http://www.kidney-international.org

basic research

& 2015 International Society of Nephrology

Fractalkine–CX3CR1-dependent recruitment and retention of human CD1c þ myeloid dendritic cells by in vitro–activated proximal tubular epithelial cells Andrew J. Kassianos1,2,3, Xiangju Wang1,2, Sandeep Sampangi1,2,3, Sadia Afrin1,2,3, Ray Wilkinson1,2,3,4,5 and Helen Healy1,2,5 1

Conjoint Kidney Research Laboratory, Pathology Queensland, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia; 2Department of Renal Medicine, Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia; 3Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Queensland, Australia and 4Medical School, University of Queensland, Brisbane, Queensland, Australia

Chemokines play pivotal roles in tissue recruitment and retention of leukocytes, with CX3CR1 recently identified as a chemokine receptor that selectively targets mouse kidney dendritic cells (DCs). We have previously demonstrated increased tubulointerstitial recruitment of human transforming growth factor-b (TGF-b)-producing DCs in renal fibrosis and chronic kidney disease (CKD). However, little is known about the mechanism of human DC recruitment and retention within the renal interstitium. We identified CD1c þ DCs as the predominant source of profibrotic TGF-b and highest expressors of the fractalkine receptor CX3CR1 within the renal DC compartment. Immunohistochemical analysis of diseased human kidney biopsies showed colocalization of CD1c þ DCs with fractalkine-positive proximal tubular epithelial cells (PTECs). Human primary PTEC activation with interferon-c and tumor necrosis factor-a induced both secreted and surface fractalkine expression. In line with this, we found fractalkine-dependent chemotaxis of CD1c þ DCs to supernatant from activated PTECs. Finally, in comparison with unactivated PTECs, we showed significantly increased adhesion of CD1c þ DCs to activated PTECs via a fractalkinedependent mechanism. Thus, TGF-b-producing CD1c þ DCs are recruited and retained in the renal tubulointerstitium by PTEC-derived fractalkine. These cells are then positioned to play a role in the development of fibrosis and progression of chronic kidney disease. Kidney International (2015) 87, 1153–1163; doi:10.1038/ki.2014.407; published online 14 January 2015 KEYWORDS: CX3CR1; fractalkine; human dendritic cells; proximal tubular epithelial cells

Correspondence: Ray Wilkinson, Conjoint Kidney Research Laboratory, Pathology Queensland, QIMR Berghofer Medical Research Institute, Level 9, Bancroft Centre, Herston, Queensland 4006, Australia. E-mail: [email protected] 5

These authors contributed equally to this work.

Received 28 August 2014; revised 23 October 2014; accepted 30 October 2014; published online 14 January 2015 Kidney International (2015) 87, 1153–1163

Dendritic cells (DCs) have a vital role in the induction and regulation of immune responses. Human DCs are defined as CD45 þ leukocytes that lack other leukocyte lineage (lin) markers and express high levels of major histocompatibility complex class II (human leukocyte antigen (HLA)-DR) (lin  HLA-DR þ ).1 Within this DC network are a heterogeneous population of cells, comprising multiple subsets with specialized functions.2–5 Human DCs mobilized to peripheral tissues during disease can be categorized into the following: (1) Inflammatory or monocyte-derived DCs (MoDCs) that differentiate from peripheral blood monocytes in response to inflammation;6 and (2) blood and lymphoid organ-resident DCs comprising CD11c  CD123hi plasmacytoid DCs (pDCs) and CD11c þ myeloid DCs (mDCs).7 The mDCs are further divided into CD141 (BDCA-3)hi and CD1c (BDCA-1) þ subsets.1,2 We have previously demonstrated that human kidneys with tubulointerstitial fibrosis, the pathological hallmark of chronic kidney disease (CKD), have significantly elevated numbers of tubulointerstitial CD141hi DCs and CD1c þ DCs compared with nonfibrotic renal tissue, with mDCs identified as key sources of the fibrogenic growth factor transforming growth factor-b (TGF-b).8 However, it remains to be determined how human mDCs are recruited and retained within the renal interstitium. Leukocyte infiltration of inflammatory sites is directed by the local expression of chemokines.9 Among these molecules, fractalkine (chemokine (CX3C motif) ligand 1 (CX3CL1)) represents a unique class of chemokine that mediates two distinct biological actions. Soluble fractalkine is a potent chemoattractant, whereas membrane-bound fractalkine functions as an adhesion molecule to retain circulating leukocytes.10 Both fractalkine-dependent chemotaxis and adhesion are mediated through a specific receptor, CX3CR1, that is not shared by any other chemokine.11 The functional impact of the fractalkine–CX3CR1 system in the progression of tubulointerstitial fibrosis has been studied in murine models of ischemia–reperfusion injury,12 1153

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

glomerulonephritis (GN),13 and lupus nephritis.14 More recently, Hochheiser et al.15 identified CX3CR1 as a chemokine receptor that targets mouse kidney DCs with relative specificity, promoting an influx of cortical DCs and progression of disease severity in a model of CKD. However, the role of fractalkine–CX3CR1 in the recruitment and retention of human kidney DC is unknown. Fractalkine is induced in human renal inflammation associated with acute allograft rejection, vasculitic GN, and crescentic GN, with minimal to no expression in normal kidney tissue or noninflammatory disease such as minimal change nephropathy.16–18 Notably, tubulointersitial fractalkine expression is significantly upregulated in biopsies from fibrotic human kidneys compared with nonfibrotic nephropathies.19 Expression of fractalkine at tubulointerstitial sites is localized to endothelial cells of the peritubular capillary network and tubular epithelial cells, with cellular infiltrates detected adjacent to fractalkine-expressing tubular epithelial cells.16–18,20 However, the proximal or distal origin of these fractalkine-expressing cells has not been defined. The tubulointersitial expression of fractalkine receptor CX3CR1 is also significantly elevated in fibrotic human kidneys compared with nonfibrotic biopsies.21 Immunohistochemical (IHC) and immunofluorescence-based studies of human diseased kidney tissue (native and renal allograft biopsies) have reported CX3CR1 expression on myeloid and lymphoid cells.21–23 However, unlike the flow cytometric staining strategy developed in our laboratory,8 these methods are not amenable to the multiparameter labeling required to unequivocally identify specific leukocyte subpopulations, in particular human DC subsets.

