s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate /s tero ids

A direct antigen heterologous enzyme immunoassay formeasuring progesterone in serum without using displacer

Anupam Basua, Tulsidas G. Shrivastava,∗, Saumen Kumar Maitrab

a Department of Reproductive Biomedicine, National Institute of Health and Family Welfare, Munirka, New Delhi 110067, Indiab Department of Zoology, Visva Bharati University, Santiniketan 731235, India

a r t i c l e i n f o

Article history:

Received 11 April 2005

Received in revised form 22

September 2005

Accepted 20 October 2005

a b s t r a c t

An antigen heterologous enzyme-linked immunosorbent assay (ELISA) for directly

measuring progesterone in serum is described. Six combinations of antigens and

enzyme conjugates were tested; the enzyme conjugate 17-�OH-progesterone-3-O-

carboxymethyloxime-alkalinephosphatase (17-�OH-P-3-CMO-ALP) and the immunogen

progesterone-3-carboxymethyloxime-bovine serum albumin (P-3-CMO-BSA) were found to

Available online 15 December 2005

Keywords:

ELISA

Antigen heterologous

Progesterone

Displacer

be best. Fifty microliters of standard or serum sample and 100 �L of the 17-�OH-P-3-CMO-

ALP enzyme conjugate were added to the antibody coated wells, and incubated for 1 h at

37 ◦C. Bound enzyme activity was measured by using p-nitrophenyl phosphate as substrate.

The sensitivity of the assay was 0.11 ng/mL, and intra- and inter-assay CVs ranged from 5.1%

to 9.6%. The analytical recoveries were 97–105%. The serum progesterone values obtained by

this method correlated well with those obtained by radioimmunoassay; r = 0.97 (n = 44). More-

over, in this ELISA no displacing agent was used or special means was required to displace

progesterone from corticosteroid binding globulin (CBG). Serum progesterone concentra-

tions of subjects, with histories of recurrent spontaneous abortions were also measured,

and correlated well with clinical history.

© 2005 Elsevier Inc. All rights reserved.

1. Introduction

Progesterone, a C21 steroid secreted by the corpus luteum,promotes the development of the endometrial lining. Serumlevels of progesterone rise during the luteal phase of themenstrual cycle. If conception occurs, levels increase dra-matically from the end of the first trimester to term duringpregnancy. Because progesterone is required for the mainte-nance of pregnancy low levels are associated with luteal phasedefect, ectopic gestation, and miscarriage. Many immunoas-say techniques have been developed for measuring proges-terone in serum. Most of these are radioimmunoassay (RIA),which involve the handling of radioactive materials. Given theinherent problems therein different non-radioactive methods

∗ Corresponding author. Tel.: +91 11 2616 6441; fax: +91 11 2610 1623.E-mail address: tgsrivastava@nihfw.org (T.G. Shrivastav).

have been developed for measuring progesterone, such aschemiluminescence immunoassays (CLIA) [1], time-resolvedfluorescence immunoassays (TRFI) [2], fluorescence polar-ization immunoassays (FPI) [3], along with enzyme-linkedimmunosorbent assays (ELISA). Although CLIA and TRFIAcompare favorably with RIA in terms of performance (sensitiv-ity), whereas, instruments required for end product measure-ment are costlier. FPIs of progesterone, on the other hand, havenot achieved desired levels of sensitivity [3,4]. Thus, the effec-tiveness of these techniques in the ordinary clinical setup islimited. Enzyme immunoassays (EIA) are still one of the mostpopular methods for measuring hormone concentrations.Homologous or site-heterologous EIAs using progesterone-11-hemisuccinate (P-11-HS) as the immunogen and P-11-HS

0039-128X/$ – see front matter © 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.steroids.2005.10.006

s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230 223

or progesterone-3-O-carboxymethyloxime (P-3-O-CMO) as theenzyme conjugate have been developed [5–9], but failed toachieve proper sensitivity. Mitsuma et al. [10] used geometricisomers of P-3-O-CMO, for progesterone assay having sensitiv-ity 0.16 pg/mL. Different homologous and heterologous EIAsfor progesterone (bridge and site) were developed to achievedesired sensitivity and specificity [11]. The highest sensitiv-ity achieved was 0.25 pg/mL with human serum using anti-body generated against 7-�-carboxyethyl-thioether proges-terone in combination with 6-�-progesterone-hemisuccinate-enzyme conjugate. Wu et al. [12] synthesized carboxylic acidderivatives of progesterone with different bridge lengths andconjugated them with ovalbumin, then tested them in an anti-gen immobilized progesterone EIA; the sensitivities rangedfrom 0.5 ng/mL to 50 ng/mL. In the above-mentioned assays,attempts were made to find combinations of immunogensand enzyme conjugates that led to progesterone EIAs effec-tive in terms of both sensitivity and specificity. However, theperformances of these EIAs were not up to mark in terms ofsensitivity [11,12].

