Chapter
II
INITIAL ASSEMBLY OF AN ALVEOLAR CELL CO-CULTURE MODEL
INTRODUCTION
Several attempts have been made to
develop in vitro systems that can be used for studying
site-specific alterations in the lungs following xenobiotic
exposure. These models can be used to study toxicological effects
in the absence of systemic processes. Techniques employing
tracheal (Cameron et al., 1989) and parenchymal explants
(Siminski et al., 1992; Fisher and Placke, 1987) as well
as microdissected explants (Plopper et al., 1991) have
been developed for studying isolated anatomical sites following
pollutant exposure. The drawback to these types of systems is the
continuous requirement for experimental animals and the
variability associated with site to site tissue differences. As
well, these systems are usually employed for studying effects at
the tissue level and do not give insight into the cellular
dynamics.
In an attempt to model these
cellular dynamics, several different co-culture models have been
used. One system uses a porous membrane, housed in a tissue
culture insert, which allows for the plating of different cell
types on either side of the membrane. Cells derived from the
epithelium and connective tissue (fibroblasts) can be used to
mimic the alveolar wall (Mangum et al., 1990). As well,
the membranes can be used to physically separate two different
cell types but maintain them in a co-culture environment. This
can be utilized to study cell-cell communication mediated by the
soluble cytokines (Adamson et al., 1991). Efforts have
been made to assemble two co-culture models incorporating
epithelial and fibroblast cells as well as isolated
bronchoalveolar macrophages. Preliminary work was done to
characterize the different lung-derived cell lines and to examine
the feasibility of assembling a useful co-culture system for the
assessment of the toxicity of particulate material.
MATERIALS AND METHODS
1. Materials and Reagents
1.1. Labware
BiocoatTM cell culture
inserts and Falcon tissue culture flasks (Becton-Dickinson;
Bedford, MA), Transwell and Transwell-COL inserts (Costar;
Cambridge, MA), CellagenTM inserts (ICN; Mississauga,
ON), Millicell-CM inserts (Millipore; Mississauga, ON), and Nunc
inserts (Gibco/BRL; Burlington, ON) were obtained as indicated.
M199 culture medium (with and
without phenol red), F12K medium (with phenol red), heat
inactivated fetal bovine serum (FBS), sodium bicarbonate, HEPES,
L-glutamine, trypsin/EDTA and phosphate buffered saline (PBS)
were from Sigma Co. (Mississauga, ON). Gentamycin was from Gibco
(Grand Island, NY). AlamarBlueTM was purchased from
Alamar Biosciences (Sacramento, CA). SRM1649 (Washington urban
dust, "organics") was purchased from the National
Institute of Standards and Technology (NIST; Gaithersberg, MD).
Thymidine (6-3H labelled) was from Amersham (Oakville,
ON).
1.2. Animals
Fisher 344, specific-pathogen-free
rats (Charles River; St. Constant, Quebec), were received in
filter cages. The animals were housed in individual cages within
a HEPA-filtered barrier unit which provided a continuous supply
of sterile air. Rats were maintained at 22 +/- 2°C and were
provided food and water ad libitum. All animals were 150 -
250 grams when used for experimentation.
2. Cell Lines
2.1. ATCC Cell Lines
Fetal, rat lung fibroblasts RFL-6
(ATCC, CCL 192, passage #9) and cat lung epithelial-like cells
AK-D (ATCC, CCL 150, passage #29) were obtained from American
Type Culture Collection (Rockville, Maryland). Upon arrival, the
ampoules containing the cells were thawed in a 37°C water bath
and seeded into T-75 flasks containing the culturemedium
suggested by ATCC (F12K - see appendix I for preparation) supplemented
with 25 mM HEPES and 25 mM sodium bicarbonate
(pH = 7.2), 25 µg/mL gentamycin, 200 mM L-glutamine
and 10% FBS (AK-D) or 20% FBS (RFL- 6). These flasks were
maintained at 37°C in a humidified atmosphere containing 5% CO2
/ 95% air. Confluent flasks were rinsed with PBS, dissociated
with trypsin-EDTA(5.0g trypsin-2.0g EDTA/L), sedimented at 1500
rpm, resuspended in culture medium and seeded into new T-75
flasks (split ratio 1:3 for AK-D cells; 1:2 for RFL-6 cells).
