Abstract 1110: Colorectal liver metastases induce a systemic phase shift of the biological clock

Circadian clock genes drive cyclic expression of 10-15% of the transcriptome. Disruption of the circadian clock may deregulate the cell's normal functions and can lead to carcinogenesis, indicating that clock genes also act as tumor suppressor genes. In previous work, we have shown that circadian rhythmicity is disrupted in C26 colon carcinoma cell-derived hepatic metastases, as compared to the non-cancerous tissue of the liver. It has been suggested that neoplastic tissue can influence the circadian clock of adjacent tissues, causing a tissue specific phase shift. A better understanding of how tumors affect the circadian system may help elucidate the role of the clock in cancer patients suffering from fatigue and sleep disorders. Therefore, we have examined the circadian clock in liver and kidney tissue of mice carrying hepatic C26 colon carcinoma cell metastases.

Male BALB/c mice were divided into two groups. In group 1, hepatic metastases were induced by injecting C26 colorectal carcinoma cells into the spleen. Group 2 was sham operated. Three weeks after (sham) surgery, animals were sacrificed over a 24-hour period (6 time points, n=4 per time point) and liver and kidneys were collected to compare circadian rhythmicity in control and tumor bearing animals. RNA was isolated and qRT-PCR was performed to determine the expression levels of 5 clock genes (Rev-Erbα, Per1, Per2, Bmal1, Cry1) and 3 clock controlled genes (Dbp, p21 and Wee1).

Liver and kidney tissue of healthy control mice showed normal 24-hour oscillations of all clock genes, consistent with normal circadian rhythmicity. Liver and kidney tissue of tumor-bearing mice, as compared to control mice, showed a marked phase difference in clock rhythms. In the liver we observed a 4-hour phase advance for Bmal1, Rev-Erbα and Dbp. In contrast in the kidney, there was a 4-hour phase delay for Bmal1, Rev-Erbα, Cry1 and DBP, and a 8 hour delay for Per2.

These data show that colorectal liver metastases induce a phase advance in the liver, and a phase delay in the kidney . This suggests a systemic effect of the tumor on peripheral clocks. This difference in clock phase between tumor-bearing mice and healthy controls may provide new clues to further understand the role of the clock for fatigue and sleep problems in cancer patients, but may also give opportunities to administer chemotherapeutic treatment in a chronotherapeutical setting .

Citation Format: Sander A. Huisman, F. Tamanini, A. Ahmadi, J.N.M. IJzermans, G.T.J. van der Horst, R.W.F. de Bruin. Colorectal liver metastases induce a systemic phase shift of the biological clock. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1110. doi:10.1158/1538-7445.AM2014-1110

Structure function analysis of mammalian cryptochromes

Members of the photolyase/cryptochrome family are flavoproteins that share an extraordinary conserved core structure (photolyase homology region, PHR), but the presence of a carboxy-terminal extension is limited to the cryptochromes. Photolyases are DNA-repair enzymes that remove UV-light-induced lesions. Cryptochromes of plants and Drosophila act as circadian photoreceptors, involved in light entrainment of the biological clock. Using knockout mouse models, mammalian cryptochromes (mCRY1 and mCRY2) were identified as essential components of the clock machinery. Within the mammalian transcription-translation feedback loop generating rhythmic gene expression, mCRYs potently inhibit the transcription activity of the CLOCK/BMAL1 heterodimer and protect mPER2 from 26S-protesome-mediated degradation. By analyzing a set of mutant mCRY1 proteins and photolyase/mCRY1 chimeric proteins, we found that the carboxyl terminus has a determinant role in mCRY1 function by harboring distinguished domains involved in nuclear import and interactions with other clock proteins. Moreover, the carboxyl terminus must cross-talk with the PHR to establish full transcription repression capacity in mCRY1. We propose that the presence of the carboxyl terminus in cryptochromes, which varies in sequence composition among mammalian, Drosophila, and plant CRYs, is critical for their different functions and possibly contributed to shape the different architecture and biochemistry of the clock machineries in these organisms.

Molecular mechanisms of the biological clock in cultured fibroblasts

In mammals, the central circadian pacemaker resides in the hypothalamic suprachiasmatic nucleus (SCN), but circadian oscillators also exist in peripheral tissues. Here, using wild-type and cryptochrome (mCry)-deficient cell lines derived from mCry mutant mice, we show that the peripheral oscillator in cultured fibroblasts is identical to the oscillator in the SCN in (i) temporal expression profiles of all known clock genes, (ii) the phase of the various mRNA rhythms (i.e., antiphase oscillation of Bmal1 and mPer genes), (iii) the delay between maximum mRNA levels and appearance of nuclear mPER1 and mPER2 protein, (iv) the inability to produce oscillations in the absence of functional mCry genes, and (v) the control of period length by mCRY proteins.

