Abstract
Hmgb3 is an X-linked member of a family of sequence-independent chromatin-binding proteins that is expressed in HSC-enriched lin−, c-kitHI, Sca-1HI, IL-7Rα− (KSIL) cells and Ter119+ erythroid cells. To define Hmgb3 function, we generated hemizygous mice (Hmgb3−/Y) using 129/SvJ ES cells. Hmgb3−/Y mice contain normal numbers of KSIL cells that are capable of normal repopulation and self-renewal. However, these mice have 1.6-fold fewer common lymphoid progenitors (CLP) and 3-fold fewer common myeloid progenitors (CMP) (p < 0.05). We hypothesized that the role of Hmgb3 in early hematopoiesis involves c-kit regulation. We observed that the level of c-kit mRNA in Hmgb3−/Y HSCs increased 30% compared to wild-type (WT) (p = 0.05). We used 5-fluorouracil (5-FU), which has been shown to down-regulate c-kit on HSCs, to characterize the interaction between Hmgb3 and c-kit. We monitored Hmgb3 expression in KSIL and lin−, Sca-1+, c-kit− cells before and after 5-FU treatment (150 mg/kg) using phenotypically normal transgenic mice containing an IRES-GFP cassette knocked into the 3′ UTR of Hmgb3. Prior to 5-FU treatment, 27% of KSIL cells were GFP+ (these cells were absent 4 days post-injection {p.i.}). In contrast, 1.8% of lin−, c-kit−, Sca-1+ cells were GFP+ before 5-FU treatment whereas 26% of lin−, c-kit−, Sca-1+ cells were GFP+ 4 days p.i. The increased proportion of GFP+ lin-, c-kit−, Sca-1+ cells after 5-FU treatment is consistent with previous findings that repopulating activity resides within the c-kit−/LO population in 5-FU treated bone marrow and our finding that Hmgb3 serves as a marker for long-term repopulating activity. To determine the time course of c-kit regulation, we compared bone marrow from 5-FU injected Hmgb3−/Y and WT mice for analysis at 2, 4, and 6 days p.i. Two days p.i., both WT and Hmgb3−/Y mice contained similar numbers of bone marrow cells (7 x 106 cells/hind limb) and the KSIL population was absent. By four days p.i., the bone marrow cellularity of WT mice declined to 5.5 ± 0.9 x 106 cells/hind limb and KSIL cells were still absent. However, in Hmgb3−/Y mice 4 days p.i., bone marrow cellularity stabilized at 7.9 ± 0.8 x 106 cells/hind limb, an increase of 43% compared to WT (p < 0.01), along with the re-emergence of the KSIL population. To determine whether the Hmgb3−/Y lin−, c-kit−, Sca-1+ population contains repopulating HSCs after 4 days of 5-FU treatment similar to WT mice, we performed repopulation assays using KSIL and lin−, c-kit−, Sca-1+ cells sorted from 4 day p.i. 5-FU treated Hmgb3−/Y mice. Recipients received either 2 x 104 KSIL or 2 x 105 lin−, c-kit−, Sca-1+ cells (Ly 5.2) from 5-FU treated Hmgb3−/Y mice along with a radioprotective dose of 3 x 105 congenic (Ly 5.1) bone marrow cells. FACS analysis performed on control recipients transplanted with congenic marrow exhibited < 1% Ly 5.2 cells in the bone marrow 16 weeks after transplant. Pre-5-FU treatment, 88% of bone marrow cells were donor derived in recipients of Hmgb3−/Y KSIL cells. There was no detectable engraftment of Hmgb3−Y lin−, c-kit−, Sca-1+ cells. In contrast to WT mice, both KSIL and lin−, c-kit−, Sca-1+ cells from 5-FU treated Hmgb3−/Y mice were capable of long-term repopulation (62–82% donor derived cells). We conclude that Hmgb3 deficiency facilitates the reemergence of c-kitHI HSCs following 5-FU treatment. Mechanisms involving either enhanced HSC self-renewal or delayed differentiation into CLPs and CMPs are both consistent with our results.
NANOG maintains self-renewal of primate ES cells in the absence of a feeder layer
