What exactly is the OD problem? (fwd)

Robert Preston rapr at MED.PITT.EDU
Fri Jul 16 22:00:16 EST 1993

>     there is a person in my lab who insists that different specs
> should give the same OD reading as any other machine, assuming that
> they are in working order.
>    So what is the exact criticism of OD? Was it specific to measuring
> cells?
> DanZ
To which I absolutely finally just HAFTA jump into this thread:
Yes, Dan, it IS specific to measuring cells, or other light-scattering
(rather than light-absorbing) suspensions of particles.  Spectro-
photometers and colorimeters compute absorbance (=optical density, OD)
as the negative base 10 logarithm of the transmittance, the latter
being the ratio of the _detected_ transmitted light to the _detected_
incident light.  The _detected_ incident light is determined on a
non-particulate reference sample, where, to a very good approximation,
the actual incident light is equal to the _detected_ light at whatever
detector is used in the instrument.  Now, with turbid suspensions,
the _detected_ transmitted light is as much a function of the instrument
optical geometry as it is of the cell or mass concentration in the
cuvette, since the turbid suspension scatters the incident light
mostly in the general direction of the detector, but not in a straight
line, as is the case with true solutions.  As a result, the _detected_
forward-scattered light varies greatly in different instruments that
have, say, different areas of detector surface or different slit-widths
in front of the detector.  This can make a HUGE difference in the
"OD" reading of the same identical suspension as reported by different
instruments.  This is totally unlike the case when true, absorbing
solutions are being assayed, which will give similar readings on
different machines (identical readings, in an ideal world) as suggested
by your lab-mate.  It turns out that many instruments happen to give
similar (within a factor of two) readings from cells at the same
density, since, by accident of design, the geometries in the various
instruments are roughly the same.  But if you're working with (or
trying to duplicate results from) an instrument that happens to have
a weirdo design (and engineers will do what they want to do, of course)
then you can be a factor of ten or more off from where you think you
are, in cell (or biomass) concentration, and end up blaming a protocol
rather than mutual ignorance of the meaning of the "OD" of turbid

That said, there's still wisdom in Fatima's comment that, in many
cases, the best way to get the right cell density is to look through
the flask at the trees or whatever, and rely on a lot of hard-earned
seat-of-the-pants lab intuition.  Unfortunately, many people, not
merely beginners only, neither have nor want to develop that kind of
intuition, and so someone got the bright idea (?) of using spectro-
photometers to do turbid suspensions.  To each his/her own, I guess,
but inter-lab reproducibility sometimes greatly favors the use of
instruments used not uncritically.

Rob Preston
UPMC Dept. Pathology
rapr at med.pitt.edu

More information about the Yeast mailing list