and AT III RFLPs by Southern blotting and by agarose gel electrophoresis of ... Acknowledgements. The authors wish to thank Professor John Wyke for critical.
Oncogene (1997) 14, 1955 ± 1964 1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
Evidence for the inactivation of multiple replicative lifespan genes in immortal human squamous cell carcinoma keratinocytes O Loughran1, LJ Clark1, J Bond2, A Baker1, IJ Berry1, KG Edington1, I-S Ly1, R Simmons1, R Haw1, DM Black1, RF Newbold1 and EK Parkinson1 1
CRC Beatson Laboratories, Beatson Institute for Cancer Research, Garscube Estate, Switchback Road, Bearsden, Glasgow G61 1BD, Scotland; 2Departments of Biology and Biochemistry, Brunel University, Uxbridge, Middlesex, UK
Human keratinocyte immortality is genetically recessive to the normal phenotype of limited replicative lifespan and appears to require the dysfunction of p53 and the cyclin D-Cdk inhibitor p16. In order to test for the inactivation of other candidate replicative lifespan genes in the immortal cells of human tumors, we developed a series of mortal and immortal keratinocyte cultures derived from neoplastic lesions of the head and neck which were amenable to molecular genetic analysis by the loss of heterozygosity (LOH) technique. The results indicate that keratinocyte immortalization in head and neck squamous cell carcinoma (SCC-HN) development involves the inactivation of at least two further pathways to senescence and four in all. Chromosomes 1, 4 and 7 carry genes representing immortality complementation groups C, B and D respectively and immortal keratinocytes showed LOH at either 4q32-q34 between D4S1554 and D4S171 (group B) or 7q31 (group D) but never 1q25 (group C). These results tentatively suggest that the genes responsible for the immortality complementation groups encode proteins on the same pathway to senescence. In addition, all of the immortal keratinocyte lines possessed high levels of telomerase activity and a suppressor of telomerase activity has been mapped to the short arm of chromosome 3p. Five out of eight lines showed LOH at 3p21.2-p21.3, a region which may carry a gene capable of suppressing SCC-HN telomerase. However, alternative mechanisms of telomerase reactivation were also suggested by our results. None of the above genetic alterations were seen in seven senescent neoplastic keratinocyte cultures. Other loci harbouring antiproliferative genes implicated in replicative lifespan showed few or no alterations and any alterations seen were additional to those described above. Keywords: keratinocyte; senescence; immortality; telomerase; squamous; cancer
Introduction Human cells from a number of lineages including keratinocytes are known to exhibit a limited proliferative capacity in vitro (Hay¯ick, 1965; Rheinwald and Green, 1975) but in contrast, many human tumors, including squamous cell carcinoma of the head and neck (SCC-HN), contain variant cells which
Correspondence: EK Parkinson Received 1 August 1996; revised 10 January 1997; accepted 10 January 1997
proliferate essentially inde®nitely and are designated immortal (Rheinwald and Beckett, 1981; Easty et al., 1981). There is persuasive evidence from the induction of immortality in human ®broblasts that cellular immortality involves the abrogation of both the p53 and RB-1 tumor suppressor gene pathways (Shay et al., 1991; Hara et al., 1991) and decreased regulation of the cell cycle in both G1 and G2 (Kaufmann et al., 1995). Although these events are necessary to induce immortality (Wright et al., 1989; Radna et al., 1989) they are still insucient and a further genetic alteration which activates telomerase (Counter et al., 1992, 1994) is usually essential. Several human chromosomes have been shown to harbour antiproliferative genes which have been mapped to 1q25 (Karlsson et al., 1996), 1q42 (Karlsson et al., 1996), 2 (Uejima et al., 1995), 3p21 (Rimessi et al., 1994), 4q32-q34 (Ning et al., 1991; O Pereira-Smith ± personal communication), 6q14-q21 (Sandhu et al., 1996), 6q21-qter (Sandhu et al., 1994), 7q31-32 (Ogata et al., 1993), 11p15 (Koi et al., 1993), 17 (Casey et al., 1993), 18 (Sasaki et al., 1994) and Xp11-ter (Wang et al., 1992) respectively. Genetic complementation groups for the immortal phenotype have been identi®ed (Pereira-Smith and Smith, 1988) and although the complementation group A locus is unknown, genes assigned to chromosomes 4 (Ning et al., 1991), 1 (Hensler et al., 1994) and 7 (Ogata et al., 1993) are thought to represent complementation groups B, C and D respectively. The chromosome 1 gene has been mapped to either the p arm or to 1q13q31 (Vojta et al., 1996), making the antiproliferative gene at 1q25 (Karlsson et al., 1996) a likely candidate. However, other studies indicate that more than one chromosome can limit replicative lifespan in the same immortal cell line, suggesting multiple pathways to cellular senescence (Sasaki et al., 1994) and some of these may not be related to the previously identi®ed complementation groups. Of the loci described above only those representing the complementation groups and carried by chromosomes 1, 4 (Dimri et al., 1995) and 7 (unpublished data) have formally been shown to reactivate speci®c markers of replicative senescence (Dimri et al., 1995) in the appropriate target cells and only chromosome 3p has so far been reported to harbour a gene capable of suppressing telomerase (Ohmura et al., 1995). Thus a de®nite link between the other loci and immortalization is less certain. Furthermore, only two studies (Ogata et al., 1993; Sandhu et al., 1994) have provided any molecular genetic evidence that the recipient cells harboured alterations in the relevant chromosomal
