Phenotypic Mutation 'Gemini' (pdf version)
Mutation Type nonsense
Coordinate120,743,883 bp (GRCm38)
Base Change C ⇒ T (forward strand)
Gene Irf8
Gene Name interferon regulatory factor 8
Synonym(s) Icsbp1, ICSBP, Myls, IRF-8
Chromosomal Location 120,736,358-120,756,694 bp (+)
MGI Phenotype NO_PHENOTYPE,Homozygotes for a targeted null mutation exhibit increased incidence of viral infections, shortened life span, deregulated hematopoiesis, and hematological neoplasias. Heterozygotes show similar, but milder, phenotypes.
Accession Number
NCBI RefSeq: NM_008320; MGI: 96395
Mapped Yes 
Amino Acid Change Glutamine changed to Stop codon
Institutional SourceBeutler Lab
Ref Sequences
Q118* in Ensembl: ENSMUSP00000040245 (fasta)
Gene Model not available
SMART Domains

IRF 3 115 8.69e-65 SMART
Blast:IRF 121 149 N/A BLAST
Blast:IRF 150 176 N/A BLAST
Pfam:IRF-3 202 380 3.3e-58 PFAM
Phenotypic Category cellular phenotype, decrease in CD8+ T cells, decrease in response to injected CpG DNA, immune system, MCMV susceptibility
Penetrance 100% 
Alleles Listed at MGI

All alleles(5) : Targeted, knock-out(1) Targeted, other(2) Spontaneous(1) Chemically induced(1)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL01314:Irf8 APN 8 120753380 missense probably damaging 1.00
IGL02944:Irf8 APN 8 120755125 missense probably benign 0.00
IGL03024:Irf8 APN 8 120753358 missense possibly damaging 0.91
ANU74:Irf8 UTSW 8 120739869 missense probably benign 0.34
R0211:Irf8 UTSW 8 120739975 missense probably damaging 1.00
R0211:Irf8 UTSW 8 120739975 missense probably damaging 1.00
R0840:Irf8 UTSW 8 120753481 missense probably benign 0.00
R1622:Irf8 UTSW 8 120739822 missense probably benign 0.28
R1715:Irf8 UTSW 8 120754388 missense possibly damaging 0.82
R2274:Irf8 UTSW 8 120753527 missense probably benign 0.01
R2875:Irf8 UTSW 8 120754463 missense probably damaging 0.96
R3743:Irf8 UTSW 8 120753571 missense probably damaging 1.00
R4209:Irf8 UTSW 8 120753469 nonsense probably null
R4729:Irf8 UTSW 8 120753439 missense possibly damaging 0.90
Mode of Inheritance Autosomal Semidominant
Local Stock Sperm
MMRRC Submission 034294-UCD
Last Updated 05/13/2016 3:09 PM by Stephen Lyon
Record Created unknown
Record Posted 08/21/2010
Other Mutations in This Stock Stock #: A4554 Run Code:
Validation Efficiency: 92/109

