. ‘ . .. . . 5.. 7‘!» ‘0!!!) '2. f... 1.3!?! .1 {'3'} . o... I.) .4 . 14“.? HA . .QX: .1 a . I... 3:} L. .5316: , 1.: I i I V III} Trutlv-IL ID 'E u ' l;- uolul; use .619". ‘ . .5 of. v ‘ :01. if. . .ulvt..'-,;!epf. t: (to. any 5.1.1- y .c 5 {itiit $5111.23. 3:; l‘ r. :‘36 v; p 753.33.15.61 15%;: at?! Citirzvliitl‘: fr: '0‘}; l! {xvrrfll \ Illal 'Ifls .J' {lit-ii???) .1 .1!!! rlvlriuozitrw - 7:. x.“ 15.11". .ilirup r lit}... .. Eirp'v ‘11!»li.l "if-8'8 CHI IGAN STAT lull/l 1|le UllllllllllIfl‘llllllllllllll 914 0025 This is to certify that the dissertation entitled STUDIES ON THE IMMUNOBIOLOGY AND SERODIACNOSIS OF EPERYTHROZOfl SUIS IN SWINE presented by Nariaki Nonaka has been accepted towards fulfillment of the requirements for PhD degreein Large Animal Clinical Sciences lell TLL flajor professor Date 1/1133 MS U is an A [firmalive Aclion/ Equal Opportunity Institution 0-12771 \ l. i 1.13315.le jMichlgen State; ; University ! l__ PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ' _JL_IL_J :JWV— - II II i ll MSU Is An Affirmative Action/Equal Opportunity Institution cMcMuna-DJ STUDIES ON THE IMMUNOBIOLOGY AND SERODIAGNOSIS OF EPERYTHROZOON’SUIS IN SWINE By Nariaki Nonaka A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Large Animal Clinical Sciences 1993 ABSTRACT STUDIES ON THE IMMUNOBIOLOGY AND SERODIAGNOSIS 0F EPERYTHROZOON SUIS IN SWINE By Nariaki Nonaka Swine eperythrozoonosis (EPE) , caused by Eperythrozoon suis (Es), is characterized by acute icteroanemia. Chronic infections are believed to be associatd with impaired growth and reproductive performance although the nature of chronic EPE is poorly understood in part because of the lack of reliable diagnostic techniques for detecting subclinical infections. This study was conducted to evaluate the current EPE serotest, indirect hemagglutination assay (IHA) and develop improved serodiagnostic assays. A Petri dish-erythrocyte in vitro culture was used to maintain Es. After intensive screening, the following conditions were found to be optimal for maintenance of erythrocyte attachment, glycolysis and metabolic incorporation of radiolabeled methionine by Es: heparin as the anticoagulant, incubation with Eagle's medium under 5 or 10% C02, supplementation with inosine and fetal calf serum, and refreshment of medium. The effects of incubation temperatures and serum absorption with normal pig erythrocytes on IRA, and the comparisons of IHA, ELISA and western blots using Es" and E3“ IHA antigens prepared from infected and noninfected blood, respectively, demonstrated that antibodies detected in IHA were primarily directed at host antigens, not Es antigens, suggesting a lack of specificity of IRA for detecting Es antibodies. Antigen preparations were developed by: 1) high speed centrifugation of Es dissociated from erythrocytes using EDTA, and Es free in plasma; and 2) differential centrifugation (500g, 10,0009 and 100,0009) of Es dissociated from hemolysates. The second method provided the most Es antigens, especially in 100,000g sediments (100K antigen), as determined by western blot analysis. Identical band appearance of Es antigens in western blots and autoradiographs of the 100K antigen prepared from radiolabeled Es confirmed that the observed antigens were Es derived. ELISA and dot blot procedures were developed using Es+ and Es“ 100K antigens. In each procedure, CrudeEs-Test and NetEs-Test, based on the Es+ antigen value only, and the subtraction of the ES' antigen value from the Es+ antigen value, respectively, showed high sensitivity with sera obtained from experimentally infected pigs. With sera from naturally exposed pigs, ELISA NetEs-Test was superior to CrudeEs-Test in detecting Es antibodies, as determined by comparison with western blot data. ACKNOWLEDGMEMES To finish this work, a lot of wonderful people have been helping me and I wish to express my deep appreciation for their heartful contributions: My advisory committee - Brad J. Thacker, Robert W. Bull, Tjaart W. Schillhorn van Veen, Robert Holland and John C. Baker for their support, advice, patience and tremendous amount of time and effort spent on my project; John M. Davis for his precise advice and suggestions in molecular biology; Alice Murphy and Ann R. Donoghue for their technical assisstance in parasitology and friendship during this study; John A. Gerlach, Peggy Bull, Eileen Thacker, Robert D. Walker, James B. Jensen, Harold Tvedten and Cunqin Han for their technical support in protein molecular biology, in vitro culture, electron microscopy and animal care; Richard Gatzmeyer for kindly providing pig sera; Masao Kamiya, Raffaele R. Roncalli and Tetsuei Ohta for providing this oppotunity for me to study abroad; My family - Isao, Kayoko, Masami, Shiho and Daiki for their countinuous support and encouragement. iv TABLE OF CONTENTS Page LIST OF TABLES ---------------------------------------------- x LIST OF FIGURES —— — ------------------------- xii LIST OF ABBREVIATIONS --- ' . xviii INTRODUCTION - - 1 CHAPTER 1 LITERATURE REVIEW 5 Cussxncnnon up Spscrns or prxrrnxozoou ------------ 6 Cnnononoorcu. Rlvrlw or Stun Ennunnozoonosrs ------- 8 Mouuonoe! m Drvxnosnnn --13 Tnnlsnr ssrou ---- — — 16 Ctr-1cm. Srols up Panama! -------------------------- 21 Conn Aoounu'rrnn — — — 1 25 Parnopnrsroroo! --------------------------------------- 27 3081' Immirn'r— — ——-—— -------é -------- 31 DIAGNOSIS ------------------- —— — 36 Tmmn m Corner. — 39 CHAPTER 2 IN VITRO CULTIVATION OF EPERYTHROZOON SUIS ----- 41 Abstract---- . -------- —— —— 42 Introduction ----------------------------- 44 Materials and Methods ---- ----46 EXPERIMENTAL INFECTIONS ----------------------- 46 Organism. —— = --------------- 46 Animals and inoculation of E. suis. ——— 46 CULTURE — 47 Media. — — — —— — ——— ——— —— 47 Incubators.——— = —— — —— --------------- 48 Procedures. --------------------------- —— 49 Measurements. — —— — 50 Experimental design and statistical analysis. ----50 Results = == —— —— — =— --52 STUDY I. PRELIMINARY SCREENING or CULTURE CONDITIONS ------ 52 EXP I-A. Anticoagulant. =——— — ------- 52 EXP I-B. Gaseous environment. -------------------- 52 Elm I-C. Medium. --------------------------------- 54 STUDY II. MEASURMENT or E. 8015 METABOLIC ACTIVITY AND FURTHER SCREENING or CULTURE CONDITIONS ----------------- 56 EXP II-A. Glucose consumption and lactate and pyruvate production. ----------------------------- 56 EXP II-B. Addition of inosine to medium. --------- 56 EXP II-C. Addition of EDTA to medium. ------------ 58 EXP II-D. Selection of serum and incubator. ------ 62 EXP II-E. Refreshment of medium. ----------------- 62 Discussion ------------------------------------------- 67 References ------------------------------------------- 77 vi CHAPTER 3 THE LACK OF A SPECIFICITY OF THE INDIRECT HEMAGGLUTINATION ASSAY FOR SWINE EPERYTHROZOONOSIS AND DEVELOPMENT OF EPERYTHROZOON SUIS SPECIFIC ANTIGENS ----------- 80 Abstract—— —— -— ————— ------- 81 Introduction — — .......... 83 Materials and Methods ------------------------------ 86 Organism.---- — -------- 86 Animals and experimental infection. -------------- 86 Serum samples. ----------------------------------- 87 Antigen preparation. ----------------------------- 87 Preparation of E. suis sensitized sheep erythrocytes (S-RBC) and glutaraldehyde treated sheep erythrocytes (G-RBC) fer IRA. ------ 90 Indirect hemagglutination assay (IRA). ----------- 91 Enzyme-linked immunosorbent assay (ELISA). ------- 92 Electrophoresis and western blot analysis of the antigens.— — — ------------------- 93 35Sqmethionine radiolabeling of E. suis. --------- 95 Statistical analysis. ---------------------------- 96 Results ---------------------------------------------- 98 STUDY I . EVALUATION or IRA ---------------------------- 98 The effect of incubation temperature on IRA results. ......................................... 93 The effect of preincubation of serum with pig erythrocytes at different temperatures. ---------- 99 IRA using 83* and Es‘ antigen preparations. ------ 99 ELISA using Es+ and ES“ IRA antigen preparations. ----------------------------------- 101 SDS-PAGE and western blot analysis of IRA antigens. --------------------------------------- 104 vii STUDY II. EVALUATION OF ANTICNS PREPARED BY THREE DIFFERENT METHODS (METHODS EDTA-ES, PLASMA-ES AND D'SPIN—ES) ----------------------------------------- 108 Antigens prepared by Method EDTA-Es. ------------ 108 Antigens prepared by Method Plasma-Es. ---------- 110 Antigens prepared by.Method D'spin-Es. ---------- 112 STUDY III. 100K ANTICN PREPARATION ------------------- 114 western blots with various sera. --------------- 114 western blots using antigens prepared at different stages after experimental infection. -------------------------------------- 114 SDS-RAGE and autoradiography'using'35s- methionine labeled antigen preparations. -------- 117 western blots with different post-infection sera of the same pigs. — — --------------- 117 Discussion ------------------------------------------ 122 References ------------------------------------------ 139 CHAPTER 4 DEVELOPMENT OF IMPROVED SERODIAGNOSTIC TESTS FOR SWINE EPERYTHROZOONOSIS ------------------- 145 Abstract ---------------------------------- 146 Introduction --------------------------------------- 148 Materials and Methods ----------------------------- 150 Organism. — ---------------------- 150 Animals and experimental infection. ------------- 150 Serum samples. ---------------------------------- 151 Indirect hemagglutination assay (IRA). ---------- 151 100K antigen preparation. ----------------------- 153 Enzyme-linked immunosorbent assay (ELISA). ------ 154 Dot blot (DB). ---------------------------------- 157 viii Comparison of specificities of the antiepig IgG and protein A reagents by western blot. ----- 159 western blot analysis of field samples. --------- 160 Statistical analysis. ---------- 162 Results — —163 Comparison of western blots using antiepig IgG and Protein A. --163 Serum samples from experimentally infected pigs. —— ----- 163 Field serum samples. ----- 170 western blot analysis of the field samples. ----- 172 Discussion ------------------------------------------ 180 REMflmTS ------------------------------------------- 180 EXPERIMENTAL INrECTION — ------------- 181 IHELDEHTA ----------------------- 186 ‘WESTERN BLOT.ANALYSIS ' --------------------- 188 CONCLUSION— — — ------------------------------- 190 References ------------------------------------------ 192 OVERALL CONCLUSION ---- --------------------- 196 OVERALL BIBLIOGRAPHY -------------------------------------- 199 Table 0-1 1-1 3-1 3-2 LIST OF TABLES Page - Swine eperythrozoonosis test (indirect hemagglutination assay) performed by the Michigan Animal Health Diagnostic Laboratory during 1983 to 1992. -------------------------------------------------- 3 Species of Eperythrozoon and their characteristics.