Abstract
Background/Aim: Recently, we demonstrated that cancer dormancy is initiated within the lymphovascular tumor embolus and consists of decreased proliferation and lower mammalian target of rapamycin (mTOR) activity. In the present study, we investigated other intersecting metabolism-signaling pathways that may ultimately determine whether the lymphovascular tumor embolus remains dormant or undergoes cell death. Materials and Methods: The present study exploited a singular patient-derived xenograft (PDX) of inflammatory breast cancer (Mary-X) that spontaneously forms high density spheroids, the in vitro equivalent of emboli. The AMPK metabolic checkpoint pathway, the mTOR nutrient-responsive cell growth pathway, the P13K/Akt intracellular quiescence regulating pathway, and the calpain-mediated E-cadherin proteolytic pathway responsible for spontaneous spheroid-genesis were also investigated, to determine their relative contributions to dormancy. Results: The levels of phosphorylated AMPK proteins (AMPKα and β subunits) decreased gradually with the formation of MARY-X spheroids in vitro. Rapamycin down-regulated mTOR activity, yet dormancy persisted. LY294002, a PI3K/Akt inhibitor, completely abolished mTOR and induced spheroid disadherence and apoptosis. Compound C (AMPK inhibitor) up-regulated mTOR and induced spheroid disadherence and apoptosis. Increasing cellular metabolism led to cell death, even in enriched medium, whereas growing the spheroids in serum-free media (starvation) did not result in further mTOR inhibition, and dormancy was maintained. Conclusion: An increase in our understanding of dormancy from the standpoint of internal signaling pathways might ultimately provide clues to the external stimuli (starvation, hypoxia or other not yet understood phenomena) that act through these pathways to maintain or disrupt dormancy.
- Inflammatory breast cancer
- analysis of variance
- patient-derived xenograft
- E-cadherin/N-terminal fragment-1
- apoptotic index
- mammalian target of rapamycin
- 5′ AMP-activated protein kinase
- phosphoinositide 3-kinase
Cancer dormancy, followed by recurrence remains a poorly understood phenomenon in both cancer biology and oncology (1-5). In patients, dormancy refers to the period between the initial detection and treatment of cancer, followed by remission, and the subsequent recurrence months or even years later (6-11). Although some cancers during their latency period are still treated with adjuvant therapy (hormonal, chemo- or immunotherapy), the vast majority of cancers during dormancy undergo only surveillance or expectant management (12).
Relapse from cancer dormancy can occur either locally (near the site of the primary cancer) or systemically (metastatic site) (13). The epicenter for both is thought to be the so-called micrometastasis, a clump of tumor cells that has escaped the confines of the primary cancer through the phenomenon of lymphovascular invasion, a step which is also poorly understood and a step which some have called, “a metastasis caught in the act” (14). Recently, we observed that cancer dormancy is indeed first initiated by the lymphovascular tumor embolus and consists of decreased proliferation and decreased mammalian target of rapamycin (mTOR) activity (15). The present study continued to exploit a patient-derived xenograft (PDX) of inflammatory breast cancer (IBC) (Mary-X) that spontaneously forms high density spheroids, the in vitro equivalent of emboli (16-22). In the present study, we investigated additional intersecting metabolism-signaling pathways that ultimately determine whether the lymphovascular tumor embolus remains dormant or undergoes cell death. We specifically investigated AMPK, a metabolic checkpoint pathway, mTOR, a nutrient-responsive cell growth pathway, P13K/Akt, an intracellular quiescence regulating pathway, and a calpain-mediated E-cadherin proteolytic pathway responsible for spontaneous spheroid-genesis, to determine their relative contributions to dormancy.
Materials and Methods
Institutional approvals. Mary-X was derived from a patient with a biopsy-proven diagnosis of IBC in the 1990s and developed into a patient-derived transplantable xenograft (PDX). Studies were conducted under UCLA’s Animal Research Committee (Certification 95-127-11). The xenograft has been stable for over 30 years of passage. Recent studies were conducted at Meharry Medical College, OLAW D16-00261 (A3420-01), IACUC protocol 24-02-1443.
ATCC patent deposits and cell identification. Mary-X and its in vitro derived spheroids were deposited in the ATCC cell repository (Manassas, VA, USA) as PTA-2737 and PTA-27376, respectfully, and recently verified and re-verified to be both novel and human in origin (STRA4993).
Initial xenograft studies. Four-week-old female athymic (nude) mice on BALB/c backgrounds, purchased from Anticancer, Inc. (San Diego, CA, USA) were derived from their respective breeding colonies.