GeneSubstitutionChr/LocMutationPredicted EffectZygosity
Asap1 T to C 15: 64,124,711 probably benign Het
Bpifb5 A to T 2: 154,227,180 Y139F possibly damaging Homo
Chd2 A to C 7: 73,480,968 V782G probably benign Homo
Chst4 G to T 8: 110,029,888 Q365K probably benign Homo
Dido1 T to C 2: 180,675,371 K548E probably benign Homo
Evpl A to G 11: 116,220,834 L2010P probably damaging Homo
Fgl2 A to G 5: 21,372,778 E21G probably benign Homo
Greb1l A to T 18: 10,532,862 M919L possibly damaging Homo
Kel T to A 6: 41,697,419 D359V possibly damaging Homo
Lmtk2 A to G 5: 144,166,317 D298G possibly damaging Homo
Masp1 A to T 16: 23,454,940 probably benign Homo
Mrgprb8 A to T 7: 48,389,408 I276F probably damaging Homo
Nde1 T to C 16: 14,188,410 probably benign Homo
Rbck1 G to T 2: 152,319,172 N385K probably damaging Homo
Senp6 G to T 9: 80,148,458 unknown Het
Tm4sf4 T to A 3: 57,437,767 probably null Homo
Ubn2 A to T 6: 38,484,110 H658L possibly damaging Homo
Vmn2r120 A to G 17: 57,525,715 F155L probably benign Homo
Vmn2r65 T to A 7: 84,946,583 T298S probably damaging Homo
Phenotypic Description
The Gemini mutation was identified in a screen for susceptibility to mouse cytomegalovirus (MCMV) in which ethylnitrosourea (ENU)-mutagenized G3 C57BL/6J mice were challenged with 105 plaque forming units (PFU) of MCMV, an inoculum that is normally harmless to wild type C57BL/6J animals (MCMV Susceptibility and Resistance Screen(1).  Mice with the Gemini mutation die between days five and six after infection with 2x105 PFU of MCMV. The degree of susceptibility is similar to that of BALB/c mice, which lack the natural killer (NK) cell-activating receptor, Ly49H. Following establishment of a homozygous stock, a total of 13 out of 13 8 week-old female Gemini mice died (or were moribund) within 7 days after infection with 1.5x105 PFU of MCMV, and 7 out of 10 8 week-old male Gemini mice died (or were moribund) within 7 days after infection with 2x105 PFU of MCMV confirming the phenotype.  
Further phenotypic analysis of Gemini mice revealed that they have a slight reduction in the frequency of peripheral CD8+ T cells, resulting in a high CD4+ to CD8+ T cell ratio. The frequency of B cells in the blood is normal, and these cells are phenotypically mature. However, in the spleen, there is a reduction in the frequency of mature follicular B cells and an expansion of the marginal zone B cell compartment. The most striking phenotype of Gemini mice is the complete absence of monocytes and mature macrophages in the blood, peritoneal cavity, and bone marrow. There is also expansion of granulocytes, most likely neutrophils, in the blood, spleen, and bone marrow of Gemini mice. The mice are unable to make type I interferon (IFN) after CpG DNA challenge in vivo (2). As plasmacytoid DCs (pDCs) are the primary type I IFN producing cell type in response to TLR9, which senses CpG DNA, it is likely that Gemini mice have a pDC defect in addition to the other immune cell phenotypes noted.
Gemini mice were outcrossed to C57BL/10J animals for mapping. The flow cytometric data for 67 F2 backcross mice is shown with approximately half of the mice lacking peripheral monocytes and macrophages as expected (Figure 1). 


