Download Use of Site-Specific Antibodies to Characterize the Circulating Form

0013-7227/02/$15.00/0
Printed in U.S.A.
The Journal of Clinical Endocrinology & Metabolism 87(8):3867–3870
Copyright © 2002 by The Endocrine Society
Use of Site-Specific Antibodies to Characterize the
Circulating Form of Big Insulin-Like Growth Factor II in
Patients with Hepatitis C-Associated Osteosclerosis
SUNDEEP KHOSLA, F. JOHN BALLARD,
AND
CHERYL A. CONOVER
Endocrine Research Unit, Mayo Clinic and Mayo Foundation (S.K., C.A.C.), Rochester, Minnesota 55905; and GroPep Ltd.
(J.B.), Thebarton, South Australia 5031, Australia
Hepatitis C-associated osteosclerosis (HCAO) is a rare syndrome of adult-onset osteosclerosis. An understanding of the
factor(s) leading to the stimulation of bone formation in these
patients may provide novel anabolic approaches for the treatment of osteoporosis. We have demonstrated that HCAO
patients have a specific increase in circulating big IGF-II
(IGF-IIE) and IGF-binding protein-2 (IGFBP-2) levels, and
that IGF-IIE and IGFBP-2 circulate together in a bioavailable,
50-kDa complex. Patients with nonislet cell tumor hypoglycemia (NICTH) also have increased circulating IGF-IIE and
IGFBP-2 levels. However, HCAO patients do not exhibit
hypoglycemia, nor do NICTH patients exhibit obvious osteosclerosis. Thus, to better understand the reason(s) for the
differing clinical manifestations of the IGF-IIE excess in the
two syndromes, we characterized IGF-IIE in HCAO and
NICTH sera using recently developed antibodies (Ab) recog-
H
EPATITIS C-associated osteosclerosis (HCAO) is a rare
syndrome of adult-onset osteosclerosis (1–7). These
patients have remarkable increases in their bone mass consequent to a marked stimulation of bone formation. Despite
its rare occurrence, HCAO is of great interest, because an
understanding of the factor(s) leading to the stimulation of
bone formation in these patients may provide novel approaches for the treatment of osteoporosis.
We have previously demonstrated that HCAO patients
have a specific increase in circulating big IGF-II (IGF-IIE) and
IGFBP-2 levels, and that IGF-IIE and IGFBP-2 circulate together in a bioavailable, approximately 50-kDa complex (8).
On the basis of these findings and in vitro studies demonstrating high avidity of the IGF-II/IGFBP-2 complex for
human osteoblast extracellular matrix, we postulated that
IGFBP-2 was targeting IGF-IIE to the skeleton in these
patients, resulting in the stimulation of bone formation (8).
Similar to HCAO patients, patients with nonislet cell tumor hypoglycemia (NICTH) have increased circulating IGFIIE and IGFBP-2 levels (9). However, HCAO patients do not
exhibit hypoglycemia, nor have NICTH patients been reported to have osteosclerosis. As IGF-II is synthesized as a
156 amino acid precursor with potential cleavage sites after
amino acid 67 (leading to mature IGF-II) as well as after
amino acids 88 and 104 (leading to E-domain extensions of
21 and 37 amino acids, respectively) (10) (Fig. 1), we postuAbbreviations: Ab, Antibodies; HCAO, hepatitis C-associated osteosclerosis; IGF-IIE, big IGF-II; IGFBP-2, IGF-binding protein-2; NICTH,
nonislet cell tumor hypoglycemia; PBS-T, PBS and 0.1% Tween 20.
nizing either the full-length IGF-IIE 89-amino acid C-terminal
extension peptide (IIE138 –156 Ab) or specific cleavage forms of
IGF-IIE (IIE78 – 88 Ab and IIE89 –101 Ab). The predominant IGFIIE form in HCAO serum migrated on SDS-PAGE as a single
band at approximately 18 kDa that reacted with the IIE89 –101
Ab. On the other hand, the predominant form in NICTH serum
migrated as a doublet of 14 and 16 kDa that reacted with the
IIE78 – 88 Ab. There results are consistent with differential processing of the IGF-IIE precursor at predicted cleavage sites
producing IGF-IIE1–104 and IGF-IIE1– 88 in HCAO and NICTH,
respectively. As these two forms may have differing biological
activities and/or targeting properties, our findings may explain at least in part the different manifestations of IGF-IIE
overproduction in the two syndromes. (J Clin Endocrinol
Metab 87: 3867–3870, 2002)
lated that the differing clinical manifestations of IGF-IIE excess in HCAO vs. NICTH may be due at least in part to
differential processing of the IGF-II prohormone in the two
syndromes. To directly test this hypothesis, we generated
specific antisera directed against the 78 – 88, 89 –101, or 138 –
156 regions of the E domain of IGF-IIE and examined HCAO
vs. NICTH serum for immunoreactivity against each of these
antisera. Our results indicate that the predominant circulating forms of IGF-IIE in HCAO and NICTH are clearly different, perhaps accounting for the development of osteosclerosis in one syndrome and hypoglycemia in the other.
