Dictyostelium chemotaxis: essential Ras activation and accessory signalling pathways for amplification.

Abstract

Central to chemotaxis is the molecular mechanism by which cells exhibit directed movement in shallow gradients of a chemoattractant. We used Dictyostelium mutants to investigate the minimal requirements for chemotaxis, and identified a basal signalling module providing activation of Ras at the leading edge, which is sufficient for chemotaxis. The signalling enzymes PI3K, TorC2, PLA2 and sGC are not required for Ras activation and chemotaxis to folate or to steep gradients of cAMP, but they provide a memory of direction and improved orientation of the cell, which together increase the sensitivity about 150-fold for chemotaxis in shallow cAMP gradients.

Full text

Page 1

Dictyostelium chemotaxis: essential Ras activation
and accessory signalling pathways for amplification
Arjan Kortholt, Rama Kataria, Ineke Keizer-Gunnink, Wouter N. Van Egmond, Ankita Khanna
& Peter J.M. Van Haastert+
Department of Cell Biochemistry, University of Groningen, Groningen, The Netherlands
Central to chemotaxis is the molecular mechanism by which
cells exhibit directed movement in shallow gradients of a
chemoattractant. We used Dictyostelium mutants to investigate
the minimal requirements for chemotaxis, and identified a basal
signalling module providing activation of Ras at the leading
edge, which is sufficient for chemotaxis. The signalling enzymes
PI3K, TorC2, PLA2 and sGC are not required for Ras activation
and chemotaxis to folate or to steep gradients of cAMP, but they
provide a memory of direction and improved orientation of the
cell, which together increase the sensitivity about 150-fold for
chemotaxis in shallow cAMP gradients.
Keywords: chemotaxis; shallow gradients; Ras; amplification;
synergy
EMBO reports (2011) 12, 1273–1279. doi:10.1038/embor.2011.210
INTRODUCTION
An important aspect of cell motility is the ability of cells to
respond to directional cues and display oriented migration (Van
Haastert & Devreotes, 2004). Gradients of diffusive chemicals give
rise to chemotaxis (Hoeller & Kay, 2007; Swaney et al, 2010).
Chemotaxis has a pivotal role in finding nutrients in prokaryotes,
forming multicellular structures in protozoa, tracking bacterial
infections in neutrophils and organizing the embryonic cells in
metazoa (Van Haastert & Devreotes, 2004). Sensitive cells, such
as Dictyostelium or neutrophils, can detect very shallow spatial
gradients of about 1% concentration difference across the cell
(Mato et al, 1975).
The signal transduction cascade for chemotaxis consists of
surface receptors, heterotrimeric and small G proteins, as well as
several signalling enzymes, leading to the local activation of the
cytoskeleton, predominantly F actin at the front and myosin
filaments in the rear of the cell. In Dictyostelium, a shallow
gradient of cyclic AMP induces the activation of cAMP receptors
and their associated G proteins Ga2bg, in a manner that is
approximately proportional to the steepness of the gradient (Xiao
et al, 1997; Jin et al, 2000). In contrast, activation of Ras is much
stronger in the front than in the rear of chemotaxing cells (Zhang
et al, 2008). Four signalling enzymes, PI3K, TorC2, PLA2 and sGC,
have been implicated in chemotaxis (Chen et al, 2007; Kamimura
et al, 2008; Veltman et al, 2008; Liao et al, 2010). Activation of
PI3K, TorC2 and sGC occurs downstream of Ras, and simulta-
neous disruption of both rasC and rasG results in the total loss
of cAMP-mediated signalling (Bolourani et al, 2006; Kortholt &
van Haastert, 2008).
A central question in chemotaxis is how these pathways allow
efficient navigation of cells in very shallow gradients. Here we
obtained quantitative data on Ras activation and chemotaxis in
cAMP and folic acid gradients and report on the minimal
requirements for chemotaxis of Dictyostelium cells.
RESULTS AND DISCUSSION
Ras activation is independent of the signalling pathways
Four signalling pathways, PI3K, TorC2, PLA2 and sGC, contribute
to chemotaxis (Chen et al, 2007; van Haastert et al, 2007;
Kamimura et al, 2008; Veltman et al, 2008). PI3K-mediated
phosphorylation of Akt/PKB is completely inhibited by 20mM
LY, whereas TorC2-stimulated phosphorylation of PKBR1 is not
inhibited by 20mM LY but is strongly inhibited by 90mM LY
(Kamimura et al, 2008; Liao et al, 2010). Here we used sgc/pla2-
null cells with 90mM LY, by which all four pathways are inhibited.
Similar results were obtained for two different cell lines in which
all four signalling pathways are inhibited: sgc/pla2/pkbR1-
nullþ90mM LY and gc-nullþ90mM LYþ2mM BPB (inhibition
of PLA2). To investigate the function of the signalling enzymes in
directional movement, we obtained quantitative data on Ras
activation and formation of F actin in the cortex, which can be
detected as the translocation of Ras-binding domain (RBD)-Raf–
GFP (green fluorescent protein) and LimEDcoil–GFP, respectively.
Mutant cells with a deletion of the single Dictyostelium gb
gene do not show any chemotaxis, actin polymerization or Ras
activation (data not shown (Wu et al, 1995; Sasaki et al, 2004)).
Cells lacking either RasC or RasG exhibit good chemotaxis,
whereas cells lacking both RasC and RasG move in random
directions at all cAMP gradients tested (supplementary Fig S1A
Received 17 June 2011; revised 16 September 2011; accepted 22 September 2011;
published online 11 November 2011
+Corresponding author. Tel: þ31 50 3634172; Fax: þ31 50 3634165;
E-mail p.j.m.van.haastert@rug.nl
Department of Cell Biochemistry, University of Groningen, Nijenborgh 7,
9747 AG Groningen, The Netherlands
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATIONEMBO reportsVOL 12 | NO 12 | 2011
scientificreport
scientificreport
1273

