Published online before print August 14, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,,1
* Center for Neurovirology and Neurodegenerative Disorders, Departments of
Pharmacology and Experimental Neuroscience and
Radiology, University of Nebraska Medical Center, Omaha, Nebraska, USA;
School of Pharmacy and Health Professions, Creighton University, Omaha, Nebraska, USA; and
¶ Baxter Healthcare Corporation, Round Lake, Illinois, USA
1 Correspondence: Department of Pharmacology and Experimental Neuroscience, University of Nebraska Medical Center, 985880 Nebraska Medical Center, Omaha, NE 68198-5880. E-mail: hegendel{at}unmc.edu
ABSTRACT
We posit that the same mononuclear phagocytes (MP) [bone marrow (BM) and blood monocytes, tissue macrophages, microglia, and dendritic cells] which serve as targets, reservoirs, and vehicles for HIV dissemination, can be used as vehicles for antiretroviral therapy (ART). Toward this end, BM macrophages (BMM) were used as carriers for nanoparticle-formulated indinavir (NP-IDV), and the cell distribution was monitored by single photon emission computed tomography (SPECT), transverse relation time (T2)* weighted magnetic resonance imaging (MRI), histology, and 
-scintillation spectrometry. BMM labeled with super paramagnetic iron oxide and/or 111indium oxine were infused i.v. into naïve mice. During the first 7 h, greater than 86% of cell label was recorded within the lungs. On Days 1, 3, 5, and 7, less than 10% of BMM were in lungs, and 74–81% and 13–18% were in liver and spleen, respectively. On a tissue volume basis, as determined by SPECT and MRI, BMM densities in spleen and liver were significantly greater than other tissues. Migration into the lymph nodes on Days 1 and 7 accounted for 1.5–2% of the total BMM. Adoptive transfer of BMM loaded with NP-IDV produced drug levels in lymphoid and nonlymphoid tissues that exceeded reported therapeutic concentrations by 200- to 350-fold on Day 1 and remained in excess of 100- to 300-fold on Day 14. These data show real-time kinetics and destinations of macrophage trafficking and demonstrate the feasibility of monitoring macrophage-based, nanoformulated ART.
Key Words: monocytes • cell trafficking • mouse • indinavir
INTRODUCTION
Peripheral blood mononuclear phagocytes [(MP), bone marrow (BM) and blood monocytes, tissue macrophages, microglia, and dendritic cells (DC)] serve as vehicles for dissemination and reservoirs for the HIV-1 infection [1 2 3 ]. The ability of BM-derived macrophages (BMM) to migrate to the sites of infection makes them attractive candidates for use as vehicles to deliver antiretroviral agents. Indeed, the tissue distribution of circulating monocytes and notably, subpopulations of monocytes (CD14/CD16), which emerge during inflammatory conditions, closely parallels sites of active viral replication [4 , 5 ].
The question we asked is whether the same MP, which disseminate virus, could be used to traffic antiretroviral drugs to tissue sites of active viral replication. Precedent is provided by cell-based systems already operative for DNA vaccines, viral gene, and drug delivery systems developed for a variety of degenerative and neoplastic diseases [6 7 8 9 ]. However, obstacles in realizing this goal center on drug cell uptake and accurate real-time longitudinal study of monocyte trafficking to tissues known to harbor HIV-1 [10 ]. The first was demonstrated in our laboratories in cell culture experiments [11 ]. The second involves cross-sectional studies of macrophage trafficking and cell distribution kinetics, which are limited in number [10 , 12 ]. Over 20 years ago, one study combining radioactive tracer and histological examinations, demonstrated solely that 45% of splenic macrophages are locally produced, and 55% are derived from monocytic extravasation [13 ]. Moreover, a single, "theoretical" report of mathematical models for the path of macrophages and their potential as vehicles for drug delivery was reported without biologic support [14 ]. Thus, we strove to use two sensitive methods, magnetic resonance imaging (MRI) and single photon emission computer tomography (SPECT), to examine BMM trafficking to test the feasibility of macrophage-based drug delivery systems. Demonstrable numbers of BMM in lymphoid tissues by MRI and SPECT analyses compelled us to examine whether macrophage delivery of nanoformulated antiretroviral agents could provide rapid and prolonged antiretroviral therapeutic concentrations in tissues, wherein HIV is harbored and actively replicates.
MATERIALS AND METHODS
Animals
Male BALB/c mice (Charles River Laboratory, Inc., Wilmington, MA), 5–8 weeks old, were used for all experiments. Animals were housed in sterile microisolator cages and maintained in accordance with ethical guidelines for the care of laboratory animals of University of Nebraska Medical Center (Omaha) and National Institutes of Health (NIH; Bethesda, MD).
