CD16 CD56 cells are a potential culprit for hematuria in IgA nephropathy

Clin Exp Nephrol (2015) 19:216–224 DOI 10.1007/s10157-014-0968-z

ORIGINAL ARTICLE

CD16+CD56+ cells are a potential culprit for hematuria in IgA nephropathy Hirotsugu Iwatani • Yasuyuki Nagasawa • Ryohei Yamamoto • Kenichiro Iio Masayuki Mizui • Arata Horii • Tadashi Kitahara • Hidenori Inohara • Atsushi Kumanogoh • Enyu Imai • Hiromi Rakugi • Yoshitaka Isaka

•

Received: 14 August 2013 / Accepted: 20 March 2014 / Published online: 6 May 2014 Ó Japanese Society of Nephrology 2014

Abstract Background Hematuria is the first manifestation of urinary abnormality in immunoglobulin A nephropathy (IgAN). Hematuria has recently been reported as a risk factor for deterioration of renal function; however, its cause remains unknown. Methods We analyzed the surface marker of peripheral blood mononuclear cells before and immediately after tonsillectomy in IgAN patients and controls (chronic tonsillitis or tonsillar hypertrophy) by flow cytometry and investigated the association with hematuria. To prove our hypothesis that NK cells induce hematuria, we administered IL-12, activator of NK cells, to HIGA mice. In addition, we transferred cultured NK cells to nude rats and transferred the CD16?CD56? cells, including NK cells,

H. Iwatani R. Yamamoto K. Iio M. Mizui H. Rakugi Y. Isaka (&) Department of Geriatric Medicine and Nephrology, Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan e-mail: [email protected] Y. Nagasawa Department of Internal Medicine, Hyogo College of Medicine, Nishinomiya, Japan A. Horii T. Kitahara H. Inohara Department of Otorhinolaryngology-Head and Neck Surgery, Osaka University Graduate School of Medicine, Suita, Japan A. Kumanogoh Department of Respiratory Medicine, Allergy and Rheumatic Diseases, Osaka University Graduate School of Medicine, Suita, Japan E. Imai Nakayamadera Imai Clinic, Takarazuka, Japan

123

that are derived from the peripheral blood of IgAN patients immediately after tonsillectomy to nude rats to assess the hematuria level and renal histology of the recipients. We also performed cytotoxicity assays against glomerular endothelial cells by NK cells. Results We found that IgAN patients who showed rapid deterioration of hematuria after tonsillectomy also displayed a significant increase in CD16?CD56? cells in the peripheral blood immediately after tonsillectomy. Exogenous administration of IL-12 to HIGA mice induced hematuria. Adoptive transfer of either cells of an NK cell line, or of CD16?CD56? cells derived from IgAN patients, into nude rats induced hematuria in the recipients. In vitro analysis showed that NK cells exert cytotoxic activity toward human glomerular endothelial cells in a dosedependent manner. Conclusions CD16?CD56? cells seem to be responsible for hematuria in IgAN. Keywords CD16?CD56? cells Hematuria IgA nephropathy NK cells Tonsil Tonsillectomy

Introduction Hematuria is the most conspicuous and important feature of immunoglobulin A (IgA) nephropathy (IgAN), which is the most common form of primary glomerulonephritis in the world [1]. Hematuria with or without proteinuria is the first manifestation of urinary abnormality in IgAN. It is also well known that IgAN patients often manifest with macroscopic hematuria after upper respiratory tract infections such as tonsillitis. However, until recently, the significance of hematuria as a prognostic factor in IgAN was not proven by multivariate analysis [2, 3], and proteinuria

