Introduction
Liver cancer, the third leading cause of cancer-related death worldwide, is a primary malignant cancer originating in the liver. There are ~60–70 million liver cancer-related deaths each year, 50% of which occur in China. Most patients are diagnosed at intermediate or advanced stages of the disease (1,2). The early detection and diagnosis of liver cancer is an important issue that must be addressed. One study has shown that detection of liver cancer using magnetic resonance imaging (MRI) has a high degree of accuracy, but there are some limitations in the method in the identification of small lesions, particularly in patients with liver lesions smaller than 1 cm, and in detecting liver cancer accompanied by liver cirrhosis (3). In order to detect small lesions more clearly, enhancement of the contrast agent in the contrast ratio and improvement of the MRI resolution of soft tissues are required. Gadolinium is effective for the imaging of the brain, kidneys and blood system, but is poor in imaging the hepatobiliary system. Therefore, it is necessary to investigate other contrast agents for imaging the liver using MRI. Superparamagnetic iron oxide (SPIO) has wide applications in the biomedical field and can be used not only as a drug-release carrier but also as a targeted probe carrier for monoclonal antibodies, polypeptides, hormones and genes (4–6). SPIO has recently garnered much attention in the field of molecular imaging research (7,8). SPIO is a specific MRI contrast agent for liver imaging and plays an important role in the diagnosis of hepatocellular carcinoma, particularly small hepatocellular carcinoma (9,10). In order to improve the targeting of SPIO and increase the efficiency of MRI in the diagnosis of hepatocellular carcinoma, modification of SPIO is necessary. In the present study, we synthesized SPIO using the chemical coprecipitation method. The nanoparticle SPIO-PEG-CTX was prepared using SPIO as the core, coating it with polyethylene glycol (PEG), and conjugating it with the targeting agent, chlorotoxin (CTX). This nanoparticle was studied using in vitro and in vivo MRI to determine its MR characteristics.
Materials and methods
Statement of ethics
The present study was reviewed and approved by the Second Xiangya Hospital Institutional Review Board. All procedures involving animals were reviewed and approved by the Institutional Animal Care and Use Committee of Second Xiangya Hospital (SYXK 2012-003).
Reagents and animals
Reagents used to prepare the nanoparticle SPIO-PEG-CTX were as follows: iron trichloride hexahydrate (99%), ferrous chloride tetrahydrate (99%) (both from Aladdin Co.), PEG [average molecular weight (Mw), 6,000; Sinopharm Chemical Reagent Co. Ltd.] and CTX (98%; Sigma-Aldrich).
Experimental animals included 8 New Zealand white rabbits of both genders weighing 2.0–3.0 kg provided by the Experimental Animal Center of Second Xiangya Hospital. One tumor-bearing rabbit was provided by Wuhan Union Hospital, Tongji Medical College, Huazhong University of Science and Technology.
MRI and SPIO hydrophobic nanoparticles
MRI was performed by Holland Philips Achieva 3.0T X Series superconducting MRI system and 16-channel SENSE XL Torso coil.
SPIO hydrophobic nanoparticles were prepared using the coprecipitation method. Iron trichloride hexahydrate (1.081 g, 4 mmol) and ferrous chloride (0.397 g, 2 mmol) were mixed in a 250-ml 3-necked flask (Fig. 1). Then, 120 ml deionized water was added into the flask. The flask was sealed, vacuumed of oxygen, and backfilled with nitrogen 3 times to create an anaerobic environment. Next, the whole iron solution system was stirred (mechanical agitation, faster than 1,500 rpm), and 0.2 ml oleic acid was added to the mixed solution drop by drop. Then, 15 ml of ammonium hydroxide (28%) was added slowly to a 3-necked flask and the mixture was heated to 80°C in an ultrasonicator. The solution turned brown-black gradually during the addition of ammonium hydroxide. After 30 min, the SPIO nanoparticles were prepared, which were dried by centrifugation, washed with anhydrous alcohol and vacuum dried at room temperature.
