SmedOB1 is Required for Planarian Homeostasis and Regeneration

The planarian flatworm is an emerging model that is useful for studying homeostasis and regeneration due to its unique adult stem cells (ASCs). Previously, planaria were found to share mammalian TTAGGG chromosome ends and telomerases; however, their telomere protection proteins have not yet been identified. In Schmidtea mediterranea, we identified a homologue of the human protection of telomeres 1 (POT1) with an OB-fold (SmedOB1). SmedOB1 is evolutionarily conserved among species and is ubiquitously expressed throughout the whole body. Feeding with SmedOB1 double-stranded RNAs (dsRNAs) led to homeostasis abnormalities in the head and pharynx. Furthermore, several ASC progeny markers were downregulated, and regeneration was impaired. Here we found that SmedOB1 is required for telomeric DNA-protein complex formation and it associates with the telomere TTAGGG sequence in vitro. Moreover, DNA damage and apoptosis signals in planarian were significantly affected by SmedOB1 RNAi. We also confirmed these phenotypes in Dugesia japonica, another flatworm species. Our work identified a novel telomere-associated protein SmedOB1 in planarian, which is required for planarian homeostasis and regeneration. The phylogenetic and functional conservations of SmedOB1 provide one mechanism by which planarians maintain telomere and genome stability to ensure their immortality and shed light on the regeneration medicine of humans.

regeneration. Inhibiting SmedOB1 expression may disrupt stem cell self-renewal causing DNA damage and apoptosis. By EMSA assays using purified proteins or endogenous nuclear extracts from planarians, we confirmed that SmedOB1 is required for telomeric DNA-protein complex association. Consistently, loss of SmedOB1 significantly increased the DNA damage as well as apoptosis signals in planarian. We also confirmed these phenotypes in another flatworm species, Dugesia japonica. In this study, we identified a novel telomere associated SmedOB1 protein that is required for freshwater planarian homeostasis and regeneration, indicating that this telomere-associated protein is crucial for planarians to maintain telomeres and genome stability to ensure their immortality.

Results
Identification of a Schmidtea mediterranea homologue of human POT1. Planarians have telomere repeat sequences similar to those of humans (TTAGGG) 13,14 . Indeed, using probes containing TTAGGG sequences, we detected telomeres in metaphase spreads of two species of planarians, Schmidtea mediterranea and Dugesia japonica ( Supplementary Fig. S1A,B). We then used the sequence from the longest isoform of human POT1 (isoform 1) to search the Schmidtea mediterranea genome database (http://smedgd.stowers.org) and found that the 333-bp genome sequence V31.002347: 3945..3613 was a significant match. Then, using 5′ and 3′ RACE PCR, a 1506-bp full-length cDNA was obtained. We named this gene SmedOB1 (Fig. 1A) (Supplementary Fig.  S2A,B).
The predicted protein has 501 amino acids with an N-terminal region (amino acids 1-147) that is highly conserved among other species (Fig. 1B,C). A domain analysis revealed the presence of one OB fold that corresponds to OB1 (Fig. 1D). Although both OB folds of the human POT1 possess ssDNA binding capacity, OB1 is indispensible for human POT1 binding to ssDNA. The C-terminus of human POT1 is responsible for associating with other shelterin complex protein such as TPP1 15 . Accordingly, we hypothesized that SmedOB1 may also be involved in ssDNA binding.

SmedOB1 is required for tissue homeostasis in Schmidtea mediterranea.