In vitro studies with Mary-X spheroids. Mary-X was placed in culture, resulting in the formation of loose aggregates in suspension. These aggregates then tightened into spheroids over the next 24 h and remained in suspension culture (16-22). Spheroids were seeded on 24-well plates in DMEM supplemented with 10% FBS and treated with different inhibitors for 24 h. Spheroids were also grown in the complete absence of FBS, mimicking cell culture conditions of starvation. The spheroids were observed with phase contrast microscopy.
Flow cytometric studies and measurement of apoptosis. Mary-X spheroids were digested using trypsin (0.05%) (Invitrogen, Carlsbad, CA, USA) for 10 min followed by neutralization in DMEM containing 10% fetal bovine serum (Invitrogen). After neutralization, single cells were collected and analyzed for apoptosis using an apoptosis assay, which included Annexin V-FLUOS (early apoptosis) and propidium iodide (late apoptosis) (Roche Diagnostics, Penzberg, Germany), followed by flow cytometric analysis. A total of 100,000 cells from 100 disadhered spheroids were analyzed on a Coulter Epics XL flow cytometer (Beckman Coulter, Inc., Fullerton, CA, USA), and the overall percentage of apoptotic cells, or Apoptotic Index (AI), was determined.
Inhibitors and antibodies. The inhibitors (pathways) used included rapamycin (mTOR), U0126 (MAPK), LY294002 (P13K/Akt), Compound C (AMPK), and calpeptin (calpain proteolysis of E-cadherin), all purchased from Thermo Fisher Scientific (Waltham, MA, USA). We used the following antibodies for western blot studies: AMPK Sampler Kit (1:1,000 dilution, Rabbit, #9957), mTOR Substrates Antibody Sampler Kit (1:1,000 dilution, Rabbit, #9862) (all from Cell Signaling Technology, Danvers, MA, USA) and E-Cadherin [rabbit anti-human ectodomain E-cadherin (H108) (Santa Cruz Biotechnology, Santa Cruz, CA, USA)].
For inhibitor treatment, stock solutions were prepared in DMSO at 1,000× the working concentrations. The inhibitors were added 24 h after seeding, and the spheroids of Mary-X were treated accordingly before being harvested for western blot analysis.
Western blot analysis. The collected cells or spheroids were washed in cold PBS and then suspended in Laemmli Sample Buffer (#1610737. Bio-Rad, Hercules, CA, USA) with β-mercaptoethanol and boiled for 10 min. Whole-cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on precast 4-20% Mini-Protean TGX gels (Bio-Rad), transferred to PVDF membranes, and probed with the indicated antibodies, followed by detection using an anti-rabbit IgG, HRP-linked antibody (Cell Signaling Technology). Rabbit mAb (13E5) was used to detect the housekeeping protein ACTB, serving as a control for lane loading (24E10; 1:1,000 dilution, Rabbit mAb #3195; Cell Signaling Technology).
Bound antibodies were detected using a chemiluminescent detection system (West Femto) (Pierce Biotechnology, Waltham, MA, USA) according to the manufacturer’s instructions.
ImageJ analysis. Quantification of protein band intensities was performed with ImageJ (NIH, Boston, MA, USA) (23).
Statistical analysis. For AI, means±SD values were determined. All experiments were performed by examining 100 spheroids, which were disadhered into 100,000 cells. Western blot studies were performed in quadruplicate, and band intensities were compared. All stated or calculated differences were considered statistically significant, as assessed by both the two-tailed Student’s t-test and ANOVA.
Results
Growth studies on Mary-X and Mary-X spheroids. Mary-X exhibited its diagnostic signature of overlying murine skin erythema (Figure 1A), which was attributed to florid lymphovascular tumor emboli, especially in the dermal lymphatics (Figure 1B). Mary-X spheroids (Figure 1C), identified by Principal Component Analysis as the in vitro equivalent of lymphovascular tumor emboli (21, 22), spontaneously formed in suspension culture. Western blot analysis revealed that these spheroids exhibited both over-expressed full length E-cadherin and a calpain-mediated proteolytic fragment, E-cad/NTF1 (Figure 1D). This fragment could be abolished by pretreating the early spheroids with calpeptin (Figure 1E).
Metabolic pathway studies of Mary-X spheroids. Progressive decrease of AMPK activity with Mary-X spheroid-genesis. AMPK activity decreased during spontaneous spheroid-genesis of Mary-X (Figure 2A and B). While both AMPKα and AMPKβ subunits were stably expressed during Mary-X spheroid-genesis, the levels of activated or phosphorylated AMPKα (Thr172) and AMPKβ (Ser182) subunits decreased at late or end-stage spheroid-genesis when the cellular density of the spheroids was highest (Figure 2A and B).