Clin Exp Nephrol (2015) 19:216–224

217

has been accepted as the sole and significant prognostic factor for future renal function in urinary findings. Due to the proteinuria-centric vision of chronic kidney disease over the last few decades, the significance of hematuria has rarely been focused on, even in studies of IgAN. There has been a recent accumulation of reports that have focused on the importance of hematuria as a risk factor. Although, Haas [4] reported that the presence of gross hematuria in IgAN correlates significantly with increased renal survival, macroscopic hematuria with a duration longer than 10 days has been reported to be a significant risk factor for incomplete recovery of renal function in IgAN [5]. Microscopic hematuria at clinical onset is associated with the progression of renal damage in IgAN patients [6]. Since hematuria is the first urinary abnormality in IgAN, the importance of hematuria in its etiology cannot be ignored. However, there have been few reports of attempts to elucidate the mechanism of hematuria in glomerulonephritis. Based on clinical observations such as temporal aggravation of hematuria after tonsillectomy, we proposed the hypothesis that some factors are released from the tonsils that aggravate hematuria in IgAN patients. Here, we found that a cell fraction that was increased in the peripheral blood immediately after tonsillectomy was associated with the rapid deterioration of hematuria. We identified this cell fraction as CD16?CD56? cells and performed in vivo and in vitro analyses to provide evidence that this cell fraction is the potential culprit behind hematuria.

Materials and methods Patients We investigated biopsy-confirmed IgAN patients who underwent tonsillectomy combined with intravenous corticosteroid administration in Osaka University Hospital. Table 1 Clinical characteristics of IgAN patients and controls IgAN

Chronic tonsillitis or tonsillar hypertrophy

Total

Early deterioration of hematuria (?)

Early deterioration of hematuria (-)

No hematuria

10

6

4

8

Age

34 ± 11

28 ± 8.9

43 ± 8.6

17 ± 11

Sex (F/M)

8/2

5/1

3/1

3/5

Intravenous corticosteroid was administered more than 7 days after tonsillectomy. After we consulted the ethics committee, we enrolled consecutive ten IgAN patients who gave informed, written consent to participate in the study. The subjects’ characteristics are shown in Table 1. As a control group (n = 8), we also investigated patients with chronic tonsillitis or tonsillar hypertrophy who underwent tonsillectomy. This study was approved by the ethics committee of Osaka University Hospital (approval number 0539). Analysis of urine Consecutive urine samples that were collected from pretonsillectomy until the morning after tonsillectomy were analyzed for hematuria using a dipstick test. Flow cytometric analysis of peripheral blood Peripheral blood was drawn from IgAN patients and controls before and immediately after the tonsillectomy. The expression of cell markers (CD20, CD3 and CD16 & CD56) on the blood cells was analyzed using a flow cytometer (SRL, Tokyo, Japan) and specific antibodies against CD3, CD16, CD56 (Beckman Coulter, Brea, CA, USA), and CD20 (Dako, Glostrup, Denmark). Exogenous administration of IL-12 to HIGA mice Murine recombinant IL-12 (PeproTec, Rocky Hill, NJ, USA) was intraperitoneally administered to HIGA mice (SLC, Hamamatsu, Japan). The IL-12 dose was as follows: 500 ng on each of days 1, 2, and 3; 750 ng on each of days 4, 5, and 6; and 1000 ng on days 7 and 8. The recipients’ urine was centrifuged at 2000g for 5 min, and sedimentation of the urine was analyzed using a dipstick test. Adoptive transfer of NK cells to nude rats Cells of the human NK cell line, NK 92MI, were cultured and labeled using the VybrantÒ CFDA SE Cell Tracer Kit (Life Technologies, Tokyo, Japan) following the manufacturer’s instructions. Labeled NK92MI cells (1–8 9 107) were transferred into nude rats (F344/N-rnu; CLEA, Tokyo, Japan) (n = 3) via the tail vein. Hematuria of the urine of the recipients was analyzed the next day using the dipstick test. Immunohistochemical staining of kidneys

n

Cr (mg/dl)

0.85 ± 0.16

0.82 ± 0.16

0.90 ± 0.16

0.60 ± 0.32

u-Pro/u-Cr (g/g Cr)

1.00 ± 0.58

1.00 ± 0.69

1.00 ± 0.50

–

The paraformaldehyde-fixed, sucrose-dehydrated, and OCT compound-embedded kidneys of the nude rats transplanted with NK92MI or those of the HIGA mice were thinly sliced using a cryostat, blocked with normal donkey serum and reacted with primary antibodies: anti-granzyme B (Gene