Nature of Mutation
Whole genome sequencing of a homozygous Gemini mouse using the SOLiD technique was performed, identifying mutations in 19 genes. 85.5% and 63% of coding/splicing sequence was covered at least 1X or 3X, respectively. Validation sequencing using the Sanger method was attempted on all nucleotides for which discrepancies were seen at 3x or greater coverage, with 92 of 109 discrepancies successfully processed. None of the mutations identified by SOLiD were previously known to have the immune phenotypes described in homozygous Gemini mice.
Based on the phenotypic data, the Irf8 gene was directly sequenced as a candidate and a C to T transition was found at position 387 of the Irf8 transcript in exon 3 of 9 total exons using Genbank record NM_008320 (Figure 2). In order to validate that the Irf8 mutation is causing the lack of peripheral monocytes and macrophages noted in Gemini mice, mapping was performed by outcrossing homozygous animals to the closely related C57BL/10J strain and backcrossing F1 animals back to Gemini mice. The DNA samples from 31 phenotypically mutant and 36 phenotypically normal animals were pooled separately and analyzed by bulk segregation analysis (BSA) using a panel of 124 single nucleotide polymorphisms (SNPs). C57BL/6J and C57BL/10J allele frequencies were calculated based on normalized capillary sequencing chromatogram peak heights, which reflect the quantity of a given nucleotide at each position in the DNA sequence. The Gemini phenotype showed strong linkage to marker B10SNPS0133 at position 126154896 bp on Chromosome 8 (P< 6.78 x10-21) with a synthetic LOD score of 20 using both mutant and wild type data (2). This marker is approximately 2.9 Mb away from the Irf8 locus.
113 -V--P--E--E--E--Q--K--C--K--L--G-
The mutated nucleotide is indicated in red lettering, and converts a glutamine at amino acid 118 to a stop codon removing 307 amino acids from the C-terminus of the encoded protein. 
Protein Prediction
In mouse and human, the Irf8 gene encodes one of nine members of the interferon regulatory factor (IRF) family of transcription factors, which regulate the transcription of type I interferons (IFN-α/β) and IFN-inducible genes during immune system development, homeostasis and activation by microbes [reviewed by (3;4)]. Mouse IRF8 contains 424 amino acids and displays 88% sequence identity to its human homologue (5;6). An additional IRF, IRF10, has been identified in chickens (7), and several viral homologs have also been described (8). Amongst all IRF family members, IRF8 is most similar to IRF4 (see the record for honey) (3;4).
As in the other IRFs, the N-terminal half of IRF8 (residues 1-121) serves as the DNA binding region ( 9;10), and is characterized by the presence of five highly conserved tryptophans (residues 13, 28, 40, 60 and 79) each separated by 10-20 amino acids (Figure 3) (11)The DNA binding region bears similarity to that of the c-Myb oncoprotein that also contains a tryptophan cluster (12), but not to any other transcription factor classes.  IRF family proteins share sequence and structural homology in their DNA binding regions, and all bind to a similar DNA motif (A/G NGAAANNGAAACT) called the IFN-stimulated response element (ISRE) (13) or IFN regulatory element (IRE) (14) that is present in the regulatory regions of interferons and interferon-stimulated genes (ISGs). By itself, IRF8 only possesses weak DNA binding affinity (10;15), but can function as both a transcriptional activator and repressor depending upon the specific promoter and binding partner (6;16-18). These binding partners include the ETS (E26-transformation specific) family members PU.1, Spi-B (19) and the transcriptional repressor TEL (20), basic helix-loop-helix(bHLH) transcription factor E2A (18), nuclear factor of activated T cells (NFAT) , as well as other IRF proteins including IRF1 (see the record for Endeka), IRF2 and IRF4 (9;15;21). IRF8 can antagonize transactivation of certain target genes by IRF1 possibly by preventing IRF1 from binding to the promoters of these genes (17), although a multiprotein complex composed of IRF8, IRF1 and PU.1 can function as a transcriptional activator (22).   With PU.1 as a partner, IRF8 binds to other DNA elements such as the Ets/IRF composite element (EICE) that is present in such genes as immunoglobin λ (Igλ), Igκ, interleukin-1β (IL-1β), CD20, Toll-like receptor 4 (Tlr4; see the record for lps3) and Tlr9 (see the record for CpG1(23;24), and the novel IRF/Ets composite site (IECS), which binds the same proteins but in the opposite orientation (25). In addition, IRF8/PU.1 can bind IFN-γ (type II interferon) activation sites (GAS) suggesting that IRF8 can amplify IFN-γ-inducible genes (26;27), and Ets/IRF response elements (EIRE) that can bind either IRF dimers or IRF/PU.1 (28). These DNA elements have been found in the promoters of a large number of genes important for immune cell development and function (27).        
Crystal structure analyses of IRF DBDs suggest that they comprise a four-stranded antiparallel β-sheet (β1-β4), three helices (α1-α3), and three long loops (L1-L3) connecting β2 to α2, α2 to α3, and β3 to β4, respectively (Figure 4) (29-33). IRF DBDs bind to DNA over a 12 base pair (bp) stretch with protein-DNA contacts in the major and minor grooves determining specificity. The structure of the IRF8 DBD has not been analyzed, but crystallization of the DNA binding domains of PU.1 and IRF4 in complex with DNA suggests that the DNA in the complex contorts into an S shape in order to juxtapose PU.1 and IRF4 and allow electrostatic and hydrophobic interactions across the central minor groove. The IRF4 and PU.1 DBDs bind to opposite faces of the DNA in a head-to-tail orientation. Loop L3 of the IRF4 DBD is actually a short 310 helix. The PU.1 α3 recognition helix lies in the major groove perpendicular to the DNA axis, while the IRF4 α3 recognition helix is tilted in the major groove with its axis almost parallel to the DNA backbone. IRF4 interacts with PU.1 in this complex with Asp 117, Leu 116, Val 111, and His 56, residues that are conserved in IRF8 (Asp 102, Leu 101, Val 95, His 42), but not in other IRF members. Amino acids in the IRF4 DBD contacting the DNA include Trp 54, Ala 57, Arg 64, Lys 94, Thr 95, Asn 102 and Lys 123. His 56, which also interacts with PU.1, specifies adenines found commonly upstream of the IRF GAAA core sequence. The outer three base pairs of the IRF core (GAAA) are recognized by Arg 98 and Cys 99. Lys 103 at the end of the recognition helix is also important for DNA recognition (31). Except for Arg 64, all of these residues are conserved in IRF8, and mutation of Lys 79, corresponding to Lys 94 of IRF4, abolishes DNA binding (34)
The C-terminal halves of all IRF family members contain either an IRF association domain 1 (IAD1) or an IAD2, with which they bind to other IRFs, other transcription factors, or self-associate.  These interactions allow the IRFs to modulate their activity and target a variety of genes. The IAD1 is approximately 177 amino acids in length, and is conserved in all IRFs except IRF1 and IRF2. IAD2 domains are found only in IRF1 and IRF2 (9;10;35). Using deletion constructs, the mouse IRF8 IAD1 domain was mapped to amino acids 200-377 (10;35), and along with the DBD is necessary for interactions with IRFs, PU.1 and other transcription factors. Mutation of Leu 368 in the closely related IRF4 protein (Leu 329 for IRF8) was sufficient to ablate the interaction of IRF4 with PU.1 (35). In addition, mutation of IRF8 Arg 294 to a cysteine in the recombinant inbred mouse strain BXH-2 abrogates binding to PU.1, SpiB and IRF2 (36). Structural studies of IRF3 and IRF5 demonstrated that the IAD forms a β-sandwich core flanked by N- and C-terminal α-helical regions (37-39), and a conserved α-helix motif exists in the IRF8 IAD domain at amino acids 361-375. Mutation of this region abolishes binding of IRF8 transcriptional complexes to DNA (34). In addition to transcription factors, IRF8 interacts with other proteins using its IAD domain including the E3 ubiquitin ligase tumor necrosis factor receptor-associated factor 6 (TRAF6) through amino acids 305-356 (40), and the COP9 signalosome complex subunit 2 (COPS2). This complex is an essential regulator of the ubiquitin (Ubl) conjugation pathway, and is also involved in phosphorylation of various proteins including IRF8. CSN leads to phosphorylation of IRF8 at an unconserved serine residue within its IAD, Ser 260. This phosphorylation event is essential for efficient association with IRF1 (41). IRF8 also contains a domain required for its repressive transcriptional ability that at least partially overlaps with the IAD (9).
IRF8 is tyrosine phosphorylated at Tyr 48 in its DBD, preventing IRF8 from binding alone to target DNA. Tyr 48 is not conserved with other IRFs except for IRF4, suggesting the mechanism behind the poor DNA binding abilities of these two proteins. IRF8 tyrosine phosphorylation also appears to be important for interactions with partner transcription factors including other IRFs (10;28;42). Phosphorylation of the more conserved Tyr 95 results in increased interaction with PU.1 and IRF1, while the tyrosine phosphatase SHP1 (see the record for spin) inhibits the ability of IRF8 to interact with these partners, leading to reduced expression of myeloid specific genes (42)

