A short history, principles, and types of ELISA, and our laboratory experience with peptide/protein analyses using ELISA

Highlights

• The invention of ELISA and its types are presented in a sequence.

• Laboratory experiences with peptide analyses are provided.

• The need to use protease inhibitors to protect peptides in peptide analyses is emphasized.

• It is underlined that ELISA kit manufacturers have to standardize kits.

• The significance of ELISA in peptide analyses is noted.

Abstract

Playing a critical role in the metabolic homeostasis of living systems, the circulating concentrations of peptides/proteins are influenced by a variety of patho-physiological events. These peptide/protein concentrations in biological fluids are measured using various methods, the most common of which is enzymatic immunoassay EIA/ELISA and which guide the clinicians in diagnosing and monitoring diseases that inflict biological systems. All the techniques where enzymes are employed to show antigen–antibody reactions are generally referred to as enzymatic immunoassay EIA/ELISA method. Since the basic principles of EIA and ELISA are the same. The main objective of this review is to present an overview of the historical journey that had led to the invention of EIA/ELISA, an indispensible method for medical and research laboratories, types of ELISA developed after its invention [direct (the first ELISA method invented), indirect, sandwich and competitive methods], problems encountered during peptide/protein analyses (pre-analytical, analytical and post-analytical), rules to be followed to prevent these problems, and our laboratory experience of more than 15 years.

Introduction

Quantitative analytical methods that show antigen–antibody reactions through the color change obtained by using an enzyme-linked conjugate and enzyme substrate and that serve to identify the presence and concentration of molecules in biological fluids are generally called enzyme immunotests [enzyme immunoassay (EIA), enzyme-linked immunosorbent assay (ELISA)] [16]. Very low-concentration molecules such as peptides/proteins, hormones, vitamins and drugs display a high level of specificity against antibodies or antigens developed for them [12], [16], [23]. This is because it is almost impossible for an antibody to be bound to a molecule other than its own antigen. Thus, this method can be used to measure even substances in very low concentrations with hardly any risk of interference. In other words, when we have the antigen which we know to be specific to a certain substance, we can identify the type and amount of its antibody and when we have the antibody, we can find out its specific antigen and the amount of antigen, using this method. All techniques and methods of analysis using enzymes to show antigen–antibody reactions are generally referred to as enzyme immunotests [12], [16].

Although the basic principle of ELISA and radioimmunoassay (RIA) techniques dates back to 1941 [11], RIA method was first used by Yalow and Berson in 1960s to measure the endogenous plasma insulin level [41]. In fact, ELISA method was invented simultaneously by two research teams at the same time [13], [39]. However, ELISA method was pioneered largely by the Swiss scientists Engvall, and Perlmann who died in 2005 [13]. These two researchers developed the ELISA method in 1971 by modifying the RIA method [13]. In other words, they devised the immunological ELISA method by conjugating the tagged antigen and antibody radioisotopes in RIA with enzymes rather than radioactive iodine 125. They employed this new method to determine the levels of IgG in rabbit serum [13]. In the same year, a different research team succeeded in quantifying human chorionic gonadotropin amounts in the urine by using horseradish peroxidase (EC 1.11.17) enzyme with the EIA method [39]. The researchers applied for a patent both in the USA and Europe.

Following the invention of ELISA, a number of researchers used it: Carlson and colleagues in 1972 [10], Holmgren and Svennerholm in diagnostic microbiology in 1973 [15], Ljungstrom and colleagues to identify the presence of trichinosis in parasitology in 1974 [26], and Voller et al. to diagnose malaria in 1975 [40]. Bishai and Galli, Leinikki et al. and Ukkonen et al. made use of the ELISA method to identify infections caused by influenza, parainfluenza and mumps viruses in 1978, 1979, and 1981, respectively [6], [22], [38]. In 1980, Siegle et al. modified the ELISA test and incorporated microtitration plates to identify the concentrations of various hormones, peptides, and proteins [35]. The method which has found different fields of application and grown beyond infancy over time has become a routinely used method in research and diagnosis laboratories around the world.

