Wnt Gradient Formation Requires Retromer Function

www.sciencemag.org/cgi/content/full/1124856/DC1
Supporting Online Material for
Wnt Gradient Formation Requires Retromer Function
in Wnt-Producing Cells
Damien Y. M. Coudreuse, Giulietta Roël, Marco C. Betist,
Olivier Destrée, Hendrik C. Korswagen*
*To whom correspondence should be addressed. E-mail: rkors@niob.knaw.nl
This PDF file includes:
Materials and Methods
Figs. S1 to S4
Tables S1 to S6
References
Published 27 April 2006 on Science Express
DOI: 10.1126/science.1124856

Supporting online material
Materials and Methods
C. elegans strains and culturing
General methods for culture, manipulation and genetics of C. elegans were as described (1).
Unless indicated, strains were cultured at 20 o C. Mutations and transgenes used in this study were
LGI, pop-1(hu9) (2), pry-1(mu38) (2, 3) and axl-1(tm1095); LGII, mab-5(e1751) (4), vps-
35(hu68) and muIs32[mec-7::gfp] (5); LGIV, vps-29(tm1320); LGIV, vps-26(tm1523), egl-
20(n585) (3) and egl-20(hu105, hu120); LGX, vps-5(tm847) and bar-1(ga80) (6). The egl-20
alleles hu105 and hu120 both change W108 into a stop, but have been isolated in two
independent mutagenesis screens. axl-1 encodes an Axin related protein (K02B12.4) and the
deletion allele tm1095 likely represents the null phenotype. The axl-1 mutation does not affect
Q.d migration on its own, but strongly enhances the pry-1(mu38) induced ectopic expression of
mab-5 in QR (Oosterveen et al., submitted)
C. elegans phenotypes, expression constructs and transgenesis
V5, HSN and Q cell phenotypes were scored in late L1 as described (7, 8). mab-5::lacZ reporter
transgene activation (4), dye filling (9) and P12 to P11 fate transformation (9) were scored as
described. To address the embryonic lethal phenotype, gravid animals were cut open, as vps-
35(hu68) and vps-26(tm1523) animals are fully egg-laying defective. Embryos were transferred
onto new plates and hatching was scored 48 hours later. RNAi was performed as described (10).
The proteinA fusion transgenes, containing two IgG binding domains of Staphylococcus aureus
1

proteinA, were injected with the dpy-20 rescuing plasmid pMH86 into dpy-20(e1282) mutants.
Extra-chromosomal arrays were integrated as described (11). A similar EGL-20 gradient was
observed with a Pegl-20::egl-20::Venus transgene (18) (Fig. S4). No gradient was detected when
the proteinA IgG binding sites were fused to the first 66 amino-acids of EGL-20 (including the
signal peptide) and expressed under the control of the egl-20 promoter (data not shown). The
Pegl-20::egl-20::proteinA transgene does not induce an egl-20 overexpression phenotype and
rescues the egl-20(n585) mutant phenotype (data not shown).
Immunostaining and microscopy
Late L1 larvae were mounted on poly-lysine-coated slides and were freeze-cracked in liquid
nitrogen. After cold methanol and acetone treatment, larvae were rehydrated in 90%, 60%, and
30% ethanol/PBS and finally PBS. Larvae were then blocked in PBS, 0.5% Tween-20, 2% milk
and stained with a 1/100 dilution of swine anti-goat Ig’s-FITC (Biosource). After washing in
PBS, 0.5% Tween-20, the slides were analysed by confocal microscopy. Identical settings were
used to compare the different strains. Series were stacked according to maximum pixel intensity.
Inducible RNAi clones, luciferase assay
HEK cells stably expressing the tetracycline repressor (HEK TR, Invitrogen) were transfected
with different pools of RNAi constructs cloned in the pTER vector (12). Stable clones were
selected and tested by northern blot for Vps35 mRNA downregulation upon doxycycline
addition. Similar results were obtained with other clones targeting different regions of Vps35.
Luciferase assays were performed as described (13). Approximately 5x10 5 cells were transfected
with 100 ng of SuperTOPFLASH or the negative control SuperFOPFLASH (14) and 10 ng of
2

