| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
From the Department of Medicine, Division of Endocrinology, University of Maryland Medical School and the Institute of Human Virology, Medical Biotechnology Center, Baltimore, Maryland 21201; and the Molecular and Cellular Endocrinology Branch, National Institute of Diabetes and Digestive and Kidney Diseases (R.W., N.G.T., J.E.T., J.E.-P.), NIH, Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: M. W. Szkudlinski, M.D., Ph.D., Laboratory of Molecular Endocrinology, Institute of Human Virology, Medical Biotechnology Center, 725 West Lombard Street, N457, Baltimore, Maryland 21201. E-mail: szkudlin{at}umbi.umd.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit and a hTSH
ß-minigene, respectively, both under the control of the polyhedrin
promoter. Expression in insect cells was 800-1000 ng/ml, 30-fold higher
than in our optimized mammalian transient transfection system using
Chinese hamster ovary (CHO) cells (20–50 ng/ml). The in
vitro activity of insect-cell expressed hTSH (IC-hTSH) was
increased 5-fold compared with CHO-hTSH, judged by the ability to
induce cAMP production in CHO cells stably transfected with the hTSH
receptor (JP09) and the rat thyroid cell line FRTL-5, as well as growth
promotion in FRTL-5 cells. Lectin binding and enzymatic desialylation
studies suggested that in contrast to CHO-hTSH, IC-hTSH lacked
complex-type oligosaccharides terminating with sialic acid but
contained predominantly high mannose-type oligosaccharides. The
in vitro activity of CHO-hTSH also increased 5- to
6-fold upon treatment of the hTSH-producing cells with the
oligosaccharide processing inhibitors swainsonine and castanospermine,
which inhibit formation of complex, terminally sialylated
oligosaccharides, and upon enzymatic desialylation. In contrast, insect
cell-expression or treatment with processing inhibitors did not affect
TSH receptor binding. Despite the higher in vitro
activity, IC-hTSH had a much lower in vivo activity than
CHO-hTSH, due to rapid clearance from the circulation. In summary, this
study shows for the first time that relatively high levels of
recombinant hTSH with high in vitro bioactivity can be
produced in a baculovirus system. Cell-dependent glycosylation is a
major factor that determines the final in vivo
biopotency of recombinant glycoproteins, a finding that should be of
general relevance for all insect cell-produced glycosylated proteins.
Although not suitable for clinical use, highly bioactive recombinant
hTSH derived from high expression in insect cells should be useful in
defining structure-function relations of hormone analogs. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit and a hormone specific ß-subunit
(1, 2, 3). They control thyroid and gonadal functions upon interaction
with specific G protein-coupled receptors that possess large
extracellular domains including leucine-rich repeats (4, 5, 6).
Elucidation of the crystal structure of human CG (7, 8) has revealed
that the glycoprotein hormones belong to the superfamily of the
cysteine-knot growth factors, which may be related to their recently
proposed nonendocrine actions (9, 10). The carbohydrate portion of
these hormones, which constitutes 20–35% of their weight, is
necessary for the in vitro biological activity. Moreover,
carbohydrates also influence the metabolic clearance rate of these
hormones and thus may cause differential effects on the in
vivo bioactivity (10, 11, 12, 13, 14). Such effects are potentially important
for therapeutic applications of glycoprotein hormones, where modulation
of in vivo hormone activity by specific carbohydrate
structures has been recently elucidated (10, 11, 12, 13, 14). In addition,
oligosaccharides are also essential for posttranslational subunit
folding and assembly, protection from intracellular degradation, and
secretion of the heterodimer (10, 11, 12, 13, 14). Therefore, production of
recombinant glycoprotein hormones requires the use of eukaryotic cells,
and attempts to achieve high level expression of glycoprotein hormones
using prokaryotic systems have not been successful (14). Recently, it
was shown that, using baculovirus systems, it was possible to express
human CG (hCG) and human FSH (hFSH) in insect cells (15, 16, 17, 18). Although
both hormones were shown to possess biological activity in
vitro, the effect of insect cell expression on their final
in vivo potency was not assessed. In contrast to the
