Clinical Genetics 1990: 38: 51-56
Chromosome breakage in lymphocytes from members of cancer families showing autosomal dominant inheritance M. N. KHOUZAM',K. TSIOUPRA' AND J. D. A. DELHANTY' 'Norfolk Norwich Hospital, Norwich, and 2Galton Laboratory, Department of Genetics and Biometry, University College, London, UK We have conducted investigations on members of three families with increased predisposition to cancer which appears to be inherited as an autosomal dominant trait. The aim of our studies overall is to provide markers for the mutant genes involved, so that gene carriers may be monitored closely for signs of malignant disease. This paper reports on studies of chromosome breakage in lymphocytes from affected and at-risk family members and control subjects. No increase in spontaneous chromosome breakage was observed in family members compared with controls. An increased sensitivity to Chromosome damage induced by the alkylating agent, N-methyl-N'-nitro-N-nitrosoguanidine(MNNG), was observed in three members from two families; one person was affected, the others at risk. These families included cases of osteosarcoma, in addition to various types of cancer of epithelial origin. Two members (one affected, one at-risk) of a third family showed increased sensitivity to the radio-mimetic agent, bleomycin. This family appeared to represent the cancer family syndrome. Whilst not conclusive at present, our results appear to justify investigation of members of cancer Families with respect to sensitivity to chromosome breaking agents. Received 24 July 1989, revised 18 January, accepted for publication 20 January 1990 Key words: cancer families; chromosome breakage; colorectal cancer.
Cancer of the colorectum is second only to lung cancer as a cause of morbidity and mortality in the Western World. Exclusive of the heritable polyposis syndromes which account for less than 5 % of all colon cancer, most cases are sporadic with a negative family history. However, that genetic factors are operational to some extent is indicated by the results of one series where 26% of patients had a positive family history (Lovett 1976) and surveys which suggest that the relatives of an index case have a 3-4 fold increased risk of developing colorectal cancer during their lifetime (Anderson 1980a). However, for some families these figures are clearly an underestimate. Pedigree stud-
ies of non-polyposis colorectal cancer patients have revealed factors which indicate a genetic predisposition to cancer in some families. These include earlier age of onset, multiple primary tumours, simultaneous or independent presence of endometrial carcinoma in a first-degree female relative, or gastric cancer in relatives of either sex and a predominance of proximal colon location. It has been suggested that in this subgroup the cancer liability behaves as an autosomal dominant trait (Anderson 1980a). This group of entities constitutes the cancer family syndrome (CFS), Lynch syndrome 11, or hereditary adenocarcinomatosis. Hereditary site-specific colonic cancer (Lynch syn-
52
KHOUZAM ET AL.
drome I) and the Torre-Muir syndrome, which features adenocarcinoma of the large and small bowel and uterus, squamous cell carcinoma of the mucous membrane and transitional cell carcinoma of the urinary system (Anderson 1980b), are also recognised as clearly due to an autosomal dominant gene. Other families are known which do not appear to be examples of the aforementioned syndromes. Two of our families come into this category, with different members presenting with cancers of various sites, some of which are of non-epithelial origin. Despite intensive research into the aetiology of colonic cancer and the use of modern surgical techniques in treatment, the prognosis has not changed over the past 4 decades (Bulow 1980). This situation could be improved by earlier detection and in the case of cancer families, identification of those individuals at increased risk. Isolation of the genes concerned would provide the most definitive markers or, alternatively, closely linked DNA polymorphisms. Prior to these becoming available, searches have been made for biomarkers of various types (Lynch et al. 1985). Boveri’s hypothesis of the chromosomal cause of cancer, which was proposed in 1911 (cited in Haluska et al. 1987), has in the last decade received much support from the discovery of the role of two classes of genes in the development of human cancer. First to be described were the human counterparts of the viral oncogenes, (proto-oncogenes), activation of which is frequently by chromosomal translocation (Klein 1987). Secondly, the more recently demonstrated recessively acting tumour suppressor genes (anti-oncogenes) (Knudson 1971). As demonstrated in the classical model, retinoblastoma, homozygous inactivation of this second class of genes leads to tumorigenesis and this is frequently accomplished by chromosomal mechanisms (Murphree & Benedict 1984). Thus it is self-evident that an
