SAPCD2 phosphorylation modulates astral microtubule stability to control spindle orientation Authors: Shuang Chen, Ying Hong, Yaping Fang, Mingxuan Li, Jianhong Xia, Jikai Wen, Feiyang Dai, Yuekai Li, Wandi Bu, Michael S. Y. Huen, Jun Jiang, Yiqun Deng Published: 2026 ABSTRACT Proper spindle orientation is essential for organogenesis and tissue homeostasis. Cortical tethering complexes, such as the Gαi-LGN-NuMA complex, anchor astral microtubules to guide spindle orientation. However, how astral microtubule stability is coupled to mitosis-dependent regulatory mechanisms remains unclear. Here, we identified suppressor APC domain-containing 2 (SAPCD2) as a cell cycle-regulated factor that functionally links astral microtubule stability to spindle orientation control, utilizing non-polarized HeLa and HaCaT cell culture models. Live-cell imaging revealed that siRNA-mediated depletion of SAPCD2 induces spindle misorientation, leading to mitotic delay and aberrant cell division. Mechanistically, SAPCD2 acts as a previously uncharacterized microtubule-associated protein that promotes microtubule stability, whereas CDK1-mediated phosphorylation at S157, together with phosphorylation at S276, generates a mitosis-specific phosphorylated isoform that destabilizes astral microtubules. Importantly, disruption of SAPCD2 phosphorylation at both S157 and S276 fails to rescue the spindle misorientation caused by SAPCD2 depletion. These findings deepen our understanding of how mitosis-dependent dynamic phosphorylation of microtubule-associated proteins orchestrates astral microtubule behavior to ensure accurate spindle orientation. Introduction In multicellular organisms, precise regulation of mitotic spindle orientation is crucial for cell fate determination, epithelial tissue maintenance, and tissue homeostasis, by determining the plane of cell division1. In polarized epithelia, polarity cues direct the spindle to orient either along the apical-basal axis for perpendicular division or parallel to the basal layer for planar cell division, thereby maintaining epithelial morphogenesis and tissue homeostasis2,3. Extensive research has demonstrated that the evolutionarily conserved Gαi-LGN-NuMA-dynein complex, where cortical Gαi-bound LGN links to astral microtubules via NuMA-dynein, is essential for spindle-cortex attachment and orientation control4–6. Except for Gαi, an increasing number of polarity cues, such as Dlg1, Par3 and Afadin have been found to guide the lateral cortical localization of LGN7–9. The mechanism underlying spindle orientation is highly intricate, and even the same protein may exhibit distinct regulatory mechanisms for spindle orientation across different cell types or tissue contexts. For example, during mouse epidermal stratification, Par3 recruits LGN to the apical domain, thereby driving perpendicular division through the LGN-NuMA-dynein-dynactin complex, while in polarized Madin-Darby canine kidney cell cysts, Par3 guides the lateral cortical localization of LGN, facilitating planar cell division8,10. Due to the greater complexity and tissue specificity in vertebrate systems, new mechanisms governing spindle orientation remain to be uncovered. Astral microtubules, a subpopulation of microtubules that emanate from centrosomes and extend to the cell cortex without kinetochore attachment, serve as the central effector for mitotic spindle orientation11. In most known mechanisms, astral microtubules mediate spindle orientation by transferring cortical forces via dynamic cortex capture (e.g., via NuMA-dynein)12. However, in non-polarized cells (e.g., HeLa cells cultured on fibronectin-coated substrates), proper mitotic spindle orientation depends on astral microtubules and actin retraction fibers, with LGN or NuMA depletion causing only mild spindle misorientation13,14. Accumulating evidence supports that astral microtubules can mediate the recruitment of force generators to the cell cortex independently of LGN, thereby ensuring proper spindle orientation15–18. Astral microtubules are dynamically unstable polymers during mitosis11. Numerous microtubule-associated proteins (MAPs), such as microtubule-associated protein CLASP1, microtubule plus-end-tracking proteins ( + TIPs) EB1, and the depolymerase KIF2, collectively modulate cortical microtubule interactions by regulating the dynamics and stability of astral microtubules, thereby controlling spindle orientation19–21. Whether other microtubule regulators with known dynamic functions on microtubules may additionally control spindle orientation remains unexplored. SAPCD2 is a known oncogene and tumor marker that is highly expressed in various tumor tissues, such as gastric cancer, colorectal cancer, and breast cancer, and is associated with tumor prognosis22–25. It promotes tumor cell migration and invasion by modulating epithelial-mesenchymal transition and activating key signaling pathways, including Hippo, PI3K/Akt, and MAPK26,27. Moreover, SAPCD2 is a cell cycle-dependent gene, which is highly expressed in G1 and M phases28. Although SAPCD2 overexpression promotes mitotic progression and induces chromosome missegregation in normal fibroblasts, and its depletion causes multinucleation and mitotic arrest in cancer cells, the molecular mechanisms underlying its mitotic function remain unknown28,29. Chiu et al. found that SAPCD2 regulates cell division in retinal progenitor cells30. During mouse retinogenesis, SAPCD2 is recruited to the apical domain of polarized epithelial cells by polarity proteins, which restricts LGN to the basolateral cortex, ensuring spindle alignment with the basal membrane for planar cell division30. Notably, during early epidermal stratification, LGN does not regulate progenitor cell division and only weakly affects spindle orientation in non-polarized cells, suggesting the presence of alternative spindle orientation mechanisms in these systems that can compensate for LGN. Whether SAPCD2 regulates spindle orientation via additional mechanisms during early epidermal stratification or in non-polarized cells is currently unclear. Here, we employ HeLa cells and human epidermal keratinocytes (HaCaT) as in vitro non-polarized cell models and demonstrate that SAPCD2 associates with microtubules to regulate astral microtubule stability and spindle orientation. Mitosis-specific phosphorylation at S157 and S276 is critical for SAPCD2 to destabilize astral microtubules, thereby maintaining proper spindle orientation. Results SAPCD2 is required for proper mitotic progression and spindle orientation To explore the function of SAPCD2 in cell cycle regulation, we first synchronized HeLa and HaCaT cells at specific cell cycle phases using various arrest methods combined with morphological characterization, and detected the expression of SAPCD2 during cell cycle. As shown in Fig. 1A, B, SAPCD2 protein levels exhibited a significant increase during mitotic entry and a decrease after G1 phase in both HeLa and HaCaT cells. The cell cycle-dependent expression pattern observed in both epithelial cell lines is consistent with previous findings in gastric cancer BGC823 cells28. Notably, we observed a shifted band of SAPCD2 on SDS-PAGE during mitosis in both cell types (Fig. 1A, B). We next assessed SAPCD2 knockdown efficiency by Western blot analysis using a pooled siRNA, which effectively reduced endogenous SAPCD2 levels in HeLa cells (Fig.1C). To investigate SAPCD2’s mitotic role, we performed live-cell imaging of siRNA-treated HeLa cells, comparing control cells with those depleted of SAPCD2 (Fig. 1D). Although SAPCD2-depleted cells were able to complete mitotic progression, they exhibited a significant delay (Fig. 1D, E, Video S1 and Video S2). Compared with control siRNA, SAPCD2-depleted cells took longer to transit from nuclear envelope breakdown (NEBD) to anaphase onset (siSAPCD2-mix: 128.80 ± 8.08 min versus siControl: 69.35 ± 4.94 min) (Fig. 1D, E). In monolayer-cultured cells, division orientation is established during prometaphase and maintained through mitosis, with the division axis typically aligned parallel to the extracellular matrix-coated substrate, ensuring proper planar orientation