Recombinant DNA Technology, Genome Editing and Sequencing
Restriction enzymes, vectors, cloning, libraries, expression systems, mutagenesis, knockouts, CRISPR and prime editing, and sequencing for CSIR NET.
- 29 min
On this page · 17 sections
- Prerequisites
- The Hook: A Molecule That Was Cut and Rejoined
- Cutting: Restriction Endonucleases
- Joining: Ligation and Its Alternatives
- Carrying: Vectors
- Libraries
- Expressing: Hosts and Tags
- Altering: Mutagenesis and Loss of Function
- Editing: Programmable Nucleases
- Reading: Sequencing Platforms
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- DNA Replication, Repair and Recombination — polymerases, ligases, NHEJ and homologous recombination are the enzymes and pathways every method on this page borrows.
- Transcription and RNA Processing — promoters, operators, terminators and splicing explain why an expression vector looks the way it does and why cDNA differs from genomic DNA.
- Isolation and Separation of Macromolecules — gels, gradients and affinity chromatography, the read-out for most cloning steps.
- Omics, Fingerprinting, Barcoding and Radiolabelling — sequencing chemistries and coverage arithmetic are developed there and used here.
- Genes, Mutations and Fine-Structure Analysis — what a mutation is, which is what mutagenesis and editing are designed to create.
The Hook: A Molecule That Was Cut and Rejoined
How CSIR NET asks it. Part B: enzyme recognition sites and end chemistry, vector features and capacities, which selection or screen reports what, blue–white screening, Clarke–Carbone library arithmetic, PAM sequences, NHEJ vs HDR outcomes, which sequencing platform senses what. Part C: restriction maps and fragment sizes from gel data, library-size calculations, choosing an expression system from a yield/activity/glycosylation table, interpreting editing outcomes from sequencing read spectra, and diagnosing why a construct failed.
Cutting: Restriction Endonucleases
Restriction enzymes are bacterial nucleases that cleave DNA at defined sequences; the host protects its own genome by methylation (the restriction–modification system).
| Type | Recognition and cleavage | Cofactors | Use in cloning |
|---|---|---|---|
| Type I | Recognises one site, cuts at a distant, random position; single protein with restriction + methylation + recognition subunits | ATP, Mg²⁺, SAM | Not used |
| Type II | Cuts within or immediately beside a short (4–8 bp), usually palindromic site; separate methylase | Mg²⁺ only | The workhorse |
| Type IIS | Asymmetric site; cuts a fixed number of bases away from it (e.g. BsaI, BsmBI, FokI) | Mg²⁺ | Golden Gate, ZFN/TALEN nuclease domain |
| Type III | Cuts about 25 bp from an asymmetric site | ATP, Mg²⁺ | Rare |
End chemistry
| Enzyme | Site (cut shown by ^) | End produced |
|---|---|---|
| EcoRI | G^AATTC | 5′ overhang (AATT), “sticky” |
| BamHI | G^GATCC | 5′ overhang (GATC) |
| HindIII | A^AGCTT | 5′ overhang (AGCT) |
| Sau3AI | ^GATC | 5′ GATC — compatible with BamHI |
| PstI | CTGCA^G | 3′ overhang |
| KpnI | GGTAC^C | 3′ overhang |
| SmaI | CCC^GGG | Blunt |
| NotI | GC^GGCCGC | 8-bp site, rare cutter |
- Compatible ends: fragments can be ligated if their overhangs are complementary, even from different enzymes (BamHI + Sau3AI; SalI + XhoI). The hybrid joint is usually no longer cut by either enzyme.
- Isoschizomers recognise the same sequence (SmaI and XmaI both read CCCGGG); neoschizomers are isoschizomers that cut it at a different position (XmaI leaves C^CCGGG, a sticky end).
- Star activity: wrong buffer, glycerol or long incubation makes an enzyme cut near-matches.
- Methylation blocks cutting. Dam methylase (GATC) and Dcm methylase (CCWGG) of E. coli block overlapping sites of some enzymes; DpnI cuts only Dam-methylated GATC, which is the trick behind QuikChange mutagenesis.
In a random sequence with equal base frequencies, a site of length \(n\) occurs on average once every \(4^{n}\) bp:
$$\text{expected fragment length} = 4^{n}\ \text{bp} \qquad (4^4 = 256,\quad 4^6 = 4096,\quad 4^8 = 65{,}536)$$Sites containing CpG (such as NotI, GCGGCCGC) are rarer than this in vertebrate genomes because CpG is depleted, which is why NotI and SfiI are used as rare cutters for pulsed-field mapping.