Generally, for the measurement of steroids directly fromthe serum, different reagents or combination of reagents wereused as displacer in enzyme conjugate or sample dilutionbuffer to displace the steroid from the specific steroid bind-ing protein. Danazol, dexamethasone, and cortisol were usedas displacer in enzyme conjugate buffer for measuring proges-terone directly from serum [13–15]. Trichloroacetic acid, dihy-duIdcitttat

2

Tc

2

PC(hu(tpaUfwP

chased from the Sisco Research Laboratory (SRL), Bombay,India.

2.1.1. Preparation of immunogen and generation ofprogesterone antibodyProgesterone-3-O-carboxymethyloxime (P-3-O-CMO) andprogesterone-11-hemisuccinate (P-11-HS) were conjugated toBSA according to the method of Basu et al. [18]. In brief, tenmilligrams (10 mg) of P-3-O-CMO or P-11-HS were dissolved in400 �L of dioxan and 400 �L of dimethyl formamide. In 200 �Lof distilled water, 20 mg of N-hydroxysuccinimide (NHS) and40 mg of 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimidehydrochloride (EDAC-HCl) were dissolved, and the aqueousmixture was added to the steroid solution. The reactionmixture was vortex-mixed and incubated overnight at 4 ◦Cto activate the –COOH group of the steroid. One hundredmilligrams of bovine serum albumin (BSA) was then dis-solved in 100 mL of distilled water. The activated steroid wasadded slowly to the aqueous solution of BSA. The reactionmixture was further vortex-mixed and incubated overnightat 4 ◦C. The steroid–BSA conjugate was then dialyzed againstdistilled water, and lyophilized. Polyclonal antibodies weregenerated in New Zealand white rabbits using P-3-O-CMO-BSA and P-11-HS-BSA immunogens as described elsewhere[18].

2.1.2. Preparation of alkaline phosphatase conjugate with

rotestosterone, saponin, and methyl isothiazolinone weresed in sample dilution buffer for the same purpose [16,17].

n the present work, for the measurement of progesteroneirectly from the serum no steroid displacer either in enzymeonjugate buffer or sample dilution buffer was added. Thiss due to absence of specific binding globulin for proges-erone, as 80% of progesterone is bound to albumin and 18%o transcortin and remaining 2% is free. The present study iso our knowledge, the first report of an antigen heterologousssay for the direct measurement of serum levels of proges-erone by EIA.

. Experimental

his study was conducted according to the Institutional Ethi-al Committee’s norms.

.1. Materials and methods

rogesterone, progesterone-3-O-carboxymethyloxime (P-3-O-MO), progesterone-11-hemisuccinate (P-11-HS), 1-ethyl-3-

3-dimethyl-aminopropyl)-carbodiimide-HCL (EDAC-HCL), N-ydroxysuccinimide (NHS), bovine serum albumin (BSA), Fre-nd’s complete adjuvant (FCA), diethyl aminoethyl-sephadex

DEAE-sephadex), and sodium azide were all purchased fromhe Sigma Chemical Company, St. Louis, MO, USA. 17-�OH-rogesterone-3-O-carboxymethyloxime (17-�OH-P-3-O-CMO)nd other steroids were purchased from Steraloids Inc.,K. Alkaline phosphatase (Cat. No. EL-1L) was procured

rom Bangalore Genei, India. Breakable strip-based micro-ell plates were procured from Thermo Labsystems India

vt Ltd. All other chemicals and buffer salts were pur-

progesterone derivativesAlkaline phosphatase was conjugated to P-3-O-CMO, P-11-HS, and 17-�OH-progesterone-3-O-CMO (17-�OH-P-3-O-CMO)according to the method of Basu et al. [18], with modification.In brief, 4 mg of each progesterone derivative were dissolvedin 50 �L of dimethyl formamide (DMF) and 50 �L dioxan. Toeach of the dissolved steroid derivatives, 100 �L of distilledwater containing 5 mg of N-hydroxysuccinimide and 10 mg of1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide hydrochlo-ride (EDAC-HCl) were added. The reaction mixture was vortex-mixed and incubated overnight at 4 ◦C. After overnight incu-bation, the reaction mixture was dried under a stream ofN2 gas. One hundred and twenty-five microliters of alka-line phosphatase (supplied in a solution of 50% glycerol,containing 5 mM Tris, 5 mM MgCl2 and 0.1 mM ZnCl2) and875 �L of distilled water were then added. The reaction mix-ture was vortex-mixed and incubated overnight at 4 ◦C, thendialyzed against 5 mM Tris–aminomethane buffer (0.6 g ofTris–aminomethane, 1 g of MgCl2, 20 g sucrose, and 1 g NaN3

