Senescence at the crossroads of postpartum remodeling and tumorigenesis

Abstract

Cellular senescence has been linked to both beneficial and detrimental functions. Chiche, Djoual, Charifou and colleagues show that senescence supports normal postpartum mammary gland remodeling, but that when oncogenic events coincide with involution, senescent cells enhance tumorigenesis by regulating plasticity, invasion and metastasis.

The biology and function of senescent cells remains puzzling. First described as a permanent growth arrest that limits the proliferative capacity of cultured cells¹, senescence was later recognized as a powerful tumor-suppressive mechanism that halts the expansion of damaged or oncogene-stressed cells². Over the past decade, however, this paradigm has been challenged. Senescence is now understood as a dynamic and context-dependent cellular program, one that under specific physiological conditions can promote disease^3,4. Although largely associated with aging, beneficial functions of senescence in embryonic development and tissue repair have also been observed. Now, Chiche, Djoual, Charifou and colleagues show that during postpartum mammary gland involution – a period of extensive tissue remodeling after pregnancy – cells enter a state of senescence that is essential for normal tissue regression. Futher, when oncogenic events coincide with postpartum mammary gland involution, these senescent cells can be co-opted to promote tumor initiation, invasion and metastasis, which reveals how a normally tightly regulated, physiological stress response can become a driver of postpartum breast cancer⁵ (Fig. 1).

Fig. 1 |. The two sides of senescence in postpartum involution.

In normal healthy breast tissue undergoing involution, p16 expression in alveolar luminal epithelial cells leads to the induction of senescence and secretion of factors that breakdown the extracellular matrix (ECM) and recruit macrophages to complete the involution process, which returns the gland back to a prepregnant state. In the presence of an oncogenic stress or insult, the senescent cells can promote tumor formation and increased tumor cell plasticity, which results in postpartum breast cancer.

Senescent cells are defined by an essentially irreversible cell-cycle arrest, which can be dependent on p21^CIP1 or p16^INK4A signaling^6,7. Either way, these cells remain metabolically active and secrete a complex mixture of paracrine signaling molecules that is collectively termed the senescence-associated secretory phenotype (SASP)⁸. This secretory phenotype enables senescent cells to communicate with their environment, which leads to the recruitment of immune cells and extracellular matrix remodeling, and influences the fate and behavior of neighboring cells. This activity underlies the contradiction of senescence: although growth arrest protects against malignant transformation, sustained or misplaced SASP signaling can promote inflammation, tissue dysfunction and tumor progression^8,9.

The mammary gland is a prime example of an organ that retains physiological plasticity in adulthood. During pregnancy and lactation, mammary epithelial cells undergo rapid expansion to generate milk-producing alveolar luminal cells. Following lactation, the gland enters postpartum involution, a highly coordinated process of epithelial cell death, immune infiltration and adipocyte repopulation that restores the tissue to a near-prepregnancy state¹⁰. Involution closely resembles wound healing in both its molecular signatures and inflammatory milieu¹⁰, which makes this stage of development associated with a transient but well-documented increase in the risk of developing aggressive breast cancer, termed postpartum breast cancer^11,12.

Armed with this information, the authors asked whether senescence might be an integral component of postpartum involution. Using the MMTV-Neu mouse model to mimic a concurrent pregnancy and BlgCre^Tg/+;p16^f/f mice to generate involuting organoids, they show that senescence is induced specifically in alveolar luminal cells early during the involution process. They further link this response to p16^INK4A-dependent senescence activation, and further show it is not associated with increases in DNA damage. This is similar to developmental senescence and the senescence response that arises during wound healing, in which senescent cells engage a transient remodeling in the absence of DNA damage^13,14, but differs in that this senescent phenotype is independent of p21 activation. Indeed, senescence during mammary involution is regulated by the withdrawal of lactogenic hormones, which indicates this phenotype is hormonally programmed rather than stress induced. Importantly, the authors show that either genetic or pharmacological inhibition of these senescent cells disrupts mammary gland remodeling and delays involution, which suggests that rather than being passive bystanders, senescent cells instead have an essential physiological role in restoring tissue homeostasis after lactation.

One might argue that these findings place postpartum involution alongside embryonic development and wound healing as contexts in which transient senescence contributes positively to tissue remodeling^13,15. However, the most salient insight from the study emerges from when the authors examined senescence in the presence of an oncogenic insult. Using a MMTV-Neu mouse model of postpartum breast cancer, the authors implicate involution-associated senescent cells as key pathogenic intermediaries that become potent drivers of malignancy. They demonstrate that elimination of these senescent cells during the involution window significantly prolongs tumor latency, which reframes senescent cells as an active architect of the tumor-permissive postpartum niche rather than a passive marker of tissue remodelling, and thereby provides direct evidence that physiological senescence can be hijacked to accelerate tumorigenesis.

The authors go on to show that, mechanistically, senescent cells exert their tumor-promoting effects through paracrine signaling. They show that SASP factors released during involution enhance the plasticity of neighboring tumor-initiating cells, which enables them to adopt more invasive and adaptable phenotypes. Cellular plasticity is a defining feature of aggressive cancers and is closely linked to invasion and metastasis⁹. Consistent with this, the authors show that the presence of involution-associated senescent cells promotes tumor invasion and metastatic dissemination in vivo, which offers a potential mechanistic explanation for the poor prognosis associated with postpartum breast cancer.

The implications of this work extend beyond the postpartum mammary gland. It provides a paradigm to explain how senescence can suppress tumor formation early on, and yet later facilitate tumor progression. As with other well-established factors with contradictive roles in tumor biology, both timing and tissue context emerge here as decisive factors. During normal involution, senescence is transient and beneficial; however, when oncogenic mutations arise within this window, the same program reshapes the microenvironment in ways that favor malignancy.

These findings also carry important translational implications, particularly for cancer prevention, as they cement the postpartum period as a window of cancer risk, which highlights the need not only for greater awareness but also potentially tailored screening strategies as part of postpartum healthcare. More broadly, these findings highlight senescence as a context-dependent determinant of cancer risk, in which physiological senescence programs engaged during development, repair or endocrine transitions can become tumor-permissive if they persist or are dysregulated. Senolytic therapies are a new class of drug that were developed to eliminate senescent cells, and have shown initial promise in preclinical and early clinical trials of age-related diseases and cancer. However, this study supports and underscores the need for caution, as widespread elimination of senescent cells could interfere with essential physiological processes such as tissue repair and remodeling. In fact, one may now argue that a more targeted approach focused on pathological senescent states may result in a more beneficial senescence modulation that avoids the disruption of important and necessary normal functions. As maternal age and postpartum breast cancer incidence rise, this study highlights a narrowly defined biological window in which prevention may be both feasible and safe. However, future therapeutics will need to not only prevent cancer initiation and progression but also, at the same time, preserve the ‘normal’ beneficial roles of senescence.

Ultimately, this research reframes how we think about senescence and positions it at the intersection of tissue repair and cancer. Rather than labeling senescence as good or bad, the authors highlight a unifying biological mechanism that links regeneration, inflammation and tumorigenesis. Their findings challenge a simplistic view of senescence and instead emphasize its context-dependent nature. By revealing how the normal postpartum process can create conditions that are permissive to cancer, this work deepens our understanding of the intimate relationship between healing and disease, and reminds us that even the body’s most protective mechanisms can – under the wrong circumstances – become accomplices to cancer.