MAEA is an E3 ubiquitin ligase promoting autophagy and maintenance of haematopoietic stem cells

Abstract

Haematopoietic stem cells (HSCs) tightly regulate their quiescence, proliferation, and differentiation to generate blood cells during the entire lifetime. The mechanisms by which these critical activities are balanced are still unclear. Here, we report that Macrophage-Erythroblast Attacher (MAEA, also known as EMP), a receptor thus far only identified in erythroblastic island, is a membrane-associated E3 ubiquitin ligase subunit essential for HSC maintenance and lymphoid potential. Maea is highly expressed in HSCs and its deletion in mice severely impairs HSC quiescence and leads to a lethal myeloproliferative syndrome. Mechanistically, we have found that the surface expression of several haematopoietic cytokine receptors (e.g. MPL, FLT3) is stabilised in the absence of Maea, thereby prolonging their intracellular signalling. This is associated with impaired autophagy flux in HSCs but not in mature haematopoietic cells. Administration of receptor kinase inhibitor or autophagy-inducing compounds rescues the functional defects of Maea-deficient HSCs. Our results suggest that MAEA provides E3 ubiquitin ligase activity, guarding HSC function by restricting cytokine receptor signalling via autophagy.

Fig. 1. MAEA expression is enriched and required in haematopoietic stem cells (HSCs) for haematopoiesis.

a Representative histograms showing MAEA expression on lineage- Sca-1 + c-kit+ progenitor cells (LSKs) and HSCs in control (Ctrl) and Maea^Csf1r-Cre bone marrow (BM). b FACS quantification of MAEA expression on control (n = 5) and Maea^Csf1r-Cre (n = 4) BM HSPCs (LMPP lymphoid-primed multipotent progenitors, CMP common myeloid progenitors, CLP common lymphoid progenitors, GMP granulocyte-macrophage progenitors, MEP megakaryocyte-erythrocyte progenitors), neutrophils (Neu), B cells (B) and monocytes (MN). Data are represented as boxes-and-whiskers with the whiskers span from minima to maxima of each dataset, the boxes extend from the 25th to 75th percentiles and the centre line indicates the mean. HSC p = 0.0202, LSK p = 0.0004, LMPP p = 0.0076, CMP p = 0.0016, CLP p = 0.0173. c The Kaplan–Meyer survival curve of control (n = 8) and Maea^Csf1r-Cre (n = 6) mice. p value was calculated by log-rank test. d White blood cell (WBC) counts in peripheral blood (n = 6 each group) and frequency of B, T and myeloid cells in total WBCs of young adult control and Maea^Csf1r-Cre mice (n = 8 each group). WBC p = 0.0001, B, T and M p < 0.0001. e Quantifications of HSC and LSK numbers in BM of control (n = 6) and Maea^Csf1r-Cre mice (n = 7). HSC p = 0.0165, LSK p = 0.024. f Quantification of myeloid progenitors in BM of control (n = 6) and Maea^Csf1r-Cre (n = 9) mice (MkP: n = 4). CMP p = 0.0146, GMP p = 0.0186, MkP p = 0.0499. g Quantification of lymphoid progenitors in BM of control (n = 6) and Maea^Csf1r-Cre (n = 9) mice^. LMPP p = 0.0027, CLP p = 0.0003. h Experimental scheme and results for evaluating lymphoid differentiation potential of HSC and LMPP at single cell level (HSC: n = 3, LMPP: n = 4 each group). p = 0.025. i Peripheral blood donor chimaerism at indicated time points after competitive BM transplantation (BMT) of equal number (1 × 10⁶) of CD45.1 wild-type (WT) competitor BM cells and CD45.2 donor BM cells from indicated genotypes into lethally irradiated CD45.1 WT recipients (fl/fl: n = 4, fl/+; Cre: n = 5; fl/fl; Cre: n = 6). 1.25 week p = 0.45, 2.25 week p = 0.027, 4 week p = 0.006, 8, 12 16 weeks p < 0.0001. j HSC donor chimaerism in the recipient BM from i at 16wks after transplant (n = 6 each group). p < 0.0001. All data are shown as mean ± sem and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by unpaired two-sided Student’s t test unless otherwise indicated.

Fig. 2. Maea deletion impairs haematopoietic stem cells (HSCs) quiescence and function in a mTOR-dependent manner.

