Oncogene (1999) 18, 79 ± 85
ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00
http://www.stockton-press.co.uk/onc
Aberrant FHIT mRNA transcripts are present in malignant and normal
haematopoiesis, but absence of FHIT protein is restricted to leukaemia
Uwe R Peters1,2, Urs Hasse1, Elisabeth Oppliger1, Mario Tschan1, S Tiong Ong5,
Feyruz V Rassool5, Bettina Borisch4, Andreas Tobler3 and Martin F Fey1,2
1
Department of Clinical Research, the 2Institute of Medical Oncology and the 3Central Haematology Laboratory, University and
Inselspital, Berne, Switzerland; the 4Department of Pathology, University of Geneva, Switzerland; the 5Section of Haematology/
Oncology, University of Chicago Medical Center, Chicago, Illinois, USA
Alterations of the recently cloned fragile histidine triad
(FHIT) gene at chromosome 3p14.2 are frequent in a
variety of solid tumours and cancer cell lines. Based on
these ®ndings, FHIT has been proposed as a putative
tumour-suppressor gene. We evaluated the mRNA
expression of the FHIT gene in samples from 55
patients with various haematological malignancies (21
AML, 8 CML, 10 CLL, seven low-grade and nine highgrade Non-Hodgkin's lymphomas), in a panel of 16
leukaemia cell lines, in normal mature haematopoietic
cells of both myeloid and lymphoid lineage, as well as in
CD34+ haematopoietic progenitor cells. Aberrant FHIT
mRNA transcripts were observed in 14/16 (88%)
leukaemia cell lines, 43/55 (78%) primary haematological neoplasms, but also in 17/22 (77%) normal
controls. 1/16 (6%) cell lines and 7/55 (13%)
neoplasms did not express any FHIT mRNA. cDNA
sequencing revealed exonic deletions, small DNA
insertions and combinations of both. Analysis of
genomic DNA showed gene deletions in two myeloid
leukaemia cell lines. In contrast to all normal types of
haematopoietic cells, FHIT protein was clearly reduced
or absent in 8/18 (44%) neoplastic samples tested. Our
data indicate that whilst aberrant FHIT mRNA
transcripts are seen both in normal and malignant cells,
lack of FHIT protein is restricted to leukaemia. Absent
FHIT protein expression might contribute to leukaemogenesis.
Keywords: FHIT; FRA3B; leukaemia; lymphoma
Introduction
Chromosome 3p often shows deletions and translocations in human cancer, particularly in carcinomas of
the breast, lung, head and neck and kidney (Brauch et
al., 1987; Kastury et al., 1996; Mitelman et al., 1997;
Zbar et al., 1987). Chromosome and DNA transfer of
di€erent chromosome 3p regions were shown to
suppress the tumour phenotype of cancer cell lines
(Killary et al., 1992; Sanchez et al., 1994). Based on
these observations the short arm of chromosome 3 is a
candidate region for the presence of tumour-suppressor
genes. Recently, the fragile histidine triad (FHIT) gene
was identi®ed at the chromosomal region 3p14.2
Correspondence: MF Fey
Received 19 March 1998; revised 2 July 1998; accepted 3 July 1998
encompassing the most common inducible human
fragile site FRA3B (Ohta et al., 1996; Ong et al.,
1997; Rassool et al., 1996; Zimonjic et al., 1997).
Fragile sites are chromosomal regions prone to genetic
instability, chromosomal breakage and structural
rearrangements and thus their disruption could
contribute to carcinogenesis (LeBeau and Rowley,
1984; Yunis and Soreng, 1984). The FHIT gene
consists of 10 exons dispersed over about 1 Mb and
encodes a 1.1 kb mRNA. The open reading frame
spans exon 5 to exon 9, and translation results in a
small 16.8 kDa protein with diadenosine triphosphate
hydrolase activity possibly involved in nucleotide
metabolism (Barnes et al., 1996; Lima et al., 1997).
Genomic alterations of the FHIT gene in tumours
include loss of heterozygosity identi®ed by polymorphic DNA markers, deletions of both alleles as
well as insertion of various small DNA sequences
between and within the gene's exons. Aberrant
transcripts of various sizes are frequent in solid
tumours, but may also be seen in normal nonneoplastic tissues which do not exhibit alterations of
the FHIT gene (Fong et al., 1997; Negrini et al., 1996;
Ohta et al., 1996; Sozzi et al., 1996). Possible genetic
mechanisms giving rise to these fragments include
abnormalities of RNA splicing and processing as well
as exonic deletions. The extent to which aberrant
transcripts are translated into FHIT protein in human
tumours is unclear at present (Druck et al., 1997;
Otterson et al., 1998; Sozzi et al., 1997).
