Download High Frequency in Vivo Loss of Heterozygosity Is Primarily a

ICANCER RESEARCH57. 1188-I 193. March 15. 9971
High Frequency in Vivo Loss of Heterozygosity Is Primarily a Consequence of
Mitotic Recombination1
Pawan K. Gupta,2Amrik Sahota,SimeonA. Boyadjiev,StevenBye, ChangshunShao, J. PatrickO'Neill,
Timothy
C. Hunter,
Richard
J. Albertini,
Peter
J. Stambrook,
and Jay A. Tischfleld3
Department ofMedical and Molecular Genetics. Indiana University School ofMedicine, Indianapolis, Indiana 46202-5251 (P. K. G.. A. S., S. A. B., S. B.. C. S., J. A. TI: Genetics
Laboratory. University of Vermont, Burlington, Vermont 05401 /1. P. 0., T. @.
H., R. J. A.): and Department of Cell Biology. Neurobiology and Anatomy, College of Medicine,
The University of Cincinnati, Cincinnati. Ohio 45267 (P. J. S.]
junctional chromosomal loss with or without reduplication as well as
gene conversion and point mutations (3—5).
We have used the adenine phosphoribosyltransferase gene (A.PRT;
A number of different assays have been developed to determine the
16q24) to investigate the mechanisms of loss of heterozygosity (LOH) in
mechanisms of point mutations and multilocus events in mammalian
normal human somatic cells in vivo. APRT-deficient (APRT@, APRT'°)
cells in vitro (5, 6). Mutagenesis studies using markers that are
T lymphocytes from the peripheral blood of four obligate APRT heterozy
functionally (e.g., X-linked HPRT) or structurally (e.g., Chinese ham
gotes (APRT@) with characterized
germ-line mutations were selected in
ster ovary Aprt) hemizygous usually detect intragenic point mutations
medium containing 100 @LM
2,6-diaminopurine. A total of 80 2,6-diamin
and relatively small deletions only, at least in aneuploid cell lines
opurine-resistant T-cell clones from 2 of the heterozygotes were analyzed
for this study. The presence or absence of LOll of proximal linked (6—8). It is believed that clones with larger deletions may not be
microsatellite repeat markers was used to divide the clones into two viable due to the loss of adjacent genes essential for survival (8). The
groups: (a) those in which LOH was likely due to localized changes In assumed limitations inherent in studying mutations at hemizygous loci
APRT (e.g., point mutations); and (b) those with LOH at additional loci. A have led to the development of assays using autosomal genes such as
total of 61 clones (76%) exhibited LOH of linked microsatellite
repeat
APRT and tk (9). Selection of mutants from cells with heterozygous
markers at different locations on 16q, which extended from the smallest
autosomal loci has been shown to occur at higher frequency than that
measured region (<5.5 cM) to the entire 16q arm. The remaining 19 from cells having a hemizygous target locus (9, 10). These results
clones (24%) had point mutations in APRT or other relatively minor
have been interpreted as evidence for a relatively closely linked
alterations. Ten clones with LOH encompassing different regions of 16q
were examined by conventional cytogenetics and by fluorescence in situ gene(s) with an essential function whose loss limits the size of
survivable deletions at the hemizygous or X-linked locus. Whereas
hybridization using an I4.PRT cosmid probe. All dones exhibited a normal
diploid karyotype, and nine exhibited two copies ofAPRT. The one clone deletion of such an essential gene will always be lethal when it is
hemizygous, lethality in diploid cells will only be observed when the
that was hemizygous for APRT had the smallest observed region of LOH
deleted essential gene is heterozygous, and its normal allele is in
in clones from that individual. These results indicate that mitotic recom
binatlon and, to a much lesser extent, deletion may be the primary
cis-configuration with the deleted normal allele of the marker gene
mechanisms for the relatively high frequency of in vivo LOH observed in (I 1—14).
normal human T cells. Became LOH leads to the expression of recessive
LOH can only be studied with heterozygous autosomal loci. In the
tumor
suppressor
genes in many cancers,
these data have significant
present study, we have used APRT as a marker to select normal human
implications for the role of LOH in the early stages of tumor development,
T cells that have undergone in vivo LOH. APRT catalyzes the syn
especially in breast cancer.
thesis of AMP from adenine and 5-phosphoribosyl-l-pyrophosphate.
A deficiency of this enzyme frequently leads to 2,8-dihydroxyadenine
stone formation in the kidney (15). The APRT gene, located at
INTRODUCTION
l6q24.3, encodes a selectable marker in cell cultures, and the se
quences of the genomic and coding regions (about 2.6 and 0.54 kb,
Allelic LOH4 plays an important role in carcinogenesis. A number
respectively) are known (16). We have analyzed germ-line mutations
of inherited cancers have been shown to be the direct consequence of
in this gene in more than 33 families (15, l7).@ Germ-line mutations
a germ-line mutation in one allele and a subsequent somatic alteration
seem to cluster at the intron 4 splice donor site and at or near codon
that inactivates the remaining functional allele (1, 2). The majority of
87 (exon 3) and consist entirely of point mutations and small frame
these germ-line mutations are intragenic point mutations that are
shifts. Analysis of in vivo APRT somatic point mutations in the T cells
generally thought to be due to DNA damage or to errors in DNA
of
three Japanese heterozygotes showed a similar clustering (17). We
repair or replication. The subsequent somatic alterations resulting in
have selected DAP-resistant T-cell clones arising in vivo in the blood
LOH are generally more complex and may include multilocus chro
of parents (obligate APRT heterozygotes) from two unrelated APRT
mosomal events such as deletion, mitotic recombination, or nondis
deficient families and have examined the entire spectrum of in vivo
APRT LOH in the clones from two of the parents. Although the
Received 9/30/96; accepted 1/19/97.
