Structural responses of pulmonary intravascular macrophages in lentivirus-infected and/or recombinant ovine interferon-τ–treated lambs

Structural Responses of PulmonaryIntravascular Macrophages in

Lentivirus-Infected and/orRecombinant Ovine

Interferon-t–Treated LambsBALJIT SINGH,1 TROY L. OTT,2,3 FULLER W. BAZER,2,3

AND ANDRES DE LA CONCHA-BERMEJILLO1*1Department of Veterinary Pathobiology, Texas A&M University Agricultural

Experiment Station, San Angelo, Texas 769012Department of Animal Sciences, Texas A&M University, College Station, Texas 778433Institute of Biosciences and Technology, Texas A&M University, Houston, Texas 77030

ABSTRACTOvine lentivirus (OvLV), a retrovirus, infects and disseminates to various

tissue organs via monocytes. The differentiation of infected monocytes intomacrophages is a prerequisite for viral replication, and the presence ofinfected macrophages in tissue organs induces chronic immunopathologysuch as lymphoid interstitial pneumonia. The pulmonary intravascularmacrophage (PIM) is a recently identified mononuclear phagocyte in domesticanimal species, including sheep. Recombinant ovine interferon-tau (roIFN-t),a type I IFN originally named as the ovine trophoblast protein, has potentantiviral activity against OvLV and human immunodeficiency virus andprevents the development of OvLV-associated lung pathology.

We investigated and compared the structural features of PIMs inOvLV-infected and/or roIFN-t–treated 1-month-old lambs using transmissionelectron microscopy. The PIMs’ numerical counts were performed in toluidineblue–stained sections of Epoxy-embedded lung tissues.

A reduction in the number of PIMs was observed with OvLV infectionand/or roIFN-t treatment of lambs as compared to the control group (P #0.05). The majority of the PIMs in OvLV-infected and/or roIFN-t–treatedgroups were devoid of their surface coat. The PIMs of OvLV-infected lambsexhibited signs of biosynthetic activation such as expanded rough endoplas-mic reticulum, prominent Golgi complexes, and accumulation of secretoryvesicles. A few PIMs contained OvLV-like structures. In roIFN-t–treatedOvLV-infected lambs, the lymphocytes had ruffled plasma membranes andwere in intimate contact with the PIMs, as is observed during cytotoxiccell-mediated killing of target cells. Most of the PIMs in roIFN-t–treatedOvLV-infected lambs appeared smaller in size.

Ovine lentivirus and roIFN-t, individually or in combination, alter theintegrity of the surface coat of PIMs and cause their disappearance from thelungs. Ovine lentivirus infection induces morphological changes that corre-late with cytotoxic cell behavior between lymphocytes and PIMs in roIFN-t–treated or placebo-treated lambs. The loss of PIMs, probably infectedwith OvLV, either through direct killing by roIFN-t or indirectly by roIFN-t–

Grant sponsor: USDA; Grant sponsor: TexasAgricultural Experi-ment Station.

Baljit Singh’s current address is Department of Physiology andCellular Biophysics, Columbia University College of Physicians,St. Luke’s–Roosevelt Hospital, 1000 Tenth Avenue, New York, NY10019.

*Correspondence to: Andres de la Concha-Bermejillo, TexasA&M University Agricultural Experiment Station, 7887 US Hwy87 North, San Angelo, TX 76901. E-mail: [email protected]

Received 10 September 1997; Accepted 20 March 1998

THE ANATOMICAL RECORD 251:472–485 (1998)

r 1998 WILEY-LISS, INC.

activated cytotoxic T lymphocytes may represent different aspects of therapeuticactions of this cytokine. Anat. Rec. 251:472–485, 1998. r 1998 Wiley-Liss, Inc.

Key words: cytotoxicity; interferon-t; lymphocytes; lymphoid interstitialpneumonia; ovine lentivirus; pulmonary intravascular macro-phages; surface coat

Ovine lentivirus (OvLV) belongs to the family Retroviri-dae and shares close structural and functional homologywith the human immunodeficiency virus (Clements andZink, 1996; (de la Concha-Bermejillo et al., 1995; de laConcha-Bermejillo, 1997). Monocytes are believed to serveas the primary cellular targets of OvLV infection as well asa vehicle for the dissemination of the virus to variousorgans in the host (DeMartini et al., 1993; Peluso et al.,1985). Differentiation of OvLV-infected monocytes intomacrophages is a prerequisite for the initiation of viralreplication (Johnstone, 1988; DeMartini et al., 1993). Thepresence of OvLV-infected macrophages in various organsattracts inflammatory cells and causes chronic immunopa-thology such as lymphoid interstitial pneumonia (Gendel-man et al., 1985, 1986; Narayan et al., 1982; Clements etal., 1994). Ovine lentivirus–infected alveolar macrophagesare believed to play a central role in the genesis oflymphoid interstitial pneumonia (Gendelman et al., 1985;Narayan et al., 1982).

