动物造模 中国实验动物信息网 2021-11-29 1258

　　Introduction: The blood-brain barrier (BBB) is the most significant obstacle to the targeted delivery of therapeutic drugs to the central nervous system (CNS). For drugs that interact with targets in the central nervous system, it is necessary to obtain an accurate kinetic relationship between the central nervous system and systemic circulation levels. It is also important to understand the factors that influence this relationship in order to plan central nervous system drug levels based on their systemic concentration-time distribution. In practice, cerebrospinal fluid (CSF) is the most accessible and widely used sampling method to measure the levels of drugs and biomarkers in the central nervous system. CSF sampling is inconvenient, so strict research design is required to minimize the number of time points that need to be collected. Some excellent reviews highlight the scientific principles and practicality of using CSF as a substitute for assessing CNS drug exposure levels. Although there are many studies on the rate of small molecule drugs crossing the blood-brain barrier (BBB) and blood-cerebrospinal fluid barrier (BCSFB). Few in-depth studies have been conducted on the kinetics of the influx of antibodies into and out of CSF, especially regarding the key factors that control CSF antibody exposure. Consider the time required for the antibody to reach the maximum CSF drug concentration (CSF T max), the CSF half-life of the antibody (CSF T) and the steady-state antibody level between CSF and the systemic circulation. We report the dynamic relationship between the serum concentration-time distribution of CSF and five humanized monoclonal antibodies after systemic administration in rats and cynomolgus monkeys, and the factors that affect CSF antibody exposure. These findings have important implications for evaluating the pharmacokinetics of antibodies in the human central nervous system.

　　Materials: In this study, we tested five human monoclonal antibodies (monoclonal antibodies). All five antibodies are highly selective and have nanomolar affinities for their respective targets. Antibody A (BIb054) preferentially binds α-synuclein, and binds human, non-human primate (NHP) and rat α-synuclein. Antibody B (399 H0/L0) recognizes the JC virus, which is present in humans but not in NHP or rats. Antibody C (BIb076) and antibody D (40E8) bind to the microtubule binding protein tau. . Antibody C recognizes the linear sequence within tau and binds to human and NHP, but does not bind to rat tau. Antibody D is specific for phosphorylated tau protein and recognizes phosphorylated human, NHP and rat tau. Antibody E (BiB033, OpICunMUB) binds to the NoGo receptor interacting protein 1 (LIGO-1) of the LRR and Ig domains, and has similar affinity to human, NHP and rat LIGO-1. Antibodies A, C, D and E are fully human antibodies, and antibody B is a humanized rabbit antibody. Antibodies A, B, C, and D contain wild-type human IgG1-Fc, while antibody E contains IgG 1-FC designed to reduce effector function. Antibody A (BIb054), Antibody C (BIb076) and Antibody E (BIb033) are currently being evaluated in clinical trials.

　　Method: Rat PK study. Adult male SD rats (250-300g) are kept at constant temperature (22±2°C) and relative humidity (50～70%), under a regular light/dark schedule (light, 7:00 to 7:00 am ). Free access to food and water. The route of administration is tail vein (mAb A, B, C, D) or intraperitoneal injection (MAB E). At different time points after administration, the animals were sacrificed by CO2 asphyxiation, and CSF was immediately collected by cisterna and blood was collected by cardiac puncture. The rats used to evaluate the E antibody were adult male brown Norway rats (average weight 150-200 grams). For brain and spinal cord collection, immediately after CSF/blood collection, 15 ml of normal saline is used for perfusion and inserted into the left ventricle through a peristaltic pump. Place the whole blood sample on the workbench for about 15-30 minutes and rotate it at 3000 revolutions/minute (about 1000×g) for 10 minutes to separate the serum. All samples are stored at -80°C until analysis. Juvenile cynomolgus monkeys (2.5-4 kg) were surgically implanted with a single lumbar intrathecal catheter (advanced into the thoracic area) approximately 1 week before the administration to allow repeated CSF sampling. Before administration, CSF was collected from the catheter for baseline measurement. The intravenous (IV) dose is administered within 1 to 2 minutes. At pre-selected sampling time points, blood (1 mL) and cerebrospinal fluid (0.5 mL) were collected from restricted conscious animals. Serum and CSF samples were prepared as described above for rat studies. Contaminated samples are excluded. Enzyme-linked immunosorbent assay (ELISA) was used to determine the concentration of antibody A-E in serum and CNS specimens. The optimized sandwich ELISA optimizes the detection of each antibody. Because the capture and detection reagents are human IgG specific rather than target specific. This test can be used to quantitatively modify human or humanized IgG molecules in non-human biological matrices (including NHP and rat samples). Determine the total monoclonal antibody concentration, including unbound and soluble ligand-bound antibodies. Briefly, the antibody A-D samples were analyzed, a 96-well ELISA plate was coated with capture antibody, and the monkey adsorbed goat anti-human IgG, 1 μg/ml. Wash after blocking and add standard (STD), control (QCS) or research samples to the plate. Wash the plate again and add the enzyme-conjugated detection antibody (goat anti-human IgG HRP) diluted 1:7500 (50 ng/ml) to the wells. After washing, add HRP substrate (TMB). Add 1N sulfuric acid to stop the reaction and read at 450nm. The average ELISA value is between 16～4400 ng/ml. The lower limit of quantification for CSF is 16 ng/ml, and for serum samples it is 50 ng/ml.

