Research Article

The protein LEM promotes CD8+ T cell immunity through effects on mitochondrial respiration

See allHide authors and affiliations

Science  29 May 2015:
Vol. 348, Issue 6238, pp. 995-1001
DOI: 10.1126/science.aaa7516

This article has a correction, but has also been retracted. Please see:

LEM gets T cells the energy they need

During an infection, T cells proliferate extensively to build a sufficient army to defeat the invading pathogen. Carefully regulated changes in metabolism let T cells do this, but the specific nature of these changes is not fully understood. Using forward genetics in mice to screen for genes that regulate T cell immunity, Okoye et al. identified a mutation in the gene that encodes a protein they named lymphocyte expansion molecule (LEM) (see the Perspective by O'Sullivan and Pearce). LEM enhanced T cell immunity, including both proliferation and memory cell generation, in response to chronic viral infection. LEM facilitated these changes through effects on mitochondrial respiration.

Science, this issue p. 995; see also p. 976

Abstract

Protective CD8+ T cell–mediated immunity requires a massive expansion in cell number and the development of long-lived memory cells. Using forward genetics in mice, we identified an orphan protein named lymphocyte expansion molecule (LEM) that promoted antigen-dependent CD8+ T cell proliferation, effector function, and memory cell generation in response to infection with lymphocytic choriomeningitis virus. Generation of LEM-deficient mice confirmed these results. Through interaction with CR6 interacting factor (CRIF1), LEM controlled the levels of oxidative phosphorylation (OXPHOS) complexes and respiration, resulting in the production of pro-proliferative mitochondrial reactive oxygen species (mROS). LEM provides a link between immune activation and the expansion of protective CD8+ T cells driven by OXPHOS and represents a pathway for the restoration of long-term protective immunity based on metabolically modified cytotoxic CD8+ T cells.

Cytotoxic CD8+ T cells (CTLs) are a central arm of the immune system responsible for protection from intracellular viruses and cancer because they kill infected or transformed cells (1). Because chronic virus infection (2) and cancer (3) are widespread diseases, it is clear that CTL immunity often fails. A major reason for this failure is that high viral (4, 5) or tumor (68) load results in either deletion or functional inactivation (known as immune exhaustion) of CTLs. The result is failure in both short-term CTL immunity and immunological memory because memory CD8 T cell development is blocked (9). Impaired expansion is an important cause of deletion and immune exhaustion and results in failure to produce sufficient numbers of protective CTLs and memory cells (5).

Retro mutant mice have increased immunity to chronic viral infection

Infection of wild-type C57BL/6 mice with the clone 13 variant of lymphocytic choriomeningitis virus (LCMV C13) is an established model for human chronic viral infection resulting in a massive viral load that causes both deletion and immune exhaustion of CTLs and a block in memory CD8 T cell development (10).

We examined the CTL response to LCMV C13 infection after germline mutagenesis to identify mutants with enhanced immunity. To this end, 430 third-generation (G3) ethyl-N-nitrosourea (ENU)–induced germline mutants were produced in a C57BL/6J background (11). G3 mice were infected with LCMV C13, and after 8 days the level of LCMV-specific CD8 T cells was measured in the spleen by staining with a tetramer for the np396 LCMV peptide and by flow cytometry (12). Three independent germline transmissible modifications, which resulted in increased levels of LCMV-specific CD8 T cells, were isolated, of which one (a semidominant) was bred to homozygosity (fig. S1A). We named this strain Retro.

