Kidney Regeneration and Stem Cells

Authors


ABSTRACT

The kidney has the capacity to recover from ischemic and toxic insults. Although there has been debate about the origin of cells that replace injured epithelial cells, it is now widely recognized that intrinsic surviving tubular cells are responsible for the repair. On the other hand, the cells, which have stem cell–like characteristics, have been isolated in the kidney using various methods, but it remains unknown if these stem cells actually exist in the adult kidney and if they are involved in kidney regeneration. This review will focus on the pathophysiology of kidney regeneration and the contribution of renal stem cells. We also discuss possible therapeutic applications to kidney disease. Anat Rec, 297:129–136. 2014. © 2013 Wiley Periodicals, Inc.

Acute kidney injury (AKI), formerly known as acute renal failure (ARF), is a common clinical complication that is associated with high morbidity and mortality. The incidence of AKI in hospitalized patients has been reported to be between 3.2 and 20%, and the risk of death increases in correlation with the severity of AKI (Murugan and Kellum, 2011). Because of the kidney's hypoxic condition, high metabolic demand, and frequent exposure to toxins and waste products, the kidney is susceptible to acute injury. After an ischemic or toxic injury, tubular epithelial cells lose their polarity and brush borders and undergo both apoptosis and necrosis. Cell–cell contacts are decreased and some of the dead cells are detached from the tubular basement membrane and are sloughed into the tubular lumen. The remaining viable cells proliferate, migrate, cover denuded areas of the basement membrane, and finally replace the injured cells (Bonventre and Yang, 2011). Growth factors, such as epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), hepatocyte growth factor (HGF), and bone morphogenic protein 7 (BMP7), play an important role in promoting the regeneration of viable cells (Nigam and Lieberthal, 2000; Zeisberg et al., 2003).

The origin of the cells that replace injured epithelial cells has been a topic of debate. Three possible sources have been considered: bone marrow stromal cells (BMSCs), intrarenal stem/progenitor cells, and endogenous surviving epithelial cells (Cantley, 2005). Early histological studies supported the hypothesis that surviving epithelial cells differentiate and proliferate after kidney injury (Bonventre, 2003), but the contribution of other cell populations to kidney repair remained unclear. Advances in stem cell biology led some researchers to put forth that bone marrow-derived cells directly replace the epithelial cells, although additional analysis revealed that earlier reports may have overestimated this phenomenon and that it may be explained by cell fusion. Other researchers identified adult stem/progenitor cells in the kidney and demonstrated that these cells are capable of regenerating the injured tubules. Genetic fate-mapping techniques provide a more direct understanding of cell fates and demonstrate that surviving tubular cells proliferate and replace the injured tubular epithelial cells. In this review, we discuss the current understanding of stem/progenitor cells in the kidney and the pathophysiology of kidney regeneration. Finally, we discuss the issues that remain in kidney regeneration research and the possible therapeutic applications for kidney disease.

KIDNEY REGENERATION

Bone Marrow Cells in Tubular Regeneration

The possible contribution of bone marrow cells to kidney repair was initially described in human transplant recipients. Poulsom et al. analyzed renal biopsy specimens from male patients who had received transplants of female kidneys. They found Y chromosome–positive cells within the tubular epithelial cells. They also confirmed that the transplantation of male bone marrow into lethally irradiated female mice resulted in the appearance of a small proportion of Y chromosome–positive cells in the recipient's tubular epithelium (Poulsom et al., 2001). Similarly, Gupta et al. reported that ∼1% of the renal tubular epithelium was Y chromosome positive in renal biopsy specimens taken from two male patients who had received transplants of female kidneys. These cells expressed epithelial cell markers but not CD45, a tyrosine phosphatase found in most hematopoietic cells (except for plasma cells and erythrocytes). Y chromosome–positive cells in the kidney were not found in patients who did not experience acute tubular necrosis following transplantation, indicating that bone marrow-derived cells migrate to the kidney in response to kidney injury (Gupta et al., 2002). To confirm the contribution of bone marrow cells to kidney repair, other researchers used transgenic approaches. Hematopoietic stem cells are lineage-uncommitted (Lin) bone marrow cells characterized by the expression of cell surface markers, stem cell antigen-1 (Sca-1), and c-kit (Morrison et al., 1997). Lin et al. purified BMSCs (Lin- Sca-1+ c-kit+) from male ROSA26 mice that express β-galactosidase ubiquitously and transplanted them into female nontransgenic mice subjected to ischemic-reperfusion injury (IRI). They identified the presence of LacZ-positive cells and Y chromosome–positive cells in proximal tubules (Lin et al., 2003). Kale et al. injected purified LacZ-positive BMSCs (Lin- Sca-1+) 12 hr before bilateral IRI, and observed LacZ-positive cells in proximal tubules. They also found that renal dysfunction was attenuated by the injection of BMSCs. Both groups concluded that bone marrow-derived cells have the capacity to differentiate into tubular epithelial cells following tubular injury (Kale et al., 2003).

