BCH 5425
Biotechnology and Molecular Biology
Dr. Michael Blaber
DNA Replication
1953
The DNA model of Watson and Crick suggested how genetic
information might be replicated: either strand of
the duplex can be used as a template to replicate
the sequence information. But, was the replication conservative
(i.e. the original parental strands remain together after replication)
or semi-conservative (one parental strand pairs with one
newly synthesized strand)?
The answer for prokaryotic organisms (i.e. lack a true membrane
bound nucleus and cellular organelles; e.g. bacteria) came from
the 1958 experiment of Meselson and Stahl.
1958
The nitrogen in ammonium salts in culture broth is incorporated
into DNA bases. The most common isotope of nitrogen is 14N.
However, 15N ammonium salts (a heavier isotope) can
also be obtained.
DNA from E. coli cells grown with 15N ammonium
salts will have a higher density than DNA grown in "normal"
(14N) ammonium salts. Such DNA will migrate differently
on cesium chloride (CsCl2) equilibrium density gradient
centrifugation. The more dense DNA will migrate as a lower
band (on this type of centrifugation the characteristic migration
position is a function of density, and is independent of DNA length).
Meselson and Stahl reasoned that if they grew E. coli in
15N salts then switched media to 14N salts
for additional rounds of replication, the mode of replication
could be deduced from the density of the DNA.
After switching to the 14N media and allowing the cells to go through a round of replication a single band of intermediate density was observed (i.e. between 14N and 15N control DNA samples).
After a second round of replication in 14N media two bands were present in approximately equimolar amounts; one was intermediate in density and the other migrated as purely 14N labeled DNA.
The results were consistent with a semi-conservative mode of
replication for DNA. Evidence of semi-conservative replication
of DNA has since been obtained with both plant and animal DNA.
DNA Replication in E. coli
E. coli DNA polymerase characteristics:
E. coli genome is circular duplex DNA of approximately
4 x 106 base pairs (i.e. 4 Mb). The genome has a single
origin of replication. DNA duplication in E. coli begins
at a specific site in the DNA called "oriC".
OriC is a region of DNA approximately 240 nucleotides long. It
contains repetitive 9-base pair and 13-base pair sequences (known
as the '9-mer' and '13-mer' regions). These sequences are AT rich
regions, which melt at lower temperatures than DNA containing
GC pairs. These regions are postulated to help melt the DNA duplex
in the oriC region for initiation of DNA replication.
The dnaA gene product: (dnaA protein)
Strains of E. coli with mutations in the dnaA gene were
able to grow at 30 °C, but not at 39-42 °C. However,
if DNA synthesis was begun at 30 °C, and then
the temperature was shifted to 42 °C, DNA synthesis continued
until the genome was replicated (and the cell divided), but no
new initiationof DNA synthesis was possible.
Conclusion: Somehow the product of the dnaA gene (i.e. the dnaA
protein) is required for initiation of DNA synthesis.
Studies of purified dnaA protein: dnaA protein binds to the '9-mer'
region in oriC and forming a multimeric complex with 10-20 protein
subunits (i.e. at a single oriC region there will be bound 10-20
dnaA protein molecules). Binding requires ATP. Further addition
of ATP was observed to result in a melting and opening up of the
DNA duplex in the oriC region. This was determined by addition
of S1 nuclease (like mung bean, but will also cut DNA at the site
of an internal nick), which resulted in cleavage of DNA at the
site of oriC.
The dnaB gene product: (dnaB protein)
The protein encoded by the dnaB gene appears to be essential
for DNA replication. The dnaB protein has been identified as a
helicase. A helicase moves along a DNA strand opening up
the duplex to melt and separate the DNA strands. It binds to the
single stranded DNA in the general region of the oriC DNA segment.
Binding requires ATP as well as the dnaC gene product (the dnaC
protein). After helicase/dnaC binds to the DNA, the dnaC protein
is released. Two helicases bind at the oriC region, one helicase
on each strand of the DNA. This stage represents the prepriming
complex:
Separated strands in the oriC region are prevented from reannealing
by the binding of single-stranded binding protein (ssb protein).
The dnaG gene protein:
The dnaG gene protein is called primase. Primase catalyzes
synthesis of short RNA molecules that function as primers
for DNA synthesis by E. coli DNA polymerase III
(pol III). Primase binds to dnaB protein at oriC and forms a primosome.
