387
Clinica Chimica Acta, 69 (1976) 387-396 0 Elsevier Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
CCA 1656
SOLID-PHASE,
MAGNETIC PARTICLE
RADIOIMMUNOASSAY
LYNN NYE 8, *, G.C. FORREST C, HELENA GREENWOOD a, JAQUELINE GARDNER C, R. JAY C, J.R. ROBERTS b and J. LANDON a
S.
Departments of a Chemical Pathology, and b Medical Electronics, St. Bartholomew’s Hospital, London E.C. 1 and C Technicon Methods and Standards Laboratory, London E.C. 1 (U.K.) (Received
November
11,1975)
Summary A solid-phase radioimmunoassay system has been developed based on the use of antibodies covalently linked to polymer-coated iron oxide (EnzacrylR). An electro-magnet is employed both to mix the particles during incubation (by switching the field on and off) and to separate the antibody-bound and free fractions. This obviates the need for vertical rotation and for the time-consuming, multiple centrifugations required with conventional solid phase procedures. The system is universally applicable and methods have been established for the assay of thyroxine, human placental lactogen and digoxin. The thyroxine assay was employed as a model and it was shown that the results obtained for serum samples correlated closely with those using a routine liquid-phase radioimmunoassay. The applicability of employing a second antibody linked to the iron oxide particles was also studied.
Introduction A variety of techniques have been employed in radioimmunoassay to separate the antibody-bound and free fractions [l] . These include adsorption procedures (such as the addition of charcoal to adsorb the free fraction), fractional precipitation of the bound fraction (with, for example, organic solvents or neutral salts), precipitation of the bound fraction by addition of a second antibody (double-antibody procedures) and solid-phase methods. Double-antibody and solid-phase systems have the advantage of universal applicability to all radioimmunoassays, and assays based on the use of antibodies covalently linked to * Requests
for reprints
should
be addressed
to Lynn
Nye.
388
ultrafine Sephadex can be operated with a precision superior to that of most other methods [2]. However, solid-phase systems of this type have two major disadvantages, namely, the need to cap tubes and mix their contents by vertical rotation throughout the period of incubation and the multiple centrifugation and washing steps required during separation, in order to reduce non-specific blank values to a minimum and obtain reproducible results. Recently, a radioimmunoassay has been described for digoxin by Hersh and Yaverbaum [ 31, employing siliconised iron oxide as a solid-phase support. The particles were mixed manually, every 5 min, and separation of the antibodybound and free fractions achieved by sedimentation of the particles with a ‘U’ shaped magnet and tipping off the supernatant fluid containing the free fraction, which was counted. The present paper describes a magnetic particle radioimmunoassay which also overcomes the problem of mixing during the incubation period as well as facilitating the separation of the antibody-bound and free fractions. Materials and methods
Thyroxine (TJ. Total Tq levels were determined in non-extracted serum by the method of Nye et al. [4], in which 8-anilino-1-naphthalene sulphonic acid (ANS) is used to block the binding sites on endogenous thyroxine-binding globulin (TBG) and polyethylene glycol (PEG) is added to separate the bound and free fractions. Human placental luctogen (HPL). Serum HPL levels were determined by the method of Letchworth et al. [ 51, in which ethanol is added to separate the bound and free fractions. IJigoxin. Serum digoxin levels were determined by the method of Greenwood et al. 163, in which dextran coated charcoal is added to separate the bound and free fractions. Liquid-phase double-antibody separation A modification of the method described by McNe~y and Hagen [‘7] for the radioimmunoassay of human chorionic gonadotrophin (HCG) was employed to study double antibody separation. 500 ~1 of phosphate buffer containing 2 percent horse serum was added to a series of tubes, followed by 50 ~1 of 12’1labelled HCG and 50 ~1 of rabbit anti-HCG serum, which gave a final antiserum dilution (1 : 48 000). At this concentration approx~ately 50 percent of the labelled hormone was bound. A series of ‘blank’ tubes was set up containing phosphate buffer in place of the antiserum. The tubes were mixed and then incubated at room temperature for 5 h. Doubling dilutions of donkey anti-rabbit serum were prepared and loo-p1 amounts added to all the tubes (in duplicate) to give final dilutions ranging from 1 : 25 to 1 : 400. Ten ~1 of normal non-immunised rabbit serum, diluted in buffer, was added as a carrier, giving a final concentration of 1 : 1000. The tubes were incubated overnight at room temperature and the antibody-bound and free fractions separated by centrifugation at 1000 X g for 30 min, followed by aspiration of the supernatant.