gonadotropins, production of hTSH has thus far been limited to
mammalian cell lines, including Chinese hamster ovary (CHO)-K1 cells,
CHO glycosylation mutant cell lines, 293 human embryonic kidney cells
and COS-7 green monkey kidney cells (19, 20, 21, 22). Transient transfections
of such cells, however, generally do not yield satisfactory amounts of
recombinant protein, and the development of stable cell lines is
cumbersome and not suitable for mass screening of recombinant
analogs. In the present study, we have tested the feasibility of using a baculovirus system to achieve high level expression of recombinant hTSH suitable for biological and structural studies. Further, we used this approach to study whether the cell type-dependent oligosaccharide processing would affect hTSH activity in vitro and in vivo because insect cells process oligosaccharides differently from mammalian cells (23, 24, 25). Insect cells lack the capacity to process carbohydrate moieties to complex-type, terminally sialylated oligosaccharides, especially when the very late polyhedrin promoter was used to express the gene(s) of interest. Consequently, glycoproteins produced under such conditions usually contained high mannose-type precursor oligosaccharides (15, 16, 23, 24, 25). In contrast, hTSH expressed in CHO cells (CHO-hTSH, Genzyme Corp., Cambridge, MA), which is presently used in clinical trials in patients with differentiated thyroid carcinoma (26), terminates with complex, sialylated oligosaccharides (27, 28). The present study describes the successful production of bioactive hTSH in insect cells and shows that IC-hTSH had a higher in vitro activity than CHO-hTSH. However, IC-hTSH had low in vivo activity because it was cleared rapidly from the circulation. Additional studies using lectin binding, the oligosaccharide processing inhibitors castanospermine and swainsonine, as well as enzymatic desialylation suggested that these different properties were related to differences in carbohydrate pattern, especially in the terminal sialylation of IC-hTSH and CHO-hTSH.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit and TSH ß-minigene transfer
vectors. A BamHI/XhoI fragment containing
the full length 621-bp h-subunit complementary DNA (cDNA) was
directionally subcloned from the eukaryotic expression vector
pcDNA/NeoI (Invitrogen, San Diego, CA) into the baculovirus
transfer vector pBacPAK8 (Clonetech, Palo Alto, CA), and a
XbaI/BamHI fragment containing the 915 bp hTSH
ß-minigene, previously constructed in our laboratory (21) was
subcloned into the transfer vector pVL1392 (Invitrogen). Expression of
both genes was under the control of the very late AcMNPV polyhedrin
promoter.
Generation of recombinant viruses. Viruses expressing the
h-subunit as well as the hTSH-ß minigene were generated
independently by cotransfecting 4 µg of linearized BacPAK viral DNA
(Clonetech) with either 4 µg of pBacPAK8/h-subunit cDNA or 4 µg
of pVL1392/hTSH ß-minigene, respectively into Spotoperda
frugiperda (SF)-9 insect cells to produce individual plaques,
using a cationic liposome formulation according to the manufacturer’s
instructions. Homologous recombination was visualized by blue color in
the presence of Bluo-Gal, and seven to ten recombinant plaques were
isolated, purified, and amplified for both constructs. Subsequently,
recombinant viruses were screened for the presence of h-subunit cDNA
or hTSH ß-minigene using the polymerase chain reaction. Positive
recombinants were amplified to a high titer viral stock and assayed via
plaque analysis to determine the plaque forming units (PFU)/ml.
Analysis of hormone expression. The insect cells SF-9 and
SF-21 were maintained in Grace’s medium (GIBCO-BRL, Gaithersburg, MD)
supplemented with 10% heat-inactivated FBS at 27 C, whereas the High
Five cells (Cell Trends, Inc., Middletown, MD) were grown in Ex-Cell
405 (GIBCO-BRL) supplemented as above. All cell lines were seeded at
300,000 cells/ml in spinner flasks at 1/3 of the rated volume per
flask. At log phase, the cells were coinfected with recombinant virus
producing the h-subunit and the hTSH ß-minigene at a multiplicity
of infection (MOI) of 1.7–5.0 PFU/ml. Supernatant was taken daily for
up to 4 days post infection for time course experiments and assayed for
hTSH immunoreactivity using dimer-specific immunoassays (Nichols
Institute, San Juan Capistrano, CA; Diagnostic Products Corp., Los
Angeles, CA; Ciba-Corning, Medfield, MA).
