Author response:
Reviewer #1 (Public review):
The study by He and colleagues aims to investigate the molecular mechanisms driving key cell potency transitions, particularly the naïve-to-primed pluripotency transition. The authors explore the relationship between cell polarity and stemness using stem cell models combined with a comprehensive panel of experiments, including pharmacological inhibition and co-culture/conditioned medium rescue approaches. Overall, the study provides interesting observations and contributes to the understanding of the molecular mechanisms dynamically regulating stem cell differentiation.
However, several conceptual and interpretational aspects could be strengthened:
(1) First, the Introduction would benefit from being more focused on what is currently known regarding cell polarity during early embryogenesis and pluripotent stem cell transitions, rather than emphasizing later neurogenesis events. Such reorientation would better match the main topic of the manuscript and improve the conceptual coherence of the study.
We thank the reviewer for this constructive suggestion. We fully agree that the Introduction should be more tightly focused on the current understanding of cell polarity during early embryogenesis and pluripotent stem cell transitions, rather than on later neurogenesis events.
Accordingly, we will revise the Introduction in the following ways:
(1) Reduce the discussion on later neurogenesis and move some of those details to the Discussion section where they more appropriate.
(2) Expand the background on early embryonic development and pluripotent stem cell transitions by citing key recent and classical references, including but not limited to: cell polarity establishment in the preimplantation embryo, apical–basal polarity during lineage specification, polarity remodeling in naïve-to-primed pluripotent stem cell transition, the role of PAR complex in early mouse development.
(3) Refocus the Introduction to clearly state: what is known about polarity in early embryogenesis and pluripotent states, what remains unknown, and how our study addresses that gap.
(2) Similarly, Figure 6, where the authors attempt to provide clinical relevance through neural organoid formation experiments, feels somewhat disconnected from the central theme of the naïve-to-primed transition. Although this section is interesting on its own, there is already extensive literature describing polarization and morphogenetic events occurring much earlier during pluripotent state transitions. Therefore, the developmental relevance of the neural differentiation phenotypes could be better contextualized in relation to earlier morphogenetic events associated with pluripotency progression.
We thank the reviewer for this insightful comment. We agree that the neural organoid experiments in Figure 6 are somewhat disconnected from the central theme of the naïve-to-primed transition, and that extensive literature already exists on polarization events occurring earlier during pluripotent state transitions.
In the revised manuscript, we will better contextualize these findings by explicitly discussing how the neural differentiation phenotypes relate to the earlier morphogenetic events associated with pluripotency progression, rather than presenting them as a standalone observation. We will also incorporate relevant references to bridge this gap and strengthen the developmental relevance of our neural organoid data.
(3) The manuscript contains a substantial amount of experimental work; however, several results would benefit from deeper discussion. For example, in Figure 1, what is the rationale behind ZO1 downregulation being observed specifically in primed PAR knockout cells but not under naïve culture conditions? In addition, in Figure 3, the authors perform co-culture and conditioned medium experiments between wild-type and knockout cells. While the authors focus on the secreted protein fraction that rescues the phenotype, they also mention that other fractions display rescuing activity. Could the authors briefly discuss what additional components may contribute to this rescue effect? For example, could other molecules within these fractions also converge on AKT signaling regulation?
We thank the reviewer for recognizing the substantial experimental work in our manuscript and for providing these thoughtful suggestions to improve the depth of our discussion. We agree that deeper discussion of several key results will strengthen the manuscript. In the revised version, we will address the specific points as follows:
(1) Regarding ZO1 expression in Figure 1:
Our primary focus is actually on ZO1 localization rather than its total expression level. In our experiments, RNA-seq and immunofluorescence analysis revealed that the total expression level of ZO1 does not change significantly in PAR knockout cells. However, ZO1 localization is markedly altered in PAR knockout primed cells. Specifically, in wild-type primed cells, ZO1 is predominantly localized at the cell membrane, whereas this specific membrane accumulation is not observed in PAR knockout primed cells. Furthermore, this phenomenon is observed specifically under primed state and does not occur under naïve culture conditions. This is likely due to the differential requirement for PAR complex components in maintaining tight junction integrity during distinct pluripotency stages.
