Download Generation of Monoclonal Antibodies to Cryptic Collagen Sites by

HYBRIDOMA
Volume 19, Number 5, 2000
Mary Ann Liebert, Inc.
Generation of Monoclonal Antibodies to Cryptic Collagen
Sites by Using Subtractive Immunization
JINGSONG XU, DOROTHY RODRIGUEZ, JENNY J. KIM, and PETER C. BROOKS
ABSTRACT
The extracellular matrix (ECM) plays a fundamental role in the regulation of normal and pathological
processes. The most abundantly expressed component found in the ECM is collagen. Triple helical collagen
is known to be highly resistant to proteolytic cleavage except by members of the matrix metalloproteinase
(MMP) family of enzymes. To date little is known concerning the biochemical consequences of collagen metabolism on human diseases. This is due in part to the lack of specific reagents that can distinguish between
proteolyzed and triple helical forms of collagen. Here we used the technique of Subtractive Immunization (SI)
to generate two unique monoclonal antibodies (MAbs HUIV26 and HUI77) that react with denatured and
proteolyzed forms of collagen, but show little if any reaction with triple helical collagen. Importantly, HUIV26
and HUI77 react with cryptic sites within the ECM of human melanoma tumors, demonstrating their utility
for immunohistochemical analysis in vivo. Thus, the generation of these novel MAbs not only identify specific
cryptic epitopes within triple helical collagen, but also provide important new reagents for studying the roles
of collagen remodeling in normal as well as pathological processes.
INTRODUCTION
H
mATRIX (ECM) was
thought to function mainly to provide mechanical support
for cells and tissues. However, it is now well accepted that the
ECM contains a wealth of biochemical information, which
helps to regulate nearly all aspects of cellular behavior.(1–5) One
of the most abundantly expressed proteins that help comprise
the ECM is collagen.(6,7) Studies have identified at least 30 separate genes that code for distinct chains of collagen, and to date
at least 19 different types of collagen have been identified.(6–8)
Collagen is composed of three chains organized in a triple helical fashion. This triple helical structure, in conjunction with
its capacity to bind numerous ECM proteins contributes to its
mechanical strength and resistance to proteolytic attack.(6–8)
Thus, collagen is uniquely suited for its role as a major structural component of the ECM.
Triple helical collagen can be cleaved by a family of proteolytic enzymes termed Matrix Metalloproteinases (MMPs).(9,10)
Specific members of this family have the ability to cleave triple
helical collagen into characteristic one-quarter and three-quarter sized fragments.(9,10) These proteolyzed forms of collagen
ISTORICALLY THE EXTRACELLULAR
are then rendered susceptible to further degradation by broadspectrum proteases.(9,10)
While cellular interactions with intact triple helical collagen
is known to regulate cellular behavior via interactions with specific integrin receptors, proteolytic remodeling of collagen is
also thought to be of great importance in the regulation of a
number of disease processes, including tumor growth and
metastasis.(11–13) To this end, immunological reagents such as
monoclonal and polyclonal antibodies have been developed to
study the expression, and distribution of collagen in vitro and
in vivo.(14–16) However, the majority of these antibodies recognize epitopes within the triple helical or noncollagenous domains of collagen. In fact, few reagents have been developed
that can selectively bind to proteolyzed or denatured collagen
yet have no reactivity to native triple helical collagen. A hallmark of invasive cellular behavior is the capacity of cells to
proteolytically remodel their immediate microenvironment.
Thus, monoclonal antibodies (MAbs) that specifically recognize distinct epitope that are selectively exposed following collagen proteolysis would be of great use in studying the roles of
collagen metabolism in invasive cellular behavior.
A major impediment to the development of MAbs directed
University of Southern California Keck School of Medicine, Department of Biochemistry and Molecular Biology, Norris Cancer Center, 1441
Eastlake Ave, Los Angeles, CA 90033.
375
376
XU ET AL.
to cryptic sites within collagen is associated with the immunodominant epitopes found within its triple helical and noncollagenous domains. These major structural epitopes likely
dominate the immune response. In fact, the majority of MAbs
developed to date are directed to epitopes normally exposed
within the intact collagen molecule. Thus, establishment of a protocol to shift the immune response away from the immu-nodominant epitopes toward cryptic sites that are only exposed following proteolysis would be of great use. To this end, we have
utilized a unique immunological tolerization approach called
Subtractive Immunization (SI)(17,18) to develop a set of novel
MAbs that specifically react with cryptic epitopes that are normally hidden within the three-dimensional structure of collagen.
Here we describe the generation and characterization of MAbs
termed HUIV26 and HUI77. These MAbs specifically recognize
denatured and proteolyzed collagen, but exhibit little if any reactivity to triple helical collagen. Moreover, these unique MAbs
are highly specific to distinct forms of collagen and do not crossreact with other ECM proteins. Interestingly, these MAbs were
shown to be useful for a number of biochemical assays including, solid phase enzyme-linked immunoadsorbent assay
(ELISA), Western blotting, and immunofluorescence analysis of
frozen tissues. Thus, the binding characteristics of MAbs
HUIV26 and HUI77 not only help to define a new set of cryptic epitopes within distinct forms of collagen, but provide immunological reagents for studying the roles of collagen degradation in normal as well as pathological processes.
