An allele separating skeletal patterning and spermatogonial renewal functions of PLZF
© Ching et al; licensee BioMed Central Ltd. 2010
Received: 27 July 2009
Accepted: 25 March 2010
Published: 25 March 2010
Skip to main content
© Ching et al; licensee BioMed Central Ltd. 2010
Received: 27 July 2009
Accepted: 25 March 2010
Published: 25 March 2010
The promyelocytic leukemia zinc finger gene Plzf (also called Zbtb16, Zfp145 or Green's luxoid) belongs to the POZ/zinc-finger family of transcription factors. It contains a BTB/POZ domain that mediates epigenetic transcriptional repression. PLZF is essential for proper skeleton patterning and male germ cell renewal. Two alleles have been reported that display similar phenotypes: a targeted knock-out, and the spontaneous nonsense mutation luxoid.
We describe a new ENU induced missense allele of Plzf called seven toes (Plzf7t). Homozygous animals exhibit hindlimb and axial skeleton abnormalities. Whereas the skeletal abnormalities are similar to those of the other alleles, Plzf7tdiffers in that it does not cause spermatogonial depletion and infertility. Positional cloning revealed a point mutation changing the evolutionarily conserved amino acid Glu44 to Gly, possibly altering the BTB domain's activity.
Plzf7tis a separation-of-function allele that reveals differential requirements for domains of PLZF in different developmental milieus.
Plzf is a widely studied gene due to its involvement in varied biological processes. It was originally discovered as being fused with the RARA (retinoic acid receptor alpha) gene in translocations underlying acute promyelocytic leukemias in humans . The most prominent phenotypic features of Plzf null mice are defects in patterning of limb and axial skeleton structures . Affected animals exhibit homeotic-like anterior-to-posterior transformations in the axial and appendicular skeleton, causing hindlimb abnormalities such as extra digits and conversion of the fibula to a tibia-like structure. The spontaneous mutation luxoid (lu), later shown to be an allele of Plzf, is characterized by similar, albeit semidominant, limb abnormalities and recessive skeletal defects [3–5].
A critical function of PLZF is in spermatogonial stem cell renewal. Both the null and luxoid alleles cause progressive male infertility due to depletion of the spermatogonial stem cell pool with age [3, 6]. In the absence of PLZF, the balance between self-renewal and differentiation of spermatogonial stem cells is disrupted. PLZF is present as nuclear foci in Type A-single (putative stem cells), A-paired (undifferentiated spermatogonia), and A-aligned (differentiating) spermatogonia, but not the more differentiated Type B spermatogonia or in subsequent meiotic spermatocytes or postmeiotic spermatids .
PLZF contains nine Krüppel-type sequence-specific zinc finger DNA binding domains and a BTB/POZ protein-protein interaction domain in its N terminus. The latter allows BTB domain-containing proteins to form homo- or hetero-dimers [7, 8]. BTB-containing proteins, including PLZF, function as transcriptional repressors that affect gene expression via chromatin remodeling [7, 9, 10]. The repressive activity of the BTB domain can be autonomous, or it can mediate interactions with co-repressors that in turn recruit histone deacetylases, resulting in epigenetic silencing .
The skeletal patterning defects of Plzf null mice (officially Zbtb16tm1Ppp; abbreviated Plzf-/-) have been linked to failed transcriptional repression of the HoxD complex genes [2, 12]. PLZF was found to bind promoter regions in the AbdB HoxD complex, directly interacting with polycomb proteins, and thereby recruiting histone deacetylases to silence gene expression . The spermatogonial stem cell renewal defects in Plzf mutants have been linked to failed transcriptional silencing of Kit . The KIT receptor is present in differentiating spermatogonia in adult mice, and is required for the first wave of prepubertal spermatogonial expansion . Therefore, KIT is a critical factor in spermatogonial differentiation, and the control of its expression affects the balance of stem cell maintenance. PLZF directly binds the Kit promoter, and PLZF-deficient mice have elevated Kit transcription , presumably resulting in a tipping of the balance from stem cell maintenance towards differentiation, and thus eventual exhaustion of the spermatogonial stem cell pool.
Here, we report a separation-of-function allele of Plzf called seven toes (Plzf7t; officially Zbtb167t), which was recovered in an ENU mutagenesis screen . This allele exhibits skeletal phenotypes similar, but not identical, to the luxoid and null Plzf alleles. Importantly, Plzf7tlacks the germ cell depletion phenotypes of all previously reported Plzf alleles. Plzf7thas a missense mutation in the BTB domain, which must disrupt a molecular activity involved in normal skeleton morphogenesis that is separate from that required for silencing of Kit in spermatogonial regeneration.