of daughter cells31. However, SAPCD2 depletion led to randomized division orientation throughout mitosis, resulting in daughter cells that partially overlapped rather than aligning within the monolayer plane, indicative of abnormal division orientation (Fig. 1D, F).Fig. 1SAPCD2 regulates proper mitotic progression and spindle orientation.A, B Western blot analysis of SAPCD2 protein level at different phases of the cell cycle in HeLa cells (A) or HaCaT cells (B). Asyn, asynchronous cells. C Western blot analysis of HeLa cells transfected with a pool of two individual SAPCD2 siRNAs (siRNA-1 and siRNA-2). D Representative images of HeLa cells treated with control siRNA or SAPCD2 siRNA at the indicated times by live imaging. a and b indicate the two daughter cells produced by division. The black dashed line indicates the direction of cell division. Scale bar, 10 µm. E, F Quantifications from the same set of live-cell imaging experiments shown in (D). E Scatter plots with the time from NEBD to anaphase onset in live HeLa cells treated with control siRNA (n = 55) or SAPCD2 mixed siRNAs (n = 63). Data were obtained from five independent experiments, and the n values indicate the total number of cells analyzed. The data are represented as mean ± SEM. ***P < 0.001, Student’s t-test. F Quantification of the misorientated division of HeLa cells imaged in (D). The data are represented as mean ± SD (n = 5), from over three independent experiments. ***P < 0.001, Student’s t test. G Scheme of the normal or abnormal spindle orientation at metaphase. Orthogonal views (x-z) of the mitosis cells with the spindle angle α were analyzed by ImageJ. H Representative images of confocal z-sections of metaphase HeLa cells treated with control siRNA or SAPCD2 siRNAs stained for γ-tubulin (green) and DAPI (blue). Scale bars, 5 µm. I Scatter plots and J radial histograms of the distributions of the spindle angles in HeLa cells treated with control siRNA (n = 115) or SAPCD2 siRNAs (siSAPCD2-1, n = 83; siSAPCD2-2, n = 72), from three independent experiments. The data are represented as mean ± SEM. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. K Western blot analysis of HeLa cells treated with control siRNA or two individual SAPCD2 siRNAs. SAPCD2 regulates proper mitotic progression and spindle orientation. A, B Western blot analysis of SAPCD2 protein level at different phases of the cell cycle in HeLa cells (A) or HaCaT cells (B). Asyn, asynchronous cells. C Western blot analysis of HeLa cells transfected with a pool of two individual SAPCD2 siRNAs (siRNA-1 and siRNA-2). D Representative images of HeLa cells treated with control siRNA or SAPCD2 siRNA at the indicated times by live imaging. a and b indicate the two daughter cells produced by division. The black dashed line indicates the direction of cell division. Scale bar, 10 µm. E, F Quantifications from the same set of live-cell imaging experiments shown in (D). E Scatter plots with the time from NEBD to anaphase onset in live HeLa cells treated with control siRNA (n = 55) or SAPCD2 mixed siRNAs (n = 63). Data were obtained from five independent experiments, and the n values indicate the total number of cells analyzed. The data are represented as mean ± SEM. ***P < 0.001, Student’s t-test. F Quantification of the misorientated division of HeLa cells imaged in (D). The data are represented as mean ± SD (n = 5), from over three independent experiments. ***P < 0.001, Student’s t test. G Scheme of the normal or abnormal spindle orientation at metaphase. Orthogonal views (x-z) of the mitosis cells with the spindle angle α were analyzed by ImageJ. H Representative images of confocal z-sections of metaphase HeLa cells treated with control siRNA or SAPCD2 siRNAs stained for γ-tubulin (green) and DAPI (blue). Scale bars, 5 µm. I Scatter plots and J radial histograms of the distributions of the spindle angles in HeLa cells treated with control siRNA (n = 115) or SAPCD2 siRNAs (siSAPCD2-1, n = 83; siSAPCD2-2, n = 72), from three independent experiments. The data are represented as mean ± SEM. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. K Western blot analysis of HeLa cells treated with control siRNA or two individual SAPCD2 siRNAs. Since spindle orientation determines the cell division axis, we further assessed the effect of SAPCD2 depletion using two independent SAPCD2 siRNAs on spindle orientation in HeLa cells by immunofluorescence and confocal microscopy. Z-stack imaging was used to quantify the spindle angle relative to the substrate (Fig. 1G). In metaphase, spindle poles were aligned at the same Z-position in control siRNA-treated cells, whereas SAPCD2-depleted cells displayed spindle poles at different Z-positions (Fig. 1H). SAPCD2-depleted cells exhibited randomly oriented mitotic spindles, whereas control cells predominantly maintained spindle alignment within 10° of the substrate plane (Fig. 1I, J). Although the two independent SAPCD2 siRNAs showed weaker knockdown efficiency than the pooled siRNA, both caused abnormal spindle orientation in HeLa cells (Fig. 1C, J, K). Consistent with the results in HeLa cells, SAPCD2 depletion in HaCaT cells similarly induced abnormal spindle orientation (Fig. S1A–D). Collectively, these findings demonstrate that SAPCD2 is essential for proper mitotic progression and spindle orientation. SAPCD2 is a novel microtubule-associated protein To investigate how SAPCD2 functions in mitosis, we first examined its subcellular localization in fixed HeLa cells during interphase and mitosis. Immunofluorescence analysis showed that exogenous SAPCD2 exhibited a diffuse cytoplasmic signal, with bright puncta reminiscent of the typical localization pattern of centrosomal proteins during interphase, and transitioned to a spindle-associated distribution at metaphase (Fig. 2A). Rectangular line scan analysis along one side of the spindle (as indicated by the white arrow) further supports its spindle localization at metaphase, showing that SAPCD2 colocalized with the spindle marker α‑tubulin, whereas no such signal was detected in control cells (Fig. 2A–D). Consistent with these observations, immunostaining of endogenous SAPCD2 in HeLa cells also confirmed its co-localization with α-tubulin upon mitotic entry (Fig. S2A, B). Given its spindle localization during mitosis, we next performed microtubule co‑sedimentation assays to test whether SAPCD2 associates with microtubules. In both HeLa and HaCaT cells, endogenous SAPCD2 co-sedimented with polymerized microtubules at 37 °C, similar to the known microtubule-binding protein EB1 (Fig. 2E, F). In vitro microtubule co-sedimentation assays further showed that recombinant MBP-SAPCD2 also bound to polymerized microtubules (Fig. 2G). Together, these results demonstrate that SAPCD2 is a previously unrecognized microtubule-associated protein.Fig. 2SAPCD2 associates with microtubules.A, B HeLa cells stably expressing SAPCD2-Flag (A) or an empty vector control (B) were labeled with anti-Flag (red) antibody and anti-tubulin (green) antibodies. Cells transfected with the empty vector served as a negative control. Scale bars, 10 µm. C, D Fluorescence intensity profiles of Flag and α-tubulin in metaphase cells from (A) and (B) were generated along the rectangular line scan using ImageJ plot profile, with the white arrow indicating the direction of analysis. E, F Microtubule co-sedimentation assay of HeLa cells (E) and HaCaT cells (F). The pellet (P) obtained from incubation at 37 °C indicates the microtubule-bound fractions, and the supernatant (S) indicates the unbound fractions. Incubation at 4 °C serves as a microtubule-free control. G Microtubule co-sedimentation assay with purified tubulin and MBP-SAPCD2 or MBP, along with a control containing MBP-SAPCD2 alone (no tubulin), all incubated at 37 °C. SAPCD2 associates with microtubules. A, B HeLa cells stably expressing SAPCD2-Flag (A) or an empty vector control (B) were labeled with anti-Flag (red) antibody and anti-tubulin (green) antibodies. Cells transfected with the empty vector served as a negative control. Scale bars, 10 µm. C, D Fluorescence intensity profiles of Flag and α-tubulin in metaphase cells