Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (216K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Broitman, S. L.
Right arrow Articles by Fresco, J. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Broitman, S. L.
Right arrow Articles by Fresco, J. R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Nucleic Acids Research, 2003, Vol. 31, No. 16 4682-4688
© 2003 Oxford University Press

Repairing the Sickle Cell mutation. III. Effect of irradiation wavelength on the specificity and type of photoproduct formed by a 3'-terminal psoralen on a third strand directed to the mutant base pair

Steven L. Broitman, Olga Amosova and Jacques R. Fresco*

Department of Molecular Biology, Princeton University, Princeton, NJ 08544, USA

*To whom correspondence should be addressed. Tel: +1 609 258 3927; Fax: +1 609 258 2759; Email: jrfresco{at}princeton.edu
Current address:
Steven L. Broitman, Department of Biology, West Chester University, West Chester, PA, USA

Received June 10, 2003; Accepted June 17, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Using a psoralen delivery system mediated by a DNA third strand that binds selectively to linear target duplexes immediately downstream from the Sickle Cell ß-globin gene mutation and the comparable wild-type ß-globin gene sequence, the kinetics of formation and yield of psoralen monoadducts and crosslinks with pyrimidine residues at and near the mutant base pair site and its wild-type counterpart were determined. By exploiting irradiation specificities at 300, 365 and 419 nm, it was possible to evaluate the orientation equilibrium of 3'-linked intercalated psoralen and to develop conditions that lead to preferential formation of each type of photoproduct in both the mutant and wild-type sequences. This makes possible the preparation of each type of photoproduct for use as a substrate for DNA repair. In this way, the base pair change(s) that each generates can be established.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Psoralen (Fig. 1a) is a tricyclic intercalator of nucleic acid duplexes, that upon irradiation can photoreact with adjacent pyrimidine residues via the double bonds present in its external pyrone and/or furan rings (1). When linked to the 5' terminus of a third strand of a nucleic acid triplex (2–4), the psoralen moiety preferentially intercalates at the duplex–triplex junction, so that its furan ring (5-membered) is situated on the homopurine strand side of the target duplex sequence (5) (Fig. 1b). This has the consequence of favoring furan ring monoadduct formation if there is a pyrimidine residue on that strand in an adjacent Watson–Crick base pair. Whereas 5'-linked psoralen produces a very narrow spectrum of photoproducts, even when there are other pyrimidine residues within reach, 3'-linked psoralen, used of necessity in the present study, results in a much wider distribution of intercalation sites (6,7; this study). Hence, the importance of optimizing the specificity of the psoralen photoreaction.



View larger version (25K):
[in this window]
[in a new window]
  		Figure 1. (a) Hydroxymethylpsoralen (HMP) linked to the 3' terminus of an oligonucleotide via (CH2)4. (b) Binding of the same third strand to the linear target duplex sequence segment of the ß-globin gene with the Sickle Cell mutation (mutant base pair in bold) and (c) with the wild-type sequence.

 
While both the pyrone and furan rings of psoralen are activated by UV irradiation, only the furan ring can be significantly activated by visible light (2,8). When such activation is followed by UVA irradiation, monoadducts can be cross-linked to adjacent pyrimidine residues in the complementary strand. Irradiation of psoralen crosslinks with 300 nm light reverses the pyrone-side linkages, thereby regenerating furan-side monoadducts (9).

These features of wavelength selectivity of photoactivation, as well as the demonstrated mutagenicity of psoralen adducts during the course of DNA bypass replication and/or repair (2,10,11), along with the enormous specificity (12) of a third strand-mediated delivery system for any tethered reagent (13) have led us to explore psoralen covalently linked to a third strand as the basis for developing a generic method for repair of genes with point substitution mutations. Towards this end, we selected as the target for gene repair the A·T->T·A transversion in the human ß-globin gene that is responsible for Sickle Cell Anemia. Moreover, having designed a third strand with the apparent desired site specificity on the gene sequence (6), we have focused on means of achieving unique covalent linkage of the tethered psoralen moiety to the duplex target.

