SEARCH

SEARCH BY CITATION

Keywords:

  • Antibody-mediated rejection;
  • donor-specific antibody;
  • HLA antibody;
  • microarray;
  • NK cell;
  • renal allograft pathology

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

To explore the mechanisms of antibody-mediated rejection (ABMR) in kidney transplants, we studied the transcripts expressed in clinically indicated biopsies from patients with donor-specific antibody (DSA). Comparison of biopsies from DSA-positive versus DSA-negative patients revealed 132 differentially expressed transcripts: all were associated with class II DSA but none with class I DSA. Many transcripts were expressed in DSA-positive ABMR but were also expressed in T-cell-mediated rejection (TCMR), reflecting shared molecular features. Removal of shared transcripts created 23 DSA selective transcripts (DSASTs). Some DSASTs (6/23) showed selective high expression in NK cells, whereas others (8/23) were expressed in endothelium or in endothelium plus other cell types (7/23). Of 145 biopsies ranked by DSAST expression, the 25 with highest DSAST expression primarily consisted of ABMR (22/25, 88%), either C4d-positive or C4d-negative. By immunostaining, CD56+ and CD68+ cells in peritubular capillaries, but not CD3+ cells, were increased in ABMR compared to TCMR, compatible with a role for NK cells, as well as macrophages, as effectors in endothelial injury during ABMR. Thus, the strategy of using DSASTs in the biopsy to identify mechanism-related transcripts in biopsies from patients with clinical phenotypes indicates the selective involvement of NK cells in ABMR.


Abbreviations: 
ABMR

Antibody-mediated rejection

TCMR

T-cell-mediated rejection

PRA

Panel reactive antibodies

DSA

Donor-specific antibodies

NK cell

Natural killer cell

DSASTs

DSA selective transcripts

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Antibody-mediated rejection (ABMR) was first defined as a syndrome of graft dysfunction, biopsy findings of microcirculation damage (glomerulitis, capillaritis, microthrombi) and circulating donor-specific antibody (DSA) (1,2) and was later found to correlate with C4d deposition in peritubular capillaries (3). This led to the realization that ABMR is responsible for many cases of late kidney transplant rejection, and that persistent ABMR leads to transplant glomerulopathy (4–7). The Banff diagnostic criteria for ABMR require C4d positivity (8), although we recently reported that ABMR is often C4d-negative, indicating that C4d staining is insensitive (6,9,10). Furthermore, C4d staining is not completely specific, as illustrated by C4d staining in stable blood group-incompatible transplants (11). The DSA status of a patient at the time of a biopsy for clinical indications (BFC) predicts the presence of microcirculation lesions of ABMR in the biopsy, but many of these cases are C4d-negative (12).

ABMR, recently highlighted in a virtual journal issue (http://www.amjtrans.com/view/0/virtual.html#ABMR), is challenging to study because it can be fulminant or subtle, early or late. The pathogenesis of ABMR is the result of damage incurred onto microvascular endothelium by DSA, and the effector mechanisms involved in the early cases may differ from those acting in the subtle late cases. DSA to class I HLA is often a feature of the early cases whereas DSA to class II HLA is predominant in the late cases, although the explanation for this puzzling finding remains elusive (13). However, the association of anti-class II DSA with many cases of late ABMR and graft loss gives this problem particular significance (4,12,14). The pathogenesis of the endothelial damage in ABMR may involve direct actions of DSA (15) or complement-mediated damage, but the presence of mononuclear cells and neutrophils in the glomerular or peritubular capillaries strongly suggests that cells could mediate the endothelial injury through their Fc receptors and/or complement receptors (9). This opens the possibility of antibody-dependent cell-mediated cytotoxicity (ADCC), which can be mediated by NK cells and to some extent monocytes and neutrophils.

