The telomerase inhibitor PinX1 is a major haploinsufficient tumor suppressor essential for chromosome stability in mice

Telomerase is activated in most human cancers and is critical for cancer cell growth. However, little is known about the significance of telomerase activation in chromosome instability and cancer initiation. The gene encoding the potent endogenous telomerase inhibitor PinX1 (PIN2/TRF1-interacting, telomerase inhibitor 1) is located at human chromosome 8p23, a region frequently exhibiting heterozygosity in many common human cancers, but the function or functions of PinX1 in development and tumorigenesis are unknown. Here we have shown that PinX1 is a haploinsufficient tumor suppressor essential for chromosome stability in mice. We found that PinX1 expression was reduced in most human breast cancer tissues and cell lines. Furthermore, PinX1 heterozygosity and PinX1 knockdown in mouse embryonic fibroblasts activated telomerase and led to concomitant telomerase-dependent chromosomal instability. Moreover, while PinX1-null mice were embryonic lethal, most PinX1^+/– mice spontaneously developed malignant tumors with evidence of chromosome instability. Notably, most PinX1 mutant tumors were carcinomas and shared tissues of origin with human cancer types linked to 8p23. PinX1 knockout also shifted the tumor spectrum of p53 mutant mice from lymphoma toward epithelial carcinomas. Thus, PinX1 is a major haploinsufficient tumor suppressor essential for maintaining telomerase activity and chromosome stability. These findings uncover what we believe to be a novel role for PinX1 and telomerase in chromosome instability and cancer initiation and suggest that telomerase inhibition may be potentially used to treat cancers that overexpress telomerase.

It has become evident that inactivation of tumor suppressor genes due to gene alterations, notably loss of heterozygosity (LOH), plays a major role in the development of common human adult cancers, with breast cancer as a notable example (1–3). Chromosome 8p23 is one of the most frequent LOH regions in common human adult epithelial malignancies, including breast, liver, lung, and gastrointestinal cancers. For example, up to 70% of hepatocellular carcinomas (4–8) and 60% of human gastric cancer (9) exhibit LOH at 8p23 near the marker D8S277. 8p23 is also a common integration site for HBV, a well-known major risk factor in liver cancer (5, 10). Similarly, LOH on 8p is found in up to 50% of breast carcinomas and is often associated with advanced tumor stage and aggressive histology (11–14). Although several tumor suppressors have been mapped to this region, including transcriptional factors Nkx3.1 at 8p21 and FEZ1/LZTS1 at 8p22 (15–17), even the combined rates of loss of these genes could not account for the extensive alterations seen in human tumors (16), indicating that major tumor suppressor gene or genes remain to be identified.

Telomerase is activated in most human cancers (18, 19). Telomerase elongates telomeres, which cap the ends of linear chromosomes and are essential for maintaining chromosome stability (20–25). Its activity is absent or very low in most normal human somatic cells so that telomeres shorten during each cell division. However, telomerase activation is critical for transforming primary human cells (26) and for enabling transformed cells to escape from crisis (27, 28). Moreover, transgenic telomerase reverse transcriptase catalytic subunit (TERT) overexpression in mice induces tumors in a telomerase RNA component–dependent (TERC-dependent) manner (29–32) and also cooperates with p53 knockout in inducing spontaneous cancer development (31), analogically to telomerase knockout (33) or telomere deprotection (34). In addition, telomerase regulates DNA damage response (35, 36) and can also promote epithelial proliferation through transcriptional activation by serving as a cofactor (37). Telomerase activation is thus important for cancer cell growth. However, there is no genetic evidence linking telomerase activation to chromosome instability, making it difficult to link telomerase activation to cancer initiation. Moreover, while transcription of the telomerase catalytic subunit TERT is well known to be activated by deregulation of many oncogenes and tumor suppressors (38–41), little is known about the inhibition of telomerase activity and its significance in oncogenesis.

In mammalian cells, the ability of telomerase to elongate telomeres is regulated by telomere-associated proteins (24), including telomeric repeat binding factor 1 (TRF1) (42) and its associated proteins (43, 44), including PIN2/TRF1-interacting, telomerase inhibitor 1 (PinX1) (45). However, unlike other TRF1-binding proteins, PinX1 is unique in that it can also directly bind to TERT and inhibit telomerase activity (45). Furthermore, inhibition of PinX1 in human cancer cells increases telomerase activity, whereas PinX1 overexpression has the opposite effect (45). Moreover, PinX1 is recruited to telomeres by TRF1 and provides a critical link between TRF1 and telomerase inhibition to help maintain telomeres at the optimal length (46). The ability of PinX1 to regulate telomerase and telomere length is conserved in yeast, rats, and fish (47–49). However, it is not known why such a telomerase inhibitor is needed in vivo.

Notably, the PinX1 gene localizes to human chromosome 8p23 near the marker D8S277 (8, 45). Furthermore, PinX1 expression is reduced in approximately 40% of HBV-related liver cancer (8, 50). Moreover, PinX1 inhibition increases, whereas PinX1 overexpression suppresses, tumorigenicity of cancer cells (45). These results suggest that PinX1 might be a putative tumor suppressor. Subsequent PinX1 studies on human cancer samples provide some supportive evidence (9, 50) and contradictory results (51, 52). Furthermore, the interpretation of these expression results is complicated because they all used RT-PCR analyses that could detect other alternatively spliced PinX1 variants and a potential PinX1 pseudogene in the genome. Therefore, expression of PinX1 in human cancer tissues remains unclear. Moreover, there is no genetic evidence for any involvement of PinX1 in tumorigenesis.

To determine the role of PinX1 in cancer, we first examined PinX1 expression in human breast cancer tissues and cells and then generated PinX1 knockout or knockdown and telomerase knockout or knockdown to examine the impact of PinX1 on tumorigenesis in vitro and in vivo. Our results provide what we believe is the first evidence for PinX1 as a major haploinsufficient tumor suppressor essential for maintaining chromosome stability and link for the first time, to our knowledge, aberrant telomerase activation to chromosome instability and cancer initiation. These findings also suggest that telomerase inhibitors may be effective in treating cancers that overexpress telomerase.

PinX1 expression is reduced in most human breast cancer tissues and cells. To examine the role of PinX1 in oncogenesis, we first used multiple methods to examine PinX1 expression in commonly used human breast cancer cell lines. PinX1 expression was readily detected in normal breast epithelial cell line MCF-10A, but its expression was reduced to variable degrees in 6 out of 7 breast cancer cell lines examined, as determined by quantitative RT-PCR (qRT-PCR) analysis of a PinX1 mRNA fragment covering all the 7 coding exons (Figure 1A) and confirmed by immunoblotting (Figure 1B) and immunostaining (Figure 1C) analyses using affinity-purified anti–C-terminal PinX1 antibodies that we generated. Immunostaining also confirmed that PinX1 was localized to nucleoli in addition to telomeres (Figure 1C), as shown (45, 47). These results indicate that PinX1 expression is reduced in most human breast cancer cell lines.

Figure 1

PinX1 expression is reduced in most human breast cancer tissues and cell lines. (A–C) PinX1 reduction in human breast cancer cell lines. Human normal and cancerous breast cell lines were subjected to qRT-PCR amplification of a 1.0-kb PinX1 mRNA fragment covering all the 7 coding exons, with GAPDH as a control (A), to immunoblotting with anti-PinX1 antibodies with actin as a control (B), and immunofluorescence staining with anti-PinX1 antibodies (C). Original magnification, ×63. (D) PinX1 reduction in human breast cancer tissues. Serial sections of tissue microarrays of 10 normal and 49 tumor breast specimens were subjected to immunohistochemistry using anti-PinX1 antibodies. In each sample, PinX1 expression was semiquantified in a double-blind manner as high, medium, or low according to the standards presented in D and summarized in Table 1. Original magnification, ×20; ×40 (insets).

Next, to examine PinX1 expression in human breast cancer tissues, we subjected serial sections of tissue microarrays of 10 normal and 49 tumor specimens to immunohistochemistry to determine PinX1 expression in a semiquantitative manner (Figure 1D). Out of 10 normal breast tissues, 9 contained high levels and 1 expressed medium levels of PinX1 (Figure 1D and Table 1). However, in breast cancer tissues, only 10% of specimens expressed high levels of PinX1, whereas 41% and 49% of samples contained medium and low levels of PinX1, respectively (Figure 1D and Table 1). The differences in PinX1 expression between normal and breast cancer tissues were highly significant, as determined by the Spearman’s rank correlation test (P < 0.01). Thus, PinX1 expression was reduced in most breast cancer tissues and cells examined.