Whereas in the accompanying paper, we attempt to limit psoralen linkage to the desired mutant base pair by varying the length of the tethering linker [Amosova et al. (7)], here we exploit additionally the wavelength dependence of photoadduct formation to limit the mutation site, and particularly to develop conditions to maximize either psoralen monoadducts or psoralen crosslinks at that site. For the present experiments, we employ as the target a linear duplex containing a segment of the ß-globin gene sequence with the Sickle Cell transversion mutation base pair (Fig. 1b). Previously, a correlation of the type of psoralen photoproduct with the several ensuing base pair changes observed was not established with any certainty (11,14,15). Yet, such a correlation is essential for our effort to develop a gene repair method to reverse the Sickle Cell mutation. We also report here the use of the same third strand to form a psoralen monoadduct with the wild-type base pair at the Sickle Cell mutation site (Fig. 1c), in the hope of thereby creating the Sickle Cell mutation in wild-type cells. We additionally provide more general information regarding multiple photoproduct formation by 3'-linked psoralen, its preferential mode of orientation upon intercalation and its relative reactivity with T versus C residues.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Deoxyoligonucleotides
These were synthesized, purified and 5'-32P-end labeled as described in the accompanying paper (7). The HMT-psoralen ring was attached via a four-atom linker, resulting in its being tethered by five atoms to the 3' end of the third strand oligomer PsT-4 (Fig. 1a).

Duplexes and triplexes
Duplexes and triplexes were formed as in Amosova et al. (7), but with duplex concentration at 0.1 µM and the third strand in 10-fold excess.

Irradiation
Most UV irradiation was performed as described previously (6), using a Blak-Ray B100 AP UV lamp (UVP Inc., Upland, CA) with peak output at 365 nm and a half bandpass of <10 nm. Samples (typically in 20 µl) were placed on parafilm in an open Petri dish on ice, and rotated under the lamp at a distance of 10 cm. The entire operation was performed in a 4°C room. Measurements with an IL 1400A calibrated radiometer (International Light, Newburyport, MA) indicated that typical energy levels were ~5 mW/ml, corresponding to power levels of ~3 J over 10 min.

Other UV irradiation was performed using a Southern New England Ultraviolet lamp (Branford, CT) with peak output at 300 nm and a half bandpass of ±15 nm.

Visible irradiation was performed using a Southern New England fluorescent source with peak output at 419 nm and a half bandpass of ±17 nm. The lamp was fit in a standard fluorescent fixture suspended ~1 cm above the sample. Evaporation was minimized over the longer exposure to visible light by maintaining samples on parafilm in plastic-covered Petri dishes (which were determined to be transparent to visible light), irradiated on ice in a 4°C room, while the samples were rotated under the light source. Radiometer measurements indicated that energy levels were maintained at ~1 mW/ml, corresponding to power levels of ~3.6 J/h.

PAGE analysis
Denaturing PAGE analysis and phosphorimager quantitation were performed as in Amosova et al. (7).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSION
 REFERENCES
 
Experimental system
Figure 1b shows the linear target duplex with the Sickle Cell transversion mutation and the third strand designed to form a triplex along this gene segment in order to position its 3'-terminal photoreactive psoralen moiety adjacent to the mutant (T11) residue on the coding strand (D-1). The rationale for this triplex-forming strategy has been described previously (6). Principally, it involves a pyrimidine-rich segment of the third strand forming a triplex with the purine-rich segment of the non-coding (D-2) strand in the parallel orientation, after which the third strand crosses over (16) and becomes a G/T segment that binds to the coding strand in the antiparallel orientation. As before (6,7), to improve triplex stability, 5-methyl C is used in the third strand instead of C to bind to target G residues in the non-coding strand, while 5-propynyl U is positioned opposite C·G inverted base pairs in the same part of the target. Finally, a six-residue duplex-forming ‘hook’ linked by four T residues to the 5' end of the third strand is utilized to improve the binding affinity of the chimeric third strand to the linear target duplex with a ‘sticky end’. Such a third strand binds comparably to the same target incorporated into a supercoiled plasmid (6,7) (that obviously has no sticky end), apparently because superhelical stress facilitates ‘breathing’ of the double helix, thereby enabling the ‘hook’ to bind by D-loop formation. The linker length was selected based upon the findings described in the accompanying paper (7).

