Exosomes and ferroptosis: roles in tumour regulation and new cancer therapies

Exosomes directly inhibit the Fenton reaction

Iron accumulation contributes to ROS production through the Fenton reaction, promoting lipid peroxidation and leading to ferroptosis. Exosomes can reduce the intracellular iron content, which may inhibit the Fenton reaction and subsequent ferroptosis (Dixon et al., 2012).

Exosomes activate ferroptosis defence pathways

Exosomes modulate ferroptosis by inhibiting other pathways

CAF-derived exosomes inhibit ferroptosis by the miR-522 /ALOX15 axis (Zhang et al., 2020). ALOX15 is synthesized via the iron-catalysed enzymatic reaction (Doll & Conrad, 2017). ALOX15 upregulation causes excess polyethylene hydroperoxides to accumulate beyond the reducing capacity of GPX4, ultimately leading to ferroptosis (Wenzel et al., 2017). CAF-derived exosomal miR-522 inhibits ALOX15 in GC cells, reducing the accumulation of lipid ROS and suppressing ferroptosis. Conversely, increased exosomal miR-522 promotes ferroptosis (Zhang et al., 2020). This process is regulated at the posttranscriptional level (Zhang et al., 2020). EPC-derived exosomes inhibit ferroptosis through the miR-30e-5p/specific protein 1 (SP1)/adenosine monophosphate-activated protein kinase (AMPK) axis (Xia et al., 2022). miR-30e-5p targets SP1, and SP1 inhibits activation of the AMPK pathway (Xia et al., 2022), which inhibits ferroptosis through phosphorylation of acetyl-CoA carboxylase (ACC) (Lee et al., 2020). EPC-derived exosomes upregulate miR-30e-5p, inhibit SP1, and activate the AMPK pathway to inhibit ferroptosis (Xia et al., 2022). In addition to miRNAs, exosomal lncRNAs also inhibit ferroptosis. Bone marrow mesenchymal stem cell (BMSC)-derived exosomal lncRNA Mir9-3 host gene (lncRNA Mir9-3hg) inhibits ferroptosis in cardiomyocytes via the pumilio RNA binding family member 2 (Pum2)/peroxiredoxin 6 (PRDX6) axis (Zhang et al., 2022). LncRNA Mir9-3hg downregulates the expression of Pum2, which binds the PRDX6 promoter to suppress PRDX6 expression (Zhang et al., 2022). PRDX6 is a negative regulator of ferroptosis, and specific PRDX6 phospholipase A2 inhibitors enhance ferroptosis (Lu et al., 2019). BMSC-derived exosomal lncRNA Mir9-3hg inhibits Pum2 and upregulates PRDX6, thereby suppressing ferroptosis (Zhang et al., 2022).

Promoting exosome-inhibited ferroptosis to enhance chemotherapy

Exosomes cause chemoresistance in several cancers, and the underlying mechanism involves ferroptosis. For example, acquired chemoresistance in NSCLC and GC is associated with exosome-induced inhibition of ferroptosis (Song et al., 2021b; Zhang et al., 2020). The first-line treatment for NSCLC is cisplatin, which induces ferroptosis (Gridelli et al., 2018; Guo et al., 2018), but long-term cisplatin therapy leads to chemoresistance (MacDonagh et al., 2018). One mechanism underlying cisplatin chemoresistance is related to ferroptosis (Song et al., 2021b). Transfer of cisplatin-resistant NSCLC–derived exosomal miR-4443 to cisplatin-sensitive NSCLC cells upregulates FSP1 expression through METTL3 in an m6A-dependent manner (Song et al., 2021b). This process inhibits ferroptosis and causes cisplatin-sensitive NSCLC cells to develop resistance to cisplatin (Song et al., 2021b). Meanwhile, silencing miR-4443 expression with inhibitors rendered A549-R cells significantly sensitive to cisplatin (Song et al., 2021b). Chemotherapy toxicity stimulates the secretion of exosomal miR-522 from CAFs, inhibiting ferroptosis in GC cells and leading to acquired chemoresistance (Zhang et al., 2020). Chemotoxicity in CAFs upregulates ubiquitin-specific protease 7 (USP7), which is a drug target for overcoming chemoresistance and antitumour therapy (Lu et al., 2021; Yao et al., 2018). USP7 regulates deubiquitination of heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1), and elevation of USP7 levels leads to elevated levels of hnRNPA1 (Zhang et al., 2020). The hnRNPA1 is involved in the exosomal secretion of multiple miRNAs (Gao et al., 2019; Liu et al., 2021a). Knockout of USP7 or hnRNPA1 decreased miR-522 levels in the extracellular environment, leading to increased cell death and reduced resistance to chemotherapy (Zhang et al., 2020). Another hnRNPA1-associated pathway, involving ferroptosis-associated lncRNA (lncFERO) and SCD1, is also involved in exosome-ferroptosis effects (Zhang et al., 2021c). Chemotoxicity targeting ferroptosis promotes exosomal lncFERO (Exo-lncFERO) secretion in GC cells via the USP7/hnRNPA1 axis. GC-derived Exo-lncFERO enters gastric cancer stem cells (GCSCs), binds to SCD1 mRNA and recruits hnRNPA1 to promote SCD1 translation upregulation. This process inhibits ferroptosis and enhances chemoresistance in gastric cancer. Knockdown of hnRNPA1 in GCSCs blocked this effect (Zhang et al., 2021c). In summary, cargoes in exosomes, miRNAs and lncRNAs, are responsible for cancer chemoresistance.

