Abstract
Background/Aim: Interferon-α (IFN-α) is produced to act locally and transiently with a relatively short circulation half-life in vivo. Hybridization of IFN-α with human immunoglobulin Fc, renamed as IFN-α-Fc, may overcome this limitation. In the present study, 131I-IFN-α-Fc and 131I-IFN-α were compared in the aspects of stability, pharmacokinetics, tissue distribution and molecular imaging quality in an animal model. Materials and Methods: Both IFN-α-Fc and IFN-α were labelled with 131I. Biodistributions and pharmacokinetics of both labelled proteins in Sprague–Dawley rats were assayed. Micro-single-photon emission computed tomography/computed tomography was used to non-invasively monitor the longitudinal distribution of both proteins. Results: 131I-IFN-α-Fc was shown to have higher stability than 131I-IFN-α in whole blood, plasma, kidney, liver and stomach from the biodistribution study. The area under curve analyzed from plasma in the phomacokinetics study was 10-fold higher for 131I-IFN-α-Fc than for 131I-IFN-α. At 0-1 h post tail-vein injection, both labelled proteins are mainly accumulated in the kidneys and liver. Notably, 131I-IFN-α-Fc is degraded more slowly than 131I-IFN-α. Conclusion: We demonstrated that 131I-IFN-α-Fc has longer blood circulation time and better biostability than 131I-IFN-α, suggesting the potential application of the immunoglobulin Fc-conjugated cytokine for long-term treatment of diseases.
- 131I-IFN-α-Fc
- biodistribution
- pharmacokinetics
- micro-single-photon-emission computed tomography/computed tomography
Interferon-α (IFN-α) was discovered 50 years ago, and has been shown to have therapeutic value for several malignant, inflammatory, chronic hepatitis C, and other viral diseases, and certain forms of chronic hepatitis B (1). Like most cytokines, IFN-α has a relatively short circulation half-life due to its in vivo requirement for functioning locally and transiently. As such, disease treatment with IFN-α requires frequent dosing and results in increasing the risk of developing neutralising antibody against IFN-α, which may be responsible for a lack of efficacy or relapse (2). Nevertheless, the formulation of IFN-α can be modified to increase its circulation half-life or to prolong its duration of release. Both dosage and administration frequency can, thus, be reduced while increasing efficacy. A recombinant IFN-α–gelatin conjugate has been reported with an extended retention time (3). In addition, pegylation and lipidization are used to modify IFN-α to improve the pharmacokinetics and pharmacodynamics (4, 5). Two pegylated forms of commercialized IFN-α, PEG-INTRON and PEGASYS, show that the prolonged half-life significantly compensates for their reduced in vitro potency (6, 7). Another approach for improving the biological half-life with less frequent dosing is to link IFN-α to the crystallisable fragments (Fc) region of a human immunoglobulin, i.e. IFN-α-Fc. Both IgG and IgM are among the most abundant proteins in human blood, with circulation half-life ranging from several days to weeks. The platform technology of Fc fragment fusion, in which the Fc region of an antibody is genetically linked to an active protein drug, is among the most successful of a new generation of bioengineering strategies (8). Fc biology also suggests the prospect of engineering products that have an extended half-life and which are able to modulate the immune system (9). We have developed an IFN-α-Fc hybrid joined by peptide linkers expected to have a much longer half-life in vivo than native IFN-α.
Pharmacokinetics play an important role in any new drug development program, and can be used to guide the dose selection for phase I studies. The bioactivity of IFN can be assayed in different tissues to evaluate the half-life and pharmacokinetics of IFN-α and IFN-α-Fc in rats. The approach, however, is laborious and subject to the inherent inaccuracy of the bioassay. INF labelled with radioiodine has been shown to improve the study of pharmacokinetics in mice due to its high detection sensitivity (10). Furthermore, radioiodine-labelled proteins are widely used in pre-clinical pharmacokinetic studies (11). Proteins labelled with 131I have been applied for the evaluation of tumor site, pre-surgical staging, immunotherapy, monitoring for recurrence, and response to treatment in the clinic (12).