The Gemini mutation results in premature truncation of IRF8 at amino acid 118, removing the IAD domain and perhaps triggering nonsense mediated decay of the transcript. Expression of the truncated protein has not been analyzed.

In both mouse and human, the expression of IRF8 is restricted to hematopoietic cells, including cells of the monocyte/macrophage lineage, B lymphocytes, activated T lymphocytes, and DC subsets (5;6;43-46)
IRF8 is expressed at low levels in hematopoietic stem cells (HSC), but its expression is increased in multipotent progenitors including common myeloid progenitors (CMPs), granulocyte-macrophage progenitors (BCMPs), common lymphoid progenitors (CLPs), and megakaryocyte-erythrocyte progenitors (MEPs) (47;48). In lymphocytes, IRF8 is expressed constitutively in B cells throughout development (43) with expression highest at the prepro-B cell stage (48). IRF8 is expressed at relatively low levels in peripheral follicular B cells, but is found at high levels in germinal center (GC) B cells of both mice and humans. Expression is then downregulated in mature plasma cells (49-51).   IRF8 expression is undetectable in thymocytes and resting T cells (43). In cells of the myeloid lineage, IRF8 expression is relatively high in macrophages and very low in granulocytes (47). IRF8 is also highly expressed in myeloid-derived CD8α+ DCs, a subpopulation of double negative (DN) DCs and lymphoid-derived pDCs (44-46). In immune cells, IRF8 localizes mainly to the nucleus, but is also present in the cytoplasm (40;43)
Although abundantly expressed in macrophages, other myeloid cells and B cells, IRF8 expression is induced by IFN-γ in mouse macrophages and T cells (5;43). This effect is mediated by a GAS element present in the IRF8 promoter, which can be bound by the signal transducer and activator of transcription 1 (STAT1) protein (see the records for domino and poison) (52-54). The macrophage-derived cytokines IFN-α and IL-12 can induce IRF8 mRNA and protein expression in human NK and T cells, although not in mouse cells (5;16;55). TLR ligands and IFN-γ synergistically induce IRF8 expression in macrophages (40;56;57)
A number of transcription factors, in addition to STAT1, are implicated in directly regulating Irf8 gene transcription. During early B cell development, the PAX5 transcription factor induces Irf8 transcription (58), while the basic helix-loop-helix transcription factor E2-2/Tcf4 directly regulates Irf8 expression during pDC development (59)Irf8 transcription is repressed by the Wilms’ tumor gene protein (WT1) in human leukemias (60), B lymphocyte-induced maturation protein-1 (Blimp1) during osteoclast differentiation, and STAT5 during the inhibition of pDC development by the granulocyte macrophage colony-stimulating factor (GM-CSF) (61).
In addition to immune cells, IRF8 mRNA and protein is found constitutively at low levels in the mouse lens and retinal pigment epithelial cells (62;63).

