In the first world people are not always aware of this, but tuberculosis is still a global epidemic, and according to estimates, about a third of the world's population is infected with it
A first impression of a person we have just met may leave a mark on his image in our eyes for many years. Even the initial encounter of a bacterium with the cells of our immune system leaves an impression on them that may manifest only many years later, and lead to an outbreak of disease. The scientists of the Weizmann Institute of Science traced those "first meetings" - and examined the "reception" of the bacteria at the level of the individual immune cell from the perspective of the overall immune system. Based on these findings, the researchers developed an algorithm that may allow in the future to predict the risk of developing infectious diseases, such as tuberculosis, many years before they break out. The study is published today in the scientific journal Nature Communications.
"Our goal in the laboratory is to look at early events at the beginning of the 'connection' between bacteria and the immune system," explains Dr. Roy Avraham from the biological control department at the institute. "After exposure, the immune system may kill the bacteria, the bacteria may overcome the immune system and cause disease or remain dormant in our body for years. We hypothesize that the decision junction for one of these scenarios is already decided in the initial meeting - 24 to 48 hours after the exposure." To test their hypotheses, the research team called a meeting in a plate between a salmonella bacterium and a human blood sample. The researchers, led by Dr. Naa Busel Ben Moshe and Dr. Sheli Chen Avivi from Dr. Avraham's group, mapped at the resolution of a single cell what each type of cell looks like after it has been exposed to Salmonella - that is, which cell it is and what its activation level is.
Sequencing genetic material at the single cell level from a human blood sample is an expertise currently available in a limited number of laboratories in the world. In contrast to available methods of mapping the genetic material in the entire sample, sequencing at the single cell level allows revealing a new world of information and complexity. Needless to say, this tool is far from being available in clinics and hospitals. How, then, can its great power be harnessed for the benefit of medicine? To this end, the researchers developed an algorithm based on the database they created following the encounter between salmonella and the immune system. The algorithm, based on a computational process of deconvolution (deconvolution), makes it possible to extract information that was not previously available in a standard analysis of blood samples - without needing sequencing at the single cell level. "Perhaps the greatest innovation in the algorithm we developed," says Dr. Bussell, "is the ability to say not only what the composition of the cells is, but also what their activity levels are - that is, not only which cells are in the sample, but also what their reaction potential is."
Our goal in the lab is to look at early events at the beginning of the 'connection' between bacteria and the immune system. After exposure, the immune system may kill the bacteria, the bacteria may overcome the immune system and cause disease or remain dormant in our body for years. We hypothesize that the decision junction for one of these scenarios is already decided in the initial meeting"
The variety of cells of the immune system in a blood sample and their activity levels, before and after exposure to the bacteria (within the circle - the types of cells, outside the circle - subtypes of these cells). With the help of the algorithm it is possible to extract this information from a blood sample without the need to sequence the genetic material at the individual cell level
To put the algorithm to the test, the researchers gathered a group of healthy subjects from the Netherlands, selected based on their medical records. The researchers infected blood samples taken from these people with salmonella, and when they looked at existing analysis methods at the expression of the genes in the sample - no significant differences were found in the response of the immune system; Only after running the algorithm, significant differences were suddenly revealed. "Using the algorithm, we were able to find differences that explain the variation between the samples in the ability to kill the bacteria," says Dr. Chen Avivi. Encouraged by the success, the researchers decided to take a step forward and test the algorithm on a completely different type of bacteria (and much more deadly): Mycobacterium tuberculosis - the bacteria that causes tuberculosis.
In the first world people are not always aware of this, but tuberculosis is still a global epidemic and according to estimates, about a third of the world's population is infected with it. Giving antibiotic treatment to hundreds of millions of people in third world countries does not constitute optimal treatment - especially in light of the fact that 95% of them will never develop the disease. Therefore, treatment is given only when the disease breaks out and symptoms appear, but at this stage the disease is already extremely dangerous, and about two million patients die from tuberculosis every year. What differentiates those who have tuberculosis from those who do not develop the disease? To check if the algorithm has solutions, the researchers turned to a British database that recorded tuberculosis carriers for two years. "The database includes people whose fate is very different - carriers, active patients and those who became carriers during that period," says Dr. Chen Avivi. Although the algorithm is based on an encounter with Salmonella, the researchers were able to discover with it a difference between the subjects in the level of activation of monocyte cells. "Already in the dormant stage, when no one could have guessed that these carriers were going to develop a disease, a significant difference in the level of activation of the monocytes between them and the other carriers is evident. This means that it is apparently possible to detect this difference at early time points, and thus possibly predict the outbreak of the disease," she adds.
"In areas affected by tuberculosis, such as China and Russia or the African continent, there are not enough resources to treat the carriers, since this is the majority of the population. On the other hand, when people are already sick it is too late - they are required to take three types of antibiotics for nine months. Moreover, the bacteria gradually develop increasing resistance to drugs," says Dr. Avraham. Follow-up studies, which will include, among other things, the expansion of the database of the algorithm for tuberculosis, may allow the development of a tool that will predict the risk of contracting the disease - and will allow the medical system to offer preventive antibiotic treatment at a stage when the bacterial load is still smaller, and the chances of recovery are improved. Later, the researchers plan to "train" the algorithm on other infectious diseases as well.
According to the World Health Organization, 10 million people contracted tuberculosis in 2017, of which approximately one million were children, and 1.6 million people died from the disease.
for the scientific article