Despite the results of these experiments, careful observations revealed rare contribution of bone marrow cells to kidney regeneration. In a follow-up study by Poulsom's group, bone marrow from male mice was transplanted into lethally irradiated female mice, and these mice were subjected to folic acid-induced acute tubular injury. The number of Y chromosome–positive tubular cells in the recipient mice was increased after the injury, but most of the cells involved in tubule repair were Y chromosome–negative endogenous tubular cells (Fang et al., 2005). Lin et al. produced transgenic mice that expressed enhanced green fluorescent protein (EGFP) specifically in renal tubular epithelial cells, and examined bromodeoxyuridine (BrdU) uptake after IRI in EGFP transgenic female mice that received bone marrow transplants from male wild-type mice. They found that only 11% of BrdU-positive cells were originated from donor bone marrow cells and that the majority of proliferating cells were derived from the recipient's tubular cells. Cells co-expressing GFP and vimentin (a marker of dedifferentiated tubular cells) were also found, indicating that tubular epithelial cells are the main source of renal repair and that these cells undergo dedifferentiation (Lin et al., 2005). Similarly, Duffield et al. showed that bone marrow-derived cells rarely contributed to renal repair. They labeled bone marrow cells derived from male mice with both β-galactosidase and GFP and injected them into female mice. They showed that, after IRI, more than 99% of GFP-labeled cells were leukocytes and that less than 1% of tubular cells were Y chromosome–positive cells. They also found that all the cells expressing β-galactosidase in the renal tubules were endogenous (Duffield et al., 2005). Recently, cell fusion between BMSCs and differentiated cells has been reported in various adult tissues (Terada et al., 2002). The previous results could be explained by the cell fusion between transplanted bone marrow-derived cells and tubular cells during renal damage. Taken together, it is now recognized that bone marrow cells rarely transdifferentiate into renal epithelial cells during kidney repair.

Lineage Tracing and Tubular Regeneration

Genetic fate-mapping is a powerful technique to trace the fate of certain cells during kidney development and injury. Cre-loxP technology enables genetic fate mapping in adult mice. This mapping involves a site-specific recombinase to activate the expression of a reporter molecule, such as β-galactosidase or fluorescent proteins. Because cell fate markers are not diluted during cell proliferation or transdifferentiation, we can trace the cell fates precisely (Duffield and Humphreys, 2011). By utilizing this technique, Humphreys et al. showed that intrinsic surviving tubular epithelial cells are responsible for the repair of injured nephrons. The Six2 gene is expressed exclusively in kidney progenitor cells that become various types of epithelial cells in the nephron, including podocytes, proximal and distal tubules, but neither cells of collecting ducts nor interstitial cells (Kobayashi et al., 2008). They crossed a Six2-Cre mouse to two different reporter mice, and successfully labeled the epithelial cells of the nephron, but not the nonepithelial interstitial cells. The authors subjected these mice to kidney injury and repair, and showed that there was no dilution of the genetic label in the tubules after the repair, indicating that surviving epithelial cells are responsible for the repair. To confirm these results, they also performed a complementary experiment by labeling interstitial fibroblasts but not epithelial cells using a FoxD1-Cre driver. After kidney injury and repair, there was no increase in labeled epithelial cells in the tubule, which suggests that an interstitial cell population does not contribute to the epithelial repair (Humphreys et al., 2008; Humphreys, 2011).

