Western Blot Guide
Protocols and technical information
Recommendations and Troubleshooting
Each antibody-antigen interaction has unique characteristics. A protocol giving good results for one antibody-antigen pair might be found to be unsatisfactory for a second antibody, even on the same sample. The interaction of antisera with protein epitopes in western blotting depends upon a number of factors, all contributing to the final signal/noise ratio. Knowledge and control of those factors allow for modifications and optimization of the procedure. Polyclonal antibodies (serum or IgY-fractions from egg yolk) usually contains a number of different antibodies interacting independently with different epitopes on the target used for immunization. This number is limited for peptide-antigens but if recombinant or native proteins have been used for immunization the signal obtained has to be regarded as cumulative (caused by several different types of antibodies present in the serum/IgY used).
To achieve reproducible results we recommend the use of gloves, forceps, clean and detergent-free laboratory materials, reagents which have not expired, and freshly prepared buffers with controlled pH. Additionally, it is important to keep protocol volumes, dilutions, and incubation times the same. In general, incubation times should be as short as possible while antibody dilutions should be as high as possible.
To ensure that the target protein is in the starting material, it is essential to prepare and store the samples under non-degrading/aggregating conditions. If the target protein is degraded or not present in sufficient amounts in the loaded sample, it can result in lack of signal detection. Epitope abundance can be enhanced under altered growth conditions, by selective tissue preparation, or fractionations of complex cellular extracts (i.e. organelle preparation). Use the extraction method providing good yields and consistent results. Agrisera has recently validated Precellys® device from Bertin Techonologies, as it fullfills such criteria. For work with organisms possessing robust cell walls like diatoms, recommendations can be found here. Useful reference: "What is the total number of protein molecules per cell volume? A call to rethink some published values. Milo, 2013.
Sample quality is of crucial importance. As shown in the example below, samples of Arabidopsis thaliana total leaf preparations from year 2008 or 2006 show greater degradation of PsbB compared to the newer 2009 samples. M-molecular weight markers.
Secondary antibody used was Agrisera goat anti-rabbit HRP (1: 50 000) and ECL Advance detection system (GE Healthcare).
Protein samples older than 1 year should not be used if possible.
Run sample extraction buffer (+loading buffer) in one lane of your gel to check for contribution of your sample extraction buffer to the background signal.
Alternative preparation methods should be considered if possible.
Include positive and negative controls from the beginning: comparison of samples prepared from material containing no target (e.g. genetic 0-mutants) or higher amounts of the target (e.g. over-expression mutants) as well as cellular fractionation (e.g. preparation of various fractions of organelles) will allow to control specificity of obtained signals.
Protein extraction protocols: seeds
(2) Protein Separation
The size-related identification of a protein-antibody interaction usually requires gel separation under fully denaturing conditions. Complete reduction of intra- and intermolecular S-S bridges ensures the accessibility of the epitope for interaction with antibodies detecting linear epitopes. In this case, reducing agents have to be added to sufficient final concentrations to both sample and (usually) running buffer. If antibodies recognize non-linear eptiopes they require conformational integrity of the target, best provided in non-denaturating PAGE systems or immuno-histochemistry applications. Increasing protein loadings might elevate epitope abundance but most often also promote non-specific cross-reactions, higher background, and impaired separation.
Avoid protein loads higher that 5-20 µg/lane for standard mini-gel systems.
Check separation with stained marker pattern or by Coomassie/Silver-staining of the gel. Mixing a pre-stained marker with a marker reacting with the secondary antibody (e.g. MagicMark, Invitrogen) will be an advantage for size-determination and can serve as a control for the visualization assay.
Some highly hydrophobic membrane proteins might require 2-8 M urea in the gels and sample buffer to be kept unfolded during separation.
If DTT is used as a reducing agent it should be freshly prepared.
If your polyacrylamide gel is not polymerizing wash glass plates carefully and make sure all grease is removed.
