Systemic immunoregulatory and proteogenomic effects of tacrolimus to sirolimus conversion in liver transplant recipients

Authors

  • Josh Levitsky,

    Corresponding author
    1. Division of Gastroenterology and Hepatology, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL
    2. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    • M. D., M. S., Division of Gastroenterology and Hepatology, Comprehensive Transplant Center, Northwestern University Feinberg School of Medicine, 676 North St. Clair Street, 19th Floor, Chicago, IL 60611
    Search for more papers by this author
    • fax: 312-695-0036

  • James M. Mathew,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    2. Department of Microbiology-Immunology, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Michael Abecassis,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Anat Tambur,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Joseph Leventhal,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Dhivya Chandrasekaran,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Nancy Herrera,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Patrice Al-Saden,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Lorenzo Gallon,

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Anmaar Abdul-Nabi,

    1. Department of Pathology, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Guang-Yu Yang,

    1. Department of Pathology, Northwestern University Feinberg School of Medicine, Chicago, IL
    Search for more papers by this author
  • Sunil M. Kurian,

    1. The Scripps Research Institute, Department of Molecular and Experimental Medicine, La Jolla, CA
    Search for more papers by this author
  • Daniel R. Salomon,

    1. The Scripps Research Institute, Department of Molecular and Experimental Medicine, La Jolla, CA
    Search for more papers by this author
  • Joshua Miller

    1. Comprehensive Transplant Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL
    2. Jesse Brown VA Medical Center, Chicago, IL
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

Immunosuppression (IS) withdrawal from calcineurin inhibitors is only possible in ∼20% of liver transplant recipients. However, mammalian target of rapamycin inhibitors (e.g., sirolimus; SRL) appear to be more immunoregulatory and might promote a tolerant state for withdrawal. Our aim was to determine whether systemic (i.e., blood, marrow, and allograft) signatures of immunoregulation are promoted by conversion from tacrolimus (TAC) to SRL. We therefore performed the following serial assays before and after SRL conversion in liver transplant recipients to test for enhanced markers of immunoregulation: (1) flow-cytometry immunophenotyping of peripheral blood mononuclear cells (PBMCs) and bone marrow aspirates for regulatory T cells (Tregs) (e.g., CD4+CD25+++FOXP3+) and regulatory dendritic cells (DCregs) (immunoglobulin-like transcript 3+/4+); (2) liver biopsy immunohistochemical staining (e.g., FOXP3:CD3 and CD4:CD8 ratios) and immunophenotyping of biopsy-derived Tregs after growth in culture; (3) effects of pre- versus postconversion sera on Treg generation in mixed lymphocyte reactions; (4) peripheral blood nonspecific CD4 responses; and (5) peripheral blood gene transcripts and proteomic profiles. We successfully converted 20 nonimmune, nonviremic recipients (age, 57.2 ± 8.0; 3.5 ± 2.1 years post–liver transplantation) from TAC to SRL for renal dysfunction. Our results demonstrated significant increases in Tregs in PBMCs and marrow and DCregs in PBMCs (P < 0.01) after conversion. In biopsy staining, FOXP3:CD3 and CD4:CD8 ratios were significantly higher after conversion and a number of biopsy cultures developed new or higher FOXP3+ cell growth. Nonspecific CD4 responses did not change. Both pre- and postconversion sera inhibited mixed lymphocyte reactions, although only TAC sera suppressed Treg generation. Finally, 289 novel genes and 22 proteins, several important in immunoregulatory pathways, were expressed after conversion. Conclusions: TAC to SRL conversion increases systemic Tregs, DCregs, and immunoregulatory proteogenomic signatures in liver transplant recipients and may therefore facilitate IS minimization or withdrawal. (HEPATOLOGY 2013)

See Editorial on Page 1

Life-long immunosuppression (IS) is generally required after liver transplantation (LT). With the advent of calcineurin inhibitors (CNIs), rejection rates have declined, yet toxicity resulting from CNI therapy has led to long-term adverse outcomes.1 Complete IS withdrawal (i.e., operational tolerance) would be ideal, although this has, thus far, been possible in only ∼20% of LT recipients.2 The inability to immunologically predict successful IS withdrawal has obligated long-term CNI maintenance at therapeutic doses, despite toxicities.

