Article

Taurine from tumour niche drives glycolysis

to promote leukaemogenesis

[Link] Sonali Sharma1,2,14, Benjamin J. Rodems1,2,14, Cameron D. Baker3,14, Christina M. Kaszuba2,4,14,

Edgardo I. Franco2,4, Bradley R. Smith5, Takashi Ito6, Kyle Swovick7, Kevin Welle7,
Received: 11 September 2023
Yi Zhang8, Philip Rock8, Francisco A. Chaves2, Sina Ghaemmaghami7,9, Laura M. Calvi2,10,
Accepted: 14 April 2025 Archan Ganguly11, W. Richard Burack8, Michael W. Becker2,10, Jane L. Liesveld2,10,
Paul S. Brookes12, Joshua C. Munger2,5, Craig T. Jordan13, John M. Ashton1,2,3 & Jeevisha Bajaj1,2 ✉
Published online: xx xx xxxx

Open access
Signals from the microenvironment are known to be critical for development, stem
Check for updates
cell self-renewal and oncogenic progression. Although some niche-driven signals
that promote cancer progression have been identified1–5, concerted efforts to map
disease-relevant microenvironmental ligands of cancer stem cell receptors have
been lacking. Here, we use temporal single-cell RNA-sequencing (scRNA-seq) to
identify molecular cues from the bone marrow stromal niche that engage leukaemia
stem-enriched cells (LSCs) during oncogenic progression. We integrate these
data with our human LSC RNA-seq and in vivo CRISPR screen of LSC dependencies6
to identify LSC–niche interactions that are essential for leukaemogenesis. These
analyses identify the taurine–taurine transporter (TAUT) axis as a critical dependency
of aggressive myeloid leukaemias. We find that cysteine dioxygenase type 1 (CDO1)-
driven taurine biosynthesis is restricted to osteolineage cells, and increases during
myeloid disease progression. Blocking CDO1 expression in osteolineage cells impairs
LSC growth and improves survival outcomes. Using TAUT genetic loss-of-function
mouse models and patient-derived acute myeloid leukaemia (AML) cells, we show
that TAUT inhibition significantly impairs in vivo myeloid leukaemia progression.
Consistent with elevated TAUT expression in venetoclax-resistant AML, TAUT
inhibition synergizes with venetoclax to block the growth of primary human AML
cells. Mechanistically, our multiomic approaches indicate that the loss of taurine
uptake inhibits RAG-GTP dependent mTOR activation and downstream glycolysis.
Collectively, our work establishes the temporal landscape of stromal signals
during leukaemia progression and identifies taurine as a key regulator of myeloid
malignancies.

Signals from the tumour microenvironment (TME) can regulate ini- previous studies have described the cellular composition of normal
tiation, progression and immune evasion of tumours1–5,7–10. While bone marrow niche13–15, their dynamic alterations during leukaemia
scRNA-seq analysis has identified cellular TME components, especially progression remain undefined. We use scRNA-seq to establish tem-
in solid tumours11,12, concerted efforts to link ligands from the changing poral changes in bone marrow microenvironment populations, and
TME landscape with cognate receptors on cancer cells have been lack- in niche-driven signals during disease progression. To define TME
ing. As cell surface proteins are particularly amenable to therapeutic ligands that are essential for leukaemogenesis, we focused on cog-
targeting, functional characterization of their interactions with the nate cell surface receptors enriched in LSCs as compared to healthy
TME are of considerable clinical interest. controls, and those essential for in vivo leukaemia progression6. This
Aggressive therapy-resistant myeloid leukaemias, such as blast- approach identified signals that are known to be critical for cancer
crisis-phase chronic myeloid leukaemia (bcCML) and AML, initiate growth, such as KIT–KITL and CD47–thrombospondin 116, as well as
and expand in a complex bone marrow microenvironment. Although multiple new signalling axes.

1
Department of Biomedical Genetics, University of Rochester Medical Center, Rochester, NY, USA. 2Wilmot Cancer Institute, University of Rochester Medical Center, Rochester, NY, USA.
3
Genomics Research Center, University of Rochester Medical Center, Rochester, NY, USA. 4Department of Biomedical Engineering, University of Rochester, Rochester, NY, USA. 5Department of
Biochemistry and Biophysics, University of Rochester Medical Center, Rochester, NY, USA. 6Department of Bioscience and Technology, Graduate School of Bioscience and Technology, Fukui
Prefectural University, Fukui, Japan. 7Mass Spectrometry Resource Laboratory, University of Rochester, Rochester, NY, USA. 8Department of Pathology and Laboratory Medicine, University of
Rochester Medical Center, Rochester, NY, USA. 9Department of Biology, University of Rochester, Rochester, NY, USA. 10Department of Medicine, University of Rochester Medical Center, Rochester,
NY, USA. 11Department of Neuroscience, University of Rochester Medical Center, Rochester, NY, USA. 12Department of Anesthesiology, University of Rochester Medical Center, Rochester, NY, USA.
13
Division of Hematology, Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA. 14These authors contributed equally: Sonali Sharma, Benjamin J. Rodems,
Cameron D. Baker, Christina M. Kaszuba. ✉e-mail: Jeevisha_Bajaj@[Link]

Nature | [Link] | 1
Article
a b Arteriolar 1 Chondro 1 Fibro 1 MSC
c d 0.6

Proportion
Arteriolar 2 Chondro 2 Fibro 2 Osteo 1 Osteo 1
Arteriolar 3 Chondro 3 Fibro 3 Osteo 2 Naive Initiation Expansion End 0.4
MSC

associated cells
Sinusoidal 1 Chondro 4 Fibro 4 Osteo 3

MSC/osteo-
Leukaemia stem
/progenitors Sinusoidal 2 Chondro 5 Pericyte Osteo 4 Chondrocyte 0.2
10 Osteo 5 10 Osteo 4
Endothelial
Osteo 3
Time points: naive,
initiation, expansion, Osteo 2
5 5 Osteo
end Osteo 5
Enzymatic
0.6

UMAP 2
digestion
UMAP 2

0 0 Arteriolar 1

Proportion
Fibroblast 0.4 Arteriolar 2

Endothelial
–5 –5 Pericyte

cells
0.2 Sinusoidal 1
Leukaemia– Osteo
CD45–TER119– –10 –10 Arteriolar 3
Stromal cells Sinusoidal 2
MSC
scRNA-seq
–10 –5 0 5 10 –10 –5 0 5 10 Naive Initiation Expansion End
UMAP 1 UMAP 1

e Naive End 20 f Naive End 40

P = 0.0003 P < 0.0001
Gated on live leukaemia–CD45–TER119–CD31– Gated on live leukaemia–CD45–TER119–CD31–
CD31–CD140a+CD51+ cells (%)

CD31–SCA1–CD51+ cells (%)

P = 0.05 P = 0.012

Leukaemia–CD45–TER119–
Leukaemia–CD45–TER119–

5 5
10 10
15 30

104 2.37 14.3 10 4

CD140a PE-Cy5

10 20

SCA1 BV421
103 103
0 0 5.30 32.1 10
5
–103 –103

–103 0 103 104 105 –103 0 103 104 105 0

0
CD51 PE CD51 PE

ve

n
ve

d
io

io
d
io

io

En
ai
En
ai

at

ns
at

ns

N
N

iti

pa
iti

pa

In
In

Ex
Ex

g Top 10 pathways downregulated in MSCs Top 10 pathways downregulated in osteo-associated cells

Padj Padj
Extracellular structure organization ER to Golgi vesicle-mediated transport
External encapsulating structure organization 0.04 Collagen fibril organization 0.04
Extracellular matrix organization 0.03 Extracellular matrix organization 0.03
Regulation of osteoblast differentiation 0.02 Protein processing in ER 0.02
Regulation of cell migration Skeletal system development
Term

0.01 0.01
Positive regulation of cell differentiation Extracellular structure organization
Pathways in cancer Prop. of External encapsulating structure organization
Regulation of transcription by RNA Pol II Aerobic electron-transport chain Prop. of
expressed expressed
Regulation of bone mineralization genes in term Mitochondrial ATP synthesis
Focal adhesion Retrograde vesicle-mediated transport genes in term
0.10 0.1
20 40 60 0.15 10 20 30 40 0.2
0.20 0.3
0.4

Top 10 pathways downregulated in arteriolar endothelial cells Top 10 pathways downregulated in sinusoidal cells
Padj Padj
RAP1 signalling pathway Regulation of endothelial cell migration
Pathways in cancer 0.015 Focal adhesion 0.04
Axon guidance 0.010 Heart trabecula morphogenesis 0.03
Focal adhesion Positive regulation of cell differentiation 0.02
0.005 Receptor-mediated endocytosis 0.01
Proteoglycans in cancer
Term

Platelet activation Regulation of osteoblast differentiation

MAPK signalling pathway Prop. of Positive regulation of endothelial cell migration Prop. of
RAS signalling pathway expressed Cellular response to growth factor stimulus expressed
Adherens junction genes in term Regulation of endothelial cell proliferation genes in term
Leucocyte transendothelial migration 0.06 Platelet degranulation
0.08 0.1
10 20 30 4 8 12 16 0.2
0.10
Number of significantly expressed 0.12 Number of significantly expressed
genes matching the term genes matching the term

Fig. 1 | Temporal scRNA-seq analysis of myeloid leukaemia bone marrow cells (top) and endothelial cells (bottom) over time. e, Representative
microenvironment. a, The experimental strategy used to determine the impact fluorescence-activated cell sorting (FACS) plots and quantification of MSC
of bcCML progression on bone-marrow microenvironment remodelling. frequency over time. f, Representative FACS plots and quantification of the
b, A uniform manifold approximation and projection (UMAP) analysis of 15,695 osteolineage cell frequency over time. For e and f, data are mean%±%s.e.m. n%=%3
non-haematopoietic cells from bone and bone marrow shows 21 distinct bone animals per timepoint. Statistical analysis was performed using one-way
marrow stromal cell clusters (n%=%3 (naive), n%=%6 (initiation), n%=%7 (expansion) analysis of variance (ANOVA). g, Unbiased Enrichr analysis showing the top 10
and n%=%9 (end) mice). The colour key indicates subclusters. Chrondro, downregulated pathways by population cluster in MSCs, and osteo-associated,
chondrocyte; fibro, fibroblast; osteo, osteo-associated cell. c, UMAP plot of arteriolar and sinusoidal endothelial populations. Blue text represents pathways
major population clusters over time (naive, 0 days; initiation, 2 and 4 days; of interest. ER, endoplasmic reticulum. The mouse image in a is adapted from
expansion, 7 and 9 days; end, 11 and 14 days after transplant; the colours represent ref. 6, Springer Nature America.
different stages of disease). d, The proportion (prop.) of MSCs/osteolineage

Of these signals, TAUT, encoded by SLC6A6, was strongly associ-

ated with poor prognosis in human leukaemias and emerged as a key Temporal changes in leukaemia niche
regulator of AML. As taurine can be neuroprotective, mitigate the To define temporal changes in non-immune bone marrow micro-
side-effects of chemotherapy17 or support anti-cancer immunity18, a environment populations during myeloid leukaemia progression,
cancer-promoting role of taurine has not been considered. We exam- we used bcCML LSCs, which can grow in unirradiated mouse recipi-
ined if blocking taurine production in the TME impairs LSC function. ents (Extended Data Fig. 1a,b). Bone marrow stromal populations
We used genetic tools to establish whether TAUT expression in cancer (leukaemia−CD45−TER119−) were isolated from leukaemic cohorts at
cells controls the growth of aggressive myeloid leukaemias. Using distinct stages of disease progression and analysed using scRNA-seq
metabolomic, proteomic and transcriptomic approaches, we iden- (Fig. 1a). Gene-expression-based clustering identified 21 lineage clus-
tify downstream mechanisms by which taurine regulates leukaemia ters with transitional subsets covering endothelial cells, chondro-
growth. cytes, fibroblasts, pericytes, mesenchymal stromal cells (MSCs) and

2 | Nature | [Link]
osteo-associated populations (Fig. 1b and Extended Data Fig. 1c,d) and TME-derived ligands of LDLR and SLC6A6 on LSC growth. One of the
showed significant remodelling with disease progression (Fig. 1c,d). primary LDLR ligands, apolipoprotein E (APOE)21, was highly expressed
Our flow-cytometry-based analyses of an independent cohort of leu- in MSCs (Extended Data Fig. 3d). To test the role of APOE from MSCs on
kaemic mice validated these changes, showing increases in MSCs, osteo- LSC function, we co-cultured bcCML LSCs with MSCs transduced with
lineage and arteriolar endothelial cells, with a concomitant decline in short hairpin RNAs (shRNAs) targeting Apoe or LacZ (control; Extended
sinusoidal endothelial populations (Fig. 1e,f and Extended Data Fig. 1e). Data Fig. 4c). Co-culture with MSCs lacking Apoe expression led to
Clustering analyses of genes expressed in major lineages identified significant reduction in both the viability and colony-forming ability of
gene sets with similar changes in expression (Extended Data Fig. 1f), LSCs (around 2.1-fold lower than the controls; Extended Data Fig. 4d–f).
suggesting that these may represent altered stromal cell fate. Pathway SLC6A6 encodes TAUT, which has high affinity for the non-essential
analysis showed that osteoblast differentiation and bone mineralization amino acid taurine, and low affinity for &-alanine (taurine Km%=%4.5%µM
were downregulated in MSCs and osteo-associated cells, and focal adhe- versus &-alanine Km%=%56%µM)22. Our experiments indicate that Slc6a6
sion and cell migration were downregulated in endothelial populations, expression can be directly induced by oncogenes (Extended Data
indicating that MSC and endothelial function may be impaired with Fig. 4g). While the leukaemia TME did not express enzymes required
disease progression (Fig. 1g and Extended Data Fig. 2a). Collectively, for &-alanine synthesis (Gadl1 and Cndp1), those needed for taurine
our temporal analysis of the TME identifies dynamic changes in gene biosynthesis were expressed in osteo-associated cells (Cdo1 and Csad;
expression of stromal subpopulations during leukaemia progression. Extended Data Fig. 4h,i). Our analysis of publicly available human bone
marrow stromal cell scRNA-seq data using surgical samples from osteo-
arthritis patients23 showed that CDO1 expression is restricted to MSC
Ligand–receptor interactome and osteolineage populations (Extended Data Fig. 4j). To test whether
To define the functional relevance of microenvironmental changes on CDO1 is expressed in the human leukaemia TME, we performed an
leukaemogenesis, we focused on proteins expressed on LSC cell surface scRNA-seq analysis of CD45− stromal cells from three myelodysplastic
that act as receptors for niche-driven signals. To identify LSC receptors syndrome (MDS) and AML bone marrow aspirates. Our analysis of the
associated with disease progression, we performed RNA-seq analysis of limited numbers of stromal cells in these samples indicate that CDO1
human AML and bcCML CD34+ cells, and healthy donor bone marrow expression is restricted to the MSC/osteolineage cells (Extended Data
CD34+ haematopoietic stem and progenitor cells (HSPCs; Extended Fig. 4k,l), as we see in mouse samples. Notably, we found a marked
Data Fig. 2b). We noted significant (adjusted P (Padj)%<%0.05) upregulation increase in CDO1 protein expression in nearly all matched biopsies of
of 1,569 genes in bcCML, 2,842 genes in AML and 2,331 genes in both patients who progressed from MDS to secondary AML, and those who
bcCML and AML compared with normal HSPCs. To identify signals relapsed after AML diagnosis (Fig. 2d–g and Extended Data Fig. 4m–p).
that are functionally relevant for disease progression, we found cell While CDO1 protein was expressed in undifferentiated human AML
surface genes19 that drop out by twofold or more in our genome-wide MSCs, its expression increased during in vitro mouse and human AML
in vivo leukaemia CRISPR screen6. This identified 13 cell surface signals MSC osteogenic differentiation (Extended Data Fig. 5a–c). Thus, CDO1
common to both bcCML and AML, 18 unique to AML and 7 unique to may be broadly expressed in immature MSCs as well as differentiating
bcCML. Of these 38 genes, 16 were misannotated as cell surface in the osteolineage cells in human bone marrow. Importantly, taurine levels in
reference dataset19 and were removed from further analysis (Methods extracellular medium increased during MSC differentiation, indicating
and Extended Data Fig. 2c,d). We used NicheNet and the published that elevated CDO1 expression is correlated with taurine production
literature to identify ligands for these receptors, especially those that (Extended Data Fig. 5d).
were significantly upregulated in our TME scRNA-seq data and in the To determine a functional role of taurine from osteolineage cells,
human AML immune microenvironment20 (Extended Data Fig. 2e). we tested the impact of inhibiting taurine biosynthesis in mouse and
This led to the exclusion of receptors with no known ligands (MR1, human leukaemia bone-marrow-derived MSCs on co-cultured leu-
TMCO3 and TSPAN15) as well as those of which the ligands were not kaemia cells. Our experiments showed that mouse LSCs co-cultured
significantly enriched in any TME population (LGALS3BP, CD96, CD274 with osteolineage cells lacking Cdo1 expression were significantly less
and CD3D). We therefore generated a unique map of TME ligands for viable (about 2-fold lower than the controls) and formed fewer colonies
15 LSC receptors that are essential for disease progression (Fig. 2a). (3.4-fold less than controls), which could be rescued by supplementing
We next determined whether disease progression could alter the cultures with taurine (Extended Data Fig. 5e–g). Similarly, inhibiting
expression of LSC-specific TME ligands, and identified four distinct CDO1 in osteolineage cultures from MSCs derived from patients with
patterns of ligand expression. While the expression of some genes AML impaired survival and colony-formation of co-cultured AML cells
remained steady, for example, Jam2 in arteriolar endothelial cells and from the same patient by 2- to 3.2-fold (Fig. 2h,i and Extended Data
Ihh in chondrocytes, the expression of Vcam1, Pcdh7 and Il7 was lost in Fig. 5h), indicating that taurine produced by osteolineage cells pro-
the TME, especially at the end point. Expression of Anxa1 first increased motes LSC growth and survival.
but then declined towards the end. The fourth category of genes was Consistent with a functional role of taurine from the TME in leu-
always expressed, but the microenvironmental populations expressing kaemia progression, taurine levels were 1.7-fold higher in leukaemic
them changed. These included Pvr, Cdo1, Apoe and Agt (Extended Data bone marrow interstitial fluid as compared to the controls (Fig. 2j). In
Fig. 3a–d). Our analysis capturing dynamic changes in the TME identi- mice, the majority of taurine is synthesized in the liver from cysteine24.
fies signals that may be necessary not only at different stages of disease To determine whether taurine produced locally in the bone marrow
progression, but also those critical during the entire course of disease. niche is essential for leukaemia growth, we generated a new Cdo1fl/fl
mouse model and crossed it to MSC/osteolineage-specific Prrx1-cre
mice (Extended Data Fig. 5i–k). We used these as recipients for LSCs and
TME signals support leukaemogenesis monitored survival, and the impact on bone marrow microenvironment
Multiple TME–LSC signalling axes identified in our interactome may populations. Our experiments showed that Cdo1fl/flPrrx1-cre+ mice lived
have a functional role in leukaemia progression. However, we were around 13.5% longer than the controls, indicating that taurine produced
interested in genes associated with unfavourable outcomes in human locally in the leukaemia TME can support disease progression, at least in
patients with AML. Of the 22 genes upregulated in human LSCs, only part (Fig. 2k). We did not find any significant changes in the composition
low-density lipoprotein receptor (LDLR) and SLC6A6 were signifi- of TME in Cdo1fl/flPrrx1-cre+ leukaemic mice as compared to the controls
cantly associated with poor prognosis in AML (Fig. 2a–c and Extended (Extended Data Fig. 5l–n), suggesting that leukaemia-driven remodel-
Data Fig. 4a,b). We therefore tested the functional requirement of ling of the bone marrow niche does not depend on taurine produced

Nature | [Link] | 3
Article
a b c P = 0.002
100 P = 0.0087
<11.14 (n = 80) 1,400
>11.14 (n = 81)
Endothelial 80
1,200

Overall survival (%)

