Molecular Human Reproduction vol.4 no.7 pp. 631–639, 1998
Glycoform composition of serum gonadotrophins through the
normal menstrual cycle and in the post-menopausal state
C.J.Anobile1, J.A.Talbot1, S.J.McCann1, V.Padmanabhan2 and W.R.Robertson1,3
1University
of Manchester, Department of Medicine, Hope Hospital, Salford, M6 8HD, UK, and 2Department of Pediatrics
and Reproductive Sciences Program, University of Michigan, 300 N.Ingalls Building, Ann Arbor, MI 48109–0404, USA
3To
whom correspondence should be addressed
The heterogeneity of follicle stimulating hormone (FSH) and luteinizing hormone (LH) was investigated in
five women aged 29.4 K 3.2 years (mean K SD) throughout their menstrual cycles and in five post-menopausal
women aged 53.8 K 5.6 years. Chromatofocusing (pH range 7–4) revealed menstrual cycle stage- and postmenopausal-related differences in the serum gonadotrophin charge. There were differences in the proportion
of FSH with an isoelectric point (pI) >4.3 across phases of the menstrual cycle (P J 0.019): midcycle (MC)
50%; early to mid-follicular (EMF) 36%; late follicular (LF) 37%, luteal (L) 29% and following the menopause
(PM) 17%. There was no significant difference in the proportion of LH with pI >6.55 between midcycle (53%)
and EMF, LF or L phases (36, 43 and 32% respectively); although all were greater than that found in the
menopause (13%). Concanavalin A chromatography revealed less (P < 0.005) complex FSH and LH glycoforms
at midcycle (63 and 13%) than in the EMF, LF and L phases (90 and 18; 90 and 20 and 93 and 24% respectively).
Menopausal gonadotrophins were least complex (FSH 34%, LH 4%). There was a direct relationship between
serum FSH and FSH pI/complexity, and less acidic FSH was associated with reduced FSH complexity. Increased
oestradiol was associated with basic FSH isoforms during the menstrual cycle and reduced follicular phase
FSH complexity. We conclude that changes in gonadotrophin glycoforms occur through the menstrual cycle
which are related to changes in the prevailing steroid environment. Following the menopause oestrogenic
loss resulted in acidic, relatively simple glycoforms.
Key words: FSH/glycoforms/isoforms/LH
Introduction
The pattern of circulating follicle stimulating hormone (FSH)
and luteinizing hormone (LH) is dynamic not only with
respect to concentration (quantity) but also with regard to the
heterogeneity (quality) of the isoform mixture present
(Chappel, 1995; Ulloa-Aguirre et al., 1995). This heterogeneity
appears to be a consequence of the asparagine-linked oligosaccharides of these αβ-heterodimeric pituitary glycoproteins
leading to the existence of numerous so called ‘glycoforms’
of these hormones (Green and Baenziger, 1988; StockellHartree and Renwick, 1992).
There is now increasing evidence that the extent and pattern
of glycosylation is under hormonal control, and originates
most likely from a combination of steroidal feedback and the
influence of gonadotrophin-releasing hormone (GnRH) (UlloaAguirre et al., 1988a, 1995; Peréz and Apfelbaum, 1996). What
is not clear is whether the change in glycoform distribution has
implications in terms of the hormones’ biological activities
and bears any direct relevance to the physiology and pathology
of the pituitary–gonadal axis. Investigations have shown that
the metabolic clearance rate of the less acidic FSH isoforms
is faster (Blum and Gupta, 1985; de Leeuw et al., 1996).
Serum half-life may therefore be a major determinant of invivo biological activity and consequently present a further
level of control of gonadal function in that slower and faster
© European Society for Human Reproduction and Embryology
acting forms of the hormones are secreted. Such a possibility
has been contested for LH isoforms by Burgon et al. (1996),
who showed that the major contributory factor in in-vivo
biopotency was not the circulating half life but rather the
activity of the isoform as assessed in in-vitro bioassays. Even
more exciting is the possibility that the different glycoforms
may encode for different functions, or alternatively have
enhanced activity for a particular function (Burgon et al.,
1993; Mitchell et al., 1994; Padmanabhan, 1995; Robertson
et al., 1997).
If the above premise proves to be true then it may have
considerable relevance to our understanding of the role of
glycoforms in the control of gonadal function. In this context
it is of interest that concurrent with changes in LH and FSH
concentration during the menstrual cycle there are marked
changes in their charge heterogeneity, such that the least acidic
isoforms predominate during midcycle (Padmanabhan et al.,
1988; Wide and Bakos, 1993; Zambrano et al., 1995). In
addition such changes in gonadotrophin heterogeneity appear
to be related to the concentration of circulating oestradiol
(Padmanabhan et al., 1988; Wide and Naessén, 1994), and
progesterone has been shown to counteract these effects in
the short term (Wide et al., 1995, 1996). Furthermore, the
complexity of the branching structure of the N-linked oligosaccharides has also been shown to vary with endocrine status
(Papandréou et al., 1993; Creus et al., 1996). Analysis of the
631
C.J.Anobile et al.
complexity and isoelectric point (pI) of the isospecies of
pharmaceutical FSH preparations have also revealed marked
differences between urinary and recombinant preparations
(Lambert et al., 1995; Harris et al., 1996; Robertson, 1997).
Further knowledge of the biological role of the different
glycoforms may influence the development of different types
of ‘designer’ recombinant forms for use in assisted reproduction
programmes.