Because the histopathology of ABMR is often subtle and C4d is insensitive, the molecules associated with DSA may be useful in defining new diagnostic criteria and providing clues to the effector mechanisms. The molecular phenotype of ABMR includes inflammation (16), interferon-γ (IFNG) effects (17) and increased expression of certain endothelial transcripts (9). Since DSA, particularly against class II, is associated with microcirculation lesions in late kidney biopsies for cause (12), we now hypothesized that the DSA status could allow us to define the transcripts selectively expressed in ABMR in the same population. This population does not represent early aggressive ABMR in cross-match-positive patients because the practice of transplanting across positive cross-matches is uncommon in our center. Since we had already shown that endothelial transcripts were altered in biopsies with ABMR (9), we were interested in determining whether an unsupervised analysis would generate transcripts for the other cell types selectively associated with DSA and ABMR lesions.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Patients and sample collection

The patient population in this study and details of the HLA antibody results have been recently described (12). The study was approved by the institutional review board of the University of Alberta (issue N0. 5299). Written informed consent was obtained from all study patients. All consenting renal transplant patients undergoing a transplant biopsy for clinical indication (deterioration in function, proteinuria, stable impaired function) as standard of care between September 2004 and March 2007 were included. Biopsies were obtained under ultrasound guidance by spring-loaded needles (ASAP Automatic Biopsy, Microvasive, Watertown, MA).

Only one biopsy—the last biopsy available—for each patient (N = 145) was included in the analyses.

Histopathology

Paraffin sections were graded according to Banff criteria (18) by a renal pathologist (BS). C4d staining was performed on frozen sections using a monoclonal anti-C4d antibody (Quidel, San Diego, CA) by indirect immunofluorescence. Diffuse linear C4d staining (>50% of biopsy area) was interpreted as positive. The criteria for C4d-negative ABMR or C4d-negative Mixed ABMR plus T-cell-mediated rejection (TCMR) were based on our previous description (10): microvascular lesions of inflammation (g, or ptc >0) or microvascular deterioration (cg >0) from a patient with detectable HLA antibody (panel reactive antibodies [PRA] positive) at the time of biopsy.

Immunohistochemistry

The number of CD3+ (T cells), CD68+ (macrophages) and CD56+ (NK cells/a minor subset of CD8+ T cells) was assessed by immunoperoxidase staining in 16–18 biopsies (depending on the marker stained) with available tissue. Antigen retrieval of formalin-fixed and paraffin-embedded tissue sections was performed with an EDTA-based buffer on a Ventana Benchmark XT. Tissue sections were incubated with anti-human CD68 (1:50, clone PG-M1, DAKO, Carpenteria, CA) or CD56 (1:25, clone 1B6, Novocastra, Australia) or rabbit polyclonal CD3 (1:75, DAKO, Carpenteria, CA) primary antibody. Absolute numbers of intraluminal-positive cells in cross-sections of five peritubular capillaries with the highest number of infiltrates were counted.

HLA antibody screening

Antibody specificities were determined by FlowPRA-specific class I and or II and/or FlowPRA single antigen I and II beads (One Lambda Canoga Park, CA). Manufacturer's instructions for staining and acquiring were followed. Beads were analyzed on a BD FACSCalibur cytometer (Becton Dickinson Biosciences, Mississauga, Ontario, Canada).

Antibody screening was performed using FlowPRA beads. These beads have HLA-A, -B, -Cw -DR -DQ and -DP antigens represented. Further testing for specificities was performed following a positive screen (≥5% PRA or clear pattern of reactivity). Single antigen beads were used to test for antibodies against HLA-A, -B, -DRB1, -DRB3, -4 and -5, -DQB1 and -DP. We did not test for specificities to Cw. Donor typing for DP was not performed and therefore DSA were not attributed to DP. De novo DSA was defined as a new DSA detected by single antigen bead technology and/or a donor-specific flow cross-match that was negative pretransplant and positive at the time of biopsy. Flow T- and B-cell cross-matches were performed as previously described (19).

HLA typing

Low-to-medium resolution HLA Class I and II typing was performed using the One Lambda Micro SSP assay as previously shown (19). Manufacturer's instructions for amplification and electrophoresis were followed.

Microarray analysis

One additional 18-gauge biopsy core was collected for gene expression analysis. The tissue was placed immediately in RNALater and stored at −20°C. RNA extraction, labeling and hybridization to the HG_U133_Plus_2.0 GeneChip (Affymetrix, Santa Clara, CA) were carried out according to manufacturer's protocols. Microarrays were scanned using GeneArrayScanner (Affymetrix) and processed with GeneChip Operating Software Version 1.4.0 (Affymetrix). Microarray data were preprocessed by robust multiarray analysis, and implemented in Bioconductor version 2.4.