After generation of sufficient numbers of cells, the AK-D and
RFL-6 cells were harvested in F12K medium containing 20% FBS and
10% DMSO, aliquotted into sterile cryovials (1.5 - 2.0 x 106
cells/vial) and frozen in the vapour phase of a liquid nitrogen
Dewar flask (AK-D - passage #31; RFL-6 -passage #12). These vials
were stored in liquid nitrogen and comprised the primary stock of
these cell lines. One vial of each cell type was thawed at 37°C
and seeded into T-75 flasks containing M199 culture medium
(containing phenol red) with 10% FBS. M199 was prepared using the
same procedure as for F12K medium (see appendix
I) and
contained the same final concentration of media supplements
(gentamycin, L-glutamine). Cells were propagated and frozen in
sterile cryovials (1.5-2.0 x 106 cells/vial, AK-D -
passage #33; RFL-6 -passage #13). These cells comprised the
working stock of these cell lines and were used between passage
#33 to 40 for AK-D and between passage #13 - 20 for RFL-6.
2.2. Derivation of a Mesothelial Cell Line (R10M)
One 10 month old Fisher 344 rat was
anesthetized with sodium pentobarbitol (Somnotol; 65 mg/kg) and
euthanized by exanguination of the abdominal aorta. The skin and
hair were removed from the abdominal and thoracic regions and the
ribcage was surgically removed. The ribcage was placed in a
sterile 60 mm2 petri dish, rinsed once with PBS and a
small amount of trypsin-EDTA applied to the inner surface of the
ribcage (pleura). This preparation was incubated at 37°C for 15
minutes, the surface of the ribcage gently scraped with a Pasteur
pipette, the solution collected and placed in a T-25 culture
flask containing M199 culture medium (with phenol red) and 10%
FBS. This was incubated for approximately two months and, once
the cells were confluent, split at a ratio of 1:3, using the
techniques previously mentioned. These cells grew to confluency
in approximately two weeks and these three flasks were then
harvested and passaged at a split ratio of 1:3 several more
times. The cells were finally harvested in M199 medium containing
20% FBS and 10% DMSO and frozen in sterile cryovials (1.5 x 106
cells per vial; passage #5). One vial of these cells were thawed
and placed in a T-75 flask containing M199 medium (with phenol
red) and 10% FBS and propagated repeatedly (split ratio of 1:2 or
1:3). Stocks of R10M cells were frozen at passage 13, 20 and 39.
The R10M cells were used between passage #25 and 45.
3. Particle Stock Preparation
The SRM1649 material was weighed
and resuspended in sterile particle solution (Tween-80,
25 µg/mL; NaCl, 0.19% w/v) to a final concentration of 10
mg/mL using a Dounce glass homogenizer (Nadeau et al.,
1987). Following vortexing, this suspension was sonicated in
ice-cold water for 20 minutes and then homogenized with 25 full
strokes of the homogenizer piston. This particle stock was then
aliquotted into sterile centrifuge tubes with O-ring seals and
sterilized at 56°C for 30 minutes. These were then frozen at
-80°C to await use.
Prior to use, particles were
routinely thawed, mixed with complete M199 medium to a make a
final particle concentration of 1-2 mg/mL and applied to the cell
cultures to generate the proper final doses. This ensured that
the final concentration of particle solution in the cell cultures
was 10-20% (v/v). Controls were prepared by exposing the cultures
to M199 medium with the same serum content and the same final
concentration of particle solution as the particle doses.
4. Initial Assembly of an In Vitro
Co-Culture System
An attempt was made to generate a
model of the mammalian alveoli using cell lines and freshly
isolated rat macrophages (Fig 2.1a). Initial work was carried out
with five different insert types: BiocoatTM, Transwell®,
Nunc, CellagenTM and Millicell-CM. Inserts were placed
in 24-well plates, the inner and outer wells were filled with PBS
and incubated at 37°C in a humidified atmosphere containing 5%
CO2 / 95% air for 24 hours. After removal of the PBS
and replacement with complete M199 containing 10% FBS, AK-D cells
were seeded inside of the inserts (1.3 x 105 cells/cm2).