Studies investigating aspects of the changes in heterogeneity
of LH and FSH with respect to charge and carbohydrate
complexity during different cycle stages have not been performed thus far. Other than the molecules circulating during the
follicular phase (Creus et al., 1996), comparative investigations
into FSH and LH complexity have not been reported for
menstrual cycle stages. While Wide and Bakos (1993) investigated the charge characteristics of both LH and FSH, they
have not characterized the glycoform distribution that has led
to such changes. Since both FSH and LH play important roles
during the menstrual cycle and FSH carbohydrate complexity
is related to charge at least for purified preparations (UlloaAguirre et al., 1988b, 1992; Harris et al., 1996), the aim of
our study was to determine if such a relationship holds true
for circulating glycoforms. To this end we examined the
relationship between charge and complexity of circulating LH
and FSH during different phases of the menstrual cycle in
individual normal subjects and compared these with the postmenopausal situation. Clearly if the relationship between
gonadotrophin heterogeneity and differential or preferential
biological activities can be elucidated further then the knowledge of changes in FSH and LH glycoforms throughout
the menstrual cycle will become important to our overall
understanding of the role of the hormone species in the
processes of folliculogenesis and ovulation.
Materials and methods
Subjects and specimens
Normal, healthy women volunteers (n 5 5; age 29.4 6 3.2 years,
mean 6 SD) who had regular (26–30 days) menstrual cycles and
were not using the contraceptive pill were recruited. Serum was
collected on days corresponding to the early to mid-follicular, late
follicular, midcycle and luteal phases of a single menstrual cycle (7,
11, 14 and 21 days following the first day of menses respectively).
The days of sampling and corresponding phases were calculated for
an average 28 day cycle. The LH, FSH, oestradiol and progesterone
contents of samples were determined to confirm the cycle stage
(Table I). Post-menopausal sera from five healthy women were
obtained from the Hospital’s routine endocrine screen and fulfilled
the following requirements (mean 6 SD): age 53.8 6 5.6 years (.12
months amenorrhoea); LH 10–60 (55.6 6 11.2) IU/l; FSH .30
(99 6 17.3) IU/l; oestradiol ,90 (all subjects ,90) pmol/l; the
subjects were not undergoing hormone replacement therapy. Serum
samples were divided into aliquots and stored at –40°C.
Hormone assays
LH bioassay
Serum LH bioactivity was measured in vitro by the modified dispersed
mouse Leydig cell assay (van Damme et al., 1974) as described in
detail by Robertson and Bidey (1990). Samples were diluted to 1–
632
8% with Dulbecco’s modified Eagle’s medium (DMEM; Sigma
Chemical Co, Poole, UK) containing 20 mM HEPES and the assay
calibrated against an LH (80/552) standard curve. Samples were
removed from duplicate wells and the end-point was measured by
in-house testosterone radioimmunoassay as described by Talbot et al.
(1990a,b). The bioassay sensitivity was 2.4 IU/l and that of the
testosterone radioimmunoassay was 1 IU/l, the inter- and intraassay coefficients of variation for both assays were ,12 and ,8%
respectively.
Immunoassays for serum hormones
Serum samples were assayed for LH, FSH, oestradiol and progesterone
using the appropriate Delfia® (Wallac Oy, Turku, Finland) timeresolved fluorescence immunoassay kits following the manufacturer’s
instructions. The sensitivities of the serum assays were defined as the
lowest kit standard: 0.6 IU/l LH, 1 IU/l FSH, 50 pmol/l oestradiol
and 1 nmol/l progesterone. The inter- and intra-assay coefficients of
variation were ,15 and ,6% for all assays respectively.
Immunoassays for LH and FSH following chromatography
Samples obtained from chromatofocusing and lectin affinity chromatography were also assayed for FSH and LH content in the appropriate
Delfia®. Standards (FSH: IRP 78/549; LH: IRP 80/552) covering the
range 0.025–50 IU/l were prepared in buffers which were matrixmatched with the elution buffer for the separation technique employed:
(i) lectin affinity chromatography, phosphate buffer, 0.05 M, pH 7.4,
0.1% (w/v) bovine serum albumin (BSA), 0.1% sodium azide; (ii)
chromatofocusing, Polybuffer 74 (Parmacia, Milton Keynes, UK)
diluted 1:35 with dH2O, pH 7.4, 0.1% (w/v) BSA.
To detect low concentrations of gonadotrophins following chromatofocusing and concanavalin A (con-A) chromatography, the LH and
FSH Delfia assay sensitivity was improved by increasing the volume
of samples and standards from the recommended 25–50 µl and by
filtration of tracer solution through a 0.2 µm syringe filter (Costar,
Cambridge, MA, USA). These assays employed five quality controls
(0.15, 0.3, 0.8, 1.5 and 20 IU/l FSH/LH) and the intra- and interassay coefficients of variation were calculated as ,6 and ,15%
respectively. The assay sensitivity (LH and FSH, phosphate and
polybuffer standards) was 0.05 IU/l defined as the lowest measurable
concentration of hormone for which the intra-assay coefficient of
variation fell ,10%. None of the buffers employed interfered in the
immunoassays as determined by maintenance of parallelism over the
range of the standard curve when samples of known FSH/LH content
were diluted with standard matrix buffer; nor did they affect the
background counts for any given assay.
Concanavalin A affinity chromatography
This was performed by employing a method modified from Harris
et al. (1996). This lectin binds with high affinity to oligosaccharides
with high mannose content or hybrid structures (designated here as
‘simple’ glycoforms) and has less affinity for biantennary and
truncated hybrids (‘intermediate’ glycoforms). Con-A will not associate with proteins bearing triantennary, tetra-antennary and bisecting
oligosaccharides (‘complex’ glycoforms). Undiluted serum (1 ml)
was applied to a pre-equilibrated mini-column containing 2.5 ml conA–agarose matrix (Sigma). The proteins were eluted consecutively
using three buffers: con-A column buffer, 7 ml (0.05 M phosphate,
pH 7.4, 0.1% (w/v) BSA); α-methyl glucopyranoside, 28ml (α-MG;
10 mM in con-A column buffer) and α-methylmannopyranoside,
13 ml (α-MM; 300 mM in con-A column buffer) to remove the
unbound, weakly bound and firmly bound proteins respectively.
