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 6
Gene and operon boundary determination by barcoded paired-end-tags and highly parallel sequencing

We want to be able to determine the transcriptional units for many species in a single sequencing run on a highly parallel sequencer (e.g. 454 or solexa).
Paired-end-tags provide an efficient mechanism to determine the 5' and 3' ends of a gene (Figure 5.18). By analyzing the sequenced tags, we should be able to get a rough idea of the transcriptional boundaries (the precise ends will probably be a little fuzzy because making full length cDNA is not always possible; Figure ). Hopefully, the sequence will also provide a new means to quantify gene expression that provides a better present/absent metric than microarrays. For the first time we'll be able to determine in if all of the hypothetical genes are really transcribed. The method should also allow transcript determination/quantification in mixed cultures of undefined composition setting the stage for metatranscriptomic (Figure Mixed Culture).
Please see the pdf version for figures
Figure 6.1: Paired-end-tags are mapped back onto the appropriate genome providing estimates of transcript boundaries (operons) and gene quantity.
By placing an error-correcting barcode on each sample, we can take advantage of the growing capacity of highly parallel sequencers to sample multiple conditions at the same time (Figure ). A barcode is just a piece of known DNA what we ligate onto our sample to identify it. For example, you could ligate AGA to the front of one some human DNA sequence and TAT to the front of some chimp DNA sequence, and then when you get the sequencing read back with AGA you know you have your human sample. The problem of course is that sequencing errors could lead to the misidentification of your sample. Because of this, I plan on using error-correcting barcodes. Error-correcting codes have been used for many years to keep things like telephones, CDs, and DVDs function properly in the prescence of noise and other problems (e.g. most CDs will work even with small stratches because there is enough redundancy built in that it can fill in the missing bits.
The simplest error correcting code would be if we replicated our code: AGA ® AAAGGGAAA; TAT ® TTTAAATTT. With this coding schema as long as you get less than one error every three base-pairs, you can recover your original sequence (e.g. AATGGGAAA = AGA). This simple method is a horribly inefficient way to do things that is good for teaching, but never used in practice. In practice, I have some software for generating a set of error-correcting DNA barcodes with a set level of misidentified barcodes (e.g. you can say you want 1 in 1,000,000,000 barcodes to fail and this places a constraint on either the length of the barcode or the number of barcodes you can make). No error correcting code is flawless, to be flawless, you'd need an infinitely long code. Since DNA has for bases = 4N barcodes of length N, it isn't too difficult to generate code codes without using too much of your sequencing read (6-10bp).
Please see the pdf version for figures
Figure 6.2: Depending on how quantitative this method is, it may be possible to use it for metatranscriptomics.
This chapter was started:
Wed Aug 23 14:13:07 EDT 2006

6.1  Planning and Goals

  1. make ds cDNA, pUC clone, TOPO clone, and sequence 10-20 (should be mostly rRNA but NOT 6S)
  2. make ds Cdna, added linkers, pUC clone, and sequence 10-20 (should be mostly rRNA but NOT 6S)
  3. make ds cDNA, added linkers, size-select pUC clone, and sequence 10-20 (should be mostly rRNA but NOT 6S)
  4. use Ambion kit to remove 16 and 22 S rRNA, make cDNA, add linkers, pUC clone, sequence 10-20 (should NOT be mostly rRNA), if rRNA is pretty rare, sequence more (see if they'll colony pick for me)
  5. take rRNA mRNA, make cDNA, A-tail? or add linkers?, circularize, amplify, cut with MmeI, PAGE purify, clone tags and sequence a few
  6. take rRNA mRNA, make cDNA, A-tail? or add linkers?, circularize, amplify, cut with MmeI, PAGE purify, clone tags, add linkers and sequence a few, if works sequence 96
size fractionate? http://wheat.pw.usda.gov/ lazo/methods/uo/pro1.html http://www.genome.ou.edu/protocol_book/protocol_partI.html pack beads into a spin-x column?
or use a cDNA fractionation colume from invitrogen ($29 apiece!)

6.1.1  Progress Reports on above enumeration

  1. done; doesn't work very well to do TA-style cloning on cDNA. Cloning was very inefficient. All plasmids submitted for sequencing failed.
  2. done, worked well. all sequences were 23S rRNA
  3. done, size-selection improved the insert size. all sequences were 23S rRNA or 16S rRNA
  4. transformation efficiency was very poor. Only 1 in 16 of picked white colonies had an insert. need to try again with more starting RNA
    Update Got it working better with higher transformation. Need to sequence more, but did get my first sequence that was NOT rRNA (1 out of 4; other three sequences were rRNA)
  5. skipped number 5 because might was well do number 6; did get the proper 70 mer tag, just didn't want to sequence it until I put on the adaptors in step 6.
  6. having troubles getting enough material to see it on a gel. I'm going to deviate a little from the Shendure at all list.

6.2  Cloning double-stranded cDNA starting from total RNA

I'm going to try and make double-stranded cDNA using slight modifications of standard approaches. In particular, I'm going to try and TA-clone the cDNA by A-tailing it with Taq and T-tailing a vector with Taq. I'm also going to try using a TOPO kit with the A-tailed cDNA. I want to use A-tailing, because that is the method employed by Shendure et.al. in their polony sequencing technique and will allow an easy transition from classic cDNA library protocols to a new polony protocol.

6.2.1  First steps

want > = 10-50 mg of total RNA (the binding capacity of the RNAeasy Plus column is 100 mg ).
For now use 5 mg which is the max amount for the SuperScript II RT protocol.
crude protocol:
  1. RNAprotect
  2. lyse with lysozyme and proteinase k
  3. RNAeasy
  4. acid-phenol (remove more genomic)
  5. LiCl (remove short RNA + clean up residual phenol)
  6. remove 16S and 22S (using Ambion kit) [skip for now]
  7. run 500 ng RNA on 1% gel
  8. first strand 5 mg total RNA (or 500 ng mRNA) using Superscript II; use 200 units per mg of (mRNA)
  9. take sample, RNA digest, ethanol precipitate, resuspend 10 ml , quantify, gel 9 ml (run all samples on same gel? 1st and 2nd strand synthesis results)
  10. second strand using Sambrook and NEB enzymes
  11. take sample, RNA digest, ethanol precipitate, resuspend 10 ml , quantify, gel 9 ml
  12. Qiagen PCR purify (to remove short stuff and enzymes)
  13. A-tag with taq and dATP (as in Shendure sequencing protocol)
  14. TOPO clone
  15. sequence
Detailed protocol and results:

Growing cells, RNAprotect

Wed Aug 30 10:33:56 EDT 2006
Grow 20 ml of LB with 1/100 dilution. Grow 20 ml of Davis with glucose with 1/25 dilution. Take samples in log phase. Add two volumes of RNAprotect. Vortex 5 sec, incubate at RT for 5 min, centrifuge at max rpm for 12 minutes.
Growing 6 samples 2 conditions: LB log phase and Davis minimal 0.5% glucose log-phase.
Started samples at 10:20 AM.
Italics indicates OD where samples were taken. For each condition I used 2 ml of culture and 4 ml of RNAprotect.
OD of cultures for 6 cDNA growth conditions
min Davis A Davis B Davis C LB A LB B LB C
40 0.099 0.101 0.101 0.035 0.035 0.037
90 0.109 0.105 0.111 0.161 0.17 0.177
120 0.113 0.107 0.113 0.302 0.323 0.317
145 0.116 0.114 0.118 0.433 0.467 0.457
170 0.121 0.119 0.122 0.676 0.74 0.721
255 0.189 0.196 0.206 1.29 1.346 1.325
320 0.307 0.322 0.336 1.373 1.606 1.577
375 0.494 0.514 0.523 1.661 1.61 1.665
395 0.498 0.523 0.552 1.44 1.683 1.624
470 0.832 0.887 0.9 1.553 1.73 1.666
Raw data in excel format.
Please see the pdf version for figures
Figure 6.3: Growth curve for the 6 samples to be used for making ds cDNA

DNA free RNA prep

I really want zero DNA (or at least completely degraded DNA). I'm going to use the RNAeasy kit, which uses a DNA binding column to get rid of genomic DNA. Then I'm going to use the TURBO DNA-free kit. Then I'm going to acid-phenol purify the solution (acid phenol moves DNA to the organic phase). Last I'm going to do LiCl precipitation, which does not precipitate the DNA (or RNA less than 200 bp).
Samples were randomized before RNAeasy preps as:
Condition Original ID Randomized ID
Davis and 0.5% glucose sample A 1 6
Davis and 0.5% glucose sample B 2 1
Davis and 0.5% glucose sample C 3 3
LB sample A 4 2
LB sample B 5 5
LB sample C 6 4
schema (much of this comes from the Ambion TOTALLY RNA kit manual):
  1. Lyse cells in 100 ml of TE with 1 mg/ml lysozyme. Incubate 2 min, vortex every minute. Add 10 ml Proteinase K. Incubate 3 more minutes, vortex every minute.
  2. add 350 ml RLT (with b-ME added) and follow the RNAeasy kit; elute with 50 ml 2 times (100 ml total)
    Sample DNA (ng/ul) 260/280 260/230 total yield
    sample 1 560.8 56.1 mg
    sample 2 538.7 53.9 mg
  3. follow DNA-free TURBO kit instructions for high-conc DNA. Briefly: add Buffer, add 1 ml DNAse, incubate 30 min, add additional 1 ml DNAse, incubate 30 more minutes. Deactivate and keep supernatant.
  4. add 200 ml TE. add 1/10 volume of sodium acetate. mix well. add 1 volume of Acid Phenol. Vortex 1 minute. Centrifuge 3 minutes at 12000 x g.
  5. transfer the upper, aqueous phase to a new eppy tube
  6. add 1 volume of isopropanol place at -20 C for 30 minutes
  7. resuspend in 50 ml of TE
  8. add 25 ml (1/2 volume) of 7.5 M LiCl; place at -20° C for 30 minutes. centrifuge at max rpm for 15 minutes
  9. wash pellet in 1 ml 70% ethanol, resuspend in RNAse free H2O .
  10. resuspend in 30 ml of RNAse free H2O .
    Sample DNA (ng/ul) 260/280 260/230 total yield super-DNA removal loss
    sample 1 638.1 19.1 mg 66%
    sample 2 347.8 10.4 mg 81%
I saved 600 ng of each to run on a gel (see Figure ).
Brief Conclusions:   RNA prep pretty good. Will know better after I run a gel. Using 2 ml of 0.5 OD gave a pretty good yield. I'd probably even bump it up to 2.5 next time, since I'm still only half way to maxing out the column (which holds 100 mg ). Loss from the DNA-removal stuff was pretty high (more than half). It isn't clear how much is due to initial DNA contamination and how much is due to all the manipulation. I'd guess most is from all the processing I did. Also, I used LiCl to remove the small RNAs. I know the Qiagen kit claims they remove them for you, but I think I probably loss quite a bit of small stuff at the LiCl stage too. Last the RNA was quite a bit dirtier after the processing. I think this is probably left over LiCl (hopefully not left-over phenol).
Brief Update Sat Sep 2 21:49:37 EDT 2006:   I should have saved some of the initial post-genomic-removal RNA. The RNA certainly looks degraded (Figure ), but I don't know where it became degraded. The 23S and 16S rRNA aren't present and the RNA smear looks like there's quite a lot of degradation. However, the cDNA lanes look longer than the rRNA that they are derived from, so maybe the RNA degraded after the cDNA was made from it. Next round I don't believe I'll use the acid-phenol step. I may move the LiCl step before the DNAse step. Next time I hope to start with more rRNA so hopefully I can use 5 mg instead of 3.5 mg

First strand synthesis of cDNA

Use Superscript II and the corresponding protocol:
Do in PCR tubes:
  1. add 1 ml of random hexamers (100 ng)
  2. add 1 ml of dNTP (10 mM each)
  3. add 3.5 mg RNA 8
  4. add H2O to 12 ml
  5. heat to 65° C for 5 minutes, chill on ice, brief centrifuge
  6. add 4 ml First-strand buffer, 2 ml DTT
  7. incubate at 25° C for 2 minutes to bind random primers
  8. add 1 ml of SuperScript II mix by flicking tube a few times
  9. incubate at 42° C for 50 minutes
  10. heat-inactivate at 70° C for 15 min
I saved 600 ng of each to run on a gel (will add RNA cocktail before I run gel). I guessed that this would be 3.4 ml but this assumes perfect efficiency.
kept on ice while adding second strand components
Brief Conclusions:   No problems will know better after gel.
Brief Update Wed Sep 6 16:54:26 EDT 2006:   The 1st strand cDNA bands are too faint to really look at. Next time I should either not run this on the gel or run more (Figure ).

Second strand synthesis of cDNA

Do in same PCR tube as first strand; no need to clean up the first strand. Keep on ice while preparing.
  1. add 66.15 ml of H2O
  2. add 10 ml of NEBuffer 2
  3. add 3 ml dNTP mix (10 mM each)
  4. add 5 ml E. coliDNA polymerase I (40 Units)
  5. add 0.25 ml RNAse H (1 Unit)
  6. incubate 2 hours at 16 C
  7. add 5 ml ligase buffer
  8. add 1 ml DNA ligase
  9. incubate 15 minutes at 16 C
  10. heat inactivate both enzymes 20 min at 75 C
  11. add 5 ml RNAse cocktail and incubate 30 min at 37 C
  12. cleaned up with Qiagen PCR clean up; eluted into 35 ml EB buffer 9
  13. spec'd DNA with 1 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield % RNA RT to DNA
    sample 1 19.7 689.5 ng 19.7%
    sample 2 25.8 903 ng 25.8%
  14. end repair with epicenter kit using 34 ml cDNA (all of it; just keep the same tube); incubated at RT 45 min
  15. heat deactivated enzymes 70 C for 10 min
  16. cleaned up with Qiagen PCR cleanup; eluted into 30 ml EB buffer
Will run 600 ng on gel.
The DNA was spec'd just before blunt cloning:
Sample DNA (ng/ul) 260/280 260/230 Yield
ds cDNA Sample 1 32.8 984 ng
ds cDNA Sample 2 29.3 879 ng
Brief Update Sat Sep 2 21:49:37 EDT 2006:  
Brief Conclusions:   The % RNA RT to DNA yield measures the amount of ds cDNA at the end relative to the 3.5 mg starting concentration of RNA. I'm not sure what this number is supposed to be. It doesn't feel too bad. However, the real % yield is actually slightly more than half this number since the 2nd strand synthesis if it were 100% efficient would double the amount of DNA from the 1st strand synthesis. RNA of course is only single-stranded.
Please see the pdf version for figures
Figure 6.4: RNA, single stranded DNA (ss cDNA), and double stranded cDNA (ds cDNA). The RNA looks pretty degraded, but the cDNA doesn't. Perhaps the RNA degraded after production of the cDNA. 1 ml of RNAse cocktail was added to the ss cDNA appx 30 min before running on a gel. Gel is 1%, 0.5 cm agarose run for 60 min at 80 V (8V/cm) in TAE with 0.5 ml of EtBr.

Preparing vector DNA

pUC19 from NEB. Cloned into DH5a, so I can have an infinite supply.
Cut 5 mg of plasmid (the DNA purchased from NEB) with SmaI at RT for 45 min. Heat deactivated 20 min at 65 C. Cleaned up with Qiagen PCR cleanup. Eluted into 30 ml . Yield:
Sample DNA (ng/ul) 260/280 260/230 Yield
pUC19 137.1 4.11 mg
T-tailing the blunt vector   Wed Sep 2, 2006
18.25 ml cut vector (2.5 mg ) was combined with 5 ml PCR buffer 25.2 ml H2O , and 1 ml dTTP (100mM). This reaction is supposed to very inefficiently add T's to the end of the sequence. I don't think it worked very well. An alternative strategy is to use terminal tranferase and add ddTTP which is more efficient. The T-tailing reaction was placed at 72° C for 90 min in a thermocycler.
The reaction was cleaned up with a Qiagen PCR clean up kit and eluted into 30 ml of EB buffer. The yields:
Sample DNA (ng/ul) 260/280 260/230 Yield
pUC19 T-tailed 67.2 2.02 mg

Cloning cDNA

Sat Sep 2, 2006
Blunt cloning the cDNA   0.5 ml vector (68 ng), 0.5 ml antarctic phosphatase, 1.7 ml phosphatase buffer, 15 ml H2O was incubated for 15 min at 37 C. The enzyme was heat-inactived for 5 min at 65 C. 2 ml T4 Ligase buffer and 1.5 ml of end-repaired cDNA (30-45 ng) were added followed by 1 ml of T4 ligase. The mixture was ligated for 2 hrs and heat deactivated for 10 min at 65C.
A-tailing the cDNA   Sat Sep 2, 2006 Taq polymerase efficiently adds a single A nucleotide to the 3' end of a double-stranded DNA piece.
Spec of A-tailed cDNA:
Sample DNA (ng/ul) 260/280 260/230 Yield
A-tailed ds cDNA Sample 1 19.2 576 ng
A-tailed ds cDNA Sample 2 19.9 597 ng
The TA-cloning reaction was done just like the blunt by first removing the phosphates and then adding the cDNA and ligase. However the TA-ligation was only for 15 minutes (if I had to do it over again, I'd do it for 2 hrs too).
For both tranformations 2 ml of the ligation mixture was placed on ice for 15 min. Followed by a 30 sec heat shock at 37 C. Cells were placed on ice for 2 min and 250 ml SOC was added before growth at 37 C for 45 min. 150 ml was plated (after x-gal and IPTG was added to the plates).
Brief Conclusions:   Only blunt Sample 1 and Sample 2 produced colonies. Sample 2 looked like satellite colonies and in the end none of sample 2 colonies grew in ampicillin.
TOPO cloning the A-tailed cDNA   Sun Sep 3, 2006
I used some of the leftover reagents from the pTrcHis TOPO kit [Invitrogen] to clone the A-tailed cDNA according to the Invitrogen protocol with TOP10 cells.
Brief Conclusions:   Many more colonies from the TOPO kit than from Blunt cloning. Still don't know if they have proper inserts.

Picking clones

The 2 TA-transformations had no colonies.
16 clones were chosen from the plates that had colonies (4 of each sample for both the TOPO kit and the Blunt cloning). The 4 clones from sample Blunt 2 didn't grow in LB and were presumed to be satellite colonies. I minipreped the 12 that grew.
Yields were:
Miniprep yields from cDNA first attempt clones
Sample ID ng/uL A260 260/280 260/230 Constant Yield (ug)
B1 a 315.07 6.301 1.95 2.09 50 15.7535
B1 b 311.39 6.228 1.95 2.05 50 15.5695
B1 c 268.3 5.366 1.95 2.06 50 13.415
B1 d 306.51 6.13 1.95 2.05 50 15.3255
T1 a 269.41 5.388 1.98 2.17 50 13.4705
T1 b 269.18 5.384 1.99 2.2 50 13.459
T1 c 283.73 5.675 1.98 2.15 50 14.1865
T1 d 237.22 4.744 1.98 2.19 50 11.861
T2 a 210.88 4.218 1.97 2.18 50 10.544
T2 b 321.16 6.423 1.99 2.19 50 16.058
T2 c 262.86 5.257 2.01 2.22 50 13.143
T2 d 236.64 4.733 2 2.18 50 11.832
Raw data in excel format
I meant to digest the 12 plasmids with EcoRI and BamHI, but I think I messed that up (see Figure ).
Please see the pdf version for figures
Figure 6.5: 1.5% gel pUC19 and pTrcHis TOPO vector digestion. Now that I think about it, I don't know what enzyme I used. I'm afriad I only added EcoRI. I meant to add EcoRI and BamHI.

Checking inserts by PCR

Thu Sep 7 11:45:41 EDT 2006
Because I'm not sure if I stuck the proper enzymes in my digestion to check for an insert (Figure 6.5), I'm going to PCR amplify the inserts. I was expecting the inserts to be realitively small anyways, so PCR is a better screening method. However, the primers just arrived today so I didn't have an opportunity to do this before.
I'll use 0.5 ml of insert (about 125 ng) and 200 nM of primer for 30 cycles with an annealing temperature of 52° C.

Sequencing clones

Wed Sep 6, 2006
Four of the Blunt clones were set to Agencourt for sequencing. B1a and B1d are being sequenced in both directions.
Sequence submission data in excel format
Brief Update Sun Sep 17 22:37:10 EDT 2006:   All of the sequences failed. Because the later projects where I used adaptors worked so much better, I don't feel any strong need to resend these out or repick,miniprep, etc them again. I do have the crap they sent me. The longest sequence was 100 bp or so and didn't match to anything related to a E. coli gene. The remaining 3 sequences were less than 10 bp. Not much to conclude for this part of the project, except that TA-style cloning of cDNA didn't seem to work that well. And neither did blunt. I'm also getting more and more disappointed in the failure rate of Agencourts sequencing service. I think it's time for a change.

6.3  Cloning double-stranded cDNA from total RNA, using adaptors

6.3.1  RNA to cDNA

RNA prep

  1. Lyse cells in 100 ml of TE with 1 mg/ml lysozyme. Incubate 2 min, vortex every minute. Add 10 ml Proteinase K. Incubate 3 more minutes, vortex every minute.
  2. add 350 ml RLT (with b-ME added) and follow the RNAeasy kit; elute with 50 ml 2 times (100 ml total)
    Sample DNA (ng/ul) 260/280 260/230 total yield
    sample 3 523.0 52.3 mg
    sample 4 482.0 48.2 mg
    I saved 1.5 ml (appx 750 ng) of each sample for a gel (see Figure ). Unfortunately, I mixed up the damn tubes. Should still give a general idea though. 97.5 ml were left for the LiCl step.
  3. add 50 ml (1/2 volume) of 7.5 M LiCl; place at -20° C for 30 minutes. centrifuge at max rpm for 15 minutes
  4. wash pellet in 1 ml 70% ethanol, resuspend in RNAse free H2O .
  5. resuspend in 50 ml of TE [Ambion]
  6. follow DNA-free TURBO kit instructions for high-conc DNA. Briefly: add Buffer, add 1 ml DNAse, incubate 30 min, add additional 1 ml DNAse, incubate 30 more minutes. Deactivate and keep supernatant.
  7. transfer the upper, aqueous phase to a new eppy tube
    Sample DNA (ng/ul) 260/280 260/230 total yield super-DNA removal loss
    sample 3 619.8 31.0 mg 40%
    sample 4 612.5 30.6 mg 37%
I saved 600 ng of each to run on a gel (see Figure ).

First strand synthesis of cDNA

Fri Sep 8 15:30 EDT 2006
Use Superscript II and the corresponding protocol:
Do in PCR tubes:
  1. add 1 ml of random hexamers (100 ng)
  2. add 1 ml of dNTP (10 mM each)
  3. add 5 mg RNA 10
  4. add H2O to 12 ml
  5. heat to 65° C for 5 minutes, chill on ice, brief centrifuge
  6. add 4 ml First-strand buffer, 2 ml DTT
  7. incubate at 25° C for 2 minutes to bind random primers
  8. add 1 ml of SuperScript II, mix by flicking tube a few times
  9. incubate at 42° C for 50 minutes
  10. heat-inactivate at 70° C for 15 min
This time I didn't save any first strand cDNA for a gel.

Second strand synthesis of cDNA

Fri Sep 8 16:30 EDT 2006
Do in same PCR tube as first strand; no need to clean up the first strand. Keep on ice while preparing.
  1. add 66.15 ml of H2O
  2. add 10 ml of NEBuffer 2
  3. add 3 ml dNTP mix (10 mM each)
  4. add 5 ml E. coliDNA polymerase I (40 Units)
  5. add 0.25 ml RNAse H (1 Unit)
  6. incubate 2 hours at 16 C
  7. add 5 ml E. coli DNA ligase buffer (NOT T4 ligase buffer)
  8. add 1 ml E. coli DNA ligase (NOT T4 ligase) 11
  9. incubate 15 minutes at 16 C
  10. heat inactivate both enzymes 20 min at 75 C
  11. this time I did not add 5 ml of RNAse cocktail, assuming instead that the RNAse H had removed enough of it to be neglegable in the spec measurements
  12. cleaned up with Qiagen PCR clean up; eluted into 35 ml EB buffer 12
  13. spec'd DNA with 1 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield % RNA RT to DNA
    sample 3 134.8 4.7 mg 94%
    sample 4 124.0 4.3 mg 86%
  14. end repair with epicenter kit using 34 ml cDNA (all of it; just keep the same tube); incubated at RT 45 min
  15. heat deactivated enzymes 70 C for 10 min
  16. cleaned up with Qiagen PCR cleanup; eluted into 30 ml EB buffer
  17. spec'd 1 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield
    sample 3 137.7 4.1 mg
    sample 4 133.4 4.0 mg
Will run 600 ng on gel (see Figure ). 13
Brief Conclusions:   Yield is much higher than last time, I don't know if that is because I got better at this or if there is still a lot of RNA that I'm measuring? I did start with almost 2x as much RNA (5 mg ). Maybe last time the RNA was too degraded for a good yield?

6.3.2  Preparing cDNA and vector for cloning

Ligation of adaptors to blunt cDNA

Fri Sep 8, 2006
Ordered BamHI adaptors from IDT. One of them was ordered with a phosphorylated 5' end (the other didn't to keep them from ligating together). Ordered 110 nmole scale. The adaptor sequence is:
BamHI adaptor 5' GATCCGAATCCGAC
                     GCTTAGGCTG-p 5'

Melting temperature is around 33° C. I resuspended each to be at 500 mM, which corresponds to 1.6 mg /ml of the short piece and 2.1 mg /ml of the long piece. I combined 20 ml of each, and placed them in a thermocycler at 60 ° C for 2 minutes. After the initial 2 minutes, I programmed the thermocycler to drop the temperature by 0.5° C every 30 seconds until it reached 4 ° C; then I transferred the annealed oligos to ice. I should be careful not to melt the annealled oligos with my fingers since the MT is lower than human body temperature. I'll use 2 ml in each reaction (appx 4.2 mg ).
  1. to the 29 ml of cleaned up DNA (1 ml was used to spec), add 3.6 ml T4 DNA ligase buffer
  2. add 2 ml (appx 4.2 mg ) of BamHI adaptor
  3. add 1 ml of T4 DNA ligase
  4. mix by flicking the tube a few times
  5. incubate for 12 hrs at 16° C 14
  6. heat inactivate T4 ligase at 65 C for 10 min
  7. add 1 ml of T4 DNA ligase buffer15
  8. add 1 ml of T4 polynucleotide kinase (no need to add ATP because it is in the ligase buffer)
  9. incubate at 37° C for 30 minutes
  10. heat inactivate for 20 minutes at 65° C
  11. clean up with Qiagen PCR purification kit, elute into 30 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield post-adaptor gain
    sample 3 162.3 4.9 mg 18%
    sample 4 156.3 4.7 mg 18%
Annealed linkers would look like:
BamHI adaptor 5' GATCCGAATCCGACGTCGGATTCG
                     GCTTAGGCTGCAGCCTAAGCCTAG 5'

GAT CCG AAT CCG ACG TCG GAT TCG GC TTA GGC TGC AGC CTA AGC CTAG
Please see the pdf version for figures
Figure 6.6: cDNA and RNA samples 3 and 4 (see section 6.2.1.1 on page pageref for condition and growth details). 1.0% gel, 80V(8V/cm), 0.5 cm, 70 minutes, 0.5 ml EtBr. Approximately 750 ng is in each of the RNA or cDNA lanes. The second pUC19 is the BamHI-cut vector used in the ligation.
Brief Conclusions:   I'm a little worried about the gain in DNA quantity from the ligation of adaptors. The gain of 18% would indicate that the original sequence was only around 54 bp in length (a 12 bp linker is 18% of a total of 66 bp). An alternative problem would be if the annealed linkers, which are around 20-28 bp long depending on how you want to look at it, got through the column even though it is supposed to eliminate short stuff. This could happen, since I'm really loading a lare amount (4 mg ) of linker through the column. The amount gained in DNA quantity would represent 80% of the short oligos being washed through. If it is the case that 20% of the primers (appx 800 ng). The gel seems to indicate that this is the case (Figure 6.6. I'm afraid all the clones are going to be this little fragment. I think I have enough to gel select (or column select, but the gel for the columns hasn't arrived yet).

Digestion of vector

Sat Sep 9 22:05:12 EDT 2006
Miniprep pUC19:
Sample DNA (ng/ul) 260/280 260/230 total yield
sample A 104.4 3.1 mg
sample B 91.9 2.8 mg
Digested 23.5 ml of pUC19 sample A (appx 2.5 mg ) with 0.5 ml BamHI (10 U) at 37 C for 45 minutes. Heat deactivated 20 min (I don't know if this helps, since BamHI can't really be heat deactivated).
Added 3.4 ml phosphatase buffer and 1 ml antarctic phosphatase to dephosphorylate for 15 minutes at 37 C followed by a 5 min heat inactivation. Reaction was cleaned up with Qiagen PCR kit:
Sample DNA (ng/ul) 260/280 260/230 total yield
sample A post digestion50.8 1.5 mg

6.3.3  cloning adaptored cDNA

Ligation of adaptored cDNA to cut vector

Sat Sep 9 22:05:12 EDT 2006
7.5 ng cDNA with 50 ng vector and 50 ng cDNA with 50 ng vector
Did in standard way like in the appendix. 2 ml ligation, 30 min on ice, 20 sec heat shock at 42 C, 2 min on ice, add 250 ml SOC, incubate 37 C rpm 225 for 60 min, plate.
Plated 150 ml with blue-white solution.
Please see the pdf version for figures
Figure 6.7: Transformation of cDNA samples 3 and 4. A = 7.5 ng of cDNA; B = 50 ng of cDNA.
Brief Conclusions:   Looks like I definitely have a lot of empty vector sequence. The blue/white ratio is horrible (see Figure 6.7). There are inserts, but they are swamped by no insert vectors. I think next time I should digest the vector longer, clean up the reaction in between the digest and the phosphatase, and run the phosphatase reaction longer (1 hr).

Insert checking

Mon Sep 11 11:49:26 EDT 2006
Last night I picked 8 white colonies (2 from each plate) to check for an insert. I'm going to Miniprep them, and then PCR the insert region of the plasmid DNA.
Ran 0.5 ml with 1 ml of 5 uM primer mix. MT = 52 C. Ran 30 cycles. Ran 9 ml on a gel (10 ml including the dye).
Please see the pdf version for figures
Figure 6.8: 1.5% gel 100 V (think was too hot; bands are kinda droopy) pUC19 vector amplification by PCR with M13-FOR and M13-REV. Plasmid DNA is also visible on the gel and the shift in plasmid size correlate well with the insert amplification size.
Brief Conclusions:   The vectors lengths look better than I guessed. Yes, there were a lot of blue colonies. And looking at the gel for the totalRNA ® cDNA (Figure 6.6), there was so much adaptor sequence, I assumed that all the inserts would be annealed adaptor sequence only. looks like 3a B and 3b A and 4b B should be sequenced. I might include 4b C and 3b B just to see what I get.
Brief Update Mon Sep 18 00:29:45 EDT 2006:   These four were sent out for sequencing. Results will be in tomorrow (hopefully).

Sequencing non-size selected inserts for samples 3 and 4

(got sequences back Mon Sep 18, 2006)
I spec'd the 5 samples sent out for sequencing. 2.5 ml of each of them was sent in the sequencing reaction along with 1 ml of 20 uM forward primer.
Sample ID ng/uL A260 260/280 260/230 blastn (nr) result
3a B 259.12 5.182 1.94 2.18 rrlD 23S
3b A 361.66 7.233 1.93 2.22 pUC19
3b B 281.49 5.63 1.94 2.2 pUC19
4b B 320.8 6.416 1.95 2.2 pUC19
4b C 172.62 3.452 2.02 2.18 pUC19
All of the sequence data and chromatagraphs from Agencourt can be found here.
Brief Conclusions:   As expected everything is rRNA. I checked if the shorter ones had inserts (e.g. the blunt ligated adaptors) or if they were just empty vectors that showed up white on the plate. 4b B is particularlly weird 3bA has 117 bp that match 16S rRNA, but the fragment is short relative to the amount of vector in the sequence so it doesn't come out as the top blast hit. 3bB matches only vector. 4bB I can't figure out. 4bC looks like there's a little adaptor seqeunce and mostly just vector.

6.3.4  size selecting cDNA for cloning

better digestion of vector?

Mon Sep 11 19:57:18 EDT 2006
Digested 23.5 ml of pUC19 sample A (appx 2.5 mg ) with 0.75 ml BamHI (15 U) at 37 C for 60 minutes. This is 5 U more than the last time, and 15 minutes longer. This time I'm cleaning up the reaction before I do the phosphatase. And I'm going to use less DNA for the phosphatase step (so the phosphatase doesn't have to do as much work) in addition to running the dephosphorylation for 60 minutes instead of 15 minutes.
Sample DNA (ng/ul) 260/280 260/230 total yield
sample A post digestion50.8 1.5 mg
I put 200 ng of this digested vector into a 10 ml phosphatase rxn for 1 hour followed by a 5 min heat deactivation. 2.5 ml (50 ng) will be used for each ligation.
Please see the pdf version for figures
Figure 6.9: Transformation of size-selected cDNA sample 3 with a better cleaned vector. I need a better way to image these plates. Even with these poor pictures, you can see there are far fewer blue colonies (though still quite a few) than last time (see Figure 6.7). The Big 3 ml plate only had 2 white colonies, and had very few colonies total. When I replated 125 ml the next day (using most of the remaining transformation cells that were sitting on my bench), I got many more colonies.
Brief Conclusions:   It's not clear what helped, but the extra digestion time, clean up after digestion, extra dephosphorylation time, and use of less vector in the dephosphorylation reaction certainly lowered the number of blue colonies (compare Figure 6.7 vs the newer protocol result in Figure 6.9). I'll use this new method from now on for the BamHI into pUC19 cloning.

size-selection of cDNA via agarose gel

Mon Sep 11 19:57:18 EDT 2006
Ran a 1.0% TAE agarose gel for 50 minutes at 90V. Stained with SYBR gold Cut two ranges of DNA: med = 500-1500 bp, big = 1500-9000 bp (see Figure ). The sample 3 lane contained 11 ml of sample 3, adaptored cDNA (162.3 * 11 = 1796.3 mg ).
Please see the pdf version for figures
Figure 6.10: Size selection gel for sample 3. 1.0% gel 90 V TAE 0.5 cm. Stained with SYBR gold for 30 minutes, washed in H2O for 5 minutes. Took 2 images in VersaDoc for a total of 2.5 secs under UV. Cut with a razor blade under the blue-light transilluminator (no UV). Although this image is pretty crappy, by-eye under the transilluminator it looked just fine (I've never been impressed with the versadoc images with SYBR gold).
The two gel slices were placed into 600 ml tubes that had a hole poked in the bottom of them with an 18 gauge needle. The 600 ml tubes were placed inside a 2 ml tube and spun at 13,000 rpm for 1 minute to macerate the gel. The macerated gel was placed in a Spin-X column (costar) and 200 ml TE was added. The column was spun for 10 minutes. 200 ml of TE was added to the 2 ml tube (the one that had the macerated gel in it) and vortexed vigorously. This TE/gel remnant mix was added to the column which was spun another 5 minutes. 40 ml of NaAcetate was added to the spin-x'd TE/DNA mixture. 1 ml of 95% EtOH was added and mixed and the solution was placed at -86C for 15 minutes, spun for 10 min at 4 C, removed EtOH, added 1 ml 70% EtOH, spun 5 minutes at 4 C, remove supernatant. Evaporated EtOH in fume hood for 10-15 minutes. Resuspended DNA in 10 ml of TE [Ambion]. The spec readings looked really crappy (not like a mountain at all) so I wouldn't really trust these yields:
Sample DNA (ng/ul) 260/280 260/230 total yield
sample 3 med, post-gel cleanup63.0 630 ng
sample 3 big, post-gel cleanup72.7 727 ng

Ligation and transformation of size-selected cDNA

Mon Sep 11 20:11:24 EDT 2006
2.5 ml of dephosphorylated/BamHI cut vector was used in all reactions. 2 med reactions were run with 1 ml (3med1) and 3 ml (3med3) of the gel cleaned sample. 1 big reaction was run with 3 ml (3big3) of the gel cleaned sample (remember med = 500-1500 bp; big = 1500-9000 bp).
Ligation was 30 min at 16 C with 10 min at 65 C to heat deactivate. Transformation was as done in the previous experiment (see section 6.3.3) except that I only plated 75 ml instead of 150 ml , which gave me too many colonies last time.
The transformation worked well with plenty of white-colonies and fewer blue than before (see Figure 6.9 and section 6.3.4.1 on page pageref).

Insert checking

Wed Sep 13, 2006
Last night I picked 16 white colonies (2 from 3med1, 10 from 3med3, and 2 from 3big3 [these were the only 3big3 white colonies]) to check for an insert. I miniprepped them, and PCR'd the insert region of the plasmid DNA.
Ran 0.5 ml with 1 ml of 5 uM primer mix. MT = 52 C. Ran 30 cycles. Ran 9 ml on a gel (10 ml including the dye).
Please see the pdf version for figures
Figure 6.11: 1.5% gel 100 V (think was too hot; also think PCR melt temp was too low bands) pUC19 vector amplification by PCR with M13-FOR and M13-REV. Plasmid DNA is also visible on the gel and the shift in plasmid size correlate well with the insert amplification size. These clones were size-selected using a 1% agarose gel before cloning. med corresponds to clones that were from the 500 bp to 1500 bp range. big were clones from the 1500 bp and above (to appx 9000 bp). Except in the cases where it looks ike there was no insert, the bands all fall roughly within the correct range. Notice that the ordering is a little funky for the 3med3 as one of the tube caps broke and when I put it back on, I stuck the tube (number 3) at the end by mistake instead of in the correct place chronologically.
Brief Conclusions:   The vector insert lengths (see Figure 6.11) are right on target for the size-selected fragments. As expected, the med clones have insert fragments between 500-1500 bp and the big clones have inserts > 1500 bp. The number of blue clonies is way down after running a longer dephosphorylation and digestion coupled with a slightly different order of those two steps. I'm going to send several out for sequencing.

sequencing size-selected cDNA sample 3

The following five clones of sample 3 were sent out for sequencing. The two big colonies were sequenced from both ends.
Sample ID ng/uL A260 260/280 260/230 blastn (nr) result
3med1A 102.67 2.053 1.97 2.24 rrlA 23S
3bigA 110.77 2.215 1.95 2.13 rrlG 23S
3bigB 139.18 2.784 1.92 2.21 rrlH 23S
3med3.1 135.34 2.707 1.97 2.17 rrsH 16S
3med3.4 167.73 3.355 1.92 2.25 rrlG 23S
The two big clones that were sequenced from both ends allowed me to see that only one fragment was inserted and that both of the adaptors were correctly ligated to the sequence. BigA for example had a 1579 bp insert of rrlG (see a compiled fasta file).
All of the sequence data and chromatagraphs from Agencourt can be found here.
Brief Conclusions:   The good: size-selection works and helps tremendously to remove non-insert clones and short-insert clones. The bad: all rRNA. Hopefully, the MICROBExpress method for removing 16S and 23S rRNA will lessen this problem.

6.4  Cloning double-stranded cDNA from mRNA, using adaptors

Using samples 5 and 6 (5=5, 6=1) from section 6.2.1.1 on page pageref.

6.4.1  RNA to cDNA

RNA prep

Tue Sep 12 10:46:43 EDT 2006
  1. Lyse cells in 100 ml of TE with 1 mg/ml lysozyme. Incubate 2 min, vortex every minute. Add 10 ml Proteinase K. Incubate 3 more minutes, vortex every minute.
  2. add 350 ml RLT (with b-ME added) and follow the RNAeasy kit; elute with 50 ml 2 times (100 ml total)
    Sample DNA (ng/ul) 260/280 260/230 total yield
    sample 5 466.2 46.6 mg
    sample 6 407.4 40.7 mg
    I saved 1.5 ml (appx 625 ng) of each sample for a gel. 97.5 ml were left for the LiCl step. Yields were a little lower than last time, hopefully the RNAprotected RNA hasn't degraded in the freezer. It's been 2 weeks (minus 1 day) since I took the samples.
  3. add 50 ml (1/2 volume) of 7.5 M LiCl; place at -20° C for 30 minutes. centrifuge at max rpm for 15 minutes
  4. wash pellet in 1 ml 70% ethanol incubate at RT 2 minutes, spin 5 minutes, dry pellet 7 minutes
  5. resuspend in 35 ml of TE [Ambion] 16
  6. follow DNA-free TURBO kit instructions for high-conc DNA. Briefly: add Buffer, add 1 ml DNAse, incubate 30 min, add additional 1 ml DNAse, incubate 30 more minutes. Deactivate and keep supernatant.
  7. transfer the upper, aqueous phase to a new eppy tube
  8. spec
    Sample DNA (ng/ul) 260/280 260/230 total yield super-DNA removal loss
    sample 5 1447.5 31.0 mg 40%
    sample 6 1160.3 30.6 mg 37%
    saved 0.75 ml (appx 950 ng) to run on gel
  9. use MICROBExpress to remove 16S and 20S from 10 mg of total RNA (max volume 15 ml ). (6.9 ml of sample 5 and 8.6 ml of sample 6) 17
  10. spec'd
    Sample DNA (ng/ul) 260/280 260/230 total yield loss to MICROBExpress
    sample 5 150.1 3.8 mg 87.7%
    sample 6 166.4 4.2 mg 86.3%
  11. save to run on gel
Finished at: Tue Sep 12 17:56:32 EDT 2006 (7 hrs 15 min from RNA prep to here)
I saved 600 ng of each to run on a gel (see Figure ).

First strand synthesis of cDNA

Tue Sep 12 10:46:31 EDT 2006
Use Superscript II and the corresponding protocol:
Do in PCR tubes:
  1. add 1 ml of random hexamers (100 ng)
  2. add 1 ml of dNTP (10 mM each)
  3. add 500 ng of mRNA18
  4. add H2O to 12 ml
  5. heat to 65° C for 5 minutes, chill on ice, brief centrifuge
  6. add 4 ml First-strand buffer, 2 ml DTT
  7. incubate at 25° C for 2 minutes to bind random primers
  8. add 1 ml of SuperScript II, mix by flicking tube a few times
  9. incubate at 42° C for 50 minutes
  10. heat-inactivate at 70° C for 15 min
Again, I didn't save any first strand cDNA for a gel.

