Content uploaded by Ryohei Furukawa
Author content
All content in this area was uploaded by Ryohei Furukawa on Jan 27, 2018
Content may be subject to copyright.
OUTSTANDING OBSERVATION
Two macrophage migration inhibitory factors regulate
starfish larval immune cell chemotaxis
Ryohei Furukawa1, Kana Tamaki2and Hiroyuki Kaneko1
Immune cell recruitment is critical step in the inflammatory response and associated diseases. However, the underlying
regulatory mechanisms are poorly understood in invertebrates. Mesenchyme cells of the starfish larvae, which allowed
Metchnikoff to complete his landmark experiments, are important model for analysis of immune cell migration. The present
study investigated the role of macrophage migration inhibitory factor (MIF)—an evolutionarily conserved cytokine that is
functionally similar to chemokines—in the larvae of the starfish Patiria (Asterina)pectinifera, which were found to possess two
orthologs, ApMIF1 and ApMIF2. ApMIF1 and ApMIF2 clustered with mammalian MIF and its homolog D-dopachrome
tautomerase (DDT), respectively, in the phylogenetic analysis. In contrast to the functional similarity between mammalian MIF
and DDT, ApMIF1 knockdown resulted in the excessive recruitment of mesenchyme cells in vivo, whereas ApMIF2 deficiency
inhibited the recruitment of these cells to foreign bodies. Mesenchyme cells migrated along a gradient of recombinant ApMIF2
in vitro, whereas recombinant ApMIF1 completely blocked ApMIF2-induced directed migration. Moreover, the expression patterns
of ApMIF1 and ApMIF2 messenger RNA in bacteria-challenged mesenchyme cells were consistent with in vivo observations of
cell behaviors. These results indicate that ApMIF1 and ApMIF2 act as chemotactic inhibitory and stimulatory factors,
respectively, and coordinately regulate mesenchyme cell recruitment during the immune response in starfish larvae. This is the
first report describing opposing functions for MIF- and DDT-like molecules. Our findings provide novel insight into the
mechanisms underlying immune regulation in invertebrates.
Immunology and Cell Biology (2016) 94, 315–321; doi:10.1038/icb.2016.6; published online 2 February 2016
The starfish is an echinoderm that is phylogenetically close to the
ancestral chordate. The transparent larvae possess a simple body
structure, which allowed Metchnikoff to complete his landmark
experiments that led to his theories of phagocytosis and
chemotaxis.1We previously described in detail the characteristics of
Patiria (Asterina)pectinifera mesenchyme cells, such as their distribu-
tion patterns and role in host defense.2,3 Most mesenchyme cells are
distributed beneath the larval body wall and form a dynamic network
structure while moving randomly. When a foreign substance is
injected into the larval blastocoel, it is completely encapsulated and/
or phagocytosed by mesenchyme cells converging from various
directions within 2 h. Based on these characteristics, mesenchyme
cells are functionally equivalent to blastocoelar cells in larval echinoids,
such as those arising from a subset of secondary mesenchyme cells in
sea urchin.4,5 Significantly, the number of recruited mesenchyme cells
is dependent on the amount and size of the foreign body,2suggesting
that this process is strictly regulated by the starfish larval immune
system.
Chemokines are essential in the mammalian immune system as
leukocyte recruitment and arrest are critical steps in the inflamma-
tory response and associated diseases.6On the other hand, little is
known about the mechanisms of chemotactic migration during
the invertebrate immune response. A recent comparative whole-
genome analysis failed to identify putative chemokine or chemokine
receptors in invertebrates,7implying that invertebrate chemotaxis is
regulated by chemokine-like function chemokines, which have func-
tions similar to those of classic chemokines but lack their typical
structure.8
Macrophage migration inhibitory factor (MIF) is a chemokine-like
function chemokine that is evolutionarily ancient and highly
conserved.9–11 In mammals, MIF directly exerts chemokine-like
functions via the CXC chemokine receptors CXCR2 and
CXCR4,12,13 and indirectly induces leukocyte arrest.14–16 Thus, MIF
acts as a pleiotropic inflammatory cytokine with critical roles in
physiological immunity as well as inflammatory diseases and cancer.8
MIF activity is shared by D-dopachrome tautomerase (DDT); this
includes binding the same receptor complex and induction of similar
cell signaling and effector functions.17–19 In contrast, the function of
invertebrate MIF and DDT—especially in the immune system—is not
well understood.
