p53, mutation frequency and apoptosis in the murine small ... - Nature

1 downloads 0 Views 300KB Size Report
p53, mutation frequency and apoptosis in the murine small intestine. Alan R Clarke1, Louise A Howard2, David J Harrison1 and Douglas J Winton2.
Oncogene (1997) 14, 2015 ± 2018  1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00

p53, mutation frequency and apoptosis in the murine small intestine Alan R Clarke1, Louise A Howard2, David J Harrison1 and Douglas J Winton2 1

CRC Laboratories, Department of Pathology, University of Edinburgh, Teviot Place, Scotland, EH8 9AG; 2CRC Human Cancer Genetics Research Group, University of Cambridge, Addenbrookes Hospital, Box 238, Hills Road, Cambridge CB2 2QQ, UK

Normal function of the p53 gene is integral to the cellular response to genotoxic stress. One prediction arising from this is that p53 de®ciency results in an increased mutation frequency. However, limited evidence has been produced in support of this idea. In order to further investigate the in vivo role of p53 in surveillance against mutation, and particularly to address the signi®cance of p53-dependent apoptosis, we scored mutation frequency at the Dlb-1 locus within cells of the intestinal epithelium of animals which were wild type, heterozygous or null for p53 and heterozygous (a/b) at the Dlb-1 locus. Using this assay we have shown that loss of a p53-dependent apoptotic pathway is associated with the detectable acquisition of mutations, but only at high levels of DNA damage. These results question the signi®cance of the immediate `wave' of p53-dependent apoptosis seen in this tissue, particularly as there was a delayed p53-independent apoptotic pathway. We conclude that loss of p53 function only becomes relevant to the in vivo acquisition of mutations and thus tumorigenesis in certain circumstances. Keywords: p53; mutation; apoptosis

Introduction Murine strains null for p53 have been used to analyse the link between DNA-damage and apoptosis. This work has shown that, following exposure to a range of genotoxic stress, p53 de®cient cells are not removed by the normal process of apoptotic induction (Clarke et al., 1993, 1994; Lowe et al., 1993; Merritt et al., 1994). The interpretation of this ®nding has been that a p53 de®cient environment results in increased survival of cells bearing DNA damage, thereby leading to an increased mutation frequency and ultimately predisposing to malignancy (Wyllie et al., 1994). The role of p53 is not, however, restricted to apoptosis, as p53 has also been shown to link DNA damage to both G1 and G2 arrest (Argarwal et al., 1995), to regulate the response of several di€erent DNA repair genes (Wang et al., 1995; Ra€erty et al., 1996), and also to directly interact with DNA (Mummenbrauer et al., 1996). The relative in vivo importance of these mechanisms is as yet unclear. However, regardless of mechanism, the prediction remains that p53 de®ciency should lead to an increased mutation frequency, either by failure to permit repair of damaged DNA or by failed deletion of mutation bearing cells. Correspondence: AR Clarke Received 18 November 1996; revised 14 January 1997; accepted 15 January 1997

Experiments designed to test this prediction suggest that p53 dependent di€erences may only become apparent following exogenous DNA damage. Two di€erent groups have used a transgene target to monitor mutation frequency, but in both cases no p53 dependent di€erence was observed in the spontaneous mutation frequency (Nishino et al., 1995; Sands et al., 1995). In contrast, experiments using either short term pre-B cell cultures (Griths et al., 1997) or a transfected ®broblast line (Yuan et al., 1995) have now provided data in support of a p53-dependent increase in the number of mutation bearing cells following genotoxic stress. Many of the phenotypic consequences of p53 de®ciency are now well documented. p53 null mice are predisposed towards spontaneous tumorigenesis (Donehower et al., 1992; Purdie et al., 1994; Jacks et al., 1994), genomic instability (Bou‚er et al., 1995; Lee et al., 1994) and also to a high level of developmental abnormality (Sah et al., 1995; Armstrong et al., 1995). The in vivo signi®cance of p53 de®ciency to tumorigenesis has been further underlined by experiments intercrossing the p53 mutant animals with a variety of other strains, either transgenic for oncogenes or mutant for other tumour suppressor genes. For example, Wnt-1 transgenic mice show more rapid mammary tumour development on a p53 null background (Donehower et al., 1995); lymphomagenesis is enhanced in mice bearing a CD2-myc transgene (Elson et al., 1995; Blyth et al., 1995); and a role for p53 in two types of pancreatic tumorigenesis has been revealed following intercrossing to mice mutant for either Rb-1 (Williams et al., 1994) or Apc (Clarke et al., 1995). These experiments clearly demonstrate that a p53 null environment predisposes to tumorigenesis, but they have also produced several apparently anomalous results. For example, why does a p53 null environment promote lymphomagenesis but not mammary carcinogenesis in the MMTV/c-myc transgene strain (Elson et al., 1995)? Why is there no increase in intestinal adenoma formation in Apc heterozygotes on a p53 null background (Clarke et al., 1995)? These results suggest that factors such as cell type and the level and type of damage may determine what in¯uence, if any, p53 de®ciency has upon neoplastic development. In order to address some of these questions, we have used p53 null mice to determine the in vivo mutation frequency at the Dlb-1 locus following exposure to girradiation. We have chosen to use the Dlb-1 assay as this will permit an in vivo assessment of any genetic change (including both point mutations and deletions) leading to dysfunction at the Dlb-1 locus. We have performed this study in one tissue type, the small intestine, for which we have also extended our previous analysis of apoptotic induction.

