| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Chemistry (K.M.F.) Union College Schenectady,
New York 12308
Division of Molecular Medicine (K.M.F.,
J.A.D., P.V.R.) Wadsworth Center Albany, New York
12201-0509
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
- and
ßsubunits of hFSH have similar folds, consisting of central
cystine-knot motifs from which three ß-hairpins extend. The two
subunits associate very tightly in a head-to-tail arrangement, forming
an elongated, slightly curved structure, similar to that of human
chorionic gonadotropin (hCG). The hFSH heterodimers differ only in the
conformations of the amino and carboxy termini and the second loop of
the ß-subunit (L2ß). Detailed comparison of the structures of hFSH
and hCG reveals several differences in the ß-subunits that may be
important with respect to receptor binding specificity or signal
transduction. These differences include conformational changes and/or
differential distributions of polar or charged residues in loops L3ß
(hFSH residues 62–73), the cystine noose, or determinant loop
(residues 87–94), and the carboxy-terminal loop (residues 94–104). An
additional interesting feature of the hFSH structure is an extensive
hydrophobic patch in the area formed by loops L1, L3, and ßL2.
Glycosylation at Asn52 is well known to be required for full signal
transduction activity and heterodimer stability. The structure reveals
an intersubunit hydrogen bonding interaction between this carbohydrate
and ßTyr58, an indication of a mechanism by which the carbohydrate
may stabilize the heterodimer. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-subunit (92 amino acids) and a unique ß-subunit (111 amino acids
in FSH) (1, 2, 3, 4). FSH acts by binding to G protein-coupled receptors that
signal, in part, through the protein kinase A pathway (5, 6). FSH
enables ovarian folliculogenesis to the antral follicle stage and is
essential for Sertoli cell proliferation and maintenance of sperm
quality in the testis. The amino acids in GPH that have been identified
as critical for receptor binding are striking in similarity, yet they
are not the residues essential for signal transduction (7).
Identification of GPH residues that are key to signal transduction
could provide for the development of molecular mimetics or antagonists
of GPH action, with clinical applications in fertility management (8)
and treatment of thyroid disorders (9).
Glycosylation of the GPH has been shown to be important in circulatory
persistence and clearance, and in bioactivity (10, 11, 12, 13, 14). Each subunit
contains two glycosylation sites: at Asn52 and Asn78 in the -
subunit and at conserved sites in the ß-subunit, Asn7 and Asn24
in human FSH (hFSH). ß-Subunit glycosylation has been reported to
affect disulfide bond formation and rate of secretion, with site 2
having a greater effect than site 1, especially on secretion (15).
Glycosylation at Asn78 appears to be important for thermal stability
(16). Deglycosylation of hFSH and hCG at Asn52 has long been
accepted to impair signal transduction while allowing full binding
activity, suggesting that receptor binding and signal transduction are
two separate functions involving different residues and that the
carbohydrate is key to signal transduction. Recent evidence suggests
that deglycosylation at Asn52 causes hCG to be metastable, and
dissociation of the subunits occurs at 37 C (17). In that study,
disulfide bonds engineered between the subunits could overcome the
effect of deglycosylation on signal transduction. Although such studies
have not been performed for hFSH, they obscure previous results with
deglycosylated GPH and demonstrate that the carbohydrate at Asn52 is
not essential for full signal transduction if the subunit association
is otherwise stabilized. It remains unclear whether the carbohydrate,
or the lack of it, affects the structure of all GPH, or whether the
formation of intersubunit disulfide bonds stabilizes hCG in a
conformation that is signal transduction competent.
Previous structural studies of GPH heteroolimers are limited to two independent reports of the crystal structure of human CG (hCG), partially deglycosylated by hydrogen fluoride treatment (3, 4), and a recent report of a low resolution structure of the ternary complex of fully glycosylated hCG with two Fv fragments (18). With the goal of determining the structure of fully active, glycosylated hFSH, we achieved high-level expression of hFSH in Hi5 insect cells and established a method for purification that produces biologically active hFSH. As part of studies of the glycosylation of hFSH, it was observed that glycosylation at Asn24 of the ß-subunit is detectable in only about half of the molecules (7). To reduce glycoform heterogeneity, in anticipation that this would facilitate crystallization, glycosylation at ßAsn24 was eliminated by site- directed mutagenesis, converting Thr26 to Ala. This isoform of hFSH was fully active and yielded crystals suitable for x-ray diffraction. Here, we report the structure of ßT26A-hFSH and compare it to that of hCG.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-APDVQDCPEC and ß-CELTNITIAI,
indicating that the -subunit signal peptide was cleaved as in
mammalian cells but that the ß-subunit lacked the two amino-terminal
residues (Asn, Ser). Purified hFSH-ßT26A was similar in activity to
hFSH expressed in insect cells, was stable during a 16-h incubation
period at room temperature, and bound receptor, effectively competing
with labeled pituitary hFSH for binding to hFSH receptors expressed in
Chinese hamster ovary (CHO) cells (Fig. 1A
). As expected, the biological activity
of hFSH-ßT26A was indistinguishable from wild-type hFSH, showing
dose-response-related stimulation of progesterone production (Fig. 1B)
and cAMP production (Fig. 1C) in Y1 cells stably transfected with
hFSHR.
