Characterization of Testudine Melanomacrophage Linear, Membrane Extension Processes-Cablepodia-By Phase and Atomic Force Microscopy

By Johnson, James C; Nettikadan, Saju R; Vengasandra, Srikanth G; Lovan, Sai; Et al

SUMMARY

Melanomacrophages (MMs) are a component of an internal, pigmented cell system in liver and splenic tissues of some fishes, anurans, and reptiles. The cells have been found in centers or aggregates in sinusoids and are associated with cells capable of producing a peptide cytokine and immunoglobulins. A unique cell extension process has been observed in turtle MMs placed into cell culture, and this process has been studied by light and atomic force microscopy. These structures, referred to as cablepodia. are uniquely straight, narrow, and unbranching and appear to originate from growth cones opposite lamellipodia. Cablepodia were found to connect with other turtle MMs and fibroblasts forming cell networks. Dividing fibroblasts to which a cablepodium attached ceased cell division. The observations collectively suggest that a principal reason for aggregations of MMs in internal organs of lower vertebrates is their ability to form interconnected networks of cell processes for trapping and processing of particulate matter, cells, and infectious organisms and, possibly, for the communication of cell signals and transfer of intracellular materials.

Key words: melanomacrophage centers; extension fibrils; in vitro cell morphologyl; cell networks: inhibition of cell division; pigmented cells.

INTRODUCTION

Melanomacrophages (MMs) are internal, melanin-hearing, pigmented phagocytes, existing individually or in aggregates referred to as MM centers or MM aggregations found in the liver, spleen, or kidneys (or all) of fishes (Ellis et al., 1976; Agius. 1980; FournierBetz et al., 2000; Agius and Roberts. 2003; Sanders et al.. 2003) and anurans (Sichel and Biondi. 1971: Cicero et al.. 1982; Sichel et al., 2002). The MMs have also been found in aggregations within the liver and spleen of reptiles (Sichel. 1984; Scalia et al.. 1988: Christiansen et al., 1996). Functions for these cells have been described to include melanin synthesis and possibly free radical capture, adherence and processing of red blood cells, and internalization of both prokaryotic and eukaryotic organisms (Johnson et al., 1999; Guida et al, 2000; Sichel et al., 2002). Functional differences between vertebrate melanocytes and MMs include synthetic pathways and transductional controls for melanin synthesis and the chemical composition of the melanin-bearing somes (melanosomes) (Sciuto et al.. 1988: Corsaro et al.. 2000; Gallone et al.. 2002: Guida et al., 2004). A biological significance of MM centers is suggested because turtle hepatic MM aggregations were found to increase in size and number with age in turtles (Christiansen et al., 1996) and by season in amphibians (Barni et al., 1999; Agius and Roberts, 2003). The MM centers were also found to contain immunoglobulin (Fournier-Betz et al.. 2000) and cytokine (granulin-1(-secreting cells (Belcourt et al., 1995). The ontogeny of MM center development remains poorly understood.

The MMs from turtle liver and spleen have been placed into culture (Rund et al., 1998), and it was observed that these cells seldom divide but phagocytize bacteria at temperatures as low as 2 C. where mammalian macrophages fail to function (Johnson et al.. 1999). This study also determined that MMs phagocytize turtle red blood cells and internalized cellular debris including apparent melanosomes from ruptured MM cells, l.amellipodia appeared to be the principal sites for attachment and engulfment of these materials and cells. Collectively, the data supported the belief that the turtle MM is a nondividing, fixed tissue histiocyte with differentiated (unctions associated with the removal and autolysis of red blood cells and engulfment with processing of invading organisms possibly for antigen presentation. This study focuses on morphologic properties of a distinct turtle

MM fibril extension process observed in culture. These spike- like projections may exceed 1000 m or 10-15 cell diameters, are remarkably straight, and are postulated to have sensor-effector functions or to facilitate movement of the cell or material between cells, or both. They form networks of interconnecting cells and may be important in the formation of MM centers, in vivo.

