Note: this is a search engine friendly version of my lab notebook, please see the pdf version of this document for a more human friendly, printer friendly version.

Chapter 7
Removal of rRNA from prokaryotic total RNA samples

THIS CHAPTER/PROJECT IS IN PROGRESS
In Chapter 6, I am trying to generate and sequence paired-end-tags from prokaryotic mRNA. All companies that offer highly-parallel sequencing machines are also working to develop mechanisms for paired-end-reads. Presumably, with their large crews of folks working on this problem, someone is going to figure out a good solution, which will then make my solution (based on the work of Shendure et.al. Science 2006) in chapter 6 obsolete.
However, one theme that has begun to dominate chapter 6 and which is a more general problem for all prokaryotes that will continue even after the development of efficient paired-reads technologies is - how to remove rRNA from total RNA in prokaryotes. I'm going to test a few different ideas related to the rRNA removal problem in this chapter.

7.1  developing a gentle, quick-lyse procedure that produces undegraded total RNA

Several of the ideas I have (or that I've received from others) for removing rRNA from total RNA require a fast gentle total RNA prep procedure that does not use harsh chemicals (e.g. those normally used to inhibit RNAses also mess up protein folding in general) and does not result in degraded RNA despite the absence of typical RNAse inhibitors. Not a small order...
I'm going to try and develop such a procedure in this section. I'll assay RNA quality by looking for the standard 23S and 16S bands on an agarose gel.

7.1.1  quick-and-simple readyLyse preparation of total RNA try 1

Dec 5, 2007
growing the cells
I added a 1:50 dilution of MG1655 overnite culture into 4 ml of LB in a 12 ml falcon tube. I grew the cells for 3hr to an OD600 of 0.8 (a little higher than I wanted).
lysing the cells
I took 1.5 ml of the OD600 0.8 culture and resuspended it into 15 ml TES (TE + 100 mM NaCl). I tried both normal and 10x ready-lyse [Epicenter] amounts (29U and 290U total). The samples were incubated for 5 minutes at RT. The samples were spun down at 13K rpm for 1 minute and the supernatant was retained as a RNA/genomic DNA mixture. 1.5 ml of Turbo-DNA-free [Ambion] was added and the samples were incubated at RT for 3 minutes to degrade the genomic DNA.
running on a gel
I ran the entire 15 ml on a 1% TAE gel with NEB ssRNA sample buffer (Figure lanes labeled 1 and 2 are the 1x and 10x ready-lyse samples respectively).
Brief Conclusions:   Rubbish. Not sure what that is on the gel (Figure ). There is certainly nucleic acid material there, but it is pretty short to be genomic DNA and doesn't have the characteristic 16S and 23S rRNA bands I'd expect to see if it were total RNA. Next time I need to add DNAse before I run the sample on a gel and to try using supernase to inhibit degradation in case what I see here is just degraded RNA.

7.1.2  quick-and-simple readyLyse preparation of total RNA try 2

Dec 12, 2007
I want to retry the quick-lyse procedure to see if I can get a decent rRNA band. First I'm going to use superase [Ambion] RNAse inhibitor. Second I'm going to try using very low levels of formaldehyde. The formaldehyde might deactivate RNAses and perhaps it will crosslink the rRNA-ribosome complex together and make it easier to spin down for removal?
I grew an 1:50 dilution of overnite MG1655 culture in 4 ml LB in a 12 ml falcon tube for 2 hr and 37 minutes the OD600 samples for the three tubes were 0.734, 0.722, 0.763 (not bkgrd subtracted). I took 1.5 ml of culture for each RNA sample. Sample 2 was mixed with 0.1% formaldehyde for 10 minutes at RT and quenched with glycine to stop the crosslinking reaction. I used 1x lysozyme and TS media (10 mM Tris, 100 mM NaCl) instead of TES in case the EDTA was inhibiting the RNAse H before. I washed all cells 1x in 100 ml TS. I resuspended them in 25 ml TS + 1 ml superase (20 Units for samples 1-3; sample 4 had no superase). I added 1 ml of 1x lysozyme and lysed for 15 minutes at RT (vortexing briefly every 3 minutes). I spun the lysed samples for 5 minuts at 13K rpm and keep the supernatant. I added 1 ml of DNA-free Turbo DNAse [Ambion] and incubated for 15 min to remove DNA from the samples. I then added 1 ml RNAse H to sample 3 and incubated for 10 minutes at 37 C.
I ran all four samples on a 1% TBE gel for 50 min at 120 V (Figure ).
Please see the pdf version for figures
Figure 7.1:
Brief Conclusions:   More rubbish... As Figure 7.1 shows (lanes 1-4) my quick lyse procedure sucks. I'm about to jump ship on this idea. I need the quick, gentle lyse procedure if I want to try to remove rRNA via ultracentrifugation, but the lyse procedure seems a long way off, since I really haven't moved forwards with any ideas that worked so far. I'm particularly disappointed that the formaldehyde sample didn't work. I figured that would deactivate the RNAses and allow me to have relatively clean total RNA with a 16S and a 23S band. Perhaps my problem is not RNAses?

7.2  problems with assaying rRNA removal

Tue Dec 11 17:12:21 EST 2007
On of the challenges with removing rRNA is that the rRNA bands are the most common way to judge the quality of the RNA purification procedures. The distinct 16S and 23S bands are the hallmark of a good RNA prep. So when I attempt to remove those two bands, it is difficult to determine if I've remove those two bands or if I just degraded those two bands into a smear of rRNA. One way to better judge this rRNA removal is with a ssRNA ladder. I have some ssRNA ladder from NEB (Part No: N0362S) to use for this purpose.

7.3  rRNA removal via RNAseH

Write strategy here.
primers
I designed the 23S primers using an alignment of E. coli 23S sequences and primer3. The 16S primers are universal 16S primers from the supplemental material of Gill et.al. Science 2006.
----- 16S primers -----
Bact-8F
5' AGAGTTTGATCCTGGCTCAG
Bact-1510R
5' CGGTTACCTTGTTACGACTT 

----- 23S primers -----
forward
5' GACTAAGCGTACACGGTGGAT
reverse
5' TTAAGCCTCACGGTTCATTAG

Note that the Bact-8F primer is NOT an exact match to E. coli 16S rRNA
 Score = 32.2 bits (16), Expect = 0.007
 Identities = 19/20 (95%)
 Strand = Plus / Plus

                                   
Query: 1       agagtttgatcctggctcag 20
               ||||||||||| ||||||||
Sbjct: 4206177 agagtttgatcatggctcag 4206196

I ordered both primers (the mismatch and the exact match primer).

7.3.1  RNAseH tests

Dec 5, 2007
As a first step to removing the rRNA using RNAseH + PCR amplicon (DNA) of rRNA, I took some total RNA from the -80C (sample 1 from the table in section 6.12.2 on page pageref and some ssRNA ladder [NEB]. I wanted to first check that the RNAse H would not degrade the RNA in the absense of DNA (NEB claims their RNAse H does not contain additional contaminating RNAses). I also wanted to see if I could degrade the RNA with RNAse H by adding a large amount of genomic DNA (which would presumably bind to the mRNA in my sample. I would not likely have enough genomic DNA to remove the rRNA bands, but perhaps everything except the rRNA bands would be degraded). I also assumed the RNA ladder would also not be degraded by RNAse H when genomic DNA was added for similar reasons to the rRNA explanation above. I used 1 mg of RNA ladder, 1 mg of total RNA, and 1 mg of genomic DNA sample 1 from section 5.1 on page pageref. The RNAse H degradation was for 30 min at 37C.
Since the RNA and genomic DNA samples were pretty old, I spec'd them again with the Nanodrop:
Sample DNA (ng/ul) 260/280 260/230
total RNA 2235.2
genomic 1 409.2
The samples were run on a 1% agarose gel (Figure ) for 35 minutes at 110 V.
Please see the pdf version for figures
Figure 7.2:
Brief Conclusions:   All the samples are such a blurry mess, its hard to say anything about this experiment (Figure 7.2). Is my RNA just being degraded? Is the gel not running properly? Why does the - control RNA ladder only sample look so bad (lane L/N/N)?