Microarray results from DSA-positive and DSA-negative classes were compared. For each probeset, the mean of the gene expression values in the two classes was compared using the Genespring GX 7.3.1(Agilent Technologies, Santa Clara, CA). False discovery rates (FDR) were calculated using the Benjamini and Hochberg test for multiple testing corrections. Transcript set scores were calculated using DSA-associated transcripts (N = 132) or DSA selective transcripts (DSASTs) (N = 23) as the geometric mean of fold changes across all probesets within each transcript set. Results for the fold increases for the individual DSASTs in each of the biopsies in the study are included in Supplementary Table S1.

Biopsies were categorized according to the patient's DSA results:

  • 1
    HLA class I DSA-positive—biopsies from patients who were HLA class I DSA-positive or HLA class I + II DSA-positive at the time of biopsy (N = 23).
  • 2
    HLA class I DSA-negative—biopsies from patients who were HLA class II DSA-positive or NDSA- or PRA-negative at the time of biopsy (N = 122).
  • 3
    HLA class II DSA-positive—biopsies from patients who were HLA class II DSA-positive or HLA class I + II DSA-positive at the time of biopsy (N = 41).
  • 4
    HLA class II DSA-negative—biopsies from patients who were HLA class I DSA-positive or NDSA- or PRA-negative at the time of biopsy (N = 104).

Rejection classifier method and statistics were performed as described previously (17).

Pathogenesis-based transcript sets (PBTs)

Transcript sets defining distinct biological processes involved in allograft rejection are detailed on our website (http://transplants.med.ualberta.ca/). Abbreviations for all transcripts conform to the Gene names (http://www.ncbi.nlm.nih.gov/sites/entrez).

Cell isolations and treatments

Purified cell populations from peripheral blood mononuclear cells isolated from whole blood of healthy volunteers as previously described (20). All leukocyte cell cultures were maintained in complete RPMI (RPMI 1640 supplemented with 10% FBS, 2 mM L-glutamine, β-mercaptoethanol, nonessential amino acids, sodium pyruvate and antibiotic/antimycotic (Invitrogen Life Technologies, Burlington, Ontario, Canada) in 5% CO2 at 37°C. Human umbilical vein endothelial cells (HUVECs) (ATCC, Manassas, VA) and renal proximal tubule epithelial cells (RPTECs) (Lonza, Inc., Allendale, NJ) were maintained in tissue culture media as recommended by the supplier.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Transcripts in the biopsy associated with circulating DSA

We studied the same cohort of 145 patients undergoing a renal transplant BFC that we previously studied for histopathology lesions associated with DSA where demographics are shown in detail (12). These biopsies included all consenting patients presenting for a BFC for whom HLA antibody results were available, and were not selected for diagnoses.

Transcripts associated with DSA were defined by comparing transcript expression in biopsies from DSA-positive versus DSA-negative patients. We found 290 transcripts more strongly expressed in biopsies from DSA-positive versus DSA-negative patients (Figure 1A) at an FDR <0.05 and 132 transcripts at FDR <0.005. To simplify biological annotation of transcripts, subsequent analysis used the 132 transcripts selected at FDR <0.005.

image

Figure 1. Algorithms used to define transcripts associated with DSA. (A) Class comparison between biopsies from DSA-positive patients to biopsies from DSA-negative (DSA neg) patients yielded the DSA-associated transcripts. (B) Class comparisons between biopsies from HLA class I DSA-positive (DSA I positive) patients to biopsies from DSA I-negative (DSA I neg) patients (left panel) and comparing biopsies from HLA class II DSA-positive (DSA II positive) patients to those from DSA II-negative (DSA II neg) patients. (C) Venn diagram showing the overlap in the number of transcripts selected by each of the algorithms in parts (A) and (B).