the internal transfection control pTKRenilla. 10ng of Wnt3A and 100 ng of TCF-β-catenin (15)
expression constructs were transfected to activate Wnt signaling 48 hours after RNAi induction.
Luciferase activities were measured 48 hours after transfection. Similar results were obtained
using 10 ng or 50 ng of a Wnt3A::proteinA fusion.
Xenopus tropicalis experiments
X. tropicalis embryos were obtained by natural mating (as described by R. Grainger at
http://faculty.virginia.edu/xtropicalis/). Except when specified, 20 and 10 ng of Vps35 MO
(Genetools, 5’GGACTGCTGGGTCGTCGGCATAATC3’) were injected to block Vps35
mRNA translation in whole and half embryos, respectively. 10 pg of human Vps35 RNA was
used to rescue Vps35 MO induced developmental arrest. 0.5 pg of Wnt1 and 20 pg of TCF3-
Armadillo RNA were ventrally injected in 4-cell stage embryos in the double axis assay. To
investigate the effect of Vps35 knock-down on endogenous Wnt target gene expression, embryos
at the 4-cell stage were injected on one side with 10 ng of Vps35 MO. In situ hybridizations were
performed as described (16). A random morpholino was used as a control in the double axis
assay.
Western blots and immunoprecipitations
Animals carrying the Pegl-20::egl-20::proteinA transgene were washed in 10 mM Tris-HCl
pH8, 150 mM NaCl, 1% Triton X-100, protease inhibitors (Roche) and lysed in the same buffer
using glass beads (Sigma #G-8772). Equal amounts of protein were separated on SDS-PAGE
gels and stained with a peroxidase-coupled anti-peroxidase antibody (Sigma). The secretion of a
functional Wnt3A::proteinA fusion in the medium of transfected mammalian cells was tested by
3

immunoprecipitation using IgG-coupled beads (Amersham) and western blot analysis. To allow
better detection, 50 ng of Wnt3A::proteinA expression construct was used. Induced and non-
induced HEK(Vps35) cells were transfected with the Wnt3A::proteinA expression construct.
After 12 hours, cells were split in different wells and the medium refreshed. 100 µg/ml of
cycloheximide was added to the medium. One well was used at each time point to
immunoprecipitate Wnt3::proteinA from the medium and prepare total cell lysate. Equal loading
was confirmed using an anti-α-tubulin antibody (Sigma).
4

Results
Mosaic analysis shows that vps-35 is specifically required in EGL-20 producing cells
We assayed rescue in transgenic animals that mosaically express vps-35 from the cdh-3
promoter, which is active in the EGL-20 producing cells F and U and several additional cells that
are located between the Q neuroblasts and the EGL-20 producing cells (17) (Table S3). We
found that 50% of randomly scored mosaic Pcdh-3::vps-35::gfp animals showed normal QL.d
migration (n=106). Similarly, when vps-35 was expressed only in F and U, the phenotype was
rescued in 54% of the animals (n=39). This rescue was not enhanced when vps-35 was also
expressed in cells located between F and U and the Q cells (53%; n=61). Finally, none of the
mosaic animals that had lost vps-35 expression in both F and U, but retained expression in other
cells, showed rescue (n=6). These results demonstrate that the presence of the retromer complex
in EGL-20 producing cells is both necessary and sufficient for its function in EGL-20 signaling.
5

Supplementary Figure Legends
Fig. S1. (A), Schematic representation of Q cell migration in C. elegans. Dorsal view. The
complete Q cell lineages and their migrations are represented on both sides. Cells are in green
and red when mab-5 expression is activated or absent, respectively. egl-20 expressing cells are
in blue. The grey circles represent the seam cells (V1 to V6). In wild type (top panel), the Q
daughter cells on the left side (L) activate mab-5 expression and migrate toward the posterior,
whereas the Q daughter cells on the right side (R) migrate in the default anterior direction.
Ectopic activation of mab-5 expression in the QR.d of pry-1(mu38), axl-1(tm1095) pry-
1(mu38) and mab-5(e1751) mutants results in posterior migration (middle panel). Loss-of-
function of positive regulators of the EGL-20 pathway such as egl-20/Wnt, lin-17/Fz, mig-
1/Fz, mig-5/Dsh, bar-1/β-catenin or pop-1/Tcf (bottom panel) results in anterior migration of
the QL.d. (B), Schematic representation of the different deletion alleles of C. elegans retromer
complex genes. Bars indicate the deletions used in this study. Numbers refer to the cDNA
sequence.
Fig. S2. (A), The retromer complex is required for Q daughter cell and HSN migration and
acts upstream of mab-5. In vps-26(tm1523) mutants, the HSN neurons could not be located
(n.l.). (B), VPS-35 acts in the EGL-20 producing cells. Wild type vps-35 was expressed in vps-
35(hu68) mutants, under the control of vps-35, myo-2, elt-2, ric-19, egl-20, myo-3 or cdh-3
promoters. Only the vps-35, myo-3, cdh-3 and egl-20 promoters showed rescue of the mutant
phenotype. The final positions of the HSN neurons (n=50), the QL.d (n=50) and QR.d (n=50)
6

