endocrine manual for reproductive assessment of ...

ENDOCRINE MANUAL FOR REPRODUCTIVE ASSESSMENT OF DOMESTIC AND NON-DOMESTIC SPECIES

Janine Brown, Ph.D. Sue Walker, M.S. Karen Steinman, B.S.

Conservation & Research Center Smithsonian's National Zoological Park 1500 Remount Road Front Royal, Virginia, 22630

Smithsonian's National Zoological Park Conservation & Research Center Endocrine Workbook

TABLE OF CONTENTS Page I.

Introduction

5

II.

Principles of Immunoassays

5

III.

Principals of Radioimmunoassays Essential components of an RIA Steps to development of an RIA

6 6 6

IV.

Types of Radioimmunoassays Single antibody RIA Double antibody RIA Solid phase RIA

8 8 8 9

V.

Principals of Enzymeimmunoassays Essential components of an EIA Steps to development of an EIA

10 10 10

VI.

Types of Enzymeimmunoassays Single antibody EIA Double antibody (sandwich) EIA Determining appropriate antibody dilution Effect of antibody dilution on standard curve

12 12 13 15 15

VII.

General Assay Terminology Antibody characteristics Assay characteristics Types of assay variation

16 16 17 17

VIII.

Monitoring Quality Control Why is quality control necessary? Proper use of the standard curve Monitoring standard curve parameters Monitoring assay quality Example – LH EIA control monitor

17 17 17 18 18 20

IX.

Laboratory Immunoassay Validation Parallelism Recovery/Accuracy check Extraction efficiency

21 21 22 23

X.

Biological Validation of Methods

25 2


XI.

HPLC Analysis of Metabolites

26

XII.

Immunoassay Troubleshooting Color and curve problems Loss of binding Shifted standard curve High non-specific binding Poor precision and high CVs

27 27 27 28 28 28

XIII.

Sample Collection/Storage Methods Blood Urine Feces Fecal dilutions

30 30 31 31 31

XIV. Hormone Extraction Methods Blood Urine Feces Conjugated steroid analysis Fecal extraction sheet

32 32 32 32 33 34

XV.

35 35 37 38 40 41 42

Assay Protocols Creatinine assay protocol LH EIA protocol LH EIA reagent preparation Cortisol EIA protocol Cortisol EIA stock preparation Sample sheet

XVI. Calculating EIA Results Reading the plate Calculating results from OD Average blank value Data matrix/table OD Converting data from feces Converting data from urine Tips for reading EIA plates

43 43 43 43 44 47 48 48

XVII. Assay Reagents and Supplies Steroid EIA recipes LH EIA recipes Creatinine assay recipes Assay reagents for steroid and LH EIA General supplies General equipment

49 49 50 51 52 53 54

3


Other

55

XVIII. Review of Steroid Metabolism

56

XIX. Review of Reproductive Physiology

65

XX.

68

Review of Adrenal Metabolism

4


I. INTRODUCTION Immunological techniques, like radioimmunoassays (RIA) and enzyme immunoassays (EIA), are used because they are capable of measuring small quantities of hormones. RIAs are highly sensitive and have been the most common immunological methods for hormone analysis used to date. However, an RIA laboratory needs to be licensed for the use of radioisotopic tracers, and gamma and beta scintillation detection equipment are comparatively expensive. By contrast, EIAs do not utilize radioactivity, equipment is less expensive and reagents are easy to prepare, are highly stable and have a long shelf-life. Many EIAs are now as sensitive as RIAs and so are gaining in popularity. The purpose of this manual is to acquaint the reader with both RIA and EIA techniques; however, the emphasis will be on EIA because of it’s universal adaptability and potential for development as field tests.

II. PRINCIPLES OF IMMUNOASSAYS The essence of an immunoassay is the competition between added labeled antigen (‘tracer’) and unlabeled antigen (i.e., hormone in the sample) binding to an antibody. Highly sensitive immunoassays rely on the use of a limited amount of antibody. If the primary antibody is in excess, there is little or no competition in binding between labeled and unlabeled antigen, thus no discrimination in measuring concentration of the unknown. With a limited amount of antibody, samples with higher concentrations of hormone have a greater chance of competing against labeled antigen for antibody binding than low concentration samples, and this relationship is proportional to the amount of unlabeled hormone added. Antigen-antibody binding in RIA and EIA follows the Law of Mass Action. The distribution between the bound and unbound phases is directly related to the total amount of antigen (Ag) in the presence of a fixed amount of antibody (Ab), where k1 and k2 denote the association (forwards reaction) and dissociation (reverse reaction) constants, respectively. k2 Ag + Ab AgAb k1 In the beginning, the rate of the reaction is greater in the forward direction (k1) until equilibrium is achieved; at equilibrium there will be no further net change in the concentrations on either side of the equation. The affinity constant:

