Articles
Complement and Complement Regulatory Proteins in
Human Tears
Mark D. P. Willcox* C. A. Morris* A. Thakur* R. A. Sack,] J. Wickson* and W. Boef
Purpose. The complement system is part of the innate defense system of the body, and it
contributes to inflammatory conditions. The current study examined tears for the presence
of complement components, the activity of the components, and the presence of regulatory
components.
Methods. The significance of a functional complement system in tears was examined in four
ways. First, the presence and concentration of complement components in tear samples (openeye, closed-eye, and reflex tears) was examined by sandwich enzyme-linked immunosorbent
assay. Second, the presence of an active pathway in each tear type was established by supplementation of complement-deficient sera. Third, Western blotting of tear samples was used to
determine whether complement components were activated in tears. Fourth, the presence of
regulatory components was examined by enzyme-linked immunosorbent assay and by the
inhibition of the ability of tears to supplement deficient sera.
Results. Components Clq, C3, factor B, C4, C5, and C9 were detected in closed-eye tears.
Only C3, factor B, and C4 were detected in open-eye and reflex tears. Tears were able to
supplement complement-deficient sera, indicating that the components were in an active
state. Complement components C3, factor B, C4, and C9 were activated in closed-eye tears.
The regulatory protein decay-accelerating factor was found only in closed-eye tears. Lactoferrin, another regulatory protein present in all tear types, was shown to inhibit complementmediated red blood cell lysis, although the inhibition by closed-eye tear lactoferrin was reduced
compared to that isolated from other tear types.
Conclusions. This study has demonstrated that the complement system in tears was functionally
active and that the concentration of all components was increased gready in closed-eye tears.
In spite of the presence of regulatory proteins, proteins of the complement cascade in tears
were shown to be activated. Invest Ophthalmol Vis Sci. 1997;38:1-8.
J. he complement system is a key element of the humoral immune system. Complement is involved in the
inflammatory response by recruiting and activating
phagocytic cells, lysing microorganisms and bystander
cells, producing vasodilation, and promoting immune
complex solubilization.1 The initial stages of the com-
From the * Cornea and Contact Lens Research Unit, Cooperative Research Centre for
Eye Research and Technology, University of New South Wales, Sydney, Australia,
and the -\ Department of Biological Sciences, State College of Optometry, State
University of New York, Neiu York.
Supported by the Australian Federal Government through the Cooperative Research
Centres (CRC) program; by Vistakon, a division of Johnson and Johnson Vision
Products, Inc.; and a grant from the Optometric Vision Research Foundation of
Australia.
Received for publication April 29, 1996; revised July 15, 1996; accepted August
16, 1996.
Proprietary interest category: N.
Reprint requests: Mark Willcox, Cornea and Contact Lens Research Unit,
Cooperative Research Centre for Eye Research and Technology, University of New
South Wales, Sydney, NSW 2052, Australia.
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
plement system involve the activation of proenzymes
in a cascade reaction, with the assembly of the membrane attack complex leading to lysis of cells. At least
30 proteins are involved in complement activation and
regulation.1
Unregulated complement activation may lead to
unwanted host tissue destruction. Therefore, two types
of regulatory proteins are present to dampen the reaction. The first are homologous restriction factors,
which consist of three membrane-bound proteins
whose role is to prevent "self" tissue damage. Decayaccelerating factor (DAF; CD55) and membrane cofactor protein (MCP; CD46) act on C3, whereas membrane attack complex inhibiting protein (CD59) acts
on the membrane attack complex.2 The second are
soluble regulatory proteins found in plasma and secretions, including vitronectin3 and clusterin (which pre-
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
vent the formation of the membrane attack complex)
and lactoferrin (which acts on soluble C3).4
The presence of complement components in tears
has been the subject of some controversy. Investigations have shown the presence of individual components,5'6 but the presence of the total pathway has not
been shown. More recently, complement components
and regulatory proteins have been demonstrated in
corneal tissue.7'8
Our studies have demonstrated that the tear film
composition changes during sleep. Open-eye and reflex tears are composed primarily of lysozyme, lactoferrin, lipocalin (prealbumin), and slgA. During
sleep, a constitutive tear fluid composed of slgA (up
to 80% of the total protein) and increased levels of
serum-derived proteins is produced.9 One of the serum-derived proteins is vitronectin,10 which may act
to dampen complement activation in the tears. Our
previous studies have shown that C3 is activated during
sleep.9
The aim of the current investigation was to determine the presence and functional activity of the complement cascade in open-eye, reflex, and closed-eye
tears and to examine those tear types for the presence
of other complement regulatory proteins.
