Microencapsulated Bovine Chromaffin Cells In Vitro: Effect of Density and Coseeding with a NGF-Releasing Cell Line

Immobilization of discrete cell clusters within a partially crosslinked matrix prevents reaggregation of primary tissues and may provide a means for long-term maintenance of encapsulated cells. Dissociated bovine adrenal chromaffin (BAC) cells were suspended throughout crosslinked polyanionic microspheres previously shown to be selectively permeable. Microcapsules approximately 500 µm in diameter were seeded with: 1) three different densities of BAC cells; and 2) BAC cells suspended in Matrigel® or coseeded with a genetically modified nerve growth factor (NGF)- releasing fibroblast cell line. Each group was analyzed in vitro at 1, 4 and 8 weeks for spontaneous and potassium-evoked release of catecholamines, and maintained in vitro for up to 12 weeks for morphological observations. Over time, release of norepinephrine (NE) and epinephrine (EPI) diminished, while dopamine (DA) remained constant from the monoseeded capsules. In the coseeded group, an increase in potassium-evoked release of DA was observed from 1 to 4 weeks, and remained at that level up to 8 weeks. Encapsulated chromaffin cells retained a rounded morphology typical of undifferentiated cells. Intact chromaffin cells with well preserved and abundant secretory granules were observed ultrastructurally after 4 weeks in vitro. Small neurites from the chromaffin cells in the coseeded group were observed at 4 weeks with light microscopy, and up to 12 weeks with electron microscopy. Under static incubation conditions, 1 mM D-amphetamine resulted in a significant increase in the output of NE and DA from the coseeded capsules 8 weeks postimplantation, as compared to microcapsules loaded with chromaffin cells alone. Encapsulation within an immobilization matrix allows manipulation of the internal environment, thereby providing the ability to pre-treat cells with various factors in a non-invasive manner, which may enhance long-term cellular viability.


INTRODUCTION
Neural implantation of polymer-encapsulated tissues has been utilized as a model system for analyzing the effectiveness of diffusive release of neuroactive agents in denervated regions of the central nervous system /1,26,32/. In animal models of parkinsonism, transplantation of encapsulated PC12 cells /26,32/ and bovine adrenal chromaffin (BAC) cells /1/ to a denervated striatum has been reported to alleviate dopamine receptor agonist-induced rotational asymmetry in rats. In the PC12 cell implants, the effects appeared to be dopamine mediated, since extracellular dopamine VOL. 3, NO. [2][3]1992115 measured by microdialysis approached control levels /26/. However, with adrenal medullary transplants, factors other than catecholamine release may be responsible for the behavioral improvements /2,3,5/. Trophic mechanisms are currently the most widely discussed possibilities /2,3,5,7,18/. However, NGF infused into the region of the adrenal medullary cell transplants has been reported to enhance survival, induce neuritic extensions, and increase the effectiveness of behavioral recovery/23/. This raises the question of whether low concentrations of NGF in a lesioned striatum /11/ contribute to the inability of the chromaffin cells to maintain a neuronal phenotype which may then limit their survival, or are other factors contributing to poor intrastriatal viability? The ability to deliver a chronic supply of NGF to chromaffin cells from a living system may improve graft survival and avoid the problems and limitations inherent in pumps or polymer systems/30/.
In the present study, polyelectrolyte-based microspheres containing either dissociated BAC cells or chromaffin cells coseeded with an NGFreleasing cell type, were maintained and characterized in vitro for: 1) catecholaminerelease for up to 8 weeks; and 2) morphology for up to 12 weeks. Polyelectrolytes are water,based ionic polymers in which a selectively permeable membrane is formed by interracial adsorption of a polycation on spherical polyanionic crosslinked matrices/8,12/. The technique described in this report immobilizes and maintains discrete cell clusters over extended periods of time in a transparent three-dimensional matrix. Furthermore, microencapsulation provides a method for pre-treating chromaffin cells with differentiating factors prior to transplantation studies without adding the insult of manipulating differentiated cultured chromaffin cells off a two-dimensional substrate.