Alterations of mTOR activity with spheroid-genesis and calpain inhibition. The activities of mTOR and its substrates, p70-S6K and 4E-BP1, showed divergent changes during the formation of spheroids: the levels of phosphorylated p70-S6K decreased, while phosphorylated 4E-BP1 levels increased (Figure 3A). However, inhibition of calpain-mediated cleavage of E-cadherin by calpeptin increased the levels of phosphorylated p70-S6K (Figure 3B).
Alterations of mTOR activity with mTOR and P13K/Akt inhibition. The PI3K/Akt pathway played an important role in the control of mTOR activity in Mary-X spheroids. PI3K/Akt inhibitor (LY294002) decreased p-mTOR(Ser2481) and p-mTOR(Ser2448) (Figure 4A) and completely abolished the activities of mTOR substrates p-p70 S6 kinase (Figure 4B), p-4E-BP1 (Figure 4C), and p-PRAS40 (Figure 4D). In contrast, the mTOR inhibitor Rapamycin led to a down-regulation in the activity of p70 S6 kinase (Figure 4B) but had a lesser effect on 4E-PB1 (Figure 4C). U0126, an inhibitor of mitogen-activated protein kinase (MAPK), showed no effect.
Alterations of mTOR activity with AMPK inhibition and starvation. The AMPK inhibitor (Compound C) blocked the activity of AMPK, which resulted in the up-regulation of the mTOR pathway and its downstream substrates (Figure 5A). However, in fetal bovine serum-free medium (starvation), the mTOR pathway was not inhibited but was instead mildly up-regulated (Figure 5B).
Net effects on dormancy with metabolic pathway disruptions. Disruptions of calpain-mediated E-cadherin proteolysis by calpeptin, inhibition of the P13K/Akt pathway by LY294002, and inhibition of AMPK by Compound C all terminated dormancy by triggering spheroid disadherence and profound apoptosis (p<0.01), whereas rapamycin inhibition of mTOR and serum starvation both sustained dormancy (p>0.1) (Figure 6A and B).
Discussion
Metastases, arising from micrometastases, can happen years or even decades after primary cancer treatment because these residual tumor cells enter dormancy and evade therapies (1-14). Our recent studies have suggested that dormancy may be first initiated within the lymphovascular embolus (15), but what triggers or induces dormancy in the first place remains unknown. Conversely, dormant tumor cell clusters may reside as small clusters of quiescent cells or alternatively small indolent micrometastases where cellular proliferation is balanced by apoptosis (24). Although it has been speculated that dormant tumor clumps can exit dormancy and begin metastatic growth when the microenvironment is altered, we really do not know what governs the release of dormancy. Understanding both the entrance to and exit from dormancy is key to designing potential therapeutic strategies that effectively prevent metastases and recurrence by targeting dormant tumor cells (25). The transition process, in which the crosstalk of tumor cell clusters within their microenvironment leads to the establishment of or exit from dormancy, is difficult if not impossible to monitor in vivo (12, 13). To date, and to our best knowledge, there are no imaging moieties to detect dormant micrometastases in patients and monitor their progression and no animal models in which to induce or extinguish dormancy.
Mary-X, a PDX of IBC that was established in our lab (16-22) and its derived spheroids is an ideal model to study the transition of tumor cells from a proliferative to a dormant state. Our previous studies had shown that a multi-enzyme cascade of E-cadherin proteolysis was required for the formation of tight structures of both spheroids in vitro and lymphovascular emboli in vivo (21, 22). Among those proteases, calpain 2-mediated E-cadherin proteolysis played a key role. Accompanied by E-cadherin proteolysis during their formation, the proliferation index of both lymphovascular emboli and Mary-X spheroids specifically decreased over time, leading to G0/G1 cell cycle arrest. Despite this, the cells did not undergo apoptosis or non-apoptotic necrosis and remained viable, retaining full tumorigenicity when reinjected into mice after periods of up to six months (15). This is the classic definition of dormancy. To our knowledge, this singular PDX model is the only one that forms florid lymphovascular tumor emboli in vivo, gives rise to spontaneous spheroid-genesis in vitro, and exhibits dormancy in both its in vivo lymphovascular tumor emboli and in vitro spheroids. Therefore, this unique PDX model is the ideal model to investigate the multiple metabolism-signaling pathways that might regulate dormancy.
Because it is expected that dormant cancer cells would exhibit reduced metabolism, we initially investigated the mTOR pathway and found that its activity was indeed decreased during spheroid-genesis. Because multiple intersecting metabolism signaling pathways (26-40) could have also played a role in regulating dormancy, we examined them in the present study.