from (A) and (B) were generated along the rectangular line scan using ImageJ plot profile, with the white arrow indicating the direction of analysis. E, F Microtubule co-sedimentation assay of HeLa cells (E) and HaCaT cells (F). The pellet (P) obtained from incubation at 37 °C indicates the microtubule-bound fractions, and the supernatant (S) indicates the unbound fractions. Incubation at 4 °C serves as a microtubule-free control. G Microtubule co-sedimentation assay with purified tubulin and MBP-SAPCD2 or MBP, along with a control containing MBP-SAPCD2 alone (no tubulin), all incubated at 37 °C. SAPCD2 regulates astral microtubule stability to control spindle orientation To elucidate the role of SAPCD2 as a microtubule-associated protein, we assessed whether loss of SAPCD2 affects microtubule organization. Cold-induced microtubule depolymerization assays were performed in HeLa and HaCaT cells to evaluate interphase microtubule integrity and organization. Compared to control cells, SAPCD2-depleted cells exhibited significantly reduced microtubule intensity (Fig. S3A, E). Quantitative analysis further showed that SAPCD2 depletion significantly enhanced cold-induced microtubule depolymerization at all time points examined (Fig. S3C, G). To further assess microtubule regrowth potential following complete depolymerization, HeLa and HaCaT cells were incubated at 37 °C to allow microtubule regrowth over a time course, and SAPCD2 deficiency continued to inhibit microtubule regrowth even after 15 min of recovery (Fig. S3B, D, F, H). Next, we explored the impact of SAPCD2 silencing on mitotic spindle structure in cells with comparable size (Fig. 3A, B). Cell size was determined as the longest cell diameter parallel to the spindle pole-to-pole axis, as previously described32. Our results showed that SAPCD2-depleted HeLa cells exhibited significantly reduced astral microtubule intensity (Fig. 3A, C). This finding prompted us to further assess astral microtubule stability by measuring length and abundance, two well-established indicators33,34. Using EB1 immunofluorescence to label growing microtubule plus ends, we quantified EB1 comet number and the distance from plus ends to the centrosome, and found that loss of SAPCD2 significantly reduced astral microtubule number and length (Fig. 3D–F). Further analysis of microtubule abundance at different lengths revealed a reduced number of long astral microtubules in SAPCD2‑depleted cells (Fig. 3G). Our data demonstrate that loss of SAPCD2 compromises astral microtubule stability, leading to astral microtubule defects. Since MCAK serves as a key microtubule depolymerase that decreases microtubule stability during mitosis35,36, we next investigated whether SAPCD2 regulates astral microtubule stability via a mechanism involving MCAK (Fig. 3I). Compared with SAPCD2 depletion alone, simultaneous knockdown of both SAPCD2 and MCAK restored astral microtubule fluorescence intensity, suggesting that SAPCD2 modulates astral microtubule stability in an MCAK‑associated manner (Fig. 3H–J). Given the essential role of astral microtubules in regulating spindle orientation, we hypothesized that SAPCD2 depletion disrupts spindle orientation by impairing astral microtubule stability. Thus, we assessed whether MCAK depletion could rescue the abnormal spindle orientation induced by SAPCD2 loss, and spindle angle analysis revealed that simultaneous knockdown of SAPCD2 and MCAK restored proper spindle orientation (Fig. 3K). To further confirm this hypothesis, we treated SAPCD2-depleted HeLa cells with low-dose paclitaxel, a microtubule-stabilizing agent. Immunofluorescence analysis showed that paclitaxel restored astral microtubule fluorescence intensity in SAPCD2-depleted cells toward control levels (Fig. 3L, M). Moreover, statistical analysis of spindle angles revealed that paclitaxel treatment partially rescued spindle misorientation in SAPCD2-depleted cells (Fig. 3N). These results suggest that SAPCD2 ensures proper spindle orientation by stabilizing astral microtubules.Fig. 3SAPCD2 tunes astral microtubule stability and subsequently regulates proper spindle orientation.A Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. B Cell diameter of HeLa cells transfected with control siRNA or SAPCD2 siRNAs. The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 30), from three independent experiments. ns, not significant. Student’s t test. C Quantification of astral microtubule (MT) fluorescence intensity in the corresponding cells shown in (B). The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 30), from three independent experiments. ***P < 0.001, Student’s t test. D Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs stained for EB1 (green) and DAPI (blue). Scale bar, 10 µm. E, F Quantification of number of microtubules (E) and mean microtubule length (F) imaged in (D). The data are represented as mean ± SD (control siRNA, n = 5; siSAPCD2-mix, n = 5). *P < 0.05, ***P < 0.001, Student’s t-test. The experiment was independently repeated three times. G Astral microtubule number distributions as a function of length imaged in (D). The data are represented as mean ± SD (control siRNA, n = 5; siSAPCD2-mix, n = 5). The experiment was independently repeated three times. H Western blot analysis of HeLa cells treated with control siRNA, SAPCD2 mixed siRNAs, MCAK siRNA or combined SAPCD2 and MCAK siRNAs. * indicates non-specific band. I Representative immunofluorescence images of HeLa cells treated with control siRNA, SAPCD2 mixed siRNAs, MCAK siRNA or combined SAPCD2 and MCAK siRNAs stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. J Quantification of microtubule fluorescence intensity imaged in (I). The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 34; siMCAK, n = 26; siSAPCD2-mix + siMCAK, n = 34), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. K Scatter plots of the distributions of the spindle angles in HeLa cells. The data are represented as mean ± SEM (control siRNA, n = 77; siSAPCD2-mix, n = 74; siMCAK, n = 79; siSAPCD2-mix + siMCAK, n = 87), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. L Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs and SAPCD2-depleted cells treated with paclitaxel stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. M Quantification of astral microtubule fluorescence intensity imaged in (L). The data are represented as mean ± SEM (control siRNA, n = 24; siSAPCD2-mix, n = 35; siSAPCD2-mix + paclitaxel, n = 33), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. N Scatter plots of the distributions of the spindle angles in HeLa cells treated with control siRNA (n = 127) or SAPCD2 mixed siRNAs (siSAPCD2-mix, n = 126) and SAPCD2-depleted cells treated with paclitaxel (siSAPCD2-mix + paclitaxel, n = 111), from over three independent experiments. The data are represented as mean ± SEM. *P < 0.05, ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. SAPCD2 tunes astral microtubule stability and subsequently regulates proper spindle orientation. A Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. B Cell diameter of HeLa cells transfected with control siRNA or SAPCD2 siRNAs. The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 30), from three independent experiments. ns, not significant. Student’s t test. C Quantification of astral microtubule (MT) fluorescence intensity in the corresponding cells shown in (B). The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 30), from three independent experiments. ***P < 0.001, Student’s t test. D Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs stained for EB1 (green) and DAPI (blue). Scale bar, 10 µm. E, F Quantification of number of microtubules (E) and mean microtubule length (F) imaged in (D). The data are represented as mean ± SD (control siRNA, n = 5; siSAPCD2-mix, n = 5). *P < 0.05, ***P < 0.001, Student’s t-test. The experiment was independently repeated three times. G Astral microtubule number distributions as a function of length imaged in (D). The data are represented as mean ± SD (control siRNA, n = 5; siSAPCD2-mix, n = 5). The experiment was independently repeated three times. H Western blot analysis of HeLa