123

218

Tex, Irvine, CA, USA), anti-RECA1 (Cosmobio, Tokyo, Japan), or rabbit anti-neural cell adhesion molecule affinitypurified polyclonal antibody (CD56; Chemicon International, Temecula, CA, USA). After three washes with phosphate-buffered saline (PBS), the samples were further stained with the appropriate secondary antibodies that were conjugated with AlexaFluor 555 or 647 (Life Technologies). After three washes with PBS, the samples were further stained with 40 ,6-diamidino-2-phenylindole (DAPI), again washed three times with PBS, and then mounted with VECTERSHIELD (Vector Laboratories, Burlingame, CA, USA). The slides were observed using a confocal microscope FV1000-D (Olympus, Tokyo, Japan) connected to a Windows computer, using the appropriate filters. Transplantation of CD16?CD56? cells from the peripheral blood of IgAN patients and healthy volunteers into nude rats Peripheral blood was drawn from two different patients immediately after tonsillectomy and from two healthy volunteers, and mononuclear cells were isolated using LymphoprepTM (AXIS-SHIELD PoC AS, Oslo, Norway) following the manufacturer’s instructions. CD16?CD56? cells were then sorted from these mononuclear cells using MACS (Miltenyi Biotec, Tokyo, Japan). The collected CD16?CD56? cells (1.0–1.2 9 106 per person) were transplanted into nude rats via the tail vein, and hematuria of the urine of the recipients was analyzed using a dipstick test within 1 week. We confirmed that the purity of the CD16?CD56? cells was more than 95 % by flow cytometry. Comparison of the ability of CD16?CD56?, CD16-CD56?, and CD56- cells from IgAN patients to induce hematuria CD56-, CD16-CD56?, and CD16?CD56? cells were serially sorted from the purified mononuclear cells of four different IgAN patients, in this order, using the MACS system as described above. The collected cells were counted, and all three fractions were adjusted to get the maximum equal cell number. For each patient, each fraction contained the same number of cells and was then transplanted into a nude rat, and the urine was analyzed 1 week after the transplant as described above. In vitro assay of NK cell cytotoxicity Human glomerular endothelial cells (hGEC; Cell Systems, Kirkland, WA, USA) were used as target (T) cells, and cells of the human NK cell line, NK92MI, were used as effector (E) cells. Following plating, hGEC was washed twice with DMEM, and 15 lmol/l of calcein-AM

123


(Dojindo, Kumamoto, Japan) was incorporated into the cells at 37 °C under 5 % CO2 for 30 min using a modified procedure [7]. After two washes with DMEM, effector cells, a maximum release control or a spontaneous release control was added to the target cells, which were incubated at 37 °C under 5 % CO2 for 4 h. The fluorescence of each cell supernatant was analyzed using SH-9000Lab (CORONA ELECTRIC, Ibaraki, Japan), with excitation at 485 nm and emission at 530 nm. A solution of 10 % formalin/DMEM was used as the maximum release control and DMEM as the spontaneous release control. Cytotoxicity was calculated as follows: cytotoxicity (% of target cell lysis) = 100 9 (fluorescence of the sample - fluorescence of spontaneous release)/(fluorescence of maximum release - fluorescence of spontaneous release). The ratios of the number of effector cells to the number of target cells assayed were 50:1, 25:1, 12.5:1, 6.25:1, and 3.125:1.

Results Tonsillectomy-induced rapid deterioration of hematuria in some patients with IgAN We examined the short-term effect of tonsillectomy on hematuria by analysis of consecutive urine samples from IgAN patients, starting from before the tonsillectomy and continuing until the next morning. Some of the IgAN patients manifested temporal macroscopic hematuria after tonsillectomy. When we closely examined the sequential hematuria level using a dipstick test, 6 patients displayed deterioration of hematuria by the next morning (day ?1), including 3 cases progressing to the level of macroscopic hematuria (Fig. 1), while 4 patients did not. We designated the former 6 patients as the early hematuria-aggravating (EHA) group and the latter 4 patients as the non-EHA group. No patients in the control group showed hematuria. Circulating hematuria-inducing cells Since we observed rapid aggravation or occurrence of hematuria at tonsillectomy in the EHA group, we hypothesized that hematuria-inducing cells are activated and released to the peripheral blood from the tonsils immediately after tonsillectomy. We therefore analyzed the surface markers of peripheral blood mononuclear cells before and immediately after the tonsillectomy to identify potential circulating hematuria-inducing cells. Flow cytometric analysis demonstrated that there was no statistically significant difference in the post:pre-tonsillectomy ratio of the CD3? T cell fraction between EHA and non-EHA (Fig. 2). However, the post:pre-tonsillectomy ratio of the CD20? B


