Both prokaryotic and eukaryotic host-vector systems can be used in GM work. This section initially concentrates on prokaryotic host-vector systems with some details of eukaryotic ones being given later.

A prokaryotic vector should:

1. Be capable of autonomous replication independent of the main bacterial chromosome, i.e. possess an origin of replication (ori).

2. Be easy to isolate, i.e. small.

3. Be non-toxic to host cells.

4. Have space for foreign inserts.

5. Have unique restriction sites for common restriction enzymes.

6. Have convenient markers for selection of transformants, e.g. antibiotic resistance genes.

7. Be relaxed, i.e. multiple copies in a host cell.

Most prokaryotic vectors are based on :

1. Plasmids

2. Bacteriophages

3. Cosmids (artificial constructions)


Plasmids are:

Circular, autonomous molecules of DNA.

Found naturally in most bacterial (and some other) species.

Size: 1.5 - 300 kilobases.

Function: carry non-essential (dispensable) genes, e.g. antibiotic resistance, toxin production.

But "cryptic" plasmids have no known function!

Plasmids can be conjugative or non-conjugative (conjugation is generally not required in GM).

Plasmids can be mobilizable or non-mobilizable (non-mobilizable plasmids are preferred as they are less likely to "escape" from host cells).

Plasmids can be relaxed (multiple copies per host cell) or stringent (1-3 copies per host cell).

For GM work we want: small, relaxed, non-conjugative, non-mobilizable plasmids with good markers and unique restriction sites.













XhoI, EcoRI,

PvuII, HincII

Tetracycline resistance


HindIII, BamHI, SalI


Tetracycline resistance





Immunity to

colicin E1

Colicin E1 production




EcoRI, BamHI

Ampicillin resistance

Colicin E1 production

1st natural plasmid is stringent (not relaxed).

2nd natural plasmid has poor marker genes.

3rd natural plasmid is too large.

The perfect plasmid doesn't exist in nature!!!!

However, some early GM experiments were done with natural plasmids.

e.g. Morrow et al. (1974)

Used frog oocyte 5S ribosome DNA (multiple copies of this gene in the oocyte cell).

DNA cut with EcoRI restriction endonuclease.

Then inserted into pSC101 plasmid.

E. coli cells transformed and transfected.

Tetracycline-resistant clones/colonies selected on tetracycline agar (these cells would contain the plasmid but not necessarily the insert).

13 out of 55 clones were transformants and contained the insert - a good result.

Therefore, eukaryotic DNA could be propagated in prokaryotic cells.

A great breakthrough!!!!!

However, natural plasmids like pSC101 were found unsuitable for much other GM work for the reasons already given.


Construct artificial plasmids with the best features derived from different natural plasmids.

Here is an example:


An example of an artificial plasmid cloning vector.

Cleavage sites shown are for those restriction enzymes that cut the plasmid only once insertional inactivation (gene disruption). The exception is EcoRI !!!

pBR322 has been completely sequenced - every base is known.

pBR322 has over 30 known unique restriction  sites (for clarity only 4 are shown in the diagram but for a more detailed map click here).

Some of these sites are within the tetracycline (Tcr) or ampicillin resistance (Apr) genes insertional inactivation.

Unfortunately, using restriction sites, e.g. EcoRI, outside the 2 genes for insertion does not lead to insertional inactivation of ampicillin or tetracycline gene and, therefore, selection of desired transformants is more difficult, although not impossible.

But if cloning with BamHI enzyme, inserts lead to:

Apr Tcs transformants (because tetracycline resistance gene is inactivated by the presence of the insert).

3 types of bacterial host cell are possible after transformation and transfection:

1. Cells with no plasmid (not required)

2. Cells with plasmid but no insert (not required)

3. Cells with plasmid and insert (required)

Therefore, select transformants on ampicillin agar plate (selects out cells not containing the plasmid, i.e. type 1 above).

Also replica plate onto tetracycline agar. 