389
Ferromagnetic particle radioimmunoassays The electromagnet. This comprises a power supply which was constructed from a heavy duty mains transformer with secondary voltages of 25 V and 50 V root mean square. These voltages are full wave rectified and the resulting unstabilised DC voltage is fed to a magnet via a switching mechanism. Initially a Technicon cam timer was employed to activate a micro-switch connected to the power supply which it switched on and off to produce a pulsed magnetic field. Later an electronic timer was substituted, which operates a relay in the primary circuit of the transformer, producing a similar pulsed magnetic field. The magnet consists of copper wire windings surrounding a mild steel core (2; X 2$ X 5 inches) and is attached to a specially constructed tube rack (Figs. 1 and 2). The rack which can hold 100 tubes, has a 0.5 inch mild steel base plate machined in a regular pattern of ridges and troughs, running parallel to one another across the plate. The ridges are 0.25 inches wide with gaps of 0.375 inches separating them to form troughs, 0.25 inches deep. The tubes are supported by ‘0’ ring clamps trapped between two aluminium sheets, and can be held in one of two positions, either on top of or suspended between the ridges. The entire system is available from ILS, 99 New Cavendish Street, London, WlM 7FQ. Preparation of ferromagnetic antibody particles. Gamma globulin fractions of sheep anti-T4, goat anti-HPL, rabbit anti-digoxin and donkey anti-rabbit sera were prepared by fractional precipitation with sodium sulphate. To 1 ml of
Fig. 1. The magnet
and rack which were specially
designed
for magnetic
particle
radioimmunoassays.
Al SUPPORT
SHEET
Fig. 2. An isometric
projection
of the magnet
and rack.
antiserum was added 180 mg anhydrous sodium sulphate, while mixing on a vortex mixer. The resulting suspension was mixed by vertical rotation at room temperature for 30 min, centrifuged at 1000 X g for 15 min and the supernatant fluid aspirated. The precipitate was washed twice with 18 percent sodium sulphate then dissolved in 0.9 percent saline and made up to 1 ml. The ferromagnetic solid-phase support was Enzacryl @ FEO-M (Koch Light Laboratories Ltd., Colnbrook, Buckinghamshire). This is a finely divided iron oxide coated with a layer of polymerised m-diaminobenzene. The free amino groups are diazotised and supplied by the manufacturer as a stabilised fluoroborate salt. Coupling of gamma globulins was achieved by incubating 500 mg EnzacrylR FEO-M with 250-500 ~1 antiserum or normal gamma globulin, for 3 days at 4°C in 7 ml barbital buffer, (0.05 M, pH 8.6). The particles were then sedimented in a magnetic field and the supernatant aspirated. Next they were washed eight times by the alternate addition of 20 ml of barbital buffer and of 1 M sucrose/l M NaCl in barbital buffer, with vortexing, magnetic sedimentation of the particles and aspiration of the supernatant. Finally the Enzacrylgamma globulin particles were washed twice in barbital buffer and once in 0.05 M phosphate buffer, pH 7.4, containing 0.01 percent Tween 20, in which they were stored at a concentration of 100 mg/ml. Ferromagnetic antibody particle (MP) dilution curves. Doubling dilutions of
Fig. 3. The position of the magnetic particles when the magnet is in the off position on position. The latter reflects the fringe fields generated from the ridged bars.
