|Home | About | Journals | Submit | Contact Us | Français|
The lamin B2 locus is the only mammalian origin whose replication initiation points (RIPs) have been mapped. Although this paper was published 8 years ago, no further mammalian RIP-mapping studies have been reported, largely due to technical difficulties of ligation-mediated (LM)-PCR used by the authors. Here, we report the development of a simple, one-way PCR-based protocol that allows one to accurately determine RIPs at mammalian origins. The procedure can be completed within 48 h from the time of cell lysis in the agarose gel. Nascent DNA is then isolated from the same gel after DNA is separated by alkaline gel electrophoresis. Subsequently, RIPs are determined by one-way PCR-based primer extension using labeled primers. Using this protocol, we have successfully mapped RIPs in the human DBF4 locus. As one-way PCR is routinely used by many scientists, this protocol will provide a powerful new tool for studying DNA replication in many organisms including mammalian cells.
Faithful replication of the entire genome once per cell cycle is critical to maintain genetic integrity. Because the origin of DNA replication (ori) plays a key role in this process, understanding the initiation mechanism is essential, for which mapping the replication initiation point (RIP) is the first step. Several RIPs have been mapped in unicellular organisms such as bacteria and budding yeast. However, RIP mapping in mammalian cells turned out to be very challenging, mainly due to technical difficulties. The lamin B2 locus is thus far the only mammalian ori whose RIPs have been mapped1. The authors determined RIPs at this locus using the technically demanding ligation-mediated (LM) PCR1. Although this paper was reported in 2000, no one has reported another mammalian RIP using LM-PCR or any other methods, underscoring the need to develop more effective and user friendly RIP-mapping techniques.
Using one-way PCR-based primer extension, Gerbi and Bielinsky successfully identified the precise transition point between continuous (leading strand) and discontinuous (lagging strand) DNA synthesis at the Saccharomyces cerevisiae ARS1 ori2–4. The original method developed by these authors2 may be divided into the following six steps:
The authors rationalized that a RIP of the leading strand coincides with the 5′ end of the shortest strand detected by the primer extension because successively larger strands are presumably generated by the ligation of Okazaki fragments to the leading strand. Since it was first reported in 1997, this RIP-mapping technique has been used to determine precise RIPs at several ori loci in Schizosaccharomyces pombe5 and the fly Sciara cropophila6. RIP mapping has also been used in archaea, such as Pyrococcus abyssi7, Sulfolobus solfataricus8 and Haloarcula sp9. However, the original RIP-mapping protocol was not considered sensitive enough to map RIPs in mammalian cells. Therefore, LM-PCR was used to map RIP at the human lamin B2 origin1.
We postulated that one-way PCR-based primer extension could be directly used for mapping RIPs in mammalian oris without the need of LM-PCR, if nascent strands can be effectively harvested. We could indeed achieve this goal by bypassing several steps required for enriching short nascent DNA strands. Importantly, this stream-lined protocol did not compromise the quality and purity of the nascent DNA sample. The steps that we could omit include cesium chloride gradient centrifugation, BND chromatography, treatment with T4 polynucleotide kinase to phosphorylate DNA molecules lacking an RNA-primer and λ-exonuclease treatments. Using this new approach, we successfully mapped RIPs at the human DBF4 locus, where we have recently identified a strong ori10. This new RIP-mappingmethod combines one-way PCR-based primer extension with several previously reported techniques to effectively isolate nascent DNA fragments as described below (see Table 1 for comparison of different methods).
We validated our one-way PCR-based RIP mapping using the lamin B2 locus model (see Supplementary Fig. 2 in ref. 10; http://www.nature.com/nsmb/journal/v15/n7/suppinfo/nsmb.1439_S1.html). In this validation experiment, however, we had to omit betaine for the reason discussed in the section “The use of betaine in PCR” below.