Joining: Ligation and Its Alternatives
- T4 DNA ligase seals a nick between a 3′-OH and a 5′-phosphate using ATP; it works on sticky and blunt ends (blunt ligation is slower and needs more enzyme and DNA). E. coli ligase uses NAD⁺ and is poor on blunt ends.
- Vector self-ligation is suppressed by dephosphorylating the vector (alkaline phosphatase) so only the insert supplies a 5′-phosphate, or by cutting with two different enzymes (directional cloning).
- Linkers and adaptors add a restriction site to a blunt fragment; terminal transferase adds homopolymer tails for A/T or G/C tailing.
- Insert : vector molar ratio is usually set at about 3:1 for a single insert.
| Strategy | How the joint is made | Scar? | Typical use |
|---|---|---|---|
| Sticky/blunt ligation | Restriction digest + T4 ligase | Site retained | Classic subcloning |
| TA cloning | Taq polymerase leaves a single 3′-A on PCR products; vector is linearised with 3′-T overhangs | Small | Fast cloning of any PCR product |
| Gibson assembly | Overlapping ends (about 20–40 bp) chewed back by a 5′→3′ exonuclease, annealed, filled by polymerase, sealed by ligase — one tube, one temperature regime | Scarless | Multi-fragment assembly, no site needed |
| Golden Gate | Type IIS enzyme (BsaI, BsmBI) cuts outside its site, so the site is removed with the stuffer; unique 4-nt overhangs set the order | Scarless | Ordered multi-part assembly, one-pot digestion–ligation |
| Gateway | Phage λ integrase/excisionase recombination between att sites: BP reaction makes an entry clone (attL), LR reaction moves the insert into a destination vector (attR) | att scar (~25 bp) | Moving one insert into many vector backbones; ccdB counter-selection kills empty-backbone transformants |
Carrying: Vectors
A vector needs an origin of replication, a selectable marker, and a cloning site; a useful one also has a screen for inserts.
| Vector | Backbone / replicon | Insert capacity | Markers and screen |
|---|---|---|---|
| pBR322 | pMB1-derived ori, low–medium copy | up to ~10 kb | ampᴿ and tetᴿ; insertional inactivation (insert in tet → Ampᴿ Tetˢ) |
| pUC19 | pMB1 with mutated control, high copy | up to ~10 kb | ampᴿ + lacZα with MCS → blue–white screening on X-gal/IPTG |
| Phage λ (insertion / replacement) | λ genome, cos ends, in vitro packaging | ~10 kb (insertion), ~9–23 kb (replacement) | Plaque morphology; Spi selection for replacement vectors |
| M13 / phagemid | Filamentous phage, single-stranded progeny | ~1–several kb | ssDNA for Sanger sequencing, Kunkel mutagenesis, phage display |
| Cosmid | Plasmid + λ cos site; packaged in vitro, then replicates as a plasmid | ~35–45 kb | Antibiotic resistance |
| BAC | F plasmid (oriS, repE, parA/parB), 1–2 copies per cell | ~100–300 kb | Low copy → very stable, little chimerism; used for the human reference genome |
| PAC | Phage P1 replicon | ~100–150 kb | Intermediate |
| YAC | Yeast: CEN + ARS + two telomeres, TRP1/URA3 | up to ~1 Mb (often a few hundred kb) | Largest capacity; prone to chimerism and rearrangement |
| Shuttle vector | Two origins and two markers (e.g. pMB1 + ampᴿ for E. coli; 2µ or ARS/CEN + URA3 for yeast) | — | Build in E. coli, test in the second host |
| Expression vector | Strong regulated promoter, operator, strong ribosome-binding site, terminator, tag, protease site | — | Protein production |
Blue–white screening logic: the MCS sits inside lacZα. An empty vector gives functional β-galactosidase by α-complementation in a lacZΔM15 host, so X-gal is cleaved and colonies are blue; an insert disrupts the reading frame, so colonies are white. White colonies still need checking — small in-frame inserts can stay blue, and deletions can give white colonies with no insert.
Libraries
| Library | Starting material | Contains | Does not contain |
|---|---|---|---|
| Genomic | Genomic DNA, partial digest (or shearing) + size selection | Promoters, introns, intergenic DNA, whole genes | Information about expression |
| cDNA | mRNA → reverse transcriptase (oligo-dT or random primers) → second strand | Only expressed sequences, spliced, tissue- and stage-specific | Introns, promoters, untranscribed regions |
- A partial digest is used for genomic libraries so that fragments overlap; a complete digest would cut every gene at the same places.
- Clarke–Carbone equation — number of independent clones \(N\) needed for probability \(P\) that a given sequence is present, where \(f\) is the fraction of the genome in one insert:
- cDNA abundance bias: a few abundant mRNAs dominate. Normalised libraries (reassociation kinetics) flatten abundance; subtractive libraries enrich transcripts present in one sample and absent in another.
