Molecular and Cellular Probes (1991) 5, 1 1 7-124
Comparison of solution hybridization efficiencies using alkaline phosphatase-labelled and 32P-labelled oligodeoxynucleotide probes Sheila Podell,* William Maske, Elaine lbariez and Edward Jablonski Molecular Biosystems, Inc ., 10030 Barnes Canyon Rd ., San Diego, CA 92121, USA (Received 25 July 1990, Accepted 5 September 1990)
The hybridization efficiencies of oligonucleotide probes directly labelled with alkaline phosphatase and probes labelled with 32P were compared by quantitating the enzyme activity or radioactivity associated with hybridization targets over time : The targets tested included both synthetic oligonucleotides (53 bases in length) and single-stranded and double-stranded cloned M13 DNA (7350 bases long) . Hybrid molecules were separated from a hhybridized probes using size exclusion FPLC. This system allowed quantitative analysis of the time course and efficiency of hybridization for both probes and targets in complex hybridization media containing protein blocking agents, formamide, and carrier DNA . Similar maximum hybridization efficiencies were attained for probes labelled with either radioactivity or alkaline phosphatase as marker . The reaction rate constant for oligonucleotide hybridization to long M13 targets was 3 . 6 x 10 5 mol - ' s - ' for a probe labelled with alkaline phosphatase, and 5 . 8 x 105 mol - ' s - ' for the same probe labelled with 32 P.
a
KEYWORDS: Alkaline phosphatase-labelled probes, hybridization kinetics, oligodeoxyribonucleotide probes, solution hybridization .
INTRODUCTION Hybridization in solution is a DNA analysis format that can be automated to obtain quantitative data
tides coupled covalently to protein markers such as alkaline phosphatase .' These complex reaction con-
for large numbers of samples . The effective use of
ditions can make it impossible to measure hybridiza-
oligonucleotide probes in this type of format requires
tion directly by monitoring changes in spectrophoto-
accurate information about how much probe, what
metric absorbance. Hydroxyapatite columns are
kinds and amounts of targets, and what lengths of time are necessary to achieve optimal hybridization .
impractical for separating molecules with mixed sin-
Many studies have been published measuring the
lease digestion of unhybridized nucleic acids' can not
reassociation kinetics of a single type of pure DNA in
be used to measure hybridization for long double-
solution,' -' but analytical hybridizations in the laboratory frequently require mixtures of DNA, containing
stranded DNA targets . A method is needed that
probes and targets which differ substantially in both
tion efficiency and kinetics under all of these non-
size and molar concentration . Hybridizations are
ideal conditions .
gle-stranded and double-stranded character . S1 nuc-
would allow quantitative measurements of hybridiza-
performed in heterogeneous solutions, containing
We have developed a flexible new method for
protein blocking agents, carrier DNA, and formamide .
making quantitative measurements of hybridization
Non-radioactive formats may employ oligonucleo-
in solution, suitable for use with a wide variety of
*Author to whom correspondence should be addressed .
0890-8508/91/020117 + 08 $03 .00/0
117
© 1991 Academic Press Limited
118
S . Podell et al .
probes, targets and buffers . This method has been employed to study the hybridization of both long cloned targets and short synthetic targets to oligonucleotide probes directly labelled with either alkaline phosphatase or 32 P .
MATERIALS AND METHODS
cally for unlabelled oligonucleotides and for purified cloned DNA, using conversion factors of 5 0l .tg/0D 2 ,,, for double-stranded M13 DNA, 40µg/OD 260 for single-stranded M13 DNA, and 37µg/0D260 for oligonucleotide DNA . The protein concentration of alkaline phosphatase conjugated probes was determined usir%,&a-commercial Coomassie blue dye assay (Bio Rad Labs ; 9) with pure alkaline phosphatase (Boehringer) as a standard .
DNA probes The sequence of the primary oligonucleotide probe used in this study was dTGT TGA CAG GTG TAG GT*C CTA . This probe was synthesized on an Applied Biosystems Inc . model 390B DNA synthesizer . The asterisk indicates the position of an eleven atom, amine terminated linker arm nucleoside, which was incorporated during synthesis at position 17 as described,,previously .' The probe was covalently coupled taialkaline phosphatase through the linker arm nucleosi,de,' and the resulting conjugate purified by. gel filtration and ion exchange chromatography . The effectiveness of purification was monitored by electrophoresis on native polyacrylamide gels .' A second probe, whose sequence was dTGT CT*C CGC TTC TTC CTt''CC, was used to verify results obtained with the primary probe (data not shown) . The linker arm on the second probe was incorporated at position number 5 . For radioactive studies, unconjugated probes were labelled at their 5' ends with y 32 P-ATP and 5' polynucleotide kinase, 10 or at their 3' ends with terminal deoxynucleotidyl transferase and a 72 P-cordycepin . 6 Unincorporated 32 P-ATP and 32 P-cordycepin were removed using SepPak C-18 cartridges from Millipore .' The specific activities of labelled probes were determined separately for each reaction by measuring the radioactivity precipitable with 5% trichloroacetic acid both before and after SepPak purification .
DNA targets A 53-base oligonucleotide (without linker arm nucleosides) was synthesized, containing the complements of both model probes . This synthetic target, along with its complement, was cloned into the polylinker region of the M13 vector mp19 . The complete M13 clone was 7350 by long . Both doublestranded (replicative form) and single-stranded (packaged phage) DNA were isolated from bacteria infected with this M13 clone, and used as hybridization targets. The original 53-base synthetic oligonucleotide was also used as a hybridization target . DNA concentrations were determined spectrophotometri-
Solution'.hyhcidization DNA targets were denatured by heating to 100 ° C for 10 min in TE (10 mm Tris, pH 7 . 5, 1 mm EDTA), followed by chilling in an ice bath for 2 min . Labelled probe and concentrated hybridization buffer were added to the chilled targets, and the mixtures were incubated at 37 ° C . The final hybridization mixtures contained 6 x SSC (0 . 93 M NaCl, 0 . 09 M sodium citrate, pH 7 . 4), 0 . 1 mg ml - ' bovine serum albumin, and 10% formamide . Some reactions also included human placental DNA (Sigma Chemical Co .), an inhibitor of non-specific hybridization, at 10 µg ml - ' or 100 µg ml^' . The reaction volume was 100 pl for probe excess experiments and 20 lal for target excess experiments . Hybridizations were stopped by freezing on dry ice. Reaction tubes were stored at - 70 ° C for periods up to 1 week, and thawed immediately prior to column chromatography by incubating at 37°C for 10 s . Elution profiles for reaction mixtures chromatographed immediately following hybridization were identical to those that had been stored frozen and thawed .
