In Vivo Quantitative Measurement of Arthritis Activity Based on Hydrophobically Modified Glycol Chitosan in Inflammatory Arthritis: More Active than Passive Accumulation

Abstract

We demonstrated that arthritis could be visualized noninvasively using hydrophobically modified glycol chitosan nanoparticles labeled with Cy5.5 (HGC-Cy5.5) and an optical imaging system. Activated macrophages expressing Mac-1 molecules effectively phagocytosed HGC-Cy5.5, which formed spherical nanoparticles under physiologic conditions. We estimated the applicability of HGC-Cy5.5 to quantitative analysis of arthritis development and progression. Near-infrared fluorescence images, captured after HGC-Cy5.5 injection in mice with collagen-induced arthritis, showed stronger fluorescence intensity in the active arthritis group than in the nonarthritis group. According to the progression of arthritis in both collagen-induced arthritis and collagen antibody-induced arthritis models, total photon counts (TPCs) increased in parallel with the clinical arthritis index. Quantitative analysis of fluorescence after treatment with methotrexate showed a significant decrease in TPC in a dose-dependent manner. Histologic evaluation confirmed that the mechanism underlying selective accumulation of HGC-Cy5.5 within synovitis tissues included enhanced phagocytosis of the probe by Mac-1-expressing macrophages as well as enhanced permeability through leaky vessels. These results suggest that optical imaging of arthritis using HGC-Cy5.5 can provide an objective measurement of disease activity and, at the same time, therapeutic responses in rheumatoid arthritis.

RHEUMATOID ARTHRITIS (RA) is a chronic inflammatory disease that primarily affects the peripheral joints. It is characterized by inflamed synovial tissues, which are markedly infiltrated by activated macrophages and lymphocytes, and shows a substantial increase in new immature and permeable blood vessels.^1–3 The quantitative measurement of disease activity in patients with RA may provide an important basis for the estimation of arthritis severity, monitoring the response to therapy, and the prediction of prognosis.^4,5 However, current arthritis imaging modalities that are able to portray the anatomic details quite well have weaknesses with regard to the acquisition of physiologic and molecular information and to the estimation of overall activity of arthritis in peripheral joints.^6,7

Recent developments in molecular imaging may provide significant tools that can be used for noninvasive imaging of molecular and cellular processes in living organisms.^8–10 In particular, imaging probes that accentuate dissimilarities between tissues, including the number of capillaries, perfusion, and permeability, have been developed for this purpose.⁸ Recently published studies include optical imaging probes that have been applied for the monitoring of inflammatory cell trafficking, apoptosis, and enzyme activity in murine models of RA.^11–14 These approaches may provide substantially improved capabilities for early diagnosis and monitoring of therapeutic responses, as well as for the investigation of specific biologic processes leading to RA.¹⁵ Nevertheless, there is still a need to project new imaging approaches that incorporate emerging methods and to develop arthritis-specific probes for assessing arthritis activity and therapeutic response.¹¹

Previous studies demonstrated that glycol chitosan modified with hydrophobic bile acid analogues can self-assemble into nanoparticles, which can then be used to visualize tumors noninvasively using optical or nuclear imaging systems.^16,17 Hydrophobically modified glycol chitosan (HGC) nanoparticles preferentially accumulate in tumor tissues via leaky vessels that reveal the pathophysiologic properties of the enhanced permeability and retention (EPR) effect. In a tumor model, in vivo tissue distribution of glycol chitosan nanoparticles shows that the molecular weight of glycol chitosan is one of the important features affecting pharmacokinetics and tumor targeting.¹⁶ Considering a large pool of activated macrophages and the exuberant angiogenesis characterized by a disorganized and permeable architecture within the synovial tissues in RA, we hypothesized that HGC nanoparticles may preferentially accumulate within arthritis joints. In the present study, we sought to investigate whether inflammatory cells take up glycol chitosan nanoparticles and whether inflamed joints can be visualized using HGC nanoparticles and an optical imaging system. We further evaluated the applicability of near-infrared fluorescence (NIRF) imaging with HGC nanoparticles to quantitative analysis of the overall activity of inflammation according to arthritis progression and after treatment with methotrexate (MTX) in murine arthritis models.

Materials and Methods

Materials

Glycol chitosan (molecular weight = 250 kDa, degree of deacetylation = 82.7%), N-hydroxysuccinimide (NHS), 5β-cholanic acid, anhydrous dimethyl sulfoxide (DMSO), MTX, and 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDAC) were purchased from Sigma (St. Louis, MO). The monoreactive hydroxysuccinimide ester of Cy5.5 (Cy5.5-NHS) was obtained from Amersham Bioscience (Piscataway, NJ). Monoclonal antibodies against human CD3 (UCHT-1), CD14 (TK4), and Mac-1 (44) molecules and against mouse CD31 (MEC13.3), F4/80 (CI:A3-1), and Mac-1 (M1/70.15) were used as primary antibodies (Serotec, UK). Secondary antibodies included fluorescein isothiocyanate (FITC)-labeled rat antimouse and rabbit antirat antibodies (Jackson Immunoresearch Laboratory, PA) and biotinylated rabbit antirat antibody (Dako, Denmark).