The pulmonary intravascular macrophage (PIM) is arecently identified mononuclear phagocyte in some domes-tic animal species, including sheep (Singh et al., 1995;Warner et al., 1990; Staub, 1994). The PIMs develop fromcirculating monocytes, form adhesions with the capillaryendothelium, and have profound phagocytic as well assecretory capabilities (Warner et al., 1990). The plasmamembrane of PIMs is decorated with a layer of globuleswith intervening electron-lucent spaces which is called thesurface coat (Atwal et al., 1992; Singh et al., 1994, 1995;Atwal et al., 1989). Surface coat globules of PIMs of sheepand other animal species are implicated in the endocytosisof tracer particles such as Monastral Blue and cationizedferritin, and Escherichia coli lipopolysaccharide (Singhand Atwal, 1997; Warner et al., 1990; Cirelli et al., 1995).The PIMs can secrete inflammatory mediators such asarachidonic acid metabolites and cytokines that contributeto sepsis-induced lung pathology (Warner et al., 1990;Cirelli et al., 1995). To date, there is only one reportimplicating PIMs in a viral infection to suggest them as invivo reservoirs of African swine fever virus (Sierra et al.,1990). To our knowledge, there is no information on the invivo behavior of PIMs in lentivirus-infected lambs.

Interferons are bioactive compounds that play a premierrole in the modulation of immune cell function and inflam-mation in response to various pathogens (Bazer andJohnson, 1991). Interferon-tau (IFN-t), first named as theovine trophoblast protein, is a recently identified noveltype I IFN (Bazer and Johnson, 1991). Recombinant ovineIFN-t (roIFN-t) exhibits robust in vitro antiviral activityagainst human immunodeficiency virus while retaininglow toxicity in comparison to IFN-t (Subramaniam et al.,1995; Dereuddre-Bosquet et al., 1996). We recently showedsuppression of in vitro OvLV virus replication in goatsynovial membrane cells and reduction of in vivo cell-associated OvLV titers in lambs (Juste et al., 1996).

Cytokines such as IFNs induce distinct morphology andcytotoxic behavior in T lymphocytes against virus-infectedcells or tumor cells in comparison to normal lymphocytes(Vollenweider and Groscurth, 1991). Recent results fromour laboratory showed increased proportions of peripheralblood CD81 (cytotoxic) and CD41 (helper) T cells inroIFN-t–treated and/or OvLV-infected lambs and suppres-sion of interstitial lung pathology (Singh et al., 1997). Thelysis or apoptosis of target cells, such as virus-infectedcells, by the cytotoxic cells is mediated through Fas and/orperforin pathways that require intimate plasma mem-brane contact (Zinkernagel, 1996; Kagi et al., 1994a,b).The process of target cell killing by the lymphocytes alsohas distinct structural features that facilitate its identifica-tion in tissues at the transmission electron microscopelevel (Vollenweider and Groscurth, 1991; Carpen et al.,1982, 1983).

The objective of this study was to obtain data on the invivo morphological behavior of PIMs and other vascularinflammatory cells in lambs infected with OvLV and/ortreated with roIFN-t. Results from this investigation showreduced number of PIMs and altered morphology of theirunique surface coat in OvLV-infected and/or roIFN-t–treated lambs. The OvLV infection induced features ofcellular activation in the PIMs. Many PIMs in roIFN-t–treated OvLV-infected lambs formed intimate membranecontacts with lymphocytes and appeared to be damaged.The significance of these findings in relation to antiviraland therapeutic effects of roIFN-t in OvLV-infected lambsis discussed.

MATERIALS AND METHODSAnimals

The experimental protocols used in this study wereapproved by Animal Care and Use Committee, Texas A&MUniversity. Twelve (N 5 12), male, 1-month-old, Rambouil-let lambs from an OvLV seronegative flock were assignedrandomly to four groups (n 5 3 each). The lambs were keptin separate cages in temperature-controlled rooms in theTexas Agricultural Experiment Station animal isolationfacility and allowed to acclimatize for 1 week prior toinitiation of the experiment. All the lambs tested negativefor the presence of OvLV antibodies by the agar gelimmunodiffusion test performed with a commercial kit(Veterinary Diagnostic Technology, Wheat Ridge, CO).

Virus and InterferonOvine lentivirus-85/34, a highly cytopathic, rapid/high

strain was grown in goat synovial membrane cells andtitrated as described previously (Juste et al., 1996). Recom-binant ovine IFN-t produced from a synthetic gene inPichia pastoris was produced and purified as describedelsewhere (Ott et al., 1991; VanHeeke et al., 1996). Twogroups of lambs were mock-infected with cell culturesupernatant, and one of these groups was treated daily

473OVLV, ROIFN-t, AND PIMS

with placebo while the other received 106 antiviral units ofroIFN-t/kilogram/day subcutaneously in the left axilla.The remaining two groups were infected intratracheallywith 106 TCID50 of OvLV. One of the OvLV-infected groupswas administered placebo, and the other received 106 anti-viral units of roIFN-t/kilogram/day subcutaneously in theleft axilla starting on the day of infection. Serum samplescollected from the lambs every week were analyzed forOvLV antibodies with the agar gel immnodiffusion test.

