Hyaluronan Enters Keratinocytes by a Novel Endocytic Route for Catabolism

Hyaluronan synthesized in the epidermis has an exceptionally short half-life, indicative of its catabolism by epidermal keratinocytes. An intracellular pool of endogenously synthesized hyaluronan, from 1 to 20 fg/cell, inversely related to cell density, was observed in cultured rat epidermal keratinocytes. More than 80% of the intracellular hyaluronan was small (<90 kDa). Approximately 25% of newly synthesized hyaluronan was endocytosed by the keratinocytes and had a half-life of 2-3 h. A biotinylated aggrecan G(1) domain/link protein probe demonstrated hyaluronan in small vesicles of approximately 100 nm diameter close to the plasma membrane, and in large vesicles and multivesicular bodies up to 1300 nm diameter around the nucleus. Hyaluronan did not co-localize with markers of lysosomes. However, inhibition of lysosomal acidification with NH(4)Cl or chloroquine, or treating the cells with the hyaluronidase inhibitor apigenin increased intracellular hyaluronan staining, suggesting that it resided in prelysosomal endosomes. Competitive displacement of hyaluronan from surface receptors using hyaluronan decasaccharides, resulted in a rapid disappearance of this endosomal hyaluronan (t(12) approximately 5 min), indicating its transitory nature. The ultrastructure of the hyaluronan-containing vesicles, co-localization with marker proteins for different vesicle types, and application of specific uptake inhibitors demonstrated that the formation of hyaluronan-containing vesicles did not involve clathrin-coated pits or caveolae. Treatment of rat epidermal keratinocytes with the OX50 monoclonal antibody against the hyaluronan receptor CD44 increased endosomal hyaluronan. However, no CD44-hyaluronan co-localization was observed intracellularly unless endosomal trafficking was retarded by monensin, or cultivation at 20 degrees C, suggesting CD44 recycling. Rat epidermal keratinocytes thus internalize a large proportion of their newly synthesized hyaluronan into non-clathrin-coated endosomes in a receptor mediated way, and rapidly transport it to slower degradation in the endosomal/lysosomal system.


INTRODUCTION
Hyaluronan is produced by the hyaluronan synthase family of enzymes (1), which directly extrude the growing glycosaminoglycan chain through the plasma membrane into the extracellular matrix or onto the cell surface. Most of the hyaluronan synthesized in tissues eventually diffuses into lymph, and is subsequently catabolized by lymph nodes (2), or the liver (3,4). However, partial degradation of radiolabeled hyaluronan injected into joints (5) and skin (6) occurs, indicating that there are also local catabolic systems. Furthermore, cultured chondrocytes (7,8), fibroblasts, smooth muscle cells (9), macrophages (10)(11)(12), and some breast cancer cell lines (13) internalize and degrade labeled exogenous hyaluronan.
A protein (HARE) associated with the endocytosis of hyaluronan in liver endothelial cells and lymph nodes has been recently cloned (14), and another (LYVE) may be involved in hyaluronan uptake by lymph vessel endothelium (15). In peripheral tissues, the ubiquitous plasma membrane protein CD44 is known to bind hyaluronan, and is also thought to contribute to hyaluronan internalization (8,12,16). Details of the molecular mechanisms of endocytosis by any of the hyaluronan receptors are largely unknown. However, it is likely that specific uptake processes are required for this macromolecule, that can exceed a molecular mass of 6 x 10 6 Da and have a 300 nm radius of gyration in solution (17), a size clearly beyond the ~100 nm diameter of a regular primary endosome.

