Anti-cancer RNA vaccines
Boris Klimovich,
Researcher at the Institute of Molecular Oncology,
University of Marburg, Germany.
The source text is published in ninavaccina's Instagram profile.
Two things motivated me to write a post about cancer vaccines:
First, of course, is the awarding of the Nobel Prize for Cancer Immunotherapy.
Secondly, the report read at the Frankfurt cancer conference by Ugur Sahin, founder of Biopharmaceutical New Technologies (BioNTech) and one of the creators of the personal cancer immunotherapy technology described below.
The text will focus only on one of the technologies for creating antitumor vaccines. There are others, but this one is still the most beautiful and, it seems to me, the closest to widespread use.
It is important to understand that in the case of antitumor vaccines, unlike conventional vaccinations, we are talking not about preventive vaccines, but about therapeutic ones.
There are only two preventive vaccines "against cancer": from hepatitis B and human papillomavirus. They are aimed at developing immunity against pathogens that cause malignant tumors, even before meeting with these pathogens in real life.
Therapeutic antitumor vaccines cannot prevent cancer in a healthy person, but they can cure it in a sick person.
To begin with, as always, a little theory.
So, everyone knows that cancer causes mutations, that is, changes in the structure of DNA. These changes "break" a number of genes in the cell. The proteins encoded by these genes are either not produced at all, or differ from normal ones. Some "wrong" proteins perform their function incorrectly, and thereby cause malignant cell growth. Others are simply broken and do not affect the cell in any way. Such "broken" proteins are called tumor antigens, or neoantigens. They are not present in normal cells, and therefore the immune system can recognize them as "alien" and kill the cells that have them.
The idea that cancer cells can be recognized by the immune system is very old, so attempts to create vaccines containing neoantigens to "train" the immune system to fight the tumor have been made for a very long time, since the 50s. However, only one vaccine, Sipuleucel-t, has reached clinical use so far (against prostate cancer).
There are several reasons for the unsuccessful introduction of antitumor vaccines.
Firstly, tumors perfectly protect themselves from the immune system. They synthesize special stop signals that tell lymphocytes: "pass by, there's nothing to see here." A Nobel Prize has just been awarded for the discovery of such signals (they are called immune checkpoints), and checkpoint inhibitors have made a real revolution in oncology (but today we are not talking about them).
The second reason for the failure of most experimental vaccines is that all tumors are different. And the immune systems of patients are also different. Even if we find the same neoantigen protein in many patients with the same type of tumor, prepare a vaccine containing this protein and inject it to patients, it turns out that in some patients the immune system will respond well to vaccination, while in others it will not. Recall that in order to develop an immune response, a foreign protein must be shown to lymphocytes by dendritic antigen-presenting cells (matchmaker cells, as one of our colleagues calls them, because these cells "introduce" lymphocytes to foreign molecules - I will also continue to use the same word). To show the protein to lymphocytes, matchmaker cells "put" its pieces on a "fork" – the MHC molecule (major histocompatibility complex, the main histocompatibility complex).
So: different people have different MHC molecules, and the same foreign protein, put on different MHC "forks", will cause a different immune response. That is, it is impossible to make a vaccine from one tumor neoantigen that will be effective for most patients.
Another important point: a vaccine against any bacterium or virus is a mixture of extremely different antigens. They are very easy to identify: it's like finding a man in a Chewbacca suit in a crowd of people in white coats. Tumor antigens are very similar to normal proteins, they are almost "their own". To find such a person is akin to searching for a Korean in a crowd of Chinese (for the European eye).
And finally: if the vaccine still worked, led to the formation of specific lymphocytes and they attacked the tumor, then it is enough for several tumor cells that do not carry this neoantigen to appear, as the tumor will immediately become invisible to the immune system again. This is called antigen escape, or antigenic avoidance, and also greatly reduces the effectiveness of vaccines.
So, let's summarize.
The creation of therapeutic vaccines against tumors is a real task, the very nature of cancer contributes to this. But such vaccines are still ineffective. Making a vaccine suitable for everyone is hindered by the fact that tumors protect themselves from immunity, are very different from each other, and they have different neoantigens. There are only some types of tumors in which the same neoantigens are often found that can be included in the vaccine. Most tumors are unique in terms of antigens. In addition, matchmaker dendritic cells in different people show the same neoantigen to lymphocytes differently effectively. Moreover, tumors easily "escape" from under immune supervision. Doesn't sound very optimistic, does it? If it were possible to overcome these obstacles, then....
And it seems that a team of scientists from Mainz (who grew up in BioNTech over several years) succeeded. They decided to kill all the birds with one stone and developed a fantastic technology for creating personal antitumor vaccines.
So, let's go.
A working group from Mainz received a biopsy sample of a patient with melanoma.
