`
`Directly labeled DNA probes using fluorescent nucleotides
`with different length linkers
`
`Zhengrong Zhu1, Jean Cha<>2, Hong Yu1,+ and Alan S.Waggoner2·*
`Center for Light Microscope Imaging and Biotechnology and Departments of 1Chemistry and
`2Biological Sciences, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA
`
`Received February 22, 1994; Revised and Accepted May 25, 1994
`
`ABSTRACT
`Directly labeled fluorescent DNA probes have been
`made by nick translation and PCR using dUTP attached
`to the fluorescent label, Cy3, with different length
`llnkers. With preparation of probes by PCR we find that
`llnker length affects the efficiency of Incorporation of
`Cy3-dUTP, the yield of labeled probe, and the signal
`Intensity of labeled probes hybridized to chromosome
`target sequences. For nick translation and PCR, both
`the
`level of Incorporation and the hybridization
`fluorescence slgnal increased in parallel when the
`length of the linker arm Is increased. Under optimal
`conditions, PCR yielded more densely labeled probes,
`however, the yield of PCR labeled probe decreased with
`greater linear density of labeling. By using a
`Cy3-modifled dUTP with the longest llnker under
`optimal conditions it was possible to label up to 28%
`of the possible substitution sites on the target DNA with
`reasonable yleld by PCR and 18% by nick translation.
`A mechanism involving sterlc Interactions between the
`polymerase, cyanlne-labeled sites on template and
`extending chains and the modified dUTP substrate is
`proposed to explaln the Inverse correlation between the
`labellng efficiency and the yield of DNA probe synthesis
`by PCR.
`
`INTRODUCTION
`
`Since its invention, (1,2) the technique of in situ hybridization
`(ISH) has become increasingly important in biological and
`diagnostic research (3). Fluorescent labels have gradually
`replaced the radioisotopic ones due to safety concerns, high spatial
`resolution, long shelf life, short detection time and simultaneous
`detection of multiple sequences in single cells. DNA probes are
`commonly labeled by nick translation (3, 4, 5). In recent years
`polymerase chain reaction (PCR) has been introduced to
`synthesize DNA probes (6, 7).
`Most fluorescent in situ hybridization (FISH) experiments have
`been done using indirectly reagents, such as avidins and
`antibodies, for visualization of probe signals (3, 4). This requires
`
`*To whom correspondence should be addressed
`
`+Present address: Abbot Diagnostic Division, Department 9MG, Building AP20.
`
`a post-hybridization labeling step, which is time consuming and
`may
`increase background fluorescence. Additional signal
`amplification steps are possible to increase the intensity of the
`signal (3, 4), but again, background fluorescence may increase.
`If fluorophores are directly attached to DNA, extra steps can be
`eliminated and signal-to-background ratio may be increased even
`if the overall fluorescence intensity is somewhat reduced
`compared with indirect labeling.
`In our previous study (8), Cy3 covalently conjugated through
`a short linker to dUTP was shown to be incorporated into DNA
`by nick translation and PCR. Other investigators have shown that
`if a linker molecule is inserted between biotin and a nucleotide
`molecule for nick translation, the labeling efficiency and
`hybridization detectability of Southern blotting experiment are
`improved (9, 10). Here we report systematic study of linker
`length on DNA probe labeling efficiency, yield of the PCR
`synthesis, and FISH signal intensity using Cy3 linked to dUTP.
`
`MATERIALS AND METHODS
`Modified dUTPs
`Cy3.29.OSu synthesized as described previously (11), contains
`6 atoms between the fluorophore ring system and the active ester
`group. For addition of 7 atoms to the linker, a mixture of 100
`mg Cy3.29.OSu and 18 mg 6-amino caproic acid (Aldrich) was
`dissolved in 15 ml sodium bicarbonate buffer (0. lM, pH = 9.4)
`and stirred at room temperature overnight. The product
`Cy3.29-13-OH (13 denotes the number of atoms between
`fluorophore ring system and carboxylic group) was isolated by
`reversed-phase (RP) chromatography (C-18) using water(cid:173)
`methanol mixture as eluent and dried down.
