ABSTRACT
We have developed a new method, designated restriction landmark cDNA scanning
(RLCS), which displays many cDNA species quantitatively and simultaneously as
two-dimensional gel spots. In this method cDNA species of uniform length were
prepared for each mRNA species using restriction enzymes. After the restriction
enzyme sites were radiolabaled as landmarks, the labeled fragments were
subjected to high resolution two-dimensional gel electrophoresis. In analyses of cDNA samples from adult
mouse liver and brain (cerebral cortex, cerebellum and brain stem) we detected
~500 and >1000 discrete gel spots respectively of various intensities at a time.
The spot patterns of the three brain regions were very similar, although not
identical, but were quite different from the pattern for the liver. RNA blot
hybridization analysis using several cloned spot DNAs as probes showed that
differences in intensity of the spots among RLCS profiles correlated well with
expression levels of the corresponding mRNA species in the brain regions.
Because the spots and their intensities reflect distinct mRNA species and their
expression level respectively, the RLCS is a novel cDNA display system which
provides a great deal of information and should be useful for systematic
documentation of differentially expressed genes.
In molecular biology it is important to identify and analyze genes whose
expression levels and patterns are different in various cell types, tissues,
developmental stages or particular conditions. Differential (
1
) and subtractive (
2
,
3
) hybridization methods have frequently been used to detect and isolate such
differentially expressed genes. Differential hybridization is easy and
reliable, but only effective for mRNAs expressed abundantly in one of two
samples. Subtractive hybridization is effective in the detection and
concentration of rare mRNA species, but is rather empirical and has poor
reproducibility. Recently Liang and Pardee developed a display system called
differential display (
4
). In this method reverse transcription-PCR is performed with a combination of anchored oligo(dT) primers and
arbitrary primers and the PCR products are separated on a sequencing gel to
detect differences in banding patterns due to differential transcription.
Compared with subtractive and differential hybridization methods the
differential display is advantageous with respect to sensitivity, simplicity
and time requirement. However, PCR-mediated amplification with arbitrary primers often gives false positive
bands that cannot detect signals in RNA blotting analysis (
5
,
6
). Furthermore, delicate control of amplification cycles and/or annealing
temperature is necessary to detect differences in gene expression levels among
samples (
7
,
8
).
Hayashizaki
et al
. have recently developed a novel method, designated RLGS (restriction landmark
genomic scanning), for the systematic analysis of genomic DNA (
9
,
10
). The principle of RLGS is based on using restriction enzyme sites as landmarks
and high resolution two-dimensional gel electrophoresis. The method enables simultaneous
visualization and quantitative determination of >1000 distinct genomic loci as
gel spots. Thus, RLGS is a very powerful technique which has been applied to
genetic mapping (
11
,
12
), systematic detection of methylatable loci (
10
,
13
-
18
) and the search for aberrations in cancer DNA (
10
,
19
-
21
). Furthermore, techniques to clone the target DNA fragments from RLGS gel spots
have also been established (
22
,
23
).
Here we report a new method for cDNA analysis in which labeled cDNA species of
uniform length are prepared for each mRNA species and applied to the two-dimensional display system in RLGS. Our method, restriction landmark cDNA
scanning (RLCS), enables quantitative and simultaneous analysis of many cDNA
species.
Anchor primer.
A biotinylated anchor primer BDT02A, 5'-biotin-GACTAGTTCTAGATCGCGAGCGGCCGCCCTTTTTTTTTTTTTTTMA-3' (M = A, G or C), was used in this experiment.
cDNA synthesis.
Total RNA was prepared from tissues of adult mice by the AGPC method (
24
). Poly(A)
+
RNA was purified from total RNA with Oligotex-dT30 (Roche). Several micrograms of poly(A)
+
RNA and 1 [mu]g anchor primer were mixed, heated to 70oC for 10 min and then chilled on ice. Double-stranded cDNA was synthesized with a cDNA synthesis kit
(SuperScripttm Lambda System; Gibco BRL). After second strand synthesis the reaction
mixture was treated with phenol/chloroform, followed by ethanol precipitation.