The primase within the primosome complex provides RNA primers
for synthesis of both strands of duplex DNA. Primase
lays down tracks of pppAC(N)7-10 (RNA).
After synthesis of the 9-12 mer RNA primer, DNA Pol III holoenzyme
enters the replication fork and is able to utilize the RNA as
a primer for DNA synthesis. As the replication fork opens up,
the leading strand synthesis can continue, but a gap develops
in the lagging strand:
DNA Pol III is a large multicomplex enzyme (holoenzyme) which
is somewhat dimeric in nature (there are two polymerase active
sites). The two active polymerase sites in Pol III could actually
function to synthesize both nacent strands at the fork. However,
the synthesis of the lagging template strand would be in the opposited
direction to the movement of the Pol III complex:
Primase can bind to the Pol III complex, but the arrangement of
the DNA strand as it passes through the Pol III/primase complex
is quite unique. It forms a loop structure such that primase and
the Pol III active site can accomplish discontinuous synthesis
of the lagging template strand even though the general direction
of the Pol III complex is opposite to the require direction of
DNA synthesis:
After primase makes another primer on the lagging template, the
adjacent Pol III active site can extend the primer (incorporating
dNTP's) by utilizing the same loop structure and feeding the template
through in the direction shown.
The lagging strand loop cannot be fed through the Pol III complex
forever, and after a nascent DNA strand is synthesized the loop
is released and a new one is formed using the opened template
DNA further up the fork:
As synthesis continues in this fashion, it results in a single
continuous DNA strand on the leading strand but a series
of short fragments on the lagging strand, containing both
RNA and DNA, called Okazaki fragments:
How are these RNA/DNA fragments converted into one long continuous
DNA strand? The RNA could be removed by a polymerase which has
5'->3' exonuclease activity, however, Pol III lacks this activity.
DNA Pol I does have 5'->3' exonuclease activity it can extend
the DNA synthesis via nick-translation. The nick-translation
activity restults in degradation of the RNA primers. The
end result is a series of "nicks" in the lagging strand,
now 100% DNA:
DNA Pol I leaves and DNA ligase then joins these discontinuous
DNA fragments to form a continuous DNA duplex on the lagging strand.
Summary of steps in E. coli DNA synthesis:
Notes from above:
Properties of E. coli polymerases (Pol I, II and III)
| 5'3' Polymerase activity | |||
| 3'5' Exonuclease activity(proof reading) | |||
| 5'3' Exonuclease activity(nick translation) | |||
| Synthesis from: | |||
| Duplex DNA | |||
| Primed single strand | |||
| Primed single strand plus ssb protein | |||
| Chain elongation rate(in vitro) bp/min | |||
| Molecules/cell | |||
| Mutation Lethal? |
Functions:
Pol III Structure and function
A "holoenzyme" complex of 10 different polypeptides,
resultant molecular weight is greater than 600 KDa (i.e. it is
a large complex).
It is structurally an asymmetric dimer - it contains two copies
of most of the polypeptides which comprise it, including two catalytic
sites for nucleotide addition (i.e. polymerization).
The various protein subunits have a variety of functions:
Termination of DNA replication
Specific termination sites of DNA replication exist in E. coli.
Termination involves the binding of the tus gene product (tus
protein). This protein may act to prevent helicase from unwinding
DNA (will therefore halt pol III and pol I action).
DNA replication produces two interlocking rings which must be
separated. This is accomplished via the enzyme topoisomerase.
E.coli cell division
E. coli cells double roughly every 30 minutes (depending
on strain, media and temperature).
Chromosomal movement requires protein synthesis. The E. coli
cell membrane appears to play an important role in chromosomal
segragation. Chromosomal attachment to the inner membrane may
play a role in chromosomal segregation.
Prior to cell division:
Signals for partitioning are not near the oriC region, because oriC plasmids are segregated randomly into daughter cells. Bacteria have no centromeres or microtubules, as are used by eukaryotic cells to segregate chromosomes. It is know that
Eukaryotic DNA replication
Lodish (208-210; 378-381)
Efforts to understand eukaryotic DNA replication have focused
on the two following areas:
SV40 virus (Simian virus 40)
An animal virus accidentally discovered in kidney cell cultures from wild monkeys used in the production of poliovirus vaccines. It is a class I DNA virus - its genome is a single molecule of circular double stranded DNA of 5.2 Kb. It replicates in the host cell nucleus. It uses host enzymes for viral DNA and mRNA synthesis. It has been utilized as a model system to understand eukaryotic DNA replication.