K = k1/k2

High K implies that the reaction is favored in the forward direction; low K implies the reaction is favored in the opposite direction. Given a constant amount of antibody of fixed K, the ratio of bound to free antigen will be related to the total amount of antigen present. Antibodies are produced against a specific antigen by immunizing an animal and collecting immune serum. • Primary antibody (or antisera) also called the “first antibody” refers to the hormonespecific antibody; the one that was produced from immunizing an animal against the hormone to be measured. Primary antisera can be produced against large proteins by direct 5


• •

•

injection of antigen solubilized in a carrier that stimulates the immune response (i.e., Freund’s adjuvant). Immunization against small proteins or steroids (i.e., haptens) requires conjugation to immune stimulating molecules (i.e., albumins, keyhole limpet antigen, etc.). Primary antibodies are often produced in rabbits, guinea pigs and monkeys. Second antibodies, also called “non-specific antibodies” are those produced by immunizing a different species like a goat or sheep against non-specific IgGs of the species that produced the primary antisera (e.g., goat anti-rabbit antisera). Polyclonal antibodies are produced by injecting an animal with a purified or partially purified antigen and collecting the immune serum. The antisera produced contains a variety of immunoglobulins against the antigen (or antigens for partially purified immunogen) or conjugate molecule (in the case of conjugated haptens). Antibody production is limited to the life span of the immunized animal and success of reimmunizations. Monoclonal antibodies are produced by immunizing an animal (e.g., mouse) and cloning individual antibody cells to produce a single variant antibody against an antigenic determinant on the hormone. Antibody production is indefinite as long as the cloned cells are maintained (frozen in liquid nitrogen).

III. PRINCIPLES OF RADIOIMMUNOASSAY RIAs depend on the assumption that an antigen can be linked to a radioactive molecule (e.g., 3H or 125 I) and retain immunological binding activity. ESSENTIAL COMPONENTS OF AN RIA: • • •

•

Antibody: An immunoglobulin produced against a specific antigen(s). Tracer antigen: An antigen that is labeled in a manner that permits its detection. The labeled antigen, or tracer, should be structurally similar to the unlabeled antigen (or standard) and generally is labeled with 125I or 3H. Standard antigen: The standard hormone is the same antigen that was injected into an animal to induce an immune response for antibody production. Standards are used in a series of different concentrations, against which 'unknown' concentrations of the antigen contained in biological fluids can be estimated. Separation method: The ability to separate bound Ab-Ag complexes from free hormone.

STEPS TO DEVELOPMENT OF AN RIA 1. Determination of antibody titer: Based on a binding inhibition curve, the optimal ‘working’ antibody dilution is one that results in 30-40% binding of the total labeled antigen. In general, increasing the antibody concentration decreases assay sensitivity, whereas decreasing antibody concentration will increase sensitivity. 2. Development of a standard curve: Incubation of a fixed amount of tracer and antibody in the presence of different concentrations of standard (unlabeled antigen). A graph is generated which depicts the relationship between the percentage of tracer bound (relative to the maximum binding, or zero tube, see below) and relative mass of standard added to 6


each assay tube. There is an inverse relationship between percentage binding of tracer to antibody and hormone mass. 3. Optimizing the RIA: Performance of the RIA can be affected by a variety of factors, including buffers, reagent volumes, tracer or antibody concentration, incubation time and temperature, and hormone mass. 1. Incubation time and temperature: The required time to achieve maximum tracer binding, assessed by counting binding in the zero tubes. Incubation temperatures range from 4 to 37˚C depending on assay characteristics. 5. Definitions: • Total Counts (TC): Total amount of radioactivity added to each tube. Usually ~30,000 cpm. • Non-specific Binding (NSB): Amount of radioactivity that remains after the separation method has been employed in a tube containing all assay components except first antibody. The amount of radioactivity remaining reflects the efficiency of the separation method and defines the 'background' counts that will be present in every assay tube independent of 'specific' binding of antigen to antibody. • Total Binding (TB): Total amount of binding that occurs between labeled Ag and Ab in the absence of competition from standard or unknown hormone. Also called “zeros”. • Unknowns: Hormone concentrations in unknown samples determined by comparing the specific binding (relative to the zero tubes) of the unknown tubes to the binding obtained using standard hormone preparations of known mass (i.e., the standard curve). • Buffers: Include phosphate, Tris or borate with pH ranges of 5.0 – 9.0 (generally pH 7.0 – 7.6) and molarities of 0.05 - 0.1 M. Phosphate buffers are most commonly used and have the advantage of the pH not being affected by temperature (unlike Tris buffers). Buffers generally are preserved by addition of sodium azide (0.1%) or thimerosal/merthiolate (0.1%). Proteins like bovine serum albumin (BSA) often are added to stabilize antisera and reduce non-specific binding. BSA can vary from batch to batch and has a limited shelf-life (~1 year refrigerated).