Colonies which grow on the ampicillin plate could be either type 2 or type 3 but those that also do not grow on the tetracycline plate are the transformants that are required . They can then be picked off from the ampicillin plate - being in the same position.

These colonies come from cells containing an insert at the BamHI site, i.e. these cells are transformants and contain plasmid and insert (i.e. type 3 above).

Click here for more detailed examples of cloning and selection using pBR322 plasmid.

There are improved derivatives of pBR322,

e.g. pBR325 - has an additional chloramphenicol resistance gene containing a unique EcoRI site. But it's a bit big. e.g. pBR328 - as pBR325, i.e. 3 antibiotic resistance genes but smaller (4.9 kb).

pBR plasmids have been used extensively but they require replica plating and negative selection (i.e. absence of growth).

Other plasmid vectors have since been constructed that use positive selection.

e.g. pUC plasmids

These contain several restriction sites within a -galactosidase [lac z] gene from the lac operon (actually only part of it to save space). The natural substrate for the enzyme -galactosidase is the disaccharide sugar, lactose.

These plasmids also contain an ampicillin resistance gene (Apr) for selection of host cells containing plasmid from those not containing plasmid.


Insertion of foreign DNA into this gene leads to loss of -galactosidase activity by insertional inactivation.

How can this be detected?

For a more detailed map of this type of plasmid (pUC18/19) click here.

e.g. pTR262 plasmid

Here the repressor protein coded for by the repressor gene 'switches off' the promoter. RNA polymerase cannot now bind to the promoter and expression of the tetracycline gene does not occur. The host cell is, therefore, tetracycline sensitive and will not grow on agar plates containing this antibiotic.

However, if an insert is added at the Hind III site, this will lead to insertional inactivation of the repressor gene and no repressor protein will be produced. This in turn will lead to expression of the tetracycline gene and the host cell will now be tetracycline resistant. Any colonies growing on tetracycline agar will be transformants containing both plasmid and insert.

This is another example of a positive selection vector and has the advantage over vectors like pBR322 that replica plating is not required. However, this particular plasmid has only one useable restriction site, which could be a disadvantage.

e.g. pET plasmids

pET plasmids are part of a powerful host-vector system for the cloning and expression of recombinant proteins in Escherichia coli. They use the strong promoter from T7 bacteriophage.

There are over 30 different types of pET vector with different characteristics and different marker genes. Over 10 different E. coli host strains can be used.

A target gene is inserted into a pET plasmid under the control of strong bacteriophage T7 transcription and translation signals. Expression is induced by providing a source of T7 RNA polymerase in the host cell. T7 RNA polymerase is so selective and active that almost all of the cell's metabolism is switched to expression of the target gene. After only a few hours, the desired product can comprise more than 50% of the total cell protein.




1. Gene expression, i.e. production of gene products.

e.g. The first mammalian product produced in bacteria was somatostatin (1977). This was done using pBR322 and E. coli. Other useful pharmaceutical products such as human insulin can also be synthesized using plasmid vectors. Click here for more details.

e.g. Gene expression using the pET system.

2. Gene cloning.

3. Cloning of complementary DNA (cDNA), i.e. DNA synthesized using mRNA as a template and reverse transcriptase.

Cloned DNA can be used for base sequencing and for probes.


1. As size of insert increases fall in transformation frequency.

2. Plasmids with large inserts are often unstable deletion of insert during growth.

Inserts >10 kb are generally impractical.


e.g. lambda (l) phage. A bacterial virus that infects E. coli.

phage1.gif (21345 bytes)             

Well understood life cycle (lytic & lysogenic phases).

Phage particle adsorbs to bacterial cell wall and DNA in head is injected via tail. In the lytic cycle DNA then circularizes and replicates. More phage particles are made and host cell then lyses to release them. Virus infection spreads to other bacterial cells plaque (clearing in a bacterial lawn).

phage2.gif (68567 bytes)


In lysogeny the phage DNA is incorporated into the main bacterial chromosome. This phase is not required for GM.

Genome size = 49 kb.

