Lamina-associated polypeptide (LAP)2α and other LEM proteins in cancer biology

Abstract

The LEM proteins comprise a heterogeneous family of chromatin-associated proteins that share the LEM domain, a structural motif mediating interaction with the DNA associated protein, Barrier-to-Autointegration Factor (BAF). Most of the LEM proteins are integral proteins of the inner nuclear membrane and associate with the nuclear lamina, a structural scaffold of lamin intermediate filament proteins at the nuclear periphery, which is involved in nuclear mechanical functions and (hetero-)chromatin organization. A few LEM proteins, such as Lamina-associated polypeptide (LAP)2α and Ankyrin and LEM domain-containing protein (Ankle)1 lack transmembrane domains and localize throughout the nucleoplasm and cytoplasm, respectively. LAP2α has been reported to regulate cell proliferation by affecting the activity of retinoblastoma protein in tissue progenitor cells and numerous studies showed upregulation of LAP2α in cancer. Ankle1 is a nuclease likely involved in DNA damage repair pathways and single nucleotide polymorphisms in the Ankle1 gene have been linked to increased breast and ovarian cancer risk. In this chapter we describe potential mechanisms of the involvement of LEM proteins, particularly of LAP2α and Ankle1 in tumorigenesis and we provide evidence that LAP2α expression may be a valuable diagnostic and prognostic marker for tumor analyses.

Introduction

The nucleus of eukaryotic cells is surrounded by a specialized internal membrane system, the nuclear envelope [1], which is composed of two membrane sheets, the inner (INM) and outer (ONM) nuclear membranes (Fig. 1). INM and ONM merge at the sites where nuclear pore complexes are inserted into the nuclear envelope, and the ONM is continuous with the endoplasmic reticulum [2]. In metazoan organisms, the nuclear envelope also includes the nuclear lamina that underlies the INM and serves as a structural scaffold for the nucleus [3,4]. It is formed by the type V intermediate filament proteins, the A- and B-type lamins, and by a number of integral and associated proteins of the INM. The nuclear envelope, and in particular the nuclear lamina, are involved in nuclear architecture and nuclear mechanical functions [3,5], in chromatin organization and gene regulation through tethering and silencing heterochromatic regions [6-8], and in signaling through recruiting transcription regulators, epigenetic modifier enzymes and signaling molecules to the nuclear periphery [9,10]. In view of the multitude of functions of the nuclear envelope it is not surprising that several components of the nuclear lamina have been linked to various diseases ranging from muscular dystrophies and cardiomyopathies over to lipodystrophies and systemic diseases like the premature ageing syndrome Hutchinson-Gilford progeria [11].

Figure 1.

Schematic overview of the mammalian LEM-protein family. Protein localization and domains are indicated.

LEM-domain containing proteins - a prominent family of nuclear (envelope) proteins

Mass spectrometric approaches revealed that the INM of mammalian cells contains over 80 integral proteins that are expressed in a tissue specific manner [12,13]. The LAP2-Emerin-MAN1 (LEM)-domain containing proteins (Fig. 1) represent one of the best studied family of INM proteins [14]. These proteins share the LEM domain, a 40 aa long bi-helical structure motif, which mediates binding to an abundant and essential chromatin protein in metazoan species termed Barrier-to-Autointegration-Factor (BAF) [15-18]. Therefore all LEM-proteins can associate with chromatin via BAF and it is generally assumed that they are involved in tethering chromatin to the nuclear periphery during interphase [19]. Most characterized INM LEM proteins bind lamins in the lamina and link the membrane to the lamina scaffold. In addition, several LEM proteins were shown to recruit and regulate signaling molecules [20] such as Smads involved in transforming growth factor beta (TGFβ) and bone morphogenetic protein (BMP) signaling [21,22], β-catenin, a transcriptional co-activator of the Wnt signaling pathway, [23], the Lmo7 transcription factor [24], the germ cell less (GCL) transcriptional repressor [25,26], and histone deacetylase 3 (HDAC3) [27].

