Technical Design of a Western Blot

If you’ve ever run a western blot, or thought about running one, you’ll know there’s a lot of choices to make when designing the experiment. What detection method? What membrane? What should you block with?

It can be so overwhelming that you might just stick with the protocol your labmate handed you — after all, it worked for them! Unfortunately, that doesn’t guarantee it’ll work for your experiment. But if you understand the technical aspects of a western blot, you'll be able to understand how to modify the protocol to one that works for you. There's a lot to cover, so let's dive in! 

Figure 1: The technical decisions to make when designing a western blot. 

Psst! Looking for a western blot protocol? Check out Addgene's western blot protocol and protocol video! 

Antibodies

The most important part of your western is your antibodies, particularly your primary antibodies. We always recommend using an antibody validated for your target in your specific assay. If validated antibodies aren't available or if your experimental conditions are different than the validation assays, we recommend performing your own validation assays. If you’re using previously published antibodies, check the manuscript and supplementary figures for validation data, or follow the references to find the original publication and ensure that it includes validation data relevant to your application — and double-check that lot number while you’re at it!

Polyclonal or monoclonal

Both polyclonal and monoclonal antibodies can work well for a western blot. Polyclonal antibodies, which bind to multiple sites on an antigen, are cheap and can help amplify a weak signal. However, they have high cross-reactivity, vary significantly from lot-to-lot, and have to be re-validated each time a new lot is purchased.

Monoclonal antibodies, on the other hand, may have lower signal, but they have higher specificity, lower cross-reactivity, and are more consistent lot-to-lot.

The highest reproducibility comes from recombinant antibodies (rAbs), which are produced from plasmids and have very little batch-to-batch variation. Confirmation of sequence can be done from relatively cheap plasmid sequencing. While rAbs are monoclonal, recombinant multiclonal antibodies — mixtures of different rAbs that target the same protein — can be used to reproduce the benefits of polyclonal antibodies. Because these mixtures are defined by the manufacturer, each rAb in a mixture can be individually validated. ‌rAbs are not as commonly available as monoclonal or polyclonal antibodies, but it’s worth a look to see if they’re available for your target!

Incubation temperature and time

Incubating at a lower temperature reduces background noise but increases the length of your assay. The most common options are 4 °C overnight or 1–2 hours at room temperature (RT). Most people choose to incubate their primary antibody overnight at 4 °C and their secondary antibody for 1–2 hrs at room temperature, which allows them to reduce background while keeping the assay run time to two days. 

Indirect or direct detection method

Indirect detection is the most common way to run a western blot. It allows you to detect relatively low levels of antigen, while providing flexibility in assay design. Conjugated secondary antibodies are relatively cheap and easy to source, making it easy to switch from one readout method to another or change primary antibodies as needed. However, more antibodies mean more optimization and likely higher background noise as well, and multiplexing is challenging and limited.

Direct westerns typically rely on primary antibodies conjugated to fluorescent proteins, instead of the more common chemiluminescent methods for detection (though the enzymes used for chemiluminescence can be conjugated to primary antibodies). Direct westerns are simpler and faster than indirect ones and can be used in highly quantitative assays. On the other hand, they are not as sensitive to lower amounts of protein, and can require more antibody or loading sample, which may not be practical. Conjugated primary antibodies can be more difficult to source, especially if you need a specific conjugate, and changing your reporter molecule requires either re-conjugating or re-purchasing your primary antibody with the new reporter. Finally, each new conjugate will have to be tested to ensure it doesn’t impact the antibody’s binding to your protein of interest.

Protein preparation

Before you can blot your proteins, you’ll need to lyse and denature them. If you are intentionally running non-denatured proteins, in what is known as a native western blot, your loading buffer will not contain any denaturing agents. However, you’ll still have to lyse your proteins. For this post, we are assuming that you are running a denaturing blot. 

There are a number of lysing methods and buffers to choose from. If you are starting from tissue, you’ll likely want to use a mechanical method before your enzyme lysis. For cell cultures, enzyme lysis is typically sufficient. 

Lysis buffer

When choosing your lysis buffer, consider the subcellular location of your protein. Proteins located in membrane-bound subcompartments, like the mitochondria or nucleus, will likely require a harsher lysis buffer than proteins easily available in the cytoplasm or whole-cell lysate. A harsher, SDS-containing lysis buffer, like RIPA, should be considered if your protein is difficult to solubilize, such as a membrane-bound protein. There are many different lysis buffers available commercially, and most common buffers have recipes available online if you prefer the DIY method. As some proteins aggregate at 95 °C, you may want to check for any available data on the best incubation temperature for your specific protein(s) of interest. 