Table 1

PinX1 expression is reduced in most human breast cancer tissues

While PinX1-null mice are embryonic lethal, PinX1 heterozygous knockout reduces PinX1 levels and increases telomerase activity and telomere length in mouse embryonic fibroblasts and mice. To determine the significance of PinX1 downregulation in oncogenesis, we generated PinX1-knockout mice. To create the PinX1 targeting construct, we subcloned 3 genomic fragments of PINX1 into the pKOII vector (Figure 2A), which was linearized and electroporated into ES cells, resulting in 2 independent ES clones that had the correct recombination (rec) of the PinX1 targeting vector at the PinX1 locus (Figure 2A), as confirmed both by PCR and Southern analyses (data not shown). Both PinX1^+/rec ES clones were injected into mouse blastocysts, giving rise to 9 and 2 chimeric mice, which were designated as A and B lines, respectively, and contained the germline transmitted PinX1^rec allele in both lines upon backcrossing. To generate PinX1-knockout mice, we crossed heterozygous PinX1^+/rec mice with CMV-Cre transgenic mice to generate a conventional PinX1-knockout allele (PinX1) (Figure 2A), as determined by Southern blot (Figure 2B) and PCR analyses (Figure 2C), as described (53). Of note, all the phenotypes as described here are found both in A and B lines of mice, indicating that it is unlikely that the phenotypes are due to a random integration of the construct.

Figure 2

PinX1 expression is gene-dosage–dependent, and PinX1 heterozygous knockout reduces PinX1 expression and increases telomerase activity in vitro and in vivo. (A) Construction of the PinX1 targeting vector and deletion of the PinX1 gene by Cre- and loxP-mediated homologous recombination. Cre-mediated deletion of the PinX1rec allele generates the PinX1^– allele. (B) Identification of PinX1-KO mice, as confirmed by genomic Southern analysis using a 5′ probe. Black and red arrows point to the expected products of WT and KO alleles. (C) Identification of PinX1 KO mice by 2 different sets of PCR amplification using 3 primers each. (D) Gene-dosage–dependent expression of PinX1 mRNA in mouse embryos, as determined by qRT-PCR amplification of 1-kb full-length PinX1 mRNA in PinX1^+/+, PinX1^+/–, and PinX1^–/– embryos at E9.5. (E and F) Gene-dosage–dependent expression of PinX1 protein in liver tissues of adult PinX1^+/+ and PinX1^+/– mice (E) and MEFs derived from PinX1^+/+ and PinX1^+/– embryos at E12.5 (F), as determined by immunoblotting using anti-PinX1 antibodies. Unrelated proteins (E, marked by asterisks) and/or actin (E and F) were used as loading controls. (G and H) Elevated telomerase activity in primary PinX1^+/– MEFs. Telomerase-containing fractions were subject to the standard TRAO assay, stained with SYBR green (G), and semiquantified as ratios between telomerase products and the internal control (IC, arrow) (H). B. lys, boiled lysates. (I and J) Elevated telomerase activity in young PinX1^+/– testes tissues. Telomerase-containing fractions isolated from 3 PinX1^+/– or PinX1^+/+ littermates were subject to the standard TRAP assay (I) and semiquantified, with PinX1^+/+ telomerase activity being set at 100% (J). Data are represented as mean ± SD.

We failed to obtain any PinX1^–/– mice from intercrosses with PinX1^+/– mice (Supplemental Figure 1E; supplemental material available online with this article; doi: 10.1172/JCI43452DS1). Although no obvious differences were observed among E8.5 embryos with different PinX1 genotypes (Supplemental Figure 1A), PinX1^–/– embryos were much smaller and very pale at E9.5 and E10.5, appearing devoid of a blood supply and often without a detectable heartbeat (Supplemental Figure 1, B and C). There were no live PinX1^–/– embryos at E11.5 or E12.5, whereas PinX1^+/– embryos appeared normal (Supplemental Figure 1, D and E). PinX1^+/– mice were viable and appeared to thrive, but their ratio to PinX1^+/+ mice was close to 1:1 (Supplemental Figure 1E). Because no neonatal lethality was detected, a fraction of the PinX1^+/– embryos had likely died in utero. Thus, PinX1 is essential for embryonic development, as are other telomere-related proteins, including TRF1 (54) and Tin2 (55).

PinX1^+/– mice were viable and appeared to thrive, but their ratio compared with PinX1^+/+ mice was only 48%:52% (Supplemental Figure 1E). Because no neonatal lethality was detected, a fraction of the PinX1^+/– embryos must have died in utero, suggesting a haploinsufficiency. To further examine this possibility, we used qRT-PCR and immunoblotting to determine levels of PinX1 mRNA and protein in mouse adult tissues, embryos at E9.5, and mouse embryonic fibroblasts (MEFs) derived from PinX1^+/+, PinX1^+/–, and PinX1^fl/fl embryos at E12.5. Compared with WT controls, PinX1 mRNA was reduced by approximately 60% in PinX1^+/– embryos, but was not detected in PinX1^–/– embryos (Figure 2D). PinX1 protein in PinX1^+/– livers and MEFs was reduced by 60%–70% (Figure 2, E and F). These results indicate gene-dosage–dependent PinX1 expression in vitro and in vivo.

To examine whether reducing PinX1 affects telomerase activity, we compared telomerase activity in PinX1^+/+ and PinX1^+/– MEFs and testis tissues using the standard telomere repeat amplification protocol (TRAP) assay. Telomerase activity in multiple independent PinX1^+/– MEF lines (Figure 2, G and H, and Supplemental Figure 2, A and B) and testis tissues from 3 mice (Figure 2, I and J) was reproducibly increased by roughly 2-fold, compared with PinX1^+/+ controls. Furthermore, similar increases in telomerase activity were observed when various amounts of proteins were added to the assays (Figure 2, G and H). This small telomerase increase might be expected given that these normal cells and tissues do not express much telomerase, since PinX1 depletion by 70% leads to an approximately 5-fold increase in telomerase activity in HT1080 cancer cells (45). Thus, PinX1 heterozygous knockout results in a reproducible increase in telomerase activity in vitro and in vivo.

To further determine whether this small telomerase activation is functionally relevant, we examined whether it affects telomere length in PinX1 heterozygous knockout MEFs and mice over time. Primary PinX1^+/+ and PinX1^+/– MEFs were continuously cultured for approximately 30 passages (~90 population doublings [PD]) using the classic 3T3 protocol (56), and then their telomere lengths were assayed at early and late passages (3–5 and 30 passages, respectively) using quantitative FISH (qFISH) and telomere Southern blot. qFISH results showed that PinX1^+/+ and PinX1^+/– MEFs had similar telomeric signal intensity at early passage (Figure 3, A–C), but PinX1^+/– cells had stronger telomeric signal intensity at late passage, with the average telomere fluorescence unit (TFU) being increased by approximately 40% (Figure 3, D–F). Although it is notoriously difficult to use telomere Southern blot to precisely measure such long mouse telomeres, especially telomere lengthening, basically similar trends in telomere elongation were also observed in telomere Southern blot. PinX1^+/– and PinX1^+/+ MEFs had again similar telomere lengths at early passage, but PinX1^+/– cells contained elongated telomeres at late passage in multiple independent clones, as evidenced by an increase in the telomeric hybridization signal and in the average TRF length by approximately 45% (Figure 3, G and H; P < 0.01). Thus, PinX1 knockout in MEFs leads to telomere elongation, further supporting telomerase activation.

Figure 3

PinX1 heterozygous knockout leads to telomere elongation in MEFs and mice. (A–F) Telomere elongation in PinX1^+/– MEFs at late passage, as determined by qFISH. PinX1^+/+ and PinX1^+/– MEFs at early (from 3 to 5) (A–C) and late (~30) (D–F) passage were fixed and hybridized with a FITC-labeled PNA (CCCTAA)₃ probe, followed by quantifying TFUs in approximately 2,000 telomeres, with a telomere distribution curve being shown in C and F. (G and H) Telomere elongation in PinX1^+/– MEFs at late passage, as determined by telomere Southern blot. Multiple independent clones of PinX1^+/+ and PinX1^+/– MEFs were cast into plug molds, lysed, and digested, followed by Southern blot analysis using a TTAGGG repeat as a probe. Prior to hybridization, the gels were stained with ethidium bromide to insure equal loading of total DNA, with a segment of the gels being shown in lower panels (G), with average TRF lengths from 3 independent MEF lines being quantified using ImageQuant (H). (I and J) Telomere elongation at old (10–13 months) (J), but not young age (6–10 weeks) (I) of PinX1^+/– mice, as determined by qFISH. (K and L) Telomere elongation at old, but not young age of PinX1^+/– mice, as determined by Southern blot analysis of splenocytes (K), with average TRF lengths from 5–6 littermates being quantified using ImageQuant (L). The lanes were run on the same gel, but were noncontiguous. **P < 0.01. Data are represented as mean ± SD.

