Hi. if you want to extract bacteria DNA indirectly, it is not problem because you don't want to extract plant DNA. you want to extract bacterial DNA from plant samples indirectly thus first you need to grow endophytic bacteria on the medium and then extract its DNA thus you do not need any liquid nitrogen 

Thank Mohsen and James. I actually want to look at the whole bacterial community inside the root, so I cannot culture them. 

If you want to extract bacterial DNA directly you must to have a sensitive detection method for identification of low number bacteria that are inside the plant root. The normal method for screening of the bacterial community that lives inside the plants is culture of bacteria in the medium, select one clone of the same bacteria in the medium and culture of that clone (bacteria) for DNA extraction. 

If you want to look into the whole bacterial community, then culture independent method should be adopted along with culturable. you can extract DNA and go for 16s identification by methods like DGGE or pyrosequencing.

If you want to look at the bacterial community you carefully tease apart the root material.  After that you soak the material in PBS while shaking for 20-40 minutes.  Take the supernatant, transfer to centrifuge tube and spin down full speed in a microfuge for 3 minutes.  You can then extract the DNA using any column based bacterial DNA extraction method (do not use phenol chloroform).  Then prep the DNA for 16S sequencing on the Next Gen platform of your choice (some of the sequencing centers have the appropriate primers etc)  This method of sequencing will give you relative abundance of the bacterial present and gets all the different types since you are not dependent any culture method.  The only problem is that part of the plant mtDNA will also be sequenced and those sequences need to be removed.  That is why I said to gently tease apart the roots, so you don't disrupt too many plant cells.  

Dear Joseph Nietfeldt,

The method that you described seems to be very good. So, it is only feasible to be carried with fresh tissues, correct? Because frozen plant tissues can be the cells disrupted by the crystals (mainly at -20°C... we know).

Other question, do you have some reference for these suggestions?

Kind regards

Sorry I cannot give you any references.  We used to do custom 454 sequencing and we did a few projects for other researchers and found this was the best way to isolate bacterial DNA from plant material with the least contamination of plant DNA.  Fresh tissue is definitely the best.  You are correct that plant cells may be disrupted in freezing which would give you more plant mtDNA sequences (I forgot to mention you will also get chloroplast sequences), it can also disrupt some bacterial cells which will be lost.  You might be able to use frozen material if it has not been thawed and refrozen and has not been frozen at -20 C for more than a few months.  It is best to store any frozen biologicals at -80 C.  

Hi Joseph,

If I just tease apart the roots and shake for 20 min as you suggested, can I still get the endophytes out of the roots? 

That has worked the best for us in getting at the bacteria while limiting the plant DNA interference.  Be sure to soak in sterile PBS.  I know others have used a 1% solution of peptone which will add contaminating DNA also.

Hello colleague!

I macerate 5 gr of sterilized roots with sterile saline solution (0.85 %) using sterile pestle and mortar. The macerate was used to extract total DNA with a soil DNA extraction kit. Then, bacterial 16S genes were amplified from this samples using special primers that avoid mitocodrial and plastids sequences.

Felix Moronta Barrios, please can you provide reference for this method you just stated. 

Dear Emmanuel, 

Please find the attached file for reference of Felix's method. But reference methods are some modified. They used commercial kit specific for plant tissue.