Footnote 1
Human epidermal keratinocytes actively synthesize hyaluronan (18), and pulse-chase experiments in skin organ cultures demonstrate a short half life ( ~1 day) for epidermal hyaluronan (19,20). Similar metabolic labeling experiments in rat epidermal keratinocytes (REKs) 1 in organotypic cultures also indicate rapid hyaluronan catabolism (21). Partial degradation of epidermal hyaluronan in organ cultures is also indicated by the decrease in the molecular weight of newly synthesized hyaluronan molecules from >4x10 6 Da to 1x10 6 and Nalge Nunc Int. Corp., Naperville, IL) precoated with FBS for 30 min at 37°C. To block hyaluronan uptake through clathrin coated pits, we: 1) added 2.5-10.0 µg/ml chlorpromazine (Sigma) for 0.5-4 h in the culture medium, or 2) used serum free, hypertonic medium with 0.4 M sucrose and 0.15% BSA for 60 min as described (29). Control cells were kept in the same medium without sucrose.
The function of lysosomes was perturbed by using either 2-10 mM ammonium chloride, or 0.5-2.5 mM chloroquine (Sigma), for 0.5-6.0 h (33). Receptor recycling was inhibited by 2-32 µM monensin (Sigma) for 0.5 to 4 h (34). Hyaluronan Oligosaccharide Analysis -Medium and intracellular samples were digested with proteinase K (250 µg/ml) at 60 °C for 4 h, boiled for 10 min, and evaporated until the volume of the sample was about 100 µl. The specimens were run on a Superdex Peptide column (Pharmacia), eluted at 0.5 ml/min with 0.012 M NH 4 HCO 3 , and 100 µl aliquots from 250 µl fractions were counted for radioactivity. The included fractions were combined in three pools (fractions 27-29, 30-32, and 33-35), and 4 µg of Healon was added as a carrier. Each pool was then evaporated to dryness, redissolved in 20 µl of 0.5 M sodium acetate, pH 6.2, digested with Streptococcus hyaluronidase, and analyzed for hyaluronan disaccharides as described above. Staining for Endogenous Intracellular Hyaluronan -The basic staining protocol was essentially as decribed before (23,38). The cells were washed once with HBSS and fixed for 20 min in 2% paraformaldehyde in 0.1 M sodium phosphate buffer, pH 7.4 (PB) at room temperature. After fixation, the cells were washed 5 x 2 min with PB, and digested with Streptomyces hyaluronidase, 5-10 TRU/ml of PB for 10 min at 37°C. After washes in PB, the cells were permeabilized for 10 min in 1% BSA containing 0.1% Triton-X100. Thereafter, a biotinylated complex of hyaluronan binding region of bovine articular cartilage aggrecan G1 domain and link protein (bHABC) (38), diluted to 3-5 µg/ml in 1% BSA, was applied to the cells, and incubated overnight at 4°C, followed by 1 h in avidin-biotin peroxidase (ABC-standard kit, Vector Laboratories, Inc., Burlingame, CA) at room temperature. The color was developed with 0.05% 3,3'-diaminobenzidine (DAB, Sigma) and 0.03% H 2 O 2 . Specimens for photography were counterstained with hematoxylin for 2 min, and embedded in Supermount (Biogenex, San Ramon, CA). Specimens for densitometry were embedded in Supermount without counterstaining. For fluorescence microscopy, the procedure was modified as follows: Either Texas Red streptavidin or FITC-avidin D (Vector, 1:1000 and 1:500 dilutions, respectively, in PB) was used instead of the ABC kit. The cells were coverslipped using Vectashield (Vector) mounting medium. The cells were washed to remove the compound from the plasma membrane, fixed, and stained for endogenous intracellular hyaluronan using Alexa Fluor 633-labeled streptavidin as a reporter (Molecular Probes).

Microscopy of Uptake Markers -
Monoclonal antibodies for transferrin receptor (Chemicon International Inc., Temecula, CA), anti-caveolin 1 (Transduction Laboratories, Lexington, KY), beta-COP (Sigma) and anti CD44 mab (OX50, Biosource, Camarillo, CA ) were used at a dilution of 1:100, and the mab against cathepsin D (Transduction Laboratories) was used at a 1:50 dilution. Texas Red-labeled anti-mouse antibody (Vector) at 1:100 dilution was used as the secondary step. In the double staining protocols the primary mab was added together with the bHABC, and the secondary antibody simultaneously with FITC-avidin.
Acidic compartments in keratinocytes were visualized by feeding the cells with 30 µM Fluid phase uptake and coated pit pathways were visualized by incubating REKs in the presence of lysine fixable, Texas Red-labeled dextran (MW 10,000, Sigma, 10 µg/ml) (41) and Texas Red labeled human transferrin (Molecular Probes, 4 µg/ml) for 10-120 min (42). The cells were then fixed and processed for microscopy as above. For dual staining of dextran or transferrin, and endogenous hyaluronan, cells were first fed with Texas Red-labeled dextran or transferrin for 10-60 min, and then fixed and stained for hyaluronan as described above.
Confocal Microscopy -Laser scanning confocal microscopy was done using an Ultraview confocal scanner, built on a Nikon TE300 microscope with a 100 x NA 1. For quantitative measurements of FL-HA binding, cells were seeded onto 6-well plates at 20,000 cells/cm 2 , grown to near confluency and incubated with FL-HA with and without the competing substances described above. Each cell layer was washed 2x2 min with PBS and stripped of surface hyaluronan by 0.025% trypsin and 0.02% EDTA for 10 min at 37°C. The cells were then pelleted and washed with PBS before lysis in distilled water. The amount of FL-HA was determined using a fluorescence plate reader (SpectraFluor, Tecan, Salzburg, Austria) calibrated with known amounts of FL-HA. cells/cm 2 during 4 days (Fig. 1). In another experiment, parallel cultures were seeded at different densities and metabolically labeled on the next day. This resulted in a tenfold higher intracellular hyaluronan content in cells seeded at 14 x 10 4 vs 88 x 10 4 cells/cm 2 ( Fig. 1, inset).