in 1 L of distilled water). The dialysate was centrifuged andthe supernatant collected; 50% ethylene glycol (v/v), 1% BSA,9% sucrose and 10% ammonium sulphate (w/v) were addedto it and the mixture was stored at −30 ◦C. The dilutionsof the ALP conjugate were prepared in 50 mM Tris buffer(pH 8.0) and working dilution was determined by checkerboard assay.

2.1.3. Immobilization of progesterone antibodies onpolystyrene wellsPolystyrene wells of the micro-titer plate were coated eitherwith P-3-O-CMO antibody or with P-11-HS antibody, accord-ing to the immuno-chemical techniques of Shrivastav et al.[19].

224 s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230

2.1.4. Checker board assayThe optimal dilutions of antibody and enzyme conjugateswere found out by checker board assay [19].

2.1.5. Preparation of progesterone standardSix progesterone-working standards (0 ng/mL, 0.5 ng/mL,2 ng/mL, 10 ng/mL, 20 ng/mL, and 50 ng/mL) were prepared in10 mM PBS supplemented with 3% BSA.

2.1.6. Sample collectionNormal serum samples were collected by venapuncture fromnon-pregnant female volunteers (having age group of 20–35years), not taking oral contraception. All of them were non-alcoholic with normal food habit.

Clinical samples were collected from patients attendingthe infertility clinic of the Institute who have had historiesof recurrent spontaneous pregnancy loss. Most of the patientswere in the age group of 25–30 years and having menstrualcycle length of 25–31 days (28 ± 3).

2.1.7. Standard displacement assayTo the antibody-coated wells, 50 �L of different concen-trations of standard were added in duplicate. One hun-dred microliters of the working dilution of the enzymeconjugate (P-3-O-CMO-ALP/P-11-HS-ALP/17-�OH-P-3-O-CMO-ALP) were added to all of the wells, which were then incu-bated for 1 h at 37 ◦C. After incubation, the contents of

The above-mentioned six assay combinations were testedin the standard displacement assay as described above.

The combination of antiserum and progesterone–enzymeconjugate, that displayed the best affinity, slope of curve, sen-sitivity, and specificity, was chosen.

2.1.9. Radioimmunoassay (RIA) of the samplesRIA of the samples was performed by the Immunotech RIAkit (Immunotech, France) utilizing a progesterone antibody-coated tube and 125I-labeled progesterone, according to themanufacturer’s guidelines.

2.2. Determination of normal reference ranges ofprogesterone in the different phases of menstrual cycle

The levels of progesterone were measured in the differentphases of the menstrual cycle of normal volunteers. The vol-unteers were in the age group of 25–35 years having normalmenstrual history (M/H 28 ± 2) and were not taking any oralcontraceptive pills. The serum progesterone levels were mea-sured in follicular phase, luteal phase, and also in the mid-luteal peak. For determination of mid-luteal value, blood sam-ples were drawn on days 18–22 (depending on the length oftheir cycles). The reference range was estimated by applyingthe statistical formula of mean ± 2S.D. (95% confidence limit).

2.3. Data analysis

the wells were decanted, and the wells washed in run-ning tap water. Finally, to measure the bound enzyme activ-ity, 100 �L of substrate solution (1 mg/mL of p-nitrophenylphosphate (PNPP) dissolved in 1% diethanolamine buffer)was added into all of the wells, and they were incu-bated 30 min at 37 ◦C. The color of the solution in thewells was measured at 405 nm by a Tecan Spectra Micro-well reader. The progesterone concentration was determinedby in-house developed computer program using logit–logmethod.

2.1.8. Selection of best combination of antiserum andenzyme conjugateThere were two progesterone antibodies generated (P-3-O-CMO-BSA and P-11-HS-BSA), and three enzyme conjugates(P-3-O-CMO-ALP, P-11-HS-ALP, and 17-�OH-P-ALP) were pre-pared. Using these antibodies and enzyme conjugates, differ-ent combinations (heterologous and homologous) of proges-terone assay were formulated as follows:

Combination 1: antibody P-11-HS-BSA and enzyme conjugateP-11-HS-ALP;Combination 2: antibody P-11-HS-BSA and enzyme conjugateP-3-O-CMO-ALP;Combination 3: antibody P-11-HS-BSA and enzyme conjugate17-�OH-P-3-O-CMO-ALP;Combination 4: antibody P-3-O-CMO-BSA and enzyme conju-gate P-11-HS-ALP;Combination 5: antibody P-3-O-CMO-BSA and enzyme conju-gate P-3-O-CMO-ALP;Combination 6: antibody P-3-O-CMO-BSA and enzyme conju-gate 17-�OH-P-3-O-CMO-ALP.