Mammalian genomes contain seven individual genes that encode LEM-domain proteins (Fig. 1): Lamina-associated polypeptide 2 (LAP2), Emerin, MAN1, LEM2, LEMD1, Ankle1 and Ankle2 [28,14,29]. In addition, Ankle1, LAP2 and LEMD1 generate various isoforms by alternative splicing. Most LEM proteins contain either one or two transmembrane domains and are integral components of the INM. However, two isoforms of the LAP2 gene (LAP2α and ζ) and Ankle1 lack a membrane-spanning domain and localize to the nucleoplasm and cytoplasm [14]. In this review we describe and discuss evidence that the LAP2 isoform LAP2α and potentially other LEM proteins may be involved in cancer development or may serve as useful diagnostic and prognostic markers for some types of cancers.

LAP2 proteins in cancer

The mammalian LAP2 (TMPO) gene, also known as thymopoietin, encodes six splice isoforms (α, β, γ, δ, ε, ζ), all sharing a common ~180 aa long N-terminal domain including the LEM-motif (interacting with the chromatin protein BAF) and an additional LEM-like motif in the very N-terminus, which interacts with DNA directly [15] (Fig. 2). While most LAP2 isoforms differ in their up to ~300 aa long C-terminus only due to the inclusion/exclusion of small, alternatively spliced domains, the ~500 aa long LAP2α C-terminus is encoded by a single exon unique for LAP2α [30]. Unlike the other major LAP2 isoforms, LAP2α’s C-terminus lacks a C-terminal transmembrane domain and folds as an extensive four-stranded antiparallel coiled coil dimer [31] that can also form higher oligomers [32]. In addition, while the membrane bound LAP2 isoforms localize at the INM and interact primarily with B-type lamins of the peripheral lamina [33], LAP2α specifically binds A-type lamins via its unique C-terminus [34] in the nucleoplasm [35]. Furthermore, LAP2α’s C-terminus mediates interaction with the cell cycle regulator and tumor suppressor, retinoblastoma protein (pRb) [36,35,37]. Recent observations that LAP2α may be involved in the regulation of pRb localization and repressor activity led to the hypothesis that LAP2α may play a role in tumorigenesis.

Figure 2.

LAP2α, lamin A/C and pRb interact directly and may form a trimeric complex. Domain organization of human LAP2α and pRb and interaction domains are shown.

Potential role of LAP2α and A-type lamins in pRb-mediated cell cycle control

The retinoblastoma protein (pRb) is one of the three pocket proteins (pRb, p107, p130) [38], which regulate cell cycle transition from G1 to S-phase, cell cycle exit and differentiation in multicellular eukaryotes [39-42]. A plethora of data has shown that pRb is one of the major tumor suppressors by preventing cell proliferation in the absence of strong mitogenic signals, and concordantly, in the majority of tumors the pRb pathway was found deregulated [43]. However, impaired pRb functions in cancer cells are rarely linked to mutations in the RB1 gene (except in the hereditary form of the childhood retinoblastoma disease), but are caused by defects in the expression or activity of upstream regulators or downstream effectors of pRb. Also certain human viruses, e.g. human papilloma virus (HPV) or cytomegalovirus (CMV) impair pRb function by expressing proteins (HPV-E7 and CMV-IE86) that bind pRb with high affinity and affect binding and activity of normal cellular pRb regulators.