Buffers were passed through the column at a flow rate of 10 ml/h at
room temperature and 1 ml fractions were collected. Fractions were
Gonadotrophin heterogeneity
Table I. Endocrine screen of volunteers. Data shown are mean 6 SD
Age
(years)
Cycle
stage
FSH
(IU/l)
29.4 6 3.2
EMF
LF
MC
L
PM
5.1
5.2
10.0
3.2
99
53.8 6 5.6
6
6
6
6
6
LH
(IU/l)
0.4
1.3
1.5a
1.2
17.3
4.2
6.9
37
3.5
55.6
6
6
6
6
6
[LH]bio
(IU/l)
1.6
1.8
6.8a
0.8
11.2
6.4
14.5
101
5.0
138
6
6
6
6
6
LH
B:I ratio
2.3
3.1
37a
1.8
31
1.7
1.9
2.8
1.4
2.5
6
6
6
6
6
0.2
0.3
0.7a
0.2
0.2
Progesterone
(nM)
1.1
1.4
2.7
36
ND
6
6
6
6
0.2
0.2
0.5
2.4
Oestradiol
(pM)
218 6 42
606 6 203
792 6 113*
455 6 91
,90
Serum samples taken during the early-mid follicular (EMF), late follicular (LF), midcycle (MC) and luteal (L) phases of the menstrual cycle of five
volunteers and five post-menopausal women (PM) were assayed for follicle stimulating hormone (FSH), luteinizing hormone (LH), progesterone and
oestradiol for confirmation of cycle stage or post-menopausal status. An oestradiol concentration of ,90 pM indicates that all subjects had oestradiol
concentrations which were less than the normal working range of the assay.
ND 5 not determined; B:I 5 ratio of biological to immunological activity.
aMidcycle value is significantly different (P , 0.005) from the values given for other cycle stages.
stored at –40°C until hormone assay. Samples were assayed without
dilution and in addition every 10th sample was assayed in duplicate
at two dilutions to assess parallelism which was maintained for all
samples. The lowest detectable concentrations of FSH and LH fell
within the working range of the immunoassays and the recoveries of
FSH and LH from the con-A column were 102 6 7% and 92 6 9%
(mean 6 SD) respectively.
Chromatofocusing
Preparation of buffers
All buffers were prepared with distilled deionized water (ddH2O).
The start buffer was bis (2-hydroxyethyl) imino-tris (hydroxymethyl)
methane (‘bis-tris’, 7.14mM, pH 7.3–7.4). The elution buffer was
Polybuffer 74 at a dilution of 1:35, pH 3.1–3.2. The pH of both
buffers was adjusted with saturated iminodiacetic acid (Sigma) prior
to degassing (4–18 h) and filtering through a 0.2 µm bottle filter
(Nalgene; Nunc International, Rochester, NY, USA).
Sample preparation
Serum samples were diluted with bis-tris and desalted on a PD-10
column (Pharmacia). A final dilution was made with bis-tris such
that 8.5 ml of the diluted sample was equivalent to 1 ml of serum.
The recoveries of FSH and LH from the PD-10 column were 90–
100%. The sample was degassed for 30 min prior to injection onto
the equilibrated Mono-P column (Pharmacia).
Chromatofocusing
Chromatofocusing was performed on a 4 ml Mono-P column in
conjunction with a fast performance liquid chromatography system
(Pharmacia). The resin was washed with NaCl, 2 M (0.5 ml) to
exchange the storage counter ion and then equilibrated with 30–50 ml
bis-tris (0.5 ml/min flow rate) until the pH of the eluent (column pH)
was equal to that of the start buffer. The desalted, degassed sample
(8.5 ml) was loaded onto the column and the pH gradient generated
by introducing the elution buffer at a flow rate of 0.5 ml/min.
Fractions (1 ml) were collected into tubes containing 50 µl of 2%
(w/v) BSA in dH2O and the pH measured. Fractions were pooled at
0.15 pH intervals between pH 7.0 and pH 3.5 and their pH adjusted
to 7.4 with dilute NaOH. The column was washed with 0.5 ml 2 M
NaCl and the proteins with pI ,3.5 eluted with 10 ml bis-tris (salt
peak). Pools were stored at –40°C until hormone assay. Samples were
assayed without dilution and in addition every 10th sample was
assayed at two dilutions in duplicate to assess parallelism which was
maintained for all samples. The lowest detectable concentrations of
FSH and LH fell within the working range of the immunoassays and
recovery from the Mono-P column was 80–95% for LH and 80–102%
for FSH. Analysis of data from chromatofocusing was performed by
measuring the proportion of material eluting above relevant pH cutoff points.
Statistical analyses
Data were normalized by expressing recoveries of glycoforms following each respective separation procedure as a percentage of the total
FSH or LH recovered. Menstrual cycle stage-related differences in
the glycoform distributions of serum gonadotrophins after separation
by chromatofocusing or con-A chromatography were determined by
repeated measures analysis of variance (ANOVA), paired t-test after
application of the Bonferroni adjustment or unpaired t-test for
comparison of menstrual cycle stages with the post-menopausal
state. Correlations between the gonadotrophins’ physico–chemical
characteristics and hormone concentrations were performed by
Spearman’s test.
Results
Endocrine changes during the menstrual cycle
The mean concentrations of the serum hormones FSH, LHimmuno, LHbio, oestradiol and progesterone for the different cycle
stages are summarized in Table I. The LH biological:immunological (B:I) was greater (P , 0.008) at midcycle than during
the early–mid-follicular and luteal phases (see Table I).
