- Email: [email protected]
Selection of high-affinity and high-specificity monoclonal antibodies for 1α,25-dihydroxyvitamin D
SELECTION OF HIGH-AFFINITY AND HIGH-SPECIFICITY MONOCLONALANTIBODIES FOR la,25-DIHYDROXYVITAMIN D E. Barbara Mawer, Jacqueline 1. Berry, Juana Bessone, Shraga Shany, h University of Manchester Medical School, Department of Medicine, Manchester Ml3 9PT, U.K. and *Department of Chemical Pathology, Hope Hospital, Salford M6 8HD, U.K. Received (Revised
7-31-85 10-21-85)
ABSTRACT
The preparation of high-affinity and high -specificity monoclonal antibodies to Ia,2&dihydroxyvitamin D is described. Monoclonal antibodies were derived from Balb-c mice imnunised with either la-hydroxy-25,26,27acid or trinor-24-cholecalcioic with la-hydroxy-26,27-dinor-cholemouse calciferol-25-oxime, and spleen cells were hybridised with myeloma cells. From six fusions nine monoclonal antibodies (MAb's) were selected from 676 antibody-secreting hybrids. Antibodies varied widely in their ability to bind la,25-dihydroxyvitamin D3 (50% displacement of radioligand ranged from 25 - 900 pg); two had particularly useful characteristics for la,25#Qdroxyvitamin D assay. MAb 5F2 has high affinity (Ka = 1.39 x IO M I and does not discriminate between la,25-dihydroxyvitamin D and D , thus enabling the two forms to be measured together. &b 163 is highly specific, having no crossreactivity with 25-hydroxy-, 24,25-dihydroxyor 25,26-dihydroxyvitamin D at concentrations found in normal human serum; this MAb has the potential to eliminate lengthy extraction procedures the involved in currently available assays for la,25-dihydroxyvitamin D. INTRODUCTION Assays currently vitamin
D,
by the
cross-reacting
food
need
to
of vitamin
August,
present
September
1985
metabolite
(1,25(OH)2D)(1). preparation at
between
in
serum
necessary
higher
to
concentrations, D
An
being
being
1,25(OH)2D3 Steroids
the
from
problem
therapy form.
do not
Volume
remove
derivat-
endogenous
assays
are
(24,25(OHj2D)
additional
derived
of
namely
the la,25-dihydroxylated
former
and the latter in use for
hormonal
(25.26(OHj2D).
D2 and D3, the
antisera
active
24,25-dihydroxyvitamin
D
distinguish
supplementation
polyclonal
sample
D (25(OH)D),
25,26-dihydroxyvitamin
ives
the D
extensive
metabolites
25-hydroxyvitamin
is the
for
la,25-dihydroxyvitamin
complicated
and
available
or Most
cross-react
46, Numbers
2,3
with
binding
assays
varying
equally
other
ing highly
specific
and provides
these methods
for
the
clones
antibodies
to problems
supply
to produce
sensitivity
would
in
cross-react
have the sensitivity
to
or no cross-reactivity
D2 or D3.
The elegant
technique
production
of monoclonal
antibodies
in culture
for
paper we describe
the required
1,25(OH)2D2
1,25(OH)2D
and show little
assay development
In this
MAb's with
(4)
already
in protein-
and presents
and 1.25(OH)2D3,
of vitamin
a constant
improving
(5,6).
the assay of
metabolites
by hybrid
with
is tedious
concentrations,
secreted
(t&b’s)
employed
(2,3)
for
and Milstein
receptors
cross-react
preparation
1,25(OH)2D2
picomolar
of Kohler
thus
and
antibody
with
detect
1,25(OH)2D3
reproducibility
An ideal
with
for
extents,
between-batch
1,25
some of the tissue
1,25(OH)2D2;
the
offers
radioimnunoassay
of antibody and precision. MAb's to
of
of defined
1,25(OH)2D
characteristics,
One of us has exploited
progesterone
the selection and
the means of select-
and testosterone
and characterisation
specificity
for
the
of
assay of
(OH)2D. EXPERIMENTAL
Reaaents Unlabelled 25(OH)D3 was obtained from Roussel, Paris, 25(OH)D2 from Upjohn, Kalamazoo, 24,25(0H) D3 from Roche Products Ltd., 25,26(0H) D from Dr. J. Redel, Par ? s, and 25,26(OH)2D2 from Dr. 0. Willia&s~ Cambridge England. 1,25(OH) D and 1';~;;;;2D3N J ant were kindly suppliid by Hoffman La R&Be Inc., la-hydroxycholecalciferol from Leo Laboratories, Princes Risbo;ough,'U.K. All vitamin D sterols were stored under nitrogen at -20" and concentrations for standard solutions were determined by U.V. spectrophotometry, using the apprgpriate extinction coefficients (X max 26! nm). 1,25(OH) [26,27-methylHID (180 Ci/nanol) was purchased from Amersham Inte 6 nationa1,U.K. la-iydroxy-25,26,26-trinor-24-cholecalcioic
S
TPIIOXDD
743
acid, kindly provided by Dr. Milan Uskokovic and la-hydroxy-26,27-dinorcholecalciferol-25-oxime, given by Dr. L. Fraher, were used to prepare imnunogens. Bovine serum albumin fraction V (BSA) was purchased from Sigma,U.K. The buffer used in the antibody evaluation procedure was 0.07 M sodium phosphate, pH 6.0,and contained 0.1% gelatin, unless specified to the contrary. Dulbecco's phosphate-buffered saline (without calcium was purchased from Flow Laboratories, Rickmansworth, and magnesium) U.K. Normal mouse serum used in the second-antibody assay method was obtained from non-immune Balb-c mice. Sheep-antimouse I gG- antiserum was produced by imnunising a Kerry-Gritstone cross-bred ewe with 500 clg mouse IgG (Sigma, U.K.) at monthly intervals and collecting serum lo-14 days after the last injection. Synthesis
of Imnunogens
(Fig.
1)
-la-Hydroxy-25,26,27-trinor-24-cholecalcioic acid (la(OH)D3-acid) and la-hydroxy-26,27-dinor-cholecalciferol-25-oxime (la(OH)D -oxime) were each conjugated to BSA using the mixed anhydride reastion (7). Ultraviolet spectroscopic analysis indicated that the complexes contained 36 and 30 steroid residues per mol BSA, respectively.
lwhydroxy-25,26,27-trinor2kcholecalcioic
Fig.
1.
Structures
la-hydroxy-26,27-dinorcholecalciferol-25-oxime
acid
of the immunogens.