Page 2

online (Bolourani et al, 2006)). Furthermore, rasC/G-null cells do
not exhibit RBD-Raf–GFP translocation to the membrane upon
cAMP stimulation and have a very poor cAMP-stimulated F-actin
response (supplementary Fig S1B online (Bolourani et al, 2006)).
These experiments confirm that Gbg and RasC/G are essential
for chemotaxis (Wu et al, 1995; Sasaki et al, 2004). In wild-type
cells, uniform cAMP induces a translocation of RBD-Raf–GFP
and LimEDcoil–GFP from the cytosol to the membrane (Fig 1;
supplementary Fig S2 online). Mutant sgc/pla2-null/LY cells
exhibit essentially the same response as wild-type cells (Fig 1;
supplementary Fig S2 online). Furthermore, even sgc/pla2-null/LY
cells incubated with 5mM actin-polymerization inhibitor Latrun-
culin A show a Ras response that is indistinguishable from that of
wild-type cells (Fig 1A).
Upon application of a cAMP gradient to wild-type cells, initially
RBD-Raf–GFP transiently translocates uniformly to the cell
boundary, which is then followed by RBD-Raf–GFP localization
at the side of the cell facing the cAMP gradient (Fig 2A;
supplementary Movie 1 online). Wild-type cells in a cAMP
gradient are very polarized and exhibit strong localization of
RBD-Raf–GFP in a well-defined crescent with a length of
4.4±0.6mm and an intensity that is 230±50% above the intensity
of the cytoplasm (mean and s.d., n¼16 cells, Fig 2A). In a steep
cAMP gradient, mutant sgc/pla2-null/LY cells are less polarized
and have a broader leading edge (Fig 2B); however, they show the
accumulation of RBD-Raf–GFP and LimEDcoil–GFP at the side of
the cell facing the cAMP gradient, and cells exhibit significant
chemotaxis (Figs 1,2; supplementary Fig S2 and Movie 2 online).
Consistent with a broader leading edge, RBD-Raf–GFP is
enriched in a larger crescent with a length of 7.6±0.6mm and
with a lower fluorescent intensity that is 130±30% above the
intensity of the cytoplasm. The amount of Ras activation, defined
as the product of crescent size and increase of intensity, is not
significantly different between wild-type (1,012±260mm%) and
sgc/pla2-null/LY cells (988±240mm%). Wild-type cells treated
with 5mM Latrunculin A exhibit strong localized Ras activation in
a cAMP gradient (Fig 1A; supplementary Movie 3 online). The
length of the crescent is 7.5±1.3mm, with an intensity that is
155±30% above the intensity of the cytoplasm (product 1,160±
300mm%), similar to the crescent of sgc/pla2-null/LY cells.
Together, these results show that Ras is activated at the leading
edge through a pathway consisting of Gabg, which does not depend
on the four signalling pathways or a functional cytoskeleton.
Chemotaxis in shallow cAMP gradients
The input signal for chemotaxis is a spatial gradient of cAMP,
dC/dx (Mato et al, 1975). We measured Ras activation and
chemotaxis of wild-type and mutant cells exposed to cAMP
gradients with different steepness; see Postma and van Haastert
(2009) for the equations that define the spatial gradient
B
C
WT
110
100
90
80
70
Raf–GFP in cytoplasm (%)
Raf–GFP in cytoplasm
(% depletion)
60
40
30
20
10
0
0 10–1010–9
cAMP concentration (M)
10–8
10–7
10–6
10–5
0
1020
Time (s)
30
sgc/pla2+LY
WT
sgc/pla2+LY
A
Buffer Uniform cAMPGradient cAMP
Wild type
RBD-Raf–GFP (activated Ras)
sgc/pla2-null
+ 90 μM LY
sgc/pla2-null
+ 90 μM LY
+ 5 μM Lat A
Fig 1|Translocation of RBD-Raf–GFP from the cytoplasm to the cell boundary in wild-type cells, sgc/pla2-null cells with 90mm LY and sgc/pla2-null/LY
cells with 5mM LatA. (A) Images of RBD-Raf–GFP-expressing cells in buffer, 3–6s after the addition of uniform cAMP or in a cAMP gradient; scale
bar, 5mm. (B) The time course of translocation after uniform stimulation with 1mM cAMP as means and s.e.m. of eight cells. (C) The response at
3–6s after addition of uniform cAMP with different concentrations; shown are the means and s.e.m. of at least eight cells. cAMP, cyclic AMP;
GFP, green fluorescent protein; LatA, Latrunculin A; WT, wild type.
Minimal requirements for chemotaxis
A. Kortholt et al
EMBO reports VOL 12 | NO 12 | 2011
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
scientificreport
1274

