Online Supplement 1 Foudi et al. 2008
This supplementary file contains:
I) Supplementary Figures 1-10
II) Supplementary References (38-42)
Nature Biotechnology: doi:10.1038/nbt.1517
ROSA26 M2-rtTA -globin polyA
SA+polyA TetOP H2B-GFP -globin polyA
+ doxycycline
ROSA26 locus
Col1A1 locus
SA
M2-rtTA
AAAAAA
AAAAAA
GFP
brightfield
-DOX +DOX
A
B CtetO-H2B-GFP
wild type
6.2kb
4kb
D
Supplementary Figure 1: Generation of ES cells and mice that permit inducible expression of H2B-GFP. (A) Alleles used to generate ES-cells and mice with doxycycline-inducible expression of H2B-GFP. (Top) The M2 reverse tetracycline transactivator (M2-rtTA) is driven by the constitutive ROSA26promoter (R26::rtTA)31,32. (Bottom) A previously characterized H2B-GFP cDNA8 was inserted downstream of the Collagen (Col) 1A1 locus under the control of atetracycline-dependent minimal CMV promoter by recombinase-mediated site-specific integration as previously described31. SA, splice acceptor; TetOP, tetracyclineoperator elements fused to CMV minimal promoter. (B) Southern blot analysis confirming integration of the H2B-GFP transgene downstream of Col1A1 locus. (C) ES-cellscontaining both the R26::rtTA and the TetOP-H2B-GFP allele turn green in the presence of the tetracycline analog doxycycline. These ES cells were used to producechimeric animals and germline offspring (data not shown). (D) Experimental scheme: mice harboring M2-rtTA and H2B-GFP were subjected to doxycycline in theirdrinking water for 6 weeks to achieve widespread acquisition of green fluorescence. Subsequently, mice were checked for the retention of H2B-GFP in the absence ofdoxycycline at defined intervals for up to 72 weeks after the end of doxycycline administration.
Nature Biotechnology: doi:10.1038/nbt.1517
3 7 18 22 26
L-K+S+
48+150
-
L-K+S+
48-150+
L-K+S+
48-150
-n=5n=4
n=2
weeks
B-cells
(B220+)
Granulocytes
(Gr1+/Mac1+)
*
Supplementary Figure 2: CD150 expression distinguishes long-term from short-termrepopulating HSCs. Analysis of CD45-isotype expression in peripheral blood lymphocytes and granulocytes aftertransplantation of highly purified early bone marrow populations (donor, CD45.2) into irradiated hosts (CD45.1) togetherwith a small number of support marrow cells (CD45.1) to ensure survival. The four plots show the proportions of donor-derived B-cells and granulocytes in groups of mice that had received 50 cells from sub-fractions of the Lineage- (L), c-Kit+ (K), Sca1+ (S)-population further resolved by SLAM markers CD48 (48) and CD150 (150) as indicated by theletter/number code in each panel (flow-cytometry of these populations is shown in Figure 1A, first panel on the right).Note: CD48-positive L-K+S+-cells did not give rise to donor-contribution (upper panels). Within the CD48-negativefractions (lower panels), CD150-negative cells (left) gave rise to transient hematopoiesis and CD150-positive cells (right)gave rise to long-term engraftment. *Persistent blood lymphoid cells in lower left panel represent re-circulating B-cells asearly bone marrow B-cells are not donor-derived in these mice (see Supplementary Figure S3 for bone marrowcontribution analysis of selected mice from lower panels). Individual data points represent means from analysis of groupsof mice, number of mice (n) given in each panel, error bars represent standard deviation in upper left panel and standarderror of mean in all other panels.
