Download Prolonged monitoring of the blood-aqueous barrier with

yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Prolonged Monitoring of the Blood-Aqueous Barrier
With Fluorescein-Lobeled Albumin
Poul G. Mitchell, Normon P. Blair, and Thomas A. Deutsch
Aqueous fluorophotometry has proved to be a useful indicator of changes in the blood-aqueous barrier
after surgical, immunologic, or laser manipulations. Previously used fluorescent tracers have been unable
to follow rapid changes continuously in the blood-aqueous barrier. Fluorescein-labeled homologous
serum albumin, however, provides extremely stable and high plasma levels of fluorescence due to active
renal reabsorption with very low levels in the normal aqueous because of its high molecular weight. This
feature allows prolonged, continuous, and highly sensitive monitoring of the blood-aqueous barrier
before, during, and after a manipulation. The usefulness of this technique is demonstrated in a model
that has been well studied with other methods: the response to argon laser iris photocoagulation. Invest
Ophthalmol Vis Sci 27:415-418, 1986
Observation of the blood-aqueous barrier (BAB) in
response to a variety of ocular manipulations has
proved to be a useful measure of induced inflammation.1"5 An increase in the aqueous humor concentration of protein or other substances can be determined
by paracentesis, but this method destroys the observed
system by causing inflammation. Although aqueous
fluorophotometry is noninvasive, it requires a suitable
tracer with a level that can be readily determined in
the aqueous and plasma. The ratio of aqueous to
plasma levels indicates the permeability of the BAB to
this tracer, which will increase with inflammation.
Sodium fluorescein is the most commonly used
tracer because of its availability, low toxicity, and high
fluorescence. However, its small molecular weight
(MW) allows it to pass through an undisturbed BAB
into the aqueous in substantial amounts. Furthermore,
its high rates of elimination and metabolism result in
rapidly changing plasma concentrations. These problems can be minimized somewhat by using a standardized dosage and determining aqueous fluorescence
after a given interval of time. This procedure is completed before and after a manipulation, with the change
interpreted as the effect of the manipulation on the
BAB. Because this method requires a complete elimination of fluorescein between measurements, some
protocols compare the aqueous fluorophotometric
measurements between the treated and fellow untreated
eye to control for variations in plasma levels of fluorescence.6'7 This modification must be used cautiously,
however, as bilateral BAB reactions to unilateral stimulations have been reported.8
Fluorescein isothiocyanate (FITC)-labeled dextran
has a higher MW, with resultant lower levels in the
anterior chamber and a slightly longer plasma half-life
than sodium fluorescein, but both tracers have plasma
kinetics that are too complex to assume a steady-state
condition in either plasma or aqueous. Therefore, neither can be used to follow rapid changes in the BAB
continuously. FITC-labeled homologous serum albumin, however, provides extremely stable and high
plasma levels of fluorescence due to active renal reabsorption and low metabolic turnover while maintaining
very low levels in the normal aqueous.
The purpose of this report is to demonstrate the
ability of this fluorescent tracer to monitor the BAB
continuously in a sensitive and prolonged fashion. A
simplified method for the synthesis and purification of
the FITC-albumin conjugate is detailed. The usefulness
of this tracer is demonstrated in a model that has been
well studied using other techniques: the response to
argon laser iris photocoagulation.'~7>9
From the Department of Ophthalmology, University of Illinois at
Chicago, Chicago, Illinois.
Supported in part by: Grant EY03106, Training Grant 7038, and
Core Grant 1792 from the National Eye Institute, Bethesda, Maryland
and by an unrestricted research grant from Research to Prevent
Blindness, Inc., New York City.
Submitted for publication: July 8, 1985.
Reprint requests: Norman P. Blair, MD, 1855 W. Taylor, Chicago,
IL 60612.
