Caloric restriction and metformin selectively improved LKB1-mutated NSCLC tumor response to chemo- and chemo-immunotherapy

Abstract

Background: About 10% of NSCLCs are mutated in KRAS and impaired in STK11/LKB1, a genetic background associated with poor prognosis, caused by an increase in metastatic burden and resistance to standard therapy. LKB1 is a protein involved in a number of biological processes and is particularly important for its role in the regulation of cell metabolism. LKB1 alterations lead to protein loss that causes mitochondria and metabolic dysfunction that makes cells unable to respond to metabolic stress. Different studies have shown how it is possible to interfere with cancer metabolism using metformin and caloric restriction (CR) and both modify the tumor microenvironment (TME), stimulating the switch from "cold" to "hot". Given the poor therapeutic response of KRAS^mut/LKB1^mut patients, and the role of LKB1 in cell metabolism, we examined whether the addition of metformin and CR enhanced the response to chemo or chemo-immunotherapy in LKB1 impaired tumors.

Methods: Mouse cell lines were derived from lung nodules of transgenic mice carrying KRAS^G12D with either functional LKB1 (KRAS^G12D/LKB1^wt) or mutated LKB1 (KRAS^G12D/LKB1^mut). Once stabilized in vitro, these cell lines were inoculated subcutaneously and intramuscularly into immunocompetent mice. Additionally, a patient-derived xenograft (PDX) model was established by directly implanting tumor fragments from patient into immunocompromised mice. The mice bearing these tumor models were subjected to treatment with chemotherapy or chemo-immunotherapy, both as standalone regimens and in combination with metformin and CR.

Results: Our preclinical results indicate that in NSCLC KRAS^mut/LKB1^mut tumors, metformin and CR do enhance the response to chemo and chemo-immunotherapy, inducing a metabolic stress condition that these tumors are not able to overcome. Analysis of immune infiltrating cells did not bring to light any strong correlation between the TME immune-modulation and the tumor response to metformin and CR.

Conclusion: Our in vitro and in vivo preliminary studies confirm our hypothesis that the addition of metformin and CR is able to improve the antitumor activity of chemo and chemoimmunotherapy in LKB1 impaired tumors, exploiting their inability to overcome metabolic stress.

Keywords: Caloric restriction; Cancer metabolism; KRAS; LKB1; Metformin; NSCLC.

Fig. 1

Treatment schedule of in vivo experiments: A Mice were treated with chemotherapy (DDP) alone or in combination with metformin and caloric restriction (CR). B Mice were treated with chemotherapy (DDP) plus immunotherapy (anti PD-1) alone or in combination with metformin and caloric restriction (CR)

Fig. 2

Flow cytometry gating strategy for the identification of immune cell subsets in spleens and tumors: Total cells were first gated on physical parameters, a forward scatter (FSC)/side scatter (SSC) plot, and then gated cells are selected as singlets on the basis of SSC-A and SSC-H. Dead cells are excluded by the Zombie Aqua™ Fixable Viability Kit and leucocytes are identified as CD45⁺. Subset identification is shown for T cells (CD3⁺), CD8⁺ T cells, CD4⁺ T cells, myeloid cells (CD11b⁺), neutrophils (Ly6G⁺), and monocytes (Ly6C.⁺)

Fig. 3

In vitro effect of metformin and CR treatments on cell viability and metabolism: Cell viability measured by MTS assay after 72 h of treatment with 2 mM of metformin (MET) and 50% of caloric restriction (CR) in KRAS^G12D (A) and KRAS^G12D/LKB1^del (B) cell lines. One-way ANOVA was used for statistical analysis (**p < 0.01, *** p < 0.001). Error bars indicate SD. Seahorse XF Real-Time Assay was done after 24 h of treatment with 2 mM of metformin (MET) and 50% caloric restriction (CR) in KRAS^G12D (C) and KRAS^G12D/LKB1^del (D) cell lines. Two-way ANOVA was used for statistical analysis (* p < 0.05, **p < 0.01, *** p < 0.001). Error bars indicate SD