Column chromatography Hybrid molecules were separated from free probe molecules by gel filtration chromatography on a Pharmacia FPLC system . Since unhybridized target molecules were not labelled with either 32 P or alkaline phosphatase, they were effectively invisible in this system . Reactions with 32 P-labelled probes were analysed on Superose 12 columns (fractionation range 1000-300,000 mw), and reactions employing alkaline phosphatase conjugated probes were analysed on Superose 6 columns (fractionation range 50005,000,000 mw) . Columns were eluted with 5 X SSC (0. 78 M NaCl, 0 . 075 M sodium citrate) at 0 . 5 ml min - ' (Superose 12) or 0 . 35 ml min - ' (Superose 6), and fractions of 0 .5 ml were collected . Radioactivity was measured by Cerenkov counting on a TM Analytic Delta 300 liquid scintillation system . Collection tubes for reactions containing alkaline phosphatase were
Comparison of solution hybridization efficiencies
119
pre-filled with 115µd of 20 mg ml - ' bovine serum albumin, to stabilize enzyme activity . The bovine serum albumin was pretreated with HCI (pH 2 . 0 for 2 h) to remove any endogenous alkaline phosphatase activity, then neutralized with NaOH .
or radioactivity hybridized was determined by FPLC chromatography on Superose gel filtration columns .
Alkaline phosphatase detection
Total recoveries ranged between 75% and 85% of the sample loaded . Typical Superose column elution profiles for target excess reactions are shown in Figs 1 and 2 . The profiles were not affected by the presence or absence of human placental DNA at 10 tg ml - ' or 100 µg ml - ' . No hybridization could be detected to control M13 DNA lacking the appropriate target sequence with either type of labelled probe . Hybridization efficiency, defined as the percentage of enzyme activity or radioactivity which eluted in the hybrid peak, was slightly lower for alkaline phosphatase-labelled probes than for 32P-labelled probes . The average efficiency for alkaline phosphatase-labelled probes was 84% ± 4% over three separate experiments. The corresponding value for 32 P-labelled probes was 92% ± 3% over four experiments .
Fractions were incubated for 30 min at 37 ° C in a mixture containing 30 lam 4-methyl umbelliferyl phosphate, 100 m m diethanolamine pH 9 . 0, 5 mm MgCl2 , and 0 . 05% bovine serum albumin . Alkaline phosphatase catalyzes the formation of 4-methyl umbelliferone, which fluoresces at an excitation maximum of 363 nm with an emission maximum of 447 nm . Fluorescence was detected using a Pandex FCA microplate reader .
RESULTS Hybridization efficiency of probes Oligonucleotide probes labelled with either alkaline phosphatase or 32 P were assayed for hybridizability using single-stranded M13 DNA as a target . For each assay, 10-15 fmol of test probe were incubated with a 10-fold excess of M13 target in a volume of 20 gl for 1 h . The percentage of alkaline phosphatase activity
All fractions collected from the columns were assayed individually for alkaline phosphatase activity or radioactivity . Recoveries and peak heights were determined by subtraction of background and simple addition of the activity in the appropriate fractions .
Hybridization efficiency of targets In experiments designed to measure the efficiency of target hybridization, target rather than probe was the
2400
Free probe peak 2000
19
OF. 1600
Hybrid peak
d 1200
800
400
0
10 20 30 40 50 Fraction number (0 . 5 ml fractions)
60
Fig. 1 . Superose 6 elution profiles, showing hybridization between an alkaline phosphatase-labelled oligonucleotide probe and a 10-fold molar excess of either single-stranded wild type M13 DNA (---0---), or single-stranded M13 target containing the complement of the probe (- •- ) .
120
S. Podell et a! . 20 000
16 000
12 000 E c U 8000
4000
10
20
30
40
50
60
Fraction number (0 5 mL fractions) Fig. 2 . Superose 12 elution profiles, showing hybridization between a '2 P-labelled oligonucleotide probe and a 10-fold molar excess of either single-stranded wild type M13 DNA (- -or single-stranded M13 target containing the complement of the probe (- •- ) .
limiting component in the reaction . To establish an optimal probe concentration for these studies, the target concentration was held constant at 250 pm, and increasing amounts of 5' end-labelled 32P probe were added . Hybridization took place in a volume of 100 µl at 37 ° C for 1 h . The amount of probe hybridized at the end of this period was measured using FPLC chromatography as described above . Raw cpm data were converted to fmols of probe hybridized based on the decay adjusted specific activity of the probe . Efficiency was calculated by dividing the moles of probe recovered in the hybrid peak by the moles of target present in the reaction . Variations in probe concentration from 1 ri to 8 nm did not affect the quantity of probe hybridized to target . In subsequent experiments, 2 nm was chosen as a standard probe concentration . Target hybridization efficiency in the presence of 2 nM probe was measured over a target concentration range of 50-500 pm, with several different types of target. No significant differences were observed between single-stranded targets 53 and 7350 nucleotides long (data not shown) . The hybridization efficiency of double-stranded M13 DNA targets did decrease dramatically at concentrations above 100 pm, probably because of target self-annealing (Fig . 3) . Decreased hybridization efficiency at higher target concentration was also observed with other plasmid DNA targets .
