Volume 18, Issue 4, Pages 522-532 (April 2016) Bile Acids Protect Expanding Hematopoietic Stem Cells from Unfolded Protein Stress in Fetal Liver  Valgardur Sigurdsson, Hajime Takei, Svetlana Soboleva, Visnja Radulovic, Roman Galeev, Kavitha Siva, L.M. Fredrik Leeb-Lundberg, Takashi Iida, Hiroshi Nittono, Kenichi Miharada  Cell Stem Cell  Volume 18, Issue 4, Pages 522-532 (April 2016) DOI: 10.1016/j.stem.2016.01.002 Copyright © 2016 Elsevier Inc. Terms and Conditions

Cell Stem Cell 2016 18, 522-532DOI: (10.1016/j.stem.2016.01.002) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 1 Higher Protein Synthesis Rate in FL-HSCs and Distinct BA Composition in FL (A) Measurement of protein synthesis rate using OP-Puro incorporation in various cell populations within BM and FL cells. Relative protein synthesis rates compared with total cells are shown. Significance is indicated by either pound signs (in comparison between BM and FL) or asterisks (in comparison with the total cells) (n = 3). (B) Intracellular staining of ER stress-related proteins in FL and BM cells. Relative mean fluorescent intensity (MFI) compared with Lin+ cells is shown (n = 3). (C–E) Gene set enrichment analysis (GSEA) revealed that FL-HSCs showed no enrichment in the expression of ER stress response genes (ER STRESS) (C) and the secretory pathway related genes (SECRETORY PATHWAY) (D). GSEA also revealed that FL-HSCs had rather lower ER chaperone genes (ER CHAPERONES) (E). Genes included in the gene sets are shown in Table S1. The assay was performed using a published database (McKinney-Freeman et al., 2012; GEO: GSE37000). NES, normalized enrichment score; nominal p, nominal p value; FDR q, false discovery rate q value. (F) Biosynthesis of primary BA and key enzymes involved in the pathways. (G) Composition of BA in FL and AL based on mass spectrometry analyses (n = 3–5) (see also Figures S1B and S1C). (H) Composition of taurine-conjugated BA within total BA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Cell Stem Cell 2016 18, 522-532DOI: (10.1016/j.stem.2016.01.002) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 2 Injection of FXR Agonist Profoundly Reduces the Number of HSPCs in FL (A) Experimental design of GW4064 injection. Pregnant mice were intravenously injected with 2.5 mg/kg of GW4064 from E10.5, and on E15.5, FL were analyzed. In the rescue experiment, either 7.5 mg/kg of TCA or 1.5 mg/kg of salubrinal was intraperitoneally injected in the same time. (B) Mass spectrometry assays for measuring BA levels of GW4064-treated FL with or without TCA. Significance is indicated by either pound signs (for total BA level) or asterisks (for TCA level). Error bars are not shown (see also Figure S2A). (C and D) Frequency of non-hematopoietic cells (CD45−Lineage−), erythrocytes (CD45−Lineage+), Lin+ cells (CD45+Lineage+), and Lin− cells (CD45+Lineage−) in FL derived from GW4064-treated mice. Representative fluorescence-activated cell sorting (FACS) profiles (C) and graphs (D) are shown. Each dot represents a single fetus (20–24 fetuses from 3 or 4 mothers). (E) Intracellular staining to measure protein levels of ER stress response genes in Lin−, KSL, and HSC populations. Values show fold change of MFI in GW4064 samples compared with vehicle-treated samples (n = 3 or 4). (F) The number of c-kit+ cells and HSCs in WT and CHOP KO FL treated with GW4064. WT and CHOP KO pregnant mice were treated with GW4064 and c-kit+ cells isolated from FL were analyzed. (G) Weight of fetuses derived from GW4064 with or without TCA/salubrinal-treated mothers (9–45 fetuses from 2–6 mothers). (H) Cellularity of FL cells. Absolute cell number per weight is shown (9–45 fetuses from 2–6 mothers). (I) Absolute cell number of HSPCs. Absolute cell numbers of Lin−, KSL, and HSC population are shown (7–30 fetuses from 3 or 4 mothers). (J) Frequency of annexin-V+ (apoptotic) cells in Lin−, KSL, and HSC populations (7–30 fetuses from 2–4 mothers). (K) Limiting dilution competitive transplantation assay of GW4064 with or without TCA/salubrinal co-injected FL (8–12 recipient mice per group). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Cell Stem Cell 2016 18, 522-532DOI: (10.1016/j.stem.2016.01.002) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 3 Dual Deletion of Cyp27a1 Both in Mother and in Fetus Severely Affects Developing FL (A) Schematic interpretation of mating set up to create Cyp27a1−/− fetuses in hetero- and homozygous mothers. (B) Mass spectrometry assays for measuring BA levels of Cyp27a1 KO FL. Significance is indicated by either pound signs (for total BA level) or asterisks (for TCA level). Error bars are not shown (see also Figure S3F). (C) Fetuses derived from heterozygous (Cyp27a1+/−) mother and homozygous (Cyp27a1−/−) mother (E15.5). Yellow arrows indicate FL. (D) Weight of fetuses (5–19 fetuses from 3 mothers). (E) Cellularity of FL. Absolute cell number per weight is shown (5–19 fetuses from 3 mothers). (F) Absolute cell number of Lin−, KSL, and HSC population in the FL (E15.5) (5–19 fetuses from 3 mothers). (G) MFI of intracellular CHOP staining within Lin−, KSL, and HSC population in the FL (E15.5) (5–19 fetuses from 3 mothers). (H) Frequency of annexin-V+ (apoptotic) cells in Lin−, KSL, and HSC population (5–19 fetuses from 3 mothers). (I) Limiting dilution competitive transplantation assay of Cyp27a1−/− FL derived from heterozygous or homozygous mother, with or without salubrinal (Salb) injection (6 or 7 recipient mice per group). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Cell Stem Cell 2016 18, 522-532DOI: (10.1016/j.stem.2016.01.002) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 4 TCA Maintains Functional HSCs In Vitro by Inhibiting Generation of Aggresomes (A) Time course uptake of 3H-TCA into c-kit+ FL cells. Disintegrations per minute (DPM) on each time point are shown (n = 3). (B) Competitive uptake of 3H-TCA versus non-labeled BA in c-kit+ FL cells. DPM of each combination after 10 min are shown (n = 3). (C) Quantification of aggregated and un- or mis-folded proteins (aggresome) using ProteoStat staining. Representative FACS plot defining ProteoStat-high population is shown. (D) Increased ProteoStat-high population in the FL-HSPCs cultured for 24 hr in vitro. Representative histogram and graph are shown. (E) Increased ProteoStat-high population in the fresh FL-HSPCs derived from Cyp27a1 KO fetuses. Representative histogram and graph are shown. (F) Experimental design of the transplantation assay. One hundred CD150+CD48−KSL cells isolated from E15.5 FL derived from Ly-5.1/5.2 fetuses were cultured for 7 or 14 days with or without 100 μM of each BA. Whole culture (donor) was transplanted into lethally irradiated Ly-5.2 mice (recipient) along with 2 × 105 total BM cells derived from Ly-5.1 mice (competitor). (G) Transplantation assay of cultured FL-HSCs. Long-term reconstitution of hematopoietic system was monitored by peripheral blood analyses. Data from 16 weeks analysis iare shown (n = 6–8). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Cell Stem Cell 2016 18, 522-532DOI: (10.1016/j.stem.2016.01.002) Copyright © 2016 Elsevier Inc. Terms and Conditions