Identification and Characterization of a Hematopoietic Stem Cell Niche in Spleen

In normal hematopoiesis, hematopoietic stem cells (HSCs) in bone marrow mainly produce various types of blood cells. HSCs have a function of self-renewal and are located in the osteoblastic and vascular niche, which regulates the stemness function as a stem cell microenvironment. Recent studies clearly show that several factors (N-Cadherin, b-Catenin, SDF-1a and Spp1) are involved in the maintenance of the stemness function in HSCs. However, it has also been shown that embryonic HSCs can proliferate and differentiate in extrameduller tissues like liver, spleen and placenta. In adults, extrameduller hematopoiesis (EMH) can be induced by a hematological malignancy and some infectious diseases. However, the mechanism of HSCs regulation and niche cells at the EMH has not been well defined.

To reveal the mechanism of HSCs in the EMH, we utilized the c-fos knockout mouse (c-fos ^−/−) as an EMH model mouse. In c-fos ^−/−, hematopoiesis in bone marrow was absent as a result of marrow spaces being occupied by increasing the number of bone forming osteoblasts; EMH then began in the spleen. First, to identify the main site of the HSC-niche in the EMH spleen, we performed HSC localization analysis using in situ hybridization of various HSC markers. Surprisingly, some CD34, Sca-1, c-kit, SCL/tal-1, and Tie-2 expressing cells were located near megakaryocyte like cells (MLCs). These MLCs were mainly located in the red pulp region in the spleen, and 3–5 cells form a syncytium. Interestingly, these MLCs express various types of osteoblastic niche related molecules, (N-Cadherin, b-Catenin, Spp1 and SDF-1a) in addition to megakaryocytic markers (CD41, CD61, and b3-integrin), suggesting that MLC is a HSC niche candidate in the EMH.

To confirm our hypothesis, we next performed CFU-S (colony forming unit-spleen) assay as a naturally induced EMH model. Total 1×10⁵ bone marrow mononuclear cells isolated from Ly5.2 or GFP-transgenic mouse were transplanted into lethally irradiated Ly5.2 mouse. In this system, MLCs were first seen in the spleen at day 1. At day 8, the number of MLCs had increased and units of 10–15 MLCs aggregated and formed syncytium. About 80% of MLCs were derived from donor (Ly5.2) cells. More interestingly, and contrary to our expectations, these aggregated MLCs with Sca-1⁺ or c-kit⁺ hematopoietic progenitor cells (HPCs) were mainly located, not inside the colony, but in the interstitial region between the developing colonies. They also expressed osteoblastic niche molecules. It has been suggested that each colonies in spleen are derived from HPCs and that HPCs exist inside the colony. Our data indicate the possibility that HPCs were transiently- located outside the colony.

To undertake a detailed characterization of MLCs, we sorted donor derived MLCs as a Lin-/CD41+ cells and performed Q-PCR analysis. In agreement with our histological analysis, the sorted MLCs expressed various types of niche molecules and cytokines, compared to megakaryocytes. As well, when we co-cultured HSCs with isolated MLCs, the numbers of HSC were significantly increased compare to with a liquid culture system. Taking these data together, we suggested that MLCs have the potential to support HSC proliferation.

In this study, we first identified a candidate for EMH-niche cells and postulated a developmental mechanism in the spleen. Our findings provide a new functional insight into HSCs outside the bone marrow, and extend a new tool that supports ex vivo expansion of HSCs.