Western Blot FAQs

What is typically done to detect IgG in a sample is to detect it by using only the secondary Ab which normally is an anti IgG antibody. For chemiluminescent detection you use a HRP conjugated or you can use a Cy-dye labeled for fluorescent detection. If you run at reduced condition and with denaturated sample the IgG or IgM molecule split into light and heavy chain. To quantify the levels I would rather run native conditions, blot for the target protein (IgG or IgM) using a labeled sec. Ab (anti IgG or IgM), re-blot for a suitable housekeeping protein (or do a multiplex detection if using fluorescent detection). In serum transferrin might be a candidate as housekeeping protein. I don´t know if transferrin levels varies or are affected by something, at least I know it is possible to detect in serum. If there is no good housekeeping protein normalization to total protein would be an option.

If you have same species the secondary Ab will then bind to both proteins and you can´t distinguish between the signals/proteins. However, if the protein bands are very distinct (no unspecific bands) and very well size separated (a large and a small protein) it might be possible to use same species but this is generally not recommended. Typical procedure is to use two different species and rabbit and mouse are two most common species.

In case there is a suitable secondary Ab available I would recommend labeling that one. It is also possible to label the primary Ab in case the protein is highly expressed. The signal gets weaker when you only have a primary Ab because you do not get the amplification effect from the secondary Ab (several secondary Ab binds to each primary Ab). There are CyDye-labeling kit available (XXX) which works very well for antibody labeling.

When do labeling it is important to have a correct balance between the protein and CyDye amounts in order to get effective labeling with minimal effect on the protein (IgG). The labeling kits and protocols are designed to have minimal effect on the antibody performance and as long you follow the recommendations regarding protein and CyDye amount it will not have any negative effect on the antigen-antibody binding.

Fluorophores are typically stable (up to 3 month) on the membrane which allows you to store them. I would not recommend freezing them. I would rather dry the membranes and store them at room temp. Protected from light (between sheets of filter paper or tissue wrapped in aluminum foil).

Typically fluorescent detection requires higher concentration of both primary and secondary antibodies at least compare to what is used for high sensitive chemiluminescent reagents. Optimal dilution varies between antibodies and is dependent of how much there is of your target protein in the sample.

Our fluorophores are very stable and minimal affected by bleaching. The signal is stable up to three month on a membrane and is not affected by normal handling in daylight or by repeated imaging. Protect membranes and antibodies from light when stored.

If the background is caused by sticky antibodies and unspecific binding you will have the same problem in fluorescent detection. The principle for protein-Ab binding and detection is the same only the light signal differs. I would rather optimize the blocking, try different blocking agents. You can also increase numbers of washing steps.

I recommend to always using fresh antibody solutions. However, I know it is common to reuse especially the primary Ab solution because primary antibodies are expensive.

At least, I recommend to not reusing the secondary Ab. For the primary Ab I would say it is okay to reuse as long it looks okay and without any bacterial growth. Avoid blocking in the primary Ab solution as this cause growth quickly (especially if using milk). You can add acid into the solution but be aware of that acid inhibit HRP.

Unspecific strong bands at lower or higher Mw may be isoforms and/or dimers of the protein and could therefore be hard to get rid of. Check in the literature if there are known isoforms and/or if it known to appear in as dimer. It might be worth to look for and try another antibody.

If there is IgG in the sample (e.g. from an immune precipitation) you get binding of the secondary Ab (anti IgG) to the heavy and light Ab chain resulting in two bands. In case the non-specific bands do not interfere with your target protein and you have an idea of what it is like IgG, dimer, isoform etc.it might be easiest to accept it rather than to spend a lot of time to get rid of them.

However, one suggestion to reduce non-specific bands is to alter high (up to 0.5M NaCl) and low salt concentration in the washing buffer. But be careful with the high salt concentration washing so you don´t lose the protein-antibody. Just wash for some minutes with the high salt.

