DNA libraries and gene cloning

Genomic library construction (in order)

1. Isolate high-molecular-weight genomic DNA.

2. Partially digest it with a frequently cutting restriction enzyme, or shear it randomly, to give overlapping fragments of clonable size. Sau3AI recognises 5′-↓GATC-3′ (a 4-bp site). The GATC overhang it leaves is compatible with the overhang of BamHI, 5′-G↓GATCC-3′, so the fragments can be ligated into a BamHI-cut vector. The hybrid junction is always recut by Sau3AI, but it re-forms a BamHI site only if the neighbouring bases happen to match. A complete digest is avoided: it gives a fixed set of non-overlapping fragments, some too large or too small to clone, and it splits any gene that contains a site.

3. Size-fractionate the fragments (e.g., by sucrose-gradient centrifugation or gel fractionation) to keep the range the chosen vector accepts (see the vector-capacity table).

4. Treat the vector with alkaline phosphatase to remove its 5′ phosphates so it cannot self-ligate, then ligate the fragments into it.

5. Package in vitro into phage heads (λ and cosmid vectors), or transform or electroporate (plasmid, BAC), or transform yeast (YAC). Plate to obtain individual clones.

Cloning vectors and approximate insert capacity (typical ranges; exact limits vary by vector)

cDNA library construction (in order)

1. Isolate total RNA and purify poly(A)+ mRNA on oligo(dT)-cellulose. This relies on the poly(A) tail of eukaryotic mRNA; bacterial mRNA generally lacks a stable poly(A) tail.

2. Make the first strand with reverse transcriptase primed by oligo(dT), or by random hexamers, which reach the 5′ regions of long mRNAs. The product is an mRNA–cDNA hybrid. Reverse transcriptase often stops before reaching the 5′ end of the mRNA, so many cDNAs are incomplete at the 5′ end.

3. Make the second strand by the replacement method (Gubler–Hoffman). RNase H nicks the mRNA strand, leaving short RNA primers; DNA polymerase I extends them by nick translation; E. coli DNA ligase seals the nicks; T4 DNA polymerase then makes the ends blunt. In the older self-priming method, a hairpin at the 3′ end of the first strand primes the second strand and S1 nuclease cuts the hairpin loop, which loses some sequence from the 5′ end of the mRNA.

4. Add ends for cloning. With linkers, first methylate internal EcoRI sites with EcoRI methylase, then ligate blunt EcoRI linkers and digest with EcoRI to create sticky ends. Adaptors already carry a sticky end, so no digestion is needed. Ligate into a vector: a plasmid, or a λ vector such as λgt10 (for hybridization screening) or λgt11 (an expression vector for antibody screening).

5. Transform or package and plate. Each clone's frequency in the library mirrors its mRNA's abundance at the time the RNA was isolated.

Genomic vs cDNA library

Colony or plaque hybridization (sequence-based screening)

1. Replica-plate the colonies, or lift the plaques, onto a nitrocellulose or nylon membrane. Keep the master plate.

2. Lyse the cells or phage in situ, denature the released DNA to single strands, and fix it to the membrane by baking or UV cross-linking.

3. Hybridize with a labelled probe. A double-stranded probe must first be denatured. The probe can be a related gene, a degenerate oligonucleotide designed from partial protein sequence, or a heterologous probe from another species.

4. Wash at a stringency (temperature and salt) suited to the expected probe–target match. Higher temperature and lower salt mean higher stringency and less tolerance of mismatches.

5. Detect the positive signals (autoradiography or non-radioactive detection) and pick the matching colony or plaque from the master plate.

Antibody (immunological) screening, for expression libraries

1. The insert must be in an expression vector, in frame and in the correct orientation (e.g., a lacZ fusion in λgt11), so that the host makes the encoded protein.

2. Transfer the proteins expressed by the colonies or plaques onto a membrane.

3. Probe the membrane with an antibody specific to the protein of interest.

4. Detect the bound antibody with a labelled secondary antibody to find the positive clone.

Functional complementation screening

1. Introduce the library clones into a host strain carrying a specific mutation, for example an auxotroph or an enzyme-deficient strain.

2. Select or screen for restoration of the wild-type phenotype, such as growth on selective medium or restored enzyme activity.

3. A clone that complements the defect carries a functional copy of the corresponding gene.

4. The host must be able to express the clone. For a eukaryotic gene tested in a yeast or bacterial mutant, this generally means a cDNA library in an expression vector, because bacteria cannot splice out introns.