n=5L-K+S+
48+150+%
Donor
(CD
45.2
)%
Donor
(CD
45.2
)
weeks
3 7 18 22 26
0
3 7 18 22 26
0
3 7 18 22 26
weeks
weeks
% D
onor
(CD
45.2
)%
Donor
(CD
45.2
)
Nature Biotechnology: doi:10.1038/nbt.1517
Supplementary Figure 3: Host and donor contributionsto bone marrow hematopoiesis after transplantation ofCD150-positive and CD150-negative CD48-L-K+S+cells.Bone marrow analysis of selected mice from lower panels inSupplementary Fig. 2. Shown are donor (CD45.2) or host (CD45.1)contributions to bone marrow hematopoiesis after transplantation ofCD150-negative (left panels) or CD150-positive (right panels) donorcells 26 weeks after transplantation. (A) Contribution to B-cells andmyeloid cells. (B) Contribution to stem cells (LKS). Note: myeloid cells(Gr1+/ Mac1+; upper panel), early B-cells (B220+/IgM-, middle panel),and L-S+K+ -cells are derived from the donor only after transplantationof CD150+ cells.
A
B
Nature Biotechnology: doi:10.1038/nbt.1517
CD150
CD
48
CD34
Flt3 H2B-GFP
H2B-GFP
H2B-GFP
EPCR H2B-GFP
Endoglin H2B-GFP
SS
CS
SC
SS
CS
SC
50.4
13
10.8
17.8
10.1
42
33.4
MNCD2 H2B-GFP
SS
C 39.6 41
38.1
A
B
C
D
E
F
88.7
93.8
CD150
41.8 15.8
1923.5
CD
48 41.8 15.8
1923.5
85.7 1.39
1.3111.6
85.7 1.4
1.311.6
CD150
CD
4846.2 7.65
13.832.4
46.2 7.6
13.832.4
CD150
CD
48
58.7
11.6 13
58.716.7
11.6 13
58.716.7
CD150
CD
48
15.2 9.7
35.439.7
15.2 9.7
35.439.7
CD150C
D48
6.11 19.2
57.217.5
6.1 19.2
57.217.5
CD150
CD
48
G
H
I
J K
L
LKS Flt3- LKS Flt3+
LKS Flt3- MNCD2+
LKS Endoglinhi
LKS EPCRhi
LKS CD34-
50.2
36.5
84.8
35.9
61.6
34.3
Supplementary Figure 4: H2B-GFP retention correlates withmultiple alternative markers forlong-term repopulating HSCs, butnot with expression of MNCD-2.(A-E) Left column shows HSCs (L-K+S+)stained with SLAM markersCD48/CD1504, CD3414, Endothelial ProteinC Receptor (EPCR)17, Endoglin18, andFlt319, all of which predict long-termrepopulation potential. (F) Left plot showsHSCs (L-K+S+Flt3-) stained with MNCD-2,an antibody that has been claimed torecognize N-cadherin20 by flow-cytometry(see Supplementary Figure 7). Dotted linesin B-F indicate arbitrary threshold set toresult in less than 5% events abovethreshold with isotype-matched controlantibody. Right plots show H2B-GFPcontent of the total population (gray shadedarea, L-K+S+ in A-E, L-K+S+Flt3- in F) orgated populations as indicated (red arrowsand red curves). Shown is onerepresentative out of three similarexperiments performed after 16 weeks ofchase (see Supplementary Fig. 1D). (G-L)SLAM-marker profiles of HSCs identifiedby alternative markers (B-F). Previousgating strategies are indicated on plots inred (as shown in B-F). (G) The majority ofCD34-negative HSCs are CD48-CD150+.(H) EPCRhi HSCs are predominantlyCD48-negative. (I) Endoglinhi HSCs arepredominantly CD48-CD150+. (J, K) Flt3-
HSCs contain the majority of the CD48-
CD150+ cells. (L) MNCD-2+ HSCs are notbiased toward a particular SLAM-markerprofile.
Nature Biotechnology: doi:10.1038/nbt.1517
A
B
LKS Flt3-
38.1
H2B-GFP
41.6
32.7
LKS Flt3-
39.5
H2B-GFP
42.6
30.1
15.1
78.1
MNCD2
SS
C
15.6
76.8
IgG2a
SS
C
Supplementary Figure 5: H2B-GFP retention ofMNCD-2 or isotype control stained HSCs usingalternative gating strategy. (A) H2B-GFP retention ofMNCD-2 stained HSCs as shown in Supplementary Figure 4F, butwith alternative gates. While we could not discern distinct expressionlevels of MNCD-2 as reported20 because of the low intensity of thestain, these gates represent our best approximation of “low” (blueframe) and “intermediate” (red frame) expression levels. Note thatthere is a small difference in H2B-GFP retention but the same is seenwith the isotype control (B).
