APC Sensitivity on the MoFlo

Summary Document

This document is a summary of message from the Purdue University Cytometry Laboratories Cytometry Discussion list.

http://www.cyto.purdue.edu/hmarchiv/index.htm

QUESTION:

APC Sensitivity on the MoFlo

Thanks to everyone who offered their insight into this issue. The Flow
community is so generous with their help, I've decided to endure all those
OUT of OFFICE REPLIES yet again in order to share the responses I got. The
names of respondents have been removed in case anyone wants to protect their
privacy.

-----------------------------------------------------------------------------------

REPLIES:
-------------------------------------------------

i have a diva and also had a solid state red laser dye 90% of its intensity
went out bd shipped me another
if you have contract ask them to send you one you install it same way as
the old and compare.

-------------------------------------------------

how is your callibration going when you set everything up with beads?

-------------------------------------------------

Your problems are all too familiar. I have the exact same problems.
The fact is, an open air system just does not, and will never have the
sensitivity of a closed flow cell system.
I also suspect the power and performance of the diode laser is less than
advertised.

-------------------------------------------------

It isn't a MoFlo problem per se, it is as you suggest a laser power
problem. The FACScan collects a lot more light than any stream-in-air
sorter, so it work reasonably with lower powered diodes. While at Scripps,
we always used mixed gas or krypton lasers at 647nm - ~ 200mW for APC and
APC tandems. The HeNe and diodes < 50 mW just are not enough power for dim
immunofluorescence on a stream-in-air in my opinion. Ask those unhappy users
for the $$ for a higher power laser... you are photon limited, and they
might be able to see to it you are not $$ limited too.

-------------------------------------------------

Our lab is experiencing similar results using a 25mW diode laser on a
Cytopeia Influx sorter. Samples run on a Calibur are routinely a log or
2 brighter and this is a result of the flowcell design and the direct
path of the laser beam. On the Influx we measured the beam intensity
from source to stream and discovered that by the time the beam traveled
through prisms and bounced off a few mirrors the laser intensity dropped
from 23mW at source to 7mW at the stream. Not much power left to excite
APC labeled cells in a jet in stream sorter. Cleaning of the optics made
no difference. One improvement that did increase separation of neg/pos
cells was to change the laser beam from a circular beam to an elliptical
beam. This resulted in a 1 decade improvement.

-------------------------------------------------

When I had this problem a while ago it was because of the actual
calibration of my red laser. Some more tweaking and it was great. The
funny thing is that it would only happen on certain stains. To solve
this, we used 8 peak Red beads in our calibration. If there is a
"hockey stick" curve to the 8 peaks, then it is still mis-aligned.

Hope that may help you out.

-------------------------------------------------

Also, get users to titer their antibody on the instrument they plan to
use for sorting. Most Calibur users get away with using less antibody
per tube and expect all instruments to generate the same results. This
is not the case.

-------------------------------------------------

We have a HeNe on our MoFlo. Our MoFlo is one of the first out in the field
so initially we had a PMT that was not as optimum (Hamamatsu-08). We
changed all our PMTs to be more sensitive in the red (-15)and this helped.
Is your MoFlo setup with -15 PMTs? But with that said, we still have better
signal to noise on the FACScalibur and our CyAn than on the MoFlo. It is
just the way it is with jet-n-air sorters. You can optimize your setup with
better PMTs and/or filters, but the cuvette flow cytometers will always out
perform. Could your user use a PE labeled antibody? On our MoFlo, PE is
the best fluorochrome.

Good Luck,

-------------------------------------------------

You'll likely get very similar response from several people, but I did want
to get one out to you.

We have 2 moflo at MSKCC, both running water cooled mixed gas lasers for red
excitation (647nm I believe). We run these at 150mW to get decent separation
in APC. On our DiVa, we had a 35mW HeNe that was losing it's ability to
separate APC, so in the next month I am replacing it with another Coherent
i305 water cooled mixed gas.

It is possible that the Diode you got from Dako is bad. they had some
problems with a couple in the CyAns people here at MSKCC have. Call your service guy.
There is a new Diode from a company called Diode Laser Concepts which is
supplying very high power lasers.

So in summary the Dako sources diodes are suspect, and they are probably too
low power for jet-n-air anyway. You could email Cytopeia and ask what they
use, since they are HUGE diode fans.