SV40 can now be replicated in vitro using 8 purified
components from mammalian cells. The functions of these proteins
are similar in nature to proteins required for replication
of E. coli DNA. The mechanistic problems involved in DNA
replication are similar in all organisms.
Replication of the SV40 chromosome is initiated at a unique
location on the SV40 DNA by interaction of 'T antigen'
(coded for by the virus). T antigen has helicase activity
and unwinds a region of DNA at the SV40 replication origin. The
opening of the SV40 chromosome at the replication origin requires
the following:
RFA is a single strand DNA binding protein (the function
is similar to the Ssb protein of E. coli).
Eukaryotic DNA replication occurs bidirectionally from
RNA primers made by a primase (similar to prokaryotes).
Leading strand synthesis is continuous, while lagging strand synthesis
is discontinuous (again, like prokaryotes).
However, there is a difference in the replication mechanism of
eukaryotes: In eukaryotes there appears to be different polymerases
associated with leading (Pol d) and
lagging (Pol a) strand synthesis.
Pol a is closely associated
with a primase (for lagging strand synthesis).
Sequence of events in SV40 replication:
1. Binding of T antigen (helicase) to SV40 DNA replication origin
region
2. T antigen helicase activity opens up DNA duplex. Replication
factor A (RFA), which is a single strand DNA binding protein binds
to the individual DNA strands and prevent reannealing
NOTE: RFC is subsequently going to bind to the Pol a/Primase
complex and allow DNA extension of RNA primer by Pol a.
One major difference between Prokaryotic and Eukaryotic DNA replication
mechanisms is that, while Prokaryotic replication utilizes the
same polymerase (Pol III) for both leading and lagging strand
synthesis, Eukaryotes utilize two different polymerases (Pol a
for lagging strand, and Pol d for leading
strand synthesis).
Bacterial chromosomes are circular, whereas eukaryotic
chromosomes are linear and have ends called telomeres.
The ends of linear chromosomes present a problem in DNA replication
with regard to the synthesis of the Lagging strand. The
end result of the synthesis is an RNA primer at incomplete replication
at the 5' end of each daughter duplex:
Since polymerase adds bases in the 5'->3' direction only, we
cannot get complete duplication of the 5' end of the daughter
strand. Therefore, with each round of replication we will lose
information at the ends of the chromosome (i.e. the Telomeres).
Telomeres consist of repetitive oligomeric sequences (e.g. Yeast
telomeres are comprised of the repeating sequence 5' -(G)1-3T-3'.
Telomerase can elongate the lagging strand template
from its 3' hydroxyl end (i.e. in a 5'->3' direction).
Telomerase- a polymerase (a reverse transcriptase)
which carries a piece of RNA. This RNA hybridizes to
the repetitive sequence at the telomeric end of the lagging strand
template:
The telomerase then extends the 3' telomeric region of the DNA
template to which it has hybridized:
Through a translocation mechanism the Telomerase and its associated
RNA can release and rebind to the newly extended telomere to repeat
the extension cycle:
Pol a/Primase complex can now repeat
lagging strand synthesis to fill in the gap at the 5' end of the
other strand.
DNA Supercoiling (Lodish 109-111, 381-385)
Unwinding of the helix during DNA replication (by the action of
helicase) results in supercoiling of the DNA ahead of the
replication fork. This supercoiling increases with the progression
of the replication fork. If the supercoiling is not relieved,
it will physically prevent the movement of helicase.
The topology of DNA can be described by three parameters:
Consider closed circular DNA:
2. Twist is a real number. It refers to the frequency
or periodicity of the turns in the DNA helix. e.g. under condition
of normal physiological salt (0.15 M NaCl) and temperature (37
°C) the helical repeat is 10.6 base pairs per turn.
This is the value of the twist parameter.
Rubber tubing "helix" experiment
Cut two lengths of 1/8" rubber tubing, each about 20"
long. Insert a smaller piece of tubing, or piece of pipette tip
in the ends to allow the ends to be connected. On one piece of
tubing, mark both ends with a sharpie. This will allow you to
maintain correct strand "orientation" when you "ligate"
the strands of the duplex.
Introduce a Linking number = +2 (two 360° right handed twists
into the duplex), then "ligate" the ends (make sure
you maintain strand orientation, i.e. you connect the appropriate
two strands).