Kinetics of hybridization It was essential to obtain accurate values for the concentrations of both targets and probes to study hybridization kinetics . Determining the exact DNA concentration of oligonucleotides covalently labelled with alkaline phosphatase was difficult, because the protein part of the conjugate interfered with spectrophotometric measurements . The oligonucleotides in the conjugates were too short to be detected by Hoechst dye assay"" and too dilute to be detected by diphenylamine assay ." Although the protein concentration of the probe-conjugate stock was readily determined, we could not rule out the possibility that purification might have failed to remove a small percentage of the unconjugated enzyme, or that some DNA molecules might be positioned in a sterically inaccessible orientation . Electrophoretic analysis of the probe conjugates did indicate that a small amount of free enzyme was present, but this amount could not be quantitated . To avoid these potential sources of inaccuracy, the exact concentration of available DNA in alkaline phosphatase-labelled probes was analysed indirectly by titration with limiting amounts of single-stranded M13 target . The probe DNA concentration was determined by calculating the percentage of total enzyme activity that could hybridize to 10 fmol of target . Reaction times were extended to 2 h and the assay
Comparison of solution hybridization efficiencies
121
70
60 Y O m n v 50 .0 T L C 40 V d
N 0 u m
30
d .0 CL 0 0
20
E . U 10
0
100
200 300 400 Target concentration (pm)
500
Fig. 3. Comparison of hybridization efficiencies using a 32 P-labelled oligonucleotide probe with single-stranded (-0-) and double-stranded (-∎-) versions of the same M13 target DNA.
was performed at probe concentrations of 2 nM, 4 nM, and 8 nM to ensure the presence of sufficient DNA to drive hybridization to completion within the alloted time . Probe concentration determinations for alkaline phosphatase-conjugated probes were verified by a second indirect measurement. A short oligonucleotide complementary to the probe-conjugate was 32P labelled at its 3' end with a32 P-cordycepin and terminal transferase . The 3' end label was necessary because y 32 P-label at the 5' end of DNA molecules was cleaved by alkaline phosphatase during hybridization to enzyme-conjugated probes . An aliquot of alkaline phosphatase-labelled probe, estimated by protein assay to be 10 fmols, was hybridized to 200 fmol of its a 32 P-labelled complement in a volume of 100 µl for 1h . The percentage of radioactivity recovered in the resulting hybrids was used to calculate the amount of available DNA actually present in the original alkaline phosphatase-labelled probe . DNA concentration determinations on enzymelabelled probes made by measuring hybridization to excess short oligonucleotides agreed well with those made by measuring hybridization to limiting amounts of long single-stranded M13 target . For the alkaline phosphatase-labelled probe used in this study, the ratio of available DNA/protein was found to be 0 . 82 using single-stranded M13 target titration and 0 .81 using oligonucleotide titration . These ratios correlate well with the probe hybridization efficiency of
84% ± 4% obtained under conditions of target excess, indicating that the fraction of unhybridized alkalinephosphatase activity was consistent between three different experimental formats . The relative contributions of free enzyme contamination and steric hindrance to incomplete hybridization could not be determined, because the exact amount of free alkaline phosphatase could not be quantitated . Time course experiments were performed at 37 ° C under standard buffer conditions (6 x SSC, 0 . 1 mg ml - ' bovine serum albumin, and 10% formamide), with 100 pm target. The most useful data was obtained with probe concentrations ranging between 1 nM and 4 nM . At higher probe concentrations, resolution of the hybrid peak on Superose columns was impaired, and at lower concentrations the sensitivity of target detection was inadequate at early time points . Control experiments showed that the length of single stranded target did not affect the time course of hybridization (Fig. 4) . The rate of hybridization was also not affected by whether the target DNA was single-stranded or double-stranded (Figs 5 and 6) . To compare the hybridization rates of alkaline phosphatase-labelled and 32 P-labelled probes, a rate constant was calculated for each hybridization reaction . The formation of hybrids (PT) from probe (P) and target (T) should follow the general formula for determining the second-order rate constant (k e) of a bimolecular reaction : 14
S . Podell et al.
1 22
100
80 v a)
a LT drn 0
60
0 40
a0 20
0
15
30
45
60
Time (minutes)
Fig. 4 . Time course of hybridization for single-stranded M13 target (7350 nucleotides ; -0-) and synthetic oligonucleotide target (53 nucleotides ; -A-) . 32 P-labelled probe was present at 2-9 nM, and targets were 100 pm each .
Fig. 5 . Time course of hybridization for single-stranded (-0-) and double-stranded ( ∎ ) M13 targets with an alkaline phosphatase-labelled oligonucleotide probe . The concentration of available DNA in the alkaline phosphatase-conjugated probe was 1 . 3 nM, and each of the targets was present at 100 pm .
Comparison of solution hybridization efficiencies
123
Fig. 6. Time course of hybridization for single-stranded (-0-) and double-stranded (-∎-) M13 targers with a 33P-labelled oligonucleotide probe . 32 P-labelled probe was present at 1 .7 nM, and targets were each 100 pm .
[P] + [Tl --* [PT] d[PT1= kb[P] IT] dt The actual probe concentrations (1 .3-3 . 5 nM) used for time course experiments were in 13-35-fold excess of target concentrations, therefore probe concentrations were assumed to remain nearly constant during the reaction . Based on this assumption, a unimolecular rate constant was defined for the hybridization reaction : .
cular) rate constant was computed by dividing the unimolecular constant by the probe concentration . The average rate constant value determined from three experiments like those illustrated in Fig . 5 was 3.6 x 105 mol - ' s - ' ± 0 .9 X 10 5 mol - ' s - ' for the alkaline phosphatase-labelled probe. The average value for three experiments like those illustrated in Fig . 6, with a 32 P-labelled version of the same probe, was 5.8 x 105 mol - ' s - ' f 0. 8 x 105 mol - ' s - ' . These data suggest that alkaline phosphatase-labelled probes hybridize at a rate that is slightly slower, but very close to that of their 32P-labelled counterparts.
k u = kb[Pl DISCUSSION If k~ is substituted into the reaction rate equation, the equation becomes the following : d[PT] = k [T] dt When this equation is solved, then for time (t), PT(t) = [PT,1(1- a -ku`) Experimental data for the time course of hybridization with labelled oligonucleotides probes was fitted to an exponential curve with the above formula by nonlinear least squares analysis . The second order (bimole-
The efficient use of oligonucleotides in solution hybridization for quantitative DNA analysis requires the optimization of a number of time and concentration parameters . The chromatographic method of measuring hybridization efficiency presented here has provided quantitative data for these parameters, under conditions where neither spectrophotometric assays nor hydroxyapatite columns would be useful . The single-stranded or double-stranded nature of long targets strongly influenced the efficiency of solution hybridization with oligonucleotide probes, especially at target concentrations exceeding 100 pm . To avoid strand re-annealing problems, double-
S. Podell et al.
124
stranded target concentrations had to be kept at or
REFERENCES
below 100 pm . The concentration of single-stranded DNA targets was not nearly as critical, up to at least 500 pm. Hybridization rate constants were calculated for the purpose of comparing the efficiency of probes labelled with alkaline phosphatase to those labelled with 32 P . The rate constants found for alkaline phosphatase-labelled probes were slightly slower than those for 32 P-labelled probes, but the difference between the two values was relatively small . The conditions used to determine second order reaction rates by gel filtration chromatrography were chosen to give an optimal signal to noise ratio for the particular probes being tested . For this reason, the rates obtained could not be compared directly with those measured for the renaturation of DNA under more ideal conditions . The temperature chosen (35°C below the T m of the probe, calculated by the method of McGraw et
al .)" and the ionic strength (1 . 0 m Na + )
were sufficient to allow hybridization at the maximum possible rate, but the presence of 10% formamide and the contribution of additional viscosity by bovine serum albumin would be expected to slow the reaction rate . The range of probe concentrations used was fairly narrow, due to technical constraints . Signal detection was difficult at probe concentrations below 1 nM, and resolution of hybrid and free probe peaks was impaired at probe concentrations above 4 nnn . Nevertheless, second order rate constants measured by gel filtration chromatography were consistent with those measured spectrophotometrically for the renaturation of a 12-base oligonucleotide sequence, reported to be 10 5 mol -1 s -1, 16 and with those measured by psoralen crosslinking for the hybridization of a 25 base oligonucleotide to M13 DNA, reported to be 1 . 5 x 106 mol -1 s -1 .17 The maximum hybridization efficiencies attainable for probes labelled with either alkaline phosphatase or 32 P were quite similar. The 5-10% difference in hybridization efficiency observed between the two types of marker was probably due to variation in the effectiveness of the purification methods used to separate labelled probe from free 32 P or unconjugated alkaline phosphatase . Chromatographic separation of hybridization mixtures by gel filtration FPLC has thus allowed quantitative comparison of the maximum hybridization efficiencies attainable by these different types of labelled probes, thus allowing for comparisons which could not have been achieved by other methods currently available .
1 . Wetmur, J . G . & Davidson, N . (1968) . Kinetics of renaturation of DNA . journal of Molecular Biology 32, 349-70. 2 . Casey, J . & Davidson, N . (1977) . Rates of formation and thermal stabilities of RNA : DNA and DNA : DNA duplexes at high concentrations of formamide . Nucleic Acids Research 4, 1539-52 . 3 . Britten, R . J . & Davidson, E . H . (1985) . In Nucleic Acid Hybridization, A Practical Approach . (Hames, B . D . & Higgins, S . J ., eds) pp . 3-15 . Washington D .C . : IRL Press . 4 . Jablonski, E ., Moomaw, E . W., Tullis, R . H . & Ruth, ) . L . (1986) . Preparation of oligodeoxynucleotide-alkaline phosphatase conjugates and their use as hybridization probes . Nuecleic Acids Research 14, 6115-29 . 5 . Durnham, D . M . & Palmiter, R . D . (1983) . A practical approach for quantitating specific mRNAs by solution hybridization . Analytical Biochemistry 131, 385-93 . 6. Tu, Chen-Pei D . & Cohen, S . N . (1980) . 3' end labeling of DNA with [all PJ-cordycepin-5'-tri phosphate . Gene 10, 177-83 . 7 . Ruth, J . L . (1984) . Chemical synthesis of non-radioactively-labeled DNA hybridization probes . DNA 3, 123 . 8 . Selby, M . J ., Barta, A ., Baxter, J . D ., Bell, G . I . & Eberhard, N . L . (1984) . Analysis of a major human chorionic somatomammotropin gene . journal of Biological Chemistry 259, 13131-8 . 9 . Bradford, M . M . (1972) . A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding . Analytical Biochemistry 72, 248-54 . 10 . Maniatis, T ., Fritsch, E. F . & Sambrook, J . (1982) . In Molecular Cloning : A Laboratory Manual . p . 122 . New York : Cold Spring Habor Laboratory . 11 . Labarca, C . & Paigen, K . (1980) . A simple, rapid, and sensitive DNA assay procedure . Analytical Biochemistry 102, 344-52 . 12 . Cesarone, C . F ., Bolognesi, C . & Santi, L . (1979) . Improved microfluorometric DNA determination in biological material using 33258 Hoechst . Analytical Biochemistry 100, 188-97 . 13 . Burton, K . (1968). Determination of DNA concentration with diphenylamine . Methods Enzymol XII, 163-6 . 14 . Young, B . D . & Anderson, M . L. M. (1985) . In Nucleic Acid Hybridization, A Practical Approach . (Names, B . D . & Higgins, S . J ., eds) pp . 47-71 . Washington D .C . : IRL Press . 15 . McGraw, R . A ., Steffe, E . K . & Baxter, S . A . (1990) . Sequence-dependent oligonucleotide-target duplex stabilities : rules from empirical studies with a set of twenty-mers . Bio Techniques 8, 674-8 . 16 . Chu, Y . C . & Tinoco, I . (1983) . Temperature-jump kinetics of the dC-G-T-G-A-A-T-T-C-G-C-G double helix containing a G-T base pair and the dC-G-C-A-G-A-A-TT-C-G-C-G double helix containing an extra adenine . Biopolymers 22, 1235-46. 17. Gamper, H . B ., Cimino, G . B . & Hearst, J. E . (1987) . Solution hybridization of crosslinkable DNA oligonucleotides to bacteriophage M13 DNA . journal of Molecular Biology 197, 349-62 .
