ABSTRACT
Several procedures have been described for fluorescent labeling of DNA and RNA.
They are based on the introduction of aldehyde groups by partial depurination of DNA or oxidation of the 3
'
-terminal ribonucleoside in RNA by sodium periodate. Fluorescent labels
with an attached hydrazine group are efficiently coupled with the aldehyde
groups and the hydrazone bonds are stabilized by reduction with sodium cyanoborohydride. Alternatively, DNA can be quantitatively split at the depurinated sites with
ethylenediamine. The aldimine bond between the aldehyde group in depurinated
DNA or oxidized RNA and ethylenediamine is stabilized by reduction with sodium
cyanoborohydride and the primary amine group introduced at these sites is used for attachment of isothiocyanate or succinimide derivatives of fluorescent dyes. The fluorescent DNA labeling can be carried out
either in solution or on a reverse phase column. These procedures provide
simple, inexpensive methods of multiple DNA labeling and of introducing one
fluorescent dye molecule per RNA, as well as quantitative DNA fragmentation and
ncorporation of one label per fragment. These methods of fluorophore attachment
were shown to be efficient for use in the hybridization of labeled RNA, DNA and
DNA fragments with oligonucleotide microchips.
Radiolabeling and fluorescent labeling have found wide application in DNA and RNA hybridization and sequence analysis. Although less sensitive, fluorescent dyes conjugated with nucleic acids offer some
essential advantages over a radioactive label (
1
). Fluorescent dyes can be detected in real time with high resolution and several can be
monitored in one experiment. Fluorescent dyes also lack radiation hazards and
the consequent problems of waste handling and disposal.
Fluorescent labeling of nucleic acids is usually carried out by enzymatic reactions. Organic fluorophores are chemically introduced into primers or nucleoside triphosphates and are then incorporated either using
PCR amplification or using DNA or RNA polymerases or terminal polynucleotide
transferase (
2
-
4
).
The direct incorporation of fluorophores into nucleic acids by chemical means (
1
,
5
) has not found wide application. Instead, the introduction of active amino or
thiol groups into synthesized oligonucleotides provides acceptors for
subsequent chemical fluorescent labeling (
1
,
6
-
8
).
Sequencing analysis by hybridization to oligonucleotide microchips is being
developed by different groups (
8
-
15
). The method can be applied to sequencing as well as to diagnostics of genetic diseases, gene polymorphism studies, quantitative analysis of gene expression, analysis of microorganisms in a sample, etc. (for review see
16
). Fluorescent labeling allows one to carry out highly sensitive and rapid
analysis of the hybridization of nucleic acids on microchips. A wide range of
possible applications for hybridization with oligonucleotide microchips requires a procedure for fluorescent labeling that should be simple, efficient and inexpensive and
also satisfy the following criteria: (i) it can be applied to both RNA and DNA,
either isolated from cells or synthesized
in vitro
; (ii) it must be compatible with the fragmentation of nucleic acids
(fragmentation is needed to decrease the formation of hairpin structures that interfere with nucleic acid
hybridization to rather short microchip oligonucleotides); (iii) to carry out quantitative hybridization analysis, the amount of
introduced label should not depend on length of the nucleic acid fragments;
(iv) several labels may be incorporated for multiple coloring, based on the introduction of aldehyde groups into DNA by partial depurination and hybridization analysis.
Here we describe several procedures for fluorescent labeling developed with
model oligonucleotides and applied to RNA and DNA. These procedures are
regularly used by our group for fluorescent labeling of DNA and RNA, before or
after their partial fragmentation. The labeled products have been used for
hybridization with oligonucleotide microchips. These procedures are based on
the introduction of aldehyde groups into DNA by partial depurination and into
RNA by oxidation of its 3'-terminal ribonucleoside with sodium periodate, followed by the
direct or indirect attachment of one fluorophore molecule to the aldehyde group
of a fragment. The hybridization of RNA and DNA with oligonucleotide microchips
demonstrated the efficiency of the labeling procedures described.
