Genotoxicity studies on HM10760A, recombinant ...

4 downloads 0 Views 598KB Size Report
Genotoxicity studies on HM10760A, recombinant human erythropoietin conjugated to globin fragment. Se Chang Kwon1, Gwan Sun Lee1, Jae Yong Han2, ...
Drug and Chemical Toxicology, 2010; 33(2): 152–159

RESEARCH ARTICLE

Drug and Chemical Toxicology Downloaded from informahealthcare.com by Seoul National University on 04/20/10 For personal use only.

Genotoxicity studies on HM10760A, recombinant human erythropoietin conjugated to globin fragment Se Chang Kwon1, Gwan Sun Lee1, Jae Yong Han2, Michael Lee3, Young-Mi Lee4, Woo Suk Koh5 and Ji-Young Kim5 Research Center, Hanmi Pharm Co., Hwaseong, Korea, 2School of Agricultural Biotechnology, Seoul National University, Seoul, Korea, 3Department of Biology, University of Incheon, Incheon, Korea 4Polar BioCenter, Korea Polar Research Institute, Incheon, Korea and 5Korea Institute of Toxicology, KRICT, Daejeon, Korea 1

Abstract HM10760A is a recombinant human erythropoietin chemically conjugated to the N-terminus of human immunoglobulin Fc fragment through a polyethylene glycol linker. HM10760A was shown to have a relatively long half-life, compared with unconjugated recombinant erythropoietin. In this study, the genotoxicity of HM10760A was investigated by using a test battery of three different methods. In the Ames assay, five strains (TA100, TA1535, TA98, TA1537, and Escherichia coli WP2 uvrA) were tested at six concentrations of 3.13, 6.25, 12.5, 25, 50, and 100 μg/plate. HM10760A did not increase the number of revertant colonies in any tester strains with and without metabolic activation by rat-liver S9 mix. Subsequently, in vitro chromosomal aberration test, using Chinese hamster lung cells, were conducted at the concentrations of 25, 50, and 100 μg/mL. HM10760A did not induce chromosomal aberrations either in the short-period (6 hours) test with or without rat-liver S9 mix or in the continuous-treatment (24 hours) test. In the in vivo bone marrow micronucleus assay using the male ICR (imprinting control region) mouse, HM10760A was subcutaneously administered twice at 24-hour intervals at doses of 0, 150, 300, and 600 μg/kg. HM10760A produced a slight, but statistically significant, increase in the frequency of micronucleated polychromatic erythrocytes at 600 μg/kg. However, no biological significance was assumed, because this value was within the historical control range. From these findings obtained from the genotoxicity assays performed in this study, it appears unlikely that HM10760A acts as a genotoxic agent in vitro and in vivo. Keywords:  Erythropoietin; Ames assay; chromosomal aberration assay; micronucleus assay; genotoxicity

Introduction Erythropoietin (EPO) is a 34-kD glycoprotein hormone that regulates red blood cell production through the promotion of survival, proliferation, and differentiation of the erythroid progenitor cells in the bone marrow (Fisher, 2003; Lacombe and Mayeux, 1999). Efforts to increase its efficacy in vivo by manipulating the protein’s structure have met with some success (Debeljak and Sytkowski, 2008). In particular, EPO has been synthetically produced by using recombinant DNA technology for use in persons with anemia due to kidney failure (Winearls et  al., 1986), anemia secondary to AZT treatment of AIDS (Revicki et  al, 1994), and anemia associated with

cancer (Bohlius. et al., 2006). Despite the worldwide use of recombinant human EPO (rhEPO), the need for frequent administration due to its short half-life has prompted the development of improved agents to stimulate erythropoiesis. Thus, the development of HM10760A with prolonged survival in the circulation would offer the advantage of more convenient dosing and may enhance compliance and clinical efficacy, compared with authentic rhEPO. This study was performed to determine the genotoxic potential in HA10706A by using a battery of three different assays recommended by the International Conference on Harmonization (ICH, 1996, 1997): Ames tests, chromosome aberrations in Chinese hamster lung (CHL) cells, and in vivo micronucleus

