Spatiotemporal upscaling of sparse air-sea pCO2 data via physics-informed transfer learning.

Global measurements of ocean pCO2 are critical to monitor and understand changes in the global carbon cycle. However, pCO2 observations remain sparse as they are mostly collected on opportunistic ship tracks. Several approaches, especially based on direct learning, have been used to upscale and extrapolate sparse point data to dense estimates using globally available input features. However, these estimates tend to exhibit spatially heterogeneous performance. As a result, we propose a physics-informed transfer learning workflow to generate dense pCO2 estimates that are grounded in real-world measurements and remain physically consistent. The models are initially trained on dense input predictors against pCO2 estimates from Earth system model simulation, and then fine-tuned to sparse SOCAT observational data. Compared to the benchmark direct learning approach, our transfer learning framework shows major improvements of up to 56-92%. Furthermore, we demonstrate that using models that explicitly account for spatiotemporal structures in the data yield better validation performances by 50-68%. Our strategy thus presents a new monthly global pCO2 estimate that spans for 35 years between 1982-2017.

MetaFlux: Meta-learning global carbon fluxes from sparse spatiotemporal observations.

We provide a global, long-term carbon flux dataset of gross primary production and ecosystem respiration generated using meta-learning, called MetaFlux. The idea behind meta-learning stems from the need to learn efficiently given sparse data by learning how to learn broad features across tasks to better infer other poorly sampled ones. Using meta-trained ensemble of deep models, we generate global carbon products on daily and monthly timescales at a 0.25-degree spatial resolution from 2001 to 2021, through a combination of reanalysis and remote-sensing products. Site-level validation finds that MetaFlux ensembles have lower validation error by 5-7% compared to their non-meta-trained counterparts. In addition, they are more robust to extreme observations, with 4-24% lower errors. We also checked for seasonality, interannual variability, and correlation to solar-induced fluorescence of the upscaled product and found that MetaFlux outperformed other machine-learning based carbon product, especially in the tropics and semi-arids by 10-40%. Overall, MetaFlux can be used to study a wide range of biogeochemical processes.

M2 isoform of pyruvate kinase rewires glucose metabolism during radiation therapy to promote an antioxidant response and glioblastoma radioresistance.

BACKGROUND: Resistance to existing therapies is a significant challenge in improving outcomes for glioblastoma (GBM) patients. Metabolic plasticity has emerged as an important contributor to therapy resistance, including radiation therapy (RT). Here, we investigated how GBM cells reprogram their glucose metabolism in response to RT to promote radiation resistance. METHODS: Effects of radiation on glucose metabolism of human GBM specimens were examined in vitro and in vivo with the use of metabolic and enzymatic assays, targeted metabolomics, and FDG-PET. Radiosensitization potential of interfering with M2 isoform of pyruvate kinase (PKM2) activity was tested via gliomasphere formation assays and in vivo human GBM models. RESULTS: Here, we show that RT induces increased glucose utilization by GBM cells, and this is accompanied with translocation of GLUT3 transporters to the cell membrane. Irradiated GBM cells route glucose carbons through the pentose phosphate pathway (PPP) to harness the antioxidant power of the PPP and support survival after radiation. This response is regulated in part by the PKM2. Activators of PKM2 can antagonize the radiation-induced rewiring of glucose metabolism and radiosensitize GBM cells in vitro and in vivo. CONCLUSIONS: These findings open the possibility that interventions designed to target cancer-specific regulators of metabolic plasticity, such as PKM2, rather than specific metabolic pathways, have the potential to improve the radiotherapeutic outcomes in GBM patients.

Intragenic suppressor mutations of the COQ8 protein kinase homolog restore coenzyme Q biosynthesis and function in Saccharomyces cerevisiae.

Saccharomyces cerevisiae Coq8 is a member of the ancient UbiB atypical protein kinase family. Coq8, and its orthologs UbiB, ABC1, ADCK3, and ADCK4, are required for the biosynthesis of coenzyme Q in yeast, E. coli, A. thaliana, and humans. Each Coq8 ortholog retains nine highly conserved protein kinase-like motifs, yet its functional role in coenzyme Q biosynthesis remains mysterious. Coq8 may function as an ATPase whose activity is stimulated by coenzyme Q intermediates and phospholipids. A key yeast point mutant expressing Coq8-A197V was previously shown to result in a coenzyme Q-less, respiratory deficient phenotype. The A197V substitution occurs in the crucial Ala-rich protein kinase-like motif I of yeast Coq8. Here we show that long-term cultures of mutants expressing Coq8-A197V produce spontaneous revertants with the ability to grow on medium containing a non-fermentable carbon source. Each revertant is shown to harbor a secondary intragenic suppressor mutation within the COQ8 gene. The intragenic suppressors restore the synthesis of coenzyme Q. One class of the suppressors fully restores the levels of coenzyme Q and key Coq polypeptides necessary for the maintenance and integrity of the high-molecular mass CoQ synthome (also termed complex Q), while the other class provides only a partial rescue. Mutants harboring the first class of suppressors grow robustly under respiratory conditions, while mutants containing the second class grow more slowly under these conditions. Our work provides insight into the function of this important yet still enigmatic Coq8 family.