4  DISCUSSION
The initial understanding of IP was limited to eyes, testes and brain, which are organs with blood tissue barriers. Later, research expanded the concept of IP, which actually exists in many tissues, including hair follicles, intestinal mucosa, lung and so on. IP actively guides and controls the immune response through various mechanisms to maintain the integrity of immune tolerance microenvironment. Fujisaki3 first revealed bone marrow IP sites by using high-resolution in vivo imaging in animal experiments. The main component cells of IP are Tregs, which are characterized by the expression of FoxP3 in the nucleus. Through the binding of membrane CXCR4 receptors with a large amount of chemokine CXCL12 in the HSCs niche, Tregs aggregate to bone marrow near the endosteum surface of trabecular bone to surround HSCs12, 13, where the transplanted allogeneic HSCs could survive for a long time (30 days) without myeloablative conditioning regimen before transplantation, which is the gold standard experiment for IP identification. Once the Tregs were removed from bone marrow by deleting the CXCR4 receptor in mice, TNF-α and IFN-γ were enhanced, while IL-10 was decreased in the bone marrow3. Then, the transplanted allogeneic HSCs described in the above experiment could not survive in the osteoblast niche. HSCs located in IP immune tolerance microenvironment are LT-HSCs3. Subsequent studies showed that LT-HSCs were over mobilized with stress induction after Tregs were depleted12, 14. Therefore, in bone marrow, IP not only provides an immune tolerance microenvironment to protect LT-HSCs from various immune factors but also reduces the overmobilization of LT-HSCs during stress and maintains the static state of LT-HSCs, which are the stem cells maintaining lifelong hematopoiesis of bone marrow.
It has been proven in AA patients that the ability of Tregs to migrate to bone marrow was impaired, and the proportion of Tregs in the bone marrow detected by flow cytometry was also decreased15,16. However, there has been no research linking the immune abnormalities of AA with local immune tolerance status of IP. Interestingly, a series of immune disorders of AA reported in the literature and in our previous studies are strikingly similar to the excessive immune status due to the abnormal IP. The increased expression of TNF-α and IFN-γ in CD4+ and CD8+ T cells was similar to the changes of bone marrow cytokines in the IP damage assay (Treg deletion)3, 17,18. The Th1/Th2 polarization detected in a previous study of an AA animal model was highly consistent with the immune disorder caused by impaired IP function19,20.
Drawing on other experimental methods studying the function of IP in local tissues 4,5,21, we tested the distribution of Tregs in the bone marrow osteoblast niche by immunohistochemistry. There were significantly fewer Tregs on the endosteal surface of bone marrow than in normal controls and MDS patients, accompanied by an increase in Th1/Th2, TNF-α, IFN-γ and IL-17 in bone marrow, while in another bone marrow failure disease, MDS, they were not. All these results suggested that IP structure and function were impaired in bone marrow of children with AA.
The LT-HSCs protected by IP are usually in a static state to maintain the hematopoietic stem cell pool3. Generally, LT-HSCs are appropriately transition into activated ST-HSCs which in turn differentiate into various blood cells22. This transition increases during stress. Recent research has shown that Tregs regulate the level of reactive oxygen species in LT-HSCs to prevent LT-HSCs from being over mobilized and exhausted during oxidative stress11,13. Our experiment showed significantly reduced LT-HSCs in AA, which can be attributed to the loss of the protective effect of IP on LT-HSCs. Various clinical evidences also support that the ability to maintain HSCs against stress in AA is reduced. Patients usually develop AA after infection, exposure to radioactive substances or certain drugs, and the severity of the disease is often aggravated during infection1, 23. In clinical practice, we found that 60%-70% of children with NSAA naturally evolved into SAA during the course of continuous exposure to various environmental stresses24, 25. After the depletion of LT-HSCs, there are not enough reserve HSCs to be mobilized, resulting in the corresponding reduction of ST-HSCs. Therefore, we also detected a decrease of ST-HSCs in the bone marrow of children with AA.