Immunisation Balb-c mice received S.C. injections of 100 pg of the BSA conjugate of either la(OH)D -acid or la(OH)D -oxime emulsified in Freund's complete adjuvant fallowed by multiple3i.p. injections of 10 ug conjugate Four days prior in Freund's incomplete adjuvant at monthly intervals. to the fusion, 10 ug conjugate in phosphate-buffered saline was given i.v. to the mouse selected on the basis of high antibody titre or of high specificity for 1,25(OH)2D3. Hybridization
and Antibody
The method for preparation in detail elsewhere (4,5).
Production of monoclonal antibodies has Briefly, spleen lymphocytes
been dgscribed (2 x 10 cells
S
744
?P-EOXDI
were fused with 1 x lo7 P3-NSI/l-Ag4 mouse myeloma cells (donated by Dr. C. Milstein) using polyethylene glycol 4000 (Merck, Darmstadt) and fused cells were distributed into 6 x 96 well plates (Costar, layers' of non-immunised mouse Northumbria Biologicals5Ltd.1 on 'feeder After 6-10 days, hybrid colonies were spleen cells (1 x 10 /well). tested for antibody secretion and positive cultures were cloned at least twice by limiting dilutiqn (8). Hybrids chosen for further study were injected i.p. (1 x 10 cells) into pristane-primed Balb-c mice and antibody in ascitic fluid was isolated lo-14 days later; the antibody was stored at -20'. Antibody
Detection
Assay
A preprecipitated double-antibody assay was performed as previously described for progesterone (4) to detect antibody in culture medium from hybrid cells. Briefly, 100 ul culture medium, mouse antiserum, or ascitic fluid was incubated with 50 ul sheep-antimouse IgG-antiserum (1:20) and 50 ul normal mouse serum (1:400) for 18 h at 6". The antibody complex was precipitated by centrifugation at 1700g for 20 min and resuspended with 2503u1 sodium phosphate buffer plus 10 ul ethanol containing 13 fmol [ H]- 1,25(OH) D . After 2h at 6' the samples were diluted with 500 ul ice-cold phosshate buffer (pH 6.0, 0.07 M) and bound fractions were separated and counted for radioactivity. Assessment
of Antibody-secreting
Hybrids
Hybrids were classified as antibody-secretors when binding of C3H]-1 25(OH) D was greater than 10% above that in lymphocytes culturid as c&&ols. Where there were large numbers of antibodysecreting hybrids, a preliminary assessment of sensitivity and specificity was carried out to select the better antibody-secreting hybrids for large scale production in vivo. Dilutions of antibody were chosen to give 25% binding of C3RR(OH) D and were incubated with sheep-antimouse IgG-antiserum as des#bed above. Displacement of radioligand by li25(OH)2Da was compared with displacement by at least 4 concentrations of e ch potentially interfering metabolite in the appropriate range 6-100,OOOpg per tube. Incubation and separation were performed as described above. Detailed
Assessment
of Monoclonal
Antibodies
in Ascitic
Fluid
Titres were established by diluting ascitic fluid with phosphate buffer containing no gelatin. Sensitivity and specificity were determined from full displacement curves derived from all available vitamin D derivatives (see Reagents) using the preprecipitated doubleantibody method described above. 8 at a given titre was defined as 4e mass of 1,25(OH) D in pg/tube r8[ uired to displace 50% of bound [ HI-1,25(OH) D . 'S$ecificity was expressed in terms of crossreactivity, c Elsulated according to Abraham (9) from the mass of each vitamin D sterol required to displace 50% of bound C3H]-I 25(DH) D Affinity constants were calculated by the method of S&chard'(%)) as modified by Chamness and McGuire (11).
S
745
TlilEOXDI
RESULTS Six fusions
were
acid
3
and
results
are
fusion
varied
itmnunogen,
performed,
derived
was the
produced
titre
hybrids
from produced
selection
each group of
produced
after
of
large
antibodies
the
using to
the
the
same
number of
The most striking
total,
hybrids
compared
of
with
antibody-secreting
6 necessitated
based on
the
antibody-secreting
number
1, 2 and
la(OH)D3-
at the time of
fusions
the fusion.
8-90% of
fusions
titre
relationship
proportion
The
with
la(OH)D3-oxime;
Mouse antibody
a direct
oxirne.
of the better
with
within
higher
in
inmtunized
1.
la(OH)D3-acid, the
mice
ionization
bore
much
with
Table
hybrids
with
0.3-1.0s
in
widely,but
antibody-secreting result
from
shown
the
3
their
a
B50
preliminary
and on
their
specificity. TABLE 1 DETAILS OF FUSIONS AND HYBRIDS PRODUCED IMMUNOGEN
la-hydroxy-26,27-dinorcholecalciferol-25-oxime
la-hydroxy-25,26,27trinor-24-cholecalcioic acid
FUSION
1
Ab TITRE AT FUSIONa NO. HYBRIDS PRODUCED (MAXIMUM 576 PER FUSION) NO. Ab-SECRETINGb HYBRIDS NO. STABLE MONOCLONAL ANTIBODIES CHARACTERISED
6
2
1:450
1:3000
3
1:15000
4
1:6000
5
1:2000
1:2500
533
399
576
478
573
572
42
104
520
5
3
2
9
2
1
0
1
2
a.
Titre (dilution of Ab binding 50% [3H]-l,25(OH)2D3) antiserum from mouse at the time of fusion.
b.
Binding
>lO% above lymphocytes
cultured
as controls.
of polyclonal
746
?E-EOIDI -
=
Fig,
2
secreted
by hybrid
having
no
preliminary
2F5 from fusion
cross-reactivity
25,26(OH12D3 The
the
shows
even with
antibody
the extent
with
characterisation 6.
with
of 22 and 1.3X,
antibody
was highly
specific
24,25(OHj2D3
25(OH)D3,
concentrations
cross-reacted
The antibody
of an
and
of 94, 23 and 29 ng,respectively. and lu-(OHID
1,25(OHj2D2
respectively.
The B5D for
to
[3H]l,25(Otl)2D3
was 85 pg.
lOOByBO 80-
60-
20-
0 10° Fig.2.
10'
Displacement antibody.