MATERIALS AND MKTHODS

Cells and media. Liver tissue from adult (15-35 yr old) painted turtles, Chrysemys picta, was used for the preparation of MM cultures. Turtles were collected under a State Permit (JLC). and tissue was collected following an IRB-approved protocol. Slices (1- 2 g) obtained from each of the lobes were processed and the cells and tissue fragments placed into culture, as described earlier (Rund et al., 1998: Johnson et al.. 1999), with the following modifications.

The MMs were harvested from plastieware surfaces by selective trypsinization 2 wk after establishment of primary total hepatocyte cell cultures. Unattached cells, aggregates, and tissue fragments were removed by decantation and a 1:5 dilution of 0.25% trypsin- ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich Inc.. St Louis. MO) into sterile phosphate-buffered saline (PBS; 10) mM phosphate buffer. pH 7.2. 137 mM NaCl and 2.7 mM KCl) was added to flasks. The hepatocytes. fat droplet-bearing cells, and most fibroblasts rounded up during the first 5-7 min of incubation at 21 C and were gently shaken from the T-flask surface while most of the MMs remained attached. The MMs detached after additional incubation or more rapidly with the addition of fresh. 30 C 0.25% trypsin- EDTA. The enriched MM fraction was collected by centriiugation at 500 g for 3 min. suspended in 15 ml complete media (CM). Roswell Park Memorial Institute -1640 containing 7.5% v/v fetal bovine serum and penicillin (100 U/ml). streptomycin (100 /ml). and fungizone (0.25 g/ml). and allowed to settle. The more rapidly settling fraction containing most of the MMs was plated at various cell densities (0.5 10^sup 3^ to 3.0 10^sup 3^ cells/cm^sup 2^) in 25- cm- T-flasks or onto sterile Thermanox. 25-mm2 polystyrene roverslips in multiwell dishes. Cultures were maintained at 28-29 C in CM with media replacements at 4- to 6-d intervals. The resultant cultures contained 70-95% MMs and could be maintained without interfering fibroblast overgrowth or loss of MMs through detachment for 30-45 d.

The Buffalo Green Monkey Kidney (BGMK) cells originally obtained from T. Metcalf (Baylor University. Houston. TX) were maintained in CM at 37 C at 5% CO2. They were split 20-fold at 80% confluency. with media replacements at 5-d intervals. Cocultures of BGMK and MMs seeded at 4000 and 1000 cells/cm-, respectively, were incubated under standard conditions for MMs. Micrographs of the interactions of these cells were recorded 3 d after seeding.

Assessment of mycoplasma contamination was conducted by both aerobic broth culture at 37 C for 8 d and mycoplasma agar plate culture maintained in an anaerobic gas pack at 37 C. Cells were incubated in CM without antibiotics for 5 d before inocula- containing media and cells were transferred to assay tubes and plates along with control positive and negative media.

Phase contrast microscopy. A Nikon TMS phase microscope with camera was used for images of viable cells in culture. All phase images wrere at 120or 240-fold magnifications at indicated times. Length measurements were obtained using a calibrated ocular micrometer. The locations of study cells were marked on the underside of the culture flasks. Cablepodia elongation studies were conducted with marked cells viewed at 24-h intervals using a mark as index point.

Atomic force microscopy. After cell attachment and incubation for 7 d, coverslip cultures were removed, and the adherent cells fixed in equal volumes of 0 -20 C acetone-methanol for 5 min after three rinses in PBS and one rinse in distilled deionized water. The coverslips were allowed to drain and dry vertically in air at room temperature. Fixation was required to immobilize cablepodia to the substratum. No general changes in cell morphology were detected by light microscopy using this cell-fixation technique: however, many cablepodia fixed to the surface in broad curves indicating that the morphology was altered by surface tension changes during dehydration. Analyzing the phase contrast images before and after fixation, we have found that -20 C methanol-acetone fixation of individual cells on Thermanox coverslips resulted in small (

A Dimension 3100 AFM (Veeco-Uigital Instruments. Santa Barbara, CA) was used for all atomic force microscopy (AEM) topography measurements. The surfaces were imaged in contact mode using silicon nitride Ultralevers (Veeco-Digital Instruments) under ambient conditions. Deflection ami height images were collected. Height information measurements, related to thickness of cell appendages and cells, are typically accurate to about 0.1 nm. Imaging was performed with a calibrated J scanner, allowing the imaging of up to 132 132 m areas. Multiple overlapping 65 65 m images of adjacent areas were required to collect data on single MMs with cablepodia. The force of the lever applied to the cells was initially 5-10 nN. Image r\esolution was set to maximize pixel density for each image on the basis of the cell dimensions ami image size. Linear measurements were made from flattened images by the section Analysis component of the Digital software package.