The single-photon-emission computerized tomography (SPECT)–computerized tomography (CT) dual imaging system provides not only metabolic and functional information, but also anatomical. Micro-SPECT/CT allows for investigation of the biological processes over time, and monitoring of the effects of interventions on disease progression (13).
Alhough the half-lives of IFN-α and IFN-α-Fc have been evaluated in in vitro studies, little information is available on their pharmacokinetics after labelling with 131I. Here the stability, pharmacokinetics, tissue distribution, and micro-SPECT/CT imaging of 131I-labelled IFN-α and IFN-α-Fc hybrid protein in rats were compared.
Materials and Methods
Animals. Sprague–Dawley rats (8 weeks old, 320-390 g) purchased from the Laboratory Animal Center of National Yang-Ming University (Taipei, Taiwan, ROC) were adapted for 7 days before the experiments. The standard diet (Lab diet; PMI Feeds, St. Louis, MO, USA) and sterilised water were provided ad libitum. The animals were maintained under the following conditions: 25±2°C, 55-60% humidity, 12-h light/dark cycle. All animal experiments were approved by the Institutional Animal Care and Use Committee of the Institute of Nuclear Energy Research, Taoyuan, Taiwan, ROC (The approval number is 96030).
Preparation of IFN-α and IFN-α-Fc. IFN-α and IFN-α-Fc were provided by the Development Center for Biotechnology of Taiwan. The preparation of IFN-α and IFN-α-Fc were as follows: total ribonucleic acid (RNA) was extracted from KG-1 human myeloid leukaemia cell line by using an RNAzol RT kit (Molecular Research Center, Inc., Cincinnati, OH, USA). The first-strand complementary deoxyribonucleic acid (cDNA) was synthesised by reverse transcription, using AMV reverse transcriptase with oligo(dT) as the 3’ primer. The reaction mixture was used directly as the template for polymerase chain reaction (PCR) to amplify IFN-α cDNA. The 5’ primer for PCR includes a HindIII site and the coding sequence of the first 21 amino acids from the IFN-α2a leader peptide. The 3’ primer includes the coding sequence of part of the linker, the last 5 amino acids of IFN-α2a, and a BamHI site integrated in the linker sequence. The cDNA of human immunoglobulin γ4 Fc was obtained by reverse transcription. The RNA was extracted from human tonsil B-cells. After sequencing and ligation of the two PCR-amplified DNA segments, full-length IFN-α-Fc cDNA was inserted into the mammalian expression vector pCDNA3 (Invitrogen, Carlsbad, CA, USA). NSO mouse myeloma cells (1×107) were mixed with 10 μg of linearised pCDNA3/IFN-α-Fc plasmid in 0.8 ml phosphate-buffered saline, and kept on ice for 5 min. After electroporation and colony selection, the clone with the highest secretion level was expanded and adapted to grow in a spinner. For large-scale preparation, the culture supernatant was collected and passed through a protein A agarose column (Sigma-Aldrich, St. Louis, MO, USA) equilibrated with phosphate-buffered saline. The protein bound to protein A was eluted with 50 mM citric acid (pH 3.0), then concentrated by lyophilisation. The purity of the recombinant protein isolated from the NSO culture medium was examined by sodium dodecyl sulphate-polyacrylamide gel electrophoresis and western blotting. The molecular weight of this protein is about 55 kDa and 110 kDa under reducing and non-reducing conditions, respectively, which are the predicted sizes of the IFN-α-Fc hybrid. The recombinant protein was recognized with anti-Fc and anti-IFN-α antibodies.