Second strand synthesis of cDNA

Tue Sep 12 10:45:43 EDT 2006
Do in same PCR tube as first strand; no need to clean up the first strand. Keep on ice while preparing.
  1. add 66.15 ml of H2O
  2. add 10 ml of NEBuffer 2
  3. add 3 ml dNTP mix (10 mM each)
  4. add 5 ml E. coliDNA polymerase I (40 Units)
  5. add 0.25 ml RNAse H (1 Unit)
  6. incubate 2 hours at 16 C
  7. add 5 ml E. coli DNA ligase buffer (NOT T4 ligase buffer)
  8. add 1 ml E. coli DNA ligase (NOT T4 ligase) 19
  9. incubate 15 minutes at 16 C
  10. heat inactivate both enzymes 20 min at 75 C
  11. this time I did not add 5 ml of RNAse cocktail, assuming instead that the RNAse H had removed enough of it to be neglegable in the spec measurements
  12. cleaned up with Qiagen PCR clean up; eluted into 35 ml EB buffer 20
  13. I skipped the spec this time to save DNA
  14. end repair with epicenter kit using 34 ml cDNA (all of it; just keep the same tube); incubated at RT 45 min
  15. heat deactivated enzymes 70 C for 10 min
  16. cleaned up with Qiagen PCR cleanup; eluted into 30 ml EB buffer
  17. spec'd 1 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield
    sample 5 20.2 606 ng
    sample 6 17.2 516 ng
Will not run 600 ng on gel as then I won't have anything left! (I'll see the cDNA when I run the gel to size select it).
Please see the pdf version for figures
Figure 6.12: RNA samples 5 and 6 (see section 6.2.1.1 on page pageref for condition and growth details). 1.0% gel, 80V(8V/cm), 0.5 cm, 70 minutes, 0.5 ml EtBr. Approximately 600 ng is in each of the RNA lanes.
Brief Conclusions:   Need to aliquot more Invitrogen dNTPs.

6.4.2  Preparing cDNA and vector for cloning

Ligation of adaptors to blunt cDNA

Tue Sep 12 10:46:04 EDT 2006
I'll use 2 ml in each reaction (appx 4.2 mg ).
  1. to the 29 ml of cleaned up, end-repaired DNA (1 ml was used to spec), add 3.6 ml T4 DNA ligase buffer
  2. add 2 ml (appx 4.2 mg ) of BamHI adaptor
  3. add 1 ml of T4 DNA ligase
  4. mix by flicking the tube a few times
  5. incubate for 12 hrs at 16° C 21
  6. heat inactivate T4 ligase at 65 C for 10 min
  7. add 1 ml of T4 DNA ligase buffer22
  8. add 1 ml of T4 polynucleotide kinase (no need to add ATP because it is in the ligase buffer)
  9. incubate at 37° C for 30 minutes
  10. heat inactivate for 20 minutes at 65° C
  11. clean up with Qiagen PCR purification kit, elute into 30 ml
    Sample DNA (ng/ul) 260/280 260/230 total yield post-adaptor gain
    sample 5 80.6 2.4 mg apprx 4x
    sample 6 80.3 2.4 mg apprx 4x
Brief Conclusions:   It is clear that the adaptors a certainly contributing a lot of DNA.

size-selection

Performed using spin-x as in previous attempt. This time I ran all 30 ml of adaptored cDNA onto the gel. I made a 0.5 cm gel with the 6 comb. The wider lane-size allowed me to fit the entire 30 ml , but it also made it a bit tight in the spin-x tube. Because of that, I didn't do the second 200 ml TE wash of the agarose this time.
The size selection gel can be seen in Figure .
Please see the pdf version for figures
Figure 6.13: Size selection gel for samples 5 and 6. 1.0% gel 90 V TAE 0.5 cm. Stained with SYBR gold for 30 minutes, washed in H2O for 5 minutes. Took 1 image in VersaDoc for a total of 1.5 sec under UV. Cut with a razor blade under the blue-light transilluminator (no UV). Thi is pretty crappy and by-eye under the transilluminator the cDNA was still pretty faint, so I think more 500 ng mRNA is needed in the 1st strand synthesis.

Ligation and transformation

Ligation and transformation were performed as in the previous attempt except that all 10 ml of the gel purified DNA was used. Also, the gel purification was resuspended in EB buffer instead of TE in case the EDTA present in 10 ml would've inhibited the reaction.

Insert checking

The normal PCR insert checking procedure was done with 0.5 ml of plasmid from 16 white colonies. Results are in Figure .
Please see the pdf version for figures
Figure 6.14: 1.5% gel. Only 2 out of 16 colonies had what looks like an insert. I need to redo the cDNA step with more RNA or use a different type a size-fractionation [like the sephacrl] or else I'm doomed with insert efficiencies like this.
Brief Conclusions:   None of the clones had good inserts (See Figure 6.14). And the only one that looked like it had an insert at all had a very short one (only 300-400 bp). I think the problem is not enough material. It is quite hard to clone into the BamHI cut pUC19 and not get a bunch of blue colonies. It would be much easier with a non-symmetric kinda thing like I'll use later. For now, a minor adjustment is that I might incubate with Antarctic phosphatase for even longer or add more. But more importantly I think I need to run with more starting material so that when I do the size-selection, I actually have something easily viewable to select (see Figure 6.13).
I'm also a little disappointed with the results of the MICROBExpress kit. It clearly depleted the 16S, but there still seems to be a lot of 23S hanging around. Maybe I need to run it through the magnetic bead protocol 2x (see Figure 6.12)? I should also try using a denaturing agarose gel instead of native.

6.5  Cloning double-stranded cDNA from mRNA, using adaptors and more RNA

Previous attempt didn't work too well with the mRNA. This time I'm bumping it up to 1.5 mg of mRNA (with is using the maximum allowable volume of 10 ml for a standard 1st strand synthesis reaction).
Most of the steps I'm doing as I did before (see section 6.4 on pages pageref-pageref). Here are the modifications: (1) using 1.5 mg instead of 500 ng of mRNA in the first-strand reaction. (2) using 2 ml of Superscript II instead of 1 ml 23 (3) adding 0.25 ml of RNAse just prior to adding the E. coli DNA Ligase.
Thur Sep 21, 2006 I ran the entire sample on an agarose gel for size-selection (see Figure ). It was cleaned up with a spinX column.
Please see the pdf version for figures
Figure 6.15: size-selection 1.0% gel for samples 3 and 5. Once again SYBR Gold visualizes crappily but it is clear as day on the transilluminator by eye. Sample 3 was size-selected for the circularization experiment below. Sample 5 was for traditional cDNA cloning.
I cloned them in the same way as before, except that the dephosphorylation setp used 2 ml of Antarctic phosphatase instead of 1 ml and I ran the reaction for 30 minutes instead of 60 minutes. Also, I used 350 ml SOC in the transformation and plated 50 ml (250 ml and 75 ml were used in the previous attempt). I ligated 5 ml of the gel-size selected cDNA (out of 15 ml total).
I got less colonies than in the past, the white-blue ratio was bad but not horrible. I picked an initial 4 colonies (2 from 5big and 2 from 5med) to check if they were all empty like last time.
Sun Sep 24, 2006
Please see the pdf version for figures
Figure 6.16: 1.5% gel EtBr. Inserts were checked by PCR using the M13 primers.
Brief Conclusions:   The inserts are actually present this time at decent lengths (compare Figure 6.16 and the previous attempt in Figure 6.14). The question now is will I have something besides 23S rRNA??? Based on Figure 6.12, I'm not too hopeful; Now I think the bias will be even stronger towards 23S since the kit seems to have done a nice job to get rid of the 16S!
Sat Sep 23, 2006 I plated another 100 ml the next day to have more colonies to pick.
Brief Update Mon Sep 25 15:20:44 EDT 2006:   Now that it looks like this mRNA derived cDNA is clonable, I need to sequence a bunch (10-20) and see if they are all rRNA still. I'll pick more colonies for miniprepping tonite.
I picked 16 colonies (2 from each plate in the previous plating above and 6 from each of the newer plates from Sep 23, 06). I miniprepped them and check the insert by PCR and agarose gel (see Figure ).
Please see the pdf version for figures
Figure 6.17: 1.5% gel sample 5 insert checks. Unfortunately, I messed up the labeling for the four samples which I picked from the Sep 23 plate. Based on the insert size I'd say that samples 3 and 4 were the big samples.
Brief Conclusions:   These additional 16 samples look good for the most part. Sample 11 is particularly interesting because the insert is more than 5kb.

Sequencing the improved insert size samples from sample 5

Thu Sep 28 15:31:49 EDT 2006
I sent four sequences out this morning (these were vectors checked in Figure 6.16). If they come back ok, I'll send more.
Here's the info for the sequencing:
Sample ID ng/uL 260/280 260/230 blastn (nr) result
5big 2A 318.1 23S rrlH
5big 2B 608.0 23S rrlA
5med 2A 539.3 tnaL - tnaA!!!
5med 2B 397.8 23S rrlC
Brief Conclusions:   Mon Oct 2 16:17:38 EDT 2006
As the table above shows, FINALLY I have an insert that is not a 23S or a 16S rRNA. One out of the four sequences sent was not an rRNA. Also, that one sequence is from the operon that contains tnaL, tnaA, tnaB. The sequence starts at the 41st bp of the leader sequence and proceeds to at least the 664 bp of tnaA (tnaA is 1431bp total). Rich Roberts mentioned to Simon that it may not be possible to get the 5' end of genes because the translation machinery in E. coli eats up (degrades) the 5' ends as it moves along. This, my first result that doesn't involve an rRNA gives at least one hint that this might not be a problem. Clearly this transcript runs across two genes, and more than that begins only 40 bp away from the start of a leader peptide not even a proper gene. Those things don't get picked up well by microarrays, but it seems that maybe this sequencing approach will catch them. Not that the sequence also did NOT read through the entire insert which is actually about 1500 bp long (see Figure 6.16). I think it would be good to sequence in the other direction, so at least I have some evidence for how long I got on this gene.
I'll probably send 10-16 more to see if I can get some stats.
Mon Oct 30 11:28:44 EST 2006
The following additional samples checked in Figure 6.17 were spec'd and sent to agencourt for sequencing. The samples had been in the fridge for about a month, so hopefully they haven't degraded.
Sample ID ng/uL 260/280 260/230 blastn (nr) result
5med 2 239.4 rrlC 23S
5big 2B 375.7 rrlG 23S
5med 2A 127.3 kdsA + intergenic
5med 2B 228.3 rrlA + 23S
5med 2B 490.8 rrlH + 23S
Here is the raw sequence data from agencourt.
Brief Conclusions:   Mon Nov 6 13:41:45 EST 2006
Now the number of rRNA to mRNA reads is 2/9. Very bad, but much better than before I used the Microbexpress kit. It is clear though that I'm either going to have to optimize that kit (e.g. annealing longer or using wather bath instead of heat block) or run the RNA through the kit 2x. However, both of the reads that I have for non-rRNA genes do yield information that would be informative to determining gene boundaries.

6.6  Circularization test with sample 3

6.6.1  Circularization adaptors and strategy

I can shorten the overhang to a 3-mer which is NOT palendromic (so that the adaptored sequences DON'T ligate to each other) by adding an extra G to the phosphorylated oligo (see section 6.3.2.1 on page pageref for the original adaptor).
BamISH adaptor 5' GATCCGAATCCGAC
                     GGCTTAGGCTG-p 5'

This primer should be much better than the BamHI for ligating to the circularization dsDNA oligo below, because it helps ensure that only one adapted RNA and one circularization probe ligate into a circularized piece (with the BamHI adaptor and a BamHI adaptor you could easily get concatamers of RNA or circularization oligo). All overhangs (on the adaptor and the circularization probe) will be phosphorylated to make the ligation more efficient 24.
Melting temperature is around 35° C. I resuspended the bottom at 500 mM, which corresponds to 1.69 mg /ml of the short piece and 2.1 mg /ml of the long piece25. I combined 20 ml of each, and placed them in a thermocycler at 60 ° C for 2 minutes. After the initial 2 minutes, I programmed the thermocycler to drop the temperature by 0.5° C every 30 seconds until it reached 4 ° C; then I transferred the annealed oligos to ice. I should be careful not to melt the annealled oligos with my fingers since the MT is lower than human body temperature. I'll use 2 ml in each reaction (appx 4.2 mg ).
dsDNA fragment for circularization
circularization adaptor   5' p-ATCGGCCAAGGCGGCCGTACG
                                  CCGGTTCCGCCGGCATGCCTA-p 5'

Please see the pdf version for figures
Figure 6.18: Schema of the circularization strategy.
Melting temperature is around 62.8° C.

6.6.2  Circularization first attempt

Circularization by ligation

Thur Sep 21, 2006
I used approximately 120 ng (0.5 ml ) of circularization oligo. Combined with 5 ml of adaptored size-selected cDNA. I used 2 ml T4 ligase, 1 ml of T4 ligase. I ran the ligation for 2 hr at 16° C followed by heat-inactivation of the ligase.

RCA amplification and MmeI digestion

Fri Sep 22, 2006
I ran RCA according to the fidelity systems protocol. 1.5 ml template, 2.5 ml annealing buffer, 1 ml fidelity systems random hexamer primer (modified to reduce exonuclease reaction). The only modification of their protocol was running the reaction for 4 hr instead of the recommended 12 hr.
This reaction certainly amplified something because in the end (after running through a Qiagen PCR cleanup), I had decent yields of DNA:
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA 3med 43.9 1.7 mg
RCA 3big 50.3 1.5 mg
I digested 15 ml of each RCA product with MmeI. I ran this digested RCA product and 15 ml of uncut RCA product (all that remained) on an agarose gel
Please see the pdf version for figures
Figure 6.19: 2% agarose gel. Doesn't look like things worked. Somewhere along the way (ligation with circularization oligo or maybe RCA) the small pieces seem to have ligated into giant pieces. 15 ml is around 850 ng of DNA.
Brief Conclusions:   It looks like somewhere along the way of ligation to circularization oligo, exonuclease digestion, and RCA amplification something went awry. You can see on the gel (Figure 6.19) for the uncut lanes there is one giant piece. I guess as a long shot, the circularized piece might move VERY slow because of some strange supercoiling. Actually, maybe it worked. Now that I read more about RCA, it should make giant pieces. However, in Shendure et.al. they cut 40 mg of the RCA material to digest with MmeI. The ran 1/4 th of this in one lane of a gel (10 mg ). This is more than 10x what I used. I'm going to try the RCA again with triple the starting concentrations and run overnite. We'll see if this is any better. I'll use SYBR gold to stain it too.

RCA amplification and MmeI digestion, try 2

Tue Sep 26, 2006
I tried multiplying the previous reaction by 3 to get more DNA. I eluted into 50 ml and had the same total concentration (so an increase in 40% or so for a 3x increase in material). I ran it on sample 3med only.
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA 3med 51.9 2.6 mg
Brief Conclusions:   After I digested it, I was ethanol precipitating the large volume digestion so that it would fit in a gel lane. Unfortunately, I was using a 600 ml tube that isn't the best fit in the centrifuge. In the final spin, the tube shattered. So I don't know if this worked or not. Nonetheless, the slight increase in DNA for a tripling of reagent and template is disappointing.

RCA amplification try 3

Wed Sep 27, 2006
I repeated the RCA more in line with what Jay used in his Science polony paper. They use less dNTP but quite a bit more enzyme. Per reaction I used: 5.25 ml 10x buffer, 2 ml dNTP, 5 ml template DNA, 2.5 ml hexamer, 35.25 ml H2O . I heated at 95C for 5 minutes (I put the sample in the thermocycler when the temperature was still at RT). I cooled them down to 4C in the thermocycler and immediately transferred the samples to ice. Then I added 2.5 ml of f29 polymerase and incubated at 30C for 12 hours followed by 10 min at 65C to deactivate the enzyme. I ran to of these 52.5 ml reactions using the circularized 3big sample for both reactions.
I combined the 2 RCA reactions into one tube and ethanol precipitated them (30 min -86C, 15 min spin 4C, 750 70%, 5 min spin). Resuspension was slow. I used 200 ml of EB buffer, 50C, and periodic vortexing. It still took a while (30-60 min) to go back into solution. There was clearly a lot of DNA. Upon initially adding the ethanol, you could see quite a lot of DNA after 30 secs or so. The yield was very high (more in line with what Jay reports in his paper):
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA 3big 650.4 130 mg
Brief Conclusions:   I think the DNA from the RCA step is incredibly long. I think the Qiagen PCR prep was selectively removing most of it, so now with EtOH precipitation I see the real yield.

MmeI digestion of RCA amplification try 3

Wed Sep 27, 2006
I digested approximately 10 mg of RCA amplified sample 3 with MmeI. The 30 ml digest was: 15 ml RCA, 5.75 ml water, 3 ml Buffer4, 1.25 ml SAM, 5 ml MmeI. I didn't clean it up. After a 30 min incubation at 37 and 10 min at 65C26, I added 5 ml dye and ran the sample on a 2% agarose gel 12V/cm for 55 minutes (see Figure ).
Please see the pdf version for figures
Figure 6.20: 2% agarose gel, sybr gold stained of circularized, RCA amplified, mmeI digested cDNA sample 3. I imaged it on the transilluminator with the $200 digital camera. There's a slight jiggle effect (i.e. my hand wasn't steady enough), but it is much clearer than the $30K Versadoc for some reason. The first lane 3 ml of uncut RCA DNA for sample 3. The second lane is a fisher 1KB ladder, the third is an NEB PCR ladder.
Brief Conclusions:   I think we're in business!!! The MmeI tag of approximately the correct size is visible (see Figure 6.20). The band is not so sharp because I ran the stuff on agarose, but nevertheless it is clearly visible. I think next time I should increase the volume of the reaction and increase the amount of enzyme or incubation time, because there are clearly going to be a LOT of MmeI sites. I don't know if things get messed up when the concentration is too high?
I ordered some PAGE gels so I can do a proper gel extraction without the fuzzy problem that comes with low MW DNA on an agarose gel (like in Figure 6.20). Then the only remaining step is to ligate the two big primers on the end and sequence it. I don't think I'm going to sequence these tags. I think I'll wait and sequence everything after I've added the two ends.

6.7  Ligating the final adaptor ends for polony sequencing

The Shendure et.al. polony approach uses two adaptors on the ends to sequence a few base pairs inwards on the MmeI digested piece. This is my last step then I can sequence them and I'm ready to go (though I still need to more efficiently remove rRNA).

6.7.1  Ligating for cloning using circularized/digested sample 3

I don't understand why the oligos they used in the Science paper were blunt. It seems using this strategy the adaptors can ligate in any orientation and then following PCR reaction allows you to select the properly oriented ones, but maybe I'm not understanding their method correctly? Here are their adaptors:
FDV2:
5' AACCACTACGCCTCCGCTTTCCTCTCTATGGGCAGTCGGTGAT
   TTGGTGATGCGGAGGCGAAAGGAGAGATACCCGTCAGCCACTA 5'

RDV2:
5' AACTGCCCCGGGTTCCTCATTCTCT
   TTGACGGGGCCCAAGGAGTAAGAGA 5'

You can see there is a big length difference between them. This is because you need to gel select the ligation produces that are as follows:
FDV2:mmeA_circularizer_mmeB:RDV2

So you can use the length differences to distinguish that correct one from the ones with either two FDV2 or two RDV2 adaptors.
I'm going to add an extra few bp to the end of the adaptors to allow me to force a direction to the adaptors and to make it easy to clone them into a vector.
FDV2_EcoRI:
5' AATTAACCACTACGCCTCCGCTTTCCTCTCTATGGGCAGTCGGTGAT
       TTGGTGATGCGGAGGCGAAAGGAGAGATACCCGTCAGCCACTA 5'

RDV2_BamHI:
5' GATCAACTGCCCCGGGTTCCTCATTCTCT
       TTGACGGGGCCCAAGGAGTAAGAGA 5'

TTGACGGGGCCCAAGGAGTAAGAGA
To try with tag: save 1 ml add phosphates? and ligate to cut puc19 save 1 ml for PCR run remaining 20 ml on gel
I think it would've been good to have phosphates on the blunt end of the oligo linkers just like with cDNA. Next time.... grow puc19

Digestion with MmeI

Tue Oct 18, 2006
As in the Shundure paper, I want to digest 40 mg of DNA (61 ml of the RCA from section 6.6.2.1 on page pageref).
I'm running the following reaction 4x (in 4 different tubes), 15 ml RCA, 5 ml NEBuffer4, 5 ml MmeI, 2 ml SAM, 23 ml H2O . This reaction was incubated at 37C for 30 minutes followed by heat deactivation at for 65C 10 minutes. I ethanol precipitated the digestion and resuspended in 80 ml of TE.
The maximum amount per 10 lane, 1.0mm Novex TBE gel is 25 ml . The 80 ml of cut RCA product was split into 4 and combined with 5 ml of loading dye (25 ml total) and run on a 6% TBE polyacrylamide gel. Two lanes of Low Molecular Weight DNA Ladder [NEB] were also run.
The gel was stained with SybrGold and imaged by hand using a the blue-light transilluminator (Figure ).
Please see the pdf version for figures
Figure 6.21: 6% TBE polyacrylamide gel showing the 80-85mer tag released from the RCA amplification by MmeI. This image has been modified using the Autocontrast feature in PhotoShop. The original can be found in the Oct, 2006 image directory.
Brief Conclusions:   Perhaps it would be useful to run the digestion through a spin column, if in the future I don't do an initial gel size-selection of the 70-mer? It would get rid of the large and small pieces. The Qiagen gel extraction kit claims it retains 70bp-10kb. The PCR retains 100bp-10kb. My tags are 80-85 bp. I think the percentage TBE gel used by Shendure et al is also pretty low. Mention long pieces site this gel and previous agarose gel
To Do!!!  Run the digestion again with 2x? Ethanol precipitate 1 and Qiagen gel extract the other. Run both on gel. Maybe run 3x and biotin select one?