The costs of publication of this article were defrayed in part by the payment of page
studies reported here also describe point mutations, the focus is
ABSTRACT
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.
, Supported
by
NIH
Grants
DK38185
(to
J.
A.
T.)
States Department of Energy Grant F6028760502
05652 (to J.A.T. and PJ.S.).
2 Present
address:
Department
of
Medicine,
and
CA3O688
(to
R.
J.
A.),
whom
requests
for
reprints
should
Division
of
Medical
Genetics,
MATERIALS
Thomas
be addressed,
at Department
of Medical
and
Molecular Genetics, Indiana University School of Medicine, 975 West Walnut Street.
1B130, Indianapolis. IN 46202-5251. Phone: (317) 274-5738: Fax: (317) 274-1069;
E-mail: [email protected].
4 The
abbreviations
used
are:
LOH,
on multilocus
chromosomal
events
as determined
by LOH of
(to R. J. A.), and NIEHS Grant ES
Jefferson University, Philadelphia, PA 19107.
3 To
mainly
flanking MR markers and by FISH.
United
loss of heterozygosity;
MR.
microsatellite
repeat;
APRT, adenine phosphoribosyltransferase; HPRT, hypoxanthine guanine phosphoribo
syltransferase; tk, thymidine kinase; DAP, 2,6-diaminopurine; TG, 6-thioguanine; CE,
AND METHODS
T.Lymphocyte Isolation and Mutant Selection. An outline of the proto
col used for lymphocyte isolation and mutant selection is given in Fig. 1.
Human peripheral blood samples were collected by venipuncture from two
APRT-deficient
cloning efficiency; TCR, T-cell receptor; FISH, fluorescence in situ hybridization.
S A.
Sahota
families,
and
J.
A.
one from
Tischfield,
Belgium
unpublished
(family
1) and the other
observations.
1188
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
from
LOSS OF HEmRoz@oosrry
AND MITOTICRECOMBINATION
clones was digested with Hind!!!, and a Southern blot was hybridized with
TCR-y and -f3probes. Clones showing different RFLP patterns were consid
ered to be of independent origin and hence were considered to represent
Blood from APRT Heterozygote
(APRT +‘@,
HPRT )
I
Isolation
independent mutations. Any clones showing identical TCR gene patterns and
the same APRT mutation were considered to represent a TCR gene-defined
mutant set and to represent the progeny of a single mutational event, but they
were considered to be of independent origin if the APRT mutations were
of T Lymphocytes
different.
SelectIon of
Microsatellite Analysis ofDAP-resistant Clones. A number of MR mark
ers and their relative positions on chromosome 16 are shown in Table 2
(21—23).All of these markers were first tested for their informativeness
Mutant Cells arising
In vlvo
APRT-'-
(heterozygosity)
HPRT
2,6-Diaminopurine
(DAP)
6-Thiogua
Inc (TG)
with putative
@4S\@
mtragenic
point mutations
from those with allelic
loss. A second
informative repeat marker, Dl6S4l3, about 1.1 CMcloser to APRT, was used
for subsequent analyses (24). Primers for PCR were either purchased from
Growthof
DAPt or TG@
(Calculate
using DNAs from the whole blood of the two heterozygotes
from family 1. One informative repeat marker, D16S305, which is approxi
mately 6.6 CMproximal to the APRT gene, was first used to distinguish clones
Colonies
Research Genetics (Huntsville, AL) or synthesized in-house. The PCR reaction
Mutant Frequency)
conditions were as follows: approximately 100 ng of DNA, 100 ng of each
primer,
I
50 mM KC1, 10 mM Tns-HC1
(pH 8.3),
1.5 nmi MgCI2, 0.2 mM each
deoxyribonucleotide, 10 pCi of [a-32PJdC'FP (specific activity, 3000 Ci'
mmol), and 0.5 unit of Taq polymerase (Perkin-Elmer Corp., Foster City, CA)
in a total volume of 20 @l.Samples were processed through 30 cycles
Analysis of T Cell Receptor (TCR)Gene Rearrangement
consisting of 45 s at 95°C,45 s at 55°C,and 30 s at 72°C.Aliquots of the
amplified DNA were mixed with 2 volumes of loading gel buffer containing
95% formamide, heat-denatured at 98°Cfor 5 mm, and electrophoresed on a
Molecular Analysis of APRT LOH and HPRT MutatIon
5%
Fig. 1. A diagram of T-lymphocyte mutant selection. DAP- and TG-resistant
lymphocytes were selected as described in “Materialsand Methods.―
T
denaturing
TaqI RFLP
Germany (family 2), and from normal controls (Table 1). The mononuclear
cell fraction was isolated by sedimentation and washed, and the cells were
enumerated.
