The second step in molecular cloning is to join the passenger DNA to the DNA of a suitable cloning vehicle. These vehicles (or vectors) have the property that they replicate themselves and any attached passenger DNA so that the passenger is amplified and can be eventually isolated. A number of different vectors have been developed for genetic engineering. Each has special distinguishing properties.
Plasmids are relatively small, double-stranded, closed-circular DNA molecules that exist apart from the chromosomes of their hosts. They are a kind of molecular parasite and are present in a wide variety of bacterial and fungal species. Naturally occurring plasmids carry one or more genes. For example, some plasmids carry genes which confer resistance to certain antibiotics. Some plasmids may bear genes that code for the restriction and modification enzymes that were discussed previously. Some may carry genes that direct the synthesis of enzymes that aid in the production of bacterial poisons or antibiotics.
However, from the viewpoint of the genetic engineer, the most important property of plasmids is that they bear a special region of DNA called an origin of replication, or more simply an origin. This region allows the plasmid to multiply within and semi-independently of its host. It confers on the plasmid the property that if a foreign DNA segment is inserted into it, the plasmid doesn't much care. It will replicate its own DNA as well as any passenger DNA that may be attached to it, producing many copies of the recombinant molecule.
In a typical cloning experiment, the circular plasmid DNA is cut once by treating it with a restriction endonuclease. This converts the circular molecule into a linear one. Then, a foreign DNA fragment adds on to the ends of the vector with the help of the enzyme DNA ligase. The ligase creates a circular molecule containing both the plasmid and its passenger. The next step is transformation -- the introduction of the recombinant DNA molecule into a single bacterium. Once inside, the plasmid produces many copies of itself as the bacteria themselves grow and reproduce.
Plasmid replication
Many of the plasmids that are commonly utilized in recombinant
technology are derivatives of pMB1, a plasmid that was originally
isolated from bacteria that had infected a human patient. pMB1 is an
example of a relaxed replicating plasmid. Relaxed in this case
refers not to a state of mind, but to the regulation of the number of
copies of the plasmid in the cell. Plasmids under relaxed control are
commonly maintained in bacteria in large numbers -- sometimes
hundreds of copies. By contrast, there are normally only one or a few
copies of stringently controlled plasmids in a cell.
Desirable properties of plasmids
Plasmids have proven extremely useful as cloning vectors,
particularly when they are modified specifically for that purpose. A
bacterial plasmid that is designed for cloning should have some or
all of the following properties:
The aim of most cloning experiments is to isolate the passenger DNA. The vector only serves to carry and amplify the passenger. A small plasmid has the advantage of contributing only a minimal amount of extraneous DNA to the plasmid/passenger construct, thereby making it easier to prepare large amounts of the passenger DNA. In addition, it is easier to get smaller pieces of DNA into bacteria than larger ones -- everything else being equal -- and the smaller the plasmid is, the less it contributes to total size. Another point is that small plasmids replicate faster and require less energy for replication than large ones. Finally, small plasmids are easier to purify than large ones because they are less fragile.
Knowledge of the DNA sequence of a plasmid allows the genetic engineer to manipulate it with the full complement of recombinant DNA techniques.
In other words, relaxed replication plasmids are most often preferred to those that are under stringent control. In most cases, having many copies makes it easier to purify the plasmid away from the chromosomal DNA, and of course increases its yield. However, there are circumstances when just one or a few copies of a plasmid are desirable. Can you think of one?
Bacterial cells that harbor plasmids don't necessarily look or act different than those that do not. It is only when the plasmid carries a gene that lends a special trait to the bacteria that the molecular engineer can tell if a bacterium contains a plasmid. Antibiotic resistance is one such trait or marker, and the genes for ampicillin resistance and tetracycline resistance are common genes carried by frequently used plasmids. Some plasmids are also readily lost from cells, and unless there is a selectable gene contributed by the plasmid and the bacteria are kept under selective conditions, the molecular engineer may end up with a plasmid-less population of cells at the end of an experiment.
