EP0223816A1

EP0223816A1 - Enzyme assay of body fluids

Info

Publication number: EP0223816A1
Application number: EP19860903511
Authority: EP
Inventors: John Kingsley Amersham Place MARTIN; Stephen Alexander Amersham Place CHARLES; Alastair Howard Amersham Place DENT
Original assignee: Amersham International PLC
Current assignee: GE Healthcare Ltd
Priority date: 1985-06-06
Filing date: 1986-06-06
Publication date: 1987-06-03
Also published as: JPS62502633A; GB8514288D0; WO1986007462A1

Abstract

Dans une analyse qui fait appel à l'incubation d'un échantillon d'un fluide somatique avec un réactif marqué avec un enzyme, les résultats peuvent être affectés par la présence dans l'échantillon d'un inhibiteur de l'enzyme. Ce problème est résolu en incluant dans le mélange d'incubation un facteur de correction pour arrêter l'action de tout inhibiteur présent dans l'échantillon. Si l'enzyme est une péroxydase de raifort (HRP), le facteur de correction peut être apo-HRP, un isoenzyme HRP acide ou un enzyme HRP carboxyméthylé. La présente invention est particulièrement bien adaptée à des immunoanalyses à luminescence accrue pour haptènes et antigènes.In an assay that involves incubating a sample of a somatic fluid with an enzyme labeled reagent, the results may be affected by the presence in the sample of an enzyme inhibitor. This problem is resolved by including in the incubation mixture a correction factor to stop the action of any inhibitor present in the sample. If the enzyme is horseradish peroxidase (HRP), the correction factor can be apo-HRP, an acidic HRP isoenzyme, or a carboxymethylated HRP enzyme. The present invention is particularly well suited to immunoassays with increased luminescence for haptens and antigens.

Description

ENZYME ASSAY OF BODY FLUIDS

Assays for various analytes in body fluids using reagents labelled with radioactive isotopes are well known and widely available commercially. Examples of such assays are radioim uno assays which employ a radioactively labelled version of the analyte; and radioimmunometric assays which employ a radioactively labelled antibody to the analyte. These assays are

_q capable of quantitating nanogra (10 g) amounts, or

-12 even picogram (10 g) amounts of analyte in a few icrolitres of body fluid.

More recently there has been interest in assays which use non-radioactive labels, for example enzyme labels. These have the advantage that special precautions for handling radioactive materials are not required. Enzyme labels have the further advantage as the enzyme is used to convert many molecules of substrate and thus result in increased observed signal intensity. Armenta R et al. describe (Analytical

Biochemistry, 146, 211-219, 1985) a homogeneous enzyme assay for ferritin. A limiting amount of specific antibody undergoes a competition reaction with ferritin from the sample and ferritin- £ -galactosidase conjugate. Addition of a macromolecular substrate of ^-galactosidase to the assay mixture results in the formation of a fluorescent product only if the ferritin-β-galactosidase conjugate is unbound by antibody. Binding of the antibody to the conjugate inhibits the approach of the macromolecular substrate to the enzyme and therefore inhibits the production of fluorescent product. Certain substances in certain patient sera bind to the Φ-galactosidase part of the conjugate thereby inhibiting the binding of the macromolecular substrate to the enzyme independently of the binding of antibody. To get round this problem the ?-galactosidase part of the conjugate has been coated with albumin and an excess of inactivated fl-galactosidase is added. It is not suggested that the addition of inactivated β-galactosidase would be useful by itself. A homogeneous assay as described depends on inactivation of the label under some circumstances, and it may be difficult to avoid unintentional inactivation when this is not desired.

We have been investigating assays which use reagents labelled with the common enzyme horseradish peroxidase (HRP). In general, we get excellent correlation between results obtained using assays with radioactive labels and results obtained using comparable assays with enzyme labels. However, in assays where a body fluid sample is incubated with a reagent carrying an HRP label, we find that a small proportion (about 2%) of samples give rise to quite misleading results. We do not know the cause of these aberrant results. But -we believe that they may be due to the presence in the sample of an unidentified substance which inhibits the enzyme action of HRP or inhibits the binding to antibody of an antigen/HRP conjugate. Whether our views as to the cause of the problem are right or wrong, we have found a way of overcoming it, which forms the subject of this invention.

