Can We Blame the Tumor for Its Bad Neighborhood?
Article May 17, 2018 | by Vijay K. Ulaganathan, Max Planck Institute of Biochemistry
Our new study suggests we may be looking in the wrong place when it comes to understanding the tumor neighborhood and tumor tissue immune evasion – clues may lie in the germline encoded protein variants.
Cancer is not a disease that occurs in isolation. Our bodies respond to rogue tumor tissues by calling into action various cell types that move into the tumor sites creating a neighborhood that is commonly referred to as the tumor microenvironment (TME).
Among the diverse assemblage of tumor neighbors, are an army of tumor killing soldier cells called cytotoxic CD8 T cells that play a crucial role in prolonging the survival of cancer patients and determining a positive therapeutic outcome. Mysteriously in a large majority of cancer cases, only a few CD8 T cells reach the tumor sites in what is technically termed as immune evasion.
Until now, tumors have been considered wholly responsible for an unpredictable and highly variable immune evading tumor microenvironment. Using genetically engineered animal models a new study, conducted by us at the Max Planck Institute of Biochemistry in collaboration with researchers at the University of Alabama at Birmingham and Duke University Medical Center, demonstrates that individual-specific receptor variants can shape the composition of the tumor neighborhood irrespective of the tumor types.
There is a large body of evidence indicating a complicated role of germline (heritable) mutations in cancer risk, cancer progression and prognosis. However, challenges in generating suitable animal models mimicking every human genetic variant hampers a proper scientific inquiry in this direction. To characterize how a human individual-specific genetic variant works, experiments in genetically engineered transgenic animal models are necessary to establish causality. However, it is a daunting task to accurately shed light on the associated biological mechanisms even with animal models, particularly when the alterations at the molecular level are unexpected and difficult to comprehend.
In a previous 2015 paper, published in the journal Nature, we revealed that a genetic mutation (rs351855-G->A) led to the exposure of a membrane-proximal STAT3-recruiting tyrosine-based sequence motif. This caused elevated levels of phosphorylated STAT3 in tumor tissues – resulting in accelerated tumor growth and poor prognosis. Through comprehensive analyses of publicly available human genome datasets, we determined that almost all of STAT3 enhancing receptor variants occur at an allele frequency of less than 1% in the general population.
This finding is profoundly relevant because unlike somatic variants that are found only in cancer cells, germline variants encoded by these alleles are also found on the surfaces of many immune cell types – as well as tumor cells.To name a few; CD118 p. R862Q (monocytes), CD319 p. E261Q (dendritic cells), CD135 p.L601fs (dendritic cells), CD307d p.C441Y (B cells) and CD334 p.G388R (regulatory T cells), are all examples of human cell surface variants that harbor a membrane-proximal STAT3-recruiting motif on the inside.
STAT3 signaling has diverse functions in different cell types – a pro-mitotic role in cancer cells and lymphoid cells, and immunosuppressive role in myeloid cells.
It is commonly believed that the tumor microenvironment is shaped and dictated by the tumor itself.
Due to my previous experience investigating the migration patterns of T cells in autoimmune animal models, I was aware that germline encoded genes determine how T cells are distributed to various tissues. However, whether cancer-associated human germline variants can play a role in mediating immune-evasion in the tumor microenvironment, independent of tumor types, was never explored before.
In cancers, tumor progression is not only dictated by events within the tumor cells. Progression is also determined by whether the surrounding niche is permissive to the growth at all stages of disease. Most anti-tumor effects of the host immune system can be broadly credited to effector CD8 T cells. Opposing the antitumor activity of CD8 T cells is a population of CD4 T cells (regulatory T cells) which is known to have a regulatory activity on the host immune response.
In the 2018 study, using the common receptor variant FGFR4 (CD334) p.G388R encoded by SNP rs351855-A, we show for the first time, that STAT3-enhancing germline variants are found in higher amounts in immunosuppressive regulatory T cells and mediate immune evasion by lowering the CD8/TREGs ratio in the tumor microenvironment.
For the study the following mouse models were used:
1. SNP knock-in transgenic mice (expressing either wild type SNP rs351855-G/G or cancer-associated SNP rs351855-A/A)
2. Fgfr4 knock-out mice
3. Genetically engineered SNP knock-in mice for two common cancer types namely breast cancer and lung cancer.
The experimental evidence presented in this study supports an important role for individual-specific cell surface protein variants in creating a tissue environment favorable for tumor growth. In addition, the study also provocatively purports the idea that cytotoxic CD8 T cell infiltration in the tumor can be predicted by examining an individual’s genome for the presence of a STAT3-enhancing receptor allele, a finding with valuable relevance in personalized treatment of cancers.
This work is just the beginning with more surprising results sure to follow – from studying human receptor variants instead of a general reference protein. The time is ripe for a new field of precision biology addressing human variant-specific aspects of molecular, subcellular and cellular events.
Written by Vijay K. Ulaganathan, first author of the 2015 and 2016 studies and corresponding author of the 2018 study.
A key statistical test in research fields including biology, economics and psychology, Analysis of Variance (ANOVA) is very useful for analyzing datasets. It allows comparisons to be made between three or more groups of data. Here, we summarize the key differences between these two tests, including the assumptions and hypotheses that must be made about each type of test.READ MORE