Does quality of life improve for clients with autoimmune or neuromuscular disorders who received stem cell replacement?

Role of Stem Cell–Like Memory T Cells in Systemic Lupus Erythematosus

Ye Ji Lee, Ji Ah Park, Hyunmi Kwon, Youn Soo Choi, Kyeong Cheon Jung, Seong Hoe Park, and Eun Bong Lee

Objective. Stem cell–like memory T (Tscm) cells are long-lived memory T cells that have multipotent capacity to differentiate into different subsets. However, the role of Tscm cells in autoimmune diseases remains unclear. Here, we performed phenotypic studies to iden- tify Tscm cells in patients experiencing systemic lupus erythematosus (SLE).

Methods. CD4+ and CD8+ Tscm cells were identi- fied in SLE patients and healthy controls (HCs). In in vitro culture systems, CD4+ Tscm cells were induced to differentiate into subsets of T cells, including follicular helper T (Tfh) cells, and cytokine production patterns were assessed after stimulation. After confirming induc- tion of transcription factors for Tfh cells, the capacity of CD4+ Tscm-derived Tfh cells to help B cells was analyzed by measuring antibody secretion.

Results. The percentages of CD4+ and CD8+ Tscm cells among the naive CD4+/CD8+ or total CD4+ T cell populations were significantly higher in SLE patients than in HCs. Stimulated Tscm cells from SLE patients could replenish themselves and differentiate into other T lym- phocyte subsets, including Tfh cells upon stimulation with T cell receptor. Production of T cell factor 1, which is an inducer of Tfh, was also increased. The differentiated Tfh cells increased antibody production by autologous B cells.

Conclusion. Taken together, these findings suggest that Tscm cells play a role in the pathogenesis of SLE by maintaining Tfh cells.

Stem cell–like memory T (Tscm) cells are multipo- tent progenitor T cells with self-renewal capacity. They can differentiate into different subsets of mature lympho- cytes when stimulated via the T cell receptor (TCR). In the progressive differentiation model, Tscm cells are regarded as the earliest and least differentiated subset of memory T cells and are derived from priming of naive T cells (1). They share surface markers of CCR7+CD45RO�CD45RA+CD62L+ with naive T cells, but they also express memory cell–like surface markers such as CD95 (Fas)high and CD122 (interleukin-2Rb+ [IL-2Rb+]), and express stem cell–associated surface mar- ker (CD127+) (2). Although expression of CD95 by human naive CD8+ Tcells is negligible, Tscm cells up-reg- ulate the expression of CD95 similarly to conventional memory T cells (1), and the CD95 molecule is one of the critical surface markers to differentiate Tscm cells from naive Tcells. Tscm cells play various roles in humans with respect to protection against or persistence of damage (3). Tscm cells generated during various microbial infec- tions have a role in maintaining the microbes’ reservoir (4–7). Recently, there has been an effort to genetically modify Tscm cells to acquire long-term anticancer capac- ity (8). In contrast, Tscm cells can also contribute to devel- opment and maintenance of immune-mediated diseases such as aplastic anemia or uveitis (9,10). However, the role of human CD4+ Tscm cells in the pathogenesis of autoim- mune diseases, especially systemic lupus erythematosus (SLE), has not been clearly defined yet.

SLE is a prototypical multiorgan systemic autoim- mune disease characterized by the presence of autoanti- bodies against self antigens such as double-stranded DNA (dsDNA) and U-rich ribonucleoproteins (11). Immune complexes composed of antinuclear antibodies (ANAs) and their cognate antigens are deposited in the skin or

Supported in part by the National Research Foundation of Korea (Basic Science Research Program), which is funded by the Ministry of Education (grant 2016R1D1A1A02937044).

Ye Ji Lee, MSc, Ji Ah Park, MSc, Hyunmi Kwon, MD, Youn Soo Choi, PhD, Kyeong Cheon Jung, MD, PhD, Seong Hoe Park, MD, PhD, Eun Bong Lee, MD, PhD: Seoul National University Col- lege of Medicine, Seoul, South Korea.

Dr. E. B. Lee has received consulting fees from Pfizer (less than $10,000) and research grants from GC Pharma and Hanmi Phar- maceutical.

Address correspondence to Eun Bong Lee, MD, PhD, Divi- sion of Rheumatology, Department of Internal Medicine, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 03080, South Korea. E-mail: [email protected].

Submitted for publication September 1, 2017; accepted in revised form April 5, 2018.

1459

ARTHRITIS & RHEUMATOLOGY Vol. 70, No. 9, September 2018, pp 1459–1469 DOI 10.1002/art.40524 © 2018, American College of Rheumatology

other internal organs such as the kidneys and nervous sys- tem, triggering immune cell activation and secretion of proinflammatory cytokines such as interferon-a (IFNa) and tumor necrosis factor (TNF), which eventually results in damage to the affected organs (12).

Despite the clinical efficacy of immunosuppres- sants, complete remission of SLE is rare. Rather, autoanti- bodies remain in SLE patients, which may be the reason the disease is incurable (13). Production of high-affinity isotype-switched antibodies by B cells is tightly regulated by follicular helper T (Tfh) cells, suggesting that increased Tfh cell differentiation and activity plays an important role in the maintenance of autoantibodies in SLE (14,15). It has been suggested that production of autoantibodies in SLE patients is positively regulated by dysregulated Tfh differentiation. However, it is unclear how this immunopathologic Tfh compartment is replenished con- tinuously in SLE patients. To identify the cellular source of differentiated Tfh cells in SLE patients, we examined the role of Tscm cells in the pathogenesis of SLE. We found that the percentage of Tscm cells in SLE patients was higher than that in healthy controls (HCs). Moreover, Tscm cells from SLE patients, when stimulated, generated Tfh cells through Tfh-relevant transcription factors. Tscm cells may be a key immunoregulator for maintenance of SLE by playing a role as a continuous supplier of Tfh cells in this disease.