5. This isolates a gene purely by the function it restores, without any prior sequence or antibody information.

Screening methods compared

Common gene isolation strategies

1. Homology-based cloning: design degenerate primers or probes from conserved regions of a known homologous protein, then use PCR or hybridization to pull out the gene.

2. RT-PCR and RACE: when part of the transcript sequence is known, amplify the cDNA with gene-specific primers. RACE (rapid amplification of cDNA ends) recovers the missing 5′ or 3′ ends.

3. Chromosome walking: use a probe from the end of one genomic clone to find the next, overlapping clone, and repeat to move along the chromosome toward a mapped gene. Chromosome jumping skips over large or unclonable stretches by circularising large fragments, so that DNA sequences far apart are brought together in one clone.

4. Positional (map-based) cloning: start from the gene's map location, defined by linked markers, and narrow it down by walking or jumping and by testing candidate genes. No prior knowledge of the gene product is needed.

5. Subtractive hybridization: hybridize cDNA from the tissue or condition of interest (the tester) with an excess of cDNA or mRNA from a reference sample (the driver), and remove the hybrids, which are the sequences common to both. This enriches for transcripts found only in, or increased in, the tester.

6. Differential screening: probe duplicate filters of the same library with labelled cDNA from two different samples, and pick clones that give a signal with only one of them.

7. Expression cloning: isolate the gene from an expression library purely by the activity or antigenicity of its protein product.

Which strategy for which situation

Common expression host systems

Strategies to improve expression, solubility and recovery

1. Promoter and induction control: use a strong, regulated promoter (e.g., lac-derived or T7 systems induced with IPTG). Keep expression tightly off before induction if the protein is toxic to the host.

2. Codon optimization: replace codons that are rare in the host with codons the host prefers, to relieve translational stalling (or supply the rare tRNAs in the host).

3. Fusion tags: fuse an affinity tag (e.g., His-tag for metal-affinity purification, GST for glutathione affinity) or a solubility partner (e.g., MBP, thioredoxin, SUMO) in frame. These aid detection and one-step purification, and often improve solubility. Include a specific protease cleavage site if the tag must be removed.

4. Secretion signals: add a signal peptide to export the protein to the periplasm or the medium. The oxidizing periplasm favours disulfide bonds, and secretion leaves fewer host proteins to purify away.

5. Controlling inclusion bodies: lower the induction temperature and inducer concentration, co-express molecular chaperones, or use a solubility tag. If inclusion bodies still form, recover the protein by solubilizing it in a denaturant (e.g., urea or guanidine hydrochloride) and refolding it under controlled conditions.

6. Host and strain choice: match the host to the protein's modification needs (a mammalian or insect system, not E. coli, when function depends on authentic glycosylation). Protease-deficient strains reduce degradation.

Transposable elements: core facts

1. Transposons are DNA segments that move from one site in a genome to another. DNA transposons are moved by a transposase, usually encoded by the element itself. Transposition does not need sequence homology between the element and its target site.

2. Insertion sequences (IS elements) are the simplest bacterial transposons: a transposase gene flanked by short terminal inverted repeats.

3. Composite transposons (e.g., Tn5, Tn10) carry a central region, often an antibiotic-resistance gene, flanked by two IS elements.

4. Insertion creates a target-site duplication: a few base pairs of host DNA, repeated directly on both sides of the element, because the transposase cuts the target with a staggered cut.

5. Conservative (cut-and-paste) transposition excises the element and reinserts it elsewhere (e.g., Tn5, Tn10). Replicative transposition copies the element and leaves the original in place (e.g., Tn3, through a cointegrate that a resolvase separates).

6. Retrotransposons move through an RNA intermediate that is reverse-transcribed into DNA and inserted (copy-and-paste). LTR retrotransposons are flanked by long terminal repeats in direct orientation.

7. Uses in cloning and mutagenesis: transposon tagging, where an insertion inactivates a gene and its marker lets the gene be recovered (e.g., the flanking DNA by inverse PCR or plasmid rescue); P-element transformation of Drosophila; engineered vertebrate transposon vectors (Sleeping Beauty, piggyBac); and Tn7 transposition used to build recombinant baculovirus shuttle vectors (bacmids).

Gene targeting by homologous recombination (a typical knockout)

1. Design a targeting construct with two homology arms matching the sequence on either side of the gene. Between the arms, a positive selectable marker (e.g., neo, which confers G418 resistance) replaces part or all of the gene.

