![[ 125 I]INSL3 Binding to, and Cell Surface Expression](https://www.researchgate.net/profile/Daniel-Scott-6/publication/6357316/figure/fig3/AS:277692030177311@1443218348038/125-IINSL3-Binding-to-and-Cell-Surface-Expression_Q320.jpg)
![[ 125 I]INSL3 Competition Binding to LGR8 and LGR8 D227A with Mutant INSL3 Peptides A, Competition of [ 125 I]INSL3 binding to LGR8 using INSL3, INSL3 Ala-B16, INSL3 Ala-B19, INSL3 Ala B12/16/20 , and INSL3](https://www.researchgate.net/profile/Daniel-Scott-6/publication/6357316/figure/fig4/AS:277692038565897@1443218349638/125-IINSL3-Competition-Binding-to-LGR8-and-LGR8-D227A-with-Mutant-INSL3-Peptides-A_Q320.jpg)
Content uploaded by Daniel J Scott
Author content
All content in this area was uploaded by Daniel J Scott
Content may be subject to copyright.
Defining the LGR8 Residues Involved in Binding
Insulin-Like Peptide 3
Daniel J. Scott, Tracey N. Wilkinson, Suode Zhang, Tania Ferraro, John D. Wade,
Geoffrey W. Tregear, and Ross A. D. Bathgate
Howard Florey Institute (D.J.S., T.N.W., S.Z., T.F., J.D.W., G.W.T., R.A.D.B.) and Department of
Biochemistry and Molecular Biology (D.J.S., T.N.W., J.D.W., G.W.T., R.A.D.B.), The University of
Melbourne, Melbourne, Victoria 3031, Australia
The peptide hormone insulin-like peptide 3 (INSL3)
is essential for testicular descent and has been
implicated in the control of adult fertility in both
sexes. The human INSL3 receptor leucine-rich re-
peat-containing G protein-coupled receptor 8
(LGR8) binds INSL3 and relaxin with high affinity,
whereas the relaxin receptor LGR7 only binds re-
laxin. LGR7 and LGR8 bind their ligands within the
10 leucine-rich repeats (LRRs) that comprise the
majority of their ectodomains. To define the pri-
mary INSL3 binding site in LGR8, its LRRs were
first modeled on the crystal structure of the Nogo
receptor (NgR) and the most likely binding surface
identified. Multiple sequence alignment of this sur-
face revealed the presence of seven of the nine
residues implicated in relaxin binding to LGR7. Re-
placement of these residues with alanine caused
reduced [
125
I]INSL3 binding, and a specific pep
-
tide/receptor interaction point was revealed using
competition binding assays with mutant INSL3
peptides. This point was used to crudely dock the
solution structure of INSL3 onto the LRR model of
LGR8, allowing the prediction of the INSL3 Trp-B27
binding site. This prediction was then validated
using mutant INSL3 peptide competition binding
assays on LGR8 mutants. Our results indicated
that LGR8 Asp-227 was crucial for binding INSL3
Arg-B16, whereas LGR8 Phe-131 and Gln-133 were
involved in INSL3 Trp-B27 binding. From these two
defined interactions, we predicted the complete
INSL3/LGR8 primary binding site, including inter-
actions between INSL3 His-B12 and LGR8 Trp-177,
INSL3 Val-B19 and LGR8 Ile-179, and INSL3 Arg-
B20 with LGR8 Asp-181 and Glu-229. (Molecular
Endocrinology 21: 1699–1712, 2007)
R
ELAXIN AND INSULIN-LIKE peptide-3 (INSL3)
are closely related peptide hormones with a con-
served two-chain (A and B chain) structure linked by
disulphide bonds (1). INSL3 is abundantly expressed
in the fetal testis and is involved in mediation of fetal
gonad translocation to the inguinal canal during de-
velopment (2, 3). Furthermore, evidence suggests that,
in adults, INSL3 is involved in reproductive function,
with studies in rats demonstrating that INSL3 can ini-
tiate oocyte maturation in the ovary, whereas it sup-
presses germ cell apoptosis in the testes (4).
The relaxin and INSL3 receptors are leucine-rich
repeat-containing G protein-coupled receptor 7
(LGR7) and LGR8, respectively (5) [also named relaxin
family peptide receptor RXFP1 and RXFP2, respec-
tively (6)]. Similar to their fellow LGR family members,
the glycoprotein hormone receptors LH receptor, FSH
and TSH receptors, LGR7, and LGR8 possess large
extracellular ectodomains containing 10 leucine-rich
repeats (LRRs), within which lies the primary hormone
binding site (7–14). It is specific residues in the B
chains of relaxin and INSL3 that bind to the LRRs, and
although INSL3 has a very poor affinity for LGR7,
relaxins from some species will bind with high affinity
to LGR8.
LGR7 and LGR8 ligand-mediated activation is com-
plex, with several crucial steps required to stimulate
signaling. Although primary hormone binding to the
LRRs is the initial stimulus of receptor signaling, at
least two secondary interactions are then required to
induce signal transduction. A lower-affinity interaction
between the hormone and the transmembrane do-
mains of the receptor is required, whereas the LDLa
module of each receptor is essential for receptor ac-
tivation (13–16). Importantly, hormone binding to the
LRRs of these receptors is not affected by the confor-
mational state of the other parts of the receptor. Pri-
mary ligand binding to LGR7 and LGR8 had been
convincingly demonstrated to be independent of the
transmembrane domains, their LDLa modules, and the
G protein coupling state of each receptor (13, 16, 17).
Thus, we were confident that the primary binding of
INSL3 to the LRRs of LGR8 could be investigated
independently of receptor signaling.
LRRs are a common structural repeat found in pro-
teins with a wide range of functions. Each LRR unit
First Published Online May 1, 2007
Abbreviations: Blast, Basic local alignment search tool;
INSL3, insulin-like peptide 3; LGR, leucine-rich repeat-con-
taining G protein-coupled receptor; LRR, leucine-rich repeat;
NgR, Nogo receptor; PDB, Protein Data Bank.
Molecular Endocrinology is published monthly by The
Endocrine Society (http://www.endo-society.org), the
foremost professional society serving the endocrine
community.
0888-8809/07/$15.00/0 Molecular Endocrinology 21(7):1699–1712
Printed in U.S.A. Copyright © 2007 by The Endocrine Society
doi: 10.1210/me.2007-0097
1699
consists of a

-strand and an antiparallel linear ex-
tended structure connected by short loops (18). These
units stack together so that the

-strands of consec-
utive LRRs lie parallel to each other, forming an arced
solenoid-like structure with a concave

-sheet lining
the inside surface (19, 20). The inner face of each LRR
generally adheres to a Lx
1
x
2
Lx
3
Lx
4
x
5
N consensus,
with L representing hydrophobic residues, forming the
strong hydrophobic interior of the structure, and x
representing any amino acid, each exposed on the
concave surface of the LRR superstructure (18). In the
glycoprotein hormone receptors, it is various x resi-
dues from multiple LRRs that form the high affinity
hormone binding sites (21–26). Recently, x residues
from LRRs IV–VIII of LGR7 were found to be important
for high-affinity relaxin binding. Three residues pro-
jecting from the
␣
-helix of the B-chain of relaxin are
involved in receptor binding (Arg-B13, Arg-B17, and
Ile-B20) and have been proposed to interact with two
acidic pockets and one hydrophobic cluster in the
LRRs of LGR7 (27). However, the residues in the LGR8
LRRs involved in INSL3 binding are unknown.
We recently demonstrated the importance of INSL3
His-B12, Arg-B16, Val-B19, and Arg-B20 for high-af-
finity binding to LGR8 (28). Together, these comprise a
relaxin-like binding cassette in INSL3. Unlike relaxin,
however, these residues are not the only amino acids
responsible for high-affinity binding. Trp-B27, at the
C-terminal tail of the B-chain of INSL3, is well charac-
terized as a critical residue for INSL3 binding to LGR8
(28–30). The unique importance of INSL3 Trp-B27 may
explain why INSL3 cannot bind to LGR7 and suggests
that INSL3 binds LGR8 in a different manner to relaxin
on LGR7. In this paper, we used homology modeling,
interspecies sequence conservation, receptor muta-
tional analyses, and competition binding assays with
mutant INSL3 peptides to define the particular LRR
residues involved in the primary INSL3/LGR8
interaction.
RESULTS
Structural Prediction of LRR Domains
BlastP (basic local alignment search tool) alignment of
the LRRs of LGR7 (residues 71–349) and LGR8 (resi-
dues 68–345), as numbered from the end of the signal
peptide, against the National Center for Biotechnology
Information protein database revealed that the Nogo
receptor (NgR) was the most homologous protein with
a known crystal structure (47% amino acid similarity
and 31% identity to the LRRs of LGR7 and LGR8)
(data not shown). This homology was predominantly
attributable to the conserved hydrophobic residues
that are staggered through the consensus sequence
for typical type LRRs (LxxLxxLxLxxNxLxxLxxxoFxx, in
which x represents any residue, o represents any non-
polar residue, and the

-strand is underlined) (18).
