Ghost bands on plasmid preps
Daniel MacArthur
I'm an undergraduate biochemistry student currently attempting to write up a
practical report on a plasmid miniprep practical. Briefly, we took two E.
coli colonies transfected with pBluescript IISK+ vectors (one containing a
fragment from an EcoRI digest of lambda phage DNA and one with no insert),
purified the plasmids using a Quantum Miniprep from Bio-Rad, digested the
plasmid DNA using EcoRI, BamHI, and a combination of the two, and then ran
out the products on an agarose gel.
The experiment worked beautifully, but I'm having a little trouble with
interpreting the results. In several of the lanes there are faint bands
running at roughly half the size of the full-length linear plasmid, which my
demonstrator called "ghost bands". She told us that these bands were
actually single-stranded plasmid molecules formed during the cell lysis step
of the miniprep, which seems reasonable given their size in relation to the
full-length plasmid. However, we were also referred to two articles by Paul
Hengen (who I believe posts/posted on this newsgroup) which seem to be
putting forward considerably more complex explanations for the existence of
these "ghost bands" (e.g. that the bands represent "double-stranded, cyclic,
coiled DNA composed of two intertwined, but permanently denatured, single
strands of plasmid DNA").
Does anyone have any suggestions regarding the nature of these bands? My
demonstrator's explanation is attractively straightforward and seems to fit
the data with respect to the plasmid without an insert (the linear plasmid
runs at ~3 kb and the ghost band at ~1.5 kb), but it doesn't quite fit with
respect to the plasmid with an insert (linear at ~7.3 kb, ghost at ~2.7 kb),
and seems to be contrary to Paul Hengen's explanations. Any advice on this
matter would be greatly appreciated.
Daniel MacArthur.
3rd Year Medical Science
University of Sydney, Australia
jw
(completely uncut plasmid). When supercoiled plasmid is run on an agarose
gel, it runs faster than the linearized plasmid. If this digest was only
done for an hour or so, and with a relatively small amount of enzyme or
large amount of DNA, I would expect to see some uncut plasmid.
One other possibility is that the buffer used for the digest did not have
enough salts for proper BamHI digestion, resulting in a little star
activity, which means the enzyme will also cut at another recognition site
slightly different that the native BamHI recognition site. BamHI is known
to show this star activity with low ionic conditions, or with high glycerol
concentration greater than 5%.
The quickest way to determine if it is supercoiled (uncut) plasmid is to
just run a small amount of your uncut plasmid in a lane next to the cut
plasmid, and compare the bands. Chances are, I feel, that this is what you
have observed.
JW
--
Reply Email:
ng...@worldnet.att.net
Dr. Hiranya S. Roychowdhury
difficult to interpret any such results. However, the explanation of
"single stranded" plasmid (if that is what your demonstrator really meant)
is not correct. single stranded nucleic acids do not run as a tight band
and certainly not at 1500bp size on a TAE or TBE gel.
The first possibility that comes to mind is that what you are observing in
the gel are "un-digested" supercoiled plasmid. This is rare, since after
overnight digestion , especially, it would be hard for any plasmid to exist
in supercoilded form.
BTW, did you run the undigested plasmid, with and without the insert,
alongside? If you did, how did those lanes compare? If you didn't, you should.
The second possibility: Although you have not described in great detail, is
it possible that the EcoRI insert from the phage DNA contains a BamHI site?
From what you have said, the insert is ~4kb. So, think in terms of an extra
BamHI site. Does the BamHI digest (single) produce any band smaller than
what is produced by single EcoRI digest?
I don't remember what Dr. Hengen had explained about "ghost bands", but I
think he was talking about undigested samples. It is a common observation
that even when a supercoiled plasmid is excised from a gel and
electrophoresed, it resolves into a relaxed form and a supercoiled form.
This phenomenon is dependent on the buffer, the gel, the gel %, the plasmid
prep, voltage applied, etc.
Hiranya
Dr. Hiranya Sankar Roychowdhury
College Asst. Prof.
