Article Text

Download PDFPDF

Extended report
Urinary T cells in active lupus nephritis show an effector memory phenotype
  1. Sebastian Dolff1,2,
  2. Wayel H Abdulahad1,
  3. Marcory C R F van Dijk3,
  4. Pieter C Limburg1,
  5. Cees G M Kallenberg1,
  6. Marc Bijl1
  1. 1Department of Rheumatology and Clinical Immunology, University Medical Center Groningen, Groningen, The Netherlands
  2. 2Department of Nephrology, University Hospital Essen, University Duisburg-Essen, Essen, Germany
  3. 3Department of Pathology and Medical Biology, University Medical Center Groningen, Groningen, The Netherlands
  1. Correspondence to Dr Sebastian Dolff, Department of Rheumatology and Clinical Immunology, University Medical Center Groningen, Hanzeplein, PO Box 30.001, Groningen 9700RB, The Netherlands; sebastian.dolff{at}


Background Systemic lupus erythematosus (SLE) is accompanied by alterations in T cell homeostasis including an increased effector response. Migrated effector memory T cells (CD45RO+CCR7; TEM) appear to be involved in tissue injury. The objective of this study was to investigate the distribution and phenotype of effector memory T cells in the peripheral blood (PB), and their presence in renal biopsies and urine of patients with SLE. The hypothesis that these TEM cells migrate to the kidney during active disease was tested.

Methods A total of 43 patients with SLE and 20 healthy controls were enrolled. CD4+TEM cells and CD8+TEM cells were analysed in PB and urine using flow cytometric analysis. In 10 patients with active lupus nephritis a parallel analysis was performed on the presence of TEM cells in kidney biopsies.

Results The percentage of circulating CD8+TEM cells in patients with SLE was significantly decreased versus healthy controls (33.9±18.3% vs 42.9±11.0%, p=0.008). In patients with active renal involvement (n=12) this percentage was further decreased to 30.4±15.9%, p=0.01. Analysis of the urinary sediment in active renal disease showed increased numbers of CD4+T cells (134±71 cells/ml) and CD8+T cells (287±220 cells/ml), respectively, while in healthy controls and patients without active renal disease almost no T cells were present. In all, 73.6±8.3% of urinary CD4+T cells and 69.3±26.0% of urinary CD8+T cells expressed the TEM phenotype. CD8+ cells were also found in renal biopsies.

Conclusions The data presented are compatible with the hypothesis that CD8+ effector memory cells migrate from the PB to the kidney and appear in the urine during active renal disease in patients with SLE. These cells could serve as an additional marker of renal activity in patients with SLE.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Systemic lupus erythematosus (SLE) is an autoimmune disease characterised by multiple organ manifestations. Inflammation of the kidney, in particular, is associated with an unfavourable prognosis.1 Although the precise pathogenesis of lupus nephritis (LN) has not been fully elucidated, disturbances in T cell homeostasis seem to contribute to the inflammatory pathology of LN.2 3

During maturation of CD45RO+ memory T cells expression of the lymph node homing chemokine receptor CCR7 distinguishes central memory T cells (TCM; CD45RO+CCR7+) from effector memory T cells (TEM; CD45RO+CCR7). In several autoimmune diseases including SLE, disturbances have been described in the distribution of these T cell subsets in the peripheral blood (PB).4,,6 Remarkably, CD8+TEM have been thought to play a crucial role in T cell homeostasis due to their ability to produce cytokines and exert cytotoxic activity.7

Further evidence for the important role of T cells in SLE is given by the observation that mononuclear cells, predominantly T lymphocytes, are a frequent histological finding in proliferative forms (International Society of Nephrology/Renal Pathology Society (ISN/RPS) class III and IV) of LN. However, data regarding CD4+ and CD8+ cell counts and their ratio in renal biopsies from patients with lupus are conflicting.2 3 Especially, the presence of periglomerular infiltrating CD8+T cells has been shown to correlate with histological activity, clinical severity and bad prognosis of LN.2 3 8 Additionally, several studies demonstrated the presence of mononuclear cells in urine of patients with active IgA nephropathy, LN and Wegener's granulomatosis.9 10 Thus analysis of urinary cells reflecting renal inflammation in SLE could be a useful tool in monitoring the course of renal disease.