- Screening: colony/plaque hybridisation with a labelled probe, antibody screening of an expression library, functional complementation of a mutant host, or PCR with gene-specific primers.
Expressing: Hosts and Tags
| Host | Promoter / induction | Strengths | Limits |
|---|---|---|---|
| E. coli | lac/tac (IPTG), T7 in BL21(DE3), araBAD (arabinose), cold-shock cspA | Fast, cheap, very high yields, easy genetics | No N-glycosylation, few post-translational modifications, inclusion bodies, codon bias (helped by Rosetta-type strains), endotoxin in the lysate |
| Yeast (S. cerevisiae, Pichia pastoris) | GAL1 (galactose); AOX1 (methanol) in Pichia | Eukaryotic folding, disulfides, secretion, cheap scale-up | Hypermannosylated glycans, unlike human |
| Insect cells (baculovirus in Sf9, High Five) | Very late polh / p10 promoters | High yields of complex, folded proteins; large inserts | Simpler (paucimannose) glycans, no complex sialylation; lytic |
| Mammalian (CHO, HEK293) | CMV, EF-1α; stable or transient | Human-like glycosylation, correct folding of secreted proteins | Slow, expensive media, lower yields |
| Cell-free | Added T7 polymerase + lysate | Toxic and unnatural products, fast tests | Small scale, costly per mg |
| Tag | Matrix | Elution |
|---|---|---|
| His₆ | Ni²⁺- or Co²⁺-NTA (IMAC) | Imidazole gradient (or low pH / EDTA) |
| GST | Glutathione–agarose | Reduced glutathione |
| MBP | Amylose resin | Maltose |
| Strep-tag II | Strep-Tactin | Desthiobiotin |
| FLAG / HA / Myc | Anti-tag antibody | Peptide competition or low pH |
| SUMO, thioredoxin, NusA | — (solubility partners) | Cleaved off afterwards |
Tags are removed with a site-specific protease: TEV (recognises ENLYFQ↓G), thrombin, Factor Xa, enterokinase or SUMO protease. Inclusion bodies are addressed by lowering the induction temperature, weakening induction, co-expressing chaperones, exporting to the periplasm (pelB, malE signal), using a solubility tag, or by denaturing and refolding.
Altering: Mutagenesis and Loss of Function
Site-directed mutagenesis
| Method | Principle | Note |
|---|---|---|
| Oligonucleotide-directed (Hutchison–Smith) | Mutagenic primer annealed to a single-stranded template (φX174 in the original work, M13 in later versions), extended, transformed | The original method; Michael Smith shared the 1993 Chemistry Nobel with Kary Mullis |
| Kunkel | Uracil-containing template made in a dut⁻ ung⁻ host; after extension with the mutagenic primer, the template is destroyed in a ung⁺ host | Very high efficiency without selection for the mutation |
| QuikChange-type PCR | Two complementary mutagenic primers amplify the whole plasmid; DpnI then digests the Dam-methylated parental plasmid only | One tube; works on double-stranded plasmid |
| Megaprimer / overlap-extension PCR | Two rounds of PCR; the first product becomes a primer | Good for insertions and deletions |
| Cassette mutagenesis | Replace a restriction fragment with a synthetic duplex | Many changes at once |
Random and semi-random: error-prone PCR (Mn²⁺, unbalanced dNTPs, low-fidelity polymerase), DNA shuffling (fragmentation and reassembly of a family of related genes), saturation mutagenesis (NNK at one codon), alanine scanning (asks which side chains matter). Random methods plus a screen or selection make up directed evolution.
Knockouts and knockdowns
graph TD
A["Loss of function wanted"] --> B["Gene knockout
DNA is changed, heritable"]
A --> C["Knockdown
mRNA is destroyed or blocked,
DNA unchanged"]
B --> D["Classical: homologous recombination
in mouse ES cells
neo-positive + tk-negative selection
chimera, germline transmission"]
B --> E["Conditional: Cre-lox / FLP-FRT
tissue-specific or drug-inducible promoter
drives the recombinase"]
B --> F["Nuclease-based: ZFN, TALEN, CRISPR
double-strand break then NHEJ or HDR"]
C --> G["RNAi: siRNA, shRNA, miRNA mimics
Dicer, Argonaute/RISC"]
C --> H["Antisense oligos, morpholinos
(blocking, not cleaving)"]- Homologous recombination in ES cells (Capecchi, Evans and Smithies; Medicine Nobel 2007): a targeting construct with two homology arms carries a positive marker (neo, G418 resistance) between them and a negative marker (tk, killed by ganciclovir) outside one arm, so random integrants are removed. Targeted ES cells are injected into blastocysts, giving chimeras; germline transmission then gives heterozygotes to intercross.