Isolated
lymphocytes
were immediately
plated
into culture
me
dium as described below for the selection of preexisting APRT-deficient
(APRr' or APRT@) T lymphocytes that, presumably, arose spontaneously
in vivo as a consequence of LOH (somatic alterations in the normal APRT
allele). Cells were incubated at a density of 1 X 106/ml in RPM! 1640
containing 20% nutrient medium HL-1, 5% bovine calf serum, and 1 mg/ml
purified phytohemagglutinin for 36—40h to promote mitogen stimulation.
This time is sufficient for optimal T-lymphocyte stimulation but is not suffi
cient for cell division. The cells were again sedimented, washed, and enumer
ated. The cells were reincubated at a density of 1—2X 104/well in the same
medium, this time containing 125 @g/mlphytohemagglutinin, optimal amounts
@
of interleukin 2 obtained from cell-free conditioned medium, irradiated (90
Gy) human lymphoblastoid feeder cells (an APRr'
cell strain designated C2
that has a single base substitution in the APRTgene) at 1 X i0@cells/well (18),
and 100
DAP. Initial experiments with cells from a homozygous APRT
deficient individual allowed optimization of the concentration of DAP for
selection ofAPRr'
cells. As shown in Fig. 1, lymphocytes were also seeded
into parallel cultures in similar medium containing 10 @.tM
TG to identify
TG-resistant clones for analysis ofHPRTgene mutations (19) and in drug-free
medium to determine the nonselection cloning efficiency.
Mutant Frequency Determinatiam. Isolated lymphocytes were assayed
for cloning efficiency by plating 1, 2, 5, and 10 cells/well in the absence of
DAP or TG and at 1—2
X iO@cells/well in the presence of DAP or TG. The
microliter dishes were incubated for 10—14
days, and growing colonies were
visualized using an inverted phase-contrast microscope. The CE was calculated
polyacrylamide
gel.
The
retention
of
CE in the presenceand absenceof selective agent(DAP or TG), respectively.
Mutant and wild-type (unselected) clones from two of the obligate APRr'
heterozygotes (the father and mother in family 1) were propagated to obtain a
sufficient number of cells for additional analyses. Genomic DNA was isolated
@
from 0.5—1.5x
cells from each of the DAP-resistant
clones. This DNA
for
in intron 2 ofAPRT
that can be assayed
by PCR (25). The mother
was not informative for the TaqI RFLP, but she was informative for a marker
(D16S303) about 2 CMdistal to APRT and for one or more polymorphic
markers located in the 5' or 3' flanking regions of the gene (26). Thus, the
presence or absence of two distinct APRT alleles in clones retaining heterozy
gosity for D16S305 or D16S413 could be confirmed. Clones that retained both
APRT alleles were analyzed for mutations by PCR amplification and sequenc
ing (either directly or after subcloning) of the APRT gene (27). LOH of
D16S305 or D16S413 suggested that there might be LOH elsewhere on
chromosome 16, and these clones were analyzed with additional MR markers.
Chromosome Analysis. Ten DAP-resistant clones from the father (family
1) with allelic loss of various regions of 16q were cultured, and metaphase
preparations and GTG-banding were produced by standard techniques. FISH
was done with an APRT-containing cosmid probe (a gift from D. F. Callan,
Adelaide, Adelaide Children's
Hospital, Australia).
Cosmids were purified
with a Qiagen kit (Chatsworth, CA) and biotin-labeled with the Bionick kit
(Life Technologies, Inc., Gaithersburg, MD). Chromosomes were denatured at
70°Cin 70% formamide in 2 X SSC for 2 mm, and hybridization was done
overnight at 37°C.Hybridization
signals were detected by applying FITC
avidin/anti-avidinlFlTC-avidin sandwich amplification. The presence of two
Table 1 APR@ and HPRT mutant cellfrequencies in human T-lymphocytes―
Themutantcellfrequencyis theratioofthe meanCEin selectivemedium(10g.u.@
TG
or 100 p@ DAP) divided by the mean CE in drug-free medium. CE was calculated by the
Poisson distribution, in which CE = InP@/N (where N equals the average number of
cells/well). The APRTgenotypes are indicated in the table. TheAPRTgerm-line mutations
in the families have been described (Refs. 15 and 17).
cell frequency
(Xl0@6
by the Poisson distribution in which CE = —InP@/N(in
which N represents the
average number of cells/well). The mutant frequency is the ratio of the mean
heterozygosity
D16S413 and D16S305 was interpreted as suggesting LOH localized to APRT
(e.g., intragenic mutation). The heterozygous father is also informative for a
DAPr1
Family
Subject (genotypes)Mutant
Father(APRT@')
Mother (APRT@')
2
Child (APRr'@)
Control (APRr'@)
Father (APRT@'@)
Mother (APRT@)
Child (APRr@)
Control (APRT@'@)10.5
was used for TCR gene rearrangement studies, mutational analysis of the
APRT gene, and analysis of LOH.
<0.27a
TCR Gene Rearrangements. Cell clonality and independenceof muta
HPRr.b
All individuals were
tions were determined by RFLP analysis of TCR gene rearrangements using drug-freemedium.
Values were approximately
two probes, TCR-@ and TCR-@ (20). Genomic DNA from DAP-resistant
I 189
TGr
21.8
7.8
6.8
49.7
4.2
6.2
9.4
1, equal to the colony-forming
153.9
21.3
097b
<0.12
24.3
115.1
1.02―
efficiency in
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
LOSS OF HETEROZYGOSITYAND MITOTICRECOMBINATION
Table 2 WH in DAP-resistant T-cell clones from APRT heterozygotes
LOHof MRmarkerson chromosome16in DAPrcells.Zygositywasassayedby PCRwithlocus-specificprimersfollowedby electrophoresisin 5%denaturingpolyacrylamide
gels. The most likely cytogenetic localization (21—24)of each repeat is shown above its designation (e.g., D16S283). A filled circle indicates no LOH, whereas an empty space indicates
observedLOH.
