PRODUCTION OF ANTISERA AGAINST HIGHLY PURIFIED HUMAN FOLLICLE-STIMULATING HORMONE, LUTEINIZING
HORMONE AND THYROID-STIMULATING HORMONE J. I. THORELL AND B. HOLMSTR\l=O"\M Department of Nuclear Medicine, University of Lund at Malmö General Hospital, S-214 01 Malmö and * Research Department AB KABI Diagnostica, S-104 25 Stockholm, Sweden
(Received 15 September 1975) SUMMARY
Antisera were produced in rabbits against highly purified preparations of human LH (2000 or 10000 i.u./mg), human FSH (5500 i.u./mg), and human TSH (7\m=.\5 i.u./mg). Most rabbits produced antisera of high titre and high avidity. Cross-reactions were minimal between human TSH and human chorionic gonadotrophin (HCG) and between human FSH and HCG but marked between human LH and HCG. TSH and FSH also showed a constant but relatively weak cross-reaction. LH cross-reacted with FSH to a higher degree than did HCG. The avidity of the antisera was high. It was concluded that much of the lack of specificity recorded for glycoprotein antisera are effects of impure immunogens. Some of the true cross-reactions are probably explained by shared antigenic determinants of the \g=b\ x=req-\ subunits. Unadsorbed antisera could be used for assay of FSH and TSH in plasma from
pregnant
women.
introduction
Production of antisera suitable for the radioimmunoassay of human glycoprotein hormones has been difficult (Franchimont, 1968; Odell, Abraham, Raud, Swerdloff & Fisher, 1969). A major problem is lack of specificity, with cross-reactions between all the glycoprotein hormones. This cross-reactivity has been ascribed to common antigenic determinants since all glycoprotein hormones share an a-subunit with almost identical structure (Shome & Parlow, 1974; Morgan, Birken & Canfield, 1975; Rathnam & Saxena, 1975). Impurities in the preparations of native hormones used for immunization may also contribute to antisera non-specificity (Odell & Hescox, 1971; Schlaff, Rosen & Roth, 1968). Therefore, it is uncertain whether the lack of specificity is due to structural similarities between the various hormones or the presence of contaminating pituitary material. If the common a-subunit were the only reason for the cross-reactions, the obvious solution to these problems should be immunization with isolated /?-subunits, since the differences in structure between the various human hormones is mainly located within this subunit. Significantly improved specificity has been reported when isolated /^-subunits of folliclestimulating hormone (FSH), luteinizing hormone (LH), thyroid-stimulating hormone (TSH) and human chorionic gonadotrophin (HCG) were used as immunogens (Vaitukaitis, Ross, Reichert & Ward, 1972; Vaitukaitis, Ross, Pierce, Cornell & Reichert, 1973; Kourides, Weintraub, Leuko & Maloof, 1974). However, many of the antisera to /?-subunits still showed considerable cross-reactions and the antisera had to be carefully selected to achieve
assays with optimal specificity. Also /?-subunits were less immunogenic than native hor¬ mones; the antisera produced had generally low titres and avidities. Fractionation of large quantities of freshly frozen human pituitary glands for the in¬ dustrial production of therapeutic growth hormone (GH) (Holmström & Fhölenhag, 1975) made available large by-fractions containing the glycoprotein hormones. From this material enough highly purified native hormones were isolated to permit immunization of several groups of animals. As reported here, many of the antisera produced were useful for specific and sensitive radioimmunoassays. materials and methods
The preparations used for immunization were prepared in our laboratories from whole fresh frozen human pituitaries according to the method described by Roos (1968). The pituitaries were extracted at pH 6-2. GH and TSH were precipitated by addition of an equal volume of saturated ammonium sulphate. After centrifugation, the ammonium sulphate concentration was increased to 70 % saturation, the precipitate contained the main part of gonadotrophin. FSH and LH were separated by chromatography on DEAE-cellulose when the mixture after dialysis was applied at low ionic strength. LH passed through the column and FSH was adsorbed. Increased ionic strength eluted an FSH fraction with low content of LH. LH
