Past exam of the biochemistry and Molecular Biology course of the University of Cambridge / 2025 / Paper 1 / Section A

The useful product is a correctly identified, active and reproducible protein preparation, not merely a large expression band. I would first examine the coding sequence, protein sequence and available biological information to identify the likely mature chain, protein domains, signal peptides, transmembrane segments, disulfide bonds and required post-translational modifications. A predicted protein structure can guide construct boundaries, but cannot establish that the expressed product will fold or function. A eukaryotic genomic sequence containing introns generally needs replacement by an intron-free Complementary DNA or synthetic coding region for expression in Escherichia coli. I would verify the chosen isoform and full sequence, since a database annotation can be wrong.

The expression host follows from the biochemical requirements. Escherichia coli is inexpensive and convenient for many soluble proteins; yeast, insect or mammalian cells may be preferable for secreted proteins or essential glycosylation. A membrane protein also requires a suitable membrane environment during expression and an appropriate detergent or membrane mimic after extraction. I would preserve biologically necessary domains and modifications. Removing a flexible region to improve yield creates a different construct whose suitability must be justified by the intended analysis.

For a bacterial preparation, I would clone the coding region into an expression vector with a controllable promoter, a suitable ribosome-binding site, a selectable marker and a removable affinity tag. A terminal polyhistidine tag is a practical initial choice, provided the terminus is compatible with activity. Correct reading frame, both junctions and the complete insert must be checked by DNA sequencing. The T7 expression system offers strong inducible expression through a phage RNA polymerase; the original work established selective expression of cloned genes, but strong transcription alone does not guarantee soluble, active product.^[ref]

I would screen small cultures before scale-up, varying induction strength, induction duration and temperature, and comparing total lysate with soluble and insoluble fractions by SDS-PAGE. Empty-vector and uninduced controls identify host bands and background expression; an immunoblot against the tag can help identify the product. Expression at lower temperature or a weaker rate can improve protein folding; persistent insolubility may require changing the construct, host or folding environment. An inclusion body is evidence of abundant insoluble material, not of correctly folded protein. Refolding may be possible, but its yield and recovered activity need independent validation.

Once a small-scale condition produces active soluble material, I would expand the culture or fermentation while monitoring growth and retaining the validated induction conditions. After harvesting, controlled lysis at low temperature, an appropriate buffer solution and protection against proteolysis preserve the product. Centrifugation removes debris; a membrane-associated target instead requires isolation and solubilization of the relevant fraction. I would use immobilized metal affinity chromatography for a polyhistidine-tagged construct, wash away nonspecific binders, and elute under conditions compatible with activity. Tag cleavage, a second affinity step to remove tag and tagged protease, and an orthogonal separation such as ion-exchange chromatography can improve purity. A final size-exclusion chromatography step removes aggregates and separates different assemblies. No single chromatographic peak proves identity or homogeneity.

At each stage I would record protein concentration, total recovered amount and a suitable activity or binding readout, rather than optimizing yield alone. Final checks would include SDS-PAGE, intact or peptide mass spectrometry, a native-size measurement and the expected biological activity. Buffer-only and host-purification controls test whether apparent activity comes from contaminants; comparing tagged and cleaved material tests tag interference. Small frozen aliquots should be used only after checking stability and freeze–thaw effects. Scale-up is justified by recovery of the desired molecular species with retained function, with any differences from the native protein documented.

The same base sequence can present different molecular surfaces in different helical conformations. A-DNA, B-DNA and Z-DNA all retain paired antiparallel strands, but differ in handedness, backbone geometry, sugar puckering and groove accessibility. The familiar dimensions are approximate averages: sequence, hydration and bound molecules change the details, so they are structural signatures rather than rigid constants.

B-DNA is a right-handed DNA double helix with about $10.5$ base pairs per turn, a rise near $0.34\mathrm{nm}$ per pair and a diameter near $2\mathrm{nm}$. Its base pairs lie approximately perpendicular to the helical axis; the sugars are predominantly C2′-endo and the glycosidic bonds are anti. A broad accessible major groove and a narrower minor groove expose different edges of the paired nucleobases. Hydrated duplex DNA commonly adopts this family of conformations, although local bends and groove widths depend on sequence.

A-DNA is also right-handed, but is shorter and wider: roughly $11$ base pairs per turn, a rise around $0.26\mathrm{nm}$ per pair, and a diameter near $2.3\mathrm{nm}$. C3′-endo sugar puckering brings successive phosphates closer along each strand. The base pairs are inclined and displaced from the axis, producing a deep narrow major groove and a broad shallow minor groove. Reduced hydration favours A-DNA. Double-stranded RNA and many DNA-RNA hybrids adopt A-like geometry because the ribose 2′-hydroxyl constrains the backbone; this is important when extending a DNA-only comparison to cellular nucleic acids.

