Cell Development and Peripheral Survival Heme Exporter FLVCR Is

Heme Exporter FLVCR Is Required for T Cell Development and Peripheral Survival Mary Philip, Scott A. Funkhouser, Edison Y. Chiu, Susan R. Phelps, Jeffrey J. Delrow, James Cox, Pamela J. Fink and This information is current as Janis L. Abkowitz of September 28, 2021. J Immunol 2015; 194:1677-1685; Prepublished online 12 January 2015; doi: 10.4049/jimmunol.1402172

http://www.jimmunol.org/content/194/4/1677 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2015/01/09/jimmunol.140217 Material 2.DCSupplemental http://www.jimmunol.org/ References This article cites 44 articles, 11 of which you can access for free at: http://www.jimmunol.org/content/194/4/1677.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

by guest on September 28, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Heme Exporter FLVCR Is Required for T Cell Development and Peripheral Survival

Mary Philip,*,† Scott A. Funkhouser,*,1 Edison Y. Chiu,* Susan R. Phelps,* Jeffrey J. Delrow,‡ James Cox,x Pamela J. Fink,{ and Janis L. Abkowitz*

All aerobic cells and organisms must synthesize heme from the amino acid glycine and the tricarboxylic acid cycle intermediate succinyl CoA for incorporation into hemoproteins, such as the cytochromes needed for oxidative phosphorylation. Most studies on heme regulation have been done in erythroid cells or hepatocytes; however, much less is known about heme metabolism in other cell types. The feline leukemia virus subgroup C receptor (FLVCR) is a 12-transmembrane domain surface protein that exports heme from cells, and it was shown to be required for erythroid development. In this article, we show that deletion of Flvcr in murine hematopoietic precursors caused a complete block in ab T cell development at the CD4+CD8+ double-positive stage, although other lymphoid lineages were not affected. Moreover, FLVCR was required for the proliferation and survival of peripheral CD4+ Downloaded from and CD8+ T cells. These studies identify a novel and unexpected role for FLVCR, a major facilitator superfamily metabolite transporter, in T cell development and suggest that heme metabolism is particularly important in the T lineage. The Journal of Immunology, 2015, 194: 1677–1685.

he role of heme as a prosthetic group in proteins involved regulated carefully (8). The feline leukemia virus subgroup C

in oxygen transport, electron transfer, and catalysis has receptor (FLVCR), a 12-transmembrane domain protein in the http://www.jimmunol.org/ T been long appreciated. Heme is critical for mitochondrial major facilitator superfamily (MFS) (9), was shown by our group oxidative phosphorylation, and heme deficiency impairs assembly to export heme (10). The gene encoding FLVCR is referred to as of the electron chain subunits (1). Heme also has important reg- FLVCR1 in humans and Mfsd7b in mice; to avoid confusion and ulatory functions. It regulates erythroid lineage differentiation by maintain consistency with the existing literature, we refer to the binding transcriptional (2) and translational regulators of globin gene and protein as Flvcr and FLVCR in this article. Conditional synthesis (3). On the organismal level, heme synthesis and the deletion of Flvcr in neonatal or adult mice caused severe anemia circadian clock are reciprocally regulated (4), and heme plays (11), similar to the erythroid failure seen in cats viremic with a role in integrating the internal circadian clock with metabolic feline leukemia virus subgroup C (FeLV-C), in which cell surface states, such as the fasting and fed states (5, 6). expression of FLVCR is inhibited by viral interference (12, 13). by guest on September 28, 2021 Although the enzymatic steps of heme synthesis and degrada- Older studies noted that cats viremic with FeLV-C had thymic tion have been well characterized (Supplemental Fig. 1), less is aplasia and lymphopenia (14), although it was not known whether known about intra- and intercellular heme trafficking (7). Free heme lymphopenia was due to cell-intrinsic loss of FLVCR or secondary causes lipid peroxidation and oxidative damage and must be to chronic viremia and/or anemia. To answer this question, we developed and studied several models in which FLVCR function could be knocked out in lymphoid cells or more specifically in † *Division of Hematology, University of Washington, Seattle, WA 98195; Clinical Re- T cells during development. search Division, Fred Hutchinson Cancer Research Center, Seattle, WA 98109; ‡Genomics Shared Resource, Fred Hutchinson Cancer Research Center, Seattle, WA 98109; xUniversity of Utah Metabolomics Core Facility, Salt Lake City, UT 84132; { Materials and Methods and Department of Immunology, University of Washington, Seattle, WA 98109 Mice 1Current address: Genetics Program, Michigan State University, East Lansing, MI. flox/flox Received for publication August 28, 2014. Accepted for publication December 11, Flvcr ;Mx-cre mice and controls were described previously (11) and 2014. were backcrossed to C57BL/6 mice (The Jackson Laboratory) for 10 generations. C57BL/6, CD45.1 (Pep3b), and Rag12/2 mice (15) were This work was supported by National Institutes of Health Grants K08 CA158069 (to obtained from The Jackson Laboratory. Homozygous Lck-cre and CD4-cre M.P.), R01 HL031823 (to J.L.A.), and R01 AI64318 (to P.J.F.) and by an American mice (16) were obtained from Taconic and crossed to Flvcrflox/flox mice to Society of Hematology Basic Science Fellow Award (to M.P.). The National Institutes of flox/flox flox/flox Health provided additional funding as follows: M.P. was supported by T32 CA009515, generate Flvcr ;Lck-cre and Flvcr ;CD4-cre mice. OT-I (17) and flox/flox flox/flox J.J.D. was supported by P30 CA015704-39, and J.C. was supported by P30 DK072437. OT-II (18) mice were crossed to Flvcr ;Mx-cre mice. OT-I;Flvcr mice were bred to Lck-cre mice to obtain OT-I;Flvcr flox/flox;Lck-cre. All mice The microarray data presented in this article have been submitted to the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/gds) under accession were bred and maintained in a specific pathogen–free barrier facility at the number GSE50202. University of Washington. Animal studies were performed according to protocols approved by the Institutional Animal Care and Use Committee of Address correspondence and reprint requests to Dr. Janis Abkowitz, Division of the University of Washington. Hematology, University of Washington, HSC K136C Box 357710, 1959 NE Pacific Street, Seattle, WA 98195-7710. E-mail address: [email protected] Noncompetitive and competitive transplants The online version of this article contains supplemental material. Flvcr+/+, Flvcr+/2,andFlvcr2/2 mice were generated by treating 6–12-wk- Abbreviations used in this article: BM, bone marrow mononuclear cell; DN, double old Flvcr+/+;Mx-cre, Flvcr+/flox;Mx-cre,orFlvcrflox/flox;Mx-cre mice with negative; DP, double positive; EGFP, enhanced GFP; FeLV-C, feline leukemia virus subgroup C; FLVCR, feline leukemia virus subgroup C receptor; ISP, immature SP; polyinosinic-polycytidylic acid (polyI:C; 0.15 mg i.p., three doses every MFS, major facilitator subfamily; polyI:C, polyinosinic-polycytidylic acid; qRT-PCR, other day; Amersham). Eight or nine days later, bone marrow mononuclear quantitative RT-PCR; SP, single positive; S1P, sphingosine-1-phosphate; WT, wild-type. cells (BMs) from the femurs and/or tibias of polyI:C-treated mice were harvested, and 2.5 3 106 BMs were injected i.v. into sublethally irradiated 2 2 Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 (6.5 Gy) Rag1 / mice. For competitive transplants, Flvcr+/+;Mx-cre, www.jimmunol.org/cgi/doi/10.4049/jimmunol.1402172 1678 FLVCR REQUIRED FOR T CELL DEVELOPMENT

