Phenotypic Mutation 'cellophane' (pdf version)
Mutation Type nonsense
Coordinate5,770,554 bp (GRCm38)
Base Change T ⇒ A (forward strand)
Gene Zeb1
Gene Name zinc finger E-box binding homeobox 1
Synonym(s) 3110032K11Rik, Tw, MEB1, Zfhx1a, Zfhep, ZEB, AREB6, Zfx1a, Tcf18, Nil2, Tcf8, [delta]EF1
Chromosomal Location 5,591,860-5,775,467 bp (+)
MGI Phenotype FUNCTION: [Summary is not available for the mouse gene. This summary is for the human ortholog.] This gene encodes a zinc finger transcription factor. The encoded protein likely plays a role in transcriptional repression of interleukin 2. Mutations in this gene have been associated with posterior polymorphous corneal dystrophy-3 and late-onset Fuchs endothelial corneal dystrophy. Alternatively spliced transcript variants encoding different isoforms have been described.[provided by RefSeq, Mar 2010]
PHENOTYPE: Mutations at this locus affect thymus organization and homozygotes exhibit severe thymic T cell deficiency. Some mutations result in eye anomalies and extensive skeletal abnormalities. Homozygotes generally die at birth due to respiratory failure. [provided by MGI curators]
Accession Number

NCBI RefSeq: NM_011546; MGI: 1344313

Amino Acid Change Tyrosine changed to Stop codon
Institutional SourceBeutler Lab
Gene Model not available
AlphaFold Q64318
SMART Domains Protein: ENSMUSP00000025081
Gene: ENSMUSG00000024238
AA Change: Y902*

low complexity region 13 30 N/A INTRINSIC
ZnF_C2H2 150 173 3.16e-3 SMART
ZnF_C2H2 180 202 3.21e-4 SMART
ZnF_C2H2 220 242 4.87e-4 SMART
ZnF_C2H2 248 268 1.86e1 SMART
low complexity region 288 304 N/A INTRINSIC
low complexity region 532 555 N/A INTRINSIC
HOX 559 621 7.53e-3 SMART
low complexity region 730 742 N/A INTRINSIC
low complexity region 766 783 N/A INTRINSIC
ZnF_C2H2 882 904 1.18e-2 SMART
ZnF_C2H2 910 932 4.4e-2 SMART
ZnF_C2H2 938 959 1.89e-1 SMART
coiled coil region 1006 1077 N/A INTRINSIC
low complexity region 1096 1112 N/A INTRINSIC
Predicted Effect probably null
SMART Domains Protein: ENSMUSP00000124395
Gene: ENSMUSG00000024238

low complexity region 13 30 N/A INTRINSIC
ZnF_C2H2 96 119 3.16e-3 SMART
ZnF_C2H2 126 148 3.21e-4 SMART
ZnF_C2H2 166 188 4.87e-4 SMART
ZnF_C2H2 194 214 1.86e1 SMART
low complexity region 234 250 N/A INTRINSIC
Predicted Effect probably benign
Predicted Effect probably benign
SMART Domains Protein: ENSMUSP00000135125
Gene: ENSMUSG00000024238

ZnF_C2H2 130 153 3.16e-3 SMART
ZnF_C2H2 160 182 3.21e-4 SMART
ZnF_C2H2 200 222 4.87e-4 SMART
ZnF_C2H2 228 248 1.86e1 SMART
low complexity region 268 284 N/A INTRINSIC
Predicted Effect probably benign
Meta Mutation Damage Score Not available question?
Is this an essential gene? Probably essential (E-score: 0.929) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status CE: no linkage results
Single pedigree
Linkage Analysis Data
Penetrance unknown 
Alleles Listed at MGI

All alleles(14) : Targeted(6) Gene trapped(7) Spontaneous(1)          