MATERIALS AND METHODS
Tears
Open-eye, reflex, and closed-eye tears were collected
from volunteers as previously described.9 Tears were
stored immediately at — 70°C until required. All tears
were collected from volunteers who were non-contact
lens wearers and had no history of ocular disease or
inflammation. Informed consent for tear collection
was obtained, and institutional human experimentation committee approval was granted. The tenets of
the Declaration of Helsinki were followed.
Tear proteins were isolated and purified partially
by high-performance liquid chromatography.11'12
Briefly, separation was carried out using an SW 300
analytical column (Waters Chromatography, Milford,
MA) with 0.5 M NaCl in phosphate buffer, pH 5, at
0.25 ml/minute flow rate. Peaks were collected, and
proteins within peaks were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE; see below, this section).
Antibodies and Sera
Polyclonal antibodies goat anti-human Clq, rabbit
anti-human C3c, rabbit anti-human C4, and rabbit
anti-human lactoferrin were obtained from Sigma (St.
Louis, MO). Polyclonal rabbit anti-human C5 and C9
and monoclonal anti-human DAF were obtained from
Wako (Wako Chemical, Richmond, VA). Polyclonal
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
sheep anti-human Clq, C3c, C4, C5, and C9 were obtained from The Binding Site (Birmingham, UK).
Polyclonal rabbit anti-human factor B was obtained
from Trace Scientific (Trace Scientific, Clayton, Australia). Sera, depleted of specific complement components, were obtained from Sigma. All sera were of
human origin except C4-deficient sera, which was
from guinea pigs.
Enzyme-Linked Immunosorbent Assay for
Individual Complement Components
The concentration of individual complement proteins
in each tear type was measured using sandwich enzyme-linked immunosorbent assays. Complement
components Clq, C3, C4, C5, C9, and factor B were
assayed in this manner.
Wells of 96-well microtiter plates (maxisorb;
Nunc, Roskilde, Denmark) were coated with 50 /il
primary antibody (2 to 50 //g/ml in 50 mM carbonate
buffer, pH 9.6). The antibody was allowed to bind
overnight at 4°C, and the wells were washed with phosphate-buffered saline (pH 7.2) containing 0.05%
(vol/vol) Tween 20 (PBST). Wells were then blocked
with PBST containing 1 % (wt/vol) bovine serum albumin (BSA) and 5% (wt/vol) lactose for 1 hour at
ambient temperature. After removal of the blocking
buffer, 50 fi\ tear samples (open, closed, or reflex;
diluted 1:200, 1:300, 1:400, 1:500, 1:1000, 1:2000 in
PBST containing 1% BSA) were added to wells and
allowed to bind for 1 hour at ambient temperature
with shaking. The plate was then washed three times
in PBST-BSA, and 50 /A of biotinylated secondary
antibody (1:10,000 dilution) was added. After incubation for 1 hour at ambient temperature with shaking,
the plate was washed three times in PBST-BSA, and
50 (A peroxidase-conjugated streptavidin (1:10,000
dilution; Sigma) was added. After incubation for 1
hour and washing (as above), 100 fi\ of 0.1 M citric
acid (pH 4) containing 0.03% (vol/vol) H2O2 and
0.3% 2.2'-azino-di-[3-acyl-benzthiazolinsulphonate]
was added to wells and incubated at ambient temperature for 10 minutes, and the optical density was measured at 405 nm.