Content of Intracellular Hyaluronan
Insert Fig. 1 In order to minimize internalization of hyaluronan during enzymatic treatment, we conducted sequential trypsin and Streptomyces hyaluronidase digestions at 4°C. This cold, sequential enzymatic treatment reduced the amount of intracellular hyaluronan to 2-4 fg/cell, 30-50% less than the amount determined when digestions were carried out at 37°C, reflecting the high rate of hyaluronan uptake. Despite the relatively small quantity of hyaluronan in the REKs, metabolic studies suggest that ~25% of the total hyaluronan synthesized is ultimately catabolized intracellularly (see Discussion, Hyaluronan Turnover in Keratinocytes). Aliquots of each sample were eluted on Sephacryl S-1000. In the fractions collected, hyaluronan was distinguished from other 3

H-labeled macromolecules as described in Experimental
Procedures. Most of the intracellular hyaluronan eluted after K av 0.5, indicating a molecular mass below 400 kDa (Fig. 2a), with the most abundant hyaluronan eluting at K av 0.67, a molecular mass of ~30 kDa. In contrast, hyaluronan in the medium and that released from the cell surface by trypsin treatment, both eluted close to the excluded volume, indicating a molecular mass greater than 2000 kDa (Fig. 2b, c). Although hyaluronan of <400 kDa size formed a minor proportion of the total hyaluronan in the extracellular compartments, the total amount in this size range was somewhat higher than that within the cell (Fig. 2b,c).

Insert Fig. 2
Superdex Peptide column chromatography demonstrated negligible amounts of hyaluronan oligosaccharides less than 5 kDa in any of the cell compartments (data not shown).

Origin and Turnover of Intracellular Hyaluronan -An experiment was designed to
show that the intracellular hyaluronan is derived from the cell surface or the culture medium.
REK cultures were incubated with a radiolabeled precursor of hyaluronan while Streptomyces hyaluronidase was present in the culture medium. At a concentration (5 U/ml) that removed >90% of the cell surface hyaluronan, there was a concomitant 85% decrease of the intracellular hyaluronan ( Fig. 3a), indicating that hyaluronan that is present within the cell had its origin at the cell surface and/or from the culture medium, compartments accessible to hyaluronidase. An almost equal reduction in the amount of intracellular hyaluronan was obtained by using a lower enzyme concentration (1 TRU/ml) (Fig. 3a). Therefore, most of the intracellular hyaluronan in the untreated cultures was probably endocytosed after its biosynthesis and extrusion across the plasma membrane.
Insert Fig. 3 The turnover time of the intracellular hyaluronan was studied by steady state radiolabeling the REK cultures, then chasing in a precursor-free medium containing Streptomyces hyaluronidase to prevent the entry of cell surface hyaluronan and that subsequently synthesized from the labeled precursor pool still remaining in the cell. The amount of intracellular radiolabeled hyaluronan was 70% of the starting value after 1 h chase and 60% after 2 h (Fig. 3b).
A plateau (~30% of the starting value) in the intracellular hyaluronan was reached in 4 h. The chase experiment thus showed that the half life of most of the intracellular hyaluronan is 2-3 hours (See Fig. 3, lower panel).

