bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Dichotomous regulation of lysosomes by MYC and TFEB controls hematopoietic stem cell fate

Laura García-Prat1,2, Kerstin B. Kaufmann1,2, Florin Schneiter3, Veronique Voisin2,4, Alex Murison1,2, Jocelyn Chen1,5, Michelle Chan-Seng-Yue1,2, Olga I. Gan1,2, Jessica L. McLeod1,2, Sabrina A. Smith1,2, Michelle C. Shoong1,2, Darrien Paris1,2, Kristele Pan1,2, Andy G.X. Zeng1,2, Gabriela Krivdova1,2, Kinam Gupta1,2, Shin-Ichiro Takayanagi1,2, Elvin Wagenblast1,2, Weijia Wang3, Mathieu Lupien1,5,6, Timm Schroeder3, Stephanie Z. Xie1,2*, and John E. Dick1,2,6*

1Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada; 2Department of Molecular Genetics, University of Toronto, Toronto, Ontario, Canada; 3Department of Biosystems Science and Engineering, ETH Zurich, 4058 Basel, Switzerland; 4The Donnelly Centre, University of Toronto, Toronto, Ontario, Canada; 5Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; 6Ontario Institute for Cancer Research, Toronto, Ontario, Canada. *Correspondence to: [email protected] and [email protected]

1 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Summary

It is critical to understand how quiescent long-term hematopoietic stem cells (LT-HSC) sense demand from daily and stress-mediated cues and transition into bioenergetically active progeny to differentiate and meet these cellular needs. Here, we show that lysosomes, which are sophisticated nutrient sensing and signaling centers, are dichotomously regulated by the Transcription Factor EB (TFEB) and MYC to balance catabolic and anabolic processes required for activating LT-HSC and guiding their lineage fate. TFEB-mediated induction of the endolysosomal pathway causes membrane receptor degradation, limiting LT-HSC metabolic and mitogenic activation, which promotes quiescence, self-renewal and governs erythroid-myeloid commitment. By contrast, MYC engages biosynthetic processes while repressing lysosomal catabolism to drive LT-HSC activation. Collectively, our study identifies lysosomes as a central regulatory hub for proper and coordinated stem cell fate determination.

Keywords lysosomes, TFEB, MYC, long-term HSC, self-renewal, erythropoiesis, myelopoiesis, TfR1, anabolism, catabolism

2 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Introduction

Human long-term hematopoietic stem cells (LT-HSC), at the apex of the hematopoietic hierarchy, must meet enormous daily demand (~1011 cells daily) while also sustaining life-long maintenance of the stem cell pool. This hierarchical organization is widely thought to protect LT-HSC from exhaustion by their maintenance in a quiescent and undifferentiated state, activating only in response to microenvironment signals to generate highly proliferative but more short-lived populations including short-term HSC (ST-HSC) and committed progenitors(1). Upon cues to exit this dormant state, HSC must respond and adapt their metabolism and nutrient uptake to meet increased bioenergetic demands for cell growth and differentiation(2). Simultaneously, the events underlying cellular and metabolic activation must also be suppressed within a subset of LT-HSC to enable re-entry to quiescence and ultimately maintaining the LT- HSC pool through self-renewal(2, 3). However, the demand-adapted regulatory circuits of these early steps of hematopoiesis are largely unknown. Sensing signals or nutrient uptake depends on that are embedded within the plasma membrane. These proteins internalize through endocytosis and can be degraded in the lysosomes or rerouted back to the cell surface and reused(4). Lysosomes are also terminal catabolic stations for autophagy where they clear cytoplasmic components, a process essential for preserving adult stem cell function(5-9). Recent work shows that lysosomes are not merely degradation stations, but also serve as major signaling centers for molecular complex assembly including mTORC1, AMPK, GSK3 and the inflammasome(10). These signaling complexes sense, integrate and facilitate cross-talk between diverse signals, and ultimately enable responses including autophagy, cell growth, membrane repair and microbe clearance. Although these distinct lysosomal activities in the stem cell context are largely unknown, lysosomes in mouse HSC were recently reported to be asymmetrically inherited, predicting future daughter cell fates(11), while pharmacological inhibition of the lysosomal v-ATPase positively impacted mouse HSC engraftment(12). Thus, we hypothesize that lysosomes coordinate the cell cycle and metabolic machinery of LT-HSC through their ability to sense and respond to diverse signaling cues to adapt their fate and lineage choices.

The MiT/TFE family of transcription factors (TFs) control lysosomes, with most studies

3 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

focusing on TFEB as a master regulator of lysosomes. TFEB can sense and respond to stress signals and metabolic cues, including nutrient starvation or mitochondrial damage, by transcriptional activation of endocytosis, autophagy and lysosomal biogenesis (13). TFEB and MYC appear to compete for binding to the same chromatin regions in HeLa cells due to a high degree of binding (14), but the biological relevance of this potential competition in a stem cell context is not understood. MYC regulates many aspects of metabolism, and plays a role in murine HSC by balancing self-renewal and differentiation(15, 16). However, the role of MYC in human HSC is unknown, and no previous studies of the role of TFEB in either human or mouse HSC function have been undertaken. Here, we uncovered a MYC-TFEB-mediated dichotomous regulation of lysosomal activity that is required to balance anabolic and catabolic processes that ultimately impact human LT-HSC fate determination.

4 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Results

MYC drives LT-HSC activation by promoting anabolism and inhibiting lysosomal activity

During mitogenic activation from quiescence, LT-HSC enter a transitional state where transcriptional and metabolic changes occur to promote G0 exit with resultant entry to cell cycle. RNA-sequencing (RNA-seq) analysis of uncultured (quiescent, qLT-HSC) vs cultured (activated, aLT-HSC) LT-HSC sorted from human cord blood (hCB) showed that genes associated with biosynthetic processes, such as ribosome and mitochondria biogenesis, were the most enriched sets in aLT-HSC (Fig. 1A;Fig. S1A). This included MYC and known MYC targets that are involved in promoting cell cycle, synthesis and mitochondrial metabolism (Fig. S1B;Fig. 1A,B;Table S1). Indeed, concordant with mouse HSC(17), MYC protein abundance was barely detectable in qLT-HSC but greatly increased in aLT-HSC at 24 and 48h of in vitro culture, when cells are growing in size while still remaining undivided(18) (Fig. 1C;Fig. S1C). Conversely, autophagy-lysosomal pathway gene sets were among the most upregulated in qLT-HSC (Fig. 1A;Fig. S1B;Table S1). Interestingly, TFEB gene expression was enriched in qLT-HSC. Additionally TFEB, a protein typically only showing nuclear localization upon stress(19), was found to be nuclear in qLT-HSC (Fig. 1B,C;Fig. S1D). By contrast, TFEB levels and nuclear localization were reduced in aLT-HSC (Fig. 1C). We noted that other MiT/TFE family members were not expressed with the same trajectories across the human hematopoietic hierarchy(20, 21) (Fig. S1E,F; Fig. S2A,B). Importantly, of the two MiT/TFE family members that bind to genetic elements of the CLEAR network (TFEB and TFE3)(22) and show nuclear localization, only TFEB changes to cytoplasmic localization upon activation (Fig. S2C-E). MYC and TFEB expression anticorrelated at the single-cell level (Fig. 1D). LAMP1 staining revealed that aLT-HSC had expanded lysosomal mass compared to qLT-HSC (Fig. 1C). However, TFEB and genes involved in lysosome activity were found to be downregulated during LT-HSC activation suggesting this increased lysosomal content may be due to decreased lysosomal turnover (23, 24). We validated this idea by blocking lysosome acidification and degradation with short-term Bafilomycin (BAF) treatment and finding that qLT-HSC had higher accumulation of lysosomes compared to aLT-HSC (Fig. 1E;Fig S3A). In agreement, aLT-HSC displayed a gradual decrease in lysosomal acidification indicating less functional lysosomes (Fig S3B).

5 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

To directly assess MYC function in LT-HSC activation, we transduced LT-HSC with a MYC overexpressing lentivirus (LV) (MYCWT-OE) (Fig. 1F). MYCWT-OE LT-HSC doubled ~3 fold more than control cells and at 6 days post-transduction they exhibited increased mitochondrial mass and expression of anabolic genes involved in ribosome biogenesis, protein translation, mitochondrial metabolism and MYC target genes (Fig. 1G-I;Fig. S3C-F;Table S2). Conversely, 4 days of MYC inhibition(17) restrained LT-HSC activation as demonstrated by reduced mitochondrial mass, ROS production and more G0-quiescent cells (Fig. S3G-H). MYCWT-OE downregulated lysosomal-related genes, consistent with our observation that lysosomal degradation is reduced in aLT-HSC (Fig. 1H,I). Importantly, MYC inhibition increased TFEB nuclear localization and lysosomal activity, which was blunted by TFEB knockdown (shTFEB) (Fig. 1J,K). Collectively, our data show that MYC is upregulated upon in vitro mitogenic stimulation causing repression of TFEB-associated lysosomal programs thereby driving LT-HSC activation and anabolism.

TFEB-induced lysosomal activity inhibits MYC-driven anabolic processes and promotes LT-HSC quiescence

To gain insight into how TFEB regulates the transition from quiescence to activation, we performed transcriptomic analysis of control (CTRL) and TFEB-overexpressing (TFEBWT-OE) LT-HSC (Fig. 2A;Fig. S4A). As expected for a transcriptional activator, 67 out of 68 differentially expressed genes were upregulated in TFEBWT-OE compared to control LT-HSC (Fig. S4B;Table S3). Lysosomal-related gene sets were the most significantly enriched, including 14 different V-ATPase subunits that are required for lysosome and endosome acidification as well as cargo degradation (Fig. 2B,C;Fig. S4C;Table S3). Moreover, endosome transit to lysosomes or receptor endocytosis genes (such as transferrin or ) were concurrently upregulated, indicating an overall induction of the endolysosomal pathway (Fig. 2B-D). ATAC-Seq profiling uncovered 116 sites uniquely acquired in TFEBWT-OE vs 438 sites uniquely acquired in control LT-HSC (Table S4). TF recognition motif analysis identified specific E-boxes recognized by the MITF family of TFs as the most enriched binding motif in open chromatin regions gained upon TFEBWT-OE, while no significantly enriched motifs were found in control (Fig. S4D,E). Genes whose promoters fell within these gained elements showed enriched expression in TFEBWT-OE LT-HSC but depletion in shTFEB LT-HSC relative to

6 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

controls providing strong evidence that the transcriptional changes observed are directly mediated by TFEB binding events (Fig. S4F). As expected, TFEBWT-OE increased LT-HSC lysosomal activity compared to control as demonstrated by the higher number, acidification and turnover of lysosomes as well as higher Cathepsin expression and activity (Fig. 2E,F;Fig. S4G,H). Notably, levels of 1 (TfR1), which is required for iron-bound transferrin uptake through -mediated endocytosis, were significantly reduced in TFEBWT-OE cells, despite only mild changes in levels of TFRC mRNA (Fig. 2G;Fig. S4I). Confocal analysis confirmed the reduction of TfR1 membrane levels upon TFEB-OE and demonstrated the existence of TfR1+/LAMP1+ late endosomes/lysosomes (Fig. 2F;Fig. S4J). BAF treatment, previously shown to block TfR1 trafficking to late endosomes/lysosomes(25), or LAMP1 knockdown (shLAMP1) resulted in increased TfR1 membrane levels, which could not be rescued by TFEBWT-OE (Fig. 2H;Fig. S4K,L). Thus, we can conclude that TFEB induces clearance of TfR1 from the membrane of LT-HSC through endolysosomal degradation.

We hypothesized that TFEB-induced endolysosomal degradation of cell surface receptors such as TfR1 is required to maintain LT-HSC quiescence, and that inhibition of this pathway could lead to enhanced TfR1 membrane levels and LT-HSC activation. Whilst qLT-HSC were mostly negative for TfR1, increased TfR1 membrane expression correlated with MYC elevation and preceded upregulation of TFRC mRNA levels during LT-HSC activation (Fig, S4M,N). Notably, transcriptomic analysis showed a large and consistent downregulation of gene sets involved in cell cycle regulation, RNA processing/transport and ribosome biogenesis upon TFEBWT-OE (Fig. 2I;Fig. S5A-C;Table S3). This group of anabolic genes included MYC and MYC-target genes and its downregulation correlated with lower levels of mitochondria and ROS (Fig. 2I;Fig. S5A-D), indicating that TFEB maintains LT-HSC in a low metabolic state during culture. TFEBWT-OE restricted LT-HSC division over 6 days of time-lapse imaging compared to controls and cells remained Ki67 and EdU negative in flow cytometry analyses (Fig. 2J,K;Fig. S5E,F), indicating that TFEBWT-OE inhibits LT-HSC quiescence exit. Furthermore, single cell tracking of TFEBWT-OE LT-HSC showed delayed division kinetics among dividing cells relative to control, resulting in an overall reduced cellular expansion (Fig. 2L;Fig. S5E;Movie S1). Of note, the induction of this quiescent/low metabolic state upon TFEBWT-OE also protected LT- HSC from cell death during in vitro culture (Fig. 2J).

7 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

In contrast to TFEBWT-OE, shTFEB resulted in decreased lysosomal and Cathepsin B activity and higher expression of anabolic pathways and the number of cycling cells (Fig. 1K;Fig. 2I;Fig. S5B,C,G-I;Table S5). Indeed, the transcriptomic programs upregulated by shTFEB and MYCWT-OE in LT-HSC were highly concordant (Fig. S5C). In agreement, genes upregulated by TFEBWT-OE and downregulated by MYCWT-OE were enriched in qLT-HSC and in single cell RNA-seq subsets of bone marrow hematopoietic stem and progenitor cells classified as more dormant (non-primed)(7, 26-28) (Fig. S6A,B). Moreover, LT-HSC exhibited increased cell surface levels of TfR1 without significant changes in TFRC mRNA levels upon MYCWT-OE, whereas MYC inhibition reduced surface expression of TfR1, consistent with the observed MYC-directed repression of the endolysosomal pathway (Fig. 1I;Fig. 2M;Fig. S4I;Fig. S6C). Importantly, TFRC knockdown (shTFRC) reduced Ki67+ LT-HSC during culture, indicating that TfR1 is instrumental for LT-HSC activation (Fig. S6D,E). Overall, our findings demonstrate that TFEB suppresses MYC activation and restrains a MYC-driven anabolic program while inducing endolysosomal degradation of membrane receptors such as TfR1, maintaining a quiescent, low-metabolic state in LT-HSC.