7.3.2  can I get a decent RNA ladder gel

Dec 6, 2007
Sanity check. Can I at least get a decent looking RNA ladder on an agarose gel? Do I have RNAse contamination somewhere that's really screwing me up? What's goin on?
I put 2 ml (1 mg ) of the RNA ladder into TE. The first sample I added ssRNA sample buffer [NEB], heated to 60C for 5 minutes; the second sample was run in standard sucrose agarose loading buffer/dye. I switched to a TBE gel and ran the gel for 40 minutes at 120V (Figure ).
Please see the pdf version for figures
Figure 7.3: Sample 1 contains NEB ssRNA sample buffer and was heated. Sample 2 contains standard sucrose loading buffer.
Brief Conclusions:   Certainly a step up. I can at least make out the bands of the ladder (compare Figure 7.3 with the lanes labeled 1 and 2 in the earlier Figure 7.2). Still not terribly good, but the ladder in the image that NEB sends as an example is also quote blury. The sample with the ssRNA sample buffer certainly looks less blurry. I'll continue to use the ssRNA sample buffer and I'll switch to TBE gels.

7.3.3  testing RNAse H on total RNA

Thur Dec 13, 2007
I've heard that RNAse H might have some contaminating RNAses which will degrade my RNA. When I was reading the superase manual, I noticed that superase has a nice feature that it inhibits most RNAses except RNAse H: beautiful. If my RNAse H enzyme has contaminating RNAses I should be able to inhibit them with the addition of superase to my total RNA, yet the RNAse H will still function.
To test this, I'm going to run standard RNAeasy total RNA preps. I'll add superase to some but not others and then test the effect of RNAse H on all of the samples. To add some complementary DNA to my samples (for the RNAse H to use), I'm going to prepare a few of the samples without using DNA removal (via LiCl precipitation).
MG1655 samples were grown 1 hr 50 min from a 1:50 dilution of overnight culture into LB. The OD600 for the two samples was 0.499 and 0.518 (not background subtracted). I used 2.5 ml of culture and 5 ml of RNAprotect. One of the two samples was placed at -20C for 30 minutes with 1/2 volume of 7.5 LiCl to precipitate the RNA (and thus remove the genomic DNA). The second sample was placed at 4C during this time.
Prior to LiCl the samples were eluted into 100 ml of RNAse free H2O , the yields were:
Sample DNA (ng/ul) 260/280 260/230 yield
sample 1 1122.5 112 mg
sample 2 759.6 76 mg
After the LiCl, sample 1 was resuspended into 50 ml of RNAse free H2O , the yield was:
Sample DNA (ng/ul) 260/280 260/230 yield
sample 1 1464.3 73.2 mg
From the above tables, sample 1 was split into 3 and becomes samples 1-3 below. Likewise, sample 2 becomes samples 4-6 below.
sample LiCl superase RNAseH
1 yes no yes
2 yes yes no
3 yes yes yes
4 no yes yes
5 no no yes
6 no yes no
1 ml (20U) of superase was used. 1 ml (5U) of RNAse H was used with an incubation at 37C for 15 min.
I ran approximately 1 mg of each sample on a 1% TBE agarose gel with ssRNA sample loading buffer [NEB] (Figure ).
Please see the pdf version for figures
Figure 7.4:
Brief Conclusions:   It looks like the RNAse H from NEB is so free of contaminating RNAses that the cool superase trick is unnecessary (Figure 7.4). Thankfully, both of the preps resulted in very clean total RNA with the characteristic 16S and 23S bands (Figure 7.4 lanes 1B and 2A). It's hard to say if the genomic DNA in samples 4-6 allowed the RNAse H to have any activity. If it did, then we'd expect sample 6 to look different than samples 4 and 5, which doesn't seem to be the case. Overall, not a bad result here though, as it looks like I don't have to worry about my RNAse H containing contaminating RNAses that will degrade the RNA.

7.3.4  testing RNAse H on total RNA with DNA oligos complementary to the rRNA

Dec 14, 2007
Initially, I intended to use PCR product of 16S and 23S from genomic DNA using the primers on page pageref. However, I tried a 100 ml reaction two times using the 16S E. coli, the 23S E. coli, and the 16S general primers, and I obtained little no DNA from these PCR reactions (reactions were for 35 cycles, annealing at 60C, extending 1 min at 72C):
Sample primers DNA (ng/ul) 260/280 260/230
1 (50 ml total volume) 16S general 6.0
2 (30 ml total volume) 16S E. coli6.6
3 (30 ml total volume) 23S E. coli24.5
4 (30 ml total volume) 16S general 0.5
5 (30 ml total volume) 16S E. coli3.3
6 (30 ml total volume) 23S E. coli-0.9
I the table above wins the prize for the worst PCR spec values I've ever had.
My thesis defense was in 3 days, and I really wanted to try and give this RNAseH idea a stab before my talk (I'm filling this part of the notebook in early Jan after the defense chaos has subsided). As a last minute, trick I realized that the MICROBExpressB kit from Ambion actually has DNA oligos that are complementary to my rRNA sequence - that's how they pull the rRNA down for removal with the magnetic beads. So I decided to try and use the MICROBExpressB oligos with my rRNA. The only trick was that with this proprietary kit, I don't known where along the rRNA the sequences bind.
I grew two cultures in 4 ml of LB in a 12 ml falcon tube starting from a 1:50 dilution of overnite culture. I grew the samples 1 hr and 53 minutes to an OD600 of 0.523 and 0.537 respectively (not background subtracted). I prepared two total RNA samples using an RNeasy kit and I removed the genomic DNA using LiCl. The samples were resuspended into 30 ml of RNAse free H2O . The yields were:
Sample DNA (ng/ul) 260/280 260/230 yield
total RNA 1 2200.0 66 mg
total RNA 2 1932.7 58 mg
Using 10 mg of the above total RNA (the maximum amount recommended by the MICROBExpressB kit), I ran samples in RNAseH buffer (1a and 1b) and in MICROBExpress binding buffer (2a and 2b and 3). I heated the samples to 70C for 10 min (I used a 25 ml volume, so the samples were placed in a thermocycler). The RNAseH buffer samples were: 5 ml RNA (2 mg /ml ), 4 ml MICROBExpressB oligo, 4.5 H2O , 1.5 ml RNAseH buffer. For the MICROBExpress binding buffer: 5 ml RNA (2 mg /ml ), 4 ml oligo, 6 ml MICROBExpress binding buffer.
I ran the samples on a 1% TBE gel. The first gel I ran with ssRNA sample loading buffer [NEB] (Figure A). The lanes ran completely wacko. I ran a second gel to see if I made the first gel wrong or something and I got the same poor migration. Before running the third gel, I did an EtOH precipitation to switch the buffers, because I had a hunch that the MICROBExpressB buffer was messing things up. The third gel (Figure B was fine.
Please see the pdf version for figures
Figure 7.5:
Brief Conclusions:   The RNAse H + MICROBExpress oligo seems to have work. The 23S and 16S bands are definitely cut in a systematic way (Figure 7.5. Looks like I need binding buffer like Ambion kit's MICROBExpress binding buffer (lanes 2a, 2b, and 3a), but I don't know what is in their buffer??? I guess they have more EDTA which prevents my sequence from getting degraded. Perhaps next run I'll try to anneal in EDTA, and then add RNAseH buffer+RNAse H. I should also run a few samples with the MICROBExpress binding buffer to test that too.

