Full-length zebrafish pin1 (NM_200748) was cloned using the SMART™ RACE cDNA amplification kit (BD Biosciences). 5′-Gene specific primer: 5′-GGCCGC​TCC​CAC​TGA​CTC​GCA​TTG​GT-3′; 3′-Gene specific primer: 5′-GAC​CCT​CGT​CCT​GGA​GAG​AGG​AGA​AC-3′.

Zebrafish were maintained at 28°C under standard conditions according to the rules of IACUC (Biopolis IACUC application #050096) and regulations of the Fish Facility of the IMCB. The embryos were staged as described (Kimmel et al., 1995).

RNA extraction from embryos at the stages indicated (Figure 2A) was performed using Trizol (Invitrogen). Subsequently, cDNA was synthesized using SuperScript reverse transcriptase (Invitrogen). The cDNAs from various embryonic stages served as templates for amplifying the zebrafish pin1 transcript. The Primers used were.

β-actin sense: 5′-GCA​CGA​GAG​ATC​TTC​ACT​CCC​CTT​G-3′;

Whole-mount in situ hybridization of zebrafish embryos was performed following the protocol described in the zebrafish book (Westerfield, 2000). Antisense/sense RNA probes were produced with a dioxygenin RNA labeling kit (Roche). For 48 hpf embryos, the hybridization temperature was reduced from 68°C to 58°C and the staining time was extended to 3–4 h. These modified conditions allowed specific detection of the zebrafish Pin1 transcript on the lateral line neuromasts.

Morpholino antisense oligonucleotides (Gene Tools Plc, United States) were injected at different concentrations ranging from 0.4 to 0.8 pmol in 1 × Danieau buffer (58 mM NaCl; 0.7 mM KCl; 0.4 mM MgSO4; 0.6 mM Ca(NO₃)₂; 5.0 mM Hepes pH 7.6). Needles for microinjection were prepared using the Sutter Micropipette puller P-97 (Sutter Instruments Co, United States). Microinjection was performed on 1-2 cell stage embryos using Picoinjector PLI-100 (Medical Systems Corp, Greenvale, NY, United States). MO sequences: pin1 MO1: 5′-ACG​GCA​GCT​TCT​CGT​CAT​CGG​ACA​T-3′; pin1 MO2: 5′-GAT​TGC​AGG​ACG​GCT​CGG​TTC​GG-3′; Standard control MO: 5′-CCT​CTT​ACC​TCA​GTT​ACA​ATT​TAT​A-3′.

Cell lines used: wild-type HEK 293T and HEK 293T stably expressing control siRNA and Pin1 siRNA, wild-type MEF and Pin1^−/− MEF cells and SH-SY5Y cells. HEK 293T and MEF cells were maintained at 37°C and 5% CO₂ in Dulbecco’s Modified Eagle Medium (DMEM) with 10% FBS (Hyclone), 2% 750 g/L NaHCO₃ (Gibco), 100 U/mL penicillin G (Gibco) and 100 μg/mL streptomycin sulfate (Gibco). SH-SY5Y cells were grown under similar conditions in a 1:1 mixture of Eagle’s Minimum Essential Medium (MEM, Hyclone) with non-essential amino acids and Ham’s F12 medium (Hyclone) supplemented with 10% FBS (Hyclone). HEK 293T cells were transfected by calcium phosphate method. MEF and SH-SY5Y cells were transfected with Lipofectamine™ 2000 (Invitrogen) according to manufacturer’s protocol.

The stable Pin1 siRNA expression is a retrovirus-mediated RNA interference targeting Pin1. The following has been added into M&M. Establish stable Pin1-siRNA HEK 293T cells. The retrovirus was generated as described in (Ryo et al., 2005). In brief, PLAT-E cells were transfected with pSuper-puro-Pin 1-ShRNA or control -ShRNA and vesicular stomatitis virus G (VSVG) vectors (kindly provided by Dr. Ryo, 2005), using Lipofectamine 2000 (Invitrogen). Culture supernatants of PLAT-E cells were collected 48 h following transfection with retroviral vectors. HEK-293T cells were infected in the presence of 10 ug/mL Polybrene. The stable clones were selected by continuous growth in 2 μg/mL puromycin (Sigma #P8833).

SH-SY5Y cells were transfected with GFP-tagged zebrafish Pin1 and HA-Nrd and seeded on coverslips. After 48 h, the cells were fixed with 4% paraformaldehyde/PBS for 30 min and blocked with 3% BSA; 0.1% Triton/PBS for 30 min at room temperature. Subsequently, the cells were incubated with polyclonal anti-GFP and monoclonal anti-HA antibodies (Santa Cruz Biotechnology) for 1 h at room temperature, followed by Alexa-Flour-488 anti-rabbit and Alexa-Fluor-568 anti-mouse antibodies (Molecular probes) for 1 h at room temperature. Nuclear staining were performed using Hoechst 33,342 (Molecular probes) for 10 min at room temperature. Finally, cells were then mounted with FluorSave™ reagent (Calbiochem) and examined under confocal fluorescence microscope (Zeiss META LSM510).

