Functionally Significant Renal Allograft ... - Wiley Online Library

C Blackwell Munksgaard 2005 Copyright

American Journal of Transplantation 2005; 5: 573–581 Blackwell Munksgaard

doi: 10.1111/j.1600-6143.2005.00719.x

Functionally Significant Renal Allograft Rejection Is Defined by Transcriptional Criteria Steven C. Hoffmanna , Douglas A. Halea,b , David E. Kleinerc , Roslyn B. Mannona , Robert L. Kampena , Lynn M. Jacobsond , Linda C. Cendalesa , S. John Swansona,b , Bryan N. Beckerd and Allan D. Kirka,b, ∗ a Transplantation Branch, NIDDK, NIH, DHHS, Bethesda, Maryland 20892 b Organ Transplant Service, Walter Reed Army Medical Center, Washington, DC c Laboratory of Pathology, NCI, NIH, DHHS, Bethesda, Maryland 20892 d Department of Medicine, University of Wisconsin, Madison, Wisconsin 53706 ∗ Corresponding author: Allan D. Kirk, M.D., Ph.D., [email protected]

Renal allograft acute cellular rejection (ACR) is a T-cell mediated disease that is diagnosed histologically. However, many normally functioning allografts have T-cell infiltrates and histological ACR, and many nonimmune processes cause allograft dysfunction. Thus, neither histological nor functional criteria are sufficient to establish a significant rejection, and the fundamental features of clinical rejection remain undefined. To differentiate allograft lymphocyte infiltration from clinically significant ACR, we compared renal biopsies from patients with ACR to patients with: sub-clinical rejection (SCR, stable function with histological rejection); no rejection; and nontransplanted kidneys. Biopsies were compared histologically and transcriptionally by RT-PCR for 72 relevant immune function genes. Neither the degree nor the composition of the infiltrate defined ACR. However, transcripts up-regulated during effector T H 1 T-cell activation, most significantly the transcription factor T-bet, the effector receptor Fas ligand and the costimulation molecule CD152 clearly (p = 0.001) distinguished the patient categories. Transcripts from other genes were equivalently elevated in SCR and ACR, indicating their association with infiltration, not dysfunction. Clinically significant ACR is not defined solely by the magnitude nor composition of the infiltrate, but rather by the transcriptional activity of the infiltrating cells. Quantitative analysis of selected gene transcripts may enhance the clinical assessment of allografts.

Introduction Clinicians continuously survey renal transplant recipients for acute cellular rejection (ACR), a T-cell mediated disease treated with chronic immunosuppression (1,2). Rejection is usually diagnosed histologically based on specific patterns of lymphocyte infiltration in biopsies prompted by renal dysfunction (3). However, histological rejection occurs in many patients without dysfunction, a condition classified as sub-clinical rejection (SCR) (3–6). Thus, lymphocyte infiltration seems necessary but insufficient for ACR, and it is unclear how lymphocytes mediating ACR differ from those merely residing in the allograft. Furthermore, many co-morbid conditions can accompany lymphocyte infiltration making even dysfunction an unreliable indicator of rejection (7). While serum creatinine remains the most commonly used test to gauge renal function, it is clear that significant damage to a kidney can occur without perceptible change in the serum creatinine. Consequently, the management of renal transplant recipients remains guided by imperfect diagnostic techniques. Several studies have demonstrated that ACR differs transcriptionally from noninfiltrated allografts (8–16). However, by not evaluating SCR, they have perhaps focused on cell infiltration rather than clinical rejection. Thus, no clinical assay accurately distinguishes immune-mediated ACR from other pathologies, and the use of immunosuppression remains largely empiric. This study was performed to determine whether ACR could be better defined by supplementing standard histological criteria for rejection with immunohistochemical and transcriptional profile analysis. We correlated renal function with routine histology, immunohistochemical analysis, and quantitative assessment of 72 relevant transcripts, studying normal nontransplanted kidneys and allografts with stable function and normal histology, SCR or ACR. We find that a transcriptional profile indicative of activated T H 1 type T cells, but not the magnitude or composition of the infiltrate, uniquely defines a functionally significant allograft rejection.