Hmgb1
s
yte

Icam1
Fib

SLC6A6 expression
Efnb2
Efnb1
roc

Efna1

Itgb1
F11r

Jam2
Jag2
rob 1,000

Calm1

Vca
Tgfb 1
Edn1
60

Nrp11
d
on las
Ch

Cl xa1
T hb
Arf1
t

Lga Ihh

Co stn1
800

L mp
Th d1 3
An
Hm mp

s1

Pk gals
ls3
40
Co o1
gb

1
Th bs
Ag bs2
600
Cd b 1
ed
ci o-

2
so te
at

Pc
Tg

1
as Os

F dh xa 20
f

t
400
Cd at1 7 An poe P = 0.0036
Cd o1 A alr

MS
Cd h2 C dh2 5 HR = 1.80 (95% CI = 1.23–2.68) 200
h1 0

C
C fna
Veg 5 E t1 b1 0 500 1,000 1,500 2,000 2,500 3,000
Tgf fb Fa sp90
H

L
Lrp b1 Time (days)

al

AM
Il7 b1

C
ap1 Itg

or

bc
Hsp

N
90b Kit b1
1 Lam p1
Hm
gb2 Lrpa d MDS sAML f AML-D AML-R
Nrp1
Immune

Calm1 Trf

Pateint 1

Pateint 1
Vcam1
Arf1
Anxa1
Anxa1 Clstn1

Pericyte
Aimp1 Itgb1
Nrp1
Adsl

Pateint 2

Pateint 2
Kitl
a3a
Sem Cd4 CDO1
g 7 CDO1
Il2r
c2
Ds
6 Ltb e 10 P = 0.049 g 15 P = 0.005
6a 4r1
Slc

L
Ep

CDO1+ BM cells in 0.5 mm2 (%)

CDO1+ BM cells in 0.5 mm2 (%)
AM
1r ha
F1 1 8
Pk
d2

bc 10
CM
La 6
Ld

nd
Fas

lr

AM
Fat

L 4
Cdon

L
1
Nt5e

CM 5
bc
2

0 0

D
DS

R
M

L-

L-
sA
M

AM

AM
h i j k Cdo1 deletion in MSC/osteolineage cells
Primary human AML P = 0.0002 P = 0.001 0.01 P = 0.002
BM aspirate-derived 25 1.5
MSC cultures WT Cdo1fl/fl/Cdo1+/+
Lin– leukaemia 100
shLacZ/ 0.008 Cdo1fl/flPrrx1-cre+
interstitial fluid (mg ml–1)

20
Relative no. of colonies

cells
No. of live cells (×104)

shCDO1
1.0 Overall survival (%) 80
Taurine in BM

15 0.006 Cdo1fl/flPrrx1-cre
or
Osteogenic Cdo1fl/fl/Cdo1+/+ 60
differentiation 6 days 0.004
10
0.5
Matched patient 40
AML cells
5 0.002
72 h Survival/ 20
BM microenvironment P = 0.0002
0 0 0 analysis HR = 0.1718
Leukaemia
0
ia

cell viability/
uk al
1

1
Z

m
DO

DO

m
C

ae

CFU assays 0 12 16 20 24
LA

LA

or
C

N
sh

sh
sh

sh

Le

Time (days)

Fig. 2 | Bone marrow microenvironmental ligands for LSC-specific cell on co-cultured AML cells (MSCs and AML cells were derived from the same
surface receptors. a, Circos plot showing leukaemia cell surface receptors and patient). CFU, colony-forming unit. i, The number of live LSCs (left; data are
cognate stromal-cell-derived ligands. b, Kaplan–Meier curves of human patients mean%±%s.e.m.; n%=%11 independent culture wells per cohort; data were combined
with leukaemia with low (<11.14, n%=%80) or high (≥11.14, n%=%81) SLC6A6 expression from two independent experiments) and their colony-forming ability (right;
(TCGA-LAML; Xena Browser; log-rank test). CI, confidence interval; HR, hazard data are mean%±%s.d.; n%=%3 independent culture wells per cohort) after coculture
ratio. c, Normalized SLC6A6 expression in CD43+ cells from normal human bone with AML MSCs. j, Taurine quantity per femur in control and leukaemic mice,
marrow (BM) samples and samples from patients with bcCML and AML. n%=%7 as determined using colourimetric analysis, 12 days after transplant. Data
(bone marrow), n%=%10 (bcCML) and n%=%11 (AML). For the box plots, the centre line are mean%±%s.e.m. n%=%5 animals per cohort. Data were combined from two
shows the median, the box limits show the interquartile range and the whiskers independent experiments. For i and j, statistical analysis was performed using
represent the minimum and maximum values, respectively. Statistical analysis unpaired two-tailed Student’s t-tests. k, Experimental strategy and survival curve,
was performed using DESeq2-implemented Wald tests. d–g, Representative showing the impact of blocking taurine production by MSC/osteolineage
IHC images (d,f) and quantification (e,g) of CDO1 expression in matched populations in vivo on bcCML progression in unirradiated recipients. n%=%18
patient bone marrow biopsies at MDS diagnosis and after AML transformation (Cdo1 fl/fl /Cdo1+/+) and n%=%14 (Cdo1 fl/flPrrx1-cre +). Data were combined from four
(d,e), or at AML diagnosis (AML-D) and relapse (AML-R) (f,g). n%=%5 independent independent experiments. Statistical analysis was performed using the log-rank
patients per cohort. Each colour represents a patient sample. Statistical test. WT, wild type. Scale bars, 50 µm (d and f). The mouse images in h and k are
analysis was performed using two-tailed ratio paired t-tests. h, The strategy adapted from ref. 6, Springer Nature America.
used to determine the impact of inhibiting CDO1 in human bone marrow MSCs

by osteolineage cells. We finally tested whether exogenous taurine disease progression in immunocompetent mice (around threefold
supplements can promote leukaemia growth. Our experiments showed higher likelihood of death relative to the controls; Extended Data
that the colony-forming ability of mouse LSCs and patient-derived AML Fig. 5r), indicating that taurine can promote leukaemic progression.
cells increased by 1.2- to 3.3-fold in the presence of taurine (Extended Collectively, these data indicate a key role of taurine and APOE from
Data Fig. 5o–q). Taurine supplements could significantly accelerate the bone marrow niche in sustaining LSCs. As SLC6A6 expression was

4 | Nature | [Link]
 a b c Slc6a6+/+ or
d 100 e f 100
g
1.5 1.5 100 40
P < 0.0001 Slc6a6–/– P = 0.0002 P = 0.007
+/+ +/+
Relative Slc6a6 expression

KLS cells 80 –/– 80 80 –/–

Relative taurine levels

30
(normalized to B2m)

Overall survival (%)

Overall survival (%)
BCR/ABL

No. of colonies
No. of colonies
1.0 1.0 NUP98/HOXA9
Sorted 500 60 60 60
BA+NH+ 48 h 20
cells 6 Gy
40 40 40
B6 Primary
0.5 0.5 bcCML
10
Sorted Lin– Cells 20 20 20
Colony
assay 6 Gy P < 0.01 P < 0.0001
B6 HR = 0.2540 HR = 0.0473
0 0 Secondary 0 0 0 0
bcCML 0 20 40 60 +/+ –/– 0 20 40 60 80 +/+ –/–
+/+ –/– +/+ –/–
Time (days) Time (days)

h i j k l m n
1.5 100 +/+ 1.5 30 Slc6a6+/+ or 100 +/+
Relative taurine levels in bcCML cells

Slc6a6+/+ or
P < 0.0001 Slc6a6 –/– –/– P < 0.0001 P = 0.002 Slc6a6 –/– –/–
KLS Cells 80 KLS Cells 80

Relative taurine levels

Overall survival (%)

Overall survival (%)

1.0 MLL-AF9 AML-ETO9a

No. of colonies
1.0 20
NRAS(G12V) 60 NRAS(G12V) 60
Equal no. of Equal no. of
48 h 48 h AE9+RAS+cells
MA9+RAS+ cells
6 Gy 40 9.5 Gy 40
0.5 B6 0.5 10 B6
20 20
P < 0.0001 P < 0.0001
Survival HR = 0.1645 Survival HR = 0.0276
0 0 0 0 0
+/+ –/– 0 20 40 60 +/+ –/– +/+ –/– 0 20 40 60
Time (days) Time (days)

o Gated on live Lin–CD150– Leukaemia stem cells

p q
leukaemia Gated on Lin– leukaemia Early Late apoptosis/ Gated on leukaemia
BM Spleen +/+ Necrosis apoptosis necrosis +/+ S 60.6
+/+ 1.11
P = 0.003 P = 0.0004 P = 0.05 P = 0.028 P = 0.005 P = 0.02
Lin–CD150–FLT3+SCA1high LSCs (%)

Percentage of annexin+7-AAD+ cells

Percentage of annexin 7-AAD cells

14 12 5 2.5 80 +/+
12 –/–

Leukaemic population (%)

–

10
Early 4 2.0
10.3 10 60
8 apoptosis G0–G1 G2–M
+

8 0.92 3 1.5 33.9 5.22

6 105 –/– 40
105 –/– 6
Necrosis 105 –/– S 72.9
1.52 2 1.0
104 4 104 4
4 10
20
103 2 103 1 0.5
2

BrdU APC
103
FLT3 PE

102 102 Early

1.92 0 0
7AAD

0 0 0 0
apoptosis G0–G1 G2–M
101 +/+ –/– +/+ –/– 101 +/+ –/– +/+ –/– G0–G1 S G2–M
1.11 –103 23.0 3.33
–103 0 103 104 105 101 102 103 104 105 0 100 200
SCA1 APC Annexin PE-Cy7 7AAD (×103)

Fig. 3 | TAUT loss impairs myeloid leukaemia initiation and propagation in independent experiments. k, Taurine in KIT+ AML cells. Data are mean%±%s.e.m.
mouse models. a, Relative Slc6a6 mRNA expression in whole bone marrow n%=%4 animals per cohort. l, CFU analysis of KIT+ AML cells. Data are mean%±%s.d.
cells from Slc6a6+/+ (+/+) and Slc6a6−/− (−/−) mice. Data are mean%±%s.d. n%=%3 n%=%3 culture wells per cohort. m,n, Experimental strategy (m) and survival
replicates per cohort. b, Taurine in normal bone marrow cells. Data are mean%±% curve (n), showing the impact of TAUT loss on de novo AML-ETO9a-driven AML.
s.e.m. n%=%6 pelvic bones from three animals per cohort; data were combined n%=%7 (+/+) and n%=%10 (−/−). Data were combined from two independent experiments.
from two independent experiments. c,d, The experimental strategy (c) and o, Representative FACS plots and quantification of the Lin−CD150 −SCA1+FLT3+
primary bcCML survival curve (d). n%=%12 (+/+) and n%=%11 (−/−). Data were combined bcCML stem cell frequency in the bone marrow (left) and spleen (right) of
from three independent experiments. e, CFU analysis of Lin− bcCML cells recipients. Data are mean%±%s.e.m. n%=%11 animals per cohort. Data were combined
from primary transplants. Data are mean%±%s.d. n%=%3 culture wells per cohort. from three independent experiments. p, Representative FACS plots and
f, Survival curve of secondary bcCML transplants. n%=%11 (+/+) and n%=%10 (−/−); quantification of early apoptosis and necrosis in bcCML at 14 days after transplant.
data were combined from two independent experiments. g, CFU of Lin− cells from Data are mean%±%s.e.m. n%=%8 animals per cohort. Data were combined from two
secondary transplants. Data are mean%±%s.d. n%=%3 independent culture wells independent experiments. q, Representative FACS plots and graph showing
per cohort. h, Taurine in Lin− LSCs. Data are mean%±%s.e.m. n%=%8 independent the frequency of in vivo BrdU incorporation in bcCML. Data are mean%±%s.e.m.
replicates per cohort from two independent samples. i,j, The experimental n%=%3 animals per cohort. Statistical analysis was performed using unpaired
strategy (i) and survival curve ( j) show the impact of TAUT loss on de novo two-tailed Student’s t-tests (b, e, g, h, k, l and o–q) and log-rank tests (d, f, j and n).
MLL-AF9-driven AML. n%=%17 (+/+) and n%=%20 (−/−); data were combined from four The mouse images in c, i and m are adapted from ref. 6, Springer Nature America.

significantly enriched in both human bcCML and AML LSCs as com- produced by the niche. TAUT loss impaired initiation of bcCML in mouse
pared to the controls, we focused our studies on SLC6A6, as it may be models (3.9-fold higher likelihood of survival; Fig. 3c,d), which could
broadly required for growth of aggressive leukaemias. not be rescued with taurine supplements (Extended Data Fig. 6b). TAUT
loss also led to functional depletion in bcCML LSCs, as indicated by a
2.3-fold reduction in their colony-forming ability (Fig. 3e), as well as
TAUT loss impairs leukaemogenesis a marked increase in the survival of mice transplanted with Slc6a6−/−
To determine whether TAUT has a functional role in leukaemia pro- LSCs (40%) relative to control LSCs (0%) in secondary transplant assays
gression we used global Slc6a6-knockout mice25 (Fig. 3a,b). While (21.2-fold higher likelihood of survival; Fig. 3f). The 8.5-fold reduction
Slc6a6-knockout mice are born at Mendelian ratios, they can develop in colony-forming ability of bcCML LSCs from secondary transplants
ageing-related defects in bone mass and retinal degeneration25–27. Our suggests that these leukaemias remain dependent on TAUT expression
experiments showed that TAUT loss in LSCs significantly reduced their and taurine uptake for their continued propagation (Fig. 3g,h). We
ability to grow when co-cultured with osteolineage cells express- noted no significant differences in bone marrow microenvironmental
ing either control shRNA or shRNA against Cdo1 (shCdo1; Extended populations of mice bearing Slc6a6−/− leukaemias as compared to the
Data Fig. 6a), indicating that they could not respond to taurine being controls, indicating that leukaemia-driven niche remodelling possibly

Nature | [Link] | 5
Article
reflects the extent of the tumour burden, and may be independent of structural analogues of taurine that inhibit uptake: 6-aminomethyl-
taurine levels within leukaemia cells (Extended Data Fig. 6c–g). 3-methyl-4H-1,2,4-benzothiadiazine-1,1-dioxide hydrochloride35
To determine whether TAUT is broadly required for de novo AML (TAG) and guanidinoethyl sulphonate36 (GES; Extended Data Fig. 8d,e).
growth, we tested the impact of its loss on initiation of leukaemia TAG and GES treatment led to a substantial reduction in colonies formed
driven by MLL-AF9/NRAS(G12V) and by AML-ETO9a/NRAS(G12V). Our by Slc6a6+/+ mouse LSCs but not Slc6a6−/− cells (Extended Data Fig. 8f–h).
experiments showed that TAUT loss markedly delays the initiation of Importantly, while TAG and GES impaired growth of primary human
MLL-driven AML relative to the control (6.1-fold higher likelihood of AML cells in colony assays by 1.8- to 20-fold, they did not impact normal
survival; Fig. 3i–k). Slc6a6−/− KIT+ AML LSCs from established disease human CD34+ HSPC colony growth (Fig. 4d,e). We next tested whether
formed 2.9-fold fewer colonies compared with the controls (Fig. 3l), venetoclax can exacerbate the impact of TAUT loss and/or inhibition
indicating that TAUT loss depleted functional LSCs. Consistent with on LSC function. Our experiments showed that venetoclax reduced the
a key role of TAUT expression in myeloid leukaemia initiation, TAUT viability of mouse Slc6a6−/− LSCs by 3.2-fold and lowered their colony
loss in disease driven by AML-ETO9a resulted in a marked increase in formation by 13.2-fold as compared to the controls (Extended Data
survival (70%) relative to the controls (0%) (36.3-fold higher likelihood Fig. 8i,j). While TAG, GES or venetoclax alone could impair mouse LSC
of survival; Fig. 3m,n). At the cellular level, TAUT loss led to a 2.4- to colony formation by 1.4- to 3.9-fold, their combination was synergistic
3.6-fold reduction in primitive Lin−CD150−FLT3+SCA1+ LSCs6 (Fig. 3o). and led to 2.3- to 7.2-fold fewer colonies as compared to the controls
Furthermore, TAUT loss increased necrosis and the Lin+ differentiated (Extended Data Fig. 8k–m). While primary human AML cells that were
cells (Fig. 3p and Extended Data Fig. 6h) and promoted cell prolifera- treated with GES, TAG or venetoclax alone formed 1.3- to 8.3-fold
tion (Fig. 3q). These data demonstrate a critical requirement for TAUT fewer colonies as compared to the controls, combining the treatments
in the initiation, self-renewal and propagation of myeloid leukaemia. substantially impaired colony formation by 2.4- to 150-fold. These
We next tested the impact of TAUT loss on normal haematopoiesis. data indicate that taurine inhibitors can synergize with venetoclax in
Our analysis indicates that TAUT loss does not severely impact bone blocking the growth of human AML cells (Fig. 4f–h).
marrow cellularity, or the total numbers of HSCs, progenitors and As TAG and GES did not effectively block taurine uptake in vivo
differentiated cells at steady-state (Extended Data Fig. 7a–e). While (Extended Data Fig. 9a–j), we used shRNA-based approaches to deter-
TAUT loss did not impair initial HSC engraftment (1 month after trans- mine the impact of inhibiting TAUT expression on human AML growth.
plant), donor chimerism dropped over time (Extended Data Fig. 7f,g). Knocking down SLC6A6 expression using two independent shRNAs
Although overall bone marrow chimerism 4%months after transplant significantly impaired taurine uptake by 2.2 to 3.4-fold, as well as the
was 2.4-fold lower in Slc6a6−/− HSC recipients as compared to Slc6a6+/+ colony-forming ability of human bcCML, AML and MDS cell lines by
HSC recipients, these Slc6a6−/− HSCs were able to contribute to all hae- 2- to 12-fold (Extended Data Fig. 9k–p). Our experiments showed that
matopoietic lineages (Extended Data Fig. 7h–k). Serial transplantation SLC6A6 knockdown in samples from patients with AML reduced their
of Slc6a6+/+ and Slc6a6−/− bone marrow cells showed a similar loss in colony-forming ability by 2.33- to 9.12-fold as compared to the con-
Slc6a6−/− engraftment over time (Extended Data Fig. 7l–q). These results trols (Fig. 4i). By contrast, SLC6A6 knockdown did not impact colonies
suggest that, while TAUT loss does not impair steady-state haemat- formed by normal human CD34+ HSPCs (Fig. 4j). Importantly, while
opoiesis, it can impact long-term HSC self-renewal and maintenance. SLC6A6 knockdown reduced the engraftment of primary AML cells
These data are consistent with previous studies showing that genetic by 1.2- to 40-fold in patient-derived xenograft models (Fig. 4k,l), it did
loss of LSC-enriched genes such as CD981, STAU26, MSI228,29, BRD430 and not significantly impair engraftment of normal human CD34+ HSPCs
BCL231 can impair HSC self-renewal. However, therapeutic inhibition of in xenograft models (Fig. 4m).
CD98 showed minimal toxicity in AML phase I trials1 and BCL2 inhibitors Collectively, our data identify a critical requirement for SLC6A6
are approved for AML therapy. It is therefore possible that non-genetic expression in primary human AML growth and indicate that blocking
approaches using small-molecule inhibitors or gene silencing may taurine transport may be of value in aggressive myeloid leukaemias.
identify a therapeutic window for TAUT targeting in human cells. We
therefore tested the impact of TAUT inhibition on growth and prolif-
eration of normal human HSPCs as well as patient-derived AML cells. Taurine drives glycolysis in leukaemia
To establish the mechanisms by which taurine uptake in leukae-
mia cells promotes disease progression, we determined metabolic
TAUT is essential for human AML growth changes in the absence of taurine. Our untargeted metabolomic
Our analyses of available gene expression datasets indicate that SLC6A6 analysis of mouse LSCs identified significant downregulation of
expression is enriched in leukaemia stem/progenitors as compared to glycolysis/TCA related pathways and metabolites such as pyruvate,
more mature blasts (Extended Data Fig. 8a). While SLC6A6 is broadly glyceraldehyde-3-phosphate and 3-phosphoglycerate in Slc6a6−/− cells,
expressed in AML irrespective of karyotype or FAB subtype (TCGA; suggesting that taurine may regulate energy metabolism (Fig. 5a–c
Extended Data Fig. 8b), increased expression correlates with venetoclax and Extended Data Fig. 10a). Consistent with this, we noted a 1.3- to
resistance (BEAT-AML; Fig. 4a). SLC6A6 expression is also enriched in 1.8-fold reduction in basal glycolysis, glycolytic capacity, maximal
reactive oxygen species (ROS)-low LSCs from human monocytic AMLs oxygen consumption and spare respiratory capacity in Slc6a6−/− LSCs
(subtype clinically associated with venetoclax resistance; CD45bright as compared to the controls (Fig. 5d,e and Extended Data Fig. 10b). To
SSChighCD117−CD11b+CD68+) compared with primitive AML (Fig. 4b; functionally test the role of glycolysis and TCA associated metabolites,
CD45medSSClowCD117+CD11b−CD68−)32. We find increased SLC6A6 we determined whether bypassing these could rescue Slc6a6−/− defects.
expression in leukaemia cells carrying RAS mutations as compared Our experiments showed that the colony-forming ability of Slc6a6−/−
to wild-type cells, consistent with data correlating RAS mutations with LSCs could be significantly rescued by pyruvate, sodium acetate and
venetoclax resistance33 (Fig. 4c). SLC6A6 is also highly expressed in lactate, but not glucagon (Fig. 5f and Extended Data Fig. 10c–e). These
relapsed AML originating from stem/progenitor like cells compared data suggest that glycolysis, and not gluconeogenesis, is the primary
with more committed populations34 (Extended Data Fig. 8c). These downstream effector of taurine in leukaemia cells.
data collectively indicate that inhibiting SLC6A6 may be of value across We next tested whether taurine contributed to any cellular metabolite
AML subtypes, including those commonly associated with venetoclax by determining 13C-taurine incorporation in K562 leukaemia cells. Our
resistance. untargeted metabolomic approach identified 13C label only in known
To determine whether small-molecule inhibitors of TAUT can taurine conjugates, N-acetyl taurine and glutaurine (Extended Data
effectively block leukaemia growth, we used two well-characterized Fig. 10f,g). Membrane-permeable taurine fully rescued colonies formed