`Cy3.29-13-OH was activated to succinimidyl ester,
`Cy3.29-13-OSu, by methods used to prepare Cy3.29.OSu (11).
`The same procedure was repeated to conjugate another 6-amino
`caproic acid with Cy3.29-13-0Su to obtain Cy3.29-20-0H and
`to form the activated Cy3.29-20-OSu. The chemical structures
`were confirmed by 1H NMR. Cy3-x-dUTP (x = 10, 17, 24)
`was synthesized as described (8) The concentrations were
`
`Columbia Ex. 2033
`Illumina, Inc. v. The Trustees
`of Columbia University
`in the City of New York
`IPR2020-01177
`
`
`
`3418-3422 Nucleic Acids Research, 1994, Vol. 22, No. 16
`
`Directly labeled DNA probes using fluorescent nucleotides
`with different length linkers
`
`Zhengrong Zhu1, Jean Cha<>2, Hong Yu1,+ and Alan S.Waggoner2·*
`Center for Light Microscope Imaging and Biotechnology and Departments of 1Chemistry and
`2Biological Sciences, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, USA
`
`Received February 22, 1994; Revised and Accepted May 25, 1994
`
`ABSTRACT
`Directly labeled fluorescent DNA probes have been
`made by nick translation and PCR using dUTP attached
`to the fluorescent label, Cy3, with different length
`llnkers. With preparation of probes by PCR we find that
`llnker length affects the efficiency of Incorporation of
`Cy3-dUTP, the yield of labeled probe, and the signal
`Intensity of labeled probes hybridized to chromosome
`target sequences. For nick translation and PCR, both
`the level of Incorporation and the hybridization
`fluorescence slgnal increased in parallel when the
`length of the linker arm Is increased. Under optimal
`conditions, PCR yielded more densely labeled probes,
`however, the yield of PCR labeled probe decreased with
`greater linear density of labeling. By using a
`Cy3-modifled dUTP with the longest llnker under
`optimal conditions it was possible to label up to 28%
`of the possible substitution sites on the target DNA with
`reasonable yleld by PCR and 18% by nick translation.
`A mechanism involving sterlc Interactions between the
`polymerase, cyanlne-labeled sites on template and
`extending chains and the modified dUTP substrate is
`proposed to explaln the Inverse correlation between the
`labellng efficiency and the yield of DNA probe synthesis
`by PCR.
`
`INTRODUCTION
`
`Since its invention, (1,2) the technique of in situ hybridization
`(ISH) has become increasingly important in biological and
`diagnostic research (3). Fluorescent labels have gradually
`replaced the radioisotopic ones due to safety concerns, high spatial
`resolution, long shelf life, short detection time and simultaneous
`detection of multiple sequences in single cells. DNA probes are
`commonly labeled by nick translation (3, 4, 5). In recent years
`polymerase chain reaction (PCR) has been introduced to
`synthesize DNA probes (6, 7).
`Most fluorescent in situ hybridization (FISH) experiments have
`been done using indirectly reagents, such as avidins and
`antibodies, for visualization of probe signals (3, 4). This requires
`
`*To whom correspondence should be addressed
`
`+Present address: Abbot Diagnostic Division, Department 9MG, Building AP20.
`
`a post-hybridization labeling step, which is time consuming and
`may
`increase background fluorescence. Additional signal
`amplification steps are possible to increase the intensity of the
`signal (3, 4), but again, background fluorescence may increase.
`If fluorophores are directly attached to DNA, extra steps can be
`eliminated and signal-to-background ratio may be increased even
`if the overall fluorescence intensity is somewhat reduced
`compared with indirect labeling.