Tissue ProcessingAll the lambs were euthanized with an overdose of

pentobarbital sodium (Euthasolt, Delmarval Labs.Inc., Midlothian, VA) 34 weeks after initiating the experi-ment. An endotracheal tube was placed in the right lung,and this lung was fixed in situ with 2.5% glutaraldehydeand 2.0% paraformaldehyde in 0.1 M sodium cacodylatebuffer (pH 7.2) for 30 min. Right-side thoracotomies wereperformed to collect lung tissues from different lobes of theright lung. Lung tissues were diced into 1 mm3 pieces andimmersion-fixed for another 3 h in the same fixative. Afterthree washings in 0.1 M Na-cacodylate buffer, the tissueswere postfixed in 1.5% osmic acid for 90 min followed bydehydration in ascending concentrations of alcohol andpropylene oxide and embedded in Epoxy resins.

Toluidine blue–stained thick sections were examined toselect representative areas for ultrathin microtomy. Theultrathin sections placed on copper grids were stainedwith uranyl acetate and lead citrate and examined in atransmission electron microscope at 80 kV.

Numerical CountsThe number of PIMs was counted in toluidine blue–

stained sections from Epoxy-embedded tissues. Six blockswere selected from each animal in each group (18 tissueblocks/group) for thick-sectioning. The PIMs were countedin ten fields/section (180 fields/group) under oil immersion.The numbers thus generated were used for between thegroup multiple comparisons with Tukey’s test at P # 0.05.

RESULTSAll lambs remained healthy during the course of this

study. The animals infected with OvLV seroconvertedwithin 2–4 weeks of infection, while the mock-infectedlambs remained seronegative for OvLV antibodies untilthe end of this study.

Light MicroscopyLight microscopic examination of toluidine blue–stained

thick sections showed a reduction in the number of PIMsin all the groups as compared to the placebo-treated,mock-infected group (mean 31.7, SE 6 1.19) (Fig. 1). Thereduction in the number of PIMs was greater in roIFN-t–treated, mock-infected (mean 14.6; SE 6 0.86) or roIFN-t,OvLV-infected lambs (mean 12.6; SE 6 0.70) than inplacebo-treated, OvLV-infected lambs (mean 18.3; SE 60.84). No differences were detected in the number of PIMsin the roIFN-t–treated, mock-infected lambs as comparedto the roIFN-t–treated, OvLV-infected lambs.

Transmission Electron MicroscopyThe ultrastructure of lung tissues from the placebo-

treated, mock-infected lambs was characteristic of a nor-

mal lung as described previously (Warner and Brain, 1990;Singh et al., 1995). Pulmonary capillaries contained numer-ous PIMs and other vascular cells such as platelets and redblood cells (Fig. 2). The PIMs were distinguished fromother vascular and extravascular lung cells by the pres-ence of unique globules on the surface as well as in thecytoplasmic vesicles (Fig. 2).

Placebo-Treated OvLV-InfectedPulmonary intravascular macrophages from the placebo-

treated, OvLV-infected group were devoid of a surface coatbut contained intracellular globules in their endosomes(Fig. 3). Many of these PIMs exhibited prominent nuclearinterchromatin granules and close association of mitochon-dria with expanded profiles of rough endoplasmic reticu-lum (Figs. 4, 5). At high magnification, the PIMs revealedelaborate arrangements of hyperplastic Golgi complexes,secretory vesicles, centrioles, and tracts of microtubules(Fig. 6). Occasionally, structures resembling OvLV werenoticed in the cytoplasm as well as on the plasma mem-brane of the PIMs of placebo-treated, OvLV-infected lambs(Fig. 7). However, in the absence of immunocytochemistryfor OvLV, no firm conclusions could be drawn regarding theprecise nature of these structures. At low magnification,many lymphocytes with ruffled plasma membranes wereseen in contact with PIMs (Fig. 8).

Lymphocytes containing electron-dense granules andpolarized Golgi complexes and resembling cytotoxic Tlymphocytes were closely apposed to the PIMs (Fig. 9).

roIFN-t–Treated Mock-InfectedThe transmission electron microscope examination of

lung tissues from roIFN-t–treated, mock-infected lambsrevealed a striking reduction in the number of PIMs frommost of the capillaries, thus in agreement with the light

Fig. 1. Multiple comparisons of numerical counts of PIMs. Groups withdifferent letters differed from each other in the number of PIMs/section,while those with the same letters were same (P # 0.5). IFN-t1NV,roIFN-t 1 no virus; IFN-t1V, roIFN-t 1 virus; PL1NV, placebo 1 no virus;PL1V, placebo 1 virus.

