Expression of human Bcl-xL (Ser49) and (Ser62) mutants in Caenorhabditis elegans causes germline defects and aneuploidy

An interesting feature of Bcl-xL protein is the presence of an unstructured loop domain between α1 and α2 helices, a domain not essential for its anti-apoptotic function and absent in CED-9 protein. Within this domain, Bcl-xL undergoes dynamic phosphorylation and dephosphorylation at Ser49 and Ser62 during G2 and mitosis in human cells. Studies have revealed that when these residues are mutated, cells harbour mitotic defects, including chromosome mis-attachment, lagging, bridging and mis-segregation with, ultimately, chromosome instability and aneuploidy. We undertook genetic experiments in Caenorhabditis elegans to understand the importance of Bcl-xL (Ser49) and (Ser62) in vivo. Transgenic worms carrying single-site S49A, S62A, S49D, S62D and dual site S49/62A mutants were generated and their effects were analyzed in germlines of young adult worms. Worms expressing Bcl-xL variants showed decreased egg-laying and hatching potency, variations in the length of their mitotic regions but not of their transition zones, appearance of chromosomal abnormalities at their diplotene stages, and increased germline apoptosis, with the exception of the S62D variants. Some of these transgenic strains, particularly the Ser to Ala variants, also showed slight modulations of lifespan compared to their controls. In addition, RNAi experiments silencing expression of the various Bcl-xL variants reversed their effects in vivo. Our in vivo observations confirmed the importance of Ser49 and Ser62 within Bcl-xL loop domain in maintaining chromosome stability.


Progeny count
Adult worms were collected and washed with M9 buffer to remove bacterial contamination. Worm pellets were treated with freshly-prepared 0.5 ml NaOH(5N) mixed with 1 ml commercial bleach for 10 min. Samples were briefly vortexed at 2-min interval, then centrifuged for 30 s at 1,300 g to pellet the released eggs. The pelleted eggs were washed with sterile H 2 O, volume reduced to 100 μl and then plated in fresh NGM. When larvae reached the L4 stage, they were placed individually in fresh NGM plates. Progeny count was started 12 h latter, once they reached the young adult stage. Eggs layed were counted for the first 3 days of adulthood, and hatched eggs were counted every 8 h. Worms were transferred into fresh NGM plates every day.
Gonad staining, MR and TZ measurements, and germ cell count Young adult worms (12 h from the L4 larvae stage) were washed with M9 buffer and placed on glass slides. Worms were fixed with 100% cold methanol (-20˚C) for 30 s and washed with M9 buffer. Gonads were stained with DAPI at a concentration of 1.0 ng/ml for 3 min and washed with phosphate buffered saline (PBS) prior to microscopy. MR/TZ or TZ/pachytene boundaries were well marked under the microscope. Germ cell numbers in the gonad were also determined by first marking MR/TZ or TZ/pachytene with the microscope, and then counting nuclei through germline width. Images were generated with a Zeiss Axio Observer Z1 automated microscope equipped with Axiovision software (v4.8.2).

Fluorescence staining and analysis
Young adult worms were dissected to expose their gonads and fixed on polylysine-coated slides with 100% cold methanol (4 min), followed by 100% cold acetone (4 min). For immunofluorescence staining, the samples were first treated with a blocking solution containing 10% donkey serum (Jackson Immuno Research Laboratories, Inc., West Grove PA) for 1 h at room temperature, then stained overnight at 4˚C in an humidified chamber with RAD-51 (14B4) Alexa Fluor (R) 488 Ab (Novus Biologicals, Oakville ON), in PBS solution containing 0.5% (v/v) Triton X-100, 1 mM EDTA pH 8.0, 0.1% (w/v) bovine serum albumine (BSA) and 0.05% (w/v) sodium azide. Samples were also stained with DAPI at a concentration of 1.0 ng/ml for 3 min. Images were generated with a Nikon microsystem mounted on a Nikon Eclipse E600 microscope with a photometric Cool-Snap HQ2 camera and Nikon NIS-Elements software 9 (v 3.8AR) or with a Zeiss Axio Observer Z1 automated microscope and Axiovision software (v4.8.2). Images were analysed by Image J software (v1.49), a Java-based processing program developed by the National Institutes of Health (USA).

