Adaptive plasticity of aspartate metabolism in succinate dehydrogenase-deficient cancer cells

Reviewer #1 (Public review):

Different studies have proposed distinct mechanisms by which succinate dehydrogenase (SDH)-deficient cells escape aspartate limitation, highlighting metabolic heterogeneity across experimental systems. In this study, the authors address these previously conflicting observations by longitudinally tracking the adaptation of multiple SDHB-knockout clones derived from the same parental cell line.

The authors identify two distinct adaptive mechanisms: complex I suppression with predominantly GOT1-dependent aspartate synthesis, and preservation of complex I activity with increased PC-GOT2-dependent aspartate synthesis. They further define shared and unique dependencies associated with these adaptive states, providing a rationale for potential therapeutic targeting strategies.

Overall, this is a strong study in cancer metabolism, integrating complementary longitudinal and mechanistic approaches, including long-term adaptation, isotope tracing, genetic perturbation, metabolomics, and functional cell growth assays. Although the study provides substantial mechanistic insight, several limitations remain.

(1) MPC is proposed as a shared dependency of both adaptive states. Testing whether MPC inhibition suppresses SDH-deficient tumor growth in vivo would substantially strengthen the therapeutic relevance.

(2) The distinction between complex I-intact and complex I-suppressed states is based mainly on the expression of two complex I subunits and the oxygen consumption. More direct assays of complex I activity or assembly are needed. Early-passage SDHB-knockout cells should also be included as controls in the OCR experiments.

(3) The two adaptive states appear to rely differentially on glucose- versus glutamine-derived aspartate synthesis. Testing the sensitivity of EP and LP clones to glucose or glutamine deprivation would further support this metabolic distinction.

(4) Since SDH is described as a tumor suppressor, the authors should clarify why SDHB loss initially inhibits hPheo1 cell proliferation.

(5) The study focuses on SDHB loss, and it remains unclear whether similar adaptive mechanisms arise following loss of other SDH subunits, including SDHA, SDHC, or SDHD, across different biological contexts. This limitation should be discussed explicitly.

Reviewer #2 (Public review):

In the manuscript entitled "Adaptive plasticity of aspartate metabolism in succinate dehydrogenase-deficient cancer cells," Sokolov et al. delineate the metabolic adaptations that succinate dehydrogenase (SDH)-deficient cancer cells undergo over time to overcome the initial aspartate limitation. To do so, the authors generated five clonal osteosarcoma SDH subunit B (SDHB) knockout cell lines using the CRISPR/Cas9 system and compared the proliferation rates of early- and late-passage cells, revealing that the latter rewired central carbon metabolism to increase aspartate levels and therefore replicate faster than their early-passage counterparts. Using a series of pharmacological and/or genetic interventions, the authors show that this rewiring can occur via two different routes: either through reduced Complex I (CI) activity, whereby glutamine is channelled towards aspartate synthesis via reductive carboxylation, or through a metabolic rewiring in which aspartate is produced from glucose via the PC-GOT2 pathway while CI activity is preserved. The CI-suppression-independent route depends on PC expression, as evidenced by an analysis of DepMap cell-line data, in which higher PC expression is associated with decreased SDH dependency. Moreover, they find that other consequences of aspartate deprivation observed in SDH-deficient cells, including impaired pyrimidine synthesis, replication stress, and DNA damage, are ameliorated in late-passage cells.

Overall, this study is interesting because it disentangles the different metabolic rewiring routes that SDH-deficient cells can undergo to reverse aspartate limitation and sheds light on previously reported, seemingly contradictory results in the field. However, the study's major premise requires further validation, and important controls are missing, diminishing the overall strength of the conclusions.

Major points:

(1) The main conclusion that two separate routes allow SDH-deficient cells to overcome aspartate limitation, defined by their CI-activity status, is not convincingly proven. Indeed, to show this dichotomous behaviour, the authors performed Western blots for two CI subunits and determined the basal oxygen consumption rate. However, these assays are insufficient to demonstrate that LP clones 2 and 3 maintain functional CI, in contrast to LP clone 1. Moreover, it is not ruled out that these clones show dysfunction in ETC complexes other than CI. To assess these points, the activities of all individual ETC complexes should be carefully measured, for instance, by Seahorse assay after permeabilization. Furthermore, given the complex nature of CI, a reduction in two subunits does not necessarily reflect a reduction in its assembly. Therefore, CI assembly should be assessed directly by BN-PAGE analysis of isolated mitochondria.

(2) It is difficult to reconcile why the authors used an NDUFA8 KO in clone 2 EP to mimic the physiological long-term CI-suppression-dependent adaptation. Indeed, this approach seems to represent an extreme scenario of Complex I loss that may induce non-physiological adaptations that override the effects of SDH KO. To assess the distinct metabolic fluxes between the two proposed routes, it would be advisable to use a more physiological model and instead compare the tracing data from LP clone 2 with those from LP clone 1, which exhibits a "natural" CI-suppressed state. Does clone 1 LP show similar metabolic changes to A8KO, including increased reductive carboxylation?

(3) It is unclear whether the loss of Complex I at late passage is an intrinsic progression of osteosarcoma cells rather than a feature specific to SDH-deficient cells. A proper comparison between SDHB-deficient cells and WT cells, both at early and late passage, should be carried out. This is essential to fully understand the adaptive trajectories of SDH-deficient cells. This comparison is essential to identify the baseline metabolic hardware of the osteosarcoma cells. Indeed, the authors state that "While wild-type 143B cells synthesize most aspartate from glutamine via oxidative TCA cycling and GOT2 activity, ..." (Page 7, third paragraph), but these data are not included in the manuscript and would represent an important control for assessing the observed metabolic changes in comparison with the wild-type context.

(4) The data showing that the PC-GOT2 pathway is mainly driven by enhanced PC activity are not fully convincing, as PC activity seems to be equally important for maintaining aspartate levels in the NDUFA8 KO compared with clone 2 LP. Moreover, the extracted expression data from DepMap suggest that increased PC expression might not be transcriptionally regulated, as only a slight association between PC mRNA levels and SDH dependency was observed. Are PC mRNA levels increased in clone 2 LP? If not, PC might be regulated post-transcriptionally. To test this, the nascent translation of PC could be assessed.

Author response:

We sincerely thank the reviewers for their time, helpful critiques and overall positive evaluation of our work. We plan to investigate the points raised by the reviewers and look forward to submitting a revised manuscript that addresses concerns about therapeutic relevance, functional distinctions between the CI-intact and CI-suppressed adapted states, and PC regulation in this system.