Key words: Acute rejection, real-time PCR, renal allotransplantation, sub-clinical rejection

Materials and Methods

Received 2 September 2004; revised 4 October 2004; accepted for publication 13 October 2004

Patients and biopsy acquisition Patients (n = 52) were enrolled in Institutional Review Board-approved protocols following informed consent. Biopsies (n = 52) were obtained

573

Hoffmann et al. Table 1: Demographic and clinical data of patient biopsy groups1

Biopsies Male/female Cadaveric/living Recipient age∗ Donor age∗ HLA match∗ Biopsy day∗∗ SCr @ biopsy∗∗ % Change SCr@ bx∗∗ SCr 1 year post bx∗∗ Subsequent ACR within 1 year Graft loss within 1 year

Group 1 NK

Group 2 SA

Group 3 SCR

Group 4 ACR

15 7/8 0/15 — 42 ± 10 — — 1.0 (0.7–1.2) — 1.1 (0.9–2.1) — —

10 6/4 3/7 39 ± 10 35 ± 11 3 ± 2.3 360 (31–1106) 1.3 (0.7–1.6) −7 (−18–6.3) 1.3 (1.1–2.4) 0 0

10 4/6 3/7 42 ± 13 37 ± 16 4 ± 1.3 263 (42–268) 1.4 (0.9–1.7) 0 (−21–9.1) 1.3 (0.8–2.5) 1 0

17 8/9 11/6 43 ± 4 40 ± 5 4 ± 0.5 173 (26–1211) 3.2 (1.4–12.4) 34 (19–65) 1.8 (0.9–4.9)∗∗∗ 6 episodes in 5 patients 2

1 SCr–Serum ∗ Mean

creatinine in mg/dL. ± standard deviation; ∗∗ Median (range); ∗∗∗ Excludes patients with graft loss.

Table 2: Immunohistochemical staining scoring Scoring is based on evaluation of cortex only Overall pattern of infiltrate judged on LCA stain: Diffuse—inflammation spread evenly through out cortex Focal—asymmetric involvement of cortex Scoring is based on evaluation of largest collection/most severe involvement, except for scores of 1 and 2, which average involvement is judged 0—No positive cells seen 1—Average of 2 or fewer positive cells per 20× field (approx 1 mm core length) 2—More than 2 positive cells per 20× field, no clustering of cells 3—Contiguous cluster(s) of positive cells, 3 tubules in size but less than width of the biopsy (1 mm) 6—Confluence of positive cells, involving the full width of the biopsy or >1 mm of biopsy length

from donors or recipients treated with standard immunosuppression: a calcineurin inhibitor (tacrolimus, target trough 10–12 ng/mL; or cyclosporine target trough 150–200 ng/mL), mycophenolate mofetil (2 g/day) or sirolimus (target trough 8–15 ng/mL), and/or prednisone (0–30 mg/day). Biopsies (16-gauge) were obtained based on protocol surveillance criteria or as clinically indicated, and were assigned to one of four groups based on clinical presentation and the Banff 1997 criteria (3). The groups did not differ significantly with regard to patient demographics, time of biopsy (Table 1) or maintenance immunosuppression (data not shown). Group 1 biopsies, normal kidney (NK, n = 15), were obtained prior to vascular cross-clamp from live donors with normal renal function undergoing open donor nephrectomy. Group 2 biopsies, stable allograft (SA, n = 10), were obtained from patients undergoing surveillance biopsy at least 1 month post-transplant without change in renal function (serum creatinine ≤10% increase over baseline without pathological proteinuria), drug toxicity, microbial infection or histological acute or chronic rejection. Group 3 biopsies, SCR (n = 10), were obtained from patients meeting the clinical criteria for Group 2 who had an acute rejection score of I-A or higher on biopsy. Group 4 biopsies, ACR (n = 17), were obtained from patients biopsied specifically to evaluate a rise in serum creatinine of at least 15% who also had a rejection score of IA or higher without other demonstrable pathology. There was no significant difference in rejection score between Groups 3 and 4. Patients with antibody mediated rejection or BK virus nephropathy were excluded from the study and all biopsies were negative for SV40 viral infection. One patient in Group 3 had diffuse C4d staining without alloantibody. No patients with borderline changes for acute rejection were included in this study. All biopsies were taken prior to any treatment for rejection.

574

Rejection on biopsy was treated in all cases. The results of protocol biopsies were considered by the Institutional Review Board to be clinically relevant and were specifically acted upon as stipulated in the protocol. Sub-clinical rejection (Group 3) was treated either by increasing the targeted levels of their maintenance immunosuppression, adding a brief steroid taper, or more typically both. Thus, this study does not address the natural history of SCR. Clinical acute rejection (Group 4) was treated initially with bolus methylprednisolone. Four patients failed to respond and were treated with rabbit anti-thymocyte globulin (n = 3) or OKT3 (N = 1). Follow-up creatinine and graft survival was determined 1 year following biopsy on all patients (Table 1).

