Author contributions: K.M. and I.S.: conception and design, Collection and assembly of data, data analysis and interpretation, manuscript writing, and final approval of manuscript; H.I., T.N., Y.M., Y.A., F.I., M.M., H.Y., G.Y., J.T., R.Y., T.S., M.-C.-i.L., K.C., and H.S.: collection and assembly of data and final approval of manuscript; Y.K. and K.T.: data analysis and interpretation and final approval of manuscript; Y.M.: collection and assembly of data, data analysis and interpretation, and final approval of manuscript.
Disclosure of potential conflicts of interest is found at the end of this article.
First published online in STEM CELLSEXPRESS July 10, 2012.
Stem cell-based therapy has been proposed as a promising strategy for regenerating tissues lost through incurable diseases. Side population (SP) cells have been identified as putative stem cells in various organs. To examine therapeutic potential of SP cells in hypofunction of exocrine glands, SP cells isolated from mouse exocrine glands, namely, lacrimal and salivary glands, were transplanted into mice with irradiation-induced hypofunction of the respective glands. The secretions from both glands in the recipient mice were restored within 2 months of transplantation, although the transplanted cells were only sparsely distributed and produced no outgrowths. Consistent with this, most SP cells were shown to be CD31-positive endothelial-like cells. In addition, we clarified that endothelial cell-derived clusterin, a secretory protein, was an essential factor for SP cell-mediated recovery of the hypofunctioning glands because SP cells isolated from salivary glands of clusterin-deficient mice had no therapeutic potential, whereas lentiviral transduction of clusterin restored the hypofunction. In vitro and in vivo studies showed that clusterin had an ability to directly inhibit oxidative stress and oxidative stress-induced cell damage. Thus, endothelial cell-derived clusterin possibly inhibit oxidative stress-induced hypofunction of these glands. Stem Cells2012;30:1925–1937
Stem cells are multipotent cells involved in organ maintenance and repair after injury. Therefore, stem cells are expected to be a promising cell source for regenerative medicine to recover damaged organ functions. Various sorts of tissue stem cells can be purified using cell-surface markers and flow cytometry. For other tissues whose stem cell-specific surface markers are unknown, a technique based on the exclusion of the DNA-binding dye Hoechst 33342 has been applied to identify and purify side population (SP) cells using fluorescence-activated cell sorting (FACS) [1, 2]. SP cells isolated from adult bone marrow can reconstitute the irradiated bone marrow of dystrophin-mutant mice and participate in muscle repair . In addition, mouse embryonic stem cells (ESCs) have the ability to efflux Hoechst . Thus, SP cells are highly enriched for stem cell activity [2, 5] and have been identified in various tissues, such as skeletal muscle, mammary glands, testis, retina, skin, heart, brain, liver, and lung [6, 7]. However, the functions of these cells remain unclear, and some controversy surrounds them [8–11]. The SP cell phenotype is determined by Bcrp1, also known as ATP-binding cassette subfamily G member 2 . Bcrp1 transports a wide range of substrates, including chemotherapeutics . The expression of Bcrp1 has been reported in several stem cell populations, including hematopoietic stem cells and ESCs. However, Bcrp1 is also expressed in a variety of differentiated cells in normal tissues, suggesting that SP cells are not necessarily a stem cell-rich population .
In this study, to clarify the function of SP cells in exocrine glands, we characterized SP cells from lacrimal and salivary glands and examined their therapeutic potential. Both lacrimal glands and salivary glands are convenient to use for such studies because they are superficially located, and their functions can be relatively easily evaluated by measuring their respective secretions. In addition, the salivary gland is also a classic model for analyzing branching morphogenesis in exocrine glands .
Here, we determined that SP cells isolated from lacrimal and salivary glands are endothelial cell-rich population and have therapeutic potential for glandular hypofunction, which is mediated by the paracrine factor, clusterin without reconstituting glands. Several types of stem and progenitor cells have been applied into clinical trials. However, it is still questionable whether the transplanted cells can effectively reconstitute the organ. On the contrary, paracrine effects of cells including stem and other cells to recover the function have been recently clarified [16, 17]. This study demonstrated that clusterin might be a new target molecule to rescue the hypofunctioning of exocrine glands.