Page 3

at different distances from a pipette. Fig 2C presents the fraction
of cells that exhibit significant enrichment of RBD-Raf–GFP at
the leading edge (significant is defined as more than 30% above
the intensity of the cytoplasm). These translocation experiments
can only detect strong responses. Weak responses are undetect-
able, because boundary pixels are on average half filled
with cytoplasm; as a result, the association of small amounts of
RBD-Raf–GFP to the membrane does not result in a fluorescent
intensity of the boundary pixel that is more than the fluorescent
intensity in the adjacent cytosolic pixels (Bosgraaf et al, 2008).
Half-maximal strong activation of Ras occurs at about 2,000pM
mm?1in both wild-type and sgc/pla2-null/LY cells (Fig 2C). The
data suggest that activation of Ras is essentially the same in wild-
type and sgc/pla2-null/LY cells, except that in the mutant
activation occurs in a broader leading edge. These results reveal
that none of the four signalling enzymes (PI3K, TorC2, PLA2 and
sGC), or potentially other enzymes inhibited by the high
concentrations of LY294002, are involved in the transient
activation of Ras to uniform cAMP, or persistent activation of
Ras at the leading edge in a cAMP gradient.
Interestingly, chemotaxis of sgc/pla2-null/LY cells occurs in
gradients with the same steepness as strong activation of Ras,
whereas chemotaxis of wild-type cells is observed in spatial
gradients that are about 150-fold smaller. Together, these experi-
ments suggest that directional movement in the absence of the four
pathways occurs without amplification of the spatial information
that is carried by the localized Ras activation, whereas the four
pathways allow chemotaxis to cAMP in very shallow gradients.
During chemotaxis in natural cAMP waves, cells are exposed to
cAMP gradients up to about 3,000pM mm?1(Tomchik & Devreotes,
1981; Postma & van Haastert, 2009). The results presented in
Table 1 reveal that many mutants will chemotax in such gradients,
despite their reduced sensitivity. This indicates that wild-type cells
have a large spare sensitivity to tolerate non-optimal conditions in
the soil; it also explains why many mutant cells show good cell
aggregation under laboratory conditions.
Synergy provides amplification of cAMP gradient sensing
We determined the steepness of the cAMP gradient that induces
half-maximal chemotaxis (dC/dx)50 in a large collection of
mutants (see supplementary Table S1 online). Wild-type cells
exhibit half-maximal chemotaxis in a cAMP gradient of magnitude
22pM mm?1. Using three different cell lines in which all four
pathways are inhibited, we obtained (dC/dx)50¼3,588±356pM
mm?1, which is about 150-fold larger than the (dC/dx)50of wild-
type cells. To delineate the contribution of the pathways in
amplification, we determined for many mutants the steepness of
the gradient that induces half-maximal chemotaxis. In a mutant
with pathway X active, the value of (dC/dx)50relative to 3,588pM
mm?1of the four-pathway null mutant indicates the increase of
sensitivity, or amplification, that is provided by pathway X.
When only one enzyme is active, chemotaxis occurs in slightly
shallower gradients compared with the four-pathway null cells
(Table 1). Expression of two pathways further increases sensitivity.
Some combinations of pathways (for example, PI3KþsGC)
exhibit a weak amplification, which in magnitude is similar
to the product of the amplifications induced by the individual
pathways, as would be expected for independent effects.
However,othercombinations
sGCþTorC2 and sGCþPLA2) induce very good chemotaxis with
strong amplification, indicating synergy of signalling pathways.
The sGC enzyme has two functions in chemotaxis: one as a
protein that localizes at the leading edge (sGCp), and one as
enzyme producing cyclic GMP (cGMP; Veltman & van Haastert,
2006).Byusingtwomutant
which prevents cGMP synthesis, and sGCDN, by which the
protein becomes cytosolic but retains the ability to synthesize
(forexample,PI3KþTorC2,
proteins, namelysGCDCat,
A
C
B
D
Wild type
*
*
Wild type
Wild type
sgc/pla2-null + LY
1.0
0.8
0.6
0.4
0.2
0
Chemotaxis index (ο,o)
1.0
0.8
0.6
0.4
0.2
0
0.01 0.1110100
0.01 0.1
dC/dx (nM μm–1)
1 10100
12.3
4.9
2.7
3.2
3.0
PI3K
TorC2
sGCp
cGMP
PLA2
10
Amplification100
1
0
25
50
75
100
0
25
50
75
100
sgc/pla2-null+LY
RBD-Raf–GFP
(activated Ras)
Percentage of cells with RBD-Raf–GFP at front (Δ,Δ)
Fig 2|Activation of Ras and chemotaxis in cAMP gradients.
(A,B) Images from a movie of wild-type and sgc/pla2-null/LY cells,
respectively. The asterisks indicate the position of the pipette; scale bar,
10mm. The chemotaxis index and the fraction of cells with significant
enhanced RBD-Raf–GFP at the leading edge were determined (defined as
cells containing a fluorescence intensity of RBD-Raf–GFP at the leading
edge that was at least 30% above the fluorescence intensity in the
cytosol). (C) Strong Ras activation occurs only in steep gradients both
in wild-type and sgc/pla2-null/LY cells. Chemotaxis of sgc/pla2-null/LY
mutant cells occurs also only in these steep gradients, whereas wild-type
cells exhibit chemotaxis in about 150-fold more shallow gradients.
(D) A cluster analysis of amplification and synergy of signalling
pathways derived from Table 1. The data points indicate the means and
s.d. of amplification, and the numbers represent synergy (both obtained
from Table 1). The tree was built by cluster analysis of the five pathways
using the neighbour-joining method with synergy and amplification as
distance measures. cAMP, cyclic AMP; GFP, green fluorescent protein.
Minimal requirements for chemotaxis
A. Kortholt et al
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATIONEMBO reports VOL 12 | NO 12 | 2011
scientificreport
1275