was further purified by gel filtration on Bio-gel P-150 and Sephadex G-100 after elimination of haemoglobin. This preparation had a potency of 0-1 mg LH/mg protein. It was used for immunization of some of the rabbits as specified in 'Results'. Refiltration on Sephadex G-100 increased the specific activity to about 0-5 mg LH/mg protein (approx. 10000 i.u./mg) and this preparation was used for immunization of another group of rabbits. The FSH eluate from the DEAE-cellulose column was gel filtered twice on a 200 cm Sephadex G-100 column resulting in a preparation with a specific activity of not less than 0-4 mg FSH/mg protein (approx. 5 500 i.u./mg). This material was used for im¬ munizations. The activities of these gonadotrophin preparations were kindly assayed by L. Wide (Uppsala), according to the method described by Wide, Nillius, Gemzell & Roos (1973). Various batches, prepared according to this procedure (Kabi MfAPl and Kabi-1765) were also used as standards and for labelling in the radioimmunoassay. TSH precipitated together with GH from the original extract of the pituitaries was purified according to the method described by Roos, Jacobson & Wide (1975). Chromato¬ graphy of the dissolved ammonium sulphate precipitate on Sephadex G-100 gave three protein peaks. The second peak contained the main portion of the TSH. This fraction was further purified on Bio-gel P-150, by hydroxyapatite chromatography and finally SPSephadex C-50 chromatography. This last purification step reduced the GH content from 18 to about 1 %. This TSH preparation, Kabi MfL 035, had a specific activity of about 7-5 i.u./mg and was used both for immunization and for labelling in the radioimmunoassay. All immunizations were performed by intradermal injections according to the technique described by Vaitukaitis, Robbins, Nieschlag & Ross (1971). Antigen was dissolved in saline to a concentration of 50-60 pg protein/ml. Four millilitres of solution were drawn up in one 10 ml disposable polypropylene syringe (A-sik, Denmark). Freund's complete adjuvant (4 ml; Difco, Detroit) was drawn into another 10 ml syringe and the tips of the two syringes were connected by a 20 mm long silicone rubber tube (diameter 2 mm). An emulsion was formed when the antigen solution was forced through the tubing into the syringe containing adjuvant. By repeatedly transferring the suspension from one syringe to the other a very stable suspension was obtained. Two millilitres of suspension were administered in 40 separate
sites on the back of each rabbit. Most animals were only injected once. Booster injec¬ tions were made using the same technique. One millilitre of Pertussis vaccine (State Bac¬ teriological Laboratories, Stockholm) was given subcutaneously to some groups of animals as stated in the 'Results'. Blood samples were collected from the central ear vein of the animals, and were allowed to clot at room temperature. Sera were separated by centrifuga¬ tion and stored at 20 °C. -
Assay of antiserum
All studies, except in cross-reaction determinations, were performed using homologous systems, e.g. 125I-labelled FSH served as radioligand in assaying FSH antiserum and com¬ petitive inhibition was studied with FSH. Labelling with125I was performed using lactoperoxidase as described previously by Thorell & Johansson (1971). A double-antibody assay was used according to the following protocol: 250/d of diluted antiserum were incubated with 50µ (008 ng) 125I-labelled antigen at 4 °C. Barbitone buffer (0075 mol/1) at pH 8-5, con¬ taining 0-25 % bovine serum albumin (Armour), was used as diluent for all reagents. After 24 h, 50 µ of 1:10 diluted goat antiserum against rabbit immunoglobulin were added, together with 50/ri normal rabbit serum, diluted 1:250. Since this precipitation procedure was optimized for assay of plasma, 100 µ of pooled human plasma were added just before addition of the goat anti-rabbit serum. The tubes were then incubated for 20 h at 4 °C, after which they were centrifuged at 4000 g for 15 min. The supernatants were decanted and the precipitates counted in a well-type scintillation detector. Specificity and avidity studies were performed using an identical protocol except that 100 µ of unlabelled standards were added. In order to assay hormone concentrations of human plasma samples, 100 µ of plasma (with EDTA as anticoagulant) were added in place of the standard. Studies of the effect of 'antigen-free plasma' (Thorell, 1968) showed that the first antibody reaction was not non-specifically influenced by the presence of plasma. At the second antibody step, 100 µ of diluent were added to the tubes containing plasma samples, in place of the plasma added to the standard tubes. The following materials were used for labelling (potencies of gonadotrophins expressed in terms of 2nd IRP-HMG): FSH, Kabi MfAPl (radioimmuno¬ assay 5200i.u./mg); LH, LER960; TSH, the preparation distributed by the National Pituitary Agency (NPA) for radioimmunoassay. The following materials were used as standards: FSH, Kabi-1765 (radioimmunoassay 5600 i.u./mg); LH, Division of Biological Standards 68/40; TSH, Kabi MfL035 (radioimmunoassay 7-5 i.u./mg against MRC Stand¬ ard A); HCG, Leo X3-42 (bioassay 10000 i.u./mg; obtained from AB Leo, Hälsingborg, Sweden); GH, Kabi DqP 11 (radioimmunoassay 2 i.u./mg against Division of Biological Standards 66/217). The following antisera, kindly supplied by the NPA were used for comparison of titre and avidity estimates: rabbit anti-hTSH and rabbit anti-hFSH batch 3. Titres were expressed as the final dilution of antiserum which bound 50 % of the labelled hormone, where 100 % was the binding to large antibody excess. Specificity was measured at dilutions of antisera which gave the most sensitive assay. All results were calculated by logit transformation of bound radioactivity versus log dose of antigen (Rodbard, Bridson & Rayford, 1969). The degree of cross-reaction was quantified from the amount (in ng) of the antigen which caused 50 % inhibition of the binding of the radioligand. It was expressed as the ratio between the amounts of homologous and hetero¬ logous antigens that caused this inhibition. Avidity was calculated from the Scatchard plot of the most sensitive assay for the homo¬ logous antigen and expressed as association constants.
RESULTS
Titres and avidity antibodies all three hormones, usually with high titres after only against produced one injection (Figs. 1-3). Peak response occurred at 2-3 months. There was marked varia¬ tion in response between individual rabbits. Titres remained low in some animals, while others which responded with high titres to the first injection continued to respond to re¬ peated injections. Booster injections after 4-8 months prolonged the antibody response for over 1 year. Of the three immunogens tested, TSH seemed most potent. TSH induced higher titres in a greater proportion of animals. The injection of Pertussis bacteria did not increase antibody production tested with LH (Fig. 2). The titres were somewhat higher in the rabbits which received the most purified LH (0-1 mg/mg) than in those given the less Animals
pure
preparation (0-5 mg/mg).
The avidity of the antiserum varied also between the animals, association constant rang¬ ing from 109 to 2 IO10 (Table 1). The association constant tended to increase after pro¬ longed immunization.
1. Human FSH antisera titres in rabbits immunized with 50 pg FSH intradermally. 125ILabelled human FSH was used as radioligand. One rabbit was given a booster injection (/ /). All animals were given Pertussis vaccine. NPA-AS denotes the titre of FSH antiserum provided by the National Pituitary Agency.
Fig.
Specificity Figures 4-6 illustrate the influence of various glycoprotein hormones and GH ing of 125I-labelled antigen to its homologous antiserum.
on
the bind¬
FSH antiserum There was no detectable cross-reactivity between FSH and GH or HCG except for a slight inhibition with large amounts of HCG in one of the low titre sera. Because of the limited availability of large amounts of LH and the known close relation between LH and HCG,
10
12
0 Months Fig. 2. Human LH antisera titres in rabbits immunized with 50 or 60/*g LH intradermally. (a) Rabbits given Pertussis vaccine at the primary immunization and (b) rabbits which did not receive Pertussis vaccine. Encircled points (O) denote animals injected with the most purified LH (10000 i.u./mg). Some animals were given booster injections (/ /). 125I-Labelled human LH was used as
radioligand.
«
104 -
Months
3. Human TSH antisera titres in rabbits immunized with 50 /¿g TSH intradermally. All animals given Pertussis vaccine. NPA-AS denotes the titre of TSH antiserum provided by the National Pituitary Agency. One animal was given a booster injection (/ /). mI-labelled human TSH was used
Fig.
were as
radioligand.
(Ka)for the study of homologous labelled antigen during early and prolonged immunization
Table 1. Association
constants
Prolonged immunization
Early immunization Antigen rabbit
Interval
no.
