The H ₆ -tagged oligonucleotides were synthesized by the convertible nucleoside approach ( 12 - 15 ) using the O ⁶ -phenyl-2'-deoxyinosine ([Phi]dI) phosphoramidite ( 16 , 17 ) along with PAC phosphoramidites (Pharmacia). The resin-bound oligonucleotide 5'-d[G([Phi]I) ₆ AGCGGATAACAATTTCACACAGG] and 5'-d[G([Phi]I) ₆ TCGTGACTGGGAAAACCCTGGCG] were deprotected by treatment with 1 ml concentrated (14 M) aqueous ammonium hydroxide at room temperature for 4 h and lyophilized to dryness on a Speed Vac (Savant). The DNA pellets were redissolved in 100 [mu]l 5 M aqueous histamine and incubated at 55^oC for 14 h. After cooling to room temperature, 300 [mu]l absolute ethanol (kept at -20^oC) was added, the mixture was chilled on crushed solid CO ₂ for 30 min, then centrifuged at 16 000 g for 30 min. The supernatants were discarded and the pellets washed with 200 ml 80% (v/v) aqueous ethanol solution (-20^oC). The pellets were dried on a Speed Vac, redissolved in 200 [mu]l formamide loading buffer (95% aqueous formamide, 20 mM EDTA, 0.05% each bromophenol blue and xylene cyanol), heated to 90^oC for 5 min, then loaded onto a 20% (19:1 acrylamide:bis) polyacrylamide gel (20 * 20 cm) containing 7 M urea. The gels were pre-run at 300-500 V in TBE buffer (90 mM Tris-borate, 2 mM EDTA, pH 8) for at least 1 h prior to loading the DNA. Following electrophoresis, the gel was removed from the glass plates, enclosed in Saran Wrap and placed over a TLC plate impregnated with fluorescent dye. The full-length DNA band was visualized using a hand-held UV lamp, excised from the gel with a sharp razor blade, placed in a 50 ml Falcon tube and crushed thoroughly using the polished end of a glass stirring rod. The crushed gel was soaked overnight at 37^oC in 10 ml 1 M triethylammonium bicarbonate (TEAB), pH 8.0. The supernatant was transferred to a new Falcon tube, and the crushed polyacrylamide was further extracted once with 5 ml 1 M TEAB. The combined TEAB solutions were loaded onto a C ₁₈ Sep-Pak cartridge (Waters), which had bee pre-washed by successive throughput of 5 ml 100% CH ₃ CN and 15 ml 50 mM TEAB. Following loading of the DNA, the Sep-Pak was washed with 2 ml 5% CH ₃ CN/95% 50 mM TEAB and eluted with 10 ml 30% CH ₃ CN/70% 50 mM TEAB into 1.5 ml Eppendorf tubes. The fractions were assayed by UV spectrophotometry, and those that contained a significant A ₂₆₀ were lyophilized to dryness in a SpeedVac. The lyophilized DNA pellets was combined in 50 [mu]l TE buffer to generate an oligonucleotide stock solution that was used directly in subsequent experiments. This procedure typically yields ~150 nmol highly purified H ₆ -tagged oligonucleotide using a 200 nmol resin; comparable yields were obtained for untagged oligonucleotides purified by the same procedure. We have not investigated purification of the crude H ₆ -tagged oligonucleotides using Ni ²⁺ -NTA chromatography, but this should work and would be sufficient for most applications.