Materials and Methods
Preparation of rabbit albumin-FITC conjugate (RAFITC) is adapted from previous reports.10"13 Rabbit
albumin, Fraction V (Sigma Chemical Co.; St. Louis,
MO), 500 mg, is dissolved in 5 ml of 0.2 M carbonatebicarbonate buffer, pH 9.4. To simplify the handling
of the fluorochrome and to aid in its dispersion, FITC
Downloaded From: on 05/13/2017
time (hours)
Fig. 1. The plasma concentration of fluorescein isothiocyanate labeled rabbit albumin following intravenous administration of 125
mg/kg body weight in the pilot study (squares) and repeat study (diamonds), six rabbits each. During the first 48-72 hr the labelled intravascular albumin equilibrates with the unlabeled extravascular albumin. The prolonged subsequent decrease is due to the slow metabolic turnover of native and labeled albumin. Substantial
fluorescence remains after 2 wk, as shown.
10% on Celite (Sigma Chemical Co.) is used, as reported
by Rinderknecht.14 A total of 50 mg of this fine powder
is added to the albumin, stirred, and allowed to react
for approximately 1 hr at room temperature. Shorter
incubation times may result in a less fluorescent conjugate. The Celite is removed easily with low-speed
centrifugation. The next steps involve the separation
of the RA-FITC conjugate from the unreacted FITC
and the exchange of carbonate-bicarbonate buffer with
physiologic 0.1% phosphate buffered saline (PBS). Both
steps are combined when using a medium-grade Sephadex G-25 (Sigma Chemical Co.) gel filtration column loaded with the PBS.
The visual separation of the two fluorescent bands
is striking; the RA-FITC conjugate elutes through the
column, while the unbound FITC remains on the top.
The resultant volume of the effluent is approximately
50% greater than the volume initially added to the column. The yield of albumin is 97%. The column is rejuvenated by running several volumes of PBS until it
is visibly clear of fluorescent material. The technique
is straightforward and easily mastered. Up to 2 g of
conjugate in 20 ml of buffer have been separated in a
2.5 cm dia X 32 cm burette containing a 24-cm column
of Sephadex. Larger volumes may be produced by using
larger or multiple columns.
These two steps may also be accomplished by using
activated charcoal to absorb the unreacted FITC before
exchanging the buffers with ultrafiltration, as described
elsewhere." This method may be suitable for larger
quantities but was cumbersome and resulted in excess
protein loss (35%) to the charcoal in our laboratory.
Vol. 27
The resultant conjugate is passed through a 0.22 micron filter (Millipore Corp.; Bedford, MA) for sterilization and stored at 0 to 5°C. The unbound fluorescence is determined by measuring the fluorescence of
a small sample before and after passage through an
ultrafiltration micropartition system (Amicon MPS-1
YMT, Danvers, MA). Less than 1% unbound fluorescence is considered acceptable. The fluorescence vs
concentration curve was linear from 1 to 900 fluorescence equivalents of 1 ng of aqueous sodium fluorescein.
Two groups of six Dutch pigmented rabbits (a pilot
and repeat study), weighing 1.5 to 2.5 kg each, were
given approximately 125 mg of RA-FITC/kg of body
weight by rapid intravenous injection in one posterior
auricular vein without adverse reaction. Plasma samples were obtained from the other ear at intervals from
1 hr to 17 days. Plasma levels of fluorescence were
determined by centrifuging whole blood in a heparinized capillary tube and withdrawing 20 p\ of plasma
with a micropipette and diluting this into 1.98 ml of
PBS. Plasma albumin concentration can be calculated
from this by using the fluorescence curve.
The aqueous fluorophotometry of both eyes were
done using a commercially available automated fluorophotometer (Coherent; Palo Alto, CA) with an anterior chamber adapter focused over the pupil, a gate
time of 100 msec, and 4 steps per millimeter, at 15,
30,60, 120, 180 min, and at 24 hr. The curvette holder
is used for diluted plasma samples.
The argon laser iris photocoagulation was carried
out at 48 to 72 hr after injection of the RA-FITC. The
right eye was treated with a single burn to the mid-iris,
with 0.5 W of energy, 0.2 sec duration, and a 200-/*m
spot size. The last .05 sec or so of the burn was often
interrupted by the animal's blink. To avoid trauma no
lid speculum was used. Because numerous medications
including topical anesthetics have an effect on the
BAB,2 no medications were used at any time after injection of the RA-FITC.