7


IV. TYPES OF RADIOIMMUNOASSAYS SINGLE ANTIBODY (CHARCOAL-DEXTRAN SEPARATION) RIA This is a common assay system used for measurement of steroid hormones. A hormone-specific antibody (or first antibody) is added to a test tube with sample (or standards) and a 3H-labeled tracer. The labeled and unlabeled antigens compete for binding sites on the antibody during the incubation phase. The unbound antigens are removed by adsorption to dextran-coated charcoal. Only free antigen is removed because the dextran coats the charcoal and blocks the larger pores to prevent Ab-Ag adsorption. After centrifugation the bound complexes are left in the supernatant which is decanted into vials containing scintillation cocktail for counting in a beta counter. The method is inexpensive, but tends to be less sensitive than other methods. Example: Y First antibody G 3H-labeled steroid 1. Incubate O O O + G G G G O O O + G G G G

Y Y Y Y Y

O Unknown (sample steroid) 7 Charcoal-dextran 2. Binding to antibody O O O O G G G G G O O G G G

Y Y Y Y Y

3. Adsorption of free steroid O7G O7O O7G O O G G G Y Y Y Y Y

4. Centrifuge and decant supernatant containing Ab-Ag complexes into vial with scintillation cocktail to enhance 3H signal. 5. Count radioactivity of liquid phase in a beta scintillation counter. The higher the radioactivity, the lower the sample hormone concentration. DOUBLE ANTIBODY RIA This technique relies on the precipitation of bound complexes with a second antibody that is specific to the IgG of the species in which the first antibody was made. In general, the first step is the incubation of first antibody with labeled (usually an 125I-labeled tracer) and unlabeled (sample) antigen. Tracer may be added simultaneously with unlabeled antigen or after the antibody and unknown antigen have incubated for a period of time. Delayed tracer addition can increase assay sensitivity. ‘Normal’ serum from the same species as the primary antibody, but not containing antigen-specific antibodies, must be included to facilitate the formation of Ab-Ag complexes. The Ab-Ag complexes are precipitated after incubation with the second antibody and centrifugation. Polyethylene glycol (PEG) can be added with the second antibody to improve precipitation efficiency because PEG decreases Ab-Ag complex solubility. After centrifugation, the supernatant is discarded and radioactivity in the pellet is counted in a gamma counter. The double antibody technique has good sensitivity, but requires an additional incubation period that can prolong assay time. It can be used for both steroid and protein assays.

8


Example: Y First antibody (e.g., rabbit IgG)

O Unknown (sample) G Labeled hormone Second antibody (e.g., goat anti-rabbit IgG)

1. Incubate

2. Binding to first Ab

O O OG G G G

O O O O G G G G G

O O OG G G G

O

O G G G

3. Incubate with second Ab O O G G O O Y GG G YY YY

Y Y Y Y Y

Y Y Y Y Y

4. Centrifuge to separate free hormone from bound complexes. 5. Decant the liquid containing free hormone (supernatant) and count the pellet containing AgAb1Ab2 complexes in a gamma scintillation counter. The higher the radioactivity in the pellet, the lower the sample hormone concentration. SOLID-PHASE RIA The hormone-specific antibody (first antibody) is attached to the solid phase (i.e., adhered to the wall of the plastic assay tube). Labeled (125I-labeled tracer) and unlabeled (sample or standard) antigens are incubated in the tube and unbound antigen is removed by decanting the liquid phase. The radioactivity remaining in the tube is counted in a gamma counter. This technique can be used for protein and steroid hormones. The advantage to this technique is that centrifugation is not required to separate bound complexes from free antigen. Example: Y First antibody

O Unknown (sample hormone)

1. Incubate with sample and tracer

G 125I-labeled hormone (tracer) 2. Decant supernatant containing free hormone

U

U

3. Count tubes in a gamma scintillation counter. The higher the radioactivity, the lower the sample hormone concentration.

