Personalized medicine: T cells fight cancer

In cancer, one of the body's cells stops playing by the rules that govern the cell cycle and divides uncontrollably. This is often due to mutations that lead to errors in the mechanisms that control cell division. Several teams of scientists at the MDC and the Charité are working on a T-cell therapy that specifically targets these mutations in the genome in order to fight tumors. In a publication in The Journal of Clinical Investigation, the scientists explain which mutations are suitable for therapy, and why cell culture experiments sometimes do not suffice.

Personalized cancer treatments tailored to individual patients are still a project for the future. This future is not too distant, though, because ever more powerful technologies are bringing us closer to this goal.

Sequencing one tumor genome costs less than a thousand Euros nowadays. Relatively cheap and fast sequencing processes like this are crucial for the development of cancer therapies that target specific mutations using genetically modified T cells. Thomas Blankenstein and Wolfgang Uckert are working in this area with their groups at the MDC, with funding from the Berlin Institute of Health (BIH ) within a program called “Targeting somatic mutations in human cancer by T-cell receptor gene therapy.” In a publication in The Journal of Clinical Investigation, the Berlin researchers explain how to assess whether mutations are suitable starting points for therapies.

The immune system has mechanisms to detect mutations


Animation: The HLA protein complex (blue) binds the antigen (magenta) in a binding cleft. This antigen complex then binds the T cell receptor. Image: MDC

The idea of targeting cancer therapies to mutations is promising: “Every cancer develops from the mutation of genes, and therefore every tumor carries mutant proteins,” explains Matthias Leisegang, the lead author of the publication. The immune system can recognize some of these mutations as changes on the surface of cancer cells. Fragments of the cell’s own proteins are embedded as “antigens” in HLA (human leukocyte antigen) molecular complexes, which in turn are presented on the cell surface. Which antigens are presented depends, among other things, on the specific version of the HLA protein a person has, and this varies a great deal among different people.

The immune system mobilizes T cells which can bind to these foreign antigens (also called neoantigens) with their T-cell receptors (TCRs). “The fact that T cells can recognize neoantigens was described for the first time in 1995,” Leisegang says.

T cells can basically kill cancer cells that carry neoantigens; obviously, however, they do not stop tumors from developing. Indeed, T cells with TCRs specific to the cancer are usually present in the tumor tissue. They can recognize tumor-specific antigens, but they become inhibited over the long term and develop tolerance for the mutant cells rather than destroying them.

In TCR gene therapy, a mutation-specific TCR is transferred into fresh T cells from the patient’s blood. The T cells that are genetically modified in this way are not functionally limited and, back in the sick person’s body, they can fight the cancer.

The neoantigen is the target structure – therapy must take careful aim

This therapeutic approach requires the neoantigen to be presented on the cell surface. A TCR that recognizes the antigen and activates the modified T cell at a sufficient level is also required. In the study in The Journal of Clinical Investigation, the authors compare different mutations in the CDK4 protein, an enzyme that regulates cell division.

“Mutations in this cell cycle protein have repeatedly been found in cancer cells,” Thomas Blankenstein says. The mutations studied by the group have been found in human skin cancer and are known to promote cancer progression. “Only one building block of the enzyme is changed. The 24th amino acid – which should be arginine – is exchanged for another amino acid, such as cysteine or leucine,” Blankenstein says.


Comparing antigens: The original CDK4 fragment (top) has an arginine (marked in blue), while in the mutants, it is changed to leucine (pink) or cysteine (yellow). Image: MDC

The research team created two lines of cancer cells with CDK4 proteins: one had a cysteine at position 24, and the other leucine. Each cell line presented protein fragments on its surface as an antigen bearing the mutation. In cell cultures, a TCR recognized both mutations equally well and reliably stimulated the “killer program” of the genetically modified T cells. Yet the subsequent animal experiments showed that cell culture tests are not a sufficient substitute for an organism with a complete immune system.

“We developed a new humanized mouse model to test human T-cell receptors and human antigens,” says Wolfgang Uckert. In this model, all cellular components (T cells, tumors) come from the mouse, while the molecules involved in antigen recognition (TCR, HLA proteins), are human-derived.