RESULTS

Melanomacrophage structure. The MMs could he cultured at low densities (>1000 cells/cm^sup 2^ seeding density) for 4 ws postisolation when the initial numbers of contaminating hepatocytes, fibroblasts, and other cells were low. Individual MMs cultured under high- or low-density conditions produced large, extending, and retracting, ruffled lamellipodia and contained variable numbers of cytoplasmic melanosomes found in the cytosol or in discrete cell organelles or, most often, in nonzeiotic, irregular, sack-like pseudopodial structures or blebs containing high densities of melanosomes (Fig. 1A-D). These membrane blebs frequently were 23- 50% of the volume ol the cell, and detachment from the adherent cell occurred with increasing culture time resulting in MMs with considerably less total pigmentation. Individual MMs varied from approximately 40 to 80 m in cross-sectional diameters when adsorbed and spread on plastic cultureware. Eventually, MMs detached from the substratum and were lost in media replacements. Many cells developed short, thick (10-20 m diameter) pseudopodial extensions to adjacent cells (Fig. 1A-D). Some of these pseudopodial extensions would end 10-50 m from the cell body with the formation of a melanosome- containing lamellipodia process that was as large in diameter as the cell from which it originated (Fig. 1A). Often other, very narrow, linear processes, many with cross-sectional diameters at points just above the resolution of the optics, were observed by changes in the light refractility during phase microscopy (Fig. 1A-D). Some of these narrow processes ended at other cells and some did not.

Phase microscopy revealed that approximately half of the MMs generated remarkable, narrow process extensions that we have termed “cablepodia” (Fig. 2A-F). Cablepodia are denned as straight (less than 2 of variance from linearity), variably narrow (midpoints of 0.5-5 m). unbranching, and variably long (15-1,000+ m) extensions derived from the cell plasma membrane. Some appeared to be tubular and contained refractile materials. A foot process or cytoplasmic fan was observed at the distal end of the cablepod that attached to the surface of the plasticware or became lodged on another cell, resulting in cell-cell bridging. Cablepodia were either nonadherent or only weakly adherent to the substratum because media currents created by tilting the culture flask would dislodge cablepodia or move free processes but neither the cell soma nor the foot process detached. The unique property of cablepodia was that of straightness from the point of origin to termination. Despite amoeboid movements of these cells and extension and contraction of lamellipodia, the cablepodium was linear until directed movement had occurred and fibroblastic overgrowth or some surface object came into contact with the cablepodium causing bending (Fig. 1A). Cablepodia processes most often originated opposite lamellipodia (Fig. 2/1, C, F) and were not observed to arise from a lamellipodium or other cytoplasmic extension as in the case of filopodial dendritic neuntes. Many cablepodia appeared to originate from at or near the melanosome- containing membrane blebs and to pass across the top of the lamellipodia as they extended outward away from the cell body (Fig. ’2B). Several of the cablepodia contained bundles of dark, pigmented, 0.5- to 2-m diameter granules resembling melanosomes (Fig. 2E. F). Our preliminary time-lapse studies suggest directed movement of this material within the cablepodium.

FIG. 1. Culture and morphology of melanomorphages (MMs). Selectively trypsinized and differentially settled liver tissue cultures were seeded at 3 10^sup 3^ cells/cm^sup 2^. Phase photographs (X240) were taken at intervals after plating. (.4) 4 d; (B) 8 d; (C) 16 d: and (D) 21 d. Legend: c, cablepodium; e, turtle erythrocyte; f, hepatic fihrnhlast: h. hepatocyte with fat droplets; k, cablepodium extending over fibroblast to MM; l, MM lamellipodium; m, melanosome-containing membrane bleb; p, pseudopodium.