Download figure to PowerPoint

We analyzed the relative contribution of HLA class I versus class II specificities to the DSA-associated transcripts. Biopsies from 41 anti-class II DSA-positive patients were compared to 104 biopsies from anti-class II DSA-negative patients and 272 transcripts were differentially expressed at an FDR <0.05 (Figure 1B). Most class II DSA-associated transcripts (198/272) were also in the total DSA-associated transcript set (Figure 1C). In contrast, comparison of biopsies from 23 anti-class I DSA-positive patients versus 122 biopsies from anti-class I DSA-negative patients yielded no transcripts (FDR <0.05).

Biological annotation of DSA-associated transcripts

We examined the biological processes represented by the 132 DSA-associated transcripts based on their annotation as members of PBTs (http://transplants.med.ualberta.ca/) or reports in the literature. Many transcripts (50/132) reflected IFNG effects, including 32 HLA transcripts (Figure 2A). Also associated with DSA were 11 macrophage transcripts; 9 transcripts selectively associated with NK cells, and 10 transcripts with high expression in NK cells as well as CD8-positive CTL. However, typical T-cell transcripts were not found (e.g. CD2, CD3D, ITK). Fourteen transcripts were associated with endothelial cells while 17 showed high expression across multiple inflammatory cells including CTL, NK cells, B cells and macrophages. Immunoglobulin transcripts (18/132) were also identified, probably reflecting plasma cells, but no previously defined B-cell selective transcripts (21) were identified. Three transcripts had not been previously annotated.

image

Figure 2. DSA-associated transcripts and the effect of time posttransplant. (A) Biological annotation of 132 transcripts associated with DSA generated in a class comparison across all 145 biopsies. Biological annotation was based on previous annotation in pathogenesis-based transcript lists, expression in a human primary cell panel or gene name. (B) Biological annotation of 40 transcripts associated with DSA generated in a class comparison across 89 late (>1 year posttransplant) biopsies. Biological annotation was based on previous annotation in pathogenesis-based transcript lists, expression in a human primary cell panel or gene name. (C) Transcript set scores were generated for each individual biopsy using 132 DSA-associated transcripts. Biopsies were arranged according to their revised diagnosis and biopsies from DSA-positive patients are indicated by blue dots, biopsies from NDSA patients as yellow dots and those from PRA-negative patients as gray dots. Normal nephrectomy samples were used as controls and are depicted as white dots.

Download figure to PowerPoint

Because DSA is more frequent in late biopsies, some DSA-associated transcripts may actually be features of late biopsies rather than truly DSA associated. To distinguish the true DSA association from associations with time, we defined transcripts associated with DSA across late biopsies only (>1 year posttransplant, N = 89). This yielded 40 transcripts (FDR <0.01), 30 of which were shared with the original DSA-associated transcripts (Figure 2B). Although most transcripts reflected the same biological processes as the original DSA-associated transcripts, immunoglobulin transcripts were eliminated. Thus immunoglobulin transcripts are associated with late biopsies (21) but are not truly associated with DSA.

We compared the mean expression of the 132 DSA-associated transcripts across all biopsies to the histological diagnoses and HLA antibody status (Figure 2C): Borderline (N = 20), C4d-positive ABMR (N = 13), C4d-negative ABMR (N = 20), TCMR (N = 16), C4d-positive Mixed (N = 3), C4d-negative Mixed (N = 5), polyoma virus nephropathy (N = 3), glomerulonephritis (GN) (N = 23) and Other (N = 43), compared to control kidneys (Figure 2C). DSA-associated transcripts were more highly expressed in biopsies with ABMR, including C4d-positive and C4d-negative ABMR and Mixed, but also highly expressed in many biopsies with TCMR. This result was expected: given the intense inflammatory disturbance shared by ABMR and TCMR (16), some DSA-associated transcripts would be shared with TCMR, particularly IFNG-induced transcripts.