Biopsy preparation Biopsies were divided at the bedside with cortex sent for RNA procurement and histology according to the Banff criteria (3). Additional sections were immunostained as previously described (14) with the following antibodies: LCA (Dako, Carpenteria, CA), CD3 (Dako), CD4 (Novocastra, Burlingame, CA), CD8 (Dako), CD68 (Dako), perforin (Kamiya, Seattle, WA) and granzyme B (Monosan, Burlingame, CA). Polyomavirus infection was excluded using the anti-SV40 large T-antigen antibody PAB-416 (Oncogene Research Products), and C4d staining was performed using the antibody Bi-RC4D (Bionet Inc, Southbridge, MA). A transplant pathologist (DK) evaluated all biopsies in a blinded fashion. The degree of infiltration for each cell phenotype was semi-quantitatively scored from 0 to 6 based on immunohistochemical staining in the renal cortex (Table 2). Tissue for complete immunophenotyping was available for all Group 1 biopsies and for 10 cases each for Groups 2, 3 and 4.

American Journal of Transplantation 2005; 5: 573–581

Transcriptional Profiling of Renal Allograft Rejection Table 3: Gene products and accession numbers for transcripts evaluated by real-time quantitative PCR1

Results

ACE AT1R AT1Rb ATR2 Bax Bcl-2 Bcl-xl C3 CD3 CD25 CD26 CD28 CD34 CD40 CD54 CD62E CD71 CD80

Clinical follow-up All patients were followed for at least 1 year from the time of biopsy (Table 1). There were no deaths during this period. Group 1 patients had stable renal function with a modestly increased baseline creatinine. There were no episodes of clinical rejection within the year following biopsy in Group 2 and there was only a single episode in Group 3. Consistent with this course, there was no graft loss in either Groups 2 or 3. Neither group had significantly worse function after 1 year of follow-up (p > 0.5 for each group compared to itself after 1 year of follow-up), and the groups were not significantly different from one another 1 year following biopsy (p = 0.637). Thus, while this study does not address the natural history or significance of SCR, in this population where SCR was detected and acted upon, there was no progressive worsening of function.

CD86 CD152 CD154 Collagen IV c-SKI ECE-1 Endothelin-1 ET-1 Rec Fas FasL Fibronectin G-CSF GM-CSF Granulysin Granzyme B HLA DR ICOS IFN-c

IjB2 IL-10 IL-12p35 IL-12p40 IL-13 IL-15 IL-17 IL-18 IL-1a IL-1b IL-2 IL-3 IL-4 IL-5 IL-6 IL-7 IL-8 iNOS

LTb MCP-1 m-CSF MIP-1a NFjB PD-1 Perforin RANTES Renin Smad-3 Smad-7 T-bet TGFb TIMP-2 TIMP-3 TNFa TNFb VEGF

1 The

entries in italics indicate those transcripts whose levels were statistically significantly increased (p < 0.05) from the levels present in normal, nontransplanted kidney in at least one of the post-transplant biopsy groups and in which the mean absolute magnitude of the deviation exceeded 5-fold.

Quantification of transcripts by real-time polymerase chain reaction (RT-PCR) Real-time PCR allows for precise relative quantitation of gene transcripts (17,18) directly from renal allograft biopsies (14,19). Samples for RNA were snap frozen in liquid nitrogen and processed for cDNA as previously described (14). cDNA (100 ng) from each biopsy was used for RT-PCR as previously described (19). Gene targets were chosen based on potential relevance to allograft biology (Table 3). Each target was analyzed in quadruplicate. In addition, primers for 18 s ribosomal RNA (an internal control for template input that we have shown does not change based on the clinical conditions under evaluation) were analyzed for each reaction. Reactions were amplified: 50◦ C for 2 min, 99◦ C for 10 min, 35 cycles at 99◦ C for 15 s and 60◦ C for 1 min. Individual samples were compared to a pooled sample of cDNA from Group 1 prepared by combining equal amounts of cDNA from each individual within Group 1 to establish a homogenous reference for normal kidney. Data were analyzed using Sequence Detector version 1.7.1 software (Applied Biosystems). Quantification was derived using the comparative threshold cycle method as previously described (14,19) and reported as an n-fold difference of the experimental sample to the pool of biopsies from Group 1. No transcript associations segregated by patient demographics or immunosuppressants.

Statistics Significance of differences between immunohistochemical staining scores was determined using the Mann–Whitney U test. Gene transcript data were normalized by log transformation, and compared between biopsy Groups 1 through 4 with one-way analysis of variance (ANOVA). Post-hoc inter-group comparisons were made using a Bonferroni correction to appropriately account for multiple comparisons. To focus on the most dynamic differences, we considered only those transcripts that differed from normal kidney by at least 5-fold in at least one group for statistical analysis (Table 3). Significance was defined as a two-sided p < 0.05.


In contrast to patients in Groups 2 and 3, patients from Group 4 had significantly worse outcome. Five patients experienced six subsequent episodes of clinical acute rejection within the first year and two patients returned to dialysis. At 1 year Group 4 patients with remaining graft function (excluding dialysis patients) had significantly worse serum creatinine than both Groups 2 (p = 0.010) and 3 (p = 0.018).

