two novel members of the globin superfamily were recently discovered, neuroglobin (Ngb) and cytoglobin (Cygb) (1, 2, 14). In vertebrates, two conventional globins, hemoglobin (Hgb) and myoglobin (Mgb), have been intensively studied. Hgb is located predominantly in erythrocytes and functions mainly in the transport of oxygen from the respiratory surfaces to the inner organs. Mgb, which is localized mainly in striated and cardiac muscle, acts as temporal oxygen store and facilitates intracellular oxygen diffusion. These globins reversibly bind gaseous diatomic ligands to a pentacoordinate heme iron atom. In contrast to these conventional globins, Ngb and Cygb display a functionally relevant hexacoordinate heme iron atom. While Ngb is predominantly expressed in nerve cells, Cygb is expressed in the fibroblast cell lineages, which are distributed in various organs (19, 29, 30).

The functional importance of Cygb is emphasized by its highest degree of sequence conservation among vertebrate globins, with mouse and human Cygb differing in only 4.7% of amino acids (2).

Many potential biological functions of Cygb have been proposed. Kawada and colleagues (14) suggested that Cygb acts as a peroxidase involved in the detoxification of reactive oxygen species (ROS). Schmidt and colleagues (29) proposed a biological function of Cygb in collagen synthesis, in which Cygb might supply oxygen to collagen prolyl-4-hydroxylase for the hydroxylation of proline residues in the procollagen molecule. Fordel and colleagues (8) showed that Cygb plays a protective role against oxidative stress in cancer cells. Based on their epigenetic studies of cancer cell lines in combination with knockdown and enforced expression of Cygb, Shivapurkar et al. (31) suggested that Cygb serves as a tumor suppressor gene. These studies were all conducted in vitro, however, and Cygb's biological function in vivo remains controversial. Localization of Cygb in the fibroblast cell lineage stimulated us to investigate a possible role of Cygb in the pathogenesis of fibrotic kidney disease.

MATERIALS AND METHODS

Animal Experiments

All experiments were conducted in accordance with the Guide for Animal Experimentation, Faculty of Medicine, University of Tokyo, with approval by the local ethical committee. To investigate the expression of Cygb in a fibrosis model of the kidney, 6-wk-old male Wistar rats (Nippon Bio-Supp. Center, Saitama, Japan) weighing 160–200 g were housed in cages in a temperature- and light-controlled environment in an accredited animal care facility, with free access to normal chow and tap water. After 1 wk of acclimatization, rats underwent subtotal nephrectomy [remnant kidney (RK)] as follows: via a midline incision under 50 mg/kg ketamine anesthesia right kidney nephrectomy was performed, followed by ligation of the posterior and one or two of the anterior branches of the left renal artery. Animals were euthanized at weeks 1, 4, and 9, and the kidneys were harvested for further analysis (n = 6 for each time point).

We administered cobalt chloride to 6-wk-old male Wistar rats (Nippon Bio-Supp. Center). Cobalt chloride (Sigma) was dissolved in phosphate-buffered saline (PBS) and administered intraperitoneally at 60 mg/kg. Control animals received PBS. After 24 h, the kidneys of the rats were harvested and mRNA of the tissue was extracted to make cDNA for measurement of the expression of Cygb.

Establishment of a Cygb-overexpressing transgenic (Cygb-Tg) rat was performed by microinjection of the ubiquitous CAGGS vector expressing Cygb into fertilized eggs of Crlj:Wistar rats, as detailed elsewhere (Nishi et al., unpublished observations). In brief, the coding region (159 bp to 731 bp) of the rat Cygb gene was obtained by PCR and cloned into the pCAGGS vector (23). Validity of the segment was verified by sequencing analysis. We engineered three lines of Cygb-Tg rats.

The Cygb-Tg rats used in this study were offspring of intersibling matings over at least three generations. The results of the studies were obtained by utilizing line A, but were confirmed in line B, an independent transgenic line. The study using line A was repeated twice, and essentially the same results were observed.