Determining linear detection range

If you are quantifying your western blot data, you’ll need to determine the linear detection range of your protein. The most common way to do this is via a Bradford assay or BCA assay. Be sure to aliquot your sample(s) before adding in your denaturing and/or loading buffers, as most Bradford assays are not compatible with detergents like SDS, while BCA assays are not with denaturing agents like DTT or β-mercaptoethanol. 

Denaturing proteins

To denature your proteins, you'll want to use SDS and either DTT or β-mercaptoethanol (BME). DTT is a stronger reducing agent than BME and has a less… distinctive smell. BME smells like rotting eggs but has a longer shelf life when stored properly. Because BME is not as strong as DTT, you’ll have to use different concentrations if you switch from one to the other. 

Type of gel

The type of gel you’ll need depends mostly on the weight of the proteins you’re interested in. Table 1 lists the most common types of gels for SDS-PAGE running conditions and will cover most western blot use cases. If you’re running native (non-reduced) proteins or are looking to preserve specific protein modifications, we recommend doing a little more research into gel chemistry and/or specialty gel options.

Table 1: Types of gels 

Gel Type	Protein sizes	Running buffer	Running conditions	Pros	Cons
Tris-glycine	6–400 kDa	Tris-glycine	100 V, 1–2 hours	Easy and cheap to handcast	Short shelf-life; can alter proteins due to high pH
Bis-tris	6–400 kDa	MES for proteins <40 kDa	180 V, 30 minutes	Good shelf life; runs at neutral pH (reduced protein alteration)	Expensive buffers
Bis-tris	6–400 kDa	MOPS for proteins > 30 kDA	200 V, 30 minutes	Good shelf life; runs at neutral pH (reduced protein alteration)	Expensive buffers
Tris-tricine	2.5–40 kDa	Tris-tricine	30 V, 1 hour or 100 V, 1–2 hours	Good separation, quality, and stability	Limited protein size range
Tris-acetate	40–500 kDa	Tris-tricine	150 V, 1–3 hours	Good for higher weight proteins	Long running time

It’s important to note that Bis and Tris gels run proteins in different patterns, due to their differing chemistries. You therefore cannot compare Bis and Tris gels or blots to each other.

Once you know your gel type, you’ll need to pick the percentage. You can get a single-percentage gel, which is cheap and useful when you’re looking at proteins all roughly the same size, or a slightly more expensive gradient gel, which will allow clear separation of proteins of different sizes. 

Table 2: Recommended gel percentages for various protein sizes

Protein Size	Target Proteins are	Gel Percentage
50–250 kDa	Similar in size	8%
20–250 kDa	Similar in size	10%
10–100 kDa	Similar in size	12%
10–250 kDa	Different sizes	4–12% gradient
4–250 kDa	Different sizes	4–20% gradient

Handcast or premade

If you need a gradient gel, the decision is probably already made: gradients are extremely difficult to handcast, so unless you have an expert around, you’ll likely want to purchase one.

If you need a single-percentage gel, you will have the option to cast your own gel. Handcast gels are significantly cheaper and generate less waste. However, they also add time to an already lengthy protocol, require the use of hazardous chemicals, and are not always consistent. Precast gels, while expensive, reduce protocol time, can be stored at 4 °C, and are very consistent gel-to-gel. 

Ladder

Whatever gel type you select, don’t forget to save a lane for your standards (ladder). You’ll want to select a pre-stained or unstained set of standards that covers a size range appropriate for the target(s) and transfer method.

Membranes

The two membrane options are PVDF and nitrocellulose. PVDF is a sturdy membrane that can be stripped and reprobed multiple times. It doesn’t degrade during long-term storage and has both high binding capacity and high protein retention. It comes in a variety of pore sizes and protein capacity, allowing you to optimize for smaller proteins or lower fluorescent backgrounds. However, PVDF does add a few (short) steps to the transfer: you’ll need to pre-wet most PVDF membranes before the transfer and allow them to dry afterwards. Unless you buy specialty low-background PVDF membrane, it also has higher background noise than nitrocellulose membranes.

Pro tip! You can visualize proteins on a PVDF membrane without staining; simply wet your membrane with 20% methanol and place it on a light box after drying.

Nitrocellulose, on the other hand, doesn’t require any pre-wetting and has lower background than PVDF. It is a little quicker to use, with high binding capacity, though it has lower retention after binding and washes. It’s less sturdy and not recommended for stripping and reprobing, unless you buy specialty nitrocellulose membranes specifically developed to allow stripping. And as a safety note, nitrocellulose is flammable, so you may want to avoid it if your lab does a lot of work around a flame.