Preparation of BMM
To obtain BM-derived mononuclear cells, femurs were excised and flushed with HBSS. RBC were lysed with ammonium chloride potassium (ACK) lysis buffer, and clumps were removed by passing the cell suspension through a 40-µm cell strainer. BM cells were cultured for 10 days in Teflon flasks [15
] at 2 x 106 cells/ml in DMEM supplemented with 10% FCS, 2 mM L-glutamine, 1% penicillin/streptomycin, and 2 µg/ml M-CSF (complete medium). All reagents for cell culture were obtained from Invitrogen (Carlsbad, CA), except for M-CSF, which was a generous gift from Wyeth, Inc. (Cambridge, MA). Half of the medium was exchanged for fresh medium every 2 days. The purity of monocytes from culture was determined by flow cytometry using PE-conjugated antibody to CD11b (BD PharMingen, San Diego, CA) and analyzed with a FACSCalibur (BD Biosciences, San Jose, CA). CD11b+ cells within the BMM cultures were always detected in excess of 98%.
Magnetic resonance tracking of Feridex-labeled macrophages
BMM were labeled with super paramagnetic iron oxide (SPIO) particles (Feridex, Berlex Inc., Wayne, NJ) by culturing at 2.5 mg Feridex/107 cells/ml complete media for 1 h at 37°C. Cells were washed twice with DMEM, and each recipient mouse received 1 x 107 SPIO-labeled BMM in 200 µl i.v. via the tail vein. Greater than 95% cells were labeled with SPIO particles as determined by enumeration of Prussian blue-stained cells from cytological preparations. The presence of SPIO-labeled BMM in tissues was evaluated by MRI. Accumulation of SPIO particles in tissue causes an increase in the magnetic relaxivity of tissue water, which is strongly field-dependent. Measures of relaxivity in our laboratories using a 7T system (Bruker 21 cm Biospec operating Paravision 3.0.2) demonstrated that relaxivity is related directly to cell density, assuming uniform labeling and no cell death leading to loss of SPIO to native macrophage activity. This approach was used to track the migration of peripheral monocytes within the liver and spleen after i.v. injection [16
, 17
]. High-resolution, three-dimensional (3D) gradient-recalled echo MRI scans of mouse body were acquired using a 25-mm birdcage volume coil covering a region from the neck to the hips with acquisition parameters of echo time (TE) = 3 ms, repetition time (TR) = 50 ms, 30% echo, flip angle = 35 degrees, number of averages (NA) = 2, field of view = 35 x 25 x 50 mm with a resolution of 256 x 128 x 128 (voxel size=137x195x390 µm), reconstructed to 256 x 256 x 128, total acquisition time = 30 min. Signal intensity was normalized to an external standard to account for signal drift over time. BMM accumulation was determined by signal loss over time within selected regions of interest. After the injection of SPIO-labeled cells, 3D gradient-recalled echo images were acquired every 30 min for 6.5 h, at 24 h, and on Days 3, 5, and 7 thereafter. Signal intensity, normalized to an external standard, was measured within anatomical regions of interest to determine the rate of labeled cell influx or efflux. Relaxation rate (R2)* (1/T2*) decreases in direct proportion to the density of labeled cells in the tissue. We used this to calculate the average cell density from normalized signal loss using the equation: Cell density 
ln(Munlabeled) – ln(MSPIO), whereby cell density is proportional to the normalized signal intensity before injection of SPIO-labeled cells (Munlabeled) minus the normalized signal intensity at each time-point after injection (MSPIO).
Monocyte tracking using SPECT
To assess cell migration by SPECT, BMM were labeled with 111indium (111In) oxyquinoline (Indium oxine, Amersham Healthcare, Arlington Heights, IL) at a dose of 600 µCi per 30 x 106 cells in 1 ml RPMI 1640/10 mM HEPES for 45 min at 37°C. Cells were washed extensively and resuspended in HBSS. Labeling efficiency, as determined by -scintillation spectrometry (Packard Instrument Co., Meriden, CT), was routinely 70–80% of total input isotope. Each recipient received 5–10 x 106 111In-labeled BMM i.v. Mice were anesthetized with 0.5–1% isoflurane delivered in a 2:1 mixture of nitrous oxide and oxygen. Image acquisition was accomplished with a -scintillation camera detector fitted with a 1-mm pinhole collimator and interfaced with image acquisition software (A-SPECT, Gamma Medica, Northridge, CA). Briefly for each animal, 64 1-min, equiangular exposures over a 360° axis of rotation were acquired at each time-point. Acquired exposures were reconstructed into a single 3D tomogram. Regions of interest (ROI) within the processed tomograms were circumscribed by electronic bit maps to contain the lungs, liver, or spleen, and relative activities for each were determined. After acquisition of SPECT images for the final time-point, animals were killed, and tissues excised, weighed, and submitted for -scintillation spectrometry to determine the intensity of 111In signal in each tissue. After scintillation spectrometry, tissue was processed for autoradiography. Frozen sections of 30 µm were obtained and exposed to X-ray film. Autoradiographic images were digitized, and intensities of 111In-labeled BMM were assessed by digital image analysis (MCID Image Analysis System, Imaging Research, Inc., St. Catherines, Ontario, Canada).