To decipher the role of SmedOB1 in planaria, we first performed whole-mount in situ hybridization (WISH) analysis using SmedWi-1, an adult stem cell marker, as a positive control 16 . SmedOB1 was expressed ubiquitously in the whole body ( Fig. 2A). To better understand the function of SmedOB1, we conducted RNAi experiments in intact Schmidtea mediterranea (Fig. 2B). The worms were fed double-stranded RNAs (dsRNAs) targeted against SmedOB1. The dsRNA targeting GFP served as a negative control. According to RT-PCR analysis, SmedOB1 was efficiently silenced after two rounds of dsRNA feeding (Fig. 2C). When we examined the RNAi-treated worms, we observed that the SmedOB1-RNAi planarians were morphologically abnormal compared with the control worms. The SmedOB1-RNAi planarians showed tissue injuries near the pharynx and behind the eyespots (darker regions in the body) ( Fig. 2D and Supplementary Fig. S3A). Over time, these worms either split at the sites of these apparent tissue injuries or died, suggesting a critical role for SmedOB1 in planarian homeostasis (Fig. 2D, right, and SmedOB1 is required for regeneration in Schmidtea mediterranea. When planarians are irradiated with X-rays, the ASCs lose their ability to proliferate, and the worms die. Also the stem cell markers like SmedWi-1 will be decreased 17 . We therefore examined the expression level of SmedOB1 following X-ray irradiation. As shown in Supplementary Fig. S4, the expression level of SmedOB1 decreased to approximately 30% after irradiation ( Supplementary Fig. S4), further linking SmedOB1 to the planarian regeneration process.
We next evaluated the effect of SmedOB1 on regeneration post-amputation. We first treated the worms with one round of dsRNA feeding. This treatment led to a 60% reduction in the SmedOB1 transcription level (Fig. 3A). The worms were then cut crosswise into three sections and further cultured (Fig. 3B). Approximately 4 days post-amputation, the control worms grew new white blastemas around the cut areas in each section, indicating successful tissue regeneration. However, in the SmedOB1-RNAi worms, there was a significant delay in the appearance and the extent of the blastema growth. This delayed regeneration phenotype was observed in all three sections (the head, trunk and tail) and was repeatable in different dsRNA feeding groups ( Fig. 3C and Supplementary Fig. S3B). The above findings indicate that SmedOB1 is involved in the tissue regeneration process in planarians.
The stem cells in planarians can be divided into three categories based on the expression of different stem cell markers. Expression of Wi2 represents the early stage, high expression of NB21.11e represents the middle stage, and expression of Agat-1 represents the late stage 17 . All three stem cell markers were downregulated in planarians after SmedOB1 knockdown (Fig. 3D). Notably, the proliferation marker PCNA was also downregulated in these worms. These findings point to a reduced number of stem cells in SmedOB1 RNAi worms, possibly due to the decreased proliferation of these cells.
SmedOB1 is required for telomeric DNA-protein complex association. Next, we tested whether SmedOB1 could also bind single-stranded telomere DNA. We prepared the nuclear extract (NE) from planarians that were fed with one round of dsRNAs targeted against GFP or SmedOB1. By the same gradient, the si-GFP NE shifted the probe while the si-OB1 NE was unable to shift the telomere probe (Fig. 4A, lanes 2-4 compared to lanes 5-7). Interestingly, there are two shifted bands in si-GFP NE group. The upper shifted bands might be a multi protein-telomere complex while the lower shift indicates as SmedOB1-telomere complex (Fig. 4A, upper and middle arrows). These results suggested that SmedOB1 may be functionally similar to human POT1 and is required for protein-telomere complex formation. Next, SFB tagged full length SmedOB1 and human POT1 were purified from eukaryotic cells (Fig. 4B) and incubated with a biotin-labelled telomere probe (TTAGGG) 8 (Fig. 4C). As shown in Fig. 4C, the human POT1 shifted the probe (Fig. 4C, lanes 6-8). With an increased gradient, the full length SmedOB1 was able to shift the oligo probe, although the signal is relatively weaker than human POT1 (Fig. 4C, lanes 2-4). SFB-tagged GFP purified from human 293T cells was used as a negative control.