Western blots
Worms were collected in maintenance (M9) buffer and washed several times until no visible bacterial contamination was observed in the washing M9 buffer. Worm pellets were froozen overnight at -80˚C, and pellets were lysed with buffer containing 150 mM NaCl, 50 mM Tris: HCl, pH 7.4, 1% (v/v) Triton X-100, 0.1% (w/v) sodium dodecyl sulfate (SDS), 1% (w/v) sodium deoxycholate (w/v), 10 μg/ml leupeptin, 10 μg/ml chymostatin and 10 μg/ml pepstatin A. Pellets were directed through 271/2G syringes (10-fold), sonicated and centrifuged at 16,000g. Protein in supernatants were collected and their concentrations measured by bicinchoninic acid assays as described in the manufacturer's protocol (Thermo Fisher Scientific). Bcl-xL(54H6) rabbit monoclonal Ab (Cell Signaling Technology, Whitby, ON), and actin (AC-15) mouse monoclonal Ab (Abcam Inc., Cambridge, MA) were tested in this study. Peroxidaselabeled secondary Ab were detected by enhanced chemiluminescence with reagent set from GE Healthcare Life Science (Mississauga, ON) or SuperSignal WestPico chemiluminescence substrates (Thermo Scientific, Rockford, IL). Densitometry analysis, were analysed by Image J software (v1.49), a Java-based processing program developed by the National Institutes of Health (USA).

Lifespan assay
Lifespan assays were peformed at 20˚C. Briefly, 45 to 50 L4 hermaphrodite transgenic larvae, and N2 (wt) larvae were placed in 400 μM 5'flurodeoxyuridine-NGM plates in triplicate, for 6 days. Day 1 was defined as the day when the worms reached adulthood. Worms were scored every 1 to 3 days. On the 6th day, they were transferred to fresh NGM plates. Strains were considered to have lost viability if they exhibited arrested development or died.