These were incubated for a period of 4 days.
Co-culture Models
Following incubation, the AK-D cells had successfully attached to only the Cellagen inserts. RFL-6 cells were then seeded (6.0 x 104 cells/cm2) onto the reverse side of the membrane. This was accomplished by inverting the inserts in a humidified vessel and allowing the seeded fibroblasts to attach for approximately 3 hours. The inserts were then reversed and allowed to incubate in fresh culture medium for 3 days.
Macrophages were isolated by bronchoalveolar lavage (appendix II), and seeded into the Cellagen inserts at 850 cells/insert (500 cells/cm2). These were allowed to incubate for 48 hours and then the entire co-culture preparations were fixed for scanning electron microscopy (SEM). The fixative (1% paraformaldehyde, 1% glutaraldehyde, 1% PVP 10000, 0.08M PIPES, 0.05% CaCl2, pH=7.2) was applied by aspirating off the medium in the surrounding well and replacing it with fixative. The fixative was allowed to incubate with the insert for one hour at room temperature. This process was repeated three times and then the inserts were placed directly into fixative, stored at 4°C, and provided to an electron microscope laboratory for evaluation. The inserts were then removed from the insert housing, the membrane cut in half, mounted for SEM and scanned (electron microscopy performed by Mr. Claude Daniel, Université du Québec à Montréal).
5. Macrophage/Fibroblast Co-Cultures
A co-culture system utilizing rat alveolar macrophages and the RFL-6 cell line was assembled for studying the dynamics of particle toxicity (Fig 2.1b). We have chosen Transwell-COL inserts for these experiments, which have similar coating material (collagen) as the Cellagen inserts, but have a known membrane porosity (0.45 µm). The Transwell-COL inserts are also transparent which makes them useful for routine light microscopy, i.e. visualize cells and particles within the inserts. Transwell-COL inserts were placed in 24-well plates, incubated in PBS for 24 hoursand then changed to complete M199 medium containing 1% FBS.
As well, RFL-6 cells were seeded into separate 24-well plates (3 x 104 cells/well) in complete M199 medium with 10% FBS (1 mL/well) and allowed to attach. After 24 hours, the cultures were rinsed with PBS and changed to complete M199 medium with 0.1% FBS. These were incubated for 48 hours (serum-starved) and the medium aspirated off and replaced with 1 mL of complete M199 containing 1% FBS. To half of the plates, 3H-thymidine (1 µCi/well) was added.
Macrophages were isolated from the lungs by lavage and resuspended to 5 x 105 cells/mL in complete M199 containing 1% FBS and seeded into half of the Transwell-COL inserts (100 µL/insert; 5 x 104 cells/insert). After attachment of the cells for one hour, SRM1649 particles were added to the inserts at final concentrations of 0, 20, 50 and 100 µg/insert. This resulted in inserts containing macrophages alone, particles alone or macrophages plus particles. These inserts were then transferred to the plates containing the RFL-6 cells. After 24 hours, 3H-thymidine (1 µCi/well) was added to the remaining 24-well plates. Cellular viability of the macrophages were measured using the alamarBlue assay and RFL-6 proliferation was measured by incorporation of 3H-thymidine. After a second 24 hour incubation, the remaining plates were processed in the same way.
The viability of the macrophages was assayed by placing the inserts from the co-culture plates in 24-well plates containing 1 mL of complete M199 medium (without phenol red) and 50 µL of alamarBlue. After incubation for 5 hours, 750 µL of the culture medium was collected, mixed with 750 µL of distilled water, and the absorbance of alamarBlue was measured at 570 and 600 nm. The results were calculated as + ARo570 hour-1 (Goegan et al., 1995).
The RFL-6 cultures were rinsed once with PBS, fixed with methanol, precipitated by three successive incubations with 10% trichloroacetic acid, and then solubilized in NaOH/SDS (0.3N/1%). These samples were then transferred to scintillation vials, mixed with 10 mL of Hionic Fluor (Beckman), and the radioactivity (DPM) was measured in a LS7500 Scintillation Counter (Beckman).