Materials and Methods
Generation of IGF-IIE region-specific Ab
Polyclonal Ab recognizing either the full-length IGF-IIE 89-amino
acid C-terminal extension peptide [IIE138 –156 antibody (Ab)] or specific
cleavage forms of IGF-IIE (IIE78 – 88 Ab and IIE89 –101 Ab; Fig. 1) were
developed. Synthetic peptides corresponding to regions of the E domain
of IGF-II were synthesized and conjugated to diphtheria toxin using a
maleimidocapryol-N-hydroxysuccinimide linker (Chiron Corp., Clayton, Australia). The sequences of these peptides are as follows: IGFIIE78 – 88, NH2-Pro-Asp-Asn-Phe-Pro-Arg-Tyr-Pro-Val-Gly-Lys-COOH;
IGF-IIE89 –101, NH2-Phe-Phe-Gln-Tyr-Asp-Thr-Trp-Lys-Gln-Ser-Thr-GlnArg-COOH; and IGF-IIE138 –156, NH2-Phe-Thr-Gln-Asp-Pro-Ala-HisGly-Gly-Ala-Pro-Pro-Glu-Met-Ala-Ser-Asn-Arg-Lys-COOH. New Zealand White rabbits were immunized with the above peptide conjugates
by repeated sc injection with Freund’s adjuvant, and immune sera were
collected according to standard techniques (11).
The reactivity of the IGF-IIE antisera with unconjugated peptides and
also recombinant IGF-II1– 67 and IGF-IIE1–156 (GroPep Ltd., Adelaide,
Australia) was tested by enzyme immunoassay. Maxisorp 96 plates
(Nunc, Naperville, IL) were coated with 50 ng peptide in PBS and then
3867
3868
J Clin Endocrinol Metab, August 2002, 87(8):3867–3870
Khosla et al. • IGF-IIE in HCAO
but did react with a recombinant IGF-IIE corresponding to
the full-length 1–156 amino acid IGF-II precursor. As expected, the three antisera were also reactive with their cognate peptides within the E domain, but did not react with the
other peptides. The ability of the three antisera to react with
IGF-IIE was also confirmed by Western blot analysis with the
mature 67-amino acid IGF-II, a 104-amino acid form of IGF-IIE,
and a 156-amino acid form (Fig. 2). It is evident that the IGF-IIE
preparations visualized with Ponceau Red staining are not single bands, but migrate as several forms, presumably due to
TABLE 1. Specificity of IGF-IIE antisera as determined by
enzyme immunoassay
FIG. 1. Potential IGF-IIE cleavage sites and Ab used for characterization of HCAO and NICTH sera.
washed with PBS and 0.1% Tween 20 (PBS-T). The wells were then
blocked with PBS-T and 0.5% BSA for 60 min at 37 C, washed twice with
PBS-T, and incubated with primary Ab at 1:1000 in PBS-T for 120 min
at 37 C. The wells were then washed with PBS-T, incubated with antirabbit horseradish peroxidase-linked secondary Ab in PBS-T, and finally
washed with PBS-T. The secondary reagent was then detected with 1
mg/ml O-phenylene-diamine chloride and 0.1% hydrogen peroxide in
a citrate-phosphate buffer at pH 5.5, the reaction was terminated with
1 m H2SO4, and the OD490 was determined.
IIE78 – 88 Ab
IIE89 –101 Ab
IIE138 –156 Ab
IGF-II
(AA 1– 67)
IGF-IIE
(AA 1–156)
AA
78 – 88
AA
89 –101
AA
138 –156
⫺
⫺
⫺
⫹
⫹
⫹
⫹
⫺
⫺
⫺
⫹
⫺
⫺
⫺
⫹
⫹ and ⫺ denote positive and negative reactions, respectively.
