Cloning insulin cDNA into pUC19
A final-year student wants E. coli to express human insulin. She has the human insulin gene as a 1.6 kb genomic clone (with both introns) and a 0.3 kb cDNA clone, each inserted in a plasmid. She plans to cut the cDNA with EcoRI and BamHI and ligate it into pUC19 cut with the same two enzymes. pUC19 carries an ampicillin-resistance gene and a lacZα gene with a multiple cloning site containing EcoRI, SacI, KpnI, SmaI, BamHI, XbaI, SalI, PstI, SphI and HindIII. The cDNA contains an internal EcoRI site but no internal BamHI, SalI or HindIII site. The plasmid preparation is a mixture of singly cut and uncut molecules. She plans to select on ampicillin plates containing X-gal and IPTG, pick white colonies, and grow them in LB with ampicillin. A senior student says white colonies are always recombinant and that the genomic clone would give a better yield because it has introns.
1. Evaluate the cloning plan. Identify two technical faults and state the corrections.
Read the worked answer
First fault: the cDNA has an internal EcoRI site, so cutting with EcoRI will cut the insert into pieces. The insert must be released with enzymes that do not cut inside it; BamHI plus SalI or HindIII would give different, non-compatible ends and allow directional cloning.
Second fault: pUC19 is a cloning vector, not an expression vector. It has a lac promoter but no strong ribosome-binding site, no terminator designed for high expression, and no way to secrete or fold insulin. The cDNA should be placed in an expression vector (e.g., a T7 or arabinose-inducible vector) with a promoter, ribosome-binding site and terminator, and with the insulin sequence fused to a signal peptide or tag if periplasmic secretion or purification is intended.
Third fault: the plasmid preparation contains uncut and singly cut molecules; gel-purify the doubly cut vector and treat with alkaline phosphatase only if a single enzyme is used. With two different enzymes, vector self-ligation is low, but uncut plasmid will transform and give false positives. Also, the genomic clone is unsuitable because E. coli cannot splice introns; the cDNA is correct. For insulin, the protein needs disulfide bonds and removal of C-peptide, so expression as proinsulin or a fusion in a suitable host, followed by refolding, is more realistic than expecting native insulin directly.
2. Explain the blue-white screen and why 'white colonies are always recombinant' is false. Give two confirmatory tests.
Read the worked answer
pUC19 carries lacZα, which encodes the α-fragment of β-galactosidase. In a host with lacZΔM15, the α-fragment complements the defective enzyme to give active β-galactosidase. X-gal is cleaved to a blue product, so colonies with intact lacZα are blue. An insert cloned into the multiple cloning site usually disrupts the α-fragment, so the colonies are white. IPTG induces the lac promoter.
The statement is false because white colonies can arise from vector molecules that self-ligated after losing a few bases, from deletions in lacZα, from host mutations, or from an insert that does not disrupt the reading frame enough to abolish α-complementation. Conversely, a small in-frame insert can leave the enzyme partly active and give blue colonies. Ampicillin only shows that the plasmid is present; it does not show that an insert is present.
Confirm by colony PCR with primers flanking the cloning site (a larger product indicates an insert), by digesting miniprep DNA with the cloning enzymes, and by Sanger sequencing across the insert–vector junctions. Sequencing is essential before claiming the insulin sequence is correct.
3. The strain makes a protein of the expected size but it is inactive. Give three biological reasons and one remedy for each.
Read the worked answer
The protein is made in the reducing cytoplasm of E. coli, so the disulfide bonds between the A and B chains do not form. Remedy: secrete the protein to the periplasm with an OmpA or pelB signal peptide, or use an oxidising cytoplasm strain, or refold inclusion bodies in vitro.
The construct encodes proinsulin or a fusion protein; the C-peptide or fusion tag must be removed and the chains correctly paired. Remedy: design a cleavage site for a specific protease (e.g., enterokinase or TEV) and purify the cleaved product, or co-express the A and B chains.
Codon usage, mRNA secondary structure or proteolysis may give a truncated or misfolded product. Remedy: codon-optimise the cDNA for E. coli, alter the ribosome-binding site, use a protease-deficient host, and check by Western blot with an anti-insulin antibody. A eukaryotic host such as S. cerevisiae, which secretes correctly folded insulin, may be the better choice. The expected size on SDS-PAGE does not prove correct folding; a biological or receptor-binding assay is needed.