The oligonucleotides 5'-d(T
6
GT
8
) and 5'-dA
8
Ur were synthesized with a 394 DNA/RNA synthesizer (Applied Biosystems); poly(dA-dT)[middot]poly(dA-dT) was purchased from Sigma.
Dimethyl sulfoxide (DMSO) was distilled
in vacuo
under calcium hydride. Pyridine and triethylamine were consecutively distilled
under toluenesulfonyl chloride and calcium hydride. Acetonitrile was distilled
under phosphorus pentoxide.
Reagents: tetramethylrhodamine (TMR) isothyocyanate, TMR succinimide ether,
fluorescein isothiocyanate and fluorescein succinimide ether (Molecular Probes
Inc.); T4 polynucleotide kinase (Promega); methanesulfonyl chloride,
toluenesulfonyl chloride, hydrazine hydrate,
N
,
N
-dimethylacetamide and sodium cyanoborohydride (Merck). All other reagents were from Aldrich.
Chromatographic sorbents: RP-18 (Merck), DOWEX-50 (Dow). Chromatographic solvent systems for TLC on kieselgel plates (glass backed; Merck) were chloroform/ethanol 9:1 (A); acetonitrile/aqueous ammonium 8:2 (B); acetonitrile/water/chloroform/triethylamine 8:1:0.5:0.1 (C). Electrophoresis was carried out in 20%
acrylamide:bisacrylamide (30:1) gels with 7 M urea in Tris-borate buffer, pH 7.8, at 300 V for 90 min. Autoradiograms were scanned with a 300A Computing Densitometer (Molecular Dynamics). UV spectra were measured with a Shimadzu UV-160A. The extent of fluorescent labeling was calculated from UV
spectra at 260 nm, where the
a
M
(molar absorbency) value for adenosine is 15 * 10
3
; for thymidine and 3-
N
-methyluridine, 10 * 10
3
; for TMR, ~30 * 10
3
; for fluorescein, 60 * 10
3
; at 490 nm the
a
M
for fluorescein is 60 * 10
3
; at 550 nm the
a
M
for TMR is 90 * 10
3
(
17
-
18
).
In formic acid.
DNA (10 [mu]g or more) in 10 [mu]l 80% formic acid was incubated at 20oC for 30 min and was then precipitated at -20oC for 15 min with 15 vol 2% lithium perchlorate in
acetone. The precipitate was centrifuged, washed twice with 100 [mu]l acetone and then with 100 [mu]l diethyl ether and dried.
In HCl.
DNA (1-10 [mu]g) was suspended in 10 [mu]l 0.2 N HCl, incubated at 37oC for 90 min, diluted with 40 [mu]l 650 mM ethylenediamine hydrochloride, pH 7.6, and used for [beta]-elimination without precipitation.
Complete depurination of d(T
6
GT
8
).
Oligonucleotide (10 pmol radioactive or 1 nmol non-radioactive) was dissolved in 10 [mu]l 80% formic acid and incubated at 60oC for 90 min. It was then precipitated at -20oC for 15 min with 15 vol 2% lithium perchlorate in acetone.
Triethylene glycol (1.5 g, 10 mmol) was dissolved in 50 ml dry pyridine and
methanesulfonyl chloride (2.35 ml, 30 mmol) was added. The reaction was carried
out at 0oC for 2 h and was monitored by TLC (solvent system A). The reaction was
stopped by dilution of the mixture with 300 ml saturated sodium bicarbonate solution. The product was extracted using chloroform (three extractions, 50 ml each). The combined organic extracts were dried over
sodium sulfate and evaporated to dryness. The residue (2.8 g) was dissolved in
30 ml
N
,
N
-dimethylacetamide and 4 ml (80 mmol) hydrazine hydrate was added. The
reaction was carried out at room temperature for 3 h and was monitored by TLC
(solvent systems A and B). The reaction mixture was diluted with 300 ml water
and passed through a DOWEX-50 column (200 ml) in the H
+
form. The column was washed with 300 ml 15% aqueous isopropyl alcohol and then
with 300 ml water and the reaction product was eluted with 500 ml 5% aqueous
ammonium hydroxide. The solution was evaporated to dryness, diluted with 5 ml
water and chromatographed on an RP-18 column (2.5 * 30 cm) eluted with water. The solution was evaporated to 50 ml
and, after filtration, chromatographed on a DOWEX-50 column (200 ml) in the NH
4
+
form. The column was washed with 300 ml water and the product eluted with a
linear gradient of 0-1% (0.5 * 0.5 l) aqueous ammonium hydroxide and then evaporated to dryness.