These investigations conformed to the ARVO Resolution on the Use of Animals in Research.
The time course of the labelled albumin plasma
concentration (in ng/m\) after intravenous injection in
two groups of six rabbits is shown in Figure 1. Each
group received a different batch of RA-FITC. Note
that the time period in the figure is more than 14 days.
Each curve very closely fits (r2 = .98) a two-compartment model with a redistribution phase, when the labelled albumin in the plasma equilibrates with the extravascular albumin over 48 to 72 hr, followed by the
Downloaded From: on 05/13/2017
No. 3
slow metabolic decay with a half-life of 8.1 days and
9.3 days for the pilot and repeat studies, respectively.
These results are in agreement with those reported previously. 101215
Since the pilot study with six animals demonstrated
varying response to argon laser iris photocoagulation
according to iris pigmentation (brown irides reacted
more than blue), we treated six uniformly pigmented
rabbits with brown irides in the repeat study. A bubble
was formed at the photocoagulation site and prompt
miosis was noted. No conjunctival hyperemia was detectable. The results from the repeat study are shown
in Figure 2. They are presented as a ratio of anterior
chamber to plasma fluorescence. The aqueous fluorophotometry scans at 30 min clearly demonstrated
high fluorescence in the anterior chamber directly anterior to the pupil, which is in agreement with the findings of others that the BAB breakdown is predominately localized to the ciliary body. u The BAB returned
to almost normal by 24 hr (3.4 ± 1.2 X 10"3, not shown
in figure). The untreated fellow eye showed no change
during the period studied.
The FITC-labeled homologous serum albumin has
been shown to be stable in vivo and in vitro. 1012 Its
biological properties are virtually identical to native
albumin, and the label does not render the conjugate
antigenic.12 The animals may receive numerous doses
of the conjugate without evidence of anaphylaxis. The
active renal resorption is the same as for native albumin
and affords the tracer a very long and stable plasma
level after the redistribution phase.
Because the plasma fluorescence declines minimally
over several hours, it allows the rapid changes in the
aqueous fluorophotometry values to be interpreted as
reflecting only rapid changes in the BAB. Methods using sodium fluorescein depend on all the BAB and
plasma fluorescence changes during the interval between its administration and the aqueous fluorophotometry. They are therefore unable to isolate and characterize the rapidly changing permeability of the BAB.
Observations over several days also are possible by correcting the aqueous fluorophotometry for the plasma
levels but do not require additional administration of
the tracer. The absence of a bilateral effect to a stimulus
need not be assumed with this technique; in fact, its
presence can be quantified.
The large MW of this tracer paradoxically makes it
a sensitive indicator of inflammation. Since little albumin is present in normal aqueous (2:1000 ratio of
aqueous to plasma fluorescence), a small influx due to
BAB breakdown becomes immediately apparent by
time post laser (min)
Fig. 2. A rapid increase in aqueous fluorescence (corrected for
plasma fluorescence) after a single argon laser iris photocoagulation
(200 ixm spot size, 0.5 W, 0.2 sec duration) is shown (squares) compared to the fellow, untreated eye (diamonds) in six uniformly pigmented rabbits. The burn is placed approximately 72 hr after injection
of RA-FITC so that the plasma concentration is virtually constant
over the 3-hr measurement period.
comparison. This sensitivity is evident in the 50-fold
increase in aqueous fluorophotometry after only a single
laser burn, far less than previously described doses,'"4l69
and one that clinicians may consider minimal.
During the study period the animals are essentially
unmedicated so we are able to eliminate the potentially
confounding effects of anesthetics and local manipulations, even trauma from a lid speculum.
Okada and Shimada16 used RA-FITC in a model
for allergic inflammation, which differs from our report
on several points. They did not wait long enough for
the plasma levels to stabilize or long enough for the
baseline amount of fluorescent protein to enter the
aqueous. The amount of inflammation caused by their
reverse passive Arthus reaction is unfamiliar compared
with that of argon laser iris photocoagulation.
