Immunosuppression Has Long-Lasting Effects on Circulating Follicular Regulatory T Cells in Kidney Transplant Recipients

Background: FoxP3+ follicular regulatory T cells (Tfr) have been identified as the cell population controlling T follicular helper (Tfh) cells and B cells which, are both involved in effector immune responses against transplanted tissue. Methods: To understand the biology of Tfr cells in kidney transplant patients treated with tacrolimus and mycophenolate mofetil (MMF) combination immunosuppression, we measured circulating (c)Tfh and cTfr cells in peripheral blood by flow cytometry in n = 211 kidney transplant recipients. At the time of measurement patients were 5–7 years after transplantation. Of this cohort of patients, 23.2% (49/211) had been previously treated for rejection. Median time after anti-rejection therapy was 4.9 years (range 0.4–7 years). Age and gender matched healthy individuals served as controls. Results: While the absolute numbers of cTfh cells were comparable between kidney transplant recipients and healthy controls, the numbers of cTfr cells were 46% lower in immunosuppressed recipients (p < 0.001). More importantly, in transplanted patients, the ratio of cTfr to cTfh was decreased (median; 0.10 vs. 0.06), indicating a disruption of the balance between cTfr and cTfh cells. This shifted balance was observed for both non-rejectors and rejectors. Previous pulse methylprednisolone or combined pulse methylprednisolone + intravenous immunoglobulin anti-rejection therapy led to a non-significant 30.6% (median) and 51.2% (median) drop in cTfr cells, respectively when compared to cTfr cell numbers in transplant patients who did not receive anti-rejection therapy. A history of alemtuzumab therapy did lead to a significant decrease in cTfr cells of 85.8% (median) compared with patients not treated with anti-rejection therapy (p < 0.0001). No association with tacrolimus or MMF pre-dose concentrations was found. Conclusion: This cross-sectional study reveals that anti-rejection therapy with alemtuzumab significantly lowers the number of cTfr cells in kidney transplant recipients. The observed profound effects by these agents might dysregulate cTfr functions.

INTRODUCTION Improvement of long-term outcomes after kidney transplantation remains a challenge (1)(2)(3)(4)(5). The most recent findings based on the United States registry data and Collaborative Transplant Study across 21 European countries, report only a slight improvement in renal allograft survival since the early 2000's (6)(7)(8). Antibody-mediated immune responses are recognized as an important factor in late kidney allograft failure (9)(10)(11). This immune response is refractory to treatment with conventional immunosuppression (12)(13)(14).
Follicular T helper (Tfh) cells play a critical role in B celldependent antibody generation (15)(16)(17). These Tfh cells colocalize with B cells in germinal centers within secondary lymphoid organs (SLOs) and are specialized in assisting antigen activated B cells to differentiate into antibody-producing plasma cells (18,19). This immune response is controlled by follicular regulatory T (Tfr) cells, a unique subset of regulatory T (Treg) cells, that inhibits Tfh and B cell responses (20,21). Tfr cells exert immune inhibitory functions through down-regulating the costimulatory molecule CD86 on B cells (22), producing inhibitory cytokines e.g., interleukin (IL)-10 and mediating cytolysis (23). Both Tfr and Tfh cells express high levels of CXC chemokine receptor 5 (CXCR5) and programmed death 1 (PD-1) (24), as well as the transcriptional factor B cell lymphoma (Bcl)-6 (25,26). Tfr cells can also express Helios a transcription factor expressed by natural Tregs, and a marker reflecting enhanced immunosuppressive functions (27). From studies in patients suffering from auto-immune disease we know that the dynamic balance between Tfr and Tfh cells is important for immune homeostasis and tolerance. An aberrant Tfr/Tfh ratio has been linked to the development of autoantibody immunitymediated diseases including systemic lupus erythematosus (28), rheumatoid arthritis (29), and myasthenia gravis (30). Also in kidney transplant recipients it was shown that antibody mediated rejection (ABMR) and chronic allograft dysfunction are associated with a disturbed balance of circulating (c)Tfh and cTfr cells (31,32).
alloantigen-activated B cells (35). From experimental transplant models we know that Tfh cells are capable of stimulating alloantibody production which in turn mediates the humoral response against the allograft (36,37). These findings highlight the importance of Tfh cells in the process of transplant rejection while the first data about Tfr cells point to their impaired function in controlling the actions of Tfh cells in the allogeneic responses after organ transplantation (32).
Immunosuppressive drugs, including the calcineurin inhibitor (CNI) tacrolimus, may decrease the number and impair the function of Tfh and Tfr cells (38). However, to what extent immunosuppressive therapies affect the biology of Tfr and Tfh cells after kidney transplantation is unknown.
In the present observational study, the absolute numbers of Tfr cells and their relation to Tfh cells in immunosuppressed renal transplant recipients and healthy controls was investigated. In addition, the impact of anti-rejection therapy on these cell populations was studied.