Lysosomal activity governs LT-HSC lineage specification

As TfR1 is required for erythropoiesis(29), we next examined whether lysosomal activity differs among subpopulations of LT-HSC with distinct TfR1 membrane levels and whether this correlates with differing lineage commitment potential. LT-HSC, sorted on the basis of higher (LysoH) lysosomal activity in culture were reminiscent of TFEB-OE LT-HSC as they displayed higher acidification and levels of Cathepsin B activity and lower TfR1 surface expression levels than LT-HSC with low activity (LysoL) (Fig. 2E-G;Fig. 3A;Fig. S6F-H). Colony-forming cell (CFC) assays showed cloning efficiency was significantly higher from LysoL than LysoH LT- HSC, predominantly due to increased BFU-E (burst-forming unit-erythroid), although total cellular output was unchanged (Fig. 3B-D;Fig. S6I,J). The percentage and total number of cells expressing the erythroid marker GlyA and/or TfR1 were reduced in the progeny of LysoH compared to LysoL LT-HSC, while LysoH LT-HSC showed increased myeloid differentiation potential (CD33+/CD15+cells) (Fig. 3C,D;Fig. S6J). Similarly, single-cell stromal (SCS) assays showed that myeloid and erythroid differentiation are associated with higher and lower lysosomal activity, respectively Fig. S6K-N). Thus, the levels of lysosomal activity are inversely

8 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

correlated with TfR1 surface expression levels and are tied to LT-HSC erythroid-myeloid lineage fate choices.

We asked whether similar regulation of TfR1-lysosomal degradation by TFEB and MYC might occur as part of the typical sequence of events that occurs during lineage commitment of LT-HSC. Interestingly, the gene expression patterns of TFEB and MYC in subpopulations downstream of LT-HSC were also mutually exclusive: megakaryocyte-erythrocyte progenitors (MEP) had low TFEB and high MYC expression, while granulocyte-monocyte progenitors (GMP) exhibited the inverse expression pattern (Fig. 3E). TfR1 membrane levels were higher in MEP, correlating with MYC and anti-correlating with TFEB expression (Fig. 3E,F). Analysis of more mature populations revealed that TFEB was distinctly suppressed in erythroid progenitors (Fig. S7A,B).

To determine functionally whether the lysosomal program is instrumental in regulating erythroid/myeloid commitment of LT-HSC, we assessed LT-HSC differentiation potential after modulation of TFEB expression. TFEBWT-OE in LT-HSC caused a complete block of erythroid differentiation as demonstrated by an almost complete absence of BFU-E colonies and GlyA+ cells in both CFC and SCS assays; similar results occurred in ST-HSC upon TFEBWT-OE (Fig. 3G,H;Fig. S7C-K). shTFEB or shLAMP1 in LT-HSC produced the opposite phenotype (Fig. S7L-Q). Importantly, shTfR1 also reduced BFU colony formation similar to TFEBWT-OE in LT- HSC, demonstrating that TfR1 is important for early erythroid-commitment (Fig. 3I;Fig. S7R,S). Remarkably, TFEBWT-OE in committed MEP subpopulations (F1, F2/F3)(21) was sufficient to suppress TfR1 membrane levels and abolish their erythroid potential while slightly increasing their myeloid output in both CFC and SCS assays (Fig. S1A;Fig. 3J,K;Fig. S7T,V;Fig. S8A-I). Transcriptomic analysis of TFEBWT-OE MEP F2/3 subsets confirmed that TFEB imposes a myeloid-associated gene program and inhibits the expression of key transcription factors essential for erythrocyte and megakaryocyte differentiation, such as GATA1 or RUNX1 (Fig. S8J-N;Table S6). Thus, TFEB expression is sufficient to alter erythroid- myeloid fate decisions even in a progenitor subpopulation committed to erythropoiesis and provides strong evidence that TFEB-mediated induction of the endolysosomal pathway controls early lineage determination in LT-HSC as well as downstream progenitors.

9 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

To determine whether TFEB may exert its functions through mechanisms other than transcriptional activation, we overexpressed either TFEBWT or a TFEB variant unable to activate transcription due to disruption of its DNA binding domain (TFEBH240R-I243N) in hCB CD34+CD38- cells (Fig. S9A-D). Flow cytometric analysis showed that overexpression of TFEBWT but not TFEBH240R-I243N increased autophagosome/lysosome formation and restrained expansion of CD34+CD38- cells in vitro (Fig. S9E,F). In CFC assays, TFEBWT strongly inhibited erythroid commitment and enhanced myeloid differentiation, whereas TFEBH240R-I243N expression produced the opposite effect (Fig. 3L,M). These results indicate that TFEB exerts its functions primarily through transcriptional activation. We speculated that TFEBH240R-I243N acts in a dominant negative manner, perhaps through formation of homodimers sequestering other TFEB molecules. These results independently support our findings with TFEBWT-OE and establish that TFEB-mediated induction of lysosomal activity is required for myelopoiesis but is incompatible with erythroid differentiation. Overall, our study has uncovered lysosomes as transcriptionally-controlled central signaling hubs that govern lineage determination.

Enhanced TFEB-driven lysosomal activity distinguishes LT-HSC from ST-HSC

ST-HSC are the immediate downstream progeny of LT-HSC and share many transcriptional and functional properties. However, they differ in their quiescence exit kinetics and self-renewal properties(18). Since our data indicate that TFEB imposes a metabolically-low quiescent state in LT-HSC and restrains their expansion while inducing endolysosomal degradation of specific membrane receptors, we hypothesized that the TFEB-controlled lysosomal program could be differentially regulated in LT- vs ST-HSC. In relation to qST-HSC, qLT-HSC showed transcriptional enrichment of lysosomal pathway genes and contained variable but overall higher levels of lysosomes (21, 30) (Fig. 4A,B;Fig. S9G,H;Table S7). The observed enrichment of lysosomal genes was associated with higher levels of nuclear TFEB and lower levels of MYC in qLT- vs qST-HSC (Fig. 4A). In short term cultures, aLT-HSC also showed enrichment of TFEB and lysosomal genes with depletion of MYC and its target genes in comparison to aST-HSC (Fig. 4C,D;Fig. S9I;Table S8). Consistent with these transcriptomic results, aLT-HSC exhibited enhanced lysosomal activity as defined by Cathepsin B activity and reduced levels of mitochondria, ROS and TfR1 compared to aST-HSC (Fig. 4E-H). TFEBWT- OE in ST-HSC induced a transcriptional program similar to that of LT-HSC, with enrichment of

10 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

endolysosomal genes and depletion of cell cycle and biosynthetic related genes (Fig. 4I,J;Fig. S9J,K;Table S9). Functionally, TFEB-OE in ST-HSC induced lysosomal activity with inhibition of quiescence exit and reduction in TfR1 membrane levels (Fig. 4K-M;Fig. S9L). MYC inhibition suppressed mitogenic and metabolic activation of ST-HSC in culture (Fig. S9M). Thus, aspects of a LT-HSC-specific program can be imposed on ST-HSC upon ectopic TFEB expression or MYC inhibition. Overall, these findings indicate that an enhanced TFEB- regulated lysosomal program distinguishes LT- from ST-HSC and actively restrains the anabolic processes of LT-HSC with resultant limitation in their cellular output upon mitogenic activation.

TFEB-regulated lysosomal activity controls LT-HSC self-renewal

We next investigated whether enhanced lysosomal activity in LT-HSC impacted on self- renewal, the hallmark trait of stem cells. TFEBWT-OE in LT-HSC resulted in an initial decrease but subsequent increase in total colonies and cells in serial re-plating CFC assays (Fig. S10A-D). In contrast, shTFEB blunted LT-HSC self-renewal ability in secondary CFC assays, prompting us to directly assess LT-HSC self-renewal capacity using gold standard xenograft assays (Fig. S10E,F). CD34+CD38- hCB cells transduced with a TFEBWT-OE lentiviral vector generated significantly smaller grafts in both NSG and NSG-W41 mice at 4 and 17 weeks compared to controls, in keeping with the restraint in total cellular output observed in vitro (Fig. 5A,B;Fig. S10G-K). By contrast, TFEBH240R-I243N-OE where mice showed only modest reductions (Fig. 5A,B;Fig. S10G-K). Remarkably, despite their lower repopulation TFEBWT, but not TFEBH240R- I243N cells, showed significant enrichment of self-renewing LT-HSC as measured in serial xenotransplantation limiting dilution assays (LDA), with a 6.6 fold increase in stem cell frequency (Fig. 5C;Fig. S10L,M). These results suggest that TFEB governs LT-HSC self- renewal and exerts its effects primarily through DNA binding.

Next, we examined the effects of modulating TFEB activity through altering the phosphorylation of the S142 and S211 sites that are known to be targeted by several kinases (mTORC1, CDK4/6, ERK1/2)(19, 31). In nutrient-rich conditions, S142 and S211 phosphorylation of TFEB results in its cytoplasmic retention, whereas upon nutrient starvation, dephosphorylated TFEB translocates to the nucleus to induce the expression of lysosomal genes(13, 19, 31). Overexpression of a constitutively nuclear form of TFEB (TFEBS142A),

11 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

generated through disruption of the S142 phosphorylation site, in CD34+CD38- hCB cells led to significant reduction of graft size in primary transplanted mice but a large 22-fold increase in the frequency of self-renewing stem cells (Fig. 5A-C;Fig. S9A-F;Fig. S10G-M). The greater enrichment of stem cell frequency in TFEBS142 compared to TFEBWT-OE cells suggests that regulation of TFEB subcellular localization is important for LT-HSC self-renewal (Fig. 5A- C;Fig. S10G-M).

Silencing TFEB in CD34+CD38- hCB cells also resulted in smaller graft size in primary mouse recipients, but in contrast to the findings with TFEB-OE vectors, the reduced engraftment in the shTFEB group was now accompanied by a 2.8-fold reduction in stem cell frequency by serial LDA (Fig. 5D-F;Fig. S11A-G). Importantly, the myeloid and erythroid bias induced by TFEB-OE and TFEB-knockdown, respectively, was recapitulated in xenotransplantation experiments (Fig. S11H-J). Interestingly, LysoH LT-HSC exhibited a 3.8-fold increase in stem cell frequency despite showing a small reduction in primary graft size compared to LysoL LT- HSC, strengthening the link between lysosomal activity and LT-HSC stemness potential (Fig. 5G; Fig. S12A-E). In summary, LT-HSC self-renewal requires the TFEB-regulated lysosomal program to limit LT-HSC metabolic activity and expansion, whereas inhibition of this program leads to LT-HSC activation followed by exhaustion.

12 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Discussion

Here, we have identified an organelle-based model of stem cell fate determination where TFEB and MYC balance the activity of lysosomes to regulate the self-renewal and differentiation properties of human LT-HSC. In an unperturbed homeostatic setting, TFEB (normally implicated in stress responses) induces constitutive lysosomal flux in LT-HSC that actively maintains quiescence, preserves self-renewal and governs lineage commitment. These effects are tied to endolysosomal degradation of membrane receptors, pointing to a role for TFEB in coordinating how LT-HSC sense environmental changes and initiate the earliest steps of their fate transitions and lineage commitment decisions. Such transitions are delicately balanced by a TFEB/MYC dichotomy where MYC is a driver of LT-HSC anabolism and activation, counteracting TFEB function by serving as a negative transcriptional regulator of lysosomes; conversely TFEB counteracts MYC activity (Fig. S12F). TFEB activation is post- transcriptionally regulated through phosphorylation and subcellular localization by upstream kinases and results in the limiting of LT-HSC quiescence and self-renewal.

Our findings point to a mechanism whereby the TFEB-induced lysosomal program actively maintains LT-HSC in a quiescent state via endosomal trafficking of external sensing machinery including signaling and nutrient-uptake receptors to lysosomes for degradation, as exemplified here by the iron transporter TfR1. We hypothesize that this endolysosomal system renders quiescent LT-HSC more refractory to specific environmental cues in order to restrain aberrant mitogenic and anabolic activation and prevent LT-HSC exhaustion, as shown by our TFEB and TfR1 knockdown experiments. Quiescent mouse HSC express lower membrane levels of TfR1 compared to progenitor cells, which require higher levels of iron-uptake for proliferation and its deletion resulted in profound reconstitution defects(32). Regulation of TfR1 by lysosomes in qLT-HSC may have a protective role by preventing iron overload and the resultant ROS production that would be detrimental to HSC self-renewal(33). Our findings on TfR1 also help explain an emerging body of work on how lineage determination often occurs already within the stem cell compartment(21). Active suppression of TFEB and its downstream lysosomal degradation of TfR1 within LT-HSC is required for commitment along the erythroid lineage: indeed, activation of TFEB can abolish erythroid differentiation even after lineage commitment

13 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

has occurred. Our transcriptomic analysis of TFEB-OE LT-HSC indicates that endolysosomal degradation of other membrane receptors such as IGF1R (insulin receptor), which is upstream of the PI3K-Akt-mTORC1 pathway, could also be taking place in qLT-HSC, a finding consistent with observed EGFR degradation in neural stem cells(34). Thus, it is highly likely that the dichotomous TFEB/MYC-mediated control of lysosomal activity will be relevant in other tissue- specific adult stem cells such as neural and muscle stem cells that are maintained in quiescence for prolonged periods of time. In summary, our work sheds on the transcriptional control of lysosomes and its crucial r ole in stemness regulation, opening new routes to explore in regenerative medicine.

Acknowledgments

We thank the obstetrics units of Trillium Health, William Osler and Credit Valley Hospitals for CB; The UHN-Sickkids Flow cytometry facility for cell sorting; The Advance Optical Microscope Facility (AOMF) for confocal microscopy, Linda Penn, Jason DeMelo and Aaliya Tamachi for the MYC overexpressing vector and all lab members of the Dick lab for critical feedback.