Surface modification of hydrophobic SPIO nanoparticles
Surface modification of hydrophobic SPIO nanoparticles was performed according to the methods of Veiseh et al (11) and Sun et al (12) as described below (Fig. 1). Step 1: PEG was added to a two-necked flask. One neck was connected to a water-separator, which was connected with a condenser tube, and the other neck was connected with a thermometer. Then, toluene was added to the flask to dissolve the PEG. The toluene/water mixture was removed by azeotroping at 110–130°C followed by cooling to room temperature to provide the product. Step 2: Succinic anhydride, the dried PEG6000 and 4-dimethylaminopyridine (DMAP) were stirred in methylene chloride for 24 h. Then, the white solid was precipitated after treatment with ether. The solid was dissolved in isopropyl alcohol. The solid byproduct was removed, and carboxyl-terminated PEG was isolated following column chromatography and condensation. Step 3: PEG acid chloride was prepared by treatment of the carboxyl-terminated PEG with thionyl chloride under a nitrogen atmosphere for 1.5 h. Step 4: Trifluoroethanol was added to the PEG acid chloride and stirred under nitrogen for 2 h. Then, the mixture was heated to reflux for 3 h, and the trifluoromethyl-functionalized PEG was purified using column chromatography. Step 5: Adenosine phosphosulfate (APS) and trifluoromethyl-functionalized PEG in a ratio of 1:2 were added to the solution made of PEG in dichloromethane. The solution was stirred overnight under nitrogen, and the PEG-modified silane was isolated following vacuum drying. Step 6: The PEG-modified silane was added to dry SPIO dispersed in cyclohexane at 50°C under ultrasound conditions to yield the hydrophilic SPIO modified by PEG-silane following removal of the cyclohexane. Step 7: An excess of ethanediamine was added to the SPIO modified by PEG-silane. After a 2-h reaction, the aminated SPIO was isolated by centrifugation following a wash with deionized water. Step 8: The carboxyl-activated iodoacetic acid was prepared by stirring iodoacetic acid, dicyclohexylcarbodiimide (DCC) and N-hydroxysuccinimide (NHS) in ethyl acetate for 3 h. The carboxyl-activated iodoacetic acid dissolved in dimethyl sulfoxide (DMSO) was added to a suspension of SPIO in water. The reaction proceeded at room temperature overnight, and the ethyl iodoacetate-terminated SPIO was prepared following the removal of excess iodoacetic acid by filtration through a semipermeable membrane with a molecular weight limitation of 3,500. Step 9: Thioglycolic acid protected by acetylated sulfhydryl was prepared by reacting thioglycolic acid and acetyl chloride under anhydrous conditions. Carboxy-activated N-succinimidyl-S-acetylthioacetate (SATA) was prepared from thioglycolic acid, DCC and NHS at 0°C. The carboxy-activated SATA was purified and then reacted with CTX (pH=8.5 in bicarbonate buffer solution, 1 μg/μl) for 4 h to yield SATA-CTX. Purification and concentration were performed by filtration through semipermeable membranes of a molecular weight range of 8,000-12,000. Step 10: Hydroxylamine hydrochloride and ethylenediaminetetraacetic acid (EDTA) was added to SATA-CTX, which reacted at room temperature for 1 h. Then, the acetyl group of SATA-CTX was released to expose the reactive sulfhydryl group and the mercapto-modified CTX was formed. The mercapto-modified CTX was purified by filtration through semipermeable membranes with a molecular weight limitation of 3,500. The product reacted with ethyl iodoacetate-terminated SPIO by stirring in an ice-bath overnight to form the final product. The characterizations of the product were analyzed using infrared spectrometer and nuclear magnetic resonance spectrometer.
