The Epidermal Growth Factor-Seven Transmembrane (EGF-TM7) Receptor CD97 Is Required for Neutrophil Migration and Host Defense

Generation of mAb

Novel mCD97 mAb were generated in Armenian hamsters (Cricetulus migratorius; Cytogen, West Roxbury, MA) as described recently (6). In short, the Armenian hamster fibroblast line ARHO12 was transfected with mCD97(EGF1,2,3,4) cDNA in pcDNA3.1/Zeo(+) (6). Cells were selected with zeocin (Invitrogen, Leek, The Netherlands) at 500 mg/ml in culture medium. Resulting stably transfected clones were tested for mCD97 expression by flow cytometry with the earlier generated mCD97 mAb 1A2 (6). One clone was selected and used for immunization. Three i.p. injections with 10 × 10⁶ irradiated (50 Gy) cells in PBS were given at weekly intervals. Eight weeks after the third injection, the hamster was boosted i.p. with 10 × 10⁶ cells. Three days later, hamster spleen cells were fused with mouse myeloma SP2/0 cells by standard hybridoma technology. Binding of hybridoma supernatants to the ARHO12 clone stably expressing mCD97(EGF1,2,3,4) was tested by flow cytometry. Flow cytometry of COS cells expressing mCD97 isoforms or chimeras between mCD97 and hCD97 was used to identify hybridomas that recognize different regions of mCD97. Selected hybridomas were subcloned until they were monoclonal and stable. The hybridomas 1B2, 1C5, and 1D2 were grown at large amounts and Ig was purified using protein A-Sepharose CL-4B (Sigma-Aldrich, St. Louis, MO).

Generation of chimeric m-hCD97 constructs

Generation of an expression construct encoding mCD97(EGF1h,2,4) by overlap extension-PCR has been described earlier (6). In a similar manner, constructs in which EGF domain 1 or EGF domains 1 and 2 of mCD97 have been exchanged by homologues EGF domains of hCD97 were made. In a first step, two overlapping PCR fragments were generated in separate PCRs. For hCD97(EGF1m,2,5), the N-terminal part of mCD97 including EGF domain 1 was amplified from mCD97 cDNA (6) using a standard T7 primer ((+) strand) and the specific primer 5′-CGTTGATGTCTTCACAGCTCTCTGCAGGGTTAG-3′ (nt 214–237 of the (−) strand; the sequence overlapping with EGF domain 2 of hCD97 is underlined). EGF domains 2 and 5 and part of the stalk region of hCD97 were amplified from hCD97 cDNA (2) using the specific primers 5′-GCTGTGAAGACATCAACGAGTGTGCAACAC-3′ (nt 261–281 of the (+) strand; the sequence overlapping with EGF domain 1 of mCD97 is underlined) and 5′-GGAATGCAGGTTCAAGGAGGC-3′ (nt 1031–1051 of the (−) strand). For hCD97(EGF1m,2m,5), the N-terminal part of mCD97 including EGF domains 1 and 2 was amplified using a standard T7 primer ((+) strand) and the specific primer 5′-CGTCCACATCTTGACATGTATTCTCACTCTCG-3′ (nt 368–390 of the (−) strand; the sequence overlapping with EGF domain 5 of hCD97 is underlined). EGF domain 5 and part of the stalk region of hCD97 were amplified using the specific primers 5′-CATGTCAAGATGTGGACGAGTGCAGCTCCGG-3′ (nt 417–438 of the (+) strand; the sequence overlapping with EGF domain 2 of mCD97 is underlined) and 5′-GGAATGCAGGTTCAAGGAGGC-3′ (nt 1031–1051 of the (−) strand). The PCR fragments were purified and mixed together to serve as template for the second step of the overlap extension-PCR performed for both constructs with a standard T7 primer and the specific primer 5′-GGAATGCAGGTTCAAGGAGGC-3′. Chimeric constructs were obtained by replacing the N-terminal part of hCD97(EGF1,2,5) cDNA in pcDNA3.1/Zeo(+) with the PCR amplicons, thereby using a HindIII site in the multiple-cloning site of the vector and a EcoRV site in the hCD97 sequence.