I have not experienced that myself but I have heard from others that they have seen it. I have no good explanation why or how to avoid it. I would try another marker. I have used our Rainbow marker (RPN 800) for quite long time and in different blots and never had problems with non-specific binding to it.

Yes there are some primary antibodies labeled with fluorophores available but probably not so many different. Another possibility is to label a primary antibody yourself and there are labeling kits available. It is a pretty simple procedure, takes about 30 minutes. This was what I did for the triplex application (slide 26), I labeled then the primary Ab for the housekeeping protein and it works fine. If you use a labeled primary Ab you do not need any secondary antibody. However, when using a labeled primary Ab the signal get weaker compare to if you have a labeled secondary Ab because the signal is amplified (several secondary Ab binds to each primary Ab). But if you have a very highly expressed protein such as a housekeeping protein the signal usually is strong enough to detect.

You can incubate with both antibodies at same time, which saves time, or you can do it one by one.

Yes I agree with you. Housekeeping proteins may vary between samples. However, housekeeping proteins such as actin and tubulin are accepted in general.

In case the target protein and housekeeping protein have very different molecular weight and are well separated and gives very clear bands it is possible. If so I suggest cutting the membrane between the two proteins expected sizes (use a pre-stained marker) after transfer and then probe for each protein.

Most common and what is typically done in this kind of studies are to normalize to its total protein (the non-phosphorylated). Another possibility is to also detect a housekeeping protein and first normalize the total protein (non-phosphorylated) against the house-keeping and then use the normalized total protein for normalization of the phosphoprotein.

Yes it is very good to measure the protein concentration in the samples before apply to Western blot. Typically you want to load equal amount of your samples and/or you need to know how much to load on the gel. BCA is a suitable reagent for protein concentration measurement.

Yes, if you do a standard curve for your protein then you can do a more direct quantitation. However, you have to do the standard curve on the same membrane as the samples because the signal intensity values will vary between blots depending on the chemiluminescence characteristics (see slide 7).

I would suggest Amersham ECL Select (GE Healthcare, RPN 2235), chemiluminescent detection. This is a very sensitive reagent with high light output optimal for low abundant proteins. I also recommend using a PVDF membrane (such as Hybond P) as those have higher protein binding capacity compare to nitrocellulose membranes.

Another option is fluorescent detection using Amersham ECL Plex system. Use the ECL Plex Cy5 labeled Ab as this gives the strongest signal intensity. For fluorescent detection it is important to use a low fluorescent PVDF membrane such as Hybond LFP (RPN 303 LFP).

I would recommend mini size format as it is easy to handle. Use 7 cm IPG strip gel for 1D separation and 2 wells mini gel for the 2D separation. Do a Cy 5 (or Cy3) labeling of total protein and run the separation. Transfer to membrane (labeled total proteins transfers) and blot for you protein of interest, use Cy3 labeled secondary Ab. Image the membrane for Cy5 total protein and Cy3 for specific protein. 2D Western has offers great opportunities such as analysis of isoforms (pI isoform resolution) phosphorylations, mapping of phosphosites, phosphorylated isoforms etc.

We have a very nice application note describing a 2D application with phosphoprotein detection

You can find it at our Web www.gelifesciences.com search on this number: 28-9042-34 AA

Yes you are correct, the stripping and re-probing is much uncontrolled and you never know if you lose some of the proteins. However, if you lose some protein you may lose similar in all lanes and as the quantitation is a relative quantitation it may not have too much impact. In case the target protein and housekeeping protein is very well size separated it is possible to detect the two proteins without stripping just do the re-probing or you can even cut the membrane between the two bands and blot them one by one. But I agree with you, stripping and re-probing giving less accurate quantitation and if you are using chemiluminescent detection you have to be very careful and do it in a proper way. Therefore I can really recommend fluorescent detection and multiplexing (simultaneous detection of target protein and housekeeping protein) when doing quantitative WB. No stripping and re-probing is required, the signal is stable and thereby you get more reliable quantitation.