In line with this, we recently reported an increase of urinary TEM (CD4+ CD45RO+CCR7) cells in patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated vascultis with active renal disease.11 Thus, analysis of urinary TEM cells seems a promising tool for detecting renal flares. Although T cell infiltration in kidneys of patients with SLE has been reported, analysis of the state of activation of urinary CD4+ or CD8+T cells has not been performed so far even though T cells lacking CCR7 have a strong ability to migrate in vitro.12

We hypothesise that T cells with effector function migrate from PB into the kidneys of patients with SLE during active renal disease. This is reflected by the presence of these T cells in the urinary sediment. To test this hypothesis of T cell migration, we analysed the PB and urine for the presence of effector memory T cells lacking CCR7 by four-colour flow cytometry and evaluated in parallel-obtained renal biopsies of patients with active renal disease for the presence of T cells.

Patients and methods

Study population

A total of 43 patients with SLE with a mean age of 41±12 years fulfilling at least 4 of the American College of Rheumatology revised criteria for SLE and 20 sex-matched healthy controls (age 36±10 years) were enrolled in this study.13 Disease activity was assessed by SLE Disease Activity Index (SLEDAI). In all, 24 patients had inactive disease (SLEDAI score ≤4) and 19 patients had active SLE (defined as SLEDAI score >4). Mean disease activity for all patients was 6.3±5.6 (table 1). A total of 25 patients had a current or former renal biopsy consistent with LN while 18 had no history of renal involvement. Currently active LN (n=12) was defined by a proliferative (class III or IV) glomerulonephritis in a parallel obtained renal biopsy (n=11) or the presence of an active urinary sediment with glomerular erythrocyturia (n=1) (table 2). Seven patients with active LN presented with the first episode of LN.

Table 1

Baseline characteristics and medication of all patients included (n=43)

Table 2

Laboratory and histological data of 12 patients with active renal disease are shown

A total of 12 patients did not receive any immunomodulating medication at the time of analysis; 6 of them were newly diagnosed as having their condition. In all, 31 patients received immunomodulating medication (table 1).


EDTA blood and fresh urine samples were collected from patients and healthy controls. Urine samples from patients that were nitrite positive on a dipstick test or with proof of bacterial contamination in the sediment were excluded.

Percentages and absolute counts of CD4+ and CD8+T cells were assessed immediately after sampling by four-colour flow cytometry in blood and urine samples.

Paraffin-embedded sections of renal biopsy specimens obtained from 11 patients were included in the present study. Informed consent was obtained from the patients after approval by the local ethics committee. The study was conducted according to the ethical guidelines of our institution and the Declaration of Helsinki.


The following antibodies were used in flow cytometry: phycoerythrin (PE)-conjugated anti-CCR7 (clone 3D12), fluorescein (FITC)-conjugated anti-CD45RO (clone UCHL-1), peridin-chlorophyll (PerCP)-conjugated anti-CD4 (clone SK3), allophycocyanin (APC)-conjugated anti-CD3 (clone UCHT1), MultiTEST four-colour antibodies (CD3-FITC, CD8-PE, CD45-PerCP and CD4-APC) and isotype-matched control antibodies of irrelevant specificity. All were purchased from Becton–Dickinson (BD) (Amsterdam, The Netherlands).