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL00833:Zeb1 APN 18 5767774 missense probably benign 0.00
IGL01139:Zeb1 APN 18 5705061 missense possibly damaging 0.69
IGL01444:Zeb1 APN 18 5767906 missense probably damaging 1.00
IGL01444:Zeb1 APN 18 5767138 missense probably benign
IGL01806:Zeb1 APN 18 5767867 missense possibly damaging 0.94
IGL01988:Zeb1 APN 18 5759037 nonsense probably null
IGL02059:Zeb1 APN 18 5766892 missense probably damaging 1.00
IGL03005:Zeb1 APN 18 5767150 missense probably benign 0.03
IGL03153:Zeb1 APN 18 5770511 missense probably damaging 1.00
Apes UTSW 18 5761394 missense probably damaging 1.00
serpens UTSW 18 5772455 missense probably damaging 1.00
N/A - 293:Zeb1 UTSW 18 5767076 missense possibly damaging 0.68
R0184:Zeb1 UTSW 18 5766808 missense probably damaging 1.00
R0488:Zeb1 UTSW 18 5772455 missense probably damaging 1.00
R0622:Zeb1 UTSW 18 5759123 nonsense probably null
R0646:Zeb1 UTSW 18 5759027 missense probably damaging 1.00
R0881:Zeb1 UTSW 18 5767138 missense probably benign
R1251:Zeb1 UTSW 18 5705089 missense probably damaging 1.00
R1257:Zeb1 UTSW 18 5772699 missense possibly damaging 0.53
R1501:Zeb1 UTSW 18 5761399 missense possibly damaging 0.95
R1547:Zeb1 UTSW 18 5767450 missense possibly damaging 0.50
R1797:Zeb1 UTSW 18 5766298 nonsense probably null
R1815:Zeb1 UTSW 18 5767898 missense probably damaging 1.00
R2090:Zeb1 UTSW 18 5766458 missense possibly damaging 0.65
R2129:Zeb1 UTSW 18 5767681 missense possibly damaging 0.92
R2875:Zeb1 UTSW 18 5772859 small insertion probably benign
R3888:Zeb1 UTSW 18 5748743 missense probably damaging 1.00
R3941:Zeb1 UTSW 18 5767799 missense probably benign 0.06
R3952:Zeb1 UTSW 18 5772716 missense probably benign 0.17
R4271:Zeb1 UTSW 18 5758985 missense probably damaging 0.99
R4512:Zeb1 UTSW 18 5759007 missense probably damaging 1.00
R4514:Zeb1 UTSW 18 5759007 missense probably damaging 1.00
R4677:Zeb1 UTSW 18 5766775 missense probably damaging 0.97
R4729:Zeb1 UTSW 18 5767286 missense probably damaging 1.00
R5839:Zeb1 UTSW 18 5767507 missense probably benign
R5913:Zeb1 UTSW 18 5766765 missense possibly damaging 0.49
R6248:Zeb1 UTSW 18 5766962 missense probably damaging 1.00
R6354:Zeb1 UTSW 18 5772743 missense possibly damaging 0.64
R6429:Zeb1 UTSW 18 5770498 missense probably damaging 1.00
R6819:Zeb1 UTSW 18 5591917 missense probably damaging 1.00
R7180:Zeb1 UTSW 18 5767867 missense possibly damaging 0.94
R7193:Zeb1 UTSW 18 5772756 missense probably damaging 0.98
R7199:Zeb1 UTSW 18 5767703 missense probably benign 0.00
R7397:Zeb1 UTSW 18 5761394 missense probably damaging 1.00
R7534:Zeb1 UTSW 18 5766611 missense probably damaging 1.00
R7702:Zeb1 UTSW 18 5766802 missense probably damaging 1.00
R7703:Zeb1 UTSW 18 5766917 missense probably benign
R7934:Zeb1 UTSW 18 5748703 missense probably benign 0.00
R8504:Zeb1 UTSW 18 5705127 missense possibly damaging 0.94
R8539:Zeb1 UTSW 18 5748784 missense probably damaging 0.99
R8716:Zeb1 UTSW 18 5767958 missense probably damaging 0.99
R8772:Zeb1 UTSW 18 5770382 critical splice acceptor site probably null
R8824:Zeb1 UTSW 18 5748680 splice site probably benign
R9082:Zeb1 UTSW 18 5772557 missense probably damaging 0.98
R9085:Zeb1 UTSW 18 5766716 missense probably damaging 1.00
R9274:Zeb1 UTSW 18 5772840 small deletion probably benign
Mode of Inheritance Autosomal Recessive
Local Stock Live Mice, Sperm, gDNA
MMRRC Submission 036726-MU
Last Updated 2016-05-13 3:09 PM by Anne Murray
Record Created 2009-11-10 12:00 AM
Record Posted 2012-08-02
Phenotypic Description
Figure 1.  Serum NP-specific IgM (C) and βGal-specific IgG (D) measured by ELISA in WT and G3 mice immunized 5 or 14 d before with NP-Ficoll and rSFV-βGal.  The cellophane index mice are indicated in blue.  Reproduced from (1).
Figure 2.  Lymphocyte development is impaired in cellophane mice.  Please see bumble and Worker for more details about these mutations.  Single-cell suspensions from WT and mutant mice were analyzed by flow cytometry. (A) IgM- B220int pro- or pre-B cells, IgM+ B220int immature B cells, and IgM+ B220high mature B cells as a percent of bone marrow lymphocytes. (B) B220+ B cells as a percent of splenocytes and IgDlow IgMhigh B200+ immature, IgDhigh IgMint B220+ mature follicular, and CD21/35high CD23low B220+ MZ B cells as a percent of splenic B220+ B cells. (C) Mean fluorescence intensity of CD44 on CD4+ and CD8+ T cells and IgM on B220+ B cells in spleen.  (D) CD19int B220high B2 and CD19high B220low/int B1 B cells as a percent of peritoneal cavity lymphocytes. (E) CD5high CD43high CD19high B220low/int B1a and CD5low CD43low CD19high B220low/int B1b B cells as a percent of peritoneal cavity B1 B cells.  (F) CD4- CD- DN, CD4+ CD8+ DP, and CD4or CD8+ SP cells as a percent of thymocytes.  (G) CD69- and CD69+ TCRβ+ CD4or CD8+ cells as a percent of CD4+ or CD8+ thymocytes, respectively. (H) CD4+ and CD8+ T cells and NK1.1+ cells as a percent of splenocytes.  Bars indicate the mean (+ range) of three to seven mice per genotype analyzed, except for A, in which the mean (+ range) is shown for two WT mice. Reproduced from (1).
Figure 3. The primary T-dependent antibody response is impaired in cellophane mice. Please see bumble and Worker for more details about these mutations.  WT and mutant mice were immunized with alum-precipitated NP-CGG.  Serum levels of NP-specific IgM (A) and total NP-specific IgG1 (B) were measured by ELISA.  Each point represents the mean (+ range) of greater than or equal to three mice per genotype.  Reproduced from (1).
Figure 4.  cellophane mice have a defect in GC formation. Please see bumble for more information on this mutant.  (A) CD38 vs. Peanut Agglutinin (PNA) binding on IgDlow B220high splenocytes from naïve WT or WT and bumble mice immunized 14 d before (Upper) or WT and cellophane mice immunized 11 d before (Lower) with alum-precipitated NP-CGG. Plots are representative of two to five naïve WT and three to seven immunized mice per genotype. (B) GC B cells as a percent of total B220high splenocytes in naïve WT or immunized WT and bumble mice. WT vs. bumble, P > 0.05, Mann–Whitney U test. (C ) GC B cells as a percent of total B220high splenocytes in naïve WT or immunized WT and cellophane mice. WT vs. cellophane, P < 0.005, Mann–Whitney U test. In B and C, each point represents data from one mouse, and the bar indicates the mean of all values. Reproduced from (1).
Figure 5.  The memory B-cell response is reduced in cellophane mice. Please see bumble and Worker for more details about these mutations.  CD19 B cells were enriched from the spleens of WT or mutant mice immunized 28 d before with alum-precipitated NP-CGG and transferred to CGG-primed WT mice. Immediately after cell transfer, the recipients were injected with NP-CGG in saline; 10 d later, serum levels of NP-specific IgM (A), total NP-specific IgG1 (B), and high-affinity NP-specific IgG1 (C ) were measured by ELISA. As controls, NP-specific IgM and IgG1 levels were measured in naïve mice and mice that were immunized with NP-CGG in saline but did not receive memory B cells (boost only). In A and B, all of the recipients made IgM and IgG1 responses equivalent to boost-only controls, indicating that memory B cells were not required for these responses. In C, only mice that received functional memory B cells and were boosted with NP-CGG made high-affinity NP-specific IgG1 responses. Each point represents data from one mouse, and the bar indicates the mean of all values. Background (indicated by the dashed line) was determined by incubating pooled WT sera on uncoated ELISA wells. Reproduced from (1).