Expression of human Bcl-xL variants in C. elegans
Several strains of transgenic worms containing human Bcl-xL wild-type (wt) and single-site (S49A, S49D, S62A, S62D) or dual-site (S49/62A) mutants were generated (S1 Table). First, 2 strains for each Bcl-xL variant and Bcl-xL wt (COP672 and COP689) were used, and N2 (wt) worms served as controls. Transgenes were confirmed by polymerase chain reaction (PCR) assays of genomic DNA ( Fig 1A). mRNA levels were evaluated by quantitative reverse-transcriptase (q-RT)-PCR using pmp-3 expression as referencence gene [25], and all transgenic worms expressed similar levels of BCL-XL variant mRNAs ( Fig 1B). Ced-9 mRNA levels were monitored in parallel in these experiments ( Fig 1C). Overall, BCL-XL mRNA expression was found slightly lower compared to ced-9 mRNA expression, with BCL-XL / ced-9 ratios ranging from 0.56 to 0.88 ( Fig 1D). On randomly selected strains representing each variant, mRNA levels in the gonads were also evaluated by quantitative reverse-transcriptase (q-RT)-PCR using pmp-3 expression as referencence gene ( Fig 1E). Expression at the protein level was confirmed by Western blotting (Fig 1F). Overall, Bcl-xL and Bcl-xL variant protein expression was found variable, with some strains expressing higher protein level compared to Bcl-xL (wt), perhaps suggesting differential stability among these proteins.
Mutations within the loop domain of Bcl-xL affect C. elegans progeny fecundity N2 (wt) hermaphrodites under laboratory conditions released embryos from the uterus during their development, and these eggs were easily counted. To determine whether the expression of Bcl-xL (wt) and mutants in C. elegans affects their progeny, eggs were counted and compared among various transgenic strains. Synchronized L4 hermaphrodite transgenic larvae were plated individually, and eggs laid were counted over a 3-day period once they have reached adulthood (about 12 h after plating). The transgenic strains expressing Bcl-xL (wt) showed no significant difference in egg-laying potency (Fig 2A) and in percentage (%) of eggs that hatched ( Fig 2B) compared to N2 (wt) worms. The number of eggs laid by transgenic worms expressing Bcl-xL mutants was decreased significantly compared to Bcl-xL (wt)-expressing worms and N2 (wt) worms ( Fig  2A). When these eggs were followed, the % that hatched was significantly affected across all the mutants with the exception of Bcl-xL (S62D) variant strain 311, that may reflect strain and/ or animal variations. Overall, embryonic lethality is presented in S1 Fig To determine if Bcl-xL mutations at Ser49 and Ser62 affect the process of mitosis, we analyzed the germline in the transgenic worms. The C. elegans germline contains an anatomicallyrestricted mitotic cell population that persists throughout life and is thought to be self-renewing [26]. The mitotic region (MR) in the gonads showed reduced lengths in all Bcl-xL variants compared to Bcl-xL (wt)-expressing worms or N2 control worms (Fig 3A) with the exception of Bcl-xL (S62D) variant. In contrast, the transition zones (TZ) in the gonads presented no overall difference in all Bcl-xL variants compared to Bcl-xL (wt)-expressing worms or N2 control worms (Fig 3B). Representative micrographs are reported in Fig 3C. ced-9 and pmp-3 and D) relative expression of BCL-XL and ced-9 in the transgenic worms. In E) mRNAs were extracted from 250-300 gonads of selected strains and expression levels assessed by qRT-PCR of BCL-XL and pmp-3. Results from 2 independent determinations. Bars are means ± variations. F) Immunoblots of Bcl-xL variant expression in various transgenic strains and control worms. Actin expression is shown as control. Bcl-xL/ACTIN ratio are indicated. https://doi.org/10.1371/journal.pone.0177413.g001 Mutations within the loop domain of Bcl-xL cause germ line aneuploidy Chromosome mis-alignment, lagging or bridging, mis-segregation and cytokinesis failure are major defects that could occur during mitosis and confer chromosome instability as well as aneuploidy [27,28]. Cells will respond in various ways including cell cycle checkpoint activation, cell cycle arrest, premature senescence or cell death and, in mammal cells, they could also enter into an immortalization or tumorigenesis path, depending on particular cellular and   [29][30][31]. To evaluate the effect of the various Bcl-xL variants on chromosome stability in C. elegans, chromosomes were analyzed at the diplotene stage in control and transgenic worms. In N2 control and in Bcl-xL (wt) worms, 6 pairs of chromosomes were clearly visible in the diplotene stage at proximal gonad. However, various Bcl-xL mutants, with the exception of Bcl-xL (S62D) variant, underwent aneuploidy and/or chromosome fragmentation compared to N2 control and Bcl-xL (wt) strains (Fig 4A). RAD51-associated nuclear foci in the germline confirmed the presence of DNA double-strand breaks in transgenic strains expressing Bcl-xL mutants, with the exception of the Bcl-xL (S62D) variant (Fig 4B). N2 animals subjected to high-dose radiation (60 Gy) were collected 4 hrs post-irradiation and used as strong reference controls (Fig 4A and 4B).

Mutations within the loop domain of Bcl-xL cause increased apoptosis in the gonads
Finally, to assess apoptosis in germlines, we took advantage of a CED-1:GFP strain [32] by crossing it with our transgenic worms. CED-1, expressed in sheath cells, is a phagocytic receptor that initiates pathways for degrading engulfed apoptotic cells and is thus a good indicator apoptotic bodies [33]. With the exception of the Bcl-xL (S62D) variant, worms expressing Bcl-xL mutants showed significantly increased apoptotic bodies compared to those expressing Bcl-xL (wt) and the N2 (wt) worms (Fig 5). Again, N2 animals subjected to high-dose radiation (60 Gy) were collected 4 hrs post-irradiation and used as strong reference controls (Fig 5).