Homozygous Retro mutant mice showed a 10-fold increase in CD8 T cells specific for LCMV np396 peptide compared with wild-type (WT) and a smaller but significant increase in the number of CTLs specific for the gp33-LCMV peptide (Fig. 1, A and B). Compared with WT mice, a smaller percentage of Retro mutant CTLs expressed the programmed cell death–1 (PD-1) immune exhaustion marker (5) (Fig. 1C). However, Retro mutant mice still harbored increased numbers of PD-1+ np396+ CD8+ cells compared with WT (Fig. 1D). In addition, Retro mutant CTLs exhibited increased functionality: There was increased CTL activity in ex vivo killing experiments (Fig. 1E), increased exocytosis of cytotoxic granules (11) (Fig. 1F), and increased production of the cytokine interferon (IFN)–γ (fig. S1B). The increase in CTL immunity resulted in a 106-fold decrease in LCMV C13 titer in the spleens of Retro mutant mice on day 8 after infection compared with WT (Fig. 1G). The increase in CTL activity was a function of cell number (fig. S1C).

Fig. 1 Retro mutant mice exhibit enhanced CD8 T cell responses to chronic LCMV infection.

(A) Flow cytometry on spleen cells from day 8 post infection (p.i.) with LCMV C13 of WT and Retro homozygous mutant (R) mice. Staining with tetramers, refolded with either np396 or gp33 peptides, and PD-1 is shown (percentage next to gate) on CD8+ gated cells. (B) Number of tetramer+ CD8+ splenocytes on day 8 p.i. (C) Percentage of PD-1+ of tetramer+ CD8+ splenocytes on day 8 p.i. (D) Number of PD-1+ tetramer+ CD8+ splenocytes on day 8 p.i. (E) Percentage of specific CTL lysis from Cr51-release assays for splenocytes on day 8 p.i. (F) Number of CD107α+ CD8+ splenocytes on day 8 p.i. (G) Titer of LCMV C13 in the spleen on day 8 p.i. (H) Kaplan-Meier survival graph after LCMV C13 infection. (I) Percentage of BrdU+ of np396+ CD8+ splenocytes on day 8 p.i. Mean ± SEM; N = 6 to 12 mice. ***P < 0.0001; **P < 0.0005; *P < 0.001. Two-tailed Student’s t tests were used for all, except that Gehan-Breslow-Wilcoxon tests were used for survival curves. Each data set is representative of at least three individual experiments.

All Retro mutant mice succumbed to LCMV C13 infection after 14 days, whereas all WT mice had survived (Fig. 1H). Retro mutant death is presumably a consequence of elevated CTL-mediated cytolysis resulting in the fatal loss of vascular integrity (13). Despite an increase in activated phenotype T cells (fig. S2), viability, blood development, and autoimmunity in Retro mutant mice was normal (fig. S3 and tables S1 to S3). In vivo bromodeoxyuridine (BrdU) labeling indicated that np396+ CD8+ cells underwent increased proliferation after LCMV C13 infection of Retro mutant mice (Fig. 1I). The increased number of antigen-specific CD8 T cells in LCMV C13–infected Retro mutant mice is likely due to increased proliferation of these cells.

Adoptive transfer of Retro mutant CD8 T cells (14) showed that the phenotype of increased CTL immunity resides in the CD8 T cell compartment (fig. S4, A to D). The phenotype of Retro mutant mice also extended to antitumor responses. After challenge with B16 F10 melanoma cells, Retro mutant mice harbored three times as many CTLs (Fig. 2A) and one-fourth as many tumors (Fig. 2, B and C).

Fig. 2 Retro improves immunity against tumors and acute viral infections.

Mice were injected intravenously with B16 F10 melanoma cells. (A) Percentage of CD3+ CD8+ CD44hi cells from lungs, day 35 p.i. (B) Foci of melanomas in the lung on day 35 p.i. (C) Number of melanomas per lung on day 35 p.i. Mice were infected with LCMV Armstrong. (D) Mean percentage of (left) np396+ and (right) gp33+ CD8+ peripheral blood leukocytes (PBL). Mean ± SEM; N = 6 to 10 mice; ***P < 0.0001. Two-tailed Student’s t tests were used for all. Each data set is representative of at least three individual experiments.