|
Structure Determination
The crystals belong to space group
P41212 with cell
parameters, a = 128.3 Å, c = 155.2 Å. Structure
determination revealed the presence of two hFSH molecules in the
asymmetric unit, with a solvent content of about 80%. The structure
was determined by multiple isomorphous replacement with anomalous
scattering, using four heavy atom derivatives, followed by solvent
flattening. Continuous electron density for more than 95% of the main
chain was observed in the initial 3.5Å electron density map (Fig. 2). The model was refined to an R value
of 0.259 and an Rfree of 0.294 (2,584
reflections, 9.7% of the total data set) for all data from 30 to 3.0
Å resolution. The final model contains residues 5 to 90 and ß3
to ß109 in hFSH molecule 1 (hFSH1) and residues 5 to 90 and
ß3 to ß108 in hFSH molecule 2 (hFSH2) in addition to 14 sugar
residues and 2 sulfate ions. No attempt was made to include other
solvent molecules because of the relatively low resolution of the
structure and the uncertainty of the location of the disordered protein
and carbohydrate moieties. Thermal parameter refinement was performed
in a blocked mode, with main chain atoms and side chain atoms of each
residue represented as separate blocks. The thermal parameters are
high, due to a combination of the high solvent content and the high
flexibility of the molecule, with an average for all protein atoms of
58 Å2 and ranging from 28
Å2 for residues in the core of the molecule to
100 Å2 in some loops. Figure 2 shows the initial
experimental electron density maps of a representative loop, residues
ß87–94. The atomic coordinates and structure factors have been
deposited with the Protein Data Bank, RCSB, entry number 1FL7.
|
) belongs
to the family of cystine-knot growth factors (20, 21) and the overall
fold is identical to that of hCG. Both the - and ß- subunits
have similar topologies, in which the cystine knot is the central
motif. In this motif, a disulfide bond between Cys 10 (ß3) and
60 (ß51) passes through a ring defined by disulfide bonds from Cys
28 to 82 and from Cys 32 to 84. Three ß-hairpins extend from the
cystine knot, two of which end in tight ß-turns at one end of the
molecule (loops L1 and L3), while the other one forms a longer, more
open loop (L2) at the opposite end. The L2 loop of the -subunit
(L2) includes the only helical segment in the molecule, a 1.5-turn
-helix that runs nearly perpendicular to the ß-strands. The
ß-hairpins are stabilized by, and associate through, disulfide
bridges. The disulfide bond pairings of hFSH, 7–31, 10–60,
28–82, 32–84, 59–87 and ß3–51, ß17–66, ß20–104,
ß28–82, ß32–84, and ß87–94, are identical to those of hCG and
contradict earlier biochemical assignments (22, 23). The two subunits
are aligned head to tail and are slightly wound around each other so
that ßL2, L1, and L3 form one end of the elongated, curved
heterodimer and L2, ßL1, and ßL3 form the other. The two
subunits are intimately associated via intermolecular contacts that
bury 32% of the total solvent-accessible surface of the monomers. A
loop at the C terminus of the ß-subunit, residues ß84–104, wraps
around ß-strands 2 and 3 of the -subunit, resulting in the
-subunit being surrounded on both sides by the ß-subunit.
Consequently, this loop has been referred to as the "seat belt"
loop in hCG (3).
|
52 and ß7 and for two GlcNAcs
and one Man at site 78. Although some additional density is observed
at all three sites in both hFSH molecules, disorder prevents modeling
of the remaining oligosaccharide.
Crystal Packing
The two molecules in the asymmetric unit sit close together with
their concave surfaces associated in clasped hands-like fashion (Fig. 4). Residues in loops L2, L3,
ßL2, and the carboxy terminus of the - subunit of hFSH2
contact the carboxy-termini of both hFSH1 subunits as well as the hFSH1
seatbelt. In total, the interactions render approximately 10% of the
surface of the heterodimer solvent inaccessible. Intermolecular
contacts with symmetry-related molecules are limited to the loops at
the ends of the molecules. This packing arrangement results in an open
crystal lattice with channels of about 70 Å in diameter. All
carbohydrate chains extend into this open channel, as would be expected
because the crystals are ultimately derived from the initial crystals
produced from the more extensively glycosylated pituitary protein.