The LRR protein sequences of LGR7 and LGR8 from
human (Homo sapiens), rhesus monkey (Macaca mu-
latta), chimpanzee (Pan troglodytes), dog (Canis famil-
iaris), mouse (Mus musculus), rat (Rattus norvegicus),
cow (Bos taurus), opossum (Monodelphis domestica),
and human NgR were aligned using ClustalW. An
alignment of the residues predicted to be on the inner
face of the LRRs of LGR7 and LGR8 (xxLxLxxN in the
consensus sequence) is presented in Fig. 1.
Residues corresponding to x positions in the xx-
LxLxxN consensus sequence, protrude out from the
inner surface of the LRR structure. Nine x residues
from LRRs IV–VIII of LGR7 were recently implicated
in binding relaxin (27). Seven of these residues, Trp-
180, Ile-182, Asp-231, Glu-233, Val-253, Glu-277,
and Asp-279, are highly conserved in LGR8 (Fig. 1).
Especially of interest are LGR7 Asp-231, Glu-233,
Glu-277, and Asp-279, which are thought to chelate
the crucial arginine residues in the B-chain of relaxin
(Arg-B13 and Arg-B17). Conservation of their coun-
terparts in LGR8, Asp-227, Glu-229, Glu-273, and
Asp-275, is potentially why relaxin can bind to LGR8
with high affinity. The strict conservation of this
relaxin binding site in LGR8, specifically Trp-177,
Ile-179, Asp-227, Glu-229, Glu-273, and Asp-275
(Fig. 1), implied that these residues may be impor-
tant for INSL3 binding.
To predict how these conserved x residues are
arranged in the LRR structure, the protein sequence
corresponding to the LRR domain of human LGR8
was submitted to the Swiss-Model server. The 1.5a
crystal structure of the NgR [Protein Data Bank
(PDB) accession code 1OZN] was chosen as the
preferred template structure because of its high se-
quence homology to LGR8. The resultant models
exhibited solenoid-like structures, typical of pro-
teins containing multiple LRRs (Fig. 2). The con-
served relaxin binding site identified in LGR8 is la-
beled in Fig. 2. The arrangement of these residues
on the inner face of the best LGR8 model was similar
to the arrangement of the homologous residues in
the LGR7 model reported previously (27). The main
difference was that the radius of curvature in our
LGR8 model was larger than the LGR7 model re-
ported by Bullesbach and Schwabe, which was
based on the ribonuclease inhibitor structure. Like
the template structure used, the NgR, the overall
curvature of the LGR8 models was more banana-like
than the horseshoe-shaped ribonuclease inhibitor
(Fig. 2).
125
I-INSL3 Binding to LGR8 Mutant Receptors
Plasmids were generated encoding mutant LGR8 re-
ceptors in which Trp-177, Ile-179, Asp-227, Glu-229,
Glu-273, and Asp-275 were individually replaced with
alanine. An additional mutant was generated in which
LGR8 Asp-227 was replaced with asparagine. These
1700 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
plasmids were transfected into HEK-293T cells, which
were subsequently used in [
125
I]INSL3 binding assays
and cell surface expression analysis.
All of the above LGR8 mutants exhibited significantly
reduced [
125
I]INSL3 binding compared with LGR8 (Fig.
3A). Presented in order from lowest INSL3 binder to
highest compared with LGR8 are as follows: LGR8
D275A (5.69 ⫾ 0.97%, P ⬍ 0.001), LGR8 I179A (5.81 ⫾
0.4%, P ⬍ 0.001), LGR8 W177A (19.55 ⫾ 0.33%, P ⬍
0.001), LGR8 D227A (50.43 ⫾ 1.81%, P ⬍ 0.001), LGR8
E229A (50.58 ⫾ 1.44%, P ⬍ 0.001), LGR8 D227N
(56.17 ⫾ 1.36%, P ⬎ 0.05), and LGR8 E273A (68.06 ⫾
3.32%, P ⬍ 0.001). Of these mutants, only LGR8 I179A
(54.4 ⫾ 3.5%, P ⬍ 0.001) and LGR8 E229A (54.4 ⫾
3.5%, P ⬍ 0.001) exhibited significantly reduced cell
surface expression compared with LGR8 (Fig. 3B). To
further define the effects of these mutations on INSL3
binding, competition binding assays were performed.
[
125
I]INSL3 Competition Binding to Mutant
LGR8 Receptors
INSL3 binds to LGR8 through the B-chain residues
His-B12, Arg-B16, Val-B19, Arg-B20, and Trp-B27.
Rosengren et al. (28) used mutant INSL3 peptides, in
which each of these residues was replaced with ala-
nine, to demonstrate their contribution to INSL3/LGR8
binding. In competition binding assays on LGR8, the
pIC
50
values of INSL3 Ala-B16 and INSL3 Ala-B19 are
10-fold lower than that of INSL3 (Table 1 and Fig. 4A)
(28). INSL3 Ala
12/16/20
displays a 100-fold decrease in
its pIC
50
compared with INSL3, whereas the pIC
50
of
INSL3 B1–26, which is missing INSL3 Trp-B27, is 500-
fold lower than that of INSL3 (Table 1 and Fig. 4A) (28).
Of note is that peptides missing INSL3 His-B12 or
Arg-B20 exhibit the same pIC
50
values as INSL3 (28).
Their contribution to LGR8 binding is only unmasked
Fig. 1. Multiple Sequence Alignment of the Inner LRR Faces of LGR7 and LGR8
Human (H. sapiens), rhesus monkey (M. mulatta), chimpanzee (P. troglodytes), dog (C. familiaris), mouse (M. musculus), rat (R.
norvegicus), cow (B. taurus), and opossum (M. domestica) LGR7 and LGR8 sequences corresponding to the inner face of their
LRRs are presented, along with the homologous sequence from the human NgR. Residues postulated to be involved in relaxin
binding to LGR7 that are conserved in LGR8 are indicated (*). Residues implicated by our model of INSL3 binding to LGR8 that
were investigated experimentally are labeled (arrow). Black shading indicates complete conservation, whereas gray indicates
there are conservative differences at the particular position.
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1701
when they are removed in combination with other res-
idues, such as in INSL3 Ala
12/16/20
. Importantly, these
mutant INSL3 peptides exhibit no significant change
to their nuclear magnetic resonance (NMR) solution
structures, indicating that mutations are revealing spe-
cific receptor interactions (28).
Here we used these same mutant peptides in com-
petition binding assays with the mutated LGR8 recep-
tors to ascertain the specific INSL3 residues interact-
ing with these mutated sites. We reasoned that, if the
binding site for one of these INSL3 residues was dis-
rupted in a particular LGR8 mutant, then the binding
contribution of that particular INSL3 residue would be
impaired or lost. This loss would result in the INSL3
mutant peptide exhibiting a similar pIC
50
value to
INSL3 on the same receptor. In contrast, a decreased
pIC
50
value would indicate that the mutated INSL3
residue was contributing to INSL3 binding at another
site on the receptor. Hence, these competition binding
studies will discern the specific INSL3 peptide side-
chain interactions with the LRR binding sites.
A set of competition binding experiments were un-
dertaken with all of the mutant LGR8 receptors except
LGR8 I179A and LGR8 D275A, which both bound
[
125
I]INSL3 at a level that was too low to conduct
competition binding. Of the others, LGR8 D227A and
LGR8 D227N produced the most interesting results.
Competition binding curves for INSL3 and the mutant
INSL3 peptides on LGR8 and LGR8 D227A are pre-
sented in Fig. 4. As expected, the INSL3 pIC
50
for
LGR8 D227A was significantly lower than LGR8 (8.72 ⫾
0.08, P ⬍ 0.05), highlighting that this site is involved in
INSL3 binding. Additionally, the rank order of pIC
50
val
-
ues for the INSL3 mutant peptides on LGR8 D227A was
Fig. 2. Homology Model of the LRRs of LGR8
Using Swiss-Model, a molecular model of the LRRs of
human LGR8 was generated using the crystal structure of the
NgR as the major template structure. Swiss-Model was un-
able to model the LRR caps of LGR8; thus models of the LRR
caps of LGR7, which display a typical LRR cap structure,
have been used and are shown in gray. x residues from the
inner face of the LRRs of LGR8 are shaded black. Residues
that have been proposed to be crucial for relaxin binding to
LGR7 and that are highly conserved in LGR8 are labeled.
Fig. 3. [
125
I]INSL3 Binding to, and Cell Surface Expression
of, LGR8 Mutants I
A, [
125
I]INSL3 binding to LGR7, LGR8, LGR8 W177A, LGR8
I179A, LGR8 D227A, LGR8 D227N, LGR8 E229A, LGR8
E273A, and LGR8 D275A. B, Cell surface expression of
LGR7, LGR8, LGR8 W177A, LGR8 I179A, LGR8 D227A,
LGR8 D227N, LGR8 E229A, LGR8 E273A, and LGR8 D275A.
*, P ⬍ 0.001 compared with LGR8. Data are the mean ⫾
SEM
of three to four individual experiments performed in triplicate.
1702 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
different from that for LGR8 (Table 1 and Fig. 4, A and B).
Importantly, no significant difference was observed be-
tween the pIC
50
of INSL3 and that of INSL3 Ala-B16
(8.30 ⫾ 0.15, P ⬎ 0.05) and INSL3 Ala
B12/16/20
(8.34 ⫾
0.07, P ⬎ 0.05) for LGR8 D227A in contrast to their lower
pIC
50
values than INSL3 on the wild-type receptor (Table
1 and Fig. 4B) (28). Furthermore, the pIC
50
values of
INSL3 Ala-B19 (6.66 ⫾ 0.08, P ⬍ 0.001) and INSL3
B1–26 (⬍ 5) were significantly lower than both the pIC
50
of INSL3 for LGR8 D227A and the pIC
50
of these pep
-
tides for the wild-type receptor, indicating that these
residues are the major contributors of INSL3 binding to
LGR8 D227A (Table 1 and Fig. 4B). Therefore, His-12,
Arg-16, and Arg-20 were not contributing to INSL3/
LGR8 D227A binding, clearly indicating that they were
somehow involved in binding interactions with LGR8
Asp-227.
To further prove that His-12, Arg-16, or Arg-20 were
involved with a specific binding interaction with LGR8
Asp-227, an LGR8 D227N mutant was generated.
LGR8 D227N bound to the INSL3 peptides in a similar
way to LGR8 D227A. The pIC
50
values of INSL3 Ala-
B16 (8.35 ⫾ 0.04) and INSL3 Ala
B12/16/20
(8.40 ⫾ 0.03)
were not significantly different from the pIC
50
of INSL3
for this receptor (8.77 ⫾ 0.16, both P ⬎ 0.05), revealing
that INSL3 His-B12, Arg-B16, and Arg-B20 were not
contributing to the INSL3/LGR8 D227N binding. Con-
versely, the pIC
50
values of INSL3 Ala-B19 (6.65 ⫾
0.09, P ⬍ 0.05) and INSL3 B1–26 (⬍5) were signifi-
cantly reduced from that of INSL3 (Table 1), indicating
that INSL3 Val-B19 and Trp-B27 were the major con-
tributors to INSL3/LGR8 D227N binding. Thus, LGR8
Asp-227 was involved in specific binding interactions
with INSL3 His-12, Arg-16, or Arg-20.
Other than LGR8 D227A and LGR8 D227N, LGR8
E229A was the only receptor tested in which INSL3
bound with a pIC
50
unchanged from its pIC
50
on LGR8
(9.42 ⫾ 0.28, P ⬎ 0.05) (Table 1). Thus, the reduced
[
125
I]INSL3 binding LGR8 E229A exhibited (Fig. 3A)
was most likely attributable to the low cell surface
expression of this receptor (Fig. 3B). Interestingly,
LGR8 E229A bound to all of the INSL3 mutant pep-
tides with LGR8-like pIC
50
values except INSL3 Ala-
B19, which bound to LGR8 E229A with significantly
lower pIC
50
compared with LGR8 (7.59 ⫾ 0.15, P ⬍
0.01) (Table 1). The other mutant LGR8 receptors,
LGR8 W177A and LGR8 E273A, exhibited reduced
INSL3 pIC
50
binding values compared with LGR8 (Ta
-
ble 1). Unlike LGR8 D227A and LGR8 D227N, how-
ever, the pIC
50
values of the mutant INSL3 peptides
for LGR8 W177A and LGR8 E273A were all shifted in
a comparative manner to what is seen when they bind
LGR8 (Table 1) (28), which indicated that no specific
INSL3 residue was losing the ability to bind to these
mutant receptors. Thus, characterization of LGR8
D227A and LGR8 D227N yielded the only conclusive
evidence for a specific side-chain interaction between
INSL3 and LGR8, and the homology model of the
LRRs of LGR8 was used to further investigate the
binding involvement of LGR8 Asp-227.
Predicting the INSL3/LGR8 Interaction
Molsoft (La Jolla, CA) ICM BrowserPro version 3.4 was
used to manually dock the NMR solution structure of
INSL3 (PDB accession code 2H8B) to the homology
model of the LRRs of LGR8 (Fig. 5). LGR8 Asp-227
was seen to be located in the center of an acidic
Table 1. pIC
50
Values of INSL3 Peptides Binding to LGR8, LGR8 D227A/N, LGR8 W177A, LGR8 E229A, and LGR8 E273A
INSL3
INSL3 Ala-B-
16
INSL3 Ala-B-
19
INSL3 Ala 12/
16/20
INSL3 B1–26
LGR8 pIC
50
9.34 ⫾ 0.02 8.49 ⫾ 0.09 8.36 ⫾ 0.11 7.33 ⫾ 0.07 6.86 ⫾ 0.06
vs. INSL3 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
LGR8 D227A pIC
50
8.72 ⫾ 0.08 8.30 ⫾ 0.15 6.66 ⫾ 0.08 8.34 ⫾ 0.07 ⬍5
vs. INSL3 ns P ⬍ 0.001 ns
vs. LGR8 P ⬍ 0.05 ns P ⬍ 0.001 P ⬍ 0.05
LGR8 D227N pIC
50
8.77 ⫾ 0.16 8.35 ⫾ 0.04 6.65 ⫾ 0.09 8.40 ⫾ 0.03 ⬍5
vs. INSL3 ns P ⬍ 0.05
vs. LGR8 P ⬍ 0.05 ns P ⬍ 0.001 P ⬍ 0.05
LGR8
W177A
pIC
50
7.38 ⫾ 0.08 6.64 ⫾ 0.19 6.52 ⫾ 0.19 ⬍5 ⬍5
vs. INSL3 ns ns
vs. LGR8 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
LGR8 E229A pIC
50
9.42 ⫾ 0.28 8.64 ⫾ 0.14 7.59 ⫾ 0.15 7.19 ⫾ 0.21 6.43 ⫾ 0.21
vs. INSL3 P ⬍ 0.05 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
vs. LGR8 ns ns P ⬍ 0.01 ns ns
LGR8 E273A pIC
50
8.79 ⫾ 0.06 7.20 ⫾ 0.14 6.76 ⫾ 0.01 6.51 ⫾ 0.17 4.97 ⫾ 0.43
vs. INSL3 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
vs. LGR8 P ⬍ 0.05 P ⬍ 0.05 P ⬍ 0.001 P ⬍ 0.05 P ⬍ 0.05
Statistical analyses were undertaken using one-way ANOVA in conjunction with Newman-Keuls multiple comparison post test.
The vs. INSL3 row displays the statistical comparison of the pIC
50
of a particular INSL3 peptide compared to the pIC
50
of INSL3
on the same receptor. The vs. LGR8 row displays the statistical comparison of the pIC
50
of a particular INSL3 peptide compared
to its pIC
50
on LGR8.
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1703
groove running across the inner face of the LRRs (Fig.
5). With INSL3 Arg-B16 aligned with LGR8 Asp-227,
the other two basic residues from the INSL3 B-chain
␣
-helix, His-B12 and Arg-B20, were aligned along this
groove.
In this crude model, with INSL3 Arg-B16 bound to
LGR8 Asp-227, INSL3 His-B12 was in close proximity
to LGR8 Trp-177, INSL3 Val-B19 was close to LGR8
Ile-179, and INSL3 Arg-20 was in the vicinity of LGR8
Glu-229 as well as Asp-181. Of considerable interest
was the close proximity of INSL3 Trp-B27, a crucial
binding residue, to LGR8 Phe-131 and Gln-133 (Fig.
5). Thus, the importance of LGR8 Asp-181, Phe-131,
and Gln-133 were investigated experimentally.
[
125
I]INSL3 Binding to LGR8 F131A and Q133A
LGR8 F131A and LGR8 Q133A mutant receptors were
produced using mutagenesis. LGR8 F131A bound
[
125
I]INSL3 at a low level compared with LGR8
(21.54 ⫾ 1.73%, P ⬍ 0.001) (Fig. 6A), and the cell
surface expression of this mutant was reduced com-
pared with that of LGR8 (85.2 ⫾ 3.6%, P ⬍ 0.001) (Fig.