Flvcr+/flox;Mx-cre, Flvcrflox/flox;Mx-cre, and CD45.1 mice were treated Adoptive cell transfer with the same polyI:C protocol as described above. Eight or nine days flox/flox +/+ later, BMs from the femurs and/or tibias of polyI:C-treated mice were Thymocytes from Flvcr ;CD4-cre or Flvcr ;CD4-cre mice were harvested, and 5 3 106 BMs from Flvcr+/+;Mx-cre or control mice were mixed 2:1 with CD45.1 thymocytes and incubated with 10 mM CFSE mixed with 5 3 106 CD45.1 BMs and injected i.v. into lethally irradiated (Invitrogen) in serum-free HBSS for 15 min at 37˚C. The reaction was (11 Gy) Rag12/2 mice or C57BL/6 3 Pep3b F1 (CD45.1/CD45.2) mice. quenched with pure FCS, and the cells were washed twice with serum-free medium; 1.5 3 107 CFSE-labeled mixed thymocytes were injected i.v. 2 2 Blood cell analysis into sublethally irradiated Rag1 / mice (6.5 Gy). A sample of the CFSE- labeled mix also was analyzed by flow cytometry for CD4 and CD8 Mice were bled retro-orbitally into EDTA anti-coagulated microtainer staining so the starting CD45.2:CD45.1 ratio of CD4 single-positive (SP) tubes (Becton Dickinson). Complete blood counts were performed on cells and CD8SP cells could be determined. Recipient mice were sacrificed a Hemavet HV950FS analyzer (Drew Scientific) programmed for mouse at the specified time points, and splenocytes were counted and analyzed by blood. flow cytometry. Flow cytometric analysis and sorting Microarray sample preparation Flow cytometric analysis was performed using a FACSCanto II and an LSR Three separate competitive transplant cohorts were generated as described II (BD Biosciences); cells were sorted using a BD FACSAria (BD Bio- above using lethally irradiated Rag12/2 mice (11 Gy) transplanted with sciences) at the Cell Analysis Facility, Department of Immunology, Uni- polyI:C-treated Flvcr+/+;Mx-cre or Flvcrflox/flox;Mx-cre BMs mixed with versity of Washington. Flow data were analyzed with FlowJo 7.6 (TreeStar). polyI:C-treated CD45.1 BMs. At 12 wk posttransplant, thymocytes were harvested from three recipients/group and pooled, and CD45.2+CD4+CD8+ Abs and reagents double-positive (DP) cells were sorted. This was repeated for the three +/+ 2/2 Fluorochrome-conjugated Abs were purchased from BD Biosciences and cohorts, resulting in three Flvcr and three Flvcr replicates. After Downloaded from eBioscience. sorting, cells were washed twice with cold PBS, and cell pellets were frozen and stored at 280˚C until all samples were collected. RNA was Tissue staining isolated using a QIAGEN RNeasy Plus Mini Kit, per the manufacturer’s instructions, and the yield was determined on a NanoDrop ND-1000 Whole thymus was fixed in 10% buffered formalin. Fixed thymus was spectrophotometer (Thermo Scientific). All samples had an appropriate paraffin embedded, sectioned, and stained with H&E by the Experimental yield (.100 ng total RNA) and were analyzed for RNA integrity using an Histopathology Facility at the Fred Hutchinson Cancer Research Center. Agilent 2100 Bioanalyzer (Agilent Technologies). RNA determined to

be of high quality was converted to cDNA and labeled with biotin for http://www.jimmunol.org/ Quantitative RT-PCR analysis microarray analysis using Ambion’s Illumina TotalPrep RNA Amplification RNA was isolated from sorted cells using an RNeasy Plus Mini Kit kit (Life Technologies). Labeled cDNA was analyzed using the MouseWG-8 (QIAGEN) and reverse transcribed using iScript reverse transcription (Bio- v2.0 Expression BeadChip Kit (Illumina). Sample labeling, array hybrid- Rad), per the manufacturers’ instructions. Quantitative RT-PCR (qRT-PCR) ization, and array scanning were performed by the Genomics Shared Re- was performed on cDNA or genomic DNA using either a TaqMan assay or source at the Fred Hutchinson Cancer Research Center. a SYBR Green assay with gene-specific primers obtained from Integrated Microarray data analysis DNA Technology. The following primer sequences were used: Flvcr (F 59- ATCTGGAACCTGTGCAGAAACA-39,R59-ATTGAATAAAATGCTC- The dataset consisting of all arrays was processed using the Bioconductor CAGTCATGAT-39,Probe59-CCCCTTTGTTCTCCTGCTGGTCAGTTATG- package lumi by using quantile normalization (22). The data were initially 39); Hmox1 (F 59-CTGCTAGCCTGGTGCAAGATACT-39,R59-GTCT- filtered by flagging probes that were below a defined signal “noise floor,” GGGATGAGCTAGTGCTGAT-39,Probe59-AGACACCCCGAGGG- which was calculated as the 75th percentile of the negative-control probe by guest on September 28, 2021 AAACCCCA-39); Alas1 (F 59-TGGTCGGTTTAGCGTCCTC-39,R59- signals within each array. A probe was retained if at least two of three GGGATAAGAATGGGCATCGG-39,Probe59-CGAGTGCCTACCGC- samples in at least one condition were not flagged by the intensity filter. CGCTTC-39); Cox6a2 (F 59-CCAGAGTTCATCCCGTATCAC-39,R59-C- We filtered the data further through the application of a variance filter, AGACATCAAGGGTGCTCATAAC-39); and b-actin (F 59-ACCTTCTAC- using the “shorth” function of the Bioconductor package genefilter. Dif- AATGAGCTGCG-39,R59-CTGGATGGCTACGTACATGG-39,Probe59-T- ferential gene expression was determined using the Bioconductor package CTGGGTCATCTTTTCACGGTTGGC-39). Gene expression was quantified limma (23), and a false discovery rate method was used to correct for as fold-change expression using the Pffafl method (19) with b-actin as the multiple testing (24). Significant differential gene expression was defined + reference gene and either C57BL/6 splenic CD8 T cells or whole thymus as |log2 (ratio)| $ 0.585 (6 1.5-fold), and false discovery rate = 5%. as the reference sample. Genomic Flvcr deletion was quantified from genomic DNA using cloned and purified wild-type (WT) and FlvcrΔexon Statistical analyses 3 DNA to generate standard curves. The following primer sequences were Statistical analyses were performed with Prism version 5.0 (GraphPad), used: genomic WT Flvcr (F 59-ATCTGGAACCTGTGCAGAAACA-39,R using an unpaired two-tailed Student t test. A p value , 0.05 was con- 59-GAGTCTAGTGCTCAAAGGTCACCAA-39, Probe 59-CCCCTTTGTT- sidered statistically significant. CTCCTGCTGGTCAGTTATG-39) and genomic deleted Flvcr (F 59-CCC- ACCCAAGTAGGATACAGAGA-39,R59-GATCATATTCAATAACCCTTA- Accession data ATATAACTTCG-39, Probe 59-ACTTGCAGGACAGACACAAGGCCT-39). The microarray data have been submitted to the Gene Expression Omnibus Rescue transplant experiments database (http://www.ncbi.nlm.nih.gov/gds) under accession number GSE50202. PCR cloning was used to create a DNA fragment containing the human FLVCR-coding region without the stop codon fused in-frame with the enhanced GFP (EGFP) coding sequence (Clontech). FLVCR-EGFP DNA Results was ligated into the pMSCVneo retroviral vector (Clontech) to generate FLVCR is required for T cell development beyond the DP stage pMSCV FLVCR-EGFP. This vector and control pMXIG (20) containing EGFP only were pseudotyped with ecotropic murine retrovirus [Plat-Eco To determine whether FLVCR plays a role in lymphocyte devel- packaging line (21) (Cell Biolabs)]. BMs were harvested from 6–12-wk- opment, we transplanted BMs from polyI:C-treated Flvcrflox/flox; flox/flox old Flvcr ;Lck-cre mice and lineage depleted using biotinylated Mx-cre mice, Flvcr+/flox;Mx-cre mice, or Flvcr+/+;Mx-cre mice lineage-specific Abs and magnetic separation (Dynal). Lineage-depleted into sublethally irradiated Rag12/2 mice, referred to henceforth BMs were prestimulated with 20 ng/ml murine IL-3, 100 ng/ml human 2/2 +/2 +/+ IL-6, and 50 ng/ml murine stem cell factor (all cytokines from PeproTech) as Flvcr , Flvcr ,orFlvcr mice. Cells with the Mx-cre in IMDM with 10% FBS for 2 d and then cultured with viral supernatant transgene express cre-recombinase after exposure to IFN-a or plus cytokines and 4 mg/ml protamine sulfate on viral-preloaded IFN-b, and the double-stranded synthetic RNA polyI:C induces RetroNectin-coated dishes (r-fibronectin CH296; Takara Bio) for 8 h. Vi- IFN production and cre-mediated gene excision in Mx-cre BMs ral supernatants were replaced by fresh media, and the cells were allowed (25). The use of sublethal irradiation permitted reconstitution of to recover overnight; the treatment was repeated once. Transduced cells 2/2 were harvested, and 2 3 106 transduced cells were transplanted into endogenous Rag1 hematopoiesis; therefore, the mice that re- 2 2 sublethally irradiated Rag12/2 mice (6.5 Gy). ceived Flvcr / BMs were not anemic (data not shown), and any The Journal of Immunology 1679

