Proteomics Analysis of Cellular Network in Human Bone Marrow Reveals Lineage Skewing Towards Megakaryocytes and Decrease in Lymphoid Development upon Aging

Background: Whereas proteomics analyses at a cellular level have been performed in murine HPC, most proteomics studies on human material have thus far focused on examining human tissues with a mixture of cell types and not at a specific cellular level. Whole proteomics analyses might reveal genuine molecular alterations und mechanisms of aging on human hematopoietic cells.

Aim: To identify the age-related molecular changes in human hematopoietic progenitor cells (HPC) as well as in the cellular elements of the human bone marrow niche, we have applied mass spectrometry-based platform for proteomics analysis of this network of bone marrow (BM) cells at a cellular level. The overarching goal is to acquire a systems understanding of the molecular mechanisms involved in aging of human HPC as well as other cellular constituents of the marrow niche, and to combine the cell biological, proteomic, and genetic studies for a better understanding of age-related diseases such as MDS.

Methods: In this study, we have simultaneously characterized at the proteomics level HPC as well as the cellular microenvironment in BM from each individual human subject. Cells were harvested from the BM of 69 healthy human subjects [Age Group (AG) 20-29 years: n=19; AG 30-39: n=16; AG 40-49: n=9; AG 50+: N=15]. BM cells were separated for nucleated cells using Ficoll, followed by sorting with specific marker constellations, i.e. HPC (CD34+), lymphoid cells (CD34-/CD45+/SSC^low), myeloid precursors (CD34-/CD45^med/SSC^high), monocyte-macrophages (CD34-/CD45+/CD14+), erythroid precursors (CD34-/CD45-), and mesenchymal stromal cells (MSC) based on plastic adherence upon culture.

Results: In 60 of these 69 healthy subjects we were able to harvest an adequate amount of material from 5 to 6 the aforementioned subsets from the respective individual, yielding altogether 342 samples for proteomics analysis. After digestion (trypsinization), isobaric labelling of peptides (TMT) of appropriate quality was successfully performed in 270 of these 342 samples derived from 69 healthy human subjects. These 270 samples underwent tandem mass-spectrometry analyses. The number of proteins identified were >12,000, covering 65% of the estimated human proteome (HPC: >7,500; Lymphoid cells: >8,500; Myeloid Precursors: >7,500; Mono/Macro: >8,500; Erythroid Precursors: >6,500; and in MSC: >9,000 proteins). We have then focused on the molecular mechanisms, specifically proteomic alterations involved in aging of these 6 cell types. Gene Set Enrichment Analysis (GSEA) using previously defined gene sets specific for common lymphoid progenitors (CLP), bi-potent granulocyte/macrophage progenitors (GMP), and megakaryocyte progenitors (MKP) revealed significant down-regulation of CLP related proteins and significant up-regulation of MKP related proteins in HPC upon aging. The same down-regulation of CLP proteins and up-regulation of MKP proteins were also found in the BM lymphoid cells. Further significant alterations of note are a decrease in cell cycle related proteins upon aging in HPC, lymphoid cells, as well as in MSC, and an increase in GMP related proteins upon aging in lymphoid cells, erythroid precursors and in MSC. Extensive pathway and network analyses are underway.

Conclusions: We have for the first time accomplished a multiplex and comprehensive analysis of a community of cells in human BM, comprising HPC and 5 other cell types that constitute the niche and have identified >12,000 proteins in this cellular network. Proteomics analyses have demonstrated a significant down-regulation of CLP related proteins in human HPC and in marrow lymphoid cells, as well as an up-regulation of MKP in all the cell types tested upon aging. Of note is also a decrease in cell cycle related proteins with age. This study has thus provided evidence that lineage skewing towards megakaryocytic and decrease in lymphoid development upon aging occur in human HPC as well as in other human BM cells.