Standard curves were run using pooled normal
human serum for which the concentration of each
complement component had been established using
radial immunodiffusion with the capture antibodies
used in the enzyme-linked immunosorbent assay
(ELISA) and against standard sera (The Binding Site).
In addition, sera deficient in the specific components
were used as negative controls.
Sodium Dodecyl Sulfate-Polyacrylamide Gel
Electrophoresis and Western Blot Analysis
SDS-PAGE was performed according to the method
of Laemmli.13 Samples were boiled in disaggregation
Complement Proteins in Human Tears
buffer (0.5 M Tris-HCl [pH 6.8], 25% glycerol, 4%
(wt/vol) SDS, 1% (vol/vol) /?-mercaptoethanol,
0.005% (wt/vol) Bromophenol blue). Gradient 4% to
15% polyacrylamide gels were used (Bio-Rad; North
Ryde, Sydney, Australia).
After SDS-PAGE, proteins were transferred to nitrocellulose membranes. Protein-binding sites were
blocked by incubating the nitrocellulose in PBST containing 5% skim milk powder for 2 hours. The nitrocellulose was incubated (2 hours, ambient temperature) in one of the following antibodies: antisera to
factor B, C3, C4, C9, DAF, and lactoferrin (all diluted
in PBST 1:500). Blots were washed three times in PBST
containing skim milk powder and were incubated (2
hours, ambient temperature) with the secondary antibody (goat anti-rabbit immunoglobulin G (IgG) coupled to biotin, 1:1000; Sigma). After washing three
times in PBST + skim milk, the blots were incubated
with peroxidase-labeled streptavidin (Sigma; 1/1000
dilution). After washing three times in PBST, fast 3,3'
diaminobenzidine substrate (Sigma) was used for detection.
Production of Antibody-Coated Sheep Red
Blood Cells
Sheep red blood cells (Commonwealth Serum Laboratories, Clayton, Victoria, Australia) were washed three
times in complement fixation diluent (CFD; Oxoid,
Basingstoke, UK) resuspended in 4 vol CFD, and the
cell concentration was titrated to 1 X 109 sheep red
blood cells per milliliter by 1:15 dilution in 0.1% (wt/
vol) Na2CO3; adjustment of the absorbance at 541 nm
was made to 0.7 with CFD.
An equal volume of hemolysin (Hunter Antisera,
Newcastle, Australia) was added to the sheep red
blood cells by the drop and gently stirred. After stirring at ambient temperature for 15 minutes, the antibody-coated sheep red blood cells (EAs) were washed
three times in CFD. The EAs were resuspended to 5
X 108 cells/ml.
Determination of Total Complement Activity in
Tears
Tear samples (open eye, closed eye, or reflex) were
used either neat or diluted 1:2, 1:5, 1:10, 1:20, 1:40,
1:80, and 1:160. Tear samples (5 /A) were added to
tubes containing 25 fx\ EA and 0.95 ml CFD. The tubes
were incubated at 37°C for 1 hour. After incubation,
0.5 ml CFD was added to each tube, and the tubes
centrifuged for 5 minutes at 3000g-. After centrifugation, the supernatant fluids were removed, and their
absorbance at 415 nm was measured in a spectrophotometer.
A standard curve of pooled normal human serum
(obtained from four volunteers) was prepared by diluting the serum (neat, 1:2, 1:5, 1:10, 1:20, 1:40, 1:80,
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
1:160) and substituting the sera dilutions for tear dilutions in the above assay. Controls of 100% EA lysis
(water substituted for the tear samples) and 0% lysis
(CFD substituted for the tear samples) were included.
The amount of lysis obtained by the tear samples was
estimated from the standard curve. This gave estimates
of the amount of total complement activity in tears as
a percentage of that in serum.
