Phenotypic Mutation 'sphinx' (pdf version)
Mutation Type missense
Coordinate48,729,543 bp (GRCm39)
Base Change G ⇒ T (forward strand)
Gene Gimap5
Gene Name GTPase, IMAP family member 5
Synonym(s) E230026N22Rik
Chromosomal Location 48,723,131-48,731,134 bp (+) (GRCm39)
MGI Phenotype FUNCTION: This gene encodes a protein belonging to the GTP-binding superfamily and to the immuno-associated nucleotide (IAN) subfamily of nucleotide-binding proteins. In humans, the IAN subfamily genes are located in a cluster at 7q36.1. [provided by RefSeq, Jul 2008]
PHENOTYPE: Mice homozygouse for a knock-out allele display defects in lymphocyte development with hematopoietic defects and reduced life span. Mice homozygous for an ENU-induced allele exhibit premature death associated with extramedullary hematopoiesis in the liver, anemia, cachexia, and colitis. [provided by MGI curators]
Accession Number

NCBI RefSeq: NM_175035; MGI: 2442232

Limits of the Critical Region 47735021 - 50663498 bp
Amino Acid Change Glycine changed to Cysteine
Institutional SourceBeutler Lab
Gene Model not available
AlphaFold Q8BWF2
SMART Domains Protein: ENSMUSP00000056820
Gene: ENSMUSG00000043505
AA Change: G38C