219

immediately after tonsillectomy in the EHA group. To test the hypothesis that NK cells rapidly aggravates hematuria, we analyzed hematuria induction in vivo in an animal model following exogenous activation of NK cells. IL-12, a potent activator of NK cells, was administered to HIGA mice, a mouse model of IgAN. On day 4 or later after the start of IL-12 administration, the centrifuged urine of treated mice showed hematuria (data not shown). PAS staining of the kidney tissue from these mice demonstrated the occurrence of crescents in the kidney as previously described [8, 9] (Fig. 3a). Immunohistochemical staining of the kidney showed a high number of CD56? cells in the glomeruli (6.7 ± 1.5/glomerulus, Fig. 3b). Adoptive transfer of cells of human NK cells to immunodeficient rats induces hematuria

Fig. 1 Perioperative hematuria level in the EHA group of IgAN patients. The x-axis indicates the time course starting from the day of tonsillectomy and the y-axis indicates hematuria level. MacroH stands for macroscopic hematuria. The plus and minus signs indicate before and after the tonsillectomy, respectively. IgAN patients in the EHA group manifested deterioration of hematuria compared to basal level (day-1 or day-0) until 1 day after the tonsillectomy (day ?1)

cell fraction was significantly lower, and that of the CD16?CD56? cell fraction was significantly higher, in EHA than in non-EHA or in control group (Fig. 2). Exogenous activation of NK cells induces mild hematuria in HIGA mice We focused on the CD16?CD56? cells, which are composed of NK cells and NK T cells, as responsible cells for hematuria aggravation, because they were increased

Since exogenous activation of NK cells was shown to result in the occurrence of hematuria in HIGA mice, we next investigated if direct transfer of human NK cells to nude rats would induce hematuria. Transplantation of NK cells, labeled with VybrantÒ CFDA SE Cell Tracer, showed that the transferred NK cells (green) were localized adjacent to the RECA1-positive endothelium within the glomerular area (0–1 NK cells per glomerulus). Of importance is that they retained cytotoxic activity, as assessed by granzyme B expression (Fig. 4a–d). Together with the infiltration of NK cells in the glomeruli, hematuria was confirmed in recipient rats 1 day after the transfer (Fig. 4e). Transplantation of CD16?CD56? cells from IgAN patients into nude rats The above results suggested that NK cells might play a causative role in aggravating hematuria in IgAN patients.

Fig. 2 Post:pre-tonsillectomy ratio of CD20?, CD3?, and CD16?CD56? cell fractions in the peripheral blood of IgAN patients or control. The EHA group of IgAN patients showed a significant increase in the post:pre-tonsillectomy ratio of the CD16?CD56? cell fraction in peripheral blood. (*p \ 0.05 Wilcoxon rank sum test)

123

220


Fig. 3 IL-12 and PBS were administered to HIGA mice. a, c The kidneys were stained with periodic acid–Schiff (PAS) stain. Crescent formation in the glomeruli can be seen in the IL-12 administered group (a), but not in controls (c). Bar indicates 30 lm. b, d The

kidneys were stained with an anti-CD56 antibody (white) and nuclei were co-stained with DAPI (blue). A high number of CD56? cells were observed in the glomeruli in the IL-12 administered group (b), but not in controls (d)

We therefore examined the effect on hematuria of transplantation of CD16?CD56? cells from two IgAN patients, into nude rats. Recipient rats exhibited hematuria 5 and 7 days after transplantation (Fig. 5). We next compared the hematuria-inducing ability of CD16?CD56? cells with that of CD16-CD56? or CD56cells using cells from four different IgAN patients. The extent of hematuria showed that CD16?CD56? cells had a much greater ability to induce hematuria than CD16-CD56? or CD56- cells (Table 2). We then examined the effect on hematuria of transplantation of CD16?CD56? cells from two healthy volunteers into nude rats. Recipient rats exhibited hematuria after transplantation (Table 3).

enough to cause hematuria, it is almost impossible to detect the sites of the breaks in vivo. To further test the hypothesis that NK cells cause hematuria, we next examined if NK cells can damage endothelial cells. Co-culture with NK cells induced cell lysis of GECs in a dose-dependent manner (Fig. 6).

