Calcium
Phosphate Transfection of Neurons in Primary Culture
Greenberg Lab
March 2001
(617) 355-4361
anne.west@tch.harvard.edu
I.
Reagents/solutions
1.
transfection media:
DMEM: Gibco/BRL
#11960-028
- we do
transfections in plain DMEM with no added serum, glutamine, or pen/strep
- imperative that DMEM is not too old. if
the color has gone pinkish, do not use. We use a new bottle for about 2 weeks,
then start fresh one. This definitely matters.
- the lab has also worked out protocols
for transfecting in MEM, which is a HEPES buffered medium, and therefore
maintains better pH in air (DMEM is carbonate based, and so works only in CO2
incubators). You can transfect in MEM, but times are much longer, and overall
transfection is less efficient. Unless there is an imperative reason to use
MEM, we recommend DMEM.
2. 2X HeBs
The quality and
accurate pH of this solution is of the utmost importance for efficient
transfection. We use the highest quality chemicals, and keep them separate from
our normal dry chemical stocks. These are labeled tissue culture only and never
are any lab implements allowed into the containers.
final
conc for
200 ml supplier
NaCl 274
mM 3.2
g Baker
# 3624-05; Mallinckrodt
KCl 10
mM 142
mg Mallinckrodt
# 6858
Na2HPO4.7H2O 1.4
mM 76
mg Mallinckrodt
# 7914; 268g/mol
dextrose
(D-glucose) 15 mM 540
mg Baker
# 1916-01; 180g/mol
Hepes (free
acid) 42
mM 2
g Calbiochem
Ultrol # 391338; 238g/mol
- add components to 180 ml. water (we use
only ddH2O straight from the purifier, and only glassware that is tissue
culture only, never having seen soap)
- pH with 1
or10N NaOH (we make special NaOH only used for this)
- pH very
carefully to 7.02 – use most accurate pH meter you have.
- bring to 200mL
with ddH20
- remove 25 mL and
save
- pH remainder
to 7.04 (with 1N NaOH); remove 50mL and save
- pH to 7.06;
remove 50mL and save
- pH to 7.08;
remove 50mL and save
- pH to 7.10;
this is the last 25 mL
- filter
sterilize each pH aliquot separately in cell culture hood
- aliquot to 1ml
aliquots (eppendorf tubes)
- store at -20C; thaw individual aliquots
as used. I usually do not reuse unless for nonneuronal cells.
- note, as for any CaP transfection,
greatest efficiency is obtained with a fine, ÒsandyÓ ppt. This is critically dependent on the pH
of the HeBS, so it is best to:
-
use an accurate pH meter
-
standardize the pH meter (with pH 4.0 and 7.0 standards) repeatedly, until standards
are read precisely
- as indicated above, make and test
multiple batches of 2X HeBS, with slightly different pHÕs (eg., pH 7.02-7.10);
test each, use best, toss the rest, or use for nonneuronal cells
- seems to be
good to make fresh 2X HeBS at least every few months
3. 10X CaCl2
Make
2.5M in ddH20. Filter sterilize and store.
-
this does not seem to be one of the finicky steps.
1. replace
culture media with plain DMEM for transfection
- prewarm DMEM to 37¡ for at least 20Õ
before putting on cells
- remove conditioned culture media;
SAVE!!!!, to return to plates after transfection. It is bad to put fresh medium
on the cells after transfection, will promote cell death
- wash cells 2X
with plain DMEM transfection medium. Be quick, donÕt let cells dry out.
- add volume for
transfection itself:
well
of 24-well plate: 250ul
12
well: 0.5ml/well
6
well: 1 ml/well
100mM:
5mLs/plate
- return plate
to 5% CO2 incubator until precipitates are ready. This should be within 30-60Õ.