7.3.5  testing RNAse H on total RNA with DNA oligos designed by primer3

Mon Jan 7, 2008
Using the internal oligo design feature of Primer3, I designed 1000 oligos to the consensus sequence of 16S and 23S rRNA. The consensus sequence was created by aligning all of these rRNA sequences from E. coliand setting the nucleotide to N where there was not an exact match for the site across all rRNA species in E. coli. I then blasted the 1000 oligos for 16S and 23S against a database of all E. colimRNA sequences using blastn. I then removed all oligos from the with an eval > 25 (the purpose was to limit the nonspecific degradation that will occur if my DNA oligos binding RNA besides the rRNA I'm trying to destroy. I used a melting temp of 60C.
The primer3 files, blast database, and primer designs are available here. The final primers are contained in the files final_16S_primers and final_23S_primers.
As an initial test, I only ordered the primers that would allow my to chop the rRNA into thirds (i.e. the 23S will become 800bp and the 16S will become 500 bp if all primers cut successfully). If these tests are successful, I'll by all of the primers in a plate which should allow my to cut the rRNA into fragments of 80bp or less. I order the 23S primers at positions 931 and 1812; I ordered the 16S primers at positions 528 and 1078. For all four primers, I ordered the reverse complements as well so I can have a positive and negative control.
Wed Jan 9, 2008
I grew MG1655 in 2x4 ml of LB from a 1:50 dilution of overnite culture. After 2hr 38min, I took 2.5 ml samples and placed them in 5 ml of RNA protect. OD600 at this time was 0.697 and 0.739 for the two samples (not background subtracted). I used the RNeasy kit, eluted into 100 ml , ran a LiCl precipitation to remove the genomic DNA, resuspended into 31 ml of TE. The yields were:
Sample DNA (ng/ul) 260/280 260/230 yield
sample 1 1921.9 59.6 mg
sample 2 1596.5 49.5 mg
   how much primer to use?
I decided to use 1.5x of each oligo relative to the rRNA. I assumed that 16S and 23S each represent half of the total RNA population (e.g. in 10 mg of total RNA, 5 mg is 16S and 5 mg is 23S).
For the 23S rRNA, which is around 2500bp, I calculated:
20/2500 ×5mg ×1.5 » 60ng
At 100 mM the primers are at around 600 ng/ml ; at 10 mM they are around 60 ng/ml .
For this first experiment, I used 1 ml of 10 mM primer for both 23S and 16S. For samples 1-3, I used 10 mg of total RNA (from sample 1 above) in 25 ml of TES (TE + 50 mM NaCl). For samples 4-6, I used 10 mg of total RNA (from sample 2 above) in 25 ml of MICROBExpressB binding buffer (since this is what I knew worked last time). I heated the 6 samples to 70C for 10 minutes. To samples 1-3, I then added 25 ml of RNAse H buffer + 1.5 ml of RNAse H. To samples 4-6, I added 25 ml of MICROBExpressB binding buffer + 1.5 ml of RNAse H. I incubated all 6 samples at 37C for 15 minutes.
After the RNAse H digestion, I cleaned up the reaction with EtOH precipitation, and resuspended the RNA pellets in 20 ml of TE. The yields for the six samples were:
Sample DNA (ng/ul) 260/280 260/230 yield
RNAse H 1 429.7 8.59 mg
RNAse H 2 418.2 8.36 mg
RNAse H 3 414.2 8.28 mg
RNAse H 4 456.7 9.13 mg
RNAse H 5 402.2 8.04 mg
RNAse H 6 399.9 8.00 mg
I ran 2 ml of each sample on
Please see the pdf version for figures
Figure 7.6: Not quite all the way degraded, but the RNAse H digestion certainly seemed to work.
Brief Conclusions:   Looks like it worked! However, it looks like I had it backwards. The reverse complement oligos allowed it to cut, but the normal oligos did not. I'm very pleased with the specificity. True there were only 4 oligos in the mix, but there were a fair amount of them and they didn't seem to cut the RNA at all (compare lanes 1 and 3 and lanes 4 and 6, which contain the wrong oligo and no oligo respectively). On a more quantitative level, there is little to no difference between the RNA yields for the samples also (if there were no specific degradation, I'd expect the yields to go down for the samples with DNA oligos). So far, this is better than I imagined. One problem is that the 16S and 23S bands are not digested to completion. I need to try different concentrations of oligo and RNAse H to figure out the optimal ratios for complete digestion.

7.3.6  optimizing RNAse H and DNA oligo concentration

Thur Jan 10, 2008
After my success with the custom-made oligos, I want to try and completely eliminate the original 23S and 16S rRNA bands. It's not clear if I need to add more oligo, more RNAse H, or if the gel in Figure 7.6 is as good as I'm gonna get.
sample oligo amount (ml of 10 mM stock) RNAseH (ml )
1 0.5 1
2 2 1
3 8 1
4 1 0.5
5 1 1.5
6 1 4.5
I ran a final sample (7) in which I only placed 1 ml of 16S oligo and 1 ml of RNAse H to make sure that the 23S rRNA remained uncut in the absense of 23S oligo. I used 5 mg of total RNA in each using the total RNA sample 2 from the section above.
The oligos and total RNA were placed in 25 ml of TES (50 mM NaCl + TE). I melted the RNA at 70C for 10 min. And then I added the 25 ml of RNAse H buffer plus the appropriate amount of RNAse H and incubated for 15 minutes at 37C. I cleaned up the reaction with EtOH and resuspended into 20 ml of TE. The yields for the 7 samples were:
Sample DNA (ng/ul) 260/280 260/230 yield
RNAse H 1 244.0 4.88 mg
RNAse H 2 246.3 4.93 mg
RNAse H 3 329.4 6.59 mg
RNAse H 4 220.0 4.40 mg
RNAse H 5 220.7 4.41 mg
RNAse H 6 205.9 4.12 mg
RNAse H 7 211.5 4.23 mg
Note that as expected, the samples with the large amount of oligo added have a boost in concentration. I'm not sure why 4-7 have higher concentrations that 1-3.
I ran 2.5 ml of each sample on a 1% TBE gel with RNA sample buffer [NEB]. The gel was run at 120V for 50 min (Figure ).
Please see the pdf version for figures
Figure 7.7: I used the wrong oligos. It was the reverse complement oligos that worked (Figure 7.6 lanes 2 and 5).
Brief Conclusions:   Neither 23S or 16S was cut in this experiment (Figure 7.7). At first my confidence in this method was shattered. But then I had a look back at the experiment that worked (i.e. Figure 7.6 lanes 2 and 5) and I saw that it was the reverse complement oligos that worked. opps! At least I know that nonspecific degradation is still undetectable using different concentrations of RNAse H and the wrong oligos. I've already ordered an entire plate of appx 60 oligos tiled along the two rRNA genes, but now I'll have to order the reverse complements too. I can use the normal oligos as a good negative control. I'm also curious to know the specificity of this RNAse H digestion trick. Will it still degrade if my oligo has one mismatch? Does the location of the mismatch matter (presumably a mutation in the middle would prevent RNA binding and degradation while a mutation on one of the ends wouldn't change anything). Ok, now I need to redo this and figure out the proper concentrations of RNAse H and oligo...