STEM CELLS IN ADULT KIDNEY

Stem cells are defined as cells that have the capacity to renew themselves and give rise to at least one type of mature differentiated cell. Progenitor cells are considered to be intermediate cells between stem cells and differentiated cells (Gupta and Rosenberg, 2008). Adult stem cells exist in many tissues and play an important role in normal cell turnover and the repair of injured tissues (Chagastelles and Nardi, 2011). It is considered that the organs previously believed to be post-mitotic, such as the heart and the brain, have their own stem cell compartments (Ott et al., 2007; Ming and Song, 2005). Recently, several groups successfully isolated and characterized adult kidney stem cells from renal tubules, the interstitium, and Bowman's capsule (Gupta and Rosenberg, 2008; Reule and Gupta, 2011; McCampbell and Wingert, 2012; Little and Bertram, 2009; Humphreys and Bonventre, 2007). These stem cells can differentiate into various cells in the kidney and have the potential to replace the injured cells. In addition to the mature tubular cells, endogenous renal stem cells may also participate in tubular regeneration.

Label-retaining Cells

One of the features that distinguishes stem cells from differentiated cells is their infrequent cell division. When these cells are labeled with markers such as BrdU and histone2B-GFP (H2B-GFP), they retain the label for a long period of time. This label retention is utilized to identify and isolate stem cells in the skin, stomach, and other organs (Yan et al., 2007), which are sometimes called label-retaining cells (LRCs). Two groups have demonstrated the existence of LRCs in the kidney. Oliver and co-workers administrated BrdU to rat and mouse neonates, chased them for more than two months, and found BrdU-retaining cells localized in the papilla. After ischemic insult, LRCs rapidly disappeared from the papilla, whereas many LRCs expressed Ki67 (a proliferation marker). The authors concluded that the renal papilla is a niche for adult kidney stem cells and that LRCs proliferate and migrate following ischemic injury (Oliver et al., 2004). In another study, they showed similar results using H2B-GFP transgenic mice (Oliver et al., 2009). H2B-GFP has several advantages over BrdU. H2B-GFP is more stable and easily detected than BrdU and it can label the cells in a cell cycle regardless of the stage. In contrast, BrdU can only label the cells in the S phase (Reule and Gupta, 2011). It was also demonstrated that H2B-GFP-positive cells proliferated and migrated to renal tubules following kidney injury (Oliver et al., 2009).

Maeshima et al. identified BrdU-retaining cells in adult rat kidney that were localized predominantly in proximal tubules. After ischemic insult, most of the BrdU-retaining cells were found in tubules and were positive for PCNA, indicating that these cells proliferate after injury. The researchers successfully isolated and cultured the population enriched with BrdU-retaining cells, and found that these cells initially expressed vimentin (a mesenchymal cell marker) and eventually became positive for E-cadherin (an epithelial cell marker) after multiple cell divisions (Maeshima et al., 2003). In the following study, they showed that LRCs have significant plasticity. When transplanted into the developing kidney of embryonic rat, LRCs became positive for proximal tubule and ureteric bud markers, indicating that LRCs were integrated into the epithelial components of nephrons in the process of kidney development (Maeshima et al., 2006).

Although BrdU labeling is a powerful technique to detect stem cells, it has considerable problems. The BrdU label may be released from dying cells and taken up by adjacent dividing cells (Humphreys and Bonventre, 2007). Vogetseder et al. demonstrated the possibility that BrdU-retaining cells are fully differentiated cells, not stem/progenitor cells. They reported that BrdU-retaining cells expressed differentiated tubule markers, such as Na-K-ATPase, NaPi-IIa, megalin, and PMP70 (Vogetseder et al., 2005). In a further study, they analyzed kidneys in healthy rats following 2 weeks of BrdU administration for label retention and cyclin D1 (G1 phase marker) expression. They found that proximal tubule cells in the S3 segment were positive for cyclin D1 and also retained BrdU (Vogetseder et al., 2008). These results showed that BrdU-retaining cells are fully differentiated cells rather than undifferentiated stem cells, and that they have the capacity of proliferation. Similarly, H2B-GFP labeling has some disadvantages because the labeling is diluted after massive cell proliferation. To obtain a better understanding of LRCs, a more accurate purification and characterization technique is needed.