T or B cells present derived from the transplanted marrow. coming CD4+CD8+ DP cells (26). The number of CD4+CD8+ DP Strikingly, we found very few T cells in the blood (Fig. 1A, 1B), thymocytes was unaffected by FLVCR loss (Fig. 1D, 1E); thus, spleen, and lymph nodes (Supplemental Fig. 2A, 2B) of Flvcr2/2 FLVCR is not required for b-selection or subsequent late DN cell mice compared with controls. Both CD4+ and CD8+ T cells were proliferation. similarly affected. B cells were present in normal numbers (Fig. DP thymocytes rearrange their TCRa-chain and undergo posi- 1B), suggesting that FLVCR loss selectively blocked development tive and negative selection to identify thymocytes capable of downstream of the common lymphoid progenitor stage. recognizing peptide–MHC complexes on APCs without excessive We next examined the thymuses in Flvcr2/2 mice and found autoreactivity before ﬁnally differentiating into CD4+ or CD8+ SP that they appeared grossly normal (Fig. 1C, left panel). Early cells (27). CD4+ and CD8+ SP thymocytes were severely de- thymic progenitors derived from bone marrow hematopoietic stem creased in Flvcr2/2 mice (Fig. 1D, 1E). On thymic histology, cells enter the thymus as CD42CD82 double-negative (DN) cells there were fewer medullary regions in Flvcr2/2 mice (lighter that subsequently progress through four substages (DN1–DN4) regions in Fig. 1C, right panel), correlating with the decreased (26). Mice had similar total numbers of DN cells (Fig. 1D, 1E). number of SP cells, typically found in the thymic medulla. Because Rag12/2 mice have DN thymocytes, we examined We examined Flvcr gene expression in sorted normal thymic Flvcr2/2 DN thymocytes obtained from mice that were lethally subsets and found that it peaked during the DN3 stage and then irradiated and transplanted with a mixture of congenically marked declined (Fig. 1F, middle panel). The expression of Flvcr during control or Flvcr-deleted BMs and WT BMs (described in greater thymic development mirrored that of the aminolevulinic acid detail in the next section) to analyze the DN1–DN4 subsets. Al- synthase1 gene (Alas1) (Fig. 1F, top panel), encoding the rate- 2 2 though there was a deﬁcit of Flvcr / DN1 cells, similar numbers limiting enzyme in heme synthesis, consistent with FLVCR’s role Downloaded from of Flvcr-deleted and control cells were present at each of the in maintaining intracellular heme homeostasis. Heme oxygenase 1 subsequent three stages (Supplemental Fig. 2C, 2D). During DN3, (Hmox1), which encodes an enzyme that breaks down heme, was thymocytes that have successfully rearranged the TCRb-chain present at high levels early during thymic development, but it was (b-selection checkpoint) undergo a proliferative burst before be- downregulated after the DN stage (Fig. 1F, bottom panel); thus, http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. FLVCR is required for T cell development beyond the DP stage. (A) Flow cytometric analysis of peripheral blood leukocytes by B220 and CD3 staining (upper panels) and CD4 and CD8 staining (lower panels) of cells within the CD3+ gate. Numbers indicate the percentage of cells within the gate. (B) Absolute numbers of B220+, CD3+, CD3+CD4+, and CD3+CD8+ PBLs. Bars show mean and SEM of three or four mice/group. *p , 0.03, **p , 0.009. (C) Representative gross thymuses from Flvcr+/+, Flvcr+/2, and Flvcr2/2 mice (left panel) and representative thymus sections from Flvcr+/+ and Flvcr2/2 mice stained with H&E (original magnification 3100) (right panel). (D) CD4 and CD8 flow staining of Flvcr+/+ and Flvcr2/2 thymocytes. (E) Enumeration of thymocyte subsets. *p = 0.05, **p = 0.02. (F) Heme regulatory gene expression quantified by qRT-PCR in thymocyte subsets relative to peripheral splenic CD8+ T cells. Mean and SEM from three mice are shown. Data are representative of two (C and F) or three (A, B, D, and E) separate experiments with three or four mice/group. 1680 FLVCR REQUIRED FOR T CELL DEVELOPMENT later-stage thymocytes may rely more on FLVCR to regulate in- CD8SP cells were found in expected proportions (Fig. 2A). tracellular free heme and prevent toxicity. Therefore, the loss of SP T cells was not due to thymic stromal cell FLVCR deficiency. The proportion of Flvcr2/2 DP thymo- Requirement for FLVCR is T cell intrinsic cytes was increased, whereas CD4SP and CD8SP thymocytes were To determine whether the block was due to FLVCR loss in APCs markedly decreased (Fig. 2B). We sorted CD45.2+ (Flvcrflox/flox; (BM and non-BM derived) or stromal cells in the thymus, rather Mx-cre-derived) B220+,Gr-1+, CD4+, and CD8+ peripheral blood than to T cell–intrinsic FLVCR loss, we performed competitive- cells and quantified Flvcr deletion through qRT-PCR. Although repopulation studies, transplanting polyI:C-treated Flvcrflox/flox; donor BMs and donor-derived peripheral B220+ and Gr-1+ blood Mx-cre or Flvcr+/flox;Mx-cre (CD45.2) BMs mixed 1:1 with con- cells were nearly completely deleted (BMs mean 97.5%, SD 2.1, genically marked (CD45.1) WT BMs into lethally irradiated n = 3; B220+ mean 99.3%, SD 0.9, n = 5; Gr-1+ mean 98.1%, Rag12/2 mice. Few CD45.2+ Flvcr2/2 CD4SP or CD8SP cells SD 1.6, n = 5), donor-derived peripheral CD4+ and CD8+ cells were found in recipients, although CD45.1+ WT CD4SP and were only partially deleted (CD4+ mean 68.9%, SD 19.2, n =5; Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 2. The requirement for FLVCR is T cell intrinsic. (A) Representative flow analysis of thymocytes from mice transplanted with a 1:1 mix of Flvcr2/2:CD45.1 BMs with CD4 and CD8 expression shown on gated CD45.1 or CD45.2 thymocytes. (B) Quantification of CD45.2 thymocyte subsets from competitively repopulated (CR) mice that received Flvcr+/+:CD45.1 (black circles) or Flvcr2/2:CD45.1 (red squares). Each symbol represents an individual mouse. *p , 0.0001. (C) Representative flow analysis with TCRgd and TCRb expression shown on gated CD45.2 Flvcr+/+ (left panels) and Flvcr2/2 (right panels) DN cells (top panels) from CR mice. Numbers indicate the percentage of TCRgd+ cells within the gate. Enumeration of the proportion and absolute number of Flvcr+/+ and Flvcr2/2 TCRgd-expressing cells in the thymus (middle panels) and spleen (bottom panels) are shown. (D) Representative flow analysis with CD69 and TCRb expression on Flvcr+/+ and Flvcr2/2 thymocytes (gated on CD45.2) from CR mice with gates indicating successive stages of development (upper left panel). CD4 and CD8 expression on gated populations with the percentage of cells in the population shown in parentheses (remaining upper panels). CD45.2:CD45.1 ratios for the gated populations are shown for individual mice (lower panels). *p = 0.02, **p = 0.001. (E)CD24andTCRb staining on gated CD4SP and CD8SP thymocytes from representative mice (left panel). Ratio of mature SP cells (CD24loTCRbhi)to immature SP cells (ISP, CD24hiTCRblo), which are precursors to the DP stage, is shown in the bar graphs (right panel). Data are representative of four separate experiments with four or five mice/group (A, B,andD) or two separate experiments with three mice/group (C). *p , 0.0001, **p =0.0006. The Journal of Immunology 1681