Defining DSA-selective transcripts

To identify which DSA-associated transcripts were selective for ABMR, we reasoned that these should be selectively expressed in rejecting cases with DSA, and not in rejecting cases lacking DSA. We defined biopsies with molecular features of rejection (ABMR or TCMR) using the previously described classifier (17), which indicated that 78/145 biopsies had rejection. We then identified transcripts preferentially expressed in DSA-positive compared to DSA-negative rejection. This identified 23 of the 132 transcripts as DSASTs (FDR <0.05)—(Figure 3A).

image

Figure 3. Algorithm for the generation of DSAST and expression of DSAST in primary human cells. (A) DSASTs were generated by starting with the 132 DSA-associated transcripts and determining how many remained differentially expressed between DSA-positive and DSA-negative groups when compared across the 78 rejection-classified biopsies. (B) DSASTs with preferential expression in human NK cells. (C) DSASTs with preferential expression in HUVECs. (D) DSASTs with high expression in HUVECs or inducible in HUVECs and also expressed in other cell types.

Download figure to PowerPoint

DSASTs are primarily expressed either in NK cells or endothelium

We studied the expression of the 23 DSASTs (for 21 unique genes) in primary human cell types; either purified or cultured (Figures 3B–D; Table 1): allostimulated CD4 and CD8 CTL, NK cells, B cells, macrophages, HUVECs and RPTECs. We also included macrophages, HUVECs and RPTECs treated with human IFN-γ. DSASTs were preferentially expressed in two cell types: endothelium and NK cells. In contrast, no DSASTs were selectively expressed in CTL, B cells, macrophages or RPTECs.

Table 1.  Detailed annotation of the transcripts making up the DSASTs (N = 23)
Affymetrix probesetGene titleGene symbolAliasGenbankBiological association
208335_s_atDuffy blood groupDARCFYNM_002036Endothelium
213620_s_atintercellular adhesion molecule 2ICAM2 AA126728Endothelium
226028_atroundabout homolog 4, magic roundabout (Drosophila)ROB04 AA156022Endothelium
204677_atcadherin 5, type 2, VE-cadherin (vascular epithelium)CDH5VE-cadherinNM_001795Endothelium
204115_atguanine nucleotide binding protein (G protein), gamma 11GNG11 NM_004126Endothelium
209373_atT-cell differentiation protein-likeMALLBENEBC003179Endothelium
211343_s_atcollagen, type XIII, alpha 1COL13A1 M33653Endothelium
201860_s_atplasminogen activator, tissuePLATtPANM_000930Endothelium
204726_atcadherin 13, H-cadherin (heart)CDH13H-cadherinNM_001257Endothelium
206702_atTEK tyrosine kinase, endothelial (venous malformations, multiple cutaneous and mucosal)TEK NM_000459Endothelium
228698_atSRY (sex determining region Y)-box 7SOX7 AI808807Endothelium
219584_atphospholipase A1 member APLA1APSPLA1NM_015900IFN-g response
221491_x_atmajor histocompatibility complex, class II, DR beta 3HLA-DRB3HLA-DR52AA807056MHC
223836_atfibroblast growth factor binding protein 2FGFBP2KSP37AB021123NK cells
1553177_atSH2 domain-containing molecule EAT2SH2D1BEAT2BC022407NK cells
213906_atv-myb myeloblastosis viral oncogene homolog (avian)-like 1MYBL1MYBL1AW592266NK cells
205898_atchemokine (C-X3-C motif) receptor 1CX3CR1Fractalkine receptorU20350NK cells
220646_s_atkiller cell lectin-like receptor subfamily F, member 1KLRF1NKp80NM_016523NK cells
205495_s_atgranulysinGNLY NM_006433NK cells/CTL
226303_atphosphoglucomutase 5PGM5 AA706788Multiple
228489_attransmembrane 4 L six family member 19TM4SF18 AI721164Multiple

Seven transcripts (six unique genes) were selective for NK cells (Figure 3B): fractalkine receptor (CX3CR1), myeloblastosis viral oncogene homolog (MYBL1), fibroblast growth factor binding protein 2 (FGFBP2, also known as KSP37), killer cell lectin-like receptor F1 (KLRF1, also known as NKp80) and SH2 domain containing 1B (SH2D1B, also known as EAT2). Transcripts encoding the cytotoxic molecule granulysin (GNLY) showed high expression in NK cells but were also expressed in CD4- and CD8-positive CTL as described previously (20). Thus, these DSASTs are selective for NK cells, although some are shared with T cells, as expected from the close relationships between these cell types (22). Nevertheless, most T-cell-specific transcripts (e.g. CD2, CD3D, etc.) were not increased in biopsies from DSA-positive patients, although they are prominent in TCMR {26865}. Thus, the expression of NK-selective DSASTs reflects the presence of NK cells, and cannot be explained by the presence of T cells.