Nucleoside composition analysis

The oligonucleotide sample (3 nmol) was digested with 0.2 U snake venom phosphodiesterase (Pharmacia) and 50 U Serratia endonuclease (`Benzonase', Merck) in 50 [mu]l buffer containing 100 mM NaCl, 14 mM MgCl ₂ , 100 mM Tris-HCl, pH 9.0 at room temperature for 2 h, then at 37^oC for 2 h. The buffer was adjusted to 100 [mu]l 0.1 mM ZnCl ₂ , 50 mM NaCl, 17 mM MgCl ₂ , 200 mM, 10 mM [beta]-mercaptoethanol, 200 mM Tris-HCl, pH 9.0, and 1 U calf intestinal alkaline phosphatase (Boehringer-Mannheim) was added, and the resulting mixture was incubated at 42^oC for 2 h. The digestion mixture was clarified by passage through a Millex 22 [mu]m filter and analyzed by reverse phase HPLC (Beckman Ultrasphere ODS, 4.6 * 250 mm) employing a photodiode array detector (Hewlett Packard LC 1090; Solvent A: 0.02 M KH ₂ PO ₄ , pH 5.6; Solvent B: 60:40 CH ₃ OH/H ₂ O; 1.5 mM/min; elution program: isocratic A for one min, 0-25% B in 10 min, 25-100% in 5 min, isocratic B for 10 min). Nucleosides were identified by comparison with authentic standards. A mock digest containing all digestion components except the oligonucleotide carried out as a control for contaminants introduced along with the buffers and enzymes.

Procedure for IMAC resolution of PCR-amplified DNA

PCR reactions, which follow standard procedures ( 4 ), employ 2.5 U Taq polymerase (Gibco/BRL), 50 pmol each primer and 0.2-0.5 [mu]g supercoiled plasmid template in a reaction volume of 100 [mu]l. To remove excess PCR reagents, 1 ml TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8) is added to the crude reaction mixture, which is then concentrated to a total volume of 50 [mu]l using a centrifugal dialysis cartridge (Centricon 30, Amicon). The DNA solution is transferred to an Eppendorf tube, to which is added 150 [mu]l binding buffer (6 M guanidine[middot]HCl, 10 mM Tris-HCl, pH 8.2). This mixture is heated for 5 min at 90^oC. Separately, to a 1.5 ml Eppendorf tube is added 250 [mu]l (bed volume) Ni ²⁺ -NTA-agarose resin and 1 ml binding buffer at room temperature. The hot DNA solution is added to the suspension of resin, and the resulting mixture is mixed for 1-1.5 min by vigorous shaking or repeated pipetting (longer mixing times result in reduced yields). The mixture is transferred to an empty 5 ml fritted column (Qiagen), and the flow-through (`unbound' fraction), which contains the unmodified strand, is collected into an Eppendorf tube. After repeated pipetting to ensure complete mixing, the unbound fractions are aliquoted into four Eppendorf tubes (300 [mu]l each) and held aside for further processing. Next, the resin is washed with 1 ml washing buffer (10 mM Tris-HCl, 5 mM imidazole, pH 8.0), which is discarded. The H6-containing strand is then eluted in 1.2 ml 200 mM aqueous imidazole solution. The imidazole eluate (`bound' fraction) is mixed thoroughly and aliquoted into four Eppendorf tubes (300 [mu]l each), to each of which is added 30 [mu]l 100 mM ethanolic 1,10-phenanthroline and 30 [mu]l 3 M aqueous NaOAc. No additional salt is added to the unbound fractions. To each of the unbound and bound fractions is added 900 [mu]l absolute ethanol (stored at -20^oC), then the tubes are vortexed briefly and chilled for 30 min on dry powdered CO ₂ . The tubes are microcentrifuged for 30 min at 16 000 g . The supernatant is removed and the pellet washed with 200 [mu]l 80% aqueous EtOH (-20^oC). Following removal of the ethanol solution, the tubes are dried by centrifugal lyophilization (SpeedVac, Savant). To each dry tube was added 50 [mu]l TE buffer. The DNA concentration was determined by UV spectrophotometry.

Dideoxy (Sanger) sequencing reactions

Sanger sequencing employed the Sequenase kit version 2.0 (US Biochemicals) using the protocol supplied by the manufacturer for ³⁵ S sequencing.

Figure 1 . Synthesis of H ₆ -tagged oligonucleotides by the convertible nucleoside approach. Treatment of an oligonucleotide bearing the convertible base O ⁶ -phenylinosine ([Phi]I) with aqueous histamine results in displacement of the O ⁶ -phenyl group to furnish 6-histaminylpurine (H). The H ₆ -tag consists of six successive 2'-deoxyribo-H nucleotide (dH) residues (box).

