing agents, including 2-MEA and DTT, resulted in poorer recovery of viable phage antibodies than spontaneous dissociation of phage by incubation in PBS (Figure 2b). In contrast, elution with high pH TEA buffer or low pH glycine buffer increased the number of viable phage recovered approximately two fold, with low pH glycine being the optimal elution buffer of those studied. This elution buffer was used for subsequent studies.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f2.jpg
Recovery of phage antibodies from yeast displayed antigens

(a) Comparison of the recovery of anti-EphA2 phage antibody 2D6 from yeast displaying EphA2 ECD (Y-EphA2) versus yeast displaying an irrelevant protein (Y-CON). A total of 10 anti-EphA2 phage antibody 2D6 were incubated with each of the yeast displayed antigens. (b) Impact of the elution buffer on the titer of EphA2 phage antibody eluted from the surface of yeast displaying EphA2 ECD. A total of 10 anti-EphA2 phage antibody 2D6 were incubated with each of the yeast displayed antigen proteins prior to elution. (c) The impact of input phage titer on eluted phage titer. The indicated titer of anti-EphA2 phage antibody 2D6 was incubated with yeast displayed EphA2 ECD or an irrelevant yeast displayed protein and the titer of eluted phage determined.

To determine the minimum frequency of a specific antibody within a library that can be enriched and selected, phage displayed anti-EphA2 antibodies were serially diluted from 10 to 10 cfu and then mixed with 10 helper phage VCSM13. Phage mixtures were incubated with 10 yeast cells displaying the EphA2 ECD or with yeast displaying the CD44 domain 1, followed by washing, elution and titration of the recovered phages. With an input of 10 specific phage particles, about 12 phage were recovered from yeast cells displaying EphA2, while an input of at least 10 phage were required before phage were present in the output when selected on CD44 (Figure 2c). The average recovery of phage antibodies was 6.5 × 10 for the specific antigen-antibody pair, and 6.5 × 10 for the mismatched pair (Table 1), respectively. This high recovery ratio for specific compared to non-specific phage suggested that it would be possible to enrich and select phage antibodies on yeast displayed antigen.

Table 1

Recovery of specific phage antibody from yeast displayed antigens

The indicated yeast displayed antigen was incubated with 10 phage and the titer of bound phage determined. Results are expressed as the ratio of output/input phage.

Ag : AbPhAb Output / Input
Y-EphA2 ECD : PhAb-EphA26 × 10−2
Y-CD44 D1 : PhAb-CD447 × 10−2
Y-EphA2 ECD : PhAb-CD443 × 10−5
Y-CD44 D1 : PhAb-EphA21 × 10−6

Y = yeast displayed antigen; PhAb = phage displayed antibody.

Selection of antigen specific phage antibodies on yeast cells displaying tumor antigens

The strategy used to select internalizing phage antibodies to specific tumor antigens is shown in Figure 3. To ensure that phage antibodies bound tumor antigens as presented on the surface of human tumor cells and could be internalized upon antigen binding, we used as a starting phage library the polyclonal phage output after the second round of selection of a non-immune human scFv phage library 26 selected for endocytosis into MDAMB231 tumor cells. Based on the fact that basal subtype breast cancer cells over-express EphA2 6 and CD44 21, the phage output of selection on the basal subtype breast cancer cell line MDAMB231 was selected independently on yeast displayed EphA2 ECD (aa 25-534) and CD44 domain 1 (aa 21-169). To remove phage antibodies that bound to irrelevant proteins on the yeast surface, 2.5 × 10 phage from the second round output of the MDAMB231 selection were incubated first with 1 × 10 yeast cells displaying the scFv 4E17 (Y-CON). Then, the depleted phage library was incubated with 2 × 10 yeast cells displaying the relevant antigen (Y-Ag) (Figure 3a).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f3.jpg
Strategy for selecting internalizing antigen specific phage antibodies

(a) After two rounds of selection for internalization on the basal breast cancer cell line MDAMB231, the pool of phage antibodies was first incubated with irrelevant control yeast to remove any yeast binding antibodies followed by panning on yeast displaying either EphA2 ECD or CD44 domain 1. (b) The binding signal of the polyclonal phage antibody pool to EphA2 ECD or CD44 domain 1 after 2 rounds of panning was measured by using flow cytometry. The irrelevant control yeast was stained with unselected phage antibody library (R0), round 1 (R1) and round 2 (R2) polyclonal phage. (c) Frequency of antigen specific phage antibodies after one and two rounds of selection on yeast displayed antigen. Binding frequency was determined by analyzing 96 randomly picked phage antibodies for binding to yeast displayed antigen by flow cytometry. The induced yeast cells displaying an irrelevant protein, EphA2-ECD and CD44 domain 1 were stained with un-selected phage library (R0), polyclonal phages from R1 and R2 respectively. The un-induced yeast cells (yeast only), irrelevant phage antibody (phage control) and the un-selected phage antibody library were used as control.

For both EphA2 ECD and CD44 domain 1 selections, the number of phage recovered from each yeast cell in the second round of selection increased over 40 fold compared to the first round (Table 2), suggesting enrichment for phage binding yeast displayed tumor antigens. This was verified by using polyclonal phage to stain the yeast displayed antigens. From both the first and second round of selections, polyclonal phage antibodies showed specific binding to the antigen domain that was used for selection, with stronger staining after the second round of selection (Figure 3b). In contrast, prior to selection, the input phage antibody library gave no signal above background on yeast cells displaying either EphA2 or CD44. Binding was specific for the yeast displayed antigen, since the polyclonal phage did not bind yeast cells displaying an irrelevant protein (the scFv 4E17) (Figure 3b). To determine the frequency of binding phage antibodies, 96 individual clones were picked, phage produced, and the phage analyzed for binding to the yeast displayed tumor antigen. After one round and two rounds of selection, 31/96 (32.2%) and 39/96 (40.6%) of the clones from EphA2 selection bound yeast cells displaying EphA2 ECD, and 11/96 (11.7%) and 21/96 (21.9%) of the clones from CD44 selection bound yeast displayed CD44, respectively (Figure 3c).

Table 2

Phage display scFv antibody selection on yeast displayed antigens

Phage input and output ratios during the first and second rounds of selection on yeast displayed antigens.

AntigenRound 1 selectionRound 2 selection

Input (cfu)Output
(cfu)
OutputPhageInput (cfu)Output
(cfu)
OutputPhage


InputYeastInputYeast
EphA2-ECD2.5 × 10113.6 × 1061.3 × 10−50.183.4 × 10112.4 × 1087 × 10−412
CD44-ECD D12.5 × 10119.8 × 1063.9 ×10−50.492.1 × 10124.3 × 1082 × 10−420

Phage antibody characterization

Individual phage antibodies from the second round of selection that bound the yeast displayed tumor antigen were analyzed by PCR fingerprinting and DNA sequencing of the scFv genes. Three unique human scFv antibodies (D2-1A7, D2-1A9 and D2-1B1) were identified which bound to EphA2, and two unique scFv (F2-1A6 and F2-1H9) were identified which bound to CD44 ECD domain 1. Characterization of each of these scFv on yeast displayed EphA2-ECD (Y-EphA2 ECD), CD44 ECD domain 1 (Y-CD44 ECD D1) and full-length ECD (Y-CD44 ECD), and scFv 4E17 (Y-CON) indicated that each scFv was specific for its target antigen (Figure 4a). Each unique phage antibody was also analyzed for its ability to bind the original selecting tumor cell line MDAMB231 by flow cytometry. Each phage antibody strongly stained MDAMB231 cells (Figure 4b). Since the initial selection of the phage antibody library aimed to target cell surface receptors specific to basal subtype breast cancer cells, we studied the binding of the selected mAbs to basal and luminal breast cancer cell lines. Each mAb was relatively specific for basal breast cancer cell lines compared to luminal breast cancer cell lines (Figure 4c).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f4.jpg
Binding specificity of monoclonal phage antibodies from yeast antigen biopanning

(a) Binding of monoclonal phage antibodies to yeast displayed antigen domains as determined by flow cytometry. The induced yeast cells displaying an irrelevant protein (Y-CON), EphA2-ECD (Y-EphA2 ECD), CD44 link domain (Y-CD44 ECD D1) and CD44 full length ECD (Y-CD44 ECD) were stained with monoclonal phage antibodies isolated from Y-EphA2 and Y-CD44 D1 selections. (b) Binding of monoclonal phage antibodies to MDAMB231 cells as determined by flow cytometry, (c) differential binding of monoclonal phage antibodies to breast cancer cell lines including luminal breast cancer cell lines SUM52PE and MCF7, and basal breast cancer cell line MDAMB231.