In contrast to solid tumours haematological
neoplasms have been less extensively characterized
with respect to alterations of the FHIT gene and its
mRNA and protein expression. Aberrant FHIT
mRNA transcripts were reported in chronic myelogenous leukaemia (CML) (Carapeti et al., 1996; Sugimoto
et al., 1997), acute myelogenous leukaemia (AML)
(Carapeti et al., 1996; Lin et al., 1997; Sugimoto et al.,
1997), acute lymphocytic leukaemia (Sugimoto et al.,
1997) and chronic lymphocytic leukaemia (CLL)
(Gayther et al., 1997; Sugimoto et al., 1997).
However, a systematic survey of various haematological neoplasms with respect to FHIT mRNA and
protein expression patterns and a comparison with
their respective normal cellular counterparts has not
been reported. We therefore investigated FHIT mRNA
and protein expression in a panel of leukaemia and
lymphoma cell lines and in clinical samples of
leukaemias and lymphomas. For comparison we
analysed mature normal haematopoietic cells of both
myeloid and lymphoid lineage as well as CD34+
haematopoietic progenitor cells.
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
80
Results
FHIT mRNA expression and gene structure in normal
haematopoietic cells
We investigated FHIT mRNA expression in ®ve
granulocyte, in ®ve monocyte and in eight lymphocyte
samples from peripheral blood of healthy blood
donors, as well as in four specimens of CD34+
haematopoietic progenitor cells. The 702-bp wild-type
FHIT transcript was present in all samples (Figure 1).
PCR products derived from this transcript invariably
gave a stronger signal than aberrant transcripts. All wt
702-bp PCR products sequenced corresponded to the
normal FHIT gene sequence. Overall we detected 23
abnormally sized di€erent FHIT mRNA transcripts
(Figure 2). At least one abnormally-sized transcript
was seen in three granulocyte, three monocyte, seven
lymphocyte and in all four CD34+ progenitor cell
samples. Only six of these aberrant transcripts had
been detected on visual inspection of ethidium
bromide-stained agarose gels prior to Southern
blotting of PCR products. Seven samples showed one
aberrant transcript only, but in ten samples multiple
smaller- and larger-than-wt transcripts were seen. The
most frequent aberrant transcript of about 780 bp was
seen in 3/10 myeloid, 5/8 lymphoid and in all four
CD34+ progenitor cell specimens. This PCR product
was isolated from agarose gels of three separate
samples (normal lymphocytes, but also one CLL and
one high-grade NHL), and sequenced. It revealed
di€erent transcripts of 771 bp or 774 bp. Nine
di€erent abnormally-sized transcripts ranging from
230 ± 851 bp in length derived from normal monoand lymphocytes were sequenced. A splice variant
lacking the ®rst 11 bases of exon 10 in the 3'
untranslated region (691-bp transcript) was detected
in three samples (Mao et al., 1996). The most frequent
aberrant fragments ranged from 742 ± 774 bp and
revealed DNA insertions of 40 ± 72 bp between exons
4 and 5. A previously described 72-bp insertion shared
a 76% homology to the Notch 4 gene (Hayashi et al.,
1997). In a normal monocyte sample we found a novel
53-bp insertion between exon 8 and 9 localized inframe within the open reading frame of the FHIT
coding region. The inserted sequence was nonrepetitive and unrelated to the FHIT gene or any
other gene sequence deposited in the EMBL data base.
Other novel insertions of 40-, 47- and 149-bp, located
outside the FHIT open reading frame, were seen in one
monocyte and one lymphocyte sample. We did not ®nd
larger bands created through heteroduplex formation
as previously reported (Fong et al., 1997; Michael et
al., 1997). Shorter FHIT transcripts resulted from loss
of entire exons generating splice junctions between
unrelated exons: exons 3 and 9 (230-bp transcript),
exons 3 and 8 (310-bp), exons 4 and 8 (394-bp) and
exons 7 and 9 (623-bp). The latter transcript was
identical to an exon 8 deletion variant observed in
normal lung and breast tissue (Ahmadian et al., 1997;
Fong et al., 1997). Assessment of genomic DNA by
Southern blotting showed no FHIT gene alterations in
any normal haematopoietic cells including those
samples displaying aberrant FHIT mRNA transcripts.