Genetic analysis of squamous cancer O Loughran et al
1956
regions, making it dicult to distinguish between gene dosage and gene replacement eects in many of the published monochromosome transfer experiments. In contrast to in vitro models, far less is known of the cellular immortalization process in human tumor cells and in particular, whether the regions of human chromosomes thought to carry replicative lifespan genes show molecular defects such as loss of heterozygosity (LOH) which would be suggestive of a recessive mutation at the locus (Cavenee et al., 1983; Vogelstein et al., 1989). Squamous cell carcinoma is an excellent system in which to address this question, since normal and malignant keratinocytes can be cultured under essentially the same conditons (Rheinwald and Beckett, 1981; Edington et al., 1995) and animal models of this tumor are available (Balmain and Brown, 1988). Our previous work has shown that immortal keratinocytes arise late during the progression of squamous cell carcinoma of the head and neck (SCC-HN, Edington et al., 1995), that the immortal phenotype is recessive to the phenomenon of limited replicative lifespan (Berry et al., 1994) and involves a high frequency of allele loss (Edington et al., 1995). This suggests that the process involves one or more recessive inactivating mutations. Two candidates are the p53 and the CDKN2A/p16ink4A genes which are both inactivated or deleted in immortal but not usually in replicatively senescent neoplastic keratinocytes (Burns et al., 1993, 1994; Edington et al., 1995; Loughran et al., 1994, 1996) and we have suggested that the loss of these two suppressor genes in concert is a necessary but insucient condition to sustain the immortal keratinocyte phenotype in human SCC-HN (Loughran et al., 1994, 1996). The aim of the present study was to use the classical molecular genetic approach of LOH (Cavenee et al., 1983; Vogelstein et al., 1989) to investigate the possible inactivation of any of the other, as yet uncloned, replicative lifespan genes in the pathogenesis of immortal malignant keratinocytes. Results Several neoplastic keratinocyte cultures of ®nite or inde®nite proliferative life span were analysed for genetic alterations on chromosomes 1, 3, 4, 6, 7, 11, 17, 18 and X by the LOH technique. We are con®dent that the changes observed did not occur in vitro because we used a culture system where malignant keratinocytes remain phenotypically stable for more than 200 population doublings (Rheinwald and Beckett, 1981) and we always perform genetic analysis on the cultures when they have completed 30 ± 50 population doublings without any observed slowed proliferation or crisis (Loughran et al., 1994; Edington et al., 1995). Furthermore, we tested lines BICR 3 and BICR 63 for LOH at several loci on chromosomes 1, 4, 6 and 7 at 30 population doublings and again at 130 population doublings and found them still to be heterozygous (data not shown), thus illustrating the genetic stability of the lines during this period. The allele losses were generally restricted to small regions of the candidate chromosomes and were often interstitial, suggesting that they were relevent to the observed phenotypes.
The results show that alterations at the complementation group loci are mutually exclusive (Table 1), since seven immortal lines displayed LOH on 4q (complementation group B) but not 7q (complementation group D) and lines BICR 7 and BICR 68 showed LOH on 7q but not 4q. No lines showed LOH at 1q25. However, LOH was also seen at other loci thought to harbour replicative lifespan genes not representing complementation groups. In particular, three regions of chromosome 3p showed frequent LOH and this is interesting given the recent evidence that a repressor of telomerase locates to this region (Ohmura et al., 1995). None of the senescent neoplastic keratinocyte cultures showed LOH at any of the above loci. The idiograms of the three complementation group chromosomes and chromosomes 3 are described in the sections below. Chromosomes 1, 4 and 7 (Complementation Groups C, B and D) LOH was very common at a locus in distal 4q which is thought to carry a gene capable of inducing replicative senescence in a variety of human tumor cells and to represent complementation group B for immortality (M Bertram and O Pereira-Smith ± personal communication, Table 1). An idiogram of chromosome 4 is shown in Figure 1a and it can be seen that while LOH is not generally common on this chromosome, LOH was detected in 6/11 lines distal to D4S1554 and proximal to D4S171 (see BICR 19 and BICR 22). The maximum LOH for the lines showing LOH on chromosome 4 occurred at D4S408/D4S1535 (4q32-q34) where 6/8 (75%) of informative cases showed loss. LOH was also found at D4S194 and D4S1549 (4q26) where 3/4 and 2/ 3 cases showed loss respectively, but there is currently no evidence for a senescence gene at this locus. Table 1 Molecular genetic alterations at loci harbouring replicative lifespan genes in neoplastic head and neck keratinocytes Line
TNM Stage