To check the specificity of this result, FLAG antibody which could recognize the FLAG in the SFB tag was used for supershift reactions. FLAG antibody retarded the protein-ssDNA complex to a higher molecular weight, for both human POT1 and SmedOB1, suggesting successful supershifts (Fig. 4C, lanes 5 and 9). Furthermore, a (TTAGCC) 8 mutant probe was used for incubation with the SmedOB1 proteins, no binding could be detected. suggesting the specificity of its telomere binding (Fig. 4D). Collectively, these results confirmed that SmedOB1 is required for the telomeric protein-DNA complex formation and it associates with telomere (TTAGGG) 8   SmedOB1 participates in the DNA damage response and cell survival. POT1 is important for controlling telomere length and protecting telomere ends. The loss of human POT1 results in DNA-damage responses at the telomere ends, telomere lengthening, cell cycle arrest, and cell death [18][19][20][21] . To better understand the role of SmedOB1 in these biological processes, we performed RNAi experiments in intact Schmidtea mediterranea (Fig. 2B,C). Then we performed the whole-mount immunofluorescence analysis using an antibody against the DNA damage marker γ H2AX. The SmedOB1 RNAi planarians exhibited a significantly increase in γ H2AX signals compared to the GFP RNAi control worms (Fig. 5A,B). In fact, these were the same areas that exhibited the tissue injuries (Fig. 2D). To further confirm this, we performed an alkaline comet assay using single cells from SmedOB1 RNAi and control planarians. Alkaline comet assay detects both single and double-strand DNA breaks. The SmedOB1 RNAi planarian cells showed significantly more comet tails than GFP RNAi cells (Fig. 5C,D). These results indicated that SmedOB1 likely plays a role in DNA damage response pathways.
The apoptosis signal is stimulated in amputated worms 22 . We next amputated the RNAi-treated worms and performed TUNEL assays. After regeneration, we detected a significant increase in the apoptotic signals in the SmedOB1-RNAi planarians compared to the control group (Fig. 5E,F). These results indicate that SmedOB1 regulates planarian homeostasis by inhibiting DNA damage and cell apoptosis.
Functional identification of DjOB1 in the planarian Dugesia japonica. The planarian Dugesia japonica also contains a TTAGGG telomere repeat sequence (Supplementary Fig. S1). Therefore, we investigated whether OB1 is conserved in the planarian species Dugesia japonica (Dj). A database search and analysis identified one Unigene 34129 23,24 , and we used PCR to clone a 508-bp partial DjOB1 sequence. We then generated dsRNAs based on this sequence (Fig. 6A) and conducted RNAi experiments in Dugesia japonica. After feeding, DjOB1 levels were dramatically reduced (Fig. 6B). Similar to the SmedOB1 RNAi worms, we observed decreased expression of the proliferation marker PCNA (Fig. 6B), and most of the DjOB1-RNAi Dugesia japonica showed tissue injuries to the head eyespots, pharynx or both, whereas the GFP-control-treated worms were completely normal (Fig. 6C,D). These results suggest that the planarian OB1 gene is functionally conserved among planarian species.

Discussion
Pot1 is an essential gene, but deletion or mutation of Pot1 is associated with various phenotypes in different species. For example, deletion of Pot1 in S.pombe leads to rapid telomere loss, while the inhibition of POT1 in humans increases telomere length 8 . Thus, when we identified a POT1 homologue in planaria called SmedOB1, it was of interest to determine whether SmedOB1 regulates telomeres in planaria and to examine the phenotypes of planaria treated with RNAi. Our results showed that the lack of SmedOB1 in the flatworm caused tissue injuries and regeneration defects at the individual level. Thus, SmedOB1 is required for planarian homeostasis and regeneration. Mechanistically, we hypothesized SmedOB1 may regulate ASC function by promoting ASC proliferation, and inhibiting DNA damage and cell apoptosis in a telomere-dependent manner. The expression pattern of SmedOB1 indicated that all of the cells in a planarian express the OB1 gene, but the RNAi phenotype was first observed in the areas around the pharynx and around the eyespots, where fewer ASCs are present 17 . We hypothesized that this difference occurred because when tissue is injured, the planarian ASC pool dispatches stem cell progenies to migrate to the injured locations, where they complete division as well as differentiation. Because the areas around the pharynx and the eyespots have few ASCs, injuries around these areas cannot heal quickly. SmedOB1 expression is not restricted to ASCs but is ubiquitous and observed in both ASCs and somatic cells; thus, it is important for whole-body homeostasis.
