0115 966 7955 Today's Opening Times 10:30 - 17:00 (BST)

Enzyme-Linked Immunosorbent Assay (ELISA) Practical

Published: Last Edited:

Disclaimer: This essay has been submitted by a student. This is not an example of the work written by our professional essay writers. You can view samples of our professional work here.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.


Commonly used in immunology, Enzyme-Linked Immunosorbent Assay (ELISA) technique takes advantage of the great specificity of antibodies for a particular antigen. In ELISA a protein (antigen or antibody) is fixed to the walls of a microtiter plate well using this specificity. In addition to this, an enzyme can be attached to either of them in order to create and amplify a coloured signal, allowing the identification and quantification of a target protein. Depending on what is used in first place, antibody or antigen, ELISA can be divided in three main types, direct, indirect and sandwich ELISA (Berg, Tymoczko & Stryer, 2007).

Direct and indirect ELISA coat the wells of the plate with an antigen. Then, an antibody specific for that antigen will be added. On direct ELISA this antibody will have an enzyme covalently attached to it that will synthesize the signal. It has the benefits of being faster and simpler than indirect but can not amplify the signal and, therefore, is not suitable for very small concentrations of the target protein. In comparison, indirect ELISA uses a secondary antibody attached to an enzyme with specificity to the primary antibody. It has the advantage of being more sensitive as there are more binding sites for more antibodies-enzyme conjugates to bind and amplify the signal(“Elisa secondary,” n.d.). As negative points, having two antibodies increases the risk of cross-reactivity with other substrates than the intended antigen, antibody-antibody-enzyme linking. This type of ELISA is intended to detect a specific antibody and it is commonly used to diagnose infectious diseases, e.g., HIV (Berg et al., 2007).

The other type of ELISA, sandwich ELISA, uses an antibody to coat the wells instead of an antigen. This antibody will detect the presence of small quantities of a particular antigen in the sample. Later, another antibody-enzyme conjugate will bind to the antigen (hence the name sandwich) synthesising the amplified coloured product that indicates antigen detection (Reed, Holmes, Weyers & Jones, 2007).

Other techniques used in immunology include Radioimmunoassay (RIA), in which a radioactive isotope is attached to an antigen. Similar to Sandwich ELISA is Immunoradiometric assay (IRMA), but in this case the secondary antibody is labelled with a radioactive isotope rather than with an enzyme. These techniques are very precise and show a linear relationship between the target protein concentration and radioactivity. As negative effects, radioactivity is a health hazard and it can not amplify the signal as ELISA. The Dye-binding Bradford method utilizes coomassie brilliant blue which binds to proteins but, it is not very specific, can not amplify the signal and the relationship between measurements and concentration is not linear (Reed et al., 2007).

Other techniques make use of ELISA to provide simple home test kits like the dipstick immunoassay (e.g., pregnancy test) or advanced procedures like western blotting that can isolate proteins by their molecular weight but utilizes ELISA’s great specificity and amplification to identify such proteins. This is useful to detect viral infections like hepatitis, HIV and gene cloning (Reed et al., 2007).


After the normal safety procedures, wash hands, lab coat, gloves and safety glasses, the microtiter plate 96 wells was put on to a layer of absorbent paper in a clean working surface. For simplicity most of the procedure will be explained as if it was a single well instead of 96, although not all wells were used (a map of the used wells and reagents applied to each of them can be found in appendix 2). Take into account that when using Gibson pipettes the same tip can be used when diluting in the same column with the same reagent, but it needs to be changed for different columns or if using a different reagent or antigen. If unsure it is better to change tips than to cross-contaminate wells.

All samples, albumin, antibody and unknown samples were tested by triplicate and the final absorbance will be the mean of those three wells. By doing this, small irregularities in pipetting can be minimized and, also, if there was a problem with one of the columns it can be detected by contrasting it with the other two columns. If there were only two tests per sample and one was found to be very different, it would not be possible to tell which one is the correct one. Therefore the experiment, or at least that sample, would need to be repeated.

The first step when running an ELISA is to coat the well with 100ul carbonate-bicarbonate buffer. This buffer has a pH of 9.4 which aids in the binding of the antigen or antibody to the well’s walls. This binding occur passively mostly by hydrophobic interactions and by the electrostatic attraction between the slightly negative unprotonated protein and the slightly positive plastic wall.