In non-tumor cells pRb is regulated by post-translational modifications (Fig. 3), among which phosphorylation by cyclin D/cyclin-dependent kinase (cdk) 4 at the G1/S phase transition, and cyclin E/cdk2 during S-phase, are the best studied ones. Only hypophosphorylated pRb binds to and represses the cell cycle-activating E2F transcription factors (E2F1, E2F2, E2F3), to allow cells to efficiently exit the cell cycle. Hyperphosphorylation of pRb by mitogen-activated cyclin dependent kinases inhibits its repressor function, since E2Fs are released from the complex, leading to transcriptional activation of E2F target genes required for S-phase progression (e.g. cyclin E, PCNA, thymidine kinase) [44]. This basic cell cycle-dependent regulation of pRb and E2Fs is fine-tuned by a complex network of proteins with pro-and anti-proliferative activities, which feed into reinforcing and attenuating signaling loops (Fig. 3). Cell cycle entry initiated by cyclin D/cdk4-dependent pRb phosphorylation and activation of E2F1 is reinforced by a positive feedback loop through E2F-dependent upregulation of E2Fs themselves and cyclins, maintaining hyper-phosphorylated pRb during S-phase [45]. On the other hand, E2Fs activate also negative cell cycle regulators [46], which provide a negative feedback loop to prevent uncontrolled proliferation. Among others, E2F1 activates transcription of RB1, anti-proliferative factors like p19^ink4d (an inhibitor of cdk4) [47], various checkpoint and DNA repair genes (e.g. p73 and ATM) and pro-apoptotic genes (e.g. APAF1, caspases) [44], as well as the transcriptional repressor E2F7 [48] which together with E2F4-6 silence promoters of cell cycle promoting genes independently of pRb [49].

Figure 3.

The pRb/E2F pathway regulates cell cycle progression. The scheme depicts major upstream regulators of pRb and downstream effectors, as well as important pro-proliferative and anti-proliferative feedback mechanisms upon transition from a resting (G0) to a proliferating state and during G1 to S-phase progression. The potential role of nucleoplasmic LAP2α-lamin A/C complexes in the pRb regulatory network is indicated.

How does LAP2α fit into this complex regulatory network modulating pRb function? Both LAP2α [36,37] and its nucleoplasmic binding partners lamin A and C (A-type lamins) [50] bind pRb directly. Several studies have revealed different mechanisms by which these proteins may affect pRb regulation (Fig. 4). LAP2α was found to preferentially interact with hypophosphorylated pRb and is required for pRb anchorage in the nucleus [37]. LAP2α overexpression in cells caused down-regulation of E2F-dependent reporter gene activity and repressed endogenous E2F target genes [36], suggesting that LAP2α promotes pRb-mediated repression of target genes. Accordingly, overexpression of LAP2α in pre-adipocytes promoted cell cycle exit and initiation of differentiation to adipocytes in vitro [36]. In contrast, fibroblasts derived from LAP2α-deficient mice showed impaired pRb repressor activity, upregulated E2F/pRb target gene expression and delayed cell cycle exit upon contact inhibition. Also at the tissue and organismal level, progenitor cells in proliferative tissues showed impaired cell cycle regulation in the mouse model. Proliferating progenitors in the paw epidermis, in colon crypts and skeletal muscle were significantly upregulated, leading to epidermal hyperplasia, increased crypt length and an increased number of fiber-associated satellite cells, respectively [35]. Overall, LAP2α seems to be important for an additional level of pRb regulation on top of the basic regulatory machinery, allowing an efficient activation of pRb repressor activity and pRb target gene repression to promote cell cycle exit (Fig. 3). Upon loss of LAP2α, pRb activity is not lost but impaired, which may explain the observation that LAP2α-deficient mice did not show a clearly increased incidence of cancer during their life time [35]. However, these mice have not been challenged yet with cancer promoting agents or treatments.

Figure 4.

Hypothetical model of the functions of the peripheral lamina and the nucleoplasmic LAP2α-lamin A/C complex in the regulation of the pRb/E2F pathway in non-proliferating and proliferating cells. In arrested cells the lamina may tether and stabilize hypophosphorylated pRb and serve as platform for efficient PP2A-dependent dephosphorylation of phospho-pRb. The nucleoplasmic LAP2α-lamin A/C complex activates repressor activity of pRb leading to E2F target gene repression. In proliferating cells, ERK may release pRb from the lamina, favoring its cdk-mediated phosphorylation. LAP2α-lamin A/C complexes dissociate from phospho-Rb, allowing E2F activation and E2F target gene expression.

