Impact of NSCLC metabolic remodeling on immunotherapy effectiveness

Glucose metabolism of tumor cells

Otto Warburg’s pioneering work displayed in the 1920s that tumor cells consume more glucose than normal cells. The phenomenon went by the name of the aerobic glycolysis or Warburg effect (Fig. 1) [22]. With the physiological condition in oxygenated environment, glucose is metabolized by cells by the course of glycolysis, the tricarboxylic acid (TCA) cycle and oxidative phosphorylation (OXPHOS), and eventually molecular oxidation is completed. This process which is dependent on oxygen is known as oxidative phosphorylation, and 32–38 ATP molecules are ultimately produced from a glucose molecule. However, pyruvate, being converted to lactate, is the byproduct of glycolysis. Merely two molecules of ATP are contained in per mol of glucose. For cancer cells，this metabolic shift includes an upregulation of biosynthetic along with bioenergetic pathways in order to maintain high proliferative rate and adapt metabolism (shown in Fig. 2).

Effect of glucose metabolism in NSCLC

There is more and more evidence suggesting metabolic remodeling is activated deeply in carcinogenesis and malignant progression in lung cancer (LC) [23]. High glucose uptake is discovered in NSCLC via positron emission tomography/computed tomography (PET/CT). Patients who has NSCLC with high glucose uptake can be identified with glucose-analog fluorodeoxyglucose (18F-FDG) PET/CT scans, which is on the rise as a potential instrument to choose patients for metabolically targeted anti-tumor treatment [24]. fLC tissue illustrates growing glucose contribution to tricarboxylic acid cycle (TCA) cycle in comparison with normal lung tissue, and lung cancer cells embody diverse glycolysis rates and mitochondrial abilities [25]. In addition, in LC cells, pyruvate carboxylase (PC) and pyruvate were overexpressed compared to normal lung tissues. Even when tumor cells are cultured without glucose, activation of alternative pathways for phosphopentose (PPP) still acts a vital part in tumorigenesis [26]. Moreover, PPP activation can trigger a good deal of glutathione and nicotinamide adenine dinucleotide phosphate (NADPH), whose oxidase activity and expression are associated with malignant biological behavior of LC. The function of inhibiting NADPH oxidase downregulates the proliferation and invasion of LC [27, 28]. Thus, Lung cancer cells alter their metabolism by taking away glucose for energy production through glycolysis, generating biomass through PPP and deprotonation, and counteracting oxidative stress through PPP.

Regulation of glucose transport and metabolism in NSCLC

There are two pathways of cellular glycolysis: oxygen-dependent and oxygen-independent pathway. These two pathways both depend on some familiar glucose transporters and glycolytic enzymes (Fig. 2). The oxygen-dependent mechanisms are mediated by transcription factor hypoxia-inducible factor 1-alpha (HIF-1α) pathway [29]. It has been shown that hypoxia promotes glycogen accumulation in cells through HIF-1α stabilization [30]. The recent study showed DPT (Deoxypodophyllotoxin) serves as an anticancer agent in NSCLC by suppressing HIF-1α activation at the protein level in NSCLC cells to reduce glycolysis [31]. Another study demonstrates that AC020978’s part in advancing cell growth and metabolic reprogramming in NSCLC, which uncovered that AC020978 could regulate PKM2-enhanced HIF-1α transcription activity [32].

Activation of the PI3K-AKT-mTOR signaling pathway mainly mediates oxygen non-dependent mechanisms of glucose utilization in LC [33, 34]. From a previous report,50–73% of NSCLC patients with poor prognosis exhibit high expression of AKT [35], while only minor patients with NSCLC embody mutations of PI3K and AKT.

Aside from activation of mutations, other molecules could also foster PI3K/AKT/MTOR signaling in LC. For instance, microRNA128 (Mir-128) plays an inhibitory role in LC progression by way of inhibition of AKT expression,thereby down-regulating glycolysis [36]. However, some other molecules have been reported activate glycolysis metabolism of LC by targeting PI3K/AKT pathway, such as oxidase 4 (NOX4) and α-enolase [37, 38].

