Key words

1 Inhibitory Chromosomes and Cancer

Cancer was first posited to have a genetic basis in pioneering work by German geneticist Theodor Boveri at the beginning of the 20th century. In his article entitled In Zur Frage der Entstehung Maligner Tumoren, Boveri investigated chromosomal segregation and abnormal growth characteristics in sea-urchin eggs [1]. Despite the fact that the notion of a “gene” was not articulated until many years later, a direct line can be drawn from Boveri’s near-prophetic predictions about the role of “chromosomen” in cancer and the discovery of countless features of cancer biology. This includes the concept that “a tumor cell that proliferated without restraint would be generated if ‘teilungshemmende chromosomen’ were eliminated” which points to the existence of inhibitory chromosomes, or as we know them today: tumor-suppressor genes [1, 2].

Precisely as predicated by Boveri in 1914, tumor-suppressor genes are now recognized as the safeguards of the cell, which suppress damaged cells from developing into full-blown tumors. Physiologically, a tumor-suppressor gene’s function is almost singularly focused on the preservation of the organism. However, at the cellular level, tumor-suppressor genes have pleiotropic roles including the maintenance, the repair of, or elimination of cells with anything less than an intact and unaltered genome. For instance, upon the occurrence of a spontaneous DNA damage event, tumor-suppressor genes are poised to “sense” the damage, directly or indirectly, and initiate an appropriate course of action prior to cell division. Tumor suppressors often respond by slowing down cell cycling which will allow other tumor suppressors time to repair the DNA, and in other cases where the damage is beyond repair, tumor suppressors initiate signaling to eliminate the damaged cell. It is increasingly apparent that tumor suppression functions as an exquisitely organized network of tumor-suppressor proteins. Critically, when even one of these genes is lost due to gross chromosomal deletion, or incapacitated due to mutation, a an essential cellular defense is disabled. Such an event can increase the chance of generating cells with uncontrollable growth characteristics, and ultimately leading to cancer.

2 PTEN and the Rise of the Tumor Suppressors

From the 1970s well into the 1980s, the oncogene-driven paradigm of cancer dominated, which suggested that it is the over-expression or abberant activation of certain genes which mainly drives transformation. However, the “seed” for the tumor suppressor era was already in place with the work of Henry Harris and his contemporaries demonstrating that the fusion of a cancer cell with a normal cell could suppress tumorigenicity. This seminal work suggested that normal cells contributed “something” to the hybrid cells which suppresses the highly malignant character of the tumor cells” [3]. The next big challenge, and the likely explanation for the long latency for the evolution of a tumor suppressor-driven model of cancer, was the identification of Harris’ “something.”

Although not immediately recognized, a breakthrough came very shortly after in the form of an epidemiological study of patients with retinoblastoma, where Alfred Knudson developed a statistical model that suggested that two distinct genes must acquire mutations for retinoblastoma tumors to develop [4]. This was soon followed by the discovery that the 13q14 locus was frequently deleted in these tumors, and the eventual discovery of the retinoblastoma (Rb) gene [5–7]. Importantly, it was the discovery that both alleles of Rb were lost, which formally validated Knudson’s two-hit hypothesis. This hypothesis proposed that mutation or loss of one allele of Rb is not sufficient to initiate tumor growth; however, deletion or disabling of the other Rb allele results in tumorigenesis [4].

The tumor suppressor era took great strides forward from the mid-1980s until well into the 1990s due to the identification of tumor-suppressor genes that were important in the pathogenesis of a variety of tumors. This was heralded by the ability to designate the localization of allelic loss to specific chromosomal regions. Candidate tumor suppressors identified in this manner included the previously described Rb gene on chromosome 13q14 [5–7], the p53 gene on chromosome 17p13 [8], the Wilms’ tumor gene on chromosome 11pl3 [9, 10], the DCC gene on 18q21 [11], and the BRCA1 gene [12, 13]. In a similar fashion, cytogenetic and molecular studies in the 1980s revealed the frequent presence of partial or complete loss of chromosome 10q23 in brain and prostate cancers [14–18]; this suggested the existence of yet another important tumor-suppressor gene on chromosome 10.