PATIENTS AND METHODS

Sample collection and clinical information. Sixty-five SLE patients and 72 healthy individuals were enrolled. Periph- eral blood was obtained, and peripheral blood mononuclear cells (PBMCs) were isolated on a Ficoll-Hypaque (GE Healthcare) gradient. All patients met the American College of Rheumatol- ogy revised classification criteria for SLE (16). Disease activity was assessed using the SLE Disease Activity Index 2000 (17). Laboratory investigations included white blood cell count, ery- throcyte sedimentation rate, and complement levels (C3 and C4; see Supplementary Table 1, available on the Arthritis & Rheuma- tology web site at http://onlinelibrary.wiley.com/doi/10.1002/art. 40524/abstract). ANA, anti-dsDNA, anti-SSA/Ro, and anti-SSB/ La antibodies were also measured. The study was approved by the Institutional Review Board of Seoul National University Hospital, and all enrolled patients and HCs provided informed consent.

Flow cytometry for surface phenotyping and cell sorting. Antibody specific for CD127 (clone A019D5) was obtained from BioLegend, and an anti-CCR7 antibody (clone 150503) was obtained from R&D Systems. Antibodies specific for CD3 (clone SK7), CD4 (SK3), CD8 (SK1), CD45RO (UCHL1), CD62L (DREG-56), CD45RA (HI100), CD28 (CD28.2), CD27 (M-T271), CD127 (A019D5), CD95 (DX2), CD122 (Mik-b3), CXCR5 (RF8B2), inducible Tcell costimulator (ICOS) (DX29), Bcl-6 (K112-91), and programmed death 1 (PD-1) (EH12.1) were obtained from BD Biosciences. After blocking Fc receptors

on PBMCs (1 9 107 cells/ml) with purified mouse anti-human IgG (BD Biosciences), the cells were stained with specific fluo- rescent antibodies to detect Tscm, central memory T lympho- cytes (Tcm), naive T lymphocytes (Tnaive), effector memory T lymphocytes (Tem), and Tfh. Accucheck counting beads (Ther- moFisher) were used to measure the absolute number of CD4+/ CD8+ Tscm cells.

Tscm cells were designated as CD3+CD4+/CD8+ CD45RO�CCR7+CD62L+CD45RA+ CD27+CD28+ CD127+ CD122+CD95high cells; Tcm cells were designated as CD3+CD4+/CD8+CD45RO+CCR7+CD45RA� cells; Tnaive cells were designated as CD3+CD4+/CD8+CD45RO�CCR7+ CD45RA+CD62L+CD27+CD95� cells; Tem cells were desig- nated as CD3+CD4+/CD8+CD45RO+CCR7�CD45RA� cells; and Tfh cells were designated as CD3+CD4+CXCR5+ICOS+ PD-1+ cells. Detailed gating procedures are explained in Fig- ure 1A. Labeled cells were detected using an LSR Fortessa flow cytometer (BD Biosciences) and sorted using a FACSAria instru- ment (BD Biosciences). All data were analyzed using FlowJo software.

Self-renewal capacity of Tscm cells. Cell proliferation was measured by staining for 10 minutes with 5 lM 5,6-carboxy- fluorescein diacetate succinimidyl ester (CFSE; Invitrogen). Tscm cells were incubated for 8 days with 30 ng/ml IL-15 (Pepro- Tech), and the number of CFSElow cells (Tscm cells) was counted by flow cytometry.

Multipotent capacity of Tscm cells. Cells were stimulated for 6 days with anti-CD3/CD28-coated beads (Dynabeads; Ther- moFisher) at a 1:1 ratio. Tnaive, Tcm, Tem, and terminal effector memory T (Temra) cells were confirmed by flow cytometry after staining with anti-CD3 (clone SK7), -CD4 (SK3), -CD8 (SK1), - CD45RO (UCHL1), -CCR7 (150503), and -CD45RA (HI100) antibodies.

Intracellular cytokine and transcription factor analysis. Sorted Tscm cells were stimulated for 5 hours with phorbol 12- myristate 13-acetate (PMA; 50 lg/ml) and ionomycin (1 mM) (both from Sigma Aldrich), in the presence of monensin (GolgiStop; BD Biosciences). Next, Fc receptors were blocked and cells were stained for CD3, CD4/CD8, CD45RA, CCR7, and CD45RO. Cells were then permeabilized with fixation/permeabi- lization solution (eBioscience) and stained for intracellular cytokines: IFNc (clone B27), IFNa (7N4-1; BD Biosciences), TNF (MAb11), and IL-2 (MQ1-17H12; eBioscience).

Differentiated Tscm-derived Tem cells were stained with antibodies against T-bet, GATA3, retinoic acid receptor–related ct (RORct), and FoxP3 to identify Th1, Th2, Th17, and Treg cell differentiation. Cells were stained for CD3, CD4, CD45RA, and CCR7, then permeabilized with fixation/permeabilization solu- tion (eBioscience) and stained for transcription factors: T-bet (4B10), GATA-3 (16E10A23; Biolegend), RORct (Q21-559; BD Biosciences), and FoxP3 (PCH101; eBioscience).

Quantitative real-time polymerase chain reaction (PCR). Transcription in Tfh cells was assessed by real-time quan- titative PCR. After extracting RNA from sorted Tscm cells using an RNeasy Mini Kit (Qiagen), complementary DNA was synthe- sized using a QuantiTect Reverse Transcription Kit (Qiagen). Levels of Tfh-associated transcripts (Bcl-6, CXCR5, ICOS, PD- 1, IL-21, B lymphocyte–induced maturation protein 1 [BLIMP- 1], lymphoid enhancer factor 1 [LEF-1], and T cell factor 1 [TCF-1) were detected with BioRad CFX 96 and SYBR Green (Qiagen), and fold changes in expression were measured using the 2�DDCt method. Gene expression was normalized to that of

1460 LEE ET AL

b-actin. Primer sequences of the above-mentioned transcripts are described in Supplementary Table 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10. 1002/art.40524/abstract. Amplicons were separated by gel elec- trophoresis in 1% agarose (Promega). Gels were imaged using GelDoc (BioRad Laboratories) and stained with RedSafe Nucleic Acid staining solution (Intron Biotechnology).

IL-21 enzyme-linked immunosorbent assay (ELISA) analysis. The culture supernatants from Tscm cells stimulated with anti-CD3/CD28 beads for 6 days were collected and stored at �20°C. IL-21 levels were measured using an ELISA (Biolegend) according to the instructions of the manufacturer. Standards and samples were measured in duplicate.