Yield of bis-1,2-(hydrazoethoxy)ethane: 1.2 g (67%). Mass: 179.1 (M+1).
Tetramethylrhodamine isothiocyanate (1 mg, 2.25 [mu]mol) was dissolved in 1 ml dry acetonitrile and a solution of bis-1,2-(hydrazoethoxy)ethane (2 mg, 11.2 [mu]mol) in 20 [mu]l dry acetonitrile and 3 [mu]l triethylamine was added. After incubation at room
temperature for 30 min (monitored by TLC, silvent system C), the reaction mixture was
chromatographed on a TLC kieselgel plate (20 * 20 cm) eluted with solvent system C. The product-containing fraction was picked up from the plate, eluted from the
sorbent with system C and lyophilized. Fluorescent dye was futher dissolved in
20% methanol to 4 mM concentration. Yield: 0.74 mg (53%). The solution of the
chemical was stored in a freezer (-20oC, 3 months) without noticeable loss of activity.
Mass fragments found:
162.9 [NH
2
NH(CH
2
CH
2
O)
2
CH
2
CH
2
NH +1];
386 [M- NH
2
NH(CH
2
CH
2
O)
2
CH
2
CH
2
NHNHC(S)NH +1];
356.7 [NH
2
NH(CH
2
CH
2
O)
2
CH
2
CH
2
NHNHC(S)NHC
6
H
3
(COO
-
) +1];
325.7 [(CH
2
CH
2
O)
2
CH
2
CH
2
NHNHC(S)NHC
6
H
3
(COO
-
) +1].
Completely depurinated [
32
P]d(T
6
GT
8
) (10 pmol) was dissolved in 10 [mu]l 0.05 M sodium acetate buffer, pH 3.75-5.8, or sodium phosphate buffer, pH 6.5-7.0 (or 10 [mu]g of depurinated DNA was dissolved in 10 [mu]l sodium acetate buffer, pH 4.0) and 1 [mu]l 4 mM TMR hydrazine (
9
) (>10-fold molar excess) in 20% methanol was added to the reaction mixture. The solution was incubated for 1 h at 37oC and then 1.5 [mu]l 0.2 M NaCNBH
3
(or NaBH
4
or PyBH
3
) in dry acetonitrile was added and the mixture was incubated at 20oC for 30 min. The mixture was diluted with 100 [mu]l water and extracted five times with 100 [mu]l
n
-butanol saturated with water. The DNA was precipitated with acetone and
dried.
Depurinated DNA (up to 10 [mu]g) or 10 pmol fully depurinated [
32
P]d(T
6
GT
8
) was dissolved in 50 [mu]l 0.5 M ethylenediamine hydrochloride, pH 7.4, and incubated at 37oC for 3 h. Four microliters of freshly prepared 0.1 M NaBH
4
was added at room temperature, followed by incubation at room temperature for 30 min. Then, 4 [mu]l 20% ethylenediamine was added to the DNA solution and the DNA was
precipitated with 1 ml 2% lithium perchlorate in acetone, washed twice with
acetone and air dried. Alternatively, after the 30 min reduction, the DNA was
diluted with 1.5 ml water, purified on a C18 Sep-Pak cartridge (Waters Corp.) eluted with 1 ml 50% methanol and
lyophilized. However, ~50% of the DNA was usually lost on the column.