It was not until 1997, almost 20 years ago now, that the identity of a frequently lost tumor suppressor on human chromosome 10q23 was revealed. That year, two independent groups, one led by Ramon Parsons and the other by Peter Steck, cloned and characterized phosphatase and tensin homologue (PTEN) or mutated in multiple advanced cancers (MMAC), respectively [19, 20]. Although Steck et al. did hint at a role for PTEN in hereditary disease, it was a collaboration between Charis Eng and Ramon Parson’s group that formally recognized PTEN as being targeted by germline mutations in patients with the cancer predisposition syndrome known as Cowden disease [21]. In that same year, in their search for novel protein tyrosine phosphatases (PTK), Li and Sun used degenerate oligonucleotide PCR or conserved sequence motif searches designed based on known phosphatase catalytic domains to identify TEPI (TGFβ-regulated and epithelial cell-enriched phosphatase), which also turned out to be PTEN [22]. With those discoveries, the race was on to identify the functions of PTEN in cancer.

3 The Race to Assign a Function to PTEN

By the end of 1997, at least 25 papers on PTEN were published. Already, these provided strong evidence that PTEN was a bona fide tumor suppressor, including a flurry of papers demonstrating that PTEN was mutated in several different cancers including glioblastoma, prostate cancer, melanoma, endometrial carcinoma, high-grade gliomas, thyroid tumors, and infrequently in breast cancer [23–29]. Concurrently, various groups validated the presence of germline mutations of PTEN in three related, inheritable, neoplastic disorders such as Cowden disease, Lhermitte-Duclos disease, and Bannayan-Zonana syndrome [21, 30–33].

Although it was reported that PTEN shared homology with protein tyrosine phosphatases (PTPs) and the cytoskeletal protein tensin in the original discovery papers [19, 20], a study from Nicholas Tonks’ group was the first to demonstrate experimentally that PTEN was functional as a dual-specificity phosphatase, and that naturally occurring PTEN mutations observed in heritable and somatic cancer specimens could disrupt that catalytic activity on synthetic substrates [34]. Additional evidence that PTEN functions as a tumor suppressor was obtained by Furnari et al. who showed that PTEN had a growth suppressor activity in glioma cells [35]. At this point, PTEN represented the first PTP to be implicated as a bona fide tumor suppressor. This was a major breakthrough, because at the time several protein tyrosine kinases (PTKs) were known to act as oncogenes by phosphorylating, and thereby activating, numerous substrates and pathways that led to proliferation, survival, and cell growth. Thus the notion that a PTP could oppose such signals was generally predicted; the discovery of PTEN finally fulfilled that foresight.

In 1998, PTEN was the focus of at least 106 publications and this trend of increasing publications continues through to today (see Fig. 1). Great leaps forward in the understanding of PTEN function were made in this year. The most important discovery, second only to the discovery of PTEN, was the nature of its substrate. The study by Tonks already clearly showed that PTEN could dephosphorylate phosphoserine, phosphothreonine, and phosphotyrosine residues on highly acidic peptide substrates, suggesting that PTEN was a PTP [34]. However, a potential problem was that such highly acidic phosphorylated domains were not common in nature.

Fig. 1
figure 1

Annual frequency of publication plot for five most published tumor-suppressor genes

Full size image

Fundamental insights into the biological function of PTEN would not be revealed until its physiological substrates were identified. Indeed, accepting a substrate contingency, Myers and Tonks debated that PTEN may be a phosphatidylinositol phosphatase in a review paper published in 1997 [36]. In a landmark PTEN paper in 1998, Tomohiko Maehama and Jack Dixon determined that purified PTEN protein was generally poor at catalyzing the dephosphorylation of phosphoproteins [37]. Instead they explored the possibility that PTEN could catalyze an acidic nonprotein substrate. They went on to demonstrate that phosphatidylinositol 3,4,5-trisphosphate (PI(3,4,5)P3) was a much better substrate for PTEN catalysis [37]. Although biologists searching for the elusive PTP tumor-suppressor gene may have been disheartened, this unexpected function for the PTEN tumor suppressor was actually very exciting and immediately shaped a working model where PTEN was in a critical cancer-associated pathway where it opposed the function of the proto-oncogene Akt, which was activated by PI(3,4,5)P3 downstream of PI-3 kinase [38]. That the lipid phosphatase activity of PTEN is critical for its tumor suppressor function was also published that year by the Tonks lab [39].