6 | Nature | [Link]
 a b c Padj < 8.41 × 10–104
d e
TAG GES TAG GES
2,500 4 1.5 1.5 2.0 1.5
P < 0.0001 P = 0.011 P < 0.0001 P < 0.0001
16 P = 0.003 P < 0.0001 P < 0.0001

SLC6A6 expression level

Variance-stabilized data

Relative no. of colonies

Relative no. of colonies

Relative no. of colonies

Relative no. of colonies
Quantiles by 2,000 P < 0.0001 1.5

Normalized SLC6A6
15 Ven response 3
1.0 1.0 1.0

expression
1
14 3 1,500
2 1.0
2
13
4 0.5 0.5 0.5
1,000
1 0.5
12

500
11 0 0 0 0.0
0
1 2 3 4 0.1 0.5 1.0 0.1 1.0 0.1 0.5 1.0 0.1 1.0 5.0

er
D+

er
SO

SO
C e

C c

at
LS iv

LS ti

12
GES (mM)

at
TAG (μM) GES (mM)

W
TAG (μM)

s
L ocy
L imit

W
DM
G

DM

W
Ven response

S
AM on
AM Pr

RA
M

N
f P < 0.0001 g P < 0.0001 h
1.5 1.5 2.0 0.1 μM Ven + 0.1 mM GES
P < 0.0001 P < 0.0001
0.1 μM Ven + 1 mM GES
P < 0.0001 P < 0.0001
1.0 0.5 μM Ven + 0.1 mM GES
P < 0.0001 P < 0.0001
Relative no. of colonies
Relative no. of colonies

1.5 0.5 μM Ven + 1 mM GES

combination Index

Venetoclax (D/Dx)
1.0 1.0

Ven + GES
Antagonism
1.0
0.5
0.5 0.5
Synergism

Ad
0.5

d
iti
ve
ef
fe
c
0 0

t
0 0
0 0.5 1.0 0 0.5 1.0
μM en
μM ter

m Ven

0 M S
0. Ve + M S
0 M S
n mM ES
m ES
ES
μM μM Ven

μM μ Ven

n
SO

0. 0 G + G

G G

G AG

0. μM + m GE
n m GE
E
μM μ Ve

Ve

V
TA TA

TA TA

n 1m G
G
G
G

Faction affected GES (D/Dx)

TA M T

w
DM

1 1 +

M
M

+
1
SO

5
0.
0.
1 1

5 .5

.1

n 1
μM Ve .1

Ve .1
1
0. 0.

DM

+
μM +
0. Ve
μM
1

5
1

5
0.

i j k l m
Patient-derived shLacZ P = 0.037 shLacZ
P < 0.0001 49.7 2.0 60
1.5 1.5 AML cells
19.3
or 40
normal CD34+ shLacZ-GFP 6

Relative bone marrow

Relative no. of colonies
Relative no. of colonies

Relative bone marrow

HSPCs shSLC6A6-GFP 1.5 4

engraftment (%)
engraftment (%)
1.0 1.0 3

1.0
Sorted 2
24 h
GFP+ cells 105 shSLC6A6 105 shSLC6A6
0.5 0.5 2.95
0.5 19.8 1
2.5 Gy 104 104
NSG
huCD45 APC

huCD45 APC
103 103
0 0 0 0
102 102

A
A

cZ
cZ
A

6-
A

6-
SL acZ
cZ

6-

La
6-

La

6A
6A

101 101
La

6A
6A

sh
L

sh

C
C
sh
sh

SL
C

SL
SL

sh
sh
sh
sh

–103 0 103 104 105 –103 0 103 104 105

Side scatter Side scatter

Fig. 4 | TAUT inhibition impairs growth of patient-derived AML cells. h, The combination index of GES and venetoclax calculated per fraction affected
a, SLC6A6 expression based on venetoclax (Ven) response (BEAT-AML). n%=%193; (left) and the normalized isobologram (right), as determined using the Chou–
dot plots show individual patients. Statistical analysis was performed using the Talalay method. n%=%2 independent primary human AML samples. Colours
DESeq2 log-rank test. b, SLC6A6 expression in primitive and monocytic ROSlow represent independent samples; shapes represent indicated venetoclax and
AML LSCs32 (Gene Expression Omnibus (GEO): GSE132511). n%=%7 (primitive) and GES combinations. D, drug dose; Dx, median-effect dose. i,j, CFU analysis of
n%=%5 (monocytic). For the box plots in a and b, the centre line shows the median, primary human AML cells (i) or normal CD34+ HSPCs ( j) that were transduced
the box limits show the interquartile range and the whiskers represent 1.5%(% with lentiviral shRNAs targeting LacZ (control) or human SLC6A6. Data are
interquartile range. c, SLC6A6 expression of human NRASG12D+ versus wild-type mean%±%s.e.m. n%=%3 independent culture wells from n%=%3 independent primary
cells (GEO: GSE253715). Statistical analysis was performed using the Seurat human patient samples. Each colour represents an independent sample.
Findmarker function with the Wilcoxon rank-sum test 33. d,e, CFU analysis of k–m, Experimental strategy (k) and representative FACS plots and graph, showing
primary human AML (d) or normal human CD34+ bone marrow HSPCs (e) treated bone marrow engraftment of primary human AML (l) or primary human normal
with dimethyl sulfoxide (DMSO)/water (control) or the indicated doses of TAG CD34+ HSPC (m) cells. The black lines show the mean. n%=%9 animals per cohort
and GES. For d and e, data are mean%±%s.e.m. n%=%3 independent culture wells from n%=%4 independent primary human AML samples (l); and n%=%13 animals per
per sample from two independent samples; each colour represents a sample. cohort from n%=%5 independent normal human CD34+ HSPC samples (m). Each
f,g, CFU analysis of human AML cells treated with DMSO/water (control) or dot represents an animal; each colour represents an independent sample.
venetoclax (ABT-199) in combination with the indicated doses of TAG (f) or GES Statistical analysis was performed using two-sample Wilcoxon tests (b), one-way
(g). Data are mean%±%s.e.m. n%=%3 independent culture wells per sample from two ANOVA (d–g) and unpaired two-tailed Student’s t-tests (i, j, l and m). The mouse
independent primary human AML samples. Each colour represents a sample. and culture well images in k are adapted from ref. 6, Springer Nature America.

by Slc6a6−/− cells (Fig. 5g), and N-acetyl-taurine partly rescued this defect mouse leukaemia cells (Fig. 5h). Our RNA-seq analysis identified 932
(Extended Data Fig. 10h), perhaps by breakdown to taurine and acetate37. downregulated and 1,158 upregulated genes in the absence of TAUT
By contrast, glutaurine was unable to rescue the Slc6a6−/− colony forma- (Padj%<%0.05). The 1,158 upregulated genes included pathways associated
tion (Extended Data Fig. 10i). These experiments indicate that taurine, with haematopoietic cell lineage and cell cycle regulation (Extended
and not a secondary metabolite, promotes leukaemia growth due to an Data Fig. 11a,b). The 932 downregulated genes primarily consti-
indirect effect on glycolysis, possibly through downstream signalling. tuted pathways associated with glycolysis (Fig. 5i and Extended Data
To identify signals downstream of taurine, we performed RNA-seq Fig. 11b–d), consistent with reduced abundance of glycolysis associated
as well as phosphoproteome and total-proteome analyses of primary metabolites in these cells (Fig. 5c). To determine effectors of glycolytic

Nature | [Link] | 7
 Article
a b Top metabolic pathways enriched in Slc6a6+/+ c G3P 3PG Pyruvate d e
P = 0.046 2.5 × 107 P = 0.004 P = 0.0005 +/+ 2.0 P < 0.0001 P = 0.0003 +/+
Slc6a6+/+ or Slc6a6–/– 4 × 105 6 × 107 250
Taurine and hypotaurine metabolism –/– –/–

(mpH per min per 50,000 cells)

(mpH per min per 50,000 cells)
Lin– leukaemia cells Histidine metabolism
Citrate cycle (TCA cycle) P 2.0 × 107 200
6 Gy 9 × 10–7 3 × 105 1.5

Normalized ECAR
Pyruvate metabolism
B6 Glycolysis/gluconeogenesis 4 × 107
7 × 10–2 1.5 × 107 150
Alanine, aspartate and glutamate metabolism

ECAR
AUC
Magnetic Lipoic acid metabolism 1 × 10–1 2 × 105 1.0
11–13 d Glyoxylate and dicarboxylate metabolism
sorting 1.0 × 107 100
Glycine, serine and threonine metabolism 2 × 107
Lin– leukaemia Cysteine and methionine metabolism
cells 1 × 105 0.5
Arginine and proline metabolism 5.0 × 106 50
Tyrosine metabolism
Primary bile and acid biosynthesis
Untargeted 0 0 0 0 0
metabolomics +/+ –/– +/+ –/– +/+ –/– 0 20 40 60

0
20
40
60
80
0

c i ic

rv c
se ti
10

ly a l

pa lyt

re oly
ty

e
co Bas
Time (min)

ca co
Enrichment ratio

si

c
ly

ly
G

G
y
gl
f g h i j
P = 0.023 Slc6a6+/+ or Slc6a6–/– Top 15 pathways downregulated by TAUT loss Hallmark pathways GSEA
P < 0.0001 P = 0.62 Lin– leukaemia cells FDR q
35 Glycolytic process through g-6-p Padj
P = 0.1 P < 0.003 1.5 P = 0.27 Canonical glycolysis 0 0.05 0.10 0.15 0.20 0.25
P < 0.0001 P < 0.0001 6 Gy Glucose catabolic process to pyruvate 0.04
30 B6 0.03
Glycolysis and gluconeogenesis
Relative no. of colonies

P < 0.0001 MTORC1 signalling

MECOM ChIP–seq KASUMI mouse 0.02
25
No. of colonies

1.0 11–13 d FACS Ribosome biogenesis 0.01

MYB ChIP–seq ERMYB mouse MYC targets v1
20 Proportion
Leukaemia ncRNA processing Unfolded protein response
Glycolysis/gluconeogenesis expressed
15 cells genes in Cholesterol
rRNA metabolic processes
0.5 rRNA processing term homeostasis
10 0.1
RNA-seq Phospho- Ribosome biogenesis in eukaryotes MYC targets v2
Gluconeogenesis 0.2
5 proteomics/ 0.3
Neutrophil degranulation Angiogenesis
proteomics Neutrophil activation in immune response 0.4
0 0 0.5
0 50 100 0 50 100 150 200 250
Py H2 O

te
Py H2 O

te

N SO

-P in l

in l

-O O
-P in l

in l
N aur eoy

ur oy

N aur eoy

ur oy
va

va

N S
al e

DMe

al e

Genes
ta mit

ta mit
DM

Number of significantly down-expressed

ru

ru

t Ol

t l
-

genes matching the term

+/+ –/–
+/+ –/–

k l m p-mTOR p-EIF4B p-S6K

n mTOR inhibition
o mTOR activation
+/+ –/–
Slc6a6+/+ or Slc6a6–/– P = 0.004 P = 0.006 P = 0.03 P < 0.0001 1.5 P < 0.0001
kDa WT1 WT2 WT3 WT4 WT5 KO1 KO2 KO3 2.0 1.5 2.0 1.5
Lin– leukaemia cells p-mTOR P < 0.0001 P < 0.0001
normalized to the loading control

250
Phosphorylated/total protein

(Ser2448) P < 0.0001

Relative no. of colonies

Relative no. of colonies
6 Gy 250 mTOR P < 0.0001
1.5 1.5
B6 p-EIF4B 1.0 1.0
1.0
75 (Ser422) P = 0.13
Magnetic 75 EIF4B 1.0 1.0 P = 0.997
11–13 d
sorting 50 Tubulin P = 0.12
Lin– leukaemia 75 p-p70S6K 0.5 0.5 0.5
cells (Thr389)
75 0.5 0.5
p70S6K

50 Tubulin
Western blots
0 0 0 0 0
+/+ –/– +/+ –/– +/+ –/– 1 10 100 1 10 100

H O

D 5
H O

5
SO

SO

48

48
M MS

M MS
nM nM

Y1

Y1
DM

DM

D
rapamycin rapamycin
+/+ –/– +/+ –/–

p Upstream regulation q r s P = 0.98

P < 0.0001 WT + vector WT + RAGA KO + vector KO + RAGA P = 0.998
Slc6a6+/+ or Slc6a6–/–
P = 0.28 Lin– leukaemia cells P = 0.005
P < 0.0001 P < 0.0001
1.5 48h Vector-mCherry or 1.0 P = 0.029
P = 0.017
–Taurine

RAGA(Q66L)-mCherry
Pearson’s correlation between
P = 0.99 P < 0.0001 P < 0.0001
mCherry+ 0.8
Relative no. of colonies

sorted cells mTOR and LAMP1 (r)

mTOR LAMP1 DAPI

1.0 0.6
Overnight X-Vivo medium +
1% FBS and cytokines 0.4
Cell-Tak-coated
0.2
0.5 coverglass chambers
0
+Taurine

30 min PBS
1.5 h –0.2
0 RPMI –AA with or without –0.4
1 mM taurine
66 A

66 A
(Q RAG r

64 EB

(Q RAG r

64 EB

Ve /+

G+

Ve /+

G+
Ve /–

G–

Ve–/–

G–
o

A
RA +/

RA –/

RA +/

RA –/
L)

L)
ct

VeL)
ct

L)

c
+

+
(Q H

(Q H
Ve

Fix and stain

+/+ –/– –Taurine +Taurine

Fig. 5 | TAUT loss impairs glycolysis and mTOR signalling in myeloid p-mTOR and p-pS6K. Data were combined from two independent experiments,
leukaemia cells. a, The untargeted metabolomics strategy. b, Unbiased Enrichr indicated by the two colours. n%=%5 (Slc6a6 +/+) and n%=%3 (Slc6a6−/−) for p-EIF4B.
analysis (MetaboAnalyst; Padj%)%0.05). TCA, tricarboxylic acid. c, Quantification For d, e and m, statistical analysis was performed using unpaired two-tailed
of glycolysis-associated metabolites. Data are mean%±%s.e.m. n%=%16 samples Student’s t-tests. n,o, CFU analysis of Lin− bcCML cells treated with rapamycin (n)
from n%=%6 Slc6a6 +/+ leukaemic mice and n%=%9 samples from n%=%3 Slc6a6−/− or 2%µM mTOR activator MHY1485 (o). Data are mean%±%s.e.m. n%=%6 independent
leukaemic mice. Statistical analysis was performed using unpaired Student’s culture wells per cohort. Data were combined from two independent experiments.
t-tests with Welch’s correction. AUC, area under the curve; G3P, glyceraldehyde-3- p, CFU analysis of Lin− bcCML cells infected with vector, RAGA(Q66L) or
phosphate; 3PG, 3-phosphoglyceric acid. d,e, Extracellular acidification (ECAR) RHEB(Q64L). Data are mean%±%s.e.m. n%=%9 independent culture wells per cohort.
curve (d) and quantification (e). Data are mean%±%s.e.m. n%=%18 independent Data were combined from three independent experiments. q–s, The strategy
culture wells per cohort from four bcCML samples. Data are combined to determine mTOR and LAMP1 co-localization in Lin− bcCML cells infected
from four independent experiments. f,g, CFU analysis of Lin− bcCML cells with RAGA(Q66L) or vector (vec) with or without taurine (q), microscopy
supplemented with 1%mM pyruvate (f) or 80%µM cell-permeable taurine images (r) and analysis (s). n%=%83 (Slc6a6 +/+, vector, −taurine), 58 (Slc6a6−/−,
conjugates (g). Data are mean%±%s.e.m. n%=%3 independent culture wells per cohort. vector, −taurine), 125 (Slc6a6 +/+, RAGA, −taurine), 43 (Slc6a6−/−, RAGA, −taurine),
Data were combined from two independent experiments. h, Experimental 78 (Slc6a6 +/+, vector, +taurine), 68 (Slc6a6−/−, vector, +taurine), 40 (Slc6a6 +/+,
strategy. i, Unbiased Enrichr analysis of transcriptomic data. ChIP–seq, RAGA, +taurine) and 23 (Slc6a6−/−, RAGA, +taurine). Data were combined from
chromatin immunoprecipitation followed by sequencing; g-6-p, glucose- two independent experiments. mTOR and LAMP1 co-colocalization is indicated
6-phosphate; ncRNA, non-coding RNA. j, Gene sets significantly enriched in by white arrows. AA, amino acids; KO, knockout. For f, g, n–p and s, statistical
Slc6a6 +/+ leukaemic mice. GSEA, gene set enrichment analysis. k–m, Schematic analysis was performed using one-way ANOVA. Scale bars, 5%µm (r). Blue text in
(k), immunoblot (l) and quantification (m) of the indicated proteins (Extended b, i and j represents pathways of interest. The mouse and culture well images in
Data Fig. 11e). Data are mean%±%s.e.m. n%=%9 (Slc6a6 +/+) and n%=%7 (Slc6a6−/−) for a, h, k and q are adapted from ref. 6, Springer Nature America.