Imagine the following scenario. A scientist is trying to insert
a passenger molecule into a plasmid. The plasmid has only a single
selectable marker. After ligation, some of the molecules contain
the passenger. Others do not. These may be plasmids that have
recircularized in the presence of DNA ligase. This mixture of
plasmids is then used to transform bacteria. Both recombinant
plasmids and recircularized plasmids contain the same marker and
cannot readily be told apart genetically. How are colonies
carrying a plasmid with a passenger distinguished from those with
a plasmid with no passenger? One way is to grow many individual
colonies, isolate plasmid DNA from each, run the DNA out on an
agarose gel, and identify the recombinant plasmid molecules
because they are bigger by virtue of the passenger that they
carry. While this procedure works and is very commonly carried
out, it is laborious, time consuming, and expensive. 
There is a better way. Suppose that a plasmid carries two different antibiotic resistance genes, A and B. What would be the consequences of cloning the passenger into the middle of gene B? In most cases, the foreign DNA will disrupt gene B and not allow it to work. On the other hand, if the plasmid simply recircularizes, gene B will be unaffected. With this arrangement it is possible to tell which cells harbor plasmids that carry passengers. Those that contain any plasmid at all will have gene A activity. And those that have a plasmid bearing a passenger, will lack activity from gene B.
Why unique? As described above, most cloning is done by inserting a fragment of DNA that is cut with a specific restriction endonuclease into a vector that is cut with the same enzyme (remember, this ensures that the two have compatible ends). If the vector contains more than one such site, it will be cut into multiple pieces by the restriction enzyme, complicating matters unnecessarily. The presence of many unique sites allows for maximum flexibility and ease in cloning.
Plasmid purification
Plasmids may be purified from bacteria by taking advantage of the
difference between the small, circular plasmid molecules and the
large, broken (hence linear) pieces of chromosomal DNA. (Chromosomal
DNA from E. coli is also circular, but during isolation its
relatively large size (six million nucleotide pairs) and consequent
fragility cause it to break easily.)
The most common method for purifying plasmid DNA involves three steps.
Upon denaturation, the two circular single-stranded chains of the plasmid DNA remain entwined and don't separate fully. When conditions are set up so that renaturation can occur, each strand rapidly finds its complement. The chromosomal DNA, on the other hand, breaks readily and therefore consists of noncircular pieces. Under these circumstances, the two strands easily denature and separate. Upon renaturation, they have difficulty finding complete copies of their complements. Often partial renaturation occurs between different sizes of single-stranded fragments and large insoluble aggregates form. These can be simply separated from the small circular plasmid DNA by high-speed centrifugation.
Some popular plasmids
The first really useful plasmid for genetic engineering, pBR322,
was pieced together by Francisco Bolivar, and others in Herbert
Boyer's laboratory in the 1970s (The "B" stands for Bolivar and the
"R" for Rodriguez, another scientist in Boyer's laboratory). What
makes pBR322 useful is that it contains an ampicillin resistance gene
and a tetracycline resistance gene. In addition it has a relaxed
origin of replication (indicated by the green arrow shown at the
bottom of the illustration) and accumulates to high numbers in E.
coli. Its entire 4363 base-pair sequence has been determined, and
21 common enzymes are available that recognize only a single site
within it. (However, only 11 are in either of the two antibiotic
resistance genes.).
More recently, a series of small plasmids (about 2.7 kilobase pairs) have been developed in Joachim Messing'e laboratory that have several properties that have made them very popular with genetic engineers. These pUC (pronounced PUCK) plasmids, exemplified by pUC18 pictured at the right, carry an ampicillin resistance gene and an origin of replication, both from pBR322. They also bear a multiple cloning site -- a sequence of DNA that carries many restriction sites (13, in the case of pUC18). The multiple cloning site of the pUC plasmids is special because it also codes for a small peptide. This peptide will correct a specific mutation in the chromosomal gene that codes for the enzyme beta-galactosidase.
When plasmids containing this sequence are cloned into a specific
E. coli strain that lacks beta-galactosidase activity, they
make the peptide and thereby begin to express active enzyme. If,
however, one of the restriction endonuclease sites in the multiple
cloning site is opened and a foreign gene inserted, the peptide is no
longer produced (because the protein coding region is disturbed), and
no beta-galactosidase activity appears.
Cells that harbor an active beta-galactosidase enzyme can be made to turn blue in the presence of certain substrates. Those colonies that have a passenger inserted at the multiple cloning site of the pUC plasmids will lack the enzyme and will be white, while those that have simply recircularized (those that don't contain a passenger) will stain blue. In this way, pUC plasmids containing a foreign insert of DNA can be distinguished from plasmids without a passenger.
Lambda is a temperate bacteriophage with a genome size of about
48.5 kilobase pairs. 