MATERIALS AND METHODS
Reagents and chemicals
Purified human collagen types I and IV were obtained from
UBI (Lake Placid, NY), types II, III, and V were purchased
from Chemicon (Temecula, CA). Purified Fibronectin,
Laminin, and Fibrinogen were obtained from Sigma (St. Louis,
MO). Vitronectin was a kind gift from Dr. David Cheresh,
Scripps Research Institute (La Jolla, CA). OCT was obtained
from VWR (San Francisco, CA). OPD substrate was obtained
from Sigma (St. Louis, MO). Polyclonal antibody to Factor
VIII-related antigen was obtained from BioGenex (San Ramon,
CA). Rhodamine-conjugated goat anti-rabbit IgG, fluorescein
isothiocyanate (FITC)-conjugated goat anti-mouse IgG and peroxidase labeled goat anti-mouse IgG were purchased from
BioSource International (Camarillo, CA). Cyclophosphamide
was obtained from Sigma (St. Louis, MO).
Cells and cell culture
SP-20 myeloma cells were purchased from ATCC (Manassas, VA). SP-20 myeloma cells were maintained in Dulbecco’s
modified Eagle’s minimal essential medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1%, Pyruvate and
Glutamate, and Pen-Strep. Cells were maintained as subconfluent cultures until used.
Subtractive immunization
The SI protocol was performed as previously described with
some modifications.(18) Female BALb/c mice, 6 to 8 weeks old,
were obtained from B&K (Fremont, CA). On Day 1, mice were
inoculated with 200 mg of triple helical human collagen type I
or IV. On Days 2 and 3, mice were injected i.p. with 150 mg/kg
cyclophosphamide in sterile phosphate-buffered saline (PBS).
Following the cycle of cyclophosphamide injections, the mice
were allowed to recover for 2 weeks. This tolerization procedure was repeated a total of three times. Following the three
cycles of tolerization to triple helical collagen, mice were injected i.p. with thermally denatured (boiled 12 min) collagen at
a concentration of 200 mg per mouse. This immunization procedure was repeated once every 3 weeks for a total of four injections. At the end of the injection cycle, sera were collected
and screened for differential reactivity to triple helical and thermally denatured collagen. The mice with the greatest immunoreactivity to thermally denatured collagen, as compared
with triple helical collagen, were selected for the fusion.
Production of hybridomas
Hybridoma fusions were completed by standard techniques(18) Briefly, the selected mice were sacrificed and the
spleens and associated lymph nodes were removed washed and
the spleen cells isolated. The spleen cells were fused (4:1 spleen
cells to myeloma cells) with mouse myeloma cell line SP20.
The cells were plated at a density of 2.5 3 104 cells in 100 mL
per well of hipoxantine, aminopterine, and thymidine (HAT)
medium. The Hybridoma cultures were allowed to grow for at
least 7 days before analysis. Conditioned medium from hybridoma cultures were screen by solid-phase ELISA assays for
reactivity to either triple helical and thermally denatured collagen types I and IV. Two clones were selected for further analysis. These hybridomas were subcloned by limiting dilution and
isotyped by standard procedures.(18) Both MAbs HUIV26 and
HUI77 were determined to be IgMs with a Kappa light chain.
MAb purification
Hybridoma clones HUIV26 and HUI77 were injected in pristine primed Balb/c mice to produce acites fluid. Acites fluid
was cleared by centrifugation and applied to an Immunopure
Mannose Binding Column for purification of IgMs, (Pierce,
Rockford, IL). Antibody purifications were carried out according to the manufacturers instructions.
Solid phase ELISA assays
Purified ECM proteins (25 mg/mL in PBS) were coated (50
mL/well) in 96-well microtiter plates for 16 h at 4°C. The plates
were washed with PBS and blocked with 1% protease free
bovine serum albumin (BSA) in PBS (100 mL/well). Purified
MAbs HUIV26 and HUI77 (1 mg/mL) in a total volume of 50
mL were incubated for 1 h at 37°C. Plates were washed as before and incubated with peroxidase-conjugated goat anti-mouse
IgG 1 IgM at a dilution of 1:3000 in 1% BSA in PBS. Finally,
plates were washed and developed by the addition of OPD (0.4
mg/mL) in 80-mM citrate-phosphate buffer (pH 5.0). Color reaction was stopped by the addition of 25 mL of 4N H2SO4. Immune reactivity was quantified by measuring the optical density (O.D.) at a wave length of 490 nm with a microtiter plate
reader. All data were corrected for nonspecific binding to BSA.
In addition, the O.D.s were also corrected for nonspecific binding of the secondary antibody.