From
curves
the preliminary Tables
the imnunogen
laIOH)D3acid had little
able cross-reaction
to vitamin
selection,
characterisation,
B50, but it
I lo3
lo2
2 and 3. (Table
other
D metabolites
nine MAb's From MAb's 21, 3Ell
cross-reactivity with
d lo'p#tube
vitamin
with
10'
using
hybrid
were chosen for produced
from fusion 1,25(OH)2D2
D metabolites;
in
2F5
further
response
to
1 had the lowest and
consider-
208 from fusion
2
747
TABLE 2 CHARACTERISATION OF MONOCLONALANTIBODIES DERIVED FROM IMMIJNISATION WITH la-HYDROXY-25,26,27-TRINOR-24-CHOLECALCIOIC ACID 1
1
1
2
2
6
3Ell
lH6
4Fl
3F6
2D8
167
1:25000
1:2500
1:5500
117500
FUSION MONOCLONALANTIBODY TITRE" DISPLACEMENT(B50)
b
40
270
700
900
1: 45000 1: 20000 60
80
SPECIFICITYC 1,25(OW2D3
100
1,2510H12D2
13
25(OH)D3
100
100
100
100
100
82
51
45
58
6
9
20
12
15
0
24,25(OH)2D3
11
21
63
75
27
0
25,26(OH)2D3
25
68
73
la(OH )D3
0.6
a.
The dilution
of antibody
b.
The concentration radioligand.
of 1,25(OH)2D3
C.
Cross-reactivity,
% compared with
0.7
needed to bind
4
0.6
0 0.5
2
50% C3H] 1,25(OH)2D3.
in pg required l,25(OHj2D3.
to displace
50% of
h h also
possessed
other
the
metabolites
with
of
characteristic
vitamin
antibody
of no cross-reactivity
03, a behaviour
which
2F5 from
fusion
from 2F5; however,this
cell
did not survive).
had also
been detected
in the polyclonal
used for
the fusion
or
hybrid
useful
24,25(OHj2D3.
(Table
3)
fusions
which
gave
MAb
3A8 (fusion
MAb's with
high-affinity
antibody
interesting
because of its
For comparison,data and
for
from
mice
lower
the
similar it
polyclonal
a calf
thymus
antibody
than either
greatly
superior
the polyclonal
has
those
for (12)
1,25(OH12D2
antibody
with
being
the
particularly
1,25(0H)2D2. a
sheep
l
polyclonal
currently
use for
in
31 has a similar
B50
cross-reaction
interfering
or the receptor.
25(OH)D3
5,
MAb167 [Table 2) has a higher
cross-reactivity
mouse
from
from fusion
MAb 5F2 (Table
and
the
la(OH)D3-oxfme
with
with
Table 3
receptor
of with
to
B50 's resulted
shown in
to the thymus receptor.
has
itmnunized
5F2 (Ka = 1.39 x 1O1' M-l)
are
produced
These characteristics
reactivity
high cross-reactivity
observed
(no MAb was
41, hhwas similar
1,25(DHj2D assay in our laboratory. to
was also
serum antibodies
had shown no cross-
Fusions
1 and 2.
antibody
line
6 (Fig.21
with
B50 but
metabolites
S
749
?faIrorr,m
TABLE CHARACTERISATION
OF MONOCLONAL ANTIBODIES DERIVEDFROM IWJNISATIONWITH la-HYDROXYCHOLECALCIFEROL-26.27-DINOR-25-OXIF1E ~._ _ _______ . -.-._- .._ - __. _- - _
FUSION
4
MONOCLONAL
3
ANTIBODY
5
3A8
TITRE'
1:2000
DISP~ACEMENT~B50)b
5
5F2
Calf e Thymus Receptor
Sheep d polyclonal
6012
1:100000 1:60000
1:200000
1:6
100
25
43
23
1,25KW2D3
100
100
100
100
1,25WH)2D2
29
95
22
2
119
25(OH1D3
15
8
13
30
0.2
24,25(OHj2D3
43
11
27
21
0.05
43
16
24
3
2
9
SPECIFICITYC
1
a (OHlO
a. The
dilution
of
b. The concentration of radioligand.
of
c.Cross-reactivity,
d. Raised conjugate, (kindly e.Prepared
antibody
needed 1,25(OH)2D3
X compared
with
0.05
4
0.6
to
bind in
100
50%
pg required
C3H] to
0.1
1.25(OH)2D3 displace
50%
1,25(OH12D3.
in
sheep to la,25-dihydroxycholecalciferol-25-hemisuccinate-~SA Currently in use in our laboratory to assay donated by Dr. J. O'Riordan). in
our
laboratory
1.25(OHj2D2simultaneously.
and
used
to
assay
1,25(OH)2D,
1,25(01i)2D3
J
and
Mb 5F2 was selected developlnent curves
for
of an assay for
in the
further
investigation
1,2510Hj203
The curves
useful
range for
assay of
ligand
at 25 pg/tube.
a view to the
and 1,25(OHj202.
range 1.25 - 160 pg/tube
are shown in Fig.3.
with
for
these
are virtually
1,25(OH12Dwith
Standard
two vitamin
superimposable
D sterols and cover
50% displacement
of
bound
80-
80.
40.
201,2502 1,25D,
OI
l-25
Fig.3.
2.5
5
10
Standard curves t0 1,25(Oh)2D2 at a dilution of 1:lOOOOO.
20
40
and
80
a
180pg/tube
1,25(OH)2D3using MAb5F2
751
DISCUSSION
The measurement binding
present
extraction
to prepare
for
that
difficult
monoclonal
to
antibodies
nine
stable
standardize.
a series
monoclonal
lengthy
assay.
with
There
is
I,26(OH)2D2
and
proteins
are tedious
In this
paper we
have shown
to overcome
these
of 676 antibody-secreting
antibodies
related
receptor
have the potential
and have produced
to the
equally
measurement;
receptor-
necessitating
prior
cross-reacts
or
of structuraTly
concentrations,
techniques
use in routine
and
by radioinrnunoassay
the presence
in much higher
antiserum
1,25(OH)2D3
ulties
by
and HPLC separation
no available
which
1,25(OH12D
assay is complicated
metabolites
that
of
diffic-
hybrids
to 1,25(OH12D3
have
from
been char-
acterised. Two different
insnunogens were compared
but the characteristics to
the imnunogen
produced
from
reactivity
with
specificity
of the resultant
used. the
same
30
immunised
cross-reactivity illustrates antibodies the
and
showed
major
in finite
produced
on
the
advantage
quantity a large
of
the
specificity
167.
polyclonal
being
in a rare
animal
D
antiserum
Antibodies
of very
metabolites. technique
animals produced. can
The
one mouse from more
monoclonal
number of
were cross-
characteristic
antibody
scale.