(3) Protein Transfer
The protein transfer to a suitable membrane (blotting) with protein-binding capacity (nitrocellulose, PVDF) is highly dependent on the biochemical properties of the target protein. Proteins of higher molecular weight (apparent molecular masses of >100 kDa) require longer blotting times than smaller proteins. At low field strengths (<15 V/cm) the mobility of proteins out of the gel will be decreased and proteins will not be completely transferred from the gel. If field strength is too high (>35 V/cm) proteins might pass through the membrane without binding. The transfer time should be as short as possible to prevent proteins (especially lower molecular weight) from passing through the membrane. Place the second membrane and check how much signal is there after incubation with primary antibody compare to the first membrane.
To minimize background staining due to non-specific membrane-binding of the antibody tested, remaining surfaces of the membrane have to be saturated (“blocked”) with proteins which are non-reactive with the antibody. Commonly used are: low-fat dry milk powder, casein, IgGs (at 5-10% but not from a species you have your antibodies derived from) or BSA, at 2-5% w/v. The shortest possible blocking-time should be determined experimentally for your system. Increasing the blocking-times unnecessarily (e.g. overnight) or incubation at low temperatures might lead to aggregation of blocking protein on the membrane and by this compromise later antibody-epitope interactions. If low background is obtained after 30-60 min of blocking, longer blocking will not improve your results. It should be determined experimentally if adding or omitting the blocking reagent in the subsequent antibody incubation steps is an improvement on the results. Normal buffer and detergent concentrations should not reduce affect of the blocking step, however, accessibility of a target protein for primary antibody can be increased by a short “block-free” wash. Each antibody-antigen pair is unique and therefore the blocking protocol should be empirically tested.
As some IgY antibodies might recognize milk proteins, BSA can give lower background than milk powder for these antibodies.
Do not use BSA as blocking agent if BSA-coupled peptides were used for immunization. BSA should be of high purity and IgG-free so as not to interfere with the assay.
As milk contains biotin, the use of milk-power for blocking is incompatible with avidin/streptavidin systems.
If serum is used for blocking, care should be taken to ensure the animal has not been exposed to (or developed antibodies to) the antigen in question. If this is the case, they may bind to the antigen and prevent binding of the primary antibody. Purified immunoglobulins are higher quality blocking reagents.
As a protein-free protein alternative you may block for 1 hr at RT with 0,2-0,5 % Tween-20 in PBS followed by incubation with the primary antibodies diluted in 0,25 % Tween-20 in PBS for 1 hr at RT.
(5) Primary Antibody
The primary antibody is the major determinant of the specificity of the target-recognition. The interaction with the primary epitope should dominate any cross-reactivity with other epitopes. Commonly used dilutions are between 1:500 to 1:20 000, depending upon the reactivity of the antibody and the detection system used. To check for specificity of the target recognition you can use (1) another antibody against your target which will bind to other epitopes on a target protein, (2) run control samples free or depleted of target-proteins, or (3) perform a peptide competition assay (for anti-peptide antibody).
If available, a second antibody known to react with the sample can be used as a control of subsequent assay steps on a parallel filter in the same experiment.
Lyophilized primary antibodies incubated for 2-4 h at 4º C after reconstitution prior to use may increase signal strength.
To diminish unspecific cross-reactions you may pre-adsorb the primary antibody (overnight, 4º C) with tissue extract which lacks your protein of interest (you can use a transferred membrane where the band/area containing your target protein has been cut out). In case of increased background signals, incubate your membrane before any protein is transferred in saturated PBS-milk followed by protein transfer.
Antibodies against carrier proteins (KLH, BSA or others) might recognize epitopes of other proteins present in the extract. Those anti-carrier antibodies can be removed by (1) incubating serum/IgY sample with a carrier protein in solution (0.1 % (w/v) or spotted on PVDF membrane or (2) passing serum/IgY sample through the column with bound carrier protein (use flow through fraction in further experiments).
Primary antibodies stored (at 4°C or -20°C) in solution with blocking reagent often lose their reactivity. Antibody dilutions should be stored in buffer only, not buffer + blocking mixture.