Molecular Biology,
Dept. of Chemistry & Biochemistry
Box 30001 - 3MLS
New Mexico State University
Las Cruces, NM 88003
Lab: (505) 646 4722
Office: (505) 646 8256
hroy...@nmsu.edu
---
ChenHA
jw wrote:
>
> This ghost band sounds to me like a small amount of supercoiled DNA
> (completely uncut plasmid).
Unlikely as this is ghost band is apparently a property of pBluescript
when used in XL1-blue This subject had been extensively discussed and
can be found in the article already mentioned:
ftp://ftp.ncifcrf.gov/pub/methods/TIBS/mar94.txt
The answer is that no one knows. There is really nothing much to add
unless someone is prepared to do more experiments on that like
sequencing the plasmid, or someone has some new insights into this
particular issue.
Nick Theodorakis
>
>Dear group,
>
>I'm an undergraduate biochemistry student currently attempting to write up a
>practical report on a plasmid miniprep practical. Briefly, we took two E.
>coli colonies transfected with pBluescript IISK+ vectors (one containing a
>fragment from an EcoRI digest of lambda phage DNA and one with no insert),
>purified the plasmids using a Quantum Miniprep from Bio-Rad, digested the
>plasmid DNA using EcoRI, BamHI, and a combination of the two, and then ran
>out the products on an agarose gel.
>
>The experiment worked beautifully, but I'm having a little trouble with
>interpreting the results. In several of the lanes there are faint bands
>running at roughly half the size of the full-length linear plasmid, which my
>demonstrator called "ghost bands".
Hengen's denatured ghost bands sounds like a reasonable explanation to me. Try
reducing the amount of time during lysis in the alkali/SDS step. The times that
I've seen them has been when I've been doing a lot of minipreps and they
probably spent too much time in the alkali.
Nick
--
Nick Theodorakis
nicholas_t...@urmc.rochester.edu
jw
ChenHA <hz...@freeuk.com> wrote:
>
> Unlikely as this is ghost band is apparently a property of pBluescript
> when used in XL1-blue This subject had been extensively discussed and
> can be found in the article already mentioned:
>
> ftp://ftp.ncifcrf.gov/pub/methods/TIBS/mar94.txt
>
> The answer is that no one knows. There is really nothing much to add
> unless someone is prepared to do more experiments on that like
> sequencing the plasmid, or someone has some new insights into this
> particular issue.
Thanks for the reference. I found this information very interesting and
helpful.
Dr. Hiranya S. Roychowdhury
>
>Unlikely as this is ghost band is apparently a property of pBluescript
>when used in XL1-blue
The phenomenon of "ghost bands" is not restricted to pBS in XL1-blue.
Michael R. Elliott
I have experienced something similar in the past, and I might be able to help.
First, I don't know that the bands you are seeing are ssDNA, as EtBr
intercalates very inefficiently to ssDNA, and whatever EtBr that does
intercalate is thought to be the result of duplex DNA formation. You can test
this ssDNA theory by running some of the undigested plasmid prep on a gel next
to the Bam/RI cut DNA-- if there's not a band in the uncut DNA that corresponds
to the digested DNA lane, then you know it is something that has occured during
the restriction digest. You should call the company that makes the enzymes and
see if any complaints have been logged for the EcoRI and BamHI. A few months
ago we had a similar problem with EcoRI from Promega. They sent us a new batch
and the problem went away. The extra bands may be due to inappropriate cleavage
of your plasmid (called "starring").
One last suggestion: the migration of DNA thru the gel is effected by the salt
concentration of the sample prior to loading in the gel. I know that when I
elute purified linear DNA in H2O, I often get smaller bands than expected. I
can correct the problem by adding 1x of any restriction enzyme buffer and
letting the sample sit for 2-3min on the bench before loading on the gel. Good
luck.
Rusty Elliott
Wake Forest University
North Carolina.
Daniel MacArthur
You said:
> I have experienced something similar in the past, and I might be able to
help.
