In previous reports, reasons for the nutritional uptake of infectious agents and their link to some common human cancers (e.g. colon, breast and prostate cancers) have been summarized (zur Hausen, 2001, 2006, 2009, 2012, 2014, 2015; zur Hausen & de Villiers, 2014, 2015; zur Hausen et al., 2017, 2019). Original epidemiological observations have been considered as a first hint suggesting a role of Eurasian dairy cattle in the transmission of those postulated agents (zur Hausen, 2012; zur Hausen & de Villiers, 2015; zur Hausen et al., 2017). This resulted into the search for candidate agents in serum and milk of dairy cattle. In initial attempts, largely based on isolation technology of single-stranded circular DNA of TT viruses (de Villiers & zur Hausen, 2009; Jelcic et al., 2004), several different single-stranded circular DNA molecules (1700–2400 nucleotides) had been isolated (Funk et al., 2014; Gunst et al., 2014; Lamberto et al., 2014; Whitley et al., 2014). They were unrelated to sequences of TT viral genomes. Simultaneously, we found two of such sequences in an autopsy lesion from a patient with multiple sclerosis. Searches of databanks revealed two related sequences in a previous publication of Manuelidis (2011, 2013), which this group had found in materials from transmissible spongiforme encephalopathies (TSEs, “prion”-linked diseases).

A striking feature of all these isolates was their obvious relatedness to specific plasmids of the bacterium acinetobacter baumannii. This raised initially the question whether our isolates could represent bacterial contaminants of the isolated clinical materials or occur during handling procedures in the laboratory. Transfection of cloned DNA into human cells, however, documented active transcription of this DNA as well as replication of re-circularized DNA in specific human cell types (Eilebrecht et al., 2018). This seemed to document an adaption of these plasmid-like sequences to human cells and, as studied later, also to specific bovine cells (Schuster, Bund et al., unpublished data). Their preferential isolation from Eurasian cattle sera and dairy products was the reason for naming them bovine meat and milk factors (BMMFs).

Early attempts to find BMMF DNA in biopsy materials of colon and breast cancers or in cell lines derived therefrom failed entirely, to our disappointment. Thus, we obtained no hints for persistence of their genomes in cells of those tumors for which we suspected a link to these infections. Reasons for their absence, in remarkable difference to presently known tumor viruses (zur Hausen, 2006; zur Hausen & de Villiers, 2014) were unraveled by staining sectioned material from colon, breast and prostate cancers (zur Hausen et al., 2019). The development of monoclonal antibodies against the main open reading frame (Rep) of the isolate, obtained from a multiple sclerosis lesion (MSB1) and staining of colon cancer biopsy sections, resulted in a somewhat surprising outcome. Intensive specific staining occurred not in the precursor cells for colon cancer (the glandular cells of Lieberkühn’s crypts), but in cells of the surrounding lamina propria in foci which were mainly composed of CD68-positive macrophages (zur Hausen et al., 2019). Fig. 1 shows the respective pattern. In more than 70 % of colon biopsies, the peripheral tissue surrounding these crypts revealed this staining pattern.

Fig. 1
figure 1

In A cells in Lieberkühn’s crypts remain unstained. Foci of surrounding cells in the lamina propria, however, are stained by monoclonal antibodies directed against BMMF1 Rep. In B the stars symbolize expected sites of interactions with oxygen radicals. C reveals the Ki-67 staining of nuclei in the crypts, indicating ongoing replicative events. Reproduced with permission

Full size image

The presence of macrophages in regions of stained foci provided evidence for inflammatory reactions and instantly raised the suspicion that the likely production of oxygen radicals may result in mutagenic events, in particular affecting cells during their replicative cycles. CD67 staining for replicative events is very evident in the bottom and medium parts of the glandular crypts of the colon. In the upper part, these cells undergo differentiation and fulfill their glandular functions. Thus, we developed a model of indirect carcinogenesis (zur Hausen & de Villiers, 2014) for colon cancers. This suggests that mutagenesis by inflammation-induced oxygen radicals preferentially affect the replicating double-stranded DNA in Lieberkühn’s crypts resulting in random mutations. In contrast, cellular replication events in the lamina propria, containing in antigen-positive foci mainly CD68-positive macrophages, occur at least 100 times less frequent (zur Hausen et al., 2019).