Eight DSASTs were primarily expressed in endothelium. Some were previously described as increased in ABMR, including cadherin 13 (CDH13), cadherin 5 (CDH5) and MALL (mal, T-cell differentiation protein-like, BENE) (9). CDH5, MALL, ROBO4 and SOX7 showed the most selective expression in HUVECs (Figure 3C). Intercellular adhesion molecule 2 (ICAM2) showed high expression in HUVECs but was also expressed in some leukocytes, particularly NK cells (Figure 3D).

Most of the remaining eight DSASTs (seven unique genes) showed high expression in HUVECs, and were expressed in other cell types. Guanine nucleotide binding protein (GNG11), tissue plasminogen activator, and transmembrane 4L six family member 18 (TM4SF18) all showed high expression in HUVECs and were also expressed in RPTECs (Figure 3D). Neither Duffy blood group (DARC) nor phosglucomutase 5 (PGM5) showed high expression in HUVECs, but DARC is expressed in postcapillary venule endothelium (23). Phospholipase A1 member A (PLA1A) was IFNG inducible in HUVECs and RPTECs, while HLA-DRB3 showed high expression in macrophages and B cells as expected.

DSASTs are selectively increased in biopsies with ABMR

We compared DSAST expression to the histopathology diagnosis of each biopsy that had previously been established, independent of molecular features. Of the 25 biopsies with the highest DSAST scores (Table 2), 22 (88%) were diagnosed by histology as ABMR, either C4d-positive or C4d-negative. Of interest, all three biopsies with high DSAST score that did not meet current criteria for histologic ABMR were nevertheless from de novo DSA-positive patients, raising the possibility that they had ABMR features that are not currently recognized. Thus every one of the 25 biopsies with the highest DSAST scores was either previously diagnosed as ABMR or was from a DSA-positive patient. Nine of 13 Banff C4d-positive ABMR biopsies were in the top 25 DSAST scores.

Table 2.  Biopsy code, revised diagnosis and DSA status of the 25 biopsies with highest DSAST scores
Biopsy codeRevised diagnosisDSA statusDSAST score
  1. DSA status: de novo DSA = de novo; preexisting DSA = pre-ex; non-DSA = NDSA.

T29C4d neg Mixedde novo4.01
A10C4d pos ABMRde novo3.54
M3C4d pos Mixedpre-ex3.12
O36C4d neg ABMRNDSA3.10
A17C4d pos ABMRde novo2.81
A12C4d pos ABMRde novo2.73
A14C4d pos ABMRde novo2.68
O10C4d neg ABMRpre-ex2.56
O32C4d neg ABMRde novo2.40
A6C4d pos ABMRde novo2.34
A4C4d pos ABMRde novo2.30
B23C4d neg ABMRde novo2.26
A15C4d pos ABMRde novo2.26
M4C4d pos Mixedpre-ex2.22
A9C4d pos ABMRde novo2.20
A8C4d pos ABMRde novo2.16
T44C4d neg Mixedde novo2.08
O100GNde novo2.07
B12C4d neg ABMRpre-ex2.05
O103C4d neg ABMRde novo2.05
O80C4d neg ABMRNDSA1.90
O115GNde novo1.86
O22C4d neg ABMRde novo1.86
O13C4d neg ABMRpre-ex1.85
T30TCMRde novo1.83

DSAST scores were examined across all biopsies with regard to histological diagnosis (Figure 4). Scores were highest for biopsies diagnosed as ABMR or Mixed, with C4d-positive ABMR biopsies carrying overall higher scores than C4d-negative ABMR. One biopsy diagnosed as C4d-positive ABMR and another diagnosed as C4d-positive Mixed showed low DSAST scores in each category; these biopsies had received steroid bolus treatment within 2 weeks prior to the biopsy. Biopsies diagnosed as Borderline, Other or polyoma virus nephropathy showed overall low DSAST scores. The DSASTs were selective for ABMR: most TCMR biopsies show a low DSAST score compared to ABMR or Mixed biopsies. Biopsies with GN show DSAST scores similar to TCMR but those with the highest scores were primarily from DSA-positive patients, suggesting that subtle forms of ABMR may be missed in some biopsies.

