that the emission wavelength of the donor fluoro–
chrorne must be excitatory to the acceptor mole–
cule. Secondly, the acceptor molecule must be in
close vicinity (
<
l
O
nm) to the donor fluorochro–
me since the rate of energy transfer is inversely
proportional to the sixthpower of the distance bet–
ween the donor and acceptor molecules. A con–
ceptual illustration is shown (Fig.l ).
If
FRET oc–
curs the intensity of donor fluorochrome is quen–
ched (double labeled cells) campared
to
donor flu–
orochrome alone in channel2. Conversely, the flu–
orescence intensity of acceptor fluorochrome is
increased (double labeled cells) as campared to
acceptor fluorochrome alone in channel 3. To ob–
tain the efficiency ofFRET, it is enough to calcu–
late donor emission quenching according to the
equation below (Koksch et al. 1995).
B
Figure
l:
Fluorescence resonance energy transfer
(FRET). In (A), both the donor and the acceptorfluoro–
chromes are closebyandFREToccurs. In (B) thedonor
and acceptor fluorochrome molenties are distant and
FRET does not occur. FRET is inverse/y proportional
to the sixth power of the distance between the two fluo–
rochromes. Also the acceptorfluorochromemust be ex–
citable by the emission wavelength ofthe donor hut not
directly at the excitation wavelength of the laser.
E =Efficiency, D =Donor,A=Acceptor
1
2
=Fluorescence intensity in channel 2
1
20
+A
= Fluorescence intensity of double !abe–
led cells in channel 2
1
2
A
= Fluorescence intensity of acceptor !abe–
led cells in channel 2
3.2 Key issues in
FRET
by flow cytometry:
FRET analysis can be useful in the broad charac–
terization and differentiation of HLA class I an–
tibodies from those directed against platelet gly-
Klinisk Kemi
i
Norden 4. /999
coproteins or even discriminate those bound via
platelet Fe receptors (Koksch et al. 1995). The an–
tibodies against glycoproteins couldbe againstgly–
coproteins such as GPilbliila or to their polymor–
phisms (HPA's). Typically platelets are incubated
with the patient's plasma (containing the antibo–
dies) tagether with unlabelled murine antibodies
with known specificity (e.g. GPIIb/IIIa). In a se–
cond step themurine antibodies as weil as the hu–
man antibodies are detected using secondary an–
tibodies labeled with the donor (PE) or acceptor
(Cy5) fluorochromes.
If
FRET occurs one can as–
sume that the patient's antibodies react at an epi–
tope very close to the added murine antibody and
thus the identity of the human antibody can be in–
ferred.While the above techniquemayappear
cam–
plex it is worth noting that in most instances cal–
culating FRET efficiency is not necessary since a
visual examination of the superimposed (donor
fluorochrome alone andwith possible quenching)
intensity curves is adequate.Koksch et al. indica–
te that direct staining gives essentially similar re–
sults as with the indirect staining method and in
addition quantitative information regarding the
antigen expression is also obtained. An important
consideration is the donor-acceptor fluorochromes
that can be used. One of the best is the PE/Cy5
combination since the spectral overlap can easily
be corrected in the presence ofFRET. Another
consideration is the possible cross-reactivity ofthe
goat anti-human andgoat anti-mouse immunaglo–
bulin sera when using indirect methods. This
shouldbeminimal with appropriate choiceofsera.
The authors also claim suitability for cross mat–
ching of stored platelets without loss of activation
dependent antigens in thepresenceofsuitable buff–
ers.
4.1 Other clinical platelet assays:
For the sake of completion it is worth mentiorring
the use of FCM for diagnosis of heparin induced
thrombocytopenia (Gordon, 1997) as well as in the
analysis ofstoragepool disorders (Wall et al, 1995;
Thom et al, 1995). These conditions are however
more frequently diagnosed using other assays.
In conclusion, some of the current thinking on
clinical FCM ofplatelets alongwith the possibili–
ties and limitations has been presented.
95
