Immunohistochemical Analysis of IRE1 and PERK Expression in Extravillous Trophoblast Cells of Placenta Accreta Spectrum and Associations With Clinicopathological Parameters

Discussion

In this study, we examined associations between expression IRS and clinicopathological variables. While neither IRE1 nor PERK correlated with baseline demographic or obstetric variables, PERK scores were lower in those with maternal complications, and elevated PERK was found in PAS cases in comparison to controls (normal placenta cases). These findings that may inform the role of ER stress pathways in PAS pathology.

Our study demonstrates a selective increase of the PERK-IRS of the UPR at the placental-myometrial interface in PAS, with significantly higher PERK immunoreactivity in EVT-rich regions compared with cesarean controls, and an association between higher PERK-IRS and maternal complications. In contrast, dichotomized IRE1 staining did not differ between groups and neither PERK-IRS nor IRE1-IRS varied by accreta, increta, or percreta category. These results fit the contemporary view that PAS arises primarily from a maternal implantation bed defect overlying a uterine scar, in which decidual deficiency and an altered myometrial interface expose anchoring trophoblasts to hypoperfusion, intermittent hypoxia-reoxygenation, oxidative stress, and inflammation, all of which can trigger UPR signaling (17-19).

The broader placental literature provides a clear framework for interpreting these findings. Placental ER stress represses transcription of placental growth factor through ATF4 and ATF6β, offering a mechanistic bridge to angiogenic imbalance in pre-eclampsia. Mizuuchi et al. showed that ATF4 and ATF6β synergistically downregulate expression of placental growth factor in human trophoblast, directly linking ER stress to reduced output of placental growth factor (18). Evidence from human placentas indicates that ER stress is prominent in early-onset pre-eclampsia even in non-labored deliveries, implicating placental malperfusion as the upstream driver (8, 12). The process of spontaneous labor itself can acutely induce placental ER stress, emphasizing the importance of non-labored controls when comparing disease groups (20, 21). Moreover, circulating factors in sera from pre-eclamptic women are sufficient to activate UPR programs in explants and trophoblast cell lines, reducing placental growth factor and increasing canonical stress readouts, which supports a causal link between maternal milieu and placental stress signaling (22).

The balance and timing of ER stress appear to differ across placental syndromes. In early-onset pre-eclampsia, multiple UPR branches are activated, with robust molecular evidence of translational inhibition and stress signaling; late-onset pre-eclampsia typically shows a milder or more variable UPR activation pattern (23, 24). Yung et al. reported differential activation of placental UPR between early and late forms, providing molecular evidence that the early phenotype is more strongly stress driven (24). Consistent with this, Burton et al. synthesized earlier findings to argue that early-onset disease reflects severe malperfusion with pronounced ER stress, while the late-onset form is more influenced by maternal constitutional factors and displays less intense placental stress signaling (25). Our observation that PERK is selectively increased at the PAS interface resembles the early-onset pre-eclampsia pattern in its focal intensity at a stressed implantation niche, although in PAS, the etiological driver is the scarred decidual bed rather than primary defects in trophoblast invasion (3, 25).

Fetal growth restriction also shows a characteristic placental ER stress signature. Classic pathology and molecular studies document increased GRP94, CHOP, and phospho-eIF2α in fetal growth restriction (FGR) placentas, supporting translational attenuation and pro-apoptotic signaling in the syncytiotrophoblast and fetal endothelium (11). A broader review likewise concluded that ER stress is central to the small-placenta phenotype in FGR and early-onset pre-eclampsia, with oxidative stress superimposed in the latter (12). When compared with our data, the overlap is the prominence of translational control pathways at sites of greatest physiological strain. In PAS, however, the stress is spatially concentrated at anchoring villi abutting myometrium rather than diffusely affecting the villous tree, which may explain why we detected a sharp PERK signal at the interface without parallel changes across depth categories (1, 19).

Gestational diabetes mellitus (GDM) provides an informative metabolic comparator. Human placental studies demonstrate ER stress in GDM, including activation of the UPR in trophoblast exposed to hyperglycemia in vitro and ER stress readouts in GDM placentas in vivo (15, 26). Yung et al. reported ER stress in GDM placentas and proposed that chemical chaperones or antioxidant vitamins warrant exploration as modulators of this pathway (26). Subsequent work has implicated inflammatory signaling under ER stress control, suggesting a CHOP-peroxisome proliferator-activated receptor α-nuclear factor-κB axis in GDM, and has shown reduced histocompatibility antigen, class I, G (HLA-G) in EVTs from GDM placentas, a finding that highlights how metabolic stress can intersect with immune tolerance at the interface (27). While not on PAS, these studies illustrate that distinct maternal stressors selectively engage different UPR branches and downstream programs in trophoblast (27, 28). Relative to these conditions, our PAS specimens showed a focal and pronounced PERK predominance at the scarred interface, consistent with sustained local stress that enforces translational control and structural remodeling rather than a global placental response.