(weeks)
FSH
59 60 61 63
NPA-antiserum* LH
55 56 57 58
79t 80t TSH
64 65 66 67
NPA-antiserum*
7 7 7 7
1/Dilution
Interval
1/Dilution
of antisera
,, 9
(weeks)
of antisera
/Caxl0s
9000 120000 75000 18000 20000
0-9 2 5 1 2
29 37 15 24
3000 75000 100000 15000
10 8 2
21 21 31
6000 5000 6000 9000
20 5 6 7
27 19 19 19
240000 240000 800000 60000
5 3 20
18000 17000 36000 30000 20000 5000 120000 170000 200000 100000 90000
10 1 7 7 5 4 1 2 2 2 3
36
7
* Antiserum provided by the National Pituitary Agency. t Rabbits immunized with purest LH (10000 i.u./mg).
cross-reactivity against LH was tested in only a few instances. A slight inhibition of a nonparallel type was observed (Fig. 4). For two antisera, the inhibition by LH became more marked in samples collected late during immunization. FSH/LH ratios of 0-5-0-2 were recorded in these sera. TSH showed a non-parallel inhibition for all antisera with FSH/TSH ratios of 0 5-0 2. This effect did not change with the purity of the TSH preparation. In two rabbits, this ratio increased to 0-5-0-2 with prolonged immunization while it remained
around 005 for the others. GH did not cross-react with any antisera. One of the antisera was used in an FSH radioimmunoassay. The mean FSH value mea¬ sured during various stages of pregnancy was 1-6 ±0-4 (s.d.) ng/ml which is significantly lower than the average basal level of 2-6 ± 0-2 ng/ml found in normal fertile women. LH antisera The cross-reactivities of LH antisera with FSH were low and non-parallel (Fig. 5). HCG inhibited radioligand binding in all sera, usually non-parallel to LH and with an LH/ HCG ratio of 0-5-0-2. In two early sera, the HCG inhibition was less prominent with a ratio of 0-02-0-01. TSH caused varying non-parallel inhibition in all antisera, with LH/TSH ratios of 0-2-0 5. GH did not inhibit the binding of LH in any antiserum. TSH-antisera The inhibition caused by HCG in the TSH-system (Fig. 6) was minimal and occurred only at concentrations of 1000 ng HCG or more. FSH showed some non-parallel inhibition. The TSH/FSH ratio was generally below 0001. One of the antisera was used for routine assays of TSH. The levels found in plasma from pregnant and non-pregnant women did not differ (1-2 ± 0-6 µ ./ml and 1-5 ± 0-7 /¿u./ml respectively; mean + s.d.). The concentration of TSH was below the limit of detection (0-5 /
HORMONE AND THYROID-STIMULATING HORMONE J. I. THORELL AND B. HOLMSTR\l=O"\M Department of Nuclear Medicine, University of Lund at Malmö General Hospital, S-214 01 Malmö and * Research Department AB KABI Diagnostica, S-104 25 Stockholm, Sweden
(Received 15 September 1975) SUMMARY
Antisera were produced in rabbits against highly purified preparations of human LH (2000 or 10000 i.u./mg), human FSH (5500 i.u./mg), and human TSH (7\m=.\5 i.u./mg). Most rabbits produced antisera of high titre and high avidity. Cross-reactions were minimal between human TSH and human chorionic gonadotrophin (HCG) and between human FSH and HCG but marked between human LH and HCG. TSH and FSH also showed a constant but relatively weak cross-reaction. LH cross-reacted with FSH to a higher degree than did HCG. The avidity of the antisera was high. It was concluded that much of the lack of specificity recorded for glycoprotein antisera are effects of impure immunogens. Some of the true cross-reactions are probably explained by shared antigenic determinants of the \g=b\ x=req-\ subunits. Unadsorbed antisera could be used for assay of FSH and TSH in plasma from
pregnant
women.