8 | Nature | [Link]
downregulation on TAUT loss, we performed gene set enrichment possibly results from downregulation of MSC differentiation signals
analysis. This identified profound downregulation of oncogenic MYC during leukaemia progression, and may explain conflicting findings
and mTOR pathways (Fig. 5j and Extended Data Fig. 11c). using candidate osteoblast markers49,50. Our data indicate that temporal
As mTOR signalling is known to regulate expression of glycolysis- expansion in arteriolar endothelial cells is accompanied by a loss in
related genes38, we tested whether TAUT loss impairs mTOR activation. signals that are essential for endothelial cell integrity and function,
Western blot analysis showed that TAUT loss results in a 2.2- to 3-fold consistent with leaky blood vessels seen by in vivo imaging of the AML
reduction in activated mTOR, p70S6k and pEIF4B as compared to the niche1,3,51. In addition to population-level changes, our studies identify
controls (Fig. 5k–m and Extended Data Fig. 11e). Our global phospho- signals such as KIT, thrombospondin, taurine and apolipoproteins
proteome and total-proteome analyses also showed reduced expres- from the TME that are essential for cancer progression. While multiple
sion of phosphorylated proteins associated with mTOR signalling, TME ligands such as APOE and PVR are detected through the disease
as well as 3.3-fold reduced BCL2 and 1.5-fold increased LC3A expres- trajectory, the populations expressing these can change over time.
sion in the absence of TAUT (Extended Data Fig. 11f,g). Like TAUT loss, Thus, therapeutic approaches aimed at targeting TME-driven signals
TAUT inhibition with TAG and GES downregulated glycolysis-related may be more effective than blocking TME remodelling or inhibiting
gene expression (Extended Data Fig. 11h,i) and impaired phosphoryla- individual stromal populations.
tion of mTOR signalling (Extended Data Fig. 12a,b). Consistent with a Consistent with clinicopathological similarities between AML and
direct impact of taurine on mTOR signalling, acute treatment in vitro bcCML52, our gene expression analysis of human AML and bcCML CD34+
and long-term taurine dosing in vivo could activate mTOR signalling LSCs identifies distinct overlap. Our unbiased approach to determine
(Extended Data Fig. 12c–h). Inhibiting mTOR with rapamycin impaired LSC-enriched cell surface receptors that are essential for disease pro-
the growth of wild-type cells but did not impact Slc6a6−/− cells in vitro gression identified multiple genes that are known to be critical for
or in vivo (Fig. 5n and Extended Data Fig. 12i,k). By contrast, mTOR acti- AML, including CD9653, CD4754 and protein kinase D255. Although these
vation with MHY148539 rescued the colony-forming ability of Slc6a6−/− are required for leukaemic progression, they are not associated with
LSCs to 60% of the control, and membrane-permeable taurine rescued poor prognosis (TCGA-LAML), in contrast to LDLR and SLC6A6 that
mTOR activation (Fig. 5o and Extended Data Fig. 12j–l). Collectively, we describe here. It is therefore possible that other cell surface recep-
these data indicate that mTOR signalling has a key role downstream of tors identified by our analysis also have a functional role in disease
taurine in leukaemia cells. progression and should be explored further to determine new thera-
Amino acid availability can be sensed by RAG GTPases to facilitate peutically relevant signals. As the APOE–LDLR and taurine–TAUT axes
mTORC1 localization to the outer lysosomal membrane40 and its inter- that we identify are known to have a role in ageing27,56, it is possible that
action with RHEB GTPases41, thereby promoting mTORC1 activation. cancer-associated signals in our TME–LSC interactome may be of broad
We hypothesized that taurine may activate mTOR by promoting its relevance in ageing-related disorders such as MDSs, as we see with TAUT.
interaction with RAG GTPases. To test this, we determined the effect Biosynthesis of taurine from cysteine is known to occur in the liver,
of expressing constitutively active RAGA(Q66L) or RHEB(Q64L) on the kidneys, adipose tissues and pancreas24. Our data identify bone mar-
colony-forming ability of Slc6a6 wild-type and null cells. Our experi- row osteolineage cells as a novel source of taurine in the leukaemia
ments showed that, while RHEB(Q64L) partially rescued Slc6a6−/− col- niche. Our studies blocking taurine produced by osteolineage cells
ony formation, this could be fully rescued by RAGA(Q66L) (Fig. 5p). establish a key role of TME-driven taurine synthesis in LSC survival and
Furthermore, mTOR phosphorylation in Slc6a6−/− cells could be rescued self-renewal. However, it is possible that taurine or &-alanine produced
by RAGA(Q66L) (Extended Data Fig. 12m). Finally, we directly tested outside the bone marrow niche also contribute to disease progression.
whether taurine could promote mTOR interactions with lysosomes Consistent with this, our data suggest that taurine supplements can
in wild-type and Slc6a6−/− leukaemia cells expressing RAGA(Q66L) accelerate myeloid leukaemia progression in mouse models. As taurine
(Fig. 5q). Our immunofluorescence-based analysis showed that, at the is a common ingredient in energy drinks, and is often provided as a
baseline, Slc6a6−/− cells have 1.2-fold lower mTOR–LAMP1 interaction, supplement to mitigate the side-effects of chemotherapy17, our work
which could be fully rescued by RAGA(Q66L). Taurine supplements suggests that it may be of interest to carefully consider the benefits of
increased mTOR–LAMP1 co-localization to levels seen in cells express- supplemental taurine in patients with leukaemia.
ing RAGA(Q66L). However, this interaction could not be increased Mechanistically, we identify a critical requirement of taurine from the
further by adding taurine to RAGA(Q66L)-expressing cells (Fig. 5r,s), microenvironment in regulating mTOR-driven glycolysis in leukaemia
consistent with our colony-forming assays (Extended Data Fig. 12n). cells (Extended Data Fig. 12o). We can rescue the growth of Slc6a6-null
Collectively, our studies show that the taurine–TAUT axis regulates LSCs by circumventing glycolysis with pyruvate or by ectopically acti-
glycolysis in myeloid leukaemia cells by promoting RAG-mediated vating mTORC1 using GTP-bound RAGA mutants. Our data showing
activation of mTOR signalling. that constitutively active Rag-GTPases can rescue mTOR interactions
with lysosomes, and rescue mTOR phosphorylation, in Slc6a6−/− LSCs
indicate that taurine levels in leukaemia cells may be detected by hith-
Discussion erto unidentified sensors, like those for arginine and leucine57,58. Our
We use scRNA-seq to define the changing landscape of the non-immune in vivo data showing a strong impact of targeting taurine uptake using
cancer microenvironment with disease progression, and to iden- genetic approaches in AML suggest that it would be of considerable
tify unique niche driven signals that promote disease progression. interest to develop stable and effective in vivo inhibitors of taurine in
scRNA-seq has been effectively used to characterize the immune micro- future studies. In light of early clinical success of glutamine inhibitors in
environment of both solid tumours and leukaemias20,42–45. Despite tech- MDS and AML59,60, our work suggests that evaluating taurine-transport
nical limitations in detecting non-immune stromal cells, a few studies inhibitors in normal and leukaemic cells may be of therapeutic interest.
have identified antigen-presenting fibroblasts and immune-suppressive
endothelial cells in lung and pancreatic cancers46–48. A reduction in
osteoblasts has been noted during leukaemia initiation in bone marrow Online content
niche damaged by irradiation13. However, dynamic changes in the TME Any methods, additional references, Nature Portfolio reporting summa-
during leukaemia progression have not been defined. ries, source data, extended data, supplementary information, acknowl-
Our temporal scRNA-seq based TME analysis identifies an expansion edgements, peer review information; details of author contributions
in MSCs and their immature osteo-associated progeny, along with a loss and competing interests; and statements of data and code availability
in mature osteo-associated cells. This skew in osteolineage populations are available at [Link]

Nature | [Link] | 9
Article
1. Bajaj, J. et al. CD98-mediated adhesive signaling enables the establishment and 35. Yarbrough, G. G., Singh, D. K. & Taylor, D. A. Neuropharmacological characterization of a
propagation of acute myelogenous leukemia. Cancer Cell 30, 792–805 (2016). taurine antagonist. J. Pharmacol. Exp. Ther. 219, 604–613 (1981).
2. Kaur, A. et al. sFRP2 in the aged microenvironment drives melanoma metastasis and 36. Huxtable, R. J. Guanidinoethane sulfonate and the disposition of dietary taurine in the rat.
therapy resistance. Nature 532, 250–254 (2016). J. Nutr. 112, 2293–2300 (1982).
3. Kwon, H. Y. et al. Tetraspanin 3 is required for the development and propagation of acute 37. Wei, W. et al. PTER is a N-acetyltaurine hydrolase that regulates feeding and obesity.
myelogenous leukemia. Cell Stem Cell 17, 152–164 (2015). Nature 633, 182–188 (2024).
4. Orimo, A. et al. Stromal fibroblasts present in invasive human breast carcinomas promote 38. Yecies, J. L. & Manning, B. D. Transcriptional control of cellular metabolism by mTOR
tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121, signaling. Cancer Res. 71, 2815–2820 (2011).
335–348 (2005). 39. Choi, Y. J. et al. Inhibitory effect of mTOR activator MHY1485 on autophagy: suppression
5. Vermeulen, L. et al. Wnt activity defines colon cancer stem cells and is regulated by the of lysosomal fusion. PLoS ONE 7, e43418 (2012).
microenvironment. Nat. Cell Biol. 12, 468–476 (2010). 40. Sancak, Y. et al. The Rag GTPases bind raptor and mediate amino acid signaling to mTORC1.
6. Bajaj, J. et al. An in vivo genome-wide CRISPR screen identifies the RNA-binding protein Science 320, 1496–1501 (2008).
Staufen2 as a key regulator of myeloid leukemia. Nat. Cancer 1, 410–422 (2020). 41. Yang, H. et al. Mechanisms of mTORC1 activation by RHEB and inhibition by PRAS40.
7. Sun, Y. et al. Treatment-induced damage to the tumor microenvironment promotes Nature 552, 368–373 (2017).
prostate cancer therapy resistance through WNT16B. Nat. Med. 18, 1359–1368 (2012). 42. Tirosh, I. et al. Dissecting the multicellular ecosystem of metastatic melanoma by single-
8. Bissell, M. J. & Radisky, D. Putting tumours in context. Nat. Rev. Cancer 1, 46–54 (2001). cell RNA-seq. Science 352, 189–196 (2016).
9. Bajaj, J., Diaz, E. & Reya, T. Stem cells in cancer initiation and progression. J. Cell Biol. 219, 43. Venteicher, A. S. et al. Decoupling genetics, lineages, and microenvironment in IDH-mutant
e201911053 (2020). gliomas by single-cell RNA-seq. Science 355, eaai8478 (2017).
10. Sharma, P. et al. The next decade of immune checkpoint therapy. Cancer Discov. 11, 44. Azizi, E. et al. Single-cell map of diverse immune phenotypes in the breast tumor
838–857 (2021). microenvironment. Cell 174, 1293–1308 (2018).
11. de Visser, K. E. & Joyce, J. A. The evolving tumor microenvironment: from cancer initiation 45. Steele, N. G. et al. Multimodal mapping of the tumor and peripheral blood immune
to metastatic outgrowth. Cancer Cell 41, 374–403 (2023). landscape in human pancreatic cancer. Nat. Cancer 1, 1097–1112 (2020).
12. Elhanani, O., Ben-Uri, R. & Keren, L. Spatial profiling technologies illuminate the tumor 46. Lambrechts, D. et al. Phenotype molding of stromal cells in the lung tumor
microenvironment. Cancer Cell 41, 404–420 (2023). microenvironment. Nat. Med. 24, 1277–1289 (2018).
13. Baryawno, N. et al. A cellular taxonomy of the bone marrow stroma in homeostasis and 47. Elyada, E. et al. Cross-species single-cell analysis of pancreatic ductal adenocarcinoma
leukemia. Cell 177, 1915–1932 (2019). reveals antigen-presenting cancer-associated fibroblasts. Cancer Discov. 9, 1102–1123
14. Tikhonova, A. N. et al. The bone marrow microenvironment at single-cell resolution. (2019).
Nature 569, 222–228 (2019). 48. Lee, J. J. et al. Elucidation of tumor-stromal heterogeneity and the ligand-receptor
15. Baccin, C. et al. Combined single-cell and spatial transcriptomics reveal the molecular, interactome by single-cell transcriptomics in real-world pancreatic cancer biopsies.
cellular and spatial bone marrow niche organization. Nat. Cell Biol. 22, 38–48 (2020). Clin. Cancer Res. 27, 5912–5921 (2021).
16. Kamijo, H. et al. Thrombospondin-1 promotes tumor progression in cutaneous T-cell 49. Frisch, B. J. et al. Functional inhibition of osteoblastic cells in an in vivo mouse model of
lymphoma via CD47. Leukemia 34, 845–856 (2020). myeloid leukemia. Blood 119, 540–550 (2012).
17. Islambulchilar, M., Asvadi, I., Sanaat, Z., Esfahani, A. & Sattari, M. Effect of taurine on 50. Hanoun, M. et al. Acute myelogenous leukemia-induced sympathetic neuropathy
attenuating chemotherapy-induced adverse effects in acute lymphoblastic leukemia. promotes malignancy in an altered hematopoietic stem cell niche. Cell Stem Cell 15,
J. Cancer Res. Ther. 11, 426–432 (2015). 365–375 (2014).
18. Cao, T. et al. Cancer SLC6A6-mediated taurine uptake transactivates immune checkpoint 51. Passaro, D. et al. Increased vascular permeability in the bone marrow microenvironment
genes and induces exhaustion in CD8+ T cells. Cell 187, 2288–2304 (2024). contributes to disease progression and drug response in acute myeloid leukemia. Cancer
19. Bausch-Fluck, D. et al. A mass spectrometric-derived cell surface protein atlas. PLoS ONE Cell 32, 324–341 (2017).
10, e0121314 (2015). 52. Hehlmann, R. How I treat CML blast crisis. Blood 120, 737–747 (2012).
20. Lasry, A. et al. An inflammatory state remodels the immune microenvironment and 53. Hosen, N. et al. CD96 is a leukemic stem cell-specific marker in human acute myeloid
improves risk stratification in acute myeloid leukemia. Nat. Cancer 4, 27–42 (2023). leukemia. Proc. Natl Acad. Sci. USA 104, 11008–11013 (2007).
21. Zhang, S. H., Reddick, R. L., Piedrahita, J. A. & Maeda, N. Spontaneous hypercholesterolemia 54. Majeti, R. et al. CD47 is an adverse prognostic factor and therapeutic antibody target on
and arterial lesions in mice lacking apolipoprotein E. Science 258, 468–471 (1992). human acute myeloid leukemia stem cells. Cell 138, 286–299 (2009).
22. Liu, Q. R., López-Corcuera, B., Nelson, H., Mandiyan, S. & Nelson, N. Cloning and 55. Mihailovic, T. et al. Protein kinase D2 mediates activation of nuclear factor κB by Bcr-Abl in
expression of a cDNA encoding the transporter of taurine and beta-alanine in mouse Bcr-Abl+ human myeloid leukemia cells. Cancer Res. 64, 8939–8944 (2004).
brain. Proc. Natl Acad. Sci. USA 89, 12145–12149 (1992). 56. Finch, C. E. & Shams, S. Apolipoprotein E and sex bias in cerebrovascular aging of men
23. Bandyopadhyay, S. et al. Mapping the cellular biogeography of human bone marrow and mice. Trends Neurosci. 39, 625–637 (2016).
niches using single-cell transcriptomics and proteomic imaging. Cell 187, 3120–3140 57. Chantranupong, L. et al. The CASTOR proteins are arginine sensors for the mTORC1
(2024). pathway. Cell 165, 153–164 (2016).
24. Ueki, I., Roman, H. B., Hirschberger, L. L., Junior, C. & Stipanuk, M. H. Extrahepatic tissues 58. Wolfson, R. L. et al. Sestrin2 is a leucine sensor for the mTORC1 pathway. Science 351,
compensate for loss of hepatic taurine synthesis in mice with liver-specific knockout of 43–48 (2016).
cysteine dioxygenase. Am. J. Physiol. Endocrinol. Metab. 302, E1292–E1299 (2012). 59. DiNardo, C. D. et al. Glutaminase inhibition in combination with azacytidine in
25. Ito, T. et al. Taurine depletion caused by knocking out the taurine transporter gene leads myelodysplastic syndromes: a phase 1b/2 clinical trial and correlative analyses. Nat. Cancer
to cardiomyopathy with cardiac atrophy. J. Mol. Cell. Cardiol. 44, 927–937 (2008). 5, 1515–1533 (2024).
26. Preising, M. N. et al. Biallelic mutation of human SLC6A6 encoding the taurine transporter 60. Liu, Y. et al. A phase 1 study of the amino acid modulator pegcrisantaspase and venetoclax
TAUT is linked to early retinal degeneration. FASEB J. 33, 11507–11527 (2019). for relapsed or refractory acute myeloid leukemia. Blood 145, 486–496 (2025).
27. Singh, P. et al. Taurine deficiency as a driver of aging. Science 380, eabn9257 (2023).
28. Ito, T. et al. Regulation of myeloid leukaemia by the cell-fate determinant Musashi. Nature Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
466, 765–768 (2010). published maps and institutional affiliations.
29. Kharas, M. G. et al. Musashi-2 regulates normal hematopoiesis and promotes aggressive
myeloid leukemia. Nat. Med. 16, 903–908 (2010). Open Access This article is licensed under a Creative Commons Attribution
30. Dey, A. et al. BRD4 directs hematopoietic stem cell development and modulates 4.0 International License, which permits use, sharing, adaptation, distribution
macrophage inflammatory responses. EMBO J. 38, e100293 (2019). and reproduction in any medium or format, as long as you give appropriate
31. Veis, D. J., Sorenson, C. M., Shutter, J. R. & Korsmeyer, S. J. Bcl-2-deficient mice demonstrate credit to the original author(s) and the source, provide a link to the Creative Commons licence,
fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, and indicate if changes were made. The images or other third party material in this article are
229–240 (1993). included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
32. Pei, S. et al. Monocytic subclones confer resistance to venetoclax-based therapy in to the material. If material is not included in the article’s Creative Commons licence and your
patients with acute myeloid leukemia. Cancer Discov. 10, 536–551 (2020). intended use is not permitted by statutory regulation or exceeds the permitted use, you will
33. Sango, J. et al. RAS-mutant leukaemia stem cells drive clinical resistance to venetoclax. need to obtain permission directly from the copyright holder. To view a copy of this licence,
Nature 636, 241–250 (2024). visit [Link]
34. Shlush, L. I. et al. Tracing the origins of relapse in acute myeloid leukaemia to stem cells.
Nature 547, 104–108 (2017). © The Author(s) 2025