To generate a mCD97(EGF1h,2h,4) construct, the N-terminal part of mCD97 including EGF domain 1 and part of EGF domain 2 was amplified using a standard T7 primer ((+) strand) and the specific primer 5′-CTTGAATGTTTTTGCCCCAGAAAC-3′ (nt 368–391 of the (−) strand). A chimeric construct was obtained by replacing the N-terminal part of mCD97(EGF1,2,4) cDNA in pcDNA3.1/Zeo(+) with the PCR amplicon, thereby using a NheI site in the multiple-cloning site of the vector and a HpaI site in the mCD97 sequence (blunt-end ligation of the 3′ end of the amplicon).

COS cell expression and mapping of mAb binding sites

COS cells were transfected with mCD97, hCD97, or chimeric m-hCD97 constructs using Lipofectamine Plus reagent (Life Technologies, Gaithersburg, MD). Three days after transfection, reactivity of mCD97 mAb 1A2, 1B2, 1C5, and 1D2 was analyzed by standard flow cytometry on a FACScan (BD Biosciences, Mountain View, CA). Control stainings were performed with hCD97 mAb CLB-CD97/1 and CLB-CD97/3 (40). PE-conjugated goat anti-hamster Ig (Southern Biotechnology Associates, Birmingham, AL) or PE-conjugated goat anti-mouse Ig (Immunotech, Marseille, France) were used as second step reagent.

Erythrocyte adhesion studies

Adhesion studies were performed as described previously (6, 30, 33) by overlaying COS cells in 12-well cell culture plates three days after transfection with 50 × 10⁶ mouse erythrocytes for 30 min at room temperature. Nonadhering cells were removed by gentle washing with PBS before examination by microscopy. To test for blocking capacity, mCD97 mAb were added to the erythrocyte suspension at a final concentration 5 μg/ml.

Induction of dextran sulfate sodium (DSS) colitis

Pathogen-free 8-wk-old female BALB/c mice were obtained from Harlan (Zeist, The Netherlands). The Animal Care and Use Committee of the University of Amsterdam (Amsterdam, The Netherlands) approved all experiments described in this manuscript. To induce DSS colitis, mice were fed 4% (w/v) DSS (TdB Consultancy, Uppsala, Sweden) in their drinking water for at least 7 days (41).

Adoptive transfer of technetium-99m (^99mTc)-labeled neutrophils and pinhole single photon emission computed tomography (SPECT)

Peritoneal granulocytes were harvested by rinsing the peritoneal cavity with 5 ml of sterile PBS 5 h after i.p. injection of 1 ml of 10% proteose peptone (Difco, Detroit, MI) in PBS. A total of 5–10 × 10⁶ cells/mice were harvested and labeled with 75 MBq of freshly prepared ^99mTc-hexamethylpropylene amine oxime (HMPAO) according to manufacturers’ guidelines (Ceretec; Amersham Health, Eindhoven, The Netherlands). After 15 min, cells were washed and resuspended in saline, containing 5 μg/ml Ab. Next to mCD97 mAb 1B2, hamster Ig (Rockland, Gilbertsville, PA) was used as control (cIg).

Within 30 min after labeling, ∼5 × 10⁶ cells, labeled with 40 MBq ^99mTc-HMPAO, were injected i.v. in a tail vein of mice with DSS colitis (n = 7–8). One hour later, scintigraphy was performed on rotating mice for three-dimensional imaging using a recently described and validated high-resolution pinhole SPECT technique (42). During the whole procedure, animals were sedated with fentanyl and fluanisone (Janssen Pharmaceutica, Beerse, Belgium) and diazepam (Roche, Mijdrecht, The Netherlands). SPECT reconstruction was done using a HERMES (Nuclear Diagnostics, Stockholm, Sweden) application program, using filtered back projection adapted to pinhole SPECT.

To determine the radioactivity uptake semiquantitatively, and to correct for differences in total amount of radioactivity administered and variable numbers labeled granulocytes, the colon uptake ratio was determined as follows. Five consecutive transverse slices with the highest colon uptake were selected and added. Regions of interest were set for the colon and abdominal background (not containing the colon) and the number of counts in each region of interest was measured. The colon uptake ratio was calculated by subtracting background activity from the colon activity and subsequently dividing the corrected colon uptake by the background to yield a specific uptake ratio: (counts colon-counts background)/(counts background).