a Experimental scheme for deleting Maea in adult mice using Mx1-Cre. (The mouse symbol in these figures were modified and recreated from Servier Medical Art (https://smart.servier.com/), an open source of medical images). b Quantification of bone marrow (BM) HSC numbers at indicated time points after poly I:C injection (day 0: n = 4; days 7 and 21: n = 5; day 14: n = 3 each group). Data are represented as floating boxes with boundaries indicate minima to maxima of each dataset and middle line indicates the mean. Day 0 p = 0.069, day 7 p = 0.0457, day 14 p = 0.168, day 21 p = 0.005. c Representative FACS plots and cell cycle profiles of control (Ctrl) and Maea^Mx1-Cre (CKO) HSCs at 21 days after 1st poly I:C injection (n = 4). p = 0.0002. d Experimental scheme for deleting Maea in 1:1 wild-type (WT) and Maea^Mx1-Cre BM chimeras after stable (8 weeks) reconstitution. e Donor chimaerism in BM LSK and HSCs and peripheral blood total leucocytes from control and Maea^Mx1-Cre mixed chimeric mice at indicated time points after poly I:C injection (n = 10 over two independent experiments). f RNA-seq and Gene Set Enrichment Analysis (GSEA) of HSCs from BM of control and Maea^Csf1r-Cre young adults at 7–12 weeks of age. Three replicates with 2000 HSCs pooled from two mice each replicate were processed and analysed for each group. Top KEGG (Kyoto Encyclopaedia of Genes and Genomes) pathways that are up-regulated and down-regulated in Maea^Csf1r-Cre HSCs are shown. g Examples of GSEA enrichment plots showing enrichment of Proteasome, Oxidative phosphorylation and mTOR signalling pathways in Maea^Csf1r-Cre HSCs. h GSEA enrichment plot showing significant downregulation of lymphoid potential related gene set in Maea^Csf1r-Cre HSCs. i Heat map showing mean expression of HSC related genes in control and Maea^Csf1r-Cre HSCs. j Experimental scheme and quantification of LSKs and HSCs in control and Maea^Mx1-Cre mice treated with vehicle (Veh), carfilzomib (CFZ), rapamycin (Rapa) or N-acetylcysteine (NAC) for 3 weeks after poly I:C induction (Veh: n = 5, CFZ: n = 6, Rapa: n = 7, NAC: n = 3 over two independent experiments). k Donor chimaerism in peripheral blood of CD45.1 lethally irradiated wild-type recipients at indicated time points after competitive BM transplantation (BMT) of equal number of CD45.1 WT competitor BM cells and CD45.2 donor BM cells from indicated groups (n = 5 each group). All data are mean ± sem unless otherwise indicated. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by unpaired two-sided t-test with 95% confidence level. ns not significant.

Fig. 3. MAEA regulates cytokine receptor ubiquitination and stability in hematopoietic stem cells (HSCs).

a Representative immunofluorescence images showing MAEA and CD150 expression in HSCs using cells from two independent experiments of four mice. Scale bar = 10 μm. b Representative histograms and FACS evaluation of Flt3 half-life in control and Maea^Csf1r-Cre LSK cells incubated in the presence of 50 μM cycloheximide (n = 4 animals). *p < 0.05, **p < 0.01, ***p < 0.001 by two-way ANOVA multiple comparisons. exact p value from left to right are 0.0325, 0.0016, 0.0003. c Experimental scheme and quantification of HSCs in control (n = 9) and Maea^Mx1-Cre (n = 10) mice treated with vehicle and RTK inhibitor (RTKi) PKC412 for 3 weeks after poly I:C induction. p = 0.0212. d Quantification of the frequencies of G0 phase HSCs in control and Maea^Mx1-Cre (n = 7 each) mice treated with vehicle and RTKi PKC412 at 3 weeks after poly I:C induction. p = 0.0033. e Quantification of BM macrophages in control (n = 9) and Maea^Mx1-Cre (n = 10) mice treated with vehicle and RTKi PKC412 at 3 weeks after poly I:C induction. p value from left to right are 0.0007, 0.0012. All data are shown as mean ± sem. ns not significant. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by unpaired two-sided t-test with 95% confidence level unless otherwise indicated.

Fig. 4. MAEA regulates cytokine receptor stability via autophagy in hematopoietic stem cells (HSCs).

a Experimental scheme and evaluation of autophagy flux in control (Ctrl) and Maea^Csf1r-Cre HSCs (n = 7). % autophagy flux is calculated as 100x(1 − (−L/N))/(+L/N). N: NH₄Cl. L: leupeptin. p = 0.0036. b Representative electron microscopy micrographs of control and Maea^Csf1r-Cre HSCs for ultrastructural analysis of the autophagic compartments (red arrows point to an autolysosome in control and an autophagosome in Maea^Csf1r-Cre). c Morphometric analysis of control and Maea^Csf1r-Cre HSCs: quantification of the numbers of autophagic vacuoles (AV) and their break down into number (left) and percentage (right) of autophagosomes (APG) and autolysosomes (AUT), per cell area. n = 27 control and 23 Maea^Csf1r-Cre cells analysed. Analysis were applied to cells blindly from two independent experiments. Unblinding was done during data plotting. p value from left to right are p = 0.035, 0.0245, 0.0306. d Frequencies of HSCs in control and Maea^Csf1r-Cre bone marrow (BM) cells before and after 3 h of starvation in culture (n = 6 each). p value from left to right are 0.0093, 0.0016. e Representative immunofluorescence images and quantification showing subcellular colocalization of FLT3 and LC3 in freshly isolated (Ctrl) and starved (cultured ex vivo in StemSpan with no cytokines but in the presence of lysosome inhibitors N/L for 3 h to induce autophagy) HSCs (n = 11 cells analysed). p = 0.0071. Scale bar = 10 μm. f Flt3 half-life in control and Maea^Csf1r-Cre lineage- Sca-1+ c-kit+ progenitor cells (LSKs) cells incubated in the presence of 50 μM cycloheximide and 10 mM lithium chloride (LiCl) (n = 4 animals). p value from left to right are 0.034, 0.0105, 0.08. g Quantification of LSKs and HSCs in control and Maea^Mx1-Cre BM treated with vehicle (Veh), LiCl or verapamil (Verap) 3 weeks after poly I:C induction (Veh: n = 3, LiCl: n = 6, Verap: n = 6). h Peripheral blood donor chimaerism in CD45.1 lethally irradiated wild-type (WT) recipients at indicated time points after competitive BM transplantation (BMT) of equal number of CD45.1 WT competitor BM cells and CD45.2 donor BM cells from indicated groups (n = 5 each group). Data are shown as mean ± sem. ns not significant. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by unpaired two-sided t-test with 95% confidence level unless otherwise indicated.