The researchers tested the two CDK4-neoantigens by treating tumors in animal models that had one of the two antigens. To accomplish this, the scientists injected the mice with T cells with the reactive TCR. The antigen with the leucine mutation appeared to be a good target, as the tumors bearing the change could be destroyed in the mouse. In contrast, the antigen with cysteine was not recognized in the mouse, and the tumors continued to grow.

Using the animal model, the scientists could therefore clearly distinguish between “suitable” and “unsuitable” neoantigens – in a system that is similar to an afflicted human body. This was not possible in the cell culture experiments, which do not sufficiently reflect the complexity of the whole organism bearing the tumor.


This mouse helps finding new cancer therapies. Image: Matthias Leisegang/MDC

Neoantigen-specific TCR gene therapy is an achievable goal

The animal model is suitable for testing the therapy compatibility of antigens and TCRs before clinical application in humans. The researchers are now working to make the animal model more flexible, as it bears only a single version of the human HLA protein. Thousands of different variants exist in the human population.

There is one more hurdle en route to the clinic. Some mutations occur early during tumor growth, others in the later stages of cancer. This means that not all cancer cells share every mutation. To be a suitable target for therapy, an antigen must be found in all tumor cells, including those that have metastasized to other tissues. This requires that different parts of the tumor are being sequenced, creating large amounts of data that must be processed using complex bioinformatic programs. In a publication in cooperation with Hans Schreiber’s research group at The University of Chicago, the Berlin researchers showed that, in principle, this problem can be overcome.

But there is another problem that can arise in cancer patients. Some of the mouse tumor cells did not have a sufficient number of the mutant antigens to be recognized by the T-cell "hit squad." Remnants of the tumor escaped, resulting in relapse – the tumor grew back. In the study with the team from Chicago, the scientists found that applying concomitant radiotherapy significantly improved the chances of a patient's recovery. Tumor cells die off due to the local radiation, which releases the neoantigen and activates the T cells more strongly.

In the experiment, the researchers demonstrated that the amount of antigen presented is crucial. “After we had programmed the cancer cells so that they presented sufficient quantities of the mutant antigen," Matthias Leisegang says, "the therapeutic T cells were effective over the long term.”

When can we expect clinical applications?

The BIH project aims to put the mutation-specific therapy into clinical practice. This therapeutic approach is already regarded as revolutionary and highly promising. The success of the treatment in curing a specific cancer, however, will probably always depend on factors related to individual cases. The tumor analysis, antigen and TCR selection systems also need to be further improved and refined. Furthermore, numerous regulatory requirements must be met before the therapy can be implemented in clinics.

Other variants of the T-cell therapy are already being clinically tested elsewhere. T cells can be targeted to the body’s own antigens, which do not exhibit mutations but are characteristic of specific tumors. This therapy can be very successful with a certain form of blood cancer. This type of antigens is rare, however, and there is a risk that the T cells may attack healthy tissue and cause severe side effects.

A direct comparison between these two T-cell therapies demonstrates the advantage of the Berlin researchers’ mutation-specific method. Since the therapeutic T cells are targeted to an individually selected mutation that is only found on the tumor, they should not produce any side effects through the destruction of healthy cells. This safer approach is technically more complex, but this will be reduced as the process becomes more efficient.

The maxim of many physicians, “no effect without side effects,” will one day be a thing of the past for cancer treatment. “The progress made in recent years is impressive," Leisegang says. "In the not-too-distant future, cancer patients could be treated with personalized and highly specific T-cell therapies."

Matthias Leisegang1,2, Thomas Kammertoens2, Wolfgang Uckert1,3 und Thomas Blankenstein2,4 (2016): Targeting human melanoma neoantigens by T cell receptor gene therapy.“ Journal of Clinical Investigation. doi:10.1172/JCI83465.

1Molekulare Zellbiologie und Gentherapie, Max-Delbrück-Centrum für Molekulare Medizin, Berlin; 2Institut für Immunologie, Charité, Berlin; 3Institut für Biologie, Humboldt-Universität Berlin, 4Molekulare Immunologie und Gentherapie, Max-Delbrück-Centrum für Molekulare Medizin, Berlin. Wolfgang Uckert and Thomas Blankenstein contributed equally to the publication.