(2) Regarding the rescue activity of other fractions in Figure 3:
In our experiments, we found that beyond the secreted protein fraction, the WT CM-Exosome fraction exhibited limited rescue efficacy, particularly during the later stages of NPT. Based on our literature review, we suggest that these exosomal components may still contribute to the observed rescue effect, potentially through the delivery of functional proteins, miRNAs, or other signaling modulators that converge on AKT signaling regulation. This discussion will provide a more comprehensive understanding of the paracrine communication between wild-type and knockout cells, while acknowledging the limited contribution of exosomes relative to the secreted protein fraction.
(4) Importantly, transitions in cell potency are frequently associated with coordinated morphogenetic changes. For example, during mouse embryogenesis, naïve pluripotent inner cell mass cells progressively polarize into a rosette-like structure with apical domain specification before lumen formation and epithelialization during progression toward the primed epiblast state. This developmental context could help strengthen the biological interpretation of the study.
We sincerely thank the reviewer for providing this valuable developmental context. The example of naïve pluripotent inner cell mass cells progressively polarizing into rosette-like structures with apical domain specification before lumen formation and epithelialization during progression toward the primed epiblast state is highly insightful and directly relevant to our study.
In the revised manuscript, in the Introduction section, we will incorporate this developmental perspective to strengthen the biological interpretation of our findings. Specifically, we will place greater emphasis on the role of Par complex-mediated cell polarity in coordinating both pluripotency transitions and morphogenetic changes during early embryogenesis. We believe this contextualization will significantly improve the framing of our study and better connect our in vitro observations to in vivo developmental processes.
(5) There are also several claims throughout the manuscript that appear to be overinterpreted or insufficiently quantified. For example, in Figure 1, the authors state that CDH1 expression is uniform; however, this is difficult to appreciate from the images shown, and quantitative analysis would be necessary to support this conclusion.
We thank the reviewer for this important comment. We agree that the claim that "CDH1 expression is uniform" in Figure 1 is overinterpreted based on the images shown, and we apologize for the lack of quantitative support.
Upon re-examination, we realize that our focus should be on CDH1 localization rather than its expression level or uniformity. In the updated manuscript, we will rephrase the statement about uniformity and instead present appropriate quantitative analysis (e.g., RNA-seq or fluorescence quantification across multiple cells) to better support our conclusions regarding CDH1 distribution. We will also adjust our data presentation to more clearly reflect the localization changes we observe.
(6) Another example appears in Figure 2, where the authors claim that "heatmap analysis revealed that transcriptomic profiles of PAR knockout cells progressively diverged from wild type from day 3 onwards". This conclusion is not fully supported by the presented data for two reasons: (1) transcriptomic divergence is more appropriately assessed through principal component analysis, clustering, or distance-based methods rather than by visual inspection of a heatmap alone; and (2) although some genes displayed in panel E begin to show genotype-associated differences from day 3, the overall transcriptomic structure shown in the PCA and heatmap remains primarily dominated by temporal progression rather than genotype.
We thank the reviewer for this careful and constructive critique. We apologize for the imprecise claim regarding the heatmap analysis in Figure 2. We agree that 1) transcriptomic divergence should be assessed by PCA, clustering, or distance-based methods rather than by visual inspection of a heatmap alone, and 2) the overall transcriptomic structure shown in PCA and heatmap remains primarily dominated by temporal progression rather than genotype.
In fact, our main point in this figure was to show that differentially expressed genes (DEGs) between PAR KO and WT become more numerous and more pronounced from day 3 onwards, and the supporting data for this claim are presented in Supplemental Figure 2 A–B. The number of DEGs between PAR knockout and wild-type cells is 480 at day 1, 523 at day 3, 1088 at day 4, and 1893 at day 6. Furthermore, we focused on specific genes within particular signaling pathways, and their expression levels began to show significant differences between PAR knockout and wild-type cells from day 3 onwards.
We realize that our original wording was misleading. In the revised manuscript, we will rephrase our conclusion to more accurately reflect what the data actually show, focusing on the timing and extent of differential gene expression rather than suggesting a global divergence of transcriptomic profiles.