The invention provides a method of performing an assay on a sample of a body fluid including a step of incubating a mixture of the sample with a reagent labelled with an enzyme, wherein the sample may contain an inhibitor for the enzyme, which method comprises including in the incubation mixture a correction factor to block the action of any inhibitor present in the sample. Preferably the assay is one which involves an immune reaction between an analyte and its specific binding partner, in which immune reaction the reagent labelled with enzyme also participates. Thus the analyte may be an antigen or hapten and the specific binding partner its associated antibody; or the analyte may be an antibody, with an antigen as its specific binding partner. Often the inhibitor is a substance, usually macromolecular, which binds to the enzyme in such a way as to inhibit participation of the labelled reagent in the immune reaction. On other occasions, the inhibitor is an enzyme poison, naturally occurring or as an added preservative, which binds to the enzyme in such a way as to reduce or destroy its enzymic activity.

Our work to date has been confined to horseradish peroxidase enzymes. However, we think it probable that our results will be valid also for enzymes in general and particularly peroxidase enzymes. Although the following description refers mainly to HRP, the invention is considered to extend to other enzymes.

The nature of the assay is not critical. It may for example be an uptake assay, or a competition or im unometric assay for total or for free analyte. The analyte may be a hormone, an enzyme, a biochemical messenger, a steroid, a drug, a drug metabolite, a polypeptide or protein, a catechol-amine, a vitamin, a tumour antigen, a toxin, an alkaloid, a mono-, di, or poly-saccharide , or a virus or virus particle. The analyte may itself be antigenic, or may be a small molecule such as a hapten (which can initiate the production of antibodies only when joined to a larger molecule); or may be an antibody.

The sample for assay may be taken from any body fluid, such as for example plasma, urine, or serum. We have coined the term "sample correction factor" or SCF for short, which term is occasionally used herein¬ after. The correction factor or SCF is a material which blocks the action of any inhibitor for the peroxidase enzyme which may be present in the assay sample. Clearly, the correction factor must be .inert to the assay reagents and must not otherwise affect the assay. For example, the correction factor should preferably not itself have enzymatic activity in order to maintain assay precision. One suitable material is the apo- protein left after gentle removal of the haem group from HRP. A satisfactory procedure for doing this, involving the use of methylethylketone to separate and dissolve the haem group, has been described (Teale, F.W.J., Biochi . Biophys. Acta, 3_5, (1959) page 543). It is envisaged that other correction factors might be employed. In particular, it is likely that HRP can be modified in different ways to destroy its enzymic activity without destroying its ability to block the action of the supposed enzyme inhibitor.

Two other materials are suitable for use as sample correction factors when an HRP label is used to generate a chemi luminescent signal which is amplified by the known techniques of enhanced luminescence. The first is an acidic form of HRP, which is not susceptible to luminescence enhancement and can hence be regarded as enzy ically inactive for the purposes of the assay. Thus Sigma Chemical Company Limited sell peroxidase isoenzymes including two acidic isoenzymes (Types VII and VIII), one basic isoenzyme (Type IX), of which only the basic isoenzyme is susceptible to luminescence enhancement. (The terminology is in accordance with the classification of Shannon L.M. et al., J. Biol. Chem., 241, 2166 (1966)).

The second SCF material for this purpose is a carboxymethylated HRP prepared by the method of Gurd F.R.N., Methods of Enzymology, Vol. 11, 532. Both these materials are suitable for use as sample correction factors: because they are inert to the assay reagents, e.g. analyte, conjugate and antibody; and because they are not susceptible to luminescence enhancement and can hence be regarded as enzymatical ly inactive for the purposes of the assay.

Aberrant samples may also contain enzyme poisons, which reduce or destroy enzyme activity under all circumstances. These poisons are distinguished from the macromolecular substances which can bind to enzymes and physically block their reaction with macromolecular substrates. Enzyme poisons may be naturally occurring, or deliberately added.

Serum controls for assay kits often contain sodium azide as a preservative. Sodium azide binds to the haem group of HRP. So serum controls often constitute aberrant samples according to this invention. And apo-HRP enzyme is not satisfactory as a SCF because it lacks the haem group to which sodium azide binds. For enhanced luminescence assays, the two materials noted above, acidic HRP isoenzyme and carboxymethylated HRP, are suitable SCF's.