The average temperature factor of hFSH2, 49 Å2,
is significantly lower than that of hFSH1, 59Å2,
possibly due to the slightly larger number of crystal contacts in which
hFSH2 is involved.
|
-carbon atoms (Fig. 5A). Although the subunits associate in
identical fashion in both heterodimers, there is a small difference in
the angle between the subunits, which results in an r.m.s.d. of 1.6 Å
between the ß-subunits when the -subunits are superimposed. This
suggests that there is some flexibility in the association of subunits.
As expected, the core region of the protein, the cystine knot motifs,
and the surrounding ß-strands have the same conformation in both
molecules, and all of the major differences occur in the loops and at
the amino and carboxy termini (Fig. 6A).
The C-terminal end of the -helix in loop 2 of the -subunit is
less helical in hFSH2, resulting in a shift in the position of the
-carbon of residue 48 by 2.4 Å. ßL2 differs most, with its
carboxy-terminal end shifted by about 5.5 Å at residue 43. This shift
appears to be correlated with a major difference in the conformation of
the carboxy terminus of the -subunit, which lies in the middle of
loop ßL2 in hFSH2 but is positioned away from the molecule in hFSH1.
The final large difference is in L3, which differs by about 3.7 Å
in the position of residue 72. While these conformational differences
between the two independent observations of the hFSH structure are
indicative of the flexibility of the hFSH heterodimer, they are small
compared with the differences between hFSH and hCG in many of the same
loops.
|
|
atoms of 0.7 Å, with maximum
main chain differences of approximately 1 Å in some of the loops. In
the absence of any systematic differences between the two structures,
the conformational analysis presented here is relevant to both
structures.
hFSH and hCG have identical folds, but significant differences occur in
the amino and carboxy termini and several loops (Figs. 5B
and 6B). As
expected, the -subunits, with r.m.s.d. values of 1.1 Å and 0.9 Å
for the least-squares fit of hCG with hFSH1 and hFSH2, respectively,
are more similar than the ß-subunits, with r.m.s.d. values of 1.5 Å
and 1.6 Å. The largest differences in the ß-subunits occur in the
three loops ßL1, ßL2, and ßL3, the cystine noose, and the
ß-subunit carboxy-terminal loop (Fig. 5B).
Loops ßL1, ßL2, andßL3.
ßL1 and ßL3, together with L2 and the C terminus of the
ß-subunit, form one end of the heterodimer, extending beyond the area
where the two subunits interact (Figs. 3 and 6). Comparison of the
structures reveals large differences in the conformations of hFSH and
hCG in this area, but no differences in the two independent molecules
of hFSH. The amino acid sequences of hFSH and hCG differ greatly in
ßL3, with only 4 identical residues in the 13-residue span between
hFSH residues ß64 and ß76. In addition the loop includes three
prolines in hCG but only one in hFSH. Consequently, the conformation of
ßL3 of hFSH is a more curved ß-hairpin than hCG, resulting in a
shift toward the concave surface of the heterodimer and nearer to the
area of the ß- subunit carboxy terminus (Fig. 7a). The distances between corresponding
residues in the loop are as large as 6.4 Å (6.8 Å) for hFSH1 (hFSH2).
The amino acid sequences in ßL1 for hFSH and hCG are quite similar
(Fig. 5C) and the conformation of the loop itself is also similar.
However, this loop moves by as much as 3.6 Å (3.9 Å) toward the
concave side of the molecule in hFSH in concert with ßL3, because the
two loops are tied together by the disulfide bridge, ßCys17-ßCys66.
These conformational changes are also correlated with conformational
changes in L2 and the C terminus of the ß-subunit, which are
discussed below.
|
), but the two hFSH
molecules differ at this locus as much in conformation from each other
as they differ from hCG (Fig. 6A). While the difference between hFSH
and hCG in part must result from differences in the sequences in this
loop, including a variation in the presence of proline residues (Fig. 5C), overall this result suggests that the loop is highly flexible in
the GPH.
Cystine Noose.
The cystine noose, or determinant loop, is a short loop between two
disulfide-linked cysteines, ß87 and ß94 in hFSH (ß93-ß100 in
hCG) (24) located toward the center of the concave surface of the
molecule, adjacent to L2. Residues in the cystine noose play a role
in determining the specificity of hCG receptor (25) and FSH receptor
(26) binding. The main chain conformation of the loop is very similar
in hFSH and hCG, although the entire loop is shifted by about 3 Å at
its tip (Fig. 7b). This shift is part of the global conformational
change at this end of the molecule, also including ßL1, ßL3, and
L2. More important than the conformational differences between the
cystine nooses are their strikingly different surface charge
characteristics. There are three negatively charged residues, Asp 88,
Asp 90, and Asp 93, and no positively charged residues in hFSH, while
hCG has two positively charged residues (Arg 94 and Arg 95) and one
negatively charged residue (Asp 99). The conserved aspartic acid
residues near the C terminus of the cystine noose, Asp 93 in hFSH and
Asp 99 in hCG, have similar conformations in the two hormones. The
structure is consistent with the biochemical data since this residue
has been shown to be essential for both hFSH activity (7, 27) and
hCG activity. The three aspartic acid residues in hFSH create a
negatively charged patch on one side of the cystine noose. In contrast,
the positively charged arginines in hCG are not arranged to form a
charged patch; rather the side chains are directed to opposite sides of
the loop (Fig. 7b). These observations are consistent with biochemical
data, showing that while neither hFSH Asp 88 nor Asp 90 is
essential for hFSH binding to receptor (27), both play a role in
discriminating between the hFSH and hCG/LH receptors (26). Sequence
comparison (Fig. 5C) shows that the cystine noose and ßL3 are the
only two areas where there is a significant charge differential between
hFSH and hCG. The only other possibly significant area is in ßL2, but
the change in this case is from a generally charged surface in hFSH
to a generally uncharged surface in hCG.