Longevity changes due to the expression of Bcl-xL mutants
The C. elegans N2 strain has an average lifespan of around 2-3 weeks at 20˚C [34]. N2 and Bcl-xL (wt) worms showed no significant differences in their lifespan (Fig 6). Strains expressing Ser to Asp variants also presented no significant differences compared to N2 controls. In contrast, overall lifespan was significantly increased in strains expressing Ser to Ala variants (Fig 6).

Reversion of the phenotypes
Finally, to confirm that the phenotypes observed were due to the expression of Bcl-xL variants in transgenic C. elegans, a serie of RNA interference experiments were conducted. Silencing BCL-XL mRNA variant expression (Fig 7A) in the transgenic worms reversed the phenotypes, including effects on germline fecundity measured as egg-laying and egg-hatching potency ( Fig  7B and 7C), mitotic region length ( Fig 7D) and transition zone length (Fig 7E), germline aneuploidy (Fig 7F), apoptotic corpse appearance in the gonads (Fig 7G) and lifespan for the Ser to Ala variants (Fig 7H). Typical micrographs are showed in S3 and S4 Figs.
A summary of observations is presented on Table 1.

Discussion
The loop domain between α1 and α2 helices of Bcl-xL in higher organisms may be due to gain-of-function domain through evolution. In most studies, the loop domain is not necessary for the Bcl-xL's anti-apoptotic activity in mammallian cells [11][12][13][14][15][20][21][22]. However, mutations within Bcl-xL (Ser49) and (Ser62) residues lead to chromosome instability and represents data obtained from a single worm; both MR and TZ were determined from the same worms. Bars are means ± s.d. Arrows on top indicate statistical significance with p<0.05 when compared to N2 control C) lowmagnification images of DAPI-stained cells.
https://doi.org/10.1371/journal.pone.0177413.g003 The presence, or absence, of a similar conserved function within C. elegans protein remains to be discovered. [35]. In this study, Bcl-xL phosphorylation in C. elegans was not determined; however the presence of conserved PLK-1 and PLK-3 activities in C. elegans raises a probability of similar phosphorylation at Ser62 and Ser49 alike human cells. Indeed, PLK-1 and PLK-3 are well conserved in C. elegans and involved in multiple process of mitosis, including spindle formation, kinetochoremicrotubule attachments, sister chromatid separation and cytokinesis [36,37]. PLK-1 is shown to be involved in nuclear envelope breakdown in the oocyte during meiosis and in structures observed at the diplotene stages at proximal gonads. B) % of cells showing RAD51-associated nuclear foci in germline nuclei (left graph). Each point in the graph represents data obtained from a single worm. Bars are means ± s.d. Arrows on top indicate statistical significance with p<0.05. Right images: typical low-magnification images of immunofluorescence illustrating RAD51-associated nuclear foci (red) and DAPI-stained cells (blue). N2 animals subjected to high-dose radiation (60 Gy) were collected 4 hrs post-irradiation and used as strong reference controls in A) and B).
https://doi.org/10.1371/journal.pone.0177413.g004 mixing maternal and paternal genomes after fertilization. Partial inactivation of PLK-1 caused failure of alignment of chromosomes at metaphase during mitosis and the nuclear membrane remains intact [38]. In contrast, PLK-3 mutations caused delay in chromosome condensation at diakinesis indicating that PLK-3 does not play a major role in meiosis [39]. Although the CED-9 protein lacks the Bcl-xL functional loop domain, we tested whether or not human Bcl-xL (Ser49) and (Ser62) variant expression in C. elegans exhibits dominant effects on mitotic behaviors, as observed previously in human cells. The proliferative properties of germlines of adult young worms [40], as well as the short and reproducible lifespan of C. elegans is well-characterized [41]. Expression of mammalian proteins is prevalent in C. elegans and viceversa is prevalent. Expression of human Bcl-2 itself partially prevents apoptosis in C. elegans [7,42], whereas CED-9 expression in monkey fibroblast COS cells and embryonic drosophila Schneider's L2 cells reveals co-localization of the 2 proteins, suggesting similar functions [43].
