Phenotypic Mutation 'elektra' (pdf version)
Mutation Type missense
Coordinate82,960,426 bp (GRCm39)
Base Change T ⇒ A (forward strand)
Gene Slfn2
Gene Name schlafen 2
Synonym(s) Shlf2
Chromosomal Location 82,955,938-82,961,504 bp (+) (GRCm39)
MGI Phenotype PHENOTYPE: Mice homozygous for an ENU-induced allele exhibit increased susceptibility to bacterial and viral infections, reduced T cell numbers, decreased T cell proliferation, and increased apoptosis of activated T cells. [provided by MGI curators]
Accession Number

NCBI RefSeq: NM_011408; MGI: 1313258

Limits of the Critical Region 44187191 - 88990222 bp
Amino Acid Change Isoleucine changed to Asparagine
Institutional SourceBeutler Lab
Gene Model not available
AlphaFold Q9Z0I6
SMART Domains Protein: ENSMUSP00000035562
Gene: ENSMUSG00000072620
AA Change: I135N

low complexity region 162 172 N/A INTRINSIC
Pfam:AlbA_2 231 365 4.5e-12 PFAM
Predicted Effect probably damaging

PolyPhen 2 Score 0.996 (Sensitivity: 0.55; Specificity: 0.98)
(Using ENSMUST00000038038)
Meta Mutation Damage Score Not available question?
Is this an essential gene? Probably nonessential (E-score: 0.051) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status loading ...
Single pedigree
Linkage Analysis Data
Penetrance 100% 
Alleles Listed at MGI
Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL02141:Slfn2 APN 11 82960670 missense probably benign 0.32
IGL02407:Slfn2 APN 11 82960402 missense probably benign 0.00
IGL03303:Slfn2 APN 11 82960293 missense possibly damaging 0.82
elektra2 UTSW 11 82960390 missense probably benign 0.05
R2270:Slfn2 UTSW 11 82960761 missense probably damaging 0.99
R2392:Slfn2 UTSW 11 82956154 missense possibly damaging 0.91
R3436:Slfn2 UTSW 11 82960390 missense probably benign 0.05
R3437:Slfn2 UTSW 11 82960390 missense probably benign 0.05
R5849:Slfn2 UTSW 11 82960402 missense probably benign 0.00
R7612:Slfn2 UTSW 11 82961089 missense probably damaging 1.00
R7738:Slfn2 UTSW 11 82960799 missense probably damaging 1.00
R8211:Slfn2 UTSW 11 82960585 missense possibly damaging 0.50
R8465:Slfn2 UTSW 11 82960487 missense probably damaging 1.00
R8986:Slfn2 UTSW 11 82960427 missense possibly damaging 0.71
R8987:Slfn2 UTSW 11 82960363 missense probably damaging 0.97
R9627:Slfn2 UTSW 11 82960367 missense
Mode of Inheritance Autosomal Recessive
Local Stock Embryos, Sperm, gDNA
MMRRC Submission 031705-UCD
Last Updated 2018-01-05 3:11 PM by Eva Marie Y. Moresco
Record Created unknown
Record Posted 2010-02-08
Phenotypic Description
The elektra phenotype was identified among ENU-induced G3 mutant mice in a screen for susceptibility to mouse cytomegalovirus (MCMV) (MCMV Susceptibility and Resistance Screen) (1)Elektra mice died six to eight days after infection with 2x105 PFU of MCMV, whereas nearly all wild type mice survived (Figure 1A). Upon infection with 1x105 PFU of MCMV, elektra mice displayed elevated splenic viral titers relative to wild type mice, although production of the cytokines interleukin (IL)-12, interferon (IFN)-γ and tumor necrosis factor (TNF)-α detected in serum was not significantly different than in wild type mice.  Natural killer (NK) cell activation and function, critical for control of MCMV (2), were intact in elektra mice: IFN-γ secretion by elektra NK cells in vitro was normal in response to IL-12, IL-12 plus IL-18, or Ly49D- or NKp46-activating antibodies. Elektra mice killed class I MHC-deficient and allogeneic NK cell targets in vivo as efficiently as wild type mice. NK cell numbers were significantly increased in the thymus, spleen and liver of elektra mice. The MCMV susceptibility of elektra mice was completely rescued by bone marrow transplantation from wild type mice, suggesting that a hematopoietic defect was responsible for this phenotype.
Elektra homozygotes were also highly susceptible to infection with Listeria monocytogenes, exhibiting increased bacterial loads in the liver 2 days post-infection relative to those of wild type mice. Similar to the MyD88 mutant pococurante, Listeria-infected elektra mice died 4-5 days post-infection, when wild type mice appeared healthy (Figure 1B). When infected with lymphocytic choriomeningitis virus (LCMV), elektra mice displayed elevated splenic viral titers, while it was effectively cleared from wild type mice by 7 days post-infection (Figure 1C).