We predicted that the starfish MIF ortholog regulates the recruit-
ment of larval mesenchyme cells in the immune response. In the
1
Department of Biology, Research and Education Center for Natural Sciences, Keio University, Kanagawa, Japan and
2
School of Fundamental Science and Technology, Graduate
School of Science and Technology, Keio University, Kanagawa, Japan
Correspondence: Dr R Furukawa, Division of Biomedical Information Analysis, Iwate Tohoku Medical Megabank Organization, Iwate Medical University, 2-1-1 Nishitokuda,
Yahaba-cho, Shiwa-gun, Iwate 028-3694, Japan.
E-mail: rfuru@iwate-med.ac.jp or furyohei@gmail.com
Received 27 October 2015; revised 7 January 2016; accepted 7 January 2016; published online 2 February 2016
Immunology and Cell Biology (2016) 94, 315–321
&
2016 Australasian Society for Immunology Inc. All rights reserved 0818-9641/16
www.nature.com/icb
present study, we identified two MIF orthologs, designated as ApMIF1
and ApMIF2, and characterized their functions in the context of
immunoreactive migration of mesenchyme cells in vivo and in vitro.
We also propose a regulatory model for mesenchyme cell recruitment
by ApMIF1 and ApMIF2 in the starfish larval immune response.
RESULTS
Structural features of ApMIF1 and ApMIF2
Over 56 000 P. pectinifera expressed sequence tags have been sub-
mitted to the DNA Data Bank of Japan (http://www.ddbj.nig.ac.jp)
without annotations. Using the protein sequences of human
macrophage MIF and its homolog DDT to search this database using
the TBLASTN program, we identified two MIF orthologs in
P. pectinifera, hereafter referred to as ApMIF1 and ApMIF2 (GenBank
accession nos. LC062600 and LC062601, respectively). Based on the
obtained expressed sequence tag sequences, we isolated full-length
complementary DNA (cDNA)-encoding ApMIF1 and ApMIF2 from
the cDNA library of mesenchyme cells, and determined that they
encode 116- and 120-amino acid polypeptides, respectively. The
deduced amino acid sequences revealed two α-helices and six β-
strands, which are characteristic of other MIF family members
(Figure 1a). To assess the evolutionary relationships within the
deuterostome MIF family, we performed phylogenetic analysis of
12 MIF orthologs by applying the widely used neighbor-joining
and maximum likelihood methods (Figure 1b). Two distinct types
of deuterostome MIFs were observed with both methods: ApMIF1
clustered with vertebrate MIF but was distantly related to ApMIF2 and
vertebrate DDT.
Human MIF is a non-cognate ligand of CXCR2 and CXCR4.12 The
molecular details of the MIF/CXCR4 interaction are not well under-
stood; however, the pseudo-(E)LR motif and N-like loop are required
for the functioning of human MIF as a non-canonical CXCR2
ligand.13,20 The pseudo-(E)LR motif in MIF is characterized by Asp-
45 and Arg-12 (Figure 1a, arrowheads), and forms a three-
dimensional structure resembling the N-terminal Glu–Leu–Arg
(ELR) motif present in a subgroup of CXC chemokines. The N-like
loop in MIF spans 10 amino acids from positions 47 to 56, but is
structurally distinct from that of CXC chemokines (Figure 1a, black
bar). Interestingly, the N-like loop of ApMIF1 was 80% identical to
that of human MIF, whereas the loop in ApMIF2 was 40 and 30%
identical to those of human MIF and DDT, respectively. In addition, a
pseudo-(E)LR motif was found in ApMIF2 but not in ApMIF1.
ApMIF1 or ApMIF2 knockdown affects the recruitment of
mesenchyme cells during the immune response
We tested the efficacy of ApMIF1 and ApMIF2 knockdown using
morpholino antisense oligonucleotides (MO) for loss-of-function
experiments. Eggs injected with each of the MOs showed normal
morphogenesis, including distribution and migration of mesenchyme
cells in the blastocoel, formation of the mouth and
Figure 1 ApMIF1 and ApMIF2 sequences and phylogenetic relationships. (a) Sequence alignment of ApMIF1 and ApMIF2 with human MIF and DDT. The
locations of α-helices and β-strands are indicated; the N-like loop is indicated by a black bar; and the pseudo-(E)LR motif in ApMIF2 is indicated by
arrowheads. Alignment of the N-like loop is also shown in a box, and hyphens denoting identical residues. (b) Phylogenetic trees of deuterostome MIFs using
either the neighbor-joining method (left) or the maximum likelihood method (right). All ambiguous positions were removed for each sequence pair. The
percentage of trees in which associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches. The collapsed branch
in the Maximum Likelihood Tree has a bootstrap value o50. Scale bars: left, evolutionary distance; right, substitutions per site.