GENERATION OF MAbs TO CRYPTIC SITES IN COLLAGEN
Proteolyzed collagen ELISA assay
Microtiter wells were coated with 25 mg/mL of triple helical collagen type IV, as described above. HUVEC conditioned
medium was added to the wells (100 mL) and allowed to incubate for 1, 6, and 24 h at 37°C. The wells were extensively
washed with PBS/ethylenediaminetetraacetic acid (EDTA) and
blocked with BSA to prevent nonspecific binding and ELISAs
were completed as described above. In control experiments,
wells were coated with HUVEC-conditioned medium in the absence of collagen and no reactivity with MAbs HUIV26 or
HUI77 was detected. To confirm the exposure of cryptic epitopes was dependent on proteolytic activity, protease inhibitors
including Aprotinin (10 mg/mL) and EDTA (50 mM) was added
to the HUVEC-conditioned medium during the incubation reactions. Under these conditions, HUIV26 and HUI77 showed
little if any reactivity. Finally, little if any changes in the amount
of immobilized collagen on the microtiter wells were detected
following proteolysis as detected by a polyclonal antibodies directed to triple helical and denatured collagen.
Western blot analysis
Immunoblotting was performed essentially as described with
some modifications.(19) Briefly, purified human collagen types
I or IV were separated by electrophoresis through a 10% sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) gel. Separated proteins were transferred to nitrocellulose membranes and probed with either MAbs HUIV26 or
HUI77 (1.0 mg/mL). Antibody reactivity was detected with
horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody and visualized with enhanced chemiluminescent substrate PS-3 according to the manufactures instructions.
Immunofluorescence analysis
Immunofluorescent analysis was performed essentially as
described with minor modifications.(20) Tissue biopsies from
metastatic human melanomas were embedded in OCT and snap
frozen in liquid nitrogen. Frozen tissue sections were cut (4.0
mm) with the use of a cryostat, as described previously.(20) Tissue sections were next blocked with 2.5% BSA in PBS for 1 h
at room temperature. The tissues were washed and incubated
with MAb HUIV26 or HUI77 (100 mg/mL) and either polyclonal antibodies directed to av integrin subunit (1:1000 dilution) or Factor VIII related antigen (1:50 dilution) in 1.0% BSA
in PBS for 2 h at 37°C. Tissue sections were washed 53 with
PBS for 5 min, followed by incubation with FITC-conjugated
goat anti-mouse and Rhodamine-conjugated goat anti-rabbit
secondary antibodies (1:400 dilution in BSA in PBS) for 1 hr.
Tissue sections were washed as before and mounted with antifade medium.(20) Immunofluorescence photomicrographs
were taken at a magnification of 2003.
RESULTS
Use of subtractive immunization to suppress the immune
response to dominant epitopes in triple helical collagen
While numerous studies have documented structural cleavage of purified ECM proteins in vitro, little direct evidence is
377
available concerning proteolytic cleavage of ECM proteins
within the in vivo microenvironment. This is due in large part
to the lack of reagents that can distinguish between the native
and proteolyzed forms of ECM proteins. To this end, we have
utilized SI to develop MAbs that recognize proteolyzed and denatured forms of genetically distinct collagen molecules, but
have little if any reaction with triple helical collagen. The SI
protocol involves tolerizing mice to the major immunodominant epitopes present on triple helical collagen (Fig. 1). The
tolerization steps were performed a total of three times to ensure suppression of the immune response to the major immunodominant epitopes accessible within triple helical collagen
molecules. Following immune suppression with cyclophosphamide, mice were injected with either denatured collagen
types I or IV, to stimulate an immune response to sites are
uniquely exposed following denaturation (Fig. 1).
Analysis of mouse sera for immunoreactivity to triple
helical and denatured collagen
To evaluate whether the subtractive immunization procedure
was effective in shifting the immune response toward the generation of antibodies that would reacted with denatured forms
of collagen, solid-phase ELISA assays were carried out. Microtiter plates were coated with either triple helical or thermally
denatured human collagen types I or IV. The microtiter wells
were blocked with BSA to prevent nonspecific interactions.
Sera from experimentally manipulated mice were tested for reactivity with either triple helical and denatured collagen types
I and IV. As shown in Table 1, sera from tolerized mice injected with denatured collagen type I showed an approximate
2- to 7-fold increase in reactivity with denatured collagen type
I as compared with triple helical collagen type I. Moreover, sera
from tolerized mice injected with denatured collagen type IV
showed an approximate 2- to 12-fold increase in reactivity with
denatured collagen type IV as compared with triple helical collagen type IV. These findings suggest that use of the SI procedure can result in a shift in the immune response in favor of
the generation of antibodies directed to cryptic epitopes exposed
following thermal denaturation. Based on these findings, mice
numbers 2 and 5 were selected for use in the fusion for the generation of hybridomas.
Characterization of MAbs HUIV26 and HUI77 binding
specificity to ECM proteins
Standard hybridoma techniques were carried out for two separate fusions for the generation on hybridomas producing antibodies directed to either denatured collagen types I or IV.(18,21)
Hybridoma clones derived from each fusion were screened for
reactivity with triple helical and denatured collagen types I and
IV by solid-phase ELISA. Based on these studies, two MAbs
termed HUIV26 and HUI77 were subcloned by limiting dilution and chosen for further analysis. Isotyping analysis of these
MAbs showed that both were IgMs with a kappa light chain
(data not shown). To evaluate the binding specificity of MAbs
HUIV26 and HUI77 to distinct ECM proteins, solid-phase
ELISA assays were performed. As shown in Fig. 2A, MAb
HUIV26 recognized thermally denatured collagen type IV,
while showing little if any reactivity to triple helical collagen
type IV. Moreover, MAb HUIV26 recognized denatured colla-
378
XU ET AL.