The further analysis of Rep-antigen positive foci revealed that 8-hydroxy-guanosine was detectable in and around theses foci, indicating a substrate of oxygen radical activity (de Villiers et al., 2019; zur Hausen et al., 2019; Bund et al., to be submitted). Laser-microdissection of Rep-positive foci and DNA extraction regularly permitted the isolation of MSBI1-like BMMF DNA (de Villiers et al. 2020). Interestingly, individual clones from the same isolate revealed different base substitutions in a few sites, probably an indication for mutagenic activity of oxygen radicals also affecting the replicating single-stranded BMMF DNA (Shigenaga & Ames, 1991, de Villiers et al., 2020).

In addition to the more direct demonstration of BMMF mutagenic activity located directly adjacent to the crypt cells, representing the target for cancer development, some remarkable additional correlations exist. They seem to fit perfectly into the arising scheme of colon carcinogenesis:

  • The emerging picture of a selective risk for meat and dairy product consumption from the aurochs-derived Eurasian dairy cattle (zur Hausen, 2012; zur Hausen & de Villiers, 2015; zur Hausen et al., 2019).

  • The preventive effect of non-steroidal anti-inflammatory drugs (NDAIDs), previously summarized (zur Hausen et al., 2019).

  • The non-detectable risk for colon and some other cancers (breast, prostate) after long-time chicken or fish consumption.

  • Up to today, but mainly prior to the application of highly active anti-retroviral (HAART) therapy for persistent human immunodeficiency virus (HIV) infections, a number of publications reported significantly reduced risks for breast, colon, and prostate cancers in long-time HIV-positive Cpersons (summarized in Akkerman et al., 2019; Coghill et al., 2018). Here the induced anti-inflammatory immune suppression repressed the formation of oxygen radicals and thus concomitantly a high mutation rate.

  • The reduced risk for breast and several other cancers after repeated and prolonged breast-feeding periods of mothers (zur Hausen et al., 2019, see below)

Prolonged breastfeeding reduces the risk for several infections and chronic diseases

The first sets of available epidemiological data, the demonstration of specific foci with antigens in the colon lamina propria, and isolation and molecular analysis of BMMF genomes directly from these foci pointed to a role of BMMFs in colon carcinogenesis. In addition, however, other previously reported data seemed to complement this assumption. A number of reports covered the preventive effect of breastfeeding for rota- and noroviruses, considered as the main reason for early childhood mortality of non-breastfed babies. The first column of Fig. 2 summarizes those infections, prevented or reduced in incidence during breast- feeding periods (zur Hausen et al., 2019). The interesting outcome of some of these studies was the discovery of preventive activity also for a number of very different chronic diseases. These include malignancies (acute lymphoblastic leukemias, Hodgkin’s disease and non-Hodgkin lymphomas), chronic metabolic disorders (early onset of diabetes types 1 and 2), one chronic neuropathy (multiple sclerosis), early onset of arteriosclerosis and autoimmune diseases. At least in several of these conditions our group demonstrated specific BMMF1 Rep antigens after staining of biopsy or autopsy materials with monoclonal antibodies. Besides colon cancer tissue, stained cells found in breast and prostate cancer biopsies and in in chronic neuropathies may point to a role of such infections in a broader spectrum of chronic diseases (zur Hausen et al., 2019). Other cancers (e.g. ovarian, lung, esophageal) reveal frequently CD68-positive foci and are reduced in incidence by long-time intake of nonsteroidal anti-inflammatory drugs (zur Hausen et al., 2019).

Fig. 2
figure 2

Prevention or reduced risks for specific infections and chronic diseases by human milk oligosaccharides

Full size image

The protective effect of breast-feeding seems to result from the presence of specific sugar molecules in human milk, which are absent in cow milk products (zur Hausen et al., 2019). The addition of such sugars to nutritional formulas at and after weaning periods seem to successfully substitute the preventive effect of human milk (zur Hausen et al., 2019, Akkerman et al,. 2019; Alisson-Silva et al., 2016).

Repeated and prolonged breastfeeding periods also reveal a reduced risk for mothers, as documented in follow-up studies for multiparous women. Significantly reduced risks for breast, colon, lung and ovarian cancers were described after multiple pregnancies and breast-feeding periods (summarized in zur Hausen et al., 2019). The specific human milk oligosaccharides (HMOs) were also present in blood and urine of pregnant women (zur Hausen et al., 2019).

