Metformin-based nanomedicines for reprogramming tumor immune microenvironment

Over the past decades, extensive studies have gradually unveiled the characteristics of the tumor microenvironment (TME), which is considered as an evolving entity composed of heterogeneous cell populations, abnormal vasculature, cytokines and extracellular matrix (ECM) 1 . There are various cell populations within the TME, including immune cells, tumor cells and stromal cells, and the specific proportion of each cell type varies in different cancer types 2 . The ECM is a non-cell component primarily secreted by cancer-associated fibroblasts (CAFs), and it serves as a reservoir for various immunosuppressive cytokines and growth factors 3 . Studies have indicated that the heterogeneous TME supports tumor progression and impedes immunotherapeutic action from multiple aspects. (I) A high interstitial pressure and a dense ECM within the TME are two physical barriers to block deep penetration of drugs and immune cells 4 ; (II) Various suppressive cytokines, which are secreted by tumor cells and immunosuppressive cells such as tumor-associated macrophages (TAMs), regulatory T (T reg ) cells and myeloid derived suppressor cells (MDSCs), facilitate tumor immunoevasion 5 ; (III) Overexpressed immune checkpoint molecules suppress T cell function and proliferation 6 ; (IV) Metabolic stress impairs anti-tumor immunity by depriving essential nutrients and allowing suppressive metabolite accumulation, leading to the exhaustion of cytotoxic T lymphocytes (CTLs) 7 and the proliferation of TAMs 8 and T reg cells 9 . It is noted that metabolic reprogramming of the TME has become as a novel approach to enhancing anti-tumor immunotherapeutic effects 10 - 15 . Metformin is commonly regarded as the primary medication for individuals with type 2 diabetes mellitus (T2DM) due to its proven ability to lower the blood sugar level as well as improve insulin sensitivity. In 1922, metformin was first synthesized, and its blood glucose-lowering function was confirmed in rabbits by 1929 16 . Finally in 1994, metformin was approved by the Food and Drug Administration (FDA) and it is currently widely used in clinical practice to treat T2DM. The most acknowledged anti-diabetic mechanisms of metformin involve in reducing hepatic gluconeogenesis and increasing intestinal secretion of glucagon-like peptide 1 (GLP1) through activating the adenosine monophosphate-activated protein kinase (AMPK) signaling pathway 17 , 18 . The extensively studied phosphatidylinositol-3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) signaling pathways are critical regulatory hubs for cell metabolism, survival, and proliferation 19 - 21 . Thus, inhibitors of PI3K and rat sarcoma (RAS) are considered as highly specific medicines approved to treat patients with cancer 22 , 23 . Interestingly, while the AMPK pathway is essential for maintaining cellular homeostasis, its tumor-suppressive mechanism remains unclear. Abundant clinical evidence suggests that diabetes occur frequently in patients with many kinds of cancer, indicating metformin as an AMPK pathway activator is a potential adjuvant in cancer management 24 . Emerging evidence shows that metformin-mediated AMPK activation is closely linked to its anti-tumor immune response. On the one hand, it directly suppresses tumor growth through inhibiting various anabolic processes, on the other hand, it reprograms the immunosuppressive TME by increasing the metabolic fitness of inflammatory immune cells 25 , 26 . Hence, metformin holds promise as an adjuvant of cancer treatment, extending its benefits beyond diabetes management. The poor pharmacokinetics and bioavailability of metformin hamper its clinical application in cancer therapy. Advances in nanotechnology promote the development of nanomedicines, which are a nano-formulation of therapeutic drugs including metformin. Nanomedicines have played an impactful role in the metabolic reprogramming approach 27 . Upon intravenous administration, vascular endothelial gaps and impaired lymphatic drainage allow nanomedicines to concentrate within tumor sites through the enhanced permeability and retention (EPR) effect 28 . Commonly overexpressed receptors on target cells can be utilized in preparing nanomedicines to target the cell type of interest via receptor-ligand interaction, thus achieving