BAP1 Malignant Pleural Mesothelioma Mutations in Caenorhabditis elegans Reveal Synthetic Lethality between ubh-4/ BAP1 and the Proteasome Subunit rpn-9/ PSMD13

Abstract

The deubiquitinase BAP1 (BRCA1-associated protein 1) is associated with BAP1 tumor predisposition syndrome (TPDS). BAP1 is a tumor suppressor gene whose alterations in cancer are commonly caused by gene mutations leading to protein loss of function. By CRISPR-Cas, we have generated mutations in ubh-4, the BAP1 ortholog in Caenorhabditis elegans, to model the functional impact of BAP1 mutations. We have found that a mimicked BAP1 cancer missense mutation (UBH-4 A87D; BAP1 A95D) resembles the phenotypes of ubh-4 deletion mutants. Despite ubh-4 being ubiquitously expressed, the gene is not essential for viability and its deletion causes only mild phenotypes without affecting 20S proteasome levels. Such viability facilitated an RNAi screen for ubh-4 genetic interactors that identified rpn-9, the ortholog of human PSMD13, a gene encoding subunit of the regulatory particle of the 26S proteasome. ubh-4[A87D], similarly to ubh-4 deletion, cause a synthetic genetic interaction with rpn-9 inactivation affecting body size, lifespan, and the development of germ cells. Finally, we show how ubh-4 inactivation sensitizes animals to the chemotherapeutic agent Bortezomib, which is a proteasome inhibitor. Thus, we have established a model to study BAP1 cancer-related mutations in C. elegans, and our data points toward vulnerabilities that should be studied to explore therapeutic opportunities within the complexity of BAP1 tumors.

Keywords: BAP1; Bortezomib; C. elegans; CRISPR-Cas; Malignant Pleural Mesothelioma; PSMD13; proteasome; rpn-9; ubh-4.

Figure 1

C. elegans BAP1 ortholog UBH-4 homology and expression pattern. (A) UBH-4 is the ortholog of human BAP1. Scheme of UBH-4, UCHL5 and BAP1 protein sequences (black lines) and conserved residues (purple bars), predominantly at the N-terminal and C-terminal domains. In the black square are the catalytic domains of these proteins. The bottom scheme shows with more detail conserved residues between UBH-4 and BAP1. Pink shadows indicate MPM-related residues modelled in this study. Table shows BLASTP analysis of C. elegans UCHs proteins, UCHL5 and BAP1. Alignments were illustrated by CLC Sequence Viewer 8.0. (B) UBH-4 is ubiquitously expressed during C. elegans development. Representative images of UBH-4::EGFP signal in developing embryo (1), L3 (2), and adult stages (3). Scale bars represent 10, 100, and 100 µm, respectively. (4) Posterior arm of the adult gonad at higher magnification (scale bar represents 100 µm) and its diagram. (5) (2) pointing out distal mitotic (a), transition (b), and meiotic zones (pachytene (c), diplotene (d) and diakinesis (e)), and spermatheca (f).

Figure 2

CRISPR-Cas generation and characterization of UBH-4 mutant alleles. (A) Molecular designs to generate the ubh-4 cancer-related alleles cer25 and cer32 by CRISPR-Cas9. Sense strands are represented. PAM sequences are shown in yellow; crRNA sequences are underlined and Cas9 cut sites are indicated by black arrows; silent mutations are labelled in light blue; mutated codons are shown in pink; nucleotide changes (T > G (cer25) and C > A (cer32)) are underlined. Aminoacidic sequences are represented below nucleotide sequences, highlighting the mutated residues in pink. (B) Phenotypes observed for indicated alleles after brood size, body length and proteasome amount and activity characterization.

