The do's and don'ts of Western Blot testing for EPM, and Sarcocystis sp.

S. neurona proteins separated on a gel and stained with Coomassie blue

If you wondered how the tests to diagnose sarcocystosis in horses were discovered you are in the right place! The early work to select an immunodominant antigen served me well for 22 years. Follow along below. A quick spoiler alert: only 0.1-1% of S. neurona infected horses (sarcocystosis) develop EPM.

Fun facts about electrophoresis

A Western Blot (WB) is an analytical molecular technique that isolates proteins, separated by their size from a tissue that is homogenized in a solution. The solution contains many proteins. To get useful information from a WB three things are done: solubilizing the protein, electrophoresis, and detection. A protein of interest is selected from the homogenate by their size using gel electrophoresis. The separated proteins in the gel are transferred to a support (or membrane) and then reagents like serum or specific antibodies are used to identify the protein of interest.

Generally, we are interested in proteins that have a specific size. One must know of the size of the ultimate target of interest— I found my target protein for Sarcocystis neurona from the literature. The published, immunodominant protein of some pathogenic protozoa included those from mice and sheep. The disease causing protozoa (from multiple species) had proteins that were in the 30 kDa range and this size protein was assoicated with pathogenicity, a function that allowed it to make animals sick. Proteins have an atomic mass unit, the Dalton, and k means thousand, so the most interesting size of pathogenic protozoa were 30 thousand kilodaltons, 30 kDa.

Electrophoresis is the process of applying an electric field to a gel. The gel acts like a sieve. The size of the pore can be changed as well as how tightly the proteins separate by the gel matrix by the buffers that are selected. Proteins are charged molecules and when an electric field is applied to the protein, the protein moves in the electric field and filters through the sieve. You can imagine that the homogenate would have multiple proteins with similar if not identical sizes, we will tackle that problem in a minute. For now, our interesting protein to identify horses that have infections with S. neurona are about 30 kDa. Of course, just the size wouldn't distinguish a 30kDa protein from a disease causing or a non-disease causing Sarcocystis.

When I separated my cultured Sarcocystis neurona parasites, I found through trial and error, that some buffers were better than others. A buffer is used to put the proteins in suspension to be applied to the gel. The buffer bis-tris was superior and gave me a sharp banding pattern especially in the 45kDa to 11 kDa range.

As anticipated, the homogenate contained several proteins at the 30 kDa size. In fact, 28-32 kDa proteins would confound many EPM researchers for years to come. I needed to find a protein that was specifically a disease-assoociated S. neurona and a protein to which horses mounted an immune response when they are infected. Incubating the gel-separated proteins that had been transferred to a membrane with serum from a horse with EPM is the tool we used to identify the important proteins. This is called immunoblotting or a Western Blot. The image above is a stained gel that shows that there are a lot of reactive proteins at 31 kDa to 17kDa, the size proteins that captivated researchers interests. Other researchers were looking at the proteins at about 11-17 kDa proteins, I stuck with the higher banding proteins at 29-32kda. I found out later that based on these data the University of Florida patented the use of this protein for diagnosis of EPM, that was a surprise! When I graduated I was able to license using the protein I had worked hard on characterizing. There was still a lot more work to do.

The next step is immunoelectrophoresis

Immunobot of S. neurona probed with sera from a horse with EPM

I immunized some rabbits and mice with the protein homogenate from S. neurona. The anitbodies collected from the immunized animals gave me some specific reagents to visualize proteins of interest. Of course, a monoclonal antibody (an antibody selected to bind only one site on a protein) that would bind the protein of interest would be a prize. I compared the reactions of sera from sick horses (with EPM) with immune rabbit sera and found some common bands. But information was still lacking. After 29 years of research by multple veterinary parasitologists and clinicians there is no protein that distinguishes sarcocystosis from EPM.

How homogenates are prepared makes significant differences to the proteins detected with immune sera on the immunoblot! My next step was realizing that some of the detergents that are used to solubilize the proteins in the homogenate decreased what I detected on my blots. The literature explained that detergents reduce the disulfide bridges in proteins., and as we would later discover, the 30kDa protein has a lot of disulfide bridges. Reduced proteins adopt a random coil conformation that easily separate by size on the gel. This detergent technique is called ‘running the gel under reducing conditions’.

If there was a protein, for example 31 kDa that had two sub-units, a heterodimer, of say 15 kDa and 17 kDa, the proteins would be observed as two bands at 15 and 17 kDA under reduced conditions. However, if the disulfide bonds were unbroken, only one 31 kDa protein would be visible. You can imagine that two labs, one using reduced conditions and one using non-reduced conditions, would be looking at immunoreactive proteins but think they were different sizes. The gel conditions change the important adaptive host responses that are detected.