The fragile X-related proteins FXR1P and FXR2P contain a functional nucleolar-targeting signal equivalent to the HIV-1 regulatory proteins

Fragile X syndrome is caused by the absence of the fragile X mental-retardation protein (FMRP). FMRP and the fragile X-related proteins 1 and 2 (FXR1P and FXR2P) form a gene family with functional similarities, such as RNA binding, polyribosomal association and nucleocytoplasmic shuttling. In a previous study, we found that FMRP and FXR1P shuttle between cytoplasm and nucleoplasm, while FXR2P shuttles between cytoplasm and nucleolus. The nuclear and nucleolar-targeting properties of these proteins were investigated further. Here, we show that FXR2P contains in its C-terminal part, a stretch of basic amino acids 'RPQRRNRSRRRRFR' that resemble the nucleolar-targeting signal (NoS) of the viral protein Rev. This particular sequence is also present within exon 15 of the FXR1 gene. This exon undergoes alternative splicing and is therefore only present in some of the FXR1P isoforms. We investigated the intracellular distribution of various FXR1P isoforms with (iso-e and iso-f) and without (iso-d) the potential NoS in transfected COS cells treated with the nuclear export inhibitor leptomycin-B. Both iso-e and iso-f showed a nucleolar localization, as observed for FXR2P; iso-d was detected in the nucleo-plasm outside the nucleoli. Further, when a labelled 16-residue synthetic peptide corresponding to the NoS of FXR1P was added to human fibroblast cultures a clear nucleolar signal was observed. Based on these data we argue that the intranuclear distribution of FXR2P and FXR1P isoforms is very likely to be mediated by a similar NoS localized in their C-terminal region. This domain is absent in some FXR1P isoforms as well as in all FMRP isoforms, suggesting functional differences for this family of proteins, possibly related to RNA metabolism in different tissues.

Dimerization and nuclear entry of mPER proteins in mammalian cells

Nuclear entry of circadian oscillatory gene products is a key step for the generation of a 24-hr cycle of the biological clock. We have examined nuclear import of clock proteins of the mammalian period gene family and the effect of serum shock, which induces a synchronous clock in cultured cells. Previously, mCRY1 and mCRY2 have been found to complex with PER proteins leading to nuclear import. Here we report that nuclear translocation of mPER1 and mPER2 (1) involves physical interactions with mPER3, (2) is accelerated by serum treatment, and (3) still occurs in mCry1/mCry2 double-deficient cells lacking a functional biological clock. Moreover, nuclear localization of endogenous mPER1 was observed in cultured mCry1/mCry2 double-deficient cells as well as in the liver and the suprachiasmatic nuclei (SCN) of mCry1/mCry2 double-mutant mice. This indicates that nuclear translocation of at least mPER1 also can occur under physiological conditions (i.e., in the intact mouse) in the absence of any CRY protein. The mPER3 amino acid sequence predicts the presence of a cytoplasmic localization domain (CLD) and a nuclear localization signal (NLS). Deletion analysis suggests that the interplay of the CLD and NLS proposed to regulate nuclear entry of PER in Drosophila is conserved in mammals, but with the novel twist that mPER3 can act as the dimerizing partner.

Oligomerization properties of fragile-X mental-retardation protein (FMRP) and the fragile-X-related proteins FXR1P and FXR2P

The absence of fragile-X mental-retardation protein (FMRP) results in fragile-X syndrome. Two other fragile-X-related (FXR) proteins have been described, FXR1P and FXR2P, which are both very similar in amino acid sequence to FMRP. Interaction between the three proteins as well as with themselves has been demonstrated. The FXR proteins are believed to play a role in RNA metabolism. To characterize a possible functional role of the interacting proteins the complex formation of the FXR proteins was studied in mammalian cells. Double immunofluorescence analysis in COS cells over-expressing either FMRP ISO12/FXR1P or FMRP ISO12/FXR2P confirmed heterotypic interactions. However, Western-blotting studies on cellular homogenates containing physiological amounts of the three proteins gave different indications. Gel-filtration experiments under physiological as well as EDTA conditions showed that the FXR proteins were in complexes of >600 kDa, as parts of messenger ribonuclear protein (mRNP) particles associated with polyribosomes. Salt treatment shifted FMRP, FXR1P and FXR2P into distinct intermediate complexes, with molecular masses between 200 and 300 kDa. Immunoprecipitations of FMRP as well as FXR1P from the dissociated complexes revealed that the vast majority of the FXR proteins do not form heteromeric complexes. Further analysis by [(35)S]methionine labelling in vivo followed by immunoprecipitation indicated that no proteins other than the FXR proteins were present in these complexes. These results suggest that the FXR proteins form homo-multimers preferentially under physiological conditions in mammalian cells, and might participate in mRNP particles with separate functions.

Different targets for the fragile X-related proteins revealed by their distinct nuclear localizations

Fragile X syndrome is caused by the absence of the fragile X mental retardation protein (FMRP). FMRP and its structural homologues FXR1P and FXR2P form a family of RNA-binding proteins (FXR proteins). The three proteins associate with polyribosomes as cytoplasmic mRNP particles. Here we show that small amounts of FMRP, FXR1P and FXR2P shuttle between cytoplasm and nucleus. Mutant FMRP of a severely affected fragile X patient (FMRPI304N) does not associate with polyribosomes and shuttles more frequently than normal FMRP, indicating that the association with polyribosomes regulates the shuttling process. Using leptomycin B we demonstrate that transport of the FXR proteins out of the nucleus is mediated by the export receptor exportin1. Finally, inactivation of the nuclear export signal in two FXR proteins shows that FMRP shuttles between cytoplasm and nucleoplasm, while FXR2P shuttles between cytoplasm and nucleolus. Therefore, molecular dissection of the shuttling routes used by the FXR proteins suggests that they transport different RNAs.

Mutations in GDI1 are responsible for X-linked non-specific mental retardation

Rab GDP-dissociation inhibitors (GDI) are evolutionarily conserved proteins that play an essential role in the recycling of Rab GTPases required for vesicular transport through the secretory pathway. We have found mutations in the GDI1 gene (which encodes uGDI) in two families affected with X-linked non-specific mental retardation. One of the mutations caused a non-conservative substitution (L92P) which reduced binding and recycling of RAB3A, the second was a null mutation. Our results show that both functional and developmental alterations in the neuron may account for the severe impairment of learning abilities as a consequence of mutations in GDI1, emphasizing its critical role in development of human intellectual and learning abilities.

Subcellular localization of the oncoprotein MTG8 (CDR/ETO) in neural cells

The t(8;21) translocation associated with acute myeloid leukemia (AML) disrupts two genes, the AML1 gene also known as the core binding factor A2 (CBFA2) on chromosome 21, and a gene on chromosome 8, hereafter referred to as MTG8, but also known as CDR and ETO. Extensive information is available on AML1, a member of the CBF family of transcription factors, containing a highly conserved domain, the runt box, of the Drosophila segmentation gene runt. This gene is essential for the hematopoietic development and is found disrupted in several leukemias. In contrast, the function of the MTG8 gene is poorly understood. The predicted protein sequence shows two unusual, putative zinc-fingers, three proline-rich regions, a PEST domain and several phosphorylation sites. In addition, we found a region encompassing aa 443-514 predicted to have a significant propensity to form coiled coil structures. MTG8 displays a high degree of similarity with nervy, a homeotic target gene of Drosophila, expressed in the nervous system. Human and mouse wild-type MTG8 are also highly expressed in brain relative to other tissues. For these reasons, we set out to investigate the expression and subcellular localization of the MTG8 protein in neural cells. Immunohistochemical experiments in a 12.5-day-old mouse embryo clearly showed that the protein was expressed in the neural cells of the developing brain and the spinal cord. In primary cultures of hippocampal neurons of 2-3 day-old mice, MTG8 was found in the nucleus, in the cytoplasm and as fine granules in the neurites. Cytoplasmic localization of the protein was observed in Purkinje cells of both human and mouse cerebellum. The molecular mass of MTG8 in total human and mouse brain was analysed by immunoblotting and determined to be between 70 and 90 kDa. Isoforms with the same molecular mass were demonstrated in synaptosomes isolated from mouse forebrain. The evidence of MTG8 in the nucleus and cytoplasm of neural cells suggests a specific mechanism regulating the subcellular localization of the protein.

Differential expression of FMR1, FXR1 and FXR2 proteins in human brain and testis

Lack of expression of the fragile X mental retardation protein (FMRP) results in mental retardation and macroorchidism, seen as the major pathological symptoms in fragile X patients. FMRP is a cytoplasmic RNA-binding protein which cosediments with the 60S ribosomal subunit. Recently, two proteins homologous to FMRP were discovered: FXR1 and FXR2. These novel proteins interact with FMRP and with each other and they are also associated with the 60S ribosomal subunit. Here, we studied the expression pattern of the three proteins in brain and testis by immunohistochemistry. In adult brain, FMR1, FXR1 and FXR2 proteins are coexpressed in the cytoplasm of specific differentiated neurons only. However, we observed a different expression pattern in fetal brain as well as in adult and fetal testis, suggesting independent functions for the three proteins in those tissues during embryonic development and adult life.

Association of FMRP with ribosomal precursor particles in the nucleolus

The fragile X syndrome, one of the most common forms of inherited mental retardation, is caused by an expansion of a polymorphic CGG repeat upstream the coding region of the FMR1 gene. These expansions are associated with hypermethylation of the FMR1 gene, which results in the absence of the gene product, the FMR1 protein (FMRP). The physiological function of FMRP remains to be determined. We studied the ultrastructural localization of FMRP at the electron microscopical level using the immunogold technique. FMRP is associated with ribosomes attached to the endoplasmic reticulum and with ribosomes free in the cytoplasm. In addition, FMRP is found in the nucleus where the protein is associated with the granular component of the nucleolus. The cellular function of FMRP is hypothesized in relation to its subcellular distribution.

FMRP is associated to the ribosomes via RNA

The FMR1 transcript is alternatively spliced and generates different splice variants coding for FMR1 proteins (FMRP) with a predicted molecular mass of 70-80 kDa. FMRP is widely expressed and localized in the cytoplasm. To study a possible interaction with other cellular components, FMRP was isolated and characterized under non-denaturing conditions. Under physiological salt conditions FMRP appears to have a molecular mass of > 600 kDa, indicating a binding to other cellular components. This interaction is disrupted in the presence of high salt concentrations. The dissociation conditions to free FMRP from the complex are similar to the dissociation of FMRP from RNA as shown before. The binding of FMRP from the complex is also disrupted by RNAse treatment. That the association of FMRP to a high molecular weight complex possibly occurs via RNA, is further supported by the observation that the binding of FMRP, containing an lle304Asn substitution, to the high molecular weight complex is reduced. An equal reduced binding of mutated FMRP to RNA in vitro was observed before under the same conditions. The reduced binding of FMRP with the lle304Asn substitution further indicates that the interaction to the complex indeed occurs via FMRP and not via other RNA binding proteins. In a reconstitution experiment where the low molecular mass FMRP (70-80 kDa) is mixed with a reticulocyte lysate (enriched in ribosomes) it was shown that FMRP can associate to ribosomes and that this binding most likely occurs via RNA.

A family of transmembrane proteins with homology to the MET-hepatocyte growth factor receptor

In hunting for unknown genes on the human X chromosome, we identified a cDNA in Xq28 encoding a transmembrane protein (SEX) of 1871 amino acids. SEX shares significant homology with the extracellular domain of the receptors encoded by the oncogenes MET, RON, and SEA [hepatocyte growth factor (HGF) receptor family]. Further screenings of cDNA libraries identified three additional sequences closely related to SEX: these were named SEP, OCT, and NOV and were located on human chromosomes 3p, 1, and 3q, respectively. The proteins encoded by these genes contain large cytoplasmic domains characterized by a distinctive highly conserved sequence (SEX domain). Northern blot analysis revealed different expression of the SEX family of genes in fetal tissues, with SEX, OCT, and NOV predominantly expressed in brain, and SEP expressed at highest levels in kidney. In situ hybridization analysis revealed that SEX has a distinctive pattern of expression in the developing nervous system of the mouse, where it is found in postmitotic neurons from the first stages of neuronal differentiation (9.5 day postcoitus). The SEX protein (220 kDa) is glycosylated and exposed at the cell surface. Unlike the receptors of the HGF family, p220SEX, a MET-SEX chimera or a constitutively dimerized TPR-SEX does not show tyrosine kinase activity. These data define a gene family (SEX family) involved in the development of neural and epithelial tissues, which encodes putative receptors with unexpected enzymatic or binding properties.

A comparative transcriptional map of a region of 250 kb on the human and mouse X chromosome between the G6PD and the FLN1 genes

The transcriptional organization of the region of the mouse X chromosome between the G6pd and the Fln1 genes was studied in detail, and it was compared with the syntenic region of the human chromosome. A cosmid contig of 250 kb was constructed by screening mouse cosmid libraries with probes for human genes and with whole cosmids. Overlapping cosmids were aligned by comparing EcoRI and rare-cutter restriction enzyme digestions. The gene order and the orientation of transcription were determined by hybridization with fragments from the 5' and 3' moieties of each cDNA. Our work demonstrates that all of the new genes identified in human are present in the mouse. The size of the region, 250 kb, is also very similar, as are gene order and gene organization: the transcriptional organization in "domains" described in human is found to be identical in the mouse. The major difference detected is the much lower content in rare-cutter restriction sites, which is related to the lower G+C and CpG content of mouse DNA. The very high conservation that we have described suggests that a potent selective pressure has contributed to such conservation of gene organization.

Transcriptional organization of a 450-kb region of the human X chromosome in Xq28

In this paper, we report the transcriptional organization of a 450-kb gene cluster in Xq28, flanked by the glucose-6-phosphate dehydrogenase and the color vision genes. CpG islands previously identified and mapped to distal Xq28 have helped in construction of a continuous contig of cosmids and in identification of cDNAs corresponding to eight transcripts. Thirteen to 16 small genes with CpG islands are clustered in a region of 250-300 kb. Many are highly expressed in muscle or brain and may be the genes responsible for muscle or neurological disorders mapped to distal Xq28. Our analysis indicates that, in this region of the genome, genes not related in sequence are organized in transcriptional domains of 100 kb and that this organization may be important for establishing and regulating gene expression in relation to tissue distribution and X chromosome inactivation.

Identification and characterization of a new gene in the human Xq28 region

A human Xqter chromosome cosmid library was screened with a mixed probe derived from porcine kidney mRNA. A new expressed gene was identified in a cosmid clone known to be part of a G6PD cosmid contig. This gene is most likely a housekeeping gene because the cDNA clone recognizes a 1 kb mRNA transcript in all cell lines and tissues tested. Hybridizing genomic DNA of several species with a cDNA probe indicated that the gene is highly conserved during evolution and that it belongs to a gene family. The genomic sequence shows a 100% homology with the recently identified QM cDNA sequence.

An archipelago of CpG islands in Xq28: identification and fine mapping of 20 new CpG islands of the human X chromosome

19 probes for CpG islands, mapping to Xq28, have been used as probes to construct a physical map of genes of this band of the human X chromosome. A total of 22 CpG islands have been precisely mapped in respect to known loci along the 9-10 Mb of Xq28. The fine mapping of such a large number of CpG islands has demonstrated that also in gene rich Giemsa light bands, like Xq28, gene distribution is non uniform: the CpG islands are clustered in the distal portion of the band in a 2 Mb region between the G6PD gene and the DXS15 locus. Moreover, 16 CpG islands were found between the G6PD and the RCP/GCP genes, a region of DNA of only about 300 kb. If this structural organization has a biological function it has yet to be determined. However, the isolation of large genomic regions enriched in gene sequences and the availability of cosmid or YAC contigs will provide the means to test the significance of such gene organization, as well as the material for large sequencing projects and gene search, for the identification of candidate genes for inherited disorders mapped to Xq28 and for comparative mapping.

Methylation and sequence analysis around EagI sites: identification of 28 new CpG islands in XQ24-XQ28

Thirty-two probes for CpG islands of the distal long arm of the human X chromosome have been identified. From a genomic library of DNA of the hamster-human cell hybrid X3000.1 digested with the rare cutter restriction enzyme EagI, 53 different human clones have been isolated and characterized by methylation and sequence analysis. The characteristic pattern of DNA methylation of CpG islands at the 5' end of genes of the X chromosome has been used to distinguish between EagI sites in CpG islands versus isolated EagI sites. The sequence analysis has confirmed and completed the characterization showing that sequences at the 5' end of known genes were among the clones defined CpG islands and that the non-CpG islands clones were mostly repetitive sequences with a non-methylated or variably methylated EagI site. Thus, since clones corresponding to repetitive sequences can be easily identified by sequencing, such libraries are a very good source of CpG islands. The methylation analysis of 28 different new probes allows to state that demethylation of CpG islands of the active X and methylation of those on the inactive X chromosome are the general rule. Moreover, the finding, in all instances, of methylation differences between male and female DNA is in very strong support of the notion that most genes of the distal long arm of the X chromosome are subject to X inactivation.