(7) In this context, it remains unclear whether PAR knockout cells truly retain a more naïve pluripotent transcriptomic identity. To support this claim, the authors should compare the knockout transcriptome directly against a naïve pluripotent population. The phenotype observed in the knockout cells may instead represent an incomplete or aberrant primed transition rather than maintenance of naïve pluripotency itself. Intermediate morphogenetic states, such as rosette-like epithelial stages, could also explain the observed phenotype.
We apologize for the confusion caused by our imprecise wording. We realize that our original manuscript may have inadvertently suggested that Par knockout cells retain a naïve pluripotent transcriptomic identity, which was not our intended claim.
To clarify, Par knockout naïve cells lose their naïve identity and differentiate toward a primed state during the NPT process described in this manuscript. Unlike wild-type primed cells, PAR-knockout primed cells exhibit altered morphology: they cannot establish or maintain the typical flat morphology, and possess distinct expression profile. In terms of naïve identity, key naïve markers (e.g., Esrrb or Oct4) are downregulated to comparable levels in both wild-type and Par knockout primed cells. Although the two cell types differ in their overall expression profiles, several core primed markers (e.g., Fgf5 or T) show normal expression in both groups. Collectively, these results indicate that Par knockout naïve cells do lose their naïve identity and undergo differentiation toward a primed state during NPT, even though the final primed states of the two cell populations are distinct.
In the revised manuscript, we will:
(1) Revisit and revise our wording to avoid any misinterpretation that Par knockout cells retain a naïve identity.
(2) Directly compare the transcriptome of Par knockout cells against a true naïve pluripotent population (e.g., naïve ESCs) to further support our conclusion that the knockout cells are not maintaining naïve pluripotency, but rather exhibit an aberrant primed state with morphological abnormalities.
(3) Discuss the possibility that the observed phenotype may represent an intermediate morphogenetic state (e.g., rosette-like epithelial stages) rather than genuine naïve pluripotency maintenance.
(8) Strengthening this aspect of the study would substantially improve its developmental and in vivo relevance, which currently appears somewhat limited. In particular, it would be interesting to determine whether this mechanism operates during embryogenesis itself. The authors could consider relatively simple but informative experiments, such as perturbing PAR signaling or Furin activity during embryo culture.
We thank the reviewer for this constructive and forward-looking suggestion. We agree that the current manuscript focuses primarily on in vitro cellular mechanisms, and we have not sufficiently explored the developmental and in vivo relevance of our findings. We acknowledge that this aspect of the study is currently somewhat limited.
In the revised manuscript, we will:
(1) Explicitly acknowledge this limitation in the Discussion section.
(2) Incorporate more background on early embryogenesis, particularly regarding pluripotency transitions and morphogenetic changes during early development, to better contextualize our in vitro observations.
(3) We will attempt to use embryo-like models to investigate whether the PAR complex–Furin–Lefty–FAK signaling axis also operates during embryogenesis itself. As the reviewer suggested, simple but informative experiments—such as perturbing PAR signaling or Furin activity during embryo culture—would be valuable next steps to determine the in vivo relevance of our proposed mechanism. We will include these as important future perspectives.
(9) Along the same lines, some statements in the manuscript appear overly speculative. For example, the statement that "these findings may reveal a developmental compensation mechanism during embryogenesis, whereby normal cells rescue defective cells or increase their own proportion" extends well beyond the experimental evidence presented. Such claims invoke concepts related to cell competition, abnormal cell recognition, or developmental quality control mechanisms in vivo, none of which are directly demonstrated in this study. The authors are encouraged either to substantially tone down these statements or move them to the Discussion as speculative possibilities.
We thank the reviewer for this important critique. We agree that our original statement—"these findings may reveal a developmental compensation mechanism during embryogenesis, whereby normal cells rescue defective cells or increase their own proportion"—is overly speculative and extends beyond the experimental evidence presented in our study. We also acknowledge that it was inappropriate to directly extrapolate from in vitro cellular mechanisms to in vivo developmental rules without proper justification.
In the revised manuscript, we will:
(1) Substantially tone down this claim from the Results section.
(2) Move this speculation to the Discussion section, where we will explicitly present it as a speculative possibility rather than a conclusion supported by our data. We will also clearly state that concepts such as cell competition, abnormal cell recognition, or developmental quality control mechanisms remain to be tested in future studies.