Measurement of immunoglobulin production in a co- culture assay. Tscm cells were stimulated for 6 days with anti- CD3/CD28 beads and then cocultured with an equal number

(5 9 104 T cells and same number of B cells) of autologous B cells (freshly isolated from peripheral blood using a negative B cell isolation kit [Miltenyi Biotec]) in the presence of RPMI 1640 (Welgene), 10% fetal bovine serum (Biowest), 1% peni- cillin/streptomycin (Gibco), and 200 ng/ml staphylococcal enterotoxin B (SEB; Toxin Technology) in 96-well, flat-bottomed plates (Corning). The supernatant and cells were harvested after an additional 7 days, and total IgG was measured in the culture supernatant using an ELISA from eBioscience, according to the instructions of the manufacturer. All samples were measured in triplicate.

For the coculture assay of Tscm-derived Tfh cells and B cells, Tscm-derived Tfh cells were isolated, by flow cytome- try, from Tscm cells, that had been stimulated with anti-CD3/ CD28 beads for 6 days. The same procedure was performed with autologous B cells as with whole Tscm cells.

Figure 1. Flow cytometry panel used to detect stem cell–like memory T (Tscm) cells and the percentage of CD4+ and CD8+ Tscm cells. A, Tscm cells designated as CD3+, CD4+ or CD8+, CD45RO�, CCR7+, CD45RA+, CD62L+, CD27+, CD28+, CD127+, CD122+, and CD95high. The pre- cise flow cytometry gating strategy is shown. B, Percentages of Tscm cells per naive T cells (CD4+ or CD8+, CD45RO�, CCR7+, CD45RA+, and CD62L+). Peripheral blood mononuclear cells from 57 healthy controls (HCs) and 52 systemic lupus erythematosus (SLE) patients were investigated. C, Percentages of Tscm cells per total CD4+ or CD8+ T cells. D, Absolute number of Tscm cells. Symbols represent individual sub- jects; bars show the mean � SEM. * = P < 0.05; *** = P < 0.001, by Student’s 2-tailed t-test. IL-7Ra = interleukin-7 receptor a; NS = not signifi- cant.

STEM CELL–LIKE MEMORY T CELLS IN SLE 1461

Statistical analysis. Data are expressed as the mean � SEM. For assessment of the significance of the differences between continuous variables, Student’s t-test (unpaired, 2-tailed) was used when the sample size was >30, while the Mann-Whitney U test was used when the sample size was <30. All graphs were constructed using Prism software (GraphPad). All statistical analyses were performed using SPSS (IBM).

RESULTS

Increased proportions of CD4+ or CD8+ Tscm cells in SLE patients versus HCs. The proportion of Tscm cells among the naive Tor total T lymphocyte populations was measured in 52 SLE patients and 57 HCs (Fig- ures 1B–D). The proportions of CD4+ and CD8+ Tscm cells (Tscm cells/naive Tcells) among the naive Tcell pop- ulations were significantly higher in SLE patients than in HCs (naive CD4+ T cells: mean � SEM 2.3 � 1.6% [SLE] versus 1.1 � 0.9% [HC]; P < 0.001; naive CD8+ T

cells: 5.4 � 4.4% [SLE] versus 3.4 � 3.4% [HC]; P = 0.013) (Figure 1B). The percentages of CD4+ and CD8+ Tscm cells among the total CD4 and CD8 T cell popula- tions were also analyzed. The Tscm compartment within the CD4 T cell population of SLE patients, but not con- trols, was selectively augmented (0.6 � 0.6% [SLE] versus 0.4 � 0.2% [HC]; P = 0.035) (Figure 1C), whereas the fre- quency of CD8+ Tscm cells among the total CD8 T cell populations was comparable in SLE patients and HCs (1.0 � 1.0% [SLE] versus 0.8 � 0.4% [HC]; P = 0.121) (Figure 1C). There was no difference in the absolute number of CD4+ or CD8+ Tscm cells between the 2 groups, even though the numbers of CD4+ or CD8+ Tscm cells were higher in some SLE patients (CD4+ Tscm cells among the CD4+ Tcells: 3,199 � 3,959 cells/ml [SLE] ver- sus 4,930 � 4,398 cells/ml [HC]; P = 0.061; CD8+ Tscm cells among the CD8+ T cells: 6,077 � 6,861 cells/ml [SLE] versus 4,075 � 2,679 cells/ml [HC]; P = 0.095)

Figure 2. Stem cell–like memory T (Tscm) cells differentiate into different T cell subsets upon T cell receptor (TCR) stimulation. A, Percentages of 5,6-carboxyfluorescein succinimidyl ester (CFSE)–diluted Tscm cells from systemic lupus erythematosus (SLE) patients and healthy controls (HCs) are shown. B, Self-renewal capacity of CD4+ Tscm cells is shown. Tscm cells (CD4+, CCR7�, CD45RO+, CD62L+, and CD95high) replicated after treatment with interleukin-15. C, Sorted CD4+ Tscm cells from 8 SLE patients and 7 HCs were stimulated with anti-CD3/CD28 beads for 6 days. After TCR stimu- lation, differentiated Tscm cells expressed a phenotype consistent with that of naive Tcells (Tnaive), central memory Tcells (Tcm), effector memory Tcells (Tem), or terminal effector memory Tcells (Temra; as discriminated by expression of CCR7 and CD45RA [gated on CD3+ and CD4+]). A summary of CD4+ Tscm differentiation patterns is also shown. D, Patterns of effector CD4+ Tcell subsets from Tscm-derived Tem cells are shown. Bars in C and D show the mean � SEM. * = P < 0.05. RORct = retinoic acid receptor–related ct.

1462 LEE ET AL

(Figure 1D). Taken together, the data indicate that the CD4+ T cell compartment in SLE patients harbors more CD4+ Tscm cells than that in HCs.