high-efficiency delivery of therapeutic drugs into tumor tissues and revitalizing anti-tumor immune response 29 , 30 . Besides, multiple therapeutic drugs can be integrated into one nanomedicine to induce immunogenic cell death (ICD) via different treatment modalities. For instance, photothermal therapy (PTT) can damage tumor cells through local heat, while chemodynamic therapy (CDT) and photodynamic therapy (PDT) generate reactive oxygen species (ROS) to induce tumor cell apoptosis 31 . Impaired or dead tumor cells release damage-associated molecular patterns (DAMPs) including ATP, calreticulin (CRT) and high mobility group box 1 (HMGB1) to activate dendritic cells (DCs) an

During tumor regression, T cells and their metabolic fitness are crucial for effective anti-tumor immunity. Importantly, the T cell metabolic profile changes dynamically during their development course. For example, naïve T cells exhibit a resting metabolic phenotype and they harness OXPHOS to produce ATP. Upon stimulation by tumor antigens and co-stimulatory molecules, glycolysis is significantly upregulated in naïve T cells and they become activated. Activated T eff cells are the major tumor-killing immune cells, and they exhibit an elevated level of glycolysis and OXPHOS 63 . After curbing tumor growth, the majority of T eff cells perish through apoptosis while a small subset survives and transforms into memory T (T m ) cells that respond to future encounters with tumor antigens for a long time 64 . Unfortunately, T cells can not compete with tumor cells for nutrients, resulting in an immunosuppressive TME. For example, glucose and glutamine deficiencies inhibit the activation of T eff cells and prevent interferon-γ (IFN-γ) production from them 7 . In addition, T m cells predominately perform OXPHOS and fatty acid oxidation (FAO) to maintain self-renewal and cope with metabolic stress 56 . The AMPK signaling pathway serves as an energy sensor in glucose-starved T eff cells, and it helps restoring a quiescent metabolic profile and facilitates T m cell generation. Therefore, promoting memory immunity of T cells through activating the AMPK pathway offers a promising strategy to enhance anti-tumor immune response. The precise impact of AMPK activation on T cell immunity remains to be unveiled. It is known that metformin activates AMPK to downregulate glycolysis in tumor cells partially through inhibiting mTORC1 and HIF-1α, and it may impair glycolysis in T eff cells in a similar manner, weakening their anti-tumor immune response. Nevertheless, AMPK is crucial for T eff cells to mount a rapid secondary immune response, as their differentiation into T m cells entails a metabolic transition of hyperactive glycolysis to quiescent OXPHOS 65 . Although glucose-derived acetyl-CoA is the primary fuel for the tricarboxylic acid (TCA) cycle, metformin-treated CAR-T cells upregulate acyl-coA synthetase short-chain family member 1 (ACSS1) to produce acetyl-CoA from acetate to fuel the TCA cycle 66 , enhancing their proliferation and tumor-killing capability through effective OXPHOS and energy production 67 . Beyond OXPHOS, FAO is essential for T m cell formation and tumor immunosurveillance. TNF receptor-associated factor 6 (TRAF6) promotes T m cell differentiation by upregulating FAO. AMPK activation and memory formation are impaired in TRAF6-deficient T cells, while they can be restored after metformin interventions, suggesting that metformin promotes T m cell generation via AMPK activation and FAO 68 . FAO and fatty acid synthesis (FAS) typically have opposite effects. Malonyl-CoA, an intermediate metabolite produced during FAS, can inhibit carnitine palmitoyl-transferase 1A (CPT1A) as a crucial enzyme involved in FAO 69 . AMPK activation induces inhibitory phosphorylation of Ser79 and Ser212 on two key FAS enzymes ACC1 and ACC2, respectively, thereby promoting FAO 70 . Interestingly, unlike T eff cells, T m cells primarily synthesize endogenous fatty acids via glycolysis rather than engulfing exogenous fatty acids via CD36. The synthesized lipids are broken down in lysosomes to produce free fatty acids for FAO in the mitochondria. Although this is an inefficient cycle for lipid metabolism in T m cells, primed fatty acids may stimulate T m cells for rapid reactivation 71 . Moreover, metformin has been reported to promote lysosomal lipolysis and mitochondrial