Regulation of chemotaxis by starfish MIF
RFurukawaet al
316
Immunology and Cell Biology
compartmentalization of the digestive tract (Supplementary Figure
S1a). However, compared with larvae injected with control MO,
ApMIF1-MO- and ApMIF2-MO-injected larvae were smaller and had
fewer cells (Supplementary Figures S1a and c). This decrease in size
was rescued by co-injecting the corresponding messenger RNA
(mRNA) (Supplementary Figure S1b), confirming the specificity of
knockdown by each MO. In subsequent experiments, ApMIF1 and
ApMIF2 loss of function was determined as a larval size of o700 μm
in length.
To examine the roles of ApMIF1 and ApMIF2 in the immune
response—especially the migration of mesenchyme cells—we injected
a large oil droplet (diameter =20 μm) into the center of the head
blastocoel of morphants, which is the area that is furthest from the
body wall (~50–80 μm from the body wall; Supplementary Figure
S1a). In vivo time-lapse observation revealed an exaggerated immune
response in ApMIF1-MO larvae, with excessive recruitment
of mesenchyme cells as compared with control-MO larvae
(Supplementary Movies S1 and S2). In contrast, fewer mesenchyme
cells were present near the oil droplet in ApMIF2-MO than in control-
MO larvae (Supplementary Movies S1 and S3). A quantitative analysis
of mesenchyme cells recruited to the oil droplet revealed that the total
number of cells attached to or surrounding the oil droplet increased
over time in control-MO larvae (Figures 2a and b); 2 h after injection,
6–8 cells were attached to the droplet (Figure 2a), consistent with
previous observations.2By comparison, the number of mesenchyme
cells surrounding the oil droplet was higher in ApMIF1-MO larvae
and reached a plateau 45 min after oil droplet injection, with similar
numbers observed up to 2 h after injection. In contrast, in ApMIF2-
MO larvae, fewer cells contacted the oil droplet as compared with
control-MO larvae 45 min after injection, and this did not change
until the end of the observation period.
Recombinant ApMIF1 and ApMIF2 affect mesenchyme cell
migration
To assess the contribution of ApMIF1 and ApMIF2 to the migration of
mesenchyme cells, we prepared the recombinant proteins rMIF1 and
rMIF2 (Supplementary Figure S2) and carried out a chemotactic
migration assay with a Zigmond chamber. To eliminate the influence
of endogenous MIFs in the assay, we prepared mesenchyme cells from
larvae in which both ApMIF1 and ApMIF2 were knocked down. Cell
migration was evaluated by tracking the movements of individual cells
in the presence of recombinant rMIF1 and rMIF2 (Figure 3). As the
fusogenic nature and network-forming ability of mesenchyme cells
made it difficult to accurately track their migration, we excluded cells
that fused to each other during the observation period.
Exposure to rMIF2 but not rMIF1 increased cell migration speed
and cell directionality within 20 min (Figures 3a and b). When rMIF2
was applied as a gradient, cells migrated toward the source (Figure 3c,
top). These migration patterns can be seen in Rose plots that show the
frequency distribution of cell orientations (Figure 3c, bottom).
Rayleigh tests confirmed that mesenchyme cells migrated randomly
when rMIF2 concentration was uniform (P40.05), similar to cells in
artificial seawater (ASW). However, when rMIF2 was applied as a
gradient, the direction of migration was strongly biased toward the
region of high-rMIF2 concentration (Figure 3c), Po0.05. On the
other hand, when an rMIF2 gradient was formed under a homo-
geneous concentration of rMIF1, mesenchyme cells maintained a
random migration pattern (P40.05, Rayleigh test), and velocity and
directionality were similar to those of migrating cells in ASW.
Figure 4 shows the change in migration speed every 10 min during
the chemotaxis assay. In the presence of an rMIF2 gradient, the
migration speed of mesenchyme cells gradually increased and attained
a maximum after 30–40 min. However, when both rMIF1 and rMIF2
were applied, no significant difference was noted in migration speed
relative to cells in ASW at any time point.
ApMIF1 and ApMIF2 mRNA expression levels are differentially
upregulated by bacterial challenge
To ascertain whether gene expression under immune stimulation can
explain the mesenchyme cell behaviors regulated by the recombinant
protein, we examined changes in temporal mRNA expression profiles
of ApMIF1 and ApMIF2 in response to bacterial challenge. ApMIF1
mRNA levels were increased about 4.3-fold at 5 min as compared with
that observed in naive mesenchyme cells at 0 min and remained
constant until 90 min, when the levels became extremely high
(41700-fold higher than baseline) (Figure 5a). ApMIF2 mRNA
expression increased ~ 100-fold within 15 min, and decreased after
90 min (Figure 5a). The mRNA expression of ApMIF2 was higher
than that of ApMIF1 until 60 min; however, after 90 min, the level
of ApMIF1 mRNA was ~ 10-fold higher than that of ApMIF2
(Figure 5b).