Expression of hTSH in eukaryotic cells
CHO, Cos-7, Cos-1, 293, and GH3 cells were transiently
cotransfected in 60 mm culture dishes at 80% confluence with the
pcDNA/NeoI h-subunit cDNA and the pLBCMV/hTSH
ß-minigene, using a previously optimized modified liposome
formulation (Lipofectamine reagent, GIBCO-BRL)-based method (29). After
subsequent culture in serum-free medium (CHO-SFM, GIBCO-BRL) for 48–72
h, supernatants were harvested and assayed for hTSH
immunoreactivity.
Lectin binding studies
IC-hTSH and CHO-hTSH binding to Concanavalin A (Con A) Sepharose
(Pharmacia) as well as to the immobilized limax flavus agglutinin (LFA,
EY Laboratories, San Mateo, CA) was studied. One milliliter of Con A
Sepharose was put into Pasteur pipettes, and after equilibration with
20 mM Tris buffer containing 0.5 M NaCl, pH
7.4, 50–100 ng of the recombinant hTSH preparations were loaded. Con
A-bound hTSH was eluted with increasing concentrations of
methyl--D mannopyranoside, and hTSH
immunoreactivity quantitated with a heterodimer-specific immunoassay
(Nichols Institute). The LFA binding studies were performed as
described for the Con A Sepharose, with the exception that 0.05
M Tris, 0.3 M NaCl, pH 7.5 buffer was used, and
0.01 M sialic acid for elution.
Oligosaccharide processing inhibition
CHO cells stably expressing hTSH (a kind gift from the Genzyme
Corp.) were grown to 70% confluence in the large T-150 flasks in DMEM
supplemented with 10% FBS. Castanospermine, an -glucosidase
inhibitor and swainsonine, an -mannosidase inhibitor were added in
medium at 200 µg/ml and 100 µM, respectively. Medium
was replaced daily for 4 days to purge the cells of wild type rhTSH. On
day 5, medium was collected from duplicate plates including from
untreated control cells, pooled, centrifuged to remove debris and
concentrated/washed using an Omega (Los Angeles, CA) Stirred Cell with
a 10K mol wt cut off (Filtron, Beverly, MA). During this processing
swainsonine (100 µM) was added to the
castanospermine-treated samples to protect the oligosaccharides from
any -mannosidase degradation.
Enzymatic desialylation
Concentrated conditioned media were incubated with 250 µU
neuraminidase attached to beaded agarose (Sigma) per 10 mg total
protein in 100 mM sodium acetate, pH 5.0, for 12 h at
room temperature, followed by 1 h at 37 C. After separation of the
neuraminidase by spinning in a microcentrifuge, media were washed,
concentrated, and reassayed for hTSH immunoreactivity.
Immunoassays of hTSH
IC-hTSH and CHO-hTSH were quantified using three different third
generation sandwich hTSH immunoassays, obtained from the Nichols
Institute, DPC, and Ciba-Corning following manufacturers’ instructions
(22, 29).
In vitro activity
cAMP production in JP09 cells and FRTL-5 cells. CHO cells
stably expressing the rhTSH receptor, clone JP09 (30), from Dr. G.
Vassart, (Brussels, Belgium), and FRTL-5 cells expressing the
endogenous rat TSH receptor (31) from Dr. L. Kohn, (Interthyr Research
Foundation, Baltimore, MD) were grown in 96-well tissue culture plates
to confluence as previously described (19). For the determination of
cAMP production, cells were incubated with serial dilutions of hTSH
preparations in a modified Krebs ringer buffer (KRB). This buffer did
not contain NaCl, but 280 mM sucrose to maintain
physiological osmolarity, and 1 mM
3-isobutyl-1-methylxanthine, an inhibitor of phosphodiesterase (29).
The amount of cAMP released into the medium was assayed by a cAMP RIA
(28).