—- — — _ ..................... 7 Differentiation of Eperythrozoon suis and E. Fame ----------------------------------------------- 12 Eperythrozoon and their possible arthropod vectors. .............................................. 19 A list of drugs tested for activity against Eperythrozoon suis. ................................... 40 The change in the percentage of parasitized erythrocytes in whole blood culture with different anticoagulants and incubators. ------------------------ 53 The change in the percentage of parasitized erythrocyte in whole blood (WB) and infected red blood cell (IRBC)‘:cultures. -------------------------- 53 The Change in the percentage of parasitized RBC's in infected red blood cell culture with various media? and whole blood culture. ----------------------- 55 Alteration of indirect hemagglutination assay (IRA) titer of sera incubated at 37 °C instead of 4 °C used in the standard IRA titer.----------------e ----- 98 Alteration Of indirect hemagglutination assay (IRA) titer of sera preincubated with pig erythrocytes at 4 °C and 37 °C --------------------------------------- 100 IRA titers of three serum samples using different antigen preparations. -------------------------------- 101 4-1 4-3 4-8 Eperythozoon suis specific bands detected in the western blots using 100K E. suis positive (Es‘) and negative (ES') antigen preparations with various pre- and post-infection sera from 4 pigs. ------------ 120 Determination of cut off values used for classification of test sera to positive or negative by ProA (protein A)- and IgG-ELISA'S and dot blot.-uu-156 The results of ProA (protein A)- and IgG-ELISAIS, dot blot and indirect hemagglutination assay (IRA) with sera obtained from 8 experimentally infected pigs at pre-infection and various post-infection tim.e —- --------- 166 The number of positive field sera detected by CrudeEs-Test or NetEs-Test in ProA (protein A)- and IgG-ELISA'S and dot blot, and by indirect hemagglutination assay (IRA). ------------ 171 P values of Chi-square contingency analysis between the positive/negative results of ProA (protein A)- and IgG-ELISA'S, dot blot and indirect hemagglutination assay (IRA). =—— ------------ 172 Field sera in combinations of positive/negative results determined by ProA (protein A)- and IgG- ELISA'S and dot blot CrudeEs-Test‘. ==—— — ----- 173 The number of positive and negative field sera as determined by ProA (protein A)- and IgG-ELISA'S, dot blot and indirect hemagglutination assay (IRA), stratified by the total number of 7 E. suis specific antigen bands detected in IgG western blots. ---------------------------- 174 Positive or negative field sera as determined by ProA (protein A)- and IgG-ELISA'S, dot blot and indirect hemagglutination assay (IRA), that reacted to each E. suis specific antigen band. --------------- 178 Chi-square contingency analysis between the positive/negative results of field samples as determined by ProA (protein A)- and IgG-ELISA'S, dot blot and indirect hemagglutination assay (IRA), and the detection of each E. suis specific antigen band in IgG western blots. --------------------------- 179 xi LIST OF FIGURES Figure Page 2-1 2-2 2-3 The percentage of parasitized erythrocytes (A) and cumulative glucose (Glc) consumption and lactate (Lac) production (B) in red blood cell (RBC) culture with reduced Eagle's medium. IRBC = Infected RBC culture, URBC = Uninfected RBC culture. Error bars represent standard deviations. -------------------------------- 57 The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (8) in red blood cell (RBC) culture with reduced Eagle's medium containing or not containing inosine (rEM and rEMI, respectively). IRBC = Infected RBC culture, URBC = Uninfected RBC culture. * = Significant difference (0.015P<0.05) between rEM and rEMI Observed in IRBC culture. Error bars represent standard deviationS.I---59 The percentage of parasitized erythrocytes (A) and cummlative glucose consumption (8) in red blood cell (RBC) culture with reduced Eagle's medium containing inosine and EDTA (rEMI-EDTA) and reduced Eagle's medium containing only inosine (rEMI). IRBC = Infected RBC culture, URBC = Uninfected RBC culture. a: = Significant difference (P<0.01) between rEMI and rEMI-EDTA in IRBC culture. Error bars represent standard deviations. ------------------- 61 xii 2-4 3-1 The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (8 and C) in red blood cell (RBC) culture with reduced Eagle's medium with inosine (rEMI) containing 10% fetal calf serum (FCS) or pig serum (PS), incubated in a 10% C02 incubator or a candle jar. 8. Cumulative glucose consumption in IRBC and URBC culture with rEMI containing 10% FCS or PS, incubated in a 10% C02 incubator. C. Cumulative glucose consumption in IRBC and URBC culture with rEMI containing 10% FCS, incubated in a 10% C02 incubator or a candle jar. *t's 8 Significant differences (P<0.01) between PCS-€02 and PS-COz (##1), between PCS-IRBC and PS-IRBC (##2), and between C02-IRBC and Candle- IRBC (##3). Error bars represent standard deviations. — __ .......................... 53 The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (B) in infected red blood cell (IRBC) culture with medium refreshment every 12 hours and 24 hours, and without changing medium. at and n = Significant differences (*=0.015P<0.05, **=P<0.01) between "No change" and "12 hours" (*1, *tl), between "No change" and "24 hours” (*2, **2) and between "12 hours" and "24 hours” (9:3). Error bars represent standard deviations. —= — ............................. 55 Changes in optical densities by serum dilution in IgM ELISA using various Eperythrozoon suis positive (Pl-4) and negative (NI-2) antigen preparations for indirect hemagglutination assay (IRA). A and D: The optical density (CD) of ELISA using post- infection sera of pig number 1 and 2. B and E: The ELISA OD using pre-infection serum Of pig number 1 and 2. C and F: The difference of OD of pre- and post-infection sera of pig number 1 and 2, rescectively. P1 = E. suis positive antigen obtained from University of Illinois. P2-4 = E. suis positive antigens prepared from experimentally infected pigs. Nl-Z = E. suis negative antigens prepared from uninfected pigs. ----------------------- 102 xiii 3-4 Comparison of the protein composition of various Eperythrozoon suis positive (33*) and negative (ES‘ ) antigen preparations for indirect hemagglutination assay (IRA) by SDS-PAGE Stained by the Coomassie brilliant blue. Each lane of a 9% gel was loaded with 40 pg of reduced conditioned antigen preparation. P1 = Es+ antigen preparation obtained from the University of Illinois, P2-4 = Es+ antigens prepared from experimentally infected pigs, N1 and 2 = 38‘ antigens prepared from uninfected pigs. MWM aIMolecular weight marker. ------ 105 western blot analysis of indirect hemagglutination assay antigen preparations. A: western blots using 40 pg of Eperythrozoon suis positive (Es+) and negative (Es') antigens with 1:100 diluted immune and nonimmune reference sera, detected by horse raddish peroxidase (RRP) conjugated anti-pig IgM and 4-chloro—1-naphthol. 8: western blots using 40 pg of Es+ and ES' antigens with 1:100 diluted immune and nonimmune reference sera and without serum, detected by RRP conjugated anti-pig IgG and 4-chloro-1-naphthol. Lane P1: Es+ antigen obtained from University of Illinois. Lanes P2-4: Es+ antigens prepared from experimentally infected animals. Lanes N1 and 2: ES' antigens prepared from uninfected pigs.—— — -------------------------- 106 western blot analysis of Eperythrozoon suis positive (ES+) and negative (Es‘) antigens prepared. by Method EDTA-Es. The blots of the 38+ and Es- antigens (8 pg in each lane) were incubated with the immune and nonimmune reference sera and detected by horse raddish peroxidase (RRP) conjugated anti-pig IgG and 4-chloro-1-naphthol. MWM= Molecular weight marker.-_- — --— ----- 109 xiv 3-6 3-8 western blot analysis of Eperythrozoon suis positive (Es+) and negative (Es‘) antigens prepared by Method Plasma-Es. The blots of the Es+ and ES' antigen preparations were cut into strips and incubated with no serum (B), and 1:100 diluted immune reference serum (I) and 4 serum samples each from 4 experimentally infected pigs with E. suis. The 4 serum samples were collected before infection, and at 2, 4 and 8 weeks post infection (from left to right). The 2 weeks post infection serum was not collected from pig number 3. The antigens were detected by horse raddish peroxidase (HRP) conjugated anti-pig IgG and 4-chloro-1- naphthol. MWM = Molecular weight marker. ------------- 111 western blot analysis of various antigens prepared by Method D'Spin-Es. The blots of 40 pg of each Eperythrozoon suis positive (Es+) and negative (Es- ) thaw, 500, 10K and 100K antigen preparations were incubated with the 1:100 diluted immune and nonimmune reference sera. The antigens were detected by horse radish peroxidase (RRP) conjugated rabbit anti-pig IgG and 4-chloro-1- naphthol. MWM = Molecular weight marker. An arrow shows a 90 Rd antigen Observed in 100K antigen preparation. ----------------------------------------- 113 western blot analysis of the 100K antigen using sera from 5 experimentally infected pigs with Eperthrozoon suis. Blots of both E. suis positive (Es+) and negative (ES') 100K antigen preparations were incubated with no serum (B), immune reference sera (I), nonimmune reference sera (N), and three sera each from five different pigs. The three sera were collected before infection, and at 2 and 4 weeks post infection (from left to right). MWM = Molecular weight marker. ----------------------------- 115 western blot analysis of the 100K antigen prepared at the various times before or after infection. The blot of 40 pg of each antigen preparation was incubated with the 1:100 diluted immune and nonimmune reference sera and detedted with horse raddish peroxidase (RRP) conjugated rabbit anti-pig IgG and 4-chloro-l-naphthol. The antigens used were prepared from blood collected before splenectomy (N1), after splenectomy and before infection (N2), at the first parasitemic phase (P1), after the first parasitemic phase (NPl), at the second parasitemic phase (P2), and after the second parasitemic phase (NP2). MWM = Molecular weight marker. ---------------------------------------------- 116 XV 3-9 3-10 Comparison of the autoradiograph of radiolabeled 100K antigen preparations and western blots of the 100K antigen preparations. Lanes 1 and 2 Show the western blots of the Eperythrozoon suis positive (Es+) and negative (ES') antigen preparations (50 pg each) incubated with the immune reference sera. Lanes 3 and 4 Show the autoradiograph of the sulfur 35 labeled Es+ and Es‘ antigen preparations (30 pg each). Lanes 5 and 6 Show the western blots of the Es+ and ES' antigen preparations (50 pg each) incubated with the nonimmune reference sera. MWM = Molecular weight marker. Left arrows Show 101, 101, 90, 81, 71, 66, 47 and 24 Rd proteins. Right arrow shows 37 kd protein. --------------------------------- 118 western blot analysis of 100K antigen preparation using sera collected weekly from an experimentally infected pig. The blots of the Eperythrozoon suis positive (Es+) and negative (Es‘) antigen preparations were incubated with no serum (B), the immune reference serum (I), the nonimmune reference serum (N), sera collected before infection (8.1.), and sera collected at 1 to 15 weeks post infection (shown by number of weeks post infection). Arrows show antigens recognized only by sera collected during late infection. MWM = Molecular weight marker.