(10) Another important conceptual point concerns the relationship between PAR complex regulation and Lefty signaling. If this mechanism indeed reflects a physiological or homeostatic process operating during embryogenesis, what would be the developmental rationale for the PAR complex regulation of Lefty? Lefty is well known for its role during gastrulation and anterior epiblast patterning. It would therefore be interesting if the authors could further discuss potential links between these developmental contexts.
We thank the reviewer for raising this important conceptual point. In our manuscript, we have indeed demonstrated that the PAR complex regulates Lefty signaling under the conditions of this study, and we are aware from the literature that Lefty signaling plays a critical role during early embryogenesis, particularly in gastrulation and anterior epiblast patterning.
However, we admit that we have not deeply considered the potential pathways and developmental rationale for PAR complex-mediated regulation of Lefty in the context of embryogenesis. This is an important gap in our current discussion.
In the revised manuscript, we will:
(1) Review and incorporate relevant literature to better understand and discuss the potential links between PAR complex regulation and Lefty signaling during early embryonic development, including possible connections to gastrulation and anterior patterning.
(2) Offer speculative but informed perspectives on the developmental rationale for such regulation, while clearly distinguishing between what our data directly show and what remains to be explored in future studies.
Minor points:
(1) The authors state that PAR knockout cells do not exhibit major differences in self-renewal capacity; however, they simultaneously claim that these cells remain in a more naïve-like state. This interpretation requires clarification, as naïve pluripotent cells are typically associated with increased clonogenicity, enhanced self-renewal, and expression of markers such as alkaline phosphatase and SSEA1 compared to primed cells. The relationship between the observed phenotype and the proposed "naïve-like" state should therefore be discussed more carefully.
We thank the reviewer for this comment, which addresses a similar concern as Point 7 mentioned above. Consistently, we do not claim that PAR knockout cells remain in a more "naïve-like" state. Our actual conclusion is that PAR knockout naïve cells undergo differentiation toward the primed state during NPT. However, due to loss of cell polarity, PAR knockout primed cells fail to establish and maintain the typical flat morphology and instead form dome-shaped colonies. Importantly, these dome-shaped colonies do not retain the characteristics of the naïve state, such as increased clonogenicity, enhanced self-renewal, or expression of alkaline phosphatase and SSEA1.
In the revised manuscript, we will:
(1) Revise our wording to avoid any misinterpretation that PAR knockout primed cells maintain a naïve-like identity.
(2) Explicitly clarify that the observed dome-shaped morphology represents an aberrant primed state rather than a naïve or naïve-like state.
(3) Discuss more carefully the relationship between the observed phenotype and the absence of typical naïve state features.
(2) The authors generated several independent knockout clones, but appear to use only one clone for downstream analyses after observing similar morphogenetic phenotypes. Is this sufficient to account for potential clonal heterogeneity? Would the use of pooled clones provide a more robust experimental system?
We thank the reviewer for raising this important concern regarding clonal heterogeneity. We agree with the reviewer that our current approach using only one representative knockout clone for downstream mechanistic analyses after confirming similar morphogenetic phenotypes across multiple independent clones is not sufficient to fully exclude potential clonal heterogeneity.
To address this issue, we will perform additional experiments in the revised study. Specifically, we will use another independent knockout clone (ParKO6) to repeat the key mechanistic analyses. The following experiments will be carried out:
(1) ParKO6 and wild-type ESCs will be subjected to NPT. During the NPT process, cells will be treated with an AKT inhibitor (MK2206), a FAK inhibitor (PF562271), or WT CM. We will observe whether the morphological defects of ParKO6 cells are rescued, and RT-qPCR will be performed to characterize the molecular features of ParKO6 cells under these conditions.
(2) After treatment with the AKT inhibitor (MK2206), FAK inhibitor (PF562271), or WT CM, immunofluorescence (IF) will be used to detect p-FAK levels in ParKO6 cells.
(3) Following the same treatments, Western blotting (WB) will be performed to detect FURIN and LEFTY protein levels in ParKO6 cells.
These additional experiments will allow us to confirm that the observed results are not due to clone-specific artifacts from the originally used clone.