Its
entire DNA sequence is known. In phage particles, the lambda genome
exists as a linear, double-stranded molecule with single-stranded,
complementary ends. These ends can hybridize with each other (and do
so when the DNA is within an infected cell) and are thus termed
cohesive. They are similar to, but longer than, the sticky ends that
were encountered when we discussed restriction enzymes.
Lysis and lysogeny
Bacteriophage lambda can enjoy one of two life styles -- that's
what's meant by the word "temperate". It can enter the lytic
cycle, replicate many times, produce more phage, and destroy
its bacterial host. Alternatively, it can take up lysogenic
growth, meaning that it integrates its DNA into the bacterial
chromosome.
The lysogenic state is highly stable, but not permanent. Although the viral DNA may be passed through hundreds of generations of bacteria while integrated into the chromosome, a number of influences can cause "induction" and a return to the lytic cycle. We're going to ignore the tendency of lambda phage to engage in the lysogenic state.
Lambda is assayed by plating phage particles on a "lawn" of bacterial colonies spread on plates so that they form a continuous group of cells. If a lambda phage infects one of the cell, it will multiply quickly and lyse that cell, releasing lambda phage. These in turn will infect adjacent cells and also cause lysis. After a while, a round hole -- technically, a plaque -- is seen in the lawn, representing the breakage of many bacterial cells and also marking a single original event: the entrance of a single virus into a single cell.
Shown below is the genetic map of phage lambda (not all the genes are included) and a scale indicating the extent of various regions, expressed as a percentage of the total genome. Notice that the genes that form the proteins of the head are gathered together toward the extreme left end of the DNA, and the tail-forming genes follow. Similarly, the genes for DNA synthesis (O and P) and those that cause bacterial lysis (S, R, and Rz) are clustered near the right end of the lambda chromosome. In the middle of the chromosome -- between the J and int genes -- is a large region whose function is unclear but apparently unessential. Also pictured at the ends of the chromosome are the two single-stranded cohesive ends.
Three reasons why bacteriophage lambda is a good cloning
vehicle
However -- and this is very important -- there is a strict size requirement for the piece of DNA that goes into the head. That is, if the distance between successive cos sites is either too long or too short (longer than about 105% or shorter than about 78% of a wild-type phage's DNA), the resultant phage will have markedly decreased viability. For example, if a piece of phage DNA of 36 kb (about 75% of 48.5 kb) is packaged, it will fail to yield significant numbers of active phage. The same thing will happen if the distance between two successive cos sites is 53.4kb, approximately 110% greater than the normal size of a lambda chromosome.
The molecular engineer can take advantage of this situation to distinguish among recombinant molecules that carry foreign DNA and those that do not. The essential idea is to begin with a vector that carries a piece of useless DNA (a so-called stuffer fragment) that can be excised with the aid of one or more restriction endonucleases. The stuffer fragment is specially designed so that when it is removed, the spacing between the cos sites will be too short for successful packaging. If a passenger is not substituted in place of the stuffer, no infective phage particles will be produced.
Phage that have a stuffer fragment are called substitution vectors because they are designed to have a piece removed and substituted with something else.
Large numbers of different lambda strains have been created that allow efficient cloning of a variety of foreign DNA's. The various strains are designed to have differing amounts of DNA removed, and they contain a variety of restriction enzyme sites for cloning. In addition, some lambda strains have a stuffer fragment that carries the beta-galactosidase gene. When it is removed or when foreign DNA is cloned within the gene, beta-galactosidase activity may be abolished. The accompanying loss of activity may be used to select recombinant clones.
Why lambda?
Why use bacteriophage lambda as a vector rather than plasmids?
Large pieces of DNA (up to about 20 kilobase pairs) can be easily
cloned in bacteriophage lambda substitution vectors. Plasmid vectors
are less useful for cloning big passengers.
But why clone large pieces of DNA in the first place? One obvious reason is that some genes are very big and it is advantageous to have them all in one piece. Interestingly, often it is not the protein-coding portion of these large genes that contributes to their size. Rather, the regulatory regions of several genes are dispersed tens of thousands of base pairs away from the structural gene.
Another reason for cloning in lambda is the efficiency it offers in DNA transformation. In general, plasmids are much less efficiently moved into bacterial cells than bacteriophage, which, after all, were designed by Nature for their ability to project their chromosome into their hosts.