Induction of pneumonia

Pathogen-free 8-wk-old female BALB/c mice were obtained from Harlan Sprague Dawley. Pneumonia was induced as described previously (43, 44, 45). Briefly, S. pneumoniae serotype 3 was obtained from American Type Culture Collection (ATCC 6303; Rockville, MD). Pneumococci were grown for 6 h to mid-logarithmic phase at 37°C in 5% CO₂ using Todd-Hewitt broth (Difco), harvested by centrifugation at 1500 × g for 15 min, and washed twice in sterile isotonic saline. Bacteria were then resuspended in sterile isotonic saline at a concentration of ∼1 × 10⁴ CFU/50 μl, as determined by plating serial 10-fold dilutions on sheep-blood agar plates. Mice were lightly anesthesized by inhalation of isoflurane (Upjohn, Ede, The Netherlands) and 50 μl were inoculated i.n.

Abs 1B2, 1C5, and cIg were given i.p. at doses of 0.5 mg 24 h before, 24 and 72 h after induction of pneumonia. The size of the groups was n = 10 for the survival study and n = 8 for the remaining assays.

Preparation of lung homogenates and determination of bacterial outgrowth

At 20 and 44 h after inoculation, mice were anesthetized with hypnorm (Janssen Pharmaceutica) and midazolam (Roche) and sacrificed by bleeding out the vena cava inferior. Blood was collected in EDTA-containing tubes. Whole lungs were harvested and homogenized at 4°C in four volumes of sterile saline using a tissue homogenizer (BioSpec Products, Bartlesville, OK). CFU were determined from serial dilutions of lung homogenates and blood, plated on blood agar plates and incubated at 37°C at 5% CO₂ for 16 h before colonies were counted.

Histologic examination of lungs

Lungs for histologic examination were harvested at 20 h and 44 h after inoculation, fixed in 10% formalin and embedded in paraffin. Four-micrometer sections were stained with H&E and analyzed by a pathologist who was blinded for groups. For granulocyte staining (46), slides were deparaffinized and endogenous peroxidase activity was quenched by a solution of methanol/0.03% H₂O₂. After digestion with a solution of pepsine 0.25% in 0.01 M HCl, the sections were incubated in 10% normal goat serum and then exposed to FITC-labeled mouse Ly-6G mAb (BD PharMingen, San Diego, CA). Slides were incubated with a rabbit anti-FITC Ab (DAKO, Glostrup, Denmark) followed by a further incubation with a biotinylated swine anti-rabbit Ab (DAKO), rinsed again, incubated in a streptavidin-ABC solution (DAKO) and developed using 1% H₂O₂ and 3.3′-diaminobenzidin-tetra-hydrochloride in Tris-HCl. The sections were mounted in glycerin gelatin with a light methylgreen counter staining and analyzed.

Cytokine and chemokine determinations

For cytokine and chemokine measurements, lung homogenates were diluted 1 to 2 in lysis buffer containing 300 mM NaCl, 30 mM Tris, 2 mM MgCl₂, 2 mM CaCl₂, 1% Triton X-100, and pepstatin A, leupeptin, and aprotinin (all 20 ng/ml; pH 7.4) and incubated at 4°C for 30 min. Homogenates were centrifuged at 1500 × g at 4°C for 15 min, and supernatants were stored at −20°C until assays were performed. Cytokines (TNF-α, IL-1β) and chemokines (KC, macrophage-inflammatory protein-2 (MIP-2)) were measured using specific ELISA (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions.

Blood cell differentiation

EDTA blood was collected from mice 24 h after i.p. administration of 0.5 mg 1B2 or cIg per mouse (n = 4). Whole blood counts were determined and the number of neutrophils was calculated from these totals, using cytospin preparations stained with modified Giemsa stain (Diff-Quick; Baxter Diagnostics, McGraw Perk, IL).