The amount of correction factor needed to block the action of the inhibitor depends on several factors, such that no overall concentration range can be given. However, the amount of correction factor required to bring an aberrant result obtained from a particular sample back to normality is easily determined by routine experiment. Factors affecting the required amount of correction factor include the following. The quantity and nature of the analyte; more analyte requires more labelled reagent, which in turn requires more correction factor. The volume of the sample; more sample may require more correction factor. The amount of the HRP labelled reagent; more labelled reagent may require more correction factor. The purity of the HRP labelled reagent, and in particular the amount of free HRP present as impurity; use of a pure labelled reagent may require use of more correction factor. Indeed, if enough free HRP is present, there may be no need for any correction factor; but then the precision would be greatly impaired.

The invention is applicable to various different kinds of assay, of which the following are examples :-

A. Competition assay for total analyte. The sample is incubated with an amount of an analyte-HRP conjugate and an .antibody to the analyte which is insolubi lised either before, during or after the incubation. The analyte and analyte-HRP conjugate compete for binding with the antibody. The amount of enzyme attached to antibody varies inversely with the amount of analyte in the sample.

Many analytes are present in body fluids, partly in free form and partly bound to naturally occurring protein binders. For example, tri-iodothyronine (T3) is present in serum 0.3% in free form and 99.7% bound to the naturally occurring binding proteins, thyroxine binding globulin (TB6), thyroxin binding pre-albumin (TBPA) and albumin (ALB). For a total analyte assay, it is necessary to remove the competition provided by these natural binding proteins. A usual way of doing this is to add to the incubation mixture a blocking agent which blocks the binding sites of the natural binding proteins and has the effect of providing all the analyte in the sample in free form. An example of a blocking agent useful with thyroid hormones is 8-ani 1 ino-naphthalene sulphonic acid (ANS). This conventional blocking agent is quite different from the correction factor of this invention. The blocking agent acts on the serum binding proteins and provides all the analyte in the serum in free form for assay; the correction factor is believed to act on some supposed inhibitor and ensures that all the HRP enzyme is available to participate in the assay.

B. A one-site im unometric assay for total analyte. The sample is incubated with an amount of added immobilised analyte and a limited amount of an HRP conjugate of an antibody to the analyte. A blocking agent may also be present if required to ensure that the analyte is all in free form. The analyte in the sample and the immobilised analyte compete for reaction with the labelled antibody. The amount of insoluble HRP label varies inversely with the amount of analyte in the sample.

C. A competition assay for free analyte. The procedure here may be as in A or B, with two important differences, both designed to avoid disturbing the free-bound equilibrium of analyte in the sample. The first difference is that the blocking agent must be absent. The second is that the amount of antibody used is insufficient to substantially disturb the free- bound analyte equilibrium; this may require the use of a rather small amount of rather high affinity antibody.

D. A simultaneous 2-site immunometric assay for total analyte. The sample is incubated with an amount of an immobilised antibody to the analyte and an amount of an HRP conjugate of an antibody to the analyte. A blocking agent may also be present if required. This method is only applicable to analytes which have at least two sites which are recognised by antibodies. The amount of immobilised label varies directly with the amount of analyte in the sample. E. An analyte uptake assay. This is a test to determine the number of available unused analyte binding sites on the natural binding proteins in the sample. For example, T3 and T4 uptake assays are commonly used as adjuncts to total T3 and total T4 assays. In this kind of test, the sample is incubated with an amount of the unlabelled analyte, an amount of an HRP conjugate of the analyte, and an amount of antibody, which is immobilised either before or after the incubation. The amount of insolubi 1 ised HRP label varies directly with the number of vacant binding sites. Other assay procedures in which a sample is incubated together with an enzyme-labelled molecule will readily occur to those skil led in the art.

The inhibitor is usually macromolecular. In assays of classes A, C and E above, our work suggests that it binds with HRP in a manner which mainly prevents subsequent binding of the analyte/HRP conjugate to antibody; and.that inhibition of the enzymic activity of HRP occurs, if at all, only to a minor degree. In assays of classes B and D, the inhibitor probably binds with HRP in a manner which mainly prevents subsequent binding of the antibody/HRP conjugate to analyte. Where an enzyme poison such as sodium azide is present, this probably binds to HRP in such a way as to reduce or destroy its enzymic activity.

Although this invention can be applied to homogeneous assays, it is better suited to heterogeneous assays, in which bound label is physically separated from not-bound label prior to observation of a signal generated by either the bound or the not-bound label. Separation may readily be achieved by the use of one assay reagent bound to solid particles or to the vial wall, in accordance with well- known techniques.