Cellular preparations
Dissociated BAC cells were isolated as previously described /19,20/ and maintained in DMEM supplemented with 6% fetal calf serum (Gibco, Grand Island Biological Co., Grand Island, NY) at 37"C in a water-saturated, ambient air atmosphere containing 5% CO 2. Approximately one week following the isolation procedure, cells were harvested by aspiration, the supernatant was collected and centrifuged at 800 g. The  with a morphometric analysis system (CUE-2, Olympus Corp., Lake Success, NY).

In vitro release kinetics
Basal and potassium-evoked catecholamine release profiles under static conditions were determined by high performance liquid chromatography (HPLC) with an electrochemical detector (LCEC). Both basal (5.4 mM K/) and potassium-evoked (56 mM K /) release was measured by analyzing 450 1 HBSS, supplemented with 0.29 mM ascorbate, incubated for 15 rain each, then treated with 50/1 of 1.1 N HC10 4 prior to column injection. The medium density chromaffin cell-loaded microcapsules and the coseeded microcapsules were also challenged with 1 mM D-amphetamine sulfate in HBSS under the same conditions after being maintained for 4 and 8 weeks in vitro. The same wells from each group were repeatedly stimulated over time. The HPLC system and the catecholamine detection limit of the chromatographic system used have been previously described/1/.
Catecholamine release was expressed as either the average release/well in ng/15 min or as per capsule (pg/capsule/15 min) obtained by dividing total catecholamine release by the number of microcapsules/well. Nissl substance with cresyl violet. After 4 and 8 weeks in vitro, the relationship and pattern of viability in the medium density chromaffin cellloaded microcapsules was visualized by simultaneous staining with fluorescein diacetate (FDA) and propidium iodide (PI). One ml suspensions from each time period with approximately 50 cell-loaded microspheres were exposed to working solutions as previously described/10/. An average of 10 microcapsules/ time period were evaluated first for PI, which labels nonviable cells, and secondly with FDA, labeling viable cells, using an epifluorescent microscope (Zeiss IM 35). For ultrastructural observations, microcapsules were fixed in 3% paraformaldehyde and 2% glutaraldehyde overnight, rinsed in 50 mM PBS and processed as previously described/31/. Ultrathin sections were stained with Reynold's lead citrate and uranyl acetate and analyzed with a Philips 410 transmission electron microscope.

Statistical analysis
Mean and standard errors of the mean (S.E.M.) were calculated for each time point and group, with the exception of the size distribution of the cell-loaded microcapsules. Student's paired t-test was used for statistical analysis.

Bioassay
The NGF released from R208N.8 cell-loaded microcapsules was shown to be biologically active by its ability to induce neuritic extensions in cultured PC12 ceils after being cocultured for 5 days in vitro (Fig. 1). Monoseeded microcapsules: Figure 2 shows the average basal and potassium-evoked release of catecholamines from three wells/group (low, medium and high) in which each group contained approximately 200 capsules/well. One week following the encapsulation procedure, norepinephrine (NE), epinephrine (EPI), and dopamine (DA) were spontaneously released in 5.4 mM HBSS from each group (Fig. 2). Following a 56 mM K + incubation, a significant increase in the output of NE, EPI and DA was observed within each group. In the high and medium density microcapsules, a greater decrease in EPI and NE released was observed from 1 to 4 weeks, with the most significant decline occurring with EPI release from the high density microcapsule group ( Fig. 2A). A small decrease in potassium-evoked release of D was observed from the high and medium density groups, whereas no such decrease over time was seen in the low density group (Fig. 2C). At the low seeding density, the chromaffin cells appeared to release, under basal conditions, a greater proportion of catecholamines compared to 56 mM K+. The medium density chromaffin microcapsules were also analyzed for 15 min, with i mM .amphetamine. No significant changes in the release characteristics were observed between the basal and amphetamine conditions after 8 weeks in vitro (Fig. 3A). The monoseeded chromaffin cells suspended in Matrigel (R) showed a potassium-evoked catecholamine release pattern similar to that observed in the medium density described previously (Fig. 4A).