Page 4

cGMP (Veltman & van Haastert, 2006), it could be shown
that thecGMPfunctionsynergizes
the localization function of sGCp protein synergizes with
TorC2 signalling (Table 1).
Cluster analysis was used to determine the interactions
between signalling pathways for optimal synergy and amplifica-
tion. The obtained tree presented in Fig 2D reveals two branches
of synergizing activities, a branch with PI3KþTorC2þsGCp, and
a branch with cGMPþPLA2, respectively. Interestingly, this tree
on chemotaxis has the same hierarchy as the tree that describes
how signalling enzymes modulate pseudopodia in gradients
of thechemoattractant(Bosgraaf
P.J.M. van Haastert, unpublished observations on TorC2 mutants).
PI3K, sGCp and TorC2 are activated in the same region of the cell
as activated Ras; that is, at the side of the cell facing the gradient
(Funamoto et al, 2002; Veltman & van Haastert, 2006; Zhang
withPLA2,whereas
& vanHaastert,2009;
et al, 2008), where they increase the probability that a pseudopod
is formed. Thus, this synergizing cluster induces orientation
of the cell towards the gradient. The second cluster consists of
cytosolic cGMP and PLA2 (Chen et al, 2007), which are not
involved in orientation but induce splitting of the current
pseudopod (Bosgraaf & van Haastert, 2009). As a daughter-
splitting pseudopod is extended in a similar direction to that
of the parent pseudopod, the cGMP/PLA2 cluster functions as a
memory of current direction.
In summary, one individual pathway does not contribute
strongly to chemotaxis, but it can synergize with other pathways,
which together increases the sensitivity about 150-fold.
Folate chemotaxis is independent of the pathways
Folate activates many signalling pathways in a way that is
similar to activation by cAMP. Folate activates a folate receptor
Table 1|Amplification of chemotaxis by synergy of signalling pathways
Active pathways (dC/dx)50(pM lm?1) Amplification Synergy*
Mean±s.d.Nz
Observedy
Expected||
Mean±s.d.
No active pathway3,588±3567 (3), a 1.00 1.00±0.10
PI3K2,414±893 (1), b 1.49
TorC21,246±2186 (2), c 2.88
PLA2 1,772±375 3 (1), d 2.03
sGC¼(sGCp+cGMP)
sGCp
1,848±273 11 (4), e1.94
2,474±744 (1), f 1.45
cGMP3,417±3114 (1), g1.05
PI3K+TorC2283±60 6 (2), h12.694.292.96±0.82**
PI3K+PLA21,082±225 3 (1), i3.323.01 1.10±0.33 NS
PI3K+sGC1,025±2584 (2), j 3.50 2.891.21±0.36 NS
TorC2+PLA2 430±1234 (1), k8.345.83 1.43±0.57 NS
TorC2+sGC220±2014 (5), l 16.285.592.91±0.71**
TorC2+sGCp322±12 4 (1), m11.16 4.172.67±0.48**
TorC2+cGMP 1,472±69 4 (1), n 2.443.02 0.81±0.16 NS
PLA2+sGC296±679 (4), o 12.113.93 3.08±1.06**
PLA2+sGCp975±1884 (1), p 3.682.941.25±0.36 NS
PLA2+cGMP 343±164 (1), q 10.472.13 4.92±1.16**
PI3K+TorC2+PLA2183±274 (1), r 19.598.662.26±0.71**
PI3K+PLA2+sGC185±144 (2), s19.434.584.24±1.05**
PI3K+TorC2+cGC171±96 (2), t20.966.51 3.22±0.67**
TorC2+PLA2+cGC86±129 (3), u41.63 8.884.69±1.51**
PI3K+TorC2+PLA2+sGC 22±26 (1), v 162.6913.1912.33±3.85**
Chemotaxis was tested to cAMP gradients with different steepness (as shown in Fig 2C). The spatial gradient that induces half-maximal chemotaxis is defined as (dC/dx)50.
Experiments were performed with various mutants and inhibitors to obtain cells with the combination of pathways active (see supplementary Table S1 online for these
conditions). The s.d. for amplification and synergy were calculated using the mean and s.d. of the (dC/dx)50data; s.d. for amplification are not shown.
*Synergy is the observed amplification divided by the expected amplification. The significance of synergy was tested relative to 1.00±0.10, the synergy for the condition ‘no
active pathway’; **Po0.001; NS, not significant at P40.05.zNumber of observations indicates total number of replicates with the number of strains in parenthesis. The letters
refer to the strains used as indicated in supplementary Table S1 online.yObserved amplification: (dC/dx)50¼3,588 of ‘no pathway’ divided by (dC/dx)50of the specified
active pathway.||Expected amplification is the product of observed amplifications of the individual pathways.
cAMP, cyclic AMP; PI3K, phosphoinositide kinase-3.
Minimal requirements for chemotaxis
A. Kortholt et al
EMBO reports VOL 12 | NO 12 | 2011
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
scientificreport
1276

