ABSTRACT
In the past, a highly sensitive and efficient method was developed to map DNA
replication origins in human cells, based on quantitative PCR performed on
nascent DNA samples. This method allowed the identification of a replication
origin in the myeloid HL-60 cell line, located on chromosome 19 within an
~
500 bp segment near the lamin B2 gene [Giacca
et al
. (1994)
Proc. Natl. Acad. Sci. USA
, 91, 7119]. The same procedure has now been further simplified and extended to
a variety of other exponentially growing human cells of different histological
derivation (three neural, one connectival and one epithelial), with a nearly
diploid chromosomal content. In all the six cell lines tested, the origin
activity within the lamin B2 gene domain was localized to the same region.
Furthermore, the lamin B2 origin was also found to be active in stimulated, but
not in quiescent, peripheral blood lymphocytes.
The eukaryotic genome is subdivided, from the point of view of replication, into
many tandemly organized units; in animal cells these `replicons' have typically
a size ranging between 5 * 10
4
and 1.5 * 10
5
bp (
1
). Regulation of DNA replication in these organisms is likely to occur by
modulation of the rate of activation of specific sequences present in each
replicon [origins of DNA replication (
ori
)], in analogy with what is observed in the better defined prokaryotic and viral
replicons. The identification of the
ori
s in metazoan cells has proven to be a much harder task than initially predicted
and has been marred by repeated failures of the traditional
ori
-searching methods, based primarily on the ability of a given sequence to
specifically sustain autonomous replication when inserted into a plasmid (
2
-
4
).
During the past few years, new techniques have been devised aimed at mapping the
ori
sites in metazoan chromosomes, based either on the analysis of newly
synthesized DNA or on the detection of replication intermediate structures
(reviewed in
5
-
8
). The application of these techniques, whose potential drawbacks are in several
cases the need for cell synchronization and/or the use of compounds altering
cell physiology to enrich for origin-enriched DNA, nevertheless led to the convincing identification of some
ori
s in higher eukaryotes (
2
,
9
-
16
).
Some controversy is still open on whether the
ori
s of animal cells are well defined, relatively short areas of the genome (few
tens of bp, like those of prokaryotic organisms and of yeast) or represent
instead broad initiation zones with multiple start sites (
17
-
19
). However, as the experimental approaches move from the analysis of 2D-gel based autoradiographic patterns to more defined molecular approaches,
such as the analysis of the abundance of markers in nascent DNA strands, the
evidence for the
ori
s being relatively short and precisely defined regions of the replicons is
definitely prevailing.
In our laboratory, over the past years, we have developed a procedure based on
the exact quantitation of the abundance of DNA markers in nascent DNA molecules
relying on a very precise and sensitive competitive PCR method (
20
). We have thus been able to identify an origin of human DNA replication within
474 bp (
21
), located in the intergenic spacer between the end of the lamin B2 gene and the
start site of another downstream located and actively transcribed gene (Fig.
1
). This replication origin was identified in the human HL-60 cells, a promyelocytic leukemia cell line. By this method, we could
also determine that in non-synchronously growing HL-60 cultures, the abundance of the sequence containing the lamin B2
ori
in the population of short (1000-1500 nt) single-stranded fragments of nascent DNA was approximately one in 30 000.
This estimation is consistent with the notion that the analyzed family of short
nascent strands corresponds to the family of
ori
s; the latter, in fact, are expected to be present one per replicon, and the
number of replicons in human cells has long been indicated to be present in the
order of 10
4
.
A question then arises whether the same
ori
is utilized also by cell lines of different derivation from the same species.
The availability of the competitive PCR mapping technique that we have
developed, allows us to address this question in asynchronous cell populations
in a relatively simple way. In this report, a further simplification of the PCR
mapping procedure is presented, together with the results obtained in five
different human cell lines and in primary human peripheral blood lymphocytes.
HL-60 (myeloid), IMR32 (neuroblastoma) and SKNMC (neuroblastoma) cells were
cultured in RPMI 1640 with 15% fetal calf serum; HeLa (epithelial), IMR90 (lung
fibroblasts) and SKNBE (neuroblastoma) cells were cultured in Dulbecco's
modified Eagle's medium with 12% fetal calf serum. Both media were supplemented
with 2 mM glutamine and 50 [mu]g/ml gentamicin.
Peripheral blood mononuclear cells were obtained by density centrifugation of
200 ml peripheral blood from a normal donor over Ficoll-Hypaque (Sigma, St Louis, MI). DNA from resting lymphocytes was obtained
from 10
8
non-adherent cells after a 30 min incubation in complete RPMI 1640 medium.