Ability of Tscm cells from SLE patients to differ- entiate into naive, central memory, or effector memory T cells. As expected, CD4+ Tscm cells could replicate them- selves. SLE Tscm cells tended to proliferate more than HC Tscm cells, resulting in more numbers of CFSE-diluted Tscm cells from SLE patients than those cells from HCs (mean � SEM 65.5 � 12.5% [SLE] versus 34.3 � 21.1% [HC]; P = 0.133) (Figure 2A). The self-renewal capacity of Tscm cells was highlighted by the appearance of Tscm cells in CFSE-diluted CD4+ T cells (Figure 2B). The propor- tion of Tscm cells within the CFSE-diluted CD4+ naive T cell population was 9.4 � 6.9%, whereas the proportion of Tscm cells within the total CD4+ Tcell population was 2.2 � 1.2% (see Supplementary Figure 1, available on the Arthritis & Rheumatology web site at http://onlinelibrary. wiley.com/doi/10.1002/art.40524/abstract).

To determine whether SLE Tscm cells can differ- entiate into daughter subsets (including Tnaive, Tcm, Tem, and Temra cells), we sorted Tscm cells from the PBMCs of 8 SLE patients and 7 HCs. These cells were then stimu- lated for 6 days at 37°C with anti-CD3/CD28 beads. While CD4+ Tscm cells from both SLE patients and HCs differentiated into Tnaive (CD3+CD4+CCR7+CD45RA+), Tcm (CCR7+CD45RA�), Tem (CCR7�CD45RA�), and Temra (CCR7�CD45RA+) cells (Figure 2B), we found that the SLE CD4+ Tscm cells tended to differentiate more into Tem cells at the expense of Tnaive and Tcm cells (Figure 2C) (Tnaive: mean � SEM 19.5 � 6.9% [SLE] ver- sus 26.6 � 8.5% [HC]; P = 0.536; Tcm: 17.2 � 3.6% [SLE] versus 26.3 � 7.1% [HC]; P = 0.397; Tem: 52.4 � 9.9% [SLE] versus 37.7 � 9.4% [HC]; P = 0.397; Temra: 10.9 � 3.0% [SLE] versus 9.4 � 4.4% [HC]; P = 0.694). Also, dif- ferentiated Tem cells presented different effector CD4+ T cell subsets, including Th1, Th2, Th17, and Treg cells (Fig- ure 2D). Differentiated Tem cells from SLE patients were

Figure 3. Differentiated stem cell–like memory T (Tscm) cells from systemic lupus erythematosus (SLE) patients secrete inflammatory cytokines. To identify cytokines secreted by activated Tscm cells, sorted CD4+ Tscm cells from 10 SLE patients and 4 healthy controls (HCs) were stimulated with T cell receptor (TCR) for 6 days. After TCR stimulation, Tscm cells were stimulated with 50 lg/ml phorbol 12-myristate 13-acetate and 1 mM ionomycin for 5 hours in the presence of GolgiStop. Symbols represent individual subjects; bars show the mean � SEM. IFNc = interferon-c; IL-2 = interleukin-2; TNF = tumor necrosis factor.

STEM CELL–LIKE MEMORY T CELLS IN SLE 1463

more skewed toward Treg cells than those from HCs (T- bet: 77.3 � 9.8% [SLE] versus 86.2 � 4.9% [HC]; P = 0.962; GATA3: 4.9 � 2.2% [SLE] versus 5.8 � 3.1% [HC]; P = 0.813; RORct: 0.8 � 0.8% [SLE] versus 1.2 � 0.6% [HC]; P = 0.800; FoxP3: 15.2 � 2.8% [SLE] versus 5.9 � 2.2% [HC]; P = 0.019). This result is consistent with the previous report that CD4+FoxP3+ T cells are more increased in SLE patients than in HCs (18).

Taken together, the data indicate that Tscm cells from SLE patients replenish themselves and differentiate

into different Tcell subsets upon appropriate stimulation, although the memory subset is favored.

Cytokine production by differentiated Tscm cells in SLE patients and controls. To better understand the nat- ure of immune responses elicited by Tscm cells, we next examined cytokine production by Tscm cells. Briefly, we stimulated Tscm cells from 10 SLE patients and 4 HCs with anti-CD3/CD28 beads for 6 days. After activating Tscm cells with PMA/ionomycin, intracellular staining was done for IFNc, IL-2, TNF, and IFNa (Figure 3). We found that

Figure 4. Stem cell–like memory T (Tscm) cells from systemic lupus erythematosus (SLE) patients differentiate into follicular helper T (Tfh) cells. To reveal whether Tscm cells can differentiate into Tfh cells, sorted Tscm cells from 17 SLE patients and 14 healthy controls (HCs) were examined for the presence of CD4+CXCR5+ICOS+PD-1+ cells after T cell receptor (TCR) stimulation with anti-CD3/CD28 beads for 6 days at 37°C. A, Tfh cells were designated as CD4+CXCR5+ICOS+PD-1+ cells. The numbers in the plots indicate the percentage of cells in each quadrant. B, Tscm cells from SLE patients (n = 18) and HCs (n = 12) were used to assess Bcl-6 expression. In A and B, symbols represent individual subjects; bars show the mean � SEM. C, RNA was extracted from CD4+ Tscm cells from 7 SLE patients and 7 HCs, cDNA was prepared, and real-time polymerase chain reaction analysis was performed. The cDNA from TCR-stimulated Tscm cells was compared with that from unstimulated Tscm cells. Bars show the mean � SEM. * = P < 0.05; ** = P < 0.01, by Mann-Whitney U test.

1464 LEE ET AL

cytokine production patterns from activated Tscm cells were similar between SLE patients and HCs but TNF tended to be higher in SLE patients than HCs (IFNc: mean � SEM 12.1 � 2.8% [SLE] versus 13.2 � 5.4% [HC]; P = 0.839; IL-2: 20.5 � 8.8% [SLE] versus 34.6 � 15.1% [HC]; P = 0.4; TNF: 17.3 � 5.5% [SLE] versus 5.3 � 2.6% [HC]; P = 0.161; IFNa: 13.2 � 6.5% [SLE] versus 24.6 � 11.1% [HC]; P = 0.730). The data suggest that Tscm cells from SLE patients can trigger different patterns of inflammatory responses.

