How to design primer for unknown DNA sequence? - (Aug/19/2006 )

Hi all,

Sequencing DNA requires primer to hybridize to one strand of the template DNA. But we are sequencing the DNA because we do not know the sequence. If we do not know the sequence, then how do we actually design the primer?? Do we use degenerated primers or..?

i have been reading about DNA sequencing for quite some time but don't have a chance to perform it (hopefully someday).

All comments and answers are welcome

Thanks in advance!

Willmch

---

DNA is typically cut with restriction enzymes or by shearing, then ligated into a replicating plasmid or phage. The plasmid is isolated for sequencing. The plasmid contains known sequences which can be used to prime the sequencing process. For short inserts, this is sufficient. For longer inserts, results of the sequencing can be used to design subsequent primers which allow sequencing further into the insert. This technique is called primer walking. Other techniques, such as transposon insertion or partial sequence deletion are also sometimes used.

---

Hi,

But to cut DNA with endonuclease, you need to know the sequence of DNA so that you know which endonuclease to use and also whether the specific sequuence is present in the DNA or not . And the problem now is we do not know the sequence

wllmch

---

is your dna a pcr product? you could clone it into a T-overhang vector (eg, pGEM T easy - promega) and then sequence with the vector derived primers (T7, SP6)

---

Other possiblity is to start with primers designed (or even proved) on DNA of near relatives of your sample species. This so called "cross species amplification" is quite often used in population genetics. Then you also know which region you use.
Instead you can also use random primers (leading to RAPDs if you want).

---

Hi,

(to aussieuk) no. i am just reading about DNA sequencing and not actually doing it.

Because most of the books and websites i found are talking about how to use primers, ddNTPs, and pcr technique to run Chain Termination sequencing, i just wonder is this the technique that is used to sequence the "first" DNA seqeunce?

Because you need to know at least "some of the sequences" of the DNA in order to sequence the others. My question is, how scientists get the "first DNA sequence ?

Thanks!

---

Ah...good question

there are many approaches. often times it begins with a created or naturally occuring phenotype, which is studied in various ways to yield the responsible gene(s). this is a very broad topic and there are many approaches. sometimes it starts with the mutant protein(s) and works backwards, sometimes it involves making mutations and studying them. for example, knockout strains can give good indications of gene function. if knockouts are created using transposon mutagenesis, it is then possible to clone the sequences contiguous to the insertions and study the 'first DNA' further.

did that help? did I make any sense?

---

Primers for unknown sequences you can design on highly conserved DNA regions, which sometimes are the same (or almost the same) even between worm and human, or yeast and chimp if you want.
If your DNA is in a vector (whose sequence you know), then you have also primers for sequencing e.g. standard M13 primers. Then you first sequence some of the known vector sequence then the unkown seq. of interest.
Hope you meant that?

---

Hi,

(to aimikins) Yea, that makes prefect sense.

Sorry, i think i didn't make the question clear enough, or maybe my title is misleading.

Actually i wanna ask:

Back to many years ago, when there was no pubmed,swiss prot or other databases, who was the first scientist that seqeunce DNA and how he/she did that?

---

Hi,

I completely understand your question as I still feel uncomfortable to answer it. As you have re-phrased your question now people would understand much easily. I'll include a reply by phage434

As we don't know the sequence so we'll do the digest with any one restriction enzyme, irrespect to whether we know the sequence for that enzyme or not it will or should at least digest twice. Why at least twice? so that we should get two bands of different kb size DNA. If the enzyme doesn't cut or it gives a very small size band on gel electropheresis which is undetectable or difficult to isolate then... simply use another restriction enzyme. And if we get more than two bands on gel electropheresis then we have option to choose of our likening. At this moment it doesn't matter if we know the restriction enzyme's site, as you've mentioned that we don't know the sequence of our 'unknown DNA'. Its just a random pick!

After choosing a band from the gel electropheresis with UNKNOWN sequence, we can use different approaches now. One could be incorporating it into a shuttle vector with known sequence. This vector has been constructed by humans or at least they know the sequence and we know where to place the insert (commonly known as multiple cloning site, MCS). There are different construct/plasmid available. After putting the UNKNOWN band in the vector, we could transformed it into bacteria to grow larger qunatities of the vector, now called a plasmid which has our UNKNOWN BAND which was isolated from the gel. We can now make a pair of primers (forward and reverse) from our plasmid backbone just on either extremes of our insert (we know the sequence of the plasmid, remember!). Then by simply doing a PCR would give larger quantity of the UNKNOWN BAND only. Now do the sequence it find the sequence. After that we don't need to repeat this process, simply use this UNKNOWN BAND (now we know the sequence) to make primers and go on....

You might ask that why didn't we do the sequencing right after isolating the UNKNOWN BAND at first place, why need to go through all this transformation and PCR steps? I'd reply to that by saying that we don't get the enough DNA after gel purification and sequencing requires stepwise optimization i.e. more than one reaction are required and we don't want to run out of our stock, do we?

I hope this was helpful to answer your question.

DNA1977

---