Heterozygous Cygb-Tg rats of the same age (n = 6) were compared with wild-type littermates (n = 4). At the end of week 9, the rats were euthanized, blood was taken by cardiac puncture, and the kidneys were removed for pathological and immunohistochemical analyses. Blood urea nitrogen (BUN) and serum creatinine levels were measured at weeks 0, 4, and 9 with commercial kits (Wako, Osaka, Japan). Urine samples were collected in metabolic cages at designated time points, and urinary protein excretion was measured with a Bio-Rad assay reagent. Blood pressure was measured by an occlusive tail-cuff plethysmograph attached to a pneumatic pulse transducer. Blood pressure was recorded as the mean value of three separate measurements obtained at each session.

Immunohistochemical Analyses

Kidney samples were embedded in OCT compound (Sakura Fine Technical, Tokyo, Japan), frozen in liquid nitrogen, and stored at −80°C until use. The sections (4 μm) were fixed with ether-ethanol (1:1) for 10 min. After endogenous biotin and peroxidase activity were quenched, samples were stained with the first antibody of anti-collagen I (1:500; Chemicon, Temecula, CA) in 1% bovine serum albumin (BSA) in PBS at 4°C overnight after use of an avidin-biotin blocking kit (Vector Laboratories, Burlingame, CA). After incubation with the first antibody, the samples were washed and reacted with a biotinylated anti-goat IgG (Dako, Carpenteria, CA) and horseradish peroxidase (HRP)-avidin (1:1,000; Vector Laboratories). Sites of immunoreactive antigens were detected with hydrogen peroxide and diaminobenzidine (DAB; Wako).

Methyl-Carnoy's-fixed paraffin-embedded sections (3 μm) were stained by the indirect immunoperoxidase method. We used rabbit anti-collagen I antibody (1:400; Abcam), goat anti-collagen IV antibody (1:500; Southern Biotech, Birmingham, AL), mouse anti-α-smooth muscle actin monoclonal antibody ASM-1 (Progen Biotechnik, Heidelberg, Germany), mouse anti-macrophage monoclonal antibody ED-1 (Chemicon), and rabbit anti-nitrotyrosine antibody (Sigma-Aldrich, St. Louis, MO) as the first antibodies.

Double Immunohistochemical Staining of Collagen IV and Anti-Cygb

We performed double immunohistochemical staining of collagen IV and Cygb. Methyl-Carnoy's-fixed paraffin-embedded tissues were reacted with goat anti-collagen IV antibody utilizing the DAB reaction without nickel chloride by the methods described above. After the procedure, we performed immunostaining with rabbit anti-Cygb polyclonal antibodies, using DAB containing nickel chloride. We raised rabbit polyclonal antibodies against the synthetic polypeptides targeting amino acid position 66–80, MEDPLEMERSPQLRK-Cys. Essentially the same observations were confirmed with the polyclonal antibody targeting a different part of Cygb (generous gift of Drs. Norifumi Kawada and Katsutoshi Yoshizato) (14). Controls included depletion of the primary antibody.

Morphometric Analysis and Semiquantitative Evaluation of Masson Trichrome Staining and Immunohistochemical Studies

Morphometric quantification of positive staining was assessed with Scion Image (Scion, Frederick, MD) and NIH Image J (version 1.42) software. At least 20 randomly selected cortical tubulointerstitial areas from each sample were evaluated.

Semiquantitative evaluation of fibrosis or positive staining in the kidneys was performed as follows (24). The fibrotic area or positive-staining area of the kidneys was measured as follows: grade 0, no fibrosis; grade 1, involvement of <10%; grade 2, 10–25%; grade 3, 25–50%; grade 4, 50–75%; and grade 5, 75–100%. More than 20 consecutive fields were examined under ×200 magnification and averaged per slide.

All evaluations were performed in a blinded manner.

RNA Isolation and Quantitative Real-time PCR

Total RNA was isolated by ISOGEN (Nippon Gene, Tokyo, Japan) according to the manufacturer's protocol, and mRNAs were evaluated by quantitative real-time PCR. We employed β-actin as an internal control. PCR was conducted in triplicate for each sample. The sequences of primers are shown in Table 1. cDNA was synthesized with random primer in a 25-μl reaction from 1 μg of total RNA and the ImProm-II reverse transcription system (Promega, Madison, WI) according to the manufacturer's protocol. One microliter of cDNA was added to SYBR Green PCR Master Mix (Bio-Rad) and subjected to PCR amplification (cycled 40 times at 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s) in the iCycler system (Bio-Rad).