Sample preparation and flow cytometry

Immediately after voiding, 100 ml of urine was diluted 1:1 with cold phosphate-buffered saline (PBS) and processed as described before.11 Briefly, isolated mononuclear cells were resuspended in wash buffer (1% bovine serum albumin in PBS) and mixed with appropriate concentrations of anti-CD45RO-FITC, anti-CCR7-PE, anti-CD4-PerCP and anti-CD3-APC for 15 min at room temperature in the dark. In parallel, blood samples were labelled with the aforementioned monoclonal antibodies. Afterwards, cells were successively treated with 2 ml diluted fluorescence-activated cell sorting (FACS) lysing solution (BD) for 10 min and samples were washed twice in wash buffer and immediately analysed by flow cytometry. Four-colour staining was analysed on FACS-Calibur (BD) and data were collected for 105 events for each sample and plotted using Win-List software package (Verity Software House, Topsham, Maine, USA). Positively and negatively stained populations were calculated by quadrant dot plot analysis, as determined by the isotype controls. Representative examples are shown in figure 1.

Figure 1

Decreased percentages of circulating CD8+ effector memory (TEM) cells in patients with active systemic lupus erythematosus (SLE). Representative dot plots of circulating naïve (Tnaïve), central memory (TCM) or effector memory (TEM) cells in the peripheral blood of healthy controls and patients with inactive and active SLE are illustrated.

Quantification of effector memory T cells

T cells were quantified in urine using TruCOUNT tubes (BD). In brief, 20 µl of MultiTEST four-colour antibodies (CD3-FITC, CD8-PE, CD45-PerCP and CD4-APC) and 50 µl of sample (urine or blood) were added to bead-containing TruCOUNT tubes. The cell suspension was processed and analysed as described elsewere.11 Afterwards, the absolute counts for TEM cells in 1 ml urine were calculated as described before.11

Analysis and scoring of renal biopsies

Biopsies taken at the time of analysis of blood and urine samples were processed. All biopsies were reviewed and classified by an experienced nephropathologist (MCRFvD) according to the revised criteria for LN. The activity index (AI) and chronicity index (CI) were calculated for each specimen with maximum scores of 24 for the AI and 12 for the CI.14 For this study, slides stained with methenamine silver (with H&E counterstaining), H&E and periodic acid-Schiff (PAS) were used.

The assessment was completed by determining the ISN/RPS 2003 classification and activity and chronicity indices for LN. For these aspects of the assessment, the definitions of the classification systems and the activity and chronicity indices were used.

Immunohistochemistry staining

All specimens were fixed in 10% neutral buffered formalin and paraffin embedded. Sections 5 μm thick were deparaffinised in xylene and rehydrated in a series of different concentrations of ethanol. EDTA buffer, pH 8.2, for heat-induced epitope retrieval was applied for 1 h, followed by neutralisation of endogenous peroxidase with 0.3% H2O2. Incubation with a monoclonal mouse anti-human CD8 (DAKO, Glostrup, Denmark) was performed. Next, sections were washed and incubated with a horseradish peroxidase-conjugated secondary antibody (EnVision; DAKO) for 30 min at room temperature. A diaminobenzidine (DAB) substrate was used for visualisation. Washing with PBS was performed after each incubation step. CD4 (Monosan, Uden, The Netherlands) was performed in the Benchmark Ultra (Ventana; Ventana Medical Systems SA, Illkirch, France) with citrate buffer heat-inducted antigen retrieval and detected with the ultraView Universal Alkaline Phosphatase Red Detection Kit (Ventana). Finally, the slides were counterstained with haematoxylin and mounted with Kaiser's glycerine gelatin (Merck, Darmstadt, Germany).

Cells were separately counted for the interstitium and glomeruli. Cells with positive staining for CD8 and CD4 were counted per high power field (40× magnification). The average value was calculated for each biopsy.

Statistical analysis

Results are presented as mean±SD and the non-parametric Mann–Whitney U test was used for comparison of values between groups. Correlation with disease activity was assessed using Spearman's rank correlation coefficient. Linear regression analysis was performed to assess associations between T cell subsets and patient ages. Two-tailed p values less than 0.05 were regarded as statistically significant.


