Phenotypic Mutation 'Glacier' (pdf version)
Mutation Type missense
Coordinate44,679,494 bp (GRCm39)
Base Change T ⇒ A (forward strand)
Gene Pax5
Gene Name paired box 5
Synonym(s) EBB-1, Pax-5
Chromosomal Location 44,524,757-44,710,487 bp (-) (GRCm39)
MGI Phenotype FUNCTION: [Summary is not available for the mouse gene. This summary is for the human ortholog.] This gene encodes a member of the paired box (PAX) family of transcription factors. The central feature of this gene family is a novel, highly conserved DNA-binding motif, known as the paired box. Paired box transcription factors are important regulators in early development, and alterations in the expression of their genes are thought to contribute to neoplastic transformation. This gene encodes the B-cell lineage specific activator protein that is expressed at early, but not late stages of B-cell differentiation. Its expression has also been detected in developing CNS and testis and so the encoded protein may also play a role in neural development and spermatogenesis. This gene is located at 9p13, which is involved in t(9;14)(p13;q32) translocations recurring in small lymphocytic lymphomas of the plasmacytoid subtype, and in derived large-cell lymphomas. This translocation brings the potent E-mu enhancer of the IgH gene into close proximity of the PAX5 promoter, suggesting that the deregulation of transcription of this gene contributes to the pathogenesis of these lymphomas. Alternative splicing results in multiple transcript variants encoding different isoforms. [provided by RefSeq, Jul 2013]
PHENOTYPE: Null mutants exhibit impaired development of the midbrain resulting in a reduced inferior colliculus and an altered cerebellar folial pattern, failure of B cell differentiation, runting, and high postnatal mortality with few survivors. [provided by MGI curators]
Accession Number

NCBI Refseq: NM_008782.2; MGI:97489

Amino Acid Change Isoleucine changed to Phenylalanine
Institutional SourceBeutler Lab
Gene Model predicted gene model for protein(s): [ENSMUSP00000014174] [ENSMUSP00000099996] [ENSMUSP00000103455] [ENSMUSP00000103457] [ENSMUSP00000103458] [ENSMUSP00000133540] [ENSMUSP00000141186] [ENSMUSP00000134370] [ENSMUSP00000139296] [ENSMUSP00000133671] [ENSMUSP00000134712] [ENSMUSP00000134391] [ENSMUSP00000128880]
AlphaFold Q02650
SMART Domains Protein: ENSMUSP00000014174
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 1e-4 SMART
Pfam:Pax2_C 279 390 6e-53 PFAM
Predicted Effect possibly damaging

PolyPhen 2 Score 0.926 (Sensitivity: 0.81; Specificity: 0.94)
(Using ENSMUST00000014174)
SMART Domains Protein: ENSMUSP00000099996
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 1e-4 SMART
Pfam:Pax2_C 276 341 1.1e-25 PFAM
Predicted Effect possibly damaging

PolyPhen 2 Score 0.905 (Sensitivity: 0.82; Specificity: 0.94)
(Using ENSMUST00000102932)
SMART Domains Protein: ENSMUSP00000103455
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 2e-4 SMART
Pfam:Pax2_C 279 356 5.7e-28 PFAM
Predicted Effect possibly damaging

PolyPhen 2 Score 0.861 (Sensitivity: 0.83; Specificity: 0.93)
(Using ENSMUST00000107825)
SMART Domains Protein: ENSMUSP00000103457
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 7e-4 SMART
low complexity region 269 281 N/A INTRINSIC
Predicted Effect probably damaging

PolyPhen 2 Score 0.980 (Sensitivity: 0.75; Specificity: 0.96)
(Using ENSMUST00000107826)
SMART Domains Protein: ENSMUSP00000103458
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 4e-4 SMART
low complexity region 298 310 N/A INTRINSIC
Predicted Effect possibly damaging

PolyPhen 2 Score 0.948 (Sensitivity: 0.79; Specificity: 0.95)
(Using ENSMUST00000107827)
SMART Domains Protein: ENSMUSP00000133540
Gene: ENSMUSG00000014030

PAX 16 140 4.92e-96 SMART
SCOP:d1ftt__ 177 211 1e-4 SMART
Pfam:Pax2_C 233 298 2.4e-25 PFAM
Predicted Effect probably benign
SMART Domains Protein: ENSMUSP00000141186
Gene: ENSMUSG00000014030
AA Change: H95L

PAX 15 108 1.2e-22 SMART
Predicted Effect probably benign

PolyPhen 2 Score 0.292 (Sensitivity: 0.91; Specificity: 0.89)
(Using ENSMUST00000186542)
SMART Domains Protein: ENSMUSP00000134370
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 1e-3 SMART
Predicted Effect probably damaging

PolyPhen 2 Score 0.996 (Sensitivity: 0.55; Specificity: 0.98)
(Using ENSMUST00000143235)
SMART Domains Protein: ENSMUSP00000139296
Gene: ENSMUSG00000014030
AA Change: I183F

PAX 15 139 8e-96 SMART
low complexity region 156 188 N/A INTRINSIC
Predicted Effect possibly damaging

PolyPhen 2 Score 0.948 (Sensitivity: 0.79; Specificity: 0.95)
(Using ENSMUST00000146335)
SMART Domains Protein: ENSMUSP00000133671
Gene: ENSMUSG00000014030
AA Change: I118F