Carbohydrate complexity of FSH and LH glycoforms
For all subjects and all cycle stages FSH contained mainly
complex carbohydrate structures (84 6 16%) with the remainder being of the intermediate type (16 6 13%). The carbohydrate moieties of LH were mostly of the intermediate type
(66 6 5%) with 19 6 5% having complex and 14 6 3%
simple forms. The complexity of both FSH and LH was
variable through the menstrual cycle, but with a consistent
reduction of complex FSH and LH glycoforms at midcycle.
Examples of elution profiles of FSH and LH following this
procedure are shown in Figure 1.
There was relatively little variability between subjects for
any particular stage of the cycle and when the data for all
subjects were combined (Figure 2) FSH complexity was
reduced (paired t-test, P , 0.005) at midcycle compared with
early to mid-follicular, late follicular and luteal phases (63 6 11,
90 6 3, 90 6 3 and 93 6 5% respectively). A concomitant
increase (P , 0.005) in the proportion of intermediate glyco633
C.J.Anobile et al.
Figure 1. Distribution of follicle stimulating hormone (FSH) and luteinizing hormone (LH) glycoforms following concanavalin A
chromatography of an individual early-mid follicular phase serum sample as measured by matrix-matched Delfia® assays. The amount of
FSH or LH is given as the total amount which was eluted in each individual 1 ml fraction. An arrow indicates a change in elution buffer
from (1) phosphate buffer, 0.1% bovine serum albumin (BSA) (column buffer) to (2) 10 mM methylglucopyranoside (in column buffer) to
(3) 300 mM methylmannopyranoside (in column buffer).
early to mid-follicular (18 6 3 and 11 6 1%), late follicular
(20 6 4 and 14 6 3%) and luteal (24 6 10 and 14 6 3%) phases.
The patterns of glycoform complexity of post-menopausal
FSH and LH were different to those observed at any stage of
the menstrual cycle. The majority of FSH glycoforms were of
the intermediate type (64 6 8%) with 34 6 9% complex
forms and 2 6 1% simple glycoforms. Similarly for LH, most
glycoforms were of intermediate complexity (73 6 5%) with
very few complex forms (4 6 1%) and 23 6 6% simple
glycoforms.
Figure 2. Changes in the proportions of complex, intermediate and
simple glycoforms of follicle stimulating hormone (FSH) (top
panel) and luteinizing hormone (LH) (bottom panel) were assessed
by concanavalin-A chromatography of serum samples obtained
from five women at the early to mid-follicular (EMF), late
follicular (LF), midcycle (MC) and luteal (L) phases of the
menstrual cycle and from five post-menopausal (PM) women. Data
shown are mean 1 SD A significant difference (P , 0.005) from
the midcycle is indicated by an asterisk.
forms (36 6 11%) was also observed at midcycle compared
with the early-mid, late follicular and luteal phases (10 6 3,
10 6 3 and 7 6 5% respectively). The glycoform distribution
of LH also changed (P , 0.017 for all phases compared with
midcycle) in that there were fewer complex and more simple
forms at midcycle (13 6 3 and 17 6 3%) compared with the
634
Analysis of gonadotrophin charge distribution
through the menstrual cycle
Separation of glycoforms of FSH and LH according to their
net charge revealed that FSH was acidic in nature with most
material eluting at ,pH 5 whereas LH was less acidic
(Figure 3). When compared with the glycoform changes
revealed by con-A chromatography, the glycoform distribution
of LH and FSH was more variable both throughout the
menstrual cycle and between individuals. An example for one
subject is shown in Figure 3. Glycoform profiles of LH were
the most variable with no clear pattern of cycle stage-related
change emerging for any subject.
Data were normalized as a percentage of total eluting with
pI .4.3 and combined data for all subjects are shown in
Figure 4. The proportion of FSH eluting with pI .4.3 at
midcycle was 50 6 6%; during the early-mid follicular, late
follicular and luteal phases this proportion was 36 6 7,
37 6 7 and 29 6 11% respectively. These changes in FSH pI
distribution were significant across the menstrual cycle
(repeated measures ANOVA, P 5 0.019; paired t-test, P ,
0.017). There was no difference in LH glycoforms with pI
.6.55 at midcycle (53 6 16%) compared with the early to
mid-follicular (36 6 14%), late follicular (43 6 24%) and
luteal (32 6 20%) phases due to the degree of variation
between subjects.
Serum gonadotrophins from post-menopausal women contained FSH and LH glycoforms of a very acidic nature, such
Gonadotrophin heterogeneity
Figure 3. Charged isoform distribution of follicle stimulating hormone (FSH) (left panels) and luteinizing hormone (LH) (right panels)
resolved by chromatofocusing of serum samples taken from one woman at different stages of her menstrual cycle. Gonadotrophin
concentrations were determined by the appropriate matrix-matched Delfia® and are shown as mIU for each pH pool. N.B. Different y-axis
scales have been employed for clarity.
that only 17 6 10% of FSH glycoforms eluted with pI .4.3
and 14 6 10% of LH glycoforms eluted with pI .6.55; P ,
0.005 compared with all menstrual cycle stages (Figure 4).
Relationships between gonadotrophin pI, complexity
and circulating hormone concentrations
There was a negative correlation between the proportion of
FSH with pI .4.3 and complex forms of FSH for all cycle
stages (r 5 –0.633, P 5 0.003) and for the follicular–
midcycle phases (r 5 –0.614, P 5 0.015). Serum oestradiol
concentrations were positively correlated with the proportion
of FSH glycoforms with pI .4.3 at all cycle stages (r 5
0.474, P 5 0.035), yet when only the follicular to midcycle
stages were analysed, oestradiol was correlated solely with a
reduction in complex FSH glycoforms (r 5 –0.543, P 5
0.037). Serum FSH concentrations were positively correlated
with the proportion of glycoforms with pI .4.3 (r 5 0.727,
P , 0.001) and negatively with the proportion of complex
FSH glycoforms (r 5 –0.745, P , 0.001) during all menstrual
cycle stages. The properties of LH glycoforms were also
635
C.J.Anobile et al.