introduction
Production of antisera suitable for the radioimmunoassay of human glycoprotein hormones has been difficult (Franchimont, 1968; Odell, Abraham, Raud, Swerdloff & Fisher, 1969). A major problem is lack of specificity, with cross-reactions between all the glycoprotein hormones. This cross-reactivity has been ascribed to common antigenic determinants since all glycoprotein hormones share an a-subunit with almost identical structure (Shome & Parlow, 1974; Morgan, Birken & Canfield, 1975; Rathnam & Saxena, 1975). Impurities in the preparations of native hormones used for immunization may also contribute to antisera non-specificity (Odell & Hescox, 1971; Schlaff, Rosen & Roth, 1968). Therefore, it is uncertain whether the lack of specificity is due to structural similarities between the various hormones or the presence of contaminating pituitary material. If the common a-subunit were the only reason for the cross-reactions, the obvious solution to these problems should be immunization with isolated /?-subunits, since the differences in structure between the various human hormones is mainly located within this subunit. Significantly improved specificity has been reported when isolated /^-subunits of folliclestimulating hormone (FSH), luteinizing hormone (LH), thyroid-stimulating hormone (TSH) and human chorionic gonadotrophin (HCG) were used as immunogens (Vaitukaitis, Ross, Reichert & Ward, 1972; Vaitukaitis, Ross, Pierce, Cornell & Reichert, 1973; Kourides, Weintraub, Leuko & Maloof, 1974). However, many of the antisera to /?-subunits still showed considerable cross-reactions and the antisera had to be carefully selected to achieve
assays with optimal specificity. Also /?-subunits were less immunogenic than native hor¬ mones; the antisera produced had generally low titres and avidities. Fractionation of large quantities of freshly frozen human pituitary glands for the in¬ dustrial production of therapeutic growth hormone (GH) (Holmström & Fhölenhag, 1975) made available large by-fractions containing the glycoprotein hormones. From this material enough highly purified native hormones were isolated to permit immunization of several groups of animals. As reported here, many of the antisera produced were useful for specific and sensitive radioimmunoassays. materials and methods
The preparations used for immunization were prepared in our laboratories from whole fresh frozen human pituitaries according to the method described by Roos (1968). The pituitaries were extracted at pH 6-2. GH and TSH were precipitated by addition of an equal volume of saturated ammonium sulphate. After centrifugation, the ammonium sulphate concentration was increased to 70 % saturation, the precipitate contained the main part of gonadotrophin. FSH and LH were separated by chromatography on DEAE-cellulose when the mixture after dialysis was applied at low ionic strength. LH passed through the column and FSH was adsorbed. Increased ionic strength eluted an FSH fraction with low content of LH. LH
was further purified by gel filtration on Bio-gel P-150 and Sephadex G-100 after elimination of haemoglobin. This preparation had a potency of 0-1 mg LH/mg protein. It was used for immunization of some of the rabbits as specified in 'Results'. Refiltration on Sephadex G-100 increased the specific activity to about 0-5 mg LH/mg protein (approx. 10000 i.u./mg) and this preparation was used for immunization of another group of rabbits. The FSH eluate from the DEAE-cellulose column was gel filtered twice on a 200 cm Sephadex G-100 column resulting in a preparation with a specific activity of not less than 0-4 mg FSH/mg protein (approx. 5 500 i.u./mg). This material was used for im¬ munizations. The activities of these gonadotrophin preparations were kindly assayed by L. Wide (Uppsala), according to the method described by Wide, Nillius, Gemzell & Roos (1973). Various batches, prepared according to this procedure (Kabi MfAPl and Kabi-1765) were also used as standards and for labelling in the radioimmunoassay. TSH precipitated together with GH from the original extract of the pituitaries was purified according to the method described by Roos, Jacobson & Wide (1975). Chromato¬ graphy of the dissolved ammonium sulphate precipitate on Sephadex G-100 gave three protein peaks. The second peak contained the main portion of the TSH. This fraction was further purified on Bio-gel P-150, by hydroxyapatite chromatography and finally SPSephadex C-50 chromatography. This last purification step reduced the GH content from 18 to about 1 %. This TSH preparation, Kabi MfL 035, had a specific activity of about 7-5 i.u./mg and was used both for immunization and for labelling in the radioimmunoassay. All immunizations were performed by intradermal injections according to the technique described by Vaitukaitis, Robbins, Nieschlag & Ross (1971). Antigen was dissolved in saline to a concentration of 50-60 pg protein/ml. Four millilitres of solution were drawn up in one 10 ml disposable polypropylene syringe (A-sik, Denmark). Freund's complete adjuvant (4 ml; Difco, Detroit) was drawn into another 10 ml syringe and the tips of the two syringes were connected by a 20 mm long silicone rubber tube (diameter 2 mm). An emulsion was formed when the antigen solution was forced through the tubing into the syringe containing adjuvant. By repeatedly transferring the suspension from one syringe to the other a very stable suspension was obtained. Two millilitres of suspension were administered in 40 separate