Elution of size-selected tags from gel

Oct 18, 2006
The tags were cut from the gel with a razor blade and all four gel pieces were transferred to the same tube containing 600 ml of elution buffer [10 mM Tris-HCL (pH 7.5), 50 mM NaCl, 1 mM EDTA (pH 8.0)] (see on page for details).
The elution was left overnight. PCP extracted and EtOH precipitated and resuspended in 20 ml of EB buffer. I was supposed to run 2 ml on a gel. I screwed up and only ran 1 ml (Figure ). The gel was to quantify the yield from the gel elution. Although, I only ran 1 ml , I (and the versadoc software) could still pick up the faint band. The versadoc software estimated this band to be 8.2 ng/ml .
Please see the pdf version for figures
Figure 6.22: 6% TBE polyacrylamide gel for quantifying yield. 1 ml
Brief Conclusions:   I gotta use EtBr with these diagnostic gels, the Sybr stuff just sucks on the versadoc. My yield was 8.2 ng/ml a little less than the 12.5 ng/ml reported in the Shendure paper.

End-repair of eluted tags

Oct 19, 2006
Used 13.75 ml of the tag (100 ng as in the Shendure paper) in a 20 ml End-Repair reaction [Epicenter]. Used 2 ml 10x buffer, 2 ml ATP, 2 ml dNTP, 0.25 ml enzyme, 0.75 ml H2O . I heat deactivated and EtOH precipitated. Resuspended in 10 ml of EB buffer.

Ligating on the linker

Oct 19, 2006
I used 2 ml T4 buffer, 10 ml tag (all), 3 ml linker 1, 3 ml linker 2, and 2 ml ligase. I let the reaction go overnite at 16C.
Oct 20, 2006
I took 0.5 ml of the ligation and ran a PCR on it to try and enrich the correctly ligated pieces using PCR. I didn't deactivate the ligase (assuming 95C from the PCR would denature it anyways). I did deactive the remaining 19.5 ml of the ligation. The PCR reaction was 15 ml EasyA, 13 ml H2O , 1 ml RDV2F primer (10 uM), 1 ml FDV2F (10 uM), 0.5 ml ligation rxn.
The PCR product and the 20 ml of ligation ligation were run on a 6% polyacrylamide gel (Figure ).
Please see the pdf version for figures Please see the pdf version for figures
Figure 6.23: a) the ligation product (No PCR) and the PCR producted from amplifying 0.5 ml of the ligation product. b) zoomed in on the No PCR lane
Brief Conclusions:   Obviously this wasn't the best looking result in the world (see Figure 6.23). The gel is painfully messy (again the SybrGold problem). The PCR reaction created a giant smear (Figure 6.23a fourth lane). The ligation without the PCR had a few bands, the strongest band is the tag, but it is surrounded by a smear, not by other bands. What I really wanted to see was three bands: 135 bp = RDV + tag + RDV, 157 = RDV + tag + FDV (band we want), 179 bp = FDV + tag + FDV.
I certainly think it would be better to add phosphates to the linkers, that will vastly increase the efficiency of the ligation and will prevent having nicks in the DNA. The downside is that it will create three concatamers: RDV+RDV, FDV+FDV, and RDV+FDV. However, this is no different than the adaptors I ligated together earlier to the cDNA, and they're so short it won't be hard to pick them out (or remove them with a microcon column).
If you look close you'll see the PCR lane has 3 bands (Figure 6.23a), but the biggest band is about 20bp shorter than I'd expect.
The gel was hard to photograph but here is the summary of my observations by eye: PCR product lane: giant smear with bands at 45bp (presumably the primer), another band at 65ish (presumably the ligation together of the two primers to each other, then 3 bands at 100, 120, 140 (perhaps the correct bands but they are 30 bp shorter than I thought they should be. they giant smear continues all the way to the edge of the gel
Ligation lane: the original band of 85 bp was clearly visible, but the rest was just a smear of different sized stuff from 500 bp down to 50 bp
I want to run just the dsPrimers together and see if they form distinct bands.
Brief Update Fri Nov 10 11:55:10 EST 2006:   I mentioned above that I was getting a bands at 45 and 65bp. Now I'm pretty sure those were my primers (see Figure ). I don't know why yet, but the primers are migrating much slower than their dsDNA size...

Running the primers only on a gel

Oct 20, 2006
I decided maybe the smearyness (Figure 6.23) is due to not having my primers purified (which would create an assortment of lengths). ssDNA is supposed to be run on a UREA gel; I didn't have one, so I ran them (and a PAGE purified 60mer that ilaria gave me) on a 6% TBE polyacrylamide gel (Figure ). The gel was stained with SYBR gold.
Please see the pdf version for figures
Figure 6.24: 6% TBE polyacrylamide gel for looking at ssDNA
Brief Conclusions:   Once again the SYBR gold make imaging impossible on the versadoc (Figure 6.24). The bands are very smeary and look like black-hole negative signals rather than positive white signals. The black spots do seem to be in the correct location. I think the secondary structure is killing me and I need to use the UREA gel. A UREA gel might even help with detecting the ligation of the linkers????

Running the primers only on a Urea gel

Wed Nov 8 16:13:14 EST 2006
I'm rerunning the primers on a gel. This time I'm using a 15% TBE Urea gel [Invitrogen]. I diluted the DNA 1/2 in TBE Urea loading dye [Invitrogen]. I did not do this with the ladder. For the ladder, I used the normal one that I use for TBE polyacrylamide gels. This certainly didn't look like it worked . I guess I need to make a ladder with Urea dye next time.
Please see the pdf version for figures
Figure 6.25: I didn't use the right kind of dye for the ladder lane. The fourth lane from the right has the wrong kind of dye and is very diffuse.
The same amounts as in the previous approach were used, except I only ran one ladder and I did not run two different concentrations of the FDV2 and RDV2. The primers were at 10 mM and I used 6 ml of each. As described in the invitrogen manual, I heated the dye/samples to 70C for 3 minutes. Then placed on ice. I also flushed the wells of the gel three times with 100ml TBE. I ran the gel at 180V for 50 minutes. I then dyed the gel for 20 minutes in 50 ml of 2 mg /ml EtBr and washed it for 10 minutes in H2O .
The amps were a little lower than the Invitrogen protocol said they should be. But the voltage was fine.
Please see the pdf version for figures
Figure 6.26: primers urea gel
Brief Conclusions:   Next time use the correct kind of dye with the ladder. It looks like there is too much DNA in both the Ilaria primer lane and in the dsDNA lanes 6.26. Perhaps the "black spot" effect that I've seen several times is due to excessive DNA? Run the next gel with less DNA, especially the dsDNA. I should also run FDV2F, FDV2R, RDV2F, RDV2R, dsFDV2, and dsRDV2, just to make sure that there isn't something wrong with the reverse primers. Maybe add a little salt to the dsDNA so that it will anneal tighter? Could probably get by with only 0.5 of ilaria's primer if I run that one again. There is some hint of impurities in the primers (shorter pieces), but it is very weak relative to the real signal

Running the primers only on a Urea gel, less DNA, more salt

Thu Nov 9 12:14:14 EST 2006
I'm going to try adding salt to the dsDNA to let them anneal. I ran the dsFDV2 and dsRDV2 with and without 50 mM NaCl2. I also want see if less DNA removes the black hole problem. So instead of the concentrated annealed DNA, I ran 3 ml of 10 mM forward and reverse in each lane (so the dsDNA should have the same amount of DNA as the ssDNA).
Please see the pdf version for figures
Figure 6.27: primers urea gel2
Brief Conclusions:   I hate when things start to become clear, and it leads you to a point where you're not quite sure what to do. What is starting to become clear is that my annealed primers with 4bp overhands don't migrate exactly according to their length (Figure 6.27). They always move slower than their single-stranded counterparts. You might conclude that it is the double-strandedness that is slowing them down, but this isn't the case when you look at the ladder whose bands match the single-stranded primers. The 42/47bp dsDNA band is migrating at 60-65bp. The 29/25bp dsDNA band is migrating at around 45 bp. So it seems like these overhangs are slowing them down by appx 15bp? Very strange but it seems to explain two of the bands that I couldn't figure out in section 6.7.1.1 on page pageref (also see Figure 6.23).
It is also clear that these things are pretty smeary. Maybe purification (HPLC or PAGE) would do some good for that problem. I'm starting to favor the following idea: Do RCA, clean with EtOH precipitation, digest with mmeI, clean up in a qiagen gel cleanup column (this will remove the very small and the very large fragments 27), Run the end repair reaction. Cleanup again with the Qiagen column (don't know if this too much and will remove too much DNA), spec on nanodrop, ligate on phosphorylated adaptors (each 10bp shorter than in previous round), heat deactivate (EtOH?), enrich with the dynal beads and the 30mer tag, wash off enriched DNA and concentrate in a microcon 30K or 50K (will remove short stuff), run on TBE urea gel and cut out the correct band (you could even synthesize the incorrect bands and the correct bands and test the enrichment?, would be expensive)
To Do!!!  buy 30mer w/ dual biotin. buy shorter primers (17mer and 30mer) with phosphates, buy one with overhang, one without (so I can run on a gel and compare for this weird shift problem); mix dsFDV2 and dsRDV2 into same lane and run. Try nusieve? buy 50K microcon. run primers through 30K and 50K and see how it does at removing them. see: http://hcgs.unh.edu/protocol/msat/CAenrich.html for dynal instructions
Bought
Dualbiotin-spacer18-CGGATCGGCCAAGGCGGCCGTACGGATCCG appx 300 bucks!
try to anneal at 0.1M rather than the recommended 1M (natasha said it's to hard to remove the strand at that conc).
when linking use 50 (molar) fold excess? checkout http://hcgs.unh.edu/protocol/msat/CAenrich.html for linker ligation ideas too (they use 2 ml of ligase)
AATTAACCACTACG  FDV3   5' CCTCCGCTTTCCTCTCTATGGGCAGTCGGTGAT
    TTGGTGATGC               GGCGAAAGGAGAGATACCCGTCAGCCACTA-P 5'

GATCAACTGCC     RDV3   5' CCGGGTTCCTCATTCTCT
    TTGACGG                  CCAAGGAGTAAGAGA-P 5'

6.7.2  Ligating ends using a PCR product

I don't want to keep wasting my material (of which I only have a little bit left before I need to make more cDNA). Ilaria had a pair of primers for amplifying the pLtet promoter. She had two forward primers and one reverse, the first forward primer yields a 81bp product. The second yields a 120bp product. These two pieces give me a nice way to test ligating the two different ends on simultaneously in a cleaner system where I can try to optimize the ratios if they need to be optimized. And I can use them with the primers to test the removal ability of the Qiagen PCR/gel columns and the microcon columns (see section on page for more on this). Last, even though the results from the primer removal weren't that promising, having shorter primers in this next set of ligations enables me to bump up the molar ratio of adaptors without raising the quantity of DNA I use. A big problem with the previous attempt was that I had so much primer I couldn't see what was going happening on the gel.

Observing linker primers

Nov 16 and 17, 2006
I wanted to have a look and make sure I got only one band when running the primers out and not a smear. The first time (Figure a), I didn't use enough primer. The second time (Figure b), the amount was plenty for the 30mer but still to short for the 15mer with the phosphates.
Please see the pdf version for figures
Figure 6.28: primers urea gel3
Brief Conclusions:   The double stranded overhang piece migrates slower. The piece that migrates as single-stranded creates a second band (see in particular Figure 6.28b columns 5 and 7).

Ligating ends first attempt

Nov 17, 2006
I'm trying three reactions: 1) primers only 2) primers + 80mer, 3) primers + 120mer. I made 50 mM stocks of the 4 primers, and use 1.8 ml of each for the ligation (this is about 25x molar excess and corresponds to 1.8 mg of the large adaptor and 900 mg of the small one). I used 200 ng of the 80mer and 120mer blunt Phusion PCR products (4 ml ).
Ligations were run for 12 hours at 16C followed by heat inactivation of the T4 ligase at 65C for 20 minutes.
Nov 19, 2006
I ran 5 ml (1/4) of the above ligation reactions on a 15% TBE Urea gel and post-stained the gel with EtBr (see Figure ).
Please see the pdf version for figures
Figure 6.29: 12 hr ligations of excess adaptor primers to an 80-mer and 120-mer PCR products were run on a 15% TBE Urea gel for 50 minutes at 190V.
Brief Conclusions:   Tue Nov 28 16:32:42 EST 2006
A look at the primer only lane shows that these primers are definitely ligating into longer pieces than they should (Figure 6.29). The primers have an overhang and are only phosphorylated on one end. So it shouldn't be possible to have long fragments and the possible bands should be much more continuous than I see here.
Next time run one lane with the same conc of primers and PCR, but without ligase, so I have something to compare to in the figure (see Figure 6.29). Above, I mentioned that the primers are about 25x molar excess. This leads to a total primer pair molar excess of 50x. I think I need to lower this if I'm going to be able to see anything besides the primers on the gel. I also think raising the ligation temperature might help make things more stringent. I should try again with unphosphorylated primers like Shendure used. I should definitely try shorter ligation times.

Ligating ends second attempt

Nov 29, 2006
I lowered the primer concentration by 1/2 in the hopes of having clear results rather than a big smear of DNA. I also used 0.5 ml (200U), instead of the typical 400U. I varied the incubation times and used unphosphorylated primers (RDV2 and FDV2). Tested primer combinations and time variations are show in Figure .
Please see the pdf version for figures
Figure 6.30: ligations of excess adaptor primers to an 80-mer run on a 15% TBE Urea gel for 50 minutes at 190V.
Brief Conclusions:   Mon Dec 4 17:56:12 EST 2006
This experiment was informative, unfortunately I'm not there yet. One thing to consider is that the 5' ends of the 80mer PCR product are NOT phosphorylated because the PCR primers aren't. The unphosphorylated priemrs like Shendure used do prevent self-ligation (see lane 3 and lane 10). However, they also don't seem able to ligate on their (at least not in the amount of time I used). On the other hand, I'd say the phosphorylated oligos are too efficient with this amount of ligase. I think 50-100U might be even better.
One thing that is wierd, the single phosphorylated dsDNA adaptors behave just as expected (lanes 1 and 2). They concatenate one time and make a second band. But when I stick the two adaptors into the same reaction (lanes 4 and 5), they ligate together in all kinds of different ways. I think I'll keep 15 minutes as the ligation time and try titrating the T4 units down.
To Do!!!  make more 80mer, keep concentrations of the primers the same. use only 15 minute ligation time, but titrate down the amount of ligase (200, 100, 50). stay at 25C. try UREA and a TBE 6% (10ml on one 10ml on the other).

Ligating ends third attempt

Mon Dec 4, 2006
I made more of the 80mer28 with 3, 100 ml PCR reactions. For each rxn I used 4 ml of 10 mM primer, 11 ng of plasmid, 45 ml H2O , 50 ml Phusion mastermix; 30 sec 98C denature, then cycle: 5 sec 98C, 15 anneal, 15 extend at 72C. Ran 30 cycles, the first 5 were annealed at 60C, the last 25 were at 67C.
Tue Dec 5 20:31:55 EST 2006
I combined the 3 rxns into 2 tubes (150 ml each) and cleaned them up with a Qiagen PCR purification kit. To one of the two cleaned up rxns, I added 5 ml of T4 ligase buffer and 1 ml of T4 polynucleotide kinase (so that the blunt PCR producted would have 5' phosphorylations). I incubated at 37C for 30 min and cleaned up the rxn with a Qiagen PCR cleanup kit. The yields from all of these steps were:
Sample DNA (ng/ul) 260/280 260/230 total yield
PCR sample 1 98.1 4.9 mg
PCR sample 2 90.8 4.0 mg
PCR sample 2 (after phosphorylation) 80.9 3.2 mg
I did my primer preparation a little differently this time.   Instead of mixing all the stuff together, heating it up, cooling it, and adding ligase. I annealed all the primers separately in STE buffer (TE plus a 50 mM NaCl). I heated the adaptors up to 95C and dropped the temperature by 2C every 30 seconds until 4C.
I ran several different combinations of primers using different amounts of ligase (200, 50, or 10 Units per rxn). Half of the 20 ml rnxs were run on a 15% TBE UREA gel and a 6% TBE gel . Contrary to the previous attempt at these ligations, I ran this ligation at 16C.
Please see the pdf version for figures
Figure 6.31: ligations of excess adaptor primers to an 80-mer run on a 15% TBE Urea gel for 50 minutes at 190V. I titrated the ligase concentration.
Please see the pdf version for figures
Figure 6.32: ligations of excess adaptor primers to an 80-mer run on a 6% TBE Urea gel for 50 minutes at 200V. I titrated the ligase concentration. Unfortunately the gel went all crooked. For some reason it started out that only one side was moving. I didn't notice until 30 minutes later. I pushed the lid on tighter and it fixed the problem but the bands were already broken beyond repair.
Brief Conclusions:   Wed Dec 6 13:06:53 EST 2006
Houston we have progress! Finally, some insights that are leading to (hopefully) finishing this crap up. So the reduction of ligase concentration definitely helped out. The contrast between 200U and 50U is quite dramatic (see Figure 6.31 lanes 1 vs 2, 3 vs 4, 8 vs 9). This ligase unit reduction does little if anything to change the amount of the ligation products we're interested in (and when it does change them it makes more not less of the correct bands, because they don't end up in long concatamers). The reduction does lead to more of the unligated adaptors (this is to be expected and is a good thing). The 10U rxn was even better and is the first time that it is clear that different adaptors have been ligated to the 80mer PCR product. The 15% TBE Urea gel doesn't do the best job of showing this 6.31, because the separation isn't so good at the size. Would also be useful to have a DNA ladder to size confirm this stuff a little better.

Ligating ends fourth attempt

Dec 6, 2006
Based off the previous results (Figure 6.31), I diluted T4 ligase even further. Using 50, 10, 5, and 2 Units. This time I ran them on 6% TBE gels only (see Figure ). I ran the first gel too long, so I ran a second one (Figure ) and replaced lanes one and two with a 25bp and a 10bp ladder respectively.
Please see the pdf version for figures
Figure 6.33: ligations of excess adaptor primers to an 80-mer run on a 6% TBE Urea gel for 45 minutes at 200V. I titrated the ligase concentration. Unfortunately, I ran the gel for too long. The band lengths I would be interested in are 30, 62, 82, 97, 112, 113, (128), 144.
Please see the pdf version for figures
Figure 6.34: ligations of excess adaptor primers to an 80-mer run on a 6% TBE Urea gel for 45 minutes at 200V. I titrated the ligase concentration. Unfortunately, I ran the gel for too long. The band lengths I would be interested in are 30, 62, 82, 97, 112, 113, (128), 144. This gel is the same as Figure 6.33 except that the first to lanes were replaced with the 25bp and 10bp ladders from Invitrogen, and the gel wasn't run so long.
Brief Conclusions:   Mon Dec 11 15:52:52 EST 2006
I ran the first gel too long. However, I can still see the size-range I'm interested in. The number of bands is correct (Figures 6.33 and 6.34), though the correct bands are too faint. The sizes (in bp) of the bands are too high based off what I'd predict: 78=62, 82=92, 97=100, 112:113=117, 128=130, 144=152. The second gel (Figure 6.34) leads me to believe that maybe I made the run PCR product as the band is at 120 bp not at 80 bp. Despite the differences in the absolute bp estimates from versadoc, the relative difference in base pairs is similar in the gel to what is expected (expect:gel1:gel2, 15:18:14, 6:8:4, 15:17:15, 16:18:16). I have no idea what that big band is. Based on gel 2, I'd guess it is the concatamerized PCR product. But for gel 1, the numbers don't work out right. Notice that the band that corresponds to two different possibilities is the brightest concatamer band (the number in both figures is in italics). This band correponds to FDV3:PCR and RDV3:PCR:RDV3.
I'm pretty much there, the most important thing is to get that band a little brighter (and thereby making the PCR product band a little less strong). I think I need to increase the concentration of each primer, try 50 and 100 U of T4 ligase and ligate for 30 minutes instead of 15. Even as it is now though at least I can see the correct band.