Preparation of HGC Nanoparticles

Glycol chitosan (250 kDa, 2 g, 8.0 μmol) was degraded in a hydrochloric acid solution (4 N, 200 mL) for 12 hours, as previously described.¹⁶ After the degradation reaction, glycol chitosan with the optimal molecular weight was purified using a simple dialysis method (molecular cutoff = 5 kDa; Spectrum, Rancho Dominquez, CA). Next, the solution was freeze-dried to produce a white powder. The purified glycol chitosan had an average molecular weight of 20 kDa, measured by gel permeation chromatograpy (PSS Hema-Bio300). The glycol chitosan was then chemically modified with hydrophobic 5β-cholanic acid in the presence of EDAC and HOSu, as previously described.¹⁶ Briefly, glycol chitosan (500 mg, 2.0 μmol) was dissolved in DMSO (120 mL), and 5β-cholanic acid (10 mg, 27.8 μmol) was added. The reaction was initiated by adding 41.7 μmol of EDAC and HOSu at 25 °C. After reacting 1 day in the dark condition, the mixture was dialyzed (molecular cutoff = 5 kDa) for 3 days in water/methanol (1:4 v/v) to remove unconjugated 5β-cholanic acid molecules and freeze-dried to produce a white powder. The glycol chitosan-5β-cholanic acid was processed so that it contained 12.8 molecules of 5β-cholanic acid per one glycol chitosan polymer as determined by a colloidal titration method. Finally, the HGC conjugates were directly labeled with NIRF dye, Cy5.5, as follows: Glycol chitosan-5β-cholanic acid conjugates (30 mg, 1.22 mmol) were dissolved in DMSO (20 mL), followed by the addition of NHS of Cy5.5 (2 mg, 1.76 μmol) in 1 mL of DMSO. After the reaction, residual Cy5.5 molecules were removed by dialysis (molecular cutoff = 5 kDa) for 2 days and freeze-dried to produce a blue powder. Under optimal conditions, HGC-Cy5.5 conjugates had 1.1 6 0.4 molecules of Cy5.5 per glycol chitosan-5β-cholanic acid conjugate as determined by measuring the extinction coefficient of Cy5.5 at 675 nm (2.5 × 10⁵ M⁻¹cm⁻¹).¹⁸

To prepare an arthritis-targeting nanoparticle probe, HGC-Cy5.5 conjugate was dissolved in distilled water or phosphate-buffered saline (PBS) and sonicated with a probe-type sonicator (Ultrasonic Processor, GEX-600) three times for 2 minutes each time at 30 W in an ice bath. After the sonication, the product formed a stable nanoparticle structure under aqueous conditions. Importantly, the HGC-Cy5.5 was well dispersed in distilled water and PBS at 37°C and remained stable up to 1 month.

The size and morphology of the HGC-Cy5.5 nanoparticles were determined by dynamic light scattering and transmission electron microscopy, respectively.

Cell Isolation and Culture

Mononuclear cells were isolated using Ficoll gradient centrifugation of peripheral blood or synovial fluid from normal controls and patients with RA who satisfied the 1987 revised criteria of the American Rheumatism Association.¹⁹ Fibroblast-like synoviocytes (FLSs) were isolated by enzymatic dispersion of synovial tissues from RA patients who underwent joint replacement surgery as previously described.²⁰ FLSs recovered between passages 3 and 8 were used for the experiments.

Differentiation and Phagocytosis of Macrophages

Monocytes were enriched by Percoll gradient fractionation of peripheral blood mononuclear cells. Murine monocytes were isolated from the bone marrow of DBA1/J mice. The purity of monocytes, determined by flow cytometry, was higher than 92% after subsequent panning for human and higher than 87% for mouse. Monocytes (1 × 10⁶ cells/mL) were inoculated onto poly-L-lysine-coated coverslips in a 12-well culture plate and cultured overnight. For differentiation into macrophages, cells were cultured in the presence of 50 ng/mL of granulocyte macrophage-colony stimulating factor (GM-CSF; R&D Systems, Minneapolis, MN) for 6 days. Phagocytosis was determined by the cells′ ability to engulf HGC-Cy5.5 (100 μg/mL) for 2 hours. For surface marker analysis, cells were fixed with PBS containing 1% paraformaldehyde at 20°C for 15 minutes, and immunofluorescence staining was performed.

Flow Cytometry Analysis

Cells were incubated with HGC-Cy5.5 at 37°C for 2 hours and washed with PBS containing 0.5% bovine serum albumin. Cells were then labeled with FITC-conjugated primary antibodies at 4°C for 30 minutes and analyzed with a flow cytometer (FACSAria BD, Franklin Lakes, NJ).

Animals and Arthritis Models

All animal care and experimental procedures were performed according to the regulations of Kyungpook National University Institutional Animal Care and Use Committee. The murine collagen-induced arthritis (CIA) model was established as previously described with minor modifications.²¹ Briefly, bovine type II collagen (CII; Chondrex, Redmond, WA) was dissolved at a concentration of 2 mg/mL in 10 mM acetic acid by stirring overnight at 4°C. Male DBA/1J mice obtained from Japan SLC, Inc. (Hamamatsu, Japan) were immunized intradermally at the base of the tail with 100 μg of CII emulsified in 1 mg/mL of Freund's complete adjuvant (Chondrex) at the age of 6 to 8 weeks. On day 21 postimmunization, the mice were given booster injections following the same procedure except that Freund's incomplete adjuvant was used. These mice were monitored daily for clinical symptoms of arthritis, from postimmunization day 22. Murine collagen antibody–induced arthritis (CAIA) was induced by intraperitoneal injection with a mixture of monoclonal antibodies to CII (4 mg/400 μL; Chondrex) in male C57BL/6 mice. Lipopolysaccharide (50 μg/50 μL; Escherichia coli 0111:B4; Chondrex) was injected intraperitoneally on day 3 to synchronize the onset of arthritis. The clinical severity of arthritis in each paw was quantified according to a graded scale from 0 to 4, as previously described.²¹ A mean arthritis score was determined by summing the scores of the four paws of all mice and dividing the result by the total number of mice in each group.