Cablepodia networks. Networks of cablepodia and other cell extensions were observed to connect MMs and liver fibroblasts (Fig. 1A-D). These networks frequently involved four to six cells but were observed to include as many as 30 cells. Within these networks, many different lengths of cablepodia were observed, some of which crossed over other cablepodia or extensions or entire fibroblasts (Fig. 1B and C). The networks imaged in panels A-D of Fig. 1 involve 5, 19, 15, and 17 cells, respectively.

Of 249 MMs from liver tissue from three turtles (2 females, 1 male), studied by phase microscopy, 115 developed cablepodia during the 14-d observation period from the onset of secondary culture from selective trypsinization. Of this population, the numbers of MMs with 1, 2, 3, 4, or 5 cablepodia were 87, 17. 9. 2. and 1, respectively. No cell developed more than 5 cablepodia.

The MMs (n = 100) with single cablepod extensions were studied by phase microscopy for a 14-d period. Most cablepodia (n = 87) ended with the distal process attached or fused to another cell. The largest numbers of attachments or fusions were to other MMs (n = 47), but many ended within fibroblasts (n = 20) or other unidentified liver cells (n = 20). In a study of cells generating two cablepodia (n = 30), observed during a 14-d period, 16 had both cablepodia attached to other cells, nine cells had one of the two cablepodia ending in one cell and the other had not reached a cell. Five cells had both cablepodia ending without cellular attachment. In several instances, two cablepodia arising from one MM attached to the same turtle hepatic (ibroblast or MM. In a study of mixed cell populations, individual BGMK cells were surrounded and engulfed by MMs as expected, but cablepodia from MMs not in contact with BGMK cells were not observed to make contact and adhere to BGMK cells (data not shown). These mixed cultures developed cablepodia that interacted with and bound to other MMs, hepatocytes, and fibroblasts in the cultures.

Typical cablepodia lengths measured using an ocular micrometer were between 50 and 500 m. The mean length of 40 cahlepodia extending from cells seeded at 300-600/cm^sup 2^ was 240 m (Table 1). The extremes were 50 and 900 m in this data set. Cablepodia lengths were a function of cell densities. When seeded at 600-1200 cells/cm^sup 2^, the mean length (n = 18) was 140 m with extremes of 30 and 400 m (Table 1). When 1200-2400 cells were seeded per cm^sup 2^, the mean length of cables (n = 18) was 75 m and the extremes were 15 and 200 m. In contrast to lamellipodia or other pseudopndiu. cablepodia were not observed to shorten or retract.

FIG. 2. The melanomacrophage (MM) cablepodia low-density culture. Trypsinized and settled MM cultures were seeded at 3 x 10^sup 2^ cells/cm^sup 2^. Photographs of single cablepodia were collected 7 d after plating. (A-C, E, F] x240; (D), x120. Legend: same as Fig. 1 and a indicates melanosomes or cell material in cablepodia. Bar, 40 m.

TABLE 1

EFFECT OF CELL SEEDING DENSITY ON CABLEPODIUM LENGTH (10 D AFTER SEEDING)a

The AFM imaging. The AFM imaging of fixed cells revealed surface features of cablepodia (Fig. 3A-F). Deflection images (Fig. 3A-C) are of an MM with cablepodium of 560 m total length, whereas images (Fig. 3D-F) are of an MM with cablepodium of 480 m total length. In image 3A and D. each cablepodium origin in the MM is shown. In Fig. 3A. there is a gradual narrowing of the origin from an apparent growth rone over a 65 m length. In Fig. 3D, the origin narrows over 20 m to a relatively constant, narrow diameter.

In Fig. 3B and E, images of the midpoints of rablepodia are shown. In Fig. 3B, the fixed cahlepodium has a cross-sectional diameter of approximately 600 nm and an average height on the basis of section analysis of 125 nm for a cross-sectional area of 0.075 m^sup 2^. In Fig. 3E, a section with increased diameter and height (160 nm) and less uniform surface features is observed. There were no projections or spikes found originating from cablepodia. Higher AFM magnifications (x and y axes of 2 m) of the midpoints of cablepodia did not reveal repeating structure or fibers.