(A) Non-senescent lines BICR 3 T2 N0 M0 BICR 63 T2 N2B M0 BICR 19 N/Aa BICR 68 T4 N0 M0 T4 N1 M0 BICR 6 BICR 56 T4 N1 M0 BICR 78 T4 N1 M0 BICR 7b T4 N2B M0 BICR 31 T4 N2B M0 BICR 18 Lymph Node Metastasis (T4 N1 M0) BICR 22 Lymph Node Metastasis (T4 N3 M0) (B) Senescent lines BICR E1 Carcinoma in situ BICR E2 Carcinoma in situ BICR E4 Carcinoma in situ BICR E5 Severe Dysplasia BICR 66 T2 N0 M0 BICR 80 T4 N2C M0 BICR 37 Lymph Node Metastasis (T4 N2C M0)
Complementation Group Loci B C D 4q32-q34 1q25 7q31 * * * * * * * * *
* * * * * * *
* * * * * * * * *
*
*
*
*
*
*
*
* * * * *
* * * * *
* * * * *
* Loss of heterozygosity; * Retention of heterozygosity; Not informative; a Not applicable, epidermal tumor; b Line enters crisis at low density
Genetic analysis of squamous cancer O Loughran et al
Examples of allele loss and retention at D4S408 are shown in Figure 1b. We are continuing to re®ne our mapping of this region to assist in the identi®cation of the candidate replicative lifespan gene in this region. Two regions of chromosome 1, 1q25 and 1q42, have been shown to harbour antiproliferative activity by monochromosome transfer (Vojta et al., 1996) and somatic cell genetics (Karlsson et al., 1996) and the 1q13-q31 region is thought to represent the complementation group C for human cellular immortality (Vojta et al., 1996) making the 1q25 region the best candidate for the complementation group C locus. No LOH was seen on any part of chromosome 1 (Figure 2) except for one microsatellite at 1q21 in line BICR 63 and the 1q42-qter region in lines BICR 18 and BICR 31 both of which were derived from very
advanced tumors. Examples of allele loss and retention at D1S245 are shown in Figure 2b. If the complementation group C locus turns out to be nearer to 1q21 than 1q25 then BICR 63 might eventually be assigned to complementation group C since this line is unlikely to be group B or D. However, the evidence presently favours 1q25 as the complementation group C locus (Karlsson et al., 1996; Vojta et al., 1996). Chromosome 7 carries a replicative senescence gene representing the complementation group D (Ogata et al., 1993) and an idiogram of LOH on chromosome 7 (Figure 3a) illustrates a speci®c region of LOH at 7q31q32 in two non-senescent SCC-HN lines BICR 7 and BICR 68. Examples of these losses are shown in Figure 3b. LOH at 4q32-q34 and 7q31 was mutually exclusive
a
b 7 N
18 T
N
22 T
N
31 T
N
68 T
N
78 T
N
T
D4S408 Figure 1 Genetic alterations on chromosome 4 in mortal and immortal neoplastic human keratinocytes. (a) Idiogram of human chromosome 4 showing a minimal region of LOH involving D4S1535 and D4S408. Note the retention of heterozygosity at D4S1554 and D4S171 but not D4S1535. (b) Autoradiograms of the microsatellite D4S408, showing examples of allele loss and retention Open symbols=Retention of heterozygosity; Solid symbols=Loss of heterozygosity; Hatched symbols=Not informative. +=Line senesced; *=Line entered crisis. N=Normal cells T=Tumor cells. The numbers above the columns and lanes correspond to the relevant BICR lines
1957
Genetic analysis of squamous cancer O Loughran et al
1958
and no line showed LOH at more than one complementation group locus. These results tentatively suggest that the genes mutated in the complementation group described by Pereira-Smith and Smith (1988) may encode products on the same molecular pathway to senescence. However, as many other human chromosomes have now been shown to harbour antiproliferative activity we tested these for their involvement in the immortal SCC-HN keratinocyte phenotype and the results are described below. Chromosome 3 An antiproliferative gene for ovarian carcinoma has been mapped to chromosome 3p21.1-p21.2 (Rimessi et al., 1994) and recently a suppressor of telomerase activity has been mapped to 3p (Ohmura et al., 1995). Our own preliminary data from monochromosome transfer experiments suggests that there is a suppressor of immortal keratinocyte telomerase activity which
maps to microsatellites D3S1478 and D3S1076 at 3p21.2-p21.3. (A Cuthbert, J Bond, EK Parkinson and RF Newbold ± unpublished data). Examination of the LOH (Figure 4) and telomerase data (Table 2) from all our cell cultures shows that ®ve out of eight informative telomerase-positive cultures showed LOH at D3S1478 and/or D3S1076. None of our senescent telomerase-negative cultures showed LOH at these loci (Table 2). Furthermore, culture BICR 7 which appears to be an example of a cell line in crisis (Edington et al., 1995), was telomerase-negative at early passage and retained heterozygosity at D3S1478 (Table 2, Figure 4a). Even at later passage at high density BICR 7 replicated poorly and possessed less than 20% of the telomerase activity of the immortal lines. The region 3p13-cen also commonly showed LOH in the nonsenescent head and neck cultures with eight out of ten informative cases showing LOH, the exceptions being BICR 6 and BICR 68. The 3p25-pter region also showed LOH in three of four informative cases
a
b E1 N
31 T
N
T
D1S245 Figure 2 Genetic alterations on chromosome 1 in mortal and immortal neoplastic human keratinocytes. (a) Idiogram of human chromosome 1 showing the complete retention of heterozygosity at 1q25 and LOH at 1q42-qter in only the lines BICR 18 and BICR 31 both of which were isolated from advanced SCC-HN. Symbols as for Figure 1. (b) Examples of allele loss and retention at D1S245 which maps to 1q42
Genetic analysis of squamous cancer O Loughran et al
