Seen part of vector get deleted postligation/transformation?

Use this category for questions directly related to DNA manipulation (isolation, purification, sequencing, etc.) and questions regarding general PCR methodologies

Moderators: mchlbrmn, r.rosati

Seen part of vector get deleted postligation/transformation?

Postby yk_lady » Dec 14 2006 12:40 am

I've been doing this for three months and have gotten nowhere. Please advise. I apologize for the long and detailed post, but I figured I'll get lots of questions subsequently anyway about every step I did. Here's the quick jump to the conclusion:

The SUMMARY
I have a vector (5 kb) and three inserts (65 bp, 1500 bp, 2600 bp) and I'm trying to ligate them to create three invididual constructs (same vector plus one of the inserts). Prior to ligation, I verify that the vector is linearized on a gel and it's at the right band size. The inserts are PCR products and are verified on a gel and then purified. Then I do ligation, transformation (DH5a), and miniprep of 4 colonies per plate. I do restriction digest analysis to separate the insert from the vector and this is what I get:

1. The 65 bp insertion works.
2. The 1500 bp inserted clones produce a weak band at ~3 kb and a very strong band at ~1500 bp.
3. The 2600 bp inserted clones produce a very weak band at ~5 kb and a very strong band at ~3000 bp.

Sequencing data of these clones show that the T7 primer sequences a part of the construct that is 2 kb away from the actual T7 primer site, suggesting that 2 kb somehow DISAPPEARED from the vector (and the insert is supposed to be inserted somewhere in this 2 kb). This is supported by the 3000 bp fragment seen on the gel from the restriction digest analysis.

So basically some time after I digest the vector to when I miniprep the ligated DNA, the vector gets chopped up. But it's still intact enough for it to grow in the bacteria.

What's going on?

And there is only one restriction site for each enzyme in the vector.
_________________________________________________
The INSERTS
1. PCR to amplify insert from cDNA - primers contain two different restriction sites (extra nucleotides at the end).
2. Run PCR product on gel - bands show up exactly where they should be.
3. Gel extraction (qiagen kit)
4. Digest insert with two enzymes (HindIII/NotI) (NEB). 15 min @ 37 degrees, 20 min at 65 degrees for inactivation. (20 uL total vol)

The VECTOR
1. pcDNA bought from Invitrogen - digest with HindIII/NotI.
2. CIP treat DNA - 1 hr @ 37 degrees.
3. Gel purify vector DNA. (bands show up where they should - cut vector runs higher than uncut vector).
4

The LIGATION
1. Use 3:1 molar ratio of insert:vector while using 50 ng of vector.
2. 1 uL Quick Ligase from NEB + 2x buffer - total vol 20 uL.
3. Incubate at RT for 15 min.

The TRANSFORMATION
1. Transformed 10 uL of ligation reaction into DH5a comp cells.
2. Plated 100 uL of each reaction.

The PLATES
1. I got anywhere from 4 to about 100 colonies on some of the plates - with 2-4 colonies on the negative control plates.
2. I picked 4 clones from each plate to screen.
Last edited by yk_lady on Dec 15 2006 1:37 pm, edited 1 time in total.
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Postby bolcsodal » Dec 14 2006 4:59 am

Hi,

What genes are located on the fragments? And, what is the copy number of the plasmid? If it is a high copy number plasmid it would be better to move onto a moderate or low copy vector.
bolcsodal
technician-in-training
technician-in-training
 
Posts: 5
Joined: Dec 14 2006 4:48 am
Location: Hungary

Postby yk_lady » Dec 14 2006 8:30 am

Here are the three inserts:

65 bp - a signal sequence
1500 - amyloid precursor protein (APP)
2600 - puromycin sensitive aminopeptidase (PSA)

Why would using a moderate or low copy plasmid help? And yes, I am using a high copy plasmid.

Thanks!
Last edited by yk_lady on Dec 15 2006 7:41 am, edited 1 time in total.
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Postby bolcsodal » Dec 14 2006 9:17 am

Sounds like they are eucaryotic cDNA-s. Even it is better to use a moderate copy number plasmid, because if the insert is too large for the plasmid, the high copy plasmids will eagerly cut out fragments form themselves. Do you know the type of the replication origin of your plasmid?