2.3.1. Preparation of standard curve and affinity constantStandard curves were prepared using MS Excel package. Theconcentration was plotted in the X-axis (log scale) and theabsorbency in the Y-axis. Values of the unknown samples,were calculated by a in-house developed personal computerprogramme written in QBASIC language using the logit–loglinear regression method according to the method of Rodbard[20]. The affinity constant of the progesterone antibody for theantigen was estimated by Scatchard plot according to methodof Feldman and Rodbard [21].

2.3.2. Comparison of ELISA with RIAMethod comparison was tested using ‘Deming Regression’model using RIA as gold standard.

2.3.3. Statistical analysisThe statistical analyses like, mean, standard deviations, cor-relation coefficient, and Deming Regression of the data weredone using the Microsoft Excel, ‘Dplot’, and the ‘Medcale’ com-puter program.

3. Results

3.1. Standard curves, sensitivity, and affinity

The standard curves and Scatchard plots of the progesteronestandard displacement assays using different combinations ofantibody and enzyme-conjugated progesterone are shown inFigs. 1–6. The slopes (m), intercepts (c), and sensitivities of theprogesterone assay using each of the different combinationsare given in Table 1.

s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230 225

Fig. 1 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-11-HS-BSA antibody and P-11-HS-ALPenzyme conjugate (plotted by MS Excel).

Fig. 2 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-11-HS-BSA antibody andP-3-O-CMO-ALP enzyme conjugate (plotted by MS Excel).

3.2. Specificity of the progesterone antibody

Based on sensitivity, affinity, and slope of curve, two combi-nations were further evaluated for cross-reaction. The cross-reactivity of the assay using the P-3-O-CMO-BSA antibodywith the P-11-HS-ALP enzyme conjugate was less than 0.1%with C18, C19, and C21 steroids, except for corticosterone(4.5%), 17-�OH-progesterone (6.5%), and pregnenolone (11%),whereas cross-reactivity of the assay using the P-3-O-CMO-BSA antibody with the 17-�OH-P-3-O-CMO-ALP enzyme con-jugate was less than 0.1% with C18, C19, and C21. steroids,

except for 17-�OH-progesterone (3.1%) and pregnenolone(15%).

3.3. Best combination of antibody and enzymeconjugate

The combination, of the P-3-O-CMO-BSA antibody and the 17-�OH-P-O-3-CMO-ALP enzyme conjugate appeared to be thebest in terms of sensitivity and specificity, affinity, and slopeof curve. This combination was further studied for analyti-cal variables like recovery, inter- and intra-assay coefficients

Fig. 3 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-11-HS-BSA antibody and17-�OH-P-3-O-CMO-ALP enzyme conjugate (plotted by MS Excel).

226 s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230

Fig. 4 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-3-O-CMO-BSA antibody withP-3-O-CMO-ALP enzyme conjugate (plotted by MS Excel).

Fig. 5 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-3-O-CMO-BSA antibody withP-11-HS-ALP enzyme conjugate (plotted by MS Excel).

Fig. 6 – Dose–response curve (a) and Scatchard plot (b) of progesterone ELISA using P-3-O-CMO-BSA antibody with17-�OH-P-3-CMO-ALP enzyme conjugate (plotted by MS Excel).

Table 1 – Slope, intercept, and sensitivity of progesterone assay, using different combination of antibodies and enzymeconjugates

No. of assaycombination

Assay combination Slope (m) and intercept (c) Sensitivity of theassay (ng/mL)

m c

1 P-11-HS-BSA antibody with P-11-HS-ALP −0.68 1.86 0.382 P-11-HS-BSA antibody with P-3-O-CMO-ALP −1.1 1.4 0.313 P-11-HS-BSA antibody with 17-�OH-P-3-O-CMO-ALP −1.5 1.97 0.184 P-3-O-CMO-BSA antibody with P-3-O-CMO-ALP −0.89 1.9 0.45 P-3-O-CMO-BSA antibody with P-11-HS-ALP −1.15 1.7 0.26 P-3-O-CMO-BSA antibody with 17-�OH-P-3-O-CMO-ALP −2.2 2.45 0.11

s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230 227

Table 2 – Recoveries of progesterone from exogenously spiked pooled serum

Serum pool no. Progesteroneadded (ng/mL)

Progesteroneexpected (ng/mL)

Progesteroneobserved (ng/mL)

Recovery (%)

Pool A 8.6Pool B 5 13.6 13.13 97Pool C 15 23.6 24 102Pool D 35 43.6 45.6 105

of variation, establishment of normal values, and correlationcoefficients with RIA.

