Altered composition of the AP-1 transcription factor in immortalized compared to normal proliferating cells.

We have investigated the expression of the AP-1 transcription factor proteins, fos, fosB, fra1, fra2, jun, junB, junD, using Western blot analysis, in several types of asynchronously proliferating cells. The latter included normal fibroblasts, immortalized but not tumourigenic fibroblasts, and two immortalized tumour cell lines. All cells expressed fos, fra1 and jun proteins and none expressed fosB. There were, however, interesting qualitative differences between the normal fibroblasts and the immortalized cells. Expression of fra2 was difficult to detect in normal cells, but was very evident in all of the immortalized cells. The normal cells only expressed a 44 kDa junB species, whereas the immortalized cells expressed both this and another 34 kDa species. All of the cells expressed the two junD proteins but the smaller 39 kDa species was more prominent in the normal cells, whereas the larger 44 kDa protein was more prominent in the immortalized cells. These data indicate that immortalized cells are not simply cells in which the ageing process has been prevented or reversed, but instead exhibit additional characteristics to those associated with young normal cells.

Altered composition and DNA binding activity of the AP-1 transcription factor during the ageing of human fibroblasts.

We have investigated the expression of AP-1 transcription factor proteins during the in-vitro ageing of human fibroblasts. The numbers of these cells that are in the cell cycle gradually decreases up to 45 cumulative population doublings (cPD), thereafter the decline is steeper, until almost all cells enter a post-mitotic state by 60 cPD. We observed that a 34 kd junB species began to replace the 44 kd junB species after 41 cPD. This was followed, after 44 cPD, by a loss of fra1 and both junD species. After 49 cPD there was a gradual decline in the levels of fos and jun proteins, but disproportionately, so that the fos/jun protein ratio also declined. Although fos and jun proteins were still clearly present at 60 cPD, utilisation of the AP-1 DNA consensus sequence could not be demonstrated after 54 cPD. These data indicate that significant changes occur in the composition of the AP-1 transcription factor during ageing, but also that alterations in its DNA binding activity may involve other factors.

Telomerase and mammalian ageing: a critical appraisal.

The telomeres that occur at the end of chromosomes are maintained by the activity of telomerase and are thought to be important protective factors in maintaining the integrity of chromosomes. It now appears that in vitro replicative senescence, which has been observed in cultured somatic cells, is due to a loss of telomere length in those cells, caused by inactivity of telomerase. This has led to the proposition that telomerase activity is an important determinant in organismal ageing. However, many cells in the body do not proliferate regularly and therefore will not lose telomere length. Cells that do proliferate frequently have now been shown to have active telomerase. Other cells, such as fibroblasts, that do not have telomerase activity but proliferate only occasionally may not reach the Hayflick limit during the lifetime of an animal. There is also no correlation between telomere length and the maximal lifespan exhibited by different species. Studies of telomerase knock-out mice have reported some aspects of accelerated ageing after three generations, but the relevance of these observations to normal ageing remains unconvincing. The role of telomerase in producing immortal tumour cells and the possibility that activation of telomerase is an important event in malignant transformation is similarly controversial and open to alternative interpretations. The significance of these and other observations, and how they define the role of telomerase in ageing, is discussed.

Differential display analysis can reveal patterns of gene expression in immortalised hepatoma cells which are similar to those observed in young adult but not old adult liver cells.

We used the differential display technique to examine whether there were any patterns of gene expression which were characteristic of both young adult rat liver and of immortalised rat hepatoma cell lines, but not of old adult rat liver. No genes were detected which appeared to be clearly expressed in young liver and immortalised cell lines, but not in old liver. However, 14 genes were detected in old liver which were down-regulated in young liver and the hepatoma cell lines. This observation lends support to the idea that immortalisation of malignant cells may involve, at least in some aspects, a reversal of the ageing process in these cells and that the genes involved have a recessive action.

Semi-quantitative fluorescence in situ hybridization analysis indicates that the myc protein is consistently stabilized both before and after transformation of low-grade follicular center to high-grade diffuse large cell lymphoma.