> First, I don't know that the bands you are seeing are ssDNA, as EtBr
> intercalates very inefficiently to ssDNA, and whatever EtBr that does
> intercalate is thought to be the result of duplex DNA formation. You can
test
> this ssDNA theory by running some of the undigested plasmid prep on a gel
next
> to the Bam/RI cut DNA-- if there's not a band in the uncut DNA that
corresponds
> to the digested DNA lane, then you know it is something that has occured
during
> the restriction digest.
I did run a lane of undigested plasmid along with the various digested
samples in the original experiment (sorry, I wasn't very clear about this in
my post). The ghost band is present at essentially identical strength in all
samples, both digested and undigested. To me this suggests two things: (1)
the problem is with some step prior to the restriction digest, and (2) the
ghost bands (whatever they are) are resistant to both EcoRI and BamHI
digestion (I used both enzymes, alone and in tandem). Perhaps the
restriction sites are absent, or there are regions of ssDNA around the
recognition sequences, or the conformation of the DNA is too tight to permit
access by enzymes. But what this all says about the identity of the bands I
have no idea. :-)
Your point about EtBr's low (absent?) affinity for ssDNA is an excellent
one, and definitely suggests that the ghost bands are not simply
single-stranded plasmid molecules denatured during the cell lysis stage of
purification.
> You should call the company that makes the enzymes and
> see if any complaints have been logged for the EcoRI and BamHI. A few
months
> ago we had a similar problem with EcoRI from Promega. They sent us a new
batch
> and the problem went away. The extra bands may be due to inappropriate
cleavage
> of your plasmid (called "starring").
Apparently the ghost bands appear routinely on gels following the
purification of this particular plasmid (pBluescript), regardless of the
restriction enzymes used, so I don't think it could be contamination of the
enzymes (although that's certainly not out of the question in this
particular case).
Out of interest, do you happen to know what caused the "starring"? Was the
batch contaminated with another restriction enzyme, or was there something
affecting the activity of the EcoRI itself?
> One last suggestion: the migration of DNA thru the gel is effected by the
salt
> concentration of the sample prior to loading in the gel. I know that when
I
> elute purified linear DNA in H2O, I often get smaller bands than expected.
I
> can correct the problem by adding 1x of any restriction enzyme buffer and
> letting the sample sit for 2-3min on the bench before loading on the gel.
I suspect that the salt balance was a bit awry as my bands were not as clean
as I would have liked. However, I'm not sure if salt balance could be used
to explain the ghost bands, as the expected bands for linear, circular and
supercoiled plasmid DNA were all present at roughly the right positions on
the gel, while the ghost bands consistently appear in every
plasmid-containing lane as a faint, fast-running band. Could the salt
concentrations have produced this, or would a salt problem have affected the
running of all of the bands equally?
I've written your suggestion down, though, as I suspect it may come in handy
at some time in the future when my gels start giving different sorts of
strange results. :-)
> Good luck.
Thanks!
Daniel.
a.kr...@biosidus.com.ar
Hi Daniel:
Did you perform Phenol-chloroform extraction on your minipreps?
Some years ago I have had the same problem and I solved the problem
shorting the SDS-NaOH step (1-2 min.) and using Phe-chlorof (I didn't
routinely use Phe in my preps)
Hope this helps
Sincerely
Alejandro Krimer
"Daniel MacArthur" <dma...@hotmail.com>@hgmp.mrc.ac.uk on 04/03/2001
06:08:38 PM
Sent by: owner-...@hgmp.mrc.ac.uk
To: met...@hgmp.mrc.ac.uk
cc:
Fax to:
Subject: Re: Ghost bands on plasmid preps
Hi Rusty,
You said:
> Good luck.
Thanks!
Daniel.
---
Eric DUFOUR
Daniel MacArthur wrote:
Just a suggestion : I remember that there is some restriction enzymes able
to cleave SS DNA (the biolabs catalog should be a good source of info), just
try one of them on either your whole plasmid prep or each on the uncuted,
gel-extracted bands. If it cleaves it then 1) you will know that it is DNA, 2)
it was probably single stranded DNA 3) you will be able to redo the same
experiment adding a denaturation/renaturation step after digestion and then
compare the gel migration of the digested DNA with the theorical restriction
pattern of your plasmid.