---------—---- ------------------------------- 121 Comparison of antibody detection by protein A.and rabbit anti-pig 196 by western blot. Western blotting were performed using Eperythrozoon suis positive (Es+) and negative (ES') 100K antigens and immune (IS) and nonimmune (NS) reference sera. Antibodies bound to antigens were detected by 125I- protein A. (105 CPM/ml of the buffer) or horse radish peroxidase (RRP) conjugated rabbit anti-pig IgG (1:2000 diluted in the buffer). 1mil-protein A treated nitrocellulose membranes (NC'S) were exposed to Kodak x-omat RP film over night at -80 °C. RRP conjugated anti-pig IgG treated NC'S were incubated with hydrogen peroxide containing 4- chloro-l-naphthol for two hours. MWM = Molecular weight marker.-_-- — —— ---------------------- 164 xvi The ProA (protein A)- and IgG-ELISA'S, dot blot and indirect hemagglutination assay (IRA) with sera of experimetally infected pigs. Pre-infection and 1, 2, 3, 4, 6, 8, 10 and 12 weeks post-infection (WPI) sera from 8 experimentally infected pigs were assayed with the ProA-ELISA (A), IgG-ELISA (B), dot blot (C) and IRA (D). In the two ELISAIS and dot blot, the mean optical densities (OD) and percent binding in CrudeEs-Test and NetEs-Test were presented in addition to the mean OD's and percent binding of the ELISA'S and dot blot using Es‘ antigen (N-Ag value). IRA titer was converted to logs (x/10). The dash lines Show the cut off values of CrudeEs-Test and NetEs-Test. Error bars represent standard deviations of each value. --------- 168 xvii LIST OF ABBREVIATIONS A. sp. ---- —— — ----Anaplasma sp. ABTS---2,2'-Azino-Bis(3-Ethy1benzthiazoline-6-Sulfonic).Acid AHDL-uuu- ———— --Animal Health Diagnostic Laboratory ATP -------------------------------------- Adenine Triphosphate B. sp. -------------------------------------------- Babesia Sp. BSA -------------------------------------- Bovine Serum Albumin CBC -------------------------------------- Complete Blood Count CPM ------------------------------------------ Count Per Minute DB —— —--------------‘ -------------------------- Dot Blot DNA ---- — ------------------- Deoxyribo Nucleic Acid E. sp. -------------------------------------- Eperythrozoon Sp. EDTA ------------------------- Ethylenediamine Tetraacetic Acid ELISA ---------------------- Enzyme-Linked Immuno Sorbent Assay EM ---- — — — —=Eagle's Miminum Essential Medium EMI ------------------------------------- EM containing Inosine EPE ----------------------------------- Swine Eperythrozoonosis Es.— —— - — — — — ---------- Eperythrozoon suis Es+V------------------------ ------- Eperythrozoon suis positive ES' ------------------------------- Eperythrozoon suis negative EXP ------------------------------------------------ Experiment FCS ------------------------------------------ Fetal Calf Serum G-RBC ------------- Glutaraldehyde treated Sheep Red Blood Cell xviii R. sp. —— —— ----Raemobartonella sp. HEPES.---N-2-Hydroxy-Ethylpiperazine-N'-2-Ethanesulfonic Acid HRP —— ——- =— --Horse Raddish Peroxidase IRA -—+ — Indirect Hemagglutination Assay IM — — --Intramuscu1ar IRBC-n— ——--Infected Red Blood Cell IV ........... _=_ __ Intraveneous MM -— — ------------------ Mycoplasma Medium MW ------ — — — -------------- Molecular weight N-Ag ------ — — ----------- Negative Antigen NC — — - ——— Nitrocellulose 0D .— — ------ Optical Density P. Sp. .—— — ------------------------------- Plasmodium sp. PAGE ------------ —POlyacry1amide Gel Electrophoresis PBS — =— — Phosphate Buffered Saline PCR-— — — Polymerase Chain Reaction ProA — — -------------------- Protein A PS -------------------------------------------------- Pig Serum PVP ------------------------------------- Polyvinyl Pyrrolidone RBC ------------------------------- Red Blood Cell rEM -----==— Reduced Eagle's Minimum Essential Medium rEMI ---- —— — — ----------------- rEM containing Inosine RNA-— = -------------- Riboxy Nucleic Acid S-RBC---- —------Sensitized Sheep Red Blood Cell SDS ------------------------------------ Sodium Dodecyl Sulfate SPF ------------------------------------ Specific-Pathogen-Free TBS -------------------------------------- Tris Buffered Saline xix URBC ----------- ———— — -------- Uninfected Red Blood Cell WB _ __ .................... —- ----Wh018 BlOOd WPI --------------------------------------- week Post-Infection INTRODUCTION Swine eperythrozoonosis is caused by Eperythrozoon suis (E. suis), a blood borne rickettsial parasite. The disease was first recognized in the early 1930's in young pigs exhibiting icterus, anemia and depression (Doyle, 1932; Kinsley, 1932). The acute disease is indeed characterized by icteroanemia, however, the majority of infections are subclinical, and recovered animals remain carriers for life. Prior to 'the 1970's, the diagnosis of swine eperythrozoonosis was based on herd and individual animal histories describing icteroanemia, the demonstration of Eperythrozoon bodies in blood smears (parasitemia) and the complement fixation test (Splitter, 1958). A low level of infection was not detected by these diagnostic techniques and was not recognized to be associated with Clinical disease at that time. E. suis was considered primarily a pathogen of feeder pigs causing a disease with a low morbidity and high case mortality (Splitter, 1950b). In the mid-1970's, an indirect hemagglutination assay (IHA) was developed (Smith & Rahn, 1975). This assay was aimed at detecting circulating antibodies induced by E. suis infection. IHA positive, non- parasitemic animals have been observed routinely and were 1 2 considered to be carriers of E. suis. A low level, Chronic infection is now believed to be associated with impaired growth and reproductive performance although these clinical signs have been widely debated. The disease is thought to be widespread in the midwestern United States. At present, IRA is the standard diagnostic test for detecting E. suis infections. A positive IHA titer indicates an increase in circulating IgM antibodies, and has been correlated with the development of anemia but not with development of detectable parasitemia (Zachary & Smith, 1985). Anemia in eperythrozoonosis has been associated with autoimmunity to altered erythrocyte membranes (Hoffmann et al., 1981; Zachary & Smith, 1985) and IgM cold agglutinin plays an important role in the development of the anemia (Schmidt et al., 1992). These results, plus the likelihood that the antigen used in the IRA test is a crude preparation that. contains host, erythrocyte. antigens, suggest. the possibility that the IgM antibodies detected in the IRA may be directed at host erythrocyte antigens as well as E. suis antigens, thus potentially reducing the specificity of IHA. Furthermore, Smith (1981) has reported that IHA negative animals can develop parasitemia after Splenectomy, indicating that the test may also lack sensitivity with respect to identifying subclinical infections and carrier animals. Monetary losses due to swine eperythrozoonosis in the United States are unknown and it is often difficult to determine losses in an individual herd because of the lack of 3 complete understanding of the nature of the disease and definitive diagnostic aids. In the past, inclusion of low levels of relatively inexpensive arsanilic acid in swine diets to control the disease was quite common. However, a change in feed medication regulations in 1986, resulting in a 300-400% increase in medication costs, has heightened producer and veterinarian awareness of swine eperythrozoonosis and related control measures. The response of many producers to the new regulations was to remove arsanilic acid from the diets rather than incur higher medication costs. Subsequently, it appears that there has been an increase in the incidence of swine eperythrozoonosis with the decreased use of arsanilic acid in swine diets. The number of requests for swine eperythrozoonosis serotesting (IRA) and the percentage of positive samples at the Michigan Animal Health Diagnostic Laboratory (ARDL) are depicted in Table 0-1 . Table 0-1. Swine eperythrozoonosis test (indirect hemagglutination assay) performed by the Michigan Animal Health Diagnostic Laboratory during 1983 to 1992. Year 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 No. sera‘ 196 294 327 384 385 385 758 648 181 278 % positive 8.2 19.7 27.2 29.2 25.0 34.8 32.6 36.0 29.3 26.3 . = Number Of sera submitted for swine eperythrozoonosis testing. 