Immunoblotting Studies

Immunoblotting analysis was performed to evaluate rat Cygb proteins in rat kidney tissues and cultured cells. Kidney cortex tissue was homogenized in sucrose buffer pH 7.4, followed by centrifugation at 700 g for 10 min at 4°C. Cultured fibroblasts were pelleted and washed in PBS and suspended in lysis buffer containing 1% Triton X, 10% glycerol, 20 mmol/l HEPES, 100 mmol/l NaCl, and protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN), and pellets were cleared by centrifugation at 14,000 rpm for 5 min at 4°C. These protein samples were separated by electrophoresis on a 12% SDS-polyacrylamide gel, followed by electrotransfer to polyvinylidene difluoride (PVDF) membranes (GE Healthcare Bio-Sciences, Little Chalfont, UK). Transfer membranes were blocked with 5% nonfat milk in Tris-buffered saline (TBS) with 0.01% Tween 20 for 60 min at room temperature. The membrane was incubated with the anti-Cygb antibody overnight at 4°C. HRP-conjugated anti-rabbit IgG (Promega) was then used as the secondary antibody. Immunoreactive protein was visualized by the chemiluminescence protocol (ECL, GE Healthcare Bio-Sciences). Anti-actin rabbit polyclonal antibody (1:1,000; Sigma-Aldrich) was used for calibration.

Cell Culture

A rat kidney fibroblast cell line, NRK49F, was obtained from RIKEN BioResource Center (Tsukuba, Japan). Cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Nissui Seiyaku, Tokyo, Japan) DMEM/F-12 at pH 7.4 supplemented with 5% fetal bovine serum (FBS; SAFH Biosciences, Lenexa, KS), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.01 mmol/l nonessential amino acid. Cells were cultured in humidified 95% air with 5% carbon dioxide at 37°C.

Primary cultures of kidney fibroblasts of Cygb-Tg rats and wild-type rats were performed as previously described (32). Primary cultures were maintained in DMEM (Invitrogen, Carlsbad, CA) with 10% FBS. To characterize primary cultured fibroblasts, immunohistochemical analysis was performed as described below with anti-vimentin monoclonal antibody V9 (Dako) or anti-α-smooth muscle actin monoclonal antibody ASM-1 (Progen Biotechnik). We also performed PCR to analyze expression of periostin, using primers described in Table 1. One microliter of cDNA was subjected to PCR amplification (cycled 30 times at 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s). Primary cultured rat vascular smooth muscle cells (12) and primary cultured rat mesangial cells (22) served as controls.

Immunocytochemical Analysis

Primary cultures from Cygb-Tg rats, NRK49F, and HEK293T were seeded on four-well Lab Tek chamber slides (Nalge Nunc International, Rochester, NY). The cells were fixed with acetone-methanol (1:1) at −20°C for 10 min and then stained with anti-Cygb polyclonal antibody at 4°C overnight. They were reacted with biotinylated anti-rabbit IgG (1:400) and HRP-avidin (described above). The chamber slides were developed with DAB.

Enzyme-Linked Immunosorbent Assay

Culture cells were incubated in culture medium without FBS for 24 h before supernatant fluid was collected for measurement. Aminopropionitrile was added to the medium because collagen I attached to the bottom of the culture dish. A 96-well microplate (MAXISORP; Nalge Nunc International) was incubated with 100 μl/well of samples at 37°C for 1 h. The plate was washed in washing buffer five times and incubated with 200 μl/well of PBS containing 0.5% BSA (Sigma-Aldrich) at room temperature for 1 h. The wells were reacted overnight at 4°C with 100 μl/well of anti-collagen I antibody (Chemicon, Billerica, MA). Development was performed with 4-nitrophenyl phosphate disodium salt hexahydrate (Sigma-Aldrich; 100 μl/well) for 10 min and stopped by the addition of 3 N NaOH (100 μl/well). Absorbance at 452 nm was measured. The level of collagen I was normalized with purified collagen type I as standard (BD Biosciences, San Jose, CA).