Proximal tubular epithelial cells (PTECs), in the perturbed disease state, have an established role in initiating the inflammatory influx of mononuclear cells via the secretion of chemokines. The in vitro expression of fractalkine by human primary PTECs can be induced by pro-inflammatory cytokines such as tumor necrosis factor-a (TNF-a).17 Soluble fractalkine secreted from cytokine-stimulated PTECs has been shown to have a role, although minor, in the in vitro recruitment of peripheral blood lymphocytes.24 Surfaceexpressed fractalkine on PTECs can also support the adhesion of a CX3CR1 þ monocytic cell line.17 However, the role of PTEC-derived fractalkine in the recruitment and retention of human DC subsets in the interstitium during renal inflammatory responses still remains unknown. Here, we use our novel flow cytometric–based approach to demonstrate human renal DC expression of CX3CR1 by TGF-b-producing CD1c þ DCs within biopsies with tubulointerstitial fibrosis. In addition, our data indicate that these CX3CR1-expressing CD1c þ DCs are recruited and retained in the renal tubulointerstitium via PTEC-derived fractalkine. RESULTS Human CD1c þ DCs are the predominant source of profibrotic TGF-b within the renal DC compartment

We have previously developed a ten-color flow cytometric gating strategy to identify and phenotype infiltrating leukocyte populations in healthy and diseased kidney tissue.8 In this study, we have expanded this methodology to 14 colors. Using a new gating strategy on diseased biopsies with interstitial fibrosis (Figure 1), we were able to separate CD45 þ leukocytes into granulocytes with higher side scatter and

Gated on live, CD45+ singlet cells: Monocytes/DC

104

104

103 MNC

102

DC 105

DC HLA-DR

105 SSC-A

SSC-A

Granulocytes 105

103 102 101

101 0

103

104

Monocytes

0 102

105

0 Lineage (CD3, 19,20,56)

50 K 10 0K 15 0K 20 0K 25 0K

103 102 0

Lymphocytes 102

104

103 104 105 CD14

FSC-A CD141hi DC

105

105

104

104 CD141

CD123

Gated on dendritic cells: pDC

103 102 0

103 104 CD11c

CD1c+ DC

102 0

MDC

0 102

103

105

0 102

103 104 105 CD1c

Figure 1 | Identification of DC subsets in human fibrotic kidney tissue. Gating strategy used to identify human dendritic cell (DC) subpopulations (CD11c  CD123hi plasmacytoid DC (pDC), CD11c þ CD141hi, and CD11c þ CD1c þ myeloid DC (mDC)) within the lineage  CD14  HLA-DR þ fraction in fibrotic kidney tissue. Representative flow cytometric data from 1 of 12 individual fibrotic renal biopsies are shown. 1154

Kidney International (2015) 87, 1153–1163

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

with minimal or no expression by pDCs and CD141hi DCs (Figure 2a and b). We next assessed the maturation state of CD1c þ DCs in healthy and diseased kidney tissue, with diseased biopsies stratified on the basis of the absence or presence of interstitial fibrosis. Although CD1c þ DC expression of the immunoregulatory molecule PD-L1, costimulatory molecule CD40, and activation marker CD83 were comparable between healthy and diseased tissue (Figure 3a-c), levels of costimulatory molecules CD80 and CD86 and integrin CD11c were significantly elevated in both fibrotic and nonfibrotic disease groups compared with healthy tissue (Figure 3d–f). Collectively, these data associate activated, pathogenic CD1c þ DCs with the development of renal interstitial fibrosis through the production of TGF-b.

mononuclear cells. These mononuclear cells were further divided into lin (CD3, CD19, CD20, and CD56) þ lymphocytes, lin  CD14 þ HLA-DR þ monocytes/macrophages, and lin  CD14  HLA-DR þ DCs. The gatings for CD14 and HLA-DR expression in Figure 1 were selected on the basis of isotype-matched control antibodies (Supplementary Figure 1). The DC compartment was fractionated into CD11c  CD123hi pDCs and CD11c þ mDCs, with the latter further divided into CD141hi and CD1c þ DC populations. An identical gating strategy was also used for healthy kidney tissue and nonfibrotic renal biopsies (data not shown). Our earlier work showed significantly elevated TGF-b levels in diseased biopsies with interstitial fibrosis compared with nonfibrotic biopsies, with monocytes/macrophages and DCs identified as the highest expressors of TGF-b latencyassociated peptide,8 a component of the TGF-b complex secreted by cells.25 We extend these findings in this current study by identifying CD1c þ DCs as the major source of TGF-b within the DC compartment of fibrotic kidney tissue,

Our flow cytometric analysis of leukocytes from human blood and fibrotic kidney tissue revealed expression of CX3CR1 on a subpopulation of lymphocytes and abundant expression on

Lymphocytes 100

80

80

80

60 40

% Of max

100

20

60 40

103

104

40

0

0

0 102

60

20

20

0 105

0 102

pDC

103

104

105

0 102

CD141hi DC 100

80

80

80

40

% Of max

100

60

60 40

0 102

103

104

105

60 40

0

0

0

104

20

20

20

103

CD1c+ DC

100

% Of max

% Of max

Monocytes

100

% Of max

% Of max

Granulocytes

Human CD1c þ DC express high levels of CX3CR1

105

0 102

104

105

0 102

103

104

105

LAP (TGF-β)

Isotype

**

*** ***

100

103

**

% TGF-β+ Cells

80 60 40 20 0 Monocytes

pDC

CD141hi DC CD1c+ DC

þ

Figure 2 | Human CD1c DCs express high levels of TGF-b in fibrotic kidney tissue. (a) Flow cytometric detection of cell surface latencyassociated peptide (LAP) (transforming growth factor-b (TGF-b)) (black unfilled) compared with isotype control (gray filled) on leukocyte populations in fibrotic kidney tissue. Representative results from one of five diseased biopsies with interstitial fibrosis are presented. (b) Percentage of TGF-b þ cells within the monocyte compartment and within DC subsets from five individual fibrotic renal biopsies; bars represent means. **Po0.01, ***Po0.001 by one-way analysis of variance with Bonferroni’s multiple-comparison test. Kidney International (2015) 87, 1153–1163

1155

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

PD-L1

CD40

1500

CD83 1500

800 600

500

1000 MFI

MFI

MFI

1000

400

500

200

0

0

0 Healthy

Healthy

–Fibrosis +Fibrosis

–Fibrosis +Fibrosis

CD80 3000

*

Healthy –Fibrosis +Fibrosis

CD86

*

10,000

*

8000

1000

8000

6000 4000

4000

0 Healthy

–Fibrosis +Fibrosis

**

2000

2000 0

***

6000 MFI

MFI

MFI

2000

CD11c

**

0 Healthy

–Fibrosis +Fibrosis

Healthy

–Fibrosis +Fibrosis

þ

Figure 3 | Human CD1c DCs in fibrotic kidney tissue display a mature phenotype. Surface expression of PD-L1 (a), CD40 (b), CD83 (c), CD80 (d), CD86 (e), and CD11c (f) on CD1c þ dendritic cells (DCs) in healthy kidney tissue and diseased biopsies without and with fibrosis. Values of mean fluorescence intensity (MFI) for individual donors are presented; bars represent means. *Po0.05, **Po0.01, ***Po0.001 by oneway analysis of variance with Bonferroni’s multiple-comparison test.