Study Population
A cohort of 211 renal transplant recipients transplanted between 2012 and 2014 was sampled cross sectionally at the timepoint of 5.0-7.7 years (median 5.3) post-transplant. All patients had a functional graft at the time of blood sampling, kidney function was assessed by estimated glomerular filtration rate (eGFR, mL/min per 1.73 m 2 ) as calculated by the CKD-EPI equation. Transplant recipients (n = 202) received induction therapy with basiliximab [Simulect R ; Novartis, Basel, Switzerland; 20 mg intravenously on days 0 and 4], rabbit anti-thymocyte globulin [Thymoglobulin R , Sanofi Genzyme, United States (n = 2) or rituximab [MabThera R ; Roche, Basel, Switzerland (n = 7) for blood group ABO incompatible kidney transplantation [rituximab 375 mg/m 2 4 weeks before transplantation; tacrolimus 0.1 mg/kg b.i.d, mycophenolate mofetil (MMF) 1,000 mg b.i.d; prednisone 20 mg once daily starting 2 weeks before transplantation]. The post-operative immunosuppressive regimen after transplantation consisted of tacrolimus (Prograf R ; Astellas Pharma, Tokyo, Japan; aiming for pre-dose concentrations of 10-15 ng/mL in weeks 1-2, 8-12 ng/mL in weeks 3-4 and 5-10 ng/mL thereafter), MMF (Cellcept R ; Roche, Basel, Switzerland; starting dose of 1 g twice a day, aiming for pre-dose concentrations of 1.5-3.0 mg/L) and prednisolone. Prednisolone was tapered to 5 mg at month 3 and withdrawn at months 4-5. In patients who experienced a rejection event, prednisolone was reintroduced to 10 mg once daily. Rejection was defined as presumed, treated acute rejection (in patients who had no biopsy or an inconclusive biopsy) or biopsy-proven rejection (BPR) as part of routine clinical care by a renal pathologist using 2 µm paraffin sections stained for HE, PAS, Jones and immunohistochemistry for C4d on 4 µm sections. After the completion of the study, all biopsies were reviewed again by a clinical pathologist (M.C.C.) in a blinded fashion per the Banff '15 classification (39). Patients suffering from vascular (Banff grade 2) or tubulo-interstitial (Banff grade 1) rejection received high-dose pulse intravenous methylprednisolon as firstline therapy (methylprednisolone 1,000 mg i.v. for 3 consecutive days), whereas patients suffering from severe or glucocorticoidresistant acute rejection were treated with alemtuzumab (one or two doses of 30 mg s.c.). Patients suffering from ABMR or mixed-type rejection (TCMR and ABMR) were treated with methylprednisolone plus intravenous immunoglobulins (IVIg) (40, 41). While the majority of patients in our study were treated according to the above mentioned guidelines, there are some instances where biopsy scores were changed after the second biopsy revision done according to the Banff '15 classification. A disparate treatment from the guidelines was also chosen for a few patients for other clinical reasons (e.g., suspicion of humoral immunity components or patient history).