CD8+ mean 58.7%, SD 12.6, n = 5). Thus, the few peripheral T cells present from polyI:C-treated donor Flvcrflox/flox;Mx-cre BMs likely expanded from a few precursors in which only one Flvcr allele was deleted (resulting in 50% deletion), reinforcing the conclusion that FLVCR is absolutely required for CD4+ and CD8+ T cell development. FLVCR is not required for postnatal gd T cell development Given that FLVCR loss impacted T cell development specifically, we also looked at the numbers and early thymic development of gd T cells. TCRb, g, and d gene rearrangements occur during the DN stage, and thymocytes that express a functional TCRgd receptor on the surface diverge from ab T cells before CD4 and CD8 upregulation and the DP stage (28). We examined the proportion of gd T cells in CD45.2 DN thymocytes from lethally irradiated mice transplanted with polyI:C-treated Flvcrflox/flox;Mx-cre or Flvcr+/flox;Mx-cre (CD45.2) BMs mixed 1:1 with WT (CD45.1)

BMs. The proportion and numbers of gd TCR-expressing DN 2 2 +/+ +/2 FIGURE 3. Positive selection is impaired in Flvcr / thymocytes

thymocytes were similar between Flvcr and Flvcr DN cells Downloaded from (Fig. 2C, top and middle panels). Moreover, the proportion and expressing transgenic TCR. CD4 and CD8 expression on thymocytes from Rag12/2 mice transplanted with Flvcr+/+ and Flvcr2/2 BMs expressing numbers of gd T cells in the spleen also were similar (Fig. 2C, transgenes encoding the MHC class I–restricted OT-I TCR (A) or the MHC bottom panel). Thus, FLVCR is required specifically for ab, and class II–restricted OT-II TCR (B). Enumeration of CD8SP and CD4SP not gd, T cell development. thymocytes [lower left panel in (A) and (B)] and of CD8+Va2+ PBLs in Flvcr-deleted thymocytes fail selection OT-I mice [(A) lower right panel] and CD4+Va2+ PBLs in OT-II mice [(B) lower right panel]. Data are representative of two separate experiments 2/2 http://www.jimmunol.org/ We next analyzed the expression of CD69 and TCRb on Flvcr with three or four mice/group. *p = 0.02, **p , 0.0001. or Flvcr+/+ thymocytes from competitively transplanted mice to more precisely identify the point at which Flvcr2/2 thymocytes failed. Although the proportion of cells in immature TCRloCD69lo be largely diluted by multiple cell divisions, resulting in little (Pop1, DN and early DP) and preselection TCRintCD69lo (Pop2) FLVCR at the DP stage and, consequently, a block in T cell de- populations was similar between Flvcr2/2 and control, the pro- velopment. In CD4-cre mice, cre is expressed in the late DN/early 2/2 int hi DP stage, and there is little subsequent proliferation (30). Therefore, portion of Flvcr thymocytes in the TCRb CD69 (Pop3) flox/flox stage and beyond was greatly diminished (Fig. 2D). Upregulation we predicted that thymic development in Flvcr ;CD4-cre mice of CD69 signifies TCR signaling and positive selection; thus, the would proceed normally through to the SP stage.