Isolation of SP Cells
SP cells were isolated as described previously . Briefly, epithelial organoids were isolated from the lacrimal and salivary glands. The cells were disaggregated by pipetting in 0.25% trypsin-EDTA (Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), followed by filtration through a 40-μm mesh (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com), to obtain a predominantly single-cell suspension. This suspension was then resuspended at 1 × 106 cells per milliliter in prewarmed (37°C) Hanks' balanced saline solution/5% fetal bovine serum (FBS). Then Hoechst 33342 (Wako Chemicals, Osaka, Japan, http://www.wako-chem.co.jp) was added at a final concentration of 5 μg/ml, and reserpine (Daiichi Pharmaceutical, Tokyo, Japan, http://www.jp-no1.co.jp/) was added to one aliquot of that dye at a final concentration of 50 μM. The cells were incubated for 90 minutes at 37°C with occasional agitation. Thereafter, they were washed with cold medium and resuspended at 1 × 106 cells per milliliter, followed by the addition of propidium iodide (Wako Chemicals) at a final concentration of 2 μg/ml. The SP and main population (MP) cells were sorted on a BD FACS Vantage SE cell sorter (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com) into tubes. The machine was equipped with two Coherent 90 C-4 argon ion lasers (Coherent, Santa Clara, CA, http://www.coherent.com), one with visible optics set to 488 nm and one with multiline UV optics (333.6–333.8 nm). Hoechst 33342 fluorescence was measured at both 424/44 nm and above 670 nm (split by a 610-nm short-pass dichroic mirror), both from UV excitation. Dead cells were excluded by the detection of propidium iodide fluorescence measured at 564–606 nm. SP and MP cells were transplanted into irradiated mice. SP and MP cells were also purified from the lacrimal or salivary glands of heterozygous enhanced green fluorescent protein-transgenic (GFP-Tg) mice in the C57BL/6 background . Two weeks after irradiation, the cells were then directly injected into the right lacrimal or submandibular gland of each experimental mouse via a microinjector. The number of isolated SP cells from lacrimal glands of 20 mice ranged from 2 × 104 to 4 × 104 cells (average 3 × 104 cells) and that from salivary glands of five mice ranged from 7 × 104 to 9 × 104 cells (average 8 × 104 cells). In addition, the exact numbers of SP and MP cells of lacrimal glands, which were injected into a lacrimal gland, were adjusted to approximately 5 × 103 cells, while those of salivary glands were adjusted to approximately 1 × 104 cells. Therefore, in every experiment, we could inject SP cells isolated from lacrimal glands into four or seven mice and SP cells isolated from salivary glands into seven or eight mice. The mice were sacrificed for analysis 8 weeks after the injection, and the engrafted cells were identified by the presence of GFP using a fluorescence microscope. Sorted SP cells were also cultured on cell culture dishes with Dulbecco's modified Eagle's medium supplemented with 10% FBS and incubated at 37°C in a humidified atmosphere of 5% CO2 air. After incubation for 1 week, the cells were observed using phase-contrast microscopy and photographed.
Each mouse was anesthetized with an intraperitoneal injection of 60 mg/kg of sodium pentobarbital. The animals in the irradiated groups were then placed in defined positions. Each mouse was exposed to a single acute 10 MV dose of x-rays (MEVATRON 74 DX40; Toshiba Medical Systems, Tokyo) at a dose rate of 3 Gy/minute and a distance of 1,000 mm. The position of each lacrimal and salivary gland was confirmed by computed-tomographic scanning under the same conditions as that for the irradiation. The effective radiation dose to the lacrimal and salivary glands was set using the percentage depth dose and was over 95% of the maximum dose delivered.
Electronic Spin Resonance Spectrometry
The scavenging ability of superoxide and hydroxyl radicals was determined by electronic spin resonance (ESR) spectrometry. Both radicals were trapped by 5-(2,2-dimethyl-1,3-propoxy cyclophosphoryl)-5-methyl-1-pyrroline N-oxide (CYPMPO), and CYPMPO adducts were measured .
Statistical evaluation was performed by one-way ANOVA or Student's t test. Data are presented as means ± SD.