Hyaluronan in Cytoplasmic Vesicles -The cytochemical hyaluronan probe (bHABC)
showed that most of the cell-associated hyaluronan was distributed in patches on plasma membrane (23) (Fig. 4a). However, some was also located in vesicle-like structures, often close to the nucleus, suggesting an intracellular location. This was confirmed by short treatment of live REK cultures with Streptomyces hyaluronidase before staining for hyaluronan. The perinuclear signal was retained, while plasma membrane staining was removed by the enzyme (Fig. 4b).
To The hyaluronan-positive vesicles varied from 60 to 600 nm in diameter, with a few even larger (800-1300 nm). The small diameter vesicles (~100 nm) were typically found close to the plasma membrane and were often elongated, forming tube-like structures (Fig. 5b,d). Larger vesicles were distributed around the nucleus with some appearing to be multivesicular bodies (Fig. 5a, Insert Fig. 5. confirm a colocalization with intracellular hyaluronan (Fig. 6a). The data suggest that intracellular hyaluronan is derived from pericellular/extracellular locations, although the intracellular origin is not totally ruled out because of possible intracellular vesicle fusion events.

Colocalization of Intracellular Hyaluronan with a Membrane Endocytic Tracer, but not with Markers of Lysosomes and Golgi Vesicles
Insert Fig. 6.
Most of the lysosomes, visualized with anti-cathepsin D staining (Fig. 6c), were hyaluronan negative, and most of the hyaluronan positive structures were negative for cathepsin D (Fig. 6c), indicating that the demonstratable intracellular hyaluronan in REKs was not localized in lysosomes. Likewise, REKs fed with DAMP, a tracer that seeks its way into acidic compartments of the cell (40), showed only rare colocalization with intracellular hyaluronan (Fig.   6b). The data suggest that most of the hyaluronan-containing vesicles had close to neutral pH, and probably represented early endosomes.
Golgi compartments were localized using an anti-COP mab (45) (Fig. 6d, red). While the COP-positive structures were grouped around the nucleus, like most of the vesicles positive for hyaluronan, they did not colocalize (Fig. 6d).

Intracellular Hyaluronan Accumulation Following Lysosome Inhibition -Lysosomal
functions of the REKs were inhibited by ammonium chloride or chloroquine treatment, followed by staining for intracellular hyaluronan. Optical density (OD) of the intracellular hyaluronan signal was also measured, and the mean OD per area as well as the relative area of intracellular hyaluronan-positive structures were determined. Both reagents increased the OD and the hyaluronan positive area ~ 3-4 times over that of controls (Table I) Insert Table I and Fig. 7.
The molecular mass distribution of metabolically labeled intracellular hyaluronan was also analyzed in ammonium chloride and chloroquine treated cultures using S-400 gel filtration ( Fig.   2d-f). In control cultures about 23% of the intracellular hyaluronan had the apparent molecular mass larger than 90 kDa, while the corresponding figures for chloroquine and ammonium chloride treated cultures were 50% and 53%, respectively, indicating a shift towards higher molecular mass.
Apigenin reportedly inhibits hyaluronidase in vitro (46), in sperm penetration assays (47), and in mammary and cervix tumors (48). When added to the REK culture medium, apigenin caused a rapid and dose-dependent increase in intracellular hyaluronan staining although the accumulation was less pronounced than with ammonium chloride (Table 1).
Taken together, these data in which lysosomal function was perturbed suggest that the intracellular hyaluronan demonstrated by the bHABC probe resides in vesicles targeted to lysosomes.

Reduction of Vesicular Hyaluronan by Hyaluronan-Oligosaccharides -Subconfluent
REK cultures were incubated with hyaluronan oligosaccharides of different sizes at 0.2 to 0.5 mg/ml concentrations. The OD measurement of the hyaluronan staining showed that oligosaccharides with 8 monosaccharide units (HA 8 ) or smaller had no effect, while HA 10 or greater decreased intracellular hyaluronan staining (Fig. 8a). The area of DAB positivity was reduced with HA 14 to 20% of the control cultures. The structurally related glycosaminoglycans chondroitin sulfate and heparan sulfate had no effect (Fig. 8a). Fig. 4e shows histological staining for intracellular hyaluronan of a culture treated with HA 10 for 4 h, exhibiting virtually no hyaluronan signal. These findings suggest that there is an uptake receptor in REKs specific for hyaluronan which requires a minimum length of HA 10 for effective binding.
The kinetics of clearance of the intracellular hyaluronan signal following addition of HA 12 into growth medium, visualized with bHABC, is shown in Fig. 8b. The OD of the DAB color was reduced to 60% of the starting value in 5 min and reached a stable basal level, ~20% of control, in 10 min. The half life of the intracellular hyaluronan detected with bHABC was thus considerably shorter than the total hyaluronan assayed by radiolabeling (Fig. 3). This observation is consistent with the hypothesis that the bHABC probe specifically demonstrates recently endocytosed hyaluronan molecules that are rapidly shifted to another pool that cannot be detected with the bHABC probe.