flox/flox by guest on September 28, 2021 decrease in CD69+ cells from the DP stage suggested that Flvcr- Flow analysis of Flvcr ;Lck-cre and control thymocytes deleted thymocytes were failing selection. In addition, Flvcr2/2 showed that early Flvcr deletion blocked development beyond the DP stage, and Flvcr flox/flox;Lck-cre mice had decreased numbers of CD4SP and CD8SP cells failed to undergo maturation, as indi- + + 2/2 CD4SP and CD8SP cells (Fig. 4A). Few peripheral CD4 or CD8 cated by downregulation of CD24. The majority of Flvcr flox/flox CD8SP cells, as well as a greater proportion of CD4SP cells T cells were seen in Flvcr ;Lck-cre mice (data not shown). compared with control SP cells, were immature SP (ISP) cells Thus, FLVCR was required for primary T cell development and hi lo not only for T cell development after bone marrow transplantation. (CD24 TCRb ), the precursor stage to the DP stage (Fig. 2E). flox/flox flox/flox To test the hypothesis that the failure of Flvcr-deleted thymo- In contrast to Flvcr ;Lck-cre mice, Flvcr ;CD4-cre cytes to undergo positive selection was due to their inability to mice had relatively normal thymic T cell development (Fig. 4B). Despite normal thymic T cell development, Flvcr flox/flox;CD4-cre rearrange and express a functional TCRa-chain, we crossed + + Flvcrflox/flox;Mx-cre and Flvcr+/+;Mx-cre mice to mice expressing mice had few peripheral CD4 or CD8 T cells (Fig. 4C, upper panel). The few CD8+ T cells present were nearly all CD44hi transgenes encoding an MHC class I–restricted (OT-I) (17) or MHC flox/flox class II–restricted (OT-II) (18) abTCR. We then transplanted BMs (Fig. 4C, lower panel); when Flvcr ;CD4-cre peripheral flox/flox T cells were sorted and genomic Flvcr deletion was assessed, from polyI:C-treated OT-I or OT-II;Flvcr ;Mx-cre mice (re- + 2/2 CD8 T cells showed 49.3% deletion (SD 0.03, n = 3). Thus, the ferred to as OT-I and OT-II Flvcr ) and controls into sublethally hi + flox/flox irradiated Rag12/2 mice. Despite the fact that both TCRs are ca- few CD44 CD8 T cells present in Flvcr ;CD4-cre mice were likely due to homeostatic proliferation of a small number of pable of providing a positive-selection signal in C57BL/6 mice, + 2/2 incompletely deleted CD8 T cells. both OT-I and OT-II Flvcr mice had a marked reduction in the flox/flox number of CD8SP or CD4SP cells in the thymus and the periphery To determine whether Flvcr ;CD4-cre SP cells were ma- (Fig. 3). The failure of transgenic TCR expression to rescue positive turing normally, we examined surface expression of CD24 and CD69 on Flvcr flox/flox;CD4-cre and control SP cells. Few CD24hi selection in Flvcr-deleted thymocytes demonstrates that the effect flox/flox of FLVCR loss is downstream of TCRa rearrangement. cells (ISP, DP precursors) were found among either Flvcr ; CD4-cre or control CD4SP or CD8SP cells (Supplemental Fig. 2E). FLVCR is required for peripheral T cell survival CD69 downregulation is required for sphingosine-1-phosphate To characterize the temporal requirement for FLVCR during T cell (S1P) receptor 1 expression and function (31). The S1P1/S1P development, Flvcr flox/flox mice were crossed to Lck-cre–orCD4- receptor 1 interaction mediates SP cell thymic emigration in re- cre–transgenic mice (16). Lck-cre mice express cre in the T cell sponse to S1P1 gradients from peripheral blood (31). Flvcr flox/flox; lineage during the early DN stage, leading to early DN Flvcr CD4-cre CD4SP and CD8SP cells evinced similar downregulation deletion. The FLVCR protein has a long half-life (29), but because of CD69 compared with controls (Supplemental Fig. 2E). Thus, there is a strong proliferative burst at the end of the DN stage, we deletion of Flvcr late during the DP stage allowed for the pro- predicted that, in Flvcr flox/flox;Lck-cre mice, FLVCR protein would duction of normal numbers of phenotypically normal SP cells. 1682 FLVCR REQUIRED FOR T CELL DEVELOPMENT Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021 FIGURE 4. FLVCR is required for peripheral T cell survival. (A) CD4 and CD8 staining of thymocytes from Flvcr flox/flox;Lck-cre and control mice (upper panel) and absolute cell numbers of thymus subsets (lower panel). *p = 0.007. (B) CD4 and CD8 staining of thymocytes of Flvcr flox/flox;CD4-cre and control mice (upper panel) and absolute cell numbers of thymus subsets (lower panel). (C) Absolute numbers of CD4+ and CD8+ peripheral blood T lymphocytes in Flvcr flox/flox;CD4-cre and control mice (upper panel). Percentages of CD4+ and CD8+ peripheral blood T lymphocytes from Flvcr flox/flox; CD4-cre and control mice with high CD44 surface expression (lower panel). *p = 0.003, **p = 0.0005, ***p # 0.0001. (D and E) CFSE-labeled Flvcr flox/flox; CD4-cre or Flvcr+/+;CD4-cre thymocytes were mixed at a 2:1 ratio with CD45.1 (WT) control thymocytes and adoptively transferred into sublethally irradiated Rag12/2 mice. Four and eight days later, spleens were harvested, and CFSE dilution of transferred thymocytes was analyzed by flow cytometry. Dot plots are gated on CD8+TCRb+ cells (D)orCD4+TCRb+ cells (E). Numbers show the frequency of cells in each gate. (F) CD45.2:CD45.1 ratio of adoptively transferred SP thymocytes and splenic lymphocytes at days 4, 8, and 21 posttransfer. CD45.2:CD45.1 ratios were normalized to the CD45.2:CD45.1 ratio of the input CD4SP cells or CD8SP cells. *p , 0.03, **p , 0.0008, ***p , 0.0001. Data in (A)–(C) are representative of four separate experiments with four to six mice/ group. Data in (D)–(F) are representative of three separate experiments with 10–12 mice/group total.

To determine the fate of Flvcr flox/flox;CD4-cre SP thymocytes in an off-target effect, we performed rescue transplants with the periphery, we adoptively transferred CFSE-labeled CD45.2 Flvcr flox/flox;Lck-cre BMs transduced with vectors encoding full- Flvcr flox/flox;CD4-cre or Flvcr+/+;CD4-cre thymocytes mixed with length FLVCR-EGFP fusion protein (FLVCR-EGFP) or EGFP WT CD45.1 thymocytes into sublethally irradiated Rag12/2 mice. alone. Re-expression of FLVCR led to increased numbers of pe- CFSE analysis of recipient splenocytes was performed (Fig. 4D, ripheral CD4+ and CD8+ T cells compared with EGFP-transduced 4E); each plot shows CFSE dilution by CD45.1 WT competitors controls (Fig. 5A), and thymic analysis showed that, although there above CD45.12 Flvcr+/+ (left panels)orFlvcr2/2 (right panels) were few CD4SP or CD8SP cells among the EGFP2 Flvcr-deleted thymocytes. At days 4 and 8 posttransfer, Flvcr2/2 CD8+ T cells thymocytes, there were normal numbers of CD4SP and CD8SP showed little CFSE dilution compared with cotransferred CD45.1 cells among the EGFP+ thymocytes in the same mouse (Fig. 5B). A WT CD8+ T cells or WT littermate Flvcr+/+ CD8+ T cells trans- recent report suggested that an alternate splice form of Flvcr leads ferred into another host (Fig. 4D, 4F). Similarly, only WT CD4+ to mitochondrial targeting of a partial FLVCR (FLVCR1b) and that T cells had proliferated by day 8 (Fig. 4E, 4F). These results suggest FLVCR1b exports heme from the mitochondria (32). The rescue that FLVCR is required for the DP to SP transition during T cell construct that we used encodes full-length FLVCR without intronic development, as well as for the survival and/or proliferation of sequences and is unlikely to be spliced; these data suggest that full- T cells in the periphery under lymphopenic conditions. length FLVCR is necessary and sufficient for T cell development. Re-expression of FLVCR in deleted BMs rescues T cell Flvcr loss does not cause global transcriptional changes in DP development thymocytes To confirm that the failure of Flvcr-deleted hematopoietic cells to The finding that a heme export protein is required for T cell de- generate T cells was due to the loss of FLVCR expression and not to velopment is surprising because, in contrast to developing erythroid The Journal of Immunology 1683