Ability of Tscm cells to differentiate into Tfh cells. Data indicate that Tscm cells from SLE patients differen- tiate into effector CD4+ T cells upon stimulation. The level of autoantibody production in SLE is strongly associ- ated with disease activity (19,20) and Tfh cells help B cells to form germinal centers, leading to production of high- affinity and isotype-switched antibodies (21). Thus, we examined whether Tscm cells from SLE patients could differentiate into Tfh cells. For this, Tscm cells from 17 SLE patients and 14 HCs were sorted and stimulated to see whether they could differentiate into Tfh cells. We found that CD4+CXCR5+PD-1+ICOS+ T cells were derived from the Tscm cell compartment (Figure 4A). Tscm cells from SLE patients tended to generate more Tfh cells than those from HCs (9.099% [SLE] versus 6.380% [HC]), although the difference was not significant (P = 0.226). The proportion of CD4+ T cells that coex- pressed Bcl-6, a critical factor for Tfh function and differ- entiation (22), was higher in SLE patients (n = 18) than in HCs (n = 12) (mean fluorescence intensity [MFI] 338.8 � 59.5 [SLE] versus 273.9 � 52.5 [HC] versus 26.9 [isotype control]) (15.1 � 3.7% [SLE] versus 7.7 � 3.5% [HC]; P = 0.035) (Figure 4B).

The capacity of Tscm cells to differentiate into Tfh cells was analyzed at the RNA level. RNA was extracted from Tscm cells that had been sorted and cul- tured for 6 days in the presence of TCR stimuli. Quanti- tative PCR was then performed, and expression of BCL6, CXCR5, PDCD1, and ICOS (genes strongly asso- ciated with Tfh differentiation and function) (14,23) was analyzed and compared with that of b-actin (Figure 4C). We found that expression of PDCD1 and ICOS was induced more strongly in Tscm cells from SLE patients than in those from controls. While both Tscm compart- ments expressed comparable levels of CXCR5 before stimulation, Tscm cells from SLE patients showed in- creased expression of CXCR5 (and the other genes) upon stimulation (Figure 4C), suggesting that Tscm cells from SLE patients harbor a genetic program that favors differentiation into Tfh cells. Gel images showing complementary DNA levels were consistent with the real-time quantitative PCR results (see Supplementary

Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.40524/ abstract). Taken together, the data demonstrate that Tscm cells contribute to the Tfh compartment via a Tfh-biased genetic program.

Tscm-derived Tfh cells from both SLE patients and HCs showed a phenotype of central memory and effector memory Tcells, consistent with published reports (24–26). Interestingly, Tscm-derived Tfh cells from 6 SLE patients tended to have more Temra phenotype than those cells from 6 HCs. However, other phenotypes were comparable (Tnaive: mean � SEM 9.6 � 1.5% [SLE] ver- sus 6.6 � 1.4% [HC]; P = 0.310; Tcm: 45.1 � 5.6% [SLE] versus 52.3 � 7.1% [HC]; P = 0.818; Tem: 38.2 � 5.5% [SLE] versus 37.7 � 6.8% [HC]; P = 0.937; Temra: 7.0 � 1.5% [SLE] versus 3.0 � 0.8% [HC]; P = 0.026) (see Sup- plementary Figure 3, available on the Arthritis & Rheuma- tology web site at http://onlinelibrary.wiley.com/doi/10. 1002/art.40524/abstract).

Molecular signatures that lead to augmented Tfh cell differentiation by Tscm cells in SLE patients. Since BCL6 is a critical factor in differentiation into Tfh cells, we investigated whether transcription factors known to regu- late BCL6 expression in CD4+ Tcells are differentially reg- ulated in Tscm cells from SLE patients and controls. IL-21 is a key cytokine that induces differentiation of Tfh cells (27). When we examined the levels of IL-21 transcripts, we found higher levels in Tscm cells from SLE patients than in those from HCs after in vitro stimulation (Fig- ure 5A). However, IL-21 levels in culture supernatant of Tscm cells from 24 SLE patients were comparable to those from 16 HCs (17.9 � 4.5 pg/ml [SLE] versus 15.2 � 3.3 pg/ml [HC]; P = 0.652) (Figure 5A). BLIMP-1, origi- nally identified as a transcriptional repressor of BCL6 in B cells (22), antagonizes Tfh differentiation (28). As such, we examined expression of PRDM1, the gene encoding BLIMP-1, in Tscm cells from SLE patients and HCs. Com- pared with Tscm cells from HCs, Tscm cells from SLE patients showed down-regulated transcription of PRDM1 upon TCR stimulation (Figure 5B). Moreover, expression of LEF1 and TCF7, which encode Lef-1 and Tcf-1, respec- tively (transcription factors recently shown to play critical roles in suppression of BLIMP-1 in mice) (29), was strongly induced in Tscm cells from SLE patients when compared with those from HCs (Figure 5C). Taken together, the data suggest that Tscm cells from SLE patients are ready to form Tfh cells when appropriately stimulated by exposure to IL-21 and repression of BLIMP-1.

Tscm-derived Tfh cells and B cell production of antibodies in vitro. Because Tscm cells from SLE patients show high expression of genes related to Tfh differentia- tion (PDCD1, ICOS, CXCR5, BCL6, LEF1, and TCF7)

STEM CELL–LIKE MEMORY T CELLS IN SLE 1465

and function (IL-21), along with strong inhibition of PRDM1 expression, we investigated whether Tscm cells can induce production of antibodies by B cells. For this, we sorted Tscm cells and stimulated them with anti-CD3/ CD28 beads for 6 days in vitro, after which autologous B cells were added in the presence of SEB, a superantigen that induces B–T interactions (30) (see Supplementary Figure 4, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.40524/ abstract). The amount of human IgG in the culture medium was then measured by ELISA after 7 more days of coculture. In accordance with the results showing that Tscm cells differentiate into Tfh cells (Figure 5), we found that autologous B cells produced more antibodies when cocultured with Tscm cells from SLE patients than when

cocultured with cells from HCs (mean � SEM 10,509.6 � 5,157.8 pg/ml [SLE] versus 1,907.8 � 671.7 pg/ml [HCs]; P = 0.034) (Figure 6A). In addition, to evaluate the direct role of Tscm-derived Tfh cells, we sorted out Tscm-derived Tfh cells on day 6 of TCR stimulation and cocultured them with equal numbers of autologous B cells and SEB for an additional 7 days. As with Tscm cells, B cells pro- duced more antibodies when cocultured with Tscm-derived Tfh cells from SLE patients than those cells from HCs (91.0 � 50.8 pg/ml [SLE] versus 13.9 � 7.6 pg/ml [HCs]; P = 0.200) (Figure 6B). Collectively, these data demonstrate that Tfh cells that have developed from Tscm cells may help B cells produce antibodies, and suggest that Tscm cells found in SLE patients can be functional by contribut- ing to production of antibodies.