The precipitate was dissolved in 70 [mu]l TE for treatment with RNase A (20 [mu]g/ml final concentration) at 37oC for 30 min, followed by phenol/chloroform extraction. The aqueous
layer was subjected to spin column treatment (Chroma Spin-100; Clontech) and cDNA in the eluate was ethanol precipitated. Complete
removal of RNA by RNase A treatment and subsequent spin column treatment
drastically decreased background noise in RLCS (data not shown).
Blocking of the RLCS sample.
Blocking was performed in 25 [mu]l buffer A [50 mM Tris-HCl, pH 7.4, 10 mM MgCl
2
, 10 mM dithiothreitol (DTT)] in the presence of 100 [mu]M each ddNTP and 6.5 U Sequenase (USB) at 37oC for 30 min. After reaction the mixture was made up to 50 [mu]l with TE and treated with phenol/chloroform. The aqueous layer was
subjected to a spin column treatment followed by ethanol precipitation.
Restriction enzyme digestion and labeling.
cDNA equivalent to 1 [mu]g starting poly(A)
+
RNA was completely digested with
Bam
HI and
Bgl
II. The reaction mixture was treated with phenol/chloroform and cDNA was ethanol
precipitated. Labeling of the digested cDNA at
the restriction enzyme sites was done in 15 [mu]l buffer A containing 100 [mu]M ddCTP, 100 [mu]M ddTTP, 50 [mu]Ci [[alpha]-
32
P]dGTP (6000 Ci/mmol) and 1.4 U Sequenase at 25oC for 5 min and the reaction was terminated by adding 35 [mu]l 25 mM EDTA. The enzyme concentration and reaction time for Sequenase
for the labeling were important: non-specific labeling of the RLCS sample with Sequenase increased in parallel
with both parameters, whereas specific labeling at the enzyme site with an
adequate concentration of Sequenase reached a maximal level within 5 min (data
not shown). The reaction mixture was treated with phenol/chloroform and
subjected to a spin column treatment.
Purification of the RLCS sample with Dynabeads.
The eluate was incubated with 0.6 mg Dynabeads M-280 streptavidin (Dynal) in 150 [mu]l STE-BSA buffer (1 M NaCl, 10 mM Tris-HCl, pH 8.0, 1 mM EDTA, 0.01% bovine serum albumin) with
gentle rotation at 37oC for 30 min. The Dynabeads were then collected magnetically and washed
once with STE-BSA buffer and twice with
Not
I buffer (50 mM Tris-HCl, pH 7.5, 10 mM MgCl
2
, 1 mM DTT, 100 mM NaCl). Next, the Dynabeads were incubated with an excess of
Not
I in 150 [mu]l
Not
I buffer with gentle rotation at 37oC for 30 min. After magnetically removing the Dynabeads the supernatant was
collected, treated with phenol/chloroform and then subjected to ethanol
precipitation. The precipitate was dissolved in 10 [mu]l TE for two-dimensional gel analysis.
The details of the two-dimensional gel electrophoresis were as described previously (
9
,
10
,
25
). Briefly, the electrophoretic apparatus for the first and second dimension
gels was purchased from Biocraft Inc. (Japan). A vertical disc gel (1% agarose
gel) in Teflon tubing (60 cm in length and 2.4 mm in inner diameter, no. 5265B;
Sanplatec Corp., Osaka, Japan) was used for the first dimension gel. A 10 [mu]l portion of RLCS sample was electrophoresed in the gel until the marker
bromophenol blue migrated ~50 cm. After electrophoresis the gel rod was taken from the Teflon tubing
and was incubated with
Hin
fI buffer twice for 10 min, followed by complete treatment with
Hin
fI for 2 h. The RLCS sample in the agarose gel rod was then subjected to a
vertical second dimension polyacrylamide gel (6%) electrophoresis. The gel rod
was transferred to the top of the second dimension polyacrylamide gel and was
then connected to the polyacrylamide gel with melted agarose. After the second
dimension gel electrophoresis, the polyacrylamide gel was dried and
autoradiographed.