were determined relative to the seam cells (V1 to V6) after their first division. Bars indicate
the percentage of cells at each position. Dashed lines indicate the wild type position.
Fig. S3. Vps35 morpholino (MO) injection has a dose dependent effect on Xenopus tropicalis
development. (A), Embryos were injected at the 2-cell stage in both blastomeres with a total of
20, 10 and 4 ng of Vps35 MO (n≥10). Body length was measured relative to an arbitrary unit
(a.u.). (B), Injection of 20 ng of Vps35 MO in 2-cell stage embryos results in developmental
arrest between stage 10 1/2 and 11, which can be rescued by co-injection of human Vps35 RNA
(hVps35) (n≥40). N.I., non-injected control.
Fig. S4. Anteroposterior concentration gradient of EGL-20 visualized in living animals using a
fusion with the enhanced YFP version Venus (18). The arrowhead indicates P11/12. The white
bar delimits the group of egl-20 expressing cells. The dashed line represents the range of the
EGL-20::Venus gradient.
7

Figure S1
A
B
Wild type
pry-1, axl-1 pry-1, mab-5(e1751)
egl-20, lin-17, mig-1, mig-5, bar-1 and pop-1
vps-5
vps-26
vps-29
vps-35
1 1419
1
1
1
tm1523
tm1320
tm847
576
hu68
1071
P11/12
R
L
R
L
R
L
2466

Figure S2
QL.d QR.d
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Wild type
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-35(hu68)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
mab-5(e1751)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-35(hu68);mab-5(e1751)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
egl-20(n585)
%
%
%
%
%
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
HSN
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-26(tm1523)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-29(tm1320)
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-5(tm847)
0
10
20
30
40
50
60
70
80
90
100
%
%
%
n.l.
V1 V2 V3 V4 V5 V6 V1 V2 V3 V4 V5 V6 V1 V2 V3 V4 V5 V6
A
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
V1 V2 V3 V4 V5 V6
QL.d QR.d
Wild type
vps-35(hu68)
vps-35(hu68);Pvps-35::vps35::gfp
vps-35(hu68);Pmyo-3::vps-35::gfp
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-35(hu68);Pcdh-3::vps-35::gfp
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-35(hu68);Pmyo-2::vps-35::gfp
%
%
%
%
%
%
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
%
vps-35(hu68);Pelt-2::vps-35::gfp
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
%
vps-35(hu68);Pegl-20::vps-35::gfp
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
HSN
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
vps-35(hu68);Pric-19::vps-35::gfp
%
V1 V2 V3 V4 V5 V6
V1 V2 V3 V4 V5 V6
B
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
egl-20(hu105)
%
0
10
20
30
40
50
60
70
80
90
100

Figure S3
A B
Body length (a.u.)
8
7
6
5
4
3
2
1
0
N.I.
Vps35 MO
% developmental arrest
100
90
80
70
60
50
40
30
20
10
0
N.I.
Vps35 MO
hVps35 RNA
Vps35 MO
+
hVps35 RNA

Figure S4
EGL-20::VENUS

Table S1. Retromer complex components in C. elegans
VPS-35
S. cerevisiae
24.3 (37.1)
12
Drosophila
41.8 (56.4)
Human
46.8 (61.1)
VPS-26 24 (34.5) 39.8 (47) 49.5 (59.8)
VPS-29 26.5 (35.7) 55 (70.2) 56.5 (70.7)
VPS-5/SNX-1 12.2 (21.1) 30.2 (45.5) 29.4 (43)
Percentage identity and similarity (between brackets) are indicated. Full-length protein
sequences were aligned using the ALIGN program.

Table S2. The retromer complex is required for Q cell migration
13
RNAi
Genotype IPTG(mM) empty vps-5 vps-26 vps-35 pry-1 pos-1 bar-1
Wild type
1
0/100/0
0/100/0
2/98/0
76/24/0
0/56/44
Wild type 0.2 0/100/0 / 3/96/1 42/58/0 / +
pop-1(hu9) 1 14/86/0 10/90/0 81/19/0 98/2/0 4/94/2 +
vps-5(tm847) 1 0/100/0 / lethal lethal 0/55/45 +
vps-5(tm847) 0.2 0/100/0 / 18/79/3 ∗ lethal / +
vps-26(tm1523) 1 90/1/9 lethal / 90/1/9 77/2/21 +
vps-29(tm1320) 1 25/75/0 75/25/0 ∗ 97/1/2 84/0/16 16/80/4 +
vps-35(hu68) 1 100/0/0 lethal 100/0/0 / 54/37/9 +
mab-5(e1751) 1 0/1/99 0/1/99 0/0/100 0/1/99 / +
Wild type † 1 0/100/0 0/100/0 2/98/0 29/71/0 / +
+
/
/
/
/
/
/
/
/
/
17/93/0
pry-1(mu38) axl-1(tm1095) † 1 0/0/100 0/0/100 0/0/100 0/0/100 / + 3/14/83
RNAi was performed at 15°C except for pry-(mu38) axl-1(tm1095) double mutants which are
not viable at this temperature. RNAi efficiency was reduced when necessary by lowering the
IPTG concentration from 1 mM to 0.2 mM. The numbers indicate the final positions of the Q
daughter cells (n > 100): % QL.d anterior / % wild type / % QR.d posterior. RNAi against the
essential gene pos-1 was used as a control for RNAi efficiency in each strain. bar-1(RNAi) was
used as a control for suppression of the pry-1(mu38) axl-1(tm1095) double mutant phenotype.
vps-29(RNAi) was found to be inefficient even in a sensitized pop-1(hu9) strain (2) carrying the
RNAi hypersensitive mutation rrf-3(pk1426) (19) (data not shown).
∗ low brood, sterility, larval arrest, embryonic lethality.
† 20 °C

Table S3. The retromer complex is required in the EGL-20 producing cells
Allele
Wild type
Transgene
promoter
-
Tissue
specificity
14
QL.d posterior HSN anterior
100
100
Wild type
V5 polarity
vps-35(hu68) - 0 6 77
vps-35(hu68) vps-35 widely 63 80 97
vps-35(hu68) myo-2 pharyngeal muscles 0 10 87
vps-35(hu68) elt-2 intestine 0 10 82
vps-35(hu68) ric-19 neurons (including the Q and
HSN neurons)
100
0 0 76
vps-35(hu68) egl-20 P11/12, K, B, F, U, mu_anal 94 94 100
vps-35(hu68) myo-3 body wall muscles, mu_sph,
mu_anal
vps-35(hu68)
cdh-3 HSN, Q, F, U, seam, exc, hyp
10/11, VC, anchor, vulva
epidermis
41 30 98
51 78 99
egl-20 expressing cells are indicated in bold. QL.d and HSN position and V5 polarity were
scored using Nomarski optics (n≥100). All numbers are percentages. Expression of egl-20
cDNA using the egl-20 promoter was sufficient to rescue the egl-20(n585) mutant phenotype
(data not shown). This promoter is not expressed in the Q lineage, as a Pegl-20::bar-1
transgene did not rescue bar-1(ga80) mutants (data not shown). Cell nomenclature is as
described (20).