Establishment of the rabbit VX2 liver carcinoma model
Establishment of the rabbit VX2 liver carcinoma model was performed according to the methods of Chen et al (13), as described below (Fig. 2). The operation was carried out under sterile condition and anaesthesia with chloral hydrate. Tissue at the edge of the tumor was removed from the tumor-bearing rabbit (Fig. 2A). Necrotic tissue, fascia and other connective tissues were removed (Fig. 2B). The remained tissue was cut into small pieces (0.5–1 mm diameter) using scissors (Fig. 2C). A vertical incision was conducted below the xiphoid of healthy rabbits to expose the liver (Fig. 2D). The middle lobe of the liver was exposed using smooth forceps, and a 15-G needle was inserted sideways and upward into the visceral surface of the liver lobe and then 0.5-ml tumor fragments were injected with a 5 ml syringe. During this process, we ensured that the tumor fragments were injected into the liver parenchyma (Fig. 2E). Finally, the puncture site was covered with gelfoam after removal of the needle, and the incision was sutured after confirming no obvious bleeding (Fig. 2F).
MRI of SPIO and SPIO-PEG-CTX in vitro and in vivo
The MRI of SPIO and SPIO-PEG-CTX in vitro and in vivo were performed as follows. All scans were performed on an MR scanner (Philips Achieva 3.0T X Series; Phillips Healthcare, The Netherlands), using a 16-element phased array of SENSE XL Torso coil. The imaging protocol included the following: axial planes sense turbo spin-echo T2-weighted MRI [repetition time (TR)/echo time (TE), 1,565/70 millisecond (msec); NSA=3; flip angle, 90°; matrix size, 252×192; field of view (FOV), 24×24 cm; section thickness: 4 mm, gap 0.4 mm]. Axial planes sense turbo spin-echo T1-weighted MRI (T1WI; TR/TE, 10/2.3 msec; NSA=3, flip angle, 15°; matrix, 252×193; FOV, 24×24 cm; section thickness: 4 mm; gap 0.4).
MRI scanning
MRI scanning was performed using the SPIO and SPIO-PEG-CTX solutions. The arrangements of the SPIO and SPIO-PEG-CTX solutions are shown in Fig. 3. The concentrations of the SPIO and SPIO-PEG-CTX stock solutions were both 10 mg Fe/ml. In samples no. 1-7, the solutions were prepared by adding 200, 100, 50, 40, 30, 20 and 10 μl SPIO stock solution separately and diluting to 2 ml in phosphate-buffered saline (PBS). PBS was used in sample no. 8. In the samples A-G the solutions were prepared by adding 200, 100, 50, 40, 30, 20 and 10 μl SPIO-PEG-CTX stock solution and diluting to 2 ml with PBS. PBS was used in sample H. The concentrations of samples no. 1-8 and A-H were: 1, 0.5, 0.25, 0.2, 0.15, 0.1, 0.05 and 0 mg Fe/ml.
MRI scanning of all 8 living rabbits was performed before and after the injection of the SPIO-PEG-CTX solution. Contrast enhanced T2 weighed images of the VX2 rabbits were acquired 30 min after intravenous bolus injection of 5 ml SPIO-PEG-CTX at 25 μg/ml.
The MR images were sent to a dedicated workstation and analyzed by 3 experienced radiologists. The region of interest (ROI) was 5 × 5 mm for the SPIO and SPIO-PEG-CTX solution or rabbit liver tumor. Partial volume effect and pixels on the margin were not taken into account.
The liver tumors were removed and fixed in formalin after MRI scanning, and the sections were prepared for hematoxylin and eosin (H&E) staining. The morphology and structure of the liver tumors were observed to determine the malignancy.
Data are presented as the mean ± standard deviation and were analyzed using the paired t-test bySPSS 17.0 software. A difference with a P-value of <0.05 was considered to indicate a statistically significant result.
Results
Characterizations of the product
The infrared spectrum of the carboxyl-terminated PEG is shown in Fig. 4. The 3,000 cm−1 broad peak and 1,700 cm−1 double peak verified the success of the carboxyl modification. Nuclear magnetic resonance spectroscopy of PEG-modified silane is shown in Fig. 5. The infrared spectrums of Fe3O4, Fe3O4-PEG-NH2 and Fe3O4-PEG-CTX are shown in Fig. 6.