FAO, indicating it may play a role in supporting T m cell development 72 . It is well established that T eff cell expansion primarily relies on glycolysis for ATP production, and their transition to quiescent T m cells triggers a metabolic shift. However, the precise role of metformin in regulating T m cell differentiation and enhancing anti-tumor immunity remains to be elucidated. Since OXPHOS and FAO occur primarily in the mitochondria, the mitochondrial fitness of immune cells is crucial against tumor progression. Dysfunctional tumor-infiltrating cells (TILs) often exhibit a metabolic disorder due to an insufficient mitochondrial mass. The loss of mitochondrial function in TILs has been ascribed to chronic tumor antigen stimulation, which persistently activates the AKT signaling pathway and suppresses the forkhead box O (Foxo)-PGC1α axis, impairing PGC1α-mediated mitochondrial biogenesis 73 . AMPK activation can enhance the PGC1α activity to promote mitochondrial biogenesis and OXPHOS in T cells, thereby supporting long-term anti-tumor immunity 74 . Beyond mitochondrial dysfunction, PD-1 on T cells interacting with PD-L1 within the TME also induces T cell exhaustion. Downregulation of PD-1 on T cells offers an alternative strategy for restoring anti-tumor immu

The effects of metformin may be exerted on other cell types within the TME ( Figure 2 ), such as T reg cells, MDSCs 84 and CAFs 85 . T reg cell, a subset of CD4 + T cells, can prevent autoimmune conditions and promote tumor progression by suppressing CTLs. The metformin-activated AMPK signaling pathway has been reported to inhibit tumor cell growth via inhibiting mTOR 86 , which could be applied to T reg cells. Indeed, AMPK activation has been reported to sustain T reg cell function via inhibiting mTOR and upregulating FAO in T1DM 87 . It was surprisingly found from one study that metformin treatment reduced the generation of tumor-infiltrating T reg cells in vitro through activating AMPK and subsequent mTOR signaling, and an elevated glycolysis/OXPHOS ratio was found to contribute to a decrease in the expression of the master transcription factor forkhead box protein P3 (Foxp3) 88 . AMPK activation to inhibit or activate mTOR in T reg cells should be investigated both in vitro and in vivo . MDSCs originate from hematopoietic stem cells, and their differentiation into mature myeloid cells is often blocked during tumor progression. Immature MDSCs contribute to an immunosuppressive TME via producing a variety of suppressive cytokines and metabolites 89 . Inhibiting the immunosuppressive activity of MDSCs represents an effective approach to reprogramming the TME. AMPK activation in MDSCs induced by metformin inhibits the signal transducer and activator of transcription 3 (STAT3) signaling pathway and its downstream ROS production, thus reducing their suppressive effect on CD4 + T cells 90 . In addition, metformin interventions result in AMPK phosphorylation to inhibit HIF-1α, thus reducing the expression of CD39 and CD73 on MDSCs and decreasing adenosine production, ultimately forming an improved immune-supportive TME 91 . Interestingly, in another study, the AMPK α1 subunit helped maintaining the immunosuppressive activity of MDSCs, while AMPK-deficient MDSCs exhibited tumoricidal activity by producing cytotoxic NO 92 . Similar to T reg cells, the role of metformin-mediated AMPK activation in modulating immune response of MDSCs is controversial, and the effects of AMPK activation may be distinguishable by investigating different subtypes of MDSCs. CAFs secrete major components of the dense ECM to form a physical barrier for infiltration of immune cells and therapeutics 93 , thus a strategy could be developed to reduce ECM formation and improve an immunosuppressive TME. In stroma-rich pancreatic ductal adenocarcinoma, the dense stroma could be effectively disrupted by metformin because it activated the AMPK pathway and inhibited the secretion of profibrogenic TGF-β to reduce ECM protein secretion by stellate cells 94 . Consequently, after metformin treatment, the dense ECM became thinner to allow the penetration of a gemcitabine-loaded nanomedicine 95 . In conclusion, metformin could act as a potent cancer therapy adjuvant, but its effects on immunometabolic modulation in other cell types within the TME remain to be unveiled through thorough investigations.