PAX 16 120 2.93e-30 SMART
SCOP:d1ftt__ 154 188 1e-4 SMART
Pfam:Pax2_C 212 290 8.7e-28 PFAM
Predicted Effect probably damaging

PolyPhen 2 Score 0.961 (Sensitivity: 0.78; Specificity: 0.95)
(Using ENSMUST00000173733)
SMART Domains Protein: ENSMUSP00000134712
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 2e-4 SMART
low complexity region 306 325 N/A INTRINSIC
Predicted Effect possibly damaging

PolyPhen 2 Score 0.861 (Sensitivity: 0.83; Specificity: 0.93)
(Using ENSMUST00000173821)
SMART Domains Protein: ENSMUSP00000134391
Gene: ENSMUSG00000014030
AA Change: I184F

PAX 16 140 4.92e-96 SMART
low complexity region 157 189 N/A INTRINSIC
SCOP:d1ftt__ 220 254 2e-3 SMART
Predicted Effect possibly damaging

PolyPhen 2 Score 0.951 (Sensitivity: 0.79; Specificity: 0.95)
(Using ENSMUST00000174242)
SMART Domains Protein: ENSMUSP00000133978
Gene: ENSMUSG00000014030
AA Change: I128F

PAX 4 85 2.44e-27 SMART
low complexity region 102 134 N/A INTRINSIC
SCOP:d1ftt__ 165 199 2e-4 SMART
low complexity region 251 270 N/A INTRINSIC
Predicted Effect unknown
SMART Domains Protein: ENSMUSP00000134119
Gene: ENSMUSG00000014030
AA Change: I128F

PAX 4 85 2.44e-27 SMART
low complexity region 102 134 N/A INTRINSIC
SCOP:d1ftt__ 165 199 7e-5 SMART
Pfam:Pax2_C 224 335 2.3e-53 PFAM
Predicted Effect unknown
SMART Domains Protein: ENSMUSP00000128880
Gene: ENSMUSG00000014030

PAX 16 140 4.92e-96 SMART
SCOP:d1ftt__ 177 211 1e-4 SMART
Pfam:Pax2_C 233 347 7.3e-55 PFAM
Predicted Effect probably benign
Meta Mutation Damage Score 0.8250 question?
Is this an essential gene? Essential (E-score: 1.000) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status loading ...
Single pedigree
Linkage Analysis Data
Alleles Listed at MGI

All Mutations and Alleles(13) : Chemically induced (ENU)(2) Gene trapped(1) Targeted(10)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL02369:Pax5 APN 4 44691919 missense probably damaging 1.00
IGL02700:Pax5 APN 4 44682722 missense probably damaging 0.99
IGL02754:Pax5 APN 4 44570059 missense probably damaging 0.96
apple UTSW 4 unclassified
Denim UTSW 4 44645661 nonsense probably null
glacier2 UTSW 4 44710407 start codon destroyed probably null 0.96
Glacier3 UTSW 4 44679526 missense probably damaging 1.00
jeans UTSW 4 44645621 missense probably benign 0.03
k2 UTSW 4 44697630 missense probably damaging 1.00
menshevik UTSW 4 44570071 missense probably damaging 1.00
Son_of_apple UTSW 4 44710583 unclassified probably benign
R0411:Pax5 UTSW 4 44609783 missense probably damaging 0.99
R0415:Pax5 UTSW 4 44691886 missense probably damaging 1.00
R0655:Pax5 UTSW 4 44537462 missense probably damaging 0.97
R1146:Pax5 UTSW 4 44697512 splice site probably benign
R1752:Pax5 UTSW 4 44609729 missense probably damaging 1.00
R1891:Pax5 UTSW 4 44691859 missense probably damaging 1.00
R4766:Pax5 UTSW 4 44679494 missense probably damaging 1.00
R4783:Pax5 UTSW 4 44570086 missense probably damaging 1.00
R5134:Pax5 UTSW 4 44710407 start codon destroyed probably null 0.96
R5341:Pax5 UTSW 4 44697630 missense probably damaging 1.00
R5458:Pax5 UTSW 4 44679526 missense probably damaging 1.00
R6281:Pax5 UTSW 4 44691955 missense probably benign 0.37
R6871:Pax5 UTSW 4 44710583 unclassified probably benign
R7025:Pax5 UTSW 4 44679501 nonsense probably null
R7204:Pax5 UTSW 4 44679485 missense possibly damaging 0.93
R7975:Pax5 UTSW 4 44537465 missense probably damaging 0.98
R8246:Pax5 UTSW 4 44570027 missense probably benign 0.08
R8527:Pax5 UTSW 4 44570071 missense probably damaging 1.00
R8542:Pax5 UTSW 4 44570071 missense probably damaging 1.00
R8836:Pax5 UTSW 4 44645621 missense probably benign 0.03
R8847:Pax5 UTSW 4 44691865 missense probably benign 0.15
R8987:Pax5 UTSW 4 44645661 nonsense probably null
R9404:Pax5 UTSW 4 44645565 missense possibly damaging 0.80
S24628:Pax5 UTSW 4 44691886 missense probably damaging 1.00
X0018:Pax5 UTSW 4 44691880 missense probably damaging 1.00
Z1176:Pax5 UTSW 4 44697678 missense probably damaging 1.00
Z1177:Pax5 UTSW 4 44697558 missense probably damaging 1.00
Mode of Inheritance Autosomal Recessive
Local Stock
Last Updated 2020-07-29 6:45 PM by External Program
Record Created 2016-11-16 8:09 AM
Record Posted 2016-12-16
Phenotypic Description
Figure 1. Glacier mice exhibit increased frequencies of peripheral B1 cells. Flow cytometric analysis of peripheral blood was utilized to determine B1 cell frequency. Normalized data are shown. Abbreviations: WT, wild-type; REF, homozygous reference mice; HET, heterozygous variant mice; VAR, homozygous variant mice. Mean (μ) and standard deviation (σ) are indicated.

