I have routinely isolated plasmid DNA from E. coli manually for cell culture transfections. I used a protocol, containing 1) alkaline lysis, 2) RNAse processing, 3) phenol/chlorophorm extraction, 4) ethanol precipitation.

DNAses are blocked by adding of EDTA, which is present in the most solutions for bacterial culture resuspension (the very first solution in the protocol). When the plasmid gets linearised at a random site it gives one band in electrophoresis, but a smear after a restriction. I think, the problem may be also in overlysing and overshaking during the alkaline lysis step - this may also give a problems with restriction (simply do not incubate with NaOH/SDS solution more than 5 min and do not shake during this step). If you would like to, I can translate my protocol for you.

Dear

You have to use method as per Shambrook. Prepare your buffers carefully. there may be some error in preparation and quality of buffer. check the efficiency of all the buffers/solutions. definitely u will get good results. as we got. Regards

I refer that,

lysozyme + EDTA with 25% Sucrose medium for cell  partial lysis (sphaeroplast)

Then  add "triton x100" (adding SDS cost chromosomal breakage) this method causing  little breakage of the (bacterial)host chromosomal breakage 

Then get a cell extract

Centrifuge it

Then u get cell debris with larger DNA fragments  (lower part)

The upper part contain clear lysate, it contain plasmid DNA

Is it gramm negative or pozitive bacteria ? Problem could be on the lysis step as well.

Anyway, here are protocols for G+/- bacteria, which I use.

Ask me if you have a quetions.