10 | Nature | [Link]
Methods
Generation and analysis of leukaemia models
Generation of experimental mice Bone marrow KLS cells were sorted from Slc6a6+/+ or Slc6a6−/− mice and
The Slc6a6 mice were bred as described previously25. For all mouse leu- cultured overnight in X-VIVO15 (Lonza) medium supplemented with
kaemia experiments, male and female Slc6a6+/+ and Slc6a6−/− mice were 10% FBS (GeminiBio), 50%µM 2-mercatpoethanol, SCF (100%ng%ml−1, R&D
used as donors and B6-CD45.1 ([Link]-PtprcaPepcb/BoyJ) or C57BL6/J Systems), TPO (10%ng%ml−1, R&D Systems) and penicillin–streptomycin
mice were used as transplant recipients. For xenograft experiments (Gibco). Cells were retrovirally infected with MSCV-BCR-ABL-IRES-GFP
with human cells, NSG mice ([Link]-Prkdcscid Il2rgtm1Wjl/SzJ) were (or -tNGFR) and MSCV-NUP98-HOXA9-IRES-YFP (or -huCD2 or -tNGFR)
used as transplant recipients. All mice were 6–16 weeks of age. All ani- to generate bcCML. AML was generated by sorting bone marrow KLS cells
mals used for in vivo studies were randomly selected to receive either from Slc6a6+/+ or Slc6a6−/− mice and culturing in RPMI medium (Gibco)
control or treatments. Animals were selected based on genotype, age supplemented with 20% FBS, 50%µM 2-mercaptoethanol, 100%ng%ml−1
and sex. No statistical method was used to predetermine sample size SCF (R&D Systems), 10%ng%ml−1 IL-3, and 10%ng%ml−1 IL-6 (R&D Systems).
for experiments. Adequate sample size was determined based on pre- Cells were retrovirally infected with MSCV-MLL-AF9-IRES-tNGFR and
vious publications1,3,6. Mice were bred and maintained in the animal MSCV-NRAS(G12V)-IRES-YFP cells were collected 48%h after infection,
care facilities at the University of Rochester. All animal experiments sorted by FACS for BCR-ABL+ and NUP98-HOXA9+ (bcCML only) and
were performed according to protocols approved by the University retro-orbitally transplanted into cohorts of sublethally irradiated (6%Gy)
of Rochester’s Committee on Animal Resources. Mice were housed C57BL/6J mice. For AML-ETOa9 and NRAS, cells were transplanted into
with the same sex in ventilated cages, under a 12%h–12%h light–dark lethally irradiated (9.5%Gy) C57BL/6J recipients along with 3%(%105 RBC
cycle with temperature (64–79%°F) and humidity (winter levels, <30%; lysed bone marrow rescue cells. For secondary transplants, Lin− cells
summer levels, >70%) control. Mice were given enrichment material from primary bcCML recipient mice were transplanted into second-
and fed a standard chow diet. Irradiated recipients were maintained ary sublethally irradiated recipients. The recipients were maintained
on acid water HYDROPAC (pH%2.5–3.0) and irradiated sulfatrim diet on acid water HYDROPAC and irradiated sulfatrim diet and evaluated
(Mod LabDiet 5P00 with 0.025% trimeth and 0.124% sulfameth). NSG daily. Recipients receiving additional taurine (T8691, Sigma-Aldrich)
recipients were housed in BSL-2 facility. Cdo1fl/fl mice were generated were maintained on regular chow without taurine (D10012Gi, Research
in collaboration with the Transgenic and Genome Editing Core Facility Diets). Taurine was dissolved in autoclaved acid water from the
at Augusta University using CRISPR–Cas9 gene editing. These mice HYDROPAC and supplied in sterile water bottles. For in vivo venetoclax
were bred to the MSC-specific Prrx1-cre ([Link]-Tg(Prrx1-cre)1Cjt/J) to treatments, venetoclax (ABT-199; Tocris Bioscience) solution was made
generate a Cdo1fl/flPrrx1-cre+ mice, where taurine production is blocked fresh daily in a solvent containing 10% ethanol with 60% Phosal 50 PG
in the MSC/osteolineage cells. As Prrx1-Cre can be expressed in the and 30% PEG-400, and delivered by oral gavage at a final dose of 50%mg
female germline, only male Prrx1-cre+ mice were used for breeding per kg. GES (Toronto Research Chemicals or MedChemExpress) and
and experiments. Blinding was not relevant to mouse experiments TAG-HCl (synthesized by Enamine) were dissolved in autoclaved acid
as researchers needed to know the conditions for each experiment. water from the HYDROPAC and supplied in sterile water bottles (GES)
Flow cytometry, FACS, western blots, Seahorse analysis, sequencing or intraperitoneally (i.p.) (TAG-HCl). For in vivo rapamycin treatments,
and imaging used analytical machines and blinding was therefore rapamycin (Selleck Chemicals) stock solution (50%mg%ml−1) was made in
not necessary. ethanol (Sigma-Aldrich). Single-use aliquots of the stock were diluted
fresh each day in vehicle containing equal parts of 10% PEG-40 with 8%
Cell isolation and FACS analysis ethanol and 10% Tween-80 solutions. Mice received i.p. 5%mg per kg
Cells were suspended in Hanks’ balanced salt solution (Gibco) with 5% rapamycin or vehicle from days 5–10 after transplant. Premorbid ani-
FBS and 2%mM EDTA. Cells were prepared for FACS analysis and sorting mals were euthanized at the indicated experimental timepoints or at
as previously described1,6. Antibodies used for defining haematopoi- the end point. For all experiments, mice were monitored closely for
etic cell populations were as follows: CD3*, CD4, CD8, GR1, CD11b, signs of disease or morbidity daily and were euthanized after visible
TER119, CD45R and CD19 (all for lineage), KIT, SCA1, CD48 and CD150. signs of hunched dorsum, failure to thrive or any signs of infection.
All antibodies were purchased from, eBioscience, BioLegend or BD These limits were not exceeded for any experiment. Relevant tissues
Biosciences. A detailed list of antibodies is provided in Supplementary were collected and analysed by flow cytometry, RNA-seq, proteomics,
Table 2. Analysis was performed on LSRFortessa (BD Biosciences), and metabolomics or fixed for histology. Apoptosis assays were done using
cell sorting was performed on the FACSAria II (BD Biosciences). Data annexin V and 7AAD (eBiosciences). Analysis of in vivo bromodeoxyur-
were analysed using FlowJo. idine (BrdU) incorporation was performed using the APC BrdU Flow Kit
(BD Biosciences) after a single i.p. injection of BrdU (2%mg at 10%mg%ml−1).
Retroviral and lentiviral constructs and virus production
Retroviral MSCV-BCR-ABL-IRES-GFP (or -tNGFR) and MSCV-NUP98- RNA extraction and RT–qPCR
HOXA9-IRES-YFP (or -huCD2 and -tNGFR) were used to generate RNA was extracted using the RNeasy Micro or Mini kits (Qiagen) accord-
bcCML. AML was generated with MSCV-MLL-AF9-IRES-tNGFR and ing to the manufacturer’s protocols. RNA concentrations were deter-
MSCV-NRAS(G12V)-IRES-huCD2 or MSCV-AML-ETO9a and MSCV- mined using NanoDrop 1000 Spectrophotometer (Thermo Fisher
NRAS(G12V)-IRES-YFP. RAGA (Q66L; Addgene, 99712) and RHEB Scientific). RNA quality was assessed with the Agilent Bioanalyser 2100
(Q64L; Addgene, 64607) were cloned into the MSCV-mCherry back- (Agilent Technologies). Quantitative PCR with reverse transcription
bone (Addgene, 52114). Lentiviral shRNA constructs were designed (RT–qPCR) was performed on the Bio-Rad CFX96 C100 Thermocycler
and cloned into the pLV-hU6-EF1a-green or pLV-hU6-EF1a-red using Bio-Rad CFX Manager 1.1 v.4.1 (Bio-Rad) or Thermo Fisher Sci-
backbone (Biosettia) according to the manufacturer’s protocol. entific Quant Studio 12K Flex Real Time PCR using Quant Studio v.1.2
A detailed list of shRNA sequences is provided in Supplementary (Thermo Fisher Scientific). RT–qPCR data were analysed using Bio-Rad
Table 1. Virus was produced in 293T cells (ATCC) transfected with CFX Manager 1.1 v.4.1 or Quant Studio v.1.2.
viral constructs along with VSV-G, Gag-Pol (retroviral production)
or pRSV-rev, phCMV and pMDlg/pRRE (lentivirus production) using Isolation of mouse stromal cells for scRNA-seq and analysis
X-tremeGENE-HP reagent (Roche). Viral supernatants were collected for Microenvironmental populations were isolated as previously des-
3 to 6%days followed by ultracentrifugal concentration at 20,000%rpm cribed61. In brief, bone and bone marrow were isolated from long bones
for 2%h. and pelvis in 1( Media 199 (Gibco) with 2% FBS (GeminiBio). Bone marrow
Article
cells from 3–5 mice per timepoint were digested for 30%min in HBSS board (IRB). To isolate the bone spicules, bone marrow aspirates were
containing 2%mg%ml−1 dispase II (Gibco), 1%mg%ml−1 collagenase type IV passed through a 40%µm cell strainer. The filter containing spicules
(Sigma-Aldrich) and 20%ng%ml−1 DNase type II (Sigma-Aldrich). Bone was washed with PBS and placed in a six-well plate. The spicules were
spicules were digested for 60%min in PBS supplemented with 2.5%mg%ml−1 digested for 1%h in collagenase (Stem Cell Technologies) at 37%°C. After
collagenase type I (Stem Cell Technologies) and 20% FBS. Digested filtering, the filtered aspirate was used to isolate bone marrow mono-
bone marrow was red-blood-cell-lysed using RBC Lysis Buffer (eBio- nuclear cells by density centrifugation (Ficoll-Paque, GE Healthcare).
science). Bone and bone marrow cells were pooled and CD45+TER119+ Digested bone cells and bone marrow mononuclear cells were pooled
haematopoietic cells were magnetically depleted on the autoMACS and stained with CD45–APC (BD Biosciences) followed by staining with
cell separator (Miltenyi Biotec). The CD45−TER119− stromal cells were anti-APC microbeads (Miltenyi Biotec). The stained cells were magneti-
either stained and analysed for candidate populations by flow cytom- cally depleted for CD45− cell fraction using LD columns (Miltenyi Bio-
etry (BD LSRFortessa) or further enriched by sorting (BD FACSAria II) tec). The CD45− cells were processed for scRNA-seq as described above.
and processed for scRNA-seq. Samples containing cells expressing COL1A1 were integrated
into a single dataset using Seurat v.4.1.0 within R v.4.1.1. Cells with
Mouse bone marrow scRNA-seq and analysis mitochondrial features making up greater than 25% of the detected
Cell suspensions were processed to generate scRNA-seq libraries using transcripts were removed. The samples were scaled and normalized
the Chromium Next GEM Single Cell 3+ GEM, Library and Gel Bead Kit together, regressing out mitochondrial and globin-related content.
v3.1 (10x Genomics) according to the manufacturer’s recommenda- Globin-related content was determined using HBB, HBA2, HBA1, HBD
tions. The samples were loaded onto the Chromium Single-Cell Instru- and HBM. Further regression was performed using Harmony v.0.1.0 to
ment (10x Genomics) to generate single-cell gel bead in emulsions reduce the impact of sample specific effects. From there, the standard
(GEMs). GEM reverse transcription (GEM-RT) was performed to pro- Seurat procedure was used, clustering to a resolution of 0.8.
duce a barcoded, full-length cDNA from polyadenylated mRNA. After A normal bone marrow reference was created using GEO GSE253355
incubation, GEMs were broken, the pooled GEM-RT reaction mixtures (ref. 23). Data from this submission were normalized, PCA was run
were recovered and cDNA was purified with silane magnetic beads using the first 50 principal components, and UMAP was run using the
(DynaBeads MyOne Silane Beads, Thermo Fisher Scientific). The puri- first 50 principal components within Seurat v.5.0.3.9911. This dataset
fied cDNA was further amplified by PCR to generate sufficient material was then used within Azimuth v.0.5.0 to create the reference using
for library construction. Enzymatic fragmentation and size selection Azimuth standard methods. The samples were then typed using this
was used to optimize the cDNA amplicon size and indexed sequencing reference with the RunAzimuth function against the L1, L2 and coarse
libraries were constructed by end repair, A-tailing, adaptor ligation annotations.
and PCR. The final libraries contained the P5 and P7 priming sites used
in Illumina bridge amplification. Mouse samples were sequenced on Primary human CD34+ cell RNA-seq analysis
the Illumina NovaSeq 6000 S2 flowcell while human samples were Total RNA was purified with Qiagen RNeasy PLUS kit according to the
sequenced across several lanes of the Illumina NovaSeq X-plus 25B manufacturer recommendations and eluted in nuclease-free water. The
flowcell. The samples were demultiplexed and counted using cellranger total RNA concentration was determined using the NanoDrop 1000
v.4.0.0 mkfastq and count, using the default parameters. The samples spectrophotometer (NanoDrop), and the RNA quality was assessed
were aligned to a custom reference containing the 10x provided mm10- using the Agilent Bioanalyser (Agilent Technologies). The TruSeq
2020-A mouse reference and an additional eGFP sequence. Stranded mRNA Sample Preparation Kit (Illumina) was used for next
Seurat v.4.1.0 within R v.4.1.1 was used for most of processing. More- generation sequencing library construction according to the manu-
over, dplyr v.1.2.0 and tidyverse v.1.3.2 were used extensively for data pip- facturer’s protocols. In brief, mRNA was purified from 200%ng total
ing and transformation. Samples were imported and cells were filtered RNA with oligo-dT magnetic beads and fragmented. First-strand cDNA
for at least 200 features captured per cell and features were filtered for synthesis was performed with random hexamer priming followed by
expression in at least 3 cells. Additional filters were applied, filtering second-strand cDNA synthesis using dUTP incorporation for strand
out all cells with higher than 5% mitochondrial content and filtering marking. End repair and 3+ adenylation was then performed on the
out all cells positive for CD45, CD71, TER119 and eGFP. All samples were double-stranded cDNA. Illumina adaptors were ligated to both ends
merged using ‘merge’ and normalized using SCTransform, regressing of the cDNA, purified by gel electrophoresis and amplified with PCR
out the impact of mitochondrial features. Principal component analy- primers specific to the adaptor sequences to generate cDNA amplicons
sis was performed using the first 40 principal components (RunPCA) of approximately 200–500%bp in size. The amplified libraries were
and clusters were generated using FindNeighbors and FindClusters hybridized to the Illumina single end flow cell and amplified using the
(resolution=0.5). UMAP dimensional reduction was also performed cBot (Illumina). Single-end reads of 100 nucleotides were generated
using RunUMAP using the first 30 principal components. Clusters were for each sample using Illumina’s HiSeq2500v4.
initially typed using scMCA v.0.2.0. Populations typed as non-stromal Raw reads generated from the Illumina basecalls were demultiplexed
(B cell, pre-B cell, pro-erythrocyte, pro-erythroblast, neutrophil and using bcl2fastq v.2.19.0. Quality filtering and adapter removal was
megakaryocyte) were filtered from the dataset. Populations were fur- performed using FastP v.0.20.1. Processed reads were then mapped
ther typed using published bone marrow stroma markers13. FindAllMark- to the human reference genome (hg38 + gencode v36; [Link]
ers was used within the context of specific populations to determine [Link]/human/release_36.html) using STAR_2.7.6a. Reads
which genes change significantly over time. DEGreport-implemented mapping to genes were counted using subread featurecounts v.2.0.1.
(v.1.30.3) degPatterns was used to tie lineage-specific expression pat- Differential expression analysis was performed using DESeq2 v.1.28.1
terns to the timepoints. Expression patterns corresponding to broadly with a Padj threshold of 0.05 within R v.4.0.2. Gene Ontology analyses
increased or decreased expression over time were passed to gene set were performed using the EnrichR-3.0 package.
enrichment against KEGG_2019, GO_Biological_Process_2021, Wiki-
Pathways_2019_Mouse and ChEA_2022 databases using EnrichR-3.0. Determining cell surface protein–ligand interactions
To determine significantly expressed receptors, the differential expres-
Human bone marrow microenvironment scRNA-seq analysis sion results from human RNA-seq described above were first filtered
Bone marrow aspirates were obtained from patients with MDS/AML for significantly changing (Padj%<%0.05) genes with a log-transformed
after written informed consent in accordance with the Declaration of fold change value of greater than 0. A list of potential cell surface
Helsinki and approval of University of Rochester institutional review receptors was generated using this gene list in conjunction with cell
surface proteins detailed within the Cell Surface Protein Atlas19 and (ATCC) were maintained in RPMI/IMDM with 10% FBS, 100%IU%ml−1
essential for LSC growth in vivo6. The resultant gene list was further penicillin–streptomycin (Gibco). These cell lines were validated by
reviewed for genes that are not truly expressed within the cell surface, vendor. MDS-L cells (from K. Tohyama) were authenticated in house
leading to the removal of CHST11, ST3GAL4, ACAA1, CLCN6, CHPF2, by flow cytometry as CD45+CD34+CD38+ and maintained in RPMI
CTSK, ATP6AP1, CTSD, PNPLA6, DMXL2, TUBB6, MAN2B2, CLN3, with 10% FBS supplemented with 20%ng%ml−1 IL-3 (PeproTech). Cell
MGAT4B, MYH9 and PIGG from further review. Ligand expression lines were not tested for mycoplasma. For colony-forming assays with
from the temporal scRNA-seq dataset corresponding to bcCML and shRNAs, leukaemia and normal cells were transduced with the indicated
AML expressed receptors was determined by differential expression lentiviral shRNAs. Cells were sorted 24%h after infection and plated in
across the populations. Differentially expressed genes were filtered for CFU assays or transplanted in sublethally irradiated NSG mice.
corresponding ligands using a combination of nichenetr (v.1.1.0)62 and
literature-supported interactions. These included NRP163, ADSL64,65, MSC isolation, osteogenic differentiation and co-culture with
CDO1 and CSAD for taurine synthesis66, GADL1 and CNDP1 for &-alanine leukaemia cells
synthesis67,68, KIT69, CD15570 and CD3371 for those not captured within MSCs were isolated from leukaemic mice and cultured in 10%cm dishes
this mapping. The Broad Institute hosted dataset20 was used as a stand in MEM- with no ascorbic acid (Gibco) supplemented with 15% FBS
in for healthy immune microenvironment within the context of AML. and 100%IU%ml−1 penicillin–streptomycin (Gibco). Then, 6 days after
Counts (RNA_soupX1.[Link]) and metadata (metadata_clustering_w_ culture initiation, the cells were sorted for CD45−CD3−B220−TER119−G
header_upd.csv) were downloaded from the Broad Single Cell Portal R1−CD31−CD51+SCA1+ MSCs. Sorted cells were expanded in the medium
([Link] descried above. For co-culture experiments, 50,000 MSCs were plated
an-inflammatory-state-remodels-the-immune-microenvironment- in a 48-well plate, and transduced with the indicated lentiviral shRNAs.
and-improves-risk-stratification-in-acute-myeloid-leukaemia) and Then, 3%days after infection, osteogenic differentiation was induced
imported using Seurat. FindMarkers was used to determine which by switching to MEM- (Gibco) supplemented 15% FBS, 100%IU%ml−1
genes were upregulated within the microenvironment in relation to penicillin–streptomycin (Gibco) along with 50%µg%ml−1 ascorbic acid
the other populations. Significantly expressed genes were then fil- (Sigma-Aldrich), 10%mM &-glycerolphosphate (Sigma-Aldrich) and
tered for ligands using the same method as the stromal microenviron- 100%nM dexamethasone (Sigma-Aldrich) for 6%days. On day 6 after
ment. Significantly expressed ligand and receptor mappings within differentiation, 100,000 Lin− bcCML cells were added in X-Vivo sup-
the stroma microenvironment, the immune microenvironment and plemented with 10% FBS, 50%µM 2-mercatpoethanol and penicillin–
within bcCML and AML human bulk RNA-seq data were visualized using streptomycin. Then, 3%days after co-culture, leukaemia cells were
circlize v.0.4.15. For illustration purposes, ligands expressed with a analysed for cell viability and plated in methylcellulose for colony-