Supplementation of Complement-Deficient Sera
With Tears
Complement-deficient sera were obtained from Sigma
and were used according to the manufacturer's instructions with minor modifications. Briefly, deficient
sera and EAs were added to tubes, and tear dilutions
(1:100 or 1:500) were made. After incubation at 37°C
for 30 minutes, 1 ml ice-cold CFD was added, the tubes
were centrifuged at 3000g for 10 minutes, and the
supernatant was removed. Absorbance at 415 nm of
the supernatant was then measured. Standard curves
of supplementation by pooled normal human sera
were obtained (according to the manufacturer's instructions). The ability of tears to supplement deficient sera was calculated from the standard curves and
was expressed as a percentage of total cell lysis.
Presence of Regulatory Proteins in Tears
The presence of substances that inhibited sections of
the complement pathway was detected by examining
the ability of the two dilutions of tears (1:100 and
1:500) to lyse EAs in the presence of deficient sera.
Increased lysis by the 1:500 dilution, compared to the
1:100 dilution, was taken as an indication of the presence of an inhibitory substance.
Complement regulatory proteins lactoferrin and
DAF also were analyzed by ELISA and Western blot
analysis. The sandwich ELISA and Western blotting
methods were identical to the methods for other complement proteins (see previous page).
Lactoferrin was purified by high-performance liquid chromatography1' and concentrated to 1.8 mg/
ml (a protein value equivalent of that in the closedeye or reflex tear samples), and individual fractions
were assayed for complement regulation by adding
the lactoferrin to a 1:40 dilution of normal human
serum in a red blood cell lysis assay. Comparison between the lactoferrin-serum mixture and serum
alone was made to determine the ability of lactoferrin
to inhibit complement-mediated EA lysis.
RESULTS
Assay for the Amount of Complement
Components in Tears
The concentrations of factor B, Clq, C3, C4, C5, and
C9 are given in Table 1, which demonstrates that the
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
TABLE l.
Average Tear Concentrations for Complement and
Regulatory Proteins*
Complement or
Regulatory Protein
Complement Clq
Complement C3
Complement C4
Complement C5
Complement C9
Factor B
Lactoferrin
Vitronectinf
Tear Type
Number of
Subjects
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
Open
Closed
Reflex
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
11
16
8
14
6
6
6
6
6
Concentration
(tig/ml)
0
0.8 ± 0.4
0
27.4
106.5
4.0
1.7
5.6
0.2
±
±
±
±
±
±
47.5
84.3
5.6
3.1
5.1
0.4
0
0.9 ± 0.4
0
0
1.2 ± 1.5
0
4.0
20.8
0.1
2500
1800
1800
0.8
± 5.0
± 8.1
± 0.1
± 900
± 100
± 400
± 0.3
3.7 ± 2.2
0.1 ± 0.0
% Serum
Concentration
—
0.5
—
3.9
15.2J
0.57
0.68
2.2
0.08
—
1.9
—
—
2.2
—
3.7
19.7J
0.1
NA
NA
NA
0.25
1.2
0.03
NA = not available.
* Different tear types were analyzed for the presence of individual complement components using a
sandwich ELISA technique and commercially available antibodies. Tears were also analyzed for the
presence of DAF. However, the ELISA assay was not able to detect DAF.
t Data from Sack et al (1993).
X Significantly different from other tear types and closed-eye values; P < 0.05, Student's <-test.
closed-eye tears contained each of the complement
components. In general, open-eye and reflex tears
contained greatly reduced amounts of each component. For Clq, C5, and C9, the levels were below the
detection limit of the assay (<10 ng/ml). For complement components Clq, C4, C5, and C9, the levels in
closed-eye tears were between 0.5% and 2.2% of the
levels found in plasma. However, for factor B and C3,
levels were between 15% and 20% of the plasma levels.
Western blots (Fig. 1) demonstrated that closedeye tears contained native forms of each complement
component as well as activated forms. For example,
C3 was in its native state in reflex tears with a major
band at Mr 195 kDa, but it was degraded to yield an
extra band at 143 kDa in closed-eye tears. Factor B
was activated in the closed-eye tears to yield fragments
of Mr 63 kDa and 30 kDa. C9 showed higher molecular
weight forms in closed-eye tears compared to reflex
tears.
were able to lyse EAs, indicating that complement eidier
was not functional or was at a very low concentration in
these tear types (<2%). Closed-eye tears did not lyse
EAs. However, closed-eye tears were seen to clump the
EAs. This clumping action could mask any complement
activity in tears by precipitating the EAs.