Plasmid Construction and Transient Transfection of Cygb

We transfected NRK49F cells with a pCAGSS/Cygb(+) construct or a control vector, utilizing Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol.

Establishment of Stable Cygb-Transfected Cell Lines

Stable cell lines with inducibly expressed Cygb were established with the T-Rex system (Invitrogen), which expresses the Tet repressor. We made a construct of pcDNA4/TO/myc-His expressing Cygb and used Zeocin to select double-transfected clones. Established cell lines were maintained with medium containing blasticidin at a concentration of 5 μg/ml. Forty-eight hours after transfection of HEK293T cells with the pcDNA4/TO/myc-His construct, the cells were split into fresh medium containing Zeocin at the appropriate concentration of 250 μg/ml. We replenished the medium every 3–4 days until Zeocin-resistant colonies were detected. About 20 foci were then selected and expanded to test for tetracycline-inducible gene expression. Cygb was induced with the appropriate concentration of tetracycline at 1 μg/ml. To confirm tetracycline inducibility, we measured the expression level of Cygb, using both RT-PCR and Western blotting. Seven stable independent cell lines were detected, which showed 10-fold induction of Cygb expression compared with the control.

Cellular ROS Assay

NRK49F cells were treated with 5(6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate acetyl ester (CM-H₂DCF-DA) according to the manufacturer's protocol to detect cellular ROS after transient exposure to 100 μM of hydrogen peroxide by fluorescence-activated cell sorting (FACS), as previously described (22).

Mutagenesis with Histidine Residue

Site-directed mutagenesis was performed as follows: substitution of histidine for His81 of the rat Cygb protein with a tyrosine residue was conducted by a PCR method (30 cycles of PCR consisting of incubation for 10 s at 98°C, 5 s at 55°C, and 6 min at 72°C) with a high-fidelity Taq polymerase, PrimeSTAR HS DNA Polymerase (TaKaRa Bio, Tokyo, Japan), with the rat Cygb gene carried in the pCAGGS plasmid. The PCR fragments were confirmed by sequence analysis. The primers are shown in Table 1.

Small Interfering RNA Transfection

NRK49F cells were transfected with Cygb-specific small interfering RNA (siRNA) and control siRNA. The rat Cygb siRNA targeted nucleotides 522 to 543 of the Cygb mRNA sequence (NM_130744). Its sequence is sense 5′-GGU GGA ACC UAU GUA CUU UAA-3′ and antisense 5′-AAA GUA CAU AGG UUC CAC CUU-3′. A nontargeting RNA duplex siRNA containing 21 nucleotides was employed as a control (Invitrogen). Cells were passaged into six-well plates and grown to 30% confluence before transfection. A total of 1.5 μl of Lipofectamine 2000 reagent (Invitrogen) and 50 pmol of siRNA duplexes per well were added in 2.5 ml of Opti-MEM (Invitrogen). At 8 h after transfection, culture media were replaced with serum-containing media, and cells were then harvested to measure the amount of collagen type I.

Statistical Analysis

All data are reported as means ± SE. Data for two groups were analyzed with the t-test, and those for more than two groups were compared with ANOVA. Nonparametric data were analyzed with the Mann-Whitney test when appropriate. Differences with P value <0.05 were considered significant.

RESULTS

Expression of Cygb in Interstitium

To study the expression of Cygb in the kidney of rats, we made rabbit polyclonal antibodies against rat Cygb. Immunohistochemical staining with these antibodies showed expression of Cygb in interstitial fibroblasts of the normal kidney (Fig. 1A, top left). Staining signals were not detected when normal rabbit sera were used as the control (data not shown).