Funding

This work to J.E.D was supported by funds from the: Princess Margaret Cancer Centre through funding provided by Ontario Ministry of Health, Princess Margaret Cancer Centre Foundation, Ontario Institute for Cancer Research through funding provided by the Government of Ontario, Canadian Institutes for Health Research (RN380110 - 409786), International Development Research Centre Ottawa Canada, Canadian Cancer Society (grant #703212), Terry Fox New Frontiers Program Project Grant, University of Toronto’s Medicine by Design initiative with funding from the Canada First Research Excellence Fund, a Canada Research Chair. LGP was supported by EMBO Long-Term Fellowship (ALTF 420-2017), Benjamin Pearl Fellowship and CIHR Fellowship (201910MFE-430959-284655). WW was supported by a SystemsX Swiss Initiative in Systems Biology Transition Postdoc fellowship and ETH Career Seed Grant. SNF grant 179490 to T.S.

14 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Author contributions

L.G.P conceived the study, performed research, analyzed data, and wrote the manuscript; F.S and W.W performed research and analyzed data; J.C, O.G, J.M, M.S, K.P, G.K, K.G, S.T and E.W performed research; V.V, A.M, M.C.S.Y and A.G.X.Z analyzed data; K.B.K and S.Z.X performed research and wrote the manuscript; and J.E.D wrote the manuscript, secured funding and supervised the study.

15 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Figure legends

Fig. 1. MYC drives LT-HSC activation by promoting anabolism and inhibiting lysosomal activity.

A. RNA-seq analysis of qLT-HSC vs cultured LT-HSC for 4 days (d) (aLT-HSC). n=3CB. B. MYC and TFEB expression in LT-HSC from (A). C. Confocal analysis of TFEB, MYC and LAMP1 in qLT-HSC and cultured LT-HSC for 24h or 48h. IntDen (Integrated Density). Scale 2 µm. n=3CB, 469-867 cells/staining. Mann-Whitney test. D. Correlation of normalized IntDen values for MYC and TFEB from (C). E. Confocal analysis of LAMP1 in quiescent or 24h- cultured LT-HSC treated with DMSO or BAF for 3h. Mann-Whitney test. n=3CB, 594 cells. F. Scheme for LT-HSC transduced with lentivirus expressing BFP (blue fluorescent protein) and CTRL or MYCWT genes. BFP+ cells were sorted at 3 or 6d for analysis. G. % BFP+ LT-HSC from (F) quantified by flow cytometry at 3 and 6d post-transduction. n=3CB. H. RNA-seq analysis of LT-HSC from (F) at 6d post-transduction. Normalized enrichment score (NES) of pathways differentially enriched in MYCWT vs CTRL LT-HSC by GSEA analysis. FDRq- value<=0.05. n=3CB. I. Expression of indicated genes in LT-HSC from (F). J. Confocal analysis of TFEB in LT-HSC treated for 4d with DMSO or MYC inhibitor. n=3CB, 272 cells. Mann- Whitney test. K. LysoSensor mean fluorescence intensity (MFI) of LT-HSC transduced with lentivirus expressing mCherry and a shRNA against Renilla (shCTRL) or TFEB (shTFEB) treated with DMSO or MYC inhibitor from 1 to 5d post-transduction. n=4CB. *P<0.05, **P<0.01, ***P<0.001. Unpaired t-test unless otherwise indicated.

Fig. 2. TFEB-induced endolysosomal degradation inhibits c-MYC-driven anabolic processes and promotes LT-HSC quiescence.

A. Scheme for LT-HSC transduced with lentivirus expressing BFP and CTRL or TFEBWT genes. BFP+ cells were sorted for analysis. B. RNA-seq analysis of LT-HSC from (A) at 3d post- transduction. NES value of pathways positively enriched in TFEBWT vs CTRL LT-HSC by GSEA analysis. FDRq-value<=0.05. n=3CB. C. RNA-seq expression of lysosomal genes in LT- HSC from (A). D. GSEA plots of indicated gene sets in LT-HSC from (A). FDRq-value<=0.05. E. LT-HSC from (A) stained with LysoTracker, LysoSensor and Magic Red and analyzed by

16 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

flow cytometry at 6d post-transduction. n=3CB. F. Confocal analysis of LAMP1 and TfR1 in LT-HSC from (A) treated for 3h with BAF or DMSO at 6d post-transduction. Scale 2 µm. n=3CB, 635 cells. Mann-Whitney test. G. TfR1 MFI analyzed by flow cytometry in LT-HSC from (A). Paired t-test. H. TfR1 MFI analyzed by flow cytometry in shCTRL and shTFEB LT- HSC treated with DMSO or BAF from 1 to 5d post-transduction. n=3CB. I. GSEA plots of indicated gene sets in LT-HSC from (A) and Fig. S5H. FDRq-value<=0.05. J. Distribution of cell fate by time-lapse imaging analysis of LT-HSC from (A), from 3 to 6d post-transduction. n=3CB. Mann-Whitney test. K. Cell cycle analysis of LT-HSC from (A) at 6d post- transduction. n=3CB. L. Time to first division analysis of LT-HSC from (A) from 3 to 6d post- transduction in hours (h). Mann-Whitney test. n=3CB. 192 cells. M. TfR1 MFI analyzed by flow cytometry in LT-HSC from Fig. 1F at 3 and 6d post-transduction. n=3CB. Paired t-test. *P<0.05, **P<0.01, ***P<0.001. Unpaired t-test unless otherwise indicated.

Fig. 3. Lysosomal activity governs LT-HSC lineage specification.

A. Experimental scheme: LT-HSC were cultured for 4d, stained with LysoTracker and sorted for the indicated differentiation assays. B,C,D. CFC colony distribution for LT-HSC from (A) scored under the microscope and GlyA+ and CD33+ lineage analysis by flow cytometry. n=3CB. E. RNA-seq normalized counts for TFEB and MYC in the indicated cell populations sorted as in Fig. S1A. n=3CB. F. TfR1 membrane expression analysis in MEP and GMP cells after 4d in culture. n=3CB. G,H. LT-HSC from Fig. S7C were plated for CFC assay as in (B,C), n=3CB. I. LT-HSC from Fig. S7R were plated for CFC assay as in (B,C), n=3CB. J,K. MEP F1 and MEPF2/3 cells from Fig. S7T were plated for CFC assay as in (B,C), n=3CB. L,M. CFC assay with CD34+CD38- cells from Fig. S9B as in (B,C). n=4CB. *P<0.05, **P<0.01, ***P<0.001. Unpaired t-test unless otherwise indicated.

Fig. 4. Enhanced TFEB-driven lysosomal activity differentiates LT from ST-HSC.

A,B. Confocal analysis of LT-HSC and ST-HSC stained for TFEB, LAMP1, MYC and DAPI. Scale 2 µm. n=3CB, 239-951 individual cells/staining. Mann-Whitney test. C,D. RNA-seq expression of indicated genes in LT- and ST-HSC cultured for 4d. n=2-3CB. E,F,G,H. Magic Red, TfR1, MitoTracker and CellROX MFI in LT- and ST-HSC cultured for 4d and analyzed by

17 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

flow cytometry. n=3-6CB. Paired t-test. I. Experimental scheme: ST-HSC transduced with lentiviral vectors expressing BFP and CTRL or TFEBWT genes. BFP+ cells were sorted at 3 and 6d for the indicated analyses. J. Expression of indicated genes in ST-HSC from (I). n=2CB. K. LysoTracker and Magic Red analyzed by flow cytometry in ST-HSC from (I). n=3CB. L. Cell cycle analysis of ST-HSC from (I). n=3CB. M. TfR1 MFI analyzed by flow cytometry on ST- HSC from (I). n=3CB. *P<0.05, **P<0.01, ***P<0.001. Unpaired t-test unless otherwise indicated.

Fig. 5. TFEB-directed lysosomal activity controls LT-HSC self-renewal.

A. Experimental scheme: CD34+CD38- cells isolated from hCB were transduced with lentiviral vectors expressing BFP and one of the following genes: CTRL, TFEBWT, TFEBS142A or TFEBH240R-I243N. At 1d post-transduction cells were intrafemorally injected into 8-10w old NSG- W41 and NSG mice. Male NSG mice were sacrificed at 4w, female NSG and male NSG-W41 mice at 17w post-xenotransplantation for analysis. CD45+/BFP+ cells were sorted from primary NSG mice at 17w for secondary limiting dilution assays (LDA) in NSG-GM3 mice. NSG-GM3 mice were sacrificed at 8w post-xenotransplantation for stem cell frequency analysis. B. From (A): LOG2 ratio of %BFP+ in hCD45+ cells at 17w of engraftment in NSG-W41 mice vs input (3d post-transduction). C. Stem cell frequency calculated from LDA. See Tables in Fig. S10L,M. D. As in (A) using lentivirus expressing mCherry and shCTRL or shTFEB. E. From (D): LOG2 ratio of %mCherry+ in hCD45+ cells at 17w of engraftment in NSG-W41 mice vs input (3d post- transduction). F. Stem cell frequency calculated from LDA. See Tables in Fig. S11F,G. G. Stem cell frequency calculated from LDA. See Tables in Fig. S12D,E. Each in vivo experiment was performed with three independent CB per experiment with 4-5 mice/CB. (RF: right femur, injected bone; BM: left femur and right and left tibiae; SP: spleen). *P<0.05, **P<0.01, ***P<0.001. Unpaired t-test unless otherwise indicated.

18 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Material and Methods

Methods Table REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies FITC–anti-CD45RA BD 555488 PE–anti-CD90 BD 555596 PECy5–anti-CD49f BD 551129 V450–anti-CD7 BD 642916 PECy7–anti-CD38 BD 335790 APC–anti-CD10 BD 340923 APCCy7–anti-CD34 BD custom made by BD BV711-CD19 BD 563036 Alexafluor700–anti-CD7 BD 561603 Alexafluor700–anti-CD10 BD 563500 BV510-CD45 BD 563891 biotin-Flt3L BD N/A Streptavidin-Qdot605 ThermoFisher Q10101MP Streptavidin-PE BD 554061 V450-anti-CD15 BD 642917 FITC-anti-CD3 349201 349201 V500-anti-CD45 BD 560777 PE–anti-CD19 BD 349209 PE–anti-GlyA Beckman Coulter A07792 BV786–anti-CD33 BD 740974 PECy5–anti-CD45 Beckman Coulter A07785 PECy7–anti-CD14 Beckman Coulter A22331 APC–anti-CD33 BD 551378 APC–anti-CD45 BD 340943 FITC–anti-CD45 BD 347463

19 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

V421–anti-CD10 BD 562902 BV605 anti-CD56 BD 562780 PE-anti-CD71 BD 555537 APCCy7 anti-CD41 Beckman Coulter 6607115 BV786-anti-CD71 BD 563768 Ki67-PE BD 556027 Anti-TFEB R&D biosystems MAB9170-100 Anti- MYC Alexa647 R&D biosystems IC36961R-100ug Anti-LAMP1 Alexa488 R&D biosystems IC7985G Anti-CD71 APC BD 551374 Purified Mouse Anti-Human CD71 BD 555534 Human Cathepsin D Antibody R&D biosystems AF1014 Goat Anti-mouse secondary antibody Alexa568 Thermo Fisher A-11004 Goat Anti-rabbit secondary antibody Alexa488 Thermo Fisher A-11008 Biological Samples Human umbilical cord blood samples Trillium, Credit N/A Valley and Brampton Civic Hospital

Chemicals, Peptides, and Recombinant Proteins DNase I Roche 11284932001 Ammonium Chloride Stem Cell 07850 Technologies FLT3 Ligand Milteny Biotec 130-096-474 IL6 Milteny Biotec 130-093-934 SCF Milteny Biotec 130-096-696 TPO Milteny Biotec 130-095-752 EPO Janssen Eprex 10,000 IU/ml IL3 Milteny Biotec 130-095-068

20 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

GM-CSF Milteny Biotec 130-093-866 AmpliTaq Gold 360 Polymerase ThermoFisher 4398813 BamH1-HF NEB R0136S Mlu1-HF Invitrogene 15432-016 KOD Hot Start DNA Polymerase Millipore-Sigma 71842 DpnI New England Biolabs R0176S Paraformaldehyde 16% Solution (methanol-free) Agar scientific AGR1026 Poly-L-lysine solution Sigma P8920 IGEPAL® CA-630 Sigma I8896-50ML DMSO Fisher Chemical D128-500 Bafilomycin A1 Sigma B1793 MYC inhibitor Milipore-sigma 475956-10MG DAPI Sigma 10236276001 Fluoromount G Thermo Fisher 00-4958-02 Tween 20 Sigma P9416-100ML Triton X-100 Sigma T8787-100 Commercial Assays StemSep™ Human Hematopoietic Progenitor Cell Stem Cell 14066 Enrichment Kit Technologies CD34 MicroBead Kit, human Milteny Biotec 130-097-047 LS Columns Milteny Biotec 130-042-401 Mouse Cell Depletion Kit Milteny Biotec 130-104-694 SuperScript VILO ThermoFisher 1754050 Power SYBR Green ThermoFisher 4367659 Magic Red™ Cathepsin B Kit BioRad ICT937 RNeasy Micro Kit Qiagen 74004 PicoPure® RNA Isolation Kit Thermo Fisher KIT0204 GatewayTM LR ClonaseTM II Enzyme Mix Thermo Fisher 11791020 CYTO-ID® Autophagy detection Enzo Life Sciences ENZ-51031- K200

21 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

RNase-Free DNase Set (50) Qiagen 79254 Click-iT® Plus EdU Alexa Fluor® 647 Flow Thermo Fisher C10634 Cytometry CellROX Deep Red ThermoFisher C10422 CellROX Orange ThermoFisher C10443 MitoTracker Green ThermoFisher M7514 LysoTracker Blue ThermoFisher L7525 LysoTracker Green ThermoFisher L7526 LysoTracker Deep Red ThermoFisher L12492 LysoSensor Green DND-189 ThermoFisher L7535 RNeasy Plus Micro Kit Qiagen 74034 Sytoxblue ThermoFisher S11348 Propidium Iodide ThermoFisher P3566 BD cytofix/cytoperm fixation solution BD 554722 BD PermWash BD 554723 Phusion High fildelity PCR master Mix with HF NEB M0531S buffer MiniElute PCR purification Kit Qiagen 28004 Nextera DNA library prep kit 96 samples Illumina 15028211 Syber Green ThermoFisher S7563 Hoechst 33342 BD 561908 Other X-vivo medium Lonza 04-380Q H5100 media StemCell 05150 Technologies Fetal bovine serum Sigma F1051-500mL Iscove's modified Dulbecco's medium (IMDM) Gibco 12440-053 MethoCultTMOptimum Stem Cell H4034 Technologies StemProTM-34 SFM Gibco 10640-019

22 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

StemPro nutrients Gibco 10641-025 L-glutamine Multicell 609-065-EL Pen/Strep Gibco 15140-122 Human LDL Stem Cell 02698 Technologies 8 Well Chamber, removable ibidi 80841 12 Well Chamber, removable ibidi

Lead contact and materials availability

Further information and requests for resources and unique/stable reagents generated in this study should be directed to and are available from the Lead Contact, John E. Dick ([email protected]), but a completed Materials Transfer Agreement may be required.