Address for Correspondence:  Ji-Young Kim,Korea Institute of Toxicology, P.O. Box 123, Yusong, Daejeon 305-600, Korea; Fax: +82-42-610-8172; E-mail: [email protected] (Received 06 May 2009; revised 07 June 2009; accepted 21 July 2009) ISSN 0148-0545 print/ISSN 1525-6014 online © 2010 Informa UK Ltd DOI: 10.3109/01480540903196824

http://www.informahealthcare.com/dct

Genotoxicity of recombinant human erythropoietin   153

formations scored in bone marrow cells of ICR (imprinting control region) mice.

Materials and methods

Drug and Chemical Toxicology Downloaded from informahealthcare.com by Seoul National University on 04/20/10 For personal use only.

Materials Most chemicals, including positive controls, were obtained from Sigma (St. Louis, Missouri, USA), except for 9-aminoacridine (9AA), which was from Merck (Darmstadt, Germany). Minimal essential medium (MEM) medium, fetal bovine serum (FBS), penicillin, and streptomycin were purchased from Gibco BRL (Grand Island, New York, USA). S9, which was prepared from male Sprague-Dawley rats induced with Aroclor 1254 or phenobarbital/5,6benzoflavone, was from Molecular Toxicology, Inc. (Boone, North Carolina, USA) and cofactor for S9 mix was from Wako Pure Chemical Industries, Ltd. (Tokyo, Japan). Preparation of HM10760A HM10760A is a highly purified chemical conjugate of recombinant human erythropoietin (rhEPO) and human immunoglobulin fragment. The rhEPO and immunoglobulin fragment were linked by a nonpeptidyl polyethylene glycol (PEG) linker. The rhEPO was produced by transformed Chinese hamster ovary (CHO) cells, using serum-free medium (animal­component–free), followed by a purification process. The immunoglobulin fragment was produced by fermentation of transformed Escherichia coli as an inclusion body, followed by a successive refolding and purification. The solubility of HM10760A was found to be higher than 37.1 mg/mL. Animals Six-week-old specific pathogen-free male ICR mice (CrljBgi:CD1, 24~31 g) and 8-week-old female normocythemic mice (CrljBgi:BDF1, 18~22 g) were purchased from Orient, Ltd. (Seoul, Korea) and used after at least 1 week of acclimation. The animals were housed in polycarbonate cages with sawdust bedding. Randomization was carried out by using the Path/Tox System (version 4.2.2.; Xybion Medical Systems Corporation, Cedar Knolls, New Jersey, USA). The animals were maintained at temperature of 23 ± 3°C, relative humidity of 50 ± 10%, air ventilation of 10~20 times/hour, and light intensity of 150~300 Lux with a 12-hour light-dark cycle. Pelleted food for

experimental animals (PMI Nutrition International, Richmond, Indiana, USA) and ultraviolet (UV)irradiated (Steritron SX-1; Daeyoung, Inc., Seoul, Korea) and ­filtrated (1 μm) tap water were given ad libitum. All procedures were approved by the Institutional Animal Care and Use Committee of the Korea Institute of Toxicology (Daejeon, Korea). Measurement of in vitro and in vivo erythroid activity In vitro cellular assay of HM10760A was performed by using 32D/hEPOR cells (mouse lymphoblast cell line, ATCC CRL-11346), which were transfected with human erythropoietin receptor (hEPOR) to proliferate in response to EPO. The EPO-BRP (biological reference preparation of EPO) was used for a reference control to compare the percent ratio of the EC50 value. The in vivo biological activity of HM10760A was assessed by using the normocythemic mice, as described in the European Pharmacopoeia (European Directorate for the Quality of Medicines and HealthCare, 2005). Eight-week-old normocythemic mice (n = 8/group) received 0.5 mL of 35.23, 17.61, and 8.84 ng/mL of HM10760A via a single subcutaneous injection. Blood samples were collected after 96 hours for reticulocyte counting, using an ADVIA® 120 Hematology analyzer (Bayer, Berkeley, California, USA). The relative potency, compared to EPO-BRP, was calculated by the parallel-line assay of the PLA2.0 (Stegmann Systems, Rodgau, Germany). Ames assay Salmonella typhimurium strains TA98, TA1537, TA100, and TA1535 and Escherichia coli WP2uvrA (Molecular Toxicology, Inc) were used as test strains, as described by Maron and Ames (1983). A 0.1-mL aliquot of HM10760A, containing 0–100 μg per plate, 0.5 mL of S-9 mix (or sodium phosphate buffer, pH 7.4), and a 0.1-mL inoculum of the test strain were added to each tube containing 2 mL of top agar. The contents of the test tubes were mixed well and poured onto the Vogel-Bonner minimal agar plates. Plates were incubated at 37°C for 48 hours. In vitro chromosomal aberration assay This assay was performed by using CHL cells (ATCC #CRL-1935), which were obtained from ATCC (Manassas, Virginia, USA). The cells were cultured in MEM, supplemented with 100 U of penicillin,