Morphology
Discrete clusters of chromaffin cells were observed throughout each capsule following the fabrication process (Fig. 5A). These clusters retained their rounded endocrine-like shape and did not reaggregate to form large clusters over time (Fig. 5B)

DISCUSSION
The present study shows that dissociated BAC cells can be maintained for at least 12 weeks in culture when immobilized in partially crosslinked microcapsules previously shown to be selectively permeable. Since the small aggregates remained isolated from one another over time, viability was similar at the different densities analyzed. When larger cell aggregates, e.g., those greater than 100/xm in diameter, are microencapsulated within polyelectrolytes, a significant degeneration of cells, i.e., greater than 50%, is observed within a 2 week maintenance period in vitro (unpublished observations). Very few of the anchorage-dependent cells in the dissociated chromaffin cell suspension, the fibroblasts, endothelial and Schwann cells, were observed at the ultrastructural level, suggesting the alginate immobilization matrix may inhibit their proliferation. In contrast, the fibroblasts continue to proliferate to a confluent monolayer when the chromaffin cell suspension is grown on tissue culture flasks (plates) in vitro. Survival of the engineered fibroblasts in a microcapsular configuration required the addition of an extracellular matrix substrate.
The immobilized dissociated BAC clusters did not require the application of exogenous NGF for survival. Basal and potassium-evoked catecholamine release patterns were consistent between the groups over time. The output of EPI, especially in the high and medium density groups, dropped in greater proportions than NE or DA. This may be due to a reduction in phenylethanolamine-N-methyltransferase (PNMT) activity, which catalyzes the formation of epinephrine from norepinephrine, whose expression is controlled by the glucocorticoids of the adrenal cortex/33/. Within each group, an increase in catecholamine release was observed following a high K + incubation, showing that the chromaffin cells retained the capacity for exocytosis. The monoseeded chromaffin cells, irrespective of the presence of Matrigel(R), also retained their rounded endocrine phenotype, in contrast to the chromaffin cells coseeded with the NGF-secreting cells.
In the coseeding experiments, the Matrigel (R) cell suspension was mixed with the alginate to provide a substrate for the anchorage-dependent NGF-secreting cells and support neuritic outgrowths from the chromaffin cells. Morphological observations revealed intact, viable fibroblasts in the microcapsules for up to 12 weeks in vitro, suggesting that the fibroblasts continue to survive. The ultrastructure of the coseeded chromaffin cells after 12 weeks in vitro together with the potassium-evoked catecholamine release profile at 4 and 8 weeks also suggests that the engineered fibroblasts retained the capacity to secrete bioactive NGF. Application of NGF to adrenal medulla cultures has been shown to elicit the selective induction of tyrosine hydroxylase and dopamine-flhydroxylase /16/, inducing a neuronal transformation, and altering the catecholamine profile. The activity of PNMT was down regulated in the coseeded microcapsules as shown by a dramatic decrease in EPI release.
Further evidence of a shift towards a neuronal transformation was observed when the coseeded microcapsules were exposed to 1 mM amphetamine after an 8 week maintenance period in vitro. A significant increase in measured quantities of NE and DA was observed as compared to the monoseeded capsules.
A previous study has revealed that intrastriatal implants of dissociated BAC cells in microcapsules consistently survived at 4 weeks and alleviated apomorphine-induced rotations in the 6-OHDA unilateral lesioned rat model/1/. These cells also appeared endocrine-like in their morphology/1/. Thus, in both cases exogenous NGF application did not appear to be required for short-term survival. Long-term survival may, however, require the addition of NGF.
Microencapsulation may provide a method for consistent chromaffin cell survival in intrastriatal transplants without exogenously applied NGF /1/. The results indicate that low concentrations of striatal NGF or the nutritional conditions cannot account for the poor survival of unencapsulated intrastriatal chromaffin cell implants. By modifying the internal milieu of the microcapsules, comparisons of striatal implants containing endocrine-like vs. differentiated chromaffin cells can be made, and thereby determine whether an altered phenotype survives better, and/or is more effective in ameliorating experimental parkinsonism.