Biotechnology & Genetic Engineering: Chapter 3 — Vectors and cloning

🔬 Key Illustrated Scientific Figure

Full-Screen Zoom

Click to enlarge on this page

3.1 What a vector must have #

3.1.1 Origin, marker, and cloning site #

A vector needs an origin, or the DNA vanishes as the cell divides. Copy number is part of the origin. pUC-type ColE1 derivatives can reach hundreds of copies; the F origin of a BAC stays near one. High copy yields more DNA and more burden, and it is a poor place for a toxic product or a huge insert.

A selectable marker, often ampicillin, kanamycin, tetracycline, or chloramphenicol resistance, keeps cells that received the vector. Resistance does not prove an insert. Yeast markers often complement an auxotroph. A multiple cloning site is a run of unique restriction sites, often inside lacZα, where the circle can be opened without destroying origin or marker. A second origin makes a shuttle vector. None of these extras saves a molecule that cannot replicate or be selected.

3.2 Vectors compared by the insert they carry #

3.2.1 Plasmids, lambda, cosmids, BACs, and YACs #

Ranges below are usual textbook values, not a guarantee for every backbone.

A cosmid is packaged like lambda and then lives as a plasmid: the phage is a syringe. YACs made long-range maps possible and often joined two unrelated fragments. BACs, smaller and cleaner, carried reference-genome projects. A 2 kb coding sequence belongs on a plasmid. A region of one or two hundred kilobases, with its regulatory DNA, belongs on a BAC.

3.3 Expression vectors and hosts #

3.3.1 Promoter, ribosome-binding site, and tag #

The promoter must suit the host and the hazard of the product. In E. coli, lac and tac are LacI-repressed, araBAD is tighter, and T7 is so strong that folding often loses. Yeast uses galactose-inducible promoters; Komagataella phaffii (formerly Pichia pastoris) uses a methanol-inducible alcohol-oxidase promoter. Baculovirus polyhedrin and p10 promoters are very late and very strong. CMV is a strong mammalian promoter that can be silenced; EF1α and CAG are alternatives.

Bacteria need a Shine–Dalgarno sequence about 6–8 nucleotides upstream of the AUG. Eukaryotes need a sensible Kozak context. A human codon served by a rare E. coli tRNA stalls the ribosome; codon optimisation changes the DNA, not the polypeptide, and can also change the mRNA in unpredicted ways. A His-tag, commonly six histidines, binds nickel or cobalt and comes off with imidazole. GST or maltose-binding protein can aid solubility and is removed with a protease such as TEV. A tag shows you have the handle. It does not show the protein is active.

3.3.2 Bacterial, yeast, insect, and mammalian systems #

Inclusion bodies are aggregates formed when synthesis outruns folding, especially for disulphide proteins in the reducing cytoplasm. Refolding sometimes works and often fails. Lower temperature, a weaker promoter, periplasmic export, or a different host are the logical responses, not recipes given here.

CHO cells are usual for licensed antibodies because they add complex N-glycans, still not identical to human ones. Engineered lines remove non-human sugars or core fucose when that is part of the drug. Immunology owns the Fc consequence. Yeast was the right host for hepatitis B surface antigen because the particles do not need a human glycan. Baculovirus works by putting the foreign gene under the polyhedrin promoter of Autographa californica nuclear polyhedrosis virus: huge yield, dying cell.

3.4 Libraries #

3.4.1 Genomic versus cDNA #

A genomic library represents total DNA, including introns and promoters. The Clarke–Carbon count of clones for probability P is N = ln(1−P) / ln(1−f), where f is insert length over genome length. For P = 0.99 and small f, N is about 4.6/f. E. coli, about 4.6 million base pairs, needs on the order of a thousand clones at 20 kb inserts. A human haploid genome of about 3 billion base pairs needs several hundred thousand at that insert size, and on the order of a hundred thousand in 150 kb BACs.

A cDNA library represents one sample’s spliced mRNA. It is what E. coli can translate, and it lacks promoters. Abundant transcripts dominate; a rare one may be absent. Normalisation is the idea of flattening that bias. An expression library adds the requirement that the host actually make a recognisable protein.

3.5 Key terms, summary, and self-check #

Key terms. Origin; copy number; selectable marker; multiple cloning site; shuttle vector; cosmid; BAC; YAC; Shine–Dalgarno sequence; His-tag; inclusion body; CHO cell; baculovirus; genomic library; cDNA library; Clarke–Carbon relation.

Summary. A vector needs an origin, a marker, and a cloning site. Insert size runs from plasmids, through lambda and cosmids, to BACs and YACs, with chimaeras the YAC’s failure. Expression needs a host-matched promoter, a ribosome-binding site, and often a tag. E. coli is fast and makes inclusion bodies without N-glycans; yeast and insect glycans are not human; CHO cells supply the glycans many medicines need. Genomic libraries are sized by genome fraction. cDNA libraries are spliced and abundance-biased.

Self-check.

1. Antibiotic-resistant cells contain no insert. What worked, and what conclusion is illegitimate?

2. Order plasmid, cosmid, BAC, and YAC by insert size. Why was a YAC a poor final map unit?

3. Why can the right polypeptide from E. coli still be the wrong glycoprotein drug?

4. A His-tag elution gives one band. What is still unknown?

5. Why will a liver cDNA library lack an embryonic pancreas promoter, and why is a genomic clone of that region still not ready to express the protein in E. coli?

—

—

—