Pfam:AIG1 27 240 5.4e-80 PFAM
Pfam:MMR_HSR1 28 151 9.5e-8 PFAM
transmembrane domain 283 305 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using ENSMUST00000055558)
Meta Mutation Damage Score Not available question?
Is this an essential gene? Non Essential (E-score: 0.000) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status loading ...
Single pedigree
Linkage Analysis Data
Penetrance 100% 
Alleles Listed at MGI
All alleles(4) : Targeted, knock-out(1) Targeted, other(2) Chemically induced(1)
Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL00899:Gimap5 APN 6 48730107 missense possibly damaging 0.80
IGL01936:Gimap5 APN 6 48729999 missense probably damaging 1.00
IGL01995:Gimap5 APN 6 48729727 missense probably damaging 1.00
IGL02371:Gimap5 APN 6 48729937 missense probably damaging 1.00
IGL02974:Gimap5 APN 6 48730311 missense possibly damaging 0.47
R0321:Gimap5 UTSW 6 48727449 splice site probably benign
R1480:Gimap5 UTSW 6 48729964 missense probably damaging 1.00
R1655:Gimap5 UTSW 6 48730110 nonsense probably null
R1761:Gimap5 UTSW 6 48730195 missense probably damaging 1.00
R7449:Gimap5 UTSW 6 48729838 missense probably damaging 1.00
R8519:Gimap5 UTSW 6 48730068 missense probably benign
R8525:Gimap5 UTSW 6 48729501 missense probably benign 0.02
Z1177:Gimap5 UTSW 6 48729819 missense possibly damaging 0.90
Mode of Inheritance Autosomal Recessive
Local Stock Embryos
MMRRC Submission 030019-UCD
Last Updated 2018-01-05 3:14 PM by Eva Marie Y. Moresco
Record Created unknown
Record Posted 2009-12-23
Phenotypic Description
Figure 1. Lymphopenia, anemia, and neutrophilia in sphinx homozygotes. (A) The percentage of CD+ T and NK1.1+ NK cells, and CD4+ and CD8+ T cells in the spleen. (B, C, D) Circulating concentrations of hemoglobin (B), lymphocytes (C), and neutrophils (D) in the blood.
The sphinx phenotype was discovered among ENU-induced G3 mutant mice in a genetic screen designed to detect defective in vivo natural killer (NK) cell and cytotoxic T lymphocyte (CTL) responses (In Vivo NK Cell and CD8+ T Cell Cytotoxicity Screen) (1). Homozygous sphinx mice failed to kill adoptively transferred NK and CTL target cells. Further analysis demonstrated a near complete absence of peripheral NK cells and CTLs and a reduction in total T cell numbers in homozygous sphinx mutants (Figure 1A). Sphinx mice developed progressive normocytic anemia [with increased variation in red blood cell width (anisocytosis) and thrombocytopenia] concurrent with the development of peripheral lymphopenia and neutrophilia (Figure 1B-D).
Figure 2. Wasting disease and colitis in sphinx mice. (A) Weight of co-housed C57BL/6J (n=6) and sphinx (n=6) males at the indicated ages. (B) H&E stained sections from the colon of 10 week old co-housed mice. Scale bar, 100 µm.
After 4 weeks of age in males and 6 weeks of age in females, sphinx homozygotes were observed to develop diarrhea and rapidly lose weight (Figure 2A). Cell infiltration and inflammation occurred in the colon by 6 weeks of age, and severe colitis developed by 10 weeks of age, likely contributing to the death of sphinx mutants by 14 weeks of age (Figure 2B).
Figure 3. Liver morphology of sphinx mice. (A) Tissue appearance at the indicated ages. (B) H&E staining of liver sections from 40 day old mice. Scale bar, 200 µm.
Sphinx livers exhibited an abnormal morphology, with extramedullary hematopoiesis associated with foci of hematopoietic cells (Figure 3). Hyperplastic nodules, containing well-differentiated hepatocytes bordered by clusters of hematopoietic cells, developed in adult sphinx mice. During embryonic development, the size and cellularity of livers from embryonic day (E)16 sphinx mice were similar to those of wild type livers. However, an increased percentage of hematopoietic cells was found at postnatal day 8, and extramedullary hematopoiesis was detected by day 14 in the livers of sphinx homozygotes. Thus, some hematopoietic cells failed to migrate from the fetal liver and underwent spontaneous development in sphinx mice. Although a disrupted splenic architecture was observed, no extramedullary hematopoiesis was found in the spleens of sphinx mice. Serum bilirubin and albumin concentrations were normal in sphinx mice at 8 weeks of age, suggesting that liver metabolic function was intact.
The bone marrow of five-week old sphinx mice contained normal numbers of Lin-Sca1+cKit+ hematopoietic stem cells (HSCs) and hematopoietic progenitor cells. However, the numbers of megakaryocyte/erythroid precursors (MEP) declined with age, in correlation with the onset of anemia and thrombocytopenia in adult Gimap5sph/sph mice. Also, despite normal numbers of common lymphoid progenitors (CLP) in the bone marrow, rapid atrophy of thymus mass and cellularity occurred with age. More dramatic differences were observed in the sphinx liver. Examination of hematopoietic progenitor cell populations revealed that a significant number of HSCs and hematopoietic precursors, representing all hematopoietic lineages, remained in the adult sphinx liver. The most abundant of these was the Lin-Sca1+cKit-IL-7Ra+ CLP and it is noteworthy that interleukin (IL)-7Ra expression was increased in this population.
Figure 4. Defects in sphinx hematopoietic cells in fetal liver chimeras. Numbers of the indicated cell types in the spleens of irradiated Rag1-/-Il2rγ-/- recipient mice reconstituted with E19 fetal liver cells of the indicated genotype and analyzed after six weeks.
The function of sphinx HSCs was investigated in bone marrow and fetal liver chimeras. Sphinx bone marrow cells poorly reconstituted the hematopoietic compartment of irradiated recipient mice. To exclude the possibility that defective sphinx HSC function was a secondary consequence of inflammation and wasting disease, irradiated Rag2-/-Il2rγ-/- recipients were reconstituted with E19 fetal liver cells. Sphinx E19 fetal liver cells failed to reconstitute the thymus; splenic CD8+ T, B, and NK cell numbers were markedly reduced, while CD4+ T cells repopulated the spleen at normal numbers (Figure 4). Recipients of sphinx E19 liver cells exhibited weight loss and wasting, extramedullary hematopoiesis, and the accumulation of granulocytes (CD11b+Gr1+ myeloid cells and Ter119+ nucleated erythrocytes) in the spleen, while recipients of control fetal liver cells appeared healthy and without hematopoietic defects. Together, these findings indicate a cell-intrinsic defect in lymphopoiesis, and suggest that early mortality of sphinx mice is at least partially due to defective hematopoiesis.