9


V. PRINCIPLES OF ENZYMEIMMUNOASSAY EIA is also known as ELISA (Enzyme Linked ImmunoSorbent Assay). EIAs depend on the assumption that an antigen can be linked to an enzyme and retain both immunological and enzymatic activity in the resultant conjugate. The soluble antigen or antibody must also be linked to an insoluble phase in a way in which the reactivity of the immunological component is retained. ESSENTIAL COMPONENTS OF THE EIA: • • • • •

• •

•

•

Solid Phase: The solid phase is the polystyrene microtiter plate. Antibody: An immunoglobulin produced against a specific antigen. Polyclonal antibodies must be affinity purified for EIA. Coating buffer: The antibody is diluted with an alkaline buffer, usually a carbonate/bicarbonate buffer of pH 9.6, which causes it to passively adsorb to the well of the microtiter plate. Wash solution: Each incubation is terminated by a washing step. The wash removes all unbound components from the plate. Enzyme conjugate (tracer): The enzyme conjugate is the component of the assay that permits detection of antigen concentration. For direct, single antibody EIAs, a common enzyme conjugate is hormone conjugated to horseradish peroxidase (HRP). For double antibody sandwich EIAs, the enzyme conjugate complex is a biotin labeled hormone that binds to peroxidase-labeled strepavidin. Assay buffer: Phosphate or Tris buffers of pH 7.0 are commonly used. Sodium azide cannot be used in buffers for single antibody EIAs because the HRP is inhibited by azide. Sample dilutions, standards and enzyme conjugate are made up in assay buffer. Standards or unlabeled antigen: The standard is usually the same antigen that was used to make the antibody and the same as the enzyme conjugate, or is structurally similar so that it crossreacts with the first antibody. Standards are used in a series of known concentrations against which unknown concentrations of antigen in the sample can be measured and calculated. Substrate: The substrate reacts with the bound enzyme conjugate and changes color. It consists of three components: buffer, chromagen, and catalyst. The buffer has an acidic pH and is either citric acid or phosphate citrate buffer. The chromagen is the color changer and is usually azino-bis-3-ethyl benzthiazoline-6-sulfonic acid (ABTS) or tetramethylbenzadine (TMB). ABTS turns a green color and TMB turns blue. The catalyst is what causes the reaction, via oxidation-reduction, and is hydrogen peroxide or sodium perborate. Stop solution: Sulfuric acid solution that stops the substrate reaction and allows the plate to be read at any time. It is used primarily in the double antibody EIA. It causes the blue substrate to turn yellow.

STEPS TO DEVELOPMENT OF AN EIA 1. Determination of antibody titer. An appropriate antibody titer is one that results in adequate color change while retaining good sensitivity. In general, increasing the antibody concentration increases the enzymatic color change, but decreases assay sensitivity. 10


2.

3.

4.

5.

Decreasing antibody concentration (more dilute) increases sensitivity, but the color change is less. Determination of enzyme conjugate dilution. Increased enzyme conjugate concentration results in a stronger color change but decreased assay sensitivity, whereas decreased conjugate concentration increases assay sensitivity but reduces color intensity. An appropriate combination of antibody and enzyme conjugate results in adequate color intensity with high assay sensitivity. Development of standard curve. Incubation of a fixed amount of enzyme conjugate and antibody in the presence of different concentrations of standard (unlabeled antigen). A graph is generated that depicts the relationship between the percentage of bound enzyme conjugate (relative to the maximum binding of the enzyme conjugate, zero well) to the concentration/mass of the standard added. The relationship of the percent binding and the standard mass is inversely proportional. Time and temperature of incubation. Incubation time can be decreased with increased temperature but antibody-antigen binding and substrate-enzyme conjugate binding can decrease if the temperature is too high. Definitions: • Total Binding (TB): Maximum binding of enzyme conjugate (labeled antigen) to the antibody in the absence of competition from unlabeled antigen (standard or unknown sample). Also called “zeroes” and should have the highest optical densities. It is assumed to be 100% binding of the enzyme conjugate. Most assay systems attempt to reach optical densities of 0.8-1.0 in the zero wells. • Unknowns: Unknown hormone concentrations in samples are determined by comparing the specific binding of the samples to the binding obtained from standard hormone concentrations of known mass. • Non-specific binding (NSB): Amount of binding that occurs in the plate that is NOT due to the antibody, but to other components of the assay. It also accounts for the amount of interference the plastic bottom of the plate produces when the light from the plate reader is refracted. The NSB wells should be less than 10% of the maximum binding wells (zeroes). The NSB results are subtracted from the sample/standard results (this is done by the microplate reader). NSB can be reduced by adding a protein blocking buffer after antibody coating, adding detergent to the assay buffer or increasing the number of washes.