6. Statistics
The data were analyzed by using two-way analysis of variance. Pairwise comparisons of group means were carried out using the Bonferroni t-test.
RESULTS
1. Cell Lines
The RFL-6 cells (Fig 2.2) were originally derived from the lung tissue of a normal 18-day gestation Sprague- Dawley rat fetus. The cells had typical fibroblast morphology (i.e. bipolar appearance) and grew quickly when propagated on tissue culture plastic. RFL-6 cells split at a ratio of 1:2 grew to confluency in a period of 2 to 3 days. If not passaged, RFL-6 cells grow to form very tight monolayers that can be very difficult to isolate from the growth surface. This indicates the production of a very tight monolayer and/or production of extracelluar matrix components that help to keep the cells firmly attached. Once cell-cell contact is made, there is no evidence of the cells piling away from the growth surface.
Phase-contrast photomicrograph of RFL-6 cells
The AK-D cells (Fig 2.3) were derived from a mass culture of fetal, feline lung cells and are described by ATCC as having epithelial-like morphology. Although the presence of lamellar bodies which are characteristic markers of type II cells has been reported in the AK-D cells, ATCC has reported that the AK-D cell line has levels of phosphatidylcholine lower than that of feline tongue fibroblasts and feline kidney epithelial cells. The AK-D cells at low density (Fig 2.3a) also have typical fibroblast morphology, although when confluent (Fig 2.3b) they have a more polygonal shape. AK-D cells also grow quickly and do not form typical domes found in some epithelial cell monolayers, which is often indicative of a transporting epithelium (Freshney, 1987).
Phase-contrast photomicrograph of AK-D cells
The mesothelial cell line (R10M) consists of predominantly small, round cells (Fig 2.4). After seeding, these cells divide to form clusters that eventually contact to produce a confluent monolayer. The later stage R10M cells (passage 39; Fig 2.4c,d) attach, divide and produce confluent monolayers much more quickly than the earlier stage cells (passage 19; Fig 2.4a,b).
Phase-contrast photomicrograph of R10M cells
2. Initial Assembly of a Co-Culture System
After a period of four days, the AK-D cells did not attach to many of the inserts. The Biocoat inserts were very difficult to see through while the Matrigel material appeared to swell and lift away from the membrane. The Millicell- CM and the Transwell inserts were also difficult to see through and were therefore inadequate for any preliminary experimentation. The Nunc inserts, although transparent, had only a few attached cells and the cells grew very slowly. The only promising inserts appeared to be the Cellagen inserts which are composed entirely of collagen, providing a surface suitable for cell attachment. After a few days the cells appeared to be growing well (Cellagen inserts) and confluent monolayers were present after 4 days. Subsequent work was carried out using only Cellagen inserts. RFL-6 cells were then seeded onto the lower surface of the membrane and allowed to grow to confluency. Bronchoalveolar macrophages were seeded inside of the inserts and the co-culture preparations were fixed and mounted for SEM.
Figure 2.5 shows micrographs of the upper (a,b) and lower (c,d) surfaces at low (a,c) and high (b,d) magnification. Both the AK-D (Fig 2.5a,b) and RFL-6 cells (Fig 2.5c,d) appear to have formed tight monolayers. Figure 2.5b shows the presence of microvilli on the surface of the AK-D cells. As well, the macrophages maintained normal, rounded morphology and have spread across the surface of the monolayer (Fig 2.5a,b).
Scanning Electron Micrograph of of an alveolar co-culture model
3. Macrophage/Fibroblast Co-Cultures
3.1. Macrophage Viability
The cytotoxicity of SRM1649 towards alveolar macrophages was determined using the alamarBlue assay (Fig 2.6). Cytotoxicity was more pronounced in the macrophages exposed to SRM1649 for 48 hours (p <0.05). The highest concentration of SRM1649 (100 µg/well), while cytotoxic, only depressed alamarBlue levels approximately 20% after a continuous 24 hour exposure. However, cytotoxicity was dose-dependent after 48 hours with 100 µg/well producing a 70-80% decrease in cellular viability. This demonstrates that, in this system, the bronchoalveolar macrophages can withstand medium to high concentrations of collected environmental particulate material for short periods of time (i.e. 24 hours) without overt cytotoxicity.