log-transformed fold change of 5 or higher were limited to 5. forming assays. Microscopy images were obtained on THE Olympus
CKX41 SYSTEM using CellSens Entry v.2.3 (Olympus).
Mouse Slc6a6 wild-type and knockout leukaemia RNA-seq
The RNeasy Plus Micro Kit (Qiagen) was used for RNA extraction. RNA Normal HSC in vivo transplantation assays
concentration was determined using the NanoDrop 1000 spectro- For bone marrow transplants, 500 HSCs were isolated from bone
photometer (NanoDrop), and the RNA quality was assessed using marrow of Slc6a6+/+ or Slc6a6−/− mice and transplanted into lethally
the Agilent Bioanalyser 2100 (Agilent Technologies). A total of 1%ng irradiated (9.5%Gy) CD45.1 mice along with 2%(%105 SCA1-depleted bone
of total RNA was pre-amplified using the SMARTer Ultra Low Input kit marrow rescue cells. For subsequent secondary transplants, 2%(%106
v4 (Clontech) according to the manufacturer’s recommendations. red-blood-cell-lysed bone marrow cells isolated from primary recipient
The quantity and quality of the subsequent cDNA was determined mice were transplanted into lethally irradiated (9.5%Gy) CD45.1 mice.
using the Qubit Fluorometer (Life Technologies) and the Agilent Bio- Peripheral blood of recipient mice was collected every 4%weeks for
analyser 2100 (Agilent). Then, 150%pg of cDNA was used to generate 4 months after transplant and bone marrow analysed at the end of
Illumina-compatible sequencing libraries using the NexteraXT library 4 months.
preparation kit (Illumina) according to the manufacturer’s protocols.
The amplified libraries were hybridized to the Illumina flow cell and Methylcellulose colony-formation assays
sequenced using the Illumina NextSeq 550 (Illumina). Single-end reads For colony assays with mouse cells, the indicated numbers of Lin−
of 100 nucleotides were generated for each sample. The mouse bulk bcCML cells or KIT+ AML cells were plated in methylcellulose medium
RNA-seq samples were processed otherwise identical to the human (M3234, StemCell Technologies). Colonies were scored at 7%days. For
bulk RNA-seq with two exceptions: m38 + gencode M27 reference colony assays with human cell lines or patient-derived AML samples,
([Link] for use cells were plated in methylcellulose medium (H4434; StemCell Technol-
within alignment and counting, and ‘-s 0’ being used within subread ogies). Colony numbers were counted 10–14 days after plating. Taurine
featureCounts. See the ‘Primary human CD34+ cell RNA-seq analysis’ antagonist or TAG (synthesized by Enamine; 92% pure), GES (G827500,
section for details on differential expression analysis. Toronto Research Chemicals), taurine (T8691, Sigma-Aldrich), mTOR
activator (MHY1485, Sigma-Aldrich) and Venetoclax (ABT-199, Tocris
Primary human patient-derived cells and human leukaemia cell Bioscience) were used as described. Venetoclax and GES Synergy quan-
lines tification was calculated using Chou-Talalay method72.
For human RNA-seq experiments and other studies, CD34+ cells were
isolated from bone marrow samples of healthy donors and samples Seahorse assays
from patients with AML and bcCML obtained under University of The Seahorse XF Glycolysis Stress Test Kit (Agilent Technologies,
Rochester institutional review board-approved protocols with writ- 103020-100) and the Seahorse XF cell mito stress test kit (Agilent Tech-
ten informed consent in accordance with the Declaration of Helsinki. nologies, 103015-100) were used to measure glycolytic flux (ECAR) and
The normal human CD34+ HSPCs used in all functional assays were oxygen consumption (OCR) respectively. Mouse Lin− bcCML cells were
purchased (Stem Cell Technologies). Cells were cultured in Iscove’s sorted and cultured for 48%h in X-Vivo supplemented with 10% FBS, SCF
modified Dulbecco’s medium with 10% FBS, 100%IU%ml−1 penicillin– and TPO. Then, 1%h before the analysis, 50,000–100,000 cells were
streptomycin (Gibco) and 55%µM 2-mercaptoethanol, 1( LDL (Sigma- seeded in Cell-Tak-coated (Corning, 324240) 96-well XF96 well plates
Aldrich) and supplemented ,-glutamine with 100%ng%ml−1 SCF and TPO in Seahorse XF medium (Agilent Technologies, 102353-100), and the
(R&D Systems). Human leukaemia cell lines K562, THP1 and M-V-411 plate was incubated at 37%°C. ECAR data were measured after sequential
Article
addition of glucose (10%mM), oligomycin (1µM) and 2-deoxyglucose were collected over a scan range of 150–2,000%m/z. Phosphoprotein
(50%mM) using the XF96 analyser (Agilent Technologies). OCR data analysis was carried out on the same Vanquish Neo UHPLC and Orbitrap
were measured after sequential addition of oligomycin (1.5%µM), car- Astral mass spectrometer system, but with several changes to the LC
bonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP) (0.5%µM), and instrument settings. To account for ratio compression inherent
rotenone and antimycin (0.5%µM) using the XF96 analyser. Data were with TMT, samples were high-pH fractionated to reduce complexity
analysed using Wave v.2.6.3 (Agilent Technologies). before MS analysis. A longer gradient of 5–30% B over 48%min was
used to further separate phosphopeptides. A data-dependent
Western blot analysis acquisition method using a FAIMS Pro Duo (Thermo Fisher Scientific)
Cell lysates prepared in 1( RIPA (Thermo Fisher Scientific) supple- was used with three compensation voltages (−40%V, −60%V, −80%V) to
mented with 1( protease, 1( phosphatase inhibitors (Cell Signaling further reduce the sample complexity. For each CV, a full scan was
Technology) and 250%IU benzonase nuclease (Millipore Sigma) were acquired over a range of 400–1,500%m/z in the Orbitrap, while MS2
separated on gradient polyacrylamide gels and transferred to nitro- scans were analysed in the Astral analyser for 1%s, after which the
cellulose blotting membrane (0.45%µM; GE Healthcare). Primary anti- instrument switched to the next CV and the process was repeated for
bodies against phosphorylated mTOR, mTOR, phosphorylated S6K, a total cycle time of 3%s. Peptides with a charge state of between 2 and
S6K, pEIF4B, EIF4B and &-actin (Cell Signaling Technology) or tubu- 6 were isolated based on intensity with a 0.5%Da isolation window, and
lin (Abcam) were used. Horse-radish peroxidase (HRP)-conjugated were fragmented with an HCD collision energy of 35%. The maximum
anti-rabbit antibodies (Cell Signaling Technology) were used to detect injection time was 20%ms, and the normalized AGC target was 100%.
primary antibodies. Immunoblots were developed using SuperSignal Dynamic exclusion was set to 15%s.
West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific).
Immunoblots were imaged using the LI-COR Odyssey M system using Data analysis. The global DIA raw data were processed with DIA-NN
Empiria Studio v.2.3 (LI-COR). Images were analysed using Empiria v.1.8.1 ([Link] using library-free analy-
Studio v.[Link] (LI-COR). Raw gel images are provided in Supple- sis mode. The library was annotated using the Mus musculus UniProt
mentary Fig. 1. database (UP000005640_9606). For precursor ion generation, the
maximum number of missed cleavages was set to 1, maximum num-
Protein extraction, sample preparation and MS analysis ber of variable modifications to 1 for Ox(M), peptide length range to
Protein extraction. Cells were lysed in 50%µl of 5% SDS, 100%mM tri- 7–30, precursor charge range to 2–3, precursor m/z range to 380–980,
ethylammonium bicarbonate (TEAB, Thermo Fisher Scientific) and and fragment m/z range to 150–2,000. The quantification was set to
50%µg of protein from each sample was reduced with dithiothrei- ‘Robust LC (high precision)’ mode with RT-dependent median-based
tol (Sigma-Aldrich). Proteins were alkylated and trapped to S-Trap cross-run normalization enabled, MBR enabled, protein inferences
micros (Protifi), digested with trypsin and extracted by sequential set to ‘Genes’ and ‘Heuristic protein inference’ turned off. Precursors
additions of 0.1% trifluoroacetic acid (TFA) in acetonitrile. Then, 1% were filtered at library precursor q-value (1%), library protein group
of each sample was used for global DIA analysis, and the remaining q-value (1%) and posterior error probability (50%). Protein quantifica-
sample was frozen and dried down in the Speed Vac (Labconco) before tion was carried out using the MaxLFQ algorithm ([Link]
TMT labelling. The samples were reconstituted in TEAB and labelled vdemichev/diann-rpackage) and the number of peptides quantified in
with TMT 10-plex reagents (Thermo Fisher Scientific) according to each protein group was determined using the DiannReportGenerator
the manufacturer’s protocol. All of the samples were combined and Package ([Link]
dried down in a speed vac before desalting with 0.1% TFA in acetoni- Further filtering, missing value imputation and statistical tests were
trile using the 130%mg C18 sorbent sep-pak attached to a 3%ml syringe performed using Perseus73. Phosphoproteome raw data were searched
(Waters). Desalted samples were frozen and dried down before phos- using the CHIMERYS within the Proteome Discoverer software platform
phorylation enrichment using the High-Select FeNTA Enrichment v.3.1 (Thermo Fisher Scientific), allowing for up to two missed cleav-
Kit (Thermo Fisher Scientific) according to the manufacturer’s pro- ages, with a fragment mass tolerance of 10%ppm. Carbamidomethyl on
tocol. Enriched phosphorylated samples were frozen, dried down cysteine, and TMT on lysine and peptide N terminus were set as fixed
and fractionated using C18 spin columns. The fractions were eluted modifications, while oxidation of methionine and phosphorylation
by stepwise addition of 10%mM AmmOH with increasing acetonitrile of serine, threonine and tyrosine were set as variable modifications.
concentrations: 3.5, 6.5, 9.5, 12.5, 15.5, 18.5, 27, 50%. The eight fractions Reporter ions were quantified using the Reporter Ions Quantifier node,
were concatenated to four by combining fractions 1/5, 2/6, 3/7, 4/8. with an integration tolerance of 20%ppm, and the integration method
Fractionated samples were frozen, dried down and reconstituted in was set to ‘most confident centroid’.
0.1% TFA for MS analysis.
Immunohistochemical staining and analysis
MS analysis. Non-TMT-tagged peptides were injected onto a 75%µm%(% Paraffin-embedded 4%µm human bone marrow biopsy sections
2%cm trap column before refocusing on a 100%µm%(%15%cm C18 column were deparaffinized in xylene and antigen epitopes were retrieved
with 1.8%µm beads (Sepax) using the Vanquish Neo UHPLC (Thermo Fish- using BOND epitope retrieval solution (pH%9) for 10%min. Endo-
er Scientific) system connected to the Orbitrap Astral mass spectrom- genous peroxidases were quenched with a 10%min incubation in 3%
eter (Thermo Fisher Scientific). Ions were introduced to the mass spec- H2O2-methanol solution. The sections were then blocked in 10%
trometer using a Nanospray Flex source operating at 2%kV. Solvent A donkey serum for 60%min. The sections were either incubated with
(0.1% formic acid in water) and solvent B (0.1% formic acid in 80% ace- anti-CDO1 (Proteintech) overnight or with anti-osterix (Abcam) for
tonitrile) formed the gradient starting at 1% B and ramped to 99% B 2%h at room-temperature. The sections were washed in PBS and stained
for a total runtime of 14%m. After each run was completed, the column with HRP-conjugated donkey anti-rabbit secondary antibody ( Jackson
was re-equilibrated with 1% B before the next injection. The Orbitrap Immunoresearch) for 1.5%h at room temperature. After three washes,
Astral was operated in data-independent acquisition (DIA) mode for colour was developed using the ImmPACT DAB Substrate kit (Vector
global proteomic analysis, with MS1 scans acquired at a resolution laboratories, SK-4105) according to the manufacturer’s protocol. The
of 240,000 with a maximum injection time of 5%ms over a range of sections were then counterstained with haematoxylin. Three different
390–980%m/z. DIA MS2 scans were acquired using 2%Da windows, a areas of each section were imaged on the Olympus BX41 microscope
3%ms maximum injection time, an HCD collision energy of 25% and with a (20 (0.5 NA) objective. Images were analysed using the IHC plugin
a normalized automatic gain control (AGC) of 500%. Fragment ions toolbox in Fiji v.1.54g.
 balanced salt solution (Gibco) with 5% FBS and 2%mM EDTA at 4%°C.
Immunofluorescence staining Leukaemia cells were collected and maintained at 4%°C all through the
Leukaemia cells were cultured in RPMI without amino acids (US Bio- process of isolation, staining and magnetic sorting to minimize meta-
logicals) in the presence or absence of taurine (Sigma-Aldrich) and bolic changes. Lin+ (CD3*−CD4−CD8−GR1−CD11b−TER119−CD45R−CD19−)
seeded on eight-well cover glass chambers (BD biosciences) coated with leukaemia cells were magnetically depleted using LD columns (Milteny
Cell-Taq (Corning) according to the manufacturer’s protocols. MSCs Biotec). Lin− leukaemia stem cell fractions were washed with PBS con-
were grown and differentiated in 35%mm glass-bottom dishes with 14%mm taining 5%mM glucose and centrifuged at 3,000g for 1%min, snap-frozen
microwell (MatTek Life Sciences). Cells were fixed with 4% PFA and and processed for metabolomics as described in the ‘LC–MS analysis’
blocked in blocking buffer (PBS with 5% donkey serum, 1% BSA and 0.1% section below.
Triton X-100) and incubated overnight in primary antibodies diluted
13
in the blocking buffer. Primary antibodies used included mTOR (Cell C-taurine tracing in leukaemia cells
Signaling Technologies), LAMP1 (DHSB) or CDO1 (Proteintech). Cells K562 cells were cultured in serum-free RPMI-1640 medium in six-well
were washed in PBS containing 0.1% Tween-20 (Sigma-Aldrich), stained plates for 48%h. The cells were then cultured in serum free RPMI-1640 sup-
with Alexa-Fluor-conjugated secondary antibodies (Thermo Fisher plemented with 200%µM taurine (1,2 13C2, 98%; Cambridge Isotope Labs)
Scientific), and mounted in Fluormount G (Thermo Fisher Scientific). or no additional taurine for 24%h. The cells were washed with PBS contain-
ing 5%mM glucose and centrifuged at 3,000g, snap-frozen and processed
Immunofluorescence imaging and analysis for metabolomics as described below in the ‘LC–MS analysis’ section.
Immunofluorescence images were acquired with the Teledyne Pho-
tometrics Prime BSI express sCMOS camera mounted on the Nikon LC–MS analysis
ECLIPSE Ti2 inverted microscope equipped with the NIS-Elements 6D Frozen cell pellets were resuspended at 2%million cells per 1%ml of 80%
imaging acquisition module (v.5.42.06). The Nikon D-LEDI fluorescence methanol by vortexing, transferred to −80%°C for 30%min and then to reg-
LED illumination system (equipped with 385%nm, 488%nm, 568%nm and ular ice for 30%min with vortexing every 10%min. Next, the samples were
621%nm excitation wavelengths) was used as the primary illumination centrifuged at 17,000g for 10%min and 90% of supernatant was dried
source. Specific illumination wavelengths were selected by combining a down in a vacuum evaporator (Thermo Fisher Scientific). The samples
large field of view quad-filter cube (DAPI/FITC/TRITC/CY5; 96378) with were reconstituted in 50% acetonitrile (A955, Thermo Fisher Scientific)
specific Lumencor emission filters (FF01-474/27-32, FF01-515/30-32, at a volume equal to 10% of the dried down volume and transferred to
FF01-595/31-32, FF02-641/75-32). MSC and osteolineage cultures were glass vials for LC–MS analysis. The metabolite extracts were analysed
imaged with the Nikon CFI60 Plan Apochromat Lambda D (20 (0.8 NA) by high-resolution MS with the Orbitrap Exploris 240 (Thermo Fisher
objective lens, and leukaemia cells were imaged using the Nikon CFI60 Scientific) system coupled to the Vanquish Flex LC system (Thermo
Plan Apochromat Lambda D (100 (1.45 NA) objective lens. Images were Fisher Scientific). Then, 2%µl of the samples was injected onto the Waters
deconvoluted using Imaris v.10.2 (Oxford instruments), and Pearson’s XBridge XP BEH Amide column (150%mm length%(%2.1%mm inner diam-
co-localization analysis was performed using the JACoP BIOP plugin eter, 2.5%µm particle size) maintained at 25%°C, with a Waters XBridge XP
in Fiji v.1.54g. VanGuard BEH Amide (5%mm ( 2.1%mm inner diameter, 2.5%µm particle
size) guard column. For positive-mode acquisition, mobile phase A was
Phosphoflow cytometry 100% LC–MS-grade H2O with 10%mM ammonium formate and 0.125%
Lineage depleted Slc6a6+/+ and Slc6a6−/− mouse leukaemia cells were formic acid. Mobile phase B was 90% acetonitrile with 10%mM ammo-
fixed using BD Cytofix/Cytoperm Fixation/Permeabilization Kit (BD nium formate and 0.125% formic acid. For negative-mode acquisition,
Bioscience) according to the manufacturer’s protocols. Cells were mobile phase A was 100% LC–MS-grade H2O with 10%mM ammonium
stained with primary antibodies against phosphorylated mTOR (Cell acetate, 0.1% ammonium hydroxide and 0.1% medronic acid (Agilent).
Signaling Technology). Cells were then stained with donkey anti-rabbit Mobile phase B was 90% acetonitrile with 10%mM ammonium acetate,
secondary antibody conjugated with Alexa Fluro 488 (Invitrogen) to 0.1% ammonium hydroxide and 0.1% medronic acid. The gradient was
detect p-mTOR. Analysis was performed on the LSRFortessa (BD Bio- 0%min, 100% B; 2%min, 100% B; 3%min, 90% B; 5%min, 90% B; 6%min, 85% B;
sciences) system. Data were analysed using FlowJo software. 7%min, 85% B; 8%min, 75% B; 9%min, 75% B; 10%min, 55% B; 12%min, 55% B;
13%min, 35% B; 20%min, 35% B; 20.1%min, 35% B; 20.6%min, 100% B; 22.2%min,
Taurine quantification 100% B; all at a flow rate of 150%µl%min−1, followed by 22.7%min, 100% B;
Bone marrow cells or bone marrow peripheral fluid were isolated from 27.9%min, 100% B at a flow rate of 300%µl%min−1, and finally 28%min, 100% B
femurs in Hanks’ balanced salt solution (Gibco) with 5% FBS and 2%mM at flow rate of 150%µl%min−1, for a total length of 28%min. The H-ESI source
EDTA. For taurine analysis of bone marrow cells after genetic loss of was operated in positive mode at spray voltage 3,500%V or negative
Slc6a6 or treatment of leukaemia cells with taurine inhibitor treat- mode at spray voltage 2,500%V with the following parameters: sheath
ments, cells were lysed in RIPA buffer (Thermo Fisher Scientific) with gas 35 au, aux gas 7 au, sweep gas 0 au, ion transfer tube temperature
benzonase nuclease (Sigma-Aldrich). For bone marrow interstitial fluid 320%°C, vaporizer temperature 275%°C, mass range 70 to 1,000%m/z,
analysis, one femur was crushed in 1%ml of buffer, filtered and centri- full scan MS1 mass resolution of 120,000 FWHM, RF lens at 70% and
fuged at 1,500g for 5%min. The supernatant was concentrated using standard AGC. LC–MS data were analysed using Compound Discover
10,000 MWCO spin columns (Corning). Then, 25%µl of concentrated (v.3.3, Thermo Fisher Scientific) and El-Maven software74 for peak-area
samples was quantified using the Taurine Assay kit (Sigma-Aldrich/ determination and compound identification. Compounds were identi-
Abcam) according to the manufacturer’s protocols on the BioTek fied by matching to LC–MS method-specific retention time values of
Synergy 2 plate reader using Gen5 v.3.11 (BioTek). The samples were external standards and MS2 spectral matching to external standards and
corrected for taurine amounts in unconditioned fresh medium or the mzCloud database (Thermo Fisher Scientific). Raw P values were
buffer. Alternatively, cell pellets were processed as described in the calculated using pairwise Mann–Whitney–Wilcoxon rank-sum tests and
liquid chromatography–mass spectrometry (LC–MS) section below, Padj values were computed using Benjamini–Hochberg false-discovery
and taurine levels were measured by LC–MS (Orbitrap Exploris 240). rate correction. Data were uploaded to the Metabolomics Workbench75.