474 SINGH ET AL.

microscopy data. Few PIMs were still attached to theendothelium and exhibited a lack of a surface coat, con-tained few intracellular globules, and showed signs ofdamage to the plasma membrane (Fig. 10). Some of thePIMs had conspicuously polarized Golgi complexes (Fig.11). Newly differentiated PIMs, as indicated by largecircular nuclei, the presence of a few surface coat globules,

and the absence of intracellular globules, were also noticed(data not shown).

roIFN-t–Treated OvLV-InfectedThe majority of the PIMs from roIFN-t–treated, OvLV-

infected group appeared to be damaged, as indicated by

Fig. 2. Placebo-treated mock-infected. A pulmonary intravascularmacrophage (PIM) attached to the capillary endothelium (single arrow)shows surface (double arrows) as well as intracellular globules in the

endosomes (E). Other vascular cells such as platelets (pt) and red bloodcells (asterisks) are devoid of similar structural features. AS, alveolarspace; C, collagen; arrowhead, elastin. 312,500.


Fig. 3. Placebo-treated OvLV-infected. A low magnification view of aPIM devoid of a surface coat except an isolated globule (double arrows).C, collagen; E, endosomal globules; En, endothelium; pt, platelet. 312,600.

Fig. 4. Placebo-treated OvLV-infected. A part of a PIM showingnuclear interchromatin granules (single arrows) and a surface coatglobule (double arrows). 362,800.

Fig. 5. Placebo-treated OvLV-infected. A high magnification view ofpart of a PIM showing profiles of rough endoplasmic reticulum (R)associated with mitochondria (Mc). Double arrows point to a cross-sectionof rough endoplasmic reticulum. E, an endosome containing globules.380,000.

Fig. 6. Placebo-treated OvLV-infected. A high magnification micrograph of a PIM revealing expanded Golgicomplexes (G) associated with secretory vesicles (arrows), centriole (Ct), and microtubules (Mt). E, endosomeswith globules; Mc, mitochondria; TGN, trans-Golgi network. 362,800.


Fig. 7. Placebo-treated OvLV-infected. This micrograph of a PIMshows a spherical structure with an electron-dense core (large arrow)resembling OvLV. A spherical bud (open arrow) on the plasma membrane

appears similar to OvLV as well. C, collagen; E, endosomes containingglobules; En, endothelium; Er, endoplasmic reticulum; double arrows,surface coat globules. 362,800.

478 SINGH ET AL.

their shrunken profiles and the paucity of cell organelles intheir cytoplasm (Fig. 12). The surface coat of most of thePIMs was disrupted. These PIMs contained large lateendosomes with partially digested cellular debris andresidual lysosomal lamellar bodies (Figs. 13, 14). Activatedlymphocytes and neutrophils formed intimate contacts viatheir plasma membrane projections with elongated andshrunken PIMs (Figs. 15, 16). Golgi complexes and mito-chondria in these lymphocytes were located in the areabetween their nuclei and the points of plasma membranecontact with the PIMs (Fig. 15).

DISCUSSIONThis study is the first to examine structural features of

PIMs and other vascular cells of lambs treated withroIFN-t and/or infected with OvLV. This is importantbecause of the recent elucidation of the impact of phago-cytic and secretory potential of PIMs on lung inflamma-tion. The number of PIMs was reduced in all the roIFN-t–treated groups as compared to placebo-treated, mock-infected lambs. The PIMs of OvLV-infected lambs containedoccasional OvLV-like particles, expanded profiles of roughendoplasmic reticulum, and Golgi complexes but weremostly devoid of a globular surface coat. In addition, manyfeatures bearing resemblance to the process of cytotoxiclymphocyte-mediated damage to the PIMs were also ob-served in OvLV-infected lambs treated with either placeboor roIFN-t.

Atwal and colleagues have previously identified a uniquestructural and functional feature on the plasma mem-brane of PIMs of ruminants and horses which consists of achain of electron-dense globules (Atwal and Minhas, 1992;Atwal et al., 1992; Singh et al., 1995). The PIMs from thelambs infected with OvLV and/or treated with roIFN-t, incontrast to those of placebo-treated, mock-infected lambs,were devoid of the surface coat. A series of studies demon-strated alterations in the integrity of surface coat globulesas the first response of PIMs to biological and chemicalchallenges (Atwal et al., 1992; Singh et al., 1994, 1995;Singh and Atwal, 1997). These studies established a stronglink between loss of the surface coat and the expression ofstructural and cytochemical features of activation in PIMs.Results from the present study add a virus, OvLV, and anovel cytokine, roIFN-t, to the list of agents that affect thesurface coat in conjunction with activation of PIMs. Ourpresent inability to isolate the coat globules, however, doesnot permit us to do in vitro experiments to determine thebiochemical significance of such changes in the surface coat.

Ultrastructural features of PIMs from OvLV-infectedlambs revealed signs of secretory activation in contrast tomock-infected animals. The hyperplastic Golgi complexesin conjunction with secretory vesicles, expanded roughendoplasmic reticulum, centrioles, and elaborate microtu-

Fig. 8. Placebo-treated OvLV-infected. This micrograph depicts threelymphocytes (L) with plasma membrane ruffles and one of them in contactwith a small PIM. AS, alveolar space; *, red blood cells. 36,600.