The identification of specific cell populations and pathways responsible for immunoregulation may give clues toward achieving tolerance in LT. Tolerance develops initially by the interaction of antigen-specific T cells with unique thymic antigen-presenting cells (APCs) or regulatory dendritic cells (DCregs), respectively resulting in either clonal deletion, anergy, or an active immunoregulatory process.3 Such DCregs are characterized by high surface expression of cluster of differentiation (CD)123 and/or immunoglobulin-like transcripts (ILTs) (e.g., ILT3 or ILT4) that inhibit antigen presentation (i.e., reflecting immunoregulation).4, 5 As mentioned above, this interaction can lead to the generation of regulatory T cells (Tregs) (e.g., CD4+CD25high) that migrate peripherally to control immune responses. These Tregs typically express an intracellular protein, forkhead box protein 3 (FOXP3), which blocks the transcription of T-cell activation molecules, such as interleukin (IL)-2, and the expression of CD127.6, 7 Moreover, gene transcripts and protein expression patterns (i.e., antibodies as well as circulating and cell proteins), as markers for immunoregulation, may also provide a window into the tolerant state.

Thus, there is strong interest in cellular (i.e., Treg and DCreg), genomic, and proteomic assays to assess immunoregulation and predict more reliably who might achieve IS withdrawal.8, 9 Because the liver appears to be the most tolerogenic transplanted organ, high numbers of such cells residing in the liver may be regulatory and protect against rejection. Similarly, the bone marrow, as a formative organ of the immune system, may also contain localized Treg niches that enhance immunoregulation.10-12 Thus, assessing tolerogenic signatures not only in the blood, but also in the allograft and bone marrow may further differentiate regulatory from alloreactive states.

In addition to the need for predictive assays, there have been few efforts to modulate the immune system to induce tolerance other than simultaneous hematopoietic cell and solid organ transplantation,13 a risky approach in patients with severe end-stage liver disease undergoing LT. One potential reason for the low efficacy of IS withdrawal is CNI therapy itself, because most patients in such studies were being treated with CNIs at withdrawal.2 CNIs may hinder immunoregulation by inhibiting IL-2 production needed for Treg development and also by not inhibiting dendritic cell (DC) maturation.14, 15 In contrast, molecular target of rapamycin (mTOR) inhibitors, such as sirolimus (SRL), might facilitate immunoregulation by increasing Treg percentages (i.e., does not block IL-2 production) and inhibiting DC maturation and function.16-18 We previously demonstrated that LT recipients on SRL monotherapy had significantly higher Treg percentages than those on tacrolimus (TAC) monotherapy.19 Thus, conversion to SRL may both minimize CNI toxicity20 and facilitate IS withdrawal by promoting immunoregulation.

We therefore hypothesized that prospectively studied LT recipients converted from CNI to SRL monotherapy would display enhanced systemic cellular, functional, and proteogenomic immunoregulatory markers. This would support larger scale studies utilizing SRL conversion as an intermediate step toward more clinically successful IS withdrawal in LT.

Abbreviations

ALT, alanine aminotransferase; ANOVA, analysis of variance; APC, antigen-presenting cell; ApoC-IV, apolipoprotein C4; BUN, blood urea nitrogen; ATP, adenosine triphosphate; CD, cluster of differentiation; CNIs, calcineurin inhibitors; CTLA-4, cytotoxic T-lymphocyte-associated antigen-4; DC, dendritic cell; DCregs, regulatory dendritic cells; eGFR, estimated glomerular filtration rate; FITC, fluorescein isothiocyanate; FOXP3, forkhead box protein 3; GFR, glomular filtration rate; HCV, hepatitis C virus; HgA1C, hemoglobin A1C tests; HLA, human leukocyte antigen; IHC, immunohistochemistry; IL, interleukin; ILT, immunoglobulin-like transcript; IS, immunosuppression; LT, liver transplantation; mAbs, monoclonal antibodies; MAPs, multianalyte profiles; MDRD, modified diet in renal disease; MIP-1α, macrophage inflammatory protein 1-alpha; MLR, mixed lymphocyte reaction; mTOR, molecular target of rapamycin; NK, natural killer; PARC, pulmonary and activation-regulated chemokine; PBMC, peripheral blood mononuclear cell; PE, phycoerythrin; PerCP, peridinin-chlorophyll protein complex; SRL, sirolimus; TAC, tacrolimus; TFF3, trefoil factor 3; TGF-β, transforming growth factor beta; Treg, regulatory T cell; TRAIL-R3, tumor necrosis factor–related apoptosis-inducing ligand receptor 3; VCAM, vascular cell adhesion molecule.

Patients and Methods

Clinical Protocol and Study Procedures.

