Hello,
as usual, the responses and help from the
list were invaluable, and made all the difference. Many, many thanks to
all who responded. The ability to sort reliably without having to spend
so much time, effort and money troubleshooting really makes a
difference. The last sort went very well, with much better recovery and
better state of the cells. Not only did the machine count agree to
within 30% (or better) of the count after washing the cells, but after
re-culturing the cells for 24 hrs the counts also kept at ~80-85% of
the number of seeded cells, so the cells recovered were in a rather
good shape (post sort count had 5-10% trypan blue positive).
What we did is this: Kept the cells in PBS + 10mM HEPES 20% FBS presort, and
sorted them into polypropylene 15 mL tubes, which were filled to 6-7 mL
with PBS + 10mM HEPES + 20% FBS. The sheath fluid was plain PBS. Before putting
the tubes in the sorter, I kept them inverted for some 30 minutes so
that the higher part of the tube would be "covered" with the FBS
containing medium. We sorted at 20 PSI (15 PSI was too unstable), 100
micron nozzle, flow rate 2-3. All tubes at all times kept cooled. We
steered the streams so they would fall on the liquid and not the tube
sides (or at least that's how it seemed). I wanted to keep the sort as
short as possible, but the first 1:30-2:00 hrs the ARIA gave us some
troubles, we switched pressures back to 20 (and laser delay), cleaned
the nozzle, changed the o-ring, etc, i.e. "the long drill", but at
least after that it stabilized and worked very well. I do wonder what
will happen to the sterility (or rather, asepticity) of the result, not
that we had much choice regarding what to do. The mask was 8/16/0. If
you're wondering who the other person in "we" is, it is Dan, the flow
core operator (who is of course actually doing the things I say "we"
did). This post would be incomplete without a huge "thank you" to Dan,
who not only goes out of his way to help, he is always cautiously
optimistic in such a cheerful way!
I have added below a summary of all the
responses I got.
Thanks again,
Uriel.
--
Uriel Trahtemberg, M.D. M.Sc.
PhD student
The Laboratory for Cellular and Molecular Immunology
The Hebrew University - Hadassah Medical Organization
Jerusalem - ISRAEL
I am not a totally useless... at least I serve as a bad example.
----------------------------
Uriel,
Ive seen the same thing. On BD's "Purity" sort mode (32/32/0) we get
anywhere from 40% to 80% of what it says it has sorted. We now often
use a 16/32/0 mask and have had much more accurate post-sort counts.
-----------------------------
Hi Uriel-
I wish all posts to the list were as complete as yours!
Two or three things I can suggest might be worth investigating:
collection
(catch) tube type, sheath fluid, and loss (waste) analysis.
I had a customer with claims of low yield (i.e. disagreement between
what the
electronics said, and what he counted in his tube afterward). The
resolution
was to use polypropylene catch-tubes, with FBS with or without a
non-buffered
saline. The sad truth is, some cells stick quite aggressively to
polystyrene.
Also, there are assertions by people wiser than I, that phosphate
buffers
(sheath solution/PBS) and carbonate buffers (RPMI, DMEM) can cause
precipitates,
lethal to some more "fragile" cells.
I have customers who run blood bank (normal) saline as sheath, to avoid
issues
such as these. If you must use RPMI or other culture buffers in your
catch
tube, you can run the same as sheath fluid - there are companies that
provide
sterile, phenyl red-free RPMI, DMEM, and various other buffers!
On MoFlo (Legacy or XDP), one has the ability to divert the waste
stream to a
catch-tube, thus enabling true loss analysis. I don't know if this is
possible
on the Aria, but I would try it if there were a way.
I hope this helps.
-----------------------------------------
Uriel,
I do have experience sorting human DC's, and yes indeed, they are
fragile.
I use a 100um nozzle at 20 psi, but unlike you, I NEVER sort above a
flow
rate of 4.0, regardless of the threshold rate. The reason is due to the
pressure differential (and higher shear stress) to the cells within the
HPLC tubing. My suspicion has always been that the higher the flow rate,
the higher the pressure differential (just going on pure physics here-at
least from what I remember from university!)-I have tried to get exact
numbers on this from BD Biosciences, but they have never been able to
give
me straight numbers.
Also, I do not know what you are referring to when you indicate it is
"next
to impossible" to align your side streams to sort into the middle of
your
collection tubes without hitting the sides. It is very easy-as I do a
test
sort, I open the sort chamber and look right down from the top to ensure
they are aligned into the middle of the tube. This becomes a little
harder
when you are sorting extremely rare populations into eppendorf tubes,
but
should be no real problem into straight facs tubes.
In addition, I would use polypropylene facs collection tubes as opposed
to
polystyrene for higher viability. I agree with you on washing the
collection tubes in FBS.
Also, you did not indicate any markers/fluorochromes that you may be
using.
Are you simply sorting on viability (Sytox Blue) and that's it? What is
your pre-sort viability (percentage) according to the Sytox Blue stain?
If
it's less than 20%, then for certain you will have difficulty in keeping
the cells alive post sort.
Hope this helps.