Figure 2. Glacier mice exhibit decreased frequencies of peripheral B1b cells in B1 cells. Flow cytometric analysis of peripheral blood was utilized to determine B1b cell frequency. Normalized data are shown. Abbreviations: WT, wild-type; REF, homozygous reference mice; HET, heterozygous variant mice; VAR, homozygous variant mice. Mean (μ) and standard deviation (σ) are indicated.

The glacier phenotype was identified among G3 mice of the pedigree R4766, some of which showed an increased frequency of total B1 cells (Figure 1) and a decrease in B1b cells in B1 cells (Figure 2), all in the peripheral blood. 

Nature of Mutation

Figure 3. Linkage mapping of the increased frequency of B1 cells using a recessive model of inheritance. Manhattan plot shows -log10 P values (Y-axis) plotted against the chromosome positions of 86 mutations (X-axis) identified in the G1 male of pedigree R4766.  Normalized phenotype data are shown for single locus linkage analysis without consideration of G2 dam identity.  Horizontal pink and red lines represent thresholds of P = 0.05, and the threshold for P = 0.05 after applying Bonferroni correction, respectively.

Whole exome HiSeq sequencing of the G1 grandsire identified 86 mutations. Both of the above anomalies were linked by continuous variable mapping to three mutations on chromosome 4 in genes Usp45, Fut9, and Pax5. The mutation in Pax5 is presumed to be causative as the phenotypes mimic those observed in Pax5 mouse mutants (see MGI for a list of Pax5 alleles and the record for Apple). The mutation in Pax5 is an A to T transversion at base pair 44,679,494 (v38) on chromosome 4, or base pair 30,947 in the GenBank genomic region NC_000070 encoding Pax5. The strongest association was found with a recessive model of linkage to the normalized frequency of total B1 cells, wherein four variant homozygotes departed phenotypically from 10 homozygous reference mice and 11 heterozygous mice with a P value of 4.234 x 10-13 (Figure 3).  A substantial semidominant effect was observed in both of the assays.

The mutation corresponds to residue 582 in the mRNA sequence NM_008782 within exon 5 of 10 total exons and to residue 284 in the cDNA sequence ENSMUST00000186542.1 within exon 3 of 3 total exons.


90    -I--L--H--Q--W--H--P--G--H--H--V- (ENSMUSP00000141186)


179   -Y--S--I--S--G--I--L--G--I--T--S- (NP_032808)

Genomic numbering corresponds to NC_000070. The mutated nucleotide is indicated in red. The mutation results in a histidine (H) to leucine (L) substitution at position 95 (H95L) in the PAX5 protein encoded by the ENSMUST00000186542.1 cDNA transcript and a isoleucine (I) to phenylalanine substitution at position 184 in the Pax5 protein (Pax5a) encoded by the NM_008782 mRNA transcript. The H95L mutation is strongly predicted by PolyPhen-2 to be benign (score = 0.415).

Illustration of Mutations in
Gene & Protein
Protein Prediction

Figure 4. Pax5 encodes multiple Pax5 isoforms. Pax5 protein domains include the highly conserved paired box DNA binding domain (Paired), an octapeptide motif (Oct), a partial homeodomain (HD), a transactivation domain (TAD), and an inhibitory domain (ID), see text for details. The amino acid numbers shown for isoform Pax5a, the primary isoform of Pax5, are from Maier and Hagman (2002). The Pax5 isoforms, Pax5d and Pax5e, are also shown. The glacier mutation results in an isoleucine (I) to phenylalanine substitution at position 184 in the Pax5a protein. It is predicted to affect most of the Pax5 isoforms. Mutations found in PAX5 are noted. Click on each mutation for more information.

Pax5 encodes Pax5 [alternatively, B-cell lineage-specific activator protein (BSAP)], a member of the paired box (PAX) family of transcription factors that have roles in cellular differentiation, migration, and proliferation (1-5). Pax5 has been further classified as a member of a Pax family subgroup that includes Pax2 and Pax8. The Pax5/2/8 subgroup share a highly conserved paired box DNA binding domain (PRD), an octapeptide motif, a partial homeodomain, a transactivation domain, and a repressor domain (1;6-12) (Figure 4). Pax5 recognizes target genes via the PRD, a 128 amino acid (aa) domain that has two helix-turn-helix motifs separated by a polypeptide linker (4;6;9;12-16). At the N-terminus, a β-hairpin binds to the DNA phosphate backbone, followed by a β-turn that contacts the minor groove of DNA targets (8;12;15-17). The N-terminus is also essential for the recombination function of Pax5 (see the “VH-to-DJH” subsection in the “Background”, below) (8). The polypeptide linker also contributes to DNA binding by contacting the minor groove (12). The centrally located octapeptide motif (TN8TCCT; N can be any combination of eight nucleotides) is a characteristic feature of most Pax proteins and is a region rich in proline, serine, threonine, and tyrosine residues [(18-20); reviewed in (21)]. The octapeptide motif is required for Pax5-Groucho-mediated gene repression; mutation or deletion of this sequence leads to an increase in the transactivation of Pax proteins (18-20;22). Also, the octapeptide motif has the potential to direct cytosine methylation to surrounding gene sequences (23). The partial homeodomain of Pax5 is not required for transcriptional activation of Pax5 target genes, but is required for Pax5-mediated activation of recombination, as deletion of this domain resulted in impaired RAG-mediated recombination (see the “VH-to-DJH” subsection in the “Background”, below) (8). Although the homeodomain is not required for activation of target genes, it is required for interaction with the TATA box binding protein, retinoblastoma protein, and the death domain-associated protein Daxx (24;25) as well as for the recruitment of the Groucho corepressor (18), and the cooperation of Pax5 and Ets-1 (17;26). The 55 amino acid proline-, serine-, and threonine-rich transactivation domain and repressor sequences are at the C-terminus of Pax5 (10;18).  The transactivation domain is active in both promoter and enhancer positions during Pax5-mediated regulation of gene expression and can be negatively regulated by the adjacent repressor sequences at the extreme C-terminus (10;18).