Statistical analysis

Differences between groups were calculated by Mann-Whitney U test. For survival analyses, Kaplan-Meier analysis followed by log rank test was performed. Values are expressed as mean ± SEM. A two-tailed p value of < 0.05 was considered to represent a significant difference.

Generation and characterization of novel mCD97 mAb

To generate mAb with different binding sites on the mCD97R, an Armenian hamster was immunized with ARHO12 cells stably expressing mCD97(EGF1,2,3,4). Screening of hybridoma supernatants for mCD97 specificity identified > 100 positive clones. mAb with disparate binding sites were detected when analyzing binding to COS cells expressing different mCD97 isoforms and m-hCD97 chimeras. The mAb 1B2, 1C5, and 1D2 were selected and analyzed in detail, together with the earlier generated mAb 1A2 (6), for binding specificity using an extended panel of m-hCD97 chimeras. As shown in Table I⇓, 1B2 binds EGF domain 1, 1A2 binds EGF domain 2, and 1C5 and 1D2 bind EGF domain 3. Accordingly, 1C5 and 1D2 recognize mCD97(EGF1,2,3,4) and mCD97(EGF1,2,X,3,4) but not the smallest isoform mCD97(EGF1,2,4).

We previously reported that the mAb 1A2 blocks adhesion of mouse erythrocytes to COS cells expressing mCD97(EGF1,2,4) or mCD97(EGF1,2,3,4) (6). Using the same assay, the ability of the novel mAb to interfere with the interaction between mCD97 and mouse CD55 was tested (Table I⇑). Whereas adhesion was blocked by 1B2, no blocking effect was seen with 1C5 and 1D2, indicating that EGF domain 1 and 2 form the binding site for mCD55. This finding is in good agreement with a recent mutational analysis study demonstrating the importance of hCD97 EGF domains 1 and 2 for human CD55 binding (34).

Binding of CD97 mAb inhibits neutrophil homing in experimental colitis

Using the newly generated mCD97 mAb, we evaluated the role of CD97 in neutrophil migration. First, we investigated the homing of radioactively labeled neutrophils that had been incubated with mAb, to the colon in mice with experimental colitis. In this model, neutrophils, recruited from the peritoneum of donor mice and labeled with ^99mTc-HMPAO, were incubated with cIg or 1B2 before adoptive transfer into the blood circulation of acceptor mice sensitized with DSS. As depicted in Fig. 2⇓, homing of neutrophils treated with 1B2 to the colon but also to other sites was significantly diminished when compared with preincubation with cIg.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Effect of preincubation with mCD97 mAb on the homing of radioactively labeled neutrophils to the colon in experimental colitis. Pinhole SPECT was performed of mice with DSS colitis 1 h after adoptive transfer of 5 × 10^{6 99m}Tc-labeled neutrophils. A, Pinhole SPECT images of the pelvic region of mice 1 h after injection of neutrophils treated with cIg (upper panels) or 1B2 (lower panels). Corresponding transverse, sagittal, and coronal slices are shown. Bladder (B), colon (C), limbs (L), pelvis (P), and spleen (S) are indicated. Whereas radioactivity uptake in colon is clearly visible with cIg-treated neutrophils, 1B2 efficiently blocked neutrophil homing. Images are representative for both groups. B, Radioactivity uptake in the colon. Specific colon radioactivity uptake was calculated of mice injected with neutrophils treated with cIg (○) or 1B2 (♦). Medians are indicated with horizontal lines. Data are pooled from two separate experiments. n = 3–4/group/experiment. ∗∗, p < 0.05.

Application of mCD97 mAb results in increased severity of S. pneumoniae-induced pneumonia

Based on the delaying effect on neutrophil migration, we expected that application of mCD97 mAb would be detrimental for anti-bacterial immunity. This assumption was tested in the well established S. pneumoniae pneumonia model (43, 44, 45). We first assessed the effect of the mAb 1B2 and 1C5 on survival of mice after i.n. inoculation with 1 × 10⁴ CFU S. pneumoniae. As shown in Fig. 3⇓A, a dramatic difference was observed between mice treated with mCD97 mAb or cIg. Whereas all mice of the control group survived to day 20, 70% of animals from the 1B2- and the 1C5-treated group died within 2 wk.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Effect of application of mAb to mCD97 on the course of pneumococcal pneumonia. Mice were inoculated i.n. with 1 × 10⁴ CFU S. pneumoniae at day 0. A, Survival study. Mortality of mice treated with cIg (○), 1B2 (♦), or 1C5 (▪) was assessed 1–2 times daily for 20 days. The p value indicates the difference between the mCD97 mAb-treated groups and the control group, n = 10/group. B, Bacterial outgrowth in lungs. CFU S. pneumoniae in lungs of mice treated with cIg, 1B2, or 1D2 20 and 44 h after inoculation. Medians are indicated with horizontal lines, n = 8/group/time point. ∗, p < 0.05; ∗∗∗, p < 0.0005.

The rapid disease progression in mCD97 mAb-treated mice was reflected by pneumococci titers cultured from the lungs as compared with the control group (Fig. 3⇑B). At 20 h after inoculation, the lung S. pneumoniae burden was similar in all three groups. In contrast, after 44 h, mCD97 mAb-treated mice had four-log greater lung S. pneumoniae titers than cIg-treated mice. Lung pneumococci titers below the detection limit, indicating bacterial clearance, were observed in 7 of 8 mice from the control group, but only in 0 of 8 mice of the 1B2-treated and in 2 of 8 mice of the 1C5-treated group. Blood cultures obtained at 20 h, were negative for S. pneumoniae in all three groups. After 44 h, pneumococci were detectable in the blood of 0 of 8 mice, 4 of 8 mice, and 2 of 8 mice of the groups treated with cIg, 1B2, and 1C5, respectively.

mCD97 mAb impair the recruitment of neutrophils to the lungs during pneumococcal pneumonia

Histological analysis revealed that at 20 h after infection with S. pneumoniae, lungs of all control mice displayed interstitial and peribronchial inflammation together with some degree of endothelialitis. Moreover, in 50% of these mice multiple foci of pneumonia were present in the lungs (Fig. 4⇓A). In contrast, mice treated with mCD97 mAb presented less inflammation in the lungs and none of the animals had developed foci of pneumonia (Fig. 4⇓, B and C). These findings were confirmed by immunostaining for granulocytes showing a prominent granulocytic infiltration in cIg-treated mice (Fig. 4⇓D) but only few granulocytes in interstitial areas in mCD97 mAb-treated mice (Fig. 4⇓, E and F). At 44 h after inoculation, all mice showed moderate inflammatory infiltrates essentially composed of monocytes and macrophages. There was no difference between control and treated mice (data not shown). The observation of similar degrees of infiltration in all three groups at 44 h may reflect a relative deficiency in neutrophil influx as the bacterial load was much higher in mCD97 mAb-treated mice at that time point.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Histopathology of lungs 20 h after induction of pneumococcal pneumonia. Mice treated with cIg (A) show interstitial and intraalveolar inflammation with endothelialitis. In contrast, mice treated with 1B2 (B) and 1C5 (C) displayed only slight interstitial inflammation. This was confirmed by immunohistochemistry for granulocytes showing massive influx of granulocytes in mice treated with cIg (D) and little granulocyte infiltration in mice treated with 1B2 (E) and 1C5 (F). A–C, H&E staining, D–F, Ly-6G immunostaining. Representative sections from all groups are shown. Magnification ×20.

We next determined expression levels of soluble inflammatory mediators implicated in pulmonary host defense (47). As depicted in Table II⇓, levels of cytokines (TNF-α, IL-1β) and neutrophil-attracting chemokines (KC, MIP-2) in the lungs did not significantly differ between the three groups. These data imply that alterations in the expression of humoral factors are unlikely to contribute to the diminished response in mCD97 mAb-treated mice during pneumococcal pneumonia.

In a final experiment, we confirmed that i.p. administration of mCD97 mAb did not result in neutropenia. One day after application of 0.5 mg Ig, peripheral blood neutrophil counts (× 10⁶ ± SEM) were 1.0 ± 0.3 in cIg-treated and 0.9 ± 0.3 in 1B2-treated mice (p = 0.4, n = 4).