Figure 5. Increased interleukin-21 (IL-21) levels, up-regulated T cell factor 1 (TCF-1) expression, and down-regulated B lymphocyte–induced mat- uration protein 1 (BLIMP-1) expression promote generation of follicular helper T (Tfh) cells. The levels of IL-21, lymphoid enhancer factor 1 (LEF-1), TCF-1, and BLIMP-1 transcripts were measured to investigate the mechanisms underlying Tfh differentiation. Stem cell–like memory T (Tscm) cells isolated from peripheral blood mononuclear cells (PBMCs) from 5 systemic lupus erythematosus (SLE) patients and 5 healthy con- trols (HCs) were stimulated with anti-CD3/CD28 beads for 6 days. Tscm cells isolated directly from PBMCs and T cell receptor–stimulated Tscm cells were compared. A, Transcription of IL-21 was assessed by real-time quantitative polymerase chain reaction (qPCR). Inset shows IL-21 levels in the culture supernatant, assessed by enzyme-linked immunosorbent assay. B, Levels of BLIMP-1 transcripts were measured by real-time qPCR. C, Levels of LEF-1 and TCF-1 transcripts were measured by real-time qPCR. Expression was calculated and compared with levels before stimula- tion. Bars show the mean � SEM. * = P < 0.05 by Mann-Whitney U test.

1466 LEE ET AL

Correlation between Tscm cells and disease activ- ity in SLE. The data so far demonstrate that Tscm cells from SLE patients not only generate more Tfh cells, but also provide IL-21 signals for antibody production by B cells. This led us to investigate the clinical significance of Tscm cells with respect to SLE disease activity. We ana- lyzed the correlation between serum levels of autoanti- bodies and the number of Tscm cells in SLE patients (see Supplementary Table 3, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.c om/doi/10.1002/art.40524/abstract). Considering errors inherent in multiple comparisons, we did not find any clinical parameters that were associated with the per- centages of CD4+ Tscm or CD8+ Tscm cells.

DISCUSSION

This study shows the presence of Tscm cells in SLE and their possible role in the maintenance of SLE through Tfh cell differentiation. The proportion of Tscm cells among T cells is higher in SLE patients than in HCs, and isolated Tscm cells can renew themselves and differentiate into different subsets of T lymphocytes, including naive, central memory, and effector memory T lymphocytes. Tscm cells can also differentiate into Tfh cells in response to IL- 21 and BLIMP-1, which are in turn regulated by TCF-1. Tscm-derived Tfh cells from SLE patients can induce autol- ogous B cells to secrete more immunoglobulins.

SLE patients have higher percentages of CD4+ Tscm and CD8+ Tscm cells (Figures 1B and C). The higher percentage of CD4+ Tscm cells in SLE patients is

important with respect to the pathogenesis of SLE, which is characterized by the persistence of autoantibodies pro- duced by activated B cells (14). In addition, CD8+ Tcells, particularly CD3+CD28� T cells, are associated with sev- eral autoimmune diseases, including SLE (31,32). In con- trast to the percentage, the absolute number of CD4+ Tscm cells in SLE patients and HCs was comparable (Fig- ure 1D). These rather contradictory results are related to leukopenia in SLE patients and suggest that inherent changes in Tscm cells, not simply the number of Tscm cells, may be important for the maintenance of disease. Thus, the relative balance between Tscm cells and other T cell populations may be important in the pathogenesis of SLE.

Inflammatory cytokines such as IFNa, IFNc, IL-6, IL-1, and TNF play critical roles in SLE pathogenesis (33). Here, we found that Tscm cells from SLE patients secreted IFNc and TNF when stimulated via TCR (Fig- ure 3). This suggests that Tscm cells from SLE patients are inherently proinflammatory in nature. The tendency of Tscm cells to form more memory type Tcells (Figure 3) may underlie secretion of proinflammatory cytokines in SLE patients. Interestingly, secretion of IFNa, which is regarded as a key cytokine in SLE, by Tscm cells from SLE patients was not higher than that by Tscm cells from HCs. However, IFNa is mainly secreted by plasmacytoid dendritic cells, not Tcells (34–36).

Tfh cells stimulate B cells in germinal centers, resulting in affinity maturation and immunoglobulin iso- type switching. Therefore, Tfh cells help the formation of memory B cells and long-lived plasma cells at germinal centers (24). In SLE, many autoreactive IgG+ memory B

Figure 6. Coculture of follicular helper T (Tfh) cells and B cells and the IgG assay. A, Secretion of human immunoglobulins. Autologous B cells from systemic lupus erythematosus (SLE) patients secreted more immunoglobulins when cocultured with stem cell–like memory T (Tscm) cells from SLE patients than when cultured with cells from healthy controls (HCs). B, Secretion of human immunoglobulins from coculture of Tscm- derived Tfh cells and B cells. Bars show the mean � SEM. * = P < 0.05 by Mann-Whitney U test.

STEM CELL–LIKE MEMORY T CELLS IN SLE 1467

cells are somatically mutated (37). Moreover, the percent- age of somatically mutated antibody-producing plas- mablasts is increased in the peripheral blood of SLE patients (38). These results suggest that Tfh cells make an important contribution to the maintenance of pathogenic B cells in SLE. We also showed that Tfh cells can be derived from Tscm cells. Maintenance of Bcl-6, which is a critical transcriptional regulator in Tfh cells (22), depends on the balance between IL-21 (inducer) and BLIMP-1 (repressor) expression (27). The observed induction of IL-21 and repression of BLIMP-1 in Tscm cells upon TCR stimulation in SLE patients (Figure 5) suggests that differentiation of SLE Tscm cells can be skewed toward Tfh cells when they are appropriately stimulated.