4 However, there is considerable controversy over the disease causing role of E. suis infection as well as over the reliability of the IRA for detecting E. suis infection. A crucial missing component in studying swine eperythrozoonosis under field conditions is the lack of specific and sensitive diagnostic tests. CHAPTER 1 LITERATURE REVIEW 6 CLASSIFICATION AND SPECIES OF EPERITHROSOOH According to Bergey's Manual of Systematic Bacteriology, the genus Eperythrozoon (Schilling, 1928) is a member of the family, Anaplasmataceae, in the order, Rickettsiales (Kreier & Ristic, 1984). General characteristics of this genus are: 1) the organisms appear as bluish or pinkish violet rings or cocci, 0.4 to 1.5 pm in diameter, in blood smears stained by Romanovsky-type stains; 2) they occur on erythrocytes and free in plasma with about equal frequency; 3) they are obligate parasites of rodents, ruminants, pigs and other vertebrate species. Eperythrozoon was first reported in mice by Schilling (1928). Since then, Eperythrozoon has been reported in many animal species, and Gothe and Kreier (1977) listed 13 species of Eperythrozoon. In addition, an Eperythrozoon associated with bovine platelets was described in Finland (Tuomi, 1966; Tuomi & von Bonsdorff, 1967). Table 1-1 lists the species of Eperythrozoon reported, literature citations, hosts and Sites of detection. A number of these listings were probably misclassifications, while others were inadequately described in the literature for meaningful evaluation. For example, the description of E. noguchii was based on the appearance of certain objects in photomicrographs published by Noguchi (weinman, 1944). E. wenyoni, E. teganodes and E. tuomii could be observed. subsequently in E. wenyoni experimentally inoculated cows (Dr. Y. Nakamura, personal communication). Table 1-1. Species of Eperythrozoon and their characteristics. '7 Sites of First recorded Species Hosts detection‘ Author (year)T E. coccoides White mouse (Mus RBC", Plasma Schilling (1928) musculus) E. dsper Dwarf mouse (Mus RBC, Plasma Bruynoghe a Vassiliadis minutus), vole (1929)! (Microtus arvall's) E. felist Cat RBC, Plasma Clark (1942) E. Ieptodactyli Frog (Leptodactylus RBC Carini (1930)5, pentadactyius) (intracytoplasmic) Brumpt (1936)! E. man'boi Flying fox (Pleropus Ewers (1971) macrotl's epularius) E. noguchii Human Lwoff & Vaucel (1930)! E. ovis Sheep RBC, Plasma Neits et al. (1934)! E. parwm Pig RBC, Plasma Splitter (19503) E. perekropovi Pike (Esox lucius) Yakimoff (1931 )5 E. suis§ Pig RBC, Plasma Splitter (19503) E. wenyoni Cow RBC, Plasma Adler 8: Ellenbogen (1934)1 E. teganodes Cow Only Plasma Hoyte (1962) E. tuomii Cow Platelet Tuomi (1966), Uilenberg (1967) E. varians Graybacked deer Tyzzer (1942) mouse (Peromyscus manicula tus gracilis) t :- Sites where Eperythrozoon were found. T .-. Author(s) who named species (published year) 1 - Haemobartonella felis may be the same species. § 3 The parasite described in Taiwan as Anaplasma taiwanensis (Sugimoto, 1935) may be the same species. Erythrocyte References are in (Weinman, 1944). 8 Kreier and Ristic (1968) suggested that E. tuomii may reflect a life-stage of E. wenyoni. Only E. coccoides, E. ovis, E. suis, E. parvum, and E. wenyoni are accepted species of Eperythrozoon in the 1984 edition of Bergey's Manual (Kreier a Ristic, 1984). Eperythrozoon like bodies have been reported in other animals including several species of rodents, goats, collared peccaries and recently in llamas (Daddow, 1979; Hannon et al., 1985; Kreier & Ristic, 1968; McLaughlin et al., 1990). CHROIOLOGICAL Rsvrsw or SWINE EDERYTEROSOOIOSIS In 1932, inclusions bodies in red blood cells, called Anaplasma like bodies, were reported from clinically ill, young pigs (Doyle, 1932; Kinsley, 1932). The main clinical and pathological features of the disease associated with these bodies was icterus and anemia. Presenting clinical signs included fever (40 to 41 °C), depression, weakness, dyspnea, and finally icterus and anemia. A scurvy like condition of the skin in the facial region was also observed (Kinsley, 1932). The disease caused a high case mortality rate (death occurred within 1 to 5 days after clinical Signs developed) and low morbidity ((10% in a herd). Microscopic examination of the blood revealed abnormal Size and shape of erythrocytes, increased number of segmented neutrophils and the presence of bodies in erythrocytes and free in the plasma (Eperythrozoon bodies were thought to be located inside erythrocytes at this time.). These bodies exhibited 9 considerable variation in morphology including coccoid, bacilliform and ring forms. Over the next two decades, several authors reported a Similar disease characterized by ictero-anemia and the presence of similarly described bodies in red blood cells of feeder pigs 2— to 8-months old and less frequently in other age groups (Campbell, 1945; Quin, 1938; Robb, 1943; Spencer, 1940). The disease was found in pigs in Illinois, Iowa, South Dakota, Missouri (Kinsley & Ray, 1934) and Kansas (Dicke, 1934). The association between the presence of the bodies in red blood cells and the occurrence of the disease led to naming the disease "anaplasmosis-like disease in swine" (Kinsley, 1939). The acute condition was also known as ”ictero-anemia" or ”yellow-belly of swine” (Splitter, 1951). Various attempts were made to transmit or to reproduce the disease by injecting the bodies into normal pigs, but experimental inoculations were unsuccessful (Kinsley & Ray, 1934; Robb, 1943; Spencer, 1940), and the etiology of the disease and the nature Of the bodies remained unknown. Robb (1943) stated that the bodies in red blood cells were artifacts produced in the process Of smearing, staining or drying, based on his finding of the bodies in affected pigs and also in normal pigs. Alternatively, Doyle (1945) suggested that the disease was more likely bartonellosis than .anaplasmosis. After several decades of controversy, the etiology of 10 the disease was clarified in the 1950's. Splitter and Williamson (1950) found that fever was associated with a high number of the bodies in the blood and that rectal temperatures returned to normal during the period when the bodies disappeared from the peripheral blood. They also described that the bodies were located on the surface of erythrocytes or occasionally free in the plasma, not within erythrocytes as previously reported. Since these features were identical to those of eperythrozoonosis in cattle and Sheep, they designated the disease swine eperythrozoonosis. Subsequently, the disease was successfully reproduced by inoculation of E. suis infected blood into splenectomized pigs or by splenectomy of carrier pigs (Splitter, 1950b). The resulting symptoms, disease process and pathology were identical to those Observed in field cases. Splitter (1950a) Observed 2 types of Eperythrozoon bodies in swine, which were different from each other in Size and pathogenicity. The larger type (0.8-2.5 pm in diameter) was designated as E. suis, the causative agent of ictero- anemia, contained more chromatin and was more pathogenic. The smaller type (0.5 pm in diameter) was designated as E. parvum, and considered relatively non-pathogenic. The failure of these Eperythrozoon bodies to cross-infect calves, sheep and mice, and the failure of E. wenyoni and E. ovis to infect splenectomized pigs provided strong evidence that these bodies were a new Eperythrozoon species. E. parvum was found only in experimentally infected pigs and was not 11 observed under field condition. The presence of E. parvum was also reported in splenectomized pigs in South Africa (Jansen, 1952) and in Great Britain (Jennings & Seamer, 1956; Seamer, 1960). Some of the differentiations of E. suis and E. parvum summarized by Splitter (Splitter, 1953) are shown in Table 1-2. Foote et a1. (1951) hypothesized that a similar icteroanemia condition in swine was caused by a filterable virus, "the virus of Farley". Because of the successful reproduction of this disease syndrome by injecting the filtrates of whole blood, serum or urine from infected swine without accompanying Eperythrozoon bodies, the disease was renamed "virus anemia of swine" and Eperythrozoon bodies were considered to be either reactionary bodies or artifacts, or were coincidental with the disease. Interestingly, the virus caused fever and anemia in a splenectomized calf. However, Splitter (1951; 1952; 1953) demonstrated that swine eperythrozoonosis was always accompanied by the appearance of Eperythrozoon bodies in blood and stated that swine eperythrozoonosis and "virus anemia of swine" are not identical. Other differences of eperythrozoonosis from "virus anemia of swine", such as no reproductivity using the urine of acutely affected pigs and in splenectomized calves, were reported by Splitter (1951; 1952; 1953). Further evidence of successful reproduction of "virus anemia of swine" and the clinical description of a Similar syndrome have not been reported. 12 Table 1-2. Differentiation of Eperythrozoon suis and E. parvum. Morphology Pathogenicity Susceptibility to neoarsphena mine Interference Filterability J. suis Large rings (0.8-1.0 pm) with much chromatin. Discoid forms and very large irregular rings (2-3 um) may be present. Consistently pathogenic in heavy infections in splenectomized or nonsplenectomized pigs. Etiological agent of acute eperythrozoonosis of swine in the field. Highly susceptible to doses of 15 mg/kg body weight. E. suis displaces E. parvum when infection is superimposed. Usually retained by the fine diatomaceous earth filters and the Seitz EK at negative pressures of 100 m m mercury. E. parvum Small rings (0.5-0.8 pm) with numerous coccoid forms of considerably smaller size. Often found in large numbers on single erythrocytes when organisms are rare in blood. Relatively nonpathogenic. May produce mild to severe anemia in some experimental cases. Not known to produce acute eperythrozoonosis in the field. Relatively resistant. Doses of 40- 45 mg/kg body weight are effective irregularly. E. parvum is suppressed in initial mixed infections with E. suis. Readily passes the fin e diatomaceous earth filters and the Seitz EK. Uncomplicated strain may be established by filtration of serum containing E. suis and E. parvum 13 With increasing concern for swine eperythrozoonosis, several review articles and case reports were published in the 1950's and 1960's (Adams et al., 1959; Biberstein et al., 1956; Kingsley & Hibbs, 1968; Preston & Greve, 1965; Stauch, 1951). The disease was reported in 13 states, including California, Georgia, Illinois, Indiana, Iowa, Kansas, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin and New York (Biberstein et al., 1956). The development of a serodiagnostic technique using the indirect hemagglutination assay (Smith & Rahn, 1975) led to the detection of low level of infections without accompanying detectable parasitemia. Since then, swine eperythrozoonosis has been clinically evident in pigs of all ages and chronic infections have been believed to be associated with impaired growth rate and reproductive performance (Henry, 1979). To date, swine eperythrozoonosis has been recognized in North and South America, Africa, Europe and Asia including Korea, Taiwan and Japan. Nonpnonoo! up Dnvntopnnn By light microscopy, Eperythrozoon exhibits polymorphic forms such as coccoid, discoid, rod, comma, ring or