Method details

Human cord blood samples Human CB samples were obtained with informed consent from Trillium Health, Credit Valley and William Osler Hospitals according to procedures approved by the University Health Network (UHN) Research Ethics Board. Mononuclear cells (MNC) from pools of male and female CB units (~4-15 units) were obtained by centrifugation on Lymphoprep medium, and after ammonium chloride lysis MNC were enriched for CD34+ cells by positive selection with the CD34 Microbead kit and LS column purification with MACS magnet technology (Miltenyi). Resulting CD34+ CB cells were viably stored in 50% PBS, 40% fetal bovine serum (FBS) and 10% DMSO at -80°C or −150 °C.

Mice Animal experiments were done in accordance with institutional guidelines approved by University Health Network Animal care committee. All in vivo experiments were done with 8- to 12-week-old female/male NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) mice (JAX) and 8- to 12-week-old female/male NOD.Cg-PrkdcscidIl2rgtm1WjlTg(CMV-

23 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

IL3,CSF2,KITLG)1Eav/MloySzJ (NSG-SGM3) that were sublethally irradiated with 225 cGy, 24 h before transplantation, or with 8- to 12-week-old male NOD.Cg- PrkdcscidIl2rgtm1WjlKitem1Mvw/SzJ (NSGW41) mice that were not irradiated. All mice were housed at the animal facility (ARC) at Princess Margaret Cancer Centre in a room designated only for immunocompromised mice with individually ventilated racks equipped with complete sterile micro-isolator caging (IVC), on corn-cob bedding and supplied with environmental enrichment in the form of a red house/tube and a cotton nestlet. Cages are changed every <7 days under a biological safety cabinet. Health status is monitored using a combination of soiled bedding sentinels and environmental monitoring.

Flow Cytometric Analysis and Sorting Sorting: CD34+ and CD34- human CB cells were thawed via slow dropwise addition of X-VIVO 10 medium with 50% FBS and DNaseI (200 μg/ml). Cells were spun at 350g for 10min at 4 °C and then resuspended in PBS+2.5% FBS. For all in vitro and in vivo experiments, the full stem and progenitor hierarchy sort was performed as described in Notta et al.(21) and shown in Fig. S1A and Fig. S7B. Cells were resuspended in 100 μl per 1x106 cells and stained in two subsequent rounds for 15 min at room temperature each. See Table 1 for antibodies. Cell cycle: Cells were cultured with EdU for 1 hour and harvested for EdU Click-it reaction following manufacturer’s instructions. Cells were then incubated with PE-anti-Ki67 antibody in PermWash solution overnight at 4°C. Prior to flow cytometry analysis cells were stained with Hoechst 33342. Cell dyes: Cells were incubated for 30 min at 37°C with the indicated dyes listed in Table 1 following manufacturer’s instructions (CellROX, 5μM; MitoTracker, 0.1μM; LysoTracker, 75nM; Magic Red, 1/260 dilution; CytoID, 1/500 dilution; LysoSensor, 1/500 dilution) in respective culture media. After staining, cells were washed once and resuspended in PBS+2.5% FBS and analyzed by flow cytometry. BD sorters FACSAria II, FACSAria III and FACSAria FUSION; and BD analyzers FACSCelesta were used.

Immunostaining analysis

24 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

Cells were spun onto Poly-L-Lysine-coated slides (200 xg, 10 min), fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton before blocking (PBS, 10% FBS, 5% BSA). Slides were incubated with primary antibodies listed in Table 1 in blocking solution O/N at 4°C. Secondary antibodies also listed in Table 1 were added in PBS, 0.025% Tween for 1.5 h at room temperature (RT). After washing, nuclei were stained with 1 μg/mL DAPI and slides were mounted (Fluoromount G). Single cell images were captured by a Zeiss LSM700 Confocal Microscope (oil, 63x/1.4NA, Zen 2012) and processed and analyzed with ImageJ/Fiji and FlowJo10. Antibody dilutions used are the following: LAMP-1 1/200; TFEB 1/50; MYC 1/100; TfR1 1/50; Cathepsin D 1/100. Analysis of TFEB-MYC correlation in single cell immunostaining data: fluorescence intensities for LAMP1, nuclear DAPI, MYC, nuclear MYC, TFEB, and nuclear TFEB were normalized to 10,000 for each cell and subject to log transformation.

Time-lapse imaging o 0.4 Time-lapse experiments were conducted at 37 C, 5% O2, 5%CO2 on μ-slide VI channel slides (IBIDI) coated with 20 μg/ml anti-human CD43-biotin antibody(37). BFP+ cells were sorted 3 days post-lentiviral transduction and cultured overnight in phenol red free IMDM supplemented with 20% BIT (Stemcell Technologies), 100 ng/mL human recombinant Stem Cell Factor (SCF), 50 ng/mL human recombinant Thrombopoietin (TPO), 100 ng/mL human Fms-related tyrosine kinase 3 ligand (Flt3L, all R&D Systems), 2 mM L-GlutaMAX (Gibco), 100 U/mL penicillin and 100 µg/mL Streptomycin (Gibco) before imaging. Brightfield images were acquired every 8 minutes for 3-4 days using a Nikon-Ti Eclipse equipped with a Hamamatsu Orca Flash 4.0 camera and a 10x CFI Plan Apochromat λ objective (NA 0.45). Single-cell tracking and fate assignment were performed using self-written software as previously described(11, 35, 36). Time to division was calculated using R 3.5.3.

Cloning of lentiviral overexpression constructs Lentiviral constructs expressing TFEB variants were constructed by GatewayTM cloning from pENTR223 donor plasmids into a lentiviral pRRL-based and GatewayTM adapted vector (pLBC2-BS-RFCA(37)) downstream of a SFFV promoter and upstream of tagBFP driven by an EFS/SV40 chimeric promoter using the GatewayTM LR ClonaseTM II Enzyme Mix according to

25 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

manufacturer’s instructions (Thermo Fisher). The original donor plasmid pENTR223-TFEB (HsCD00373101) was obtained from PlasmID (DF/HCC DNA Resource Core at Harvard Medical School). By standard site-directed PCR mutagenesis a STOP-codon was inserted (fw primer: 5’-TAG ACC CAG CTT TCT TGT ACA AAG TTG-3’; rev primer: 5’-TCA CAG CAC ATC GCC CTC C-3’) and subsequently pENTR223-TFEB-S142A (fw primer: 5’- CAC CCA TGG CCA TGC TGC AC-3’; rev primer: 5’- CAT TGG GAG CAC TGT TGC CAG C-3’) and pENTR223-TFEB- H240R_ I243N (fw primer: 5’- CGG AAC TTA AAT GAA AGG AGA CGA AGG; rev primer: 5’- ATT GTC TTT CTT CTG CCG CTC C-3’) were generated using KOD Hot Start DNA Polymerase (Millipore-Sigma) followed by DpnI (NEB) digestion. The negative control (CTRL) vector for overexpression encodes gp91phoxP415H (catalytic inactive gp91phox/CYBB; (35) that was previously cloned into the same lentiviral GatewayTM adapted vector (pLBC2-BS-RFCA(37)). MYC overexpressing vector was kindly provided by Linda Penn’s lab.

Cloning of lentiviral knockdown constructs The negative control (CTRL) vector for knockdown is a shRNA directed against Renilla luciferase (shCTRL)(37). Additional shRNA sequences were predicted using the Sherwood algorithm as in(38) and ordered as Ultramer DNA oligos (IDT). Subsequently, shRNAs were amplified using AmpliTaq Gold 360 Polymerase (ThermoFisher) using shRNA amplification Forward and Reverse primers. The PCR product was subsequently digested with BamH1 and Mlu1 and cloned into a UltramiR scaffold (miR30) within a pRRL-based vector downstream of a SFFV promoter and upstream stream of mCherry (pLBC2-mCherry). shRenilla: TGCTGTTGACAGTGAGCGCAGGAATTATAATGCTTATCTATAGTGAAGCCACAGATG TATAGATAAGCATTATAATTCCTATGCCTACTGCCTCGGA1 shTFEB-2: TGCTGTTGACAGTGAGCGAACGATGTCCTTGGCTACATCATAGTGAAGCCACAGATG TATGATGTAGCCAAGGACATCGTCTGCCTACTGCCTCGGA shTFEB-3: TGCTGTTGACAGTGAGCGAGATGATGTCATTGACAACATATAGTGAAGCCACAGAT GTATATGTTGTCAATGACATCATCCTGCCTACTGCCTCGGA

26 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

shTFEB-4: TGCTGTTGACAGTGAGCGCAGAAAGACAATCACAACTTAATAGTGAAGCCACAGAT GTATTAAGTTGTGATTGTCTTTCTTTGCCTACTGCCTCGGA shLAMP-1: TGCTGTTGACAGTGAGCGATGGACGAGAACAGCATGCTGATAGTGAAGCCACAGAT GTATCAGCATGCTGTTCTCGTCCAGTGCCTACTGCCTCGGA shLAMP-2: TGCTGTTGACAGTGAGCGCCACGAGAAATGCAACACGTTATAGTGAAGCCACAGAT GTATAACGTGTTGCATTTCTCGTGATGCCTACTGCCTCGGA shLAMP-3: TGCTGTTGACAGTGAGCGACAGCTCATGAGTTTTGTTTAATAGTGAAGCCACAGATG TATTAAACAAAACTCATGAGCTGGTGCCTACTGCCTCGGA shLAMP-4: TGCTGTTGACAGTGAGCGCGACTGTGGAATCTATAACTGATAGTGAAGCCACAGAT GTATCAGTTATAGATTCCACAGTCTTGCCTACTGCCTCGGA shTFRC-1: TGCTGTTGACAGTGAGCGCGAACCAGATCACTATGTTGTATAGTGAAGCCACAGATG TATACAACATAGTGATCTGGTTCTTGCCTACTGCCTCGGA shTFRC-2: TGCTGTTGACAGTGAGCGCCCAGATCACTATGTTGTAGTATAGTGAAGCCACAGATG TATACTACAACATAGTGATCTGGTTGCCTACTGCCTCGGA shTFRC-3: TGCTGTTGACAGTGAGCGCCTCAATGATCGTGTCATGAGATAGTGAAGCCACAGATG TATCTCATGACACGATCATTGAGTTGCCTACTGCCTCGGA shTFRC-4: TGCTGTTGACAGTGAGCGCCAGCCAACTGCTTTCATTTGATAGTGAAGCCACAGATG TATCAAATGAAAGCAGTTGGCTGTTGCCTACTGCCTCGGA

Lentiviral Production and Transduction VSV-G pseudotyped lentiviral vector particles were produced and titers were determined as

27 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

previously described(37). Unless stated otherwise, after 16-20 hours of pre-stimulation in low cytokine media (see descriptioon below) cells were transduced with lentiviral vectors described above at matching multiplicity of infection(37) aiming at mid-range (20-40%) transduction efficiencies but without lentiviral preparation exceeding 20% of total culture volume. Transduction efficiency (%BFP+ or %mCherry+) was determined at day 3 post-transduction by flow cytometry on a BD Celesta, which served as initial input estimate for xenotransplantation assays.

Cell Culture For in vitro experiments sorted cells were cultured in 96 well-plate round bottom with the indicated cell media. - high cytokine media: StemPro-34 SFM media with the supplied supplement, 1x L-Glutamine, 1x Pen/Strep, 0.02% Human LDL and cytokines FLT3L (20 ng/mL), GMCSF (20 ng/mL), SCF (100 ng/mL), TPO (100 ng/mL), EPO (3 U/mL), IL-3 (10 ng/mL), IL-6 (50 ng/mL). This media was used to culture cells with the presence of DMSO, MYC inhibitor or BAF. - low cytokine media: X-Vivo 10, 1% BSA, L-Glutamine, Pen/Strep and cytokines SCF (100 ng/ml), Flt3L (100 ng/ml), TPO (50 ng/ml) and IL7 (IL-7; 10 ng/ml). This media was used to transduce cells as indicated above. For drug studies of untransduced cells, DMSO or MYC inhibitor (53.3 μM) were added at the time of seeding in high cytokine media and cells were harvested 4 days later for indicated analysis. For drug studies of transduced cells, DMSO, MYC inhibitor (53.3 μM) or Bafilomycin (20 nM) were added in high cytokine media 1day after transduction with lentiviral vectors. Cells were harvested 4 days later for indicated analysis. For lysosomal turnover studies, Bafilomycin (20 nM) or equivalent amount of DMSO was added to the cultures for 3 hours in StemPro-34 SFM media (1x L-Glutamine, 1x Pen/Strep, 0.02% ) without cytokines.

Colony-Forming Cell Assays For LysoH and LysoL cells: ~ 20K LT-HSC were cultured for 4 days in high cytokine media and stained with LysoTracker as described above. For each group (based on LysoTracker

28 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

fluorescence) 150 cells were sorted directly into 2 mL methylcellulose, supplemented with FLT3 Ligand (10 ng/mL) and IL6 (10 ng/mL) and plated onto 2x35 mm dishes as duplicates.