The cellophane phenotype was identified in the T-dependent humoral response screen and in the T-independent B cell response screen.  Cellophane mice were unable to mount T-independent IgM responses following immunization with (4-hydroxy-3-nitrophenylacetyl)-Ficoll (NP-Ficoll) (Figure 1, left), and made suboptimal T-dependent IgG responses to a recombinant Semliki Forest virus (rSFV)-encoded β-galactosidase (rSFV-βGal) (Figure 1, right(1).  The frequency of mature B cells in the bone marrow and spleen of cellophane mice (Figure 2A and B) was normal, although B cells from cellophane mice expressed lower surface levels of IgD than B cells from wild type mice (1).  However, cellophane mice had reduced frequencies of splenic marginal zone (MZ) B cells (Figure 2B) and peritoneal B1 B cells (Figure 2D); the frequencies of B1a and B1b B cells were equally affected (Figure 2E(1)Cellophane mice had small, hypocellular thymi, and significant changes in thymic cell subsets (i.e. DN and SP thymocyte frequencies were increased; DP thymocyte frequencies were decreased) (Figure 2F(1).  The frequency of splenic T cells was not affected, but splenic NK cells were reduced (Figure 2H). Cellophane mice had normal levels of all Ig subtypes assayed with the exception of IgG2a, which was slightly reduced (1).