sites on the back of each rabbit. Most animals were only injected once. Booster injec¬ tions were made using the same technique. One millilitre of Pertussis vaccine (State Bac¬ teriological Laboratories, Stockholm) was given subcutaneously to some groups of animals as stated in the 'Results'. Blood samples were collected from the central ear vein of the animals, and were allowed to clot at room temperature. Sera were separated by centrifuga¬ tion and stored at 20 °C. -
Assay of antiserum
All studies, except in cross-reaction determinations, were performed using homologous systems, e.g. 125I-labelled FSH served as radioligand in assaying FSH antiserum and com¬ petitive inhibition was studied with FSH. Labelling with125I was performed using lactoperoxidase as described previously by Thorell & Johansson (1971). A double-antibody assay was used according to the following protocol: 250/d of diluted antiserum were incubated with 50µ (008 ng) 125I-labelled antigen at 4 °C. Barbitone buffer (0075 mol/1) at pH 8-5, con¬ taining 0-25 % bovine serum albumin (Armour), was used as diluent for all reagents. After 24 h, 50 µ of 1:10 diluted goat antiserum against rabbit immunoglobulin were added, together with 50/ri normal rabbit serum, diluted 1:250. Since this precipitation procedure was optimized for assay of plasma, 100 µ of pooled human plasma were added just before addition of the goat anti-rabbit serum. The tubes were then incubated for 20 h at 4 °C, after which they were centrifuged at 4000 g for 15 min. The supernatants were decanted and the precipitates counted in a well-type scintillation detector. Specificity and avidity studies were performed using an identical protocol except that 100 µ of unlabelled standards were added. In order to assay hormone concentrations of human plasma samples, 100 µ of plasma (with EDTA as anticoagulant) were added in place of the standard. Studies of the effect of 'antigen-free plasma' (Thorell, 1968) showed that the first antibody reaction was not non-specifically influenced by the presence of plasma. At the second antibody step, 100 µ of diluent were added to the tubes containing plasma samples, in place of the plasma added to the standard tubes. The following materials were used for labelling (potencies of gonadotrophins expressed in terms of 2nd IRP-HMG): FSH, Kabi MfAPl (radioimmuno¬ assay 5200i.u./mg); LH, LER960; TSH, the preparation distributed by the National Pituitary Agency (NPA) for radioimmunoassay. The following materials were used as standards: FSH, Kabi-1765 (radioimmunoassay 5600 i.u./mg); LH, Division of Biological Standards 68/40; TSH, Kabi MfL035 (radioimmunoassay 7-5 i.u./mg against MRC Stand¬ ard A); HCG, Leo X3-42 (bioassay 10000 i.u./mg; obtained from AB Leo, Hälsingborg, Sweden); GH, Kabi DqP 11 (radioimmunoassay 2 i.u./mg against Division of Biological Standards 66/217). The following antisera, kindly supplied by the NPA were used for comparison of titre and avidity estimates: rabbit anti-hTSH and rabbit anti-hFSH batch 3. Titres were expressed as the final dilution of antiserum which bound 50 % of the labelled hormone, where 100 % was the binding to large antibody excess. Specificity was measured at dilutions of antisera which gave the most sensitive assay. All results were calculated by logit transformation of bound radioactivity versus log dose of antigen (Rodbard, Bridson & Rayford, 1969). The degree of cross-reaction was quantified from the amount (in ng) of the antigen which caused 50 % inhibition of the binding of the radioligand. It was expressed as the ratio between the amounts of homologous and hetero¬ logous antigens that caused this inhibition. Avidity was calculated from the Scatchard plot of the most sensitive assay for the homo¬ logous antigen and expressed as association constants.
RESULTS
Titres and avidity antibodies all three hormones, usually with high titres after only against produced one injection (Figs. 1-3). Peak response occurred at 2-3 months. There was marked varia¬ tion in response between individual rabbits. Titres remained low in some animals, while others which responded with high titres to the first injection continued to respond to re¬ peated injections. Booster injections after 4-8 months prolonged the antibody response for over 1 year. Of the three immunogens tested, TSH seemed most potent. TSH induced higher titres in a greater proportion of animals. The injection of Pertussis bacteria did not increase antibody production tested with LH (Fig. 2). The titres were somewhat higher in the rabbits which received the most purified LH (0-1 mg/mg) than in those given the less Animals
pure
preparation (0-5 mg/mg).
The avidity of the antiserum varied also between the animals, association constant rang¬ ing from 109 to 2 IO10 (Table 1). The association constant tended to increase after pro¬ longed immunization.