In vitro analysis of the cytotoxicity of NK cells toward human GECs Hematuria in IgAN patients is thought to be caused by breaks in the glomerular capillary walls. Because just a few breaks among one million nephrons in one kidney are

123

Discussion In this study, we demonstrated that CD16?CD56? cells in the peripheral blood appeared to be potent hematuriainducing cells in IgAN patients. The key findings are as follows: (1) CD16?CD56? cell numbers were significantly increased immediately after tonsillectomy in the EHA group of IgAN patients; (2) exogenous administration of IL-12, a potent activator of NK cells that are the major component of the CD16?CD56? cell fraction, into HIGA mice induced hematuria; (3) adoptive transfer of NK cells into nude rats induced hematuria; (4) adoptive transfer of CD16?CD56? cells from IgAN patients into nude rats


Fig. 4 Dipstick analysis of the urine and immunohistochemical analysis of the kidneys of rats that were transplanted with NK92MI cell line. Rat kidneys were analyzed for the expression of CFDA-SE (green) (a), RECA1 (red) (b), granzyme B (white) (c), and nuclei (blue, using DAPI staining). A merged image of CFDA-SE, RECA1, granzyme B, and DAPI is shown in d. CFDA-SE-expressing NK cells

221

are present in the glomeruli of the recipient’s kidney and they retain granzyme B expression, which is important for NK cell cytotoxic activity. Dipstick analysis of the urine of recipient rats showed positive hematuria (e). Green signal indicates positive hematuria. There is no expression of CFDA-SE in the glomeruli of the control rats with no transplant (f)

123

222


Fig. 5 Induction of hematuria in recipient rats transplanted with CD16?CD56? cells from IgAN patients. Positive hematuria was observed in recipient rats that were transplanted with CD16?CD56? cells from IgAN patients. Green signal indicates positive hematuria

Table 2 Comparison of the hematuria-inducing ability of CD16?CD56? cells from IgAN patients with that of CD16-CD56? and CD56- cells IgAN patient

Transplanted number of cells

1

Post-transplant hematuria CD16?CD56? cells

CD16-CD56? cells

CD56cells

2.8 9 105

(3?)

(-)

(-)

2

1.3 9 105

(1?)

(-)

(-)

3

1.0 9 105

(1?)

(±)

(±)

4

6.2 9 104

(±)

(1?)

(-)

75 % (3/4)

25 % (1/4)

0 % (0/4)

Hematuria-inducing frequency

Hematuria levels of (1?), (2?), and (3?) were classified as positive. The percentage of the 4 rats that were positive for hematuria following transplantation of each cell phenotype was then calculated

Table 3 Transplantation of CD16?CD56? cells from healthy volunteers into nude rats Healthy volunteer

Transplanted number of CD16?CD56? cells

Post-transplant hematuria

1

1.0 9 105

(3?)

2

5

(1?)

1.0 9 10

Hematuria-inducing frequency

100 % (2/2)

induced heavy hematuria in the recipients compared with CD16-CD56? or CD56- cells; (5) in vitro analysis showed that NK cells were cytotoxic toward human GECs in a dose-dependent manner. We therefore concluded that CD16?CD56? cells induce hematuria in IgAN patients. We demonstrated that the CD16?CD56? cell number increased immediately after tonsillectomy in EHA group of

123

Fig. 6 Dose-dependent cytotoxicity of NK cells toward human glomerular endothelial cells. The dose-dependent cytotoxicity of NK cells toward human GECs was analyzed using an in vitro cytotoxicity assay. The cytotoxicity rate (% of target cell lysis) was calculated as described in ‘‘Materials and methods’’. Data are expressed as mean ± standard deviation. *p \ 0.05, **p \ 0.01, ***p \ 0.005 (Student’s t test)