A-type lamins have been suggested to be required for pRb protein stabilization (Fig. 4) by preventing its proteasomal degradation [51,52]. Another study has suggested that A-type lamins form a platform allowing efficient PP2A-dependent dephosphorylation of pRb, which is required for efficient TGFβ-induced cell cycle arrest of fibroblasts [53]. Yet another study reported a complex of pRb-lamin A in non-proliferating cells, keeping pRb in a hypophosphorylated state [54]. Mitogenic signal-dependent activation of MAPK signaling leads to translocation of ERK to the nucleus, where it displaces pRb from lamin A, which in turn becomes hyper-phosphorylated by cyclin/cdks. Interestingly, displacement of pRb from lamins did not require ERK’s kinase activity.

Overall, these results indicate that LAP2α and A-type lamins can activate pRb repressor activity and thereby act in an anti-proliferative manner (Figs 3 and 4). The molecular mechanisms how these proteins affect pRb activity are not yet clear. One could imagine various ways, such as stabilization of pRb protein, stable tethering of pRb to chromatin and/or promoters, preventing pRb phosphorylation or mediating efficient pRb dephosphorylation. Interestingly, lamin A has been shown to have repressive activity when artificially tethered to reporter gene promoters [55]. Furthermore, a recent study in C. elegans showed that a muscular dystrophy-linked lamin mutant impaired tissue-specific gene regulation during worm development [7]. Thus, lamin A may not only increase pRb-dependent gene repression, but may have more general roles maybe through recruiting epigenetic modifiers or changing overall chromatin state at promoters.

Another open question concerns the functional relationship between LAP2α and A-type lamins in pRb-mediated gene regulation. A-type lamins are assumed to exist in 2 different sub-compartments in the nucleus: An estimated 90% of total A-type lamins localize to the nuclear envelope as a component of the nuclear lamina scaffold in a LAP2α-independent pool, while ~10% localize throughout the nucleoplasm in a mobile and dynamic pool, most likely in association with LAP2α [56,57,35]. It is unclear whether pRb binds preferentially to a LAP2α - lamin complex in the nucleoplasm or to the peripheral lamina network (Fig. 4). Most studies on the role of A-type lamins in pRb-mediated cell cycle control do not discriminate between these two lamin pools, or assume, without clear experimental evidence that the peripheral nuclear lamina is involved. We favor a predominant role of nucleoplasmic lamins in the regulation of pRb-dependent gene expression for the following reasons: i) Peripheral localization of pRb, as predicted by a preferential docking of pRb to the lamina, has not been observed [58]. ii) Both knockout of LAP2α in mice, which leads to specific loss of the nucleoplasmic lamin pool only, and total lamin A/C knockout, which affects both the peripheral and nucleoplasmic lamin A, showed the same misregulation of pRb and hyperproliferation phenotype in epidermal progenitor cells [35].

LAP2α expression during the cell cycle and in the context of tumorigenesis

A number of studies have shown that LAP2α is highest expressed in proliferating cells and is down-regulated upon cell cycle exit and differentiation [59,37,60,35] or is differentially expressed during the cell cycle in proliferating cells [61]. LAP2α transcripts were also up-regulated during liver [62] and muscle regeneration in vivo [63,64], processes that involve transient controlled proliferation of progenitor cells. Independent studies on the LAP2α promoter, based on chromatin immunoprecipitation and microarray techniques identified E2F1 and c-Myc [65], E2F1 and E2F4 [66], E2F3b [67] and E2F7 [68] on the LAP2 promoter, suggesting that the expression of the LAP2 gene is under direct control of major cell cycle regulators, such as E2Fs. Interestingly both, cell cycle driving (i.e. c-Myc, E2F1) and repressing transcription factors (i.e. E2F3b, E2F4, E2F7) seem to regulate the LAP2 promoter, which points to a complex feedback mechanism. From these studies it remains unclear, though, whether all LAP2 isoforms are similarly regulated. In line with its high expression in proliferating cells, LAP2α was found to be overexpressed in various human tumor samples and cancer-derived cell lines at transcript and protein levels (examples summarized in Table 1).