Detection of Anti-HLA Antibodies
The complement-dependent cytotoxicity cross-match was negative before transplantation in all patients. Serum samples from recipients were screened for the presence of HLA antibodies using the Lifecodes Lifescreen Deluxe (LMX) kit, according to the manufacturer's manual (Immucor Transplant Diagnostics Inc. Stamford, CT, USA) 5-7 years post-transplant at the time of blood sampling. Anti-HLA class I (HLA-A, HLA-B, or HLA-C) or HLA class II (HLA-DR or HLA-DQ) antibodies were further analyzed with a Luminex Single Antigen assay using LABscreen HLA class I and class II antigen beads (One Lambda, Canoga Park, GA, USA), as described in our previous study (42). The presence of donor-specific antibodies (DSA) was determined by comparing the various HLA specificities with donor HLA typing.

Statistical Analyses
Data are presented as median with interquartile range (IQR), unless otherwise specified. The Mann-Whitney U-test was used for comparisons between two groups. The Kruskal-Wallis test with Dunn's test for multiple comparisons was used for comparisons between three or more groups. The association between two variables was analyzed by Spearman's correlation and graphically represented by scatter plot. Statistical analysis was performed by IBM SPSS version 25 (Armonk, NY, USA). P-values with a two-sided α < 0.05 were considered statistically significant.   An overview describing the number of patients who developed a rejection and received anti-rejection therapy is given in Figure 1.

Numbers of cTfr and cTfh in Patients With Anti-HLA Antibodies and Donor-Specific Antibodies
In this study, anti-HLA antibodies were measured for all patients at the time of blood sampling. Out of 211 patients, 34 (16.1%) patients had detectable anti-HLA antibodies, of which 11 (5.2%) were DSA positive. The association between the numbers of cTfr and cTfh cells with the presence of anti-HLA antibodies or DSA was analyzed. No significant differences were observed between absolute numbers of cTfr, cTfh, and cell ratios (Figure 4) in patients with anti-HLA antibodies or DSA at the time of blood sampling.

cTfr and cTfh in Patients With a History of Rejection and the Effect of Anti-rejection Therapy on Their Cell Numbers
To define whether alloreactivity and subsequent anti-rejection therapy affect cTfr and cTfh cell numbers, we first grouped the samples according to the rejection history and later according to the type of anti-rejection therapy (Figure 1). The 211 patients were divided into a group with a history of rejection (n = 5 presumed, treated rejection, and n = 44 BPR) and a group with no previous rejection (n = 162). Compared to patients without rejection, patients with a history of rejection had 43.9, 39.4, and 39% lower numbers of cTfr, PD-1 + cTfr, and Helios + cTfr cells,