Thymocyte and lymphocyte development were examined in homozygous sphinx mice. Largely normal thymocyte development was observed, including development of CD4+ T cell, CD8+ T cell, γδ T cell, Foxp3+ regulatory T cell, and iNKT cell lineages.  Thymi of sphinx mice contained normal numbers of total thymocytes and normal frequencies of CD4-CD8- double negative (DN, stages 1-4), CD4+CD8+ double positive (DP), CD4+ single positive (SP), and CD8+ SP cells. However, sphinx CD8+ SP, but not CD4+ SP, thymocytes failed to upregulate IL-7Rα or downregulate CD24 and CD69, hallmarks of terminal thymocyte maturation. CD4+ T cells from the spleen of sphinx mice also displayed reduced IL-7Rα expression, and failed to proliferate in response to TCR stimulation. In addition, as lymphopenia progressed with age, sphinx CD4+ T cells began to express cell surface markers characteristic of cells undergoing lymphopenia-induced proliferation, including low CD62L and high CD69. The bone marrow contained a high frequency of NK cells expressing the activation marker CD69.
When the sphinx mutation was bred onto the H-Y antigen-specific TCR transgenic background, H-Y-specific thymocytes were appropriately deleted in male mice, but female mice had reduced H-Y-specific CD8+ SP thymocytes relative to wild type. This finding suggests that sphinx thymocytes are less sensitive to positive selection, or more sensitive to negative selection. H-Y-specific CD8+ T cells were found in reduced numbers and frequencies in the spleens of male and female sphinx mice, indicating that lymphopenia occurred irrespective of TCR specificity for self-antigen. 
Figure 5. Reduced iNKT survival and aberrant αGalCer cytokine responses in sphinx mice. (A) iNKT cells in the spleen of six-week old sphinx mice were identified by αGALCer-tetramer binding and quantified by flow cytometry. (B) Serum cytokine concentrations of IFN-γ and IL-4 were measured by ELISA 90 min after injection of the iNKT cell agonist αGalCer. (B) Cytokine production was also measured on a per cell basis by intracellular staining for IFNγ, TNF-α, IL-4, or IL-13 expression among αGalCer-tetramer iNKT cells. Same color scheme applies in (B) and (C).
Peripheral iNKT cells in sphinx homozygotes declined in number with age, and expressed reduced amounts of NK1.1, which is required at a late step in iNKT differentiation following export from the thymus (Figure 5A). These iNKT cells responded aberrantly to αGalCer stimulation, producing little interferon (IFN)-γ or tumor necrosis factor (TNF)-α, and reduced amounts of IL-4 and normal amounts of IL-13 on a per cell basis (Figure 5B, C).
Figure 6. B cell defects in sphinx mice. (A) Percentage of IgMhighIgDint naive and IgMlowIgDhigh mature B cells among CD19+ splenocytes from 6 week old mice. (B) Mice were immunized with 50 µg NP36-CGG with alum/LPS or 50 µg of NP50-Ficoli to assess T-dependent (left) or T-independent (right) antibody responses. NP-specific serum antibodies were measured 14 days after immunization by ELISA. (C) Cell cycle progression of B cells in the presence or absence of LPS measured by propidium iodide incorporation after 48 hours of incubation.
CD19+ splenic B cells are reduced in number in sphinx mice (Figure 6A), but to a lesser extent than T cells or NK cells. The percentages of marginal zone B cells and follicular B cells were similar to those of C57BL/6J control mice.  B1 peritoneal B cells were nearly absent in sphinx mice. Upon immunization, sphinx homozygotes failed to mount T-dependent antigen-specific IgG1 or T-independent antigen-specific IgM responses (Figure 6B). Homozygous sphinx B cells failed to proliferate in response to BCR stimulation or treatment with PMA/ionomycin, but expanded normally in response to LPS treatment or CD40 ligation. Normal degradation of the NF-κB inhibitor IκB and normal phosphorylation of the ERK1/2, JNK, p38 MAP kinases, and Akt occurred in PMA/ionomycin-activated homozygous sphinx B cells. Ex vivo cell cycle analysis of sphinx B cells showed an increased percentage of B cells in S phase and a reduced number of B cells in G2 phase (Figure 6C). When proliferation was induced with LPS, an equivalent number of sphinx B cells entered G2 phase, yet an increased percentage of cells remained in S phase (Figure 6C).
Localized inflammation in the gastrointestinal tract of sphinx mice suggested the possibility that aberrant responses to intestinal flora may be involved the lymphopenia or wasting disease of these animals. When treated continuously with antibiotics from birth, intestinal inflammation, weight loss, and the accumulation of CD11b+ myeloid cells were prevented. However, lymphopenia, acquisition of a CD44highCD62Llow phenotype by CD4+ T cells, hepatic extramedullary hematopoiesis, and hepatocyte hyperplasia were not affected by antibiotic treatment.
To investigate the involvement of lymphocytes in immune pathology of sphinx mice, adoptive transfer of wild type splenocytes into young sphinx mice, prior to the onset of wasting disease, was performed. Recipients did not succumb to early death or develop wasting disease. Adoptively transferred splenocytes promoted the maintenance of endogenous sphinx CD4+ T cells, but not CD8+ T cells. In the CD4+ T cell compartment, cells of sphinx origin retained a CD44highCD62Llow phenotype. Adoptive transfer of wild type cells into sphinx mice also reduced extramedullary hematopoiesis, accumulation of common myeloid progenitors in liver and bone marrow, and prevented anemia and thymic atrophy.  No reduction in mortality or wasting disease was observed when sphinx mice received Rag-/- splenocytes, indicating that a lymphocyte population is required to prevent wasting disease. As adoptively transferred Cd4-/-, Cd8-/- or Ja18-/- (iNKT cell-deficient) splenocytes, or anti-NK1.1 antibody depleted splenocytes, could each prevent weight loss and early mortality, multiple lymphocyte populations might contribute to the suppression of wasting disease.