For transduced cells: At day 3 post-transduction, 150 LT-HSC BFP+ or mCherry+, 200 ST-HSC BFP+, 300 MEP F1 BFP+ or 300 MEP F2/3 BFP+ cells were sorted directly into 2 mL methylcellulose, supplemented with FLT3 Ligand (10 ng/mL) and IL6 (10 ng/mL) and plated onto 2x35 mm dishes as duplicates. Colonies were allowed to differentiate for 10-11 days and then morphologically assessed in a blind fashion by a second investigator. Subsequently, colonies from replicate plates were pooled, resuspended in PBS/5% FBS and stained for flow cytometry analysis. For serial CFC assays, 0.5% of progeny from LT-HSC was added to fresh methylcellulose as above, replated and scored after 10-11 days.

Single-Cell Stromal Assays Single cell in vitro assays were set up as described previously(39) with low passage murine MS- 5 stroma cells seeded at a density of 1500 cells per 96-well and grown for 2-4 days in H5100 media. One-day prior to coculture initiation, the H5100 media was removed and replaced with 100 μl erythro-myeloid differentiation media: StemPro-34 SFM media with the supplied supplement, 1x L-Glutamine, 1x Pen/Strep, 0.02% Human LDL and the following cytokines: FLT3L (20 ng/mL), GMCSF (20 ng/mL), SCF (100 ng/mL), TPO (100 ng/mL), EPO (3 U/mL), IL-2 (10 ng/mL), IL-3 (10 ng/mL), IL-6 (50 ng/mL), IL-7 (20 ng/mL), and IL-11 (50 ng/mL). Sorted single-cells were deposited in each well (80 wells/96-well plate). Colonies were scored after 15-17 days under the microscope and every individual well containing a visible colony was stained with antibodies listed in Table1 and analyzed by flow cytometry using a BD FACSCelesta instrument equipped with a high throughput sampler (HTS).

Xenotransplantation The progeny of 10K CD34+CD38- transduced cells one day after transduction were intra-femoral injected in aged and gender matched 8-12 wk old male and female recipient NSG, NSG-SGM3 or NSG-W41 mice. At indicated time points, mice were euthanized, injected femur and other long bones (non-injected femur, tibiae) were flushed separately in Iscove’s modified Dulbecco’s

29 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

medium (IMDM)+5%FBS and 5% of cells were analyzed for human chimerism along with BFP or mCherry and antibodies listed in Table 1. Sick and miss-injected mice were excluded from analysis. For purification of human cells from xenotransplanted mice, fresh or thawed BM from individual or from pools of 2-5 mice were mouse depleted (Mouse Cell Depletion Kit, Miltenyi) according to manufacturer’s protocol and CD45+/BFP+ or CD45+/mCherry+ were sorted for serial transplantation by intra-femur injection in NSG-SGM3 mice at the indicated cell doses. A mouse was considered engrafted if % of human CD45+ cells > 0.1. For LDA experiments LT- HSC frequency was estimated using the ELDA software (http://bioinf.wehi.edu.au/software/elda/). For LysoH and LysoL LT-HSC: ~ 20K LT-HSC were cultured for 4 days in high cytokine media and stained with LysoTracker as described above. For each group (based on LysoTracker fluorescence) 1K LT-HSC were intra-femoral injected in aged and gender matched 8-12 wk old male and female recipient NSG-W41 or NSG-SGM3 mice and. Engraftment analysis of primary and secondary xenograft was performed as described above.

RNA-seq processing and analysis Freshly sorted populations from 3-5 independent CB pools or transduced cells in vitro after a second sort (PI-BFP+ or PI-mCherry+) on day 3 or 6 post-transduction were directly resuspended and frozen (-80°C) in PicoPure RNA Isolation Kit Extraction Buffer. RNA was isolated using the PicoPure RNA Isolation Kit (Thermo Fisher) according to manufacturer’s instructions. Samples that passed quality control according to integrity (RIN>8) and concentration as verified on a Bioanalyzer pico chip (Agilent Technologies) were subjected to further processing by the Center for Applied Genomics, Sick Kids Hospital: cDNA conversion was performed using SMART-Seq v4 Ultra Low Input RNA Kit for Sequencing (Takara) and libraries were prepared using Nextera XT DNA Library Preparation Kit (Illumina). Equimolar quantities of libraries were pooled and sequenced with 4 cDNA libraries per lane on a High Throughput Run Mode Flowcell with v4 sequencing chemistry on the Illumina 30 HiSeq 2500 following manufacturer’s protocol generating paired-end reads of 125-bp in length to reach depth of 55-75 million reads per sample. Reads were then aligned with STAR v2.5.2b against hg38 and annotated with ensembl v90. Default parameters were used except for the following: chimSegmentMin 12; chimJunctionOverhangMin 12; alignSJDBoverhangMin 10; alignMatesGapMax 100000;

30 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

alignIntronMax 100000; chimSegmentReadGapMax parameter 3; alignSJstitchMismatchNmax 5 -1 5 5. Read counts were generated using HTSeq v0.7.2 and general statistics were obtained from picard v2.6.0. Differential gene expression was performed using edgeR_3.24.3 following recommended practices. For the qLT-HSC vs aLT-HSC comparison (Supplementary Table S1) the LT-HSC from the dataset GSE125345 were compared to the cultured CTRL LT-HSC from the TFEB-OE experiment. For the aLT-HSC vs aST-HSC comparison (Supplementary Table S8), CTRL LT-HSC and ST-HSC from the TFEB-OE experiment were compared. These data were used in a single sample gene set variation analysis using TFEB-OE and MYC- OE up and downregulated genes (top 100, 250 and 500 genes) using the gsva function of the R package GSVA_1.30.0. Conditional Quantile Normalization (CQN) from the cqn R package was applied on TMM normalized raw counts in order to correct for gene length effect. Expression of TFEB, TFEC, TFE3 and MITF was extracted from the data and plotted in each population using R stripchart.

Transcriptomic analysis of published datasets Several transcriptomics data containing HSC and other hematopoietic populations from cord blood (CB), bone marrow (BM), mobilized peripheral blood cells (mPB) or fetal liver (FL) tissues were obtained from Gene Expression Omnibus data portal (GEO). GSE42414 (Laurenti et al. 2013, PMID: 23708252)(30) Illumina HumanHT-12 beadchip expression data from lineage-depleted CB cells were downloaded from GEO. Normalized log2-transformed signal was used in a moderated t test in the R limma_3.38.3 package to compare HSC1 (Lin- CD34+ CD38- CD45RA- CD90+ CD49f+) and MPP Lin- CD34+ CD38- CD45RA- CD90- CD49f-). A file was obtained by ordering all genes using the t statistics from upregulated to downregulated genes in HSC1 compared to MPP and GSEA_4.0.3 was run using default parameters using the KEGG lysosome as gene-set. GSE76234 (Notta et al. 2016, PMID: 26541609)(21) FPKM expression data from Illumina HiSeq 2000 RNA sequencing aligned on hg19 were downloaded from GEO. Median gene expression for HSC, MPP(+-), MPP(++), and MPP(--) lineage-depleted CB populations were available and used for this analysis. These data were used in a single sample gene set variation analysis with the KEGG-Lysosome gene-set using the gsva

31 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

function of the R package GSVA_1.30.0. They were also used to plot the expression of TFEB, TFEC, TFE3 and MITF in each population. GSE109093 (Cesana et al. 2018, PMID: 29625070)(20) FPKM expression data from Illumina HiSeq 2000 RNA sequencing aligned on hg19 were downloaded from GEO. Median gene expression for HSC (CD34+ CD38– CD90+ CD45RA–) and committed CD34+CD38+ progenitor populations (PROG) from BM, FL and CB were available and used for this analysis. These data were used to plot the expression of TFEB, TFEC, TFE3 and MITF in each population. GSE17054, GSE19599, GSE11864, E-MEXP-1242 (Hemaexplorer) Log2 batch corrected expression signal from Affymetrix U133 Plus 2.0 Array were downloaded from Bloodspot (http://servers.binf.ku.dk/bloodspot/). It includes normal hematopoietic populations from BM and peripheral blood. Expression of TFEB, TFEC, TFE3 and MITF was extracted from the data and plot in each population using R stripchart.

Pathway enrichment analysis and visualization Pathway enrichment analysis and visualization was performed as described previously(40). Briefly, a score to rank genes from top up-regulated to down-regulated was calculated using the formula -sign(logFC) * -log10(pvalue). The rank file from each comparison was used in GSEA analysis (http://software.broadinstitute.org/gsea/index.jsp) using 2000 permutations and default parameters against indicated gene sets. All gene sets were obtained from a pathway database http://baderlab.org/GeneSets. EnrichmentMap version 3.1.0 in Cytoscape 3.7.0 was used to visualize enriched gene-sets with indicated FDR-q value and NES and a Jaccard coefficient set to 0.375. Full results in the comparison of CTRL vs TFEBWT-OE LT-HSC can be seen in Supplementary Table 3. g:profiler (https://biit.cs.ut.ee/gprofiler/gost) was performed using the upregulated genes by TFEB-OE (FDRq-value<=0.01), results can be seen in Supplementary Table 3.

scRNA-seq Signature Enrichment Using signatures derived from over-expression of TFEB and MYC in LT-HSCs (FDR < 0.05), we scored single CD34+CD38- HSPCs from three publicly available datasets for their relative expression of each signature. We then compared signature enrichment between cell cycle-primed

32 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

HSPCs and non-primed HSPCs, as defined in Xie et al.(7) The MYC OE signature was higher in the cycle-primed HSPCs while the TFEB OE signature was higher in the non-primed HSPCs, and these results were consistent across all three datasets(26,27,28).

Comparison of TFEB-regulated genes in MEP F2/3 with uncultured CB populations

The Gene Expression Omnibus dataset GSE125345 contains standardized RNA-Seq gene expression data of distinct hematopoietic cell states from uncultured cord blood. MEP, EryP, GMP, Gr, Mono and MEP populations were selected to be compared with the genes differentially expressed in TFEBWT vs CTRL MEP F2/3 cells. The top 250 genes enriched in MEP F2/3 TFEBWT and the top 250 genes enriched in MEP F2/3 CTRL were selected to be used as the reference signatures. Firstly, bar graphs were created by selecting the MEP F2/3 TFEBWT reference signature, and calculating the number of scaled data that were above (>0) or below (<0) the mean for each population, corrected by the number of samples per population and 1,000 random permutations. Secondly, enrichment of both the TFEBWT and CTRL MEP F2/3 reference signatures were estimated in each sample of the uncultured cord blood population using ssgsea() from the GSVA_1.30.0 R package.

ATAC-seq processing and analysis Transduced BFP+ LT-HSC: Library preparation for ATAC-Seq was performed on 1000-5000 cells with Nextera DNA Sample Preparation kit (Illumina), according to previously reported protocol(41). 4 ATAC-seq libraries were sequenced per lane in HiSeq 2500 System (Illumina) to generate paired-end 50-bp reads. Reads were mapped to hg38 using BWA (0.7.15) using default parameters. Duplicate reads, reads mapped to mitochondria, an ENCODE blacklisted region or an unspecified contig were removed (Encode Project Consortium, 2012). MACS (2.2.5) was used to call peaks in mapped reads. A catalogue of all peaks was obtained by concatenating all peaks and merging any overlapping peaks. Peaks were considered unique to one condition or another if they were present in at least 2 out of three replicates but not in the contrasting condition. Homer (4.11.1) was used to calculate enrichment of subsets of peaks using default parameters plus the catalogue of called peaks as background.

33 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

References 1. E. Laurenti, B. Gottgens, From haematopoietic stem cells to complex differentiation landscapes. Nature 553, 418-426 (2018). 2. K. Ito, M. Bonora, K. Ito, Metabolism as master of hematopoietic stem cell fate. Int J Hematol 109, 18-27 (2019). 3. L. Garcia-Prat, P. Sousa-Victor, P. Munoz-Canoves, Proteostatic and Metabolic Control of Stemness. Cell Stem Cell 20, 593-608 (2017). 4. W. Palm, C. B. Thompson, Nutrient acquisition strategies of mammalian cells. Nature 546, 234-242 (2017). 5. K. Ito et al., Self-renewal of a purified Tie2+ hematopoietic stem cell population relies on mitochondrial clearance. Science 354, 1156-1160 (2016). 6. T. T. Ho et al., Autophagy maintains the metabolism and function of young and old stem cells. Nature 543, 205-210 (2017). 7. S. Z. Xie et al., Sphingolipid Modulation Activates Proteostasis Programs to Govern Human Hematopoietic Stem Cell Self-Renewal. Cell Stem Cell 25, 639-653 e637 (2019). 8. L. Garcia-Prat et al., Autophagy maintains stemness by preventing senescence. Nature 529, 37-42 (2016). 9. D. S. Leeman et al., Lysosome activation clears aggregates and enhances quiescent neural stem cell activation during aging. Science 359, 1277-1283 (2018). 10. R. E. Lawrence, R. Zoncu, The lysosome as a cellular centre for signalling, metabolism and quality control. Nat Cell Biol 21, 133-142 (2019). 11. D. Loeffler et al., Asymmetric lysosome inheritance predicts activation of haematopoietic stem cells. Nature 573, 426-429 (2019). 12. R. Liang et al., Restraining Lysosomal Activity Preserves Hematopoietic Stem Cell Quiescence and Potency. Cell Stem Cell 26, 359-376 e357 (2020). 13. G. Napolitano, A. Ballabio, TFEB at a glance. J Cell Sci 129, 2475-2481 (2016). 14. I. Annunziata et al., MYC competes with MiT/TFE in regulating lysosomal biogenesis and autophagy through an epigenetic rheostat. Nat Commun 10, 3623 (2019). 15. A. Wilson et al., c-Myc controls the balance between hematopoietic stem cell self- renewal and differentiation. Genes Dev 18, 2747-2763 (2004). 16. E. Laurenti et al., Hematopoietic stem cell function and survival depend on c-Myc and N- Myc activity. Cell Stem Cell 3, 611-624 (2008). 17. N. Cabezas-Wallscheid et al., Vitamin A-Retinoic Acid Signaling Regulates Hematopoietic Stem Cell Dormancy. Cell 169, 807-823 e819 (2017). 18. E. Laurenti et al., CDK6 levels regulate quiescence exit in human hematopoietic stem cells. Cell Stem Cell 16, 302-313 (2015). 19. R. Puertollano, S. M. Ferguson, J. Brugarolas, A. Ballabio, The complex relationship between TFEB transcription factor phosphorylation and subcellular localization. EMBO J 37, (2018). 20. M. Cesana et al., A CLK3-HMGA2 Axis Impacts Human Hematopoietic Stem Cell Molecular Identity throughout Development. Cell Stem Cell 22, 575-588 e577 (2018). 21. F. Notta et al., Distinct routes of lineage development reshape the human blood hierarchy across ontogeny. Science 351, aab2116 (2016).