-------------------------------------------------

I had your same fustration last year with mine. As you correctly
concluded, stream-in-air sorters are not as sensative as benchtop
analyzers. I wish that it wasn't that way, but that's the price that we
pay to be able to sort using that technology. I've found that for the APC
measurements, it is very important to make sure you are adjusting
alignment beads to their maximum intensity and then tighten the CV. A good
CV alone is insufficient. Also, I believe that ~10% or so of a signal is
lost going through an optical filter. Make sure that APC signals are
passing through the minimum amount of optical filters. If you have a
krypton or argon/krypton laser, tune it to its 647.1nm line and put that
in the same pathway as the 633nm line of the HeNe so you can excite APC
off of those 2 lines. You'll have to change the band pass filter to a
670/14nm. I am in the process of replacing my 35mW HeNe laser with a mixed
gas argon/krypton laser because I can get so much more power out of the
647 line and I also require the laser line selection that a mixed gas
laser affords. You may wish to get a laser power meter so you can be
certain of the level of power of your lasers. I have the pocket one by
Coherent
(http://www.coherent.com/Lasers/index.cfm?fuseaction=show.page&id=250&loc=830)
though I am sure that there are others out there that are equally as
suitable. It's comforting to know the level of power that you're getting
whenever things are being questioned. I found out the my I-90 power meter
is reading 30% less than the actual value and that my 35mW HeNe is
actually providing 40mW.
Moral of the story: I've learned to live with the situation but I have
made some modifications.

-------------------------------------------------

Put a more sensitive PMT on that detector location. I used to run a
MoFlo and saw this same thing only in our PE-Cy5 channel. There's a
nice little section in Shapiro (p.165 on PMT's) or you can pull a
couple of PMT's out of the MoFlo for comparison. If I remember
correctly, the "-8" PMTs have a higher sensitivity than the "-6" or
"-4's."

-------------------------------------------------

The diode laser on the MoFlo can be problematic (actually on any
machine) due to the nature of diode lasers. I have 2 MoFlos - one
with a 35mW He-Ne which works fine and the other with the red
diode. Initially I was unhappy with the red diode. The problem is
that a red diode puts out wavelengths other than the one you think
about and want- 635-638nm in this case. There are other longer
wavelength lines coming out. Unless you do something these lines get
into the lower end of the APC channel (APC-Cy5.5 and APC-Cy7 are not
much of a problem since they are longer) and increases the signal on
the negatives thus giving you a signal to noise problem (the laser
light is relatively bright even though these lines are much lower
power than the 635/638 line). You need to reduce the noise. How can
you do this? You can re-configure your PMT path so that you have the
emission light go through a dichroic filter before the APC that lets
these unwanted laser emissions go through (I use a 650nm SP dichroic)
to an unused PMT location. You also need to look at the APC filter
and keep it rather tight (I use a 665/20). You also should put a
bandpass filter in front of the laser before it gets to the
cells. This filter would be a tight bandpass laser emission filter
at 635-638nm. Doing these 2 things helps and we do APC all the time
with acceptable signals (e.g. CD8 gives about 2 logs separation) but
the He-Ne does better because you don't have to do these things since
from a gas laser the emission is monochromatic. I also have
purchased a Coherent Radius laser that I am going to try but it is
essentially a diode laser also. Talk to Dako and see if they can
help with this. See if they will make you a red diode lasr assembly
with the filter included. The best solution at the moment is to get
a He-Ne but that will cost more than what I state above although 35mW
He-Nes are less than $5K from Melles-Griot (Coherent used to market,
under their label, the Melles-Griot laser but no longer do).

-------------------------------------------------

Hey I have to say it is tough to be a sorter and pleasing everyone. I
had to actually chuckle, because I am having the exact same issue, where
my customer has run APC on the Calibur and gets great separation, but on
the sorter...I couldn't see signal. You say you have the MoFlo, which
is interesting in that I am using a Cytopeia Influx, and they were
designed by the same person essentially. One thing I think is my
problem is that the discrimination between the Calibur and my Cytopeia
are completely different. Maybe this applies to you? I think I am
picking up extremely small debris that is perhaps killing my total
signal?? I have thought about it and thought about it till I am blue in
the face. I don't know if you are having the same issue, but it seems
to me that I can barely get the samples with APC above the axis (which
may be something entirely different causing this). If you hear any good
ideas, would you be so kind to fill me in? Also, one problem that we
fixed and it helped in the several times we have tried these experiments
on the instrument is that our dichroic mirror was a little off. That
helped a great deal unfortunately not enough to convice them that I am
setting up my instrument right. Remember that beads may not always tell
you if your mirrors are off. That is what I was told. Well these were
some of the things I have found so far in this really long journey. If
you find anything else I would be interested to hear.