11


VI. TYPES OF ENZYMEIMMUNOASSAYS SINGLE ANTIBODY EIA (e.g., cortisol) A hormone-specific antibody (or first antibody) is passively adsorbed (i.e., coated) to a polystyrene microtiter plate. Unabsorbed antibody is washed away. Known (standards) and unknown (samples) concentrations of hormone (unlabeled antigen) and the hormone-specific enzyme conjugate (HRP) (the labeled antigen) are added to the well. The labeled and unlabeled antigens compete for binding sites on the antibody during the incubation phase. The unbound components are washed away. The substrate is added and reacts with the bound enzyme conjugate and changes color. The more color change in the well, the more enzyme conjugate is bound, meaning less hormone. The relationship of color to hormone concentration is inversely proportional. The zero wells contain only enzyme conjugate so there is no competition for antibody binding. The zero wells represent the maximum binding of the labeled antigen and, hence, have the most color change. Example:

Y

First antibody Free hormone

HRP-labeled hormone Substrate

1) Antibody binding to solid phase (known as coating)

YYYYYYYY 2) Competition for antibody binding sites by labeled and sample hormone

YY Y Y YY YY 3) Wash away excess unbound hormone

YYYYY YYY 4) Substrate binding to HRP-labeled hormone

YYYYY YYY YYYYYY YY

Y Y Y Y YY Y Y

Zero standard Maximum color

High standard Minimum color

12


SECOND ANTIBODY (SANDWICH) EIA (e.g., luteinizing hormone) In this antibody system a ‘second antibody’ that recognizes the first antibody is used to coat the microtiter plate. Following incubation, unbound second antibody is decanted and a blocking buffer, usually containing a protein such as BSA is added to reduce non-specific binding. After incubation, plates are washed and any unbound components removed. The first antibody and unlabeled antigens (sample or standards) are added to the wells. The first antibody binds to the second antibody and the free hormone then binds to the first antibody. Reagents are allowed to incubate followed by incubation with the biotin-labeled antigen. Plates are washed, removing any unbound biotin-labeled antigen, and streptavidin-peroxidase is added. The distinguishing feature of the avidin-biotin system is the extremely high affinity of the avidin (from egg white or a bacteria -Streptomyces avidinii) for biotin (a water-soluble B vitamin). The speed and the strength of binding between these two molecules is used to provide an amplification of the enzyme signal. This binding also is not influenced as much by extreme temperatures or pH levels. The peroxidase enzyme is incorporated into the streptavidin and after binding to the biotin, forms the enzyme conjugate complex. Following plate washing substrate (TMB) is added. The chromagen within the substrate reacts with the bound enzyme conjugate and changes color. Similar to the direct single antibody competitive EIA, the more color change in the well the more enzyme conjugate bound, meaning less hormone. The relationship of color to hormone concentration is inversely proportional. Example:

Y

Second antibody

BSA

First antibody Free hormone

Biotin-labeled hormone Substrate Streptavidin-peroxidase (enzyme)

1) Second antibody binding to solid phase (coating)

2) BSA is added to block non-specific binding sites

13


3) Addition of hormone specific (first) antibody, biotin-labeled hormone, and free hormone

Y Y Y

Y

4) Wash away unbound components

Y Y Y Y 5) Addition of streptavidin-peroxidase (enzyme)

Y Y Y Y 6) Addition of substrate

Y Y Y Y

Y Y Y Y

Y Y Y Y

Zero Maximum color

High standard Minimum color

14


HOW TO DETERMINE APPROPRIATE ANTIBODY DILUTION FOR RIA AND EIA 100

80

60 % Binding (TB/TC) Want Ab to bind ~30% of labeled hormone

40

20 Use Ab at 1:120,000 dilution 0 1

10

100

1000

Ab Dilution (x 1000)

Appropriate first antibody concentration is determined by running a titration curve which involves incubating serial dilutions of first antibody with a constant amount of tracer. Calculate the % binding of the antibody to tracer and plot as a linear-log curve. The best antibody dilution (one that provides good sensitivity with adequate detectibility) is ~30% (range 20-50%).

EFFECT OF ANTIBODY DILUTION ON STANDARD CURVE CHARACTERISTICS 120

Correct

100

too much not enough

80

% B/TB 60 40 20 0 1

10

100

1000

Standard Concentration

Inappropriate first antibody concentration will result in a poor standard curve that is too ‘flat’ on the top or bottom portion. Too much antibody also will reduce assay sensitivity.

15


VII. GENERAL ASSAY TERMINOLOGY ANTIBODY CHARACTERISTICS Sensitivity:

The ability to detect small quantities of antigen. It is the lowest concentration of antigen that can be statistically distinguished from a sample with no antigen. - objective determination - calculated as the value 2 SD from the mean response of the blank or zero (Bo) tube. - subjective determination - the value at 90% or 95% of maximum binding.