In Fig. 3C and F, images of the foot process are shown. In Fig. 3C. the foot process ends on a fibroblast. the nucleus of which is observed at the image periphery. The uniform or smooth membrane edges in this image suggest a flowing together or fusing of the cablepodium with the fibroblast. The oval region above the fan may have been the location of a large melanosome-containing bleb lost during fixation. In Fig. 3F, the fan or foot process of the end of the cablepod is observed. This fan is not associated with another cell. A portion of a raised, melanosome-containing bleb is observed opposite the spreading fan of cytoplasm. Fan margins contained sublight microscopic, filopodia-like extensions (linear or branched) of up to 25 m length and cross sections of 60-180 nm (Fig. 3F; Table 2). These may be involved in adherence and cell recognition. Such fibrillar structures were not observed in cablepodia fans attached to other cells (Fig. 3C).

FIG. 3. The AFM imaging of cablepodia structure. The melanornacrophages from selectively trypsmized and settled populations were seeded onto coverslips, incubated for 7 d. and fixed, as described in Materials and Methods. Height and deflection images were collected using contact mode AFM. Deflection images of two cablepodia were collected. Origins (A arid D): Central region (approximately midway from the cell body and distal end) (B and E): and fan or foot process (C and F) are presented. Image x and y axes are each 65 m. Legend\: same as in Figs. 1 and 2; fl indicates foot process filaments; n, nucleus of fibroblast: db, region of detached melanosome-containing membrane bleb; o, origin of cablepodium. Bar. 10 m.

TABLE 2

CHARACTERIZATION OF CABLEPODIA ANCHORAGE PAD FIBERS IDENTIFIED BY AFM

Cablepodia midpoint diameters as determined by AFM section analysis varied by a factor of 1.4, whereas length varied by more than a factor of 60 (Table 3). Length was influenced by the spatial arrangement of adjacent cells. With cablepodia of 80 m or longer (n = 18), the mean diameter 20 m from the cell origin was 970 nm with extremes of 750 and 1350 nm. At the midpoint of these same filaments. the mean diameter was 610 urn and the extremes were 020 and 820 nm. At 20 m from the insertion into u cell or from the end of the cable, if there was no insertion, the mean diameter was 590 nm with extremes of 530-710 nm. Some variability occurred because of bulges occurring at random intervals along the cablepodia; otherwise, the filaments were quite uniform (Fig. 3B and E).

Rate of cablepodia growth and inhibition of fibrobalst division. The mean cell generation time for fibroblasts in mixed cultures of MMs, fibroblasts, and hepatocytes, under the selected culture conditions, was 32 h as determined by studies of individual adherent cells marked on cultureware. As found in our previous studies. MMs were not observed to undergo mitosis. Fibroblasts with single, attached cablepodia (n = 80) were studied for cell division and morphology changes. Among these cablepodia-attached fibroblasts, none underwent cell division during a 10-d observation period, although the remaining population of fibroblasts in the same flask divided regularly. Morphologically, the fibroblastic cells became senescent but did not become apoptotic. Cell zeiotic blebbing suggestive of apoptosis was not observed in any of the cablepodia- attached fibroblasts.

The rate of growth of cablepodia was determined from ocular micrometer measurements with time (Table 4). Fields of MMs were studied after selective trypsinization and plating. Study fields were selected on the basis of cell density and examined for the initiation of cablepodia. Three cells were selected and the length of a single cablepodium from each cell determined in repeated measurements as a function of time. The maximum lengthening was 115 m in a 24-h period when studied during a period of 5 d beginning 2 d after seeding. Minimum lengthening was 15 m. None of these selected cablepodia terminated at another cell during the study period. The extension was maximal on the fifth day of culture after seeding.

TABLE 3

AFM DETERMINED CABLEPODIA DIAMETERS IN CABLEPODIA LONGER THAN 80 m. (RANGE 80-870 m) n = 18a

TABLE 4

CABLEPODIA EXTENSION FROM REFERENCE INDEX POINT(a)

DISCUSSION ?