-------------------------------------------------

That is an issue with quantum physics rather than laser power. The drop off
in signal intensity is expected. I can concede about a 1/2 log differential.
Higher wavelength lower energy photons efficiency drops greater in stream in
air versus quartz cuvette flow cell, where much more of the incident photons
are conserved.

-------------------------------------------------

Yes, we have the same problem with our MoFlo. APC pmt is a H957-15 and this
is important but still we have bad resolution and everyone has given up and
uses Alexa 647.

-------------------------------------------------

This is tough when you can never get the same sensitivity as
your Calibur. I also had someone not so happy when they compared our
Calibur to our BD Vantage sorter. I'm assuming your detector is fine so
I'd try increasing the sensitivity of the APC detector even if it means
pulling the negatives out a little more, maybe this will resolve the two
peaks.

-------------------------------------------------

The Caliburs have optically coupled collection lenses which make it's
sensitivity greater than stream in air systems.

Diodes can loose power over time just like any other laser. Some do it
rapidly, then stabilize at a lower power level. An inexpensive power meter
can be had for about $300 that will read your red line up to 50mW.

Bluesky Research makes a 30mW diode package that we use for a replacement to
the Spectra Physics model 127(also a 30mW laser). We call it the LRC-2 for
"little red compatible #2". It is 635nm with a collminated beam instead of
an elliptical beam. It makes a tight, bright intersect point. It seems to
match the SP127 performance very well from what the users tell me.

-------------------------------------------------

The first conversion I observed using the Bluesky component was done by Mark
Curry at U of New Mexico on a MoFlo system. It is still in use going on 4+
years I think.

We recently converted a VantageSE for Wayne Green at U of Utah core lab
using the LRC-2 system. I have not heard any reports of poor APC
performance.

APC is a good dye on our MoFlo (I used two of them in recent years) even
with the flimsy 10 or less mW laser. I assume your laser alignment is right
so the source of the problem is either the PMT, the PMT position
(misaligned) or the dichroic in the optical path. I would go through the
optical bench alignment process with APC or some multi line beads.

You may want to measure the laser's output. This is almost unlikely to be
the source.

I had APC problems (high base line noise) because my nozzle tip was too
close to the laser beams. I was told many years ago to align the blue laser,
lower the tip until it hits the beam then pull it back just outside the
beam. This way you minimize the noise (on the blue line) caused by the
stream's vibration. If you do this the red laser (even with the beam shaping
optics) is too close. Move the tip just a smudge higher.

The PMT must be a red sensitive one "-15" model.

Is the filter the same as in the Calibur (or similar?). APC emits around the
mid 600 nm range.

Finally, the MoFlo will never be as sensitive as the Calibur due to its less
efficient light collection and much faster velocity (~ 10x shorter dwelling
time in the laser beam). That can be compensated by more powerful laser
output but still no the same. The ARIA sorter, on the other hand, is as good
or better than the Calibur.

PS I would start with twisting the PMT around a little while running beads.

-------------------------------------------------

Speaking of APC on the MoFlo...we are having a recurring problem ever
since Dako replaced a board on our MoFlo. We get a shadow on our
negative populations. In other words, the negatives are essentially
bimodal, with one population on the axis and one dimly positive around
the first decade on a log scale.
Dako is now INSISTING that this is normal....despite the MoFlo never
having done this before, nor has any cytometer I have ever used for the
past 30 years. I suspect a problem with the log amp, or some type of
baseline offset issue.

1) Anyone know what this is and how I can fix it?

2) HAS ANYONE RECEIVED DECENT FIELD SERVICE ON A MOFLO IN THE PAST FEW
MONTHS? Apparently all their field engineers who had ANY flow experience
have left...and the new guys just don't have a clue.

-------------------------------------------------

Finally, the service rep manger for Cytomation suggested that reading the
APC off the second pin hole instead of the third might increase sensitivity.
Remeber the second detector cluster runs straight back whereas the first and
third are reflected at right angles thus losing whatever light a mirror
can't reflect.

--------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------