In assays which use enzyme labels, the enzyme can be made to generate an observable signal in various ways. For example, HRP can be caused to catalyze the the oxidation of luminol with the production of light via chemi luminescence. Or HRP enzyme can be used to generate colour by addition of o-phenylenediamine. The nature of the system used to generate an observable signal is not critical to the present invention, and any known and appropriate system can be used.

Reference is directed to the accompanying drawings in which:-

Figure 1 is a graph comparing the concentration of T3 in serum samples by two different techniques. The X axis shows the concentration, expressed in nanomoles/litre, determined by a commercially available radioimmuno assay. The Y axis shows the concentration on a similar basis, determined by a conventional enzyme immunoassay (using enhanced luminescence) without the use of a correction factor.

Figure 2 represents a graph which is the same as Figure 1 except that the Y axis gives results obtained by an enzyme immunoassay (using enhanced luminescence) including a correction factor according to this invention.

In Figure 1, the X axis represents total T3 concentrations, in nanomoles per litre of serum, determined using the commercially available Amerlex RIA kit, which uses latex beads coated with antibody. The Y axis represents total T3 concentrations of the same sera determined using a conventional enzyme immunoassay (enhanced luminescence) test, the protocol of which is set out below. In the graph, the letter H refers to a hypothyroid sample. The letter E to a Euthyroid sample, and the letter P to a sample obtained in pregnancy, while the letter T refers to hyperthyroid samples. As can be seen, most of the results are quite accurately on a straight line passing through the origin, as is to be expected if the two assay techniques are comparable. But there are four aberrant results indicated by stars. In these, the indicated T3 concentration is too high, which means that the activity of immobilised HRP enzyme was too low. This is consistent with part of the enzyme having been inhibited by some unknown inhibitor in the sample.

In the results shown in Figure 2, all the enzyme assays have been repeated using a comparable protocol but including a correction factor, namely apo-HRP enzyme in the incubation mixture. Of the four aberrant results, three have been normalized and the fourth substantially so. These graphs thus show that the use of a correction factor can be effective to normalize aberrant results sometimes obtained with HRP enzyme immunoassays. Example 1

The four samples giving rise to aberrant results in Figures 1 and 2 were further analysed, together with three other aberrant samples by a variety of assay techniques. The assay techniques are set out below; techniques 1 , 2 and 7 were those used to generate

Figures 1 and 2. The results of seven samples are set out in the table which follows; of these, samples A, B, C and D were those giving rise to aberrant results in Figures 1 and 2. 1. Total T3 Amerlex RIA (Amersham International pic, Amersha , Bucks., U.K.)

The protocol detailed in the pack leaflet of the IM2000 kit was used. 2. Total T3 Enhanced Luminescence Immunoassay

The following reagents were incubated together in the wells of white microtitre strips (Dynatech Laboratories Limited, Bii 1 ingshurst, U.K.), to which had been attached donkey anti-sheep antibodies, for 60 minutes at 37°C:

25 microlitres of a standard solution of T3 in human serum.

100 microlitres of a solution of T3 coupled to

HRP in a barbitone buffer, pH 8.8. 100 microlitres of a solution of sheep anti-T3 in a phosphate buffer, pH 7.

After the incubation, the contents of each well were aspirated and the wells washed thoroughly. Immobilised HRP enzyme was determined by enhanced luminescence involving a luminol/perborate system.

3. Total T3 Colorimetric Immunoassay

The procedure differed from that described in 2 above in that clear microtitre strips were used. After washing, immobilised HRP enzyme was detected by a chromogenic reaction by adding o-phenylenediamine.

4. Total T3 RIA

The procedure differed from that described in 2 above as follows. The T3 coupled to HRP was replaced with a T3 solution labelled with 125-1. After washing, bound activity in each well was measured for ten minutes in a gamma counter.

5. Total T3 by a commercially available Colorimetric enzyme method

The protocol detailed in the pack leaflet of the 204528 (Boehringer Mannheim, Lewes, U.K.) kit was used. - 12 -

6. Total T3 Colorimetric immunoassay plus SCF

The protocol for this assay was the same as for the Boehringer assay 5 above, except that a correction factor (apo-HRP enzyme) was included in the solution of T3-HRP conjugate at a concentration of 10mg/1.