E1A-binding protein (p300), a histone acetyltransferase that regulates transcription through chromatin remodeling, interacts with the Pax5 C-terminus, and can subsequently facilitate the acetylation of several lysines in the Pax5 N-terminal PRD (e.g., Lys67, Lys87, Lys94, Lys98, and Lys103) upon p300 overexpression (7).  Mutations of Lys67, Lys87, Lys89, or Lys103 altered the induction of Pax5 target genes (e.g., CD19 and Blnk (B cell linker; see the record for busy)) in murine pro B cells [see Table 2 for the functional significance of these Pax5 targets; (7)], indicating that Pax5 activity is regulated through acetylation.

Pax5 generates multiple isoforms from two known alternative promoters to generate transcripts with distinct first exons (exon 1A and 1B) and otherwise similar in-frame sequences [(9;11;27-29); reviewed in (30)]. Alternative splicing also occurs throughout the transcript to generate multiple protein isoforms [(9); reviewed in (30)]. In both murine spleen and B-lymphoid cell lines, Pax5 generates four isoforms by alternative splicing that are expressed at detectable levels in normal B cells: Pax-5a (i.e., full-length Pax5), Pax5b, Pax5d, and Pax5e (28;31) (Figure 4).  Pax-5d lacks exons 6-10 and Pax-5e lacks exons 2 and 6-10 (31). As a result, Pax5a and Pax5d, but not Pax5e, possess an intact DNA-binding domain; Pax5d and Pax5e do not have transactivation, repression, or partial homeodomain homology regions, but do contain a novel 128 base pair sequence with an unknown function (28;31). In a transient transfection system, the Pax5d and Pax5e isoforms worked to either compete or enhance the transactivating function of Pax5a: Pax5d had a transactivating function opposite of Pax5a, while isoform Pax5e increased the activity of Pax5a (31). Lowen et al. also showed that the ratio of Pax5d to Pax5e correlated with the proliferation state of the lipopolysaccharide-activated B cell: increased levels of Pax-5e led to increased proliferation, while increased Pax5d levels inhibited cell growth (31). The function of Pax-5b was not addressed. Some studies have indicated that the differential expression of PAX5 isoforms can be disease-specific in B cell malignancies (11;32;33). However, the findings have been inconsistent as isoform expression patterns differed between the studies, even in normal cells (11;32;33). One study showed that there were no significant differences in PAX5 transcript patterns between normal and cancerous B-lymphocytes (11). In contrast, another study showed a higher number or leukemia-specific isoforms in B-cell precursor acute lymphoblastic leukemia (BCP-ALL) (33). Robichaud et al. identified the expression of five PAX5 isoforms that were generated through alternative splicing in the 3’ region in healthy B-lymphocytes, B cell lines, and primary lymphoma samples (9).  The isoforms excluded different combinations of exons 7-9 resulting in changes at the C-terminus of the protein (i.e., shorter protein variants) (9). In three of the isoforms, there was a shift in the reading frame that resulted in novel protein sequences that contained a higher ratio of proline, serine, and threonine residues (9). In this study, all of the PAX5 isoforms could recognize and bind the same DNA targets, but with different transactivation potentials (9). Furthermore, in individual lymphoma samples, Robichaud et al. noted one predominant isoform, either the full length PAX5 or PAX5Δ8, an isoform that lacks 34 amino acids within the transactivating domain; healthy samples expressed multiple isoforms (9). This study also determined that upon activation of distinct cellular pathways with mitogenic agents, different isoform expression patterns were induced (9). In another study, on ‘aged’ normal lymphocytes, researchers found that transcripts in these lymphocytes corresponded to isoforms already described (9;11), including those with various exon deletions (e.g., Δ2, Δ7/8, Δ2/7/8, Δ2/7/8/9, Δ2/4/8, and Δ2/3/4/8) (32). Nebral et al. found that there was a time-dependent accumulation of splice variants, indicating that the conflicting finds in other studies may be due to unintentional ‘ageing’ of lymphocytes or cold shock conditions (32).

Figure 8. The crystal structure of the Pax5/Ets-1 complex bound to DNA. Amino acids 19-142 of the Pax5 PRD was solved in complex with amino acids 335-436 of the Ets domain of Ets-1; the DNA fragment is a 25 base pair Mb-1 DNA promoter fragment. Figure based on PDB: 1K78.  See text and (17for details.

