ZNF692 organizes a hub specialized in 40S ribosomal subunit maturation enhancing translation in rapidly proliferating cells

SUMMARY

Increased nucleolar size and activity correlate with aberrant ribosome biogenesis and enhanced translation in cancer cells. One of the first and rate-limiting steps in translation is the interaction of the 40S small ribosome subunit with mRNAs. Here, we report the identification of the zinc finger protein 692 (ZNF692), a MYC-induced nucleolar scaffold that coordinates the final steps in the biogenesis of the small ribosome subunit. ZNF692 forms a hub containing the exosome complex and ribosome biogenesis factors specialized in the final steps of 18S rRNA processing and 40S ribosome maturation in the granular component of the nucleolus. Highly proliferative cells are more reliant on ZNF692 than normal cells; thus, we conclude that effective production of small ribosome subunits is critical for translation efficiency in cancer cells.

Graphical Abstract

In brief

Lafita-Navarro et al. characterize the zinc finger protein ZNF692 as a MYC-induced nucleolar scaffold that enhances 18S rRNA processing to generate translation-efficient ribosomes in highly proliferative cells such as cancer cells.

INTRODUCTION

The nucleolus is the largest membrane-less organelle that is responsible for nearly every step in the production of ribosomes.¹ Dysregulation of nucleolar activity is linked to defects in ribosome biogenesis and can lead to a variety of human diseases including Alzheimer’s, Parkinson’s, and Huntington’s diseases, Diamond-Blackfan anemia (DBA), and Treacher Collins syndrome.^2,3 Conversely, increased nucleolar size and activity are associated with augmented ribosome production in cancer cells. Indeed, having more and larger nucleoli is the most distinguishable morphological alteration of a cancer cell and is frequently used by pathologists to grade solid tumors.⁴

The major steps in ribosome biogenesis are spatially organized in three distinct nucleolar compartments: the fibrillar center (FC), the dense fibrillar component (DFC), and the granular component (GC); all are named based on their morphological appearance under electron microscopy. Three out of four ribosomal RNAs (rRNAs) are generated as a single precursor (47S pre-rRNA) at the border of the FC and DFC by the activity of the RNA polymerase I (RNAPolI). The 47S pre-rRNA precursor undergoes maturation in a multi-step process that is initiated in the DFC and is completed in the GC, generating the 18S, 5.8S, and 28S mature rRNA.⁵ The 18S rRNA comprises the 40S small ribosomal subunits, while the 5.8S and 28S rRNAs, together with 5S rRNA, which is synthesized from a separate locus, comprise the 60S large ribosomal subunits.

The 90S pre-ribosome, also called the small subunit (SSU) processome, is a large multiprotein complex, which is co-transcriptionally assembled on pre-rRNA.⁶ This complex is responsible for coordinating the multiple cleavage steps that occur during maturation of the pre-rRNA. The first pre-rRNA cleavage step gives rise to two precursor complexes that later generate rRNAs that comprise mature small 40S and large 60S ribosomal subunits.^6,7 While this process is accomplished exclusively through endonucleolytic cleavages in yeast, it requires the combined action of both endonucleases and exonucleases in human cells.^6,8

The exosome complex, comprising nine structural subunits (EXOSC1–EXOSC9) organized in a ring structure and an exonuclease,⁹ harbors 3ʹ–5ʹ exoribonucleolytic activity that can degrade or process RNA in the nucleolus and in the cytoplasm.¹⁰ Nucleolar exosomes, which are associated with the 3ʹ–5ʹ exonuclease EXOSC10, are required for the maturation steps that generate 18S and 5.8S rRNA in human cells.^11,12 The near mature 18S-E pre-RNA, in complex with ribosomal proteins, is exported to the cytoplasm where it matures into 18S rRNA within the 40S small ribosome subunit.⁸ This subunit can then interact with mRNA and translation initiation factors forming the 43S preinitiation complex (PIC) that recruits the 60S ribosomal subunit to form actively translating ribosomes.¹³

Numerous studies have shown that the transcription factor MYC promotes ribosome biogenesis^14,15 by driving rDNA transcription and the transcription of structural and regulatory components of the ribosome and by amplifying RNAPolIII-mediated 5S rRNA production.^16–18 Here, we describe the identification and characterization of ZNF692, a protein that evolved in chordates, as a MYC-induced nucleolar scaffold responsible for increasing the efficiency of small ribosomal subunit maturation. Our work shows that ZNF692 localizes in the GC of the nucleolus where it facilitates the formation of a complex that anchors the exosome complex to the SSU processome, promoting 21S pre-rRNA trimming and thus 18S maturation. We propose that by facilitating the final steps of 40S subunit biogenesis, MYC-induced ZNF692 increases translation efficiency to match protein synthesis with the overall increase in mRNAs that are transcribed by MYC in proliferative cells.

RESULTS

MYC promotes the transcription of the evolutionarily conserved nucleolar factor ZNF692

To dissect the regulatory network underlying cell growth, we performed RNA-seq comparing Rat1 myc−/− fibroblasts expressing empty vector (E.V.) or human MYC and searched for transcription factors with increased expression in MYC-expressing cells.^19,20 We identified three understudied zinc finger-containing genes whose expressions were upregulated by MYC (Figures S1A and S1B). Among these, Zfp692 (ZNF692 mouse orthologue) was the only zinc finger protein gene to also be induced by MYC in published datasets of MYC-driven liver tumors²¹ and lymphoma²² (Figures S1C and S1D). qRT-PCR and western blots (WBs) in Rat1 fibroblasts and in human retinal pigment epithelial cells (ARPE-19) confirmed Zfp692 and ZNF692 induction by MYC (Figures 1A–1D). Knocking down MYC in human MYC-reconstituted myc−/− fibroblasts decreased Zfp692 (Figures 1A and S1E). We identified MYC binding sites (E-box) on ZNF692 and Zfp692 gene promoters (Figure S1F). In agreement, ENCODE chromatin immunoprecipitation (ChIP)-seq data revealed that MYC bound to the ZNF692 and Zfp692 promoters (Figure S1F). ZNF692 was increased in colon cancer cell lines with elevated MYC (Figure S1G). Knocking MYC down in these cells reduced ZNF692 expression (Figures 1E and S1H). Thus, colon cancer cell lines were used to elucidate the cellular and molecular functions of ZNF692.

Figure 1. MYC promotes the transcription of ZNF692, an evolutionarily conserved nucleolar factor.

(A) Zfp692 mRNA expression in Rat1 myc−/− cells expressing empty vector or human MYC upon control siRNA (siCtrl) or MYC siRNA transfection. Biological replicates n = 3 with two technical replicates.

(B) WB for Zfp692 in Rat1 myc−/− cells expressing empty vector or human MYC.

(C) ZNF692 mRNA expression in ARPE cells expressing empty vector or MYC. Biological replicates n = 4.

(D) WB for ZNF692 in ARPE cells expressing empty vector or MYC.

(E) WB for ZNF692 in control siRNA (siCtrl) or MYC siRNA-transfected HCT116 cells.

(F) IF for ZNF692 in ARPE cells expressing empty vector or MYC.

(G) Predicted nucleolar localization sequence (NoLS) in the N terminus of ZNF692.

(H) Identification of a NoLS in ZNF692 using nucleolar localization sequence detector (NoD).

(I) IF of endogenous ZNF692 colocalizing with NPM1.

(J) IF of NPM1 in WT and ΔNoLS GFP-ZNF692-transfected cells.

(k) IF of RPA40 in WT and ΔNoLS GFP-ZNF692-transfected cells. Deletion of the NoLS prevented ZNF692 colocalization with NPM1 or RPA40. *p < 0.05, two-tailed unpaired Student’s t test statistical analysis. Graphs with mean ± SD. Scale bars represent 10 μm.

Immunofluorescence (IF) confirmed that MYC promoted ZNF692 expression, which was surprisingly concentrated in the nucleoli (Figures 1F and S1I) despite being previously proposed to function as an RNA polymerase II-related transcription factor.²³ Using nucleolar localization prediction tools,²⁴ we identified a nucleolar localization signal (NoLS) in the ZNF692 N-terminal (Nt) region (Figures 1G and 1H). Confirming its nucleolar presence, ZNF692 colocalized with nucleophosmin 1 (NPM1) (Figures 1I and S1I). ZNF692 antibody specificity was validated by IF upon ZNF692 knockdown (KD) (Figure S1J). GFP-ZNF692 also localized in the nucleolus (Figure 1J), while NoLS deletion re-localized GFP-ZNF692 into nucleoplasmic droplets that no longer colocalized with NPM1 (Figure 1J) and RPA40 (RNAPolI component) (Figure 1K).