Histology
Tissues were collected from mice after MRI tracking, fixed in 4% parafarmaldehyde for 24 h, embedded in paraffin, and 5 µm-thick sections were cut for analysis. Slide-mounted sections were deparaffinized, rehydrated, and reacted for 30 min in 2% potassium ferrocyanide in 3.7% hydrochloric acid to visualize ferric iron particles by Prussian blue. Stained sections were washed and counter-stained with nuclear fast red. Replicate serial sections were stained with H&E to provide histological cell distributions. Images of stained sections were imported into Image-Pro Plus, Version 4.0 (Media Cybernetics, Silver Spring, MD) for quantifying Prussian blue-stained cells. As a control for tissue distributions obtained by cell-free Feridex, Feridex alone was administered i.v. to naïve recipients at 0.25 mg for each mouse, equivalent to the amount used to load 10 x 106. Spleen, liver, lung, and lymph nodes were collected on Day 5 after Feridex administration, paraffin-embedded, sectioned, and stained for Prussian blue.
Indinavir (IDV) loading of BMM, adoptive transfer, and detection in tissues
BMM were incubated with 5 x 10–4 M nanoparticle-formulated IDV (NP-IDV; Baxter Healthcare Corp., Round Lake, IL) in complete medium for 12 h of incubation at 37°C. Cells were washed and adoptively transferred into recipient mice via tail vein. Animals were killed on Days 1 and 14, and tissues were collected for analysis of IDV concentrations by reverse phase (RP)-HPLC, as modified by the method of Jayewardene et al. [18
]. Briefly, tissues were homogenized in 60% methanol, maintained at 4°C overnight, and clarified by centrifugation at 20,000 g for 15 min at 4°C. Supernatants were collected and added to glass tubes containing 1 ml diethyl ether. The tubes were mixed for 30 s and maintained at –20°C for 30 min. The ether layer was evaporated to dryness under a nitrogen stream at room temperature. The residue was reconstituted in 150 µL mobile phase [10 mM ammonium dihydrogen phosphate with 1 mM 1-heptanesulfonic acid at pH 4.8 mixed with acetonitrile 65:35 (v/v)]. The rehydrated samples were clarified by centrifugation at 20,000 g for 5 min. Triplicate 35 µl aliquots of each sample were injected for RP-HPLC analysis (Shimadzu Corp., Columbia, MD). A C4 RP column with 5 µm particle size packing (Phenomenex, Torrance, CA) was used, and analytes were measured at 210 nm. The data were analyzed using chromatographic software (EZStart, Shimadzu Corp.), and peak area was integrated. Concentrations of IDV were determined compared with a standard concentration curve of IDV in mobile phase. Processing and analyses were validated by spiking known concentrations of IDV in homogenized tissue samples from naïve animals.
NP-IDV was tagged with LissamineTM rhodamine B 1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt (rhodamine DHPE; Invitrogen). BMM were loaded with rhodamine-labeled NP-IDV (rh-NP-IDV) and adoptively transferred i.v. to naïve recipient mice. Frozen sections of lymphoid and nonlymphoid tissues were reacted with anti-CD11b (Serotec, Raleigh, NC) and Alexa Fluor 488 (Molecular Probes, Eugene, OR)-labeled secondary antibody and evaluated by fluorescence microscopy to detect rh-NP-IDV (red) and CD11b+ (green) cells.
Statistical analyses
Normative data from SPECT, MRI, -scintillation spectrometry, and counts from histology were evaluated by Student’s t-test and ANOVA with least significant difference post-hoc test using the statistical software packages, SPSS, v13.0 (SPSS, Inc., Chicago, IL) and Statistica, v7.1 (StatSoft, Inc., Tulsa, OK). Data are expressed as means ± SEM for 4–6 animals per group. Significance levels were chosen as P 
0.05.