IgAN patients compared within the non-EHA group of IgAN patients or in the controls. This observation is consistent with previous reports showing that CD16? and CD56? cells are markers of active IgAN [10] and that redistribution of NK cells is observed in IgAN patients [11]. Bacteremia is reported to occur briefly in the process of tonsillectomy [12]. This temporal bacteremia, which is probably due to mechanical stimulation of the tonsils during the operation, may stimulate the increase in CD16?CD56? cells in the circulation of IgAN patients. The observed decrease in CD20? B cell number in the EHA group following tonsillectomy may be a secondary event as a result of the increase in another cell fraction such as that of CD16?CD56? cells. A recent report has focused on the involvement of innate immunity in immunological disorders such as atopic dermatitis [13]. Considering that IgAN is also an immunological disorder which is involved in innate immunity [14] and that NK cells are a key player of innate immunity, the rapid increase in the CD16?CD56? cell number immediately after tonsillectomy in the EHA group of IgAN patients is understandable. Cox et al. [14] recently reported that NK cells or NK T cells increase in IgAN patients with macroscopic hematuria compared to those with microscopic hematuria, and their data also support our observations. We showed that exogenous administration of IL-12 into HIGA mice, which is a mouse model of IgAN [15], induced mild hematuria. IL-12 is a potent activator of NK cells and these cells are the major component of the CD16?CD56? cell fraction. HIGA mice exhibit a higher concentration of plasma IgA and glomerular deposition of IgA as they age. Although these mice do not manifest hematuria, the IL-12-treated HIGA mice did show mild


hematuria, suggesting the occurrence of GEC injury. In addition, administration of IL-12 induced infiltration of CD56? cells into glomeruli, suggesting that infiltrated NK cells induced GEC injury, thereby resulting in hematuria. We also showed that adoptive transfer of NK cells into nude rats induced hematuria. Using nude rats enabled us to perform this cell transfer without any immunosuppression which would attenuate the cytotoxic ability of the transferred cells. The transferred NK cells were detected adjacent to GECs and were positive for granzyme B, proving that infiltrated NK cells were capable of causing cytotoxic injury to endothelial cells. In addition, intravenously transplanted CD16?CD56? cells that were purified from IgAN patients induced hematuria in nude rats. In contrast, the ability of CD16-CD56? cells or CD56- cells from IgAN patients to induce hematuria was much less than that of CD16?CD56? cells. Highly cytotoxic activity of CD16?CD56? NK cells in general [16] also supports our experimental data. These observations support our hypothesis that CD16?CD56? cells induce hematuria in IgAN patients. NK cells could induce GEC injury in vitro. Moreover, the cytotoxic activity of NK cells toward human GECs was dose dependent. As to the mechanism of endothelial injury by NK cells, one possibility would be through p-selectin and its ligand. In general, NK cells express p-selectin glycoprotein ligand-1 (PSGL1), and NK cells that are activated by IL-12 administration adhere to selectin on endothelial cells through PSGL1 [17]. Moreover, p-selectin is reported to be upregulated in the GECs of IgAN patients [18], suggesting that activated NK cells can adhere to GECs through p-selectin and its ligand. Another highly possible mechanism would be CX3CR1-fractalkine axis. Membrane fractalkine expression can be reportedly induced on mature endothelial cells by proinflammatory cytokines. This fractalkine expression on endothelial cells is reported to mediate NK cell-mediated lysis, because NK cells express fractalkine receptor, CX3CR1 [19]. Moreover, deposition of fractalkine is reported in glomerular endothelium in IgAN patients [14]. These facts reinforce the hypothesis that NK cells mediate injury via CX3CR1fractalkine axis in glomerular endothelial cells, which leads to hematuria in IgAN. However, further research is necessary to determine the mechanism of NK cell attachment to endothelial cells. Our hypothesis that NK cells induce hematuria is further supported by a previous suggestion that there is persistent endothelial injury in IgAN patients accompanied by hematuria. Thus, the number of circulating endothelial cells, which is considered to be a marker of endothelial damage, has been shown to be elevated in IgAN patients compared to healthy controls [20].

223

Conclusions CD16?CD56? cells are a potential culprit behind hematuria in IgAN. Further research is necessary to determine the precise mechanism by which CD16?CD56? cells attach to glomerular endothelial cells so that they can exert their cytotoxic effect. Acknowledgments This study was supported by a Grant-in-Aid for Young Scientists B (18790562). Conflict of interest interest.