`In our previous study (8), Cy3 covalently conjugated through
`a short linker to dUTP was shown to be incorporated into DNA
`by nick translation and PCR. Other investigators have shown that
`if a linker molecule is inserted between biotin and a nucleotide
`molecule for nick translation, the labeling efficiency and
`hybridization detectability of Southern blotting experiment are
`improved (9, 10). Here we report systematic study of linker
`length on DNA probe labeling efficiency, yield of the PCR
`synthesis, and FISH signal intensity using Cy3 linked to dUTP.
`
`MATERIALS AND METHODS
`Modified dUTPs
`Cy3.29.OSu synthesized as described previously (11), contains
`6 atoms between the fluorophore ring system and the active ester
`group. For addition of 7 atoms to the linker, a mixture of 100
`mg Cy3.29.OSu and 18 mg 6-amino caproic acid (Aldrich) was
`dissolved in 15 ml sodium bicarbonate buffer (0. lM, pH = 9.4)
`and stirred at room temperature overnight. The product
`Cy3.29-13-OH (13 denotes the number of atoms between
`fluorophore ring system and carboxylic group) was isolated by
`reversed-phase (RP) chromatography (C-18) using water(cid:173)
`methanol mixture as eluent and dried down.
`Cy3.29-13-OH was activated to succinimidyl ester,
`Cy3.29-13-OSu, by methods used to prepare Cy3.29.OSu (11).
`The same procedure was repeated to conjugate another 6-amino
`caproic acid with Cy3.29-13-0Su to obtain Cy3.29-20-0H and
`to form the activated Cy3.29-20-OSu. The chemical structures
`were confirmed by 1H NMR. Cy3-x-dUTP (x = 10, 17, 24)
`was synthesized as described (8) The concentrations were
`
`
`
`PCR
`PCR was generally carried out according to procedure of Yu (8).
`The PCR reaction mixture contains l .5mM of MgC12•
`When a certain amount of modified dUTP was used instead
`of dTTP, the total concentration of modified dUTP and dTTP
`was kept the same as the concentration of each of other 3
`nucleotides because the maximum fidelity of polymerization
`requires a balanced concentration of each of the 4 dNTPs (12).
`The template for all nick translation and PCR experiments was
`a 900bp chromosome 1 insert (ATCC cat. no. 59863) in plasmid
`DNA. The primers were two ends of the vector (8), which leads
`to amplification of the entire 900bp insert. For quantification of
`DNA synthesis and Cy3 incorporation, the 5 '-end of each primer
`was labeled with Cy5.18.OSu as previously described (13).
`PCR was performed in a DNA Thermal Cycler (Perkin Elmer
`Cetus). The temperature cycling parameters were similar to the
`ones used by Ried et al. (14). After an initial denaturation at
`94°C for 4 min, 30 cycles (except cycle number study) of PCR
`were carried out with denaturation at 94 °C for 1 min, annealing
`at 52°C for 2 min and extension at 72°C for 3 min. Prep-A(cid:173)
`Gene method (BioRad) was used for purification to remove
`unreacted Cy3-x-dUTP and excess Cy5 labeled primers. The
`molecular weight of PCR product was examined using a 1 %
`agarose gel with 0.5 µg/ml ethidium bromide.
`In order to obtain the labeling efficiency and the yield of PCR,
`the product was dissolved in 140 µl sterile water for fluorescence
`measurement of Cy3 and Cy5 with a Spex Fluorolog 2
`spectrofluorometer (Spex Industries, Inc.). Calibration of
`
`0 3S~ ~ so ;
`VN~1~N0
`
`01 )
`[
`HN VHN
`
`n
`
`J_
`O
`0
`0
`0
`N
`11
`11
`_
`11
`0-P-O-P-O-P-O~
`o
`I_
`I -
`I -
`0
`0
`0
`
`OH
`
`n = 0 Cy3-10-dUTP
`n = 1 Cy3-17-dUTP
`
`n = 2 Cy3-24-dUTP
`
`Figure 1. Chemical structures of Cy3-modified dUTPs.