(3) The rescue experiments using pathway inhibitors are interesting; however, the interpretation again relies primarily on colony morphology. Readers may question whether these experiments truly represent rescue of the naïve-to-primed transition itself without additional transcriptomic or molecular characterization.
We thank the reviewer for this important comment. We apologize for the lack of clarity in our original manuscript, which may have led to the misunderstanding that our interpretation of the rescue experiments relied solely on colony morphology.
In fact, we did perform molecular characterization on a subset of cells rescued by pathway inhibitors, and these data are presented in Supplemental Figure 2 D–E. We realize that our description of these results was insufficiently clear, and we failed to properly highlight this molecular evidence in the main text.
In the revised manuscript, we will revise our wording to clearly state that the rescue effects are supported not only by morphological observations but also by molecular characterization.
(4) In Figure 4, the manuscript could be strengthened by integrating transcriptomic analyses from pharmacological treatments with the secreted-factor and co-culture datasets.
We thank the reviewer for this constructive suggestion.
In our current manuscript (Figure 4), we have indeed performed an integrated transcriptomic analysis comparing pharmacological treatment and secreted-factor treatment, and we demonstrated that both treatments converge on the FAK signaling.
Regarding the co-culture dataset, we did not include it in the integrated analysis presented in Figure 4. This is because, based on our data in Figure 3, we concluded that the rescue effect observed in co-culture is primarily mediated through secreted factors. Therefore, the secreted-factor transcriptomic data already capture the key signaling pathways responsible for the co-culture rescue effect.
We will clarify this rationale explicitly in the revised manuscript to avoid any confusion.
(5) The authors could better clarify the context of Furin downregulation in the knockout cells. Is this a direct consequence of altered transcriptional regulation by the PAR complex, or could it instead represent a secondary consequence of impaired progression through the primed pluripotent transition?
We thank the reviewer for this important mechanistic question.
Based on our experimental data, we conclude that Furin downregulation in PAR knockout cells is a direct consequence of altered transcriptional regulation by the PAR complex, rather than a secondary consequence of impaired progression through the primed pluripotent transition. Our evidence is as follows:
(1) Transcriptomic analysis revealed that PAR knockout leads to a significant reduction in Furin RNA levels.
(2) Western blot analysis confirmed that PAR knockout also results in a significant reduction of FURIN protein levels.
(3) Importantly, treatment with an AKT inhibitor (upstream of the proposed pathway) significantly upregulated both Furin RNA and protein levels in PAR knockout cells. In contrast, treatment with a FAK inhibitor or WT CM (downstream) did not significantly alter Furin expression.
These data collectively indicate that Furin downregulation is directly linked to PAR complex-mediated transcriptional regulation, rather than being an indirect consequence of defective primed state transition. We will clarify this rationale in the revised manuscript.
Reviewer #2 (Public review):
Summary:
The study demonstrated that Par, but not other polarity genes, Crumbs or Scrib, regulates cell polarity during PSC transition to primed state as well as neural tube formation.
Strengths:
The use of KO convinces the role of Par in NPT. Scrib and Crumbs KO data are informative to the field. The conditioned medium experiment is informative. They suggested the potential secreted factors over 50kDa are responsible for maintaining the polarity of NPT in Par KO.
Weaknesses:
(1) Most importantly, how Par is important for PSC maintenance and differentiation is not clear. The data provided are dome shape formation, endoderm lineage tendency, and neural tube formation reduction. The manuscript lacks a core message of the physiological importance of Par. Is Par critical of PSC maintenance? Is Par critical for neural system development?
We thank the reviewer for this critical comment, which helps us better articulate the core message of our study.
In our manuscript, we have provided clear evidence regarding the role of the PAR complex in pluripotent stem cell (PSC) maintenance and differentiation:
(1) Regarding PSC maintenance:
The PAR complex is not critical for PSC maintenance under self-renewing conditions. Specifically, PAR knockout does not significantly affect the expression levels of pluripotency genes (Figure 1 B–C and Supplemental Figure 1 C). Moreover, PAR knockout PSCs can be continuously cultured for at least 30 passages without notable changes in cell morphology or proliferation capacity (Figure 1 D–F). These findings are consistent with previous literature, which demonstrates that the core function of the PAR complex is to establish and maintain cell polarity, rather than directly regulating the transcriptional network of pluripotency genes.