Increased Expression of Cygb in Remnant Kidney

We next investigated the expression of Cygb in fibrotic kidneys by using the remnant kidney (RK) model, a model of renal fibrosis. RK rats developed severe chronic tubular damage, including tubular dilation, fibrosis, and cast formation. Interestingly, immunohistochemical analysis showed the upregulation of Cygb in RK rats (Fig. 1A, top right). Expression of Cygb was localized in the interstitial cells in the disease model, as was observed in the normal kidney. We then examined the temporal profile of Cygb expression for 9 wk after the onset of RK. As shown in Fig. 1A, Cygb expression increased with time. Morphometric analysis revealed that both the area and number of Cygb-positive cells increased in a time-dependent manner (Fig. 1, B and C), in parallel with progression of fibrosis estimated by Masson trichrome staining (Fig. 1A, bottom, and D). Our Western blotting analysis of control animals and rats with RK showed a single band corresponding to Cygb, excluding cross-reaction with other members of the globin family such as Hgb, Mgb, and Ngb, and confirming specificity of the antibody (Fig. 1E).

To confirm the upregulation of Cygb in RK rats seen in our immunohistochemical analysis, we performed quantitative analysis of Cygb utilizing real-time quantitative RT-PCR (Fig. 2A) and Western blot (Fig. 2, B and C) analyses. Expression levels of Cygb in RK rats were significantly increased as the disease progressed at both the mRNA and protein levels.

Association of Upregulation of Cygb with Fibrosis of Kidney

The association of Cygb upregulation with fibrosis was confirmed by using immunohistochemistry for the detection of collagens. Accumulation of collagen I occurred in a time-dependent manner (Fig. 3A). Our morphometric analysis using collagen I staining of paraffin-embedded tissues showed that upregulation of Cygb was associated with an increase in collagen I in the tubulointerstitium of RK rats (Fig. 3B). Accumulation of collagen I was confirmed by using a different antibody that worked only in frozen sections (Fig. 3C).

The association of the upregulation of Cygb and the progression of fibrosis was also demonstrated in studies of collagen IV accumulation in the tubulointerstitium of the RK (Fig. 4, A and B). Because the anti-collagen I antibody that works in paraffin-embedded sections was a rabbit polyclonal antibody, we used collagen IV deposition as a marker of fibrosis in our double-staining studies. We observed the expression of Cygb stained with black in interstitial cells adjacent to collagen IV deposition areas stained with brown (Fig. 4C), suggesting that Cygb expression was correlated with collagen synthesis in interstitial cells in RK rats.

We speculated that the upregulation of Cygb expression in RK rats was due to local hypoxia in the fibrotic kidney. We confirmed hypoxia-inducible factor (HIF)-dependent upregulation of Cygb in the kidney in vivo by cobalt chloride treatment (Supplemental Fig. S1). ¹

Induction of RK in Cygb Transgenic Rats

To investigate the biological role of Cygb in vivo, we established a new transgenic (Tg) rat that overexpressed rat Cygb under the control of a ubiquitous CAGGS promoter and induced RK in these animals. Cygb was expressed in various organs in these rats, including brain, heart, liver and kidney. Histological and laboratory analysis of blood and urine samples did not show any differences between the wild-type and Cygb-Tg animals under normal conditions, and both developed kidney injury when RK was induced. In wild-type rats with RK, the number of Cygb-positive interstitial cells markedly increased in parallel with disease progression (Fig. 5A, top). In RK-induced Cygb-Tg rats, in contrast, expression of Cygb remained high throughout the experimental course (Fig. 5A, bottom), and even at the end of the course, when expression peaked in wild-type animals, we confirmed that Cygb expression increased more in Tg rats with RK than in wild-type rats at both the mRNA and protein levels (2.15 ± 0.67-fold increase vs. wild-type RK rats at 9 wk, P < 0.05) (Fig. 5, B and C).

Preservation of Renal Function in Cygb Transgenic Rats with RK

Importantly, deterioration of renal function over 9 wk as estimated by serum creatinine levels was ameliorated in Cygb-Tg rats with RK compared with wild-type rats with RK (Fig. 6A). The amount of proteinuria was also less in Cygb-Tg rats with RK than in wild-type animals, although the difference did not reach statistical significance (Fig. 6B). While systolic blood pressure tended to be lower in Cygb-Tg rats with RK than in wild-type animals, the difference did not reach statistical significance throughout the experimental course (Fig. 6C). Body weight did not differ between the two groups at 4 or 9 wk (Fig. 6D).