The oligonucleotide or DNA with attached ethylenediamine was dissolved in 10 [mu]l absolute DMSO and then 1 [mu]l freshly prepared 30 mM fluorescein (or tetramethylrhodamine) isothiocyanate (or succinimide ether) in dry DMSO and 0.2 [mu]l triethylamine were added. The reaction mixture was incubated at
room temperature for 1 h and then 80 [mu]l 0.1 M sodium acetate buffer, pH 4.0, was added. Unreacted fluorescein was
extracted five times with 100 [mu]l water-saturated
n
-butanol. The DNA was precipitated with acetone and dried.
For column labeling, 20 [mu]g poly(dA-dT)[middot]poly(dA-dT) was incubated in 200 [mu]l 80% formic acid at 20oC for 30 min. The reaction mixture was diluted
with 3 ml water, loaded onto the C-18 column (Alltech; 500 mg, high load) and the column washed with 5 ml
water. All loading and washing procedures were performed with a syringe. The column was filled with 200 [mu]l 500 mM ethylenediamine hydrochloride, pH 7.4, and incubated in a water bath at 37oC for 3 h. The column was washed with 2 ml water and then with 0.5 ml of a 20 mM solution of ethylenediamine hydrochloride in 100 mM sodium acetate buffer, pH 4.2. Nine
microliters of 1 M sodium cyanoborohydride in dry acetonitrile was dissolved in
300 [mu]l of a 20 mM solution of ethylenediamine hydrochloride in 100 mM sodium
acetate buffer, pH 4.2, and loaded immediately onto the column. The column was
incubated at room temperature for 0.5 h and washed with 10 ml water. Fresh
saturated solution of fluorescein isothyocyanate (~50-100 [mu]g) in 200 [mu]l sodium carbonate/bicarbonate buffer, pH 9.1, was loaded,
the column was washed with 1.5 ml water to distribute the dye throughout the
whole column and then the column was washed with 1 ml sodium
carbonate/bicarbonate buffer, pH 9.1. The column was incubated at room
temperature for 5 h, washed with 10 ml water until the eluted solution became
colorless, dried, washed with 2 ml acetone and dried again. The poly(dA-dT)[middot]poly(dA-dT) was eluted with 50% acetonitrile. The fourth to the tenth drops were collected and labeled product was
precipitated with 1.3 ml acetone. The poly(dA-dT)[middot]poly(dA-dT) was dissolved in 100 [mu]l water, extracted three times with 100 [mu]l water-saturated
n
-butanol and precipitated with acetone.
Freshly prepared 0.1 M NaIO
4
(1 [mu]l) was added to RNA (up to 20 [mu]g) or dA
8
Ur (10 pmol [
32
P]dA
8
Ur or up to 1 nmol dA
8
Ur) in 5 [mu]l water and the solution incubated at 20oC for 20 min. RNA or oligonucleotides were precipitated with 2% lithium perchlorate in acetone followed by two washings with acetone. Alternatively, the
excess of NaIO
4
was reduced with 1 [mu]l 0.2 M sodium hypophosphite for 20 min at room temperature. Then 8 [mu]l 0.1 M sodium acetate, pH 4.0, and 4 mM TMR hydrazine in 20% methanol
at 1 [mu]l/0.5 nmol oligonucleotide or fragmented RNA were added. Coupling was carried out at 37oC for 1 h, then 1.5 [mu]l 0.2 M NaCNBH
3
in dry acetonitrile were added and reduction was carried at room temperature
for 30 min. The solution was diluted with 100 [mu]l water, free fluorescent label was extracted with
n
-butanol and RNA was precipitated as described above for DNA.
To quantitatively separate fluorescently labeled oligonucleotides from unlabeled ones, the reaction solution was extracted four times with 50 [mu]l phenol saturated with 1 M Tris-HCl, pH 8.5. Fluorescently labeled oligonucleotides were precipitated
from the combined phenol solutions with a 10-fold excess of 2% LiClO
4
in acetone.