studied at D3S1273 and in all three at D3S1038. In all four dierent regions of LOH at 3p13-cen, 3p14-p21.3, 3p22-p24 and 3p25-pter were identi®ed in the SCC-HN lines with no detectable LOH on 3q. One or more regions of chromosome 3p may well carry genes which repress telomerase but the total retention of heterozygosity at all informative 3p markers in the telomerase-positive line BICR 68 suggests that there is at least one alternative mechanism of telomerase reactivation to the loss of the 3p suppressor(s). Examples of allele loss and retention at D3S1067 are shown in Figure 4b. LOH on the X chromosome and other candidate loci The other chromosomal loci known to have antiproliferative activity in immortal human cells were tested for LOH in immortal SCC-HN. In addition to the two lines which showed LOH at 1q42-qter, described above, BICR 19 and BICR 31 showed LOH at 6q13-q16, BICR 78 showed LOH at 11p15.5 and BICR 18 and BICR 22 at 18q22-qter. However, there was no LOH at Xp11-pter which is known to
carry an antiproliferative gene for hamster ®broblasts (Wang et al., 1992), although study of the X chromosome was hampered by the lack of informative cases, most especially in male lines. Similarly we could not detect LOH on chromosome 5q which harbours the mortalin gene (Kaul et al., 1995) or on chromosome 17q21 which harbours the prohibitin gene (White et al., 1991). We also failed to detect any LOH on chromosome 21. These results illustrate that LOH is not generally common in our SCC-HN cell lines and emphasise the possible importance of the more frequent alterations on chromosomes 3 and 4 described above. Furthermore, the LOH on chromosomes 1q42-qter, 6q13-q16, 11p15.5 and 18q22-qter were all additional to the losses described at the other loci and were always found in cell lines derived from advanced tumors. Discussion Our previous studies have indicated that the immortal phenotype of neoplastic head and neck keratinocytes is
a
b 7 N
68 T
D7S496
N
T
D7S530
Figure 3 Genetic alterations on chromosome 7 in mortal and immortal neoplastic human keratinocytes. (a) Idiogram of human chromosome 7 showing evidence of LOH at 7q31 involving the microsatellites D7S496 in line BICR 7, and D7S530 in line BICR 68. Symbols as for Figure 1. (b) Auttoradiograms of the microsatellites D7S496, and D7S530 showing examples of the LOH described above
1959
Genetic analysis of squamous cancer O Loughran et al
1960
inactivation of any of the above replicative lifespan genes in the immortal cells of human tumors. In order to elucidate the mechanisms of cellular immortality and its role in human cancer we developed a series of mortal and immortal human keratinocyte lines from SCC-HN which were amenable to molecular genetic analysis by the LOH technique (Edington et al., 1995). The results show that the most commonly altered loci thought to harbour replicative lifespan genes in immortal SCC-HN keratinocytes were the regions 3p21.2-p21.3 (63%) and 4q32-q34 (the complementation group B locus, 63%). Furthermore, two of the lines not showing LOH at 4q32-q34 did show LOH at 7q31 (the complementation group D locus) and LOH at 4q32-q34 and 7q31 was mutually exclusive. There was no LOH at any of the genetic loci investigated in
associated with p53 dysfunction (Burns et al., 1993; Edington et al., 1995), CDKN2A/p16ink4A dysfunction (Loughran et al., 1994, 1996) and a high frequency of allele loss at other genetic loci indicative of multiple suppressor gene loss (Edington et al., 1995, see also Table 2). Therefore, since many other human chromosomal loci have been shown to have antiproliferative eects reminiscent of replicative senescence when transferred into immortal cells (Vojta et al., 1996; Karlsson et al., 1996; Uejima et al., 1995; Rimessi et al., 1994; Ning et al., 1991; Sandhu et al., 1994, 1996; Ogata et al., 1993; Koi et al., 1993; Casey et al., 1993; Sasaki et al., 1994; Wang et al., 1992) we decided to investigate the role of these loci in neoplastic human keratinocyte immortalization. Furthermore, prior to our study there was no molecular genetic evidence to support the a
b 6 N
18 T
N
19 T
N
31 T
N
63 T
N
T
68 N T
78 N
T
D3S1067 Figure 4 Genetic alterations on chromosome 3 in mortal and immortal neoplastic human keratinocytes. (a) Idiogram of human chromosome 3 showing extensive LOH at 3p13-cen and 3p21.1-p21.3 with consistent LOH also seen in all four informative cases at 3p25-pter. Symbols as for Figure 1. (b) Autoradiogram of the microsatellite D3S1067, showing examples of the allele loss and retention described above
Genetic analysis of squamous cancer O Loughran et al
1961
Table 2 Line Immortal Lines BICR 3 BICR 6 BICR 18 BICR 19 BICR 22 BICR 31 BICR 56 BICR 63 BICR 68 BICR 78 Mortal Cultures BICR 7 BICR E1 BICR E2 BICR E4 BICR E5 BICR 37 BICR 66 BICR 80
CDKN2A Alteration
LOH at D3S1478 and/or D3S1076
Telomerase
Activity Relative to Standard
Codon 282 arg->pro1 Codon 192glu->stop1 No protein2 Exon 1 deleted1 Codons 332+ out of frame Exon 8/9 splice site1 19 bp deletion Codons 308+ out of frame Codons 173, 174 36p deletion1val, arg->gly Codons 126-132 21 bp deletion1 7aa deleted Codon 255 iso->phe High levels of protein, mutation not yet characterised Codon 176 cys->phe
No transcript4 Deleted4 Deleted4 No transcript4
Yes Not informative No Yes
+ + + +
1.4 2.6 1.8 1.9
Deleted4
No
+
2.1
Deleted4
Yes
+
1.1
No protein4
Yes
+
1.9
Deleted No transcript4
Not informative No
+ +
1.8 1.8
Deleted4
Yes
+
2.5
Codon 151 pro->his1 Wild type3 Wild type3 Wild type3 Wild type3 Wild type2 Wild type2 Wild type2
Deleted4 Wild type4 Not detemined4 Wild type4 No protein4 Wild type4 Wild type4 Wild type4
NO Not informative No No No No No No
7a 7 Not determined 7 7 7 7 7