HA-tagged Nrd were co-transfected either with FLAG-tagged zebrafish Pin1 or FLAG-vector control. Cell lysates were collected in mammalian cell lysis buffer (50 mM Hepes pH 7.4; 10% glycerol; 1% Triton X-100; 100 mM NaCl and 0.5 mM MgCl₂) supplemented with protease and phosphatase inhibitors. Supernatants were incubated with FLAG-M2 beads (Sigma) for 1.5 h at 4°C. Bound proteins were eluted by 2 × SDS loading dye. The samples were analyzed on SDS-PAGE and Western blotting with polyclonal anti-HA antibody (Zymed) and monoclonal anti-FLAG antibody (Santa Cruz Biotechnology).

The recombinant GST or GST-tagged zebrafish Pin1 proteins were immobilized on Glutathione Sepharose 4B beads (Amersham Biosciences). The beads were then incubated with HEK 293T cell lysates overexpressing HA-tagged Nrd or Nrd mutants at 4°C for 3 h. Bound proteins were eluted by 2 × SDS loading dye and analyzed by Western blotting with polyclonal anti-HA antibody (Zymed).

For stability assays, HA-Nrd/HA-Nrd5A were transfected into HEK 293T control siRNA and HEK 293T Pin1 siRNA cells; or MEFs WT and Pin1^(−/−) cells, respectively. The cells were treated with cycloheximide (100 μg/mL; Sigma) after 24 h of transfection to inhibit de novo protein synthesis. The cells were then harvested using mammalian cell lysis buffer supplemented with protease and phosphatase inhibitors at the different time points. Equal amounts of total proteins for each time point were loaded onto SDS-PAGE and analyzed by Western blotting against monoclonal anti-HA antibody (Santa Cruz Biotechnology) and monoclonal anti α-tubulin antibody (Sigma-Aldrich).

To investigate if re-introduction of Pin1 could stabilize Nrd in Pin1 depleted cells, HA-Nrd was co-transfected with FLAG-tagged zebrafish Pin1 or infected with adenovirus encoding human Pin1 into HEK 293T control siRNA and Pin1 siRNA cells. The cells were treated with cycloheximide and harvested as above to be analyzed by Western blotting with monoclonal anti-HA (Santa Cruz Biotechnology), monoclonal anti-FLAG (Santa Cruz Biotechnology) and monoclonal antibody α-tubulin (Sigma Aldrich) antibodies.

Densitometry was performed on scanned immunoblot images using the ImageJ gel analysis tool.

To study pin1 function in zebrafish, a full-length zebrafish pin1 was cloned using Rapid Amplification of cDNA Ends (RACE) from total mRNA of adult zebrafish. The coding sequence of zebrafish pin1 consists of a 480-bp open reading frame, an 81-bp 5′UTR and a 342-bp 3′UTR (Figure 1A). The cDNA of zebrafish pin1 encodes a 159-amino acid protein that shares 79%, 79% and 78% identity with human, mouse and Xenopus Pin1, respectively (Figure 1B). Given the substantial similarity observed, it is probable that zebrafish Pin1, akin to other proteins within this family, will adopt a similar structural arrangement. This arrangement is expected to comprise a WW domain, serving as the binding/recognition domain, and a PPIase domain, functioning as the catalytic domain, which in turn will likely enable the isomerization of phosphorylated substrates.

Semiquantitative RT-PCR analysis of pin1 in zebrafish embryos showed the presence of this transcript as early as 1-cell stage, suggesting that pin1 is maternally deposited and remains relatively constant until 72 hpf (Figure 2A). Western blot analysis of Pin1 distribution in adult zebrafish tissues revealed that its ubiquity, with higher levels observed in the brain and testis (Figure 2B). In addition, whole-mount in situ hybridization (WISH) of pin1 in zebrafish embryos showed that although the expression was ubiquitous early on, it became largely restricted to the developing brain by 24 hpf and 48 hpf (Figures 2C–E). These expression results are consistent with those in the previous report (Ibarra et al., 2017). The expression of pin1 is also detected in the PLL neuromasts at 48 hpf (Figure 2F). Notably, sense pin1 mRNA transcripts were not detected in the PLL neuromasts of 48 hpf embryos (Supplementary Figure S1). The specific expression of Pin1 in proliferating cells and neuromasts prompted us to investigate the role of Pin1 in neuronal development.