`
`Nucleic Acids Research, 1994, Vol. 22, No. 16 3419
`
`fluorescence intensities was accomplished by measuring the
`fluorescence of known concentration of Cy3 and Cy5, both in
`a purified carboxylic acid form. Using the calibration data, the
`Cy3 and Cy5 emission intensities of labeled DNA were converted
`to Cy3 and Cy5 molar concentrations. Since Cy5 concentration
`is identical to the concentration of DNA (use of Cy5 labeled
`primers),
`it was possible to calculate from fluorescence
`measurements the amount of DNA produced in PCR, which was
`taken as the yield of the reaction. The Cy3/Cy5 molar ratio
`corresponds to the number of Cy3 molecules per DNA chain
`of PCR product. Assuming 1/4 of bases of this DNA are thymine,
`the Cy3/Cy5 molar ratio divided by 225 (~) can be used as
`an estimate of the degree of substitution of Cy3-U (henceforth
`U*) for T in product DNA, defined as SProd. Thus, SProd
`= u?:T. The quantum yield (Q. Y.) of Cy3-labeled DNA was
`determined as previously described (15).
`Since the 900 bp labeled DNA probes are too long for optimal
`FISH, the PCR product was incubated with 1/350 unit of DNase
`I in nick translation buffer at 37°C for 5 min. A 1 % agarose
`gel with 0.5 µg/ml ethidium bromide was used to check the chain
`length (Optimal length is 200-400 bp (3, 4).). The DNA
`fragments were ethanol-precipitated and reconstituted in 150 µl
`sterile water. The ratio of substitution and amount of DNA
`fragments were determined again by absorption spectroscopy with
`a HP 8452 diode array spectrophotometer (Hewelett-Packard).
`The average extinction coefficient of a base in double stranded
`DNA at 260 nm was 6500 L/mol.cm (16).
`
`Labeling probes by nick translation
`Nick translation was carried out under sequential reaction
`conditions found to be optimal for cyanine-modified dUTPs (8).
`
`In situ hybridization and signal analysis
`A. common procedure was used for FISH and an imaging
`rmcroscope system was used to quantify hybridiz.ation signals (8).
`For each slide, at least 24 FISH signals were collected.
`Hybridiz.ation signals were analyzed by 'spot analysis' software.
`First an algorithm was used to define the boundaries of each
`hybridiz.ation spot. Then fluorescence signals within the spot were
`quantified and an average background measurement was made
`in an area outside of the spot boundary where the spot signal
`was no longer significant. Two parameters were used to describe
`the FISH signals. The net spot intensity (NSI) was the sum of
`all pixel intensities within the spot minus the background net
`intensity of an area equivalent to the spot. The signal-to(cid:173)
`background ratio (Sib) is the NSI divided by the background
`intensity of an area equivalent to the spot.
`
`RESULTS AND DISCUSSION
`
`PCR and nick translation reactions vary from day-to-day
`depending on the activity of the enzyme and reaction conditions.
`Therefore experiments presented in the same figure or same table
`have been done in the same batch for maximum consistency. The
`same trends were observed for data obtained on different days
`The chemical structures of Cy3-modified dUTPs (dU*TPs) are
`shown in Fig. 1. In this work we are concerned with optimizing
`the substitution efficiency and the yield of DNA probes directly
`labeled by PCR or nick translation with dU*TPs. The key
`variables under our control are the length of the linker arm
`between the ring structure of the fluorescent label and the uracil
`
`
`
`3420 Nucleic Acids Research, 1994, Vol. 22, No. 16
`
`ring, and the ratio of substitution of dU*TP for dTTP in the
`reaction mixture. We define the ratio of substitution of modified
`
`uracil for thymine i.e., u~:T in the reaction mixture and in the
`
`product as SRxn and SProd, respectively. Thus, the efficiency of
`labeling is quantified by SProd and a value of 1 means that all
`possible sites for incorporation of a thymine contain a modified
`uracil after nick translation or PCR.