Target spot DNAs with a
Bam
HI site at one end were cloned by the PCR-mediated method with minor modifications (
23
). Briefly, the electroeluted target spot DNAs were ligated with
Bam
HI and
Hin
fI adapters consisting of pre-annealed double-stranded oligonucleotides 5'-CGCCAGGGTTTTCCCAGTCACGACG-3' and 5'-pATCCGTCGTGACTGGGAAAACCCTGGCG-3' for the
Bam
HI site and 5'-CGCCAGGGTTTTCCCAGTCACGACG-3' and 5'-pANTCGTCGTGACTGGGAAAACCCTGGCG-3' for the
Hin
fI site. Because [[alpha]-
32
P]dGTP was incorporated at the
Bam
HI site of target spot DNAs, a G nucleotide was omitted at the 5'-end of the
Bam
HI adapter. The spot DNAs were then purified in spin columns and PCR amplified
(94oC 1 min, 60oC 1.5 min and 72oC 2 min for 20-35 cycles) using a M13 forward HT primer, 5'-CGCCAGGGTTTTCCCAGTCACGACG-3'. The amplified DNA fragments were purified
using acrylamide gel electrophoresis followed by the Wizard PCR Preps DNA Purification System (Promega). The
purified fragments were digested with
Bam
HI, followed by ligation with a
Bam
HI-dT vector, which had both a
Bam
HI end and a 3'-dT protruding end and which was prepared from the pT7 Blue T-vector (Novagen) by
Bam
HI digestion.
Poly(A)
+
RNA derived from each tissue was denatured with formaldehyde and
electrophoresed in a 1% agarose gel. RNA was blotted onto Hybond-N
+
(Amersham) nylon membrane. Hybridization was performed with Quik-Hyb (Stratagene) using the cloned spot DNAs as probes.
To display various cDNA species as discrete spots in a two-dimensional gel, every cDNA for a paticular mRNA species must be of
uniform length. Figure
1
A represents our strategy for preparing RLCS samples (see also Materials and
Methods). We prepared an oligo(dT) anchor primer designated BDT02A, which has a
biotin residue at the 5'-end, multiple restriction enzyme sites, including for
Not
I, a 15mer dT stretch and an additional 2 nt MA at the 3'-end (M = A, G or C). The primer is designed to anchor to roughly
25% of the mRNA species at the upstream end of the poly(A)
+
tail, so as to make the poly(A)
+
tail of the cDNA uniform in length (
26
), and is designed to recover the final products easily. cDNA is synthesized
using this anchor primer by conventional methods and is blocked by ddNTPs to
prevent most non-specific labeling in the following step (Fig.
1
A, a and b). cDNA is then digested with restriction enzyme A, which creates
protruding cohesive 5'-ends (c) and is radiolabeled at the enzyme sites by Sequenase.
Thus, for individual mRNA species the labeled cDNA fragments generated will be
uniform in length downstream of the enzyme sites, whereas the labeled cDNA
fragments upstream of the enzyme sites are of various sizes (d). The 3'-end-radiolabeled cDNA fragments are then recovered using
streptavidin-conjugated magnetic beads (e). cDNA fragments are released from the beads
by
Not
I digestion and used as RLCS samples (f).
In this report we describe a new method for cDNA display, RLCS, which is based
on restriction landmarking and two-dimensional gel electrophoresis. Using cDNA samples prepared from mouse
liver and brain, we could detect more than several hundred spots in an
analysis. Because one mRNA species theoretically corresponds to one spot, the
results show that expression of more than several hundred genes can be
simultaneously detected in an RLCS analysis. Although the analysis in this
experiment was performed using an anchor primer with MA residues at the 3'-end,
Bam
HI and
Bgl
II for restriction enzyme A and
Hin
fI for restriction enzyme B, expression of other genes should also be detectable
using other anchor primers with MG, MC or MT at the 3'-end and/or different sets of restriction enzymes.
We further show that the difference in intensity of the spots among RLCS
profiles correlated with expression level of the corresponding mRNA species
among samples used for the RLCS. Therefore, it is possible to search for
differentially expressed genes among multiple samples by comparing their RLCS
profiles. Once such spots are identified, established methods can be used to
clone the spot DNAs (
22
,
23
). In the present study, using the PCR-mediated cloning method (
23
), we could easily clone spot DNA fragments derived from cDNA species
differentially expressed in the three brain regions (S3-S5 in Fig.