Ability of Tears to Supplement ComplementDeficient Sera
Because of the presence of the clumping factor in
tears and the relatively low levels of complement components, an alternative and more sensitive assay was
required to demonstrate whether the complement
components identified in tears were functional. This
was provided by assessing the ability of tears to supplement complement-deficient sera.
Table 2 demonstrates that all tear types were able
to supplement each deficient serum. Tears were able to
supplement C3-deficient serum to the greatest extent,
as measured by the largest percentage of total lysis.
Determination of the Total Complement
Activity in Tears
Figure 2 demonstrates the standard curve of complement activity in serum. Neither open-eye nor reflex tears
Presence of Inhibitory-Regulatory Substances
in Tears
Inhibitory substances were active in tears after the C3
stage of complement activation. When tears were used
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Complement Proteins in Human Tears
C3c
C4
FACTOR B
(kD)
' (kD)
-210
FIGURE l. Western blots for
complement components in
tears. Tear proteins in reflex
or closed-eye tears were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and probed with
specific antisera to complement components. Molecular weights of the major
bands seen are given. RT =
reflex tears; CE = closed-eye
tears.
-93
-63
-30
RT
CE
CE
to supplement C3-deficient serum (Table 2), a 1:500
dilution of reflex tears did not yield one fifth the
amount of activation compared to a 1:100 dilution.
On the other hand, supplementation of sera by Clq
did not show any inhibitory activity. The finding that
a 1:500 dilution of tears was able to supplement complement-deficient serum but did not show complement activity when used directly indicates that the supplementation of deficient sera was a very sensitive
assay.
Tears also were examined for the presence of
known complement regulatory components. Sandwich ELISA demonstrated the presence of lactoferrin
in all tears types, with the concentration remaining
similar in the different tear types (1.8 to 2 mg/ml;
Table 1). No DAF was found in tears using the sandwich ELISA technique. Western blot analysis demon0.5
A
0.4 •
Serum
|o.3Closed eye tears
o0.2 a
0.1 •
[
00
Reflex tears
-sOpen eye tears
)^^^t=20
ft
^-
__
40
60
80
Concentration of tears/serum {%)
a
•
too
2. Standard curve of complement lysis of sheep red
blood cells. Comparison of serum lysis and tear lysis. Sheep
red blood cells were coated widi antibodies, and dilutions
of either normal human serum or different tear types were
added. After incubation for 1 hour at 37°C, lysis was determined by measuring the increase in absorbance at 415nm,
None of the tear types were able to induce cell lysis.
FIGURE
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
RT
CE
RT
CE
RT
strated the presence of lactoferrin and DAF (Fig. 3),
with DAF present only in closed-eye tear samples. We
demonstrated previously that vitronectin was present
in all tear types using both Western blotting and sandwich ELISA (Table I). 10 Lactoferrin from closed-eye
and reflex tears, purified by high-performance liquid
chromatography, was shown to inhibit lysis of EAs by
complement (Table 3). The lactoferrin purified from
closed-eye tears was far less potent as an inhibitor of
complement than that from reflex tears.
DISCUSSION
This study has demonstrated clearly that tears, particularly tears derived during 8 hours of sleep, contain
complement and that the complement components
are functionally active. These results are similar to
those of previous findings,5'6 although the previous
investigators have not analyzed tears for functionally
active complement, and some investigators failed to
find specific components.14 Most of the complement
components were present in closed eye tears between
0.5% and 2.2% of their levels in plasma. This may
indicate that these components are derived from the
leakage of plasma through the conjunctival blood vessels during sleep. Indeed, this has been proposed in
previous studies that have examined the level of fibronectin and vitronectin in tears.10'13 The exception
to this rule was for the components C3 and factor B.