p53 and mutation AR Clarke et al

2016

Results The Dlb-1 assay is a speci®c locus assay in which somatic mutations a€ecting a polymorphic genetic locus, Dlb-1, determining a lectin binding site are detected in mouse intestinal epithelium. The assay is autosomal with the locus located on mouse chromosome 11. The Dlb-1b allele speci®es binding of the lectin Dolichos bi¯orus agglutinin to intestinal epithelium while the Dlb-1a allele speci®es binding to vascular endothelium. In heterozygous Dlb-1b/Dlb-1a mice any inactivating mutation of the single b allele results in the a€ected cell losing the ability to bind the lectin. Following cell proliferation and clonal expansion mutations in the intestinal stem cell population result in the formation of a clone which does not bind a peroxidase conjugate of DBA. These unstained clones are easily identi®ed and quanti®ed in stained wholemount preparations of the small intestine (Winton et al., 1988). Following exposure to g-irradiation, Dlb-1 mutations were scored as non lectin binding clones along the length of the villus using standard techniques 20 days after exposure (Winton et al., 1988). Mutation frequencies following increasing doses of g-irradiation are given in Figure 1. p53 status did not a€ect the spontaneous mutation rate, con®rming for an endogenous locus the ®ndings of two independent experiments using an exogenous lacI target transgene in addition to an inactivating p53 mutation (Nishino et al., 1995; Sands et al., 1995). A p53 dependent di€erence in mutation frequency was seen following exposure to g-radiation, but only at the highest dose used, 6 Gy. At this dose, the mutation rate at the Dlb-1 locus was elevated 5 ± 6fold in p537/7 animals compared to that in p53 wild type animals. The same preparations were used to investigate whether p53 status a€ected the width of mutant clones to allow, if necessary, a correction in mutation frequency arising from di€erential survival of the target stem cells in wild type and p537/7 animals. However, no such di€erence was apparent. Mean clone widths were as follows: in +/+ mice, 6.09 (s.e. 0.66, n=41); in 7/7 mice, 6.47 (s.e.=0.68, n=44). To further investigate the relationship between the incidence of apoptosis and mutation in the intestine, we extended our previous analysis of apoptosis following g-irradiation. Four hours following 5 Gy whole body irradiation, p53-dependent apoptosis was observed, comparable to that previously reported (Clarke et al., 1994; Potten, 1990) (Figure 2a). However, p53-independent apoptosis was also observed in p537/7 samples beyond 4 h. Levels of apoptosis in the 7/7 samples were maximal 36 h following exposure, at this point being approximately half the highest level observed in the +/+ samples. To investigate whether the occurrence of delayed, p53-independent apoptosis was a general feature of cells which normally engage p53-dependent apoptosis, we scored cell death in a number of other cell types. Mantle zone lymphocytes of the spleen di€ered from intestinal cells in that there was only a small p53 independent increase, which was at no point greater than 11% of that observed in the wild type samples (Figure 2b). Other tissues did not show strong

induction of apoptosis over basal levels (Table 1) in response to g-irradiation. Some tissue types, notably the liver and kidney cortex, showed no evidence of apoptosis.