The mechanism of this preventive effect of HMOs seems to result from blocking of specific cell membrane receptors containing N-5-glycolyl-neuraminic acid (Neu5Gc) as the terminal component of lectin or ganglioside receptors (Samraj et al., 2014). This glycan is absent in humans, due to a deletion in one of the exons of the converting enzymes as outlined in Fig. 3. It is present in most animal products (not in chicken meat and only in re-arranged form in several fishes). Consumption of Neu5Gc-containing meat or of dairy products results in its uptake and incorporation into the respective membrane receptors and changes their properties for binding specific ligands.

Fig. 3
figure 3

Enzymatic defect prevents in humans the enzymatic conversion of Neu5Ac to Neu5Gc. Nutritional uptake of the latter results in modifications of glycan receptors and in different binding properties of those receptors

Full size image

In addition to the more direct demonstration of BMMF mutagenic activity located directly adjacent to the crypt cells, representing the target for cancer development, some remarkable additional correlations exist. They seem to fit perfectly into the arising scheme of colon carcinogenesis, fitting also to some other human cancers (e.g. breast cancer, zur Hausen et al., 2019):

  • The emerging picture of a selective risk for meat and dairy product consumption from the aurochs-derived Eurasian dairy cattle, previously interpreted as the result of chemical carcinogens arising during the roasting, frying or other modes of meat-processing (zur Hausen, 2012; zur Hausen & de Villiers, 2015; zur Hausen et al., 2019)

  • The preventive effect of non-steroidal anti-inflammatory drugs (NSAIDs), previously summarized (zur Hausen et al., 2019)

  • The non-detectable risk for colon and some other cancers (breast, prostate) after long-time chicken or fish consumption (zur Hausen et al., 2019)

  • Prior to the highly active anti-retroviral (HAART) therapy of persistent human immunodeficiency virus (HIV) infections a measurable reduced risk for breast, colon, and prostate cancers in long-time persistently HIV infected persons (summarized in Coghill et al., 2018). Here the induced anti-inflammatory immune suppression repressed the formation of oxygen radicals and thus as a consequence a high mutation rate.

  • The reduced risk for breast and several other cancers after repeated pregnancies and prolonged breast-feeding periods of mothers (zur Hausen et al., 2019).

BMMFs, do they contribute to the aging process?

This caption is considered here as a hypothesis. Presently relatively firm evidence exists for a role of BMMFs in human colon cancer. Rep antigens, BMMF genomes, their RNA discovered in foci surrounding the Lieberkühn’s crypts (cells in the latter may convert to malignant growth), CD68-mediated chronic local inflammation and demonstration of oxygen radical activity, they all support the concept of an indirect carcinogenic role (zur Hausen, 2006; zur Hausen & de Villiers, 2014; zur Hausen et al., 2019) of persisting BMMF infection. One other common human malignancy shares many of the epidemiological characteristics of colon cancer: cancer of the breast (zur Hausen, 2012, 2015; zur Hausen & de Villiers, 2015; zur Hausen et al., 2017, 2019). In biopsies of breast and prostate cancer we find concomitant BMMF Rep-staining and CD68 macrophage infiltration. The studies on these two tumors require further extensions.

Colon, breast and prostate cancers contribute to approximately one third of the global cancer burden. Assuming at present a total global cancer rate of ~ 15.000.000 new cancer cases annually, this amounts to about 5.000 000 cases of those three cancers. Successful prevention and curing of the latter diseases, would this result in a significant reduction of the cancer rate and in an increase of lifespan? Secondary prevention of early cases, patient survival and mortality substantially differ worldwide in different geographic regions, yet, the likelihood for a very significant effect on longevity would probably be small and requires an analysis by professional statisticians. Effective prevention and targeted treatment of persisting infections with the agents discussed here, however, will clearly have one beneficial effect: it will reduce the percentage of patients suffering and, as we trust, even pave the way for a drastic reduction of patients with these malignancies.

Additional aspects emerge as potentially novel: the appearing scenario of a link of chronic neuropathies (multiple sclerosis, Alzheimer disease, Parkinson’s disease, amyotrophic lateral sclerosis) to similar infections with long-lasting inflammations in part for several decades should provide a new focus of medical attention. If we include metabolic diseases (diabetes mellitus), arteriosclerosis and some autoimmune diseases in this spectrum, this may cause substantial reconsiderations of our disease perceptions, relieve us from preconceived dogmas, and provide fresh ideas for prevention, diagnostics and therapeutic approaches.

Aging goes along with a number of burdens: back pain, joint problems, mental deficiencies, endocrine dysfunctions and other deficiencies. Many of them go along with chronic inflammatory reactions, most of those not fully understood by us. An unbiased approach to look into these problems and to understand their causes may still provide us with surprises and new insights.