Figure 4. The proportion of follicle stimulating hormone (FSH)
isoforms eluting with an isoelectric point (pI) .4.3 (top panel) and
luteinizing hormone (LH) isoforms eluting with pI .6.55 (bottom
panel) after chromatofocusing of serum samples from five women
during the early to mid-follicular (EMF), late follicular (LF),
midcycle (MC) and luteal (L) phase of the menstrual cycle and
from five post-menopausal women (PM). Data are shown as
mean 6 SD. *Significantly different (P , 0.05) from midcycle.
analysed in this manner. The complexity of LH was negatively
correlated with serum bioactive and immunoactive LH concentration (r 5 0.529, P 5 0.017; r 5 –0.509, P 5 0.022,
respectively).
Discussion
We have investigated the changes in the physico–chemical
characteristics of both FSH and LH at different stages of the
normal human menstrual cycle and examined the relationships
between charge, carbohydrate complexity and the concentrations of circulating hormones. Ideally such a study is best
performed with a larger subject group, sampled on each day
of the menstrual cycle. Nonetheless, in some respects our
study is more comprehensive than previous investigations
(Padmanabhan et al., 1988; Papandrèou et al., 1993; Wide and
Bakos 1993; Zambrano et al., 1995; Creus et al., 1996) in
that changes in both charge characteristics and glycoform
complexity of the two gonadotrophins have been investigated
concurrently for all cycle stages within plasma samples from
the same subject. Finally we compared the glycoforms of
normal women during different phases of their cycle with
those from women during the post-menopausal state.
The complexity of the glycoforms was dependent upon the
physiological state (i.e. cycle stage and cessation of cyclicity).
Essentially, FSH and LH glycoforms were less complex at
636
midcycle than during other stages of the menstrual cycle. Least
complex and more simple glycoforms were observed following
the menopause than were present during midcycle. In contrast
to LH, where there was no cycle stage-related change in
isoform distribution, glycoforms of FSH became less acidic at
midcycle compared with the other cycle stages. The finding
that the B:I ratio for LH increased at midcycle was indicative
of changes in the quality of molecular species circulating
at this time. However, no inferences regarding changes in
biopotency can be made from this type of assessment due to
the likelihood of differential recognition of glycoforms, i.e.
the ‘I’ component is not constant (Robertson et al., 1997). Postmenopausal gonadotrophin glycoforms were of a relatively
‘simple’ yet very acidic nature.
Analysis of relationships between the varying structural
characteristics and other parameters, e.g. circulating gonadotrophin concentrations, revealed some interesting correlations
between structure and ovarian activity. Increased serum oestradiol concentrations were associated with less acidic FSH
glycoforms for all cycle stages, but the relationship between
oestradiol and reduced FSH complexity was only apparent
during the follicular to midcycle stages. Complexity of FSH
was negatively correlated with the proportion of less acidic
forms, confirming that the degree of sialylation is dependent
upon the extent of branching (number of terminal residues) of
the carbohydrate component. No relationships were shown
between LH physico–chemical characteristics and physiological function.
Our findings are the first to define FSH and LH complexity
throughout the entire menstrual cycle in addition to following
the menopause. Previous estimates of FSH complexity are
available only for the late follicular phase (Creus et al., 1996)
and the menopause for FSH and LH glycoforms (Papandréou
et al., 1993; Creus et al., 1996). Comparisons of our late
follicular and post-menopausal data with these earlier studies
revealed differences that may be attributed to the methodologies
employed. Clearly the advantage of our study was that the
glycoform distribution and complexity were compared across
cycle stages within the same individual utilizing the same
methodology. While our study showed that midcycle FSH
comprises 63% complex, 36% intermediate glycoforms, Creus’
study conducted during the late follicular phase (day 13)
showed 41% complex and 59% intermediate glycoforms. The
proportion of complex FSH glycoforms in Creus’ study was
intermediate between the late follicular phase (studied on day
11/12) and midcycle phase of our study. It may be argued that
the glycosylation changes observed at midcycle were beginning
to be established during the latter stages of the follicular phase.
The change in the FSH complexity from late follicular (90%)
phase to midcycle (63%) was rapid, occurring over 2–3 days.
Studies into glycoform complexity were complemented by
those of charge heterogeneity, and the finding that FSH
glycoforms became less acidic at midcycle are in agreement
with others (Padmanabhan et al., 1988; Wide and Bakos, 1993;
Zambrano et al., 1995). In this study, the amount of FSH
eluting with pI .4.3 was greatest at midcycle and was also
associated with increased serum oestradiol. In addition to the
proportion of FSH isoforms with pI .4.3, analyses were
Gonadotrophin heterogeneity
performed employing other cut-offs (pH 4.45 and 4.6) and the
observation that FSH glycoforms became less acidic at midcycle was confirmed (data not shown). No shift in LH glycoforms
was found, which was in contrast to the finding of Wide and
Bakos (1993) who studied median charge. However, the
charged LH glycoform profiles varied considerably from subject to subject and an increased population size may have
revealed more conclusive differences in this investigation. The
acidic nature of menopausal gonadotrophins was in agreement
with Wide and Bakos (1993) and Beitins and Padmanabhan
(1991).