and when it is in the
anti-TJ, anti-digoxin and anti-HPL magnetic particles containing from 10.0 to 0.04 mg of Enzacryl per tube, were incubated with hormone free serum, “‘Ilabelled hormone and, in the case of the anti-T4 particles, ANS. The tubes were then incubated for 2 h at room temperature, during which time they were efficiently and reproducibly mixed in a pulsed magnetic field, between the ridged bars of the rack. The mild steel base of the rack modifies the field generated by the electromagnet so that the ridges of the base act as individual magnets with similar poles being adjacent to one another, creating a magnetic field characteristic of two opposing, like poles. Thus when the magnet is switched on the MP move from their resting position at the base of the tube to occupy positions along the magnetic field lines at distances up to 2 cm from the base (Fig. 3). The MP are mixed efficiently by switching the magnet on for 2 s (seconds) in every 4, causing them to move up and down the tube regularly, for which 25 volts DC was adequate. Separation of the antibody-bound and free fractions was obtained by moving the top of the rack so that the tubes were directly over the ridges and hence over the contrived magnetic poles. The magnet was switched off and the MP resuspended in 1 ml of 0.9 percent saline containing 0.01 percent Tween 20 (Fig. 4a), it was then switched on to sediment the MP (Fig. 4b). This latter process was most efficient, requiring approximately 3 min, when the magnet was energised from 50 volts DC for 10 s in every 15 s. The power was then held on continuously during aspiration of the supernatant fluid (Fig. 4~). This was repeated three times and took approximately 15 min for 100 tubes.
Fig. 4. Use of the magnet to separate the antibody-bound and free fractions. (A) Following addition of 1 ml of 0.9% saline with the magnet switched off. (B) Rapid sedimentation of the particles when the magnet is switched on. (C) Aspiration of the supernatant with the magnet switched on.
Ferromagnetic particle radioimmunoassays. Reagents were added in the order and volumes summarised in Table I. The MP were kept in suspension during their addition by the use of an overhead paddle stirrer. The tubes were incubated for 2 h at room temperature, during which time mixing was achieved in the pulsed magnetic field. The bound and free fractions were then separated as above. Double-antibody magnetic particle (DAMP) separation This was investigated using the same reagents, reagent volumes and blanks as for the liquid-phase HCG radioimmunoassay. After 5 h incubation at room temperature, doubling dilutions of the Enzacryl-donkey anti-rabbit serum conTABLE
I
ORDER AND VOLUMES RADIOIMMUNOASSAYS
OF ADDITION
Sample or standard (serum) l2 5 I-labelled antigen Antibody-Enzacryl in Tween 20 buffer (final concentration of Tween 20, 0.01%)
FOR
THE
Tq. HPL AND DIGOXIN
MAGNETIC
PARTICLE
T4
HPL
Dlgoxin
100 /ll 100 cl1 (1 ng) 100 /.I1 (4 rag)
25 ~1 50 fil (0.5 ng) 175 ~1 (1.25 mg)
100 /.I1 100 /ll (100 pg) 100 /.ll (0.3 mg)
393
Bi
/ T x 100
20 --
10 __ 0.
n-
.
10
.
0.
5
.
0
_ ‘rn_
-
o----o-z-~ -.- _._ 1 I
2: 5
Fig. 5. Antibody dilution and T4 (0). The interrupted
0 _
i. 25
_
&_?
0:;2-
-
=--b
0
-c_&=___ I
- 0.31
0: 16
*= I
___‘
dO8
0.01
curves employing dilutions of Enzacryl conjugated lines illustrate the corresponding ‘blank’ values.
mg Enzacryl to digoxin
I tube p),
HPL (0)
jugate were added and mixed overnight in the pulsed magnetic field. The bound and free fractions were then separated as above. Calculations After subtraction of the blank (Sb), the antibody-bound fraction (BJ was expressed as a percentage of the total radioactivity ((Si-Bb)/T X 100). The best cubic curve was computed for the standards by the method of least squares and the amount of T4 in the test samples was interpolated automatically using a suitable computer programme. Results The antibody dilution curves for anti-T,, anti-HPL and anti-digoxin linked Enzacryl are shown in Fig. 5. Standard curves for T4, HPL and digoxin employing a suitable concentration of the magnetic particles were then compared with the corresponding liquid phase assays. Both methods cover similar ranges (Fig. 6). The ‘within-batch’ coefficients of variation for ten simultaneous analyses of a single pooled serum were 3.9 and 4.5 percent for the digoxin and T4 assays, respectively. Similar T4 levels were obtained when 50 serum samples were assayed by the PEG and MP techniques (Fig. 7) with r = 0.9572. Assessment of the donkey anti-rabbit gamma globulin serum before and after
394
bU *
i 50
c1u
!