Untargeted metabolomics of Lin− leukaemia cells Statistical analysis

For untargeted metabolomics, spleens from mice bearing Slc6a6+/+ or Statistical analyses were performed using GraphPad Prism software
Slc6a6−/− leukaemias were quickly dissected and dissociated in Hanks’ v.6.0 (GraphPad). Data are mean%±%s.e.m. One-way ANOVA, two-way
Article
ANOVA, unpaired two-sided Student’s t-tests, multiple unpaired t-tests ([Link] and was gener-
corrected with the Benjamin–Hochberg method, ratio-paired t-tests, ated using ShinyCell (v.2.1; [Link]
and log-rank tests were used to determine statistical significance. Com-
bination index and isobologram plots were made using CompuSyn72. 61. Kaszuba, C. M. et al. Identifying bone marrow microenvironmental populations in
myelodysplastic syndrome and acute myeloid leukemia. J. Vis. Exp. [Link]
10.3791/66093 (2023).
Reporting summary 62. Browaeys, R., Saelens, W. & Saeys, Y. NicheNet: modeling intercellular communication by
Further information on research design is available in the Nature Port- linking ligands to target genes. Nat. Methods 17, 159–162 (2020).
folio Reporting Summary linked to this article. 63. Fujisawa, H. Discovery of semaphorin receptors, neuropilin and plexin, and their functions
in neural development. J. Neurobiol 59, 24–33 (2004).
64. Jurecka, A., Zikanova, M., Kmoch, S. & Tylki-Szymańska, A. Adenylosuccinate lyase
deficiency. J. Inherit. Metab. Dis. 38, 231–242 (2015).
Data availability 65. Sowa, N. A., Taylor-Blake, B. & Zylka, M. J. Ecto-5′-nucleotidase (CD73) inhibits nociception
by hydrolyzing AMP to adenosine in nociceptive circuits. J. Neurosci. 30, 2235–2244
The scRNA-seq and bulk RNA-seq data for this publication are available (2010).
under GEO accessions GSE226372 (human bulk RNA-seq), GSE227082 66. Kang, Y. P. et al. Cysteine dioxygenase 1 is a metabolic liability for non-small cell lung
cancer. eLife 8, e45572 (2019).
(mouse bulk RNA-seq), GSE226644 (mouse temporal scRNA-seq) and
67. Mahootchi, E. et al. GADL1 is a multifunctional decarboxylase with tissue-specific roles in
GSE288862 (scRNA-seq of human MDS and AML bone marrow micro- &-alanine and carnosine production. Sci. Adv. 6, eabb3713 (2020).
environment). A shiny app hosting the mouse temporal scRNA-seq is 68. Teufel, M. et al. Sequence identification and characterization of human carnosinase and a
closely related non-specific dipeptidase. J. Biol. Chem. 278, 6521–6531 (2003).
available online ([Link]
69. Amyere, M. et al. KITLG mutations cause familial progressive hyper- and hypopigmentation.
Receptor and ligand interactions were determined using the NicheNet62, J. Invest. Dermatol. 131, 1234–1239 (2011).
Cell Surface Protein atlas19, AML and healthy immune microenviron- 70. Fuchs, A., Cella, M., Giurisato, E., Shaw, A. S. & Colonna, M. Cutting edge: CD96 (tactile)
promotes NK cell-target cell adhesion by interacting with the poliovirus receptor (CD155).
ment20 datasets. The link between SLC6A6 and LDLR expression and
J. Immunol. 172, 3994–3998 (2004).
AML prognosis was determined using UCSC Xena Browser (https:// 71. Läubli, H. et al. Lectin galactoside-binding soluble 3 binding protein (LGALS3BP) is a
[Link]/). Wild-type and Slc6a6−/− proteomics data are avail- tumor-associated immunomodulatory ligand for CD33-related Siglecs. J. Biol. Chem.
289, 33481–33491 (2014).
able at ProteomeXchange (PXD062322). The library was annotated
72. Chou, T. C. Drug combination studies and their synergy quantification using the
using M. musculus UniProt ‘one protein sequence per gene’ database Chou-Talalay method. Cancer Res. 70, 440–446 (2010).
(UP000005640_9606, downloaded April 2021). WT and Slc6a6−/− 73. Tyanova, S. et al. The Perseus computational platform for comprehensive analysis of
(prote)omics data. Nat. Methods 13, 731–740 (2016).
metabolomics data (ST003835) and 13C-taurine tracing data (ST003836) 74. Melamud, E., Vastag, L. & Rabinowitz, J. D. Metabolomic analysis and visualization engine
are available online. Source data are provided with this paper. for LC-MS data. Anal. Chem. 82, 9818–9826 (2010).
75. Sud, M. et al. Metabolomics Workbench: an international repository for metabolomics
data and metadata, metabolite standards, protocols, tutorials and training, and analysis
tools. Nucleic Acids Res. 44, D463–D470 (2016).
Code availability
Mouse and human scRNA datasets were processed using Seurat (v.4.1.0;
[Link] within R Acknowledgements We thank H. Land for scientific discussions and advice; J. Tran for
technical support; K. Tohyama for MDS-L cells; W. Pear and A. M. Pendergast for the BCR-ABL
(v.4.1.1; [Link] Time-course construct; D.G. Gilliland for the NUP98-HOXA9 construct; C. Counter for the NRAS(G12V)
analysis was performed using DEGreport (v.1.30.3; [Link] construct; S. Armstrong for the MLL-AF9 construct; S. Lowe for the AML-ETO9a construct;
[Link]/DEGreport/[Link]). Pathway enrichment was performed the staff at the Cytometry, Genomics, Imaging and Radiation, and Metabolomics Shared
Resources supported in part by University of Rochester Wilmot Cancer Institute Support Grant
using EnrichR (v.3.0; [Link] P30CA272302; the staff at the URMC Histology, Biochemistry and Molecular Imaging facility
chR/). Sample integration within human scRNA datasets was performed supported by P30AR069655; and the members of the Metabolomics Workbench/National
using Harmony (v.0.1.0; [Link] Metabolomics Data Repository supported by U2C-DK119886 and OT2-OD030544. Open access
funding was provided by Wilmot Cancer Institute. C.M.K. was supported by NIH Training Grant
mony). Integrating external datasets into the human scRNA dataset was T32AR076950; and C.T.J. by the Nancy Carroll Allen Chair in Hematology Research and
done using Seurat (v.5.0.3.99911; [Link] a Leukemia and Lymphoma Society SCOR grant (7033-24). J.B. is a recipient of American
Society of Hematology Scholar Award, and Leukemia Research Foundation New Investigator
releases/tag/v5.0.3) and Azimuth (v.0.5.0; [Link]
Award. This work was supported by NIH grants R01AG079556 awarded to L.M.C. and J.L.L.;
jalab/azimuth) within R (v.4.3.1; [Link] R35CA242376 to C.T.J.; and R01DK133131 and R01CA266617 to J.B.
tag/v4.3.1). Mouse and human bulk RNA datasets were trimmed and
quality-filtered using FastP (v.0.20.1; [Link] Author contributions S.S. performed most of the cell culture experiments and the
biochemical analyses. B.J.R. and C.M.K. carried out the in vivo leukaemia progression
fastp/releases/tag/v0.20.1). Read data were aligned using STAR experiments. C.D.B. performed all of the mouse and human microenvironment scRNA-seq
(v.2.7.6a; [Link] analyses. E.I.F. provided experimental data and performed image analysis. B.R.S., P.S.B. and
J.C.M. carried out the metabolomic analysis and provided experimental advice. T.I. provided
and counted using subread-featureCounts (v.2.0.1; [Link]
the Slc6a6-knockout mice. K.S., K.W. and S.G. performed the proteomic analyses. Y.Z., P.R. and
[Link]/). Differential expression analysis was performed F.A.C. helped to process the patient samples. L.M.C., M.W.B., W.R.B. and J.L.L. provided the
using DESeq2 (v.1.28.1; [Link] patient samples. A.G. carried out the imaging experiments. J.M.A. and C.T.J. performed the
human leukaemia sequencing experiments and analyses. J.B. conceived of the project,
bioc/html/[Link]) within R (v.4.0.2; [Link]
planned and guided the research, and wrote the manuscript.
releases/tag/v4.0.2). Pathway enrichment analysis within the bulk RNA
datasets was performed using EnrichR (v.3.0, above). Ligand–receptor Competing interests The authors declare no competing interests.
visualization was performed using NicheNetR (v.1.1.0; [Link]
Additional information
com/saeyslab/nichenetr) and Circlize (v.0.4.15; [Link] Supplementary information The online version contains supplementary material available at
ergoo/circlize) within R (v.4.1.1, above). Ggplot2 ([Link] [Link]
tidyverse/ggplot2) and dplyr ([Link] Correspondence and requests for materials should be addressed to Jeevisha Bajaj.
Peer review information Nature thanks Peter van Galen and the other, anonymous, reviewer(s)
were used for figure generation and data manipulation throughout. for their contribution to the peer review of this work.
A shiny app hosting the mouse temporal scRNA-seq is available online Reprints and permissions information is available at [Link]
Extended Data Fig. 1 | Temporal changes in leukaemia bone marrow temporal changes in the proportion of all major lineages, and in sub-clusters
microenvironment. a, Survival curve shows bcCML progression in of chondrocyte and fibroblasts. e, Representative FACS plots and graph show
unirradiated recipients, indicating time course for scRNA-seq experiment changes in BM sinusoidal and arteriolar endothelial cell frequency over time
(naïve, initiation, expansion, and end). b, Representative FACS plots and (mean%±%s.e.m.; n%=%3 animals per timepoint, two-way ANOVA). f, Gene clusters
graph show changes in leukaemia engraftment over time (2 or 3 mice pooled associated with changes in MSC, chondrocyte, and arteriolar endothelial
per timepoint). c, Heatmap of significantly expressed marker genes for populations during disease progression.
non-haematopoietic Bone Marrow (BM) populations. d, Line graphs show
Article

Extended Data Fig. 2 | Signals from bone marrow niche interacting with expressed on cell surface19, and those that drop-out by 2-fold or more in the
human LSC cell surface receptors. a, Unbiased Enrichr analysis shows top 10 leukaemia in vivo CRISPR screen6. d, Heatmap shows r-log normalized RNA
upregulated pathways by population cluster in MSCs, osteo-associated, expression of cell surface receptors upregulated in AML, bcCML or both.
arteriolar, and sinusoidal endothelial populations. b, PCA-plot shows the e, Average expression of ligands for cell surface proteins in primary human
distribution of 7 CD34+ healthy donor BM HSPCs, 10 bcCML CD34+ LSCs, and 11 adult and paediatric AML cells, human AML immune microenvironment
AML CD34+ LSCs from human samples. c, Overlap between genes upregulated populations, and in normal cells20.
in bcCML and AML CD34+ cells compared to normal CD34+ cells, proteins
Extended Data Fig. 3 | Temporal changes in niche-driven signals. a-d, UMAP plot of indicated gene expression in microenvironmental populations over time
(naïve: 0%d, initiation: 2 and 4 days, expansion: 7 and 9 days and, end:11 and 14 days post-transplant).
Article

Extended Data Fig. 4 | CDO1 expression in human leukaemia g, Slc6a6 expression in KLS cells infected with BCR-ABL and NUP98-HOXA9
microenvironment. a, Kaplan-Meier curves of human leukaemia patients (BA-NH) oncogenes for 48%h (mean%±%s.d.; n%=%3 technical replicates per cohort).
with high (<9.32, n%=%80) or low (≥9.32, n%=%81) LDLR expression (TCGA-LAML; h, Taurine biosynthesis pathway. i, Dot plot of Cdo1 and Csad expression during
Xena Browser; log-rank test). b, Normalized LDLR expression in CD34+ cells disease progression. j, UMAP plot of CDO1 and CSAD expression in normal
from human bcCML, AML, or normal BM (n%=%7 normal BM, n%=%10 bcCML, n%=%11 human BM23. k, UMAP plot of microenvironmental cells in three human MDS
AML; central line, box, and whiskers represent median, interquartile range or or AML BM aspirates. l, UMAP plot of CDO1 and CSAD in human MDS and AML
IQR, minimum/maximum within 1.5*IQR respectively; DESeq2 implemented BM aspirates. m-p, Representative IHC images and quantification of Osterix
Wald test). c, d, Experimental strategy (c) and relative Apoe expression (d) in expression in matched human BM biopsies at MDS diagnosis and AML
MSCs transduced with shRNAs targeting LacZ or Apoe (mean%±%s.d.; n%=%3 technical transformation (m, n) or AML diagnosis and relapse (o,p) (n%=%5 independent
replicates per cohort). e,f, Number of live leukaemia cells (e) and their colony patients per cohort; each colour represents a patient sample; two-tailed ratio
forming ability (CFU) (f) post co-culture with MSCs (mean%±%s.e.m.; e, n%=%8 paired t-test). The mouse image in c is adapted from ref. 6, Springer Nature
independent culture wells; f, n%=%6 independent culture wells, data combined America.
from two independent experiments; unpaired two-tailed Student’s t-test).
Extended Data Fig. 5 | Inhibiting taurine synthesis in osteolineage cells expression (k) in Cdo1 fl/fl and Cdo1 fl/fl;Prrx1-Cre + MSCs following osteogenic
impairs LSC growth. a, b, Alizarin red staining (a), and Cdo1 and Csad expression (b) differentiation (mean%±%s.d.; n%=%3 technical replicates per cohort). l-n, BM
in MSCs undergoing osteogenic differentiation (mean%±%s.d.; n%=%3 technical stroma in leukaemic control and Cdo1 fl/fl;Prrx1-Cre+ mice: MSCs (l), osteolineage
replicates per cohort). c, CDO1 expression in murine leukaemia and patient cells (m), and, endothelial cells (n) (mean%±%s.e.m.; n%=%7 Cdo1 fl/fl and n%=%3 Cdo1 fl/fl;
AML MSCs undergoing osteogenic differentiation (mean%±%s.e.m.; n%=%3 replicates Prrx1-Cre +; data combined from three independent experiments; unpaired
per time point). d, Taurine in MSC culture media during osteogenic differentiation two-tailed Student’s t-test). o-q, CFU of murine cKit+ AML cells (o), murine
(mean%±%s.e.m.; n%=%3 replicates; secreted over 48–72%h). e,f, Experimental Lin− LSCs (p), or primary human AML cells (q) with taurine (mean%±%s.e.m.; n%=%3
strategy (e) and Cdo1 expression (f) in MSCs transduced with shRNAs targeting independent culture wells from n%=%2 murine AML, n%=%2 murine bcCML, and
LacZ or Cdo1 (mean%±%s.d.; n%=%3 technical replicates per cohort). g, Live leukaemia n%=%3 primary human AML samples; data combined from two independent
cells (left), and CFU (right), post coculture with bcCML MSCs (mean%±%s.e.m.; experiments; one-way ANOVA). r, Impact of taurine supplements on murine
n%=%7 independent culture wells per cohort (live); n%=%6 (CFU); data combined bcCML progression in unirradiated recipients (n%=%14 no taurine and 0.01%mg/ml
from 2 independent experiments; one-way ANOVA). h, CDO1 expression in taurine, n%=%15 10%mg/ml taurine; data combined from three independent
293%T transduced with shRNAs targeting LACZ or CDO1 (mean%±%s.d.; n%=%3 experiments; log-rank test). The mouse image in e is adapted from ref. 6, Springer
technical replicates per cohort). i, Cdo1 fl/fl model. j,k, PCR of Cdo1 ( j) and Cdo1 Nature America.
Article

Extended Data Fig. 6 | Impact of TAUT loss on leukaemia development from two independent experiments). d-g, Graph (d) shows number of
in murine models. a, Number of live LSCs (left) and their CFU (right) post microenvironmental cells per mouse leg. Representative FACS plots and
co-culture with leukaemic MSCs transduced with shCdo1 or shLacZ (mean%±% quantification of MSC (e), osteolineage (f), and endothelial (g) frequency (left)
s.e.m.; n%=%6 independent culture wells; data combined from 2 independent and total cell number (right) at initiation and end point (mean%±%s.e.m.;
experiments; one-way ANOVA). b, Survival curve shows the impact of taurine n%=%3 +/+ and n%=%5 −/−; data combined from two independent experiments).
supplements on murine leukaemia progression in irradiated recipients (n%=%10 h, Representative FACS histogram and quantification of Lin+ frequency in
per cohort; data combined from two independent experiments; log-rank test). bcCML BM (mean%±%s.e.m.; n%=%14 animals per cohort; data combined from
c, Representative FACS plots and quantification of BM engraftment in cancer at three independent experiments; c, e-h unpaired two-tailed Student’s t-test).
initiation and end point (mean%±%s.e.m.; n%=%3 +/+ and n%=%5 −/−; data combined
Extended Data Fig. 7 | The role of TAUT in normal haematopoietic stem cell combined from two independent experiments). f-g, Experimental strategy (f),
function. a, Average number of BM cells in +/+ and −/− mice (mean%±%s.e.m.; donor engraftment in peripheral blood over time (g). h-k, Average donor
n%=%5 animals per cohort). b-d Representative FACS plots, and quantification of chimerism (h), frequency of KLS (Lin−cKit+Sca+), HSCs, and MPPs (i), committed
frequency (left) and cell number (right) of hematopoietic stem cells (b, HSC: Lin− progenitors GMP, CMP, and MEP ( j) and, differentiated haematopoietic cells (k)
cKit+Sca+CD150+CD48−) and multipotent progenitors (b, MPPs: Lin−cKit+Sca+ in primary recipient BM four months post-transplant (f-k: mean%±%s.e.m.; n%=%9
CD150 −CD48−), committed granulocyte–macrophage progenitors (c, GMP: Lin− +/+ and n%=%10 −/−; data combined from two independent experiments). l, Donor
IL7Ra−Kit+Sca1−CD34+CD16/3+), common myeloid progenitors (c, CMP: Lin−IL7Ra− engraftment in peripheral blood of secondary recipients (mean%±%s.e.m.; n%=%6
Kit+Sca1−CD34+CD16/32−), megakaryocyte–erythroid progenitors (c, MEP: Lin− animals per cohort; g, l, two-way ANOVA). m-q, Average donor chimerism (m),
IL7Ra−Kit+Sca1−CD34−CD16/32−), and differentiated haematopoietic cells (d) frequency of KLS, HSCs, and MPPs (n), committed progenitors GMP, CMP, and
in the BM of +/+ and −/− mice (mean%±%s.e.m.; n%=%5 animals per cohort; data MEP (o), and differentiated haematopoietic cells (p, q) in BM of secondary
combined from four independent experiments). e, Complete blood count of recipients four months post HSC transplant (mean%±%s.e.m.; n%=%6 animals per
indicated cells and haemoglobin content in the peripheral blood of age and cohort; a-e, h-k, m-q unpaired two-tailed Student’s t-test). The mouse images
sex-matched 8-week-old littermates (mean%±%s.e.m.; n%=%11 +/+ and n%=%13 −/−; data in f are adapted from ref. 6, Springer Nature America.
Article

Extended Data Fig. 8 | Taurine inhibition synergizes with venetoclax culture wells per cohort; data combined from two independent experiments
in vitro. a, SLC6A6 expression in human AML stem cells, progenitors and for f,h). i, j, Viability (i) and CFU ( j) of murine Lin− bcCML cells treated with
mature blasts (Gene Expression Commons). b, SLC6A6 expression by karyotype 0.5%mM venetoclax (ABT-199, Ven) for 48%h (n%=%3 independent culture wells per
or FAB subtype (TCGA-AML, n%=%146 karyotype and n%=%149 FAB subtype; dot cohort; data combined from two (i) or three ( j) independent experiments;
represents an individual patient). c, SLC6A6 expression in relapse origin- two-way ANOVA). k, CFU of Lin− bcCML cells with DMSO/water, Ven, or indicated
committed (ROC) and relapse origin-primitive (ROP) AML 34 (n%=%29 ROC and doses of TAG (left) and GES (right), or indicated combinations (mean ± s.e.m.;
n%=%10 ROP); (b-c central line, box, and whiskers represent median, interquartile n%=%3 per independent culture wells; data combined from 2 or 3 independent
range, minimum/maximum point within 1.5*IQR respectively; unpaired two- experiments). l, CFU of cKit+ AML LSC with DMSO/water, Ven, GES, or indicated
samples Wilcoxon test). d, Taurine in Lin− bcCML cells treated with DMSO or combinations (mean±s.e.m.; n%=%3 independent culture wells; data combined
TAG (mean±s.e.m.; n%=%8 independent wells combined from 2 experiments). from 2 independent experiments, f-h, j-l one-way ANOVA). m, Combination
e, Taurine in Lin− bcCML cells treated with GES or water (mean±s.e.m.; n%=%5 index of GES and Ven per fraction affected (left) and isobologram (right) by
independent culture wells; d-e unpaired two tailed Student’s t-test). f-h, CFU of Chou-Talalay method (n%=%3 independent culture wells; data combined from
Lin− bcCML (f), ckit+ AML LSC (g) with DMSO or indicated doses of TAG, or Lin− 3 independent experiments; colours represent sample type, shapes represent
bcCML with water or indicated doses of GES (h)(mean±s.e.m.; n%=%3 independent Ven/GES combinations).
Extended Data Fig. 9 | Taurine inhibition in vivo. a, Survival curve showing of indicated tissues in treated animals. j, Intracellular taurine levels by LC/MS
the impact of treating mice transplanted with Lin− bcCML cells with GES (2.5%), in BM cells from treated animals (mean%±%s.e.m.; n%=%8 independent replicates
Ven (50%mg/kg), or their combination (n%=%10 control, n%=%12 GES, n%=%6 Ven, n%=%8 from n%=%3 control, n%=%10 from 4 GES, n%=%9 from 3 TAG treated animals; each
GES/Ven); data combined from three independent experiments; lines below colour represents a mouse). k, l, SLC6A6 expression (k) and taurine levels (l) in
graph represent days of treatment; log-rank test). b, Taurine in 107 BM cells K562 cells transduced with shRNAs targeting LacZ and SLC6A6 (mean%±%s.d.;
from leukaemic mice treated with indicated amounts of GES for 13–16%d n%=%3 technical replicates per cohort). m-o, CFU of human leukaemia cell lines
(mean%±%s.e.m.; n%=%3 technical replicates per cohort). c, d, Experimental transduced with lentiviral shRNAs targeting LacZ or SLC6A6, K562 (m), THP1 (n),
strategy (c) and mass (d) of mice during treatment with GES, TAG, or control and MV-4-11 (o) (mean%±%s.d.; n%=%3 independent culture wells per cohort).
(mean%±%s.e.m.; n%=%3 animals per cohort; two-way ANOVA). e,f, Viability (e) p, Relative SLC6A6 expression (left; mean%±%s.d.; n%=%3 technical replicates per
and average number of BM cells (f) (mean%±%s.e.m.; n%=%6 mice per cohort; data cohort) and CFU (right) in MDS-L cells transduced with shRNAs targeting Luc
combined from two independent experiments). g-h, Representative FACS plots and SLC6A6 (mean%±%s.e.m.; n%=%9 independent culture wells per cohort; data
(left), and number (right) of HSC and MPP (g), and differentiated haematopoietic combined from 3 independent experiments; b, e-h, j, l-p one-way ANOVA).
cells (h) (mean%±%s.e.m.; n%=%6 animals per cohort; data combined from two The mouse image in c is adapted from ref. 6, Springer Nature America.
independent experiments; two-way ANOVA). i, Haematoxylin & Eosin staining
Article

Extended Data Fig. 10 | TAUT loss impairs energy metabolism in myeloid lactate (d), or indicated amounts of glucagon (e) on the CFU of Lin− bcCML cells
leukaemia. a, Quantification of taurine and glycolysis associated metabolites (mean%±%s.e.m.; n%=%3 independent culture wells per cohort; data combined from
in +/+ and −/− Lin− bcCML cells (mean%±%s.e.m.; n%=%16 samples from n%=%6 +/+ 2 independent experiments; one-way ANOVA). f, Experimental strategy used
leukaemic mice and n%=%9 samples from n%=%3 −/− leukaemic mice; unpaired to determine 13C taurine tracing in K562 cells by untargeted metabolomics.
two-tailed Student’s t-test with Welch’s correction). b, Curve and quantification g, Incorporation of 13C label from taurine in indicated metabolites (mean%±%s.e.m.;
of normalized of oxygen consumption rate (OCR) in Lin− bcCML cells (mean%±% n%=%5 independent biological replicates). h,i, Colony-forming ability of Lin− bcCML
s.e.m.; n%=%10 independent culture wells per cohort from 3 bcCML samples; cells in the presence of 160 mM N-acetyl taurine (h) or indicated amounts of
data combined from three independent experiments; unpaired two-tailed glutaurine (i) (mean%±%s.e.m.; n%=%3 independent culture wells per cohort;
Student’s t-test). c-e, Impact of supplementing 5%mM sod. acetate (c), data combined from two independent experiments; one-way ANOVA).
Extended Data Fig. 11 | Transcriptomic and proteomic analysis of n%=%4 independent disease end point samples/cohort). f, Volcano plot of global
leukaemia in the absence of TAUT. a, Top 20 upregulated pathways in −/− time proteome (upper panel) and phosphoproteome (bottom panel) with differential
matched bcCML cells as compared to +/+ controls. b, Heatmap of glycolysis abundance in +/+ and −/− leukaemia cells from time matched recipients (n%=%5
associated genes in bcCML samples. c, GSEA showing downregulation of genes animals per cohort). g, Normalized abundance of mTOR pathway proteins
associated with Myc and mTORC1 pathways on TauT loss (a-c, n%=%3 animals per (mean%±%s.e.m.; n%=%5 animals per cohort; unpaired two-tailed Student’s t-test).
cohort). d, Relative expression of glycolysis associated genes in +/+ and −/− Lin− h, Expression of glycolysis related genes in +/+ Lin− bcCML treated with indicated
bcCML cells (mean%±%s.d.; n%=%3 technical replicates per cohort). e, Western blot doses of TAG for 72%h (mean%±%s.d.; n%=%3 technical replicates per cohort).
shows phospho-mTOR (pmTOR), mTOR, phosopho-p70S6k (p-p70S6K), p70S6K, i, Expression of glycolysis related genes in +/+ Lin− bcCML treated with 1%mM
and actin protein expression in +/+ and −/− Lin− bcCML cells (also see Fig. 5k–m, GES for 72%h (mean%±%s.d.; n%=%3 technical replicates per cohort).
Article