7.3.7  optimizing RNAse H and DNA oligo concentration using reverse complement oligos

Fri Jan 11, 2008
The last section failed, because I used the forward oligos when I should've used the reverse complement ones. This experiment will be exactly the same 7 samples as above. However, I ran out of RNAse H, so that will come from a new tube and I ran out of total RNA, so I'll prep more and use the new sample.
I prepped two samples from overnite cultures using LiCl to remove the genomic DNA (see sections above for more info). Samples were resuspended into 30 ml of TE. I only spec'd the sample I used:
Sample DNA (ng/ul) 260/280 260/230
sample A 1795.7
The remaining sample was placed at -20C and I'll spec it if I use it later.
Again the conditions for all 7 samples are the same as the previous section. The yields of the 20 ml of sample after EtOH precipitation were:
Sample DNA (ng/ul) 260/280 260/230 yield
RNAse H 1 295.1 5.90 mg
RNAse H 2 322.8 6.46 mg
RNAse H 3 354.3 7.09 mg
RNAse H 4 254.3 5.09 mg
RNAse H 5 232.5 4.65 mg
RNAse H 6 264.4 5.29 mg
RNAse H 7 254.3 4.65 mg
2.5 ml samples were run on a 1% TBE agarose gel (Figure ) for 50 minutes at 120 V.
Please see the pdf version for figures
Figure 7.8: Different primer and RNAse H concentrations. Increasing primer concentration (lanes 1-3) seems more important than increasing RNAse H concentrations (lanes 4-6).
Brief Conclusions:   Experiment worked much better when I placed the proper oligos in the tube. I attained practically complete degradation of the original 23S band and mostly complete degradation of the 16S rrrNA band (Figure 7.8 lane 3) when 8 ml of each 10 mM primer was added. So for practical purposes, I probably need to have my primers at higher concentration when I move to much higher numbers of primers. RNAse H seems like its good to go with 1 ml per rxn. Last, when I only placed the 16S oligo, I only degraded the 16S rRNA (lane 7) which is reassuring. Can't wait for the tiled oligo plate to arrive.

7.3.8  testing RNAse H on total RNA with a PLATE DNA oligos complementary to the rRNA

Thu Jan 17 13:29:10 EST 2008
Both the reverse complement and the forward rRNA oligo plates have arrived. Again info about the primers can be found here: The primer3 files, blast database, and primer designs are available here. The final primers are contained in the files final_16S_primers_w_rc and final_23S_primers_w_rc. Brief information about the scripts is in the file Notes.txt;
Today, I'll test the forward and reverse primer plates on the total RNA samples to see the effect on rRNA removal. There are 17 16S primers and 30 23S primers. The density of the primer tiling is such that it should chop the rRNA into fragments of 80bp or less.
I'm combining the primers into two separate tubes (one for 16S and one for 23S), so I don't have to individual add the large cocktail of primers one-at-a-time. I'm placing 5 ml of each 16S primer + 15 ml of TE for 100 ml of 10 mM 16S rRNA oligo stock. Unfortunately, the number of oligos for the 23S is so high that I can't create a 10 mM stock (I'd need to have the oligos delivered at a higher concentration, but this was the highest concentration they'd give me for a 25nM scale synthesis). For the 23S, I'm placing 5 ml of each for 150 ml of 6.6666 mM 23S rRNA oligo stock.
quantifying the total RNA
I made the genomic-DNA-free RNA in the previous section, but I never quantified it. I spec'd it with the nanodrop before I started the RNAse H experiment.
Sample DNA (ng/ul) 260/280 260/230 yield
sample B 2485.7 74.6 mg
In the previous section, I found that adding more oligo help increase the amount of rRNA that was cut. I'm going to test the same thing in this experiment using all of the oligos form the rRNA oligo plates. Samples 1 and 2 will use the reverse complement plate, while 3 and 4 will use the forward plate. The two concentrations tested are the best performing from the previous section (i.e. 8 ml of 10 mM per rxn) and one-quarter of this optimal amount.
sample plate used oligo amount (ml of 10/6.66 mM stock) RNAseH (ml )
1 revcomp 8,12 1
2 revcomp 2,3 1
3 forward 8,12 1
4 forward 2,3 1
The primers and total RNA were combined with NaCl to make a total volume of 25 ml TES. The TES mixture was heated to 70C for 10 minutes. A 1 ml RNAse H + 24 ml RNAse H buffer mixture was added to all four samples, and they were incubated at 37C for 15 minutes.
After the RNAse H digestion, 1 ml of glycoblue was added to each sample and they were cleaned up with EtOH, resuspended in 20 ml of TE, and spec'd with the nanodrop:
Sample DNA (ng/ul) 260/280 260/230 yield
1 1222.1 24.4 mg
2 510.8 10.2 mg
3 1328.8 26.6 mg
4 507.2 10.1 mg
I prepared 2.5 ml of each sample to run on a 1% TBE gel (i.e. I mixed with RNA sample buffer [NEB]). I then added 1 ml DNAse-Turbo [Ambion] to samples 1 and 3, followed by a 10 min incubation at RT to see if I could get rid of some of the DNA oligo before running the samples on a gel. These were very unfavorable conditions for a DNAse reaction, since the sample was in TE, but we'll see. If it works in this situation, it should work in easier situations too (i.e. with less chelator).
I ran all 6 samples (4 normal + DNAse treated sample 1 and 3) on a 1% TBE gel for 1 hour (Figure ).
Please see the pdf version for figures
Figure 7.9: As expected the rRNA was degraded with the reverse complement oligo plate (lanes 1,2,1*).
Brief Conclusions:   Clearly the DNA is making up the bulk of my reaction now as even after cleanup I have up to 5x more nucleic acid than starting RNA (see the four sample spec table above). However, it appears that the lower oligo amount performs as well as the high amount (compare lanes 1 and 2), so we can stick with this lower amount in the future. The rRNA bands are completely destroyed with the reverse complement oligos (Figure 7.9 lanes 1, 2, and 1*). However, the rRNA didn't get degraded all the way down to 80bp; there is a strong smear from 500bp-50bp. I think it the RNAseH rRNA removal technique is going to be the sole technique for rRNA removal, it'll need to be a two pass where I run the digestion, remove short RNA/DNA oligos, then run a second digestion with a second oligo plate (double the density, bringing it down to at best a 40bp window); with that tight a window, I'm actually going to be tiling 1/3 of the rRNA. Perhaps this is unnecessary if I do a two pass with a pull down (e.g. MICROBExpress) followed by a RNAseH digestion like this one. I need those Ambion RNA columns to come in; they're supposed to remove small RNA pretty efficiently. I could also remove the short DNA downstream after the cDNA step. I do it anyways for the gel size-selection step. But I'd prefer to remove the short RNA earlier.
removing the short RNA with LiCl precipitation
Sat Jan 19, 2008
Rather than wait for the Ambion MEGACLEAR columns to arrive, I decided to precipitate the RNA from the above experiment with LiCl, which can also remove short RNA. Before starting a respec'd all of the samples to make sure the RNA hadn't degraded.
Sample DNA (ng/ul) 260/280 260/230
1 1247.5
2 519.1
3 1244.6
4 520.2
The specs were quite similar to the previous specs, so looks like there's no problem with RNA degradation. I precipitated all four samples with LiCl (added 90 ml of H2O and 50 ml of 7.5M LiCl) and eluted into 15 ml TE.
After the LiCl the specs were:
Sample DNA (ng/ul) 260/280 260/230 yield (ng)
1 5.0 75
2 33.5 502.5
3 37.0 555
4 46.9 703.5
I ran the LiCl precipitated samples on a TBE agarose gel alongside some of the original sample for comparison (Figure ).
Please see the pdf version for figures
Figure 7.10: As expected the rRNA was degraded with the reverse complement oligo plate (lanes 1,2,1*).
Brief Conclusions:   The LiCl certainly got rid of the DNA oligos (Figure 7.10 compare asterisk lanes vs non-asterisk lanes). It's hard to say if the precipitation also removed the short RNA fragment. It does appear like the RNA is less enriched at around 80bp after the LiCl (lanes 1* and 2*), but then again, the LiCl lanes are also quite a bit fainter, so it could be just do to the lane having less RNA overall. Hopefully the MEGAClear columns will do a better job and be easier.