Pax5 interacts with partner proteins to increase Pax5 specificity for target genes [Table 1(16)]. Pax5 can form ternary complexes between Pax5, Ets proto-oncogene family proteins, and DNA [Figure 5; (26)]. The crystal structure of the ternary complex of the Pax5 PRD (aa 19-142) and the Ets domain (aa 335-436) of Ets-1 bound to a 25 base pair Mb-1 DNA promoter fragment has been solved at a resolution of 2.25 Å [Figure 5(17); PDB: 1K78]. Pax5 binds DNA similar to Pax6 [PDB: 6PAX; (14)] in that the N-terminal domain (NTD) and C-terminal domain (CTD) of the Pax5 PRD each have three α helices with a helix-turn-helix motif that contact the DNA major groove (17).  Furthermore, the NTD has a short β hairpin at its N-terminus, followed by a β turn; the linker that connects the NTD and CTD binds in the DNA minor groove in an extended conformation (17). The Pax5 PRD has several direct contacts with the DNA bases as well as contacts with the sugar phosphate backbone (17). For example, in helix C (i.e., the recognition helix), His62 directly contacts G9 (and indirectly contacts T10 and A11 through water-mediated interactions) and Lys67 directly contacts T11 (17). Previous studies had determined that His62 is essential for Pax5 binding, as mutations at this base results in 95% loss in Pax5 binding (16;26).  Within the β turn, Asn29 and Gly30 contact the minor groove through direct and water-mediated interactions (17). Residues Ile83, Gly84, Gly85, and Ser86 within the linker domain contact the minor groove through main chain or side chain atoms (17).  Within the CTD, Arg137 is in van der Waals contact with T21 and S133 is in a hydrogen bond with G23 in the major groove (17). Pax5 recruits Ets-1 to a Pax5-Ets binding site to subsequently enhance its DNA-binding ability at the promoter of the Mb-1 (alternatively, Ig-α; see the record for crab)) gene; Gln22 in Pax5 is an essential residue for this interaction [(7;12;16;17;26;34-36); reviewed in (30)]. The interaction of Pax5 with Ets-1 counteracts Ets-1 autoinhibition to subsequently increase the binding of Ets-1 to the Mb-1 promoter >1000-fold (34). In addition to Ets-1, Pax5 can recruit other members of the Ets family (i.e., Fli-1, GABPα (together with the ankyrin-repeat protein GABPβ), and PU.1) to the Mb-1 promoter in vitro (16;37-39); reviewed in (21)]. In the case of PU.1, the interaction with Pax5 can result in either a cooperative function (e.g., recruiting Groucho proteins such as Grg4 (7;24;38)) to the Igh locus (38)), or a repressive function (e.g., antagonizing activity at the Igk locus (37)).

Table 1.  Additional Pax5 interacting proteins.  See (30) and (21) for more information.

Pax5 interacting partner




Coactivator of Pax5 in mature B cells; inhibits Pax5 in HS-Sultan (Burkitt lymphoma) cells and J558L plasma cells 



Interacts with Pax5 to activate the Rag-2 promoter in immature B cells


Retinoblastoma (Rb) protein

Currently unknown; proposed to silence the 3’ enhancer of the immunoglobulin heavy chain locus and to repress lymphoid-specific PD-1 and J-chain genes


TATA-binding protein (Tbp)

Facilitates Pax5 interaction with the basal transcription machinery

Ada2, GcN5, and CBP

Transcriptional co-activators with Pax5



Co-expression with Pax5 enhanced Pax5-mediated induction of a Luc-CD19 reporter



Pax5 promotes β-catenin localization to intercellular junctions to increase adhesion



Pax5 is expressed in the embryonic central nervous system (transiently in the mesencephalon (i.e., the midbrain), in the midbrain-hindbrain boundary preceding the development of the midbrain and cerebellum, and spinal cord) and fetal liver, B-lymphocytes, and adult testis [(3;4;43-45); reviewed in (21;30)]. In B-lymphocytes, Pax5 expression is initiated in pre-pro-B cells and expressed by the B220+ pro-B cell stage in the bone marrow (46;47). Expression persists to the mature B cell stage in peripheral lymphoid organs where it is subsequently repressed in terminally differentiated plasma cells [(7;44;48-51); reviewed in (21;52)].

Several factors regulate Pax5 expression. For example, in adult B lymphopoiesis, the transcription factors E2A and early B cell factor (EBF1) initiate Pax5 expression [(53;54); reviewed in (55)].  Also, Pax2 activates Pax5 expression via binding to an enhancer sequence within the Pax5 gene (45).  Metastasis-associated protein 1 (MTA1) has also been identified as a regulator of Pax5 transcription [(56); reviewed in (30)]. The JAK2/STAT5 pathway regulates Pax5 activity and may also regulate its expression: phosphorylated STAT5 (see record for domino for information on STAT1) enhances activation of the Pax5 promoter in pre-B cells (57). Also, JAK2 activation can induce the DNA-binding ability and nuclear localization of Pax5 in breast carcinoma cells (58).