Immunoblotting of IGF-II, IGF-IIE1–104, and IGF-IIE1–156
Three-microgram samples of each growth factor were run on 7.5–20%
SDS-PAGE under reducing conditions (100 mm dithiothreitol), transferred to nitrocellulose, and then stained with Ponceau Red. The growth
factors used in these studies were produced using recombinant DNA
techniques (GroPep Ltd.). The membranes were then cut into thirds to
match the loading protocol (lanes 1–3, 4 – 6, and 7–9) before being developed with one of the three primary Ab. After blocking with either
nonfat dry milk or BSA in Tris-buffered saline/0.1% Tween 20, the
membranes were incubated overnight at 4 C in primary Ab, washed, and
incubated in secondary Ab (goat antirabbit) for 1 h at room temperature.
After subsequent washing, they were visualized by an enhanced chemiluminescent detection system (Amersham Biotech, Piscataway, NJ).
Immunoblotting of serum
After informed consent and approval of the local institutional review
board were granted, serum was obtained from normal healthy subjects,
two patients with HCAO who have been previously described (9), and
patients with NICTH. The HCAO sera were obtained at a time when the
patients still had elevated bone formation rates, as described in our
previous publication (5). The NICTH sera were all obtained while the
subjects still had their tumors. NICTH serum was provided by Dr.
Raymond Hintz (Stanford University, Stanford, CA; one sample) and Dr.
Naomi Hizuka (Tokyo Women’s Medical College, Tokyo, Japan; six
samples). For the immunoblotting studies, a 1-␮l serum sample was run
on 7.5–20% SDS-PAGE under reducing conditions (100 mm dithiothreitol), transferred to nitrocellulose, and blocked with either nonfat dry
milk or BSA in Tris-buffered saline/0.1% Tween 20. Membranes were
incubated in primary and secondary Ab, and visualization systems were
as described above.
Results
Immune sera directed against each of the epitopes in the
E domain of IGF-IIE were generated as described in Materials
and Methods. The specificities of these antisera were determined by enzyme immunoassay and are summarized in
Table 1. The three antisera did not react with mature IGF-II,
FIG. 2. SDS-PAGE of IGF-II (lanes 1, 4, and 7), IGF-IIE1–104 (lanes
2, 5, and 8), and IGF-IIE1–156 (lanes 3, 6, and 9) first stained with
Ponceau Red (A) and then immunoblotted with the three site-specific
Ab shown in Fig. 1 (B).
Khosla et al. • IGF-IIE in HCAO
partial glycosylation (Fig. 2A). As expected, subsequent Western blotting with the different Ab (Fig. 2B) demonstrated that
the IIE78 – 88 and IIE89 –101 antisera detected both of the IGF-IIE
forms, but not mature IGF-II. The IIE138 –156 antiserum only
detected IGF-IIE1–156, again as expected.
Initial studies comparing normal serum to HCAO and
NICTH serum demonstrated that whereas all three samples
had approximately equal amounts of mature IGF-II based on
immunoblotting with an IGF-II-specific Ab (RDI, Flanders,
NJ), there were marked differences in the profiles of big
IGF-II in these sera (Fig. 3). Thus, the IIE78 – 88 Ab detected
strong bands in the NICTH serum at approximately 14 –16
kDa. A weak signal corresponding to the lower of the two
bands in NICTH serum was also present in normal and
HCAO sera. In addition, the IIE78 – 88 Ab weakly detected a
higher molecular mass band at approximately 18 kDa in
HCAO serum. This higher molecular mass band in HCAO
serum reacted strongly with the IIE89 –101 Ab. The IIE138 –156
Ab detected a weak band at an even higher molecular mass
in HCAO serum as well as some lower molecular mass bands
in all three sera, which may represent nonspecific binding by
this Ab.
Figure 4 shows data from six additional NICTH sera and
the HCAO serum used in Fig. 3 as well as an additional
HCAO sample using the IIE78 – 88 Ab. As shown, there were
multiple bands in the NICTH samples at 14 –16 kDa. These
were present to a much lesser extent in normal or HCAO
serum, although the latter again had the higher molecular
mass form of approximately 18 kDa.
Further characterization of HCAO and NICTH serum using IIE89 –101 Ab is shown in Fig. 5. Again, HCAO serum
reacted strongly with this Ab, giving a band of approximately 18 kDa. This was not present to any significant extent
in NICTH serum.
Discussion
We report the generation and characterization of Ab recognizing various cleavage forms of IGF-IIE and the use of
these Ab in defining possible differences in the circulating
forms of IGF-IIE in HCAO and NICTH. Based on the known
FIG. 3. Comparison of normal, HCAO, and NICTH sera immunoblotted with the three site-specific IGF-IIE Ab shown in Fig. 1. Numbers indicate the positions of the respective molecular mass markers
in kilodaltons.