not be related
to 0% with
this
vitamin
in a large
a suitable
4Fl
only
desirable
of HAb's
la(OH)D3-acid,
of the
however,
interfering
can be raised of
using
to that
could
differing
from 20% with
the fusion;
with a
chances
produced
of widely
of MAb 167 was similar
the preparation
antibodies
inznunogen;
25(OH)D3 varied
of the mouse used for than
Antibodies
in
thus
This in
that
increasing
Such an antibody
then be
produced
low
from
imnortalised imnunisation
S
752
h(OH)Dj-oxime
~4th similar 5F2
also
to the majority
was
produced
T331OXDI
varied
in
of those
which
resulting
differed
cross-reactivity
same
imnunogen
differed
widely
while
showing
broadly
similar
is
with
difficult,
relationship
also
specificities,
therefore,
between
to
the ilrrnunogen
were
but MAb the desired
derived
from
the
from
fusion
1,
had displacement
that
been
confer
predicted
specificity
Ia(
which
reactivity
on
antibodies
substituted
feature
that
values
conclusions
antibodies
the sterol
MAb.
the
However,
substituents,
metabolites
cross-reacted
with
the
it
might
of the sterol synthetic
and
would
derivative little
crbss-
conversely,
strongly
except
region
was an
sidechain
the Both
chain,and
showed very
imnunogen,
on
produced.
side
A ring
from either
interaction
firm
the unrestricted
has no sidechain
with
suggesting
the
draw
and the
have
side-
with
167,
important
of most of the antibodies. The
most
striking
of antibody-secreting hybrids
fusions
that
has yet
which
the
total
resulted
10 from oxime fusions;
number
of
from la(OH)D3-
no explanation
for
this
been found.
a rapid
early
from the vitamin
produced
screening
second-antibody
might
Although
666 secretors
number of hybridomas
preprecipitated interference
only
in the innnunogens was in the number
produced.
in each case,
but
The large
difference
hybrids
was similar
difference
media
in having
in B50; e.g.,those
to B5A through
meant
3A8,
from lu(OH)D3-acid,
fundamentally
inmaunogens were linked
acid
like
from 40 to 900 pg of 1,25(OH)2D3.
It
chain
some,
1.25(OH)2D2. Antibodies
high
varying
specificity;
method
method D-binding
have masked specific
was
in
some of
had
antibody
fusions
to be developed.
used to
globulin
the
present binding.
eliminate in
serum
A any and
Preliminary
S testing ing
for
affinity
hybrids
TEEOXDR
and specificfty cloned
were
and
ensured
that
grown in vivo
only
for
the most
promis-
large-scale'
antibody
production. The
purpose
with
suitable
Until
now
of the present
characteristics only
one
1,25(OHJ2D3; this reported with
further,
other
We have displacement
described
is
in having
enormous
advantages
receptor
needs a source
age,
receptor.
(12).
assay for
1,25(OH)2D
of bound
method used for MAb 167,
although
of 1,25(OH)2D3
the
polyclonal
for
1,25(OH12D2
titre
calf
MAb 5F2 radioligand
the characterisation
over
higher a useful
to been
strongly
differing
in
valuable
antisera
higher
the
B50 than range,
described
than in
low
because
dilution,
of a routine
charcoal
separ-
the second-antibody the
MAb 5F2, still and,
The
of a partic-
on the development
rather
than
supply.
from animals
a
far
MAb possesses
and unrestricted
and involving
so
1,25(OHj2D3
and
a B50 slightly
glands
Work is now in progress
with
MAb's
has to be made at extremely
utilising
not
two show particularly
having
of high
of fresh
and free
has
Mb's
cross-reacted
Compared to the receptor,
and the preparation
1:6 - 1:8
ation
affinity
in terms
raise
that
nine
which
all
displacement,
thymus
described
to
1,25(DH12D.
assay. to
equal
the calf
account
characterising of
superior
for
(13).
1,25(OH12D
and has the requisite
ular
in
a range of Mb's
assays
described
preliminary
specificity, for
been
the antibodies
succeeded
characteristics
was to produce
improved
has
D metabolites
and
MAb 5F2
in a
and
vitamin
for
attempt
was
study
of
present permits its
zero
study. assay cross-
S
754
reaction an can
with
assay be
in
simplified
other which
v itamin the or
TmEOfDI
D
metabolites.
h
may
perm
it
the
fication
development
of
procedures
omitted. ACKNOWLEDGMENTS
This study was supported by awards to the Departments of Chemical Pathology and Medicine from the North Western Regional Health Authority (Research Infrastructure) and by a Project Grant awarded to EBM by the Medical Research Council. Dr. S. Shany was a Bruno Mendel Fellow of the Royal Society. We are grateful to Professor J.G. Ratcliffe and Dr. D-C. Anderson for their helpful suggestions and to Dr.G.N. Smith for chemical advice and assistance. Excellent technical assistance was provided by Mrs. S.M. Hall and Mr. P. Cundall. NOTES
1.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
AND
REFERENCES
Systemic and trivial names of vitamin D secosterols abbreviations used in this paper are as follows: (5Z,7E)-(3S)-9,10-seco-5,7,10(19),22-ergostatetraen-3-o1;
and their vitamin D v?;amin
la(OH )D3-acid; la-hydroxy-26,27-dinorcholecalciferol-25-oxime, la(OH )D3-oxime. Jongen, M.J.M., Van Ginkel, F.C., van der Vijgh, W.J.F., Kuiper, S., Netelenbos, J.C. and Lips, P., CLIN. CHEM. 30, 399 (1984). Fraher, L.J., Adami, S., Clemens, T.L., Jones, G. and O'Riordan, J.L.H., CLIN. ENDOCRINDL. 19, 151 (1983). Kohler, G. and Milstein, C.,mATURE 256, 495 (1975). White, A., Anderson, D.C. and Daly, Jr, J.CLIN.ENDOCRINOL. METAB. 54, 205 (1982). WhXe, A., Gray, C. and Corrie, J.E.T., J. STEROID BIOCHEM.(1985). Erlanger, B.F., Borek, F., Beiser, S.M. and Lieberman, S., J. BIOL. CHEM. 228, 713 (1957). Goding, m., J. IMMUNOL. METHODS39, 285 (1980). Abraham, G.E., J. CLIN. ENDOCRImL. METAB. 29, 866 (19691. Scatchard, G., ANN.N.Y. ACAD. SCI. 51, 660 (1949. Chamness, G.C. and McGuire, W.L., SmOIDS 26, 538 (1975). Reinhardt, T.A., Horst, R.L., Orf, J.W. and Ellis, B.W., J. CLIN. ENDDCRINOL. METAB. 58, 91 (1984). Perry, H.M., Chappel, J.Cy Clevinger, B-L., Haddad, J.G. and Teitelbaum, S.L., BIOCHEM. BIOPHYS. RES. COMM. -’112 431 (1983).