Fig. 4

In vitro effect of metformin and CR treatments on mTOR and MAPKs proliferative pathways: Representative immunoblot analysis of total and phosphorylated forms of extra-cellular signal-regulated kinase (ERK 1/2) and S6 ribosomal protein (S6) in KRAS^G12D (A) and KRAS^G12D/LKB1^del(B) cell lines. C-F Bar graphs of the quantification analysis of the active forms of ERK and S6 normalized on RAN, used as housekeeping. Proteins were extracted after 72 h of treatment with 2 mM of metformin (MET) and 50% caloric restriction (CR). Non-treated cells were used as control. One-way ANOVA was used for statistical analysis (**p < 0.01). Error bars indicate SD

Fig. 5

In vitro effect of metformin and CR treatments on mTOR proliferative pathway: Representative immunoblot analysis of total and phosphorylated forms of P70 S6 kinase (P70) in KRAS^G12D (A) and KRAS^G12D/LKB1^del(B) cell lines. C, D Bar graphs of the quantification analysis of the active forms of P70 normalized on RAN, used as housekeeping. Proteins were extracted after 72 h of treatment with 2 mM of metformin (MET) and 50% caloric restriction (CR). Non-treated cells were used as control. One-way ANOVA was used for statistical analysis and no statistically significant results were highlighted. Error bars indicate SD

Fig. 6

In vitro effect of metformin and CR treatments on PI3K/AKT proliferative pathway: Representative immunoblot analysis of total and different phosphorylated forms of the serine/threonine kinase AKT in KRAS^G12D (A) and KRAS^G12D/LKB1^del(B) cell lines. C-F Bar graphs of the quantification analysis of the different active forms of AKT normalized on RAN, used as housekeeping. Proteins were extracted after 72 h of treatment with 2 mM of metformin (MET) and 50% caloric restriction (CR). Non-treated cells were used as control. One-way ANOVA was used for statistical analysis and no statistically significant results were highlighted. Error bars indicate SD

Fig. 7

Effect of metformin and CR plus chemotherapy in subcutaneous K and KL tumors: Tumor growth inhibition in KRAS^G12D (A, B) and KRAS^G12D/LKB1^del (C, D) subcutaneously injected in immunocompetent mice (N = 9). For three weeks metformin (MET) was administered daily at 300 mg/kg, cisplatin (DDP) 5 mg/kg once a week and caloric restriction (CR) consisted of 36 h of fasting, once a week. Two-way ANOVA was used for statistical analysis (**p < 0.01, ****p < 0.0001). Error bars indicate SEM. The most statistically significant results have been reported in panels A and C. Single treatments, reported in panels B and D, did not induce a statistically significant reduction in tumor growth when compared to control or DDP groups

Fig. 8

Graph of body weight of mice bearing K and KL tumors: Body weight of KRAS^G12D (A) and KRAS^G12D/LKB1^del (B) mice used in Fig. 7 submitted to CR alone or in combination with DDP or/and metformin. The treatment schedule was well-tolerated, as the body weight lost during the fasting period was regain within 24 h

Fig. 9

Graph of body weight of mice bearing K and KL tumours: Body weight of KRAS^G12D (A) and KRAS^G12D/LKB1^del (B) mice used in Fig. 10. Untreated mice (CTRL) were compared to mice treated with DDP or DDP/MET/CR. The treatment schedule was well-tolerated, as the body weight lost during the fasting period was regain within 24 h. Intramuscular inoculation of KL cells, induced a massive weight loss in immunocompetent mice

Fig. 10

Effect of metformin and CR plus chemotherapy in intramuscular K and KL tumors: Tumor growth inhibition in KRAS^G12D (A, B) and KRAS^G12D/LKB1^del (C, D) intramuscularly injected in immunocompetent mice (N = 9). For three weeks metformin (MET) was administered daily at 300 mg/kg, cisplatin (DDP) 5 mg/kg once a week and caloric restriction (CR), consisted of 36 h of fasting, once a week. Two-way ANOVA was used for statistical analysis (**p < 0.01, *** p < 0.001, ****p < 0.0001). Error bars indicate SEM. The most statistically significant results have been reported in panels A and C. Single treatments, reported in panels B and D, did not induce a statistically significant reduction in tumor growth when compared to control or DDP groups