The study was a prospective investigation of the changes in immunoregulatory markers in the blood, bone marrow, and liver allograft in recipients converted from TAC monotherapy to SRL monotherapy for clinical indications (e.g., TAC toxicity). Inclusion criteria were as follows: age ≥18 years; ≥6 months post-LT; TAC monotherapy ≥1 month before SRL monotherapy conversion for nephrotoxicity (glomular filtration rate [GFR] 30-60 cc/min by modified diet in renal disease [MDRD]) or other indication; ≥6 months without a rejection episode; no lymphocyte depletion therapy for ≥1 year; normal liver-function tests; and no rejection or fibrosis on preconversion liver biopsy. Exclusion criteria were as follows: previous liver or multiorgan transplant; previous immune or viral liver disease unless hepatitis C virus (HCV) RNA was undetectable; proteinuria (≥0.5 g/day); estimated glomerular filtration rate (eGFR) ≤30 cc/min; ≥2 rejections post-LT; history of hepatic artery thrombosis; hematological abnormalities or severe hypertriglyceridemia; active infection or malignancy; and inadequacy for follow-up. All patients signed informed consent and were followed for 7 months after SRL conversion. The protocol conformed to the Declaration of Helsinki guidelines and was approved by the Northwestern Institutional Review Board (Northwestern University Feinberg School of Medicine, Chicago, IL).

History and physical exams, complete blood counts, comprehensive metabolic panels, fasting lipids, hemoglobin A1C tests (HgA1C), and spot urine protein:creatinine ratios were performed before and 3 and 6 months after conversion. Bone marrow aspirations and percutaneous liver biopsies were performed once before and 6 months after conversion. For conversion, SRL at 2 or 3 mg (< or ≥100 kg body weight) daily was initiated with subsequent weekly SRL trough-level monitoring. When these reached ≥5 ng/mL, TAC was discontinued followed by weekly laboratory tests and SRL trough levels (goal, 5-8 ng/mL) for 1 month, then monthly. Prospective liver- and renal-function tests, lipid levels, urine protein:creatinine ratios, and any new SRL toxicities were recorded.

Assay Methods.

Peripheral blood

Treg immunophenotyping (twice before conversion and 3, 4, 6, and 7 months after conversion): Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized samples on Ficoll-Hypaque gradients. Tregs were enumerated utilizing extracellular immunofluorescent staining with CD3-FITC (fluorescein isothiocyanate), CD4-PerCP (peridinin-chlorophyll protein complex), CD8-PerCP, CD25-APC, and CD127-FITC (BD Biosciences, San Diego, CA). After fixation and permeabilization, the cells were washed and incubated with anti-human FOXP3-PE (phycoerythrin) or rat immunoglobulin G2a-PE isotype control (eBioscience, San Diego, CA) (21, 22). Samples were acquired using the FACSCalibur flow cytometer (BD Biosciences) and were analyzed by gating based on CD4+CD127CD25+ expression against FOXP3 to calculate the percentage of Tregs.

Cell subset analysis: Peripheral blood was also labeled with monoclonal antibodies (mAbs) for T-cell subsets (e.g., CD3, CD4, CD8, and CD25), B cells (CD19), monocytes (CD14), and natural killer (NK) cells (CD56). After red blood cell lysing, samples were acquired on the FACSCalibur and the absolute cell-subset numbers/μL blood were calculated.

DC surface markers (twice before conversion and 3, 4, 6, and 7 months after conversion): PBMC isolated by Ficoll-Hypaque gradients were incubated with FITC-labeled mAbs to lineage markers (e.g., CD3, CD14, CD16, CD19, CD20, and CD56) for negative selection and markers to distinguish between monocytoid (CD11c) and plasmacytoid (CD123) DCs (Beckman-Coulter, Miami, FL).19, 23 Cells were further stained with CD205-PE and CD83-PC5 as DC antigen uptake/presentation markers or ILT3-PC5 and ILT4-PE as tolerogenic DC markers. Four-color multiparameter flow cytometry was performed using a Coulter FC500 instrument (Beckman-Coulter) compensated with single fluorochromes.