N/A/0.2a N/A N/A N/A N/A N/A N/A N/A
p53 Alteration
4
1
Data from Burns et al., 1993; 2 Data from Edington et al., 1995; 3 Data from Burns et al., 1994; 4 Data from Loughran et al., 1996; +=Telomerase detected; 7 =Telomerase not detected; a Detectable at later population doubling level only; N/A=Not applicable
neoplastic keratinocytes which were senescent in vitro, consistent with the notion that the genes mapping to chromosomes 3, 4 and 7 may be involved in the control of keratinocyte replicative lifespan but with the reservation that the neoplastic keratinocytes which senesce show a very low frequency of allele loss overall (Edington et al., 1995). It has been suggested that multiple genetic routes to senescence exist (Pereira-Smith and Smith, 1988; Sasaki et al., 1994) and dierent recessive mutations are predicted to exist within the same tumor type (Pereira-Smith and Smith, 1988). These mutations may be mutually exclusive (Pereira-Smith and Smith, 1988) or additive (Sasaki et al., 1994). Our results show that LOH at the complementation group B and D loci is mutually exclusive, suggesting that the genes responsible encode products on the same pathway to replicative senescence. Our data also suggest that several pathways to senescence exist (Sasaki et al., 1994) since the immortal SCC-HN phenotype appears to depend on the inactivation of multiple genes. Line BICR 7, is non-senescent, but has a low cloning eciency and resembles a cell population in crisis (Edington et al., 1995). This line has lost both p53 (Burns et al., 1993) and CDKN2A/p16ink4A (Loughran et al., 1996) function, in addition to showing LOH at 7q31 (the complementation group D locus). However, the phenotype of BICR 7 is most likely explained by recent results showing it to have very low levels of telomerase (Table 2) whereas all immortal SCC-HN keratinocytes tested had very high levels. Thus abrogation of at least four pathways involving p53, CDKN2A, the complementation group genes and the regulation of telomerase appear essential to maintain the immortal phenotype of
SCC-HN keratinocytes. Others have also noted the consistent inactivation of the p53 and pRb pathways and the upregulation of telomerase in a range of immortal human tumor cell lines belonging to the dierent complementation groups and concluded that the genes responsible for the complementation groups must reside on a dierent pathway to senescence from the other three (Whitaker et al., 1995). The idea that multiple pathways to senescence exist is not necessarily at variance with the observation that genetic complementation of the immortal phenotype can be achieved (Pereira-Smith and Smith, 1988), since only one of the four pathways presently known may be inactivated by complementary mutations. A gene mapping to chromosome 3p has been reported to inhibit telomerase, cause renewed telomeric attrition and result in delayed proliferation arrest (Ohmura et al., 1995). Furthermore, antiproliferative genes thought to cause the senescence of ovarian cancer cells have been mapped to 3p21.1-p21.2 and 3p24 (Rimessi et al., 1994). Our results showed several regions of LOH on 3p in our neoplastic head and neck cultures that evaded normal senescence, namely 3p13cen, 3p14-p21.3, 3p22-p24 and 3p25-pter. All of these regions might be considered as candidates for the telomerase repressor since high levels of telomerase have been detected in all the immortal head and neck cell lines we have tested but not in the mortal cultures (Table 2). The monochromosome transfer of an intact chromosome 3 into line BICR 31 appears to result in the repression of telomerase and deletion analysis of immortal, telomerase-positive, segregant colonies from these experiments is being carried out to map the position of the repressor. Preliminary data suggests
Genetic analysis of squamous cancer O Loughran et al
1962
Table 3 List of Microsatellites Chromosome 5
Chromosome 6
Chromosome 11
Chromosome 17
Chromosome 18
Chromosome 21
X Chromosome
Microsatellite D5S117 D5S11 D5S107 D5S318 MCC CSF1R Microsatellite D6S105 D6S510 D6S243 D6S239 D6S251 D6S252 D6S87 IGF2R Microsatellite THO 1 D11S861 D11S873 D11S488 Microsatellite D17S520 THRA 1 D17S702 D17S855 D17S902 D17S750 D17S183 D17S579 D175293 NME 1 Microsatellite D18S456 D18S535 D18S475 D18S455 D18S487 D18S465 Microsatellite D21S49 D21S168 D21S171 Microsatellite MIC2 DXS996 KAL DXS987 DXS989 DXS538 DXS556 DXS7 MAOB DXS426 ALAS2 DXS559 DXS453 DXS227 DXS454 COL45A DXS424 DXS1254 DXS297 DXS113 DXYS154
Locus 5p15.1-p15.3 5p13.1-p14.1 5q11.2-q13.3 5q15-q23 5q21-q22 5q33.4-q34 Locus 6p21.1-21.3 6p21.1-p21.2 6p21.2-p21.2 6q13 6q14-q16 6q14-q16 6q22.1-q23.1 6q25.1-q27 Locus 11p15.5 11p15.1 11q14 11q25-q25 Locus 17p12 17q21.1 17q21.1 17q21.1 17q21.1 17q21.1 17q21.1 17q21.1 17q21.1 17q21.1 Locus 18q12.1 18q12.1 18q12.3 18q21.1 18q21.2 18q22.1 Locus 21q22.3 21q22.3 21q22.3-qter Locus Xp22.3 Xp22.3 Xp22.3 Xp22.1 Xp22.1 Xp11.21-p21.1 Xp11.21-p11.4 Xp11.21-p11.4 Xp11.21-P11.4 Xp11.21-p11.4 XP11.21-p11.4 Xq12-q13 Xq12-q13 Xq12-q13 Xq22.3 Xq22.3 Xq24 Xq26 Xq27 Xq27-q28 Xq28
that the telomerase suppressor maps close to microsatellites D3S1478 and D3S1076. However, the lack of LOH at these loci in lines BICR 18, BICR 22 and BICR 68 which have high levels of telomerase, suggests alterative mechanisms for the reactivation of the enzyme, other than the inactivation of a suppressor gene on 3p, as also concluded by Ohmura et al. (1995) in their study of renal cell carcinoma.