FHIT protein analysis in normal haematopoietic cells
The 16.8 kDa FHIT protein was detected by Western
blot analysis in all normal mature haematopoietic cells
of both myeloid and lymphoid lineage (three granulo-
Figure 1 RT ± PCR agarose gel (a) and PCR product blot hybridization (b) from corresponding malignant and normal
hematopoietic fresh cell samples and cell lines. Both the 702-bp wild-type (wt) FHIT transcript as well as a variety of aberrant
transcripts are seen in normal and in neoplastic cells. The sensitivity of detecting rare aberrant transcripts by Southern blot
hybridization of RT ± PCR products is higher
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
81
Figure 3 Western analysis of FHIT protein expression in normal
and malignant hematopoietic cells using a polyclonal rabbit antiFHIT IgG antibody. Human embryonal kidney cell line 293
(positive control; lane 1), gastric adenocarcinoma cell line Kato
III (negative control; lane 2). (a) Myeloid cell samples including
cell lines K562 (Lane 3), U937 (Lane 4), MV4-11 (Lane 5), and
cases AML 27 (Lane 6), AML 49 (Lane 7), AML 52 (Lane 8),
CML 135 (Lane 9), and normal granulocyte and monocyte
samples (Lanes 10 and 11). Several myeloid leukaemias do not
show any protein. (b) Cell samples of lymphoid lineage including
cell lines Raji (Lane 3), CEM (Lane 4) and Nalm6 (Lane 5), cases
CLL 2 (Lane 6), CLL 4 (Lane 7) and healthy donor lymphocyte
samples (Lanes 8-10). The two lymphoid leukaemias as well as the
normal lymphoid cells show the 16.8 kDa FHIT protein
Figure 2 Schematic representation of aberrant FHIT mRNA
transcripts sequenced from our series of normal and malignant
hematopoietic cells. Transcript sizes seen at PCR are indicated to
the left. Numbered boxes represent FHIT exons. The wild-type
FHIT cDNA including positions of primers, start and stop codon
and the normal 702-bp PCR product are depicted at the top.
Representative samples from which the data were obtained are
listed to the right. Normal controls are underlined
(Ly=lymphocytes; Mo=monocytes). Slashes indicate exon
deletions and shaded boxes represent DNA sequence insertions.a=alternative splice of exon 10 (absence of ®rst
11 bp);b=FRA3B sequence;c=77% homology to the ataxia
telangiectasia (ATM) gene;d=®rst 103 bp of FHIT intron
5;e=DNA sequences of uncertain origin;f=76% homology to
the NOTCH4 gene;g=duplication of FHIT exon 5
cyte, one monocyte and four lymphocyte samples). No
abnormal protein bands were seen in any of the cases
including samples with aberrant FHIT mRNA transcripts (Figure 3). No Western blots were performed on
CD34+ progenitor cells due to lack of suitable
material.
FHIT mRNA expression and gene structure in
haematological malignancies
The 702-bp wt transcript was present in 15/16
leukaemia cell lines (94%) and in 48/55 patients
(87%). All 702-bp FHIT transcripts sequenced
corresponded to the expected wt sequence without
any mutations. No FHIT message was detected in the
cell line Nalm6 and in fresh samples from one AML,
four Lg-NHL and two Hg-NHL. Overall we detected
33 di€erent abnormally sized FHIT transcripts ranging
from 240 ± 1700 bp. Such aberrant FHIT mRNA
transcripts consisting of both smaller- and/or largerthan-wt sizes were seen in 14/16 leukaemia cell lines
(nine myeloid, ®ve lymphoid), 18/21 AML, 7/8 CML,
9/10 CLL, 2/7 Lg-NHL, as well as in 7/9 Hg-NHL.
The vast majority of cases harboured more than one
aberrant transcript (range 1 ± 9). All types of aberrant
FHIT transcripts were seen in conjunction with the wt
702-bp product. Except for two myeloid cell lines
(K562 and U937) the band intensity of the wt fragment
was always stronger than or equal to aberrant
transcripts. As in normal cells the 780-bp mRNA
fragment was most frequent (38/55 cases). In three
cases (K562, Hg-NHL 5, Hg-NHL 9) we found the
same splice variant of exon 10 seen in normal
haematopoietic cells. Sequencing of aberrant transcripts revealed a broad variety of FHIT cDNA
deletions or splice variants as well as DNA sequence
insertions and combinations of both (Figure 2).
Smaller-than-wt transcripts were due to exon deletions
(241-bp, 394-bp, 634-bp transcripts) or contained small
DNA insertions replacing FHIT exons (464-bp, 476bp, 547-bp and 558-bp transcripts). The 547- and 558bp transcripts resulted from use of a cryptic splice site
103 bp within intron 5. Larger-than-wt cDNAs
contained insertions of di€erent size and origin
adjacent to exons 3, 4, 5 and 7. The same 72-bp
insertion seen in one of our lymphocyte samples was
also detected in a CLL and a high-grade NHL.
Southern blotting of genomic DNA did not reveal
any abnormal restriction fragments in neoplasms with
abnormal FHIT mRNA expression and/or with
undetectable FHIT message (one cell line, one AML,
two Hg-NHL, three Lg-NHL) suggesting an intact
FHIT gene locus. As an exception, DNA from the
myeloid leukaemic cell lines K562 and U937 showed
marked reduction of the intensity or virtual absence of
restriction fragments. This points to the presence of
large deletions including most exons. XbaI digests
revealed an extra 9.4 kb fragment in two normal and
two leukaemia samples suggesting the presence of a
pseudogene (Druck et al., 1997).