34 bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

22. J. A. Martina, H. I. Diab, H. Li, R. Puertollano, Novel roles for the MiTF/TFE family of transcription factors in organelle biogenesis, nutrient sensing, and energy homeostasis. Cell Mol Life Sci 71, 2483-2497 (2014). 23. J. P. Luzio, Y. Hackmann, N. M. Dieckmann, G. M. Griffiths, The biogenesis of lysosomes and lysosome-related organelles. Cold Spring Harb Perspect Biol 6, a016840 (2014). 24. A. P. Mutvei et al., Rap1-GTPases control mTORC1 activity by coordinating lysosome organization with amino acid availability. Nat Commun 11, 1416 (2020). 25. H. Cao, B. Schroeder, J. Chen, M. B. Schott, M. A. McNiven, The Endocytic Fate of the Transferrin Receptor Is Regulated by c-Abl Kinase. J Biol Chem 291, 16424-16437 (2016). 26. L. Velten et al., Human haematopoietic stem cell lineage commitment is a continuous process. Nat Cell Biol 19, 271-281 (2017). 27. D. Pellin et al., A comprehensive single cell transcriptional landscape of human hematopoietic progenitors. Nat Commun 10, 2395 (2019). 28. P. van Galen et al., Single-Cell RNA-Seq Reveals AML Hierarchies Relevant to Disease Progression and Immunity. Cell 176, 1265-1281 e1224 (2019). 29. I. C. Moura, O. Hermine, C. Lacombe, P. Mayeux, Erythropoiesis and transferrin receptors. Curr Opin Hematol 22, 193-198 (2015). 30. E. Laurenti et al., The transcriptional architecture of early human hematopoiesis identifies multilevel control of lymphoid commitment. Nat Immunol 14, 756-763 (2013). 31. Q. Yin et al., CDK4/6 regulate lysosome biogenesis through TFEB/TFE3. J Cell Biol 219, (2020). 32. S. Wang et al., -mediated iron uptake plays an essential role in hematopoiesis. Haematologica, (2019). 33. Y. Muto, M. Nishiyama, A. Nita, T. Moroishi, K. I. Nakayama, Essential role of FBXL5- mediated cellular iron homeostasis in maintenance of hematopoietic stem cells. Nat Commun 8, 16114 (2017). 34. T. Kobayashi et al., Enhanced lysosomal degradation maintains the quiescent state of neural stem cells. Nat Commun 10, 5446 (2019). 35. O. Hilsenbeck et al., Software tools for single-cell tracking and quantification of cellular and molecular properties. Nat Biotechnol 34, 703-706 (2016). 36. O. Hilsenbeck et al., fastER: a user-friendly tool for ultrafast and robust cell segmentation in large-scale microscopy. Bioinformatics 33, 2020-2028 (2017). 37. K. B. Kaufmann et al., A stemness screen reveals C3orf54/INKA1 as a promoter of human leukemia stem cell latency. Blood 133, 2198-2211 (2019). 38. S. R. V. Knott et al., A computational algorithm to predict shRNA potency. Mol Cell 56, 796-807 (2014). 39. E. Wagenblast et al., Functional profiling of single CRISPR/Cas9-edited human long- term hematopoietic stem cells. Nat Commun 10, 4730 (2019). 40. J. Reimand et al., Pathway enrichment analysis and visualization of omics data using g:Profiler, GSEA, Cytoscape and EnrichmentMap. Nat Protoc 14, 482-517 (2019). 41. J. D. Buenrostro, P. G. Giresi, L. C. Zaba, H. Y. Chang, W. J. Greenleaf, Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA- binding proteins and nucleosome position. Nat Methods 10, 1213-1218 (2013).

35 ✱✱✱✱ ✱✱✱✱ ✱✱✱✱ ✱✱✱✱ ✱✱✱✱ ns e 5 ✱✱✱✱ A NDUFAB1 B2 C 10 ✱✱✱✱ NOP16 ✱✱✱✱ ODC1 4 CDC20 4 ✱✱ e 10 NDUDUTFAB1 ✱✱✱ 10 NOP1RAN 6 2 Scor 4

e

e ODCPCN1A 02 − 2 10 ✱✱✱✱ CDC2CCNA02 autophagosome assembly DUCCTT 5 Scor Scor autophagosome assembly −

Scor RABUBN3 0 TFEB−

w Z

− PCNGNLA3 0 w Z 0

o MAD2L1 w Z autophagosome organization CCNA2 o

-2 R autophagosomCCTPRDXe 5organizatio3 n o 4 -2 R 10 BHPUBR3T1 R -2

GNLMCM37 w Z lysosome MAD2LSNRPD11 -2 o PRDXRRM13 lysosome TYMS HPRT1 R MYC MYC targets V1 MCMVDAC71 SNRPDNME1 1 3 3 RRMIDMYCS 1 targets V1 10 10 TYMSLC11AS 1 CB mitochondrial translation elongation VATP6V0ADAC1 1 CB NMELAPTM41 A qLT-HSC -2 e CB bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720mitochondrial translationIDGGAS 1 elongation ; thisr version posted February 24, 2021. The copyright holder3 for this preprint ARSA 10

SLC11A1 o

cell cycle checkpoints ACP5 MYC IntDen AMAN2BTP6V0A11 CB

(which was not certified by peer review) is theLAPTM4A author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Lamp1 IntDen cell cyclGGATCIRGe 1checkpoint1 GP91s aLT-HSC IDUA Sc 3 translational.termination ARSNAPSA A 10 TFEB IntDen total

0 − ASCARBCP5 2

availableMAN2B1 under aCC-BY-NC-NDZ GP91 4.0 International license. CTSK TFEB IntDen nuclear translational.terminationGP91 TCIRGCTSS 1 ribosome biogenesis IDHGSNUA AT MEP_T NAPSTPP1A w SCARB2 GAA 2 o 2 2 ribosomCTSKeF biogenesisGP91 2 mitochondrial gene expression CTSS 10 10 10

PSAP R HGSNCTSOAT 1 1 1 B B B TPPGABARAPL1 2 C C C 0h 24h 48h 0h 24h 48h mitochondrial genGAULKAe2 expression 0h 24h 48h 0h 24h 48h GP9 GP9 CTSMAP1LC3FGP9 A MYC PSAP ATG2A 1 1 1 B B B

CTSSTX1O2 C C C GMP GABARAPLATG14 2 targets ULKUBQLN2 1 ✱✱✱✱ GP9 GP9 GP9 MAP1LC3GABARAPLA 1 MYC D E 3.5 AMAP1LC3TG2A B DMSO 10 STX1WDR452 B MEP ATP53INPTG1✱✱✱✱4 1 targets ns MEP_G UBQLNATG16L12 BAF GABARAPLGABARAP 1 MAP1LC3ULK1 B EryDMSOP WDR45PACS2 B lysosome/auto- 3 TP53INPUBQLN21 10 AWDR4TG16L52 GABARASMURF1P phagosome Merge ULKIRG1M Mono MYC TFEB LAMP1 DAPI PMAP1LC3BACS2 2lysosome/auto- BAF UBQLN2 WDR45 MER SMURF1 phagosome

1 1 2.5 1 B B B 10 IRGM Gr MAP1LC3B2 C C C

GP9 GP9 GP9 GE 1 1 1 B B B IntDen 0h

C C C 102 Lamp1 IntDen GP9 GP9 GP9

r = -0.53 1.5

TFEB 10 p < 2.2e-16 101 0h 24h MYC IntDen 24h

F + G ✱ Sort BFP cells Sort BFP+ cells 4 CTRL

5 5 WT 10 10 MYC LV WT cells) 4 4 MYC 10 10

CTRL LV + LT-HSC 3 3 3 10 10 Sorting cells +

0 0 BFP BFP 2 0 50K 100K 150K 200K 250K 0 50K 100K 150K 200K 250K 48h

NAP1L1 e hCB GSPT1 2 3d 6d 1 TFDP1 RPL6

(6vs 3 days) CLNS1A Scor SF3BCCT35 e RPL1CCT64 A 02 − 1 day %BFP FC SSCCTB 7 Analysis Analysis PA2GCCT42

w Z GOPTGEST2 3 Scor

o -2 − pre-stimulation 0 MAD2LPCNA1 0

TUFCCTM3 R Fold change (%BFP NOLCCCT14

ABCETCP1 w Z

NES SSBPUBE21 N -2 o POLEPARP31 Lysosome MYC targets R RRMHNRNP1 U -3 -2 -1 0 1 2 3 NOP5DEK6 LT-HSC GNLGNL3 3 CORX5PAA2 LT

e

H I HDAC2 e 2 KPNANAP1L21 CTRL LYSOSOME%KEGG%HSA04142 ATP6V0D1 DHX1GTPBPGSPT5 14 2 E RSL1DNUCKSTFDP11 1 RPL6 LT Scor HNRNPMYCLNS1C DA WT VACUOLAR ACIDIFICATION%GO:0007035 GUSB Scor 0 − HNRNPASF3BURK3BR MYC 0 GP91 − CTSA EEF1BRFCRPL1424 HALLMARK_INTERFERON_ALPHA_RESPONSE%MSIGDB_C2 RPSDNMTSS5B 1

w Z PA2G4 LT

GLORFC1 2 w Z ATP6V0A2 -2 o GOT2 REGULATION OF PHOSPHATIDYLINOSITOL 3-K SIGNALING%GO:0014066 STSMC1ARDMAD2L7A1 -2 o R SET MYC USPTUF1M R REGULATION OF RESPONSE TO WOUNDING%GO:1903034 CTSW CULRIFNOLC11 1 -2LT e USPCDCABCE1 61 r CTPSSSBP1 1

GNPTG RAD51 o HALLMARK_MYC_TARGETS_V1%MSIGDB_C2 POLE3 TFEBOE ETFRFCRRM1 31 ABCB9 DDX2TFRNOP51C 6 LT NOP1MSHGNL623 HALLMARK_MYC_TARGETS_V2%MSIGDB_C2 LT Sc SREXOSC1CMOX5A 0 FUCA1 E PRDXHDA4C2 0 − MSHDHX165 E RRNA PROCESSING%REACTOME%R-HSA-72312.3 Z EIF4DHXE9 RSL1D1 HEXA NDUTRIM2HNRNPFAB81D TRANSLATION%GO:0006412 XPHNRNPOT R MEP_T CDK9 w MANBA SRSFMPHOSPHEEF1B1 2 8

LT RPS5 o RIBOSOME BIOGENESIS%GO:0042254 UBAKMT22 A 2LT SMARCCGLO1 1 AP1S2 MYC CHEKSTARD17 IMPDH2 MYC R MITOCHONDRIAL GENE EXPRESSION%GO:0140053 BRCASET 1 NPC2 PHAKTCULB 11 USP1 CCNARFC25 MITOCHONDRIAL TRANSLATION_GO:0032543 EIF2SCTPS1 1 AP1M1 TIMELESETF1 S PHBANKRD1DDX22 1 7 OXIDATIVE PHOSPHORYLATION_GO:0006119 CLTB HPERCCNOP1RT1 61 GMP SYNCRITHOCSRM 1P PRDX4 KPNBMAD2L1 2 ATP SYNTHESIS COUPLED ELECTRON TRANSPORT%GO:0042773 NMEEIF41 E ATP6V0A1 KPNA1 NCBPNDU1FAB1 SPIDXPOTR MEP CELLULAR AMINO ACID BIOSYNTHETIC PROCESS%GO:0008652 PWP1 NEU1 CCDC88SRSF1 A TXNL4UBA2A DCP2 MEP_G HSPESMARCC1 1 HALLMARK_G2M_CHECKPOINT%MSIGDB_C2 UHRF1 CTSD PSMCIMPDH6 2 POLPHBQ EryP DDX18 SMGCCNA52 APC C-DEGRADATION OF CELL CYCLE PROTEINS%REACTOME174143 HGSNAT XPO1 SPIEIF2S1 1 SRPKPHB12 GBA PRPSFHPOXMR2T1 shCTRL+DMSO LAS1TP5SYNCRIL3 P Mono TGFBKPNB11 AP4M1 WDR4NME31 EIF1ABAXX MAPKAPKNCBP1 5 NAGLU HNRNPWPP1A3 shCTRL+MYC inhibitor FBALTXNL4CD A Gr 15000 SUPV3LDDX1HSPE111 p=0.08 PSMC6 DNASE2 PPRCAUNI1P CEBPDDX1G8 TARDBXPO1P J DMSO K ACP2 TERF1 DAPI Merge TFEB shTFEB+DMSO MRSRPKTO4 1 NIPESCOPRPS7 2 PLA2G15 NOPCDTLAS121L ✱✱ WDR43 NDUSLFN1FAF14 TIPIEIF1AN X shTFEB+MYC inhibitor CTNS RCLHNRN1 PA3 MYC inhibitor MEN1 PRMTFBL3 p=0.054 CTSF PESCHEKSUPV3L1 2 1 PSMDDSCCPPRC3 1 TARDBP p=0.057 PSMANEK12 ✱✱✱✱ CTSO BLMRMTO4 2.0 SORNIPD7 TFB2RNF16NOPM2 8 4 MCOLN1 RABEPPNDUARPBFKAFP4 shCTRL+DMSO FAM120USPRCL71 A 10 PRKCPRMTQ3 ARSA PABPCPES14 USP37 DMSO MYBBP1PSMD3A 10000 IDUA HKPRKCPSMA2 D1 UTP2WRAP5SOR0D 3 ns shCTRL+MYC inhibitor PUSTERFTFB21 M2 CTSK OGGRABEP1 K 1.5 PTGESFAM1203 A CCTPOLPABPC3 H 4 SLC11A1 CCTDNAMYBBP12 2 A C1QBWRHKN2P shTFEB+DMSO CLTCL1 PRDXPRKDUTP23 02 PUS1 CANTICRPTGESX R 3 CLN3 HNRNPMAP3KCCT3U4 EIF3FCCTAM168B 2 A SMPD1 VDCTCC1QBAC11 P shTFEB+MYC inhibitor SERBPNAFPRDX113 CANX