A. Father, family
locationl6pl3.3qll.2qll.2
ql3q24.2q24.3q24.3q24.3No.
of DAP@clones
(5398)(8422)(5305)(5413)(APRT)1
No. of independent mutations―
118
•••I
•6
••17
•5
...4
....15
•••••B.
17
1
6
16
5
4
15
Cytogenetic
q12.lq12.l
(5261)(S3O8)
(S283)(S298)(SPN)
•••
S••
•••
...
...
•I•
ql3
(Sl86)
••
..
..
.
.
•
•S
Mother, family Il6pll.2
No. of DAP@clones
(S303)4
No. of independent mutations@'
...I
••••I
••I•I
•I
•••4
S•••.a
(5420)ql2.l
4
I
I
I
I
4
q22.l
(5421)
(S265)Cytogenetic
(S261)ql3
••••
••••
....
location
q24.2
(S402)q24.3 (S305)q24.3 (5413)q24.3(APRT)q24.3
q24.2
(S422)
•
.
Number of independent mutations by TCR analysis (65 of 67).
b Number of independent mutations by TCR analysis (13 of 13).
copies of the chromosome
Thirteen DAP-resistant
16 centromeric region was confirmed by FISH with
DI6Z2, a chromosome 16-specific a-satellite probe (Oncor, Gaithersburg,
MD) as described above (28). At least 10 metaphases for each clone were
visualized by FISH.
RESULTS
Comparison of APRT and HPRT Mutant Frequencies. The
APRT and HPRT mutant frequencies in two unrelated APRT-deficient
families (father, mother, proband, and control) are shown in Table 1.
As expected, T cells from both the APRT'
children exhibited about
the same colony-forming efficiency in DAP medium as in drug-free
medium, indicating optimal selection conditions for APRT'
clones.
The 4 heterozygotes showed frequencies ranging from 21 X l0_6 to
154 X lO_6, a difference of about 7-fold. The frequencies of HPRT
clones in the same individuals (except for the father in family 2, who
exhibited an unusually high frequency of 50 X lO_6) varied within a
2—5-foldrange and are consistent with extensive previous data (19,
20). Thirteen clones from the mother and 67 clones from the father
from family I were propagated for LOH and mutation analyses.
Analysis of Heterozygosity. LOH of D16S413 was observed in 48
clones from the father, and retention of heterozygosity was observed
in the remaining 19 clones. Subsequent analysis of the 19 clones by
TaqI RFLP indicated the retention of both APRT alleles in 15 clones
and the retention of only 1 allele in 4 clones, suggesting a chromo
somal LOH of less than 5.5 CMin the latter group. Therefore, these 4
clones were grouped with the 48 DAP-resistant clones with LOH for
at least D16S413 (e.g., multilocus LOH). The 15 clones retaining both
APRT alleles were analyzed for intragenic point mutations (see be
low). The 52 clones (48 with LOH of D16S413 and 4 with LOH for
APRT) were tested with additional MRs. As shown in Table 2A, LOH
in these clones encompassed variable lengths of l6q (23). When LOH
was observed at a particular locus, the same allele was lost in all
instances, consistent with LOH of only those alleles that are physi
cally linked to the normal APRT in this particular heterozygote.
Analysis of ICR gene RFLPs indicated that 65 of 67 clones from the
father are likely to be the result of independent mutations (Table 2A).
analyzed
as described
clones from the mother in family 1 were
above.
Because
D16S303
was informative
in
this individual, we derived data concerning LOH distal to APRT. One
clone did not show LOH for Dl6S4l3 (which is 4 CMproximal to
APRT), and four clones did not exhibit LOH for either APRT or
D16S3O3. DNA sequencing indicated point mutations in all four
clones without APRT LOH (see below). As shown in Table 2B, LOH
in the remaining nine clones encompassed variable lengths of l6q (22,
23). When LOH was observed at a particular locus, the same allele
was lost in all instances,
as was the case with the father.
TCR gene
analysis indicated that each of the 13 clones is the result of an
independent mutation (Table 2B).
Chromosome
Analysis of Clones with Allelic Loss. GTG
banding revealed a normal karyotype (46,XY) in all 10 DAP-resistant
clones from the father in family 1 (data not shown). FISH analysis
indicated the presence of 2 APRT alleles in 9 clones that exhibited
LOH for APRT and D16S413 and the presence of only 1 allele in the
remaining clone (29 of 33 metaphases). The clone exhibiting only one
APRT allele (clone Fl6) was one of four in which LOH did not
include D16S413. Fig. 2 shows a typical FISH metaphase from clone
Fl6. Because D16S303, which is distal to APRT, was not informative
in this individual, we could not determine the extent of allelic loss in
the telomeric region of 16q. However, karyotype analysis of these 10
clones did not suggest terminal deletions. These results, along with the
MR LOH data, suggest that interstitial deletion occurred in only 1 of
the 10 clones that were cytogenetically analyzed and that multilocus
LOH in the other 9 clones was not a consequence of l6q deletion or
deletion and translocation.