7-31-85 10-21-85)
ABSTRACT
The preparation of high-affinity and high -specificity monoclonal antibodies to Ia,2&dihydroxyvitamin D is described. Monoclonal antibodies were derived from Balb-c mice imnunised with either la-hydroxy-25,26,27acid or trinor-24-cholecalcioic with la-hydroxy-26,27-dinor-cholemouse calciferol-25-oxime, and spleen cells were hybridised with myeloma cells. From six fusions nine monoclonal antibodies (MAb's) were selected from 676 antibody-secreting hybrids. Antibodies varied widely in their ability to bind la,25-dihydroxyvitamin D3 (50% displacement of radioligand ranged from 25 - 900 pg); two had particularly useful characteristics for la,25#Qdroxyvitamin D assay. MAb 5F2 has high affinity (Ka = 1.39 x IO M I and does not discriminate between la,25-dihydroxyvitamin D and D , thus enabling the two forms to be measured together. &b 163 is highly specific, having no crossreactivity with 25-hydroxy-, 24,25-dihydroxyor 25,26-dihydroxyvitamin D at concentrations found in normal human serum; this MAb has the potential to eliminate lengthy extraction procedures the involved in currently available assays for la,25-dihydroxyvitamin D. INTRODUCTION Assays currently vitamin
D,
by the
cross-reacting
food
need
to
of vitamin
August,
present
September
1985
metabolite
(1,25(OH)2D)(1). preparation at
between
in
serum
necessary
higher
to
concentrations, D
An
being
being
1,25(OH)2D3 Steroids
the
from
problem
therapy form.
do not
Volume
remove
derivat-
endogenous
assays
are
(24,25(OHj2D)
additional
derived
of
namely
the la,25-dihydroxylated
former
and the latter in use for
hormonal
(25.26(OHj2D).
D2 and D3, the
antisera
active
24,25-dihydroxyvitamin
D
distinguish
supplementation
polyclonal
sample
D (25(OH)D),
25,26-dihydroxyvitamin
ives
the D
extensive
metabolites
25-hydroxyvitamin
is the
for
la,25-dihydroxyvitamin
complicated
and
available
or Most
cross-react
46, Numbers
2,3
with
binding
assays
varying
equally
other
ing highly
specific
and provides
these methods
for
the
clones
antibodies
to problems
supply
to produce
sensitivity
would
in
cross-react
have the sensitivity
to
or no cross-reactivity
D2 or D3.
The elegant
technique
production
of monoclonal
antibodies
in culture
for
paper we describe
the required
1,25(OH)2D2
1,25(OH)2D
and show little
assay development
In this
MAb's with
(4)
already
in protein-
and presents
and 1.25(OH)2D3,
of vitamin
a constant
improving
(5,6).
the assay of
metabolites
by hybrid
with
is tedious
concentrations,
secreted
(t&b’s)
employed
(2,3)
for
and Milstein
receptors
cross-react
preparation
1,25(OH)2D2
picomolar
of Kohler
thus
and
antibody
with
detect
1,25(OH)2D3
reproducibility
An ideal
with
for
extents,
between-batch
1,25
some of the tissue
1,25(OH)2D2;
the
offers
radioimnunoassay
of antibody and precision. MAb's to
of
of defined
1,25(OH)2D
characteristics,
One of us has exploited
progesterone
the selection and
the means of select-
and testosterone
and characterisation
specificity
for
the
of
assay of
(OH)2D. EXPERIMENTAL
Reaaents Unlabelled 25(OH)D3 was obtained from Roussel, Paris, 25(OH)D2 from Upjohn, Kalamazoo, 24,25(0H) D3 from Roche Products Ltd., 25,26(0H) D from Dr. J. Redel, Par ? s, and 25,26(OH)2D2 from Dr. 0. Willia&s~ Cambridge England. 1,25(OH) D and 1';~;;;;2D3N J ant were kindly suppliid by Hoffman La R&Be Inc., la-hydroxycholecalciferol from Leo Laboratories, Princes Risbo;ough,'U.K. All vitamin D sterols were stored under nitrogen at -20" and concentrations for standard solutions were determined by U.V. spectrophotometry, using the apprgpriate extinction coefficients (X max 26! nm). 1,25(OH) [26,27-methylHID (180 Ci/nanol) was purchased from Amersham Inte 6 nationa1,U.K. la-iydroxy-25,26,26-trinor-24-cholecalcioic
S
TPIIOXDD
743
acid, kindly provided by Dr. Milan Uskokovic and la-hydroxy-26,27-dinorcholecalciferol-25-oxime, given by Dr. L. Fraher, were used to prepare imnunogens. Bovine serum albumin fraction V (BSA) was purchased from Sigma,U.K. The buffer used in the antibody evaluation procedure was 0.07 M sodium phosphate, pH 6.0,and contained 0.1% gelatin, unless specified to the contrary. Dulbecco's phosphate-buffered saline (without calcium was purchased from Flow Laboratories, Rickmansworth, and magnesium) U.K. Normal mouse serum used in the second-antibody assay method was obtained from non-immune Balb-c mice. Sheep-antimouse I gG- antiserum was produced by imnunising a Kerry-Gritstone cross-bred ewe with 500 clg mouse IgG (Sigma, U.K.) at monthly intervals and collecting serum lo-14 days after the last injection. Synthesis
of Imnunogens
(Fig.
1)
-la-Hydroxy-25,26,27-trinor-24-cholecalcioic acid (la(OH)D3-acid) and la-hydroxy-26,27-dinor-cholecalciferol-25-oxime (la(OH)D -oxime) were each conjugated to BSA using the mixed anhydride reastion (7). Ultraviolet spectroscopic analysis indicated that the complexes contained 36 and 30 steroid residues per mol BSA, respectively.
lwhydroxy-25,26,27-trinor2kcholecalcioic
Fig.
1.
Structures
la-hydroxy-26,27-dinorcholecalciferol-25-oxime
acid
of the immunogens.