RESULTS
Tracking 111In-labeled BMM migration by SPECT
To assess real-time migration of macrophages, 5–7 x 106 111In-labeled BMM were adoptively transferred to naïve recipients, and migration was assessed by SPECT between 5 and 8 h post-transfer and on 1, 3, 5, and 7 days thereafter. Within 5 h post-transfer (Day 0), the majority of radiolabeled BMM was detected on tomographic images within the lungs with lower intensity signals emanating from the spleen and liver (Fig. 1A
). By Day 1 post-transfer and times thereafter, the most intense signals were found within the spleen and liver. Congruous signal intensities were demonstrated by all animals. The progressively decreasing signal intensities exhibited on Days 3, 5, and 7 reflect the decay characteristics of 111In (t1/2=2.8 days). To quantify the density of BMM within each tissue, ROI were drawn with electronic bit maps, and the relative counts and volumes were determined and corrected for radioactive decay. In agreement with tomographic images, the number of counts/cm3 in lungs was significantly greater at 5–8 h post-transfer compared with those in spleen and liver; however, by Day 1 and thereafter, counts decreased in lungs and increased significantly in spleen and liver, but no significant differences in densities of BMM were discernible between liver and spleen (Fig. 1B)
. Evaluation of percent distribution of BMM indicated that after 5 h post-transfer, 66% ± 2% of the labeled BMM were in the lung, 25% ± 2% in the liver, and 9% ± 1% in the spleen (Fig. 1C)
. Later, significantly greater levels of BMM (74–81%) remained in the liver, 13–17% in spleen, and 6–9% in the lungs.
 View larger version (46K): [in a new window] |
Figure 1. SPECT analysis to track BMM after adoptive transfer to naïve mice. Mouse BMM were labeled with 111In oxine [half-life (t1/2)=2.8 days], washed, and 1 x 107 were adoptively transferred by i.v. injection to naïve, syngeneic recipients. Trafficking of BMM was evaluated by SPECT analysis for each animal after 6 h (Day 0) and 1, 3, 5, and 7 days post-transfer. (A) SPECT images of all four mice on Days 0, 1, 3, 5, and 7, showing congruous intensity of 111In-labeled BMM in lung (Lg), liver (Lv), and spleen (Sp). (B) For each lung, liver, and spleen within a tomographic image, a ROI was circumscribed by electronic bit maps to include each tissue and saved as a subimage. The tomographic subimage was sectioned transaxially, and the relative number of counts and volumes was determined by image analysis software provided by the manufacturer. Relative counts were corrected for decay, and the mean counts/cm3 ± SEM (n=4) for each tissue are plotted as a function of time. (C) The distribution of labeled BMM was determined by normalizing the total relative counts as a percentage of total counts within all tissues for each animal, and the mean distribution ± SEM (n=4) was evaluated as a function of time after transfer of BMM. aP 0.05, for spleen and liver compared with lung.
|
 View larger version (46K): [in a new window] |
Figure 2. Tracking BMM migration by MRI and coregistration with SPECT images. (A) Time series of spleen (Spl), kidneys (Kid), and bowel (Bwl; upper panels) and liver (Liv; lower panels) from mice before (Day 0) and on 1, 3, and 5 days after i.v. injection of SPIO (Feridex)-labeled BMM, demonstrating visible signal loss after injection of Feridex-labeled BMM. Quantitation of in vivo tracking of mouse BMM labeled with Feridex using T2* weighted MRI during (B) the initial 6.5 h after adoptive transfer and (C) on Days 1, 3, 5, and 7 thereafter. Results shown are measures of Feridex-labeled cell densities in spleen, liver, and kidney (mean±SEM from n=4 mice per group). aP 0.05, compared with kidney. (D) BMM were dual-labeled with Feridex and 111In and transferred i.v. to recipient mice. Recipients were anesthetized and positioned in custom-built, MRI/SPECT-compatible holders with attached fiducial markers (not shown), and serial acquisitions of MRI and SPECT scans were performed on Day 1 after transfer. SPECT data were interpolated to the resolution of the MRI data (256x256x128), and fiducial markers were used for scaling and alignment. Reconstructed SPECT images (green) were coregistered and overlaid to MRI scans (red) using the Analyze package (AnalyzeDirect, Inc., Lenexa, KS). Displayed, coregistered images are 390 µm slices containing spleen and kidney (left panel) and liver (right panel) and show areas of radioactive intensities of 111In-labeled BMM in SPECT images (green), which correspond to loss of MRI signal as a result of Feridex-labeled BMM.