To determine whether PinX1 knockout affects telomere-mediated DNA damage, PinX1^+/+ and PinX1^+/– cells at early and late passages before and after γ-radiation were subjected to double staining with anti-p53BP1 and qFISH with a telomeric PNA probe, as described (57). Before irradiation, there were not any obvious p53BP1 foci in PinX1^+/+ and PinX1^+/– cells at either early or late passage (Supplemental Figure 4A). In contrast, radiation induced prominent p53BP1 foci in these cells (Supplemental Figure 4B). However, the number or intensity of p53BP1 foci was not significantly different between PinX1^+/+ and PinX1^+/– cells either at early or late passage, and very few p53BP1 foci were colocalized with telomeres (Supplemental Figure 4B). These results suggest that PinX1 heterozygous knockout may not significantly increase telomere-mediated DNA damage. These results might be expected because these cells contained elevated telomerase activity (Figure 2, G and H) and telomerase does not induce DNA damage, but rather improves DNA repair (35, 36).

Since elevated telomerase activity was also found in PinX1 heterozygous knockout mouse tissues (Figure 2, I and J), we determined whether the time-dependent effects of PinX1 on telomere lengths also occur in mice by comparing telomere lengths in spleen cells isolated from both young (6–10 weeks) and old (10–13 months) PinX1^+/+ and PinX1^+/– littermates. As recently reported (58), both qFISH (Figure 3, I and J) and Southern blot (Figure 3, K and L) analyses consistently showed that older PinX1^+/+ mice had shorter telomeres than younger mice and, more importantly, telomeres in PinX1^+/– mice did not show age-dependent shortening, but rather lengthening. Young PinX1^+/+ and PinX1^+/– mice had similar TFUs (Figure 3I) and TRF lengths (Figure 3, K and L). However, older PinX1^+/– mice had increased TFUs and longer TRF lengths, with their average increases by 51% (Figure 3J) and 56% (Figure 3, K and L), respectively. In addition, PinX1 knockdown in SV40-immortalized MEFs (Figure 4D) also significantly increased telomerase activity (Figure 4E and Supplemental Figure 2C) and telomere length (Figure 4F), as described below. Of note, unlike PinX1^+/– or PinX1 shRNA MEFs at late passages (Figure 3F and Figure 4F), the entire telomere distribution curve in old PinX1^+/– splenocytes was not shifted to the right (Figure 3J), which might be expected given that total splenocytes contain many different cell populations and many of them are differentiated and do not have active telomerase. Thus, reducing PinX1 by either knockout or knockdown increases telomerase activity leading to telomere elongation in cells and mice, as shown in cancer cells or yeasts, respectively (45, 47).

Figure 4

PinX1 heterozygous knockout or knockdown leads to anaphase bridges, lagging chromosomes, and chromosome instability in MEFs. (A–C) PinX1 knockout leads to anaphase bridges and chromosome instability at late passage. MEFs derived from PinX1^+/+ and PinX1^+/– littermates were continuously cultured using the 3T3 protocol and fixed at early (from 3 to 5) or late (~30) passage, followed by staining with DAPI to score for the frequency of anaphase bridges and/or lagging chromosomes (A and B) and by counting the number of chromosomes at metaphase chromosome spreads to score for aneuploidy (C). (D–G) PinX1 knockdown leads to telomerase activation, telomere elongation, anaphase bridges, and chromosome instability at late passage. SV40 immortalized MEFs were stably infected with PinX1 shRNA or control viruses, followed by RT-PCR analysis (D) and TRAP assay using various amounts of telomerase-containing fractions (ng) (E) and semiquantified (Supplemental Figure 2C). Cells were continuously cultured, followed by analyzing telomere length using qFISH (F), anaphase bridges and/or lagging chromosomes at late (~20) passage (G). (H–J) Aneuploidy (H) and chromosome instability in PinX1^+/– MEFs (I and J) at late passages. Insets show examples of chromosome translocations (left to right), the display color, the inverted DAPI counterstain, and the classification pseudocolor. Arrows in I and J point to chromosome fusions, while red arrows in insets point to centrosomes. Numbers indicate the chromosomal origin of each fragment. Insets in I show 3 examples of translocations between short arms of 2 different chromosomes, with each containing a centromere, while inset in J shows an example of a translocation between long arms of 2 different chromosomes, with 1 containing a centromere.

PinX1 heterozygous knockout or knockdown leads to anaphase bridges and chromosome instability in MEFs. Given that PinX1 heterozygous knockout in MEFs activates telomerase activity, leading to telomere elongation, we wondered whether these phenotypes are associated with any cell-cycle defects. To examine this possibility, we compared PinX1^+/+ and PinX1^+/– primary MEFs at early and late passages and observed a strikingly notable phenotype in PinX1^+/– MEFs only at late passage; approximately 16% of these cells displayed prominent anaphase bridges and/or lagging chromosomes (Figure 4, A and B). In contrast, these phenotypes were very rarely found in PinX1^+/+ MEFs at the same late passage or PinX1^+/– MEFs and PinX1^+/+ MEFs at early passage (Figure 4, A and B). Consistent with such obvious abnormal chromosome separation, approximately 80% of PinX1^+/– MEFs at late passage also displayed prominent aneuploidy, having more or less than the normal 40 chromosomes (Figure 4C). The frequency of aneuploidy was approximately 10-fold higher than that in PinX1^+/– MEFs at early passage or in PinX1^+/+ MEFs at early and late passages (Figure 4C). Thus PinX1 heterozygous knockout leads to abnormal chromosome separation at late passage.

To independently confirm that reducing PinX1 function affects chromosome separation during mitosis, we examined the effects of PinX1 knockdown in MEFs that have been immortalized by SV40. After being infected with lentiviruses expressing PinX1 shRNA or control viruses, stable cell pools were selected. PinX1 shRNA effectively knocked down PinX1 mRNA in stable pools (Figure 4D). Significantly, PinX1 knockdown increased telomerase activity by approximately 3-fold at various concentrations examined (Figure 4E and Supplemental Figure 2C). Consistent with telomerase activation, telomeres were elongated after 20 passages, with the average TFUs being increased by approximately 98%, as compared with vector control (Figure 4F). The findings that PinX1 shRNA was more potent in activating telomerase and elongating telomeres (Figure 4, E and F) than PinX1 heterozygous knockout (Figure 2, G and H, and Figure 3F) are consistent with the fact that PinX1 shRNA was more effective in reducing PinX1 expression (Figure 4D and Figure 2D). More importantly, although these SV40 immortalized MEFs had a slightly higher rate of anaphase bridges than primary MEFs (Figure 4, B and G), as expected, PinX1 knockdown also significantly increased anaphase bridges at 20 passages, with about 33% of cells displaying anaphase bridges, 4- to 5-fold higher than vector controls (Figure 4G). These results together indicate that reducing PinX1 levels by heterozygous knockout or knockdown leads to abnormal chromosome separation.

To further confirm abnormal separation of mitotic chromosome in PinX1^+/– MEFs, we performed multicolor FISH (M-FISH) analysis because this procedure allows us to identify all individual chromosomes and their overall structures using a distinctly identifiable color spectrum. The number of a given chromosome varied from 0 to 4 depending on the individual cells and specific chromosomes (Figure 4H), further confirming aneuploidy (Figure 4C). Importantly, chromosome translocations were readily detected in PinX1^+/– MEFs at late passage (Figure 4, I and J). For example, chromosomes were translocated between the short arms of 2 different chromosomes, with each containing a centromere (Figure 4I), or between the long arms of 2 different chromosomes, with 1 containing a centromere (Figure 4J). Importantly, such chromosome abnormalities were rarely found in PinX1^+/+ MEF at the same passage or PinX1^+/– MEFs at early passage (data not shown), as expected from a very low frequency of anaphase bridges and aneuploidy in these cells (Figure 4, B and C). Thus, PinX1 heterozygous knockout not only increases telomerase activation, but also leads to anaphase bridges and chromosome instability in MEFs.

TERT knockdown or knockout rescues telomerase activation and telomere elongation and also abrogates anaphase bridges and chromosome instability in PinX1^+/– MEFs. The above findings indicate that inhibiting endogenous PinX1 through gene knockout or knockdown in cells increases telomerase activity, leading to telomere elongation, anaphase bridges, and chromosome instability. Given that PinX1 is a telomerase inhibitor (45, 47–49), a key question is whether telomerase is essential for PinX1 ablation to affect telomere length and chromosome stability.

To address this question, we first knocked down TERT in PinX1^+/– MEFs to examine its effects on telomere length and chromosome instability. PinX1^+/+ and PinX1^+/– MEFs were stably infected with lentiviruses expressing 2 different TERT-shRNA constructs at passage 3, before any chromosome instability phenotypes were observed. Both TERT-shRNA constructs effectively knocked down TERT mRNA (Figure 5A) and telomerase activity (Figure 5B and Supplemental Figure 2, D and E) in PinX1^+/+ and PinX1^+/– MEFs. To examine the effects of telomerase knockdown on telomere length and chromosome instability, we continuously cultured stable cell pools in vitro for approximately 20 passages. Telomerase knockdown in PinX1^+/+ MEFs did not have any detectable effects on telomere length, anaphase bridges, aneuploidy, or overall DNA content profile within the 20 passages examined (Figure 5C, Supplemental Figure 3, and data not shown). Importantly, although viral infection itself did not have any obvious effects on PinX1^+/+ or PinX1^+/– MEFs (Figure 5, C and D, and Supplemental Figure 3), both TERT shRNAs almost fully suppressed telomere elongation (Figure 5, E and F), anaphase bridges (Figure 5G), and aneuploidy (Figure 5H) in PinX1^+/– cells. Moreover, TERT-silenced PinX1^+/– cells had DNA content similar to that of control diploid PinX1^+/+ MEFs (Figure 5, K and L vs. I), whereas the DNA content in control PinX1^+/– cells lay between those of diploid and tetraploid cells (Figure 5, J vs. I), which is consistent with M-FISH analysis (Figure 4, H–J). These results show that TERT knockdown almost completely abrogates the ability of PinX1 knockout to induce telomerase activation, telomere elongation, and chromosome instability in MEFs.

Figure 5

TERT knockdown or knockout rescues telomerase activation and telomere elongation and abrogates anaphase bridges and chromosome instability in PinX1^+/– MEFs. (A and B) TERT knockdown. PinX1^+/+ and PinX1^+/– MEFs were stably infected with 2 different TERT shRNA or control lentiviruses, followed by quantitative TERT RT-PCR analysis with actin as a control (A) and TRAP assay (B, Supplemental Figure 2D, and E). Data are represented as mean ± SD. (C–F) TERT knockdown rescues telomere lengthening in PinX1^+/– MEFs at late passage. Stable cells were continuously cultured for 20 passages and then subjected to telomere qFISH. TERT knockdown had no effects on telomere length in PinX1^+/+ MEFs under the same conditions (Supplemental Figure 3). (G and H) TERT knockdown rescues anaphase bridges, lagging chromosomes, and aneuploidy in PinX1^+/– MEFs. PinX1^+/+ and PinX1^+/– MEFs expressing TERT shRNAs or control vector were fixed at 20 passages, followed by scoring for the frequency of anaphase bridges and/or lagging chromosomes (G) or aneuploidy (H). (I–L) TERT knockdown rescues abnormal DNA content in PinX1^+/– MEFs. PinX1^+/+ and PinX1^+/– MEFs expressing TERT shRNAs or control were fixed at 20 passages, followed by flow cytometry to determine the DNA content. Green and red arrows point to the DNA contents expected for diploid and tetraploid cells, respectively. (M) Generation of TERT and PinX1 double-knockout MEFs. 2 MEF lines of each group derived from Tert^–/– and PinX1^+/+ or PinX1^+/– embryos at E12.5 were subjected to PCR genotyping and confirmed by RT-PCR. (N and O) TERT knockout prevents telomere elongation in PinX1^+/– MEFs at late passage, as determined by qFISH. (P) TERT knockout prevents anaphase bridges and/or lagging chromosomes in PinX1^+/– MEFs at late passage.

Since shRNA constructs could sometimes have off-target effects, we needed to use an independent method to examine the role of telomerase in mediating the PinX1 phenotypes. For this purpose, we crossed PinX1^+/– mice and Tert^–/– mice to isolate Tert^–/– MEFs in the presence or absence of PinX1 heterozygous knockout (Figure 5M), followed by measuring telomere length and anaphase bridges at early (from 3 to 5) and late (from 25 to 30) passages. PinX1 heterozygous knockout failed to induce any increase in telomere length (Figure 5, N and O) and anaphase bridges and/or lagging chromosomes (Figure 5P) in Tert^–/– MEFs even at late passage. Similarly, PinX1 knockdown failed to induce any obvious increase in anaphase bridges and/or lagging chromosomes in G1 Terc^–/– MEFs (Supplemental Figure 5). These results indicate that telomerase is essential for PinX1 to affect telomere length and chromosome instability and demonstrate that reducing PinX1 activates telomerase and leads to telomerase-dependent telomere elongation and chromosome instability.

Most PinX1 heterozygous knockout mice develop malignant tumors of varied histopathology, which are unusual in mice, but are known to have LOH at 8p23 in humans. The above results show that PinX1 is reduced in most human breast cancer tissues and cells and that reducing PinX1 increases telomerase activity, leading to telomere elongation and chromosome instability in cells. Although further experiments are needed to elucidate whether chromosome instability is related to telomere-dependent, and/or -independent functions of telomerase or PinX1, we have focused our effort on determining whether reducing PinX1 levels leads to cancer development. This question is important because PinX1 is located at 8p23 in humans, a frequent LOH region in many epithelial cancers, and depleting PinX1 increases tumorigenicity of cancer cells (8, 9, 45, 50). Moreover, it might offer new insights into the development and treatment of human cancers.

To address this question, we first monitored a group totaling 52 PinX1^+/– and 15 PinX1^+/+ mice generated from intercrosses of young PinX1^+/– mice for tumor phenotypes (Supplemental Figure 6 and Supplemental Table 1). Whereas only 1 of 15 PinX1^+/+ mice developed a liver tumor, strikingly, 49 out of 52 PinX1^+/– mice developed tumors at relatively young ages, mostly between 9 and 18 months (Table 2, Supplemental Figure 6, and Supplemental Table 1). Most tumors were epithelial carcinomas arising in organs that are known to develop common cancers in humans and also known to have frequent heterozygous loss at 8p23 in humans (Table 2). The most common tumors were lung cancer (37% of all tumors), followed by liver cancer (18%), mammary cancer (13%), and then tumors of the gastrointestinal tract (10%), including stomach, small intestine, colon, and rectum (Table 2 and Supplemental Table 1). Tumors were found in both sexes of PinX1^+/– mice that were generated from 2 independent clones of targeted ES cells (Supplemental Table 1). Most tumors showed features commonly seen in advanced human carcinomas such as nuclear atypia, desmoplasia, stromal invasion, and/or distant metastasis (Figure 6).

Figure 6

Varied histopathology of PinX1^+/– tumors. (A–J) 2 different tumors (A–H) or metastatic tumors (I and J) in the same PinX1^+/– animals. (K–M) Different growth patterns of mammary adenocarcinoma in different PinX1^+/– mice. (N–P) Different growth patterns and tumor grades of mammary adenocarcinoma in the same PinX1^+/– mammary glands. (N) Acinar and tubular, high grade; (O) acinar and cyst papillary, intermediate grade; (P) tubular and scirrhous, low grade. (Q–T) Different tumor grades next to each other in the same section of lung cancer (Q). (R) Highly undifferentiated and aggressive (high grade) adenocarcinoma cells with huge nuclei; (S) intermediately differentiated and aggressive (intermediate grade) adenocarcinoma cells with intermediate nuclei; (T) well-differentiated and less aggressive (low grade) adenocarcinoma cells with small nuclei. Red arrows point to different sizes of nuclei; blue arrows point to mitotic cancer cells. Original magnification, ×40 (A–Q); ×60 (R–T).

Table 2

PinX1 knockout mice develop a range of epithelial cancers

Notably, about one-fifth of PinX1^+/– mice developed more than 1 type of tumor, with 3 and 7 mice having 3 and 2 tumor types in the same animals, respectively (Supplemental Table 1 and Figure 6, A–J). Furthermore, there were diverse histopathologies observed in the same types of tumors among different mice or even in the same tumors (Figure 6). For example, in mammary adenocarcinomas, there were many different growth patterns (Figure 6, K–P). Furthermore, even in the same mice, mammary cancer cells displayed different differentiation states and tumor grades, including high (Figure 6N), intermediate (Figure 6O), and low (Figure 6P) grades of cancer cells. Similarly, even in the same lung cancer section, cancer cells with high (Figure 6R), intermediate (Figure 6S), and low (Figure 6T) tumor grades were located right next to each other. Finally, PinX1^+/– tumors developed lung metastasis from lung cancer (Figure 6I) or mammary cancer (Figure 6J). These results together indicate that almost all PinX1^+/– mice spontaneously develop aggressive epithelial tumors that are unusual in mice, but are common in humans, demonstrating strong tumor-suppressing function for PinX1.

Since telomerase knockout (33) or telomere deprotection (34) shifts the tumor spectrum of p53 mutant mice from mainly lymphoma to include epithelial carcinomas, similarly to TERT overexpression (31), we crossed PinX1^+/– mice with p53 knockout mice and examined the effects of PinX1 heterozygous knockout on the tumor spectrum of p53 knockout mice. As expected, p53 knockout mice mainly developed lymphoma (Table 3). However, in the presence of PinX1 heterozygous knockout, the frequency of epithelial carcinomas such as lung, liver, breast, and gastrointestinal cancers was dramatically increased both in p53 heterozygous and homozygous knockout mice, with most p53^–/– and PinX1^+/– mice dying of cancer between 6–9 months and most p53^+/– and PinX1^+/– mice between 9–12 months (Table 3), when the majority of PinX1^+/– mice just started to develop tumors (Supplemental Figure 6). These results indicate that PinX1 heterozygous knockout shifts the tumor spectrum of p53 mutant mice toward epithelial carcinomas.

Table 3

PinX1 knockout shifts the tumor spectrum of p53 mutant mice toward epithelial carcinomas

PinX1 heterozygous knockout cancer cells express reduced PinX1 and display obvious telomere elongation, anaphase bridges, and chromosome instability. Given that PinX1 heterozygous knockout reduces PinX1 expression and activates telomerase, leading to aggressive epithelial malignancies in mice, we asked whether PinX1^+/– cancer cells have any PinX1 protein and any evidence of telomere elongation and chromosome instability.

To determine whether PinX1^+/– cancer cells express any PinX1 protein, we first analyzed PinX1 protein and mRNA levels in a large number of tumor samples and surrounding noncancerous samples of liver, lung, and mammary tissues in PinX1^+/– mice. Compared with levels in normal tissues of PinX1^+/+ mice, levels of PinX1 mRNA and protein were reduced to a similar extent (>50%) in cancer tissues and the surrounding noncancerous tissues in the PinX1^+/– mice (Figure 7, A–D, and data not shown), further supporting PinX1 haploinsufficiency.

Figure 7

PinX1^+/– cancer cells express reduced PinX1 and display telomere elongation, anaphase bridges, lagging chromosomes, and chromosome instability. (A–D) Reduced PinX1 expression in PinX1^+/– tumors (T) at levels similar to those in the surrounding noncancerous tissues (N) of liver (A and B), lung (C), or mammary gland (D), or in PinX1^+/+ normal control (C), as assayed by immunoblotting (A) or by qRT-PCR (B–D), with actin as a control. (E and F) Anaphase bridges and/or lagging chromosomes in PinX1^+/– lung and mammary tumors. H&E-stained series sections of lung (E) or mammary tumors (F) were examined for anaphase bridges (yellow arrows) and/or lagging chromosomes (red arrows). Original magnification, ×100. (G and H) Telomere elongation in PinX1^+/– lung cancer cells and mammary cancer cells. PinX1^+/– primary lung (G) or mammary (H) cancer cells as well as their respective PinX1^+/+ normal cells were subjected to qFISH. (I and J) Chromosome (Chr.) translocation in PinX1^+/– tumors. Genomic DNAs from PinX1^+/– tumors and PinX1^+/+ normal tissues were subjected to aCGH analysis. Each sample was repeated twice by performing dye-swap experiments to make sure that red and blue signals showed reproducible mirror images. Aberration calls identified by the ADM-1 algorithm are shown. There were widespread chromosomal imbalances with regions of gain or loss, and altering gene copy number, with a 1 representative shown (Supplemental Figure 7). The concomitant gains of 6qA1-A3.1 and 8qE2-8qD1 and remaining chromosome 8 (I) or of 4qE1-18qA1 with breakpoints outside of these regions (J) are consistent with chromosome fusions between 6q and 8q or 4q and 18q.

To detect chromosome instability phenotypes in PinX1^+/– tumors, we first looked for anaphase bridges in lung and mammary tumors after staining their series sections with H&E. Anaphase bridges and/or lagging chromosomes were readily found in both tumors (Figure 7, E and F), with up to 10% of mitotic cells, depending on the angle of sectioning, suggesting that PinX1 mutant tumors might have chromosome instability. To further examine this possibility, we established primary cell cultures from PinX1^+/– lung and mammary tumors and their respective normal controls using the procedure we described earlier (59). Both types of cancer cells displayed telomere elongation, with the average TFU increased by approximately 2-fold (Figure 7, G and H), but also anaphase bridges and/or lagging chromosomes in approximately 30% of cells (data not shown). These results indicate that PinX1^+/– tumors displayed prominent anaphase bridges and/or lagging chromosomes, suggesting chromosome instability.

To obtain further evidence of chromosome instability in PinX1^+/– tumors, we performed genome-wide oligonucleotide-based array comparative genome hybridization (aCGH) analysis of 6 PinX1^+/– tumor samples and WT controls. Each sample was repeated twice by performing dye-swap experiments to ensure that red and blue signals showed reproducible mirror images, as expected for true chromosome imbalance, as shown previously (60). We observed widespread chromosomal imbalances with regions of gain or loss (Supplemental Figure 7), which were different among the samples analyzed. These results are consistent with the diverse chromosome instability in PinX1^+/– MEFs (Figure 4, H–J) and also with the diverse histopathologies of PinX1^+/– tumors (Figure 6). Importantly, mapping breakpoints with concomitant gain or loss among different chromosomes revealed many possible chromosome fusions. For example, the concomitant gains of 6qA1–A3.1 and 8qE2–8qD1 and remaining chromosome 8 (Figure 7I) or of 4qE1–18qA1 with breakpoints outside of these regions (Figure 7J) are consistent with chromosome fusions between 6q and 8q or between 4q and 18q, similar to chromosome fusions obtained from M-FISH analysis of PinX1^+/– MEFs (Figure 4, H–J). Thus, PinX1^+/– cancer cells display prominent anaphase bridges and chromosome instability, as seen in PinX1^+/– MEFs, indicating that telomerase activity and chromosome instability likely contribute to tumorigenesis in PinX1 mutant mice. These findings together indicate that PinX1 heterozygous knockout increases telomerase activity and induces chromosome instability, eventually leading to the development of aggressive epithelial cancers in most mice.

PinX1 is a conserved potent telomerase inhibitor, but its physiological or pathological function is largely unknown. We here show that PinX1 expression is reduced in most human breast cancer tissues and cell lines examined. Importantly, reducing PinX1 expression via heterozygous knockout or knockdown increases telomerase activity and leads to concomitant telomerase-dependent telomere elongation and chromosomal instability. Moreover, PinX1 heterozygous knockout causes most mice to spontaneously develop a range of malignant tumors displaying evidence of telomere elongation and chromosomal instability. Notably, the majority of cancers in these mutant mice are carcinomas and share tissues of origin with human cancer types linked to 8p23 alterations. In addition, PinX1 knockout also shifts the tumor spectrum of p53 mutant mice toward epithelial carcinomas. These results are, to our knowledge, the first demonstration that PinX1 is a major haploinsufficient tumor suppressor and also provide what we believe is the first genetic evidence linking aberrant telomerase activation to chromosome instability, at least under the conditions where its endogenous inhibitor PinX1 is reduced. These findings not only uncover what we believe is a novel role of PinX1 and telomerase in chromosome instability and cancer initiation, but also suggest that telomerase inhibitors may be potentially effective in treating PinX1-related cancers.

PinX1 is a major haploinsufficient tumor suppressor at human chromosome 8p23. Human PinX1 gene is located at 8p23 near the marker D8S277 (8, 45). This region has been widely investigated because it is a frequent LOH region in common human adult epithelial cancers (4–14), but major tumor suppressor or suppressors at this region remain to be identified. Our new results indicate that PinX1 is a strong candidate for such a major tumor suppressor. We show that PinX1 is reduced in most human breast cancer tissues and cell lines examined (Figure 1) and that PinX1 expression is gene-dosage dependent; ablation of 1 allele reduces protein level by 60%–70% in vitro and in vivo (Figure 2). Furthermore, PinX1 heterozygous knockout or knockdown not only increases telomerase activity (Figures 2 and 4E and Supplemental Figure 2, A–C) and telomere length (Figures 3 and 4F), but also leads to chromosome instability in cells (Figure 4). Furthermore, the resulting chromosome instability is dependent on aberrant telomerase activation (Figure 5). The significance of these findings is further substantiated by the demonstration that nearly all PinX1 heterozygous knockout mice develop a range of epithelial malignancies (Figure 6 and Table 2) and that PinX1 mutant cancer cells express reduced PinX1 and display evidence of chromosome instability (Figure 7).

Although most human cancers are epithelial carcinomas, common tumor suppressor mutant mice mainly develop lymphomas and soft tissue sarcomas, with a few exceptions (61, 62). Notably, PinX1^+/– tumors are among the most common tumors and are also known to have frequent LOH at 8p23 in humans, including lung, breast, liver, and gastrointestinal cancers (Table 2). Moreover, many mutant mice developed 2 or even 3 different cancer types and metastases, and even within a given cancer type, histopathology varied among individuals and between and within tumors (Figure 6). This suggests that PinX1^+/– cancers likely originate from multiple cells and behave aggressively. These results are consistent with the findings that PinX1 heterozygous knockout MEFs (Figure 4) and cancer cells (Figure 7) display widespread chromosome instability. Finally, PinX1 heterozygous knockout also shifts the tumor spectrum of p53 mutant mice from lymphoma toward epithelial carcinomas (Table 3), similar to telomerase overexpression (31), telomerase knockout (33), or telomere deprotection (34). Notably, anaphase bridges, chromosome instability, and aneuploidy are very common in human cancers, especially epithelial tumors (63, 64). In addition, although telomere loss in cancers has been well studied, telomere elongation is quite common in human cancers, e.g., in more than 40% of liver cancer (65), esophageal cancer (66), and brain tumors (67), and also correlates with advanced stages and/or poor survival in some cancers (65, 66, 68, 69). Together with previous findings that reducing PinX1 potently increases tumorigenicity of human cancer cells (45) and that PinX1 is reduced in the majority of liver and gastric cancers (8, 9, 50), these results indicate that PinX1 is a major haploinsufficient tumor suppressor whose reduction may contribute to tumorigenesis by activating telomerase and promoting chromosome instability. Given telomerase activation in most human cancers, loss of PinX1 function likely contributes to the pathogenesis of many cancers and might have important therapeutic implications.

PinX1 is essential for inhibiting chromosome instability during tumorigenesis. The ability of PinX1 to regulate telomerase is highly conserved from humans to yeasts (45–49). However, it is not known whether PinX1 is a rate-limiting physiological regulator of telomerase and why such a negative regulatory mechanism is needed. We have now addressed these questions using PinX1 knockout or knockdown in vitro and in vivo. Reduced PinX1 expression by gene deletion or knockdown in MEFs increases telomerase activity, leading to telomere elongation, anaphase bridges, lagging chromosomes, and chromosome instability (Figures 2–4). Moreover, TERT knockdown or knockout fully prevents PinX1 deletion from not only activating telomerase activation and elongating telomeres, but also from inducing anaphase bridges and chromosome instability (Figure 5). Importantly, nearly all PinX1^+/– mice develop aggressive epithelial cancers that display elongated telomeres, anaphase bridges, lagging chromosomes, and chromosomal instability (Figures 6 and 7; Table 2). These results provide evidence for an essential role of PinX1 in inhibiting telomerase activation and chromosome instability during tumorigenesis in vitro and in vivo. This is consistent with the critical role of PinX1 in linking between TRF1 and telomerase inhibition to prevent telomere elongation and help maintain telomere homeostasis (46).

Telomerase is essential for PinX1 reduction to induce chromosome instability. At the present time, we do not yet know how reducing PinX1 function leads to chromosome instability. Since PinX1 localizes to nucleoli in addition to telomeres (45, 47) and affects some ribosomal RNA maturation in yeast (70), it could be possible, but less unlikely, that PinX1 regulates chromosome stability via affecting some nucleolar function. Recently, Yuan et al. reported that exogenous PinX1 binds to microtubules and localizes to kinetochores at mitosis and its almost complete knockdown in transient experiments causes chromosome mis-separation and micronuclei formation (71). At the first glance, these PinX1 functions could explain chromosome instability in our PinX1-reduced cells. However, we observed abnormal cell division only when PinX1 was almost completely knocked down or both PinX1 alleles were deleted by infecting PinX1^fl/fl MEFs with CMV-Cre, and these cells died in a few days (data not shown). In contrast, PinX1 heterozygous knockout or stable knockdown cells grew normally for an extended period before displaying anaphase bridges and chromosome instability. Therefore, PinX1 action on the basic cellular structures might explain the essential function of PinX1 for cell survival in MEFs and mice, and could contribute to chromosome instability, but cannot fully explain the phenotypes that we observed here due to partial PinX1 reduction.

Notably, the PinX1 knockout phenotypes including telomerase activation, telomere elongation, anaphase bridges, aneuploidy, and chromosome instability are fully suppressed by knockdown or knockout of TERT (Figure 5) or TERC (Supplemental Figure 5), indicating that telomerase is essential for PinX1 reduction to induce chromosome instability. Moreover, it takes time for PinX1-induced telomerase activation to induce telomere elongation (Figure 3) and chromosome instability (Figure 4) when PinX1 is knocked out or down. Notably, our PinX1 and p53 double-mutant mice have a tumor spectrum similar to that found in TERC and p53 double-mutant mice due to telomere loss (33) or in TPP1/ACD and p53 double-mutant mice due to telomere deprotection (34). Furthermore, abnormal telomere elongation is common and also correlates with advanced stages and/or poor survival in some cancers (65, 66, 68, 69). Moreover, TERC is required for the tumor-promoting effects of TERT overexpression in transgenic mice (32). These results together suggest that abnormal telomerase activation and telomere elongation due to loss of PinX1 might have effects on the development of epithelial cancers similar to those of telomere shortening or telomere deprotection. Given that PinX1 directly binds to and inhibits TERT (45) and is targeted by TRF1 to telomeres to prevent abnormal telomere elongation by telomerase (46), it is conceivable that when PinX1 is inhibited, telomerase is aberrantly activated without a proper brake and eventually leads to chromosome instability, possibly via inducing aberrant telomere elongation to compromise telomere function. However, telomerase has other telomere-independent functions, such as is found in DNA damage response (35, 36) and activation of β-catenin (37). In addition, PinX1 might have nontelomeric functions such as in RNA maturation (70) and cell division (71). Therefore, further experiments are needed to define how PinX1 controls chromosome stability via telomere-dependent and/or -independent telomerase and/or other mechanisms unrelated to telomerase. Nevertheless, we have further demonstrated a major role of PinX1-induced chromosome instability in tumorigenesis.

We show that PinX1 is reduced in most human breast cancer tissues and cells (Figure 1) and that reducing PinX1 levels leads to telomerase activation, telomere elongation, and chromosome instability (Figures 2–4). Moreover, almost all PinX1 heterozygous knockout mice develop a range of epithelial malignancies, with multiple tumor types in the same mice, diverse cell morphologies/grades in 1 tumor type among mice, or even within individual tumors (Figures 6, and 7; Table 2). PinX1 heterozygous knockout also shifts the p53 mutant tumor spectrum to epithelial carcinomas (Table 3). Importantly, PinX1^+/– cancer cells also display chromosome instability (Figure 7) similar to that in PinX1-inhibited cells (Figure 4). Thus, PinX1 mutant tumors are likely derived from multiple epithelial cells, presumably due to chromosome instability. These results are consistent with previous findings that PinX1 potently controls tumorigenicity of cancer cells (45) and that telomerase overexpression is not as oncogenic as that in PinX1 knockout shown here. Analogous situations have been documented previously. For example, CDK subunit cyclin D1 overexpression in mice (72) is much less potent than CDK inhibitor p16 knockout in inducing tumorigenesis (73). Finally, our results also suggest that telomerase inhibition might be used to treat PinX1-related tumors.

In summary, we find that PinX1 is often reduced in human breast cancer, that reducing PinX1 leads to chromosome instability in a telomerase-dependent manner, and that PinX1 heterozygous knockout causes all mice to develop a range of malignancies with evidence of chromosome instability. Given that PinX1 is located at frequent LOH regions in many common human cancers and its expression is actually reduced in many breast, liver, and gastrointestinal tumors, these results indicate that low PinX1 can contribute to tumorigenesis by activating telomerase and inducing chromosome instability and suggest novel options for cancer treatments.

View Supplemental data

We are grateful to L. Cantley, P.P. Pandolfi, B. Neel, and T. Hunter for constructive discussions; W. Hahn for many critical reagents and much advice; S. Chang for G1 Terc^–/– MEFs; E. Li and W. Yang for advice on generating PinX1-knockout mice; C. Lee’s Cytogenetics Core for mFISH and CGH analyses; S. Hagen for Confocal Microscopy Core (NIH grant S10 RR017927); and the members of the Lu and Zhou laboratories for help or advice, especially J. Driver for editing the manuscript. The work was supported by NIH grant RO1CA122434 and Susan G. Komen for the Cure grant KG100958 to X.Z. Zhou.

Address correspondence to: Xiao Zhen Zhou or Kun Ping Lu, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, CLS 0408, Boston, Massachusetts 02215, USA. Phone: 617.735.2017 (X.Z. Zhou), 617.735.2016 (K.P. Lu); Fax: 617.735.2050; E-mail: xzhou@bidmc.harvard.edu (X.Z. Zhou), klu@bidmc.harvard.edu (K.P. Lu).

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J Clin Invest. 2011;121(4):1266–1282. doi:10.1172/JCI43452.

See the related article at PinX1 the tail on the chromosome.

1. Foulkes WD. Inherited susceptibility to common cancers. N Engl J Med. 2008;359(20):2143–2153.
   View this article via: PubMed CrossRef Google Scholar
2. Nathanson KL, Wooster R, Weber BL. Breast cancer genetics: what we know and what we need. Nat Med. 2001;7(5):552–556.
   View this article via: PubMed CrossRef Google Scholar
3. Hartman AR, Ford JM. BRCA1 and p53: compensatory roles in DNA repair. J Mol Med. 2003;81(11):700–707.
   View this article via: PubMed Google Scholar
4. Emi M, et al. Frequent loss of heterozygosity for loci on chromosome 8p in hepatocellular carcinoma, colorectal cancer, and lung cancer. Cancer Res. 1992;52(19):5368–5372.
   View this article via: PubMed Google Scholar
5. Becker SA, Zhou YZ, Slagle BL. Frequent loss of chromosome 8p in hepatitis B virus-positive hepatocellular carcinomas from China. Cancer Res. 1996;56(21):5092–5097.
   View this article via: PubMed Google Scholar
6. Nagai H, Pineau P, Tiollais P, Buendia MA, Dejean A. Comprehensive allelotyping of human hepatocellular carcinoma. Oncogene. 1997;14(24):2927–2933.
   View this article via: PubMed CrossRef Google Scholar
7. Pineau P, et al. Identification of three distinct regions of allelic deletions on the short arm of chromosome 8 in hepatocellular carcinoma. Oncogene. 1999;18(20):3127–3134.
   View this article via: PubMed CrossRef Google Scholar
8. Liao C, Zhao M, Song H, Uchida K, Yokoyama KK, Li T. Identification of the gene for a novel liver-related putative tumor suppressor at a high-frequency loss of heterozygosity region of chromosome 8p23 in human hepatocellular carcinoma. Hepatology. 2000;32(4 pt 1):721–727.
   View this article via: PubMed CrossRef Google Scholar
9. Kondo T, et al. Loss of heterozygosity and histone hypoacetylation of the PINX1 gene are associated with reduced expression in gastric carcinoma. Oncogene. 2005;24(1):157–164.
   View this article via: PubMed CrossRef Google Scholar
10. Kahng YS, Lee YS, Kim BK, Park WS, Lee JY, Kang CS. Loss of heterozygosity of chromosome 8p and 11p in the dysplastic nodule and hepatocellular carcinoma. J Gastroenterol Hepatol. 2003;18(4):430–436.
    View this article via: PubMed CrossRef Google Scholar
11. Yokota T, et al. Localization of a tumor suppressor gene associated with the progression of human breast carcinoma within a 1-cM interval of 8p22-p23.1. Cancer. 1999;85(2):447–452.
    View this article via: PubMed CrossRef Google Scholar
12. Loo LW, et al. Array comparative genomic hybridization analysis of genomic alterations in breast cancer subtypes. Cancer Res. 2004;64(23):8541–8549.
    View this article via: PubMed CrossRef Google Scholar
13. Bhattacharya N, et al. Three discrete areas within the chromosomal 8p21.3-23 region are associated with the development of breast carcinoma of Indian patients. Exp Mol Pathol. 2004;76(3):264–271.
    View this article via: PubMed CrossRef Google Scholar
14. Varma G, et al. Array comparative genomic hybridisation (aCGH) analysis of premenopausal breast cancers from a nuclear fallout area and matched cases from Western New York. Br J Cancer. 2005;93(6):699–708.
    View this article via: PubMed Google Scholar
15. Hamaguchi M, et al. DBC2, a candidate for a tumor suppressor gene involved in breast cancer. Proc Natl Acad Sci U S A. 2002;99(21):13647–13652.
    View this article via: PubMed CrossRef Google Scholar
16. Lai J, Flanagan J, Phillips WA, Chenevix-Trench G, Arnold J. Analysis of the candidate 8p21 tumour suppressor, BNIP3L, in breast and ovarian cancer. Br J Cancer. 2003;88(2):270–276.
    View this article via: PubMed CrossRef Google Scholar
17. Ishii H, et al. The FEZ1 gene at chromosome 8p22 encodes a leucine-zipper protein, and its expression is altered in multiple human tumors. Proc Natl Acad Sci U S A. 1999;96(7):3928–3933.
    View this article via: PubMed CrossRef Google Scholar
18. Kim NW, et al. Specific association of human telomerase activity with immortal cells and cancer. Science. 1994;266(5193):2011–2015.
    View this article via: PubMed CrossRef Google Scholar
19. Broccoli D, Young JW, de Lange T. Telomerase activity in normal and malignant hematopoietic cells. Proc Natl Acad Sci U S A. 1995;92(20):9082–9086.
    View this article via: PubMed CrossRef Google Scholar
20. Blackburn EH. Switching and signaling at the telomere. Cell. 2001;106(6):661–673.
    View this article via: PubMed CrossRef Google Scholar
21. Shay JW, Wright WE. Telomerase: a target for cancer therapeutics. Cancer Cell. 2002;2(4):257–265.
    View this article via: PubMed CrossRef Google Scholar
22. Maser RS, DePinho RA. Connecting chromosomes, crisis, and cancer. Science. 2002;297(5581):565–569.
    View this article via: PubMed CrossRef Google Scholar
23. Dong CK, Masutomi K, Hahn WC. Telomerase: regulation, function and transformation. Crit Rev Oncol Hematol. 2005;54(2):85–93.
    View this article via: PubMed CrossRef Google Scholar
24. Smogorzewska A, de Lange T. Regulation of telomerase by telomeric proteins. Annu Rev Biochem. 2004;73:177–208.
    View this article via: PubMed CrossRef Google Scholar
25. Blasco MA. Telomere length, stem cells and aging. Nat Chem Biol. 2007;3(10):640–649.
    View this article via: PubMed CrossRef Google Scholar
26. Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, Brooks MW, Weinberg RA. Creation of human tumour cells with defined genetic elements. Nature. 1999;400(6743):464–468.
    View this article via: PubMed CrossRef Google Scholar
27. Bodnar AG, et al. Extension of life-span by introduction of telomerase into normal human cells. Science. 1998;279(5349):349–352.
    View this article via: CrossRef PubMed Google Scholar
28. Vaziri H, Benchimol S. Reconstitution of telomerase activity in normal human cells leads to elongation of telomeres and extended replicative life span. Curr Biol. 1998;8(5):279–282.
    View this article via: CrossRef PubMed Google Scholar
29. Gonzalez-Suarez E, et al. Increased epidermal tumors and increased skin wound healing in transgenic mice overexpressing the catalytic subunit of telomerase, mTERT, in basal keratinocytes. EMBO J. 2001;20(11):2619–2630.
    View this article via: PubMed CrossRef Google Scholar
30. Artandi SE, et al. Constitutive telomerase expression promotes mammary carcinomas in aging mice. Proc Natl Acad Sci U S A. 2002;99(12):8191–8196.
    View this article via: PubMed CrossRef Google Scholar
31. Gonzalez-Suarez E, Flores JM, Blasco MA. Cooperation between p53 mutation and high telomerase transgenic expression in spontaneous cancer development. Mol Cell Biol. 2002;22(20):7291–7301.
    View this article via: PubMed CrossRef Google Scholar
32. Cayuela ML, Flores JM, Blasco MA. The telomerase RNA component Terc is required for the tumour-promoting effects of Tert overexpression. EMBO Rep. 2005;6(3):268–274.
    View this article via: PubMed CrossRef Google Scholar
33. Artandi SE, et al. Telomere dysfunction promotes non-reciprocal translocations and epithelial cancers in mice. Nature. 2000;406(6796):641–645.
    View this article via: PubMed CrossRef Google Scholar
34. Else T, et al. Genetic p53 deficiency partially rescues the adrenocortical dysplasia phenotype at the expense of increased tumorigenesis. Cancer Cell. 2009;15(6):465–476.
    View this article via: PubMed CrossRef Google Scholar
35. Masutomi K, et al. The telomerase reverse transcriptase regulates chromatin state and DNA damage responses. Proc Natl Acad Sci U S A. 2005;102(23):8222–8227.
    View this article via: PubMed Google Scholar
36. Sharma GG, et al. hTERT associates with human telomeres and enhances genomic stability and DNA repair. Oncogene. 2003;22(1):131–146.
    View this article via: PubMed CrossRef Google Scholar