Cytotoxicity of SRM1649 toward bronchoalveolar macrophages in an alveolar co-culture model as measured by the alamarBlue assay
3.2. Macrophage/Particle Mediated Alterations in RFL-6 Activity
Figure 2.7 outlines the results obtained from the co-culture of serum-starved RFL-6 cells with Transwell-COL inserts containing 1) macrophages alone, 2) SRM1649 particles alone or 3) macrophages plus SRM1649 particles. Very little effect was seen in the cultures incubated for 24 hours (Fig 2.7a) with the macrophages with or without particles, while a slight effect could be detected in the co-cultures incubated for 48 hours (Fig 2.7b). When compared to controls, increasing concentrations of SRM1649 in the Transwell-COL inserts caused a slight decrease in the levels of 3H-thymidine incorporation in the RFL-6 cell cultures (right slant fill) after 48 hours. This suggests the presence of a soluble or ultra-fine particle fraction that is able to cross the 0.45 µm membrane to reach the RFL-6 cells. As well, compared to controls, the presence of macrophages plus an increasing concentration of SRM1649 caused a slight increase in the levels of 3H-thymidine incorporation (open fill). By calculating a cumulative effect, meaning the stimulation of 3H-thymidine incorporation by the presence of macrophages and SRM1649 in combination with the absolute decrease in the 3H-thymidine uptake caused by SRM1649 cytotoxicity alone, a slight increase with increasing dose could be visualized (left slant fill).
Response of RFL-6 cells following co-culture with bronchoalveolar macrophages and SRM1649
This type of experiment demonstrated that it may be possible to generate a working co-culture system, although it is not a very sensitive measure considering the high concentrations of particulate matter used (100 µg/insert) and the low levels of response measured in the RFL-6 cells.
DISCUSSION
In an attempt to test the feasibility of assembling a co-culture model, work was carried out with several different cell lines derived from pulmonary tissue. The RFL-6, AK-D and R10M cell lines were selected as potentially useful cell lines as these cells are all derived from the mammalian lung. The Cellagen inserts, which have membranes composed entirely of collagen, provided the best surface for cellular attachment as it was possible to seed different cell types on opposite sides of the collagen membrane. Seeding the AK-D epithelial cells on the upper surface and the RFL- 6 fibroblasts on the lower surface produced a system that resembled, in some aspects, the mammalian alveolar wall. In the alveoli, the type II epithelial cells and the underlying fibroblast cells are in very close proximity. This has important implications in cellular communication as mediated by production of soluble cytokines or by direct cell-cell contact (Adamson et al., 1991). This culture system recreates the alveolar microenvironment and has potential for the study of effects of pollutants towards the mammalian lungs.
It is possible to harvest and seed bronchoalveolar macrophages into the inner well (epithelial side) of these cell culture inserts. The macrophages maintained their characteristic rounded appearance for 48 hours in culture (Fig 2.4 a,b) indicating this environment may be suitable for long term maintenance of bronchoalveolar macrophages. Following pollutant exposure, it should be possible to study macrophage mediated toxicity. Phagocytosis of insoluble material from airborne particles can cause the alveolar macrophage to release large amounts of soluble cytokines (Driscoll et al., 1993; Perkins et al., 1993; Schapira et al., 1991; Bauman et al., 1990) as well as the superoxide anion(Vallyathan et al., 1992; Mossman et al., 1989). These particles can also be directly cytotoxic to the macrophages themselves (Kondo et al., 1993; Schimmelpfeng and Seidel, 1992; Nadeau et al., 1987; Tasat and de Rey, 1987). All of these processes can modulate the toxicity of particulate preparations.
Preliminary work was done to study the effects of particle exposure on two of the cell types. Bronchoalveolar macrophages were seeded into Transwell-COL inserts and RFL-6 cells were seeded into the bottom of the 24 well plate. The response of the RFL-6 cells were analyzed following incubation with 1) macrophages alone, 2) particles alone or 3) macrophages and particles. Response of the RFL-6 cells was measured by changes in the levels of 3H-thymidine incorporation (DNA synthesis). Therefore, the ability of the bronchoalveolar macrophage to modulate the toxicity of the particles towards the RFL-6 cells, was studied. As well, the bronchoalveolar macrophage has the potential to produce factors (cytokines, reactive oxygen species) that can cause up/down regulation of RFL-6 activity.