Binding specificity of phage antibodies

The binding of the identified EphA2 and CD44 antibodies to native antigens was also confirmed by immunoprecipitation of the receptors from cell extracts of MDAMB231 cells followed by Western blotting with murine monoclonal antibodies specific to EphA2 and CD44 (Figure 5a). The specificity of the EphA2 mAbs was further studied. Each anti-EphA2 scFv was evaluated for its ability to compete with the natural ligand, Ephrin A1 for binding to EphA2 on the surface of MDAMB231 cells. Although the IC50 of Ephrin A1 for phage antibody D2-1A7 and D2-1A9 at the given concentration differed by nine fold, 1 µg/ml of Ephrin A1 can fully block the cell binding of both D2-1A7 and D2-1A9 phage antibodies (Figure 5b), indicating that these two antibodies bind epitopes which overlap with Ephrin A1.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f5.jpg
Characterization of scFv antibodies by western-blot and flow cytometry

(a) Anti-EphA2 scFv 2D6 or D2-1A7 and anti-CD44 scFv F2-1A6 were used to immunoprecipitate their target antigen from MDAMB231 cells. Antigen was detected by Western blotting using either anti-EphA2 antibody D7 or anti-CD44 antibody Ab-4. (b) EphA2 antibodies D2-1A7 (□), D2-1A9 (◊), and 2D6 (○) compete with ephrin A1 for binding to MDAMB231 cells. Ability of phage antibodies D2-1A7, D2-1A9, and 2D6 binding to MDAMB231 cells in the presence of increasing concentrations of EphA2 ligand, Ephrin A1 was determined by flow cytometry.

Phage antibodies are internalized by MDAMB231 cells

Since the phage antibodies identified by yeast display antigen biopanning were originally selected for the ability to be endocytosed into MDAMB231 cells (Fig 3a), it was anticipated that they would be internalized efficiently. To determine whether D2-1A7, D2-1A9 and F2-1A6 phage antibodies were endocytosed, the phage antibodies were incubated with MDAMB231 cells at 37°C to allow receptor mediated endocytosis, the surface bound phage removed by stripping with low pH buffer, and the internalized phage stained with anti-fd antibody and visualized using confocal microscopy (Fig 6). D2-1A7 and D2-1A9 anti-EphA2 antibodies gave strong intracellular staining while the control phage gave no staining. F2-1A6 anti-CD44 antibody also gave intracellular staining, most clearly visible in the higher magnification panel E. The staining pattern was more diffuse than the classic endosomal pattern seen for the EphA2 phage antibodies. These results indicate that all three phage antibodies were endocytosed by MDAMB231 cells.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f6.jpg
Phage antibodies specific to EphA2 and CD44 are endocytosed by MDAMB231 cells

Cultured cells were incubated with irrelevant phage (A), anti-EphA2 phage D2-1A7 (B) and D2-1A9 (C), anti-CD44 phage F2-1A6 (D, E) for 3 hr at 37°C followed by glycine buffer wash. Panels D and E are from the same experiment, but E is imaged at higher magnification. Endocytosis was visulaized by detection of intracellular phage with anti-fd antibody, and analyzing by confocal microscopy.

Display of tumor associated antigens on the surface of yeast

Two TAA (CD44 and EphA2) known to be overexpressed in basal breast cancers 19; 20; 21; 22; 23; 24 were selected for display on the surface of Sachromyces cerevisiae. For yeast surface display, the full-length extracellular domain (ECD) of EphA2 (aa 1-510) and the link domain of CD44 (aa 1-149) (domain 1) were cloned into the yeast display vector pYD2 for (C-terminal) fusion to Aga2 25. Vector DNA was used to transform EBY100, and cell surface display was induced. Both extracellular domains of CD44 and EphA2 were well displayed on the yeast surface as quantitated by a monoclonal antibody to a C-terminal epitope tag (Figure 1). Specific binding to yeast displayed EphA2 and CD44 extracellular domains of the EphA2 natural ligand Ephrin A1, and antibodies to EphA2 and CD44 suggests that the domains are not only displayed but displayed in a form that can be specifically recognized by antibodies (Figure 1).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f1.jpg
Display of antigen domains on the surface of yeast

(a) The extracellular domain (ECD) of receptor EphA2 was displayed on yeast surface and recognized by anti-EphA2 antibody and recombinant mouse Ephrin A1 (R&D) as determined by flow cytometry analysis. (b) The link domain of CD44 (domain 1, or D1) was displayed on the yeast surface and recognized by anti-CD44 rabbit monoclonal antibody as determined by flow cytometry analysis. Both anti-EphA2 and anti-CD44 antibodies did not recognize an irrelevant protein displayed on the yeast surface.

Efficient recovery of antigen specific phage antibodies from yeast cell surface displayed antigen

A scFv phage antibody that specifically bound to EphA2 was used to study the ability to select phage antibodies on yeast displayed antigen. Approximately 10 phage particles displaying anti-EphA2 human scFv 2D6 (YZ and JDM, unpublished) were incubated with 10 yeast cells displaying the target antigen EphA2 ECD (Y-EphA2). As a control, an identical number of anti-EphA2 phage antibodies were incubated with 10 yeast cells displaying an irrelevant scFv (Y-CON). The recovery of anti-EphA2 phage antibody from yeast displaying the EphA2 ECD was more than 10 fold higher than the recovery of anti-EphA2 phage antibody from yeast displaying the scFv (Figure 2a). To determine the optimal buffer to elute phage antibodies from the yeast surface, different buffers (PBS, 2-MEA, DTT, triethylamine (TEA) and glycine) were evaluated. Although yeast surface display of TAA results from the disulfide linkage between the Aga2 and Aga1 proteins on the yeast surface, reducing agents, including 2-MEA and DTT, resulted in poorer recovery of viable phage antibodies than spontaneous dissociation of phage by incubation in PBS (Figure 2b). In contrast, elution with high pH TEA buffer or low pH glycine buffer increased the number of viable phage recovered approximately two fold, with low pH glycine being the optimal elution buffer of those studied. This elution buffer was used for subsequent studies.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f2.jpg
Recovery of phage antibodies from yeast displayed antigens

(a) Comparison of the recovery of anti-EphA2 phage antibody 2D6 from yeast displaying EphA2 ECD (Y-EphA2) versus yeast displaying an irrelevant protein (Y-CON). A total of 10 anti-EphA2 phage antibody 2D6 were incubated with each of the yeast displayed antigens. (b) Impact of the elution buffer on the titer of EphA2 phage antibody eluted from the surface of yeast displaying EphA2 ECD. A total of 10 anti-EphA2 phage antibody 2D6 were incubated with each of the yeast displayed antigen proteins prior to elution. (c) The impact of input phage titer on eluted phage titer. The indicated titer of anti-EphA2 phage antibody 2D6 was incubated with yeast displayed EphA2 ECD or an irrelevant yeast displayed protein and the titer of eluted phage determined.

To determine the minimum frequency of a specific antibody within a library that can be enriched and selected, phage displayed anti-EphA2 antibodies were serially diluted from 10 to 10 cfu and then mixed with 10 helper phage VCSM13. Phage mixtures were incubated with 10 yeast cells displaying the EphA2 ECD or with yeast displaying the CD44 domain 1, followed by washing, elution and titration of the recovered phages. With an input of 10 specific phage particles, about 12 phage were recovered from yeast cells displaying EphA2, while an input of at least 10 phage were required before phage were present in the output when selected on CD44 (Figure 2c). The average recovery of phage antibodies was 6.5 × 10 for the specific antigen-antibody pair, and 6.5 × 10 for the mismatched pair (Table 1), respectively. This high recovery ratio for specific compared to non-specific phage suggested that it would be possible to enrich and select phage antibodies on yeast displayed antigen.

Table 1

Recovery of specific phage antibody from yeast displayed antigens

The indicated yeast displayed antigen was incubated with 10 phage and the titer of bound phage determined. Results are expressed as the ratio of output/input phage.