2. Optionally, place a negative selectable marker (e.g., HSV-tk) outside one homology arm. Correct homologous recombination exchanges only the DNA between the arms, so the outside marker is lost. Random (non-homologous) integration usually inserts the whole construct, so the marker is normally retained.

3. Introduce the linearized construct into the target cells, e.g., mouse embryonic stem cells, by electroporation.

4. Apply positive selection (G418, which keeps any stable integrant) and negative selection (ganciclovir, which is toxic to cells expressing HSV-tk) together. This enriches for correctly targeted clones.

5. Confirm correct targeting in the surviving clones by PCR or Southern blot. In mammalian cells, homologous recombination is much rarer than random integration, so enrichment is not proof of targeting. In yeast, homologous recombination is efficient.

Site-specific recombinases and targeted nucleases

1. Cre–loxP: Cre recombinase, from bacteriophage P1, recombines two 34-bp loxP sites. Each site is two 13-bp inverted repeats around an asymmetric 8-bp spacer. Two loxP sites in the same orientation on one molecule lead to excision of the DNA between them, leaving one loxP. Sites in opposite orientations lead to inversion. Flanking an exon with loxP sites ('floxing') and expressing Cre only in one tissue or at one time gives a conditional knockout.

2. FLP–FRT: the yeast FLP recombinase acting on FRT sites works the same way; it is often used to remove selection markers.

3. Nuclease-assisted targeting: a double-strand break made at the chosen locus greatly raises the rate of targeted change. In CRISPR–Cas9, the Cas9 nuclease is guided by a guide RNA whose roughly 20-nt spacer pairs with the target, next to a PAM (5′-NGG-3′ for the commonly used Streptococcus pyogenes Cas9). Repair by non-homologous end joining (NHEJ) leaves small insertions or deletions, often frameshifts, which gives a knockout. Repair by homology-directed repair (HDR) copies a supplied donor template and gives a precise edit or insertion. Zinc-finger nucleases and TALENs make targeted breaks with protein-based DNA recognition.

Transposon mutagenesis vs targeted knockout

Gateway cloning (site-specific recombination at att sites of phage λ)

1. The gene of interest is flanked by attB sites, by PCR or an earlier cloning step.

2. BP reaction: BP Clonase (λ integrase, Int, plus E. coli integration host factor, IHF) recombines the attB sites with the attP sites of a donor vector. The product is an entry clone flanked by attL sites, plus a by-product carrying attR sites.

3. LR reaction: LR Clonase (Int, λ excisionase Xis, and IHF) recombines the attL sites of the entry clone with the attR sites of a chosen destination vector. The product is an expression clone with attB sites regenerated, plus a by-product with attP sites.

4. The donor and destination vectors carry the toxic ccdB gene between their att sites. Unreacted vectors therefore kill standard E. coli, while recombined clones, which have lost ccdB, survive.

5. Recombination is directional and conservative, so one entry clone can be moved into many destination vectors (different promoters, tags or hosts) without re-cloning the gene.

Homologous-recombination cloning

1. λ Red recombineering in E. coli: three phage λ proteins carry out recombination. Exo is a 5′→3′ exonuclease that exposes single-stranded 3′ ends. Beta binds single-stranded DNA and anneals it to its complementary target. Gam inhibits the host RecBCD nuclease, protecting the linear DNA. Short homology arms, typically about 35–50 bp added on PCR primers, are enough to insert, delete or modify sequence in plasmids, BACs or the chromosome.

2. Yeast gap-repair cloning: a linearized vector and an insert whose ends match the vector ends are co-transformed into S. cerevisiae. The yeast's efficient homologous recombination joins them into a circular plasmid.

Gibson assembly (isothermal, overlap-based, seamless)

1. Design the fragments (inserts and linearized vector) so that the ends to be joined share overlapping sequence, typically about 20–40 bp.

2. Mix the fragments with one enzyme cocktail containing a 5′→3′ exonuclease, a DNA polymerase and a DNA ligase, and incubate at a single temperature, commonly 50 °C.

3. The exonuclease chews back the 5′ ends, exposing single-stranded 3′ overhangs at the overlaps, which anneal to their partner fragments.

4. The polymerase fills the gaps and the ligase seals the nicks. Several fragments are joined in their designed order in one reaction, with no restriction-site scar.

Comparing cloning approaches