30
60
!
100
200 403 0
c--c---c1
2
Fig. 6. Comparison (C) digoxin.
4
6
HPt rnj 1
T4 11nol I
of liquid phase (0) and magnetic
particle
10
0
(0) assays for (A)
thyroxine,
(B) HPL, and
I60
+d! t
40
.
l
~40
80
!
120
!
160
!
200
!
240
!
280
hlP Assay T4 nmoli
Fig. 7. Comparison of the results obtained for 50 serum samples by the liquid-phase assay and by the magnetic particle assay for thyroxine.
I
polytheylene
glycol
395
10
Fig.
8 ,\*
10
5
2. 5
1.25
0.625 mq Enracryl
I25
1 50
1.100
1 200
1400
8.
second
FIR& dilution
i
tube DARS
Comparison of magnetic particle second antibody separation antibody separation (0) in the radioimmunoassay for HCG.
(0) and of conventional
liquid-phase
linkage to the Enzacryl indicated that approximately 60 percent had been conjugated, with 0.6 ~1 of antiserum being linked to a mg of the solid phase. Fig. 8 shows second antibody dilution curves comparing separation of the antibodybound and free fractions in the HCG assay by the liquid-phase system (which employs normal rabbit serum as a carrier at a final dilution of 1 : 1000) and by the DAMP technique (where no carrier serum was required). Between 1.25 and 2.5 mg of DAMP gave similar results to those obtained using the second antibody at a final dilution of 1 : 100 in the liquid system. Thus the liquid-phase system required about ten times the amount of second antibody required in the DAMP technique. Discussion While this manuscript was being prepared, Hersh and Yaverbaum [3] published preliminary findings concerning a magnetic particle solid-phase radioimmunoassay for digoxin, with its advantage of substituting a magnet in place of centrifugation. The present study extends this approach to the assay of T4 and HPL as well as digoxin and demonstrates that the multiple washing steps essential in a solid phase system can be performed more simply and rapidly. Furthermore, magnets are less expensive than centrifuges, enable the whole assay to be performed at the bench and avoid the need for sample transfer. We have shown, for the first time, that use of such magnetic particles has the additional advantage, compared with many solid phase systems, of avoiding the need to cap tubes and mix their contents by vertical rotation during incubation. Thus efficient mixing can be obtained by means of a pulsed magnetic field. The magnetic support, Enzacryl, employed in this study is available commercially and can be covalently linked to gamma globulins relatively simply, following the manufacturer’s instructions. The magnetic particle-antibody conjugates are stable for at least three months when stored at 4°C. As an alternative to having to conjugate several antisera to the particles, experiments were undertaken in which a double-antibody system (DAMP) was employed. This proved satisfactory for the assay of HCG and had the advantage, compared with a liquid-phase double-antibody system, of requiring about ten times less antiserum.
596
References 1 Ratcliffe, J.G. (1974) Br. Med. Bull. 30, 32 2 Wide, L. and Porath, J. (1966) Biochim. Biophys. Acta 130, 257 3 Hersh. L.S. and Yaverbaum, S. (1975) Clin. Chim. Acta 63, 69 4 Nye, L., Hassan, M., Wilmott, E. and Landon, J. (1976) J. Clin. Pathol.. in press 5 Letchworth, A.T.. Boardman, R., Bristow, C., Landon. J. and Chard, T. (1971) J. Obstet. Gynaecol. Br. Commonw. 78.535 6 Greenwood, H., Howard, M. and Landon, J. (1974) J. Clin. Pathol. 27, 490 7 McNeilly, AS. and Hagen, C. (1974) Clin. Endocrinol. 3, 427
Clinica Chimica Acta, 69 (1976) 387-396 0 Elsevier Scientific Publishing Company,
Amsterdam
- Printed
in The Netherlands
CCA 1656
SOLID-PHASE,
MAGNETIC PARTICLE
RADIOIMMUNOASSAY
LYNN NYE 8, *, G.C. FORREST C, HELENA GREENWOOD a, JAQUELINE GARDNER C, R. JAY C, J.R. ROBERTS b and J. LANDON a
S.