Fig. 9. Placebo-treated OvLV-infected. This micrograph shows closeassociation between an activated lymphocyte (L) in contact with a PIMthat is devoid of surface coat globules. Note the polarized arrangement ofcell organelles such as the Golgi complex (G) in the lymphocyte in relationto its nucleus. AS, alveolar space; C, collagen; arrow, cytoplasmicgranules. 312,600.


Fig. 10. roIFN-t–treated mock-infected. A view of a PIM showing largenucleus (N), rarefied cytoplasm without cellular organelles, and only a fewisolated coat globules (double arrows). AS, alveolar space; pt, platelet; *,red blood cells. 312,600.

Fig. 11. roIFN-t–treated mock-infected. The PIM in this micrographshows the polarized arrangement of Golgi complexes (G) and mitochon-dria and occasional surface coat globules (double arrows). AS, alveolarspace; *, red blood cells. 316,000.

480 SINGH ET AL.

Figs. 12–14. (Legend, overleaf).


bules indicate enhanced biosynthetic activity in these cells(Rothman and Orci, 1992). We noticed spherical structuresresembling OvLV in the cytoplasm as well as on theplasma membrane of a few PIMs. These structures appearsimilar to the ones recently reported in in vitro transmis-sion electron microscope studies on maedi visna virus-infected macrophages (Lee et al., 1996). However, a preciseidentification of these particles awaits ultrastructuralimmunocytochemistry using antibodies against one of theviral proteins. Because the proportion of OvLV-infectedlung cells is low, the possibility of virus detection at thetransmission electron microscope level remains minimaleven after immunogold labeling (DeMartini et al., 1993;Peluso et al., 1985). Furthermore, the limited amount oftissue that can be examined with a transmission electronmicroscope also diminishes the probability of directlyobserving OvLV particles.

Ovine lentivirus localizes to various organs in the hostanimal via infected monocytes (Clements et al., 1994;Gendelman et al., 1985, 1986). Alveolar macrophages,following their development from infected monocytes, actas the principal infected cells in the lungs and assume acrucial role in the pathogenesis of lymphoid interstitialpneumonia (Gendelman et al., 1985; Brodie et al., 1995;Clements and Zink, 1996). Since the PIM is a recentlyidentified mononuclear phagocyte, there is no informationon the possible contributions of this cell in the OvLV-induced disease process. The PIMs are believed to developfrom circulating vascular monocytes (Warner and Brain,1990; Staub, 1994). Logically, as well as based on observa-tions from the present study, maturation of some of theOvLV-infected monocytes into PIMs would turn these cellsinto a vascular reservoir of infection. Moreover, PIMs havealready been shown to serve as target cells for replicationof African swine fever virus (Sierra et al., 1990). Wepropose that some of the PIMs may be a site of OvLVreplication in infected lambs. The PIMs have previouslybeen demonstrated to secrete proinflammatory substancessuch as arachidonic acid metabolites, TNF-a and IL-1(Cirelli et al., 1995; Chitko-McKown et al., 1991). There-fore, the OvLV-infected and activated PIMs, similarly toalveolar macrophages, may secrete inflammatory sub-stances to generate a local cytokine-rich environment andmodulate adhesion molecules on the pulmonary capillaryendothelium to aid extravasation of lymphoid cells into thepulmonary interstitium.

Pulmonary intravascular macrophages have been re-ported only in some domestic animal species (Singh et al.,

1995; Warner et al., 1990). Lungs of normal humans orrats do not contain a significant population of such mono-nuclear phagocytes (Dehring and Wismar, 1989). However,PIMs have been shown to develop in the lungs of ratsunder certain conditions such as bile duct ligation andbacteremia (Chang and Ohara, 1994; Singh et al., inpress). Because the OvLV-induced pulmonary pathology issimilar to the one caused by the human immunodeficiencyvirus, it would be interesting to investigate if the lungs ofacquired immunodeficiency syndrome patients containhuman immunodeficiency virus–infected intravascularmacrophages.

The roIFN-t treatment as well as OvLV infection in-duced a striking reduction in the number of PIMs. There isno information on such an effect of an IFN on other fixedvascular macrophages such as Kupffer cells. We can onlyspeculate on the causes of this behavior of PIMs inresponse to roIFN-t administration. The roIFN-t mayremove the PIMs by inducing shape changes as a conse-quence of their activation that might interfere with theirattachment to capillary endothelium. Second, roIFN-ttreatment of OvLV-infected or mock-infected lambs maycause direct damage to PIMs via apoptosis or indirectdamage via lymphocyte-mediated cytotoxicity. It is pos-sible that the antiproliferative effects of roIFN-t alsocontribute to a decrease in the number of PIMs by prevent-ing their replacement by newly differentiated monocytes(Bazer and Johnson, 1991). We did not examine theKupffer cells to see if these cells were also affected byroIFN-t treatment or OvLV infection. Removal of a pre-mier cell, such as a macrophage, by roIFN-t would haveprofound implications for the inflammatory processes dueto their phagocytic and immunomodulatory functions.This contention is supported by the suppression of OvLV-induced interstitial lung pathology in conjunction withgadolinium chloride–mediated removal of PIMs (Singhand de la Concha-Bermejillo, in press). Similar effectshave previously been documented in endotoxemic ratsby abrogating Kupffer cell function with gadolinium chlo-ride (Pendino et al., 1995; Liu et al., 1994; Roland et al.,1996).