Quantitative Evaluation of Arthritis Activity by In Vivo NIRF Imaging

The clinical arthritis index was determined by the summation of scores from the four paws of each mouse. To evaluate the activity of arthritis quantitatively using HGC-Cy5.5 and optical imaging, CIA mice were classified into four groups according to the development of arthritis: no arthritis (n = 5), 1 week (n = 5), 3 weeks (n = 5), and 5 weeks (n = 5) after the booster immunization. CAIA mice were classified into three groups: no arthritis (n = 5), day 5 (n = 6), and day 10 (n = 5) after collagen antibody injection. After injection with 5 mg/kg of HGC-Cy5.5 into the lateral tail veins, images of the mice were captured by the eXplore Optix system (Advanced Research Technologies Inc., Montreal, QC). The fluorescence intensity in the paws of the mice was calculated by the region of interest (ROI) function of the Analysis Workstation software (Advanced Research Technologies). The total fluorescence intensity of the four paws of each mouse was determined to assess the overall degree of the arthritis. Additionally, fluorescence reflectance images were captured by a 12-bit charge-coupled device (CCD) camera (Image Station 4000 MM, Kodak, New Haven, CT) with a near-infrared emission filter (680–720 nm; Omega Optical).

In Vivo Monitoring of Therapeutic Responses of MTX-Treated Mice

To evaluate the therapeutic response to MTX in the CIA model, mice were distributed randomly into three groups, with one group as the control that received PBS and the other two groups receiving a different dosage of MTX dissolved in PBS (0.5 and 10 mg/kg, intraperitoneally twice a week). At 1 and 5 weeks after the second immunization, five mice in each group received 5 mg/kg of HGC-Cy5.5. Imaging of arthritic mice was performed according to the same protocols described above.

Histologic Evaluation

Immediately after taking NIRF images, the mice were euthanized and joint tissues were prepared for histologic examination. Joint tissues were fixed in 10% phosphate-buffered formalin for 24 hours and subsequently decalcified in 10 mM ethylenediaminetetraacetic acid (EDTA) buffer for 2 weeks, paraffin embedded, and cut into 3 mm thick sections. The histologic inflammation score was determined as described previously after hematoxylineosin staining.²² For immunohistochemical staining, tissue sections were rehydrated and deparaffinized and underwent antigen retrieval. After treatment with 0.3% hydrogen peroxide for 15 minutes to inhibit endogenous peroxidase, sections were incubated with primary and isotype control antibodies at 4°C overnight, followed by incubation with biotinylated secondary antibodies. The reaction was visualized with horseradish peroxidase-labeled streptavidin using a Vectastatin Kit (Vector Laboratories, CA) and a 3,3'-diaminobenzidine tetrahydrochloride (DAB) substrate and counterstained with a hematoxylin solution. For immunofluorescence staining, tissue sections were incubated with primary antibodies and then with FITC-labeled rabbit antirat immunoglobulin, followed by 4',6-diamidino-2-phenylindole (DAPI) staining for nuclear identification and mounting with Vectashield (Vector Laboratories). Slides were viewed using a confocal microscope in fluorescence mode (Leica Microsystems, Germany). Excitation and emission wavelengths were 675 and 695 nm for Cy5.5, 492 and 520 nm for FITC, and 358 and 461 nm for DAPI, respectively. A cooled CCD camera adapted with a bandpass filter was used for image capture. Images of FITC, Cy5.5, and DAPI filters were overlaid to identify cells phagocytosing HGC-Cy5.5. To quantify accumulated HGC-Cy5.5 within synovial tissues, tissue sections were stained with DAPI and viewed with a confocal microscope in phase contrast and fluorescence modes. ROI for the measurement of Cy5.5 intensity were selected after DIC, Cy5.5, and DAPI images were overlaid. The total fluorescence intensity of Cy5.5 from nine fields (magnification, 630×) of each section was calculated using ImageJ software (NIH Image, National Institutes of Health, Bethesda, MD).

Statistical Analysis

Statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL). Differences between the two groups were examined using the Student unpaired t-test, and differences between more than two groups were compared using one-way analysis of variance followed by a Bonferroni test. Simple univariate correlations were calculated based on baseline values using Spearman rank correlation. All p values less than .05 were considered significant.

Results

Characteristics of Cy5.5-Labeled HGC

To prepare an arthritis-homing nanoparticle probe, we synthesized Cy5.5-labeled glycol chitosan nanoparticles that accumulated in the joint (Figure 1). Briefly, each glycol chitosan polymer (20 kDa) was chemically modified with 12.8 molecules of 5β-cholanic acid in the presence of EDAC and HOSu as catalysts. The amphiphilic glycol chitosan–5β-cholanic acid conjugates formed stable nanoparticles, which were well dispersed under aqueous physiologic conditions in vitro (PBS, pH 7.4) at 37°C. For noninvasive optical imaging, the HGC nanoparticles were labeled with 1.1 ± 6 0.4 molecules of the NIRF dye, Cy5.5 (see Figure 1A). Under optimal reaction conditions, the HGC-Cy5.5 had an average diameter of 233 ± 15 nm with a spherical morphology (see Figure 1B). HGC-Cy5.5 as a nanoparticle probe produced a substantial NIRF signal that could be viewed using an optical fluorescence imaging system with a Cy5.5 filter set.