Two anti-sense morpholino oligonucleotides (MOs) were designed and used to target the translation initiation region (pin1 MO1) and the 5′–UTR region (pin1 MO2) of the pin1 transcripts, aiming to knockdown Pin1 function in zebrafish embryos (Figure 3A). We firstly conducted an initial screen using both MO1 (Figure 3B) and MO2 (Supplementary Figure S2). Briefly, we observed that all zPin1 MO-injected embryos displayed developmental delay, and this delay was found to be dose-dependent. Increasing the injected morpholino amount from approximately 0.3 pmol–0.6 and even 0.8 pmol resulted in more evident phenotypes of developmental delay. Although both pin1 MOs, but not the standard control MO, reduced the level of Pin1 in zebrafish embryos (Figure 3B; Supplementary Figure S2). However, when the injected morpholino amount reached 0.8 pmol, all embryos exhibited toxic effects with global apoptosis, malformed brain, and edema. As a result, we had to limit the amount of morpholino to a maximum of 0.6 pmol. On the other hand, MO2 did not exhibit as pronounced of a knockdown effect at 0.6 pmol; therefore, we primarily used MO1 to study the phenotypes. MO2 was mainly utilized as a morpholino specificity control.

Depletion of Pin1 in zebrafish resulted in severe developmental delay of up to 2 h in 24 hpf embryos and 6 h in 48 hpf embryos (Figures 3C–J). Since p53 is a known substrate of Pin1, the developmental delay observed in morphant embryos could be attributed to the role of Pin1 in cell cycle regulation and apoptosis (Mantovani et al., 2004; Becker and Bonni, 2007; Ryo et al., 2007). In addition, p53 is also involved in some off-target effects of MO. Concurrent knockdown of p53 specifically attenuates the off-targeting cell death induced by MO while p53 MO does not affect specific loss of gene function (Robu et al., 2005). To eliminate the potential secondary effects of p53, subsequent analyses using pin1 MOs were performed in the presence of p53 MO (Eisen and Smith, 2008).

Co-injection of pin1 MO/p53 MO resulted in developmental delay of 2 h in 24 hpf embryos and reduced the developmental delay to 4 h in 48 hpf embryos (Figures 4A, B, D, E) without any hint of off-target effects associated with MOs. Importantly, morphants displayed a reduced number of posterior neuromasts (Figures 4C, F), suggesting a role of Pin1 in neuromasts formation. We next proceeded to knock down pin1 in two Tol2 enhancer trap lines, ET4 and ET20, with GFP expression in the hair cells and mantle cells, respectively (Parinov et al., 2004). The pin1 MO/p53 MO injected ET4 fish displayed a marked decrease in the number of anterior neuromast hair cells, and a complete loss of these cells along the PLL at 48 hpf (Figure 4I). However, ET20 injected with pin1 MO/p53 MO displayed the full complement of neuromasts mantle cells (Figure 4L). In both ET lines, p53 MO had no observable effects in neuromasts development (Figures 4H, K).

The specific loss of hair cells, but not mantle cells by 72 hpf was clearly demonstrated when pin1 MO/p53 MO were injected into double transgenic embryos derived from crossing ET4 and ET20 (ET4x20, Figures 4M–P). Furthermore, WISH for atoh1a transcripts at 36 and 48 hpf showed that both atoh1a positive proneuromasts and deposited neuromasts were present in Pin1 morphants (Supplementary Figure S3), suggesting that initial specification of neuromasts was not affected by the attenuation of Pin1 function. Taken together, the data suggest that Pin1 is required for the specification of hair cells, but not the mantle cells. Similar to pin1, loss of function of Nrd specifically resulted in the loss of hair cells in zebrafish (Sarrazin et al., 2006). Taken together, our data suggest a possible link between Pin1 and Nrd activity in regulating PLL hair cells formation in zebrafish.

To evaluate whether neuron determination process was affected, we utilized glial fibrillary acidic protein (GFAP) antibody to stain embryos. GFAP is an intermediate filament primarily expressed in astrocytes and radial glial cells of the central nervous system (CNS). Radial glial cells are neural stem cells from which most of neurons in brain are derived, either directly or indirectly. Supplementary Figure S4A and B showed an indicatical staining pattern, indicating that early neuron specification or determination remained unaffected. However, we suspected that Pin1 might interfere with neuron differentiation process. To further characterize zPin1 knockdown neuronal phenotypes, we performed in situ hybridization analysis using several markers, including neurogenin1 (ngn1), neuroM, her4, neuroD. The choice of neuroD for analysis was due to the observed interaction between zPin1 and NeuroD, as well as the key role of neuroD as a neuron differentiation factor. The in situ hybridization results, presented in Supplementary Figure S4, showed a defective neuroD expression pattern, particularly with the loss of neuroD transcripts, but not other markers, in the midbrain and hindbrain region. These findings suggest that zPin1 may impact neuroD expression, thereby influencing the process of neuron differentiation.