Studies have shown that there is a massive recruitment
of polymorphonuclear leukocytes (PMNs) into the
closed-eye tear film,9 and PMNs contain complement
C3 and factor B as part of their granular components.16'17 Thus, we propose that C3 and factor B are
released from the PMNs after their recruitment into
the tear film, and we provide evidence that the PMNs
are activated during sleep. An alternative hypothesis
Investigative Ophthalmology 8c Visual Science, January 1997, Vol. 38, No. 1
TABLE 2. Ability
of Tears to Supplement Complement Deficient
Sera: Percentage of Total Lysis of EAs*
Reflex Tears
Complement
Deficient Sera
Clq
C4
C3
C5
C9
1/100
Open-Eye Tears
1/500
5(4)f
0
2 (2)
8(4)
5(4)
14 (10)
1 (1)
3(1)
2(1)
12 (13)
1/100
2(2)
1 (0)
1 (0)
1 (0)
9(3)
Closed-Eye Tears
1/500
1/100
1/500
Inhibitor
Activity
1 (2)
2 (3)
2(0)
2 (1)
4(3)
8 (4)
0
2(2)
25 (15)
7(5)
18 (15)
No
No
Yes
Yes
Yes
2(2)
13 (12)
5(4)
8(6)
* 1/100 and 1/500 are tear dilutions.
f Mean of four tear samples from different individuals; standard deviation in parentheses. Tears were
added to specific complement component deficient sera in the presence of EAs. The ability to
supplement a deficient serum was measured by lysis of EAs and compared with deficient serum
supplemented with normal human serum.
is that C3 and factor B are locally synthesized, perhaps
by the epithelial cells of the conjunctiva or cornea.
Indeed, epithelial cells of the kidney and uterus of
animals can express C3 and factor B.18'19 Along with
the possible release of complement components by
PMNs, these leukocytes also release proteolytic enzymes on activation. These proteolytic enzymes may
generate complement cleavage products seen in the
current study, although the aggregation of C9 observed, and the specific cleavage products of C3 and
factor B demonstrated, by Western blotting indicate
that activation was probably caused by the normal
complement pathway.
(kD)
80
Lactofe rin
f
P
DAF
(kD)
75
Relatively large amounts of C3 and factor B and
low amounts of Clq indicate that the alternative pathway of complement activation was probably predominant in the tears. We were precluded from measuring
the hemolytic alternative pathway activity because of
the large volume of tears required for the assay. This
pathway was activated primarily by the deposition of
C3 onto microbial surfaces,1 and it did not require
the presence of specific Ig G. It was interesting to note
that IgA has been claimed to activate the alternative
pathway,20 not the classical IgG-mediated complement
pathway, and that slgA is the predominant immunoglobulin in close-eye tears.
The current study demonstrates that the complement components are functionally active, and the fact
that they were activated during sleep is of particular
interest. Previously, we had denned the closed-eye tear
film as in a state of subacute inflammation because of
the presence of PMNs and large amounts of slgA.9 An
active complement system lends further weight to this
argument. A potential role for complement in the
defense of the eye during sleep was established by
the Western blots in the current investigation, which
indicated that complement had been activated during
sleep. It has been reported 21 that there was an increase
in the levels of the normal ocular microbiota during
sleep. Complement activation by the surfaces of this
TABLE 3.
Inhibition of Complement Activity
by Lactoferrin
CERT
Lactoferrin Type (1.8 mg/ml)
CE RT
FIGURE 3. Western blots of lactoferrin and Decay-accelerating factor (DAF) in closed-eye and reflex tears. Tears were
separated by sodium dodecyl sulfate-polyacrylamide gel
electrophoresis and probed with antisera to human lactoferrin and DAF. DAF was only present in closed eye tears,
whereas lactoferrin was present in both tear types. RT =
reflex tears; CE = closed-eye tears.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Reflex tear
Closed-eye tear
Commercial (Sigma)!