Table S4. The retromer complex mainly affects EGL-20/Wnt signaling
Genotype
Wild type
V5 polarity
reversal
0
Embryonic
lethality
9
15
Dye filling
defect
1
P12 → P11
mom-2(or309) - 100 - -
lin-44(n1792) 0 9 84 48
egl-20(n585) 18 12 5 0
egl-20(hu105) 67 n.d. 5 2
vps-35(hu68) 23 13 4 6
vps-29(tm1320) 0 12 2 0
vps-26(tm1523) 34 32 0 17
vps-5(tm847) 0 16 0 0
lin-17(n671) 0 13 81 23
bar-1(ga80)
0 9 3 96
V5 polarity and P12 to P11 transformation were scored using Nomarski
optics at the appropriate stages. Dye filling was scored on synchronized
populations of young adults (9). n≥100 in each assay. All numbers are
percentages. n.d : not determined.
0

Table S5. Loss of retromer function affects the polarity of V5 division
Genotype
Wild type
V5 polarity reversal
0 ± 0
egl-20(hu120) 62 ± 3
egl-20(hu105) 54 ± 7
vps-35(hu68) 14 ± 2
vps-26(tm1523)
37 ± 3
V5 polarity was scored at 25°C (21). vps-26(tm1523) was scored at 20°C, as adults
were sterile at 25°C. We found that the V5 polarity defect of vps-26(tm1523) is not
temperature sensitive (data not shown). Numbers are percentages (mean ± SD,
n=300).
16

Table S6. Overexpression of β-catenin or TCF3-Armadillo rescues the Vps35
morpholino-induced downregulation of XmyoD and Xslug
Vps35 MO
XmyoD
0 (20)
17
Xslug
4 (25)
Vps35 MO + 100 pg β-cat. 69 (13) 0 (9)
Vps35 MO + 200 pg β-cat. 94 (16) 37.5 (8)
Vps35 MO + 250 pg β-cat 100 (10) 22 (9)
Vps35 MO + 500 pg β-cat. 100 (8) 75 (8)
Vps35 MO + 100 pg TCF3-Arm. 87 (16) 30 (10)
Vps35 MO + 200 pg TCF3-Arm.
86 (15) 22 (9)
Percentages of animals expressing XmyoD or Xslug are indicated. Numbers of animals
scored are between brackets.

References and notes
1. J. A. Lewis, J. T. Fleming, Methods Cell Biol 48, 3 (1995).
2. H. C. Korswagen et al., Genes Dev 16, 1291 (2002).
3. J. N. Maloof, J. Whangbo, J. M. Harris, G. D. Jongeward, C. Kenyon, Development 126,
37 (1999).
4. S. J. Salser, C. Kenyon, Nature 355, 255 (1992).
5. Q. Ch'ng et al., Genetics 164, 1355 (2003).
6. D. M. Eisenmann, J. N. Maloof, J. S. Simske, C. Kenyon, S. K. Kim, Development 125,
3667 (1998).
7. J. Harris, L. Honigberg, N. Robinson, C. Kenyon, Development 122, 3117 (1996).
8. J. Whangbo, C. Kenyon, Mol Cell 4, 851 (1999).
9. M. A. Herman, H. R. Horvitz, Development 120, 1035 (1994).
10. R. S. Kamath, M. Martinez-Campos, P. Zipperlen, A. G. Fraser, J. Ahringer, Genome
Biol 2, RESEARCH0002 (2001).
11. C. Mello, A. Fire, in Caenorhabditis elegans: Modern Biological Analysis of an
Organism H. F. Epstein, D. C. Shakes, Eds. (Academic, San Diego, 1995) pp. 451-482.
12. M. van de Wetering et al., EMBO Rep 4, 609 (2003).
13. M. van de Wetering et al., Cell 88, 789 (1997).
14. M. T. Veeman, D. C. Slusarski, A. Kaykas, S. H. Louie, R. T. Moon, Curr Biol 13, 680
(2003).
15. F. J. Staal, B. M. Burgering, M. van de Wetering, H. C. Clevers, Int Immunol 11, 317
(1999).
18

16. R. M. Harland, Methods Cell Biol 36, 685 (1991).
17. J. Pettitt, W. B. Wood, R. H. Plasterk, Development 122, 4149 (1996).
18. T. Nagai et al., Nat Biotechnol 20, 87 (2002).
19. F. Simmer et al., Curr Biol 12, 1317 (2002).
20. J. E. Sulston, H. R. Horvitz, Dev Biol 56, 110 (1977).
21. J. Whangbo, J. Harris, C. Kenyon, Development 127, 4587 (2000).
19