Photoaddition of third strand to the Sickle Cell target
Figure 2 shows a denaturing PAGE analysis of triplex irradiated with visible (419 nm) or UVA (365 nm) light. For these experiments, either the coding (D-1) or the non-coding (D-2) strand of the target duplex was end labeled with 32P. As previously observed (6), UVA irradiation for 10 min converts a substantial fraction of D-1 strands to photoproducts, both monoadducts [several monoadducts, in contrast to what was observed for 5'-linked psoralen (3–5)] and interstrand crosslinks, as well as a significant fraction (~16%) of D-2 strands to multiple photoproducts. In contrast, irradiation at 419 nm for 4 h produces monoadducts almost exclusively; a substantial amount on the coding strand (23%), but almost none along the non-coding strand. In general, the reproducibility from experiment to experiment was very good and, as can be seen from the crosslink values measured on the complementary strands, even the internal consistency of low values was reasonable. These results show that while UVA irradiation produces much more total photoproduct, visible light is much more specific in targeting the coding strand.



View larger version (37K):
[in this window]
[in a new window]
  		Figure 2. Differences between photoproducts formed after irradiation at 365 nm (10 min) and at 419 nm (4 h). (a and b) Denaturing PAGE of photoproducts. (c) Percentage conversion of Sickle Cell duplex target-coding (D-1) and non-coding (D-2) strands to monoadducts (MA) and crosslinks (XL) by psoralen-linked third strand.

 
Kinetics of photoproduct formation with visible light
To assess the kinetics of photoproduct formation with visible light, a ‘saturation experiment’ was performed at 419 nm (Fig. 3). Quantitation of the photoproducts, determined after PAGE analysis, indicated that saturation occurs at ~4 h of irradiation, with 26% of the coding strand converted to monoadductss, barely detectable levels (<1%) of crosslinks, and only 3.7% of the non-coding strand converted to photoproducts (3% monoadducts and 0.7% crosslinks). Thus, significant specificity of monoadduct formation, exclusively along the coding strand, can be achieved even after 4 h of irradiation at 419 nm, which selectively activates the furan side of psoralen (1). Hence, the furan side of psoralen must preferentially intercalate adjacent to the all-purine strand (the coding strand of the ß-globin target duplex), and its pyrone side adjacent to the all-pyrimidine strand, similar to what was observed for 5'-linked psoralen (2–5).



View larger version (89K):
[in this window]
[in a new window]
  		Figure 3. Denaturing PAGE analysis of the kinetics of photoproduct formation after irradiation at 419 nm.

 
Sequential multiple wavelength irradiation
UVA-induced crosslinks have been shown to occur via a sequential process of furan ring activation to form monoadducts, followed by pyrone activation to complete them (1,17). Thus, UVA irradiation principally forms monoadducts at early time points, which are gradually converted into interstrand crosslinks. We attempted to exploit this knowledge to better control the nature of the photoproducts, and therefore the specificity of photoproduct formation. To that end, we explored a sequential dual-wavelength irradiation protocol, first using lengthy 419 nm irradiation to extensively form monoadducts along the coding strand, and then short exposure at 365 nm to efficiently convert those monoadducts to crosslinks.

Results obtained with this approach are shown in Figure 4. First, 3 h of visible irradiation converted 23% of the coding strand into monoadducts only, with no detectable (<1%) cross-links or non-coding strand monoadducts (Fig. 3). This was followed by irradiation at 365 nm for times up to 30 min, which resulted in a maximum of 45% of the coding strand crosslinked in addition to 7% of D-1 monoadducts and 11% of D-2 monoadducts (not shown). By comparison, direct irradiation at 365 nm for 30 min (lane 1) led to 37% crosslinks, 12% D-1 monoadducts, and again 11% D-2 monoadducts (not shown). Thus, it is apparent that sequential irradiation significantly improves the ratio of crosslinks to coding strand monoadducts, i.e. 6.5:1 versus 3:1. Moreover, at shorter times of sequential irradiation at 365 nm, the differential is larger (not shown). It is emphasized that all the foregoing percentages represent yields of photoproducts, i.e. percentage conversion of duplex into particular photoproducts, and not their proportions of total photoproducts formed.



View larger version (59K):
[in this window]
[in a new window]
  		Figure 4. Denaturing PAGE analysis of coding strand photoproduct formation following sequential irradiation of the Sickle Cell target with bound psoralen-containing third strand, first at 419 nm for 3 h, and then at 365 nm for 30 min.