Figure 6. B cell differentiation and Pax5.  Conventional B cell differentiation occurs in the bone marrow and spleen. In the bone marrow, development progresses from stem cells through the pro-B cell, pre-B cell and immature pre-B cell stages. Steps upstream of the pro-B stage are not shown for simplicity. During this differentiation, VDJ rearrangements at the immunoglobulin locus result in the generation and surface expression of the pre-B cell (pre-BCR), comprising of an Igm heavy (H) chain and surrogate light chains (VpreB and λ-5). Pre-BCR signaling is necessary for proliferation and further differentiation resulting in expression of a muture BCR, composed of rearranged heavy and light (L) chains, that is capable of binding antigen. Cells successfully completing ths checkpoint leave the bone marrow and proceed through two transitional (T) stages before becoming mature follicular B cells or marginal-zone B cells. The gray boxes list membrane markers and the stage of immunoglobin chain rearrangement associated with each step. Also listed are mouse knockouts (see text) and the resulting block in B cell develpment. D, diversity; J, joining; V, variable; SLC, surrogate light chain. The blue bar below the diagram designates the timing of Pax5 expression. Factors activated by Pax5 are in green, while factors inhibited by Pax5 are in red.

The Pax protein family regulates pattern formation, morphogenesis, cellular differentiation, and organogenesis by activating (or repressing) genes that encode secreted proteins, cell surface receptors, cell cycle regulators, and transcription factors [(3;16); reviewed in (21)]. Pax5/BSAP was identified as a regulator of B cell lineage development (7;48;49) that also plays a role in neural development (44;50) and cancer (1;11;48). In order to facilitate gene regulation, Pax5 can either induce active chromatin (to activate target genes) or eliminate active chromatin (to repress genes) (59). Pax5 induces chromatin changes by recruiting both histone-modifying complexes and chromatin-remodeling complexes to the target genes (48;59). ChIP profiling of histone modifications in Pax5-deficient cells determined that Pax5 induces H3K4 methylation and H3K9 acetylation at regulatory elements of target genes [(59;60); reviewed in (52)]. In the case of repressed genes, Pax5 eliminates active histone modifications (59).

Pax5 and B Cells

Pax5 plays a role in several aspects of B cell biology including the initiation of B cell lineage commitment in the bone marrow, maintenance of the B cell fate in more mature cells, V-DJ recombination of the Igh locus, and B cell migration [(25;44;48;61); reviewed in (62)].

Early B cell differentiation

Mutation of Pax5 leads to B cell developmental arrest at an early progenitor B cell stage in the bone marrow [(50); Figure 6]. During B cell differentiation from a common hematopoietic stem cell progenitor to a mature B cell, Pax5 and other lineage-specific B lymphoid transcription factors such as EBF and E2A function to both activate B lineage-specific genes as well as to repress the transcription of other lineage-inappropriate genes (1;9;18;21;47-50;52;60;61;63-67). Pax5 can bind to a large part of the cis-regulatory genome (i.e., ~8000 Pax5 target genes) in pro-B and mature B cells (47). Microarray analysis of both wild-type and Pax5-deficient pro-B cells found that Pax5 repressed 110 genes and activate 170 genes (46;60;61).  The Pax5-regulated genes included those that encode regulatory and structural proteins that have diverse functions such as transcriptional control, adhesion, migration, immune function, cellular metabolism, and receptor signaling [Table 2; (47;59-61;68); reviewed in (52;55).

Table 2.  Select B cell genes regulated by Pax5. See (47;59-61;68-70) for more Pax5 targets.

Pax-5 mediated regulation

Gene symbol




Ebf1, LEF-1, Mycn (alternatively, N-myc), XBP-1

Transcription factors required to generate lineage-restricted pro-B cells



Transcription factor that regulates B-cell trafficking


Sdc4, Bst1 (alternatively, CD157), CD97, CD55

Cell surface receptors involved adhesion and migration

[(60;61); reviewed in (21)]

Rasgrp3, Prkd2, Lrmp, Stk3, Stk4, and Wnk1

Receptor signaling

Nedd9, calpain2, Eps8

Intracellular signaling components


Cell surface component of the pre-BCR and in the mature B cell



Tyrosine kinase


Blnk (see record for busy)

Component of the pre-BCR


Mb-1 (alternatively, Cd79a or Igα; see record for crab)

B-cell receptor component in pro- and mature B cells


λ5 and VpreB

Surrogate light chains/component of the pre-BCR



Catalytic subunit of the telomerase reverse transcriptase in B cells



Pdcd1 (alternatively, PD-1)

Cell surface protein



Chemokine involved in osteoclast formation



Inhibits T cell fate


Flt3 (see record for warmflash)

Required for early hemopoiesis; must be silenced by Pax5 for B lymphopoiesis to proceed


Csf1r (alternatively, M-CSFR or c-fms)

Represses myeloid fates


Mature B cell function, identity, and survival

In addition to its role in early B cell lymphopoiesis, Pax5 also functions in mature B cells to downregulate B cell-specific genes and reactivate lineage-inappropriate genes (e.g., X-box-binding protein 1 (Xbp1) (71;81), B lymphocyte-induced maturation protein 1 [Blimp-1; (68;81)], immunoglobulin J chain (Igj) (82), and the Igh 3′ enhancer (83-85)) that were previously repressed during B cell maturation to maintain B cell identity throughout B lymphopoiesis (i.e., to repress the plasma cell pathway [(47;60;68;78;86); reviewed in (21)]. Pax5 also functions to control the migration of committed B cells by activating genes that code for adhesion receptors as well as adaptors and effectors downstream of integrin signaling [(59;60); reviewed in (52)]. In a study with conditional knockout of Pax5 in mature B cells, loss of Pax5 caused dedifferentiation of mature B cells into uncommitted progenitors in the bone marrow and the subsequent conversion of those progenitors to functional T cells (78;87). Figure 7 shows an overall view of BCR signaling; factors controlled by Pax5 are bolded.