Functional assays (once before and once 6 months after conversion): The Immune Cell Function Assay (Cylex Inc., Columbia, MD) assay was performed per the manufacturer's instructions, detecting CD4 responses by adenosine triphosphate (ATP) production in whole blood after 18 hours of incubation with phytohemagglutinin stimulation. In the Treg-MLR (mixed lymphocyte reaction) assay, using healthy human leukocyte antigen (HLA)-typed volunteer PBMCs, responding cells were stimulated with X-irradiated HLA2-DR matched stimulating allogeneic cells.21, 22 To these cultures, LT recipient sera, containing trough levels of TAC (preconversion) versus SRL (postconversion), were added and compared with the addition of similar volumes of human AB sera (Invitrogen, Carlsbad, CA) versus media controls. After 7 culture days, lymphoproliferation was assessed by tritium-labeled thymidine (3H-TdR) incorporation and immunophenotyping was performed for CD3, CD4, CD8, CD25, and FOXP3 markers. Stimulation indices were calculated from the counts per minute measured with a beta counter.24

Gene-expression microarrays and protein multianalyte profiles (once before and once 6 months after conversion): Peripheral blood was collected in PaxGene tubes (Qiagen, Valencia, CA) for gene microarrays and Baxter PPT tubes for multianalyte profiles (MAPs).25, 26 RNA was extracted using the PaxGene blood RNA kit. Whole blood human genome profiling was performed with Affymetrix GeneChip 1.0 ST arrays (Affymetrix, Santa Clara, CA), following standard protocols. Plasma proteomics were performed using a proprietary Luminex Bead technology testing the 189 protein Human DiscoveryMAP v1.0 (Rules Based Medicine, Austin, TX).

Bone marrow aspirates

Treg and DC immunophenotyping (once before and once 6 months after conversion): Under local anesthesia as outpatients, 10 cc of marrow were aspirated from the posterior iliac crest using a 16-gauge needle. After 5 minutes of pressure, normal daily activities resumed. Leukocytes were isolated by Ficoll-Hypaque gradients and characterized by flow cytometry for Treg and DC subsets identical to the peripheral blood methods, as described above.

Liver biopsy

Routine histolog;: Treg immunophenotyping by immunohistochemical staining and after culture (once before and once 6 months after conversion): A 2-cm core was obtained for hematoxylin and eosin, trichrome, and immunohistochemistry (IHC). IHC staining of formalin-fixed tissue was performed with streptavidin/biotin/peroxidase using dual-staining antibodies to FOXP3, CD3, CD4, and CD8.27 The number of CD3- and FOXP3-positive and CD4- and CD8-positive lymphocytes were counted in a 400× power field. Ratios of FOXP3:CD3 and CD4:CD8 were calculated, and an average of three portal-tract ratios were recorded. A second core was obtained for flow immunophenotyping after 14 days of culture in media (50 U/mL of recombinant IL2 + 50% MLR supernatant) that reliably expands cells already activated in vivo.28, 29

Statistical Analysis.

Pre- versus postconversion measurements of immune assays (e.g., PBMC, marrow, and biopsies) and clinical outcomes were performed using the appropriate paired analysis (i.e., paired t test and Wilcoxon's signed-rank test) or the chi-squared/Fisher's exact test for continuous or categorical measures, respectively. For microarray and MAP comparisons, P values were calculated using a two-way analysis of variance (ANOVA) model by the method of moments,30 using the Partek Genomics Suite (Partek Inc., St. Louis, MO). A false discovery rate correction of ≤10% (q-values) was used for the proteomic data. A paired ANOVA was used for the gene-expression changes, because the samples represented two time points from the same individual. Analyses were performed using SAS 9.2 software (SAS Inc., Cary, NC).

Results

Patient Characteristics.

Twenty-seven LT recipients were initially considered candidates for TAC to SRL conversion because of renal dysfunction. Two were excluded before conversion: 1 because of elevated alanine aminotransferase (ALT) at screening and 1 with interface hepatitis on the preconversion biopsy. Five were excluded as they were converted back to TAC within 1 month after SRL conversion because of cost (n = 1), SRL intolerability (1 foot ulcer and 1 nausea), or mild rejection on biopsy (n = 2, each resolved with TAC reversion). Other than biopsy IHC staining in the 2 with rejection, these 5 patients were withdrawn from the study and followed clinically because it was not considered necessary (i.e., no longer on SRL) or ethical to continue the serial sample collections. Thus, 20 were successfully converted and completed the study (Table 1).

Table 1. Preconversion Clinical Characteristics (n = 20)
 Mean ± SDMedian (Range)
  • *

    Other immunosuppressive therapy (e.g., prednisone and mycophenolate mofetil) was discontinued ≥3 months before study enrollment (see inclusion criteria).

  • Abbreviations: SD, standard deviation; M, male; F, female.