6C). Similarly, LGR8 Q133A bound to [
125
I]INSL3 at a
significantly lower level than LGR8 (9.67 ⫾ 0.52%, P ⬍
0.001) (Fig. 6A) and also exhibited significantly re-
duced cell surface expression compared with LGR8
(33.9 ⫾ 4.6%, P ⬍ 0.001) (Fig. 6C).
Because both LGR8 Phe-131 and Gln-133 were
predicted to be potential INSL3 Trp-B27 interactors, a
mutant LGR8 receptor was produced in which both of
these residues were mutated to alanine. LGR8 Ala-
131/133 was also combined with LGR8 D227A to form
the triple mutant LGR8 Ala-131/133/227.
Both LGR8 Ala-131/133 and LGR8 Ala-131/133/
227 displayed [
125
I]INSL3 binding that was not sig
-
nificantly higher than background using both 100 p
M
[
125
I]INSL3 (0.42 ⫾ 0.49%, P ⬎ 0.05 and ⫺3.52 ⫾
1.22%, P ⬎ 0.05, respectively) and 500 p
M
[
125
I]INSL3 (0.7 ⫾ 1.0%, P ⬎ 0.05 and 3.6 ⫾ 0.5%,
P ⬎ 0.05, respectively) (Fig. 6, A and B). The cell
surface expression of LGR8 Ala-131/133 and LGR8
Ala-131/133/227, however, was only moderately re-
duced compared with LGR8 (53.1 ⫾ 9.7%, P ⬍
0.001 and 61.3 ⫾ 2.4%, P ⬍ 0.001, respectively)
(Fig. 6C). Such a loss in INSL3 binding would be
Fig. 4. [
125
I]INSL3 Competition Binding to LGR8 and LGR8 D227A with Mutant INSL3 Peptides
A, Competition of [
125
I]INSL3 binding to LGR8 using INSL3, INSL3 Ala-B16, INSL3 Ala-B19, INSL3 Ala
B12/16/20
, and INSL3
B1–26 (missing Trp-B27). B, Competition of [
125
I]INSL3 binding to LGR8 D227A using INSL3, INSL3 Ala-B16, INSL3 Ala-B19,
INSL3 Ala
B12/16/20
, and INSL3 B1–26 (missing Trp-B27). Data are the mean ⫾ SEM of three to four individual experiments
performed in triplicate.
1704 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
expected if LGR8 lost its ability to bind to INSL3
Trp-B27.
To confirm that INSL3 Trp-B27 is specifically inter-
acting with LGR8 F131 and LGR8 Gln-133, competi-
tion binding assays with mutant INSL3 peptides and
the individual mutant receptors were performed.
Based on the results above, we would expect that
each of these mutant receptors would demonstrate a
partial disruption in INSL3 Trp-B27 binding. As ex-
pected, INSL3 bound to LGR8 F131 with a reduced
pIC
50
compared with LGR8 (8.75 ⫾ 0.20, P ⬍ 0.05)
(Fig. 7A and Table 2). Importantly, INSL3 Ala-B16
(7.00 ⫾ 0.20, P ⬍ 0.001), INSL3 Ala-B19 (6.54 ⫾ 0.17,
P ⬍ 0.001), and INSL3 Ala
B12/16/20
(5.98 ⫾ 0.50, P ⬍
0.01) all displayed reduced pIC
50
values compared
with their binding to LGR8, whereas INSL3 B1–26
(6.68 ⫾ 0.17, P ⬎ 0.05) exhibited a similar pIC
50
value
on both of these receptors (Table 2). This indicated
that INSL3 His-B12, Arg-B16, Val-B19, and Arg-B20
are the major contributors to INSL3 binding to LGR8
F131A and thus that the lower pIC
50
of INSL3 for LGR8
F131A was attributable to a partial disruption in INSL3
Trp-B27 binding.
LGR8 Q133A also bound INSL3 with a significantly
lower pIC
50
than its pIC
50
for LGR8 (8.69 ⫾ 0.18, P ⬍
0.05) (Fig. 7B and Table 2). Importantly, LGR8 Q133A
bound to all of the INSL3 mutant peptides in a similar
way to LGR8 F131A. The pIC
50
values of INSL3 Ala-
B16 (7.22 ⫾ 0.01, P ⬍ 0.001), INSL3 Ala-B19 (7.47 ⫾
0.20, P ⬍ 0.01), and INSL3 Ala
B12/16/20
for LGR8
Q133A were all significantly lower than their pIC
50
values for LGR8 (Fig. 7B and Table 2). However, as
with LGR8 F131A, INSL3 B1–26 exhibited a similar
pIC
50
for LGR8 Q133A as it does for LGR8 (pIC
50
of
6.80 ⫾ 0.13, P ⬎ 0.05) (Fig. 7B and Table 2). The very
similar peptide pIC
50
profiles of LGR8 Q133A and
LGR8 F131A suggested that LGR8 Phe-131 and Gln-
133 are both involved in binding INSL3 Trp-B27. Thus,
the loss of INSL3 binding exhibited by LGR8 Ala-131/
133 was attributable to the complete loss of INSL3
Trp-B27 binding.
[
125
I]INSL3 Binding to LGR8 D181N
INSL3 Arg-B20 was predicted to be in close proximity to
both LGR8 Asp-181 and Glu-229 (Fig. 5). An LGR8
D181N mutant was produced and was also combined
with LGR8 D229A to form the double mutant LGR8
D181N/E229A. LGR8 D181N bound to [
125
I]INSL3 at a
reduced level compared with LGR8 (35.7 ⫾ 0.6%, P ⬍
0.001) (Fig. 6A) and significantly disrupted cell surface
expression (67.8 ⫾ 3.0%, P ⬍ 0.001) (Fig. 6C). Although
LGR8 D181N/E229A was expressed on the cell surface
at a similar level to LGR8 D181N (59.9 ⫾ 1.4%, P ⬍
0.001 compared with LGR8) (Fig. 6C), this double mutant
exhibited very low [
125
I]INSL3 binding (5.21 ⫾ 0.50%,
P ⬍ 0.001) compared with LGR8 (Fig. 6A).
To determine whether the loss of [
125
I]INSL3 binding
observed above was a result of the loss of a specific
INSL3 interaction, competition binding was under-
taken. LGR8 D181N bound INSL3 with a similar pIC
50
to LGR8 (9.45 ⫾ 0.10, P ⬎ 0.05) (Table 2). Additionally,
LGR8 D181N bound most of the mutant INSL3 pep-
tides with similar pIC
50
values to LGR8 (Table 2); how
-
ever, like LGR8 E229A, it exhibited a lower INSL3
Ala-B19 pIC
50
compared with LGR8 (7.54 ⫾ 0.09, P ⬍
0.01) (Table 2). This sensitivity to the loss of the INSL3
Val-B19 implied that this INSL3 residue may be com-
pensating for the loss of a minor interaction in these
two receptor mutants. Importantly, INSL3 Ala-B20
binds to LGR8 with a similar pIC
50
to INSL3, and its
binding contribution is only revealed when coupled to
another INSL3 mutation. Hence, the increased INSL3
Ala-B19 pIC
50
shift seen with LGR8 D181N and LGR8
E229A strongly suggested that a minor binding site
such as that of INSL3 Arg-B20 was disrupted in both
of these mutants.
DISCUSSION
LGR7 and LGR8 ligand-mediated signaling involves at
least three steps. Primary hormone binding is the initial
Fig. 5. Surface-Rendered Homology Model of the LRRs of
LGR8 Colored According to Electrostatic Charge
Red represents negative charge areas, and blue repre-
sents positively charged regions. Of note is the predicted
INSL3 binding site, which is composed of a negatively
charged (red) groove in the concave surface of the LRRs of
LGR8 with Asp-227 (bold) at its center. A putative interaction
conformation is highlighted using the NMR solution structure
of the B-chain of INSL3 (black) with INSL3 Arg-B16 (dark
blue) docking to LGR8 Asp-227. The location of INSL3 Trp-
B27 (green) in the structure inferred that LGR8 Phe-131 and
Gln-133 might be involved in binding to this INSL3 residue.