Different tumour cell lines have different sensitivities to ferroptosis (Hangauer et al., 2017; Tsoi et al., 2018). A critical mechanism underlying these differences at the cellular level is caused by differences in prom2 expression (Brown et al., 2019). Ferroptotic stress (e.g., interference with GPX4, isolated cells, and extracellular matrix) induces prom2 expression in breast carcinoma cells (Brown et al., 2019). prom2Prom2 promotes the binding of intraluminal vesicles to iron-containing ferritin to form ferritin-containing multivesicular bodies. Transporting ferritin out of the cell via exosomes inhibits ferroptosis (Strzyz, 2020).

Some strategies that promote exosome-inhibited ferroptosis may prevent chemoresistance in cancer cells; for example, (1) inhibiting the secretion of specific exosomes in chemoresistance, (2) regulating FSP1 m6A modification (Song et al., 2021b), (3) decreasing the level of miR-522/lncFERO (Zhang et al., 2020; Zhang et al., 2021c), and (4) targeting prom2 (Brown et al., 2019). Promoting exosome-inhibited ferroptosis could be a novel approach to reduce the development of chemoresistance in tumour cells and improve the efficacy of chemotherapy.

Encapsulating exosomes with ferroptosis inducers to confer anticancer effects in target cells

Experimentally engineered exosomes hold vast therapeutic potential (Cheng et al., 2018; Feng et al., 2021; Shi et al., 2020b). Folic acid (FA)-modified exosomes targeting ferroptosis can be used for clinical applications. FA-modified exosomes containing the ferroptosis inducer erastin (Erastin@FA-Exo) target triple-negative breast cancer cells, and confer antitumour effects (Yu et al., 2019b). Erastin induces ferroptosis by inhibiting cystine/glutamate antitransporters (Kwon et al., 2020). The Erastin@FA-Exo complex improves the cellular uptake of erastin and suppresses cell proliferation better than Erastin@Exo and free erastin. Erastin@FA-Exo promotes ferroptosis by depleting cellular GSH and overproducing ROS (Yu et al., 2019b). Exosomal targeting of ferroptosis combined with immune modification and photodynamic therapy (PDT) effectively induces antitumour effects in HCC cells (Du et al., 2021a). PDT is a new tumour treatment that injects a photosensitizer such as rose bengal (RB) to accumulate in tumour tissue. Then, the tumour cells are irradiated with a specific laser wavelength to activate the photosensitizer (Wang et al., 2014a), thereby producing monooxygenase ions that specifically destroy tumour cells (Dolmans, Fukumura & Jain, 2003). CD47 is loaded into donor-cell exosomes to create Exos-CD47, which effectively evades mononuclear phagocyte–mediated phagocytosis (Du et al., 2021a). Erastin and RB were then encapsulated into Exos-CD47 (Erastin/RB@Exos-CD47). Erastin inhibited system Xc- and blocked cystine uptake into the cells, leading to GSH depletion and decreased GPX4 activity (Cao & Dixon, 2016). This process significantly increased lipid ROS accumulation, and induced ferroptosis in HCC cells. Erastin/RB@Exos-CD47 effectively exerts anti-HCC effects in in vivo and in vitro assays, and has much lower liver toxicity than the control group (Erastin/RB@Exos) (Du et al., 2021a).