SP Cells Are Present in Mouse Lacrimal and Salivary Glands
We identified SP cells in the mouse lacrimal and salivary glands to characterize their functions. Cells isolated from these glands were stained with Hoechst 33342 and sorted by FACS. Density dot-plot analysis (Hoechst red vs. Hoechst blue) confirmed the presence of SP cells in both the lacrimal and the salivary glands (Fig. 1A, 1B). Isolation of the SP cell population by this procedure was prevented by treatment with reserpine, a potent functional inhibitor of several ABC transporters, including Bcrp1. SP cells comprised approximately 0.5% (range, 0.2%–1.0%) of the total cell population in both glands. The SP cells also expressed Bcrp1 protein (Fig. 1C). Next, we examined the percentage of CD45-positive or hematopoietic lineage marker-expressing cells in the SP cell population to account for possible contamination of the glandular SP cells by SP cells derived from peripheral blood and bone marrow. The frequency of these contaminating SP cells was negligible because the percentage of cells positive for both markers was quite low (Fig. 1D). Next, to examine the distribution of Bcrp1-expressing cells in detail, immunofluorescence staining with a ductal cell-specific marker, cytokeratin 18 (CK18), and an endothelial cell-specific marker, CD31, was performed because Bcrp1 is expressed in some epithelial cells and microvessel endothelial cells [5, 14]. Positive staining for Bcrp1 was found in most microvessel endothelial cells around the small ducts and some ductal cells (Fig. 1E).
SP Cells Display Higher Expression of Sca-1 than MP Cells
Sca-1 expression has been reported in SP cells from various tissues . By RT-PCR analysis, SP cells demonstrated higher expression of Bcrp1 and Sca-1 mRNAs than MP cells in both the lacrimal (Fig. 2A, left panel) and the salivary (Fig. 2A, right panel) glands. In addition, gene expression of Sca-1 was also examined by real-time polymerase chain reaction (PCR). Expression level of Sca-1 in SP cells in lacrimal glands was approximately 3.2-folds higher than that of MP cells and that in SP cells in salivary glands was approximately 4.5-folds higher than that of MP cells (Fig. 2B). By double immunofluorescence analysis with Sca-1 and CK18, CD31, or Bcrp1, we determined that the Sca-1-positive cells were also localized in most of the microvessel endothelial cells around the small ducts and some ductal cells, displaying a similar distribution as Bcrp1-positive cells (Fig. 2C, 2D).
SP Cell Transplantation Recovers the Irradiation-Induced Hypofunction of Lacrimal and Salivary Glands
To verify whether SP cells from the lacrimal and salivary glands may also have therapeutic potential and/or stem cell activity (e.g., the potential for cell reconstitution), we isolated SP cells from both glands in transgenic mice expressing GFP and transplanted them into mice with irradiation-induced hypofunction of the respective glands. Irradiation causes dysfunction of hematopoietic and melanocytic stem cells [22, 23]. Irradiated patients with head and neck cancer present with salivary gland hypofunction due to atrophy of the glands . Therefore, we expected that irradiation would induce dysfunction of the endogenous stem cells in the glands and examined whether SP cell transplantation could restore their functions. The secretory ability of the exocrine glands was quantified by pilocarpine stimulation after localized irradiation (15 Gy) to the head and neck region, including lacrimal and salivary glands. SP cells from both glands were transplanted into each gland 2 weeks after irradiation. Both the lacrimal and the salivary secretions were restored in the SP cell-transplanted mice 4 and 8 weeks after transplantation compared with their functions in the control and MP cell-transplanted mice (Fig. 3A).