Growth assay in FRTL-5 cells. FRTL-5 cells, grown for 6 days at 30% confluence in the absence of bovine TSH (bTSH) were incubated for 48 h with various concentrations of IC-hTSH or CHO-hTSH. Subsequently, 1 µCi 3H-thymidine (DuPont, Wilmington, DE) per well was added, and after an additional 24 h, cells were washed, solubilized and radioactivity incorporated into the DNA was measured by liquid scintillation spectrometry (Beckman Instruments, Columbia, MD) as described (29, 32).
RRA of hTSH
The receptor-binding activity of the recombinant hTSH
preparations was determined by their ability to displace
125I-bTSH from a solubilized porcine thyroid membrane
receptor preparation (Kronus, Dana Point, CA), using a 0.01
M Tris-Cl Buffer, pH 7.4, containing 0.05 M
NaCl (28, 32).
In vivo bioactivity
The in vivo bioactivity of IC-hTSH in comparison with
CHO-hTSH was assessed with a bioassay recently developed in our
laboratory (33). Briefly, endogenous TSH secretion of male albino Swiss
Crl:CF-1 mice was suppressed by adding 3 µg/ml T3 (Sigma)
to their drinking water for 5–6 days. Six hours after ip hTSH
injection, blood samples for determination of T4
(T4 Kit, Nichols Institute) and hTSH values was drawn from
the orbital sinus.
Metabolic clearance rate
The metabolic clearance rate of IC-hTSH and CHO-hTSH was
determined in the rat after iv injection of both preparations and
subsequent determination of hTSH serum levels at defined intervals from
1–120 min. Experimental details of this procedure are given elsewhere
(22, 28).
Experimental animals
Animal experiments using rats and mice were conducted in accord
with the standards of animal care, as outlined in the NIH guidelines
for care and use of experimental animals.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows that coinfection these cell lines with a 1:1
ratio of - and ß-subunit producing virus at an MOI of 5.0 PFU/ml
yielded the highest levels of hTSH in the High Five cell line. Further
time course experiments in the High Five cells (Fig. 1b) showed that
expression of hTSH reached maximal levels 4–5 days after infection,
which is typical for the use of a very late polyhedrin promoter (23, 24). Infection with - to ß-subunit producing virus at a ratio of
1:1 resulted in higher hTSH levels than with an -subunit excess at a
3:1 ratio, which was opposite from our experience with mammalian cells
(22, 29, 32), indicating that system-dependent factors may be
responsible for these differences. More importantly, maximal expression
reached 800-1000 ng/ml and was thus more than 30-fold higher than in
our optimized transient transfection system using CHO cells (29), where
levels only ranged from 20–50 ng/ml, demonstrating the superiority of
the IC-cell system compared with an eukaryotic transient transfection
system using CHO cells. This was not related to low expression peculiar
to CHO cells, as these cells yielded higher hTSH levels than other
mammalian cell lines tested, including Cos-7, Cos-1, 293 embryonic
kidney, GH3 cells (data not shown).
|
-D-mannopyranosyl and -D-glucopyranosyl
residues, whereas LFA recognizes terminal sialic acids (22). Equimolar
amounts of the different hTSH preparations were applied to lectin
columns and eluted with the appropriate competitor, 10–1000
mM methyl--D mannopyranoside for Con A and
10 mM sialic acid for LFA lectin columns, respectively. As
expected, CHO-hTSH, which bears 1.8–2.2 sialic acid residues per chain
(27, 28), bound to the LFA column better than IC-hTSH (51.1 ±
5.7% vs. 23.7 ± 4.5% of total applied,
P = 0.001). Following enzymatic desialylation, binding
of the CHO-hTSH decreased to similar levels (23.2 ± 6.6%) as for
IC-hTSH, indicating that the absence of sialic acids was responsible
for the lower binding of the IC-hTSH. Binding of the IC-hTSH did not
change after neuraminidase treatment (27.2 ± 2.2%), further
supporting the absence of sialic acid on the IC-hTSH. Conversely,
IC-hTSH bound better to the Con A Sepharose than CHO-hTSH (99.7 ±
0.25% vs. 74.8 ± 2.0% P < 0.001.
Enzymatic desialylation increased the binding of CHO-hTSH to Con A to
86.0 ± 3.0%, but as expected, not to the level of IC-hTSH, which
remained unchanged upon enzymatic desialylation (99.7 ± 1.0%).