Confirm that the correct linking number has been introduced by
opening the duplex on one side and forcing all turns into a small
region of the duplex (easy to count this way). Confirm that looking
down the helix at the turns that they are "right handed"
(does not matter which way you look down the helix).
Note that the "duplex" when held between thumb and forefinger
and allowed to hang, prefers a "supercoiled" topology,
as opposed to "relaxed". Confirm that the "supercoil"
twists are actually "left handed" as you look down the
supercoil twists (regardless of direction down the twists).
The "supercoil" twist is most likely a full 360°
twist, rather than a 180° twist. In any case, hold the ends
of the duplex so that a 360° "supercoil" lefthanded
twist is present. Now, count how many times the strands of the
"duplex" cross each other. In this conformation the
strands of the "duplex" will not actually
cross each other (Note: you may have one strand crossing and then
later uncrossing, for a net result of no crossing). Thus, in response
to the introduction of +2 Linking number, the "duplex"
adopted +2 (180°) "supercoils", such that the resulting
"Twist" is zero (Note: that the supercoil is considered
positive, in response to positive twist, although it is "left
handed").
Remove one "positive" supercoil by unwinding by 180°.
Now hold the ends of the "duplex" and count the twists.
There will be a single right handed twist. Thus, a single
180° "positive" supercoil has the effect of removing
a single "positive" twist. The Writhe
number refers to the number of supercoils present.
Although it may seem that the consequence of introducing supercoiling
(Writhe) is changing the Linking number, it is not.
The consequence of Writhe is that the Twist is altered
(increased or decreased). DNA has a prefered Twist value (about
10.6 bp/turn) and Writhe is introduced to achieve this value for
a given (fixed) Linkage number for a given length of DNA. Linkage
number does not change with supercoiling (it can only change by
breaking the duplex) Writhe has the effect of changing the apparent
Linkage number. One supercoil is defined as being able to change
the apparent linkage number by +/- 1.
In our rubber tubing "duplex" the relationship between
Linkage number, Twist and Writhe can be stated as:
Linkage Number = Twist + Writhe
For example, we introduced a Linkage number of +2. If we prevent
supercoiling from occuring:
2 = 2 + 0 i.e. we get two 360° Twists in the duplex.
If we allow a single (180°) positive supercoil:
2 = 1 + 1
Thus, Twist and Writhe can vary, but linkage number remains the
same. Our rubber tubing "duplex" apparently prefers
a Twist value of zero, but DNA prefers a Twist value of about
10.6. For DNA of a certain number of base pairs in length, if
Twist is defined as base pairs per turn, we can write:
Linkage Number = (size of DNA in base pairs)/(Twist) + Writhe
For example, for "relaxed" DNA (i.e. with no supercoiling)
Writhe would equal zero. For a 5300 base pair duplex with ideal
Twist (i.e. 10.6 bp/turn) the expected Linkage number would be:
Linkage Number = (5300 bp)/(10.6 bp/turn) + 0
= 500 turns
SV40 genome
The SV40 genome is a circular, closed, double stranded DNA genome.
For the purposes of this discussion, it has 5300 bases. We expect
that under physiological conditions the DNA will exhibit 10.6
base pairs per turn (i.e. Twist = 10.6 bp/turn). In this case
the Linking number would be:
Linking number = 5300 bp/(10.6 bp/turn)
Linking number = 500 turns
i.e. we would expect 500 360° turns of the DNA strands over
the length of the circular genome. This form (with 10.6 base pairs
per turn) represents the "standard", or undistorted,
DNA helix. This is also known as the "relaxed"
form of DNA, and the duplex could physically be laid out flat
on a surface:
However, when the replication of SV40 is initially completed it
is observed that there remains an open duplex region in the DNA:
The result is that there are about 475 turns of the helix within
the duplex DNA (i.e. the Linking number = 475). The DNA is said
to be underwound. An open area is energetically
unfavorable. The covalently closed molecule cannot adjust
for this by increasing the Linking number. That is, it cannot
spontaneously break one or both strands of the duplex,
introduce another 25 turns into the duplex (increase the Linking
number by 25) and re-ligate the duplex. The DNA has three
choices:
Thus for the 5300 bp SV40 genome, with a Linking number of 475,
to maintain a Twist of 10.6 bp/turn, a total of 25 negative supercoils
(Writhe=-25) are needed:
475 = (5300/10.6) + Writhe
-25 = Writhe
That is, 25 negative supercoils (twenty five 180°
turns of the DNA duplex, right handed as you look down
the supercoiling).