Patients with SLE have decreased percentage of circulating CCR7–CD45RO+CD8+ effector memory T cells during active disease and even less during active renal disease

We determined the percentages of naïve (CCR7+CD45RO; Tnaïve), central memory (CCR7+CD45RO+; TCM) and effector memory (CCR7CD45RO+; TEM) subsets of CD4+ and CD8+T cells in PB of healthy controls and patients with SLE. No differences in the percentages of circulating Tnaïve, TCM or TEM CD4+T cells were found between healthy controls and patients with SLE. There was a significant difference in the percentages of TCM and TEM CD8+T cells between healthy controls and patients with SLE (TCM: 5.8±3.8% vs 10.9±10.3%, p=0.02; TEM: 42.9±11.0% vs 33.9±18.3%, p=0.008, figure 2A,B). TCM were increased in patients with SLE while TEM CD8+T cells were decreased as compared to healthy controls. There was no difference between the percentages of circulating naïve CD8+T cells in PB of healthy controls and patients with SLE. Within the patient group there was no difference in the percentages of circulating naïve CD8+T cells in PB between those with or without immunomodulating medication.

Figure 2

CD4+ and CD8+ naïve and memory T cell subsets in the peripheral blood of patients with systemic lupus erythematosus (SLE) (n=43) and healthy controls (n=20). The percentages of naïve (Tnaïve), central memory (TCM) or effector memory (TEM) T cells within the CD4+ (A) and CD8+T cell population (B) in 43 patients with SLE (SLE) and 20 healthy controls (HC) are illustrated. The open symbols within the patient group represent patients without immunomodulating medication (n=12), solid symbols represent patients with immunomodulating medication. A subanalysis was performed for patients with active (SLE active, n=19) and inactive SLE (SLE inactive, n=24) according to the SLE Disease Activity Index (SLEDAI) score (C/D). (E,F) Data of patient groups analysed according to renal disease activity. Patients with active renal lupus nephritis (SLE with aLN, n=12) were compared to patients without active lupus nephritis (SLE without aLN, n=31) and healthy controls. Bars represent mean percentages. p Values were calculated using the non-parametric Mann–Whitney U test.

In addition, subsets were compared between patients with SLE with active and inactive disease. No differences were present between the percentages of circulating naïve (Tnaïve), central memory (TCM) or effector memory (TEM) CD4+T cells of healthy controls as compared to patients with active and inactive SLE, respectively (figure 2C). Within the CD8+T cell populations, a significant increase of circulating TCM cells was observed in patients with SLE with active disease as compared to healthy controls (TCM: 14.4±13.3% vs 5.8±3.8%, p=0.009). The TEM population was significantly decreased in active disease as compared to healthy controls (TEM: 31.6±14.9% vs 42.9±11.0%, p=0.004, figure 2D).

Next, we assessed the percentages of CD4+ and CD8+T cell subsets in patients with SLE in relation to the presence of active renal disease. The percentage of circulating TCM cells in patients with SLE without active LN was significantly increased compared to healthy controls (TCM: 10.1±9.0% vs 5.8±3.8%, p=0.02) whereas circulating peripheral TEM cells were decreased in patients with inactive renal disease as compared to healthy controls (TEM: 35.2±19.2% vs 42.9±11.0%, p=0.03). They were even more decreased in patients with active renal disease (TEM: 30.4±15.9% vs 45.1±9.4%, p=0.01, figure 2F) but not statistically significant compared to inactive disease. There was no difference between patients with a first episode of LN and those with relapsing renal disease regarding circulating memory T cell subsets (data not shown).

Urinary T cells with effector memory phenotype are associated with active renal disease in SLE

In parallel to the analysis of PB we collected urine to quantify the absolute numbers of CD4+ and CD8+ cells (figure 3A). The absolute count of CD4+T cells was significantly increased in patients with SLE with active LN as compared to patients with SLE without active LN and healthy controls, respectively (134±71 cells/ml vs 15±30 cells/ml, p=0.0001 and vs 2±4 cells/ml, p=0.002). Furthermore, the absolute count of CD8+T cells was significantly increased in patients with SLE with active LN as compared to patients with SLE without active LN and healthy controls, respectively (287±220 cells/ml vs 22±28 cells/ml, p<0.0001 and vs 1±1 cells/ml, p=0.002). Increased CD4+T cell and CD8+T cell counts/ml in urine were both associated with active renal disease and correlated with disease activity as assessed by SLEDAI (CD4+: r=0.62, p<0.001; CD8+: r=0.68, p<0.001). There was no correlation between these cell counts and other renal parameters in the subgroup of patients with active renal disease, such as serum levels of creatinine or 24 h proteinuria. Also, histological scores as the AI or CI did no correlate with the absolute cell count (table 2).