　　Result: The pharmacokinetic properties of five human antibodies in serum and cerebrospinal fluid were evaluated in rats and NHPS. The two-compartment first-level pharmacokinetic model conforms to the concentration-time curves of all five antibodies. After intravenous administration in rats, serum and cerebrospinal fluid concentration-time distribution of antibodies A (10 mg/kg), B (10 mg/kg), C (20 mg/kg) and D (20 mg/kg). The serum concentration profile showed a multi-exponential decline. All 4 antibodies appeared slowly in CSF, and CSF T max was between 6 and 48 h. The concentration-time curves of serum and cerebrospinal fluid reached parallelism after CSF Tmax. The cerebrospinal fluid antibody concentration is about 1000 times lower than the serum concentration. Intraperitoneal injection (ip) of antibody E to brown Norway rats, respectively, to draw the concentration-time curves of cerebrospinal fluid, serum and CNS tissue of brain and spinal cord. After IP (30 mg/kg) administration, serum T max is reached 6 to 24 hours after administration. The cerebrospinal fluid T max, brain T maximum and spinal cord maximum T value are reached 1 to 2 days after serum T max. The concentration-time curves of serum and cerebrospinal fluid were also parallel to CSF T max after IP administration. The concentration-time curve in the cerebrospinal fluid (antibody A and E) and spinal cord (antibody E) are parallel to the time curve in CSF, although the antibody concentration in these central nervous system tissues is slightly higher than the antibody concentration in the cerebrospinal fluid. No evidence of target-mediated clearance was observed, which is consistent with the lack of cross-reactivity of antibodies to rodent targets (antibodies B and C) or low levels of antibodies in the periphery (antibodies A, D, and E). As observed in rats, the serum concentration distribution of each antibody in NHP showed a multi-exponential decrease. Antibodies gradually appear in CSF, reaching a peak 24-96 hours after IV administration. The concentration-time curves of serum and cerebrospinal fluid paralleled after CSF Tmax, and the CSF antibody concentration was about 1000 times lower than the serum concentration on average. The serum concentration is proportional to the antibody dose, and the CSF concentration is proportional to the antibody serum concentration. The pharmacokinetic characteristics of antibodies A, C and D are similar. Although the subject-to-subject changes in the serum concentration-time distribution of all three antibodies are relatively small, the ratio of the cerebrospinal fluid concentration to the serum concentration varies approximately tenfold. To better understand the variability in CSF measurement, we analyzed the serum and CSF concentration-time curves of individual animals from three NHP studies. The change in the ratio of CSF to serum is random and occurs in all studies, regardless of the expected target and dose intensity. These five antibodies showed high serum exposure (SuruMAuC) and long half-life (T-) in serum and CNS tissues. After IV administration, CSF gradually appeared antibodies, reaching the maximum concentration of CSF in rats at 16 to 48 hours, and reaching the maximum concentration of CSF in cynomolgus monkeys in 2 to 3 days. For all five antibodies, the serum and cerebrospinal fluid concentration-time curves became parallel 2 to 3 days after dosing, and the CSF concentration was roughly a thousand times lower than the corresponding serum concentration, regardless of the dose.

　　Conclusion: From the pharmacokinetic analysis of five kinds of human antibodies in rats and monkeys, we found that the antibodies slowly enter the CNS about 24 to 72 hours after IV administration, and the Tmax of cerebrospinal fluid is about 24 to about 24 hours after IV administration. At 72 hours, the CSF and serum concentration-time curves were parallel after CSF Tmax, and the average ratio of CSF to serum was 0.1-0.2%. Before CSF Tmax is obtained, the ratio of CSF to serum is not an accurate indicator of CNS uptake or the ratio of steady-state CSF to systemic concentration. These studies contribute to the design of future studies.