There are different methods and protocols for stripping. Time varies between methods, usually shorter time for harder conditions and shorter time if milder conditions are used. In our Western blotting handbook there are different methods described that might give you some guidance. The handbook can be downloaded via this link

www.gelifesciences.com/researchsolutions/wb

Yes you are right, if the two proteins are well separated and have different size you may not need to strip. Another possibility is to cut the membrane between the two proteins (use a pre-stained marker to see where to cut) and blot them one by one, put the membranes together for detection.

I have no experience myself but have heard about it. Addition of H2O2 “kill” the enzyme and thereby its activity. This may be an option instead of stripping but I feel a bit concern about if the proteins are affected, but I do not know. It might work; I would rather go for stripping.

Glycosylated proteins are difficult to separate nicely. The glycosylation cause size heterogeneity which often results in vertically broad bands and vertical streaks which is hard to avoid.

I would suggest using a broad range gradient gel, such as 4-20% if you have very small and very large proteins. Other gradient may be 8-16 %, 4-12%.

There are useful tables available showing how proteins separate in different densities (search on the web, I think BioRad and Invitrogen has guides).

There is pre-packed HiTrap columns for depletion of albumin and IgG in serum and plasma available (GE Healthcare, prod no. 28-9466-03)

Use a low density gel, preferably a gradient gel 4-8% (such as Novex NUPAGE 4-8% gel from Life technologies). Run the gel for longer time than you normally do for smaller proteins and use a pre-stained molecular weight marker or monitor the separation. Add 0, 1% SDS in transfer buffer which improve migration of large proteins from the gel to the membrane. You can also decrease the methanol concentration to 10%. Increase transfer time compare to what you run for smaller proteins (suggestion: 100V for 2 hours at +4 degrees Celsius).)

Towbin and Bjerrum Schaffer Nelson buffers differs in tris and glysin content and have different pH. Bjerrum Scaffer Nelson buffer is developed for semi-dry transfer but I have no experience if they differ in performance of small proteins. Small proteins are usually easier to transfer under semidry conditions than large proteins. General recommendations for transfer of small proteins are 20% methanol and no addition of SDS in transfer buffer.

I would compare them side by side to see if there is any for your specific protein.

It is recommended to use 20% MeOH and no SDS in the transfer buffer for transfer of small proteins. I would recommend equilibrating the gel in transfer buffer for at least 10-15 minutes. You may also run the transfer for shorter time compare to what is used for larger proteins.

Blow through means when proteins pass through the membrane during transfer. This can happen, especially for small proteins, and if you run the transfer too long time or at too high voltage or current.

For small proteins we recommend to use a membrane with 0.2 µm pore size.

It is important that a certain protein transfers equal across the whole membrane (in all lanes). However if there are a difference in transfer efficiency between two, your target protein and the housekeeping protein (within the lane) it does not affect the quantitation as it is a relative quantitation you do as long the transfer is equal between the lanes.

It is more important that proteins of same molecular weight (same protein) transfers equally across the membrane than proteins of different molecular (transfer in same lane).

To check transfer uniformity across the membrane you can stain the membrane after transfer or you can label you sample with a fluorescent dye such as Cy5 or Cy3 before loading on gel. The gel can then be imaged before transfer, after transfer and the membrane can be imaged to visualize the transfer efficiency.

I usually mark the membrane by cutting a specific pattern for a certain membrane. I think that is very convenient. However, to mark carefully with pencil woks as well, but do not use a ballpoint pen. This is an example how you can mark with cuts.

You should not use PBS-T for blotting of phosphoproteins, use TBS-T instead.

I am sorry if the webinar didn’t meet your expectations. The webinar was not aimed to be a sales pitch for our fluorescent reagents. The aim was to show the benefits of fluorescent detection in quantitative WB and how you can overcome some of the challenges. The last part of the presentation covered some hints&tips which I think can be useful to understand how to get good WB results. If you have any specific issues please feel free to contact me.

Susanne.grimsby@ge.com

There are many reasons for background issues. One of the most likely reasons is non-optimal antibody dilutions, typically too high concentration of the secondary antibody but also too high primary Ab concentration may cause background. Another reason can be non-sufficient blocking and/or washing. X-ray film is very sensitive and gives therefore also more sensitive to background. CCD imagers are also very sensitive but usually have an inbuilt function for background reduction.

Thanks for this advice.

Also shorter and more frequent washing is effective.

No one blocking is optimal for all blots so it is good to try some different. However, our ECL Prime blocking (GE Healthcare, RPN 418) is a very good blocking as it a mixture of different blocking compounds. It works fine for both chemiluminescence and fluorescence.

One reason to high background is non-optimal antibody concentrations. I would recommend you to dilute your antibodies further especially the secondary. A good start may be to increase the dilution at least 2-3 times. You can also add some additional washing steps. In case the background is unspecific bands I would recommend trying another primary antibody that is more specific. Unspecific bands can also appear if there are isoforms of the target protein that the antibody binds to. Another way to reduce unspecific bands is to alter high (up to 0.5M NaCl) and low salt concentration in the washing buffer. You have to be careful with the high salt concentration and not wash too long time (just some minutes) to not lose the proteins/antibodies.

Standard recipes recommend 0.05-0.1% Tween20. Higher concentrations may be too stringent and you risk losing proteins. Some stripping buffers contain higher Tween20 concentration.

Remember that Tween20 gives high background in fluorescent detection and therefor needs to be washed away before scanning.

This is only for fluorescence. Tween tends to give unwanted background fluorescence. The reason is to avoid background.

Limit of detection varies between different proteins and is dependent of antibody affinity and quality as well as imaging equipment. There are also a lot of different reagents on the market with different light output and thereby provides different detection sensitivity. Amersham ECL Select (GE Healthcare RPN 2235) is one of the most sensitive detection reagents on the market today. The limit of detection is around 10 times higher or even more comparing to a standard ECL (such as Amersham ECL). To give you an example, in a WB for ERK ½ in HeLa cell lysate you can detect ERK ½ protein in 5 ng of total protein loaded (only a small part of this 5 ng is ERK1/2) when using ECL Select while with ECL you can only detect it in 80 ng (16 times difference). In a WB for a pure protein, using transferrin, it is possible to detect down to low pico gram levels when using ECL Select.

Theoretically yes, if the proteins are very well sizes separated and have different sizes and the protein-antibody binding is specific.

I do not have exact prices but the fluorescent labeled antibodies are slightly more expensive and you usually need to have higher concentrations of them compare to for high sensitive chemiluminescence. In terms of that it is more expensive, but you do NOT need any additional ECL detection reagents (which are quite expensive). All in all I would say that the two detection methods are comparable in price.

The presentation covers both chemiluminescence and fluorescence. Some are based on chemiluminescence but all application examples are done with fluorescent detection using Amersham ECL Plex

I do not recommend Poseau-S staining. This may give background fluorescence. It is better to use a fluorescent staining that do not spectrally interfere with the CyDye

Fluorescent detection is typically characterized by very high sensitivity. However there are ECL detection reagents with very high sensitivity such as Amersham ECL Select.

To detect such low levels I would recommend trying Amersham ECL Select. I also recommend using a PVDF membrane as this type of membrane has higher protein binding capacity compare to nitrocellulose membrane. Another option is to use a fluorescent detection system such as Amersham ECL Plex. You should then use a low fluorescent PVDF membrane (Hybond LFP).

No, HRP is the enzyme used for chemiluminescence.

Chemiluminescence is one kind of signal, a light of 425 nm.

Alkaline phosphatase is used for ECF (chemiFLUORESCENCE) which gives another light signal.

Theoretically it might be possible to use these two systems to detect two proteins but I would not recommend it. I would rather choose fluorescence, much easier.