Figure 3

Absolute numbers of urinary T cells. Absolute counts of urinary CD4+ and CD8+ of patients with systemic lupus erythematosus (SLE) with active lupus nephritis (LN) (n=11), patients with SLE without active LN (n=24) and healthy controls (n=5) are shown. Bars represent mean±SD. Significant differences are indicated (**p<0.005, ***p<0.001) (A). Urinary CD8+ cells (B and C) show an effector memory phenotype (D).

Additionally, urinary T cells were assessed for the expression of CD45RO and CCR7 in order to differentiate between naïve T cells, central memory T cells and effector memory T cells. The majority of urinary CD4+T cells as well as CD8+T cells were CD45RO+CCR7 effector memory cells (59±23% and 62±26%, respectively). As shown in figure 4D almost no Tnaïve and TCM were present in the urine of these patients with SLE.

Figure 4

Immunohistochemical staining of CD8+ cells in a renal biopsy of a patient with lupus nephritis. Periglomerular (A) and interstitial (B) T cell infiltrates are positive for CD8 (brown).

CD8+ cells in renal biopsies

To determine the presence and localisation of CD8+T cells and CD4+T cells in renal biopsies, immunohistochemistry staining with anti-CD8 and anti-CD4 was performed.

CD8+T cells were present in all renal biopsies investigated. The average amount of CD8+ cells was 6.8±6.2 cells high power field. This was significantly higher than the average amount of CD4+T cells (2.6±4.9 cells/high power field; p=0.02). In ∼90% of specimens CD8+T cells were distributed as peritubular infiltrates, in 50% of renal biopsies CD8+T cells were localised periglomerularly as well (figure 4). Intraglomerularly almost no CD8+T cells could be found. The amount of infiltrated CD8+T cells did not correlate with the number of urinary CD8+T cells.


This study demonstrates disturbed frequencies of CD8+TEM cells and TCM cells in the PB of patients with SLE with active disease, especially a decrease of CD8+TEM cells in active renal disease, consistent with increased numbers of urinary CD8+TEM cells during renal flares. Moreover, infiltrating CD8+T cells could be observed in renal biopsies of patients with active renal disease. These data support the hypothesis that effector memory T cells migrate during active renal disease from the PB to the kidney and appear in the urinary sediment.

Previous studies have also reported aberrant CD8+ memory T cell populations in the PB of patients with SLE.6 15 Sen et al15 described increased percentages of CD8+TCM cells producing Th2 cytokines (IL-4 and IL-5) in patients with active SLE as compared to patients with inactive disease and healthy controls. A decrease of CD8+TEM cells was not reported and a subanalysis of patients with SLE with renal involvement was not performed. In accordance with our results these alterations were restricted to the CD8+T cells. In contrast to the latter and our findings, a recent study by Fritsch et al6 reported a shift in the CD4+ memory T cell balance towards an increased CCR7CD4+ memory population in SLE independent of disease activity. However, in that study no information was given on the characteristics of disease activity and SLEDAI was relatively low (3.9±1.6). As we included 12 patients with active LN, differences in lymphocyte subsets between both studies might be explained by differences in disease activity and/or organ involvement.