 
Crosslink and pyrone monoadduct photoreversal
Since a goal of this investigation was to develop conditions that maximize selective formation of each type of photoproduct, experiments were performed to exploit crosslink photoreversal by irradiation at 300 nm. Such irradiation reverses pyrone-side psoralen photoproducts (9). Mono adducts were first formed with 3 h of irradiation at 419 nm (Fig. 5, lane 3). As shown above, this produces coding strand monoadducts nearly exclusively, with no crosslinks, and almost no monoadducts on the non-coding strand. This was followed by irradiation at 365 nm for 30 min (lane 5), which generated yields of 45% crosslinks, 7% monoadducts on the coding strand, and 11% monoadducts on the non-coding strand (not shown). Subsequent irradiation of this mix of photoproducts at 300 nm for 30 min (lane 7) resulted in 40% coding strand monoadducts, only 6% non-coding strand monoadducts (not shown) and 4% crosslinks. These results show that such an irradiation protocol generates a much higher level of desired monoadduct than could be achieved by visible irradiation alone.



View larger version (70K):
[in this window]
[in a new window]
  		Figure 5. (a) Denaturing PAGE analysis of kinetics of cross-link photoreversal on coding strand D-1 after irradiation at 300 nm. (b) High-resolution denaturing PAGE showing an additional monoadduct band that appears after 365 nm irradiation and disappears after 300 nm irradiation. Lane 8, 419 nm, 3 h irradiation; lane 9, 419 nm, 3 h, followed by 365 nm, 30 min; lane 10, 419 nm, 3 h, followed by 365 nm, 30 min, followed by 300 nm, 15 min.

 
Major and minor orientations of intercalated psoralen
These data also provide insight into the relative proportions of psoralen orientation when it intercalates. In the linear duplex D-1·D-2, the main residues that are psoralen modified were identified earlier (6), and confirmed in the accompanying paper (7). In summary, there are always two monoadducts formed on strand D-1, a major one at the Sickle Cell mutation site T11, and a very minor one at T9, as well as two on strand D-2 at C24 and C23 (Figs 1b and 2). Typical yields of these photoproducts and their fraction of total photoproduct formation are summarized in Table 1. High-resolution denaturing gels sometimes also reveal an additional minor monoadduct band formed with D-1 after 365 nm irradiation, that has a mobility intermediate between T11 and T9; this band disappears after 300 nm reversal (Figs 2 and 5b). Similarly, a minor cross-link band appears and disappears (Fig. 5a). We explain these minor monoadduct and crosslink bands as the result of the small population of pyrone-side monoadducts formed with the coding strand by a low level of alternatively oriented intercalated psoralen, such that its furan ring is bonded to the all-pyrimidine D-2 strand of the duplex. This interpretation is consistent with finite, i.e. non-zero, monoadduct formation on the non-coding strand after irradiation at 419 nm.


View this table:
[in this window]
[in a new window]
  		Table 1. Photoproduct formation at various target residues after 10 min of UVA irradiation
 
Photoaddition of third strand-linked psoralen to the wild-type ß-globin gene target
With the hope of exploiting psoralen photochemistry to create the Sickle Cell mutation in wild-type human cell lines, triplexes were formed between the third strand developed for Sickle Cell gene repair and a duplex target containing the wild-type ß-globin gene sequence. As seen in Figure 1c, this sequence only differs from the D-1·D-2 duplex by having the wild-type A11·T25 base pair in place of the T11·A25 transversion base pair. In that case, if the equilibrium geometry of psoralen intercalation remains the same, i.e. with the furan ring intercalated predominantly on the coding strand side, then the coding strand (H-1) should give only a minor photoaddition product at T9 after irradiation with visible or UVA light since it lacks T11. On the other hand, since the non-coding strand (H-2) has a T at position 25, that residue should react efficiently with the pyrone ring of psoralen upon UVA irradiation, but only poorly when irradiated with visible light.