This new one-way PCR-based protocol is certainly sensitive enough to determine mammalian RIPs, even relatively weak oris such as the one at the lamin B2 locus. Nevertheless, this mapping method may not be ideal for the detection of RIPs at extremely weak mammalian ori loci.
Finally, this new protocol can also be very useful in determining RIPs in nonmammalian organisms because it can bypass several rather tedious steps. Thus, the RIP-mapping protocol described here can greatly facilitate research in the regulation of DNA replication of many different organisms including mammalian cells.
To carry out RIP mapping at the DBF4 ori locus, we used HeLa cells as the source of nascent DNA. HeLa cells were maintained in RPMI-1640 medium containing 10% (vol/vol) fetal bovine serum (Hyclone, Logan, UT), 2.05 mM L-glutamine, 100 μg ml−1 streptomycin and 100 Uml−1 penicillin. Although we could easily detect DBF4 RIPs from exponentially growing HeLa cells, the results were much more pronounced when nascent DNA fragments were isolated from synchronous HeLa cells at 1 h in S phase10. This was achieved by releasing HeLa cells synchronized at the G1/S border by a double-thymidine treatment into complete medium for 1 h as described previously10.
One of the most serious problems of determining a RIP is that ligation of Okazaki fragments to a leading strand complicates the interpretation of the true initiation point of leading strand synthesis. In yeast cells, one strategy to circumvent this problem is to compare RIPs in wild-type and DNA ligase I-deficient mutants3. Since this method is difficult to apply in mammalian cells, we and others used emetine, which inhibits new production of lagging strands without affecting leading strand synthesis1,11. Although concern was raised previously regarding drug treatments and the generation of potential artifactual results2, we did not observe any notable problemwhenwe determined DBF4 RIPs with cells maintained in the presence and absence of emetine10. However, we cannot completely rule out the potential effects of emetine on RIP-mapping results because we have not systematically compared RIP patterns of cells grown in the absence and presence of emetine.
Several different approaches have been developed to isolate nascent DNA without contamination by randomly nicked DNA. Isolation of DNA molecules containing RNA primer has been a preferred method for RIP mapping2. However, yielding nascent strands by this approach may not be sufficiently high because many tedious steps are required. We found that the simplest and most promising approach of isolating nascent DNA strands is: (i) to lyse cells directly in the wells of alkaline agarose gels; (ii) to separate DNA by electrophoresis under denaturing conditions and (iii) to isolate short single-stranded DNA by gel extraction12. This procedure minimizes sample handling and has been previously employed by others to map oris in mammalian cells13,14. This simple method also appears to minimize DNA damage because origin enrichment was not found with nonreplicating control DNA10. If substantial damage occurred during nascent strand preparation, we would have observed the enrichment of origin sequence from the control DNA.
Another modification we made is the use of betaine in our one-way PCR. This was particularly important for the RIP mapping at the DBF4 locus, where G+C content is very high (~70%). Betaine has been previously shown to prevent the formation of secondary structures in template DNA and promote polymerization of GC-rich sequences15. Interestingly, adding betaine to the RIP-mapping reaction for the human lamin B2 ori resulted in a decrease in efficiency10, perhaps because of the high A+T content in this locus. The protocol presented here is optimal for the detection of RIPs at the human DBF4 locus where the G+C content is high. Because we had to omit betaine in the reaction for the RIP mapping at the lamin B2 ori10, a slight adjustment for mapping protocol at different oris may be necessary.
Although Vent (exo−) DNA polymerase has been the polymerase of choice for RIP mapping1–3, this enzyme was not ideal under our experimental conditions. The possible reason for this is that the activity of this polymerase may not be compatible with betaine. In our hands, IDPol polymerase was the best for one-way PCR-based primer extension.
The primers we use for RIP mapping are at least 19-bases long with a melting temperature close to≥60 °C. As ~800 bases might be the maximum length that can be efficiently amplified by IDPol polymerase in the one-way PCRs10, primer distance to the potential replication start sites should be <800 bases, preferably between 100 and 500 bases.