3.4. Validity of the progesterone ELISA using theP-3-O-CMO antibody with the 17-˛OH-P-3-O-CMO-ALPenzyme conjugate

3.4.1. RecoveriesTo the three fractions of human pooled serum, different con-centrations of progesterone were added externally. After spik-ing, the concentration of progesterone was measured andrecovery was calculated for each fraction of the pool. Therecoveries of the three spiked samples ranged from 97% to105% (Table 2).

3.5. Parallelism

Serum from pregnant women of fifth week of gestation wasdiluted in the ratio of 1:2, 1:4, 1:8, with charcoal strippedpooled human serum, for the study of parallelism of the devel-oped ELISA. Each diluted fractions were run in assay alongwith standard doses. The slope of the standard curve anddilution curve of serum sample on logit–log transformed lin-ear regression were found as −1.45 and −1.56, respectively(Fig. 7).

3.5.1. Assay precisionImaf

3FsEsaR

Fig. 7 – Parallelism of progesterone ELISA. Large line isstandard curve and small line is sample dilution curve.(Plotted by ‘DPLOT’ software using logit–log followed bylinear regression method. X-axis is in log scale and Y-axisis in linear scale.)

equation:

Y (ELISA) = 1.13X (RIA) − 1.03

The regression graphs are presented in Fig. 8.

3.6. Determination of normal profile of progesterone inthe menstrual cycle

The normal ranges of progesterone in the different phases ofthe menstrual cycle of normal subjects as determined by thepresent ELISA are given in Table 4.

3.7. Assay of progesterone in the clinical subjects

The serum progesterone levels of 17 clinical subjects, all withhistories of recurrent pregnancy loss, as measured by the

f four internal control samples as estimated by developed

Inter-assay variation

CV (%) Mean progesterone(ng/mL)

S.D. (n = 8) CV (%)

9.6 3.79 0.36 9.47.5 9.76 0.71 7.18.0 14.83 0.77 5.1

nter- and intra-assay coefficients of variation were deter-ined for the four internal control samples. Intra-assay vari-

tion ranged from 5.1% to 9.6%; inter-assay variation rangedrom 5.1% to 9.4% (Table 3).

.5.2. Correlation with RIAorty-one human serum samples (normal as well clinicalamples) were assayed simultaneously by this progesteroneLISA and by the Immunotech RIA kit. The Deming Regres-ion analysis was performed. The percent coefficient of vari-tion (%CV) of ELISA was 7.5% and %CV of RIA was 6.8%.egression analysis of the samples yielded the following

Table 3 – Intra- and inter-assay variation of progesterone oELISA

Sample number Intra-assay variation

Mean progesterone(ng/mL)

S.D. (n = 8)

Sample 1 4.25 0.41Sample 2 9.86 0.74Sample 3 14.66 1.18Sample 4 27.18 1.43

5.1 27.55 1.65 5.8

228 s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230

Fig. 8 – Method comparison by Deming Regression between RIA and ELISA (a) and Residual plotting (b). Values of the RIAplotted in X-axis, whereas values of ELISA plotted in Y-axis. The CV of method ‘X’ is 6.8% where as CV of method ‘Y’ is 7.5%(by ‘MedCale’ software).

Table 4 – Reference ranges of progesterone in thedifferent phases of menstrual cycle as determined by thedeveloped ELISA

Phases of menstrual cycle Normal range of progesterone(mean ± 2S.D.)a (ng/mL)

Follicular phase (n = 51)b 0.3–2.3 (1.3 ± 0.5)a

Luteal phase [excludingmid-luteal values] (n = 50)b

0.7–19.9 (10.3 ± 4.8)a

Mid-luteal value (n = 50)b 8.1–30.1 (19.1 ± 5.5)a

a Denotes mean ± 2S.D.b Data within bracket are sample numbers.

present ELISA are shown in Table 5. These values, correlatewell with the findings of other clinical tests, such as endome-trial biopsy (wherever available) for the diagnosis of lutealphase defect.

