Abstract

We have previously shown that a distinctive feature of umbilical cord blood (UCB) CD34+ hematopoietic progenitor cells (HSPC) as compared to bone marrow (BM) CD34+ is a higher expression of transcription targets and components of the nuclear factor kappa B (NF-κB) pathway. NFKB2 and RELB are sub-units of the transcription factor (TF) that specifically mediates the constitutive NF-κB signaling pathway and their increased levels could be related with the primitive state of the newborn’s HSPC. However, BM and UCB CD34+ HSPC differ in their sub-population compositions, and a higher proportion of more primitive cells among the CD34+ cells could account for those differences. CD133 is a surface marker expressed on a more primitive sub-population of CD34+ cells that are highly enriched in long-term culture-initiating cells, NOD/SCID-repopulating cells. We used flow cytometry, oligonucleotide microarray gene expression profiling and real time quantitative PCR to better characterize immunomagnetically sorted CD34+ and CD133+ HSPC derived from BM and UCB. We found that UCB CD34+ cells contain a larger proportion of CD133+ cells (around 70%), differing from BM CD34+ cells (around 30%). Cluster analysis of the expression profiles, encompassing 10.000 genes, showed that UCB CD133+ are more similar to UCB CD34+ than to BM CD133+ cells. Furthermore, a statistically significant higher expression of NFKB2 and RELB was demonstrated by quantitative PCR on UCB CD133+ HSPC, compared to BM. Overall this indicates that despite distinct compositions of the cells from UCB or BM, UCB HSPC display intrinsic molecular differences related to their ontological age. The comparison of the gene expression profiles of the CD133+ with the CD34+ populations revealed the higher expression of many well known factors related to more primitive HSPC and hemangioblasts. In fact, TFs such as RUNX1/AML1, GATA3, USF1, TAL1/SCL, HOXA9 and HOXB4 were all present at higher levels in CD133+ HSPC. In an attempt to identify a key TF that could be responsible for the expression of these important factors, we carried a promoter analysis for the set of highly expressed TF found in the CD133 cells. A frequency of TF binding sites significantly higher than the expected was observed for the NF-κB TFs, including potential NF-κB targets such as RUNX1, GATA3 and USF1. Measurements of GATA3, NFKB2 and RELB expression by real-time PCR showed a higher expression of the three genes in CD133+ samples (both from BM and UCB), as well as a correlation of the expression levels of NFkB2 and RELB with one another and with GATA3 (Sperman’s correlation), indicating that GATA3 could be, in fact, regulated by NF-κB. To further test this hypothesis, we used interference RNA (RNAi) against NFKB2 in HSPC. Levels of NFKB2, GATA3 and RELB (a known target of NFKB2/RELB dimmers) were down-modulated, in comparison with cells transfected with control RNAi. Taken together, our data indicates that constitutive NF-κB signaling may act up-regulating transcription factors related to a more primitive state of HSPC.

This work was supported by FAPESP, CNPq and FINEP.

Author notes

Disclosure: No relevant conflicts of interest to declare.