SP Cells Have No Ability to Reconstitute the Lacrimal and Salivary Glands
We next examined the distribution of GFP-positive cells after SP and MP cell transplantation. The GFP-positive transplanted cells were only sparsely distributed in stromal tissue of SP cell-injected lacrimal and salivary glands and there was no apparent outgrowth of these cells, whereas GFP-positive cells were not detected in either MP cell-injected lacrimal or salivary glands (Fig. 3B). In addition, most of the GFP-positive cells showed positive staining for CD31 but not CK18 and formed neither microvessel structure with an apparent lumen nor epithelial structure consisting of acinar and duct. These results suggested that while the SP cells from both glands had therapeutic potential, they may not be a population containing epithelial stem cells. To confirm this, expression of CK18, which is a marker for epithelial cells, was examined on SP cell-injected lacrimal and salivary glands. Indeed, there was no significant difference of CK18 expression between SP cell-injected and phosphate-buffered saline (PBS)-injected lacrimal and salivary glands (Fig. 3C). In fact, SP cells are a heterogeneous population consisting of epithelial stem cells, mesenchymal stem/stromal cells (MSCs), or endothelial cells, differing with organ type [10, 25]. To further examine whether the SP cells were enriched in stem cell activity, the BrdU label retention approach was used because long-term BrdU-retaining cells (LRCs) are plausible stem cells [26, 27]. Most of the BrdU-positive staining was localized in the nuclei of the ductal cells of both glands, and little positive staining was found in the nuclei of the acinar cells (Supporting Information Fig. S1A). Thus, the percentages of LRCs in the SP and MP cells from the lacrimal glands were 4.0% and 6.9%, and 4.7% and 2.4% from the salivary glands (Supporting Information Fig. S1B and Table S1). There was no significant difference between these percentages in the SP cells and MP cells in either gland. In addition, to examine whether SP cells in both glands are derived from the epithelial lineage including epithelial stem/progenitor cells, we examined the expression of an epithelial progenitor marker, cytokeratin-5 (CK5), and a ductal cell marker, CK18. Importantly, the expression levels of both markers in SP cells in these glands were not detectable or were much lower than those in MP cells in these glands (Fig. 4A). In addition, few SP cells in lacrimal and salivary glands expressed CK5 and CK18 on the cytospun specimens, whereas a lot of MP cells expressed them (Fig. 4B). Therefore, the SP cells in these glands may be derived from a mesenchymal lineage. SP cells in the adult lung have been reported to have MSC potential . Additionally, we have shown that endometrial SP cells exhibit endothelial cell markers . Therefore, to examine whether SP cells in lacrimal and salivary glands include these mesenchymal populations, the expression of human MSC markers, CD105, CD73, and CD90, and an endothelial cell marker, CD31, was examined by flow cytometry. We also searched for double positive cells for Sca-1 and PDGFalpha, which we have recently reported as a mouse MSC marker . Most SP cells in the lacrimal and salivary glands were CD105-, CD31-, and Sca-1-positive, but CD73-negative (Fig. 4C). The percentage of double positive cells for Sca-1 and PDGFRalpha was less than 1% in both glands (Fig. 4D). In addition to these findings, our immunohistochemical analysis indicated that most Bcrp1-positive cells in the lacrimal and salivary glands were CD31-positive endothelial cells (Fig. 1E). Thus, SP cells in both glands mainly consist of endothelial cells and not MSCs.
The SP Cell-Specific Secretory Protein Clusterin Is an Essential Factor for the Recovery of Hypofunctioning Glands
Microvessel protection can prevent irradiation-induced salivary hypofunction . Because SP cells in both glands contained endothelial cells, production of microvessels by transplanted SP cells might recover the irradiation-induced hypofunction of lacrimal and salivary glands without reconstitution of the glands. After 1 week of culture, most SP cells from both glands did not grow, although several colonies of proliferating cells formed sheet-like structures on tissue culture dishes freshly (Supporting Information Fig. S2A). However, isolated SP cells demonstrated no tube-like structures on Matrigel, although a mouse vascular endothelial cell line formed tube-like structures (Supporting Information Fig. S2B). In addition, to examine the ability of SP cells to produce microvessels in vivo, microvessel density (MVD) was measured in each