Mutants must be 78-105% of this for the lytic cycle.

Central region controls lysogeny. This can be deleted for GM work and inserts added.

cos (cohesive, sticky) ends circularization.


(showing genes associated with different regions)

cos _________________________cos

A ....Head ...| ..Tail ...| Lysogeny | Regulation | ...Lysis... R



Wild type l has multiple restriction sites and is, therefore, no use for GM (except as a standard for gel electrophoresis).

Several types of artificial l are now available which have:

1. 1 or 2 restriction sites only.

2. Deletions to allow room for inserts.

3. Selection system(s).

With replacement vectors the DNA is cut twice and piece removed allowing space for insert. Insertion vectors have a deletion which leaves space for an insert. Up to 20 kb size inserts are possible.

e.g. Charon 16A

Selection is on Xgal plates.

Blue plaques - phage DNA but no insert

White plaques - phage DNA with insert (lac z gene has been disrupted)

This type of vector is named after a boatman in Greek mythology who ferried the souls of the dead across the river Styx into the underworld (Hades or Hell). To find out more about Charon click here. Also, click on the following thumbnail image to see a painting of Charon in his boat with some dead souls and with Cerberus, the three-headed dog, waiting on the other side of the river:

Charon_painting.jpg (748859 bytes)


Very useful feature of the l system.

With in vivo packaging:

Transduction using l 105 plaques per g DNA (transduction frequency).

But after manipulation this can fall 103 - 104 plaques per g DNA.

With in vitro packaging:

Mix together:

phage DNA (with appropriate cos ends and insert)

+ head protein

+ tail protein

+ packaging enzymes

complete phage particles

These artificially produced phages are then able to infect host cells.

In vitro packaging usually gives a higher transduction frequency (i.e. number of plaques produced per g DNA) than in vivo packaging.


1. Large inserts possible (up to 20 kb).

2. Stable propagation (C.F. plasmids where large inserts may get deleted).

3. Efficient entry into cell using in vitro packaging.

4. Multiple copies of insert and strong promoters.


1. Plaques are produced not colonies.

Useful for cloning DNA but no use for gene expression, e.g. production of gene products in biotechnology.

2. 20 kb is still not large enough for many eukaryotic genes. For larger inserts cosmids or eukaryotic vectors can be used.


1. For cloning DNA for further analysis.

2. For use to store DNA in gene and cDNA libraries (random collection of fragments of all the genome/mRNA of a species in a host-vector system).


Most l vectors are disabled and will only grow in special host cell strains. e.g. amber (nonsense) mutations present in essential genes. Only host bacteria with a nonsense suppressor can support growth of the phage vector. Non-laboratory strains do not have the suppressor gene and, therefore, will not support phage growth.

Another phage vector:


M13 is a small filamentous E. coli bacteriophage cloning vector which behaves like a plasmid for part of its replication cycle

It is rather unusual in that  it produces both double-stranded (dsDNA) and single-stranded DNA (ssDNA) in different phases of its replication cycle. When in replicative form (RF) inside the host cell, M13 acts rather like a plasmid, the DNA being double stranded (ds), but, when packaged up in capsids to form phage particles to be released from the host cell, the DNA is single stranded.

M13 (unlike lambda phage in its lytic cycle) does not cause lysis and death of the E. coli host cells. Instead, infected host cells excrete M13 phage particles containing ssDNA in large numbers.

But can only take relatively small inserts up to 0.5kb in size.


Probes and primers are always single stranded.


And another phage vector:


Vectors based on this bacteriophage can take larger inserts than lambda vectors shoehorning up to 110kb of DNA into the capsid.



Artificial constructions.

Consist of the cos ends of phage with plasmid DNA.

In vitro packaging is used to introduce the DNA into a host cell.

Any DNA between 2 cos sites will be packaged in the phage head. Phage injects the DNA which then behaves like a plasmid once inside the host cell.

But such "phages" are not virulent and do not cause lysis or produce plaques. Colonies are produced. Therefore gene expression and production is possible.