The bioactivity of the recombinant protein was determined by an antiviral assay. Human lung carcinoma cells (A549, ATCC number CCL-185) (ATCC, Manassas, VA, USA) were seeded in 96-well plates at a density of 40,000 cells/well and incubated at 37°C for 24 h. Serial diluted IFN-α-Fc or native IFN-α (NIH reference: Gxa01-901-535) (NIH, Bethesda, Maryland, USA.) was added and incubated at 37°C for 24 h. Every sample was processed in triplicate. The culture medium was replaced with fresh medium containing encephalomyocarditis virus (ATCC number VR-129B) (ATCC, Manassas, VA, USA) at about 0.1 multiplicity of infection/cell, and incubated at 37°C for another 48 h. The attached cells were fixed in 4% formaldehyde followed by Giemsa staining. The stained cells were dissolved by using methanol, and the optical density of the samples was measured at 595 nm by spectrophotometry.
Radiolabelling. IFN-α and IFN-α-Fc were labelled with 131I by using the Iodogen (1,3,4,6-tetrachloro-3α,6α-diphenylglycoluril) method (14). In brief, Iodogen (100 μg)-precoated tubes (Pierce, Rockford, IL, USA) were rinsed with 200 μl Tris iodination buffer (25 mM Tris-HCl, pH 7.5, and 0.4 M NaCl), then decanted. 37-224 MBq of Na[131I] (PerkinElmer, Boston, Massachusetts, USA) in 100 μl NaOH (0.1 M) Tris iodination buffer was added to each tube. The reaction was allowed to proceed for 6 min at room temperature with concomitant mixing every 30 s, followed by the addition of 100 μg of IFN-α or IFN-α-Fc in 50 μl Tris iodination buffer for 6-9 min at room temperature for 30 s. The samples were transferred to new tubes to stop the reaction. The radiochemical yield was determined by radio-thin-layer chromatography (radio-TLC). Radio-TLC was performed by spotting the samples on instant TLC silica gel strips (Gelman Sciences, Inc., Ann Arbor, MI, USA), and developing the strips in 85% methanol as the mobile phase, they were then scanned with an imaging scanner for the calculation of labelling efficiency (Bioscan, Inc., Washington, DC, USA).
Analyses of tissue distribution and excretion. Male rats (5 per group) were intravenously (i.v.) injected with 0.2 ml of 131I-IFN-α or 131I-IFN-α-Fc (25 μCi, 25 μg). At intervals (5 min, 30 min and 1, 3, 5, 24, 48 and 72 h), the animals were sacrificed by CO2 asphyxiation. The organs of interest were removed and weighed, the radioactivity was measured by the total radioactivity assay (RA) (15). The decay correction of all radioactivity data was based on the day of the injection (day 0). The percentage of injected dose per gram (%ID/g) and percentage of injected dose (%ID) were calculated by comparison with the standards representing the injected dose per animal. The results were expressed as the mean±standard error (S.E.) of the mean. Ten conscious male rats were used to evaluate excretion. The rats were kept in metabolic cages individually up to 84 h post 131I-IFN-α and 131I-IFN-α-Fc injection, respectively. Food and water were adequately provided. Urine and faeces were collected continuously in each metabolic cage and counted with a gamma-counter.
Trichloroacetic acid (TCA) precipitation and gamma counting. Tissue samples (plasma, liver, kidney and stomach) used for the biodistribution study were extracted by precipitation (TCA-RA) (16) using a Polytron homogeniser (Kinematica, Inc., Bohemia, NY, USA). Organ homogenates were dissolved in 0.1 M phosphate buffer at pH 7.2 (500 mg liver/ml, 300 mg kidney or stomach/ml). Organ extracts were obtained by low-speed (868×g) and ultra-high-speed (160,000 × g) centrifugation at 4°C for 10 min. Tissue extracts (50 μl) were placed in 12×75 mm polypropylene tubes containing 10% TCA (4 ml). After incubation on ice for 15 min, the extracts were filtered through a 0.45 μm membrane (Millipore, Billerica, MA, USA). The filtrates were counted with a gamma counter. The percentage of protein-bound radioactivity was calculated using the following formula:
The distributions in the plasma, liver, kidney and stomach, respectively, were calculated as %ID/g by the TCA-RA.