increased predisposition to chromosome breakage will lead to an increased risk of cancer. Investigation of the spontaneous rate of chromosome breakage and rearrangement in members of cancer families may establish whether there is an increased rate in affected persons, as has already been demonstrated for adenomatous polyposis coli (Gardner et al. 1982, Delhanty et al. 1983). By determining the comparative sensitivity to different types of chromosome damaging agents, we may detect individuals whose cells are particularly sensitive to environmental agents. In order to search for informative biomarkers, a survey of families with colorectal and other cancers has been undertaken. Two families have been studied in detail and a third one is being recruited. This paper presents the results from chromosome studies on lymphocytes. Further studies on skin fibroblasts are being carried out. Linkage analysis is being performed where possible. Material and Methods
Families The first family, D, appears to represent the cancer family syndrome. The proband, E.D., developed mucoid carcinoma of the rectum at the age of 55. At the age of 57 an adenocarcinoma of the caecum and 2 polypi were detected, and at 61 a poorly differentiated mucus-secreting gastric adenocarcinoma was diagnosed. A long-standing keratoacanthorna of the chin was removed at 57 and a hyperkeratolic acantholytic skin lesion, Darner’s disease, of the right axilla and chest became apparent at 55, with later development of squamous cell carcinoma. He is still alive. Cancers diagnosed in other family members are marked on pedigree 1 (Fig. 1). The second family, P, is one of the examples of a familial aggregate of cancers of various sites and tissues. S.P. is the proband,
C H R O M O S O M E B R E A K A G E IN C A N C E R FAMILIES
53
several types, with at least one known case of osteosarcoma (Pedigree 3, Fig. 1). Investigations were carried out on two affected patients (ED and SP); 7 at-risk family members (PD, ID, JA, SS, MAP, MKP and BW) and 3 spouses or other controls (TP, RA & AJ). RA had been adopted into family P.
Ill
IY
I
Fig. 1. Pedigrees of cancer families. Cancer sites: BL Bladder; Br Breast; Co Colon: Cx Cervix; Le Leukaemia: Li Liver; Lu Lung; Oes Oesophagus; 0 Ostecsarcoma; Ov Ovary; P Pancreas; R Rectum; Sk Skin; St Stomach; Ut Uterus; U Unknown.
who presented with cancer of the cervix uteri, aged 40. The distribution of malignancies in related and unrelated family members can be seen in Pedigree 2 (Fig. 1). Most members of the third family, W, live in Ireland; so far, we have obtained samples only from BW, the proband, who is herself at risk. This family resembles family P in having members afflicted with cancers of
Lymphocyte Culture and Chromosome Analysis Peripheral blood samples were collected for lymphocytic chromosome culture and DNA preparation. Small skin biopsies (3 mm') were taken under local anaesthetic and established in long-term tissue culture for chromosomal analysis of fibroblasts. Skin fibroblasts were also frozen in liquid nitrogen for future use in linkage studies as more DNA markers become available. Chromosomes were prepared by standard methods and slides were coded before analysis. Spontaneous aberration rates were determined by complete karyotype analysis of 50 orcein-stained metaphases per individual, using 48-h cultures. Two agents were used to investigate susceptibility to induced chromosome damage. Bleomycin, a radio-mimetic anti-tumour agent, and the mutagenic alkylating agent, MNNG. The effect of bleomycin (30 pg ml-*) was monitored at the G2 stage of the cell cycle (the last 4 h before harvesting). MNNG at 5 pg ml-', or the equivalent amount of solvent, (DMSO), was present for the final 24 h of culture. Fifty cells per treatment were scored on coded slides for chromosome and chromatid breaks and exchanges, dicentrics and acentric fragments, and the results expressed as breaks per cell. Due to variables such as culture media and drug batches, it is considered that only experiments carried out at the same time can be compared in absolute terms; results are thus divided into two groups, according to
K H O U Z A M ET A L .
54
when cultured. Significancewas assessed by means of the two sample t-test.