Comparison of solution hybridization efficiencies using alkaline phosphatase-labelled and 32P-labelled oligodeoxynucleotide probes Sheila Podell,* William Maske, Elaine lbariez and Edward Jablonski Molecular Biosystems, Inc ., 10030 Barnes Canyon Rd ., San Diego, CA 92121, USA (Received 25 July 1990, Accepted 5 September 1990)
The hybridization efficiencies of oligonucleotide probes directly labelled with alkaline phosphatase and probes labelled with 32P were compared by quantitating the enzyme activity or radioactivity associated with hybridization targets over time : The targets tested included both synthetic oligonucleotides (53 bases in length) and single-stranded and double-stranded cloned M13 DNA (7350 bases long) . Hybrid molecules were separated from a hhybridized probes using size exclusion FPLC. This system allowed quantitative analysis of the time course and efficiency of hybridization for both probes and targets in complex hybridization media containing protein blocking agents, formamide, and carrier DNA . Similar maximum hybridization efficiencies were attained for probes labelled with either radioactivity or alkaline phosphatase as marker . The reaction rate constant for oligonucleotide hybridization to long M13 targets was 3 . 6 x 10 5 mol - ' s - ' for a probe labelled with alkaline phosphatase, and 5 . 8 x 105 mol - ' s - ' for the same probe labelled with 32 P.
a
KEYWORDS: Alkaline phosphatase-labelled probes, hybridization kinetics, oligodeoxyribonucleotide probes, solution hybridization .
INTRODUCTION Hybridization in solution is a DNA analysis format that can be automated to obtain quantitative data
tides coupled covalently to protein markers such as alkaline phosphatase .' These complex reaction con-
for large numbers of samples . The effective use of
ditions can make it impossible to measure hybridiza-
oligonucleotide probes in this type of format requires
tion directly by monitoring changes in spectrophoto-
accurate information about how much probe, what
metric absorbance. Hydroxyapatite columns are
kinds and amounts of targets, and what lengths of time are necessary to achieve optimal hybridization .
impractical for separating molecules with mixed sin-
Many studies have been published measuring the
lease digestion of unhybridized nucleic acids' can not
reassociation kinetics of a single type of pure DNA in
be used to measure hybridization for long double-
solution,' -' but analytical hybridizations in the laboratory frequently require mixtures of DNA, containing
stranded DNA targets . A method is needed that
probes and targets which differ substantially in both
tion efficiency and kinetics under all of these non-
size and molar concentration . Hybridizations are
ideal conditions .
gle-stranded and double-stranded character . S1 nuc-
would allow quantitative measurements of hybridiza-
performed in heterogeneous solutions, containing
We have developed a flexible new method for
protein blocking agents, carrier DNA, and formamide .
making quantitative measurements of hybridization
Non-radioactive formats may employ oligonucleo-
in solution, suitable for use with a wide variety of
*Author to whom correspondence should be addressed .
0890-8508/91/020117 + 08 $03 .00/0
117
© 1991 Academic Press Limited
118
S . Podell et al .
probes, targets and buffers . This method has been employed to study the hybridization of both long cloned targets and short synthetic targets to oligonucleotide probes directly labelled with either alkaline phosphatase or 32 P .
MATERIALS AND METHODS
cally for unlabelled oligonucleotides and for purified cloned DNA, using conversion factors of 5 0l .tg/0D 2 ,,, for double-stranded M13 DNA, 40µg/OD 260 for single-stranded M13 DNA, and 37µg/0D260 for oligonucleotide DNA . The protein concentration of alkaline phosphatase conjugated probes was determined usir%,&a-commercial Coomassie blue dye assay (Bio Rad Labs ; 9) with pure alkaline phosphatase (Boehringer) as a standard .
DNA probes The sequence of the primary oligonucleotide probe used in this study was dTGT TGA CAG GTG TAG GT*C CTA . This probe was synthesized on an Applied Biosystems Inc . model 390B DNA synthesizer . The asterisk indicates the position of an eleven atom, amine terminated linker arm nucleoside, which was incorporated during synthesis at position 17 as described,,previously .' The probe was covalently coupled taialkaline phosphatase through the linker arm nucleosi,de,' and the resulting conjugate purified by. gel filtration and ion exchange chromatography . The effectiveness of purification was monitored by electrophoresis on native polyacrylamide gels .' A second probe, whose sequence was dTGT CT*C CGC TTC TTC CTt''CC, was used to verify results obtained with the primary probe (data not shown) . The linker arm on the second probe was incorporated at position number 5 . For radioactive studies, unconjugated probes were labelled at their 5' ends with y 32 P-ATP and 5' polynucleotide kinase, 10 or at their 3' ends with terminal deoxynucleotidyl transferase and a 72 P-cordycepin . 6 Unincorporated 32 P-ATP and 32 P-cordycepin were removed using SepPak C-18 cartridges from Millipore .' The specific activities of labelled probes were determined separately for each reaction by measuring the radioactivity precipitable with 5% trichloroacetic acid both before and after SepPak purification .
DNA targets A 53-base oligonucleotide (without linker arm nucleosides) was synthesized, containing the complements of both model probes . This synthetic target, along with its complement, was cloned into the polylinker region of the M13 vector mp19 . The complete M13 clone was 7350 by long . Both doublestranded (replicative form) and single-stranded (packaged phage) DNA were isolated from bacteria infected with this M13 clone, and used as hybridization targets. The original 53-base synthetic oligonucleotide was also used as a hybridization target . DNA concentrations were determined spectrophotometri-
Solution'.hyhcidization DNA targets were denatured by heating to 100 ° C for 10 min in TE (10 mm Tris, pH 7 . 5, 1 mm EDTA), followed by chilling in an ice bath for 2 min . Labelled probe and concentrated hybridization buffer were added to the chilled targets, and the mixtures were incubated at 37 ° C . The final hybridization mixtures contained 6 x SSC (0 . 93 M NaCl, 0 . 09 M sodium citrate, pH 7 . 4), 0 . 1 mg ml - ' bovine serum albumin, and 10% formamide . Some reactions also included human placental DNA (Sigma Chemical Co .), an inhibitor of non-specific hybridization, at 10 µg ml - ' or 100 µg ml^' . The reaction volume was 100 pl for probe excess experiments and 20 lal for target excess experiments . Hybridizations were stopped by freezing on dry ice. Reaction tubes were stored at - 70 ° C for periods up to 1 week, and thawed immediately prior to column chromatography by incubating at 37°C for 10 s . Elution profiles for reaction mixtures chromatographed immediately following hybridization were identical to those that had been stored frozen and thawed .