Figure 1.

A, The chemical structure of Cy5.5-labeled hydrophobically modified glycol chitosan (HGC-Cy5.5). B, Transmission electron microscopy image of self-assembled HGC nanoparticles in aqueous conditions.

Enhanced Phagocytosis of HGC-Cy5.5 by Differentiated Mac-1-Expressing Macrophages

To define a subpopulation of cells that may take up HGC-Cy5.5, we first cultured mononuclear cells isolated from peripheral blood and synovial fluid and FLSs isolated from RA synovial tissues with the HGC probe. After a 2-hour incubation, lymphocytes, including T and B cells, did not show cytoplasmic fluorescence (data not shown). FLSs, which are an important class of effector cells comprising hyperplastic synovial tissues of RA, did not take up the HGC probe (data not shown). In contrast, monocytes from synovial fluid, most of which displayed an activated phenotype, phagocytosed the HGC probe more efficiently than monocytes from peripheral blood (Figure 2A). These results led to the hypothesis that HGC-Cy5.5 might be more readily phagocytosed by activated macrophages than by monocytes in vitro. CD14-expressing monocytes isolated from human peripheral blood did not efficiently phagocytose HGC-Cy5.5 after a 2-hour incubation with the probe. However, after monocytes were differentiated into macrophages, surface Mac-1 expression was highly upregulated compared to CD14 expression, and Mac-1-expressing macrophages revealed conspicuous intracellular NIRF signal after incubation with HGC-Cy5.5 (Figure 2B). We then tested whether in vitro differentiation of macrophages also enhanced phagocytosis of the HGC probe in murine monocytes isolated from spleen. Murine macrophages differentiated with GM-CSF stimulation strongly expressed Mac-1 on the surface and phagocytosed HGC-Cy5.5 more efficiently than monocytes (Figure 2, C and D).

Figure 2.

Enhanced phagocytosis of HGC-C5.5 by Mac-1-expressing macrophages. A, Monocytes from peripheral blood (PB) and synovial fluid (SF) of normal controls and patients with rheumatoid arthritis (RA) were incubated with HGC-Cy5.5 for 2 hours and analyzed by flow cytometry. B, Monocytes and differentiated macrophages were incubated with HGC-Cy5.5 (red) for 2 hours at 37°C and stained with primary antibodies against CD14 and Mac-1 followed by FITC-labeled antirat IgG antibody (green). For nuclear identification, cells were stained with DAPI (blue). Images of the different fluorescence channels were merged. C, Phagocytosis of HGC-Cy5.5 by monocytes isolated from bone marrow of DBA/1J mice and differentiated macrophages for 7 days was analyzed by flow cytometry. D, Murine monocytes and differentiated macrophages were cultured with HGC-Cy5.5 (red) for 2 hours and stained with primary antibodies against Mac-1 (green) and DAPI (blue). Original magnification, 630×.

Measurement of Arthritis Activity Using HGC-Cy5.5

In a pilot in vivo study, we confirmed that the total photon count (TPC), which is the summation of photon counts from the ROI of four paws, was increased within 60 minutes of intravenous injection with HGC-Cy5.5. Furthermore, the maximum TPC was maintained up to 2 hours after injection in mice with arthritis, whereas the TPC rapidly decreased after 1 hour of injection in normal mice (Figure 3, A and B). The photon count at 2 hours after injection was significantly correlated with the clinical arthritis score (r² = .91, p < .05).

Figure 3.

Measurement of arthritis activity with HGC-Cy5.5. A and B, In vivo kinetics of the fluorescence intensity in a murine collagen-induced arthritis (CIA) model. A, Color-coded near-infrared fluorescence (NIRF) images obtained at 60, 90, 120, and 150 minutes after injection of HGC-Cy5.5 in normal and active arthritis (56 days after the first immunization) mice were superimposed on white light images of mice from the eXplore Optic system. B, The total photon counts were determined at the region of interest selected on the four paws after intravenous injection of HGC-Cy5.5 (5 mg/kg) in normal (n = 4) and arthritic (n = 6; 56 days after the first immunization) mice. C, NIRF images (upper) obtained 2 hours after injection of HGC-Cy5.5 (5 mg/kg). Color-coded NIRF images (lower) superimposed on white light images of control, active arthritic, and long-standing, deformed (100 days after the first immunization) arthritic mice. D, Quantitative measurement of NIRF intensity from the four paws of normal (n = 4), actively arthritic (n = 6), and long-standing, deformed (n = 3) arthritic mice. Values are the mean and SEM; *p < .05 versus normal group; **p < .01 versus normal group. E, Histologic evaluation of HGC-Cy5.5 distribution from inflamed joint tissues using a confocal microscope. The box in the hematoxylin-eosin (H&E)-stained section indicates the selected area for fluorescence images. Images from Cy5.5 (red) and DAPI (blue) channels were merged (H&E stain, ×200 original magnification; fluorescence microscopy, 630× original magnification. F, Colocalization of Cy5.5 (red) and macrophage markers (green) including F4/80 and Mac-1 within the synovial tissues (×630 original magnification).