2. make
calcium phosphate/DNA precipitate:
a. variables
- volume of ppt
24
well: 12.5-30ul
12
well: 25-100ul
6
well plate: 50-200ul
100mM
plate: 300-1200ul
- amount of DNA
:
24
well: 0.5-4ug (eg., for immunostaining)
12
well: 1-4ug (eg., for luciferase assay)
6
well: 2-5ug (eg., for RNAase protection)
100mM
plate: 5-20ug (RNAse protection/expression)
- for each cell type, you should test ppt
volume, DNA amt, and time of transfection to optimize. For cortical and
hippocampal neurons, I generally use the low end of the volume range (25uL on
12 well plate for my luciferase assays) and intermediate DNA amounts (For each
well of a 12 well dish, I have been using 0.5ug of a Gal4 UAS firefly
luciferase reporter plasmid, and about 1ug Gal4-transcriptional activator, with
0.01 ug of a renilla luciferase reporter plasmid for a transfection control).
By contrast in cerebellar granule neurons we use 30uL ppt for a well of a 24
well plate, with up to 4ug of DNA transfected per well.
b. recipe: eg.,
for 25ul precipitate/well (scale up or down accordingly):
- generally we do triplicates of any
condition, and I make the precipitates for all three wells in one tube
together, then distribute onto the three wells. Some people in the lab make
each precipitate in separate tubes. This increases uniformity of transfection
(each successive pipette dip into the tube seems to reduce the ppt formation
just a little bit) but increases the complexity of setting up the precipitates.
- I have written the recipe below for an
example using 12 well plates with 25uL total ppt per well, and IÕve made enough for 3.5 wells
– always make .5 wells more than you need, as pipetting is rarely exact
(I speak for myself at leastÉ)
1) in a 7.5ml
polystyrene pop-cap tube (Falcon 2057), mix DNA, ddH2O, and CaCl2 .
1.75ug
Gal4-luciferase reporter (0.5ug/well x 3.5 wells)
3.5ug
Gal4-transcription factor (1 ug/well x 3.5 wells)
35ng
renilla luciferase plasmid (10ng/ well x 3.5 wells)
ddH2O
(tissue culture sterile, direct from purifier) to 35uL
8.75uL
10X CaCl2 (2.5M) (1/10 final
volume)
(total
volume = 43.75uL)
2) aliquot an
equal volume 2X Hepes-Buffered Saline (HeBS) to second tube for each ppt
(43.75uL, in this case)
3) thoroughly
mix DNA/CaPO4 mixture and add to 2X HeBS dropwise with pipetman.
There are many schools of thought on
exactly how you should make this mixture. Some people bubble the DNA through
the HeBS, some swirl the tube. I am in the camp that holds the pipetter with
DNA mix above the HeBS and then just slowly drops one drop at a time on top of
the HeBS. Different mixing styles give different size precipitates that will influence
the amount of ppt and time of transfection you should use. Therefore, each
person must optimize these parameters his/herself since everyone mixes slightly
differently.
4) let ppt form
in dark, 25Õ, room temp
For
neurons, we always let the ppt form for longish periods of time before putting
on the cells. For nonneuronal cells, however, mixing the ppt then applying
directly to cells seems to work best. If you are in the mood to optimize you
could test for your cells the effects of changing the time of ppt formation.
As
to forming the ppt in the dark, we know of no good reason why that is supposed
to matter, though all books say that it does. We frequently form ppts in the
light and have not noticed a difference.
3. Add ppt to
plates
- drip evenly
over surface, with pipetman
- careful as you
return pipetter to tube for each aliquot. Try to minimally disturb ppt.
- return plate
to 5% CO2 incubator
- leave ppt on
cells 20 -40 min
The
time of transfection is one of the most critical variables. This time varies
widely between cell types. Cortical and hippocampal neurons are more sensitive
and will transfect within about 20Õ. Cerebellar granule cells require longer
transfection times, up to 30-40Õ. If you are doing the transfection for the
first time, you should definitely try a range of times. At times too short you
will see no transfection, at times too long, the cells will die.
You
can follow the transfection by following the appearance of the precipitate on
the cells. It should be sandy, not found in large clumps, but this is difficult
to describe. Most useful is if you watch the appearance of the precipitate on
your cells as you conduct the time course, then once you view the outcome,
match your memory of the ppt appearance to the optimal transfection time, and
always make it look like that.