7.3.9  running low on total RNA

I need to make more total RNA to keep pursuing these RNAse H based rRNA removal optimizations.
Mon Jan 21, 2008
I grew up 4 samples from a 1:50 dilution of overnite MG1655 culture. The cultures were grown in LB to an OD600 of 0.674, 0.661, 0.654, 0.586 respectively (OD not background subtracted). I added 2.5 ml of culture to 5 ml RNA protect, ran through a RNeasy RNA purification kit and used LiCl precipitation to remove genomic DNA.
I'll spec the samples as I use them for particular experiments downstream.

7.3.10  testing RNAse H on total RNA with a PLATE DNA oligos complementary to the rRNA and MEGAClear

Tue Jan 22, 2008
The Ambion MEGAClear columns have arrived, so hopefully they'll do the job of removing the low-MW RNA thats been chopped with RNAse H.
I'm running RNAse H samples, MicrobeExpress samples, and RNAse H + MicrobeExpress samples. The total RNA for samples 1-3 is from sample B on page pageref. The total RNA from samples 4-7 is from sample A in section 7.3.9 above. All MicrobeExpress samples used 50 ml beads and were run in PCR tubes in a thermocycler.
Sample DNA (ng/ul) 260/280 260/230
sample A from section 7.3.9 2064
The set up is (RC = reverse complement):
1 5 mg total RNA; 2 ml 16S RC oligo mix; 3 ml 23S RC oligo mix; cleanup with EtOH; resuspend in 15 ml TE
2 5 mg total RNA; 2 ml 16S RC oligo mix; 3 ml 23S RC oligo mix; cleanup with MEGAClear; cleanup with EtOH; resuspend in 15 ml TE
2b 5 mg total RNA; 1 ml 16S RC oligo mix; 1.5 ml 23S RC oligo mix; cleanup with MEGAClear; cleanup with EtOH; resuspend in 15 ml TE
3 5 mg total RNA; 2 ml 16S forward oligo mix; 3 ml 23S forward oligo mix; cleanup with MEGAClear; cleanup with EtOH; resuspend in 15 ml TE (this shouldn't cut because it uses the forward primers)
4 run 10 mg total RNA through MicrobeExpress in 100 ml (50 ml binding + 50 ml beads); clean up directly in MEGAClear; EtOH; elute in 12 ml
5 run 10 mg total RNA through MicrobeExpress in 100 ml (50 ml binding + 50 ml beads); add 0.6 ml 16S RC oligo, 0.9 ml 23S RC oligo; heat to 70C for 10 min; bind at 37C and add 1 ml RNAse H; clean up directly in MEGAClear; EtOH; elute in 12 ml
6 run 10 mg total RNA through MicrobeExpress in 200 ml (150 ml binding + 50 ml beads); EtOH; elute in 12 ml
7 run 10 mg total RNA through MicrobeExpress in 200 ml (150 ml binding + 50 ml beads); add 0.6 ml 16S RC oligo, 0.9 ml 23S RC oligo; heat to 70C for 10 min; bind at 37C and add 1 ml RNAse HEtOH; elute in 12 ml
yields from the first four samples were:
Sample DNA (ng/ul) 260/280 260/230 yield (mg ) loss
1 705.5 10.58 -112%
2 245.1 3.68 26%
2b 279.1 4.19 16%
3 305.9 4.59 8%
I ran 2.5 ml of each sample on a 1% TBE gel at 120V for 50 min (Figure ).
Please see the pdf version for figures
Figure 7.11:
yields from the second four samples were (estimated using Qubit values):
Sample DNA (ng/ul) 260/280 260/230 Qubit (ng/ml ) yield (mg ) loss
4 72.3 67.7 0.81 92%
5 46.1 56 0.67 93%
6 174.1 120 1.44 86%
7 309.0 91 1.09 89%
I ran all of each sample on a 1% TBE gel at 120V for 50 min (Figure ).
Please see the pdf version for figures
Figure 7.12:
Brief Conclusions:   The hybrid approach of using MICROBExpress followed by RNAse H looks like it might have real potential (Figure 7.12 lanes 5 and 7). The RNAse H looks like it helps remove more of the 23S and 16S bands. I'm not sure why the low-volume binding (lanes 4 and 5) doesn't have any of the high molecular weight RNA?

7.3.11  cloning and sequencing to test the RNAse H based rRNA removal methods

The results in the previous section look promising enough that I want to try and clone some mRNA enriched cDNA to get some counts on the proportion of the remaining RNA that is still rRNA.
Wed Jan 23, 2008
All experiments with oligos used 1 ml 16S RC oligo mix and 1.5 ml of 23S oligo mix. All samples were resuspended into 12 ml of 0.5X TE. For the samples that required removal or loss of large amounts of RNA, I pooled multiple samples to have enough RNA to make cDNA and clone it. The number of samples pooled for each rRNA removal method is indicated in the table below.
The total RNA for this experiment was sample A from the section above and sample B (see spec below).
Sample DNA (ng/ul) 260/280 260/230
sample B from section 7.3.9 2060
description pooled
1 10 mg total RNA; oligos; cleanup with MEGAClear; cleanup with EtOH 1
2 10 mg total RNA; 100 ml MICROBExpress (50 ml beads; 50 ml sample); oligos; cleanup with MEGAClear; cleanup with EtOH 2
3 10 mg total RNA; 200 ml MICROBExpress (50 ml ; 150 ml sample); oligos; cleanup with EtOH 2
4 10 mg total RNA; 200 ml MICROBExpress (50 ml beads; 150 ml sample); EtOH; oligos; cleanup with MEGAClear; cleanup with EtOH 3
The yields for the 4 samples (after pooling):
sample Qubit RNA conc. (ng/ml ) yield (ng)
1 > 200 -
2 97.8 1173
3 122 1464
4 61.4 736.8
I spec'd sample 1 with the nanodrop since it was too concentrated for the Qubit RNA dye:
Sample DNA (ng/ul) 260/280 260/230 yield (ng)
1 660.6 7927
I made the cDNA as specified in the Preparation of PET libraries Protocol (section on page ). I used all 11 ml for samples 2-5 with 1.5 ml superscript III; I used 5 ml of sample 1 with 3 ml superscript III. The following alterations were made to the PET cDNA library protocol: During 1st strand synthesis, I incubated at 50C for 45 minutes instead of 60. During 2nd strand synthesis, I incubated at 16C for 1hr 30 min instead of 2 hr. I did not heat inactivate the 2nd stand enzymes, rather I cleaned up the reaction directly with a Qiagen PCR purification column and eluted into 34 ml . I also did not heat inactivate the end-repair, as I also cleaned up this reaction directly with a Qiagen PCR purification kit and eluted into 30 ml . After which I quantified the DNA with the dsDNA HS Qubit kit:
sample Qubit RNA conc. (ng/ml ) yield (ng)
1 35.2 1056
2 36.6 1098
3 89.6 2688
4 21.6 648
I used 1 ml (appx 2.1 mg ) of BamHI adaptor.