CD71 BV786 MFI 1.0 CLTA ODCPINXHNRNP1 1 U PABPCERCCEIF3B14 CCTCHTF1VD7AC18 ATP6AP1 CCTPIFSERBP41 1 SLC25AODC13 KLHL1PABPC51 ASAH1 EIF4G2 E2FCCT77 5000 PGKWICCTZ1 4 3 LAPTM4A HSPDHELSLC25AB1 3 RAEREEIF4GN G 2 10 0.5 CCTPGK5 1 TPP1 GZMHSPDA1 TCPSLXRA1N4 CD63 CNBSICCTRPT65 HNRNKDM4TCPP1A2BD 1 nuclear TFEB IntDen nuclear NPMFGFRCNB1 P4 LAPTM5 PPIHNRNA PA2B1 RNPMTEL11 HSP90ABHSP90ABPPIA 1 1 LysoSensor Green MFI MYC inhibitor MYC CD164 LDHHMGBHSP90ABA 1 1 0.0 HNRNLDHA PA2B1 HSP90AA1 1 2 3 3 1 2 1 2 3 3 1 2 1 2 3 3 1 2 1 2 3 1 2 3 T_HSC_E L T_HSC_E_ T_HSC_E_ T_HSC_E L L L T_HSC_E_ T_HSC_E_ T_HSC_E L L L T_HSC_MYC_ T_HSC_MYC_ T_HSC_MYC_ T_HSC_E_ T_HSC_E_ L L L L L TFEBOE_ TFEBOE_ TFEBOE_ Figure 1 T_HSC_MYC_ T_HSC_MYC_ T_HSC_MYC_ L L L T_HSC_MYC_ T_HSC_MYC_ T_HSC_MYC_ L L L TFEBOE_GP9_ TFEBOE_GP91_ TFEBOE_GP91_ NES A TFEBWT LV Sort BFP+ cells Sort BFP+ cells B C Lysosome + 5 5 10 10 0 1 2 3 Transferrin endocytosis

4 4 CTRL LV 10 10 LYSOSOME%KEGG%HSA04142 WT -2 e CTRL TFEB 3 3 LT-HSC 10 10 r INSULIN RECEPTOR RECYCLING_R-HSA-77387.3 ATP6V0E1 2 e

o BFP Sorting BFP TRANSFERRIN ENDOCYTOSIS AND RECYCLING_R-HSA-917977.1 ATP6V0D1 0 0

Scor

Sc GLYCOLIPID METABOLIC PROCESS_GO:0006664 ATP6V1F 0 − 0 50K 100K 150K 200K 250K 0 − 0 50K 100K 150K 200K 250K

PH REDUCTION_GO:0045851 Z ATP6V1B2

w Z

MEMBRANE LIPID METABOLIC PROCESS_GO:0006643 LIPA -2 o MEP_T w

R bioRxiv preprint doi:3d https://doi.org/10.1101/2021.02.24.4327206d ; this versionHALLMARK_HYPOXIA%MSIGDB_ posted February 24, 2021. The copyright 2holder o for this preprint TPP1 hCB CARBOHYDRATE DERIVATIVE CATABOLIC PROCESS_GO:1901136 R ATP6V0B (which was not certified by peer review) is the author/funder, who hasVACUOLAR granted TRANSPORT_GO:0007034 bioRxiv a license to display the preprint in perpetuity. It is made ATP6V1G1 1 day Analysis Analysis available under aCC-BY-NC-NDORGANELLE MEMBRANE 4.0 FUSION_GO:0090174 International license. HEXB GMP M6PR LT pre-stimulation ENDOSOME TO LYSOSOME TRANSPORT_GO:00083331.51.5 ✱✱✱✱✱✱ CTRLCTRL ATP6AP1 GP91 cells) cells) MEP CTSB + MEP_G + TFEBTFEBWTWT CTSD EryP ATP6V1H LT D E ✱✱✱✱ 1.01.0 ✱✱ GGA2 Transferrin Endocytosis Endocytic genes 5000 3000 ✱ TFEBOE 30000 Mono ATP6V1A CTSS 4000 AP1S2 NES 2.13 NES 2.29 Gr 2000 0.50.5 ATP6V0A2 3000 20000 CTSA WT WT ATP6V1D TFEB TFEB GM2A 2000 1000 ATP6V1C1

% of Max 0.0 Fold change (%BFP 0.0 10000 NEU1 Fold change (%BFP CTRL NPC2 CTRL 1000 Magic Red MFI ATP6V1E1 (Cathepsin B activity) LysoSensor Green MFI GLB1 LysoTracker Green MFI 0 0 0 SCARB2 SUMF1 LysoTracker Green MANBA F SLC11A2 GNPTG DAPI Merge TfR1 LAMP1 ✱✱✱✱ 1.51.5 NAGLU ✱✱✱✱✱✱ CTRL GBA ✱✱✱ G CTRL cells)

cells) DNASE2 + ✱✱ + TFEBTFEBWTWT SORT1 104 ATP6V0A1 ✱✱✱✱ 4 1200012000 10 1.01.0 ✱✱✱✱ NAGPA ✱✱✱✱ HEXA CTRL AP1B1 GNS 0.5 GALNS 0.5 GAA 100 103 103 60006000 SGSH MCOLN1 80

TfR1 MFI CTNS TfR1 MFI TfR1 IntDen WT Lamp1 IntDen 0.0 % of Max GALC Fold change (%BFP 0.0 Fold change (%BFP 60 +BAF ATP6V1E2 CTSH 102 102 PSAP 40 0 0 TFEB CD63

Normalized To Mode WT WT WT PPT1 CTRL 20 CTRL CTRL ASAH1 TFEB TFEB TFEB TfR1-BV786 1.2 1 2 3 1 2 3 0 CTRL+DMSO 1.0 0 2 4 6 10 10 10 10 p=0.063 Cell death H ✱ CTRL+BAF I J Comp-BV786-A :: CD71 0.8 Dividing

MYC targets V1 MYC targets V2 TFEBOE_ TFEBOE_ TFEBOE_ ✱ TFEBWT+DMSO 1.2 100 100 0.6 Non-Dividing ✱✱✱✱ WT TFEB +BAF BAF TFEBOE_GP9_ 60000 DMSO TFEBOE_GP91_ 1.0TFEBOE_GP91_ ✱✱✱ 0.4 p=0.063 Cell death 80 80 NES -2.45 NES 2.49 NES -1.94 NES 1.71Frequency 0.8 0.2 Dividing 40000 60 60 WT TFEB shTFEB-4 TFEBWT shTFEB-4 0.6 ✱✱✱✱ Non-Dividing 0.0 40 40 0.4 Frequency

Normalized To Mode Normalized To Mode WT TfR1 MFI 20000 % of Max shCTRL 20 20 CTRL shCTRL CTRL CTRL 0.2 TFEB ✱✱✱✱ 0 0 0.0 0 0 2 4 6 0 2 4 6 10 10 10 10 10 10 10 10 WT Comp-BV786-A :: CD71 Comp-BV786-A :: CD71 TfR1-BV786 CTRL TFEB 1.51.5 K ✱✱✱100✱✱✱ L M CTRLCTRL WT ✱✱✱✱ CTRLCTRL cells) cells) CTRL TFEB 1.51.5 120

+ ✱✱✱

+ ✱✱✱ 80 WTWT 20000 40000 WTWT 80 ✱✱✱ TFEBTFEB CTRLCTRL ✱ MYCMYC

cells) CTRL✱ cells) 100 + 1.01.0 + WTWT 6 days 60 TFEBTFEB WT 30000 MYC 60 ✱✱✱✱ 80 1.01.0 15000 40

Normalized To Mode 60

0.5 40 PE 20000

0.5 - 20 % Cells

40 TfR1 MFI 0.50.5 TfR1 MFI 10000

✱ Ki67 20 0 10000 % of Max 0 2 4 6

10 10 10 10 Time to first division (h)20 0.0 Fold change (%BFP 0.0 Fold change (%BFP Comp-BV786-A :: CD71 0 0.0 0 5000 0 Fold change (%BFP 0.0 G0 G1 S/G2/M Fold change (%BFP Hoechst WT 3 days 6 days TfR1-PE CTRL TFEB Figure 2 ✱

L H A Lyso Lyso Colony-Forming 50Cell B C ✱✱✱ -20% -20% (CFC) assay Mix 120 L H ✱ + LT-HSC ✱✱✱ GlyA Lyso Lyso (10-11days)40 120 Sorting 50 BFU100 + Mix CD33GlyA+ 100 + 30 40 GM 80 BFU CD33 M GM 80 30 60 20 60 bioRxiv4 days preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version postedG FebruaryM 24, 2021. The copyright holder for this preprint #Colonies

20 %Cells 40 G

(which was not certified by peer review) is the author/funder, who#Colonies has granted bioRxiv a license to display the preprint in perpetuity. It is made hCB #Colonies 40 10 CD33 CD33 BV786 available under aCC-BY-NC-ND 4.0 International% positive cells license. 10 20 20 Single-Cell Stromal 0 0 0 0 (SCS) assay LysoHLysoL H L H L LysoTracker H L GlyA PE ✱✱✱✱ (16-17 days) Lyso Lyso Lyso Lyso Lyso 8 8 Lyso 9.0 ✱✱✱ 7 CB1 7 D LysoL LysoH E Counts Normalized F 60 6 6 H CB2 80 ✱✱✱✱ 80 Lyso H 8.5 5 5 Lyso ✱✱✱ CB3 L L TFEB 8.0 40 4 cells 4 LysoLyso cells + 60 60 + 3 MLP GMP CMP MEP 3 LT-HSC ST-HSC cells cells cells 12 + + 2.02.0 + 20 40 40 % of Max %TfR1 %CD71 1.5 10 1.5 MYC

%TfR1 20 %CD15

%CD71 20 1.01.0 CD15 FITC 0 MLP GMP CMP MEP MEP GMP 0.50.5 LT-HSCST-HSC 0 0 TfR1 BV786 % Human engraftment % Human engraftment 0.0 0.0 ✱ TfR1 BV650 ✱ ✱

50 50 50 ✱✱✱ Mix 60 ✱ Mix Mix G 120 H WT BFU MEP F1 MEP F2/3 + CTRL TFEB I J 40 40 20 ✱ BFU GlyA 40 GMBFU ✱✱ BFU 100 150 Mix 120 100 ✱ ✱✱ CD33+ Mix GM BFU GlyA M GM Mix GM 30 30 40 100 30 15 80 GM CD33 G 80 BFU BFU M M M M 100 80 GM GM 20 60 G 20 20 60 10 G G G #Colonies M M #Colonies #Colonies 60 #Colonies

# Colonies 20 %Cells 40 10 40 50 10 10 G G #Colonies #Colonies #Colonies 40

#Colonies 5 CD33 CD33 BV786 % positive cells 20 20 20 0 0 0 0 H L 0 0 H 0 L 0 LysoHLyso0 L Lyso Lyso WT Lyso Lyso L WT CTRL H GlyA PE WT WT TFEB CTRL TFEB shCTRL shTFRC CTRL CTRL Lyso Lyso ✱ TFEB TFEB

K L M ✱✱✱ 50 MIX ✱✱✱ 120 120 ✱ Mix + 120 ✱✱ * BFU GlyA + 120 ✱✱✱ WT 120 GlyA 40 100 100 + H240R-I243N ✱✱✱✱ + BFUGM CD33 CTRL + TFEB TFEB 100 GlyA + GlyA 100 CD33+ 100 + CD33 80 GMM 80 CD33 80 30 *** 80 M G 80 60 60 60 20 60 G 60 #Colonies

%Cells 40 %Colonies 40 %Cells 4040 40

10 * % positive cells 20 % positive cells 2020 20 20 CD33 CD33 BV786 0 0 ** 0 H0 L 0 0 H L WT WT Lyso Lyso H L WT CTRL CTRL WT CTRL Lyso Lyso GlyA PE TFEB TFEB TFEB CTRL Lyso Lyso H240R-I243N TFEB H240R-I243N TFEB TFEB Figure 3 A B ✱✱✱✱ Lysosome Merge c-MYC TFEB LAMP1 DAPI ✱✱✱✱ ✱✱✱ Merge 104 104 104 NES 1.71 HSC - MEP MEP

LT LT-HSC _ _ bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version 10posted3 February 24, 32021. The copyright3 holder for this preprint

G 10 T 10 (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made ST-HSC

available under aCC-BY-NC-ND 4.0MYC IntDen International license. Lamp1 IntDen

e HSC 2 -

e MYC

2 HDGF TFEB IntDen nuclear

MYC Scor ABCE1 CCT5 e HDGF 0 − 2 2 2 ST ETF1 2 Scor CCT6A ABCE1 10 10 10 CCT5 e CTPSCCT71 ETF1 02 − CCT6A DDX2CCT21 w Z CTPSCCT71 PTGES3 Scor

