UBR2 of the N-end rule pathway is required for chromosome stability via histone ubiquitylation in spermatocytes and somatic cells

Abstract

The N-end rule pathway is a proteolytic system in which its recognition components (N-recognins) recognize destabilizing N-terminal residues of short-lived proteins as an essential element of specific degrons, called N-degrons. The RING E3 ligases UBR2 and UBR1 are major N-recognins that share size (200 kDa), conserved domains and substrate specificities to N-degrons. Despite the known function of the N-end rule pathway in degradation of cytosolic proteins, the major phenotype of UBR2-deficient male mice is infertility caused by arrest of spermatocytes at meiotic prophase I. UBR2-deficient spermatocytes are impaired in transcriptional silencing of sex chromosome-linked genes and ubiquitylation of histone H2A. In this study we show that the recruitment of UBR2 to meiotic chromosomes spatiotemporally correlates to the induction of chromatin-associated ubiquitylation, which is significantly impaired in UBR2-deficient spermatocytes. UBR2 functions as a scaffold E3 that promotes HR6B/UbcH2-dependent ubiquitylation of H2A and H2B but not H3 and H4, through a mechanism distinct from typical polyubiquitylation. The E3 activity of UBR2 in histone ubiquitylation is allosterically activated by dipeptides bearing destabilizing N-terminal residues. Insufficient monoubiquitylation and polyubiquitylation on UBR2-deficient meiotic chromosomes correlate to defects in double strand break (DSB) repair and other meiotic processes, resulting in pachytene arrest at stage IV and apoptosis. Some of these functions of UBR2 are observed in somatic cells, in which UBR2 is a chromatin-binding protein involved in chromatin-associated ubiquitylation upon DNA damage. UBR2-deficient somatic cells show an array of chromosomal abnormalities, including hyperproliferation, chromosome instability, and hypersensitivity to DNA damage-inducing reagents. UBR2-deficient mice enriched in C57 background die upon birth with defects in lung expansion and neural development. Thus, UBR2, known as the recognition component of a major cellular proteolytic system, is associated with chromatin and controls chromatin dynamics and gene expression in both germ cells and somatic cells.

Figure 1. The localization of Ub conjugates on meiotic chromosomes in comparison with UBR2.

Surface-spread meiotic chromosomes were coimmunostained with FK2 antibody (red) which recognizes both monoubiquitin and polyubiquitin conjugates and an antibody to UBR2 (green). (A) Leptotene. (B) Mid−/late zygotene. (C) Early pachytene. (D) Mid-pachytene. (E) Mid-late pachytene. (F, G) The comparison of FK2 and UBR2 signals near the telomere region. Both Ub and UBR2 signals are enriched in specific regions of meiotic chromosomes until mid-pachytene, with significant colocalization along unsynapsed axial regions. At mid-pachytene, both signals are drastically induced in the majority of chromosomal regions except that the chromatin domain of sex chromosomes is relatively devoid of the UBR2 staining. The spermatocytes were staged based on the localization profile of UBR2 and the topology of the sex chromosomes. Scale bar: 10 µm (A–E), 5 µm (F, G).

Figure 2. The localization of polyubiquitin conjugates on meiotic chromosomes in comparison with UBR2.

Surface-spread meiotic chromosomes were coimmunostained with FK1 antibody (red) which specifically recognizes polyubiquitin conjugates (poly-Ub) and an antibody to UBR2 (green). (A) Zygotene. (B) Early pachytene. (C) Mid-pachytene. (D) Mid-late pachytene. Different from the FK2 staining, polyubiquitin conjugates are enriched in sex chromosomes until mid-pachytene. At mid-pachytene, both polyubiquitin and UBR2 signals are drastically induced in the majority of chromosomal regions except that the chromatin domain of sex chromosomes is relatively devoid of the UBR2 staining. Scale bar: 10 µm.

Figure 3. Genome-wide polyubiquitylation activities on meiotic chromosomes are reduced in UBR2^−/− spermatocytes.

Surface-spread meiotic chromosomes were coimmunostained with FK1 antibody (red) which specifically recognizes polyubiquitin conjugates and an antibody to SCP3 (green), a component of the synaptonemal complex. (A) Zygotene. (B) Early pachytene. (C) Mid-pachytene. Arrowhead indicates the sex chromosome. Note that polyubiquitin signals are virtually nondetectible in UBR2^−/− chromosomes. Scale bar: 10 µm.