Cloning the ds cDNA

Thur Jan 24, 2008
preparing the vector
I cut 2 mg of pUC19 with 1 ml BamHI, 2 ml 10x BSA, 13 ml H2O , and 2 ml NEB3 buffer for 45 min at 37C. I cleaned up the digestion with a Qiagen PCR purification kit, eluted into 30 ml and spec'd the cut plasmid:
Sample DNA (ng/ul) 260/280 260/230
cut puc19 51.3
I dephosphorylated 10 ml of the pUC19 with 7 ml H2O , 2 ml antararctic phosphatase buffer, and 1 ml antaratic phosphatase at 37C for 30 min followed by heat inactivation at 65C for 5 minutes.
size-selecting the cDNA
I cleaned up the adaptored cDNA with a Qiagen PCR purification kit, eluted into 30 ml and ran all of it on a 65 ml TAE sybrsafe gel for 25 min at 90V with a PCR ladder and a 10-well wide comb. I gel purified with 550 ml of QG buffer and a Qiagen column.
I measured the final size-selected cDNA yields with a Qubit dsDNA HS kit:
sample Qubit RNA conc. (ng/ml ) yield (ng)
1 3.65 110
2 2.73 81.9
3 10.0 300
4 5.26 158
Brief Conclusions:   Once again, I was wishing I'd run the gel a little longer or a little hotter; Those adaptors are really in the way. Perhaps I should either use less adaptor or use the Purelink HS kit from invitrogen to try and eliminate more of the adaptor.
ligating the cDNA
I ligated the dephosphorylated vector to the adaptored cDNA using 10 ml of the gel purified cDNA, 2 ml of T4 ligase buffer, 2 ml of the dephosphorylated vector, 5 ml H2O , and 1 ml of T4 DNA ligase at 16S for 30 minutes.
transforming the ligation
I transformed 2 ml of each of the four ligation into DH5alpha competent cells. I plated 50 ml and 100 ml of each on amp plates containing Xgal.
picking colonies
Fri Jan 25, 2008
I picked 24 colonies. I planned on picking 10 of each of the four samples, but I had no colonies for sample 3 and only three colonies for sample one.
second transformation
Since, I didn't have enough colonies, I retransformed an additional 2 ml of the ligation from the previous day. This time I used higher efficiency competent cells (oneshot TOP10 [Invitrogen]). I plated 100 ml and 200 ml of each of the four transformations. The TOP10 cells actually expired in Sep 07, hopefully they're still ok.
first minipreps
Sat Jan 26, 2008
I miniprepped all 24 samples.
second picking colonies
The higher efficiency TOP10 transformations worked well. Plenty of colonies for all four samples for the initial sequencing and for a potential downstream plate of sequencing. I picked 18 colonies (sample 1 and 3 since that was what was missing).
second minipreps
Sun Jan 27, 2008
I minipreped the 18 colonies picked yesterday.
spec/digest/sequence
I spec'd all 42 samples with the Nanodrop. The nomenclature: first number = sample ID (from one of the four rRNA removal techniques); second number = plate colony was picked from (e.g. 100ml is the plate with 100 ml from the DH5alpha transformation; T = TOP10); final letter = colony picked from the plate (in alphabetical order; colony 1 = a, colony 2 = b, etc...).
I digested 6 ml of 2 of each sample type with 1 ml of HindIII and 1 ml of EcoRI using EcoRI buffer just to verify that the inserts were of a decent size (Figure ).
Please see the pdf version for figures
Figure 7.13:
The trimmed sequenced reads were blasted against the E. coligenome to determine the cDNA match. The raw sequence reads are available here and here. Note that I screwed up the nomenclature for the sample 4 when I submitted the sequence names to agencourt. I used 4-T100-a through 4-T100ul-j; The sample names in the table below are correct.
RNAse H rRNA removal results
Sample ID ng/ul 260/280 260/230 yield (ug) blast match
1.50ul.a 286.39 1.95 2.26 14.3 rrlE 23S
1.50ul.b 233.78 1.93 2.2 11.7 rrlE 23S
1.50ul.c 214.58 1.97 2.23 10.7 rrsE 16S
1.T100ul.a 134.47 1.95 2.09 6.7 rrsE 16S
1.T100ul.b 218.73 1.97 2.25 10.9 rrlE 23S
1.T100ul.c 127.08 1.98 2.26 6.4 rrsA 16S
1.T100ul.d 171.97 1.97 2.26 8.6 rrsE 16S
1.T100ul.e 195.41 1.94 1.94 9.8 rrlA 23S
1.T100ul.f 179.58 1.99 2.21 9.0 rrsE 16S
1.T100ul.g 279.68 1.94 2.24 14.0 rrlE 23S
2.100ul.a 227.17 1.95 2.26 11.4 rrlE 23S
2.100ul.b 223.11 1.94 2.11 11.2 rrsE 16S
2.100ul.c 371.54 1.91 2.18 18.6 rrlA 23S
2.100ul.d 411.8 1.91 2.23 20.6 rrlA 23S
2.100ul.e 220.31 1.93 2.06 11.0 rrlE 23S
2.100ul.f 199.03 1.94 2.01 10.0 rrsE 16S
2.100ul.g 328.28 1.93 2.25 16.4 rrlE 23S
2.100ul.h 234.85 1.95 2.27 11.7 rrlE 23S
2.100ul.i 265.44 1.94 2.07 13.3 rrsE 16S
2.100ul.j 214.93 1.94 2.22 10.7 rrlE 23S
3.T100ul.a 152.71 1.94 1.88 7.6 rrlE 23S
3.T100ul.b 245.84 1.95 1.98 12.3 rrlE 23S
3.T100ul.c 158.01 1.98 2.2 7.9 rrlE 23S
3.T100ul.d 174.04 1.95 1.98 8.7 rrsE 16S
3.T100ul.e 87.93 1.95 2.05 4.4 -
3.T100ul.f 177.74 1.95 2.2 8.9 rrlE 23S
3.T100ul.g 159.02 1.95 2.17 8.0 rrlE 23S
3.T100ul.h 176.91 1.99 2.02 8.8 rrlE 23S
3.T100ul.i 116.5 1.97 2.13 5.8 rrlE 23S
3.T100ul.j 174.56 1.96 2.06 8.7 rrsE 16S
3.T100ul.k 257.99 1.94 2.22 12.9 -
4.50ul.a 261.6 1.94 2.25 13.1 rrlE 23S
4.50ul.b 386.9 1.91 2.22 19.3 rrlE 23S
4.50ul.c 257.63 1.94 2.25 12.9 rrlE 23S
4.50ul.d 306.41 1.93 2.19 15.3 rrlE 23S
4.50ul.e 328.23 1.93 2.23 16.4 rrlE 23S
4.100ul.a 261.63 1.95 2.25 13.1 mukF
4.100ul.b 274.39 1.95 2.22 13.7 mukF
4.100ul.c 218.56 1.97 2.28 10.9 rrlA 23S
4.100ul.d 315.31 1.89 1.81 15.8 cheA
4.100ul.e 280.3 1.94 2.27 14.0 rrlE
4.100ul.f 203.29 1.97 2.22 10.2 -
I thought it was weird that there were two mukF matches for sample 4. After an alignment (below), I think that I either picked the same colony twice or somehow two identical colonies were right next to each other on the plate (rather than two independent ligation events, which is what I'm really interested in).
CLUSTALW ALIGNMENT OF the two mukF cDNA matches