Retro mutant mice have increased memory T cell development

After acute infection of Retro mutant mice with LCMV Armstrong, there was an increase in the clonal burst size of both np396+ CD8+ and gp33+ CD8+ cells (Fig. 2D) compared with WT. Longitudinal analysis in the blood revealed a corresponding increase in the number of LCMV-specific CD8 T cells during the memory phase in Retro mutant mice (15). After 70 days, there was a fourfold increase in the number of LCMV-specific central memory T CD8 cells (Tcm) in the spleen as identified by the CD62L+ CD127+ CD8+ cell surface phenotype (16) (fig. S4, E and F).

We boosted mice with a second dose of LCMV Armstrong and measured the expansion of secondary CTLs during the memory response. The numbers of secondary CTLs were about 10 times as high compared with WT (fig. S4, G and H), indicating that Retro mutant mice have an enhanced memory phenotype. During acute immune responses the pool of CTLs contains not only short-lived effectors but also memory cell precursors (17, 18). The increase in the total number of CD8 T cells resulted in an increase in the number and percentage of both memory cell precursors and short-lived effectors in Retro mutants compared with WT (fig. S5). Therefore, the Retro mutation increases the level of both primary short-lived effectors and Tcm by facilitating a global increase in the expansion of antigen-activated CD8 T cells.

The Retro mutation identifies lymphocyte expansion molecule

We used high-throughput exome sequencing (19) to identify the Retro mutation. Comparison with the reference C57BL/6J genome identified several homozygous single-nucleotide variants (SNV) in homozygous Retro mutant mice (fig. S6A). Upon resequencing and comparison with the Charles River substrain of C57BL/6J used for our ENU mutagenesis, four homozygous SNVs were identified as being associated with Retro homozygous mutant mice. Only one homozygous SNV correlated with the Retro homozygous phenotype (fig. S6B). This was an A to G transition of nucleotide 1304 (A1304G) in exon 2 of the BC055111 gene on chromosome 4 (fig. S7A). BC055111 encodes an orphan protein with no substantial homology to any other mouse protein (BLASTP, E value > 0.38). The BC055111Retro allele (R) was semidominant for increased CTL immunity to LCMV C13 (fig. S7, B to D).

We generated C57BL/6 mice with a BC055111 knock-out (KO) allele (20) (fig. S8). Homozygosity for the BC055111 KO allele results in embryonic lethality, so we used heterozygous KO mice to validate the gene candidate (12). There was decreased BC055111-encoded protein in CTLs from heterozygous BC055111 KO mice compared with WT (Fig. 3, A and B). We observed decreased np396+ CD8+ expansion (Fig. 3C), increased LCMV titer (Fig. 3D) and decreased CTL activity (Fig. 3E) in BC055111WT/KO compared with BC055111WT/WT mice. Therefore, BC055111 directly controls CTL immunity to LCMV. We named the BC055111 encoded orphan protein “lymphocyte expansion molecule” (LEM).

Fig. 3 Retro mutation identifies LEM.

BC055111WT/WT and BC055111WT/KO mice were infected with LCMV C13. (A) Western blot on CTLs for BC055111 (49 kD) on total CD8 T cells from the spleen on day 4 p.i. (B) Bar graph of BC055111 signal strength from (A). BC055111WT/WT and BC055111WT/KO mice were infected with LCMV C13. (C) Number of np396+ CD8+ splenocytes on day 8 p.i. (D) LCMV titer in the spleen on day 8 p.i. (E) Percentage of specific CTL lysis from Cr51-release assays for splenocytes on day 8 p.i. (F) Bar graph for mean Lem mRNA in total CD8 T cells from the spleens of WT or R mice after infection with LCMV C13. Level of mRNA normalized to the t = 0 WT level. (G) Bar graphs for mean Lem mRNA in CD8 T cells cultured for 3 days in cytokines, then incubated with actinomycin D or emitine. Level of mRNA normalized to the t = 0 WT level. [(A) to (G)] Mean ± SEM; N = 6 mice. Each data set is representative of at least three individual experiments. (H) Bar graphs for mean relative Lem or Gapdh mRNA from two independent immunoprecipitation experiments performed on N = 2 mice each time. mRNA levels are relative to WT and are from CD8 T cells cultured for 3 days in cytokines. (I) Western blot for SF2 (27 kD) after immunoprecipitation in (H). (J) Mean number of green fluorescent protein–positive (GFP+) CTLL-2 cells cultured in IL-2 (N = 4 wells) from three independent transductions with MIGR1 alone (V-black), or MIGR1 encoded Lem ORF (WT, blue sequence; R, red sequence). ***P < 0.005; **P < 0.01; *P < 0.05. Two-tailed Student’s t tests were used.