In SLE patients, transcription of TCF-1 signifi- cantly increased during differentiation of Tfh cells from stimulated Tscm cells (Figure 5). In a study by Choi et al, transcription factors LEF-1 and TCF-1 (involved in Wnt/ b-catenin signaling) caused activated CD4+ T cells to dif- ferentiate into Tfh cells in a mouse model (29). They showed that Tcf7 (encoding TCF-1)�/� or Lef1 (encoding LEF-1)�/�Tcf7�/� mice had significantly lower numbers and percentages of germinal center (GC) Tfh cells (Bcl- 6+CXCR5+PD-1high) and GL7+Fas+ GC B cells than their control littermates. In vitro data from humans also show binding of TCF-1 to Bcl-6 in 293T human embryonic kidney cells (39). Up-regulation of TCF-1 promotes dif- ferentiation toward Tfh cells by repressing BLIMP-1 (encoded by PRDM-1). Therefore, we can infer that increased expression of TCF-1 in stimulated Tscm cells suppresses the BLIMP-1 transcription factor and increases the level of Bcl-6, which subsequently leads to generation of Tfh cells.

Serum levels of IL-6 are significantly higher in SLE patients than in HCs (40,41). IL-6 favors develop- ment of Tfh cells. IL-6 is also required for development of early Bcl-6+CXCR5+ Tfh cells (42). Another study showed that activation of STAT-3 by pretreatment with IL-6 increases TCF-1 and b-catenin levels (43). There- fore, it can be inferred that increased TCF-1 levels in SLE patients after TCR stimulation may be favored by an environment containing high levels of IL-6.

Neither disease activity nor autoantibody levels in SLE patients were associated with the percentages of Tscm cells. These results suggest that Tscm cells con- tribute to the development and maintenance of SLE, while disease activity is determined by other factors that affect the stimulation status of the involved immune cells.

In conclusion, the percentage of Tscm cells in SLE patients is higher than that in HCs. Also, Tscm-derived Tfh cells contribute to increased humoral autoimmunity by helping B cells produce antibodies.

ACKNOWLEDGMENTS

We thank the patients and healthy controls for provid- ing blood for this study.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Lee had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Y. J. Lee, E. B. Lee. Acquisition of data. Y. J. Lee, J. A. Park, E. B. Lee. Analysis and interpretation of data. Y. J. Lee, J. A. Park, Kwon, Choi, Jung, S. H. Park, E. B. Lee.

REFERENCES

1. Gattinoni L, Lugli E, Ji Y, Pos Z, Paulos CM, Quigley MF, et al. A human memory T cell subset with stem cell-like properties. Nat Med 2011;17:1290–7.

2. Hong H, Gu Y, Sheng SY, Lu CG, Zou JY, Wu CY. The distri- bution of human stem cell-like memory T cell in lung cancer. J Immunother 2016;39:233–40.

3. Gattinoni L, Speiser DE, Lichterfeld M, Bonini C. T memory stem cells in health and disease. Nat Med 2017;23:18–27.

4. Marraco SA, Soneson C, Cagnon L, Gannon PO, Allard M, Maillard SA, et al. Long-lasting stem cell-like memory CD8(+) T cells with a naive-like profile upon yellow fever vaccination. Sci Transl Med 2015;7282ra48.

5. Gattinoni L. The dark side of T memory stem cells. Blood 2015; 125:3519–20.

6. Buzon MJ, Sun H, Li C, Shaw A, Seiss K, Ouyang ZY, et al. HIV-1 persistence in CD4(+) T cells with stem cell like proper- ties. Nat Med 2014;20:139–42.

7. Nagai Y, Kawahara M, Hishizawa M, Shimazu Y, Sugino N, Fujii S, et al. T memory stem cells are the hierarchical apex of adult T-cell leukemia. Blood 2015;125:3527–35.

8. Sabatino M, Hu JH, Sommariva M, Gautam S, Fellowes V, Hocker JD, et al. Generation of clinical-grade CD19-specific CAR-modified CD8(+) memory stem cells for the treatment of human B-cell malignancies. Blood 2016;128:519–28.

9. Hosokawa K, Muranski P, Feng XM, Townsley DM, Liu BY, Knickelbein J, et al. Memory stem T cells in autoimmune disease: high frequency of circulating CD8(+) memory stem cells in acquired aplastic anemia. J Immunol 2016;196:1568–78.

10. Roederer M, Quaye L, Mangino M, Beddall MH, Mahnke Y, Chattopadhyay P, et al. The genetic architecture of the human immune system: a bioresource for autoimmunity and disease pathogenesis. Cell 2015;161:387–403.

11. Gensous N, Schmitt N, Richez C, Ueno H, Blanco P. T follicular helper cells, interleukin-21 and systemic lupus erythematosus. Rheumatology (Oxford) 2017;56:516–23.

12. Lech M, Anders HJ. The pathogenesis of lupus nephritis. J Am Soc Nephrol 2013;24:1357–66.

13. Wilhelm TR, Magder LS, Petri M. Remission in systemic lupus erythematosus: durable remission is rare. Ann Rheum Dis 2017; 76:547–53.

14. Xu H, Liu J, Cui X, Zuo Y, Zhang Z, Li Y, et al. Increased fre- quency of circulating follicular helper T cells in lupus patients is associated with autoantibody production in a CD40L-dependent manner. Cell Immunol 2015;295:46–51.

15. Blanco P, Ueno H, Schmitt N. T follicular helper (Tfh) cells in lupus: activation and involvement in SLE pathogenesis. Eur J Immunol 2016;46:281–90.

1468 LEE ET AL

16. Hochberg MC, for the Diagnostic and Therapeutic Criteria Com- mittee of the American College of Rheumatology. Updating the American College of Rheumatology revised criteria for the classi- fication of systemic lupus erythematosus [letter]. Arthritis Rheum 1997;40:1725.

17. Gladman DD, Ibanez D, Urowitz MB. Systemic Lupus Erythe- matosus Disease Activity Index 2000. J Rheumatol 2002;29:288–91.

18. Bonelli M, von Dalwigk K, Savitskaya A, Smolen JS, Scheinecker C. Foxp3 expression in CD4(+) T cells of patients with systemic lupus erythematosus: a comparative phenotypic analysis. Ann Rheum Dis 2008;67:664–71.