Clones with Point Mutations. As indicated above, 15 clones from
the father and 4 clones from the mother in family 1 did not show LOH
for APRT or flanking markers. DNA sequencing of the previously
wild-type APRT alleles in these clones revealed point mutations in
nine clones from the father and in all four clones from the mother. The
results are summarized in Table 3. There were four pairs of clones that
exhibited the same point mutation, but in each case, TCR gene
rearrangement analysis indicated that the mutations were of independ
1190
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
LOSS OF HETEROZYGOSITYAND MITOTICRECOMBINATION
4
Fig. 2. Metaphase from clone Fl6 that exhibits
only APRT LOH. Only a single hybridizing region
was observed (arrowhead). FISH with an APRT
containing cosmid was performed as described in
“Materialsand Methods.―
a
ent origin. The entire APRT gene was sequenced in one of the six
clones from the father that exhibited no mutation. In the other five
clones that exhibited no mutation, only the coding regions and splice
junctions were sequenced.
Clonal Origin of Mutant Clones. A comparison of TCR gene
rearrangement RFLPs and the nature of mutation in DAP-resistant
clones revealed that most clones (78 of 80) were the result of inde
pendent mutations. Several groups of clones, however, merit com
ment. For example, two clones yielding identical TCR gene rearrange
ment patterns exhibited two different mutations (point mutation and
allelic loss, respectively). Among three clones with identical TCR
RFLP patterns, allelic losses of two different sizes and a point muta
tion were observed. These five clones (6%) may represent oligo
clonality of mutants (two or more mutants with the same TCR gene
rearrangement pattern). This estimate of oligoclonality is similar to
that reported for HPRT mutations (29). There were two pairs of clones
(5%) with identical TCR gene rearrangement RFLPs and identical
mutations, and they are likely to represent true sibling clones.
markers along the length of chromosome 16. Because these heterozy
gotes are the parents of APRT-deficient probands, the germ-line
mutation in each is known (15, 17). The focus of this study is in vivo
somatic mutations that arise spontaneously in the T cells of these
heterozygotes and produce LOH. T lymphocytes are well suited to
these studies because they can be easily obtained from peripheral
blood and propagated in vitro as clones from a single cell that has
previously undergone in vivo somatic mutation. Furthermore, as was
shown in this study and previously (6), T lymphocytes grown in vitro
maintain a normal karyotype. We have also measured the frequency of
mutation
locus
in the same
individuals
We have developed an assay for determining the mechanisms of in
vivo LOH in normal human T cells based on individuals that are
heterozygous for a selectable telomeric marker (APRT) and MR
that produce
in vivo LOH
and thus
provide
insight
mutations in DAP-resistant
T-cell
LOHofAPRT.
clones were analyzed for either an informative Taql restriction fragment polymorphism (25) or flanking polymorphism (26) to determine whether or not there had been
fromthe In 13 of 19 clones without LOH ofAPRT, specific point mutations could be identified within APRT. No mutations were observed in six clones. F and M indicate clones
father
(27).Clone
and mother in family 1, respectively. The intron/exon, codon, and nucleotide numbers of the protein coding regions of APRT are as described previously
ChangeP2,
No. of independent mutations―
Mutation
@
StopM5,
FlO
2
C
@
ThrF33
M14
FrameshiftMl
errorM4
1
error?F28,
2
1
1
1
MetF30
F42
@
@
@
errorF63
MetF19,
liea
F4l
T
for
we have used polymor
into
molecular mechanisms.
Among 80 DAP-resistant T-lymphocyte clones from two heterozy
gotes, 61 (76%) were a consequence of allelic loss. Of the remaining
19 clones, 13 were demonstrated to be and 6 are likely to be the result
of point mutations in APRT. Despite a 7-fold higher frequency of
DISCUSSION
Table 3 APRTpoint
HPRT
to the APRT data. Furthermore,
phic MR markers to initially identify DAP-resistant clones with
either chromosomally extensive or localized LOH. Subsequently,
the clones were further characterized with additional CA repeat
markers, FISH analysis, and DNA sequencing ofAPRT. Therefore,
this assay could discriminate between most types of mutational
changes
clonesDAPr
at the X-linked
comparison
Intron/exon no.
Codon
El
7
G
A
del CC
A @‘
C
C @‘
A
E2
E2
E2
13
37
56-57
62
2
G
A
E3
I
1
2
G
0
C
@‘
A
A
T
E3
E4
E5
Nucleotide no.
19
CAG @s
TAG, Arg
272
331—332
349
133 1
GCC
ACC, Ala
84
1406
GTG
ATG. Val
107
I 26
135
1477
1806
2063
GTG
ACC
Splice
ATG, Val
ATC, Thr
@s
Splice
Splice
Number of independent mutations by TCR analysis.