These expectations were fully met, as shown by the PAGE analysis in Figure 6. Thus, both visible and UVA irradiation generated only a very small amount of H-1 (coding strand) monoadduct, which was identified by primer extension arrest to be at T9 (7). The proportion of that monoadduct, 4% with UVA and 0.3% with 419 nm irradiation, correlates well with the monoadduct formation at T9 on the Sickle Cell D-1 strand (7). The same amount of UVA irradiation gave a much higher yield of photoproducts along the H-2 (non-coding) strand (51%) than along the Sickle Cell D-2 strand (24.5%), consistent with the added presence of T25 in the H-2 strand. On the other hand, visible irradiation gave only a very slight increase in photoproduct formation along the wild-type H-2 strand (3.4 versus 2.7% along D-2 after 90 min of irradiation), which is again consistent with the predominant orientation of the furan ring of psoralen on the coding strand side. Hence, these data confirm that the pyrone ring of the intercalated psoralen is predominantly positioned adjacent to the non-coding strand of the duplex even in the wild-type sequence, which lacks adequate psoralen adduct-forming pyrimidines on the coding strand. In fact, UVA irradiation for 10 min produced three major H-2 monoadducts due to photoactivation of the pyrone ring, the most prevalent being at the wild-type T25 residue. The other two were at C24 and C23, which also form in the Sickle Cell D-2 strand.



View larger version (79K):
[in this window]
[in a new window]
  		Figure 6. Denaturing PAGE analysis affording a comparison of photoproducts formed by wild-type (WT) and Sickle Cell (SC) ß-globin target sequences. Strands D-1 and DH-1 are listed as 1; strands D-2 and DH-2 as 2.

 
Neither visible nor UVA irradiation generated any interstrand cross-links with the wild-type target (Fig. 6). This is consistent with the absence of any nearby pyrimidine residues in the H-1 coding strand (T9 is located too far away). These results suggest that the same third strand designed to repair the Sickle Cell mutation can probably also serve to create it starting with the wild-type ß-globin gene sequence.

Previously, primer extension arrest was used to identify the monoadduct at T11 from strand D-1 (6). Comparable data for the other monoadducts formed by the Sickle Cell and wild-type triplexes are presented in the accompanying report (7). As expected, only the T9 monoadduct is formed on the wild-type coding strand. Three major monoadduct bands are formed with non-coding strand DH-2 (Fig. 6), each differing in mobility. Monoadduct bands MA-1 and MA-2 in Figure 6, which form as well with the Sickle Cell D-2 strand, are similarly identified as monoadducts of residues C24 and C23. MA-3 is the monoadduct band that only forms with the wild-type duplex and so must derive from residue T25, as was confirmed by primer extension arrest. The same is true for MA-4, a very minor, slightly faster moving fourth band, that was not obtained in an amount sufficient for primer extension analysis. Since this monoadduct is derived only from the wild-type target and its mobility is faster than that of the major T25 monoadduct band, it can be presumed to be a minor furan-side monoadduct of residue T25, and thereby serve as another demonstration of alternative psoralen intercalation. With this assumption, all monoadducts formed with both the Sickle Cell and wild-type target sequences are identified. Taken together, these results provide a consistent picture of third strand-linked psoralen intercalation equilibrium and photoinduction.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSION
 REFERENCES
 
In this investigation, visible and/or UVA irradiation-induced addition of psoralen linked to the 3' end of a third strand bound to either Sickle Cell or wild-type target DNA was explored. The principal aims of the work were to develop irradiation protocols that maximize the specificity of either monoadduct or crosslink photoproduct formation with residue T11 on the coding strand of the Sickle Cell ß-globin gene, and with residue T25 on the non-coding strand of the wild-type allele. These goals have been met.

More generally, the results enhance our understanding of third strand-mediated delivery of reactive agents to specific sites on a DNA duplex. Thus, insights were obtained regarding the orientation of the intercalating psoralen linked to the 3' end of the third strand. In addition, the results indicate that 3'-terminal psoralen on the third strand forms photoproducts with both T and C residues with similar efficiency, in contrast to free psoralen, which strongly prefers T residues (18).

Sites of psoralen photoaddition
The N-4 psoralen linker to the third strand used in this work allows the psoralen moiety to intercalate readily within ±1 bp of the duplex–triplex boundary. This is apparent from the identity of the photoproducts formed with the wild-type target and their distribution [Table 1 and Amosova et al. (7)]. Up to 50% of residues T11 and C24 of the D-1·D-2 Sickle Cell target duplex can be converted to monoadduct or crosslink (Fig. 4), while wild-type duplexes form only monoadducts with up to 30% of T25 and up to 12% of C24 at the duplex–triplex boundary, and with rather more difficulty with residues T9 and C23 on either side of the boundary. A maximum of only 4% of either target duplex forms monoadducts (but no crosslinks) at T9 under UVA irradiation; C23 is modified in 5% of the Sickle Cell target duplexes and in almost 10% of the wild-type ones (Table 1) (6,7).