Departments of a Chemical Pathology, and b Medical Electronics, St. Bartholomew’s Hospital, London E.C. 1 and C Technicon Methods and Standards Laboratory, London E.C. 1 (U.K.) (Received
November
11,1975)
Summary A solid-phase radioimmunoassay system has been developed based on the use of antibodies covalently linked to polymer-coated iron oxide (EnzacrylR). An electro-magnet is employed both to mix the particles during incubation (by switching the field on and off) and to separate the antibody-bound and free fractions. This obviates the need for vertical rotation and for the time-consuming, multiple centrifugations required with conventional solid phase procedures. The system is universally applicable and methods have been established for the assay of thyroxine, human placental lactogen and digoxin. The thyroxine assay was employed as a model and it was shown that the results obtained for serum samples correlated closely with those using a routine liquid-phase radioimmunoassay. The applicability of employing a second antibody linked to the iron oxide particles was also studied.
Introduction A variety of techniques have been employed in radioimmunoassay to separate the antibody-bound and free fractions [l] . These include adsorption procedures (such as the addition of charcoal to adsorb the free fraction), fractional precipitation of the bound fraction (with, for example, organic solvents or neutral salts), precipitation of the bound fraction by addition of a second antibody (double-antibody procedures) and solid-phase methods. Double-antibody and solid-phase systems have the advantage of universal applicability to all radioimmunoassays, and assays based on the use of antibodies covalently linked to * Requests
for reprints
should
be addressed
to Lynn
Nye.
388
ultrafine Sephadex can be operated with a precision superior to that of most other methods [2]. However, solid-phase systems of this type have two major disadvantages, namely, the need to cap tubes and mix their contents by vertical rotation throughout the period of incubation and the multiple centrifugation and washing steps required during separation, in order to reduce non-specific blank values to a minimum and obtain reproducible results. Recently, a radioimmunoassay has been described for digoxin by Hersh and Yaverbaum [ 31, employing siliconised iron oxide as a solid-phase support. The particles were mixed manually, every 5 min, and separation of the antibodybound and free fractions achieved by sedimentation of the particles with a ‘U’ shaped magnet and tipping off the supernatant fluid containing the free fraction, which was counted. The present paper describes a magnetic particle radioimmunoassay which also overcomes the problem of mixing during the incubation period as well as facilitating the separation of the antibody-bound and free fractions. Materials and methods
Thyroxine (TJ. Total Tq levels were determined in non-extracted serum by the method of Nye et al. [4], in which 8-anilino-1-naphthalene sulphonic acid (ANS) is used to block the binding sites on endogenous thyroxine-binding globulin (TBG) and polyethylene glycol (PEG) is added to separate the bound and free fractions. Human placental luctogen (HPL). Serum HPL levels were determined by the method of Letchworth et al. [ 51, in which ethanol is added to separate the bound and free fractions. IJigoxin. Serum digoxin levels were determined by the method of Greenwood et al. 163, in which dextran coated charcoal is added to separate the bound and free fractions. Liquid-phase double-antibody separation A modification of the method described by McNe~y and Hagen [‘7] for the radioimmunoassay of human chorionic gonadotrophin (HCG) was employed to study double antibody separation. 500 ~1 of phosphate buffer containing 2 percent horse serum was added to a series of tubes, followed by 50 ~1 of 12’1labelled HCG and 50 ~1 of rabbit anti-HCG serum, which gave a final antiserum dilution (1 : 48 000). At this concentration approx~ately 50 percent of the labelled hormone was bound. A series of ‘blank’ tubes was set up containing phosphate buffer in place of the antiserum. The tubes were mixed and then incubated at room temperature for 5 h. Doubling dilutions of donkey anti-rabbit serum were prepared and loo-p1 amounts added to all the tubes (in duplicate) to give final dilutions ranging from 1 : 25 to 1 : 400. Ten ~1 of normal non-immunised rabbit serum, diluted in buffer, was added as a carrier, giving a final concentration of 1 : 1000. The tubes were incubated overnight at room temperature and the antibody-bound and free fractions separated by centrifugation at 1000 X g for 30 min, followed by aspiration of the supernatant.