`The amount of incorporation of Cy3 into probe can be
`determined by two methods (see 'Materials and Methods' for
`details). In one method, an absorption spectrum of purified
`
`a
`
`0.8
`
`0.6
`1
`0.4
`
`l:'l.l
`
`0.2
`
`0
`
`0
`
`b
`
`12
`
`10
`
`a.I
`
`8 = E
`'ii ;;; 4
`
`~ 6
`"Cl
`
`2
`
`0
`
`,(cid:127) 2L
`
`• • • (cid:127)'
`
`elL
`
`.... ·· ...... --
`·····:.---
`
`........ :.:.1•:.:.:. -
`
`OL
`
`0.2
`
`0.4
`
`0.6
`SRxn
`
`0.8
`
`... _ _....
`......
`
`\\
`. ....--•---411.
`OL
`
`..,!LI\
`IL
`' \ ,
`
`\
`
`'
`' \
`
`\
`
`0
`
`0.2
`
`0.4
`
`0.6
`situ
`
`0.8
`
`1
`
`Figure 2. a. s_ vs. ~ of PCR for modified dUTPs. We found that the high
`base concentration does not significantly improve the yield and higher base
`concentrations are known to cause lower fidelity of polymeri7.ation (12). Therefore
`a low base concentration, (i() mM, was used for all other PCR experiments. S(cid:173)
`and the yield were detennined by fluorescence measurement. OL = Cy3-10-dUTP;
`IL = Cy3-17-dUTP; 2L = Cy3-24-dUTP. b. Yield vs. ~ of PCR for
`modified dUTPs. The yield was determined by fluorescence measurement. Same
`samples as a.
`
`product yields the absorbance of nucleotide bases at 260 nm and
`the absorbance of Cy3 at 550 nm. The dye/base ratio can be
`calculated from this data with knowledge of the extinction
`coefficients. A second method involving measurement of
`fluorescence is useful only for PCR but it is more sensitive and
`requires less material. Primers labeled at the 5'-end with a
`different fluorophore, Cy5, are used in this method to determine
`the number of copies of DNA synthesized during amplification.
`The ratio of Cy3 fluorescence to Cy5 fluorescence allows
`calculation of the number of Cy3 molecules incorporated per copy
`of probe DNA. Quantification of SProd and probe yield by
`fluorescence measurement requires
`that the fluorescence
`efficiency be independent of the linear density of Cy3 molecules
`on the probe and that the Cy5 molecule is not involved in energy
`transfer from Cy3 molecules on the probe. With 100%
`substitution of U* for T there should be an average of one Cy3
`molecule every 4 bases. In our experiments, there is no indication
`of significant ( > 1.5 x) fluorescence quenching at labeling
`densities up to 38 % substitution of U* for T. And determined
`SProd relative order is not affected (data not shown). According
`to previous calculations, the energy transfer from Cy3 to Cy5
`can be neglected because the length of the primer sequence that
`separates them (8). Thus SProd and the yield of labeled DNA
`determined by fluorescence measurement in this study should be
`reasonably accurate.
`
`Incorporation of Cy3-modified dUTP by PCR
`The yield of Cy3-labeled probe generated by PCR comes to a
`plateau region after approximately 20 cycles. The yield of PCR
`product in a reaction using only dTTP and no dU*TP reaches
`a similar plateau. Although we have not tested thermal stability
`of dU*TP, this result shows that dU*TP is almost as stable as
`dTTP. Otherwise we will expect to see a much higher plateau
`of the reaction using only dTTP and no dU*TP. Because there
`was not much improvement in the yield after 30 cycles, all other
`PCR experiments were carried out at 30 cycles. SProd does not
`change with the cycle number .