The antigen utilized in the ELISA method is bound to a solid phase. Tubes and microplates made of rigid polystyrene, polyvinyl and polypropylene are used as the solid phase. The microplates used must be able to appropriately adsorb the antigen and the antibody, but not adsorb the components in the other phases [13], [41]. The enzymes that can be employed in ELISA include beta galactosidase, glucose oxidase, peroxidase, and alkaline phosphatase. Alkaline phosphatase can be stored at 4 °C with its conjugate sodium azide. Alkaline phosphatase and P-nitro-phenyl phosphate are used as substrates, are available in safe tablet forms, and produce a yellow color in positive reactions. For the peroxidase conjugate, 5 amino salicylic acid and orthophenylenediamine are used as the substrates and the production of a brown color is considered a positive reaction. If beta galactosidase is used, the sample must be read in a fluorometer. The catabolic effects of enzymes determine both the acceleration and the specificity of the immunological reaction during the enzyme-substrate reaction [12]. The enzyme-substrate reaction is usually completed within 30–60 min. The reaction can be stopped using sodium hydroxide (NaOH), hydrochloric acid (HCl) or sulfuric acid (H₂SO₄) [16]. The results are read on a spectrophotometer and at 400–600 nm depending on the characteristics of the conjugate used.

Enzymatic immunoassay methods are considered under two general headings as homogeneous enzymatic immunoassay methods and heterogeneous enzymatic immunoassay methods [27] (Fig. 1). In the homogeneous enzymatic immunoassay methods, enzymes become inactivated when they are bound to the antibody, and thus, there is no stage (washing) where the antigen is separated from the medium. Homogeneous enzymatic immunoassay method is usually employed to measure substances in small quantities, like therapeutic drugs [27]. Homogeneous method is expensive and has low sensitivity. The only advantage it possesses is its ease of use.

As heterogeneous enzymatic immunoassay methods are more commonly used [27], the methods and types of this method are detailed in the following paragraphs. In this method, in order to the prevent interference of any molecule in the medium with it after the binding of the antigen and the antibody, the antigen–antibody complex is bound to the walls of the experiment tubes and anything other than the complex is removed from the medium through washing procedures. In other words, in heterogeneous enzymatic immunoassay methods, it is essential to have a washing stage to separate the bound antigen from the free antigen after the antigen–antibody interaction. Since the heterogeneous method is more sensitive than the homogeneous one, it is more commonly used. ELISA is a heterogeneous immunoassay technique used to detect specific antibodies and soluble antigens, and since the structure and the characteristics of the substances to be measured are not always the same, a variety of ELISA types have been developed to increase the specificity of measurement [27]. Schematic description of the homogeneous enzymatic immunoassay and heterogeneous enzymatic immunoassay methods is presented in Fig. 2.

The technique was simultaneously developed in 1971 by Engvall and Perlmann [13] and by Van Weemen and Schuurs [39], the technique pioneered other ELISA types. Direct ELISA method is suitable for determining the amount of high molecule-weight antigens. The surface of the plate is coated directly with the antibody or antigen. An enzyme tagged antibody or antigen enables the measurement. Incubation is followed by washing which removes the unbound antigens or antibodies from the medium. Then the appropriate substrate is added to the medium to produce a signal through coloration. The signal is measured to determine the amount of the antigen or antibody [12], [16].

The technique was developed in 1978 Lindström and Wager [25], who were inspired by the direct ELISA method. The researchers reported measuring porcine IgG using this method. The reason why this method is called the indirect method is that what determines and separates the antigen to be measured is not the primary antibody, but another antibody that is placed in the medium. In this method, the diseased serum is added to the antigen-coated wells and the plates are incubated. During this incubation, the antibodies formed against the antigens in the diseased serum plaque produce an antigen–antibody complex. In order to render the antigen–antibody complex visible, a secondary antibody that recognizes the antibody in the serum and that is tagged with the enzyme is added. Then the substrate of the enzyme is added to the medium to produce color and the concentration is determined. This method utilized to identify antigens is used more commonly in endocrinology.