Further studies with the sphinx mouse found that CD4+ T cells are required for the development of colitis (2).  In addition, the CD4+ T cells from the sphinx mice had a LIP phenotype as well as a loss of Forkheadbox group O (Foxo)1, Foxo3, and Foxo4 expression. Examination of all lymphocyte populations from the spinx mouse found that, similar to the T cells, all cells lost Foxo expression.  The loss of Foxo expression subsequently led to reduced amounts of (and loss of function in) Foxp3+ Treg cells. Examination of CD4+ T cells from the sphinx mice found that at around 8 weeks the T cells are unable to proliferate, but they can produce proinflammatory cytokines (2)

Nature of Mutation
The sphinx mutation was mapped to Chromosome 6 and corresponds to a G to T transversion at position 352 of the Gimap5 transcript, in exon 3 of 3 total exons.
33  -G--K--S--G--C--G--K--S--A--T--G-
The mutated nucleotide is indicated in red lettering, and causes a glycine to cysteine substitution at amino acid 38 of the Gimap5 protein.
Illustration of Mutations in
Gene & Protein
Protein Prediction
Figure 7. Domain structure of mouse Gimap5. The extent of the AIG1 domain, which contains G1-G5 motifs, the conserved box (CB), and the IAN domain, is indicated with a dark blue box. The two colied coil motifs are indicated with hatched boxes. The position of the sphinx mutation is shown with a red asterisk.

The eight members of the mouse Gimap (GTPase of the immunity-associated protein) or IAN (immune-associated nucleotide-binding protein) family are GTP-binding proteins that regulate the survival of immune cells through interactions with apoptotic regulatory proteins [reviewed in (3)]. The family, not present in invertebrates or yeast, is encoded by eight genes located in a tight cluster on mouse Chromosome 6 and human Chromosome 7, an arrangement conserved in zebrafish, chicken, and rat (4-6). A cluster of Gimaps is also found in plants; two plant Gimaps, AIG1 and AIG2, are induced during the defensive response to bacterial infections (4;7). Among the eight mouse Gimaps, Gimap3 and Gimap5 show the highest similarity, with 83.8% identical amino acids (4;5).

Gimap5 encodes a 308-amino acid protein (Figure 7). Like the other Gimap family members, Gimap5 contains an AIG1 (avrRpt2-induced gene 1) domain (amino acids 27-238) in its N terminal half, which consists of a GTP-binding domain, a conserved box (CB) domain, and an IAN domain (4;5). The GTP-binding domain, common to other GTP/GDP-binding proteins, is composed of five short motifs designated G1-G5. Gimap5 has not been demonstrated to either bind GTP/GDP or hydrolyze GTP. The CB domain (amino acids 105-126) is a hydrophobic sequence of amino acids (LSxPGPHALLLVxQLGRθTxEψ; θ=aromatic, ψ=acidic) located between G3 and G4, and has unknown function. The secondary structure of the CB domain is predicted to be an extended sheet flanked by random coiled regions, and has been proposed to help maintain the three dimensional structure of the AIG1 domain, similar to the corresponding region between the G3 and G4 motifs of the GTPase H-ras [discussed in (5)]. The IAN domain (amino acids 183-190), also of unknown function, consists of sequence RxxxθNN R/K A/E. There is partial overlap between the G5, IAN domain, and a coiled coil domain (amino acids 187-221). A second coiled-coil motif, typically used for oligomerization, lies adjacent and C terminal to the AIG1 domain (amino acids 239-265). The C terminus of Gimap5 contains a hydrophobic region that may serve as a transmembrane domain (amino acids 283-305) (4;5).
The sphinx mutation results in a glycine to cysteine substitution at position 38, which is lies within the G1 sequence of the AIG1 domain. A glycine at this position is conserved in all eight mouse Gimap proteins, and in all currently annotated AIG1 domain-containing genes.

In general, Gimap5 mRNA is predominantly expressed in immune tissues in both mice and humans. Northern blot analysis demonstrates high levels of human GIMAP5 mRNA expression in spleen, lymph nodes, lung, and placenta (5;8). Human GIMAP5 transcript is also detectable at lower levels in digestive tract, skeletal muscle and heart. Peripheral blood leukocytes, thymus and bone marrow show only weak expression, and leukemia and lymphoma cell lines express no human GIMAP5 mRNA. In mice, Gimap5 transcript is abundantly expressed in spleen, lymph nodes, and lung, with a low level of expression in the thymus, as measured by RT-PCR (4). In rats, Gimap5 mRNA is detected by Northern blot in the spleen, thymus and lymph nodes, but not in the kidney (6). Gimap5 mRNA is highly expressed in CD4+, CD8+, and B220+ mouse lymphocytes from spleen and thymus, but not in Mac1+ myeloid cells or in NK cells (4;9). In mice, Gimap5 mRNA expression increases during the development of DN into DP thymocytes, and during the transition from DP to CD4+ or CD8+ SP thymocytes (4;9). In primary human T cells, Gimap5 mRNA is upregulated upon activation with phytohemagglutinin (PHA) plus IL-2 (10).

When overexpressed in 293T cells or in thymocyte cell lines, epitope-tagged Gimap5 displays a subcellular localization to the endoplasmic reticulum, Golgi and mitochondria, where Gimap5 is anchored by its C-terminal hydrophobic domain (4;8;10;11). Gimap5 has also been found at the centrosomal region (11). In rat primary T lymphocytes, endogenous Gimap5 is associated with a sedimentable subcellular fraction distinct from mitochondria and ER, as detected with a polyclonal antibody against amino acids 2-15 of rat Gimap5 (12).

