The recombinant CTX was prepared from a GST-6′His-CTX precursor by enzymatic cleavage and in vitro refolding, according to our previously described procedure (34). The recombinant Onc was prepared through heterologous expression of an N-terminally 6xHis-tagged Onc (6xHis-Onc) precursor by enzymatic cleavage, N-terminal cyclization and in vitro refolding, according to our previously described procedure (35).

To label CTX with N-Succinimidyl 3-(2-pyrdydithio)propionate (SPDP; Thermo Fisher Scientific, Rockford, IL, USA), the lyophilized recombinant CTX was dissolved in 1.0 mM aqueous HCl (Sinopharm, Beijing, China)and the SPDP reagent in dimethylsulfoxide (DMSO). To start the reaction, they were mixed together at equal concentrations (400 μm each) in the reaction buffer (50 mM phosphate and 150 mM NaCl; pH 8.0) and incubated at room temperature for 1 h. Subsequent to the reaction mixture being acidified to pH 3.0 by trifluoroacetic acid (Merck KGaA, Darmstadt, Germany), high performance liquid chromatography (Agilent Technologies, Santa Clara, CA, USA) was applied to the mixture, according to our previously described chromatography methods (35). The labeled CTX fractions were eluted from a C18 reverse-phase column (Agilent Technologies) by an acidic acetonitrile gradient, manually collected, and lyophilized.

To label Onc with 2-iminothiolane (Thermo Fisher Scientific), the recombinant Onc was dissolved in 1.0 mM aqueous HCl and the reagent 2-iminothiolane in DMSO. To start labeling, Onc (50 μM at final) and 2-iminothiolane (1 mM at final) were mixed in the reaction buffer (50 mM phosphate and 150 mM NaCl; pH 8.0) and incubated at room temperature for 1 h. Thereafter, the reaction mixture was acidified to pH 3.0 by acetic acid and applied to gel filtration. The Onc fraction was eluted from a Sephadex^® G-25 column (Sinopharm) by 10% aqueous acetic acid, manually collected and lyophilized.

To prepare the CTX-Onc conjugate, the SPDP-labeled CTX and the 2-iminothiolane labeled Onc were dissolved in 1.0 mM aqueous HCl, respectively. To start conjugation, they were mixed together at equal concentrations (50 μM each) in the reaction buffer (50 mM NaPO₄ and 150 mM NaCl; pH 8.0) and incubated at room temperature for 0.5 h. Thereafter, the reaction mixture was acidified to pH 3.0 by acetic acid and applied to gel filtration. The CTX-Onc conjugate was eluted from the Sephadex^® G-25 column by 10% acetic acid and lyophilized.

Human glioma U251 and SHG-44 cells were purchased from the Cell Bank, Shanghai Institutes For Biological Sciences, Chinese Academy of Sciences (Shanghai, China) and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and antibiotics, at 37°C in a CO₂ incubator. For the cytotoxicity assay, the cells were seeded into a 96-well plate (5×10³ cells/well). The subsequent day, the cells were changed into the medium containing different concentrations of CTX-Onc conjugate or the physical mixture of CTX and Onc (CTX + Onc), and were continuously cultured for 36 h. Thereafter, the cell viability was assayed using the MTT method (MTT assay kiy; Solarbio, Beijing, China).

All animal experiments were approved by the Animal Care and Use Committee of Tongji University (Shanghai, China). The five-week-old male athymic mice (BALB/c) were obtained from Shanghai Laboratory Animal Center (Shanghai, China). The cultured U251 cells (5×10⁷ cells/mouse) were suspended in 200 μl phosphate-buffered saline (PBS) containing 50% Matrigel™ (BD Biosciences, Franklin Lakes, NJ, USA) and subcutaneously injected into the nude mice. The dimensions (length and width) of the tumors were measured by calipers, and the tumor burden was calculated using the following formula: 0.5 × length × width². Once the tumor had grown to 200–300 mm³, it was removed and cut into uniform tablets (size, 10–20 mm³) subsequent to the mouse being sacrificed in a CO₂ chamber, and then the tumor tablets (one tablet/mouse) were subcutaneously injected into the nude mice. The cultured SHG-44 cells (1×10⁷ cells/mouse) were suspended in 200 μl PBS and subcutaneously injected into the nude mice.

Two months after inoculation with the tablets, the nude mice bearing the U251 tumors were randomly divided into three groups (n=3 mice/group), and two of the groups were intravenously injected at 2.5 mg/kg with 0.2 ml CTX + Onc mixture dissolved in PBS and CTX-Onc conjugate dissolved in PBS, respectively, twice a week for ~1 month. The other group was intravenously injected with 0.2 ml PBS by the same method as a control. The nude mice bearing SHG-44 tumors were only divided into three groups (n=5 mice/group) two months after inoculation, and were then given the same treatment as the nude mice bearing the U251 tumors. The dimensions of the tumors and the weights of the mice were measured at least twice a week and recorded during the treatments.

As CTX can selectively bind malignant gliomas, the present study attempted to use it to target Onc toward gliomas. To prepare the CTX-Onc conjugate, an efficient conjugation procedure was established by which CTX and Onc were covalently crosslinked through reversible disulfide bond(s) (Fig. 1). Firstly, active disulfide(s) were introduced into the recombinant CTX by chemical modification using the primary amine-specific reagent SDPD, while free thiol(s) were introduced into the recombinant Onc by chemical modification using the primary amine-specific reagent 2-iminothiolane. Secondly, CTX-Onc conjugates were formed via reaction of the introduced free thiol(s) of the modified Onc with the introduced active disulfide bond(s) of the modified CTX. Owing to the presence of the reversible disulfide crosslinking in the present CTX-Onc conjugate, free Onc is likely to be released from the conjugate and to enter the glioma cells to exert its cytotoxicity, since the disulfide crosslinking may be reduced by the reductive redox potential at the cell membrane.

CTX has four primary amine moieties, including one N-terminal α-amine and three internal ɛ-amines of the side-chain of lysine (Lys) residues (Fig. 1A). These primary amine moieties would all react with the primary amine-specific modification reagent SPDP during chemical modification. Thus, seven major peaks were eluted from the C18 reverse-phase column following SPDP modification, although the ratio of CTX to SPDP was carefully controlled (Fig. 2A). Mass spectrometry analyses showed that the first peak (indicated by a star) was the unlabeled CTX (measured molecular mass, 3,998.0; theoretical value, 3,996.8); peaks 2–4 were the mono-labeled CTXs (indicated by a letter ‘m’) carrying one active disulfide bond (measured molecular mass, 4,193.0; theoretical value, 4,194.8); peaks 5–7 were the di-labeled CTXs (indicated by a letter ‘d’) carrying two active disulfide bonds (measured molecular mass, 4,391.0; theoretical value, 4,391.8). In the subsequent conjugation reaction, the mono-labeled and di-labeled CTXs were mixed together and reacted with the modified Onc.

Although Onc has no N-terminal α-amine moiety due to the formation of a pyroglutamate residue, it has 12 internal ɛ-amine moieties of the side-chain of Lys residues (Fig. 1B). All these internal primary amine moieties would react with the reagent 2-iminothiolane, thus the reaction condition was optimized by controlling the ratio of Onc to 2-iminothiolane during the modification. Subsequently the excess 2-iminothiolane was removed by gel filtration (Fig. 2B) and the eluted Onc fraction was analyzed by mass spectrometry; 1–3 free thiol moieties were introduced into the majority of Onc molecules (measured molecular mass, 11,920.0, 12,022.0 or 12,121.0). In the subsequent conjugation reaction, the modified Onc mixture was used for conjugation with the SDPD-modified CTX.

To form the CTX-Onc conjugate, the SPDP-modified CTX and the 2-iminothiolane-modified Onc were mixed at a molar ratio of 1:1. Following the reaction, the CTX-Onc conjugate was eluted from a gel filtration column by 5% aqueous acetic acid (Fig. 2C) and analyzed by tricine SDS-PAGE (Fig. 2D). The CTX-Onc conjugate (lane 3) showed smear bands as it contained different numbers of CTX and Onc. Subsequent to treatment by the reducing reagent dithiothreitol, the conjugate (lane 4) showed two sharp bands corresponding to the modified CTX (lane 1) and the modified Onc, respectively. Thus, the conjugate was sensitive to the reducing condition and the Onc would be released from the conjugate under reducing conditions.

To test the anti-glioma effect of the CTX-Onc conjugate, its cytotoxicity to the cultured glioma cells was first measured. To test the targeting effect of CTX, the CTX + Onc mixture at equal molar ratio was used as a control, designated as CTX + Onc. Subsequent to treatment either by CTX-Onc conjugate or by CTX + Onc mixture, the viability of the glioma cells were measured by MTT assay. As shown in Fig. 3, the CTX-Onc conjugate had a dose-dependent cytotoxicity to the U251 and SHG-44 cells, with the concentration to inhibit cell survival to 50% (SF50) of ~20 μg/ml to the two cell types. Furthermore, the conjugate showed much higher (>10-fold) cytotoxicity than the CTX + Onc mixture, suggesting that CTX targeted more Onc to glioma cells and significantly enhanced its cytotoxicity.

As the CTX-Onc conjugate had a significant anti-glioma effect on cultured cells, its effect was further tested on nude mice bearing either U251 or SHG-44 tumors. As shown in Fig. 4A, when the mice bearing subcutaneous U251 tumors were treated with CTX-Onc conjugate, CTX + Onc mixture or PBS for a month, the CTX-Onc conjugate showed the highest inhibition effect on the tumor growth, suggesting that CTX also had a targeting effect in vivo. Subsequent to the mice being sacrificed at the end of treatment, the mean tumor weight of the CTX-Onc group was also lower than that of the CTX + Onc group, confirming the CTX targeting effect (Fig. 4B).

The effect of the CTX-Onc conjugate on the mice bearing the subcutaneous SHG-44 tumors was also tested. After ~1 month of treatment, the CTX-Onc conjugate also showed the highest inhibition effect on tumor growth (Fig. 5A), and it also suggested that CTX had a targeting effect in vivo. Subsequent to the mice being sacrificed at the end of treatment, the mean tumor weight of the CTX-Onc group was also lower than that in the CTX + Onc group, confirming the CTX targeting effect again (Fig. 5B). As shown in Figs. 4C and 5C, the mean weights of the CTX-Onc group were not significantly affected during the treatment, suggesting that the CTX-Onc conjugate exhibited a low toxicity in vivo.