Better to decrease the cultivating tempereture around 28Co even if it is a high copy plasmid. For example pBluescript (ColEI) lacks the replication regulator from the plasmid (rom) and at high temperatures like 38-40Co the copy number could be several thousand copies/cell. Ofcours at such a high copy number, the frequency of mutations, and plasmid rearrangements are higher.
bolcsodal
technician-in-training
technician-in-training
 
Posts: 5
Joined: Dec 14 2006 4:48 am
Location: Hungary

Re: Can ligation, transformation, miniprep cleave a vector?

Postby relaxin » Dec 14 2006 9:38 am

yk_lady wrote:Sequencing data of these clones show that the T7 primer sequences a part of the construct that is 2 kb away from the actual T7 primer site, suggesting that 2 kb somehow DISAPPEARED from the vector (and the insert is supposed to be inserted somewhere in this 2 kb). This is supported by the 3000 bp fragment seen on the gel from the restriction digest analysis.


Is the T7 primer site closed to the cloning site? The restriction sites seem to be intact and you can cut out the insert. What is this 2 kb deletion?

Can you check the presence of insert by PCR using gene-specific primers as well as primers based on the sequences flanking the cloning site?
Not affiliated with any company. Mention of a specifc product does not imply my endorsement of the product. No conflict of interest or guarantee to work on the advice given. Do as I say, not as I do. Not liable to the loss of your valuable samples.
relaxin
PI of Posters
PI of Posters
 
Posts: 6724
Joined: Jan 11 2006 12:40 pm
Location: Mauna Kea

Postby yk_lady » Dec 14 2006 10:18 am

in response to bolcsodal, the plasmid has a pUC origin and an f1 origin ... is this what you mean? is a 1500 or 2600 bp insert to big for a 5.6 kb vector?

in response to relaxin, yes the T7 primer site is close to the cloning site (like 10 bp away). the thing is - the insert is supposed to be in the multiple cloning site but the T7 primer sequencing data don't show base pairs until like site ~3300 of the vector.

The exact vector I'm using is pcDNA3.1/Hygro(+) from Invitrogen.
Image

The WEIRD thing is that the sequencing data (using T7 primer) for the 1500 and 2600 inserted constructs show that the beginning of the sequence starts right after that SV40 pA site - at the EXACT spot for both of them. You'd think that if it were random cutting due to the comp cells or whatnot, I should get random sized vectors with no insert then. But 4 clones screened from the 1500 bp constructed vector and 4 from the 2600 bp constructed vector plates give different bands from each other, but the same within the 4 clones themselves (sorry, don't know how to say that more clearly). Again, if it were random, I shouldn't be getting the same result within the same ligation/transformation reaction right?

So from the sequencing data - it seems like the Hygro resistance, SV40 pA (i don't even know what that is), the SV40 ori, f1 ori and BGH pA should all be cut out. Is this a correct assumption?

I've also done the PCR test for the inserts - got nothing. The only one that worked is that 65 base pair one. Sequencing data came back and it was in there and PCR analysis of the construct showed the insert was in there as well.

I'm stumped and so is my supervisor!
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Postby mchlbrmn » Dec 14 2006 11:07 am

I don't know. It sounds like everything described is correct, although there are some short cuts you might not do since you're having trouble (but they are approved shortcuts, like the short ligation and rest. digestion times). I might say maybe one of the REs cutting the vector didn't work, but the 65bp construct worked fine. Hopefully you mimimized time the DNA was in UV, although I don't know if that could cause such a problem.

I'm thinking it's possible you lost your multiple cloning site withe the restriction sites you used. Both clones produce a stronger smaller looking band because it's the supercoil form (does the gel band look rounder than the other possibly?) and a fainter, larger relaxed circular or linear form. One construct attempt produces smaller plasmids than the other. I dunno why more than you do. Maybe there was some insert sequence specific interaction?? Maybe the T7 itself moved?? Maybe the cells are playing games with you?