An important advantage of our approach is that this new protocol is designed to use nonradioactive labeling. Although radioactively labeled oligonucleotides have been the choice for primer extension reactions in previous RIP-mapping studies1–3,5,6, we used digoxigenin-labeled primers, followed by immunoblotting with antidigoxigenin antibodies. We have found that this nonradioactive-labeling method is sensitive enough to detect RIPs in mammalian cells. Therefore, the entire work described in our protocol can be completed on the laboratory bench, eliminating the need for a specially shielded working area or disposing radioactive wastes.
To detect potential artifacts resulted by the reaction mix (including primers), each one-way PCR experiment should include a no-template control. In addition, a control RIP-mapping reaction should be carried out, in parallel with themain experiment, using nonreplicating DNA isolated from cells in mitosis10 or cells treated with a high concentration of replication inhibitors such as aphidicolin (Fig. 1a). This is necessary to rule out the possibility that the products observed after primer extension are resulted from amplification of randomly nicked DNA, instead of replication intermediates (i.e., leading strands).
To prepare 100 ml solution, add 18.61 g of disodium EDTA-2H2O to ~70 ml distilled water. Adjust pH to 8.0 initially with NaOH pellets and then 10 N NaOH. Adjust the final volume to 100 ml and autoclave. The solution may be stored for several months at room temperature (~21 °C), provided that it remains free of contamination.
Dissolve 40 g of NaOH in the final volume of 100 ml distilled water. It may be stored for up to 6 months at room temperature.
To make 1 l of solution, dissolve 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4 and 2.4 g of KH2PO4 in ~800 ml in distilled water. Adjust pH to 7.4–7.5, and then the final volume to 1 l with distilled water, followed by autoclaving. The buffer may be stored for several months at room temperature.
Prepare freshly at the time of use by mixing 100 μl 10× PBS with 800 μl sterile water and 100 μl glycerol. Add 5 μl of 10% (wt/vol) bromophenol blue to serve as tracking dye.
To prepare 300 ml solution, dissolve 36.34 g Trizma base in ~250 ml distilled water. Adjust pH to 8.0, and the final volume to 300 ml, followed by autoclaving. It may be stored for several months at room temperature.
Dissolve 2 g of SDS in sterile water tomake the final volume of 20 ml solution. It may be stored at room temperature for several months.
Mix 500 μl of 1MTris-HCl, pH 8.0, 100 μl of 0.5MEDTA, pH 8.0, and 500 μl of 10% (wt/vol) SDS. Adjust the final volume to 10 ml solution with sterile water. Itmay be stored at 4 °C for up to 1 month.
To prepare 100 ml buffer, mix 76 ml distilled water with 20 ml 1 M Tris-HCl, pH 8.0, and 4 ml of 0.5 M EDTA, pH 8.0. Autoclave and store at 4 °C for up to several months.
Mix 2 g Ficoll 400, 2 ml 0.5 M EDTA, pH 8.0, 1 ml 10% (wt/vol) SDS and 250 μl 10% (wt/vol) bromophenol blue. Adjust the final volume to 10 ml solution with sterile water. Aliquot and store at −20 °C for up to several months.
Mix 20 μl of 500 μg ml−1 stock DNA ladder (New England BioLabs) with 10 μl 20× TE buffer, 20 μl 10× DNA-loading buffer and 150 μl sterile water. The diluted ladder may be stored for several months at −20 °C.
Prepare a fresh batch at the time of use by mixing 40 μl of 1-kb DNA marker solution, 19 μl sterile water, 3.3 μl 1 N NaOH (the final concentration is 50mM) and 2.7 μl 0.5MEDTA, pH 8.0 (the final concentration is 20 mM).
Prepare a fresh batch at the time of use by mixing 993 ml of distilled water with 5 ml of 10 N NaOH and 2 ml 0.5 M EDTA, pH 8.0.