Taken together, these data support the predominance of nonsialylated
high mannose-type oligosaccharides in our IC-hTSH, similar to what had
been reported for other insect cell-expressed glycoprotein hormones as
well as other glycosylated proteins (14, 23, 25). However, there are
major limitations in lectin analysis of intact glycoproteins compared
with labeled glycopeptides and these data can only be considered
semiquantitative. Additional studies with other exo- and
endoglycosidases as well as more direct analysis will be required to
confirm these potential structures when larger amounts of purified
IC-hTSH are available.
In vitro activity
The in vitro activity of IC-hTSH was 5-fold higher than
that CHO-hTSH, as shown by a 5-fold decrease in the EC50
(0.7 ± 0.3 ng/ml for IC-hTSH vs. 3.3 ± 0.8 ng/ml
for CHO-hTSH, P = 0.002) for cAMP stimulation in CHO
cells stably transfected with the human TSH receptor (JP09) (Fig. 2a). A similar 5- to 6-fold decrease in the
EC50 of the IC-hTSH was also observed, if cAMP production
(Fig. 2b), or 3H-thymidine incorporation as a marker of
cellular proliferation (Fig. 2c) in the rat cell line FRTL-5 was
assessed, indicating that these results were neither species-specific
nor system-dependent. In accordance with previous studies assessing
carbohydrate composition on insect cell-expressed glycoproteins (14, 15) and supported by our own data using lectin binding, the most likely
explanation for these differences in activity would be differences in
carbohydrates between the hTSH of the two cell types.
|
).
Castanospermine is an indolizidine alkaloid from the Australian
chestnut tree Castanosperun australe, a potent inhibitor
of both - and ß-glucosidases including glycoprotein processing
glucosidases, glucosidase I and II (34). In animal cell culture,
castanospermine therefore causes production of Glc3
Man7–9-(GlcNac)2 glucosylated high
mannose-type N-linked oligosaccharides. Swainsonine, initially isolated
from the Australian plant Swainsona canescens inhibits Golgi
mannosidase II, thus giving rise to hybrid-type oligosaccharide
structures (34). After treatment of the hTSH-producing CHO cells with
both inhibitors (Fig. 4a), the activity of hTSH from the
treated cells increased to almost similar levels as the IC-hTSH,
supporting previous findings (22) that complex sialylated
oligosaccharides attenuate signal transduction at the TSH receptor. The
EC50 of hTSH from the swainsonine-treated cells was
slightly, but reproducibly, higher than from castanospermine treated
cells (1.7 ± 0.8 ng/ml vs. 0.7 ± 0.3 ng/ml), as
could be expected because this compound allows for the formation of
complex chains on one antenna of the hybrid. Figure 4b shows that this
inhibitory effect of the complex oligosaccharides appeared to be due to
their terminal sialylation because enzymatic desialylation increased
the activity of CHO-hTSH 7-fold (EC50 0.45 ± 0.2
ng/ml) to similar levels as for IC-hTSH. In contrast, the activity of
IC-hTSH remained unchanged after neuraminidase treatment
(EC50 0.6 ± 1.5 ng/ml), thus indirectly supporting
the absence of sialic acids on the IC-hTSH. Remarkably, these findings
on the inhibitory role of sialic acids on the in vitro
activity of hTSH are in contrast to findings for the gonadotropins
because insect cell-expressed or enzymatically desialylated hCG or hFSH
either had unchanged or even decreased in vitro activity
compared with the sialylated hormone (14, 15, 16, 17, 18, 35).
|
|
). This was in good agreement with our previous studies
on hTSH using sequential enzymatic deglycosylation (36), glycosylation
mutant cell lines (22) as well as with studies on hFSH and hCG
(11, 12, 13), showing that oligosaccharides modulate glycoprotein hormone
signal transduction at a postreceptor binding step. The relatively low
hTSH binding affinity (half-maximal binding inhibition 45 ± 8
ng/ml) compared with its EC50 for cAMP induction (3.3
± 0.8 ng/ml) has been consistently observed in our previous studies on
hTSH structure-function relationships (22, 28, 29, 32) and by others
(30). These findings are in accord with the concept of signal
amplification along the signal transduction cascade common for the
glycoprotein hormones. Moreover, they may be related to the higher
binding affinity of bovine TSH (used a the tracer) compared with human
TSH (28, 29, 30), and/or quantitative or qualitative differences in the TSH
receptor.