Fluorescent detection is recognized by very high sensitivity and for providing low limit of detection. However, today there are very high sensitive ECL reagents available as well such as Amersham ECL Select (GE Healthcare RPN 2235) providing a limit of detection similar to fluorescence. But as you probably know the detection sensitivity varies a lot between proteins and blots and there might cases when chemiluminescence is better or vice versa.

I agree with you, antibody optimization is time consuming (and boring), unfortunately there is no difference between chemiluminescence and fluorescence. To obtain best results both systems requires optimal antibody dilutions. You can find some guidance in our product instructions where we give recommended dilution ranges.

I do not have exact prices but the fluorescent labeled antibodies are slightly more expensive and you usually need to have higher concentrations of them compare to for high sensitive chemiluminescence. In terms of that it is more expensive, but you do NOT need any additional ECL detection reagents (which are quite expensive). All in all I would say that the two detection methods are comparable in price.

As I can see it there are two ways either to quantify each isoform separately (normalize each isoform to housekeeping) or summarize the signal intensity values for the isoforms and normalize that to the housekeeping.

Another possibility when having isoforms is to run 2D Western. You can do the 1D separation using 7 cm IPG strips and the 2D separation on 2-wells mini gels. Transfer and probing same as for standard WB. 2D WB gives very good separation, both pI and Mw.

I can recommend this application note about 2D western blotting: www.gelifesciences.com search on this number: 28-9042-34 AA

We provide software called ImageQuant TL. This software has analysis tools for WB and 1D gel electrophoresis analysis as well as colony counting and array analysis. From the analysis you obtain signal values for the protein bands which you then export to excel for normalization and quantitation calculations.

It is not absolute quantitative (as long as you don´t do a standard curve for your protein on each blot) but very widely used for relative quantitation. However, it is important to consider all aspects and do it in a proper way. Fluorescent detection is preferred as detection method. The signals are stable, more reproducible and you can do multiplexing which improves quantitative analysis.

There is analysis software available such as IQTL from GE Healthcare with Mw calibration function. Using this type of software you can calculate the protein size using the Mw as a reference. I recommend loading a molecular weight marker in each outer lane. I also recommend using a pre-stained Mw marker with which you can see the migration on the gel as well as the marker bands on the membrane, which also confirms the transfer. With our new LAS 500 CCD camera you can get a picture of both the marker (colored bands) and the target protein bands in the same image. The camera takes an image of the marker and of the chemiluminescent signals and creates an overlay image. Very useful.

Yes, if you do it in a proper way and consider following. Fluorescent detection is preferable as the signals are stable and varies less between blots. Run the blots side-by-side and treat them exactly the same.

If you are using chemiluminescence it is important to treat all membranes in the same way and do it side-by-side. Do not handle and detect more than 2-4 membranes at the same time. Incubate with ECL reagent for same time and expose both membranes at the same time together and with same exposure time. If you have samples such as cell lysates I recommend blotting for a housekeeping protein as well and normalize the target protein values against that. If you consider these things it would be possible to have samples on different blots. However you should always be aware of that there might be some variation especially if using chemiluminescence.

Best is to use a CCD camera for imaging and an analysis software where you get a signal intensity value. CCD camera imaging systems has a known max intensity value where the signal is not saturated. If the signal intensity is above the max value you should expose for shorter time. A CCD camera also typically has a function where a saturated signal is shown in red. With x-ray film it is more difficult to know as you don’t get a signal value. But if you get a very strong and big/broad band it is probably saturated.

Chemiluminescence as well as fluorescence is used in ELISA. The principle is similar based on protein-antibody binding. The difference is that in WB the proteins are size separated which gives further confirmation of identity. In ELISA you measure all signal even signals coming from unspecific binding.

My name is Susanne Grimsby

if you have any further questions please contact me at

Susanne.grimsby@ge.com

www.gelifesciences.com/researchsolutions/wb