Sustained TGF beta exposure suppresses Smad and non-Smad signalling in mammary epithelial cells, leading to EMT and inhibition of growth arrest and apoptosis.

To better understand the dual, tumour-suppressive and tumour-promoting function of transforming growth factor-beta (TGFbeta), we analysed mammary epithelial NMuMG cells in response to short and long-term TGFbeta exposure. NMuMG cells became proliferation-arrested and apoptotic after exposure to TGFbeta for 2-5 days, whereas surviving cells underwent epithelial-mesenchymal transition (EMT). After chronic TGFbeta exposure (2-3 weeks), however, NMuMG cells became resistant to proliferation arrest and apoptosis, showing sustained EMT instead (TD cells). EMT was fully reversed by a pharmacologic TGFbeta-receptor-I kinase inhibitor or withdrawal of TGFbeta for 6-12 days. Interestingly, both cell cycle arresting/proapoptotic (Smads, p38 kinase) and antiapoptotic, proliferation and EMT-promoting signalling pathways (PI3K-PKB/Akt, ERK) were co-suppressed to low, but significant levels. Except for PI3K-Akt, TGFbeta-dependent downregulation of these signalling pathways in transdifferentiated (TD) cells was fully reversed upon TGFbeta withdrawal, together with partial re-induction of proliferation arrest and apoptosis. Co-injection of non-tumorigenic NMuMG cells with tumour-forming CHO cells oversecreting exogenous TGFbeta1 (CHO-TGFbeta1) allowed outgrowth of epithelioid cells in CHO-TGFbeta1 cell-induced tumours. These epithelial islands enhanced CHO-TGFbeta1 tumour cell proliferation, possibly due to chemokines (for example, JE/MCP-1) secreted by NMuMG/TD cells. We conclude that suppression of antiproliferative, proapoptotic TGFbeta signalling in TD cells may permit TGFbeta-dependent proliferation, survival and EMT-enhancing signalling pathways to act at low levels. Thus, TGFbeta may modulate its own signalling to facilitate switching from tumour suppression to tumour progression.

Evolutionarily plastic regions at human 3p21.3 coincide with tumor breakpoints identified by the "elimination test".

We have previously found with the microcell hybrid-based "elimination test" that human chromosome 3 transferred into murine or human tumor cells regularly lost certain 3p regions during tumor growth in SCID mice. The most common eliminated region, CER1, is approximately 2.4 Mb at 3p21.3. CER1 breakpoints were clustered in approximately 200-kb regions at both telomeric and centromeric borders. We have also shown, earlier, that tumor-related deletions often coincide with human/mouse synteny breakpoints on 3p12-p22. Here we describe the results of a comparative genomic analysis on the CER1 region in Caenorhabditis elegans, Drosophila melanogaster, Fugu rubripes, Gallus gallus, Mus musculus, Rattus norvegicus, and Canis familiaris. First, four independent synteny breaks were found within the CER1 telomeric breakpoint cluster region, comparing human, dog, and chicken genomes, and two independent synteny breaks within the CER1 centromeric breakpoint cluster region, comparing human, mouse, and chicken genomes, suggesting a nonrandom involvement of tumor breakpoint regions in chromosome evolution. Second, both CER1 breakpoint cluster regions show recent tandem duplications (seven Zn finger protein family genes at the telomeric and eight chemokine receptor genes at the centromeric side). Finally, all genes from these regions underwent horizontal evolution in mammals, with formation of new genes and expansion of gene families, which were displayed in the human genome as tandem gene duplications and pseudogene insertions. In contrast the CER1 middle region contained evolutionarily well-conserved solitary genes and a minimal amount of retroposed genes. The coincidence of evolutionary plasticity with CER1 breakpoints may suggest that regional structural instability is expressed in both evolutionary and cancer-associated chromosome rearrangements.

Human wig-1, a p53 target gene that encodes a growth inhibitory zinc finger protein.

We previously identified a novel p53-induced mouse gene, wig-1, that encodes a 290 amino acid zinc finger protein (Varmeh-Ziaie et al., 1997). Here we have identified and characterized the human homolog of mouse wig-1. The human wig-1 protein is 87% identical to the mouse protein and contains three zinc finger domains and a putative nuclear localization signal. Human wig-1 mRNA and protein is induced following activation of wild type p53 expression in our BL41-ts p53 Burkitt lymphoma cells. Wig-1 is also induced in MCF7 cells following treatment with the DNA-damaging agent mitomycin C. Northern blotting detected low levels of wig-1 mRNA in normal human tissues. Fluorescence in situ hybridization mapped wig-1 to human chromosome 3q26.3-27. FLAG-tagged human wig-1 localizes to the nucleus. Ectopic overexpression of human wig-1 inhibits tumor cell growth in a colony formation assay. These results suggest that human wig-1 has a role in the p53-dependent growth regulatory pathway.

The position of t(11;22)(q23;q11) constitutional translocation breakpoint is conserved among its carriers.

The t(11;22)(q23;q11) translocation is the most common recurrent balanced translocation described in humans. Carriers are phenotypically normal and often go undetected until diagnosis as a result of infertility investigations or following the birth of chromosomally unbalanced offspring. Efficient diagnostics of t(11;22) is important for children born to carriers of the translocation and for prenatal and pre-implantation diagnosis. The translocation breakpoint on chromosome 22 is located within a region containing low copy repeats, and this site is one of the last unfilled gaps in the sequence of this chromosome. This autosome harbors multiple other low copy repeats, which have been entirely sequenced. We report a combined sequencing and fiber FISH breakpoint characterization in five translocation carriers. From one carrier a cosmid library was constructed, and two chimeric cosmids (cos4_der11 and cos6_der22) were sequenced, which showed that strong palindromes (or inverted repeats) occur on both chromosomes. The translocation breakpoints occur at the tip of both inverted repeats. The palindrome on chromosomes 22 and 11 is composed of 852 and 166 bases, respectively. Four additional carriers were studied using fiber FISH with a resolution limit of 2 kb. Analysis of breakpoints on the DNA sequence level, or at the level of fiber FISH, indicate that they occur at the same position on both chromosomes in all five carriers. Using cos6_der22, PAC 158L19 and BAC 3009A19, we demonstrate that FISH is an attractive alternative in molecular diagnostics of t(11;22), as PCR assays are not reliable, due to the presence of numerous copies of low copy repeats.

The LZTFL1 gene is a part of a transcriptional map covering 250 kb within the common eliminated region 1 (C3CER1) in 3p21.3.

Deletions on 3p have been described in a large number of human tumors, suggesting the presence of a tumor suppressor gene(s). Using the elimination test, we previously defined a 1-Mb segment from human 3p21.3 (C3CER1). Genomic sequencing allowed us to construct a transcription map covering 250 kb containing five genes. We have characterized a human leucine zipper containing gene, leucine zipper transcription factor-like 1 (LZTFL1), and its mouse orthologue (Lztfl1), which was also mapped to mouse chromosome 9F. The LZTFL1 gene has two transcript isoforms displaying alternative polyadenylation. We have localized the human orthologue of the yeast SAC1 (suppressor of actin) gene as well as characterized and mapped the mouse Sac1 gene. Furthermore, the XT3 gene was characterized, encoding a member of the Na(+)/Cl(-) neurotransmitter superfamily. It has been shown that the XT3 gene had an alternatively spliced brain-specific isoform, predicted to remove 1 of 12 putative transmembrane domains. The transcription map also includes the CC chemokine receptor 9 gene (CCR9) and the LIM domain containing gene 1 (LIMD1). This work partially defines the gene content of C3CER1 that is a prerequisite for delineation of its role in tumorigenesis.

Similar regions of human chromosome 3 are eliminated from or retained in human/human and human/mouse microcell hybrids during tumor growth in severe combined immunodeficient (SCID) mice.

By passaging microcell hybrids (MCHs) containing human chromosome 3 (chr3) on A9 mouse fibrosarcoma background through severe combined immunodeficient (SCID) mice (elimination test), we have previously defined a 1-Mb-long common eliminated region 1 (CER1) at 3p21.3, a second eliminated region (ER2) at 3p21.1-p14 and a common retained region (CRR) at 3q26-qter. In the present work, chr3 was transferred by microcell fusion into the human nonpapillary renal cell carcinoma line KH39 that contained uniparentally disomic chr3. Four MCHs were generated. Compared with KH39, they developed fewer and smaller tumors, which grew after longer latency periods in SCID mice. The tumors were analyzed in comparison with corresponding MCHs by chr3 arm-specific painting, 19 fluorescent in situ hybridization (FISH) probes, and 27 polymorphic markers. Three MCHs that maintained the intact exogenous chr3 in vitro lost one 3p copy in all 11 tumors. Seven of 11 tumors lost the exogenous 3p, whereas four tumors contained mixed cell populations that lacked either the exogenous or one endogenous KH39 derived 3p. In one MCH the exogenous chr3 showed deletions within CER1 and ER2 already in vitro. It remained essentially unchanged in 8/9 derived tumors. The third, exogenous copy of the 3q26-q27 region (part of CRR) was retained in 16/20 tumors. It can be concluded that the human/human MCH-based elimination test identifies similar eliminated and retained regions on chr3 as the human/murine MCH-based test.