|
). Intraperitoneal injection of
CHO-hTSH led to a 3-fold stimulation hTSH with the highest dose tested,
which was 1000 ng/animal injected ip (P < 0.01). In
contrast, there was no significant T4 stimulation with the
IC-hTSH, indicating a low in vivo activity. Concurrent
determination of hTSH serum levels in the mice showed that, 6 h
after ip injection, 7.4–11% of the injected CHO-hTSH was detectable
in the serum, compared with less than 0.1% of the IC-hTSH, suggesting
that the IC-hTSH was cleared from the circulation more rapidly than the
CHO-hTSH.
|
). These data are best explained
by the fact that terminal sialic acids protect CHO-hTSH from clearance
by the hepatic galactose receptor (37), whereas IC-hTSH is susceptible
to clearance by the hepatic mannose receptor (38).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our baculovirus system, immunoreactive hTSH was efficiently secreted from the insect cells, and expression levels were considerably higher than levels obtained with our optimized transient transfection system using CHO and other eukaryotic cells (22, 29, 32). Remarkably, the IC-hTSH had a higher in vitro activity than the hTSH produced from CHO cells, whereas its in vivo activity was reduced. Our results suggest that the disparate effects in the in vitro and in vivo activity were related to the different glycosylation patterns, and predominantly to the different degrees of terminal sialylation of IC-hTSH and CHO-hTSH. This was supported by treatment of hTSH-producing CHO cells with glycosylation processing inhibitors, enzymatic desialylation of the hTSH preparations, as well as Con A and LFA lectin binding studies.
Unlike thyrotrophs of the anterior pituitary gland, CHO cells do not
express GalNAc-transferase and GGnM-4-sulfotransferase (13, 39).
Therefore, in contrast to pituitary hTSH, recombinant CHO-hTSH does not
contain sulfated oligosaccharides but terminates in sialylated chains
Sia2-3Galß1-4GlcNAcß1-2Man (13, 28, 39). Only very few
studies have analyzed insect cell-expressed glycoprotein carbohydrates
by compositional analysis. These studies suggested that the
glycosylation machinery of insect cells does not possess the same
capacity to process N-linked oligosaccharides to complex-type
carbohydrates as mammalian cells (14, 25). Alternative or complementary
explanations have been proposed, including lack or down-regulation of
appropriate glycosyltransferases, relative insufficiency of the insect
cell processing machinery for the secretory load, and direct cytopathic
effects of the virus (23, 24, 25). Accordingly, our lectin binding studies
before and after desialylation indicated that, in comparison with
CHO-hTSH, the IC-hTSH lacked complex sialylated oligosaccharides and
predominantly had high mannose-type precursors. Likewise, IC-hCG
expressed from a polyhedrin promoter was found to contain high
mannose-type oligosaccharides by carbohydrate compositional analysis,
whereas sialic acid residues were not detected (15, 16). Evidence that
the temporal nature of the baculovirus promoter may affect
glycosylation came from a recent study that showed that the hTSH
receptor was less glycosylated if expressed from a very late compared
with a late promoter (40).
Compared with CHO-hTSH, the IC-hTSH had a higher in vitro activity in all the different assay systems used in this study. Further, the in vitro activity of CHO-hTSH also increased upon treatment of the hTSH-expressing CHO cells with the glycosylation inhibitors castanospermine and swainsonine, which both inhibit full processing to complex, terminally sialylated chains: castanospermine inhibits glucosidase I and II and causes the formation of glucosylated high mannose Glc3 Man7–9(GlcNAc)2 structures, whereas swainsonine inhibits mannosidase II and causes the accumulation of hybrid chain oligosaccharides (34). The fact that the in vitro activity of CHO-hTSH increased upon enzymatic desialylation indicated that the terminal sialic acid residues are primarily responsible for the reduced in vitro signal transduction of hTSH bearing complex-type oligosaccharides. This is in accord with our previous findings on expression of hTSH in CHO-glycosylation mutant cell lines as well as sequential deglycosylation using exoglycosidases (22, 36). In this respect, pituitary human TSH, which physiologically occurs in a variety of glycoforms and is predominantly sulfated, was shown to have a 3-fold higher in vitro activity than CHO-hTSH (41). Although we did not compare pituitary and IC-hTSH directly, it thus appears that pituitary hTSH is slightly less active in vitro than IC-hTSH, which could be due to the partial sialylation of the pituitary hTSH.