IRF8 was first identified as a protein that bound to interferon response elements and was initially known as the IFN consensus sequence binding protein (ICSBP) (5)Irf8-/- animals display deregulated hematopoiesis with impaired macrophage development, and development of a chronic myelogenous leukemia-like (CML) syndrome characterized by hyperproliferation of abnormal myeloid, histocytic, and lymphocytic cells.  Recent findings indicate that IRF8 binds directly to the Bax promoter region to activate Bax transcription (64).  Furthermore, IRF8-mediated regulation of Bax expression is thought to maintain myeloid cell sensitivity to Fas-mediated apoptosis, although further study of IRF8 and myeloid cell apoptosis is needed to verify these findings.  Heterozygous animals displayed less dramatic manifestations of the same phenotypes (65). More recently, IRF8 was found to be important for the differentiation of B cells, DCs, and eosinophils (Figure 4) [reviewed by (4;66)], and IRF8-deficient mice display osteoporosis due to an increased numbers of myeloid-derived osteoclasts (67).  In addition, IRF8-deficient mice are susceptible to infection with various pathogens, the control of which requires IFN-γ-mediated immunity.  These pathogens include vaccinia and lymphocytic choriomeningitis viruses (65), such bacteria as Listeria monocytogenes Yersinia enterocolitica, Mycobacterium bovis and Mycobacterium tuberculosis (68-71), and such parasites as Leishmania major and Toxoplasma gondii (72;73).  However, Irf8-/- animals were able to survive infection with vesicular stomatitis virus (VSV) or influenza A (65;72), which are controlled primarily by type I IFNs and humoral immunity, respectively.


The most striking phenotype of IRF8-deficient mice is the marked expansion of primarily granulocytes in the spleen, lymph node, and bone marrow, suggesting that IRF8 plays a critical role in the differentiation of myeloid cells. All blood cell types originate from multipotent, self-renewing hematopoietic stem cells (HSCs), which first differentiate into non-renewing multipotent progenitors (MPPs) through two intermediate steps known as long-term HSCs (LT-HSCs) and short-term HSCs (ST-HSCs) [reviewed by (74;75)]. Factors implicated in stem cell renewal and differentiation include the receptor tyrosine kinases Flt3 (FMS-like tyrosine kinase 3) and c-Kit (see the records for warmflash and Pretty2). Historically, MPPs have been divided into common myeloid progenitors (CMPs) and common lymphoid progenitors (CLPs) in the bone marrow (BM). CMPs give rise to cells of the myeloid lineage, while CLPs give rise to B, T and NK (natural killer) cell lymphocytes. PU.1 has been shown to be important for the generation of early lymphoid progenitors, but high levels of PU.1 also facilitate myeloid lineage differentiation (76). Interestingly, IRF8 is able to directly repress PU.1 levels by binding to the PU.1 promoter, thus restricting the production of myeloid lineage precursors (48). In addition, IRF8 strongly inhibits cell growth and positively regulates apoptosis in myeloid cells (34;47;77). During myeloid cell differentiation, IRF8 promotes monocyte/macrophage over granulocyte differentiation (34;47). This contrasts with IRF1, which promotes granulocyte development (78). However, IRF1 and IRF8 act together to regulate the transcription of a multitude of genes involved in macrophage maturation, function and TLR stimulation including the genes encoding both subunits (p35 and p40) of the IL-12 cytokine (79;80). IL-12 governs production of IFN-γ in NK cells and CD4+ T cells. In turn, IFN-γ priming of cells greatly enhances the inducibility of the IL-12 p40 gene. These data, along with the ability of IRF8 to be induced by IFN-γ and to enhance the transcription of IFN-γ response genes (27), explains the susceptibility of IRF8-deficient animals to a variety of pathogens that require IFN-γ for their control (65;68-73).