6.7.3  Ligating ends using an RCA product again

Mon Dec 11 19:06:17 EST 2006
I'm starting to get the hang of things with the PCR product. Now I want to go back to the RCA product and see if I can get things working with the MmeI digested, end-repaired tag.
A modification of my earlier approach   I'm leaning towards the idea of not size-selecting before I end-repair and ligate on the two different adaptors. I feel like I lose too much and make the DNA too junking by selecting the thing from the polyacrylamide gel. Instead I'm going to use a Qiagen PCR cleanup kit to remove the large undigested RCA stuff (which I know from previous experience it does 29) and the small nucleotide stuff that would use up the enzymes in the end-repair kit and lead to ligation junk problems in the adaptor ligation steps.
Then I'll ligate my adaptors to all of the Qiagen cleaned up RCA products. This should result in a smear with 5 bands (82, 97, 112.5 [big band], 128 [band we want], 144). I might also try use the dual-biotin oligo to capture and further enrich the 128mer tag prior to running the 6% gel (might need to be UREA if I use the biotin-capture) and size-selecting the correct band for PCR amplification.

testing the Qiagen cleanup and biotin oligo selection procedures

Mon Dec 11 19:21:48 EST 2006
I'm digested 15 ml (10 mg ) of RCA product (the one from page pageref) in 5 ml NEB4, 5 ml MmeI, 2 ml SAM (diluted 1/20 from the 32 mM stock), 23 ml H2O . The digestion was for 30 min 37C followed by heat deactivation for 10 minutes at 65C.
I left 12.5 ml and cleaned up the remaining 37.5 ml with a Qiagen PCR cleanup kit to remove the very large and very small DNA. I eluted into 31 ml and spec'd the resulting DNA on the nanodrop:
Sample DNA (ng/ul) 260/280 260/230 total yield
digested RCA after cleanup 56.7 1.76 mg
Biotin-oligo DNA capture: first try  
Based on a recommendation from Natalia Broude, I'm not going to use the recommended Dynal salt concentration of 1M30. I'm using a Binding and Wash buffer (BW) of 100 mM final NaCl concentration (10 fold lower than recommended). Binding and wash buffer is just TE with salt added.
I'm going to use the following protocol:
  1. wash 5 ml of Dynal beads in 5 ml of 2x BW buffer
  2. repeat step 1
  3. resuspend beads in 12.5 ml of 2x BW buffer
  4. add 12.5 ml of oligo plus the dual-biotin 30mer (this will be 1 ml of the 50 mM working stock) 31.
  5. heat the 25 ml soln to 98C for 2 minutes
  6. flick tube
  7. incubate in 50C for 1 hr to binding oligos; give tube a few flicks every 15 minutes or so
  8. wash 2x in 1x BW buffer
  9. resuspend in 25 ml of 0.1N NaOH at 50C for 5 minutes to remove captured oligo
  10. place on magnetic stand one minute to capture beads; keep the supernatant (the solution containing the now freed oligo)
  11. add 25 ml of 1M Tris (to adjust the pH back)
  12. add 425 ml of TE (to make the near the maximum volumn for the microcon column)
  13. concentrate with YM30 microcon; spin 12min at 14000g 32
I loaded the uncleaned up RCA, the Qiagen cleaned RCA, the oligo selected RCA and a NEB lowMW ladder and a Invitrogen 25 bp ladder (0.5 ml = 0.5 mg ) onto a 6% TBE polyacrylmide gel. The original Qiagen cleaned DNA floated right out of the well (Figure , maybe it has something to do with EB buffer?). So I added 2.5 ml TE and 3 ml of TBE Hi Density buffer to the last remaining part of the Qiagen cleaned sample and loaded it into the final lane.
Please see the pdf version for figures
Figure 6.35: The Qiagen EB buffer plus TBE Hi Density buffer would not settle into the bottom of the well and diffused out.
I ran the gel for 35 minutes at 200V. Stained 20 min in EtBr and 10 min destain in H2O (Figure ).
Please see the pdf version for figures
Figure 6.36:
Brief Conclusions:   Wed Dec 13 19:19:08 EST 2006
The float away lane was faint as expected (Figure 6.36 lane 3). But lanes 2 (uncleaned digestion) and lanes 6 give the cleanest view so far of our PET. Running one-fourth of one digestion seems a lot clearer than previous fuzzy bands on hard to interpret gels as occured in my previous attempts of loading the entire digestion on one lane. Also using EtBr was a big benefit over Sybr Gold with the versadoc (see Figures 6.22, 6.21, and 6.20 for previous gels with the PET tags).
The biotinylated oligo selection either didn't work or didn't use enough DNA so that not enough DNA was recovered to make a visible band (Figure 6.36 lane 4). Overall, I think things look pretty promising for adding the adaptors to the RCA products.
try to add adaptors   Dec 13, 2006
The above experiment showed me that the PET tag is fairly clean when I don't have too much DNA (Figure 6.36). Now I want to see if I can get the adaptors on there. Maybe in previous attempts I just had too much DNA?
I'm going to do the same digestion: 15 ml (10 mg ) of RCA product (the one from page pageref) in 5 ml NEB4, 5 ml MmeI, 2 ml SAM (diluted 1/20 from the 32 mM stock), 23 ml H2O . Digest for 30 min at 37C followed by heat deactivation for 10 minutes at 65C. After cleaning the rxns with a Qiagen PCR cleanup kit the yields were:
Sample DNA (ng/ul) 260/280 260/230 total yield
digested RCA after cleanup 63.3 2.2 mg
Please see the pdf version for figures
Figure 6.37: The tag band is appx 82 bp. After that we expect to see: 97, 112:113, 128, 144.
Brief Conclusions:   Thu Dec 14 23:01:12 EST 2006
I can't seem to get a good polyacryamide gel any more. They've always been finicky but this is ridiculous. I don't know how well this worked because the gel sucks (Figure 6.37). It doesn't look horrible (the rxn not the gel); it looks like I had too much adaptor primer.
To Do!!!   use less DNA (split into 3), use less primers, clean DNA after ligation, keep same amount of ligase, run 2 two parallel ligations tomorrow, 1) use polyacrylamide, 2) use TAE and Nusieve; run for a LONG time
try to add adaptors again  
Thur Dec 14, 2006
Using the same protocol as yesterday, but today I'm running two digestions, one for a polyacrylamide gel and one for a Nusieve gel.
RCA digestion yields were:
Sample DNA (ng/ul) 260/280 260/230 total yield
digested RCA sample 1 after cleanup 78.0 2.7 mg
digested RCA sample 2 after cleanup 70.9 2.5 mg
After end repair and clean up of the end repair with a Qiagen PCR kit, yields were:
Sample DNA (ng/ul) 260/280 260/230 total yield
end-repaired PET sample 1 after cleanup 71.9 2.2 mg
end-repaired PET sample 2 after cleanup 71.5 2.2 mg
For the ligations I used 1/3 of the cleaned up sample 1 from the table above for each of the two reactions (2/3 total). I also ran 1/6 of the unligated RCA digest on a polyacrylamide gel with the two ligations. I cleaned the 20 ml ligations with a Qiagen PCR kit and eluted into 30 ml . I ran 20 ml (because the entire 30 ml wouldn't fit) on a polyacrylamide gel (Figure ).
Please see the pdf version for figures
Figure 6.38: 20 ml of the 30 ml Qiagen cleaned ligation reactions were run on a 6% polyacrylamide gel.
Once again having problems with consistent running of the Novex precast gels, I decided to run the remaining 10 ml of each ligation and the remaining 1/6 of digested RCA onto a 3.5% Nusieve TBE gel Figure .
Please see the pdf version for figures
Figure 6.39: 10 ml of the 30 ml Qiagen cleaned ligation reactions (the left overs from Figure 6.38) were run on a 3.5% Nusieve TBE. The image on the right has been background subtracted using the Versadoc software.
Brief Conclusions:  
Well, there are some nice conclusions that we can make here: 1) Novex polyacrylamide gels either suck or I suck at using them (Figure 6.38); 2) Nusieve is much better (Figure 6.39); 3) Nusieve requires much more DNA per lane (compare the 25 bp ladder between the two figures); 4) the exACTGene 50 bp ladder is a beautiful addition because it has a band at 100 and at 112 bp which is very similar to two of the bands I'm really interested in (because if I can find those two then I know that the band I want comes next); 5) add the 25 bp ladder and you have pretty much a ladder band corresponding to every adaptored PET I am interested in (see my bp annotations on the side of the bk subtract gel). 6) If you look really close (and I stress the really in the background subtracted image I think the adaptored PET band of interest might be faintly visible
Brief Update Tue Dec 19 15:11:48 EST 2006:   When I ran the Nusieve gel (Figure 6.39) the mAmps was very high and I had trouble reaching a decent voltage. The next time I ran a Nusieve gel, I used 0.5x TBE and the mAmps were much lower and it seemed like the gel didn't get as hot.
try to add adaptors again, again; this time we'll cut the gel slice  
I used sample 2 from the above RCA digestion. I ligated it in the same way as above and ran 20 ml on a Nusieve gel (previous Nusieve gel only had 10 ml on it).
Please see the pdf version for figures
Figure 6.40: 20 ml of the 30 ml Qiagen cleaned ligation reactions were run on a 3.5% Nusieve TBE. The appx 128 bp band was extracted.
I cut out the band for 50U and 100U at approximately 128 bp.
Brief Conclusions:   Things are looking up since I switched to Nusieve (Figure 6.40). I've heard heard MetaPhor has even better resolution but is very fragile and not low-melt (so you have to do some sort of gel cleanup). I think I still need to run more DNA. I also need to use more DNA in the ligation. Last, I would forgo the EtBr and just use sybr gold and the crappy imaging system. I post-stained the gel with SybrGold after post-staining it with EtBr for imaging. Sybr gold isn't nearly as bright if EtBr was on the DNA first. I cut out the bands at the correct sizes. The bands were pretty dang faint though, so hopefully next time I have a slightly more obvious band to cut out.

6.8  Amplifying the adaptored tags

6.8.1  PCR Primers

I want to make a variety of primers for amplifying the adaptored DNA so I can clone it or do what ever else I like. Below are the primers I'm using for now
::
Primers from Shendure et al:

Unmodified forward primer
5' CCTCTCTATGGGCAGTCGGTGAT

Reverse primer
5' CTGCCCCGGGTTCCTCATTCTCT

-------------------------------------------------

Forward primer for directional cloning EcoRI
5' TAGAATTCCCTCTCTATGGGCAGTCGGTGAT


Reverse primer for directional cloning BamHI
5' ATGGATCCCTGCCCCGGGTTCCTCATTCTCT


-------------------------------------------------
Forward primer for directional cloning USER
5' CCTCTCTATGGGCAGTCGGTGAT

Reverse primer for directional cloning USER
5' CTGCCCCGGGTTCCTCATTCTCT

PCR amplification of cut adaptored tag   Mon Dec 18, 2006
I PCR amplified the band taken from 6.7.3.1 on page pageref (Figure 6.40). I heated the gel slice up to 65C for appx 20 minutes and used 5 ml in a 50 ml PCR with Phusion Taq. I ran one reaction with each of the two gel slices taken and one rxn with each of the primers sets 1) EcoRI/BamHI added primers AND 2) blunt primers. The blunt primers used an annealing temperature of 63C for all cycles. The sticky end primers used 63C for the first 5 rounds and 66C for the remaining 25 rounds.
Used 10 mM primer, melt 10 sec 98C, extend 15 sec 72C. PCR rxns were run out on a gel .
Please see the pdf version for figures
Figure 6.41: 5 ml of Nusieve GTG agarose gel slice (128bp) was amplified for 30 cycles in a 50 ml rxn.
Brief Conclusions:   It appears that a band of approximately the correct 128 bp size is visible. But several other bands are present as well.
PCR amplification of cut adaptored tag: try 2   Thu Jan 4, 2007
I repeated the PCR from the gel as I did on Dec 18. I wanted to try and determine why the blunt rxn didn't work. I ran everything like above except I didn't use the two step proceedure with the sticky ends. I keep all the rxns at 63C for the entire time. The PCR was run out on a 2% SB gel .
Please see the pdf version for figures
Figure 6.42: 5 ml of Nusieve GTG agarose gel slice (128bp) was amplified for 30 cycles in a 50 ml rxn.
Brief Conclusions:   I looks like the blunt primers consistently fail (Figures 6.41 and 6.42.

6.9  Cloning the adaptored tags

Now that it looks like I might have the correctly adaptored PET tag band cut and amplified, I'm gonna try to clone and sequence it.

6.9.1  cloning the sticky PCR product into pUC19

Fri Jan 5, 2007
The sticky PCR was the only one to work consistently (Figures 6.41 and 6.42), so that's what I'm gonna try to clone into pUC19.
I cleaned the PCR reactions with a gel clean up kit, because the 5 ml in the PCR made the PCR rxn pretty gelly. In retrospect, I don't know why I didn't cut the band out since there were clearly multiple bands in the PCR? The specs of these clean ups looked pretty dirty. The yields from the clean up of all four rxns were:
Sample DNA (ng/ul) 260/280 260/230 total yield
blunt 50 44.2 1.33 mg
sticky 50 42.5 1.28 mg
blunt 100 38.5 1.16 mg
sticky 100 104.0 3.12 mg
I cut 2 mg of pUC19 with EcoRI and BamHI and cleaned it up with a Qiagen PCR cleanup kit. I cut all of the purified sticky 50 and sticky 100 PCR products with EcoRI and BamHI and cleaned them in the same manner.
I ligated the sticky 50 and sticky 100 in two different rxns consisting of: 1 ml T4 ligase, 2 ml T4 ligase buffer, 8 ml cut PCR, 2 ml cut PCR (I meant to use 1 ml but screwed up), 8 ml H2O . 2 ml of each ligation was tranformed into DH5alpha compentent cells.

6.9.2  checking the clones

Mon Jan 8, 2007
I picked five sticky-50 clones (samples 2a-e) and 6 sticky-100 clones (samples 4a-f). I PCR amplified them with the M13 primers. The primers should add 80-90 bp, so the final piece should be around 200 bp. I used 1 ml of miniprep in each rxn. The first nine rxns were run on an agarose gel (Figure ).
Please see the pdf version for figures
Figure 6.43: Miniprepped pUC19 cloned 128bp sticky adaptored PET tags were amplified with M13 primers
Tue Jan 9, 2007
I spec'd five of the miniprepped samples to send them to agencourt for sequencing: 2b, 2c, 2d, 4a, 4b.
Sample DNA (ng/ul) 260/280 260/230 total yield
tag 2b (2) 235.2 11.8 mg
tag 2c (3) 190.3 9.51 mg
tag 2d (4) 272.0 13.6 mg
tag 4a (7) 104.0 10.2 mg
tag 4b (8) 104.0 12.4 mg
600-1000 ng of each of the five samples was sent to agencourt for sequencing using the M13F primer.
Brief Conclusions:   Didn't work, no tags were in the pUC19 vector.

6.10  Cloning the unadaptored tags

Many moons ago, when this project was going so well, I skipped sequencing the unadaptored PET tags because I thought it was a waste of time, given that everything looked correct. Not that nothing works, and I haven't made forward progress since that day, I'm taking a step back to try and figure out what went wrong.

6.10.1  blunt coloning unadaptored tags try 1

Tue Jan 30, 2007
I cut the tags from RCA2b (15 ml = 10 mg ), ran them on a gel , and cloned them into an Invitrogen Zero Blunt PCR for Sequencing kit. I sent five of the clones out for sequencing today. Although now that I look at the length of the inserts in detail, unfortunately, it looks like the PCR check of insert verification is NOT long enough.
Please see the pdf version for figures
Figure 6.44: RCA2b tag was cut from lanes 3 and 4 and combined into one Qiagen gel cleanup
The yields for the minipreps were (sequenced ones are in italics)
Sample DNA (ng/ul) 260/280 260/230 total yield
(1) 327.3 16.4 mg
(3) 68.8 3.4 mg
(4) 133.9 6.7 mg
(5) 328.0 16.4 mg
(6) 539.7 27 mg
(7) 493.5 24.7 mg
Please see the pdf version for figures
Figure 6.45: Miniprepped Zero Blunt PETs, inserts don't seem to be long enough (PCR minus insert should 170bp, with insert should be 250 bp)
I'm trying again with more RCA DNA. This is the last drop I have of RCA DNA, so hopefully this works...
Brief Conclusions:   The sequences came back with no inserts. Technically, having the same vector close on itself is impossible with the Zeroblunt kit, because that would result in there being expression of the lethal ccdB toxin. However, in all the sequences I got back there was no insert, but rather one or two base-pairs had been added/removed33 from the vector resulting in a ccdB gene that was frame-shifted and therefore not a properly folded lethal protein.

6.10.2  blunt coloning unadaptored tags try 2

Wed Jan 31 17:24:57 EST 2007
I'm trying again, but this time I'm digesting more of the RCA. Actually, I'm digesting all of the remaining RCA, so hopefully this works! I only had about 30 ml (20 mg ) of the RCA. Which I digested with 15 ml of MmeI at 37C for 15min and heat deactivated at 65C for 10 minutes. Instead of a Qiagen Cleanup, I did an EtOH precipitation, and eluted into 15 ml of TE, added 2.5 ml loading dye and ran it all in one lane of a 3.5% TBE gel with Sybr Safe.
Thur Feb 1, 2007
I cut the gel on the transilluminator (at jpeg picture here) and cleaned it up with a Qiagen gel clean up kit, eluting into 34 ml of EB buffer. I then repaired the ends with a new End-it End repair kit (note: I should throw away the old one, it's been freeze-thawed too many times). I increased the volume to 400 ml with TE cleaned up the end-it reaction with Phenol/Chloroform and then with Phenol. Then I concentrated the DNA with EtOH precipitation and eluted into 15 ml of TE.
I used 4ml of 15 ml of end-repared in the ZeroBlunt reaction [Invitrogen].
Fri Feb 2, 2007
I picked eight colonies from the ZeroBlunt cloning.
Sat Feb 3, 2007
5 of the 8 colonies grew in kanamycin (the plates I used for the cloning were amp, but this vector has two resistance genes). I minipreped the 5 and did a PCR with the M13 primer set.
Mon Feb 5, 2007
Ran the PCRs on a gel (Figure ). Same junk with the ZeroBlunt, looks like no good inserts.
Please see the pdf version for figures
Figure 6.46: PET tags were end-repaired and cloned into the ZeroBlunt kit for the secone time. It looks like once again the insert failed to go in there. (See Figure 6.43 for the other failed attempt)
Brief Conclusions:   I've lost a little faith in the folks at Invitrogen and their ZeroBlunt kit. Clearly I've got an appx 85 bp piece that should be blunt that just won't go into there superdupper vector.

6.10.3  blunt coloning unadaptored tags try 3, going old-school

Mon Feb 5, 2007
Clearly the ZeroBlunt kit and I aren't getting along. I'm going to try this the old fashion way. I digested 2mg of pUC19 with SmaI for 30 min at 25C and 20 min at 65C (to deactivate) to make a blunt vector. I cleaned the rxn with a Qiagen PCR cleanup and eluted into 30 ml of EB. I took 10 ml of the cleaned up digestion and aded 1.2 ml Antarctic phosphatase buffer and 1 ml antarctic phosphatase. I incubated this 30 min at 37C and 5 min at 65C.
For the ligation, I used 1 ml of the dephosphorylated vector, 4 ml of the end-repaired, gel selected PET tags (this was from the 11 ml remaining after I used 4ml in the previous experiment with the ZeroBlunt kit). I used 2 ml of T4 ligase buffer, 12 ml H2O , and 1 ml high concentration T4 DNA ligase [NEB]. I ran the ligation for 10 min at RT and deactivated for 20 min at 65C.
I transformed 2 ml into DH5alpha, incubated on ice 5 minutes, heat shock 30 sec at 42C, back on ice, add 250 ml LB. Shake 200 rpm for 1hr at 37C. Plated 100 ml .
Tue Feb 6, 2007
Only had 3 white colonies, so I picked them all. I also replated the remaining ligation, so I can have more colonies.
Wed Feb 7, 2007
All 3 grew and I miniprepped, PCR'd and ran them on a gel (Figure ). With the replate from yesterday, unfortunately I forgot to add Xgal!!!!!! :( So I initially thought, I'd just have to pick them and hope I had inserts. However, I talked Jamey for a while and he had many ideas from copying the plate with a piece of nitrocellulose membrane to picking them all in replating them with a stamper on a new plate (note the replate had probably around 120 colonies so at least that was good). However, I tried out his riskiest idea. I lifted the gel with a sterilized spatula and squirted around 300 ml of sterile H2O mixed with Xgal. I did this 4 times, breaking the plate into four quarters. For each quarter, I used the amount of Xgal you'd normally use for one gel (so in the end I used 4x the normal amount). The idea (which I pretested on a test gel) was that the liquid would diffuse up to the colonies and allow them to change color without the problem of colony mixing that would occur of you added the liquid to the top of the plate (which would probably screw everything up). I hacked plate in the 37C incubator for about and hour and the blue-white colonies were clear as day. Very nice recovery of a big-screwup, thanks to Jamey!!! I picked 8 colonies and grew them overnight in LB.




Please see the pdf version for figures
Figure 6.47: PET tags were end-repaired and cloned into dephosphorylated pUC19 at the smaI site (first three lanes). It appears the inserts are the proper size given the size of the insert (85bp) and the size PCR band given the additional stuff amplified outside this 85 bp by the M13 primers. The additional two lanes were just (successful) tests of two sets of RT-gtp primers for use in a different project.
Thur Feb 8, 2007
I ran 8 minipreps and PCRs from the picked colonies. I didn't have enough time to run them on a gel.
Fri Feb 8, 2007
I didn't have time to check the insert lengths on the replated colonies before the agencourt courier got to BU. I spec'd all 11 samples and sent the 3 original pUC19-PET samples plus two of the replated ones. Yields for all of the minipreps are (r indicates from the replate; those sent out for sequencing are in italics):
Sample DNA (ng/ul) 260/280 260/230 total yield
a 389.8 19.5 mg
b 511.8 25.6 mg
c 387.3 19.3 mg
r.a 195.3 9.8 mg
r.b 65.0 3.3 mg
r.c 144.6 7.2 mg
r.d 208 10.4 mg
r.e 196.3 9.8 mg
r.f 191.8 9.6 mg
r.g 304.7 15.2 mg
r.h 191.1 9.6 mg
-----   SEQUENCE tag a   -----

 --- TAG_FRONT ---
(4034053..4034070) rrsA, 16S ribosomal RNA (rrsA) 

 Score = 36.2 bits (18), Expect = 4e-04
 Identities = 18/18 (100%)
 Strand = Plus / Minus
                                 
 Query: 1       cacggagttagccggtgc 18
                ||||||||||||||||||
 Sbjct: 4034070 cacggagttagccggtgc 4034053

 --- TAG_BACK ---
 (4033583..4033594) rrsA, 16S ribosomal RNA (rrsA) 

 Score = 24.3 bits (12), Expect = 1.4
 Identities = 12/12 (100%)
 Strand = Plus / Plus

 Query: 7       tgaacgctggcg 18
                ||||||||||||
 Sbjct: 4033583 tgaacgctggcg 4033594

-----   SEQUENCE tag b   -----
no match (only looked by hand, very much not thorough)

-----   SEQUENCE tag c (awfully short)   -----
(4208234..4208252) rrlE, rrlE 23S ribosomal RNA 

 Score = 38.2 bits (19), Expect = 1e-04
 Identities = 19/19 (100%)
 Strand = Plus / Minus
                                  
 Query: 1       cccggttcgcctcattaac 19
                |||||||||||||||||||
 Sbjct: 4208252 cccggttcgcctcattaac 4208234

(4208100..4208117) rrlE, rrlE 23S ribosomal RNA 

 Score = 36.2 bits (18), Expect = 4e-04
 Identities = 18/18 (100%)
 Strand = Plus / Plus

                                 
 Query: 1       ggcagtcagaggcgatga 18
                ||||||||||||||||||
 Sbjct: 4208100 ggcagtcagaggcgatga 4208117

-----   SEQUENCE tag r.a   -----
no match (only looked by hand, very much not thorough)

-----   SEQUENCE tag r.b   -----
couldn't uniquely map tag

Raw data in Word format.
Brief Conclusions:   Finally got that stupid little tag cloned into a vector (Figure 6.47)! Unfortunately, I have less than the agencourt minimum of 5 sequences.

