APRIN is a unique Pds5 paralog with features of a chromatin regulator in hormonal differentiation

Abstract

Activation of steroid receptors results in global changes of gene expression patterns. Recent studies showed that steroid receptors control only a portion of their target genes directly, by promoter binding. The majority of the changes are indirect, through chromatin rearrangements. The mediators that relay the hormonal signals to large-scale chromatin changes are, however, unknown. We report here that APRIN, a novel hormone-induced nuclear phosphoprotein has the characteristics of a chromatin regulator and may link endocrine pathways to chromatin. We showed earlier that APRIN is involved in the hormonal regulation of proliferative arrest in cancer cells. To investigate its function we cloned and characterized APRIN orthologs and performed homology and expression studies. APRIN is a paralog of the cohesin-associated Pds5 gene lineage and arose by gene-duplication in early vertebrates. The conservation and domain differences we found suggest, however, that APRIN acquired novel chromatin-related functions (e.g. the HMG-like domains in APRIN, the hallmarks of chromatin regulators, are absent in the Pds5 family). Our results suggest that in interphase nuclei APRIN localizes in the euchromatin/heterochromatin interface and we also identified its DNA-binding and nuclear import signal domains. The results indicate that APRIN, in addition to its Pds5 similarity, has the features and localization of a hormone-induced chromatin regulator.

Figure 1

The chromosomal, contig and exon maps of the mouse APRIN gene. Mouse chromosome 5 is depicted with G-banding, the scale is in centimorgans (cM). Panel A, The full pPAC417G6 clone map (“mouse BRCA2-APR area” map line), the identified genes are depicted on a kilobase scale. The arrows indicate the direction of transcription. Exons were mapped on the mouse genomic supercontig, Mm5_WIFeb01_97. Panel B, APRIN mRNA and the open reading frame are shown. Panel C, the human and mouse APRIN chromosomal areas are compared (human chromosome 13 q12.3 and the corresponding mouse chromosome 5 region, 82-84 cM). The corresponding gene names and positions are indicated. Positions are given in million base pair units in the human chromosome segment and in centimorgan units in the mouse chromosome 5 segment

Figure 2

APRIN protein expression, nuclear transport and chromatin localization. Panel “a”, Western blot analysis of mouse prostate proteins using the anti APRIN-1370 antibody. The protein markers on the right are in kDa, the arrow indicates the mouse APRIN protein band. Panels “b” and “c”, anti-APRIN-1370 immuno histochemistry of mouse prostate sections, peroxidase-DAB staining, 100× (“b”) and 400× (“c”) magnifications. Panels “d”-“i”, expression and nuclear localization of transiently transfected FLAG-tagged APRIN protein. Panels “d”-“f”, nuclear immunofluorescence of MCF7-AR1 cells, transfected with an N-terminal FLAG-tag APRIN fusion construct. They were stained with mouse anti-FLAG and Alexa 594-labeled secondary antibody (red). Hoechst33258 was used as counterstain (blue) (2000×). Panels “g”-“i”, transfections with C-terminal truncated APRIN constructs (cytoplasmic). The host cell line and staining were the same as in the previous panels (1000×). Panels “j”-“l”, confocal microscopy of MCF7-AR1 cell nuclei transfected with the N-terminal FLAG-tag APRIN fusion construct. The anti-FLAG/FITC-labelled secondary antibody staining is green and the DNA counterstaining with propidium iodide is red. The overlap of FLAG expression and DNA staining indicates co-localization of APRIN with the nuclear DNA (yellow). The three panels show various compartmentalization patterns of the APRIN-FLAG fusion protein in the nucleus (2000×). Panels “m”-“o”, high resolution fluorescence immunohistochemistry of the native APRIN protein using the anti APRIN-1434 antibody. A single MCF7-AR1 nucleus is shown (6000×). Panel “m”, Anti APRIN-1434 only (green). Panel “n”, colocalization of APRIN with the chromatin. The DAPI counter-stain (blue) labels the heterochromatin. Panel “o”, the middle section of panel “n” is electronically magnified to show APRIN colocalization with the euchromatin/ heterochromatin interface. Panels “p”-“s”, APRIN expression in vivo, in prostate cancer. The panels depict formaldehyde-fixed, paraffin-embedded human prostate tissue sections in 400×, 1000× and 2000× magnifications, respectively. The sections were stained with anti-APRIN-1434 and with anti-rabbit peroxydase-DAB, using hematoxylin counter staining. The left panel shows a disorganized acinus in neoplasia, the middle and right panels show the same nuclei in increasing magnifications.