image

Figure 4. DSAST scores plotted according to HLA antibody status. DSAST set scores were generated for each individual biopsy and biopsies were arranged according to their revised diagnosis. Biopsies were arranged according to their revised diagnosis and biopsies from DSA-positive patients are indicated by blue dots, biopsies from NDSA patients as yellow dots and those from PRA-negative patients as gray dots. Normal nephrectomy samples were used as controls and are depicted as white dots.

Download figure to PowerPoint

NK cells and macrophages are selectively increased in peritubular capillaries in ABMR

We performed cell counting in sections stained for CD3, CD68 and CD56 in biopsies where tissue was available (six C4d-positive ABMR, six C4d-negative ABMR, six TCMR). Many biopsies for cause did not have adequate tissue remaining for these immunostaining studies. The average number of CD56+ NK cells (p = 0.006) and CD68+ macrophages (p = 0.03) in peritubular capillaries was higher in C4d-positive or C4d-negative ABMR biopsies versus those with TCMR (Figure 5). In contrast, CD3+ T cells in peritubular capillary were not elevated in ABMR (p = 0.09). Representative figures of intracapillary CD3+, CD68+ or CD56+ cells are shown (Figures 5 C–E). Therefore, NK cells, as well as macrophages, are found in areas of antibody-mediated microcirculation injury.

image

Figure 5. Intraluminal cell types in peritubular capillaries by immunohistochemistry. (A) A representative of peritubular capillaritis in a kidney transplant biopsy (periodic acid-Schiff, original magnification × 600). (B) Comparison of mean numbers of intraluminal CD56+, CD68+ or CD3+ cells in five peritubular capillaries in biopsies with antibody-mediated rejection (ABMR) versus T-cell-mediated rejection (TCMR). (C–E) Representative figures for (C) intracapillary CD56+ NK cells, (D) CD68+ macrophages or (E) CD3+ T cells (immunoperoxidase, original magnification × 600).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

We identified the transcript associations in BFCs from DSA-positive patients at the time of biopsy. Comparison of biopsies from DSA-positive versus negative patients revealed 132 transcripts more strongly expressed in BFC from DSA-positive patients (FDR <0.005) primarily reflecting IFNG effects, NK cells and endothelium. Thus, DSA in the recipient at the time of a BFC was associated with transcripts reflecting a strong inflammatory state in the biopsy, some features of which were shared with TCMR. After removing transcripts expressed in rejecting biopsies from DSA-negative patients, 23 transcripts were selectively expressed in rejecting biopsies from DSA-positive patients—the DSASTs. DSASTs were not only expressed by endothelium, as expected from previous analyses (9), but also by NK cells. The selectivity of DSASTs for ABMR is evidenced by the highest 25 DSAST scores being from BFC from HLA antibody-positive patients, and 88% had already been diagnosed as ABMR or mixed based on histologic criteria. Immunostaining showed CD56+ and CD68+ cells selectively increased in biopsies with ABMR. The results associate NK cells with ABMR but not with TCMR, and are compatible with a role for NK cells in the mechanism of microcirculation injury in ABMR.

The expression of some DSA-associated transcripts in TCMR as well as ABMR does not imply that the mechanisms of rejection in TCMR and ABMR are similar, or that ABMR is always associated with TCMR. These two forms of rejection share inflammation features including strong IFNG effects in the tissue (16,17). The association of NK cells with ABMR in our study raises the possibility that IFNG effects observed in ABMR reflect IFNG production by NK cells triggered by Fc receptors in ABMR (25), potentially contributing to pathogenesis, for example induction of increased class II expression.

The DSAST algorithm did not find B-cell transcripts or immunoglobulin transcripts, which are associated with time posttransplant but are features of injury and scarring rather than associations with DSA or ABMR (12,21). We did not confirm previous reports of association of B-cell and plasma-cell features with steroid-resistant rejection and graft loss (28), once time of the biopsy for cause was factored in. It is like that the high-affinity antibody required for ABMR originates from plasma cells in the bone marrow or spleen, not inflamed sites (29). The association of immunoglobulin transcripts with time posttransplant (21) and scarring (26) is independent of the underlying diseases, reflecting plasma cell homing to damaged tissues rather than an active role in ABMR.