-----------------------------------------------------
Hi Uriel
I usually have the people using my core block the 5 ml BD Falcon tube
with 4 mls of their media they wish to sort the cells into, no more then
a 30% FCS concentration. If I remember correctly, a person at a job I
was at did a test and found that to be optimal. If they can't use serum,
I have them use whatever media it is in the coating process of the
tubes. I have them do it for at least a couple of hours at 4 degree but
tell them overnight is better. I have them remove 3 mls, leaving 1 ml
at the bottom for the cells. Unfortunately, with the way the tube
holders are in the Aria, your cells, until the media hits a certain
level in the tube, will hit the side of the tube to some degree. You can
always try to leave more volume in the collection tube if that is a
concern to you.
I don't know how effective my strategy is (would like to think it's
great) but most where I am at don't work with your specific cell type.
The closest I get is with Macrophage/grans and I have had no complaints
with any counts being that far off.
Good Luck
------------------------------------------------
Hi Uriel,
I
worry about sorting DC cells in PBS into a carbonate buffer such as
RPMI.
Question: What is the sample buffer, and is there any HEPES in it or the
catch media? Besides buffering, HEPES appears to help many cells through
the process by helping to stabilize the membrane (I am not sure anyone
has documented just how this works). Generally, most sorts of this kind
do well using Dulbecco's PBS (DPBS) or HBSS + 10 to 25 mM HEPES +
protein
(sometimes dialyzed serum, sometimes BSA 0.5 to 2 %) as a base sample
buffer.
Generally, carbonate buffers are poor sample buffers without adding
HEPES
and EDTA (phenol minus, of course). Similarly, sorting into RPMI can
cause
trouble if the pH is not optimal, since one is essentially mixing a
phosphate
buffer with a carbonate buffer with fragile cells present.
Best,
---------------------------------------------------
Dear Uriel,
I sort DCs on an Aria at 60 psi, frequency ~88K, 70 um nozzle.
I have a custom sort mask Yield 0, Purity 32, Phase 0.
My event rate is <20,000/sec (preferably 16-18,000) Efficienct
should by >85%.
To avoid issues about slamming cells into walls of tubes collect in
15ml blue top centrifige tubes, (round bottom, 10ml tubes with yellow
lids fit to). The angle of the larger tubes, plus their larger
diameter makes it easier to get the sorted cells hitting the media (I
recommend ~3ml of whatever makes the cells happiest).
Recovery is usually ~80% of the instrument count, you can get a high %
of dead cells (50%).
Try to keep your sample concentration high - 10e7-10e8/ml, ideally your
"rate" should be 5 or less, but I am usually happy with <7. The
sample should also be in whatever media makes the cells happy (although
indicator free is prefered).
I have also been successful at 35 psi, 70 um nozzle. Same sort mask.
But the event rate is ~10-12,000/sec.
Arias seem to sort best at an event rate of <1/4 of the frequency.
Hope this helps.
Regards
----------------------------------------------------------
Hi Uriel,
Here are a few ideas:
1] Your instrument parameters seem fine. My only suggestion in this
regard would be to try and concentrate your cells more in the pre-sort
tube and see if you can run at a lower rate, i.e.e keep it below 4.
You'll have to check how the DCs respond at a higher concentration.
2] Since these are fragile cells, why not sort into 50 ml tubes? And
make sure the stream doesn't hit the side. This would also allow you to
sort into a large volume of medium which might imporve cell viability.
3] make sure your pre-sort count is done right before sorting. I have
experienced where the count is done at the early stages of processing
and what is in the tube is actually a lot less.
4] I would look at the concentration of serum in your pre-sort RPMI, I
have seen cells actually clump in the pre-sort tube during the sort.
5] Maybe try and vary what medium you are sorting into, e.g. RPMI with
different concentrations of serum.
Good luck,
------------------------------------------------
What is the pH of your RPMI? As you do not control CO2 you would need to
stabilize is somehow, probably RPMI-hepes. Whilst serum adds essential
proteins and reduces stickiness it is also based on CO2 (bicarbonate)
buffering
Sticking the stuff on ice will also effect your pH
In addition you have to test your osmolarity depending if it is designed
to be used with or without the addition of bicarbonate.
When I was trained on our Aria I thought that the wisdom was not to go
above
an event rate of 5.
----------------------------------------------------
When my cell counts were at odds with the sort count, it was most often
due
to the users counting after spinning (as you say you're doing) or
because
they were sticking to the tube (you don't say which plastic you're
using) -
try using polypropylene tubes, to which most cells have a hard time
adhering
(I've found these to be much more useful than serum-coated or silanised
glass or polystyrene)- and count the cells before you spin them down -
spinning doesn't always go to completion even though you see a pellet
there
may well be cells in suspension also you may well inadvertently
resuspend
and discard a fraction of the pelleted cells (and in polystyrene tubes a
proportion of the cells may be forced into sticking to the bottom of the
tube). It's simple to test if your loss is due to centrifugation - just
dilute some cells to the sorted concentration, pipette a known number
into
some tubes, spin them and decant as you normally would after a sort,
then
count them- if the loss is similar to what you've observed, voila!
Count the
supernatant too, the missing cells are likely to be there, if not then
use
microscope to see if they're sitting on the surface of the tube.
-------------------------------------------------------------
Received on Wed Jun 25 13:58:00 2008
This archive was generated by hypermail 2.1.8
: Wed Jan 31 2007 - 03:12:00 EST