Lots of small spots visible on a membrane during development can be due to fat in a given serum. To improve such blots, use milk based blocking reagents as well as increasing Tween concentration in your buffers.
(6) Secondary Antibody
The secondary antibody has to be reactive against the primary antibody (e.g. use anti-rabbit to detect primary antibodies raised in a rabbit) and usually is coupled to an enzyme or dye that allows subsequent visualization. Thus, any non-target binding of the secondary antibody will result in background (if bound to the membrane due to insufficient blocking) or false-positive recognition of non-target proteins present on the filter (“cross-reactions”). Usually secondary antibodies are used at dilutions of 1:20 000-1:500 000 depending on the sensitivity of the visualization method (e.g. enhanced chemi-luminescense, ECL or alkaline phosphatase, AP). The optimal dilution of the secondary antibody has to be determined experimentally for the detection system used. Different secondary antibodies may even result in different recognition patterns when applied to the same sample.
To check for contribution of the secondary antibody for the result you may cut a suitable are of your filter and run it as a parallel control where the primary antibody is omitted (see example for trouble shooting).
If an Ig-reactive marker has been used, the signals obtained from the marker can serve as a control of the function of the secondary antibody as well as for size determination.
Sensitivity of secondary antibodies can differ between various manufacturers
10 μg of mitochondrial fraction from Arabidopsis thaliana (1,3) and Arabidopsis thaliana leaf extract (2,4) were separated on 10% gel and blotted on nitrocellulose membrane using wet transfer (0.22% CAPS, pH 11). Filters where blocked (1.5h) in 5% milk in TBST (1X TBS, 0,1% Tween 20), incubated with 1: 1000 anti-COXII antibodies (2h in TBST) followed by incubation with 1: 10 000 secondary anti-rabbit (1h) HRP-coupled antibodies from Agrisera (left panel) and other manufacture (right panel) and visualized with standard ECL on Kodak autoradiography film for 5 s. Antibody in left panel detects target protein also in total cell extract (2) and can be used in higher dilution than applied 1: 10 000. Agrisera goat anti-rabbit HRP conjugated antibody (AS09 602) can be used at following dilutions: 1: 50 000 -1: 90 000 (ELISA), 1 : 75 000 with enhanced ECL and 1: 25 000 with regular ECL (WB), 1: 500 -1: 5000 (IHC).
After steps 4, 5 and 6, excess amounts of blocking or antibody need to be diminished by washing the membrane in buffer, usually the same buffer used in the creation of blocking or incubating solutions. Typically they will include a detergent, such as 0.05 % - 0.5 % of Tween20 (a detergent). Adding Blocking protein in a 1:10 dilution can be added to the wash buffer to help minimize background. In some protocols, the multiple wash steps omit the detergent in the last step. For reasons of reproducibility it is recommended to keep volumes and times constant. The intensity of washing steps can be elevated by (a) increased times and volumes, (b) additional changes of buffer volume, (c) higher detergent concentrations, (4) use of stronger detergents (e.g. SDS instead of Tween-20, use high-purity detergents as they might contain high amounts of peroxides and contribute to increased background). It is good practice to verify that stocks of buffers (including wash buffer) do not contain microbial growth (visible as cloudiness that settles), as this can contribute to increased background noise.
Enzymatic detection systems are most popular, which employ usage of secondary antibodies conjugated with Alkaline phosphatase (AP or ALP) or horseradish peroxidase (HRP). In our hands alkaline phosphatase detection system is at least 10x less sensitive compared to chemiluminescence based system (like ECL). Therefore using ALP is not recommended when working with low expression level proteins in the tissue. In such cases, more sensitive detection systems should be employed (like TMA-6 from Lumigen). Some laboratories use fluorophore-conjugated antibodies (like DyLight®) and in such cases there is no substrate development step in the assay. The protocol is shorter and due to advances of digital imaging multiple fluorophores can be used in the same assay.
(9) Membrane Storage
Membranes (nitrocellulose or PVDF) can be stored after western blotting has been completed for up to 6 months, at room temperature between sheets of tissue paper (like Kleenex or KimWipes).