While association of DSA with endothelial transcripts is expected (9,30), the association with NK transcripts with biopsies from patients with DSA has novel implications that will guide more mechanistic studies in the future. The model proposed is that NK cells are triggered by DSA bound to the microcirculation endothelium, triggering IFNG production and ADCC when CD16 engages Fc portions of high-affinity IgG DSA bound to donor HLA (25). Although other cells including macrophages and granulocytes may also interact with endothelium coated with antibody (and complement), the NK cell probably has unique functions, due to its granule-associated cytotoxicity and IFNG release. This model must be further explored, despite the challenge of limiting material, and complexity of the reagents against NK markers, reflecting the sharing of markers between CD8 T cells and NK cells including CD56 and CD57 (20,31,32). Thus, the immunohistochemical staining for CD56 in the present studies must be interpreted in the context of CD3 in order to rule out CD8 T cells with NK markers. Interestingly, although a role for NK cells in allorecognition has been postulated (33), NK cells are rare in TCMR and mouse models with NK cells but no T cells never reject organ allografts (34).

The NK-cell transcripts among the DSASTs highlight candidate mechanisms of endothelial cell injury for further study. SH2D1B is a member of the signaling lymphocytic activation molecule (SLAM)-associated protein family of signaling molecules potentially controlling NK-cell activation (35,36). KLRF1 is a coreceptor for natural cytotoxicity receptors on NK cells (37), potentially cooperating with SH2D1B. It is thus conceivable that SLAM receptors participate in the interaction of NK cells with DSA-coated capillary endothelium. FGFBP2 is expressed in CD16-positive NK cells in vivo (38) and involved in lymphocyte cytotoxic functions (39). CX3CR1 could be involved in recruitment of NK cells to allograft endothelium when DSA is present (40). GNLY is of particular interest because GLNY, while expressed in CTLs, has very high expression in NK cells. These results thus point to molecules as possible mechanisms operating in ABMR and targets for intervention.

The potential use of DSASTs and other NK markers in diagnosis of ABMR faces the challenge that there is no true gold standard for active ABMR. High DSAST scores-selected biopsies independently diagnosed as ABMR without knowledge of their DSA or pathology status. Even the three biopsies with high DSAST scores that lacked histologic criteria for ABMR may be subtle ABMR: all were from patients with de novo DSA (12), suggesting that histologic changes may evolve later. The fact that NK cells are found selectively in peritubular capillaries in ABMR suggests that the NK markers could be developed as diagnostic aids. Some complexity should be anticipated: DSAST expression is unlikely to correspond to all histologic lesion-based ABMR diagnoses, because transcripts reflect disease activity, whereas lesions such as transplant glomerulopathy require sustained microcirculation injury over time and persist when ABMR is inactive. DSAST scores probably do correlate with activity, being higher in C4d-positive ABMR than C4d-negative ABMR and perhaps reflecting greater activity in C4d-positive cases (41). Thus the emergence of the DSASTs, and potentially other markers, offers an opportunity to reclassify ABMR on the basis of intensity and activity.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

The authors thank Dr. Zija Jacaj for help with collection of the clinical data; and Kara Allanach, Vido Ramassar and Anna Hutton for technical support.

Funding: This research has been supported by funding and/or resources from Genome Canada, Genome Alberta, the University of Alberta, the University of Alberta Hospital Foundation, Roche Molecular Systems, Hoffmann-La Roche Canada Ltd., the Alberta Ministry of Advanced Education and Technology, the Roche Organ Transplantation Research Foundation, the Kidney Foundation of Canada and Astellas Canada. Dr. Halloran also holds a Canada Research Chair in Transplant Immunology and the Muttart Chair in Clinical Immunology. The authors have no competing financial interests.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Table S1: Normalized values for each transcript in the DSAST list for 145 clinically indicated biopsies

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
AJT_3201_sm_TableS1.xls81KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.