Deciphering the molecular mechanism of ineffective erythropoiesis in MDS-5q

（请保持文字，不要PDF） 大约80%的MDS患者存在贫血，这是发病的主要原因。贫血通常是巨红细胞性的，总是反映无效的红细胞生成。然而，为什么红系细胞在骨髓中成熟时死亡还不确定，促使这项调查。尤其不清楚的是为什么贫血发生在MDS的早期（即在介绍时），通常50-75%的骨髓被肿瘤细胞取代，许多正常（非肿瘤）细胞仍然存在。我们以前已经表明，CFU-E/proerythroblasts死亡时，他们的细胞内血红素（酶合成的有毒螯合物）超过珠蛋白（蛋白质）。这可能是由于血红素输出失败（Flvcr 1缺失小鼠）或核糖体蛋白单倍不足、翻译受损和珠蛋白合成减慢（MDS-5 q和Diamond Blackfan贫血（DBA）患者）。由于血红素合成正常启动，但球蛋白翻译减慢，血红素超过FLVCR的输出能力，并诱导高水平的ROS和细胞死亡。为了描述导致成熟MDS-5 q和DBA细胞红系细胞死亡的共同机制，我们用CITE-seq（细胞表面蛋白的抗体条形码）和RNA测序研究了单个骨髓细胞。我们将总转录组与SCVelo伪时间分析进行比对，其中包括不完全加工的mRNA，然后将细胞的表面蛋白表达与其独特的转录组联系起来。在初步研究中使用这种方法，我们表明MDS-5 q患者骨髓中的所有红系细胞具有涉及血红素毒性的转录变化。几乎所有的骨髓红细胞生成都发生在成红细胞岛（EBI）内，EBI是一种由中央巨噬细胞（“护士细胞”）和约10-50个成熟红细胞组成的结构。我们怀疑，中央巨噬细胞的作用是安全和有效地回收血红素从CFU-E/proerythroblasts到邻近的铁亲后期红系细胞。有趣的是，5 q+（非肿瘤性）细胞也受到损害，不能扩增和重建红细胞生成。我们假设，大量血红素从肿瘤性（5 q-）红细胞前体流入中央巨噬细胞超过了可以代谢成铁，然后储存为铁蛋白或通过ferroportin输出。这损害了巨噬细胞支持共粘附正常（5 q+）细胞成熟的能力，因此正常和肿瘤性红细胞前体的分化都失败。为了验证这一假设，我们将首先展示在这项由R56资助的研究中使用EBI培养系统来模拟人类EBI的可行性和相关性。这将提供所需的数据，以证明血红素铁运输及其对MDS-5 q患者红细胞分化的影响的研究，也许其他低和低中等风险MDS患者致残性贫血。

(PLEASE KEEP IN WORD, DO NOT PDF) Approximately 80% of MDS patients present with anemia, which is the major cause of morbidity. The anemia is generally macrocytic and always reflects ineffective erythropoiesis. However, why erythroid cells die while maturing in the bone marrow is uncertain, prompting this investigation. Especially unclear is why anemia occurs early in MDS (i.e. at presentation) when often 50-75% of the marrow is replaced by neoplastic cells and many normal (non-neoplastic) cells remain. We previously have shown that CFU-E/proerythroblasts die when their intracellular heme (a toxic chelate synthesized enzymatically) exceeds globin (a protein). This can result from failed heme export (Flvcr1-deleted mice) or from ribosomal protein haploinsufficiency, impaired translation, and slowed globin synthesis (MDS-5q and Diamond Blackfan anemia (DBA) patients). Since heme synthesis initiates normally, but globin translation is slowed, heme exceeds the export capacity of FLVCR and induces high levels of ROS and cell death. To delineate the shared mechanisms that lead to the death of maturing MDS-5q and DBA cells erythroid cells, we studied single marrow cells with CITE-seq (antibody barcoding of cell surface proteins) and RNA sequencing. We aligned total transcriptomes with SCVelo pseudotime analysis, which includes incompletely processed mRNAs, and then linked a cell’s surface protein expression to its unique transcriptome. Using this approach in preliminary studies, we showed that all erythroid cells in MDS-5q patient marrow have transcriptional changes implicating heme toxicity. Nearly all marrow erythropoiesis takes place within erythroblastic islands (EBIs), a structure comprised of a central macrophage (“nurse cell”) and ~10-50 maturing red cells. We suspect that the role of the central macrophage is to safely and efficiently recycle heme from CFU-E/proerythroblasts to neighboring iron-avid later erythroid cells. Interestingly, the 5q+ (non-neoplastic) cells are also compromised and fail to expand and reconstitute erythropoiesis. We hypothesize, that the large quantities of heme influx from neoplastic (5q-) erythroid precursors into the central macrophage exceeds that which can be metabolized to iron then stored as ferritin or exported via ferroportin. This compromises the macrophage’s ability to support the maturation of co-adherent normal (5q+) cells, and thus the differentiation of both normal and neoplastic erythroid precursors fail. In order to test this hypothesis we will first show the feasibility and the relevance of using an EBI culture system to model human EBI in this R56-funded study. This would provide the data needed to justify studies of heme-iron trafficking and its impact on red cell differentiation in MDS-5q patients, and perhaps other low and low-intermediate risk MDS patients with disabling anemia.