Careful analysis of early B cells in the BM of Irf8 mutant mice found that the prepro-B cell numbers significantly reduced, suggesting that IRF8 plays a role in the differentiation of CLP progenitors to B cells. The decreased commitment of CLPs to develop into B cells was associated with reduced expression of important B cell lineage factors including the transcription factors E2A, EBF, and PAX5. IRF8, along with PU.1, was shown to directly regulate the expression of EBF (48) and may also directly regulate the expression of IKAROS, a transcription factor involved in the generation of primitive lymphoid progenitors (81). EBF is responsible for the activation of several genes involved in B cell lineage commitment including Pax5. As mentioned above (Expression and Localization), PAX5 directly activates Irf8 transcription and also regulates EBF expression. In addition, PU.1 and EBF are autoregulatory (82;83). Thus, many of the transcription factors important in B cell lineage priming form a mutual, autoregulatory system that allows both lymphoid and myeloid specification in early progenitor cells, a process in which IRF8 plays a critical role.


During later stages of B cell development in the BM, IRF8 has been found to act in concert with IRF4 [reviewed by (3;4)]. The development of B cells in the bone marrow is characterized by the differential expression of marker proteins such as CD45 (B220) (see the record for belittle) and CD43, and sequential recombination of Ig gene loci [reviewed in (84)]. Following the prepro-B cell stage, pro-B cells rearrange their immunoglobulin heavy (IgH) chains in a process known as VDJ recombination mediated by the RAG1 (recombination activating gene 1)-RAG2 complex (see the record for maladaptive). Successful VDJ recombination gives rise to the Igμ chain, two of which combine with the surrogate light chains (SLCs), λ-5 and Vpre5, and the signaling subunits Igα and Igβ to form a pre-BCR complex initiating proliferation and further differentiation in pre-B cells. Subsequent rearrangement of the immunologlobin light (IgL) chain forms the BCR (or surface IgM) characteristic of immature B cells.   Both IRF4 and 8 are expressed in pre-B cells (51), and there is a redundant role for these factors in promoting the transition from pre-B cells to IgM+ B cells. B cells from Irf4-/-;Irf8-/- double knockout mice arrest at the cycling pre-B stage (85), and either IRF4 or IRF8 expression can rescue this defect (86). IRF4 and IRF8 induce Ig light chain gene transcription and rearrangement, and also inhibit the expression of pre-BCR components by inducing the expression of IKAROS and the related protein AIOLOS (81;85-88). These double knockout cells are also hyperproliferative, suggesting that IRF4 and 8 are critical for limiting pre-B cell expansion, which is also necessary for further B cell differentiation (81;85;87). IRF4/IRF8-induced IKAROS and AIOLOS attenuate IL-7 signaling, which is essential for pre-B cell proliferation and survival (81). Unlike IRF4, IRF8 is not required for receptor editing, which is used by B cells to revise antigen receptors and maintain self-tolerance (89).  


IRF8 is also involved in the germinal center (GC) program. GCs are sites within the lymph nodes and spleen where B lymphocytes rapidly proliferate, differentiate, and undergo somatic hypermutation and class switching during antibody responses. In Irf8 -/- mice, GCs show less organized morphology and Irf8 -/- B cells express reduced expression levels of the Aicda and Bcl6 genes. IRF8 has been found to directly regulate the induction of these two critical genes during the GC reaction, both in human and mouse. The Aicda gene encodes activation-induced cytidine deaminase (AID), which is required for both class switch recombination (CSR) and somatic hypermutation, and the Bcl6 gene encodes B cell lymphoma 6 (BCL6) protein, a Krüppel-type zinc finger transcriptional repressor that functions as a master regulator in the GC program (49). Interestingly, IRF8 is downregulated in mature plasma cells concomitant with high levels of IRF4 expression. IRF4 expression is upregulated by CD40 stimulation of immature cells in the GC, leading to downregulation of BCL6 and terminal differentiation to post-GC lymphocytes (90).