4. Discussion

In developing the present progesterone ELISA, six combi-nations of antibodies and enzyme conjugates were tested.The combination of the antibody P-3-O-CMO-BSA and theenzyme conjugate 17-�OH-P-3-O-CMO-ALP was selected forassay development on the basis of its sensitivity, affinity andspecificity.

The developed assay is simple, direct, and convenient touse, as it permits the direct addition of the serum sample into the assay, and it requires only one and a half hours to com-plete. Only 50 �L of standard or serum samples and 100 �Lof enzyme conjugate are required, for addition to the pro-gesterone antibody-coated micro-well. After incubation andsubsequent washing, another 30 min of incubation with thesubstrate is required to get the result. The sensitivity of theassay was found to be 0.11 ng/mL.

Table 5 – Serum progesterone levels and endometrial biopsy study of the subjects having history of recurrent pregnancyloss

Patent no. Serum progesterone (ng/mL)at 20–22 days of cycle

Endometrial biopsy: clinical impression Remark

1 7.9 Early secretary enodmetruim havingcycle days 16–17

LPDa

2 3.6 Early secretary enodmetruim LPD3 2.6 Early secretary e4 7.1 Early secretary e5 5.4 Early secretary e6 5.6 Early secretary e7 2.0 Early secretary e8 1.7 Early secretary e9 7.3 Early secretary e

10 2.0 Early enodmetru11 1.0 Early enodmetru12 1.8 Early enodmetru13 6.8 Early secretary e14 5.4 Early secretary e15 6.4 Early secretary e16 16.8 Mid-secretary e

cycle days 21–2217 18.4 Mid-secretary e

2–23

cycle days 2

a LPD: luteal phase defect.

nodmetruim LPDnodmetruim LPDnodmetruim LPDnodmetruim LPDnodmetruim LPDnodmetruim LPDnodmetruim LPDim LPDim LPDim LPDnodmetruim LPDnodmetruim LPDnodmetruim LPDnodmetruim having No endometrial defect

nodmetruim having No endometrial defect

s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230 229

Achieving specificity along with sensitivity is a well-recognized problem in progesterone enzyme immunoassaydevelopment [22]. To this end, different combinations of anti-bodies and enzyme conjugates have been tested. A homol-ogous assay using a P-11-HS bridge was developed, havinga sensitivity of 0.5 pg/well and a 32% cross-reaction with11-�OH-progesterone and a 62.5% cross-reaction with 5-�-pregnane-3,20-dione [23]. Different heterologous combina-tions have also been tried. The EIA comprised of the P-11-HS antibody with the P-11-glucuronide enzyme conjugateshowed a sensitivity of 25 pg/well, with an 8% cross reac-tion with 5-�-pregnane-3,20-dione [24]. Using the P-11-HSantibody with the P-3-O-CMO enzyme conjugate, Saha andDas [9] developed a progesterone assay having a sensitivityof 3.8 pg/well, whereas using the same combination, Munroand Stabenfeldt [8] had previously developed an assay hav-ing a sensitivity 0.25 pg/well, with a 21.4% cross-reactionwith 11-�OH-progesterone, and a 62.5% cross-reaction with5-�-pregnane-3,20-dione. Hatzidakis et al. [11] tested differ-ent homologous assays (P-3-O-CMO, P-11-HS, P-6-�-HS, andP-7-�-carboxethyl-thioether bridge) and heterologous assays(P-3-O-CMO-BSA with P-6-�-HS-HRP, P-11-HS-BSA with P-3-O-CMO-HRP, P-6-�-HS-BSA with P-3-O-CMO-HRP and P-7-�-carboxethyl-thioether-BSA with P-6-�-HS-HRP). Among them,the combination of the antibody P-7-�-carboxethyl-thioether-BSA with P-6-�-HS-HRP was selected, with a sensitivity of0.25 ng/mL, and a 10% cross-reaction with 5-�-pregnane-3clmAs0thPcawpgCtg

udpwagtbsaotga

mixture of different non-related steroids is also used as dis-placing agents in direct assays of steroids like testosteroneand estradiol. These exogenous steroids are potential sourcesof cross-reaction. The complex mixtures of these displacingagents in direct immunoassay interfere with the performanceof the assay and the sensitivity is affected. In the presentstudy, a direct ELISA of progesterone was developed for themeasurement of progesterone in serum, using neither any“displacing agent” nor any special means and no measure-ment bias was noticed. This may be due to the fact thatwhile in circulation only 18% of progesterone is bound withtranscortin where as 80% is bound with serum albumin andrest is free form. As the major percentage of the progesteroneis non-specifically bound with albumin, it may be readilyavailable for antibody during immunoassay without the helpof displacer.