7t arose in an N-ethyl-N-nitrosourea (ENU) mutagenesis screen that employed a balancer chromosome to detect mutations on proximal mouse chromosome 5 . Affected animals exhibited recessive limb patterning defects (the most overt being one or two extra digits on one or both hindlimbs; described below in greater detail) without other obvious abnormalities. F1 intercrosses indicated Mendelian segregation of the limb phenotype. However, it was evident that 7t was unlinked to Chromosome 5.
Complementation testing and comparative phenotypes of luxoid and 7t mice
Comparison of Plzf-/- and Plzf7t/7tskeletal phenotypes
digit I>II transformation
preaxial digit resembles digit V or digit I absence
anterior C1 to navicular
C1 to cuboid bone
metatarsal bony outgrowths
separation of tibia and fibula
tibia to fibula transformation
shortened zeugopod/absent fibula
duplication or absence of patella
absence of meniscus
14 thoracic vertebrae
extra pair of ribs attached to sternum
malformed xiphoid process
T10 to T11 vertebrae shift
homeotic transformation of sacral region
Shortening and kinked tail
In the zeugopod region, the tibia and fibula failed to fuse in 47% of 7t/7t mice (vs. 96% of null animals; Fig. 2f). Additionally, 40% of fibulae were thickened and 40% of tibiae were transformed to fibulae (compared to 32% and 38%, respectively, in Plzf-/-; Fig. 2f). There was a shortening of zeugopod/absence of fibula in 33% of limbs, but no animals develop duplication or absence of the patella or meniscus (data not shown).
All of the 7t/7t animals (N4-N5 backcross generations into strain C3HeB/FeJ) had 13 thoracic vertebrae as in wild-type, (Figs. 2g, h), but 70% of Plzf-/- mice were reported to have 14 . The additional segment of Plzf-/- vertebrae is associated with an extra pair of ribs that also attached to the sternum. Unlike the wild type, 73% of 7t/7t mice had elongated 8th and 9th ribs that attached to the sternum (Fig. 2i, j). All Plzf-/- mice had malformed xiphoid processes (lower part of sternum) compared to only 40% of the 7t/7t mice (data not shown). Unlike the nulls, none of the 7t homozygotes had a truncated and kinked tail. These results demonstrated that 7t causes a similar but milder axial phenotype than the reported knock-out allele, but the polydactyly phenotypes are essentially identical with higher penetrance. Table 2 summarizes the skeletal phenotypes of null and 7t/7t mice.
Whereas lu/lu and Plzf-/- males undergo male germ cell depletion and infertility due to defective spermatogonial stem cell renewal, 7t homozygotes were normally fertile. In matings to C3HeB/Fej (C3H) females, C3H semi-congenic males (N4-N5) sired an average of 6.5 ± 1.9 pups compared to 5.6 ± 2.3 for C3H control males (p = 0.036), indicating the 7t does not impact fertility negatively. Notably, lu/lu males co-isogenic on C3H/He (the strain on which lu arose) were found to be infertile (Green, 1955), thus making it unlikely that genetic background is responsible for the disparate lu and 7t fertility phenotypes.
lu/lu and Plzf-/- adults have small testes and markedly decreased numbers of germ cells in seminiferous tubules. This Sertoli Cell Only-like syndrome becomes more prominent with age as the stem cell pool declines [3, 6]. 7t/7t males did not exhibit small testes at any age. Nevertheless, we performed histological analysis of their testes to reveal potential germ cell depletion. There were no agametic tubules in young mice, unlike the null and luxoid alleles (Fig. 3a). Only at older ages (> 10 months) did some mice exhibit Sertoli Cell only tubules (< 2% of tubules; Fig. 3b) compared to <1% in wildtype (Fig. 3a). To quantify any possible minor effects on the germ cell pool, testis immunostaining for proliferating cell nuclear antigen (PCNA), a marker of cells in early G1 and S phases of the cell cycle, was performed on 7t/7t mice at 4, 6, 8 and 12 months, lu/lu mice at 3 months, and wild types at 8 months of age. Twenty randomly selected tubules from each testis were counted for PCNA+ cells. PCNA+ cells remained present in most tubules of 7t/7t testes at 8 months (data not shown) of age and older, with no evidence of quantitative decline (Fig. 3e). This contrasts with the lu/lu and null mutants, where over 50% of tubules were PCNA-negative at 8 months of age (Fig. 3c) [3, 6]. The 7t allele also rescued the male germ cell depletion phenotype of lu, as lu/7t had normal testes (Fig. 3d). These data indicate that the 7t allele does not appreciably affect spermatogonial stem cell maintenance.