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1705
driver of receptor activation and occurs between the
B-chain of relaxin or INSL3 and the LRRs of LGR7 or
LGR8, respectively (13, 14, 16). At least two secondary
interactions are then required to stimulate each recep-
tor. A lower-affinity interaction between the hormone
and the transmembrane domains of the receptor is
required, whereas the LDLa module of each receptor
is essential for receptor activation (13–16). Impor-
tantly, experiments have shown that primary hormone
binding is completely independent of the other steps
involved in receptor activation. When the ectodomains
of LGR7 and LGR8 are tethered to a single membrane
spanning CD8 domain, the resultant proteins can bind
their hormone partners with similar affinity to the wild-
type receptors (13, 17). Furthermore, LGR7 and LGR8
receptors missing their LDLa module can also bind
their ligands with unchanged affinity but cannot signal
(16). Thus, we were confident that the results pre-
sented here relate to the primary binding of INSL3 to
the LRRs of LGR8 and that the signaling characteris-
tics of the various mutant receptors were irrelevant in
this context.
A key finding in this study was that seven of the
nine residues identified to be crucial for LGR7 to
bind relaxin (15) were highly conserved in LGR8 (Fig.
1), potentially explaining why relaxin binds LGR8
with high affinity. Recent work, including the solved
solution structure of human INSL3 (28), has estab-
lished that the determinants of INSL3 binding lie in
the B-chain, specifically His-B12, Arg-B16, Val-B19,
Arg-B20, and Trp-B27 (28, 29, 31). Of particular
interest is the “relaxin-like” receptor binding cas-
sette (His-B12, Arg-B16, and Val-B19) on the B-
chain
␣
-helix of INSL3, which offered a possible
explanation for the highly conserved relaxin binding
site in the LRRs of LGR8.
Fig. 6. [
125
I]INSL3 Binding to, and Cell Surface Expression of, LGR8 Mutants II
A, [
125
I]INSL3 (100 pM) binding to LGR7, LGR8, LGR8 F131A, LGR8 Q133A, LGR8 Ala-131/133, LGR8 Ala-131/133/227, LGR8
D181N, and LGR8 D181N/E229. B, [
125
I]INSL3 (500 pM) binding to LGR7, LGR8, LGR8 Ala-131/133, and LGR8 Ala 131/133/227.
C, Cell surface expression of LGR7, LGR8, LGR8 F131A, LGR8 Q133A, LGR8 Ala-131/133, LGR8 Ala-131/133/227, LGR8 D181N,
and LGR8 D181N/E229. *, P ⬍ 0.001 compared with LGR8. Data are the mean ⫾
SEM of three to four individual experiments
performed in triplicate.
1706 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
The NMR solution structure of INSL3 reveals that
the basic residues along the B-chain
␣
-helix, His-B12,
Arg-B16, and Arg-B20, form a positively charged clus-
ter that is most probably the initial driver of INSL3
binding to LGR8 (28). In the solution structure of
INSL3, the C terminus of the B-chain, containing Trp-
B27, loops back to be close to Val-B19. This region
was reported to be flexible, and it was postulated that
binding of the B-chain basic residues to LGR8 would
stabilize Trp-B27, enabling it to interact with its bind-
ing partner(s) (28). The low LGR8 binding affinity of
INSL3 B1–26, which lacks Trp-B27, indicates how im-
portant this interaction is for INSL3 binding to LGR8
(28). Although it was tempting to postulate that relaxin
binding to LGR8 is attributable to basic residues along
its B-chain (Arg-B13 and Arg-B17), the low affinity
exhibited by rhesus monkey relaxin and rat relaxin for
LGR8 (13), which both lack a C-terminal tryptophan in
their B-chains, indicated that the tryptophan interac-
tion is a unique and crucial characteristic of ligand
binding to LGR8.
The conservation of relaxin binding residues in
LGR8, Trp-177, Ile-179, Asp-227, Glu-229, Glu-273,
and Asp-275, suggested that these residues were im-
portant for INSL3 binding to LGR8, and thus they were
individually mutated to alanine. All of the mutant re-
ceptors exhibited disrupted [
125
I]INSL3 binding. How
-
ever, this loss of binding did not give an indication into
the cause of the observed loss. Rosengren et al. (28)
used competition binding assays with INSL3 mutant
peptides to demonstrate which INSL3 residues were
contributing to wild-type LGR8 binding. Here we used
these same mutant peptides in competition binding
assays with the mutated LGR8 receptors to ascertain
the specific INSL3 residues interacting with these mu-
tated sites. We reasoned that, if the binding site for
one of these INSL3 residues was disrupted in a par-
ticular LGR8 mutant, then the binding contribution of
that particular INSL3 residue would be impaired or
lost. This loss would result in the INSL3 mutant pep-
tide exhibiting a similar pIC
50
value to INSL3 on the
same receptor. In contrast, a decreased pIC
50
value
would indicate that the mutated INSL3 residue is con-
tributing to INSL3 binding at another site on the re-
ceptor. Hence, these competition binding studies
Fig. 7. [
125
I]INSL3 Competition Binding to LGR8 F131A and LGR8 Q133A with Mutant INSL3 Peptides
A, Competition of [
125
I]INSL3 binding to LGR8 F131A using INSL3, INSL3 Ala-B16, INSL3 Ala-B19, INSL3 Ala
B12/16/20
, and
INSL3 B1–26 (missing Trp-B27). B, Competition of [
125
I]INSL3 binding to LGR8 Q133A using INSL3, INSL3 Ala-B16, INSL3
Ala-B19, INSL3 Ala
B12/16/20
, and INSL3 B1–26 (missing Trp-B27). Data are the mean ⫾ SEM of three to four individual experiments
performed in triplicate.
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1707
would discern the specific INSL3 peptide side-chain
interactions with the LRR binding sites.
We used this paradigm to determine the interactions
(if any) that were occurring at the conserved relaxin
binding site residues Trp-177 (LGR8 W177A), Asp-227
(LGR8 D227A), Glu-229 (LGR8 E229A), and Glu-273
(LGR8 E273A) of LGR8. These experiments could not
be conducted for LGR8 Ile-179 (LGR8 I179A) and Asp-
275 (LGR8 D275A) because their maximum INSL3
binding ability was too low to accurately conduct com-
petition binding assays.
Of the initial group of receptor mutants, only LGR8
D227A and LGR8 D227N mutations provided evi-
dence of a specific side-chain interaction with INSL3.
The similar pIC
50
values of INSL3, INSL3 Ala-B16, and
INSL3 Ala
B12/16/20
indicated that the binding contribu
-
tions of INSL3 His-B12, Arg-B16, and Arg-B20 were
lost in INSL3 binding to LGR8 D227A and LGR8
D227N because the interaction site for INSL3 His-B12,
Arg-B16, and Arg-B20 was disrupted. Although LGR8
Asp-227 may interact with more than one of these
INSL3 residues, it is likely that only Arg-B16 interacts
with this residue, and the removal of Asp-227 causes
a disruption of charge that perturbs the conformation
of the surrounding residues as well. Such a chain
reaction would explain why the binding of INSL3
His-12 and Arg-20 was also perturbed in LGR8 D227A
and LGR8 D227N. The latter possibility became more
plausible when the homology model of the LRRs of
LGR8 was revisited, especially when this model was
visualized in parallel with the solution structure of
INSL3. LGR8 Asp-227 was located in the middle of a
negatively charged groove (Fig. 5), and, when the
structure of INSL3 was crudely docked to our model of
LGR8, the
␣
-helix of the B-chain of INSL3 was able to
fit along this groove with its basic side chains protrud-
ing into the acidic pockets. We predicted that INSL3
Arg-B16, the largest contributor of these three resi-
dues to INSL3 binding, was the most likely residue that
interacts with LGR8 Asp-227 (Fig. 7). This prediction
was based on the central position of Arg-B16 in the
␣
-helix of the B-chain of INSL3 and the competition
binding results above that highlight the interaction of
INSL3 Ala-B16 with LGR8 Asp-227.
This binding conformation was purely hypothetical
and did not take into account the flexible nature of the
peptide and receptor. However, it allowed us to pre-
dict some more INSL3 binding residues in LGR8. The
location of INSL3 Trp-B27, close to LGR8 Phe-131
and Gln-133 (Fig. 5), was enough evidence to investi-
gate these residues experimentally. LGR8 F131A and
Q133A mutant receptors were produced, and the
same experimental paradigms were applied to them.
As expected, both of these receptors displayed re-
duced [
125
I]INSL3 binding compared with LGR8, and
competition binding assays ascertained that the loss
of binding exhibited by these receptors was attribut-
able a partial, but not complete, disruption in INSL3
Trp-B27 binding. Hence, it was postulated that both of
these residues were contributing to the Trp-B27 inter-
action. This was confirmed by removing both LGR8
Phe-131 and Gln-133, which resulted in complete loss
of [
125
I]INSL3 binding at both 100 and 500 pM ligand
concentration, indicative of the complete loss of INSL3
Trp-B27 binding (Fig. 6, A and B), confirming that
INSL3 Trp-B27 binds LGR8 Phe-131 and Gln-133.