4. Stop
transfection
- aspirate media, wash twice with warmed
plain DMEM. Cells may be slightly loose at this stage, so wash carefully,
perhaps pulling medium off with a pipetter rather than an aspirator. Send wash
medium down the side of the well not directly onto the cells.
- add back conditioned media (hopefully
you saved this!!) If you are slightly short for the number of well, you can top
off with a little bit of fresh medium.
- optional: to the conditioned medium,
add extra glucose at the level of 0.6mL of 45% glucose to each 50mL conditioned
medium. The extra glucose is especially helpful at promoting the survival of
cerebellar granule cells after transfection.
1. Amount of
precipitate (ppt) per plate, length of time ppt on plates
- The most
critical parameters for efficient, nontoxic calcium phosphate (CaP)
transfection of primary neurons are the amount of CaP/DNA precipitate added to
the cells (per volume of media), and the length of time the precipitate is left
on the cells. The length of time
the precipitate is left on is decided based on its efficiency of
formation/accumulation; this is affected by a number of factors including the
pH of the HeBS, the quality of the DNA, and the transfection media composition.
The above parameters should yield a layer of precipitate, starting at 15 -30Õ ,
and becoming heavy by 30 - 60Õ.
The neurons tolerate a heavy layer of ppt., and apparently are preferentially
transfected by such a short exposure (15-60 min), compared to glia.
- With too
little ppt, transfection efficiency is low; with too much, toxicity occurs
(either immediate, or with one day delay); therefore, for a new neuronal cell
type, media conditions, or culture age, we first do a pilot transfection, eg.,
with three ppt volumes, and stopping the transfection after 10-40Õ. We typically transfect a GFP expression
vector which can be well seen within 12-20 hours to assess transfection efficiency. We achieve approximately 1% - 5%
transfection efficiency.
- During the
transfection, plates are checked periodically to assess ppt accumulation and
possible toxicity; the transfection is stopped early if an unusually large
amount of ppt has accumulated, or if there is any toxicity (though this is hard
to visually assess).
- Usually, a
layer of CaP ppt remains on plate even after washes; for cells cultured with
serum, this disappears with time, with no apparent ill effect on cells.
HOWEVER, for cells cultured without serum, leaving too much ppt on the plate
after transfection can be devastating. When cells are returned to their
serum-free medium, the ppt may continue to accumulate O/N and next day all the
cells will be dead. If you grow your cells serum-free (neurobasal with B27
supplements, for example) make the last wash before returning to conditioned
medium in some other medium that contains serum (for example, DMEM with 10%
CS). Leave the cells in this medium for 5-10Õ before washing it off and
replacing with the serum-free conditioned medium. This serum wash seems to kill
the precipitate and protect the cells.
For your first
pass, I would recommend something like the following test:
on 24 well
plates
pH 7.02, 7.04,
7.06, 7.08, 7.10
for each
condition use 1ug GFP/well
for each pH make
total precipitates volumes of 12.5uL, 15uL, 20uL, 25uL
for each pH and
volume condition try leaving ppt on for 10Õ, 15Õ, 20Õ, 25Õ, 30Õ
Admittedly that
is a total of nearly five complete 24 well plates. But you will have your
answer! Once you have optimal volume, pH, and time, then go back to alter total
DNA and potentially time of ppt formation.
2. The wash with
transfection media before transfection seems to be important for removing a
residual media component that inhibits precipitate formation/accumulation. This component may be serum, therefore
the washes must be in serum-free medium.
3. Channel
inhibitors
- for at least
some neuronal types (in particular, more developed cortical neurons), the
inclusion of kynurenate and APV in the protocol appears to reduce
toxicity. However, these
inhibitors are not essential, particularly if the cells are not very sensitive
to excitotoxicity, and they can also reduce transfection efficiency. Under some conditions these reagents
may be useful, but in the majority of cases, we have achieved better
transfection without them.