To evaluate whether HGC-Cy5.5 can be used to measure the activity of arthritis, CIA mice at different stages were divided into three groups: normal, active arthritis (56 days after the first immunization), and long-standing arthritis (100 days after the first immunization). In raw reflectance images captured 2 hours after HGC-Cy5.5 injection, paws from the active arthritis group showed significantly stronger fluorescence intensity than those from the nonarthritis group. Although paws from the long-standing arthritis group were severely distorted, the fluorescence intensity detected from those paws was lower than what was seen in the active arthritis group (Figure 3C). The fluorescence images obtained by the eXplore Optix system showed that the TPC in the active arthritis group [(10.90 ± 1.33) × 10⁵] was 3.1-fold (p < .01) and 2.3-fold (p < .05) higher than that in the nonarthritis group [(3.34 ± 0.06) × 10⁵] and in the long-standing arthritis group ([4.68 ± 0.30) × 10⁵], respectively (Figure 3D).

Fluorescence microscopy revealed considerable accumulation of HGC-Cy5.5 within the joint tissues of CIA mice. A portion of the probe was also found to localize within the interstitial of the inflamed synovial tissues. In addition, a specific population of cells showed concentrated fluorescence within the cytoplasm (Figure 3E). The intracellular NIRF signal within the synovial tissues of active arthritis predominantly colocalized with Mac-1-positive rather than F4/80-expressing macrophages, of which only a portion showed cytoplasmic NIRF (Figure 3F). These results demonstrated that both rapid phagocytosis of HGC probe by activated macrophages and the EPR effect of inflamed synovial tissues were involved in the imaging of arthritis.

Monitoring of Arthritis Progression in CIA and CAIA Models with NIRF Imaging

To determine whether HGC-Cy5.5 may be useful in monitoring the progression of arthritis in the CIA model, fluorescence images were obtained at different stages of disease progression, specifically 1, 3, and 5 weeks after a booster immunization. TPCs began to increase at 1 week (p < .05 versus the nonarthritis group) and continued to increase until 5 weeks (p < .001 versus the nonarthritis group) after a booster immunization. TPCs were (3.35 ± 0.15) × 10⁵ before the development of arthritis, (3.72 ± 0.10) × 10⁵ at 1 week, (5.64 ± 0.19) × 10⁵ at 3 weeks, and (7.66 ± 0.83) × 10⁵ at 5 weeks after a booster immunization (Figure 4A). The clinical arthritis index of these mice began to rise as early as 1 week and peaked at 5 weeks (Figure 4B), whereas TPC showed only a marginal but significant change at 1 week. Histologic evaluation using fluorescence microscopy revealed that accumulation of HGC-Cy5.5 in the synovial tissues increased as inflammation in the paws progressed (Figure 4C).

Figure 4.

Monitoring inflammation during arthritis progression with HGC-Cy5.5 in murine collagen-induced arthritis (CIA) and collagen antibody-induced arthritis (CAIA) models. A, Near-infrared fluorescence (NIRF) intensity was assessed in the four paws of nonarthritis and CIA mice at 1, 3, and 5 weeks after booster immunization (n = 5 per each group). B, The clinical arthritis index of nonarthritis and CIA mice at 1, 3, and 5 weeks after the second immunization was determined (n = 5 per each group). C, Distribution of HGC-Cy5.5 probe was examined using a confocal microscope in joint tissues from the paws of CIA mice, which were fixed immediately after NIRF imaging with HGC-Cy5.5 probe (red) followed by DAPI staining (blue) (×630 original magnification). D, Color-coded NIRF images were superimposed on white light images of nonarthritis (n = 5) and arthritis mice at 5 (n = 6) and 10 (n = 5) days after collagen antibody injection (4 mg/mouse). E and F, NIRF intensity and clinical arthritis index were assessed in the four paws of CAIA mice at 0, 5, and 10 days after injection with collagen antibody. Values are the mean and SEM; *p < .05 versus normal group; **p < .01 versus normal group; ***p < .001 versus normal group.

We then tested whether NIRF imaging with HGC-Cy5.5 is applicable to other arthritis models using the murine CAIA model, which provided an opportunity to evaluate imaging of acutely developing arthritis where macrophages and neutrophils are the major mediators²³ of inflammation (Figure 4D). Fluorescence intensity began to increase significantly by day 5 [(5.5 ± 1.1) × 10⁵ versus (3.5 ± 0.4) × 10⁵ on day 0, p < .01] and substantially increased by day 10 [(10.3 ± 1.4) × 10⁵ versus (3.5 ± 0.4 × 10⁵ on day 0, p < .001] in parallel with an increase in the clinical arthritis score (Figure 4, E and F).

Evaluation of Therapeutic Responses in MTX-Treated CIA Mice

To examine whether HGC-Cy5.5 could discriminate therapeutic responses, CIA mice were treated with either 0.5 or 10 mg/kg of MTX twice weekly beginning on day 21 after the first immunization, and NIRF imaging was performed 1 and 5 weeks after the second immunization. At the fifth week, the ankles in untreated mice were severely swollen, and intense reflectance fluorescence was observed. Treatment with MTX suppressed joint inflammation in a dose-dependent manner, which was reflected in raw NIRF images. Quantitative analysis of NIRF images showed that TPCs in the MTX-treated group (10 mg/kg) were significantly lower than those in the untreated group after 5 weeks of treatment (Figure 5A). Optical quantification of arthritis severity with HGC-Cy5.5 resulted in a very similar pattern compared to the clinical arthritis index (Figure 5B). We observed accumulation of HGC-Cy5.5 within the synovial tissues and measured the fluorescence intensity of Cy5.5 (Figure 5C). Quantitative analysis of the integrated density of fluorescence in nine different fields per each section obtained from CIA mice revealed that total fluorescence intensity paralleled with the severity of histologic inflammation and optical signal intensity detected in in vivo imaging (Figure 5, D and E). We propose a hypothetical model of selective accumulation of the HGC-Cy5.5 nanoparticle probe within arthritic tissue in Figure 6.