Specificity:

The ability of the antibody to discriminate between antigens.

Crossreactivity:

The ability of an antibody to have immunoreactivity with a more than one antigen. Crossreactivity of an antibody with other hormones is determined using binding inhibition curves. It is expressed as the standard concentration at 50% binding divided by the concentration of the competitive antigen at 50% binding, expressed as a percentage.

Examples: • • •

Hormone C is detected but the antibody is not as specific for it. The crossreactivity is less than 100%. The antibody for progesterone crossreacts better with hormone D than the standard. The crossreactivity is over 100%. Hormones A and B only bind the antibody at high concentrations and crossreactivity cannot be calculated. Crossreactivity of Different Antigens on a Progesterone EIA Progesterone Standard Hormone A Hormone B Hormone C Hormone D

100 80 60 % B/TB 40 20 0 0.1

1

10

100

Standard Concentration (ng)

16

1000


ASSAY CHARACTERISTICS Accuracy: The degree to which the measured concentration corresponds to the true concentration of a substance. This is related in part to the specificity of the assay. Precision: Refers to the repeatability of a measured value or the consistency of results. It is a measure of random error defined as the variation among replicate measurements of a defined sample. It is expressed as the Coefficient of Variation (%CV) which is the (Standard Deviation/Mean)*100. TYPES OF ASSAY VARIATION Intra-assay variation: The variation within assays determined by repeated analysis of samples, typically low, medium and high, within a single assay. Inter-assay variation: The variation between assays determined by repeated analysis of the same sample in several assays. These samples are called controls and should bind at 30% and 70%. Note: A sample can be precise, but not accurate. This occurs when the measured value deviates from the true value as a result of using a non-specific first antibody or to systematic error, like improperly calibrated weighting or measuring equipment, improper technique (pipetting) or sample problems.

VIII. MONITORING QUALITY CONTROL WHY IS QUALITY CONTROL NECESSARY? • • • • •

Proper validation and standardization of an assay is only the first step towards establishing a reliable endocrine monitoring program. Subsequent assessment of assay quality and consistency is absolutely necessary to assure the biological relevance of results. For every assay system there is an inherent level of error which must be accepted. A quality control program indicates when that level of error becomes unacceptable. Quality control samples only have value if their analysis provides a reasonable confidence in data for the whole assay, or reflects a true error in the method. It is an ongoing process; the value of quality control values increases with time.

PROPER USE OF THE STANDARD CURVE •

Assay range (the 20-80% rule): A subjective method of sample exclusion which eliminates the use of values in which the binding is less than 20% or greater than 80% of maximum binding. All outliers are re-analyzed after appropriate modification (i.e., taking more or less sample to the assay). This rule is based on the assumption that most standard curves are linear between 20 and 80% binding, and therefore the dose response is linear. The binding cut-off limits may vary, however, and should be appropriate for each assay. 17


MONITORING STANDARD CURVE PARAMETERS •

•

•

Total binding (%B/Bo): Indicates the maximum binding of label in the given assay system. Given that the assay parameters are unchanged, the maximum binding should remain relatively constant from assay to assay. A decrease in binding suggests one or more assay factors are not optimized (see Troubleshooting). Non-specific binding (%NB/Bo): The amount of binding due to factors other than specific antibody binding. It should remain constant from assay to assay. It also should be kept to a minimum, generally less than 5% of the total counts added. An increase in NSB again suggests one or more assay factors are not optimized (see Troubleshooting). Effective dose values (ED20, ED50, ED80): Gives a good estimate of the overall ‘shape’ of the curve. The ED50 is especially useful for indicating changes in assay sensitivity. The ED20 and ED80 can also indicate changes in the slope of the curve.

MONITORING ASSAY QUALITY •

•

How is quality control monitored? Assay consistency is monitored by analyzing ‘internal control samples’ which are treated as unknowns, but are run in every assay. - how many to run? Usually 2 - 3 controls, assayed in duplicate or triplicate. Much of this depends upon the variability of the assay. - what concentrations to run? Controls should provide an estimate of variability over the working range of the standard curve. Recommend controls be run at ~30% (high concentration), 50% (medium) and 70% (low) of maximum binding. - what samples can be used as controls? Almost any biological material containing the relevant antigen, including serum pools, fecal extracts, purified steroids diluted in buffer or other fluid (provided it does not interfere with the assay), pituitary extracts, urine pools, purchased standards, etc. i. prepare a large pool projected to last for several years. ii. divide the material into small aliquots to avoid repeated freeze-thawing. iii. keep the aliquots in a safe freezer (preferably one with a temperature alarm). iv. store all materials, including samples, in a no frost-free freezer. v. do not allow controls to run out before making up new controls. Assay coefficients of variation: (%CV = standard deviation/mean times 100) - intra-assay CV - determines the within assay error (the error associated with running the same sample in one assay). Most accurately determined by calculating the variation in assaying multiple replicates of one sample throughout the assay (e.g., n = 10 replicates). More typically, the average intra-assay CV is calculated from the internal controls assayed at the beginning of the assay. A third method is to calculate the average CV of all unknowns run in an assay. If more than one assay has been run for a particular study, then the mean intra-assay CVs for those assays should be averaged. - inter-assay CV - determines the between assay error (the error observed when the same sample is run in different assays). Determined by calculating the variation in values for samples run in every assay. Within a study can calculate the individual inter-assay CVs for each internal control and then average those numbers.