Lysates from a mouse NKT cell line, C1498 (13), were separated using density separation and were immunoblotted for GIMAP5 (14).  Wong et al. observed that the distribution of GIMAP5 through the gradient was similar to lysosomal makers.  Similar to the findings stated above, the distribution of GIMAP5 did not match marker proteins for either the ER or mitochondria.  Examination of the localization of human GIMAP5 in CEM cells mimicked the mouse findings in that there was strong colocalization of GIMAP5 with lysosomal proteins (i.e. LAMP1 and LAMP2).  Electron microscopy following immunogold labeling of GIMAP5 showed GIMAP5 on multivescular bodies.  Wong et al. speculate that GIMAP5 is localized in the lysosome to provide protection against cell death (14).


Gimap5 (originally designated the lymphopenia or lyp locus) was identified by positional cloning as a gene mutated in the diabetes-prone BioBreeding (BBDP or BB) rat first described in 1978 (6;15), and known to develop progressive lymphopenia as well as autoimmune diabetes. In these animals, a single nucleotide deletion in the codon for amino acid 85 causes a frameshift that results in the addition of nineteen aberrant amino acids followed by a premature stop codon (6)

BB rats display cell-intrinsic T lymphopenia, with reduced numbers of CD4+ T cells, severely reduced numbers of NK and NKT cells, and a complete lack of CD8+ T cells (16-19).  BB rats, and rats on the F344 background carrying the mutant lyp locus, have a reduced frequency of DP thymocytes and an increased frequency of DN thymocytes (17;20;21). Knockdown of Gimap5 expression in mouse immature DN thymocytes disrupts the generation of DP thymocytes in fetal thymus organ culture (4). As mentioned above, Gimap5 expression increases during the T cell transitions from the DN to DP stage, and from the DP to SP stage (4;9). Thus, the reduced numbers of CD4+ and CD8+ T cells may be caused in part by defective intrathymic maturation and selection. In addition, reduced thymic output of T cells (17;22), and increased apoptosis of mature T cells (see below) likely contribute to T lymphopenia in BB rats.  Although the population of peripheral T cells in BB rats is greatly reduced, it is also relatively stable due to a combination of high mitotic activity and rapid turnover relative to control T cells (16;22;23)

Figure 8. Lymphocyte apoptosis. Death receptors, including Fas or TNFR1, are triggered by death ligands and activate caspase-8/10 through FADD and TRADD. In type I cells characterized by high levels of caspase-8, caspase-8/10 activates caspase-3/6/7, leading directly to cell apoptosis. Type II cells have lower levels of active caspase-8 and require an additional amplification loop. In type II cells, caspase-8/10 leads to the cleavage of BID and assembly of Bax and Bak into heterodimeric pores in the outer mitochondrial membrane. This causes the release of cytochrome c, which assembles with APAF1 into the apoptosome complex. Caspase-9 is activated by the apoptosome, leading to caspase-3/6/7 activation. Gimap5 is localized to the mitochondrial membrane, where it may regulate the function of Bcl-2, Bcl-xL, and Bax through direct interactions. 

Several reports suggest that Gimap5 is a negative regulator of apoptosis in lymphocytes (Figure 8). Low or absent Gimap5 mRNA expression is found in several leukemia and lymphoma cell lines, and the cluster of human GIMAPs is located within a region on chromosome 7q35–7q36.1 frequently deleted in acute myeloid leukemias (5;24). When cultured in vitro, peripheral T cells and mature SP thymocytes from BB rats display increased rates of apoptosis relative to control cells (16;25-27).  Enhanced apoptosis is also observed after knockdown of Gimap5 expression in either Jurkat T cells or in the IL-2-dependent 23-1-8 T cell line (4;27). Conversely, Gimap5 expression can prevent apoptosis induced by okadaic acid (11). Mitochondrial dysfunction causes apoptosis, and mitochondrial membrane potential is reduced in T cells from BB rats (27). However, one group has reported observing apoptosis when Gimap5 was overexpressed in Jurkat T cells, or in naïve, but not in activated, primary human T cells (10). Gimap5 may exert distinct effects depending on the activation status of lymphocytes and availability of growth factors.

