Comparison of Deprotection Methods for the Phthalimidyl Amino CPGs

By Richard Hogrefe, Ph.D., Paul Imperial, and Alexandre Lebedev, Ph.D.; TriLink BioTechnologies.

One of the concerns regarding the use of the phthalimidyl 3'-amino CPG is the efficiency of the cleavage reaction from support by ammonium hydroxide. Although the accompanying article on these molecules offer data using deprotection conditions of concentration ammonium hydroxide at 55° C for 15 hours, we do not feel that we adequately addressed the question regarding efficiency of cleavage. Also, we wanted to compare the cleavage efficiencies of the phthalimidyl amino supports under a variety of other, commonly used oligonucleotide deprotection conditions. Many conditions have been developed to deprotect oligonucleotides that contain modifications.

We also wanted to explore the AMA (ammonium hydroxide/methylamine) system, which is a newer fast deprotection method. It is becoming the reagent of choice for many high throughput synthesis laboratories with its fast 10-minute deprotection rate.

As a control we compared our results to the deprotection of a thymidyl 20mer prepared from a standard, ester linked, thymidine CPG.

Table 1: Coupling Efficiencies
Support
Avg. Coupling Effciency
Thymidine (1)
98.8%
Thymidine (2)
98.7%
Phthalimidyl-C-3-amino (1)
97.4%
Phthalimidyl-C-3-amino (2)
97.8%
Phthalimidyl-C-6-amino (1)
96.9%
Phthalimidyl-C-6-amino (2)
97.8%

Results

The experiment was carried out by synthesizing a thymidyl 20mer on a 15 μmole scale on each of the following supports; phthalimidyl-C-3 amino-CPG, phthalimidyl-C-6 amino-CPG, and T-CPG. The initial dimethoxytrityls were collected so that the exact starting scale could be determined, as well as the coupling efficiencies.

After synthesis, the supports were dried and carefully weighed out in 1 μmole scale aliquots into 4 mL screw top vials. Aliquots were then subjected, in duplicates, to the deprotection conditions shown in Table 2. The product was carefully isolated and the yield determined by absorbance at 260 nm. Table 2 shows the results of these deprotections, averaged and normalized to account for differences in the actual amount deprotected and for differences in coupling efficiencies. The results are given as ratios to the highest yielding deprotection off the thymidine support, which was potassium carbonate in methanol.

Table 2: Deprotection Conditions and Yield Determination (norm.) NH4OH- 30% aq. sol.; MeNH2- 40% aq. soln.
Method
Temp °C
Time
C-3 Pth
C-6 Pth
T-20 Std
NH4OH RT 24 hrs 0.92 0.92 0.99
NH4OH RT 48 hrs 0.80 0.87 1.00
NH4OH 55° 15 hrs 0.83 0.82 0.97
NH4OH 65° 4 hrs 0.72 0.79 0.96
3/1 NH4OH/EtOH RT 48 hrs 0.87 0.91 0.95
1/1 MeNH2/NH4OH1 55° 10 min 1.04 1.11 0.90
1/1 MeNH2/NH4OH 55° 15 hrs 1.13 1.18 0.98
0.4 M NaOH in 4/1 MeOH/H2O RT 15 hrs 0.97 1.00 0.90
0.05 M K2CO3 in MeOH3 RT 15 hrs 0.15 0.13 1.00

The products were analyzed by PAGE for purity. The products all looked exceptional with little n-1 evident in any of the syntheses. No evidence of any partially deprotected species as shown in the Figure 1 was observed by mass spectral analyses.

 

Comparison
Figure 1: Partially protected species

Discussion

The results confirm the anecdotal knowledge that the phthalimidyl amino supports do not fully cleave using ammonium hydroxide, although the differences of 10% to 20% seem hardly worth consideration given the benefits shown in the last article. In general, the various ammonium hydroxide conditions all yielded equal amounts of amino labeled product which was 80-90% of the yield from the thymidine support, after taking into account the slightly lower coupling efficiencies.

The actual yields were about 70% of the yield recovered from the thymidine support. An additional 10% or more was lost due to the lower coupling efficiencies of about 1% per cycle as shown in Table 1. We also see similar lowered efficiencies with almost all of the non-nucleosidyl modified supports. We have no proven explanation for the phenomenon, although one could suggest that the non-nucleosidyl derivatives do not extend away from the support properly.

Since we see no partially protected species as shown in Figure 1, we can only assume that the two phthalimidyl amide bonds cleave well before the linkage to the support. The aromatic amide link apparently does not cleave under these conditions, or we would see some of those side products in the crude product mix.We were pleased to find that the AMA system resulted in the highest yield with our linker. We believe that the extra absorbance is due to released aromatic linker side product. Although the longer deprotection with AMA gave slightly better results, the 10-minute deprotection with AMA is an excellent choice for this support. It is convenient to prepare and well suited for today's high throughput requirements. We are exploring this reagent more fully.

One disappointment was that the potassium carbonate deprotection conditions do yield considerably less product and should not be considered for use with the phthalimidyl protected amino supports. This deprotection condition is useful for phosphoramidite Cy5 dye. In conclusion, we found that with one exception, K2CO3, the phthalimidyl-3'- amino supports can be deprotected with any of the common deprotection conditions with good results.

References

  1. Reddy, M. P.; Farooqui, F.; Hanna, N. B. (1995) Tetrahedron Letters 36 (49): 8929-32.
  2. Kim, M.; Guengerich, F. P. (1997) Chemical Research in Toxicology 10: 1133-1143.
  3. 3. Schnetz-Boutaud, N. C.; Mao, H.; Stone, M. P.; Marnett, L. J. (2000) Chemical Research in Toxicology 13 (2): 90-95.