Drug and Chemical Toxicology Downloaded from informahealthcare.com by Seoul National University on 04/20/10 For personal use only.

154   Se Chang Kwon et al.

100 μg/mL of streptomycin, 2 mM of L-glutamine, and 10% FBS. The assay consisted of short-term and continuous (22-hour) treatments. Cells were seeded approximately 72 hours before treatment at 4 × 104 cells/5 mL/25 cm2 flask. For short-term treatment, HM10760A was treated for 6 hours, followed by a recovery period of 18 hours, with and without rat-liver S9. Colchicine (1 μM) was added 2 hours before cell harvesting. The cells were trypsinized and incubated in a 75-mM hypotonic KCl solution for 20 minutes at 37°C, fixed with acetic-acid– ethanol (1:3 by volume), and then spread onto clean glass slides. Each slide was stained with 2.5% Giemsa solution (pH 6.8). To determine cytotoxicity, cell numbers were counted by a Coulter counter (Beckman Coulter, Inc., Fullerton, California, USA). Observation criteria for chromosomal aberrations were based on the Atlas of Chromosomal Aberrations Induced by Chemicals (JEMS/MMS, 1988). The results were expressed as the number of findings per 200 metaphases. An additional 100 metaphases were examined to determine the frequency of polyploid (PP) and endoreduplication (ER). In vivo micronucleus assay HM10760A was administered subcutaneously to groups of 6 male ICR mice twice at 24-hour intervals at doses of 150, 300, and 600 μg/kg. The positive control received a single intraperitoneal injection of cyclophosphamide at 70 mg/kg. Mice were sacrificed by CO2 gas inhalation at about 24 hours after the final administration for femur collection. The bone marrow was expelled from the cavity by repeated gentle aspirations and flushings with FBS and centrifuged at approximately 1,000 rpm for 5 minutes in a clinical centrifuge. The pellet was suspended in a small volume of serum and used for making two slides. The air-dried slides were stained with May-Grunwald and Giemsa. The slides were then examined under 1,000X magnification. Small round or oval-shaped bodies, the size of which ranged from about 1/5 to 1/20 of the diameter of polychromatic erythrocyte (PCE), were counted as micronuclei. A total of 2,000 PCEs were scored per animal by the same observer for determining the frequencies of MNPCEs. PCE/ (PCE+NCE) ratio, an indicator of cytotoxicity, was calculated by counting 500 cells. Statistical analysis The statistical analyses for in vitro chromosomal aberration results, according to Richardson et  al.

(1989), were performed by using the Statistical Analysis System (SAS) program (SAS, Inc., Cary, North Carolina, USA). The number of aberrant metaphases (excluding gaps, according to OECD guidelines) and [PP+ER] were analyzed. Statistical evaluation of the in vivo micronucleus results was performed according to Lovell et  al. (1989), with a minor modification. The study was accepted when all of the PCE/(PCE+NCE) ratio were greater than 0.1 (Heddle et  al., 1984). The result of the statistical evaluation was regarded significantly when the P-value was less than 0.05. No statistical analyses were performed on the Ames results.

