Persistence of tick-derived Anaplasma marginale in cultured bovine turbinate and endothelial cells

Des monocouches de cellules bovines des cornets nasaux (cellules CN) et de cellules bovines endothéliales ont été inoculées avec des Anaplasma marginale dérivés de glandes salivaires de Dermacentor andersoni. Des passages des couches ont été faits à des intervalles de 2 ou 4 semaines et examinés aux microscopes classique et électronique, ainsi que par une sonde d’ADN spécifique pour A. marginale. Des inclusions intracellulaires ont été observées dans les cellules CN aorès 2-4 semaines. Le nombre de cellules avec des inclusions auamentait au cours de 1-2 semaines, ensuite il y avait une disparition graduelle. Un fragment du gène msplB du stade érythrocytaire d’A. &arginaZe, marqué par isotope radioactif, a hybridé avec de I’ADN extrait de cultures de cellules CN jusqu’à 7 semaines après inoculation (passage 4). Des rickettsies individuelles ont été observées au microscope électronique dans des prélèvements faits à ce moment. Des veaux sensibles inoculés avec des cultures susoectes n’ont aas montré d’anaplasmose clinique, mais ont développé des titres sig&catifs d’anticorps détectés par ELISA. De I’ADN de cultures de cellules endothélides 9 semaines aurès inoculation s’est également lié à la sonde spécifique pour A. mar&ale. II semble qu’Ana~lasma marginale des glandes salivaires de D. andersoni persiste en culture dans des cellules bovines CN et endothéliales, mais qu’il n’y a pas de développement normal ni infectiosité pour les bovins. When established mammalian and insect cell lines were incubated with infected erythrocytes, cells and organisms were taken up by some cell types and survived for extended periods of time ; further development was not apparent (8, 12). The tick gut stage of A. marginale was used to infect an embryonic tick cell line (3). This stage was reported to infect and grow within these cells but cultures were not infective for cattle, and host cells eventually destroyed the intracellular parasites through lysosomal digestion (3, 4). In this study salivary glands infected with A. marginale were used as inoculum for cultured cells. This stage is most likely transmitted to the vertebrate host during tick feeding. Manipulation of the feeding schedule of D. andersoni males resulted in large numbers of A. marginale colonies in salivary glands (7). We have recently shown that the salivary gland stage Will infect bovine erythrocytes in vitro, although further development does not occur (2). This report describes results of attempts to establish infections of A. marginale using the salivary gland stage to inoculate bovine turbinate and endothelial cell cultures.

When established mammalian and insect cell lines were incubated with infected erythrocytes, cells and organisms were taken up by some cell types and survived for extended periods of time ; further development was not apparent (8, 12). The tick gut stage of A. marginale was used to infect an embryonic tick cell line (3). This stage was reported to infect and grow within these cells but cultures were not infective for cattle, and host cells eventually destroyed the intracellular parasites through lysosomal digestion (3, 4). In this study salivary glands infected with A. marginale were used as inoculum for cultured cells. This stage is most likely transmitted to the vertebrate host during tick feeding. Manipulation of the feeding schedule of D. andersoni males resulted in large numbers of A. marginale colonies in salivary glands (7). We have recently shown that the salivary gland stage Will infect bovine erythrocytes in vitro, although further development does not occur (2). This report describes results of attempts to establish infections of A. marginale using the salivary gland stage to inoculate bovine turbinate and endothelial cell cultures.

INTRODUCTION
Anaplasma marginale is a tick-borne rickettsial organism (Rickettsiales : Anaplasmataceae) that infects erythrocytes of cattle and causes significant mottality and production losses in many parts of the world (11). The life cycle of A. marginale also involves developmental stages in the tick vector where a complex developmental sequence occurs in gut, gut muscle and salivary gland cells (5). One of the major constraints in anaplasmosis research has been lack of a continuous in vitro culture system. Most attempts to cultivate the organism have involved the erythrocytic stage of A. marginale. Organisms were found to be viable and retain infectivity in whole erythrocyte cultures but replication or futther development did not occur after approximately 48 hours (8).