Glucose transporters (GLUTs) belong to a protein family which is beneficial to the transferring of glucose into the blood. Several studies have shown that higher level of GLUT-1 protein was spotted to be remarkably connected with resistance to radiotherapy and poor disease particular overall survival of lung cancer [39, 40]. However, another study performed by Osugi et al. indicated that NSCLC patients with GLUT-1 expression failed to independently display poorer overall survival in comparison to GLUT1-negative patients [41]. Furthermore, GLUT-4 was also identified to be an appropriate potential target for epigenetic treatment or metabolic targeting in the management and NSCLC therapy [42].

Hexokinase (HK) is the first rate-limiting enzyme when cells begin glycolysis. HK is composed of four isoforms featured with diverse functions along with cellular positions. HK-II serves as an enzyme which catalyzes the phosphorylation of glucose, and it is the first step of glycolytic rate [43]. The expression levels of the HK-II protein in aggressive cancer cells greatly exceed those in normal cells [44]. Recently，supramolecular assemblies of new-type amphiphilic cell-penetrating peptides in order to target cancer cell mitochondria stand for an emerging instrument for suppressing tumor growth. The adopted strategy is designed to amplify the apoptotic stimuli by impairing the mitochondrial VDAC1 (voltage-dependent anion channel-1)-hexokinase-II (HK-II) interaction [45].

NSCLC cells have overexpressed phosphofructokinase (PFK) (Fig. 1) as well, driving glycolytic flux to grow. Among the members of PFKFB, PFKFB3 appears higher level of (740-fold) kinase activity and lesser level of bisphosphatase activity. Based on previous studies, PFKFB3 was widely expressed in diverse organs and tumor cells, including lung, gastric, breast, ovarian, and thyroid carcinomas. It causes shifts in metabolism, giving rise to the proliferation and survival of tumor cells [46, 47]. A recent study also demonstrates that targeting PFKFB3 restrained cell viability and glycolytic activity, which may be a promising therapeutic strategy in treating lung adenocarcinoma [48]. Another PFK-1 family member, PFKP, was found to be up-regulated in both NSCLC tissues and cell lines and is correlated with lung cancer cell proliferation and patient prognosis [49].

The M2 isoform of pyruvate kinase (PKM2), as a glycolytic terminal enzyme, catalyzes the last step of glycolysis, shifting the phosphate from phosphoenolpyruvate (PEP) to adenosine diphosphate (ADP) [43]. It has become a vital element that adjusts aerobic glycolysis in cancer cells [50]. The strategies of PKM2 inhibition or silencing [51] as well as activation [52, 53] have shared equal debates in literature owing to potential of treatment in limiting tumor growth.

Gopinath Prakasam found that knockdown of AMPK in cells silenced for PKM2 or PKM1 took on inhibiting growth and contributed to apoptosis [54]. While another study carried out by Li.et al. showed that a potential PKM2 activator, 0089–0022, serves as a promising anti-cancer therapy candidate in NSCLC [55].

Due to the Warburg effect, a large amount of lactic acid is synthesized from pyruvate [22]. Lactate generated by LDH is shifted by monocarboxylate anion transporters (MCT) so that an alkaline internal environment is maintained. This will bring benefits to metabolism [56]. As is well-known to all that MCT4 transfers lactate out of the cell and MCT1 moves the entry of lactate to tumor cells [57]. A previous study had shown cellular expression levels of MCT1 and MCT4 were in relation to invasion activity. This suggests inhibitors of the MCT maybe offer a new-model strategy in order to hinder cancer metastasis [58].