The data in Table 1 can be used to calculate the amount of photoproduct formed within the two base pairs immediately adjacent to the triplex–duplex junction, i.e., T11·A25 + G12·C24 for D1·D2 and A11·T25 + G12·C24 for H-1·H-2. These photoproducts include monoadducts or crosslinks in the case of D-1·D-2 or just monoadduct in the case of H-1·H-2, for which the absence of nearby pyrimidines on strand H-1 prohibits the possibility of forming a crosslink. To compare the extent of modification, we sum the yields of crosslinks and monoadducts (Table 1, columns 1 and 3) involving C24 and T11 for D-1·D-2, and T25 and C24 monoadducts for H-1·H-2. It is noteworthy that the total photoproducts formed by the base pairs adjacent to the triplex–duplex junction are the same within experimental error for both mutant and wild-type targets, as can be calculated from the data in Table 1, i.e. C24 + T25 + MA-4 for H-1·H-2, and C24 + T11 furan + T11 pyrone + T11 XL for D-1·D-2.

Minimal pyrone-side monoadduct is formed by visible irradiation
Furan-side photoproducts form upon both visible (419 nm) and UVA (365 nm) irradiation, although with somewhat better yield and higher efficiency with UVA. In contrast, the yield of pyrone-side photoproducts is very much greater with UVA than with visible irradiation (18). Since crosslinks inevitably combine both types of photoproduct, they provide a convenient measure of pyrone-side photoproduct formation. Up to 40% of the Sickle Cell target can be converted to cross-links by short UVA irradiation. On the other hand, only 1% of target duplexes form crosslinks after several hours of visible irradiation. This information enables us to estimate that pyrone-side photoproduct formation with visible light, while finite, is at least 40 times less efficient than with UVA irradiation, consistent with what was previously observed for 5'-linked psoralen (2).

Preferred orientation and equilibrium of intercalated 3'-linked psoralen
Previously, it had been found that psoralen linked to the 5' terminus of third strands almost exclusively forms furan-side photoproducts with the all-purine strand of the target duplex whenever there are pyrimidine residues within reach on that strand (2–4). In the present work, the 3' terminus of the third strand had to be used to deliver psoralen to the Sickle Cell mutant T11 residue. This provided an opportunity to evaluate the orientation of intercalated psoralen as a consequence of its altered location on the third strand.

For the all-purine segment on strand D-1 immediately downstream from the mutant T11 residue, there is only an ~2-fold difference in total photoproduct yield between 365 and 419 nm, i.e. >40% after 10 min versus 20% after 4 h, respectively (Table 1 and Fig. 3). In contrast, the difference is >10-fold, i.e. 34% versus 3% for non-coding strand D-2. Thus, the pyrone side of the third strand-linked psoralen must predominantly face the non-coding strand D-2, since, as noted, pyrone-side photoproduct formation is very much less efficient with visible light. Comparable changes in photoproducts formed occur for the non-coding strand of the wild-type duplex, showing that the equilibrium of psoralen intercalation is very similar in that case. Hence, the preferred psoralen orientation is with the furan side facing the all-purine segment of strand D-1, just as had been found for 5'-linked psoralen.

The foregoing results show that the pyrone ring of the intercalated psoralen is predominantly positioned adjacent to the non-coding strand of the duplex in both the mutant and the wild-type sequence, even though the wild-type strand lacks the principal psoralen adduct-forming pyrimidine T11 on the coding strand. To quantitate the extent of this predominance, photoreversal of pyrone photoproduct at 300 nm was exploited. The pyrone-side monoadduct with T11 on the D-1 strand (7), observed as extra monoadduct and crosslink bands formed with UVA (Fig. 5), is completely reversed with 300 nm irradiation. That pyrone photoproduct accounts for ~10% of total photoproduct formed with strand D-1. This suggests that at least 10% of the psoralen moieties intercalate with their pyrone side facing the purine strand. This estimate is consistent with the residual amount of D-2 photoproducts that remain after 300 nm photoreversal (not shown) and with a very low level of photoproduct formation on the non-coding strand by visible irradiation (Fig. 3).