Nature Biotechnology: doi:10.1038/nbt.1517
B220hiIgM+
Mature B
MNCD2
4.7
99.2
40.1 7.32
17.834.9
Sca-1
c-K
it
CD150
CD
48
Lin- Flt3- PI- gated LKS+ gated
MNCD2 MNCD2
LKS+ Flt3- LKS+ CD48-CD150+
5.2
37.2
4.6
29
17.8
MNCD2
B220lowIgM+
Immature B4.8
45.2
IgG
2a
MNCD2
B220+IgM-
PreB and ProB4.9
28.5
4.5
26
10.8
4.93
IgM
B220
Scatter PI- gated
10.8
4.926
Lin + PI
MN
CD
2
7.3
Lin + PI
0.4
A B
C
D
MNCD2
LKS-
24.5
4.8
33
Supplementary Figure 6: Characterization of MNCD-2 staining in the adult bone marrow. (A, B) Dim staining with MNCD-2 isdetectable on HSCs (Flt3-L-K+S+ (A), CD150+CD48-L-K+S+ (B), red curves) and myeloid progenitors ((A), blue curve) compared to isotype control (gray shadedareas). (C) Bright staining with MNCD-2 (left panel) but not isotype control (right panel) on lineage positive cells. (D) Progressive up-regulation of MNCD-2during B-cell maturation in the bone marrow. Note that this staining profile does not represent N-cadherin expression (shown in Supplementary Figure 7).
Nature Biotechnology: doi:10.1038/nbt.1517
18.7 89.5Isotype
4.7
MNCD2 MNCD2
Isotype
5
LKS+Lin+
Con
trol
Con
trol
KO
KO
floxedwt
deleted
A B
C
Flt3-L-K+S+
Ncad KO
B220hiIgM+
Control96.5Isotype
5.3
MNCD2
17.4Isotype
4.3
Flt3-L-K+S+
Control
MNCD2
B220hiIgM+
Ncad KO
Supplementary Figure 7: Analysis of MNCD-2 binding in the adult bone marrow cells two months after pIpCmediated disruption of a conditional N-cadherin allele using Mx-Cre reveals that MNCD-2 staining isindependent of N-cadherin expression. (A, B) MNCD-2 staining of HSCs (Flt3-L-K+S+) (A) and mature B-cells (B) in pIpC-treated control (left panels) and conditional N-cadherin knockout bone marrowsupplementary ref. 38 (right panels). Note that the MNCD-2 stainwas not affected by disruption of N-cadherin. (C) Conditional N-cadherin was excised in sorted HSCs (LKS+) and sorted lineage marker+
cells from conditional knockout mice (Genotype: Mx-Cre, N-cadherin floxed/floxed) but not pIpC treated controls (Genotype: No Cre, N-cadherin floxed/ wildtype) (primers as describedsupplementary ref. 38). Note: These data demonstrate that MNCD-2 does not recognize N-cadherin on hematopoietic cells when used in flow-cytometry. Thus the ligand for MNCD-2 in hematopoiesis is dubious, even if previousstudies have shown that MNCD-2 binds to N-cadherin in Western blot analysis of neuronal tissue supplementary ref. 39,40.
Nature Biotechnology: doi:10.1038/nbt.1517
75985
127563
127563
59686
57572
117258
A B
n=21 n=42
Supplementary Figure 8: H2B-GFP of label-retainingHSCs is distributed equally to daughter cells at the timeof division. Highly purified HSCs (CD150+CD48-L-K+S+) from H2B-GFP mice, 20 weeks after the pulse, were single-cell sorted intoindividual culture wells containing stem cell factor, thrombopoetin, andinterleukin-3 to induce divisionsupplementary ref. 41,42. (A) Fluorescence wasquantified before and after the first division. Pixel intensity of individualcells is shown in white numbers, pixel intensity of all cell is shown inyellow. (B) Bar graphs show mean fluorescence of daughter cells (rightbar) as the percentage of the fluorescence of the parental cells (left bar,set at 1); number of cells analyzed and standard deviations are givenbelow.
