Dear Xin-Hua,
The questions are everything but stupid. But obviously, it is not the
first time that, under one or the other form, they are asked to the
community. If I had time I wold compile a FAQ !
So, for point 1 to 4, cumulated TS lab wisdom is that the
concentration of antibiotics to be used (and this is VERY true for
G418, quite true for hygromycin, and much less for Blasticidin)
depends on the medium used and the confluence of the cells. When we
use commercial HL5c (Formedium, which works for 90% of our strains) or
a traditional, rich recipe for HL5 (that works for the other 10% of
the strains), we vary the concentration by a factor TWO. In Formedium
HL5c, 5 microg/ml, and in HL5, 10 microg/ml are sufficient for G418
and BlastS. But crucially, this depends on the cell density. The
denser he cells, the harder hey are to kill. Therefore, my suggestion
is to finetune your electroporation conditions (see previous DictyNews
correspondence) so as to kill about half of the cells, and plate them
on 2 to 3 10 cm dishes, so that hey will be between 20 and 80%
confluent max the next day. Next day, decant medium and dead cells,
and apply antibiotics at 3 different concentrations, for example 5,
7.5 and 10 microg/ml. This ensures that at least one condition of
concentration/density will be perfect and it also increases the
chances to obtain obviously independent clones. Then, after 3 to 10
days, pick the forming colonies and ransfer to 12-well plates. Most of
the time, this leads to clonal populations, as judged by the pattern
of GFP expression, or PCR for KOs. If needed, sometime we clone on
Klebsiella, but we are not so happy about that, because it is hard to
keep selection. Nevertheless, it is possible to use plates containing
as high as 50 microg/ml of antibiotics, a lawn of bacteria and
dilutions of Dictys.
Good luck, but please no Voodoo!
Thierry
On 4 Nov 2009, at 21:35, Liao, Xin-hua (NIH/NIDDK) [F] wrote:
> Hi, dear all,
> Since I joined Dicty community, I encountered a lot of technique
> problems or confusions. Some of them may be not a real problem, and
> the question may be stupid, but they all bother me for a long time.
> I put them all here, hope you can help me clarify them.
>
> 1. Blasticidin selection: for knock out, after transformation
> for around 5 days under blasticidin selection, I transfer single
> colonies under medium to 96-well plate. For 96 colonies, only a few
> of them (usually around 10) can really grow up in the wells again.
> It looks to me a lot of "fake" colonies survive through the drug
> selection. If the recombination rate happens to be low, this
> phenomenon will make the screening very difficult to me. 5ug/ml of
> blasticidin is used. (This concentration is enough to stop WT cells
> growing and kill them under shaking condition, though).
> 2. G418 selection: for overexpression, after transformation for
> around 5 days under G418 selection, I spread single cell on KA
> plate, and then pick single colony to culture in 12-well plate. A
> lot of them can not expand in the well anymore. Again, I suspect
> some WT cells survive through the drug selection. 20ug/ml of G418 is
> used. Even for the real colonies growing up, sometimes only a few of
> them express target protein, others do not express but are G418
> resistant. This will make me picking a stable overexpression cell
> line is as difficult as screening a knock-out strain!
> 3. Colonies on KA plate: for picking single colony on KA plate,
> I found if I pick too much KA bacteria with dicty cells to the
> plate, then the well surface will be occupied by the bacteria
> (although there are antibiotics in the medium), and the dicty cells
> would not eat bacteria to proliferate. Instead, they looks "trapped"
> individually by the surrounding bacteria, and starve to die after a
> few days! If I pick much less bacteria with dicty cells, then dicty
> cells will expand! Is it possible bacteria secreting something to
> anti dicty?
> 4. Colonies size on KA plate: the colonies sizes are not
> uniform at all even for cells from the same strain. And the size is
> also dependent on the physiological stage of the cells. For example,
> old, unhealthy cells give smaller colonies, comparing with normal
> cells. So, like screening strains giving different size of colony,
> or presenting data showing KO cells giving smaller colony size, does
> not make a lot of sense to me. Am I wrong?
> 5. Chemotaxis: to show a mutant strain has chemotaxis defect,
> we would quantify the chemotaxis speed and directionality. However,
> chemotaxis is also dependent on how cells are polarized. For
> example, pi3k- cells would not show a lot of chemotaxis defect if
> they are pulsed long enough. The thing is "pulsed enough" or
> "polarized well" is an arbitrary concept. If polarizing is not
> paralleling, how meaningful to calculate the absolute number of
> "chemotaxis index"? At least we should not expect to get the same
> number from different labs, right?
> 6. Loading control of northern blot during development: to show
> a gene expression changes during development, usually we show an
> mRNA northern blot normalized to "total RNA". However, we all know
> that the cells size shrinks a lot during cells starvation and
> development, and the "total RNA" also decrease a lot in a cell. The
> real question in our mind actually is "how this gene changes in a
> cell during development", or "how this gene changes in a aggregate/
> slug/fruiting body during development". So I think normalization to
> total RNA is not accurate at all. Should we normalize RNA-seq of
> dictyexpress to "cell number" instead of "total RNA"?
>
> Thanks, Xin-Hua Liao
--
================================================
Dr. Thierry Soldati, Senior Lecturer
Department of Biochemistry
University of Geneva
30 quai Ernest Ansermet, Sciences II
CH-1211-Genève-4, Switzerland
Tel: +41-22-379-6496
Fax: +41-22-379-3499
email: [log in to unmask]
http://www.unige.ch/sciences/biochimie/Research/Soldati.html
================================================
|