Nanomedicine is rapidly revolutionizing cancer treatment methods because of its multifunctionality and accommodation of multiple drugs within one nanocarrier, and it opens a new avenue for treating advanced cancers such as triple-negative breast cancer. Great efforts have been made to pursue smart nanomedicines which could preferably accumulate in tumor sites and specifically recognize targeted cells, thereby reducing the damage to normal cells 121 . Meanwhile, a variety of combination therapies can be realized in synergy with nanomedicine, including immunotherapy, radiotherapy, chemotherapy, CDT, PDT and PTT ( Figure 3 ). Metformin-derived nanomedicines are predominantly developed for targeting tumor cells, and they have been recently used to target immune cells and reprogram the immunosuppressive TME. In this section, we discuss the use of metformin-derived nanomedicines to reprogram a suppressive TME and promote tumor regression ( Table 2 ). Improving metformin delivery efficiency via nanocarriers Oral metformin is predominately absorbed by gastrointestinal tracts, and it exhibits unspecific biodistribution due to the broad presence of its corresponding transporters 146 . Besides, high hydrophilicity and rapid renal elimination of metformin lead to poor cellular permeation and short half-time in blood 147 , 148 . All of these above factors restrain the entrance of metformin to cancer cells in which it could interact with mitochondria and activate the AMPK pathway for tumor-suppressive effects. Achieving effective therapeutic outcomes requires enhanced pharmacological properties and sufficient drug accumulation in tumor tissues 149 . As a result, there is growing interest in developing nanomedicines that improve pharmacokinetics, targeting, and bioavailability of metformin by encapsulating it in nano-scale carriers 150 . Nanomedicines could prolong half-life and improve tissue distribution of metformin 151 . The EPR effect has been the primary mechanism for passive accumulation of nanomedicines in tumor tissues via leaky blood vessels and abnormal lymphatics since 1987 152 . A novel concept, active transport and retention (ATR) is proposed. During ATR, tumor endothelial cells can actively transport nanomedicines from the bloodstream to tumor sites through transcytosis 153 . Additional active transfer mechanisms include vesicular transport and blood vessel leakage facilitated by neutrophil extravasation, which help nanomedicines enter tumor tissues 154 . However, ATR is not universally accepted for different types of solid tumors, thus personalized nanomedicines have been developed to tailor to specific biological characteristics of different cancer types 155 . Active targeting molecules for specific biological characteristics of tumors, such as antibodies, peptides, and aptamers, have been identified and incorporated into nanomedicines to increase their binding affinity to tumor cells, thus enhancing tumor accumulation 156 . Similar active targeting strategies could hold promise in targeting immune cells 157 , 158 . It is important to note that evidence from the past decade supports that only 0.7% of nanoparticles successfully reach tumor tissues due to their elimination via the mononuclear phagocytic system 159 . Thus, optimization of the size, charge, and surface coating of nanocarriers is crucial to reduce their uptake by mononuclear phagocytes in the liver and spleen 160 . Therefore, nanocarriers with optimized physio-chemical properties could act as an effective delivery system for metformin. Nanomedicines could also enhance overall biocompatibility and pharmacokinetics of metformin while achieving a high localized concentration at tumor sites, resulting in improved efficacy and reduced toxicity 29 , 159 . Over the past decades, a variety of advanced nanocarriers have been developed to deliver therapeutics to tumor sites 161 - 163 . Polymeric nanoparticles are nanoscale biomaterials made from polymers, and poly lactic-co-glycolic acid (PLGA) nanoparticles are the most widely used due to their excellent biocompatibility 164 - 167 . PLGA nanoparticles have been reported to protect metformin from early release and realize sustainable release over 160 hours 168 . It is noted that polymetformin (Polymet) is a polymeric form of metformin with positive charge. Polymet has been employed to deliver siRNA and form nano-composites with therapeutic photothermal agents for tumor treatment 124 , 125 . Liposomes are an effective drug delivery system with improved pharmacokinetics and bioavailability 169 . It was reported that liposomes could improve the entrapment efficiency of metformin (~65%) 170 . Micelles are self-assembled nanoscale spheres composed of amphiphilic molecules, and they have a hydrophobic tail and a hydrophilic head 171 . Metformin-containing polymeric micelles could release metformin via pH-responsiveness and exhibit more potent cytotoxicity against breast cancer 172 . Hydrogels are highly bioco