Recently, some researchers have argued against the role of stem cells in kidney regeneration. Humphrey et al. showed that renal progenitor cells are not involved in the repair of injured proximal tubules. They used a DNA analog-based approach. They administrated two types of thymidine analogue (first CldU [5-chloro-2-deoxyuridine] and second IdU [5-iodo-2-deoxyuridine]) sequentially after IRI. If progenitor cells exist within a proximal tubule, the new epithelial cells during repair should have been positive for both CldU and IdU. However, they found that double-positive cells were uncommon, thus denying the contribution of progenitor cells in tubular regeneration (Humphreys et al., 2011). Song et al. used another approach. They marked papilla epithelial cells using mTert (mouse telomerase reverse transcriptase). mTert is the enzyme that regulates telomere length and is expressed in ES cells and several adult stem cells. Using mTert-GFP mice, they found that mTert-GFP-positive cells exist in the papilla and that these cells neither divide nor migrate out of the renal papilla during kidney repair. These data suggest that cells expressing mTert are not progenitor cells and are not involved in kidney repair (Song et al., 2011).

Stem Cell Surface Marker

Another approach for isolating putative stem cells is utilizing stem cell markers, including CD133, CD24, and Sca-1. CD133 is generally expressed in hematopoietic and progenitor cells as well as embryonic kidney (Miraglia et al., 1997; Corbeil et al., 2000). Bussolati and co-workers isolated CD133+ cells from normal adult human kidney. These cells expressed Pax-2 (an embryonic kidney marker), but not CD45 (hematopoietic lineage marker), c-Kit, or CD90 (stem cell markers), suggesting that they were of renal origin. These cells were capable of self-expansion and limited self-renewal, and could be differentiated in vitro and in vivo into epithelial and endothelial cells. When CD133+ cells were injected into mice 3 days after glycerol-induced AKI, they homed to the injured kidney and were incorporated into tubules (Bussolati et al., 2005). Sagrinati et al. found that a subset of parietal epithelial cells in adult human kidney exhibited co-expression of CD24 and CD133. These cells also expressed the stem cell–specific transcription factors Oct-4 and BmI-1, and comprised between 0.5% and 4% of cortical renal cell population. These cells had a high self-renewal potential and could be differentiated into proximal and distal tubules, as well as osteocytes and adipocytes in vitro. When these cells were injected into severe combined immune deficiency (SCID) mice with glycerol-induced AKI, they were incorporated into regenerating tubules and renal function was improved 7 to 10 days after injection (Sagrinati et al., 2006). Dekel and co-workers used the presence of stem cell marker Sca-1 and the absence of hematopoietic lineage markers to isolate nontubular multipotent stem cells from mouse kidneys. These cells exist in the interstitium of the kidney and are capable of differentiation into myogenic, osteogenic, adipogenic, and neural lineages in vitro. When injected into the renal parenchyma following IRI, these cells adopt a tubular phenotype (Dekel et al., 2006).

Selective Culture Condition

Selective culture conditions are useful for isolating stem cells (Bertoncello and Williams, 2004). Gupta et al. enzymatically isolated a unique population of cells from rat kidneys by using culture conditions similar to those used for bone marrow-derived cells. The surviving cells were called multipotent renal progenitor cells (MRPCs) and are characterized by long-term self-renewal without senescence and the expression of vimentin, Pax-2, and Oct-4 without any markers of MHC class or other differentiated cell markers. MRPCs exhibit plasticity and can be induced to multiple lineages in vitro, including endothelial cells, hepatocytes, and neurons. The cells can differentiate into tubular cells when injected into uninjured kidney or IRI kidney. It is proposed that MRPCs participate in the regenerative response of the kidney to acute injury (Gupta, 2006). However, this method has limitations due to the possible contamination of differentiated kidney cells and blood cells.