Neither the magnitude nor the composition of the inflammatory infiltrate predicts the functional status of renal allografts nor distinguishes sub-clinical from clinical rejection All Group 1 biopsies were histologically normal (data not shown). Biopsies from Group 2 contained modest focal interstitial infiltrates of CD3+ T-cells and scattered macrophages without tubulitis, vasculitis or glomerulitis. Consistent with our previous studies (14), Group 2 biopsies had increased numbers of LCA+ , CD3+ , CD4+ , CD8+ , CD68+ , perforin+ and granzyme B+ cells (p = 0.009, 0.02, 0.0002, 0.03, 0.008 and 0.04 respectively, Figure 1, Table 4) compared to normal kidney. Thus, a mild inflammatory infiltrate distinguished stable allografts without rejection from normal kidney. Groups 3 and 4 showed a cellular infiltrate that greatly exceed that in Groups 1 and 2 with prominent tubulitis and interstitial inflammation consisting of macrophages and T cells. Despite the different clinical presentations (Group 3 with normal function and Group 4 with deteriorating renal function), neither their Banff nor immunohistochemical scores differed, including the degree of infiltration or tubulitis, or the presence of CD3+ , CD4+ , CD8+ , CD68+ , perforin+ or granzyme B+ cells (Figure 1, Table 4). Furthermore, there were no differences in the regional distribution of the infiltrates that distinguished between Groups 3 575

Hoffmann et al. Table 4: Immunohistochemical and Banff scoring of patient biopsies by groups1 Group 2 SA Biopsies 10 Banff scoring Tubulitis (t) 0.0 ± 0.0 Inflammation (i) 0.2 ± 0.4 Immunoperoxidase staining 2.6 ± 0.8 LCA+ CD3+ 2.7 ± 0.9 CD4+ 2.8 ± 0.8 CD8+ 2.4 ± 0.5 CD68+ 2.0 ± 0.9 Perforin+ 1.3 ± 0.9 1.4 ± 0.9 Granzyme B+

Group 3 SCR

Group 4 ACR

10

17

2.3 ± 0.5∗ 1.7 ± 0.5∗

2.0 ± 0.7∗ 1.9 ± 0.8∗

5.3 ± 0.5∗ 5.1 ± 0.3∗ 5.1 ± 0.6∗ 3.5 ± 0.5∗ 3.5 ± 0.5∗ 2.2 ± 0.9∗ 2.5 ± 0.9∗

4.8 ± 0.8∗ 4.6 ± 0.8∗ 4.8 ± 0.8∗ 3.7 ± 0.5∗ 2.9 ± 0.6 1.6 ± 0.5 2.2 ± 0.8

are presented as mean values ± standard deviations. (p < 0.05) compared to Group 2. There were no significant differences between Groups 3 and 4.

1 Results

not significantly elevated. The down-regulatory inducible costimulation molecule CD152 was generally not detected in Group 3. Significant elevations of chemotactic molecules MIP-1a (p = 0.002) and m-CSF (p = 0.004) were also observed, as were transcript levels for complement (C3, p = 0.001). Several transcripts were increased in both Groups 2 and 3 in a progressive fashion that only reached significance when Group 3 was considered against Group 1. These included (p-values shown for Group 3 vs. Group 1) the T-cell transcripts Fas ligand (FasL, p = 0.001), CD28 (p = 0.001), granulysin (p = 0.003), granzyme B (p = 0.001), and TGFb (p = 0.001). Thus, despite equivalent clinical function, allografts with SCR were distinguished histologically and transcriptionally from allografts without SCR.

∗ Significance

and 4. These two groups were not distinguishable by histological or immunohistochemical criteria. Therefore, in this cross-sectional analysis of patients including individuals whose biopsies were not obtained for dysfunction, histological assessment of the infiltrate was not predictive of clinical dysfunction. Biopsies from patients with normal function had histological and immunohistochemical evidence of ACR indistinguishable from that detected in patients with clinical rejection. Transcriptional analysis distinguishes sub-clinically rejecting allografts from normal allografts As expected, both Groups 1 and 2 biopsies were transcriptionally quiescent. Group 2 biopsies occasionally had insignificantly increased levels of T-cell or macrophage transcripts, including CD3, CD25, CD28, RANTES and granulysin consistent with their mild infiltrate detected histologically (Figure 2). There was a trend for consistently upregulated CD154 (p = 0.078), although the mean absolute increase in fold expression (6.5) was modest. Despite the stable function observed in Group 3 patients, the transcriptional pattern of their biopsies was markedly different compared to Groups 1 and 2, and was indicative of considerable inflammation (Figure 2). Group 3 biopsies showed significant increases (p-values vs. Group 2) in transcripts for the inflammatory cytokine TNF-a (p = 0.001) and consistent increases in the T-cell transcripts CD3 (p = 0.001) and CD25 (p = 0.001). There was a bias for a T H 1 phenotype shown by increases in IFN-c (p = 0.001), T-bet (p = 0.001) and RANTES (p = 0.034). Perforin transcripts were present but not significantly elevated. Inducible costimulatory transcripts were prominently upregulated including CD80 (p = 0.001), CD154 (p = 0.001) and ICOS (p = 0.024), although constitutively expressed costimulatory molecules CD40 and CD86 transcripts were 576