Prior to assembling the macrophage/RFL-6 co-cultures, the RFL-6 cells were incubated in M199 containing only 0.1% FBS for a period of 48 hours. This was an attempt to put the RFL-6 cells into a state of quiescence (maintenance of cellular viability and function but stopping the active cell cycle - i.e. GO phase). It has been shown that the bronchoalveolar macrophage can produce large quantities of cytokines following particulate exposure and plays an important role in causing fibroproliferation in the mammalian lung (Driscoll et al., 1993; Perkins et al., 1993; Kumar et al., 1992; Adamson et al., 1991). It was hypothesized that exposure of the RFL-6 cells to macrophages and/or particles following this quiescent state would allow the cells to be more sensitive to small changes in the culture environment (production of growth factors, cytokines).
Following incubation with SRM1649, cellular viability of the bronchoalveolar macrophages was measured using the alamarBlue assay. SRM1649 produced much higher levels of cytotoxicity after a 48 hour incubation than after a 24 hour incubation (Fig 2.5). SRM1649 caused a dose-dependent depression of cellular viability following exposure for 48 hours while all doses (20, 50 and 100 µg/insert) caused similar levels of cytotoxicity (20% decrease) after 24 hours. This indicates that the macrophages have the ability to withstand relatively high concentrations of particulate material (SRM1649) for periods of up to 24 hours without overt cytotoxicity. It is necessary to use doses and incubation times that result in stimulation of the macrophage without causing overt cell death. This allows the macrophage time to phagocytose the particles resulting in solubilization of surface material, engulfment and breakdown of the particle, biotransformation of promutagens, production of stimulatory/inhibitory cytokines and production of reactive oxygen species.
As well, the cellular activity of the RFL-6 cells was measured by the incorporation of 3H-thymidine. Very little response with respect to particle concentration could be detected after 24 hours in co-culture, while slight effects could be measured after 48 hours. Compared to controls, i.e. empty inserts incubated with RFL-6 cells, the presence of macrophages plus an increasing concentration of SRM1649 caused a slight increase in the levels of 3H-thymidine incorporation (open fill) in the RFL-6 cells. Increasing concentrations of SRM1649 alone (in the Transwell-COL inserts) caused a slight decrease in the levels of 3H-thymidine incorporation in the RFL-6 cell cultures (right slant fill) after 48 hours. Taken together these results suggest that the SRM1649 preparation contains soluble or ultrafine material that can cross the cell culture insert membrane and cause a depression on 3H-thymidine incorporation of the RFL-6 cells. The slight increase in RFL-6 activity following exposure to macrophages plus particles may be attributable to the production of stimulatory cytokines by the macrophages or a depression in the production of inhibitory cytokines. The results for wells containing both macrophages and particles may be explained by the combination of two processes. One is the increase of RFL-6 activity due to the presence of the macrophages, and the other is the decrease of RFL-6 activity due to the presence of SRM1649. Taking the increase in RFL-6 activity due to the presence of macrophages plus particles and adding it to the absolute decrease caused by the presence of particles alone, one generates a net effect seen in the RFL-6 cells (left slant fill). It can be seen that with increasing particle concentration (0 - 100 µg/insert) there is a slight increase in the extent of this net effect. This suggests that the presence of SRM1649 and macrophages caused a stimulation in 3H-thymidine incorporation in serum starved RFL-6 cells.
The results indicate that it is possible to assemble a co-culture model of the alveoli using cultured cells and freshly isolated bronchoalveolar macrophages. The major drawback of this system is the low level of sensitivity. Macrophages are exposed to high levels of particles (100 µg/insert) and very little response is seen in the RFL-6 reporter cells. Before continuation with this type of modelling, work had to be carried out to characterize a cell reporter system which is more sensitive and responsive to specific stimuli. As well, it appeared that some of the material in the particulate suspensions was readily soluble in an aqueous environment. Prior to conducting further work on the assembly of a working cell culture model, the effects of different particulate preparations on individual cell lines were investigated.