ß-Carboxy-Terminal Loop.
The carboxy terminus of the hFSH ß-subunit, residues 95–108, adopts
a different conformation than in hCG, with maximum distances between
equivalent residues of 7.0 Å (7.0 Å) (Figs. 5B and 6B) at hFSH
residue ß99. Specifically, the loop spanning residues 95–103 in hFSH
makes a tighter turn than the equivalent loop from 101–109 in hCG.
Both hFSH molecules have identical conformations in this area, despite
the fact that this loop in hFSH1 is involved in more intermolecular
contacts than in hFSH2. Therefore, differences between hFSH and hCG are
caused by divergent sequences in this area (Fig. 5C). Biochemical data
indicate that the ß-subunit carboxy terminus plays a role in hFSH
receptor binding, because alanine substitution of the three residues
Arg97-Gly98-Leu99 (27), as well as replacement of the hFSH residues
95–100 with the corresponding residues of human LH (hLH) (26),
diminishes hFSH receptor binding. However, swapping hLH residues for
hFSH residues at this locus does not allow for LH receptor binding by
the hFSH-LH chimera (26). The biochemical data are consistent with the
structure, in that these residues clearly adopt a different
conformation, allowing for differential recognition by the appropriate
receptor. However, the changes in this region alone are not sufficient
to change the receptor specificity of the hormone.
Glycosylation.
Comparison of the structures of hFSH and hCG indicate that
glycosylation has no global effect on the glycohormone conformations.
Although the conformation of the insect cell-expressed, and fully
active, hFSH differs significantly from that of more extensively
deglycosylated, HF-treated hCG in several loops, none of these
differences can be directly correlated with differences in the
glycosylation. The structure of hFSH reveals one likely mechanism by
which the oligosaccharide contributes to heterodimer stability (17).
The nitrogen atom of the acetamido group of the Asn-proximal GlcNAc at
Asn52 forms a hydrogen bond with the hydroxyl group of ßTyr58,
thereby adding an additional intersubunit contact. In hCG, the
equivalent residue is Phe64, incapable of forming a similar hydrogen
bond, but able to make a hydrophobic interaction with the hydrophobic
side of the sugar ring. Indeed, Lapthorn et al. (3) report
contacts between the Asn52 carbohydrate and residues ßTyr59,
ßVal62, ßPhe64, ßAla83, and ßThr97. While the single additional
hydrogen bond observed in the hFSH structure may not be adequate to
explain the higher stability of the glycosylated heterodimer, it is
quite possible that sugar groups not fully defined in the electron
density map make additional contacts.
Hydrophobic Patch.
A propeller-shaped triad of aromatic residues,
Phe17, Phe74, and ßTyr39, together with Pro16, Phe18, and
Met71, forms a solvent-exposed hydrophobic patch at the end of the
hFSH heterodimer composed of loops L1, L3, and ßL2 (Fig. 7c).
This region includes only three charged residues (Lys75, ßLys40,
and ßAsp41), all of which are directed away from the hydrophobic
patch. In hCG, ßLeu45 replaces hFSH ßTyr 39 and the patch is
smaller because of a difference in the conformation of the ßL2 loop.
This area may play a major role in receptor binding. Although the
relevance of the L1 loop in hFSH (residues 14–27) has not been
evaluated by mutagenesis studies, epitope mapping with monoclonal
antibodies revealed a discontinuous immunoneutralizing epitope
comprised of residues 11–27 and 61–92 (29, 30). Mutational
analysis of Phe74 revealed only a modest decrease in hFSH binding to
receptor. In contrast to the importance of L1 and L3, alanine
scanning mutagenesis has shown that ßL2 (residues 33–53) is not
essential for hFSH binding to receptor (31). Rather, ßL2 appears to
be important in stabilizing the heterodimer association (32).
-Subunit Carboxy Terminus.
The -subunit carboxy terminus has been implicated in receptor
binding in hCG (33, 34, 35), hFSH (36), and TSH (37). Clearly, this region
is very flexible in hFSH because it adopts different conformations in
hFSH1 and hFSH2 and electron density is lacking for the last two
residues, 91–92. This makes it difficult to draw conclusions about
how this region is involved in receptor binding. The only clear
difference between hFSH and hCG is the formation of a hydrogen bond
between the side chain NH of Arg95 from the cystine noose of hCG
with the carbonyl oxygen of 89, constraining the conformational
flexibility of hCG. This bond is not possible in the hFSH structure,
because the hFSH -carboxy terminus is not as close to the cystine
noose. This difference in conformation is interesting in the context of
receptor interaction, as the residues in the carboxy terminus of the
-subunit are essential for hFSH and hCG binding (34, 35, 36).
Conclusion
The overall structures of hFSH and hCG are similar, but several
intriguing differences are observed in specific loops, especially in
the ß-subunit. The largest difference is at one end of the molecule
where ßL1 and ßL3 move together toward the concave side of the
molecule in hFSH compared with hCG. In addition, the ßL3 loop has a
very different conformation as a result of major differences in the
amino acid sequence. Currently, no biological data are available
regarding the importance of these loops in receptor binding or signal
transduction. The three other areas with different conformations, the
ß carboxy-terminal loop, the cystine noose, and the hydrophobic patch
area between loops L1, L3, and ßL2 also have different surface
characteristics. In addition, several residues in these areas have been
shown, by scanning alanine mutagenesis or epitope mapping, to play a
role in receptor binding (Fig. 8).
Interestingly, the ß-carboxy-terminal loop, the cystine noose, and
the ßL1 and ßL3 loops are located on the concave side of the
heterodimer (Fig. 8), resulting in a face that is very different in the
two gonadotropins. These differences are therefore very likely to be
important for discrimination between hFSH and hCG by their respective
receptors.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Production of Recombinant hFSH
Production of recombinant hFSH in insect cells was carried out
using procedures as previously described (26, 31). Each production run
consisted of about 30 roller bottles, seeded with 1 x
108 Hi5 cells and kept at 27 C for 48 h. At
that time, virus was added directly to each flask [1.0 MOI
(multiplicity of infection) of hFSHßT26A virus and 3.5 MOI of hFSH
virus] and flasks were rotated for an additional 4–5 days. Next,
media were collected by centrifugation in 1-liter bottles (2,500
x g, 10 min). The clarified media were pooled and made 1.0
mM with phenylmethylsulfonyl fluoride, and 0.1%
with sodium azide. Typically, recombinant hFSH-ßT26A was expressed at
levels of 2.7 mg/liter (n = 9) as determined by enzyme-linked
immunosorbent assay (ELISA) (32). Media (typically 4 liters) were then
concentrated at 4 C utilizing an Amicon (Waltham, MA) radial flow
cartridge (10,000 mwc), to a volume of about 200 ml, and then frozen
until processed. The concentrate was thawed, clarified by
centrifugation (16,000 x g, 30 min) and then applied
to the affinity column without further processing.
Preparation of the Affinity Support
An affinity column was prepared using monoclonal antibody (mAb)
46.3H6.B7. This antibody binds to both the
monomeric ß-subunit of native hFSH, and the heterodimeric hormone,
but has no measurable cross-reactivity with the -subunit or with hLH
(19, 38).
The cell line producing mAb 46.3H6.B7 was expanded as ascites tumors in mice (Animal Welfare Committee Approval was obtained for these studies). Approximately 38 ml of 46.3H6.B7 ascitic fluid were diluted 1:3 with PBS and subjected to ammonium sulfate precipitation as follows. Saturated ammonium sulfate (pH 7.2) was added to the diluted ascitic fluid in a ratio of 4.5 volumes of ammonium sulfate solution to 5.5 volumes of antibody. The mixture was stirred (4 C, 30 min), and the antibody precipitate was collected by centrifugation (5, 858 x g, 15 min). Each precipitate was dissolved in PBS and reprecipitated twice as before. Wet pellets were dissolved and dialyzed against 0.01 M potassium phosphate, pH 7.2. Dialyzed samples were clarified by centrifugation, the pH and conductivity were adjusted to the buffer values, and the sample was applied to a 1.9 x 18 cm diethylaminoethyl (DEAE) Sephacel column. Antibody was eluted with a gradient of 0.0–0.2 M NaCl. The procedure used for coupling of the DEAE Sephacel-purified antibody (100 mg) to CNBr-activated Sepharose was exactly as described by the manufacturer (Pharmacia Biotech, Piscataway, NJ).
Purification of hFSH ßT26A
Concentrates of conditioned media, collected from cells
producing hFSH ßT26A, were applied directly to the affinity support.
