Development of Monoclonal Antibodies and Characterization of an ELISA Platform Against Kallikrein-Related Peptidase 6 as a Tumor Biomarker

Materials

The expression vector pET28a (5369bp) was acquired from Novagen (Darmstadt, Germany), and Escherichia coli BL21 was acquired from Takara (Dalian, China). Primers were synthesized Genewiz (Suzhou, China), and DMEM medium and fetal bovine serum were purchased from Gibco (Guangzhou, China). KLK3, KLK6, KLK10, KLK11 proteins, and the Rapid Mouse Isotyping Kit-Gold Series, LFM-ISO, were purchased from RayBiotech (Guangzhou, China). BCA Protein Assay Kit was purchased from Thermo Scientific (Waltham, MA). HiTrap Protein A HP was purchased from GE Healthcare (Atlanta, GA). All other chemicals and reagents used throughout the experiments were from Sigma-Aldrich (St. Louis, MO).

Animal experiments

All animal work was performed according to relevant national and international guidelines. The experiments were approved by the Institutional Animal Care and Use Committee at South China Agricultural University (certification no. CNAS BL0011).

Cloning of KLK6 gene

The cloning procedure was carried out according to standard techniques.⁽¹⁹⁾ A KLK6 PCR product was generated using forward primer (5′-AAACATATG GAGGAGCAGAATAAGTTG-3′) and reverse primer (5′-AAACTCGAGTCA CTTGGCCTAATGGTTT-3′), which included the desired fragment of KLK6 protein and Nde I and Xho I excision sites to allow insertion in the pET28a expression vector. The construct pET28a/KLK6 was verified by PCR with specific primers (5′-AAACATATGGAGGAGGAGAATAAGTTG, 5′-AAACTCGAGTCACTTGGCCTGAATGGTTTT) and DNA sequenced to ensure 100% identity with the expected DNA sequence, in-frame orientation of KLK6, and inclusion of the histidine tag (His6).

Protein expression

Two colonies were selected for small-scale expression analysis. After induction with 0.8 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 4 h at 37°C, the bacterial lysates were sonicated, separated by high speed centrifugation (20,000 g, 20 min at 4°C), and analyzed by SDS-PAGE (12% polyacrylamide gel).⁽²⁰⁾ The recombinant construct pET28a/KLK6 was transformed into E. coli (DE3), the transformants were inoculated into 0.5 L Luria-Bertani (LB) broth medium supplemented with 50 μg/mL kanamycin, and then grown at 37°C for large-scale protein expression. The culture was induced by the addition of 0.8 mM IPTG until the A₆₀₀ reached 0.6. The cells were harvested after incubation at 37°C for 4 h; pellets were washed twice with cold PBS, and then kept frozen at −80°C until use.

Protein purification

Bacterial cell pellets were resuspended (2.5 mL/g wet weight of pellet) in lysis buffer containing lysozyme and DNase. After 20 min on ice, the pellet lysate was subjected to sonication to break cells and centrifuged at 20,000 g for 30 min at 4°C to remove cell debris. The precipitate was successively dissolved in 4 and 8 M urea with stirring at room temperature for 30 min and centrifuged at 20,000 g for 10 min at 4°C. The soluble fraction was loaded to a His-Bind Nickel column (GE) to purify the protein according to the manufacturer's recommendations. The eluted fractions were pooled and desalted immediately through dialysis in desalting buffer (20 mM Tris-HCl [pH 8.0], 0.5 M NaCl). The purity of the KLK6 protein was determined by SDS-PAGE.

Preparation of anti-KLK6 monoclonal antibodies

The immunization protocol consisted of small immunogen doses occurring three times over 6 weeks. Female BALB/c mice (6–8 weeks old; Guangdong Medical Laboratory Animal Center, China) were subcutaneously injected with 100 μg KLK6 protein emulsified with equal volume of Freund's complete adjuvant for initial immunization. The mice received the same injections with 50 μg antigen mixed with incomplete Freund's adjuvant at intervals of 2 weeks following primary immunization. Immunity was assessed by indirect ELISA after the third immunization. Finally, a booster immunization was given intraperitoneally with 50 μg antigen at 3 days before cell fusion, when titers of anti-serum were highest.

Immunized mouse splenocytes and myeloma sp2/0 cells (Center of Laboratorial Animals in South China Agricultural University, Guangzhou, China) were washed twice with serum-free DMEM medium, mixed at a ratio of 5:1 in a 50 mL centrifuge tube, and then had a slow addition of 1 mL 50% polyethylene glycol (PEG). Cells were incubated for 1 min, at which point 30 mL DMEM medium were added gently to quench PEG activity. Subsequently, the cells were centrifuged and resuspended in HAT medium containing 20% serum and then added to 96-well plates. Three days after fusion, half of culture supernatant was replaced with HAT medium; 7 days later, the entire supernatant was switched to HT medium containing 20% serum. After 3 days in HT medium, the culture supernatants were tested for activity in an indirect ELISA. Positive hybridoma cells were subcloned by limiting dilution. The resulting clones were validated for KLK6 specificity, expanded in culture, and stored in liquid nitrogen until needed.