4-T100ul-g.trim      ----------GAGACACTTGCCGTGT-CAAAACCAGACAAGTGCCGCTGG 39
ecoli_genome         ---------------------------CAAAACCAGACAAGTGCCGCTGG 23
4-T100ul-f.trim      ATCCGACCGAAGACAACTTGCCGTGTACAAAACCAGACAAGTGCCGCTGG 50
                                                ***********************

4-T100ul-g.trim      ATCTTGGTCTGGTGGTACGCGAATATCTGTCACAGTATCCGCGTGCACGT 89
ecoli_genome         ATCTTGGTCTGGTGGTACGCGAATATCTGTCACAGTATCCGCGTGCACGT 73
4-T100ul-f.trim      ATCTTGGTCTGGTGGTACGCGAATATCTGTCACAGTATCCGCGTGCACGT 100
                     **************************************************

4-T100ul-g.trim      CACTTTGACGTTGCGCGTATTGTTATTGATCAGGCGGTACGTCTTGGCGT 139
ecoli_genome         CACTTTGACGTTGCGCGTATTGTTATTGATCAGGCGGTACGTCTTGGCGT 123
4-T100ul-f.trim      CACTTTGACGTTGCGCGTATTGTTATTGATCAGGCGGTACGTCTTGGCGT 150
                     **************************************************

4-T100ul-g.trim      AGCGCAAGCAGATTTCACCGGACTGCCAGCGAAATGGCAGCCGATTAATG 189
ecoli_genome         AGCGCAAGCAGATTTCACCGGACTGCCAGCGAAATGGCAGCCGATTAATG 173
4-T100ul-f.trim      AGCGCAAGCAGATTTCACCGGACTGCCAGCGAAATGGCAGCCGATTAATG 200
                     **************************************************

4-T100ul-g.trim      ATTACGGAGCCAAGGTACAGGCGCATGTCATCGACAAATATTGAACAAGT 239
ecoli_genome         ATTACGGAGCCAAGGTACAGGCGCATGTCATCGACAAATATTGAACAAGT 223
4-T100ul-f.trim      ATTACGGAGCCAAGGTACAGGCGCATGTCATCGACAAATATTGAACAAGT 250
                     **************************************************

4-T100ul-g.trim      GATGCCGGTTAAGCTGGCGCAGGCGCTGGCGAATCCGTTATTTCCGGCGC 289
ecoli_genome         GATGCCGGTTAAGCTGGCGCAGGCGCTGGCGAATCCGTTATTTCCGGCGC 273
4-T100ul-f.trim      GATGCCGGTTAAGCTGGCGCAGGCGCTGGCGAATCCGTTATTTCCGGCGC 300
                     **************************************************

4-T100ul-g.trim      TGGACAGCGCCTTACGTTCAGGACGCCATATTGGCCTCGACGAACTGGAT 339
ecoli_genome         TGGACAGCGCCTTACGTTCAGGACGCCATATTGGCCTCGACGAACTGGAT 323
4-T100ul-f.trim      TGGACAGCGCCTTACGTTCAGGACGCCATATTGGCCTCGACGAACTGGAT 350
                     **************************************************

4-T100ul-g.trim      AATCATGCATTCCTGATGGATTTTCAGGAATATCTGGAAGAGTTTTACGC 389
ecoli_genome         AATCATGCATTCCTGATGGATTTTCAGGAATATCTGGAAGAGTTTTACGC 373
4-T100ul-f.trim      AATCATGCATTCCTGATGGATTTTCAGGAATATCTGGAAGAGTTTTACGC 400
                     **************************************************

4-T100ul-g.trim      GCGTTATAACGTTGAGCTTATTCGCGCACCAGAAGGGTTCTTCTATTTAC 439
ecoli_genome         GCGTTATAACGTTGAGCTTATTCGCGCACCAGAAGGGTTCTTCTATTTAC 423
4-T100ul-f.trim      GCGTTATAACGTTGAGCTTATTCGCGCACCAGAAGGGTTCTTCTATTTAC 450
                     **************************************************

4-T100ul-g.trim      GCCCACGTTCCACCACGCTGATCCCTCGTTCCGTCTTGTCGGAACTGGAT 489
ecoli_genome         GCCCACGTTCCACCACGCTGATCCCTCGTTCCGTCTTGTCGGAACTGGAT 473
4-T100ul-f.trim      GCCCACGTTCCACCACGCTGATCCCTCGTTCCGTCTTGTCGGAACTGGAT 500
                     **************************************************

4-T100ul-g.trim      ATGATGGTCGGGAAAATCCTCTGTTATCTCTATCTCAGCCCGGAACGGCT 539
ecoli_genome         ATGATGGTCGGGAAAATCCTCTGTTATCTCTATCTCAGCCCGGAACGGCT 523
4-T100ul-f.trim      ATGATGGTCGGGAAAATCCTCTGTTATCTCTATCTCAGCCCGGAACGGCT 550
                     **************************************************

4-T100ul-g.trim      GGCGAATGAGGGGATTTTCACCCAGCAGGAACTGTACGACGAACTGCTCA 589
ecoli_genome         GGCGAATGAGGGGATTTTCACCCAGCAGGAACTGTACGACGAACTGCTCA 573
4-T100ul-f.trim      GGCGAATGAGGGGATTTTCACCCAGCAGGAACTGTACGACGAACTGCTCA 600
                     **************************************************

4-T100ul-g.trim      CCCTGGCCGATGAAGCAAAACTGCTGAAACTGGTGAACAACCGTTCAACC 639
ecoli_genome         CCCTGGCCGATGAAGCAAAACTGCTGAAACTGGTGAACAACCGTTCAACC 623
4-T100ul-f.trim      CCCTGGCCGATGAAGCAAAACTGCTGAAACTGGTGAACAACCGTTCAACC 650
                     **************************************************

4-T100ul-g.trim      GGTTCAGACGTTGACCGTCAGAAGTTGCAGGAGAAAGTACGTTCTTCGCT 689
ecoli_genome         GGTTCAGACGTTGACCGTCAGAAGTTGCAGGAGAAAGTACGTTCTTCGCT 673
4-T100ul-f.trim      GGTTCAGACGTTGACCGTCAGAAGTTGCAGGAGAAAGTACGTTCTTCGCT 700
                     **************************************************