Fig. 11

Effect of metformin and CR plus chemotherapy in subcutaneous KL PDX tumor: A, B Tumor growth inhibition in KL PDX implanted subcutaneously in immunocompromised mice. For three weeks metformin (MET) was administered daily at 300 mg/kg, cisplatin (DDP) 5 mg/kg once a week and caloric restriction (CR), consisted of 24 h of fasting, once a week. Caloric restriction alone or plus metformin and chemotherapy reduced (C) body weight and (D) blood glucose in immunecompromised mice. Two-way ANOVA was used for statistical analysis (**p < 0.01, *** p < 0.001). Error bars indicate SEM. The most statistically significant results have been reported in panel A. Single treatments, reported in panels B, did not induce a statistically significant reduction in tumor growth when compared to control or DDP groups

Fig. 12

Effect of metformin and CR plus chemo-immunotherapy in intramuscular K and KL tumors: Tumor growth inhibition in KRAS^G12D (A, B) and KRAS^G12D/LKB1^del (C, D) intramuscularly injected in immunocompetent mice (N = 10). For five weeks metformin (MET) was administered daily at 300 mg/kg, cisplatin (DDP) 5 mg/kg once a week, anti PD-1 at 100 μg/dose twice a week, while the caloric restriction (CR) consisted of 36 h of fasting once a week. Two-way ANOVA was used for statistical analysis (**p < 0.01, ****p < 0.0001). Error bars indicate SEM. A, B Non statistically significant effects were observed in K tumours. For KL tumours, the most statistically significant results have been reported in panels C. D CR increases the effect of DDP/anti PD-1 combination (DDP/anti PD-1 vs DDP/anti PD-1/MET **p < 0.01) but (C, D) the best effect were reached with the DDP/anti PD-1/MET/CR combination

Fig. 13

Graph of body weight of mice bearing K and KL tumors: Body weight of KRAS^G12D (A) and KRAS^G12D/LKB1^del (B) mice used in Fig. 12. No-treated mice (CTRL) were compared to mice treated with DDP/anti PD-1 or DDP/anti PD-1/MET/CR. The treatment schedule was well-tolerated, as the body weight lost during the fasting period was regain within 24 h. Intramuscular inoculation of KL cells, induced a massive weight loss in immunocompetent mice

Fig. 14

Best treatment effect obtained in intramuscular K and KL tumours in immunocompetent mice: KRAS^G12D (A) and KRAS^G12D/LKB1^del (B) tumor volumes on days 35 and 31 of treatment respectively, the time-point when treatment benefit was greatest. Two-way ANOVA was used for statistical analysis (*p < 0.05, **p < 0.01, **** p < 0.0001). Error bars indicate SEM

Fig. 15

Quantification of GPX4 expression in K and KL tumors: Representative image of immunohistochemical staining for GPX4 of K (A) and KL (B) tumors. C, D The percentage of negative cells are represented as the mean of three different samples. Unpaired t test was used for statistical analysis (* p < 0.05). Error bars indicate SD

Fig. 16

TME and spleen immune characterization of K and KL untreated mice: A, B TME immune profile of KRAS^G12D and KRAS^G12D/LKB1^del tumours injected intramuscularly; C, D spleen immune profile of immunocompetent mice bearing KRAS^G12D and KRAS^G12D/LKB1^del tumours; samples from the experiment in Fig. 12 were collected and processed for FACS analysis of lymphoid and myeloid infiltrating immune cells (see Fig. 2 for the gate strategy). One-way and two-way ANOVA were used for statistical analysis (**p < 0.01, ****p < 0.0001). Error bars indicate SD

Fig. 17

TME and spleen immune characterization of KL treated mice: A, B TME and (C, D) spleen immune modulation in mice bearing KL tumors; samples from the experiment in Fig. 12 were collected and processed for FACS analysis of lymphoid and myeloid infiltrating immune cells (Fig. 2 for the gate strategy). One way and two-way ANOVA were used for statistical analysis (*p < 0.05, *** p < 0.001 ****p < 0.0001). Error bars indicate SD