FIG. 1. Schematic representation of the Subtractive Immunization (SI) protocol. The SI protocol can be organized into four
steps: Step I involves the tolerization of mice by completing three cycles of injections of the antigen to be tolerized to (triple helical collagen), followed by cyclophosphamide injections at 24 and 48 h following the antigen. Step II involves immunizing the
mice with thermally denatured collagen. The third step includes fusion of the spleen cells from selected mice with myeloma cells.
The final step includes screening individual hybridoma clones for those that produce antibodies that are selective for thermally
denatured collagen as compared with triple helical collagen.
gen type IV in a dose-dependent and saturable manner (data not
shown). Importantly, MAb HUIV26 showed no cross-reactivity to other ECM proteins, including fibronectin, fibrinogen,
laminin, or vitronectin. In similar experiments, MAb HUI77
specifically bound thermally denatured collagen type I, while
showing little if any binding to triple helical collagen type I
(Fig. 2B). In addition, HUI77 also reacted with denatured col-
lagen type I in a dose-dependent manner (data not shown). MAb
HUI77 also showed no reactivity to other ECM proteins, including fibronectin, fibrinogen, laminin, or vitronectin. Neither
of these MAbs showed any reactivity with thermally denatured
forms of other ECM proteins, further demonstrating their specificity for collagen (data not shown). Finally, MAbs HUIV26
and HUI77 were also shown to recognize denatured collagen
TABLE 1. ELISA SCREEN
OF
MOUSE SERA
Mouse
number
Thermally denatured
collage I
Triple helical
collagen I
ELISA ratio
denatured/triple
helical
Mouse #1
Mouse #2
0.254 (6 0.002)
0.264 (6 0.009)
0.110 (6 0.002)
0.036 (6 0.006)
2.3
7.3
Mouse
number
Thermally denatured
collage I
Triple helical
collagen I
ELISA ratio
denatured/triple
helical
Mouse #3
Mouse #4
Mouse #5
0.420 (6 0.011)
0.455 (6 0.032)
0.526 (6 0.061)
0.098 (6 0.008)
0.178 (6 0.063)
0.044 (6 0.099)
14.2
12.6
12.0
Microtiter wells were coated with either triple helical or denatured collagen type
I or IV. Sera (1: 100 dilution) form experimentally treated mice were examined for
reactivity to collagen in solid phase ELISA assays. Data represent mean optical
density (O.D.) 6 standard deviations from triplicate wells.
379
GENERATION OF MAbs TO CRYPTIC SITES IN COLLAGEN
A
C
B
FIG. 2. ELISA analysis of MAb reactivity to
purified ECM protein. Microtiter wells were
coated with 25 mg/mL of purified ECM proteins
collagen types I and IV, fibrinogen, fibronectin,
laminin, and vitronectin. Purified MAb antibodies HUIV26 (A) and HUI77 (B) were added to
the wells (1.0 mg/mL) in a total volume of 50 mL
and allowed to incubate for 1 h. The wells were
washed and incubated with peroxidase labeled
goat anti-mouse secondary antibody. Immunoreactivity was detected following addition of OPD
substrate by measuring the optical density (O.D.)
at a length of 490 nm. All data were corrected for
nonspecific reactivity with secondary antibody.
Data bars represent the mean O.D. 6 standard
deviation from triplicate wells. (C). Western blot
analysis of purified collagen type I and IV using
MAbs HUIV26 and HUI77. Top panel, purified
collagen probed with MAb HUIV26. Bottom
panel, purified collagen probed with MAb
HUI77. Lane 1, purified collagen type 1. Lane 2,
purified collagen type IV.
type I and type IV in Western blot analysis of purified collagen (Fig. 2C). While MAb HUIV26 recognized its epitope
within one a-chain of collagen type IV (Fig. 2C, top panel),
MAb HUI77 recognized epitopes within a-chains of collagen
types I and IV (Fig. 2C, bottom panel). These results confirm
the specificity and utility of these MAbs in Western blotting
(Fig. 2C). Taken together, these findings suggest the MAbs
HUIV26 and HUI77 identify unique cryptic epitopes of colla-
gen, which are specifically exposed following denaturation, but
which are normally hidden within the triple helical molecule.
Differential reactivity of MAbs HUIV26 and HUI77 to
genetically distinct forms of collagen
Collagen represents the most abundantly expressed ECM
protein in the body.(6–8) In fact, at least 19 distinct forms of col-
380
lagen have been identified.(6–8) Therefore, to determine whether
the cryptic epitopes defined by MAbs HUIV26 and HUI77 were
specific to collagen types I and IV or whether these epitopes
were present within other collagen molecules, solid-phase
ELISA assays were performed. To facilitate these studies microtiter plates were coated with triple helical or thermally denatured collagen types I, II, III, IV, and V. The microtiter wells
were blocked with BSA to prevent nonspecific binding and purified MAb HUIV26 and HUI77 were tested for reactivity. As
shown in Fig. 3A, MAb HUIV26 readily bound to thermally
denatured collagen type IV while showing little if any reactivity to triple helical collagen. Importantly, MAb HUIV26
showed little if any reactivity with other distinct forms of collagen, including types I, II, III, or V. In contrast, MAb HUI77
reacted with the denatured forms of all the genetically distinct
forms of collagen tested (Fig. 3B). However, MAb HUI77
showed little if any reactivity to the triple helical forms of these
collagen molecules. Taken together these findings indicate that
while the HUIV26 epitope appears specific for collagen type
IV, the HUI77 cryptic epitope is present within all the genetically distinct forms of collagen tested.