The information obtained with regard to changes in glycoform distribution and complexity thus provides important clues
regarding the glycosylation status of the gonadotrophins. The
reduction in acidic, complex glycoforms at midcycle are most
probably related to increased GnRH pulse frequency and/or
oestradiol concentrations that characterize this phase of the
cycle, although the exact involvement of these endocrine
factors remains to be resolved (Dharmesh and Baenziger, 1993;
Pérez and Apfelbaum, 1996). In the present investigation
increased FSH and LH concentrations at midcycle were correlated strongly with reduced complexity (FSH and LH) and
acidity (FSH alone) and one may postulate that the increased
GnRH pulsatility at this time (Sollenberger et al., 1990) is
having an effect on the degree of glycosylation as well as the
rate of synthesis. Alternatively, the midcycle forms may be
the result of increased rates of synthesis and secretion such
that the molecules released contain less mature carbohydrate
structures. In the luteal phase, high progesterone concentrations
may counteract the effects of oestradiol affecting the glycosylation pattern – in oestradiol-treated post-menopausal women
progesterone administration can result in the short term production of acidic forms (Wide et al., 1995, 1996). Studies of
Zambrano et al. (1995) show that the changes in pI of
circulating FSH are attributable to secretory alterations and
not to structural modifications. The exact proportion of gonadotrophin glycoforms with short half life, i.e. least sialylated
and most basic, identified in any such study may be an
underestimate as a considerable amount may have been lost
through metabolic clearance prior to sampling.
While it is established that the in-vivo half life is affected
by sialic acid content (Blum and Gupta, 1985; de Leeuw et al.,
1996), whether glycoforms with reduced metabolic clearance
have greater in-vivo biological activity is still a matter of
controversy (see Chappel, 1995; Ulloa-Aguirre et al., 1995).
There is some evidence to support this premise (Galway et al.,
1990; Fares et al., 1992). For instance, although sialic acid
dictates the rate of clearance for normal isospecies, the invivo biological activity of LH glycoforms appears to be more
dependent on their interaction with their target cells than the
in-vivo half life (Burgon et al., 1996).
The less acidic glycoforms that predominate during the
midcycle may have greater in-vitro biological activity according to B:I ratio estimates of the various glycoforms (CerpaPoljak et al., 1993; Flack et al., 1994; Lambert et al.,
1995). Increasing evidence for variable immunoreactivity of
glycoforms (Burgon et al. 1993; Simoni et al. 1994; Chappel,
1995) casts doubt on such B:I estimates. However the recent
investigations of Burgon et al. (1996) and Stanton et al. (1996),
working on a molar basis, have demonstrated that glycoforms
of FSH and LH do indeed display sialic acid content-related
variations in in-vitro and in-vivo biological activity in addition
to acute in-vivo clearance and immunological activity. Whether
different glycoforms have different functions, e.g. steroidogenesis versus growth, or enhanced activity for a particular
function, is another matter of contention. Some evidence has
been provided for FSH (Padmanabhan, 1995), LH (Nomura
et al., 1988, 1989; Mitchell et al., 1994) and throid stimulating
hormone (TSH) (Sergi et al., 1991; Pickles et al., 1992; Medri
et al., 1994). Furthermore, artificially-induced glycosylation/
deglycosylation patterns of FSH may confer different intrinsic
properties with respect to the signal transduced following
receptor activation (Arey et al., 1997). As with preferential
bioactivity, more detailed investigations need to be performed
before such hypotheses can be established unequivocally.
The possibility of different glycoforms having differential
or preferential biological activities may have implications for
the changing gonadotrophin glycoform profiles observed during
the menstrual cycle in this and other studies. Such observations
are not limited to FSH and LH in the context of the menstrual
cycle and changes in the glycosylation pattern of human
chorionic gonadotrophin (HCG) through pregnancy have also
been reported (Diaz-Cueto et al., 1996). It has been suggested
that each FSH or LH glycoform type may have a specific role
to play in the recruitment and development of the dominant
follicle and then in ovulation (see Lambert et al., 1995).
Should this be proven, it is possible to produce ‘designer’
recombinant gonadotrophins with more physiological characteristics for use within individualized assisted reproduction
treatments. For example protein products can be synthesized
with longer half life or different/enhanced biological activity
due to changes in glycosylation or overall protein structure
(Fares et al., 1992; La Polt et al., 1992; Sugahara et al., 1995,
1996; Benmenahem and Boime, 1996).
In summary, we have shown that, during the normal menstrual cycle, the glycoforms of FSH and LH become less
complex (more simple) at midcycle compared with the follicular and luteal phases; the charge of FSH glycoforms also
changed at midcycle such that fewer acidic forms were present.
Following the menopause the gonadotrophins are of a very
acidic, ‘simple’ nature. The changes in glycoform characteristics appear to be related to the prevailing oestrogen milieu.
Changes in the carbohydrate complexity and the charge of the
gonadotrophins, as regulated by the neuroendocrine milieu
throughout the menstrual cycle, may therefore provide a
secondary level of control by which gonadal function is
regulated.
Acknowledgements
This study was funded by Organon Laboratories Ltd and Salford
Royal Hospitals NHS Trust. The authors also wish to thank Dr Andy
Pickersgill and Sister Helen Jones for volunteer recruitment and
sample collection.
References
Arey, B.J., Stevis, P.E., Deecher, D.C. et al. (1997) Induction of promiscuous
G-protein coupling of the follicle stimulating hormone (FSH) receptor: a
637
C.J.Anobile et al.
novel mechanism for transducing pleiotropic actions of FSH isoforms. Mol.
Endocrinol., 11, 517–526.
Beitins, I.Z. and Padmanabhan, V. (1991) Bioactivity of gonadotrophins. In
Styne, D. (ed.), Endocrinology and Metabolism Clinics and North America.
Vol. 20. W.B.Saunders Co., Philadelphia, USA, pp. 85–120.
Benmenahem, D. and Boime, I (1996) Converting heterodimeric gonadotropins
to genetically linked single chains – new approaches to structure–activity
relationships and analog design. Trends Endocrinol. Metab., 7, 100–105.
Blum, W.F.P. and Gupta, D. (1985) Heterogeneity of rat FSH by
chromatofocusing-studies on serum FSH, hormone released in vitro and
metabolic clearance rates of its various forms. J. Endocrinol., 105, 29–37.