Immunometabolic modulation of T cells via metformin has been extensively explored. Polymeric metformin has been used to enhance T cell infiltration via AMPK activation and improve the TME, and it has been combined with ICI therapy in colorectal cancer treatment 129 . Tumor relapse is a predominant indicator of poor outcomes in cancer patients, while long-living T m cells can prevent tumors from recurrence. It is well established that maintenance of T m cells is principally dependent on mitochondrial metabolism, such as OXPHOS and FAO 188 . Therefore, Chao et al. constructed a hydrogel scaffold to store and gradually release chimeric antigen receptor T (CAR-T) cells and metformin, and implanted the scaffold in the post-resection tumor site to evaluate the anti-tumor effect 67 . Interestingly, released metformin from the scaffold inhibited glycolysis and OXPHOS in tumor cells, while it helped strengthening the TCA cycle and mitochondrial respiration in the T cells. In CAR-T cells, upregulation of ACSS1 boosted acetyl-CoA production, fueling the TCA cycle and OXPHOS. Metabolic reprogramming of CAR-T cells enhanced their proliferation and secretion of effector cytokines. The scaffold fine-tuned CAR-T cells into a more persistent and memory-like phenotype, resulting in significant inhibition of primary and metastatic tumors in the HGC-27 tumor-bearing mice ( Figure 6 ). However, distinct metabolic influences by metformin on cancer cells and T cells remain unknown. We believe metformin improves T cell viability via altering their engagement with cancer cells, since metformin fails to exert such effects in the absence of cancer cells. Interestingly, different concentrations of metformin can exert absolutely different effects on the mitochondrial activity 189 , indicating the importance of designing a proper dose of metformin to tumor cells and T cells for activating effective anti-tumor effects. The exposure of tumor antigens to T cells is required for T cell activation, but long-term stimulation by tumor antigens as anti-tumor vaccines can exhaust T cells and block their differentiation to a memory-like phenotype. Metformin can downregulate the expression level of immunosuppressive PD-1 and prolong T m cell survival by strengthening their mitochondrial function and elevating their FAO level. Luo et al. developed a metformin-based tumor vaccine to induce generation of T m cells 126 . They initially treated the tumor lesion with PTT to acquire anti-tumor immunogenicity. The tumor antigens were collected and co-encapsulated with metformin and hollow gold nanoparticles as a photothermal agent into biodegradable PLGA microspheres to obtain TA-Met@MS as a vaccine. TA-Met@MS released pulsed tumor antigens and metformin under near infrared radiation to promote T cell activation and subsequent central T m cell formation via interfering with FAO. Thus, TA-Met@MS pronouncedly inhibited tumor growth and metastasis and a great number of CD8 + T m cells were produced in the 4T1 and B16F10 tumor-bearing mice. Since effectiveness of cancer vaccines relies on rapid response from T m cells, central T m cells could be endowed with a hyper proliferative ability in response to tumor antigen reencounter 190 . Notably, the metformin-activated AMPK signaling pathway promotes catabolism and T m cell differentiation, but it could impair T eff cell differentiation 46 , therefore, it is suggested to induce T m cell differentiation via metformin during the T cell contraction phase, which could maximize in maintaining anti-tumor immune response. Novel T cell-targeting nanomedicines have emerged for cancer treatment. For example, engineered T cells anchored with a nanomedicine resolved the issue of vasculature extravasation of nanomedicines and they precisely acted on T cells without any physiological changes 191 . More recently, a tri-specific nano-antibody has been reported to simultaneously bind to tumor cells via targeting PD-L1, and T cells and NK cells via targeting 4-1BB and natural killer group 2 member A (NKG2A) 192 . The tri-specific nano-antibody exhibited more effective targetability and better anti-tumor immunity compared with clinical monoclonal antibodies and bispecific monoclonal antibodies. Metformin could be combined with these innovative and effective strategies in the nano-medicinal format for advanced solid tumors. Emerging immunological evidence indicates precursor exhausted T cells in lymph nodes are emerging as novel targeting candidates, since these precursor cells can respond to ICIs and then proliferate and replenish functional T cells at the tumor site 193 . However, upon prolonged stimulation by tumor antigens, precursor exhausted T cells inevitably generate terminally exhausted progeny 194 . Mitochondrial dysfunction has been found to be linked to T cell exhaustion 195 , 196 , and studies have confirmed that PGC-1α-overexpressed T cells exhibit significant improvements in mitochondrial biogenesis and expansion 197