Another thing, detergents stretch proteins out, using detergents limits detection of the amino acids to those that sit side by side on a protein. The associations in the heterodimer described above would be lost.

S neurona antigens probed with immunized mouse sera

By using a different protocol, I was able to separate proteins in their native condition. Amino acids that are in close proximity in a folded protein are what the host detects. Techniques that keep the disulfide bridging ensures the native structure that is important in a hosts reaction would be available for analysis. Native conformation allows proteins that are not linear to become detectable. Horses produce antibodies to non-linear proteins when they are infected. Non-linear sections on proteins are called conformational epitopes.

Two dimensional WB of S. neurona antigens reacted with sera from a horse with EPM

Protein chemists get quite fancy with their analysis. They use native conditions to separate proteins first by their isoelectric point, this is called the first dimension. Once the proteins are seperated by isoelectric point they are then electrophoresed separating them by size, the second dimension. This technique was used to determine that the 31 kDa band seen on an immunoblot was composed of several proteins, some were immunologically useful, and some were not. The useful ones are detected by serum or cerebrospinal fluid from diseased horses. The two dimensional analysis is called a 2-D Western Blot.

Now the fun part, immunoprecipitation!

Another analytic analysis is a technique commonly used to concentrate proteins of interest. The term for this is immunoprecipitation. Antibodies are incubated with the original, non-reduced, protein homogenate and the antibody hooks on, or captures, interesting proteins. The antibody/protein combination is purified. This way one can define a set of proteins that would identify infected host’s adaptive response to the parasite when compared to a non-infected host. I bet you can already see that comparing the serum and CSF from a diseased horse would show important differences in the immune reactions between the periphery (blood) and central nervous system!

Use the linked button to read the paper, Molecular characterization of a major 29 kDa surface antigen of Sarcocystis neurona and you can see the effect of reducing conditions, it is demonstrated in Figure 3. When you read other scientific papers that describe WB detection of antibodies in serum and CSF from infected horses, be sure to note the conditions used to prepare the blots!

A major 29 kDa surface antigen of S. neurona

Based on the science outlined above, we isolated, purified, sequenced, and cloned the SAG1 protein of S. neurona, the immunodominant protein of the most common serotype to infect horses. Recombinant proteins give us one protein in solution to continue developing a diagnostic test. It is much easier and more meaningful to detect an immune response to a pure protien that it is detecting a tiny amount of a protein in a homoginate. Not surprisingly, epitopes (host binding sites to the recombinant protein) were destroyed by sulfhydryl reduction shown in Figure 7 of my paper. Here is our evidence that conformational epitopes are highly important in diagnosing host reactions to S. neurona infections.

The ELISA tests that I developed use the recombinant proteins of SAG 1, 5 or 6. These proteins are three, mutually exclusive, immunodominant proteins of S. neurona that infect horses. The sulfhydryl bonds are maintained when we run the ELISA. By now, you realize why it is important. Also, using three seperate ELISA's allow us to serotype the infecting strain. An opossum can carry three strains or a combination of the three. It can be important to have this information.

Antibodies against S. neurona are found in most horses. Most horses get a gut infection and resolve it without getting protozoa in the central nervous system (CNS). Antibodies can be found in the CNS but very few horses have organisms there. We believe these infected horses get inflammatory responses that result in clinical signs that can be confused with EPM.

When you examine other ELISA test systems be sure and note the important factors: the protein used, if the protein is prepared under reducing or non-reducing conditions, if the protein represents a pathogenic organism. For example, the 17 kDa protein is one that is part of all apicomplexans, not specific to S. neurona. We found that common proteins were not useful to identify diseased animals. If I combined the SAG’s into one large chimera, fusing the proteins together, I would lose important host-reactive epitopes that we use for case analysis. A chimera would be a Frankenstein protein that horses never see. It would be very difficult to incorporate conformational epitopes into the chimera.

Journeyman's Guide

Now it's your turn!

See if you can answer this question:

If your horse has EPM due to a SAG 1 infection, will a SAG 5 ELISA be positive or negative?

Great job if your answer was negative!

And you can move to the head of the class if you can answer this next question (you’d have to understand the last blog to get this right, go back and read it if you skipped it):

Go Back: EPM Testing: what does a test measure?

If you used proteins that were common to all Apicomplexa for your analysis, would your horse be positive or negative on an ELISA if it had been exposed to S. fayeri and not S. neurona?

(Hint: S. fayeri doesn’t have S. neurona SAG 1, 5, or 6 on its surface but it does share some common antigens.)

Great job if you answer your horse would be positive! One way to overcome this issue is to dilute the sera, trying to dilute out the S. fayeri antibodies on the ELISA. You would report the results at the cutoff dilution, probably about 240.

Send us your questions if you have any! We are happy to explain.