Lymph

CD1c+ DC

80

80

80

80

80

40

60 40

60 40

60 40

60 40

20

20

20

20

20

0

0

0

0

0

0 102 103 104 105

0 102 103 104 105

0102 103 104 105

100

80

80

40

40

60 40

% Of max

100

80

% Of max

100

80

% Of max

100

80

60

60 40

20

20

20

0

0

0

0

0

0 102 103 104 105 Isotype

0102 103 104 105

0 102 103 104 105 100

40

20 0 102 103 104 105

0

60

20 0 102 103 104 105

40

0102 103 104 105

100 60

60 20

% Of max

60

% Of max

80

% Of max

100

% Of max

100

% Of max

100

% Of max

% Of max

CD141hi DC

pDC

100

0 102 103 104 105

Fibrotic kidney

Mono

100 % Of max

% Of max

Peripheral blood

Gran 100

80 60 40 20 0

0102 103 104 105

0 102 103 104 105

CX3CR1

CX3CR1

100

MFI

400

200

0 Healthy

–Fibrosis

+Fibrosis

þ

Figure 4 | Human CD1c DCs express high levels of CX3CR1. (a) Flow cytometric detection of cell surface CX3CR1 (black unfilled) compared with isotype control (gray filled) on granulocytes (Gran), lymphocytes (Lymph), monocytes (Mono), and dendritic cell (DC) subsets from human peripheral blood and fibrotic kidney tissue. Representative results from one of four peripheral blood samples and one of seven diseased biopsies with interstitial fibrosis are presented. (b) Surface expression of CX3CR1 on CD1c þ DC in healthy kidney tissue and diseased biopsies without and with fibrosis. Values of mean fluorescence intensity (MFI) for individual donors are presented; bars represent means.

monocytes and, within the DC compartment, on CD1c þ DCs (Figure 4a). Expression of CX3CR1 on pDCs and CD141hi DCs was minimal or absent (Figure 4a). The levels of 1156

CX3CR1 on CD1c þ DCs were comparable between healthy and diseased fibrotic and nonfibrotic tissue (Figure 4b), ruling out a disease-driven regulation of CX3CR1 expression in fibrosis. Kidney International (2015) 87, 1153–1163

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

Aquaporin-1

Fractalkine

CD1c

Fractalkine 1000

*

pg/ml

800 600 400 200 0 –Fibrosis

+Fibrosis

Figure 5 | Colocalization of human CD1c þ dendritic cells with fractalkine-expressing proximal tubular epithelial cells. (a) Levels of soluble fractalkine in dissociation supernatants of diseased biopsies without and with interstitial fibrosis, as measured by enzyme-linked immunosorbent assay; bars represent means. *Po0.05 by two-tailed t-test. (b–d) Serial section immunohistochemical staining of fibrotic kidney tissue stained for aquaporin-1 (b), fractalkine (c), and CD1c (d) and lightly counterstained with hematoxylin (purple). Scale bars ¼ 200 mm (top panels) and 60 mm (bottom panels). Representative results from one of three individual fibrotic renal biopsies are shown.

Colocalization of human CD1c þ DCs with fractalkine-expressing PTECs

We postulated that the elevated CD1c þ DC recruitment in renal fibrosis that we have previously reported8 may be mediated via a fractalkine–CX3CR1-dependent mechanism. Supporting this hypothesis, levels of soluble fractalkine in supernatants of dissociated fibrotic biopsies were significantly increased compared with nonfibrotic biopsies (Figure 5a). The source of this fractalkine in fibrotic renal biopsies was assessed using IHC, with strong fractalkine expression detected in renal tubules (Figure 5c). Fractalkine-expressing tubular epithelial cells were also shown, for the first time, to stain positive for aquaporin-1 (Figure 5b), a marker restricted to the proximal regions of the renal tubules.26 Serial section analysis highlighted a tubulointerstitial accumulation of CD1c þ cells (with dendritic morphology) adjacent to these fractalkine-expressing PTECs (Figure 5d). The specificity of IHC staining was confirmed with isotype-matched control antibodies on adjacent serial sections to the test antibodies (Supplementary Figure 2a-c). These findings suggest that PTEC production of fractalkine may be pivotal in the trafficking and retention of CX3CR1-expressing CD1c þ DCs in the diseased renal interstitium.

significantly elevated expression of fractalkine was observed when IFN-g and TNF-a were applied in combination (Figure 6a-d), demonstrating synergy between the two cytokines. Thus, for all subsequent in vitro functional studies, PTECs were stimulated with both IFN-g and TNF-a. Activated PTECs chemoattract human CD1c þ DCs via a fractalkine-dependent mechanism

We next investigated the functional capacity of PTEC-derived soluble fractalkine to chemoattract CX3CR1-expresssing human blood CD1c þ DCs (Figure 7a) in a transwell assay. Quantification of migrated cells with Flow-Count Fluorospheres (Beckman Coulter, Brea, CA) revealed significantly enhanced CD1c þ DC chemotaxis in response to recombinant human fractalkine (Figure 7b) and activated PTEC supernatants (Figure 7c). The addition of fractalkineblocking antibody in the lower chamber significantly reduced this chemotactic response (Figure 7b and c). The presence of fractalkine-blocking antibody did not alter CD1c þ DC viability (Figure 7d). Taken together, these in vitro findings indicate that activated PTECs produce functionally active fractalkine that is capable of promoting human CD1c þ DC migration.

PTECs significantly upregulate fractalkine in response to inflammatory cytokines

Cell surface–expressed fractalkine promotes adhesion of human CD1c þ DCs to activated PTECs

These IHC findings prompted us to investigate the in vitro production of intracellular, cell-surface, and soluble fractalkine by PTECs in response to interferon-g (IFN-g) and TNF-a, both key pro-inflammatory cytokines in renal inflammatory conditions.27,28 An upregulation of intracellular, surface, and secreted fractalkine was induced by IFN-g, whereas TNF-a had comparatively negligible effects (Figure 6a-d). Notably,

In a final set of experiments, we examined the role of fractalkine as an adhesion molecule on the cell surface of activated PTECs. Human blood CD1c þ DCs were fluorescently labeled with calcein-acetoxymethyl ester dye (Figure 8a) and cocultured with unactivated or activated PTECs. Adhesion of CD1c þ DCs to activated PTECs was significantly elevated compared with unactivated PTECs (Figure 8b and c).