gland 8 weeks after SP or MP cell transplantation. As shown in Supporting Information Fig. S2C, there was no significant difference in MVD between SP cell-, MP cell-, and PBS-injected glands. In addition, to evaluate microvessel promotion by SP cell transplantation, gene expression of CD31 in SP cell-injected salivary glands was compared with that in PBS-injected salivary glands. There was no significant difference in expression of CD31 between SP cell-injected and PBS-injected salivary glands (Supporting Information Fig. S2D). Thus, recovery effect by SP cell transplantation was not mediated through production of microvessels. Therefore, to explain the restored functions of the damaged glands following SP cell transplantation, we postulated that instead of cell reconstitution by the SP cells, soluble factors secreted by the SP cells restored the irradiation-induced damage of the cellular molecules involved in secretion. To identify these soluble factors in the SP cells of the lacrimal and salivary glands, we used microarray glass slides with an array of 15,000 cDNAs (Supporting Information Experimental Procedures). The selected lists of differentially expressed genes between the SP and MP cells in the lacrimal and salivary glands are shown in Supporting Information Tables S2 and S3. Importantly, significantly greater expression of clusterin was observed in the SP cells compared with the MP cells obtained from both the lacrimal and the salivary glands. To verify the relative expression levels of the genes derived from the microarray analysis, we performed RT-PCR and quantitative real-time RT-PCR analysis on the cDNA obtained from the SP and MP cells (Fig. 5A, 5B). The results confirmed the expression of the clusterin gene. Immunofluorescence analysis also revealed clusterin-positive signals in the SP cells but not in the MP cells (Fig. 5C). In addition, immunohistochemical analysis demonstrated that capillary endothelial cells stain positively for clusterin (Fig. 5D). To confirm this result, CD31-positive and CD31-negative cells were sorted from mouse salivary glands and the expression of clusterin mRNA by real-time RT-PCR analysis. The percentage of CD31-positive cells was 2.57% (Fig. 5E). The expression level of clusterin in CD31-positive cells was approximately sixfold higher than that in CD31-negative cells (Fig. 5F).
Next, it was important to ascertain whether the SP cells continued to express clusterin even after SP cell transplantation because we postulated that clusterin secreted from SP cells was one of the protective factors against irradiation-induced cell damage. For this evaluation, real-time RT-PCR analysis was conducted, which confirmed the expression of clusterin mRNA in the SP cell-injected glands. Significantly, more pronounced expression of clusterin was found in the SP cell-injected lacrimal and salivary glands than in the PBS-injected glands (Fig. 5G). The expression level in the PBS-injected irradiated glands was also slightly increased compared with that in the nonirradiated glands, suggesting that irradiation may also induce endogenous expression of clusterin, as reported previously .
Next, to examine whether clusterin is indispensable for the recovery of hypofunctioning glands, SP and MP cells isolated from the salivary glands of clusterin-deficient mice were injected into the glands of irradiated mice. Clusterin-deficient mice display a normal phenotype . Histological analysis of the lacrimal and salivary glands in these animals did not reveal any significant differences between the clusterin-deficient and wild-type mice (Supporting Information Fig. S3). Density dot-plot analysis (Hoechst red vs. Hoechst blue) confirmed the presence of SP cells in the salivary glands of the clusterin-deficient mice at 0.62% (Supporting Information Fig. S4A). However, intracellular reactive oxygen species (ROS) in SP cells in salivary glands of clusterin-deficient mice showed slight increase compared with that of wild-type mice (Supporting Information Fig. S4B). Moreover, interestingly, the percentage of Sca-1-and CD31-positive cells in SP cell population was not different between them, but fluorescent intensity of Sca-1 on SP cells in salivary glands of clusterin-deficient mice was lower than that of wild-type mice (Supporting Information Fig. S4C). This means the possibility that clusterin regulates cell-surface expression of Sca-1. SP cells and MP cells isolated from these mice were transplanted into irradiated salivary glands. Importantly, transplantation of neither SP nor MP cells from the clusterin-deficient mice produced recovery of the hypofunctioning salivary glands (Fig. 6A).