Figure 2 Recruitment of mesenchyme cells to injected oil droplets. (a) Number of mesenchyme cells attached to oil droplets. (b) Number of mesenchyme
cells within 20 μm of the surface of the oil droplet without contact. **Po0.01, ***Po0.001 (ttest; n=27 (control-MO), 37 (ApMIF1-MO) and 37
(ApMIF2-MO)).
Regulation of chemotaxis by starfish MIF
R Furukawa et al
317
Immunology and Cell Biology
DISCUSSION
In this study, we investigated the functional properties of two starfish
MIF orthologs, ApMIF1 and ApMIF2, in the context of immune cell
recruitment in larvae. ApMIF1 and ApMIF2 clustered with MIF and
DDT, respectively, in the phylogenetic analysis. In contrast to the
functional similarity between mammalian MIF and DDT, the in vitro
chemotactic migration assay revealed that rMIF1 and rMIF2 had
antagonistic functions, with the former inhibiting and the latter
inducing directed mesenchyme cell migration. This result was con-
sistent with in vivo observations that ApMIF2 knockdown abrogated
mesenchyme cell migration toward a foreign body, whereas ApMIF1
deficiency led to excessive cell recruitment. Moreover, the expression
levels of ApMIF1 and ApMIF2 mRNA, which were upregulated by
bacterial stimulation, mirrored mesenchyme cell behaviors in vivo.
These results indicate that ApMIF1 and ApMIF2 act as chemotaxis
Figure 3 Regulation of chemotactic mesenchyme cell migration by ApMIF1 and ApMIF2. (a–c) Cells seeded in the Zigmond chamber were tracked by time-
lapse microscopy for 1 h in ASW, a homogeneous concentration of rMIF1 (rMIF1 constant), a homogeneous concentration of rMIF2 (rMIF2 constant), a
gradient of rMIF2 (rMIF2 gradient) or a gradient of rMIF2 in a homogeneous concentration of rMIF1 (rMIF1 constant+rMIF2 gradient). Cell velocity (that is,
accumulated distance/observation time) (a) and directionality (that is, Euclidean distance/accumulated distance) (c) were calculated from tracks o f individual
migrating cells (plotted in the upper row of c). **Po0.01 (ttest). Rose diagrams (bottom row of c) were also plotted to visualize the distribution of migration
angles. The Rayleigh test was used to evaluate unimodal distribution of cell direction at end points. P40.05 was considered as a uniform distribution
(random migration).
Figure 4 Changes in mesenchyme cell migration speed. Migration speed per
10 min was calculated from tracks of individual migrating cells shown in
Figure 3c. **Po0.01 vs ASW control (ttest).
Regulation of chemotaxis by starfish MIF
RFurukawaet al
318
Immunology and Cell Biology
inhibitory and stimulatory factors in the larval immune system,
respectively.Thisisthefirst report describing opposing functions
for MIF- and DDT-like molecules in the immune response.
Based on these findings, we generated a model of mesenchyme
cell recruitment by ApMIF1 and ApMIF2 (Figure 6). The first
contact between randomly migrating cells and a foreign body triggers
a dramatic upregulation of ApMIF2 gene expression and
ApMIF2 secretion, creating an ApMIF2-rich environment around
the foreign substance. In ApMIF1 morphants, the number of recruited
mesenchyme cells reached a plateau 45 min after the oil droplet
injection, suggesting that all cells within the range of ApMIF2 activity
had been recruited by that time. This activity range was estimated to
be 70 μm based on the migration speed of rMIF2-treated mesenchyme
cells (1.4–2.2 μmmin−1; Figure 4), which corresponds to the distance
from the body wall to the central area of the blastocoel where the oil
droplet was injected (50–80 μm, Supplementary Figure S1a). This
indicates that ApMIF2 can recruit a sufficient number of mesenchyme
cells from the body wall where they are mainly distributed irrespective
of where the foreign body is located. On the other hand, ApMIF1 is
secreted by cells undergoing chemotactic migration to inhibit the
activity of ApMIF2 and prevent the recruitment of more distant cells.