(2) Regarding PSC differentiation:
The PAR complex is important for proper differentiation. PAR knockout leads to multiple differentiation defects, including: Failure to establish normal cell morphology during (NPT) (Figure 1 G–K). Impaired formation of proper three-germ-layer structures during embryoid body (EB) and teratoma differentiation (Figure 5 F–G). In particular, the type and quantity of ectodermal tissues are significantly reduced. Consistent with our findings, previous literature has reported that PAR complex deficiency leads to neural developmental defects in mouse embryos, resulting in mid-gestation embryonic lethality.
(3) Regarding neural system development:
The PAR complex is critical for neural development. During neural stem cell (NSC) differentiation, PAR knockout cells exhibit a significantly reduced efficiency of Nestin-positive cells and fail to form the classical rosette structures (Supplemental Figure 5 B–C). During neural tube organoid induction, PAR knockout cells show significantly impaired lumen formation and spontaneous elongation efficiency. Moreover, during subsequent maturation, PAR knockout cells fail to differentiate into neurons, leading to a marked reduction in neural tube organoid maturation efficiency (Figure 6 B–E).
These findings are consistent with previous literature showing that in zebrafish embryonic development, mislocalization of the PAR complex leads to neural tube abnormalities while PAR complex deficiency results in severe hydrocephalus; in mouse embryonic development, PAR complex deficiency causes neural developmental defects leading to embryonic lethality; and disruption of the PAR complex impairs the formation of apical tight junctions in the neuroepithelium and subsequent neuroepithelial tissue polarization, resulting in neural tube closure defects in humans.
In the revised manuscript, we will incorporate classical literature to discuss the essential roles of the PAR complex in early embryonic development, thereby providing a broader developmental context for our findings.
(2) Secondly, AKT-FURIN-...... axis still lacks supportive data. Various inhibitors were used to rescue the Par KO. But the link between each component in the axis is missing and rather superficial.
We thank the reviewer for this critical comment. We acknowledge that the proposed AKT–FURIN–LEFTY–ECM-integrin–FAK signaling axis has certain limitations, particularly that the connection between LEFTY and ECM-integrin lacks direct experimental support. Therefore, in the revised manuscript, we will de-emphasize the role of ECM and integrin and revise the signaling axis to AKT–FURIN–LEFTY–FAK.
We believe the current data and previous publications support this revised signaling axis well. Accordingly, we have summarized the relevant information as follows. In addition, we plan to perform additional experiments to further support the new signaling axis, which are also included in the following text.
(1) AKT-FAK
We found that Par KO cells exhibit defects during NPT, and these defects can be rescued by AKT inhibitor (MK2206), FAK inhibitor (PF562271), and WT CM. Through transcriptomic analysis, we found that both AKT inhibitor and WT CM share similar expression profiles with WT and converge on FAK signaling. Notably, through Western blotting analysis, we found that Par KO led to upregulated p-AKT levels, which were effectively suppressed by MK2206 treatment, but WT CM did not decrease p-AKT levels. In contrast, through immunofluorescence analysis, we found that FAK signaling was hyperphosphorylated in Par knockout primed cells compared to WT primed cells, and MK2206, WT CM, and PF562271 all effectively reduced p-FAK levels. Given that both MK2206 and WT CM attenuated the elevated p-FAK, we propose that all three treatments restore the flat monolayer morphology by regulating FAK signaling homeostasis, with WT CM acting downstream of AKT signaling. The relevant data are presented in Figure 1G-I, Figure 2F, Figure S2C, Figure 3H-I, Figure 4A-D, and Figure S4A.
(2) AKT-LEFTY
Through integrated proteomic and transcriptomic analysis, we identified a set of functional proteins. Overexpression screening revealed that LEFTY exhibited the most significant rescue effect in Par KO cells during NPT. Proteomic analysis revealed that the protein levels of LEFTY were significantly higher in WT CM compared to KO CM, suggesting that WT cells modulate FAK signaling via secretion of LEFTY proteins. It is therefore reasonable to infer that MK2206 rescues the defects in Par KO primed cells through upregulation of LEFTY expression. Western blotting analysis confirmed this, showing that MK2206 significantly increased LEFTY protein levels in Par KO primed cells. The relevant data are presented in Figure 4E, Figure 4H and Figure S4C-D.