Moreover, to determine whether overexpression of clusterin alone can produce recovery of the hypofunctioning salivary glands, we directly injected a recombinant lentivirus vector expressing clusterin (Lenti-Clu) or GFP as a control (Lenti-U6i) into the irradiated salivary glands (Supporting Information Fig. S5A). The efficiency of infection was evaluated by the percentage of GFP-positive cells 1 week after Letni-U6i infection. Approximately 16% of the total cells expressed GFP (Supporting Information Fig. S5B). Furthermore, real-time RT-PCR analysis revealed a 3.8-fold higher expression of clusterin in the Lenti-Clu-injected salivary glands than that in the Lenti-U6i-injected salivary glands (Supporting Information Fig. S5C). At 4 weeks, the secretion of saliva in the Lenti-Clu-injected mice was significantly higher than that in the Lenti-U6i-injected mice (Fig. 6B). Next, we examined the expression of aquaporin-5 and M3-muscarinic acetylcholine receptor (M3R) (both of which are involved in water secretion) and the secretory protein amylase and β2-adrenergic receptor (β2-AdR) involved in protein secretion between PBS-, SP cell-, Lenti-Clu-injected, and Lenti-U6i-injected salivary glands. The expression levels of M3R, β2-AdR, and amylase in SP cell- and Lenti-Clu-injected glands were significantly higher than those in PBS- and Lenti-U6i-injected glands (Fig. 6C). Furthermore, the expression levels of M3R and β2-AdR genes in SP cell-injected glands were higher than those in Lenti-Clu-injected glands. Therefore, Lenti-Clu-injection only partly recovered the hypofunction of the glands compared with SP cell-injection. Taken together, these data indicate that clusterin expression in the SP cells is indispensable for the recovery of hypofunctioning salivary glands.
Clusterin Inhibits ROS-Induced Cell Damage As a Radical Scavenger
Clusterin, also known as apolipoprotein J or testosterone-repressed prostate message 2, is a ubiquitous secretory glycoprotein that occurs in two forms: the nonglycosylated intracellular form (cClu) and the highly glycosylated secretory form (sClu). sClu is a heterodimer consisting of α- and β-chains and has various functions, for example, extracellular chaperoning and antiapoptotic function in response to genotoxic agents such as irradiation [33, 34]. Irradiation-induced impairment of secretory function is caused by ROS, including free radical-mediated damage of secretion-related proteins such as ion transporters, without any associated loss of cells at an early period [35–37]. However, the mechanism underlying the effect of clusterin in reversing the damage caused by irradiation-generated ROS still remained obscure. To clarify the functions of SP cell-derived clusterin, we established mouse embryonic fibroblast cells stably expressing clusterin. As shown in Figure 7A, clusterin-expressing STO cells (STOclu) and control cells (mock STO) were generated by transfection of a clusterin expression vector and empty vector, respectively. Western blot analysis revealed the presence of cClu in the whole-cell lysates and of sClu in the cell culture medium of STOclu, whereas neither endogenous cClu nor sClu was detected in the mock STO. Under reducing conditions, sClu was dissociated into α- and β-chains. To examine the ability of clusterin to inhibit the cellular damage induced by ROS, we treated STOclu and mock STO cells with various concentrations of H2O2 and then measured cell viability. At 0.1 mM H2O2, the cell viability was approximately 80% in STOclu and 20% in mock STO (Fig. 7B). Additionally, a recent report states that clusterin-knockdown cells are significantly more sensitized to oxidative stress induced by H2O2 , although the underlying mechanisms were not clarified. We postulated that clusterin might inhibit intracellular ROS production and were subsequently able to validate this hypothesis. To examine whether inhibition of cell death in the STOclu line was mediated by scavenging of intracellular ROS produced by exposure to H2O2, the accumulation of ROS was measured by FACS after staining the cells with a fluorescent indicator dye, CM-H2DCFDA. STOclu had significantly lower levels of ROS compared with mock STO cells (Fig. 7C). Consistent with this observation, the extent of protein carbonylation in STOclu after treatment with H2O2 was much lower than that in the mock STO, suggesting the possibility of ROS-mediated damage of secretion-related proteins, such as ion transporters (Fig. 7D). Next, to clarify the role of sClu in ROS-induced cell death, recombinant clusterin was produced in a baculovirus system (Supporting Information Fig. S6). The intracellular ROS level and cell death in NIH3T3 cells cultured in the presence of various concentrations of recombinant clusterin were examined. Interestingly, clusterin rescued the cells from death in a dose-dependent manner (Fig. 7E). By