Macrophage targeting within the kidney using USPIO was first proposed for a model of nephrotic syndrome, induced by intravenous injection of puromycin aminonucleoside in rats (Hauger et al. 1999). This model induces both lesions of glomerular epithelial cells and a glomerular and tubulointerstitial infiltration by macrophages. MR images showed a diffuse decrease of signal intensity predominantly within the outer medulla 24 h after the IV injection of Sinerem^® (Guerbet Group, Aulnay-sous-Bois, France). The degree of signal decrease was correlated with the number of macrophages within each renal compartment and to the amount of iron within the tissue.

A diffuse interstitial macrophagic infiltration of renal parenchyma is a well-documented feature of chronic ureteral obstruction (Schreiner et al. 1988). This cellular influx peaks the second day after the obstruction. Delayed USPIO-enhanced MR imaging demonstrated a diffuse but significant decrease of signal intensity in the three renal compartments, slightly more pronounced in the cortex (Hauger et al. 2000) (Fig. 2).

In acute proliferative types of GN, such as Goodpasture syndrome or vascular nephritis, macrophages also play a role in the development of glomerular inflammation (Cattell 1994). They accumulate in the Bowman space as a primary feature in the development of advanced cellular crescents which is a known as a feature of rapidly progressive GN which is associated with a poor prognosis. A model of anti-glomerular basal membrane (GBM) GN, comparable to Goodpasture syndrome in humans, was evaluated with MR imaging (Hauger et al. 2000). This model was induced in rats by IV injection of sheep anti-rat-GBM serum. The kinetics of signal intensity followed the biphasic evolution of the disease with a decrease of signal intensity within the cortex only at days 2 and 14 (Fig. 3a) and no change at day 21. This effect was due to endocytosis of USPIO by macrophages at day 14 and by activated mesangial cells at both phases (Fig. 3b). The degree of signal intensity decrease was strongly correlated with the degree of proteinuria.

Intrinsic acute renal failure (ARF) is a multifactorial disease with concomitant ischemic, nephrotoxic, and septic components (Devarajan 2006), and ATN is its pathological counterpart. Besides alterations in hemodynamics, tubule dynamics, tubule cell metabolism and structure, the intrarenal inflammatory response plays a major role in the ischemic ARF. Neutrophils are the first leukocytes to accumulate in the postischemic kidney and macrophages are the next. Macrophages migrate into the outer medulla of rat kidneys and accumulate within peritubular capillaries, interstitial space, and even within tubules (Friedewald and Rabb 2004; Ysebaert et al. 2000). In an ischemia-reperfusion model in rats, USPIO-enhanced MR imaging demonstrated the same endocytosis pattern with a decrease of signal intensity within the outer medulla only from 24 to 120 h after the injection of particles (Jo et al. 2003) (Fig. 4). Signal intensity decrease was also correlated with the level of renal function.

After renal transplantation, acute and chronic rejection episodes play a major role on long-term graft survival and involve T cells and macrophages. Two studies evaluated a rat model of acute rejection and showed a decrease of signal intensity within the cortex and medulla, increasing with iron dosage (Ye et al. 2002) and inhibited by immunosuppressive treatments (Zhang et al. 2000). Beckmann et al. (2006) used SPIO to label macrophagic infiltration in a model of chronic rejection. They showed a dose-dependent decrease in cortical signal intensity in allografts between 8 and 16 weeks after transplantation (Fig. 5). The relative cortical signal intensity in the grafts was negatively correlated to the Banff score 6 weeks after transplantation, and changes in creatinine clearance occurred only after 28 weeks. Groups treated with immunosuppressive and antiproliferative drugs showed lesser degree of signal intensity change (Beckmann et al. 2003).

All these experimental results showed that USPIO-enhanced MR imaging could demonstrate intrarenal internalization of particles by macrophages or by glomerular cells gaining endocytic activity, i.e., mesangial cells, and could precisely localize this endocytic activity in the different kidney compartments. Different types of experimental renal diseases showed different patterns of signal intensity changes, but the degree of renal dysfunction always appeared correlated to the degree of endocytic activity, which may have significant implications in clinical practice.

