failed PCR on DNA extract from blood - (Jun/24/2009 )

Hi,

I have been trying to PCR the bacterial DNA gene from real human blood. Briefly, I collect the buffy coat (where my bugs reside), next extract the bacterial DNA with Qiagen DNA Blood kit, then use the eluted DNA directly for PCR.

Problems I have encountered:

1) I have been able to amplify only small percentage of the samples, say out of 10 samples, only a few are positive. I assume that there may be too few of the copy number of the bacterial DNA. Do you have any suggestion for improvement of the sensitivity?

2) In some samples, besides my expected band on the gel, I also get the non-specific band also and I believe it is human DNA. There are also cases where only human DNA is amplified and no specific bacterial band at all. How can I get rid of this non-specific? I have tried the optimization with annealing temp, Mg2+, and primer concentration but still no luck. It's either with the co-amp of human DNA or no specific bacterial band at all.

However, I don't have any problems whatsoever when I amplify from cell culture-extracted DNA where I think there's high bacterial load and primers may have better chance at finding the bacterial DNA.

I appreciate any suggestion you may have. Thank you so much in advance. :-)

I'm sorry if I'm asking stupid questions, but I don't work with eukaryotes and I'm not a native English speaker.

What do you mean with "buffy coat"? Are the bacteria in the plasma or inside the blood cells?

If they are in the plasma you should be able to seperate bacterial cells from blood cells by centrifugation.

And "DNA Blood Kit" sounds like it is for extracting eukaryotic genomic DNA. Wouldn't it be better to use a kit which is specially made for the isolation of bacterial genomic DNA?

I'd think about lysing the white blood cells in a hypotonic solution and doing differential centrifugation step to separate the WBC nucleus from the bacteria. That should reduce the human DNA contamination.

swanny on Jun 25 2009, 08:37 AM said:

Thanks so much for your suggestion. I'd really love to follow your advice but .. unfortunately the starting materials that I have right now is just DNA already extracted from buffy coat. It is impossible to obtain the patients blood samples again.

What about the reactions where there's no amplification at all? Do you think it can be attributed to the too low of a sensitivity of PCR? Because some patients do have very very few copy numbers of the bugs.

logix on Jun 25 2009, 12:29 PM said:

OK, because you can't get fresh tissue (a real pain, that), you could do a couple of things.
1. increase the reactions volume to 100 or 150 ul, and use more template DNA. Yes, this will increase the noise level, and yes, it might inhibit the reaction a bit, but I have done similar stuff, looking for a single viral genome in 2.6 ug gDNA (I kid you not).

Better than this, however, is to do a nested PCR. Do your first round and take 5 ul into a fresh 50 ul reaction with the inner primers (there's no problem of interference this time because it's a "clean" PCR reaction.

What are your conditions again? Have you done a positive control reaction with a housekeeping gene to show the DNA is high quality, and that there are no significant inhibitors?

swanny on Jun 26 2009, 08:18 AM said:

Hi and thanks again! I think my boss would not like it so much if I were to use "more" template DNA for the optimization since the buffy coat is like the most precious thing and in fact, we are already running low on them (obviously, because of my too many attempts to amp the bacterial DNA). I might do a serial dilution of a positive control and optimize it from there but then again to make it as real as possible I might have to draw my own blood and spike it with these dilutions. But it's frustrating if I have to do all this, it's really going to be a hassle and I don't have that much time. Ugh.

As for nested PCR, again we have this published primer pair to amplify the full-length ORF of this gene and we want to get the entire gene out of PCR, if I used the "inner" primers, I wouldn't get the full-length then.

They don't do the internal housekeeping gene here (as far as I'm concerned, I'm pretty new here). However, I'm not worried too much about the DNA integrity as the majority of the DNA samples are positive when we amp a different gene.

Again, thanks so much for trying to help! I appreciate it a lot! :-)

Hi, I have done very similar work PCRing bacterial DNA from blood and used the same method, had some good positives and others zero (either too low or no bacteria!). Have you made sure the eubacterial primers are good? ie, tried them against a range of bacterial DNA, the ones you're most likely gonna find in the blood? There are lots of eubacterial primers out there and some are good and some really arent at all, therefore maybe have a look around?

Hope you get it sorted soon
KT