RESULTS AND DISCUSSION

Employed widely in affinity purification of proteins, IMAC relies on the ability of a strongly metal-coordinating polypeptide sequence ( 18 ) such as hexa-histidine (His ₆ ) ( 19 , 20 ) to bind specifically to a chromatography matrix containing a tightly chelated metal ion, usually Ni ²⁺ ( 21 - 23 ). Importantly, IMAC can be carried out in the presence of urea and guanidinium hydrochloride, because these additives do not interfere with the specific metal-ligand interactions on which the technique is based. In principle, this insensitivity to denaturants makes IMAC uniquely suited to the problem of DNA strand resolution. Implementation of an IMAC-based DNA purification strategy required the development of a polynucleotide equivalent of His ₆ .

Figure 2 . HPLC analysis of the products from enzymatic digestion of 1a with snake venom phosphodiesterase, benzonase and alkaline phosphatase. X is a contaminant peak that is introduced along with snake venom phosphodiesterase. The UV spectrum of X does not match that of any ordinary 2'-deoxynucleoside. The retention times and UV spectra of the four native nucleosides and the modified nucleoside (dH) were identical to those of authentic standards (not shown). The authentic standard for dH was generated by histamine displacement on O ⁶ -phenyl-2'-deoxyinosine (15).

We reasoned that a run of six nucleotides, each containing a suitably attached imidazole ligand, might serve as a ligand tag analogous to His ₆ . As an initial test of this concept, we synthesized DNA sequencing primers containing six successive 6-histaminylpurine (H) residues (H ₆ -tag) added onto the 5'-end (Fig. 1 ) through a post-synthetic modification procedure known as the convertible nucleoside approach ( 12 - 17 ). Briefly, resin-bound oligonucleotides containing O ⁶ -phenyl-2'-deoxyinosine ([Phi]dI) in place of the dH residue were deprotected from the resin by mild ammonia treatment (conc. NH ₄ OH, rt, 4 h), lyophilized to dryness, then treated with 5 M aqueous histamine (55^oC, 14 h) to convert the [Phi]dI residues to dH. The crude oligonucleotides were purified by denaturing polyacrylamide gel electrophoresis and then analyzed for constituent nucleosides by enzymatic digestion and HPLC. Figure 2 shows the HPLC trace for the nucleoside composition analysis of the H ₆ -tagged oligonucleotide 1a . In addition to the four peaks arising from the native nucleosides (dC, dG, T, dA), a peak due to the histamine-modified nucleoside dH is evident. The peak denoted X is a contaminant present in the mock digest, having been introduced along with the snake venom phosphodiesterase. Importantly, a peak at the retention time of [Phi]dI was not observed, thus indicating complete conversion of [Phi]dI to dH. To assess the ability of the H ₆ -tag to endow DNA with selective retentivity on a Ni ²⁺ -charged chelate resin, the H ₆ -tagged oligonucleotides ( 2a / 2b ) and their unmodified counterparts ( 1b / 2b ) were 5'-end labeled with ³² P, then passed in parallel through a Ni ²⁺ -nitrilotriacetic acid (NTA)-agarose resin, commercially available from Qiagen (Chatsworth, CA) ( 17 , 18 ). Whereas <5% of the radioactivity derived from the unmodified oligonucleotides bound the resin, >95% of that from the H ₆ -tagged oligonucleotides was adsorbed by the resin. All but a trace amount of the bound radioactivity could be eluted from the resin upon washing with 200 mM aqueous imidazole. These data demonstrate that H ₆ -tagged oligonucleotides can be selectively retained on Ni ²⁺ -NTA agarose, and efficiently eluted from the matrix by aqueous imidazole.