Indeed, the upregulation of ApMIF1 and ApMIF2 mRNA expression
followed similar trends. Finally, when an appropriate number of
mesenchyme cells had been recruited, ApMIF1 completely inhibited
ApMIF2 activity. According to our model, the activities of the two
MIFs can regulate the recruitment of a sufficient number of
mesenchyme cells for the amount and size of the foreign body.2
Intriguingly, the cooperative regulation of the chemotaxis by
ApMIF1 and ApMIF2 implies a trade-off between efficiency and risk
avoidance. The requirement for inhibition by ApMIF1 indicates that
the immune system of starfish larvae is at risk for excessive or
insufficient accumulation of mesenchyme cells if there is an imbalance
in the levels of ApMIF1 and ApMIF2. Risks of this nature are a feature
of various cytokine systems; for example, sepsis development is closely
related to an imbalance between the levels of pro- and anti-
inflammatory cytokine including MIF.21 On the other hand, we
previously reported that foreign materials can easily invade the
blastocoel of starfish larvae through the body wall;2given this
susceptibility, the inhibitory effect of ApMIF1 might contribute to
preserving part of the structural network, providing protection against
sequential invasion of multiple foreign bodies. Although the antag-
onistic activities of cytokines can be very effective for a subset of
immune cells to localize immune responses, there is a constant risk of
immunological disequilibrium even in a relatively primitive animal.
We recently identified the P. pectinifera ortholog of mammalian
dedicator of cytokinesis (DOCK)2, an intracellular regulator of
mesenchyme cell migration in the immune response.3DOCK2 is
activated downstream of the non-cognate MIF receptors CXCR2 and
CXCR412 to regulate immune cell chemotaxis.22,23 Our previous work
revealed that loss of ApDOCK function also resulted in the failure of
mesenchyme cell recruitment,3similar to ApMIF2-MO larvae
Figure 5 Temporal changes in ApMIF1 and ApMIF2 mRNA expression in cultured mesenchyme cells following bacterial challenge. (a) Relative expression
levels of ApMIF1 and ApMIF2 mRNA in bacteria-challenged and naive mesenchyme cells (0 min). (b) Temporal changes in the ratio of ApMIF2 to ApMIF1
mRNA expression (n=3).
Figure 6 Model of the control of chemotaxis by ApMIF1 and ApMIF2.
Mesenchyme cells move randomly before an immune response. Within
20 min of invasion of a foreign body (that is, an oil droplet) (a), a nearby
mesenchyme cell makes initial contact with the foreign body and secretes
ApMIF2 (b). Adjacent cells are activated by ApMIF2 binding and move along
the ApMIF2 gradient toward the source (c). Migrating cells secrete ApMIF1,
which inhibits chemotaxis in other more distal mesenchyme cells (c).
Consequently, an immune response is initiated by an appropriate number of
mesenchyme cells (d).
Regulation of chemotaxis by starfish MIF
R Furukawa et al
319
Immunology and Cell Biology
(Figure 2 and Supplementary Movie S3), implying that ApMIF2
triggers while ApMIF1 inhibits ApDOCK-mediated signaling. Despite
the evolutionarily conserved roles of MIF and DOCK family members
in immune cell recruitment, to date there have been no orthologs of
mammalian MIF receptors identified in invertebrates. This raises the
possibility of conserved canonical MIF receptor(s); given that ApMIF1
and ApMIF2 have motifs for chemokine receptor binding, other types
of G protein-coupled receptor superfamily members may be con-
sidered as candidates.
In summary, the identification of starfish MIF receptor(s) not only
provides insight into the evolutionary conservation and diversity of
MIF function, but it also clarifies the mechanism by which ApMIF1
and ApMIF2 exert their opposing functions. In addition, the starfish
MIF receptor(s) can provide a useful tool for investigating treatment
strategies for MIF-related diseases.
METHODS
Animals
Adult P. pectinifera were collected along the coast of Japan and maintained at
15 °C in a glass aquarium filled with ASW (MarineArt SF-1; Tomita
Pharmaceutical, Tokushima, Japan). Developing larvae were obtained as
previously described.2In brief, mature eggs prepared by treatment with 1-
methyladenine (Sigma-Aldrich, Tokyo, Japan) were fertilized with diluted
sperm, and were allowed to develop in ASW at 20 °C.
Cloning and identification of ApMIF1 and ApMIF2
To identify starfish orthologs of MIF family members, we carried out
TBLASTN searches of456 000 expressed sequence tags of P. pectinifera eggs
and gastrulae deposited in the DNA Data Bank of Japan (GenBank accession
no. DB384948–DB441778) using amino acid sequences of human MIF
(NP_002406) and DDT (NP_001346) as queries with e-value scoreso1e −5.