In the present study, we observed that human Bcl-xL (Ser49) and (Ser62) variant expression in C. elegans interfered with germline fertility, effects that correlated with MR length variations, the appearance of chromosomal aberrations, RAD51-associated foci and increase apoptosis with the exception of Bcl-xL (S62D) variants. It is interesting to note that the Bcl-xL (S62D) strain that express higher protein level compared to other variants, showed undetectable cell division errors (summary in Table 1). Most likely, reduced fecundity could resulted from cell division errors in germlines and embryos, resulting in chromosome aberrations, aneuploidy and augmented apoptosis. MR length variations were seen as being decreased in individual worms and strains bearing Bcl-xL variants. Perhaps defects in mitosis and elevated  apoptosis could account for shortened length. All experiments were performed on proliferative germline in the gonads. In the near future, time-lapse imaging on embryos could further document mitotic behaviours and chromosome stability/instability in dividing embryos.
Lifespan modulation also has been observed in Ser to Ala variants, a possible consequence of aneuploidy, DNA damage and increased apoptosis of germlines. Indeed, repeated ultra-violet electromagnetic radiation exposure has been shown to severely reduce lifespan in C. elegans [44], whereas mutations in the nucleotide excision repair proteins ERCC-1 and XPF-1 extend lifespan in daf-2 worms. Fecundity also decreased in worms expressing mutant ERCC-1, XPF-1 and XPG-1 compared to wt proteins [45]. In the long-term, it would be interesting to monitor gene expression profiles in various strains to identify genes whose expression could be altered as well.
The exact mechanisms by which Bcl-xL exerts its function on chromosome stability are unknown, but Ser49 and Ser62 are 2 essential residues associated with this activity in human cells. Previous studies have revealed the presence of phospho-Bcl-xL(Ser49) and (Ser62) in centrosomes with γ-tubulin during G2 and mitosis, respectively [20,22]. In addition, phospho-Bcl-xL(Ser62) interacts in mitotic cytosol with some SAC signaling proteins during prometaphase/metaphase, including the Mad2-, BubR1-, Bub3-and Cdc20-bound complexes [22], while phospho-Bcl-xL(Ser49) is found in mid-zone bodies during telephase/cytokinesis [20]. In C. elegans, most of these key players in the SAC signaling pathway are functionally and structurally conserved, including MAD1, MAD2, MAD3/BubR1, BUB1, BUB3 and FZY-1/ Cdc20 [46]. Proliferating germ cells have functional SAC [47], but SAC function in C. elegans embryos is unclear due to their lack of apparent mitotic arrest phenotype [48]. SPDL-1, a C. elegans homolog of kinetochore-specific dynein recruiter protein [49] that senses the microtubule attachment status of kinetochore and functions upstream of MAD1MDF-1, is part of the kinetochore receptor of the MAD1MDF-1-MAD2MDF-2 complex that regulate APC/C activity [49,50]. Whether or not, and how, human Bcl-xL interplays with these C. elegans components remains to be elucidated.
Bcl-xL expression in human cancers is often associated with poor prognosis and chemotherapy resistance [51][52][53]. Current efforts are being made to develop and test new drugs targeting the conventional BH1-, BH2-, BH3-forming hydrophobic pocket domain of Bcl-2 antiapoptotic members including Bcl-xL [54][55][56][57][58][59]. Future perspectives should also focus on the loop domain of Bcl-xL and Bcl-2 for therapeutic evaluation. These in vivo transgenic strains will be an important tool to screen and evaluate the effects of future putative new compounds targeting this function.