Nature Biotechnology: doi:10.1038/nbt.1517
25.8
23.2
39.3
29
43.3 90.3
EPCR
CD34H2B-GFP CD34 CD34
EPCR EPCR
SS
C
SS
C
SS
C
SS
C
SS
C
SS
C
29.828.527
Endoglin
SS
C
Endoglin
SS
C
Endoglin
SS
C
29
3923.2
I
II
III
Gate I
H2B-GFPneg
Gate II
H2B-GFPint
Gate III
H2B-GFPhiL-K+S+ 48-150+
28± 7
48± 9
83± 9
44± 2
19± 4
7± 2
30± 4
31± 5
29± 2
Gate IH2B-GFPneg
Gate II
H2B-GFPint
Gate IIIH2B-GFPhi
Gate I
H2B-GFPneg
Gate II
H2B-GFPint
Gate III
H2B-GFPhi
A B
C
D
Supplementary Figure 9: CD34 expression and loss of EPCR correlate with loss of H2B-GFP inCD150+CD48-L-K+S+ HSCs. HSCs (CD150+CD48-L-K+S+) from bone marrow of H2B-GFP mice 16 weeks after a doxycylinepulse were analyzed for H2B-GFP expression (A) and the expression of CD3414 (B), EPCR17 (C), and Endoglin18 (D) was correlated withabsence of H2B-GFP (left panels), intermediate levels of H2B-GFP (middle panels), and high levels of H2B-GFP (right panels).Representative panels from three identical experiments are shown, numbers are averages ± standard deviations, n=3 for all panels. Note:there in a striking inverse correlation between CD34 expression and H2B-GFP retention, and H2B-GFP retention correlates with highEPCR levels. Endoglin shows no correlation with H2B-GFP expression.
Nature Biotechnology: doi:10.1038/nbt.1517
CD34S
SC
CD
48
L-K+S+ L-K+S+ 48-150+
31.8 6.81
2140.5 21
54.654±2
EPCR
SS
C
24±1
84±2
57±2
78±7
57±2
H2B-GFP
H2B-GFPCD150
L-K+S+ 48-150+
40.4
31.8 6.8
Supplementary Figure 10: CD34 and EPCR enhance thecorrelation with H2B-GFP retention in CD150+CD48-L-S+K+
HSCs. HSCs (CD150+CD48-L-K+S+) from bone marrow of H2B-GFP mice16 weeks after a doxycyline pulse were analyzed for H2B-GFP expression incombination with CD3414 (upper panels) or EPCR17 (lower panels). Greyshaded areas and grey numbers in right panels show fluorescence of the entireCD150+CD48-L-K+S+-population; red curves and numbers show fluorescenceof CD34-negative (upper) or EPCRhi HSCs (CD150+CD48-L-K+S+). Numbersare averages ± standard deviations (n=3).
Nature Biotechnology: doi:10.1038/nbt.1517
Online Supplement 12 Foudi et al. 2008
II) Supplementary References 38. Kostetskii, I. et al. Induced deletion of the N-cadherin gene in the heart leads to
dissolution of the intercalated disc structure. Circ Res 96, 346-54 (2005). 39. Matsunami, H. & Takeichi, M. Fetal brain subdivisions defined by R- and E-
cadherin expressions: evidence for the role of cadherin activity in region-specific, cell-cell adhesion. Dev Biol 172, 466-78 (1995).
40. Radice, G. L. et al. Developmental defects in mouse embryos lacking N-cadherin. Dev Biol 181, 64-78 (1997).
41. Ema, H., Takano, H., Sudo, K. & Nakauchi, H. In vitro self-renewal division of hematopoietic stem cells. J Exp Med 192, 1281-8 (2000).
42. Takano, H., Ema, H., Sudo, K. & Nakauchi, H. Asymmetric division and lineage commitment at the level of hematopoietic stem cells: inference from differentiation in daughter cell and granddaughter cell pairs. J Exp Med 199, 295-302 (2004).
Nature Biotechnology: doi:10.1038/nbt.1517