Page 5

(De Wit & Bulgakov, 1985), predominantly the G protein Ga4bg
(Hadwiger et al, 1994), Ras (Lim et al, 2005) and the signalling
enzymes PI3K, mTOR and guanylyl cyclase (Mato et al, 1977),
leading to the formation of F actin and myosin filaments in the
cortex (McRobbie & Newell, 1985). In a gradient of folate, Ras is
activated at the leading edge with similar kinetics to that observed
for cAMP (Fig 3A,B). The onset and magnitude of the response
upon uniform stimulation are similar in wild-type and mutant cells
lacking the four signalling enzymes, but recover faster in mutant
cells (Fig 3A,B). Compared with cAMP chemotaxis, the response
to folate requires steeper gradients, and the maximal chemotaxis
index is smaller (Fig 3C; about 0.9 for cAMP versus about 0.6 for
folate; the half-maximal response is about 0.3, and the gradient
that induces a half-maximal response is defined as (dC/dx)50). For
wild-type cells, we observed half-maximal chemotaxis to folate with
a gradient of (dC/dx)50¼2.8±1.4nM mm?1(Fig 3D). Unexpectedly,
inhibition of all four signalling enzymes had no effect on folate
chemotaxis (Fig 3D). Mutant sgc/pla2/pkbR1-null cells in the
presence of 20mM LY or gc-null cells in the presence of BPB and
LY exhibit chemotaxis in folate gradients that is at least as great as
that of wild-type cells in the absence or presence of 20mM LY.
Although these pathways are activated well by uniform folate, and
probably at the leading edge in a folate gradient (Mato et al, 1977;
McRobbie & Newell, 1985; Lim et al, 2005; Liao et al, 2010),
apparently this activity does not support chemotaxis. Possibly, the
target enzymes are not activated strongly in vegetative cells.
Model for chemotaxis
Fig 4 combines previous and current experiments to yield a model
for chemotaxis. Our results show that activation of Ras by cAMP is
independent of the four signalling enzymes sGC, PLA2, PI3K and
TorC2 and actin formation. Furthermore, none of the pathways
are essential for chemotaxis to folate or to steep gradients of
cAMP. As chemotaxis is completely lost in cells lacking Gb (Wu
et al, 1995) or RasC/G (Bolourani et al, 2006), we suggest a basal
signalling module consisting of heterotrimeric and monomeric G
proteins. This basal module provides activation of Ras at the
leading edge, which is sufficient for chemotaxis. The next
challenge will be to identify further components of this basal
pathway and to determine the mechanism by which hetero-
trimeric G proteins induce Ras activation. As activation of Ras
results in rapid actin polymerization, the basal module most
probably will contain regulators of the cytoskeleton such as Rho
family GTPases, ARP 2/3, WASP WAVE/Scar and mDIA (Stradal
et al, 2004). The second module consists of the Ras-activated
enzymes that induce amplification of gradient sensing. Activation
of these enzymes at the leading edge has little effect on folate
chemotaxis and is also not strictly essential for cAMP chemotaxis,
but is very important to allow chemotaxis in shallow cAMP
gradients. Important, but only partly known, are the interaction
of these pathways with the cytoskeleton and the machinery of
pseudopod formation. Positive and negative feedback loops
between the different components in these two modules probably
110
AX3
sgc/pla2/pkbR1-null + LY
Raf–GFP in cytoplasm (%)
100
90
80
70
60
5 10
Time (s)
150
B
WT
(dC/dx)50 (nM μm–1)
1 10100 1,000 10,000 100,000
rasC/G-null
g?4-null
g?-null
gc-null + BPB +LY
sgc/pla2/pkbR1-null + LY
D
Buffer
Uniform folate Gradient folate
Wild type
RBD-Raf–GFP (activated Ras)
sgc/pla2/pkbR1-
null+LY
A
1.0
Chemotaxis index
0.8
0.6
0.4
0.2
0
0.001 0.010.1
dC/dx (nM μm–1)
1 10100
cAMP
Folate
C