Transcriptional analysis distinguishes sub-clinically rejecting allografts from clinically rejecting allografts The elevated transcript levels in biopsies from patients with clinical allograft dysfunction (Group 4) were qualitatively similar but quantitatively distinguished from Group 3 in the substantial elevation of three transcripts. Two T H 1 effector T-cell transcripts were detected in log-fold higher amounts (Figure 2). Specifically, FasL (p = 0.001) and T-bet (p = 0.001) were significantly elevated compared to Group 3, and the down-regulatory costimulation molecule CD152 was present in all biopsies from Group 4 and significantly elevated (p = 0.001) over all other groups. Other prominent T-cell transcript levels including CD3, CD25 and RANTES were numerically higher in Group 4 continuing a trend for sequential increases established for Groups 1, 2 and 3 respectively. This trend distinguished Group 4 from Group 2 but failed to reach statistical significance when compared to Group 3 (Figure 2). Despite the considerable clinical difference between Groups 3 and 4, transcripts associated with other aspects of the inflammatory response were equivalent including the cytokines interferon-c and IL-10 (Figure 2). Despite the unique elevation of CD152 in Group 4, the other costimulatory transcripts were equivalently elevated in Groups 3 and 4, with no significant or suggestive differences between Groups 3 and 4 for CD28, CD40, CD80, CD86, CD154, ICOS or PD-1. Thus, when cDNA was derived from biopsies associated with a clinically significant event, T H 1-associated T-cell transcripts were increased, particularly T-bet and FasL, above all other allograft states, and transcripts suggesting downregulatory costimulation (CD152) were present. The remainder of the transcriptional milieu remained equivalent to that measured in patients without renal dysfunction. The transcriptional activation state of the infiltrate, not the number or type of cells present, was uniquely indicative of a functionally significant rejection. American Journal of Transplantation 2005; 5: 573–581

Transcriptional Profiling of Renal Allograft Rejection

Figure 1: Infiltrating cell phenotype and histological analysis of human renal allograft biopsies. Representative histologic and immunophenotypic changes in sub-clinical and clinical rejection. Each column shows a representative case from Groups 2, 3 and 4. The biopsies showing sub-clinical (Group 3, Banff grade IB) and clinical (Group 4, Banff grade IA) rejection demonstrate focally intense lymphohistiocytic infiltrates associated with tubulitis (H&E, 200×). The Group 2 graft shows no significant interstitial inflammation and no tubulitis. Immunostaining for T cells (CD3, 400×) and macrophages (KP-1, 400×) shows that the majority of the infiltrating cells in both rejection cases are T cells, although small clusters of macrophages are present, particularly at the edges of inflammation. The Group 2 graft shows only scattered inflammatory cells. Immunostains for perforin (600×) and granzyme B (600×) show scattered cells containing cytotoxic granules in the Groups 3 and 4 cases. Perforin and granzyme B positive cells are also present in the Group 2 biopsy, but are rare.


577

Hoffmann et al. 1000

100

10

1

0.1

0.01

T-bet*

FasL*

CD3

CD25

IFN-γγ

Rantes

Granzyme B

Perforin

1000

100

10

1

0.1

0.01

CD28

CD40

CD80

CD86

CD152*

CD154

ICOS

PD-1

Figure 2: Transcript levels distinguishing the status of human renal allografts. The relative transcript prevalence for each patient biopsy sample is shown for those transcripts with significant deviation from normal kidney. Individual, log-scale, quantitative fold-levels of RNA transcripts are displayed for Group 2 (SA, open circles), Group 3 (SCR, gray circles) and Group 4 (ACR, black circles) biopsies referenced to the pool of Group 1 (NK, reference line value of 1) biopsies. Circles located below the graph baseline indicate undetectable levels. Bars represent the mean fold-level of RNA transcripts for each biopsy group. Targets are separated into functional categories: (A) factors typically associated with T-cell activation and effector function (T-bet, FasL, CD3, CD25, Interferon-c , RANTES, granzymeB and perforin); (B) costimulatory molecules (CD28, CD40, CD80, CD86, CD152, CD154, ICOS and PD-1); (C) other cytokines and factors typically associated with general inflammation or macrophage presence (TNF-a, TNF-b, TGF-b, IL-10 and granulysin), or elements mediating chemotaxis (C3, Mip1-a and m-CSF). All of the T-cell associated transcripts in Figure 2A distinguished ACR from SA biopsies. An asterisk (∗ ) denotes those transcripts that significantly distinguish SCR from ACR. See the text for other levels of significance.