Intracellular Hyaluronan and CD44
-REKs reacted with the mab OX50 against the hyaluronan receptor CD44 showed a strong plasma membrane signal, especially concentrated at areas of cell-cell contact and in patches on the apical surface, a distribution similar to that of hyaluronan (Fig. 7a, and ref. (23)). Very little CD44 was detected intracellularly by confocal microscopy, and dual staining of CD44 and intracellular hyaluronan exhibited only rare colocalizations close to the plasma membrane (Fig. 7a-c). Immunoelectron microscopy demonstrated CD44 enrichment on plasma membrane projections (Fig. 5b). CD44-positive, small diameter vesicles or tubules close to plasma membrane were occasionally ultrastructurally seen ( Fig. 5b, h). These structures were also hyaluronan positive (Fig. 5b,h). CD44 was not found in coated pits or coated vesicles (Fig. 5e,f).
Perturbation of lysosomal activity with ammonium chloride or chloroquine, which markedly increased the amount of vesicular hyaluronan, caused a minor increase in the intracellular CD44 staining (data not shown). However, treatments which have been shown to slow down receptor recycling, e.g. monensin (34) (Fig. 7d-f), or incubation at room temperature (49) (Fig. 7g-l), caused accumulation of intracellular CD44 in addition to hyaluronan.
Monensin-treated cultures frequently exhibited large circular structures with CD44 on the perimeter and hyaluronan inside (Fig. 7d-f). Confocal analyses confirmed that these structures with a ring-like distribution of CD44 were within the cells. REKs incubated at room temperature also showed a colocalized signal for CD44 and hyaluronan, typically close to the plasma membrane ( Fig. 7g-l), in addition to vesicles farther removed from the surface ( Fig. 7i and j, 4 and 2 µm from the bottom, respectively).
It was previously shown that OX50, a monoclonal anti-CD44 antibody increases pericellular hyaluronan staining on REKs (23). The consequences of this pericellular accumulation were studied by an OD assay of intracellular hyaluronan after a 4 h incubation with OX50. This antibody caused a consistent, dose-dependent increase in the intracellular hyaluronan-positive area (Fig. 9). A normal, isotypically matched mouse IgG, (data not shown) and Hermes 3, an anti CD44 antibody against human CD44, which does not recognize rat CD44, had little or no effect on the level of intracellular hyaluronan (Fig. 9).

Endogenous Hyaluronan Uptake and Coated Pits, Caveolae and Pinocytosis -Two
established mechanisms of receptor mediated endocytosis, utilizing clathrin-coated pits and caveolae, were studied. The avid internalization of transferrin, a widely used marker of the coated pit pathway (42), confirmed that this endocytosis pathway operated in REKs ( Fig. 6e-g). The formation of coated pits is selectively inhibited by chlorpromazine and hyperosmotic medium (50). Both of these inhibitors significantly reduced the uptake of Texas Red-or FITC-labeled transferrin (data not shown), indicating that the treatments perturb the coated pit pathway in REKs. However, neither of these treatments decreased the amount of intracellular endogenous hyaluronan in REK cultures ( Table 2). In fact, chlorpromazine caused an increase in the intracellular hyaluronan staining in all experiments ( Table 2). Higher chlorpromazine concentrations and shorter and longer treatment times (30 min -4 h) also caused a consistent increase in the intracellular hyaluronan staining (data not shown).
Insert Table II.
Cells with very active transferrin uptake usually exhibited low intracellular hyaluronan content. The converse was also true with cells exhibiting abundant intracellular hyaluronan (53). REKs stained with a monoclonal antibody against caveolin 1 showed a diffuse plasma membrane and cytoplasmic signal (Fig. 6l). Occasional cells with more intense punctate caveolin 1 staining were seen (Fig. 6l). The distribution patterns of caveolin 1 and hyaluronan were clearly different (Fig. 6l). Filipin and nystatin, reported to inhibit the formation of caveolae in endothelial cells (30), failed to cause any reduction in the intracellular hyaluronan staining ( Table   2). In contrast, both increased intracellular hyaluronan positivity ( Table 2). Lower and higher concentrations of inhibitors, and longer and shorter treatment times, were also tested, and they gave similar results (data not shown).