389
Ferromagnetic particle radioimmunoassays The electromagnet. This comprises a power supply which was constructed from a heavy duty mains transformer with secondary voltages of 25 V and 50 V root mean square. These voltages are full wave rectified and the resulting unstabilised DC voltage is fed to a magnet via a switching mechanism. Initially a Technicon cam timer was employed to activate a micro-switch connected to the power supply which it switched on and off to produce a pulsed magnetic field. Later an electronic timer was substituted, which operates a relay in the primary circuit of the transformer, producing a similar pulsed magnetic field. The magnet consists of copper wire windings surrounding a mild steel core (2; X 2$ X 5 inches) and is attached to a specially constructed tube rack (Figs. 1 and 2). The rack which can hold 100 tubes, has a 0.5 inch mild steel base plate machined in a regular pattern of ridges and troughs, running parallel to one another across the plate. The ridges are 0.25 inches wide with gaps of 0.375 inches separating them to form troughs, 0.25 inches deep. The tubes are supported by ‘0’ ring clamps trapped between two aluminium sheets, and can be held in one of two positions, either on top of or suspended between the ridges. The entire system is available from ILS, 99 New Cavendish Street, London, WlM 7FQ. Preparation of ferromagnetic antibody particles. Gamma globulin fractions of sheep anti-T4, goat anti-HPL, rabbit anti-digoxin and donkey anti-rabbit sera were prepared by fractional precipitation with sodium sulphate. To 1 ml of
Fig. 1. The magnet
and rack which were specially
designed
for magnetic
particle
radioimmunoassays.
Al SUPPORT
SHEET
Fig. 2. An isometric
projection
of the magnet
and rack.
antiserum was added 180 mg anhydrous sodium sulphate, while mixing on a vortex mixer. The resulting suspension was mixed by vertical rotation at room temperature for 30 min, centrifuged at 1000 X g for 15 min and the supernatant fluid aspirated. The precipitate was washed twice with 18 percent sodium sulphate then dissolved in 0.9 percent saline and made up to 1 ml. The ferromagnetic solid-phase support was Enzacryl @ FEO-M (Koch Light Laboratories Ltd., Colnbrook, Buckinghamshire). This is a finely divided iron oxide coated with a layer of polymerised m-diaminobenzene. The free amino groups are diazotised and supplied by the manufacturer as a stabilised fluoroborate salt. Coupling of gamma globulins was achieved by incubating 500 mg EnzacrylR FEO-M with 250-500 ~1 antiserum or normal gamma globulin, for 3 days at 4°C in 7 ml barbital buffer, (0.05 M, pH 8.6). The particles were then sedimented in a magnetic field and the supernatant aspirated. Next they were washed eight times by the alternate addition of 20 ml of barbital buffer and of 1 M sucrose/l M NaCl in barbital buffer, with vortexing, magnetic sedimentation of the particles and aspiration of the supernatant. Finally the Enzacrylgamma globulin particles were washed twice in barbital buffer and once in 0.05 M phosphate buffer, pH 7.4, containing 0.01 percent Tween 20, in which they were stored at a concentration of 100 mg/ml. Ferromagnetic antibody particle (MP) dilution curves. Doubling dilutions of
Fig. 3. The position of the magnetic particles when the magnet is in the off position on position. The latter reflects the fringe fields generated from the ridged bars.