Discussion We have shown that neither the magnitude nor composition of a renal allograft cellular infiltrate is independently predictive of clinically significant cellular rejection. Instead, transcripts indicative of T H 1 pathway maturation (delayed578

type hypersensitivity and cellular immunity as opposed to humoral immunity) are most closely associated with the event’s clinical import. This suggests that T-cell chemotaxis is necessary but insufficient for rejection providing that T-cell effector maturation is relatively suppressed. Accordingly, adequate immunosuppression might be more American Journal of Transplantation 2005; 5: 573–581

Transcriptional Profiling of Renal Allograft Rejection 1000

100

10

1

0.1

0.01

α TNF-α

β TNF-β

β TGF-β

IL-10

Granulysin

C3

α MIP-1α

m-CSF*

Figure 2: Continued.

accurately defined by transcriptional indicators of effector T-cell maturation, than by evidence of cellular infiltration, chemotaxis, antigen presentation or costimulation. Prior studies evaluating allograft histology have been predicated on biopsies obtained to evaluate worsening function. With a bias toward dysfunction, histological findings have been similarly biased toward association with poor function. Indeed, a rejecting allograft does have an infiltrate that greatly exceeds that in a normal kidney. However, this does not establish a causal link between the histological findings and dysfunction. The current study underscores the incomplete reliability of histology alone and suggests that several factors intimately associated with effector T-cell function may be more precise indicators of rejection. Furthermore, given the myriad factors potentially altering renal function, it seems likely that an elevated creatinine might be due to nonimmune causes even in the presence of histological rejection. We have initiated study in states such as calcineurin inhibitor toxicity. In the absence of an infiltrate, toxicity is void of the factors that we have associated with SCR and ACR (unpublished observations). We suggest that in the presence of an infiltrate without an ACR-like transcriptional milieu, other causes of dysfunction should not be discounted. These data support the need for a trial prospectively investigating the use of transcriptional analysis in tailoring a patient’s immunosuppressive burden, and assays of this type may provide a more biologically relevant metric for immunosuppressive management. There has been polarizing debate between those who view SCR as a benign condition, and those recommending it be viewed as ACR and treated irrespective of its immediate functional consequence (7,20–22). Furthermore, it has been unclear whether SCR represents a less severe form of ACR, perhaps a product of sampling error, or if it is pathogenically a different, or even beneficial, process. Our data are most consistent with the notion that differences American Journal of Transplantation 2005; 5: 573–581

between SCR and ACR are more quantitative than qualitative and that they are fundamentally part of the same process. Clinical stability does not indicate immune quiescence so much as it speaks to the efficacy of modern drugs to prevent T-cell maturation. The markers most predictive of allograft rejection in this study are those closely associated with activated T H 1 Tcells, specifically T-bet, a prototypical transcription factor unique to cytotoxic T-cell maturation (23,24), and FasL, a molecule mediating T-cell mediated target cell apoptosis. This is the first report of T-bet as an indicator of allograft rejection, and the use of this parameter for diagnostic purposes should be investigated. The cytotoxic molecules granzyme B, perforin and FasL, seen in our study, are consistent with markers detecting clinically apparent rejection by less invasive techniques (25–28), as were markers known to be related to T-cell function including CD3 and CD25. However, prior studies have not specifically considered the issue of SCR (16). That these effector molecules are also present in SCR suggests that rejection is not predicated on their presence, but rather on some more central activation signal precipitating their employment. They may be effective mediators of rejection without initiating the process. We believe that, among other things, the transcription factor T-bet is an excellent candidate gene for this central signal, and the data would support this. Indeed, T-bet and FasL distinguish all allograft states with the most statistical power in this study. However, it is important to underscore that no transcript should be viewed as exclusively indicative of rejection. Rather, rejection progresses along a spectrum that involves many activation genes. It remains to be shown where the line should be drawn with regard to clinical relevance of an immune response towards the allograft. It is well established that costimulatory signals are necessary for naive T-cell activation and are important for 579

Hoffmann et al.