T cell survival may be influenced by Gimap5 through several possible mechanisms. Disruption of mitochondrial outer membrane integrity leads to the release of pro-apoptotic proteins and signaling resulting in apoptosis (28). Gimap5 is localized to the mitochondrial membrane, and can be coimmunoprecipitated from 23-1-8 T cells together with the anti-apoptotic Bcl-2 and Bcl-xL (4). In cells stimulated to undergo apoptosis, Gimap5 coimmunoprecipitates with the pro-apoptotic protein Bax. Treatment of adult thymic organ cultures from BB rats with a general caspase inhibitor prevents apoptosis of thymocytes (25). Thus, Gimap5 may physically associate with and regulate the components of the mitochondria-mediated apoptosis pathway. A recent report indicates that calcium influx through the plasma membrane calcium release-activated calcium (CRAC) channel is reduced by 30-40% in T cells or thymocytes from BB rats after treatment with thapsigargin or TCR crosslinking (29). Self-peptide: MHC complexes that normally stimulate relatively small Ca2+ responses and promote cell survival may elicit an insufficient Ca2+ response to support T cell survival in BB rats (29).

Gimap5 has also been proposed to control T cell survival by regulating the activation state of T cells.  T cells in the BB rat (mostly CD4+ thymocytes and the few remaining peripheral T cells) exist in a partially activated state characterized by decreased expression of CD62L, and increased expression of MHC class I, MHC class II, and CD28 (30). These cells readily progress through the cell cycle to S phase, but fail to divide, and instead undergo apoptosis. Strongly activating BB T cells through TCR/CD28 stimulation in vitro rescues them from spontaneous death and stimulates proliferation (16;27;30).  These data suggest that Gimap5 is required to maintain T cells in a state of quiescence, and without Gimap5 cells enter a partially activated state driving them to abortive death rather than proliferation. The partial activation phenotype appears to be due in part to activation of a MEK-IKK-NF-κB pathway (31).

Mutation of Gimap5 appears to cause a predisposition to autoimmune disease. The BB rat spontaneously develops autoimmune insulin-dependent diabetes, which is T cell dependent and typically accompanies T lymphopenia (32-34). It is now known that lymphopenia and diabetes are genetically separable phenotypes, and genes other than Gimap5 have protective effects against diabetes in BB rats (20;35).  For example, the MHC RT1 allotype lv1 allele is sufficient to protect lymphopenic rats from diabetes, while the RT1 u allele is permissive for diabetes (36). When the lyp and u loci are bred to homozygosity on the rat PVG background, the animals develop a spontaneous, progressive inflammatory bowel disease resembling human eosinophillic gastroenteritis, with increased serum IgE and production of IgG autoantibodies (37). Impaired development of regulatory T cells may account for the development of autoimmune diabetes in BB rats. While recent emigrant naïve T cells die by apoptosis, autoreactive T cells may be spared from death by TCR-mediated activation and a scarcity of regulatory T cells. In support of this hypothesis, RT6-expressing peripheral regulatory T cells are lacking in BB rats, and depletion of RT6-expressing cells together with IFN-α treatment of control rats results in T lymphopenia and development of diabetes (38;39). BB rats also exhibit reduced numbers of peripheral CD4+CD25+Foxp3+ regulatory T cells, and adoptive transfer of wild type regulatory T cells protects BB rats from diabetes (40).

In humans, a single nucleotide polymorphism (rs6598) located in the polyadenylation signal of GIMAP5 has been associated with the presence of significant levels autoantibodies against the islet cell protein IA-2, in a case-control study of type I diabetes patients (41). This polymorphism was not found to be associated with type I diabetes itself, with autoantibodies against insulin or the islet protein glutamic acid decarboxylase (GAD) 65, or with age at clinical onset. Another polymorphism within the polyadenylation signal has been associated with a significant increase in risk for systemic lupus erythematosus (SLE) (42). The polymorphism produces an inefficient polyadenylation signal resulting in increased frequency of non-terminated transcripts in homozygous individuals. 

A targeted Gimap5-null mouse mutant has recently been generated (43). It recapitulates the lymphopenia observed in BB rats, with 20-fold and 2-fold reductions of splenic CD8+ and CD4+ T cells, respectively, and greater reductions in lymph node. Lymphopenia was demonstrated to stem from cell-intrinsic, impaired intrathymic maturation and increased apoptosis of peripheral (splenic but not blood) CD4+ and CD8+ T cells. The frequencies of DN1-4 and DP thymocytes are normal, but there are reduced CD4-CD8+CD69lowQa2high cells and increased CD4-CD8+CD69lowQa2low cells, indicating a defect during the final stages of intrathymic CD8+ thymocyte maturation. The residual CD8+ and CD4+ T cells of Gimap5-null mice have reduced CD62L and increased CD69 expression, demonstrating an activated phenotype similar to that observed in T cells from BB rats. The percentage of CD4+CD25+Foxp3- lymphocytes is increased 3-fold and 12-fold, respectively, in the spleen and inguinal lymph node, consistent with a state of partial effector T cell activation. However, normal to slightly increased frequencies of CD4+CD25+Foxp3+ T cells are found in spleen and inguinal lymph node. Gimap5-/- mice display an 80% reduction of NK cell numbers (CD122+CD3- or NK1.1+DX5+), as well as a significant reduction of NKT cells (CD122+CD3ε3+ or CD1-αGalCer+CD3ε+) in the bone marrow and spleen, a cell-extrinsic defect that can be corrected by transfer of bone marrow into a wild type environment. The residual NK cells display reduced levels of NKG2D and CD11b, and lack expression of Ly49A, Ly49D, or Ly49G2. Gimap5-/- mice display chronic hepatic hematopoiesis and hepatocyte apoptosis leading to liver failure, which are not observed in BB rats. Gimap5-/- livers become discolored and develop nodules, and contain greatly increased numbers of mononuclear and polymorphonuclear cells and hematopoietic progenitors. The liver phenotype cannot be transferred with Gimap5-/- bone marrow into wild type recipients, and does not require T or B cells. Gimap5-/- animals exhibit a reduced lifespan, with animals dying as early as 5 weeks, and most between 10 and 20 weeks of age.