If you can't do that I thing a very short digest of the ghosts bands with
an unspecific endonuclease followed by digestion with a specific restriction
enzyme (one with at least four sites in your plasmid) will answer the question.
Not beautifull but should work...
If you don't have enough DNA you cna try to southern blot the digests with
a probe covering the whole plasmid...
Hope It helped
Eric Dufour
Eric DUFOUR
I just found that explanation in the Quiagen plasmid purification handbook :
Extra DNA bands on analytical gel -a Dimer form of plasmid -b Plasmid has
formed denatured supercoils "this species run faster than closed circular and
are resistant to restriction digest ". Since their protocol is based on
alcaline lysis... They also state as it as been said before that it is due to
too long incubation in NaOH/SDS (too long lysis). I think there is a comment
on that in maniatis also...
Eric Dufour
R. John Lye
Daniel MacArthur wrote:
> I did run a lane of undigested plasmid along with the various digested
> samples in the original experiment (sorry, I wasn't very clear about this in
> my post). The ghost band is present at essentially identical strength in all
> samples, both digested and undigested.
That sure sounds like the infamous "form 4" DNA. It is resistant to
most restriction digestion. Make sure that you're not leaving the
prep in the alkaline lysis step for too long, that's where this band
comes from. It is irreversibly denatured, and won't cut, but can
actually transform bacteria. In my experience, keeping the time
from addition of the alkaline SDS to the addition of the neutralising
solution less than five minutes (preferably less than four minutes)
seems to minimize this ghost band.
Good luck,
John
Claudia Sutton
<Eric....@cmm.ki.se> wrote:
I discovered the same thing years ago, the band is denatured DNA. In my
case it was plasmid+insert-dependent (one clone gave me ONLY denatured
form). It can be almost completely solved by decreasing [NaOH] to 0.15M in
solution 2.
Claudia
Michael L. Sullivan
when we did sequencing on single stranded templates, I always quanitified
template by running it on an agarose gel and comparing band intensities.
Whether this staining is due to double stranded regions of the molecule,
intercalation of the dye in single stranded regions, or both, I'm not sure.
But single-stranded DNA will stain with ethidium, so you can't really use
staining to rule that out.
Mike
>
>Your point about EtBr's low (absent?) affinity for ssDNA is an excellent
>one, and definitely suggests that the ghost bands are not simply
>single-stranded plasmid molecules denatured during the cell lysis stage of
>purification.
>
Michael L. Sullivan, Ph.D
U.S. Dairy Forage Research Center
1925 Linden Drive West
Madison WI, 53706
(608) 264-5144 Phone
(608) 264-5147 Fax
---
Daniel MacArthur
> Did you perform Phenol-chloroform extraction on your minipreps?
> Some years ago I have had the same problem and I solved the
> problem shorting the SDS-NaOH step (1-2 min.) and using Phe-chlorof
No, I used an SDS/NaOH lysis solution, which may well have been the cause of
the problem. If it's not too much trouble, could you give me a little more
detail on the composition of the phenol-chloroform solution? I'd be quite
keen to try it out and see what happens.
Thanks in advance,
Daniel.
Daniel MacArthur
> That sure sounds like the infamous "form 4" DNA. It is resistant to
> most restriction digestion. Make sure that you're not leaving the
> prep in the alkaline lysis step for too long, that's where this band
> comes from. It is irreversibly denatured, and won't cut, but can
> actually transform bacteria. In my experience, keeping the time
> from addition of the alkaline SDS to the addition of the neutralising
> solution less than five minutes (preferably less than four minutes)
> seems to minimize this ghost band.
eliminating it altogether (another poster in this forum had some great
suggestions regarding alternatives that I'm quite keen to try out).
Now this might be a stupid question, but what exactly is meant be the term
"irreversibly denatured"? When I read Paul Hengen's 1996 article on ghost
bands [1] (derived from discussions on this newsgroup) I was confused by his
description of the bands as "double-stranded, cyclic, coiled DNA composed of
two intertwined, but permanently denatured, single strands of plasmid DNA".