chain. It is often difficult or even impossible to differentiate Eperythrozoon bodies from stain deposit, cell fragments or other artifacts in Romanovsky stained preparations. Therefore, it is not clear by light microscopy alone that various shapes of the organism demonstrated are due to 14 polymorphism or artifact. In earlier studies, Kreier and Ristic (Kreier & Ristic, 1963) failed to observe discoid and ring forms with wet preparations in phase contrast microscopy and indicated that the parasite morphology was different from that observed in dry preparations. Later, coccoid, discoid, ring and chain forms were found by scanning and transmission microscopes (Augsten, 1982; Keeton & Jain, 1973; McKee et al., 1973), and parasite polymorphism in shape was clearly established. As demonstrated by electron microscopy, E. suis is enclosed by a single membrane with a diameter 375 to 500 nm. Internal structures include microtubules beneath the cytoplasmic layer; ribosome-like electron-dense granules and short filaments in the cytoplasma; and vacuoles, 110 nm long and 70 nm wide with no apparent internal structures. No distinct cell organelles nor structures resembling a nucleus were identifiable (Pospischil & Hoffmann, 1982). However, RNA and DNA appears to be present in cytoplasma as small granules and filaments, respectively, as indicated by Tanaka et al. (1965). The presence of RNA and DNA in E. coccoides was suggested by Peters and Wigand (1955) and .in Haemobartonella felis by studies using nucleic acid specific fluorescent dye (acridine orange) (Small & Ristic, 1967). The various shapes of the organism appear to be closely related with developmental stages. Gulland et a1. (1987) classified the shapes of E. ovis into 6 forms (spheres, commas, rods, rings, chains and multiple forms) and found a 15 linear relationship between the forms and the degree of parasitemia. The proportion of sphere and rod forms declined as parasitemia increased, whereas the proportion of multiple forms increased as parasitemia increased. Zachary and Basgall (1985) proposed a developmental cycle for E. suis. They hypothesized that E. suis exhibits three developmental stages, immature, juvenile and mature forms, which correspond to coccoid, discoid and ring forms, respectively, which were previously observed by light microscopy. The immature form was between 0.2 to 0.5 pm in diameter, indirectly attached to the erythrocyte membrane surface and initially produced no erythrocyte membrane deformation. As the immature form enlarged, intracellular vacuoles developed and the organism became embedded in a deep cup-like invagination on the erythrocyte surface, resulting in significant erythrocyte membrane deformation. The juvenile form developed from the immature form by dorsal-ventral flattening or by developing umbilicated centers, - possibly as a result of rupturing intracellular vacuoles. The mature form had a thin concave central area with numerous small depressions, irregularly thickened outer edges, and an increased diameter. Budding of the small immature forms appeared to be a major mechanism of multiplication and occurred in all developmental stages but was most commonly observed from the thickened edges of juvenile and mature forms. Bud formation has also been described in E. wenyoni and E. ovis (Ichijo et al., 1982b; Keeton & Jain, 1973). Alternatively, structures indicative 16 of binary fission were observed in E. wenyoni and Haemobartonella muris (Keeton & Jain, 1973; Tanaka et al., 1965). Even though a developmental cycle during the acute parasitemic phase has been postulated, the mechanism of E. suis survival/development in carrier animals without having detectable parasitemia is not clearly understood. One potential hypothesis is that multiplication of the organism takes place in bone marrow. E. ovis and Haemobartonella muris were found on immature erythrocytes (reticulocytes and normochromic erythroblasts) in bone marrow (Ichijo et al., 1982b; Tanaka et al., 1965). In ovine eperythrozoonosis, Ichijo et al. (1982b) observed 35% parasitemia in bone marrow, while the degree of parasitemia in peripheral blood was only 4%, and suggested that bone marrow may be a main site of parasite multiplication. Further investigations are required to assess this hypothesis in swine eperythrozoonosis. Tnnnsurssron Since the discovery of swine eperythrozoonosis, the mechanisms of transmission have been widely debated. For successful experimental transmission with reproduction of the disease, splenectomy of pigs is required (Splitter, 1950b) as attempts at reproducing the disease with intact pigs by injecting parasitemic blood resulted in failure (Kinsley & Ray, 1934; Robb, 1943). Splenectomy is known to induce 17 clinical disease with various hemoparasites including the genus Eperythrozoon (Marmoston, 1935). With splenectomized pigs, E. suis can be transmitted via inoculation of infected blood by parenteral routes (IN, IV) and orally (Williams a McCain, 1982). The minimum infective dose appears to be low. In an experiment with E. ovis, the minimum infective dose was found to be equivalent to one parasitized erythrocyte (Mason & Statham, 1991). Because of the seasonal occurrence of the disease, primarily summer and fall, arthropod vectors appear to play an important role in transmission. Hog lice, mosquitoes, stable flies and mange mites have been suggested as potential vectors (Claxton & Kunesh, 1975; Dicke, 1934; Quin, 1938; Robb, 1943; Smith, 1981; Spencer, 1940; Splitter, 1950b; Williams a McCain, 1982). Dicke (1934) reported that the incidence of the disease was lower in herds with an adequate 'hog lice control program compared to herds with no program. Experimental transmission by various arthropod vectors has been attempted, but no clear understanding has been achieved. Robb (1943) conducted the first attempt to transmit the disease using hog lice (Haematopinus suiS). A clinically normal, intact pig was housed with a clinically ill pig that was heavily infested with hog lice. The contact pig exhibited some evidence of depression and fever 2 days after contact, but apparently the clinical signs observed in this pig were more likely due to the sudden heavy infestation of lice rather than E. suis infection. One reason for 18 failure of transmission in this experiment may have been that an intact, non-splenectomized pig was used. The importance of mosquitoes (Culex pipiens and Aedes sp.) and stable flies (Stomoxys calcitrans) in transmission has been suggested by the correlation of the occurrence of blood feeding diptera and E. suis infections in swine herds (Prullage, 1989). Subsequently, direct mechanical transmission via mosquitoes (Aedes aegypti) and stable flies was demonstrated (Prullage, 1989; Williams a McCain, 1982). Ingestion of the organism by lmost blood-sucking arthropods, including hog lice, has been -observed using a fluorescent antibody assay to detect E. suis in the arthropods (Prullage, 1989). Experimental transmission via intramuscular injection of a ground debris of lice which infested parasitemic pigs for 7 days was demonstrated; however, direct transfer of infested lice did not transmit the disease in this study (Dr. R. E. Williams, personal communication; Williams & McCain, 1982). The failure to transmit E. suis by direct transfer of infested lice may be partly because the 7 day feeding period may not have been long enough for E. suis to completely develop. For example, Anaplasma, another rickettsial organism, has been shown to develop slowly in ticks, requiring at least 6 to 9 days before the organisms appeared in the salivary glands for subsequent transmission (Kocan et al., 1992). Arthropods demonstrated to transmit Eperythrozoon species are listed in Table 1-3. Most of the successful 19 Table 1-3. Eperythrozoon and their possible arthropod vectors. Eperythrozoon Vector species References E. suis mosquito (Culex pipiens) Williams 81 McCain (1982) stable fly (Stomoxys calcitrans) Prullage (1989) E. parvum hog louse (Haematopinus suis) Jansen (1952) Seamer (1960) E. ovis tick (Haemaphysalis plumbeum and Nikolskii & Splipchenko (1969) Rhipicephalus bursa) sheep ked (Melophagus ovinus) Avakyan et al. (1973), Overas (1969) mosquito (Aedes camptorhynchus) Howard (1975) mosquito (Culex annulirostn's) Daddow (1980) stable fly (Stomoxys calcitrans) Overés (1969) E. coccoides louse (Polyplax serrata) Eliot (1936), Berkenkamp (1988) 20 transmissions were probably due to mechanical transmission except possibly for transmission of .E. parvum 1x) splenectomized pigs, which was achieved by transferring hog lice after 15 days of infestation on an infected pig and by oral inoculation of ground hog lice (Jansen, 1952; Seamer, 1960). There are no observations available of E. suis development in any arthropod vectors. Strictly speaking, no species of arthropods has been experimentally demonstrated to be a biological vector in Eperythrozoon transmission. Mechanical transmission via needles, ear notching, tail docking and other routine surgical procedures may also play a role in transmission. The occurrence of acute disease has been often observed in conjunction with hog cholera vaccination (Kinsley, 1939; Quin, 1938; Spencer, 1940; Splitter & Williamson, 1950) and with castration (Stauch, 1951). The possibile transmission of E. suis via contaminated hog cholera vaccine was rejected because phenolized E. suis could not be transmitted (Splitter, 1950b). However, the observation that the disease was not reported following vaccination or other surgical procedures during the winter (Splitter, 1950b) suggested that arthropod vectors were more important in E. suis transmission. In utero transmission may be another mode of transmission. The death of new born pigs (24 to 48 hours of age) from severe anemia with E. suis infection has been observed by Berrier and Gouge (1954). Preston and Greve (1965) detected the disease in 4-week-old pigs from a louse 21 free farm where flies and mosquitoes were not present in significant number and concluded that in utero transmission had occured. Claxton and Kunesh (1975) reported that in utero transmission was demonstrated by Dr. A. R. Smith (University of Illinois), who observed that new born pigs delivered via hysterotomy from a known E. suis carrier sow developed parasitemia one week after birth. These pigs had no contact with the sow or other infected pigs. These findings suggest the possibility of in utero transmission; however, detection of E. suis in fetal pig blood or antibody production against E. suis in fetal pigs has