Dissolve 121 g of Trizma base, 9.31 g disodium EDTA-2H2O and 28.6 ml glacial acetic acid, and adjust the final volume to 500 ml with sterile distilled water. The pH of the solution should normally be between 8.4 and 8.5 without further adjustment. It may be stored at room temperature for several months.
To prepare 10 ml solution, mix 9.8 ml deionized formamide, 200 μl 0.5M EDTA, pH 8.0, 0.01 g xylene cyanol and 0.01 g bromophenol blue. Store at 4 °C in 1 ml aliquots up to several months.
Dissolve 27 g Trizma base and 13.75 g boric acid in ~400 ml sterile distilled water. Add 10ml of 0.5MEDTA pH 8.0. Adjust the final volume to 500 ml with sterile distilled water. The pH should be 8.2–8.5 without further adjustment. Store at room temperature for up to one month. Discard if a precipitate is visible.
Dissolve 95 g acrylamide and 5 g bisacrylamide, and adjust the final volume to 250 ml with distilled water. Filter and store at 4 °C protected from light for up to several months.
Dissolve 0.5 g of ammonium persulfate in sterile water, and then adjust the final volume to 5 ml. The solution may be stored at 4 °C for several weeks. However, freshly prepared solution should be used to ensure effective gel polymerization.
Mix 7.5 ml of 40% acrylamide solution, 27.72 g urea and 6 ml 5× TBE, and adjust the final volume to 60 ml. Heat slightly and stir until all the urea has dissolved (make sure the solution cools down to room temperature before proceeding to the next step). Add 600 μl of 10% (wt/vol) ammonium persulfate and 36 μl TEMED, mix gently and promptly cast the gel. Allow the gel to solidify overnight at room temperature before running samples.
Mix 2 μl of DIG-labeled DNA molecular weight marker stock (Roche), 14 μl sterile water and 14 μl sequencing gel-loading buffer. Store at−20 °C for up to several weeks. Heat denature immediately before use.
Dissolve 11.61 g maleic acid and 8.77 g NaCl in ~800 ml distilled water. Adjust pH to 7.5, initially with NaOH pellets, and then 10 N NaOH. Adjust the final volume to 1 l with distilled water, and then autoclave. The buffer may be stored at room temperature for several months.
Dissolve 5 g of blocking reagent in 40ml ofmaleic acid buffer by gently heating it (do not allow the solution to boil). Once completely dissolved, adjust the final volume to 50ml withmaleic acid buffer. Store at 4 °C for up to amonth and dilute 1/10 in maleic acid buffer at the time of use.
0.03% (vol/vol) Tween 20 in maleic acid buffer. Prepare a fresh batch at the time of use.
Dissolve 3.05 g Trizma base and 1.46 g NaCl in~200 ml distilled water. Adjust pH to 9.5, and the final volume to 250 ml, and autoclave. Store at room temperature for up to several months.
Figure 1a shows a typical result of RIP mapping at the DBF4 ori using the protocol presented here. The size of the top band shown in Figure 1a (lane ‘−’) is estimated to be 504 nucleotides. Since the start site of primer pC is −131 (i.e., 131 nucleotide upstream from the ATG translation start site), the RIP of this leading strand was determined to be the +373 position (Fig. 1b). This position was later confirmed by sequencing gel electrophoresis10. As expected, no bands were observed in the control sample lacking template DNA (Fig. 1a, lane ‘N’). Similarly, primer extension signal was greatly reduced when the nascent template DNA was isolated from aphidicolin-treated cells (Fig. 1a, lane ‘+’).
By carrying out detailed RIP mapping with many different primers, we discovered that DBF4 ori contains two replication initiation zones that are regulated by an asymmetric bidirectional replication mechanism10. This finding was possible because we could construct a fine RIP map at this locus using the one-way PCR-based primer extension method described here. It should be noted that a nascent strand abundance assay alone, which is commonly used for mapping replication ori, could not detect two replication initiation zones10. Therefore, the protocol presented here is a very powerful tool in unraveling the regulation of replication oris.
Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/