The present ELISA correlated well with RIA when testedwith normal and clinical subjects. The normal ranges ofthe progesterone in the different phases of the menstrualcycles were estimated. Serum samples of clinical subjectshaving histories of recurrent abortions were also assayedin the mid-luteal phase for the diagnosis of luteal phasedefects. Luteal phase defects (LPD) are defined as abnormali-ties of corpus luteum function with insufficient progesteroneproduction [25,26]. As reported in most patients with LPD,peak mid-luteal progesterone levels remain below 10 ng/mL[27]. Progesterone values of the subjects having histories

r

,20-dione. Different progesterone-4-succinimidyl-ovalbuminonjugates with varying length linkers (4, 11, and 18 atomengths) were synthesized and used in different surface plas-

on resonance-based immunoassays (SPR) and ELISAs [12].lthough the sensitivities of these SPR assays were good, theame was not true for the ELISAs, (sensitivity was greater than.5 ng/mL). Thus progesterone ELISAs with both good sensi-ivity and high specificity have been difficult to design. Butere, using as an antigen heterologous combination of the-3-O-CMO antibody with the 17-�OH-P-3-CMO-ALP enzymeonjugate we have developed a sensitive (0.11 ng/mL) as wells specific assay, having a cross-reaction of less than 0.1%ith C18, C19, and C21 steroids, with the exception of 17-�OH-rogesterone (3.1%) and pregnenolone (15%). The reason ofetting better sensitive and specific assay using 17-�OH-P-3-MO as enzyme conjugate than other progesterone deriva-

ives may be due to differential kinetic interaction with pro-esterone antibody and this can be further studied.

In the present study, direct assay was developed withoutsing any steroid displacing or blocking agent. Earlier for theevelopment of direct assays of progesterone, different dis-lacing agents such as danazol, dexamethasone, and cortisolere used [13–16]. It is reported that danazol, dexamethasone

nd cortisol, which have affinities for corticosteroid bidinglobulin (CBG), have been used in direct assay for proges-erone. Other steroids, e.g. cortisol can be displaced from CBG,y protein binding agents such as 8-anilino-1-naphthaleneulphonic acid (8-ANS) and salicylate, by proteolytic enzymet low pH or by heat treatment. The disadvantage of the usef these displacing agents is that in their blocking concentra-ion, some of them reduce the specific binding of the anti-en with antibody. In some cases, these displacing agentsre also cross-react with the antibody. In several cases a

of repeated spontaneous abortions were measured by thepresent ELISA, and some of the values clearly showed the signsof luteal phase defect and, later confirmed by endometrialbiopsy.

The ELISA, presented here, then, is both sensitive and spe-cific in measuring serum progesterone concentration and itsefficacy has been validated both analytically and clinically.

Acknowledgements

The authors are grateful to HOD, RBM, Prof. K. Kalivani, andDirector, Prof. N.K. Sethi, for their support, also express sin-cere gratitude to Prof. Somnath Roy for his valuable suggestionthroughout the study. We sincerely acknowledge the due con-tribution of Dr. Pikee Saxena, Dr. Saroj Goel, Mr. H.C. Pandeyand Mr. A.K. Mazumder of Institute’s Infertility Clinic, in theirdifferent capacities. Technical support from Mrs. S. Bala is alsogratefully acknowledged.

e f e r e n c e s

[1] Miller SA, Morton MS, Turkes A. Chemiluminescenceimmunoassay for progesterone in plasma incorporatingacridinium ester labelled antigen. Ann Clin Biochem1988;25(Pt. 1):27–34.

[2] Kakabakos SE, Khosravi MJ. Direct time-resolvedfluorescence immunoassay of progesterone in seruminvolving the biotin-streptavidin system and theimmobilized-antibody approach. Clin Chem 1992;38:725–30.

[3] Choi MJ, Choi J, Yoon DY, Park J, Eremin SA. Fluorescencepolarization immunoassay of progesterone. Biol PharmBull 1997;20:309–14.

230 s t e r o i d s 7 1 ( 2 0 0 6 ) 222–230

[4] Hong JY, Choi MJ. Development of one-step fluorescencepolarization immunoassay for progesterone. Biol PharmBull 2002;25:1258–62.

[5] Dray F, Andrieu JM, Renaud F. Enzyme immunoassay ofprogesterone at the picogram level usingbeta-galactosidase as label. Biochim Biophys Acta1975;403:131–8.

[6] Joyce BG, Wilson DW, Read GF, Riad-Fahmy R. Animproved enzyme immunoassay for progesterone inhuman plasma. Clin Chem 1978;24:2099–102.