Albumin is a ubiquitous plasma protein with many binding sites for different molecules. If the secondary antibody binds to it means it is not as specific for the antigen as it should be and, therefore, cross-reactivity is to be expected to some extend. Now 100ul of rabbit albumin antigen is pipetted into the wells (as indicated in the appendix 3) and diluted in the same column passing 100ul to the next well down and discarding 100ul from the last one indicated. To ensure appropriate mixing is achieve aspirate 3 times using the pipette in that well before diluting 100ul to the next.

Repeat the process above with the rabbit IgG antibody in the appropriate wells (See appendix 3 for a map of wells and instructions). Repeat again (without dilutions) with the unknown samples. When the above is done, the microtiter plate is cover with parafilm to avoid contamination during manipulation and incubation for a week.

After incubation the plate is washed (overflowed) with PBS-Tween to get rid of any antigen not bind to the walls (see appendix 1). To do so the excess liquid is dumbed upside-down a couple of times with an energetic movement and blotted by tapping enthusiastically on to a piece of absorbent blotting paper at least three times. Now, 125ul of casein protein were added to cover any uncoated surfaces in the well were the secondary antibody could bind and artificially increase colour production.

Before adding the anti-rabbit IgG peroxidase conjugate the plate is washed again 3 times with PBS-Tween and blot. Now 100ul was added to the appropriate wells and incubated for a further 30 min at 25C to give time to the antibody to bind. Once this was done, the plate was washed and blotted and the substrate for the conjugated enzyme, ortho-phenilene diamine (OPD) was added. OPD can be degraded by light and must be kept in darkness, therefore it must be pipetted quickly to not to affect the results. The plate was then incubated for 10 to 30 minutes at 25C. Now it was time to stop the reaction by adding 50ul (2.5M) of sulphuric acid. This will drop the pH, denaturizing the enzyme and stopping the coloured product synthesis.

Now the plate is ready for reading the absorbance of each well in the spectrophotometer. The results were produced in an excel file that can be uploaded to “Myassay” webpage for the analysis of the results and creation of the standard curve.

A standard curve is a graph that helps to appreciate the relationship between the absorbance of a sample and its concentrations in an easy visual way. Generally, the higher the concentration of antigen, the more secondary antibody-enzyme conjugate will bind to it and, consequently, more coloured product will be created, given that enough substrate is available. If this is done with standards of known concentration, a line can be draw between the points of absorbance “Y” against concentration “X”. Now, by measuring the absorbance of an unknown concentration samples, it is possible to go to the “Y” axis, select the unknown result on the scale and draw a horizontal line that touches the curve. From this point, a line can be draw vertically until it reaches the “X” axis, revealing the concentration of the unknown sample (“Calculating,” 2011).

The expected results if the study has been correctly performed are a coefficient of variance (CV) equal or inferior to 10. CV is a measure of how similar the concentrations are for the three samples of every triplicated sample. Therefore, it measures how precise is the pipetting technique of the researcher. Another important measure is the coefficient of determination, or R2, which compares how good the data fits to standard curve or the level of variation. R2 is a value between 0 and 1, being values of 0.99 considered a good fit (Gauchez, Ravanel, Villemain & Brand, 2008).


The absorbance results from reading the microtiter plate on the spectrophotometer were analysed using the webpage “Myassay”. In the outputs, the graph at the top is the standard curve of absorbance against concentration of antigen. The first table of the outputs showed the standards’ results (albumin figures 1 and 3, and antibody figures 2 and 4) by well triplets at each concentration. The second table of the outputs show the results belonging to the unknown samples, again by well triplets at each concentration.

In first place, the standard curve of the rabbit albumin was calculated in figure 1. The results were inconclusive. Concentrations, coefficient of variation (CV%) between samples, which tests the precision, and coefficient of determination ( R2) testing the replicability , could not be calculated as the readings were outside the range of the curve (0.0174-0.2106). Because the curve did not fit the curve different scales were tried without improvement and the data had to be discarded, although the outputs are included in figures 1 and 2. Probably the standards and reagents were not storage as recommended in manufacturer instructions. It was also possible that the

As a result of all the above, another set of data was utilized for this practical and the outputs included in figures 3 and 4.

Figure 1: Rabbit albumin antigen ELISA.

Unfortunately the results of the rabbit IgG antibody were the opposite of the expected; the higher the absorbance the lower the concentration in the unknown samples, as can be seen in figure 2.