In addition to B cells, IRF4 and IRF8 appear to coordinate the development of DCs, professional antigen-presenting cells that link the innate and adaptive immune systems. In response to pathogens, DCs secrete cytokines, and can regulate other immune cells including B, T and NK cells. They are composed of distinct subsets that are classified according to phenotypic markers, functional features, localization and lineage [reviewed by (91)]. In the mouse, two major subsets of DCs have been identified based on the levels of the cell surface integrin subunit CD11c or the receptor protein tyrosine phosphatase CD45, CD11chighCD45- myeloid-derived conventional DCs (cDCs) and CD11clowCD45+ lymphoid-derived plasmacytoid DCs (pDCs). Conventional DCs can be further distinguished by expression levels of CD8α, CD4 and CD11b. The spleen contains three cDC subsets: CD4+CD8α-CD11bhigh, CD4-CD8α+CD11blow and CD4-CD8α-CD11bhigh double negative (DN) DCs. The thymus is missing the CD4+CD8α- DCs. Lymph nodes contain two additional DC subsets that are derived from epidermal DCs (Langerhans cells) and interstitial DCs. In humans, two major blood-derived DC subsets are characterized as CD11c+ myeloid DCs (mDCs) and pDCs, but the markers used to distinguish these cells are generally not conserved in mouse DC subsets. In vitro, several growth factors are capable of stimulating DC differentiation in a subset-specific manner. GM-CSF supports the generation of CD11b+ cDCs from bone marrow (BM), while Flt3 ligand (Flt3L) can expand both CD11b+ cDCsand pDCs (92-95). Analysis of DC subsets in mice lacking IRF4 and IRF8 suggests that the pattern of expression of these IRFs in DC subsets strictly correlates with their requirement for DC development. Thus, IRF4, but not IRF8, is necessary for CD4+ DC differentiation, both IRFs support the development of DN DCs, and IRF8 is important for both CD8α+, pDC, Langerhans and interstitial DC differentiation (44-46;96-98). IRF4 also supports pDC development, but to a lesser extent (96). Thus, IRF8-deficient mice are devoid of pDC and CD8α+ DCs and subsequently have impaired production of type I IFN and IL-12p40 (46). IRF4 is important for GM-CSF-mediated DC differentiation, while Flt3L-dependent development requires IRF8 (96;97). In addition to pDC development, IRF8 supports pDC function by being involved in the second, amplifying phase of type I IFN transcription in response to TLR stimulation and viral infections, including MCMV. IRF8, along with IRF7 (see the record for inept), is recruited to the promoters of IFNα/β genes and is necessary for sustained RNA polymerase II activation of type I IFN response genes (99). IRF8-deficient pDCs do not activate the NF-κB pathway in response to TLR9 stimulation, and fail to produce the proinflammatory cytokines tumor necrosis factor (TNF)-α and IL-6 (100).


IRF8 promotes Th1 differentiation (72;73). CD4+ T cells can differentiate into the T helper (TH) cell subsets TH1, TH2, TH3, TH17 and follicular helper (TFH) cells (please see the record for sanroque), as well as immune-suppressive regulatory T cells (Tregs). TH cells are involved in activating and directing other immune cells, and do so by producing cytokines that are specific to each subset. The differentiation of CD4+ T cells into various TH subtypes depends on the environment encountered by the naïve TH cell. TH1 cells produce IFN-γ and tumor necrosis factor β (TNF-β) and are important in stimulating macrophages and cytotoxic CD8+ T cells [reviewed by (101)]. Irf8-/- fail to mount TH1 responses (72;73), but this defect is likely due to the inability of IRF8-deficient macrophages and DCs to promote TH1 differentiation due to impaired production of the major TH1-promoting cytokine IL-12, rather than a defect in T cells (69;79;80). TH17 cells are involved in autoimmunity (e.g. inflammatory bowel diseases, asthma, rheumatoid arthritis, and multiple sclerosis) and tissue inflammation.  A recent study has shown that IRF8 is essential in silencing TH17 differentiation (102).  A conventional IRF8 knockout model found that there were no defects on TH1 or TH2 cells, but that TH17 cell differentiation was enhanced through the IRF8-mediated targeting of RORγt, a nuclear receptor involved in TH17 development (103).


Patients with chronic myeloid leukemia (CML; OMIM #608232) have reduced levels of IRF8 (104), and ectopic expression of IRF8 can rescue an induced mouse model of murine myeloid leukemia (105). Along with the CML-like disease observed in animals with mutations in Irf8 and the role of IRF in inhibiting cell growth and promoting apoptosis (34;47;77), this suggests that IRF8 acts as a tumor suppressor in myeloid progenitor cells and may be important to suppress human CML. CML is often treated by IFN-α, and a positive correlation was found between patients that had a good response to this treatment and displayed increased levels of IRF8 (106)Irf4 transcript levels are also significantly low in CML patients (107) implying that IRF4 has an activity similar to IRF8 in myeloid cell development and CML pathogenesis, as in the case of DC and B cell development. Mice deficient in both IRF4 and IRF8 develop a more aggressive CML-like disease than mice deficient in IRF8 alone, correlating with a greater expansion of granulocyte-monocyte progenitors. However, these animals succumb to a B-lymphoblastic leukemia/lymphoma. Combined losses of IRF4 and IRF8 therefore may contribute to both myeloid and lymphoid tumors (108).


For an overview of hematopoiesis, click here.