Sorry I can't really help. Good luck.
mchlbrmn
ModSquad
ModSquad
 
Posts: 3581
Joined: Dec 13 2005 1:03 pm
Location: Boston, USA

Postby bolcsodal » Dec 14 2006 11:54 am

I must agree with mchlbrmn. The gel picture, what you described is more closely to an uncut vector. Anyhow my question is why you are using such a complicated vector, for the cloning procedure? This vector contains the full set of SV40 promoter, and replication togeather vith f1 origin (so single stranded DNA can be purified due the f1 viral replication), and another set of replication/promoter and finally the Pcmv (wich is I guess, the cythomegalovirus promoter). If you do not want to use tissue culture transfection, at this step of cloning, I think it would be better to use a more simple cloning vector - like pBluescript, and a very simple method to determine the quality of the ligation/transformation. This method is the alpha complementation (Blue/White selection). And after you will have the correct plasmid, and the insert, you will be able to clone it into this special one.
bolcsodal
technician-in-training
technician-in-training
 
Posts: 5
Joined: Dec 14 2006 4:48 am
Location: Hungary

Postby yk_lady » Dec 14 2006 1:49 pm

to mchlbrmn: i'm thinking the stronger lower band in my restriction digest analyses can't be supercoiled and uncut because they appear just slightly curved upwards. but when i run the same DNA right next to it UNDIGESTED, it appears to curve downwards. so i think it IS cut, just somehow the vector is missing a part of itself and there's no insert.

But if you want something even weirder to think about - if you look back at my first post - I show the different sizes of the bands that appear for the respective constructs. For the construct that should have a 1500 bp in it, the strong band appears at 1500, with the weaker one above it. And then for the construct that should have a 2600 insert in it, the strong band appears right about there (I use 1 Kb plus ladder that only labels 2 or 5 kb, so i have to estimate in between) with a weaker band above it. Is that something significant?? Seriously, these vectors must be playing games with me. I even considered the possibility that for some reason - the inserts themselves were able to relinearize, somehow get Amp resistance and then grow themselves up inside the bacteria. It's magic.

But DNA sequencing data came back and I got no insert in the data ...


to bolcsodal: yes i am going to do transfection with these constructs (ultimately create stable cell lines) and i need a mammalian expression vector. The inserts themselves ARE already in pBluescript and I'm just trying to transfer them to a mammalian vector.

So do you guys believe my interpretation that it's not supercoiled DNA and that something seriously disappeared from the vector during ligation or transformation? Haha.
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Postby relaxin » Dec 14 2006 2:19 pm

I think the inserts got deleted along with the MCS. Try to screen more colonies by PCR using primers flanking the cloning site as well as insert-specific primers.

I cannot explain how you got the insert band by restriction digestion.
Not affiliated with any company. Mention of a specifc product does not imply my endorsement of the product. No conflict of interest or guarantee to work on the advice given. Do as I say, not as I do. Not liable to the loss of your valuable samples.
relaxin
PI of Posters
PI of Posters
 
Posts: 6724
Joined: Jan 11 2006 12:40 pm
Location: Mauna Kea

Postby mchlbrmn » Dec 14 2006 2:36 pm

Yes, I accept your interpretation: it's magic.
I also noticed that 2.6 is close to 3.0 kb, but in my gel and marker set the difference is clear. I assume this is coincidence.
Supercoil usually runs faster than linear. When you ran them side by side where was the supercoil native form?
From your sequence data you know the plasmid got messed up. I was wondering if there's some kind of homology between the inserts and the vector that promotes some kind of rearrangement/deletion?

If the size difference is large enough, maybe you could screen a large number more easily by running out a crude lysate without digesting the plasmid. There should be a good size difference between a plasmid with an insert, and one that lost -did you say it was 3kb? The bacterial pellet can just be heated in a "cracking" buffer to lyse them, and the cell lysate run out directly on a gel. To grow the cultures easily 100 ul LB-Ap in a microtiter plate can be innoculated and incubated stationary 37 deg overnight. When I did this many years ago, the cell titer was very high (> 10^9) even though it's not shaken. Any good clones can then be grown up with a conventional miniprep for restriction digestion and sequencing. This would make it easier to screen more colonies to find a good one.
mchlbrmn
ModSquad
ModSquad
 
Posts: 3581
Joined: Dec 13 2005 1:03 pm
Location: Boston, USA

Postby yk_lady » Dec 14 2006 2:56 pm

relaxin: i agree the MCS SHOULD have been cut out .... but then how come when i go to do the restriction digests, the bands appear to be linearized? in that they run curved upwards rather than curved downwards like supercoiled DNA would ...