(3) LEFTY-FAK
Proteomic analysis indicated that WT CM treatment supplied extracellular LEFTY to Par KO ESCs, thereby rescuing the phenotypic defects of Par KO primed cells, and significantly reduced p-FAK levels in these cells. Concordantly, LEFTY overexpression also reduced p-FAK in Par KO primed cells. These results are consistent with the reported role of LEFTY in suppressing FAK signaling (Alowayed et al., 2016). The relevant data are presented in Figure 4D, Figure 4F, and Figure S4D-E.
(4) AKT-FURIN
LEFTY proprotein requires FURIN-mediated cleavage for secretion and function (Dubois et al., 2001). Through transcriptomic analysis, we found that Par KO downregulated Furin mRNA expression, while MK2206 treatment restored its expression levels. Through Western blotting analysis, we found that MK2206 increased FURIN protein abundance and cleaved LEFTY levels. The relevant data are presented in Figure 4G-H.
(5) FURIN-LEFTY
To validate the role of FURIN in LEFTY maturation, we treated WT cells with BOS318, a highly specific and potent inhibitor of FURIN that irreversibly binds to the protease by mimicking its natural substrate (Ivachtchenko et al., 2024). BOS318 induced WT primed cells to adopt a dome-shaped morphology resembling Par KO primed cells, confirming that inhibition of FURIN prevents LEFTY secretion and function, leading to defective primed cell morphology. The relevant data are presented in Figure 4I-J. To further strengthen the role of FURIN in regulating LEFTY, we will treat wild-type cells with BOS318 and examine the expression changes of LEFTY.
(6) ECM/integrin
Integrated analysis of both transcriptomic and proteomic data revealed that Par KO leads to significant enrichment of pathways associated with ECM and integrin (Figures 2D, 3F, 3K, and S4B). Notably, both MK2206 and WT CM treatment co-upregulated the ECM-receptor interaction pathway (Figure 4C). The FAK signaling pathway serves as a central node that integrates upstream inputs from both PKC and AKT pathways while transducing extracellular cues derived from ECM-integrin interactions into intracellular signaling cascades (Sakthivel et al., 2025). We therefore propose that secreted LEFTY acts as an extracellular signal that activates specific ECM receptors and modulates integrin complexes, thereby regulating FAK phosphorylation and maintaining normal cell adhesion and morphology. However, this speculation still lacks direct experimental evidence. We will endeavor to perform additional experiments to support this proposed connection in the future. Nevertheless, we have decided to de-emphasize the role of ECM and integrin in the AKT–FURIN–LEFTY–FAK signaling axis in the current manuscript.
References
Alowayed, N., Salker, M. S., Zeng, N., Singh, Y., & Lang, F. (2016). LEFTY2 Controls Migration of Human Endometrial Cancer Cells via Focal Adhesion Kinase Activity (FAK) and miRNA-200a. Cellular Physiology and Biochemistry, 39(3), 815-826. https://doi.org/10.1159/000447792
Dubois, C. M., Blanchette, F., Laprise, M.-H., Leduc, R., Grondin, F., & Seidah, N. G. (2001). Evidence that Furin Is an Authentic Transforming Growth Factor-β1-Converting Enzyme. The American Journal of Pathology, 158(1), 305-316. https://doi.org/10.1016/s0002-9440(10)63970-3
Ivachtchenko, A. V., Khvat, A. V., & Shkil, D. O. (2024). Development and Prospects of Furin Inhibitors for Therapeutic Applications. International Journal of Molecular Sciences, 25(17). https://doi.org/10.3390/ijms25179199
Sakthivel, K., Kotowska, A., Fan, Z., Portner, E. J., Merry, C., Nordenfelt, P., Simonsen, A. C., Wright, A. J., & Swaminathan, V. S. (2025). Integrin‐Piezo1 Axis Drives ECM Remodeling and Invasion of 3D Breast Epithelium. Advanced Science. https://doi.org/10.1002/advs.202509932