Figure 5.

Quantification of therapeutic response in the murine collagen-induced arthritis (CIA) model using HGC-Cy5.5. CIA mice (n = 5 for each group) were treated with vehicle (□) or 0.5 mg/kg (▨) and 10 mg/kg (▪) of methotrexate (MTX) twice a week after a booster immunization. A, The total photon count and, B, the clinical arthritis index were measured 1 week and 5 weeks after a booster immunization. C, Hematoxylineosin (H&E)-stained histology and fluorescence microscopic images with HGC-Cy5.5 (red) and DAPI staining (blue) were obtained (×200 and ×630 original magnification, respectively). D, Total fluorescence intensity was calculated from nine fields of each section from the paws of CIA mice using ImageJ software. E, Histologic scores of inflammation within the joint tissues were determined as previously described.²² Values are expressed as the mean and SEM; *p < .05 versus the control; **p < .01 versus the control group, and ***p < .001 versus the control group.

Figure 6.

A hypothetical model of selective infiltration of HGC-Cy5.5 nanoparticles into arthritis tissue. EC = endothelial cells.

Discussion

Optical-based imaging technology is a new approach that provides an opportunity for noninvasive monitoring of cellular and molecular events in arthritis.⁸ The present study demonstrated that a lower-molecular-weight HGC probe that forms spherical nanoparticles can be efficiently phagocytosed by activated macrophages expressing Mac-1.

It further showed that a quantitative analysis of joint inflammation with optical imaging using the HGC probe can provide an objective measurement for disease activity in acute and chronic arthritis models, which also can be applied for the monitoring of therapeutic responses in a murine model of RA. Histologic evaluation demonstrated that the accumulation of HGC-Cy5.5 increased in parallel with the severity of joint inflammation and confirmed that the mechanism underlying selective accumulation of HGC-Cy5.5 within synovitis tissues included enhanced phagocytosis of the probe by Mac-1-expressing macrophages and increased permeability through leaky vessels.

The glycol chitosan–based nanoparticles modified with hydrophobic bile acid analogues self-assembled in aqueous conditions with hydrophobic cores of bile acid analogues and hydrophilic shells of glycol chitosan. In vitro experiments revealed that GM-CSF-induced differentiation of macrophages, either from human peripheral blood or murine bone marrow, remarkably facilitated phagocytosis of HGC-Cy5.5 with upregulation of Mac-1 expression on the cell surface, but this was not observed in other subpopulation of cells, including lymphocytes and FLSs. The exact mechanism underlying enhanced phagocytosis of HGC-Cy5.5 by these cells is not known, although it may involve positively charged characteristics of the probe.^17,24 In vivo experiments showed that HGC-Cy5.5 accumulates preferentially within hypervascularized tissues such as tumors with low toxicity and acceptable biodegradability and biocompatibility.^16,18,25,26 In the present study, we demonstrated that HGC-Cy5.5 selectively accumulated within inflamed joints, resulting in higher fluorescence intensities of active arthritis compared to normal and deformed but burned-out joints. A conspicuous similarity in the pattern of quantitative change was found between the fluorescence signals and clinical arthritis index from the paws of arthritic mice. The sensitivity of HGC-Cy5.5-based NIRF imaging was sufficient for discriminating between the different stages of arthritis progression except for the very early stage of arthritis development and to monitor the response to MTX treatment. There have been recently developed approaches that use the fluorescence reflectance imaging system for noninvasive imaging of inflammation-related biologic processes in animal arthritis models, with either near-infrared fluorophore-labeled ligands targeting a specific population of cells such as macrophages or protease-activatable NIRF probes.^11,12,14 By specifically tagging inflammatory cells or selected ligands, optical imaging has been applied to define specific aspects of the inflammatory processes.¹⁵ For the quantitative analysis of inflammation severity, however, compartmental distribution coupled with selective accumulation by a dominant effector mechanism may have advantages over specific targeting of a single molecular or cellular event.

The pathophysiologic events that induce rapid and selective accumulation of HGC-Cy5.5 within inflamed joint tissues may include at least two pathways. Increased angiogenesis begins at an early stage during the development of RA and results in a disorganized and highly permeable vascular architecture, which shares common features with tumor vasculature.^27–30 Compared to the larger intercellular openings of tumor vessels ranging from 200 nm to 2 mm in size, intercellular gaps produced by inflammatory mediators are usually uniform and range up to 0.5 μm.³¹ This is large enough to permit the concentration-dependent passage of spherical HGC-Cy5.5 particles, which have a mean diameter of 233 nm. Transcellular holes of the endothelium in inflamed vessels may also facilitate the transport of HGC-Cy5.5.³² Finally, an imbalance between angiogenesis and lymph angiogenesis within the rheumatoid synovium may play an additional role in accumulation of the probe.³³