Extended Data Fig. 12 | Taurine activates mTOR to promote leukaemia (mean%±%s.e.m.; n%=%4 samples per cohort; data combined from two independent
growth. a,b, Immunoblot and quantification of mTOR pathway proteins in +/+ experiments). k,l, Expression of glycolysis genes in +/+ Lin− bcCML cells treated
Lin− bcCML treated with indicated doses of TAG (a, mean%±%s.d.; n%=%2 replicates with rapamycin for 24%h (k; mean%±%s.d.; n%=%3 technical replicates per cohort) or
per cohort) or 1%mM GES (b, mean%±%s.e.m.; n%=%4 samples per cohort). 2%mM MHY1485 for 48%h (l; mean%±%s.e.m.; n%=%3 technical replicates per cohort;
c-f, Experimental strategy (c), immunoblot (d, f), and quantification (e) of data combined from two independent experiments). m, Histogram and MFI of
indicated proteins in Lin− bcCML cells +/− 200%mM taurine (mean%±%s.e.m.; n%=%3 p-mTOR in Lin− bcCML cells infected with vector or RagA(Q66L) (mean%±%s.e.m.;
biological replicates; gels processed in parallel; tubulin run on each gel as sample n%=%6 samples per cohort; data combined from two independent experiments).
processing control). g,h, Immunoblots (g), and quantification (h) of mTOR n, CFU of Lin− bcCML cells infected with vector or RagAQ66L and treated with
pathway proteins in Lin− bcCML cells from leukaemic mice supplemented with indicated amounts of taurine (mean%±%s.e.m.; n%=%3 independent culture wells
taurine for 10 days (mean%±%s.e.m.; n%=%4 control and n%=%5 taurine; b, e, h unpaired per cohort; data combined from two independent experiments; j, l-n one-way
two-tailed Student’s t-test). i, Survival curve shows impact of 5%mg/kg rapamycin ANOVA). o, Schematic shows taurine from BM osteolineage cells promotes
treatment for 6%d on leukaemia progression. Line represents treatment days RagA-dependent mTOR activation and glycolysis in leukaemia cells to drive
(n%=%9 +/+ and n%=%5 −/−; data combined from two independent experiments; disease progression. The culture well image in c is adapted from ref. 6, Springer
log-rank test). j, Histogram and geometric Mean Fluorescence Intensity (MFI) Nature America.
of p-mTOR in Lin− bcCML cells treated with DMSO or N-Oleoyl taurine for 36–40%h.
 nature portfolio | reporting summary
Corresponding author(s): Jeevisha Bajaj
Last updated by author(s): Apr 5, 2025

Reporting Summary
Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.

A description of all covariates tested

A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)

For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.

Software and code

Policy information about availability of computer code
Data collection Flow cytometry-based analysis was performed on LSR Fortessa and cell sorting was carried out on BD FACS Aria II using FACSDiva v9.0 (BD
Biosciences).
qRT-PCR was carried out on BioRad CFX96 C100 Thermocycler using BioRad CFX Manager 1.1 v4.1 (BioRad) and Thermofisher Scientific Quant
Studio 12K Flex Real Time PCR using using Quant Studio v1.2 (Thermofisher Scientific).
RNA and DNA concentrations were determined using the NanoDrop 1000 Spectrophotometer (Thermofisher Scientific). RNA quality was
assessed with the Agilent Bioanalyzer 2100 (Agilent Technologies)
Western blots were performed using LI-COR Odyssey M using Empiria Studio v2.3 (LI-COR).
Taurine assay absorbance was measured using BioTek Synergy 2 plate reader using Gen5 v.3.11 (BioTek).
Seahorse assays were carried out on Agilent Technologies XF96 Analyzer using Wave v2.6.3 (Agilent Technologies).
Microscopy images were obtained on Olympus CKX41 using CellSens Entry v2.3 (Olympus).
Immunohistochemical images were obtained with Olympus BX41 and deconvolution was completed with Imaris v10.2 .
Immunoflurescence images were obtained with with a Teledyne Photometrics Prime BSI Express sCMOS Camera mounted on a Nikon ECLIPSE
Ti2 inverted microscope equipped with the NIS-Elements 6D Imaging Acquisition Module (v5.42.06)
scRNA-Sequencing data was generated using NovaSeq 6000 (Illumina) and samples were counted using cellranger-4.0.0.
Bulk RNA Sequencing data was generated using Illumina HiSeq2500 v4 or llumina NextSeq 550.
Mass Spectrometry data was obtained Vanquish Neo UHPLC (Thermofisher Scientific) connected to Orbitrap Astral Mass Spectrometer
April 2023

(Thermofisher Scientific) . Data-dependent acquisition method utilized a FAIMS Pro Duo (Thermofisher Scientific).
LC/MS data was analyzed by high resolution mass spectrometry with an Orbitrap Exploris 240 (Thermofisher Scientific) coupled to a Vanquish
Flex liquid chromatography system (Thermofisher Scientific).

1
 Data analysis Statistical analyses were performed using GraphPad Prism v9 and v10.
Flow cytometry data was analyzed using FlowJo software v10.9.0.

nature portfolio | reporting summary

Seahorse data was analyzed using Wave V2.6.3 (Agilent Technologies)
qRT-PCR data was analyzed using BioRad CFX Manager 1.1 v4.1 and Quant Studio v1.2 (Thermofisher Scientific)
Western blot images were analyzed using Empiria Studio v3.2.0.186 (LI-COR)
Immunohistochemical images were analyzed using IHC plugin toolbox in Fiji v1.54g.
Immunioflurescence Images were deconvoluted using Imaris (Oxford instruments), and Pearson’s colocalization analysis was done using the
JACoP BIOP plugin in Fiji v1.54g. .
Mouse and human single cell RNA datasets were processed using Seurat (v4.1.0, [Link]
within R (v4.1.1, [Link]
Time course analysis was performed using DEGreport (v1.30.3, [Link]
Pathway enrichment was perfomed using EnrichR (v3.0, [Link]
Sample integration within human single cell RNA datasets was performed using Harmony (v0.1.0, [Link]
harmony).
Integrating external datasets into the human single cell RNA dataset was done using Seurat (v5.0.3.99911, [Link]
seurat/releases/tag/v5.0.3) and Azimuth (v0.5.0, [Link] within R (v4.3.1, [Link]
tag/v4.3.1).
Mouse and human bulk RNA datasets were trimmed and quality filtered using FastP (v0.20.1, [Link]
tag/v0.20.1).
Read data was aligned using STAR (v2.7.6a, [Link] and counted using subread-
featureCounts (v2.0.1, [Link]
Differential expression was performed using DESeq2 (v1.28.1, [Link] within R
(v4.0.2, [Link]
Pathway enrichment within the bulk RNA datasets was performed using EnrichR (v3.0, above). Ligand receptor visualization leveraged
NicheNetR (v1.1.0, [Link] and Circlize (v0.4.15, [Link] within R (v4.1.1,
above).
Ggplot2 ([Link] and dplyr ([Link] were leverage for figure generation and data
manipulation throughout.
Mass Spectrometry phosphoprotein data was analyzed with Vanquish Neo UHPLC (Thermofisher) connected to Orbitrap Astral Mass
Spectrometer (Thermofisher). global DIA raw data was processed with DIA-NN version 1.8.1 using library-free analysis mode. Protein
quantification was carried out using the MaxLFQ algorithm and the number of peptides quantified in each protein group with the
DiannReportGenerator Package. Phosphoproteome raw data was searched using the CHIMERYS within the Proteome Discoverer software
platform v3.1 (ThermoFisher)
LC-MS data were analyzed by Compound Discover (v3.3, Thermo Scientific) and El-Maven software
Statistical analyses were carried out using Graphpad Prism Software v6.0 (Graphpad Software Inc.)
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and
reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Portfolio guidelines for submitting code & software for further information.

Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A description of any restrictions on data availability
- For clinical datasets or third party data, please ensure that the statement adheres to our policy

Raw data is available under source data.

The data for this publication is available under the following accessions:
GSE226372 (human bulk RNA-Seq; [Link]
GSE227082 (mouse bulk RNA-Seq; [Link]
GSE226644 (mouse temporal scRNA-Seq; [Link]
GSE288862 (ScRNA-Seq of Human MDS and AML Bone Marrow Marrow Microenvironment; [Link]
PXD062322 (WT and TauT-/- proteomics data). The library was annotated using Mus musculus UniProt ‘one protein sequence per gene’ database
(UP000005640_9606, downloaded April 2021).
ST003835 (WT and TauT-/- metabolomics data)
ST003836 (13C taurine tracing data).
A shiny app hosting the mouse temporal scRNA-Seq is hosted at [Link] and was generated using ShinyCell (v2.1, https://
[Link]/SGDDNB/ShinyCell)

Additional Datasets:
1. Ligands were determined using NicheNet that can be accessed here:
Browaeys, R., Saelens, W. & Saeys, Y. NicheNet: modeling intercellular communication by linking ligands to target genes. Nat Methods 17, 159-162 (2020). https://
[Link].1038/s41592-019-0667-5
2. Cell Surface Protein atlas can be accessed here:
Bausch-Fluck, D. et al. A mass spectrometric-derived cell surface protein atlas. PLoS One 10, e0121314 (2015). [Link]
3. AML and Healthy Immune Microenvornment Datasets can be accessed here:
April 2023

Lasry, A. et al. An inflammatory state remodels the immune microenvironment and improves risk stratification in acute myeloid leukemia. Nat Cancer 4, 27-42
(2023). [Link]

2
 nature portfolio | reporting summary
Research involving human participants, their data, or biological material
Policy information about studies with human participants or human data. See also policy information about sex, gender (identity/presentation),
and sexual orientation and race, ethnicity and racism.
Reporting on sex and gender Studies on human samples not considered human subjects research (per NIH guidelines)

Reporting on race, ethnicity, or Studies on human samples not considered human subjects research (per NIH guidelines)
other socially relevant
groupings

Population characteristics Studies on human samples not considered human subjects research (per NIH guidelines)

Recruitment Studies on human samples not considered human subjects research (per NIH guidelines)

Ethics oversight Studies on human samples not considered human subjects research (per NIH guidelines)

Note that full information on the approval of the study protocol must also be provided in the manuscript.

Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see [Link]/documents/[Link]

Life sciences study design

All studies must disclose on these points even when the disclosure is negative.
Sample size No statistical method was used to predetermine sample size for experiments. We did not have any preconceived assumption regarding
experiments sample size, data was collected depending on the nature of the experiments and proper statistical analysis. Adequate sample size
were determined based on previous publications:

Bajaj, J., Hamilton, M., Shima, Y. et al. An in vivo genome-wide CRISPR screen identifies the RNA-binding protein Staufen2 as a key regulator of
myeloid leukemia. Nat Cancer 1, 410–422 (2020). [Link]

Bajaj J, Konuma T, Lytle NK, Kwon HY, Ablack JN, Cantor JM, Rizzieri D, Chuah C, Oehler VG, Broome EH et al (2016) CD98-mediated adhesive
signaling enables the establishment and propagation of acute myelogenous leukemia. Cancer Cell 30: 792–805

Kwon HY, Bajaj J, Ito T, Blevins A, Konuma T, Weeks J, Lytle NK, Koechlein CS, Rizzieri D, Chuah C, et al. (2015). Tetraspanin 3 Is Required for
the Development and Propagation of Acute Myelogenous Leukemia. Cell Stem Cell 17, 152–164. 10.1016/[Link].2015.06.006.

Data exclusions No data was excluded

Replication Each experiment was repeated 2 to 4 times as stated in the legend of each figure.

Randomization All animals used for in vivo studies were randomly selected to receive either control or treatments. Animals were selected based on genotype.
Age, Sex, and environment were controlled for. In vitro experimental samples (cell lines and primary samples) were randomly selected for
control group or treatments. All other environmental factors were controlled for.

Blinding Blinding was not relevant to this study as researchers needed to know the conditions for each experiment. Flow cytometry, FACS, western
blots, seahorse, sequencing, and imaging obtain objective measurements using analytical machines therefore blinding was not necessary.

Reporting for specific materials, systems and methods

We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
April 2023

3
Materials & experimental systems Methods

nature portfolio | reporting summary

n/a Involved in the study n/a Involved in the study
Antibodies ChIP-seq
Eukaryotic cell lines Flow cytometry
Palaeontology and archaeology MRI-based neuroimaging
Animals and other organisms
Clinical data
Dual use research of concern
Plants

Antibodies
Antibodies used A detailed list of antibodies used for flow cytometry and IF (including vendor, catalog number, and dilution) is provided in
Supplementary Table 2.

Validation All antibodies were purchased from commercial vendors and have been validated by the manufactures for use in the species and
assays utilized in this study. The validation statements and published references are on the manufacturers' websites. Further in-
house validation was performed with appropriate negative and positive controls for each antibody.

Eukaryotic cell lines

Policy information about cell lines and Sex and Gender in Research
Cell line source(s) HEK293 (ATCC CRT-3216), K562 (ATCC CCL-243), THP-1(ATCC TIB-202), MV-4-11 (ATCC CRL-9591) cells were purchased from
ATCC. MDS-L cells were from K. Tohyama.

Authentication Cell lines obtained directly from ATCC were validated by vendor. MDS-L cells were authenticated in house via flow cytometry
as CD45+CD34+CD38+ on 2/27/23.

Mycoplasma contamination Cell lines were not tested for mycoplasma.

Commonly misidentified lines No commonly misidentified lines were used.

(See ICLAC register)

Animals and other research organisms

Policy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research, and Sex and Gender in
Research

Laboratory animals All mice used were between 6 and 16 weeks old and both male and female mice were used.
Slc6a6(TauT) were maintained on the B6 background.
B6-CD45.1 (Strain: [Link]-PtprcaPepcb/BoyJ), C57BL6/J, and NSG mice ([Link]-Prkdcscid Il2rgtm1Wjl/SzJ) mice were maintained as
homozygous strains.
Cdo1fl/fl;Prrx1-Cre+ were maintained by crossing female Cdo1fl/fl with male Prrx1-Cre/+

Wild animals The study does not involve wild animals.

Reporting on sex Both male and female mice were used for experiments.

Field-collected samples The study does not involve field-collected samples.

Ethics oversight All animal experiments were performed according to protocols approved by the University of Rochester Institutional Animal Care and
Use Committee. Premorbid animals were euthanised at indicated experimental time points or at end point. For all experiments, mice
were monitored closely for signs of disease or morbidity daily and were sacrificed for visible signs of hunched dorsum, failure to
thrive, or any signs of infection. These limits were not exceeded for any experiment.
Note that full information on the approval of the study protocol must also be provided in the manuscript.
April 2023

4
Flow Cytometry

nature portfolio | reporting summary

Plots
Confirm that:
The axis labels state the marker and fluorochrome used (e.g. CD4-FITC).
The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers).
All plots are contour plots with outliers or pseudocolor plots.
A numerical value for number of cells or percentage (with statistics) is provided.

Methodology
Sample preparation For KLS, bone marrow was recovered from femurs, tibias, and pelvis. Bones were crushed using a mortar and pestle, cells
were suspended in Hanks’ balanced salt solution (HBSS) (Gibco, Life Technologies) containing 5% (vol/vol) fetal bovine serum
and 2 mM EDTA and filtered using a 70um filter. Red blood cells were lysed using RBC Lysis Buffer (eBioscience) and ckit cells
were enriched by positive selection using magnetic beads (Miltenyi Biotec) on a AutoMACS Pro (Miltenyi Biotec) before
staining for lineage markers. All antibodies were purchased from BD Pharmingen, eBioscience or BioLegend.

Analysis of leukaemic mice was similarly done. Briefly, cells, spleens and/or bone marrow from sick mice were prepared as
described above to generate single cell suspensions that were then incubated with FACS antibodies.

Mesenchymal stromal cells (MSCs) were isolated from leukaemic mice and cultured in 10cm dishes in MEM α with no
ascorbic acid (Gibco) supplemented with 15% FBS and 100 IU/mL Penicillin/Streptomycin (Gibco). 6 days post culture
initiation, the cells were sorted for MSCs.

For BM stromal cells, bone and bone marrow (BM) were isolated from long bones and pelvis in 1x Media 199 (Gibco) with 2%
fetal bovine serum (GeminiBio). BM was digested for 30 minutes in HBSS containing 2mg/mL Dispase II (Gibco), 1mg/mL
Collagenase Type IV (Sigma-Aldrich), and 20ng/mL DNase Type II (Sigma-Aldrich). Bone spicules were digested for 60 minutes
in PBS supplemented with 2.5mg/mL Collagenase Type I (Stem Cell Technologies) and 20% FBS. Digested bone marrow was
RBC lysed using RBC Lysis Buffer (eBioscience). Bone and BM cells were pooled and CD45+ Ter119+ hematopoietic cells were
magnetically depleted on an autoMACS cell separator (Miltenyi Biotec). The CD45-Ter119- stromal cells were either stained
an analyzed for candidate populations by flow cytometry (BD LSRFortessa) or further enriched by sorting (BD FACSAria II) and
processed for single cell RNA-Sequencing.

Lineage depleted TauT+/+ and TauT-/- murine leukeamia cells were fixed using BD Cytofix/Cytoperm Fixation/
Permeabilization Kit (BD Bioscience) per manufacturer’s protocols. Cells were stained with primary antibody against
phospho-mTOR (Cell Signaling Technologies). Cells were then stained with donkey anti-rabbit secondary antibody conjugated
with Alexa Fluro 488 (Invitrogen) to detect mTOR. Analysis was performed on LSRFortessa (Becton Dickinson). Data was
analyzed using FlowJo software.

Instrument Analysis was carried out by LSRFortessa and cell sorting was carried out on a FACSAria II (BD Biosciences).

Software FACSDiva v9.0 (BD Biosciences) was used to collect data and data analysis was done using FlowJo v10.9.0. (Tree Star Inc.).

Cell population abundance The purity of post-sort KLS from TauT +/+ and TauT-/- mice was higher than 90% as determined by flow cytometry.
The purity of post-sort transduced KLS cells was greater that 95% as determined by flow cytometry.

Gating strategy KLS sort: morphology (FSC/SSC) -> live cells (PI-) -> lineage negative (lin-) -> ckit + & Sca1 + (ckit/Sca1).
BCR-ABL & NUP98-HOXA9 transduced KLS: morphology (FSC/SSC) -> live cells (PI-) -> NGFR+ (BCR-ABL vector) & CD2+
(NUP98-HOXA9 vector).
Primary and Secondary bcCML Sort: morphology (FSC/SSC) -> live cells (PI-) -> NGFR+/GFP (BCR-ABL vector) & CD2+/YFP
(NUP98-HOXA9 vector) -> lineage negative (lin-).
Secondary AML Sort: morphology (FSC/SSC) -> live cells (PI-) -> NGFR+ (MLL-AF9 vector) & YFP (NRAS vector) -> Ckit positive
(Ckit+).
shRNA transduced bcCML and human samples: morphology (FSC/SSC) -> live cells (PI-) -> GFP+ (shRNA vector).
Stromal cell Sort/analysis: morphology (FSC/SSC) -> live cells (PI-) -> GFP- (Cancer Negative) ->CD45-/Ter119- -> CD31-
(endothelial) -> CD51+(osteo) -> Sca-1+ or CD140a
MSC sort/analysis: morphology (FSC/SSC) -> live cells (PI-) ->CD45-/Ter119 -> CD31-(endothelial) -> CD51+(osteo) -> Sca-1+
Flow cytometry-based pMTOR expression: morphology (FSC/SSC) -> mCherry/RagA+ (if infected) -> pMtor+
Boundaries between "positive" and "negative" populations were defined using a fluorescence minus one or unstained
samples.
See Supplementary Table 2 for antibody panels used.

Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.
April 2023