The MRI results of SPIO and SPIO-PEG-CTX in vitro
The MRI signal values of the SPIO and SPIO-PEG-CTX solutions are shown in Table I. The calibration curves of the concentrations and MRI signal values of the SPIO and SPIO-PEG-CTX solutions are shown in Fig. 7. There was a linear relationship between the concentrations and MRI values of the SPIO and SPIO-PEG-CTX solutions. With increasing SPIO and SPIO-PEG-CTX concentrations, the T1WI value increased, and the T2WI values gradually decreased. The MRI value variations of the two solutions were basically identical. The MR images of the SPIO and SPIO-PEG-CTX solutions are shown in Fig. 8.
The MRI results of SPIO-PEG-CTX in vivo
Thirteen hepatic tumors were dissected from 8 rabbits. The MRI values of the tumors and the adjacent normal hepatic tissues on T2WI are shown in Table II. STumor-0 and SNormal-0, respectively, indicate the T2 signal value of the hepatic tumor and the normal adjacent hepatic tissue before injection of the SPIO-PEG-CTX solution. STumor-1 and SNormal-1, respectively, indicate the T2 signal value of the hepatic tumor and the adjacent normal hepatic tissue after injection of the SPIO-PEG-CTX solution. The data in Table II were analyzed by paired samples t-test using SPSS 17.0 software to evaluate differences between STumor-0 and STumor-1, SNormal-0 and SNormal-1. A P<0.05 was considered to indicate a significant difference. The T2 signal values of VX2 liver tumors and adjacent normal hepatic tissue before and after injection of the SPIO-PEG-CTX solution are shown in Table III. The MR images of a rabbit are shown in Fig. 9. Fig. 9A is the T1WI image. Fig. 9B is the plan-scanning T2WI image. Fig. 9C is the enhanced-scanning T2WI image. The VX2 liver tumor is indicated by an arrow. The tumor demonstrated hypointensity on plan-scanning T2WI, hyperintensity on T2WI, and the T2 signal intensity increased on enhanced-scanning T2WI.
The mean T2 value of the 13 VX2 tumors from 8 rabbits before injection of the SPIO-PEG-CTX solution was 1,346±300.5, and the mean value after injection was 1,161±331.5. Thus, the enhanced-scanning T2 value was smaller than the plan-scanning (P=0.004<0.05). The mean T2 values of the normal hepatic tissues near the 13 tumors before and after injection of SPIO-PEG-CTX were 563.4±67.8 and 480.6±165.1, respectively (P=0.202>0.05). Thus the T2 value of the normal hepatic tissue had no obvious change after injection of SPIO-PEG-CTX.
The histological findings as shown in <xref rid="f10-or-35-05-3059" ref-type="fig">Fig. 10</xref>
Fig. 10A shows the MR image of VX2 liver carcinoma on T1WI before the injection of SPIO-PEG-CTX (indicated by an arrow). The carcinoma was in the front of the liver and demonstrated hypointensity. Fig. 10B shows general sample observation of the VX2 liver carcinoma (indicated by an arrow). There was a circumscribed whitish nodule at the edge of the liver. Fig. 10C is the pathological image of H&E staining at a magnification, x100; and Fig. 10D is the pathological image of H&E staining at a magnification, ×400. The tumor cells were morphologically diverse, with obviously heterogeneous and trachychromatic nuclei; karyokinesis was also observed (indicated by an arrow in Fig. 10D).
Discussion
Hepatocarcinoma is able to be detected by MRI in the clinical. However, there are some difficulties in the detection of small lesions, particularly in patients with liver lesions smaller than 1 cm or liver cancer accompanied by liver cirrhosis (3). The development of molecular imaging provides a new way for the early diagnosis of hepatocarcinoma. SPIO has a broad application in the biomedical field, and currently has become a hot spot in the field of molecular imaging research (7,8).
Compared to gadolinium, SPIO can heighten the proton relaxation rate and shorten the T2 relaxation time. The reticuloendothelial system contains Kupffer cells, which can uptake SPIO and decrease the signal intensity of the liver. Since there are few or no Kupffer cells in the majority of malignant tumors of the liver, the signal intensity of malignant tumors remains unchanged when the liver takes up SPIO. In this case, the contrast in T2WI between the lesion and liver increases, which leads to negative enhancement (6).
