LCL vs whole blood samples? (shearing differences)

sehrrot · 09-19-2010, 11:23 PM

Dear all,
I sheared lymphoblastoid cell line (LCL) samples and whole blood samples by covaris S2. Then, I ran bioanalyzer to check their length.
Interestingly, unlike whole blood samples that showed typical graph after covaris shearing, sheared LCL samples looked different
First, a bit wider than whole blood samples
Second, low fluorescent unit -> I measured conc' again with nanodrop but same as whole blood.
Third, their graph started from little bit higher than baseline.

I assumed that fragmentation of episome is a reason why a bunch of small fragments are. But I cannot understand low concentration and wider distribution.

upenn_ngs · 09-20-2010, 08:24 AM

What is the downstream application? Looks like the LCL sample concentration was lower than expected, so although the same covaris profile was used the two samples run on the bioanalyzer much differently.

Also, if you are sequencing be aware LCL samples have a high number of SNP artifacts from transformation.

sehrrot · 09-20-2010, 04:15 PM

Quote:

Originally Posted by upenn_ngs

What is the downstream application? Looks like the LCL sample concentration was lower than expected, so although the same covaris profile was used the two samples run on the bioanalyzer much differently.

Also, if you are sequencing be aware LCL samples have a high number of SNP artifacts from transformation.

I used same protocols, including time, intensity, and volume, and application (covaris S2). Also, I ran bioa' with other samples in same condition.

These LCL samples are transformed by EBV. I haven't found any papers that EBV change the host genome. If you know any papers for EBV integration, could you cite them on here?

upenn_ngs · 09-21-2010, 08:03 AM

Good question-- I presume the variability is due to a single cell, or subset of cells taking over the culture during transformation, and additional mutations during division. For whole exome resequencing, we have seen 10,000-30,000 more SNPs pass filter in LCL samples vs. blood.

sehrrot · 09-21-2010, 11:56 PM

Quote:

Originally Posted by upenn_ngs

Good question-- I presume the variability is due to a single cell, or subset of cells taking over the culture during transformation, and additional mutations during division. For whole exome resequencing, we have seen 10,000-30,000 more SNPs pass filter in LCL samples vs. blood.

Yes, we used two LCL samples, which of one has been cultured for a long time. We are also expecting EBV-transformed one has some SNPs but I'm just wonder why it was shown broader. Maybe LCL-transformed cell's genome might have become fragile cause of virus integration (consistently it would take more CNV I think).

Hamid · 09-27-2010, 11:06 AM

Hi Sehrrot,

I think the diffrences you are noticing are simply due to concentration differences of the samples. I think it is essential not only to carry out a nanodrop analysis, but also a Qbit or similar picogreen analysis for concentration. My guess is that if you run a Qbit analysis you will notice that the concentration of the two samples will be different unlike the uv/vis analysis on the nanodrop.
what settings, volume, buffer type, and tube type are you using for the fragmentation?

Thank you

Hamid

sehrrot · 09-27-2010, 08:24 PM

Hi Hamid,

Thank you for your reply.
I did QBit, nanodrop and Epoch and also did 5 different concentration in bioa' analysis.
Even Qbit gave more accurate results, all three method showed almost same result.

But my question is why LCL transformed sample showed little bit higher start region (10bp~80bp) and why LCL transformed samples generally showed lower concentration than whole blood sample.

We used recommended covaris setting with various setting, including 100, 120 ul, 130 ul shearing and TE buffer, MiliQ water shearing.

josdegraaf · 10-04-2010, 02:08 AM

your lcl samples fragmented to much smaller sizes, and you lost a lott of those during cleanup?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Whole Blood RNA-Seq Libraries and globins	jbrowne73	Sample Prep / Library Generation	11	10-16-2013 01:20 PM
full sequencing on refrigerated blood	xyzzy123	General	4	01-03-2012 02:30 PM
Agencourt RNAdvance Blood kit	bethedelblute	Sample Prep / Library Generation	0	09-16-2011 08:46 AM
Cufflinks pipeline for splicing differences between animal samples	seqhorn	Bioinformatics	0	07-08-2011 02:23 AM
RNA from blood fractions	cjharrys	Illumina/Solexa	0	10-29-2010 03:05 PM

09-20-2010, 08:24 AM	#2
upenn_ngs Member Location: philadelphia Join Date: Sep 2009 Posts: 70	What is the downstream application? Looks like the LCL sample concentration was lower than expected, so although the same covaris profile was used the two samples run on the bioanalyzer much differently. Also, if you are sequencing be aware LCL samples have a high number of SNP artifacts from transformation. Last edited by upenn_ngs; 09-20-2010 at 08:28 AM.

09-21-2010, 08:03 AM	#4
upenn_ngs Member Location: philadelphia Join Date: Sep 2009 Posts: 70	Good question-- I presume the variability is due to a single cell, or subset of cells taking over the culture during transformation, and additional mutations during division. For whole exome resequencing, we have seen 10,000-30,000 more SNPs pass filter in LCL samples vs. blood.

09-27-2010, 11:06 AM	#6
Hamid Senior Member Location: San Diego, CA Join Date: Sep 2009 Posts: 105	Hi Sehrrot, I think the diffrences you are noticing are simply due to concentration differences of the samples. I think it is essential not only to carry out a nanodrop analysis, but also a Qbit or similar picogreen analysis for concentration. My guess is that if you run a Qbit analysis you will notice that the concentration of the two samples will be different unlike the uv/vis analysis on the nanodrop. what settings, volume, buffer type, and tube type are you using for the fragmentation? Thank you Hamid

09-27-2010, 08:24 PM	#7
sehrrot Member Location: USA Join Date: Jul 2010 Posts: 58	Hi Hamid, Thank you for your reply. I did QBit, nanodrop and Epoch and also did 5 different concentration in bioa' analysis. Even Qbit gave more accurate results, all three method showed almost same result. But my question is why LCL transformed sample showed little bit higher start region (10bp~80bp) and why LCL transformed samples generally showed lower concentration than whole blood sample. We used recommended covaris setting with various setting, including 100, 120 ul, 130 ul shearing and TE buffer, MiliQ water shearing.

10-04-2010, 02:08 AM	#8
josdegraaf Member Location: Germany Join Date: Mar 2010 Posts: 33	your lcl samples fragmented to much smaller sizes, and you lost a lott of those during cleanup?