FHIT protein analysis in haematological malignancies
We studied FHIT protein expression in six myeloid cell
lines (K562, U937, NB4, MV4-11, KG1, HL60), three
lymphoid cell lines (Raji, CEM, Nalm6) and fresh
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
82
samples from six AML, one CML and two CLL
patients (Figure 3). However, no protein extracts
suitable for Western blotting were available from the
lymphomas. The FHIT protein of the expected size was
only present in neoplasms expressing the wt 702-bp
transcript. It was detected in 4/6 myeloid and 2/3
lymphoid cell lines, as well as in 4/6 AML, in the CML
and in both CLL samples. However, FHIT protein
expression was very weak in one myeloid leukaemia
cell line (MV4-11), one CML and two AML cases.
Three cell lines (K562, U937, Nalm6) and one AML
patient did not express any detectable FHIT protein.
We did not observe any abnormally-sized FHIT
protein bands with the antibody used.
Discussion
Since the FHIT gene is usually expressed at very low
levels (Ohta et al., 1996), we used a sensitive yet highly
speci®c nested touchdown RT-PCR technique to screen
both normal and malignant haematopoietic cells for
aberrant FHIT mRNA expression patterns. Southern
blot hybridization of ampli®cation products with a fulllength FHIT cDNA probe permitted us in most of our
samples to trace additional FHIT transcripts which
had escaped detection on inspection of agarose gels.
Since Southern blot hybridization of PCR products has
only rarely been performed in a systematic way, we
conclude that the frequency of aberrant transcripts of
the FHIT gene is underreported in the current
literature (Michael et al., 1997; Panagopoulos et al.,
1997; Wang et al., 1998).
We show that aberrant FHIT mRNA expression is
frequent in di€erent types of haematological neoplasms
as reported for many types of solid tumours. The
majority of aberrant transcripts seen in our series of
leukaemias and lymphomas were previously described
in various carcinomas (Fong et al., 1997; Negrini et al.,
1996; Ohta et al., 1996; Sozzi et al., 1996). This may
point to a common mechanism of disturbed gene
transcription operative in both groups of malignancies.
In addition, we found several novel types of aberrant
transcripts resulting from the insertion of short unique
DNA sequences between di€erent exons. The mechanisms responsible for the generation of aberrant FHIT
transcripts are not yet fully elucidated. Our cDNA
sequencing data would support the notion that
aberrant or alternative splicing is a prime mechanism
of generating aberrant FHIT transcripts in haematological neoplasia, as reported for the candidate tumour
suppressor gene TSG101 in normal lymphocytes, CLL,
melanoma, gynaecological tumours and lung cancer
(Gayther et al., 1997; Lee and Feinberg, 1997).
Interestingly, aberrant transcripts were also described
in non-malignant tissues including human lung, liver,
peripheral blood lymphocytes and other cell types
(Druck et al., 1997; Fong et al., 1997; Gayther et al.,
1997; Panagopoulos et al., 1997; van den Berg et al.,
1997). We now con®rm and extend these ®ndings in
our survey of various types of normal haematopoietic
cells which frequently showed both smaller- and largerthan-wt transcripts. These aberrant transcripts in
normal cells most likely resulted from aberrant
splicing and/or small DNA insertions similar to their
malignant counterparts. Thus aberrant FHIT mRNA
transcripts are not tumor-speci®c and may be observed
both in normal and neoplastic cells.
Aberrant FHIT mRNA processing or lack of FHIT
gene expression in a cell might point to altered FHIT
protein expression (Druck et al., 1997; Greenspan et
al., 1997; Sozzi et al., 1997). In contrast to data from
solid tumours we never detected truncated FHIT
protein in normal or neoplastic haematopoietic cells
with the antibody used in our survey (Yeung et al.,
1997). This observation is corroborated by our detailed
analysis of sequenced aberrant FHIT transcripts.
Disturbances of the open reading frame precluding
translation were identi®ed in about one half of all
aberrant transcripts seen in our samples. In about one
half of aberrant transcript types with a maintained
open reading frame, translation (if performed at all)
would produce abnormally-sized proteins of uncertain
stability. Our sequencing data would thus explain the
absence of truncated proteins in the majority of cells
with aberrant transcripts identi®ed in our series. We
did, however, note interesting di€erences in the
expression of wild-type FHIT protein between normal
and neoplastic haematopoietic cells. In all normal
haematopoietic cell types FHIT protein was readily
detected. However, we now report the novel ®nding
that some haematological neoplasms mostly of myeloid
lineage lacked both wt FHIT mRNA and FHIT
protein. Absence or very low expression of FHIT
protein may clearly distinguish some leukaemic from
normal haematopoietic cells and suppression of FHIT
protein production might contribute to leukaemogenesis.