Ag : AbPhAb Output / Input
Y-EphA2 ECD : PhAb-EphA26 × 10−2
Y-CD44 D1 : PhAb-CD447 × 10−2
Y-EphA2 ECD : PhAb-CD443 × 10−5
Y-CD44 D1 : PhAb-EphA21 × 10−6

Y = yeast displayed antigen; PhAb = phage displayed antibody.

Selection of antigen specific phage antibodies on yeast cells displaying tumor antigens

The strategy used to select internalizing phage antibodies to specific tumor antigens is shown in Figure 3. To ensure that phage antibodies bound tumor antigens as presented on the surface of human tumor cells and could be internalized upon antigen binding, we used as a starting phage library the polyclonal phage output after the second round of selection of a non-immune human scFv phage library 26 selected for endocytosis into MDAMB231 tumor cells. Based on the fact that basal subtype breast cancer cells over-express EphA2 6 and CD44 21, the phage output of selection on the basal subtype breast cancer cell line MDAMB231 was selected independently on yeast displayed EphA2 ECD (aa 25-534) and CD44 domain 1 (aa 21-169). To remove phage antibodies that bound to irrelevant proteins on the yeast surface, 2.5 × 10 phage from the second round output of the MDAMB231 selection were incubated first with 1 × 10 yeast cells displaying the scFv 4E17 (Y-CON). Then, the depleted phage library was incubated with 2 × 10 yeast cells displaying the relevant antigen (Y-Ag) (Figure 3a).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f3.jpg
Strategy for selecting internalizing antigen specific phage antibodies

(a) After two rounds of selection for internalization on the basal breast cancer cell line MDAMB231, the pool of phage antibodies was first incubated with irrelevant control yeast to remove any yeast binding antibodies followed by panning on yeast displaying either EphA2 ECD or CD44 domain 1. (b) The binding signal of the polyclonal phage antibody pool to EphA2 ECD or CD44 domain 1 after 2 rounds of panning was measured by using flow cytometry. The irrelevant control yeast was stained with unselected phage antibody library (R0), round 1 (R1) and round 2 (R2) polyclonal phage. (c) Frequency of antigen specific phage antibodies after one and two rounds of selection on yeast displayed antigen. Binding frequency was determined by analyzing 96 randomly picked phage antibodies for binding to yeast displayed antigen by flow cytometry. The induced yeast cells displaying an irrelevant protein, EphA2-ECD and CD44 domain 1 were stained with un-selected phage library (R0), polyclonal phages from R1 and R2 respectively. The un-induced yeast cells (yeast only), irrelevant phage antibody (phage control) and the un-selected phage antibody library were used as control.

For both EphA2 ECD and CD44 domain 1 selections, the number of phage recovered from each yeast cell in the second round of selection increased over 40 fold compared to the first round (Table 2), suggesting enrichment for phage binding yeast displayed tumor antigens. This was verified by using polyclonal phage to stain the yeast displayed antigens. From both the first and second round of selections, polyclonal phage antibodies showed specific binding to the antigen domain that was used for selection, with stronger staining after the second round of selection (Figure 3b). In contrast, prior to selection, the input phage antibody library gave no signal above background on yeast cells displaying either EphA2 or CD44. Binding was specific for the yeast displayed antigen, since the polyclonal phage did not bind yeast cells displaying an irrelevant protein (the scFv 4E17) (Figure 3b). To determine the frequency of binding phage antibodies, 96 individual clones were picked, phage produced, and the phage analyzed for binding to the yeast displayed tumor antigen. After one round and two rounds of selection, 31/96 (32.2%) and 39/96 (40.6%) of the clones from EphA2 selection bound yeast cells displaying EphA2 ECD, and 11/96 (11.7%) and 21/96 (21.9%) of the clones from CD44 selection bound yeast displayed CD44, respectively (Figure 3c).

Table 2

Phage display scFv antibody selection on yeast displayed antigens

Phage input and output ratios during the first and second rounds of selection on yeast displayed antigens.

AntigenRound 1 selectionRound 2 selection

Input (cfu)Output
(cfu)
OutputPhageInput (cfu)Output
(cfu)
OutputPhage


InputYeastInputYeast
EphA2-ECD2.5 × 10113.6 × 1061.3 × 10−50.183.4 × 10112.4 × 1087 × 10−412
CD44-ECD D12.5 × 10119.8 × 1063.9 ×10−50.492.1 × 10124.3 × 1082 × 10−420

Phage antibody characterization

Individual phage antibodies from the second round of selection that bound the yeast displayed tumor antigen were analyzed by PCR fingerprinting and DNA sequencing of the scFv genes. Three unique human scFv antibodies (D2-1A7, D2-1A9 and D2-1B1) were identified which bound to EphA2, and two unique scFv (F2-1A6 and F2-1H9) were identified which bound to CD44 ECD domain 1. Characterization of each of these scFv on yeast displayed EphA2-ECD (Y-EphA2 ECD), CD44 ECD domain 1 (Y-CD44 ECD D1) and full-length ECD (Y-CD44 ECD), and scFv 4E17 (Y-CON) indicated that each scFv was specific for its target antigen (Figure 4a). Each unique phage antibody was also analyzed for its ability to bind the original selecting tumor cell line MDAMB231 by flow cytometry. Each phage antibody strongly stained MDAMB231 cells (Figure 4b). Since the initial selection of the phage antibody library aimed to target cell surface receptors specific to basal subtype breast cancer cells, we studied the binding of the selected mAbs to basal and luminal breast cancer cell lines. Each mAb was relatively specific for basal breast cancer cell lines compared to luminal breast cancer cell lines (Figure 4c).

An external file that holds a picture, illustration, etc.
Object name is nihms244789f4.jpg
Binding specificity of monoclonal phage antibodies from yeast antigen biopanning

(a) Binding of monoclonal phage antibodies to yeast displayed antigen domains as determined by flow cytometry. The induced yeast cells displaying an irrelevant protein (Y-CON), EphA2-ECD (Y-EphA2 ECD), CD44 link domain (Y-CD44 ECD D1) and CD44 full length ECD (Y-CD44 ECD) were stained with monoclonal phage antibodies isolated from Y-EphA2 and Y-CD44 D1 selections. (b) Binding of monoclonal phage antibodies to MDAMB231 cells as determined by flow cytometry, (c) differential binding of monoclonal phage antibodies to breast cancer cell lines including luminal breast cancer cell lines SUM52PE and MCF7, and basal breast cancer cell line MDAMB231.

Binding specificity of phage antibodies

The binding of the identified EphA2 and CD44 antibodies to native antigens was also confirmed by immunoprecipitation of the receptors from cell extracts of MDAMB231 cells followed by Western blotting with murine monoclonal antibodies specific to EphA2 and CD44 (Figure 5a). The specificity of the EphA2 mAbs was further studied. Each anti-EphA2 scFv was evaluated for its ability to compete with the natural ligand, Ephrin A1 for binding to EphA2 on the surface of MDAMB231 cells. Although the IC50 of Ephrin A1 for phage antibody D2-1A7 and D2-1A9 at the given concentration differed by nine fold, 1 µg/ml of Ephrin A1 can fully block the cell binding of both D2-1A7 and D2-1A9 phage antibodies (Figure 5b), indicating that these two antibodies bind epitopes which overlap with Ephrin A1.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f5.jpg
Characterization of scFv antibodies by western-blot and flow cytometry

(a) Anti-EphA2 scFv 2D6 or D2-1A7 and anti-CD44 scFv F2-1A6 were used to immunoprecipitate their target antigen from MDAMB231 cells. Antigen was detected by Western blotting using either anti-EphA2 antibody D7 or anti-CD44 antibody Ab-4. (b) EphA2 antibodies D2-1A7 (□), D2-1A9 (◊), and 2D6 (○) compete with ephrin A1 for binding to MDAMB231 cells. Ability of phage antibodies D2-1A7, D2-1A9, and 2D6 binding to MDAMB231 cells in the presence of increasing concentrations of EphA2 ligand, Ephrin A1 was determined by flow cytometry.