37. Park JI, et al. Telomerase modulates Wnt signalling by association with target gene chromatin. Nature. 2009;460(7251):66–72.
    View this article via: PubMed CrossRef Google Scholar
38. Wang J, Xie LY, Allan S, Beach D, Hannon GJ. Myc activates telomerase. Genes Dev. 1998;12(12):1769–1774.
    View this article via: PubMed Google Scholar
39. Greenberg RA, et al. Telomerase reverse transcriptase gene is a direct target of c-Myc but is not functionally equivalent in cellular transformation. Oncogene. 1999;18(5):1219–1226.
    View this article via: PubMed CrossRef Google Scholar
40. Wu KJ, et al. Direct activation of TERT transcription by c-MYC. Nat Genet. 1999;21(2):220–224.
    View this article via: PubMed CrossRef Google Scholar
41. Lin SY, Elledge SJ. Multiple tumor suppressor pathways negatively regulate telomerase. Cell. 2003;113(7):881–889.
    View this article via: PubMed CrossRef Google Scholar
42. van Steensel B, de Lange T. Control of telomere length by the human telomeric protein Trf1. Nature. 1997;385(6618):740–743.
    View this article via: PubMed CrossRef Google Scholar
43. de Lange T. Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev. 2005;19(18):2100–2110.
    View this article via: PubMed CrossRef Google Scholar
44. Lee TH, et al. Essential role of Pin1 in the regulation of TRF1 stability and telomere maintenance. Nat Cell Biol. 2009;11(1):97–105.
    View this article via: PubMed CrossRef Google Scholar
45. Zhou XZ, Lu KP. The Pin2/TRF1-interacting protein PinX1 is a potent telomerase inhibitor. Cell. 2001;107(3):347–359.
    View this article via: PubMed CrossRef Google Scholar
46. Soohoo CY, Shi R, Lee TH, Huang P, Lu KP, Zhou XZ. Telomerase inhibitor PINX1 provides a link between TRF1 and telomerase to prevent telomere elongation. J Biol Chem. 2011;286(5):3894–3906.
    View this article via: PubMed CrossRef Google Scholar
47. Lin J, Blackburn EH. Nucleolar protein PinX1p regulates telomerase by sequestering its protein catalytic subunit in an inactive complex lacking telomerase RNA. Genes Dev. 2004;18(4):387–396.
    View this article via: PubMed CrossRef Google Scholar
48. Oh BK, Yoon SM, Lee CH, Park YN. Rat homolog of PinX1 is a nucleolar protein involved in the regulation of telomere length. Gene. 2007;400(1–2):35–43.
    View this article via: PubMed CrossRef Google Scholar
49. Sun C, Wu Z, Jia F, Wang Y, Li T, Zhao M. Identification of zebrafish LPTS: a gene with similarities to human LPTS/PinX1 that inhibits telomerase activity. Gene. 2008;420(1):90–98.
    View this article via: PubMed CrossRef Google Scholar
50. Park WS, et al. Genetic analysis of the liver putative tumor suppressor (LPTS) gene in hepatocellular carcinomas. Cancer Lett. 2002;178(2):199–207.
    View this article via: PubMed CrossRef Google Scholar
51. Oh BK, Chae KJ, Park C, Park YN. Molecular analysis of PinX1 in human hepatocellular carcinoma. Oncol Rep. 2004;12(4):861–866.
    View this article via: PubMed Google Scholar
52. Hawkins GA, et al. Mutational analysis of PINX1 in hereditary prostate cancer. Prostate. 2004;60(4):298–302.
    View this article via: PubMed Google Scholar
53. Lewandoski M, Martin GR. Cre-mediated chromosome loss in mice. Nat Genet. 1997;17(2):223–225.
    View this article via: PubMed Google Scholar
54. Karlseder J, et al. Targeted deletion reveals an essential function for the telomere length regulator Trf1. Mol Cell Biol. 2003;23(18):6533–6541.
    View this article via: PubMed CrossRef Google Scholar
55. Chiang YJ, Kim SH, Tessarollo L, Campisi J, Hodes RJ. Telomere-associated protein TIN2 is essential for early embryonic development through a telomerase-independent pathway. Mol Cell Biol. 2004;24(15):6631–6634.
    View this article via: PubMed CrossRef Google Scholar
56. Liou YC, et al. Loss of Pin1 function in the mouse causes phenotypes resembling cyclin D1-null phenotypes. Proc Natl Acad Sci U S A. 2002;99(3):1335–1340.
    View this article via: PubMed CrossRef Google Scholar
57. Takai H, Smogorzewska A, de Lange T. DNA damage foci at dysfunctional telomeres. Curr Biol. 2003;13(17):1549–1556.
    View this article via: PubMed CrossRef Google Scholar
58. Flores I, Canela A, Vera E, Tejera A, Cotsarelis G, Blasco MA. The longest telomeres: a general signature of adult stem cell compartments. Genes Dev. 2008;22(5):654–667.
    View this article via: CrossRef PubMed Google Scholar
59. Wulf G, Garg P, Liou YC, Iglehart D, Lu KP. Modeling breast cancer in vivo and ex vivo reveals an essential role of Pin1 in tumorigenesis. EMBO J. 2004;23(16):3397–3407.
    View this article via: PubMed CrossRef Google Scholar
60. Perry GH, et al. The fine-scale and complex architecture of human copy-number variation. Am J Hum Genet. 2008;82(3):685–695.
    View this article via: PubMed CrossRef Google Scholar
61. Wu X, Pandolfi PP. Mouse models for multistep tumorigenesis. Trends Cell Biol. 2001;11(11):S2–S9.
    View this article via: PubMed Google Scholar
62. Hakem R, Mak TW. Animal models of tumor-suppressor genes. Annu Rev Genet. 2001;35:209–241.
    View this article via: CrossRef PubMed Google Scholar
63. Rudolph KL, Millard M, Bosenberg MW, DePinho RA. Telomere dysfunction and evolution of intestinal carcinoma in mice and humans. Nat Genet. 2001;28(2):155–159.
    View this article via: PubMed CrossRef Google Scholar
64. Stewenius Y, et al. Defective chromosome segregation and telomere dysfunction in aggressive Wilms’ tumors. Clin Cancer Res. 2007;13(22 pt 1):6593–6602.
    View this article via: PubMed Google Scholar
65. Oh BK, et al. High telomerase activity and long telomeres in advanced hepatocellular carcinomas with poor prognosis. Lab Invest. 2008;88(2):144–152.
    View this article via: PubMed CrossRef Google Scholar
66. Gertler R, Doll D, Maak M, Feith M, Rosenberg R. Telomere length and telomerase subunits as diagnostic and prognostic biomarkers in Barrett carcinoma. Cancer. 2008;112(10):2173–2180.
    View this article via: PubMed CrossRef Google Scholar
67. Nurnberg P, Thiel G, Weber F, Epplen JT. Changes of telomere lengths in human intracranial tumours. Hum Genet. 1993;91(2):190–192.
    View this article via: PubMed Google Scholar
68. Bisoffi M, Heaphy CM, Griffith JK. Telomeres: prognostic markers for solid tumors. Int J Cancer. 2006;119(10):2255–2260.
    View this article via: PubMed CrossRef Google Scholar
69. Gertler R, et al. Telomere length and human telomerase reverse transcriptase expression as markers for progression and prognosis of colorectal carcinoma. J Clin Oncol. 2004;22(10):1807–1814.
    View this article via: PubMed CrossRef Google Scholar
70. Guglielmi B, Werner M. The yeast homolog of human PinX1 is involved in rRNA and small nucleolar RNA maturation, not in telomere elongation inhibition. J Biol Chem. 2002;277(38):35712–35719.
    View this article via: PubMed Google Scholar
71. Yuan K, et al. PinX1 is a novel microtubule-binding protein essential for accurate chromosome segregation. J Biol Chem. 2009;284(34):23072–23082.
    View this article via: PubMed CrossRef Google Scholar
72. Wang TC, Cardiff RD, Zukerberg L, Lees E, Arnold A, Schmidt EV. Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature. 1994;369(6482):669–671.
    View this article via: PubMed CrossRef Google Scholar
73. Sharpless NE, et al. Loss of p16Ink4a with retention of p19Arf predisposes mice to tumorigenesis. Nature. 2001;413(6851):86–91.
    View this article via: PubMed CrossRef Google Scholar
74. Wulf GM, Ryo A, Wulf GG, Lee SW, Niu T, Lu KP. Pin1 is overexpressed in breast cancer and potentiates the transcriptional activity of phosphorylated c-Jun towards the cyclin D1 gene. EMBO J. 2001;20(13):3459–3472.
    View this article via: PubMed CrossRef Google Scholar
75. Liu Y, et al. The telomerase reverse transcriptase is limiting and necessary for telomerase function in vivo. Curr Biol. 2000;10(22):1459–1462.
    View this article via: PubMed CrossRef Google Scholar