The technique was developed in 1977 Kato and his co-workers [20]. In this ELISA method, the wells are coated with a capture antibody and blocked. The sample is added to the microplate wells coated with the antibody; then, the plate is incubated for some time and washed. Washing removes the unbound antigens. When the antigen specific to the bound antibody is found, these antigens cannot be removed. Following the washing step, antibodies that are tagged with the enzyme specific to the antigen are added and incubated. After incubation and washing, if there are antigens in the medium, these cannot be removed as the enzyme-tagged antibodies are bound to them. In order to reveal the enzyme activity, enzyme substrate is added to the medium and coloration is ensured. Coloration shows a positive result, while lack of coloration indicates lack of enzymes, or a negative result. As the relevant protein is stuck between two antibody molecules, this method is called Sandwich ELISA. Sandwich ELISAs have been reported to be 2–5 times more sensitive than all other ELISAs.

The technique was developed in 1976 Yorde and his coworkers [42]. In this method, the surface of the wells is coated with the antigen-specific antibody or antibody-specific antigen. The sample to be measured and the enzyme-tagged antigen or antibody are placed into the well simultaneously. The tagged and untagged antigen (patient antigen) or antibody molecules compete with each other to bind to the antibody or antigen in the wells. After the wells are washed and enzyme substrate is added, the resulting coloration enables quantifying the concentration. There is an inverse proportion between the analyte concentration and the intensity of resulting coloration. To put it differently, when the amount of the antigen or the antibody analyzed in the serum is low, high absorbance is obtained, while greater quantities produce low absorbance.

In recent years, ELISA test has been used very commonly in peptide and protein analyses [12]. This is a sensitive and specific test that rapidly produces results. It has a wide area of application due to its ease of use and speed [12]. Besides, it is more practical as there is no need to study two serum samples. The ELISA test is almost as sensitive as RIA and does not require special equipment or radioactive labels. However, its reliability is low, in comparison to RIA [31]. The advantages and differences of ELISA tests are presented in Table 1.

We have used the ELISA kits of many companies in our laboratory. The major manufacturers of ELISA kits are presented in Table 2 [23]. However, as the most commonly used ELISA kit in our laboratory is the enzyme immune assay kit of Phoenix Company, general protocol steps for this product are presented [19]. Although it is simple, the ELISA method consists of a number of steps. For a better understanding of the method, the steps such as the preparation of the washing solution, analysis, reading, etc. are explained under different subheadings. The materials needed for these experiments can be seen in Table 3.

For the preparation of standards, 5 propylene eppendorf tubes of 1.5 mL should be placed on a rack and numbered. Then each of these should be added a diluent in the amount identified in the manufacturing firm's catalog. After that, an appropriate concentration of the stock standard should be taken using an automatic pipette, added to tube 5, and mixed with a vortex for about 10 s. Then, an appropriate concentration of this number 5 standard should be taken with the automatic pipette whose head has been changed and put into tube 4. The process should be repeated up to tube 1, using a new pipette head in each transfer, and the standards should be prepared through this process of serial dilution.

In order to dilute 50 mL of 20-fold concentrated washing solution provided by the manufacturing firm, 950 mL distilled water should be put into a flask. Then by adding the 50 mL 20-fold concentrated washing solution to the distilled water, you can have a 1-fold concentrated washing solution.

(1) The blank well chambers in the plate should be kept empty; (2) An adequate amount of EIA buffer should be added to NSB wells; (3) Appropriate quantities of standard solution should be added to standard wells; (4) A suitable amount of the relevant quality control solutions should be placed in Quality Control 1 (QC1) and QC2 wells; (5) Sample wells should be added the relevant sample diluents; (6) All wells except the blind should be added an amount of primary antibody; (7) The plate should be covered with parafilm and left to incubate at room temperature for a period of time; (8) All wells should be washed in an ELISA washer using an appropriate amount of washing buffer; (9) They should be shaken with an Orbital shaker (350 rpm) for 5 min at room temperature, and (10) washed in the ELISA washer with the washing buffer as indicated on the kit catalog, (11) A certain amount of biotinylated peptide should be added to all wells, except the blank, (12) The plate should be covered with parafilm and left to incubate at room temperature for a period of time stated on the kit catalog, (13) The parafilm on the plate should be removed and the well contents should be emptied, (14) They should be washed in an ELISA washer using the washing buffer, which should then be discarded, (15) SA-HRP solution should be added to all wells, (16) The plate should be covered with parafilm again and left to incubate in the Orbital shaker (350 rpm) at room temperature for a certain period of time, (17) After the parafilm over the plate is removed, all wells should be washed in an ELISA washer using an appropriate amount of washing buffer, (18) TMB solution should be added to all wells, which will then be incubated; at this stage, the plate should be covered to protect it from light, (19) The plate will be covered with parafilm again and incubated, (20) The parafilm should be removed and acid (2N HCl) should be added to each well to stop the reaction; after this procedure (addition of the stop solution), the blue color will start turning to yellow [19]. All the procedures listed above are shown in Fig. 3.