Phage antibodies are internalized by MDAMB231 cells

Since the phage antibodies identified by yeast display antigen biopanning were originally selected for the ability to be endocytosed into MDAMB231 cells (Fig 3a), it was anticipated that they would be internalized efficiently. To determine whether D2-1A7, D2-1A9 and F2-1A6 phage antibodies were endocytosed, the phage antibodies were incubated with MDAMB231 cells at 37°C to allow receptor mediated endocytosis, the surface bound phage removed by stripping with low pH buffer, and the internalized phage stained with anti-fd antibody and visualized using confocal microscopy (Fig 6). D2-1A7 and D2-1A9 anti-EphA2 antibodies gave strong intracellular staining while the control phage gave no staining. F2-1A6 anti-CD44 antibody also gave intracellular staining, most clearly visible in the higher magnification panel E. The staining pattern was more diffuse than the classic endosomal pattern seen for the EphA2 phage antibodies. These results indicate that all three phage antibodies were endocytosed by MDAMB231 cells.

An external file that holds a picture, illustration, etc.
Object name is nihms244789f6.jpg
Phage antibodies specific to EphA2 and CD44 are endocytosed by MDAMB231 cells

Cultured cells were incubated with irrelevant phage (A), anti-EphA2 phage D2-1A7 (B) and D2-1A9 (C), anti-CD44 phage F2-1A6 (D, E) for 3 hr at 37°C followed by glycine buffer wash. Panels D and E are from the same experiment, but E is imaged at higher magnification. Endocytosis was visulaized by detection of intracellular phage with anti-fd antibody, and analyzing by confocal microscopy.

Discussion

Breast cancers have been characterized by expression profile into several subtypes including luminal, basal, and HER2 positive, among which the basal subtype is related to reduced disease free survival, resistance to multiple drugs, and poor prognosis. Although multiple genes have been identified from gene expression arrays that are associated with this aggressive breast cancer subtype, generation of specific antibodies to these targets remains challenging, especially for cell surface receptors. Relatively large proteins must be expressed and purified in their native conformation and used for antibody generation. Traditional antibody generation approaches, such as hybridomas or direct selection from antibody libraries using recombinant protein, may generate antibodies that do not bind the receptor as expressed on cells. As an alternative, we developed an approach where phage antibody libraries can be directly selected on the target cell type not only for binding but for binding in a manner that results in antibody internalization 9; 27. We and others have used this approach to make cell targeting antibodies binding HER2, EGFR, transferrin receptor, integrin α3β1, CD9P-1, CD166, CD146, and other antigens 9; 10; 11; 13; 18; 28; 29. One useful aspect of such antibodies is that they are internalized upon cell binding and thus can be used for targeted drug, toxin or nucleic acid therapeutic approaches 14; 15; 30; 31; 32; 33; 34; 35; 36. Major limitations of this approach include the fact that one cannot direct the selection to specific antigens and identification of the antigen bound by what appear to be useful antibodies can be quite challenging using approaches such as immunoprecipitation and mass spectrometry sequencing 11; 13; 18.

Here we have shown that these limitations can be overcome by using yeast displayed antigens known to be associated with a cell type to select the phage output after several rounds of selection on a mammalian cell line. The antigen bound by the resulting antibodies is thus known and the antibodies bind the mammalian cell surface protein and undergo receptor mediated endocytosis. Expression and purification of the antigen used for selection is not necessary, simplifying antibody generation. We did not attempt to generate antibodies to TAA by directly selecting the primary phage antibody library on yeast displayed TAA. While this approach may prove successful, potential problems include generation of non-internalizing antibodies, generation of antibodies binding yeast displayed but not mammalian TAA, and dominance of antibodies to yeast antigens other then the displayed TAA.

While bacteria have been used to display antigens for phage antibody selection 37, the use of a simple eukaryote for display may result in a greater proportion of antigens being displayed in their proper conformation 38. Besides the two antigens described here, EGFR 39, HER2 domains (YZ and JDM, unpublished), cancer-testis antigen NY-ESO-1 40 and breast cancer-related antigens 41 have been successfully displayed on the surface of yeast. Thus it appears that this approach may be generalizable to many mammalian cell surface proteins, with the caveat that there would be a low probability of generating antibodies to specific post-translational modifications, such as glycosylation, as these are quite different in yeast compared to mammalian cells. Finally, it may be possible to apply this approach in a more parallel manner, selecting for antibodies to multiple antigens at the same time. Bowley et al have proposed that a library of scFv antibodies on yeast can be panned against a library of cDNA displayed on phage to generate cognate pairs 42. Reversing this approach, so that antibodies are displayed on phage and antigens on yeast would have two potential advantages. First, a greater probability of successful antigen display in the eukaryotic yeast compared to bacteria, and second the ability to display antigen as a C-terminal fusion protein, eliminating issues with the stop codon in cDNA. Such yeast cDNA display has been reported by Bidlingmaier as a means to isolate proteins binding phosphtidyl inositols 43. The possibility of such highly parallel approaches remains to be proved. In the meantime, the use of yeast displayed mammalian surface receptors appears to offer a simple means of generating internalizing antibodies to cell surface receptors from phage antibody libraries.

Materials and Methods

Cell lines, media, antibodies and full-length cDNA clones

Breast cancer cell lines MCF7, T47D, MDAMB453, MDAMB231, human mammary epithelial cell (HMEC), and SUM52PE were obtained from the ATCC and Clontech (HMEC), or from collections developed in the laboratories of Dr. Steve Ethier (SUM52PE). The cell lines were cultured using conditions described previously 6. Yeast strain EBY100 was grown in YPD medium (Current Protocols in Molecular Biology, John Wiley and Sons, Chapter 13.1.2). EBY100 transfected with expression vector pYD2 25 was selected on SD-CAA medium (Current Protocols, Chapter 13). The Aga2p antigen fusion was expressed on the yeast surface by induction in SG-CAA medium (identical to SD-CAA medium except the glucose is replaced by galactose) at 20°C for 24~48 hr as described previously 44. Bacteria strain E. coli DH5α and TG1 were used for the preparation of plasmid DNA and the expression of soluble scFv antibodies respectively. SV5 antibody was purified from hybridoma supernatant using Protein G and directly labeled with Alexa-488 or Alexa-647 using a kit provided by the manufacturer (Invitrogen; Carlsbad, CA). Biotin conjugated rabbit anti-fd bacteriaphage was purchased from Sigma and used to detect phage antibodies. Monoclonal antibody D7 against EphA2 ECD was purchased from Upstate Biotech, polyclonal goat anti-EphA2 and recombinant mouse Ephrin A1 with human Fc fusion protein from R&D Systems, anti-CD44 antibody for Western Blotting from NeoMarkers, and monoclonal anti-CD44 recognizing link domain from Abcam. The full-length cDNA of EphA2 and CD44 was obtained from the ATCC.

Antigen and antigen domains displayed on the yeast surface

Primers annealing to antigen cDNA and having a 25-mer overlapping sequence with pYD2/NcoI-NotI-digested vector were designed to amplify antigen domains by PCR using Pfu polymerase. After gel purification, the amplified antigen fragment and NcoI-NotI digested pYD2 vector were used to transform LiAc-treated EBY100 cells by gap repair 45; 46. The transformation mixes were cultured and subcultured in SD-CAA, and induced by culturing in SG-CAA medium for 24–48 hours at 18°C. To validate antigen display, anti-EphA2 (R&D) and recombinant mouse Ephrin A1 (R&D) were analyzed for binding to yeast displayed EphA2 ECD, and anti-CD44 antibody (Abcam) was analyzed for binding to CD44 domain 1 by flow cytometry. Briefly, the induced yeast cells (10 cells) with specific displayed antigen domains were incubated with monoclonal or polyclonal antibodies (1 µg/ml) for 1 h at 4°C, detected using anti-goat PE conjugate for anti-EphA2, anti-human (Fc specific) for rEphrinA1-human Fc fusion protein, and anti-rabbit PE for anti-CD44 respectively, and co-stained with SV5-Alexa-647.