cells treated with control siRNA, SAPCD2 mixed siRNAs, MCAK siRNA or combined SAPCD2 and MCAK siRNAs. * indicates non-specific band. I Representative immunofluorescence images of HeLa cells treated with control siRNA, SAPCD2 mixed siRNAs, MCAK siRNA or combined SAPCD2 and MCAK siRNAs stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. J Quantification of microtubule fluorescence intensity imaged in (I). The data are represented as mean ± SEM (control siRNA, n = 28; siSAPCD2-mix, n = 34; siMCAK, n = 26; siSAPCD2-mix + siMCAK, n = 34), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. K Scatter plots of the distributions of the spindle angles in HeLa cells. The data are represented as mean ± SEM (control siRNA, n = 77; siSAPCD2-mix, n = 74; siMCAK, n = 79; siSAPCD2-mix + siMCAK, n = 87), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. L Representative immunofluorescence images of HeLa cells treated with control siRNA or SAPCD2 mixed siRNAs and SAPCD2-depleted cells treated with paclitaxel stained for α-tubulin (green) and DAPI (blue). Scale bar, 10 µm. M Quantification of astral microtubule fluorescence intensity imaged in (L). The data are represented as mean ± SEM (control siRNA, n = 24; siSAPCD2-mix, n = 35; siSAPCD2-mix + paclitaxel, n = 33), from three independent experiments. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. N Scatter plots of the distributions of the spindle angles in HeLa cells treated with control siRNA (n = 127) or SAPCD2 mixed siRNAs (siSAPCD2-mix, n = 126) and SAPCD2-depleted cells treated with paclitaxel (siSAPCD2-mix + paclitaxel, n = 111), from over three independent experiments. The data are represented as mean ± SEM. *P < 0.05, ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. SAPCD2 undergoes mitosis-specific phosphorylation Studies have demonstrated that proper regulation of astral microtubule stability is critical for ensuring correct spindle orientation during mitosis33,37. Either excessive length or insufficient formation of astral microtubules can lead to abnormal spindle positioning. Given that SAPCD2 is highly expressed during mitosis, we investigated how it contributes to proper astral microtubule stability while preventing excessive stabilization. Our earlier results revealed that highly expressed SAPCD2 exhibited a mobility shift on SDS-PAGE gel during mitosis, suggesting potential post-translational modifications (Fig. 1A, B). Considering the extent of molecular weight changes, we assessed phosphorylation, acetylation, lactylation, and glycosylation of SAPCD2. Western blot and immunoprecipitation experiments indicated that SAPCD2 was modified by phosphorylation at Ser/Thr sites in both asynchronous and mitotic cells (Fig. 4A). We next treated the lysates from mitotic cells with thermosensitive alkaline phosphatase and found that the mobility shift of SAPCD2 was inhibited by dephosphorylation (Fig. 4B). These results confirm that SAPCD2 undergoes Ser/Thr phosphorylation during mitosis, which is responsible for the mitotic mobility shift of SAPCD2.Fig. 4SAPCD2 is phosphorylated during mitosis.A Immunoprecipitation of exogenous SAPCD2 to detect its post-translation modifications in asynchronous and mitotic HeLa cells. M, mitotic cells. B Western blot analysis of SAPCD2 modifications in mitotic HeLa cells following incubation with Fast AP at the indicated time. Fast AP, fast thermosensitive alkaline phosphatase. C Phos-tag gel analysis of SAPCD2 phosphorylation state at different phases of the cell cycle in HeLa cells. Black arrows highlight the two weaker mitosis-specific phosphorylated SAPCD2 bands. White arrow indicates the phosphorylated SAPCD2 band persisting into the early G1 phase. Magenta arrow highlights the major mitosis-specific phosphorylated SAPCD2 band. Blue arrow indicates the unphosphorylated SAPCD2 band. D Western blot analysis of SAPCD2 mobility shift in HeLa cells treated with inhibitors RO3306 or BI2536. E Immunoprecipitation of exogenous CDK1 to detect its interaction with SAPCD2. F In vitro kinase assays using ATPγS and CDK1/Cyclin B1 were performed to assess thiophosphorylation (Thiop) of MBP-SAPCD2. The magenta arrow indicates the thiophosphorylatd MBP-SAPCD2 and the black arrow indicates the thiophosphorylatd SAPCD2. The magenta asterisks indicate non-specific bands. G Phos-tag gel analysis of SAPCD2 mitosis phosphorylation state in HeLa cells treated with 5 μΜ RO3306. The black arrows highlight the two weaker mitosis-specific phosphorylated SAPCD2 bands. White arrow indicates the phosphorylated SAPCD2 band persisting into the early G1 phase. The magenta arrow highlights the major mitosis-specific phosphorylated SAPCD2 band. Blue arrow indicates the unphosphorylated SAPCD2 band. H Mass spectrometry identified the major specific phosphorylation sites of SAPCD2 during mitosis. The amino acids highlighted in magenta represent potential phosphorylation sites of SAPCD2. I In vitro kinase assays using ATPγS and CDK1/Cyclin B1 were performed to assess the thiophosphorylation (Thiop) of wild-type MBP-SAPCD2 and its phosphorylation-site mutants (S157A, S276A, and 2 A). J Immunoprecipitation of wild-type SAPCD2 (SAPCD2-WT) and SAPCD2-2A mutant to assess phosphorylation in mitotic HeLa cells. SAPCD2 is phosphorylated during mitosis. A Immunoprecipitation of exogenous SAPCD2 to detect its post-translation modifications in asynchronous and mitotic HeLa cells. M, mitotic cells. B Western blot analysis of SAPCD2 modifications in mitotic HeLa cells following incubation with Fast AP at the indicated time. Fast AP, fast thermosensitive alkaline phosphatase. C Phos-tag gel analysis of SAPCD2 phosphorylation state at different phases of the cell cycle in HeLa cells. Black arrows highlight the two weaker mitosis-specific phosphorylated SAPCD2 bands. White arrow indicates the phosphorylated SAPCD2 band persisting into the early G1 phase. Magenta arrow highlights the major mitosis-specific phosphorylated SAPCD2 band. Blue arrow indicates the unphosphorylated SAPCD2 band. D Western blot analysis of SAPCD2 mobility shift in HeLa cells treated with inhibitors RO3306 or BI2536. E Immunoprecipitation of exogenous CDK1 to detect its interaction with SAPCD2. F In vitro kinase assays using ATPγS and CDK1/Cyclin B1 were performed to assess thiophosphorylation (Thiop) of MBP-SAPCD2. The magenta arrow indicates the thiophosphorylatd MBP-SAPCD2 and the black arrow indicates the thiophosphorylatd SAPCD2. The magenta asterisks indicate non-specific bands. G Phos-tag gel analysis of SAPCD2 mitosis phosphorylation state in HeLa cells treated with 5 μΜ RO3306. The black arrows highlight the two weaker mitosis-specific phosphorylated SAPCD2 bands. White arrow indicates the phosphorylated SAPCD2 band persisting into the early G1 phase. The magenta arrow highlights the major mitosis-specific phosphorylated SAPCD2 band. Blue arrow indicates the unphosphorylated SAPCD2 band. H Mass spectrometry identified the major specific phosphorylation sites of SAPCD2 during mitosis. The amino acids highlighted in magenta represent potential phosphorylation sites of SAPCD2. I In vitro kinase assays using ATPγS and CDK1/Cyclin B1 were performed to assess the thiophosphorylation (Thiop) of wild-type MBP-SAPCD2 and its phosphorylation-site mutants (S157A, S276A, and 2 A). J Immunoprecipitation of wild-type SAPCD2 (SAPCD2-WT) and SAPCD2-2A mutant to assess phosphorylation in mitotic HeLa cells. To gain further insights into SAPCD2 phosphorylation dynamics, we performed phos-tag gel electrophoresis to analyze SAPCD2 phosphorylation at different cell cycle phases in HeLa cells. As cells transitioned into mitosis, SAPCD2 exhibited at least four distinct phosphorylated bands, with three being unique to the mitotic phase (the bottom band represents the major phosphorylation form in M phase) and one persisting into the early G1 phase (Fig. 4C). These suggest that SAPCD2 undergoes multisite phosphorylation, hyperphosphorylation and potentially sequential phosphorylation events during the mitotic phase. Next, we sought to identify the kinase responsible for mitotic phosphorylation of SAPCD2. Treatment with CDK1 inhibitor (RO3306) but