The conclusive interactions, INSL3 Arg-B16 with
LGR8 Asp-227 and INSL3 Trp-B27 with LGR8 Phe-
131 and Gln-133, were used as constraints to trian-
gulate where the other INSL3 B-chain residues may be
binding to LGR8 (Fig. 7). The first of these residues,
INSL3 His-B12, was predicted to interact with LGR8
Trp-177 (Fig. 7). LGR8 W177A exhibited the largest
INSL3 affinity shift of all of the mutant receptors (Table
1). The parallel pIC
50
shifts seen with all of the mutant
INSL3 peptides indicated that the binding contribution
Table 2. pIC
50
Values of INSL3 Peptides Binding to LGR8, LGR8 F131A, LGR8 Q133a, and LGR8 D181N
INSL3
INSL3
Ala-B-16
INSL3
Ala-B-19
INSL3
Ala 12/16/20
INSL3
B1–26
LGR8 pIC
50
9.34 ⫾ 0.02 8.49 ⫾ 0.09 8.36 ⫾ 0.11 7.33 ⫾ 0.07 6.86 ⫾ 0.06
vs. INSL3 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
LGR8 F131A pIC
50
8.75 ⫾ 0.20 7.00 ⫾ 0.20 6.54 ⫾ 0.17 5.98 ⫾ 0.50 6.68 ⫾ 0.17
vs. INSL3 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
vs. LGR8 P ⬍ 0.05 P ⬍ 0.05 P ⬍ 0.001 P ⬍ 0.01 P ⬍ 0.05
LGR8
Q133A
pIC
50
8.87 ⫾ 0.13 7.22 ⫾ 0.01 7.47 ⫾ 0.20 6.40 ⫾ 0.23 6.80 ⫾ 0.13
vs. INSL3 P ⬍ 0.001 P ⬍ 0.01 P ⬍ 0.001 P ⬍ 0.001
vs. LGR8 P ⬍ 0.05 P ⬍ 0.001 P ⬍ 0.01 P ⬍ 0.05 P ⬍ 0.05
LGR8
D181N
pIC
50
9.45 ⫾ 0.10 8.18 ⫾ 0.36 7.54 ⫾ 0.09 7.49 ⫾ 0.04 6.53 ⫾ 0.18
vs. INSL3 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001 P ⬍ 0.001
vs. LGR8 P ⬍ 0.05 P ⬍ 0.05 P ⬍ 0.01 P ⬍ 0.05 P ⬍ 0.05
Statistical analyses were undertaken using one-way ANOVA in conjunction with Newman-Keuls multiple comparison post test.
The vs. INSL3 row displays the statistical comparison of the pIC
50
of a particular INSL3 peptide compared to the pIC
50
of INSL3
on the same receptor. The vs. LGR8 row displays the statistical comparison of the pIC
50
of a particular INSL3 peptide compared
to its pIC
50
on LGR8.
1708 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
of no particular INSL3 residue had been lost and,
coupled with the low INSL3 pIC
50
, was indicative of a
gross structural change in this receptor. Such a
change, in which all of the individual interaction points
were disrupted, would remove the sensitivity required
to elucidate the loss of the minor His-B12 interaction.
A similar story may apply to the adjacent LGR8 resi-
due, Ile-179, which, when replaced with alanine, re-
sulted in a receptor that bound INSL3 at a very low
level. Although this was likely the result of a similar
structural perturbation, LGR8 Ile-179 was predicted to
interact with INSL3 Val-B19, which is a significant
contributor to the INSL3 binding site (Fig. 8). The large
loss of INSL3 binding exhibited by LGR8 I179A (Fig.
3A) meant that we were unable to investigate this
potential interaction experimentally. Importantly, the
location of LGR8 Trp-177 and Ile-179 in our model,
both in the center of the predicted INSL3 binding site
(Fig. 7), may bolster the explanation for their drastically
reduced INSL3 binding.
The final minor contributor to the INSL3 binding site
is INSL3 Arg-B20, which was predicted to be interact-
ing with LGR8 Glu-229 or Asp-181 or both (Fig. 7).
Direct evidence for this interaction could not be ob-
tained experimentally, mainly because this is one of
the minor contributors to INSL3 binding (28). As would
be expected if these residues are involved in minor
binding contributors, INSL3 bound both LGR8 E229A
and LGR8 D181N with pIC
50
values not significantly
different from that on LGR8 (Tables 1 and 2). This was
consistent with the affinity of INSL3 Ala-20 for LGR8,
which was unchanged from that of INSL3 (28). Only
when coupled to the mutation of another INSL3 resi-
due is the importance of INSL3 Arg-B20 revealed (28).
The increased sensitivity of both LGR8 E229A and
LGR8 D181N to the loss of INSL3 Val-B19 (Tables 1
and 2) strongly supported our predicted binding
model. This result suggested that Val-B19 was com-
pensating for the loss of another minor interactor, such
as Arg-B20, when INSL3 bound LGR8 D181N and
LGR8 E229A (Fig. 8). The predicted role of these res-
idues in INSL3 binding was further justified by the
removal of LGR8 Asp-181 and Glu-229 in LGR8
D181N/E229A, which resulted in low [
125
I]INSL3 bind
-
ing (Fig. 6A). Unfortunately, this low level of INSL3
binding meant that competition binding assays could
not be undertaken to further define the reason for this
loss. Overall, the characterization of LGR8 D181N,
LGR8 E229A, and LGR8 D181N/E229A revealed that
LGR8 Asp-181 and Glu-229 were involved in minor
contributions to INSL3 binding. This, together with the
location of these residues in the acidic groove of the
LRRs of LGR8, indicated that the most probable role
these residues are playing was the binding of INSL3
Arg-B20.
Interestingly, five of the seven residues that have
been implicated by our model as being crucial INSL3
interaction points in LGR8 (Gln-133, Trp-177, Ile-179,
Asp-227, and Glu-229) were totally conserved in LGR7
(Fig. 1). The residue corresponding to Phe-131 in
LGR8 was not well conserved in LGR7, being Tyr-134
in human, rhesus monkey, chimpanzee, and opossum
LGR7, but is Cys-134 in dog, mouse, rat, and cow
LGR7 (Fig. 1). Considering that INSL3 Trp-27 poten-
tially interacts at this site in LGR8, Phe-131 may be a
crucial residue in determining the specificity of LGR8
for INSL3. The binding of INSL3 to LGR8 is reliant on
five B-chain residues, whereas relaxin interacts with
LGR7 through only three residues. We postulated that,
although the LRRs of LGR7 contain most of the resi-
dues needed for INSL3 binding, these residues are not
arranged in the specific surface topology needed to
accommodate INSL3. In contrast, relaxin can likely
bind to LGR8 through interaction of Arg-B13 and Arg-
B17 with the acidic grove, which would then allow the
Trp-B27 of relaxin to mimic the role of INSL3 Trp-B27.
In summary, the results presented here have con-
clusively defined the LGR8 LRR residues that are in-
volved in binding to the two most important INSL3
residues, Arg-B16 and Trp-B27. Additionally, they al-
lowed the prediction of the LGR8 residues involved in
the minor binding residues in INSL3, His-B12, Val-
B19, and Arg-B20. Understanding the molecular de-
terminants of primary INSL3 binding to LGR8 is the
Fig. 8. INSL3 Binding to LGR8
Our model interaction of INSL3 with the LRRs of LGR8 is
shown with INSL3 folded back from the predicted binding
site. The confirmed interactions, INSL3 Arg-B16 with LGR8
Asp-227 and INSL3 Trp-B27 with LGR8 Phe-131 and Gln-
133, are indicated with bold arrows. More inferred pairings
are indicated with dashed arrows and include INSL3 His-B12
with LGR8 Trp-177, INSL3 Val-B19 with LGR8 Ile-179, and
INSL3 Arg-B20 with LGR8 Asp-181 and Glu-229.
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1709
first step to thoroughly characterize the mechanism of
LGR8 signaling. A B-chain only mimetic of INSL3 was
recently characterized as an LGR8 antagonist in vitro
that was able to block testicular function in vivo in male
rats (32). It was able to do this by competing with
INSL3 for the primary binding site in LGR8. Knowledge
of the residues involved in INSL3 binding to LGR8 will
allow additional development of such peptides, which
can be made to target the Trp-B27 and Arg-B16 sites
in LGR8. Recently, the N terminus of the A-chain of
INSL3 and the N-terminal LDLa module of LGR8 were
shown to be crucial for LGR8 activation but not pri-
mary binding (15, 16). Combining the results pre-
sented here with those from future efforts to charac-
terize the roles of the A-chain of INSL3 and the
receptor LDLa module will lead to a more complete
understanding of how these unique receptors function
and thus greatly aid in the development of agonists
and antagonists.