Fine mapping of the constitutional translocation t(11;22)(q23;q11).

Translocation t(11;22)(q23;q11) is the most common constitutional reciprocal translocation in man. Balanced carriers are phenotypically normal, except for decreased fertility, an increased spontaneous abortion rate and a possible predisposition to breast cancer in some families. Here, we report the high resolution mapping of the t(11;22)(q23;q11) breakpoint. We have localised the breakpoint, by using fluorescence in situ hybidisation (FISH) walking, to a region between D11S1340 and WI-8564 on chromosome 11, and D22S134 and D22S264 on chromosome 22. We report the isolation of a bacterial artificial chromosome (BAC) clone spanning the breakpoint in 11q23. We have narrowed down the breakpoint to an 80-kb sequenced region on chromosome 11 and FISH analysis has revealed a variation of the breakpoint position between patients. In 22q11, we have sequenced two BACs (BAC2280L11 and BAC41C4) apparently mapping to the region; these contain low copy repeats (LCRs). Southern blot analysis with probes from BAC2280L11 has revealed different patterns between normal controls and translocation carriers, indicating that sequences similar/identical to these probes flank the translocation breakpoint. The occurrence of LCRs has previously been associated with genomic instability and "unclonable" regions. Hence, the presence of such repeats renders standard translocation breakpoint cloning techniques ineffective. Thus, we have used high resolution fiber-FISH to study this region in normal and translocation cases by using probes from 22q11, LCRs and 11q23. We demonstrate that the LCR containing the gap in 22q11 is probably substantially larger than the previous estimates of 100 kb. Using fiber-FISH, we have localised the breakpoint in 22q11 to approximately 20-40 kb from the centromeric border of the LCR (i.e. the telomeric end of AC006547) and have confirmed the breakpoint position on 11q23.

Combined LOH/CGH analysis proves the existence of interstitial 3p deletions in renal cell carcinoma.

We have recently developed an allele titration assay (ATA) to assess the sensitivity and influence of normal cell admixture in loss of heterozygosity (LOH) studies based on CA-repeat. The assay showed that these studies are biased by the size-dependent differential sensitivity of allele detection. Based on these data, we have set up new criteria for evaluation of LOH. By combining these new rules with comparative genome hybridization (CGH) we have shown the presence of interstitial deletions in renal cell carcinoma (RCC) biopsies and cell lines. At least three out of 11 analysed RCC cell lines and three out of 37 biopsies contain interstitial deletions on chromosome 3. Our study suggests the presence of several regions on human chromosome 3 that might contribute to tumor development by their loss: (i) 3p25-p26, around the VHL gene (D3S1317); (ii) 3p21. 3-p22 (between D3S1260 and D3S1611); (iii) 3p21.2 (around D3S1235 and D3S1289); (iv) 3p13-p14 (around D3S1312 and D3S1285). For the first time, AP20 region (3p21.3-p22) was carefully tested for LOH in RCC. It was found that the AP20 region is the most frequently affected area. Our data also suggest that another tumor suppressor gene is located near the VHL gene in 3p25-p26.

Inactivation of the human fragile histidine triad gene at 3p14.2 in monochromosomal human/mouse microcell hybrid-derived severe combined immunodeficient mouse tumors.

We have previously shown that inoculation of human chromosome 3 (chr3)/A9 mouse fibrosarcoma microcell hybrids (MCHs) into severe combined immunodeficient (SCID) mice was followed by the regular elimination of some 3p regions whereas a 3q region was retained even after prolonged mouse passage. Using this approach, referred to as the elimination test (Et), we have defined a common eliminated region (CER) of approximately 7 cM at 3p21.3 that was absent in all of the 27 tumors generated from five MCHs. Later, CER was reduced to a 1-Mb region, designated as CER1. Another eliminated region (ER2) at 3p21.1-p14.2 was absent in 21 of the 27 tumors. ER2 borders at but does not include the fragile histidine triad (FHIT) gene, considered as a putative tumor suppressor gene. In the present work, two new and two previously studied MCHs, and 13 derived SCID mouse tumors were analyzed by fluorescence in situ hybridization (FISH) chromosome painting and by PCR, using 72 chr3p-specific and 11 chr3q-specific markers. Nine tumors generated from three MCHs that carried cytogenetically normal chr3, remained PCR-positive for all of the chr3 markers tested. Designated as "PCR+" tumors, they were examined by reverse transcription (RT)-PCR, together with four of six previously studied tumors derived from MCH910.7, which carried a del(3)(pter-p21.1), for the expression of 14 human genes: 5 genes within CER1 (LIMD1, CCR1, CCR2, CCR3, CCR5), 5 genes located within regions that were homozygously deleted in a variety of carcinomas (ITGA4L, LUCA1, PTPRG, FHIT, DUTT1), and 4 other genes in chr3p (VHL, MLH1, TGM4, UBE1L). We found that VHL, MLH1, ITGA4L, LIMD1, UBE1L, LUCA1, PTPRG, and DUTT1 were expressed in the MCH lines in vitro and also in the derived SCID tumors. No transcripts that originated from the four CCR genes or from TGM4 could be detected in any of the MCH lines. Alone among the 14 genes examined, FHIT showed a tumor growth-associated change. It was expressed in vitro in five of seven MCH lines. Nine of 13 derived tumors had no FHIT transcript. The remaining 4 expressed a truncated mRNA and a reduced amount of the full-length mRNA. We have previously found that FHIT was deleted at the DNA level in 17 of 21 tumors derived from four MCHs. The remaining 4 of 21 had no FHIT transcript. Our compiled data show that FHIT was either physically or functionally impaired in all 34 of the 34 analyzed tumors. Variants with deleted or down-regulated FHIT have a selective growth advantage.

A 1-Mb PAC contig spanning the common eliminated region 1 (CER1) in microcell hybrid-derived SCID tumors.

We have developed an elimination test to identify chromosomal regions that contain tumor inhibitory genes. Monochromosomal human/mouse microcell hybrids are generated and passaged through SCID mice. Derived tumors are then analyzed for deletions on the transgenomic chromosome. Using this strategy, we have previously identified a 1.6-cM common eliminated region 1 (CER1) on human 3p21. 3. We now report that CER1 contains 14 markers that are deleted in 19 SCID-derived tumors. A 1-Mb PAC contig that spans CER1 was assembled. Five chemokine receptor genes (CCR1, CCR3, CCR2, CCR5, and CCR6) were localized in CER1 in a 225-kb cluster. The lactotransferrin gene (LTF, or lactoferrin, LF), which reportedly has tumor inhibitory activity, also maps to CER1. Our results create a basis for characterization and further functional testing of genes within CER1.

TOM1 genes map to human chromosome 22q13.1 and mouse chromosome 8C1 and encode proteins similar to the endosomal proteins HGS and STAM.

The avian tom1 (target of myb 1) gene has been previously characterized from v-myb-transformed cells. We report here cloning of the human and mouse tom1 orthologs. Both genes are expressed ubiquitously, with the highest levels in skeletal muscle, brain, and intestines, as assessed by Northern blot and mRNA in situ hybridization. The N-terminal domain of the TOM1 protein shares similarity with HGS (hepatocyte growth factor-regulated tyrosine kinase substrate) and STAM (signal-transducing adaptor molecule), which are associated with vesicular trafficking at the endosome. A putative coiled-coil domain was also detected in the central part of the TOM1 protein. This domain structure suggests that TOM1 is another member of a family of genes implicated in the trafficking regulation of growth-factor-receptor complexes that are destined for degradation in the lysosome. We also show that a human paralog of TOM1 (TOM1-like gene 1) exists. Furthermore, we provide a transcription map over a 190-kb contig of the TOM1 region. This map includes its distal neighbors HMOX1 and MCM5 and two proximal novel genes, one of which is a HMG-box-containing gene (HMG2L1), and the other of unknown function. Using a genomic PAC clone, we demonstrate that the mouse Tom1 and Hmox1 genes are part of an as yet undescribed syntenic group between mouse chromosome 8C1 and human chromosome 22q13.1.

Visualization of alternative Epstein-Barr virus expression programs by fluorescent in situ hybridization at the cell level.

Epstein-Barr virus (EBV) transforms human B lymphocytes into immortalized lymphoblastoid cell lines (LCLs). They regularly express six virally encoded nuclear proteins (EBNA1 to EBNA6) and three membrane proteins (LMP1, LMP2A, and LMP2B). In contrast, EBV-carrying Burkitt lymphoma (BL) cells in vivo and derived type I cell lines that maintain the BL phenotype express only EBNA1. During prolonged in vitro culturing, most EBV-carrying BL lines drift toward a more immunoblastic (type II or III) phenotype. Their viral antigen expression is upregulated in parallel. We have used fluorescent in situ hybridization to visualize viral transcripts in type I and III BL lines and LCLs. In type I cells, EBNA1 is encoded by a monocistronic message that originates from the Qp promoter. In type III cells, the EBNA1 transcript is spliced from a giant polycistronic message that originates from one of several alternative Wp or Cp promoters and encodes all six EBNAs. We have obtained a "track" signal with a BamHI W DNA probe that could hybridize with the polycistronic but not with the monocistronic message in two type III BL lines (Namalwa-Cl8 and MUTU III) and three LCLs (LCL IB4-D, LCL-970402, and IARC-171). A BamHI K probe that can hybridize to both the monocistronic and the polycistronic message visualized the same pattern in the type III BLs and the LCLs as the BamHI W probe. A positive signal was obtained with the BamHI K but not the BamHI W probe in the type I BL lines MUTU I and Rael. The RNA track method can thus distinguish between cells that use a type III and those that use a type I program. The former cells hybridize with both the W and the K probes, but the latter cells hybridize with only the K probe. Our findings may open the way for studies of the important but still unanswered question of whether cells with type I latency arise from immunoblasts with a full type III program or are generated by a separate pathway during primary infection.

Stability of a functional murine satellite DNA-based artificial chromosome across mammalian species.

A 60-Mb murine chromosome consisting of murine pericentric satellite DNA and two bands of integrated marker and reporter genes has been generated de novo in a rodent/human hybrid cell line (mM2C1). This prototype mammalian artificial chromosome platform carries a normal centromere, and the expression of its beta-galactosidase reporter gene has remained stable under selection for over 25 months. The novel chromosome was transferred by a modified microcell fusion method to mouse [L-M(TK-)], bovine (P46) and human (EJ30) cell lines. In all cases, the chromosome remained structurally and functionally intact under selection for periods exceeding 3 months from the time of transfer into the new host. In addition, the chromosome was retained in three first-generation tumours when L-M(TK-) cells containing the chromosome were xenografted in severe combined immunodeficiency mice. These data support that a murine satellite DNA-based artificial chromosome can be used as a functional mammalian artificial chromosome and can be maintained in vivo and in cells of heterologous species in vitro.

Genomic structure and chromosomal localization of the mouse persyn gene.

Synucleins are a family of small intracellular proteins expressed mainly in the nervous system. The involvement of synucleins in neurodegeneration and malignancy has been demonstrated, but the physiological functions of these proteins remain elusive. Further studies including generation of animals with modified persyn expression are necessary to clarify the functions of these proteins and the mechanisms of their involvement in human diseases. We cloned and determined the organization and chromosomal localization of the mouse gene coding for persyn, a member of the synuclein family. The gene is composed of five exons, and its general structure is very similar to that of the human persyn gene. Using fluorescence in situ hybridization, we assigned the persyn gene to the boundary of bands B and C on mouse chromosome 14. We found a fragment of the gene that directs expression of the persyn protein in sensory neurons and could be used for generation of transgenic animals.

The human LARGE gene from 22q12.3-q13.1 is a new, distinct member of the glycosyltransferase gene family.

Meningioma, a tumor of the meninges covering the central nervous system, shows frequent loss of material from human chromosome 22. Homozygous and heterozygous deletions in meningiomas defined a candidate region of >1 Mbp in 22q12.3-q13.1 and directed us to gene cloning in this segment. We characterized a new member of the N-acetylglucosaminyltransferase gene family, the LARGE gene. It occupies >664 kilobases and is one of the largest human genes. The predicted 756-aa N-acetylglucosaminyltransferase encoded by LARGE displays features that are absent in other glycosyltransferases. The human like-acetylglucosaminyltransferase polypeptide is much longer and contains putative coiled-coil domains. We characterized the mouse LARGE ortholog, which encodes a protein 97.75% identical with the human counterpart. Both genes reveal ubiquitous expression as assessed by Northern blot analysis and in situ histochemistry. Chromosomal mapping of the mouse gene reveals that mouse chromosome 8C1 corresponds to human 22q12.3-q13.1. Abnormal glycosylation of proteins and glycosphingolipids has been shown as a mechanism behind an increased potential of tumor formation and/or progression. Human tumors overexpress ganglioside GD3 (NeuAcalpha2,8NeuAcalpha2, 3Galbeta1,4Glc-Cer), which in meningiomas correlates with deletions on chromosome 22. It is the first time that a glycosyltransferase gene is involved in tumor-specific genomic rearrangements. An abnormal function of the human like-acetylglucosaminyltransferase protein may be linked to the development/progression of meningioma by altering the composition of gangliosides and/or by effect(s) on other glycosylated molecules in tumor cells.

A novel gene containing LIM domains (LIMD1) is located within the common eliminated region 1 (C3CER1) in 3p21.3.

Chromosomal deletions on 3p have been described in a large number of human tumors, suggesting the presence of a tumor suppressor gene(s). Using an experimental system, called the elimination test, we previously identified a 1 Mb segment, the common eliminated region 1 (C3CER1). C3CER1 was also covered by a PAC contig. Using the sequence of two overlapping PACs from C3CER1, we localized the human KIAA0028 cDNA, encoding the precursor of mitochondrial leucyl-tRNA synthetase. We also characterized a novel human LIM domain-containing gene (LIMD1) and its mouse ortholog (Limd1). LIM domains consist of a cysteine-rich consensus sequence containing two distinct zinc-binding subdomains, which mediate protein-protein interactions. The predicted protein sequences of the human and mouse genes reveal three LIM domains located at the C-terminal end, which indicates that they belong to the group 3 of the gene family encoding LIM motifs. We characterized the genomic structure of the human LIMD1 gene and assigned the mouse Limd1 gene to the chromosome 9F subtelomeric region. Both genes are ubiquitously expressed at the mRNA level. The LIM motif has been previously identified in many developmentally important factors from various eukaryotes. These factors have been shown to play a role in intracellular signaling, transcriptional regulation and cellular differentiation during development. The human C3CER1-located LIMD1 gene should therefore be further studied for its possible role in tumor suppression.

A minimal glycine-alanine repeat prevents the interaction of ubiquitinated I kappaB alpha with the proteasome: a new mechanism for selective inhibition of proteolysis.

The Epstein-Barr virus nuclear antigen 1 contains a glycine-alanine repeat that inhibits in cis MHC class I-restricted presentation. We report here that insertion of a minimal glycine-alanine repeat motif in different positions of I kappaB alpha protects this NF-kappaB inhibitor from signal-induced degradation dependent on ubiquitin-proteasome, and decreases its basal turnover in vivo resulting in constitutive dominant-negative mutants. The chimeras are phosphorylated and ubiquitinated in response to tumor necrosis factor alpha, but are then released from NF-kappaB and fail to associate with the proteasome. This explains how functionally competent I kappaB alpha is protected from proteasomal disruption and identifies the glycine-alanine repeat as a new regulator of proteolysis.

Organization, expression and polymorphism of the human persyn gene.

Persyn is a recently identified member of the synuclein family with a distinct pattern of expression during pre- and postnatal development of the mouse peripheral and central nervous systems. As with other synucleins, persyn is believed to be involved in the pathogenesis of human neurodegenerative diseases. However, in contrast to other synucleins, high levels of persyn mRNA expression were also found in advanced breast carcinomas, suggesting an involvement of the encoded protein in breast tumour progression. Here we have used an antibody specific to human persyn to demonstrate that the level of this protein is increased in ageing cerebral cortex and in breast tumours. We cloned, characterized and sequenced the human persyn genomic locus and localized it to the long arm of chromosome 10 in the q23.2-q23.3 region. Sequence information was used to search for specific mutations in the protein coding regions of persyn mRNA and the persyn gene in breast tumours and tumour cell lines. No tumour-specific mutations were found, but two linked polymorphisms in the coding region were detected, both in mRNA and exons III and IV of the gene. These results suggest that development of breast tumours correlates with overexpression of the wild-type persyn protein. Detailed characterization of the human persyn locus is important for further studies of the involvement of persyn in neurodegeneration and malignancy.