|
Tissue distribution of 111In-labeled BMM
To validate SPECT analysis, we assessed tissue distribution and density of 111In-labeled macrophages by -scintillation spectrometry of lymphoid (spleen and lymph nodes) and nonlymphoid (liver, lung, and kidney) tissues harvested immediately after the last SPECT data acquisition on Days 1 and 7. After i.v. administration of BMM, spleen, liver, and lung retained 96% of the tissue counts (62.5, 21.3, and 12.1%, respectively) after Day 1 and were essentially unchanged by Day 7 (68.2, 20.6, and 6.7%, respectively). Rank order analysis of BMM density (relative counts/mg tissue weight) and distribution of counts from spleen and liver indicated significant differences (P<0.002) between those two tissues and from other tissues; however, no significant differences were detected among other tissues. As the densities and tissue distributions of BMM were similar on Days 1 and 7 post-transfer, this suggested that after initial migration into tissues, the bulk of macrophages is retained in those tissues. On Days 1 and 7, distribution of BMM in lymph nodes (brachial, cervical, and inguinal) accounted for 2% ± 1.6% and 1.5% ± 0.7%, respectively, of the total radioactive signal, suggesting that macrophages possess the capacity to migrate to and remain within lymph nodes. Despite negligible signals in MRI tracking, -scintillation spectrometry indicated that kidneys retained 1.9% ± 1.4% of total 111In-labeled BMM measured in tissues, demonstrating that BMM are capable of migrating through nonlymphoid tissues as well.
Histological confirmations of Feridex-labeled BMM
To validate migration of Feridex-labeled macrophages after adoptive transfer, tissues were harvested from recipient mice after the last MRI examination at 24 h and from another set after the last MRI examination on Day 7 post-transfer. Sections from lung, liver, spleen, and lymph node were processed and stained to detect Feridex-labeled BMM as Prussian blue-stained cells. Tissues from uninjected control animals without Feridex were stained for Prussian blue to detect endogenous ferrum (data not shown), and this baseline stain was used for comparison and analysis. Histological examination of tissues following acquisition of MRI images confirmed that the majority of macrophages migrated to spleen, liver, lung, (Fig. 3
), and lymph nodes (Fig. 4
). When quantified using Image-Pro software, spleens showed the greatest density of Prussian blue-stained cells on Days 1 and 7 without detectable, significant differences (555±25 and 473±35 Prussian blue cells/mm2, respectively; P=0.07). Prussian blue-stained BMM were distributed primarily within the resident macrophage areas surrounding the germinal centers (Fig. 3)
. Lungs showed an initial accumulation of Feridex-labeled cells by 24 h post-transfer (237±19 cells/mm2), but virtually no cells could be detected by Day 7 (1.1±0.3 cells/mm2), demonstrating a significant reduction in BMM in the lung over the final 6 days after transfer (P<0.0001). Feridex-labeled cells were detected in the sinusoids of the liver. The density of BMM in the liver on Day 1 (81±9 cells/mm2) was significantly less than densities in the spleen and lung at the same time (P<0.0001) but significantly increased by Day 7 (121±10 cells/mm2, P=0.009) exceeding levels in lung (P=0.001), although remained less than BMM density in spleen at that time (P<0.0001).
 View larger version (113K): [in a new window] |
Figure 3. Feridex or NP-IDV-loaded macrophages within lymphoid and nonlymphoid tissues after transfer to naïve recipients. BMM were loaded with Feridex or rhodamine-labeled NP-IDV (rh-NP-IDV) and injected i.v. into naïve recipients. Spleen, lung, and liver were harvested and evaluated by immunohistological analysis. After acquisition of final MRI images after Day 1 post-transfer, tissues were stained to detect Feridex-labeled macrophages by Prussian blue staining (blue) and counter-stained with nuclear fast red. Serial sections were stained with H&E for histological comparison. Feridex-labeled BMM were detected in areas of resident interfollicular macrophages surrounding the germinal centers. Tissues from mice receiving rh-NP-IDV-loaded BMM were collected at Day 5 after adaptive transfer, processed as frozen sections for immunohistology, and examined by fluorescent microscopic analysis. CD11b (green)-positive macrophages containing rhodamine-labeled NPs (red) are shown in spleen, liver, and lung. Original, large micrograph magnifications, 200 µm; original bar, 50 µm; original inset magnifications, x1000; bar, 5 µm.
|
 View larger version (120K): [in a new window] |
Figure 4. BMM migration to the lymph nodes, which were extracted from mice at (A–D) Day1 and (E–H) Day 7 after adaptive transfer of Feridex-labeled BMM. (A, C–E, G, H) Sections were stained to detect Feridex using Prussian blue staining and counterstained with nuclear fast red. (B, F) Serial sections were stained with H&E for histological comparison to show the afferent lymphatics marked by the arrow. (A, B, E, F) Micrographs are x100 original magnifications. (C, D) Micrographs are x1000 original magnifications of the areas marked in A showing (C) Feridex-labeled cells and (D) many degenerating macrophages faintly labeled with Feridex. Panels are (G) x400 original magnification and (H) x1000 original magnification of the boxed area shown in E containing released Feridex and Feridex-labeled cells in the afferent lymphatics.