[7] Sauer MJ, Foulkes JA, Cookson AD. Directenzymeimmunoassay of progesterone in bovine milk.Steroids 1981;38:45–53.

[8] Munro C, Stabenfeldt G. Development of a microtitre plateenzyme immunoassay for the determination ofprogesterone. J Endocrinol 1984;101:41–9.

[9] Saha B, Das C. Development of a highly sensitive enzymelinked immunosorbent assay for human serumprogesterone using penicillinase. J Immunoassay1991;12:391–412.

[10] Mitsuma M, Kambegawa A, Okinaga S, Arai K. A sensitivebridge heterologous enzyme immunoassay of progesteroneusing geometrical isomers. J Steroid Biochem 1987;28:83–8.

[11] Hatzidakis G, Stefanakis A, Krambovitis E. Comparison ofdifferent antibody-conjugate derivatives for thedevelopment of a sensitive and specific progesteroneassay. J Reprod Fertil 1993;97:557–61.

[12] Wu Y, Mitchell J, Cook C, Main L. Evaluation ofprogesterone-ovalbumin conjugates with different lengthlinkers in enzyme-linked immunosorbant assay andsurface plasmon resonance-based immunoassay. Steroids

time-resolved fluoroimmunoassay of progesterone withlabeled analyte. Clin Chem 1988;34:501–4.

[17] Williams GT, Groskopf WR, Baker HN, Agdeppa DA.Determination of steroids by competitive immunoassay.US Patent (US 6,201,141). Assignee: Abbott Park, IL: AbbottLaboratories; 2001.

[18] Basu A, Shrivastav TG, Kariya KP. Preparations of enzymeconjugate through adipic acid dihydrazide as linker andits use in immunoassays. Clin Chem 2003;49:1410–2.

[19] Shrivastav TG, Basu A, Kariya KP. Substitution ofcarbonate buffer by water for IgG immobilization inenzyme linked immunosorbent assay. J ImmunoassayImmunochem 2003;24:191–203.

[20] Rodbard D. Satistical quality control and routine dataprocessing for radioimmunoassays andimmunoradiometric assays. Clin Chem 1974;20:1255–70.

[21] Feldman H, Rodbard D. In: Odell WD, Daugaday WM,editors. Principles of competitive protein binding assay.Philadelphia: J.B. Lipincott; 1971. p. 158.

[22] Grover PK, Odell WD. Specificity of antisera to sexsteroids. I. The effect of substitution and stereochemistry.J Steroid Biochem 1977;8:121–6.

[23] Wang HX, Stott RA, Thorpe GH, Kricka LJ, Holder G, RuddBT. An enhanced chemiluminescence enzymeimmunoassay for serum progesterone. Steroids1984;44:317–28.

[24] Sauer MJ, Foulkes JA, Worsfold A, Morris BA. Use ofprogesterone 11-glucuronide-alkaline phosphataseconjugate in a sensitive microtiter-plateenzymeimmunoassay of progesterone in milk and itsapplication to pregnancy testing in dairy cattle. J Reprod

2002;67:565–72.[13] Ratcliffe WA, Corrie JE, Dalziel AH, Macpherson JS. Direct

125I-radioligand assays for serum progesterone comparedwith assays involving extraction of serum. Clin Chem1982;28:1314–8.

[14] Carriere PD, Lee B. Direct radioimmunoassay ofprogesterone in bovine plasma using danazol(17-alpha-2,4-pregnadien-20-yno(2,3-d)isoxazol-17-ol) as adisplacing agent. Can J Vet Res 1994;58:230–3.

[15] Barnard G, Osher J, Lichter S, Gayer B, De Boever J, LimorR, et al. The measurement of progesterone in serum by anon-competitive idiometric assay. Steroids 1995;60:824–9.

[16] Dechaud H, Bador R, Claustrat F, Desuzinges C, Mallein R.New approach to competitive lanthanide immunoassay:

Fertil 1986;76:375–91.[25] Jones GES, Madrigal-Castro V. Hormonal findings in

association with abnormal corpus luteum function in thehuman: the luteal phase defect. Fertil Steril 1970;21:1–13.

[26] Soules MR, McLahlan RI, Ek M, et al. Luteal phasedeficiency: characterization of reproductive hormones overthe menustrual cycle. J Clin Endocrinol Metab1989;69:804–12.

[27] Hull MGR, Savage PE, Bromhan DR, et al. The value of asingle serum progesterone measurement in themid-leuteal phase as a criterion of a potentially fertilecycle (‘ovulation’) derived from treated and untreatedconception cycles. Fertil Steril 1982;37:355–60.