Interestingly, in contrast to our findings with IC-hTSH, insect cell-expressed hCG or hFSH had either a similar or even decreased in vitro activity compared with their sialylated counterparts (15, 16, 17, 18). Likewise, enzymatic desialylation (35) or expression in CHO glycosylation mutant cell lines deficient in terminal sialylation (42, 43) led to a decreased in vitro activity of hCG or hFSH, again contrary to similar studies using CHO-hTSH (22, 36). Thus, terminal sialic acid residues affect the in vitro activity of the closely related glycoprotein hormones in a different fashion. In fact, terminal sialylation, introduced by expression in heterologous systems, attenuates the in vitro activity of glycoprotein hormones that are physiologically predominantly sulfated, such as TSH and LH (44). In contrast, terminal sialic acid residues are necessary for full in vitro activity of the physiologically sialylated CG and FSH.
The present findings also support the notion that the carbohydrates modulate signal transduction at a postreceptor binding step, as all the different hTSH preparations had similar receptor binding affinities. The precise molecular basis how the oligosaccharides influence signal transduction remains to be elucidated, as there is no structural information on ligand-receptor complexes. An indirect mechanism involving a conformational change appears more likely than a direct activation of the receptor by the oligosaccharide moiety (45) because a lectin-like component identified in the hCG receptor is not present in the hTSH receptor (11).
Unlike the increase in in vitro activity, IC-hTSH had a much lower in vivo activity than CHO-hTSH. This was related to its much more rapid clearance that superseded its higher in vitro activity. Likewise, clearance rates of hTSH glycosylation isoforms had a larger impact on the in vivo activity than the in vitro activity (28), emphasizing that clearance contributes to the final in vivo activity to a greater extent than does in vitro activity. The increased clearance rate of the IC-hTSH further supported the absence of or reduction in terminal sialic acid residues of the IC-hTSH because terminal sialic acid residues protect glycoprotein hormones from carbohydrate-specific hepatic receptor-mediated clearance mechanisms (46), which include the asialoglycoprotein-, N-acetylgalactosamine/sulfate or mannose receptors (37, 38, 47). Therefore, it is likely that susceptibility to the hepatic mannose receptor explained the rapid clearance of the IC-hTSH. In this respect, sialylated CHO-hTSH was shown to be predominately distributed to the kidneys even in the earliest phase of clearance, whereas enzymatically desialylated CHO-hTSH was cleared in a similarly rapid fashion to IC-hTSH in this study by the liver, with only minor involvement of other organs (46). The importance of terminal sialic acid residues in the maintenance of serum half life has also been established for the other members of the glycoprotein hormone family CG, LH, and FSH (11, 12, 13, 47). Taken together, these findings indicate that terminal sialylation has similar effects on the clearance and thus in vivo activity of all glycoprotein hormones but modulates in vitro activity in a hormone-dependent fashion. In this respect, control of terminal sialylation has proven, in addition to the classical negative TSH/T3 feedback regulation, to constitute a physiologically relevant mechanism to modulate TSH activity in the human (2, 48). To our knowledge, there have only been very few reports investigating the clearance or in vivo activity of insect cell-expressed glycosylated proteins. Interestingly, the insect cell-expressed glycoprotein metalloproteinase-1 tissue inhibitor was cleared from the circulation within min after iv injection and distributed mainly to the liver (49), suggesting that insect cell-expressed glycoproteins may generally be susceptible to rapid clearance. It is tempting to speculate that the unusual efficiency of high mannose- and asialo-glycoprotein clearance mechanisms may have evolved in vertebrates to eliminate or reduce the in vivo activity of glycoproteins with incompletely processed oligosaccharide chains.