Typically a flow rate of 0.6 ml/min was used. The mAb 46.3H6.B7 column
dimensions were 0.9 x 10 cm. The sample buffer and column buffer
were 0.1 M potassium phosphate, made 0.3 M with
NaCl, pH 7.0. The absorbance (275 nm) of the fractions was determined
during the procedure. After a decrease of absorbance to baseline,
elution buffer (0.1 M sodium acetate, pH 2.0, 0.5
M NaCl) was pumped through the column. Fractions were
collected into tubes containing 2.0 M Tris base and mixed
to neutralize each fraction.
Biological Characterization of hFSH-ßT26A Heterodimer
Recombinant hFSH-ßT26A was compared with wild-type hFSH in a
RRA using CHO cells, stably expressing hFSH receptors (36, 39). Signal
transduction induced by hFSH-ßT26A was determined by measuring
progesterone (36) or cAMP (38) production by Y1 cells that stably
express hFSH receptors (39).
Protein Crystallization
Crystals of ßT26A hFSH were grown by macroseeding into drops
containing 3 µl protein (9 mg/ml in 10 mM Tris, pH 7) and
1 µl reservoir solution. The reservoir solution was 100
mM glycine, pH 9.0, and 0.9–1.2 M ammonium
sulfate.
Data Collection and Structure Determination
The crystals were transferred to a solution consisting of the
crystallization buffer enriched with 25% wt/vol sucrose before flash
cooling in liquid nitrogen. The crystals were highly sensitive to the
addition of heavy atom compounds. Crystals tolerated soaking in heavy
atom solutions for no more than about 10 h, with some heavy atom
compounds, especially platinum and mercury compounds resulting in
significant degradation of the diffraction quality after as little as
2 h. Data were collected at Stanford Synchrotron Radiation
Laboratory (SSRL) Beamline B9.1, processed with MOSFLM (40), and then
scaled and merged with SCALA (40) (Table 1). Initial MIRAS phases were calculated
using SOLVE (41) (Table 1). The correct space group and
enantiomer were determining by examining the figure-of-merit
(FOM) for both space groups with each hand. This clearly indicated that
P41212 was the correct
space group, FOM 0.57 vs. 0.43, and that the positive
hand was correct. All four derivatives had relatively weak phasing
power due to low occupancy, and the sites of the platinum and osmium
derivatives were similar. However, solvent flattening [CCP4 program DM
(40)] greatly improved the quality of the phases because of the very
high solvent content of the crystals (
80%). MIRAS phasing in
program SOLVE resulted in an overall FOM of 0.57. After solvent
flattening, the electron density clearly showed two molecules in the
asymmetric unit.
|
Structure Comparisons/Surface Area Calculations
Protein structures were compared using LSQMAN (45), and values
quoted in the text are the results of comparing regions of the
structure after the entire structure was superimposed by least-squares
fitting of the -carbon atoms only. Solvent-accessible surface area
was calculated using CNS with a 1.4-Å probe (42).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
This research was supported by NIH Grants HD-18407 (J.A.D.) and GM-50431 (P.V.R.) and a grant from ARES Advanced Technologies/Serono. The x-ray crystallography facilities of SSRL are funded by the Department of Energy and the National Institutes of Health.
Received for publication June 30, 2000. Revision received August 29, 2000. Accepted for publication September 19, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
subunit of human chorionic
gonadotropin. Biochem Biophys Res Commun 232:117–120[CrossRef][Medline]
and
ß-subunits. J Mol Biol 289:1375–1385[CrossRef][Medline]
using monoclonal antibody 3A
identifies a potential receptor binding sequence. Endocrinology 128:1485–1495-subunit of human follicle stimulating hormone. Mol Cell
Endocrinol 85:41–52[CrossRef][Medline]
-subunit: contributions to receptor binding
and signaling in human chorionic gonadotropin. Mol Endocrinol 6:914–919 subunit play common and different roles in human
choriogonadotropin and follitropin. J Biol Chem 268:13034–13042-subunit
are crucial for receptor binding and hormone action of
follicle-stimulating hormone (FSH) and play hormone-specific roles in
FSH and human chorionic gonadotropin. Endocrinology 136:2948–2953[Abstract]
- subunit C terminus
collaborates with a ß-subunit cystine noose and an -subunit loop
to assemble a receptor-binding domain competent for signal
transduction. Biochemistry 37:1762–1768[CrossRef][Medline]
-subunit in the expression and bioactivity of human
thyroid-stimulating hormone. Mol Endocrinol 9:948–958This article has been cited by other articles:
 |
|
 |
|
