In genetic engineering, a vector is a vehicle used to transfer foreign genetic material into a host cell. There are many vectors available to molecular biologists, and the vector chosen usually corresponds to the amount of genetic material involved. Types of vector include plasmids, bacteriophages, cosmids, phagemids and artificial chromosomes.

Vectors must have three essential properties:

1. The ability to replicate in the host cell (i.e. they must be replicons or contain multiple replicons)
2. The ability to be readily introduced into a host cell
3. The ability to have foreign DNA spliced into it

The host cells which receive the vector must also possess important properties:

1. Deficiency in restriction enzymes (in E. coli this involves deletion of the hsdR gene to inactivate the K restriction-modification system)
2. Stable maintenance of foreign DNA, without rearrangement or recombination (often done by mutating recombination genes such as recA and recF)
3. Disablement of the host outside of the laboratory environment to prevent the spread of a pathogenic strain in the wild (often done by using an auxotroph which requires a nutrient medium supplied only in the lab)


Plasmids are a common choice of vector. Often synthetic plasmids are used that contain an origin of replication, a selectable marker (often for antibiotic resistance) and cloning site(s) which can be cut by restriction enzymes for the insertion of foreign DNA. Plasmid regions with multiple adjacent cloning sites (i.e. a cluster of restriction sites) are called polylinkers.

Natural plasmids can have a copy number ranging from 1 to several thousand, and a size ranging from <1kbp to >100kbp in size (the larger ones are hard to distinguish from the bacterial chromosome). Because many plasmids code for their own transfer to neighbouring cells, using them can pose a safety risk (hence the need for auxotrophic growth).

Plasmids are relatively easy to isolate by alkaline lysis:

1. Grow a bacterial culture with antibiotic
2. Harvest the cells by centrifugation
3. Break the cells open under conditions favouring plasmid isolation
4. Denature genomic DNA to single stranded form in pH12 conditions (plasmid DNA is supercoiled; does not denature)
5. Bring the solution to pH7 to create a tangled aggregate of linear genomic DNA
6. Centrifuge to isolate the aggregated genomic DNA in a pellet, which can then be discarded
7. Treat the supernatant with RNase to remove RNA
8. Use ethanol precipitation to isolate plasmids

Once a plasmid has been isolated, a recombinant plasmid is made like so:

1. Linearise the plasmid with a restriction enzyme that has a recognition site in the polylinker
2. De-phosphorylate the plasmid using alkaline phosphatase (to prevent self-ligation)
3. Inactive the phosphatase with heat so that it doesn't de-phosphorylate the DNA to be ligated
4. Mix the foreign DNA and digested plasmid with ligase to generate recombinant products (cutting with the same restriction enzymes provides complementary sticky ends to aid in ligation)
5. Transform bacteria. Bacteria are made competent (capable of transformation) either by chemical treatment with CaCl2 or by electroporation (the latter being more efficient). Note that most bacteria are not transformed – only a million colonies per µg DNA – so transformation is a huge limiting step
6. Allow plasmids to replicate and screen recombinants by plating on antibiotics

Plasmids are straightforward to isolate and exist in high copy numbers, making them suitable vectors, but the transformation process is limiting and the possible range of insert sizes is only 0-10kb; thus, for example, they would not be suitable when making clones for a genomic library.

Bacteriophage lambda:

Bacteriophage lambda is a good alternative vector. The problem of transformation is avoided because bacteriophages naturally inject their DNA into host cells. The genome of lambda is double-stranded and composed of 49kb, or thereabouts. The terminal ends of these two strands are single-stranded and complementary: they form cohesive (cos) sites. When the genome is in low concentration, the cohesive ends self-associate and the molecule circularises. When the genome is found in higher concentration, it is easier for multiple genomes to link by their cos sites, forming a long linear structure called a concatemer.

When a bacteriophage injects its genome into a bacterium, the genome either circularises and is replicated and transcribed to produce more phages which eventually break out of the cell and spread (lysis – desirable in gene cloning), or they are integrated into the bacterial genome and silenced by a repressor molecule (lysogeny – undesirable in gene cloning). In the latter situation, the phage genome is called a prophage. Lytic growth can be visualised in beakers and on agar plates. Where bacterial lysis occurs, the beaker liquid (or plaques of cells) appear paler than where lysogeny has occurred (because the bacterial cells normally create a cloudy solution, and when lysed they are dead so cannot do this).