Topoisomerases
The enzymes that control DNA topology are critical to DNA replication
and transcription. As the replication fork opens up, the region
of the duplex in front of the fork is subject to positive supercoiling.
The linking number has not changed, but the length of DNA which
contains all the turns is effectively shorter. To maintain 10.6
bp/turn in that region, the DNA will adopt positive supercoils.
For example, during SV40 replication, the duplex may open up such
that the Linking number (500) is effectively distributed over
only 4505 bases:
500 = (4505/Twist) + 0
Twist = 9.01
Thus, if no supercoiling is introduced, the DNA must adopt a Twist
of 9.01 base pairs/turn of the helix within the region ahead of
the replication fork. This is energetically unfavorable, and one
option for the DNA is to adopt a supercoiled configuration to
achieve 10.6 bp/turn:
500 = (4505/10.6) + Writhe
75 = Writhe
Thus, movement of the growing fork causes the DNA to adopt
positive supercoils. In this case the DNA has adopted
75 (180° left handed supercoils).
Type I Topoisomerase
Type I topoisomerases cut one strand of the DNA (i.e. it "nicks"
the DNA duplex). The 5' phosphate of the nicked strand is covalently
attached to a tyrosine in the protein. The 3' end of the nick
then passes once through the duplex. The nick is then resealed,
and a single supercoil is removed. In E. coli,
type I topoisomerase can only relieve negatively supercoiled
DNA (negative supercoiling is the end result of newly replicated
DNA genome). In eukaryotes, type I topoisomerase can also relieve
positively supercoiled DNA.
Type II Topoisomerases
Type II topoisomerases actually cleave the duplex DNA in changing
the linkage number. Type II topoisomerases can convert a single
positive supercoil into a negative supercoil. Thus the linkage
number changes (reduced) by two in a single step. Type
II topoisomerases are involved in both decatenation of daughter
chromosomes, and relieving the positive supercoiling ahead of
the replication fork. E. coli DNA gyrase
is an example of a type II topoisomerase.
Drug resistance mechanisms and expression vectors.
Including an origin of replication into a circular plasmid is a mechanism to have an extrachromosomal element in either a prokaryotic or eukaryotic cell.
In this case the plasmid uses the host cell machinery (i.e. polymerases,
helicases, dNTP's etc.) to direct replication. Since the added
work of replicating the extrachromosomal element is a load on
a cell, it will be outcompeted by other cells which do not contain
the plasmid. Since in prokaryotic cells the segregation of plasmids
is a random event, daughter cells can arise which do not contain
the plasmid and these grow faster (outcompete) the parent cell.
In other words, in the absence of other pressures, after a period
of time the population of cells in a culture will be those which
have "lost" the plasmid.
In organisms with more than one chromosome (eukaryotes) there are a variety of mechanisms to ensure that proper segregation of chromosomes occurs, i.e. to make sure that daughter cells contain equal numbers of all the chromosomes.
One basic mechanism is that each chromosome contains essential genes, and if these are lost, the cell cannot survive. This mechanism has been used in E. coli cells as well. For example, there is an extrachromosomal element called an F episome, or F factor which can exist in E. coli.
The F factor is a circular DNA element, with an origin of replication,
which is naturally involved in bacterial "sex". It has
been engineered by researchers to contain a gene essential for
Proline biosynthesis. Inserted into E. coli hosts which
are defective in Proline biosynthesis - and grown on media devoid
of proline - forces the host E. coli to maintain this extrachromosal
element.
e.g. The E. coli strain "JM101" genotype:
supE D(lac-proAB) [F' traD36 proAB+
lac Iq lacZDM15]
Interpretation:
The host genome contains an amber suppressor that inserts glutamine at UAG codons.
It also has a deletion of the chromosomal segment spanning the lac operon and neighboring genes coding for enzymes involved in proline biosynthesis.
This E. coli strain also harbors an F episome.
This episome contains a mutation that suppresses conjugal transfer of the F episome.
It also contains the region of the bacterial chromosome coding for enzymes involved in proline biosynthesis.
The laq Iq gene codes for the lac operon repressor protein and the 'q' indicates it is a mutant which makes a lot of the repressor protein (about 10x over the normal gene).