Membranes come in a variety of pore sizes. If your protein of interest is small, you may want to consider a smaller pore size (0.2 µm). Conversely, if your protein of interest is large, look for a larger pore size (0.45 µm). 

Types of transfers

Your transfer options are wet, semi-dry, or dry/electroblot.

A wet transfer is the most efficient method, meaning that it will transfer the most protein from your gel to your membrane. It requires anywhere from sixty minutes to overnight to run, and is recommended for very large proteins. A semi-dry transfer is faster — 3–30 minutes — and uses less buffer, but it is not as efficient. It is recommended for small to medium proteins. Finally, there are dry/electroblot transfer systems available from various manufacturers. These proprietary systems are fast (often <10 minutes) and simple, utilizing individually packaged membrane ‘sandwiches’. They work well for proteins of most sizes and are worth considering if your proteins of interest vary wildly in size. However, these systems are expensive, generate plastic waste, and limit the amount of optimization that can be done. 

Note that each type of transfer requires different equipment. Wet transfers can be done in an insert in your gel box, while semi-dry transfers require a semi-dry transfer cell. Dry/electroblot transfers only work in the proprietary hardware sold by the manufacturer.

Blocking solutions

Your blocking options come in two categories: protein-based and chemical-based. Before selecting one, check your antibody information. Many antibodies will have a recommended blocking buffer and a recommended concentration. That makes the choice easy!

Protein-based blockers include dry non-fat milk or serums like FBS, PBS, and BSA. For nitrocellulose membranes, which require a lower concentration of blocker, you can also use gelatin. 

Protein-based blockers are inexpensive and easily made in the lab. They can be stored at your bench during the procedure and at 4 °C when you’re not running a western. They are not very stable and will noticeably degrade after several months at 4 °C. 

Pro tip! If you get a speckled image, your protein-based blocker is either degraded or not completely mixed together.

Commercially available chemical-based blockers, while much more expensive, are consistent and stable at room temperature for 1–2 years and are available in formulas specific to your choice of detection methods. They are very useful for troubleshooting or optimizing western blots. If you need to preserve sample, check the ingredients before using, as some contain Tween-20, which can strip proteins from the membrane.

There are also commercially available blockers that work as signal enhancers. These blockers amplify signal without increasing background noise and can reduce the amount of primary antibody required.

Reporters

There are two main classes of reporters: chemiluminescence enzymes and fluorescent proteins. Chemiluminescence enzymes are cheap, fast, shelf-stable before activation, and light-insensitive. They can detect femtogram levels of protein.

These enzymes cannot be used for multiplexing, and the solutions used to provide the chemilumescence may be light-sensitive. They also have a poor dynamic range, meaning that if your blot contains one protein in the femtogram range and another in the microgram range, they cannot be read with equal accuracy. Finally, they degrade quickly and need to be read fairly soon after the reaction has started.

Types of chemiluminescence enzymes

The two most common enzymes are horseradish peroxidase protein (HRP) and alkaline phosphates. HRPs are usually preferred because of their higher specific activity, slightly lower cost, and smaller size. However, HRP is not compatible with sodium azide, a common microbial agent used in antibody buffers, which may be a concern if you’re conjugating your antibodies yourself. Alkaline phosphates have a linear reaction rate and are compatible with antimicrobial agents.

Chemiluminescence kits

When activating your enzyme, make sure to choose a chemilumescence kit appropriately sensitive for the amount of protein you are expecting. 

Fluorescent proteins

Fluorescent proteins, while more expensive, allow for multiplexing. They have a ten-fold greater dynamic range than chemiluminescence enzymes, making it easier to accurately visualize proteins of wildly different amounts in the same blot, and better linearity within detection limits (important if you want to quantify your blots). Fluorescent proteins are stable, so blots can be stored for weeks to months without loss of signal. While not as sensitive as chemiluminescence enzymes, they are more likely to be compatible with stripping and probing.  

Imaging

At the final step of your western, you’ll have two choices: x-ray film or a CCD camera system.

X-ray film is only compatible with chemiluminescence reporters and requires access to a darkroom and a developer. Film is highly sensitive but has a limited dynamic range, which means it may take multiple exposures for a blot with varying signals on it. It requires long exposure times (up to thirty minutes!) to reach the limit of detection. Film also needs to be digitized for quantitative work — and with the advent of online lab notebooks, it’ll likely need to be digitized even if you aren’t quantifying your blot.

CCD cameras have a much higher upfront cost, but require no consumables, produce digital images, reach their limit of detection quickly (~1 minute), and have a linear response over a broad dynamic range that typically spans 2–5 orders of magnitude, which includes a wide dynamic range for fluorescent proteins. They do tend to have increased background with higher exposure times, although newer models can compensate for this by binning.