The PIMs in roIFN-t–treated, OvLV-infected lambs wereof smaller size and showed a lack of cell organelles incytoplasm and signs of plasma membrane damage. ThesePIMs exhibited signs of enhanced phagocytic activity, asindicated by the presence of large endosomes containingcellular debris and lysosomal lamellar bodies. This couldbe an outcome of direct or indirect activation of PIMs byroIFN-t. Many lymphocytes and neutrophils appeared tobe activated, as suggested by the ruffled plasma mem-branes and polarized location of cell organelles in placebo-or roIFN-t–treated, OvLV-infected lambs. Plasma mem-brane processes of these lymphocytes and neutrophilsformed intimate and extensive contacts with the PIMs.These lymphocytes contained nuclei with characteristicdeep invaginations, and their Golgi complexes were local-ized in the vicinity of the area of contact with the targetPIMs. These features are similar to the previously de-scribed contacts of small membrane projections of effectorlymphocytes with surface modifications in the target cells(Vollenweider and Groscurth, 1991; Carpen et al., 1982,1983). The relocation and polarization of the Golgi complexis an early and crucial event during activation, post-injuryregeneration, and movement of cells to direct the flow ofcellular secretions to a particular area to facilitate plasma

Figs. 12–14: Fig. 12. roIFN-t–treated OvLV-infected. This low mag-nification view of a capillary shows a shrunken, possibly damaged,pulmonary intravascular macrophage (P) without any cell organellesexcept globules. Other pulmonary cells such as red blood cells (*) andendothelium (En) appear to be normal. AS, alveolar space; C, collagen.35,000.

Fig. 13. roIFN-t–treated OvLV-infected.An electron micrograph depict-ing a PIM with a large late endosome (LE) containing partially digestedcellular debris and impinging on the nucleus. AS, alveolar space; C,collagen; double arrows, a fuzzy coat on plasma membrane. 312,600.

Fig. 14. roIFN-t–treated OvLV-infected. A PIM showing a lysosomallamellar body (LB) and polarized arrangement of mitochondria (Mc). Thesurface coat is composed of small globules (double arrows) as comparedto those of the control PIMs. AS, alveolar space; C, collagen; E,endosomes containing globules; En, endothelium; *, red blood cells.312,000.

482 SINGH ET AL.

membrane modifications (Kelly, 1990; Singer and Kupfer,1986). Such plasma membrane modifications will be re-quired for closer contact between lymphocytes and thePIMs to execute efficient killing. It is possible that these

PIMs were infected with the virus and expressed viralantigen in association with MHC class I complex, thusoffering a target to roIFN-t–activated T lymphocytes(Zinkernagel, 1996; Kagi et al., 1994a).

Fig. 15. roIFN-t–treated OvLV-infected. A lymphocyte (L) with direc-tional arrangement of the Golgi complex (G) and mitochondria (Mc) is inclose contact with an elongated and shrunken PIM. AS, alveolar space;R, rough endoplasmic reticulum; double arrows, surface coat globules.312,500.

Fig. 16. roIFN-t–treated OvLV-infected. A neutrophil (Nt) showingGolgi elements (G) forms a broad contact with a possibly damaged PIMthat contains an endosome (E) with cellular debris. AS, alveolar space; pt:platelet; double arrows, surface coat; *, red blood cells. 312,500.


roIFN-t, in addition to having direct antiviral effects,also increases the proportions of cytotoxic T cells, reducescell-associated OvLV viremia, and prevents developmentof lymphoid interstitial pneumonia (Subramaniam et al.,1995; Juste et al., 1996). The reduction in the number ofPIMs in roIFN-t–treated, OvLV-infected lambs was greaterthan in placebo-treated, OvLV-infected lambs, possibly asa consequence of roIFN-t–mediated increased cell killingby cytotoxic T lymphocytes. This reduction of fixed vascu-lar macrophages, PIMs, by roIFN-t represents anotherhitherto unknown function of this novel cytokine. Thedirect or indirect loss of PIMs in roIFN-t–treated, OvLV-infected lambs could result in the reduction of virus loadand the source of inflammatory mediators in the lungduring infection with OvLV and thereby add to an array ofcellular and therapeutic actions of this cytokine.

ACKNOWLEDGMENTSWe thank Mr. Miles Frey, Image Analysis Laboratory,

Texas A&M University, College Station, for excellent tech-nical support. We thank Mr. Octavio Ramos, Mr. PascualHernandez, and Mr. Rene Flores for animal care andhandling. This work was supported by funds from theUSDA under the Animal Health Act of 1977, Public Law95–113, and the Texas Agricultural Experiment Station.

LITERATURE CITEDAtwal OS, Minhas KJ. In vivo interaction of cationised ferritin with

the surface coat and endocytosis by pulmonary intravascular macro-phages: A tracer kinetic study. J. Anat. 1992;181:313–325.