1191
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
LOSS OF HEmRoz@oosrrY
AND MITOTICRECOMBINATION
mutant clones in the father as compared to the mother in family 1
(Table 1), the fraction of clones due to large chromosomal events was
relatively constant (0.77 and 0.69, respectively) in each. These results
are similar to our observations in T cells from three Japanese APRT
heterozygotes (17). The frequency of loss of APRT alleles is consist
ent with that observed at other loci. LOH of approximately 73% was
observed in primary retinoblastomas (30). Spontaneous tk-deficient
clones from a heterozygous human TK6 lymphoblastoid cell line
exhibited a 71% frequency of LOH (31). As expected, hemizygous
Aprt in CHO cells exhibited a lower (43%) spontaneous frequency of
allelic loss (32). Although the proportion of mutant cells due to allelic
loss may be similar in vitro and in vivo, the relative frequencies of
mutants due to particular molecular mechanisms may differ. For
example, karyotypic instability, including chromosome loss and trans
location, may be a major cause of DAP-resistant clones in APRT@'
human heteroploid MR12-1 fibrosarcoma cells (33)•6
High frequency
of APRT LOH has also been reported in human TK6 lymphoblasts,
but in 36 of 38 clones it began at 16ql2, and translocations were
detected in 20% of clones (34).
Linked recessive lethal alleles likely play a key role in limiting the
extent of allelic loss in cells. A study of two clonally related cell lines
indicated that one had a much higher rate of LOH (96% of clones)
than the other (47% of clones). Furthermore, LOH was observed to
extend 98 CMin the former but only 4 CMin the latter. These results
were explained by postulating a relatively essential heterozygous gene
whose normal allele is linked to the normal tk allele in the cells with
reduced LOH (12—14).The initial observation of no deletions distal to
APRT in human 5W620 cells and an open reading frame downstream
of APRT that was noted by our group (Genbank accession number,
U04709)7 led to the description of a unique heterozygous transcribed
gene downstream ofAPRT, in the opposite orientation (1 1). This gene
is postulated to provide a function that is essential to cellular survival.
Allelic loss extending through all regions of 16q in our T-cell clones
would suggest the absence of a heterozygous recessive lethal gene
whose normal allele is linked to the functional APRT allele in these
two individuals. Also, the similar proportions of clones with point
mutations or larger chromosomal events from the mother and father in
family 1 suggest that the observed 7-fold difference in mutation
frequency
is not due to linked
recessive
breast cancers (36—38).LOH for Dl6S4l3 was shown to be 25% for
ductal carcinoma in situ, a preinvasive lesion, and 43% for invasive
ductal carcinoma, suggesting that LOH of 16q24.3 is an early event in
the development of breast cancer (39). There are a number of common
fragile sites located on l6q that have been implicated in allelic loss
during carcinogenesis. For example, in human hepatocellular carci
noma, LOH is common between l6q22.l and 16q22.3—q23.4 (40).
We speculate that LOH in l6q in T cells is without phenotypic
consequence because T cells are inherently apoptotic or because genes
on 16q that are involved in breast cancer are not active in T cells.
Because LOH encompasses different regions of 16q in our DAP
resistant
of mitotic
recombination
that
REFERENCES
1. Knudson, A. G., Jr. Hereditary cancer, oncogenes and antioncogenes. Cancer Res.,
45: 1437—1443,1985.
2. Loeb, L. A. Mutator phenotype may be required for multistage carcinogenesis. Cancer
Res.,51:3075—3079,
1991.
3. Stanbridge, E. J. Human suppressor genes. Annu. Rev. Genet., 24: 615—657, 1990.
4. Levine, A. J. The tumor suppressor gene. Annu. Rev. Biochem., 62: 623—651, 1993.
5. Meuth, M. The structure of mutation in mammalian cells. Biochim. Biophys. Acts.
1032: 1—17,
1990.
6. Albertini, R. J., Nicklas, J. A., O'Neill, J. P., and Robison, S. H. in vivo somatic
mutations in humans: measurement and analysis. Annu. Rev. Genet., 24: 305—326,
1990.
7. Phear, G., Armstrong, W., and Meuth, M. Molecular basis of spontaneous mutation
at the Aprt locus of hamster cells. J. Mol. Biol., 209: 577—582,1989.
8. de Jong, P. J., Grosovsky, A. J., and Glickman, B. W. Spectrum of spontaneous
mutation at the APRT locus of Chinese hamster ovary cells: an analysis at the DNA
sequence level. Proc. Natl. Acad. Sci. USA, 85: 3499—3503,1988.
9. Yandell, D. W., Dryja, T. P., and Little, J. B. Somatic mutations at a heterozygous
lethal genes.
Mitotic recombination seems to be the major pathway for in vivo
LOH. All 10 DAP-resistant clones with variable allelic loss, encom
passing primarily APRT and the distal portion of 16q up through most
of l6q, showed normal chromosomes, and nine clones showed two
APRT alleles by FISH. The only plausible explanation for these data
is mitotic recombination because gene conversion is not known to
affect such large regions. At least one maternal clone, however, is
likely due to whole-chromosome loss or translocation because it also
exhibited LOH for a marker on l6p (Table 2B). Mitotic recombination
also seems to produce about 30% of lymphocytes that have lost the
expression of one codominant HLA-A allele on chromosome 6 (35).
Furthermore, we have observed that mouse primary fibroblasts exhibit
a high frequency of Aprt LOH that is mostly due to mitotic recom
bination.8 Thus, the phenomenon is not restricted to T lymphocytes.
The observation of high-frequency mitotic recombination in normal
cells has significant implications for the etiology of cancer. It suggests
that LOH caused by mitotic recombination may trigger carcinogenesis
or at least be an early event in this process. Several groups have
reported LOH for at least two regions of 16q in about 50% of primary
T-cell clones as a consequence
seems to begin within different chromosomal regions, a set of such
clones could be used to establish the order ofpolymorphic loci on l6q.