- Cre-lox: Cre recombinase acts on 34-bp loxP sites (two 13-bp inverted repeats flanking an 8-bp asymmetric spacer). Two loxP sites in the same orientation → excision of the intervening DNA; opposite orientation → inversion. Tissue-specific promoters (or tamoxifen-activated CreER, or tet-on/tet-off systems) decide where and when. FLP–FRT from yeast is the equivalent second system, used with Cre for complex alleles.
- RNAi gives a knockdown: a 21–23-nt siRNA guides Argonaute in RISC to cleave the complementary mRNA. It is partial, reversible, usually transient (stable with shRNA), and prone to off-target effects and to saturating the miRNA machinery. A knockout removes the gene; a knockdown lowers the message — a distinction Part C tests directly.
- Gene trap and transposon (Sleeping Beauty, piggyBac, P element) insertion give large collections of disruptions without designing each one.
Editing: Programmable Nucleases
All first-generation editors work the same way: make a double-strand break at a chosen site and let repair finish the job.
| Platform | How the DNA is recognised | Nuclease | Key constraints |
|---|---|---|---|
| ZFN | Zinc-finger array; each finger reads about 3 bp | FokI, which must dimerise → used as a pair on opposite strands | Hard to design, context-dependent specificity |
| TALEN | TALE repeats, one repeat per base; RVD code NI = A, NG = T, HD = C, NN = G (or A) | FokI pair | Easy to design but large, repetitive constructs |
| CRISPR-Cas9 (SpCas9) | 20-nt spacer of an sgRNA (crRNA + tracrRNA fused) base-pairs the protospacer; needs a 5′-NGG-3′ PAM immediately 3′ of it | One protein; HNH cuts the target strand, RuvC the non-target strand → blunt break about 3 bp from the PAM | Only a 20-nt RNA has to be changed per target; off-targets at mismatched sites |
| CRISPR-Cas12a (Cpf1) | Single crRNA, no tracrRNA; T-rich PAM (TTTV) on the 5′ side | RuvC only, cuts in a staggered way → 5′ overhangs | Useful where NGG is absent; sticky ends favour defined insertions |
| Cas13 | Targets RNA, no PAM (a protospacer-flanking preference) | HEPN domains | Knockdown and nucleic-acid detection, not genome editing |
What repair does with the break
| Pathway | Outcome | When |
|---|---|---|
| NHEJ (Ku70/80, DNA-PKcs, LigIV–XRCC4) | Small indels → frameshift and knockout; the default | All cell-cycle phases; dominant in most cells |
| MMEJ / alt-EJ | Deletions between microhomologies | Predictable deletion alleles |
| HDR | Precise edit from a supplied donor (plasmid or ssODN) | S/G2 only, low efficiency; boosted by NHEJ inhibition, cell-cycle timing, donor design |
Editing without a double-strand break
| Editor | Architecture | Conversion | Notes |
|---|---|---|---|
| Cytosine base editor (CBE) | nCas9/dCas9 + cytidine deaminase + uracil glycosylase inhibitor | C•G → T•A | Acts on a small window in the non-target strand’s exposed ssDNA |
| Adenine base editor (ABE) | nCas9 + laboratory-evolved TadA adenosine deaminase | A•T → G•C | No natural DNA adenosine deaminase existed; it was evolved |
| Prime editor (PE) | Cas9 nickase (H840A) fused to a reverse transcriptase, guided by a pegRNA (spacer + primer-binding site + RT template) | Insertions, deletions and all base changes, template-written | No donor DNA and no double-strand break needed; PE3 adds a second nick to bias repair |
Off-target editing is measured by targeted deep sequencing of predicted sites, GUIDE-seq-type unbiased capture, or whole-genome sequencing; specificity is improved with high-fidelity Cas variants, truncated guides, paired nickases and ribonucleoprotein (rather than plasmid) delivery. The 2020 Chemistry Nobel went to Emmanuelle Charpentier and Jennifer Doudna for CRISPR-Cas9 as a genome-editing tool.