Fig 3|Folate chemotaxis. (A) Images of RBD-Raf–GFP-expressing cells in buffer, 3–6s after the addition of folate or in a folate gradient; scale bar,
5mm. (B) The time course of translocation after uniform stimulation with 1mM folate as means and s.e.m. of minimal seven cells. (C) The chemotaxis
index of wild-type cells exposed to different spatial gradients of cAMP or folate. (D) The gradient inducing half-maximal folate chemotaxis (dC/dx)50
was determined for the mutants shown; mutants rasC/G-null and gb-null do not exhibit chemotaxis in the steepest gradient tested. The data shown are
the means and s.d. of at least three experiments with at least 12 cells in each. Data for cAMP and folate chemotaxis were determined in cells starved
for 5 and 0.5h, respectively. cAMP, cyclic AMP; GFP, green fluorescent protein.
Minimal requirements for chemotaxis
A. Kortholt et al
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATION EMBO reports VOL 12 | NO 12 | 2011
scientificreport
1277

Page 6

further enhance chemotaxis in shallow gradients. Therefore, this
model provides a conceptual framework that can be used to
identify critical components and missing links in chemotaxis,
which might be useful to fine-tune and deepen our understanding
of chemotaxis.
METHODS
Cell culture and preparation. The strains and conditions used
in the chemotaxis experiments are described in supplementary
Table S1 online (these strains have been described; see Veltman
et al, 2008). The sgc/pla2/pkbR1-null strain was obtained by the
inactivation of the pkbR1 gene in sgc/pla2-null cells.
Cells were transformed with the plasmid pDm115RafRBD
expressing RBD-Raf–GFP (amino acids 50–134 of RAF1) that
binds to the activated GTP-bound state of Ras proteins (Kortholt &
van Haastert, 2008), or plasmid LB15B expressing LimEDcoil–GFP
(amino acids 1–145 of LimE) that binds to F actin. Cells
were grown in HL5-C medium including glucose (ForMedium),
containing 50mgml?1HygromycinB (Invitrogen) for selection.
Cells were collected and suspended in 10mM KH2PO4/Na2HPO4,
pH 6.5 (phosphate buffer (PB)).
Chemotaxis assays. Chemotaxis was measured with the small
population assay as described (Konijn, 1970), and with micro-
pipettes (supplementary Table S1 online). To obtain mutant cells
cAMP
cAR1 FAR
RasGEF
RasGEF
RasGDP
RasGDP
RasGTP
RasGTP
RasGAP
RasGAP
PLA2sGC
TorC2 PI3K
PIP3
sGCpcGMP AA?
Splitting
Memory
Persistent at front
Persistence
at front
Orientation
Orientation
F-actin
Pseudopod
F-actin
Pseudopod
ChemotaxisChemotaxis
Gα2βγ
Gα2βγ
Gα4βγ
Gα4βγ
FA
cAMP FA
cAR1 FAR
Essential
Ras activation
chemotaxis
Amplification
strong in cAMP
weak in FA
Fig 4|Model for chemotaxis. A basal essential module provides prolonged activation of Ras and F actin at the leading edge of the cell, thereby
inducing chemotaxis. This response occurs only in steep gradients. Amplification: activated Ras induces the activation of four signalling pathways.
Strong synergy of PI3K, TorC2 and sGC protein at the leading edge induces the upgradient orientation of the cell. Synergy of cGMP and PLA2 induces
splitting of the pseudopod, and as a daughter-splitting pseudopod is extended in a similar direction to that of the parent pseudopod, the cGMP/PLA2
cluster functions as a memory of current direction. Together, memory and orientation allow chemotaxis in more shallow cAMP gradients but have
little effect on folate chemotaxis. AA, arachidonic acid; cAMP, cyclic AMP; FA, folic acid; FAR, folic acid receptor; PI3K, phosphoinositide kinase-3;
PIP, phosphoinositides; PLA2, phospholipase A2; sGC, soluble guanylyl cyclase.
Minimal requirements for chemotaxis
A. Kortholt et al
EMBO reportsVOL 12 | NO 12 | 2011
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATION
scientificreport
1278