Burgon, P.G., Robertson, D.M., Stanton, P.G. and Hearn, M.T.W. (1993)
Immunological activities of highly purified isoforms of human FSH correlate
with in vitro bioactivities. J. Endocrinol., 139, 511–518.
Burgon, P.G., Stanton, P.G. and Robertson, D.M. (1996) In vivo bioactivities
and clearance patterns of highly purified human luteinizing-hormone
isoforms. Endocrinology, 137, 4827–4836.
Cerpa-Poljak, A., Bishop, L.A., Hort, Y.J. et al. (1993) Isoelectric charge of
recombinant human follicle stimulating hormone isoforms determines
receptor affinity and in vitro bioactivity. Endocrinology, 132, 351–356.
Chappel, S.C. (1995) Heterogeneity of follicle stimulating hormone – control
and physiological-function. Hum. Reprod. Update, 1, 479–487.
Creus, S., Pellizzari, E., Cigorraga, S.B. and Campo, S. (1996) FSH isoforms –
bio- and immuno-activities in postmenopausal and normal menstruating
women. Clin. Endocrinol., 44, 181–189.
De Leeuw, R., Mulders, J., Voortman, G. et al. (1996) Structure–function
relationship of recombinant follicle stimulating hormone (Puregon). Mol.
Hum. Reprod., 2, 361–369.
Dharmesh, S.M. and Baenziger, J.U. (1993) Estrogen modulates expression
of the glycosyltransferases that synthesize sulfated oligosaccharides on
lutropin. Proc. Natl. Acad. Sci. USA, 90, 11127–11131.
Diaz-Cueto, L., Barrios de Tomasi, J., Timossi, C. et al. (1996) More in-vitro
bioactive, shorter-lived human chorionic gonadotrophin charge isoforms
increase at the end of the first and during the third trimesters of gestation.
Mol. Hum. Reprod., 2, 643–650.
Fares, F.A., Suganuma, N., Nishimori, K. et al. (1992) Design of a longacting follitropin agonist by fusing the C-terminal sequence of the chorionic
gonadotrophin β subunit to follitropin β subunit. Proc. Natl. Acad. Sci.
USA, 89, 4303–4308.
Flack, M.R., Bennet, A.P., Froehlich, J. et al. (1994) Increased biologicalactivity due to basic isoforms in recombinant human follicle stimulating
hormone produced in a human cell-line. J. Clin. Endocrinol. Metab., 79,
756–760.
Galway, A.B., Hsueh, A.J. W., Keene, J.L. et al. (1990) In vitro and in vivo
bioactivity of recombinant human follicle stimulating hormone and partially
deglycosylated variants secreted by transfected eukaryotic cell lines.
Endocrinology, 127, 93–100.
Green, E.D. and Baenziger, J.U. (1988) Asparagine-linked oligosaccharides
on lutropin, follitropin, and thyrotropin.1. Structural elucidation of the
sulfated and sialylated oligosaccharides on bovine, ovine, and human
pituitary glycoprotein hormones. J. Biol. Chem., 263, 25–35.
Harris, S.D., Anobile, C.J., McLoughlin, J.D. et al. (1996) Internal carbohydrate
complexity of the oligosaccharide chains of recombinant human follicle
stimulating hormone (Puregon, Org 32489): a comparison with Metrodin
and Metrodin-HP. Mol. Hum. Reprod., 2, 807–811.
Lambert, A., Rodgers, M., Mitchell, R. et al. (1995) In vitro biopotency and
glycoform distribution of recombinant human follicle stimulating hormone
(Org-32489), Metrodin and Metrodin-HP. Hum. Reprod., 10, 1928–1935.
La Polt, P.S., Nishimori, K., Fares, F.A. et al. (1992) Enhanced stimulation
of follicle maturation and ovulatory potential by long acting follicle
stimulating hormone agonists with extended carboxyl terminal peptides.
Endocrinology, 131, 2514–2520.
Medri, G., Sergi, I., Papandréou, M.-J. et al. (1994) Dual activity of human
pituitary thyrotrophin isoforms on thyroid cell growth. J. Mol. Endocrinol.,
13, 187–198.
Mitchell, R., Bauerfeld, C., Schaefer, F. et al. (1994) Less acidic forms of
luteinizing-hormone are associated with lower testosterone secretion in
men on haemodialysis treatment. Clin. Endocrinol., 41, 65–73.
Nomura, K., Tsunasawa, S., Ohmura, K. et al. (1988) Renotropic activity of
ovine luteinizing hormone isoform(s). Endocrinology, 123, 700–712.
Nomura, K., Ohmura, K., Nakamura, Y. et al. (1989) Porcine luteinizing
hormone isoform(s): Relationship between their molecular structures and
renotropic versus gonadotropic activities. Endocrinology, 124, 712–719.
Padmanabhan, V. (1995) Neuroendocrine control and physiologic relevance
of FSH heterogeneity. [Abstr. no. S3] J. Reprod. Fertil., 15,
638
Padmanabhan, V., Lang, L.L., Sonstein, J. et al. (1988) Modulation of
serum follicle stimulating hormone bioactivity and isoform distribution by
estrogenic steroids in normal women and in gonadal dysgenesis. J. Clin.
Endocrinol. Metab., 67, 465–473.
Papandréou, M.-J., Asteria, C., Pettersson, K. et al. (1993) Concanavalin A
affinity chromatography of human serum gonadotropins: evidence for
changes of carbohydrate structure in different clinical conditions. J. Clin.
Endocrinol. Metab., 76, 1008–1013.
Perez, G.T. and Apfelbaum, M.E. (1996) GnRH-stimulated glycosylation
(proximal and distal) of luteinizing hormone by cultured rat pituitary cells.
Neuroendocrinology, 64, 456–461.