With a sufficient number of loci, a map of the distances between loci,
based on mitotic recombination, could be produced and compared to
the meiotic map. Our T-cell data suggest an order for polymorphic CA
repeats that is consistent with their known chromosomal locations, as
indicated in the top row of Table 2, A and B (22). Whether or not
events other than mitotic recombination skew the data is not known.
It would be of interest to examine LOH due to mitotic recombination
in individuals with different heterozygosity for flanking markers and
in different tissues.
autosomal locus in human cells occur more frequently by allele loss than intragemc
structural alteration. Somatic Cell Mol. Genet., 12: 255—263,1986.
10. Amundson, S. A., Fortunato, J. E., and Liber, H. L. Heritable alterations at the
adenine phosphoribosyltransferase (APRT) locus in human lymphoblastoid cell lines.
Mutat. Res., 284: 287—295,1992.
11. Harwood, J., and Meuth, M. Deletion mapping of highly conserved
transcribed
sequence downstream from APRT locus. Somatic Cell Mol. Genet., 21: 151—160,
1995.
12. Dobo, K. L., Giver, C. R., Eastmond, D. A., Rumbos, H. S., and Grosovsky, A. J.
Extensive loss of heterozygosity accounts for differential mutation rate on chromo
some 17q in human lymphoblasts. Mutagenesis, 10: 53—58,1995.
13. Amundson, S. A., and Liber, H. L. A comparison of induced mutation at homologous
alleles of the fit locus in human cells. Mutat. Res., 247: 19—27,1991.
14. Amundson S. A., and Liber H. L. A comparison of induced mutation at homologous
alleles of the tk locus in human cells. II. Molecular analysis of mutants. Mutat. Res.,
267: 89—95,
1992.
15. Simmonds, H. A., Sahota, A., and Van Acker, K. J. Adenine phosphoribosyltrans
ferase deficiency and 2,8-dihydroxyadenine lithiasis. In: C. R. Scriver, A. L. Beaudet,
W. S. Sly, and D. Valle (eds.), The Metabolic and Molecular Bases of Inherited
Disease, 7th edition, pp. 1707—1724.
New York: McGraw-Hill, Inc., 1995.
16. Broderick, T. P., Schaff, D. A., Bertino, A. M., Duals, M. K., Tischfield, J. A., and
Stambrook, P. J. Comparative anatomy of the human APRT gene and enzyme:
nucleotide sequence divergence and conservation of a nonrandom CpG dinucleotide
arrangement. Proc. Nati. Aced. Sci., USA, 84: 3349—3353,1987.
17. Chen, J., Sahota, A., Martin, G. F., HakOda,M., Kamatani, N., Stambrook, P. J., and
Tischfield, J. A. Analysis of germline and in vivo somatic mutations in the human
APRT gene: mutational hot spots at the intron 4 splice donor site and at codon 87.
Mutat. Res., 287: 217—225,1993.
6 C. Shao,
P. K. Gupta,
Y. Sun,
A. Sahota,
and J. A. Tischfield,
Complex
chromosomal
mechanisms lead to APRT loss of heterozygosity in heteroploid cells. Cytogenet. Cell
Genet., in press.
7 5. A.
S C.
Boyadjiev,
Shao
and
J.
A.
A.
Sahota,
Tischfield,
and
J. A.
unpublished
Tischfield,
unpublished
18. Klinedinst, D. K., and Drinkwater, N. R. Reduction to homozygosity is the predom
inant spontaneous mutational event in cultured human lymphoblastoid cells. Mutat.
Ret., 250: 365—374,1991.
19. O'Neill, J. P., McGinniss, M. J., Berman, J. K., Sullivan, L. M., Nicklas, J. A., and
Albertini, R. J. Refinement of a T-lymphocyte cloning assay to quantify the in vivo
thioguanine-resistant mutant frequency in humans. Mutagenesis, 2: 87—94,1987.
observations.
observations.
1192
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
LOSSOF nnnozy@osn'y
4@mM@FOTlC
RECOMBINATION
20. Nicklas, J. A., O'Neill, J. P., and Albetini, R. J. Use of TCR gene probe to quantify
the in vivo HPRT mutations in human 1-lymphocytes. Mutat. Ret., 173: 67—72,
1986.
21. Gyapay. G., Morissette, J., Vignal, A.. Dib, C.. Fizames, C., Millasseau, P., Marc. S.,
Bernardi, G., Lathrop, M., and Weissenbach, J. The 1993-1994 Genethon human
genetic linkage map. Nat. Genet., 7: 246—339,1994.
30. Cavenec, W. K., D'yjs. T. P., Phillips, R. A., Benedict, W. F., Godbout, R., Gallie,
B. L., Murphree, A. L, Strong, L C., and White, R. L Expression ofrecessive alleles
by chromosomal mechanisms in retinoblastoma. Nature (Lond.), 305: 779-784,
1983.
31. Li, C-Y., Yandell, D. W., and Little, J. B. Molecular mechanisms of spontaneous and
22. Doggen, N. A., Goodwin, L A., Tesmer, L J., Meincke, L 1., Bruce, D. C., and
induced loss ofheterozygosity in human cells in vitro. Somatic Cell Mol. Genet., 18:
Clark, L. M., Altherr, M. R., Ford, A. A., @hi,
H. C., Marrone, B. L., Longmire, J. L..
77—87,1992.