Elektra mice displayed normal cellularity of the spleen, thymus, lymph nodes, and peripheral blood. B cells from elektra mice proliferated normally in response to BCR stimulation or CpG treatment. However, antibody production in response to MCMV infection was delayed. Elektra mice lacked NKT (CD3+NK1.1+) cells in the thymus and spleen, and had reduced numbers in the liver. When elektra NKT cells were stimulated with α-galactosylceramide tetramer, they produced greatly reduced levels of IFN-γ and IL-4 compared to wild type cells. Elektra mice displayed a 50% reduction in Foxp3+ regulatory T cells.
Low percentages of CD4 and CD8 T cells were evident both in the spleen and lymph nodes of elektra homozygotes. The percentage of CD8 T cells was markedly reduced in blood, while the percentage of CD4 T cells was slightly reduced (Figure 2A). However, thymic T cell populations were normal as assessed by double negative (CD4-CD8-), double positive (CD4+CD8+), and single positive cell ratios, as well as total thymocyte numbers. Consistent with their failure to restrict the proliferation of LCMV, elektra homozygotes showed a reduction in CD8 T cell numbers in response to LCMV infection (Figure 2B, upper panel).  Re-stimulation of splenocytes from LCMV-infected elektra homozygotes ex vivo using LCMV-derived peptides (representing immunodominant epitopes of both envelope and nuclear protein antigens) revealed a severe reduction in the number of IFNγ-producing CD8 cells relative to wild type (Figure 2B, lower panel).
Elektra CD8+ T cells failed to proliferate upon stimulation with CD3ε and CD28 antibodies, with PMA and ionomycin, or with IL-12 in vitro (Figure 3A). However, a higher percentage of elektra CD8 T cells were observed to incorporate BrdU relative to wild type cells 24 hours after stimulation with CD3ε/CD28 antibodies, indicating that elektra T cells are not growth arrested. After 48 hours of stimulation, fewer elektra CD8 T cells replicated DNA relative to wild type cells (Figure 3B, left panel). Under steady state conditions, an elevated percentage of elektra CD8 T cells incorporated BrdU (Figure 3B, right panel). Annexin V staining 48 hours after activation revealed massive apoptosis of mutant CD8 T cells (Figure 3C). However, γ-irradiation-induced apoptosis was equivalent in homozygous elektra and wild type CD8 T cells (Figure 3D), indicating that the elektra mutation selectively causes death in response to activation.
Examination of TCR signaling events in elektra CD8 T cells revealed intact calcium influx as well as normal NFAT-dephosphorylation, NF-κB nuclear translocation, and ERK and AKT phosphorylation. T cells from elektra homozygotes also exhibited normal induction of the activation markers CD25 and CD69 24 hours after TCR activation. However, p38-MAPK and JNK, which are activated through phosphorylation, were constitutively phosphorylated under basal conditions, and further phosphorylated upon TCR stimulation in homozygous elektra CD8 T cells (Figure 3E).
When CFSE-labeled elektra splenocytes were adoptively transferred into sublethally irradiated congenic wild type recipient mice, they failed to proliferate and were undetectable in the spleen 7 days after infusion. When spleens were collected 2 and 4 days after adoptive transfer, a marked excess of the elektra homozygous mutant cells were found to be apoptotic as indicated by Annexin V staining (Figure 4A). Elektra B cells reconstituted the spleen normally. Thymectomy was performed in order to block replenishment with new thymic emigrants and follow the fate of existing peripheral T cells.  Wild type T cell numbers declined by 30% following thymectomy and then remained stable for more than 60 days.  In contrast, homozygous elektra T cells declined dramatically within 12 days after thymectomy, and almost completely disappeared from the blood by 60 days after thymectomy (Figure 4B).
Elektra CD8 T cells displayed impaired homeostatic expansion. Injection of IL-7 together with non-neutralizing IL-7 antibody mimics the lymphopenic condition and results in expansion of both T cells and B cells. In wild type mice, IL-7/anti-IL-7 injection resulted in expansion of both B and T cells, and a reduction in the percentage of Annexin V positive CD8 T cells after 6 days. In elektra homozygotes, whereas B cells expanded normally, both CD8 and CD4 T cells failed to proliferate in response to IL-7/anti-IL-7 treatment (Figure 4C). In addition, in contrast to wild type mice, a higher percentage of CD8 T cells from homozygous elektra mice were positive for Annexin V 6 days after IL-7/anti-IL-7 than after PBS injection (Figure 4C, lower panel), suggesting that they die in response to signals induced by IL-7.