Pickles, A.J., Peers, N., Robertson, W.R. and Lambert, A. (1992) Different
isoforms of human pituitary thyroid-stimulating hormone have different
relative biological activities. J. Mol. Endocrinol., 9, 251–256.
Robertson, W.R. and Bidey, S.P. (1990) The in vitro bioassay of peptide
hormones. In Siddle, K. and Hutton, J. (eds), Peptide Hormone Secretion.
Oxford University Press, Oxford, UK, pp. 150–154.
Robertson, W.R. (1997) Gonadotrophin isoform patterns in different
pharmaceutical preparations. In Kahn, J. (ed.), Gonadotrophin Isoforms –
Facts and Future. Serono Fertility Series 2. pp. 53–60.
Robertson, W.R., Lambert, A., Mitchell, R. and Talbot, J.A. (1997) The
interpretation of immuno- and bioassays of human FSH. In Fauser, B.C.J.M.
(ed.), FSH Action and Intraovarian Regulation. Studies in Profertility
Series 6. Parthenon Publishing Group, Carnforth, UK, pp. 33–50.
Sergi, I., Papandréou, M.-J., Medri, G. et al. (1991) Immunoreactive and
bioactive isoforms of human thyrotrophin. Endocrinology, 128, 3259–3268.
Simoni, M., Jockenhovel, F. and Nieschlag, E. (1994) Polymorphism of human
pituitary FSH – analysis of immunoreactivity and in vitro bioactivity of
different molecular-species. J. Endocrinol., 141, 359–367.
Sollenberger, M.J., Carlsen, E.C., Johnson, M.L. et al. (1990) Specific
physiological regulation of luteinising hormone secretory events throughout
the menstrual cycle: new insights into the pulsatile mode of gonadotropin
release. J. Neuroendocrinol., 2, 845–852.
Stanton, P.G., Burgon, P.G., Hearn, M.T.W. and Robertson, D.M. (1996)
Structural and functional characterisation of hFSH and hLH isoforms. Mol.
Cell. Endocrinol., 125, 133–141.
Stockell Hartree, A. and Renwick, A.G. (1992) Molecular structures of
glycoprotein hormones and functions of their carbohydrate components.
Biochem. J., 287, 665–679.
Sugahara, T., Pixley, M.R., Minami, S. et al. (1995) Biosynthesis of a
biologically active single peptide chain containing the human common α
and chorionic gonadotrophin β subunits in tandem. Proc. Natl. Acad. Sci.
USA, 92, 2041–2045.
Sugahara, T., Sato, A., Kudo, M. et al. (1996) Expression of biologically
active fusion genes encoding the common alpha subunit and the follicle
stimulating hormone beta subunit - role of a linker sequence. J. Biol.
Chem., 271, 10445–10448.
Talbot, J.A., Rodger, R.S. and Robertson, W.R. (1990a) Pulsatile bioactive
luteinising hormone secretion in men with chronic renal failure and
following renal transplantation. Nephron, 56, 66–72.
Talbot, J.A., Rodger, R.S.C., Shalet, S.M. et al. (1990b) The pulsatile secretion
of bioactive luteinising hormone in normal adult men. Acta Endocrinol.,
122, 643–650.
Ulloa-Aguirre, A., Espinoza, R., Damianmatsumura, P. and Chappel, S.C.
(1988a) Immunological and biological potencies of the different molecular
species of gonadotropins. Hum. Reprod., 3, 491–501.
Ulloa-Aguirre, A., Espinoza, R., Damián-Matsumura, P. et al. (1988b) Studies
on the microheterogeneity of anterior pituitary follicle-stimulating hormone
in the female rat. Isoelectric focusing pattern throughout the estrous cycle.
Biol. Reprod., 38, 70–78.
Ulloa-Aguirre, A., Damián-Matsumura, P., Jiménez, M. et al. (1992) Biological
characterization of the isoforms of urinary human follicle-stimulating
hormone contained in a purified commercial preparation. Hum. Reprod., 7,
1371–1378.
Ulloa-Aguirre, A., Midgley, Jr., A.R., Beitins, I.Z. and Padmanabhan, V.(1995)
Follicle stimulating isohormones: characterization and physiological
relevance. Endocr. Rev., 16, 765–787.
Van Damme, M.P., Robertson, D.M. and Diczfalusy, E. (1974) An improved
in vitro method for measuring luteinising hormone (LH) activity using
mouse Leydig cell preparations. Acta Endocrinol., 77, 665–671.
Wide, L. and Bakos, O. (1993) More basic forms of both human folliclestimulating hormone and luteinizing hormone in serum at midcycle
compared with the follicular or luteal phase. J. Clin. Endocrinol. Metab.,
76, 885–889.
Gonadotrophin heterogeneity
Wide, L., Naessen, T., Eriksson, K. and Rune, C. (1996) Time-related effects
of a progestogen on the isoforms of serum gonadotropins in 17-β-estradiol
treated postmenopausal women. Clin. Endocrinol., 44, 651–658.
Wide, L. and Naessén, T. (1994) 17β-oestradiol counteracts the formation of
the more acidic isoforms of follicle-stimulating hormone and luteinizing
hormone after menopause. Clin. Endocrinol., 40, 783–789.
Wide, L., Naessén, T. and Phillips, D.J. (1995) Effect of chronic daily oral
administration of 17β-oestradiol and norethisterone on the isoforms of
serum gonadotrophins in post-menopausal women. Clin. Endocrinol., 42,
59–64.
Zambrano, E., Olivares, A., Mendez, J.P. et al. (1995) Dynamics of basal
and gonadotropin-releasing hormone-releasable serum follicle-stimulatinghormone charge isoform distribution throughout the human menstrual cycle.
J. Clin. Endocrinol. Metab., 80, 1647–1656.
Received on October 16, 1997; accepted on April 28, 1998
639