SP Cells

Stem cells extrude Hoechst dye and Rhodamine via ATP binding cassette (ABC) transporters (Challen and Little, 2006). These cells occupy a unique location in fluorescent-activated cell sorting (FACS) analysis and are called side population (SP) cells. SP cells are characterized as Hoechst-low cells, are isolated from many different tissues (including the liver, skeletal muscle, and heart), and contain multipotent stem cells (Challen and Little, 2006). Several groups isolated and characterized SP cells from the kidney. Iwatani and co-workers isolated SP cells from adult rat kidney that comprised approximately 0.03 to 0.1% of the kidney cells. However, these cells did not participate in kidney regeneration following gentamicin-induced nephropathy (Iwatani et al., 2004). Hishikawa and co-workers isolated SP cells from adult mouse kidneys that express Musculin/MyoR, a transcription factor found in skeletal muscle precursors. Musculin/MyoR-positive cells reside in the renal interstitium and are decreased following cisplatin-induced AKI. Infusion of kidney SP cells improved renal function in AKI models, but not in two chronic kidney disease models. It was also confirmed that kidney SP cells expressed a high level of leukemia inhibitory factor (LIF) and renoprotective factors, such as HGF, vascular endothelial growth factor (VEGF), and BMP7 (Hishikawa et al., 2005). Challen and co-workers isolated SP cells from adult mouse kidney that comprised 0.1 to 0.2% of kidney cells; these SP cells express genes involved in Notch signaling. These cells demonstrated multilineage differentiation potential in vitro, but were heterogeneous and included renal macrophages. Infusion of these cells improved renal function in adriamycin-induced nephropathy models (a focal segmental glomerulosclerosis model) (Challen et al., 2006). SP cells are considered to release renoprotective factors in a paracrine fashion; however, the exact function or composition of SP cells remains unclear.

Parietal Epithelial Cells

Mature podocytes are highly differentiated nondividing cells. It remains unclear whether podocyte injury in adulthood can be repaired. Glomerular epithelial stem cells may reside in the kidney and may be capable of regenerating podocytes. Some researchers hold that glomerular epithelial stem cells exist in the parietal epithelial cells (PECs) (Lasagni and Romagnani, 2010). Sagrinati et al. found that some PECs localized in the Bowman's capsule of adult human kidney have the features as stem cells (Sagrinati et al., 2006). This group divided PECs from the Bowman's capsule of human kidney into three subpopulations. Cells localized at the urinary pole expressing CD133 and CD24, but not podocalyxin (PDX), a differentiated podocyte marker, (CD133+, CD24+, PDX cells), could regenerate both tubular cells and podocytes. In contrast, cells localized between the urinary pole and the vascular pole expressing both progenitor and podocytes markers (CD133+, CD24+, PDX+ cells) could only regenerate to podocytes. Finally, cells localized at the vascular pole did not exhibit progenitor markers, but displayed phenotypic features of differentiated podocytes (CD133, CD24, PDX+ cells). Furthermore, the researchers evaluated the in vivo properties of these cells. Injection of CD133+, CD24+, PDX cells, but not CD133+, CD24+, PDX+ or CD133, CD24, PDX+ cells, into mice with adriamycin-induced nephropathy reduced proteinuria and improved chronic glomerular damage, suggesting that CD133+, CD24+, PDX cells are potential therapeutic targets for glomerular disorders characterized by podocyte injury (Ronconi et al., 2009). Appel et al. showed that the CD133+ PECs have the capacity to proliferate, migrate along the glomerular tuft, and differentiate to mature podocytes after podocyte injuries by using immunostaining, label retention, and transgenic mice approaches. They examined the cells at the base of the vascular pole adjacent to the podocytes and found that they were stained with both the parietal epithelial cell marker (claudin-1) and podocyte-specific markers (nestin, dipeptidyl peptidase IV, aminopeptidase A). BrdU staining of rat PECs supported the idea that PECs migrate to become podocytes. They performed genetic tagging of PEC utilizing PEC specific promoter (hPODXL1), and demonstrated the recruitment of podocytes from PECs in juvenile mice. They also demonstrated that the cells at the urinary pole of the Bowman's capsule migrate down to the proximal tubules and are involved in the turnover of tubular epithelium. Finally, they concluded that the cells localized at the border of different compartments have features of progenitor cells (Appel et al., 2009).

Although stem cells are promising candidates for therapeutic applications, there are serious concerns about their adverse effects. For example, myeloproliferative diseases are characterized as the aberrant proliferation of hematopoietic stem cells (Lasagni and Romagnani, 2010). Smeets and co-workers showed that the majority of cells present in the hyperplastic lesions of patients with collapsing glomerulopathy or crescentic glomerulonephritis exhibit CD133 and CD24 (glomerular epithelial stem cell markers). These results indicate that glomerular hyperplastic lesions may be derived from renal progenitor cells in the Bowman's capsule (Smeets et al., 2009).