4-T100ul-g.trim      CAACCGTCTGCGTCGTTTAGGCATGGTGTGGTTGTCGGATTCGGATCCTC 739
ecoli_genome         CAACCGTCTGCGTCGTTTAGGCATGGTGTGGTT----------------- 706
4-T100ul-f.trim      CAACCGTCTGCGTCGTTTAGGCATGGTGTGGTTGTCGGATTCGGATCCTC 750
                     *********************************                 

4-T100ul-g.trim      TAGAGTCGACCTGCAGGCATGCAAGCTTGGCGTAATCATGGTCATAGCTG 789
ecoli_genome         --------------------------------------------------
4-T100ul-f.trim      TAGAGTCGACCTGCAGGCATGCAAGCTTGGCGTAATCATGGTCATAGCTG 800
                                                                       

4-T100ul-g.trim      TTTCCTGTGTGAAATTGTTATCCGCTCACAATTCCACACAACATACGAGC 839
ecoli_genome         --------------------------------------------------
4-T100ul-f.trim      TTTCCTGTGTGAAATTGTTATCCGCTCACAATTCCACACAACATACGAGC 850
                                                                       

4-T100ul-g.trim      CGGAAGCATAAAGTGTAAAGCCTGNGGTGCCTAATGAGTGAGCTAACTCA 889
ecoli_genome         --------------------------------------------------
4-T100ul-f.trim      CGGAAGCATAAAGTGTAAAGCCTGGGGTGCCTAATGAGTGAGCTAACTCA 900
                                                                       

4-T100ul-g.trim      CATTAATTGCGTTGCGCTCACTGCCCGCTTTCCAGTCGGGAAACCTGTCG 939
ecoli_genome         --------------------------------------------------
4-T100

Brief Conclusions:   That really sucked!
I was hoping to have a range of success rates with each of the four rRNA removal strategies. But three of them failed to produce a single mRNA in 10 sequencing reactions (see table above). Method 4, using 200 ml MICROBexpress, EtOH; RNAse H + oligos; cleanup with MEGAClear; and cleanup with EtOH performed the best with 2 in 9 samples being mRNA (22%). I'm going to sequence a half a plate or more of these samples to try and determine more precisely the true rRNA proportion. Since I didn't run any samples without the RNAse H, its hard to know if the RNAseH really helped or not (it appeared to do so in the gel in the previous section, but
Thoughts for next round:
try: 1) two rounds of MICROBExpress with EtOH (as I did before to get around 28%); 2) two round of MICROBExpress with EtOH followed by RNAseH and MEGAClear (i.e. is the RNAse H helping at all?); 3) one round of MICROBExpress followed by 2 rounds of RNAse H (using the two different RC plates and a MEGACLEAR in between); 4) three rounds of MICROBExpress with EtOH
In addition, I'm working on biotinylating the oligos from the oligo plate to see how well it works to try MICROBExpress followed by a second type of pull down with a different oligo set (or perhaps 3 different oligo sets).
All of this will require quite a large number of sample poolings to ensure I have enough cDNA to clone.

7.3.12  specificity of RNAse H oligos

I want to know how much non-specific degradation I'm getting from the oligos. I purposely chose oligos that had the least similarity with genomic regions, but I still don't know how well an oligo has to match to allow the RNAse H to cut. To gain a little insight into this potential degradation, I developed 10 versions of the center-most 16S oligo from the first reverse complement oligo plate. Successful directed RNAse H degradation at the oligo should roughly break the 16S in half which makes for an easy gel based assay of degradation ability. Each variant has a mutation at one location along the primer. The changed nucleotide position is labeled with a * and the randomized order number for the primer is presented on the left:
# original
4) TTTACGGCGTGGACTACCAG

# two center changes
        # change G to anything else (one at a time)
2)      TTTACGGCGTCGACTACCAG
6)      TTTACGGCGTAGACTACCAG
9)      TTTACGGCGTTGACTACCAG
                  *

        # change T to anything else (all together)
1)      TTTACGGCGVGGACTACCAG
                 *

# 5' changes
        # 5 from the end
7)      TTTADGGCGTGGACTACCAG
            *

        # 3 from the end
10)     TTVACGGCGTGGACTACCAG
          *

        # 1 from the end
3)      VTTACGGCGTGGACTACCAG
        *

# 3' changes
        # 6 from the end
11)     TTTACGGCGTGGACVACCAG
                      *

        # 3 from the end
5)      TTTACGGCGTGGACTACDAG
                         *

        # 1 from the end
8)      TTTACGGCGTGGACTACCAH
                           *

1 center T®V
2 center G®C
35' 1bp
4original
53' 3bp
6center G®A
75' 5bp
83' 1bp
9center G®T
105' 3bp
113' 6bp
should I try shorter lower MT oligos?

7.3.13  initial testing of the 16S mutation oligos

Thur Feb 7, 2008
I tested all 11 variants (including the non-mutation version) using the standard RNAse H protocol I've been following: 4 ml of 10 mM oligo, 5 mg of total RNA, in 25 ml total volume TES; 10 min 70 C; add 25 ml RNAse H buffer, H2O , with 1 ml RNAse H; 37 C for 15 minutes.
For the total RNA, I spec'd and used sample C from section 7.3.9:
Sample DNA (ng/ul) 260/280 260/230
sample C from section 7.3.9 2982.5
I cleaned up all samples with EtOH and glycoblue and eluted into 15 ml of TE:
Sample DNA (ng/ul) 260/280 260/230
1 282.3
2 290.6
3 256.3
4 271.1
5 332.7
6 258.0
7 270.1
8 269.5
9 265.7
10 264.6
11 287.1
I ran 2.2 ml of all 11 samples on a 1% TBE agarose gel for 50 min at 120V (Figure ).
Please see the pdf version for figures
Figure 7.14: mutation 16S oligos with RNAse H; what's going on with the oligo in lane 5 (3' 3bp)? The lanes are: 1) center T®V; 2) center G®C; 3) 5' 1bp; 4) original; 5) 3' 3bp; 6) center G®A; 7) 5' 5bp; 8) 3' 1bp; 9) center G®T; 10) 5' 3bp; 11) 3' 6bp
Brief Conclusions:   The original unmutated primer cut as expected (Figure 7.14 lane 4) - although it didn't cut as completely as I expected. If the mutation is in the middle (i.e. position 10 or 11 of the 20-mer oligo) the oligo doesn't noticably cut the 16S band regardless of the substitution (Figure 7.14 lanes 1, 2, 6, 9). On the other extreme, a subtitution on the last basepair on either end of the oligo (i.e. position 1 or 20) results in cutting (lanes 3 and 8). The remaining primers all seem to have cut to some extent except the 5' 3bp mutation (lane 10), which is weird because the 5' 5bp mutation (which should be less able to bind than the 3bp version) did cut.
The complete odd ball is lane 5 where the 3' 3bp completely sheared the RNA. Did some RNAse get in there? Is the oligo perhaps poorly synthesized so that it contains a large number of smaller oligos that bind everywhere (and result in non-specific RNA degradation?). Is this type of thing common with any large set of synthesized oligos, so that I need to individually screen my oligo library for the "good" ones?

7.4  rRNA removal via ultracentrifugation

7.5  rRNA removal via column chromatography