Pharmacokinetic analysis. Blood samples (300-500 μl) from 10 rats were collected by heart puncture at 5 min, 30 min and 1, 3, 5, 24, 48 and 72 h post i.v. injection of 131I-IFN-α and 131I-IFN-α-Fc, respectively. The collecting tubes were rinsed with 200 IU/ml heparin (Sigma Chemical, St. Louis, MI, USA), and the blood samples were centrifuged at 868 × g to acquire the plasma. The radioactivity of 131I-IFN-α and 131I-IFN-α-Fc in the plasma was normalized to %ID/g of tissue. The data were fitted to a non-compartment model (17), and the pharmacokinetic parameters were derived by using WinNonlin 5.0 software (Pharsight Corporation, Mountain View, CA, USA).
Micro-SPECT/CT imaging analysis. Rats were i.v. injected with 0.2 ml of 131I-IFN-α or 131I-IFN-α-Fc (25 μg total protein/200 μl/rat) with high specific activity (500 μCi/200 μl). The rats were anaesthetised with 2% isoflurane and imaged by micro-SPECT/CT (fusion X-SPECT/CT dual modality animal imaging system; Gamma Medica, Northridge, CA, USA) at 0, 24, 48 and 72 h post injection as previously described (18). The SPECT images were acquired by using a parallel-hole collimator. The radius of rotation (ROR) was 1 cm, and the field of view (FOV) was 1.37 cm. Image acquisition was accomplished by using 64 projections at 120 s per projection. The SPECT imaging was followed by CT acquisition. The operating current and voltage of the X-ray tube were set at 0.5 mA and 50 kV, respectively. Each scan had 256 projections with a 0.5 s exposure per projection. The software for CT, SPECT and SPECT/CT image fusion was COBRA (Exxim Computing Corporation, Pleasanton, CA, USA), LumaGEM (Gamma Medica) and IDL 6.0, respectively. The SPECT images were reconstructed to sizes of 56×56×56 voxels with a resolution of 0.2 mm. The CT images were reconstructed to sizes of 512×512×512 voxels with a resolution of 0.3 mm.
Results
Radiolabelling efficiency. The labelling efficiencies of 131I-IFN-α and 131I-IFN-α-Fc were 96.6±0.8% and 97.9±0.1%, respectively (n=5).
Tissue distribution and excretion. The distributions of 131I-IFN-α and 131I-IFN-α-Fc in various tissues are demonstrated in Tables I and II. The radioactivity of 131I-IFN-α at 5 min and 72 h was 2.24±0.22 and 0.03±0.00 %ID/g, respectively, in whole blood versus 3.91±0.39 and 0.05±0.00 %ID/g, respectively, in the plasma. Notably, the radioactivity of 131I-IFN-α-Fc at 5 min and 72 h was 4.35±0.14 and 0.24±0.01 %ID/g, respectively, in whole blood versus 7.80±0.43 and 0.49±0.02 %ID/g, respectively, in the plasma. 131I-IFN-α was almost completely excluded at 72 h after injection. The radioactivity of 131I-IFN-α reached the plateau in the kidneys at 5 min after injection (10.5±0.67 %ID/g), whereas the distribution of 131I-IFN-α-Fc in the kidneys was 0.67±0.04 %ID/g at the same time point.
The radioactivity of 131I-IFN-α and 131I-IFN-α-Fc in the faeces at 84 h after injection was 2.32±0.06 and 2.52±0.31 %ID, respectively, as shown in Figure 1. The radioactivity of 131I-IFN-α and 131I-IFN-α-Fc in the urine was 70.4±3.31 and 54.0±2.00 %ID, respectively, at 84 h after injection. Therefore, both 131I-IFN-α and 131I-IFN-α-Fc were excreted mainly via the urinary system.