Kidney International (2015) 87, 1153–1163

1157

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

Intracelluar fractalkine

Surface fractalkine

**

**

30 20 10

IFN-γ

TNF-α

***

30 20 10

IFN-γ + TNF-α

50

TNF-α

ng/ml

105

104

103

100 80 60 40 20 0

IFN-γ + TNF-α

***

***

30 20 10 0 IFN-γ

TNF-α

IFN-γ + TNF-α

105

104

Nil

0 102

% Of max

***

40

IFN-γ +TNF-α

105

104

103

IFN-γ

Soluble fractalkine

0 102

% Of max 105

104

103

0 102

100 80 60 40 20 0

TNF-α 100 80 60 40 20 0 0 102

Nil

IFN-γ

Nil 100 80 60 40 20 0

103

Nil

% Of max

*

40

0

0

% Of max

***

50 % Fractalkine+ cells

% Fractalkine+ cells

40

Figure 6 | Human PTECs express high levels of fractalkine in response to IFN-c and TNF-a stimulation. (a and b) Flow cytometric detection of intracellular (a) and cell surface (b) fractalkine on human primary positive proximal tubular epithelial cells (PTECs) left unstimulated (Nil) or stimulated with interferon-g (IFN-g), tumor necrosis factor-a (TNF-a), or a combination of both cytokines for 24 h. Percentage values of fractalkine þ PTECs from three different donors are presented; bars represent means. (c) Cell surface fractalkine (black unfilled) compared with isotype control (gray filled) on human primary PTECs. Representative histograms were replicated with PTECs from three different donors. (d) Soluble fractalkine shed by human primary PTECs left unstimulated (Nil) or stimulated with IFN-g, TNF-a, or a combination of both cytokines for 24 h. Soluble fractalkine levels were measured by enzyme-linked immunosorbent assay of PTEC culture supernatants; bars represent means. *Po0.05, **Po0.01, ***Po0.001 by one-way analysis of variance with Bonferroni’s multiple-comparison test.

Furthermore, this adhesion was significantly inhibited when the activated PTECs were pretreated with a fractalkineblocking antibody (Figure 8b and c). These in vitro data suggest that fractalkine-dependent adhesion may be pivotal for the retention of pathogenic human CD1c þ DCs within the tubulointerstitium. DISCUSSION

DCs are key players in kidney diseases. In murine studies, DCs have a protective function in acute immune-mediated kidney disease,29 but they have a pathogenic role in models of CKD.30–34 Although the fractalkine–CX3CR1 system contributes to mouse DC trafficking into the inflamed kidney cortex,15 the processes that drive the tubulointerstitial recruitment of human DCs during renal inflammation have not been well characterized, apart from the role of chemerin in the recruitment of pDCs in lupus nephritis.35 In our study, we highlight, for the first time, a potential role for PTECderived fractalkine in the recruitment and retention of TGF-b-producing, CX3CR1-expressing CD1c þ DCs within the tubulointerstitium during fibrotic kidney disease. Collectively, these data provide evidence of functional parallels between mouse and human renal DC trafficking. 1158

We showed that, in contrast to other kidney DC subsets (pDC and CD141hi DC), CD1c þ DCs expressed high levels of TGF-b in diseased biopsies with interstitial fibrosis. With TGF-b established as the principal fibrogenic growth factor in renal scarring,36,37 this highlights a potentially pivotal and pathogenic role for CD1c þ DC in CKD progression. Interestingly, a recent study of human blood DCs demonstrated that CD1c þ DCs, but not CD141hi DCs, express membrane-bound TGF-b protein.38 Furthermore, human blood and lung CD1c þ DCs display higher expression of molecules involved in TGF-b activation compared with CD141hi DCs of the same origin.38 These data suggest that TGF-b production by human CD1c þ DCs is conserved across different human tissues and may represent an inherent characteristic of this DC subset. Consistent with this putative pathogenic role, CD1c þ DCs from diseased kidney biopsies also displayed increased surface expression of costimulatory/maturation molecules CD80 and CD86 compared with healthy kidney tissue. This acquisition of a mature phenotype concurs with previous findings in a mouse model of CKD that pathogenic kidney DCs express elevated levels of CD80 and CD86.33 Previous IHC/immunofluorescence-based studies of human renal diseases have reported CX3CR1 expression on Kidney International (2015) 87, 1153–1163

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

Gated on live, singlet cells:

–anti-Fract

** **

104 103 102 0

30,000 20,000 10,000 0

0 102 103 104 105 Lineage (CD3, 14, 19, 20, 56) –anti-Fract

**

40,000

0 ng/ml rhFract

+anti-Fract

**

25 ng/ml rhFract

–anti-Fract

*

+anti-Fract

100

30,000

% Viability

DC counts

+anti-Fract

**

40,000

97.4% DC counts

HLA-DR

105

20,000

50

10,000 0 Nil PTEC S/N

0 25 ng/ml rhFract

IFN-γ +TNF-α PTEC S/N

IFN-γ +TNF-α PTEC S/N

Figure 7 | Chemoattraction of human CD1c þ DCs by PTEC-derived fractalkine. (a) Human CD1c þ dendritic cells (DCs) were isolated by positive immunomagnetic selection from whole blood–derived buffy coats. Purified CD1c þ DCs were identified as lineage (CD3, 14, 19, 20, 56)  HLA-DR þ cells. The dot plot shown is from a single representative donor. (b and c) Chemotactic responses of CD1c þ DCs to recombinant human fractalkine (rhFract) (b) and unactivated (Nil) or activated (interferon-g þ tumor necrosis factor-a (IFN-g þ TNF-a)) positive proximal tubular epithelial cell (PTEC) supernatants (c). Chemoattractants were preincubated in the absence (  anti-Fract) or presence (þanti-Fract) of a fractalkine-blocking antibody. Migrated CD1c þ DCs (lineage (CD3, 14, 19, 20, 56)  HLA-DR þ CD11c þ CD1c þ cells) were identified and enumerated by flow cytometry. Results represent mean±s.e.m. of four (b) and three (c) individual experiments. *Po0.05, **Po0.01 by oneway analysis of variance with Bonferroni’s multiple-comparison test. (d) Viability of CD1c þ DCs in the absence (  anti-Fract) or presence ( þ anti-Fract) of a fractalkine-blocking antibody was determined using a LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit with flow cytometry. Results represent mean±s.e.m. of four and three individual experiments assessing recombinant human fractalkine (rhFract) and activated (IFN-g þ TNF-a) PTEC supernatants, respectively.