All authors have declared no competing

References 1. Pettersson E. IgA nephropathy: 30 years on. J Intern Med. 1997;242(5):349–53. 2. Haas M. IgA nephropathy histologically resembling focal-segmental glomerulosclerosis: a clinicopathologic study of 18 cases. Am J Kidney Dis. 1996;28(3):365–71. 3. Espinosa M, Ortega R, Gomez-Carrasco JM, et al. Mesangial C4d deposition: a new prognostic factor in IgA nephropathy. Nephrol Dial Transplant. 2009;24(3):886–91. 4. Haas M. Histologic subclassification of IgA nephropathy: a clinicopathologic study of 244 cases. Am J Kidney Dis. 1997;29(6):829–42. 5. Gutierrez E, Gonzalez E, Hernandez E, et al. Factors that determine an incomplete recovery of renal function in macrohematuria-induced acute renal failure of IgA nephropathy. Clin J Am Soc Nephrol. 2007;2(1):51–7. 6. Manno C, Strippoli GF, D’Altri C, et al. A novel simpler histological classification for renal survival in IgA nephropathy: a retrospective study. Am J Kidney Dis. 2007;49(6):763–75. 7. Ayalon O, Hughes EA, Cresswell P, et al. Induction of transporter associated with antigen processing by interferon gamma confers endothelial cell cytoprotection against natural killer-mediated lysis. Proc Natl Acad Sci USA. 1998;95(5):2435–40. 8. Nogaki F, Muso E, Kobayashi I, et al. Interleukin 12 induces crescentic glomerular lesions in a high IgA strain of ddY mice, independently of changes in IgA deposition. Nephrol Dial Transplant. 2000;15(8):1146–54. 9. Kobayashi I, Nogaki F, Kusano H, et al. Interleukin-12 alters the physicochemical characteristics of serum and glomerular IgA and modifies glycosylation in a ddY mouse strain having high IgA levels. Nephrol Dial Transplant. 2002;17(12):2108–16. 10. Hotta O, Seki S, Taguma Y, et al. CD16? and CD56? cells as markers of necrotizing and crescentic glomerulonephritis and active IgA glomerulonephropathy. Nephron. 1991;57(2):239–40. 11. Antonaci S, Polignano A, Ottolenghi A, et al. Redistribution of natural killer (NK) cell frequency and NK cytotoxic activity in primary IgA nephropathy. Cytobios. 1992;69(276):27–34. 12. Yildirim I, Okur E, Ciragil P, et al. Bacteraemia during tonsillectomy. J Laryngol Otol. 2003;117(8):619–23. 13. Maintz L, Novak N. Modifications of the innate immune system in atopic dermatitis. J Innate Immun. 2011;3(2):131–41. 14. Cox SN, Sallustio F, Serino G, et al. Activated innate immunity and the involvement of CX3CR1-fractalkine in promoting hematuria in patients with IgA nephropathy. Kidney Int. 2012;82(5):548–60.

123

224 15. Muso E, Yoshida H, Takeuchi E, et al. Enhanced production of glomerular extracellular matrix in a new mouse strain of high serum IgA ddY mice. Kidney Int. 1996;50(6):1946–57. 16. Beziat V, Duffy D, Quoc SN, et al. CD56brightCD16? NK cells: a functional intermediate stage of NK cell differentiation. J Immunol. 2011;186(12):6753–61. 17. Yago T, Tsukuda M, Fukushima H, et al. IL-12 promotes the adhesion of NK cells to endothelial selectins under flow conditions. J Immunol. 1998;161(3):1140–5.

123

Clin Exp Nephrol (2015) 19:216–224 18. Segawa C, Wada T, Takaeda M, et al. In situ expression and soluble form of P-selectin in human glomerulonephritis. Kidney Int. 1997;52(4):1054–63. 19. Yoneda O, Imai T, Goda S, et al. Fractalkine-mediated endothelial cell injury by NK cells. J Immunol. 2000;164(8):4055–62. 20. Erdbruegger U, Grossheim M, Hertel B, et al. Diagnostic role of endothelial microparticles in vasculitis. Rheumatology (Oxford). 2008;47(12):1820–5.