For the present IMAC-based strategy to be of significant practical value, it must be capable of resolving PCR-amplified duplex DNA into its two component strands. To test this application directly, we used one H ₆ -tagged primer plus one unmodified primer to amplify a 183 base-pair segment of the plasmid pUC18-mARRE2 ( 24 ; Fig. 3 a and b). pUC18-mARRE2 was derived from the commercial cloning vector pUC18 by inserting a segment of the murine interleukin-2 enhancer into the Bam HI site. We chose a pUC18 derivative for these experiments because such plasmids are widely employed in cloning and other subsequent manipulation that involve single-stranded DNA. Primers 1b and 2b, commercially available oligonucleotides that are commonly used to sequence pUC18 derivatives, were used in combination with the corresponding H ₆ -tagged oligonucleotides 1a and 2a to generate PCR products having H ₆ -tagged on any one strand. Namely, the PCR product generated using primers 1a + 2b will bear the H ₆ -tag on the `top' strand only, whereas that generated using 2a + 1b will bear the H <sub>6</sub> -tag on the `bottom' strand only (the designation of `top' and `bottom' strands refers to the sequence as illustrated in Figure 3 a). The yield of PCR product formed in these reactions was no less than that observed in parallel reactions with only unmodified primers (data not shown), thus indicating that the H ₆ -tag does not affect PCR amplification adversely. The PCR products were denatured in 6 M guanidinium[middot]HCl, then incubated batchwise with Ni ²⁺ -NTA-agarose. Following removal of the supernatant, which contains unbound DNA, the resin was washed and the bound DNA eluted with 200 mM imidazole. Aliquots of the crude PCR product and the two DNA-containing fractions from the IMAC step were 5'-end labeled with ³² P and analyzed by denaturing polyacrylamide gel electrophoresis (PAGE). The use of PAGE as an assay for strand resolution was made possible by the slightly reduced mobility of H ₆ -tagged DNA strands relative to their complementary strands generated by PCR (Fig. 4 , lanes 1 and 3). The retarded mobility of the H ₆ -tagged strand relative to the unmodified strand may arise from the partial positive charge of its six pendant imidazole moieties at pH 8 and a greater length, due to incomplete polymerization of the T ₆ -stretch opposite H ₆ during PCR. The identities of the strands were independently verified by PAGE analysis of PCR reactions that employed one ³² P-labeled primer and one non-radioactive primer. Although this slight difference in mobility provides a convenient means of assaying the strand resolution by IMAC, it is not sufficiently large to permit preparative resolution of relatively long PCR products by gel electrophoresis and extraction. The slight difference in intensity of the two bands in lanes 1 and 4 of Figure 4 results from differential kinetics of end-labeling, rather than differences in quantity of the DNA. The unbound fractions from IMAC (lanes 2 and 5) contained primarily the faster-migrating species, which corresponds to the respective unmodified strands, whereas the imidazole-eluted bound fractions (lanes 3 and 6) contained primarily the respective H ₆ -tagged strands. Quantitative phosphorimaging analysis from repeated runs of the strand resolution procedure revealed that the unmodified strand typically comprise >90% of the unbound DNA and the H ₆ -tagged DNA comprise >95% of the bound fractions. Thus we conclude that IMAC cleanly resolves a singly H ₆ -tagged PCR product into its two component strands.

Figure 3 . Overview of procedure for production of long single-stranded DNA molecules using the polymerase chain reaction (PCR) followed by immobilized metal affinity chromatography (IMAC). ( a ) Primers and double-stranded PCR template used in this study. Primers 1a / 1b contain a sequence that is identical to a stretch (shaded) of the `top' strand of the plasmid pUC18-mARRE2; they are extended in the rightward direction (5' -> 3') during PCR. Primers 2a / 2b contain a sequence that is identical to a stretch (shaded) of the `bottom' strand of pUC18-mARRE2; they are extended in the leftward direction. The shaded sequence in primers 1a and 2a denotes the H ₆ -tag. E and H denote the Eco RI and Hin dIII sites, respectively, of the pUC18 polylinker; a stretch of the murine IL-2 enhancer (mARRE2, hatched) was inserted into the Bam HI site to generate pUC18-mARRE2. ( b ) Schematic illustration of the procedure for PCR amplification to produce duplex DNA containing a H ₆ -tag on one strand, and resolution of the two constituent strands using IMAC.

Figure 4 . Denaturing polyacrylamide gel electrophoresis analysis of DNA strand resolution using IMAC. Each panel of three lanes represents the results obtained using a pair of PCR primers, one of which ( 1a or 2a ) contains an H ₆ -tag and the other of which ( 2b or 1b , respectively) is unmodified. C (lanes 1 and 4), controls showing the mixture of strands obtained directly by PCR, prior to IMAC resolution; U (lanes 2 and 5), unbound fractions from IMAC; B (lanes 3 and 6), bound fraction from IMAC, after elution with 200 mM imidazole.

To assay the biological activity of single-stranded DNA generated through the PCR-IMAC sequence, we employed the resolved strands as templates in Sanger dideoxy sequencing runs. Initially, the unmodified strands were found to be excellent sequencing templates, but the H ₆ -tagged strands failed to support template-directed polymerization. Eventually, it was discovered that the H ₆ -tagged DNA solutions contain adventitious Ni ²⁺ in amounts sufficient to cause profound inhibition of the DNA polymerase enzyme. This problem was overcome simply by adding 1,10-phenanthroline to the imidazole-containing eluate prior to precipitation with ethanol. Single-stranded H ₆ -tagged templates prepared in this manner give uniformly high-quality DNA sequence data. Figure 5 shows the results of sequencing runs for all four single-stranded templates generated in this study, two of which ( 1b and 2b ) represent unbound fractions from Ni ²⁺ -NTA-agarose and two of which ( 1a and 2a ) represent bound fractions eluted with 200 mM imidazole. DNA templates in Figure 4 are designated according to the PCR primer that they incorporate.

Figure 5 . Autoradiogram of ³⁵ S sequencing reactions using single-stranded template DNA generated by the described separation technique: 2b , unbound DNA from PCR/IMAC using primers 1a and 2b ; 1a , bound DNA from PCR/IMAC using primers 1a and 2b ; 1b , unbound DNA from PCR/IMAC using primers 1b and 2a ; 2a , bound DNA from PCR/IMAC using primers 1b and 2a .

CONCLUSIONS

In this report, we have described a method for the affinity purification of DNA using immobilized metal affinity chromatography. The method centers on the attachment to DNA of a ligand tag consisting of six successive 6-histaminylpurine residues (H ₆ -tag), which mediates selective adsorption onto a Ni ²⁺ -charged chelate resin. To incorporate the 6-histaminylpurine moiety into DNA, we relied on a post-synthetic method known as the convertible nucleoside approach ( 12 - 17 ); alternatively, the same operation could presumably be accomplished using standard DNA synthesis chemistry and a suitably protected dH phosphoramidite. The H moiety can be viewed as an adenine residue containing a tethered imidazole ligand, and was chosen for the present study primarily for reasons of synthetic expediency. It seems likely that imidazole ligands similarly tethered to other DNA bases or other DNA-like scaffolds will provide alternative reagents for use in PCR-IMAC. We envision that this chemistry will facilitate a wide variety of operations in molecular biology, including automated DNA sequencing, hybridization screening of DNA libraries, assembly of gene cassettes, run-off transcription, site-directed mutagenesis and footprinting of protein-DNA complexes by template-directed interference footprinting ( 25 ). The chemistry outlined here has been adapted for use in IMAC purification of RNA (Allerson,C.R., Chen,S. and Verdine,G.L., submitted for publication; 26 ).

ACKNOWLEDGEMENTS

This research was supported by NIH grant GM44853. We thank O. D. Schärer and C. R. Allerson for a critical reading of the manuscript and members of the Verdine group for helpful comments.

REFERENCES

1 Watson,J.D., Gilman,M., Witkowski,J. and Zoller,M.J. (1992) Recombinant DNA. W.H. Freeman, New York.

2 Saiki,R.K., Scharf,S., Faloona,F., Mullis,K.B., Horn,G.T., Erlich,H.A. and Arnheim,N. (1985) Science, 230, 1350-1354.

3 Saiki,R.K., Gelfand,D.H., Stoffel,S., Scharf,S.J., Higuchi,R., Horn,G.T., Mullis,K.B. and Erlich,H.A. (1988) Science, 239, 487-491.

4 Erlich,H.A. (1989) PCR Technology. Stockton Press, New York.

5 Gyllensten,U.B. and Erlich,H.A. (1988) Proc. Natl. Acad. Sci. USA, 85, 7652-7656.

6 Hayashibara,K.C. and Verdine,G.L. (1992) Biochemistry, 31, 11265-11273.

7 Hayashibara,K.C. and Verdine,G.L. (1991) J. Am. Chem. Soc., 113, 5104-5106.

8 Mascareñas,J.L., Hayashibara,K.C. and Verdine,G.L. (1993) J. Am. Chem. Soc., 115, 373-374.

9 Mitchell,L.G. and Merril,C.R. (1989) Anal. Biochem., 178, 239-242.

10 Bowman,B.H. and Palumbi,S.R. (1993) In Zimmer,E.A., White,T.J., Cann,R.L. and Wilson,A.C. (ed.), Methods in Enzymology. Academic Press, New York, Vol. 224, pp. 399-405.

11 Porath,J., Carlsson,J., Olsson,I. and Belfrage,G. (1975) Nature, 258, 598-599.

12 MacMillan,A.M. and Verdine,G.L. (1990) J. Org. Chem., 55, 5931-5933.

13 MacMillan,A.M. and Verdine,G.L. (1991) Tetrahedron, 47, 2603-2616.

14 Ferentz,A.E. and Verdine,G.L. (1994) In Eckstein,F. and Lilley,D.M. J. (ed.), Nucleic Acids and Molecular Biology. Springer-Verlag, Berlin, Vol. 8, pp. 14-40.

15 Ferentz,A.E. and Verdine,G.L. (1992) Nucleosides Nucleotides, 11, 1749-1763.

16 Ferentz,A.E. and Verdine,G.L. (1991) J. Am. Chem. Soc., 113, 4000-4002.

17 Ferentz,A.E., Keating,T.A. and Verdine,G.L. (1993) J. Am. Chem. Soc., 115, 9006-9014.

18 Smith,M.C., Furman,T.C., Ingolia,T.D. and Pidgeon,C. (1988) J. Biol. Chem., 263, 7211-7215.

19 Hochuli,E. and Piesecki,S. (1992) Methods, 4, 68-72.

20 Hochuli,E., Bannwarth,W., Döbeli,H., Gentz,R. and Stüber,D. (1988) BioTechnology, 6, 1321-1325.

21 Sulikowski,E. (1989) BioEssays, 10, 170-175.

22 Wong,J.W., Albright,R.L. and Wang,N.-H.L. (1991) Sep. Purif. Methods, 20, 49-106.

23 Arnold,F.H. (1991) BioTechnology, 9, 151-156.

24 Chen,L., Jain,J., Oakley,M.G., Glover,J.N.M., Dervan,P.B., Hogan,P.G., Rao,A. and Verdine,G.L. (1995) Curr. Biol., 5, 882-889.

25 Min,C., Cushing,T.D. and Verdine,G.L. (1995) J. Am. Chem. Soc., 118, 6116-6120.

26 Allerson,C.R. and Verdine,G.L. (1995) Chem. Biol., 2, 667-682.

---

Return

^* To whom correspondence should be addressed
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				E. Boopathi, N. Lenka, S. K. Prabu, J.-K. Fang, F. Wilkinson, M. Atchison, A. Giallongo, and N. G. Avadhani Regulation of Murine Cytochrome c Oxidase Vb Gene Expression during Myogenesis: YY-1 AND HETEROGENEOUS NUCLEAR RIBONUCLEOPROTEIN D-LIKE PROTEIN (JKTBP1) RECIPROCALLY REGULATE TRANSCRIPTION ACTIVITY BY PHYSICAL INTERACTION WITH THE BERF-1/ZBP-89 FACTOR J. Biol. Chem., August 20, 2004; 279(34): 35242 - 35254. [Abstract] [Full Text] [PDF]

This Article Abstract Print PDF (125K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (20) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Min, C Articles by Verdine, G. Search for Related Content PubMed PubMed Citation Articles by Min, C Articles by Verdine, G. Social Bookmarking          What's this?