|
Animals were also injected with Feridex alone and tissues collected on Day 5. When stained for Prussian blue, tissues from mice administered Feridex alone showed uniquely different patterns of Feridex distribution compared with tissues from mice administered Feridex-labeled BMM (Fig. 5 ). Livers from mice injected with Feridex showed higher accumulations of Prussian blue-staining cells than those injected with Feridex-loaded BMM. Spleens showed robust staining only from animals treated with Feridex-labeled BMM.
 View larger version (145K): [in a new window] |
Figure 5. Tissue distributions from mice treated with BMM loaded with Feridex or Feridex alone. Animals were injected i.v. with 10 x 106 BMM loaded with 0.25 mg Feridex or 0.25 mg Feridex alone. Tissues were obtained on Day 5 postadministration and paraffin-embedded, and sections were subjected to Prussian blue staining. Feridex was distributed uniquely when injected as a cell-free formulation compared with BMM-loaded Feridex showing greater accumulation of Feridex in liver and less in spleen. Original micrograph magnifications, x100; bar, 50 µm.
|
-scintillation spectrometry confirmed the migration and presence of 111In-labeled BMM in lymph nodes (Fig. 6C)
.
 View larger version (97K): [in a new window] |
Figure 6. Distribution of 111In/Feridex-labeled BMM, which were dual-labeled with Feridex and 111In, and 10 x 106-labeled cells were adoptively transferred to recipient mice i.v. Tissues were collected on Day 5 after BMM administration, and 30 µm frozen sections were obtained and exposed to X-ray film for autoradiography. Autoradiographic images were digitized and assessed by digital image analysis. (A) Original autoradioraphic images (top panels) and enlarged areas (center panels) show radioactive signal intensities from 111In-labeled BMM in spleen, liver, and lung. Higher radioactive signal is observed in splenic interfollicular areas around germinal centers. The same sections used for autoradiography were visualized for Feridex by Prussian blue staining (bottom panels) and nuclear fast red counterstain. A splenic germinal center is depicted (circle) surrounded by Prussian blue-stained BMM in interfollicular areas (arrows). Original magnifications for bottom panels are x100. (B) Intensities of radioactive signals in tissues from the images collected from six animals were quantified and expressed as mean ± SEM. (C) Radioactivity in lymph nodes collected from animals on Day 5 after BMM administration was measured by -scintillation spectrometry and graphed as cpm/mg tissue weight (mean±SEM of six animals). In LN, inguinal lymph node; Br LN, brachial lymph node; Cer LN, cervical lymph node.
|
37.5 µM or 75% of the initial NP-IND concentration [21
]. To evaluate the potential for delivery of NP-IDV by macrophages and assess whether therapeutic concentrations of IDV can be attained in immune tissues, we loaded BMM with NP-IDV, as shown in Figure 7B
, and adoptively transferred the NP-IDV-loaded BMM i.v. to naïve, recipient mice. IDV concentrations were evaluated by RP-HPLC in spleen, lymph nodes, lung, liver, and kidney on Days 1 and 14 after transfer (Fig. 7C)
. Of foremost significance is the finding that for all tissues examined, IDV concentrations, on a tissue-weight basis, were at least one order of magnitude in excess of the reported therapeutic concentration range in human plasma [19
, 20
] and were at least 2 and 3 orders of magnitude higher, respectively, in the lymph nodes and spleen; both lymphoid tissues implicated in harboring HIV-1 reservoirs. One day after transfer of NP-IDV-loaded BMM, IDV concentrations were significantly highest in the lung compared with other tissues, reaching levels that were in excess of 3.5 orders of magnitude greater than the therapeutic concentration range of IDV in plasma. It is notable that significant levels of IDV were detected in spleen and lymph nodes. By Day 14 after transfer, IDV concentrations were significantly higher in lymphoid tissues of the spleen and lymph nodes compared with other nonlymphoid tissues of lung, liver, and kidney. Although IDV concentrations were not changed significantly in lymph nodes between Days 1 and 14, levels increased significantly in spleen by Day 14 and diminished in lungs, liver, and kidney.