Atwal OS, Minhas KJ, Ferenczy BG, Jassal DS, Milton D, Mahad-evappa VG. Morphology of pulmonary intravascular macrophagesin ruminants: Ultrastructural and cytochemical behaviour of densesurface coat. Am. J. Anat. 1989;186:285–299.

Atwal OS, Singh B, Staempfli H, Minhas KJ. Presence of pulmonaryintravascular macrophages in the equine lung: Some structuro-functional studies. Anat. Rec. 1992;234:530–540.

Bazer FW, Johnson HM. Type I conceptus interferons: Maternalrecognition of pregnancy signals and potential therapeutic agents.Am. J. Reprod. Immunol. 1991;26:19–22.

Brodie SJ, Pearson LD, Zink MC, Bickle HM, Anderson BC, MarcomKA, DeMartini JC. Ovine lentivirus expression and disease: Virusreplication, but not entry, is restricted to macrophages of specifictissues. Am. J. Pathol. 1995;146:1–13.

Carpen O, Virtanen I, Saksela E. Ultrastructure of human naturalkiller cells: Nature of the cytolytic contacts in relation to cellularsecretion. J. Immunol. 1982;128:2691–2697.

Carpen O, Virtanen I, Lehto VP, Saksela E. Polarization of NK cellcytoskeleton upon conjugation with sensitive target cells. J. Immu-nol. 1983;131:2695–2698.

Chang S, Ohara N. Chronic biliary obstruction induces pulmonaryintravascular phagocytosis and endotoxin sensitivity in rats. J. Clin.Invest. 1994;94:2009–2019.

Chitko-McKown CG, Chapes SK, Brown RE, Phillips RE, McKowanRD, Blecha F. Porcine alveolar and pulmonary intravascular macro-phages: Comparison of immune function. J. Leukoc. Biol. 1991;50:364–372.

Cirelli RA, Carey LA, Fisher JK, Rosolia DL, Elassser TH, CapernaTJ, Gee MH, Albertine KH. Endotoxin infusion in anesthetisedsheep is associated with intrapulmonary sequestration of leuko-cytes that immunohistochemically express tumor necrosis factor-alpha. J. Leukoc. Biol. 1995;57:820–826.

Clements JE, Zink MC. Molecular biology and pathogenesis of animallentivirus infections. Clin. Microbiol. Rev. 1996;9:100–117.

Clements JE, Zink MC, Narayan O, Gabuzda DH. Lentivirus infectionof macrophages. Immunol. Ser. 1994;60:589–600.

Dehring DJ, Wismar BL. Intravascular macrophages in pulmonarycapillaries of humans. Am. Rev. Respir. Dis. 1989;139:1027–1029.

de la Concha-BermejilloA. Maedi-visna and ovine progressive pneumo-nia. Vet. Clin. North Am. Food Anim. Pract. 1997;13:13–33.

de la Concha-Bermejillo A, Juste RA, Kretschmer R, Aguilar Setien A.Ovine lentivirus infection: An animal model for pediatric HIVinfection. Arch. Med. Res. 1995;26:345–354.

DeMartini JC, Brodie SJ, de la Concha-Bermejillo A, Ellis JA,Lairmore MD. Pathogenesis of lymphoid interstitial pneumonia innatural and experimental ovine lentivirus infection. Clin. Infect.Dis. 1993;17(Suppl. 1):S236–242.

Dereuddre-Bosquet N, Clayette P, Martin M, Mabondzo A, Fretier P,Gras G, Martal J, Dormont D. Anti-HIV potential of a new inter-feron, interferon-t (trophoblastin). J. Acquir. Immune Defic. Syndr.1996;11:241–246.

Gendelman HE, Narayan O, Molineaux S, Clements JE, Ghotbi Z.Slow, persistent replication of lentiviruses: Role of tissue macro-phages and macrophage precursors in bone marrow. Proc. Natl.Acad. Sci. U. S. A. 1985;82:7086–7090.

Gendelman HE, Narayan O, Kennedy-Stoskopf S, Kennedy PE,Ghotbi Z, Clements JE, Stanley J, Pezeshkpour G. Tropism of sheeplentiviruses for monocytes: Susceptibility to infection and virus geneexpression increase during maturation of monocytes to macro-phages. J. Virol. 1986;58:67–74.

Johnstone RB Jr. Monocytes and macrophages. N. Engl. J. Med.1988;318:747–752.

Juste RA, Ott TL, Kwang J, Bazer FW, de la Concha-Bermejillo A.Effects of recombinant interferon-t on ovine lentivirus replication. J.Interferon Cytok. Res. 1996;16:989–994.

Kagi D, Ledermann B, Burki K, Seiler P, Odermatt B, Olsen KJ,Podack ER, Zinkernagel RM, Hengartner H. Cytotoxicity mediatedby T cells and natural killer cells is greatly impaired in perforin-deficient mice. Nature 1994a;369:31–37.