The lambda genome replicates by rolling circle replication. This involves ‘spinning off’ a concatemer of replicated DNA, joined at cos sites, which can later be broken apart (by phage-encoded endonucleases) into individual genomes that re-circularise. The circularised genomes are packaged into heads which join with pre-assembled tails. A fixed length of DNA is always packaged – there are mechanisms to prevent packaging of too small or large an amount – this is called the packaging constraint.

Two problems may be encountered when using phage lambda as a cloning vector:

1. The presence of too many restriction sites, which must be removed by site-directed mutagenesis
2. Only DNA whose length is 78-105% of the wild-type phage genome can be packaged. This only leaves room for the insertion of about 3kb of foreign DNA

The packaging constraint in phage lambda can be overcome by replacing the genes for lysogeny (integration and excision; not needed) with 20kb of foreign DNA. Lambda insertion vectors involve the deletion of the central ‘lyosgeny’ genes and then the ligation of the remaining left and right ‘arms’ of the genome to produce a unique restriction site, which can then be used to insert foreign DNA. However, you cannot delete more than 25% of the wild-type lambda DNA, or the genome will not be effectively packaged and replicated. The size of fragment that can be inserted into the insertion vector depends on the extent of the deletion of the lysogenic genes - thus, there is still an upper limit in that if not much of the lysogenic genes were deleted, then little foreign insert can be added. Insertion vectors are generally suitable for inserts that are 0-10kb in size.

Lambda replacement vectors involve replacing the central, non-essential lysogenic genes with ‘stuffer DNA’ that contains multiple restriction sites to facilitate its own 'chopping up' and prevent its own reinsertion during cloning. This makes the vector big enough to grow on its own, but also easily removed when we want to replace it with cloned DNA. Replacement vectors are generally used to clone larger fragments of DNA (10-20kb) because the entire complement of lysogeny genes has been removed, rather than an unknown proportion as in insertion vectors. However, the insert still cannot be larger than the upper limit according to the packaging constraint, but also not so small that packaging is not possible (i.e. if it is <10kb, it is probably better to use an insertion vector than a replacement vector).

When using bacteriophage lambda as a cloning vector, concatemers of recombinant phage DNA are made, and then packaged in vitro into empty phage heads and tails (bought industrially). The packaged concatemers are ready to infect cultures of E. coli.

Bacteriophage M13:

Bacteriophage M13 is a filamentous phage: a single-stranded DNA molecule (wild-type: 6.4kb) surrounded by (predominantly) the same coat protein to form a long filament under the microscope. It is often said to be ‘male-specific’ as it binds to strains of E. coli that are F+ and have a sex pilus for it to dock to. Because the DNA is single-stranded, there is much less of a packaging constraint when using M13. This is a major advantage over lambda. However, in practice, when the genome is over 10kb, there are often genome rearrangements, so there is a packaging constraint of sorts.

The M13 phage circularises once inside E. coli and then forms a double-stranded ‘replicative form’ (RF), which replicates to form more double-stranded RFs that can ultimately become single-stranded again and be packaged with coat proteins.

Besides the lesser packaging constraint of M13, there are many other advantages, including the ease of DNA purification and restriction digestion (as is the case with plasmids). Because of its single-stranded and lengthy nature, M13 was originally used in DNA sequencing projects. Nowadays, M13 is typically used in phage display, which is a technique used in the study of protein-protein, protein-peptide, and protein-DNA interactions that uses bacteriophages to connect proteins with the genetic information that encodes them.

The insert size range for bacteriophage M13 is 0-10kb, as in plasmids and in lambda insertion vectors. Thus, although using M13 overcomes some packaging problems and difficulties with transforming E. coli, there is still a limit on the insert size which would prevent cloning of large fragments for genomic libraries.

The initial part of cloning with M13 is just like using a plasmid vector – including the physical transformation of E. coli. Like lambda, it is then necessary to plate on a lawn of E. coli, but as the production of new M13 phage particles does not kill the cells the plaques will appear turbid (a plaque of slow-growing E. coli).


Both plasmids and bacteriophages put an upper limit of around 22kb on the size of inserts. One solution is the use of cosmid vectors. Cosmids are plasmids that contain the bacteriophage cos sites – they contain no more phage DNA than this. In vitro packaging extracts will package any molecule that has cos sites with 37-52kb between them.

The cosmid vector is packaged as a phage, then propagated as a plasmid once inside the cell – so there is no lysis of E. coli (and therefore no plaques formed). There is only replication of the cosmid within the cell.

The cosmid is made into a recombinant in much the same way as a plasmid: digestion with restriction enzymes and then ligation of the foreign insert. The cosmid is often a recombinant concatemer of many inserts and many cos sites. The lack of genes for phage packaging and lysis means more room is available for foreign DNA, so recombinant cosmids have an advantage over recombinant bacteriophage genomes.