18


•

•

Causes of variation: - intra-assay CV - generally is the result of the presence of unequal amounts of sample, tracer or antibody (i.e., poor pipetting, or incomplete mixing of sample or reagents), but also can be due to inconsistencies in counting or decanting. - inter-assay CV - caused by system-related problems like reagent instability, procedural variation or changes in standards (do not allow old standards to run out before making up new standards!). What is an acceptable level of error? - sample rejection - a level of error needs to be established that objectively determines when a sample needs to be re-analyzed. However, the amount of error tolerated is subjective and determined, in part, by how critically the data need to be interpreted. For example, the error rate would be lower in a human endocrinology laboratory where health status and potential treatments hinge on accurate measurements. Conversely, more error might be tolerated in assays conducted to define hormonal trends (i.e., the absolute values are not as important as the profile). - rules of thumb for defining acceptable levels of error i. fixed percentage of the CV - often designated as 10% such that all samples with a CV >10% are re-analyzed. ii. assay rejection criteria - often subjective, but can involve exclusions based on control values falling within 2 SD of the mean of previous values, or a set proportion of sample CVs being below 10%.

120 100

The same 5% error in binding results in a greater %CV at each end of the curve

80

% B/TB 60 40 20 0 1

10

100

1000

10000

Standard Concentration • •

Proportional error: The CVs are higher and dose calculations are not proportional at the extreme ends of the standard curve, due to the non-linearity of the curve (see above figure). External quality control procedures: Commonly used in clinical laboratories to ensure that the quality of data are independent of the laboratory which generated it. Involves analysis of samples provided by external sources. 19


LH EIA Control Monitor

Date 4/22/02 6/27/02 6/28/02 6/29/02 6/30/02 7/1/02 7/1/02 7/1/02 7/1/02 7/3/01 7/3/01 7/3/01 7/3/01 7/19/02 7/19/02 7/19/02 7/19/02 7/31/02 7/31/02 10/23/02 10/23/02

98.0 100.2 96.6 97.8 100.3 98.4 95.2 103.7 96.6 98.8 102.4 98.6 94.2 97.8 100.0 100.0

Mean 69.7 Count 21.0 Std Dev 3.2 CV 4.6 Mn +1SD 72.9 Mn-1SD 66.4 Min 63.0

55.7 21.0 5.9 10.6

47.5 105.6 97.9 21.0 21.0 21.0 3.6 5.2 2.8 7.6 4.9 2.9

98.6 20.0 2.5 2.5

Max

75.7

500.0

100.0 92.6 98.5 99.0

250.0

100.8 94.6 98.6 100.5

125.0

100.8 93.7 97.4 97.2 97.1 95.1 98.5 93.6 96.1 98.8 98.1 97.1 104.0 95.9 101.3 103.1 96.7 94.6 97.4 100.0 100.0

62.5

QCL %B 53.7 48.0 46.6 47.8 44.8 44.5 49.1 50.7 48.2 47.0 45.7 46.3 48.7 42.0 46.3 43.0 40.3 49.8 48.4 51.4 55.1

31.3

QCH Dose 64.4 57.4 58.6 57.1 58.6 68.5 59.8 44.8 61.3 50.0 52.1 54.4 62.7 53.4 53.3 47.4 50.7 59.6 50.1 54.0 50.8

15.6

7.8

QCH %B 74.8 69.8 66.4 70.5 68.3 63.0 68.4 73.4 67.3 71.6 69.7 71.9 68.8 64.8 68.1 69.6 66.7 68.6 72.3 73.4 75.7

Tech sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw sw

QCL Dose 114.9 106.6 104.0 106.6 114.0 114.1 102.5 94.3 104.0 102.8 102.3 108.7 107.7 104.6 106.5 102.3 112.0 105.8 104.4 103.2 96.9

