Dual role of ER stress in response to metabolic co-targeting and radiosensitivity in head and neck cancer cells

Arginine deprivation therapy (ADT) is a new metabolic targeting approach with high therapeutic potential for various solid cancers. Combination of ADT with low doses of the natural arginine analog canavanine effectively sensitizes malignant cells to irradiation. However, the molecular mechanisms determining the sensitivity of intrinsically non-auxotrophic cancers to arginine deficiency are still poorly understood. We here show for the first time that arginine deficiency is accompanied by global metabolic changes and protein/membrane breakdown, and results in the induction of specific, more or less pronounced (severe vs. mild) ER stress responses in head and neck squamous cell carcinoma (HNSCC) cells that differ in their intrinsic ADT sensitivity. Combination of ADT with canavanine triggered catastrophic ER stress via the eIF2α-ATF4(GADD34)-CHOP pathway, thereby inducing apoptosis; the same signaling arm was irrelevant in ADT-related radiosensitization. The particular strong supra-additive effect of ADT, canavanine and irradiation in both intrinsically more and less sensitive cancer cells supports the rational of ER stress pathways as novel target for improving multi-modal metabolic anti-cancer therapy. Electronic supplementary material The online version of this article (10.1007/s00018-020-03704-7) contains supplementary material, which is available to authorized users.

(A) Representative Western blot and semi-quantitative densitometric analysis of relative intrinstic ASS protein level in the HNSCC cell line panel when grown as exponential monolayer culture under standard conditions.α-tubulin (α-tub) was used as loading control.Data represent means +SD of N=3 independent experiments.
(B) Representative RT-PCR analysis of ASS1 and ASL genes in HNSCC cell lines grown according to (A).ACTB mRNA level was recorded as reference.
(C) Correlation analysis of the expression of ASS1 gene and protein levels in the HNSCC cell lines determined by RT-PCR and Western blot analysis.
(D) Growth behavior of HNSCC cells under mono-ADT in the presence of Cit.Cells were exposed to Arg-free medium (-Arg) containing physiological Cit concentrations (0.04 mM) for up to 9 days.Selected samples were assessed for growth recovery by exchanging the supernatant to standard complete medium (+Arg) after 1d, 3d, 5d, 7d or 9d of exposure to the Arg-deprived condition; cell counts were measured 3 days later.Graphs show mean values +SD from N=3 independent experiments.

Figure S2
High level of ASS1 expression is a marker of poor prognosis for HNSCC patients and radioresponse.
(A) Correlation analysis of the expression of ASS1 gene and 0.1% surviving fraction (SF) doses of HNSCC lines determined in vitro.
(B) Correlation analysis of 0.1% SF doses determined in vitro and TCD 50 (dose required to control the disease in 50% of tumor-bearing animals) of the corresponding xenograft models of HNSCC lines published previously [20].
(C) Oncoprint of human HNSCC with upregulated expression of ASS1 gene based on the data in The Cancer Genome Atlas (TCGA; 517 non-randomized tumors; 1.5 z-score).
(D) Kaplan-Meier analysis to assess the correlation of ASS1 expression with overall and relapsefree survival of HNSCC patients from the TCGA (517 non-randomized tumors) database.Data were stratified by ASS1 expression level: ASS1 median and ASS1 high (≥3 f.c upregulation; correspondent 1.5 z-score) and comparative analysis by log-rank test was performed between these two different groups.(B) Representative data sets of ER stress response genes in SAS and FaDu spheroids analyzed by RT-PCR.Spheroids were incubated in Arg-free medium for up to 120 h or left untreated (0 h). ACTB mRNA level was determined as reference.
(C) Representative Western blots of proteins from SAS spheroids upon mono-ADT.Cell lysates were probed with specific antibodies against the indicated proteins; α-tubulin (α-tub) was used as loading control.
Figure S1 -see legend on next page concentrations is cytotoxic for HNSCC cells only in an Arg-free environment.(A) Surviving fractions (±SD) of SAS and FaDu cells irradiated under standard culture conditions (+Arg) in the absence or presence of 0.04 mM Cit and 0.01-0.1 mM Cav (N=3, n≥3).Data were fitted with a linear-quadratic model as depicted in Materials and Methods.(B) Colony forming capacity of SAS and FaDu cells treated by comb-ADT with 0.04 mM Cit and 0.05-0.1 mM Cav upon 0-6 Gy single dose irradiation.(C) Representative RT-PCR analyses of spliced XBP1 expression in monolayer SAS and FaDu cells after defined times of Cav (0.1 mM) exposure.The spliced XBP1 mRNA (lower band) is apparent already after 4 h of comb-ADT; ACTB is shown as reference control and FaDu cells treated with tunicamycin (Tn; 3 µg/ml) served as positive control (right).
Figure S4Single dose irradiation alone leads to spheroid growth delay and control; mono-ADT less effectively triggers ER stress in the 3-D environment than in 2-D culture (cf.Figs.6 and 2).(A)Spheroid volume growth is delayed after irradiation of 3-D cultures with single doses that do not yet abrogate growth recovery (up to 12.5 Gy for SAS and 10 Gy for FaDu spheroids); mean values (± SD) are shown for n≥28 spheroids.
Figure S5see legend on next page