and when it is in the
anti-TJ, anti-digoxin and anti-HPL magnetic particles containing from 10.0 to 0.04 mg of Enzacryl per tube, were incubated with hormone free serum, “‘Ilabelled hormone and, in the case of the anti-T4 particles, ANS. The tubes were then incubated for 2 h at room temperature, during which time they were efficiently and reproducibly mixed in a pulsed magnetic field, between the ridged bars of the rack. The mild steel base of the rack modifies the field generated by the electromagnet so that the ridges of the base act as individual magnets with similar poles being adjacent to one another, creating a magnetic field characteristic of two opposing, like poles. Thus when the magnet is switched on the MP move from their resting position at the base of the tube to occupy positions along the magnetic field lines at distances up to 2 cm from the base (Fig. 3). The MP are mixed efficiently by switching the magnet on for 2 s (seconds) in every 4, causing them to move up and down the tube regularly, for which 25 volts DC was adequate. Separation of the antibody-bound and free fractions was obtained by moving the top of the rack so that the tubes were directly over the ridges and hence over the contrived magnetic poles. The magnet was switched off and the MP resuspended in 1 ml of 0.9 percent saline containing 0.01 percent Tween 20 (Fig. 4a), it was then switched on to sediment the MP (Fig. 4b). This latter process was most efficient, requiring approximately 3 min, when the magnet was energised from 50 volts DC for 10 s in every 15 s. The power was then held on continuously during aspiration of the supernatant fluid (Fig. 4~). This was repeated three times and took approximately 15 min for 100 tubes.
Fig. 4. Use of the magnet to separate the antibody-bound and free fractions. (A) Following addition of 1 ml of 0.9% saline with the magnet switched off. (B) Rapid sedimentation of the particles when the magnet is switched on. (C) Aspiration of the supernatant with the magnet switched on.
Ferromagnetic particle radioimmunoassays. Reagents were added in the order and volumes summarised in Table I. The MP were kept in suspension during their addition by the use of an overhead paddle stirrer. The tubes were incubated for 2 h at room temperature, during which time mixing was achieved in the pulsed magnetic field. The bound and free fractions were then separated as above. Double-antibody magnetic particle (DAMP) separation This was investigated using the same reagents, reagent volumes and blanks as for the liquid-phase HCG radioimmunoassay. After 5 h incubation at room temperature, doubling dilutions of the Enzacryl-donkey anti-rabbit serum conTABLE
I
ORDER AND VOLUMES RADIOIMMUNOASSAYS
OF ADDITION
Sample or standard (serum) l2 5 I-labelled antigen Antibody-Enzacryl in Tween 20 buffer (final concentration of Tween 20, 0.01%)
FOR
THE
Tq. HPL AND DIGOXIN
MAGNETIC
PARTICLE
T4
HPL
Dlgoxin
100 /ll 100 cl1 (1 ng) 100 /.I1 (4 rag)
25 ~1 50 fil (0.5 ng) 175 ~1 (1.25 mg)
100 /.I1 100 /ll (100 pg) 100 /.ll (0.3 mg)
393
Bi
/ T x 100
20 --
10 __ 0.
n-
.
10
.
0.
5
.
0
_ ‘rn_
-
o----o-z-~ -.- _._ 1 I
2: 5
Fig. 5. Antibody dilution and T4 (0). The interrupted
0 _
i. 25
_
&_?
0:;2-
-
=--b
0
-c_&=___ I
- 0.31
0: 16
*= I
___‘
dO8
0.01
curves employing dilutions of Enzacryl conjugated lines illustrate the corresponding ‘blank’ values.
mg Enzacryl to digoxin
I tube p),
HPL (0)
jugate were added and mixed overnight in the pulsed magnetic field. The bound and free fractions were then separated as above. Calculations After subtraction of the blank (Sb), the antibody-bound fraction (BJ was expressed as a percentage of the total radioactivity ((Si-Bb)/T X 100). The best cubic curve was computed for the standards by the method of least squares and the amount of T4 in the test samples was interpolated automatically using a suitable computer programme. Results The antibody dilution curves for anti-T,, anti-HPL and anti-digoxin linked Enzacryl are shown in Fig. 5. Standard curves for T4, HPL and digoxin employing a suitable concentration of the magnetic particles were then compared with the corresponding liquid phase assays. Both methods cover similar ranges (Fig. 6). The ‘within-batch’ coefficients of variation for ten simultaneous analyses of a single pooled serum were 3.9 and 4.5 percent for the digoxin and T4 assays, respectively. Similar T4 levels were obtained when 50 serum samples were assayed by the PEG and MP techniques (Fig. 7) with r = 0.9572. Assessment of the donkey anti-rabbit gamma globulin serum before and after
394
bU *
i 50
c1u
!