1. Human FSH antisera titres in rabbits immunized with 50 pg FSH intradermally. 125ILabelled human FSH was used as radioligand. One rabbit was given a booster injection (/ /). All animals were given Pertussis vaccine. NPA-AS denotes the titre of FSH antiserum provided by the National Pituitary Agency.
Fig.
Specificity Figures 4-6 illustrate the influence of various glycoprotein hormones and GH ing of 125I-labelled antigen to its homologous antiserum.
on
the bind¬
FSH antiserum There was no detectable cross-reactivity between FSH and GH or HCG except for a slight inhibition with large amounts of HCG in one of the low titre sera. Because of the limited availability of large amounts of LH and the known close relation between LH and HCG,
10
12
0 Months Fig. 2. Human LH antisera titres in rabbits immunized with 50 or 60/*g LH intradermally. (a) Rabbits given Pertussis vaccine at the primary immunization and (b) rabbits which did not receive Pertussis vaccine. Encircled points (O) denote animals injected with the most purified LH (10000 i.u./mg). Some animals were given booster injections (/ /). 125I-Labelled human LH was used as
radioligand.
«
104 -
Months
3. Human TSH antisera titres in rabbits immunized with 50 /¿g TSH intradermally. All animals given Pertussis vaccine. NPA-AS denotes the titre of TSH antiserum provided by the National Pituitary Agency. One animal was given a booster injection (/ /). mI-labelled human TSH was used
Fig.
were as
radioligand.
(Ka)for the study of homologous labelled antigen during early and prolonged immunization
Table 1. Association
constants
Prolonged immunization
Early immunization Antigen rabbit
Interval
no.
(weeks)
FSH
59 60 61 63
NPA-antiserum* LH
55 56 57 58
79t 80t TSH
64 65 66 67
NPA-antiserum*
7 7 7 7
1/Dilution
Interval
1/Dilution
of antisera
,, 9
(weeks)
of antisera
/Caxl0s
9000 120000 75000 18000 20000
0-9 2 5 1 2
29 37 15 24
3000 75000 100000 15000
10 8 2
21 21 31
6000 5000 6000 9000
20 5 6 7
27 19 19 19
240000 240000 800000 60000
5 3 20
18000 17000 36000 30000 20000 5000 120000 170000 200000 100000 90000
10 1 7 7 5 4 1 2 2 2 3
36
7
* Antiserum provided by the National Pituitary Agency. t Rabbits immunized with purest LH (10000 i.u./mg).
cross-reactivity against LH was tested in only a few instances. A slight inhibition of a nonparallel type was observed (Fig. 4). For two antisera, the inhibition by LH became more marked in samples collected late during immunization. FSH/LH ratios of 0-5-0-2 were recorded in these sera. TSH showed a non-parallel inhibition for all antisera with FSH/TSH ratios of 0 5-0 2. This effect did not change with the purity of the TSH preparation. In two rabbits, this ratio increased to 0-5-0-2 with prolonged immunization while it remained
around 005 for the others. GH did not cross-react with any antisera. One of the antisera was used in an FSH radioimmunoassay. The mean FSH value mea¬ sured during various stages of pregnancy was 1-6 ±0-4 (s.d.) ng/ml which is significantly lower than the average basal level of 2-6 ± 0-2 ng/ml found in normal fertile women. LH antisera The cross-reactivities of LH antisera with FSH were low and non-parallel (Fig. 5). HCG inhibited radioligand binding in all sera, usually non-parallel to LH and with an LH/ HCG ratio of 0-5-0-2. In two early sera, the HCG inhibition was less prominent with a ratio of 0-02-0-01. TSH caused varying non-parallel inhibition in all antisera, with LH/TSH ratios of 0-2-0 5. GH did not inhibit the binding of LH in any antiserum. TSH-antisera The inhibition caused by HCG in the TSH-system (Fig. 6) was minimal and occurred only at concentrations of 1000 ng HCG or more. FSH showed some non-parallel inhibition. The TSH/FSH ratio was generally below 0001. One of the antisera was used for routine assays of TSH. The levels found in plasma from pregnant and non-pregnant women did not differ (1-2 ± 0-6 µ ./ml and 1-5 ± 0-7 /¿u./ml respectively; mean + s.d.). The concentration of TSH was below the limit of detection (0-5 /