Immediately after adding the stop solution, results must be read within 10 min at 450 nm. Before reading, the concentrations of the standards must be entered to the ELISA reader. Thus, the device can draw a standard curve graph and automatically calculate the concentrations of the samples. This calculation can be printed out. In our laboratory, an ELX800 ELISA device is used for reading Readers, printers and washers used in ELISA methods are presented in Fig. 4.

An examination of the sources of errors in peptide and protein analyses demonstrates that the errors occur more in the pre-analytical period, than in the analytical and post-analytical periods. There are a number of changeable and unchangeable factors affecting peptide and protein values. Peptide/protein analyses with ELISA have three phases, namely, pre-analytical, analytical, and post-analytical phases [24], [28], [33], [37].

The most common mistake in peptide analyses is the failure to add protease inhibitors to collection containers or biochemistry tubes when collecting biological samples to protect peptides against proteases. Peptide and protein concepts do not have definite boundaries. Some researchers categorize structures with less than 50 amino acids as peptides, while some others classify substances with fewer than 100 amino acids as peptides. The amounts of peptides and proteins in biological fluids vary depending on diseases [18]. Diseases also affect the amount of proteases in biological fluids. These proteases whose quantities vary depending on diseases easily break down substances in peptide structure [2]. Therefore, it is necessary to show whether the change in the amount of the analyzed peptide was due to disease or due to its destruction by proteases whose amount increased because of the disease. It was reported that there are more than 700 proteases in circulation in humans [30]. Thus, the most frequent pre-analytical error in the analysis of hormones with peptide structure is the failure to use protease inhibitors while collecting biological samples (like saliva, blood, urine, etc.) or to use test tubes that do not contain protease inhibitors [2].

There are a number of protease inhibitors in the market like phenylmethylsulfonyl fluoride, aprotinin, and protease cocktail inhibitors. Currently, commercial firms have started supplying biochemistry tubes containing aprotinin and EDTA. It was reported that 500 kallikrein unit of aprotinin was protective for 1 mL of biological sample and care should be taken to use the indicated amount when collecting biological samples [17]. Additionally, our laboratory experience shows that it is not appropriate to use different proteases in the same experiment and that aprotinin is more protective than phenylmethylsulfonyl fluoride and protease cocktails.

Although the use of protease inhibitors in peptide analyses has recently become common, protease inhibitors are not used in some research laboratories due to lack of experience or expertize, and thus, the amounts of the same peptide in the same disease vary among laboratories. In order to overcome this problem, the importance of using protease inhibitors in peptide measurements should be emphasized in the congresses of the world peptide association. It is even more desirable, if possible, to reach an agreement on the use of a single type of protease inhibitor to protect peptides against proteases and to ensure standardization. Otherwise, different researchers will continue publishing different results for the same peptide in the same disease. The need to use protease inhibitors for the optimal stabilization of peptides was previously expressed by Blatnik et al. [7], [8].

Besides, it should be remembered that some peptides like salusins (alpha and beta) have specific biological sample collection requirements, in addition to protease inhibitors. For instance, as salusin-β adheres to the wall of polypropylene and polystyrene test tubes, its measurement poses certain problems. To avoid these problems, it is recommended that low doses of NP-40 or Tween-20 be added to the test tubes [5], [34].