This is probably the most limiting of steps, as you may only have access to one or another. Most institutes have moved towards CCD cameras over the past few years.

Pro tip! You only need a resolution of ~150 µm to image a blot. This is because proteins diffuse semiradially during the transfer, making membranes inherently “fuzzier” than gels.

Quantification

If you want to quantify the amounts of protein in your blot, you have two options. It’s fairly easy to decide! 

If your protein of interest is available in pure form, you can use it to create a standard curve on your blot to calculate the amount in your sample. (Do ensure your blot is optimized so your standard curve is within your limit of detection!) You’ll need to have enough lanes available for your standard curve, your standard, and your sample(s), as you can only compare samples to a standard curve generated on the same blot.

If your protein of interest is not available in a pure form, you can perform relative quantification, which measures protein amounts and compares the amount of one protein to another. Since antibody binding varies wildly based on proteins, and their differing amounts of available epitopes for that particular antibody, this truly does provide a relative, not absolute, readout.

Normalization

Overall protein expression varies from sample to sample. To compare one sample to another, you’ll need to normalize your protein of interest to overall protein expression for each sample and protein. While ubiquitously expressed “housekeeping” proteins — like GAPDH or actin — are often used, it is far more accurate to use total protein loading.

Total protein loading measures overall protein expression and does not rely on the assumption that a specific protein’s expression has not changed due to ‌experimental conditions. You then normalize your protein of interest to total protein expression. Total protein expression can be measured using a protein stain like Coomassie Blue or Ponceau S. It is far less prone to error than normalizing to so-called housekeeping genes (Aldridge et al., 2008). 

Woah! That was a lot of information to cover. But don’t worry — the next time you’re designing a western blot, you can check out our handy reference table to help you quickly decide what approaches best suit your experimental needs.

Table 3: Optimizing the technical design of a western blot

If you want to…	Consider using…	But the method has/is…
Save time	Direct detection	Decreased signal
	Nitrocellulose membrane	Not recommended for strip/reprobe
	Semi-dry transfer	Decreased efficiency
	Electroblot transfer	Increased cost; decreased efficiency
	CCD camera	High initial (purchase) cost
	Precast gel	Increased cost
Save money	Polyclonal antibodies	Higher non-specific binding; lot-to-lot variability
	Some recombinant antibodies	Not always available; not always cheaper.
	Indirect detection	Increased time; higher background
	Chemiluminescence reporter	Smaller dynamic range; no multiplexing
	Protein-based blocker	Degrades quickly
	Wet transfer	Increased time
	Handcast gel	Increased time; have to make as-needed
Increase specificity	Monoclonal antibodies	Increased cost
	Recombinant antibodies	Not always available; cost varies widely
Increase signal	Polyclonal antibodies	Higher non-specific binding; lot-to-lot variability
	Recombinant multiclonal antibodies	Not always available
	Indirect detection method	Increased time; higher background
	Chemiluminescence reporter	Smaller dynamic range; no multiplexing
	Wet transfer	Increased time
	Specialty chemical-based buffers	Increased cost
	X-ray film imaging	Increased time; increased consumables cost
Decrease background	Direct detection method	Less sensitive; usually more expensive
	Nitrocellulose or low-fluorescence PVDF membranes	Increased fragility (nitrocellulose) or increased cost (specialty PVDF)
	Chemical buffers	Increased cost; may be harsh
Multiplex	Fluorescent proteins	Less sensitive; more expensive
Strip/Reprobe	PVDF membrane	Increased time
Calculate absolute quantification	Running a standard curve	Uses many wells; purified protein not always available
Calculate relative quantification	Normalizing to ubiquitously expressed genes	Not as accurate; increased antibody costs; increased optimization
	Normalizing to total protein loading	Increased time; increased reagents

Good luck and happy blotting!

Many thanks to Addgenies Ashley Waldron, Meghan Rego, and Amrita Rhodes for their help with this piece. 

---

Resources and references

More resources on the Addgene blog

Antibodies 101: The Basics of Western Blotting
Antibodies 101: Selecting the Right Antibody

Antibodies 101: Secondary Antibodies

Resources on Addgene.org

Addgene's Western Blot Protocol

Addgene's Western Blot Protocol Video

References

Aldridge, G. M., Podrebarac, D. M., Greenough, W. T., & Weiler, I. J. (2008). The use of total protein stains as loading controls: An alternative to high-abundance single protein controls in semi-quantitative immunoblotting. Journal of Neuroscience Methods, 172(2), 250–254. https://doi.org/10.1016/j.jneumeth.2008.05.00