Table 2 Chromosome damage induced b y MNNG in lymphocytes Mean no. breaks per cell f SD
Results
All subjects had a normal chromosome complement. No difference was apparent in the spontaneous aberration rate between patients, or at-risk family members and controls (Table 1). Following treatment with MNNG, patients and at-risk individuals in Group 1 were compared with control, AJ, who was a member of the laboratory staff (Table 2). BW (at-risk) showed a significantly higher number of breaks per cell than the control, while other at-risk individuals together with patient ED did not differ significantly from the control value. Family P constituted Group 2, with spouse TI' acting as control for comparison purposes, although since RA is adopted he can a:so be considered a In this group, sp (affected) and JA (mother of an affected person and hence gene carrier) Table 1 Spontaneous
chromosome aberrations Iymp hocytes Total breaks
Cancer patients SP ED At-risk relatives MKP MAP JA
t
Per cent abnormal cellst 6 0
4 6
14 4
ss
BW Controls TP RA
in
Significance'
Group 1
co2p1
0.36f1.16
Cancer Datient
ED
0.18f0.60
NS
PD BW
0.68f 1.52 0.08f 0.27 1.26f2.06
NS NS p
Chromosome breakage in lymphocytes from members of cancer families showing autosomal dominant inheritance M. N. KHOUZAM',K. TSIOUPRA' AND J. D. A. DELHANTY' 'Norfolk Norwich Hospital, Norwich, and 2Galton Laboratory, Department of Genetics and Biometry, University College, London, UK We have conducted investigations on members of three families with increased predisposition to cancer which appears to be inherited as an autosomal dominant trait. The aim of our studies overall is to provide markers for the mutant genes involved, so that gene carriers may be monitored closely for signs of malignant disease. This paper reports on studies of chromosome breakage in lymphocytes from affected and at-risk family members and control subjects. No increase in spontaneous chromosome breakage was observed in family members compared with controls. An increased sensitivity to Chromosome damage induced by the alkylating agent, N-methyl-N'-nitro-N-nitrosoguanidine(MNNG), was observed in three members from two families; one person was affected, the others at risk. These families included cases of osteosarcoma, in addition to various types of cancer of epithelial origin. Two members (one affected, one at-risk) of a third family showed increased sensitivity to the radio-mimetic agent, bleomycin. This family appeared to represent the cancer family syndrome. Whilst not conclusive at present, our results appear to justify investigation of members of cancer Families with respect to sensitivity to chromosome breaking agents. Received 24 July 1989, revised 18 January, accepted for publication 20 January 1990 Key words: cancer families; chromosome breakage; colorectal cancer.
Cancer of the colorectum is second only to lung cancer as a cause of morbidity and mortality in the Western World. Exclusive of the heritable polyposis syndromes which account for less than 5 % of all colon cancer, most cases are sporadic with a negative family history. However, that genetic factors are operational to some extent is indicated by the results of one series where 26% of patients had a positive family history (Lovett 1976) and surveys which suggest that the relatives of an index case have a 3-4 fold increased risk of developing colorectal cancer during their lifetime (Anderson 1980a). However, for some families these figures are clearly an underestimate. Pedigree stud-
ies of non-polyposis colorectal cancer patients have revealed factors which indicate a genetic predisposition to cancer in some families. These include earlier age of onset, multiple primary tumours, simultaneous or independent presence of endometrial carcinoma in a first-degree female relative, or gastric cancer in relatives of either sex and a predominance of proximal colon location. It has been suggested that in this subgroup the cancer liability behaves as an autosomal dominant trait (Anderson 1980a). This group of entities constitutes the cancer family syndrome (CFS), Lynch syndrome 11, or hereditary adenocarcinomatosis. Hereditary site-specific colonic cancer (Lynch syn-
52
KHOUZAM ET AL.
drome I) and the Torre-Muir syndrome, which features adenocarcinoma of the large and small bowel and uterus, squamous cell carcinoma of the mucous membrane and transitional cell carcinoma of the urinary system (Anderson 1980b), are also recognised as clearly due to an autosomal dominant gene. Other families are known which do not appear to be examples of the aforementioned syndromes. Two of our families come into this category, with different members presenting with cancers of various sites, some of which are of non-epithelial origin. Despite intensive research into the aetiology of colonic cancer and the use of modern surgical techniques in treatment, the prognosis has not changed over the past 4 decades (Bulow 1980). This situation could be improved by earlier detection and in the case of cancer families, identification of those individuals at increased risk. Isolation of the genes concerned would provide the most definitive markers or, alternatively, closely linked DNA polymorphisms. Prior to these becoming available, searches have been made for biomarkers of various types (Lynch et al. 1985). Boveri’s hypothesis of the chromosomal cause of cancer, which was proposed in 1911 (cited in Haluska et al. 1987), has in the last decade received much support from the discovery of the role of two classes of genes in the development of human cancer. First to be described were the human counterparts of the viral oncogenes, (proto-oncogenes), activation of which is frequently by chromosomal translocation (Klein 1987). Secondly, the more recently demonstrated recessively acting tumour suppressor genes (anti-oncogenes) (Knudson 1971). As demonstrated in the classical model, retinoblastoma, homozygous inactivation of this second class of genes leads to tumorigenesis and this is frequently accomplished by chromosomal mechanisms (Murphree & Benedict 1984). Thus it is self-evident that an