In summary, this study shows for the first time that it is possible to produce biologically active hTSH in insect cells. The high in vitro activity of IC-hTSH, together with the relatively high expression levels compared with mammalian expression systems, make this approach attractive for in vitro structure-function studies of recombinant hTSH analogs. However, further modification of the IC-hTSH to prolong its half life, such as the in vitro processing of oligosaccharides, modification of the baculovirus promoter (40) or coinfection with glycosyltransferases will be necessary to increase its in vivo potency. This should also be relevant for other insect cell- expressed glycoproteins. At the same time, this study highlights the dual function of oligosaccharides, particularly of the terminal sialic acids, for the in vitro and in vivo activity of hTSH.
|
Footnotes |
|---|
Received September 20, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
subunit.
Glycobiology 5:3–10- subunit. J Biol Chem 268:21613–21617-subunit in the expression and bioactivity of human
thyroid-stimulating hormone. Mol Endocrinol 9:948–958[Abstract]
-subunit reveals different structural
requirements for heterodimer expression among the glycoprotein hormones
and suggests that cAMP production and growth promotion are potentially
dissociable functions of hTSH. Mol Endocrinol 10:769–779[Abstract]
that mediates rapid
clearance of lutropin. Cell 67:1103–1110[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  S. Donadio, A. Pascual, J. H.H. Thijssen, and C. Ronin Feasibility Study of New Calibrators for Thyroid-Stimulating Hormone (TSH) Immunoprocedures Based on Remodeling of Recombinant TSH to Mimic Glycoforms Circulating in Patients with Thyroid Disorders Clin. Chem., February 1, 2006; 52(2): 286 - 297. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Gunawan, D. Simard, M. Gilbert, A. L. Lovering, W. W. Wakarchuk, M. E. Tanner, and N. C. J. Strynadka Structural and Mechanistic Analysis of Sialic Acid Synthase NeuB from Neisseria meningitidis in Complex with Mn2+, Phosphoenolpyruvate, and N-Acetylmannosaminitol J. Biol. Chem., February 4, 2005; 280(5): 3555 - 3563. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Aumiller, J. R. Hollister, and D. L. Jarvis A transgenic insect cell line engineered to produce CMP-sialic acid and sialylated glycoproteins Glycobiology, June 1, 2003; 13(6): 497 - 507. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. A. Oliveira, L. Persani, P. Beck-Peccoz, and J. Abucham Investigating the Paradox of Hypothyroidism and Increased Serum Thyrotropin (TSH) Levels in Sheehan's Syndrome: Characterization of TSH Carbohydrate Content and Bioactivity J. Clin. Endocrinol. Metab., April 1, 2001; 86(4): 1694 - 1699. [Abstract] [Full Text]
|
|
|
|
|
|
Y. Nagayama, H. Namba, N. Yokoyama, S. Yamashita, and M. Niwa Role of Asparagine-linked Oligosaccharides in Protein Folding, Membrane Targeting, and Thyrotropin and Autoantibody Binding of the Human Thyrotropin Receptor J. Biol. Chem., December 11, 1998; 273(50): 33423 - 33428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Grossmann, B. D. Weintraub, and M. W. Szkudlinski Novel Insights into the Molecular Mechanisms of Human Thyrotropin Action: Structural, Physiological, and Therapeutic Implications for the Glycoprotein Hormone Family Endocr. Rev., August 1, 1997; 18(4): 476 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Lawrence, K. A. Huddleston, L. R. Pitts, N. Nguyen, Y. C. Lee, W. F. Vann, T. A. Coleman, and M. J. Betenbaugh Cloning and Expression of the Human N-Acetylneuraminic Acid Phosphate Synthase Gene with 2-Keto-3-deoxy-D-glycero- D-galacto-nononic Acid Biosynthetic Ability J. Biol. Chem., June 2, 2000; 275(23): 17869 - 17877. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
, N-acetylglucosamine; 
, mannose; 
,
N-acetylgalactosamine; 
, sialic acid.
counter. Radioactivity
precipitated in the presence of concentrated medium from mock
transfections was defined as 100%. Values of a representative
experiment are shown as mean ± SEM of triplicate
determinations (see legend to Fig. 2