30
60
!
100
200 403 0
c--c---c1
2
Fig. 6. Comparison (C) digoxin.
4
6
HPt rnj 1
T4 11nol I
of liquid phase (0) and magnetic
particle
10
0
(0) assays for (A)
thyroxine,
(B) HPL, and
I60
+d! t
40
.
l
~40
80
!
120
!
160
!
200
!
240
!
280
hlP Assay T4 nmoli
Fig. 7. Comparison of the results obtained for 50 serum samples by the liquid-phase assay and by the magnetic particle assay for thyroxine.
I
polytheylene
glycol
395
10
Fig.
8 ,\*
10
5
2. 5
1.25
0.625 mq Enracryl
I25
1 50
1.100
1 200
1400
8.
second
FIR& dilution
i
tube DARS
Comparison of magnetic particle second antibody separation antibody separation (0) in the radioimmunoassay for HCG.
(0) and of conventional
liquid-phase
linkage to the Enzacryl indicated that approximately 60 percent had been conjugated, with 0.6 ~1 of antiserum being linked to a mg of the solid phase. Fig. 8 shows second antibody dilution curves comparing separation of the antibodybound and free fractions in the HCG assay by the liquid-phase system (which employs normal rabbit serum as a carrier at a final dilution of 1 : 1000) and by the DAMP technique (where no carrier serum was required). Between 1.25 and 2.5 mg of DAMP gave similar results to those obtained using the second antibody at a final dilution of 1 : 100 in the liquid system. Thus the liquid-phase system required about ten times the amount of second antibody required in the DAMP technique. Discussion While this manuscript was being prepared, Hersh and Yaverbaum [3] published preliminary findings concerning a magnetic particle solid-phase radioimmunoassay for digoxin, with its advantage of substituting a magnet in place of centrifugation. The present study extends this approach to the assay of T4 and HPL as well as digoxin and demonstrates that the multiple washing steps essential in a solid phase system can be performed more simply and rapidly. Furthermore, magnets are less expensive than centrifuges, enable the whole assay to be performed at the bench and avoid the need for sample transfer. We have shown, for the first time, that use of such magnetic particles has the additional advantage, compared with many solid phase systems, of avoiding the need to cap tubes and mix their contents by vertical rotation during incubation. Thus efficient mixing can be obtained by means of a pulsed magnetic field. The magnetic support, Enzacryl, employed in this study is available commercially and can be covalently linked to gamma globulins relatively simply, following the manufacturer’s instructions. The magnetic particle-antibody conjugates are stable for at least three months when stored at 4°C. As an alternative to having to conjugate several antisera to the particles, experiments were undertaken in which a double-antibody system (DAMP) was employed. This proved satisfactory for the assay of HCG and had the advantage, compared with a liquid-phase double-antibody system, of requiring about ten times less antiserum.
596
References 1 Ratcliffe, J.G. (1974) Br. Med. Bull. 30, 32 2 Wide, L. and Porath, J. (1966) Biochim. Biophys. Acta 130, 257 3 Hersh. L.S. and Yaverbaum, S. (1975) Clin. Chim. Acta 63, 69 4 Nye, L., Hassan, M., Wilmott, E. and Landon, J. (1976) J. Clin. Pathol.. in press 5 Letchworth, A.T.. Boardman, R., Bristow, C., Landon. J. and Chard, T. (1971) J. Obstet. Gynaecol. Br. Commonw. 78.535 6 Greenwood, H., Howard, M. and Landon, J. (1974) J. Clin. Pathol. 27, 490 7 McNeilly, AS. and Hagen, C. (1974) Clin. Endocrinol. 3, 427