Infection of ticks
Dermacentor andersoni males, reared at Oklahoma State University, were placed in orthopedic stockinettes attached to donor calves when parasiternia reached 3-5 %. Ticks were allowed to feed for seven days, after which they were removed and placed in a humidity chamber (90-98 % RH) at 25" C for 5 days. Ticks were then allowed to feed on a second, susceptible calf for 12 days and removed. Uninfected male D. andersoni were fed on a separate, susceptible calf for 12 days in similar fashion to provide uninfected control ticks. Samples from all batches of ticks were collected and salivary glands were examined by light and electron microscopy (LM and EM) for presence of A. marginale. Salivary glands were dissected from individual ticks and placed in sterile Minimum Essential Medium (MEM) on ice. Glands were transferred to a sterile glass tissue grinder and suspended in complete medium to a volume of 1 ml medium/ten pairs of glands and homogenized.
Crude homogenate was transferred to centrifuge tubes and spun at 1000 x g for 10 min at 4 "C. The supernatant was removed and immediately inoculated onto cell monolayers. Samples of supernatants and pellets were fixed for EM to confirm presence and morphology of A. marginale.

Inoculation of cell lines
Established monolayers of bovine turbinate cells obtained from Washington State University (T. Crawford) and bovine endothelium (aorta) from ATCC (# CPAE CCL209) were passaged three days prior to inoculation into 25 cm 2 flasks and maintained in Glasgow MEM supplemented with 15 % calf serum, 5 % tryptose phosphate broth and penicillin/streptomycin (100 lu/100 ug/ml) at pH 6.8-7.0. The volume of medium in each flask was reduced to 2 ml to which 1 ml supernatant from infected or uninfected salivary gland centrifuged homogenate was added. Flasks were placed on a rotator in a 37°C incubator and shaken slowly (80 rpm) for 30 min to disseminate the inoculum. After 2 h medium was removed from cultures and monolayers were washed and then replenished to 5 ml with fresh medium.

Maintenance of cultures
Culture medium was replenished in cultures every 3 days by replacing 2/3 of the old medium. Cell layers were passaged at 2 or 4 week intervals by trypsinizing and splitting monolayers 1:3.

Light microscopy
Culture flasks were examined daily for the presence of inclusion bodies. Smears were made from samples taken from monolayers with a sterile needle and stained with a modified Wright stain prior to being passaged.

Electron microscopy
Portions of monolayers were removed from flasks with a cell scraper and fixed in cold 2 % glutaraldehyde in 0.1 M sodium cacodylate buffer and processed according to procedures of KOCAN et al. (6). Ultrathin (silver-reflecti-~ ve) sections were tut on a Sorval MT 5000 ultramicroto-~ me with a Diatome diamond knife and stained with uranyl acetate and lead citrate (13). Sections were observed on a JEOL 100 CX electron microscope.

DNA hybridization
The 965 bp Hind IIIIXhol fragment (1) from within the msplB gene of the erythrocytic stage of A. marginale (Virginia isolate) was amplified by the polymerase chain reaction and cloned into the vector pBluescriptll SK (Stratagene). For nucleic acid hybridizations, the 965 bp fragment was purified from low melting temperature agarose and radiolabeled with 32P-dCTP. Culture samples (approximately 1000 cells) were centrifuged (2000 x g, 10 min), and cell pellets were washed three times with sterile phosphate buffered saline. Cells were lysed by incubating with 0.35 ml digestion buffer (10 mM Tris-HCL (pH 8), 1 mM EDTA, 0.5 % SDS, 0.2 mg/mI proteinase K) at 56 "C for 3 h. Protein and cell debris were removed by phenollchloroform extraction. DNA was precipitated in 100 ul TE (10 mM Tris-HCL, 1 mM EDTA, pH8).
Serial dilutions of each sample were prepared, SO that three dilutions were analyzed by DNA hybridization (l:l, l:lO, 1:lOO). DNA samples were treated and bound to nitrocellulose using a slot blot manifold (Schleicher &. Schuell lnc.), according to manufacturer's instructions. Blots were hybridized with the radiolabeled mspll3 fragment and washed under stringent conditions (at 72"C), as described by MURPHY and DALLAS (9), then exposed to X-Ray film overnight.