ESR, recombinant clusterin was shown to directly scavenge O2− and HO· (Fig. 7F, 7G), suggesting that sClu may have both the ability to inhibit cell death by scavenging extracellular ROS generated by Fenton's reaction in the culture medium and the ability to inhibit damage to the cell membrane. Next, to examine whether SP cell transplantation inhibits oxidative stress-induced cell damage in vivo, immunohistochemical detection of 8-hydroxy-2′deoxyguanosine (8-OHdG) in PBS- and SP cell-injected salivary glands of irradiated mice was performed. The intensity of 8-OHdG-positive stainings was apparently decreased in SP cell-injected lacrimal and salivary glands compared with that in PBS-injected glands although positive nuclear staining was found in ductal and acinar cells of both glands (Fig. 7H, left panel), suggesting that SP cell transplantation restored the hypofunction of salivary glands in irradiated mice through inhibition of oxidative stress-induced cell damage. In addition, to examine whether clusterin also can inhibit the oxidative stress-induced cell damage, immunohistochemical detection of 8-OHdG in Lenti-Clu- and Lenti-U6i-injected salivary glands of irradiated mice was performed. Although positive nuclear stainings were found in ductal and acinar cells of the glands, the intensity of 8-OHdG-positive stainings was apparently decreased in Lenti-Clu-injected salivary glands compared with that in Lenti-U6i-injected glands (Fig. 7H, right panel). These results suggest that the effect of SP cell transplantation on the recovery of hypofunction of irradiated glands is possibly mediated through clusterin.
In this study, we characterized SP cells in mouse lacrimal and salivary glands. SP cells are reported to contain multipotent stem cells in many tissues. Therefore, we expected SP cells in both glands to be stem cell-enriched cell populations. However, SP cells in lacrimal and salivary glands did not display stem cell activity, that is, the ability to reconstitute host glands after their transplantation. We then expected SP cells in both glands to be differentiated epithelial cells because both glands are solid tissues composed of epithelial cells, including acini and ducts. However, SP cells in lacrimal and salivary glands expressed Bcrp1, Sca-1, CD105, and CD31, but not CK5 or CK18, suggesting that they predominantly consisted of endothelial cells. Indeed, this finding is consistent with the lack of reconstitution of glands by SP cells. Similar populations were also found in SP cells from several tissues, including brain, lung, and heart [5, 25, 39, 40]. However, interestingly, SP cells from these tissues contain bone marrow-derived SP cells to some degree, while SP cells in lacrimal and salivary glands have few bone marrow-derived SP cells. In addition, lung SP cells also contain MSCs but lacrimal and salivary glands do not. Therefore, the characteristics of SP cells appear to have organ specificity, and their therapeutic potential is expected to be different.
Microvessel protection prevented irradiation-induced salivary hypofunction . In addition, SP cells from dental pulp displayed vasculogenesis in an ischemic model . Therefore, SP cells in lacrimal and salivary glands, which were enriched in endothelial cells, were expected to recover the irradiation-induced hypofunction of the glands by vasculogenesis. However, MVD was not affected by SP cell transplantation. Endothelial cells have also been reported to have therapeutic potential in paracrine systems . In our study, a SP cell-derived secretory protein, clusterin, was identified as an indispensable factor to recover the hypofunction of the glands. In addition, clusterin expression by CD31-positive cells was much higher than that of CD31-negative cells, suggesting that SP cells expressing clusterin are endothelial cells. Thus, SP cell-derived clusterin in lacrimal and salivary glands was from CD31-positive endothelial cells. Consistent with this result, a SAGE library of vascular endothelial cells from normal mouse brain displays comparatively high expression of clusterin and Bcrp1 (http://cgap.nci.nih.gov/SAGE/SAGELibInfo?LID=2186&ORG=Mm). By contrast, those from kidney exhibit comparatively low expression of both (http://cgap.nci.nih.gov/SAGE/AbsLL?FORMAT=text&LID=2188&MIN=0&MAX=0&ORG=Mm&METHOD=SS10,LS10). Therefore, the clusterin expression level of endothelial cells may depend on the tissue of origin. We expected that the function of SP cells in these glands would be similar to the function of SP cells in other organs. However, the expression of clusterin was specific to the SP cells of the lacrimal and salivary glands and was not found in the SP cells of other organs (Supporting Information Fig. S7). This finding is also consistent with previous reports, which describe that endothelial cells within each organ are functionally different [17, 42].