These results led to the first pilot clinical study (Hauger et al. 2007), based on 12 patients exhibiting a dysfunction of either their native kidneys or their transplant. As the blood half-life of USPIO is 36 h in humans, MR imaging was performed 3 days after USPIO injection (Sinerem^®, Guerbet Group) to ensure avoiding signal changes from vascular blood volume. Patients with chronic and fibrotic disease, without inflammatory component on biopsy, did not show any significant change (Fig. 6a, b). Three patients with ATN (two transplanted kidneys and one native kidney) showed a significant decrease of signal intensity within the medulla only (Fig. 6c, d). All patients but one with an inflammatory component on cortical biopsy showed a significant decrease of signal intensity after USPIO injection (Fig. 6e, f). These preliminary clinical findings seem to corroborate the experimental results, with the same topographic patterns, and call for larger multicenter clinical trials, and the evaluation of imaging at 2 days after injection to reduce delay in the diagnosis.

Recovery of renal function after acute nephrotoxic or ischemic insult is dependent on the replacement of necrotic tubular cells by functional tubular epithelium (Gupta et al. 2002). This cellular regeneration may originate from renal resident cells or from extrarenal ones. Mesenchymal stem cells (MSCs) have been shown to be able to repair damaged tissues, whether genetically modified or not (Baksh et al. 2004; Pittenger and Martin 2004). They also per se have potential therapeutic effects such as those with regard to the degradation of the extracellular matrix (ECM) in the experimental models of fibrosis (Fang et al. 2004) or the facilitation of the grafting of transplanted bone marrow progenitor cells by providing a competent stroma (Bacigalupo 2004). They have a well-established ability to differentiate into the mesoderm lineage, which makes them potentially useful in strategies aiming at targeting the kidney mesangium for instance.

Recently, the possibility for bone marrow-derived MSC to differentiate into mesangial cells (Imasawa et al. 2001; Ito et al. 2001) and for hematopoietic stem cells into tubular cells was demonstrated in vivo (Lin et al. 2003), bringing great therapeutic promise for the future. Noninvasive imaging techniques allowing in vivo assessment of the location of stem cells could be of great value for experimental studies in which these cells are transplanted. It provides a tool to immediately verify if the grafted cells have reached the target organ, to estimate the number of cells that were seeded, and to assess the permanence of these cells over time with sequential imaging. Using SPIO preparations to magnetically label the cells, several groups have demonstrated the feasibility of grafting and subsequent visualization of the progenitor of different organs (Bulte et al. 1999; Hoehn et al. 2002; Kraitchman et al. 2003). A strong R2* effect was observed in vitro after MSC labeling for 48 h with 50 μg Fe/mL and increased linearly with the iron dose (Bos et al. 2004). After intravascular administration, the renal distribution of SPIO-labeled MSC has been investigated recently. When administered into the renal artery of normal kidneys, labeled MSC could be detected in vivo within the cortex as long as 7 days after injection at 1.5 T (Fig. 7) and iron-loaded cells were identified in renal glomeruli using histology (Bos et al. 2004). The same experiment was repeated at 3 T (Ittrich et al. 2007) with similar results in normal kidneys. In kidneys with acute renal injury (40′ ischemia-reperfusion), the improvement of renal function was accelerated in rats treated by cell graft: lower plasma creatinine levels at days 2 and 3 (Ittrich et al. 2007). In a model of acute experimental glomerulopathy (Thy1 + PAN), a homing effect was identified ex vivo, at 9.4 T, when the magnetically labeled MSCs were administered intravenously. A large proportion of cells were trapped within the liver precluding their detection in the kidneys in vivo at 1.5 T. When comparing the enhancing renal segments in ex vivo images and pathological lesions in histological sections, the areas of low signal intensity correlated well with alpha-actin and Prussian blue stains, indicating that MSCs specifically homed to injured tissues (Hauger et al. 2006) (Fig. 8).