Gated on live, singlet cells: 80 60 40 20 0 0 102 103 104 105 Calcein

** **

1200 800 400 0

+anti-Fract

–anti-Fract

*

** **

40 % Adhesion

Fluorescence units

% Of max

100

+anti-Fract

–anti-Fract

**

30 20 10 0

Nil PTEC

IFN-γ+TNF-α PTEC

Nil PTEC

IFN-γ+TNF-α PTEC

Figure 8 | Fractalkine-dependent adhesion of human CD1c þ DCs to activated PTECs. Purified blood CD1c þ dendritic cells (DCs) were fluorescently labeled with calcein-acetoxymethyl ester dye (black unfilled histogram compared with gray filled histogram of unlabeled control) (a) and assayed for adhesion to unactivated (Nil) or activated (interferon-g þ tumor necrosis factor-a (IFN-g þ TNF-a)) positive proximal tubular epithelial cells (PTECs). Fractalkine-mediated adhesion was evaluated by preincubation of PTECs in the absence (  anti-Fract) or presence ( þ anti-Fract) of a neutralizing antibody to fractalkine. After threefold washing, the fluorescent signal from adherent cells was recorded and is presented as both fluorescent units (b) and percent adhesion (c). Results for b and c represent mean±s.e.m. of three individual experiments, with each experiment consisting of triplicate samples. *Po0.05, **Po0.01 by one-way analysis of variance with Bonferroni’s multiplecomparison test.

CD3 þ lymphocytes and CD68 þ monocytes/macrophages infiltrating the glomerulus and the interstitium22 and on putative interstitial DCs, identified on the basis of their expression of CD11c21 or CD209/DC-SIGN.23 However, recent advances in the characterization of human DC Kidney International (2015) 87, 1153–1163

subsets report that these markers are not DC-specific, with broader expression of both antigens on monocyte/ macrophage lin cells.6,39,40 This issue further highlights the importance of our multicolor, flow cytometric methodology for human DC subset identification, and phenotyping. Our 1159

basic research

flow cytometric analysis of human blood and kidney DCs showed CX3CR1 expression on CD1c þ DCs from both sources, but not pDCs and CD141hi DCs. These results are consistent with previous DC studies of human blood and lung41,42 and point to the recruitment of CD1c þ DCs from the blood into the kidney. Fractalkine, the sole ligand for CX3CR1, is present in human glomerular and interstitial renal disease.16–18 Notably, Koziolek et al.19 reported a significant increase in tubulointersitial fractalkine mRNA expression in fibrotic human kidneys compared with nonfibrotic nephropathies. Our study has corroborated these data at the protein level, with significantly elevated levels of soluble fractalkine in supernatants of dissociated fibrotic biopsies compared with nonfibrotic biopsies. Previous human studies have demonstrated in vivo expression of fractalkine protein by renal tubular epithelial cells.16–18,20 However, we extend these reports by providing in vivo evidence that PTECs, identified on the basis of their expression of aquaporin-1, are a major source of fractalkine in fibrotic kidney disease. Fractalkine expression by human PTECs in vitro has been shown to be induced by protein overload43 and TNF-a.17 In our study, both IFN-g and TNF-a in combination were required for the optimal in vitro induction of intracellular, cell surface, and soluble fractalkine by human primary PTECs. These results mirror previous reports that IFN-g and TNF-a cooperate in the induction of fractalkine in human endothelial cells,44 smooth muscle cells,45 and astrocytes.46 To our knowledge, this report is the first to demonstrate chemotaxis of CD1c þ DCs toward recombinant human fractalkine. Importantly, we also demonstrated that soluble fractalkine derived from activated PTEC cultures was biologically active, as it was found to be responsible, in part, for the chemoattraction of CD1c þ DCs by PTEC-conditioned medium. Our human data support and extend previous evidence in mice that CX3CR1 contributes to entry of DCs into the inflamed kidney.15 We also show that membrane-associated fractalkine on the surface of activated PTECs functions as an adhesion molecule for CX3CR1-expressing CD1c þ DCs. Using a similar in vitro system, Chakravorty et al.17 previously reported that fractalkine–CX3CR1 interactions accounted for only modest levels of adhesion between TNF-a-treated PTECs and both monocytic THP-1 cells and peripheral blood natural killer cells. This minor role of fractalkine–CX3CR1 may be reflective of the low level of fractalkine surface expression on TNF-a-treated PTECs reported in both their study17 and our current report. However, it may also be indicative of a specialized fractalkine–CX3CR1 dependence of human CD1c þ DCs in comparison with other infiltrating leukocyte populations. On the basis of our results, we propose that the expression of fractalkine by activated PTECs represents a mechanism for the tubulointerstital recruitment and retention of pathogenic CD1c þ DCs that may drive the development of interstitial fibrosis and subsequent loss of renal function in CKD. 1160

AJ Kassianos et al.: Recruitment and retention of human renal DCs

Importantly, our current study also provides the first human evidence supporting the mouse data of Hochheiser et al.15 that the CX3CR1-dependent recruitment of interstitial DCs into the inflamed kidney promotes disease progression. Future clinical studies are now required to assess the therapeutic benefit of targeting the fractalkine–CX3CR1 system, and therefore migration and adhesion of human CD1c þ DCs, in fibrotic renal disease. MATERIALS AND METHODS Kidney tissue specimens Renal cortical tissue was obtained with informed patient consent from the macroscopically/microscopically healthy portion of tumor nephrectomies or native diseased biopsies, following approval by the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (2002/011 and 2006/072). Healthy cortical tissue was obtained from 8 donors (4 female/4 male) of mean age 55±10 years, whereas diseased clinical biopsies were obtained from 24 donors (14 female/10 male) of mean age 53±18 years. A range of primary diagnoses was sampled, including 13 glomerular immunemediated (crescentic GN, membranoproliferative GN, necrotizing GN, pauci-immune GN, IgA nephropathy, membranous nephropathy, and minimal change disease—4/13 fibrotic), 7 glomerular nonimmune-mediated (amyloidosis, focal segmental glomerulosclerosis, and diabetic nephropathy—4/7 fibrotic), and 4 nonglomerular (interstitial nephritis and hypertensive nephropathy—4/4 fibrotic) etiologies. Fresh biopsies were divided for the following: (1) tissue dissociation (1–5mm of a core biopsy); (2) freezing in Tissue-Tek OCT compound (Sakura, Torrance, CA) for IHC analysis; and (3) fixation in formalin for assessing the levels of interstitial fibrosis/tubular atrophy by renal histopathologists. For assessment of renal interstitial fibrosis, formalin-fixed 4-mm sections were stained with Masson’s trichrome, and the proportion of fibrotic area in the cortex was quantified over 20 high-power fields. Biopsies displaying X5% interstitial fibrosis were deemed fibrotic, on the basis of the Banff 97 working classification of renal pathology.47 According to this criterion, diseased specimens were then grouped into biopsies without (n ¼ 12; 7 female/5 male; mean age 48±18 years; mean estimated glomerular filtration rate (eGFR) 78±24 ml/min per 1.73 m2) or with interstitial fibrosis (n ¼ 12; 7 female/5 male; mean age 58±17 years; mean eGFR 34±26 ml/min per 1.73 m2). Kidney function (eGFR) was calculated using the MDRD method by AUSLAB (Queensland Health, Brisbane, Australia). Tissue dissociation for flow cytometric analysis Healthy kidney tissue and diseased biopsies were digested with 1 mg/ml collagenase P (Roche, Mannheim, Germany) in the presence of 20 mg/ml DNase I (Roche; 250-ml volume) for 15 min. Supernatant from this initial dissociation step was collected for cytokine/ chemokine analysis. Dissociated tissue was then further digested with 10 mg/ml trypsin/4 mg/ml EDTA (Invitrogen, Grand Island, NY) (500-ml volume) for 10 min. Flow cytometry Single-cell suspensions were initially stained with LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit (Invitrogen) to allow exclusion of nonviable cells. Cells were then stained on ice for 30 min with combinations of test (0.25 mg per antibody) (Table 1) or isotypematched control antibodies in cold FACS buffer (0.5% bovine serum Kidney International (2015) 87, 1153–1163