Putative Mechanism

No Gimap5 protein can be detected by Western blot in bone marrow or B cells from homozygous sphinx mice, suggesting that the sphinx mutation, a glycine to cysteine change within the G1 sequence of the AIG1 domain, destabilizes the protein and leads to its degradation. Accordingly, the sphinx phenotype is essentially identical to that of the Gimap5-null mouse.

The phenotypes of mice and rats with Gimap5 mutations are also in general agreement. Mutants of both species display peripheral T lymphopenia, particularly of CD8+ T cells, arising from impaired intrathymic maturation, increased apoptosis in the periphery, and possibly impaired positive selection. However, several phenotypic variations exist between species. The stage at which intrathymic maturation is impaired is earlier in rats than in mice, and the proliferative response to TCR activation of CD4+ T cells is normal in rats, but absent in mice. Additionally, bone marrow cells from rats with Gimap5 mutation are capable of normal reconstitution of the B cell compartment (16), whereas HSCs from sphinx mice are not. More significantly, Gimap5 mutant mice do not develop diabetes and BB rats do not develop nodular hyperplasia of the liver. Development of diabetes in BB rats has been shown to be a consequence of modifier loci (20;36), which are presumably of a genotype capable of preventing diabetes in mice. For unknown reasons, hepatic hematopoiesis and hyperplasia are apparently mouse-specific phenotypes.

B cells from sphinx mice fail to proliferate in response to BCR stimulation or PMA/ionomycin treatment, and fail to generate antigen-specific IgG or IgM responses to immunization. However, PMA/ionomycin induces normal IκB degradation, and normal ERK, JNK, p38MAP kinase, and Akt activation in sphinx B cells. These data suggest that Gimap5 modulates BCR-induced proliferation and Ig responses independently of canonical BCR signaling pathways. When cultured in vitro, sphinx B cells accumulate in S phase, raising the possibility that Gimap5 promotes cell cycle progression required for BCR-induced proliferation.

Examination of the sphinx phenotype suggests that Gimap5 functions to promote quiescence in lymphoid cells and HSCs. In support of this hypothesis, thymic negative selection thresholds are reduced, HSCs are aberrantly activated in the fetal liver, and bone marrow NK cells display an activated phenotype. It is possible that aberrant activation of sphinx HSCs in the neonatal liver may prevent their outward migration and induce extramedullary hematopoiesis. HSCs are known to express Toll-like receptors (TLRs) and can undergo myelopoiesis at extramedullary sites in response to LPS-induced inflammation (45;46). Thus, Gimap5 might act as a negative regulator of HSC activation initiated by TLR agonists or other stimuli.

Previous work suggests that Gimap5 also promotes cell survival. A reduced expression of IL-7Rα on the surface of sphinx CD8+ SP thymocytes and peripheral CD4+ T cells may contribute to the diminished survival of these cell types. IL-7Rα signalling promotes expression of Bcl2, which can directly inhibit the activity of the pro-apoptotic protein Bim, and IL-7Rα signalling is largely dispensable for T cell survival in Bim-/- mice (47).  However, IL-7-independent T cell survival signals also require Gimap5 function, since no rescue of lymphopenia was seen in Gimap5sphinx/sphinxBim-/- mice.

Intestinal inflammation dependent upon commensal microbes was found to contribute to weight loss and early death of homozygous sphinx mice. An abundance of IL-13 producing NKT cells are found in the intestinal lamina propria of human ulcerative colitis patients (48) and IL-13 producing iNKT cells can drive colitis in a mouse model induced by the chemical oxazolone (49)iNKT cells in sphinx mice are polarized towards the production of IL-4 and IL-13, suggesting the involvement of these cells in intestinal inflammation. Another possibility is that progressive loss of non-CD4+ lymphocytes in sphinx mice might promote lymphopenia-induced proliferation of CD4+ T cells, shown to cause intestinal inflammation and wasting disease in the T cell transfer model of colitis utilizing C57BL/6 Rag-/- recipients (50).  In support of this hypothesis, CD4+ T cells in sphinx mice gradually acquired low CD62L and high CD69 expression, markers characteristic of lymphopenia-induced proliferation. 