FIGURE 3 | Changes in cTfr and cTfh cell numbers in all kidney transplant patients. The above image shows the absolute numbers (cells/µL) of cTfr cells and its subsets (A), absolute numbers (cells/µL) of cTfh and its subset (B)
, and cell ratios of cTfr to cTfh and ratios of their subsets (C). Boxplots represent the median and interquartile ranges and the whiskers show the 95% confidence interval. KTx, kidney transplantation; HCs, healthy controls. Mann-Whitney-U-test, **p < 0.01; ***p < 0.001; and ****p < 0.0001.
respectively. Furthermore, in the rejection patients cTfh and PD-1 + cTfh cell numbers were 42.8 and 39.2% lower, respectively. Due to the significant decrease of both cTfr and cTfh, all cell ratios in patients with or without a history of rejection were comparable ( Figure 5A). Additionally, we examined whether there were any significant differences in absolute cell numbers between patients with different types of BPR. Patients with rejection were categorized into three subgroups according to their latest biopsy results: TCMR (n = 29), ABMR (n = 15), mixed TCMR and ABMR (n = 5). We found no significant differences in cTfr and cTfh absolute cell numbers between patients with different types of rejection ( Figure S2).
Patients with a history of rejection received antirejection therapy at a median time of 4.9 years before blood sampling. Since there could be a difference in the effects of the various anti-rejection treatments on follicular cell numbers, patients were grouped accordingly. Patients were treated with high-dose pulse intravenous methylprednisolon only, high-dose pulse intravenous methylprednisolon with intravenous immunoglobulin or high-dose pulse intravenous methylprednisolon with alemtuzumab at a median time of 5.1 (range: 2.1-7.1), 2.5 (0.4-6.5), and 5.0 (2.5-6.1) years before blood sampling, respectively. Table 2 provides an overview of the cohort of 49 patients who received anti-rejection therapy before blood sampling. As expected, alemtuzumab treatment had the strongest effect, and it was the only anti-rejection treatment associated with significantly lower numbers of both cTfr and cTfh cells, including their subsets. When compared to patients who had not been treated with anti-rejection therapy, the median absolute numbers of total cTfr in patients treated with previous pulse methylprednisolone decreased by 30.6%, pulse methylprednisolone + IVIg led to 51.2% decrease and pulse methylprednisolone + alemtuzumab led to 85.8% decrease. Absolute numbers of total cTfr, PD-1-cTfr, and Helios + cTfr in patients who had received pulse methylprednisolone + alemtuzumab were significantly lower than cell numbers of patients who had received no anti-rejection treatment ( Figure 5B). Treatment with other anti-rejection therapies did not result in a significant measurable decrease of total cTfr or activated cTfr absolute numbers compared to no anti-rejection treatment.
Similarly, absolute numbers of total cTfh in patients treated with previous pulse methylprednisolone decreased by 42.7%, pulse methylprednisolone + IVIg resulted in 32.6% decrease and pulse methylprednisolone + alemtuzumab resulted in 66.8% decrease compared to total cTfh numbers in patients who had not been treated with anti-rejection therapy. This decrease in absolute numbers of total cTfh was only significant for patients treated with pulse methylprednisolone + alemtuzumab, while PD-1 + cTfh were not significantly lower compared to patients who did not receive antirejection therapy (Figure 5C). The ratio of total cTfr to cTfh and Helios + cTfr to PD-1 + Tfh were also significantly lower in patients who received pulse methylprednisolone + alemtuzumab anti-rejection therapy (Figure 5D). We found a similar decrease in CD3 + , CD4 + , and T conventional cells in patients treated with pulse methylprednisolone + alemtuzumab anti-rejection therapy. Interestingly, Treg cell numbers were not significantly lower in patients who received pulse methylprednisolone + alemtuzumab anti-rejection therapy ( Figure S3).