Column chromatography Hybrid molecules were separated from free probe molecules by gel filtration chromatography on a Pharmacia FPLC system . Since unhybridized target molecules were not labelled with either 32 P or alkaline phosphatase, they were effectively invisible in this system . Reactions with 32 P-labelled probes were analysed on Superose 12 columns (fractionation range 1000-300,000 mw), and reactions employing alkaline phosphatase conjugated probes were analysed on Superose 6 columns (fractionation range 50005,000,000 mw) . Columns were eluted with 5 X SSC (0. 78 M NaCl, 0 . 075 M sodium citrate) at 0 . 5 ml min - ' (Superose 12) or 0 . 35 ml min - ' (Superose 6), and fractions of 0 .5 ml were collected . Radioactivity was measured by Cerenkov counting on a TM Analytic Delta 300 liquid scintillation system . Collection tubes for reactions containing alkaline phosphatase were
Comparison of solution hybridization efficiencies
119
pre-filled with 115µd of 20 mg ml - ' bovine serum albumin, to stabilize enzyme activity . The bovine serum albumin was pretreated with HCI (pH 2 . 0 for 2 h) to remove any endogenous alkaline phosphatase activity, then neutralized with NaOH .
or radioactivity hybridized was determined by FPLC chromatography on Superose gel filtration columns .
Alkaline phosphatase detection
Total recoveries ranged between 75% and 85% of the sample loaded . Typical Superose column elution profiles for target excess reactions are shown in Figs 1 and 2 . The profiles were not affected by the presence or absence of human placental DNA at 10 tg ml - ' or 100 µg ml - ' . No hybridization could be detected to control M13 DNA lacking the appropriate target sequence with either type of labelled probe . Hybridization efficiency, defined as the percentage of enzyme activity or radioactivity which eluted in the hybrid peak, was slightly lower for alkaline phosphatase-labelled probes than for 32P-labelled probes . The average efficiency for alkaline phosphatase-labelled probes was 84% ± 4% over three separate experiments. The corresponding value for 32 P-labelled probes was 92% ± 3% over four experiments .
Fractions were incubated for 30 min at 37 ° C in a mixture containing 30 lam 4-methyl umbelliferyl phosphate, 100 m m diethanolamine pH 9 . 0, 5 mm MgCl2 , and 0 . 05% bovine serum albumin . Alkaline phosphatase catalyzes the formation of 4-methyl umbelliferone, which fluoresces at an excitation maximum of 363 nm with an emission maximum of 447 nm . Fluorescence was detected using a Pandex FCA microplate reader .
RESULTS Hybridization efficiency of probes Oligonucleotide probes labelled with either alkaline phosphatase or 32 P were assayed for hybridizability using single-stranded M13 DNA as a target . For each assay, 10-15 fmol of test probe were incubated with a 10-fold excess of M13 target in a volume of 20 gl for 1 h . The percentage of alkaline phosphatase activity
All fractions collected from the columns were assayed individually for alkaline phosphatase activity or radioactivity . Recoveries and peak heights were determined by subtraction of background and simple addition of the activity in the appropriate fractions .
Hybridization efficiency of targets In experiments designed to measure the efficiency of target hybridization, target rather than probe was the
2400
Free probe peak 2000
19
OF. 1600
Hybrid peak
d 1200
800
400
0
10 20 30 40 50 Fraction number (0 . 5 ml fractions)
60
Fig. 1 . Superose 6 elution profiles, showing hybridization between an alkaline phosphatase-labelled oligonucleotide probe and a 10-fold molar excess of either single-stranded wild type M13 DNA (---0---), or single-stranded M13 target containing the complement of the probe (- •- ) .
120
S. Podell et a! . 20 000
16 000
12 000 E c U 8000
4000
10
20
30
40
50
60
Fraction number (0 5 mL fractions) Fig. 2 . Superose 12 elution profiles, showing hybridization between a '2 P-labelled oligonucleotide probe and a 10-fold molar excess of either single-stranded wild type M13 DNA (- -or single-stranded M13 target containing the complement of the probe (- •- ) .
limiting component in the reaction . To establish an optimal probe concentration for these studies, the target concentration was held constant at 250 pm, and increasing amounts of 5' end-labelled 32P probe were added . Hybridization took place in a volume of 100 µl at 37 ° C for 1 h . The amount of probe hybridized at the end of this period was measured using FPLC chromatography as described above . Raw cpm data were converted to fmols of probe hybridized based on the decay adjusted specific activity of the probe . Efficiency was calculated by dividing the moles of probe recovered in the hybrid peak by the moles of target present in the reaction . Variations in probe concentration from 1 ri to 8 nm did not affect the quantity of probe hybridized to target . In subsequent experiments, 2 nm was chosen as a standard probe concentration . Target hybridization efficiency in the presence of 2 nM probe was measured over a target concentration range of 50-500 pm, with several different types of target. No significant differences were observed between single-stranded targets 53 and 7350 nucleotides long (data not shown) . The hybridization efficiency of double-stranded M13 DNA targets did decrease dramatically at concentrations above 100 pm, probably because of target self-annealing (Fig . 3) . Decreased hybridization efficiency at higher target concentration was also observed with other plasmid DNA targets .