We have investigated the expression of the MYC gene at both the mRNA and protein levels to determine how these parameters are related in lymphoma cells and in nonmalignant lymphoid cells. To do this we have adopted a multicolor fluorescence in situ hybridization methodology, which has allowed us to investigate the expression of different genes at the same time in the same cell. We have made use of the digital imaging capabilities of a charge-coupled device camera system to quantify the hybridization signals for the MYC gene and, by comparing these to the expression of a control gene (glyceraldehyde-3-phosphate dehydrogenase; GAPDH), have obtained relative quantitations of MYC mRNA and protein levels. In this study we have compared cells both within and outside the germinal centers in control tissues (reactive lymph nodes and tonsils) and in low-grade follicular center lymphomas, as well as cells in high-grade diffuse large cell lymphomas. The MYC/GAPDH mRNA hybridization signal ratios were calculated and found to be higher in cell populations containing a majority of malignant cells (p < 0.04). However, when the myc/GAPDH protein hybridization signal ratios were calculated, these were significantly higher in malignant cells from all lymphomas than the ratios observed in the nonmalignant cells (p < 0.0005). These observations indicate that the environment in a malignant cell may contribute to the stabilization of the myc protein, thus enabling it to function for a longer time period than in nonmalignant cells.

Identification of different gene expression patterns in low and high grade non-Hodgkin's lymphomas by differential display.

We have compared the patterns of gene expression in non-Hodgkin's lymphoma (NHL) biopsy samples from patients with either low grade or high grade disease, by the polymerase chain reaction (PCR) based technique of differential display. By using a combination of 30 primer pairs we estimate that we were able to survey over 3,000 genes expressed in these tissues. In this study we compared a group of three low grade follicular centre lymphomas with a group of two high grade diffuse large cell lymphomas and scored only those PCR products that were represented in all samples of each group. In doing so we were able to avoid many of the problems associated with the occurence of false PCR-positives. 139 differences were then scored as representing genes which may be differentially expressed during the transformation from low to high grade disease. However, as many of these might simply reflect changing populations of cells, we focused on only those genes which appeared to be expressed exclusively in either low grade or high grade disease. We have identified 14 such genes, of which 10 were low grade specific and 4 were high grade specific. This approach therefore appears to offer a systematic method for the identification and characterisation of differentially expressed genes, which are characteristic of different NHL sub-types.

Fluorescence in situ hybridization analysis of the fos/jun ratio in the ageing brain.

We have examined the expression of the fos and jun genes in the cerebellum of the rat brain during ageing, by use of a semi-quantitative fluorescence in situ hybridization (FISH) method. In these experiments we have utilised the digital imaging capabilities of a cooled CCD camera system to store the fluorescence intensities of individual cells and to compare the data from each target (fos or jun) gene with that of a control (beta-actin) gene. In this way we have been able to obtain a relative quantitation of fos and jun mRNA levels. Purkinje cells were analysed in brain from Sprague-Dawley rats of 6, 13 and 23 months of age. Data obtained in this way demonstrated that the level of fos expression decreased significantly during ageing but, in contrast, that of jun increased between 6 and 13 months and thereafter remained constant. We subsequently carried out a further comparison of fos/jun ratios in purkinje cells in Wistar rats and also observed a highly significant fall in the ratio between 6 and 23 months. This change in the fos/jun ratio has important implications for the composition of the AP-1 transcription factor and for the expression of genes that it regulates.

Cytogenetic analysis of non Hodgkin's lymphomas by ratio-painting and comparative genomic hybridization reveals unsuspected chromosomal abnormalities.

Cytogenetic analysis of cancer cells has proven to be a powerful tool in understanding malignant evolution and in providing clinically useful markers. In recent years the advent of new fluorescence in-situ hybridization (FISH) methods such as ratio-painting and comparative genomic hybridization (CGH) have enabled much more accurate karyotypes of malignant cells to be detected. In this study, we have examined the chromosomes present in malignant cells from a series of 6 low grade follicular centre and 2 high grade diffuse large cell non-Hodgkin's lymphomas (NHL) using conventional G-banding. In all cases chromosome abnormalities were observed, including the presence of marker chromosomes in six cases. The NHL cells were then subjected to the FISH method of ratio-painting. This provided a more accurate understanding of the origins of derivative chromosomes and identified the origins of all of the marker chromosomes. It also revealed hitherto unsuspected abnormalities. For example, in one case four abnormal chromosomes were demonstrated to contain material from chromosome 8, which had not been previously suspected from G-banding. Regions of amplification and deletion on the chromosomes were also investigated by CGH, which identified further unsuspected chromosomal abnormalities. For example, in case L124, trisomy of chromosome 7 was confirmed by CGH, but an unsuspected amplification of 3(p12) was also revealed. These approaches demonstrate the power of FISH technology in providing a more precise analysis of malignant cell chromosomes, and in doing so have produced comprehensive karyotypes of the NHL under study.

Differential display analysis of gene expression indicates that age-related changes are restricted to a small cohort of genes.

It is clear that there is a genetic component associated with the ageing process. Although evolutionary theory has suggested that the activity of certain genes may facilitate ageing by favouring resource utilisation by the germ cells at the expense of somatic cells, there is reason to believe that the senescent phenotype, which is the endpoint of the ageing process, may be due to alterations in the levels of expression of other genes. To investigate this situation we have used the differential display technique to survey gene expression during ageing of the rat brain, heart and liver. By optimising this technique it is possible to identify up to 10000-14000 PCR products, which represent genes expressed in the tissue under study. Interestingly, only a relatively small cohort (approximately 2%) of these genes appear to show significant changes in their levels of expression during ageing. Characterisation of the latter has so far revealed certain genes, such as glial fibrillary acidic protein, which are associated with the senescent phenotype. It has also revealed that the level of fos, a component of the AP-1 transcription factor, decreases with age, which has implications for AP-1 regulated genes. The differential display technique has also revealed an increase in mitochondrial RNA during ageing of the heart, which may be due to a gene dosage effect caused by the presence of increased numbers of mitochondrial genomes in myocytes in old age. The differential display technique therefore appears to offer a powerful tool for identifying genes which contribute to the emergence of a senescent phenotype.

Investigation of the effect of BB-94 (batimastat) on the colonization potential of human lymphoma cells in SCID mice.

We have assessed the effectiveness of the metalloproteinase inhibitor BB-94 (batimastat) in reducing the colonization potential of the human Burkitt lymphoma Namalwa cell line. In this study Namalwa cells were injected intraperitoneally into SCID mice and their spread to the spleen, liver and lung studied over a 3 week period. The human cells were detected in the mouse tissues by polymerase chain reaction (PCR) amplification of a human alu repeat sequence. Comparison of BB-94-treated animals with an untreated control group provided no evidence for a significant reduction in the colonization of mouse tissues by the human lymphoma cells in the presence of the drug. Tumour growth, after subcutaneous injection of the Namalwa cells into SCID mice, was similarly unaffected by BB-94. The significance of these results is discussed.

Investigation of the RB-1 tumour suppressor gene in a United Kingdom series of non-Hodgkin's lymphomas.

We have investigated the RB-1 tumour suppressor genes in a series of 20 non-Hodgkin's lymphomas (NHL). Polymerase chain reaction (PCR) amplification of polymorphic alleles indicated that there was evidence of allelic imbalance around 13q14, the site of the RB-1 gene, in at least 5 NHL. Immunohistochemical analysis of the RB-1 protein demonstrated wide variations in the percentage of cells exhibiting positive staining, but these usually correlated with differences in the proliferation index as indicated by staining of Ki67. Only 3/35 NHL exhibited significantly fewer cells expressing RB-1 protein than expressed Ki167. A comprehensive analysis of the mutation status of RB-1 in 20 NHL was carried out using PCR based strategies involving single strand conformational polymorphism (SSCP) gels. Most of the protein coding region was studied by analysing cDNA derived from its mRNA and the remaining 5'-end of the coding region investigated by analysing exon I of the gene. We also examined the promoter region of the gene. In none of the 20 NHL investigated were we able to identify a mutation: the only abnormal migrating fragment observed proved to be a polymorphism in exon I of the gene in 5 NHL. In one other case we detected instability at an intron repeat sequence, which had occurred during progression of the disease, but again no mutation of the protein coding region was found. The low levels of RB-1 protein expression that we had observed in a few of our NHL therefore did not appear to be due to mutation of the gene. These data suggest that mutation of RB-1 is not a common event in the evolution of NHL, but that there may be another, as yet unidentified, tumour suppressor gene near the RB-1 locus which is associated with NHL.

Age-related changes in gene expression in the rat brain revealed by differential display.