increased predisposition to chromosome breakage will lead to an increased risk of cancer. Investigation of the spontaneous rate of chromosome breakage and rearrangement in members of cancer families may establish whether there is an increased rate in affected persons, as has already been demonstrated for adenomatous polyposis coli (Gardner et al. 1982, Delhanty et al. 1983). By determining the comparative sensitivity to different types of chromosome damaging agents, we may detect individuals whose cells are particularly sensitive to environmental agents. In order to search for informative biomarkers, a survey of families with colorectal and other cancers has been undertaken. Two families have been studied in detail and a third one is being recruited. This paper presents the results from chromosome studies on lymphocytes. Further studies on skin fibroblasts are being carried out. Linkage analysis is being performed where possible. Material and Methods
Families The first family, D, appears to represent the cancer family syndrome. The proband, E.D., developed mucoid carcinoma of the rectum at the age of 55. At the age of 57 an adenocarcinoma of the caecum and 2 polypi were detected, and at 61 a poorly differentiated mucus-secreting gastric adenocarcinoma was diagnosed. A long-standing keratoacanthorna of the chin was removed at 57 and a hyperkeratolic acantholytic skin lesion, Darner’s disease, of the right axilla and chest became apparent at 55, with later development of squamous cell carcinoma. He is still alive. Cancers diagnosed in other family members are marked on pedigree 1 (Fig. 1). The second family, P, is one of the examples of a familial aggregate of cancers of various sites and tissues. S.P. is the proband,
C H R O M O S O M E B R E A K A G E IN C A N C E R FAMILIES
53
several types, with at least one known case of osteosarcoma (Pedigree 3, Fig. 1). Investigations were carried out on two affected patients (ED and SP); 7 at-risk family members (PD, ID, JA, SS, MAP, MKP and BW) and 3 spouses or other controls (TP, RA & AJ). RA had been adopted into family P.
Ill
IY
I
Fig. 1. Pedigrees of cancer families. Cancer sites: BL Bladder; Br Breast; Co Colon: Cx Cervix; Le Leukaemia: Li Liver; Lu Lung; Oes Oesophagus; 0 Ostecsarcoma; Ov Ovary; P Pancreas; R Rectum; Sk Skin; St Stomach; Ut Uterus; U Unknown.
who presented with cancer of the cervix uteri, aged 40. The distribution of malignancies in related and unrelated family members can be seen in Pedigree 2 (Fig. 1). Most members of the third family, W, live in Ireland; so far, we have obtained samples only from BW, the proband, who is herself at risk. This family resembles family P in having members afflicted with cancers of
Lymphocyte Culture and Chromosome Analysis Peripheral blood samples were collected for lymphocytic chromosome culture and DNA preparation. Small skin biopsies (3 mm') were taken under local anaesthetic and established in long-term tissue culture for chromosomal analysis of fibroblasts. Skin fibroblasts were also frozen in liquid nitrogen for future use in linkage studies as more DNA markers become available. Chromosomes were prepared by standard methods and slides were coded before analysis. Spontaneous aberration rates were determined by complete karyotype analysis of 50 orcein-stained metaphases per individual, using 48-h cultures. Two agents were used to investigate susceptibility to induced chromosome damage. Bleomycin, a radio-mimetic anti-tumour agent, and the mutagenic alkylating agent, MNNG. The effect of bleomycin (30 pg ml-*) was monitored at the G2 stage of the cell cycle (the last 4 h before harvesting). MNNG at 5 pg ml-', or the equivalent amount of solvent, (DMSO), was present for the final 24 h of culture. Fifty cells per treatment were scored on coded slides for chromosome and chromatid breaks and exchanges, dicentrics and acentric fragments, and the results expressed as breaks per cell. Due to variables such as culture media and drug batches, it is considered that only experiments carried out at the same time can be compared in absolute terms; results are thus divided into two groups, according to
K H O U Z A M ET A L .
54
when cultured. Significancewas assessed by means of the two sample t-test.
Table 2 Chromosome damage induced b y MNNG in lymphocytes Mean no. breaks per cell f SD
Results
All subjects had a normal chromosome complement. No difference was apparent in the spontaneous aberration rate between patients, or at-risk family members and controls (Table 1). Following treatment with MNNG, patients and at-risk individuals in Group 1 were compared with control, AJ, who was a member of the laboratory staff (Table 2). BW (at-risk) showed a significantly higher number of breaks per cell than the control, while other at-risk individuals together with patient ED did not differ significantly from the control value. Family P constituted Group 2, with spouse TI' acting as control for comparison purposes, although since RA is adopted he can a:so be considered a In this group, sp (affected) and JA (mother of an affected person and hence gene carrier) Table 1 Spontaneous
chromosome aberrations Iymp hocytes Total breaks
Cancer patients SP ED At-risk relatives MKP MAP JA
t
Per cent abnormal cellst 6 0
4 6
14 4
ss
BW Controls TP RA
in
Significance'
Group 1
co2p1
0.36f1.16
Cancer Datient
ED
0.18f0.60
NS
PD BW
0.68f 1.52 0.08f 0.27 1.26f2.06
NS NS p