Sequence homology analyses demonstrated that ZNF692 emerged in chordates, suggesting an evolved role of this phylum (Figure S1K). ZNF692 contains five conserved zinc fingers in the C terminus (Ct). The Nt (including the NoLS) was also conserved in all species, indicating that ZNF692 nucleolar localization is evolutionary conserved (Figure S1K in red). The ZNF692 central domain showed poor conservation among the analyzed species, where only a core of glutamic acid (E) amino acids was conserved (Figure S1K in red).

ZNF692 depletion reduces cell growth

Nucleolar function is highly regulated by various cellular stimuli. FBS stimulation, which promotes cell proliferation, caused an increase in MYC and ZNF692 including ectopically expressed ZNF692 in colon cancer cells (Figures 2A and 2B). While the expression of ZNF692 was induced by ectopic MYC in ARPE cells, both ZNF692 and MYC were not affected by FBS stimulation (Figure S2A). These results suggest that ZNF692 levels may be controlled by complementary mechanisms during growth stimulation that include transcription by MYC and increase mRNA/protein stability. ZNF692 increase correlated with cyclin D1 (Figure 2A), which occurs through G1-S phase when cells are growing and requiring more ribosomes for protein synthesis. Using siRNA for ZNF692, we demonstrated that ZNF692 KD decreased the proliferation of colon cancer cells (Figures 2C and 2D) supporting previous data in lung, cervical, and colon cancer cells.^23,25,26 ZNF692 KD in MYC-expressing ARPE (ARPE-MYC) cells reduced their proliferation, while more modest effects were observed in ARPE cells expressing E.V. (Figures 2E and 2F), suggesting an increased demand for ZNF692 in MYC-transformed cells. Prolonged ZNF692 KD reduced proliferation also in ARPE (Figures S2B and S2C). Surprisingly, stable shRNA-mediated KD of ZNF692 or stable overexpression of ZNF692 did not affect cell proliferation (Figures 2G and 2H) or size (Figures S2D–S2F), suggesting an adaptation to the changes in ZNF692 expression in long-term cultures. Like stable KD of ZNF692, CRISPR knockout (KO) cells did not display a reduction in growth in vitro (Figure 2I). However, growth of these cells xenotransplanted in mice (Figure 2L) led to the formation of smaller tumors than control cells (Figures 2L, 2M, and S2I–S2K), demonstrating the importance of ZNF692 for tumor growth in vivo.

Figure 2. Knockdown of ZNF692 reduced the viability of proliferative cells without affecting the expression of cell-cycle regulators.

(A) WB of HCT116 expressing empty vector (E.V.) or ZNF692 4, 6, 24 or 48 h after FBS stimulation.

(B) Heatmap showing quantification of (A) related to tubulin.

(C) Relative proliferation of HCT116 cells 3 days after transfection with control or ZNF692 siRNAs. Biological replicates n = 4.

(D) WB of HCT116 cells transfected with control or ZNF692 siRNA.

(E) Relative proliferation of ARPE cells expressing empty vector or MYC 3 days after transfection with control or ZNF692 siRNA. Biological replicates n = 3.

(F) WB of ARPE cells expressing empty vector (E.V.) or MYC 3 days after transfection with control or ZNF692 siRNA.

(G) Relative proliferation of HCT116 cells stably expressing ZNF692 or shRNA for ZNF692. Biological replicates n = 3.(H) WB of HCT116 cells stably expressing ZNF692 or shRNA for ZNF692.

(I) Relative proliferation of DLD1 cells CRISPR KO for ZNF692 or control. Biological replicates n = 4.

(J) WB of DLD1 cells CRISPR KO for ZNF692 or control.

(K) Schematic representation of xenograft experiment.

(L) Representative pictures of tumors collected at endpoint of the xenografts experiment. (See also Figure S3H.)

(M) Tumor weight from xenograft experiment. SgCtrl n = 9, sgZNF692–3 n = 8.

(N) Previously reported ZNF692 target genes in our RNA-seq. Biological replicates n = 3.

(O) ZNF692 expressing levels in normal and tumor tissues of TCGA COAD samples.

(P) Patient survival correlation with 25% highest and lowest ZNF692 mRNA levels of TCGA COAD samples.

(Q) ZNF692 mRNA expression in tumor vs normal tissues from patients of different TCGA tumor types. *p < 0.05, two-tailed unpaired Student’s t test statistical analysis. Graphs with mean ± SD.

Previous studies suggested that ZNF692 regulates p27 and cyclin D1 levels to promote cell growth.^23,25 However, our RNA-seq comparing control and ZNF692 KO DLD1 cells (Figures S2G and S2H; Table S2) did not identify growth-promoting pathways as transcriptionally regulated by ZNF692. Importantly, PCK1, CDKN1A, CCND1, or CCNA1 were not affected by ZNF692 KO (Figure 2N; Table S2). WB for p27 and cyclin D1 upon ZNF692 up- or downregulation confirmed that ZNF692 does not regulate p27 or cyclin D1 (Figures 2D, 2F, 2H, 2J, and S2L). Furthermore, KD, KO, or overexpression of ZNF692 had no effect on other growth regulators such as p21, p53, and cyclin A1 (Figures 2D, 2F, 2H, 2J, and S2L; Table S1). ZNF692 expression was dramatically elevated in colorectal adenocarcinoma (COAD) and other tumor types when compared to normal tissues deposited in The Cancer Genome Atlas (TCGA), and elevated ZNF692 mRNA correlated with shorter survival of patients with COAD (Figures 2O–2Q).

ZNF692 regulates nucleolar morphology and protein synthesis

Consistently with ZNF692 nucleolar localization, tumors generated by ZNF692 KO cells contained smaller and rounder nucleoli (Figures 3A, 3B, S3A, and S3B), indicating reduced nucleolar activity and thus ribosome biogenesis. In non-perturbed cultures, cells expressing elevated levels of ZNF692 (top 25%) had large nucleoli, while cells expressing low levels of ZNF692 (bottom 25%) had smaller nucleoli (Figures 3C and S3C). Cells expressing low levels of ZNF692 displayed rounder nucleoli (Figures 3C and S3C). Electron microscopy also demonstrated that acute ZNF692 KD caused a reduction in nucleolar perimeter (Figures 3D–3E and S3D) as observed for xenografted cells.

Figure 3. ZNF692 regulates nucleolar morphology and protein synthesis.

(A) H&E staining from ZNF692 KO or control DLD1 xenograft tumors. White dotted line outlines the nucleolus. (See also Figure S3G.) Scale bars represent 10 μm.

(B) Quantification of nucleolar area and circularity (more circular the closer to 1) of ZNF692 KO or control DLD1 xenograft tumors. (See also Figure S3H.) Number of nucleoli sgCtrl n = 108, sgZNF692–1 n = 124, sgZNF692–3 n = 212.

(C) Correlation between levels of ZNF692 measured by IF with nucleolar size and circularity (more circular the closer to 1) in HCT116 cells. (See also Figure S3C.) Number of nucleoli top 25% n = 220, bottom 25% n = 218. Representation of one biological replicate out of three with similar results.

(D) Electron microscopy of two examples of HCT116 cells after 3 days of infection with empty vector (PLKO) or shRNA for ZNF692 (#2 and #5) containing virus (See also Figure S2B.) Scale bars represent 2 μm.

(E) Quantification of nucleolar area and circularity (more circular the closer to 1) from cells in C. ZNF692 KD reduces the perimeter. Number of nucleoli pLKO n = 25, shZNF692–2 n = 34, shZNF692–5 n = 30.

(F) Schematic representation of the puromycylation-based assays to measure protein synthesis.

(G) Puromycylation of control or ZNF692 siRNA-transfected DLD1 cells.

(H) Puromycylation of control or ZNF692 shRNA transiently infected HCT116 cells.

(I) Puromycylation of control or ZNF692 shRNA stably infected HCT116 cells. Three replicates are shown.

(J) Puromycylation of WT or ZNF692 KO (sg1 and sg3) DLD1 cells.

(K) Puromycylation of HCT116 stably expressing ZNF692. Three replicates are shown.(L) Schematic representation of the harringtonine and puromycylation-based assay to measure translation elongation.

(M) Inhibition of translation initiation with harringtonine followed by puromycylation chase in DLD1 ZNF692 KO or control. *p < 0.05, two-tailed unpaired Student’s t test statistical analysis. Graphs with mean ± SD.

The nucleolar localization of ZNF692 and its correlation with nucleolar size and morphology prompted us to determine whether ZNF692 plays a role in ribosome biogenesis and/or protein synthesis. Using puromycylation (Figure 3F),²⁷ we demonstrated that ZNF692 KD by siRNA (Figures 3G and S3E) or shRNA either transiently (Figure 3H) or stably (Figures 3I and S3F) decreased protein synthesis. ZNF692 KO also reduced protein synthesis (Figure 3J). Conversely, ectopic expression of ZNF692 promoted protein synthesis (Figure 3K). Treating control or ZNF692 KO cells with harringtonine (translation initiation inhibitor) prior to puromycylation (Figure 3L) confirmed that translation elongation was reduced in ZNF692 KO cells (Figure 3M). ZNF692 up- or downregulation did not affect the expression of nucleolar regulators such as UBF, RPA40, FBL, and NPM1 (Figures S3G–S3J).

ZNF692 resides in the granular component of the nucleolus and interacts with ribosome biogenesis factors

ZNF692 was proposed to function as a transcription factor due to the presence of zinc fingers in its Ct (Figures 1G and S1K). However, our RNA-seq found that the expression of some ribosome factors was increased but not decreased upon ZNF692 KO, indicating that the nucleolar alterations observed upon ZNF692 KO are not due to transcriptional changes. Moreover, ChIP experiments confirmed that ZNF692 does not bind rDNA. We measured ZNF692 binding to rDNA regions near the transcription start site (H1) and the transcription end site (H13). rDNA intergenic region (H32) where transcription machinery does not bind was used as a negative control.¹⁶ While RNAPolI efficiently bound to rDNA at regions H1 and H13, but not to the H32 region as expected, ZNF692 did not bind to rDNA (Figure S4A). Neither RNAPolI nor ZNF692 bound to LDHA promoter (negative control) (Figure S4A). Immunoprecipitation (IP) and WB demonstrated that ZNF692 antibody efficiently immunoprecipitated ZNF692 in formaldehyde-fixed lysates (Figure S4B). In agreement, pre-rRNA levels as measured by qRT-PCR (Figures 4A and 4B), anti-rRNA IF (Figure 4C), or rRNA 5ʹ external transcribed spacer (5ʹ-ETS) RNA-FISH (Figure 4D) were not altered upon ZNF692 depletion. qPCR of newly transcribed pre-rRNA of samples labeled with 4sU followed by biotinylation and streptavidin pull-down in wild-type (WT) or ZNF692 KO cells demonstrated that ZNF692 does not regulate pre-rRNA synthesis (Figure S4C). Additionally, we evaluated rDNA promoter activity in control and ZNF692 KD cells using a reporter that contains rDNA regulatory region cloned upstream of the luciferase gene. This experiment found that ZNF692 KD does not decrease rDNA promoter activity (Figure S4D). Collectively, these results show that ZNF692 does not regulate rDNA transcription.

Figure 4. ZNF692 resides in the granular component of the nucleolus where it interacts with ribosomal proteins.

(A) Pre-rRNA levels of control or ZNF692 shRNA stably infected HCT116 cells. pLKO n = 3, shZNF692–2 n = 1, shZNF692–5 n = 3, with two to three technical replicates each. Graphs with mean ± SD.

(B) Pre-rRNA in ZNF692 KO or control DLD1 xenograft tumors from Figures S2J–S2L. SgCtrl n = 6, sgZNF692–1 n = 5, sgZNF692–3 n = 6. Graphs with mean ± SD and p value.

(C) rRNA and ZNF692 IF showing equal rRNA levels in control or ZNF692 shRNA stably infected HCT116 cells. Scale bars represent 10 μm.

(D) RNA-FISH for 5ʹ-ETS (targeting 47S pre-rRNA) in ZNF692 KO or control DLD1 cells. Scale bars represent 10 μm.

(E) GFP, GFP-tagged WT ZNF692, and ΔNt constructs and their localization when expressed in HCT116 cells. Scale bars represent 5 μm.

(F) GFP WB of anti-GFP nanotrap bead immunoprecipitants in HCT116 cells transfected with GFP, GFP-ZNF692, or GFP-ZNF692ΔNt. Immunoprecipitants were analyzed by mass spectrometry to identify ZNF692’s interactome.

(G) Gene Ontology of ZNF692 partners (see also Figures S4E–S4F).

(H) Representation of a nucleolus and its contained subcompartments.

(I) Representation of the expected localization of RPA40 in the fibrillary centers (FCs), fibrillarin (FBL) in the dense fibrillar component (DFC), and NPM1 in the granular component (GC) of the nucleolus.

(J) IF of ZNF692-transfected DLD1 cells showing the colocalization of ZNF692 with NPM1 in the GC through double staining of ZNF692 (red) with RPA40, FBL, or NPM1 (green). Scale bars represent 5 μm.

(k) Structured illumination microscopy (SIM) in ZNF692-transfected DLD1 cells showing that ZNF692 is colocalized with NPM1 but not RPA40 and FBL. (See also Figure S4I.) Scale bars represent 2 μm.

To investigate the molecular functions of ZNF692 in the nucleolus, we characterized its interactome. Lysates of cells transfected with GFP, GFP-ZNF692 WT, or GFP-ZNF692 ΔNt (Figure 4E) were immunoprecipitated with GFP nanotrap beads, and the interacting proteins were subjected to mass spectrometry (Figure 4F). To identify nucleolar-specific ZNF692 interactors, only proteins that lost their binding to ZNF692 when the NoLS was deleted were considered. This approach identified 334 proteins as ZNF692 interactors with most of these (275) dependent on the NoLS. Consistent with the nucleolar localization of ZNF692, Gene Ontology analysis determined that the majority of ZNF692 interactors are ribosomal proteins and rRNA processing factors (Figures 4G and S4E–S4H; Table S3).

By performing co-IF of ZNF692 and RPA40 (a marker of the FC), fibrillarin (FBL, a marker of the DFC), or NPM1 (a marker of the GC) (Figures 4I and 4J), we found that ZNF692 was excluded from the FC and the DFC, and it predominantly colocalized with NPM1 in the GC (Figure 4J). Structure illumination microscopy (SIM) found that ZNF692 was consistently separated from RPA40 and FBL while mostly colocalized with NPM1 (Figures 4K and S4I), confirming that the GC, where most of the final steps in ribosome assembly occur, is the primary residence of ZNF692. Overexpression of ZNF692 did not alter the nucleoli structure (Figure S4J). The absence of ZNF692 from the FC is in agreement with its inability to bind rDNA and to regulate its transcription. The localization of ZNF692 in the GC of the nucleolus and its interaction with ribosome biogenesis factors suggested that ZNF692 is involved in ribosome maturation (Figure 4H).

ZNF692 interacts with components of the small subunit processome and the exosome complex

ZNF692 interacts with 16 regulators of rRNA processing and ribosome assembly (Figure 5A; Table S4). Eleven of these interactors are involved in 18S rRNA processing and are components of the SSU processome in yeast,²⁸ including the 18S rRNA processing factors KRR1, the KRR1 interactor RCL1, IMP3, and DIMT1. Using IP, we confirmed that ZNF692 was bound to NOP2, KRR1, EXOSC7, and EXOSC8 (Figure 5C) but not to NHP2, BUD23, or DIMT1 (Figure S5A). Furthermore, we confirmed that ZNF692 interacted with additional nucleolar proteins identified as ZNF692 interactors in databases such as Biogrid and Bioplex (Figure S5B; Table S5), thus suggesting that the nucleolar interactome of ZNF692 goes beyond factors identified in this study.

Figure 5. ZNF692 interacts with the components of the exosome complex.

(A) ZNF692 interactors and their function in ribosomal subunit assembly. (See also Table S4.).

(B) ZNF692 recombinant proteins purified from insect cells to be used for in vitro assays.

(C) IP of recombinant ZNF692 using GFP nanotrap beads and DLD1 cells nuclear extracts followed by WB for NOP2, KRR1, EXOSC7, and EXOSC8.

(D)IF of NOP2, KRR1, EXOSC7, or EXOSC8 with NPM1 showing their co-colocalization in the granular component of the nucleolus. Scale bars represent 10 μm.

(E) IP of EXOSC7, EXOSC8 NOP2, KRR1, and ZNF692 from HCT116 cells stably overexpressing ZNF692 followed by WB for ZNF692, NOP2, KRR1, EXOSC7, and EXOSC8. Left panel: control; right panel: lysates were treated with 20 μg/mL RNase A prior to IP.

(F) RNA-IP of ZNF692 in DLD1 cells on rRNA; ZNF692 binds rRNA. n = 3 (two technical replicates each). *p < 0.05, two-tailed unpaired Student’s t test statistical analysis. Graphs with mean ± SD and p value.

(G) Puromycylation in control or NOP2, KRR1, EXOSC7, or EXOSC8 shRNA transiently infected DLD1 cells 3 days after infection.

(H) WB for NOP2, KRR1, EXOSC7, and EXOSC8 in cell myc−/− fibroblasts expressing empty vector (E.V.) or human MYC.

(I) WB for MYC, NOP2, KRR1, EXOSC7, and EXOSC8 in control or MYC siRNA DLD1 cells.

(J) Heatmap of ZNF692, KRR1, EXOSC7, EXOSC8, and NOP2 expression in tumor vs normal tissues of TCGA COAD samples.

Deletion of the Nt domain reduced ZNF692 binding to NOP2, KRR1, EXOSC7, and EXOSC8, while disruption of the zinc fingers abolished it. In agreement, colocalization with NPM1 determined that NOP2, KRR1, and EXOSC7 localized in the GC where ZNF692 resides (Figure 5D). EXOSC8 was present in the cytoplasm and nucleus (Figure 5D). NHP2 did not localize in the GC (Figure S5C), thus validating its lack of co-IP with ZNF692. Reciprocal co-IPs for endogenous EXOSC7, EXOSC8, NOP2, and KRR1 with ZNF692 in ZNF692-overexpressing cells confirmed that ZNF692 interacts with EXOSC7, EXOSC8, NOP2, and KRR1 (Figure 5E). EXOSC7, EXOSC8, NOP2, and KRR1 also interacted with each other (Figure 5E), suggesting the presence of a complex containing the exosome and the SSU.

Given that deletion of the zinc fingers reduced the binding of ZNF692 to its interactors (Figure 5C), we asked whether the presence of rRNA was required for ZNF692 to bind rRNA processing factors. The interaction between ZNF692, EXOSC7, EXOSC8, and NOP2 depended on the presence of RNA because treating the lysates with RNase A prevented their interaction (Figure 5E). Conversely, the interaction between ZNF692 and KRR1 remained unaffected even when RNA was degraded (Figure 5E). RNA-IP with an anti-ZNF692 using UV-crosslinked cells overexpressing ZNF692 demonstrated that ZNF692 interacted with rRNAs containing the 18S and 28S regions (Figure 5F). NPM1, which is known to bind rRNA, was used as a positive control (Figure S5D). Furthermore, inhibition of rDNA transcription with 0.05 μg/mL actinomycin D led to ZNF692 re-localization to the nucleoplasm (Figures S5E and S5F), suggesting that the presence of rRNA contributes to the retention of ZNF692 in the nucleolus.

Acutely knocking down EXOSC7, EXOSC8, NOP2, and KRR1 led to a reduction in protein synthesis (Figure 5G). However, stable KD had variable effects on cell proliferation, indicating that reduction in protein synthesis caused by downregulation of these genes is not a consequence of cell death (Figure S5G). KRR1, EXOSC7, and EXOSC8 were upregulated in MYC-expressing fibroblasts and downregulated upon MYC KD in colon cancer cells (Figures 5H and 5I). These results suggest that the components of the complex containing ZNF692 and ribosome maturation factors are co-upregulated by MYC. Moreover, the expression of KRR1, EXOSC7, EXOSC8, and NOP2 was also up-regulated in TCGA colon cancer samples (Figure 5J). ZNF692 was the most upregulated gene in this group when compared to the basal levels of each gene in normal tissues. Interestingly, while high ZNF692 mRNA levels correlated with poor survival of patients with colon cancer (COAD) (Figure 2P), the expression of KRR1, EXOSC7, EXOSC8, and NOP2 was not (Figure S5H), indicating that ZNF692 is the nucleolar adaptor that enhances ribosome biogenesis specifically in hyperproliferative cells.

ZNF692 regulates 18S maturation

Given that ZNF692 interacted with rRNA processing factors, we investigated the importance of ZNF692 for rRNA production. Northern blots of control or ZNF692 KD cells using the rRNA probes 5ʹ-ITS1 (maps the internal transcribed spacers between 18S and 5.8S rRNA) and 5ʹ-ITS2 (maps the internal transcribed spacers between 5.8S and 28S rRNA) (Figure 6A)²⁹ showed that ZNF692 and EXOSC10 KD caused an accumulation of the 21S pre-rRNA, the precursors of 18S rRNA (Figures 6B–6E and S6A–S6G). In contrast, while EXOSC10 KD led to accumulation of 12S pre-rRNA as previously documented,³⁰ ZNF692 KD did not alter the rRNA processing steps mapped with the 5ʹ-ITS2 probes (Figures 6B–6C and S6A). Mature rRNA levels were not altered upon ZNF692 or EXOSC10 depletion (Figure S6H). These results indicate that ZNF692 enhances specific steps in the late maturation of the 18S rRNA (Figure 6F). The interaction (Figure 5E) of ZNF692 with the exosome complex and KRR1, a component of the SSU processome necessary for small ribosome biogenesis in yeast and stem cells,^31–33 and the accumulation of the 21S pre-rRNA upon ZNF692 KD indicated that ZNF692 promotes efficient 18S pre-rRNA processing (Figure 6F), defining a mechanism for the decreased protein synthesis upon ZNF692 KD or KO (Figures 3G–3M).

Figure 6. ZNF692 enhances the last steps of 18S rRNA processing.

(A) Representation of the pre-rRNA processing pathway showing the functions of ZNF692 interactors.

(B) Northern blot of ZNF692, EXOSC10, or control siRNA transfected ARPE-MYC cells.

(C) Quantification of 5ʹ-ITS1and 5ʹ-ITS2 by ratio analysis of multiple precursors (RAMP) for siZNF692 in B. N = 3.

(D) Northern blot of ZNF692 or control siRNA-transfected DLD1 cells.

(E) Quantification of 5ʹ-ITS1 by RAMP of D. N = 3.

(E) Schematic representation of the role of ZNF692 on 18S rRNA maturation.

siCtrl., siRNA control; siZ-3, siRNA for ZNF692 #3; siEX10, siRNA for EXOSC10. *p < 0.05, two-tailed unpaired Student’s t test statistical analysis. Graphs with mean ± SD and p value.

ZNF692 enhances small ribosomal subunit assembly and translation initiation

To map the molecular functions and localization of ZNF692 more accurately, we used the three-step Pre-ribosome Sequential Extraction (PSE) method³⁴ that allows for nucleolar sub-fractionation. This method produces three fractions (Figure 7A): the SN1, containing cytoplasmic and nuclear fractions; the SN2, containing the outer layer of the nucleolar GC; and the SN3 containing inner nucleolar GC, DFC, and FC. Performing PSE in DLD1 cells that were either WT or ZNF692 KO, and in HCT116 cells expressing E.V. or ZNF692 (Figures 7B, 7E, and S7C), we found that ZNF692 predominantly localized in the outer layer (SN2) of the GC. All ZNF692 interactors where localized in both inner and outer layers of the GC (Figures 7B, 7E, and S7C). Similarly, super-resolution microscopy showed the presence of ZNF692-enriched compartments in the GC of the nucleolus (Figures 7C, 7D, and S7B). To confirm that ZNF692 colocalized in the SN2 with unprocessed rRNA, we performed northern blot of SN1 and SN2 fractions in HCT116 overexpressing ZNF692. Our results showed that SN2 is enriched in late 18S rRNA precursors such as 30S, 26S, 21S, and 18S-E rRNA (Figure 7F), demonstrating that ZNF692 is in proximity with these precursors and that it likely takes part in the final steps of 18S rRNA maturation. ZNF692 colocalized with 5ʹ-ITS1 in the nucleolus (Figure S7D).

Figure 7. ZNF692 enhances small ribosomal subunit assembly and translation initiation.

(A) Representation of cellular/nucleolar fractionation using the three-step Pre-ribosome Sequential Extraction (PSE) protocol. Adapted from Nieto et al., 2021.³⁵

(B) WB of the cellular/nucleolar fractionation for ZNF692 and ZNF692 interactors in control or ZNF692 KO DLD1 cells.

(C and D) 3D reconstruction of SIM IF images (Figure S7A) for NPM1, ZNF692, and FBL in ZNF692 KO DLD1 cells stably overexpressing ZNF692 showing the surface (C) and the volume (D) of ZNF692, NPM1, and FBL in the nucleolus. (See also Figure S7B.).

(E) WB of the cellular/nucleolar fractionation for ZNF692 and EXOSC10 in E.V. or ZNF692-overexpressing HCT116 cells.

(F) Northern blot of SN1 and SN2 fractions of ZNF692-overexpressing HCT116 cells.

(G) Polysome profiling of control or ZNF692 KO DLD1 cells.

(H) RNA bioanalyzer analysis of 40S, 60S, and 80S cytoplasmic fractions collected from (G).

(I) Proteomics analysis heatmap showing proteins whose abundance changed in 40S, 60S, and 80S single ribosomes of control or ZNF692 KO DLD1 cells from (G).

(J) WB for ribosomal proteins of 40S, 60S, and 80S fractions (G) of control or ZNF692 KO DLD1 cells. Actin was used as loading control.

(K) Click-IT AHA experiments in DLD1 ZNF692 KO cells (−/+ ectopic ZNF692) or control cells.

Graphical abstract: ZNF692 and SSU processome components promotes 18S rRNA processing and small ribosomal subunit assembly. Model for the function of ZNF692 in driving small ribosome subunit maturation and translation initiation in highly proliferative cells. MYC drives the expression of ZNF692 as well as components of the SSU processome such as KRR1 and the exosome complex promoting small ribosomal subunit maturation.

We sought to characterize the importance of ZNF692 in the composition of cellular ribosomes. For that, we performed polysome profiling upon ZNF692 KO and confirmed a decrease in polysomes accompanied by an increase in 80S monosomes indicating reduced protein translation (Figure 7G). Disassociation of ribosomal subunits with EDTA treatment confirmed that absolute amounts of 40S and 60S did not change (Figure S7D). Proteomics of ribosomes subunits 40S (small), 60S (large), and 80S ribosomes (Figures 7H and 7I) showed that translation initiation factors abundance was altered in ZNF692 KO cells. For example, eIF2B3, eIF2S3, eIF3M, and eIF4G1 and ribosomal proteins RPS12, RPS14, RPS16, RPS17, and RPS18, which are enriched in 40S fractions, were reduced in the 40S ribosomes upon ZNF692 KO (Figure 7I; Table S6), supporting ZNF692’s role in 18S rRNA processing. RPL11, RPL27, RPL18, RPL35, RPL36, and RPLP2, which are enriched in 60S fractions, were decreased in the 60S ribosomes in ZNF692 KO cells (Figure 7I; Table S6), suggesting that ZNF692 might also affect large ribosome subunit maturation. RPS24 and RPS26 were enriched in 60S fractions, and these were decreased in ZNF692 KO cells (Figure S7F; Table S6). RPS27L and RPL29 were decreased in the 80S fractions when ZNF692 was knocked out (Figure 7I; Table S6). WB validated that RPS14, RPS16, RPS17, and RPS18 were decreased in ZNF692 KO 40S fractions in comparison with control (Figure 7J). Importantly, the expression of these proteins was not altered by ZNF692 KO (Figures S7G and S7H). RPL15 was only present on the 60S and 80S as expected, and its abundance in these fractions did not change in ZNF692 KO cells (Figure 7J).

Delayed maturation of the small ribosomal subunit may lead to defects in translation initiation and in overall protein synthesis. We propose that the reduction of eIF2B3, eIF2S3, eIF3M, and eIF4G1 and RPS proteins in the 40S fraction of ZNF692 KO cells (Figure 7I) suggests that the formation of the 43S PIC is more efficient in cells containing ZNF692. Cap-dependent translation initiation in eukaryotes involves the formation of the 43S PIC, which contains the 40S subunit, initiation factors including eIF1, eIF1A, eIF3, and eIF5, and the ternary complex eIF2-GTP-Met-tRNA.³⁶ After 43S PIC formation, eIF4F complex (containing eIF4E, eIF4A, and eIF4G) recruits mRNA, and the 40S scans it to find the initiation codon where the 60S subunit binds to start protein elongation.³⁷ Click-iT AHA (L-azidohomoalanine) experiments were used as an orthogonal approach to quantify translation in the absence of ZNF692 (Figure 7K). The absence of ZNF692 reduced AHA incorporation, and reconstitution of ZNF692 rescued it (Figures 7K and S7I). Altogether, our data suggest that ZNF692, by facilitating the maturation of the small ribosomal subunit, enhances translation initiation and thus protein synthesis in highly proliferative cells.

DISCUSSION

With the identification of zinc finger domains in its Ct, ZNF692 had initially been suggested to function as a potential RNAPol II-related transcription factor.^23,25 Nonetheless, our investigation has revealed a different role for ZNF692 as a nucleolar scaffold, thus demonstrating the necessity of experimental approaches in delineating protein functionalities and cautioning against overly relying on domain similarities for functional inference.

We discovered that ZNF692 is key to physically and functionally connect conserved multiprotein complexes necessary for ribosome biogenesis, increasing small ribosome subunit maturation efficiency and, hence, translation initiation. This mechanism aligns translation efficiency with the increased transcription promoted by MYC. While most nucleolar proteins are constitutively expressed, our data show that ZNF692 senses and responds to growth signals to fine-tune nucleolar activity with cell growth state, thus making ZNF692 an interesting entry point to better understand nucleolar activity in hyperproliferative mammalian cells.

Given the high demands for protein synthesis of highly proliferative cells, it is not surprising that ribosome biogenesis and tumorigenesis are intimately linked. MYC is related to tumor initiation and maintenance in part due to its capacity to boost ribosome biogenesis.¹⁴ MYC upregulates genes involved in all of the steps of ribosome biogenesis from rDNA transcription to rRNA maturation and genes encoding ribosomal proteins.^38,39 This ribosome biogenesis and tumorigenesis connection is also apparent in genetic disorders with tumor development predisposition. For example, DBA has been correlated with several ribosomal protein gene mutations. Patients with DBA have a significantly higher risk of developing cancers of various types, including acute myeloid leukemia, colon cancer, and osteosarcoma.^40,41 Mutations in ribosomal proteins such as RPS19, RPL11, RPL5, or RPL35a are linked to cancer development.⁴² Given the essential roles of ribosomal proteins and MYC, targeting these proteins for cancer treatment could have detrimental effects in normal cells and tissues. Our study demonstrates that ZNF692 is a nucleolar regulator highly induced by MYC to enhance protein synthesis. ZNF692 increases translation efficiency of highly proliferating cells, possibly becoming an Achilles’ heel in tumors.

Limitations of the study

While our study has unveiled a molecular mechanism contributing to small ribosome subunit biogenesis via ZNF692, cryo-EM techniques would be instrumental in determining the structural implications of ZNF692’s involvement within the SSU processome. To comprehensively characterize ZNF692’s role in cancer development, it would be important to generate and cross Zfp692 KO mouse models with cancer-prone mouse models. Interestingly, our observations indicate that while acute ZNF692 KD diminishes cancer cell proliferation, stable ZNF692 KO cells remain viable, suggesting potential compensatory mechanisms, such as heightened transcription of ribosome biogenesis or other zinc finger protein genes (as detailed in Tables S2 and S8), may sustain ribosome biogenesis in the prolonged absence of ZNF692. It’s worth noting that our discovery of ZNF692 interactors relied on immunoprecipitating overexpressed ZNF692, given the limited quality of commercially available ZNF692 antibodies for immunoprecipitation or immunoblotting endogenous ZNF692. Consequently, future experiments may require endogenously tagging ZNF692 or developing high-quality antibodies tailored for efficient immunoprecipitation of endogenous ZNF692 to confirm its interactions.

STAR☆METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact: Maralice Conacci-Sorrell (Maralice.ConacciSorrell@UTSouthwestern.edu).

Materials availability

All reagents generated in this study are available without restriction from the lead contact.

Data and code availability

• Proteomics and RNA-seq data have been deposited at MASSIVE and GEO databases and are publicly available as of the date of publication. Accession numbers are listed in the Key resources table.

• This paper does not report original code

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
ZNF692	Abcam	ab204595; RRID: AB_3068305
RPA40	Santa Cruz	sc-374443; RRID:AB_10991310
FIBRILLARIN	Cell Signaling	#2639; RRID:AB_2278087
FIBRILLARIN	Novus Biologicals	NBP2-26151; RRID: AB_3068306
NPM1	Santa Cruz	sc-47725; RRID:AB_628034
FLAG	Sigma	F3165; RRID:AB_259529
MYC	Abcam	ab32072; RRID:AB_731658
Puromycin	Sigma	MABE343; RRID:AB_2566826
α-Tubulin	Sigma	T6199; RRID:AB_477583
β-Actin	Cell Signaling	#4970; RRID:AB_2223172
NOP2	Proteintech	10448-1-AP; RRID:AB_2282772
NHP2	Proteintech	15128-1-AP; RRID:AB_2267290
EXOSC7	Proteintech	25292-1-AP; RRID:AB_2880011
EXOSC8	Proteintech	11979-1-AP; RRID:AB_2101981
KRR1	Sigma	HPA043433; RRID:AB_10960988
WBSCR22 (BUD23)	Thermo Fisher	PA5-62566; RRID:AB_2649539
DIMT1	Thermo Fisher	PA5-60170; RRID:AB_2640601
RPS14	Novus Biologicals	NBP2-22319; RRID: AB_3068307
RPS16	Novus Biologicals	NBP1-80025; RRID:AB_11002635
RPS17	Novus Biologicals	NBP2-93721; RRID: AB_3068308
RPS18	Novus Biologicals	NBP2-93632; RRID: AB_3068309
RPL15	ProteinTech	16740-1-AP; AB_2301193
Normal Rabbit IgG	Cell Signaling	#2729; RRID:AB_1031062
Chemicals, peptides, and recombinant proteins
Actinomycin D	Sigma	A9415
GFP-nanotrap	Chromotek	Gta-10
Harringtonine	Abcam	ab141941
RNAse OUT	Thermo Fisher	10777-019
RNasin® Ribonuclease Inhibitor	Promega	N2115
RNase A	Sigma	10109169001
DNase I	Sigma	4716728001
Click-IT™ AHA (L-Azidohomoalanine)	Thermo fisher	C10102
Biotin Alkyne (PEG4 carboxamide-Propargyl Biotin)	Thermo fisher	B10185
Click-iT™ Protein Reaction Buffer Kit	Thermo fisher	C10276
4-Thiouridine (4sU)	Sigma	T4509
EZ-link HPDP Biotin	Thermo fisher	PIA35390
Streptavidin Magnetic Beads	Pierce	88816
HRP-Conjugated Streptavidin	Thermo Scientific	N100
SSC (20X), RNase-free	Invitrogen	AM9763
UltraPure Ethidium Bromide	Invitrogen	15585011
Amersham Hybond-N+ membrane	GE Healthcare	RPN303B
Thermo Scientific™ SuperSignal™ West Dura Extended Duration Substrate	Thermo Scientific	PI34076
Critical commercial assays
NorthernMax Kit	Invitrogen	AM1940
Deposited data
Proteomics	This study	Massive: MSV000089271
RNA-seq	This study	GEO: GSE215838
Experimental models: Cell lines
Rat1 fibroblasts	Dr. John Sedivy (Brown University)	N/A
ARPE-19	Sandra Schmid lab	N/A
HCT116	ATCC	CCL-247
DLD1	ATCC	CCL-221
RKO	ATCC	CRL-2577
ARPE EV (pBabe-puro-HA)	This study	N/A
ARPE MYC (pBabe-puro-HA-MYC)	This study	N/A
HCT116 EV (pCMV-entry)	This study	N/A
HCT116 ZNF692 (pCMV-entry-ZNF692)	This study	N/A
HCT116 shControl (pLKO)	This study	N/A
HCT116 shZNF692-2	This study	N/A
HCT116 shZNF692-5	This study	N/A
DLD1 sgControl	This study	N/A
DLD1 sgZNF692-1	This study	N/A
DLD1 sgZNF692-3	This study	N/A
Experimental models: Organisms/strains
NOD/SCID mice	Jackson lab	001303
Oligonucleotides
universal siRNA controls #1	Sigma mission	SIC001
siRNA ZNF692-1	Sigma mission	SASI_Hs01_00035045
siRNA ZNF692-2	Sigma mission	SASI_Hs01_00035047
siRNA ZNF692-3	Sigma mission	SASI_Hs02_00350781
siRNA MYC-1	Sigma mission	SASI_Hs01_00222676
siRNA MYC-2	Sigma mission	SASI_Hs01_00222677
Primer and Northern probes	This study	Table S7
Recombinant DNA
pCMV-entry-ZNF692-Myc-DDK	Origene	RC200163
pIRES-puro-GFP-ZNF692	This study	N/A
pOCC29-TEV-GFP-ZNF692 FL	This study	JWV59
pOCC29-TEV-GFP-ZNF692-ΔNt	This study	H079
pOCC29-TEV-GFP-ZNF692-ΔZNF2-4	This study	H080
pOCC29-TEV-GFP-ZNF692-ΔNt-ΔZNF2-4	This study	H081
pLKO (control vector)	Sigma mission	SHC002
ZNF692 shRNA-2	Sigma mission	TRCN0000229714
ZNF692 shRNA-5	Sigma mission	TRCN0000219054
EXOSC7 shRNA-1	Sigma mission	TRCN0000051070
EXOSC7 shRNA-2	Sigma mission	TRCN0000051071
EXOSC8 shRNA-1	Sigma mission	TRCN0000303393
EXOSC8 shRNA-2	Sigma mission	TRCN0000051639
KRR1 shRNA-1	Sigma mission	TRCN0000072294
KRR1 shRNA-2	Sigma mission	TRCN0000072296
NOP2 shRNA-1	Sigma mission	TRCN0000157484
NOP2 shRNA-2	Sigma mission	TRCN0000154296
Scrambled sgRNA CRISPR/Cas9 All-in-One Lentivector	Applied Biological Materials	K010
ZNF692 sgRNA CRISPR/Cas9 All-in-One Lentivector (Human) (Target 1)	Applied Biological Materials	K2725906
ZNF692 sgRNA CRISPR/Cas9 All-in-One Lentivector (Human) (Target 3)	Applied Biological Materials	K2725908
Software and algorithms
FIJI	ImageJ software	N/A
Prism 9	Graphpad	N/A
Nucleolar localization sequence detector	Geoff Barton lab	N/A
GEPIA	http://gepia.cancer-pku.cn/about.html	Tang, Z. et al., 201743
PrDOS: Protein DisOrder prediction System	https://prdos.hgc.jp/cgi-bin/top.cgi	Ishida T, Kinoshita K., 200744
PyMOL	Schrodinger	N/A
Other
RNA-seq TCGA	The Cancer Genome Atlas Program	N/A
RNA-seq Rat1 fibroblasts MYC OE	Maralice Conacci-Sorrell lab	Lafita-Navarro et al., 201819
RNA-seq EμMyc mice	Bruno Amati lab	Sabo et al., 201422
RNA-seq tet-off liver tumor mouse model	Bruno Amati lab	Kress et al., 201621
ChemiDoc Imaging System	BIO-RAD	17001401
CFX96 Touch Real-Time PCR Detection System	BIO-RAD	1855195
BioLogic LP low-pressure chromatography system	BIO-RAD	7318300
Zeiss LSM780 Inverted confocal microscope	Zeiss	N/A
OMX SR Super-resolution Microscope	DeltaVision	N/A
Bioruptor Standard Sonication	Diagenode	UCD-200

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Cell culture

Cells (Key resources table) were cultured in DMEM with 5–10% fetal bovine serum (FBS) and 100 U/mL Penicillin-Streptomycin.

Mice

Female NOD/SCID mice (Jackson lab). All procedures were approved by the Institutional Animal Care and Use Committee of the University of Texas Southwestern Medical Center (approval number: 2017–101798).

METHOD DETAILS

Cell proliferation

Cells were seed at equal amounts in dividing states. Cells were washed with PBS, fixed with methanol for 10 min, and then stained with crystal violet solution (1% acetic acid, 1% methanol, 1% (w:v) crystal violet dye) for 10 min. After washing with water, the plates were scanned, and dye intensity was quantify using FIJI. Results were normalized to control condition.

Plasmids

Human ZNF692 cDNA was amplified from pCMV6-entry-ZNF692 plasmid by PCR and cloned to pIRES-puro-GFP vector with restriction enzymes FseI and AscI. For in vitro purified protein, ZNF692 cDNA was synthesized and cloned to pOCC29-TEV-GFP vectors with restriction enzymes NotI and AscI. To make the ZNF692 mutant plasmids, corresponding cDNA sequence were amplified from full-length plasmids and cloned using same restriction enzymes.

Transient transfections

siRNA was reverse transfected with Lipofectamine RNAiMAX (Invitrogen). 1×10⁵ cells/well were seeded in 12-well plates and transfected with 2 μL siRNA (20 μM) and 2 μL Lipofectamine previously mixed in OPTI-MEM for 15 min. ZNF692 overexpression was performed with Lipofectamine 3000 (Invitrogen). One day before transfection, 1×10⁶ cells/well were seeded in 6-well plates. The next day, 2.5 μg DNA plasmid and 5 μL Lipofectamine were mixed and incubated for 15 min in OPTI-MEM, then added to cells. Cells were collected after 24–72h post-transfection. siRNAs/plasmids used are listed in Key resources table.

Viral transduction

Cells were seeded and mixed with 10 μg/mL polybrene and lentiviral vectors (Key resources table) made in 293T cells. For transient experiments, cells were harvested for WB analysis or to assess proliferation at the indicated times. To generate stable cell lines, 48 h after lentiviral particles addition, media was replaced and puromycin added for selection. Puromycin was used at 10 μg/mL for DLD1 and 5 μg/mL for HCT116 cells. Plasmids are listed in Key resources table. KD or KO of the specific genes was confirmed by WB.

RNA-seq

Method was run as described.⁴⁵ Briefly, the qualities of sequencing reads were evaluated using NGS QC Toolkit (v2.3.3) and aligned to human reference genome (hg38) using STAR (v2.5.2b). SAMtools (v1.9) was used to sort the alignments, and HTSeq Python package was used to count reads per gene. DESeq2 R Bioconductor package was used to normalize read counts and identify differentially expressed (DE) genes. Gene ontology was performed with GSEA-MSigDB using ‘‘GO biological processes’’ setting. RNA-seq data has been deposited in GEO: GSE215838.

RT-qPCR and rDNA quantification

RNA was extracted with Tri-Reagent (Sigma) and reverse transcribed to cDNA with High-Capacity cDNA Reverse Transcription Kit (Life Technologies). RNA levels were measured by qPCR with iTaq Universal SYBR Green Supermix (BIO-RAD) and the BioRad CFX96 device. For qPCR analysis, 2^−ΔΔCt method was used, and Actin was used as a housekeeping gene. qPCR primers are listed in Key resources table.