Optimization of elution buffer for phage antibody selection

Different elution buffer including phosphate buffered saline, pH 7.4 (PBS), 40 mM 2-mercaptoethylamine (2-MEA), 1 mM dithiolthreitol (DTT), 100 mM triethylamine (TEA) and 100 mM Glycine/150 mM NaCl/0.1% BSA/0.5% Tween 20 were evaluated for their ability to elute bound phage form the yeast surface. The elution time was 1 hour at 37°C for PBS, 2-MEA and DTT, and 2 minutes at RT for TEA and glycine. After neutralizing with 10 mM cysteine for 2-MEA and DTT elution, and ½ volume of 1M Tris-HCl (pH 7.4) for TEA and glycine elutions, the eluted mixture was used to infect exponentially growing E. coli TG1 cells, and the titer of phage determined by serial dilution and plating on tetracycline resistant media.

Selection of phage antibodies specific to yeast displayed antigen domains

Human mammary epithelial cell (HMEC), luminal breast cancer cell line SUM52PE, T47D, and MDAMB453 were used to deplete the phage library of nonspecific binders by incubating 10 phage particles 47; 48 with 10 cells for 4 h at 4°C. The depleted phage library was then incubated with 5×10 basal breast cancer cell line MDAMB231 cells for 1 h at 4°C, followed by washing with cold PBS and incubation with 37°C-prewarmed medium/10% FBS for 30 min at 37°C to enable the receptor mediated endocytosis of phage particles. The cell surface was stripped by three incubations of five minutes with 4 ml of glycine buffer (150 mM NaCl, 0.1 M glycine, pH 2.5). The cells were then trypsinized, washed with PBS, lysed with 1 ml of 100 mM TEA for four minutes at 4°C and neutralized with 0.5 ml of 1M Tris (pH 7.4). Internalized phage (TEA lysate) was amplified for further selections.

After two rounds of selection on MDAMB231 cells, the polyclonal phage antibodies were used to select phage antibodies specific to yeast displayed antigens EphA2 (Y-EphA2) and CD44 link domain (Y-CD44 D1). The induced yeast cells displaying an irrelevant protein were used to deplete the non-specific binders by incubating 2.5×10 phage particles with 10 yeast cells for 2 h at 4°C. The filtered supernatant containing the depleted phage library was then incubated with 2×10 yeast-cells displaying specific antigen domain for 1 h at 4°C. Yeast cells were washed with cold PBS ten times and pelleted by centrifugation. The bound phage antibodies were eluted by incubating yeast cells with 1 ml of 100 mM Glycine/150 mM NaCl/0.1% BSA/0.5% Tween 20, neutralized with 0.5 ml of 1M Tris-HCl (pH 7.4), and amplified for another round of selection. In the second round of selection, 2×10 yeast cells were used for both antigens, while 2.1×10 phage particles from the first round selection were used for CD44 domain 1 compared to 3.4×10 used for EphA2. Two rounds of selection were performed.

Characterization of phage antibodies

After two rounds of selection, individual phage antibodies were prepared by growing single colonies in 96-well microtiter plates as described 26. Binding of each phage antibody to yeast displayed antigen was determined by incubation of 10 yeast cells with 100 µl phage supernatant diluted in FACS buffer (PBS with 1 mM MgCl2, 0.1 mM CaCl2 and 0.3% BSA) for 2 h at 4°C in conical 96-well microtiter plates, followed by incubation with biotinylated anti-fd antibody and streptavidin-phycoerythrin conjugate (PE) (Jackson), and analyzed using a FACS LSRII (Becton Dickinson). The number of unique phage antibodies was determined by patterns of BstNI digestion of 18 scFv genes amplified by PCR from phage-infected bacteria 49 and confirmed by DNA sequencing.

For binding to breast cancer cells and Ephrin A1 competition experiments, 5×10 MDAMB231 cells were incubated with 10 phage antibodies in the presence of recombinant mouse Ephrin A1 (R&D) at concentration ranging from 0 to 1000 ng/ml for 2 h at 4°C. The bound phage antibodies were detected by incubating cells with biotin conjugated anti-fd antibody (1µg/ml) (Sigma) for 30 min at 4°C and streptavidin-PE (Jackson) followed by flow cytometry analysis.

Immunoprecipitation and Western blot using scFv antibodies

MDAMB231 cell extracts were prepared using 1 ml of lysis buffer per T75 culture flask, containing 0.5 % NP40, 50 mM Tris (pH 7.4), 150 mM NaCl, 2 mM DTT, and protease inhibitor cocktail (Sigma). Soluble scFv antibodies with a (His)6 tag at the C-terminal were generated by subcloning scFv genes from the phage vector into the expression vector pUC119mycHis 50, followed by purification from the periplasmic fraction of E. coli TG1 by IMAC 51using a Ni-NTA column (Qiagen) and gel filtration 52. Cell extracts were incubated with scFv at 26 µg/ml for 2 h at 4°C before the immune complexes were captured on Ni-NTA agarose. The agarose captured immune complexes were then washed 5 times in lysis buffer and heated to 94 °C for 4 min in non-reducing protein loading buffer. Immunoprecipitates were resolved by SDS-PAGE and analyzed by Western Blotting using anti-EphA2 (Upstate) and anti-CD44 (NeoMarkers) antibodies.

Immunofluorescence

MDAMB231 cells were grown on coverslips to 70% of confluence in 12 well-plates and incubated with 10 phage antibodies for three hours at 37°C. The coverslips were washed once with PBS, three times for five minutes with glycine buffer (50 mM glycine (pH 2.5), 150 mM NaCl), neutralized with PEM (80 mM Potassium PIPES (pH 6.8), 5 mM EGTA (pH 7), 2 mM MgCl2), and fixed with PEM containing 4% (W/V) paraformaldehyde for 30 min on ice. Cells were quenched with 0.1 M NH4Cl, permeabilized with 0.5 % Triton X-100, and blocked with 5% non-fat dry milk in TBS-T buffer overnight at 4°C. After blocking endogenous biotin with Avidin-Biotin Kit (Lab Vision), intracellular phages were detected with biotinylated anti-fd polyclonal antibody (Sigma) and streptavidin Texas Red (Amersham). Coverslips were inverted on a slide on mounting medium and microscopic images were taken with a Zeiss LSM 510 laser scanning microscope (Zeiss, Germany).

Cell lines, media, antibodies and full-length cDNA clones

Breast cancer cell lines MCF7, T47D, MDAMB453, MDAMB231, human mammary epithelial cell (HMEC), and SUM52PE were obtained from the ATCC and Clontech (HMEC), or from collections developed in the laboratories of Dr. Steve Ethier (SUM52PE). The cell lines were cultured using conditions described previously 6. Yeast strain EBY100 was grown in YPD medium (Current Protocols in Molecular Biology, John Wiley and Sons, Chapter 13.1.2). EBY100 transfected with expression vector pYD2 25 was selected on SD-CAA medium (Current Protocols, Chapter 13). The Aga2p antigen fusion was expressed on the yeast surface by induction in SG-CAA medium (identical to SD-CAA medium except the glucose is replaced by galactose) at 20°C for 24~48 hr as described previously 44. Bacteria strain E. coli DH5α and TG1 were used for the preparation of plasmid DNA and the expression of soluble scFv antibodies respectively. SV5 antibody was purified from hybridoma supernatant using Protein G and directly labeled with Alexa-488 or Alexa-647 using a kit provided by the manufacturer (Invitrogen; Carlsbad, CA). Biotin conjugated rabbit anti-fd bacteriaphage was purchased from Sigma and used to detect phage antibodies. Monoclonal antibody D7 against EphA2 ECD was purchased from Upstate Biotech, polyclonal goat anti-EphA2 and recombinant mouse Ephrin A1 with human Fc fusion protein from R&D Systems, anti-CD44 antibody for Western Blotting from NeoMarkers, and monoclonal anti-CD44 recognizing link domain from Abcam. The full-length cDNA of EphA2 and CD44 was obtained from the ATCC.

Antigen and antigen domains displayed on the yeast surface

Primers annealing to antigen cDNA and having a 25-mer overlapping sequence with pYD2/NcoI-NotI-digested vector were designed to amplify antigen domains by PCR using Pfu polymerase. After gel purification, the amplified antigen fragment and NcoI-NotI digested pYD2 vector were used to transform LiAc-treated EBY100 cells by gap repair 45; 46. The transformation mixes were cultured and subcultured in SD-CAA, and induced by culturing in SG-CAA medium for 24–48 hours at 18°C. To validate antigen display, anti-EphA2 (R&D) and recombinant mouse Ephrin A1 (R&D) were analyzed for binding to yeast displayed EphA2 ECD, and anti-CD44 antibody (Abcam) was analyzed for binding to CD44 domain 1 by flow cytometry. Briefly, the induced yeast cells (10 cells) with specific displayed antigen domains were incubated with monoclonal or polyclonal antibodies (1 µg/ml) for 1 h at 4°C, detected using anti-goat PE conjugate for anti-EphA2, anti-human (Fc specific) for rEphrinA1-human Fc fusion protein, and anti-rabbit PE for anti-CD44 respectively, and co-stained with SV5-Alexa-647.