MATERIALS AND METHODS
Hormones
H2 relaxin was kindly provided by BAS Medical (San Mateo,
CA). INSL3 was synthesized as described previously (33) and
labeled with
125
I by Dr. Pierre Demeyts (Hagedorn Research
Institute, Gentofte, Denmark) or Steve Sutton (Johnson and
Johnson Pharmaceutical Research and Development, La
Jolla, CA). INSL3 Arg3Ala-B16 (INSL3 Ala-B16), INSL3
Val3Ala-B19 (INSL3 Ala-B19), INSL3 His3Ala-B12 ⫹
Arg3Ala-B16 ⫹ Arg3Ala-B20 (INSL3 Ala
B12/16/20
), and the
truncated INSL3 B1–26 were synthesized as described pre-
viously (28).
Secondary Structure Prediction and
Molecular Modeling
The protein sequences of the LRR subdomains of LGR7 and
LGR8 were submitted into the PROF secondary structure
prediction server (34), and the outputs were used to estimate
the positions of each

-strand. The identity of each

-strand
was further clarified by aligning LGR7 and LGR8 protein
sequences to that of their closest structural homolog, the
NgR, using BlastP and by judging sequence conformity to the
typical LRR

-strand consensus, Lx
1
x
2
Lx
3
Lx
4
x
5
N. Molecular
models of the LRRs of LGR7 and LGR8 were generated by
submitting the relevant protein sequences to Swiss-Model
using the first approach mode (35–37). Swiss-Model outputs
were opened with the program DeepView 3.7 (38) for quality
assessment. LRR domain models were analyzed using UCSF
Chimera (39) and Molsoft BrowserPro version 3.4 molecular
visualization software packages. Structural alignments were
performed with the MatchMaker function in UCSF Chimera
(39). The solution structure of INSL3 (PDB accession no.
2H8B) was manually docked to the best LGR8 LRR model
using Molsoft BrowserPro version 3.4.
Sequence Conservation Analysis
The amino acid sequences of LGR7 and LGR8 from the
human (H. sapiens) (Q9HBX9 and Q8WXD0), mouse (M. mus-
culus) (AAR97515 and Q91ZZ5), and rat (R. norvegicus)
(AAR97516 and AAW84088) were retrieved from the Gen-
Bank database at National Center for Biotechnology Infor-
mation (http://www.ncbi.nlm.nih.gov). Sequence similarity
searches of the available genomes at Ensembl (http://www.
ensembl.org) using TBlastN (40) using a representative full-
length sequence of each receptor were conducted. Predicted
orthologs of LGR7 and LGR8 from the rhesus monkey
(M. mulatta), chimpanzee (P. troglodytes), dog (C. familiaris),
cow (B. taurus), and opossum (M. domestica) genomes were
identified (Wilkinson, T., G. W. Tregear, T. P. Speed, R. A. D.
Bathgate, unpublished data). Sequences were aligned using
ClustalW (41) with default parameters and shaded using Box-
shade. All LGR7 and LGR8 protein sequences are numbered
from the predicted signal peptide cleavage sites.
Site-Directed Mutagenesis
LGR8 mutants were generated using the QuikChange II site-
directed mutagenesis kit (Stratagene, La Jolla, CA) using a
pcDNA3.1/zeo plasmid encoding N-terminal FLAG-tagged
human LGR8 as the template for each reaction. Mutagenic
primer pairs were designed following the protocol described
previously (42). Reactions were undertaken as described by
the manufacturer, incubated with 1
l Dpn1 for 60 min before
1
l was transformed into XL-1 supercompetent cells (Strat-
agene), which were then grown overnight on 100
g/ml am-
picillin containing agar plates. Plasmid DNA was extracted
from selected clones and sequenced.
Ligand Binding Assays
HEK-293T cells were transfected with plasmids encoding the
receptor of interest, and [
125
I]INSL3 binding assays were
conducted as described previously (43). Briefly, 300,000 cells
were seeded into each well of poly-
L-lysine-coated 48-well
plates and left to attach overnight before receptor transfec-
tion the following morning using Lipofectamine 2000 (Invitro-
gen, Carlsbad, CA). Twenty-four hours later, the media were
aspirated from each well, and the cells washed with PBS, and
treated with solutions containing 100 p
M [
125
I]INSL3 together
with a selected concentration of the unlabeled peptide of
interest. Double LGR8 mutants that displayed very low INSL3
binding using 100 p
M [
125
I]INSL3 were assayed again using
500 p
M [
125
I]INSL3. Data are expressed as mean ⫾ SEM of
percentage specific binding of triplicate measurements
pooled from at least three independent experiments. Data
were analyzed using Prism (GraphPad Software, San Diego,
CA), and a nonlinear regression one-site binding model was
used to plot curves and calculate pIC
50
values. Final pooled
pIC
50
and total binding data were analyzed using one-way
ANOVA coupled to Newman-Keuls multiple comparison test
for multiple group comparisons.
Cell Surface Expression Assays
All of the receptors produced in this study contained an
N-terminal FLAG epitope (Sigma, St. Louis, MO), which does
not affect the pharmacology of LGR7 or LGR8 (5, 13). This tag
enabled us to quantitate the cell surface expression of each
receptor. Cell surface expression assays were performed as
described previously (16). Empty vector transfected cells
were used to determine nonspecific background of this cell
surface expression assay. All data points were performed in
triplicate, and the data are expressed as the mean ⫾ SEM from
triplicate measurements pooled from three separate experi-
ments. These values were compared using one-way ANOVA
coupled to Newman-Keuls comparison test for multiple
group comparisons.
Acknowledgments
We thank Prof. Pierre De Meyts and Dr. Steve Sutton for
125
I-labeling of INSL3, Sharon Layfield for technical assis
-
1710 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
tance, and Assoc. Prof. Paul Gooley for assistance with the
manual docking.
Received February 20, 2007. Accepted April 26, 2007.
Address all correspondence and requests for reprints to:
Ross A. D. Bathgate, Howard Florey Institute, The University
of Melbourne, Melbourne, Victoria 3001, Australia. E-mail:
r.bathgate@hfi.unimelb.edu.au.
This work was supported by Australian National Health and
Medical Research Council Project Grants 30012 and 454375
(to R.A.D.B., J.D.W., and G.W.T.) and 350245 (to J.D.W. and
R.A.D.B.). D.J.S. is a recipient of an Australian Postgraduate
Award.
Disclosure Statement: The authors have nothing to
disclose.
REFERENCES
1. Wilkinson T, Scott DJ, Hopkins E, Bathgate RAD 2005
Modern perspectives on the structure, function and evo-
lution of the relaxin-like peptides and their receptors.