J Clin Endocrinol Metab, August 2002, 87(8):3867–3870 3869
FIG. 4. Additional NICTH and HCAO sera immunoblotted with the
IIE78 – 88 Ab. Numbers indicate the positions of the respective molecular mass markers in kilodaltons.
FIG. 5. Further characterization of HCAO and NICTH sera using the
IIE89 –101 Ab. Numbers indicate the positions of the respective molecular mass markers in kilodaltons.
cleavage sites for IGF-IIE (10), these Ab are capable of specifically recognizing either IGF-IIE1– 88, IGF-IIE1–104, or fulllength IGF-IIE1–156.
Our characterization of HCAO vs. NICTH serum demonstrates that the predominant form of IGF-IIE in HCAO serum
migrated on SDS-PAGE as a single band at approximately 18
kDa and reacted with the IIE89 –101 Ab. On the other hand, the
predominant form in NICTH serum migrated as a doublet of
approximately 14 and 16 kDa and reacted with the IIE78 – 88
Ab. These results are consistent with differential processing
of the IGF-IIE precursor at predicted cleavage sites producing IGF-IIE1–104 and IGF-IIE1– 88 in HCAO and NICTH, respectively. Our findings using these specific Ab are consistent with previous studies attempting to identify the form of
IGF-IIE that is elevated in patients with NICTH. Thus, Hizuka et al. (12) performed Western analysis of serum from 10
patients with NICTH using an anti-IGF-II Ab and found that
most of the IGF-II immunoreactivity migrated between 11
and 18 kDa. This was reduced in size to approximately 9.5
kDa after neuraminidase and O-glycosidase digestion, consistent with the predominant circulating form of IGF-IIE in
NICTH sera being IGF-IIE1– 88 (12). We also found that the
IIE78 – 88 Ab detected multiple bands in NICTH sera, presumably corresponding to variable glycosylation of the IGF-IIE.
Finally, the IIE138 –156 Ab failed to recognize a specific band
in either HCAO or NICTH sera, indicating that intact IGFIIE1–156 is not present in any significant amount in the circulation in either of these conditions. We did attempt to
deglycosylate the IGF-IIE in HCAO serum using neuraminidase and O-glycanase, but were unable to show any clear
changes in the size of the observed band (data not shown).
This would suggest that either the IGF-IIE in HCAO is not
glycosylated (which appears to be an unlikely possibility), or
3870 J Clin Endocrinol Metab, August 2002, 87(8):3867–3870
that for some reason this peptide in HCAO serum is extremely resistant to standard deglycosylation methods.
A clear limitation of this study is the relatively small number of HCAO (n ⫽ 2) and NICTH (n ⫽ 7) subjects analyzed,
because both are rare syndromes, and samples are difficult
to obtain. Nonetheless, as noted earlier, our findings in the
NICTH patients are entirely consistent with the apparent size
of IGF-IIE previously identified in these patients (12) attesting (to some degree) to the robustness of the data. In addition,
the major finding of our study has to do with qualitative
differences (i.e. clearly different patterns of immunoreactivity with site-specific Ab) in IGF-IIE in HCAO vs. NICTH,
rather than quantitative differences in IGF-IIE levels in the
two syndromes, making issues of sample size somewhat less
of a concern. Nonetheless, we recognize this limitation, and
additional studies characterizing the IGF-IIE peptide in a
larger number of HCAO and NICTH sera are clearly warranted to further validate our findings.
As the IGF-IIE1–104 and IGF-IIE1– 88 isoforms may have
differing biological activities and/or targeting properties,
our findings may explain at least in part the different clinical
manifestations of IGF-IIE overproduction in HCAO vs.
NICTH. In particular, a major abnormality in NICTH, due
perhaps directly to the excess IGF-IIE in the serum of these
patients, is the disruption of the approximately 150-kDa
ternary complex with a consequent shift of the IGFs and
IGFBP-3 to the approximately 50-kDa binary complex and an
increase in free IGFs, with resultant hypoglycemia (9). By
contrast, we have previously shown that the ternary complex
remains intact in HCAO serum, and as such, there is no
increase in free IGFs in the circulation of these patients (8),
probably explaining the absence of hypoglycemia in HCAO
patients. Whether the disruption of the ternary complex in
NICTH and its preservation in HCAO are due to the different
forms of IGF-IIE present in the circulation of these patients
is an issue that requires further study.