The planarian flatworm is an ideal model for studying regeneration and stem cells, and many hypotheses that were previously based on the findings of in vitro experiments can be confirmed using in vivo experiments in planaria. Further, because planaria have the same telomere repeat sequence as humans, we believe that planaria will also be suitable for telomere studies. We studied SmedOB1 in planaria because its sequence is conserved among all of the shelterin complexes, while no other telomeric proteins could be predicted in planarians through BLAST analysis of conserved domains. Here we found that SmedOB1 associates with the telomere TTAGGG sequence in vitro and is required for planarian homeostasis and regeneration. The phylogenetic and functional conservations of the SmedOB1 provide one of the mechanisms by which planarians maintain telomeres and genome stability to ensure their immortality. Interestingly, we also found that besides SmedOB1 itself, there may be a potential multi protein-telomere complex in planarian, suggesting other un-identified telomere associating proteins existing in this species (Fig. 4A). Further studies elucidating these proteins or complexes would make great efforts to this field.
SmedOB1 is the first telomere-associated protein identified in planaria so far. The telomere structure is complicated and the maintenance of telomere structure, length and function require multiple proteins as well as TERRA 5 . By using high-throughput biochemical methods such as proteomics of isolated chromatin segments (PiCH), DNA-protein affinity assays and quantitative proteomics, unknown planarian telomeric proteins will be identified in the near future. A complete telomeric protein network will provide additional insights regarding telomere regulation in planarian adult stem cell biology. Further studies will shed light on the regeneration medicine of humans. Methods Planarian cultures. The asexual strain of Schmidtea mediterranea was maintained in 1X Montjuic water at 20 °C as previously described 25 . A clonal line of the mixoploid asexual strain of Dugesia japonica was derived from a single worm collected from Beibei Mountain in Chongqing and was maintained in filtered tap water at 22 °C. The worms were fed with beef liver once per week and starved for at least one week prior to all of the experiments. For irradiation, the planarians were exposed to X-rays at 5 Gy or 10 Gy and then maintained in culture medium for 4 days before RNA extraction.
Identification and cloning of SmedOB1 and DjOB1. A BLAST-based reciprocal best-hit method, in combination with protein sequence alignment, was used to identify orthologous POT1 genes in planarians as previously reported 25 . Briefly, the human POT1 isoform 1 was used to perform a tBLASTn search in the Schmidtea mediterranea genome database SmedGD 2.0 (http://smedgd.stowers.org) with an e-value of 1e-08. The genomic sequence (V31.002347: 3945..3613) was then translated and compared with human sequences using NCBI BLAST tools.
To clone SmedOB1, RACE primers were designed accordingly, and 5′ RACE and 3′ RACE were performed using the FirstChoice ® RLM-RACE Kit (Thermo Scientific). The PCR products were TA-cloned using the Scientific RepoRts | 6:34013 | DOI: 10.1038/srep34013 488-conjugated secondary antibody and mounted using fluorescent mounting medium (Dako). Alkaline comet assay was performed according to the manufacturer's protocol (Trevigen). The DNA was stained by Gel-Red (Beyotime) and photographed in Zeiss SteREO Discovery V.20. The quantification of data was analysed in CASP software.
TUNEL assay. TUNEL assays using the planarians were performed as described previously 25 . Briefly, the worms were sacrificed in 10% n-acetyl cysteine (Sigma), fixed in 4% formaldehyde, and permeabilised in 1% SDS (for 20 min) before being bleached overnight in 6% H 2 O 2 (diluted in 1x PBST). Following the washes in 1xPBST and 1xPBS, the worms were incubated with the terminal transferase enzyme (Roche) for 1 h at 37 °C, and worms treated with DNase I (NEB) for 10 min at RT were used as positive controls. The planarians were incubated with DAPI for 10 min before mounting. The signals were photographed in Zeiss SteREO Discovery V.20. The quantification of data was analysed in ImageJ software.
Data analysis. All the statistic data are presented as means ± SDs. The statistical significance of differences between two means was calculated using Student's t-test. The probability level accepted for significance was P < 0.05 (*) and P < 0.01 (**).