In this study, we investigated the AMP-activated protein kinase (AMPK) pathway, a nutrient-responsive metabolic checkpoint pathway coordinating cell growth with energy status (26-30), the mammalian target of rapamycin (mTOR) pathway, a pathway that is a highly conserved regulator of cell growth found in all eukaryotes (31-36) and the phosphoinositide 3-kinase (PI3K) pathway, a pathway stimulated by diverse oncogenes and growth factor receptors and a pathway generally thought to exhibit increased activity in most cancers (37-40).
Because our previous Mary-X studies demonstrated transcriptome equivalence between xenograft-generated spheroids in vitro and lymphovascular emboli in vivo, with both structures demonstrating E-cadherin over-expression and specific proteolytic processing that produces distinct E-cadherin fragments – particularly the calpain 2-generated E-cad/NTF1 present during late spheroid-genesis (21, 22) – we also investigated the relationship between calpain-mediated E-cadherin proteolysis and various metabolism-signaling pathways to gain insight into the mechanisms of dormancy initiation. The outcome of our investigations revealed that the high-density Mary-X spheroids either maintained or abolished dormancy depending on specific perturbations of the multiple intersecting metabolism signaling pathways.
Inhibiting calpain proteolysis of E-cadherin and the generation of E-cad/NTF1 resulted in profound apoptosis. Since the high-density Mary-X spheroids also demonstrated a decrease in AMPK activity and since it had been reported that E-cadherin inhibits the activity of LKB1, a protein upstream of AMPK, which activates both AMPK and several AMPK-related kinases (26-30), the decrease in AMPK activity in the MARY-X spheroids was likely due to the homophilic full length E-cadherin interactions, further strengthened by E-cad/NTF1, which mediates the formation of tight spheroid aggregates. Inhibiting calpain with calpeptin, which disrupts the generation of E-cad/NTF1, weakened the adhesion of the Mary-X spheroids, promoted disadherence, and triggered apoptosis. Whether these changes altered metabolism or were the result of altered metabolism remain unanswered.
Inactivating the P13K/Akt pathway through the inhibitor LY294002 completely abolished the activities of both p70 S6 kinase and 4E-BP1, the substrates of mTOR. The treated spheroids dissociated, and the cells underwent a profound degree of apoptosis. In this case, the complete block of mTOR activity reduced metabolism to a threshold incompatible with cell survival.
Inactivating the AMPK pathway with Compound C up-regulated the mTOR pathway and increased the downstream activities of both p70 S6 kinase and 4E-BP1. This meant that the cells within the Mary-X spheroids had increased metabolism; however, this increased metabolism paradoxically led to the abolishment of dormancy and resulted in cell death.
Ironically, inhibiting mTOR with rapamycin led to down-regulation of the activity of p70 S6 kinase and had a lesser effect on 4E-BP1 activity. This resulted in reduced metabolism but the maintenance of dormancy. Similarly, growing Mary-X spheroids under starvation (no FBS), which activated AMPK and led to decreased intracellular ATP levels and increased intracellular AMP, had minimal effects on mTOR. Cellular metabolism remained low, and dormancy was maintained.
Conclusion
Alterations in the first three metabolic-signaling pathways either increase mTOR activity and cellular metabolism, paradoxically causing cell death, or decrease mTOR to levels too low to sustain critical metabolic functions, resulting in cell death. Alterations in the fourth and fifth metabolic-signaling pathways only minimally affect metabolism, allowing dormancy to be maintained. An increase in our understanding of dormancy from the standpoint of internal signaling pathways might ultimately provide insights into how external stimuli (starvation, hypoxia or other not yet understood phenomena) interact with these pathways to maintain or disrupt dormancy.
Acknowledgements
The Authors wish to thank Meharry Medical College Instructional and Informational Technology Services for facilitating videoconferencing coauthor communications during the duration of the study.
Footnotes
Authors’ Contributions
All Authors made an intellectual contribution to the work. Yin Ye carried out the vast majority of the in vitro spheroid-genesis experiments and provided a draft of the manuscript. Justin Wang carried out the in vitro apoptosis experiments. Jordan Dillard carried out serial propagations of the PDX. Sanford H. Barsky supervised all of the experiments and re-wrote portions of the manuscript, which was reviewed by all of the Authors.
Conflicts of Interest
The Authors declare that they, at the present time, have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. None of the sources of support listed influenced the collection, analysis, and interpretation of data, the generation of the hypothesis, the writing of the manuscript or the decision to submit the manuscript for publication.
Funding
This work was supported by the Department of Defense Breast Cancer Research Program Grants BC990959, BC024258, BC053405. The work was also supported by the Dr. Carolyn S. Glaubensklee Endowment as well as Meharry Medical College funds and its Translational Pathology Shared Resource Core, supported by NIH U54CA163069.
- Received July 18, 2024.
- Revision received August 13, 2024.
- Accepted August 17, 2024.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).