rejection (29–31). However, our study suggests that their detection is more indicative of cell presence than cell function. Indeed, CD154 appears to be the single most sensitive indicator of cell infiltration as indicated by its presence even in Group 2 biopsies. This suggests that T-cell activation and CD154 up-regulation precedes chemotaxis, and that costimulation is proximal in the events leading to rejection, perhaps explaining why costimulation blockade has effectively prevented rejection in na¨ıve animal, but not in sensitized animals (32–34). It is also interesting to note that the costimulatory molecule most typically related to T-cell anergy, CD152 (35–37), was most prominent in ACR kidneys, suggesting that regulatory factors may indeed be induced during a functionally significant rejection to limit the destructive process. Many other transcripts previously associated with ACR were clearly represented in our patients including the monocyte/macrophage-derived chemotactic transcripts TNF-a, TNF-b, MIP-1a, m-CSF and RANTES. However, these markers did not predict dysfunction. Chemotaxis leads to cell infiltration, but is not sufficient to induce renal dysfunction. Another interesting observation is the increased presence of transcripts for complement in SCR and ACR biopsies. Complement has long been recognized as an opsonin and chemoattractant following injury and has recently been suggested to be required for initiating an alloimmune response (38). As an initial validating experiment, this study has evaluated patients with unambiguous histological diagnoses. In the future it will be important to apply this technique to ambiguous diagnostic conditions and indeed, this may be an important use of this type of analysis. However, prospective use of transcriptional data to help guide clinical care can only be attempted after the technique has been defined in the context of established diagnostic techniques. We have shown herein that allograft rejection cannot be equated with cell presence, and that clinical stability is not synonymous with immune inactivity. We also find no indication that SCR infiltrates are uniquely regulatory or beneficial, but rather they are the product of appropriate reduction in T-cell activation. Novel technology such as transcriptional profiling provides a method to detect subtle changes in allograft status, and may provide a more precise clinical tool for guiding the use of an increasingly complex immunosuppressive arsenal, preempting complications rather than responding to them.

Acknowledgments This work was supported by the Division of Intramural Research of the National Institute of Diabetes, Digestive and Kidney Diseases, and in part (BNB) by National Institutes of Health grants RO1 AI49285-04 and 1K24 DK616962-02. The authors gratefully acknowledge the nurses and transplant coordinators of the Organ and Tissue Transplant Research Center of the Warren G. Magnuson Clinical Center for their excellent patient care

580

skills; Ms. Terri Wakefield and Ms. Rebecca Muehrer for their research administrative skills, Ms. Beverly Niles for her meticulous data management skills, Ms. Christine Chamberlain for her clinical pharmacology support, and Applied Biosystems for supplying the RT-PCR reagents and for their generous technical support.

References 1. Suthanthiran M, Strom TB. Renal transplantation. N Engl J Med 1994; 331: 365–376. 2. Pascual M, Theruvath T, Kawai T, Tolkoff-Rubin N, Cosimi AB. Strategies to improve long-term outcomes after renal transplantation. N Engl J Med 2002; 346: 580–590. 3. Racusen LC, Solez K, Colvin RB et al. The Banff 97 working classification of renal allograft pathology. Kidney Int 1999; 55: 713– 723. 4. Rush DN, Henry SF, Jeffery JR, Schroeder TJ, Gough J. Histological findings in early routine biopsies of stable renal allograft recipients. Transplantation 1994; 57: 208–211. 5. Colvin RB, Cohen AH, Saiontz C et al. Evaluation of pathologic criteria for acute renal allograft rejection: reproducibility, sensitivity, and clinical correlation. J Am Soc Nephrol 1997; 8: 1930–1941. 6. Legendre C, Thervet E, Skhiri H et al. Histologic features of chronic allograft nephropathy revealed by protocol biopsies in kidney transplant recipients. Transplantation 1998; 65: 1506–1509. 7. Halloran PF. Call for revolution: a new approach to describing allograft deterioration. Am J Transplant 2002; 2: 195–200. 8. Kirk AD, Bollinger RR, Finn OJ. Rapid, comprehensive analysis of human cytokine mRNA and its application to the study of acute renal allograft rejection. Hum Immunol 1995; 43: 113–128. 9. Strehlau J, Pavlakis M, Lipman M et al. Quantitative detection of immune activation transcripts as a diagnostic tool in kidney transplantation. Proc Natl Acad Sci 1997; 94: 695–700. 10. Suthanthiran M. Molecular analyses of human renal allografts: differential intragraft gene expression during rejection. Kidney Int Suppl 1997; 58: S15–S21. 11. Lipman ML, Shen Y, Jeffery JR et al. Immune-activation gene expression in clinically stable renal allograft biopsies: molecular evidence for subclinical rejection. Transplantation 1998; 66: 1673– 1681. 12. Kirk AD, Jacobson LM, Heisey DM, Radke NF, Pirsch JD, Sollinger HW. Clinically stable human renal allografts contain histological and RNA-based findings that correlate with deteriorating graft function. Transplantation 1999; 68: 1578–1582. 13. Strehlau J, Pavlakis M, Lipman M, Maslinski W, Shapiro M, Strom TB. The intragraft gene activation of markers reflecting T-cellactivation and -cytotoxicity analyzed by quantitative RT-PCR in renal transplantation. Clin Nephrol 1996; 46: 30–33. 14. Hoffmann SC, Kampen RL, Amur S et al. Molecular and immunohistochemical characterization of the onset and resolution of human renal allograft ischemia-reperfusion injury. Transplantation 2002; 74: 916–923. 15. Sarwal M, Chua MS, Kambham N et al. Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. N Engl J Med 2003; 349: 125–138. 16. Flechner SM, Kurian SM, Head SR et al. Kidney transplant rejection and tissue injury by gene profiling of biopsies and peripheral blood lymphocytes. Am J Transplant 2004; 4: 1475–1489. 17. Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Res 1996; 6: 995–1001. 18. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res 1996; 6: 986–994.