 View larger version (132K): [in a new window] |
Figure 7. IDV levels in lymphoid and nonlymphoid tissues after i.v. administration of NP-IDV-loaded BMM. Mouse macrophages (BMM) were incubated for 12 h at 37°C in 5% CO2 with NP-IDV at a final concentration of 5 x 10–4 M. Light micrographs of (A) unloaded macrophages and (B) NP-IDV-loaded macrophages are shown at x400 original magnification. (C) NP-IDV-loaded BMM were administered i.v. to recipient mice. On Days 1 and 14 postadministration, tissues were harvested, weighed, homogenized, and submitted for RP-HPLC analysis for levels of IDV. Means of IDV concentrations in ng/g tissue wet weight (±SEM) of six mice per time-point were determined and plotted as a log function. Therapeutic concentration range for IDV in human plasma (150–675 ng/ml plasma) [19
, 20
] is indicated as the translucent gray band parallel to the x-axis. Comparison of IDV concentrations between tissues within each time-point: P 0.05 is significant compared with aspleen; blymph nodes; clung; and dliver. eP 0.05, kidney, is significant for comparison of IDV concentrations between Day 1 and Day 14 for each respective tissue.
|
Macrophages have been of great interest as gene and drug delivery vehicles [9 , 22 ]. However, relatively few in vivo studies have assessed the ability of ex vivo-transferred macrophages to migrate to the target sites. For optimal use of macrophage delivery systems, such studies are necessary to accurately assess whether cells and drugs reach a targeted distribution. In this study, we explored the feasibility of using macrophages for delivery of nanoformulations of the anti-HIV-1 protease, IDV, with the expressed intent to improve the efficacy of the drug by targeted distribution. A necessary step toward implementing this strategy required the measurement of macrophage migration and homing to ensure the adequacy of cellular penetration into tissues wherein HIV is harbored and replicates.
For these studies, we used syngeneic BMM as sources originating from normal progenitor cells, which were expanded by in vitro cultivation. For tracking studies, we used BMM as radiolabeled (111In) macrophages or macrophages laden with Feridex. The former were tracked by SPECT and the latter by T2* MRI. Labeled BMM were adoptively transferred i.v. to naïve recipient mice, and recipients were subjected to MRI and SPECT analyses. Images were initially obtained every 30–60 min during the first 8 h post-transfer and then, 24 h and every other day thereafter. Clearly, as assessed by MRI and SPECT, the majority of BMM initially remained in the lung but infiltrated to the spleen by 24 h and remained within lymphoid tissue for up to 7 days. -Scintillation spectrometry and histological examination of spleen confirmed the presence of labeled BMM and further demonstrated their presence in lymph nodes. Although the majority of BMM was retained in the liver as a result of the mass of the tissue, the densities of BMM in spleen and liver on a volume basis were not significantly different from each other, as determined by SPECT and MRI. Several studies reported similar pattern of distribution of systemically administered macrophages [10
, 12
]. Migration of macrophages into the spleen was shown to be dependent on their activation state. Murine peritoneal macrophages elicited by Brewer’s thioglycollate medium when injected i.v. localized in the lung with minimal migration to spleen even after 72 h [10
]. Macrophage-like cell lines were shown to have no migration capability but were rapidly cleared from the circulation.
SPECT and MRI provide invaluable tools for noninvasive longitudinal assessment of migration and distribution of cell trafficking after transfer of leukocytes in animals and humans. SPECT is a sensitive method to follow the time-course of migration of transplanted cells using 111In-oxine-labeled cells [23 ]. 111In has been used in studies to find biodistribution of the cells after transplantation, but animals typically had to be killed at each time-point for autoradiography or radioactivity measurement [24 ]. SPECT scanning allows in vivo real-time tracking of cell trafficking in the same animal or patient. 111In-oxine labeling of peripheral blood leukocytes to assess foci of inflammation has long been used in clinical studies of inflammatory disorders [25 , 26 ]. Moreover, the use of 111In-labeled macrophages to study the distribution of mononuclear cells in tumors was demonstrated as an elegant procedure for tracking the distribution of macrophages and DC [27 28 29 30 ]. Similarly, MRI provides another noninvasive method for longitudinally studying in vivo, the fate of transplanted cells labeled with a paramagnetic probe such as Feridex [31 ], therefore rendering cell-specific imaging an increasingly important field of MRI [16 , 32 33 34 ].