Immunisation Balb-c mice received S.C. injections of 100 pg of the BSA conjugate of either la(OH)D -acid or la(OH)D -oxime emulsified in Freund's complete adjuvant fallowed by multiple3i.p. injections of 10 ug conjugate Four days prior in Freund's incomplete adjuvant at monthly intervals. to the fusion, 10 ug conjugate in phosphate-buffered saline was given i.v. to the mouse selected on the basis of high antibody titre or of high specificity for 1,25(OH)2D3. Hybridization
and Antibody
The method for preparation in detail elsewhere (4,5).
Production of monoclonal antibodies has Briefly, spleen lymphocytes
been dgscribed (2 x 10 cells
S
744
?P-EOXDI
were fused with 1 x lo7 P3-NSI/l-Ag4 mouse myeloma cells (donated by Dr. C. Milstein) using polyethylene glycol 4000 (Merck, Darmstadt) and fused cells were distributed into 6 x 96 well plates (Costar, layers' of non-immunised mouse Northumbria Biologicals5Ltd.1 on 'feeder After 6-10 days, hybrid colonies were spleen cells (1 x 10 /well). tested for antibody secretion and positive cultures were cloned at least twice by limiting dilutiqn (8). Hybrids chosen for further study were injected i.p. (1 x 10 cells) into pristane-primed Balb-c mice and antibody in ascitic fluid was isolated lo-14 days later; the antibody was stored at -20'. Antibody
Detection
Assay
A preprecipitated double-antibody assay was performed as previously described for progesterone (4) to detect antibody in culture medium from hybrid cells. Briefly, 100 ul culture medium, mouse antiserum, or ascitic fluid was incubated with 50 ul sheep-antimouse IgG-antiserum (1:20) and 50 ul normal mouse serum (1:400) for 18 h at 6". The antibody complex was precipitated by centrifugation at 1700g for 20 min and resuspended with 2503u1 sodium phosphate buffer plus 10 ul ethanol containing 13 fmol [ H]- 1,25(OH) D . After 2h at 6' the samples were diluted with 500 ul ice-cold phosshate buffer (pH 6.0, 0.07 M) and bound fractions were separated and counted for radioactivity. Assessment
of Antibody-secreting
Hybrids
Hybrids were classified as antibody-secretors when binding of C3H]-1 25(OH) D was greater than 10% above that in lymphocytes culturid as c&&ols. Where there were large numbers of antibodysecreting hybrids, a preliminary assessment of sensitivity and specificity was carried out to select the better antibody-secreting hybrids for large scale production in vivo. Dilutions of antibody were chosen to give 25% binding of C3RR(OH) D and were incubated with sheep-antimouse IgG-antiserum as des#bed above. Displacement of radioligand by li25(OH)2Da was compared with displacement by at least 4 concentrations of e ch potentially interfering metabolite in the appropriate range 6-100,OOOpg per tube. Incubation and separation were performed as described above. Detailed
Assessment
of Monoclonal
Antibodies
in Ascitic
Fluid
Titres were established by diluting ascitic fluid with phosphate buffer containing no gelatin. Sensitivity and specificity were determined from full displacement curves derived from all available vitamin D derivatives (see Reagents) using the preprecipitated doubleantibody method described above. 8 at a given titre was defined as 4e mass of 1,25(OH) D in pg/tube r8[ uired to displace 50% of bound [ HI-1,25(OH) D . 'S$ecificity was expressed in terms of crossreactivity, c Elsulated according to Abraham (9) from the mass of each vitamin D sterol required to displace 50% of bound C3H]-I 25(DH) D Affinity constants were calculated by the method of S&chard'(%)) as modified by Chamness and McGuire (11).
S
745
TlilEOXDI
RESULTS Six fusions
were
acid
3
and
results
are
fusion
varied
itmnunogen,
performed,
derived
was the
produced
titre
hybrids
from produced
selection
each group of
produced
after
of
large
antibodies
the
using to
the
the
same
number of
The most striking
total,
hybrids
compared
of
with
antibody-secreting
6 necessitated
based on
the
antibody-secreting
number
1, 2 and
la(OH)D3-
at the time of
fusions
the fusion.
8-90% of
fusions
titre
relationship
proportion
The
with
la(OH)D3-oxime;
Mouse antibody
a direct
oxirne.
of the better
with
within
higher
in
inmtunized
1.
la(OH)D3-acid, the
mice
ionization
bore
much
with
Table
hybrids
with
0.3-1.0s
in
widely,but
antibody-secreting result
from
shown
the
3
their
a
B50
preliminary
and on
their
specificity. TABLE 1 DETAILS OF FUSIONS AND HYBRIDS PRODUCED IMMUNOGEN
la-hydroxy-26,27-dinorcholecalciferol-25-oxime
la-hydroxy-25,26,27trinor-24-cholecalcioic acid
FUSION
1
Ab TITRE AT FUSIONa NO. HYBRIDS PRODUCED (MAXIMUM 576 PER FUSION) NO. Ab-SECRETINGb HYBRIDS NO. STABLE MONOCLONAL ANTIBODIES CHARACTERISED
6
2
1:450
1:3000
3
1:15000
4
1:6000
5
1:2000
1:2500
533
399
576
478
573
572
42
104
520
5
3
2
9
2
1
0
1
2
a.
Titre (dilution of Ab binding 50% [3H]-l,25(OH)2D3) antiserum from mouse at the time of fusion.
b.
Binding
>lO% above lymphocytes
cultured
as controls.
of polyclonal
746
?E-EOIDI -
=
Fig,
2
secreted
by hybrid
having
no
preliminary
2F5 from fusion
cross-reactivity
25,26(OH12D3 The
the
shows
even with
antibody
the extent
with
characterisation 6.
with
of 22 and 1.3X,
antibody
was highly
specific
24,25(OHj2D3
25(OH)D3,
concentrations
cross-reacted
The antibody
of an
and
of 94, 23 and 29 ng,respectively. and lu-(OHID
1,25(OHj2D2
respectively.
The B5D for
to
[3H]l,25(Otl)2D3
was 85 pg.
lOOByBO 80-
60-
20-
0 10° Fig.2.
10'
Displacement antibody.