Another type of vector is a phagemid (sometimes called a phasmid). This is a type of cloning vector developed as a hybrid of the filamentous phage, M13, and plasmids to produce a vector that can grow as a plasmid, and also be packaged as single stranded DNA in viral particles. Phagemids contain an origin of replication for double stranded replication, as well as an f1 ori to enable single stranded replication and packaging into phage particles.

Many commonly used plasmids contain an f1 ori and are thus phagemids. Similarly to a plasmid, a phagemid can be used to clone DNA fragments and be introduced into a bacterial host by a range of techniques (transformation, electroporation). However, infection of a bacterial host containing a phagemid with a 'helper' phage, for example VCSM13 or M13K07, provides the necessary viral components to enable single stranded DNA replication and packaging of the phagemid DNA into phage particles. These are secreted through the cell wall and released into the medium. Filamentous phage retard bacterial growth but, in contrast to lambda phage, are not generally lytic.

Artificial chromosomes:

Multiple artificial chromosomes may be used as vectors, including:

Bacterial artificial chromosomes (BACs) - to clone fragments typically between 150 and 350kb, but sometimes closer to 700kb. BACs are based on functional fertility (F) plasmids. F-plasmids are important because they contain partition genes that ensure their even distribution between daughter cells during binary fission. BACs are often used to carry DNA fragments used in genome sequencing by shotgun sequencing.

P1-derived artificial chromosomes (PACs) – to clone fragments between 100 and 300kb in length; typically 150kb in length. It is based on a bacteriophage (P1) genome and used to infect E. coli cells.

Yeast artificial chromosomes (YACs) - to clone fragments between 100 and 3000kb in length. These are constructed by breaking a circular plasmid into two linear fragments using restriction enzymes and then using the linearised plasmid as a template to which a gene of interest can be hybridised (using DNA ligase). YACs are constructed to contain telomeres, centromeres and origins of replication that are necessary for their preservation, and then they are transfected into yeast cells. YACs are preferable to BACs when expressing eukaryotic proteins that require post-translational modifications, although YACs have reduced stability to compared to BACs.

Recombinant selection procedures:

Once a vector has been used to transform a colony of potential hosts, various selection procedures are required to ascertain which individuals have been successfully transformed. These include:

1. Antibiotic resistance/sensitivity

2. LacZ complementation (‘blue-white selection’). The lacZ gene encodes an enzyme called β-galactosidase, the N-terminal end of which is incorporated into the cloning vector. The vector with this gene also carries regulatory sequences that bind lac repressors which keep the lacZ gene repressed in the absence of an inducer (IPTG). If the host has been engineered to express the C-terminal end of lacZ, then only when the plasmid is present will both parts be present – which will reconstitute beta-galactosidase activity. We can detect that activity by using an enzyme substrate (“X-gal”) that turns blue after catalysis. Here, cells that have taken up pUC19 (without any inserts) will form blue colonies in the presence of IPTG and X-gal; where the insert is present (i.e. in the case of recombinants) only white colonies will form.

3. Selection of lambda recombinants by spi. Although lysed plaques tell us that E. coli has been infected, we need to distinguish between those lysed by recombinant phages and those lysed by normal phages that simply self-ligated. Plaques produced from cloning experiments involving replacement vectors nearly always contain true recombinants. This is because if the arms re-ligated without the insert, the resulting genome would have been too small to package. However, there is a need to distinguish between the required insert and the ‘stuffer’ DNA.

Phage lambda cannot grow normally in E. coli cells that are lysogenic for the bacteriophage P2. This is due to the presence, within the lambda genome, of red and gam genes. Phages that are red+gam+ are sensitive to the P2 lysogenic repressor molecule (spi+). Spi selection means ‘sensitive to P2 inhibition’. If the genes are removed (red-gam-) then it can grow on a P2 lysogen. Some replacement vectors are designed to incorporate red and gam genes into the stuffer region - these regions are removed during cloning, and phage that have aquired an insert will therefore be able to grow on lawn of P2-lysogenic E. coli. So spi selection can only be used with replacement vectors, and not with insert vectors.

4. Selection of lambda recombinants by cl mutations. The formation of lysogens by phage lambda requires a phage-encoded cl repressor protein. Phages in which cl is inactivated are incapable of lysogeny. Lysogens may be distinguished from lytically infected cells by plaque morphology: plaques containing lysogenic cells appear turbid, in the mutant cl repressor only lysis can occur, resulting in clear plaques. So unlike the spi selection approach, the cl approach can be used with both insertion and replacement vectors.