i am currently in the process of screening more clones - simply by picking them off the plate and then using the same primers i used to make the PCR product to amplify the insert if it is in there.

and you called it my "insert band" but it most likely isn't, heh. it's just coincidentally at that location??

mchlbrmn: to answer your question, when i run the supercoil native form next to the digested constructs, they look EXACTLY the same except for the direction of the curve of the band. so in the supercoiled native sample lane, i also get the strong lower band with the weaker one on top. so this i think means my clones are being cut (which HAS to mean the MCS is still there??? because those sites only occur in the MCS) but they're just shortened somehow.

as for sequence homology, this could be the case ... but then i get this same bizarre result when i use different inserts - with the 1500 APP insert and and 2600 PSA insert. how likely is it that they both can have homology to the vector such that the SAME exact sequence of the vector gets deleted??

i really appreciate all the responses! and the mystery continues ....
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Postby bolcsodal » Dec 15 2006 3:15 am

Do you know the exact sequence of the plasmid, and the inserts? It would make easy to create the ligation in silico. If you know the seq, you will be able to check wether are there any homology or not. For the screening I would use t7 and one of the specific primer that is at the opposite direction. Instead of CIP I would use SAP, which can be heat incativated. Or maybe it doesnt make sense to use phosphatase, just use a large excess of insert DNA. Exactly..why you have used CIP? It is an oriented ligation with two different ends. In this situation it wont increase the efficiency of the ligation.

If the insert is already presented in pBLuescript, try other RE pairs for cloning.
bolcsodal
technician-in-training
technician-in-training
 
Posts: 5
Joined: Dec 14 2006 4:48 am
Location: Hungary

Postby mchlbrmn » Dec 15 2006 10:11 am

Bolcsodal, phosphatase does, I believe, actually help ligation efficiency somewhat in most cases. It prevents the vector from ligating to a second molecule of vector.

YK, If the uncut and cut bands have the same sizes I'd conclude the enzyme did not cut. Maybe you're more experienced at looking at the shape of bands than me, but I just use that as a little secondary data whereas the band migration rate is primary. Perhaps other factors in the enzyme reaction and storage buffers affect the conformation of the plasmids and the shape of the gel bands sometimes?
mchlbrmn
ModSquad
ModSquad
 
Posts: 3581
Joined: Dec 13 2005 1:03 pm
Location: Boston, USA

Postby yk_lady » Dec 15 2006 1:34 pm

bolcsodal: i did do the PCR screening with the T7 primer and the insert's anticoding primer (the same one i used to make the insert). i even did it side by side with the insert's own coding primer (along with the anticoding primer), so that on a gel the bands should be staggered a little since the T7 site is a bit upstream. only the 65 bp insert construct showed exactly what i expected (the T7 primer sample band being higher than the specific primer sample). All the others showed nothing where i expected the insert to be (i ran the PCR product itself alongside to show where it should be).

i actually started out using SAP but because i saw so many colonies on my negative control plates (linearized vector) and absolutely zero colonies on my experimental plates, i just switched back over to CIP. after using CIP i did get colonies on my experimental plate (with no more than 4 on my negative control - as opposed to hundreds before) - but that led me to the problem i'm curently having with a part of the vector being deleted (as proven by sequencing data).

mchlbrmn: i agree that looking at the shape of the band is not very "scientific" ... but the sequencing data also showed that the vector was missing ~2kb ... so sure - those bands could be uncut - but certainly shouldn't be running at around ~1500 or ~3000. The vector alone is 5 kb. if there's an insert it in, it should be 6 or 7 kb. would an uncut ~5kb vector run so low? i've run the vector itself (straight from the invitrogen tube - so no insert) uncut and it runs at just slightly below the 5 kb mark. when linearized, it runs a bit above the 5 kb mark.

ok - and assuming that the enzyme didn't cut - how would you recommend that i move forward? how can i analyze my DNA then to see if it is what i want it to be? i've already sequenced it - and it shows 2 kb missing. i've also tested each of the enzymes on the vector itself (again straight from the invitrogen tube) and they all work efficiently.

thanks for all the ideas!
yk_lady
technophile
technophile
 
Posts: 20
Joined: Dec 13 2006 11:22 pm

Next

Return to DNA and General PCR Methods

Who is online

Users browsing this forum: No registered users and 2 guests