1. The main problem: how to detect which neoantigens are contained in a particular patient's tumor? The BioNTech team solved this problem simply: they sequence ("read") the entire genome (DNA) of the tumor (or rather, the part of it that encodes proteins), as well as the entire RNA of the tumor (more on this below). In parallel, DNA is read from normal cells. Now it can be done for a few hundred dollars in a couple of days. Then the bespectacled bioinformatics come into play. They compare the genome of a normal cell with the genome of a tumor and look for differences (yes, as in the children's magazines "find 10 differences"). This is how they get a list of all mutations contained in the tumor, a "molecular portrait" that distinguishes a tumor cell from a normal one.
2. The human genome contains about 20 thousand genes (according to new data, about as many genes encode proteins, and the same number of RNA, so the total number of genes necessary for the human body to work is almost 50 thousand), but not all genes work in every cell.
This means that not all mutations found in the tumor are in active genes and lead to the production of neoantigens. It is necessary to identify those mutations that are in the active genes from which the modified proteins are read. To do this, bioinformatics uses RNA sequencing data: if there is a gene RNA in a cell, it means that the gene is active and this protein is produced in the cell. After the real neoantigens are found, you need to choose the best ones.
3. Then complex algorithms come into play, the purpose of which is to determine which antigens are able to cause the best immune response in this patient. Recall that matchmaker dendritic cells show T-lymphocytes not whole proteins, but small pieces (peptides) 10-20 amino acids long. And they do this by "putting" the peptide on the MHC receptor fork. MHC molecules are different for all people, each person carries several variants of MHC genes from mom and several from dad, and there are several hundred variants in total. Each MHC variety can better represent some types of peptides to lymphocytes, and others are worse. Information about which varieties of MHC bind which peptides better and which ones bind worse was obtained by immunologists as a result of a fantastic amount of work, and is now contained in special databases. Therefore, by taking a set of neoantigens, as well as information about which MHC genes the patient has (we know his genome!), we can simulate which antigens are best suited to specific MHC forks of dendritic cells, and thereby select only those neoantigens that have the greatest chance of triggering an immune response. BioNTech selects the top 10 neoantigens for each tumor. This leads to a very important consequence: the vaccine will contain a LOT of antigens. Even if the immune system remains indifferent to several antigens, others will be able to trigger an immune reaction. Also, remember about antigen escape? The tumor can easily escape from the immune attack by losing one neoantigen. It is almost impossible to lose 5 or 8 at once. There is too little chance that 5 genes will turn off in one cell at once. Therefore, it will be much more difficult for her to avoid the attention of the immune system.
4. So, bioinformatics geniuses from Mainz have compiled a complete genetic portrait of their enemy, and have chosen the most noticeable features by which it will be easiest for the immunity of a particular patient to identify cancer cells. What's next? Then we need to make a vaccine. And here lies the main difficulty. We need to somehow prepare 10 small proteins, purify them in large quantities, perhaps attach them to a carrier or adjuvant (if you remember, the immune system does not recognize small proteins and other biomolecules well if they are not attached to something large and immunogenic (for example, vaccines against meningococcus or Hib – hemophilic type b infections). In them, not peptides, but polysaccharides are sewn to the carrier, but the essence is the same. An additional problem is that proteins are difficult to work with. They are all unique, each protein requires an individual purification procedure, they are capricious, easily precipitate, it is very expensive to synthesize them chemically, it is also not easy to work out in bacteria or yeast. And the entire production process needs to be developed and adapted anew for each patient. Impossible, right?
Truth. BioNTech scientists understood this perfectly well, so they came up with an incredibly elegant solution. They thought: what if we give the antigen-presenting cells (matchmakers) not the proteins themselves, but instructions for their production? All information about the structure of proteins in a cell is recorded in DNA. RNA is read from it, and it is the final instruction for protein production. If you want, DNA is the production documentation in the archive of the plant, and RNA is the drawing that lies on the machine of the worker who is grinding the part.
How is RNA better than protein?
RNA is very easy and cheap to synthesize and purify in large quantities. The chemical properties of RNA practically do not depend on what information is recorded in it, it is very simple to make a unique RNA for each patient, you just need to change a few hundred "letters" in the genetic code, and you can do it in a couple of days. And, very importantly, it is quite simple to deliver it to the cage. Then the cell will make a complex protein out of it itself. How is such a vaccine produced?
5. Biotechnologists have come up with a special "framework" of RNA, in which it remains only to insert genetic information about those 10 peptides-neoantigens from the patient's tumor. This framework contains special signals telling the cell where to start reading the protein, where to stop, as well as another special signal-a label that gives the cell the command: "after synthesis, put this protein on the fork-MHC and pull it out to its surface." To begin with, the whole structure is assembled in a test tube in the form of a plasmid – a small circular DNA molecule (it costs about $ 150 and any smart master's student can do it). Then this plasmid is propagated in bacteria, isolated and mixed with the enzyme RNA polymerase. RNA polymerase sits on DNA and copies the entire gene sequence in the form of an RNA molecule. That's it, the instructions are ready, the drawing got out of the printer, it remains to deliver it to the machines to the matchmaker cages.