Figure 3

RNAi for synthetic lethal interactions with ubh-4 mutants. (A) RNAi screen workflow is illustrated. After candidate genes are selected (1), clones from RNAi libraries were picked aleatorily and genotyped to generate a 150-gene RNAi sublibrary for screening (2). For the RNAi screen, 150-RNAi cultures were IPTG-induced in RNAi 24-well plates. Synchronized L1 animals were seeded and checked every day for 120 h at 20 °C, to score both wild-type and cer27 animals (3). Finally, candidate genes were validated in larger (55 mm-diameter) RNAi plates (4). (B) Representative images of reduced brood size in ubh-4(cer27) background after 120 h rpn-9(RNAi) treatment (20 °C). The chart quantifies hatched larvae, dead embryos, and blister animals under different RNAi conditions. Statistics were performed by one-way ANOVA (Kruskal–Wallis and Dunn’s tests). ** p < 0.01 comparing hatched larvae between wild-type and cer27 under rpn-9(RNAi). No significant differences were found between the other observed phenotypes. An average of 20 animals were scored. The experiment was performed once. (C) Schematic model of the C. elegans 26S proteasome. 26S complex is formed by two regulatory particles 19S and one core particle 20S [39]. Structure subunit forms are represented by colors and detailed subunit proteins by coding gene names. Rings conforming to the 20S core and 19S structure are represented linearly. Black arrows point out rpn-9 subunits. Illustrations were drawn using Inkscape 1.2.2. (Retrieved from https://inkscape.org).

Figure 4

ubh-4 deletion and [A87D] significantly reduces body length and survival in the rpn-9 mutant background. (A) Representative images showing smaller body length in ubh-4 and rpn-9 single and double mutants (i, ii, iii, or use top and bottom rows). Pictures were taken at day 1 of adulthood growth at 1 °C. Scale bar represents 0.5 mm. Below, the chart quantifies the body length in wild-type (WT), single, and double mutants. Bars represent median and interquartile range, and dots the measured length of individual animals. The same experiment was performed three times, measuring an average of 100 animals per replicate. **** mean p < 0.0001. Statistics were analyzed by one-way ANOVA (Kruskal–Wallis and Dunn’s tests). (B) Survival curves of WT, ubh-4, and rpn-9 simple and double mutants at 1 °C and 2 °C. Percent survival is indicated for each allele. An average of 80 animals were censored through two independent experiments for lifespan analysis at both temperatures. Statistical analysis by Log-rank (Mantel–Cox) test was performed to compare rpn-9(gk401) and double mutants. * and **** indicate p < 0.1 and p < 0.0001, respectively.

Figure 5

UBH-4 and RPN-9 expression patterns suggest these proteins cooperate, particularly in the germline. (A) Single and merged UBH-4::EGFP and RPN-9::wrmScarlet signals evidence ubiquitously co-expression of both proteins at different stages (embryo, young adult and adult stages). Scale bar represents 100 µm. Representative images showing endogenous levels of UBH-4 and RPN-9 during spermatogenesis (B) and oogenesis (C).

Figure 6

DAPI-stained nuclei in gonads at day one adulthood. (A) DAPI staining of dissected gonads from WT, simple and double ubh-4 and rpn-9 mutants. (-) indicates deletion. Scale bars represent 50 µm. (B) Whole-body DAP staining of a double mutant ubh-4(cer150); rpn-9(gk104) showing abundant fragmented nuclei in the germline. The highlighted region is shown at higher magnification (63× lens) at the upper part. Scale bar represents 100 µm.

Figure 7

Dose-dependent effect of proteasome inhibitor Bortezomib on ubh-4 deletion mutant. Graphs represent the health status of both wild-type and CER535 animals under different Bortezomib dosages after two (A) and three (B) days of exposure at 20 °C. Animals were transferred to Bortezomib-containing plates at the L4 stage and were observed on day 2 and day 3 of adulthood. Animals were scored on basis of mobility and visual appearance as normal, sick (defective movement and appearance), very sick (highly defective movement and appearance), and dead. Results are the mean of three independent experiments (~30 animals in each experiment). Statistical analysis by t-test were preformed between wild-type and ubh-4 mutant (CER535) within each corresponding phenotypic group (normal, sick, very sick, and dead). Error bars, STD, * p < 0.05, ** p < 0.01 and *** p < 0.001.