Transcriptional Profiling of Renal Allograft Rejection 19. Hoffmann SC, Pearl JP, Blair PJ, Kirk AD. Immune profiling: molecular monitoring in renal transplantation. Front Biosci 2003; 8: E444–E462. 20. Rush D. Protocol biopsies should be part of the routine management of kidney transplant recipients. Am J Kidney Dis 2002; 40: 671–673. 21. Shapiro R, Randhawa P, Jordan ML et al. An analysis of early renal transplant protocol biopsies—the high incidence of subclinical tubulitis. Am J Transplant 2001; 1: 47–50. 22. Gough J, Rush D, Jeffery J et al. Reproducibility of the Banff schema in reporting protocol biopsies of stable renal allografts. Nephrol Dial Transplant 2002; 17: 1081–1084. 23. Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 2000; 100: 655–669. 24. Mullen AC, High FA, Hutchins AS et al. Role of T-bet in commitment of TH1 cells before IL-12-dependent selection. Science 2001; 292: 1907–1910. 25. Li B, Hartono C, Ding R et al. Noninvasive diagnosis of renalallograft rejection by measurement of messenger RNA for perforin and granzyme B in urine. N Engl J Med 2001; 344: 947–954. 26. Strom TB, Suthanthiran M. Prospects and applicability of molecular diagnosis of allograft rejection. Semin Nephrol 2000; 20: 103– 107. 27. Dugre FJ, Gaudreau S, Belles-Isles M, Houde I, Roy R. Cytokine and cytotoxic molecule gene expression determined in peripheral blood mononuclear cells in the diagnosis of acute renal rejection. Transplantation 2000; 70: 1074–1080. 28. Pascoe MD, Marshall SE, Welsh KI, Fulton LM, Hughes DA. Increased accuracy of renal allograft rejection diagnosis using com-


29. 30.

31.

32.

33. 34.

35. 36.

37.

38.

bined perforin, granzyme B, and Fas ligand fine-needle aspiration immunocytology. Transplantation 2000; 69: 2547–2553. Yamada A, Sayegh MH. The CD154-CD40 costimulatory pathway in transplantation. Transplantation 2002; 73 (1 Suppl): S36–S39. Sayegh MH, Turka LA. The role of T-cell costimulatory activation pathways in transplant rejection. N Engl J Med 1998; 338: 1813– 1821. Harlan DM, Kirk AD. The future of organ and tissue transplantation: can T-cell costimulatory pathway modifiers revolutionize the prevention of graft rejection? JAMA 1999; 282: 1076–1082. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 1999; 401: 708–712. Welsh RM, Selin LK. No one is naive: the significance of heterologous T-cell immunity. Nat Rev Immunol 2002; 2: 417–426. Adams AB, Williams MA, Jones TR. Heterologous immunity provides a potent barrier to transplantation tolerance. J Clin Invest 2003; 111: 1887–1895. June CH, Vandenberghe P, Thompson CB. The CD28 and CTLA-4 receptor family. Chem Immunol 1994; 59: 62–90. Riley JL, Mao M, Kobayashi S et al. Modulation of TCRinduced transcriptional profiles by ligation of CD28, ICOS, and CTLA-4 receptors. Proc Natl Acad Sci 2002; 99: 11790– 11795. Carter LL, Carreno BM. Cytotoxic T-lymphocyte antigen-4 and programmed death-1 function as negative regulators of lymphocyte activation. Immunol Res 2003; 28: 49–59. Pratt JR, Basheer SA, Sacks SH. Local synthesis of complement component C3 regulates acute renal transplant rejection. Nat Med 2002; 8: 582–587.

581