Developing therapies that combine exosomal inhibitors and ferroptosis inducers

Complexes containing exosomal PD-L1 derived from tumour cells (e.g., melanoma) suppress T cell activity and lead to resistance to tumour therapy (Page et al., 2014; Poggio et al., 2019). Ferroptosis is involved in T cell immunity and tumour resistance (Wang et al., 2019a; Wang et al., 2019b; Wang et al., 2019c). Wang and colleagues developed HACA-Fe@GW4869 nanoparticles (HGF NPs), which combine an exosome inhibitor (GW4869) with a ferroptosis promoter (Fe³⁺) to stimulate an antitumour response in melanoma cells (Wang et al., 2021). GW4869 inhibits the secretion of exosomal PD-L1, which triggers T cell activation and promotes interferon gamma (IFN-γ) secretion. Subsequently, SLC7A11 and SLC3A2 in the tumour cell cytosol are inhibited by IFN- γ, cystine is reduced, GSH levels are decreased, and GPX4 is suppressed, thereby promoting ferroptosis. Fe³⁺ directly promotes ferroptosis. The addition of the ferroptosis inhibitor liproxstatin reduced this effect. Subsequently, this team used modified semiconductor polymers, Fe³⁺, and GW4869 to develop novel phototheranostic metal-phenolic networks (PFG MPNs) (Xie et al., 2022). PFG MPNs also possess GW4869 (to block exosomal PD-L1) and Fe³⁺ (to promote ferroptosis) (Xie et al., 2022). Moreover, PFG MPNs promote dendritic cells maturation upon integrated laser irradiation (Xie et al., 2022), enhancing antitumour therapy. PFG MPNs have excellent near-infrared (NIR) type II fluorescence/photoacoustic imaging performance under NIR laser irradiation (Xie et al., 2022). Together with photothermal therapy, PFG MPNs may be used for precise malignancy immunotherapy. These results demonstrate that the combination of exosomes and ferroptosis for tumour therapy has excellent potential for clinical applications (Table 2).

Mechanism underlying exosomal inhibition of ferroptosis

The upregulation of GPX4 expression and activity has been reported in ferroptosis, but the effects on other Xc-GSH-GPX4 axis genes, such as SLC7A11 and SLC3A2, have been studied less thoroughly. Exosomal miR-4443 modulates FSP1 m6A modification–mediated ferroptosis and facilitates cisplatin resistance in NSCLC. Furthermore, exosomal miR-4443 may regulate ferroptosis-related genes other than FSP1, so the specific pathways of action remain to be established (Song et al., 2021b). There are no reports on exosomal regulation of the third ferroptosis defence system, the GCH1-BH4 axis. ALOX15, the AMPK pathway, and PRDX6 are involved in the exosome-ferroptosis effect, but the other ferroptosis pathways remain to be explored. Future research should conduct a deep pathway study to achieve a more comprehensive molecular understanding of exosomal inhibition of ferroptosis.

Exosome-based drug delivery systems

Cisplatin-resistant NSCLC–derived exosomal miR-4443 promotes cisplatin resistance in NSCLC by regulating FSP1 m6A modification–mediated ferroptosis (Song et al., 2021b). This suggests that future studies could reduce cisplatin resistance and develop new anticancer strategies by restoring METTL3/FSP1–mediated ferroptosis in tumour cells.

Cells release iron-containing exosomes by expressing prom2, which transports iron out of the cell and thereby suppresses ferroptosis (Strzyz, 2020). Recent studies have attempted to reverse ferroptosis in cancer cells by inhibiting prom2 transcription (Brown et al., 2021). Heat shock factor 1 (HSF1) positively regulates Prom2 transcription. This suggests that Prom2 transcription may be blocked with HSF1 inhibitors, thereby sensitizing chemoresistant cancer cells to drugs that induce ferroptosis. However, the hypothesis still needs to be studied and more prom2 inhibitors must be tested.

Exosomes have become an active topic of current research as a drug delivery system (Patil, Sawant & Kunda, 2020). Studies on exosomes loaded with anticancer drugs targeting ferroptosis are limited to erastin acting on the system Xc⁻. In the future, additional anticancer drugs may be developed to target different ferroptosis pathways. For example, a first-line therapeutic agent for glioblastoma (temozolomide) may induce ferroptosis by targeting DMT1 expression in glioblastoma cells, which partially inhibits cell growth (Song et al., 2021a).

Exosomes for cancer diagnosis and prognosis

Exosomes have an important role in liquid biopsies for early detection and prognosis prediction of cancer (Li et al., 2021b). Altered ferroptosis markers in exosomes may be useful biomarkers for cancer screening, such as the early detection of HCC (Sanchez et al., 2021) and PC (Yi et al., 2021). The lipid composition of HCC and PC cell-derived exosomes is altered, and pathway analysis implicates ferroptosis. Lipidomic profiling in plasma exosomes plays a role in the early detection of HCC in patients with cirrhosis (Sanchez et al., 2021), and molecular lipids in urinary exosomes can be used as biomarkers for PC (Skotland et al., 2017). Ferroptotic pancreatic ductal adenocarcinoma cells (PDACs) with exosomal KRAS^G12D may provide information about the prognosis of pancreatic cancer (Dai et al., 2020). Oxidative-stressed PDACs produce autophagy-dependent ferroptosis, releasing KRAS^G12D, which is packaged extracellularly as Exo-KRAS^G12D (Dai et al., 2020). Exo-KRAS^G12D activates signal transducer and activator of transcription 3 (STAT3)-dependent fatty acid oxidation pathways, polarizing tumour-associated macrophages to an M2-like native phenotype and leading to poor prognosis in pancreatic cancer patients (Dai et al., 2020). This implies that binding ferroptosis to exosomes holds potential in liquid biopsies of tumours.