Page 7

that were in an optimal state for chemotaxis towards cAMP in the
micropipette assays, unlabelled wild-type cells were mixed with
equal amounts of green fluorescent protein (GFP)-tagged mutant
cells and starved as a confluent layer (about 106cells per cm2) on
plastic support for 5–7h until streams started to form. Cells were
collected in PB and incubated on a glass support at a lower density
of approximately 4?104cells per cm2. The distance between
adjacent cells is about 5 cell lengths, where cells do not form
streams, and chemotaxis can be measured in the absence of strong
interactions between mutant and wild-type cells. Chemotaxis was
observed using a confocal fluorescent microscope, detecting
wild-type and mutant cells in the phase-contrast and fluorescent
channels, respectively.
For chemotaxis to cAMP, we used pipettes with a tip opening
of 0.5mm, containing 0.1mM cAMP and operated at 0 or 25hPa.
For folate chemotaxis, the conditions were 1mM or 1mM folate, a
pipette tip opening of 3mm and a pressure of 1, 4 or 16hPa. The
actual gradient was measured with the fluorescent dye alexa 594
that was added to the chemoattractant solution, and the local
concentration was recorded in the red channel of the confocal
microscope (Postma & van Haastert, 2009). Most chemotaxis data
were recorded at a distance x of 30–100mm from the pipette. It
holds that the concentration is given by C(x)¼xdC/dx (Postma &
van Haastert, 2009).
Confocal images were recorded using a Zeiss LSM 510 META-
NLO confocal laser scanning microscope equipped with a Zeiss
plan-apochromatic ?63 numerical aperture 1.4 objective. The
chemotaxis index and speed were determined as described
(Veltman & van Haastert, 2006). The fluorescent intensity of
RBD-Raf–GFP in the cytoplasm was determined in large area
selections of the cytoplasm using ImageJ. The crescent of RBD-
Raf–GFP at the leading edge is defined as the area with fluorescent
intensity that is at least 1.3-fold the fluorescent intensity of
the cytoplasm. The fluorescent intensity of RBD-Raf–GFP in the
crescent was determined with curved lines using ImageJ; the
length of the crescent is defined as the length of the curved line.
Supplementary information is available at EMBO reports online
(http://www.emboreports.org).
ACKNOWLEDGEMENTS
We thank P. Devreotes, R. Firtel and the Dictyostelium stock center
for providing Dictyostelium mutants. We thank D. Veltman for initial
experiments on the F-actin response of sgc/pla2-null cells.
CONFLICT OF INTEREST
The authors declare that they have no conflict of interest.
REFERENCES
Bolourani P, Spiegelman GB, Weeks G (2006) Delineation of the roles played
by RasG and RasC in cAMP-dependent signal transduction during
the early development of Dictyostelium discoideum. Mol Biol Cell 17:
4543–4550
Bosgraaf L, Keizer-Gunnink I, van Haastert PJ (2008) PI3-kinase signaling
contributes to orientation in shallow gradients and enhances speed in
steep chemoattractant gradients. J Cell Sci 121: 3589–3597
Bosgraaf L, van Haastert PJ (2009) Navigation of chemotactic cells by
parallel signaling to pseudopod persistence and orientation. PLoS ONE
4: e6842
Chen L, Iijima M, Tang M, Landree MA, Huang YE, Xiong Y, Iglesias PA,
Devreotes PN (2007) PLA2and PI3K/PTEN pathways act in parallel
to mediate chemotaxis. Dev Cell 12: 603–614
De Wit RJ, Bulgakov R (1985) Guanine nucleotides modulate the ligand
binding properties of cell surface folate receptors in Dictyostelium
discoideum. FEBS Lett 179: 257–261
Funamoto S, Meili R, Lee S, Parry L, Firtel RA (2002) Spatial and temporal
regulation of 3-phosphoinositides by PI 3-kinase and PTEN mediates
chemotaxis. Cell 109: 611–623
Hadwiger JA, Lee S, Firtel RA (1994) The Gasubunit Ga4 couples to
pterin receptors and identifies a signaling pathway that is essential for