Seven days post-immunization with NP-CGG, cellophane mice had undetectable levels of NP-specific IgM (Figure 3A) and IgG1 (Figure 3B(1).  At day 14 post-immunization, the cellophane NP-specific IgG1 response was reduced compared to wild-type animals (Figure 3B), while NP-specific IgM remained undetectable (1).  The IgG1 in the cellophane mice was determined to be low affinity at 28 days post-immunization (Figure 3C), indicating that these mice do not support functional germinal centers (GCs).   The frequency of GC B cells was reduced in the cellophane mice (Figure 4A and C(1).  In addition, splenic B cells from immunized cellophane mice were unable to transfer NP-specific memory to naïve wild type mice (Figure 5).  

Nature of Mutation
Figure 6.  Bulk segregation analysis (BSA) results from the cellophane mice.  The mutation was localized to chromosome 18 at position 15,408,257 bp (LOD score = 7.9).
Figure 7.  Trace file of the cellophane mutation.

The cellophane mutation was mapped by bulk segregation analysis (BSA) of F2 intercross offspring using C57BL/10J as the mapping strain (n=18 with mutant phenotype, 82 with normal phenotype).  The mutation showed strongest linkage with the marker at position 15408257 bp on Chromosome 18 (synthetic LOD=7.9) (Figure 6). Whole genome SOLiD sequencing identified a T to A transversion at base pair 5770552, 9.6 Mb from the lineage peak, on Chromosome 18 in the genomic region NC_000084.  This position corresponds to base pair 2725 of the Zeb1 transcript, in exon 7 of 8 exons (Figure 7).



897  -L--L--R--H--K--Y--E--H--T--G--K-


The mutated nucleotide is indicated in red lettering, and results in conversion of a tyrosine codon at position 902 to a premature stop codon.

Illustration of Mutations in
Gene & Protein
Protein Prediction
Figure 8. Domain structure of Zeb1. The cellophane mutation results in the conversion of a tyrosine codon to a premature stop codon at position 902 of the encoded protein. The Zeb1 incidental mutation observed in the aoba mice is also shown at position 529 (H529L).  RD, repression domain; p300 P/CAF, p300 P/CAF interaction; ZF, Zinc finger; NZF, N-terminal zinc finger cluster; SID, Smad interaction domain; HBD, homeobox domain; CID, CtBP interaction domain; NC2, NC2 binding; CZF, C-terminal zinc finger cluster; AD, activation domain. This image is interactive; click to show other mutations in Zeb1.
Figure 9. 3D structure of the Zeb1 homeodomain. UCSF Chimera model is based on PDB 2E19. Click on the 3D structure to view it rotate.

Zeb1 and Zeb2 form the mammalian Zeb family of transcription factors.  Their effects are typically repressive, although they have also been shown to activate the transcription of certain genes.  Mouse Zeb1 and Zeb2 are 42% identical in sequence overall, and approximately 90% identical across their two zinc finger arrays (see below).  Human and mouse Zeb1 are 88% identical.


Zeb1 contains seven C2H2-type zinc finger domains, each approximately 23 amino acids in length (2-5) (Figure 8).  Four are clustered near the N-terminus (NZF, spanning aa 150-272) and three near the C-terminus (CZF, spanning aa 882-959) of the 1117 amino acid protein (all amino acid numbering corresponds to the mouse sequence).  Near the center of the protein is a homeobox domain (aa 559-618) that bears similarity to POU homeodomains in the helical regions, but lacks a basic amino acid cluster usually found on the N-terminal side of the domain.  An NMR structure of the isolated Zeb1 homeodomain has been deposited in the Protein Data Bank (PDB 2E19) (Figure 9). 


The zinc finger arrays of Zeb1 bind to DNA, but the homeodomain does not (6;7).  Each isolated cluster of zinc fingers bound independently to the consensus sequence CAC(C/G)(T/G) (G/T), with highest affinity for CACCT(G) (5-7).  These sequences encompass a subset of E boxes (consensus CANNTG), which often occur in tandem and are typically bound by activating transcription factors of the basic helix loop helix (bHLH) family.  Some E box-like sequences can negatively regulate gene expression.  The isolated Zeb1 NZF and CZF domains appear to have similar binding specificity (5), but the CZF domain had higher binding affinity for target DNA sequences relative to the NZF domain (6).  The presence of both intact zinc finger clusters was required for DNA binding and repressor activity in the full Zeb1 protein (6).  In another study, Zeb1 was shown to regulate transcription either positively or negatively depending on alternative DNA-binding modes, through either the NZF or CZF domain alone (7).  Although the homeodomain of Zeb1 had no specific DNA-binding activity, it has been shown to interact with the NZF domain of Zeb1 itself (7)


Like Zeb1, the 1215 amino acid Zeb2 has an NZF with four individual zinc fingers (three C2H2 and one CCHC), and a CZF with three C2H2 zinc fingers (8).  The DNA binding specificity of Zeb2 appears to mirror that of Zeb1 for CACCT sequences.  Similarly to Zeb1, the presence of intact NZF and CZF was required for high-affinity binding to DNA and optimal repressor activity (9).  Zeb2 has a centrally located homeobox domain that is 42% identical to that of Zeb1.