The Retro mutation stabilizes Lem mRNA

The kinetics of Lem mRNA accumulation in Retro mutant CD8 T cells after LCMV C13 infection was increased compared with WT (Fig. 3F). The increased level of Lem mRNA was also observed when transcription was inhibited by actinomycin D, indicating that the Retro mutation stabilized Lem mRNA in CD8 T cells (Fig. 3G).

Lem mRNA contains introns in the 3′ untranslated region (3′UTR), which may make it susceptible to nonsense-mediated decay (NMD) (21). This was confirmed when blockade of NMD by emetine increased the stability of wild-type Lem mRNA (Fig. 3G). The Retro mutation is predicted to abolish an alternative splice factor/splice factor 2 (ASF/SF2) binding site in Lem (12), which is thought to be required for NMD (22). Immunoprecipitation (IP) confirmed decreased levels of SF2 associated with Retro mutant Lem message (Fig. 3H) in cells with normal SF2 levels (Fig. 3I). This suggests that the A1304G mutation in Retro mutant mice partially protects Lem mRNA from NMD, resulting in stabilization. The Lem open reading frame (ORF) harboring the Retro A1304G mutation was no better than WT ORF at driving interleukin-2 (IL-2)–induced proliferation of T cells (fig. S9 and Fig. 3J). Therefore, the increased expression of Lem rather than any qualitative alteration in protein activity is responsible for the phenotype of Retro mutant mice. We conclude that WT LEM is a positive modulator of CTL expansion that is up-regulated in Retro mutant mice.

The Retro mutation increases LEM expression

LEM protein was up-regulated in total CD8 T cells after LCMV C13 infection with Retro mutant mice displaying increased kinetics compared with WT (Fig. 4, A and B). Intracellular staining (fig. S10) indicated that LEM expression was specific for activated CD8 T cells (CD44hi CD62Llo CD8+) and was elevated in Retro mutant cells (Fig. 4, C and D). LEM was up-regulated 2 days after LCMV C13 infection despite a low frequency of CD8 cells expressing LCMV-specific T cell receptors (TCRs) (fig. S11A and Fig. 4, A and C). In addition to stimulation through the TCR (via antibody to CD3) (Fig. 4E), treatment with either IL-2 (Fig. 4F) or IL-15 (Fig. 4G) resulted in LEM up-regulation, which was enhanced in Retro mutant CD8 T cells (Fig. 4, E to G). Therefore, LEM expression can be induced by bystander activation by pro-proliferative cytokines as well as antigen-dependent TCR signals. The level of LEM controlled the levels of Retro mutant CD8 T cells both in vivo (fig. S11, B to E) and in vitro (fig. S12). PD-1 up-regulation by Retro mutant CTLs was unaffected in vitro after TCR stimulation, suggesting that LEM does not directly down-regulate PD-1 expression (fig. S11, F and G).

Fig. 4 The Retro mutation increases LEM expression.