We have used the polymerase chain reaction (PCR)-based technique of differential display to analyse changes in gene expression during ageing of the rat brain. In this approach we have compared three young adult (6 months) with three old adult (20 months) animals. RNA preparations from the homogenised brains were subjected to reverse transcriptase (RT)-PCR using 36 different combinations of primer pairs. Any PCR product which was consistently found to be more prominent in the three young brains compared to the three old brains, and vice versa, was scored as potentially representing a gene which was differentially expressed during the ageing of this tissue. Out of a possible 2000+ PCR products we identified 44 that might represent genes that exhibit differential expression during ageing of the rat brain. An initial screen of these fragments, by Southern-blotting the PCR products and hybridising them with cDNA probes derived from either young or old brain RNA preparations, indicated that 40% of them represented genes that were differentially expressed. This approach is likely to prove invaluable for identifying cohorts of genes that show differential expression during the ageing process.

Analysis of the colonization of unirradiated and irradiated SCID mice by human lymphoma and non-malignant lymphoid cells.

We have evaluated the severe combined immunodeficient (SCID) mouse as an in-vivo model for the study of non-Hodgkin's lymphomas (NHL). Characterization of the immune system of the animals in our SCID mouse colony was carried out to assess the numbers of lymphoid cells present, to determine natural killer (NK) cell activity as a function of age and to examine the histology of the lymphoid organs. In this study four human NHL established cell lines (Daudi, Namalwa, U937, MC116), lymphoma cells from four fresh NHL biopsies and normal peripheral blood mononuclear cells (PBMC) and bone marrow cells were investigated, after intraperitoneal injection into the mice. The presence of the human NHL cells in the peritoneum and spleen was assessed by FACS analysis. The colonization potential was investigated in a range of tissues by polymerase chain reaction (PCR) amplification of human repetitive sequences. These studies revealed clear differences in the abilities of the NHL cell types to colonize the SCID mice. Namalwa, Daudi and U937 cells demonstrated the highest efficiency of colonization and readily formed tumours, whereas MC116, the NHL biopsy cell populations and the non-malignant lymphoid cells showed little ability to survive and colonize other tissues in the SCID mice. Whole body irradiation of the SCID mice appeared to improve the survival of human PBMC, NHL biopsy cells and MC116 cells in the peritoneum, but had little effect on their colonization potential. The significance of these studies is discussed.

Gestational choriocarcinoma of the ovary diagnosed by analysis of tumour DNA.

Gestational choriocarcinoma of the ovary is a rare form of malignancy which can be difficult to distinguish from primary ovarian choriocarcinoma. The ability to make such a diagnosis could, however, have important implications for therapy. We report here a case of choriocarcinoma whose origins were difficult to determine and which behaved clinically more like a primary rather than a gestational choriocarcinoma. We have analysed DNA from this tumour by using polymerase chain reaction (PCR) amplification of a range of polymorphic alleles and have demonstrated that the tumour was in fact gestational. Furthermore, the lack of chromosome Y sequences and the presence of heterozygosity of the spouse's alleles, indicated that this tumour arose as a result of dispermic fertilisation of an empty ovum by sperm carrying the X chromosome.

Analysis of 14q+ derivative chromosomes in non-Hodgkin's lymphomas by fluorescence in-situ hybridization.

The most common chromosomal abnormality observed in non-Hodgkin's lymphomas (NHL) involves the structural alteration of the q arm of chromosome 14. It is not always possible, however, to fully analyse such derivative chromosomes by Giemsa-banding. Therefore, we have applied the fluorescence in-situ hybridization (FISH) technique of chromosome painting to elucidate the origins of the der(14) chromosomes in 8 cases of NHL. In 2 NHL the der(14) appeared to be the product of the t(14;18)(q32;q21) translocation, but were not accompanied by the reciprocal der(18) chromosome. In 3 cases the breakpoint was at 14q32 but the translocated material appeared not to be from chromosome 18 and in 2 cases the breakpoint was centromeric to 14q32. One case with a t(14;18)(q32;q21) was also analysed as a control. Dual painting was carried out with paints for chromosome 14 and either chromosome 3, 8, 10, 11, 18 or 19. In the control and 2 other cases the translocated material was demonstrated to be from chromosome 18, in two cases it was from chromosome 3 and in 1 case there was an unusual insertion of chromosome 11 material. We were unable to identify the origins of the translocated material in 1 NHL and in the final case the apparent der(14) was demonstrated not to contain chromosome 14 material. These data demonstrated the utility of the FISH technique for analysing malignant cell karyotypes, and in particular indicated the potential of this approach for identifying cases containing putative NHL associated oncogenes that may have been translocated adjacent to the immunoglobulin locus at 14q32.

Activation of MYCN in a case of non-Hodgkin's lymphoma.

We have examined a series of non-Hodgkin's lymphomas (NHL) for evidence of expression of the MYC gene family. Northern blot analysis of RNA samples derived from 11 non-malignant reactive lymphoid tissues and 33 NHL was used to investigate expression of MYC, MYCL and MYCN. As expected MYC expression was detected in all samples. The levels of MYC expression were quantified by densitometry and appeared to be 3-8 fold higher in high grade NHL than in the low grade NHL or non-malignant lymphoid tissue. No expression of MYCL was detected in any sample. Expression of MYCN was however observed in one sample, which had been diagnosed as a T-cell high grade NHL. A detailed cytogenetic analysis of this sample proved difficult to obtain but, by using fluorescence in-situ hybridization (FISH), we were able to demonstrate that on one of the chromosomes 2 the MYCN gene was localised to a translocation breakpoint region. It therefore appears that in NHL it is possible for MYCN, like MYC in Burkitt lymphoma, to be activated as a result of a chromosome translocation event.

MDR-1 expression in non-Hodgkin's lymphomas is unrelated to treatment intensity or response to therapy.

Over-expression of the MDR-1 gene, which codes for P-glycoprotein, is thought to be an important mechanism in the drug resistance exhibited by many tumours. A number of chemotherapeutic agents which induce MDR-1 expression are also components of combination chemotherapies that are used in the treatment of high grade non-Hodgkin's lymphomas (NHL). We have therefore examined expression of MDR-1 in a series of NHL by Northern blot analysis as well as investigated the localization of P-glycoprotein by immunohistochemistry. The series included 11 hyperplastic reactive nodes and tonsils, 17 low grade NHL and 15 high grade NHL. The levels of MDR-1 mRNA were quantified by scanning densitometry and comparison with levels of glucose-6-phosphate dehydrogenase (G6PD). The MDR-1 mRNA was observed in both non-malignant and NHL tissues. Immunohistochemical staining revealed that expression of MDR-1 mRNA in reactive nodes was related to the presence of P-glycoprotein in lymphocytes, however, P-glycoprotein was apparent in both the reactive lymphocytes and tumour cells in the NHL samples. Elevated mRNA levels (2-3 fold increase) were observed in some low grade and high grade NHL relative to those observed in reactive lymphoid tissue. There appeared to be little correlation, however, between expression of the MDR-1 gene and either treatment intensity or response to therapy. The drug resistance that is often encountered in NHL patients is therefore likely to involve mechanisms other than over-expression of P-glycoprotein.

Elevated levels of MDM-2 and p53 expression are associated with high grade non-Hodgkin's lymphomas.

The role of p53 in the evolution of non-Hodgkin's lymphomas (NHL) is unclear. Mutations of the p53 gene appear to be relatively uncommon but stabilized p53 protein, as detected by immunohistochemistry, has indicated a more frequent involvement of p53. As dysfunction of p53 protein has also been suggested to occur after overexpression of the mdm-2 protein, we have therefore investigated a series of non-malignant hyperplastic reactive lymphoid tissues and NHL to examine whether the levels of expression of MDM-2 correlated to positivity of p53 protein staining. Northern blot analysis of MDM-2 expression was compared to glucose-6-phosphate dehydrogenase (G6PD) expression by densitometry to quantify the relative levels of MDM-2 expression. Consistent low levels of MDM-2 expression were observed in non-malignant lymphoid tissue and in low grade NHL, however, 13/15 high grade NHL exhibited a 2-15-fold increase in MDM-2 expression. Interestingly similar elevations in p53 mRNA expression were also observed in 6/15 high grade NHL. Positive staining of the p53 protein did not, however, correlate with elevated mRNA levels of either MDM-2 or p53. The significance of these observations is discussed.

2D-gel analysis of protein synthesis profiles of different stages of chronic lymphocytic leukaemia.

Protein synthesis profiles of leukaemic cells from 15 chronic lymphocytic leukaemia (CLL) patients were analysed by 2D-electrophoresis of 35S-methionine labelled proteins. This series of CLL included patients with stage A (7), B (4) and C (4) disease. Although the protein synthesis profiles were similar in all cases, some consistent differences were noted between the different stages. The levels of synthesis of three proteins (approximately 35 kD size) were of particular interest. Two of these were always expressed in stage C CLL but either infrequently or not at all in stage A or B CLL. By contrast a third protein was expressed at a much reduced level in stage C compared to stages A or B. This type of analysis could prove invaluable for identifying proteins whose expression was intimately associated with the evolution of CLL.