Is there any way to design a PCR primer if region of interest is highly repetitive (tens of kilobases only SINE's & LINE's)?
This paper describes some PCR strategies with LINE and SINE PCR identification (Shedlock and Okada. SINE Insertions: powerful tools for molecular systematics. BioEssays (2000) 22:148-160.).
I have no experience with PCR amplification of SINEs or LINEs, however I can think of two strategies right now.
1) You may be able to find a unique 18-20 nt region flanking those regions. If you can, great. If not, perhaps there is a unique site somewhere in the middle of these regions, then you can amplify from the middle outward to both upstream and downstream regions.
2) If there are no unique sites that you can exploit, that would imply a highly regular (ie, repetitive) pattern, but you can take advantage of this too. You can pick primers such that they are on the extremes of the repetitive subunit within these structures. For example, if your sequence looks like this:
Then you can pick primers A and B for PCR use. You will have non-specific primers, and therefore when you amplify the DNA, you will get all possible permutations of DNA that these primers will produce. Specifically, you will get:
and so forth.
This last approach is obviously inefficient, because your possible products scale exponentially, and will be produced with equal likelihood, minimizing the DNA yield of any particular segment.