The question of whether the FHIT gene acts as a
classical tumour-suppressor is still controversial (Le
Beau et al., 1998; Mao, 1998). Some data argue against
this notion: (a) at the genomic level point mutations
possibly inactivating one allele appear to be extremely
rare (Ahmadian et al., 1997; Gemma et al., 1997;
Thiagalingam et al., 1996); (b) aberrant mRNA
transcripts are mostly accompanied by an often
predominant wt mRNA fragment, and abnormal
FHIT mRNA transcripts are also found in various
normal tissues (Gayther et al., 1997; Panagopoulos et
al., 1997; van den Berg et al., 1997; Wang et al., 1998);
(c) aberrant transcripts are not consistently translated
into truncated protein; and (d) replacement of FHIT
protein in human carcinoma cells does not always
suppress tumour cell growth implying that FHIT is
perhaps involved in tumorigenesis along pathways
distinct from the classical tumour suppressor mechanism (Otterson et al., 1998). On the other hand some
data suggest that the wt FHIT gene product might
exert tumour-suppressing activity: (a) in contrast to
normal cells which invariably express normal FHIT
protein, the protein may be absent in solid tumours
(Druck et al., 1997; Greenspan et al., 1997; Sozzi et al.,
1997). We now show that in contrast to normal
haematopoietic cells some haematological cancers lack
FHIT protein; (b) introduction of the FHIT gene into
several FHIT-negative tumour cell lines may result in
their loss of tumorigenicity in nude mice (Siprashvili et
al., 1997); and (c) recent evidence suggests that cancerspeci®c FHIT deletions may originate through breaks
and recombinations within the FRA3B fragile site
(Inoue et al., 1997; Michael et al., 1997; Ong et al.,
1997; Zimonjic et al., 1997). At present, the precise role
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
of FHIT in tumorigenesis is thus not ®rmly clari®ed.
Based on our results and those of others we nevertheless believe that FHIT gene disruptions and
particularly lack of FHIT protein expression may be
relevant to tumorigenesis including the biology of
leukaemia development. Further studies, particularly
FHIT gene knock-out experiments will be of great
interest to advance our understanding of the contribution of this interesting gene to cancer.
Materials and methods
Normal cells, leukaemic cell lines and clinical tumour samples
Normal haematopoietic cells were isolated from peripheral
blood of healthy blood donors. Granulocytes and mononuclear cells were separated by Lymphoprep gradient
centrifugation (Nycomed Pharma AS, Oslo, Norway),
and subsequently monocytes and lymphocytes were
isolated from the mononuclear cell fraction by Percoll
gradients (Pharmacia Biotech AB, Uppsala, Sweden)
(Uguccioni et al., 1996). CD34+ progenitor cells were
obtained from patients with lymphoid neoplasms in
remission who were candidates for high dose chemotherapy consolidation and reinfusion of peripheral blood stem
cells. Cells were mobilized using chemotherapy and
granulocyte-colony-stimulating factor and harvested
through leukapheresis. Apheresis products were labelled
with a biotinylated anti-CD34 monoclonal antibody
(CellPro, Wezembeek-Oppen, Belgium) and tagged cells
were separated with a CEPRATE SC Cell Concentrater
(CellPro, Wezembeek-Oppen, Belgium). The purity of the
CD34+ population obtained was 490% as assessed by
immunocytochemical analysis of cytospin preparations.
The panel of cell lines studied included 10 myeloid (HEL,
HL60, U937, K562, KG1, Kasumi, EM3, THP1, NB4, MV411) and six lymphoid (CEM, Jurkat, Molt4, Raji, Daudi,
Nalm6) leukaemic cell lines obtained from the ATCC
(American Type Culture Collection, Rockville, USA) or the
DSMZ (Deutsche Sammlung von Mikroorganismen und
Zellkulturen, Braunschweig, Germany). The adenovirus 5 Tantigen-transformed human embryonal kidney cell line 293
(kindly provided by N Barron, Manchester, UK) and the
gastric adenocarcinoma cell line Kato III (kindly provided by
SA Hahn, Bochum, Germany) served as positive and negative
controls for FHIT protein analysis, respectively. Cell lines
were cultured either in McCoy's medium (Gibco BRL, Grand
Island, USA) or in RPMI 1640 (Life Technologies, Paisley,
Scotland, UK) supplemented with 10 ± 15% fetal bovine
serum (Nabi, Miami, USA) in a humi®ed atmosphere
containing 5% CO2 at 378C.
We investigated samples from 55 patients with haematological malignancies including 21 acute myeloid leukaemias
(AML), eight chronic myeloid leukaemias (CML), 10 chronic
lymphocytic leukaemias (CLL), as well as seven low-grade
(Lg-NHL) and nine high-grade B-cell Non-Hodgkin's
lymphomas (Hg-NHL) diagnosed according to the REALclassi®cation. Mononuclear cells from untreated leukaemias
containing 490% malignant cells as assessed by morphology
were enriched by Ficoll density gradient centrifugation
(Histopaque 1.077 g/L; Sigma, St Louis, USA). All
lymphoma biopsies were snapfrozen and stored at 7708C
until use.
