Abstract
Hematopoietic stem cells (HSCs) are effectively expanded in fetal liver (FL), while they are maintained in a dormant state in adult bone marrow (BM). However, developmental mechanisms allowing this have not been fully explained. BM-HSCs have the lowest protein synthesis rate within the blood hierarchy, even under forced self-renewal divisions. In addition, HSCs are vulnerable to and quickly activate endoplasmic reticulum (ER) stress responses fueled by accumulation of unfolded / misfolded proteins (Miharada et al., Cell Rep. 2014). Of note, we have seen that FL-HSCs have low levels of ER stress related genes despite their high proliferation status without an increase in heat shock protein levels, strongly indicating that other factor(s) block ER stress elevation. This raises the question how HSCs deal with the higher protein-folding requirement during expansion in the FL. Here we demonstrate that bile acids (BAs) are required to eliminate ER stress in the FL and are essential for proper expansion of FL-HSCs.
Measurement of protein synthesis rate using OP-puro incorporation revealed that protein synthesis was enhanced in FL-HSCs, whereas BM-HSCs have half the rate of other populations in BM. Mass spectrometry analyses showed that BAs in the FL were all taurine conjugated while 30% of BA in the adult liver was taurine-conjugated, and the main proportion was taurocholic acid (TCA) that is known for its low toxicity. In the FL we also detected secondary BAs (e.g. TDCA), requiring intestinal bacteria in the production process, suggesting that FL BAs are a mixture of fetal and maternal BAs.
Reduction of BA levels using GW4064, a chemical inhibitor of BA synthesis, significantly decreased the number of HSCs (6.6 fold decrease compared to vehicle treatment). This decrease was due to increased apoptosis caused by elevated ER stress levels. Similarly, dual deletion of Cyp27a1, a key BA synthetic enzyme, in both mother and fetus severely decreased total cellularity (2.0 fold decrease compared to littermate heterozygotes) and number of HSCs (6.8 fold decrease) in FL due to increased ER stress and subsequent apoptosis. Interestingly, FL of homozygotes grown in heterozygous mothers did not show any significant differences compared to littermate heterozygotes, suggesting that the contribution of maternal BA in FL is critical for HSCs. In both models, ER stress-oriented apoptosis and reduction in cellularity were most pronounced within the HSC population, indicating that stem cells are particularly sensitive to BA levels during development in FL. Importantly, injection of TCA or Salubrinal, an ER stress inhibitor, rescued the effects of BA reduction in both models. These data strongly suggest that BAs are required to block ER stress elevation in expanding FL-HSCs.
ER stress and protein aggregation are closely linked together in number of pathological diseases like AlzheimerÕs- and HuntingtonÕs disease. Quantification of aggregated proteins (aggresomes) revealed that Cyp27a1 KO FL-HSCs from homozygote mothers contained significantly higher amount of aggresomes (2.0 fold), while KO FL-HSCs from heterozygote mothers showed no increase. Higher levels of aggregated proteins were most pronounced within the HSC population and BA suppressed formation of aggresomes during in vitro culture. This leads to reduction of ER stress and the maintenance of functional HSCs. Finally, transplantation assay showed that TCA can support functional HSCs ex vivo for up to 14 days.
These findings propose a novel role for BA as a critical part of fetal hematopoiesis supporting expansion of HSC. Maternal and fetal BA coordinately contribute to this natural chaperone regulation.
No relevant conflicts of interest to declare.