In short, there is evidence demonstrates that LC employs HIF-1α to improve glycolytic flux and to neutralize ROS via the upregulated activity of glucose importers (GLUT1, GLUT4), glycolytic enzymes (PFK, PK), and lactate transporters (MCT1, MCT4). Suppressing metabolic enzymes activity involved in glycolysis and lactate production may become promising new alternative therapeutic targets for lung cancer. Up to date, the majority of prior available evidence was concentrated on small single studies, which only involve partly LC metabolism in a limited number of primary human LC cell lines. Before transforming into clinical trials, more comprehensive preclinical investigations on LC metabolism in divergent phases of the disease are in demanded to enhance the effectiveness of these findings.

Glucose and glycolysis

The tumor microenvironment (TME) plays a vital role in tumor behavior and therapeutic effect [59]. Meanwhile, cancer cells can regulate a well-characterized metabolic phenotype that can deeply affect TME [60]. In addition, an emerging theme is that metabolic phenotypes are tissue specific and cancer subtype specific. As a highly heterogeneous disease covering a heterogeneous population, NSCLC is blessed with a complicated system to identify the disease state and progression [61]. Tumor-induced TME metabolic reprogramming as shown in Fig. 3 appears to be more common.

With the increase of metabolic activity of tumor cells, glucose and amino acids in tumor microenvironment were significantly deficient. Glucose utilization by tumors metabolically limits T cells, which paralyzes their ability to maintain aerobic glycolysis and anti-tumor activity [62–64]. This is partly attributed to T cells, like tumor cells, being metabolically dependent on glycolysis. Further, T cell metabolism is regulated by HIF-1α, PPARy signaling and the transcription factor C-MYC-associated pathway and nuclear receptor family pathway [65, 66].

However, tumor-restricted glucose utilization would not totally make T cells disrupted, presumably because T cells seek for available metabolic sources. For example，mitochondrial activation of CD8 + T cells by PPAR-γ agonists strengthens the anti-tumor immunity in T cells during PD-1 blockade [66]. Another experiment has demonstrated that reactivation of depleted T cells relies on reserve of lipids by fatty acid oxidation in T lymphocytes receiving PD-1 signals [67]. Thus, blocking these pathways with checkpoint inhibitors can partially rescue glycolysis and biosynthesis, thereby reversing the effector function of CD8 + T cells (Fig. 4).

On the basis of the previous studies, declining metabolic burden on effector T cells, that lies in tumor microenvironment, might cause durable and steady anti-tumor immune responses. For instance, inhibition of GLUT1 receptors may contribute to more useful T cells and strengthen anti-tumor immune responses, or under the circumstance of glycolysis inhibitors like 2-deoxyglucose (2-DG), antitumor function can be enhanced in primed T cells [68, 69].However，several other studies have demonstrated that inhibiting nutrient transporters and enzymes got involved in glucose metabolism of CD8 + T cells could regulate T cell differentiation and inhibit CD8 + T cell function under low-glucose conditions [70].

The Treg cells accumulate in the TME and play a critical role in dampening antitumor effect. According to growing evidence, Tregs are able to differentitate and survive due to low glucose availability imposed by tumor condition. This extreme environment requires Tregs to utilize oxidative phosphorylation (OXPHOS) as a source of energy. In fact, lactate and kynurenin, metabolic waste products of glycolysis pathways, inhibit conventional T cell activation and cytotoxicity in Treg cells [71–73]. Reprogramming tumor cells, for example, inhibits the infiltration of effector T cells (Teffs) or induces apoptosis, enhances the differentiation of regulatory T cell (Tregs). As a result of reprogramming tumor cells, lactic acid accumulates and carbon dioxide is released, thereby suppressing the immune system [74].

Contrary to CD8+ and CD4+ T cells, Treg metabolism dependes on external factors which include nutrient availability as well as TCR triggering and cytokine milieu, [75, 76]. It is likely that Treg cells are not affected by glucose competition in the tumor site because they are able to utilize alternative to glucose for energy [77]. Rather than glycolysis, the metabolism of Tregs is primarily based on the oxidation of fatty acids [78, 79]. Tregs rely primarily on FAO for self-maintenance, and they exhibit low mTOR activity, so these fatty acids provide the perfect soil for Treg maintenance [80].