The existence of a repressor domain in Zeb1 has been substantiated by experiments in which portions of Zeb1 were tethered to heterologous DNA binding domains and tested for their effect on transcription of a reporter construct.  However, different groups have attributed repressive activity to distinct regions of the Zeb1 protein.  Repressor domains have been identified at the N-terminus (aa 19-127) (6), in a large central region between the two zinc finger clusters (aa 303-902) (10), and in two separate regions within aa 303-902 (aa 302-542 and 760-902) (11).  The region between the CZF domain and the C terminus is highly rich in glutamic acid residues (38%), and has been characterized as an activation domain (aa 1011-1124) (12).


Discoveries of interactors have led to the identification of several protein binding domains in Zeb1.  (i) Three five-amino acid sequences (PLNLC, PLDLS, PLNLS) between the homeodomain and the CZF domain bind to the co-repressor CtBP, and this region (spanning aa 685-749) is designated the CtBP interaction domain (CID) (13).  CtBP exists in a complex containing Zeb1 and Zeb2 and multiple histone-modifying proteins that act to downregulate gene expression (14).  (ii) A region between the NZF and homeobox domains of Zeb1 was identified as a Smad interaction domain (aa 377-456) for Smad1, Smad2, and Smad3 (15).  Smad transcription factors are activated upon stimulation of receptors for TGF-β or BMP, and translocate to the nucleus to activate transcription of target genes.  Zeb1 synergizes with TGF-β/BMP in transcriptional activation by aiding in the recruitment of p300 and P/CAF (histone acetylases) through a direct interaction involving the N-terminal region of Zeb1 (16).  (iii) Finally, the negative cofactor NC2 binds to amino acids 726-829 of Zeb1 in yeast two-hybrid assays, and can mediate Zeb1-dependent repression of a reporter in vitro (12).  NC2 binds to TATA-binding protein and prevents assembly of the transcription initiation complex (17).


Zeb1 is both phosphorylated and sumoylated.  Phosphorylation at serine and threonine residues, but not tyrosine residues, results in two distinctly migrating species upon gel electrophoresis (18).  These species are differentially expressed among several cell lines.  Zeb1 is sumoylated in vitro on lysine 327 (IK327TE) and lysine 752 (AK752KE) (19), but the effect of this posttranslational modification on Zeb1 function is unknown.  Sumoylation of Zeb2 by polycomb protein Pc2 has been shown to attenuate the repressive effect of Zeb2 at the E-cadherin promoter (19).


The cellophane mutation truncates the Zeb1 protein after amino acid 901, resulting in a protein lacking the C-terminal 216 amino acids that encompass part of the CZF domain and the glutamic acid-rich activation domain.


By Northern analysis, Zeb1 mRNA was detected in heart, brain, placenta, and muscle, and at low levels in lung, liver, and kidney (20).  In the brain, Zeb1 mRNA was in cerebellum, medulla, spinal cord, occipital lobe, frontal lobe, temporal lobe, and putamen.  Among muscle tissues, Zeb1 mRNA was found in skeletal muscle, heart, and in smooth muscle of the uterus, colon, and bladder (20;21)


Immune system tissues expressing Zeb1 mRNA were spleen, lymph node, thymus, and bone marrow (20).  However, another group reported Zeb1 transcript in thymocytes and at very low levels in bone marrow cells, but not in splenocytes consisting of only mature B and T cells; they concluded that expression of Zeb1 is downregulated upon migration of cells to the periphery (22).  Zeb1 mRNA was detected in the transformed human B cell line Namalwa (5).


By immunohistochemistry and in situ hybridization, the embryonic expression of Zeb1 was observed in mesodermal tissues (e.g., notochord, somite, limb bud mesenchyme, heart); in neural crest derivatives (e.g., dorsal root ganglia, cephalic ganglia); in neuroectoderm; and in parts of the central nervous system (hindbrain, motor neurons in the spinal cord).  This pattern was similar in mouse (22;23) and chicken embryos (2), except that Zeb1 was detected in the lens of the eye in chicken but not mouse embryos.


Zeb1 is found in the nucleus (2).  The hamster homolog of Zeb1 was shown to translocate from nucleus to cytoplasm in response to serum starvation of cells in culture (24).


Physiological functions of Zeb1

Zeb1 was discovered independently as δEF1 in chicken (2); Zfh1 in fly (3); Nil-2 (25), ZEB (5), and AREB6 (26) in humans; BZP in hamster (24); δEF1 (27) and MEB1 (4) in mouse; and Zfhep in rat (28).  Studies of mutant organisms deficient in Zeb1 demonstrated a role for Zeb1 in embryonic development, and in the differentiation of certain tissues. 