Age (years)59.2 ± 7.958 (44-71)
Gender13 M; 7 F 
Caucasian race (%)19 (95) 
Cause of liver disease (%)  
 HCV (non-viremic)4 (20) 
 Alcohol7 (35) 
 Cryptogenic/fatty Liver8 (40) 
 α-1 antitrypsin deficiency1 (5) 
Months post-LT42.8 ± 25.140.8 (9.6-104.6)
Months on TAC monotherapy before conversion24.4 ± 26.739.6 (1.5-100.1)
Mean trough TAC level at conversion (ng/mL)6.3 ± 1.55.6 (2.3-7.7)
Other immunosuppressive therapy* (%)  
 Induction at LT (alemtuzumab)5 (25) 
 Number on prior prednisone post-LT15 (75) 
 Months on prednisone post-LT5.4 ± 3.26 (3-12)
 Number on mycophenolate mofetil  post-LT14 (70) 
 Months on mycophenolate mofetil  post-LT20.7 ± 1616.9 (6.1-59.4)

Clinical and Histological Outcomes.

SRL was generally well tolerated. There were no infectious complications. Side effects (e.g., 4 diarrhea, 3 anemia, 4 lower extremity edema, and 1 rash) resolved by targeting lower SRL troughs (∼5 ng/mL). Although renal parameters (e.g., eGFR by MDRD equation, blood urea nitrogen [BUN], creatinine, and potassium levels) improved, urine protein:creatinine ratios increased with SRL conversion (Table 2). Biochemical changes included minor decreases in bilirubin and increases in ALT. Significant increases in low-density lipoproteins and triglycerides occurred. Liver biopsies did not demonstrate significant histological changes, other than mild steatosis and increased portal lymphocytes later characterized as staining FOXP3+ (see below).

Table 2. Clinical and Histological Outcomes of TAC to SRL Conversion (n = 20)
 Preconversion (TAC)Postconversion (SRL)P Value
Blood pressure (mmHg)   
 Systolic130.9 ± 12.8135.4 ± 20.70.15
 Diastolic76.5 ± 8.478.8 ± 10.80.25
Complete blood count   
 White count (103/μL)5.0 ± 1.84.7 ± 1.80.26
 Hemoglobin (g/dL)12.9 ± 1.412.9 ± 1.70.49
 Platelet (103/μL)159.7 ± 56.1164.3 ± 58.70.46
Renal/electrolyte   
 eGFR (mL/min/1.73 m2) by MDRD45.9 ± 7.949.5 ± 10.80.02
 BUN (mg/dL)26.6 ± 11.322.9 ± 9.20.001
 Creatinine (mg/dL)1.49 ± 0.31.42 ± 0.20.03
 Urine protein:creatinine ratio0.08 ± 0.040.29 ± 0.460.05
 Serum potassium (mEq/L)4.6 ± 0.64.3 ± 0.460.03
Liver function   
 Total bilirubin (mg/dL)0.76 ± 0.230.6 ± 0.250.0001
 ALT (units/L)25.8 ± 11.431.4 ± 14.60.001
 Alkaline phosphatase (units/L)73.2 ± 33.076.7 ± 30.10.16
HgA1c (%)5.4 ± 0.65.6 ± 0.60.15
Mean trough SRL level (ng/mL)n/a6.4 ± 2.1n/a
Cholesterol (mg/dL)   
 Total157 ± 34.1179.2 ± 40.80.07
 Low-density lipoprotein92.8 ± 32.8114.5 ± 34.50.03
 High-density lipoprotein39 ± 8.940.5 ± 11.90.51
 Triglycerides126.1 ± 73.8180.2 ± 98.70.0003
Body mass index (kg/m2)29.9 ± 5.730.6 ± 5.30.62
Liver histology (%)   
 Normal10 (50)6 (30)0.33
 Steatosis <30%5 (25)5 (25)0.1
 Mild portal lymphocytes5 (25)7 (35)0.73
 Steatosis <30% + mild portal lymphocytes04 (20)0.11

Immunophenotyping: Peripheral Blood, Bone Marrow, and Allograft.

To clarify the rationale for these assays, changes in Treg and DCreg percentages after SRL conversion were assessed because high percentages of these cells were formerly reported in tolerant LT recipients.5, 8, 9, 31 Also, in previous studies, we have safely and repeatedly performed outpatient marrow aspirations, demonstrating the role of bone marrow cells in controlling antidonor immune responses.10, 11, 32 Recently, bone marrow Tregs have been shown to establish an immunoregulatory niche in supporting stem cells and protecting against immune injury.12 Because SRL inhibits DC function in vitro, it was also questioned whether SRL conversion might affect the percentage of ILT3, ILT4, and CD123, all of which are markers of regulatory DCs.5, 18, 33 We therefore measured bone marrow immunophenotypes (e.g., Treg and DCreg) before and after conversion to determine whether changes similar to those observed in PBMCs occurred. In addition, liver biopsy IHC staining has been utilized in previous studies demonstrating high Treg numbers in tolerant LT recipients.8, 27 We therefore performed both liver biopsy IHC staining and allograft culture immunophenotyping, previously validated approaches,8, 27-29 to characterize the percentage of Tregs residing within the graft before and after conversion.