% Inhibition ofEA Lysis
53
7
* The commercial lactoferrin was isolated from human
colostrum. Lactoferrin (1.8 mg/ml) was added to EAs in the
presence of a 1/40 dilution of normal sera. The amount of lysis
was measured and compared with that obtained with normal
human serum.
Complement Proteins in Human Tears
resident microbiota would opsonize the bacteria and
increase their phagocytosis by PMNs,1 thus preventing
the microbiota from growing to levels that would be
pathogenic. In addition, a functional complement system in the closed-eye tears may mediate some of the
inflammatory response seen with contact lens-induced
adverse responses, such as contact lens-induced acute
red eye (CLARE). CLARE occurs during sleep as a
consequence of massive numbers of gram-negative
bacteria adhering to the contact lens.22 Complement
in the closed-eye tear film would be activated by these
adherent bacteria, leading to the production of the
anaphylatoxins C3a and C5a. C3a and C5a have numerous inflammatory sequelae, among them the production of vasodilation, chemotaxis of PMNs and macrophages, and activation of mast cells. This may mediate partially the acute red eye reaction seen with
CLARE.22 Previous studies have shown that Cl, C4,
C3, and C5 are present at higher than normal levels
in the tears of patients with corneal ulceration. 5
The complement system is kept under tight control and, during uncomplicated sleep, probably is not
overtly stimulated. Regulatory components of the
complement system include those that act at the level
of C3—both deposited C3 through the membranebound compounds DAF and MCP2 and soluble C3
through tear-borne lactoferrin 4 —and those that act
at the level of membrane attack complex, membranebound CD59,2 and tear-borne vitronectin.3'10 The current study has demonstrated that DAF was released
during sleep into the tear fluid, presumably from
sloughed off epithelial cells. Lass et al8 demonstrated
that DAF could be identified in reflex tears. The current study did not confirm this finding, and it indicated that DAF was only present in closed-eye tears.
These differences may reflect differences in the
method of tear collection or differences in detection
methods (ELISA in the current study and radioimmunoassay in the study by Lass et al 8 ). Studies have shown
that lactoferrin blocks C3 cleavage by inhibiting the
classical C3 convertase of the complement system23
and that vitronectin in tears,10 which acts on the terminal complement proteins, also may be functionally
active. Whether DAF, a normally membrane-associated protein, is functional once released into tears is
not known. Of all the complement inhibitors, lactoferrin was present in the tears in the highest concentration (1.8 mg/ml) and maybe the most potent complement inhibitor. In the current study, there was evidence of inhibition of complement activity in reflex
tears that was probably mediated by lactoferrin. Of
particular interest was the finding that the lactoferrin
in closed-eye tears was less able to inhibit complement,
which may indicate that C3 was activated more readily
during sleep. The reason for the relative lack of inhibition by closed-eye lactoferrin is unknown. Western
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
blots of lactoferrin from reflex and closed-eye tears
showed no obvious degradation of lactoferrin during
sleep. However, it is possible that the amount of Fe 3+
bound to lactoferrin changes during sleep and this
may influence functions of lactoferrin.2'1 The ability of
tears to supplement C3-deficient serum in the presence of lactoferrin may indicate that the inhibition of
complement by lactoferrin is not particularly efficient
or that the level of supplementation demonstrated
may be less than would occur in the absence of lactoferrin.
In conclusion, this study has demonstrated that
tears, particularly closed-eye tears, contain a functional complement system and regulatory compounds.
The regulatory compounds may not be able to suppress complement activation in closed-eye tears. The
presence of complement during sleep is further evidence of the subinflammatory nature of the closedeye tears and may indicate a mechanism for the production of the inflammatory response associated with,
for example, contact lens-induced acute red eye reactions.
Key Words
closed-eye, complement, regulatory components, tears
References
1. Rother K, Till GO, eds. The Complement System. New
York: Springer-Verlag; 1988.
2. Lachmann PJ. The control of homologous lysis. Immunol Today. 1991; 12:312-315.
3. Preissner KT. Structure and biological function of
vitronectin. Ann Rev Cell Biol. 1991;7:275-310.
