Article in Seminars in Cancer Biology · April 2008 doi: 10. 1016/j semcancer

An update on the Tissue Organization Field Theory

The TOFT predicts that neoplastic phenotypes are potentially reversible through cell-cell and/

or tissue-tissue interactions. This has been verified experimentally [38-40]. A now classical

example is the normalization of teratocarcinoma cells injected into blastocysts [41]. Moreover,

nuclei from a variety of cancer cells transplanted into enucleated oocytes could support normal

pre- and post-implantation development [42,43]. Also, modification of extracellular matrix

components resulted in normalization of the neoplastic phenotype [44-46].

The ability of normal tissues to reverse the neoplastic phenotype is subject to the age and

physiologic status of the host. For instance, mammary gland stroma from mature and

multiparous rats prevented neoplastic development and resulted, instead, in normal ductal

growth of grafted epithelial cancer cells [47]. This tumor development pattern suggests a

parallel to the phenomenon of age- and parity-dependent susceptibility and resistance to

chemical carcinogens in mammary gland neoplasia. As susceptibility to carcinogenesis

decreased with age, the ability of the stroma to normalize neoplastic epithelial cells increased.

This tissue-mediated, age-dependent normalization showed an inverted age-dependent pattern

when rat liver carcinoma cells were tested [48]. In addition, human metastatic melanoma cells

injected into zebra fish embryos acquired a non-neoplastic phenotype, while they formed

tumors when injected into zebrafish once organogenesis was completed [46]. Altogether, these

experiments suggest that the cancer phenotype is an adaptive, emergent phenomenon taking

place at the tissue level of organization.

Experiments designed under a theory-neutral strategy showed that the tissue recombination of

mammary gland stroma exposed to a carcinogen with unexposed, normal mammary epithelial

cells resulted in adenocarcinomas. However, the reverse combination did not [49]. This

observation suggests that the stroma, rather than the epithelium, was the target of the carcinogen

[50] and challenges the notion that carcinogens cause mammary gland adenocarcinomas by

mutating the DNA of the epithelial cells.

In sum, alterations in tissue architecture can and do induce neoplasms, and those neoplasms,

like the sporadic ones in humans, may end up showing aneuploidy [51] and even mutations

[10]. But, as Prehn remarked, “it may be more correct to say that cancers beget mutations than

it is to say that mutations beget cancers.” [52]. Nevertheless, some have recently proposed to

switch the focus of investigation to the search for hypothetical cancer-causing mutations now

in stromal cells [39,53]. It remains unclear how this alternative would overcome the

shortcomings of the SMT [54].