Up to now, MR imaging of transgene expression has been limited to the demonstration of proof-of-principle experiments, using a change of signal intensity under the control of the expression of a reporter gene. As with PET, two approaches have been proposed with MR imaging using either an enzymatic pathway or a receptor pathway. The first method is based on the expression of a reporter gene (Lac-z) coding for an enzyme (β-galactosidase), which is able to activate a Gd-chelate by giving access of water to the first coordination sphere of the Gd ions (Louie et al. 2000). The second method is based on the targeting, with iron oxide particles, of cells overexpressing the transferrin receptors on their membrane (Moore et al. 2001; Weissleder and Mahmood 2001). None of these methods have been applied in the field of urogenital diseases.

MR imaging could also play a role in controlling gene expression in space and in time. The possibility to limit the effect of a therapeutic gene to a target diseased tissue and during a predefined time window is a key challenge for the future of gene therapy. For such a purpose, expression of a transgene can be controlled by a heat-sensitive promoter, such as the promoter of the inducible heat shock protein systems, HSP70 (Rome et al. 2005). These promoters can activate gene expression several thousandfold in response to hyperthermia (Dreano et al. 1986). Heating of the deep parts of the body can be achieved with high-intensity focused ultrasound (HIFU), and temperature control is made possible by using temperature-sensitive MR sequences based on the proton resonance frequency (PRF) method (Quesson et al. 2000) and by using an automatic monitoring of the acoustic power.

This approach has been recently experimented in vivo within the rat kidney. First, modified MSC expressing the luciferase reporter gene under the control of the hsp70B promoter was administered through the left renal artery. Luciferase expression in MR-guided HIFU heated regions was first demonstrated in vitro, using immunostaining (Letavernier et al. 2007). More recently, using modified C6 cells with hsp70B-luc expression injected in the artery of superficialized rat kidneys, local luciferase expression was demonstrated for the first time in vivo, by using bioluminescence imaging (Eker et al. 2010). The optimal heating protocol was found to be 43° during 5 min to get reproducible significant expression without parenchymal deleterious effect of heating.

Molecular MR imaging of cell receptors is a growing field and requires the development of specific contrast agents. The most developed strategy for targeting cells using MR imaging is to bind a contrast agent with a monoclonal antibody, an aptamer, or a peptide, able to recognize specifically a surface antigen expressed on the cell membrane surface. Up until now, the main applications focused on targeting inflammatory receptors, such as P-selectin (Chaubet et al. 2007), E-selectin (Reynolds et al. 2006), VCAM (Nahrendorf et al. 2006), ICAM (Choi et al. 2007), or tumor-specific receptors, such as HER2/neu (Daldrup-Link et al. 2005). To our knowledge, no application of these strategies involved the kidney.

Apoptosis and its regulatory mechanisms contribute to cell number regulation in ARF (Ortiz et al. 2003). Tubular cell apoptosis is promoted by both exogenous factors such as nephrotoxic drugs and bacterial products and endogenous factors such as lethal cytokines. Conversely to necrosis, which is a nonreversible process, apoptotic pathways are potentially accessible to therapeutic modulation (Rana et al. 2001). MR imaging for the detection of apoptosis requires labeling of apoptotic cells using USPIOs as a marker and phosphatidylserine as a target. This necessitates linking the superparamagnetic nanoparticle to annexin V or synaptotagmin I both of which bind to the phosphatidylserine present on the outer leaflet of the plasma membrane of apoptotic cells (Hakumaki and Brindle 2003). Several preclinical applications using MR imaging have been reported to target apoptotic cells within the heart (Hiller et al. 2006) or within tumors during chemotherapy (Zhao et al. 2001), but, to our knowledge, this approach has never been applied to any kidney model.

Whereas enzyme activity is highly involved in renal physiopathology of many renal parenchymal diseases and urinary cancers, nothing has been done in that field using MR imaging.