Reading: Sequencing Platforms
| Platform | Chemistry | What is detected | Read length | Characteristic error |
|---|---|---|---|---|
| Sanger (chain termination) | ddNTP terminators, capillary electrophoresis | Fluorescence of the terminator in a size-separated ladder | up to ~1 kb | Very accurate; low throughput, one template per reaction — still the confirmation standard |
| Illumina | Bridge amplification on a flow cell; reversible terminators, one base per cycle | Fluorescence imaged per cluster per cycle | ~50–300 bp (often 2 × 150 paired-end) | Highest throughput and per-base accuracy; substitutions; GC and homopolymer bias; short reads struggle with repeats |
| Ion Torrent | Sequencing by synthesis in wells on a semiconductor chip | H⁺ released on nucleotide incorporation → pH change | ~200–400 bp | Homopolymer length errors (indels) |
| PacBio SMRT | Single polymerase in a zero-mode waveguide; phospholinked fluorescent nucleotides | Real-time fluorescence pulses | tens of kb; HiFi reads from circular consensus | Random errors, averaged out by consensus; resolves repeats, isoforms and structural variants |
| Oxford Nanopore | DNA (or RNA) threaded through a protein pore by a motor enzyme | Ionic-current changes as bases pass | ultra-long (up to Mb) | Highest raw error; reads native molecules, so base modifications and direct RNA are detectable |
- Second generation = short, amplified, massively parallel (Illumina, Ion Torrent). Third generation = single molecule, long read, amplification-free (PacBio, Nanopore).
- Coverage \(C = LN/G\) (read length × number of reads / genome size); hybrid assemblies use long reads for structure and short reads for base accuracy.
- Sanger’s earlier rival, Maxam–Gilbert chemical sequencing, is of historical and exam interest only.
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| Genomic library: introns, promoters, whole genome | vs | cDNA library: expressed, spliced, tissue-specific |
| Partial digest (overlapping genomic fragments) | vs | Complete digest (every site cut, genes split identically) |
| Isoschizomer (same site) | vs | Neoschizomer (same site, different cut position) |
| Type II (cuts in the site) | vs | Type IIS (cuts a fixed distance outside it) |
| Insertional inactivation (pBR322 tet) | vs | Blue–white screening (pUC lacZα) |
| Selection (only transformants survive) | vs | Screening (all grow; you look for a phenotype) |
| Cosmid (packaged like λ, replicates like a plasmid) | vs | Phagemid (plasmid that can be packaged as ssDNA by a helper phage) |
| BAC (F replicon, 1–2 copies, stable) | vs | YAC (yeast chromosome, largest inserts, chimerism) |
| Knockout (DNA removed/disrupted, heritable) | vs | Knockdown (mRNA lowered, partial, usually transient) |
| loxP same orientation → excision | vs | loxP opposite orientation → inversion |
| Cas9: NGG 3′ PAM, blunt cut, sgRNA | vs | Cas12a: TTTV 5′ PAM, staggered cut, crRNA only |
| Base editing (deaminase, transition only, no DSB) | vs | Prime editing (RT + pegRNA template, any small change, nick only) |
| Illumina (fluorescence, short, amplified) | vs | Nanopore (ionic current, long, native molecule) |
| Ion Torrent (H⁺, homopolymer errors) | vs | PacBio (fluorescence in a ZMW, consensus accuracy) |
Common Mistakes
Practice
All questions below are original practice questions written in CSIR NET style. They are not previous-year questions.
Level 1: Foundation
- Which enzyme class cuts a fixed distance outside its recognition sequence, and which cloning method exploits it?
- What is the PAM requirement of Streptococcus pyogenes Cas9, and where does it lie relative to the protospacer?
- Name the two markers used for positive–negative selection in classical mouse gene targeting.
Answers
- Type IIS enzymes (BsaI, BsmBI, FokI); Golden Gate assembly (and the FokI domain of ZFNs/TALENs).
- 5′-NGG-3′, immediately 3′ of the 20-nt protospacer, on the non-target strand (the strand that carries the protospacer sequence); the blunt cut falls about 3 bp upstream of the PAM.
- neo (positive, G418 resistance, between the homology arms) and tk (negative, ganciclovir sensitivity, outside one arm).
Level 2: Part B Style
Q1 (practice). A genomic library is made in a BAC vector with a mean insert size of 150 kb from a genome of 3.0 × 10⁹ bp. The number of independent clones needed so that a given single-copy sequence is present with 99% probability is closest to:
- (a) 2 × 10⁴
- (b) 9 × 10⁴
- (c) 2 × 10⁵
- (d) 9 × 10⁵
Answer and explanation
(b). \(f = 1.5\times10^{5}/3.0\times10^{9} = 5\times10^{-5}\). \(N = \ln(0.01)/\ln(1-5\times10^{-5}) = -4.605/(-5.0\times10^{-5}) \approx 9.2\times10^{4}\). Note the shortcut: for small \(f\), \(N \approx 4.6/f\) at P = 0.99 and \(N \approx 3.0/f\) at P = 0.95.