Beside significantly reduced numbers of CCR7CD45RO+ TEM cells in the PB of patients with active SLE, in particular during active renal disease, which have not been reported so far, this is the first study investigating the effector phenotype of memory T cells with special regard to renal disease. There is increasing evidence from animal and human studies that T cells, in particular CD8+T cells, contribute to the pathogenesis of LN.3 16 Analysis of 26 renal biopsies of patients with SLE with the aim to identify renal infiltrating leucocytes showed a CD4/CD8 ratio of 0.71 due to a relative increase of the mean percentage of CD8+ cells in LN biopsies which was found to correlate with histological activity.3 This finding was confirmed in a recent study where predominantly infiltrating CD8+ cells were found in kidneys of patients with SLE. Moreover, the authors showed that a high number of periglomerular CD8+ cells were a characteristic feature of severe and active forms of LN.8

Based on the assumption that these histopathological changes are reflected by the appearance of effector cells in urine, analysis of urine has become an attractive goal over the last few years. Microscopic examination of urine shows a significantly increased amount of CD3+ cells in patients with SLE with active renal disease.10 A deeper insight in urinary T cells was provided by an elegant and cohesive study of CD4+T cells in the urine and in renal biopsies of patients with active LN.17 The authors concluded from the presence of these cells in biopsies and the urinary sediment that CD4+T cells were potentially recruited to the kidney via CXCR3 and finally appeared in the urine. A selective accumulation of CD4+ cells expressing CCR4 was described as well in renal biopsies of patients with LN suggesting that this chemokine receptor plays a pivotal role in recruiting T cells to the kidney.18 These studies are confirmed by the presence of CD4+T cells in kidney biopsies as well as urinary CD4+T cells in our cohort. However, in contrast to the aforementioned investigations, we observed a predominant appearance of CD8+T cells in biopsies and the urine of patients with active renal disease. This is a new finding in SLE and corresponds with a previous study in which mainly CD8+T cells were detected in the urine of patients with IgA nephropathy, Henoch–Schönlein purpura nephritis and anti-neutrophil cytoplasmic antibody associated GN.9 Additionally, urinary CD8+T cells and also urinary CD4+T cells were correlated with disease activity. The phenotypic analysis of these urinary CD8+T cells and the observation of renal infiltrating CD8+T cells provide additional evidence that TEM cells might enforce inflammatory processes in LN. Remarkably, an overwhelming amount of T cells in the urine displayed an effector memory cell type, in contrast to the PB, supporting the hypothesis of selective migration of TEM cells into the inflamed tissue, in particular the kidney.

The migratory abilities as well as the harmful cytotoxic properties of TEM cells are not completely understood so far, but recent investigations may reveal underlying mechanisms. Evidence for the high migrating capacity of TEM cells comes from in vitro experiments demonstrating selective accumulation of TEM cells at the site of inflammation.12 19 20 Roberts et al19 demonstrated an enrichment of TEM cells in the peritoneal cavity of patients on peritoneal dialysis. Gattorno et al21 observed an enrichment of TEM cells in the synovial fluid of patients with juvenile arthritis as compare to the PB. The authors concluded that these TEM cells migrate selectively from PB into the inflamed tissue.

Urinary and infiltrating TEM cells are presumed to have cytotoxic capacities. A characteristic feature of CD8+TEM cells is the expression of perforin and granzyme B which mediate tissue injury.22 23 In renal transplant tissue the amount of infiltrated CD8+TEM cells, being strongly positive for granzyme B, has been found to be associated with the severity of graft damage.24 Apart from perforin and granzyme B, cytotoxic effector T cells are able to release interferon γ upon stimulation.25 Further investigations are necessary to reveal the mechanisms underlying migration and cell-mediated cytoxicity of CD8+TEM cells in the pathogenesis of LN.

The present study is the first investigating effector memory T cells in the PB, renal biopsy and urine of patients with SLE. The data provide strong evidence that CD8+TEM cells migrate from the PB to the kidney and appear in the urine during active LN. Therefore, CD8+TEM cells could be a useful monitoring tool in patients with SLE with renal involvement. Additionally, based on the potential pathological role of these CD8+TEM cells, they could represent a new therapeutic target.


This study was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG), DO 1419/1-1 (SD).



  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of The Medical Ethical Committee of the University Medical Center Groningen (UMCG).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Competing interests None declared.