Once HGC-Cy5.5 infiltrates the interstitium, activated phagocytes, which are abundant within inflamed synovial tissues, may take up the nanoparticles. Previous studies have shown that macrophages play a pivotal role in RA because the abundance and activation of macrophages in inflamed synovial tissues, especially at the cartilage-pannus junction, significantly correlate with the severity of RA.^34–37 Considering that not all the macrophages within the rheumatoid synovial membrane have activated phenotypes, the severity of inflammation may be more closely reflected by the multitude of activated macrophages that have highly effective phagocytic activity.^36,37 We found that most of the cells with cytoplasmic fluorescence for Cy5.5 expressed Mac-1, whereas only a part of these were positive for F4/80 in the inflamed synovial tissues. In vitro experiments revealed that GM-CSF-induced differentiation of macrophages, either from peripheral blood of human or murine bone marrow, remarkably facilitated phagocytosis of the HGC-Cy5.5 with upregulation of Mac-1 expression on the surface. The exact mechanism underlying the enhanced phagocytosis of HGC-Cy5.5 by these cells is not known, although it may include the positively charged characteristics of the probe.^17,24

Radioactive tracer detection methods including positron emission tomography (PET) and single-photon emission computed tomography (SPECT) offer greater tissue penetration and quantitative integrity but have the main drawbacks of short half-lives of the isotopes and the requirement for larger equipments and facilities.^38,39 The NIRF imaging system used in this study is relatively inexpensive, safe, and easily accessible. HGC-Cy5.5, which has the appropriate physicochemical characteristics for both in vivo and in vitro applications, is easy to prepare and is stable under physiologic conditions for a prolonged period.^16,39 Although tissue penetration of near-infrared light is limited to a few centimeters, most of the peripheral joints that are included for the assessment of disease activity in the clinical settings are located superficially beneath the skin, thus having the privilege to be a preferred candidate for the clinical application of HGC-Cy5.5. The current limitations of optical imaging with NIRF include a low target to background ratio, limited depth of penetration, and surface-weighted image information. The development of new near-infrared fluorophores and technological advancements in photon sources, imaging processes, and detection techniques stand a good chance of overcoming the limitations.^10,40,41

In previous studies, HGC nanoparticles have been used as a drug delivery system for hydrophobic drugs and proved to be therapeutically effective, with much less toxicity in several tumor models.^18,26,42,43 It may be applied to increase the therapeutic index through the targeted delivery of active antirheumatic drugs into the inflamed synovial tissues. It is hoped that the combination of targeted drug delivery and NIRF imaging based on the HGC nanoparticle may provide a new strategy to improve and, at the same time, monitor therapeutic responses in RA.

Footnotes

Acknowledgments

We thank Hyun Min Cho for technical assistance with the in vitro studies.

Financial disclosure of authors: This work was supported by the Intramural Research Program of the KIST, the Seoul R&BD program in Korea, the Global Research Laboratory Project, and a Korea Science and Engineering Foundation grant (No. R01-2007-000-11155-0) funded by the Korean government (MOST).

Financial disclosure of reviewers: None reported.

References

Firestein

. Evolving concepts of rheumatoid arthritis. Nature 2003;423:356–61, doi:10.1038/nature01661.

Koch

Distler

. Vasculopathy and disordered angiogenesis in selected rheumatic diseases: rheumatoid arthritis and systemic sclerosis. Arthritis Res Ther 2007;9 Suppl 2:S3, doi:10.1186/ar2187.

McInnes

Schett

. Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol 2007;7:429–42, doi:10.1038/nri2094.

Welsing

Landewe

van Riel

. The relationship between disease activity and radiologic progression in patients with rheumatoid arthritis: a longitudinal analysis. Arthritis Rheum 2004;50:2082–93, doi:10.1002/art.20350.

Pincus

Sokka

. Quantitative measures for assessing rheumatoid arthritis in clinical trials and clinical care. Best Pract Res Clin Rheumatol 2003;17:753–81, doi:10.1016/S1521-6942(03)00077-9.

Sullivan

. Molecular imaging in oncology. Ann Oncol 2006;17 Suppl 10:x287–92, doi:10.1093/annonc/mdl275.

Weissleder

. Molecular imaging in cancer. Science 2006;312:1168–71, doi:10.1126/science.1125949.

Wunder

Straub

Gay

. Molecular imaging: novel tools in visualizing rheumatoid arthritis. Rheumatology (Oxford) 2005;44:1341–9, doi:10.1093/rheumatology/keh709.

Weissleder

Mahmood

. Molecular imaging. Radiology 2001;219:316–33.

10.

Weissleder

. A clearer vision for in vivo imaging. Nat Biotechnol 2001;19:316–7, doi:10.1038/86684.

11.

Chen

Mahmood

Weissleder

. Arthritis imaging using a near-infrared fluorescence folate-targeted probe. Arthritis Res Ther 2005;7:R310–7, doi:10.1186/ar1483.

12.

Hansch

Frey

Sauner

. In vivo imaging of experimental arthritis with near-infrared fluorescence. Arthritis Rheum 2004;50:961–7, doi:10.1002/art.20112.

13.

Wunder

Schellenberger

Mahmood

. Methotrexate-induced accumulation of fluorescent annexin V in collagen-induced arthritis. Mol Imaging 2005;4:1–6.

14.

Wunder

Tung

Muller-Ladner

. In vivo imaging of protease activity in arthritis: a novel approach for monitoring treatment response. Arthritis Rheum 2004;50:2459–65, doi:10.1002/art.20379.

15.

Biswal

Resnick

Hoffman

. Molecular imaging: integration of molecular imaging into the musculoskeletal imaging practice. Radiology 2007;244:651–71, doi:10.1148/radiol.2443060295.

16.

Park

Kim

Nam

. Effect of polymer molecular weight on the tumor targeting characteristics of self-assembled glycol chitosan nanoparticles. J Control Release 2007;122:305–14, doi:10.1016/j.jconrel.2007.04.009.

17.