Relative psoralen photoproduct formation with T and C residues
Comparison of photoproduct formation with Sickle Cell and wild-type targets also provided insight into the relative reactivities of psoralen with T and C residues. Since the data show that the equilibrium distribution of intercalating psoralen is unaffected by the difference between mutant and wild-type sequences, the total photoproduct yields on those strands and their distribution among individual photoproducts provide a basis for such an estimate (Table 1). The overall yield is 24% in strand D-2, but 51% in strand H-2 (Fig. 6). This difference is largely accounted for by the extra 24% of monoadduct formed with residue T25 of H-2. Its fraction represents 47% of the total photoproduct yield on H-2. The distribution of individual photoproducts for strand D-2 indicates that most of the psoralen is intercalated adjacent to C24, which accounts for ~80% of the total photoproducts on that strand when monoadducts and crosslinks formed at that site are totaled.

Base pair A11·T25 in the wild-type duplex and its counterpart T11·A25 in the Sickle Cell duplex are positioned at the duplex–triplex junction, which, as shown above, is a preferred 3'-linked psoralen intercalation site. Whereas 41% of T11 on D-1 forms photoproducts, only 24% of T25 on H-2 does so. This difference is probably due to the interaction of residue T25 with the pyrone double bond of psoralen rather than with that on the furan side, which interacts with residue T11. Also, given that in the wild-type sequence the difference in photoreactivity between T25 at the triplex–duplex junction and C24 is only 2-fold (Table 1), the intrinsic reactivity of T and C residues cannot be too dissimilar, in fact, much less so than the >20-fold difference that was observed for free psoralen with these residues (19).


   CONCLUSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
REFERENCES
 
Having shown that the T11 monoadduct is almost exclusively formed with visible irradiation, which can be subsequently converted almost exclusively to crosslinks with UVA irradiation, we should be able to directly correlate the nature of the T11 photoproduct in the mutant gene, the T25 photoproduct in the wild-type allele and any of the other photoproducts with the type of base pair change that they each induce when confronted by the cellular DNA repair systems.


   ACKNOWLEDGEMENTS
 
We are grateful to Francis Gasparro for helpful discussions, to Yakov Varganov for technical assistance, and to Dmitry Klimov for assistance with figure preparation. This work was supported by NIH grant 1R01 HL63888 from the National Heart, Lung, and Blood Institute.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 

  1. Cimino,G.D., Gamper,H.B., Isaacs,S.T. and Hearst,J.E. (1985) Psoralens as photoactive probes of nucleic acid structure and function: organic chemistry, photochemistry and biochemistry. Annu. Rev. Biochem., 54, 1151–1193.[CrossRef][ISI][Medline]

  2. Gasparro,F.P., Havre,P.A., Olach,G.A., Gunther,E.J. and Glazer,P.M. (1994) Site-specific targeting of psoralen photoadducts with a triple helix-forming oligonucleotide: characterization of psoralen monoadduct and crosslink formation. Nucleic Acids Res., 22, 2845–2852.[Abstract/Free Full Text]

  3. Bates,P.J., Macaulay,V.M., McLean,N.J., Jenkins,T.C., Reszka,A.P., Laughton,C.A. and Niedle,S. (1995) Characteristics of triplex-directed photoadduct formation by psoralen-linked oligodeoxynucleotides. Nucleic Acids Res., 23, 4283–4289.[Abstract/Free Full Text]

  4. Oh,D.H. and Hannawalt,P.C. (2000) Binding and photoreactivity of psoralen linked to triple helix-forming oligonucleotides. Photochem. Photobiol., 72, 298–307.[CrossRef][ISI][Medline]

  5. Takasugi,M., Guendouz,A., Chassignol,M., Decouit,J.L., Lhomme,J., Thuong,N.T. and Hélène,C. (1991) Sequence-specific photo-induced cross-linking of the two strands of double-helical DNA by a psoralen covalently linked to a triple helix-forming oligonucleotide. Proc. Natl Acad. Sci. USA, 88, 5602–5606.[Abstract/Free Full Text]

  6. Broitman,S., Amosova,O., Dolinnaya,N.G. and Fresco,J.R. (1999) Repairing the Sickle Cell mutation. I. Specific covalent binding of a photoreactive third strand to the mutated base pair. J. Biol. Chem., 274, 21763–21768.[Abstract/Free Full Text]