6. How can the vaccine be delivered to dendritic cells? It turned out that it was very simple: it was enough to inject RNA directly into the lymph node under the control of ultrasound. The main job of matchmaker cells is to grab everything from the surrounding space and present these "treasures" to lymphocytes. They will also be happy to capture the RNA floating around. And immediately they will begin to produce what is drawn in this drawing, namely, a chain of 10 neoantigens, exactly the same as in the tumor. Having made this chain, they will immediately drag it to the surface and attach it to the fork-MHC II (remember, our miracle engineers from Mainz attached a special label for this?). In addition to delivering ready-made instructions for the production of tumor antigens to the matchmaker cell, RNA performs another very important role. The matchmaker cell has special sensors (TLRs) that respond to an excess of RNA inside the cell, because such an excess usually means that the cell is infected with a virus. So, our RNA vaccine "simulates" infection of a cell with a virus. This literally "turns on" the cell, it activates, puts additional red flags next to the fork of the main histocompatibility complex and does its best to show passing lymphocytes that it is infected. The lymphocyte cannot ignore such a matchmaker. Only now the matchmaker will show him not pieces of the virus, but pieces of tumor neoantigens.
7. Everything. It's done. An immune response begins to form. Since this is not happening in the tumor itself, but in a lymph node far from it, where the stop signals released by the tumor do not reach, the immune system is not inhibited and an army of T-lymphocytes is gathering there, ready to destroy the tumor. Brilliant, isn't it?
And what, does it really work, you ask?
Article by Sahin et al. Personalized RNA mutanome vaccines mobilize poly-specific therapeutic immunity against cancer with data from the first human trials of the technology was published in Nature in July 2017 (2 years after the article with tests on mice, this is a fantastic speed!). The authors selected for the first trial 13 patients with stage III-IV melanoma, with multiple metastases (these are patients with poor or very poor prognosis). An individual RNA vaccine was made for each of them, and they received at least 8 doses of the vaccine during the year. Everyone rescheduled the procedure well. Two patients, unfortunately, died during the follow-up period as a result of the progression of melanoma. No relapse was registered in 10 patients during the entire follow-up period (12-24 months). Moreover, after vaccination, multiple progressive metastases resistant to radiation disappeared in one patient (without the vaccine, this person would have been doomed).
In short, the vaccine has shown excellent efficacy in the first clinical trials.
By the way, why melanoma? The fact is that melanomas form on the skin. The skin is constantly exposed to solar ultraviolet; UV is a mutagen, so melanomas have the largest number of mutations of all tumors, that is, there are many neoantigens, so melanomas are very immunogenic.
What's next, you ask? And then a small scientific group from Mainz grew into that terrible FARM, which they love to blame for all their sins and are afraid of. You've probably read something like: it's not profitable for pharma to create effective cancer drugs, because then it will be cured, and they won't be able to sell their terrible chemistry, so they hide all breakthrough therapies. Familiar, right?
Yes, yes, pharma hides breakthrough technologies. In fact, a year after the publication of the first clinical data: BioNTech employs more than 800 employees, they received more than $ 200 million in investments (the figure was reported to the author by an employee of the company in a private conversation), they are conducting a huge clinical study and want to test the technology for 1,500 (!!!) patients (let me remind you, there were 13 a year ago), and not only on melanoma, but on all types of cancer. Therefore, they are introducing night shifts (!) – they do not have time to scale production and quality control yet.
If the first vaccines for patients were produced for about 6 months (not every patient will live), now their goal is 6 weeks (from receiving a biopsy to delivering the vaccine), and they claim that they are close to it. Plus, they are conducting research aimed at finding the causes of unsuccessful vaccinations.
Further prospects? It is necessary to wait for the end of clinical trials. Potentially, it is a technology capable of treating any type of cancer, and relatively inexpensive in mass production. Although, like any technology, there will certainly be its drawbacks and limitations. For example, they found that in one of the deceased patients, despite the effective work of the vaccine, the tumor was still able to outwit the immune system.
What is a fable without morality?
1. If you've ever thought (or read and agreed with it): "oh, we know so little about immunity, it's so complicated to interfere with terrible vaccines" – stop thinking like that. Now and forever. They know so much about immunity that they have learned to create in a matter of weeks a completely individual, tailored to the individual characteristics of the immune system, a vaccine that can cure stage IV melanoma. Conventional vaccines are a sandbox compared to this.
2. There are no wonderful "unconventional" ways to treat cancer. A person claiming the opposite is very likely either a charlatan or a militant idiot (with a small probability he is just very poorly informed). Only science, technology and modern medicine can defeat cancer. The methods of immunotherapy that are developing now are outwardly little distinguishable from a miracle.
Stick to the scientific approach. Do not waste your attention, time and money on bloggers, healers and other naturopaths, especially in serious cases. It can be dangerous for your health.