Across phyla, reduced nicotinamide adenine dinucleotide phosphate (NADPH) transfers intracellular reducing power to thioredoxin reductase-1 (TrxR1) and glutathione reductase (GR), thereby supporting fundamental housekeeping and antioxidant pathways. Here we show that a third, NADPH-independent, pathway can bypass the need for TrxR1 and GR in mammalian liver. Most mice genetically engineered to lack both TrxR1 and GR in all hepatocytes (“TR/GR-null livers”) remain long-term viable. TR/GR-null livers cannot reduce oxidized glutathione disulfide but still require continuous glutathione synthesis. Inhibition of cystathionine gamma-lyase causes rapid necrosis of TR/GR-null livers, indicating that methionine-fueled trans-sulfuration supplies the necessary cysteine precursor for glutathione synthesis via an NADPH-independent pathway. We further show that dietary methionine provides the cytosolic disulfide reducing power and all sulfur amino acids in TR/GR-null livers. Although NADPH is generally considered an essential reducing currency, these results indicate that hepatocytes can adequately sustain cytosolic redox homeostasis pathways using either NADPH or methionine.

Cre-responsive dual-fluorescent alleles allow in situ marking of cell lineages or genetically modified cells. Here we report a dual-fluorescent allele, ROSAnT-nG, that directs nuclear accumulation of tdTomato in Cre-naïve lineages. Cre converts the allele to ROSAnG, which drives nuclear EGFP accumulation. Conditions were established for analyzing marked nuclei by flow cytometry on the basis of red-green fluorescence and ploidy, with a particular focus on liver nuclei. Hydrodynamic delivery of a Cre-expression plasmid was used to timestamp arbitrary hepatocytes for lineage-tracing. The distinct green fluorescence of nuclei from Cre-exposed lineages facilitated analyses of ploidy transitions within clones. To assess developmental transitions in liver nuclei, ROSAnT-nG was combined with the hepatocyte-specific AlbCre transgene, facilitating discrimination between hepatocyte- and non-hepatocyte-nuclei. Nuclei extracted from postnatal day 2 (P2) livers were 41% green and 59% red, and reached a stable level of 84% green by P22. Until P20, green nuclei were >98% diploid (2N); at P40 they were ~56% 2N, 43% 4N, and <1% 8N; and by P70 they reached a stable distribution of ~46% 2N, 45% 4N, and 9% 8N. In conclusion, ROSAnT-nG will facilitate in vivo and ex vivo studies on liver and will likely be valuable for studies on tissues like muscle, kidney, or brain, in which cells are refractory to whole-cell flow cytometery, or like trophectoderm derivatives or cancers, in which cells undergo ploidy transitions.

Besides helping to maintain a reducing intracellular environment, the thioredoxin (Trx) system impacts bioenergetics and drug-metabolism. We show that hepatocyte-specific disruption of Txnrd1, encoding Trx reductase-1 (TrxR1), causes a metabolic switch in which lipogenic genes are repressed and periportal hepatocytes become engorged with glycogen. These livers also overexpress machinery for biosynthesis of glutathione and conversion of glycogen into UDP-glucuronate; they stockpile glutathione-S-transferases and UDP-glucuronyl-transferases; and they overexpress xenobiotic exporters. This realigned metabolic profile suggested that the mutant hepatocytes might be preconditioned to more effectively detoxify certain xenobiotic challenges. Hepatocytes convert the pro-toxin acetaminophen (APAP, paracetamol) into cytotoxic N-acetyl-p-benzoquinone imine (NAPQI). APAP defenses include glucuronidation of APAP or glutathionylation of NAPQI, allowing removal by xenobiotic exporters. We found that NAPQI directly inactivates TrxR1, yet Txnrd1-null livers were resistant to APAP-induced hepatotoxicity. Txnrd1-null livers did not have more effective gene expression responses to APAP challenge; however their constitutive metabolic state supported more robust GSH biosynthesis-, glutathionylation-, and glucuronidation-systems. Following APAP challenge, this effectively sustained the GSH system and attenuated damage.

Thioredoxin reductase-1 (TrxR1)-, glutathione reductase (Gsr)-, and Nrf2 transcription factor-driven antioxidant systems form an integrated network that combats potentially carcinogenic oxidative damage yet also protects cancer cells from oxidative death. Here we show that although unchallenged wild-type (WT), TrxR1-null, or Gsr-null mouse livers exhibited similarly low DNA damage indices, these were 100-fold higher in unchallenged TrxR1/Gsr–double-null livers. Notwithstanding, spontaneous cancer rates remained surprisingly low in TrxR1/Gsr-null livers. All genotypes, including TrxR1/Gsr-null, were susceptible to N-diethylnitrosamine (DEN)-induced liver cancer, indicating that loss of these antioxidant systems did not prevent cancer cell survival. Interestingly, however, following DEN treatment, TrxR1-null livers developed threefold fewer tumors compared with WT livers. Disruption of TrxR1 in a marked subset of DEN-initiated cancer cells had no effect on their subsequent contributions to tumors, suggesting that TrxR1-disruption does not affect cancer progression under normal care, but does decrease the frequency of DEN-induced cancer initiation. Consistent with this idea, TrxR1-null livers showed altered basal and DEN-exposed metabolomic profiles compared with WT livers. To examine how oxidative stress influenced cancer progression, we compared DEN-induced cancer malignancy under chronically low oxidative stress (TrxR1-null, standard care) vs. elevated oxidative stress (TrxR1/Gsr-null livers, standard care or phenobarbital-exposed TrxR1-null livers). In both cases, elevated oxidative stress was correlated with significantly increased malignancy. Finally, although TrxR1-null and TrxR1/Gsr-null livers showed strong Nrf2 activity in noncancerous hepatocytes, there was no correlation between malignancy and Nrf2 expression within tumors across genotypes. We conclude that TrxR1, Gsr, Nrf2, and oxidative stress are major determinants of liver cancer but in a complex, context-dependent manner.

Cre-responsive fluorescent marker alleles are powerful tools for cell lineage tracing in mice; however their utility is limited by regulation of Cre activity. When targeting hepatocytes, hydrodynamic delivery of a Cre-expression plasmid can convert Cre-responsive alleles without inducing the intracellular or systemic antiviral responses often associated with viral-derived Cre-expression vectors. In this method, rapid high-volume intravenous inoculation induces hepatocyte-targeted uptake of extracellular molecules. Here we tested whether hydrodynamic delivery of Cre protein or Cre fused to the HIV-TAT cell-penetrating peptide could convert Cre-responsive reporters in hepatocytes of mice. Hydrodynamic delivery of 2 nmol of either Cre or TAT-Cre protein converted the reporter allele in 5 to 20% of hepatocytes. Neither protein gave detectable Cre activity in endothelia, non-liver organs, or non-hepatocyte cells in liver. Using mice homozygous for a Cre-responsive marker that directs red- (Cre-naïve) or green- (Cre-converted) fluorescent proteins to the nucleus, we assessed sub-saturation Cre-activity. One month after hydrodynamic inoculation with Cre protein, 58% of hepatocyte nuclei that were green were also red, indicating that less than half of the hepatocytes that had obtained enough Cre to convert one marker allele to green were able to convert all alleles. For comparison, one month after hydrodynamic delivery of a Cre-expression plasmid with a weak promoter, only 26% of the green nuclei were also red. Our results show that hydrodynamic delivery of Cre protein allows rapid allelic conversion in hepatocytes, but Cre-activity is sub-saturating so many cells will not convert multiple Cre-responsive alleles.