`The observation that the yield of probe using a 50:50 mixture
`of dTTP and dU*TP (SRxn = 0.5) is almost the same as the
`yield when dTTP alone is used indicates that under the right
`conditions incorporation of dU*TP does not interfere with the
`formation of complete PCR product. However, this does not
`imply that dU*TP is as good a substrate as dTTP. If the reactivity
`of the modified nucleotide is the same as natural nucleotide in
`PCR, there should be a linear relationship between SProd and
`~ with a slope of 1. In Fig. 2a, even the highest ~ is below
`the line with a slope of 1, which means that Tag polymerase
`favors the natural nucleotide over the modified nucleotide even
`under the best reaction conditions. The Tag enzyme normally
`shows not much preference for uracil over thymine ( 17).
`Therefore the fluorophore and linker must account for the reduced
`capability of dU*TP to act as a substrate.
`The efficiency of labeling DNA with Cy3 in a PCR reaction
`is increased when SRxn is increased (Fig. 2a). Thus, with
`increased SRxn, the probability of substitution of dU*TP for T
`during incorporation is greater and leads to higher SProd. The
`increased linear density of Cy3 molecules on product that occurs
`with higher SRxn is also reflected in the decreased mobility of
`these PCR products on electrophoresis gels. It has been reported
`that the incorporated digoxigenin can cause the retardation of
`mobility of PCR products on the electrophoresis gel (18, 19).
`
`
`
`Nucleic Acids Research, 1994, Vol. 22, No. 16 3421
`
`Our early study of nick translation with a different template (a
`Sstl fragment) yielded similar results (20). This Sstl fragment
`is a moderately repetitive sequence of2.5 kb appearing in tandem
`arrays on human chromosome 19 and less frequently on
`chromosome 4 (21, 22). Others have reported that as the distance
`between biotin and nucleotide was increased, the labeling
`efficiency by nick translation first increased and then stayed
`constant or decreased (9, 10). Apparently steric interactions with
`the fluorophore on a shorter tether lead to less efficient
`incorporation of U* in the nick translation reaction.
`Since a DNA probe made by nick translation is composed of
`extending fragments and residual fragments and only extending
`fragments can be labeled, overall SProd of a DNA probe depends
`upon both SProc1 of extending fragments and the ratio of
`extending fragments to residual fragments. In carrying out these
`reactions dU*TP completely replaces dTTP, so that in the
`extending fragments all T-sites contain U*. On the other hand,
`for PCR, the whole probe except for the short primer, can be
`labeled. Thus, with equally efficient incorporation by the
`polymerase, PCR should produce probes with an overall higher
`linear density of labeling. However, we have shown that there
`can not be complete substitution of dU*TP for dTTP in
`polymerase reaction. Even though SRxn is 1 in nick translation
`and SRxn is less than 1 in PCR, the optimal SProc1 of PCR is still
`higher than that of nick translation.
`
`Fluorescence in situ hybridization (FISH) with Cy3-labeled
`probes
`After having examined the labeling efficiency and yield of
`fluorescent DNA by PCR and nick translation, we now turn to
`the utility of such probes in FISH. Target DNA of chromosome
`1 in interphase nuclei of HeLa cells was hybridized with
`Cy3-modified probes. FISH signals were quantified by image
`cytometry.
`Table 1 shows that both the net spot intensity and the signal(cid:173)
`to-background ratio of signals follows the same trend as SProo
`of probes. This means that the long linker does not seem to affect
`probe mobility and entry to the nuclei. Since the background is
`not increased as fast as the signal intensity, the fidelity of
`
`Increasing the linker length also favors greater incorporation
`of dU*TP (Fig. 2a). It is possible that longer linkers reduce steric
`interactions between the Cy3 molecule and the polymerase-DNA
`complex thus making dU*TP with longer linkers better substrates.