AJ Kassianos et al.: Recruitment and retention of human renal DCs

Table 1 | Antibodies used for flow cytometric staining Antigen

Clone

Fluorochrome

Source

CD45 CD3a CD19a CD20a CD56a CD14 HLA-DR CD11c CD123 CD1c CD141 CD40 CD80 CD83 CD86 PD-L1 CX3CR1 LAP (TGF-b1)

HI30 HIT3a HIB19 2H7 HCD56 M5E2 L243 3.9 7G3 L161 AD5-14H12 5C3 2D10 HB15e IT2.2 29E.2A3 2A9-1 27232

Brilliant Violet 510 PerCP/Cy5.5 PerCP/Cy5.5 PerCP/Cy5.5 PerCP/Cy5.5 Alexa Fluor 700 Brilliant Violet 785 Brilliant Violet 711 PE-CF594 FITC or BV421 APC PE/Cy7 Brilliant Violet 650 PE Brilliant Violet 605 Brilliant Violet 421 PE/Cy7 PE

Biolegend Biolegend Biolegend Biolegend Biolegend Biolegend Biolegend Biolegend BD Biolegend Miltenyi Biotec Biolegend Biolegend Biolegend Biolegend Biolegend Biolegend R&D Systems

Abbreviations: APC, allophycocyanin; HLA, human leukocyte antigen; LAP, latencyassociated peptide; PE, phycoerythrin. a Lineage/lymphocyte mix.

albumin (Sigma-Aldrich, St Louis, MO) and 0.02% sodium azide (Sigma-Aldrich) in phosphate buffered saline). Cell acquisition was performed on an LSR Fortessa (BD, San Jose, CA), and data were analyzed with the FlowJo software (TreeStar, Ashland, OR). IHC staining Frozen 7-mm tissue sections from three fibrotic renal biopsies were fixed with 25% ethanol:75% acetone at room temperature for 5 min. Endogenous peroxidase activity was blocked with 1% H2O2 for 10 min, followed by blocking of endogenous biotin activity with an Avidin/Biotin Blocking Kit (Vector Laboratories, Burlingame, CA) and a protein block with Background Sniper Blocking Reagent (Biocare Medical, Concord, CA). Serial sections from each fibrotic renal biopsy were probed with anti-aquaporin-1 (Rabbit polyclonal IgG; Santa Cruz, Dallas, TX), anti-fractalkine (Goat polyclonal IgG; R&D Systems, Minneapolis, MN), and biotinylated anti-CD1c (Clone L161; Biolegend, San Diego, CA), or isotype-matched control antibodies at room temperature for 1 h. Tissue sections were washed and rabbit or goat horseradish peroxidase (HRP) polymer systems (Biocare Medical) or HRP-conjugated streptavidin (Invitrogen) were applied according to the manufacturers’ instructions. Peroxidase activity was developed with DAB substrate for 10 min. Sections were lightly counterstained with hematoxylin and mounted using DPX Mounting Medium. Isolation and culture of human primary PTECs PTECs were purified from the healthy portion of tumor nephrectomies according to the method of Glynne and Evans48 and cultured in defined medium (DM) as previously described.5 Activation of human primary PTECs PTECs were cultured in DM until they reached 70–80% confluence, and then they were further cultured for 24 h alone or in the presence of IFN-g (100 ng/ml) and/or TNF-a (20 ng/ml) (both from R&D Systems). Supernatants were collected for cytokine/chemokine analysis Kidney International (2015) 87, 1153–1163

basic research

and chemotaxis assays, whereas PTECs were stained for fractalkine expression. For surface fractalkine staining, PTECs were resuspended in cold FACS buffer and labeled on ice for 1 h with 0.25 mg of antifractalkine-PE (Clone 51637; R&D Systems) or isotype-matched control antibody. For intracellular staining, PTECs were fixed in 4% paraformaldehyde at room temperature for 10 min and washed/ permeabilized in a 0.1% saponin balanced salt solution before fractalkine labeling at room temperature for 30 min. Cell acquisition was performed on an LSR Fortessa, as described previously. Soluble fractalkine detection Soluble fractalkine was detected using the Quantikine Human CX3CL1/Fractalkine Immunoassay (R&D Systems) according to the manufacturer’s recommendations. Cytokine values for dissociation supernatants were normalized to the volume of 5 mm of renal core biopsy tissue. Human blood CD1c þ DC isolation Leukocyte-rich buffy coats were obtained from healthy blood donors (Australian Red Cross Blood Service, Brisbane, Australia). Mononuclear cells were isolated from buffy coats using Ficoll-Paque Plus density gradient centrifugation (Amersham Biosciences, Uppsala, Sweden). CD1c þ DCs were isolated by positive immunomagnetic selection using the CD1c (BDCA-1) þ DC isolation kit (495% lin (CD3, 14, 19, 20, 56)  HLA-DR þ cells; Miltenyi Biotec, Auburn, CA). DC cultures were performed in complete medium, as previously described.49 Chemotaxis assay Migration assays were performed using 24-well Costar Transwell plates (5-mm pore size; Sigma-Aldrich). Human blood CD1c þ DCs (200,000 cells in Complete Medium, 100-ml volume) were seeded in the upper chamber and 600 ml of DM alone or with recombinant human fractalkine (25 ng/ml; R&D Systems) or PTEC supernatants (diluted in DM at a preoptimized 1:5 ratio) were added to the lower chamber. All chemotaxis media in the lower chamber were supplemented with 10% heat-inactivated fetal bovine serum (FBS). For selected samples, the chemotaxis medium was preincubated in the presence of 20 mg/ml anti-fractalkine (Clone 81506; R&D Systems) at 37 1C for 30 min. After 3 h of incubation at 37 1C, migrated cells were recovered from the lower well and enumerated by flow cytometry. Absolute CD1c þ DC numbers (lin (CD3, 14, 19, 20, 56)  HLA-DR þ CD11c þ CD1c þ cells) were determined using Flow-Count Fluorospheres according to the manufacturer’s recommendations. Briefly, CD1c þ DC concentrations (cells per ml) were calculated as follows: total number of CD1c þ DCs counted/total number of Fluorospheres countedconcentration of Flow-Count Fluorospheres. This value was then multiplied by the total sample volume to obtain absolute CD1c þ DC counts for each sample. Cell adhesion assay Human primary PTECs were cultured in DM in Costar 96-well clear bottom, black plates (Sigma-Aldrich) until 70–80% confluency and then further cultured for 24 h alone or in the presence of IFN-g (100 ng/ml) and TNF-a (20 ng/ml). For selected samples, PTECs were then preincubated with anti-fractalkine (20 mg/ml) at 37 1C for 30 min and washed twice with Hank’s Buffered Salt Solution (HBSS) þ 1% FBS. 1161