4. Kijlstra A. The role of lactoferrin in the nonspecific
immune response on the ocular surface. Reg Immunol.
1991;3:193-197.
5. Mondino BJ, Zaidman GW. Hemolytic complement
in tears. Ophthalmic Res. 1983; 15:208-216.
6. Yamamoto GK, Allansmith MR. Complement in tears
from normal humans. Am J Ophthalmol. 1979; 88:758763.
7. Bora NS, Gobleman CL, Atkinson JP, Pepose JS,
Kaplan HJ. Differential expression of the complement
regulatory proteins in the human eye. Invest Ophthalmol Vis Sci. 1993;34:3579-3584.
8. Lass JH, Walter El, Burris TE, et al. Expression of two
molecular forms of the complement decay-accelerating factor in the eye and lacrimal gland. Invest Ophthalmol Vis Sci. 1990;31:1136-1148.
9. Sack RA, Tan KO, Tan A. Diurnal tear cycle: Evidence
for a nocturnal inflammatory constitutive tear fluid.
Invest Ophthalmol Vis Sci. 1992; 33:626-640.
10. Sack RA, Underwood PA, Tan KO, Sutherland H,
Morris CA. Vitronectin-possible contribution to the
closed-eye external host-defense mechanism. Ocul Immunol Inflam. 1993; 1:327-336.
11. Bogart B, Sack RA, Beaton A, Lew G, Kim HC. slgA,
glycoproteins and soluble mucin in reflex and closed
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
8
12.
13.
14.
15.
16.
17.
eye tears: Does the epithelium shed its membranebound mucin? ARVO Abstracts. Invest Ophthalmol Vis
Sci. 1994; 34:: 1560.
Boonstra A, Kijlstra A. Separation of human tear proteins by high performance liquid chromatography.
CurrEyeRes. 1984;3:1461-1469.
Laemmli UK. Cleaving of structural proteins during
the assembly of the head of bacteriophage T4. Nature.
1970; 227:680-685.
Chandler JW, Leder R, Kaufman HE, Caldwell JR.
Quantitative determinations of complement components and immunoglobulins in tears and aqueous humor. Invest Ophthal Vis Sci. 1974; 13:151-157.
Fukuda M, Fullard RJ, Willcox MDP, et al. Fibronectin
in the tear film. Invest Ophthalmol Vis Sci. 1996; 32:459467.
Okuda T. Murine polymorphonuclear leukocytes synthesize and secrete the third component and factor
B of complement. Intern Immunol. 1991;3:293-296.
Faried HF, Tachibana T, Okuda T. The secretion of
the third component of complement (C3) by human
polymorphonuclear leucocytes from both normal and
systemic lupus erythematous cases. Scand J Immunol.
1993;37:19-28.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
18. Ault BH, Colten HR. Cellular specificity of murine
renal C3 expression in two models of inflammation.
Immunology. 1994; 81:655-660.
19. Hasty LA, Lambris JD, Lessey BA, Pruksananonda K,
Lyttle CR. Hormonal regulation of complement components and receptors throughout the menstrual cycle. AmJObstet Gynecol. 1994; 170:168-175.
20. Stad RK, Bogers WM, Thoomas-van der Sluys ME,
van Es LA, Daha MR. In vivo activation of complement
by IgA in a rat model. ClinExp Immunol. 1992;87:138143.
21. Ramachandran L, Sharma S, Sankaridurg PR, et al.
Examination of the conjunctival microbiota after
eight hours of eye closure. CLAOJ. 1995;21:195-199.
22. Holden BA, Grant T, La Hood D, et al. Gram negative
bacteria can induce a contact lens related acute red
eye (CLARE). CLAOJ. 1995;22:47-52.
23. Kijlstra A, Jeurissen SHM. Modulation of classical C3
convertase of complement by tear lactoferrin. Immunology. 1982; 47:263-270.
24. Kievits F, Kijlstra A. Inhibition of C3 deposition on
solid-phase bound immune complexes by lactoferrin.
Immunology. 1985; 54:449-456.