Mice were infected with LCMV C13. (A) Western blots for LEM (49 kD) on total CD8 T cells from the spleen. (B) Bar graph of signal strength from (A). (C to G) Mean fluorescence intensity (MFI) after staining with antibody to LEM in gated (C) CD44hi CD62Llo CD8+ or (D) CD44lo CD62Lhi CD8+ splenocytes after infection with LCMV C13, or on total CD8 T cells that were activated in vitro with either (E) antibody to CD3/CD28, (F) IL-2, or (G) IL-15. [(A) to (G)] Mean ± SEM; N = 6 mice. (H) Western blots for human LEM, 49 kD on antibody to CD3/CD28-activated human CD8 T cells. (I) Bar graph of signal strength from (H). (J) Mean number of GFP+ human CD8 T cell blasts (N = 4 wells) after three independent transductions with MIGR1 alone (V-black), or MIGR1 encoded hLEM ORF (gray). ***P < 0.0001; **P < 0.0005; *P < 0.001. Two-tailed Student’s t tests were used. Each data set is representative of at least three individual experiments.

Functional equivalence of human LEM

C1ORF177 encodes the human homolog of LEM (BLASTP, 72% of complete amino acid identity with mouse LEM; E value = 0), which is up-regulated in activated human CD8 T cells (Fig. 4, H and I). Ectopic expression of human LEM resulted in about a 15-fold increase in the expansion (Fig. 4J). Therefore, LEM can modulate CD8 T cell expansion in man as well as mice.

LEM interacts with CRIF1 to control the activity of OXPHOS proteins

Analysis of LEM amino acid sequence predicts the presence of at least two intrinsically unstructured (IU) protein domains (12, 23). Therefore, we investigated whether LEM modulates CD8 T cell proliferation through specific protein:protein interactions. Yeast two-hybrid (Y2H) analysis identified nine different LEM-interacting proteins (24) (fig. S13). Coimmunoprecipitation (co-IP) with antibody to LEM (fig. S10) demonstrated that one target—CR6 interacting factor (CRIF1)—interacts with LEM in CTLs (Fig. 5, A and B).

Fig. 5 LEM interacts with CRIF1 to control the activity of OXPHOS proteins.

WT and R mice were infected with LCMV C13, and on day 4 p.i. total CD8 T cells were purified. (A) Western blots directly on cell extracts (input) or after immunoprecipitation for LEM (LEM, 49 kD; CRIF1, 25 kD; β actin, 42 kD). (B) Bar graph of signal strengths from (A). (C) CIM from staining with antibodies against LEM (red) or CRIF1 (green). Nuclei stained with 4′,6-diamidino-2-phenylindole (blue). Examples of colocalization (yellow) are indicated by arrows. Scale bar, 2.5 μm. (D) Colocalization of LEM and CRIF1 as measured by Pearson’s correlation coefficient in multiple optical cell slices (N = 62). Mean = 0.68Embedded Image 0.11. WT mice were infected with LCMV C13, and on day 4 p.i. total CD8 T cells were purified. (E) Western blots directly on cell extracts (input) or after immunoprecipitation with LEM antibody or with control rabbit immunoglobulin G (MRPL23, 18 kD; ND1, 36 kD; COX1, 57 kD; ATP5A1, 60 kD). LemWT/WT, LemR/R, or LemWT/KO mice were infected with LCMV C13, and on day 4 p.i. total CD8 T cells were purified. (F) Western blots for OXPHOS proteins (UQCRC2, 49 kD.) (G) Bar graph of signal strength from (F). WT and R mice were infected with LCMV C13, and on day 4 p.i. total CD8 T cells were purified. (H) Immunocapture-enzyme assays for OXPHOS complexes I and IV. (I) Bar graphs for mean specific activity (signal strength per mg protein) from (H). Donor CD8 T cells harboring vector alone of Crif1-AS were purified from WT recipient mice on day 4 p.i. with LCMV C13. (J) Immunocapture-enzyme assays for OXPHOS complexes I and IV. (K) Bar graphs for mean specific activity from (J). Mean ± SEM; N = 6 mice. ***P < 0.005; **P < 0.01; *P < 0.05. Two-tailed Student’s t tests were used. Each data set is representative of at least three individual experiments.