The main ingredients of SPIO are Fe3O4 and some γ-Fe2O3 (14). Fe3O4 is black and strongly ferromagnetic, while γ-Fe2O3 is brown and has strong magnetic permeability. There are many methods by which to prepare magnetic nanoparticles, such as mechanical ball mill, coprecipitation, emulsification, sol-gel method, liquid-phase microwave dielectric heating, vapor deposition and hydrothermal method (15,16). Chemical coprecipitation is the most common method, which is simple and inexpensive. Thus, stable and uniform SPIO was prepared using this method in the present study.
Previous studies have demonstrated that matrix metal-loproteinase-2 (MMP-2) is expressed weakly in the liver tissues of chronic hepatitis and liver cirrhosis, but strongly in hepatocellular carcinoma (17,18). The expression of MMP-2 is related to the differentiation, invasion, ability to metastasize and tendency of the tumor to recur in liver cancer (19). Thus, MMP-2 may be a new target that can be combined with a new molecular probe of hepatocarcinoma. CTX is a 36-amino acid peptide containing 1 tyrosine residue, 8 cysteine residues and 4 disulfide bonds. It is purified from the venom of the giant Israeli scorpion (Leiurus quinquestriatus). CTX can combine with MMP-2 selectively and specifically (20), which is strongly expressed in hepatocellular carcinomas. PEG is a cheap and commonly used surface-modification agent with good hydrophilicity and biocompatibility, but without immunogenicity and toxicity (21–24). It is soluble in water and many types of organic solvents and can be bound to nanoparticles by blocking or grafting. The nanoparticles modified with different functional groups display hydrophilicity and stability, can prolong the blood circulation time and can reduce phago-cytosis. Therefore, we combined PEG as medium with SPIO and CTX to prepare SPIO-PEG-CTX nanoparticles. These nanoparticles can target hepatocellular carcinoma through the combination of CTX and MMP-2.
There are many dangling bonds on the surface of SPIO nanoparticles, which can be modified by targeted molecular probes with different functional groups such as carboxyl or amino groups or monoclonal antibodies. This type of modification can make magnetic nanoparticles reach the targeted organization and then amplify an imaging signal and improve image quality, effectively (25,26). Surface modification is defined as changing the physicochemical properties of particles such as surface chemical structure, reaction characteristics, surface hydrophobicity, or surface chemical adsorption by physical or chemical methods. The unmodified nanoparticles have large specific surface area, surface hydrophobicity and static electricity. They can bring about precipitation and combine with plasma protein easily and then be cleared by the reticuloendothelial system in vivo, which may affect biomedical application of the particles. Therefore, SPIO was modified using ligand exchange in the present study. Whether the magnetic performance of SPIO could change after modification was investigated. In the present study, MRI signals of the SPIO and SPIO-PEG-CTX solutions had the same variations in vitro. These results suggest that SPIO-PEG-CTX can be used as a negative MRI contrast agent in T2WI sequence similar to SPIO.
The reticuloendothelial system contains Kupffer cells, which can uptake SPIO and decrease T2 signal intensity of the liver, while liver tumors do not uptake SPIO for the lack of Kupffer cells. In the present study, MR images of VX2 hepatocarcinoma rabbits injected with the SPIO-PEG-CTX (contrast) agent indicated an uptake of SPIO-CTX in the liver tumors which induced a decrease in the T2 signal intensity which was not observed in healthy regions of the liver. This suggests that CTX can alter the targeting of SPIO, and SPIO-PEG-CTX was taken up by the liver tumors, but not by the normal liver tissue. Thus, the SPIO-PEG-CTX nanoparticles have potential for use in the early diagnosis of hepatocellular carcinoma. However, there were some limitations to the present study. The concentration of SPIO-PEG-CTX was not changed as an MR agent, and the time of enhanced-scanning was invariable. Thus, further investigation is warranted.