The functions of the antiproliferative genes at 1q42qter, 6q13-q16, 11p15 and on chromosome 18 are not known and it is unclear whether these genes are really related to the process of replicative senescence. Our data shows that these loci were only altered in a few lines derived from advanced tumors and the alterations were additional to the ones previously described. It therefore still remains to be established whether these loci harbour true senescence genes. The genes encoding the proteins mortalin (Kaul et al., 1995) and prohibitin (White et al., 1991) have also been implicated in the process of replicative senescence but neither locus showed LOH in our immortal keratinocyte lines indicating that these genes are not common targets for inactivation in SCC-HN. Another issue which has been the subject of much debate is the idea that limited replicative lifespan represents a barrier to tumor development or progression (Newbold, 1985) and that the molecules underpinning the process could be encoded by tumor suppressor genes (Sager, 1989). With reference to this, it is notable that chromosomal imbalances suggestive of LOH have been noted in SCC-HN in vivo at 3p21p24 (Maestro et al., 1993), 4q (Nawroz et al., 1994) and 7q31 (Nawroz et al., 1994; Zenklusen et al., 1995) but the cloning of the relevant genes at these loci and the demonstration of loss of function mutations in vivo are necessary to consolidate this hypothesis. Nevertheless, these reports support our assertion that the genetic alterations we have observed are more likely to have occurred in vivo than in vitro. In summary, our data provide further molecular genetic evidence that replicative lifespan genes may be altered in the immortal cells of a naturally occurring human malignancy, SCC-HN. Our results also support the hypothesis that there are at least four pathways involved in the limitation of replicative lifespan (Sasaki et al., 1994) but that the genes representing the dierent complementation groups (Pereira-Smith and Smith, 1988) may be on the same pathway. A further understanding of the candidate genes described in this article will be advanced greatly by their identi®cation and characterization but the keratinocyte should prove an excellent cell type in which to study their role in cellular immortalization and carcinogenesis. Materials and methods Cell cultures All the neoplastic keratinocytes were cultured using 3T3 feeder layers as described elsewhere (Edington et al., 1996). Immunoreactivity with keratin and involucrin con®rmed that the cultures were keratinocytes and reduced terminal maturation coupled with numerical chromosome aberrations con®rmed that even the senescent tumor-derived keratinocytes were neoplastic (Edington et al., 1995, 1996). The source of normal DNA was lymphocytes and/or ®broblasts taken from the same patient as the keratinocytes, and DNA ®ngerprinting con®rmed that the cultures were unique and derived from the appropriate donors (Edington et al., 1996). A list of the cell lines and cultures used in the study is given in Tables 1 and 2. De®nition of immortality, senecence and crisis An immortal culture was classi®ed as one which proliferated through at least 130 population doublings. A
Genetic analysis of squamous cancer O Loughran et al
senescent culture was de®ned as one which had a cloning eciency of less than 0.1%, did not double in number for 4 weeks, and was composed of large, ¯at and terminally mature keratinocytes. The BICR 7 culture did not senesce normally but reproducibly ceased dividing after 40-45 population doublings. The BICR 7 colonies were composed of small vacuolated cells which eventually detached from the culture dish. This culture appeared to display the features of cells in crisis (Edington et al., 1995). Measurement of telomerase activity Telomerase was measured by a modi®cation of the telomere repeat ampli®cation protocol (TRAP) technique (Kim et al., 1994). Neoplastic keratinocytes and six strains of normal epidermal keratinocytes were seeded at 105 cells per 5 cm dish together with lethally irradiated 3T3 feeder cells. When the cultures were around 50 ± 60% con¯uent the medium was changed and 16 ± 24 h later the feeders were removed with 0.02% EDTA and the dishes washed three times with phosphate-buered saline (PBS). The keratinocytes were then scraped into 1 ml of PBS centrifuged and snap frozen prior to storage at 7708C and extraction of the proteins for TRAP assay. As a control lethally irradiated 3T3 cells were seeded without keratinocytes (mock plates) and were treated in exactly the same way as the test cultures. The mock plates were always telomerase negative. Normal keratinocytes were also cultured using Clonectics Keratinocyte Growth Medium (Tissue Culture Services, UK). In most cases the normal keratinocytes were completely negative for telomerase at the protein concentrations used but in one strain, on one occasion, a faint ladder could be detected at the highest protein concentration. Cell pellets were lysed for 30 min on ice in 200 ml of ice cold lysis buer (10 mM Tris-HCl, pH 7.5), 1 mM MgCl2, 1 mM EGTA, 0.5% CHAPS, 10% glycerol, 5 mM bmercaptoethanol, 0.1 mM PMSF, Piatyszek et al., 1995). Centrifugation of the lysate at 14 000 g for 30 min at 48C pelleted the cell debris, allowing 160 ml of the supernatant to be removed, snap frozen and stored at 7808C. Protein concentrations were measured with a Coomassie protein assay kit. Telomerase activity was determined from 0.5 mg protein, using the one tube PCR-based protocol. Wax barrier tubes were prepared in advance and stored for up to 3 weeks at room temperature. 100 ng of the CX oligonucleotide in Tris-HCl pH 8.3) was dehydrated and sealed beneath a Perkin Elmer Ampliwax barrier (PCR cycles of 838C for 5 min to melt the wax gem and slow cooling to 258C at a rate of 18C 5 s). Telomerase mediated extension of the TS primer was achieved during a 30 min incubation at room temperature using a 50 ml reaction mixture containing reaction buer (20 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 68 mM KCl, 0.5% Tween 20, 1 mM EGTA), 100 ng TS primer, 0.5 mM T4 gene 32 protein, 50 ml each dinucleoside triphosphate, 2 U Taq polymerase, 2 mCi each of 10 mCi/ml [a32P]CTP and [a32P]TTP 3000 Ci/mM. A 150 bp myogenin cDNA internal standard (a gift from Lauren S Gollahan) was included at ®ve attograms per reaction (Wright et al., 1995). The wax barrier was melted at 908C for 2 min and the reaction subjected to 30 PCR cycles of 948C for 30 s, 508C for 30 s, 728C for 45 s, followed by one cycle of 948C for 30 s, 508C for 30 s and 728C for 1 min. An RNase control tube was included to determine speci®city. Reaction tubes were frozen for at least 1 h to removal of the wax plug. 25 ml of the PCR products were visualised on a 17613.5 cm 10% polyacrylamide non-denaturing gel, run in 0.56TBE buer for 45 min at 175 V and 105 min at 280 V until the xylene cyanol band is 5 cm from the bottom of the gel. Gels were ®xed in 50% ethanol, 0.5 M NaCl and 40 mM sodium acetate (pH 4.2) for 20 min and dried for 1 h prior to