CRIF1 is required for the translation and insertion of oxidative phosphorylation (OXPHOS) polypeptides into the inner membrane of mitochondria after they emerge from the 39S subunit of the mitoribosomes (25). Confocal immunofluorescence microscopy (CIM) revealed that LEM colocalizes with CRIF1 in CTLs (Fig. 5, C and D, and fig. S14). Both LEM and CRIF1 were localized to mitochondria by costaining with the mitotracker red dye (figs. S15 and S16). Co-IP localized LEM to the mitoribosome by revealing an association with the 39S subunit protein MRPL23 and several OXPHOS proteins such as NADH ubiquinone oxidoreductase chain 1 (ND1), ubiquninol cytochrome c oxidoreductase chain 2 (UQCRC2), and cytochrome c oxidoreductase chain 1 (COX1) (25) (Fig. 5E). The OXPHOS protein adenosine triphosphate (ATP) synthase α subunit 1 (ATP5A1) does not require CRIF1 for translation and did not coimmunoprecipitate, suggesting that interaction of LEM with OXPHOS proteins is via CRIF1. Through its interaction with CRIF, LEM is part of a complex that mediates the translation and insertion of OXPHOS proteins into the mitochondrial inner membrane.

Retro mutant CTLs harbored increased levels, and heterozygous LEM KO CTLs reduced levels of OXPHOS proteins (Fig. 5, F and G). Immunocapture confirmed that the specific activities of OXPHOS complexes were elevated in Retro mutant CTLs (Fig. 5, H and I), although LEM did not affect the expression of Oxphos genes or the number or mass of mitochondria (fig. S17, A to C). Therefore, LEM controlled the activity of OXPHOS complexes by most likely facilitating the production of subunit proteins.

As was observed in CRIF1 KO cells (25), when we knocked down Crif1 mRNA (fig. S18, A to C), OXPHOS-specific activity was reduced (Fig. 5, J and K). Knock-down of Crif1 message in Retro mutant CTLs also abolished the increases in OXPHOS activity due to LEM up-regulation. We conclude that LEM interaction with CRIF1 within mitochondria determines OXPHOS activity.

CRIF1 is required for cell proliferation during embryo development (26). Crif1 message knock-down resulted in the impaired proliferation and expansion of anti-LCMV CTLs (fig. S18, D and E). The increase in CTL expansion caused by LEM up-regulation required CRIF1 expression because Crif1 message knock-down diminished the increased expansion of Retro mutant CTLs. We conclude that LEM requires interaction with CRIF1 to drive the expansion of CTLs.

LEM is a positive modulator of mitochondrial ROS–driven proliferation

We examined the consequences of LEM control of OXPHOS activity on CTL metabolism and expansion. Oxygen consumption studies demonstrated that Retro mutant CTLs had significantly higher respiratory levels compared with WT (Fig. 6A) (27), but glycolysis was not increased (fig. S17D). Conversely, respiratory activity (Fig. 6B) of LEM heterozygous KO CTLs, but not glycolysis (fig. S17E), was decreased.

Fig. 6 LEM controls production of mitogenic mROS.

LemWT/WT, LemR/R, or LemWT/KO mice were infected with LCMV C13, and on day 4 p.i. total CD8 T cells were purified from the spleen. (A and B) Oxygen consumption rate (OCR) after sequential injection of oligomycin, FCCP, and antimycin A/rotenone. (C) LemWT/WT, LemR/R, or LemWT/KO mice were infected with LCMV C13, then on day 8 p.i. np 396+ CD8+ splenocytes were gated, then the MFI of MitoSOX red staining was determined. R mice were infected with LCMV C13, then over the course of 8 days were injected with MnTBAP or not. Gated np 396+ CD8+ splenocytes were examined on day 8 p.i. for (D) MitoSOX red staining and (E) % BrdU-positive cells. In the same spleens: (F) Number of tetramer-positive cells and (G) Titer of LCMV. Mean ± SEM. For (A) to (F), N = 6 mice, and for (G), N = 9 mice. ***P < 0.001; **P < 0.005; *P < 0.01. Two-tailed Student’s t tests were used. Each data set is representative of at least three individual experiments.