Exposure of the HUIV26 and HUI77 cryptic epitopes
following proteolysis
Our previous studies have indicated that thermal denaturation can expose the cryptic epitopes defined by MAbs HUIV26
and HUI77. However, it is possible that structural changes conferred by thermal denaturation may be different than that resulting from proteolytic cleavage. Moreover, proteolytic cleavage of collagen likely represent a more physiologically relevant
mechanism of ECM remodeling than thermal denaturation.
Therefore, we examined whether the cryptic epitopes defined
by MAbs HUIV26 and HUI77 could be exposed by proteolytic
activity. To facilitate these studies, mictotiter wells were coated
with triple helical collagen type IV. The coated wells were next
incubated with concentrated serum free Human Umbilical Vein
Endothelial Cell (HUVEC) conditioned medium for 1, 6, and
24 h. Endothelial cell condition medium has been shown to contain a variety of proteolytic enzymes, including serine and
matrix metalloproteinases known to cleavage collagen.(22,23)
Following incubation of the immobilized collagen with
HUVEC-conditioned media, exposure of the cryptic epitopes
were monitored by incubation with purified MAbs HUIV26 and
HUI77. As shown in Figs. 4A and B, the cryptic epitopes defined by MAbs HUIV26 and HUI77 were specifically exposed
in a time-dependent manner. In control experiments, serum-free
conditioned media coated on plates in the absence of collagen
showed no reactivity with these MAbs (data not shown). Importantly, both serine (Aprotinin) and matrix metalloproteinase
inhibitors (EDTA) were shown to block the exposure of these
cryptic epitopes, thus confirming a role for MMPs and serine
proteases in the exposure of these unique sites (data not shown).
The minimal reactivity observed with these antibodies to triple
helical collagen in this assay was likely due to a small amount
of thermal denaturation that may have occurred as a result of
the time of incubation at 37°C. These findings provide further
evidence for the cryptic nature of the HUIV26 and HUI77 epitopes and furthermore, indicate that these sites can be exposed
by either proteolysis or thermal denaturation.
XU ET AL.
Exposure of the HUIV26 and HUI77 cryptic epitopes
within human tumor tissues
Our studies indicate that MAbs HUIV26 and HUI77 can recognize cryptic epitopes of purified collagen that were specifically exposed following proteolysis and thermal denaturation.
However, collagen in vivo does not exist in isolation, but is
rather interconnected within complex networks of many ECM
molecules. Therefore, to determine whether these unique cryptic epitopes are exposed during invasive human tumor growth
in vivo, we examined malignant human melanoma tumors by
immunofluorescence staining with MAbs HUIV26 and HUI77.
Tissue sections were prepared from biopsies of metastatic human melanoma tumors. Because collagen type IV is primary
expressed as a component of the basement membranes of epithelial sheets and blood vessels, tumor sections were co-stained
with a polyclonal antibody directed to factor VIII related antigen to mark the tumor blood vessels and with MAb HUIV26.
MAb HUI77 recognizes cryptic epitopes present in a variety of
collagen molecules, including collagen type I, which is expressed in the interstitial matrix surround the tumor cells. Thus,
visualize the cellular components of these tumors, melanomas
were co-stained with a polyclonal antibody directed to the av
integrin subunit, which is known to be highly expressed within
metastatic melanomas cells, and with MAb HUI77. As shown
in Fig. 5A, the cryptic HUIV26 epitope (Green) was detected
within human metastatic melanoma tumors. Importantly, the
HUIV26 epitope appear to be primarily associated with the
basement membrane of tumor-associated blood vessels. In contrast, the cryptic HUI77 epitopes (Green) were observed within
the interstitial matrix surrounding tumor nodules, as well as surrounding many individual tumor cells (Fig. 5B). Interestingly,
little if any reactivity was detected with these MAbs in frozen
sections or normal human skin (data not shown). Taken together,
these findings provide evidence for the exposure of unique cryptic collagen sites within human malignant melanoma. Moreover,
our results also suggest that during human melanoma tumor
growth in vivo, the three-dimensional structure of collagen is altered allowing exposure of unique sites defined by MAbs
HUIV26 and HUI77. Further studies are know underway to more
precisely identify the exact amino acid sequence of these cryptic epitopes and to determine if exposure of these cryptic collagen epitopes play a functional role in human melanoma tumor
growth.
