Click to return to homepage

CBRL Core Technologies

Computationally Optimized DNA Assembly (CODA)

CBRL Computing ClusterCorrect self-assembly of CODA genes from chemically synthesized DNA oligonucleotides is achieved by making silent substitutions in the DNA code to strengthen correct hybridizations and disrupt erroneous ones in a way that produces a large DNA melting temperature gap between the correct versus all possible incorrect hybridizations.

While computationally demanding, this process ensures a temperature range within which, with virtual certain thermodynamic probability, correct hybridizations form and incorrect ones melt. The figure below shows that before the application of CODA melting temperature histograms for matches and mis-matches totally overlap; but, after sequence optimization by the CODA algorithm, correct assembly hybridizations are designed at temperatures far above those for all possible incorrect assemblies.

Hybridization melting temperatures and assembly of the Integrase (IN) gene of the Saccharomyces cerevisiae transposable element, Ty3. A, Computational melting temperatures for correct and incorrect hybridizations of oligonucleotides and intermediate DNA gene fragments without CODA melting temperature optimization. Every amino acid used the most optimal codon in highly expressed E. coli genes (21). B, Computational melting temperatures for correct and incorrect hybridizations of oligonucleotides and intermediate DNA gene fragments with CODA melting temperature optimizations. Computational melting temperatures were calculated with Mfold for [Na+] = 0.01 M and [Mg++] = 0.0015M with a folding temperature = 50oC. Solid lines represent correct hybridizations of oligonucleotides. Dashed lines represent correct hybridizations of intermediate DNA gene fragments. Dot-dash lines represent incorrect hybridizations of oligonucleotides. Dotted lines represent incorrect hybridizations of intermediate DNA gene fragments. C, lanes 1-10, IN intermediate DNA fragments 1 through 10 assembled from oligonucleotides optimized for codon usage only; lanes 11 and 13, molecular weight markers; lane 12, assembly of full-length IN gene (1,640 bp) from intermediate IN DNA fragments optimized for codon usage only. D, lanes 1-10, IN intermediate DNA fragments 1 through 10 assembled from CODA designed oligonucleotides; lanes 11 and 13, molecular weight markers; lane 12, assembly of full-length IN gene (1,640 bp) from CODA designed intermediate IN DNA fragments optimized for self-assembly and codon usage.

This insures that each oligo hybridizes only with its nearest overlapping neighbor(s) to self assemble into a correct full-length, single product, gene sequence.

The hierarchical assembly of overlapping intermediate DNA gene fragments and the full length Ty3 integrase gene products of the hybridization mixtures obtained with ~10 overlapping oligonucleotides for each intermediate gene fragment (lanes 1-10) optimized for codon usage but not CODA optimized for self assembly is illustrated in panels C and D. When the oligonucleotides optimized for codon usage but not CODA optimized for self-assembly are mixed and primer extended, a broad range of product sizes are observed (Panel C, lanes 1-10), and no full-length, 1,640 bp, product is apparent (Panel C, lane 12).

However, when the same procedures are performed with CODA designed oligonucleotides optimized both for codon usage and self-assembly single product intermediate gene fragments of the correct sizes are formed (panel D, lanes 1-10). Furthermore, when these 10 overlapping intermediate gene fragments are mixed and primer extended. A single 1,640 bp full-length gene product is observed ( Panel D, lane 12).

The hierarchical assembly of overlapping intermediate DNA gene fragments and the full length Ty3 integrase gene products of the hybridization mixtures obtained with ~10 overlapping oligonucleotides for each intermediate gene fragment (lanes 1-10) optimized for codon usage but not CODA optimized for self assembly is illustrated in panels C and D. When the oligonucleotides optimized for codon usage but not CODA optimized for self-assembly are mixed and primer extended, a broad range of product sizes are observed (Panel C, lanes 1-10), and no full-length, 1,640 bp, product is apparent (Panel C, lane 12). However, when the same procedures are performed with CODA designed oligonucleotides optimized both for codon usage and self-assembly single product intermediate gene fragments of the correct sizes are formed (panel D, lanes 1-10). Furthermore, when these 10 overlapping intermediate gene fragments are mixed and primer extended. A single 1,640 bp full-length gene product is observed ( Panel D, lane 12).

Translation Engineering

CODA-based Translation Engineering technologies are built around our ability to manipulate pause signals to control the speed of protein translation, a key factor in determining protein yield and functionality. Living organisms from bacteria to fungi to higher plants and animals all use translational pause signals to regulate the amount and quality of protein translated from a given mRNA. These explicit translation signals are encoded as pairs of codons that specify the species-specific identities of adjacent isoacceptor tRNA molecules that, in turn, control individual translation step-times. This phenomenon is independent of species-specific codon usage related to tRNA levels.

While the presence of these pause signals is universal, the actual codon pairs used to encode pauses vary widely from organism to organism. Thus, moving an ORF into a non-native host scrambles these signals, and results in randomly placed pause signals. This is why many attempts to improve protein yield and expression with foreign genes in non-native hosts fail. CODA-based Translation Engineering technologies avoid this complication and enable high level expression of any open reading frame in a non-native host, opening up a range of new research possibilities During the design and synthesis of each CODA gene we: