Research

Albeit we currently know a large set of cancer risk factors such as direct carcinogens and associated (viral) cancer driver genes, cancer driver mutations and a large selection of systemic/environmental risk factors, we still fail to understand the early causative steps in tumourigenesis of the largest part of cancers. Although uniting features – e.g. the involvement of chronic inflammation and the interplay of several indirect cancer risk factors in tumourigenesis - are broadly accepted, recent synergism analyses pointing e.g. at known indirect carcinogens or the microbiome seem to be incomplete. We specifically addressed the epidemiologic association of certain types of cancer, in particular chronic inflammation (CI)-linked colorectal cancer (CRC), breast cancer (BRCA), prostate cancer (PRAD), liver cancer (HCC), pancreatic cancer (PDAC) and others with the consumption of bovine meat and milk products. The link of high cancer incidences and high meat and milk consumption on the regional level suggests a causative contribution of zoonotic-linked infectious agents - such as Bovine Meat and Milk Factors (BMMF) broadly identified in bovine meat and milk - in indirect carcinogenesis (e.g. zur Hausen, 2012, Int J Cancer, zur Hausen & de Villiers, 2014, Sem Oncol, zur Hausen, 2015, Clin Oncol, zur Hausen, 2017, Curr Top Microbiol Immunol, zur Hausen H et al, 2018, Int J Cancer).
BMMFs are frequently isolated from bovine meat and milk samples but also from human material (de Villiers et al, 2019, Emerg Microbes and Infect) and might represent bioactive risk factors for the development of some of the most common human diseases like e.g. breast and colon cancer, but also neurodegenerative disease like Multiple Sclerosis, Alzheimer's and Parkinson's Disease. Positive results might result in novel options for diagnosis and specific disease prevention and treatment (zur Hausen and de Villiers, 2021, Cancer).
We apply basic research for the development of new translational tools and technology for diagnosis and intervention which we test in preclinical models with the aim to generate real-world cancer prevention.

The group "Episomal Persistent DNA in Cancer- and Chronic Diseases" isolates plasmid-like episomal DNAs mainly from bovine meat (serum) and milk but also from human sample materials (e.g. Gunst et al, Whitley et al, Funk et al, Lamberto et al, Falida et al, 2014-2017, Genome Announc, de Villiers et al, 2019, Emerg Microbes and Infect). The isolates have been termed Bovine Meat and Milk Factors or shortly BMMFs. Epidemiology correlates the consumption of bovine meat and milk with a higher prevalence of e.g. breast and colon cancer on a global scale, pointing towards a zoonotic-linked infection as possible risk factor for disease development (e.g. zur Hausen, 2012, Int J Cancer, zur Hausen & de Villiers, 2014, Sem Oncol, zur Hausen, 2015, Clin Oncol, zur Hausen, 2017, Curr Top Microbiol Immunol, zur Hausen H et al, 2018, Int J Cancer). Importantly, BMMFs were also isolated from patients suffering Multiple Sclerosis. Additionally, degradation-resistant DNAs comparable to BMMFs were found in scrapie-particle preparations (Manuelidis, 2011, J Neurovirol) whereas BMMF-related proteins were identified in mouse and hamster CNS as well as human glioblastoma (Yeh et al, 2017, Proc Natl Acad Sci U S A) indicating a possible link with neuronal diseases. More recently, additional plasmid-like DNA elements with high sequence similarity to BMMF were isolated from goat and sheep milk and milk from water buffalos living in German zoos (König et al, 2021, Viruses).

We assessed bioactivity of specific BMMFs in human cells. Therefore, transfection experiments with different types of BMMF were performed in a human cell line model showing replication, transcription and translation of BMMF genes (Eilebrecht at al, 2018, Scientific Reports). Analysis of phenotypic changes in the human host cells were characterized together with specific host transcriptome analysis based on RNA Seq indicating the use of "humanized" transcription/translation control elements.