Stability. The stability of 131I-IFN-α and 131I-IFN-α-Fc in the rats as determined by the TCA-RA is shown in Table III. 131I-IFN-α was undetectable 5 h after the injection. Notably, 98% of 131I-IFN-α-Fc was still detectable up to 72 h post-injection. These results show that 131I-IFN-α-Fc is more stable than 131I-IFN-α in the plasma, kidneys, liver, and stomach. Table IV shows the distributions of 131I-IFN-α and 131I-IFN-α-Fc in the plasma, kidneys, liver and stomach, as determined by the RA and TCA-RA. The distributions of these proteins were similar at 5 min post-injection, and were only demonstrated by the TCA-RA at later time points. In addition, only low radioactivity of 131I-IFN-α was detected by the TCA-RA. These results further confirm the better stability of 131I-IFN-α-Fc than that of 131I-IFN-α.
Radioactivity of 131I-radiolabelled interferon-α (131I-IFN-α) in rat organs evaluated by total radioactivity assay (RA). The data are expressed as the percentage of injected dose per gram (%ID/g).
Radioactivity of 131I-radiolabelled interferon-α/human immunoglobulin Fc fusion protein (131I-IFN-α-Fc) in rat organs evaluated by total radioactivity assay (RA). The data are expressed as the percentage of injected dose per gram (%ID/g).
Pharmacokinetics. A non-compartmental data analysis was conducted to determine the parameters of pharmacokinetics as shown in Table V. The time to peak concentration (Tmax) was 5 min post-injection for both 131I-IFN-α and 131I-IFN-α-Fc, and the maximum concentrations (Cmax) at 5 min were 3.91 and 7.80 %ID/g, respectively. The area under the plasma concentration-versus-time curve (AUC)0→72 h of 131I-IFN-α and 131I-IFNα-Fc was 12.1 and 111.6 min×%ID/g, respectively. In addition, the AUC0→INF of 131I-IFN-α and 131I-IFNα-Fc was 15.7 and 132.4 min × %ID/g, respectively.
Micro-SPECT/CT imaging. The coronal images of the abdomen and leg regions by micro-SPECT/CT are shown in Figure 2. These images indicate the accumulation of 131I-IFN-α in the kidneys is at 0-1 h post-injection, whereas 131I-IFN-α-Fc is dispersed mainly in the liver. In addition, the radioactivity of 131I-IFN-α is undetectable at 48 h post-injection, whereas 131I-IFN-α-Fc still could be visualised at 72 h post-injection. The results suggest that the degradation of 131I-IFN-α-Fc is slower than that of 131I-IFN-α.
Excreted amounts from faeces (A) and urine (B) after intravenous injection of 25 μg 131I-radiolabelled interferon-α (131I-IFN-α) and 131I-radiolabelled interferon-α–human immunoglobulin Fc fusion protein (131I-IFN-α-Fc) to rats. The data represent the mean±standard error of the mean (n=5).
Radioactivity of 131I-radiolabelled interferon-α (131I-IFN-α) and 131I-radiolabelled interferon-α/human immunoglobulin Fc fusion protein (131I-IFN-α-Fc) in organ extracts evaluated by radioactivity assay after precipitation with 10% trichloroacetic acid (TCA-RA). The data are expressed as percentage (%).
Discussion
Cytokines are generally small proteins with relatively short half-lives, and dissipate rapidly among various tissues and organs. Small quantities of some cytokines may cross the blood–brain barrier and enter the central nervous system, thereby causing severe neurological toxicities (19, 20).
IFN-α was among the first cytokines to be produced by using recombinant DNA technology (21). Many pharmacokinetic studies show that IFN-α has a short half-life in most animals and humans (22–24). On the contrary, IFN-α-Fc is expected to have a much longer half-life in vivo than native IFN-α. In the present study, the AUC of IFN-α-Fc is about 10-fold larger than that of IFN, and shows the higher uptake in the plasma post injection. The radioactivity of 131I-IFN-α and 131I-IFN-α-Fc in the urine was 70.4±3.31 and 54.0±2.00 %ID, respectively, at 84 h post-injection (Figure 1), suggesting that both proteins are mainly excreted via the urine. To our knowledge, this may be the first study to evaluate the pharmacokinetics of IFN-α and IFN-α-Fc labelled with 131I.