Table 1. Tumors and tumor derived cell lines, which show upregulated LAP2 expression.

Cancer samples and cell lines with upregulated LAP2 expression	Reference
cervix carcinoma; human	[78,111]
colorectal cancer cell lines; human	[112]
colon cancer; human	[70]
hepatocellular carcinoma; TKO mice	[71]
breast cancer cell line (MCF7); human	[73]
myeloma; human	[69]
pancreatic cancer; human	[113]
gastric cancer; human	[114]
medulloblastoma; human	[115]
lymphoma; human	[81,97]
larynx, stomach, colon, lymphoma, sarcoma; human	[116]

Is cancer-related LAP2α overexpression cause or consequence of proliferation?

Numerous studies found LAP2 or LAP2α overexpressed in tumor cells (Table 1), frequently correlated with the up-regulation of tumor-relevant signaling pathways such as the NEK2-[69], Gli- [70], Notch- [71] and Estrogen- [72,73] pathways. Thus, many studies have shown a consistent correlation between active proliferation and up-regulation of LAP2 transcription during physiological processes and in a variety of cancers. These results are inconsistent with an anti-proliferative activity of LAP2α, predicted from the studies in mice and several cell lines mentioned above. On the one hand LAP2α seems to be up-regulated in most if not all proliferating, particularly cancer cells: on the other hand LAP2α expression is predicted to have an anti-proliferative function. How can one solve this apparent discrepancy? Is LAP2α causally involved in promoting proliferation of cancer cells or is its upregulation a consequence of the high proliferation activity of these cells?

A potential causal positive relation between LAP2α levels and proliferation is only supported by a handful of reports consistent with a pro-proliferative effect of the LAP2α-lamin A complex. Both RNA interference-mediated knockdown of lamin A and LAP2α caused cell cycle arrest in primary human dermal fibroblasts [58]. Furthermore muscular dystrophy-linked homozygous mutations in the LMNA gene in patient-derived fibroblasts, leading to loss of lamin protein expression, severely impaired cell proliferation [74]. Additionally, fibroblasts derived from Hutchinson-Gilford progeria patients expressing the very different lamin A variant progerin, lacking 50 amino acids and permanently farnesylated, caused passage-dependent proliferation defects in culture [75]. Similarly postnatal fibroblasts, but not embryonic fibroblasts, from a progeria mouse model showed proliferative arrest and cell death [76]. However in all these cases, it seems likely that the anti-proliferative effect is caused indirectly by DNA damage and/or deregulation of signaling pathways, which lead to the activation of cell cycle checkpoints. For example progeria cells have been shown to accumulate DNA damage during in vitro passage [75], and mouse fibroblasts from a progeria model have a defective Wnt signaling causing misregulation of extracellular matrix components [76].

Overall, it seems very likely that LAP2α and lamin A have an anti-proliferative activity through promoting pRb repressor activity as described above. The increased LAP2α protein levels in proliferating versus non-proliferating cells and in tumor versus normal cells may simply be a consequence of the mitogen-induced increase in E2F1 activity. We propose that in normal cells the E2F1-dependent up-regulation of LAP2α provides a negative feedback loop that ensures efficient LAP2α-mediated activation of pRb repressor activity causing inhibition of E2F-dependent transcription and cell cycle exit. Cancer cells acquire multiple changes impacting on the pRb and p53 checkpoint pathways. If these changes impair the pRb pathway, LAP2α overexpression would be unable to activate the pRb-mediated negative cell cycle feedback loop. In line with this hypothesis, several studies reported that viral oncoproteins affect LAP2 expression levels via inactivation of pRb [77,78] and p53 pathways [78]. For example, expression of HPV proteins E6 and E7, which are known to inhibit p53 and pRb by direct binding [79], is linked to an upregulation of LAP2α in cervical cancer cells [78]. Knockdown of E6 or E7 in these cells restored p53 and pRb activity, respectively, and caused down-regulation of LAP2α. Conversely, knockdown of p53 in normal human fibroblasts increased LAP2α levels [78]. Similarly, expression of CMV IE86, an inhibitor of pRb led to up-regulation of LAP2 levels [77], and the LAP2 promoter was shown to be regulated by the pRb-p16ink4a pathway [80] and by E2F7, which is a target of p53 [68]. Unfortunately, these studies did not look at the relationship between p53/pRb-mediated LAP2 expression control and tumorigenic behavior of cells.

Overall we conclude that LAP2α overexpression in cancer cells is possible only upon additional changes that promote cell cycle progression, like amplification of c-myc, hyperactivity of mitogenic signaling or loss of repressive factors. This may also explain why overexpression of LAP2α was linked to a worse patient prognosis [81,82]. LAP2α may thus be a useful diagnostic and prognostic gene in tumorigenesis.

LAP2α at telomeres - a link to DNA damage repair?

Several microscopy studies revealed that LAP2α is highly dynamic and changes its localization during the cell cycle. While it localizes throughout the nucleoplasm in interphase and disperses in the cytoplasm upon nuclear envelope breakdown in prophase, it associates with (sub-)telomeric regions on chromosomes during late anaphase and telophase [56] (Fig. 5). In addition, a proteomic approach identified LAP2α in a complex with the telomere repeat binding factor 1 (TRF1) [83], a component of the telomeric shelterin complex that protects and regulates telomeres. The physiological relevance of these findings is still unclear, but several recent observations are consistent with a potential role of LAP2α in a DNA damage-response pathway at telomeres, which is known to be required for functional telomeres [84].

Figure 5.

LAP2α transiently localizes to telomeric regions during late anaphase-telophase (top) and is detected on the chromosome tips in metaphase spreads (bottom). Confocal fluorescence images showing the localization of ectopic, fluorescently tagged LAP2α and Histone 2B (upper panel) or stained for LAP2α and DNA (courtsey of T. Dechat and A. Gajewski, MFPL). Bars, 5μm. Arrow indicates localization of YFP-LAP2α at chromosome tips in late anaphase. Right panel shows a hypothetical involvement of LAP2α in telomere stability based on recently reported protein interactions. See text for details.

Several studies showed that lamin A is involved in DNA repair and telomere maintenance [85-87]. LAP2α was found to associate with Werner helicase, WRN [88], a protein well-known for its role in telomere maintenance and DNA repair [89]. Both Werner helicase and LAP2α were found in two independent studies in a complex with Ku86 [90], a key protein involved in non-homologous end joining DNA repair pathways and in telomere protection [91-93]. LAP2α also appeared among the top hits in an interaction screen of proteins modified with Poly(ADP-ribose) scaffolds [94], which are generated by poly(ADP-ribose) polymerases (PARPs) at sites of DNA damage and serve as docking site for DNA damage signaling and repair proteins. Interestingly, the PARP tankyrase localizes to telomeres and targets TRF1 in S/G2 phase, thereby triggering the release of TRF1 from telomeres and allowing access for telomerase and telomere-processing enzymes.

Overall, LAP2α may be involved in telomere maintenance or protection pathways. However, as LAP2α deficient mice did not show any of the phenotypes associated with dysfunctional telomeres (e.g. premature aging, infertility, increased occurrence of tumors), it seems unlikely that LAP2α is essential for telomere maintenance. Having said this, one has to take into account that telomere biology is substantially different in rodents and humans, including the broader expression of telomerase in mouse versus human cells and tissues and the about 5-10 times longer telomeres in murine versus human cells. Therefore it is still possible that aberrant LAP2α function may have an impact on telomere maintenance in human cancer.