Subtype identification of antibody

Antibody subclasses were screened via the Rapid Mouse Isotyping Kit-Gold Series (LFM-ISO, RayBiotech). Fifty μL hybridoma culture supernatant was added onto the sample pad where indicated or the strip was physically dipped into the culture supernatant. After 10 min, the strip was removed from the fluid and placed on a flat and clean surface to allow development of the red bands to indicate the respective Ig isotypes, as described in the manufacturer's manual.

Preparation of ascites and purification of MAb

Preparation of MAb was done by the ascites method, with the hybridoma injected into the peritoneum of mice. Briefly, female BALB/c mice (8–10 weeks old) were intraperitoneally treated with 0.5 mL saxoline, then 7 days later implanted with 1 × 10⁶ hybridoma cells. Ascites fluid was then collected by syringe and centrifuged to remove any impurities. At 4°C, ammonium sulfate (50% saturation) was added slowly with constant stirring. Solution was incubated at 4°C, stirred slowly for 30 min, and then allowed to stand at 4°C for 1 h. Sample was centrifuged again to remove supernatants and resuspended with 50 mL PBS. Purification was done by HiTrap Protein A HP affinity columns (GE Healthcare), according to the manufacturer's instructions.

Western blot analysis

Western blot analysis was performed as previously described⁽²¹⁾ using the Li-Cor Odyssey infrared imaging system (Lincoln, NE). 0.1 μg of KLK6 protein or 40 μg of cell and tissue lysates were mixed with 5x sample loading buffer, heated at 100°C for 10 min, and separated by 12% SDS-PAGE.⁽²⁰⁾ Separated proteins in the gels were electrophoretically transferred onto nitrocellulose membranes at 200 mA for 2 h and incubated in blocking buffer (Odyssey) for 1 h. The membranes were then incubated with the anti-KLK6 antibodies at 1:1000 overnight at 4°C in PBS-T (PBS containing 0.05% Tween-20 [pH 7.4]). Membranes were then washed in PBS-T washing buffer three times. IRDye 800CW Donkey anti-mouse IgG (Li-Cor) was used as a secondary antibody at 1:15,000 in PBS-T and incubated for 1 h. Li-Cor Odyssey Imaging System was then used for detection of targets according to the manufacturer's instructions.

Sandwich ELISA

Two antibodies with distinct antigen binding sites were then identified. Polystyrene EIA/RIA flat bottom plates were coated overnight at 4°C with 100 ng of anti-KLK6 MAb. Plates were washed with PBS-T washing buffer, then blocked with PBS-T plus 5% skim milk (blocking solution) at 37°C for 2 h. Plates were washed and incubated for 1 h with sample or KLK6 recombinant protein diluted in the blocking solution. Following washing, biotin-MAb was added and incubated for 1 h, washed, then incubated with conjugated HRP-Streptavidin for 1 h. Plates were washed again and TMB substrate solution was added. The reaction was stopped 10 min later with 2 M sulfuric acid (100 μL/well, 0.5 mol/L). Absorbance at 450 nm was measured in a microplate reader (Biotek).

Construction and expression of KLK6 vector

The constructed expression vector pET28a/KLK6 is depicted in Figure 1A, and contains a His-tag and the indicated genomic sequence. The KLK sequence we chose to use was Glu17-Lys244 based on the Uniprot Accession #Q92876. This sequence was confirmed by single colony PCR (Fig. 1B) and DNA sequencing. DNA sequences of KLK6 cloned and PET28a vector were digested and connected with Nde I and Xho I simultaneously. Finally, this gene was cloned into a PET28a vector containing a His-tag and expressed in E. coli.

OPEN IN VIEWER

FIG. 1.

Construction and confirmation of the recombinant plasmid pET28a/KLK6. (A) Diagram of pET28a (+) / KLK6 expression vector. KLK6 sequence is inserted in the dark black region, with DNA sequence displayed in top box. (B) Identification of recombinant plasmid pET28a/KLK6 by PCR. Lane M, DNA marker; lanes 1–3, positive clone; lane 4, positive control.

The validated and E. coli inserted constructs were then subjected to a pilot expression test to determine both the level and purity of the KLK6 protein. As shown in Figure 2A, KLK6 protein was indeed expressed at reasonable levels in E. coli, and was expressed predominantly in the insoluble fraction (pellet, lanes 3 and 4), but not in the soluble fraction (culture supernatant, lanes 1 and 2). Further purification of the insoluble fraction via affinity chromatography resulted in recombinant KLK6 purity reaching at least 95%, as assayed by SDS-PAGE and Image J software analysis (Fig. 2B).

OPEN IN VIEWER

FIG. 2.

Expression and purification of KLK6 recombinant protein. (A) SDS-PAGE gel and stained with Coomassie brilliant blue. Lane M, molecular mass marker; lanes 1, 2, soluble fractions; lanes 3, 4, insoluble fractions. (B) SDS–PAGE analysis of fusion protein after purification. The observed band size is 27 kDa, which is consistent with the predicted band size.