We have also previously found that the IGF-IIE in HCAO
patients circulates bound to IGFBP-2 in the approximately
50-kDa binary fraction (8). Moreover, our previous in vitro
studies indicate that upon binding IGF-II, IGFBP-2 has
greatly enhanced avidity for the osteoblast extracellular matrix (8). Thus, we have postulated that the IGF-IIE/IGFBP-2
complex accumulates in bone in HCAO patients, and the
subsequent local release of IGF-IIE in the bone microenvironment results in the stimulation of bone formation
and osteosclerosis observed in these patients. Further
studies more directly testing this hypothesis are currently
underway. Indeed, based on the findings of the present
study, IGF-IIE1–104 appears to be the IGF-IIE fragment that
should be examined in combination with IGFBP-2 in these
Khosla et al. • IGF-IIE in HCAO
studies, because this the IGF-IIE isoform that is elevated
in vivo in the HCAO patients.
In summary, we describe the generation and characterization of Ab specific for various potential cleavage forms of
IGF-IIE. These Ab have provided powerful tools to characterize the forms of IGF-IIE that are elevated in HCAO vs.
NICTH serum. The insights from these studies may help
both in explaining the differing clinical manifestations of
the IGF-IIE excess in the two syndromes as well as in
selecting the optimal form of IGF-IIE to use in future
studies aimed at developing novel anabolic approaches to
treating osteoporosis.
Acknowledgments
We thank Laurie Bale, Phillip Elliott, and Leanne McGrath for technical assistance. We also thank Drs. Raymond Hintz and Naomi Hizuka
for providing NICTH sera.
Received November 28, 2001. Accepted May 9, 2002.
Address all correspondence and requests for reprints to: Sundeep
Khosla, M.D., Mayo Clinic, 200 First Street SW, 5-194 Joseph, Rochester,
Minnesota 55905. E-mail: [email protected].
This work was supported by NIA Grant AG-04875. This paper was
presented in part at the 82nd Annual Meeting of The Endocrine Society,
Toronto, Canada, 2000.
References
1. Villareal DT, Murphy WA, Teitelbaum SL, Arens MQ, Whyte MP 1992
Painful diffuse osteosclerosis after intravenous drug abuse. Am J Med
93:371–381
2. Beyer SH, Parfitt AM, Shih MS, Anderson Q, Heath III H 1990 Idiopathic
acquired diffuse osteosclerosis in a young woman. J Bone Miner Res 5:1257–
1263
3. Whyte MP, Teitelbaum SL, Reinus WR 1996 Doubling skeletal mass during
adult life: the syndrome of diffuse osteosclerosis after intravenous drug abuse.
J Bone Miner Res 11:554 –558
4. Whyte MP, Reasner CA 1997 Hepatitis C-associated osteosclerosis after blood
transfusion. Am J Med 102:219 –220
5. Hassoun A, Nippoldt TB, Tiegs RD, Khosla S 1997 Hepatitis C-associated
osteosclerosis: an unusual syndrome of acquired osteosclerosis in adults. Am J
Med 103:70 –73
6. Diamond T, Depczynski B 1996 Acquired osteosclerosis associated with intravenous drug use and hepatitis C infection. Bone 19:679 – 683
7. Shaker JL, Reinus WR, Whyte MP 1998 Hepatitis C-associated osteosclerosis:
late onset after blood transfusion in an elderly woman. J Clin Endocrinol Metab
83:93–98
8. Khosla S, Hassoun AAK, Baker B, Liu F, Zein NN, Whyte MP, Reasner CA,
Nippoldt TB, Tiegs RD, Hintz RL, Conover CA 1998 Insulin-like growth
factor system abnormalities in hepatitis C-associated osteosclerosis: potential
insights into increasing bone mass in adults. J Clin Invest 101:2165–2173
9. LeRoith D, Clemmons D, Nissley P, Rechler MM 1992 Insulin-like growth
factors in health and disease. Ann Intern Med 116:854 – 862
10. Duguay SJ 1999 Post-translational processing of insulin-like growth factors.
Horm Metab Res 31:43– 49
11. Harlow E and Lane D, eds 1988 Antibodies: a laboratory manual. Cold Spring
Harbor: Cold Spring Harbor Laboratory
12. Hizuka N, Fukuda I, Takano K, Asakawa-Yasumoto K, Okubo Y, Demura
H 1998 Serum high molecular weight form of insulin-like growth factor II from
patients with non-islet cell tumor hypoglycemia is O-glycosylated. J Clin
Endocrinol Metab 83:2875–2877