The cell surface glycoprotein CD44 normally accumulates on T cells as they mature in the periphery: naïve T cells that have newly emigrated from the thymus express low levels of CD44 while mature, expanding T cells express high levels of CD44. In vitro, experimental signals for homeostatic expansion also induce surface expression of CD44 (3-6). When CD44 expression was examined with Annexin V expression in T cells, most Annexin V positive cells were from the CD44high population, regardless of genotype. While similar small percentages of wild type and elektra CD44low cells were Annexin V positive, the percentage of homozygous elektra Annexin V positive CD44high cells was significantly elevated over wild type levels (Figure 4D). Together, these findings demonstrate that homozygous elektra CD44high CD8 T cells die in response to homeostatic expansion signals.
Homeostatic lymphoid cell death is mediated both by the intrinsic apoptotic signaling pathway (controlled by the balance of pro- and anti-apoptotic Bcl2 family members), and the extrinsic apoptotic signaling pathway (controlled by signals from receptors for TNF, FasL, TRAIL). No rescue of CD8 or CD4 T cells was observed in homozygous elektra; Faslpr/lpr double mutant mice, excluding the Fas-mediated extrinsic pathway as the mechanism of death of elektra T cells. However, CD4 and CD8 T cells death was completely rescued in homozygous elektra mice expressing a Bcl2 transgene (Figure 5A). The Bcl2 transgene could also partially restore the ability of homozygous elektra CD4 and CD8 T cells to proliferate in vitro in response to TCR stimulation (Figure 5B).  Consistent with the finding that apoptosis of homozygous elektra T cells is increased in the CD44high population, Bcl2 expression was specifically reduced in the CD44high population (Figure 5C). These results demonstrate that death of homozygous elektra T cells is mediated by the intrinsic apoptotic pathway through the action of Bcl2 family members, and that T cells from elektra homozygotes have the capacity to proliferate once their propensity for apoptosis is blocked.
The CD44high population encompasses both recently activated and memory phenotype cells, which may be discriminated on the basis of IL-2Rβ (CD122) expression (elevated in cells with a memory phenotype) (7). In wild type spleen, the CD44high CD8 T cell population was composed mostly of CD122+ cells, but in homozygous elektra spleen, most of the CD44high CD8 and CD4 T cells were CD122- (Figure 6A, left panel). Induction of homeostatic T cell expansion in homozygous elektra mice by injection of IL-7 and anti-IL-7 antibody increased the percentage of cells with this altered CD122 expression (Figure 6A, right panel). This CD44highCD122- population showed complete shedding of CD62L, absent expression of IL-7Rα (CD127), and low CD5 expression, suggesting that these cells were recently activated (Figure 6B). In addition, the CD44low (naïve) population of CD8 and CD4 cells in elektra homozygotes showed low expression of both IL-7Rα and CD62L (Figure 6B). Low expression of CD62L, as opposed to complete shedding of the CD62L molecule, has been shown to occur in response to continuous TCR stimulation (8;9). Similarly, low expression of IL-7Rα is associated with T cell activation. However, homozygous elektra CD8 T cells (both CD44high and CD44low) expressed normal levels of the CD69 activation marker (Figure 6B) suggesting that they are not fully activated. CD8 T cells from elektra mice were also not exhausted, as indicated by normal levels of PD1 expression (Figure 6B). These findings strongly suggest that homozygous elektra T cells exist in a partially activated state that predisposes them to apoptosis in response to activation stimuli.
Surprisingly, although the Bcl2 transgene rescued T cell death in elektra homozygotes, it did not prevent the development of a semi-activated phenotype by either CD44high or CD44low T cells from elektra homozygotes (Figure 6C, D). Moreover, in non-lymphopenic mixed bone marrow chimeras [homozygous elektra (Ly5.2) and wild type (Ly5.1) cells mixed 1:1 and transplanted into lethally irradiated Cd3 deficient recipients], elektra T cells showed the same semi-activated phenotype (Figure 6E, F). This indicates that the semi-activated state, although exacerbated by lymphopenic conditions, is not primarily driven by the lymphopenic environment in homozygous elektra mice. Rather, it is intrinsic to elektra T cells and apoptosis is a consequence.