The major drawback of our technique is the current
lack of commercially available RA-FITC. Attempts to
substitute commercially available FITC-labeled bovine
serum albumin caused anaphylactic shock. Fortunately, the methods described above can be carried out
with commonly available laboratory equipment in
several hours.
The emphasis of this work is not the reaction to iris
photocoagulation, which has been reported previously,
but rather than unique way in which it was demonstrated. To our knowledge, no other technique has the
ability to measure repeatedly the BAB bilaterally in a
rapid sequence with such sensitivity over a prolonged
period of time in an unmedicated animal.
Key words: blood-aqueous barrier, aqueous fluorophotometry, FITC, FITC-labelled albumin, argon laser iris photocoagulation
Downloaded From: on 05/13/2017
Timothy Lesar, PharmD, assisted with the statistical analysis of the plasma fluorescence decay curves.
1. Unger WG, Perkins ES, and Bass MS: The response of the rabbit
eye to laser irradiation of the iris. Exp Eye Res 19:367, 1974.
2. Unger WG and Bass MS: Prostaglandin and nerve-mediated response of the rabbit eye to argon laser irradiation of the iris.
Ophthalmologica 175:153, 1977.
3. Denffer H, Erhardt W, and Neiss A: Fluorescein angiography
and changes in aqueous humor protein after argon laser photomydriasis in rabbits. Graefes Arch Clin Exp Ophthalmol 211:
155, 1979.
4. Schrems W, vanDorp HP, Mechler W, and Krieglstein GK: The
time course of laser-induced disruption of the blood-aqueous
barrier in the rabbit. Graefes Arch Clin Exp Ophthalmol 221:
65, 1983.
5. Schrems W, vanDorp HP, Wendel M, and Krieglstein GK: The
effect of YAG laser iridotomy on the blood aqueous barrier in
the rabbit. Graefes Arch Clin Exp Ophthalmol 221:179, 1984.
6. Sanders DR, Joondeph B, Hutchins R, Schwartz D, Yeh T, and
Peyman GA: Studies on the blood-aqueous barrier after argon
laser photocoagulation of the iris. Ophthalmology 90:169, 1983.
7. Sanders DR, Spigelman A, Kraff C, Agouros P, Goldstick B, and
Vol. 27
Peyman GA: Quantitative assessment of postsurgical breakdown
of the blood-aqueous barrier. Arch Ophthalmol 101:131, 1983.
Miyake K, Asakura M, and Maekubo K: Consensual reactions
of the human blood aqueous barrier to implant operations. Arch
Ophthalmol 102:558, 1984.
Schrems W, vanDorp HP, Mager S, and Krieglstein GK: The
effect of prostaglandin inhibitors of the laser-induced disruption
of the blood-aqueous barrier in the rabbit. Graefes Arch Clin
Exp Ophthalmol 221:61, 1983.
Nairn RC: Fluorescent Protein Tracing. London, Livingston,
McDonagh PF and Williams SK: The preparation and use of
fluorescent-protein conjugates for microvascular research. Microvasc Res 27:14, 1984.
Schiller AA, Schayer RW, and Hess EH: Fluorescein-conjugated
bovine albumin. Physical and biological properties. J Gen Physiol
36:489, 1953.
Katora ME and Hollis TM: A simple fluorescent method for
quantitative determination of aortic protein uptake. J Appl Physiol 39:145, 1975.
Rinderknecht H: Ultra-rapid fluorescent labelling of proteins.
Nature 193:167, 1962.
Sterling K: The turnover rate of serum albumin in man as measured by 1131 -tagged albumin. J Clin Invest 30:1228, 1951.
Okada M and Shimada K: The continuous and quantitative observation of permeability changes of the blood-aqueous barrier
in allergic inflammation of the eye. Invest Ophthalmol Vis Sci
19:169, 1980.
Downloaded From: on 05/13/2017