not PLK1 inhibitor (BI2536) prevented the mobility shift of SAPCD2 on SDS-PAGE during mitosis, suggesting CDK1 as the primary kinase (Fig. 4D). Immunoprecipitation further confirmed a direct interaction between SAPCD2 and CDK1 (Fig. 4E). To further confirm CDK1-mediated phosphorylation of SAPCD2, we performed in vitro kinase assays under two conditions. MBP-SAPCD2 was thiophosphorylated in the presence of ATPγS and CDK1/Cyclin B1, but not MBP alone (Fig. 4F). In a separate assay using ATP instead of ATPγS, phos-tag analysis also showed phosphorylation of SAPCD2, which was significantly reduced by the CDK1 inhibitor RO3306 (Fig. S4). These results confirm that CDK1 directly mediates the phosphorylation of SAPCD2 during mitosis. To identify CDK1-mediated phosphorylation sites on SAPCD2, phos-tag gel was first performed to analyze the phosphorylation of SAPCD2 after treatment with RO3306. Interestingly, inhibition of CDK1 only suppressed the three M-phase-specific bands but had no effect on the band that appeared early in G1 phase (Fig. 4G). Meanwhile, RO3306 treatment also completely abolished the band shift of SAPCD2 during mitosis (Fig. 4G). These results further validate that SAPCD2 is phosphorylated by CDK1 specifically in M phase, accounting for its mitotic mobility shift. Further mass spectrometry analysis of the major mitosis-specific phosphorylated bands revealed that SAPCD2 undergoes phosphorylation at the S157 and S276 residues (Figs. 4H and S5A). To further validate the CDK1-targeted phosphorylation sites on SAPCD2, we generated phosphorylation-deficient mutants of SAPCD2, including S157A, S276A, and the double mutant (S157A/S276A, hereafter referred to as 2 A). Mutation of S157 significantly impaired CDK1-mediated thiophosphorylation of SAPCD2, with the S157A mutant showing similar levels to the 2 A mutant, whereas mutation at S276 alone had little effect, indicating that S157 serves as a CDK1 phosphorylation site on SAPCD2 (Fig. 4I). Notably, CDK1-mediated thiophosphorylation was not entirely abolished in either the S157A or the 2 A mutant, indicating that SAPCD2 harbors additional phosphorylation sites targeted by CDK1 (Fig. 4I). Additionally, we performed immunoprecipitation and phos-tag gel analysis to assess the phosphorylation of these sites during mitosis. Compared to wild-type SAPCD2, phosphorylation of the 2 A mutant was markedly decreased during mitosis (Figs. 4J and S5B, C). These results demonstrate that SAPCD2 predominantly undergoes mitosis-specific phosphorylation at S157 and S276. SAPCD2 mitosis-specific phosphorylation ensures accurate spindle orientation by destabilizing astral microtubules To test whether phosphorylation of SAPCD2 disrupts its association with microtubules, we performed an in vivo microtubule co-sedimentation assay to examine the microtubule-binding ability of the phosphorylation-deficient SAPCD2 mutant in mitotic HeLa cells (Fig. 5A). Compared to wild-type SAPCD2, the 2 A mutant was significantly enriched in the microtubule-bound fractions, indicating enhanced association with microtubules (Fig. 5A, B). To investigate the functional role of SAPCD2 phosphorylation in microtubule regulation, we first assessed the effects of SAPCD2 phospho-mutants on interphase microtubule organization. Using cold-induced microtubule depolymerization assays, we observed that, compared with wild-type SAPCD2, the SAPCD2-2D mutant significantly decreased the resistance of interphase microtubules to depolymerization in HeLa cells (Fig. S6A–C). Next, we constructed stable cell lines expressing wild‑type SAPCD2 and its phospho‑mutants in SAPCD2‑depleted HeLa cells to explore the impact of SAPCD2 phosphorylation on astral microtubule stability (Fig. 5C). Rescue experiments revealed that astral microtubule intensity in SAPCD2-depleted cells reconstituted with either wild-type SAPCD2 or SAPCD2-2D was comparable to that in control cells, whereas cells expressing SAPCD2-2A exhibited significantly stronger astral microtubule intensity than control cells (Fig. 5D, E). These results indicate that re-expression of phosphorylation-mimetic SAPCD2 restores normal astral microtubule organization in SAPCD2-depleted cells, whereas phosphorylation-deficient SAPCD2 results in excessively stabilized astral microtubules relative to control cells. Rescue experiments were also performed to investigate whether SAPCD2 phosphorylation regulates spindle orientation (Fig. 5F). Spindle angle analysis showed that SAPCD2-WT expression restored normal spindle orientation, with most cells maintaining a spindle angle below 10° (Fig. 5G, H). Unlike SAPCD2-2A, which failed to rescue spindle misorientation, SAPCD2-2D successfully rescued these defects, indicating that SAPCD2 phosphorylation is essential for mitotic spindle positioning (Fig. 5G, H). Taken together, these results indicate that SAPCD2 destabilizes astral microtubules through mitosis-specific phosphorylation, which is crucial for maintaining proper spindle orientation.Fig. 5Mitosis-specific phosphorylation of SAPCD2 is required for accurate spindle orientation and astral microtubule stability.A Microtubule co-sedimentation assay of mitotic HeLa cells transfected with empty control, SAPCD2 wild-type or its mutant (SAPCD2-2A). The pellet (P) obtained from incubation at 37 °C indicates the microtubule-bound fractions, and the supernatant (S) indicates the unbound fractions. B Quantification of the relative enrichment of SAPCD2 in the pellet fraction, normalized to total SAPCD2 and tubulin levels in the pellet. Data are presented as (SAPCD2 in pellet/total SAPCD2) normalized by tubulin in the pellet (A). The data are represented as mean ± SD (n = 3). ***P < 0.001, Student’s t test. The experiment was independently repeated three times. C Western blot analysis of control cells or SAPCD2-depleted HeLa cells transfected with empty control, SAPCD2 wild-type or mutants. D Representative immunofluorescence images of control cells or SAPCD2-depleted HeLa cells stably expressing an empty control, SAPCD2 wild-type or mutants stained for α-tubulin (green), Flag (red) and DAPI (blue). Scale bar, 10 µm. E Quantification of astral microtubule fluorescence intensity imaged in (D). The data are represented as mean ± SEM (shControl, n = 27; shSAPCD2, n = 29; shSAPCD2 + SAPCD2-WT, n = 28; shSAPCD2 + SAPCD2-2A, n = 31; shSAPCD2 + SAPCD2-2D, n = 28), from three independent experiments. ns, not significant. **P < 0.01, ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. F Representative images of confocal z-sections of control cells or SAPCD2-depleted HeLa cells transfected with empty control, SAPCD2 wild-type or mutants stained for γ-tubulin (green), Flag (red) and DAPI (blue). Scale bars, 5 µm. G, H Scatter plots (G) and radial histograms (H) of the distributions of the spindle angles in HeLa cells imaged in (F). The data are represented as mean ± SEM (shControl, n = 117; shSAPCD2, n = 115; shSAPCD2 + SAPCD2-WT, n = 93; shSAPCD2 + SAPCD2-2A, n = 116; shSAPCD2 + SAPCD2-2D, n = 104), from over three independent experiments. ns, not significant. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. Mitosis-specific phosphorylation of SAPCD2 is required for accurate spindle orientation and astral microtubule stability. A Microtubule co-sedimentation assay of mitotic HeLa cells transfected with empty control, SAPCD2 wild-type or its mutant (SAPCD2-2A). The pellet (P) obtained from incubation at 37 °C indicates the microtubule-bound fractions, and the supernatant (S) indicates the unbound fractions. B Quantification of the relative enrichment of SAPCD2 in the pellet fraction, normalized to total SAPCD2 and tubulin levels in the pellet. Data are presented as (SAPCD2 in pellet/total SAPCD2) normalized by tubulin in the pellet (A). The data are represented as mean ± SD (n = 3). ***P < 0.001, Student’s t test. The experiment was independently repeated three times. C Western blot analysis of control cells or SAPCD2-depleted HeLa cells transfected with empty control, SAPCD2 wild-type or mutants. D Representative immunofluorescence images of control cells or SAPCD2-depleted HeLa cells stably expressing an empty control, SAPCD2 wild-type or mutants stained for α-tubulin (green), Flag (red) and DAPI (blue). Scale bar, 10 µm. E Quantification of astral microtubule fluorescence intensity imaged in (D). The data are represented as mean ± SEM (shControl, n = 27; shSAPCD2, n = 29; shSAPCD2 + SAPCD2-WT, n = 28; shSAPCD2 + SAPCD2-2A, n = 31; shSAPCD2 + SAPCD2-2D, n = 28), from three independent experiments. ns, not significant. **P < 0.01, ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. F Representative images of confocal z-sections of control cells or SAPCD2-depleted HeLa cells transfected with empty control, SAPCD2 wild-type or mutants stained for γ-tubulin (green), Flag (red) and DAPI (blue). Scale bars, 5 µm. G, H Scatter plots (G) and radial histograms (H) of the distributions of the spindle angles in HeLa cells imaged in (F). The data are represented as mean ± SEM (shControl, n = 117; shSAPCD2, n = 115; shSAPCD2 + SAPCD2-WT, n = 93; shSAPCD2 + SAPCD2-2A, n = 116; shSAPCD2 + SAPCD2-2D, n = 104), from over three independent experiments. ns, not significant. ***P < 0.001, one-way ANOVA with Tukey’s post hoc test. Discussion In retinal development, SAPCD2 has been reported to regulate planar spindle orientation through the conserved Gαi-LGN-NuMA pathway30. Here, using non-polarized HeLa and HaCaT cells as in vitro models, we uncovered an additional mechanism in which SAPCD2 functions as a previously unrecognized microtubule-associated protein that regulates proper astral microtubule stability, thereby controlling spindle orientation. Astral microtubule stability can be controlled by MAPs localized at the centrosome or at the microtubule plus ends. For example, MAP1S and γ-TuRC are recruited to the centrosome to stabilize astral microtubules, and a large number of microtubule plus-end tracking proteins, such as EB1 and MCAK, localize to the growing plus ends of microtubules to regulate astral microtubule stability19,38–40. Since co-depletion of SAPCD2 and MCAK effectively rescued the astral microtubule defects observed in SAPCD2-depleted cells, we propose that SAPCD2 and MCAK cooperate in maintaining a proper and functional astral microtubule population by contributing opposing roles on astral microtubule stability. Considering that SAPCD2 and MCAK play opposing roles in astral microtubule stability, it is plausible that SAPCD2 preferentially regulates rescue frequency or growth rate of microtubules, rather than catastrophe frequency as MCAK does41, although our current data do not directly assess these specific parameters of microtubule dynamic instability. Additionally, it has been established that mitotic phosphorylation of GTSE1 disrupts its interaction with EB1, leading to its dissociation from microtubule plus ends and the consequent relief of MCAK inhibition37,40. Overexpression of GTSE1 similarly restored the length and number of astral microtubules in SAPCD2-depleted cells (Fig. S7A–D), suggesting that SAPCD2 and GTSE1 may function in a common pathway. However, the subcellular localization of SAPCD2 does not support its accumulation at microtubule plus ends, posing a challenge to understanding its mechanism in stabilizing astral microtubules. Notably, the spindle pole localization of SAPCD2 during mitosis resembles that of ASK1, another regulator of astral microtubule stability. ASK1 colocalizes with EB1 at spindle poles and phosphorylates EB1 to enhance its plus-end targeting, thereby stabilizing astral microtubules42. Overexpression of ASK1 also rescued the defects in astral microtubule number and length in SAPCD2-depleted cells (Fig. S7A–D). These findings suggest that SAPCD2 plays an indirect role in regulating astral microtubule stability through interactions with other MAPs. Although MCAK and EB1 were identified as potential SAPCD2-interacting proteins, they exhibited low peptide coverage and few unique peptides in our mass spectrometry data (Fig. S2C and Table S1). Future studies focusing on the dynamic composition of SAPCD2-interacting proteins will be essential to elucidate its precise molecular function within the microtubule stability regulatory network. A notable feature of this work is the dynamic, multisite phosphorylation of SAPCD2 across the cell cycle, culminating in hyperphosphorylation during mitosis. CDK1 mediates three M-phase-specific phosphorylated forms of SAPCD2, which likely explain the M phase band shift. However, additional undetected phosphorylated forms cannot be ruled out. Dynamic multisite phosphorylation, a common feature of M-phase regulators, is achieved through the synergistic action of multiple kinases and phosphatases43–45. For example, mitotic kinases spatially control NuMA via site-specific phosphorylation: CDK1-mediated phosphorylation at T2055 promotes NuMA localization to spindle poles to ensure proper spindle orientation, whereas Aurora A-dependent phosphorylation at S1969 suppresses phase separation to enhance cortical localization, thereby regulating spindle length and dynamics46–48. Here, we identified a predominant mitosis-specific dual-phosphorylated form of SAPCD2 at S157 and S276 and suggested that this phosphorylation reduces SAPCD2’s affinity for microtubules, thereby decreasing astral microtubule stability and ensuring proper spindle orientation. S157 is a direct CDK1 target, but our data indicate that CDK1 phosphorylates SAPCD2 at additional sites, whose identities and functional relevance remain unknown. The kinase responsible for S276 phosphorylation has yet to be identified, underscoring the need for further dissection of upstream regulatory mechanisms. In addition, we observed a distinct SAPCD2 phosphor-form that appears in M phase and persists into G1 phase but is not phosphorylated by CDK1, suggesting the involvement of alternative kinases such as Chk2, which is active during M and G1 phases49. Notably, the dual-phosphorylated (pS157/pS276) form exhibited the fastest electrophoretic mobility among all detected isoforms, implying that phosphorylation at these sites may initiate a cascade of additional phosphorylation events during mitosis. Indeed, large-scale phosphoproteomic analyses have identified other candidate phosphorylation sites in SAPCD2, including T219 and S28450. Future studies should investigate additional regulatory kinases and their phosphorylation sites on SAPCD2 to fully elucidate its cell cycle-dependent modifications and functions. Phosphorylation of many MAPs can regulate their interaction with microtubules, likely through phosphorylation-induced changes in electrostatic charge or conformation51–53. We also found that mitosis-specific phosphorylation of SAPCD2 reduces its association with microtubules and decreases astral microtubule stability. Upon mitotic entry, SAPCD2 is rapidly upregulated, which could potentially lead to excessive microtubule binding and hyperstabilization of astral microtubules, as seen with the non‑phosphorylated 2 A mutant. However, mitotic phosphorylation of SAPCD2 at S157 and S276 reduces its association with microtubules and prevents excessive microtubule stability while retaining sufficient SAPCD2 function for proper astral microtubule population, as complete loss is also detrimental. Proper spindle orientation requires tight regulation of astral microtubule stability, and excessive stability is detrimental, as supported by findings that CDK1-dependent phosphorylation of GTSE1 destabilizes long astral microtubules to ensure accurate spindle orientation33,37. Thus, our data suggest that the physiological role of SAPCD2 phosphorylation may be to fine‑tune its microtubule interaction to an appropriate level, ensuring proper astral microtubule dynamics and correct spindle‑cortex attachment. This mechanism counteracts the detrimental effects of SAPCD2 upregulation and ultimately ensures precise spindle orientation (Fig. 6).Fig. 6A model for SAPCD2 phosphorylation dynamics and its impact on microtubule stability during the interphase-to-mitosis transition.In interphase, non-phosphorylated SAPCD2 binds to microtubules and enhances cytoplasmic microtubule stability. Upon mitotic entry, SAPCD2 undergoes phosphorylation at S157 and S276, with CDK1 specifically mediating the