SRSF1 -2 o DDX2CCT21 PCNA 0 − w Z

Scor KPNB1 PTGES3 CCT3 R SRSF1 -2 o PCNA 0 − SMARCCCCT4 1 KPNB1

R CCT3 PA2GTCP41 w Z SMARCC1 CCT4 PLKUBE21 N -2 o LT-HSC PA2GTCP41 w Z PARP1 LT-HSC LT-HSC PPM1G R ST-HSC PLKUBE21 N -2 o HNRNPU ST-HSC ST-HSC PARP1 EIF2SDEK 1 PPM1G R HNRNPU HNRNGNL3PA3 EIF2S1 LT- DEK PSMDRPA28 HNRNPA3 GNL3 LT- POLEKPNA32 GP91LT-HSC 1500 C PSMDRPA28 D Lysosome GTPBP4 E F PHB2 MYC targets ✱ POLEKPNA32 GP91LT-HSC NUCKS1 LT GTPBP4 TRIM2MYC 8 LT-HSC LT-HSC PHB2 NUCKS1 LT NOLCAURK1B ST-HSC

TRIM28 GP91 e MYC TFDPRFC14 2 NOLCAURK1B ST-HSC DNMT1 GP91 TMEM97 ST MYC ST-HSC ST-HSC TFDPRFC14 RFC2 e GSPTATP6V0D1 1 HDGF 2 Scor TMEM9DNMT17 ST SMC1A ABCE1

GP91 − RFC2 CCNAUSPASAH12 1 ETF1 0 ✱✱ GSPT1 SMC1A GP91 SF3BRIF13 LT Scor CTPS1

− CCNA2 e DDX21 w Z USP1 e TOMM7CDCTPP6 0 1 -20 1400

o 2 r SRSF1 ✱ SF3BRIF13 LT RAD51 -2 LT-HSC MYC SUPV3L1 TFEBOE KPNB1 LIPA R 2500 CDC6 RFC3 w Z LT-HSC TOMM70 o HDGF SMARCC1

Scor RAD51 1 TCOFTFRC1 -2 o SUPV3L1 ABCE1 TFEBOE − HEXB PA2G4 1000 RFC3 ETF1 0 IFRDMSH12 R TFEBTFEB PLK1 TCOFTFRC1 CTPS1 FAM120EXOSC1A 0 Sc DDX21 PPM1G IFRD1 0 w Z GUSB ST-HSC MSH2 MSH6 EIF2S1 0 − SRSF1 -2 o RCL1 FAM120EXOSC1A 0 DHX9 HNRNPA3 2000 1200 KPNB1 Z R LT- ST-HSC BYSAP3SL 1 MYCMYCRCLMSH1 6 SMARCC1 -1 TRIM28 PSMD8 DHX9 PA2G4 PABPCCDK94 POLE3 GP91 BYSL AP1M1 LT TRIM28 MEP_T PLK1 PRMTMPHOSPH3 8 w PHB2 PABPCCDK94 PPM1G KMT2A TRIM28 TFB2GNPM TAB 2GP91 o PRMT3 EIF2S1 NOLC1 MPHOSPH8 HNRNPA3 CHEK1 1500 LT- CDK2 TFDP1 TFB2KMT2MA PSMD8 BRCAATP6V11 H R PPRC1 TMEM97 ST CHEK1 POLE3 GP91 AKT1 1000 CDK2 GSPT1 BRCA1 PHB2 CARFCDNPC5 2 GP91 -2 CCNA2

PPRC1 0 TRIM28 2 AKT1 TARDBTIMELESP S Green MFI MitoTracker NOLC1 SF3B3 CARFCD 5 CTSS 2 0 -2 TFDP1 PLKANKRD14 7 ST TOMM70 1000 500 TARDBTIMELESP S TMEM97 ST ERCC1 GMP SUPV3L1 ERHAP3M2 GP91 % of Max PLKANKRD14 7 GSPT1 GP91 THOC1 TCOF1 ERCC1 CCNA2 EIF3MAD2LJ 2 IFRD1 ERH SF3B3 SCARB2 THOC1 NDUKPNAFAF14 FAM120A 800

G TOMM70 M G Ery ME EIF3J RCL1

MAD2L2 SUPV3L1 ORCSPID2R MEP Magic Red MFI NDUFAF4 FUCA1 BYSL 500 KPNA1 TCOF1 SORCCDC88D A

r PABPC4 M on ORCSPID2R IFRD1 DCP2 MEP_G RAD23CTSB D PRMT3 CCDC88A Row Z−Score FAM120A UHRF1 SORD RCL1 TFB2M (Cathepsin B activity)

P P AIMP2 DCPP 2 POLSUMFQ 1 EryP RAD23B BYSL CDK2 MitoTracker Green MFI MitoTracker o UHRF1 WDR7SMG54 AIMP2 PABPC4 PPRC1 0 POLQ PRMT3 UBE2ESPIA1G1A CAD 600 WDR74 TFB2M SMG5 PUSFOXM1 1 TARDBP UBE2ESPI1 1 CDK2 TP5GNPT3 G PLK4 PPRC1 SLC19A1 Mono FOXM1 ERH PUS1 CAD TGFB1 NOC4SOL RT1 EIF3J TP53 TARDBP BAX SLC19A1 NDUFAF4 TGFB1 PLK4 UTP2MAPKAPK0 5 Magic Red SLC17A5 ORC2 NOC4BAX L ERH ACD RRP12 Gr SORD UTP2MAPKAPK0 5 EIF3J DDX11 NDUFAF4 HK2CTSF RAD23B RRP1ACD2 AUNIP ORC2 MYBBP1A AIMP2 HKDDX12 1 SORD CEBPDNASEG 2 WDR74 ✱✱ AUNIP RAD23B EIF4ATERF11 MYBBP1A UBE2E1 CEBPG AIMP2 U2AFESCONA1 2GPA PUS1 EIF4ATERF11 WDR74 CDT1 SLC19A1 UBE2E1 FBL NOC4L 30000 U2AFESCO1 2 PUS1 SLFN1ACP1 2 FARSA UTP20 FBCDTL 1 SLC19A1 TIPIN LAS1L RRP12 SLFN11 NOC4L MENGALN1 S G H LT-HSC FARSA HK2 TIPIN UTP20 HSPECHEK12 LT-HSC MYBBP1A LAS1MENL1 RRP12 DSCCMAN2B1 1 HK2 NOP2 EIF4A1 HSPECHEK12 NEK2 MYBBP1A MRGATO4 A U2AF1 ST-HSC NOPDSCC2 1 EIF4A1 BLM WDR43 FBL MNEKRTO24 U2AF1 RNF16CTS8O FARSA ST-HSC BLM FBL NCBPPARPB1 P LAS1L WDR43 RNF168 FARSA SRPKUSP71 HSPE1 LAS1L PLA2G15 20000 ✱✱ NCBP1 NOP2 ✱✱✱✱ PARPBP HSPE1 DDX1PRKC8Q SRPK1 MRTO4 USP7 NOP2 XPUSP3OMCOLNT 7 1 DDX1PRKC8Q MRTO4 PRKCD WDR43 30000 XPO1 NCBP1 USP37 WDR43 WRAP53 40000 XPOT CTSG SRPK1 PRKCD NCBP1 SRSFTERF22 XPO1 SRPK1 DDX18 WRAP53 DDX18 SERBPOGGARS11 A SRSF2 XPOT TERF2 XPOT DHX1POLH5 XPO1 SERBPOGG11 XPO1 DNALGM2 N SRSF2 LT-HSC SRSF2 EIF4H ST-HSC POLH WRN SERBP1 DHX15 SERBP1 DNA2 EIF3PRKDCTSB 2 Z DHX15 10000 EIF4H DHX15 30000 EIF4H WRN EIF4H HTICRDAC2R EIF3B SGSH EIF3B PRKD2 EIF3B MCMMAP3K4 4 20000 CellROX MFI HTICRDAC2R HDAC2 FAM168A HDAC2 STARDACP7 5 MCM4 MCMMAP3K4 4 MCM4 CTC1 STARD7 SYNCRIP STARD7 FAM168A STARD7 SYNCRIP NAFLAMP1 3 SYNCRIP CTC1 TCP1 SYNCRIP TCP1 PINX1 TCP1 NAF1 SET SEERCCTLAMP4 2 SET 20000 TCP1 PINX1 HNRNPD HNRNPCHTF1D8 HNRNPD VDAC1 SEERCCT 4 PIFC1LTA VDAC1 0 G3BP1 VDAC1 G3BP1 HNRNPCHTF1D8 ODC1 KLHL15 G3BPLAPTM41 A ODC1 VPIFDAC11 CCT7 E2F7 TfR1 MFI 10000 KLHL15 CCT2 ODCWIZ1 CCT7 % of Max G3BP1 E2F7 PSMA7 CCTHELA7TP6APB 1 CCT2 PSMA7 ODCWIZ1 CANX EREG 10000 HNRNPU CCT2 CANX CCTHEL7B LAPTM4B HSPD1 PSMAGZMA7 HNRNPU CCTERE2G EIF4G2 SLX4 CANPPTX 1 HSPD1 PSMAGZMA7 HNRNPA2B1 SIRT6 EIF4G2 SLX4 HSP90AB1 HNRNPKDM4DU HNRNPA2B1 CellROX Orange MFI CANX CD164 SIRT6 HSPDFGFR14 HSP90AB1 HNRNPU 1 2 KDM4D 1 2 3 EIF4GRTEL21 0 0 1 2 HSPDFGFR14 1 2 3 HSP90AB1 1 2 HNRNPA2B1 1 2 3 EIF4GRTEL21 HMGB1 HSP90AB1 HNRNHSP90ABPA2B11 WT HNRNPA2B1 HMGB1 TFEB LV + HSP90AA+ 1 HSP90ABHNRNPA2B1 1 Sort BFP cells Sort BFP cells 1 2 I 1 2 3 HSP90AA1 TfR1-BV786 HSC_GP91_ HSC_GP91_ HSC_GP91_ MPP_GP91_ MPP_GP91_ 5 5 1 2 3 1 2 10 3 10 1 2 1 2 3 HSC_GP91_ HSC_GP91_ HSC_GP91_ MPP_GP91_ MPP_GP91_ HSC_GP91_ HSC_GP91_ HSC_GP91_

CTRL LV MPP_GP91_ MPP_GP91_ 1 2 3 1 2 3 4 4 ST-HSC 10 10

3 3 Sorting 10 10 TFEBOE_ TFEBOE_ TFEBOE_ BFP BFP HSC_GP91_ HSC_GP91_ HSC_GP91_ MPP_GP91_ MPP_GP91_

TFEBOE_ TFEBOE_ TFEBOE_ 0 0 HSC_GP91_ HSC_GP91_ HSC_GP91_ MPP_GP91_ MPP_GP91_

0 50K 100K 150K 200K 250K TFEBOE_GP9_ 0 50K 100K 150K 200K 250K TFEBOE_GP91_ TFEBOE_GP91_ TFEBOE_GP9_ TFEBOE_GP91_ TFEBOE_GP91_ hCB 3d 6d 1 day Analysis Analysis CTRL pre-stimulation CTRL K L M TFEBWT J TFEBWT Transferrin endocytosis CTRL ✱ Lysosome ✱✱ 80 40000 ns 15000 150000 TFEBWT p=0.054 NES 2.47 NES 2.05 ✱ 30000 60 14000 WT TFEB TFEBWT 13000 20000 40 100000 ✱ BV786MFI % Cells 12000 CTRL CTRL - 10000 20 Magic Red MFI 11000 CD71-BV786 MFI TfR1 (Cathepsin B activity) 50000 0 0 10000 LysoTracker Deep Red MFI G0 G1 S/G2/M Figure 4 p=0.052 ✱✱✱✱ 4 ✱✱✱✱ A B RF BM SP CTRL ✱✱ ✱✱ CTRL LV TFEBS142A LV WT ✱✱✱✱ ✱✱✱✱ ✱✱ TFEB ✱✱✱✱ 3 ✱✱✱✱ ✱✱✱✱ S142A WT TFEB TFEB LV TFEBH240R-I243N LV 2 2 2 ✱✱✱✱ vs input)

+ H240R-I243N + - TFEB

CD34 CD38 cells cells 2 cells + + Sorting 0 0 + bioRxiv preprint doi: https://doi.org/10.1101/2021.02.24.432720; this version posted February 24, 2021. The copyright holder for0 this preprint /CD45 + (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprintFold change in perpetuity. It is made /CD45 /CD45 /CD45 + + available under aCC-BY-NC-ND 4.0 International -2license. -2 1 + hCB (BFP -2 vs BFP+ input) vs BFP+ input) 1 day vs BFP+ input) 1o engraftment 2o engraftment -4 -4 pre-stimulation 0 10K CD34+CD38- cells Limiting Dilution Assay LOG2 (BFP LOG2 (BFP 4 and 17 weeks LOG2 (BFP 8 weeks -6 -6 -4 C D CTRL vs. TFEBWT CTRL vs. TFEBS142A CTRL vs. TFEBH240R-I243N shTFEB-4 LV shCTRL LV

+ - engrafted CD34 CD38 - Sorting

x6.6 fold x22 fold x1.5 fold hCB 1 day 1o engraftment 2o engraftment p=0.0012 p=6.6E-06 p=0.56 + - pre-stimulation 10K CD34 CD38 cells Limiting Dilution Assay 4 and 17 weeks 8 weeks Log fraction #mice non

Dose (number of cells injected) E F G LysoH vs LysoL RF BM SP shCTRL vs. shTFEB-4 Secondary LDA LysoLow vs. LysoHigh

shCTRL 0.0 ✱✱✱✱ ✱✱✱ 1 2 shTFEB-3shCTRL 0.2 cells engrafted cells ✱✱✱✱ 1 − L - cells + + Lyso + engrafted shTFEB-4shTFEB-3 - 0 1 0.4 engrafted − 0 − /CD45

/CD45 shTFEB-4 /CD45 + + H -1 + Lyso 0.6

0 − -1 # mice non log fraction -2 x3.8 fold 0.8 -1 x-2.8 fold − p=0.009 -3 -2 p=0.03 vs mCherry+ input) vs mCherry+ input) vs mCherry+ input) 1.0 Log fraction #mice non -4 -2 -3 − Log fraction #mice non 0 50000 100000 150000 200000 250000 LOG2 (mCherry LOG2 (mCherry LOG2 (mCherry dose [number of cells injected] Dose (number of cells injected) Dose (number of cells injected) Figure 5