Inflammatory monocyte recruitment is essential for defense against L. monocytogenes in mice (10). Since elektra homozygotes were highly susceptible to L. monocytogenes, inflammatory monocytes in these animals were examined before and during infection.  Prior to infection, elektra homozygotes showed a normal percentage of inflammatory monocytes in the bone marrow but slightly reduced percentages of these cells in the blood and spleen compared with wild type mice (Figure 7A, gated R1 population).  Following infection, the fraction of inflammatory monocytes increased far less in the blood and spleen of elektra homozygotes relative to the increase observed in wild type mice (Figure 7A).  No accumulation of inflammatory monocytes was observed in the bone marrow (Figure 7A), indicating that a migration problem did not cause this aspect of the elektra phenotype. In contrast to inflammatory monocytes, a normal percentage of neutrophils was found in elektra homozygotes under steady state conditions. An increased percentage of neutrophils was found in elektra homozygotes following Listeria infection, (Figure 7A, gated R2 population). The exaggerated accumulation of neutrophils after infection may be attributed to the elevated bacterial burden in elektra homozygotes, as previously shown for CCR2 deficient mice (10). Macrophage numbers and recruitment to the peritoneal cavity were normal in elektra homozygotes, as were macrophage responses to TLR ligands in vitro (data not shown). Dendritic cells (DCs) were present in normal numbers (data not shown), and when stimulated with unmethylated CpG oligonucleotides, these cells survived and upregulated the costimulatory molecule CD40 as effectively as wild type cells. In mixed bone marrow chimeras [homozygous elektra (Ly5.2) and wild type (Ly5.1) cells mixed 1:1 and transplanted into lethally irradiated Cd3 deficient recipients], both T cells and inflammatory monocytes derived from elektra homozygotes were markedly disfavored compared to cells of wild type origin, while equal percentages of B cells and neutrophils of homozygous elektra and wild type origin were observed (Figure 7B). Together, these findings indicate a cell-intrinsic, lineage-specific effect of the elektra mutation.
To determine whether the CD11b+Ly6Chi monocytes from homozygous elektra mice are more prone to apoptosis in response to bacterial stimuli, CD11b+Ly6Chi Ly6G- monocytes from the bone marrow of uninfected mice were cultured for 3 days in vitro in the presence of IFNγ and heat-killed L. monocytogenes, and then examined for activation and apoptosis markers by flow cytometry. After treatment, wild type monocytes displayed an increase in MHC class II expression (Figure 7C, upper panel) and nitric oxide (NO) secretion (Figure 7C, lower panel), and in their forward scatter/side scatter (FSC/SSC) profile, indicating that they became activated and differentiated (Figure 7D). The percentage of wild type Annexin V positive cells also decreased after treatment (Figure 7E). In contrast, the majority of homozygous elektra monocytes failed to upregulate MHC class II (Figure 7C, upper panel), and NO could not be detected in the culture medium (Figure 7C, lower panel). The percentage of Annexin V positive elektra monocytes increased, and indeed most appeared to be dead, as indicated by their low FSC/SSC profile (Figure 7D, E). These results demonstrate that similar to homozygous elektra T cells, monocytes also undergo apoptosis in response to activation.
Nature of Mutation
The elektra mutation was mapped to Chromosome 11, and corresponds to a T to A transversion at position 586 of the Slfn2 transcript, in exon 2 of 2 total exons.
130 -E--G--Y--I--Y--I--Y--V--K--S--W-
The mutated nucleotide is indicated in red lettering, and results in an isoleucine to asparagine change at amino acid 135.
Illustration of Mutations in
Gene & Protein
Protein Prediction
Schlafen2 is one of ten Schlafen (Slfn) protein family members that exist in mice and are thought to regulate thymocyte maturation and T cell activation. Except for schlafen-like 1 (Slfnl1), the genes encoding all known mouse Slfns are located in a cluster on Chromosome 11. Slfn genes have viral homologues in several orthopoxviruses, including vaccinia, variola, cowpox, camelpox, ectromelia, monkeypox and taterapox viruses (11;12). In vaccinia and variola virus strains, v-Slfn genes are fragmented (i.e. several ORFs encode portions of the gene), and no protein is detected with α-v-Slfn antibodies (12). The Slfn family has been subdivided into three groups based on overall amino acid sequence homology and protein size (11;13). Subgroup I consists of the three shortest members, Slfn1, Slfn2, and Slfn-like 1, which are about 350 amino acids in length. Subgroup II proteins, Slfn3 and Slfn4, are approximately 200 amino acids longer than subgroup I proteins, and are of intermediate length. Members of subgroup III, Slfns 5, 8, 9, 10, and 14, are 300 to 400 amino acids longer than subgroup II members due to large C-terminal extensions. Slfns 1-4 contain a core region of 120 residues that are nearly identical, with divergent sequences at the N terminus. 
Figure 8. Domain structure of SLFN2. The COG2685 domain partially overlaps the divergent AA domain. No other domains have been identified in SLFN2. The position of the elektra mutation is indicated.
