There are several DNA-binding domains, allowing the identification of specific sequences or motifs:
Basic Leucine Zipper Proteins These alpha helices have leucine residues repeating every seven amino acids, causing a positive charge to be present on one side of the helix. This allows interaction with the negative DNA, enabling the protein to bind and interact. Other residues change to allow specificity with the target DNA sequence, allowing transcription factors (amongst other DNA-binding proteins) to specifically interact with their intended region.
Zinc-Finger Domains Zing-finger domains use a coordinated zinc ion to fold the protein around. This provides a $$\beta ,\beta ,\alpha$$ fold, forming a recognition helix with specific amino acid residues, enabling the identification of specific nucleotide sequences. This was dominant prior to the advent of CRISPR, using the zinc-finger domain to identify the desired region before a catalytic domain is used to modify the DNA. CRISPR, due to the use of a guide RNA of approximately 20 nucleotides, allows for greater specificity with gene editing - zinc-fingers only allow specific binding with one nucleotide triplet. Multiple domains can be used to increase specificity, however this greatly increases the complexity of the protein being used.
Homeodomain Fold This is a structural motif found in specific transcription factors (sTFs), interacting with the major groove of the DNA. Using a dimeric homeodomain fold increases the size of the protein, and subsequently makes it more specific. An alpha helix has specific residues to interact with the desired DNA sequence.