To measure newly synthesized rRNA, nascent RNA was tagged with 10 μM 4-Thiouridine (4sU) for 3 h in control or ZNF692 KO DLD1 cells and total RNA was extracted with Tri-Reagent (Sigma). 4sU tagged RNA was then labeled with EZ-link biotin-HPDP (2 μg/μg RNA) and subsequently pulled down with Streptavidin Magnetic Beads (Pierce). RNA levels were measured by qPCR as above.

For rDNA promoter activity quantification, HCT116 cells stably expressing ZNF692 shRNA or control were transfected with rDNA promoter internal ribosome entry site (IRES)-luciferase and Renilla plasmids as previously described.⁴⁶ The ratio of pHrD-IRES-luciferase/Renilla activity was calculated to control for transfection efficiency.

Western Blot

Total protein was extracted with lysis buffer containing 50 mM Tris-HCl (pH 7.7), 150 mM NaCl, 0.5% Nonidet P-40 or RIPA buffer in the presence of protease and phosphatase inhibitors. Proteins were separated by SDS–polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes, and probed with specific antibodies (Key resources table). Molecular marker labels were added to the left side of each Western blot panel.

Puromycylation

Puromycin was added at 20 μg/mL for DLD1 and 10 μg/mL for HCT116. Cells were serum starved overnight, complete media (containing FBS) was added the next day for 6 h, followed by puromycin for 2 h. For Figure 5G, cells were seeded and fresh media with puromycin was added for 20 min on the next day. After puromycin incubation, cells were lysed with RIPA buffer for WB using anti-puromycin antibody (Key resources table). For harringtonine experiments, cells were seeded in 6-well plates, and harringtonine was added for 1, 2, and 8 min the next day. Following incubation with harringtonine, puromycin was added for 20 min, and then cells were processed for WB.

Immunofluorescence staining and microscopy imaging

Cells grown on glass coverslips were fixed with 4% paraformaldehyde in PBS for 15 min, permeabilized with 0.1% Triton X-100 for 20 min and blocked with 5% BSA in PBS for 30 min to 1 h. Primary antibodies (Key resources table) in 5% BSA-PBS were added and incubated overnight at 4°C. Cells were washed with PBS and incubated with secondary antibodies in 5% BSA-PBS for 1 h at room temperature (RT). Then, coverslips were washed with PBS and mounted with Mowiol mounting media. DAPI was used to stain the nuclei. Images were acquired with a Zeiss LSM780 or Nikon CSU-W1 SoRa fluorescent microscope. For SIM images (Figure 4K), DeltaVision OMX SR Super-resolution Microscope was used (GE Healthcare). 3D reconstructions from OMX SR Super-resolution microscopy were obtained with Imaris software. Images were processed with FIJI. For Figures 3C and S3C, to quantify ZNF692 nucleolar intensity as well as nucleolar circularity and perimeter, a mask using ZNF692 staining was created and then each particle was measured using the analyze particle setting.

In silico analyses

Nucleolar Localization Sequence Detector, and Protein DisOrder prediction System were used to predict nucleolar localization sequences, and disordered regions of human ZNF692. For ribosome reconstructions, the human 80S ribosome structure (PDB: 4V6X) was obtained from Protein DataBank (https://www.rcsb.org/) and ribosomal proteins interacting with ZNF692 were highlighted with PyMOL software. RNA expression and patient survival analysis from The Cancer Genome Atlas (TCGA) RNA-seq experiments was performed with the GEPIA web server. ZNF692 and Zfp692 promoter sequences were scanned for canonical (CACGTG) and non-canonical E-boxes (CATGTG; CATGCG; CACGCG; CACGAG; CAACGTG; CACATG). ChIP-seq data was obtained from ENCODE database and visualized in UCSC genome browser.

Electron microscopy

Three days after infection with lentiviral particles, HCT116 cells were fixed on MatTek dishes with 2.5% (v/v) glutaraldehyde in 0.1 M sodium cacodylate buffer. After three rinses in 0.1 M sodium cacodylate buffer, samples were post-fixed in 1% osmium tetroxide and 0.8% K3[Fe(CN6)] in 0.1 M sodium cacodylate buffer for 1 h at RT. Samples were rinsed with water and stained with 2% aqueous uranyl acetate overnight. Samples were then dehydrated with increasing concentration of ethanol, infiltrated with Embed-812 resin and polymerized in a 60°C oven overnight. Blocks were sectioned with a diamond knife (Diatome) on a Leica Ultracut UCT (7) ultra-microtome (Leica Microsystems) and collected onto copper grids, post stained with 2% uranyl acetate in water and lead citrate. Images were acquired on a JEOL 1400 Plus (JEOL) equipped with a LaB6 source using a voltage of 120 kV. FIJI was used to measure nucleolar perimeter, and circularity.

Polysome fractionation

Cells at 60%–80% confluency were harvested and resuspended in lysis buffer (20 mM Tris pH 7.4, 5 mM MgCl₂, 100 mM NaCl in DPEC-treated dH₂O + 100 μg/mL CHX + RNAse inhibitor RNAse OUT with protease inhibitor and 0.1% NP-40) and incubated on ice for 15 min. Samples were centrifuge at 12,000g 4°C 10 min. Supernatants were collected. RNA amounts were quantified using nanodrop and the same RNA amounts were used for each condition. Gradients were made by solubilizing different sucrose amounts in buffer 0% (20 mM Tris pH 7.4, 5 mM MgCl₂, 100 mM NaCl in DPEC-treated dH₂O + 100 μg/mL CHX + RNAse inhibitor RNAse OUT). The sucrose gradient columns were kept at −80°C until use. Lysates were loaded on columns and centrifuged in a swinging bucket rotor at 34,000 rpm for 2 h at 4°C, acc = 8, dec = 0. BioLogic LP low-pressure chromatography system (BIO-RAD) was used to analyze and collect fractions. Samples were run at 1 mL/min, and the UV recorded. 0.5 mL fractions were collected. Fractions corresponding to single 40S, 60S, and 80S ribosomes according to UV profile were used for proteomics, WB, and bioanalyzer analysis. This experiment was performed twice with similar results. For Figure S7E, lysates were treated with 30 mM EDTA before loading to column.

Xenografts experiments

Either 1 × 10⁶ DLD1 control or ZNF692 CRISPR KO cells (sg1 and sg3) were injected into the flank of female NOD/SCID mice (Jackson lab). Mice were sacrificed when largest tumors reached 2 cm. At the end of the experiment, tumors were harvested, weighed and snap frozen in liquid nitrogen and processed for protein and RNA extraction. Xenograft experiment was performed twice. First time with n = 5 or 6 mice per group. Second time with an n = 8 or 9 per group. All procedures are approved by the Institutional Animal Care and Use Committee of the University of Texas Southwestern Medical Center (approval number: 2017–101798).

Three-step Preribosome Sequential Extraction protocol

The protocol was followed as described^34,35 with slight modification in SN3 buffer. Cells were lyse with SN1 buffer (20 mM HEPES-NaOH [pH 7.5], 130 mM KCl, 10 mM MgCl₂, 0.05% NP-40) followed by SN2 buffer (10 mM HEPES-NaOH [pH 7.5], 10 mM NaCl, 5 mM MgCl₂, 0.1% NP-40, 0.5 mg/mL heparin) and followed by SN3 (50 mM Tris-HCl [pH:8], 5mM EDTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl). All buffers were supplemented with protease inhibitors 1:100 and 600 U/mL RNasin (Promega). SN2 and SN3 buffers were also supplemented with DNAse I.

Nuclear extracts

Nuclear extracts were obtained by lysing cells with lysis buffer A (10 mM Tris-HCl pH 8.0, 50 mM NaCl, 0.1% Nonidet P-40) + proteinase inhibitors for 20 min on ice. After incubation, cells were centrifuged 1200 rpm for 5 min. Supernatants were discarded and nuclear pellets were lysed using buffer A + proteinase inhibitors for 20 min and sonicated at maximal intensity (30 s ON, 30 s OFF) for 5 min. Then lysates were centrifuge at 15,000 g at 4°C. Supernatant were collected as nuclear extracts.

Protein immunoprecipitation (IP)

Cells were collected and lysed using lysis buffer A (10 mM Tris-HCl (pH 8.0), 50 mM NaCl, 0.1% Nonidet P-40) and proteinase inhibitor for 20 min on ice. Lysates were sonicated for 5 min and centrifuged at 15,000 g at 4 °C for 20 min. Supernatants were incubated with primary antibodies or GFP nanotrap magnetic beads overnight at 4°C. For IP with primary antibodies, magnetic beads were added to the lysate-antibody mixture for 3 h rotating at 4°C. For IP with GFP-ZNF692 purified proteins, 1 μL of 5 μM purified protein was incubated with GFP nanotrap magnetic beads and equal amounts of nuclear extracts overnight at 4°C. After incubation, the beads containing the immunocomplexes were washed with lysis buffer A 3 times for 10 min each. Immunoprecipitants were eluted by boiling antibody-beads complexes in 2X SDS Laemli sample buffer. Supernatants were collected and subjected to WB. Antibodies and reagents are listed in Key resources table.