OXPHOS results in the production of ROS through stepwise reduction of O2 (27). The modulation of OXPHOS by LEM resulted in altered production of mROS. Staining with the redox-sensitive dye MitoSOX red revealed that compared with WT, Retro mutant CTLs had increased levels of mROS and LEM heterozygous KO CTLs had decreased mROS (Fig. 6C). The production of mROS is required for the activation and expansion of T lymphocytes (27). Injection of Retro mutant mice with the antioxidant MnTBAP (28) reduced the level of mROS in CTLs (Fig. 6D), resulting in decreased proliferation (Fig. 6E) and expansion of CTLs (Fig. 6F) with a corresponding increase in LCMV C13 titer (Fig. 6G). We conclude that mROS resulting from OXPHOS activity drives expansion and CTL immunity in Retro mutant mice.

We have used unbiased forward genetics to discover LEM at the heart of a pathway that, when up-regulated, not only restores CTL immunity to chronic viral infection and tumor challenge but also increases memory cell development. Although other molecular interventions either restore primary CTL immunity (5) or divert development toward the memory lineage (29), the LEM pathway is important because it controls the expansion of both short-term effectors and memory cells (9). Up-regulation of LEM in Retro mutant mice increased both the overall number of CTLs and also the number of PD-1+ CTLs. Therefore, LEM therapy has the potential both to globally expand CTLs and to increase the number of PD-1+ CTLs available for derepression by antibodies to PD-1 (8).

At the peptide tunnel exit of 39S mitoribosomes, LEM is likely recruited by CRIF1 to facilitate the insertion of OXPHOS polypeptides into the inner mitochondrial membrane, thereby controlling OXPHOS activity (25) (fig. S19). We propose that CTL expansion and memory cell development are controlled by mROS produced after the up-regulation of LEM upon immune activation (27, 30)

Supplementary Materials

www.sciencemag.org/content/348/6238/995/suppl/DC1

Materials and Methods

Figs. S1 to 19

Tables S1 to S3

References (3138)

References and Notes

  1. Materials and methods are available as supplementary materials on Science Online.
  2. Acknowledgments: We thank C. Bangham and R. Shattock for helpful comments on the paper. The data reported in this manuscript are tabulated in the main paper and in the supplementary materials. The Lem gene sequence is deposited in GenBank accession KP939367. P.G.A.-R. is the inventor on international patent application PCT/GB2014/051603 (International Publication Number WO 2014/188220) A1. The patent covers the use of LEM nucleic acids and polypeptides to activate T cell immunity to cancer and viral infection and agents that down-regulate LEM for the treatment of autoimmune and inflammatory disease. Retro mutant mice are available from P.G.A.-R. under a material transfer agreement with Imperial College London. We thank J. Hehl and S. Stoma from the Scientific Center for Optical and Electron Microscopy ScopeM of the Swiss Federal Institute of Technology ETHZ for technical support. The research was supported by the ETH Zurich, the Swiss National Science Foundation (310030_146140), Medical Research Council (MRC) grant G0700795, and NIH grant AI091930. I.O. and research were supported by a grant from the Wellcome Trust; L.W. and D.H. and research were supported by a grant from Cancer Research UK (A9995); O.C. and research were supported by NIH grant AI45108. R.H. was supported by a MRC Ph.D. studentship, and C.M. was supported by a British Heart Foundation Fellowship (FS/12/38/29640).
View Abstract

Subjects

Navigate This Article