Lymphocytes were activated by incubation in culture medium containing 1 [mu]g/ml phytohemagglutinin (PHA, Sigma) and 10% interleukin-2 (IL-2, Cellular Products Inc., San Diego, CA). Five days after
stimulation, DNA was extracted from 10
8
cells.
When appropriate (see below) cells (1 * 10
8
) were pulse-labeled for 10 min at 37oC either with 1 [mu]M (final concentration) [
3
H]deoxycytidine (21.5 Ci/mmol, Amersham, UK) and 100 [mu]M (final concentration) cold 5'-bromodeoxyuridine (BrdUrd; Boehringer Mannheim GmbH, Mannheim,
Germany), as described (
21
) or with 1 [mu]M (final concentration) [
3
H]thymidine (15.1 Ci/mmol, Amersham).
Total DNA was extracted from exponentially growing, asynchronous cell cultures
by standard procedures (
22
), denatured by a 10 min incubation in boiling water, and size-separated (300 [mu]g per gradient) on 35 ml of 5-30% neutral sucrose gradients for 20 h at 26 000 r.p.m. in a
Beckman SW28 rotor at 20oC. In parallel, a reference tube with a double-stranded size marker DNA (containing four different sized DNA
fragments in the range of 500-5000 bp) was also run on an identical gradient. Fractions of 1 ml each
were collected and those from the gradient with the markers were run on 1%
agarose gel at 30 V overnight. Using the separation pattern of the marker DNA
on the gel as a sedimentation velocity reference, the fractions containing
ssDNA of ~1 kb were selected in each case and dialyzed against Tris-EDTA (0.5 M Tris pH 8; 0.01 M EDTA) for at least 8 hours. These
fractions were then used to quantify ssDNA by competitive PCR using different
sets of primers in the lamin B2 region. Only in the case of SKNBE cells, a 10
min incubation of the cells in 100 [mu]M BrdUrd and 1 [mu]M [
3
H]deoxycytidine in complete RPMI 1640 medium was performed prior to DNA
extraction, and the size-selected fractions were purified by immunoaffinity chromatography using
anti-BrdUrd antibodies (
23
) before proceeding to quantitative PCR (this is exactly the same procedure as
reported in
21
; Fig.
2
A). The variations to this standard procedure (i.e. omission of the
immunoaffinity step, omission of the BrdUrd labeling step, labeling with [
3
H]thymidine 1 [mu]M, etc.) are reported in Table
1
. Analysis of all samples in Figure
4
was according to the simplified protocol outlined in Figure
2
B.
The primers used for competitive PCR experiments in the lamin B2 genomic area
(primer sets B13, B48, BN1, SB12, SE10 and SE17) and within the [beta]-globin gene (primer set PCO) have already been described (
21
). These primers amplify segments of 100-300 bp; their position in the lamin B2 genomic area is shown in Figure
1
. For each primer set, competitor DNA segments were constructed carrying the
corresponding genomic sequence with the addition of 20 extra nucleotides in the
middle, to allow gel electrophoretic resolution of the amplification products.
The procedure for competitor construction was the same as described by Diviacco
et al.
(
20
).
The original method for DNA replication origin mapping (
21
), as described in Figure
2
protocol A, consisted of: (i) bromodeoxyuridine (BrdUrd) and [
3
H]deoxycytidine pulse-labeling of the newly synthesized DNA; (ii) DNA extraction and size
fractionation of short nascent DNA strands (~1 kb); (iii) further purification of newly synthesized, BrdUrd-substituted DNA by immunoaffinity chromatography using anti-BrdUrd antibody coupled to Sepharose beads; and (iv)
quantitation of DNA segment abundance in the purified sample by competitive
PCR. By this procedure, nascent DNA (i.e. short DNA stretches emanating from
the origin) is firstly purified from total newly synthesized DNA (deriving from
all regions in the genome, the cells being unsynchronized) according to its
size and then further enriched from the bulk unreplicated genomic DNA by
affinity chromatography. Since the origin mapping procedure by fragment
abundance detection relies in principle on size selection only, we attempted a
simplification of this procedure by omitting the affinity chromatography step.
Accordingly, nascent DNA strands of ~1 kb in size isolated from exponentially growing asynchronous cells of the
neuroblastoma cell line SKNBE were directly submitted to quantitative PCR
analysis (Fig.
2
B) using six sets of primers spanning the region of interest (Fig.
1
) and one control primer set (primers PCO3 and PCO4), which detects an unrelated
region in the [beta]-globin gene. As shown in Table
1
(rows A and B), the omission of the BrdUrd-purification step did not give any appreciable differences neither in the
detectability of DNA segments by competitive PCR nor in the relative abundances
of the markers. On the contrary, omission of this step, besides simplifying the
procedure, also avoids the introduction of a possible bias in fragment
selection on the anti-BrdUrd affinity column and reduces the possibility of DNA fragmentation by
photodamage.