Proteomics

For Figure 4E, HCT116 transfected with GFP, GFP-ZNF692, or GFP-ZNF692 ΔNoLS plasmids were immunoprecipitated with GFP nanotrap after 48 h. Eluted lysates were run on a gel, and proteins were extracted for proteomics analysis. Gel samples were digested overnight with trypsin (Pierce) following reduction and alkylation with DTT and iodoacetamide (Sigma–Aldrich). The samples then underwent solid-phase extraction cleanup with an Oasis MCX plate (Waters) and the resulting samples were injected onto an Orbitrap Fusion Lumos mass spectrometer coupled to an Ultimate 3000 RSLC-Nano liquid chromatography system. The mass spectrometer operated in positive ion mode with a source voltage of 1.8–2.2 kV and an ion transfer tube temperature of 275°C. MS scans were acquired at 120,000 resolutions in the Orbitrap and up to 10 MS/MS spectra were obtained in the ion trap for each full spectrum acquired using higher-energy collisional dissociation (HCD) for ions with charges 2–7. Raw MS data files were analyzed using Proteome Discoverer v2.4 SP1 (Thermo), with peptide identification performed using Sequest HT searching against the human protein database from UniProt. Abundance of each protein was determined by summing the peak intensities for all peptides matched to that protein. The FDR cutoff was 1% for all peptides. To identify ZNF692’s interactome, first, a fold change WT vs GFP >10, and ΔNt vs GFP >10 was applied. Then, to identify the ZNF692 interactors dependent on the NoLS region, we applied a fold change WT vs ΔNt > or = 4. Conversely, to identify ZNF692 interactors that were not dependent on the NoLS, we applied a fold change WT vs. ΔNt <4. We used a stringent cut off of 4 given that the levels of immunoprecipitated WT ZNF692 was about 4 times higher than the ΔNt-ZNF692 (Figure S4B). For Figure 7F, lysates from 40S, 60S and 80S fractions of DLD1 WT or ZNF692 KO were run on a gel and submitted for proteomics analysis as described above. For each condition, the abundance of each peptide was normalized by the sum of all peptide peaks found in each condition. Gene ontology was performed with GSEA-MSigDB using ‘‘GO cellular components’’ setting. Proteomics data were submitted to MASSIVE : MSV000089271.

ChIP-qPCR

Cells at 70% confluency were fixed with 1% formaldehyde, nuclear extracts were lysed with RIPA buffer for 10 min on ice. DNA was sonicated to ~500 bp fragments using a the diagenode sonicator at high intensity (30 s ON, 30 s OFF). Lysates were centrifuged at 15,000 rpm for 15 min at 4°C. Supernatants were collected and immunoprecipitated with anti-ZNF692 or POLR1 antibody (Key resources table) and magnetic beads. Normal rabbit IgG was used a negative control. Beads were washed with lysis buffer A (10 mM Tris-HCl [pH 8.0], 50 mM NaCl, 0.1% Nonidet P-40) 3 times. To reverse DNA crosslinking, beads were resuspended with 200 μL of elution buffer (1% SDS; 50 mM Tris-HCl pH 8) and incubated with 12 μL of 10 mg/mL RNAse A and 24 μL 5M NaCl overnight at 65°C. The next day, 4 μL of 10 mg/mL proteinase K, 4 μL 0.5 M EDTA and 8 μL 1M Tris-HCl pH 6.8 were added to the samples and incubated for 3 h at 45°C. After incubation, beads were removed with the magnet, and the DNA purified from the supernatants for qPCR. Fold change enrichment was normalized by comparing the amount of DNA immunoprecipitated with the specific antibodies in comparison with normal IgG. Primers are listed in Key resources table.

RNA-IP

Cells at 70% confluency were UV crosslinked at 254 nm at 400 mJ/cm². Nuclear extracts were lysed with RNA-IP lysis buffer (50mM Tris-HCl pH 7.4, 100mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate [protected from light]) + proteinase and RNAse inhibitors for 20 min on ice. Lysates were centrifuged 15,000 rpm for 15 min at 4°C. Supernatants were collected, and the same amounts of nuclei lysates were used for immunoprecipitation. Magnetic beads were incubated without antibody (negative control) or antibodies for ZNF692 or NPM1 for 45 min at RT. Next, nuclear lysates were added to the antibody–bead complexes and incubated overnight at 4°C. Beads were washed 3 times with high salt wash buffer (50mM Tris-HCl pH 7.4, 1 M NaCl, 1mM EDTA, 1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate [protected from light]) and 3 times with wash buffer #2 (20 mM Tris-HCl pH 7.4, 10 mM MgCl₂, 0.2% Tween 20). Beads were then resuspended in 100 μL of wash buffer #2 containing DNAse I and incubated for 1–2 h at 37°C. RNA was extracted using Tri-Reagent and the RNA purification RNeasy Kit (Qiagen). To quantify the amount of RNA immunoprecipitated, half of the RNA eluted from the column was reverse transcribed to cDNA, and the other half was incubated with the buffer containing no retro-transcriptase as a negative control. After cDNA synthesis, qPCR was performed as indicated above. The 2^−ΔΔCt method was used. The amount of RNA in each condition was first normalized by the amount of that condition without retro-transcriptase. Then, each condition compared with the control conditions to determine the fold change enrichment. Primers are listed in Key resources table.

Northern blot

Cells were starved overnight and fresh media containing FBS was added the following morning for 24 h. Then, the RNA was extracted with Tri-Reagent (Sigma). Northern blots were performed following Northern Max Kit (Invitrogen) instructions. Total RNA (7.5–10 mg) was loaded onto a 1.5% agarose denaturing gel (Northern Max Kit, Invitrogen) and electrophoresed at 70 V for 3 h. RNAs in the gel were visualized using a UV imager. Next, RNAs were transferred to a nylon membrane and crosslinked using UV light (120000 μJ× 3 pulses). Membranes were then prehybridized with ULTRAhyb oligo buffer (Invitrogen) for 30 min at 37°C. For Figures 6B, 6E, and S6F, membranes were hybridized overnight at 42°C with probe that was labeled with γ-³²P-ATP (PerkinElmer Cat # NEG035C; 6000 Ci/mmol) and membranes were exposed after washing with 2X SSC and 0.5% SDS for 3 times. For Figures S6C, membranes were hybridized overnight at 37°C and washed twice with Stringency wash buffer (2X SSC and 0.5% SDS) and Wash buffer (0.5% SDS and 1X PBS). Then, membranes were blocked with Blocking buffer (0.5% SDS, 1% BSA, and 1X PBS) for 30 min, followed by incubation with HRP-conjugated streptavidin (Thermo Scientific, 1:4,000) for 1 h at 37°C. Membranes were washed with Wash buffer 3 times and 1X PBS 3 times. Northern signals were detected by chemiluminescence. Northern blots were quantified following Rapid Analysis of Multiple Precursors method.⁴⁷

RNA-FISH

Cells grown on glass coverslips were fixed with 4% paraformaldehyde PBS for 15 min and permeabilized with 0.1% Triton X-100 for 20 min. Coverslips were rinsed with RT PBS and then hybridized with NorthernMax Prehybridization/Hybridization Buffer (Thermo Fisher) in a humid chamber for 1 h at 65°C. Then prehybridization buffer was removed. Hybridization buffer (20 μL) containing 55 nM biotin-labeled 5ʹ-ETS probes (Table S7) were added overnight at 65°C in a humid chamber. Coverslips were washed 3 times with 23 SSC at 37°C and twice with 13 SSC at RT. Slides were then fixed in 4% formaldehyde PBS for 15 min, washed twice with PBS and blocked with PBS-BSA for 30 min at RT. Streptavidin-Alexa 568 was used at 1:200 dilution for 1 h at RT in the dark. Coverslips were washed twice with PBS, and one final wash with PBS+DAPI before mounting with Mowiol. Images were acquired with a Zeiss LSM780. Images were process using FIJI software.

QUANTIFICATION AND STATISTICAL ANALYSIS

All statistical analyses were performed using two-tailed unpaired T-student statistical analysis, p < 0.05 was considered statistically significant. All values are reported as mean ± SD in each figure.

Highlights.

• MYC promotes the expression of ZNF692 in highly proliferative cells

• ZNF692 is predominantly localized in the granular component of the nucleolus

• ZNF692 and the SSU processome interact to promote small ribosome subunit biogenesis

• ZNF692 is needed for efficient translation and growth of hyperproliferative cells