DISCUSSION
In the present study, we investigated whether immunosuppression affects the absolute numbers of cTfr and cTfh cells and the balance between cTfr to cTfh cells in kidney transplant recipients. We found that tacrolimusbased maintenance immunosuppression results in decreased absolute numbers of cTfr cells, without affecting cTfh cells when compared to healthy individuals. Other studies in renal transplant recipients, demonstrated increased cTfh cell numbers before the development of DSA, in association with ABMR or with the development of anti-HLA antibodies (43)(44)(45). Moreover, decreased numbers of Tfr cells were observed in both kidney transplant tissue and peripheral blood of patients with ABMR (32). The majority of patients in our cohort were on a combined maintenance treatment with tacrolimus and MMF. MMF is a known inhibitor of lymphocyte proliferation (46) and calcineurin inhibitors have an adverse effect on the differentiation and function of regulatory T cells (47)(48)(49). Circulating Tfr cells are primarily derived from Tregs (50) so it stands to reason that if Treg differentiation is inhibited then this may also have a direct effect on cTfr cell numbers.
We also found that patients with anti-HLA antibodies or DSA at 5-7 years post-transplant had similar cTfr and cTfh cell numbers as patients without antibodies. This is somewhat unexpected as other studies have shown elevated numbers of cTfh in patients with anti-HLA antibodies or DSA (45,51). In this study, 13 of the 34 patients with anti-HLA antibodies had a history of rejection and anti-rejection therapy. This suggests that the immunosuppression regimen and/or anti-rejection therapy administered to our patients is able to prevent the expansion of cTfh cells which are commonly associated with the presence of anti-HLA antibodies. However, it may not be sufficient to completely suppress previously formed Tfh and memory B cells from developing anti-HLA antibodies.
As demonstrated in this study, treatment with alemtuzumab led to significantly lower numbers of total and subsets of cTfr and total cTfh cells. After administration of alemtuzumab, T cells repopulation occurs through the clonal expansion of T cells as opposed to repopulation though thymopoiesis, in a process known as homeostatic proliferation (52)(53)(54). Our findings suggest that the homeostatic proliferation of cTfr cells is impaired after treatment with alemtuzumab. This is in line with the findings of Macedo et al., who showed that Tregs are significantly decreased compared to HCs several years after alemtuzumab induction (55). Moreover, the resulting shift in balance between Tfr to Tfh cells after treatment with alemtuzumab may result FIGURE 5 | Absolute numbers and ratios of cTfr, cTfh, and their subsets in healthy controls and patients who received different types of anti-rejection therapy. The above image shows the ratio of cTfr to cTfh in patients with and without rejection (A). Also shown are the absolute numbers (cells/µL) of cTfr cells and its subsets (B), absolute numbers (cells/µL) of cTfh and its subset (C), and cell ratios of cTfr to cTfh and ratios of their subsets (D). Boxplots represent the median and interquartile ranges and the whiskers show the 95% confidence interval. HC, healthy controls; ART, anti-rejection therapy; MP, methylprednisolone; IVIg, intravenous immunoglobulin; Alem, alemtuzumab. Kruskal-Wallis with Dunn's multiple comparisons. **p < 0.01; ***p < 0.001; and ****p < 0.0001.
in increased formation of memory B cells and plasma blasts thereby increasing the risk for development of ABMR. Clinical studies have reported that alemtuzumab induction therapy, is associated with a higher incidence of chronic ABMR in kidney transplant recipients (56,57). It is unclear whether alemtuzumab therapy has an effect on follicular T cells and B cells which reside in SLOs. If these cells remain present in SLOs, T and B cell cross-talk can still result in the formation of memory B cells and antibody-producing B cells. It is also interesting to note that despite successful treatment of rejection, the patients in our cohort have a higher incidence of anti-HLA antibodies and DSA up to 5 years after treatment with anti-rejection therapy. Nonetheless, the results presented in this study should be interpreted in light of the study limitations. The patient cohort described here is heterogeneous and includes patients treated with different modalities of anti-rejection therapy and differences in the timing of treatment. While this heterogeneity is representative of daily clinical practice, it provides a challenge when interpreting our results. Another limitation is the fact that cell counts prior to treatment with anti-rejection therapy are not available. Ideally, serial measurements both before and after treatment with said therapies would provide interesting data. As comparisons with pre-treatment samples are not possible to conclude with certainty whether cTfh cell numbers are less affected by anti-rejection therapy or whether there is a faster repopulation of this subset compared to that of cTfr cells. And finally, the challenges in acquiring sufficient numbers of follicular T cells in transplant recipients results in a lack of assays to describe the functionality of these cells after treatment with maintenance and anti-rejection therapy. The use of a mouse animal model might be valuable method to study the mechanisms of follicular T cells in vivo. Moreover, by using a mouse monoclonal antibody against the CD52 molecule it would also be possible to study the effect of alemtuzumab on follicular T cells and B cells present both in the circulation and in SLOs.
In conclusion, compared to healthy controls, numbers of cTfr cells in peripheral blood of kidney transplant recipients on tacrolimus-based therapy are significantly lower up to 7 years after transplantation. Furthermore, previous treatment with alemtuzumab has a significant and long-lasting effect on the numbers of these cells.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by METC Erasmus Medisch Centrum Rotterdam. The patients/participants provided their written informed consent to participate in this study.

ACKNOWLEDGMENTS
We are grateful for the contributions of our lab members.