Another peptide that has special sample collection requirements is ghrelin [2], [7], [17]. As it is known, ghrelin is found in four different forms in biological tissues and samples. One of these forms is the ghrelin with 8-carbon fatty acids attached to the third serine amino acid at the N-terminal (to the third amino acid threonine at the n-terminal in frogs); the other is the 10-carbon fatty acid ghrelin attached to the third amino acid serine at the n-terminal; another is the 10-carbon ghrelin with a single unsaturated fatty acid attached to the serine amino acid; and the last is the one from which these fatty acids are detached and which has only 28 amino acids. This last form is called desacyl ghrelin [32]. Ghrelins with fatty acids, on the other hand, are called acylated ghrelins. In order to enhance the stability of acylated ghrelins, the sample should be added HCl at a volume ratio of 1/10 per milliliter from a 1 mL normal HCl [2], [17]. Besides, the researchers must indicate which form of the ghrelin they studied in their reports. Variety of form is seen not only in ghrelins [17], but in a number of peptides, including apelines, among others [4]. Therefore, it is important for peptide researchers to state which form they examined for the purposes of standardization.

Another important issue encountered in peptide analyses is that some peptides may remain below the detection limit of the kit or stay low because of certain diseases. In this case, standard addition–subtraction methods should be used [36]. That is, a known quantity of peptide must be added to all biological samples. In order to identify the actual concentration of peptides, the added concentration should be subtracted from the concentration found at the end of the experiment. Thus, it becomes possible to measure peptides found in very low concentrations in biological samples.

As peptides are found in a number of biological fluids, ELISA method has been used recently to identify peptide amounts in the supernatants of biological tissues, as well as the biological fluids. If the peptides are to be quantified in the supernatants of biological tissues, the wet weight of the tissue should be determined immediately after it is removed from the organism (200–300 mg of tissue will be adequate) and the tissue should be boiled in boiling water for 5 min [21]. Thus, proteases are inactivated and peptides are protected against proteases. Besides, when the tissues are to be homogenized, 20–30 μl of aprotinin out of 500 KIU aprotinin should be added [17]. If possible, the supernatant obtained after the centrifuge must be studied simultaneously with other biological samples. If the supernatant will not be studied shortly, then it must be stored at −20 °C or −80 °C until the time of analysis. Our laboratory experience shows that samples stored as such will stay intact for 3 years at −20 °C and for 7 years at −80 °C.

The patient's position while the blood is collected is important and can be one of the pre-analytical errors [1], [29]. For instance, the blood volume of a healthy adult in standing position is 600–700 mL (10%) lower than the volume in lying position. In an erectly sitting position, the liquid without protein is transferred to the tissues through capillaries and this causes a significant difference in the plasma volume. As a result, peptides and proteins, hormones with protein structure, drugs carried by binding to proteins, calcium, enzymes and bilirubin concentrations will be elevated. Exercise also causes pre-analytical errors. Exercise-associated increases have been reported in aspartate aminotransferase (AST), lactate dehydrogenase (LDH), creatinine kinase (CK), urea, creatinine, transferring [24], [28], [33], [37] and irisin levels [9].

The patient's condition [fasting (fasting for about 10–12 h is appropriate for taking samples), satiety, drug use, pre-examination, etc.], incorrect sample collection (from the arm of the infusion), collecting the blood sample into the wrong tube, collecting more or less than the appropriate amount, effects of food (drinks containing caffeine like tea, coffee, coke, fat, protein and carbohydrate rates), alcohol consumption (changes many analytes depending on short-term or long-term effects), effects of fever, age and sex (cause changes in reference values in biochemical and hematological tests), pregnancy, diurnal rhythm (the release and metabolism of some analytes change over the day), seasonal changes (for instance, higher vitamin D levels in summer and higher triglyceride and total cholesterol levels during summer, in comparison to winter; therefore, this should be considered, when a correlation between vitamin D and peptides is explored), altitude (when blood is collected from individuals living in higher altitudes, hemoglobin, hematocrite and CRP will be higher; this should be considered when a correlation is examined between these molecules and peptides), different body mass indices (relationships have been reported between many peptides and BMI), use of tobacco and tobacco products, inappropriate transfer (samples which kept waiting for too long), freezing and thawing, and the measurement interval of the kits being sensitive enough are among the most common causes of pre-analytical errors [1], [24], [28], [29], [37]. All these conditions should be taken into consideration in sound scientific research.