basic research

Freshly isolated blood CD1c þ DCs were resuspended at 2106 cells/ml in HBSS þ 1% FBS and labeled with 5 mM calcein-acetoxymethyl ester dye (Invitrogen) at 37 1C for 30 min. Excess dye was removed by centrifugation, and the labeled cells were resuspended in HBSS þ 1% FBS. Labeled CD1c þ DCs (100,000 cells in HBSS þ 1% FBS, 100 ml volume) were added to the PTEC monolayer and incubated at 37 1C for 3 h. The plate was gently washed three times to remove nonadherent cells, and the fluorescent signal was measured in a Synergy H4 plate reader (excitation wavelength 485 nm; emission wavelength 528 nm; Biotek, Winooski, VT). Percent adhesion was calculated as 100bound fluorescent units (FU)/(total FU  spontaneous FU). Statistics Comparisons between two groups were performed using two-tailed t-tests and multiple comparisons using one-way analysis of variance. Statistical tests were performed using Prism 5.0 analysis software (GraphPad Software, La Jolla, CA). P-values p0.05 were considered statistically significant. DISCLOSURE

AJ Kassianos et al.: Recruitment and retention of human renal DCs

12.

13.

14.

15.

16. 17.

18. 19.

20.

21.

All the authors declared no competing interests. 22.

ACKNOWLEDGMENTS

The work was funded in part by Pathology Queensland. AJK was supported by a RBWH Post-doctoral Fellowship. We thank the tissue donors and renal histopathologist Dr Leo Francis (Queensland Health) for assessment of interstitial fibrosis levels in kidney biopsies. SUPPLEMENTARY MATERIAL Figure S1. Identification of monocytes and DC in human fibrotic kidney tissue. Figure S2. IHC negative control staining of fibrotic kidney tissue. Supplementary material is linked to the online version of the paper at http://www.nature.com/ki

23.

24.

25.

26. 27.

REFERENCES 1.

Kassianos AJ, Jongbloed SL, Hart DN et al. Isolation of human blood DC subtypes. Methods Mol Biol 2010; 595: 45–54. 2. Jongbloed SL, Kassianos AJ, McDonald KJ et al. Human CD141 þ (BDCA3) þ dendritic cells (DCs) represent a unique myeloid DC subset that cross-presents necrotic cell antigens. J Exp Med 2010; 207: 1247–1260. 3. Radford KJ, Turtle CJ, Kassianos AJ et al. CD11c þ blood dendritic cells induce antigen-specific cytotoxic T lymphocytes with similar efficiency compared to monocyte-derived dendritic cells despite higher levels of MHC class I expression. J Immunother 2006; 29: 596–605. 4. Kassianos AJ, Hardy MY, Ju X et al. Human CD1c (BDCA-1) þ myeloid dendritic cells secrete IL-10 and display an immuno-regulatory phenotype and function in response to Escherichia coli. Eur J Immunol 2012; 42: 1512–1522. 5. Kassianos AJ, Sampangi S, Wang X et al. Human proximal tubule epithelial cells modulate autologous dendritic cell function. Nephrol Dial Transplant 2013; 28: 303–312. 6. Segura E, Touzot M, Bohineust A et al. Human inflammatory dendritic cells induce Th17 cell differentiation. Immunity 2013; 38: 336–348. 7. MacDonald KP, Munster DJ, Clark GJ et al. Characterization of human blood dendritic cell subsets. Blood 2002; 100: 4512–4520. 8. Kassianos AJ, Wang X, Sampangi S et al. Increased tubulointerstitial recruitment of human CD141hi CLEC9A þ and CD1c þ myeloid dendritic cell subsets in renal fibrosis and chronic kidney disease. Am J Physiol Renal Physiol 2013; 305: F1391–F1401. 9. Liu Y. Cellular and molecular mechanisms of renal fibrosis. Nat Rev Nephrol 2011; 7: 684–696. 10. Bazan JF, Bacon KB, Hardiman G et al. A new class of membrane-bound chemokine with a CX3C motif. Nature 1997; 385: 640–644. 11. Imai T, Hieshima K, Haskell C et al. Identification and molecular characterization of fractalkine receptor CX3CR1, which mediates both leukocyte migration and adhesion. Cell 1997; 91: 521–530.