Kinetics of hybridization It was essential to obtain accurate values for the concentrations of both targets and probes to study hybridization kinetics . Determining the exact DNA concentration of oligonucleotides covalently labelled with alkaline phosphatase was difficult, because the protein part of the conjugate interfered with spectrophotometric measurements . The oligonucleotides in the conjugates were too short to be detected by Hoechst dye assay"" and too dilute to be detected by diphenylamine assay ." Although the protein concentration of the probe-conjugate stock was readily determined, we could not rule out the possibility that purification might have failed to remove a small percentage of the unconjugated enzyme, or that some DNA molecules might be positioned in a sterically inaccessible orientation . Electrophoretic analysis of the probe conjugates did indicate that a small amount of free enzyme was present, but this amount could not be quantitated . To avoid these potential sources of inaccuracy, the exact concentration of available DNA in alkaline phosphatase-labelled probes was analysed indirectly by titration with limiting amounts of single-stranded M13 target . The probe DNA concentration was determined by calculating the percentage of total enzyme activity that could hybridize to 10 fmol of target . Reaction times were extended to 2 h and the assay
Comparison of solution hybridization efficiencies
121
70
60 Y O m n v 50 .0 T L C 40 V d
N 0 u m
30
d .0 CL 0 0
20
E . U 10
0
100
200 300 400 Target concentration (pm)
500
Fig. 3. Comparison of hybridization efficiencies using a 32 P-labelled oligonucleotide probe with single-stranded (-0-) and double-stranded (-∎-) versions of the same M13 target DNA.
was performed at probe concentrations of 2 nM, 4 nM, and 8 nM to ensure the presence of sufficient DNA to drive hybridization to completion within the alloted time . Probe concentration determinations for alkaline phosphatase-conjugated probes were verified by a second indirect measurement. A short oligonucleotide complementary to the probe-conjugate was 32P labelled at its 3' end with a32 P-cordycepin and terminal transferase . The 3' end label was necessary because y 32 P-label at the 5' end of DNA molecules was cleaved by alkaline phosphatase during hybridization to enzyme-conjugated probes . An aliquot of alkaline phosphatase-labelled probe, estimated by protein assay to be 10 fmols, was hybridized to 200 fmol of its a 32 P-labelled complement in a volume of 100 µl for 1h . The percentage of radioactivity recovered in the resulting hybrids was used to calculate the amount of available DNA actually present in the original alkaline phosphatase-labelled probe . DNA concentration determinations on enzymelabelled probes made by measuring hybridization to excess short oligonucleotides agreed well with those made by measuring hybridization to limiting amounts of long single-stranded M13 target . For the alkaline phosphatase-labelled probe used in this study, the ratio of available DNA/protein was found to be 0 . 82 using single-stranded M13 target titration and 0 .81 using oligonucleotide titration . These ratios correlate well with the probe hybridization efficiency of
84% ± 4% obtained under conditions of target excess, indicating that the fraction of unhybridized alkalinephosphatase activity was consistent between three different experimental formats . The relative contributions of free enzyme contamination and steric hindrance to incomplete hybridization could not be determined, because the exact amount of free alkaline phosphatase could not be quantitated . Time course experiments were performed at 37 ° C under standard buffer conditions (6 x SSC, 0 . 1 mg ml - ' bovine serum albumin, and 10% formamide), with 100 pm target. The most useful data was obtained with probe concentrations ranging between 1 nM and 4 nM . At higher probe concentrations, resolution of the hybrid peak on Superose columns was impaired, and at lower concentrations the sensitivity of target detection was inadequate at early time points . Control experiments showed that the length of single stranded target did not affect the time course of hybridization (Fig. 4) . The rate of hybridization was also not affected by whether the target DNA was single-stranded or double-stranded (Figs 5 and 6) . To compare the hybridization rates of alkaline phosphatase-labelled and 32 P-labelled probes, a rate constant was calculated for each hybridization reaction . The formation of hybrids (PT) from probe (P) and target (T) should follow the general formula for determining the second-order rate constant (k e) of a bimolecular reaction : 14
S . Podell et al.
1 22
100
80 v a)
a LT drn 0
60
0 40
a0 20
0
15
30
45
60
Time (minutes)
Fig. 4 . Time course of hybridization for single-stranded M13 target (7350 nucleotides ; -0-) and synthetic oligonucleotide target (53 nucleotides ; -A-) . 32 P-labelled probe was present at 2-9 nM, and targets were 100 pm each .
Fig. 5 . Time course of hybridization for single-stranded (-0-) and double-stranded ( ∎ ) M13 targets with an alkaline phosphatase-labelled oligonucleotide probe . The concentration of available DNA in the alkaline phosphatase-conjugated probe was 1 . 3 nM, and each of the targets was present at 100 pm .
Comparison of solution hybridization efficiencies
123
Fig. 6. Time course of hybridization for single-stranded (-0-) and double-stranded (-∎-) M13 targers with a 33P-labelled oligonucleotide probe . 32 P-labelled probe was present at 1 .7 nM, and targets were each 100 pm .
[P] + [Tl --* [PT] d[PT1= kb[P] IT] dt The actual probe concentrations (1 .3-3 . 5 nM) used for time course experiments were in 13-35-fold excess of target concentrations, therefore probe concentrations were assumed to remain nearly constant during the reaction . Based on this assumption, a unimolecular rate constant was defined for the hybridization reaction : .
cular) rate constant was computed by dividing the unimolecular constant by the probe concentration . The average rate constant value determined from three experiments like those illustrated in Fig . 5 was 3.6 x 105 mol - ' s - ' ± 0 .9 X 10 5 mol - ' s - ' for the alkaline phosphatase-labelled probe. The average value for three experiments like those illustrated in Fig . 6, with a 32 P-labelled version of the same probe, was 5.8 x 105 mol - ' s - ' f 0. 8 x 105 mol - ' s - ' . These data suggest that alkaline phosphatase-labelled probes hybridize at a rate that is slightly slower, but very close to that of their 32P-labelled counterparts.