not been experimentally demonstrated. Duncan. Stars m Parsonos! Swine eperythrozoonosis can be generally classified into three types (acute, subacute and chronic) based on the time course after infection. The acute type was characterized by a rapid onset of clinical signs followed by death one to three days later (Kinsley & Ray, 1934; Quin, 1938; Robb, 1943; Spencer, 1940). The subacute type lasted 4 to 7 days and was usually followed by recovery (Kinsley & Ray, 1934; Robb, 1943; Spencer, 1940). The chronic type represented the majority of infections, was subclinical and affected animals became carriers for life (Smith, 1981). All three types have been recognized in pigs of all ages; however, Henry (1979) reported that pigs less than 5-days-old are the most likely group within a herd to have clinical signs. Common 22 clinical signs/syndromes recognized in different age groups included: 1) acute or subacute disease as evident by anemia, mild icterus and weakness in newborn pigs, 2) Chronic disease associated with impaired growth in finishing pigs; 3) acute or subacute disease with the classic icteroanemia in stressed feeder pigs; 4) chronic disease associated with impaired reproductive performance in gilts and sows. Acute and subacute diseases were generally characterized by icterus and anemia (Kinsley, 1932; Splitter, 1950b; Splitter & Williamson, 1950), in conjunction with many of the other clinical signs. Affected animals exhibited fever (Doyle, 1945; Kinsley, 1932; Smith, 1981), depression, weakness, dyspnea (Doyle, 1932), constipation (Robb, 1943), gait abnormality (Campbell, 1945) and even death (Kinsley & Ray, 1934; Robb, 1943; Spencer, 1940). Purpura, cyanosis and gangrenous necrosis of extremities associated with autoantibody production were also described (Hoffmann et al., 1981; Kinsley & Ray, 1934). Hemoglobinuria, a characteristic of autoimmune hemolytic anemia, has not been observed in swine eperythrozoonosis (Smith, 1977; Splitter, 1950b). There is considerable controversy on this issue in ovine eperythrozoonosis. Ovine eperythrozoonosis has been described as a disease not associated with hemoglobinuria because of the extravascular destruction of erythrocytes 23 (Blood et al., 1983; Sheriff, 1978; Valli, 1985). However, hemoglobinuria in ovine eperythrozoonosis has been reported (Ichijo et al., 1982a; Inada et al., 1985; Mason et al., 1981; Neitz, 1937; Sonoda et al., 1977) and intravascular hemolysis was observed in the kidneys (Sutton, 1979b). Intravascular hemolysis has not been described in swine eperythrozoonosis. At present, there are not enough data available to evaluate this issue. Clinical signs of acute disease in sows include anorexia, fever, occasional mammary gland and vulva edema, icterus, reduced milk production and abnormal maternal behavior. Acute episodes at weaning apparently resulted in poor conception at the first return-to-estrus after weaning (Brownback, 1981). The occurrence of acute disease in sows is usually associated with stressful conditions such as movement into the farrowing house or regrouping for breeding. Recovered animals continue with a chronic infection and may exhibit general icterus, no or delayed return to estrus, poor conception, early embryonic death, late-gestation abortion and small litter size. Coincidental respiratory and enteric disease, even death from secondary infections, has been noted in chronically affected sows (Henry, 1979). zinn et al. (1983) classified sows by their indirect hemagglutination assay (IHA) titer and compared the reproductive performance between sows with IHA titer (1:40 versus 21:40, and also between sows with IHA titer <1:80 versus 21:80. Litters from sows with IHA titer 21:40 or 21:80 exhibited significantly 24 lower birth weights, increased numbers of pigs born dead, lower packed cell volumes and lower hemoglobin values, compared to litters from the sows with IHA titer <1:40 or <1:80. The associations between IHA titer and other reproductive failures, delayed return-to-estrus, early embryonic death or late-gestation abortion, were not observed in this study. However, the ability of IHA in detecting true E. suis infections is controversial; thus true associations between E. suis infection and specific antibody levels, and various reproductive parameters as investigated in this study must also be questioned. Blood collected from affected animals showed spontaneous agglutination and elevated sedimentation rate (Doyle, 1932; Hoffmann et al., 1981; Robb, 1943). Analysis of thin smears revealed varied size and shape of erythrocytes and erythrophagocytosis (Doyle, 1932). In addition, hyperglobulinemia and hypoglycemia have been observed (Heinritzi, 1989). Postmortem examination of acutely affected pigs revealed a general icterus, thin and pale blood, hemorrhages in lung, heart and kidney, enlarged and soft spleen, engorged and yellowish liver (Doyle, 1932; Kinsley, 1932), hydrothorax, hydropericardium, ascites, bile staining of stomach and intestinal contents and mucosa (Robb, 1943), swollen, edematous lymph nodes and pulmonary edema (Quin, 1938; Spencer, 1940). Microscopic examination of spleen and liver revealed hyperplasia of the reticuloendothelial cells, 25 hemosiderosis and congestion in the spleen, central atrophy of liver lobules, hepatic erythrophagocytosis and hemosiderosis, lymphocyte and macrophage infiltration in liver, and albuminous degeneration of the hepatic cells (Kingsley & Hibbs, 1968; Quin, 1938; Robb, 1943; Spencer, 1940; Splitter, 1950b). Cow Aocrunnx Production of cold agglutinins to erythrocytes appears an important characteristic in Eperythrozoon affected animals. Since discovery of Eperythrozoon, spontaneous agglutination of blood taken from animals with acute eperythrozoonosis, especially at cold temperature, has been observed repeatedly. Although it has been postulated that Eperythrozoon bodies themselves could bridge erythrocytes and induce hemagglutination (Iralu & Ganong, 1983), anti- erythrocytic antibodies are a more likely cause of this phenomenon (Kreier & Ristic, 1968). Antibodies directed against erythrocytes were demonstrated in E. coccoides infected mice and .E. suis infected pigs by: 1) hemagglutination of noninfected red cells by incubation with immune serum; 2) hemolysis of noninfected red cells by ‘incubation with immune serum and complement; 3) the presence of immunoglobulin on red cells taken from infected animals (direct Coombs test); and 4) the detection of immunoglobulin on noninfected red cells after incubation of noninfected red cells and immune serum (Cox & Calaf-Iturri, 1976; Hoffmann et 26 al., 1981). Schmidt et al. (1992) isolated a cold agglutinin from E. suis infected pigs and demonstrated temperature dependent reactivity. Further biochemical analysis indicated that the isolated cold agglutinin consisted of IgM exclusively. Detection of cold agglutinins has been“ reported in .Mycoplasma pneumoniae infection, mononucleosis (Pruzanski & Shumak, 1977a; Pruzanski & Shumak, 1977b) and many other diseases including hemoparasitic diseases such as haemobartonellosis, anaplasmosis, malaria and babesiosis which are associated with anemia and splenomegaly (Bellamy et al., 1978; Oki & Miura, 1970; Schroeder & Ristic, 1965; Soni & Cox, 1974; Soni & Cox, 1975; Thoongsuwan & Cox, 1973; Zulty & Kociba, 1990). Similar to eperythrozoonosis, the majority of cold agglutinins observed in other diseases are IgM and directed at glycoproteins (I antigen in human) of the erythrocyte membrane and possibly similar antigens located on lymphocytes, monocytes, macrophages and normal tissue cells (Lau & Rosse, 1975; Pruzanski & Shumak, 1977a; Pruzanski & Shumak, 1977b). In haemobartonellosis and eperythrozoonosis, alteration of the erythrocyte surface membrane by neuraminidase or glutaraldehyde treatment was required to induce hemagglutination of noninfected red cells, indicating that the cold agglutinins were directed at cryptic or hidden antigens of the erythrocyte membrane which are exposed to the immune system by parasite infection (Zachary & Smith, 1985; Zulty & Kociba, 1990). Interestingly, Eperythrozoon and 27 Anaplasma infection appear to induce new glycoproteins in erythrocyte membranes. Nordelo and Ysern-Caldentey (1982) observed new glycoproteins in A. marginale infected erythrocyte membranes by SDS-PAGE analysis. Gdff et al. (1986) observed that the agglutinating ability of erythrocytes via lectins increased with E. wenyoni or A. marginale infections. Lectins are saccharide binding proteins and it was suggested that a new glycoprotein or glycolipid was synthesized or exposed to the erythrocyte surface by E. wenyoni or A. marginale infection. Elucidating an association of these new glycoproteins and production of cold agglutinin may provide information on the mechanism of cold agglutinin induction by Eperythrozoon or Anaplasma infection . thaopnrsronos! The anemia observed in swine eperythrozoonosis has been classified as an acquired autoimmune hemolytic anemia (Hoffmann et al., 1981). Autoantibodies including cold agglutinins have been suggested to act as opsonins in the process of erythrophagocytosis or to activate complement dependent lysis of red cells (Cox & Calaf—Iturri, 1976; Zachary & Smith, .1985). Deformation of erythrocyte surface membrane may be a primary initiator of the autoimmunity (cold agglutinin production) induced by E. suis infection. Even though direct contact of E. suis to erythrocyte membranes was not observed 28 and E. suis and erythrocytes were separated by a distance of 30 nm (Pospischil & Hoffmann, 1982), a thread like structure located between the organisms and erythrocyte membrane appears to be an association tool of the organisms for erythrocytes (Ichijo et al., 1982b). The area of the erythrocyte membrane associated with the parasite has been observed to be more electron dense than non-parasitized areas. Furthermore, membrane deformation persisted after parasites disappeared following tetracycline treatment. These observations suggest an alteration of erythrocyte cytoskelton proteins or an alteration of the relationship of structural proteins with integral proteins of the inner membrane surface (Zachary & Basgall, 1985). The degree of association of Eperythrozoon to the erythrocyte membrane may explain variations in the degree of pathogenicity with respect to anemia induction. For example, it was observed that E. wenyoni which generally induces no or slight anemia in cows caused only slight depression of erythrocyte membranes and did not cause electron density alteration of the erythrocyte membrane (Keeton & Jain, 1973). Cyanosis of the skin, especially in the extremities, has been occasionally observed in affected pigs, presumably the result of cold agglutinin (autoantibody) production. The cold agglutinins induced by E. suis infection cause agglutination, restricted circulation and necrosis in the skin because the skin temperature is lower than the core body temperature (Heinritzi et al., 1990a). 29 Little is known about the pathophysiology of impaired growth resulting in "delayed marketing syndrome". Eperythrozoonosis is considered to be associated with stress, especially social stress, as is seen in certain protozoal diseases in which the host parasite relationship is guided by premunition. However, Nicholls et al. (1989) failed to show stress induced by feed or exercise restriction affected the growth rate of E. ovis infected sheep. Hypoglycemia, a common feature in acutely affected pigs, may partially explain the impaired growth in chronically infected finishing pigs and also the weakness observed in acutely affected baby pigs. Investigation of metabolic changes in experimentally infected, splenectomized animals found decreased concentration of serum glucose, increased concentration of blood pyruvate and lactate, and decreased blood pH (Heinritzi et al., 1990b; Heinritzi et al., 1990c), indicating activation of glycolysis. A decreased blood glucose concentration (up to 25% at 23 days post infection) was also observed in experimentally infected, intact pigs (Heinritzi et al., 1990b). Similar observations were reported in E. ovis infected sheep (Ilemobade & Blotkamp, 1978b; Sutton, 1976; Sutton, 1977). These observations suggested that Eperythrozoon themselves have a high glycolytic activity or Eperythrozoon may have a profound effect on erythrocyte glycolytic activity, despite their semi-extracellular location. High glucose utilization, either as a result of parasite metabolism or enhanced 30 erythrocyte metabolism caused by the parasite, may cause generalized hypoglycemia (Wensing et al., 1974) and affect growth rates. This seems especially important in piglets which are known to have a precarious energy balance at birth (English a Smith, 1975) and may partly explain the mechanisms of the perinatal problems (stillbirth, weak pigs) observed in swine eperythrozoonosis. Impaired erythrocyte function could be postulated as a mechanism for impaired growth. Lactate production as a result of massive glycolysis leads to metabolic acidosis, which is usually combined with respiratory acidosis caused by impaired pulmonary gas exchange (Heinritzi et al., 1990b). Low blood pH leads to shifts in the oxyhemoglobin curve and, subsequently, the delivery of oxygen to peripheral tissue may be impaired. In ovine eperythrozoonosis, reduced glutathione levels were observed in E. ovis infected sheep (Sutton, 1979a). Glutathione is an essential component of erythrocytes for removing peroxide which is produced as methemoglobin is formed. Therefore, the effect of E. ovis on glutathione concentration might interfere to some extent with erythrocyte function. Hypoglycemia may also be related with some aspects of impaired reproductive performance. For example, it has been reported that insulin regulates ovulation rate (Matamoros et al., 1991). Insulin administration affects follicular dynamics by reducing follicular atresia and increasing the number of potential ovulatory follicles. It could be 31 expected that hypoglycemia induced by E. suis infection may lead to low endogenous insulin levels. Perhaps, the impaired reproduction observed in swine eperythrozoonosis, such as delayed puberty in gilts and delayed return-to-estrus in sows, may be explained by low insulin levels due to chronic hypoglycemia. At present, insulin levels in E. suis infected sows nor the pathophysiology of swine eperythrozoonosis in reproducing sows has not been studied although such studies appear to be warranted. Hos:- Ixxunn'! As observed with other hemoparasitic diseases such as malaria and haemobartonellosis (Maede, 1978; weiss et al., 1986), the reticuloendothelial system plays an important role in eliminating Eperythrozoon species from the circulation. Splenomegaly is an important characteristic of Eperythrozoon infections and E. coccoides has been called the ”Spleen Weight Increasing Factor" (Stansly, 1965). Splenectomy induces severe parasitemia both in primary infected animals and in carrier state animals (Marmoston, 1935; Rouse & Johnson, 1966; Splitter, 1950b). Therefore, the spleen plays a critical role in the sequestration and elimination of the Eperythrozoon organisms (Baker et al., 1971). Detailed studies of splenic function have been conducted using electron microscopy in swine and ovine eperythrozoonosis. Reported observations included: 1) cordal reticular cells and macrophages phagocytized both parasitized 32 and non-parasitized red cells; 2) pseudopodia extending from the reticular cells and macrophages appeared to engulf and remove Eperythrozoon organisms without destroying red cells; 3) parasitized red cells did not pass through basement membrane fenestrations and were likely to remain in the cord, thus facilitating phagocytosis in the cord (Hung & Lloyd, 1989; Pospischil & Hoffmann, 1982). Removal of organisms from erythrocytes by extended pseudopodia of reticular cells and trapping of organisms by pseudopodia at the basement membrane fenestrations were also reported in feline haemobartonellosis (Maede, 1979). Maede (1978) observed that E. felis infected red cells labeled with.5LCr were highly sequestered in the spleen of intact animals, and in the liver, lung and bone marrow of splenectomized animals. These findings indicated that the spleen is a major organ for eliminating the organism from circulation, and liver, lung and bone marrow might also contribute to some extent (Baker et al., 1971; Maede, 1979). Mononuclear phagocytic activity has been noted to be markedly increased during infection with Eperythrozoon and Haemobartonella species as measured by carbon clearance from blood (Elko & Cantrell, 1968; Gledhill et al., 1965). Mice inoculated with E. coccoides or rats inoculated with H. muris and subsequently challenged with mouse hepatitis virus (Gledhill & Dick, 1955), lactate dehydrogenase virus (Riley, 1964; Riley et al., 1964), and lymphocytié choriomeningitis virus (Seamer et al., 1961) exhibited more severe clinical 33 signs and lesions than animals inoculated with virus alone. Baker et al. (1971) suggested that rapid replication of macrophages and increased phagocytosis due to hemotropic parasitic infection made it easier for viruses to enter and replicate in cells. Interferon responses in the host may contribute to these observations. E. coccoides infection suppressed further interferon production in mice in response to viral infection (Glasgow et al., 1974; Glasgow et al., 1971). Conversely, E. coccoides infection by itself induced low levels of interferon production (Suntharasamai & Rytel, 1973). This alternate finding may partially explain observations that E. coccoides infected mice resisted superinfection with Plasmodium berghei (Peters, 1965), P. chabaudi (Ott et al., 1967; Voller & Bidwell, 1968) and P. vinckei (Cox, 1966), possibly due to the antiparasitic effect of interferon. The importance of protective antibodies in host defense mechanisms in swine eperythrozoonosis is not clearly understood although it has been suggested by studies of murine and ovine eperythrozoonosis. Observations which support the presence of protective antibodies include: 1) infected animals were resistant to reinfection; 2) clearance of parasitemia was correlated with the appearance of circulating antibodies; 3) transfer of immune serum to parasitemic animals resulted in the clearance of organisms; and 4) inoculation of Eperythrozoon bodies after incubation with immune serum (neutralization) resulted in a prolonged 34 prepatent period and reduction of parasitemia (Glasgow et al., 1974; Hung & Lloyd, 1985; Hyde et al., 1972; Thurston, 1955). However, Hung and Lloyd (1985) failed to demonstrate an opsonin effect of antibodies~ in an in *mitro erythrophagocytosis test with E. ovis infected erythrocytes. Although infected animals become resistant to reinfection, recurring parasitemia after the initial clearance of circulating parasites has been frequently observed (Hung & Lloyd, 1985; Thurston, 1955). Recurring parasitemia has been observed in animals infected with hemoparasites such as Babesia and Eperythrozoon, and was proposed to be associated with new antigenic variants (Hung & Lloyd, 1985; Rogers, 1974; Thoongsuwan & Cox, 1973). However, detailed studies on antigenic variants of Eperythrozoon species have not been reported. Passive immunity via colostrum was observed in sheep with E. ovis and as expected, transplacental transfer of passive antibodies could not be demonstrated in sheep (Daddow, 1982). A study of lymphocyte function in eperythrozoonosis was conducted by Zachary and Smith (1985). Using a lymphocyte blastogenesis assay, they found polyclonal T cell suppression and polyclonal B cell activation during E. suis infection. These findings may explain the often reported increased susceptibility of Eperythrozoon infected pigs to other infectious diseases. Induction of parasitemia using immunosuppressive agents 35 is a controversial issue. Immunosuppression after administration of cortisone was generally ineffective for inducing detectable parasitemia as observed in H. muris (Laskowski et al., 1954; Scheff et al., 1956) and E. coccoides infected animals (Thurston, 1955). However, some modification of infections by corticosteroids has been reported. Daddow (1977) used betamethasone in E. ovis infected sheep to hasten the development of parasitemia and Martin et al. (Martin et al., 1988) found that dexamethasone treatment enhanced the severity of clinical disease in sheep. Antilymphocyte serum and cyclophosphamide were ineffective for inducing parasitemia of H. muris and E. coccoides (Baker et al., 1971). Conversely, induction of parasitemia in latently infected animals using anti-splenic agents was successful. Successful induction of parasitemia has been reported with: 1) anti-rat spleen serum (Pomerat et al., 1947; Thomas et al., 1949); 2) ethylpalmitate, which depresses reticuloendothelial activity (Finch et al., 1968); and 3) polonium (21°Po), which selectively injures the spleen (Scott & Stannard, 1954). In summary, splenic function appears to be the core defense mechanism against E. suis infection. Activation of the reticuloendothelial system appears to be a major component of eliminating E. suis from the circulation. However, the role of humoral immunity in E. suis elimination is not well understood, although protective antibodies appear to be produced following infection. 