phosphorylation at S157. Dual phosphorylation at S157 and S276 attenuates SAPCD2’s association with microtubules, facilitating astral microtubule destabilization. This phosphorylation switch, which dynamically regulates astral microtubule stability through both phosphorylated and non-phosphorylated forms of SAPCD2, is essential for proper spindle orientation. Created in BioRender. shuang, C. (2026) https://BioRender.com/f7qb9qv. A model for SAPCD2 phosphorylation dynamics and its impact on microtubule stability during the interphase-to-mitosis transition. In interphase, non-phosphorylated SAPCD2 binds to microtubules and enhances cytoplasmic microtubule stability. Upon mitotic entry, SAPCD2 undergoes phosphorylation at S157 and S276, with CDK1 specifically mediating the phosphorylation at S157. Dual phosphorylation at S157 and S276 attenuates SAPCD2’s association with microtubules, facilitating astral microtubule destabilization. This phosphorylation switch, which dynamically regulates astral microtubule stability through both phosphorylated and non-phosphorylated forms of SAPCD2, is essential for proper spindle orientation. Created in BioRender. shuang, C. (2026) https://BioRender.com/f7qb9qv. SAPCD2, an embryonically expressed gene, has been shown to regulate planar spindle orientation in mouse retinal progenitor cells30. Our data extend this finding by demonstrating that SAPCD2 also controls spindle orientation in human epidermal keratinocytes, likely through regulating astral microtubule stability. We therefore propose that SAPCD2 plays a critical role in epidermal morphogenesis; however, validation using KO mouse models will be essential to confirm this hypothesis. SAPCD2 expression is generally minimal or undetectable in normal adult tissues, yet it is aberrantly upregulated in multiple cancers, where elevated levels correlate with poor tumor differentiation and advanced stage. Given the established link between spindle misorientation and tumorigenesis54–57, it will be important to investigate the phosphorylation status of SAPCD2 and its potential role in promoting tumor development through dysregulated mitotic spindle orientation. In conclusion, we identified SAPCD2 as a novel microtubule-associated protein that controls spindle orientation by stabilizing astral microtubules. Furthermore, we characterized two mitosis-specific phosphorylation sites on SAPCD2: CDK1-mediated phosphorylation at S157 and an additional phosphorylation at S276. Dual phosphorylation at these sites promotes the dissociation of SAPCD2 from microtubules, antagonizes the excessive stability of astral microtubules caused by increased SAPCD2 expression, and thereby helps maintain proper spindle orientation. These findings provide new insights into the molecular mechanisms governing mitotic spindle orientation and highlight a previously unrecognized role for SAPCD2 in cell cycle regulation. Materials and methods Cell culture HeLa (ATCC, CCL-2) cells were cultured at 37 °C in DMEM (11965092, Thermo Fisher Scientific) containing 10% FBS (FSP500, ExCell). HaCaT (BMCR, 1101HUM-PUMC000373) cells were cultured at 37 °C in MEM (11095080, Thermo Fisher Scientific) containing 15% FBS. Synchronization and drug treatments To enrich cells in different phases of the cell cycle, cells were synchronized in G1/S phase using a double thymidine block. Briefly, cells maintained in cultured medium with 2 mM thymidine (T9250, Sigma-Aldrich) for 18 h were replaced with fresh medium for 8 h before being treated with 2 mM thymidine again for 18 h. Following a 4 h release from a double thymidine block, cells were treated with 5 µM RO3306 (SC6673, Beyotime) for 9 h (cells synchronized in metaphase) before being released 3 h later to synchronize cells in early G1 phase. After a double thymidine block, cells were released for 4 h to synchronize them in S phase. To enrich cells in G2 phase, cells were treated with 10 µM RO3306 for 4 h after 4 h release from a double thymidine block. Finally, to enrich cells in M phase, cells were treated with 332 nM nocodazole (S1765, Beyotime) for 8 h after 4 h release from a double thymidine block. For spindle orientation or astral microtubules analyses, cells were released for 8.5 h after 24 h treatment with 2.5 mM thymidine. To analyze the impact of astral microtubules on spindle angle, cells were treated with 0.25 nM paclitaxel (SC0213, Beyotime) for 30 min following an 8 h release from a single thymidine block. RNAi, plasmids and transfection The target sequences of human SAPCD2 siRNAs were designed as follows: siSAPCD2-1, 5’-GCUGCAGGGUUUGGAGAUG-3’; siSAPCD2-2, 5’-CAAGGAAACAAUACACUAU-3’. The target sequences of human MCAK siRNA were designed as follows: siMCAK, 5’- CAUUGUAUGUAACAAUAAA-3’. The siRNA was transfected into cells using lipofectamine 3000 (L3000015, Thermo Fisher Scientific) to a final concentration of 50 nM according to the manufacturer’s instructions. SAPCD2 shRNA was inserted downstream of the U6 promoter in pLKO.1 vector. The targeting sequence (5’-CAAGGAAACAATACACTATGA-3’) was designed against the SAPCD2 3’‑UTR. Rescue constructs express only the coding sequence and are therefore resistant to this shRNA. Human SAPCD2 cDNA and CDK1 cDNA were constructed into the pCMV-N-Flag vector. The phosphorylation site mutant of SAPCD2 was constructed based on the plasmid pCMV-N-Flag-SAPCD2 using site-directed mutagenesis. The plasmid encoding SAPCD2‑Myc was constructed in our previous study58. Plasmids pCMV-N-3xFlag-ASK1 and pCMV-N-3xFlag-GTSE1 were purchased from Miaoling Biology Co., Ltd. (Wuhan, China). Plasmids for E. coli expression of MBP-tagged wild-type and mutant SAPCD2 were generated using Gateway® cloning technology (Thermo Fisher Scientific). The coding sequences of SAPCD2 and its mutants were first cloned into the pDONR™201 entry vector via BP recombination, and subsequently transferred into an N-terminal MBP-tagged destination vector through LR recombination. To generate plasmids for stable cell lines, SAPCD2 wild-type and mutants were constructed into the pSIN-C-Flag-Puro vector. Plasmids were transfected into cells using lipofectamine 2000 (11668027, Thermo Fisher Scientific). Immunoprecipitation, western blotting and antibodies HeLa cells transfected with SAPCD2-Flag or mutants were collected in lysate buffer (20 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% NP40, pH 7.5) on ice for 15 min before centrifugation at 13,000 × g for 10 min at 4 °C. The lysate supernatant was incubated with Flag-beads (Sigma) for 6 h. The beads were washed four times with lysate buffer before being eluted with SDS sample buffer. For the phos-tag assay, cells were lysed in EDTA-free buffer. Protein samples were resolved by SDS-PAGE or phos-tag SDS-PAGE and then transferred to a PVDF membrane (Millipore). The membrane was blocked in 5% milk for 1 h before being incubated with appropriate primary and secondary antibodies. Chemiluminescence was used to visualize protein signals. The antibodies used in this study are as follows: Anti-SAPCD2 (Homemade; WB 1:1000, IF 1:50)58, anti-Cyclin B1(AF1606, Beyotime; WB 1:2000), anti-GAPDH (D190090, BBI; WB 1:3000), anti-γ-tubulin (A9657, ABclonal; IF 1:200), anti-α-tubulin (ab4074, Abcam; IF 1:100), anti-α-tubulin (sc32293, Santa Cruz; IF 1:100), anti-α-tubulin (A6830, ABclonal; WB 1:5000, IF 1:200), anti-EB1 (17717-1-AP, Proteintech; WB 1:1000, IF 1:50), anti-MCAK (A9140, ABclonal; WB 1:1000), anti-Flag (AF5051, Beyotime; WB 1:2000, IF 1:200), anti-pan-phosphoTyrosine (AP0905, ABclonal; WB 1:1000), anti-phosphoSerine/Threonine (PP2551, ECM Biosciences; WB 1:1000), anti-Acetylated-Lysine (9441, Cell Signaling Technology; WB 1:1000), anti-L-lactyllysine (PTM-1401RM, PTM; WB 1:1000), anti-O-GlcNAc (sc-59623, Santa Cruz Biotechnology; WB 1:1000), anti-phospho-Erk1/Erk2 (AF1891, Beyotime; WB 1:1000), anti-MBP (sc-376995, Santa Cruz Biotechnology; WB 1:2000), anti-Myc tag (AF0033, Beyotime; WB 1:1000), anti-H3S10ph (ab14955, Abcam; WB 1:2000) and anti-Thiophosphate ester (ab92570, Abcam; WB 1:1000). Live-cell imaging and immunofluorescence For live-cell imaging, HeLa cells were plated in glass-bottom dishes coated with fibronectin (5 μg/mL, Sigma-Aldrich) and cultured at 37 °C with 5% CO2 in DMEM containing 10% FBS. Cells were observed using a Leica TCS SP8 confocal microscope equipped with a ×40 (Oil, NA = 1.3) objective. For the spindle orientation assay, cells cultured on fibronectin-coated dishes were fixed in methanol at −20 °C for 3 min follow by being fixed in PBS with 4% paraformaldehyde for 2 min at room temperature. For endogenous SAPCD2 localization analysis, cells were fixed in PBS with 10% trichloroacetic acid. For exogenous SAPCD2 localization analysis, cells were permeabilized in cytoskeleton buffer with 0.1% Triton X‑100 for 1 min, followed by fixation in cytoskeleton buffer containing 4% paraformaldehyde for 10 min at room temperature. For EB1 comets analysis, cells were fixed with pre‑chilled methanol for 10 min at −20 °C. For microtubules visualization analysis, cells were permeabilized in cytoskeleton buffer with 0.1% Triton X-100 for 2 min before being fixed in cytoskeleton buffer with 4% paraformaldehyde for 10 min. The fixed cells were then blocked with 3% BSA for 1 h before being incubated with the primary and Alexa-labeled secondary antibodies. Lastly, DAPI was utilized to stain the DNA. Spindle angle, image analysis and quantification Images of metaphase cells stained with γ-tubulin were acquired as Z-stacks with a step size of 0.4 µm using a Leica TCS SP8 confocal microscope equipped with a ×63 (Oil, NA = 1.40) objective. The angle tool from Fiji was then used to measure the angle of the mitotic spindle in the Z-stacks projection. To quantify the fluorescence intensity of microtubules, the maximum intensity projections of Z-stack images of cells stained with α-tubulin were acquired using a Leica Stellaris 5 confocal microscope equipped with ×63 (Oil, NA = 1.40) and ×100 (Oil, NA = 1.40) objectives. The fluorescence intensity of microtubules in all images was calculated using Fiji. Astral microtubule intensity was quantified, as mentioned in another study16. Fiji was used to measure the total microtubule intensity (I total) and spindle microtubules intensity (I spindle) in each cell. The relative astral microtubule intensity (I astral, relative) was then computed using [(I total – I spindle)/I total]. To measure astral microtubule length based on EB1 comet signals, Z-stack images of EB1-stained cells were acquired using a Leica Stellaris 5 confocal microscope equipped with a ×100 (Oil, NA = 1.40) objective. Astral microtubule lengths and numbers were determined from EB1 comet signals as previously described59. In short, EB1 was used to label microtubule plus ends, and the number of EB1 comets represents the number of microtubules. EB1 comets were identified using the Spots function in IMARIS with a diameter threshold of 0.5 μm, and their positions were recorded. The angle between each EB1 comet and the spindle axis was calculated. Comets with an angle larger than that between the spindle axis and the outermost spindle microtubules were identified as astral microtubule comets. The distance from these comets to the nearest spindle pole was defined as the astral microtubule length. Accordingly, the number of EB1 comets at the plus ends of these astral microtubules represents the number of astral microtubules. Microtubule co-sedimentation assay For the in vivo microtubule co-sedimentation assay, cells were broken with glass beads in PIPES buffer containing protease inhibitors, subsequently centrifuged twice at 20,000 × g for 20 min each, at 4 °C. The supernatant was added with 1 mM GTP and 20 μM paclitaxel before being incubated at either 4 °C or 37 °C for 30 min to promote tubulin polymerization. The samples were centrifuged again at 20,000 × g for 30 min at 4 °C or 37 °C, respectively and then the pellets and supernatants were collected for immunoblotting assay. For the in vitro microtubule co-sedimentation assay, porcine brain tubulin and purified MBP or MBP-SAPCD2 protein were co-incubated in PIPES buffer containing 1 mM GTP and 20 μM paclitaxel for 30 min at 37 °C to promote tubulin polymerization. As a control, purified MBP-SAPCD2 protein alone (without tubulin) was incubated under the same conditions. The subsequent steps were the same as for the in vivo microtubule co-sedimentation assay. In vivo microtubule depolymerization/regrowth assays For in vivo microtubule depolymerization assays, cells were incubated on ice for different time points, then permeabilized with 0.1% Triton X-100 followed by fixation with 4% paraformaldehyde. For in vivo microtubule regrowth assays, cells were incubated on ice for 40 min to introduce complete depolymerization of microtubules, and then the cells were returned to 37 °C for culturing to allow microtubules to regrow for different durations. Cells were permeabilized with 0.1% Triton X-100 followed by fixation with 4% paraformaldehyde. The fixed cells were labeled with anti-α-tubulin antibody. Residual/regrowth microtubules were quantified by measuring α-tubulin intensity. In vitro kinase assays For in vitro kinase assays using ATP (D7378, Beyotime), 1 μg of purified MBP or MBP-SAPCD2 wild type was incubated with 0.1 μg CDK1/Cyclin B1 in 30 μL of kinase reaction buffer (50 mM Tris-HCl, 1 mM EGTA, 0.01% Brij 35, 10 mM MgCl2, 100 μM ATP, pH 7.5) at 30 °C. After incubation for 30 min or 60 min, samples were boiled in SDS loading buffer for 10 min and analyzed by phos-tag gel. For in vitro kinase assays using ATPγS (ab138911, Abcam), 1 μg of purified MBP, MBP-SAPCD2 wild type or mutants was incubated with 0.1 μg CDK1/Cyclin B1 in 30 μL of kinase reaction buffer (50 mM Tris-HCl, 0.1 mM EDTA, 2 mM DTT, 0.01% Brij 35, 10 mM MgCl2, 100 mΜ ATPγS, pH 7.5) at 30 °C for 30 min. Then, the products were alkylated by adding 2.5 mM PNBM (ab138910, Abcam) at 25 °C for 90 min. Samples were boiled in SDS loading buffer for 10 minutes, separated by SDS-PAGE, and analyzed by immunoblotting using an anti-thiophosphate ester (anti-ThioP) antibody to detect thiophosphorylated proteins. In-gel digestion, mass spectrometry and data analysis SAPCD2-Flag was pulled down from HeLa cells with Flag-beads and then separated by phos-tag SDS-PAGE. After coomassie brilliant blue staining, phosphorylated SAPCD2-Flag bands were cut and subjected to in-gel destained, reduced, alkylated and digested with trypsin at 37 °C overnight. Peptide samples were reconstituted in 0.1% formic acid (FA) containing 2% acetonitrile (ACN) and separated using a reversed-phase analytical column (Acclaim PepMap RSLC C18, 5 μm, 100 Å, nanoViper; 75 μm × 150 mm i.d.) with an EASY-nLC 1200 system. Mass spectrometric analysis was conducted on a Q Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific) equipped with a nano-electrospray ionization (NSI) source. Protein identification was performed using MASCOT software by searching the Uniprot human protein database with carbamidomethyl (C) as a fixed modification, oxidation (M) as a variable modification, trypsin digestion, 1 missed cleavage, 20 ppm peptide mass tolerance, 0.6 Da fragment mass tolerance, monoisotopic mass values, and a significance threshold of 0.05. Statistics and reproducibility The normality of data distribution was assessed using Shapiro–Wilk or Kolmogorov–Smirnov tests. Statistical comparisons between two groups were performed using two-tailed unpaired Student’s t tests, while one-way ANOVA followed by Tukey’s post hoc test was applied for multi-group comparisons. All analyses were conducted using Origin 2019 (OriginLab Corporation, Northampton, MA, USA). P values are shown as: *P < 0.05, **P < 0.01 and ***P < 0.001. Protein expression was typically assessed in at least three independent experiments. For all other experiments, experimental replicates and sample sizes (n) are stated in the corresponding figure legends. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Description of Additional Supplementary files Supplementary Data 1 Supplementary Video 1 Supplementary Video 2 Reporting Summary Transparent Peer Review file Supplementary Information Description of Additional Supplementary files Supplementary Data 1 Supplementary Video 1 Supplementary Video 2 Reporting Summary Transparent Peer Review file