Curr Med Chem Immunol Endocr Metabolic Agents
5:369–381
2. Kubota Y, Nef S, Farmer PJ, Temelcos C, Parada LF,
Hutson JM 2001 Leydig insulin-like hormone, guber-
nacular development and testicular descent. J Urol 165:
1673–1675
3. Kubota Y, Temelcos C, Bathgate RA, Smith KJ, Scott D,
Zhao C, Hutson JM2002 The role of insulin 3, testoster-
one, Mullerian inhibiting substance and relaxin in rat gu-
bernacular growth. Mol Hum Reprod 8:900–905
4. Kawamura K, Kumagai J, Sudo S, Chun SY, Pisarska M,
Morita H, Toppari J, Fu P, Wade JD, Bathgate RA, Hsueh
AJ2004 Paracrine regulation of mammalian oocyte mat-
uration and male germ cell survival. Proc Natl Acad Sci
USA 101:7323–7328
5. Hsu SY, Nakabayashi K, Nishi S, Kumagai J, Kudo M,
Sherwood OD, Hsueh AJ2002 Activation of orphan re-
ceptors by the hormone relaxin. Science 295:671–674
6. Bathgate RA, Ivell R, Sanborn BM, Sherwood OD, Sum-
mers RJ 2006 International Union of Pharmacology LVII:
recommendations for the nomenclature of receptors for
relaxin family peptides. Pharmacol Rev 58:7–31
7. Braun T, Schofield PR, Sprengel R 1991 Amino-terminal
leucine-rich repeats in gonadotropin receptors determine
hormone selectivity. EMBO J 10:1885–1890
8. Nagayama Y, Wadsworth HL, Chazenbalk GD, Russo D,
Seto P, Rapoport B 1991 Thyrotropin-luteinizing hor-
mone/chorionic gonadotropin receptor extracellular do-
main chimeras as probes for thyrotropin receptor func-
tion. Proc Natl Acad Sci USA 88:902–905
9. Osuga Y, Kudo M, Kaipia A, Kobilka B, Hsueh AJ 1997
Derivation of functional antagonists using N-terminal ex-
tracellular domain of gonadotropin and thyrotropin re-
ceptors. Mol Endocrinol 11:1659–1668
10. Hong S, Phang T, Ji I, Ji TH 1998 The amino-terminal
region of the luteinizing hormone/choriogonadotropin re-
ceptor contacts both subunits of human choriogonado-
tropin. I. Mutational analysis. J Biol Chem 273:
13835–13840
11. Vischer HF, Granneman JC, Bogerd J 2003 Opposite
contribution of two ligand-selective determinants in the
N-terminal hormone-binding exodomain of human go-
nadotropin receptors. Mol Endocrinol 17:1972–1981
12. Vischer HF, Granneman JC, Noordam MJ, Mosselman S,
Bogerd J 2003 Ligand selectivity of gonadotropin recep-
tors. Role of the

-strands of extracellular leucine-rich
repeats 3 and 6 of the human luteinizing hormone recep-
tor. J Biol Chem 278:15505–15513
13. Halls ML, Bond CP, Sudo S, Kumagai J, Ferraro T, Lay-
field S, Bathgate RA, Summers RJ2005 Multiple binding
sites revealed by interaction of relaxin family peptides
with native and chimeric relaxin family peptide receptors
1 and 2 (LGR7 and LGR8). J Pharmacol Exp Ther 313:
677–687
14. Sudo S, Kumagai J, Nishi S, Layfield S, Ferraro T, Bath-
gate RA, Hsueh AJ 2003 H3 relaxin is a specific ligand for
LGR7 and activates the receptor by interacting with both
the ectodomain and the exoloop 2. J Biol Chem 278:
7855–7862
15. Bullesbach EE, Schwabe C 2005 LGR8 signal activation
by the relaxin-like factor. J Biol Chem 280:14586–14590
16. Scott DJ, Layfield S, Yan Y, Sudo S, Hsueh AJ, Tregear
GW, Bathgate RA 2006 Characterization of novel splice
variants of LGR7 and LGR8 reveals that receptor signal-
ing is mediated by their unique low density lipoprotein
class A modules. J Biol Chem 281:34942–34954
17. Yan Y, Cai J, Fu P, Layfield S, Ferraro T, Kumagai J, Sudo
S, Tang JG, Giannakis E, Tregear GW, Wade JD, Bath-
gate RA 2005 Studies on soluble ectodomain proteins of
relaxin (LGR7) and insulin 3 (LGR8) receptors. Ann NY
Acad Sci 1041:35–39
18. Kobe B, Kajava AV 2001 The leucine-rich repeat as a
protein recognition motif. Curr Opin Struct Biol 11:
725–732
19. Fan QR, Hendrickson WA 2005 Structure of human fol-
licle-stimulating hormone in complex with its receptor.
Nature 433:269–277
20. He XL, Bazan JF, McDermott G, Park JB, Wang K,
Tessier-Lavigne M, He Z, Garcia KC 2003 Structure of
the Nogo receptor ectodomain: a recognition module
implicated in myelin inhibition. Neuron 38:177–185
21. Smits G, Campillo M, Govaerts C, Janssens V, Richter C,
Vassart G, Pardo L, Costagliola S 2003 Glycoprotein
hormone receptors: determinants in leucine-rich repeats
responsible for ligand specificity. EMBO J 22:2692–2703
22. Song YS, Ji I, Beauchamp J, Isaacs NW, Ji TH 2001
Hormone interactions to Leu-rich repeats in the gonad-
otropin receptors. I. Analysis of Leu-rich repeats of hu-
man luteinizing hormone/chorionic gonadotropin recep-
tor and follicle-stimulating hormone receptor. J Biol
Chem 276:3426–3435
23. Song YS, Ji I, Beauchamp J, Isaacs NW, Ji TH 2001
Hormone interactions to Leu-rich repeats in the gonad-
otropin receptors. II. Analysis of Leu-rich repeat 4 of
human luteinizing hormone/chorionic gonadotropin re-
ceptor. J Biol Chem 276:3436–3442
24. Galet C, Ascoli M 2005 The differential binding affinities
of the luteinizing hormone (LH)/choriogonadotropin re-
ceptor for LH and choriogonadotropin are dictated by
different extracellular domain residues. Mol Endocrinol
19:1263–1276
25. Thomas D, Rozell TG, Liu X, Segaloff DL 1996 Mutational
analyses of the extracellular domain of the full-length
lutropin/choriogonadotropin receptor suggest leucine-
rich repeats 1–6 are involved in hormone binding. Mol
Endocrinol 10:760–768
26. Bernard MP, Myers RV, Moyle WR 1998 Lutropins ap-
pear to contact two independent sites in the extracellular
domain of their receptors. Biochem J 335:611–617.
27. Bullesbach EE, Schwabe C 2005 The trap-like relaxin-
binding site of the leucine-rich G-protein-coupled recep-
tor 7. J Biol Chem 280:14051–14056
28. Rosengren KJ, Zhang S, Lin F, Daly NL, Scott DJ,
Hughes RA, Bathgate RA, Craik J, Wade JD 2006 Solu-
tion structure and characterization of the LGR8 receptor
binding surface of insulin-like peptide 3. J Biol Chem
281:28287–28295
29. Bullesbach EE, Schwabe C 1999 Tryptophan B27 in the
relaxin-like factor (RLF) is crucial for RLF receptor-bind-
ing. Biochemistry 38:3073–3078
Scott et al. • LGR8 Residues Involved in Binding INSL3 Mol Endocrinol, July 2007, 21(7):1699–1712 1711
30. Bullesbach EE, Schwabe C 2004 Synthetic cross-links
arrest the C-terminal region of the relaxin-like factor in an
active conformation. Biochemistry 43:8021–8028
31. Bullesbach EE, Schwabe C 2006 The mode of interaction
of the relaxin-like factor (RLF) with the leucine-rich repeat
G protein-activated receptor 8. J Biol Chem 281:
26136–26143
32. Del Borgo MP, Hughes RA, Bathgate RA, Lin F,
Kawamura K, Wade JD 2006 Analogs of insulin-like pep-
tide 3 (INSL3) B-chain are LGR8 antagonists in vitro and
in vivo. J Biol Chem 281:13068–13074
33. Bathgate RA, Lin F, Hanson NF, Otvos Jr L, Guidolin A,
Giannakis C, Bastiras S, Layfield SL, Ferraro T, Ma S,
Zhao C, Gundlach AL, Samuel CS, Tregear GW, Wade
JD2006 Relaxin-3: improved synthesis strategy and
demonstration of its high-affinity interaction with the re-
laxin receptor LGR7 both in vitro and in vivo. Biochem-
istry 45:1043–1053
34. Ouali M, King RD 2000 Cascaded multiple classifiers for
secondary structure prediction. Protein Sci 9:1162–1176
35. Arnold K, Bordoli L, Kopp J, Schwede T 2006 The
SWISS-MODEL workspace: a web-based environment
for protein structure homology modelling. Bioinformatics
22:195–201
36. Schwede T, Kopp J, Guex N, Peitsch MC 2003 SWISS-
MODEL: an automated protein homology-modeling
server. Nucleic Acids Res 31:3381–3385
37. Guex N, Peitsch MC 1997 SWISS-MODEL and the
Swiss-PdbViewer: an environment for comparative pro-
tein modeling. Electrophoresis 18:2714–2723
38. Kaplan W, Littlejohn TG 2001 Swiss-PDB Viewer (Deep
View). Brief Bioinform 2:195–197
39. Pettersen EF, Goddard TD, Huang CC, Couch GS,
Greenblatt DM, Meng EC, Ferrin TE 2004 UCSF Chime-
ra—a visualization system for exploratory research and
analysis. J Comput Chem 25:1605–1612
40. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z,
Miller W, Lipman DJ1997 Gapped BLAST and PSI-
BLAST: a new generation of protein database search
programs. Nucleic Acids Res 25:3389–3402
41. Thompson JD, Higgins DG, Gibson TJ 1994 CLUSTAL
W: improving the sensitivity of progressive multiple se-
quence alignment through sequence weighting, position-
specific gap penalties and weight matrix choice. Nucleic
Acids Res 22:4673–4680
42. Zheng L, Baumann U, Reymond JL 2004 An efficient
one-step site-directed and site-saturation mutagenesis
protocol. Nucleic Acids Res 32:e115
43. Muda M, He C, Martini PG, Ferraro T, Layfield S, Taylor
D, Chevrier C, Schweickhardt R, Kelton C, Ryan PL,
Bathgate RA 2005 Splice variants of the relaxin and
INSL3 receptors reveal unanticipated molecular com-
plexity. Mol Hum Reprod 11:591–600
Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost
professional society serving the endocrine community.
1712 Mol Endocrinol, July 2007, 21(7):1699–1712 Scott et al. • LGR8 Residues Involved in Binding INSL3