We have used a protocol for MM isolation from liver tissue of lnrtles that results in enriched cultures of MMs and reduced contamination by hepatocytes, fibroblasts, and other cells. The MMs produced by these culture methods produced long, slender cell extension processes, resembling cables, that we have termed “cablepodia.” These cell processes have not been identified through histological procedures for these cells, in vivo, but because of their size and structure, they may not be visible in typical light microscopy of fixed, stained tissue or because of the close association of cells within MM centers, cablepodia may not form.

The principal physical properties of cablepodia include straightness, variable lengths to approximately 1000 m. elongation continuing through 7 d of culture, lack of branching, a distal foot process involved in anchorage to cells, little or no adherence to the cultureware substratum, midpoint diameters as small as 500 nm, and occasional bulges apparently containing cell materials.

The cablepodia differ morphologically from filopodia of neurites and skin melanocytes by not being branched and by the cell location where initiation occurs. Whereas filopodia often are produced at the distal edge of lamellipodia, cablepodia initiate most often from the cell body, opposite or sometimes adjacent to active lamellipodia. Filopodia are relatively short extensions, whereas cablepodia may extend for great distances (10-15 cell diameters or more) before they attach and insert into another cell or end with a small foot process adherent to the substratum. Mammalian Kupffer cell filopodia may also be branched and have also been postulated to have nonanchorage functions including that of communication between nearby Kupffer cells (Muto, 1975; Polliack et al., 1978). In preliminary immunohistochemical studies with antiactin and anti- myosin, we found that MM lamellipodia contained material that was bound by these antibodies, but the cablepodia were not labeled and the fibrils found at the margins of the foot processes could not be identified by light microscopy for this type of study.

The rate of growth of microtubules of Heliosoma neurons has been measured at 10-14 m/h (Zhou et al., 2002). The average growth of rat hippocampal axons was determined to be 115 m/d. Similarly, cablepodia elongation was found to be between 15 and 115 m/24 h period with the greatest extension found for 3 cablepodia on the fifth day of culture. Neunte elongation is accompanied by an extensive remodeling of the cytoskelelon. a process as yet unknown in MM cablepodia.

Other dendritic and pigmented dendritic cells have been imaged by AFM revealing both external and internal structure. Gluteraldehyde- fixed and living neurons and glial cells were imaged to study neural growth cones and the structure of resultant filopodia (Parpura et al., 1993). Human skin melanocytes were imaged to reveal the structure of the branched network of dendritic filopodia bifurcating close to the cell soma (Zhang et al., 2004). They observed melanosomes in melanocyte filopodia and suggested that movement of melanosomes occurs through the filopodia to the distal margin where exocytosis occurred to keratinocytes. This was suggested as a means for melanocyte-keratinocyte transfer of melanosomes but may be only one of several mechanisms for melanosome transfer (Scott et al., 2002). Invadopodia are also slender, straight, and long processes derived from immortal, transformed glioma cells, which bear metalloprotease functions (Chen and Wang, 1999; Fillmore et al.. 2003). These processes were postulated to be involved in metastases of glioma cells by dissolving collagens thereby permitting the cells to enter into different organ compartments.

Functional properties of cablepodia are suggested by the foot process anchorage to other cells and the cessation of contacted cell mitosis, by the possible directed movement of material within the cablepodium. and by the observation that MMs form centers. From this study, the MM cablepodia may communicate signals, which switch off hepatocyte-fibroblast mitosis and possibly transport materials or melanosomes to other cells and recognize conspecific BGMK cells. Granulocyte-macrophage colony-stimulating factor, macrophage colony- stimulating factor, or recombinant alpha interferon were not observed to result in enhanced colony formation or survival of turtle hepatocytes, fibroblasts, or MMs. and these materials did not result in the division of cablepodia-attached fibroblasts in culture (Rund et al., 1998). In another study, interleukin-2 was not found to enhance proliferation of turtle peripheral blood lymphocytes (Ulsh et al., 2000). It is probable that membrane receptors for testudine cytokines poorly recognize these human recombinant molecules. However, homologous sera did stimulate division of turtle lymphocytes suggesting growth factors exist and may be isolated. Nonetheless, the MM to fibroblast interaction through cablepodia resulting in inhibition of fibroblast mitosis may become a useful model for study of contact inhibition of cell division that is not density dependent. Thus, the list of functions ascribed to MMs now include those of superoxide and free radical removal, melanin synthesis, phagocytosis of bacteria and eukaryotic organisms, end processing of red blood cells, formation of centers, formation of cablepodia, and a contact-dependent but nondensity-dependent inhibition of hepatic fibroblast mitosis. Just how these properties and cells interact within the context of the testudine immune system to promote turtle survival and relative freedom from cancers remain to be established.