Optimization of elution buffer for phage antibody selection

Different elution buffer including phosphate buffered saline, pH 7.4 (PBS), 40 mM 2-mercaptoethylamine (2-MEA), 1 mM dithiolthreitol (DTT), 100 mM triethylamine (TEA) and 100 mM Glycine/150 mM NaCl/0.1% BSA/0.5% Tween 20 were evaluated for their ability to elute bound phage form the yeast surface. The elution time was 1 hour at 37°C for PBS, 2-MEA and DTT, and 2 minutes at RT for TEA and glycine. After neutralizing with 10 mM cysteine for 2-MEA and DTT elution, and ½ volume of 1M Tris-HCl (pH 7.4) for TEA and glycine elutions, the eluted mixture was used to infect exponentially growing E. coli TG1 cells, and the titer of phage determined by serial dilution and plating on tetracycline resistant media.

Selection of phage antibodies specific to yeast displayed antigen domains

Human mammary epithelial cell (HMEC), luminal breast cancer cell line SUM52PE, T47D, and MDAMB453 were used to deplete the phage library of nonspecific binders by incubating 10 phage particles 47; 48 with 10 cells for 4 h at 4°C. The depleted phage library was then incubated with 5×10 basal breast cancer cell line MDAMB231 cells for 1 h at 4°C, followed by washing with cold PBS and incubation with 37°C-prewarmed medium/10% FBS for 30 min at 37°C to enable the receptor mediated endocytosis of phage particles. The cell surface was stripped by three incubations of five minutes with 4 ml of glycine buffer (150 mM NaCl, 0.1 M glycine, pH 2.5). The cells were then trypsinized, washed with PBS, lysed with 1 ml of 100 mM TEA for four minutes at 4°C and neutralized with 0.5 ml of 1M Tris (pH 7.4). Internalized phage (TEA lysate) was amplified for further selections.

After two rounds of selection on MDAMB231 cells, the polyclonal phage antibodies were used to select phage antibodies specific to yeast displayed antigens EphA2 (Y-EphA2) and CD44 link domain (Y-CD44 D1). The induced yeast cells displaying an irrelevant protein were used to deplete the non-specific binders by incubating 2.5×10 phage particles with 10 yeast cells for 2 h at 4°C. The filtered supernatant containing the depleted phage library was then incubated with 2×10 yeast-cells displaying specific antigen domain for 1 h at 4°C. Yeast cells were washed with cold PBS ten times and pelleted by centrifugation. The bound phage antibodies were eluted by incubating yeast cells with 1 ml of 100 mM Glycine/150 mM NaCl/0.1% BSA/0.5% Tween 20, neutralized with 0.5 ml of 1M Tris-HCl (pH 7.4), and amplified for another round of selection. In the second round of selection, 2×10 yeast cells were used for both antigens, while 2.1×10 phage particles from the first round selection were used for CD44 domain 1 compared to 3.4×10 used for EphA2. Two rounds of selection were performed.

Characterization of phage antibodies

After two rounds of selection, individual phage antibodies were prepared by growing single colonies in 96-well microtiter plates as described 26. Binding of each phage antibody to yeast displayed antigen was determined by incubation of 10 yeast cells with 100 µl phage supernatant diluted in FACS buffer (PBS with 1 mM MgCl2, 0.1 mM CaCl2 and 0.3% BSA) for 2 h at 4°C in conical 96-well microtiter plates, followed by incubation with biotinylated anti-fd antibody and streptavidin-phycoerythrin conjugate (PE) (Jackson), and analyzed using a FACS LSRII (Becton Dickinson). The number of unique phage antibodies was determined by patterns of BstNI digestion of 18 scFv genes amplified by PCR from phage-infected bacteria 49 and confirmed by DNA sequencing.

For binding to breast cancer cells and Ephrin A1 competition experiments, 5×10 MDAMB231 cells were incubated with 10 phage antibodies in the presence of recombinant mouse Ephrin A1 (R&D) at concentration ranging from 0 to 1000 ng/ml for 2 h at 4°C. The bound phage antibodies were detected by incubating cells with biotin conjugated anti-fd antibody (1µg/ml) (Sigma) for 30 min at 4°C and streptavidin-PE (Jackson) followed by flow cytometry analysis.

Immunoprecipitation and Western blot using scFv antibodies

MDAMB231 cell extracts were prepared using 1 ml of lysis buffer per T75 culture flask, containing 0.5 % NP40, 50 mM Tris (pH 7.4), 150 mM NaCl, 2 mM DTT, and protease inhibitor cocktail (Sigma). Soluble scFv antibodies with a (His)6 tag at the C-terminal were generated by subcloning scFv genes from the phage vector into the expression vector pUC119mycHis 50, followed by purification from the periplasmic fraction of E. coli TG1 by IMAC 51using a Ni-NTA column (Qiagen) and gel filtration 52. Cell extracts were incubated with scFv at 26 µg/ml for 2 h at 4°C before the immune complexes were captured on Ni-NTA agarose. The agarose captured immune complexes were then washed 5 times in lysis buffer and heated to 94 °C for 4 min in non-reducing protein loading buffer. Immunoprecipitates were resolved by SDS-PAGE and analyzed by Western Blotting using anti-EphA2 (Upstate) and anti-CD44 (NeoMarkers) antibodies.

Immunofluorescence

MDAMB231 cells were grown on coverslips to 70% of confluence in 12 well-plates and incubated with 10 phage antibodies for three hours at 37°C. The coverslips were washed once with PBS, three times for five minutes with glycine buffer (50 mM glycine (pH 2.5), 150 mM NaCl), neutralized with PEM (80 mM Potassium PIPES (pH 6.8), 5 mM EGTA (pH 7), 2 mM MgCl2), and fixed with PEM containing 4% (W/V) paraformaldehyde for 30 min on ice. Cells were quenched with 0.1 M NH4Cl, permeabilized with 0.5 % Triton X-100, and blocked with 5% non-fat dry milk in TBS-T buffer overnight at 4°C. After blocking endogenous biotin with Avidin-Biotin Kit (Lab Vision), intracellular phages were detected with biotinylated anti-fd polyclonal antibody (Sigma) and streptavidin Texas Red (Amersham). Coverslips were inverted on a slide on mounting medium and microscopic images were taken with a Zeiss LSM 510 laser scanning microscope (Zeiss, Germany).

Acknowledgements

This work was supported by National Cancer Institute Specialized Programs of Research Excellence (SPORE) in Breast Cancer (P50-CA58207).

Department of Anesthesia and Pharmaceutical Chemistry, University of California, San Francisco Rm 3C-38, San Francisco General Hospital, 1001 Potrero Ave, San Francisco, CA 94110
Corresponding author: Department of Anesthesia and Pharmaceutical Chemistry, University of California, San Francisco Rm 3C-38, San Francisco General Hospital, 1001 Potrero Ave, San Francisco, CA 94110 USA, tel: 415-206-3256, FAX: 415-206-3253, ude.fscu.aisehtsena@jskram

Abstract

A number of approaches have been utilized to generate antibodies to cancer cell surface receptors which can be used as potential therapeutics. A number of these therapeutic approaches, including antibody-drug conjugates, immunotoxins, and targeted nucleic acid delivery, require antibodies that not only bind receptor, but that also undergo internalization into the cell upon binding. We previously reported the ability to generate cancer cell binding and internalizing antibodies directly from human phage antibody libraries selected for internalization into cancer cell lines. While a number of useful antibodies have been generated using this approach, limitations include the inability to direct the selections to specific antigens and identifying the antigen bound by the antibodies. Here we show that these limitation can be overcome by using yeast displayed antigens known to be associated with a cell type to select the phage antibody output after several rounds of selection on a mammalian cell line. We used this approach to generate several human phage antibodies to yeast displayed EphA2 and CD44. The antibodies bound both yeast displayed and mammalian cell surface antigen and were endocytosed upon binding to mammalian cells. This approach is generalizable to many mammalian cell surface proteins, results in the generation of functional internalizing antibodies, and does not require antigen expression and purification for antibody generation.