exposure to a phosphor screen for an hour. The intensity of the telomerase-derived ladders and internal standards were determined on a PhosphorImager using ImageQuant software. Quanti®cation was attained as log10 of the sum of the peaks of each sample normalised to the signal from the internal standard. DNA isolation, Southern blotting and RFLP analysis The details of DNA isolation and analysis of the pMUC7 and AT III RFLPs by Southern blotting and by agarose gel electrophoresis of restriction enzyme digested PCR fragments respectively has been described previously (Loughran et al., 1994; Edington et al., 1995). Microsatellite analysis All PCRs of microsatellite sequences were carried out in the same way. Total reaction volumes were 25 ml containing 40 ± 100 ng DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 10% dimethylsulfoxide, 140 ± 350 ng of each primer, 200 mM concentrations of each deoxynucleotide triphosphate, and 1 ml of [a32P]dCTP, 0.1 mBq/ml. The mixture was heated to 948C, at which point 2.5 units of Taq polymerase (Perkin Elmer Cetus, Emeryville, CA) was added. Reactions were then subjected to six cycles for 948C for 30 s and 608C for 30 s, followed by 28 cycles of 948C, 30 s; 558C, 30 s; and 728C, 30 s. After this was completed, the reactions underwent further extension of 7 min at 728C and were cooled to 48C. Thermocyclers were Perkin Elmer Cetus type 9600. In all cases cell line and normal DNA were ampli®ed simultaneously and each experiment was repeated at least once to ensure the validity of the results. The radiolabeled reaction products were separated on 4 ± 10.5% polyacrylamide gels under nondenaturing conditions. The gels were then exposed to X-ray ®lm to visualize the resolved reaction products. In some informative cases, two shadow bands were also visible in addition to the major bands. The shadow bands migrated at a slightly higher molecular weight than the major bands, and in the case of LOH both the major band and the shadow band representing the lost allele both disappeared. Thus, the presence of the shadow bands did not interfere with the interpretation of the result. All primers were either obtained from Research Genetics, Inc. (Huntsville, AL) or synthesized on an Applied Biosystems 381 DNA Synthesizer or 392 DNA/RNA Synthesizer using the manufacturer's protocols and reagents from Cruachem (Glasgow, Scotland, UK). A list of the microsatellite markers used, but not illustrated in the Figures, is given in Table 3. The chromosomal loci were obtained from the Human Genome Mapping Project.
Abbreviations The abbreviations used are: SCC, squamous cell carcinoma; SCC-HN, squamous cell carcinoma of the head and neck; LOH, loss of heterozygosity; RFLP, restriction fragment length polymorphism; PCR, polymerase chain reaction. Acknowledgements The authors wish to thank Professor John Wyke for critical reading of the manuscript and the Cancer Research Campaign and the Association of International Cancer Research for ®nancial support. We would also like to thank Olivia Pereira-Smith and Mike Bertram for the communication of results prior to publication.
1963
Genetic analysis of squamous cancer O Loughran et al
1964
References Balmain A and Brown K. (1988). Adv. Cancer Res., 51, 147 ± 182. Berry IJ, Burns JE and Parkinson EK. (1994). Mol. Carcinog., 9, 134 ± 142. Burns JE, Baird MC, Clark LJ, Burns P, Edington K, Chapman C, Mitchell R, Robertson G, Soutar D and Parkinson EK. (1993). Br. J. Cancer, 67, 1274 ± 1284. Burns JE, Clark LJ, Yeudall WA, Mitchell R, MacKenzie K, Chang S and Parkinson EK. (1994). Br. J. Cancer, 70, 591 ± 595. Cavanee WK, Dryja TP, Phillips RA, Benedict WF, Godbout R, Gallie BL, Murphree AL, Strong LC and White RL. (1983). Nature, 305, 779 ± 784. Casey G, Plummer S, Hoetge G, Scanlon D, Faschling C and Stanbridge EJ. (1993). Hum. Mol. Genet., 2, 1921 ± 1927. Counter CM, Avilion AA, Le Feuvre CE, Stewart WG, Greider CW, Harley CB and Bacchetti S. (1992). EMBO J., 11, 1921 ± 1929. Counter CM, Botelho EM, Wang P, Harley CB and Bacchetti S. (1994). J. Virol., 68, 3410 ± 3414. Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C, Medrano EE, Linskens M, Rubelj I, Pereira-Smith O, Peacocke M and Campisi J. (1995). Proc. Natl. Acad. Sci. USA, 92, 9363 ± 9367. Easty DM, Easty GC, Carter RL, Monaghan P and Butler LJ. (1981). Br. J. Cancer, 43, 772 ± 785. Edington KG, Berry IJ, O'Prey M, Burns J, Clark LJ, Mitchell R, Robertson G, Soutar D, Coggins LW and Parkinson EK. (1996). RI Freshney (ed.) Culture of Tumor Cells, Culture of Specialized Cells. Vol 4, Wiley-Liss Inc: New York, in press. Edington KG, Loughran OP, Berry IJ and Parkinson EK. (1995). Mol. Carcinog., 13, 254 ± 265. Hara E, Tsurui A, Shinosaki S, Nakada S and Oda K. (1991). Biochem. Biophys. Res. Commun., 179, 528 ± 534. Hay¯ick L. (1965). Exp. Cell. Res., 37, 614 ± 636. Hensler PJ, Annab LA, Barrett JC and Pereira-Smith OM. (1994). Mol. Cell. Biol., 14, 2291 ± 2297. Karlsson C, Stenman G, Vojta PJ, Bongcam-Rudlo E, Barrett JC, Westermark B and Paulsson Y. (1996). Cancer Res., 56, 241 ± 245. Kaufmann WK, Levedakou EN, Grady HL, Paules RS and Stein GH. (1995). Cancer Res., 55, 7 ± 11. Kaul SC, Wadhwa R, Matsuda Y, Hensler PJ, Pereira-Smith OM, Komatsu Y and Mitsui Y. (1995). FEBS Lett., 361, 269 ± 272. Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PL, Coviello GM, Wright WE, Weinrich SL and Shay JW. (1994). Science, 266, 2011 ± 2014. Koi M, Johnson LA, Kalikrin LM, Little PFR, Nakamura Y and Feinberg AP. (1993). Science, 260, 361 ± 364. Loughran O, Edington KG, Berry IJ, Clark LJ and Parkinson EK. (1994). Cancer Res., 54, 5045 ± 5049. Loughran O, Malliri D, Owens D, Gallimore PH, Stanley MA, Ozanne B, Frame MC and Parkinson EK. (1996). Oncogene, 13, 561 ± 568. Maestro R, Gasparotto D, Vukosavlijevic T, Barzan L, Sulfaro SM and Boiocchi M. (1993). Cancer Res., 53, 5775 ± 5779.