Receptor-Mediated and Bulk Phase Uptake of Exogenous Hyaluronan -Exogenous
FITC-labeled macromolecular hyaluronan (size > 100 kDa) given to the REKs, was internalized into vesicular structures ( Fig. 4h and 6i). Cells showing efficient uptake of exogenous hyaluronan often had high endogenous intracellular hyaluronan content, and the two signals showed partial colocalization (Fig. 4h). To distinguish between receptor mediated versus bulk phase endodocytosis of exogenous hyaluronan, hyaluronan oligosaccharides were added to inhibit the binding and uptake of FITC-labeled hyaluronan. The amount of FITC-hyaluronan bound to the cell surface at 4°C, and that associated with the cells after 30 min chase at 37°C, are shown in Fig. 10. The oligosaccharides partially inhibited FITC-hyaluronan binding on the cell surface, while sulfated glycosaminoglycans at the same concentration were less effective as competitors (Fig. 10). The HA 14 oligosaccharide also reduced the internalization of high molecular weight hyaluronan by ~70%, while the other glycosaminoglycans were less effective (Fig. 10). The uptake of exogenous hyaluronan in the presence of competing oligosaccharides and the fact that exogenous hyaluronan showed partial colocalization with dextran ( Fig. 6i) suggest that exogenous hyaluronan can be endocytosed both via receptor-mediated (HA 14 affected) and non-specific mechanisms.

Short hyaluronan oligosaccharides (HA ≥10 ) can displace hyaluronan from its receptors on
REKs (23). To study whether they were also internalized, oligosaccharides were labeled with the fluorescent AMAC group at their reducing end to avoid possible interference of receptor binding along the relatively short chain. The biological properties of the tagged oligosaccharides were similar to those of unlabeled oligosaccharides in that they were equally effective in displacing intact hyaluronan from the cell surface (data not shown). Throughout the HA 4 -HA 40 size range examined, the labeled oligosaccharides were avidly internalized by REKs when incubated at 37°C (Fig. 6j). However, this internalization was not inhibited by unlabeled oligosaccharides (Fig. 6k).
Unlike macromolecular hyaluronan, the labeled oligosaccharides were not bound to the cells when incubated at 4°C, or internalized upon subsequent chase at 37°C (data not shown). This suggests that these small hyaluronan oligosaccharides have a low affinity for the hyaluronan receptor, and that their uptake is not receptor mediated. The labeled oligosaccharides were taken up into the same vesicles as the Texas Red labeled-dextran ( Fig. 6j and k), supporting the idea of bulk phase endocytosis as their primary route for internalization. shift of the size towards higher molecular mass (see diagram in Fig. 11). Exogenous hyaluronan is partly taken up via receptor mediated uptake, and partly via bulk phase endocytosis, because: 1)

Origin of Intracellular Hyaluronan
its uptake was only partially reduced with hyaluronan-oligosaccharides, and 2) it showed a marked colocalization with bulk fluid phase endocytosis markers. of CD44 back to the cell surface from early endosomes (Fig. 11).

