Figure 1.
(A) Structure of a simplified nine-member routing gene library. The ssDNA consists of 20-base noncoding regions (black lines Z1–Z4) and 20-base coding positions (colored bars [a,b,c]1–3). All library members contain the same four DNA sequences at the four noncoding regions. At each of the three coding positions, three mutually exclusive codons, (a,b,c)n, are present for a total of twenty-seven different routing genes. Resin beads coated with an oligonucleotide complementary to one codon (anticodon beads; gray ball at left) capture by hybridization ssDNAs containing the corresponding codon.
(B) To travel through the network, the ssDNA library starts on one or multiple DEAE columns (black column on left) and is hybridized to a set of anticodon columns (red, green, and blue columns) corresponding to the set of codons in the first coding position. The genes are thus physically partitioned into subpools based on sequence identity and can be processed accordingly. Each subpool is subsequently transferred to a distinct DEAE column, completing the first step through the network. The hybridization splitting, processing, and transfer are repeated for all subsequent coding regions. After completion of the final step, the library is concentrated on a reverse-phase column (RP; black-and-white column on right) and eluted for solution manipulation.
Figure 2.
Construction and Diversification of Routing Gene Populations
(A) Overlapping complementary oligonucleotides that span an entire gene (for example [Z–a]1–8 and a1–8′–Z2–9′) were assembled into full gene products (“all a,” “all b,” etc.) by primerless PCR and subcloned. Equivalent amounts of the ten resulting plasmids (a1–8, … , j1–8) were mixed and used as template for eight separate PCR reactions with noncoding region primer pairs (Zi/Zi+1′) that flanked a single coding position. The eight degenerate PCR products (Zn−xn−Zn+1) were assembled into a library of 108 different genes by primerless PCR (right).
(B) To generate ssDNA, a T7 promoter (pT7) was appended to the 3′ end of the double strand DNA library. The minus strand of the library was transcribed using T7 RNA Polymerase (T7 RNAP), and reverse transcribed from a Z1 primer using MMLV Reverse Transcriptase (MMLV RT) in a coupled reaction. The resulting DNA/RNA heteroduplex was treated with sodium hydroxide to hydrolyze the RNA, providing ssDNA.
Figure 3.
(A) Jeffamine 1500 (compound 1) was reacted with glutaric anhydride, and the singly acylated linker (compound 2) was purified over a HiTrap SP column. Purified compound 2 was coupled to NHS-activated Sepharose (gray ball). Treatment of the linkered resin compound 3 with TBTU/NHS, and subsequent incubation with a 5′-amino modified oligonucleotide (NH2-DNA), completed the synthesis of an anticodon column.
(B) Refractive index FPLC chromatograms of PEG compounds 1 and 2 before and after purification by cation-exchange chromatography. Linker compound 1 migrates as a bisamine (green trace) while compound 2 migrates as a monoamine (red trace).
(C) HPLC chromatograms of a 5′-aminated 20-base oligonucleotide (NH2-20mer) and a nonaminated ten-base oligonucleotide control (10mer) incubated with TBTU/NHS activated resin compound 3. Chromatograms of the starting material (black) and supernatant after 12 h (red) are shown. An unknown side-product of the coupling reaction (NH2-20mer side-product) is labeled.
Figure 4.
Linker Effects on Hybridization
The hybridization to anticodon columns of a ten-base noncomplementary oligonucleotide and a 20-base complementary oligonucleotide was analyzed by HPLC. Chromatograms of the hybridization load (blue), flow-through (red), and elute (black) are shown.
(A) shows the anticodon column that was synthesized by coupling the anticodon oligonucleotide directly to NHS-activated Sepharose.
(B) shows the anticodon column that was synthesized by coupling the anticodon oligonucleotide to NHS-activated Sepharose through a PEG linker.
Figure 5.
(Left) Genes are transferred from the “n−1” step DEAE columns to the “n” step anticodon columns by connecting all columns in series and cyclically pumping a high-salt buffer through the system with a peristaltic pump (gray box) for 1 h at 70 °C and 1 h at 46 °C. (Right) Genes are transferred from an “n” step anticodon column to an “n” step DEAE column by connecting the two columns in series and cyclically pumping 50% DMF through the system for 1 h at 45 °C. Arrows indicate the direction of flow.
Table 1.
340-Base ssDNA Hybridization Efficiencies and Specificities