NaIO
4
-oxidized RNA or dA
8
Ur (10 pmol [
32
P]dA
8
Ur or up to 1 nmol dA
8
Ur) were dissolved in 10 [mu]l 50 mM sodium acetate buffer, pH 4.0, and 3 [mu]l 20 mM ethylenediamine hydrochloride, pH 7.2, were added. After incubation at 37oC for 1 h, 1.5 [mu]l 200 mM NaCNBH
3
in dry acetonitrile was added at room temperature. Reduction was carried out for 30 min and then RNA (oligonucleotide) was precipitated with 200 [mu]l of a 2% solution of LiClO
4
in acetone and dried. The precipitate was dissolved in 10 [mu]l dry DMSO and then 1 [mu]l of a 30 mM solution of fluorescein isothiocyanate in dry DMSO
contaning 0.2 [mu]l triethylamine was added. The mixture was incubated at room temperature for 60 min and then diluted with 80 [mu]l 100 mM sodium acetate buffer, pH 4.0. The excess fluorescent label along with traces of fluorescently labeled ethylenediamine were extracted five times with water-saturated
n
-butanol. The RNA was precipitated with acetone and dried.
Samples of 0.5 [mu]g single-stranded DNA or 5 [mu]g RNA, each ~200 nt long and complementary to the DQ[alpha] region of the HLA gene, were prepared as described (
15
; Drobishev
et al.
, unpublished data). A microchip with immobilized 10-, 20- and 30mer was manufactured as described (
15
). RNA and DNA, fluorescently labeled as described above, were hybridized with
the microchip in 10 [mu]l 50 mM sodium phosphate buffer, pH 7.4, containing 1 M NaCl, 10 mM EDTA and
50% formamide at 4oC for 18 h. The microchip was washed once with the same hybridization
solution at 4oC and fluorescence imaging of the hybridized microchip was carried out with
a multicolor fluorescence microscope equipped with a CCD camera and the
necessary software (
15
).
Highly reactive aldehyde groups can be easily formed in DNA after its partial
depurination at acidic pH. DNA depurination has been well studied and has been
used for DNA fragmentation (
19
-
21
), to detect DNA regions shielded by proteins or other ligands (
21
), in the Maxam-Gilbert sequencing method (
22
-
24
) and for DNA-protein crosslinking (
25
,
26
).
Figure
1
shows that the active aldehyde groups can be directly used for DNA labeling by
reaction with fluorophores containing aldehyde-specific (e.g. hydrazine) groups. This reaction causes partial DNA fragmentation at depurinated sites. Alternatively, DNA can be quantitatively fragmented at depurinated sites through [beta]-elimination reactions catalyzed by ethylenediamine (
27
). In the presence of reducing agents, ethylenediamine forms a stable secondary
amine bond with the depurinated site. Then activated fluorescent dyes can be
attached to the second primary amine group of the bound ethylenediamine. These
procedures can be applied to any abasic DNA (Fig.
1
).
Dialdehyde groups are easily incorporated into RNA by oxidation of the 3'-terminal ribonucleotide with NaIO
4
(Fig.
1
). Such oxidized RNA has been used for fractionation (
28
,
29
) and for incorporation of a fluorescent label (
30
-
33
). These procedures were partly modified and applied for fluorescent labeling of RNA in the same
way as described for DNA.
Though quantitative analysis of the reaction yields (Table
1
) was carried out only on the model oligonucleotides, a wide range of RNAs, as
well as single-stranded and double-stranded DNAs up to 1500 nt long, were labeled by the described
methods and used for hybridization with oligonucleotide microchips.
A high excess of activated fluorescent label must be added in the condensation
reaction so that we can ignore the traces of ethylenediamine, contaminated RNA and DNA after condensation. We have found that butanol extraction from acid aqueous solutions completely
separates the labeled product from the excess of unreacted fluorophore without
laborious procedures such us dialysis. Figure
5
shows that butanol extraction produces pure fluorescently labeled poly(dA-dT)[middot]poly(dA-dT) with no admixture of other fluorescent products.