RNA extraction, RT ± PCR and sequencing
From all samples total cellular RNA was extracted using
either RNAzolB1 (Tel-Test Inc., Friendswood, USA) or
RNeasy1 (Qiagen, Hilden, Germany). Random-primed
®rst-strand cDNA was synthesized from total RNA (0.5 ±
1 mg/20 ml reaction) using the 1st Strand cDNA Synthesis
Kit1 (Boehringer, Mannheim, Germany). The entire coding
region of the FHIT gene was ampli®ed using a sensitive
nested touchdown RT ± PCR. 5 ± 10 ml of cDNA were
subjected to a ®rst PCR round using primers 5U2 and
1038R according to GenBank accession number U46922
(Fong et al., 1997; Ohta et al., 1996). After an initial
denaturation step at 948C for 5 min primer annealing
started at 708C, and subsequently the annealing temperature was decreased by 18C per cycle for the ®rst ®ve cycles
until the de®nitive annealing temperature of 658C used for
the following 25 cycles was reached. The other PCR
conditions remained constant throughout including denaturation at 948C for 15 s, and primer extension at 728C for
45 s. Reactions were carried out in a volume of 30 ml with
10% glycerol under standard conditions using Taq DNA
polymerase (Boehringer, Mannheim, Germany) on a DNA
Thermal Cycler 9600 (Perkin Elmer, Foster City, USA).
After a 20-fold dilution 1 ml of the reaction product was
ampli®ed in a nested PCR reaction with primers 203F and
904R for another 30 cycles to increase the sensitivity (Fong
et al., 1997). Since this highly sensitive nested PCR might
favour the ampli®cation of small aberrant FHIT mRNA
transcripts, selected samples were reanalysed using a single
step touchdown PCR with primers 203F and 904R for 40
cycles under the conditions described above. No di€erences
in the band patterns of ampli®cation products were noted.
RNA integrity was veri®ed by RT ± PCR ampli®cation of
b-actin RNA (Schwaller et al., 1997). All PCR products
were separated on 1.5% agarose gels and stained with
ethidium bromide. Bands of interest visible on agarose gels
were sequenced directly after excision and spin columnpuri®cation (Qiagen, Hilden, Germany). Other bands
including those visible only on autoradiographs after
hybridization were analysed through cloning of 1 ml of
PCR product into a pCR-TOPO vector (Invitrogen, NV
Leek, The Netherlands) and subsequent screening of the
colonies. Sequencing was performed with primers 203F and
904R using the Dye terminator DNA cycle sequencing kit
and sequencer 373A (Applied Biosystems, Foster City,
USA). Sequencing data were analysed with the Wisconsin
Package Version 9.0 (Genetics Computer Group, Madison,
USA).
Southern blot hybridization of PCR products
RT ± PCR products were electrophoresed on 1.5% agarose
gels, Southern blotted and hybridized with a PCRgenerated digoxigenin-labeled cDNA probe containing
exons 1 ± 10 (positions 140 ± 1038) of the FHIT gene. The
signals were detected with chemiluminescent alkaline
phosphatase substrate (Boehringer, Mannheim, Germany)
after exposure to X-ray ®lms. In order to rule out nonspeci®c hybridization of the cDNA probe PCR products
which were unrelated to the FHIT gene were co-blotted on
the ®lters and thus served as negative controls. As a
positive control a sequenced PCR amplicon from the
normal FHIT gene was included.
Analysis of genomic DNA by Southern blotting
Five mg of genomic DNA from malignant and normal cells
was digested with the restriction enzymes BamHl, HindIII
and XbaI (Boehringer, Mannheim, Germany) selected
according to published restriction data (Druck et al.,
1997). The fragments which included the entire gene were
size-fractioned in 0.7% agarose gels and Southern blotted.
The PCR-generated cDNA probe spanning FHIT exons
1 ± 10 was [a32P]dCTP by random priming (Boehringer,
Mannheim, Germany). Filters were hybridized, washed
under stringent conditions and autoradiographed. Equal
DNA loading was veri®ed by cross-hybridization with a
83
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
84
DNA probe recognizing the 5'bcr region on chromosome
22 (Gro€en et al., 1984).