7. Total T3 Enhanced Luminescence immonoassay plus SCF The protocol for this assay was the same as for assay number 2 above, except that there was included in the T3-HRP conjugate solution, a correction factor (apo-HRP enzyme) at a concentration of 10mg/l, sufficient to give 1 microgram per test. The correction factor was prepared by the method outlined above (Teale, F.W.J., Biochim. Biophys. Acta, 3_5, 543 [1959]).

Table 1 Total T3(nmol/l) of aberrant Serum Samples by various

Assay Techniques.

Assay Method

Sample 1 2 3 4 5 6 7

A 1.06 4.73 4.07 1.10 4.65 - 0.96

B 2.22 13.50 8.35 2.74 8.53 2.62 2.59

C 1.49 14.10 9.66 0.59 20.76 3.25 2.66

D 1.24 6.29 7.21 1.25 - - 1.10

E 0.51 1.85 - - - - 0.59

F 0.71 4.45 - - - - 0.80

G 1.91 7.92 - - - - 1.94

The figures in column 1 were obtained by a commercial RIA and are considered for the purposes of this invention as accurate.

The figures in column 2 were obtained by a conventional type of HRP enzyme enhanced luminescent immunoassay; they are all substantially greater than the corresponding figures in column 1.

The figures in column 3 were obtained by a conventional HRP enzyme colorimetric immunoassay; they too are greater than the corresponding figures in column 1.

The figures in column 4 were obtained by a RIA employing a solid phase identical to that used in methods 2) and 3) except that the HRP enzyme was

125 replaced by an I tracer; they are comparable to those in column 1 and suggest that the aberrant results in columns 2 and 3 are due to the nature of the enzyme label used, rather than to any other factor, such as the nature of the solid phase.

The figures in column 5 were obtained using a commercially available HRP enzyme colorimetric immunoassay; they also are much greater than the corresponding figures in column 1.

Columns 6 and 7 represent assays performed in accordance with the present invention. Column 6 should be compared to column 5, and also column 1, to show the effect of adding a correction factor to a commercial colorimetric HRP enzyme immunoassay. Column 7 should be compared to column 2, and also to column 1, to show the effect of adding a correction factor to a conventional enhanced luminescent HRP enzyme immunoassay.

Example 2

Tests were carried out to determine how much correction factor needed to be added to cope with aberrant samples in a total T3 HRP enzyme immunoassay. The assay protocols were as described in Example 1 under 1 and 7. Samples of two sera, one normal and one abnormal, were tested. The results are set out in table 2.

Table 2 Total T3 (nmol/1) of two Serum Samples with SCF

SCF Abnormal Normal

(micrograms Serum Serum per assay) Sample Sample

0 7.92 0.92

0.1 2.89 1.05

1 1.73 1.09

2.5 1.58 1.23

10- ^• 1.84 ^' 0.92

Amerlex RIA value 1.91 1.10

In this particular assay, 1 microgra of SCF per test was sufficient to "normalize" the result from the abnormal serum sample. Addition of SCF to assays of the normal serum sample did not markedly affect the results obtained.

Example 3

This experiment employed two assay techniques in addition to those listed in Example 1 :-

8. Total T enhanced luminescence assay without SCF, as 2 except that an ¹²⁵I-label led HRP/T₃ conjugate in place of the unlabelled HRP/T₃ conjugate.

The 125 I-label was not expected to make any difference to the luminescent results obtained.

9. As 8, except that immobilised 125 I was measured instead of the HRP.

The following results were obtained on two aberrant samples, (Total I- in nmol/1)

Table 3

Assay Method

Sample 7 2 8 9

H 2.07 4.59 5.31 5.29

I 1.76 >12 >12 >12

The results of method 7, enhanced luminescence immunoassay with SCF, are regarded as correct for the purpose of this experiment.

The results of method 2, enhanced luminescence immunoassay without SCF, are much higher thus demonstrating that samples H and I are aberrant.

The results of method 8 are, as expected, essentially the same as method 2.

125

The rt-ults of method 9, in which immobilised I was measured instead of immobilised HRP, are also essentially the same as methods 2 and 8. This indicates that in this case the inhibitor has acted mainly to prevent HRP/T- conjugate from binding to immobilised antibody. There is no evidence that the inhibitor has reduced the enzymatic activity of HRP in the HRP/T₃ conjugate. Example 4 This experiment employed two more assay techniques in addition to those listed in Example 1 :-

10. Total T enhanced luminescence assay without SCF, as 2 except that rabbit anti-HRP antibody was added to the sample.

11. Total T₃ enhanced luminescence immunoassay with SCF, as 7 except that rabbit anti-HRP antibody was added to the sample.

