Ribonomic analysis of human DZIP1 reveals its involvement in ribonucleoprotein complexes and stress granules

Abstract

Background: DZIP1 (DAZ-interacting protein 1) has been described as a component of the Hh signaling pathway with a putative regulatory role in ciliogenesis. DZIP1 interacts with DAZ RNA binding proteins in embryonic stem cells and human germ cells suggesting a role in mRNA regulation.

Results: We investigated DZIP1 function in HeLa cells and its involvement in ribonucleoprotein complexes. DZIP1 was predominantly located in granules in the cytoplasm. Under oxidative stress conditions, DZIP1 re-localized to stress granules. DZIP appears to be important for the formation of stress granules during the stress response. We used immunoprecipitation assays with antibodies against DZIP1 and microarray hybridization to identify mRNAs associated with DZIP1. The genetic networks formed by the DZIP1-associated mRNAs were involved in cell cycle and gene expression regulation. DZIP1 is involved in the Hedgehog signaling pathway. We used cyclopamine, a specific inhibitor of this pathway, to analyze the expression of DZIP1 and its associated mRNAs. The abundance of DZIP1-associated mRNAs increased with treatment; however, the silencing or overexpression of DZIP1 in HeLa cells had no effect on the accumulation of the associated mRNAs. Polysomal profile analysis by sucrose gradient centrifugation demonstrated the presence of DZIP1 in the polysomal fraction.

Conclusions: Our results suggest that DZIP1 is part of an RNP complex that occupies various subcellular locations. The diversity of the mRNAs associated with DZIP1 suggests that this protein is a component of different RNPs associated with translating polysomes and with RNA granules.

Figure 1

DZIP1 is located predominantly in the cytoplasm and shows a granular distribution. (A) Indirect immunofluorescence staining of DZIP1 (green) in HeLa cells. (B) Nuclei counterstained with DAPI (blue) and (C) merged image. HeLa (D-F), HEK293 (G-I) and hTERT-RPE1 (J-O) cells were transfected with pDZIP1-GFP. (J-O) Ciliary axonemes were labeled with anti-acetylated tubulin antibodies (red). (M-O) are magnified sections from (K) (white boxes). Arrow = potentially compatible with a basal body localization. Scale bar: 10 μm.

Figure 2

DZIP1 is recruited to stress granules in cells under oxidative stress. (A-B) Indirect immunofluorescence staining was carried out to examine the colocalization of DZIP1 (green) and TIA1 (red) in HeLa cells. Nuclei were counterstained with DAPI (blue). (A) Control cells without arsenite sodium (scale bar: 10 μm) and (B) cells treated with arsenite sodium to induce oxidative stress. (C) Images show magnified sections from B (white boxes). (D) Fluorescence intensities of DZIP1 (green channel) and TIA1 (red channel) were measured on selected regions (a and b) in the merged images (left) of HeLa cells under oxidative stress. The region (a) shows no colocalization of fluorescent signals with higher fluorescence intensity in the green channel than in the red channel. The region (b) shows DZIP1 and TIA1 colocalization with equivalent fluorescence intensities in green and red channels. (E) Indirect immunofluorescence staining of DZIP1 (green) and DCP1 (red) antibody was used to detect colocalization with p-bodies. (F) Detail of colocalization of DZIP1 and DCP1 in a single cell. (B-F) Confocal microscopy.

Figure 3

Identification of DZIP1-associated mRNAs in HeLa cells. (A) Western blot of HeLa cell proteins immunoprecipitated with a specific anti-DZIP1 antibody and a control IgG, probed with a specific anti-DZIP1 antibody; (B) mRNAs associated with DZIP1. Microarray analysis of the immunoprecipitated RNA fractions. Rows correspond to individual transcripts and the color code indicates the degree of enrichment. Three independent assays, with anti-DZIP1 antibody and negative controls, are shown. (C) Quantitative RT-PCR analysis of transcript levels in DZIP1-IPs (DZIP1 immunoprecipitate) versus negative IPs (IP-Control), normalized to concentration. Enrichment of the following mRNAs was analyzed: IFT80, PUM1, SNX2, BRD8, PTCH, DISP1, NPC1, CSNK1E, MYO5B, CEP164, STK36 and ASH1L. The mean of technical triplicates is shown. (D) Quantitative RT-PCR of transcript levels in DZIP1-GFP IPs versus negative IPs (IP-GFP), normalized to concentration. Enrichment of the following mRNAs was analyzed: IFT80, PUM1, SNX2, BRD8, PTCH, DISP1, NPC1, CSNK1E, MYO5B, CEP164, STK36 and ASH1L. The mean of technical triplicates is shown.