After labeling dA
8
Ur with TMR hydrazine and butanol extraction of the unreacted fluorescent dye, the labeled product can be separated from non-labeled oligonucleotide by phenol extraction. This procedure transfers
>90% of the labeled, but no unlabeled, 9mer into the phenol phase because of
the hydrophobic nature of the fluorophore residue. The procedure works only for
rather short oligonucleotides.
Single-stranded DNA was first depurinated and then labeled with TMR hydrazine or
was fragmented with ethylenediamine and labeled with activated fluorescein. RNA
was enzymatically labeled or was oxidized and labeled with TMR hydrazine.
Hybridization of the mixture of differently labeled DNAs and then RNA samples
successively with the microchip (containing 10-, 20- and 30mers that are complementary to the RNA and DNA; Fig.
6
A) is shown in Figure
6
B. As expected, the fragmented DNA (6a) was hybridized more efficiently and
produced stronger signals than the unfragmented DNA. The hybridization signals
were higher with longer immobilized oligonucleotides. The unfragmented DNA (6b)
and RNA (6c) was also preferentially hybridized on the periphery of the gel
pads due to retarded diffusion of long nucleic acids into the gel and the
possible formation of secondary structures in it. RNA labeled enzymatically
(Fig.
6
B, d) or after oxidation with either TMR hydrazine (Fig.
6
B, c) or with activated fluorescein through a ethylenediamine linker (Gushin
et al.
, unpublished data) showed similar hybridization patterns. The hybridization shown in Figure
6
B was carried out at 4oC. However, hybridization was also carried out at higher temperatures (up
to 60oC) for up to several hours without any sign of degradation of the
fluorescence label (Drobishev
et al.
, Gushin
et al
., unpublished data).
Depurination was carried out under conditions when ~1-2% of the bases were excised from DNA. The random removal of 2-4 bases within, for example, 200 base long DNA molecules did not
appear to significantly affect the efficiency of hybridization. Depurination within the hybridization sequence has a low probability when hybridizing the labeled DNA with short oligonucleotides, although it significantly destabilizes the duplexes. Hybridization with longer oligonucleotides will increase the probability of
depurination within the hybridized regions, but the destabilization effect of
the excised base will be less strong.
Different fluorescent labels can be introduced into different samples of nucleic
acids. The simultaneous hybridization of these samples in a mixture and
differentiation of their hybridization pattern with a multicolor fluorescence microscope (shown in Fig.
6
B, a and b) provides the possibility of more accurate quantitative comparison of
the hybridization of these samples.
It appears that the described methods can provide efficient fluorescence
labeling of DNA and RNA for their use in hybridization and for other procedures
as well.
The procedures described for fluorescent labeling of DNA in solution or on a
column and of RNA in solution are efficient, simple, inexpensive and based on commercially available reagents. They can be applied to samples extracted from cells or prepared by enzymatic or
chemical methods. Fluorescent labeling of DNA can be carried out in parallel
with quantitative or partial fragmentation. RNA can be labeled before or after
standard procedures of fragmentation with acids, alkalis or metals and followed by treatment with phosphatase to remove the 3'-terminal phosphate group. One fluorescent dye is introduced per RNA or DNA
molecule after its fragmentation and this allows us to quantitatively analyze
the labeling and hybridization procedures. These methods have been routinely
used by our group at both ANL and EIMB for fluorescent labeling and hybridization of DNA and RNA with oligonucleotide microchips.
We are indebted to Prof. A.Krayevsky, Prof. V.Florentiev, Dr V.Shick and Dr
P.Hemken for their support of this work. This work was supported by grant DE-FG02-93ER61538 of the US Department of Energy, by grants 558 and 562 of
the Russian Human Genome Program and grant 96-04-49858 of the Russian Foundation of Fundamental Research.
*To whom correspondence should be addressed at present addresss. Tel: +1 630 252
3161; Fax: +1 630 252 3387; Email: amir@everest.bim.anl.gov
+
Present address: Center for Mechanistic Biology and Biotechnology, Argonne
National Laboratory, 9700 South Cass Avenue, Argonne, IL 60439, USA
{
1996 Oxford University Press
REFERENCES
Return