The following results were obtained (total To in nmol /l ) :-

Table 4

Assay Method

Sample 7 10 11

J 0 .92 0 .83 > 12 1 .03 K 2.08 2.04 > 12 2. 1 3 L 5 .20 5 . 15 > 12 5.38 M 3 .24 3 .54 > 1 2 3 .46

The similarity between the results of methods 2 and 7 indicates that samples J, K, L and M were not aberrant samples.

In method 10 (without SCF) and method 11 (with SCF), the samples were spiked with anti-HRP antibodies. The results of method 10, but not method 11 , are high compared to methods 2 and 7. This demonstrates that anti-HRP antibodies can mimic the inhibitor for the purpose of this invention. Example 5

Serum controls for assay kits often contain sodium azide as a preservative. Sodium azide binds to the haem group of HRP. So serum controls often constitute aberrant samples requiring the use of a SCF according to this invention.

This experiment employed two more assay techniques in addition to those listed in Example 1:-

12. Total To enhanced luminescence assay, as 2 except that acidic HRP was added at a concentration of 100ng/assay.

13. Total T_o enhanced luminescence assay, as 2 except that carboxymethylated HRP was added at a concentration of 100ng/assay.

The following results were obtained (total T_o in nmol/1). The samples were three commercial serum controls containing sodium azide:-

Ta b l e 5

As s ay Method

Samp l e 1 7 12 1 3

N 0.69 2. 16 0.70 0.48

0 1 .76 4.72 1 .73 1 .65

P 5. 15 9.63 5.08 5.50

The results of method 1, a commercial RIA, are considered for the purposes of this experiment as accurate. The results of method 7, enhanced luminescence immunoassay using apo-HRP enzyme as SCF, are higher than those of method 1. This demonstrates that the samples are aberrant, and that the apo-HRP enzyme was not effective as a SCF.

The results of methods 12 and 13, enhanced luminescence immunoassay using different SCF's are essentially the same as those of method 1. This demonstrates that the SCF's chosen were effective for their purpose.

Claims

C LA IMS

1. A method of performing an assay on a sample of a body fluid including a step of incubating a mixture of the sample with a reagent labelled with an enzyme, wherein the sample may contain an inhibitor for the enzyme, which comprises including in the incubation mixture a correction factor to block the action of any inhibitor present in the sample.

2. A method as claimed in claim 1, wherein the assay is one which involves an immune reaction between an analyte and its specific binding partner, in which immune reaction the reagent labelled with enzyme also participates.

3. A method as claimed in claim 2, wherein the inhibitor is a substance which binds to the enzyme in such a way as to inhibit participation of the labelled reagent in the immune reaction.

4. A method as claimed in any one of claims 1 to 3, wherein the enzyme is a peroxidase enzyme.

5. A method as claimed in claim 4, wherein the correction factor is a derivative or form of the peroxidase enzyme which is inert to the assay reagents and which does not otherwise affect the assay.

6. A method as claimed in claim 4 or claim 5, wherein the peroxidase enzyme is a horseradish peroxidase (HRP) enzyme.

7. A method as claimed in claim 6, wherein the sample correction factor is an apo-HRP enzyme.

8. A method as claimed in claim 6, wherein the inhibitor is an enzyme poison.

9. A method as claimed in claim 6 or claim 8, wherein the sample correction factor is an acidic HRP isoenzyme or a carboxymethylated HRP enzyme.

10. A method as claimed in any one of claims 1 to 9, wherein the assay is selected from: competition assay for total analyte, one-site immunometric assay for total analyte, competition assay for free analyte, simultaneous two-site immunometric assay for total analyte, and analyte uptake assay.

11. A method as claimed in claim 10, wherein the antibody is insolubi lised before, during or after the incubation, the portion of the analyte/enzyme conjugate bound to antibody is physically separated from the not- bound portion, and a signal generated by the bound portion or the not-bound portion is observed.

12. A method as claimed in claim 11, wherein the enzyme is caused to catalyse an oxidation reaction with the production of light via chemi luminescence.

13. A method as claimed in claim 12, wherein the assay is an enhanced luminescence assay.

14. A method as claimed in any one of claims 10 to 13, wherein the analyte is a thyroid hormone.