Mechanistically, PERK phosphorylates eIF2α to reduce global translation while enabling adaptive ATF4 programs; under sustained stress, CHOP induction can promote apoptosis. In trophoblast models, pro-inflammatory cytokines and pharmacological stress lower matrix metallopeptidase-2 and impair invasion in a PERK-dependent manner, while PERK pathway inhibition restores matrix metallopeptidase-2 and partially rescues invasive capacity (20). PERK also coordinates actin remodeling by interacting with filamin-A at ER-plasma membrane contact sites, providing a structural route by which local stress can alter cell-matrix engagement and adhesion (29). Taken together, these data offer a cogent explanation for our findings: in the decidual-deficient, hypoperfused implantation bed of PAS, sustained PERK engagement may favor firm adhesion and shallow but pathological anchoring of villi to the myometrium, increasing the risk of abnormal placental separation and hemorrhage without invoking an intrinsically hyper-invasive trophoblast phenotype (1, 2).

Of note, in the first-trimester, extravillous trophoblasts (EVTs) exhibit enhanced IRE1-XBP1s signaling compared with villous cytotrophoblasts. Partial IRE1 inhibition with 4μ8C during EVT differentiation reduced the proportion of surface HLA-G-positive cells without broadly altering lineage markers, suggesting that IRE1 supports antigen synthesis or trafficking required for immune tolerance (14). This baseline necessity suggests that both under- and overactivation might be detrimental depending on the context. Moreover, a study of human placentae from pre-eclamptic pregnancies found subtype-specific engagement of UPR branches. Early-onset pre-eclampsia (often with FGR) showed stronger PERK-eIF2α-ATF4/CHOP activity together with IRE1-dependent XBP1 splicing and ATF6 activation than late-onset disease (30).

The PAS interface emerges as a microanatomical zone where sustained local stress preferentially engages PERK. Our finding of correlation between higher PERK scores and maternal complications suggests that the magnitude of local UPR engagement may associated with operative complexity or hemorrhagic risk, a hypothesis that requires confirmation in larger cohorts with standardized grading and clinical endpoints (1, 31).

Methodologically, our targeted sampling of distal trophoblast cell columns and basal plate, combined with normal placental tissues controls, is a strength because ER stress is induced by labor and the invasive front is the biologically relevant site for PAS (21). Limitations include modest sample size, dichotomization of IRE1 and lack of activation-specific readouts such as phospho-PERK, phospho-eIF2α, ATF4, CHOP, phospho-IRE1, or spliced XBP1, as well as the absence of paired functional markers such as matrix metallopeptidase-2 or HLA-G in the same tissue regions (20). The lack of a PAS-control difference in our IRE1 analysis may reflect limited power, coarse categorization, or unsuitable of the current experimental process or the antibody used, rather than a true absence of IRE1 involvement at the interface. Future studies should incorporate multiplexed activation mapping across EVT, decidua, and immune compartments and should integrate UPR readouts with quantitative PAS pathology frameworks to test whether stress signatures predict hemorrhage or surgical difficulty beyond depth grading (1).

Beyond placentation, oncological data underscore the clinical relevance of UPR signaling. In breast cancer, immunohistochemical studies report that high IRE1 and PERK expression levels are associated with aggressive tumor characteristics and reduced survival, highlighting the importance of the UPR in carcinogenesis and disease progression (32). A contemporary synthesis further links ER stress programs, including PERK-mediated translational control and IRE1 to XBP1s signaling, to tumor growth, immune escape, and treatment resistance in a context-dependent manner (33). These findings underscore the UPR as a central cellular pathway implicated across diverse diseases, and indicate that activation-resolved, cell type-specific studies will deepen pathophysiological insight and guide the development of targeted therapies.

Finally, these data carry translational implications. UPR markers, particularly PERK at the interface, may serve as tissue adjuncts for biological stratification rather than diagnostic surrogates. Interventions that modulate placental stress remain experimental; however, mechanistic work in GDM and pre-eclampsia suggests that targeted mitigation of ER stress can influence trophoblast function and angiogenic signaling. Any attempt to modulate UPR in PAS must be carefully calibrated to avoid undermining IRE1-dependent antigen handling and EVT immune tolerance (14, 26).

In conclusion, by situating our PAS data within the comparative landscape of placental stress in early- and late-onset pre-eclampsia, fetal growth restriction, and gestational diabetes, we propose that PERK-dominant UPR activation is a defining feature of the stressed implantation interface in PAS. This aligns with a maternal-side pathophysiology rooted in decidual deficiency and scar biology and motivates mechanistic and translational studies that measure activated UPR components (e.g. phospho-PERK-eIF2a and spliced XBP1-XBP1s) and functional outputs at the true invasive front.