`Although the efficiency of incorporation of modified bases by
`PCR improves as linker length or SRxn increases, the yield of
`labeled probe generated by multiple PCR cycles is reduced, and
`when 100% dU*TP was used instead of dTTP, there is no
`amplification (Fig. 2b). We found that the longer extension time
`does not raise SProo and the yield actually decreases. This may
`be caused by the loss of enzyme activity during long incubation
`at 72°C. An extension time as short as 1 minute does not affect
`the labeling efficiency and yield.
`Why does the yield of Cy3-labeled probe in a PCR reaction
`decrease when the labeling density increases? One explanation
`is based on the steric considerations. When more U* is
`incorporated into a DNA probe, there is a greater possibility of
`modified bases appearing in close proximity to one-another. In
`the next cycle, the labeled chain serves as the template. The
`presence of several Cy3 fluorophores in a small space on the
`template (or the template and extending chain) may inhibit the
`progression of the polymerase and lead to termination of the probe
`fragment. Since incomplete chains can not be used as templates
`in the next cycle, the yield is decreased. Finckh et al. noted a
`similar decrease in yield of labeled probe when the substitution
`of biotin modified dUTP was increased ( 19).
`According to this model, the yield is reduced because of a high
`labeling density that sterically inhibits translation of the labeled
`template and that can be generated by either using a high SRxn
`or by using probes with longer linkers. However, there is a
`compensating effect that appears with longer linkers that reduce
`steric constraints from the fluorophores. This effect shows up
`in Fig. 2b. Since the labeling density for Cy3-24-dUTP is greater
`than for Cy3-17-dUTP when SRxn is in the range of 0.5-0.9
`(Fig. 2a), it would be expected that the yield of reaction of
`Cy3-24-dUTP would be smaller. However, Fig. 2b shows that
`there is a greater yield of Cy3-24-dUTP than Cy3-17-dUTP over
`this range of SRxn. It is apparent that the longer linker facilitates
`chain extension even when the template is more heavily labeled.
`Therefore Cy3-24-dUTP is the best substrate for DNA probe
`labeling by PCR because it can give the highest labeling density
`with reasonable yield.
`
`Incorporating Cy3-modified dUTP by nick translation
`Table 1 shows that under optimal nick translation conditions
`SProc1 values for Cy3-17-dUTP and Cy3-24-dUTP are nearly
`equivalent but are about two-fold higher than for Cy3-10-dUTP.
`
`Table 1. Sproc1 and FISH result of DNA probes
`
`NT.0L
`NT.IL
`NT.2L
`PCR.0L
`PCR.IL
`PCR.2L
`
`SProd
`
`0.ll
`0.18
`0.18
`0.01
`0.14
`0.28
`
`NSI
`(IO")
`
`0.5 ± 0.3
`1.5 ± 0.6
`1.2 ± 0.6
`0.2 ± 0.1
`1.7 ± 1.3
`2.3 ± 1.6
`
`Sib
`
`0.12 ± 0.05
`0.24 ± 0.07
`0.22 ± 0.o7
`0.o7 ± 0.04
`0.23 ± 0.ll
`0.29 ± 0.08
`
`SProd was determined by absorbance measurement. PCR probes were the same
`as those for Fig. 3a after DNase digestion. NT = Nick translation; 0L =
`Cy3-10-dUTP; IL = Cy3-l7-<IUTP; 2L = Cy3-24-dUTP. The FISH parameters
`are described in • Materials and methods' section.
`
`Figure 3. FISH images by different DNA probes. The upper images were obtained
`using probes made by nick translation. The bottom images were obtained using
`probes made by PCR. From left to right: Cy3-10-dUTP, Cy3-17-dUTP,
`Cy3-24-dUTP.
`
`
`
`3422 Nucleic Acids Research, 1994, Vol. 22, No. 16
`
`incorporating modified nucleotide is high enough to serve the
`purpose of preparing DNA probes . Thus improving probe
`labeling is a way to optimize FISH.