In the fly

In the fly, Zeb1 (Zfh-1) is expressed in the mesoderm of early embryos, in a number of mesodermally-derived structures of late embryos, and in many motor neurons of the developing CNS (29).  Flies with loss of function mutations of zfh-1 displayed local errors in cell fate or positioning, but normal segregation of the mesoderm and differentiation of mesodermally derived tissues (30).  Zfh-1 mutant embryos also showed mild alterations in the number and position of embryonic somatic muscles (30;31).  In addition, zfh-1 is required for germ cell migration and gonadal mesoderm development (32), and for formation of a subset of cells in the developing heart (33).


In the mouse

Mutant mice with homozygous null mutation of Zeb1 developed to term but died shortly after birth (23).  Overall, Zeb1-/- embryos appeared growth retarded and had short limbs beginning around E15. Shortened distal maxillae and mandibles, curled tails, edema, exencephaly, internal bleeding in the nasal region, and failure of spinal cord closure were observed in embryos at varying frequencies.  Many skeletal defects were observed in embryos, including craniofacial abnormalities (e.g. cleft palate, cartilage hyperplasia, dysplasia of nasal septum), limb defects (e.g. shortening and broadening of long bones, fusion of bones and joints), fusion of ribs, sternum defects, and hypoplasia of intervertebral discs.  Defects in chondrogenesis may be due in part to dysregulation of the transcription of genes encoding Indian hedgehog (Ihh) (34) and collagen II α1 (Col2a1) (35).  Aberrant transcriptional regulation of the collagen I α1 gene in osteoblasts may contribute to the bone abnormalities of Zeb1-/- embryos (36)Zeb1-/- embryos also had small hypocellular thymi with no histological distinction between medulla and cortex, and one tenth the number of thymocytes relative to wild type mice.  Female Zeb1+/- mice displayed increased fat mass relative to wild type female mice when fed either regular chow or high-fat chow (37).


In contrast, mice with a truncation of the C terminus of Zeb1 following amino acid 727 (Zeb1ΔC727/ΔC727), and therefore lacking the cluster of C-terminal zinc fingers and the Glu-rich domain, displayed none of the skeletal defects of Zeb1-/- mice (22).  Approximately 80% of Zeb1ΔC727/ΔC727 mice died within two days after birth.  Like those of Zeb1-/- mice, thymi of surviving Zeb1ΔC727/ΔC727 mice were smaller and contained 0.2% to 1% the number of thymocytes found in wild type mice (22).  The medulla and cortex were indistinguishable upon histological analysis of Zeb1ΔC727/ΔC727 thymus sections.  Spleen size and cellularity were similar between Zeb1ΔC727/ΔC727 and wild type mice, although there was a trend toward reduced cellularity in Zeb1ΔC727/ΔC727 mice.  Lymph node cellularity was 10% of that in wild type mice.  Cell populations in thymus, spleen, and lymph node were analyzed by flow cytometry (Table 1).  There was a depletion of c-kit+ thymocytes.  Of those T cells that reached maturity in Zeb1ΔC727/ΔC727 mice most were CD4+, consistent with the existence of an E-box sequence in a Cd4 enhancer to which Zeb1 binds to repress transcription (38).  Forward light scatter analysis showed a significant population of DP thymocytes was larger, and a population of DN thymocytes was smaller in Zeb1ΔC727/ΔC727 mice than in Zeb1ΔC727/+ mice, although the significance of this observation is unknown.  Zeb1ΔC727/ΔC727 splenic T cells proliferated normally upon Con A stimulation in vitro.


Table 1. Percentages of cell types in thymus, spleen, and LN of Zeb1ΔC727/ΔC727 mice





CD4+CD8+ (DP)












DN CD44+CD25-



DN CD44+CD25+



DN CD44-CD25+



DN CD44-CD25-






















Lymph node

















Similar numbers of B cells existed in Zeb1ΔC727/ΔC727 and Zeb1ΔC727/+ mice.  The population of myeloid cells was reported to be normal in Zeb1ΔC727/ΔC727 mice.