beyond the DP stage. Through competitive transplantation, we showed that FLVCR’s role in developing thymocytes is T cell intrinsic. Flvcr-deleted thymocytes failed selection, as shown by the failure of FLVCR-deﬁcient thymocytes to upregulate CD69, as well as by the failure of TCR-transgene expression to rescue development of Flvcr-deleted thymocytes. Through the use of mouse strains in which FLVCR could be deleted early or late during thymic T cell development, we showed that early deletion of Flvcr during the DN stage led to a block in T cell development, whereas late deletion of Flvcr during the early DP stage allowed for the formation of CD4 and CD8SP. However, these SP cells were unable to proliferate and/or survive in the periphery. All aerobic cells require heme for oxidative phosphorylation and other key functions; however, excess intracellular free heme is toxic and must be carefully controlled. We analyzed expression of the genes encoding key heme regulators during T cell development and found that expression of Flvcr parallels that of Alas1, supporting the notion that FLVCR may act as a safety valve to export excess

heme. Moreover, HMOX1, which breaks down free heme, is Downloaded from downregulated after the DN stage; therefore, FLVCR may be particularly critical for heme regulation after the DN stage in FIGURE 5. FLVCR re-expression in Flvcr flox/flox;Lck-cre BMs rescues thymocytes. T cell development. (A) Absolute numbers of CD4+ and CD8+ PBLs. Red The finding that FLVCR is required for T cell development bars show numbers of EGFP2 (nontransduced) Flvcr flox/flox;Lck-cre lym- beyond the DP stage and peripheral T cell survival is surprising, + phocytes. Blue bars show numbers of EGFP lymphocytes transduced (tx) because it supports a novel and specific role for heme metabolism http://www.jimmunol.org/ either with EGFP only (open) or FLVCR-EGFP (filled). The bars show the in T cell development. Because heme can regulate gene expression mean and SEM of three or four mice/group. *p = 0.02, EGFP transduced by binding directly to transcription factors, we carried out tran- versus FLVCR-EGFP transduced. (B) Representative flow plots of thy- flox/flox scriptional profiling. One explanation for the paucity of tran- mocytes from a recipient of Flvcr ;Lck-cre BMs transduced with scriptional alterations in Flvcr-deleted DP cells compared with FLVCR-EGFP. CD4 versus CD8 staining of the EGFP2 (nontransduced) controls is that FLVCR loss blocks T cell development by non- thymocytes (left panel). EGFP+ (FLVCR-EGFP transduced) thymocytes from the same mouse (right panel). The data are representative of two transcriptional means. Metabolite profiling and heme measure- experiments with three to five mice/group. ment on whole thymus from deleted or control mice revealed no significant differences (data not shown), and metabolite and heme cells, whose main function is to synthesize heme and hemoglobin, levels from sorted Flvcr-deleted DP and control thymus were by guest on September 28, 2021 there is no known specific role for heme in T cell development below the limit of detection, likely because the DP subset is or function. Heme binds to and inhibits the transcription factor metabolically and transcriptionally quiescent as a result of c-myc BACH2, a negative regulator of BLIMP1, the master regulator of downregulation (37). We are working to develop more sensitive plasma cell differentiation, thus affecting humoral immunity (33). reporter assays to measure heme in vivo and in ex vivo cells to A recent study showed loss of regulatory T cells in Bach22/2 mice determine heme levels during T cell development. Nonetheless, (34), although other CD4+ and CD8+ T cell subsets were present the increased expression of Cox6a2 seen in Flvcr-deleted DP cells in normal numbers. Thus, FLVCR function during T cell devel- suggests that heme levels are increased in Flvcr-deleted thymo- opment is unlikely to be mediated by BACH2. However, this cytes. This raises the interesting question of why careful heme raised the question whether FLVCR loss/heme dysregulation alters regulation might be required in T cells but not in other lymphoid transcription in DP thymocytes. Therefore, we carried out tran- lineages. One avenue for future studies is to leverage the differ- scriptional profiling on sorted Flvcr2/2 and Flvcr+/+ DP thymo- ential requirement for FLVCR in ab, but not gd, T cells to identify cytes; surprisingly, there were very few transcriptional differences specific functions or characteristics of ab thymocytes/T cells that (Supplemental Fig. 3A). The increased expression of Cox6a2 seen require more stringent heme metabolism. by profiling (Supplemental Fig. 3B, Supplemental Table I) was There is growing appreciation that metabolites modify the validated in independent cohorts by qRT-PCR (Supplemental Fig. epigenome, thereby regulating cell development and differentia- 3C). COX6A2 is a subunit of respiratory cytochrome c oxidase tion and contributing to oncogenesis (6, 38, 39). Recently, it was (complex IV) expressed in striated muscle, and it was shown to demonstrated that, in cancers with succinate dehydrogenase play a role in respiratory uncoupling in muscle (35). Studies in mutations, the tricarboxylic acid intermediate and heme building yeast demonstrated that COX6 expression is upregulated by heme block succinate accumulates and acts as an oncometabolite, (36); thus, the increase in Cox6a2 expression in Flvcr-deleted causing DNA hypermethylation and transcriptional alterations thymocytes may result from increased heme levels. The lack of (Supplemental Fig. 1) (40, 41). Epigenetic changes often precede transcriptional alterations in Flvcr-deleted DP subsets implies that transcriptional changes in development (42), and one possibility is FLVCR loss abruptly disrupts development in the late DP stage, that FLVCR loss and heme accumulation cause epigenetic alter- rather than the block occurring at this