1162

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

Furuichi K, Gao JL, Murphy PM. Chemokine receptor CX3CR1 regulates renal interstitial fibrosis after ischemia-reperfusion injury. Am J Pathol 2006; 169: 372–387. Feng L, Chen S, Garcia GE et al. Prevention of crescentic glomerulonephritis by immunoneutralization of the fractalkine receptor CX3CR1 rapid communication. Kidney Int 1999; 56: 612–620. Inoue A, Hasegawa H, Kohno M et al. Antagonist of fractalkine (CX3CL1) delays the initiation and ameliorates the progression of lupus nephritis in MRL/lpr mice. Arthritis Rheum 2005; 52: 1522–1533. Hochheiser K, Heuser C, Krause TA et al. Exclusive CX3CR1 dependence of kidney DCs impacts glomerulonephritis progression. J Clin Invest 2013; 123: 4242–4254. Cockwell P, Chakravorty SJ, Girdlestone J et al. Fractalkine expression in human renal inflammation. J Pathol 2002; 196: 85–90. Chakravorty SJ, Cockwell P, Girdlestone J et al. Fractalkine expression on human renal tubular epithelial cells: potential role in mononuclear cell adhesion. Clin Exp Immunol 2002; 129: 150–159. Furuichi K, Wada T, Iwata Y et al. Upregulation of fractalkine in human crescentic glomerulonephritis. Nephron 2001; 87: 314–320. Koziolek MJ, Muller GA, Zapf A et al. Role of CX3C-chemokine CX3C-L/fractalkine expression in a model of slowly progressive renal failure. Nephrol Dial Transplant 2010; 25: 684–698. Durkan AM, Alexander RT, Liu GY et al. Expression and targeting of CX3CL1 (fractalkine) in renal tubular epithelial cells. J Am Soc Nephrol 2007; 18: 74–83. Koziolek MJ, Schmid H, Cohen CD et al. Potential role of fractalkine receptor expression in human renal fibrogenesis. Kidney Int 2007; 72: 599–607. Segerer S, Hughes E, Hudkins KL et al. Expression of the fractalkine receptor (CX3CR1) in human kidney diseases. Kidney Int 2002; 62: 488–495. Hoffmann U, Bergler T, Segerer S et al. Impact of chemokine receptor CX3CR1 in human renal allograft rejection. Transpl Immunol 2010; 23: 204–208. Cockwell P, Calderwood JW, Brooks CJ et al. Chemoattraction of T cells expressing CCR5, CXCR3 and CX3CR1 by proximal tubular epithelial cell chemokines. Nephrol Dial Transplant 2002; 17: 734–744. Miyazono K, Hellman U, Wernstedt C et al. Latent high molecular weight complex of transforming growth factor beta 1. Purification from human platelets and structural characterization. J Biol Chem 1988; 263: 6407–6415. Bedford JJ, Leader JP, Walker RJ. Aquaporin expression in normal human kidney and in renal disease. J Am Soc Nephrol 2003; 14: 2581–2587. Noronha IL, Kruger C, Andrassy K et al. In situ production of TNF-alpha, IL-1 beta and IL-2R in ANCA-positive glomerulonephritis. Kidney Int 1993; 43: 682–692. Noronha IL, Eberlein-Gonska M, Hartley B et al. In situ expression of tumor necrosis factor-alpha, interferon-gamma, and interleukin-2 receptors in renal allograft biopsies. Transplantation 1992; 54: 1017–1024. Scholz J, Lukacs-Kornek V, Engel DR et al. Renal dendritic cells stimulate IL-10 production and attenuate nephrotoxic nephritis. J Am Soc Nephrol 2008; 19: 527–537. Castellano G, Trouw LA, Fiore N et al. Infiltrating dendritic cells contribute to local synthesis of C1q in murine and human lupus nephritis. Mol Immunol 2010; 47: 2129–2137. Tucci M, Ciavarella S, Strippoli S et al. Oversecretion of cytokines and chemokines in lupus nephritis is regulated by intraparenchymal dendritic cells: a review. Ann NY Acad Sci 2009; 1173: 449–457. Iwata Y, Furuichi K, Sakai N et al. Dendritic cells contribute to autoimmune kidney injury in MRL-Faslpr mice. J Rheumatol 2009; 36: 306–314. Hochheiser K, Engel DR, Hammerich L et al. Kidney dendritic cells become pathogenic during crescentic glomerulonephritis with proteinuria. J Am Soc Nephrol 2011; 22: 306–316. Heymann F, Meyer-Schwesinger C, Hamilton-Williams EE et al. Kidney dendritic cell activation is required for progression of renal disease in a mouse model of glomerular injury. J Clin Invest 2009; 119: 1286–1297. De Palma G, Castellano G, Del Prete A et al. The possible role of ChemR23/Chemerin axis in the recruitment of dendritic cells in lupus nephritis. Kidney Int 2011; 79: 1228–1235. Yamamoto T, Noble NA, Miller DE et al. Sustained expression of TGF-beta 1 underlies development of progressive kidney fibrosis. Kidney Int 1994; 45: 916–927. Zhu Y, Usui HK, Sharma K. Regulation of transforming growth factor beta in diabetic nephropathy: implications for treatment. Semin Nephrol 2007; 27: 153–160. Kidney International (2015) 87, 1153–1163

basic research

AJ Kassianos et al.: Recruitment and retention of human renal DCs

38.

39.

40.

41.

42.

43.

Yu CI, Becker C, Wang Y et al. Human CD1c þ dendritic cells drive the differentiation of CD103 þ CD8 þ mucosal effector T cells via the cytokine TGF-beta. Immunity 2013; 38: 818–830. Soilleux EJ, Morris LS, Leslie G et al. Constitutive and induced expression of DC-SIGN on dendritic cell and macrophage subpopulations in situ and in vitro. J Leukoc Biol 2002; 71: 445–457. Masten BJ, Olson GK, Tarleton CA et al. Characterization of myeloid and plasmacytoid dendritic cells in human lung. J Immunol 2006; 177: 7784–7793. Schlitzer A, McGovern N, Teo P et al. IRF4 transcription factor-dependent CD11b þ dendritic cells in human and mouse control mucosal IL-17 cytokine responses. Immunity 2013; 38: 970–983. Haniffa M, Shin A, Bigley V et al. Human tissues contain CD141hi crosspresenting dendritic cells with functional homology to mouse CD103 þ nonlymphoid dendritic cells. Immunity 2012; 37: 60–73. Donadelli R, Zanchi C, Morigi M et al. Protein overload induces fractalkine upregulation in proximal tubular cells through nuclear factor kappaB-

Kidney International (2015) 87, 1153–1163

44. 45.

46.

47. 48.

49.

and p38 mitogen-activated protein kinase-dependent pathways. J Am Soc Nephrol 2003; 14: 2436–2446. Fraticelli P, Sironi M, Bianchi G et al. Fractalkine (CX3CL1) as an amplification circuit of polarized Th1 responses. J Clin Invest 2001; 107: 1173–1181. Ludwig A, Berkhout T, Moores K et al. Fractalkine is expressed by smooth muscle cells in response to IFN-gamma and TNF-alpha and is modulated by metalloproteinase activity. J Immunol 2002; 168: 604–612. Yoshida H, Imaizumi T, Fujimoto K et al. Synergistic stimulation, by tumor necrosis factor-alpha and interferon-gamma, of fractalkine expression in human astrocytes. Neurosci Lett 2001; 303: 132–136. Racusen LC, Solez K, Colvin RB et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713–723. Glynne PA, Evans TJ. Inflammatory cytokines induce apoptotic and necrotic cell shedding from human proximal tubular epithelial cell monolayers. Kidney Int 1999; 55: 2573–2597. Wilkinson R, Wang X, Roper KE et al. Activated human renal tubular cells inhibit autologous immune responses. Nephrol Dial Transplant 2011; 26: 1483–1492.

1163