The results show that the CV% was ranging 17.8-82.9, which implies that the pipetting was not accurate (must be <10). The R2 value of 0.6965 indicates poor fit of the data to the standard curve and poor replicability (values around 0.99 are considered good).

Figure 2: Flawed Rabbit IgG antigen ELISA without column 5.

The experiment was repeated by an experimented researcher showing a typical standard curve for both, rabbit albumin and rabbit antibody IgG. The results can be seen in Figure 3 and 4 respectively.

Figure 3: Correct rabbit albumin antigen ELISA.

The standard curve for rabbit albumin showed a typical positive correlation between absorbance and albumin concentration in the standards. This may indicate cross-linking between secondary antibody and albumin. The CV% was too high in unknowns 8, 9 and 12 (51.1, 20.8 and 36.2), but unknowns 10 and 11 had good CV’s of 2.68 and 0.696. Still the pipetting accuracy was inconsistent enough to be careful when drawing conclusions from the data. R2, although not perfect, was relatively high (0.9565), suggesting an acceptable fit of the absorbance readings to the curve.

Next, the output of the rabbit antibody IgG is presented in figure 4.

Figure 4: Correct rabbit IgG antigen ELISA.

The standard curve for rabbit IgG antibody shows a positive correlation between absorbance and rabbit antibody concentration. This suggests good specificity of the secondary antibody for the primary antibody, although certain level of cross-reactivity can not be discarded given the results from the albumin standard curve. All the unknown samples CV are below 10 but for unknown 12 that scored 24.9. This suggests that most of the pipetting was accurate. An R2 of 0.9854 suggests a good fit of the measures to the curve. Now, from the standard curve in figure 4 the unknown concentration was calculated and plotted in table 1.

Table1: Rabbit antibody concentrations of the unknown samples with their absorbances.


Absorbance in nm

Concentration in ug/ml

Unknown 1



Unknown 2



Unknown 3



Unknown 4



Unknown 5*



*CV >10 lacks accuracy



The inclusion of an enzyme attached to the antibody provides ELISA with great sensitivity as a single enzyme will create many product molecules This amplification of the signal makes ELISA capable to detect very small amount of protein in a sample (“ELISA”, 2001). Moreover, the stability of the agents has greatly improved facilitating the researches work. Additionally, equipment like the spectrophotometer has been automated, using microtite plates that can test up to 96 wells at once, simplifying extensive research (Reed, et al., 2007).

As limitations of ELISA, manual pipetting is usually not precise enough and automated equipment might be expensive. Cross-reactivity between the antibody and other binding sites can also distort the results producing artificially high readings (“A practical guide,” n.d.). This is a problem, as the relationship between the amount of target protein and the measurement (whether if it is a coloured product or a fluorescent marker) is indirect and not linear in most cases, making it difficult to make precise concentration estimations. In contrast, techniques like RIA are very precise and enjoy a direct and linear relationship between concentration of target protein and measurements (Reed, et al., 2007).

As a personal experience, the first time you use a 96 well microtitre plate is quite challenging. You must be methodical to remember if you have already put a particular reagent in a particular well. What I did was to make notes in a paper copy of the plate and use it as a map, writing down what wells have this or that reagent. I was a bit slower than other people but doing that gave me confidence in my work. Regardless my caution I did miss putting the casein protein in column 5. Had the study not been flawed, the lack of casein should had the effect of increasing the cross-reactivity of the secondary antibody, as there will be empty spaces on the wells’ walls for it to bind. Another interesting thing is how difficult is to do accurate pipetting. I am used to Gibson pipettes and never thought there was a major problem with my technique but the CV results were appalling. Calibrating the pipettes by choosing a smaller volume first to then get to the desired volume did not work well enough with these tiny volumes. I realize I need more practice.

In conclusion, on my own results the standard curve did not fit the curve, commonly due to poor dilutions or bad standards storage (“A practical guide,” n.d.). In addition to this, CVs were too high and highly inconsistent at best, which means that the pipetting technique was defective. The albumin ELISA did not show cross-reactivity in figure 1 but the readings from the standards and unknowns were constantly high in figure 1 and 2. If I remember well there was a bit of colour already in the substrate solution (OPD) prior to its addition to the wells. This might indicated the OPD was not correctly stored or was exposed to light, hence the high readings.

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Request Removal

If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal:

More from UK Essays

We can help with your essay
Find out more
Build Time: 0.0025 Seconds