Figure 4

The expression of DZIP1 and its mRNA targets is affected by blockade of the Hh pathway. (A-D) Indirect immunofluorescence staining was used to detect DZIP1 (red – alexa546), and nuclei were counterstained with DAPI (blue). (A-B) HeLa cells not treated with cyclopamine. (C-D) HeLa cells treated with 300 nM cyclopamine for 24 h. (E-H) Quantitative RT-PCR analysis of DZIP1, GLI1, BRD8 and PTCH1 mRNA levels in HeLa cells treated with cyclopamine for 24, 48 or 72 h. GAPDH was used as an internal housekeeping gene control. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. Scale bar: 10 μm.

Figure 5

DZIP1 knockdown and overexpression do not affect the accumulation or stability of mRNAs associated with DZIP1-containing complexes but modify the quantity of stress granules per cell. (A) Protein extracts from DZIP1 knockdown (siDZIP1) and negative control (siNC1) cells after 72 h of transfection were analyzed by western blotting with antibodies against DZIP1 and GAPDH. (B) Western blots shown in (A) were used for quantitation and are representative of results from two experiments. (C) Quantitative RT-PCR analysis of DZIP1, GLI1, PTCH1, BRD8, SNX2 and IFT80 mRNA levels in DZIP1 knockdown cells. (D) Quantitative RT-PCR analysis of DZIP1, GLI1, PTCH1, BRD8, SNX2 and IFT80 mRNA levels in HeLa cells transfected with pDZIP1-GFP or pGFP alone. GAPDH was used as an internal housekeeping gene control. (E) Mean half-life for some of the mRNA targets of DZIP1 (IFT80, SNX2, BRD8 and PTCH1) in HeLa cells transfected with siDZIP1 or siNC1. Triplicates were performed for qRT-PCR and the mean half-life was obtained from two independent experiments. (F) DZIP1 knockdown cells and control (NC1) were subjected to oxidative stress and the formation of stress granules was determined by TIA1 staining. The granules were counted and statistically analyzed. At least 5 fields per slide (technical triplicate) were counted. *P ≤ 0.05.

Figure 6

Polysomal distribution of DZIP1-GFP. (A) Lysates were prepared from HeLa cells treated with cycloheximide and were separated in a 10–50% sucrose gradient. Distribution of DZIP1-GFP was examined by western blotting. (B) HeLa cells were treated with puromycin, and extracts were prepared and processed as in (A). (C) Quantitation of the distribution of DZIP1-GFP after cycloheximide and puromycin treatment. Western blots shown in (A) and (B) were used for quantitation. (D) Protein extracts of Hela cells transfected with pDZIP1-GFP, pGFP alone (control sample), siDZIP1 or siNC1 were analyzed by western blotting with antibodies against SNX2, GAPDH1 and GFP. Quantitation by integrated density of SNX2, DZIP1-GFP (α-GFP ~120 kDa) or GAPDH1 is shown in the table above.

Figure 7

Model of ribonucleoprotein complexes containing DZIP1 in stress granules. DZIP1 is located mainly in the cytoplasm, where it is involved in ribonucleoprotein complexes related to mRNA networks involved primarily in controlling cell cycle and gene expression in HeLa cells. Inhibiting the Hedgehog pathway with cyclopamine blocks its translocation to the nucleus. DZIP1 is associated with polysomes and when HeLa cells are exposed to oxidative stress, DZIP1 is translocated to stress granules but not to processing bodies (p-bodies).