Q2 (practice). In a 100 kb stretch of DNA with equal base frequencies and no sequence bias, the expected numbers of BamHI (6-bp site) and NotI (8-bp site) sites are approximately:
- (a) 24 and 1.5
- (b) 16 and 4
- (c) 400 and 24
- (d) 24 and 6
Answer and explanation
(a). \(10^{5}/4^{6} = 10^{5}/4096 \approx 24.4\); \(10^{5}/4^{8} = 10^{5}/65{,}536 \approx 1.5\). Real genomes deviate because of base composition and CpG suppression; NotI (GCGGCCGC) contains two CpG dinucleotides and is rarer still in vertebrate DNA.
Q3 (practice). A plasmid is mutagenised by whole-plasmid PCR with two complementary mutagenic primers, then treated with DpnI before transformation. The purpose of DpnI is to:
- (a) linearise the PCR product so it can be ligated
- (b) destroy the Dam-methylated parental plasmid, which is unmethylated in the PCR product
- (c) remove the primers
- (d) nick the product so that ligase can seal it
Answer and explanation
(b). DpnI cuts GATC only when Dam-methylated. The template came from E. coli and is methylated; the PCR product, made in vitro, is not, so it survives. This is the selection step of QuikChange-type mutagenesis — without it, parental plasmid swamps the mutant.
Q4 (practice). Which pairing is incorrect?
- (a) MBP tag — amylose resin — maltose elution
- (b) His₆ tag — Ni-NTA — imidazole elution
- (c) GST tag — glutathione agarose — imidazole elution
- (d) Strep-tag II — Strep-Tactin — desthiobiotin elution
Answer and explanation
(c). GST is eluted with reduced glutathione; imidazole belongs to IMAC (His tags), where it competes with the histidine imidazole rings for the metal.
Q5 (practice). Two loxP sites flank an exon. In tissues expressing Cre, the exon is inverted rather than deleted. The most likely explanation is:
- (a) only one loxP site is functional
- (b) the two loxP sites are in opposite orientations
- (c) Cre requires FRT sites for excision
- (d) the exon is longer than Cre can excise
Answer and explanation
(b). loxP is a 34-bp site with an asymmetric 8-bp spacer, so it has direction. Same orientation → excision of the intervening DNA as a circle; opposite orientation → inversion. This is a standard cloning error in making conditional alleles.
Level 3: Part C Style
Q1 (practice). A 6.0 kb circular plasmid is digested and run on an agarose gel:
| Digest | Fragment sizes (kb) |
|---|---|
| EcoRI alone | 6.0 (linear) |
| HindIII alone | 4.0, 2.0 |
| EcoRI + HindIII | 3.0, 2.0, 1.0 |
(i) How many sites does each enzyme have? (ii) Taking the single EcoRI site as position 0 and numbering around the circle in the direction that meets the nearer HindIII site first, give the positions of the HindIII sites. (iii) Which statements are correct?
(A) EcoRI has one site and HindIII has two. (B) With EcoRI at position 0 and numbering in that direction, the HindIII sites lie at 1.0 kb and 3.0 kb. (C) The 2.0 kb fragment in the double digest also appears in the HindIII-alone digest, so it contains no EcoRI site. (D) A partial EcoRI digest of this plasmid would leave some molecules uncut, giving a 6.0 kb circular species alongside the 6.0 kb linear.
- (a) A, B and C
- (b) A, C and D
- (c) A, B, C and D
- (d) B and C only
Answer and worked solution
(c).
- (i) EcoRI linearises the circle without cutting it further → one site. HindIII gives two fragments from a circle → two sites (4.0 + 2.0 = 6.0, matches).
- (ii) The double digest gives 3.0 + 2.0 + 1.0 = 6.0, matches, from three cuts (1 EcoRI + 2 HindIII). The 2.0 kb band survives both digests, so the EcoRI site must lie inside the 4.0 kb HindIII fragment, splitting it into 3.0 + 1.0. Put EcoRI at 0: one HindIII site is 1.0 kb away in one direction, the other is 3.0 kb away in the other direction, i.e. at positions 1.0 kb and 3.0 kb on a 0–6.0 kb map. Check: 1.0 → 3.0 is the 2.0 kb fragment; 3.0 → 6.0/0 → 1.0 is the 4.0 kb fragment, cut by EcoRI at 0 into 3.0 and 1.0. Correct.
- (C) True, and it is the deduction that fixes the map.
- (D) True: incomplete digestion leaves uncut circles, which migrate differently from the linear of the same length.