In total, 57 candidate sequences were obtained from the TBLASTN hits that
encoded two different sequences. Based on these sequences, we designed the
following two primer pairs: 5′-CAG CGA ATC TTC CTG AAT TGT CAA
G-3′/5′-CAG TTT TCA CGG CTA CAC TAA CTT G-3′and 5′-GTG
GAG AAA TTG CTT GTC GTA-3′/5′-GGG AGT TGG GGT GGT GCA
G-3′.Amplified cDNAs were subcloned from a cDNA library of cultured
mesenchyme cells into the pGEM-T vector (Promega, Tokyo, Japan) and
sequenced as previously described.24 Deduced amino acid sequences were
analyzed with the BLAST algorithm, and we identified one ortholog each of
MIF and DDT. Phylogenetic trees were reconstructed with the neighbor-joining
and maximum likelihood methods based on CLUSTALW-aligned MIF/DDT
sequences using MEGA6 software.25 All positions containing gaps and missing
data were eliminated. For the neighbor-joining tree, evolutionary distances were
computed with the p-distance amino acid substitution model using uniform
rates among sites and pairwise deletion. For the maximum likelihood tree, we
applied the LG+G model, which was the best fit substitution model estimated
by the lowest Bayesian Information Criterion scores, and an initial tree for the
heuristic search was generated automatically by applying Neighbor-Joining and
BioNJ algorithms to a matrix of estimated pairwise distances. Both methods use
the bootstrap method with 1000 replicates for phylogeny testing.
Morpholino oligonucleotides
MOs were purchased from Gene Tools (Philomath, OR, USA). ApMIF1- and
ApMIF2-MO sequences were 5′-TCTACGATCGGCATCTTGACAATTC-3′
and 5′-TCG CAC AGAGGCATGTTGATTCTTG-3′, respectively, and were
complementary to mRNA sequences containing the translation start site
(underlined). Standard Control Morpholino Oligo (Gene Tools) was used as
the control-MO. MOs were dissolved in sterile water at 1 mMto obtain a stock
solution, which was diluted to 240 μMbefore injection. For simultaneous
knockdown of ApMIF1 and ApMIF2, equal volumes of each MO at 480 μM
were mixed. In brief, follicle cells were removed from immature eggs using
calcium-free seawater (Jamarin, Osaka, Japan), and the eggs were injected with
each MO in a microinjection chamber.26 The detailed injection procedure has
been described elsewhere.27
Microinjection and in vivo analysis of mesenchyme cell migration
Silicone oil (Shin-Etsu Chemical Co., Tokyo, Japan) was injected into larvae as
previously described.2Injected larvae were immobilized and observed at 20 °C
by time-lapse using an IX71 light microscope (Olympus, Tokyo, Japan)
controlled with DP controller software (Olympus).3Images were acquired
every 2 min for 2 h. The larvae were then labeled with a monoclonal antibody
against MC5—a membranous metalloproteinase of the astacin family27 and
mesenchyme cell marker—combined with propidium iodide nuclear staining
(0.1 μgml
−1; Wako Pure Chemical Industries, Osaka, Japan).2Specimens were
visualized by laser confocal microscopy with FluoView software (Olympus).
Preparation of recombinant proteins
DNA fragments encoding ApMIF1 or ApMIF2 were amplified using the
following forward and reverse primer pairs: for rMIF1, 5′-CAC CAT
GCC GAT CGT AGA-3′and 5′-TCA TAT TTT GCC AGC AAG TTT
GTC-3′and for rMIF2, 5′-CAC CAT GCC TCT GTG CGA GCT GAA-3′and
5′-TTA TTT CAG TTG GCT GGC GAG TTT GCC-3′. PCR products were
subcloned into the pET100/D-TOPO vector (Invitrogen, Tokyo, Japan), and
the recombinant proteins (rMIF1 and rMIF2) were expressed in Escherichia coli
strain BL21 and purified by metal-chelate affinity chromatography using the
His60 Ni Gravity Column Purification kit (Clontech, Shiga, Japan) according to
the manufacturer’s instructions.
Chemotaxis assay
Mesenchyme cells were obtained from morphant or control larvae as previously
described.3The chemotaxis assay was performed using a Zigmond chamber
(Funakoshi, Tokyo, Japan). The assembled chamber contained a gap of exactly
5μm between the surface of the bridge and coverslip on which cells were
cultured. An rMIF2 gradient was formed by placing ASW and 10 nMrMIF2-
containing ASW in separate wells. In the case of no gradient, both wells were
filled with ASW or 10 nMrMIF-containing ASW. To obtain a stable gradient,
the chamber was allowed to stand for 30–60 min at 20 °C under humid
conditions before observation. Time-lapse images were acquired every 2 min
for 1 h. Mesenchyme cells were tracked in two independent experiments using
the Manual Tracking plugin of ImageJ software (National Institutes of Health,
Bethesda, MD, USA). Tracks were analyzed using the Chemotaxis and
Migration Tool (Image J) to obtain cell speed and directionality, which
represents a ratio of Euclidean distance to accumulated distance and shows
the direction of migration. The Rayleigh test for unimodal direction clustering
was used to determine whether the cell population showed directionality along
the chemotactic gradient;28 the distribution of directions was considered
uniform for P40.05.