Slfn proteins contain several motifs identified based on sequence similarity to other proteins (Figure 8). All Slfns possess a region, designated COG2685 (NCBI Conserved Domain Database), with similarity to a domain present in several ATP-dependent DNA helicases, transcriptional regulators, as well as several hypothetical proteins from both prokaryotes and eukaryotes (13). The domain, of unknown function but predicted to contain a helix-turn-helix DNA binding motif, has been referred to as the “Slfn box” (14). In Slfn2, this domain is positioned between amino acids 258-280.
Figure 9. Crystal structure of a model AAA domain, the second AAA domain of N-ethylmaleimide-senitive fusion protein (NSF). The N-terminal Rossman fold is the lower wedge-shaped structure and the C-terminal helical domain is in the upper left. UCSF Chimera model is based on PDB 1D2N, Lenzen et al., Cell 94, 525-536 (1998). Click on the 3D structure to view it rotate.
Adjacent and overlapping the COG2685 domain, Slfns contain a divergent AAA domain, which is related to the AAA (ATPases associated with various cellular activities) domain present in proteins of the AAA-containing family (11).  AAA-containing proteins are present in all kingdoms, and mediate a wide variety of cellular processes, including protein unfolding and degradation, disassembly of protein complexes, DNA replication, and microtubular transport [reviewed in (15)]. During cycles of nucleotide binding and hydrolysis, AAA domains undergo conformational changes that are transmitted to substrate proteins, either translocating or changing the conformation of these targets. The classical AAA domain is a 200-250 amino acid ATP-binding domain that assembles into oligomers (often hexamers) in a ring-shaped structure with a central pore. Crystallographic studies demonstrate that the AAA domain consists of two subdomains, an N-terminal α/β Rossman fold, and a C-terminal α-helical subdomain (Figure 9, PDB ID 1D2N). The functional diversity of AAA-containing proteins is provided by attachment to various other protein domains, interaction with accessory or adapter proteins, variable organization within oligomeric assemblies, and by changes in key amino acids in the AAA domain itself. The divergent AAA domain of Slfn2 is located between residues 259-359, and is presumed to have similar function to the AAA domain in binding ATP.
Although absent from Slfn1, Slfn2 and Slfn-like 1, other Slfn proteins contain a conserved “SWADL” domain of unknown function, having within it the five amino acid sequence Ser-Trp-Ala-Asp-Leu (14). In their C-terminal extensions, subgroup III Slfns also have eight short sequence motifs with weak homology and similar spacing as sequences found in the DNA/RNA helicase superfamily I, suggesting that they possess DNA/RNA helicase or structure remodeling activity (13).
Slfn2 contains 378 amino acids. The elektra mutation, which converts isoleucine 135 to asparagine, is found in a region of Slfn2 for which no functional domains have yet been identified.
Slfn genes are expressed at relatively low basal levels as detected by Northern blot analysis. Using RT-PCR, Slfn2 transcript is found most highly expressed in the thymus, spleen, and lymph node, with lower levels in the lung and heart (11). Slfn1 has a similar expression pattern. Among differentiating thymocytes, Slfn2 expression increases 5-10 fold during the CD4+CD8+ double positive (DP) to single positive (SP) transition. Slfn1 is upregulated 100-fold during this transition. Slfn1 and Slfn2 expression in T cells decreases upon activation with α-CD3 and α-CD28 antibodies. Subcellularly, overexpressed FLAG-tagged Slfn2 protein is exclusively cytoplasmic, similar to Slfn1 and Slfn4 (14;16). Slfn1 can also translocate to the nucleus (17). Subgroup III Slfns are exclusively nuclear.
The gene encoding the prototypic Slfn protein, Slfn1, was identified through subtractive hybridization between thymus cDNA libraries from transgenic mice in which thymocyte maturation was halted at the DP stage, and mice in which maturation was skewed toward CD4+ SP selection (11). Slfn1 is upregulated in positively selected, mature CD4+ and CD8+ thymocytes compared to immature progenitors. BLAST searches for similar gene and protein sequences identified eight more mouse Slfn genes, encoding Slfns 2, 3, 4, 5, 8, 9, 10, and 14 (11;13). Slfn-like 1 was found through EST database searches for tissue-specific genes (18).