Figure 1 Dlb-1 Mutation frequencies following irradiation. Mutation frequencies were determined from intestinal wholemounts prepared 20 days after treatment as described in Materials and methods. Columns represent means at the radiation dosage shown. Black, p53 wild type; dark crosshatch, p53 heterozygotes; grey p53 nulls; light crosshatch p53 wild type historical controls. The number of animals per group is as follows: wild type; 0 rads (7), 200 rads (4), 400 rads (6), 600 rads (4): heterozygotes; 0 rads (14), 200 rads (10), 400 rads (7), 600 rads (5): homozygotes; 0 rads (9), 200 rads (7), 400 rads (10), 600 rads (10). Standard errors are shown. Analysis of variance was applied to groups receiving the same radiation dose. An overall di€erence was only observed between groups receiving 600 rads. The student t test was then used to compare wild type and heterozygote groups to p53 nulls at this dose and were found to be signi®cantly di€erent (P50.001, t values 4.92 and 5.09 for wild type and heterozygotes respectively, degree of freedom 10)

Figure 2 Apoptosis following irradiation. (a) Apoptosis scored (as described in Materials and methods) within the crypts of the small intestine of p53+/+ (~) and p537/7 (*) animals at a series of time points up to 60 h following whole body exposure to 4 Gy g-irradiation. (b) Apoptosis of T cells within the mantle zone of the spleens of the same animals from which the intestinal data was derived. ~ p53+/+; * p537/7

p53 and mutation AR Clarke et al

2017

Table 1 Apoptosis in different tissues p53 Status +/+ +/+ +/+ +/+ +/+ +/+ +/+ 7/7 7/7 7/7 7/7 7/7 7/7 7/7

Tissue Pancreas Lung Kidney Liver Pancreas Lung Kidney Liver

Cell type Bronchus Alveoli Glomerulus Cortex Medulla Islet Cells Bronchus Alveoli Glomerulus Cortex Medulla

Mock irradiated

0

4

0.1 0 0 0 0 0 0 0 0.16 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0.25 0

0.25 0.35 0 0 0 1.0 0 0 0 0 0 0 0 0

Time (h) following 5 Gy g-irradiation 8 24 36 0.25 0.3 0 0 0 1.2 0 0 0.6 0 0 0 0.7 0

0.5 0.2 0 0 0 1.2 0 0 0.4 0 0 0 0 0

0 0 0 0 0 0.2 0 0 0 0 0 0 0 0

48

60

0.1 0.9 0.4 0.3 0 0 0 0.1 nd nd 0 0 0 0

0.1 0.2 0 0.2 0 0.4 0 0 nd nd 0 0 0 0

Apoptosis scored as described and expressed as a percentage of the total number of cells counted. nd; not done

Discussion We have shown that p53 de®ciency leads to an increase in mutation frequency at a marker allele, Dlb-1, but only at high levels of genotoxic stress. The inference from this is that loss of p53 function may not predispose to spontaneous tumorigenesis in the murine intestine; a ®nding supported by data from the Min mouse model, where p53 de®ciency fails to lead to an increased adenoma burden, as might be predicted, but rather to pancreatic neoplasia (Clark et al., 1995). These observations are in contrast to other tissue types, where germline loss of p53 predisposes to spontaneous tumorigenesis in both human (Malkin et al., 1990) (predominated by breast cancer) and mouse (Purdie et al., 1994; Jacks et al., 1994) (predominated by lymphoma). Clearly, in these tissues, p53 plays a signi®cant role at levels of DNA damage arising spontaneously. It is possible that such tissue speci®c di€erences arise as a consequence of di€erent apoptotic responses. We have demonstrated a highly tissue speci®c pattern of both p53 dependent and independent apoptosis, with the latter delayed and reduced by comparison to the p53-dependent response. The murine intestine was the only tissue studied to show a clear p53-independent wave of apoptosis, perhaps sucient to render any p53-dependent response redundant at spontaneous and low levels of DNA damage. It is clear that a combination of factors (including di€erent genetic requirements for neoplastic change and di€erent repair capacities) contribute to produce tissue speci®c predisposition to malignancy. We now propose that such predisposition is further in¯uenced by di€erences in the patterns of reliance upon p53 dependent and independent pathways. However, the exact relevance of such pathways to di€erent patterns of malignancy remains to be determined. For example, we have shown here that the ease by which a cell type engages p53 dependent apoptosis is not a good indicator of the role of that pathway in preventing the acquisition of mutations. In conclusion, p53-de®ciency leads to an increase in the mutation rate at a marker allele, the Dlb-1 locus, from which we infer that mutations within other genes critical for tumorigenesis are more likely to occur in a

p53 null background. However, it is only at the highest dose of 6 Gy that a protective e€ect due to p53 is observed. We have also demonstrated the presence of p53-independent mechanisms of apoptosis, exempli®ed by the delayed induction of apoptosis, and highlighted the tissue speci®c nature of both this and the p53 dependent response. Taken together, these ®ndings show that a variety of factors modulate the in¯uence of p53 pathways upon mutation rate, and that the ability of p53 to act as a `guardian of the genome' is strongly cell type speci®c. Materials and methods Mice p53 mutant animals were from an outbred colony segregating for Ola/129 and SWR genomes. Experimental cohorts were derived by crossing the p53 mutants to a congenic B1/6 strain homozygous for the Dlb a allele, and subsequently intercrossing these progeny from this cross to generate mice heterozygous at the Dlb allele and segregating for all possible p53 genotypes. Dlb-1 Mutation assay This was performed as previously described (Winton et al., 1988) and results are presented as number of mutations per 104 villi. The absolute mutation frequency can be calculated as follows: the values given in Table 1 indicate that the clones are on average around four cells wide (one cell=1.5 graticule units=7.5 mm). From the published villus row count of 75 cell rows at the 25% point of the length of the small intestine it can be calculated that there are on average 300 cells/mutant clone, equivalent to a production rate of 195 cells/mutant clone/day based on a villus transit time of 37 h (Wright and Nafussi, 1982). As the crypt cell production rate is 394 cells/day this represents 195/3956100% of the daily output of the crypt. Therefore at 6 Gy the average mutant clone corresponds to around 50% of crypt output implying that there are two surviving stem cells/crypt. With a crypt to villus ratio of 10 : 1 (reference 20), this allows a calculation of the Dlb-1 mutant frequency in cellular terms of 1.16104 and 6.16104 for wild type and p537/7 mice respectively. Ribbon width measurement Clone widths were measured by dissecting individual small intestinal villi containing unstained, vertical oriented

p53 and mutation AR Clarke et al

2018

ribbons from stained gut wholemounts using watchmaker forceps. Dissected villi placed in a drop of glycerol on a microscope slide and immobilised using a cover slip. Ribbon widths were measured microscopically using an eyepiece graticule with a 0 ± 100 arbitrary scale and a 625 objective. The graticule scale at this magni®cation was calibrated using a slide containing a 1 mm scale subdivided into 100 mm divisions. Each graticule unit was calculated to be 5 mm. Ribbons were dissected from the same region of the wholemount as used for mutation frequency determination in which villi and ribbons were mainly of constant width along their length. Some clones arising from a common centre gave rise to ribbons on two villi or to two parallel ribbons on a single villus. In these cases the values of the individual ribbons were totalled. Occasionally ribbons were not of constant width along their length and in this case the value obtained was taken to be mid way between the widest and narrowest measurements. g-irradiation In order to determine the pattern of induction of apoptosis following exposure to 5 Gy ionising irradiation, animals

were exposed to 5 Gy from a Cs 137 source. At various time points after irradiation, animals were killed and subject to full necropsy. Two mice were sampled for each time point. All tissues were ®xed in bu€ered formol saline except for small intestine, which was mounted en face and ®xed overnight in methacarn as previously described (Clarke et al., 1994). Apoptosis was scored visually in haematoxylin and eosin stained sections. A minimum of 50 half crypts were scored. Figures for these values represent the average apoptotic index (number of apoptotic cells as a percentage of the total number of cells) per half crypt. For all other tissue types a minimum of 250 cells were scored per tissue section, although where the cell type population was high (liver and spleen), this number was increased to a minimum of 1000.

Acknowledgements This work was supported by the Cancer Research Campaign. ARC is a Royal Society University Research Fellow. We are grateful to Professors AH Wyllie and BAJ Ponder for many useful discussions.

References Agarwal ML, Agarwal A, Taylor WR and Stark GR. (1995). Proc. Natl. Acad. Sci. USA, 92, 8493 ± 8497. Armstrong JF, Kaufman M, Harrison DJ and Clarke AR. (1995). Curr. Biol., 5, 931 ± 936. Blyth K, Terry A, O'Hara M, Baxter EW, Campbell M, Stewart M, Donehower LA, Onions DE, Neil JC and Cameron ER. (1995). Oncogene, 10, 1717 ± 1723. Bou‚er SD, Kemp CJ, Balmain A and Cox R. (1995). Cancer Res., 55, 3883 ± 3889. Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC, Hooper ML and Wyllie AH. (1993). Nature, 362, 849 ± 852. Clarke AR, Gledhill S, Hooper ML, Bird CC and Wyllie AH. (1994). Oncogene, 9, 1767 ± 1773. Clarke AR, Cummings MC and Harrison DJ. (1995). Oncogene, 11, 1913 ± 1920. Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA, Butel JS and Bradley A. (1992). Nature, 356, 215 ± 221. Donehower LA, Godley LA, Aldaz CM, Pyle R, Shi YP, Pinkel D, Gray J, Bradley A, Medina D and Varmus HE. (1995). Genes Dev., 9, 882 ± 895. Elson A, Deng C, Campos-Torres J, Donehower LA and Leder P. (1995). Oncogene, 11, 181 ± 190. Griths SD, Clarke AR, Healy LE, Ford AM, Hooper ML, Wyllie AH and Greaves MF et al. (1997). Oncogene, 14, 523 ± 531. Jacks T, Remington L, Williams BO, Schmitt EM, Halamachi S, Bronson RT and Weinberg RA. (1994). Curr. Biol., 4, 1 ± 7.4. Lee MJ, Abrahamson JL, Kandel R, Donehower LA and Bernstein A. (1994). Oncogene, 9, 3731 ± 3736. Lowe SW, Schmitt ES, Smith SW, Osborne BA and Jacks T. (1993). Nature, 362, 847 ± 849. Malkin D, Li FP, Strong LC, Fraumeni JF, Nelson CE, Kim DH, Kassel J, Gryka MA, Bischo€ FZ, Tainsky MA and Friend SH. (1990). Science, 250, 1233 ± 1238.

Merritt AJ, Potten CS, Kemp CJ, Hickman JA, Balmain A, Lane DP and Hall PA. (1994). Cancer Res., 54, 614 ± 617. Mummenbrauer T, Janus F, Muller B, Wiesmuller I, Deppert W and Grosse F. (1996). Cell, 85, 1089 ± 1099. Nishino H, Knoll A, Buettner VL, Frisk CS, Maruta Y, Haavik J and Sommer SS. (1995). Oncogene, 11, 263 ± 270. Potten CS. (1990). Int. J. Radiat. Biol., 58, 925 ± 973. Purdie CA, Harrison DJ, Peter A, Dobbie L, White S, Howie SEM, Salter D, Bird CC, Wyllie AH, Hooper ML and Clarke AR. (1994). Oncogene, 9, 603 ± 609. Ra€erty JA, Clarke AR, Sellapan D, Koref MS, Frayling IM and Margison GP. (1996). Oncogene, 12, 693 ± 697. Sah VP, Attardi LD, Mulligan GJ, Williams BO, Bronson RT and Jacks T. (1995). Nature Genetics, 10, 175 ± 180. Sands AT, Suraokar MB, Sanchez A, Marth JE, Donehower LA and Bradley A. (1995). Proc. Natl. Acad. Sci. USA, 92, 8517 ± 8521. Wang XW, Yeh H, Schae€er L, Roy R, Moncollin V, Egly JM, Wang Z, Freidberg EC, Evans MK, Ta€e BG et al. (1995). Nature Genetics, 10, 188 ± 195. Williams BO, Remington L, Albert DM, Mukai S, Bronson RT and Jacks T. (1994). Nat. Genet., 7, 480 ± 484. Winton DJ, Blount MA and Ponder BAJ. (1988). Nature, 333, 463 ± 466. Wright NA and Nafussi A. (1982). Cell and Tissue Kinetics, 15, 611 ± 621. Wyllie AH, Carder PJ, Clarke AR, Cripps KJ, Gledhill S, Harrison DJ, Hooper ML, Morris RG, Purdie CA and Bird CC. (1994). Cold Spring Harbor Symposium LIX: Molecular Genetics of Cancer, pp. 403 ± 409. Yuan J, Yeasky TM, Harve PA and Glazer PM. (1995). Carcinogenesis, 16, 2295 ± 2300.