Bacterial challenge
Cultured E. coli were adjusted to a concentration of OD600=0.5 in ASW and
boiled for 5 min. A 100-μl volume of E. coli-ASW was added to cultured
mesenchyme cells in a plastic culture dish (diameter, 6 cm; BD, Tokyo, Japan)
with 5 ml ASW.
Real-time PCR
Total RNA isolation and a cDNA synthesis were performed as previously
described.24 A140-bpfragmentoftheApMIF1 gene and a 137-bp fragment of
the ApMIF2 gene were amplified using the primer pairs 5′-TGC CCA
AGT CAT CAC AGA AA-3′/5′-TTT TGC CAG CAA GTT TGT CA-3′and
5′-CAG CGG ATT CAG CAG TAA ACA C-3′/5′-TTG CCT CAA TGA
GAT GAA AGG A-3′, respectively. As a reference gene, a 114-bp fragment of
the P. pectinifera 18 S rRNA gene was amplified using the previously reported
18 S-3-f and 18 S-3-r primers.24 Real-time PCR was carried out using the
KAPA SYBR FAST qPCR master mix (Kapa Biosystems, Woburn, MA, USA)
and a StepOne Real-Time PCR system (Applied Biosystems, Foster City,
CA, USA).
Regulation of chemotaxis by starfish MIF
RFurukawaet al
320
Immunology and Cell Biology
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGEMENTS
We thank members of the Asamushi Marine Biological Station of Tohoku
University for supplying the starfish. We also thank members of the Tateyama
Marine Laboratory, Marine Coastal Research Center, Ochanomizu University
for maintaining the adult starfish. We are particularly grateful to Drs Hiroshi
Takeda, Motonori Hoshi, Midori Matsumoto, Ritsu Kuraishi and Kazuya
Kobayashi for critical discussions. This work was supported by Keio Gijuku
Academic Development Funds (to RF and HK).
1MetchnikoffE.Lectures on the Comparative Pathology of Inflammation, Delivered at
Pasteur Institute: Dover: New York, 1891 (reprinted in 1968).
2 Furukawa R, Takahashi Y, Nakajima Y, Dan-Sohkawa M, Kaneko H. Defense system by
mesenchyme cells in bipinnaria larvae of the starfish, Asterina pectinifera. Dev Comp
Immunol 2009; 33:205–215.
3 Furukawa R, Funabashi H, Matsumoto M, Kaneko H. Starfish ApDOCK protein
essentially functions in larval defense system operated by mesenchyme cells. Immunol
Cell Biol 2012; 90:955–965.
4 Tamboline CR, Burke RD. Secondary mesenchyme of the sea urchin embryo: ontogeny
of blastocoelar cells. J Exp Zool 1992; 262:51–60.
5 Silva JR. The onset of phagocytosis and identity in the embryo of Lytechinus variegatus.
Dev Comp Immunol 2000; 24:733–739.
6 Ley K, Laudanna C, Cybulsky MI, Nourshargh S. Getting to the site of inflammation: the
leukocyte adhesion cascade updated. Nat Rev Immunol 2007; 7:678–689.
7 DeVries ME, Kelvin AA, Xu L, Ran L, Robinson J, Kelvin DJ. Defining the origins and
evolution of the chemokine/chemokine receptor system. J Immunol 2006; 176:
401–415.
8 Tillmann S, Bernhagen J, Noels H. Arrest functions of the MIF ligand/receptor axes in
atherogenesis. Front Immunol 2013; 4:115.
9 Sato A, Uinuk-ool TS, Kuroda N, Mayer WE, Takezaki N, Dongak R et al. Macrophage
migration inhibitory factor (MIF) of jawed and jawless fishes: Implications for its
evolutionary origin. Dev Comp Immunol 2003; 27:401–412.
10 Vermeire JJ, Cho Y, Lolis E, Bucala R, Cappello M. Orthologs of macrophage migration
inhibitory factor from parasitic nematodes. Trends Parasitol 2008; 24:355–363.
11 Li F, Huang S, Wang L, Yang J, Zhang H, Qiu L et al. A macrophage migration inhibitory
factor like gene from scallop Chlamys farreri: involvement in immune response and
wound healing. Dev Comp Immunol 2011; 35:62–71.
12 Bernhagen J, Krohn R, Lue H, Gregory JL, Zernecke A, Koenen RR et al. MIF is a
noncognate ligand of CXC chemokine receptors in inflammatory and atherogenic cell
recruitment. Nat Med 2007; 13:587–596.
13 Kraemer S, Lue H, Zernecke A, Kapurniotu A, Andreetto E, Frank R et al. MIF-
chemokine receptor interactions in atherogenesis are dependent on an n-loop-based 2-
site binding mechanism. FASEB J 2011; 25:894–906.
14 Gregory JL, Leech MT, David JR, Yang YH, Dacumos A, Hickey MJ. Reduced leukocyte-
endothelial cell interactions in the inflamed microcirculation of macrophage migration
inhibitory factor-deficient mice. Arthritis Rheum 2004; 50: 3023–3034.
15 Fan H, Hall P, Santos LL, Gregory JL, Fingerle-Rowson G, Bucala R et al. Macrophage
migration inhibitory factor and CD74 regulate macrophage chemotactic responses via
MAPK and rho GTPase. JImmunol2011; 186:4915–4924.
16 Santos LL, Fan H, Hall P, Ngo D, MacKay CR, Fingerle-Rowson G et al. Macrophage
migration inhibitory factor regulates neutrophil chemotactic responses in inflammatory
arthritis in mice. Arthritis Rheum 2011; 63:960–970.
17 Coleman AM, Rendon BE, Zhao M, Qian M-WW, Bucala R, Xin D et al. Cooperative
regulation of non-small cell lung carcinoma angiogenic potential by macrophage
migration inhibitory factor and its homolog, D-dopachrome tautomerase. JImmunol
2008; 181: 2330–2337.
18 Xin D, Rendon BE, Zhao M, Winner M, Coleman AM, Mitchell RA. The MIF homologue
D-dopachrome tautomerase promotes COX-2 expression through beta-catenin-
dependent and - independent mechan isms. Mol Cancer Res 2010; 8:1601–16 09.
19 Merk M, Zierow S, Leng L, Das R, Du X, Schulte W et al. The D-dopachrome
tautomerase (DDT) gene product is a cytokine and functional homolog of macrophage
migration inhibitory factor (MIF).”.Proc Natl Acad Sci USA 2011; 108:E577–E585.
20 Weber C, Kraemer S, Drechsler M, Lue H, Koenen RR, Kapurniotu A et al. Structural
determinants of MIF functions in CXCR2-mediated inflammatory and atherogenic
leukocyte recruitment. Proc Natl Acad Sci USA 2008; 105: 16278–16283.
21 Schulte W, Bernhagen J, Bucala R. Cytokines in sepsis: potent immunoregulators
and potential therapeutic targets—an updated view. Mediators Inflamm 2013; 2013:
165974.
22 Reif K, Cyster J. The CDM protein DOCK2 in lymphocyte migration. Trends Cell Biol
2002; 12:368–373.
23 Sai J, Raman D, Liu Y, Wikswo J, Richmond A. Parallel phosphatidylinositol 3-kinase
(PI3K)-dependent and Src-dependent pathways lead to CXCL8-mediated Rac2 activa-
tion and chemotaxis. JBiolChem2008; 283:26538–26547.
24 Furukawa R, Matsumoto M, Kaneko H. Characterization of a scavenger receptor
cysteine-rich-domain-containing protein of the starfish, Asterina pectinifera:ApSRCR1
acts as an opsonin in the larval and adult innate immune systems. Dev Comp Immunol
2012; 36:51–61.
25 Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular evolutionary
genetics analysis version 6.0. MolBiolEvol2013 ; 30:2725–2739.
26 Keihart O. Microinjection of echinoderm eggs: apparatus and procedures. Methods Cell
Biol 1982; 25:13–31.
27 Hamanaka G, Matsumoto M, Hoshi M, Kaneko H. Studies on function of MC5
molecule that is a novel membrane-type metalloproteinase of astacin family
during morphogenesis of the starfish, Asterina pectinifera. In: Harris LG, Bottger SA,
Walker CW, Lesser MP (eds). Echinoderms. Taylor & Francis Group: London. 2 010,
pp 457–462.
28 Zicha D, Dunn GA. Are growth factors chemotactic agents? Exp Cell Res 1995; 221:
526–529.
The Supplementary Information that accompanies this paper is available on the Immunology and Cell Biology website (http://www.nature.com/icb)
Regulation of chemotaxis by starfish MIF
R Furukawa et al
321
Immunology and Cell Biology