The function of this newly discovered protein family has been investigated in several studies, but so far relatively little is known. Initial reports suggested that Slfns 1-3 have growth inhibitory properties. Ectopic expression of Slfn2 or Slfn3 in fibroblast cell lines greatly reduced cell proliferation, while Slfn1 modestly reduced proliferation (11;13;17). Slfn1 was shown to block the G1 to S phase transition of the cell cycle by inhibition of cyclin D1 at the transcriptional level (11;19). Nuclear localization of Slfn1, mediated by the chaperone DnaJB6, is reportedly required for Slfn1 to block the cell cycle (8). In vitro, DnaJB6 binds to Slfn1, stabilizes it, and promotes its translocation to the nucleus where it can inhibit cyclin D1. In yeast two-hybrid experiments, Slfn2 interacted with cyclin D-type binding protein-1 (Ccndbp1), a cyclin D1 inhibitor; cyclin dependent kinase 2-associated protein-1 (Cdk2ap1); protein inhibitor of activated STAT3 (PIAS3); Flt3 interacting zinc finger protein-1 (Fiz1); zinc finger protein 296 (Zpf496); and transforming acidic coiled-coil protein 3 (TACC3) (B. Neumann, Thesis Abstract, University of Queensland, Australia). Slfn2 colocalizes with Ccndbp1 and Cdk2ap1, and may cooperate with Ccndbp1 to inhibit cell proliferation.
In vivo, thymocyte-specific transgenic expression of Slfn1 cDNA (which does not result in overexpression, but alters the kinetics of expression such that DP thymocytes express levels normally found in SP thymocytes) resulted in a severe reduction of overall thymus size, with thymus cellularity reduced 70-90% and 97-99% in Slfn1-transgene+/- and Slfn1-transgene+/+ mice, respectively, relative to control animals (11). In addition, homozygous transgenic mice had greatly reduced numbers of SP thymocytes and splenic T cells due to disrupted progression through the cell cycle that prevents cell division (as measured by reduced BrdU incorporation), as well as increased apoptosis (as measured by increased TUNEL staining and reduced mitochondrial membrane potential). Despite these findings, another study demonstrated that Slfn1 and Slfn2 have no effect on cell proliferation, cell cycle progression, or cyclin D1 levels or promoter activity in fibroblast and myeloid cell lines (16). Slfn1-deficient mice have no obvious phenotype, with normal numbers of CD4+ and CD8+ thymocytes and peripheral T cells (11). Thus, the role of Slfns in regulating cell growth remains uncertain.
Transcripts encoding Slfns 1, 2, 4, 5, 8, 9, and 10 are expressed in macrophages, where they are upregulated upon treatment and activation with lipopolysaccharide (LPS), CpG oligodeoxynucleotides, or IFN-γ in vitro, or after infection with Listeria monocytogenes in vivo (13;20;21). The promoter region of Slfn2 is reported to contain binding sites for NF-κB and AP-1, which are necessary for Slfn2 induction by LPS or CpG treatment in peritoneal macrophages (20). Slfn2 is specifically upregulated in RAW264.7 cells upon infection with Brucella abortus (22). Increased mRNA expression of Slfns 1, 2, 4, 5, 8, 9, and 10 is also detected in terminally differentiated ER-MYB, FDB1, and M1 myeloid cell lines relative to their expression during the undifferentiated, proliferative state (13;14;16). These data suggest that Slfns contribute to myeloid differentiation and activation.
Several orthopoxviruses contain a Slfn orthologue, present either as a full length gene, or broken into fragments. The full length camelpox v-Slfn is most similar to mouse Slfn1 and Slfn2, with approximately 30% amino acid identity in the conserved region (12). v-Slfn has no effect on cell proliferation when stably expressed in NIH3T3 cells. A recombinant vaccinia virus strain expressing camelpox virus v-Slfn exhibits similar titers, plaque formation and plaque morphology as the wild type virus in cells infected in vitro (12). However, this recombinant virus is attenuated relative to the wild type virus when mice are infected intranasally, resulting in reduced weight loss and more rapid recovery. Clearance of the v-Slfn-expressing virus appeared to be accelerated relative to wild type virus, and increased numbers of lymphocytes were found in the lungs.
Although no human diseases have been associated with any of the Slfn genes, one study found that transcription of the Slfn gene cluster is increased in the joints of mouse models of rheumatoid arthritis, although the significance of this finding is unknown (23). The paternal gene of the mouse DDK syndrome (named for the mouse strain in which it is observed), located within the Ovum mutant locus, has been mapped to the Slfn gene cluster (24). The DDK syndrome, not found in humans, is characterized by lethality of early embryos from crosses between females of the DDK inbred strain and non-DDK males due to incompatibility between a maternal DDK factor and a non-DDK paternal gene.
Putative Mechanism
Figure 10. The semi-activated phenotype of homozygous elektra T cells predisposes them to apoptosis in response to activation signals.
