Vaccines are the safest, cheapest and most effective way to protect against infectious diseases  but to make a good one remains a challenge, and traditional approaches are now stretched to the limit while fatal diseases, like HIV and malaria, remain without vaccine. 
 But a major breakthrough that turns vaccine design on its head has now been published in Nature on the 6th of February - a new computational method that, from the protective antibodies of patients, can design the vaccine specific to induce them (and protect against the disease).

Showing the potential of the new design  Bruno Correia from the Instituto Gulbenkian Ciência and Instituto de Tecnologia Química e Biológica (IQTB) in Portugal and colleagues from the Department of Biochemistry at the University of Washington and The Scripps Research Institute designed a vaccine for the human-infecting respiratory syncytial virus (RSV). The vaccine was tested in rhesus monkeys (which have a very similar immune system to us), and proved to induce protective antibodies. RSV was a particular good example of the vaccine potential,  because not only it causes an often deadly respiratory infection among very young children so it is a dangerous virus, but is also one with which scientists have struggled to make a vaccine for a long time without success.

So how do vaccines normally work, and why there are some more difficult to make?

Nature is full of disease-inducing agents, like viruses or bacteria (collectively known as pathogens or germs) and it is easy to get infected. If we do, our immune system (the cells and organs that protect us against disease) mounts a protective response that, once the pathogen is eliminated, will leave behind a protective immune memory. This memory, if we reencounter the pathogen, can now trigger a much faster and effective attack (called secondary response) that eliminates the threat before disease develops. That is why often we only have a disease once.

Vaccines work similarly, the difference being that that first encounter is not with a live infectious pathogen, but instead with a vaccine that contains a dead, attenuated (weaken) or partial pathogen. Without giving disease these are enough, nevertheless, to create an immune memory that protects the individual if he/she ever comes in contact with the “real thing”.

Image: Post World War II United Kingdom poster promoting vaccination against diphtheria.

But despite all vaccines already developed, some serious diseases, in particular some by fast changing/mutating viruses, like HIV or hepatitis C, remain without protection. The problem is that these viruses change so fast that vaccines (and the immune memory they trigger) become obsolete very quickly. Unless that it is, if they are against those epitopes (the parts of the pathogen targeted by the immune system) crucial for viral survival, what means that they cannot be changed. This is why flu vaccines only work for one year - because the flu virus (influenza) has an extremely high mutation rate.

But even if we use the right epitopes, there is still no certainty that a new vaccine will lead to a protective immune response. The problem is that we still do not fully understand how the protection works - neutralizing antibodies (antibodies capable of blocking the effects of the pathogen) are crucial, but the rest is much less clear. This means that at the moment vaccines are developed empirically (by observation/experimentation), and when this fails we are stuck.

A possible solution (although so far only in theory) are “epitope-focused vaccines”, which turn the process backwards and use the end product – the neutralizing antibodies –to create the vaccine that induces them.

For that it is necessary to identify the epitope targeted by the antibody, and then construct a protein scaffold (a holding structure) to carry it just like in the virus, to be sure to trigger the right antibody response. But despite its potential and many attempts, until now this has not been attained

In the study now published Correia and colleagues develop a computational program called “Rosetta - Fold from Loops” to design new protein scaffolds that - contrary to previous attempts - are flexible, meaning that they can be fitted around the epitope to better mimic the natural viral site. They chose an  antibody used to treat RSV infection, that targets a known viral epitope.As the crystal structure of the virus and the neutralizing antibody bond together was found, this informed researchers of what shape should the scaffold be.

Next came the hard part: to design a structure that held the epitope, while having the right biophysical and structural characteristics to induce the protective antibodies. 

With the new software Correia produced 40.000 possible structures that were then screened in the computer and a few (8) were characterized in the laboratory until only 6 of best final designs were chosen. All the designed epitope-scaffolds bound to the neutralizing antibody, and all had similar, although not identical, scaffold.

But are the new constructs clinically relevant for the disease?

To test that, the researchers looked into individual that had the disease (so have RSV-specific antibodies), testing their sera with the new epitope-scaffolds. Impressively several individuals had antibodies that recognized the “new vaccines”, while none reacted to scaffolds alone

Now could these “new vaccines” induce neutralizing antibodies?

To test for this the different “epitope-scaffolds” formulations were injected into mice and rhesus macaques, which were then tested for a RSV immune response. While some mice could produce antibodies against the virus, none of their antibodies could neutralize the virus. In contrast, after a few vaccination rounds with the new structures the majority of the macaques were producing neutralizing antibodies what was very promising considering the similarity between our immune systems

“Actually the macaques were producing antibodies more potent than the prophylactic antibody that is used to treat high risk patients, the one from where we started,” says Correia, “This when the natural infection exposes multiple viral epitopes, while our scaffolds only have a single epitope supposedly triggering a much more limited immune response. The fact that the results obtained are so good proves the ability of our method to fully explore the immune system abilities when producing therapeutic antibodies. “

What is particularly remarkable in Correia and colleagues’ latest work is how close we become to a real life result as Correia explains, “obviously our structures still need to be optimized and tested in humans, but these results are the first part of the protocol to develop a vaccine against RSV putting us in a good position to create a cheap and effective RSV vaccine.”

This is even more important because of RSV characteristics - a virus responsible for about 7% of all deaths among children between 1 month and 1 year that presents multiple challenges for vaccine design. In fact, not only RSV mutates very fast, but even its non-live vaccines (usually the safest alternative) can not be tested in very young children – the highest priority target population –- since clinical trials had to be stopped because vaccinated infants went to develop a more severe disease instead of being protected. Synthetic vaccines with just one protective epitope,, like the one here described, should however be much safer.

These remarkable new results prove the viability of epitope-focused and scaffold-based vaccine design, opening the door to use these strategies on other illnesses, including HIV. In a world where infectious diseases caused 18.5% of all human deaths and 23% of disability-adjusted life years as recently as 2010, this is no doubt very good news.

Citation: Bruno E. Correia, John T. Bates, Rebecca J. Loomis, Gretchen Baneyx, Chris Carrico, Joseph G. Jardine, Peter Rupert, Colin Correnti, Oleksandr Kalyuzhniy, Vinayak Vittal, Mary J. Connell, Eric Stevens, Alexandria Schroeter, Man Chen, Skye MacPherson, Andreia M. Serra, Yumiko Adachi, Margaret A. Holmes, Yuxing Li, Rachel E. Klevit, Barney S. Graham, Richard T. Wyatt, David Baker, Roland K. Strong, James E. Crowe, Philip R. Johnson, William R. Schief, 'Proof of principle for epitope-focused vaccine design', Nature 5 February 2014 doi:10.1038/nature12966