Results Erythroid activity of HM10760A In vitro cellular assay, using 32D/hEPOR cells, showed that the EC50 of HM10760A was 46.72 ng/mL, while that of BRP-EPO, an international standard for EPO, was 2.89 ng/mL (Table 1). An ­approximately 16-fold higher concentration of HM10760A was required to induce the proliferation of 32D/ hEPOR, compared with EPO-BRP. In vivo activity of HM10760A, measured with the inducing capability of blood reticulocytes in normocythemic mice, was 3.74 × 106 IU/mg, which is 26-fold higher than that of BPR-EPO (Table 1) Ames assay of HM10760A As the test article is a protein-based substance with specific biological activity, the maximum dose level was determined according to the highest Cmax value obtained in a PK/PD study and referring the previously reported data on EPO (Che et al., 1999; Kang et al., 1998). The highest Cmax value obtained in our PK/PD study was approximately 100 ng/mL. The highest dose of 100 μg/plate corresponds to 37 μg/mL in top agar mixture, which is about 370-fold higher than the Cmax and >3,700-fold higher than the anticipated Cmax in human clinical trials. An 100 μg/plate was selected for the all strains as the Table 1.  Erythroid activity of HM10760A. In vitro activity (EC50, ng/mL) HM10760A 46.72 ± 4.84a b 2.89 ± 0.35 BRP-EPO a Mean ± standard deviation (n = 3). International standard for erythropoietin. Determined from one sample (n = 1).

In vivo activity (x 106 IU/mg) 3.736 ± 0.147 0.141c

Genotoxicity of recombinant human erythropoietin   155 Table 2.  Ames assay for HM10760A. Without S-9 mix With S-9 mix Colonies/plate(mean) [factor]a Dose (μg/plate) Colonies/plate(mean) [factor]a 0 116 ± 10 124 ± 21 3.13 110 ± 10 [0.9] 121 ± 3 [1.0] 6.25 119 ± 9 [1.0] 121 ± 10 [1.0] 12.5 104 ± 4 [0.9] 119 ± 15 [1.0] 25 103 ± 2 [0.9] 113 ± 15 [0.9] 50 106 ± 3 [0.9] 114 ± 21 [0.9] 100 105 ± 5 [0.9] 107 ± 9 [0.9] TA1535 HM10760A 0 11 ± 4 11 ± 2 3.13 9 ± 1 [0.8] 9 ± 3 [0.8] 6.25 10 ± 3 [0.9] 9 ± 5 [0.8] 12.5 7 ± 1 [0.6] 9 ± 3 [0.8] 25 9 ± 1 [0.8] 6 ± 2 [0.5]* 50 12 ± 3 [1.1] 11 ± 4 [1.0] 100 8 ± 2 [0.7] 7 ± 3 [0.6] TA98 HM10760A 0 34 ± 3 41 ± 6 3.13 30 ± 3 [0.9] 35 ± 4 [0.9] 6.25 38 ± 9 [1.1] 44 ± 3 [1.1] 12.5 39 ± 1 [1.1] 41 ± 4 [1.0] 25 35 ± 4 [1.0] 44 ± 4 [1.1] 50 38 ± 6 [1.1] 41 ± 5 [1.0] 100 28 ± 3 [0.8] 35 ± 10 [0.9] TA1537 HM10760A 0 5 ± 1 18 ± 5 3.13 5 ± 2 [1.0] 13 ± 5 [0.7] 6.25 5 ± 2 [1.0] 14 ± 2 [0.8] 12.5 5 ± 1 [1.0] 11 ± 2 [0.6] 25 6 ± 1 [1.2] 14 ± 1 [0.8] 50 8 ± 2 [1.6] 9 ± 4 [0.5]* 100 6 ± 1 [1.2] 15 ± 2 [0.8] Escherichia coli HM10760A 0 17 ± 2 20 ± 2 WP2uvrA 3.13 12 ± 2 [0.7] 20 ± 1 [1.0] 6.25 16 ± 2 [0.9] 17 ± 5 [0.9] 12.5 14 ± 1 [0.8] 19 ± 2 [1.0] 25 15 ± 4 [0.9] 18 ± 3 [0.9] 50 14 ± 4 [0.8] 17 ± 3 [0.9] 100 14 ± 2 [0.8] 15 ± 1 [0.8] TA100 SA 0.5 392 ± 18 [3.4] TA1535 SA 0.5 333 ± 25 [30.3] TA98 2-NF 2 393 ± 13 [11.6] TA1537 9-AA 50 191 ± 14 [38.2] WP2 uvrA 4NQO 0.5 204 ± 10 [12.0] TA100 BP 2 1249 ± 69 [10.1] TA1535 2-AA 2 14 ± 1 [1.3] 228 ± 3 [20.7] TA98 BP 2 39 ± 3 [1.1] 391 ± 9 [9.5] TA1537 BP 2 192 ± 2 [10.7] WP2 uvrA 2-AA 4 246 ± 15 [12.3] a No. of colonies of treated plate/no. of colonies of vehicle control plate. SA, sodium azide; 2-NF, 2-nitrofluorene; 9-AA, 9-aminoacridine; 2-AA, 2-aminoanthracene; 4NQO, 4-nitroquinoline 1-oxide, BP, benzo(a)pyrene.