"lacZDM15" is a deletion
mutant of the lacZ gene (b-galactosidase)
where a short amino terminal region of the protein has been deleted
(if this non-functional mutant protein is mixed with a peptide
representing this amino terminal region functionality will be
restored).
Drug resistance
Although an extrachromosomal element can be maintained by incorporation
of essential genes (in conjunction with a deletion host), by far
the most common approach to the maintenance of plasmids is through
the incorporation of drug resistance genes. These are also
known as selectable markers, i.e. we can select for their
presence by including antibiotics in the growth media.
Ampicillin
Ampicillin binds to and inhibits a number of enzymes in the bacterial membrane that are involved in the synthesis of the cell wall. Therefore, proper cell replication cannot occur. The ampicillin resistance gene (ampr) codes for an enzyme (b-lactamase) that is secreted into the periplasmic space of the bacterium where it catalyzes hydrolysis of the b-lactam ring of the ampicillin. Thus, the gene product of the ampr gene destroys the antibiotic.
Over time the ampicillin in a culture medium or petri plate may
be substantially destroyed by b-lactamase.
When this occurs, cell populations can arise which have "lost"
the plasmid.
Tetracycline
Tetracycline binds to a protein of the 30S subunit of the ribosome and inhibits ribosomal translocation (i.e. the drug interferes with normal translation - production - of proteins). The tetracycline resistance gene (tetr) encodes a 399 amino acid outer membrane associated protein that prevents the antibiotic from entering the cell.
Thus, this drug resistance gene does not destroy the antibiotic.
Pressure will be maintained throughout the cell culture process
to keep the plasmid.
Chloramphenicol
Chloramphenicol binds to the ribosomal 50S subunit and inhibits protein synthesis. The chloramphenicol resistance gene (Cmr) codes for a protein known at the cat protein. The cat protein is a tetrameric cytosolic protein that, in the presence of acetyl coenzyme A, catalyzes the formation of hydroxyl acetoxy derivatives of chloramphenicol that are unable to bind to the ribosome.
Like with ampicillin, the Cmr gene product destroys
the antibiotic. Additionally, the expression of cat protein is
influenced (down regulated) by the presence of glucose in the
media.
Kanamycin and neomycin
Bind to ribosomal components and inhibits protein synthesis. The Kanr gene codes for a protein which is secreted into the periplasmic space and interferes with the transport of these antibiotics into the cell.
Like tetracycline resistance, the Kanr gene does not
destroy the antibiotic.
Colicin E1
This is a member of a general class of substances known as bacteriocins.
Colicin E1 causes lethal membrane changes in bacteria. The drug
resistance gene (cea) codes for a protein that interferes
with the action of colicin in an unknown manner.
Prokaryotic plasmids:
pBR322 (4.36 Kb)
One of the original cloning plasmids. Contains both ampicillin
and tetracycline resistance genes (markers). Contains unique restriction
sites inside and outside of these markers. Low copy number (10-30)
PUC18/19 (2.69)
Lacks the rop gene near the oriC region. This gene is involved in regulation of copy number (replication). Thus, this plasmid tends to accumulate in high copy number (100-200). This vector contains only the ampicillin resistance marker.
This vector also contains a transcription promoter region from
the lac operon, which allows foreign genes to be inserted and
transcribed/translated. Sites for gene insertion exist in a polylinker
region. This region of DNA contains sequences for a variety of
unique restriction sites. It is in one orientation in pUC18, and
another orientation in pUC19. Thus, with the two vectors unique
insertion orientations may be accomplished.
Eukaryotic selectable markers
A selectable marker for eukaryotes which has been used successfully
involves the biosynthesis of the pyrimidine deoxy Thymidine monophosphate
(dTMP). In eukaryotic cells this essential nucleotide can be synthesized
from dUMP via the action of a pair of enzymes - dihydrofolate
reductase and deoxythymidilate synthetase:
dTMP can also be synthesized from deoxythymidine by thymidine kinase. This is known as the "salvage" pathway. Methotrexate inhibits the action of dihydrofolate reductase. If methotrexate is given to cells which are defective in thymidine kinase they die - they have no means of synthesizing the essential pyrimidine dTMP.
Mutant cells (tk-) which harbor a plasmid containing a functional
thymidine kinase gene (tk+) will rely upon the thymidine kinase
for salvage pathway synthesis of dTMP when they are grown in the
presence of methotrexate.