Immunoprecipitation and Western blot analysis
For immunoprecipitation, 10 ± 206106 cells were lysed in
ice-cold NP40 bu€er containing 150 mM NaCl, 50 mM
Tris-HCL (pH 8.0), 0.5% NP40 and complete protease
inhibitors (Boehringer, Mannheim, Germany). The lysates
were then centrifuged to remove debris. 1 ± 2 mg of total
protein was incubated with 2 ml of a polyclonal rabbit antiFHIT IgG antiserum (kindly provided by Dr K Huebner,
Philadelphia, USA) (Barnes et al., 1996) on a rotator at
48C for 2 h. Immune complexes were collected on 35 ml
protein A-Sepharose (Pharmacia Biotech AB, Uppsala,
Sweden) after incubation for 1 h and were washed three
times in 0.5 ml ice-cold NP40 bu€er. The beads were
resuspended in 35 ml of 26sample bu€er. For Western blot
analysis 5 ± 106106 cells were lysed in electrophoresis
sample bu€er (80 mmol/L Tris-HCl [pH 6.8], 2% SDS,
2% 2-Mercaptoethanol, 10% glycerol). Both immunoprecipitated and total protein samples were boiled for 5 min,
subjected to electrophoresis in 12% SDS-polyacrylamide
and transferred to nitrocellulose ®lters. The membranes
were blocked with 2.5% dried milk in TBST for 1 h,
washed three times with TBST and incubated overnight
with the FHIT antibody (1 : 3000 dilution in TBST). After
three ®nal TBST washes the FHIT protein was detected
with a peroxidase-labeled anti-rabbit antibody (1 : 2000
dilution in TBST) using the ECL system (Amersham,
ZuÈrich, Switzerland).
Acknowledgements
We would like to thank Dr Nicole Schmitz and Dr Pius
LoÈ tscher for their help in establishing some of the
techniques used, Dr F Joncourt for her assistance in
sequencing, and Madeleine Oestreicher, Elisabeth Ischi as
well as Karin ZbaÈren for excellent technical assistance.
This work was supported by grants from the Swiss
National Foundation (31-43458.95), the Swiss Cancer
League (KFS 156-9-1995) and the Bernese Foundation
for Clinical Cancer Research.
References
Ahmadian M, Wistuba II, Fong KM, Behrens C, Kodagoda
DR, Saboorian MH, Shay J, Tomlinson GE, Blum J,
Minna JD and Gazdar AF. (1997). Cancer Res., 57, 3664 ±
3668.
Barnes LD, Garrison PN, Siprashvili Z, Guranowski A,
Robinson AK, Ingram SW, Croce CM, Ohta M and
Huebner K. (1996). Biochemistry, 35, 11529 ± 11535.
Brauch H, Johnson B, Hovis J, Yano T, Gazdar A, Pettengill
OS, Graziano S, Sorenson GD, Poiesz BJ, Minna J et al.
(1987). N. Engl. J. Med., 317, 1109 ± 1113.
Carapeti M, Goldman J and Cross N. (1996). Blood, 88
Suppl I, 578a.
Druck T, Hadaczek P, Fu TB, Ohta M, Siprashvili Z, Ba€a
R, Negrini M, Kastury K, Veronese ML, Rosen D,
Rothstein J, McCue P, Cotticelli MG, Inoue H, Croce
CM and Huebner K. (1997). Cancer Res., 57, 504 ± 512.
Fong KM, Biesterveld EJ, Virmani A, Wistuba I, Sekido Y,
Bader SA, Ahmadian M, Ong ST, Rassool FV, Zimmerman PV, Giaccone G, Gazdar AF and Minna JD. (1997).
Cancer Res., 57, 2256 ± 2267.
Gayther SA, Barski P, Batley SJ, Li L, de Foy KA, Cohen
SN, Ponder BA and Caldas C. (1997). Oncogene, 15,
2119 ± 2126.
Gemma A, Hagiwara K, Ke Y, Burke LM, Khan MA,
Nagashima M, Bennett WP and Harris CC. (1997). Cancer
Res., 57, 1435 ± 1437.
Greenspan DL, Connolly DC, Wu R, Lei RY, Vogelstein JT,
Kim YT, Mok JE, Munoz N, Bosch FX, Shah K and Cho
KR. (1997). Cancer Res., 57, 4692 ± 4698.
Gro€en J, Stephenson JR, Heisterkamp N, de Klein A,
Bartram CR and Grosveld G. (1984). Cell, 36, 93 ± 99.
Hayashi S, Tanimoto K, Hajiro-Nakanishi K, Tsuchiya E,
Kurosumi M, Higashi Y, Imai K, Suga K and Nakachi K.
(1997). Cancer Res., 57, 1981 ± 1985.
Inoue H, Ishii H, Alder H, Snyder E, Druck T, Huebner K
and Croce CM. (1997). Proc. Natl. Acad. Sci. USA, 94,
14584 ± 14589.
Kastury K, Ba€a R, Druck T, Ohta M, Cotticelli MG, Inoue
H, Negrini M, Rugge M, Huang D, Croce CM, Palazzo J
and Huebner K. (1996). Cancer Res., 56, 978 ± 983.
Killary AM, Wolf ME, Giambernardi TA and Naylor SL.
(1992). Proc. Natl. Acad. Sci. USA, 89, 10877 ± 10881.
Le Beau MM, Drabkin H, Glover TW, Gemmill R, Rassool
FV, McKeithan TW and Smith DI. (1998). Genes Chrom.
Cancer, 21, 281 ± 289.
LeBeau MM and Rowley JD. (1984). Nature, 308, 607 ± 608.
Lee MP and Feinberg AP. (1997). Cancer Res., 57, 3131 ±
3134.