k u = kb[Pl DISCUSSION If k~ is substituted into the reaction rate equation, the equation becomes the following : d[PT] = k [T] dt When this equation is solved, then for time (t), PT(t) = [PT,1(1- a -ku`) Experimental data for the time course of hybridization with labelled oligonucleotides probes was fitted to an exponential curve with the above formula by nonlinear least squares analysis . The second order (bimole-
The efficient use of oligonucleotides in solution hybridization for quantitative DNA analysis requires the optimization of a number of time and concentration parameters . The chromatographic method of measuring hybridization efficiency presented here has provided quantitative data for these parameters, under conditions where neither spectrophotometric assays nor hydroxyapatite columns would be useful . The single-stranded or double-stranded nature of long targets strongly influenced the efficiency of solution hybridization with oligonucleotide probes, especially at target concentrations exceeding 100 pm . To avoid strand re-annealing problems, double-
S. Podell et al.
124
stranded target concentrations had to be kept at or
REFERENCES
below 100 pm . The concentration of single-stranded DNA targets was not nearly as critical, up to at least 500 pm. Hybridization rate constants were calculated for the purpose of comparing the efficiency of probes labelled with alkaline phosphatase to those labelled with 32 P . The rate constants found for alkaline phosphatase-labelled probes were slightly slower than those for 32 P-labelled probes, but the difference between the two values was relatively small . The conditions used to determine second order reaction rates by gel filtration chromatrography were chosen to give an optimal signal to noise ratio for the particular probes being tested . For this reason, the rates obtained could not be compared directly with those measured for the renaturation of DNA under more ideal conditions . The temperature chosen (35°C below the T m of the probe, calculated by the method of McGraw et
al .)" and the ionic strength (1 . 0 m Na + )
were sufficient to allow hybridization at the maximum possible rate, but the presence of 10% formamide and the contribution of additional viscosity by bovine serum albumin would be expected to slow the reaction rate . The range of probe concentrations used was fairly narrow, due to technical constraints . Signal detection was difficult at probe concentrations below 1 nM, and resolution of hybrid and free probe peaks was impaired at probe concentrations above 4 nnn . Nevertheless, second order rate constants measured by gel filtration chromatography were consistent with those measured spectrophotometrically for the renaturation of a 12-base oligonucleotide sequence, reported to be 10 5 mol -1 s -1, 16 and with those measured by psoralen crosslinking for the hybridization of a 25 base oligonucleotide to M13 DNA, reported to be 1 . 5 x 106 mol -1 s -1 .17 The maximum hybridization efficiencies attainable for probes labelled with either alkaline phosphatase or 32 P were quite similar. The 5-10% difference in hybridization efficiency observed between the two types of marker was probably due to variation in the effectiveness of the purification methods used to separate labelled probe from free 32 P or unconjugated alkaline phosphatase . Chromatographic separation of hybridization mixtures by gel filtration FPLC has thus allowed quantitative comparison of the maximum hybridization efficiencies attainable by these different types of labelled probes, thus allowing for comparisons which could not have been achieved by other methods currently available .
1 . Wetmur, J . G . & Davidson, N . (1968) . Kinetics of renaturation of DNA . journal of Molecular Biology 32, 349-70. 2 . Casey, J . & Davidson, N . (1977) . Rates of formation and thermal stabilities of RNA : DNA and DNA : DNA duplexes at high concentrations of formamide . Nucleic Acids Research 4, 1539-52 . 3 . Britten, R . J . & Davidson, E . H . (1985) . In Nucleic Acid Hybridization, A Practical Approach . (Hames, B . D . & Higgins, S . J ., eds) pp . 3-15 . Washington D .C . : IRL Press . 4 . Jablonski, E ., Moomaw, E . W., Tullis, R . H . & Ruth, ) . L . (1986) . Preparation of oligodeoxynucleotide-alkaline phosphatase conjugates and their use as hybridization probes . Nuecleic Acids Research 14, 6115-29 . 5 . Durnham, D . M . & Palmiter, R . D . (1983) . A practical approach for quantitating specific mRNAs by solution hybridization . Analytical Biochemistry 131, 385-93 . 6. Tu, Chen-Pei D . & Cohen, S . N . (1980) . 3' end labeling of DNA with [all PJ-cordycepin-5'-tri phosphate . Gene 10, 177-83 . 7 . Ruth, J . L . (1984) . Chemical synthesis of non-radioactively-labeled DNA hybridization probes . DNA 3, 123 . 8 . Selby, M . J ., Barta, A ., Baxter, J . D ., Bell, G . I . & Eberhard, N . L . (1984) . Analysis of a major human chorionic somatomammotropin gene . journal of Biological Chemistry 259, 13131-8 . 9 . Bradford, M . M . (1972) . A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding . Analytical Biochemistry 72, 248-54 . 10 . Maniatis, T ., Fritsch, E. F . & Sambrook, J . (1982) . In Molecular Cloning : A Laboratory Manual . p . 122 . New York : Cold Spring Habor Laboratory . 11 . Labarca, C . & Paigen, K . (1980) . A simple, rapid, and sensitive DNA assay procedure . Analytical Biochemistry 102, 344-52 . 12 . Cesarone, C . F ., Bolognesi, C . & Santi, L . (1979) . Improved microfluorometric DNA determination in biological material using 33258 Hoechst . Analytical Biochemistry 100, 188-97 . 13 . Burton, K . (1968). Determination of DNA concentration with diphenylamine . Methods Enzymol XII, 163-6 . 14 . Young, B . D . & Anderson, M . L. M. (1985) . In Nucleic Acid Hybridization, A Practical Approach . (Names, B . D . & Higgins, S . J ., eds) pp . 47-71 . Washington D .C . : IRL Press . 15 . McGraw, R . A ., Steffe, E . K . & Baxter, S . A . (1990) . Sequence-dependent oligonucleotide-target duplex stabilities : rules from empirical studies with a set of twenty-mers . Bio Techniques 8, 674-8 . 16 . Chu, Y . C . & Tinoco, I . (1983) . Temperature-jump kinetics of the dC-G-T-G-A-A-T-T-C-G-C-G double helix containing a G-T base pair and the dC-G-C-A-G-A-A-TT-C-G-C-G double helix containing an extra adenine . Biopolymers 22, 1235-46. 17. Gamper, H . B ., Cimino, G . B . & Hearst, J. E . (1987) . Solution hybridization of crosslinkable DNA oligonucleotides to bacteriophage M13 DNA . journal of Molecular Biology 197, 349-62 .