The simplified procedure described in the previous paragraph was applied to
different cell lines. Figure
3
shows the polyacrylamide gels obtained after a competitive PCR experiment
performed with a nascent DNA sample obtained from the human lung fibroblasts
cell line IMR90. A fixed volume of sample DNA was amplified by seven primer
sets (SB12, SE10, BN1, B48, SE17, B13 in the lamin B2 genomic region, and PCO
in the [beta]-globin gene) in the presence of different quantities of the
respective competitors, as indicated at the top of each gel. According to the
principles of competitive PCR (
20
), the ratio between the competitor and genomic template amplification products
(shown at the bottom of each gel) is linearly related to the number of
competitor molecules initially added to the reaction. Since the latter amount
is known, the concentration of each investigated genomic segment in the sample
could be calculated easily from the interpolation of the regression line.
In addition to the results reported above, which were obtained in established
cell lines that replicate at a continuous rate in the absence of inhibitory
treatments, the lamin B2
ori
activity was studied in peripheral blood lymphocytes from a normal individual.
These primary cells are largely quiescent in the absence of stimulation. When
size-selected single-stranded DNA was analyzed from these cells, no
ori
activity could be detected in the lamin B2 region, as expected (Fig.
5
A). However, when the PCR mapping experiment was performed on nascent DNA
samples extracted from the same lymphocytes at 5 days after stimulation
(obtained by the addition of phytohemoagglutinin and interleukin-2), a clear evidence of activation of the lamin B2
ori
could be observed, peaking at the same region as in the cell lines. This
observation, besides providing an important control to the results obtained in
the established cell lines, represents the first observation of origin usage in
primary human cells.
Figure
The use of BrdUrd pulse-labeling of newly replicated DNA and of the anti-BrdUrd antibodies in the process of template DNA preparation for
origin-detection experiments was originally introduced (
23
,
24
) with the aim of obtaining a high signal-to-noise ratio. We show here that the method can be further simplified
and that the use of BrdUrd and immunoaffinity chromatography can be omitted.
This result can be attributed to careful size selection of the nascent DNA
strand pool to be used as well as to the reduction in the background (i.e. pool
contamination by aspecifically broken DNA) resulting from the elimination of
the BrdUrd-substituted areas in the DNA which are prone to breakage. The value of ~3 * 10
4
different
ori
s inferred by our data in human cells agrees with the occurrence of an origin
every ~100 kb, as originally suggested by the fiber autoradiographic studies on
the mammalian DNA replication fork movement (
25
). This new simplified procedure for origin mapping now provides a simple tool
for the study of the process of DNA replication in single copy domains of
untreated, asynchronously growing cells. Its extensive application will be
suitable for the identification of novel origins in mammalian cell DNA and will
assist in the understanding of the functional organization of the genome.
As far as the lamin B2
ori
in the chromosome 19p13.3 location is concerned, the results reported here show
that this origin is used to initiate DNA replication in a variety of human
cells. Of particular interest is the observation that this
ori
is activated also in primary peripheral blood lymphocytes, providing further
support to the physiological significance of the observations obtained in the
established cell lines. The absence of origin activity in unstimulated,
quiescent lymphocytes goes in parallel to our previously reported observation
where also HL-60 cells, when differentiated
in vitro
to a non-proliferating state, gave no evidence of lamin B2
ori
activity (
21
). Once again, these observations reinforce the validity of the developed method
for origin identification.
The universal use of this origin in all the analyzed cells is not surprising.
Given the close relationship between transcriptional activity and initiation of
DNA replication (
5
), possibly related to the accessibility of chromosomal domains to proteins
involved in both processes, the observation that this origin lies in a
constitutively expressed gene domain, coding for a house-keeping protein (
26
) predicts origin usage by all cells. Further support to this notion is the
observation that, in synchronized cells, this
ori
fires within the first minutes of S phase (
27
).
The observation that the same precise and rather narrowly defined region of the
chromosome works as an
ori
in very different cell types gives still more weight to the conception that
ori
s in higher organisms are constituted by
cis
-acting sites of spatially and temporally programmed interactions with
trans
-acting specific protein factors. A study of the nature of the proteins
involved in the definition and activation of the human lamin B2
ori
has been initiated (
28
).
This work was supported by the Human Capital and Mobility Programme of the
European Commission and by the Progetto Finalizzato `Ingegneria Genetica' of
the Consiglio Nazionale delle Ricerche, Italy.

REFERENCES
Return