Since Pin1 specifically regulates the phosphorylated proteins through their pSer/pThr-Pro motifs, to investigate the potential Pin1 binding sites of Nrd, we performed sequence analysis of Nrd and identified five potential Pin1-binding Ser/Thr-Pro motifs (asterisks, Figure 5A). Among these motifs, the consensus Ser157, Ser259, and Ser267 residues (highlighted in red, Figure 5A) were known to be phosphorylated by ERK2 or GSK3β in Xenopus (Marcus et al., 1998; Moore et al., 2002; Khoo et al., 2003). This led us to hypothesize that Pin1 could interact physically with Nrd to regulate its function during differentiation of mechanoreceptors. To test this hypothesis, we co-transfected GFP-tagged zebrafish Pin1 and HA-tagged Nrd were transiently co-transfected into neuroblastoma SH-SY5Y cells. The results, as shown in Figure 5B, revealed that NeuroD was predominantly localized in nucleus (Figure 5B, labelled in red). As for Pin1 (Figure 5B, labelled in green), the images showed mostly of GFP-Pin1 resided in nucleus, with some signal being detected in cytoplasm. The merged image revealed that these two proteins co-localized in the nucleus, suggesting that they had the required proximity to interact.

Subsequently, we conducted an in vivo co-immunoprecipitation (Co-IP) FLAG pulldown experiment was performed. HEK 293T cells were co-transfected with FLAG-zPin1 and HA-NeuroD, and then we performed the FLAG pulldown assay using harvested whole cell lysates. We utilized anti-FLAG and anti-HA antibodies to detect zPin1 and NeuroD, respectively. As depicted in Figure 5C (arrow), our results demonstrated that FLAG-zPin1 successfully pulled down HA-NeuroD, indicating that an in vivo interaction of between zPin1 and NeuroD. This suggests that zPin1 may play a role in post-translational regulation of NeuroD and forming the basis for all our subsequent studies. Next, we utilized recombinant glutathione S-transferase (GST)-zebrafish Pin1 to further demonstrate the interaction. The results showed that GST-Pin1 could bind to HA-tagged Nrd overexpressed in HEK293T cells, providing evidence of an in vitro interaction between zebrafish Pin1 and Nrd (Figures 5C, D).

To further investigate the phosphorylation-dependent interaction between Pin1and Nrd via Ser/Thr-Pro motifs, we next conducted a screening using site-directed mutagenesis. Substituting all five Ser-Thr/Pro motifs in Nrd with alanine (HA-Nrd5A) significantly attenuated zebrafish Pin1 binding, indicating a motif specific interaction between zebrafish Pin1 and Nrd (Figure 5E). Notably, substantial binding between Pin1 and Nrd persisted even when up to four of the Ser/Thr-Pro motifs in Nrd were substituted with alanine (Supplementary Figure S5). In addition, the interaction between zebrafish Pin1 and HA-Nrd was reduced in HEK 293T cell lysates treated with calf intestine alkaline phosphatase (CIAP) (Supplementary Figure S6). Taken together, these results collectively demonstrate that zebrafish Pin1 interacts with Nrd via all five Ser/Thr-Pro motifs in Nrd and that this interaction is phosphorylation dependent.

The interaction between Pin1 and its substrates has been demonstrated to be crucial in regulating substrates stability and turnover (Lu and Zhou, 2007). To explore the impact of Pijn1 on Nrd, we next examined the stability of HA-Nrd in HEK 293T cells with stably expressing human Pin1 siRNA and control siRNA, treated with cycloheximide. HA-Nrd in Pin1-depleted HEK 293T cells showed reduced stability with a higher turnover rate compared to control cells (Figure 6A). Similar results were observed in Pin1 knockout MEF cells (Supplementary Figure S7A). Additionally, the stability of HA-Nrd5A, which has compromised for binding with zebrafish Pin1, was protected from degradation in the absence of Pin1 (Figures 6B, E). We observed slight variations in the levels of Pin1 (under siRNA treatments) between Figures 6A, B, which can be attributed to differences in the exposure time of the films during experimentation. However, it is important to note that these variations fall within an acceptable range of variation. Overall, our results and conclusions remain consistent. Importantly, the stability of HA-Nrd was rescued by re-introduction of FLAG-tagged zebrafish Pin1 (Figures 6C, D) and adenovirus encoding human Pin1 (Supplementary Figure S7B) in HEK 293T Pin1 siRNA cells. These findings indicate that Pin1 plays a direct and critical role in modulating Nrd stability. Taken together, our data suggest that Pin1 functions in Nrd dependent neuronal specification events by regulating Nrd stability.