Having shown by SPECT, MRI, -scintillation spectrometry, and histological examination that adoptive transfer of unladened and NP-ladened BMM provided detectable and prolonged levels of macrophage migration into lymphoid tissues, whether adoptive transfer of macrophages ladened with antiretroviral agents can be used for efficacious, antiretroviral drug delivery has not been assessed. Adoptive transfer of macrophages has been shown to be a relatively risk-free procedure in humans with only mild side effects [35
, 36
] and should provide a unique, efficacious delivery platform by which to introduce antiretroviral agents deeper into lymphoid tissues. To assess the efficacies of antiretroviral distribution into tissues where HIV replicates, strategies were initiated to deliver antiretroviral drugs in macrophage vehicles, cells that naturally migrate to tissues associated with HIV infection and harboring HIV reservoirs. Our data demonstrated that with one administration of macrophages loaded with NP-IDV, therapeutic levels of IDV in lymphoid and nonlymphoid tissues were attained rapidly by 1-day postadministration and were maintained in the spleen, lung, and lymph nodes for up to 14 days thereafter. These observations are supported by investigations in humanized mice infected with HIV-1. Here, NP-IDV administered in BMM demonstrates antiretroviral responses and protection of human CD4+ T cells [21
]. It is interesting that although IDV levels were highest in the lung on Day 1 compared with other tissues, BMM in the lung were relatively low as determined by SPECT for 111In-labeled BMM and by MRI using Feridex-labeled BMM. This suggested that by Day 1 after transfer, a sufficient number of NP-IDV-loaded BMM remained in the lung to release such high levels of IDV or high levels of NP-IDV or free IDV remained in the lung after cells had immigrated, possibly as a result of release from BMM and subsequent uptake by endogenous lung macrophages. The kinetics and biodistribution of free IDV given orally have been shown to be short-lived and require multiple doses per day to attain and maintain therapeutic levels [37
]; nanoformulated IDV delivered via macrophage vehicles afforded rapid and prolonged attainment of therapeutic levels of IDV in tissues where HIV is harbored and replicates. Although the levels reported here resulted from a single administration of NP-IDV-loaded IDV, the potential for higher concentrations remains possible upon multiple doses. Therefore, these findings strongly support the feasibility of using macrophages as drug delivery vehicles for nanoformulated drugs to enhance the efficacy of antiretroviral therapeutics. This strategy has been shown to reduce markers of infectious HIV-1 in serum and lymphoid tissues of a human/mouse chimera model of HIV-1 infection [21
].
Antiretroviral therapy (ART), which includes IDV as a potent protease inhibitor, has been effective in reducing plasma viral load in HIV-infected patients. The efficacy of ART is compromised typically as a result of the nature of complex regimens requiring multiple daily dosing, diligent adherence to the regimen, and the limited biodistribution of an orally administered drug [38 ]. The efficacy of IDV use is also compromised, as administration of an unformulated drug results in low levels in deep lymphoid organs [39 ], where HIV infection is prevalent as a result of trapped HIV particles on the follicular DC in the germinal centers [40 41 42 ]. Viral replication in lymphoid tissues has been shown to be ten- to one hundred-fold higher than that in PBMC [41 ]. Using higher doses of antiretroviral agents can increase the drug concentration within lymphoid tissues; however, the drawback is that this strategy for attainment of therapeutically effective concentrations produces toxic side effects by many antiretroviral drugs and often complicates HIV drug therapies [38 , 43 ]. Such adverse effects could be minimized by nanoformulated drug delivery modalities that result in selective uptake of drug by macrophages [9 , 44 , 45 ] and provide site-specific delivery to the lymphoid tissues, thus preventing HIV replication and the establishment of HIV reservoirs during a clinical latency period.
ACKNOWLEDGEMENTS
This work was supported by NIH Grants P01 NS31492, R01 NS34239, R37 NS36126, P20 RR15635, P30 A142845, NS 43011, T32 NS07488, P01 NS11766, and P01 MH64570 (to H. E. G.), R1049264 (to R.L.M.), P01 NS43985 (to M. B. and H. E. G.) and Baxter Healthcare Corporation. The authors thank Robin Taylor of the University of Nebraska Medical Center for administrative assistance.
Received February 21, 2006; revised June 19, 2006; accepted June 20, 2006.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  H. Dou, C. B. Grotepas, J. M. McMillan, C. J. Destache, M. Chaubal, J. Werling, J. Kipp, B. Rabinow, and H. E. Gendelman Macrophage Delivery of Nanoformulated Antiretroviral Drug to the Brain in a Murine Model of NeuroAIDS J. Immunol., July 1, 2009; 183(1): 661 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Montaner, S. M. Crowe, S. Aquaro, C.-F. Perno, M. Stevenson, and R. G. Collman Advances in macrophage and dendritic cell biology in HIV-1 infection stress key understudied areas in infection, pathogenesis, and analysis of viral reservoirs J. Leukoc. Biol., November 1, 2006; 80(5): 961 - 964. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