Amelioration of Fibrosis in Cygb Transgenic Rats with RK

The better physiological parameters in Cygb-Tg rats were associated with the improvement of histological injury in the kidney. Deposition of collagen I and IV was less severe in Cygb-Tg rats than in wild-type littermates, as estimated by immunostaining of collagen I and IV (Fig. 7, A and B). As described above, evaluation of fibrosis of the RK by morphometric analysis of collagen I and IV accumulation (Fig. 7, C and D) showed that transgenic overexpression of Cygb ameliorated fibrosis of the kidney. Protection against fibrosis in Cygb-Tg rats was confirmed by Masson trichrome stain (Fig. 7, E and F).

We also used histological injury markers to evaluate kidney injury. Use of anti-α-smooth muscle actin antibody to identify activated myofibroblasts in the interstitial area showed that areas positive for the staining were reduced in Cygb-Tg rats (Fig. 8, A and B). Furthermore, infiltration of macrophages was decreased in Cygb-Tg rats (Fig. 8, C and D).

Suppression of Oxidative Stress in RK of Cygb Transgenic Rats

Given that Cygb is a member of the globin superfamily, we speculated that it may play a role as an antioxidant. To clarify an antioxidative role of Cygb, we transfected the kidney fibroblast cell line NRK49F to overexpress Cygb, as described below. When Cygb-overexpressing NRK49F cells were treated with hydrogen peroxide, they displayed increased ROS scavenging activity compared with cells expressing a mock vector (Fig. 9).

We then speculated that Cygb may protect the kidney against fibrosis via a reduction in oxidative stress. To investigate the mechanism of amelioration of fibrosis by Cygb in vivo, we used nitrotyrosine as a marker of oxidative stress. Immunohistochemical analysis revealed a decrease in nitrotyrosine accumulation in tubules in RK in Cygb-Tg rats (Fig. 10, A and B). We also measured urinary 8-hydroxy-2′-deoxyguanosine (8-OHdG) as another marker of oxidative stress in wild-type and Cygb-Tg rats with RK of 9-wk duration (Fig. 10C). Urinary 8-OHdG in Tg rats was significantly reduced compared with wild-type animals.

Decrease in Collagen Production by Overexpression of Cygb

Primary cultured kidney fibroblasts isolated from Cygb-Tg rats.

To further clarify the mechanism by which renal fibrosis is ameliorated in Cygb-Tg rats, we established primary cultured fibroblasts from the kidneys of wild-type and Cygb-Tg rats. These primary cultured fibroblasts were stellate and spindle shaped (Fig. 11A) and showed positive staining for vimentin, a marker for fibroblasts, but no staining for α-smooth muscle actin (data not shown). These cells also expressed periostin, another marker of fibroblasts (33) (Fig. 11B). Cygb-Tg rat-derived fibroblasts showed overexpression of Cygb at both the mRNA and protein levels compared with cells derived from wild-type animals. We established an ELISA to measure collagen I production by these primary cultured cells. Synthesis of collagen I was reduced in fibroblasts from Cygb-Tg rats at both mRNA (73.8%) and protein (43.5%) levels compared with cells from wild-type rats (Fig. 11, C and D).

Inducible transfectant of Cygb gene in kidney cells.

To investigate the functional effect of Cygb on collagen synthesis, we also established stable transfectants in HEK293T cells, in which Cygb can be overexpressed in an inducible manner (Fig. 11, E and F). The amount of collagen I production measured by ELISA was reduced in these cells when Cygb expression was induced (Fig. 11G).

Transient Cygb transfectant of rat kidney fibroblast cells.

We also employed a rat kidney fibroblast cell line, NRK49F. We transiently overexpressed Cygb in this line with a transfection efficiency of 53% as estimated by immunocytochemical analysis (Fig. 12A). Collagen I synthesis in the transfectants was significantly reduced compared with that in mock-transfected cells with control plasmid vector (Fig. 12B).