Peripheral blood and bone marrow immunophenotyping

In both the PBMC and marrow aspirates, percentages of CD4+CD25+FOXP3+ and CD4+CD25highFOXP3+ phenotypic Tregs significantly increased after SRL conversion (Fig. 1A; Supporting Table 1). PBMC CD3+ (total T cells), CD14+ (monocytes), and CD56+ (NK cells) cell numbers all statistically decreased after conversion, although the absolute changes in number/uL whole blood were minor (Supporting Table 1). Also, the percentage of DC (CD123+ and CD11c+) expressing ILT3 and ILT4 increased significantly in the peripheral blood (P < 0.01; Fig. 1B), but not in the bone marrow (Supporting Table 2). Other than decreased total HLA-DR+ cells and DCs and increased CD11c+/11c+83+ cell percentages, no differences were observed in other DC subsets (Supporting Table 2).

Figure 1.

Increase in phenotypic regulatory T cells and regulatory DCs with TAC to SRL conversion. Immunophenotyping flow cytometry was performed on peripheral blood and bone marrow aspirate specimens before (0, horizontal axis) and after (means of 12 and 14 weeks and of 24 and 26 weeks for PBMC, horizontal axis) conversion from TAC to SRL. Note the significant increase in the percentage of phenotypic Tregs (e.g., CD4+CD25+FOXP3+ and CD4+CD25highFOXP3+) in the peripheral blood and marrow (A), as well as the number of peripheral blood CD123+ILT3+ILT4+ and CD11c+ILT3+ DCregs (B), after SRL conversion. *P < 0.05 and **P < 0.01, compared to time point 0 (preconversion on TAC). More details are given in Supporting Tables 1 and 2.

Allograft immunophenotyping

The ratio of FOXP3:CD3 positive cells on IHC slide staining increased significantly after SRL conversion (0.19 ± 0.1), compared to preconversion (0.11 ± 0.1; P = 0.01) or rejection controls (2 from this study and 5 randomly selected from our pathology database: 0.09 ± 0.01; P = 0.005). A representative example is shown in Fig. 2. The ratio of CD4:CD8-positive cells also increased significantly (1.2 ± 0.5), compared to preconversion (0.86 ± 0.2; P = 0.02) or those with rejection (0.9 ± 0.1; P = 0.01). For the liver biopsy cultures, there was significant variability in cell growth, precluding an appropriate pre- and postconversion statistical analysis (Supporting Table 3). Of the biopsies that had growth pre- and postconversion, two had decreases, six had no change, and three had increases in Treg percentages. In the others, two lost growth, but seven had new Treg growth after conversion, perhaps suggesting a trend toward increased intragraft Tregs after culture. However, given the variability of culture growth, these data are not fully conclusive.

Figure 2.

Increased FOXP3+ cells in liver biopsies after TAC to SRL conversion. Immunohistochemical staining was performed on liver biopsy specimens before and after conversion from TAC to SRL (representative example shown). Note the increase in the number of FOXP3+ cells (in dashed-line circles) after SRL conversion.

Functional Assays.

There has been recent interest in functional assays (Cylex ImmuKnow; Cylex, Inc.) assessing nonspecific CD4 responses to distinguish alloreactive from immunosuppressed states.34 In the present study, mean ATP values did not change after SRL conversion (266 ± 132 to 274 ± 149 ng/mL; P = 0.15), suggesting that SRL conversion and Treg generation did not appear to lead to nonspecific over-immunosuppression.

We have also recently reported on a novel in vitro immune monitoring assay in humans (the Treg MLR) demonstrating favorable immunoregulatory effects of SRL versus TAC when added directly to MLR cultures.21, 22 As another functional measure, we therefore questioned whether the addition of patient sera containing TAC versus SRL and, possibly, other resulting regulatory molecules might suppress lymphoproliferation and enhance Treg generation.21, 22 Both pre- and postconversion sera equally suppressed MLR lymphoproliferation (stimulation indices) below media controls (n = 13; P < 0.05) (Fig. 3A). However, TAC sera also suppressed CD4+CD25highFOXP3+ cell generation (n = 13; P < 0.01) (Fig. 3B), whereas SRL sera did not.