Figure 4. Total ubiquitylation activities are significantly reduced in UBR2^−/− spermatocytes at pachytene.

Surface-spread meiotic chromosomes were coimmunostained with FK2 antibody (red) which recognizes both monoubiquitin and polyubiquitin conjugates and an antibody to SCP3 (green), a component of the synaptonemal complex. (A, B) Leptotene. (C, D) Zygotene. (E, F) Early pachytene. (G, H) Mid-pachytene. (I, J) Late pachytene. Compared with controls, the FK2 staining in UBR2^−/− chromosomes is relatively weak in the XY body at mid-pachytene and throughout the entire chromosomal regions at mid-late pachytene. Scale bar: 10 µm.

Figure 5. UBR2 mediates monoubiquitylation and polyubiquitylation of H2A and H2B but not H3 and H4.

(A) In vitro ubiquitylation assay (20 µL) with 1 µg of histone H2A, H2B, H1, or H3. The reaction contains 100 ng E3-F (or E3-V) prepared from rat testes, 30 ng UbcH2, and Ub activating reagents, including 1 µg flag-Ub and 100 ng E1. E3-F and E3-V represent protein mixtures that have been captured by Phe-peptide and Val-peptide, respectively. (B) In vitro binding assay in which UBR2 (as a mixture with UBR1) immobilized on Phe-peptide-beads was mixed with histone H2A, H2B, or H3 in the presence of HR6B, E1, and Ub activating reagents, followed by immunoblotting of histones retained by X-peptide (X = Phe or Val). (C) The screening of E2s which can support E3-F-mediated ubiquitylation of H2A. In vitro ubiquitylation assays were performed as (A) with different E2s indicated above. In this screening, UbcH2 and UbcH5b showed reproducibly the E2 activity in H2A ubiquitylation. (D–F) Allosteric modulation, an additional E2, and synthetic ligands for UBR2. (D) The interaction between UBR2 and HR6B is cooperatively accelerated by H2A and Arg-Ala. UBR2 (0.2 µg) from 10 mg rat testes extracts were immobilized with Phe-peptide conjugated with beads. Precipitated E3-peptide beads were mixed with 60 ng HR6B, 1 µg H2A, and/or 2 mM Arg-Ala, followed by immunoblotting analysis. (E) The HR6B-H2A interaction is cooperatively facilitated by UBR2 and Arg-Ala. GST-pulldown assays were done with 200 ng GST-HR6B, 200 ng UBR2, 1 µg H2A, and/or 2 mM Arg-Ala. (F) UBR2-dependent H2A ubiquitylation is synergistically activated by type-1 and type-2 N-end rule ligands.

Figure 6. Pachytene arrest of UBR2^−/− spermatocytes at stage IV.

(A–D) Testis sections from 8-week (A–C) and 3-week (D) old UBR2^−/− tubules. (A) Tubules that did not yet reach epithelial stage IV. A large number of spermatocytes (thick arrow) are present, indicating that the spermatogonial compartment keeps forming spermatocytes. Arrowhead, diplotene spermatocyte; thin arrow, round spermatid. The predecessors of these cells survived the stage IV arrest. (B) Tubules in epithelial stage IV as evidenced by the presence of large, G2 phase intermediate (In) spermatogonia (blue arrow) about to or dividing into B spermatogonia (yellow arrowhead). There is massive apoptosis of spermatocytes (asterisks). (C) Tubules after stage IV. A variable number of spermatocytes survive the passage through stage IV. The left tubule shows only one spermatocyte (arrowhead) and a few round spermatids (thin arrow) that stem from spermatocytes that survived stage IV one epithelial cycle earlier. The tubule on the right shows more spermatocytes (arrowhead) and round spermatids (black arrow) and even a few elongated spermatids (yellow arrow). (D) Stage IV arrest at the age of 3 weeks. The lower tubule shows massive apoptosis (asterisk). The upper tubule is after stage IV and shows two surviving spermatocytes (arrowhead), indicating that the arrest was already present before three weeks. (E) Surface-spread chromosomes of 781 control and 691 UBR2^−/− spermatocytes isolated from mice at P17 were stained with SCP3 and staged based on the morphology of SCP3-positive chromosomes. (F) Surface-spread chromosomes of 344 +/+ and 161 UBR2^−/− pachytene spermatocytes were substaged.