Animal inoculation
Cultures of turbinate cells from which samples hybridized to the A. marginale DNA probe or that were found to harbor rickettsial organisms were used to inoculate susceptible calves to determine infectivity of culture material. Cells in 2 ml of MEM medium from 2 suspect flasks were inoculated IV into calves. Five animals were inoculated with cultures collected at different times post inoculation (PI) (table 1). Calves were monitored for presente of intraerythrocytic parasites in blood smears and for clinical signs of anaplasmosis.
Serum samples were collected before and after inoculation and evaluated by the anaplasmosis complement fixation test (Oklahoma Animal Disease Diagnostic Lab.). An ELISA test developed previously to detect antibody to erythrocyte, tick gut and tick salivary gland stages of A.marginale was used to screen serum samples collected before and after inoculation (10). Fifty male D. andersoni were fed on one of the inoculated calves (PA 145) at 10 weeks PI, to attempt tick infection, and subsequently refed on a susceptible calf (PA 163) to test, for the transmission STVM-93 of A. marginale. All animals were challenged-exposed with either infected A. marginale blood or ticks to prove their susceptibility. observed in either infected or uninfected tick salivary glands.

Infection of ticks
Salivary glands from all groups of infected ticks used were found to be infected with colonies of A. marginale by LM and EM. Calves on which these ticks fed developed clinical anaplasmosis. Colonies were not observed in uninfected control ticks nor did calves they fed on develop anaplasmosis. Symbiotic rickettsiae were not Inoculum Supernatant from infected glands collected after centrifugation contained many individual A. marginale as well as mitochondria, secretory granules and other tick cell components (photo la, b). A few intact A. marginale colonies, as well as smaller groups of rickettsiae were also seen in each inoculum sample. Anaplasma marginale in the inoculum appeared to be morphologically intact and binaty fission was apparent. Colonies of A. marginale were seen in pellets recovered after centrifugation, along with tick tissue and organelles. Anaplasma were not seen in inocula from any uninfected gland preparations. Light microscopy and electron microscopy observed after 18 hours PI. Rickettsiae became polymorphic and changed from reticulated forms to denser forms. At two weeks PI round intracytoplasmic inclusions were In samples collected at 60 min PI rickettsiae were seen observed in turbinate cells (photo 3a). These inclusions attached to host cell membranes (photo 2a). Colonies increased in number for 1-2 weeks, contained several were also observed in association with host cells with subunits and were often observed in close association individual rickettsiae adhered to the host cell membrane with the nucleus (photo 3a, b). With EM, inclusions were (photo 2b). Changes in rickettsial morphology were found to contain a variable number of subunits with dense Photo 2 : Electron micrographs qf bovine turbinate ce11 cultures at 60 min PI with tick salivary gland material infected with A. marginale. a) An individual Anaplasma organism (A) in contact with a host ce11 membrane (x 60,780). 6) A colony (arrows) of A. marginale in association with a bovine turbinate ce11 membrane (x 17, 720).

STVM-93
interna1 material (photo 4). In some cultures these inclusions were not seen until 6 weeks Pl. After the initial increase, the number of inclusions decreased as cultures were passaged and eventually could be found only occasionally. In older cultures inclusions and subunits appeared to be devoid of interna1 structure but rickettsial organisms were occcasionally seen within the inclusion (photo 5). Phagocytic vacuoles were abundant in endothelial cultures and obscured any other material in the host cell cytoplasm during the first l-2 weeks. After approximately 10 days inclusions were found but did not increase in numbers following passage as was observed in turbinate cells. Similar inclusions were not observed in control cultures. Colony formation or rickettsial replication were not observed during culture passage. Individual rickettsiae were identified in some turbinate cultures with EM ; these organisms had a reticulated tore with a denser peripheral area and were variable in size (photo 6). The plasma membrane and cell wall were separated and a single large vacuole was often present within rickettsiae (photo 6). The number of organisms identified in each sample was small but persisted after passage and organisms were identified after 6 weeks Pl. Organisms were not seen in some samples that hybridized to the A. marginale-specific DNA probe, however amorphous inclusions with reticulated areas resembling rickettsial DNA were seen in several of these samples (photo 7).

