Identification of the Target Terminal Exon and PCR Amplification

Human adenoviruses have a well-conserved genome organization, and the terminal exons of the MLTU are well defined (Fig. 1). Using readily available sequence databases, it is a straightforward matter to identify the location of the target capsid protein in your serotype of interest. We will go through a strategy designed for generating an Ad5-based vector that replaces the Ad5 fiber with the tandem fibers of Ad41. Fiber is coded within the L5 terminal exon (Hexon in L3, Penton in L2). We use DNA Strider as a simple software package for basic sequence analysis.

1. Knowing the relative location of the target gene, it is useful to identify the proper open reading frame (ORF) for the target capsid protein. Identify the translation initiation codon ATG. The termination codon is revealed when doing a primary translation of the DNA in frame with this ATG. It is always important to confirm several times that you are targeting the right sequence. An old carpenter's rule is: "measure twice and cut once."

2. The procedure that we describe is a terminal exon replacement of fiber. To accomplish this we must identify the essential RNA-processing elements that define the fiber terminal exon; these elements include the fiber splice acceptor and the fiber poly A site (L5).

a. The poly A site is identifiable by the highly conserved AATAAA hexameric sequence that is immediately 3' to the ORF termination site. The poly A cleavage site is usually 15-20 nucleotides 3' to the AATAAA, and an additional 20 nucleotides 3' to the cleavage site usually contributes to cleavage efficiency (Fig. 2). PCR primers should be created such that the downstream domains are included in the amplified terminal exon domain.

b. The 5' splice acceptor (SA) site does not have as highly a conserved sequence motif, but can be easily identified, and is located 5' to the ATG of the fiber ORF. The 5' SA site has a highly conserved AG:A trinucleotide where cleavage occurs between the G:A residues. Immediately upstream of the conserved trinucleotide is a polypyrimidine tract and a branch point A. (In a remarkable

A

F5

F41L

F41S

F7

F3

F12

F4

F17

B

FS

F41L

—i

F41S

—i

F7

Oo

F3

F12

F4

F17

TCCG C AC C C ACTATCTTC ATGTTGTTGC ¡iGATG GGCACAAT CTTCGC ATTTC TTTTTTTCC ^G, ^TG

CCCTGAC C C ACGATCTTC AT CTTG---C BiGiTG

CCCAC AAT CTTCA---TGTCTTTCTTCC W31TG

CCCACAATCTTC--ATTTCTTTATCCCCiGATG TCCC TAG C C ACCATCTTC AT CTTTTC - T W3ATG CCCTCAAT CTTCATTTTTTATTTCTATTM3 VTG TCCACAATTTTCATTGTC T-TCCCTCTC M3 ^TG

^TTGCCCAAGAA GTAACAGAACAA ^TAACTGAACAA IGAC TGA

rGA CAACAAAAATAAAGTTCAACATTTTTTATTGAAATTCCTTTTACA

TAA TAA TAA

C AAAAT AAAGTTTAAC TG T TTTATTTAAAATCA

AGAAT CGTTTGTGTTATGTTTCA AATATTGTTGTTTTTG TTTTTATA AATC ATTGCAGGCACAAT CTTCGCA

^TTACCCAAGAA

"TGAAACAAAAATAAAGTTCAAGTGTTTTATTGATTCA UTTGCCCAACAA

AAACACACACAAAACACAAATTGCGTA

AAGACCAATAAACGTGTTTTTTATTTCA

Fig. 2. Sequence conservation of fiber splicing and 3'-processing element sites. (A) Fiber DNA sequences upstream of Hie corresponding translation initiation codon (AUG) were aligned using ClustalW. Conserved residues are denoted with an asterisk. Sequences corresponding to pre-mRNA splicing elements are indicated as follows: splice acceptor (boxed), polypyrimidine tract (dashed line), putative branch site (solid line). (B) 3'-Fiber sequences were aligned manually at the consensus hexanucleotide AATAAA (*). Downstream elements include sequences corresponding to U/GU-rich regions in fiber pre-mRNA and putative adenine cleavage sites. Fiber translation stop codons are underlined.

example of efficiency, in many Ad serotypes, the fiber splice acceptor sequence is coincident with the ATG translation initiation codon, and the polyadenylation AATAAA is coincident with the translation termination codon. PCR primers should be created such that the polypyrimidine tract and the branch point A are included in the amplified terminal exon domain.)

3. Splice acceptor 5' PCR primer oligo: the 5' oligo is designed to complement the specific sequence that begins at the branchpoint A of the 5' splice acceptor. In the case of the fiber 41 tandem genes, the sequence upstream of the initiation ATG is: CCTTTTTACCCTGACCCACGATCTTCATCTTGCAGATG. We selected the sequence immediately upstream of the branchpoint A for the Ad41 short 5' splice acceptor: CCTTTTTACCCTGACCC. For cloning purposes, we add sequences that will allow cloning into both the expression cassette and into the viral recombination cassette. In our case we have added sequences for Bam HI and PacI sites and a few buffer nucleotides to protect the ends of the PCR products: cccggatccttaattaa. The 5' splice acceptor primer for F41 tandem is 5'-cccggatccttaattaa (CCTTTTTACCCTGACCC) and corresponds to the coding strand of fiber.

4. Poly A site 3' PCR primer oligo: based on the position of the AATAAA hexanucleotide located in the coding strand of viral DNA and in order to include the majority of sequence that influences polyadenylation, simply count 50 nucle-otides downstream (3') of the AATAAA. This should give sufficient sequence downstream of the AATAAA to include the authentic mRNA cleavage site (at approx 20 nts) as well as sequence that contributes to the efficiency of the 3' RNA cleavage event. In this instance the 3' primer would correspond to the reverse complement of the coding strand. In the case of Ad41, AATAAAATA TTGTTGTTTTTGTTTTTATAACTTTATTGATCATTTTACAGAATT represents the coding sequence of interest. The sequence that we select for our PCR primer is indicated by bold/italicized letters. The reverse complement of this sequence 5'-AATTCTGTAAAATGATCAATAAAG corresponds to the primer necessary for amplification from the coding strand. To the 5'-end of this primer, restriction enzyme sites should be added for the purpose of cloning. We have added a BamHI linker (GGATCC) as well as buffer sequence to facilitate efficient digestion. The final 3' primer is cccggatcc(AATTCTGTAAAATGA TCAATAAAG).

5. PCR amplification from miniprep viral DNA: amplification from viral miniprep DNA is straightforward and should be accomplished using standard procedures. For routine 50-|L amplifications, 30 ng of viral DNA is sufficient template.

0 0

Post a comment