The mouse thymic cortical epithelial reticular cells (1308.1) were kindly provided by Dr Barbara B. Knowles (The Jackson Laboratory) (15). The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM), containing 10% (v/v) fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin (all from Gibco Life Technologies, Grand Island, NY, USA) at 37°C in 5% CO₂ incubator.

The coding sequences of mouse EGFL8 total cDNA were isolated and amplified by PCR and cloned into pcDNA3.1 (Life Technologies, Carlsbad, CA, USA). They were amplified by PCR using the following oligonucleotide primers: EGFL8 forward, 5′-TTT CAA AGA GAG TTT GGG AGT G-3′ and reverse, 5′-CAC CAC GTG T′T CTG TGG TA-3′ to create the Nco1 and Xho1 restriction sites at the start and stop codon sites. The PCR product was cloned into the pET28a vector, which carries a C-terminal His·Tag/thrombin/T7·Tag configuration plus an optional C-terminal His·Tag sequence.

The E. coli bacteria were cultured for 2 h, stimulated by the addition of 0.2 mM IPTG, and then cultured for an additional 4 h. The bacteria were harvested immediately by centrifugation at 6,000 rpm for 8–10 min, and the pellet was either frozen at −80°C until purification or directly resuspended in 60 ml lysis buffer [50 mM Tris-HCl (pH 8.0), 100 mM NaCl and 5 mM EDTA]. Subsequently, 0.5% Triton X-100, 0.1 mM phenylmethylsulfonyl fluoride (PMSF) and 1 mM dithiothreitol (DTT) were added to this pellet. Subsequently, the pellet was sonicated and centrifuged at 12,000 rpm for 15 min. The supernatant was discarded and the pellet was resuspended in lysis buffer. The following steps were repeated as above except for the addition of 10 mM MgCl₂, 0.01 mg/ml DNase and 0.1 mg/ml lysozyme to the Triton X-100, PMSF and DTT mixture. The pellet mixture was incubated for 20 min at room temperature, sonicated and centrifuged as described above. The supernatant was discarded and the pellet was resuspended in 60 ml lysis buffer. Subsequently, only PMSF and DTT were added, sonicated and centrifuged as described above. The supernatant was discarded, and the pellet was resuspended in 40 ml of 8 M urea [100 mM Tris-HCl (pH 8.0), 50 mM glycine]. The pellet was then slowly shaken for 1 h at room temperature to completely solubilize the proteins. After a short centrifugation to remove non-solubilized proteins, the protein concentration was determined by a NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA). The proteins (1–2 mg/ml) were dialyzed through a dialysis bag. Two liters of dialysis buffers [20 mM Tris-HCl (pH 8.0), 150 mM NaCl and 0.1 mM DTT] were used at 4°C for 12 to 15 h and the process was repeated. The third dialysis buffer was used under the same conditions except for the absence of DTT.

Purification of His-tagged proteins for the recombinant protein was then performed. The column was washed with distilled water and then with 6X Ni-NTA washing buffer. The protein mixture was added to the Ni-NTA column and slowly passed through the column. The bound EGFL8 protein was eluted with 6X Ni-NTA elution buffer. The eluted protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) after determining its concentration.

The rEGFL8 protein was subjected to 10% SDS-PAGE, stained with Coomassie blue, and the protein bands were excised from the gel. The excised gel pieces were transferred to microcentrifuge tubes containing 0.5 ml of distilled water. The protein was then digested with trypsin. The single band was used for matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS/MS).

C57BL/6 mice (Dae Han Bio Link, Chungbuk, Korea) were intravenously (i.v.) injected with rEGFL-8 (100 μg) and were sacrificed 12 h, 1, 2 and 3 days after injection. Animal care and all experimental procedures were conducted in accordance with the ‘Guide for Animal Experiments’ published by the Korean Academy of Medical Sciences.

The mice were injected intraperitonealy (i.p.) with BrdU (BD Biosciences Pharmingen, San Diego, CA, USA) or with PBS. The thymus was isolated by dissection 2 h after BrdU injection. The isolated cells were stained for CD4, CD8, CD25, CD44 and BrdU according to the BrdU Flow kit manual (BD Biosciences Pharmingen). For the apoptosis assay, the isolated cells were stained for CD4, CD8, CD25, CD44 and Annexin V according to the Annexin V detection kit protocol (BD Biosciences Pharmingen).

The following fluorochrome-conjugated monoclonal antibodies (mAbs) were purchased from BD Biosciences Pharmingen: Pacific Blue-conjugated anti-CD4 (RM4-5), allophycocyanin (APC)-Cy7-conjugated anti-CD8 (53-6.7), PE-labeled anti-CD25 (PC61) and APC anti-CD44 (IM7). Flow cytometric analysis was performed using a FACSCanto II flow cytometer (BD Biosciences Pharmingen), and the acquired data were analyzed using the FlowJo software (Tree Star, Ashland, OR, USA).

For electron microscopic investigation, all specimens were processed according to a standard procedure. The ultra-thin sections were examined under a JEOL-1200 EXII transmission electron microscope.

The TECs were incubated with 100 ng/ml rEGFL8 for 12 h. The thymocytes were isolated from the mice injected with 100 μg of rEGFL8 i.v., after 12 h, 1 and 2 days. Following treatment with rEGFL8, the thymocytes and TECs were washed with cold PBS and the total proteins were then extracted from the cultured cells using a protein extraction solution (iNtRON Biotechnology, Seongnam, Korea) supplemented with a protease inhibitor mixture (Sigma-Aldrich, St. Louis, MO, USA). Protein concentrations were measured using the Bradford protein assay kit (Bio-Rad, Hercules, CA, USA). Anti-Hes-1 (sc-166378), anti-Hey-1 (sc-28746) (both from Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-β-actin (Abcam, Cambridge, UK) antibodies were used for immunoblot analysis. The goat anti-mouse IgG-HRP (7076) and goat anti-rabbit IgG-HRP (7074) (both from Cell Signaling Technology, Danvers, MA, USA) were used as the secondary antibodies. Immunoreactivity was detected and quantified using a LAS-3000 imaging system (Fujifilm, Tokyo, Japan).

Data are expressed as the means ± SD. Statistical analyses were performed using the Student’s t-test. A value of P<0.05 was considered to indicate a statistically significant difference.

rEGFL8 was expressed in E. coli and purified. The harvested rEGFL8 protein was stained with Coomassie blue in 10% SDS-PAGE (Fig. 1A). We produced high-purity proteins after the third lysis step, suggesting that most of the secreted fusion proteins and most of the other bacterial proteins were removed by the lysis procedure. The dialysis and refolding steps were performed in the presence of urea, mainly as EGFL8 is prone to precipitation at high concentrations and is also prone to protein interactions with the exchange groups on the columns. However, to recover the biological activity of rEGFL8, urea needs to be removed and the protein should be refolded properly. The dialysis protein was analyzed on SDS-PAGE in comparison with cell lysates and supernatants to confirm the successful recovery of the recombinant protein after solubilization with 8 M urea. Fig. 1A illustrates an exact size of the desired protein for this study with no contaminated product from the E. coli cells. This result also verified that most or all of the inclusion bodies of rEGFL8 protein were solubilized in 8 M urea. The Ni-NTA-purified protein was loaded onto 10% SDS-PAGE for total protein analysis (Fig. 1B). The total Ni-NTA column-purified protein showed a single protein band with a specific size, which confirmed the successful rEGFL8 protein purification through the Ni-NTA column. The Ni-NTA-purified protein was also analyzed by western blot analysis for further confirmation (Fig. 1B). The western blot analysis result also showed a single and specific band of rEGFL8 protein. For further characterization, the rEGFL8 protein was run on 10% SDS-PAGE and the band was cut and trypsinized before loading onto the spectrometer (Fig. 2). All liquid chromatography (LC)-MS/MS data were submitted to a Mascot search; the peaks from the purified rEGFL8 protein sample are shown in Fig. 3, with the masses indicated on the top of each peak. The identified molecular weight of purified rEGFL8 measured by MS was 30.09 kDa, which was almost identical to the theoretical value.

To investigate the biological activity of rEGFL8 protein in mouse thymocytes, 100 μg of rEGFL8 protein were injected i.v. into the mice, and the weight of the thymus and the total number of thymocytes were determined (Fig. 4). The weight of the thymus and the total number of thymocytes were significantly decreased after the rEGFL8 injection. The weight of the thymus and the number of thymocytes diminished from 1 day after rEGFL8 injection, and these effects reached a peak value at 3 days. The observation that the number of thymocytes markedly decreased after the administration of rEGFL8 raised the question of whether the subsets of thymocytes were differentially affected by treatment with rEGFL8. To address this issue, mouse thymocytes were freshly isolated 1, 2 and 3 days after rEGFL8 injection, and were stained with anti-CD4 and anti-CD8 mAbs (Fig. 5). Of note, there was a drastic decrease in both the percentage and number of CD4⁺CD8⁺ double-positive (DP) thymocytes 3 days after rEGFL8 injection (Figs. 5 and 6). Moreover, the CD4⁺ single-positive (SP) and CD8⁺ SP thymocyte subsets were also significantly reduced in number compared with the wild-type subset, although to a lesser degree than the CD4⁺CD8⁺ DP thymocytes on day 3 (Fig. 6).

The reduction in the number of thymocytes follwing treatment with rEGFL8 also raised the question of whether these changes occur due to the inhibition of cell proliferation or the promotion of apoptosis of thymocytes by rEGFL8. To shed light on this matter, the effect of rEGFL8 on cell proliferation was first investigated 12 h, 1 and 3 days after rEGFL8 injection. Freshly isolated thymocytes from the thymus 2 h after BrdU injection were stained for anti-CD4, anti-CD8, anti-CD25, anti-CD44 and anti-BrdU, and subsequently flow cytometric analysis was performed (Fig. 7). Remarkably, the BrdU⁺ thymocyte numbers were robustly reduced in the DN, DP, CD4⁺ SP and CD8⁺ SP thymocyte subsets 12 h and 1 day after rEGFL8 injection, indicating that rEGFL8 profoundly inhibited the cell proliferation of all 4 major thymocyte subsets. The number of BrdU⁺ thymocytes returned to a level similar to that of the normal control mice at 3 days post-treatment with rEGFL8 (Fig. 7). Subsequently, to determine whether rEGFL8 induces apoptosis during thymocyte development, thymocytes were stained with Annexin V antibody, as well as anti-CD4 and anti-CD8 mAbs after rEGFL8 injection for 12 h, 1 and 3 days. A relatively low but significant level of apoptosis was induced in total thymocytes 12 h after rEGFL8 injection (Fig. 8). The number of Annexin V⁺ cells reached peak values 12 h following treatment with rEGFL8 (Fig. 8). These results clearly indicate that EGFL8 is not a potent inducer of apoptosis in mouse thymocytes.

To determine whether EGFL8 causes morphological changes in the mouse thymus at the ultrastructural level, transmission electron microscopy was performed after the injection of rEGFL8. The thymus from the control mice showed normal ultrastructural features (Fig. 9A). However, electron micrographs of the thymus 12 h after the administration of rEGFL8 revealed apoptotic thymus cells, although their proportion was relatively small (Fig. 9B). Apoptotic cells (Fig. 9B and C) had typical morphological changes; the chromatins were condensed and aggregated into large dark, compact masses. Most of the apoptotic thymic cells appeared to be thymocytes, whereas the apoptosis of TECs and other thymic stromal cell types was rarely observed. In the thymus 1 day after the administration of rEGFL8, apoptotic thymocytes were observed to a much lesser extent than in at 12 h (Fig. 9C). Within 3 days after the rEGFL8 injection, apoptotic cells were rarely visible and the thymus exhibited an almost normal ultrastructural appearance (Fig. 9D).

To determine the effect of rEGFL8 on the Notch signaling pathway in mouse TECs, TECs were treated with rEGFL8 protein and the expression of Hes1 and Hey1 was then assessed by western blot analysis. Remarkably, the expression of Hes1 and Hey1 was downregulated in the TECs (Fig. 10A). To determine the effect of EGFL8 on the Notch signaling pathway in mouse thymocytes, rEGFL8 protein was injected into the mice, and the expression of Hes1 and Hey1 was assessed by western blot analysis. The expression of Hes1 and Hey1 was significantly downregulated in the mouse thymocytes (Fig. 10B). The inhibitory effect of rEGFL8 on Hes1 expression was more pronounced 12 h after rEGFL8 injection. The decrease in Hey1 expression was most evident 1 day following treatment with EGFL8.