4. Preparation
of DNA
Though original
work with the protocol was done using DNA prepared by double cesium banding, we
have repeatedly proven in our lab that DNA prepared on columns is equally efficient
for transfection, and may be somewhat less toxic (perhaps due to cesium or EtBR
contamination of banded DNA). We have not found improved transfection or
reduced toxicity using more expensive Òmycoplasma freeÓ columns, and have even
achieved good transfection with quick mini-column prepared DNA.
Some aspects of
DNA preparation do seem to matter however. It is important to have a prep that
is relatively free of protein. This can be achieved by following the
manufacturers directions for the columns and not overloading them with
bacteria. For most high copy plasmids, 100mL of an overnight culture should be
sufficient to give maximal DNA yield (500ug) on a MAXI prep column. If your DNA
prep contains high protein levels, you can phenol/cholorform and ppt the DNA.
Make sure to do 1-2 chloroform washes before precipitation to fully remove the
phenol as residual phenol can be toxic to the neurons.
DNA
concentration can also affect your transfection. This is because irregularities
in your DNA prep will matter more if the DNA volume constitutes a larger
percentage of the total transfection mixture. Therefore we try to keep our DNA
concentrations at least 1ug/uL so that they do not constitute a large
proportion of the total ppt volume.
Finally, many
column preps suggest that you resuspend your DNA in TE, while many of us will
resuspend instead in ddH2O. Because of the buffer in TE, this could have a
small effect on the exact pH of the final precipitate. Given that pH is a very
sensitive factor influencing transfection efficiency, if you do resuspend your
DNA in TE you should make sure that the total volume of TE is exactly the same
in every precipitate you make for an experiment. When calculating the recipe
for the DNA/CaCl2 mixture, just normalize all your DNA for TE at the minimum
level possible, then top off with ddH2O to the final volume. Alternatively,
resuspend your DNA in ddH2O and store at –20¡C to enhance stability.
5. Cell culture
- this CaP
protocol has been used successfully for primary cultures of rat neurons from
cortex, hippocampus, striatum, spinal cord, and cerebellum
- for cortical
cultures from E17/18 fetuses, or newborns, this protocol has worked well for
cells on or after 3DIV (day of culture being 0 DIV); if done on 1 or 2 DIV,
there can be high toxicity and low transfection efficiency. Transfection
efficiency seems to fall off after about 10DIV, and can be quite low after
14DIV.
- cell density
at seeding: cerebellar
(P4-6)
24
well: 5 X 105 cell/well
60mm: 5 X 106 cells/plate
For
cortical/hippocampal neurons used for transcription assays we usually use 12
well plates with 18 x 106 cells/dish (so 1.5 x 106 per well). We scale
proportionately for 24 well or 6 well. 24 well plates work less well for some
reason and 6 well use too many cells. One well of a 12 well plate will be lysed
in 100uL lysis buffer for luciferase assay, and 20uL of the solution is used
for the luciferase measurement.
For 100mM plates
for RPA, we use 10-12X106 cells/dish.
For cell
staining, we have been able to use CaPO4 transfection on cells grown at lower
densities in serum-free supplemented medium, but only to densities of about
20,000/24 well plate. At very low densities (Banker cultures, isolated cells)
CaPO4 transfection is too hard on the cells and they lift off.
- growth medium:
this protocol has worked for cells grown in a variety of different culture
media; however, the particular culture conditions can change the optimal day of
transfection. In addition, please see the notes above about washing off the ppt
if cells are grown in serum free conditions. For transcription assays we grow
cortical neurons in BME with 10% calf serum, pen/strep, and glutamine. We
transfect 4-6DIV and assay 2 days later. Note, for depolarization activation of
transcription, cells respond better when they have been grown in serum rather
than serum-free, potentially because synapses develop slower with serum,
keeping endogenous activity low, enhancing response to exogenous
depolarization.
IV. Ref :
Xia, Z, Dudek,
H, Miranti, C and Greenberg, ME J. Neurosci. 16, 5425-5436, 1996.