DISCUSSION
Collagen, the most abundant protein expressed in the body
has been shown to play critical roles as a mechanical scaffold
as well as regulating tissue morphogenesis and differentiation.(6–8) In fact, the crucial functions of collagen in biological
systems can be demonstrated by the fact that a variety of human diseases are characterized by genetic defects in genes coding for collagen, such as Osteogenesis Imperfecta, Epidermolysis Bullosa, and Alports Syndrome.(24) Thus, an in-depth
understanding of the cellular and biochemical functions of collagen on physiological processes is of great importance. While
numerous studies have demonstrated the functional importance
of intact triple helical collagen on biological processes, recent
studies have indicated that collagen remodeling plays a critical
GENERATION OF MAbs TO CRYPTIC SITES IN COLLAGEN
381
FIG. 3. ELISA analysis of MAb reactivity to genetically distinct forms of collagen. Microtiter wells were coated with 25 mg/mL
of triple helical or thermally denatured purified collagen including, types I, II, III, IV, and V. Purified MAb antibodies HUIV26
(A) and HUI77 (B) were added to the wells (1.0 mg/mL) in a total volume of 50 mL and allowed to incubate for 1 h. The wells
were washed and incubated with peroxidase labeled goat anti-mouse secondary antibody. Immunoreactivity was detected following addition of OPD substrate by measuring the optical density (O.D.) at a length of 490 nm. All data were corrected for
nonspecific reactivity with secondary antibody. Data bars represent the mean O.D. 6 standard deviation from triplicate wells.
role in a number of pathological events including, tumor growth
and metastasis, arthritis, and diabetic retinopathy.(8,25,26) However, the cellular and biochemical mechanisms by which collagen remodeling contributes to these diseases is not completely
understood. This gap in our knowledge is due in part to the lack
of reagents that can specifically detect proteolytically remodeled collagen.
Recently, some success has been achieved in the generation
of antibodies that can recognize denatured forms of collagen
type II.(27–30) These antibodies were developed by immunizing
382
XU ET AL.
FIG. 4. MAbs HUIV26 and HUI77 react with cryptic sites within collagen following proteolysis. Microtiter wells were coated
with 25 mg/mL of triple helical collagen types IV. Concentrated (203) serum-free-conditioned medium from HUVEC was added
to the wells (100 mL) and allowed to incubate for 1, 6, and 24 h at 37°C. The wells were extensively washed with PBS/EDTA
and blocked with BSA to prevent nonspecific binding. Purified MAb antibodies HUIV26 (A) and HUI77 (B) were added to the
wells (1.0 mg/mL) in a total volume of 50 mL and allowed to incubate for 1 h. The wells were washed and incubated with peroxidase labeled goat anti-mouse secondary antibody. Immunoreactivity was detected following addition of OPD substrate by
measuring the optical density (O.D.) at a length of 490 nm. All data were corrected for nonspecific reactivity with secondary antibody. Data bars represent the mean O.D. 6 standard deviation from triplicate wells.
GENERATION OF MAbs TO CRYPTIC SITES IN COLLAGEN
383
FIG. 5. Detection of cryptic sites defined by MAbs HUIV26 and HUI77 within invasive human melanoma tumors. Biopsies
of human metastatic melanoma tumors were embedded in OCT, snap frozen, and 4-mm-thick cryosections were prepared. Tumor sections were blocked with 2.5% BSA in PBS to prevent nonspecific binding. (A) Tumor sections were co-stained with MAb
HUIV26 and a polyclonal antibody directed to factor VIII related antigen. Red indicates human blood vessels and green indicates exposure of the HUIV26 cryptic epitope. Yellow represents co-localization. (B) Tumor sections were co-stained with MAb
HUI77 and a polyclonal antibody directed to the av integrin subunit. Red indicates cellular expression of av integrin and green
indicates exposure of the HUI77 cryptic epitope. Yellow indicates co-localization. (Photomicrographs were taken at a magnification of 2003.)
animals with isolated peptides or cleavage fragments of collagen.(27–30) These antibodies were used to detect soluble fragments of collagen from body fluids and tissue extracts.(27–30)
Moreover, these antibodies were able to detect denatured collagen type II within osteoarthritis and rheumatoid arthritic tissues in vivo.(27–30) These findings suggest that degradation of
collagen type II may play an important role in the pathogenesis of arthritis. In other studies, antibodies were developed
against the propeptide domains of collagen type I. These antibodies have been used to detect soluble propeptides in serum
and urine.(31–33) Interestingly, the levels of these soluble peptides have been shown to correlate with inflammation, hypertensive heart disease, and tumor metastasis.(32,34) These antipropeptide domain antibodies, however, do not detect
degradation of triple helical collagen because these propeptides
are cleaved as a normal process of collagen fibril formation and
thus are more indicative of collagen maturation.(31–33) Thus,
few antibodies exist that specifically react with proteoloyzed
collagen types-I and IV. Therefore, highly specific antibodies
to proteolyzed forms of collagen could have great utility not
only in disease diagnosis and prognosis, but also in the study
of the biochemical and molecular mechanisms contributing to
cellular invasion.
In this regard, SI was used to develop a set of MAbs to cryptic collagen epitopes. Similar SI protocols has been used successfully to develop MAbs directed to rare and low abundant
epitopes.(35–38) In fact, SI has been used to generate specific
MAbs directed to a variety of antigens, including sheep red
blood cells and metastasis-associated antigens expressed on the
surface of human tumor cells.(18,35–38) However, to our knowl-
edge, this immunological approach has never been used to generate specific MAbs that recognize a conformationally dependent epitope within complex macromolecules such as collagen.