FIgure 7. Pre-BCR and BCR signaling. Factors controlled by Pax5 are bolded. Pre-BCR engagement results in the activation of SYK (spleen tyrosine kinase), which together with Src-family kinases (LYN, FYN, BLK), phosphorylates many substrates and triggering signaling pathways that are involved in both proliferation and differentiation of pre-B cells. These include activation of phosphoinositide 3 kinase (PI3K) by phosphorylating the coreceptor CD19 and/or the adaptor protein B-cell PI3K adaptor (BCAP). PI3K activation results in the generation of phosphatidylinosital-3,4,5-triphosphate (PIP3), which recruits plekstrin-homology domain signaling molecules to the membrane including the serine threonine protein kinase B (PKB) and its activating kinase 3- phosphoinositide-dependent protein kinase 1 (PDK1). Signaling through this pathway pathway suppresses recombination-activating gene 1 (RAG1) and RAG2 expression, blocks Igk (the k chain of the immunoglobin light chain) gene recombination, and induces cell proliferation. SYK also phosphorylates SH2-domain containing leukocyte protein of SLP65, resulting in the organization of a molecular complex that includes Bruton’s tyrosine kinase (BTK) and phospholipase Cγ2 (PLCγ2). This complex controls downregulation of l-5, a component of the surrogate light chain (SLC), and upregulates the expression of RAG proteins and the interferon-regulatory factor 4 (IRF4). IRF4 positively regulates Igk recombination. SLP65 also modulates PKB activity either directly or by altering the activity of SYK, CD19, or PI3K. Alternatively, SLP65 may regulate PKB activity by activating lipid phosphatases such as SH2-domain containing inositol-5 phosphate (SHIP) and altering PIP3 levels. Multiple downstream signaling pathways are activated by BCR stimulation and lead to a multitude of cellular responses. Following aggregation and localization of BCR molecules, the tails of Igα and Igβ become phosphorylated by Src family kinases (typically Lyn) and by SYK . These phosphotyrosines then act as docking sites for the SH2 domains of SYK, resulting in SYK phosphorylation and activation. SYK phosphorylates a number of downstream targets including BLNK, PLC-g2 and protein kinase C β (PKCβ). BCR stimulation also activates phosphatidylinositol 3 kinase (PI3K) resulting in the generation of 3′-phosphorylated phosphoinositides. One of these lipids, phosphatidylinositol-3,4,5-triphosphate (PIP3), binds selectively to the pleckstrin homology (PH) domain of Btk, facilitating membrane recruitment of the kinase. Phosphorylated BLNK also provides docking sites for Btk, as well as PLC-g2, which results in the additional phosphorylation and activation of PLC-γ2 by Btk leading to the hydrolysis of phosphatidylinositol-3,4-diphosphate (PIP2) to inositol-1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). The recruitment of Vav, Nck and Ras by BLNK to the BCR activates MAP kinase cascades such as JNK, p38 and extracellular signal regulated kinase (ERK). Together, these signals allow the activation of multiple transcription factors, including nuclear factor of activated T cells (NF-AT), nuclear factor (NF)-κB and AP-1, which subsequently regulate biological responses including cell proliferation, differentiation, and apoptosis as well as the secretion of antigen-specific antibodies. Other molecules that play important roles in BCR signaling include Bcl10, mucosa-associated lymphoid tissue translocation gene 1 (MALT1), and caspase recruitment domain family, member 11 (CARMA1 or CARD11), which are involved in NF-κB activation along with PKCβ. The image is interactive, please click to see other BCR-associated mutations.

VH-DJH rearrangement

In Pax5−/− progenitors, VH-DJH recombination of the distal VHJ558 gene family is reduced by ∼50-fold [(50); reviewed in (52)]. Expression of Pax5 in developing T lineage cells in vitro can induce rearrangement of DH-proximal VH7183 genes to the DJH region (85;88). In situ hybridization determined that Pax5 mediates Igh locus contraction and subsequent distal VHJ558 gene recombination (83-85). Subsequent findings determined that Pax5 binds to the VH coding regions and subsequently facilitates RAG1-RAG2 (see the records for maladaptive and snowcock, respectively)-mediated VH-to-DJH recombination by physically interacting with RAG1-RAG2 (8). Zhang et al. propose two Pax5 functions in regulating Igh recombination (8). One, Pax5 binds to the Igh enhancer regions to establish the proper chromosomal configuration for VH-to-DJH rearrangements. Two, the binding of Pax5 to VH coding regions as well as to the RAG1-RAG2 complex mediates RAG-mediated recombination of VH genes (8).

Midbrain and cerebellum development

Studies have found that the Pax5-Pax2 complex is necessary for development of the midbrain and cerebellum (89).  In a mouse model in which only one copy of both Pax2 (Pax2+/-) and Pax5 (Pax5+/-) were expressed, 20% of the compound heterozygotes had missing inferior colliculi and disrupted vermis of the cerebellum (89). A compound model with ablation of Pax5 expression (Pax5-/-) and Pax2 heterozygosity had complete loss of the posterior midbrain and cerebellum by embryonic (E) day 9.5 (89). In the chick brain, changes in Pax5 expression via transfection resulted in swelling of the anterior tectum and changed the diencephalon into a tectum-like structure (90). This study proposed a role of Pax5 in regulating the activity of the isthmus organizer (90).