Figure 3

Protein conservation analysis between mouse and human proteins; APRIN is highly conserved. The positions of the boxes reflect the percentage of conservation between mouse and human in a decreasing order. The gene-boxes are aligned with the conservation scale on the y-axis. The x-axis shows the abbreviations of the names of genes. The boxes also indicate the sizes of the polypeptides in amino acids and the percent identities between the murine and human sequences. Groups of genes of particular functions are indicated, APRIN is highlighted.

Figure 4

Conservation-based domain map of the APRIN orthologs. The three large panels represent the graphic displays of conservational calculations between the species indicated. The numbered boxes at the top represent the identified conservation domains. The vertical scale shows protein identity in percents, the horizontal scale represents the amino acid positions along the APRIN polypeptide, in 20 amino acid units. The dotted lines identify the major domain or subdomain boundaries. The graph at the bottom (APRIN v. Pds5) represents the conservation map of the human paralog Pds5. The comparison is based on combined sequences from the KIAA0648 [72], (Acc#: AB014548) and the SCC-112 [73] (Acc#: NP_065015) sequences, which overlap and represent different isolates of the human Pds5 cDNA

Figure 5

Comparison of the APRIN conserved domain pattern with APRIN-related sequences in the Pfam domain database. Genes names (in bold) and organisms (in parentheses) are indicated on the left. The numbers at the top label the three major Pfam domain families recognized in the APRIN sequence. The scale and the conserved domains in the AS3 line are the same as in Figure 4. The boxes below the APRIN domains are EST data representing partial vertebrate sequences with homology to the human APRIN sequence. (EST accession numbers from N-terminal to C-terminal direction: bovine, BE756042, AW483768 and BF043498; chicken: BG712953, BG625340, BG625239 and AJ392316; Xenopus, BI448976, BG360016+BG346620, BG016343 and BI095461; zebrafish, BG515319, AI558326, AW175091, AW128709 and BI472753). The interruptions within the EST sequences signal regions of low conservation. The dotted lines between the boxes represent regions where sequence information was not available.

Figure 6

AT-hook domains and nuclear localization signals in the unique C-terminal sequence of APRIN. The C-terminal APRIN polypeptide sequences of the human (top sequence) and the mouse (bottom sequence) are shown in alignment, the numbers indicate amino acid positions. Domains #8, #9 and #10 are marked. The identified AT-hook motifs are boxed and labeled. The boxes also contain sequences of high similarity with other DNA binding proteins (Dd_AAC2, “AAC-rich mRNA” gene from Dictyostelium discoideum, acc.# P14196; Sc_SUM1, “suppressor of mar1 mutation” gene from Saccharomyces cerevisiae, acc # P46676; Dm_CPD1, “chromosomal protein D1” gene from Drosophila melanogaster, acc # P22058; and Ce_T11A5.1, “protein 1 from region T11A5” from Caenorhabditis elegans). The two classes of AT-hook domains (Class II and Class III) with close similarity to the APRIN elements are also shown [46]. The lysine (K) which is critical in Class III (highlighted) is missing in the APRIN sequence, classifying it to a Class II motif. The acidic domains adjacent to the AT-hooks are indicated by the ellipses. The sequences indicated by the bold long lines below and above the alignment represent the two bipartite nuclear localization signals (NLS1 and NLS2). The two positively charged elements within the localization signals are underlined. The single arrows above the human sequence indicate casein kinase II (CKII) and protein kinase C (PKC) targets. The amino acids in the recognition sequences are in bold. The double arrows indicate putative amidation and asparagine glycosylation sites.