Alignment of mouse PLZF amino acid sequence surrounding residue 44 indicated that this region is highly conserved amongst vertebrates including zebrafish (Danio rerio) and chicken (Gallus gallus) (Fig. 4b). However, the particular mutated amino acid (GLU44 in WT mouse PLZF) is not absolutely conserved in the BTB domains of all proteins of the BTB-ZF (BTB domain-containing zinc finger proteins) family (Fig. 4b).
The BTB domain structure of human PLZF was determined at 1.9A resolution, revealing a tightly intertwined dimer . Amino acid residue 44 is located in the 3rd β-sheet of the BTB domain, in a charged pocket that facilitates dimerization . It has been shown that mutations that cause misfolding of the BTB domain result in failed dimerization and therefore nonfunctional protein products . The proper charge alignment in the BTB pocket is required for PLZF transcriptional repression via histone deacetylase (HDAC) recruitment . Therefore, the Glu44Gly mutation may alter both the conformation and the charge of the BTB domain, thus affecting its function.
Through a forward genetics approach, we isolated a new allele (7t) of Plzf. The 7t/7t mice displayed skeleton abnormalities similar to the null and nonsense (lu) alleles, without the male germ cell depletion phenotype observed in the two previously reported alleles. A Glu44Gly mutation was found in the evolutionarily conserved BTB domain, which mediates protein-protein interactions. The Glu44Gly may weaken, disrupt or change the conformation of interactions between PLZF and itself or other BTB/POZ domain-containing proteins, thereby affecting the expression of genes involved in skeletal patterning. The expression of Hoxd genes, Hoxa13 and bone morphogenetic proteins (BMPs) are misregulated in Plzf-/- mice, however, expression of Shh and Fgf4 is not altered . Therefore, Plzf is likely to mediate aspects of anterior-posterior pattering in both the axial and limb skeleton through the BTB/POZ domain.
The lack of a germ cell depletion phenotype indicates that 7t is a separation-of-function mutation that can provide insight into the functional domains of PLZF. Given existing information, there are two likely mechanistic explanations for why the mutation specifically affects skeletal patterning. One is that alteration of the BTB domain abolishes co-repressor(s) binding in cell types involved in skeletal development, but does not alter binding of co-repressor(s) in germ cells required for silencing Kit. A second possibility is that the BTB domain is not required for PLZF's role in repressing Kit expression in spermatogonia. We hypothesize that since the zinc finger domains remain intact in the 7t allele, that sequence-specific binding of PLZF7tto the Kit promoter is unaffected by the BTB domain disruption. Interestingly, a human patient with skeletal defects and genital hypoplasia was identified as carrying a missense allele of PLZF in trans to a deletion . The nonconservative Met617Val alteration occurred in a zinc finger motif, causing functional disruption in reporter assays. These data suggest that the Zn finger domains (and presumably, DNA binding) are important for both skeletal and germ cell functions. This patient also had phenotypes that were not apparent or tested in any of the mouse Plzf mutant models, including severe mental retardation, short stature, dystrophy, microcephaly, cranial dysmorphisms, and cryptorchidism. It is likely that many of these phenotypes are related to the deletion, rather than homozygous loss of PLZF, since other patients with deletions in this region of 11q share the non-genital phenotypes .
The 7t/7t mutation was induced on strain C57BL/6J (B6) and backcrossed into C3H for five generations (N5). In the earlier generations (N1 to N3), the hindlimbs displayed more severe abnormalities (especially in the zeugopod region) than in later generations (N4 & N5). This may explain why the limb and axial skeleton phenotypes of 7t/7t mice were not as severe as in null mice (studied in a 129xB6 mixed background). During the positional cloning process, Plzf7twas outcrossed to several different inbred strains, including CAST/Ei. More than 147 F2 offspring from the CAST/Ei crosses were recovered, yielding the expected number of F2 homozygotes, but none of these displayed polydactyly. This presumably dominant modifier effect was not seen in similar crosses to strains BALB/cJ or DBA/2J. A similar modifier effect was observed in Twist mutants  and Strong's luxoid . These observations open the door to the potential identification of modifiers that function in the same developmental pathway as PLZF. Homozygous 7t males were fertile on all strain backgrounds; thus the phenotype modifications appear limited to skeletal defects.
This work demonstrates the utility of forward genetic approaches for the dissection of functional domains within a protein, and the roles of those domains in various cell types and developmental processes. The 7t separation of function mutation will be informative for eventual mechanistic elucidation of how this protein acts in different milieus, namely, skeletal development and in the germ lineage.