Decrease in Collagen Synthesis by Overexpression of Cygb Recovered in Mutant Cygb

The heme component of Cygb might be the main contributor to radical scavenging. To test this, we engineered mutant cells with impaired ability to bind ligands at the heme of Cygb. Design of these mutants was assisted by reference to the pathogenesis of human methemoglobinemia, in which a genetic mutation of Hgb results in the inability to bind oxygen molecules and severe systemic cyanosis. The distal histidine is a classic site of mutation in hereditary methemoglobinemia, and on the basis of a shared structural homology between Hgb and Cygb we substituted the distal histidine (His81 in the rat Cygb protein) with a tyrosine residue (Fig. 12C). In contrast to cells overexpressing intact Cygb, the amount of collagen I synthesis by kidney fibroblasts expressing mutated Cygb did not differ from that by control cells (Fig. 12B).

Knockdown of Cygb by siRNA and Collagen Type I Production

To investigate the antifibrotic mechanism of Cygb, we performed knockdown of Cygb in NRK49F using RNA interference (RNAi) technology, as shown in Fig. 13A. The knockdown efficiency of Cygb was >90% on average. In contrast to the decrease in collagen production by overexpression of Cygb in three different in vitro systems, we failed to observe an increase in collagen production by knockdown of Cygb at both mRNA and protein levels (Fig. 13B).

DISCUSSION

Chronic hypoxia in the kidney is the final common pathway to end-stage kidney disease (3, 4, 13), and expression of Cygb is known to be regulated in an oxygen tension-dependent manner via HIF (5, 10). RKs suffer from hypoxia owing to both hemodynamic and structural changes (18, 35), and hypoxia of the kidney with renal artery ligation has been demonstrated with immunohistochemical analysis of HIF, accumulation of pimonidazole, and hypoxia-sensing transgenic rats (18, 25, 35, 36). Upregulation of HIF-1α has also been demonstrated (27, 28). Moreover, the pathogenic role of hypoxia in the RK model was further emphasized by the beneficial effects of HIF activation (33, 36). Here we demonstrate that Cygb is upregulated in the interstitial region of fibrotic kidneys in RK model rats. Fibrotic tissues experience hypoxia as a result of the distortion and loss of capillaries and reduction in oxygen diffusion efficiency. It is therefore likely that the upregulation of Cygb expression in RK is due to this induced decrease in oxygen tension. HIF-dependent upregulation of Cygb in the kidney was confirmed by cobalt chloride treatment of animals to stimulate HIF in vivo.

The next question was whether the increase in Cygb in fibrotic areas was an adaptive mechanism against fibrosis or a cause of the accumulation of extracellular matrix. To answer this, we utilized novel Cygb-overexpressing Tg rats. Importantly, we observed that Cygb overexpression ameliorated RK-induced kidney fibrotic damage, and that this effect was associated with better physiological parameters.

Furthermore, we showed that the synthesis of collagen I was inhibited when Cygb was upregulated in NRK49F, a cell line of kidney fibroblasts, and in primary cultured kidney fibroblasts. Collagen I synthesis was also inhibited by inducible Cygb overexpression in HEK293T, an embryonic kidney cell line. Our results from three different assays, together with staining data for collagen I and IV and Masson trichrome staining of Cygb-Tg rats, confirmed that overexpression of Cygb decreased the synthesis of collagen. These results suggest that upregulation of Cygb in fibrotic tissues may serve as a defensive mechanism against hypoxia in the kidney.

Mammalian Cygb has the highest degree of sequence conservation among the vertebrate globins (2). Mammalian Cygb is larger than most globins, covering 190 amino acids instead of the typical 140–150 amino acids. Cygb folding consists of eight α-helices, and important globin-specific residues involved in oxygen binding are conserved in the Cygb protein. While other members of the globin family Hgb and Mgb are known to form a tetramer and a monomer, respectively, the quaternary structures of Cygb and Ngb are not known yet.