From
curves
the preliminary Tables
the imnunogen
laIOH)D3acid had little
able cross-reaction
to vitamin
selection,
characterisation,
B50, but it
I lo3
lo2
2 and 3. (Table
other
D metabolites
nine MAb's From MAb's 21, 3Ell
cross-reactivity with
d lo'p#tube
vitamin
with
10'
using
hybrid
were chosen for produced
from fusion 1,25(OH)2D2
D metabolites;
in
2F5
further
response
to
1 had the lowest and
consider-
208 from fusion
2
747
TABLE 2 CHARACTERISATION OF MONOCLONALANTIBODIES DERIVED FROM IMMIJNISATION WITH la-HYDROXY-25,26,27-TRINOR-24-CHOLECALCIOIC ACID 1
1
1
2
2
6
3Ell
lH6
4Fl
3F6
2D8
167
1:25000
1:2500
1:5500
117500
FUSION MONOCLONALANTIBODY TITRE" DISPLACEMENT(B50)
b
40
270
700
900
1: 45000 1: 20000 60
80
SPECIFICITYC 1,25(OW2D3
100
1,2510H12D2
13
25(OH)D3
100
100
100
100
100
82
51
45
58
6
9
20
12
15
0
24,25(OH)2D3
11
21
63
75
27
0
25,26(OH)2D3
25
68
73
la(OH )D3
0.6
a.
The dilution
of antibody
b.
The concentration radioligand.
of 1,25(OH)2D3
C.
Cross-reactivity,
% compared with
0.7
needed to bind
4
0.6
0 0.5
2
50% C3H] 1,25(OH)2D3.
in pg required l,25(OHj2D3.
to displace
50% of
h h also
possessed
other
the
metabolites
with
of
characteristic
vitamin
antibody
of no cross-reactivity
03, a behaviour
which
2F5 from
fusion
from 2F5; however,this
cell
did not survive).
had also
been detected
in the polyclonal
used for
the fusion
or
hybrid
useful
24,25(OHj2D3.
(Table
3)
fusions
which
gave
MAb
3A8 (fusion
MAb's with
high-affinity
antibody
interesting
because of its
For comparison,data and
for
from
mice
lower
the
similar it
polyclonal
a calf
thymus
antibody
than either
greatly
superior
the polyclonal
has
those
for (12)
1,25(OH12D2
antibody
with
being
the
particularly
1,25(0H)2D2. a
sheep
l
polyclonal
currently
use for
in
31 has a similar
B50
cross-reaction
interfering
or the receptor.
25(OH)D3
5,
MAb167 [Table 2) has a higher
cross-reactivity
mouse
from
from fusion
MAb 5F2 (Table
and
the
la(OH)D3-oxfme
with
with
Table 3
receptor
of with
to
B50 's resulted
shown in
to the thymus receptor.
has
itmnunized
5F2 (Ka = 1.39 x 1O1' M-l)
are
produced
These characteristics
reactivity
high cross-reactivity
observed
(no MAb was
41, hhwas similar
1,25(DHj2D assay in our laboratory. to
was also
serum antibodies
had shown no cross-
Fusions
1 and 2.
antibody
line
6 (Fig.21
with
B50 but
metabolites
S
749
?faIrorr,m
TABLE CHARACTERISATION
OF MONOCLONAL ANTIBODIES DERIVEDFROM IWJNISATIONWITH la-HYDROXYCHOLECALCIFEROL-26.27-DINOR-25-OXIF1E ~._ _ _______ . -.-._- .._ - __. _- - _
FUSION
4
MONOCLONAL
3
ANTIBODY
5
3A8
TITRE'
1:2000
DISP~ACEMENT~B50)b
5
5F2
Calf e Thymus Receptor
Sheep d polyclonal
6012
1:100000 1:60000
1:200000
1:6
100
25
43
23
1,25KW2D3
100
100
100
100
1,25WH)2D2
29
95
22
2
119
25(OH1D3
15
8
13
30
0.2
24,25(OHj2D3
43
11
27
21
0.05
43
16
24
3
2
9
SPECIFICITYC
1
a (OHlO
a. The
dilution
of
b. The concentration of radioligand.
of
c.Cross-reactivity,
d. Raised conjugate, (kindly e.Prepared
antibody
needed 1,25(OH)2D3
X compared
with
0.05
4
0.6
to
bind in
100
50%
pg required
C3H] to
0.1
1.25(OH)2D3 displace
50%
1,25(OH12D3.
in
sheep to la,25-dihydroxycholecalciferol-25-hemisuccinate-~SA Currently in use in our laboratory to assay donated by Dr. J. O'Riordan). in
our
laboratory
1.25(OHj2D2simultaneously.
and
used
to
assay
1,25(OH)2D,
1,25(01i)2D3
J
and
Mb 5F2 was selected developlnent curves
for
of an assay for
in the
further
investigation
1,2510Hj203
The curves
useful
range for
assay of
ligand
at 25 pg/tube.
a view to the
and 1,25(OHj202.
range 1.25 - 160 pg/tube
are shown in Fig.3.
with
for
these
are virtually
1,25(OH12Dwith
Standard
two vitamin
superimposable
D sterols and cover
50% displacement
of
bound
80-
80.
40.
201,2502 1,25D,
OI
l-25
Fig.3.
2.5
5
10
Standard curves t0 1,25(Oh)2D2 at a dilution of 1:lOOOOO.
20
40
and
80
a
180pg/tube
1,25(OH)2D3using MAb5F2
751
DISCUSSION
The measurement binding
present
extraction
to prepare
for
that
difficult
monoclonal
to
antibodies
nine
stable
standardize.
a series
monoclonal
lengthy
assay.
with
There
is
I,26(OH)2D2
and
proteins
are tedious
In this
paper we
have shown
to overcome
these
of 676 antibody-secreting
antibodies
related
receptor
have the potential
and have produced
to the
equally
measurement;
receptor-
necessitating
prior
cross-reacts
or
of structuraTly
concentrations,
techniques
use in routine
and
by radioinrnunoassay
the presence
in much higher
antiserum
1,25(OH)2D3
ulties
by
and HPLC separation
no available
which
1,25(OH12D
assay is complicated
metabolites
that
of
diffic-
hybrids
to 1,25(OH12D3
have
from
been char-
acterised. Two different
insnunogens were compared
but the characteristics to
the imnunogen
produced
from
reactivity
with
specificity
of the resultant
used. the
same
30
immunised
cross-reactivity illustrates antibodies the
and
showed
major
in finite
produced
on
the
advantage
quantity a large
of
the
specificity
167.
polyclonal
being
in a rare
animal
D
antiserum
Antibodies
of very
metabolites. technique
animals produced. can
The
one mouse from more
monoclonal
number of
were cross-
characteristic
antibody
scale.