Later that year, Tamura et al. published two papers describing that PTEN was able to mitigate cell migration through an ability to act as a protein phosphatase by reducing tyrosine phosphorylation on focal adhesion kinase (FAK) [40, 41]. These findings once again opened up the possibility that PTEN may function as both a lipid and a protein phosphatase. Despite a hand full of putative protein substrates, it is still, to this day, unclear whether this is truly the case. In fact, there is compelling data demonstrating that PTEN may actually be its own substrate [42]. Needless to say, it is still debated as to whether it is also a protein phosphatase, mainly because of the large majority of data demonstrating that PTEN dephosphorylates inositol phospholipids in vivo, as demonstrated by mouse models later that year.

The first of these mouse models of Pten was reported by Pier Paolo Pandolfi, where it was determined that Pten is an essential gene that demonstrated tumor suppressive qualities by controlling cellular differentiation and anchorage-independent growth [43]. Shortly thereafter, Tak Mak’s group utilized their own iteration of Pten knockout mice to demonstrate experimentally that PTEN exerts its role as a tumor suppressor by negatively regulating the PI3K/PKB/Akt signaling pathway [44], and confirm that Pten was indeed a bona fide tumor suppressor as mice heterozygous for the Pten gene had an increased susceptibility to develop various types of tumors [45]. A third mouse developed in Ramon Parsons lab confirmed the importance of Pten tumor suppression in multiple tissue types [46]. Curiously, none of the mouse models recapitulated any of the features of Cowden disease, Lhermitte-Duclos disease, or Bannayan-Zonana syndrome. Nor did the mice develop brain tumors, despite the frequent loss of PTEN in glioblastoma. These data point to possible species-specific differences in mice and humans.

By the next year the group of Nikola Pavletich solved the crystal structure of PTEN, leading to important functional insights into how PTEN was regulated [47]. This confirmed the previous work demonstrating that PTEN was indeed a phosphoinositide phosphatase, and added functional and structural characterization of the tumor-derived mutations in this gene. With this newfound understanding of the catalytic site the question of a putative phosphoprotein substrate was again suggested as the catalytic pocket of PTEN is deep and wide enough to accommodate phosphotyrosine or a phosphoserine/threonine residue. Moreover, biochemical inroads exploring downstream pathways of PTEN function were continuously being reported including an ability to activate apoptosis or anoikis, regulate cell cycle and mitigate migration, and regulate cell growth and metabolism (reviewed in ref. 48). Importantly, in addition to all the functional discoveries, a furious stream of studies were published describing more human cancers and hereditary tumor syndromes that were found to have defective or mutated PTEN gene, firmly establishing PTEN as one of the most commonly mutated genes in human cancer.

4 PTEN and the Future

PTEN continues to be an important and oftentimes unique paradigm for tumor suppression. For instance, although PTEN was discovered during the height of Knudson’s two-hit hypothesis, it has been an instrumental example in leading the discussion about the more recent concept of haploinsufficiency and tumor suppression [49]. The concept that PTEN dosage has a profound effect on tumor initiation and progression was demonstrated in a series of hypomorphic Pten mice, in which different levels of PTEN inactivation dictated prostate and breast cancer progression, incidence, latency, and biology in a dose-dependent fashion [50, 51].

Given that the focus was mostly on mutation and function in early days, we have only recently begun to appreciate other critical features of PTEN. For instance the exquisite transcriptional, posttranscriptional, and posttranslational regulation of PTEN is now understood to occur through numerous different mechanisms [52]. PTEN relocalization and its specific function in the nucleus and in extracellular space have identified novel paradigms for PTEN function [53, 54] [see Chapter 2 by Gorbenko and Stambolic]. Moreover, the importance of PTEN in integral cellular processes including cellular senescence, stem cell maintenance, DNA damage repair, and cellular metabolism have also only recently come to light and add to its more established roles in cell cycle, cell survival, and cellular polarity [55]. Important roles of PTEN in non-tumor-associated diseases such as diabetes and autism are also emerging [48].

5 Closing Remarks

To conclude, the histories of PTEN and other tumor suppressors have origins near the beginning of the previous century from the work of Boveri; however a specific understanding of genes with tumor suppressor function accelerated in the last quarter of the century and into the new millennium with the innovative techniques developed by pioneers of the tumor-suppressor gene field including Henry Harris and Alfred Knudson. The research to date indicates that PTEN is a protein that has a function which is exquisitely controlled at numerous levels of cellular regulation. This is not surprising for a protein that plays such a critical cellular role in normal function and in the suppression of tumorigenesis. We await many more major discoveries in the understanding of PTEN biology and expect the trajectory of the PTEN field to continue on its upward trend for years to come.