Z-DNA is a left-handed duplex with a zigzag backbone and a two-base-pair conformational repeat. It has approximately $12$ base pairs per turn, a rise near $0.37\mathrm{nm}$ per pair and a diameter around $1.8\mathrm{nm}$. Alternating purine–pyrimidine sequences, especially alternating G and C, favour it under suitable conditions. Purines commonly adopt syn glycosidic bonds and C3′-endo sugars, whereas pyrimidines remain anti with C2′-endo sugars. Its minor groove is narrow and deep, while the conventional major groove is flattened and poorly defined. Atomic-resolution analysis of an alternating CG oligomer established that a duplex can adopt this left-handed fold.^[ref] High salt in a crystal experiment is evidence of an accessible conformation, not proof that most cellular DNA is in that state; negative supercoiling and protein binding can instead favour local transitions.

These geometries determine how proteins read sequence. In B-DNA, the pattern of hydrogen bond donors, acceptors and hydrophobic groups in the major groove distinguishes the four oriented base pairs without strand separation. The minor groove carries less direct chemical information: A–T and T–A present similar patterns there, as do G–C and C–G. Specificity can nevertheless arise from DNA shape readout, including groove width, electrostatic potential, deformability and hydration, combined with contacts to the phosphate backbone. The chemical accessibility of base edges underlies the classic analysis of sequence-specific recognition of duplex nucleic acids.^[ref]

An alpha helix from a transcription factor can fit a B-DNA major groove, but the same docking arrangement need not fit the deep narrow groove of A-DNA. Many double-stranded RNA-binding interactions therefore emphasize backbone or shape, although A-form geometry does not make sequence-specific RNA recognition impossible. A Z-DNA-binding domain can recognize the characteristic left-handed backbone, making recognition partly conformation-specific rather than simply recognition of an exposed base code. In cells, nucleosomes, competing proteins, torsional stress and local hydration further determine access. Binding assays should compare sequences and conformations independently, using structural or spectroscopic controls: a sequence mutation can change both the chemical recognition pattern and the probability of adopting the recognized structure.

Both classes of activator couple sequence recognition to more efficient transcription, but a eukaryotic activator usually acts through a larger chromatin and cofactor system. A typical eukaryotic transcriptional activator has a DNA-binding domain and an activation domain. The former positions it at a regulatory sequence; the latter recruits or regulates other components. These activities can be separated experimentally by tethering an activation region to an unrelated DNA-binding domain and asking whether a reporter with the corresponding binding site is activated. Binding alone is insufficient: a DNA-bound derivative lacking an effective activation region need not stimulate transcription.

In eukaryotes, the target DNA is packaged in chromatin. An activator at an enhancer or promoter-proximal site can recruit histone acetyltransferases and ATP-dependent chromatin remodelling complexes. Histone modification and nucleosome repositioning can increase access to regulatory and promoter sequences; acetylated residues can also recruit reader proteins. These are context-dependent routes, not a universal requirement that every activated gene first undergo the same histone change. Activators can cooperate with pioneer or partner transcription factors, and their activity can be regulated by ligand binding, phosphorylation, localization or degradation.

Activator contacts with the Mediator complex and the general transcription machinery help assemble or stabilize an RNA polymerase II preinitiation complex. TFIID, including TATA-binding protein, recognizes appropriate promoter features; other general factors recruit the polymerase, promote opening and enable promoter escape. Regulatory contacts may additionally influence release of paused polymerase and productive elongation. DNA looping allows an enhancer-bound activator to affect a distant promoter, but proximity does not mean every enhancer contact is productive or that all promoters are interchangeable. Biochemical reconstitution showed that a mediator fraction was required for activated transcription in vitro, providing evidence that activator–polymerase communication can require an intermediate rather than a direct DNA-to-polymerase contact.^[ref]

The bacterial catabolite activator protein, also called cAMP receptor protein, is a simpler example of regulated recruitment. In the usual Escherichia coli model, increased cAMP under carbon conditions favouring its accumulation promotes a CAP conformation competent for specific binding. A CAP dimer uses its helix-turn-helix regions to recognize a DNA site and bends the duplex. At a class I CAP-dependent promoter such as the canonical lac arrangement, a site centred near $-61.5$ positions an activating surface to contact the C-terminal domain of the RNA polymerase alpha subunit. This stabilizes polymerase recruitment; the CAP–alpha-domain–DNA structure supplies a direct molecular basis for that contact.^[ref]

CAP is not limited to that single geometry. At class II promoters, a site near $-41.5$ overlaps the promoter region and permits additional contacts with polymerase, affecting more than initial recruitment. Promoter position, orientation and other repressors therefore matter. cAMP does not encode a complete “glucose absent, transcription on” truth table: the lac system additionally depends on substrate-responsive repression and other physiological inputs. An activating protein can improve a limiting step without determining the entire expression pattern.

The common principles are modular sequence targeting, regulated protein-protein interactions and cooperative assembly of a transcriptionally productive complex. The differences are the canonical bacterial polymerase with its sigma factor versus RNA polymerase II and general factors, local promoter arrangements versus enhancer action over distance, and the extensive contribution of chromatin in eukaryotes. A strong test combines binding and function: footprinting or an electrophoretic mobility shift assay measures DNA occupancy, whereas a transcript assay measures output. An activation-surface mutant that retains DNA binding but loses transcriptional stimulation separates recruitment from recognition. In a reconstituted system, removing and restoring a required cofactor tests causality; in cells, changes in chromatin or reporter activity alone do not identify the direct protein contact.