Kagi D, Vignaux F, Ledermann B, Burki K, Depraetere V, Nagata S,Hengartner H, Golstein P. Fas and perforin pathways as majormechanisms of T cell-mediated cytotoxicity. Science 1994b;265:528–530.

Kelly RB. Microtubules, membrane traffic and cell organization. Cell1990;61:5–7.

Lee WC, McConnell I, Blacklaws BA. Electron microscope studies ofthe replication of a British isolate of maedi visna virus in macro-phages and skin cell lines. Vet. Microbiol. 1996;49:93–104.

Liu M, Xu J, Tanswell AK, Post M. Inhibition of mechanical strain-induced fetal rat lung cell proliferation by gadolinium, a stretch-activated channel blocker. J. Cell. Physiol. 1994;161:501–507.

Narayan O, Wolinsky JS, Clements JE, Strandberg JD, Griffin DE,Cork LC. Slow virus replication: The role of macrophages in thepersistence and expression of visna viruses of sheep and goats. J.Gen. Virol. 1982;59:345–356.

Ott TL, van Heeke G, Johnson HM, Bazer FW. Cloning and expressionin Saccharomyces cerevisiae of a synthetic gene for the type Itrophoblast interferon ovine trophoblast protein-1: Purification andantiviral activity. J. Interferon Res. 1991;11:357–364.

Peluso R, Haase AT, Stowring L, Edwards M, Ventura P. A Trojan horsemechanisms for the spread of visna virus in monocytes. Virology1985;147:231–236.

Pendino KJ, Meidhof TM, Heck DE, Laskin JD, Laskin DL. Inhibitionof macrophages with gadolinium chloride abrogates ozone-inducedpulmonary injury and inflammatory mediator production. Am. J.Respir. Cell Mol. Biol. 1995;13:125–132.

Roland CR, Naziruddin B, Mohanakumar T, Flye MW. Gadoliniumchloride inhibits Kupffer cell nitric oxide synthase. J. Leukoc. Biol.1996;60:487–492.

Rothman JE, Orci L. Molecular dissection of the secretory pathway.Nature 1992;355:409–415.

Sierra MA, Carrasco L, Gomez-Villamandos JC, Martin de las MulasJ, Mendez A, Jover A. Pulmonary intravascular macrophages inlungs of pigs inoculated with African swine fever virus of differingvirulence. J. Comp. Pathol. 1990;102:323–334.

Singer SJ, Kupfer A. The directed migration of eukaryotic cells. Annu.Rev. Cell Biol. 1986;2:337–365.

Singh B, Atwal OS. Ultrastructural and immunocytochemical study ofthe pulmonary intravascular macrophages of Escherichia coli lipo-polysaccharide-treated sheep. Anat. Rec. 1997;247:214–224.

484 SINGH ET AL.

Singh B, de la Concha-Bermejillo A. Gadolinium chloride removespulmonary intravascular macrophage and curtails the degree ofovine lentivirus-induced lymphoid interstitial pneumonia. Int. J.Exp. Pathol. (in press).

Singh B, Minhas KJ, Atwal OS. Surface coat of pulmonary intravascu-lar macrophages of sheep is reconstituted following brefeldin A–me-diated endocytosis. J. Submicrosc. Cytol. Pathol. 1994;27:235–249.

Singh B, Atwal OS, Ireland WP, Minhas KJ. Surface coat of sheeppulmonary intravascular macrophages: Reconstitution, and implica-tion of a glycosyl-phosphatidylinositol anchor. Anat. Rec. 1995;243:466–478.

Singh B, Ott TL, Bazer FW, de la Concha-Bermejillo A. Recombinantovine interferon-tau modulates lymphocyte subsets and ovine lenti-virus-induced lung pathology. FASEB J. 1997;11:A104 (abstract).

Singh B, Doanne K, Niehaus GD. Ultrastructural and cytochemicalevaluation of sepsis-induced changes in the rat pulmonary intravas-cular mononuclear phagocytes. J. Anat. 1998;192:13–23.

Staub NC. Pulmonary intravascular macrophages. Annu. Rev. Physiol.1994;56:47–67.

Subramaniam PS, Khan SA, Pontzer CH, Johnson HM. Differentialrecognition of the type I interferon receptor by interferons tau andalpha is responsible for their disparate cytotoxicities. Proc. Natl.Acad. Sci. U. S. A. 1995;92:12270–12274.

VanHeeke G, Ott TL, Strauss A, Ammaturo D, Bazer FW. High yieldexpression and secretion of the pregnancy recognition hormoneovine interferon by Pichia pastoris. J. Interferon Res. 1996;16:119–126.

Vollenweider I, Groscurth P. Ultrastructure of cell mediated cytotoxic-ity. Electron Microsc. Rev. 1991;4:249–267.

Warner AE, Brain JD. The cell biology and pathogenic role ofpulmonary intravascular macrophages.Am. J. Physiol. 1990;258:L1–L12.

Zinkernagel RM. Immunology taught by viruses. Science 1996;271:173–178.


Documents

Structural responses of pulmonary intravascular macrophages in lentivirus-infected and/or recombinant ovine interferon-τ–treated lambs