Although no muscle defects have been reported in Zeb1 mutant mice, several groups have provided evidence that Zeb1 regulates muscle cell differentiation.  The transcription factor p73 is transcriptionally activated by muscle regulatory factors MyoD, myogenin, Myf5 and Myf6, which bind to E-box sequences of enhancers.  Zeb1 competes with these bHLH transcription factors for binding to E-box sites, and has been shown to inhibit MyoD-induced myogenesis in vitro (10;39).  Zeb1 also opposes MyoD/Myf5- and MyoD/Myf6-mediated transcriptional activation of p73 by binding to a silencer element in the first intron of p73 (40;41).  Zeb1 also negatively regulates integrin α7 expression in myoblasts during skeletal muscle myogenesis by binding to the negative regulatory region in the promoter (42).  In vascular smooth muscle cells (SMC), Zeb1 represses expression of collagen I α2 by competing with transcriptional activator Nkx2.5 for binding to an enhancer element (43).  Zeb1 also promotes transcription during differentiation of vascular SMC through cooperation with serum response factor (SRF) and Smad3 downstream of TGF-β signaling (21).  Zeb1 binds to Smad1, Smad2, and Smad3 (15), and promotes the recruitment of histone acetylases p300 and P/CAF, which act to increase the accessibility of DNA to the transcriptional apparatus (16).


Mechanisms of transcriptional repression and activation

Figure 10. Zeb1 acts as both a transcription repressor and activator.  See text for more details.

Zeb1 has been shown to repress transcription by several mechanisms (Figure 10).  First, it may recruit co-repressors, such as CtBP (14), NC2 (12), and BRG1 (44), to target genes to modify chromatin and histone organization and/or to block assembly of the transcriptional machinery.  Second, it competes with transcriptional activators for binding to promoters and/or enhancers, as for example to the Ig heavy chain enhancer (5), the CD4 promoter (38), and the integrin α7 promoter (42).  Third, Zeb1 may directly repress transcription through repressor domains located at its N and C termini; mutants lacking one of these repressor domains displayed reduced repressive activity despite retaining DNA binding activity (6;10).


Zeb1 also acts as a transcriptional activator, as has been demonstrated for the genes encoding the vitamin D3 receptor, cyclin G2, and p130 (45;46).  Zeb1 can aid in the recruitment of histone acetylases, as mentioned above, and may also directly activate transcription through its activation domain.


Zeb1 in cancer

Through its ability to promote the ‘epithelial to mesenchymal transition’ (EMT), Zeb1 has been implicated in cancer invasion and metastasis.  During the EMT, epithelial cells lose polarity and contact with neighboring cells, and adopt a mesenchymal phenotype characterized by the acquisition of migratory and invasive properties (47).  This transition is necessary during many developmental processes, such as gastrulation, neural crest formation, and heart development.  EMT also plays an important role in tumor progression by supporting tumor cell extension, detachment, and invasion into the adjacent stroma (47).  Zeb1 induces EMT in many human cancers, including prostate, colon, breast, and pancreatic cancers, and is known to do so by suppressing the expression of basement membrane components (48) and cell polarity factors (49).  Zeb1 also represses the cell adhesion molecule E-cadherin through binding to conserved E-boxes in the E-cadherin promoter [(50), reviewed in (51)].  Downregulation of E-cadherin is a key feature of EMT, and E-cadherin is repressed during malignant transformation.  Zeb1 (and Zeb2) are also the most prominent targets of the miR-200 family of microRNAs, which has been shown to revert EMT and induce epithelial differentiation [(52), reviewed in (53)].  miR-200 family members are divided into subgroup I (miR-200a and miR-141) and subgroup II (miR-200b, miR-200c and miR-429).  The Zeb1 3’ UTR contains eight miR-200 binding sites, five for subgroup II and three for subgroup I.  Conversely, Zeb1 directly inhibits the transcription of miR-200 family microRNAs (54).


Heterozygous null mutations of ZEB1 in humans are linked to posterior polymorphous corneal dystrophy-3 (PPCD3) (OMIM #609141), and have been estimated to cause one-third to one-half of all PPCD cases (55;56).  PPCD is an autosomal dominant corneal disorder usually affecting both eyes and characterized by hyperplasia of corneal endothelial cells, which adopt an epithelial phenotype and gene expression pattern, and produce an abnormal basement membrane (Descemet’s membrane).  The severity of PPCD symptoms varies, and may include opacities, irodocorneal adhesions (abnormal connection of the iris and cornea), corneal edema, corectopia (displacement of the pupil from its normal, central position) and secondary glaucoma.  Mutations in ZEB1 causative for PPCD are associated with abdominal and inguinal hernias (57).  In contrast to null mutations, hypomorphic mutations of ZEB1 cause late-onset Fuchs corneal dystrophy, in which guttae form and Descemet’s membrane thickens, leading to corneal edema (58).


Similar to humans with PPCD, Zeb1 homozygous null mouse embryos and heterozygous adult mice displayed thickened corneas with hyperplasia and epithelialization of the endothelium (59).  Corneal endothelia of mutant embryos aberrantly expressed epithelial markers including cytokeratin, E-cadherin, and collagen IV α3 (COL4A3) (59).  Mutations in human COL4A3, COL4A4, and COL4A5 cause basement membrane abnormalities resulting in the renal disease Alport syndrome (OMIM #301050; #203780) (see aoba, a mouse Col4a4 allele (59)).  A binding site for ZEB1 was identified in the promoter of COL4A3, and ectopic expression of COL4A3 was observed in the corneal endothelium of PPCD patients (55).  Thus, ZEB1 mutations may result in de-repression of COL4A3 leading to the observed defects in Descemet’s membrane in PPCD.