19. Andrejevic S, Jeremic I, Sefik-Bukilica M, Nikolic M, Stojimirovic B, Bonaci-Nikolic B. Immunoserological parameters in SLE: high-avidity anti-dsDNA detected by ELISA are the most closely associated with the disease activity. Clin Rheumatol 2013;32: 1619–26.

20. Touma Z, Gladman DD, Tulloch-Reid D, Toloza SM, Ibanez D, Fortin PR, et al. Burden of autoantibodies and association with disease activity and damage in systemic lupus erythematosus. Clin Exp Rheumatol 2010;28:525–31.

21. Weinstein JS, Hernandez SG, Craft J. T cells that promote B-cell maturation in systemic autoimmunity. Immunol Rev 2012;247: 160–71.

22. Yu D, Rao S, Tsai LM, Lee SK, He YQ, Sutcliffe EL, et al. The transcriptional repressor Bcl-6 directs T follicular helper cell lin- eage commitment. Immunity 2009;31:457–68.

23. Ramiscal RR, Vinuesa CG. T-cell subsets in the germinal center. Immunol Rev 2013;252:146–55.

24. Choi JY, Ho JH, Pasoto SG, Bunin V, Kim ST, Carrasco S, et al. Circulating follicular helper–like T cells in systemic lupus erythe- matosus: association with disease activity. Arthritis Rheumatol 2015;67:988–99.

25. He J, Tsai LM, Leong YA, Hu X, Ma CS, Chevalier N, et al. Cir- culating precursor CCR7loPD-1hi CXCR5+ CD+ T cells indicate Tfh cell activity and promote antibody responses upon antigen reexposure. Immunity 2013;39:770–81.

26. Crotty S. T follicular helper cell differentiation, function, and roles in disease. Immunity 2014;41:529–42.

27. Linterman MA, Beaton L, Yu D, Ramiscal RR, Srivastava M, Hogan JJ, et al. IL-21 acts directly on B cells to regulate Bcl-6 expression and germinal center responses. J Exp Med 2010;207: 353–63.

28. Johnston RJ, Poholek AC, DiToro D, Yusuf I, Eto D, Barnett B, et al. Bcl6 and Blimp-1 are reciprocal and antagonistic regulators of T follicular helper cell differentiation. Science 2009;325:1006–10.

29. Choi YS, Gullicksrud JA, Xing SJ, Zeng ZH, Shan Q, Li FY, et al. LEF-1 and TCF-1 orchestrate T-FH differentiation by regu- lating differentiation circuits upstream of the transcriptional repressor BcI6. Nat Immunol 2015;16:980–90.

30. Morita R, Schmitt N, Bentebibel SE, Ranganathan R, Bourdery L, Zurawski G, et al. Human blood CXCR5+CD4+ T cells are counterparts of T follicular cells and contain specific subsets that differentially support antibody secretion. Immunity 2011;34: 108–21.

31. Zabinska M, Krajewska M, Koscielska-Kasprzak K, Klinger M. CD3+CD8+CD28- T lymphocytes in patients with lupus nephritis. J Immunol Res 2016;2016:1058165.

32. Wang EC, Lawson TM, Vedhara K, Moss PA, Lehner PJ, Borysiewicz LK. CD8high+ (CD57+) T cells in patients with rheumatoid arthritis. Arthritis Rheum 1997;40:237–48.

33. Ohl K, Tenbrock K. Inflammatory cytokines in systemic lupus ery- thematosus. J Biomed Biotechnol 2011;2011:432595.

34. Kassim SH, Jordan J, Schreiter J, Adhikarakunnathu S, Baribaud F, Mateo LS. Systematic identification of novel SLE related autoantibodies responsible for type I IFN production in human plasmacytoid dendritic cells. Cell Immunol 2013;284:119–28.

35. Sacre K, Criswell LA, McCune JM. Hydroxychloroquine is associ- ated with impaired interferon-a and tumor necrosis factor-a produc- tion by plasmacytoid dendritic cells in systemic lupus erythematosus. Arthritis Res Ther 2012;14:R155.

36. Rowland SL, Riggs JM, Gilfillan S, Bugatti M, Vermi W, Kolbeck R, et al. Early, transient depletion of plasmacytoid dendritic cells ameliorates autoimmunity in a lupus model. J Exp Med 2014;211: 1977–91.

37. Tiller T, Tsuiji M, Yurasov S, Velinzon K, Nussenzweig MC, Wardemann H. Autoreactivity in human IgG+ memory B cells. Immunity 2007;26:205–13.

38. Jacobi AM, Odendahl M, Reiter K, Bruns A, Burmester GR, Radbruch A, et al. Correlation between circulating CD27high

plasma cells and disease activity in patients with systemic lupus erythematosus. Arthritis Rheum 2003;48:1332–42.

39. Xu LF, Cao Y, Xie ZY, Huang QZ, Bai Q, Yang X, et al. The transcription factor TCF-1 initiates the differentiation of T-FH cells during acute viral infection. Nat Immunol 2015;16:991–9.

40. Galil SM, Ezzeldin N, El-Boshy ME. The role of serum IL-17 and IL-6 as biomarkers of disease activity and predictors of remission in patients with lupus nephritis. Cytokine 2015;76:280–7.

41. Solus JF, Chung CP, Oeser A, Li C, Rho YH, Bradley KM, et al. Genetics of serum concentration of IL-6 and TNFa in systemic lupus erythematosus and rheumatoid arthritis: a candidate gene analysis. Clin Rheumatol 2015;34:1375–82.

42. Choi YS, Eto D, Yang JA, Lao C, Crotty S. Cutting edge: STAT1 is required for IL-6-mediated Bcl6 induction for early follicular helper cell differentiation. J Immunol 2013;190:3049–53.

43. Pramanik KC, Fofaria NM, Gupta P, Ranjan A, Kim SH, Srivastava SK. Inhibition of b-catenin signaling suppresses pancreatic tumor growth by disrupting nuclear b-catenin/TCF-1 complex: critical role of STAT-3. Oncotarget 2015;6:11561–74.

STEM CELL–LIKE MEMORY T CELLS IN SLE 1469

Copyright of Arthritis & Rheumatology is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.