Pax5 in cancer

Pax5 has been proposed to act as either an oncogene (11;48) or a tumor suppressor (1;87) in human cancers. PAX5 point mutations, monallelic loss of PAX5, and PAX5 fusions with other genes that cause increased Pax5 expression [e.g., PAX5 fusion to IGH (27;91-94)] have been identified in acute lymphoblastic leukemia (ALL) cell lines (i.e., REH and a B-lymphoid cell line derived from a patient) and in lymphoplasmocytoid lymphoma [(10;95); reviewed in (52)]. Overexpression or aberrant somatic hypermutations of PAX5 has also been identified in non-Hodgkin lymphoma [(27;96;97); reviewed in (30)].  In addition to B cell-associated cancers, PAX5 deregulation has been linked to increased malignancy in nonlymphoid cancers such as astrocytomas, medulloblastomas, neuroblastomas, oral carcinomas, colorectal carcinomas, cervical carcinomas, bladder carcinomas, breast cancer, and small-cell lung carcinomas by altering cancer proliferation, apoptosis, and phenotype transitioning [(42;58); reviewed in (30)]. A recent study on hepatocellular cancer (HCC) showed that ectopic expression of PAX5 suppressed HCC growth through the induction of p53-mediated apoptosis indicating that, in the case of hepatocarcinogenesis, PAX5 functions as a tumor suppressor (1).

Figure 8. Pax5 functions in B cell lineage commitment. Pax5-/- uncommitted B cells can differentiated into several hematopoietic cell types either in vitro in the presence of indicated cytokines or in vivo after transplantation. Conditional knockout of Pax5 results in dedifferentiation of the B lymphocyte to an uncommitted progenitor cell.  Figure adapted from Cobaleda et al. (2007).

Pax5 animal models

Pax5 knockout (Pax5-/-) mice exhibit growth retardation and early death by three weeks of age (44). The ~5% of Pax5-/- mice to survive to adulthood were fertile, but were runted (44). Pax5-/- mice are osteopenic, are missing ~60% of their bone mass, and their spleens produced ~5 times the amount of osteoclasts compared to controls (3). Examination of the brains of Pax5-/- mice determined that the inferior colliculus in the auditory midbrain region was underdeveloped near the midline by as early as embryonic day (E) 16.5, but hearing was not affected (44;98). In the Pax5-/- mice, B cell development is arrested at the pro-B cell stage resulting in a lack of immunoglobulin in the serum [(3;25;44;50;86); reviewed in (21)]. In the Pax5-/- fetal liver, B cell development halts before the appearance of B220+ progenitors (50); in the Pax5-/- bone marrow, development progresses to a c-Kit+B220+ progenitor cell stage [(44;50); reviewed in (52)]. Pro-B cells from Pax5-/- mice can differentiate into almost all hemopoietic cell lineages both in vitro and in vivo [Figure 8(18;67;99;100)].  Retroviral transduction of Pax5 causes the Pax5-deficient pro-B cells to progress along the B-lymphoid lineage (18;67;99). DH-to-JH rearrangement proceeds normally in the Pax5-/- mice along with rearrangement of a subset of the VH7183 genes; rearrangement of DH-distal VHJ558 genes is severely compromised (8;101). 

Putative Mechanism

Similar to previous studies using conditional knockout or Pax5 (78;87) and the Pax5-/- mice (18;67;99;100), aberrant Pax5 expression/function in glacier is proposed to cause a conversion of B cell progenitors to other lymphoid cells. Similar to the Pax5-/- mice, arrest of B cell development in the glacier mice leads to a reduction in the amount of immunoglobulin in the serum [(3;25;44;50;86); reviewed in (21)]. As the glacier mutation is within the octapeptide motif, it may result in an increase in the transactivation of Pax proteins.

Primers PCR Primer

Sequencing Primer

PCR program

1) 94°C 2:00
2) 94°C 0:30
3) 55°C 0:30
4) 72°C 1:00
5) repeat steps (2-4) 40x
6) 72°C 10:00
7) 4°C hold

The following sequence of 457 nucleotides is amplified (chromosome 4, - strand):

1   atccttgagc tggcacttgc ccttctgcag cagttgtacc cagcccagta tatctagtga
61  gcacctctgt gacaggccgg cacccctctc cttctacagt gtctacaggc tccgtgacgc
121 aggtgtcatc ggtgagcacc gactccgcgg gctcctcata ctccatcagt ggcatcctgg
181 gcatcacgtc ccccagtgcc gacaccaaca aacgcaagag ggatgaaggt aagagatggc
241 tgcgcatccc cacccggaat gcacctagca gggtctgtgt ggaaccctcg cagaaggggg
301 aaacaaaagt gcagtacgac atgcttagct gcagaggacc gggcttccca gggtcctagc
361 ctttcctgag acagggaaga tggacagggg cctgtatggg aggactcctt cctggctggt
421 ctcctcccat ctttactcct tgggtgagaa aagccat

Primer binding sites are underlined and the sequencing primers are highlighted; the mutated nucleotide is shown in red.

Science Writers Anne Murray
Illustrators Katherine Timer
AuthorsXue Zhong and Bruce Beutler