Inserts up to 40 kb can be inserted.

Selection can be by adding ampicillin resistance gene to insert - positive selection on ampicillin plates.

However, cosmids can be unstable.



Artificial constructions of plasmid and l attachment site. The plasmid can insert into the l phage genome. Reversal of the process releases the plasmid vector. Phasmids can exist in bacterial cells as a plasmid or as a phage.


This type of vector is based on the F (fertility) plasmid involved in conjugation in some bacterial species. BACs can take inserts up to 300kb in size.


This type of vector combines features of P1 phage vectors with BACs and, like the latter, can take inserts up to 300kb in size.


Several types:

e.g. Yeast episomal plasmids (2 m)

YIP's (yeast integrative plasmids - bacterial plasmids which integrate into yeast chromosome)

YRP's (yeast replicative plasmids - include an ori and can replicate independently of the main yeast chromosome)

YAC's (yeast artificial chromosomes - can accept very large inserts (e.g. 2000kb) and become more stable the larger the insert - unlike virtually every other type of vector!)

"Shuttle vectors" can replicate in two types of cell, e.g. both yeast and bacterial cells. They can ferry DNA between the two.

Yeasts are eukaryotic and, therefore, often better than prokaryotic vectors for the expression of eukaryotic genes. Factors that should be considered in the choice of a host-vector system for gene expression (i.e. if a product is required) and especially if a eukaryotic gene is inserted into a prokaryotic host-vector system include:

Some potential problems in achieving successful gene expression are highlighted by the early example of somatostatin synthesis.

You should look up further details about gene expression and production in one of the recommended textbooks. 


Used to introduce foreign DNA into animal cells,

e.g. SV40 (simian virus), adenoviruses, papillomaviruses, poxviruses, and baculoviruses (for insect cells). The latter have the advantage over mammalian systems that the host cells, being insect, are not so fastidious about temperature and, therefore, are easier to culture. Also, insect cell lines do not require such expensive and highly accurate incubators as mammalian cell lines, making their culture cheaper.


e.g. Ti plasmids derived from a plasmid found naturally in the bacterium Agrobacterium tumefaciens which produces galls on some plants. Insertion of genes into this plasmid can be used to alter the genetic make-up of the host plant to improve its characteristics for agriculture.

e.g. Most plant viruses contain RNA as their genetic material. RNA cannot be manipulated directly and, therefore, this makes plant viruses poor vectors for GM work. However, a few plant viruses contain DNA which can be manipulated in an attempt to introduce genetic modifications into the host plant itself. An example of a DNA-containing plant virus is cauliflower mosaic virus.


For more details of these vectors, see your textbook (Brown or Primrose et al.). 



































Replica plating

This technique entails making an exact copy, after growth and incubation, of the colony pattern on the first agar plate on another plate. This used to be done by using a wooden block covered in sterile velvet to pick up a small amount of the colonies on the first plate and then pressing this gently onto the second plate which was then incubated to give an exact copy of the colony pattern of the first plate (less any colonies that were inhibited by the agar of the second plate). Nowadays a sterile membrane is used to transfer the colonies from the first to the second plate.























With vectors carrying antibiotic resistance genes, e.g. pBR plasmid vectors, insertion of DNA into a gene will disrupt it (insertional inactivation).

Selection is "negative" in that the desired transformants/recombinants will not grow on agar containing the appropriate antibiotic. However, the desired clones can be obtained by replica plating at the same time onto another agar.


A. Using a restriction enzyme that cuts within the tetracycline resistance gene of pBR322, e.g. BamHI.

Selection is first done on ampicillin agar (only possibilities 2 and 3 below will grow). Replica plating onto tetracycline agar will enable identification of transformants containing plasmid and insert (possibility 3) because these will not grow but can be obtained off the original ampicillin plate. The absence of colonies on the tetracycline plate will indicate the position of the desired colonies/clones on the original ampicillin plate.