Use of kits from different firms in the same study is among the most common analytical errors. There are several ELISA manufacturers. The ELISAs of these different companies can measure the same sample in different amounts. For instance, ELISA kit of the Phoenix firm measures ghrelin amounts 10 times lower than that of the Linco firm, although linearity was reported [14]. Accordingly, for the purposes of standardization, it is important for a laboratory to use the kits of the same firm. It is all the better if the kits of the same firm and with the same lot number can be used in the same study.

Research laboratories depend on grants to continue operating. The institution which offers the grant requires that the products of the firm which gives the best offer be bought or asks for an explanation as to why certain products are not preferred (this is the common procedure). If the ELISA kit you have already been using in your laboratory gives the best offer in the bid, then there is no problem. Otherwise, you have to conduct the assay validity experiments for the new ELISA kit. The assay validity test procedures for an ELISA kit were used in our laboratory and the relevant details can be found in the article cited here. Although firms manufacturing ELISA kits provide the assay validity parameters on their kit catalogs, it is sometimes difficult to obtain these data. Therefore, laboratories must test the assay validity parameters when they use different ELISA kits. Furthermore, the kits are usually designed to measure peptides in blood. Consequently, if the kit is going to be used for another purpose, for instance to detect peptides in tissue supernatants, ELISA assay validity parameters must be certainly tested to ensure the reliability of results, as briefly explained as follows [3].

The intra-assay (within-day) and inter-assay (between-days) variation should be determined for biological samples. The coefficient of variation (CV) can be calculated as: CV = Standard Deviation (SD)/Mean. The CV values used in testing are generally less than 15% for clinical use.

Biological samples should be diluted with distilled water and assayed. After biological sample dilution, concentration of biological sample constituents must indicate perfect linearity on serial dilution.

Biological samples should be enriched with increasing amounts of standard samples. Then, the percentage recovery can be calculated as follows: observed value-baseline value/amount added 100×. The obtained results would verify whether or not the used kit would detect peptides/proteins quantitatively in other biological fluid, beside serum [3].

A laboratory director cannot do all the work on their own. However, they must not reveal the title of the research in question to the research assistant and lab technician to avoid biases; that is, the staff should be blinded to the study [17]. In addition, if a new staff member is hired, the same sample should be studied by several people. Thus, it can be checked whether the same sample is measured in the same way.

Among the most common analytical errors are the degradation of reagents, equipment errors, and pipettes that do not measure correctly, and not enough biological samples. Serum/plasma is a yellow liquid that is around 55% total volume of blood. 100 μl of serum/plasma is necessary for ELISA. So that some one can estimate how much blood they need for ELISA test. Another error is the failure to follow the washing procedures and to keep to the time reactions must be ended. Errors are also encountered when the antigen and the antibody used to coat the solid phase is not at the right concentration or when planting the standards, using only one plate to analyze a number of samples at the same time (7–8 standards prepared for each plate should be added at all times, because this is the most important step showing whether the experiment was correctly conducted. Theoretically expected concentrations of the standards should more or less coincide with the concentrations found in the experiment). Failure to arrange for the optimum incubation duration or temperature may cause errors. Other errors may include the following: failure to choose the most appropriate substrate, using reagents with different lot numbers during the study, accidental manual contact with the bottom of microplates, not maintaining the room temperature (below or over 20–23 °C), not determining the appropriate dilution factors, presence of unacceptably high levels of interfering proteins in the samples, the tubes having air bubbles in them and neglecting to change the tubes at every step, differences in pipette techniques (to ensure standardization, the same analysis must be carried by one person; more than one person should not be involved in the analysis), making an error in incubation time (longer or shorter than the appropriate duration) and differences in washing techniques.

Although the ELISA reader provides the absorbances and concentrations of the samples to the researcher, the concentrations of some samples may occasionally be higher or lower than expected. In this case, the researcher may be tempted to calculate the concentration of the studied peptide from the absorbance using the curve expect program. Our experiences show that this must never be done, as we have seen that, when samples with high concentrations are diluted, studied again with the same kit, and compared with the calculated concentration using the curve expect program, the results show considerable differences. Among the other post-analytical errors are wrong or incomplete statistical evaluation and incorrect interpretation.