3
).
The sensitivity of RLCS seems to be high enough to detect very rare mRNA
species. It is estimated, according to the sensitivity of RLGS, as follows. In
RLGS, 1 [mu]g genomic DNA, which is equivalent to 3 * 10
5
copies of genomic DNA in mammalian cells, is usually used per analysis. Because
some spots with intensities several-fold weaker than those of the main observed spots are detectable in RLGS
profiles, we estimate that at least 10
5
copies of a labeled DNA fragment can be detected as a spot in the two-dimensional gel system. In RLCS analysis, when 1 [mu]g poly(A)
+
RNA was used, 0.2-0.3 [mu]g cDNA with an estimated mean size of 1000 bp was normally
synthesized in our experiment (data not shown). Therefore, the total number of
cDNA molecules is calculated as [(0.2-0.3 * 10
-6
)/(660 * 1000 bp)] * (6 * 10
23
) = 1.8-2.7 * 10
11
(copies). Since the population of very rare mRNA species is thought to be
0.0001% of total mRNA (
27
), the number of cDNA corresponding to such mRNA species is calculated as 1.8-2.7 * 10
11
* 0.0001% = 1.8-2.7 * 10
5
. This value is comparable with the estimated limiting number of 10
5
copies for detection. Thus, very rare mRNA species may be detectable in RLCS
analysis, although sensitivity for the spots with a first dimension size of
>1000 bp could be lower, since synthesis efficiency of longer cDNA is lower
than that of shorter cDNA. However, it is possible to enhance such sensitivity
by using cDNA samples corresponding to several micrograms of poly(A)
+
RNA for an analysis, since 1-2 [mu]g DNA can be applied to the two-dimensional gel system.
RLCS is a cDNA display system and should be useful for searching for
differentially expressed genes. Of course, RLCS has its disadvantages, the
major one being that it cannot be used when the total RNA source is limited, as
it requires several micrograms of poly(A)
+
RNA. The differential display developed by Liang and Pardee (
4
) is advantageous in this respect as it requires only a very small amount of
RNA. The operation of RLCS is also technically more complicated than
differential display. However, RLCS is advantageous in several other ways.
First of all, it makes more than several hundred spots visually detectable in
an analysis and newly obtained RLCS profiles can be compared with previous ones
to detect additional differences in mRNA expression, since corresponding spots
among RLCS profiles are easily identified. Second, mRNA is simply converted to
cDNA in an RLCS sample without PCR amplification, in which saturation of
amplification sometimes occurs. Therefore, in RLCS the level of gene expression
should more directly reflect the intensities of the corresponding spots than in
PCR-mediated display systems. This enables detection of subtle differences in
the level of gene expression among the samples examined. The results of the RNA
blot hybridizations in the present study confirm this idea. Furthermore, false
positive PCR products that sometimes appear in the differential display are
excluded in RLCS. Third, most of the restriction enzymes which can be used for
spotting cDNA species two-dimensionally in RLCS cleave DNA strictly with high reproducibility. This
reaction is simple and does not need delicate controls, such as number of
amplification cycles or annealing temperature of the arbitrary primers in the
PCR reactions of differential display (
7
,
8
). In this respect Ivanova and Belyavsky sucessfully reported another cDNA
display method based on consecutive restriction enzyme digestion of 3'-end cDNA fragments to produce fingerprints of gene expression (
28
). However, it may be difficult for their method to detect subtle differences in
the level of gene expression, because it uses PCR amplification. Thus, RLCS is a method with higher reproducibility and has many other advantages compared with other cDNA display systems.
We offer RLCS as a novel method to obtain a great deal of information about
expression of mRNA species. It should be useful for systematically searching
for differentially expressed genes among samples.
We are grateful to Drs Y.Hinuma and M.Hatanaka for their encouragement. We also
thank Dr Y.Hayashizaki for his valuable comments and Dr M.A.El-Farrash for reading the manuscript.
REFERENCES
Return