Keywords: asaloid breast cancer, single chain Fv antibody, phage display, yeast display
Abstract

Abbreviations used

TAAtumor associated antigen
EGFRepidermal growth factor receptor
ECDextracellular domain
FACSfluorescent activated cell sorting
IMACimmobilized metal affinity chromatography
IgGimmunoglobulin G
ILsimmunoliposomes
IPTGIsopropyl-β-D-thiogalactopyranosid
KDdissociation equilibrium constant
2-MEA2-Mercaptoethylamine
DTTDithiothreitol
TEAtriethylamine
PBSphosphate buffered saline
TBS-TTris-buffered saline Tween-20
Ni-NTANickel-nitrilotriacetic acid
PEphycoerythrin
scFvsingle chain Fv
mAbmonoclonal antibody
HMEChuman mammary epithelial cell
Abbreviations used

Footnotes

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Footnotes

References

  • 1. Baselga J, Tripathy D, Mendelsohn J, Baughman S, Benz CC, Dantis L, Sklarin NT, Seidman AD, Hudis CA, Moore J, Rosen PP, Twaddell T, Henderson IC, Norton LPhase II study of weekly intravenous recombinant humanized anti-p185HER2 monoclonal antibody in patients with HER2/neu-overexpressing metastatic breast cancer. J Clin Oncol. 1996;14:737–744.[PubMed][Google Scholar]
  • 2. Hainsworth JD, Burris HA, 3rd, Morrissey LH, Litchy S, Scullin DC, Jr, Bearden JD, 3rd, Richards P, Greco FARituximab monoclonal antibody as initial systemic therapy for patients with low-grade non-Hodgkin lymphoma. Blood. 2000;95:3052–3056.[PubMed][Google Scholar]
  • 3. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, Santoro A, Bets D, Mueser M, Harstrick A, Verslype C, Chau I, Van Cutsem ECetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med. 2004;351:337–345.[PubMed][Google Scholar]
  • 4. Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, Pollack JR, Ross DT, Johnsen H, Akslen LA, Fluge O, Pergamenschikov A, Williams C, Zhu SX, Lonning PE, Borresen-Dale AL, Brown PO, Botstein DMolecular portraits of human breast tumours. Nature. 2000;406:747–752.[PubMed][Google Scholar]
  • 5. Sorlie T, Perou CM, Tibshirani R, Aas T, Geisler S, Johnsen H, Hastie T, Eisen MB, van de Rijn M, Jeffrey SS, Thorsen T, Quist H, Matese JC, Brown PO, Botstein D, Eystein Lonning P, Borresen-Dale ALGene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A. 2001;98:10869–10874.[Google Scholar]
  • 6. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, Clark L, Bayani N, Coppe JP, Tong F, Speed T, Spellman PT, DeVries S, Lapuk A, Wang NJ, Kuo WL, Stilwell JL, Pinkel D, Albertson DG, Waldman FM, McCormick F, Dickson RB, Johnson MD, Lippman M, Ethier S, Gazdar A, Gray JWA collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer Cell. 2006;10:515–527.[Google Scholar]
  • 7. Cai X, Garen AAnti-melanoma antibodies from melanoma patients immunized with genetically modified autologous tumor cells: selection of specific antibodies from single-chain Fv fusion phage libraries. Proc Natl Acad Sci U S A. 1995;92:6537–6541.[Google Scholar]
  • 8. de Kruif J, Terstappen L, Boel E, Logtenberg TRapid selection of cell subpopulation-specific human monoclonal antibodies from a synthetic phage antibody library. Proc Natl Acad Sci U S A. 1995;92:3938–3942.[Google Scholar]
  • 9. Poul MA, Becerril B, Nielsen UB, Morisson P, Marks JDSelection of tumor-specific internalizing human antibodies from phage libraries. J Mol Biol. 2000;301:1149–1161.[PubMed][Google Scholar]
  • 10. Heitner T, Moor A, Garrison JL, Marks C, Hasan T, Marks JDSelection of cell binding and internalizing epidermal growth factor receptor antibodies from a phage display library. J Immunol Methods. 2001;248:17–30.[PubMed][Google Scholar]
  • 11. Gao C, Mao S, Ronca F, Zhuang S, Quaranta V, Wirsching P, Janda KDDe novo identification of tumor-specific internalizing human antibody-receptor pairs by phage-display methods. J Immunol Methods. 2003;274:185–197.[PubMed][Google Scholar]
  • 12. Liu B, Conrad F, Cooperberg MR, Kirpotin DB, Marks JDMapping tumor epitope space by direct selection of single-chain Fv antibody libraries on prostate cancer cells. Cancer Res. 2004;64:704–710.[PubMed][Google Scholar]
  • 13. Goenaga AL, Zhou Y, Legay C, Bougherara H, Huang L, Liu B, Drummond DC, Kirpotin DB, Auclair C, Marks JD, Poul MAIdentification and characterization of tumor antigens by using antibody phage display and intrabody strategies. Mol Immunol. 2007;44:3777–3788.[Google Scholar]
  • 14. Sapra P, Allen TMInternalizing antibodies are necessary for improved therapeutic efficacy of antibody-targeted liposomal drugs. Cancer Res. 2002;62:7190–7194.[PubMed][Google Scholar]
  • 15. Nielsen UB, Kirpotin DB, Pickering EM, Hong K, Park JW, Refaat Shalaby M, Shao Y, Benz CC, Marks JDTherapeutic efficacy of anti-ErbB2 immunoliposomes targeted by a phage antibody selected for cellular endocytosis. Biochim Biophys Acta. 2002;1591:109–118.[PubMed][Google Scholar]
  • 16. Pastan I, Hassan R, Fitzgerald DJ, Kreitman RJImmunotoxin therapy of cancer. Nat Rev Cancer. 2006;6:559–565.[PubMed][Google Scholar]
  • 17. Lewis Phillips GD, Li G, Dugger DL, Crocker LM, Parsons KL, Mai E, Blattler WA, Lambert JM, Chari RV, Lutz RJ, Wong WL, Jacobson FS, Koeppen H, Schwall RH, Kenkare-Mitra SR, Spencer SD, Sliwkowski MXTargeting HER2-positive breast cancer with trastuzumab-DM1, an antibody-cytotoxic drug conjugate. Cancer Res. 2008;68:9280–9290.[PubMed][Google Scholar]
  • 18. Liu B, Conrad F, Roth A, Drummond DC, Simko JP, Marks JDRecombinant full-length human IgG1s targeting hormone-refractory prostate cancer. J Mol Med. 2007;85:1113–1123.[PubMed][Google Scholar]
  • 19. Matsumura Y, Tarin DSignificance of CD44 gene products for cancer diagnosis and disease evaluation. Lancet. 1992;340:1053–1058.[PubMed][Google Scholar]
  • 20. Marhaba R, Zoller MCD44 in cancer progression: adhesion, migration and growth regulation. J Mol Histol. 2004;35:211–231.[PubMed][Google Scholar]
  • 21. Hamilton SR, Fard SF, Paiwand FF, Tolg C, Veiseh M, Wang C, McCarthy JB, Bissell MJ, Koropatnick J, Turley EAThe hyaluronan receptors CD44 and Rhamm (CD168) form complexes with ERK1,2 that sustain high basal motility in breast cancer cells. J Biol Chem. 2007;282:16667–16680.[Google Scholar]
  • 22. Andres AC, Reid HH, Zurcher G, Blaschke RJ, Albrecht D, Ziemiecki AExpression of two novel eph-related receptor protein tyrosine kinases in mammary gland development and carcinogenesis. Oncogene. 1994;9:1461–1467.[PubMed][Google Scholar]
  • 23. Zelinski DP, Zantek ND, Stewart JC, Irizarry AR, Kinch MSEphA2 overexpression causes tumorigenesis of mammary epithelial cells. Cancer Res. 2001;61:2301–2306.[PubMed][Google Scholar]
  • 24. Macrae M, Neve RM, Rodriguez-Viciana P, Haqq C, Yeh J, Chen C, Gray JW, McCormick FA conditional feedback loop regulates Ras activity through EphA2. Cancer Cell. 2005;8:111–118.[PubMed][Google Scholar]