The potential relevance of other LAP2 isoforms in cancer

Two studies have reported a potential link between cancer and LAP2β, the largest membrane-bound LAP2 isoform. LAP2β was upregulated in various digestive tract cancers (stomach, liver, pancreas and bile duct) [95]. Knockdown of LAP2β in cancer cells reduced — whereas ectopic expression of LAP2β increased — cell motility, but had no effect on cell proliferation. LAP2β knockdown also induced significant changes in gene expression [95], which is likely linked to the previously reported interaction of LAP2β with histone deacetylase 3 (HDAC3) and its involvement in gene regulation at the nuclear envelope [96]. LAP2β upregulation was also reported in lymphoma patient samples and in normal human lymphocytes upon mitogenic stimulation with phytohaemagglutinin [97]. Since also LAP2α was found upregulated in phytohaemagglutinin stimulated cells this is likely a consequence of proliferation dependent activation of the LAP2 promoter.

Links of other LEM proteins to cancer

LEMD1, a germline LEM protein is re-expressed in cancer cells

LEMD1 is a mammalian-specific LEM-protein of the INM expressed exclusively in testis as six alternatively spliced isoforms. Interestingly, LEMD1 was initially described as a component overexpressed in colorectal tumor samples [98], prostate cancer [99] and lymphoma cells [100]. Hence, LEMD1 was postulated to belong to the cancer/testis antigens [98], a group of about 40 germ line-specific genes, that are re-expressed in tumors originating from unrelated cell types [101]. However, neither the biological functions of LEMD1 nor its contribution to tumorigenesis are known.

Polymorphisms in the Ankle1 gene are linked to breast and ovarian cancer risk

In breast and ovarian cancer, female and male carriers of mutant alleles of the breast cancer associated genes BRCA1 and BRCA2 bear a dramatically increased risk to develop cancer compared to the reference population. Inherited BRCA1 and BRCA2 mutations account for approximately 5% of all cases of breast and 14% of ovarian cancers [102]. In families with inherited predisposition for breast cancer or with a combined risk for breast and ovarian cancers, the frequency of BRCA mutations is 40% and > 80%, respectively [103]. Some other cases of families with an inherited predisposition for breast cancer show no mutations in the coding sequences of BRCA1 and BRCA2. Therefore, it is assumed that other high risk breast cancer genes exist, or that a combination of genetic variants of low penetrance genes exert additive effects manifesting in a significantly increased risk to develop breast or ovarian cancer [102]. Several recently published studies employed high-throughput genomic analyses of breast and ovarian cancer patient samples aiming at the identification of loci containing such low penetrance genes that may modulate cancer risk [104-108].

Intriguingly, two single nucleotide polymorphisms (SNPs) locating at chromosomal locus 19p13.11 consistently appeared in all of these studies showing a statistically significant link to breast cancer susceptibility. These SNPs are localized within the coding sequence of the gene encoding Ankyrin and LEM-domain containing protein 1 (Ankle1) and lead to amino acid changes within the polypeptide. Ankle1 is a conserved gene in metazoan species encoding an unusual LEM protein described in two recent studies in C. elegans and mammalian cells [109,110]. Ankle1 lacks a transmembrane domain and shuttles between the nucleoplasm and cytoplasm in human cells [109]. Intriguingly, Ankle1 contains a GIY-YIG-type endonuclease domain, which was shown to cleave DNA in vitro and in vivo [109,110]. While Ankle1 overexpression in mammalian cells activates DNA damage response pathways, a C. elegans strain carrying a point mutation in the lem-3 gene — the C. elegans ortholog of Ankle1 — was hypersensitive towards DNA damaging agents. Altogether, these studies suggest that Ankle1 may be an enzyme involved in DNA repair pathways.

Considering that the vast majority of mutations predisposing carriers to breast cancer were identified in genes involved in DNA damage signaling and repair (e.g. BRCA1, BRCA2, Rad51, Chek2, ATM, p53), the endonuclease Ankle1 may indeed be a relevant factor for tumorigenesis.

Abbreviations

CMV

cytomegalovirus

HPV

human papilloma virus

INM

inner nuclear membrane

ONM

outer nuclear membrane

PARP

poly(ADP-ribose) polymerase

SNP

single nucleotide polymorphism