Glucose Metabolism and Aging of Hematopoietic Stem and Progenitor Cells

Abstract

Comprehensive proteomics studies of human hematopoietic stem and progenitor cells (HSPC) have revealed that aging of the HSPC compartment is characterized by elevated glycolysis. This is in addition to deregulations found in murine transcriptomics studies, such as an increased differentiation bias towards the myeloid lineage, alterations in DNA repair, and a decrease in lymphoid development. The increase in glycolytic enzyme activity is caused by the expansion of a more glycolytic HSPC subset. We therefore developed a method to isolate HSPC into three distinct categories according to their glucose uptake (GU) levels, namely the GU^high, GU^inter and GU^low subsets. Single-cell transcriptomics studies showed that the GU^high subset is highly enriched for HSPC with a differentiation bias towards myeloid lineages. Gene set enrichment analysis (GSEA) demonstrated that the gene sets for cell cycle arrest, senescence-associated secretory phenotype, and the anti-apoptosis and P53 pathways are significantly upregulated in the GU^high population. With this series of studies, we have produced a comprehensive proteomics and single-cell transcriptomics atlas of molecular changes in human HSPC upon aging. Although many of the molecular deregulations are similar to those found in mice, there are significant differences. The most unique finding is the association of elevated central carbon metabolism with senescence. Due to the lack of specific markers, the isolation and collection of senescent cells have yet to be developed, especially for human HSPC. The GU^high subset from the human HSPC compartment possesses all the transcriptome characteristics of senescence. This property may be exploited to accurately enrich, visualize, and trace senescence development in human bone marrow.

Keywords: aging; central carbon metabolism; glycolysis; hematopoietic stem and progenitor cells; senescence signature.

Figure 1

Prominent proteome changes upon aging in six individual cell types in human bone marrow (modified from Hennrich et al. [23]). Selected pathways that were significantly altered, i.e., either up- or downregulated with age are depicted. The size of each circle represents the percentage coverage of a pathway with proteins assessed by LF quantification in our dataset. The color indicates the direction of change, with red being upregulated and blue being downregulated with age.

Figure 2

Changes in central carbon metabolism of HSPC upon aging (modified from Hennrich et al. [23]). (a) Glucose metabolism pathway and the tricarboxylic acid (TCA) cycle are depicted. The respective enzymes are shown in capital letters and the color encodes for changes upon aging. Glycogen is broken down and the products are fed into glycolysis. Glucose-6 phosphate is metabolized to dihydroxyacetone-phosphate (DHAP) and glyceraldehyde-3-phosphate (GA3P). The enzymes GPD2 and DAK that feed the pool of DHAP and GA3P increase upon aging as well as TALDO1. (b) A specific set of enzymes involved in the mitochondrial beta-oxidation of fatty acids increase in abundance with age. (c) Volcano plots of all proteins quantified in the respective cell populations. The elevation of enzymes involved in the central carbon metabolism is found only in the HSPC.

Figure 3

Glycogen accumulation and glucose uptake capacity of the CD34+ HSPC (modified from Poisa et al. [49]). (a) Results of the semi-quantitative assessment of intensity of PAS reaction as surrogate marker for glycogen in CD34+ cells from young versus old human subjects. The average glycogen content was significantly higher in old subjects than young subjects < 35 years (one-sided t-test, p = 1.2 × 10⁻⁶, *** p < 0.001). (b,c) Glucose uptake capacity of the total CD34+ cells was assessed by Cayman’s Glucose Uptake Assay Kit. Incubation for 30 min with 1.75 μg/mL 2-NBDG yielded a dose-dependent uptake of glucose into the CD34+ cells. The latter could then be separated according to their respective levels of glucose uptake by a FAC-Sorter into three distinct subpopulations according to the glucose uptake (GU) capabilities: GU^low, GU^inter, and GU^high. (b) An example of separation of CD34+ cells from a human subject > 50 years according to GU capacity. (c) An example of separation of CD34+ cells from a young subject (<35 years). In young subjects only, GU^inter and GU^low, but scarcely any GU^high cells could be detected.