Nawroz H, van der Riet P, Hruban R, Koch W, Ruppert JM and Sidransky D. (1994). Cancer Res., 54, 1152 ± 1155. Newbold RF. (1985). Barret JC and Tennent RW (eds.) Carcinogenesis: A Comprehensive Survey, Vol. 9, Raven Press: New York pp. 17 ± 28. Ning Y, Weber JL, Killary AM, Ledbetter DH, Smith JR and Pereira-Smith OM. (1991). Proc. Natl. Acad. Sci. USA, 88, 5635 ± 5639. Ogata T, Auysawa D, Namba M, Takahashi E, Oshimura M and Oishi M. (1993). Mol. Cell. Biol., 13, 6036 ± 6043. Ohmura H, Tahara H, Suzuki M, Ide T, Shimizu M, Yoshida MA, Tahara E, Shay JW, Barrett JC and Oshimura M. (1995). Jpn. J. Cancer Res., 86, 899 ± 904. Pereira-Smith OM and Smith JR. (1988). Proc. Natl. Acad. Sci. USA, 85, 6042 ± 6046. Piatyszek MA, Kim NW, Weinrich SL, Hiyama K, Hiyama E, Wright WE and Shay JW. (1995). Meth. Cell Sci., 17, 1 ± 15. Radna RL, Caton Y, Jha KK, Kaplan P, Li G, Traganos F and Ozer HL. (1989). Mol. Cell. Biol., 9, 3093 ± 3096. Rheinwald JG and Beckett MA. (1981). Cancer Res., 41, 1657 ± 1663. Rheinwald JG and Green H. (1975). Cell, 6, 331 ± 343. Rimessi P, Gualandi F, Morelli C, Trabanelli C, Wu Q, Possati L, Montesi M, Barrett JC, Barbanti-B and Rodano G. (1994). Oncogene, 9, 3467 ± 3474. Sager R. (1989). Science, 246, 1406 ± 1412. Sandhu AK, Hubbard K, Kaur GP, Jha KK, Ozer HL and Athwal RS. (1994). Proc. Natl. Acad. Sci. USA, 91, 5498 ± 5502. Sandhu AK, Kaur GP, Reddy DE, Rane NS and Athwal RS. (1996). Oncogene, 12, 247 ± 252. Sasaki M, Honda T, Yamada H, Wake N, Barrett JC and Oshimura M. (1994). Cancer Res., 54, 6090 ± 6093. Shay JW, Pereira-Smith OM and Wright WE. (1991). Exp. Cell. Res., 196, 33 ± 39. Uejima H, Mitsuya K, Kugoh H, Horikawa I and Oshimura M. (1995). Genes Chrom. Cancer, 14, 120 ± 127. Vogelstein B, Fearon ER, Kern SE, Hamilton SR, Preisinger AC, Nakamura Y and White R. (1989). Science, 244, 207 ± 211. Vojta PJ, Futreal PA, Annab LA, Kato H, Pereira-Smith OM and Barrett JC. (1996). Genes Chrom. Cancer, 16, 55 ± 63. Whitaker NJ, Bryan TM, Bonne®n P, Chang AGM, Musgrove EA, Braithwaite AW and Reddel RR. (1995). Oncogene, 11, 971 ± 976. Wang XW, Lin X, Klein CB, Bhamra RK, Lee Y-W and Costa M. (1992). Carcinogenesis, 13, 555 ± 561. White JJ, Ledbetter DH, Eddy RL, Shows TB, Stewart DA, Nuell MJ, Friedman V, Wood CM, Owens GA, McClung JK, Danner DB and Morton CC. (1991). Genomics, 11, 228 ± 230. Wright WE, Pereira-Smith OM and Shay JW. (1989). Mol. Cell. Biol., 9, 3088 ± 3092. Wright WE, Shay JW and Piatyszek MA. (1995). Nucleic Acids Res., 23, 3794 ± 3795. Zenklusen JC, Thompson JC, Klein-Szanto AJP and Conti CJ. (1995). Cancer Res., 55, 1347 ± 1350.