Studies of elektra mice demonstrate that a Slfn2 mutation causes immunodeficiency as a consequence of T cell and monocyte apoptosis secondary to a semi-activated phenotype (Figure 10). Multiple findings strongly suggest that this semi-activated state in Slfn2elektra/elektra T cells stems from a loss of cellular quiescence. First, homozygous elektra T cells display an elevated percentage of cells replicating DNA both in the steady state and during the initial response to TCR stimulation, suggestive of an increased readiness to enter S phase. Second, unstimulated homozygous elektra T cells display a unique phenotype in which naïve (CD44low) T cells express low levels of CD62L and IL-7Rα, and mature (CD44high) T cells express no CD62L, IL-7Rα, or CD5, but normal levels of CD69, indicative of a semi-activated phenotype.
We hypothesize that Slfn2 maintains the quiescent state by promoting the expression of quiescence genes and/or repression of genes required for proliferation/differentiation. Previous work, primarily on the related Slfn1 protein (53% identical), supports this hypothesis. Thymocyte-specific transgenic expression of Slfn1 cDNA in mice resulted in severe reduction of overall thymus size due to impaired development of the CD4/CD8 double negative population in the thymus (11)In vitro, ectopic expression of Slfn1 in fibroblasts repressed cell growth and arrested cells in the G1 phase by causing transcriptional downregulation of cyclin D1 (11;13;18).  Notably, relative to Slfn1, Slfn2 overexpression caused much stronger growth repression (11). The Cyclin D genes are targets of FoxO transcription factors important in the regulation of quiescence (25), raising the possibility that function of Slfn2 intersects with the pathways controlling FoxO function.
Several studies suggest that lymphocytes must actively maintain a quiescent state in order to remain resistant to apoptosis. IL-15 has been shown to exert quiescence-inducing effects which are important for its anti-apoptotic properties (26). Members of the FoxO family of transcription factors (25) and LKLF (27;28) have also been shown to promote quiescence and loss of their activity leads to induction of apoptosis. However, several recent studies have challenged the idea that FoxO and LKLF regulate the quiescent state per se, presenting evidence that these proteins instead control expression of homing and survival proteins (29-31). These reports suggest that impaired trafficking and survival of T cells create a lymphopenic environment that leads to the loss of quiescence in peripheral T cells, and raise the question of whether Slfn2 actually controls survival or quiescence of T cells. Death of homozygous elektra T cells was rescued by transgenic expression of the anti-apoptotic Bcl2 gene. However, although T cells in these mice were not subject to a lymphopenic environment, they still showed the same semi-activated phenotype as T cells from homozygous elektra mice. Similarly, in mixed bone marrow chimeras, elektra T cells underwent apotosis and displayed a semi-activated phenotype, despite being in a normal, non-lymphopenic environment. Thus, the failure of mature homozygous elektra T cells to engage the anti-apoptotic machinery of the cell stems from loss of quiescence and acquisition of a semi-activated phenotype by naïve elektra T cells, and is not a consequence of impaired survival signaling or activation by a lymphopenic environment. IL-7 and IL-15 are critical T cell growth and survival factors, and the absence of their signaling due to loss of IL-7Rα and IL-2Rβ expression in CD44high T cells may be the trigger that activates the intrinsic apoptotic pathway in homozygous elektra T cells.
The activation of several pathways, including the NF-κB, NFAT, AKT, ERK1/2, JNK, and p38-MAP kinase pathways, regulates T cell survival. In the case of homozygous elektra mutants, we presume that as a result of a loss of quiescence, the balance between these pathways is altered to favor apoptosis rather than expansion. Activity of p38-MAPK and JNK kinases, which are constitutively activated in unstimulated elektra T cells, is associated with T cell apoptosis (32;33), and may represent a factor contributing to apoptosis of elektra T cells. The enhancement of p38-MAPK and JNK phosphorylation observed upon TCR stimulation of elektra T cells, and the fact that inflammatory monocytes (of the myeloid lineage) also undergo apoptosis upon activation, together suggest that the proliferative signals negatively regulated by Slfn2 may be distinct from the TCR signaling pathways that lead to the development of activated T cell effector functions and full T cell activation.
Homozygosity for Slfn2elektra confers susceptibility to two very different viral infections (LCMV and MCMV) in mice.  Susceptibility to LCMV is likely related to impairment of CD8 T cell function. Interestingly, the mutation has no effect on NK cell numbers or function, although these cells are known to be essential for defense against MCMV.  It is possible that inflammatory monocytes fulfill an essential function as well, given that lymphocytes are not needed for survival during the first several weeks following inoculation with MCMV (2).  Indeed, monocyte/macrophage cells are suggested to be protective in influenza infections (34), and inflammatory monocytes are essential for defense against L. monocytogenes infection (10;35). Recently, a non-redundant role was directly demonstrated for inflammatory monocytes in controlling viral infection (35).