Metformin, a first-line treatment drug for T2DM to lower the blood glucose concentration through AMPK activation, has recently gained attention as a potential adjuvant in cancer therapy due to its direct and indirect anti-tumor effects. Metformin-mediated AMPK activation can induce PD-L1 degradation and interfere with anabolic processes in tumor cells via inhibiting mTOR. Additionally, inhibition of mitochondrial complex I by metformin can reduce oxygen consumption in tumor cells and remodel a hypoxic TME. Inspiringly, metformin reprograms the TME via enhancing the effector activity of anti-tumor immune cells. For example, AMPK activation in T cells favors energy metabolism and promotes memory phenotype formation, and it also contributes to anti-tumor repolarization of TAMs. However, direct administration of metformin has shown limited benefits in cancer therapy, which may be due to significant variations in the local concentration of metformin at tumor sites. To address this issue, metformin-containing nanomedicines have been developed to increase its localized concentration in tumor tissues, thereby revealing its role as an anti-tumor adjuvant and confirming its tumor-suppressing effects. The introduction of the nanomedicine formulation endows metformin with improved targetability and bioavailability through diversified strategies for targeting and responsive release in response to characteristic stimuli in the TME, in this way, the toxicity of metformin can be significantly diminished. A variety of nanomaterials have been explored to encapsulate metformin in nanomedicines. More encouragingly, metformin combined with other therapeutic drugs can be incorporated into one single nanomedicine to realize combination therapy. In summary, repurposing metformin in a nano-medicinal formulation holds great potential for effectively reprogramming an immunosuppressive TME and comprehensively activating anti-tumor immunity. Although metformin-containing nanomedicines have shown promise in reprogramming an immunosuppressive TME, there are very few fundamental mechanistic studies in this area. To fully leverage the potential of metformin in cancer nanomedicines, the following future research directions are highly recommended ( Figure 9 ). (I) Conducting systematical studies on the AMPK signaling network to identify specific targets that induce tumor-suppressive effects in different immune cells; (II) Enhancing efficiencies in accumulation and penetration of nanomedicines in the TME by optimizing their size, charge, coating, and ligands. In addition, exploring and designing immune-regulating nanomaterials for metformin-containing nanomedicines is conducive to enhancing anti-tumor immune response; (III) Improving active targetability and reducing off-target effects of metformin-based nanomedicines via exploring high-affinity peptides, antibodies, aptamers and developing cell-conjugated nanomedicines; (IV) Developing personalized nanomedicines through patient stratification by tissue biomarkers and biological characteristics of different cancer types; (V) Finally, reducing the cost, simplifying the preparation process, selecting the best administration routes, and mitigating toxicities from nano-formulations so that the nanomedicines can be readily translated into clinical use. In conclusion, metformin in a nanomedicine formulation has emerged as a promising adjuvant to activate the AMPK signaling pathway which is a key regulator in cellular metabolic activities, energy production and immune response, and it can reprogram the TME by remodeling an immunologically cold TME into an immunologically hot one to reactivate the suppressed anti-tumor immunity. In the nanomedicine formulation, metformin can work in synergy with other therapeutic agents or modalities to eliminate malignancies. The mechanisms of regulating immunometabolism within the TME via AMPK activation by metformin remain mysterious because of the complex interplay between the components of the AMPK signaling pathway and unveiling of these mechanisms could offer great potential in restoring suppressed anti-tumor immunity within the TME. Overall, repurposing metformin in a nanomedicine formulation to reprogram the TME will advance cancer treatment and benefit cancer patients.