I'd like to get the sequence from an antibody because I want the variable region of the antibody to get expressed in my cells. 

1. How to get the sequence of the Ab?

Check sequence databases for the name of the antibody (Genbank, IMGT), trace back publications and patents to find whether its sequence has already been published, or if the hybridoma generating this antibody is available. If necessary, ask the providers of the antibody/the researchers who first published the antibody. 

If you can get the hybridoma, but not the sequence - clone antibody sequences and sequence  the DNA. 

If you only have the monoclonal antibody: affinity-purify the immunoglobulin on your antigen and sequence the variable domains

If you have a polyclonal antibody - This will contain a wild mixture of different antibodies, and you won't be able to get the sequence.

2. After knowing the sequence, how can I get the DNA for the sequence? 

Reverse-translate the protein sequence into a DNA sequence, taking into account the codon preferences of the cells you want to use to express your antibody. Select an appropriate expression plasmid, and design your DNA sequence  with the appropriate secretion signals, restriction sites and so on. Once you are satisfied with your sequence, you can order the DNA from a specialized DNA synthesis company.

Amended on 26 September 2017

I would like to express my opinion.   Binding protein such as avidin binds D-biotin (vitamin H), D-desthiobiotin, lipoic acid (thioctic acid), and many amino acids (affinity is stronger to L-type than D-type in the case of amino acid) (please see files; Thioctic acid Determination, and D-Asp avidin).   Further, D-biotin binds stronger than D-desthiobiotin (please see file; Netherlands biotin).

DNA sequence (or protein's amino-acid sequence) is quite different between avidin (chicken) and streptavidin (bacteria).   Purified human serum biotinidase shows many N-terminal sequences by using Edman degradation analysis (please see file; Structure BIN (in Japanese)).

We have recently found that the affinity to substrate (Km) is changed by changing the glyco-chain of eucaryotic enzyme (please see file; JMBT Alopecia 1).   The affinity of eucaryotic IgG may be considerably effected by glyco-chain structures.   Thus, monoclonal antibody is not so specific than polyclonal antibody (personal communication from Dr. Chihiro Yabe-Nishimura, M.D., Ph.D. (Kyoto Prefectural University of Medicine, Department of Pharmacology, Kyoto, Japan)).   Therefore, ELISA is unexpectedly not so specific method, and the positive or negatve falses (false binding reactions) are frequently occur.

Then, I am considering that the affinity is mainly regulated by the glycochain structure and the state of multimerization in eucaryote, though affinity of procaryote proteins may be related to DNA sequence and the state of multimerization (please see file; The Fascio effect).

I'd like to get the sequence from an antibody because I want the variable region of the antibody to get expressed in my cells. 

1. How to get the sequence of the Ab?

Check sequence databases for the name of the antibody (Genbank, IMGT), trace back publications and patents to find whether its sequence has already been published, or if the hybridoma generating this antibody is available. If necessary, ask the providers of the antibody/the researchers who first published the antibody. 

If you can get the hybridoma, but not the sequence - clone antibody sequences and sequence  the DNA. 

If you only have the monoclonal antibody: affinity-purify the immunoglobulin on your antigen and sequence the variable domains

If you have a polyclonal antibody - This will contain a wild mixture of different antibodies, and you won't be able to get the sequence.

2. After knowing the sequence, how can I get the DNA for the sequence? 

Reverse-translate the protein sequence into a DNA sequence, taking into account the codon preferences of the cells you want to use to express your antibody. Select an appropriate expression plasmid, and design your DNA sequence  with the appropriate secretion signals, restriction sites and so on. Once you are satisfied with your sequence, you can order the DNA from a specialized DNA synthesis company.

@Annemarie, Thanks for the reply! Elegant answer.

And I have a follow up question: if I take a variable region of an antibody from a patent, and then recombine the region with other domains. Can I publish this on journal?

Commercial services are available just for that. For example at rapidnovor.com, we routinely sequence monoclonal antibody proteins without the need to access the cell line.

I'd like to quickly follow on my previous comment.

Antibody Sequencing using mass spectrometry is a good solution for this problem - here are some articles about Antibody Sequencing (https://www.rapidnovor.com/antibody/).

The advantages are that:

(1) you can determine the antibody sequence directly from a small 0.2mg protein sample of antibody, quickly

(2) there’s no need to reference an antibody sequence database, use DNA or access the cell line or hybridoma.

There are also commercial services available for getting the sequence of an antibody:

Antibody Sequencing Services (https://www.rapidnovor.com/antibody-sequencing-service/)

Antibody Sequencing plus Expression Services (https://www.rapidnovor.com/recombinant-antibody-expression-service/)

Thank you.

@Artur Thanks! !