Abstract
Single-cell analysis reveals sequential genetic events from hematopoietic stem cell differentiation hierarchy in multiple myeloma patients with FGFR3/IGH translocation
Jiaojiao Guo1*, Yv Wang2*, Dehui Zou2*, Zhenhao Liu1, Dan Wu2, Lu Xie3, Fenghuang Zhan4, Guoji Guo5,Bing Liu6, Lugui Qiu2, Jiaxi Zhou2, Tao Cheng2#, Wen Zhou1#
1Cancer Research Institute,Central South University, Changsha 410078, China.2State Key Laboratory of Experimental Hematology, Chinese Academy of Medical Science,Tianjin, China.3Shanghai Center for Bioinformation Technology, Shanghai , China.4Department of Internal Medicine, Division of Hematology, University of Iowa, Iowa City, IA 52242, USA. 5Center for Stem Cell and Regenerative Medicine, Zhejiang University School of Medicine, Hangzhou, Zhejiang , China.6307-Ivy Translational Medicine Center, Laboratory of Oncology, Academy of Military Medical Sciences, Beijing , China.
E-mail: wenzhou@csu.edu.cn; chengtao@ihcams.ac.cn
Background: It is generally believed that multiple myeloma (MM) cells develop from either B cells or plasma cells in the bone marrow (BM). However, the nature of the earliest precursors of MM cells is still poorly defined and existing data are conflicting regarding the origin of MM cells. How these genes cooperate with other MM-related factors to induce MM oncogenic transformation in the hematopoietic hierarchy needs to be further defined.
Materials and Methods: In this study, we employed single-cell gene expression analysis of HSC-PC lineage including HSC, MLP, B cells, Plasmablasts (PBs) and Plasma cells (PC) from one health donor (HD), one paired monozygotic twin(one is HD and another one suffered with MM), MM patients at diagnosis (AD) and relapse (RE) with FGFR3/IGH translocation. 672 individual cells from the 8 population were sorted for single cell real-time PCR reaction.
Results: We focused on the HSC-PC lineage because of its implication in MM and identified several differentially expressed genes (DEGs) in the HSC compartment with continuous activation throughout HSC-PCs. Among those, BACH2FA and DDIT3 showed consistent up-regulation, while FOXO3 and IL1 were down-regulated continuously. 7 DEGs (SIPA1L2, RBM34, SPI1, MDM2, AKAP12, DKK1 and c-Myc) showed overlap among the three conditions (HD, AD and RE, respectively). Kaplan-Meier analysis revealed HSC-associated signatures (HAS) in overlapping and non-overlapping DEGs with poor survival outcome, p=0.000). To compare genetic alterations between sporadic and hereditary MM patients, we analyzed two groups of DEGs - one from sporadic MM patients (AD) and HD; the other from the hereditary MM patient (AD) and the healthy counterpart of its paired monozygotic twin- along the HSC differentiation hierarchy. We identified both overlapping and non-overlapping DEGs. For example, MDM2 and c-Myc were over-expressed at the HSC and the MLP stage, respectively, in both groups. However, ARUKA was not activated in the monozygotic twin group, presumably because of the lack of FGFR3 translocation. By using fluorescence in situ hybridization (FISH) analysis, we detected IGH and FGFR3 translocation in 50-60% and 11% MM patients, respectively. These chromosomal rearrangements have been used to identify tumor plasma cells. FGFR3-translocated was found in 18%, 23% and 44% for B cells, PBs and plasma cells, respectively. Our work extends previous discoveries of chromosome translocation in malignant plasma cells, suggesting existence of heterogeneity in cells with FGFR3 translocation. Furthermore, FGFR3 translocation was accompanied by alteration of ARUKA. While B cell differentiation associated genes begin to express in the plasmablast stage, expression of multi-drug resistance genes occurs only in plasma cells in relapsed patients. Together, these molecular alterations likely constitute key elements of a sophisticated genetic program underlying the occurrence of MM.
Conclusion: Our findings reveal the existence of genetically distinct cellular origins of MM cells from the HSC differentiation hierarchy within each MM tumor mass. FGFR3 translocation was accompanied by alteration of ARUKA, indicating FGFR3 translocation might cause CIN gene activation. Finally, we find that a small percentage of HSCs from MM patients could harbor oncogenic activation, thereby pointing to the need of a more vigorous quality control in autologous HSC transplantation for MM patients.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