Q2 (practice). A human secreted glycoprotein is expressed in four hosts. Yield is mg of purified protein per litre; specific activity is relative to the native protein (set to 100).
| Host | Yield (mg/L) | Soluble fraction | Specific activity | Glycans |
|---|---|---|---|---|
| E. coli cytoplasm | 320 | 8% (rest in inclusion bodies) | 5 | none |
| E. coli periplasm (pelB) | 45 | 85% | 40 | none |
| Pichia pastoris (AOX1) | 180 | secreted | 70 | high-mannose, hypermannosylated |
| CHO (stable pool) | 35 | secreted | 98 | complex, sialylated |
(i) Why does moving the E. coli product to the periplasm raise specific activity eightfold while cutting yield? (ii) Which host would you choose for a crystallography project needing 50 mg of correctly folded protein, and which for a therapeutic candidate? (iii) Which statements are correct?
(A) The low activity in the E. coli cytoplasm is explained mainly by misfolding and absent disulfide bonds, not by the lack of glycans alone. (B) Pichia glycans are human-like, so its product can substitute for the CHO product in the clinic. (C) Total activity (yield × specific activity) is highest for Pichia among the four. (D) Refolding the inclusion-body material could raise the usable amount from the cytoplasmic E. coli culture.
- (a) A, C and D
- (b) A and B only
- (c) B, C and D
- (d) A and D only
Answer and worked solution
(a).
- (i) The periplasm is oxidising and contains the Dsb disulfide-bond machinery, so a secreted, disulfide-rich protein folds correctly there; the cytoplasm is reducing, so the same chain aggregates into inclusion bodies. Export capacity is limited, hence the lower yield.
- (ii) Crystallography: Pichia or periplasmic E. coli — both give ample correctly folded material, and glycan heterogeneity is often removed or trimmed before crystallisation. Therapeutic: CHO, for human-like complex sialylated glycans, which govern serum half-life and immunogenicity.
- (iii) (A) true — activity rises eightfold in E. coli with no change in glycosylation, so folding and disulfides are the limiting factor. (B) false — Pichia glycans are hypermannosylated and not human-like. (C) true: total activity ≈ 320 × 0.08 × 5 = 128 (cytoplasm, counting only soluble), 45 × 0.85 × 40 ≈ 1,530, 180 × 70 = 12,600, 35 × 98 = 3,430 → Pichia highest. (D) true — denaturation and controlled refolding is the standard route to recover inclusion-body protein.
Q3 (practice). A single sgRNA plus Cas9 is delivered to a cycling human cell line, with or without a single-stranded oligonucleotide donor carrying a 1-bp change plus a silent restriction site. Amplicon deep sequencing of 480 reads per condition gives:
| Condition | Unmodified | Small indels (1–15 bp) | Precise donor-derived edit | Large deletions (> 50 bp) | Total |
|---|---|---|---|---|---|
| Cas9 only | 180 | 288 | 0 | 12 | 480 |
| Cas9 + donor | 168 | 257 | 43 | 12 | 480 |
| Cas9 (H840A nickase) + donor | 446 | 7 | 27 | 0 | 480 |
(i) Calculate the editing efficiency (any modification) and the HDR efficiency for “Cas9 + donor”. (ii) Why does the nickase give far fewer indels but still some precise edits? (iii) Which statements are correct?
(A) Overall editing efficiency with Cas9 + donor is 65%. (B) HDR accounts for about 9% of all reads and about 14% of all edited reads. (C) The nickase result shows that a double-strand break is not strictly required for donor-templated repair. (D) Using a base editor instead would be a better route if the desired change were a C•G → T•A transition and no restriction site had to be inserted.
- (a) A, B, C and D
- (b) A and C only
- (c) B, C and D
- (d) A, B and D
Answer and worked solution
(a).
- (i) Edited reads = 480 − 168 = 312, so overall editing = 312/480 = 65%. HDR = 43/480 = 8.96% ≈ 9% of all reads, and 43/312 = 13.8% ≈ 14% of edited reads. (A) and (B) are both correct.
- (ii) A nick is repaired largely by faithful, ligation-based pathways and does not engage NHEJ end-joining the way a double-strand break does, so indels collapse. A nick can still be used as a substrate for donor-templated repair, at low efficiency — the principle that prime editing is built on.
- (iii) (C) follows directly from the nickase row. (D) is correct: base editors give transitions efficiently without a break and without a donor, but they cannot write in a new restriction site or any insertion — for that you need HDR or prime editing. All four statements hold.
Q4 (practice). A 12 kb bacterial operon with three tandem near-identical 1.1 kb repeats is sequenced three ways: Illumina 2 × 150 bp at 120× coverage, PacBio HiFi at 25×, and Nanopore at 30× with a mean read length of 18 kb. The Illumina assembly collapses the repeats into one copy; both long-read assemblies resolve three copies, but the Nanopore consensus carries five single-base errors inside the repeats, while HiFi carries none.
(i) Why does high Illumina coverage not resolve the repeat number? (ii) Which platform would you use to confirm the final sequence of one repeat junction, and why? (iii) Which statements are correct?