Aksoylar et al. link the impaired activity of Tregs and colitis in the sphinx mice to a loss of Foxo expression in CD4+ T cells (2).  The sphinx phenotype is similar to the phenotype found in Foxo1 and/or Foxo3 deficiencies (51-54).  Although there is an observed loss of Foxo expression in the sphinx lymphocytes, the method of interaction between Gimap5 and Foxo remains to be elucidated. 

Primers Primers cannot be located by automatic search.
The sphinx mutation destroys an Aci I restriction enzyme site in the Gimap5 genomic DNA sequence. Sphinx genotyping is performed by amplifying the region containing the mutation using PCR, followed by Aci I restriction enzyme digestion.
PCR mix
Red Taq                                2.5 μL
dNTP mix                              2.5 μL
10X RedTaq Buffer                 5.0 μL
Primer F (50μM)                    0.5 μL
Primer R (50μM)                    0.5 μL
Genomic DNA                        2.0 μL
ddH20                                   37 μL
NOTES: Use SIGMA Red Taq, associated buffers and dNTPs; no product is amplified by Accu Taq.  PCR reaction mix can be scaled down to a volume of 25 μL if the genomic DNA is of high quality.
PCR program
1) 94°C             10:00
2) 94°C             0:30
3) 55°C             0:30
4) 68°C             1:00
5) repeat steps (2-4) 32X
6) 68°C             10:00
7) 4°C              ∞
The following sequence of 556 nucleotides (from Genbank genomic region NC_000072 for Gimap5 linear genomic sequence) is amplified:
6154                                     cccttgg ctgacttcca ataggcacgc  
6181 tgggaaagtc aagagccacc taaaattata ggcaaggagt tgggaacatt aaatgatgcc
6241 tatttttctg ggctagaggg ccattgcttt gaagagctat ccacaaggac catgacatca
6301 aagtcagaat tactaagtga aacacaacct tctcttgtct acaggaccag aagcccactg
6361 tgtacaagaa tctagctgcc tgaggatcct cctggtgggc aaatctggct gcggtaaaag
6421 cgccacaggg aacagcatcc tccgacgacc agcattccag tccaggctca gaggccagtc
6481 tgtgaccagg accagccagg cagagacagg cacatgggag gggaggagca tcctagtggt
6541 agacacaccc cccatctttg agtcaaaggc ccagaaccaa gacatggaca aggacatcgg
6601 agactgctac ctgctgtgtg ccccaggacc ccatgtgttg ttactggtga cccagctggg
6661 acgcttcaca gctgaagatg ccatggctgt gaggatggtg aaggaggtc
Primer binding sites are underlined; the mutated G is indicated in red.
Restriction Digest
10 μL PCR reaction
17 μL ddH2O
3 μL NEB Buffer 3
0.4 μL AciI
Incubate 2 hours – overnight at 37°C; heat inactivate at 75°C for 10 minutes.
Run on 3% agarose gel with heterozygous and C57BL/6J controls.
Products: sphinx allele- 556 bp.  Wild type allele- 258 bp, 298 bp.
  41. Shin, J. H., Janer, M., McNeney, B., Blay, S., Deutsch, K., Sanjeevi, C. B., Kockum, I., Lernmark, A., Graham, J., Swedish Childhood Diabetes Study Group, Diabetes Incidence in Sweden Study Group, Arnqvist, H., Bjorck, E., Eriksson, J., Nystrom, L., Ohlson, L. O., Schersten, B., Ostman, J., Aili, M., Baath, L. E., Carlsson, E., Edenwall, H., Forsander, G., Granstrom, B. W., Gustavsson, I., Hanas, R., Hellenberg, L., Hellgren, H., Holmberg, E., Hornell, H., Ivarsson, S. A., Johansson, C., Jonsell, G., Kockum, K., Lindblad, B., Lindh, A., Ludvigsson, J., Myrdal, U., Neiderud, J., Segnestam, K., Sjoblad, S., Skogsberg, L., Stromberg, L., Stahle, U., Thalme, B., Tullus, K., Tuvemo, T., Wallensteen, M., Westphal, O., and Aman, J. (2007) IA-2 Autoantibodies in Incident Type I Diabetes Patients are Associated with a Polyadenylation Signal Polymorphism in GIMAP5. Genes Immun.. 8, 503-512.
Science Writers Eva Marie Y. Moresco, Anne Murray
Illustrators Diantha La Vine
AuthorsMichael J. Barnes, Philippe Krebs, Carrie N. Arnold, Kasper Hoebe, Bruce Beutler
Edit History
2011-07-01 4:22 PM (current)
2011-07-01 4:13 PM
2011-06-30 4:14 PM
2011-01-07 9:47 AM
2010-11-08 10:13 AM
2010-07-08 1:39 PM
2010-07-08 10:36 AM
2010-02-03 4:40 PM