Cho

Park

Han

. In vivo tumor targeting and radionuclide imaging with self-assembled nanoparticles: mechanisms, key factors, and their implications. Biomaterials 2007;28:1236–47, doi:10.1016/j.biomaterials.2006.10.002.

18.

Min

Park

Kim

. Hydrophobically modified glycol chitosan nanoparticles-encapsulated camptothecin enhance the drug stability and tumor targeting in cancer therapy. J Control Release 2008;127:208–18, doi:10.1016/j.jconrel.2008.01.013.

19.

Arnett

Edworthy

Bloch

. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24, doi:10.1002/art.1780310302.

20.

Nam

You

. Up-regulated transforming growth factor beta-inducible gene h3 in rheumatoid arthritis mediates adhesion and migration of synoviocytes through alpha v beta3 integrin: regulation by cytokines. Arthritis Rheum 2006;54:2734–44, doi:10.1002/art.22076.

21.

Brownlie

Myers

Wooley

. Treatment of murine collagen-induced arthritis by the stress protein BiP via interleukin-4-producing regulatory T cells: a novel function for an ancient protein. Arthritis Rheum 2006;54:854–63, doi: 10.1002/art.21654.

22.

Guo

Holmdahl

. Contrasting roles of plasminogen deficiency in different rheumatoid arthritis models. Arthritis Rheum 2005;52:2541–8, doi:10.1002/art.21229.

23.

Nandakumar

Svensson

Holmdahl

. Collagen type II-specific monoclonal antibody-induced arthritis in mice: description of the disease and the influence of age, sex, and genes. Am J Pathol 2003;163:1827–37, doi:10.1016/S0002-9440(10)63542-0.

24.

Ishii

Okahata

Sato

. Mechanism of cell transfection with plasmid/chitosan complexes. Biochim Biophys Acta 2001 1514. 51–64.

25.

Park

Hong

Moon

. A new atherosclerotic lesion probe based on hydrophobically modified chitosan nanoparticles functionalized by the atherosclerotic plaque targeted peptides. J Control Release 2008;128:217–23, doi:10.1016/j.jconrel.2008.03.019.

26.

Kim

Park

. Antitumor efficacy of cisplatin-loaded glycol chitosan nanoparticles in tumor-bearing mice. J Control Release 2008;127:41–9, doi:10.1016/j.jconrel.2007.12.014.

27.

Carmeliet

. Angiogenesis in health and disease. Nat Med 2003;9:653–60, doi:10.1038/nm0603-653.

28.

Folkman

. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995;1:27–31, doi:10.1038/nm0195-27.

29.

Paleolog

. Angiogenesis in rheumatoid arthritis. Arthritis Res 2002;4 Suppl 3:S81–90, doi:10.1186/ar575.

30.

Koch

. Review: angiogenesis: implications for rheumatoid arthritis. Arthritis Rheum 1998;41:951–62, doi:10.1002/1529-0131(199806)41:6<951::AID-ART2>3.0.CQ;2-D.

31.

Hashizume

Baluk

Morikawa

. Openings between defective endothelial cells explain tumor vessel leakiness. Am J Pathol 2000;156:1363–80, doi:10.1016/S0002-9440(10)65006-7.

32.

Baluk

Hirata

Thurston

. Endothelial gaps: time course of formation and closure in inflamed venules of rats. Am J Physiol 1997;272:L155–70.

33.

Zhang

Proulx

. Increased lymphangiogenesis in joints of mice with inflammatory arthritis. Arthritis Res Ther 2007;9:R118, doi:10.1186/ar2326.

34.

Kinne

Brauer

Stuhlmuller

. Macrophages in rheumatoid arthritis. Arthritis Res 2000;2:189–202, doi:10.1186/ar86.

35.

Tak

Smeets

Daha

. Analysis of the synovial cell infiltrate in early rheumatoid synovial tissue in relation to local disease activity. Arthritis Rheum 1997;40:217–25, doi:10.1002/art.1780400206.

36.

Mulherin

Fitzgerald

Bresnihan

. Synovial tissue macrophage populations and articular damage in rheumatoid arthritis. Arthritis Rheum 1996;39:115–24, doi:10.1002/art.1780390116.

37.

Kinne

Stuhlmuller

Burmester

. Cells of the synovium in rheumatoid arthritis. Macrophages. Arthritis Res Ther 2007;9:224, doi:10.1186/ar2333.

38.

Dustin

. In vivo imaging approaches in animal models of rheumatoid arthritis. Arthritis Res Ther 2003;5:165–71, doi:10.1186/ar768.

39.

Kim

Kwon

Park

. Physicochemical characterizations of self-assembled nanoparticles of glycol chitosan-deoxycholic acid conjugates. Biomacromolecules 2005;6:1154–8, doi:10.1021/bm049305m.

40.

Ntziachristos

Tung

Bremer

. Fluorescence molecular tomography resolves protease activity in vivo. Nat Med 2002;8:757–60, doi:10.1038/nm729.

41.

Weissleder

Ntziachristos

. Shedding light onto live molecular targets. Nat Med 2003;9:123–8, doi:10.1038/nm0103-123.

42.

Kim

. Hydrophobically modified glycol chitosan nanoparticles as carriers for paclitaxel. J Control Release 2006;111:228–34, doi:10.1016/j.jconrel.2005.12.013.

43.

Kim

Park

. Self-assembled glycol chitosan nanoparticles for the sustained and prolonged delivery of antiangiogenic small peptide drugs in cancer therapy. Biomaterials 2008;29:1920–30, doi:10.1016/j.biomaterials.2007.12.038.