REGENERATIVE MEDICINE

A large body of evidence has showed that MSCs can ameliorate renal injury (Humphreys and Bonventre, 2008; Tögel and Westenfelder, 2008). MSCs are fibroblast-like cells of stromal origin and can be isolated from virtually any tissue, including bone marrow, adipose tissue, and umbilical cord blood. Embryonic stem cells (ESCs) were the initial candidates of cellular therapies, but the clinical use of ESCs has some problems, including ethical issues, immunological rejections, and the risk of developing teratoma. On the other hand, MSCs are attractive for clinical use because of their availability from tissue, easy expansion in culture on a large scale, and the ability to differentiate into a variety of tissues. Furthermore, MSCs do not express blood-group antigens, MHC class I and II antigens, or major co-stimulatory molecules, and are immunologically tolerant. Previously, the renoprotective activity of MSCs had been explained by the direct contribution of MSCs to tubular regeneration (Morgi et al., 2004), but the direct contribution of MSCs is now considered to be very rare (Lin et al., 2005; Duffield et al., 2005; Tögel et al., 2005). Paracrine actions exerted by MSCs are considered to be the main mechanism protecting kidney injury (Tögel et al., 2007).

Bone marrow contains both hematopoietic stem cells (HSCs) and MSCs. Bone marrow–derived cells have been shown to be renoprotective after kidney injury (Kale et al., 2003). Morigi et al. showed that injection of MSCs, but not HSCs, significantly protected the rise in serum blood urea nitrogen (BUN) after cisplatin-induced renal injury in C57/B6 mice. They also found an increase in Ki67-positive tubule epithelial cells, indicating that MSCs protect against AKI by accelerating tubular proliferation (Morgi et al., 2004). Similar results were shown in other experimental models, including IRI (Duffield et al., 2005; Tögel et al., 2005; Lange et al., 2005), glycerol-induced nephropathy (Herrera et al., 2007), and glomerulonephritis induced by anti Thy1.1 antibody (Kunter et al., 2007).

When MSCs are administered into the suprarenal aorta or intravenously, they efficiently home to the injured kidney. Lange et al. labeled MSCs with iron oxide nanoparticles and administrated them systemically to rats following IRI. They could identify these cells in the kidney using magnetic resonance imaging (Lange et al., 2005). In a more detailed analysis, Tögel et al. fluorescently labeled MSCs and administrated them systemically to rats after IRI. They observed rat kidney using two-photon laser in vivo microscopy, and found that MSCs localized at both glomeruli and peritubular capillaries within 10 min of the injection (Tögel et al., 2005). The signals that regulate MSCs homing to the injured kidney are now under investigation, and some candidates, such as stromal-derived factor-1 (SDF-1) and CD44, have been reported (Herrera et al., 2007; Tögel et al., 2005). Tögel et al. examined the expressions and functions of SDF-1 and its receptor CXCR4 in the normal and injured kidney. SDF-1 is mostly expressed in distal tubules, and is upregulated in the kidney after renal injury. CXCR4 is expressed in both MSCs and the kidney in a similar distribution to that of SDF-1. In vitro and in vivo migration and homing assays showed that SDF-1 is an important mediator of MSCs homing to the injured kidney (Tögel et al., 2005). Another promising mediator of MSC homing is CD44, which mainly binds to hyaluronic acid. Herrera and co-workers showed that CD44 is expressed on MSCs, and the injection of MSCs together with an anti-CD44 blocking antibody or the injection of CD44 knockout MSCs resulted in reduced migration to the injured kidney compared to the controls (Herrera et al., 2007).

Previously, exogenously administrated MSCs were considered to differentiate directly into tubular cells (Morgi et al., 2004). However, more recent studies have demonstrated that the contribution of exogenous MSCs to tubular incorporation is rare (Lin et al., 2005; Duffield et al., 2005; Tögel et al., 2005). Many researchers believe that MSCs protect against tubular injury via paracrine effects (Tögel et al., 2007). MSCs are known to secrete a variety of growth factors, such as VEGF, IGF-1, and HGF (Tögel et al., 2007; Imberti et al., 2007; Tögel et al., 2009). Tögel et al. found significant levels of VEGF, HGF, and IGF-1 in MSC-conditioned media, which was capable of enhancing endothelial cell proliferation and differentiation (Tögel et al., 2007). Imberti et al. showed that MSCs protect proximal tubular epithelial cells (PTECs) from cisplatin-injured injury via IGF-1. They found that knocking down IGF-1 expression in MSCs by specific antibody or small interfering RNAs (siRNAs) resulted in a significant decrease in PTEC proliferation and increased apoptosis (Imberti et al., 2007). Tögel et al. showed that MSCs-derived VEGF is also important in protecting against kidney injury after AKI (Tögel et al., 2009).