  7. Amosova,O., Broitman,S.L. and Fresco,J.R. (2003) Repairing the Sickle Cell mutation. II. Effect of psoralen linker length on specificity of formation and yield of third strand-directed photoproducts with the mutant target sequence. Nucleic Acids Res., 31, 4673–4681.[Abstract/Free Full Text]

  8. Kanne,D., Straub,K., Hearst,J.E. and Rapoport,H. (1982) Isolation and characterization of pyrimidine–psoralen–pyrimidine photodiadducts from DNA. J. Am. Chem. Soc., 104, 6754–6764.[CrossRef]

  9. Cimino,G.D., Shi,Y. and Hearst,J.E. (1986) Wavelength dependence for the photoreversal of a psoralen–DNA cross-link. Biochemistry, 25, 3013–3020.[CrossRef][Medline]

  10. Havre,P.A., Gunther,E.J., Gasparro,F.P. and Glazer,P.M. (1993) Targeted mutagenesis of DNA using triple helix-forming oligonucleotides linked to psoralen. Proc. Natl Acad. Sci. USA, 90, 7879–7883.[Abstract/Free Full Text]

  11. Sandor,Z. and Bredberg,A. (1994) Repair of triple helix directed psoralen adducts in human cells. Nucleic Acids Res., 22, 2051–2056.[Abstract/Free Full Text]

  12. Fossella,J.A., Kim,Y.J., Shih,H., Richards,E.G. and Fresco,J.R. (1993) Relative specificities in binding of Watson–Crick base pairs by third strand residues in a DNA pyrimidine triplex motif. Nucleic Acids Res., 21, 4511–4515.[Abstract/Free Full Text]

  13. Strobel,S.A., Doucette-Stamm,L.A., Riba,L., Houseman,D.E. and Dervan,P.B. (1991) Site-specific cleavage of human chromosome 4 mediated by triple-helix formation. Science, 254, 1639–1642.[Abstract/Free Full Text]

  14. Wang,G. and Glazer,P. (1995) Altered repair of targeted psoralen photoadducts in the context of an oligonucleotide-mediated triple helix. J. Biol. Chem., 270, 22595–22601.[Abstract/Free Full Text]

  15. Barre, F-X., Asseline,U. and Harel-Bellan,A. (1999) Asymmetric recognition of psoralen interstrand crosslinks by the nucleotide excision repair and the error-prone repair pathways. J. Mol. Biol., 286, 1379–1387.[CrossRef][ISI][Medline]

  16. Jayasena,S. and Johnston,B.H. (1992) Intramolecular triple-helix formation at (PunPyn).(PunPyn) tracts: recognition of alternate strands via Pu·PuPy and Py·PuPy base triplets. Biochemistry, 31, 320–327.[CrossRef][Medline]

  17. Shi,Y. and Hearst,J.E. (1987) Wavelength dependence for the photoreactions of DNA–psoralen monoadducts. 2. Photo-cross-linking of monoadducts. Biochemistry, 26, 3792–3798.[CrossRef][Medline]

  18. Hearst,J.E., Isaacs,S.T., Kanne,D., Rapoport,H. and Straub,K. (1984) The reaction of the psoralens with deoxyribonucleic acid. Q. Rev. Biophys., 17, 1–44.[ISI][Medline]

  19. Kanne,D., Straub,K., Rapoport,H. and Hearst,J.E. (1982) Psoralen–deoxyribonucleic acid photoreaction. Characterization of the monoaddition products from 8-methoxypsoralen and 4,5',8-trimethylpsoralen. Biochemistry, 21, 861–871.[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Proc. Natl. Acad. Sci. USAHome page
O. Amosova, R. Coulter, and J. R. Fresco
Self-catalyzed site-specific depurination of guanine residues within gene sequences
PNAS, March 21, 2006; 103(12): 4392 - 4397.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
O. Amosova, S. L. Broitman, and J. R. Fresco
Repairing the Sickle Cell mutation. II. Effect of psoralen linker length on specificity of formation and yield of third strand-directed photoproducts with the mutant target sequence
Nucleic Acids Res., August 15, 2003; 31(16): 4673 - 4681.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (216K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Broitman, S. L.
Right arrow Articles by Fresco, J. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Broitman, S. L.
Right arrow Articles by Fresco, J. R.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?