Deficiency of clusterin increases cardiac damage induced by experimental autoimmune myocarditis and renal injury induced by ischemia-reperfusion [32, 43]. Thus, clusterin protects cells from stress-induced damage, including inflammation and oxidative stress. Clusterin has two forms, cClu and sClu, which have different functions. cClu is localized in the cytoplasm and inhibits apoptosis induced by genotoxic agents via interactions with Bax, while sClu also protects against oxidative stress-induced cell death [33, 44, 45]. Irradiation transiently increases cellular and extracellular ROS levels through water hydrolysis, which can cause DNA damage and mitochondrial dysfunction . Mitochondrial dysfunction causes continuous generation of ROS . Therefore, SP cell-secreted clusterin possibly scavenges ROS after irradiation. Consistent with this, in this study, intracellular ROS level of SP cells isolated from clusterin-deficient mice increased compared with that of SP cells isolated from wild-type mice and recombinant clusterin directly scavenged ROS. However, the possibility remains that clusterin-related proteins are involved in functional recovery (as described in the kidney) because clusterin can induce the expression of various factors, such as insulin-like growth factor-1 (IGF-1) . Interestingly, some growth factors, including IGF-1 and keratinocyte growth factor, suppress radiation-induced salivary gland dysfunction [49, 50]. Therefore, clusterin may recover the glands from irradiation-induced hypofunction through IGF-1. In addition, extracellular clusterin might bind to specific receptors, such as Gp330/megalin/LPR2, which is a target molecule of clusterin and inhibits death signals downstream of ROS . Thus, extracellular clusterin maintains cellular functions via receptor-mediated signaling pathways. Consistent with these findings, SP cells in the kidney recover renal dysfunction via renoprotective factors, such as hepatocyte growth factor, vascular endothelial growth factor, and bone morphogenetic protein 7 , suggesting that SP cells in lacrimal and salivary glands may have the potential to secrete cell-protective factors via a clusterin-related pathway. Furthermore, Sca-1 expression of SP cells isolated from clusterin-deficient mice decreased compared with that of wild-type mice, indicating that clusterin possibly affects SP cell function. To validate this hypothesis, the gene expression profile of clusterin-deficient SP cells should be compared with that of wild-type SP cells. This approach may also help us to understand the mechanism underlying the radioresistance of clusterin-expressing cancer cells , which should be tested in future studies.
Inhibition of ROS production is the mechanism underlying the ability of stem cells to maintain their self-renewal capability . The clusterin-expressing NIH3T3 cells we established promote colony formation when corneal epithelial cells are used as feeder cells . In addition, an ESC-conditioned medium containing clusterin can inhibit apoptosis, suggesting that clusterin may be important for the maintenance of the cellular microenvironment or stem cell niche . c-Kit- or c-Kit- and Sca-1-positive cells are multipotent stem cells in salivary glands, and transplantation of c-Kit-positive cells alleviates irradiation-induced salivary dysfunction [57–59]. Therefore, a combination of SP cell and c-Kit-positive cell transplantation is expected to be a promising method for the treatment of irradiation-induced hypofunction of glands because SP cells may promote engraftment of c-Kit-positive cell as a niche. Oxidative stress also has an important role in aging and the development of aging-related diseases [60, 61]. Thus, clusterin seems to be an important factor in the maintenance of the microenvironment in both glands.
Here, we report the possibility of the clinical application of SP cells and SP cell-related factors to restore the organ dysfunction caused by incurable and ROS-related diseases.
We thank K. Yamada and S. Yamamoto for their technical assistance and J. Nishino for helpful discussions. We also thank Dr. H. Miyoshi (Subteam for Manipulation of Cell Fate, BioResource Center, RIKEN Tsukuba Institute, Tsukuba, Japan) for providing us with an HIV vector packaging system and Dr. H. Niwa (Laboratory for Pluripotent Cell Studies, RIKEN Center for Developmental Biology) for his kind donation of the 129/SvJ mouse ESCs. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan and Grants-in-Aid for Scientific Research from the Ministry of Health, Labor, and Welfare.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.