Routes of Hyaluronan Endocytosis -It is puzzling that clathrin-coated pits, a common
vehicle for wrapping and detaching receptor bound cargo into endosomal vesicles, and also shown to be active in REKs, seems not to operate in hyaluronan uptake. This is suggested by: 1) the lack of transferrin and transferrin receptor colocalizations with hyaluronan, 2) the inability of the inhibitors of coated pit formation to reduce intracellular hyaluronan although they effectively reduced the uptake of transferrin, and 3) the lack of electron microscopically detectable coats around hyaluronan-containing vesicles. There is previous evidence for the involvement of coated pits with the hyaluronan uptake mediated by the receptor HARE on liver endothelial cells (60) although non-saturable, fluid-phase endocytosis has also been reported in these cells (4). Coated pits were not associated with the CD44-associated hyaluronan endocytosis in chondrocytes (7,61), nor was CD44 found in the coated pits of the present keratinocytes. Keratinocytes are known to use unusual mechanisms not corresponding to the classical endocytic routes to internalize subdomains of their plasma membrane. Thus, hemidesmosomal and desmosomal plaques are internalized by a pathway not involving clathrin coated pits (62,63). A similar mechanism might operate in the uptake of membrane bound hyaluronan.
Caveolae, the other common mechanism of receptor mediated endocytosis, were also excluded as the hyaluronan-uptake route because: 1) inhibitors of caveola formation did not reduce intracellular hyaluronan content, and 2) immunostaining for caveolae showed no colocalization with hyaluronan. The fact that inhibitors of endocytosis through coated pits and caveolae actually lead to accumulation of intracellular hyaluronan may reflect the pleiotropic effects of the inhibitors, or activation of alternate uptake systems when one route is blocked (64).
While the hyaluronan endocytosed by specific receptors, and labeled dextran taken up by bulk phase endocytosis, resided in separate vesicles, both were often abundant in the same cells, suggesting that the two processes are somehow linked. Both uptake systems may correlate with the rate of membrane turnover. High molecular mass hyaluronan when present in the forming endosome may sterically exclude other macromolecules like dextran, thereby explaining the lack of colocalization of hyaluronan and dextran in most vesicles.
Degradation of Hyaluronan Following Endocytosis -The molecular mass of the intracellular hyaluronan was strongly skewed towards the lower end, with most of the material below 400 kDa and a peak at ~20 kDa, in contrast to that on the cell surface and in the medium, mostly exceeding 2000 kDa. This suggests that either: 1) the intracellular material has already undergone degradation after endocytosis, or 2) there is selective uptake of low molecular mass hyaluronan from the cell surface. The fact that perturbation of lysosomal activity caused a shift towards higher molecular mass species of intracellular hyaluronan supports the idea that the intracellular hyaluronan detected in control samples was partly degraded inside the cells; quite likely by hyaluronidases in the early endosomal compartment. Interestingly, one of the recently described hyaluronidases (Hyal 2) was reported to produce ~20 kDa hyaluronan fragments as its end product (65). The rapid clearance of the bHABC positive vesicles by the receptor competing oligosaccharides also fits with the hypothesis that these endosomes represent dynamic, transitory vehicles in which hyaluronan is moved towards lysosomes and in which its initial fragmentation takes place. The 20-90 kDa hyaluronan probably does not give a signal in contrast to the high molecular mass hyaluronan, since it may not fix well enough to be retained throughout the staining procedure. The relatively long 2-3 h half life of the total intracellular hyaluronan is best explained by gradual endoglycosidase action in the prelysosomal endosomes (Fig. 11), a process similar to that in heparan sulfate catabolism (66). The final degradation in lysosomes may be rapid, since no oligosaccharide size intermediates were detected.
Hyaluronan Size and Endocytosis -The alternative that there was preferential uptake of low molecular mass hyaluronan cannot be completely excluded, however, since there was sufficient low molecular mass material in the medium, despite its low proportion of the total, to supply the intracellular pool. Selective endocytosis of relatively low molecular mass material is also in agreement with a study on cultured endothelial cells, fibroblasts, and vascular smooth muscle cells, showing the highest internalization rate in the molecular mass range ≤30 kDa (9).
The size of pericellular hyaluronan may also influence the probability of its engulfment into a vesicle, since a long chain of hyaluronan simultaneously bound to distant cell surface sites would be expected to resist vesicle budding and detachment. Extracellular and intracellular hyaluronan showed a molecular mass up to 6000 kDa and less than 400 kDa, respectively, corresponding to 300 nm and 65 nm radius of gyration (17). It has been shown that high molecular mass dextran (150 kDa) is taken up by macropinocytosis, i.e. vesicles larger than 200 nm in diameter, while small dextran molecules (4-10 kDa) can enter both in macropinosomes and micropinosomes. Therefore, accommodation of the large hyaluronan species in the micropinosomes of ~120 nm diameter vesicles (41) seems difficult, and fragmentation of cell surface attached hyaluronan could facilitate its uptake. In chondrocytes, the increased content of hyaluronan, and its liberation from other matrix molecules enhances CD44 turnover and internalization (7