(A) Repeat resolution depends on read length relative to repeat length, not on coverage depth. (B) Nanopore’s residual errors reflect its higher raw per-base error rate; more coverage and better consensus can reduce them. (C) A hybrid approach — long reads for structure, short or HiFi reads for base accuracy — is the standard solution. (D) Sanger sequencing of a PCR product spanning the junction is a reasonable confirmation step.
- (a) A, B, C and D
- (b) A and C only
- (c) B, C and D
- (d) A, B and C
Answer and worked solution
(a).
- (i) A 150-bp read (or a ~500-bp fragment with its mate) cannot span a 1.1 kb repeat, so no read is anchored in unique sequence on both sides. The assembler sees identical paths and merges them. Adding depth multiplies identical, equally ambiguous reads — depth fixes base accuracy, read length fixes structure.
- (ii) For one junction, Sanger sequencing of a PCR amplicon with primers in the flanking unique sequence is the cheapest and most defensible confirmation; among the high-throughput options, PacBio HiFi combines span and accuracy.
- (iii) All four are correct. (B) is the standard remedy for nanopore consensus errors; (C) describes the usual assembly pipeline; (D) is routine practice.
Quick Revision Table
| Item | Fact |
|---|---|
| Restriction types | I (distant, random cut), II (within site), IIS (fixed distance outside), III (~25 bp away) |
| End chemistry | EcoRI/BamHI/HindIII 5′ overhangs; PstI/KpnI 3′ overhangs; SmaI blunt; NotI 8-bp rare cutter |
| Site frequency | Once per \(4^{n}\) bp: 256 (4-bp), 4,096 (6-bp), 65,536 (8-bp) |
| Ligation | T4 ligase (ATP; sticky and blunt); dephosphorylate the vector; insert:vector about 3:1 |
| Scarless assembly | Gibson (exonuclease + polymerase + ligase, 20–40 bp overlaps); Golden Gate (Type IIS, one-pot); Gateway (λ att, ccdB counter-selection) |
| Capacity ladder | Plasmid ~10 kb → λ ~23 kb → cosmid ~45 kb → BAC ~300 kb → YAC ~1 Mb |
| Screens | pBR322 insertional inactivation (Ampᴿ Tetˢ); pUC lacZα blue–white (X-gal/IPTG) |
| Library size | \(N = \ln(1-P)/\ln(1-f)\); \(N \approx 4.6/f\) at P = 0.99 |
| Genomic vs cDNA | Genomic: introns + promoters, partial digest. cDNA: expressed only, spliced, abundance-biased |
| Hosts | E. coli (no glycans, inclusion bodies, T7/BL21(DE3)); Pichia (AOX1, hypermannose); baculovirus/Sf9; CHO/HEK293 (human-like glycans) |
| Tags | His₆–Ni-NTA–imidazole; GST–glutathione; MBP–maltose; Strep-tag II–desthiobiotin; TEV cleaves ENLYFQ↓G |
| Mutagenesis | Kunkel (uracil template, dut⁻ ung⁻); QuikChange + DpnI; error-prone PCR and DNA shuffling for directed evolution |
| Knockout | ES cells + homologous recombination; neo positive, tk negative; chimera → germline |
| Conditional | loxP 34 bp: same orientation → excision, opposite → inversion; FLP–FRT; CreER (tamoxifen); tet-on/off |
| Knockdown | siRNA/shRNA → Dicer → Argonaute/RISC; partial, reversible, off-target risk |
| Editors | ZFN (3 bp/finger, FokI pair); TALEN (NI=A, NG=T, HD=C, NN=G); Cas9 NGG 3′ PAM, blunt; Cas12a TTTV 5′ PAM, staggered |
| Repair | NHEJ → indels (default, all phases); HDR → precise (S/G2, inefficient); MMEJ → microhomology deletions |
| No-DSB editing | CBE C•G→T•A; ABE A•T→G•C (evolved TadA); prime editing = nickase + RT + pegRNA |
| Sequencing | Sanger ddNTP ~1 kb; Illumina reversible terminators, short, most accurate at scale; Ion Torrent H⁺ (homopolymers); PacBio ZMW HiFi; Nanopore current, ultra-long, native modifications |
Key Takeaways
What’s Next
- Animal and Agricultural Biotechnology — the same cloning, editing and marker tools applied to transgenic animals, crops and tissue culture.
- Marine and Environmental Biotechnology — where engineered and natural organisms are put to work on products and pollutants.
- Enzyme and Microbial Technology — taking a cloned enzyme from a shake flask to immobilisation and process use.
- Sequence Alignment, BLAST and FASTA — what you do with the reads once a platform has produced them.
- Quick Revision — the whole unit on one page.