32. Ward, M. A., Yu, M., Glickman, B. W., and Orosovsky, A. J. Loss of heterozygosity
in mammalian cell mutagenesis: molecular analysis of spontaneous mutations at the
Lane, S. A., Whitmore, S. A., Lowenstein, M. 0., Sutherland, R. D., Mundt, M. 0.,
K.iiill, E. H., Bruno, W. J., Macken, C. A., Torrey, D. C., Wu, J-R., Griffith, J.,
Sutherland, G. R., Deaven, L L, Callen, D. F., and Moyzis, R. K. An integrated
APRT locus in CHO cells. Carcinogenesis (Lond.), 11: 1485—1490,
1990.
physical map of human chromosome 16. Nature (Lond.), 377 (suppl.): 335-365,
33. Thu. Y., Stambrook, P. J., and Tischfield, J. A. Loss of heterozygosity:
B lymphoblastoid cell line. Somatic Cell Mol. Genet., 19: 515—527,1993.
35. Morley. A. A., Grist, S. A., Turner, D. R., Kutlaca, A., and Bennett, G. Molecular
Richards, R. I. Isolation and characterization of(AC)n microsatelhte genetic markers
from human chromosome 16. Genomics, 13: 402—408,1992.
nature of in vivo mutations in human cells at the autosomal HLA-A locus. Cancer
Res., 50: 4584—4587,1990.
25. Chen, J., Saints, A., Laxdal, T., Scrine, M., Bowman, S., Cal, C., Stambrook, P. 1.,
andTischfield,J. A.Identificationofa singlemissensemutationin theAPRTgenein
five Icelandic patients and a British patient. Am. J. Hum. Genet.. 49: 1306—1311,
36. Tsuda, H., Callen, D. F., Fukutomi, T., Nakamura, Y., and HirOhaShi,S. Allele loss
on chromosome 16q24.2-qter occurs frequently in breast cancers irrespectively of
1991.
differences in phenotype and extent of spread. Cancer Ret., 54: 513—517,1994.
37. Cleton-Jansen, A. M., Moerland, E. W., Kuiperes-Dijkshoorn, N. J., Callen, D. F.,
Sutherland, G. R., Hansen, B., Devilec, P., and Corndisse, C. J. At least two different
26. Boyadjiev, S. A., Sahota, A., and Tischfield, J. A. Identification and application of
polymorphisms flanking the human adenine phosphoribosyltransferase gene. Hum.
Mutat., 8: 214—215,1996.
regions are involved in alleic imbalance on chromosome arm l6q in breast cancer.
Genes Chromosomes & Cancer, 9: 101-107, 1994.
27. Chen, J., Saints, A., Stambrook, P. J., and Tischfield, J. A. Polymerase chain reaction
amplification and sequence analysis of human mutant adenine phosphoribosyltrans
ferase genes: the nature and frequency of errors caused by Taq DNA polymerase.
Mutat. Ret., 249: 169—176,1991.
38. Dorion-Bonnet, F., Mautalen, S., Hostein, I., and Longy, M. Alleic imbalance study
of l6q in human primary breast carcinomas using microsatellite markers. Genes
Chromosomes & Cancer, 14: 171-181, 1995.
39. Aldaz,C. M., Chen,T., Sahin,A., Cunningham,J., and Bondy,M. Comparative
28. Greig, G. M., England, S. B., Bedford, H. M., and Willard, H. F. Chromosome
specifica-satelliteDNAfromthecentromereofhumanchromosome16.Am.J. Hum.
allelotype of in situ and invasive human breast cancer: high frequency of microsat
ellite instability in lobular breast carcinomas. Cancer Res., 55: 3976—3981, 1995.
Genet., 45: 862—872,1989.
29. O'Neill, J. P., Nicklas, J. A., Hunter, T. C., Batson, 0. D., Allegretta, M., Falls. M. T.,
Brands, R. F., and Albcrtini, R. J. The effect of T-lymphocyte ‘clonality'on the
calculated HPRT mutant frequency occurring in vivo in humans. Mutat. Res., 313:
215-225,1994.
the most
frequent cause of recessive phenotype expression at the heterozygous human adeninc
phosphoribosyltransferase locus. Mol. Carcinog., 8: 138—144,1993.
34. Smith, L. E., and Grosovsky, A. 1. Genetic instability on chromosome 16 in a human
1995.
23. Weber, J. L. Informativeness ofhuman (dC-dA)n.(dG-dT)n polymorphisms. Genom
ics, 7: 524—530, 1990.
24. Thompson, A. D., Shen, Y., Holman, K., Sutherland, G. R., Callen, D. F., and
40. Tsuda, H., Zhang, W. D., Shimosato, Y., Yokota, J., Terada, M., Sugimura, T.,
Miyamura, T., and Hirohashi, S. Allele loss on chromosome 16 associated with
progression of human hepatoccllular carcinoma. Proc. Natl. Aced. Sci. USA, 87:
6791—6794,1990.
1193
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.
High Frequency in Vivo Loss of Heterozygosity Is Primarily a
Consequence of Mitotic Recombination
Pawan K. Gupta, Amrik Sahota, Simeon A. Boyadjiev, et al.
Cancer Res 1997;57:1188-1193.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/57/6/1188
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on May 10, 2016. © 1997 American Association for Cancer Research.