3.9

Percent Binding For Standards 2.0

Controls

99.0 101.0 91.5 99.2 96.7 98.0 93.9 104.1 96.5 96.6 101.1 97.2 96.3 97.7 102.0 103.0

101.2 88.7 90.4 92.2 91.6 90.9 94.0 88.7 91.1 92.6 91.0 94.7 94.2 87.4 88.8 91.5 85.3 88.1 88.2 91.7 91.4

89.6 82.2 80.6 83.9 81.0 80.7 83.0 79.8 82.0 83.4 81.8 83.1 87.1 77.7 79.7 78.6 78.2 81.0 81.7 85.2 86.2

74.7 67.5 65.1 67.1 67.7 66.0 67.2 64.7 66.4 63.4 64.6 66.8 69.9 60.8 64.0 64.0 62.8 67.9 68.1 70.4 69.8

51.9 42.2 42.0 42.7 42.5 42.2 43.0 42.0 42.4 41.6 38.7 41.4 43.9 36.3 44.9 35.4 36.8 44.8 42.6 45.3 48.0

23.6 16.9 14.8 16.1 15.7 15.7 17.0 16.3 15.8 16.7 15.2 16.4 16.7 13.3 12.7 13.1 13.1 20.6 18.8 17.7 17.4

9.2 6.4 5.9 6.2 5.7 5.2 6.7 6.0 6.0 6.3 5.1 6.1 4.8 3.1 3.8 4.0 4.3 12.8 11.3 5.6 4.0

98.2 20.0 3.2 3.3

91.1 21.0 3.3 3.6

82.2 21.0 3.0 3.6

66.6 21.0 3.1 4.6

42.4 21.0 3.7 8.8

16.4 21.0 2.5 15.5

6.1 21.0 2.4 38.7

61.6

51.1 110.8 100.8 101.1 101.4 94.4

85.2

69.7

46.1

18.9

8.5

49.7

43.9 100.4 95.1

79.2

63.5

38.7

13.8

3.7

44.8 68.5

40.3 94.3 93.6 94.2 91.5 85.3 77.7 55.1 114.9 104.0 103.7 104.1 101.2 89.6

60.8 74.7

35.4 51.9

12.7 23.6

3.1 12.8

96.2

20

94.9

87.8


IX. LABORATORY IMMUNOASSAY VALIDATION PARALLELISM Definition: Parallelism is a way of determining if the assay is actually measuring what it should be measuring. It can also tell what dilution of sample to use for the assay. •

Samples to test should reflect the range of normal concentrations expected (i.e., estrous cycle and pregnancy, seasonal samples).

•

Pool an equal amount of urine or fecal extract from each sample and dilute serially twofold in assay buffer (neat, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64, etc.). For example add 200 ul of buffer into tubes labeled from 1:2 to 1:8192 - then take 200 ul of the neat sample and mix with the 200 ul in the “1:2” tube, vortex the “1:2” - then take 200 ul out of the 1:2 and place in the “1:4” and so on. This will result in 200 ul in each tube except the last dilution which will have 400 ul.

•

Plot the % binding of samples by choosing an arbitrary concentration for the neat sample and halving the concentration for each dilution.

•

If the sample curve parallels the standard curve the sample hormone is immunologically similar to the standard and can be measured proportionately.

100 Parallel displacement 80

1:64 1:32

% B/TB

60 Use 1:16 40

1:8 1:4

20

1:2 neat

0 0.1

1

10

Standard curve

100

1000

Standard Concentration or Dilution

The best (most accurate) dilution to run samples is at ~50% binding (see above figure). The pool shows parallelism with the standard curve and is therefore valid for use in the assay. Samples should be run at a 1:16 dilution based on the binding inhibition observed at ~50%.

21


120 No displacement

100 80

% B/TB

60

Low concentration Non-parallel displacement

40 20

Standard curve

0 0.1

1

10

100

1000

Standard Concentration or Dilution

In the graph above, samples that produced non-parallel or no displacement cannot be run in this assay system because they do not demonstrate immunoactivity of endogenous antigen similar to the assay standards. The sample with a low concentration of immunoactive antigen demonstrated limited parallelism, but this type of sample could only be run ‘neat’ and there likely would be problems with limited detection within the working range of the assay (i.e., 90% of gonadal steroids into feces, whereas baboons excrete >80% of gonadal steroids into urine. Steroids vary in the extent to which they are metabolized before excretion, both within and among individuals and species. The time course of steroid excretion and the degree to which steroids are excreted in urine or feces are determined by infusing unlabeled or radiolabeled steroid and quantifying hormonal metabolites in excreta. The lag-time from steroid production/secretion to appearance in excreted urine is generally