  • 25. Razai A, Garcia-Rodriguez C, Lou J, Geren IN, Forsyth CM, Robles Y, Tsai R, Smith TJ, Smith LA, Siegel RW, Feldhaus M, Marks JDMolecular evolution of antibody affinity for sensitive detection of botulinum neurotoxin type A. J Mol Biol. 2005;351:158–169.[PubMed][Google Scholar]
  • 26. O’Connell D, Becerril B, Roy-Burman A, Daws M, Marks JDPhage versus phagemid libraries for generation of human monoclonal antibodies. J Mol Biol. 2002;321:49–56.[PubMed][Google Scholar]
  • 27. Becerril B, Poul MA, Marks JDToward selection of internalizing antibodies from phage libraries. Biochem Biophys Res Commun. 1999;255:386–393.[PubMed][Google Scholar]
  • 28. An F, Drummond DC, Wilson S, Kirpotin DB, Nishimura SL, Broaddus VC, Liu BTargeted drug delivery to mesothelioma cells using functionally selected internalizing human single-chain antibodies. Mol Cancer Ther. 2008;7:569–578.[Google Scholar]
  • 29. Bidlingmaier S, He J, Wang Y, An F, Feng J, Barbone D, Gao D, Franc B, Broaddus VC, Liu BIdentification of MCAM/CD146 as the target antigen of a human monoclonal antibody that recognizes both epithelioid and sarcomatoid types of mesothelioma. Cancer Res. 2009;69:1570–1577.[Google Scholar]
  • 30. Park JW, Hong K, Carter P, Asgari H, Guo LY, Keller GA, Wirth C, Shalaby R, Kotts C, Wood WI, et al Development of anti-p185HER2 immunoliposomes for cancer therapy. Proc Natl Acad Sci U S A. 1995;92:1327–1331.[Google Scholar]
  • 31. Beers R, Chowdhury P, Bigner D, Pastan IImmunotoxins with increased activity against epidermal growth factor receptor vIII-expressing cells produced by antibody phage display. Clin Cancer Res. 2000;6:2835–2843.[PubMed][Google Scholar]
  • 32. Mamot C, Drummond DC, Greiser U, Hong K, Kirpotin DB, Marks JD, Park JWEpidermal growth factor receptor (EGFR)-targeted immunoliposomes mediate specific and efficient drug delivery to EGFR- and EGFRvIII-overexpressing tumor cells. Cancer Res. 2003;63:3154–3161.[PubMed][Google Scholar]
  • 33. Eliaz RE, Nir S, Szoka FC., Jr Interactions of hyaluronan-targeted liposomes with cultured cells: modeling of binding and endocytosis. Methods Enzymol. 2004;387:16–33.[PubMed]
  • 34. Hayes ME, Drummond DC, Hong K, Park JW, Marks JD, Kirpotin DBAssembly of nucleic acid-lipid nanoparticles from aqueous-organic monophases. Biochim Biophys Acta. 2006;1758:429–442.[PubMed][Google Scholar]
  • 35. Zhou Y, Drummond DC, Zou H, Hayes ME, Adams GP, Kirpotin DB, Marks JDImpact of single-chain Fv antibody fragment affinity on nanoparticle targeting of epidermal growth factor receptor-expressing tumor cells. J Mol Biol. 2007;371:934–947.[Google Scholar]
  • 36. Roth A, Drummond DC, Conrad F, Hayes ME, Kirpotin DB, Benz CC, Marks JD, Liu BAnti-CD166 single chain antibody-mediated intracellular delivery of liposomal drugs to prostate cancer cells. Mol Cancer Ther. 2007;6:2737–2746.[PubMed][Google Scholar]
  • 37. Bradbury A, Persic L, Werge T, Cattaneo AUse of living columns to select specific phage antibodies. Biotechnology (N Y) 1993;11:1565–1569.[PubMed][Google Scholar]
  • 38. Boder ET, Wittrup KDYeast surface display for screening combinatorial polypeptide libraries. Nat Biotechnol. 1997;15:553–557.[PubMed][Google Scholar]
  • 39. Cochran JR, Kim YS, Olsen MJ, Bhandari R, Wittrup KDDomain-level antibody epitope mapping through yeast surface display of epidermal growth factor receptor fragments. J Immunol Methods. 2004;287:147–158.[PubMed][Google Scholar]
  • 40. Piatesi A, Howland SW, Rakestraw JA, Renner C, Robson N, Cebon J, Maraskovsky E, Ritter G, Old L, Wittrup KDDirected evolution for improved secretion of cancer-testis antigen NY-ESO-1 from yeast. Protein Expr Purif. 2006;48:232–242.[PubMed][Google Scholar]
  • 41. Wadle A, Mischo A, Imig J, Wullner B, Hensel D, Watzig K, Neumann F, Kubuschok B, Schmidt W, Old LJ, Pfreundschuh M, Renner CSerological identification of breast cancer-related antigens from a Saccharomyces cerevisiae surface display library. Int J Cancer. 2005;117:104–113.[PubMed][Google Scholar]
  • 42. Bowley DR, Jones TM, Burton DR, Lerner RALibraries against libraries for combinatorial selection of replicating antigen-antibody pairs. Proc Natl Acad Sci U S A. 2009;106:1380–1385.[Google Scholar]
  • 43. Bidlingmaier S, Liu BConstruction and application of a yeast surface-displayed human cDNA library to identify post-translational modification-dependent protein-protein interactions. Mol Cell Proteomics. 2006;5:533–540.[PubMed][Google Scholar]
  • 44. Feldhaus MJ, Siegel RW, Opresko LK, Coleman JR, Feldhaus JM, Yeung YA, Cochran JR, Heinzelman P, Colby D, Swers J, Graff C, Wiley HS, Wittrup KDFlow-cytometric isolation of human antibodies from a nonimmune Saccharomyces cerevisiae surface display library. Nat Biotechnol. 2003;21:163–170.[PubMed][Google Scholar]
  • 45. Gietz RD, Schiestl RHApplications of high efficiency lithium acetate transformation of intact yeast cells using single-stranded nucleic acids as carrier. Yeast. 1991;7:253–263.[PubMed][Google Scholar]
  • 46. Orr-Weaver TL, Szostak JWYeast recombination: the association between double-strand gap repair and crossing-over. Proc Natl Acad Sci U S A. 1983;80:4417–4421.[Google Scholar]
  • 47. Sheets MD, Amersdorfer P, Finnern R, Sargent P, Lindquist E, Schier R, Hemingsen G, Wong C, Gerhart JC, Marks JDEfficient construction of a large nonimmune phage antibody library: the production of high-affinity human single-chain antibodies to protein antigens. Proc Natl Acad Sci U S A. 1998;95:6157–6162.[Google Scholar]
  • 48. Huie MA, Cheung MC, Muench MO, Becerril B, Kan YW, Marks JDAntibodies to human fetal erythroid cells from a nonimmune phage antibody library. Proc Natl Acad Sci U S A. 2001;98:2682–2687.[Google Scholar]
  • 49. Marks JD, Hoogenboom HR, Bonnert TP, McCafferty J, Griffiths AD, Winter G. By-passing immunization. Human antibodies from V-gene libraries displayed on phage. J Mol Biol. 1991;222:581–597.[PubMed]
  • 50. Schier R, Marks JD, Wolf EJ, Apell G, Wong C, McCartney JE, Bookman MA, Huston JS, Houston LL, Weiner LM, et al In vitro and in vivo characterization of a human anti-c-erbB-2 single-chain Fv isolated from a filamentous phage antibody library. Immunotechnology. 1995;1:73–81.[PubMed][Google Scholar]
  • 51. Hochuli ELarge-scale chromatography of recombinant proteins. J Chromatogr. 1988;444:293–302.[PubMed][Google Scholar]
  • 52. Schier R, Bye J, Apell G, McCall A, Adams GP, Malmqvist M, Weiner LM, Marks JDIsolation of high-affinity monomeric human anti-c-erbB-2 single chain Fv using affinity-driven selection. J Mol Biol. 1996;255:28–43.[PubMed][Google Scholar]
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.