What is the chemical basis for this denaturation? Is the DNA covalently
modified? If not, what stops it from reforming dsDNA? If it is covalently
modified, how can it still transform bacteria?
Sorry if the answer is really obvious, but I've been thinking about it for a
while and can't seem to figure it out. Any suggestions would be very welcome.
Daniel.
Nick Theodorakis
>
>John:
>
>> That sure sounds like the infamous "form 4" DNA. It is resistant to
>> most restriction digestion. Make sure that you're not leaving the
>> prep in the alkaline lysis step for too long, that's where this band
>> comes from. It is irreversibly denatured, and won't cut, but can
>> actually transform bacteria. In my experience, keeping the time
>> from addition of the alkaline SDS to the addition of the neutralising
>> solution less than five minutes (preferably less than four minutes)
>> seems to minimize this ghost band.
>
>OK - I think next time I'll try cutting down the length of the lysis step or
>eliminating it altogether (another poster in this forum had some great
>suggestions regarding alternatives that I'm quite keen to try out).
Generally, once the alkali lysis solution has cleared you can go on to the
neutralization step.
>
>Now this might be a stupid question, but what exactly is meant be the term
>"irreversibly denatured"? When I read Paul Hengen's 1996 article on ghost
>bands [1] (derived from discussions on this newsgroup) I was confused by his
>description of the bands as "double-stranded, cyclic, coiled DNA composed of
>two intertwined, but permanently denatured, single strands of plasmid DNA".
>What is the chemical basis for this denaturation? Is the DNA covalently
>modified? If not, what stops it from reforming dsDNA? If it is covalently
>modified, how can it still transform bacteria?
I don't think anybody really knows for sure (OK, maybe I should say I don't
know). My guess is that it's not covalently modified, but kinetically trapped in
a conformation unfavorable to re-annealling. The bugs have a lot enzymes for
dealing with messed up DNA, so that can probably unwind it and fix it if you
give it to them.
I've seen it myself on occasion, and it's true that it doesn't seem to cut with
any restriction enyme. Interestingly, I never saw it in the old days of CsCl
preps; I guess it doesn't band it the bouyant density of ccc superciled DNA.
Nick Theodorakis
>
>Alejandro:
>
>> Did you perform Phenol-chloroform extraction on your minipreps?
>> Some years ago I have had the same problem and I solved the
>> problem shorting the SDS-NaOH step (1-2 min.) and using Phe-chlorof
>> (I didn't routinely use Phe in my preps).
>
>No, I used an SDS/NaOH lysis solution, which may well have been the cause of
>the problem. If it's not too much trouble, could you give me a little more
>detail on the composition of the phenol-chloroform solution? I'd be quite
>keen to try it out and see what happens.
I don't believe the phenol/chloroform gets rid of the "form 4" DNA, but I do use
it sometimes on minipreps in which I don't use column purification.
Recipe:
for crytalline phenol, heat to 60C to melt, saturate with water. Allow the
phases to separate, remove the water (upper phase), and re-equilibate once or
twice with 0.1M Tris pH8; check the pH of the upper phase with pH paper and make
sure it's >7. Then remove upper layer and replace with TE. Some people also
prefer to add 8-hydroxyquinoline to 0.1% or so to retard oxidation; I have
noticed thet neutral to basic phenol tends to off a little quicker than acidic
phenol. Or just aliquot it and freeze the aliquots.
Or you can just buy it already equilibrated.
For phenol:CHCl3, just mix an equal amount of the above phenol and chloroform.
Some people prefer to use a 50:1 mix of chloroform and iso-amyl alcohol, but I
haven't noticed that it's necessary.
NB: chloroform dissolves polystyrene plastic; use polypropylene or glass.
NBB: Phenol is very caustic; use appropriate eye, face, and body protection when
handling, especially when equilibrating it.
I'm also probably obligated to mention that handling open bottles of chloroform
outside of a fume hood probably generates air levels of CHCl3 that exceed OSHA
standards of safety. But if you wait a few months, the current adminstration
might roll back those standards. ;-)
--
Nick Theodorakis
nicholas_t...@urmc.rochester.edu