  R Nunez Miguel, J Sanders, D Y Chirgadze, J Furmaniak, and B Rees Smith Thyroid stimulating autoantibody M22 mimics TSH binding to the TSH receptor leucine rich domain: a comparative structural study of protein-protein interactions J. Mol. Endocrinol., May 1, 2009; 42(5): 381 - 395. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R Nunez Miguel, J Sanders, D Y Chirgadze, T L Blundell, J Furmaniak, and B Rees Smith FSH and TSH binding to their respective receptors: similarities, differences and implication for glycoprotein hormone specificity J. Mol. Endocrinol., September 1, 2008; 41(3): 145 - 164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ghadami, S.A. Salama, N. Khatoon, R. Chilvers, M. Nagamani, P.J. Chedrese, and A. Al-Hendy Toward gene therapy of primary ovarian failure: adenovirus expressing human FSH receptor corrects the Finnish C566T mutation Mol. Hum. Reprod., January 1, 2008; 14(1): 9 - 15. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jablonka-Shariff, T. R. Kumar, J. Eklund, A. Comstock, and I. Boime Single-Chain, Triple-Domain Gonadotropin Analogs with Disulfide Bond Mutations in the {alpha}-Subunit Elicit Dual Follitropin and Lutropin Activities in Vivo Mol. Endocrinol., June 1, 2006; 20(6): 1437 - 1446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Okada, J. L. Ellsworth, D. M. Durnam, H. S. Haugen, J. L. Holloway, M. L. Kelley, K. E. Lewis, H. Ren, P. O. Sheppard, H. M. Storey, et al. A Glycoprotein Hormone Expressed in Corticotrophs Exhibits Unique Binding Properties on Thyroid-Stimulating Hormone Receptor Mol. Endocrinol., February 1, 2006; 20(2): 414 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P N Themmen An update of the pathophysiology of human gonadotrophin subunit and receptor gene mutations and polymorphisms Reproduction, September 1, 2005; 130(3): 263 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.C. Low, S.L. Nunes, A.J. Bitonti, and J.A. Dumont Oral and pulmonary delivery of FSH-Fc fusion proteins via neonatal Fc receptor-mediated transcytosis Hum. Reprod., July 1, 2005; 20(7): 1805 - 1813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Legardinier, M. Duonor-Cerutti, G. Devauchelle, Y. Combarnous, and C. Cahoreau Biological activities of recombinant equine luteinizing hormone/chorionic gonadotropin (eLH/CG) expressed in Sf9 and Mimic insect cell lines J. Mol. Endocrinol., February 1, 2005; 34(1): 47 - 60. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Kleinau, H. Jaschke, S. Neumann, J. Lattig, R. Paschke, and G. Krause Identification of a Novel Epitope in the Thyroid-stimulating Hormone Receptor Ectodomain Acting as Intramolecular Signaling Interface J. Biol. Chem., December 3, 2004; 279(49): 51590 - 51600. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xing, W. Lin, M. Jiang, D. Cao, R. V. Myers, M. P. Bernard, and W. R. Moyle Use of Protein Knobs to Characterize the Position of Conserved {alpha}-Subunit Regions in Lutropin Receptor Complexes J. Biol. Chem., October 22, 2004; 279(43): 44427 - 44437. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Bernard, W. Lin, D. Cao, R. V. Myers, Y. Xing, and W. R. Moyle Only a Portion of the Small Seatbelt Loop in Human Choriogonadotropin Appears Capable of Contacting the Lutropin Receptor J. Biol. Chem., October 22, 2004; 279(43): 44438 - 44441. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xing, R. V. Myers, D. Cao, W. Lin, M. Jiang, M. P. Bernard, and W. R. Moyle Glycoprotein Hormone Assembly in the Endoplasmic Reticulum: I. THE GLYCOSYLATED END OF HUMAN {alpha}-SUBUNIT LOOP 2 IS THREADED THROUGH A {beta}-SUBUNIT HOLE J. Biol. Chem., August 20, 2004; 279(34): 35426 - 35436. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xing, R. V. Myers, D. Cao, W. Lin, M. Jiang, M. P. Bernard, and W. R. Moyle Glycoprotein Hormone Assembly in the Endoplasmic Reticulum: II. MULTIPLE ROLES OF A REDOX SENSITIVE {beta}-SUBUNIT DISULFIDE SWITCH J. Biol. Chem., August 20, 2004; 279(34): 35437 - 35448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xing, R. V. Myers, D. Cao, W. Lin, M. Jiang, M. P. Bernard, and W. R. Moyle Glycoprotein Hormone Assembly in the Endoplasmic Reticulum: III. THE SEATBELT AND ITS LATCH SITE DETERMINE THE ASSEMBLY PATHWAY J. Biol. Chem., August 20, 2004; 279(34): 35449 - 35457. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xing, R. V. Myers, D. Cao, W. Lin, M. Jiang, M. P. Bernard, and W. R. Moyle Glycoprotein Hormone Assembly in the Endoplasmic Reticulum: IV. PROBABLE MECHANISM OF SUBUNIT DOCKING AND COMPLETION OF ASSEMBLY J. Biol. Chem., August 20, 2004; 279(34): 35458 - 35468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jablonka-Shariff and I. Boime Luteinizing Hormone and Follicle-Stimulating Hormone Exhibit Different Secretion Patterns from Cultured Madin-Darby Canine Kidney Cells Biol Reprod, March 1, 2004; 70(3): 649 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Sohn, H. Youn, M. Jeoung, Y. Koo, C. Yi, I. Ji, and T. H. Ji Orientation of Follicle-stimulating Hormone (FSH) Subunits Complexed with the FSH Receptor: {beta} SUBUNIT TOWARD THE N TERMINUS OF EXODOMAIN AND {alpha} SUBUNIT TO EXOLOOP 3 J. Biol. Chem., November 28, 2003; 278(48): 47868 - 47876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. D. Cohen, J. T. Bariteau, L. M. Magenis, and J. A. Dias Regulation of Follitropin Receptor Cell Surface Residency by the Ubiquitin-Proteasome Pathway Endocrinology, October 1, 2003; 144(10): 4393 - 4402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ulloa-Aguirre, C. Timossi, J. Barrios-de-Tomasi, A. Maldonado, and P. Nayudu Impact of Carbohydrate Heterogeneity in Function of Follicle-Stimulating Hormone: Studies Derived from in Vitro and in Vivo Models Biol Reprod, August 1, 2003; 69(2): 379 - 389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Perlman, B. van den Hazel, J. Christiansen, S. Gram-Nielsen, C. B. Jeppesen, K. V. Andersen, T. Halkier, S. Okkels, and H. T. Schambye Glycosylation of an N-Terminal Extension Prolongs the Half-Life and Increases the in Vivo Activity of Follicle Stimulating Hormone J. Clin. Endocrinol. Metab., July 1, 2003; 88(7): 3227 - 3235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. B. Fralish, B. Dattilo, and D. Puett Structural Analysis of Yoked Chorionic Gonadotropin-Luteinizing Hormone Receptor Ectodomain Complexes by Circular Dichroic Spectroscopy Mol. Endocrinol., July 1, 2003; 17(7): 1192 - 1202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. B. Fralish, P. Narayan, and D. Puett Consequences of Single-Chain Translation on the Structures of Two Chorionic Gonadotropin Yoked Analogs in {alpha}-{beta} and {beta}-{alpha} Configurations Mol. Endocrinol., April 1, 2003; 17(4): 757 - 767. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Schubert, P. Narayan, and D. Puett Specificity of Cognate Ligand-Receptor Interactions: Fusion Proteins of Human Chorionic Gonadotropin and the Heptahelical Receptors for Human Luteinizing Hormone, Thyroid-Stimulating Hormone, and Follicle-Stimulating Hormone Endocrinology, January 1, 2003; 144(1): 129 - 137. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Garcia-Campayo, T. R. Kumar, and I. Boime Thyrotropin, Follitropin, and Chorionic Gonadotropin Expressed as a Single Multifunctional Unit Reveal Remarkable Permissiveness in Receptor-Ligand Interactions Endocrinology, October 1, 2002; 143(10): 3773 - 3778. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. C. Layman, A. L. A. Porto, J. Xie, L. A. C. R. da Motta, L. D. C. da Motta, W. Weiser, and P. M. Sluss FSH{beta} Gene Mutations in a Female with Partial Breast Development and a Male Sibling with Normal Puberty and Azoospermia J. Clin. Endocrinol. Metab., August 1, 2002; 87(8): 3702 - 3707. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Smits, C. Govaerts, I. Nubourgh, L. Pardo, G. Vassart, and S. Costagliola Lysine 183 and Glutamic Acid 157 of the TSH Receptor: Two Interacting Residues with a Key Role in Determining Specificity toward TSH and Human CG Mol. Endocrinol., April 1, 2002; 16(4): 722 - 735. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Vlaeminck-Guillem, S.-C. Ho, P. Rodien, G. Vassart, and S. Costagliola Activation of the cAMP Pathway by the TSH Receptor Involves Switching of the Ectodomain from a Tethered Inverse Agonist to an Agonist Mol. Endocrinol., April 1, 2002; 16(4): 736 - 746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Walton, V. T. Nguyen, V. Y. Butnev, V. Singh, W. T. Moore, and G. R. Bousfield Characterization of Human FSH Isoforms Reveals a Nonglycosylated {beta}-Subunit in Addition to the Conventional Glycosylated {beta}-Subunit J. Clin. Endocrinol. Metab., August 1, 2001; 86(8): 3675 - 3685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Schmidt, R. MacColl, B. Lindau-Shepard, D. R. Buckler, and J. A. Dias Hormone-induced Conformational Change of the Purified Soluble Hormone Binding Domain of Follitropin Receptor Complexed with Single Chain Follitropin J. Biol. Chem., June 22, 2001; 276(26): 23373 - 23381. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
-level. The refined average
temperature factors for the residues included in the figure were 70Å
for the main chain atoms and 85Å for the side chain atoms. Figures 2