Figure 3.

Effect of patient preconversion (TAC) and postconversion (SRL) sera on lymphoproliferation and Treg generation in MLR. First, 5 × 105 carboxyfluorescein succinimidyl ester–labeled responding PBMCs from healthy volunteers were cultured with 5 × 105 PKH26-labeled X-irradiated HLA-2DR matched stimulator cells in the presence of the indicated heat-inactivated serum samples (final, 50%). Standard 3H-TdR incorporation assays were also performed with 1 × 105 each of the responders and stimulators in the presence of the indicated sera. The results from peak responses on day 7 are depicted in (A). In parallel, flow-cytometric analysis was also performed on day 7 and is depicted in (B). Both preconversion (TAC) and postconversion (SRL) sera inhibited lymphoproliferation by 3H-TdR incorporation (A; *P < 0.05), but only the preconversion sera (TAC) significantly inhibited the generation of new CD4+CD127CD25HighFOXP3+ Tregs (**P < 0.01) (B). To minimize variation between experiments, the data are depicted as percentage of medium control. Only 13 MLRs were performed because the amount of pre- and post-conversion sera available from the remaining seven patients was inadequate to conduct these experiments.

Proteogenomic Signatures.

Genomic, proteomic, and cytokine signatures may have the potential to predict tolerance.9, 35 In previous reports, transcripts for cell-proliferation arrest proteins and T- and NK-cell receptors have been identified as putative LT tolerance signatures, correlating with increased circulating Tregs. We examined whether similar signatures of immunoregulation might be also observed after SRL conversion. In the present study, several gene transcripts (n = 288; Supporting Table 4) and plasma proteins (n = 22; Table 3), many involved in immunoregulatory pathways, were found to be significantly different after SRL conversion (P < 0.005). Within the heat map displayed in Fig. 4 were up-regulated transcripts of FOXP3, CD25, and cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4), transforming growth factor beta (TGF-β), and CD4 and down-regulated transcripts of chemokine (C-C motif) receptor 3, apolipoprotein C4 (ApoC-IV) and collagen type IV (Supporting Table 4). Also, a number of proteins known to be involved in lymphocyte and DC activation (e.g., IL-3, IL-7, IL-13, macrophage inflammatory protein 1-alpha [MIP-1α], and CD40), lymphocyte trafficking (e.g., vascular cell adhesion molecule [VCAM-1] and pulmonary and activation-regulated chemokine [PARC]), and renal injury (e.g., tumor necrosis factor–related apoptosis-inducing ligand receptor 3 [TRAIL-R3], MIP-1α, and trefoil factor 3 [TFF3]) were down-regulated after conversion (Table 3). The relevance of the changes in the remaining proteins (Table 3) to the present study is not known.

Figure 4.

Heat map of immune/inflammatory gene transcripts (n = 93) before and after TAC to SRL conversion. The bottom half of the heat map (yellow bar on the left) represents the peripheral blood gene transcripts present in patients preconversion (TAC), whereas the top half (red bar on the left) represents the transcripts postconversion (SRL). Below the heat map is the color coding of gene transcripts: red signifies up-regulation, blue down-regulation, and gray unchanged. The color gradation represents the intensity of the signal in log value. More details are given in Supporting Table 4.

Table 3. Proteomic Profile Changes With TAC to SRL Conversion
 Fold Change (Post Verus Pre)P Value
IL-7−8.830.0037
IL-3−3.490.0053
IL-13−2.240.0026
PARC−1.822.15E-06
Ferritin−1.660.0115
TFF3−1.430.0020
CD5−1.430.0104
MIP-1α−1.389.14E-05
Sex hormone-binding globulin−1.360.0066
IL-18−1.320.0049
CD40 antigen−1.310.0095
TRAIL-R3−1.250.0008
AXL receptor tyrosine kinase−1.200.0097
VCAM-1−1.160.0107
Complement C31.220.0022
Haptoglobin1.370.0057
Serum amyloid P-component1.380.0001
Cancer antigen 1251.400.0050
Factor VII1.430.0011
Matrix metalloproteinase-31.436.42E-05
Glutathione S-transferase alpha2.070.0020
Pancreatic polypeptide2.650.0068