6.11  New circularization adaptors

circularization adaptor 2  5' p-ATCGCA GCATCG ACG
                                   CGT CGTAGC TGCCTA-p 5'

circularization adaptor 3  5' p-ATCGCA AGAGAG ACG
                                   CGT TCTCTC TGCCTA-p 5'


6.12  Concatenating tags for Sanger and Pyro sequencing

Fri Feb 16 17:18:28 EST 2007
I've had so many problems getting two different adaptors to ligate on opposing ends at the same time, that I've decided to try and do this with Sanger sequencing and/or pyro sequencing. The strategy will be pretty much the same for both. Rather than adding two different adaptors, I'm going to ligate a single adaptor (with the blunt side phosphorylated) to the blunt PET tags (Figure ). I'll probably need to prepare a LOT of PET DNA. Or else add an additional ligation step where I add a blunt 20-30mer and do single-primer PCR to amplify this stuff up. I'd prefer to skip PCR.
For pyrosequencing I'll try to time the ligation reaction so that the concatamers are 300-500bp. For the sanger sequencing, I'll aim for > = 800bp inserts.
Please see the pdf version for figures
Figure 6.48: PET tags will be adaptored and concatenated so that multiple tags can be read with a single sequencing read (6-10 tags per Sanger, 1-4 per pyro).

6.12.1  The adaptor sequence for concatenation

Fri Feb 16 17:26:33 EST 2007
I'm making a HindIII based adaptor (see below). Only one end is phosphorylated to prevent concatamers. All that should be possible is a blunt/blunt ligation of the adaptors to themselves plus an adaptor-PET-adaptor ligation (the one we want). The former will be appx 20 bp and the latter should be appx 105 bp. The critical parameters will be the ligation time and the amount of ligase, but they should be easier to optimize than before when I was trying to add two adaptors at the same time. Since the adaptors are so short, I want to do the ligations at 16C. The increase in length from 85 bp to 105 bp should be enough for me to gel select the correct ones. Plus the ones that aren't adaptored will not concatenated nearly as fast in the subsequent concatenation step. As with the adaptoring of the cDNA, I will use a vast excess of adaptor and run the reaction a long time.
HindIII concatenation adaptor 
5' AGCTTGCGAGCG
       ACGCTCGC-p 5'

Brief Update Fri Mar 23 13:13:54 EDT 2007:   I've designed a new HindIII adaptor. Using this one, I might not need to do the initial end-repair reaction (see new design below).
HindIII concatenation degenerate adaptor
5' AGCTTGCGAGCGNN
       ACGCTCGC-p 5'

6.12.2  Preparing cDNA for the concatenation approach

Growing cells   Mon Feb 19, 2007
I grew cells from a 1:100 dilution of overnite E. coli MG1655 culture in LB. Three samples were grown in a baffled flask with 20 ml of LB, and three additional samples were in 20 ml LB + 75 ng/ml norfloxacin antibiotic (Figure ).
I put 2.5 ml of OD 600   0.5 in 5 ml of RNAprotect [Qiagen], vortexed 5 sec, incubated 5 min, and spun down 12 min at 4000 rpm. The pellets were put in the -20C.
Please see the pdf version for figures
Figure 6.49: Cells were grown after 1:100 dilution from an overnight culture. Samples 1-3 were grown in 20 ml LB, samples
Brief Conclusions:   No problems so far, as expected the growth of the cells in norfloxacin started to slow after 0.5 OD600 (Figure 6.49).
making ds cDNA from RNAprotect pellet   Tue Feb 19, 2007
cDNA was made from samples 1 (LB) and 4 (LB + nor) according to the protocol in section on page .
These are the spec readings for SAMPLE POINT A (total RNA + contaminating genomic DNA: 100 ml ):
Sample DNA (ng/ul) 260/280 260/230 total yield
1 587.0 58.7 mg
4 1275.2 127.5 mg
These are the spec readings for SAMPLE POINT B (total RNA without contaminating genomic DNA: appx 35 ml ):
Sample DNA (ng/ul) 260/280 260/230 total yield
1 1559.1 54.6 mg
4 1704.3 60.0 mg
Used appx 6 ml (10 mg ) of each sample for the MICROBExpress mRNA kit.
These are the spec readings for SAMPLE POINT C (mRNA appx 16 ml ):
Sample DNA (ng/ul) 260/280 260/230 total yield
1 171.6 2.7 mg
4 222.2 3.6 mg
Please see the pdf version for figures
Figure 6.50: RNA was extracted with an RNAeasy kit (a), treated with LiCl and TURBO DNA-free [Ambion] (b), and run through the MICROBExpress kit (c)
Brief Conclusions:  
The RNA samples look pretty good (Figure ). 1a, 4a, 2b, 4b are just what I expect to see. 1c and 4c, post-rRNA removal is still a little disappointing. Once again, the 16S rRNA is removed well, but the 23S is not. I think that this time 23S is a little better removed than last time (Figure 6.12) where the 23S band was even stronger. The only difference this time is that I used water-baths for the MICROBExpress kit rather than heat blocks.
These are the spec readings for SAMPLE POINT D (double stranded, end-repaired cDNA: appx 30 ml ).
Sample DNA (ng/ul) 260/280 260/230 total yield
1 187.8 5.6 mg
4 277.5 8.3 mg
Brief Conclusions:   The 260/280 and 230/260 values were really high here????
size selecting adaptored-cDNA   Feb 22, 2007
I ran the adaptoed cDNA on a low-melt agaroase gel (Figure ). And cut the bands for 500-1500 and > 1500 for both cDNA sample 1 and cDNA sample 4.
agarase preparation of cDNA   Feb 23, 2007
I digested the 4 gel slices with b-agarase [NEB] using the NEB protocol. I think there was too much agarose, because it really didn't seem to digest too well. There was a lot of gel in the bottom after spinning down. I should call NEB and see if this is normal. I cleaned the agarases digestions with isopropanol. Then I eluted into 15 ml TE.
Please see the pdf version for figures
Figure 6.51: Thur Mar 22, 2007. cDNA samples 1 and 4 on a 1% SeqPlaque lowMT gel and two lanes of 1kb ladder. I know, pretty much all you can see are the adaptors. Maybe next time I'll use the Qiagen PCR kit to help remove them before running the gel.
circularize size-selected cDNA   Feb 23, 2007
I used 5 ml of the cDNA (1/3), 120 ng of the circularizer DNA (pre-annealed in STE), and 1 ml ligation buffer in a 10 ml ligation at 16C for 2hr. I did 2 reactions with cDNA sample 1 big (i.e. > 1500bp), I used circularizer 2 (which creates a 77mer). With cDNA sample 4 big, I used circularizer 1 (the original, which creates an 83mer).
RCA the circularized cDNA   Sat Mar 24, 2007
5.25 ml 10x buffer, 2 ml dNTP, 5 ml template DNA, 35.25 H2O , 2.5 ml hexamer. heat to 95C for 5 min, cool to 4C, transfer to ice. add 2.5 ml f29 polymerase, incubate at 30C for 12 hr, heat deactivate 10 min at 65C.
Mon Mar 26, 2007
I EtOH precipitated the RCA and resuspended the DNA into 200 ml of TE. The yields were (number each cDNA sample, b = big = > 1500bp):
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA 1b 359.8 72 mg
RCA 4b 340.1 68 mg
Brief Conclusions:   Mon Mar 26, 2007 The yields above are about half of what I expected based on the previous result (see section 6.6.2.1 on page pageref). However, I realize now that I forgot to do the exonuclease reaction to remove the non-circular DNA. 60 ml of each of the above RCA = 20 mg , which is what I want for my MmeI digestions.
MmeI digestion of the RCA'd circularized cDNA   Mon Mar 26, 2007
I ran the following digestion: 60 ml (20 mg ) RCA, 10 ml MmeI, 4 ml SAM (diluted 1/20), 10 ml NEB4, 16 ml H2O at 37C for 15 min and 10 min at 65C to deactivate the MmeI. I EtOH precipated the rxn and eluted into 15 ml TE.
Nusieve of MmeI digested RCA   Tues Mar 27, 2007
I ran the MmeI digested 20 mg of RCA'd cDNA for samples 1b and 4b on to a 3.5% TBE gel (Figure ).
Please see the pdf version for figures
Figure 6.52: 3.5% Nusieve gel with sybr safe run at 80V for 90 minutes. Gel shows the MmeI digested RCA products from two cDNA samples.
Brief Conclusions:   Although it's hard to see in Figure 6.52, the two barcodes were faintly visible and the correct sizes (i.e. the new one looked appx 5bp shorter).
RCA the circularized cDNA   Tues Mar 27, 2007
In case it mattered, I added the Exonuclease I and the Exonuclease III to the circularized cDNA to remove non-circular DNA. The protocol was: add 2 ml Exonuclease I and 0.4 ml ExonucleaseIII, incubated 45 min at 37C, heat-deactivate 80C for 20 minutes.
After removing the linear DNA, I make the following RCA rxn: 5.25 ml 10x RCA buffer, 4 ml dNTP, 5 ml template, 33.25 H2O , 2.5 ml hexamer; incubated 95C for 5 min, to ice, added 2.5 ml f29, 30C for 12 hr, 10 min 65C to deactivate.
Wed Mar 28, 2007
After an EtOH, spec'd the two RCA reactions to see the following yields:
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA 1b 2722.2 544 mg
RCA 4b 2670.4 534 mg
Brief Conclusions:   Wow, much better yield than the last RCA. Almost 10x higher yield. I changed two parameters from last time: 1) I doubled the amount of dNTP, 2) I removed the linear DNA. You wouldn't think that doubling dNTP would give an order of magnitude more yield, so I guess digesting the linear DNA away is really important.

6.12.3  Ligating tags into circles for RCA amplification

I want to be able to ligate the PETs into a circle that is > = 500 bp. The circle provides an easy way to amplify concatenated tags, since I think it is unlikely to be able to meet the 454 pyrosequencing requirement of 5 mg of clean DNA. So the protocol will go: 1) make tags, 2) concatenate into big circles, 3) amplify into long, linear RCA DNA, 4) send to 454 for shearing and sequencing. The downside of this requirement is that you end up with partial PET sequences for the start of your sequencing read (depending on where the shearing breaks your DNA).
The persistence length of DNA is around 150 bp under normal salt conditions. This means the circles are unlikely to be less than 150 bp in length, but we want to push that towards larger circles. The basic approach comes from some suggestions from Ravi Sachidanandam (my old boss at CSHL and a former physicist that gets a kick outta this kinda thing).
Ravi's rules for biasing towards larger circles  
The overall strategy of make larger circles it to create conditions where each PET end finds another PET end more often than it finds its own end (self-end).

tests on fake tags (coming from amplified pUC19 with HindIII overhangs)

I'm using a 77mer and an 83mer (the current sizes of the tags I'm using) to try and approximate the real results of ligating the tags without having to go through the long process of making and gel-selecting the tags each time. I'm using 2 tags of each size amplified from regions of pUC19 that 1) do not have a HindIII site (since HindIII is the adaptor sequence) and 2) are not in regions that might have strong secondary structure (e.g. promoters and transcription terminators). The features of pUC19 are below:
Features:
  469- 146      lacZ alpha CDS (start 469, complementary strand)
  519- 514      Plac promoter -10 sequence (TATGTT)
  543- 538      Plac promoter -35 sequence (TTTACA)
  575- 563      CAP protein binding site
  396- 452      multiple cloning site (EcoRI-HindIII)
 1455- 867      origin of replication (counterclockwise)
                  (RNAII -35 to RNA/DNA switch point):
 1273-1278        RNAI transcript promoter -35 sequence (TTGAAG)
 1295-1300        RNAI transcript promoter -10 sequence (GCTACA)
 1309-1416        RNAI transcript
 1419- 867        RNAII transcript (complementary strand)
 1434-1429        RNAII transcript promoter -10 sequence (CGTAAT)
 1455-1450        RNAII transcript promoter -35 sequence (TTGAGA)
 2486-1626      beta-lactamase (bla; amp-r) CDS
                  (start 2486, complementary strand)
 2486-2418      beta-lactamase signal peptide CDS
                  (start 2486, complementary strand)
      2521      bla RNA transcript start (complementary strand)
 2535-2530      bla promoter -10 sequence (GAGACA)
 2556-2551      bla promoter -35 sequence (TTCAAA)

 1 LEFT PRIMER       2046   20   60.21   60.00  4.00  0.00 GGTTAGCTCCTTCGGTCCTC
   RIGHT PRIMER      2128   20   59.72   50.00  8.00  1.00 TATGCAGTGCTGCCATAACC
   PRODUCT SIZE: 83, PAIR ANY COMPL: 5.00, PAIR 3' COMPL: 1.00

 2 LEFT PRIMER       1878   20   59.97   50.00  6.00  1.00 TTGCCGGGAAGCTAGAGTAA
   RIGHT PRIMER      1960   20   60.60   55.00  5.00  2.00 GTGACACCACGATGCCTGTA
   PRODUCT SIZE: 83, PAIR ANY COMPL: 3.00, PAIR 3' COMPL: 2.00

 3 LEFT PRIMER       1878   20   59.97   50.00  6.00  1.00 TTGCCGGGAAGCTAGAGTAA
   RIGHT PRIMER      1959   20   60.74   55.00  3.00  2.00 TGACACCACGATGCCTGTAG
   PRODUCT SIZE: 82, PAIR ANY COMPL: 4.00, PAIR 3' COMPL: 3.00

 4 LEFT PRIMER       1216   20   60.14   55.00  6.00  2.00 GCAGCCACTGGTAACAGGAT
   RIGHT PRIMER      1298   20   60.65   55.00  4.00  0.00 TAGCCGTAGTTAGGCCACCA
   PRODUCT SIZE: 83, PAIR ANY COMPL: 4.00, PAIR 3' COMPL: 1.00

ORDERED 1 and 2 (I refer to these as 83.1 and 83.2 in my experiments)

AND a 77 bp product almost exactly like 1
   LEFT PRIMER       2048   18   55.43   55.56  4.00  0.00 TTAGCTCCTTCGGTCCTC
   RIGHT PRIMER      2124   20   60.69   50.00  4.00  2.00 CAGTGCTGCCATAACCATGA
   PRODUCT SIZE: 77, PAIR ANY COMPL: 3.00, PAIR 3' COMPL: 2.00

AND a 77 bp product almost exactly like 2
   LEFT PRIMER       1879   20   59.61   55.00  4.00  0.00 TGCCGGGAAGCTAGAGTAAG
   RIGHT PRIMER      1955   19   60.28   52.63  3.00  2.00 ACCACGATGCCTGTAGCAA
   PRODUCT SIZE: 77, PAIR ANY COMPL: 5.00, PAIR 3' COMPL: 1.00

ORDERED 1 and 2 (I refer to these as 77.1 and 77.2 in my experiments)

add AATTAAGCTT to LEFT
add ATATAAGCTT to RIGHT

Final sequences (minus the HindIII adaptor) are:

77.1
TTAGCTCCTTCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTTATCACTCATGGTTATGGCAGCACTG

77.2
TGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCCATTGCTACAGGCATCGTGGT


83.1
GGTTAGCTCCTTCGGTCCTCCGATCGTTGTCAGAAGTAAGTTGGCCGCAGTGTTATCACTCATGGTTATGGCAGCACTGCATA 

83.2
TTGCCGGGAAGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTGCCATTGCTACAGGCATCGTGGTGTCAC


The "add" sequences above were added to the 5' end of the left and right primers respectively. They allow the addition of a HindIII site to the 83mer and the 77mer just as we'll have with the real PETs. In addition they add 4 bp just to allow the HindIII to cut. The cutting enzymes will leave the phosphates, so that the cut PCR products can just be ligated directly after cutting (perhaps with an initial clean up).
PCR amplifying the pUC19 based tags  
Thur Mar 8, 2007
I made 200 ml rxns of each of the four primer pair sets. I cleaned them with Qiagen PCR purification kits and eluted into 30 ml . The yields were:
Sample DNA (ng/ul) 260/280 260/230 total yield
77.1 73.8 2.2 mg
77.2 76.4 2.3 mg
83.1 78.5 2.4 mg
83.2 72.7 2.2 mg
Brief Conclusions:   Pretty low yields considering I used 200 ml for my PCRs (no in one tube, that's 2 x 100 ml rxns. I don't know if the column is not catching the short DNA well or if the short DNA PCR just doesn't result in as much product (e.g. if PCR give you a certain number of molecules than a 200mer product should yield 1/10th the amount of DNA [by weight] of a 2000bp piece).
83.2 tag digestion try 1  
Sat Mar 10, 2007
I digested all 30 ml of the PCR product for tag 83.2 using the following rxn conditions: 30 ml cleaned PCR, 4 ml NEBuffer2, 0.5 ml HindIII, 5.5 ml H2O . 37C for 30 min, followed by heat deactivation. The rxn was cleaned with a Qiagen PCR purification kit with the following yield (into 34 ml of EB):
Sample DNA (ng/ul) 260/280 260/230 total yield
83.2 36.8 1.3 mg
83.2 tag ligation try 1  
Sat Mar 10, 2007
I ran three ligation reactions just to try things out the rxns were all 20 ml total volume with 10 ml of DNA (368 ng) and 1 ml of T4 ligase. All rxns were at 16C with a final heat activation at 65C for 10 minutes.
The differences for each were:
  1. rxn 1: ligation for 15 minutes with 100% T4 buffer
  2. rxn 2: ligation for 60 minutes with 100% T4 buffer
  3. rxn 3: ligation for 60 minutes with 1/4 diluted T4 buffer
Wed Mar 14, 2007
I ran the three ligations on a 2% TAE gel (Figure ).
Please see the pdf version for figures
Figure 6.53: 83.2 was amplified from pUC19, cut with HindIII, and ligated in 3 different conditions. The resulting concatentations are multiples of the original length of around 93 bp (83 bp tag + 10 bp for the HindIII site on each end).
Brief Conclusions:   First concatenation doesn't look too bad (see Figure 6.53). Looks like the DNA is all linear (just a guess, but I figure if it were non-linear the band lengths wouldn't be just multiples of the original 93 mer, since circular DNA migrates differently). The only worrying thing is that the 83.2 unligated band looks a little bit too long (it appears longer than the 100 bp band in the ladder, but it should be 7bp or so shorter). Also, in the next attempt I definitely need to ligate longer, and hopefully I'll have enough DNA to try the Exonuclease step as well.
PCR amplifying the pUC19 based tags try 2   Sat Mar 10, 2007
I ran two 500 ml PCRs: (0.5 ml pUC19 = 250 ng), 10 ml primer 250 ml Taq master mix, 239 ml H2O . The reactions were concentrated with EtOH and resuspended into 15 ml TE. The primer pair used was 83.2
cutting/ligating the pUC19 based tags try 2   Mon Mar 26 17:30:08 EDT 2007
I digested the first of the two 500 ml PCR reactions: 15 ml of tag 83.2, 2 ml buffer, 0.5 ml HindIII, and 2.5 ml H2O . Incubated at 37C for 30 min and deactivated the HindIII for 20 min at 65C.
Sample DNA (ng/ul) 260/280 260/230 total yield
9.1 36.8 0.28 mg
Brief Conclusions:   I think the Qiagen columns, poor as they are, might be working better than EtOH with this short mer (maybe use Glyco-blue [Ambion] or switch back to Qiagen).
PCR amplifying the pUC19 based tags try 3   Tue Mar 27, 2007
I ran a 600 ml PCR (300 ml Master mix, 12 ml pUC19 (12 ng), 15 ml primer (250 nM), 273 ml H2O . Ran 35 cycles (first 5 annealed at 54C, last 30 cycles annealed at 59C); I cleaned the entire reaction and ran it through a Qiagen PCR purification kit (I used the vacuum and just kept adding more PBI buffer mixed with PCR product until I was out of PCR product). The yield was:
Sample DNA (ng/ul) 260/280 260/230 total yield
83.1 107.3 3.3 mg
I cut all 3 mg of the PCR product in a 40 ml digestion with 0.5 ml of HindIII for 30 min at 37C and 20 min at 65C. I cleaned up the digestion with a Qiagen PCR clean up kit to get the following yield:
Sample DNA (ng/ul) 260/280 260/230 total yield
83.1 (post cut) 68.8 2.1 mg
Brief Conclusions:   Compared to try1 where I also used the Qiagen cleanup, the yields were higher. But in try1, I used only 200 ml for the PCR! Maybe I wrote that down wrong? Tomorrow, I'm going to try 400 ml and 1 ml PCR. I could be that somehow these short DNA fragments are exhausting the Qiagen column? In which case, I'll have to pick the optimal PCR volumn and just run a bunch of them.
ligating the 83.1 try 3   Tue Mar 27 19:40:00 EDT 2007
I ran 3 ligations (similar to try 1, but with more DNA, longer incubations, and more T4 ligase):
  1. rxn 1: ligation for 2 hours with 100% T4 buffer and 400 U T4 ligase
  2. rxn 2: ligation for 12 hours with 100% T4 buffer and 400 U T4 ligase
  3. rxn 3: ligation for 12 hours with 100% T4 buffer and 800 U T4 ligase
Wed Mar 28, 2007
All 20 ml of the ligations were run on a 2% TAE gel for 50 minutes (Figure ).
Please see the pdf version for figures
Figure 6.54: 83.1 was amplified from pUC19, cut with HindIII, and ligated in 3 different conditions. The resulting concatentations are multiples of the original length of around 93 bp (83 bp tag + 10 bp for the HindIII site on each end).
Brief Conclusions:   The 2hr incubation time worked much better than the 1hr and the 15 minute incubation times (compare previous result in Figure 6.53 with the current result in Figure 6.54) The strongest band is still the PCR tag. I'm surprised the 2x ligase didn't help remove more of that band. Next time I want to try the high concentrate ligase that we have. Also, it is possible (unlikely?) that the things are circularizing at that size already? Maybe that's why I can't get rid of that band? I don't know. I'll try high conc ligase AND exonuclease (but maybe not together) on the next round.
PCR amplifying the pUC19 based tags try 4   Wed Mar 28 18:19:32 EDT 2007
I ran 2x 1000 ml (1-2) and 2x 500 ml (3-4) PCR reactions just as in try 3 to try and figure out where I'm losing DNA.
Sample DNA (ng/ul) 260/280 260/230 total yield
83.1 (1) 40.7 1.2 mg
83.1 (2) 60.6 1.8 mg
83.1 (3) 24.3 0.73 mg
83.1 (4) 25.4 0.76 mg
Brief Conclusions:   This sucks. Come on, how hard can it be to purify this little piece. I might try using EtOH again, but with Glycoblue. Alternatively, I might make a 160 mer with a HindIII site in the middle, then I'd make a double tag and cut it (3x: once on each end and once in the middle).
cutting and ligating PCR try 4   Fri Mar 30 15:22:17 EDT 2007
Based off the results in Figure , we need to work harder to coerce these ligations into circles. In this experiment, I'm going to continue with the 2 hr ligation. After those 2 hrs, I'm going to dilute the ligation 1:10 (to make my 40 ml rxn 400 ml ). This lower concentration should help push things towards circles rather than ligating with each other. I'm also going to add more ligase at the same time in case the initial ligase is exhausted. Last, I'm going to try this using normal and high concentration T4 ligase (400U and 2000U).
I digested the two 1ml PCR reactions from try 4 (81.1 (1) and 81.1 (2)) in separate 40 ml reactions with 0.5 ml of HindIII at 37C for 45 minutes followed by at 20 min heat deactivation at 65C. The two digestions were cleaned with a Qiagen PCR clean up kit and eluted into 30 ml of EB buffer. 4 ml of T4 ligase buffer was added to both elutions for a 40 ml ligation. 81.1 (1) used 400 U of T4 ligase (1 ml of normal concentration) and 81.1 (2) used 2000 U of T4 ligase (1 ml of high concentration). The reactions were run at 16C for 2 hrs and the ligase was NOT deactivated. Then I diluted the ligations 1:10 in H2O + T4 ligase buffer (this created one 400 ml reaction for each ligase concentration and required breaking the reactions into two PCR tubes. To each 400 ml ligation reaction one additional microliter of ligase was added (normal conc for 81.1 (1) and high conc for 81.1 (2)). The ligations were run for 2 hr at 25C (hopefully the higher temperature will aid circularization). The ligase was then heat deactivate at 65C for 10 minutes.
Sat Mar 31 16:48:12 EDT 2007
The two 400 ml ligation reactions were concentrated with EtOH precipitation and 1 ml of glycoblue. The tubes were placed at -20C for 20 minutes and then at -85C for 20 minutes. They were spun for 20 minutes at 4C followed by washing with 750 ml 70% EtOH. The pellet was dried for 5 minutes and resuspended in 30 ml of TE.
To one-half of the two ligation reactions (and to 250 ng of a pUC19 control), I added 2 ml Exonuclease I and 0.4 ml Exonuclease III. I added 1.8 ml of NEBuffer1. The 3 rxns were digested for 30 min at 37C followed by 20 min at 80C. All 6 reactions (including the two pUC19 controls) were run on a 2% gel .
Please see the pdf version for figures
Figure 6.55: 2% gel run for 50 minutes.
Brief Conclusions:   I think the problem is actually the opposite of what I thought it was. We can't get the damn DNA to circularize. All it will do is concatenate into longer linear fragments (see attached image). So I guess now I just try the opposite of all the things we were doing to push towards linear: 1) lower the concentration of DNA (that's what I tried in the attached picture. I lowered it 1:10, apparently that's not enough. Unfortunately, 1:18 is about as low as I can go. 2) increase temperature this one I also max'ed out on the attached gel (I increased T by 9C); I might be able to switch to the Taq ligase (like the Taq polymerase it works at really high T), but won't this also increase diffusion and make the linearization more likely? 3) increase salt, this might be possible, I'll have to check 4) decrease viscosity, not really possible, I never added anything viscous
PCR amplifying the pUC19 based tags try 5   Thu Mar 29, 2007
Ran 2 x 500 ml PCRs and 2 x 200 ml PCRs (using same conc of puc19 and primer as before). I cleaned them up with ethanol precipitation using glycoblue as a carrier. Hopefully, this carrier increases my yield (which with my last EtOH of this short piece was appx 0 ng/ml ). With one of the duplicates I used 1 ml of carrier and with the other I used 5 ml . I let the reaction stay at -20C for 30 min and at -85C for 30 min. I spun at 4C for 30 min, washed with 750 ml of 70% EtOH, spun 10 more min at 4C. Dryied the pellet for 5 minutes and resuspended in 30 ml of TE. I also made two blank of 30 ml TE with 1 ml and 5 ml of glycoblue respectively. The yields where:
Sample DNA (ng/ul) 260/280 260/230 gel estimate
blank (1 ml glyco) 6.6 0
blank (5 ml glyco) 41.3 0
83.1 (500 ml ; 1 ml glyco) 432.2 56.4
83.1 (500 ml ; 5 ml glyco) 595.0 66.8
83.1 (200 ml ; 1 ml glyco) 386.9 24.5
83.1 (200 ml ; 5 ml glyco) 539.4 30.6
The values looked fishy, so I ran 1 ml of each reaction on a gel (I ran the blanks too, but as expected they were not visible on the gel) for gel based quantification. The software estimated values for this quantification are shown last column in the table above.
Please see the pdf version for figures
Figure 6.56: 1 ml of each EtOH/glycoblue precipitated PCR rxn was run on a gel for quantification (since the spec seemed inaccurate, perhaps due to dNTPs and primers).
Brief Conclusions:   The gel-based quantification was definitely better than the spec. However, I think the gel-based values are pretty large underestimates, because the bands get fuzzy at the end. I'm going to switch to the NEB 2-log ladder, which provides bands with a larger dynamic range that should help quantification.
cutting and ligating the pUC19 based tags try 5   Apr 4, 2007
I cut 83.1 (500 ml ; 1 ml glyco) with HindIII. I cleaned up the digestion with a Qiagen PCR cleanup kit and got:
Sample DNA (ng/ul) 260/280 260/230 yield mg
cut 83.1 (500 ml ; 1 ml glyco) 46.7 14.0
I'm divided the digested tags into 3 groups each using a different two stage ligatino procedure
  1. (stage 1) 10 ml cut DNA, 2 ml T4 buffer 8 ml H2O , 1 ml T4 16C for 2 hr; (stage 2) add 179 ml H2O + T4 buffer, 1 ml T4 ligase 16C for 2 hr
  2. (stage 1) 10 ml cut DNA, 2 ml T4 buffer 8 ml H2O , 1 ml T4 16C for 2 hr; (stage 2) add 179 ml H2O + Taq ligase buffer, 1 ml Taq ligase 45C for 2 hr
  3. (stage 1) 10 ml cut DNA, 2 ml Taq ligase buffer 8 ml H2O , 1 ml Taq ligase 45C for 2 hr; (stage 2) add 179 ml H2O + Taq ligase buffer, 1 ml Taq ligase 45C for 2 hr
I digested half of the each of the 3 reactions with exonuclease I and III. I ran them all on a 2% agarose gel (Figure ). I cut out the bands at around 500 bp from the non-exonuclease gel.
Please see the pdf version for figures
Figure 6.57: 2% gel run for 45 minutes.
Apr 5, 2007
I cleaned two gel slices (lanes 1 and 2 in Figure 6.57) using a Qiagen column-based gel extraction kit and a QiaexII gel extraction kit (to see which method was better).
I then made the assumption that if I added ligase, at least a few of the cut pieces would ligate into a circle. Allowing me to amplify my tags even if I wasn't able to see the circles on a gel (because so far they are way to few relative to the uncircularized pieces). I used the entire gel clean up reactions in a 40 ml ligation with 4 ml T4 buffer, 1 ml T4 ligase, and 5 ml H2O (I did one of these reactions for each of the two samples). I incubated 12hr at 16C and heat deactivated at 65C for 10 minutes.
I took the (hopefully) circularized ligation and digested the linear fragments using Exo I and Exo III. I used 5 ml of this in an RCA reaction to amplify the circles.
Sat Apr 7, 2007
I did an EtOH precipitation of the RCA.
Sun Apr 8, 2008
I spec'd the RCAs:
Sample DNA (ng/ul) 260/280 260/230 yield (mg )
83.1 Qiaex 2388.9 119.4
83.1 Qiagen column 2694.2 134.7
I sheared 4 ml of each RCA (appx 10 mg ) in 125 ml TE at 10% power. I cleaned up the reaction with a Qiagen PCR cleanup:
Sample DNA (ng/ul) 260/280 260/230 yield (mg )
83.1 Qiaex sheared 71.4 2.42
83.1 Qiagen column sheared 20.3 609
I ligated the sheared DNA with 50 ng of blunt (SmaI) cut, dephosphorylated puc19, using 1 ml of concentrated T4 for 10 min at RT, and then heat deactiviated at 65C for 10 min. I cloned this into DH5a competent cells.
Mon Dec 9, 2007
To make sure the RCA and ligation worked (well I know the RCA amplified something because I got a ton of DNA, I just don't know if it amplified the correct thing), I digested 10 mg of the RCA with HindIII. 4 ml RCA DNA, 2 ml NEB2, 13 ml H2O , 1 ml HindIII. I ran the digestion for 45 min at 37C followed by deactivation at 65C for 10 minutes. I ran all of the digestion reaction onto a 2% agarose gel.
Please see the pdf version for figures
Figure 6.58: 2% gel run for 45 minutes. Unfortunately, I don't see the expected band at 83 bp.
Brief Conclusions:   I doesn't look like this worked too well. For sure there was circular DNA in the RCA reaction because I get a huge amount of amplification (perhaps some sorta contaminating circular DNA?). However, when I gut that amplified DNA with HindIII, I should be back to my 83mer. And that doesn't seem to be the case (Figure 6.58). Esther had better luck when she tried the same thing using a one stage ligation and no gel extraction (but there we still don't know if the circles were big enough. perhaps we can size select the circles using the ChargeSwitch PCR cleanup kit?). I'm also considering using a PCR-plug like is used in LM-PCR. I can just something like:
5' p-AGCTGCTAGC
         CGATCG

www.flychip.org.uk/protocols/chip/lm_pcr.pdf
http://nar.oxfordjournals.org/cgi/content/full/27/18/e23

In excess, the plug would halt the ligation/concatenation (by blunting the ends with an unphosphorylated primer). And it could be used together with the HindIII site to amplify the tags via PCR rather than rolling circle.

BEGIN WORK BY Esther Rheinbay
Mar 29, 2007
Exonuclease digestion of linear and circular DNA: do we have ligated circles?  
Esther just did a 50 ml PCR to get started on this project of trying to create definable sized circular DNA from these 83mers and 77mers. As an initial check to see if the smaller DNA was forming circles during the ligation (and to make sure that the exonuclease really does remove the non-circular DNA), she digested known circular DNA (pUC19) and the ligated 83.2 fragments. She ran the resulting four lanes (cut/uncut for both) on a 2% gel for 45 min (Figure ).
She used 2 ml Exonuclease I and 0.4 ml Exonuclease III, incubated at 37C for 45 min and heat deactivated at 80C for 20 minutes. 100 ng of puc19 DNA was used in the control.
Please see the pdf version for figures
Figure 6.59: 2% gel run for 45 minutes. The perfect example of circular DNA (lanes 1 and 2) surviving Exonuclease I and III digestion (lane 2) while linear DNA (lanes 3 and 4) gets completely digested (lane 4).
Brief Conclusions:   A few nice things can be concluded from this gel. 1) the exonuclease trick definitely works as the circular plasmid was just slightly digested (presumably due to nicks and other DNA breaks) while the linear DNA was completely removed. 2) we certainly can get linear DNA higher than the 150bp persistance length of DNA. The problem might even be how to we get the ligated pieces to form circles at all. At least initially, we can try the opposite of the tricks to bias away from circles. So after the initial 2 hr ligation, we can dilute a lot (1:10? or maybe 1:100?), but keep the salt conc high and incubate at higher temp (25C). We'll also need to add more ligase, because it seems to be exhausted after that initial 2hrs.
Esther also ran the following (similar to my try 5 but worked better): digest 83.2 with HindIII, ligate 2 hr with T4 ligase. Digest with Exonuclease I and III to remove all non-circular DNA (she actually skipped this step?). RCA. Digest RCA with HindIII. She did this for two reactions, one worked and the other failed (Figure ). The unfortunate thing is that she forgot to do the Exonuclease digestion. Would be much more confident in the result if she'd done that.
Please see the pdf version for figures
Figure 6.60: 2% gel run for 45 minutes. Seems lane 1 worked and lane 2 didn't.
END WORK BY Esther Rheinbay

6.12.4  amplification of tags with LM-PCR

Wed May 2, 2007
Previously I've been ligating the tags into a circle for amplification (after which they could be sheared and sequenced). I wanted to try Ligation Mediated PCR (LM-PCR) as an alternative. The plan is to do the initial ligation for 2hr, then to add a blunt-dephosphorylated plug
5' p-AGCTGCTAGC
         CGATCG

As I mentioned a few sections before, in excess, the plug would halt the ligation/concatenation (by blunting the ends with an unphosphorylated primer). And it could be used together with the HindIII site to amplify the tags via PCR rather than rolling circle.
Wed May 2, 2007
I amplified 77.1, 77.2, 83.1, and 83.2. I also ran a separate reaction with the 83.1 only, so I could look for discrete bands whereas mixing the two sizes might make things a little smeary. I did a 2 hr ligation with 1 ml of T4 at 16C. I then added 500 ng (1 ml in STE buffer) of annealed plug plus 1 ml of T4 buffer and 1 ml of T4 ligase and I incubated another 2 hr at 16C followed by deactivation at 65C.
Thur May 3, 2007
I ran a PCR using the single primer (1 ml of 10 mM ). I used 5 ml ligation product and annealed at 60C. The PCR was for 30 cycles. Unfortunately, yesterday I forgot to make a control with no plugs. I ran the ligation and the LM-PCR from the ligation onto an agarose gel (Figure ).
Please see the pdf version for figures
Figure 6.61: 1.5% gel run for 45 minutes.
Brief Conclusions:   Looks way to smeary (Figure 6.61). I don't think that was succesful. The circular method seems much better for the moment.

6.13  Further removal of rRNA

Tues Jul 3, 2007
I want to have fewer rRNA samples in my sequencing results. Without using MicrobeExpress, the sequences were virtually 100% rRNA. After one round of MicrobExpress I sequenced a couple of real genes for the first time. Still 80% of the reads were 23S (though 16S seemed to be be largely removed on the gel and from the sequencing results; see section 6.5.0.1 on page pageref for more details). I want to try and remove more of the rRNA. Preferably, it would be completely removed or into the single-digit percentile of my sequencing reads.

6.13.1  Further removal of rRNA: strategy

Going from 100% to 80% was a drastic reduction. I think the move from 80% down will be easier. The plan is to run two (or more) samples through the MicrobExpress kit. Then pool the two and run them through a second time. Presumably this second run will contain much less rRNA so the sample loss will likewise be much less.

6.13.2  Further removal of rRNA: first try

Mon Jul 2, 2007
I'm growing up 6 samples in LB. The samples were grown in 5 ml of LB shaking at 300 rpm from a 1:100 dilution. Cells were grown to an OD600 of around 0.6.
Tues Jul 3, 2007
I followed the protocol Preparation of PET libraries in the appendix with the following modications: I put 1 ml ready lyse in 500 ml TE. I used 100 ml of this for the lysis rather than measuring the dry lysozyme. I didn't take a sample at SAMPLE POINT A, but I did after SAMPLE POINT B (total volume was 35 ml here):
Sample DNA (ng/ul) 260/280 260/230 total yield
sample A 1770.0 62.0 mg
sample B 1225.0 43.0 mg
At SAMPLE POINT B, I also saved 1 ml of each sample to run on a gel (see the first two lanes in Figure ).
After SAMPLE POINT B, I took 3 samples of 10 mg from sample A and sample B (6 samples total). One of the three samples was run through the MICROBExpress kit normally. The other two were run through as well, but before the final EtOH precipitation, I pooled the two into a single tube and ran an Isopropanol precipitation rather than an EtOH precipitation. I then took the pooled samples and ran them through again the rRNA removal step a second time. I eluted all of the samples into 16 ml TE [Ambion]. I also measured the RNA concentrations using the Qubit, which is probably more accurate, since it should be less influenced by salts and other junk that have made their way into my RNA over these many steps.
Sample RNA (ng/ml ) Qubit (ng/ml )260/280 260/230 yield
A 105.0 74.6 1.68 mg
A.next (pooled 1st enrichment) 111.9 - 1.79 mg
B 103.4 62.2 1.65 mg
B.next (pooled 1st enrichment) 119.2 - 1.91 mg
A.2x (pooled 2nd enrichment) 92.6 67.0 1.48 mg
B.2x (pooled 2nd enrichment) 76.2 54.8 1.22 mg
I saved 2.5 ml of A, B, A.2x, and B.2x for a gel (see the last four lanes of Figure ).
Please see the pdf version for figures
Figure 6.62: total RNA was run through the MICROBExpress kit one time and two times (2x) to remove the rRNA. this is a 1.5% TAE agarose gel
Brief Conclusions:   From the gel (Figure 6.62), it looks like the rRNA was more depleted this time than in an earlier attempt (Figure 6.12). Actually even only doing 1x rRNA removal looks better this time, but there is a noticable amount of 23S left. After 2x the 23S on the gel looks completely eliminated. It's hard to tell about the 16S, since there might mRNA at that length too. 16S seemed to be pretty much removed last time I used the MICROBExpress kit. We'll know better after sequencing.
Wed Jul 4, 2007
For the 1st strand synthesis, I used 11 ml of each of the four samples (A, B, A.2x, B.2x)+ 1.5 ml of SuperScript II (I ran out of SuperScript III, which is why I used SuperScript II). Unfortunately, I screwed up and didn't follow the superscript manual exactly. I added the buffer before heating to 65C rather than after. I followed the SuperScript II instructions not the Superscript III instructions (e.g. I incubated my random hexamers at RT for 10 minutes rather than 5).
After the 2nd strand synthesis and end-repair, I cleaned up with a Qiagen PCR purification rather than phenol:chloroform. Since I know have the Qubit for measuring low DNA quantities (and also importantly, quantifying DNA in the presence of RNA), I quantified the amount of cDNA using the HS dsDNA reagent:
Sample Amount (ng/ml ) yield (ng)
A 17.12 513.6
B 11.48 344.4
A.2x 14.74 442.2
B.2x 14.54 436.2
Because, I could quantify the DNA, I realized I've been using too much adaptor for the adaptor ligation. This time I used only 1 ml of BamISH adaptor (2.1 mg ) rather the previous 2 ml .
Thur Jul 5, 2007
I ran the end-repaired cDNA on a TAE Sybr Safe 1% gel for 20 minutes for size-selection. I purposely ran it this short amount of time, so I wouldn't have to cut to large a chunk of gel to cleanup. I cut from > =300bp for each of the four samples. I cleaned up the samples using a Qiagen Gel cleanup column, eluting into 30 ml . I didn't quantify the DNA, instead I assumed (based on some experiments I did right before this experiment) that the loss from the kit would be about 50%.
Fri Jul 6, 2007
Circularization
I circularized all for cDNA samples (a, a.2x, b, b.2x) using 10 ml of the 30 ml cDNA from the gel extraction above. I used 0.5 ml of circularizer DNA (50 ng), 2 ml T4 ligase buffer, 1 ml T4 ligase, and 6.5 ml H2O . I used the original dsDNA circularize for samples a and a.2x. I used circularizer 2 for samples b and b.2x. The ligation was for 2hr at 16C, deactivated for 10 min at 65C.
remove linear DNA
I removed the linear DNA with 2 ml of Exonuclease I and 0.4 ml of Exonuclease III for 45 min at 37 C. The reaction was heat deactivated at 80 C for 20 minutes. I did NOT clean up the reaction.
RCA
I ran the following RCA rxn to amplify my circularized cDNA: 5.25 ml 10x RCA buffer [epicenter], 4 ml dNTP, 5 ml template (i.e. uncleaned circular DNA from above), 333.25 ml H2O , 2.5 ml RCA hexamer. I heated it to 95C for 5 minutes, cooled it back to RT and added 2.5 ml f29 polymerase [epicenter]. The rxn was incubated at 30C for 12hr followed by 10 min of deactivation at 65C.
Sun Jul 8, 2007
EtOH the RCA
I EtOH precipitated the 4 RCA rxns; I didn't do a phenol:chloroform extraction. I resuspended the nice bulky white DNA pellets in 60 ml of TE. I was trying to get the DNA a little more concentrated than before.
The yields were measured with the nanodrop:
Sample DNA (ng/ul) 260/280 260/230 total yield
sample a 2022.1 121 mg
sample a.2x 2261.1 136 mg
sample b 2160.7 130 mg
sample b.2x 2374.8 142 mg
cut/gel the RCA to make/see PETs
I cut 10 ml (appx 20 mg ) of each RCA sample with 7.5 ml MmeI using 1.25 ml of 1:10 SAM, 2.5 ml of buffer 4, and 3.75 ml of H2O . After starting the rxn, I realized that the glycerol concentration was probably way too high, since over a forth of my rxn was enzyme (and enzymes come in glycerol). A quick chat with Ilaria from my lab and the NEB website, and I had two more verifications that this was a bad idea. Nonetheless I kept going assuming that it would cut at least to some extent.
I ran a 3.5% nusieve TAE sybrsafe gel . I ran at 120V for 1hr. I cut out all four bands for gel extraction (this is why the image doesn't look so nice. I just used the lab's cheap digital camera, so I didn't have to use UV). I cleaned up all four gel slices with a Qiagen column-based gel cleanup kit.
Please see the pdf version for figures
Figure 6.63: cut PETs from a, a.2x, b, b.2x imaged on a transilluminator
Brief Conclusions:   A and A.2x look fine. Looks like something might be funky with B and B.2x. Maybe I just ran the gel too hot and they diffused away? Next time I'll try less voltage. I also need to remember to use a 25 bp ladder next time in addition to the 2-log.
cloning the PETs
Mon Jul 9, 2007
I end-repaired the gel-cleaned PETs [Epicenter] for samples A and A.2x. I cut 2 mg of pUC19 with SmaI for 30 min at RT followed by cleaning up with a Qiagen PCR cleanup kit. I eluted into 30 ml EB. Yield was 41.0 ng /ml , 1.86 (280/260), 2.05 (230/260). I made a 15 ml dephosphorylation rxn with 1.5 ml antarctic phosphatase buffer, 1 ml antarctic phosphatase, 190 ng (4.6 ml ) cut pUC19, and 7.9 ml of H2O . I ran this reaction at 37C for 60 min followed by a 5 min deactivation at 65C. I ligated as follows: 2 ml T4 buffer, 1 ml high concentration ligase, 2 ml dephos vector (appx 25 ng), 10 ml insert, 5 ml H2O . I ran the rxn for 15 min at RT followed by 65C for 10 min. I transformed 2 ml of the ligation into DH5a. I plated 100 ml of each tranformation onto an amp plate with X-gal.
Unfortunately, the plates were filled with blue colonies. I think either the smaI or the antarctic phosphatase aren't working anymore. This also explains why Steve in the lab (and Esther too) couldn't get their blunt clones to work. Both of the enzymes were really old. I bought new ones. And I'll try again.