Drug and Chemical Toxicology Downloaded from informahealthcare.com by Seoul National University on 04/20/10 For personal use only.

Tester strain TA100

Chemical treated HM10760A

highest concentration of this study. The results of the Ames assay of HM10760A are shown in Table 2. HM10760A did not induce more than twice the number of revertant colonies, in comparison

with the negative control, in any tester strains with or without S9 mix. In addition, potential antibacterial effects, such as decrease in the number of colonies, were not observed in all of the strains tested.

156   Se Chang Kwon et al.

Drug and Chemical Toxicology Downloaded from informahealthcare.com by Seoul National University on 04/20/10 For personal use only.

The positive controls induced a marked increase in the numbers of revertant colonies, compared to the vehicle control. The number of revertant colonies in both the negative and positive controls was within the historical control ranges obtained at the Korea Institute of Toxicology. As a result, HM10760A was considered to be nonmutagenic in bacteria under the conditions of this assay. In vitro chromosomal aberration assay of HM10760A In order to give a high enough concentration of HM10760A, 100 μg/mL was selected as the highest concentration of in vitro chromosomal aberration assay. The Cmax obtained in our PK/PD study was about 100 ng/mL, which is 1,000-fold lower than the highest exposure concentration in the assay. For the continuous-treatment test, structural ­chromosomal aberrations and the polyploidy induction rates were less than 2% at any concentration (Table 3). In the short-term treatment, total aberrant metaphases, excluding gaps of HM10760A, were less than 1.5% without S-9 mix and less than 1% with S-9 mix, respectively. Since structural aberrations did not exceed 2% in either treatment, HM10760A was concluded to be nonclastogenic in this assay. No

reduction in relative cell count was found up to the highest concentration of HM10760A treated in this study. In vivo micronucleus assay of HM10760A The results of the bone marrow micronucleus test in mice are shown in Table 4. As the test article is a substance with specific biological activity at the low nontoxic dose, the maximum tolerated dose of test article to mice was not determined. Instead, the HM10760A was tested for in vivo clastogenicity at 100 times the expected clinical dose exposure (6 μg/kg/weekly). There were no significant changes in ­micronuclei ­frequencies at 150 and 300 μg/kg of HM10760A. However, there was a weak, but statistically significant, increase in the frequencies of MNPCEs in the 600-μg/kg–treated group, when compared to the vehicle group. However, no biological significance was assumed because this value was within the historical control range. No remarkable effects of HM10760A were seen on the ratio of PCE to total erythrocytes, an indicator of cytotoxicity. The positive control group (CPA 70 mg/kg) induced a significant increase in MNPCE (73.17; P