Based on the current limited understanding of Slfn protein function, the elektra phenotype is somewhat unexpected. The conclusion that Slfn1 and Slfn2 inhibit cell cycle progression and cell proliferation, while still controversial, is not supported by the elektra phenotype, which indicates that Slfn2 is required for normal proliferation of CD8+ T cells in response to TCR stimulation, PMA/ionomycin treatment, and LCMV infection. Transgenic expression of Slfn1 in T cells results in increased apoptosis. In contrast, elektra T cells fail to repopulate irradiated congenic recipient mice, and are observed to undergo increased apoptosis both in vivo and in vitro. Furthermore, while transgenic Slfn1 expression was shown to block thymocyte maturation at the CD25+CD44- stage (11), thymic T cell development appears normal in elektra mice, as evidenced by normal percentages of CD25+, CD44+, or CD3+ thymocytes, and single and double positive thymocytes.
Whether Slfn proteins have enzymatic, scaffolding, adapter, or other functions is unknown. Slfn1- (11) and Slfn3-deficient mice show a normal immune system phenotype. Thus, Slfn2 has a non-redundant role in regulation of peripheral T cell and inflammatory monocyte survival, while Slfn1 and Slfn3 do not.
Primers Primers cannot be located by automatic search.
Elektra genotyping is performed by amplifying the region containing the mutation using PCR, followed by sequencing of the amplified region to detect the single nucleotide transition.
PCR program
1) 95°C             2:00
2) 95°C             0:30
3) 56°C             0:30
4) 72°C             1:00
5) repeat steps (2-4) 29X
6) 72°C             7:00
7) 4°C               ∞
Primers for sequencing
elektra_seq(F): 5’- AATTCGGGAGGTGGAGCAATC -3’
The following sequence of 1310 nucleotides (from Genbank genomic region NC_000077 for linear genomic sequence of Slfn2) is amplified:
4214               gctgggt ctaaatctcg gagcaatcac ttttggagag aaggatagga
4261 agaaaatgaa gaattctcac ctcagaaaac aggagaatgc aaacatctct ctagctgtat
4321 gtgctctcct gaattcggga ggtggagcaa tcaaggttaa aattgaaaat gaaaattata
4381 gtctcactag ggatggcctg ggactagatt tggaagcctc tctttgtaaa tgtctgccct
4441 ttgtccagtg gcacctggac ttcacggaga gcgaaggcta catttatatc tacgtgaaat
4501 cgtggagcca agaaatcttt gggctgccta ttggcaccct aagaaccaat ttgtatgtaa
4561 ggagcatgtc atcttctgta caagtgagcg ccgctgctgc cctggaattt ctccaggacc
4621 tggaggaaac tggagggaga ccctgtgtca gaccagagtt gcctgcaagc atagccttcc
4681 ctgaagtgga aggagaatgg cacctggagg atttggctgc tgcattgttt aacaggacag
4741 aatttcagta cgaggaaact ttccccttta ccagatccag atatgttgaa gttacattgc
4801 tttcagcgaa acgcctgcga aaacgcatca aagagctcct ccctcaaact gtttctgctt
4861 ttgcaaacac ggatggggga tttttgttca ttggtttgga tggcaaaacc cagcaaatta
4921 ttggttttga agcagagaag agcgatctcg tgcttctaga gagtgaaata gaaaagcaca
4981 tccggcagct gcctgtcact cacttctgtg aggagaagga gaagatcaag tacacgtgca
5041 aattcatcga agtgcacaaa tccggagctg tgtgtgcata tgtgtgtgcg ctcagagtgg
5101 agagattctg ctgtgcagta ttcgctgcag agcccgagtc ctggcatgtg gaaggcggct
5161 gtgtgaagag gtttaccaca gaggaatggg tgaaactcca gatgaatgcc ccatcaggtt
5221 gaagggggat taataatctg tatgagggcg gcagagatgg ctcagtggct aagagcactg
5281 actgctcttc cagagatctg agctgtaacc cccccccccc agcctgaaac ctggttgctc
5341 aggtttcact gttagcaaga agagagcatg ggagaggggg gaggagcctg ccgccctatc
5401 gttcatccta ccctttgtcc acccgagact agaggggtgt gtccgttcca cgcagacaaa
5461 ggggcccaga gggtctgtgg cagtggtcac ctgaaactgg actccaggat gatttggcgg
5521 gaa
Primer binding sites are underlined; sequencing primer binding sites are highlighted in gray; the mutated T is indicated in red.
Science Writers Eva Marie Y. Moresco
Illustrators Diantha La Vine
AuthorsMichael Berger, Bruce Beutler
Edit History
2011-05-19 1:44 PM (current)
2011-01-17 4:10 PM
2011-01-04 3:19 PM
2010-12-30 4:05 PM
2010-12-23 1:43 PM
2010-03-15 10:59 AM
2010-02-08 12:01 PM