Growth on ampicillin agar

Growth on tetracycline agar

1. No plasmid



2. Plasmid alone (no insert)



3. Plasmid + insert



B. Using a restriction enzyme that cuts within the ampicillin resistance gene of pBR322, e.g. PstI.

Selection is first done on tetracycline agar (only possibilities 2 and 3 below will grow). Replica plating onto ampicillin agar will enable identification of transformants containing plasmid and insert (possibility 3) because these will not grow but can be obtained off the original tetracycline plate. The absence of colonies on the ampicillin plate will indicate the position of the desired colonies/clones on the original tetracycline plate.


Growth on tetracycline agar

Growth on ampicillin agar

1. No plasmid



2. Plasmid alone (no insert)



3. Plasmid + insert































Using agar plates containing Xgal (5-bromo-4-chloro-3-indolyl--D-galactoside) or ONPG (2-nitrophenyl--D-galactopyranoside) and the antibiotic, ampicillin, coloured colonies.

The natural substrate for the enzyme -galactosidase is the disaccharide sugar, lactose, but degradation of this does not produce a coloured product and this makes it difficult to detect. Using artificial substrates like Xgal and ONPG has the advantage that when these are degraded by -galactosidase the product is coloured (blue and yellow respectively) and easily detectable.

On Xgal plates:

Blue colonies - plasmid present but no insert (not required)

White colonies - plasmid present with insert (required)

Why ampicillin??

Ampicillin selects out those cells with no plasmid since they will be sensitive to the antibiotic and unable to grow.

















In 1977 (Itakura et al.) somatostatin - a small human peptide only 14 amino acids in size - was synthesized using E. coli. This was the first example of a human product made by bacteria and also the first use of a synthesized gene - the somatostatin gene used was a chemically synthesized DNA sequence. Thus two breakthroughs in one go!

How was the DNA sequence of the somatostatin gene determined?

In this case, by first determining the amino acid sequence of the gene product and then working  'backwards' to extrapolate the original DNA codon sequence. This is the reverse of what happens in nature where the DNA codon sequence is transcribed into mRNA and then translated into an amino acid sequence. However, remember that the genetic code is redundant and that more than one codon can code for a single amino acid. Therefore, determination of the actual gene sequence is not that straightforward and may require some 'presumption' (what non-scientists call 'guesswork') and further testing. Somatostatin was a good example to choose for this pioneer work because it consists of only 14 amino acids and, therefore, is coded for by a very short gene of only 14 codons or 42 nucleotides (14 x 3 = 42).

However, a gene is not sufficient on its own for synthesis of a product. For gene expression to occur various other factors have to be taken into account, e.g. a promoter must be present in the correct position relative to the gene sequence.


1. The promoter sequence from the lac operon was inserted into pBR322 plasmid by cutting, insertion and ligation using ligase enzyme.

2. Two superfluous EcoRI sites were then removed (to prevent circle falling apart when this restriction enzyme was used in the next stage). This produced a smaller plasmid called pBH 20.

3. The chemically synthesized somatostatin gene was then inserted into the plasmid by removing a small fragment using two restriction enzymes (EcoRI and BamHI) followed by ligation. The plasmid now looked like this:

4. E. coli cells were then transformed and transfected with the plasmid. But on growth, no somatostatin production was detected! The reasons for this failure were not immediately obvious.

In E. coli small peptides are degraded. Somatostatin was actually being synthesized by the host cells but they contained enzymes which were then digesting the product .

5. Plasmid was then reconstructed to replace the lac operon with a larger fragment containing the promoter and most of the lac z (-galactosidase) gene. This made the product larger and, therefore, not prone to degradation by host cell enzymes. The reading frame now needed correction.

6. On transformation and transfection of E. coli, a "fused" product was produced which consisted of part of the -galactosidase product attached to the somatostatin peptide.

7. The fused product was then separated into its two component parts using cyanogen bromide which cuts the amino acid sequence at a methionine between the somatostatin and -galactosidase chains.


Human insulin can also be synthesized using a pBR322-like plasmid and a similar method to this involving fused products.