Lima CD, Klein MG and Hendrickson WA. (1997). Science,
278, 286 ± 290.
Lin PM, Liu TC, Chang JG, Chen TP and Lin SF. (1997). Br.
J. Haematol., 99, 612 ± 617.
Mao L, Fan YH, Lotan R and Hong WK. (1996). Cancer
Res., 56, 5128 ± 5131.
Mao L. (1998). J. Natl. Cancer Inst., 90, 412 ± 414.
Michael D, Beer DG, Wilke CW, Miller DE and Glover TW.
(1997). Oncogene, 15, 1653 ± 1659.
Mitelman F, Mertens F and Johansson B. (1997). Nat.
Genet., 15, 417 ± 474.
Negrini M, Monaco C, Vorechovsky I, Ohta M, Druck T,
Ba€a R, Huebner K and Croce CM. (1996). Cancer Res.,
56, 3173 ± 3179.
Ohta M, Inoue H, Cotticelli MG, Kastury K, Ba€a R,
Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T and
et al. (1996). Cell, 84, 587 ± 597.
Ong ST, Fong KM, Bader SA, Minna JD, Le Beau MM,
McKeithan TW and Rassool FV. (1997). Genes Chrom.
Cancer, 20, 16 ± 23.
Otterson GA, Xiao GH, Geradts J, Jin F, Chen WD,
Niklinska W, Kaye FJ and Yeung RS. (1998). J. Natl.
Cancer Inst., 90, 426 ± 432.
Panagopoulos I, Thelin S, Mertens F, Mitelman F and Aman
P. (1997). Genes Chrom. Cancer, 19, 215 ± 219.
Rassool FV, Le Beau MM, Shen ML, Neilly ME, Espinosa
R, Ong ST, Boldog F, Drabkin H, McCarroll R and
McKeithan TW. (1996). Genomics, 35, 109 ± 117.
Sanchez Y, el-Naggar A, Pathak S and Killary AM. (1994).
Proc. Natl. Acad. Sci. USA, 91, 3383 ± 3387.
Schwaller J, Pabst T, Bickel M, Borisch B, Fey MF and
Tobler A. (1997). Br. J. Haematol., 99, 896 ± 900.
Siprashvili Z, Sozzi G, Barnes LD, McCue P, Robinson AK,
Eryomin V, Sard L, Tagliabue E, Greco A, Fusetti L,
Schwartz G, Pierotti MA, Croce CM and Huebner K.
(1997). Proc. Natl. Acad. Sci. USA, 94, 13771 ± 13776.
Sozzi G, Tornielli S, Tagliabue E, Sard L, Pezzella F,
Pastorino U, Minoletti F, Pilotti S, Ratcli€e C, Veronese
ML, Goldstraw P, Huebner K, Croce CM and Pierotti
MA. (1997). Cancer Res., 57, 5207 ± 5212.
Sozzi G, Veronese ML, Negrini M, Ba€a R, Cotticelli MG,
Inoue H, Tornielli S, Pilotti S, De Gregorio L, Pastorino
U, Pierotti MA, Ohta M, Huebner K and Croce CM.
(1996). Cell, 85, 17 ± 26.
FHIT expression in malignant and normal haematopoiesis
UR Peters et al
Sugimoto K, Yamada K, Miyagawa K, Hirai H and Oshimi
K. (1997). Stem Cells, 15, 223 ± 228.
Thiagalingam S, Lisitsyn NA, Hamaguchi M, Wigler MH,
Willson JK, Markowitz SD, Leach FS, Kinzler KW and
Vogelstein B. (1996). Cancer Res., 56, 2936 ± 2939.
Uguccioni M, Loetscher P, Forssmann U, Dewald B, Li H,
Lima SH, Li Y, Kreider B, Garotta G, Thelen M and
Baggiolini M. (1996). J. Exp. Med., 183, 2379 ± 2384.
van den Berg A, Draaijers TG, Kok K, Timmer T, Van der
Veen AY, Veldhuis PM, de Leij L, Gerhartz CD, Naylor
SL, Smith DI and Buys CH. (1997). Genes Chrom. Cancer,
19, 220 ± 227.
Wang L, Darling J, Zhang JS, Qian CP, Hartmann L,
Conover C, Jenkins R and Smith DI. (1998). Oncogene,
16, 635 ± 642.
Yeung R, Xiao G, Kisielexski A, Klein-Szanto A, Sina J,
Bleicher K, Cheng J, Testa J and Goodrow T. (1997). Proc
AACR, 38, 422.
Yunis JJ and Soreng AL. (1984). Science, 226, 1199 ± 1204.
Zbar B, Brauch H, Talmadge C and Linehan M. (1987).
Nature, 327, 721 ± 724.
Zimonjic DB, Druck T, Ohta M, Kastury K, Croce CM,
Popescu NC and Huebner K. (1997). Cancer Res., 57,
1166 ± 1170.
85