While the biological role of Cygb has remained controversial, several potential functions have been proposed. Schmidt and colleagues (29) speculated that Cygb might supply oxygen to collagen prolyl-4-hydroxylase for the hydroxylation of proline residues in the procollagen molecule and proposed a biological function of Cygb in collagen synthesis. In contrast, accumulating evidence suggests that one potential function of Cygb is an antioxidative stress effect (17). Kawada and colleagues (14) suggested that Cygb acts as a peroxidase in the detoxification of ROS. Antioxidative stress effects of Cygb were also shown in human neuroblastoma cells (6, 8, 16). Xu et al. (39) showed that overexpression of Cygb protected against both toxic and cholestatic models of liver injury by scavenging toxic species, including hydrogen peroxide. In the present study, we showed that oxidative stress was reduced in RK of Cygb-Tg rats. Our in vitro study utilizing NRK49F overexpressing Cygb also supported the antioxidative stress effects of Cygb. Taking these findings together, we speculate that Cygb protects the kidney against the fibrotic changes induced by RK via a reduction in oxidative stress.

Oxidative stress increases collagen production (25) and contributes to the progression of kidney fibrosis (15). To clarify the mechanism of the antifibrotic effects of Cygb, we examined collagen I synthesis in kidney fibroblasts overexpressing mutated Cygb at the heme region. Heme is a porphyrin molecule with diverse biological functions, including scavenging oxidative stress, and is common to all members of the globin superfamily of proteins. In this study, His81 mutant-oriented disruption of heme attenuated the inhibitory effect of collagen I synthesis by Cygb, indicating that heme is responsible for the antifibrotic role of Cygb via a reduction in oxidative stress.

Overexpression of Cygb resulted in milder hypertension in RK rats, albeit that the difference between Cygb-Tg and control rats did not reach statistical significance. Lower blood pressure levels in Cygb-Tg rats may have been no more than a consequence of the improvement in kidney injury by Cygb overexpression. However, high blood pressure is a major exacerbator of kidney injury, and oxidative stress plays a pathogenic role in hypertension (37, 38). It is likely that the decrease in oxidative stress improved kidney injury via the amelioration of hypertension, in addition to the direct protection of tissues in Cygb-Tg rats.

Although the notion of oxidative stress under hypoxic conditions sounds paradoxical, hypoxic cells do in fact suffer from energy depletion and oxidative stress (9, 20). The complexity of the relationship between hypoxia and oxidative stress is highlighted by the fact that hypoxia aggravates oxidative stress and vice versa, forming a vicious cycle of hypoxia and oxidative stress. It is therefore likely that the upregulation of antioxidative protein in fibrotic organs, which suffer from hypoxia due to the loss of capillaries and inefficiency in O₂ diffusion, can serve as an effective defense.

Although this is the first study to show a protective effect of Cygb against kidney fibrosis, several limitations of our study warrant mention. We performed the knockdown of Cygb by using siRNA and failed to observe an increase in collagen production. This may be due to redundant effects of other antioxidative enzymes, which may have compensated for a decrease in Cygb in the loss-of-function studies.

We did not precisely establish the mechanism by which Cygb alters the fibrosis in the interstitium of the kidney. The measurement of collagen I production by NRK49F did not show a significant increase by hydrogen peroxide treatment (data not shown), and we speculate that the cell toxicity of hydrogen peroxide may have cancelled out the stimulatory effect of superoxide on collagen production. We have to keep in mind that it remains to be clarified how Cygb might mediate the reduction of collagen I production directly or indirectly. The molecular mechanism by which Cygb reduces fibrosis is a subject for future study. Another important source of collagen in the kidney is pericytes (11). However, because of the lack of an appropriate culture system for renal pericytes, we could not perform in vitro studies to study collagen expression in pericytes.

In conclusion, we demonstrated for the first time that Cygb plays a protective role against fibrotic changes in the kidney both in vitro and in vivo via the amelioration of oxidative stress. Cygb is a fascinating target for therapeutic approaches in a variety of fibrotic diseases, including chronic kidney disease.

GRANTS

I. Mimura is a Research Fellow of the Japan Society for the Promotion of Science (JSPS). This work was supported by Grant-in-Aid for JSPS Fellows 09J07041 (to H. Nishi) and Grant-in-Aids for Scientific Research 19590939 and 22590880 (to R. Inagi) and 19390228 and 2139036 (to M. Nangaku) from the Japan Society for the Promotion of Science.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

FOOTNOTES

• ¹Supplemental Material for this article is available online at the Journal website.