MSCs have also been shown to play a renoprotective role in chronic kidney disease models. Ninichuck and co-workers evaluated the effect of MSCs that were injected into mice deficient for the α3 chain of type IV collagen (COL4α3), a model of hereditary progressive glomerulonephritis, Alport syndrome. In COL4α3-null mice injected with MSCs weekly from 6 to 10 weeks of age, interstitial fibrosis and peritubular capillary loss were prevented, whereas there were no differences in renal functions as measured by serum BUN and creatinine levels and proteinuria. They also found that VEGF mRNA levels increased in the kidneys of MSCs-treated mice (Ninichuk et al., 2006). Semedo et al. used another chronic kidney disease model (5/6 nephrectomy remnant kidney model in rat) and showed that interstitial fibrosis was attenuated in MSC-treated rats. Furthermore, expression of interleukin (IL)−6, interferon-γ, transforming growth factor-β (TGFβ), and tumor necrosis factor-α (TNFα) was significantly decreased, whereas expression of IL-4 and IL-10 was increased after MSC treatment, indicating that MSCs have prominent immunomodulatory functions (Semedo et al., 2009).

While the injection of MSCs has been shown to be renoprotective in several kidney disease models, several side effects have recently been reported. Krunter et al. analyzed the long-term side effects of MSC administration in a rat model of glomerulonephritis, and showed that ∼20% of the glomeruli contained single adipocytes or adipocyte clusters with pronounced surrounding fibrosis. This finding indicates MSCs' maldifferentiation to adipocytes and subsequent fibrosis (Kunter et al., 2007). Other unwanted side effects of MSCs were found in the lungs, where injected MSCs were trapped and formed cysts containing collagen deposits (Anjos-Afonso et al., 2004).

Despite these findings, cellular therapy using MSCs is considered basically safe (Tögel et al., 2009), and several clinical trials are now ongoing. In some diseases, the effectiveness and safety of MSCs in humans has been demonstrated. Severe graft-versus-host disease (GVHD) is a life-threatening complication after allogeneic transplantation with hematopoietic stem cells. Patients with steroid-resistant severe acute GVHD were treated with MSCs in a phase II trial. Fifty-five patients were treated with an injection of MSCs expanded in vitro. Thirty patients had a complete response (disappearance of all symptoms of acute GVHD) and nine showed improvement following infusion. No patients had serious side effects during the treatment (Le Blanc et al., 2008). The effect of MSCs on AKI is now under examination. A phase I clinical trial has been conducted to investigate the safety and efficacy of MSCs for patients at high risk of developing AKI after open-heart surgery (Anna Gooch et al., 2008).

CONCLUSIONS

Many researchers have come to realize that intrinsic surviving tubular epithelial cells are responsible for the repair of injured nephrons. Stem cells in the kidney have been isolated and characterized by different methods, but the features of these cells are not yet fully understood. Due to the lack of definitive markers for stem cells in adult kidney, it is unclear whether these cells actually contribute to kidney regeneration in vivo. Further exploration of the specific cell surface markers and analysis of the exact localization of renal stem cells is essential.

Unlike mammalian nephrons, the nephrons of zebrafish increase throughout their lifespan and can be regenerated after injury (Zhou et al., 2010). Recently, nephron stem/progenitor cells that have kidney regeneration capacity have been identified in zebrafish (Diep et al., 2011), and some groups have identified certain pathways and compounds responsible for the stem/progenitor cell expansion, which might be a candidate drug for the treatment of AKI in mammalian kidney (Hukriede, 2011). Zebrafish are a useful model for studying the behavior of stem/progenitor cells because of their transparent tissues, easy manipulation of their well-understood genome, and direct applicability to high-throughput screening. Further utilization of the model may provide important insights into the development of novel therapeutic approaches for human kidney disease.

Ancillary