To investigate the causative role of BMMF in early tumourigenesis, we developed and evaluated a set of DKFZ-built BMMF antibodies for detection of a conserved BMMF protein (Rep) (Bund et al, 2021, Proc Natl Acad Sci U S A) from a BMMF representative, for which bioactivity (replication, transcription and translation of BMMF genes) was shown in human BMMF cell models, before (Eilebrecht et al, 2018, Sci Rep). A partial structure of the Rep antigen, which was used for the DKFZ-built antibody production, was resolved by X-ray crystallography allowing mapping of antibody epitopes and indicating Rep oligomerization (Kilic et al, 2019, Acta Crystallogr D Struct Biol). The antibodies were systematically applied in histochemical staining on CRC tumour and peritumour tissues and revealed Rep expression in peritumour tissues stained by immunohistochemistry (IHC), out of which BMMF DNA was isolated by laser microdissection (de Villiers et al, 2019, Emerg Microbes Infect).

Rep expression was increased in the peritumour mucosa of CRC patients when compared to healthy controls and linked with increased levels of CD68+ macrophages (basic marker for chronic inflammation (CI)) and 8OHdG (surrogate marker for ROS/RNS radical-induced DNA oxidation) indicating a causative contribution of BMMF to CRC based on CI and (random) DNA mutation (e.g. zur Hausen et al, 2018, Int J Cancer, Bund et al, 2021, Proc Natl Acad Sci U S A). The model of BMMF-induced CI, diffusing ROS/RNS, DNA mutation in replicating crypt cells as precursors for polyps and adenocarcinoma is also supported by BMMF expression identified in the stroma of pre-cancerous polyps (zur Hausen et al, 2018, Int J Cancer and previous publications).

To further stratify the structure of BMMF targets in human cancer tissues, we applied immuno electron microscopy (pre-embedding and structure-retaining cryo-embedding) in CRC, LUAD, PDAC tissues, which were previously tested positive for Rep expression in histochemistry. In all three cancer types, we identified pleomorphic vesicular structures of 50-200 nm in size in the cytoplasm, most likely of macrophages, with association to membranes (Nikitina et al, 2022, Int J Cancer). This and the accessibility of the epitopes suggest intervention strategies based on blocking, immobilization or neutralization of BMMF targets.

Since long, EBV is known as a risk factor for MS. Current studies strongly support an early stage, causative contribution to MS (Bjornevik et al, 2022, Science), but at the same time conclude that a second (strong) risk factor must exist due to the high sero-positivity of EBV in individuals which never develop MS. We proposed the model of a dual infection of the same brain cell with EBV and BMMF (Borkosky et al, 2012, PLoS One, zur Hausen et al, 2017, Curr Top Microbiol Immunol). Upon vitamin D deficiency and concurrent upregulation of and TGF-β, EBV is (re-)activated and stimulates replication of BMMF and synthesis of BMMF proteins. The BMMF antigens will induce an increased concentration of BMMF antigen-reactive antibodies in the peripheral blood of MS patients. We developed a highly sensitive and specific bridging ELISA based on BMMF Rep antigens for the detection of BMMF specific antibodies (support by ORYX Alpha, unrestricted research grant). The test showed significantly increased levels of BMMF antibodies in MS patients when compared to age- and gender-matched non-MS donors. The test aims to overcome existing limitations of MS diagnosis due to the low capacity and high cost of magnetic resonance imaging as the major diagnostic readout and, in addition, allow MS therapy monitoring.

In BMMF mouse models we test different types of BMMF exposure (DNA, protein, milk, uptake as nutrition and/or anal exposure, project funded by the World Cancer Research Fund International, in cooperation with Mathias Heikenwälder and Rene Jackstadt) and any possible induction of inflammation in the relevant target organs (colon, lung, liver, etc.). In addition, the effect of BMMF+ immune cells (in particular macrophages) will be tested based on cell culture/organoid models (e.g. colon, liver and lung organoids), to test an effect on the viability and migration of tumour cells. In all pre-clinical models, we will additionally test the effect of a specific class of sugars (Neu5Gc) described as putative BMMF receptor and possible intervention of BMMF-receptor-binding by specific human milk oligosaccharides (HMO) (de Villiers & zur Hausen, 2021, Cancers).