Preparation of antibody

With the purified protein in hand, we set out to immunize mice in order to create monoclonal B cell populations. Mice were immunized, and antigen boosted for four rounds to allow the development and expansion of B cell populations against KLK6. Two mice were selected and tail bled to determine the KLK6 antibody titer by indirect ELISA. ELISA testing determined the titer of the mice to be 40,000 and 80,000, respectively (Fig. 3). The mouse with the highest KLK6 titer was then selected for B cell purification and fusion techniques to generate monoclonal hybridomas. Following B cell hybridoma fusion and three rounds of limiting dilutions to isolate out individual clones, five hybridomas, clones 1D8, 2E3, 3B7, 5G8, and 5H12 respectively, were confirmed to be stable secreters of anti-KLK6 antibodies.

OPEN IN VIEWER

FIG. 3.

Antisera titers of immunized mice as measured by indirect ELISA. Serum serially diluted began at 1:5000 and added in recombinant KLK6 protein-coated 96-well ELISA plates. Number indicates the titer of immunized mice. Serum from pre-immune mice served as negative control.

Isotype characterization

A commercial rapid mouse isotyping antibody array kit (RayBiotech) was used to determine the antibody isotypes of the generated antibodies. Cell culture supernatant from each hybridoma clone was added onto the sample pad of strip 2 (green, IgG (IgG1 /IgG2a/IgG2b/IgG3) isotyping strip) and allowed to incubate until colored bands appear, which indicate the corresponding isotype. The IgG isotype strips indicated that the isotype of MAb 2E3 was IgG2b, while the other four clones were of the IgG1 isotype (Fig. 4).

OPEN IN VIEWER

FIG. 4.

Classes and subclasses of MAbs were determined by Rapid Mouse Isotyping Kit-Gold Series. Each strip was dipped in the indicated hybridoma culture supernate and allowed to develop 10 min. Isotype was determined by comparison to the provided chart, showing positive control “line” and the isotype identified “line.”

Characterization of antibodies

Following validation of our hybridomas, these cells were expanded so that large-scale ascites antibody production could begin. Mice were sensitized with saxoline and implanted with hybridoma cells (clones 1D8 and 5G8) intraperitoneally. Following abdominal swelling, ascites was collected from the peritoneal cavity and purified by saturated ammonium sulfate and Hi-Trap protein affinity columns. The concentration of the collected and purified antibody was 4 mg/mL, as measured by a BCA assay. One clone, 5G8, was then confirmed for specificity by Western blot analysis against KLK6. Recombinant protein, ovarian cancer patient serum or plasma, or mouse liver tissue lysate were all probed with this MAb for antigen binding. As expected, the purified ascites showed significant KLK6 binding at 1:1000 dilution, confirming the presence of anti-KLK6 antibodies produced from the hybridomas (Fig. 5A). To determine the specificity of our generated monoclonal antibodies against other KLK family proteins, we tested them by indirect ELISA for cross-reactivity. Microtiter plates were coated with 1 μg/mL of KLK6, or other KLK6 family members (KLK3, KLK10, and KLK11). These wells were then probed with our MAb. As seen in Figure 5B, the cross-reactivity of our KLK6 monoclonal antibody (5G8) was overall negative for related proteins, implying a strong specificity for KLK6.

OPEN IN VIEWER

FIG. 5.

Characterization of monoclonal antibody 5G8. (A) Western blot detection of KLK6 with 5G8 MAb. Lane M, protein size marker; lane 1, KLK6 protein (0.1 μg); lane 2, ovarian cancer serum (40 μg); lane 3, plasma (40 μg); lane 4, mouse liver lysate (40 μg). (B) Specificity analysis of 5G8 MAb by indirect ELISA. Bar graph shows level of specific binding of the MAb to the indicated coated protein. Results are a compilation of data from four independent experiments.

Sandwich ELISA development and validation

Finally, with our production of several monoclonal antibodies to KLK6, we set out to develop a sandwich ELISA pair for the quantitative detection of KLK6 in various sample types. In order to identify a potential sandwich antibody pair, unlabeled capture antibody (clone1D8) was coated onto microtiter plates and a secondary antibody was chosen to be the detection antibody (5G8) to attempt to make a valid pair. The result, as shown in Figure 6, identified a good correlation between antigen levels and OD seen in the developed ELISA, with an R² = 0.999. Sensitivity of the assay was determined by 20 times the average zero concentration OD, plus 2 standard deviations. As seen in Table 1, the minimum limit of KLK6 that could be detected in our custom ELISA was in the 10 ng/mL range.

OPEN IN VIEWER

FIG. 6.

Establishment of sandwich ELISA. 1D8 MAb was used as capture antibody, and detecting one is 5G8. Recombinant KLK6 protein was serial dilution from 10,000 ng/mL to 1 ng/mL. Curve equation was Y = 3.652 – (0.016 + 3.582) / (1 + (X/5003.055) × 0.554); R2 = 0.999.