In this report, we describe the use of SI to generate two MAbs
(HUIV26 and HUI77) that recognize cryptic epitopes that are
normally hidden within the three-dimensional structure of collagen. The epitopes defined by these unique MAbs are only exposed following thermal denaturation or proteolysis. While
MAb HUIV26 is highly specific for a cryptic epitope within
collagen type IV, MAb HUI77 recognizes a cryptic epitope that
appears common within a number of genetically distinct forms
of collagen including collagens types I, II, III, IV, and V. These
MAbs can detect their respective epitopes in solid-phase
ELISAs and Western blotting. Moreover, the cryptic epitopes
defined by MAbs HUIV26 and HUI77 were exposed in invasive human melanoma tumors. Interestingly, the cryptic
HUIV26 epitope was expressed primarily within the basement
membrane of tumor-associated blood vessels and was not detected within the interstitial matrix. In contrast, the HUI77 cryptic epitope exhibited a more broad distribution pattern, showing expression mainly within the interstitial matrix and
surrounding individual tumor cells. These findings are in agreement with the known distribution patterns of basement membrane collagen type IV and the various forms of interstitial collagen.(6–8) Our findings of exposure of distinct cryptic epitopes
within malignant tumor with little if any detected in normal tissues, is consistent with the notion that increased proteolytic remodeling occurs during invasive tumor growth.(9–13) Our results raise many important questions such as, whether exposure
of these unique cryptic sites are restricted to sites of melanoma
384
XU ET AL.
invasion or whether they are also exposed in other diseased tissues? Does exposure of these cryptic sites correlate with a poor
clinical prognosis? Moreover, do these cryptic collagen sites
play a functional role in disease processes or are they simply a
result of enhanced proteolytic activity? We are now in the
process of performing experiments to answer these important
questions. Taken together, our findings indicate that the SI protocol can be used to generate unique immunological reagents
to conformational and cryptic epitopes of complex ECM proteins such as collagen. More importantly, our studies have generated a novel set of MAbs with unique binding properties that
can be used to study the roles of collagen remodeling in human
diseases.
ACKNOWLEDGMENTS
P.C. Brooks was supported by NIH grants CA74132-01 and
ROI CA086140-01. The authors would like to thank Dana Jones
for her expert technical assistance and Kathryn Carner for her
help in the preparation of this manuscript.
16.
17.
18.
19.
20.
21.
22.
23.
REFERENCES
1. Timpl R: Structure and biological activity of basement membrane
proteins. Eur J Biochem 1989;180:487–502.
2. Zetter BR, and Brightman SE: Cell motility and the extracellular
matrix. Curr Opin Cell Biol 1990;2:850–856.
3. Assoian RK, and Marcantonio EE: The extracellular matrix as a
cell cycle control element in atherosclerosis and restenosis. J Clin
Invest 1996;11:2436–2439.
4. Adams JC, and Watt FM: Regulation of development and differentiation by the extracellular matrix. Development 1993;117:
1183–1198.
5. Juliano RL, and Haskill S: Signal transduction from the extracellular matrix. J Cell Biol 1993;120:577–585.
6. van Der Rest M, and Garrone R: Collagen family of proteins.
FASEB J 1991;5:2814–2823.
7. Prockop DJ, and Kivirikko KI: Collagens: molecular biology, diseases, and potentials for therapy. Ann Rev Biochem 1995;64:403–
434.
8. Olsen BR: New insights into the function of collagens from genetic analysis. Curr Opin Cell Biol 1995;7:720–727.
9. Chambers AF, and Matrisian LM: Changing views of the matrix
metalloproteinases in metastasis. J Natl Cancer Inst 1997;89:1260–
1270.
10. Mignatti P, and Rifkin DB: Biology and biochemistry of proteinases in tumor invasion. Physiol Rev 1993;73:161–185.
11. Liotta LA, Steeg PS, and Stetler-Stevenson WG: Cancer metastasis and angiogenesis: an imbalance of positive and negative regulation. Cell 1991;64:327–336.
12. Hiraoka N, Allen E, Apel IJ, Gyetko MR, and Weiss SJ: Matrix
metalloproteinases regulate neovascularization by acting as pericellular fibrinolysins. Cell 1998;95:365–377.
13. Brooks PC, Stromblad S, Sanders LC, von Schalscha TL, Aims
RT, Stetler-Stevenson WG, Quigley JP, and Cheresh DA: Localization of matrix metalloproteinase MMP-2 to the surface of invasive cells by interaction with integrin avb3. Cell 1996:85:683–693.
14. Timpl R, Wick G, and Gay S: Antibodies to distinct types of collagen and procollagens and their applications in immunohistology.
J Immunol Methods 1977;18:165–182.
15. Scheinman JI, and Tsai C: Monoclonal antibody to type-IV colla-
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
gen with selective basement membrane localization. Lab Invest
1984;50:101–112.
Monteagudo C, Merino MJ, San-Jaun J, and Stetler-Stevenson WG:
Immunohistochemical distribution of type-IV collagenase in normal, benign, and malignant breast tissue. Am J Pathol
1990;136:585–592.
Williams CV, Stechmann CL, and McLoon SC: Subtractive immunization techniques for the production of monoclonal antibodies to rare antigens. BioTechniques 1992;12:842–847.