Figure 7. Defective DSB repair in UBR2^−/− spermatocytes.

(A) Prolonged retention of the γH2AX staining in pachytene chromosomes of UBR2^−/− spermatocytes. Wild type and UBR2^−/− spermatocytes were stained with γH2AX (green) and SCP3 (red). Scale bars: 10 µm. (B) The near absence of MLH1 (green) foci in UBR2^−/− spermatocytes at pachytene. (C) Asynapsis in UBR2^−/− spermatocytes at pachytene as indicated by unpaired autosomes segregated near the X-Y pair. Control and UBR2^−/− chromosomes were stained for RNA polymerase II, SCP3, or BRCA1.

Figure 8. UBR2 is associated with chromatin during cell cycle of somatic cells.

(A) UBR2 is enriched in the nucleus of MEFs. MEFs were stained for UBR2 without (top) and with a peptide that has been used to raise the antibody. (B) UBR2 is associated with chromatin in MEFs. Control and UBR2^−/− cells were separated into cytosolic, nuclear soluble, and chromatin-bound fractions in the presence or absence of iodoacetamide, followed by immunoblotting for proteins indicated. (C) Chromatin association of UBR2 is cell cycle-dependent. HeLa cells were synchronized at the G1-S border using the double thymidine block, released from G1-S arrest, and subjected to time-course fractionation and immunoblotting. Cell cycle stages were verified using flow cytometry and based on behaviors of cell cycle regulators, including down-regulation of chromatin-associated cyclin A and CDC6.

Figure 9. UBR2 is involved in ubiquitylation of chromatin-associated proteins in DNA-damaged somatic cells.

(A) MEFs were treated with 0.1 µg/ml mitomycin C (MMC) or irradiated with UV at 20 J/m². After 24 hrs later, cells were immunostained for UBR2 (red) and gH2AX (green). (B) HeLa and U2OS cells were treated with 0.1 µg/ml mitomycin C for 24 hrs and immunostained for UBR2. (C) Immunoblotting analysis of +/+ and UBR2^−/− MEFs treated with mitomycin C, UV, or MG132. (D) UBR2 is enriched in the nucleus of MEFs treated with 5 µM MG132. (E) Control and UBR2 knockdown U2OS cells were treated with 0.1 µg/ml mitomycin C for 24 hrs, followed by immunostaining with FK2 antibody.

Figure 10. Chromosome instability and hypersensitivity to DNA damage of UBR2-deficient somatic cells.

(A) UBR2-knockdown induces hyperproliferation in HeLa cells. (B) Metaphase chromosomes of UBR2^−/− MEFs show increased chromosomal aberrations, including breaks and fragmentations, compared with control cells. Arrowhead, break; Arrow, fragmentation (C) Quantitation of chromosomal abnormalities (breaks and fragments) observed in metaphase chromosomes from +/+ and UBR2^−/− MEFs. (D) UBR2^−/− MEFs are hypersensitive to hydroxyurea, or methyl methanesulfonate (Sigma).

Figure 11. Neonatal lethality in UBR2^−/− newborn pups associated with defects in lung expansion and neural development.

(A) The majority of UBR2^−/− mice in the C57 genetic background die neonatally. Shown is gross morphology of neonatal pups enriched in the C57 genetic background at P1. Surviving UBR2^−/− neonates weighed slightly less than their +/+ and UBR2^+/− littermates of the same gender. No gross morphological differences were observed between UBR2^−/− and control mice during embryogenesis and after birth. Arrowheads mark stomachs with or without milk, which indicate the feeding from mother. (B) The lungs of UBR2^−/− newborn pups are not properly expanded. Shown are H&E-stained cross sections from +/+ and UBR2^−/− lungs at P0. Arrowheads indicate alveoli. (C) In situ hybridization of UBR2 mRNA on cross sections of +/+ and UBR2^−/− brains at P1. (D) Dilated ventricles in UBR2^−/− brain (arrowhead). Shown are Nissl-stained cross sections of +/+ and UBR2^−/− brains at P1. (E) Defective gliogenesis and neuronal differentiation in hippocampus of UBR2^−/− brains at P1. Cross sections of a mildly affected UBR2^−/− brain, together with its littermate control, were subjected to Nissl or DAPI staining, or immunofluorescent staining of GFAP or NeuN.