ACKNOWLEDGMENTS

We acknowledge the support of Kori Radke, Curtis Mosher, and Juntan Xu for helpful discussions and critical readings of this manuscript. This work was supported in part by a grant from the Iowa Space Grant Consortium. Ames, Iowa.

REFERENCES

Agius, C. Phylogenetic development of meleno-macrophage centers in fish. J. Zool. 191:11-31; 1980.

Agius, C.; Roberts, R. J. Melano-macrophage centres and their role in fish pathology. J. Fish. Dis. 26:499-509; 2003.

Barni, S.; Bertone, V.; Croce, A. C.; Bottiroli, G.; Bernini, K: Gerzeli, G. Increase in liver pigmentation during natural hibernation in some amphibians. J. Anat. 195:19-25; 1999.

Belcourt, D. R.; Okawara, Y.; Fryer, J. N.; Bennett, H. P. Immunocytochemical localization of granulin-1 to mononuclear phagocytic cells of the teleost fish Cyprinus carpio and Garassius auratus. J. Leukoc. Biol. 57:94-100; 1995.

Chen, W. T.; Wang, J. Y. Specialized surface protrusions of invasive cells. invadopodia and lamellipodia, have differential MT1- MMP, MMP-2. and TIMP-2 localization. Ann. N. Y. Acad. Sci. 878:361- 371; 1999.

Christiansen, J. L; Grzybowski, J.; Kodama, R. M. Melanomacrophage aggregations and their age relationships in the yellow mud turtle. Kinosternon flaveseens (Kinosternidae). Pigment Cell Res. 9:185-190; 1996.

Cicero, R.; Scuito, S.; Chillemi, R.; Sichel, G. Melanosynthesis in the Kupffer cells of amphibia. Comp. Biochem. Physiol. A. 73:477- 479; 1982.

Corsaro, C.\; Scalia, M.; Leotta, N.; Mondio, F.; Sichel, G. Characterization of Kupffer cells in some Amphibia. J. Anat. 196:249- 261; 2000.

Ellis, A. E.; Monroe, A. L. S.; Roberts, R. J. Defense mechanisms in fish. 1. A study of the phagocytic system and the fate of intraperitoneally injected particulate material in the plaice (Pleuronectes platessa L.). J. Fish. Biol. 8:67-78; 1976.

Fillmore, H. L.; Cha-siotis, I.; Cho, S. W.; Gillies, G. T. Atomic force microscopy observations of tumor cell invadopodia: novel cellular nanomorphologies on collagen substrates. Nanotechnology 14:73-76; 2003.

Fournier-Betz, V.; Quentel, C.; Lamour, F.; LeVen, A. Immunocytoehemical detection of Ig-positive cells in blood, lymphoid organs and the gut associated lymphoid tissue of the turbot (Scophthalmus maximus). Fish Shellfish Immunol. 10:187-202: 2000.

Gallone, A.; Guida, G.; Maida, I.; Cicero, R. Spleen and liver pigmented macrophages of Rana esculenta L. A new melanogenic system? Pigment Cell Res. 15:32-40; 202.

Guida, G.; Gallone, A.; Maida, I.; Boffoli, D.; Cicero R. Tyrosinase gene expression in the Kupffer cells of Rana esculenta L. Pigment Cell Res. 13:431-435; 2000.

Guida, G.; Zanna, P.; Gallone, A.; Argenzio, E.; Cicero, R. Melanogenic response of the Kupffer cells of Rana esculenta L. to melanocyte stimulating hormone. Pigment Cell Res. 17:128-134; 2004.