multicellular development in Dictyostelium. Proc Natl Acad Sci USA
91: 10566–10570
Hoeller O, Kay RR (2007) Chemotaxis in the absence of PIP3 gradients.
Curr Biol 17: 813–817
Jin T, Zhang N, Long Y, Parent CA, Devreotes PN (2000) Localization of
the G protein bg complex in living cells during chemotaxis. Science 287:
1034–1036
Kamimura Y, Xiong Y, Iglesias PA, Hoeller O, Bolourani P, Devreotes PN
(2008) PIP3-independent activation of TorC2 and PKB at the cell’s
leading edge mediates chemotaxis. Curr Biol 18: 1034–1043
Konijn TM (1970) Microbiological assay of cyclic 30,50-AMP. Experientia
26: 367–369
Kortholt A, van Haastert PJ (2008) Highlighting the role of Ras and Rap during
Dictyostelium chemotaxis. Cell Signal 20: 1415–1422
Liao XH, Buggey J, Kimmel AR (2010) Chemotactic activation of
Dictyostelium AGC-family kinases AKT and PKBR1 requires separate
but coordinated functions of PDK1 and TORC2. J Cell Sci 123:
983–992
Lim CJ, Zawadzki KA, Khosla M, Secko DM, Spiegelman GB, Weeks G
(2005) Loss of the Dictyostelium RasC protein alters vegetative cell size,
motility and endocytosis. Exp Cell Res 306: 47–55
Mato JM, Losada A, Nanjundiah V, Konijn TM (1975) Signal input for a
chemotactic response in the cellular slime mold Dictyostelium
discoideum. Proc Natl Acad Sci USA 72: 4991–4993
Mato JM, van Haastert PJ, Krens FA, Rhijnsburger EH, Dobbe FC, Konijn TM
(1977) Cyclic AMP and folic acid mediated cyclic GMP accumulation
in Dictyostelium discoideum. FEBS Lett 79: 331–336
McRobbie SJ, Newell PC (1985) Effects of cytochalasin B on cell movements
and chemoattractant-elicited actin changes of Dictyostelium. Exp Cell
Res 160: 275–286
Postma M, van Haastert PJ (2009) Mathematics of experimentally generated
chemoattractant gradients. Methods Mol Biol 571: 473–488
Sasaki AT, Chun C, Takeda K, Firtel RA (2004) Localized Ras signaling at
the leading edge regulates PI3K, cell polarity, and directional cell
movement. J Cell Biol 167: 505–518
Stradal TEB, Rottner K, Disanza A, Confalonieri S, Innocenti M, Scita G (2004)
Regulation of actin dynamics by WASP and WAVE family proteins.
Trends Cell Biol 14: 303–311
Swaney KF, Huang CH, Devreotes PN (2010) Eukaryotic chemotaxis:
a network of signaling pathways controls motility, directional sensing,
and polarity. Annu Rev Biophys 39: 265–289
Tomchik KJ, Devreotes PN (1981) Adenosine 30,50-monophosphate
waves in Dictyostelium discoideum: a demonstration by isotope
dilution–fluorography. Science 212: 443–446
van Haastert PJ, Keizer-Gunnink I, Kortholt A (2007) Essential role of
PI3-kinase and phospholipase A2 in Dictyostelium discoideum
chemotaxis. J Cell Biol 177: 809–816
Van Haastert PJM, Devreotes PN (2004) Chemotaxis: signalling the way
forward. Nat Rev Mol Cell Biol 5: 626–634
Veltman DM, Keizer-Gunnik I, van Haastert PJ (2008) Four key signaling
pathways mediating chemotaxis in Dictyostelium discoideum. J Cell Biol
180: 747–753
Veltman DM, van Haastert PJ (2006) Guanylyl cyclase protein and cGMP
product independently control front and back of chemotaxing
Dictyostelium cells. Mol Biol Cell 17: 3921–3929
Wu LJ, Valkema R, Vanhaastert PJM, Devreotes PN (1995) The G-protein
b-subunit is essential for multiple responses to chemoattractants in
Dictyostelium. J Cell Biol 129: 1667–1675
Xiao Z, Zhang N, Murphy DB, Devreotes PN (1997) Dynamic distribution
of chemoattractant receptors in living cells during chemotaxis and
persistent stimulation. J Cell Biol 139: 365–374
Zhang S, Charest PG, Firtel RA (2008) Spatiotemporal regulation of Ras
activity provides directional sensing. Curr Biol 18: 1587–1593
Minimal requirements for chemotaxis
A. Kortholt et al
&2011EUROPEAN MOLECULAR BIOLOGY ORGANIZATIONEMBO reportsVOL 12 | NO 12 | 2011
scientificreport
1279