The mice used in this study were maintained first at The Jackson Laboratory, and later at Cornell University. The use of mice for these experiments was approved by the Institutional Animal Care and Use Committees (IACUC) of both institutions.
Skin and internal viscera were removed from mice, fixed overnight in 100% ethanol, followed by 4 days in acetone. The specimens were stained with Alcian blue and Alizarin red for 10 days followed by overnight treatment with 2% KOH. Samples were stored in a solution of 40% glycerol, 40% ethanol and 20% benzoalcohol.
Plzf7tarose on the strain C57BL/6J. Homozygous mutants were crossed to C3HeB/FeJ to generate F1 heterozygotes, which were intercrossed to yield 411 F2 progeny. Affected offspring were PCR genotyped with a panel of polymorphic microsatellite markers distributed across the genome, and association of the phenotype with homozygosity for B6 alleles at microsatellite loci revealed linkage.
Denaturing HPLC analysis using a Transgenomic WAVE machine was employed to scan for mutations in candidate genes within the 7t critical region. PCR primers were designed to amplify either cDNA or genomic sequences from heterozygous experimental and control animals. If sequence polymorphisms were detected in the dHPLC analysis, those amplicons were sequenced to identify the exact nucleotide change.
Testes were fixed in Bouin's, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). For immunohistochemistry, the sections were deparaffinized and dehydrated through Xylene/Ethanol/PBS gradients. Slides were subjected to antigen retrieval by boiling at 95°C for 40 minutes in 10 mM Tris-sodium citrate/PBS (pH = 6.0) with 0.05% Tween 20, followed by quenching and blocking in 3% H2O2/methanol for 15 min. A M.O.M. kit (Vector Laboratories) was used for blocking according to the manufacturer's instructions. The sections were incubated with a mouse monoclonal anti-PCNA antibody (clone PC10; Sigma; diluted 1:200) at 4°C overnight. Histostain-SP Kit (AEC, Broad Spectrum, Zymed®) was used to visualize binding.
Epididymides were minced with small scissors in 1 ml PBS then incubated at 37° for 15 minutes. A portion of the sperm suspension was diluted, mixed 1:1 with trypan blue, and live sperm were counted on a hemacytometer.
the structure of BTB was retrieved from the Entrez's 3D-structure database http://www.ncbi.nlm.nih.gov/Structure/Entrez/mmdb.shtml, and visualized using the Cn3D application http://www.ncbi.nlm.nih.gov/Structure/CN3D/cn3d.shtml
Tissues were disrupted in the Tissue Protein Extraction Reagent (T-per; PIERCE), centrifuged at 4°C for 10 minutes at 10,000 × G, and the supernatant was retained and stored at -80°. Protein concentrations were determined by the BCA method (PIERCE), using a Nanodrop (NanoDrop Technologies, LLC) spectrophotometer. 15 μg protein was mixed with loading buffer containing β-mercaptoethanol, boiled 3 min., and electrophoresced on 10% PAGE gels. Proteins were electrotransferred onto Immobilon-P PVDF membranes (Millipore). The membrane were blocked 1 h at RT in 1×TBS containing 5% skim milk and 0.05% Tween-20. Blots were probed at 4°C overnight in blocking solution with mouse anti-PLZF monoclonal D-9 (Santa Cruz Biotechnology, Inc; 1:1,000 dilution). The blots were then probed with an anti-mouse secondary antibody conjugated to horseradish peroxidase at a 1:3,000 dilution in blocking solution. Blots were washed three times for 5 min after the primary and the secondarychemiluminescence (ECL; PIERCE). Blots were stripped at room temperature in 0.1 M glycine-HCl (pH = 2.5-3.0) for 30 min, before reprobing with 1:10,000 anti β-actin (AC-15, Abcam) mouse monoclonal antibody and goat anti-mouse secondary antibody conjugated to horseradish peroxidase at 1:3,000 dilutions in blocking solution. Blots were washed three times for 5 min after the primary and the secondary antibody probing, and then visualized by chemiluminescence (ECL; PIERCE). Blots were stripped at room temperature in 0.1 M glycine-HCl (pH = 2.5-3.0) for 30 min, before reprobing with 1:10,000 anti β-actin (AC-15, Abcam) mouse monoclonal antibody and goat anti-mouse secondary antibody conjugated to horseradish peroxidase.
Luxoid mice were provided by R. Braun. This work was funded by NIH grant R01 HD035984 to JCS.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.