DNA hybridization
Culture samples with inclusion bodies, as well as those with distinct rickettsial organisms, contained DNA that hybridized to the A. marginale-specific DNA probe. Probe-positive DNA was present in endothelial cultures after four passages at 6 weeks PI (photo 8) and beyond 9 weeks PI in unpassaged cultures. In turbinate cultures probe-positive DNA was present after four passages at seven weeks PI (photo 8). Approximately 10 % of turbinate culture flasks sampled and 15l % of endothelial cultures hybridized to the A. marginale DNA probe. Control samples of bovine erythrocytes and tick salivary glands infected with A. marginale were DNA probe positive while uninfected erythrocytes and tick salivary glands, and turbinate and endothelial cell cultures, inoculated with uninfected material were negative.

Calf inoculation
Calves that were inoculated with bovine turbinate culture material did not develop clinical anaplasmosis, their sera were negative by the complement fixation test and they proved to be fully susceptible to challenge-exposure (table 1). Small numbers of intraerythrocytic organisms were seen in each calf as early as 7 days PI and calves had a decreased percentage in packed cell volume (PCV) of 4-27 % after inoculation (table 1). When sera from these animals were evaluated with ELISA, 3 of 5 animals had significant antibody titers to different stages of A. marginale (table Il). Anaplasma marginale was not seen in ticks which had fed on PA 145 for attempted tick infection and they did not cause clinical anaplasmosis in calf (PA 163) used for transmission feeding of these ticks although small numbers of intraerythrocytic inclusions were seen for 4 weeks. PA 163 proved to be susceptible upon challenge ; however peak parasiternia was Iow (17 %) and percent reduction in PCV was only 17 %.

DISCUSSION
Monitoring salivary glands of D. andersoni with LM and DNA probe for presence of A. marginale colonies after feeding and during isolation assured that all cultures received organisms in inoculum. Isolation of individual rickettsiae from colonies and other host cell material initially involved longer preparation time, differential and density gradient centrifugations, and sonication. Although more of the salivary gland material was extracted, the number of A. marginale recovered after each step decreased. Host cell fragments and organelles were present in semi-purified preparations. With extended preparation time, rickettsiae appeared to change morphologically ; the central chromatin became more diffuse and disappeared in some cases. Prolonged manipulation of released parasites may have had a detrimental effect on the rickettsiae by reducing their infectivity. Mechanical disruption salivary glands was sufficient to release many individual rickettsiae, leaving some larger groups, a few intact colonies, and host cell material. Rickettsiae recovered using minimal prepa- ration were found to be morphologically intact and, within colonies, evidence of replication by binary fission was apparent.
It is likely that only organisms which had entered host cells persisted in culture since the inoculum was removed after 2 hours and monolayers were washed and later trypsinized. The appearance and subsequent proliferation of intracellular inclusion bodies in turbinate cells suggested a developmental stage of the organism. The extended period before appearance (2-6 weeks PI) may be necessary for transition and accommodation of the parasite to the host cell. Although inclusions and their subunits were not typical of Anaplasma, samples containing these forms hybribized with the A. marginale-specific DNA probe. Some of these inclusions did contain distinct rickettsiae which may have originated from inclusions. Because distinct transitional stages were not identified it is also possible that rickettsia from the inoculum survived in host cells without further development.
Parasites may also change form in culture. Amorphous inclusions, represented by photo 7, contained areas resembling reticulated chromatin and were the only suspicious bodies found in some culture samples that hybridized with the A. marginale-specific DNA probe.
Inoculation of susceptible calves with infected cultures did not result in clinical anaplasmosis or serologic conversion as determined by the anaplasmosis CF test but intraerythrocytic A. marginale were observed in all 5 animals. However, 3 calves had significant levels of specific antibodies to different stages of the parasite, suggesting that they were exposed to A. marginale antigens. Since we know that infected salivary gland material Will produce cli-