stage as a result of cumu- ations in developing thymocytes. DP thymocytes are transcrip- lative deleterious effects during earlier stages. tionally and metabolically quiescent but are epigenetically poised to reactivate transcription and initiate CD4+ or CD8+ T lineage Discussion differentiation once the cells pass positive and negative selection; In this article, we demonstrate that FLVCR, an MFS protein FLVCR loss may disrupt epigenetic programming in these cells so previously shown to transport heme and to be required for erythroid that even DP cells with a TCR capable of selection cannot tran- development, is absolutely required for ab T cell development sition to the SP state. Even in a model in which Flvcr-deleted 1684 FLVCR REQUIRED FOR T CELL DEVELOPMENT thymocytes could develop to the SP stage, these SP cells were not 8. Tracz, M. J., J. Alam, and K. A. Nath. 2007. Physiology and pathophysiology of heme: implications for kidney disease. J. Am. Soc. Nephrol. 18: 414–420. able to survive or proliferate in the periphery in a lymphopenic 9. Reddy, V. S., M. A. Shlykov, R. Castillo, E. I. Sun, and M. H. Saier, Jr. 2012. The host. This finding raises the possibility that FLVCR is required for major facilitator superfamily (MFS) revisited. FEBS J. 279: 2022–2035. T cells to transition between environments (e.g., between different 10. Quigley, J. G., Z. Yang, M. T. Worthington, J. D. Phillips, K. M. Sabo, D. E. Sabath, C. L. Berg, S. Sassa, B. L. Wood, and J. L. Abkowitz. 2004. thymic niches or between lymphoid niches). Although Flvcr- Identification of a human heme exporter that is essential for erythropoiesis. Cell deleted SP cells appeared to mature normally, as determined by 118: 757–766. immunophenotyping, there may be underlying gene expression 11. Keel, S. B., R. T. Doty, Z. Yang, J. G. Quigley, J. Chen, S. Knoblaugh, P. D. Kingsley, I. De Domenico, M. B. Vaughn, J. Kaplan, et al. 2008. A heme changes that lead to the demise of these cells in the periphery that export protein is required for red blood cell differentiation and iron homeostasis. could be interrogated through future microarray experiments. Our Science 319: 825–828. 12. Riedel, N., E. A. Hoover, P. W. Gasper, M. O. Nicolson, and J. I. Mullins. 1986. current models did not allow us to conclusively determine the role Molecular analysis and pathogenesis of the feline aplastic anemia retrovirus, of FLVCR in mature peripheral T cells, but the requirement for feline leukemia virus C-Sarma. J. Virol. 60: 242–250. FLVCR at the DP–SP transition and at the SP cell–peripheral 13. Abkowitz, J. L. 1991. Retrovirus-induced feline pure red blood cell aplasia: pathogenesis and response to suramin. Blood 77: 1442–1451. T cell transition raises the question of whether FLVCR is required 14. Anderson, L. J., W. F. Jarrett, O. Jarrett, and H. M. Laird. 1971. Feline leukemia- in mature T cells and, in particular, whether FLVCR is required virus infection of kittens: mortality associated with atrophy of the thymus and for naive T cell differentiation upon Ag stimulation. We are de- lymphoid depletion. J. Natl. Cancer Inst. 47: 807–817. 15. Mombaerts, P., J. Iacomini, R. S. Johnson, K. Herrup, S. Tonegawa, and veloping models in which to address these questions. V. E. Papaioannou. 1992. RAG-1-deficient mice have no mature B and Heme synthesis has been regarded as a generic housekeeping T lymphocytes. Cell 68: 869–877. 16. Lee, P. P., D. R. Fitzpatrick, C. Beard, H. K. Jessup, S. Lehar, K. W. Makar, function that occurs in all cells (43); however, our finding that M. Pe´rez-Melgosa, M. T. Sweetser, M. S. Schlissel, S. Nguyen, et al. 2001. A FLVCR is specifically required for ab T cell and erythroid de- critical role for Dnmt1 and DNA methylation in T cell development, function, Downloaded from velopment, but not that of other hematopoietic lineages, suggests and survival. Immunity 15: 763–774. 17. Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, and that there are important lineage-specific aspects of heme metab- F. R. Carbone. 1994. T cell receptor antagonist peptides induce positive selec- olism. Although there is now great interest in metabolic control of tion. Cell 76: 17–27. cell function and differentiation, much of the focus has been on 18. Barnden, M. J., J. Allison, W. R. Heath, and F. R. Carbone. 1998. Defective TCR expression in transgenic mice constructed using cDNA-based alpha- and beta- enzymatic pathways, and less is known about the intra- and in- chain genes under the control of heterologous regulatory elements. Immunol. tercellular transport of metabolites. Recently, it was discovered Cell Biol. 76: 34–40. http://www.jimmunol.org/ 19. Pfaffl, M. W. 2001. A new mathematical model for relative quantification in real- that mutations in a ubiquitously expressed magnesium cation time RT-PCR. Nucleic Acids Res. 29: e45. importer, MAGT1, caused T cell immunodeficiency and defects in 20. Persons, D. A., J. A. Allay, E. R. Allay, R. A. Ashmun, D. Orlic, S. M. Jane, T and NK cell cytotoxic function (44, 45). This revealed a more J. M. Cunningham, and A. W. Nienhuis. 1999. Enforced expression of the + GATA-2 transcription factor blocks normal hematopoiesis. Blood 93: 488–499. specific role for Mg2 in T cell and NK cell signaling, in addition 21. Morita, S., T. Kojima, and T. Kitamura. 2000. Plat-E: an efficient and stable to this cation’s universal role as a cofactor in nucleic acids and system for transient packaging of retroviruses. Gene Ther. 7: 1063–1066. enzyme activity (46). Similarly, heme may have a specific and 22. Du, P., W. A. Kibbe, and S. M. Lin. 2008. lumi: a pipeline for processing Illu- mina microarray. Bioinformatics 24: 1547–1548. unique function in ab T cells in addition to its other global 23. Smyth, G. K. 2005. Limma: linear models for microarray data. In: Bio- functions. Our finding that FLVCR specifically blocks T cell de- informatics and Computational Biology Solutions Using R and Bioconductor.