It has been demonstrated that fatty acids, combined with in vitro inhibition of glucose uptake and glucose oxidation, lead to Treg differentiation [81]. Moreover, lipid uptake and oxidation are mandatory for the expression of Foxp3, as demonstrated in murine models [81]. In hypoxic tumors, hypoxia-inducible factor (HIF)-1* elicits pyruvate to exit mitochondria with OXPHOS, causing Tregs reliant upon fatty acids for mitochondrial metabolism. As a result, FAO plays an important role in cancer metabolism of Tregs [82].

Effects of acidic extracellular microenvironment on immune cells

One main driving force of the metabolic remodeling appearing in the TME is without doubt rendered by hypoxia and accumulation of lactate. Lactate production is able to be higher level (40-fold) in tumor cells, and lactate dehydrogenase (LDH) display a positive relationship with tumor volume and clinical severity, as well as prognosis [83, 84].

Table 1 summarizes the effects of tumor-derived lactic acid on tumor-infiltrating lymphocytes in TME.

As antigen presentation cells, tumor associated dendritic cells (TADCs) initiate and enhance antitumor immune responses. TADCs promote antitumor immune surveillance, as they are able to exhibit neoantigens to T cells, thereby initiating a T-cell-mediated immune response [85]. Studies have shown that lactic acid accumulation has a direct immunosuppressive effect on immune cells. Lactic acid inhibits monocyte activation and dendritic cell antigen presentation [85, 86].

Myeloid-derived suppressor cells (MDSC) are key components of protumor immune responses and their accumulation in immune organs relate to immunosuppression and cancer aggressiveness. Tumoral lactate production was also found to increases MDSCs and suppresses natural killer (NK) cell function, further accelerating the immunosuppressive microenvironment [87]. A recent study conducted by Baumann found that MDSC blocks T cells by transferring the glycolytic metabolite methylglyoxal, which acts as an immunosuppressant by consuming L-arginine in CD8+ T cells [88].

Tumor-associated macrophages (TAMs) serve as main components in TME [89].TAM, being highly plastic, is able to polarize to two primary phenotypes: the antitumor M1 (TAM1) together with the protumor M2 (TAM2). Several experiments and clinical trials demonstrate that TAMs are primarily of the M2 phenotype, which drive tumor progression and metastasis [90, 91] as well as suppression of antitumor immune responses [92].

Tumor-cell-derived lactic acid facilitate M2-polarization of TAMs through improved arginase and HIF-1α stabilization [93, 94]. After we treated TAMs from patients either lactate or conditioned medium from two tumor cell lines: the Lewis lung carcinoma (LLC) and the melanoma cell lines, TAMs show enhanced expression of HIF-1α and M2-polarization [93].

Therefore, lactic acid produced by cancer cells has a direct immunosuppressive effect, further initiating MDSC-mediated immunosuppression and driving M2 polarization by inhibiting the differentiation of monocyte derived dendritic cells. Tumor-derived lactate can also decrease T cell function by reducing lactate export through MCT1, which inhibits their capability to retain aerobic glycolysis [95]. Additionally, inhibition of effector T cell and boost of Treg in TME by lactate lead to enhanced immunosuppressive microenvironment [96, 97]. Hence, neutralization of an acidic environment is probably to have a significant implication on further the efficacy of ICB (Fig. 3).

Meanwhile, in a recent preclinical study, buffering the TME with bicarbonate administration able to limit tumor growth and improve antitumor responses in animals when merged with either anti-PD-1antibodies (Abs), anti-CTLA or adoptive T cell transfer in a melanoma model [98] .Although molecular mechanism underlying this combination were not clear，these results imply that reversing the acidic TME may increase the selectivity of treatments and meanwhile improve ICB therapy by increasing tumor infiltration by activated T lymphocytes. Fig. 4.

Ethics approval and consent to participate

This article does not contain any studies with human participants or animals performed by any of the authors.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.