Putative Mechanism

The Gemini mutation results in premature truncation of IRF8, which would remove more than half of the protein. Although the expression of the truncated protein has not been determined, it is probable that the Gemini mutation results in a null allele due to instability of the truncated protein product. In addition, the IRF8 IAD domain is crucial for IRF8 function as demonstrated by BXH2 mice, which carry the R249C missense mutation in the IAD and have an almost identical phenotype to Irf8-/- animals (70). Interestingly, these mice do not completely recapitulate the knockout phenotype as pDC development and type I IFN production are retained in these animals, although the development of the other DC subsets and production of IL-12 remains impaired (36). By contrast Gemini animals display a similar phenotype to Irf8 nulls including an impairment of type I IFN production, further suggesting that this mutation results in a null allele. 


In addition to deregulated hematopoiesis, animals with mutations in Irf8 display susceptibility to pathogenic infections including viruses due to profound defects in macrophage and DC function. The identification of the Gemini strain as an MCMV susceptible mutant is consistent with these observations, although IRF8-deficient animals are able to control certain viral infections (65).  NK cell cytolytic function plays a critical role during the innate immune response to MCMV, and IRF8-deficient animals displayed a lack of IFN-γ production by NK cells following LPS stimulation suggesting that NK cell activity is affected in these mice. The production of IFN-γ by NK cells is dependent on IL-12 production by macrophages and CD8α+ DCs, which is compromised in Irf8-/- mice. Interestingly, the presence of CD8α+ DCs in the spleen is required for the late expansion of Ly49H+ NK cells, a subset crucial for controlling MCMV infection in certain mouse strains including C57BL/6J (109). Considering their impairment of macrophage development, it is also possible that IRF8-deficient mice lack inflammatory monocytes, which have been shown to be important in controlling viral infections (110).

Primers Primers cannot be located by automatic search.
Gemini 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               8
Primers for sequencing
The following sequence of 824 nucleotides (NCBI Mouse Genome Build 37.1, Chromosome 8, bases 123,267,336 to 123,268,159) is amplified:
agaggacttt gccaaagctg catcctaggc tgtggggagc ctgggttcaa attccagctt       
tgctgagtgt tttgggggac ccttacccag ctcttgtctg ttgagataca gtggtcttcc      
tgaagtccaa ttgggtgctg accttggaca agggttgatg gccaggaaag gctctggggg      
aagggggcgg gcggtatgca gccggccttt gagttgaact ctgtggtttg tgtccctctg      
tgtttactga gtttctctct cgttcttcag gcctgggcag tttttaaagg gaagtttaaa      
gagggagaca aagctgaacc agccacgtgg aagacgaggt tacgctgtgc tctgaacaag      
agcccagatt ttgaagaagt gactgaccgg tcccagctgg acatttctga gccatataaa      
gtttaccgaa ttgtccccga ggaagaacaa aaatgtaact atctgttggg accaggaagc      
cctctaggaa cctccttcac cccccccccc cccacgctta gctcctgaag ccctcagtct      
ccacagggct agagcagaca tgccacccac agggtttggg tgcctggttt tagctcacca      
gagttgttct ctcttgtgta agcagacata gagagctgct ctgctgacct catacgccat      
gtgctgtcaa gtggcagctt cctgcagagg gagtggccca gtggaggcag ggaagggtgg      
tggactctgg gtcatggcag ccagaagctg gagggagcga gtccccggag ctcagcataa      
actcctccaa atgggtttga ggctctgctg aagtctgtga tgca
Primer binding sites are underlined; sequencing primer binding sites are highlighted in gray; the mutated C is indicated in red.
Science Writers Nora G. Smart
Illustrators Diantha La Vine
AuthorsCarrie N. Arnold, Amanda L. Blasius, Bruce Beutler
Edit History
06/29/2011 1:24 PM (current)
06/15/2011 3:28 PM
01/07/2011 7:43 AM
12/02/2010 9:16 AM
12/01/2010 9:56 AM
12/01/2010 9:56 AM
12/01/2010 9:55 AM
11/12/2010 10:01 AM
11/08/2010 10:46 AM
09/07/2010 12:25 PM
08/27/2010 11:20 AM
08/27/2010 10:22 AM
08/27/2010 10:21 AM
08/27/2010 10:20 AM
08/27/2010 10:18 AM
08/27/2010 10:17 AM
08/27/2010 10:06 AM
08/27/2010 10:04 AM
08/27/2010 10:01 AM
08/27/2010 9:44 AM
08/21/2010 3:20 PM