R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, and W. Huber, eds. Springer, by guest on September 28, 2021 velopment at the DP stage, together with the temporally regulated New York, p. 397–420. models of FLVCR function in T cell development that we gen- 24. Reiner, A., D. Yekutieli, and Y. Benjamini. 2003. Identifying differentially erated, will provide the tools needed to determine the unique roles expressed genes using false discovery rate controlling procedures. Bio- informatics 19: 368–375. for FLVCR and heme in T cell development and, ultimately, lead 25. Kuhn,€ R., F. Schwenk, M. Aguet, and K. Rajewsky. 1995. Inducible gene tar- to new insights into the metabolic regulation of T cell develop- geting in mice. Science 269: 1427–1429. ment and function. 26. Rothenberg, E. V., J. E. Moore, and M. A. Yui. 2008. Launching the T-cell- lineage developmental programme. Nat. Rev. Immunol. 8: 9–21. 27. Klein, L., B. Kyewski, P. M. Allen, and K. A. Hogquist. 2014. Positive and negative selection of the T cell repertoire: what thymocytes see (and don’t see). Acknowledgments Nat. Rev. Immunol. 14: 377–391. We thank R.T. Doty for technical support and helpful discussion, 28. Prinz, I., A. Sansoni, A. Kissenpfennig, L. Ardouin, M. Malissen, and T. Mittelstaedt and M. Black from the University of Washington Flow B. Malissen. 2006. Visualization of the earliest steps of gammadelta T cell de- Cytometry Facility, and R. Basom and A. Dawson from the Fred Hutchinson velopment in the adult thymus. Nat. Immunol. 7: 995–1003. 29. Yanatori, I., Y. Yasui, K. Miura, and F. Kishi. 2012. Mutations of FLVCR1 in Cancer Research Center Genomics Shared Resource. posterior column ataxia and retinitis pigmentosa result in the loss of heme export activity. Blood Cells Mol. Dis. 49: 60–66. 30. Seitan, V. C., B. Hao, K. Tachibana-Konwalski, T. Lavagnolli, H. Mira- Disclosures Bontenbal, K. E. Brown, G. Teng, T. Carroll, A. Terry, K. Horan, et al. 2011. A The authors have no financial conflicts of interest. role for cohesin in T-cell-receptor rearrangement and thymocyte differentiation. Nature 476: 467–471. 31. Cyster, J. G., and S. R. Schwab. 2012. Sphingosine-1-phosphate and lymphocyte egress from lymphoid organs. Annu. Rev. Immunol. 30: 69–94. References 32. Chiabrando, D., S. Marro, S. Mercurio, C. Giorgi, S. Petrillo, F. Vinchi, V. Fiorito, 1. Kim, H. J., O. Khalimonchuk, P. M. Smith, and D. R. Winge. 2012. Structure, S. Fagoonee, A. Camporeale, E. Turco, et al. 2012. The mitochondrial heme ex- function, and assembly of heme centers in mitochondrial respiratory complexes. porter FLVCR1b mediates erythroid differentiation. J. Clin. Invest. 122: 4569–4579. Biochim. Biophys. Acta 1823: 1604–1616. 33. Watanabe-Matsui, M., A. Muto, T. Matsui, A. Itoh-Nakadai, O. Nakajima, 2. Furuyama, K., K. Kaneko, and P. D. Vargas. 2007. Heme as a magnificent K. Murayama, M. Yamamoto, M. Ikeda-Saito, and K. Igarashi. 2011. Heme molecule with multiple missions: heme determines its own fate and governs regulates B-cell differentiation, antibody class switch, and heme oxygenase-1 cellular homeostasis. Tohoku J. Exp. Med. 213: 1–16. expression in B cells as a ligand of Bach2. Blood 117: 5438–5448. 3. Chen, J. J. 2007. Regulation of protein synthesis by the heme-regulated 34. Roychoudhuri, R., K. Hirahara, K. Mousavi, D. Clever, C. A. Klebanoff, eIF2alpha kinase: relevance to anemias. Blood 109: 2693–2699. M. Bonelli, G. Sciume`, H. Zare, G. Vahedi, B. Dema, et al. 2013. BACH2 4. Kaasik, K., and C. C. Lee. 2004. Reciprocal regulation of haem biosynthesis and represses effector programs to stabilize T(reg)-mediated immune homeostasis. the circadian clock in mammals. Nature 430: 467–471. Nature 498: 506–510. 5. Handschin, C., J. Lin, J. Rhee, A. K. Peyer, S. Chin, P. H. Wu, U. A. Meyer, and 35. Quintens, R., S. Singh, K. Lemaire, K. De Bock, M. Granvik, A. Schraenen, B. M. Spiegelman. 2005. Nutritional regulation of hepatic heme biosynthesis and I. O. Vroegrijk, V. Costa, P. Van Noten, D. Lambrechts, et al. 2013. Mice de- porphyria through PGC-1alpha. Cell 122: 505–515. ficient in the respiratory chain gene Cox6a2 are protected against high-fat diet- 6. Feng, D., and M. A. Lazar. 2012. Clocks, metabolism, and the epigenome. Mol. induced obesity and insulin resistance. PLoS ONE 8: e56719. Cell 47: 158–167. 36. Trawick, J. D., R. M. Wright, and R. O. Poyton. 1989. Transcription of yeast 7. Khan, A. A., and J. G. Quigley. 2011. Control of intracellular heme levels: heme COX6, the gene for cytochrome c oxidase subunit VI, is dependent on heme and transporters and heme oxygenases. Biochim. Biophys. Acta 1813: 668–682. on the HAP2 gene. J. Biol. Chem. 264: 7005–7008. The Journal of Immunology 1685

37. Mingueneau, M., T. Kreslavsky, D. Gray, T. Heng, R. Cruse, J. Ericson, structure marks key developmental genes in embryonic stem cells. Cell 125: S. Bendall, M. H. Spitzer, G. P. Nolan, K. Kobayashi, et al; Immunological 315–326. Genome Consortium. 2013. The transcriptional landscape of ab T cell differ- 43. Bishop, D. F., A. S. Henderson, and K. H. Astrin. 1990. Human delta- entiation. Nat. Immunol. 14: 619–632. aminolevulinate synthase: assignment of the housekeeping gene to 3p21 and 38. Katada, S., A. Imhof, and P. Sassone-Corsi. 2012. Connecting threads: epi- the erythroid-specific gene to the X chromosome. Genomics 7: 207–214. genetics and metabolism. Cell 148: 24–28. 44. Li, F. Y., B. Chaigne-Delalande, C. Kanellopoulou, J. C. Davis, H. F. Matthews, 39. Kaelin, W. G., Jr., and S. L. McKnight. 2013. Influence of metabolism on epi- D. C. Douek, J. I. Cohen, G. Uzel, H. C. Su, and M. J. Lenardo. 2011. Second genetics and disease. Cell 153: 56–69. messenger role for Mg2+ revealed by human T-cell immunodeficiency. Nature 40. Letouze´, E., C. Martinelli, C. Loriot, N. Burnichon, N. Abermil, C. Ottolenghi, 475: 471–476. M. Janin, M. Menara, A. T. Nguyen, P. Benit, et al. 2013. SDH mutations establish 45.Li,F.Y.,B.Chaigne-Delalande,H.Su,G.Uzel,H.Matthews,and a hypermethylator phenotype in paraganglioma. Cancer Cell 23: 739–752. M. J. Lenardo. 2014. XMEN disease: a new primary immunodeficiency af- 41. Killian, J. K., S. Y. Kim, M. Miettinen, C. Smith, M. Merino, M. Tsokos, fecting Mg2+ regulation of immunity against Epstein-Barr virus. Blood 123: M. Quezado, W. I. Smith, Jr., M. S. Jahromi, P. Xekouki, et al. 2013. Succinate 2148–2152. dehydrogenase mutation underlies global epigenomic divergence in gastroin- 46. Chaigne-Delalande, B., F. Y. Li, G. M. O’Connor, M. J. Lukacs, P. Jiang, testinal stromal tumor. Cancer Discov. 3: 648–657. L. Zheng, A. Shatzer, M. Biancalana, S. Pittaluga, H. F. Matthews, et al. 2013. 42.Bernstein,B.E.,T.S.Mikkelsen,X.Xie,M.Kamal,D.J.Huebert,J.Cuff, Mg2+ regulates cytotoxic functions of NK and CD8 T cells in chronic EBV B. Fry, A. Meissner, M. Wernig, K. Plath, et al. 2006. A bivalent chromatin infection through NKG2D. Science 341: 186–191. Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021