Micro-single-photon emission computed tomography/computed tomography (Micro-SPECT/CT) fused coronal images after i.v. injection of 131I-radiolabelled interferon-α (131I-IFN-α) (A) and 131I-radiolabelled interferon-α–human immunoglobulin Fc fusion protein (131I-IFN-α-Fc) (B) to the rats. The fused abdominal and leg region scans were performed at 0, 24, 48 and 72 h after i.v. injection of 25 μg 131I-IFN-α and 131I-IFN-α-Fc via tail veins. The white, green and yellow arrows indicate the kidney, stomach and liver, respectively. The highest and lowest amounts of radioactivity are indicated in white and black, respectively. Similar results were obtained from three different rats.
Longitudinal monitoring of radiolabelled proteins in vivo can be achieved with multimodalities of molecular imaging. These non-invasive strategies not only reduce the time required for imaging, but enable the spatial and temporal monitoring of tissue distribution and excretion of novel drugs (25, 26). Several studies have demonstrated that the kidney is the main site for the degradation of IFN (27-29). Since the liver is the primary site for removal of soluble immune complexes, and the Kupffer cells are rich with Fc receptors, IFN-α conjugated with an immunoglobulin Fc region, i.e. IFN-α-Fc, will target the liver tissues more efficiently than IFN-α (30). This inference is confirmed by micro-SPECT/CT imaging, which shows that the main distribution sites of 131I-IFN-α and 131I-IFN-α-Fc are indeed the kidneys and liver, respectively (Figure 2).
Although the total radioactivity can be determined by the RA, the decomposed products of 131I-IFN-α and 131I-IFN-α-Fc cannot be distinguished. The TCA-RA, on the other hand, separates 131I and low-molecular-weight decomposed products from the parent proteins, and may reflect the tissue-distribution characteristics of 131I-IFN-α and 131I-IFN-α-Fc compared to that of the RA. As such, the TCA-RA should be more optimal for indicating the iodinated protein deposition. In the present study, a low percentage of TCA-precipitated radioactivity of 131I-IFN-α in stomach extracts (Table III) suggests that more 131I or low-molecular-weight metabolites are retained in this organ. On the contrary, more than 98% of TCA-precipitated radioactivity of 131I-IFN-α in the plasma extracts at 72 h post-injection were found, suggesting that IFNs may have more complicated biological functions through the circulation.
Radioactivity of 131I-radiolabelled interferon-α (131I-IFN-α) and 131I-radiolabelled interferon-α/human immunoglobulin Fc fusion protein (131I-IFN-α-Fc) in rat organs assessed by total radioactivity assay (RA) and radioactivity assay after precipitation with 10% trichloroacetic acid (TCA-RA). The results are expressed as the percentage of injected dose per gram (%ID/g).
Conclusion
IFN-α-Fc is chemically more stable, with a longer circulation than IFN-α, suggesting its potential application for the long-term treatment of systemic diseases. The modified TCA-RA seems to be a more optimal method for determining the iodinated protein deposition. SPECT/CT molecular imaging might be useful for non-invasive theranostic monitoring for radiolabelled IFN therapy.
Pharmacokinetic parameters measured after intravenous injection of 131I-radiolabelled interferon-α (131I-IFN-α) and 131I-radiolabelled interferon-α–human immunoglobulin Fc fusion protein (131I-IFN-α-Fc). Each value represents the mean of 10 rats.
Acknowledgements
This work was financially supported by the Technology Development Program (No. 102-EC-17-A-02-04-0499) of Ministry of Economic Affairs of Taiwan. The Authors thank CH Yeh, LC Chen, WC Lee, YH Wu, CL Ho, CY Yu, YJ Chang, and TJ Chang for their technical assistance.
Footnotes
Conflicts of Interest
None of the Authors have any conflicts of interest with regards to funding or support of any kind of this study.
- Received April 15, 2015.
- Revision received May 17, 2015.
- Accepted May 19, 2015.
- Copyright © 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved