Parasite Immunology, 2011, 33, 34–44 DOI: 10.1111/j.1365-3024.2010.01249.x

Dynamics of bovine cell populations during the acute response to Babesia bovis infection: an immunohistological study

D. A. SCHNEIDER,1 H. YAN,2 R. G. BASTOS,2 W. C. JOHNSON,1 P. R. GAVIN,3 A. J. ALLEN,3 G. M. BARRINGTON,3 L. M. HERRMANN-HOESING,1 D. P. KNOWLES1 & W. L. GOFF1*

1Animal Disease Research Unit, USDA-ARS, Pullman, WA, USA, 2Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA, USA, 3Department of Veterinary Clinical Medicine and Surgery, Washington State University, Pullman, WA, USA

SUMMARY INTRODUCTION The spleen is a critical organ in defence against haemopara- Babesiosis is a tick-borne disease affecting cattle in much sitic diseases like babesiosis. Many in vitro and ex vivo stud- of the world, with Babesia divergens, B. bigemina and ies have identified splenic cells working in concert to activate B. bovis the economically important species. Babesia bovis mechanisms required for successful resolution of infection. is the most virulent, often causing death in susceptible ani- The techniques used in those studies, however, remove cells mals because of the development of anaemia, cerebral vas- from the anatomical context in which cell interaction and cular congestion and pulmonary and renal failure (1). The trafficking take place. In this study, an immunohistological virulent nature of the disease is attributed in part to the approach was used to monitor the splenic distribution of sequestration of parasitized erythrocytes to capillary endo- defined cells during the acute response of nave calves to thelium, but overproduction of inflammatory cytokines Babesia bovis infection. Splenomegaly was characterized by has also been suggested (2–4). Young calves are relatively disproportionate hyperplasia of large versus small leucocytes resistant to clinical infection (5) and demonstrate a strong and altered distribution of several cell types thought to be innate immunity composed of a type-1 inflammatory important in mounting an effective immune response. In par- response (6). In comparison with adult cattle, we have pre- ticular, the results suggest that the initial crosstalk between viously demonstrated that the immune response of calves NK cells and immature dendritic cells occurs within the mar- involves early IL-12 expression with consequent IFN-c ginal zone and that immature dendritic cells are first redi- production, a nitric oxide burst and modulation by IL-10 rected to encounter pathogens as they enter the spleen and (6–9). This age-related immunity is dependent upon cellu- then mature as they process and migrate to T-cell- lar events within the spleen as splenectomy of calves ren- rich areas. The results of this study are remarkably similar ders them equally susceptible (5,10). to those observed in a mouse model of malarial infection, Our studies have utilized a technique to marsupialize suggesting these dynamic events may be central to the acute the spleen of calves (11) so that cells could be acquired for response of nave animals to haemoparasitic infection. ex vivo analysis (microplate assays and flow cytometry) (12–16). Such analyses have proven valuable in determin- Keywords Babesia bovis, dendritic cells, immunohistochemis- ing the function of various splenic cell phenotypes but try, , macrophages, NK cells, spleen, cd T cells lack the ability to place these cell populations within their anatomical context which include the marginal zone, red Correspondence: David A. Schneider, Animal Disease Research and (17). Amongst many factors that comprise Unit, USDA-ARS, 3003 ADBF, Washington State University, an effective immune response to haemoparasitic infection, Pullman, Washington 99164-6630 USA trafficking and interaction of cells within such domains (e-mail: [email protected]). Disclosures: None. are central (18). Received: 19 March 2010 Intravital imaging techniques have been used to dynami- Accepted for publication: 06 July 2010 cally study such factors within superficial lymphoid organs *Retired Research Immunologist, Animal Disease Research Unit (19,20) and, to a limited extent, also within deeper struc- USDA-ARS, 4181 W. Upriver Drive, Coeur d’Alene, ID 83814, tures including the spleen of mice (21). But current tech- USA. niques are not well suited to study the spleen of large

34 2010 Blackwell Publishing Ltd Volume 33, Number 1, January 2011 Spleen cell dynamics during babesial infection mammals because of the limits on depth resolution (22). sequence. This sequence resulted in a hyperintense spleen An approach readily applied to the spleen of large mam- on a low intense background. The volume was calculated mals is the serial analysis of the distribution of pheno- by tracing the outline of the spleen for the area on each typed cells in tissue sections. Similar to a recent study on slice and multiplying by the number of slices plus gap the acute immune response of nave mice to haemoparasit- thickness (3D-DOCTOR; Able Software Corporation, ic infection (23), we have applied this technique to the Lexington, MA, USA). Each calf’s spleen volume was cal- spleen of nave calves infected with Babesia bovis. The culated on the day prior to infection and then at 11 or results document acute change in the distribution of sev- 12 dpi, 2 calves each. eral cells thought to be important to the spleen-dependent Immediately following each MRI procedure, a 1 cm3 response of nave calves to B. bovis and serve to under- biopsy of marsupialized spleen was removed under local score common themes in the acute response to haemopar- lidocaine anaesthesia for determining differential cell asitic infections. In addition, this is the first documented counts. Each biopsy was immediately processed into a sin- use of magnetic resonance imagery to measure spleen vol- gle cell suspension using a tissue grinder (Tenbroek; Bellco ume in calves. Glass, Inc., NJ, USA), suspended in 50 mL of PBS and enumerated for differential cell counts by standard meth- ods used for whole blood (28). MATERIALS AND METHODS

Animals and experimental B. bovis infection Immunohistochemistry (IHC) Twelve Holstein–Friesian steer calves were obtained at Six inoculated calves were euthanized by captive bolt and 8 weeks of age, vaccinated against pathogenic Clostridium jugular exsanguination for collection of spleen tissue: one species, castrated and dehorned. All animals were cELISA calf each on dpi 7, 8, 9 (fever day 1) and 14 (fever day 5), seronegative for Anaplasma marginale (VMRD, Pullman and two calves at 13 dpi (fever days 4 and 5). In this way, WA, USA) and B. bovis and B. bigemina (24–26). The care the from three calves each were examined from and use of these calves were approved by the Institutional two periods: a period just prior to, or including, the initia- Animal Care and Use Committee at Washington State tion of fever (7, 8 and 9 dpi) and a period several days University (Pullman, WA, USA). At 12 weeks of age, all after fever initiation (13 and 14 dpi). Spleen tissue from calves underwent a surgical procedure to marsupialize the two uninfected calves was similarly collected. Multiple spleen (11). When necessary, spleen cell aspirates were 15 · 15 · 5 mm sections of spleen were collected from obtained under local lidocaine anaesthesia into 60cc syrin- each calf immediately posteuthanasia. Each section was ges containing ACD and prepared for in vitro studies as placed into a cryostat mould containing Tissue-Tek previously described (14,27). O.C.T. Compound (Sakura Fineteck USA, Inc., Tor- Ten of the twelve calves were inoculated intravenously rance, CA, USA), snap frozen by floating on liquid nitro- 5 with 1 · 10 erythrocytes infected with the T2Bo virulent gen, and stored at )80C. Cryostat sections (15 lm) were isolate of B. bovis (7). The acute response to infection was mounted on standard SuperFrost Plus slides (Electron studied 7–14 days postinfection (dpi). As described below, Microscopy Services, Hatfield, PA, USA), fixed in 95% repeated measures of spleen volume and cell content were EtOH for 10 min and allowed to air dry overnight at made in four inoculated calves whereas change in regional room temperature. Formalin-fixed, paraffin-embedded distribution of phenotyped cells was determined by samples of spleen were also collected from each calf and sequential euthanasia of six inoculated calves in compari- routinely stained in haematoxylin and eosin (H&E). son with two un-inoculated calves. Immunolabelling was carried out at room temperature in a humidified chamber. A Super PAP Pen HT (Research Products International Corp., Mt. Prospect, IL, Determination of spleen volumes and differential cell USA) was used to create a hydrophobic margin to retain counts fluid reagents on slides. Thin sections on slides were pre- Magnetic resonance imagery was performed with a 1Æ0 blocked for 1 h using 10% goat serum ⁄ 0Æ1% Triton-X Tesla machine (Philips Intera, Andover, MA, USA). 100 ⁄ PBS. After three 5-min washes in PBS, thin sections Sequences were acquired in a dorsal plane. The area were exposed (2–4 h) to primary antibody (Table 1) imaged was from the spine to the ventral abdominal wall. diluted in 10% goat serum ⁄ PBS. Unbound primary anti- A 40 cm field-of-view ensured that the entire spleen could body was removed with three 5-min washes in PBS and be visualized. One-centimetre-thick slices with a 2 mm gap then exposed (2 h) to fluorophore-conjugated secondary were acquired using a short tau inversion recovery (STIR) antibody, all diluted 1 : 200 in 10% goat serum ⁄ 0Æ1%

2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 35 D. A. Schneider et al. Parasite Immunology

Table 1 Primary antibodies

Working concentration Target Name Species isotype (lg ⁄ mL) Source

CD3 MM1a Mouse IgG1 2Æ5 VMRD, Inc., Pullman, WA, USA CD4 IL-A11 Mouse IgG2a 2Æ5 VMRD, Inc., Pullman, WA, USA CD335 (NKp46) MCA2365 (clone AKS1) Mouse IgG1 5 AbD Serotec, Oxford, UK CD13 (aminopeptidase N) MCA2338 (clone CC81) Mouse IgG1 2Æ5 AbD Serotec, Oxford, UK CD172a (SIRPa) DH59b Mouse IgG1 5 VMRD, Inc., Pullman, WA, USA MCA2041G (clone CC149) Mouse IgG2b 2Æ5 AbD Serotec, Oxford, UK TcR1-N24 (d chain) GB21a Mouse IgG2b 5 VMRD, Inc., Pullman, WA, USA WC1 IL-A29 Mouse IgG1 2Æ5 VMRD, Inc., Pullman, WA, USA MCA838G (clone CC15) Mouse IgG2a 5 AbD Serotec, Oxford, UK Babesia bovis MSA-1 BABB35 Mouse IgG2a 5 Goff WL, et al. (29)

Triton-X 100 ⁄ PBS. After three 5-min washes in PBS, the the term ‘progressive’ is used to indicate appreciation of slides were coverslipped using ProLong Gold antifade an ordered change over time. Measurements of the splenic mounting media with DAPI (Molecular Probes, Inc., marginal zone included the region extending from its folli- Eugene, OR, USA). DAPI staining aided in follicle locali- cle junction (indicated in figures by a dashed curved line) zation, especially in the presence of a greatly expanded red to a width of 100 lm, and measurements of the pup postinfection. included regions furthest away from neighbouring white Immunohistochemistry (IHC) controls for these experi- pulp. IHC measurements must be considered approximate ments included substitution of primary or secondary anti- as uncontrolled changes in tissue dimensions are expected bodies with antibody diluent, and substitution of primary to have occurred during euthanasia and preparation of antibodies with isotype-matched irrelevant antibodies. thin frozen sections. All data were tabulated in Microsoft Dual-labelling experiments were performed by co-incuba- Office Excel 2003 and are reported as mean € standard tion of primary antibodies followed by co-incubation of error. Splenic volume (MRI) and differential cell count selective secondary antibodies. Nonspecific staining and data were analysed for significant (P <0Æ05) postinfection cross-reactions between secondary antibodies or between a increases by paired T-test (SAS for Windows 9.2; SAS primary antibody and nonrelevant secondary antibody Institute Inc., Cary, NC, USA). were not observed. Note: Attempts were made to localize CD8+ cells by IHC (primary antibody = BAQ111a, iso- RESULTS type = IgM; VMRD, Inc., Pullman, WA, USA). CD8 localization was precluded, however, by significant back- As previously described (7), intravenous inoculation of 105 ground mediated by anti-IgM secondary antibody. T2Bo isolate of B. bovis into 6 month-old nave Holstein calves consistently induced fever (>39Æ5C) between 8 and 10 dpi. The rare presence of B. bovis-infected erythrocytes Image and data analysis was noted in each animal by examination of Giemsa Immunohistochemistry (IHC) slides were viewed and pho- stained blood films just prior to euthanasia. Although tographed using an Axio Imager M1 microscope (Carl calves were necropsied at different intervals, each was Zeiss Microimaging, Thornwood, NY, USA) equipped experiencing a decrease in haematocrit from their normal with an LED illuminator for bright field microscopy and pre-infection levels. At 7 dpi the haematocrit was an X-Cite 120 Fl Illuminating system (EXFO Photonic decreased 19% and by 13–14 dpi had decreased 45 € 6Æ7% Solutions, Mississauga, ON, Canada) for epi-fluorescence (n = 3). microscopy. Digital images were captured using an Axio- The spleen of nave calves doubled in volume by Cam MRc5 digital camera connected to a desktop com- 11–12 dpi and was associated with significant increases in puter running AxioVision (version 4.7.1.0) and prepared the total splenic content of small leucocytes (approxi- for presentation using Photoshop Elements (version 4.0; mately twofold), large leucocytes (approximately eightfold) Adobe Systems Inc., San Jose, CA, USA). Figure images and total leucocytes (approximately twofold) (Table 2). As are representative, and variation within or between time determined by FACS analysis (data not shown), the large points (dpi) is noted in the Results section. In particular, leucocyte population included monocytes, macrophages,

36 2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 Volume 33, Number 1, January 2011 Spleen cell dynamics during babesial infection dendritic cells (DCs) (12) and large granular natural killer were larger angular cells. In the uninfected calf, WC1+ (NK) cells (15). As viewed in H&E sections, splenomegaly cells densely populated the marginal zone (900– 7–14 dpi was associated with a progressive basophilic 2500 cells ⁄ mm2, see ‘{’ in Figure 2j) but were relatively hyperplasia within the red pulp and histological reduction scarce in the red pulp (100–150 cells ⁄ mm2) whereas ) in the white pulp (w) and trabeculae (t) elements brightly fluorescent TcR1+ ⁄ WC1 cells were predominately (Figure 1, 1Æ25·), and also a loss in zonal distinction observed within the red pulp, often appearing clustered between marginal zone and red pulp (Figure 1, 10·). The (see arrow, Figure 2g). During acute infection, the density ) regional distributions of phenotyped cells were further of TcR1+ ⁄ WC1 cells progressively increased in the red investigated by IHC. pulp and marginal zone (Figure 2g–i) whereas the density of marginal zone WC1+ cells progressively dropped to a similar density as observed in the red pulp (Figure 2j–l). Small leucocyte populations of the spleen: CD3+, CD4+ cells and cd T cells Large leucocyte populations of the spleen: CD335+ (NK Examples of the splenic cellular immunoreactivity to cells), CD172+ (monocytes ⁄ macrophages) and CD13+ monoclonal antibodies specific for CD3 and CD4 are (immature DC) shown in Figure 2a–f. Two cell populations were clearly evident in this dual-labelling experiment: CD3+ ⁄ CD4+ and Fluorescent immunoreactivity mediated by a CD335-spe- ) CD3+ ⁄ CD4 cells. In the uninfected calf, CD3+ ⁄ CD4+ cific antibody, a specific marker for natural killer (NK) cells were always most dense within the periarteriolar lym- cells, is shown in Figure 3a–c. In the uninfected calf, phatic sheath (PALS; see ‘[’ in Figure 2a,d). A band of CD335+ cells were typically present as a dense marginal ) CD3+ ⁄ CD4+ and CD3+ ⁄ CD4 cells was consistently pres- zone band extending approximately 250 lm from the folli- ent within the marginal zones of uninfected spleens, cle–marginal zone junction (600–1400 cells ⁄ mm2, see ‘{’ in extending 185 € 29 lm away from the follicle [see ‘{’ in Figure 3a) and were less dense in the red pulp (140– Figure 2a,d]. Both populations were relatively scarce 480 cells ⁄ mm2). By 7 dpi and continuing through 14 dpi, within the red pulp. During the acute response to infec- the density of CD335+ cells within the marginal zone was tion, the distinctive presence of this marginal zone band reduced to approximate that found in the red pulp (Fig- was obscured by a progressive red pulp increase in ure 3b,c). ) CD3+ ⁄ CD4 cells and a more modest increase in MCA2338 is a monoclonal antibody directed towards CD3+ ⁄ CD4+ cells (Figure 2b,c,e,f). CD13, a marker for immature splenic dendritic cells The localization of cd T cells in the spleen is shown in (iDCs) (12). In all calves the vast majority of CD13+ cells Figure 2g–l. Two major cd T-cell phenotypes were were ‘dendritic’ in shape; however, thin parallel CD13+ observed in this dual-labelling experiment: TcR1+ cells structures resembling small-vessel walls were occasionally that were either WC1+ or WC1).WC1+ cells were gener- observed but were not further evaluated. In the uninfected ally small and round in appearance whereas WC1) cells calf (Figure 3d), CD13+ cells were mostly organized as a

Table 2 Effects of acute Babesia bovis infection on nave calf spleen volume and leucocyte contenta

Difference t-value (paired, Spleen Pre-infection Postinfection (post–pre) upper-tail, df =3) P-value

Volumeb (cm3) 506 € 49 996 € 91 491 € 49* 9Æ97 0Æ0011 Leucocyte concentrationc (·103 cells ⁄ cm3) Small 57Æ5€6Æ753Æ3€6Æ1 )4Æ3€5Æ9 )0Æ72 0Æ7385 Large 1Æ4€0Æ45Æ5€1Æ34Æ1€1Æ5* 2Æ71 0Æ0365 Total 58Æ9€7Æ058Æ8€6Æ0 )0Æ1€6Æ0 )0Æ02 0Æ5078 Leucocyte contentd (·106 cells ⁄ spleen) Small 29Æ3€4Æ852Æ8€7Æ523Æ5€6Æ1* 3Æ84 0Æ0156 Large 0Æ7€0Æ25Æ8€1Æ75Æ2€1Æ8* 2Æ84 0Æ0327 Total 30Æ0€4Æ758Æ8€8Æ628Æ8€7Æ0* 4Æ10 0Æ0131 aValues reported are mean € standard error; * indicates ‘difference’ significantly >0. bDetermined by MRI. cDetermined by differential cell counts on dissociated splenic biopsies. dCalculated as the paired product of spleen volume and differential cell count.

2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 37 D. A. Schneider et al. Parasite Immunology

Figure 1 Splenomegaly in calves following acute infection with Babesia bovis is associated with expansion and increased cellularity of the red pulp and loss of zonal definition around white pulp as shown in representative macroscopic (1Æ25· objective) and microscopic (10· objective) images. Haematoxylin and eosin; r, red pulp; w, white pulp; cyan [, periarteriolar lymphatic sheath; yellow outline, central arteri- ole; f, follicle; mz, marginal zone; dashed curve, follicle–marginal zone junction; {, band of cells within marginal zone; t, trabeculae; bars = 1 mm (1Æ25·), 200 lm (10·); dpi, days postinfection. discontinuous honeycomb-like network that spanned the apparent density of CD172a+ cells increased from 7 to red pulp and marginal zone with little zonal distinction. 14 dpi, and progressively obscured distinction between More sparsely stained CD13+ cells were also located at marginal zone and red pulp. the follicle–marginal zone junction and occasionally within the PALS. An unambiguous change in the distribution of Immunoreactivity for merozoite surface antigen-1 CD13+ cells was already evident at 7 dpi and persisted to (MSA-1) 14 dpi (Figure 3e,f), wherein the majority of CD13+ cells formed a distinct band at the follicle–marginal zone MSA-1 was localized in the spleen of B. bovis-infected junction. Sparsely stained CD13+ cells were also observed calves using monoclonal antibody BABB35 (29,30). within the PALS and outer margin of follicles between Immunoreactivity for BABB35 was not observed in unin- 7 and 14 dpi. Postinfection CD13+ cells surrounding fected or 7–9 dpi spleens. At 13 and 14 dpi, BABB35 the PALS were more sparsely stained and scattered by immunoreactivity was consistently observed within the 14 dpi. outer margins of all splenic follicles, being most dense Immunoreactivity specific for the myeloid marker near its junctions with the marginal zone and PALS CD172a (12) is shown in Figure 3g–i. CD172a+ cells were (Figure 4a). BABB35 immunoreactivity was generally numerous in the red pulp of the uninfected spleen. The punctate and appeared to be distributed along fine

38 2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 Volume 33, Number 1, January 2011 Spleen cell dynamics during babesial infection

(a) (d) (g) (j)

(b) (e) (h) (k)

(c) (f ) (i) (l)

Figure 2 Changes in the parafollicular distribution of phenotyped small leucocyte populations in the spleen of calves during acute infection with Babesia bovis. Shown are spleen sections dually immunolabelled for the T-cell markers CD3 (a–c) and CD4 (d–f), or for the cd T-cell markers TcR1-N24 (d chain) [TcR1; (g–i)] and WC1 (j–l). In particular, note the presence of bands (demarked by ‘{’) of CD3+ ⁄ CD4+ cells, ) CD3+ ⁄ CD4 cells, and TcR1+ ⁄ WC1+ cells within the marginal zone of the uninfected spleen (a ⁄ d, g ⁄ j). Arrows (g ⁄ j) indicate the red pulp ) ) location of TcR1+ ⁄ WC1 cells in uninfected spleen. Postinfection, note the progressive hyperplasia of CD3+ ⁄ CD4 and CD3+ ⁄ CD4+ cells ) (b ⁄ e and c ⁄ f) and TcR1+ ⁄ WC1 cells (h ⁄ k and i ⁄ l) which obscures these zonal distinctions. Also note the relative loss of TcR1+ ⁄ WC1+ cells from the marginal zone. Cyan [, margins of periarterial lymphatic sheath; yellow outline, central arterioles; dashed curve, follicle–mar- ginal zone junction; {, band of cells within marginal zone; bars = 200 lm; dpi, days postinfection.

‘dendritic’ structures (Figure 4b) but never clearly high- in these leucocyte measurements probably results from lighted any round cell bodies. Immunoreactivity for sampling a relatively small piece of spleen and the dispro- BABB35 was frequently co-distributed with structures portionate increase in red pulp postinfection. Despite this having sparse immunoreactivity for CD13. In contrast, limitation, these results are consistent with a local immune well-labelled CD13+ cells at the follicle–marginal zone response of nave animals to acute haemoparasite infec- junction were not immunoreactive for BABB35. tion and reflect the central importance of large leucocytes – monocytes ⁄ macrophages, DCs and NK cells – to the spleen-dependent immune response of nave calves to DISCUSSION B. bovis infection (31). The results of this study demonstrate that the spleen of In addition to these gross changes in splenic composi- calves doubles in volume and total cell content by 11– tion, changes were also observed in the distribution of 12 dpi. While retention of erythrocytes is a functionally phenotyped cells within and between the domains of the important contributor to splenomegaly, here we document spleen (summarized in Table 3). Regarding the splenic dis- that an acute hyperplasia of nucleated cells, which tribution of large leucocyte populations during acute obscures histological appreciation of zonal boundaries, B. bovis infection, observations with functional implica- also occurs and is similar to a recent report of the acute tions include the following: (i) loss of the relative accumu- response of nave mice to infection with Plasmodium lation of NK cells (CD335+) within the marginal zone, (ii) chabaudi (23). The acute hyperplasia in calves was charac- unambiguous early redistribution of iDCs (CD13+) to the terized by an approximate fourfold greater increase in junction of the marginal zone with follicles and PALS, (iii) large versus small splenic leucocytes. The variation noted subsequent juxtaposed appearance of an immunologically

2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 39 D. A. Schneider et al. Parasite Immunology

(a) (d) (g)

(b) (e) (h)

(c) (f) (i)

Figure 3 Changes in the parafollicular distribution of phenotyped large leucocyte populations in the spleen of calves during acute infection with Babesia bovis. Shown are representative sequential sections immunolabelled by a marker for NK cells (CD335; a–c) and a marker for immature dendritic cells (CD13; d–f). Also shown are sections immunolabelled by a marker of monocyte ⁄ macrophage cells (CD172a; g–i). (a–c) Note the postinfection decrease in CD335+ cells present as a band [see ‘{’ in (a)] within the marginal zone. (d–e) Note the postinfec- tion replacement of the vast honeycomb-like network of CD13+ cells by the appearance of a distinct row of CD13+ cells at the marginal zone junctions with the follicle and periarterial lymphatic sheath. (g–i) Finally, note the progressive postinfection increase in red pulp CD172a+ cells with expansion into the marginal zone. Cyan [, margins of periarterial lymphatic sheath; yellow outline, central arterioles; dashed curve, follicle–marginal zone junction; {, band of cells within marginal zone; bars = 200 lm; dpi, days postinfection.

(a) (b)

Figure 4 Parafollicular distribution of merozoite surface antigen-1 (MSA-1) in the spleen of a calf 14 dpi with Babesia bovis. MSA-1 was localized using monoclonal antibody BABB35. (a) Note the distribution of punctate Babb35 immunoreactivity (red) within the outer mar- gin of the follicle juxtaposed to the marginal zone and periarterial lymphatic sheath. Higher magnification (b) [corresponds to red box in (a)] shows BABB35 immunoreactivity is not associated with brightly fluorescent CD13+ (green) cells located at the follicle–marginal zone junction but instead follows a similar ‘dendritic’ course as the more weakly CD13-stained cell processes. Cyan [, margins of periarterial lymphatic sheath; yellow outline, central arterioles; dashed curve, follicle–marginal zone junction; bars = 100 lm (a), 25 lm (b). important B. bovis surface antigen (MSA-1+), and (iv) The marginal zone is a complex environment in which restriction of monocyte ⁄ macrophage (CD172a+) hyperpla- cell trafficking and interaction with blood-borne foreign sia to within the red pulp. takes place (32,33). Within the spleen of unin-

40 2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 Volume 33, Number 1, January 2011 Spleen cell dynamics during babesial infection

Table 3 Changes in the distribution of phenotyped leucocytes importance placed on NK ⁄ DC crosstalk. As such, early within the spleen of nave calves during acute Babesia bovis activation with narrow redistribution to these junctions a infection by 7 dpi may optimally position iDCs to encounter

c B. bovis merozoites and infected erythrocytes as they 7–9 dpi (pre-fever 13–14 dpi enter the parenchyma of the spleen. Cell Splenic days and (fever days Furthermore, immunoreactivity for merozoite surface phenotypes zonesb first fever day) 4 and 5) antigen-1 (MSA-1) – an antigen important to infectivity (40) – was first observed at 13–14 dpi and only in close CD3+ ⁄ CD4+ MZ flfl d proximity to these junctions with the marginal zone. It PALS ne ne RP ›› was noted that the punctate immunostaining for MSA-1 ) CD3+ ⁄ CD4 MZ ›› was accompanied by sparse CD13 staining and always in PALS ne ne juxtaposition to redistributed iDCs. We have previously RP ››› shown that maturation of splenic iDC from nave calves in TcR1+ ⁄ WC1+ MZ flflfl vitro results in the loss of CD13 expression and gain in PALS RP capacity to present antigen (12,41). Thus, similar to the ) TcR1+ ⁄ WC1 MZ ››› P. chabaudi model in mice (23), these results support the PALS hypothesis that iDC mature during processing of the para- RP ››› site and migrate as antigen-presenting cells to - CD335+ MZ flfl rich domains. PALS RP The spleen-dependent innate response of nave calves to CD13+ MZ fl overall but fl overall but infection with B. bovis is also characterized by early IL-12 with › at with › at production with subsequent IL-10 modulation (6), the junctions with junctions with major sources of which in cattle are iDCs and mono- follicle and follicle and cytes macrophages, respectively (8,14,42). We have also PALS PALS ⁄ PALS ›, weakly labelled ›, weakly labelled shown that monocytes ⁄ macrophages of cattle can produce RP flflfl NO with direct babesiacidal activity (14,27,43). It was CD172a+ MZ › interesting to note that following haemoparasitic infection, PALS intense acute hyperplasia of monocytes ⁄ macrophages is RP ››› restricted to the red pulp of both mice (23) and calves aArrows indicate change in phenotyped cellularity relative to the (present study). Thus, in addition to regulatory function uninfected spleen. Blank cells indicate no appreciable change through cytokine production, our collective findings are noted. consistent with monocytes ⁄ macrophages acting as effector b MZ, marginal zone; PALS, periarteriolar sheath; RP, red pulp. cells in close juxtaposition with infected erythrocytes as cdpi, days postinfection. they enter the splenic sinuses. dne, not examined (cellularity too dense for evaluation). Regarding the distribution of small leucocytes, dual- fected calves, iDCs were present as a network-like distri- labelling experiments demonstrated acute progressive accu- bution covering the marginal zone and red pulp whereas mulation of numerous CD3+ CD4) cells and TcR1+ NK cells appeared to accumulate only within the mar- WC1) cells within the red pulp. Thus, it is likely that at ginal zone. Crosstalk between NK cells and DCs is cru- least a portion of these accumulated lymphocytes were cial to the innate immune response (34–37), and in mice WC1) cd T cells. The role of these cells is still not clear involves secretion of IL-15 by DCs (38), which primes but as bovine WC1) cd T cells express CD2 and CD8, can NK cells to produce IFN-c, which in turn increases DC produce IFN-c in response to cytokine stimulation, and activity (39). Similarly, we have previously shown up-regu- are found in largest proportion in the spleen and intestine lation of IL-15 mRNA in the spleen of calves early after (15,16,44,45), it is intriguing to consider the possibility infection with B. bovis (15) and in vitro crosstalk that that cells with this phenotype might be the bovine func- requires NK cell-iDC contact (13). Given the initial tional equivalent of NKT cells (46–48). If so, then the co-population of the marginal zone with CD335+ and observed accumulation of these cells in the red pulp of CD13+ cells, it is possible that the first 4–6 days of nave calves infected with B. bovis is consistent with their B. bovis infection in calves involves critical NK ⁄ iDC cros- expected role in the transition from innate to acquired stalk and activation. The unambiguous redistribution of immunity. iDCs to the junction between marginal zones, follicles Our results are in agreement with previous reports and PALS is consistent with the immunological (49,50) that demonstrate relatively small accumulations of

2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 41 D. A. Schneider et al. Parasite Immunology

WC1+ cd T cells within the splenic marginal zones of REFERENCES uninfected calves. The splenic decrease in WC1+ cd T cells during the acute response of calves to B. bovis infection 1 Wright IG, Goodger BV & Clark IA. Immunopathophysiology may indicate their activation within the marginal zone is of Babesia bovis and Plasmodium falciparum infections. Paras- itol Today 1988; 4: 214–218. followed by redistribution to effector sites outside of the + 2 Clark IA & Jacobson LS. Do babesiosis and malaria share a spleen. Indeed, several reports indicate WC1 cd T cells common disease process? Ann Trop Med Parasitol 1998; 92: are most numerous and reactive within the blood of young 483–488. calves (45,49,51–53). 3 Gale KR, Waltisbuhl DJ, Bowden JM, et al. Amelioration of The observations made in this study were initiated by virulent Babesia bovis infection in calves by administration of the nitric oxide synthase inhibitor aminoguanidine. Parasite events occurring within the first few hours and days of Immunol 1998; 20: 441–445. inoculation as Babesial antigen entered the spleen via the 4 Clark IA & Alleva LM. Is human malarial coma caused, or circulation. The marginal sinus is an important route by merely deepened, by sequestration? Trends Parasitol 2009; 25: which blood-borne particles and nonlymphoid cells first 314–318. enter the spleen (17). Our observations in nave calves are 5 Trueman KF & Blight GW. The effect of age on resistance of cattle to Babesia bovis. Aust Vet J 1978; 54: 301–305. consistent with recent intravital imaging studies in rodent 6 Goff WL, Johnson WC, Parish SM, Barrington GM, Tuo W models (54–56) which document the early interactions and & Valdez RA. The age-related immunity in cattle to Babesia trafficking of several marginal zone cell types and the bovis infection involves the rapid induction of interleukin-12, importance of these events to the splenic immune interferon-gamma and inducible nitric oxide synthase mRNA responses. Our results, however, do not exclude the poten- expression in the spleen. Parasite Immunol 2001; 23: 463–471. 7 Goff WL, Johnson WC & Cluff CW. Babesia bovis immunity. tial relevance of initial antigen interaction with other In vitro and in vivo evidence for IL-10 regulation of IFN- zonal cell populations (e.g., PALS lymphocytes) to the gamma and iNOS. Ann N Y Acad Sci 1998; 849: 161–180. acute response of nave calves to B. bovis. 8 Goff WL, O’Rourke KI, Johnson WC, Lacy PA, Davis WC In summary, the results of this immunohistological & Wyatt CR. The role of IL-10 in iNOS and cytokine mRNA investigation have demonstrated dynamic change in the expression during in vitro differentiation of bovine mononu- clear phagocytes. J Interferon Cytokine Res 1998; 18: 139–149. distribution of several cell types thought to be impor- 9 Goff WL, Johnson WC, Tuo W, et al. Age-related innate tant to the acute spleen-dependent response of calves to immune response in calves to Babesia bovis involves IL-12 B. bovis infection. In particular, unambiguous redistribu- induction and IL-10 modulation. Ann N Y Acad Sci 2002; 969: tion of iDC to regions where parasites first enter the 164–168. spleen and evidence for further maturation and antigen 10 Riek RF. In Garnham PCC, Pierce AE & Roitt I (eds): Immu- nity to protozoa; a symposium of the British Society for Immu- processing seem noteworthy. The remarkable similarity of nology In Immunity to Babesiosis. Oxford, Blackwell Scientific these acute splenic responses of calves to B. bovis and Pub, 1963; 160–179. those reported in mice responding to P. chabaudi indicates 11 Varma S & Shatry AM. A technique for partial marsupialisa- that redistribution of splenic cells is central to the tion of the spleen in calves. Vet Rec 1980; 106: 127–128. acute immune response of nave animals to haemoparasite 12 Bastos RG, Johnson WC, Brown WC & Goff WL. Differential response of splenic monocytes and DC from cattle to micro- infection. bial stimulation with Mycobacterium bovis BCG and Babesia bovis merozoites. Vet Immunol Immunopathol 2007; 115: ACKNOWLEDGEMENTS 334–345. 13 Bastos RG, Johnson WC, Mwangi W, Brown WC & Goff WL. This work was supported by USDA-ARS-CWU-5348- Bovine NK cells acquire cytotoxic activity and produce IFN- gamma after stimulation by Mycobacterium bovis BCG- or 32000-010-00D. The authors especially recognize the Babesia bovis-exposed splenic dendritic cells. Vet Immunol expert technical contributions of Sallie Bayly who Immunopathol 2008; 124: 302–312. assisted in the splenic transposition surgeries, Tom Trus- 14 Goff WL, Johnson WC, Parish SM, et al. IL-4 and IL-10 inhi- cott for immunohistochemical advice, and Thomas Wil- bition of IFN-gamma- and TNF-alpha-dependent nitric oxide kinson and Rob Houston for MRI techniques. We thank production from bovine mononuclear phagocytes exposed to Babesia bovis merozoites. Vet Immunol Immunopathol 2002; 84: Duane Chandler and Amy Hetrick for their contributions 237–251. to the care and use of the animals. The authors thank 15 Goff WL, Johnson WC, Horn RH, Barrington GM Dr William C. Davis for his critical review of the manu- & Knowles DP. The innate immune response in calves to script. Mention of trade names or commercial products Boophilus microplus tick transmitted Babesia bovis involves or enterprises in this article is solely for the purpose of type-1 cytokine induction and NK-like cells in the spleen. Parasite Immunol 2003; 25: 185–188. providing specific information and does not imply recom- 16 Goff WL, Storset AK, Johnson WC & Brown WC. Bovine mendation or endorsement by the US Department of splenic NK cells synthesize IFN-gamma in response to IL-12- Agriculture.

42 2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 Volume 33, Number 1, January 2011 Spleen cell dynamics during babesial infection

containing supernatants from Babesia bovis-exposed monocyte 35 Degli-Esposti MA & Smyth MJ. Close encounters of different cultures. Parasite Immunol 2006; 28: 221–228. kinds: dendritic cells and NK cells take centre stage. Nat Rev 17 Mebius RE & Kraal G. Structure and function of the spleen. Immunol 2005; 5: 112–124. Nat Rev Immunol 2005; 5: 606–616. 36 Ferlazzo G. Natural killer and dendritic cell liaison: recent 18 Wykes MN & Good MF. What really happens to dendritic insights and open questions. Immunol Lett 2005; 101: 12–17. cells during malaria? Nat Rev Microbiol 2008; 6: 864–870. 37 Reschner A, Hubert P, Delvenne P, Boniver J & Jacobs N. 19 Cahalan MD, Parker I, Wei SH & Miller MJ. Two-photon tis- Innate lymphocyte and dendritic cell cross-talk: a key factor in sue imaging: seeing the in a fresh light. Nat the regulation of the immune response. Clin Exp Immunol Rev Immunol 2002; 2: 872–880. 2008; 152: 219–226. 20 Sumen C, Mempel TR, Mazo IB & von Andrian UH. Intravi- 38 Koka R, Burkett P, Chien M, Chai S, Boone DL & Ma A. tal microscopy: visualizing immunity in context. Immunity Cutting edge: murine dendritic cells require IL-15R alpha to 2004; 21: 315–329. prime NK cells. J Immunol 2004; 173: 3594–3598. 21 Cahalan MD & Parker I. Choreography of cell motility and 39 Lucas M, Schachterle W, Oberle K, Aichele P & Diefenbach interaction dynamics imaged by two-photon microscopy in A. Dendritic cells prime natural killer cells by trans-presenting lymphoid organs. Annu Rev Immunol 2008; 26: 585–626. interleukin 15. Immunity 2007; 26: 503–517. 22 Theer P & Denk W. On the fundamental imaging-depth limit 40 LeRoith T, Berens SJ, Brayton KA, et al. The Babesia bovis in two-photon microscopy. J Opt Soc Am A Opt Image Sci Vis merozoite surface antigen 1 hypervariable region induces sur- 2006; 23: 3139–3149. face-reactive antibodies that block merozoite invasion. Infect 23 Leisewitz AL, Rockett KA, Gumede B, Jones M, Urban B Immun 2006; 74: 3663–3667. & Kwiatkowski DP. Response of the splenic dendritic cell 41 Zhuang Y, Mwangi W, Brown WC, Davis WC, Hope JC population to malaria infection. Infect Immun 2004; 72: 4233– & Palmer GH. Characterization of a Phenotypically Unique 4239. Population of CD13+ Dendritic Cells Resident in the Spleen. 24 Goff WL, McElwain TF, Suarez CE, et al. Competitive Clin Vaccine Immunol 2006; 13: 1064–1069. enzyme-linked immunosorbent assay based on a rhoptry- 42 Hope JC, Thom ML, McCormick PA & Howard CJ. Interac- associated protein 1 epitope specifically identifies Babesia tion of antigen presenting cells with mycobacteria. Vet Immu- bovis-infected cattle. Clin Diagn Lab Immunol 2003; 10: 38–43. nol Immunopathol 2004; 100: 187–195. 25 Goff WL, Molloy JB, Johnson WC, et al. Validation of a com- 43 Johnson WC, Cluff CW, Goff WL & Wyatt CR. Reactive oxy- petitive enzyme-linked immunosorbent assay for detection of gen and nitrogen intermediates and products from polyamine antibodies against Babesia bovis. Clin Vaccine Immunol 2006; degradation are Babesiacidal in vitro. Ann N Y Acad Sci 1996; 13: 1212–1216. 791: 136–147. 26 Goff WL, Johnson WC, Molloy JB, et al. Validation of a com- 44 MacHugh ND, Mburu JK, Carol MJ, Wyatt CR, Orden JA petitive enzyme-linked immunosorbent assay for detection of & Davis WC. Identification of two distinct subsets of bovine Babesia bigemina antibodies in cattle. Clin Vaccine Immunol gamma delta T cells with unique cell surface phenotype and 2008; 15: 1316–1321. tissue distribution. Immunology 1997; 92: 340–345. 27 Goff WL, Johnson WC, Wyatt CR & Cluff CW. Assessment of 45 Wyatt CR, Madruga C, Cluff C, et al. Differential distribution bovine mononuclear phagocytes and for induced of gamma delta T-cell receptor lymphocyte subpopulations in L-arginine-dependent nitric oxide production. Vet Immunol blood and spleen of young and adult cattle. Vet Immunol Immunopathol 1996; 55: 45–62. Immunopathol 1994; 40: 187–199. 28 Jain NC. Examination of the Blood and . In 46 Bannai M, Kawamura T, Naito T, et al. Abundance of uncon- Essentials of Veterinary Hematology. Philadelphia, Lea & Febiger, ventional CD8(+) natural killer T cells in the large intestine. 1993: 1–18. Eur J Immunol 2001; 31: 3361–3369. 29 Goff WL, Davis WC, Palmer GH, et al. Identification of Babe- 47 Osman Y, Kawamura T, Naito T, et al. Activation of hepatic sia bovis merozoite surface antigens by using immune bovine NKT cells and subsequent liver injury following administration sera and monoclonal antibodies. Infect Immun 1988; 56: 2363– of alpha-galactosylceramide. Eur J Immunol 2000; 30: 1919– 2368. 1928. 30 Hines SA, Palmer GH, Jasmer DP, Goff WL & McElwain TF. 48 Kitamura H, Iwakabe K, Yahata T, et al. The natural killer T Immunization of cattle with recombinant Babesia bovis mero- (NKT) cell ligand alpha-galactosylceramide demonstrates its zoite surface antigen-1. Infect Immun 1995; 63: 349–352. immunopotentiating effect by inducing interleukin (IL)-12 31 Brown WC, Norimine J, Knowles DP & Goff WL. Immune production by dendritic cells and IL-12 receptor expression on control of Babesia bovis infection. Vet Parasitol 2006; 138: NKT cells. J Exp Med 1999; 189: 1121–1128. 75–87. 49 Clevers H, MacHugh ND, Bensaid A, et al. Identification of a 32 Schmidt EE, MacDonald IC & Groom AC. Comparative bovine surface antigen uniquely expressed on CD4-CD8- aspects of splenic microcirculatory pathways in mammals: the receptor gamma ⁄ delta+ T lymphocytes. Eur J Immunol 1990; region bordering the white pulp. Scanning Microsc 1993; 7: 20: 809–817. 613–628. 50 Hein WR & Mackay CR. Prominence of gamma delta T cells in 33 Steiniger B, Timphus EM & Barth PJ. The splenic marginal the ruminant immune system. Immunol Today 1991; 12: 30–34. zone in humans and rodents: an enigmatic compartment and 51 Brown WC, Davis WC, Choi SH, Dobbelaere DA & Splitter its inhabitants. Histochem Cell Biol 2006; 126: 641–648. GA. Functional and phenotypic characterization of WC1+ 34 Della Chiesa M, Sivori S, Castriconi R, Marcenaro E & gamma ⁄ delta T cells isolated from Babesia bovis-stimulated T Moretta A. Pathogen-induced private conversations between cell lines. Cell Immunol 1994; 153: 9–27. natural killer and dendritic cells. Trends Microbiol 2005; 13: 52 Okragly AJ, Hanby-Flarida M, Mann D & Baldwin CL. 128–136. Bovine gamma ⁄ delta T-cell proliferation is associated with

2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44 43 D. A. Schneider et al. Parasite Immunology

self-derived molecules constitutively expressed in vivo on 55 Aoshi T, Carrero JA, Konjufca V, Koide Y, Unanue ER mononuclear phagocytes. Immunology 1996; 87: 71–79. & Miller MJ. The cellular niche of Listeria monocytogenes 53 Price SJ, Sopp P, Howard CJ & Hope JC. Workshop cluster infection changes rapidly in the spleen. Eur J Immunol 2009; 1+ gammadelta T-cell receptor T cells from calves express high 39: 417–425. levels of interferon-gamma in response to stimulation with 56 John B, Harris TH, Tait ED, et al. Dynamic Imaging of interleukin-12 and -18. Immunology 2007; 120: 57–65. CD8(+) T cells and dendritic cells during infection with 54 Aoshi T, Zinselmeyer BH, Konjufca V, et al. Bacterial entry to Toxoplasma gondii. PLoS Pathog 2009; 5: e1000505. the splenic white pulp initiates antigen presentation to CD8+ T cells. Immunity 2008; 29: 476–486.

44 2010 Blackwell Publishing Ltd, Parasite Immunology, 33, 34–44