Johnson, J. C.; Schwiesow, T.; Ekwall, A K.; Christiansen, J. L. Reptilian melanomacrophages function under conditions of hypothermia: observations on phagocytic behavior. Pigment Cell Res. 12:376-382; 1999.

Muto. M. A. Scanning electron microscopic study on endothelial cells and Kupffer cells in rat liver sinusoids. Arch. Histol. Jpn. 37:369-389; 1975.

Parpura, V.; Haydon, P. G.; Henderson, E. Three-dimensional imaging of living neurons and glia with the atomic force microscope. J. Cell Sci. 104:427-132; 19911

Polliack, A.; Gamliel, H.; Hershko, C.; Knook, D. L.; Slevster, E. C. Surface morphology and ultrastructure of isolated hepatic Kupffer and endothelial cells. Biomedicine 29:268-272; 1978.

Rund, C. R.; Christiansen, J. L.; Johnson, J. C. In vitro culture of melanomacrophages from the spleen and liver of turtles: comments on melanomacrophage morphology. Pigment Cell Res. 11:114-119; 1998.

Sanders, G. E.; Batts, W. N.; Winton, J. K. Susceptibility of zebrafish (Danio rerio) to a model pathogen, spring viremia of carp virus. Comp. Med. 53:514-521; 2003.

Scalia, M.; Geremia, E.; Corsaro, C.; Santoro, C.; Scuito, S.; Sichel, G. The extracutaneous pigmentary system: evidence for the melanosynthesis in amphibia and reptilia liver. Comp. Biochem. Physiol. 89B:715-717; 1988.

Sciuto, S.; Chillemi, R.; Patti, A.; Sichel, G.; Sealia, M. Melanosomes from liver and skin of Rana esculenta L. A comparative chemical study. Comp. Biochem. Physiol. B. 90:397-400; 1988.

Scott, G.; Leopardi, S.; Printup, S.; Madden, B. C. Filopodia are conduits for melanosome transfer to keratinocytes. J. Cell Sci. 115:1441-1451; 2002.

Sichel, G. Qualche osservazione sulle cellule pigmentate del fegato del Rettili. Mon. Zool. Italy 57:3-7; 1984.

Sichel, G.; Sealia, M.; Corsaro, C. Amphibia kupffer cells. Microsc. Res. Tech. 57:477-490; 2002.

Sichel, G.; Biondi, E. Osservazioni sulle cellule pigmentate del fegato di Rana esculenta L. Boll Acead Gioenia Sci Nat Catana 10:875- 877; 1971.

Ulsh, B. A.; Congdon,. J. D.; Hinton, T. G.; Whicker, F. W.; Bedford, J. S. Culture methods for turtle lymphocytes. Methods Cell. Sci. 22:28.5-972; 2000.

Zhang, R. Z.; Zhu, W. Y.; Xia, M. Y.; Feng, Y. Morphology of cultured human epidermal melanocytes observed by atomic force microscopy. Pigment Cell Res. 17:62-65; 2004.

Zhou, F-Q.; Waterman-Storer, C. M.; Cohan, C. S. Focal loss of actin bundles causes microtubule redistribution and growth cone turning. J. Cell Biol. 157:839-849; 2002.

JAMES C. JOHNSON,1 SAJU R. NETTIKADAN, SRIKANTH G. VENGASANDRA, SAI LOVAN, JAMES MUYS, ERIC HENDERSON, AND JAMES CHRISTIANSEN

Department of Microbiology, Des Moines University, Des Moines, Iowa 50312 (J. C. J.). BioForce Nanosciences Inc.. 1615 Golden Aspen Drive, Suite 101. Ames, loma 50010 (S. R. N.. S. G. V., J. M., E. H.). Department of Biology. Drake University. Des Moines. Iowa 50314 (S. L., J. C.). and Department of Genetics. Development, and Cell Biology State University, Ames, Iowa 50011 (E. H.)

(Received 26 April 2005; accepted 22 May 2005)

1 To whom correspondence should he addressed at E-mail: James. johnson@dmu.edu

Copyright Society for In Vitro Biology Jul/Aug 2005