`Since the FISH signal intensities obtained with PCR-generated
`probes are higher than those of probes made by nick translation
`(Table 1), and since PCR can label and amplify DNA probes
`simultaneously, PCR is a good labeling technique for preparation
`of directly labeled fluorescent DNA probes. Even higher labeling
`densities are achievable by PCR but at the price of low yield.
`FISH images of 6 kinds of probe are shown in Fig. 3. The
`signal intensity of each image is close to the respective average
`intensity of each kind of probe.
`
`ACKNOWLEDGEMENTS
`We would like to thank Dr Ratnakar Mujumdar and Mrs Swati
`Mujumdar for their help in organic synthesis. We are also grateful
`to Mr Alasdair Dow for his help in image instrumentation. This
`work was supported by NSF Grant BIR-8920118 and State of
`Pennsylvania Contract #93-110-0026 to Center for Light
`Microscope Imaging and Biotechnology.
`
`REFERENCES
`1. Pardue, M. L. and Gall, J. G. (1969) Proc. Natl. Acad. Sci. USA, 64,
`(i(l()-ti04.
`2. John, H., Bimstiel, M. and Jones, K. (1969) Nature, 223, 582-587.
`3. Keller, G. H. and Manak, M. M. (1989) DNA Probes. Stockton Press, New
`York.
`4. Trask, B. J. (1991) Methods in Cell Biology, 35, 3-35.
`5. Wetmur, J. G. (1991) Crit. Rev. Biochem. Mol. Biol., 26, 227-259.
`6. Ried, T., Balclin, A., Rand, T. C. and Ward, D. C. (1992)Proc. Natl. Acad.
`Aci. USA, 89, 1388-1392.
`7. Celecla, D., Bettag, U. and Cremer, C. (1992) BioTechniques, 12, 98-102.
`8. Yu, H., Chao, J., Patek, D., Mujumdar, R, Mujumdar, S. and Waggoner,
`A. S. Nucleic Acids Res., submitted.
`9. Leary, J. J., Brigati, D. J. and Ward, D. C. (1983) Proc. Natl. Acad. Sci.
`USA, 80, 4045-4049.
`10. Gebeyehu, G., Rao, P. Y., SooChan, P., Simms, D. A. and Klevan, L.
`(1987) Nucleic Acids Res., 15, 4513-4534.
`11. Mujumdar, R. B., Ernst, L. S., Mujumdar, S. R., Lewis, C. J. and
`Waggoner, A. S. (1993) Bioconjugate Chemistry, 4, 105-111.
`12. Eckert, K. A. and Kunkel, T. A. (1991) In McPherson, M. J., Quirke, P.
`and Taylor, G. R. (ed.), PCR. IRL, Oxford. 225-244.
`13. Yu, H., Ernst, L., Wagner, M. and Waggoner, A. (1992) Nucleic Acids
`Res., 20, 83-88.
`14. Ried, T., Baldini, A, Rand, T. C. and Ward, D. C. [1992) Proc. Natl. Acad.
`Sci. USA, 89, 1388-1392.
`15. Ernst, L.A., Gupta, R. K., Mujumdar, R. B. and Waggoner, A. S. (1989)
`Cytometry, 10, 3-10.
`16. Fasman, G. D., (1975) Handbook of Biochemistry and Molecular Biology.
`CRC Press, Cleveland, Vol I.
`17. Slupphaug, G., Alseth, I., eftedal, I., Volden, G. andKrokan, H. E. (1993)
`Anal. Biochem. 211, 164-169.
`18. Lanzillo, J. J. (1991) Anal. Biochem. 194, 45-53.
`19. Finckh, U., Lingenfelter, P. A. and Myerson, D. (1991) BioTechniques,
`10, 35-39.
`20. Zhu, Z. and Waggoner, A. S. unpublished result.
`21. Epstein, N. D., Karlsson, S., O'Brien, S., Moulton, A. and Nienhuis, A.
`W. (1987) Nucleic Acids Res., 15, 2327-2341.
`22. Samulski, R. J. and Xiao, X. personal communication.
`
`