Putative Mechanism
Figure 11.  Zeb1 is required for B-cell proliferation after B-cell receptor cross-linking.  Please see bumble for more information on Nfkbid.  Induction of (A) Nfkbid and (B) Zeb1 measured by real-time PCR in WT splenic B cells stimulated for up to 24 h with the indicated stimuli.  Each point represents the mean for triplicate samples. (C) Histograms showing carboxyfluorescein succinimidyl ester dilution peaks from mock-treated WT B cells and WT or mutant B cells stimulated with the indicated concentrations of F(ab')2αIgM, CpG, or LPS for 72 h.  Results are representative of triplicate cultures.  (D) Percentage of B cells that divided at least one time based on carboxy-fluorescein succinimidyl ester dilution.  Bars (+ range) indicate the mean of triplicate cultures. Reproduced from (1).

The cellophane mutation is believed to result in a truncated Zeb1 protein with hypomorphic function.  In support of this idea, cellophane mice displayed a phenotype similar to that of Zeb1ΔC727/ΔC727 mice (22) in that their thymi were small and hypocellular, they had fewer DP thymocytes and expanded SP thymocytes, and mature B cell frequencies were normal in spleen and bone marrow.  Furthermore, the phenotypes of Zeb1ΔC727/ΔC727 mice and cellophane mice were considerably less severe than that of homozygous null mice, which displayed numerous skeletal defects.


In wild type B cells, Zeb1 mRNA was upregulated within 30 minutes after BCR cross-linking (Figure 11B(1).   This finding suggests that Zeb1 regulates early cellular events following BCR activation.  Consistent with this hypothesis, B cells from cellophane mice displayed reduced proliferation in response to BCR crosslinking, although proliferation was normal after treatment with CpG oligonucleotides (Figure 11C and D).  We propose that the reduced ability of cellophane B cells to proliferate in response to BCR stimulation leads to impaired antibody responses and GC formation following immunization.  More studies are needed to identify the targets of Zeb1 regulation in B cells, and to determine the mechanism by which Zeb1 controls B cell proliferation during humoral immune responses.

Primers Primers cannot be located by automatic search.

Cellophane genotyping is performed by amplifying the region containing the mutation using PCR, followed by sequencing of the amplified region to detect the single nucleotide change.






PCR program

1) 95°C             2:00

2) 95°C             0:30

3) 56°C             0:30

4) 72°C             1:00

5) repeat steps (2-4) 29X

6) 72°C             7:00

7)  4°C              ∞


Primers for sequencing

cellophane_seq(F): 5’- GTGTAGATCCCAGGGATTCAACTC -3’

cellophane_seq(R): 5’- CACCAAGGCATTAAAGGCG -3’


The following sequence of 670 nucleotides (from Genbank genomic region NC_000084.5 for linear genomic sequence of Zeb1, sense strand) is amplified:


178206      tcagg tggagggctt cacatctaac atgaaatttc tctcttgtgc atctcttgac

178261 tttcctacat ttattgactt attggcagtg ggtgctttgt gtgccagctc tcacacgtgg

178321 aggtccgagg acaacttgtg ggagtcagtt ctctcctttc accgtgtaga tcccagggat

178381 tcaactctgc tcatctttct ttgtggcaaa tagctttgtc tgctgagcca tctcagtgac

178441 cttatctctt gggaatcttt ttctttgaca tttaatcttc tttttccact taggatgaaa

178501 gacaagacac tagctcagaa ggagtctcca ctgtggagga ccagaatgac tctgactcca

178561 cgccacccaa aaagaaaact cggaagacag agaatggaat gtatgcatgt gacctgtgtg

178621 acaagatatt tcagaagagc agctcactgt tgagacacaa atatgagcac acaggtgtgt

178681 gggggacctg ggcacgaggt tctaaaggtg cctgtggccc agtgcacatg aaacatgccc

178741 atagtgtgtg acgttctcag ctctgctccg tggtctccac ttttcacata ttcctactgc

178801 ggagcagtgg cttggctgtc tgctgcctag actcccgttt cgcctttaat gccttggtgt

178861 ctatgctgac agagc


Primer binding sites are underlined; sequencing primer binding sites are highlighted in gray; the mutated T is indicated in red.

Science Writers Eva Marie Y. Moresco, Anne Murray
Illustrators Diantha La Vine
AuthorsCarrie N. Arnold, Elaine Pirie, Bruce Beutler