Discussion

This study complements previous reports supporting the immunoregulatory properties of mTOR inhibitor therapy.16-18, 22, 36 First, Treg populations increased prospectively in all key immune compartments studied (e.g., blood, marrow, and allograft) after SRL conversion. This provides a more robust classification of patients that might be “tolerance prone,” rather than single time-point analysis of peripheral blood.8, 19 Speculatively, these intragraft FOXP3+ cells promoted by SRL might regulate immune responses and facilitate tolerance.8, 37 Second, the serial paired data on monotherapy conversion provide support that these changes were directly caused by SRL. Third, to our knowledge, this study is the first to analyze serial changes in DC profiles and supports SRL therapy promoting tolerogenic DCs.18, 38, 39 Finally, the functional and proteogenomic regulatory signatures coincided with the phenotypic cellular markers of immunoregulation after SRL conversion. Sera from patients on SRL inhibited lymphoproliferation alloreactivity, but did not inhibit Treg generation as did TAC sera, possibly because of the action of SRL itself or other regulatory serum proteins.40 Many of the genes/proteins were up-regulated (immunoregulatory pathways) or down-regulated (kidney injury pathways) by conversion, supporting their combined use as surrogate tolerance signatures.9, 26

Previous reports have demonstrated differences in the effects of CNIs versus mTOR inhibitors on Tregs and DCregs.14, 16, 17, 41 Tacrolimus inhibits cytokines (e.g., IL-2) important in FOXP3 expression and Treg function.7 In contrast, SRL inhibits postactivation signaling (e.g., phosphatidylinositol 3-kinase/mTOR pathways) and does not block IL-2 production or other cascades (e.g., signal transducer and activator of transcription 5) involved in Treg generation.14, 17, 39, 42 Likewise, we demonstrated allo-specific inhibition and Treg generation by SRL versus TAC in vitro (22) and reported higher blood Tregs in LT recipients on SRL versus TAC.19

Moreover, while CNIs have little effect on DC function, SRL impairs DC maturation, generation, and costimulation.18 The resulting immature DCs are less capable of allo-stimulation and more proficient in generating allo-specific Tregs.39, 43 Alternatively, the reverse regulatory process might occur, given that FOXP3+ suppressor T cells can induce ILT3/4 on DCs, rendering them tolerogenic.44 Though our study did not demonstrate an increase in plasmacytoid (CD123+) to myeloid (CD11c+) DC ratios observed previously in tolerant LT patients,5 an increase in negative costimulatory molecules (e.g., ILT3 and ILT4) occurred after conversion. In addition, the plasma costimulatory CD40 protein was reduced, which might also signify impairment of DC allo-activation and enhanced immunoregulation.

Supporting the immunophenotyping data, SRL conversion led to enhanced peripheral blood transcript expression of known Treg markers (e.g., CD4, FOXP3, CD25, and CTLA-4) and Treg-enhancing cytokines (e.g., TGF-β). In addition, proteins involved in lymphocyte responses (e.g., IL-3, IL-7, and IL-13), trafficking and adhesion (e.g., VCAM-1 and PARC), and DC development and costimulation (e.g., MIP-1α and CD40) were down-regulated (Table 3). Interestingly, other genes (e.g., ApoC-IV and collagen type IV) and proteins (e.g., TFF3, factor VII, TRAIL-R3, and MIP-1α) often associated with renal dysfunction were down-regulated, correlating clinically with eGFR improvement. Two (e.g., TFF3 and factor VII) were associated with chronic kidney disease after LT in our recent report.26 This might be merely related to CNI withdrawal, although SRL has antifibrotic effects45 that might improve renal function.

Several limitations in this report need to be mentioned. First, our final enrollment was only 20 patients, although our tolerability was somewhat better than trials in which ∼30% could not tolerate SRL. This is possibly the result of our monotherapy patients being further out from LT and targeted for lower trough levels. Second, many of the biopsies did not grow in culture media designed to expand Tregs, precluding a pre- and postconversion statistical analysis. However, several biopsies had new or higher Treg percentages after conversion. Also, the increased CD4+ and FOXP3+ cells (i.e., putative Tregs) on IHC staining might be more directly indicative of the regulatory cell changes within the allograft. Third, because donor cells were not available, we could not assess the effect of SRL conversion on donor-specific immunoregulation observed in vitro.21 Finally, our preliminary results, particularly the proteogenomic analysis, need to be validated in larger, prospective patient cohorts.

In conclusion, this study is consistent with the notion that CNI to SRL conversion after LT could take advantage of the regulatory properties of SRL and allow more successful subsequent IS minimization and/or full withdrawal.

Ancillary