Improving the yield of digested PET

Tues Jul 10, 2007
I want to mess around a little to try and get more PET from my RCA. Am I using too much MmeI, too little? Too much DNA? Too little?
Unlike last time, I'm going to use a much higher volume rxn to prevent excess glycerol in my rxn. I'm trying four 100 ml digestions. All four contain 10 ml of NEB4 plus 1.6 ml of 1:10 SAM. I used RCA sample A from above (2.02 mg /ml ). All rxns were incubated 1hr at 37C.
digestion RCA (ml ) MmeI (ml ) H2O (ml )
A 5 3.75 79.65
B 5 5.00 78.4
C 5 5.00 73.4
D 10 10 68.4
I cleaned up the reactions with a Qiagen PCR kit and eluted into 30 ml of EB buffer. I added 5 ml of 1:20 fisher dye and loaded them into a 4% TAE 60 ml agarose gel with SybrSafe (Figure ). I also ran one well with NEB 2log ladder + 4 ml of full strength fisher dye to determine the approximate migration times of the three dyes for future reference. I ran the gel for 80 minutes at 110 V, which placed the red dye at the end of the gel. In hindsight, I think 110V was a little hot; go with 100 V next time.
One big problem with this experiment was that my DNA seemed to float/diffuse upwards when I loaded it into the well. This happens sometimes when I have DNA cleaned with Qiagen PCR cleanup kits. What causes this? Residual ethanol? A lack of salt? A lack of EDTA? It is highly annoying to watch your hard-earned sample float away, so I investigated this a little further to hopefully prevent this in the future. See section on page for details.
Please see the pdf version for figures
Figure 6.64: PETs from cuts A, B, C, and D imaged on a transilluminator
Brief Conclusions:   Unfortunately, 5 ml of RCA (10 mg ) doesn't seem like enough to be able to detect a different in intensity between my bands. 10 ml of RCA (20 mg ) was certainly better, and I should use that as my starting point in the future. If anything, I'd say that the lesser quantities of mmeI still cut well.
improving PET digestion yield using more RCA DNA   Wed Jul 11, 2007
I tried two 100 ml rxns with 10 ml (20 mg ) of RCA using 10 ml and 5 ml of MmeI. I also tried one 200 ml rxn using 20 ml RCA and 20 ml enzyme. Digestions were for 1hr at 37C. Instead of using the Qiagen PCR purification kit, I used a microcon YM-30 to concentrate the digestion and hopefully avoid the problem of having my sample diffuse away after I loaded it. Plus with the microcon, you can get lower volumes (10 ml ), which would allow me to use less agarose (cheaper, better picture, easier to gel extract). I ran the gel at 90V until the purple dye was at the very edge to try and get even better separation than before. The gel was junk and I didn't bother taking a picture. Two things could've happened (and I think it was a mixture of the two): 1) the microcon didn't recover a big percentage of my sample; 2) I ran the gel so long that the small DNA started to diffuse too much.
Brief Conclusions:   Next time: try minElute from Qiagen, elute with TE. Add a little salt to the eluted DNA in TE? run gel for less time

6.13.3  further removal of rRNA: 2nd try focus on ESTs

Fri Jul 20 19:58:35 EDT 2007
I'm in a bit of a time crunch trying to prepare some slides for a talk. I don't want to mess with cloning the blunt PETs when I can know straight away if the MICROBExpress 2x trick worked with the easy to sequence ESTs. So I'm going to make a couple new samples and attach a BamHI adaptor instead of a BamISH adaptor. I'll then clone and sequence a few clones to see the frequency of rRNA.

growing the cells

Fri Jul 20, 2007
I grew up two samples labeled e and f from a 1:100 dilution from an overnite that Ilaria made for me (I was still in West Palm Beach, FL). The cells were grown at 300 rpm in LB for appx 3 hr to an OD of 0.692 and 0.645 (background LB subtracted) for samples e and f respectively. 2.5 ml of this culture as added to 5 ml of RNAprotect [Qiagen], vortexed 5 sec, incubated at RT for 5 min, and spun at 4000 rpm in a bucket centrifuge for 15 minutes. The pellets were placed at -20C.

RNA preparation

Sat Jul 21 21:13:33 EDT 2007
RNA was prepared as per my PET protocol in the appendix but with the following modifications. Like the previous RNA prep, I used 1 ml of ReadyLyse [Epicenter] in 500 ml of TE rather than weighing out the power lysozyme. I didn't take a SAMPLE A. I spec'd the RNA at SAMPLE POINT B:
Sample DNA (ng/ul) 260/280 260/230 total yield
sample E 1826.6 63.9 mg
sample F 1676.7 58.5 mg
I took 0.5 ml of each sample to run on a gel.
I used 5.5 ml and 6.05 ml of sample E and sample F respectively (appx 10 mg ) for the MICROBExpress rRNA removal. As in my previous attempt at 2x rRNA removal, I ran 3 rRNA removal rxns for each sample. Unlike the previous time, I did not combine the two samples to be used for a second round of MICROBExpress, as I felt like the isopropanol precipitation caused me to lose some of my sample relative to the EtOH precipitated samples. I eluted the normal samples in 15 ml of TE. I eluted the samples to becombined into 10 ml each (20 ml total combined RNA).
Sample RNA (ng/ml ) Qubit (ng/ml )260/280 260/230 yield
E 123.2 102 1.85 mg
E.next (pooled 1st enrichment) 166.4 134 3.33 mg
F 104.0 88.2 1.56 mg
F.next (pooled 1st enrichment) 150.9 120 3.02 mg
E.2x (pooled 2nd enrichment) 118.2 95.4 1.89 mg
F.2x (pooled 2nd enrichment) 108.9 86.6 1.74 mg
Brief Conclusions:   The yields for the pooled 1st enrichment are much closer to what I'd expect for pooling two samples together. They are much closer to having two times the total yield of the unpooled samples. Compare this table with the one in section 6.13.2 on page pageref.
Sun Jul 22, 2007
I used 11 ml of all of the samples for 1st strand synthesis of cDNA (E, F, E.2x, F.2x), which is approximately 1.5 mg of each. My Superscript III arrived, so I followed the standard protocol this time (incubate 50C 1hr). I used 1.5 ml of Superscript III.
The yields of ds-cDNA prior to adaptoring (as measured by Qubit HsDNA) were:
Sample Qubit (ng/ml ) yield (ng)
E 30.4 912
E.2x 20.8 624
F 27.8 834
F.2x 15.46 463.8
I'm using BamHI adaptor NOT BamISH adaptor, since I plan to clone into pUC19. I performed the blunt adaptor ligations with BamHI adaptor just like I did with the BahISH adaptor in the previous MICROBExpress 2x attempt.
Mon Jul 23, 2007
I ran a gel of the total RNA and mRNA sample (Figure ); everything looks normal enough.
Please see the pdf version for figures
Figure 6.65: Total RNA (e.B, f.B), 1x MICROBExpressB (e.C, f.C, e.1x, f.1x), 2x MICROBExpress (e.2x, f.2x). Note that e.2x and f.2x are derived from e.1x and f.1x (i.e. after running e.1x and f.2x through the kit a second time, I had enough 1x left over to throw some on the gel and compare to the 2x samples).
I also ran two gels to size-select the cDNA prior to cloning. For e and e.2x, I loaded the samples directly onto a wide-comb (6-well) 45 ml 1% gel and ran it 20 min at 90V. For f and f.2x, I first ran the samples through a PCR purification to try and remove some of the adaptor (didn't really help there was still tons of adaptor on the gel). I ran this gel in the exact same way as the e and e.2x gel except I used a 10-well comb and a 65 ml gel. Both gels were SybrSAFE. Next time I should run the gels for 25-30 minute to help increase the separation away from the adaptor; after 20 minutes they're still pretty close.
To Do!!!  It would be nice to know if you can aid in the removal of the short adaptor relative to the longer cDNA by running the PE buffer across the sample multiple times during the Qiagen PCR purification process.
To clone the adaptored cDNA, I cut 2 mg of pUC19 using the following rxn: 2 ml pUC19, 2 ml 10x BSA, 2 ml NEB3, 13 ml H2O , 1 ml BamHI enzyme. I ran the rxn for 45 min at 37C. I cleaned up the reaction with a Qiagen PCR purification kit. I ran two reactions because the first one a screwed up and used NEB2 rather the optimum NEB3, normally I wouldn't care, but since I'm only cloning using one-cutter, I want to make sure I cut this plasmid well. The yield of cut DNA was:
Sample DNA (ng/ul) 260/280 260/230 total yield
pUC19 NEB2 59.3 1.78 mg
pUC19 NEB3 62.9 1.89 mg
I ran a phosphotase reaction to prevent plasmid self-ligation: 10 ml cut pUC19 (the NEB3 sample), 7 ml H2O , 2 ml antarctic phosphatase buffer, 1 ml antartartic phosphatase.
Finally, I prepared the ligation of the cut, dephosphorylated buffer with the adaptored phosphorylated size-selected cDNA: 2 ml T4 ligase buffer, 10 ml cDNA gel purified, 2 ml dephosphorylated pUC19, 5 ml H2O , 1 ml T4 ligase. I ran the rxn 30 min at 16C. I cloned the ligation into DH5a and plated 50 ml of each sample (e, e.2x, f, f.2x).
Tues July 24, 2007
I picked 24 colonies and grew them up in LB. After > 8hrs of growth, I miniprepped all 24 samples.
Wed July 25, 2007
I digested all 24 minipreps with 2 ml EcoRI buffer, 10 ml plasmid, 0.5 ml EchoRI, 0.5 ml HindIII, 7 ml H2O . The rxns were run 15 minutes at 37C followed by heat deactivation at 65C for 10 minutes. I ran the digestions on an agarose gel (actually two because I didn't have enough lanes). Unfortunately, I goofed a couple different ways. First, I tried using the matrix-impact 2 variable-spacing-electronic-multichannel-pipettor to put the first four samples in at the same time. A nice time saving idea, but I learned that electronic multichannel pipettors are a very bad way to load a gel. The problem is that as soon as you push the dispense button on a multichannel, all of the sample is going to come out whether or not you're got the pipettes accurately centered or not. With the manual multichannel pipettor, you can dispense little-by-little - making sure to optimize the amount of sample that falls to the bottom of the gel well as you go along. However, there is no manual multichannel with variable tip spacing. So in short, I lost the first 4 digestion samples into the gel buffer. The second problem is that I'm filling this section out in the middle of September and I didn't mark on a sheet of paper the order of the samples on the gel. I'd assume it goes e5, e.2x 1, ..., f1, ..., f.2x 1. However, it doesn't matter too much because the gel (Figure ) looks just fine, and virtually all inserts are > 500bp, which is what I wanted to see.




Please see the pdf version for figures
Figure 6.66: Cut ESTs cloned into pUC19 are in general longer than the 500 bp gel selection length.
Since the gel looked fine (Figure 6.66), I spec'd 20 samples (the first five from every sample type) and sent them out for sequencing. The spec values and the gene the sequenced EST best matches on the genome is shown in the table below.
Sample Time ng/ul 260/280 260/230 Top Hit
e1 11:56 AM 370.89 1.96 2.26 16S
e2 11:57 AM 379.76 1.96 2.44 23S
e3 11:59 AM 209.03 1.98 2.25 ?
e4 11:59 AM 308.71 1.97 2.24 23S
e5 12:00 PM 415.84 1.94 2.28 23S
e.2x 1 12:01 PM 140.57 2.01 2.46 mdlA
e.2x 2 12:03 PM 349.38 1.96 2.29 23S
e.2x 3 12:03 PM 431.42 1.93 2.25 23S
e.2x 4 12:04 PM 512.19 1.91 2.23 tnaC
e.2x 5 12:05 PM 513.86 1.9 2.22 ?
f 1 12:05 PM 302.78 1.95 2.29 23S
f 1 12:06 PM 274.55 1.95 2.27 mglB?
f 3 12:06 PM 298.16 1.95 2.25 23S
f 4 12:07 PM 505.84 1.88 2.19 23S
f 5 12:08 PM 382.68 1.94 2.26 16S
f.2x 1 12:09 PM 290.61 1.98 2.29 16S
f.2x 2 12:10 PM 376.87 1.94 2.24 23S
f.2x 3 12:10 PM 275.03 1.98 2.25 23S
f.2x 4 12:11 PM 530.4 1.89 2.2 ?
f.2x 5 12:11 PM 248.21 2.01 2.31 proS
The three questions marks in the table above are the sequencing rxns that failed. Agencourt ran them again and they worked. However all three mapped back to rRNA (can't remember if it was 16S or 23S, but who cares, I hate them both).
To Do!!!  I just realized with the MicrobeExpressB kit there might be a faster way to cycle the procedure. I currently run through the kit, concentrate with EtOH, and start over. The EtOH precipitation step takes up the majority of the time. Rather than cleaning up, I can just take the elution (ignor the wash step, hopefully won't result in too much loss) and start over again with new oligo and beads. Clearly runing 2x through the kit is helping. I'd guess I'll get dimminishing returns in terms of rRNA removal as I go through 3x, 4x, ..., but it is worth a shot. Another strategy, use some 1st strand cDNA to run a PCR reaction to amplify full length rRNA (get general primers to all E. coli rRNA [is all E. coli rRNA sequence exactly the same?]). During the amplification, use a biotinylated primer for the reverse strand. Now catch the biotin on a dynal bead and strip away the other strand. Whenever you finish with other first strand reactions in the future, add the biotinylated rRNA plus a dsDNA exonuclease to zap the first strand cDNA complement to the rRNA. This should work if the dsDNA exonuclease doesn't also attack RNA:DNA hybrids. If it does attach hybrids, I'd need to do a RNAse digestion, clean up the cDNA, and then use the exonuclease trick.

results and conclusions from the further removal of rRNA focus on ESTs

Tue Sep 18 18:57:34 EDT 2007
Unfortunately, single-read Sanger DNA sequencing is still kinda expensive to be running hundreds of test samples. So I'm going to just calculate some crude stats with the 20 or so samples that I do have.
For this round alone, I ran 10 samples with 1x MicrobeExpressB and 10 samples with 2x MicrobeExpressB. As I found out many months ago, 1x MicrobeExpressB removes a drastic amount of rRNA (I've was never able to sequence mRNA from a sample that was not run through MicrobeExpressB, it has all been rRNA). From the gel alone it appears that the second round removes addition amounts of rRNA (see Figures 6.62 and 6.65). In the table above 1 in 10 (10%) samples in the 1x MicrobeExpressB was mRNA, while 3 in 10 (30%) samples were mRNA for the 2x MicrobeExpressB. Seventy-percent unwanted sequence still sucks, but it is a hell-of-a-lot better than ninety-percent. I'd like to get at least 50% mRNA. Hopefully the MicrobeExpress cycling idea or the dsDNA exonuclease idea will give me that extra boost.
With my previous results with 1x MicrobeExpressB, the ratio was quite a bit better. The overall ratio for 1x MicrobeExpress is 3 in 20 (15%) mRNA.
One final thing to note, tnaA has now been cloned two independent times. Given that I've only sequenced six mRNAs so far (and lots of rRNA), it's kinda odd that one gene has shown up 2x already.

6.14  How many samples must be pooled for one 454 run

After all the work trying to recircularize the PETs (or do some sorta single-primer PCR), I've decided to consider brute force accumulation of PETs. It's clear that I don't get a huge amount of PET DNA when I run the MmeI cut the RCA amplified cDNA (see for example Figure 6.44). However, I want to pool the samples anyways, because I don't want to just run one species, one condition in my pyrosequencing reaction.
So Tim suggested I forget all of the fancy 2nd amplification strategies, and just pool a bunch of samples together. The question then is: how much DNA do I get from one MmeI digestion and PET purification.

6.14.1  Amplifying the PET DNA

Wed, Sep 19, 2007
I ran the following RCA reaction on samples A.1x and A.2x from section 6.13.2 on page pageref. 5.25 ml 10x RCA buffer, 2 ml dNTP, 5 ml template, 2.5 ml hexamer, 35.25 ml H2O . Incubate at 95C for 5 minutes, place on ice, add 2.5 ml f29 enzyme and incubate at 30C for 12 hr followed by heat inactivation at 65C for 10 minutes.
Thur, Sep 20, 2007
I concentrated the RCA reaction with EtOH and resuspended in 60 ml TE. The yields from these reactions were:
Sample DNA (ng/ul) 260/280 260/230 total yield
RCA sample A.1x 898.5 53.9 mg
RCA sample A.2x 995.2 59.7 mg
Brief Conclusions:   The yields were half what I got when I performed the exact same reaction back on July 4 (see the table in section 6.13.2 on page pageref). Maybe something is going bad with my DNA, enzymes, or dNTPs?
Fri, Sep 21, 2007
I cut approximately 40 mg of RCA samples A.1x and A.2x from above. For each digestion, I used 40 ml RCA DNA, 30 ml MmeI, 14.1 ml SAM, 45 ml NEB4, 320.9 ml H2O for a total volume of 450 ml . I ran the reaction at 37C for 30 minutes in a water bath. I then precipitated each reaction in EtOH plus 1 ml of glycoblue. I resuspended the cut DNA into 20 ml of TE. I ran all 20 ml of the two samples (plus an extra 5 ml of loading dye) out on a Nusieve 4% agarose gel (Figure ). The gel was SYBR Safe, I used 1:20 fisher dye, I ran the gel 45 min at 90V and 45 min at 120V. The gel looked great by eye. And although not perfect when photographed by a point-and-shoot camera, it wasn't too bad either.
Please see the pdf version for figures
Figure 6.67: The lanes from top to bottom are MmeI cut A.2x, bank, 4 ml 25 bp ladder (invitrogen), NEB 2-log ladder, 2 ml 25 bp ladder (invitrogen), MmeI cut A.1x.
I cut extremely thin slices around the PETs A.1x and A.2x. For controls, I cut the 125 bp fragment from the Invitrogen ladder, and I cut another band at around 100 bp in a lane with no DNA as a negative control. I cleaned the reactions up with a Qiagen column-based gel purification kit, and I eluted into 30 ml of EB. I quantified the DNA using 20 ml of the purified DNA with the hsDNA Qubit kit.
Sample DNA (ng/ul) total yield (ng)
gel purified PET A.1x 0.538 16.14
gel purified PET A.2x 0.576 17.28
125 bp band from ladder 1.956 58.68
gel slice from DNA-free lane 0.0254 0.762
Brief Conclusions:   Well, the yields weren't so great, but they're much better than background. The empty gel-slice resulted in 0.762 ng yield of DNA (presumably this is just background noise from the Qubit), while the purified PET gel-slices resulted yields of around 20 ng. I'm not sure why the 125 bp piece had such a high yield, because on the gel it actually looks weaker than the PET (Figure 6.67). Maybe the PET is a little short, so it doesn't purify as well? Maybe I was too conservative when I cut the band out (most likely?)34. As things stand, I need 2 mg of DNA to give to the pyrosequencing folks. That would require pooling 100 samples (ouch!). However, I could cut more RCA, since I still have quite a bit left from each RCA run. Perhaps, I need to try other gel purification methods or optimize my digestion to use less MmeI, or this will cost quite a bit per sequencing run on MmeI enzyme.