Brooks PC, Lin J, French DL, and Quigley JP: Subtractive immunization yields monoclonal antibodies that specifically inhibit
metastasis. J Cell Biol 1993;122:1351–1359.
Brooks PC, Silleti S, von Schalscha TL, Friedlander M, and
Cheresh DA: Disruption of angiogenesis by PEX, a noncatalytic
metalloproteinase fragment with integrin binding activity. Cell
1998;92:391–400.
Brooks PC, Stromblad S, Klemke R, Visscher D, Sarkar FH, and
Cheresh DA: Antiintegrin avb3 human breast cancer growth and
angiogenesis in human skin. J Clin Invest 1995;96:1815–1822.
Galfre G, and Milstein C: Preparation of monoclonal antibodies:
Strategies and procedures. Method Enzymol 1981;73:3–46.
Schnaper WH, Grant DS, Stetler-Stevenson WG, Friedman R,
D’orazi G, Murphy AN, Bird RE, Hoythya M, Fuerst TR, French
DL, et al.: J Cell Physiol 1993;156:235–246.
Pepper MS, and Montesano R: Proteolytic balance and capillary
morphogenesis. Cell Diff Dev 1990;32:319–328.
Paepe A: The ehlers-danlos syndrome: a heritable collagen disorder as cause of bleeding. Thromb Haemost 1996;75:379–386.
Petitclerc E, Stromblad S, von Schalscha TL, Mitjans F, Piulats J,
Montgomery AMP, Cheresh DA, and Brooks PC: Integrin avb3
promotes M21 melanoma growth in human skin by regulating tumor cell survival. Cancer Res 1999;59:2724–2730.
Petitclerc E, Boutaud A, Prestayko A, Xu J, Sado Y, Ninomiya Y,
Sarras MP, Hudson BG, and Brooks PC: New functions for noncollagenous domains of human collagen type-IV: Novel integrin
ligands inhibiting angiogenesis and tumor growth in vivo. J Biol
Chem 2000;275:8051–8061.
Srinivas GR, Barrach HJ, and Chichester CO: Quantitative immunoassays for type-II collagen and its cyanogen bromide peptides. J Immunol Methods 1993;259:53–62.
Hollander AP, Pidoux I, Reiner A, Rorabeck C, Bourne R, and
Poole R: Damage to type-II collagen in aging and osteoarthritis
starts at the articular surface, originates around chondrocytes, and
extends into the cartilage with progressive degeneration. J Clin Invest 1995;96:2859–2869.
Billinghurst RC, Dahlberg L, Ionescu M, Reiner A, Bourne R,
Rorabeck C, Mitchell P, Hambor J, Diekmann O, Tschesche H, et
al. Enhanced cleavage of type-II collagen by collagenase in osteoarthritis articular cartilage. J Clin Invest 1999;99:1534–1549,
Otterness IG, Downs JT, Lane C, Bliven ML, Stukenbrok H, Scampoli DN, Milici AJ, Mezes PS: Detection of collagenase-induced
damage of collagen by 9A4, a monoclonal C-terminal neoepitope
antibody. Matrix Biol 1999;18:331–341.
Fleischmajer R, Perlish JS, and Olsen BR: The carboxylpropeptide
of type-I procollagen in skin fibrillogenesis. J Invest Dermatol
1987;89:212–215.
Miyahara M, Njieha FK, and Prockop DJ: Formation of collagen
fibrils in vitro by cleavage of procollagen with procollagen proteinases. J Biol Chem 1982;257:8442–8448.
Jukkola A, Tahtela R, Tholix E, Vuorinen K, Blanco G, Risteli L,
and Risteli J: Aggressive breast cancer leads to discrepant serum
levels of the type-I procollagen propeptides PINP and PICP. Cancer Res 1997;57:5517–5520.
Diez J, and Laviades C: Monitoring fibrillar collagen turnover in
hypertensive heart disease. Cardiovasc Res 1997;35:202–205.
GENERATION OF MAbs TO CRYPTIC SITES IN COLLAGEN
35. Prigozhina TB, and Fontalin LN: Mechanisms of antigen-induced
blockade of immune response and cyclophosphamide-promoted
tolerance to Salmonella typhi Vi antigen. Eur J Immunol 1980;10:
641–646.
36. Many A, and Schwartz RS: On the mechanism of immunological
tolerance in cyclophosphamide-treated mice. Clin Exp Immunol
1970;6:87–99.
37. Ker-hwa S, McDonald C, and Patterson PH: Comparison of two
techniques for targeting the production of monoclonal antibodies
against particular antigens. J Immunol Methods 1991;145:111–118.
38. Matthew WD, and Sandrock AW: Cyclophosphamide treatment
used to manipulate the immune response for the production
of mono-clonal antibodies. J Immunol Methods 1987;100:
73–82.
385
Address reprint requests to:
Peter C. Brooks
New York University School of Medicine
Kaplan Cancer Center
Department of Radiation Oncology and Cell Biology
Rusk Building Rm 812
400 East 34 Street
New York, NY 10016
E-mail: [email protected]
Received for publication June 12, 2000. Accepted for publication July 13, 2000.