not be related
to 0% with
this
vitamin
in a large
a suitable
4Fl
only
desirable
of HAb's
la(OH)D3-acid,
of the
however,
interfering
can be raised of
using
to that
could
differing
from 20% with
the fusion;
with a
chances
produced
of widely
of MAb 167 was similar
the preparation
antibodies
inznunogen;
25(OH)D3 varied
of the mouse used for than
Antibodies
in
thus
This in
that
increasing
Such an antibody
then be
produced
low
from
imnortalised imnunisation
S
752
h(OH)Dj-oxime
~4th similar 5F2
also
to the majority
was
produced
T331OXDI
varied
in
of those
which
resulting
differed
cross-reactivity
same
imnunogen
differed
widely
while
showing
broadly
similar
is
with
difficult,
relationship
also
specificities,
therefore,
between
to
the ilrrnunogen
were
but MAb the desired
derived
from
the
from
fusion
1,
had displacement
that
been
confer
predicted
specificity
Ia(
which
reactivity
on
antibodies
substituted
feature
that
values
conclusions
antibodies
the sterol
MAb.
the
However,
substituents,
metabolites
cross-reacted
with
the
it
might
of the sterol synthetic
and
would
derivative little
crbss-
conversely,
strongly
except
region
was an
sidechain
the Both
chain,and
showed very
imnunogen,
on
produced.
side
A ring
from either
interaction
firm
the unrestricted
has no sidechain
with
suggesting
the
draw
and the
have
side-
with
167,
important
of most of the antibodies. The
most
striking
of antibody-secreting hybrids
fusions
that
has yet
which
the
total
resulted
10 from oxime fusions;
number
of
from la(OH)D3-
no explanation
for
this
been found.
a rapid
early
from the vitamin
produced
screening
second-antibody
might
Although
666 secretors
number of hybridomas
preprecipitated interference
only
in the innnunogens was in the number
produced.
in each case,
but
The large
difference
hybrids
was similar
difference
media
in having
in B50; e.g.,those
to B5A through
meant
3A8,
from lu(OH)D3-acid,
fundamentally
inmaunogens were linked
acid
like
from 40 to 900 pg of 1,25(OH)2D3.
It
chain
some,
1.25(OH)2D2. Antibodies
high
varying
specificity;
method
method D-binding
have masked specific
was
in
some of
had
antibody
fusions
to be developed.
used to
globulin
the
present binding.
eliminate in
serum
A any and
Preliminary
S testing ing
for
affinity
hybrids
TEEOXDR
and specificfty cloned
were
and
ensured
that
grown in vivo
only
for
the most
promis-
large-scale'
antibody
production. The
purpose
with
suitable
Until
now
of the present
characteristics only
one
1,25(OHJ2D3; this reported with
further,
other
We have displacement
described
is
in having
enormous
advantages
receptor
needs a source
age,
receptor.
(12).
assay for
1,25(OH)2D
of bound
method used for MAb 167,
although
of 1,25(OH)2D3
the
polyclonal
for
1,25(OH12D2
titre
calf
MAb 5F2 radioligand
the characterisation
over
higher a useful
to been
strongly
differing
in
valuable
antisera
higher
the
B50 than range,
described
than in
low
because
dilution,
of a routine
charcoal
separ-
the second-antibody the
MAb 5F2, still and,
The
of a partic-
on the development
rather
than
supply.
from animals
a
far
MAb possesses
and unrestricted
and involving
so
1,25(OHj2D3
and
a B50 slightly
glands
Work is now in progress
with
MAb's
has to be made at extremely
utilising
not
two show particularly
having
of high
of fresh
and free
has
Mb's
cross-reacted
Compared to the receptor,
and the preparation
1:6 - 1:8
ation
affinity
in terms
raise
that
nine
which
all
displacement,
thymus
described
to
1,25(DH12D.
assay. to
equal
the calf
account
characterising of
superior
for
(13).
1,25(OH12D
and has the requisite
ular
in
a range of Mb's
assays
described
preliminary
specificity, for
been
the antibodies
succeeded
characteristics
was to produce
improved
has
D metabolites
and
MAb 5F2
in a
and
vitamin
for
attempt
was
study
of
present permits its
zero
study. assay cross-
S
754
reaction an can
with
assay be
in
simplified
other which
v itamin the or
TmEOfDI
D
metabolites.
h
may
perm
it
the
fication
development
of
procedures
omitted. ACKNOWLEDGMENTS
This study was supported by awards to the Departments of Chemical Pathology and Medicine from the North Western Regional Health Authority (Research Infrastructure) and by a Project Grant awarded to EBM by the Medical Research Council. Dr. S. Shany was a Bruno Mendel Fellow of the Royal Society. We are grateful to Professor J.G. Ratcliffe and Dr. D-C. Anderson for their helpful suggestions and to Dr.G.N. Smith for chemical advice and assistance. Excellent technical assistance was provided by Mrs. S.M. Hall and Mr. P. Cundall. NOTES
1.
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
AND
REFERENCES
Systemic and trivial names of vitamin D secosterols abbreviations used in this paper are as follows: (5Z,7E)-(3S)-9,10-seco-5,7,10(19),22-ergostatetraen-3-o1;
and their vitamin D v?;amin
la(OH )D3-acid; la-hydroxy-26,27-dinorcholecalciferol-25-oxime, la(OH )D3-oxime. Jongen, M.J.M., Van Ginkel, F.C., van der Vijgh, W.J.F., Kuiper, S., Netelenbos, J.C. and Lips, P., CLIN. CHEM. 30, 399 (1984). Fraher, L.J., Adami, S., Clemens, T.L., Jones, G. and O'Riordan, J.L.H., CLIN. ENDOCRINDL. 19, 151 (1983). Kohler, G. and Milstein, C.,mATURE 256, 495 (1975). White, A., Anderson, D.C. and Daly, Jr, J.CLIN.ENDOCRINOL. METAB. 54, 205 (1982). WhXe, A., Gray, C. and Corrie, J.E.T., J. STEROID BIOCHEM.(1985). Erlanger, B.F., Borek, F., Beiser, S.M. and Lieberman, S., J. BIOL. CHEM. 228, 713 (1957). Goding, m., J. IMMUNOL. METHODS39, 285 (1980). Abraham, G.E., J. CLIN. ENDOCRImL. METAB. 29, 866 (19691. Scatchard, G., ANN.N.Y. ACAD. SCI. 51, 660 (1949. Chamness, G.C. and McGuire, W.L., SmOIDS 26, 538 (1975). Reinhardt, T.A., Horst, R.L., Orf, J.W. and Ellis, B.W., J. CLIN. ENDDCRINOL. METAB. 58, 91 (1984). Perry, H.M., Chappel, J.Cy Clevinger, B-L., Haddad, J.G. and Teitelbaum, S.L., BIOCHEM. BIOPHYS. RES. COMM. -’112 431 (1983).