36 Drncnosrs Prior to 1970, the diagnosis of swine eperythrozoonosis was based on herd and individual animal histories describing ictero-anemia, the demonstration of Eperythrozoon bodies in blood smears and the complement fixation serotest (Splitter, 1958). Detection of the organism provides definite diagnosis; however, the majority of E. suis infections are subclinical and parasitemia can be detected only in acutely affected animals. Moreover, artifacts in blood smears are often difficult to distinguish from E. suis bodies in many cases (Henry, 1979). A low level, chronic infection could not be detected by these classical techniques. In the mid 1970's, Smith and Rahn (1975) developed an indirect hemagglutination assay (IHA). This assay was aimed at detecting antibodies induced by E. suis infection. IHA positive, non-parasitemic pigs have been observed routinely and are considered to be carriers of E. suis. The IHA test involves glutaraldehyde attachment of a partially purified Eperythrozoon antigen, prepared from the plasma of E. suis infected blood, to sheep erythrocytes. Serial serum dilutions are reacted with the treated sheep erythrocytes to determine the serum titer. A positive IHA titer, associated 'with IgM antibodies to E. suis, is correlated with anemia but not with detectable parasitemia, and indicates either a recent or persistent/chronic infection (Zachary & Smith, 1985). The IHA test is widely used for diagnosis although 37 clinical interpretations are often confusing. The time course of IHA titer development and decline is poorly understood. As the test measures IgM antibodies, a transient titer rise would be expected from 2-6 weeks following exposure. Experimentally, this transient rise has been observed. However, persisting high titers are frequently observed in naturally infected pigs as well as splenectomized pigs (Baljer et al., 1989). These results, along with the facts that the antigen used in IHA test is a crude preparation that may include host antigens, and that E. suis infection induces autoantibody production (Hoffmann et al., 1981) suggests the possibility that IgM antibodies measured in the IHA test may be directed at host erythrocyte antigens, thus potentially reducing the test's specificity; Furthermore, Smith (1981) reported that infected animals which are serologically negative by the IHA test can be carriers because they developed a parasitemia after splenectomy, thus indicating that the test may lack sensitivity with respect to identifying carrier pigs. Various other serological techniques have been proposed for Eperythrozoon spp. antibody detection. Kreier and Ristic (1963) developed an indirect fluorescent test using intact Eperythrozoon bodies as the antigen for ovine and bovine eperythrozoonosis. Subsequently, antibodies .against Eperythrozoon bodies were detected by indirect fluorescent tests in mice (Baker et al., 1971) and sheep (Ilemobade & Blotkamp, 1978a; Nicholls & veale, 1986). In addition, 38 complement-fixation tests for murine (Wigand, 1956), ovine (Daddow, 1977) and swine eperythrozoonosis (Splitter, 1958), an indirect hemagglutination assay for bovine eperythrozoonosis (Finerty et al., 1969), and enzyme linked immunosorbent assays (ELISA) for ovine (Lang et al., 1987), bovine (Kawazu et al., 1990) and swine eperythrozoonosis (Hsu et al., 1992; Schuller et al., 1990) have been developed using partially purified antigens prepared from erythrocyte lysates. Considering autoantibody production iJI Eperythrozoon infection, a major concern with these immunoassays is the inclusion of erythrocyte antigens in the antigen preparations. Lang et al. (1986) described the successful separation of E. ovis bodies from hemolysates. Following this method, Hall et al. (1988) failed to completely separate E. suis bodies from erythrocyte antigens by lysing erythrocytes. Alternatively, they used di-sodium ethylenediamine tetraacetic acid (EDTA) to release the organisms from the erythrocytes. They separated the free organisms from the erythrocytes by low speed centrifugation. The organisms were then concentrated by high speed centrifugation yielding a high concentration of E. suis bodies. Recently, E. suis DNA detection techniques have been developed using whole organism DNA hybridization (Oberst et al., 1990b) and recombinant DNA hybridization (Oberst et al., 1990a). These techniques detected E. suis specific DNA from the high salt lysate of whole blood samples, and apparently 39 offer high sensitivity and specificity for detecting E. suis infection. These techniques could potentially lead to the development of a more powerful diagnostic technique using the DNA polymerase chain reaction (PCR), in which very low level of infection could be detected by amplification of E. suis specific DNA. Tmmnm m Conmxon Laboratory and field studies have demonstrated the beneficial effects of treatment with tetracyclines and/or arsenicals (Henry, 1979; Rosenkrans et al., 1984; Splitter, 1950c; Splitter & Castro, 1957). Individual drugs tested for activity against Eperythrozoon species are listed by Gothe and Kreier (1977). Table 1-4 shows drugs tested for activity against swine eperythrozoonosis. Some of these medications have been Shown to reduce the clinical signs of affected pigs; however, no drug has been shown to completely clear infection. Although no drug has been approved for the treatment and control of swine eperythrozoonosis at present in the United States, inclusion of low levels of arsanilic acid in swine diets appeared to effectively control the disease (Smith, 1981). 40 Table 1-4. A list of drugs tested for activity against Eperythrozoon suis Drug Dose Action Reference Arsanilic acid 90 g/ton in Prophylaxis (Smith, 1981 ) feed Chlortetracycline 45 g/ton in Prophylaxis (Anthony et al., 1962) feed ‘ Chlortetracycline 200 mg/gal in Prophylaxis (Anthony et al., 1962) water Neoarsphenamine 45 mg/kg IV Cleared parasites; (Splitter, 1950c) relapses occured Oxytetracycline 3 mg/lb IM Cleared parasites; (Splitter & Castro, 1957) relapses occured Sodium cacodylate 0.3 mg/lb IV No effect (Splitter, 1950a) Tetracycline 3 mg/Ib lM Cleared parasites; (Splitter 8: Castro, 1957) relapses occured m2 IN VI TRO CULTIVATION OF EPERYTHROZOON SUIS 41 42 Abstract In vitro culture of Eperythrozoon suis was attempted using a Petri dish-erythrocyte culture system. In preliminary experiments, the following conditions were found to be optimal for maintaining E. suis attachment to erythrocytes in cultutre: anticoagulation with heparin or citrate solution, incubation with 5 or 10% CO; at 37°C, and incubation with reduced or non-reduced Eagle's minimum essential medium. Using heparin, a C02 incubator and reduced Eagle's medium (rEM), E. suis metabolic activity was evaluated by measuring glucose consumption, and lactate and pyruvate production. Glucose consumption and lactate production were apparent while pyruvate production was not detected. Erythrocyte integrity was improved by the addition of inosine, an energy source of swine erythrocytes, although no effect was observed on maintenance of E. suis attachment to erythrocytes or the rate of glucose consumption. To determine whether the observed glucose consumption in culture was due to E. suis glycolytic activity or enhanced erythrocyte glycolytic activity, the effect of E. suis killing by EDTA addition to medium was evaluated using rEM containing inosine (rEMI). Glucose consumption decreased proportionally with the decline in the percentage of parasitized erythrocytes induced by EDTA, indicating glucose consumption was due to E. suis. In a subsequent experiment, 43 the effect of different types of serum (pig or fetal calf serum) and different gaseous environments (5% C02 incubator or candle jar) were evaluated using rEMI. Glucose consumption by E. suis was significantly increased by the addition of fetal calf serum; however, no difference in the maintenance of E. suis attachment to erythrocytes and in E. suis glycolytic activity was observed with a 5% C02 incubator and candle jar. Finally, the effect of medium refreshment (rEMI containing fetal calf serum) was evaluated. Maintenance of E. suis parasitism on erythrocytes and E. suis glycolytic activity were significantly improved by frequent medium refreshment. The developed culture system could be used for metabolic radiolabeling of E. suis for further protein/antigen analysis. 44 Introduction Swine eperythrozoonosis is caused by the rickettsial organism, Eperythrozoon suis. The disease is characterized by acute ictero-anemia and is associated with impaired weight gains and reproductive performance (Henry, 1979) . Swine eperythrozoonosis has been reported world wide including North and South America, Africa, Europe and Asia (Henry, 1979; Hsu et al., 1990; Sisk et al., 1980; Smith, 1977; Smith, 1981). Swine eperythrozoonosisis poorly understood in part because of the lack of an E. suis in vitro culture system. Organism propagation is typically done by inoculating splenectomized pigs with infected whole blood and then collecting blood from the inoculated pigs as parasitemia develops. The whole blood is used immediately or frozen for future use. The lack of an in vitro culture system is a critical barrier preventing systematic serological, immunological and molecular-biological studies of E. suis. In comparison, a successful in vitro culture system established for Plasmodium falciparum (Trager & Jensen, 1976), has contributed significantly to understanding the disease by facilitating studies on antiparasitic compounds (Geary, 1983), and parasite ultrastructure, development, biochemistry, immunology and serology (Jensen, 1983). The purpose of this study was to develop an in vitro culture system for E. suis. In vitro culture could provide a 45 uniform method for propagating E. suis in a controlled environment, decrease the need for animal experimentation in .E. suis investigations and facilitate radiolabeling techniques for antigen analysis. 46 Materials and Methods EXPERIMENTAL INFECTIONS Organism. E. suis was initially isolated from a parasitemic, febrile boar housed at the Swine Research Center, Michigan State University, and has been maintained by passage through splenectomized pigs. Infected blood also was cryopreserved by adding 6.7% and 10% (v/v) of 40% polyvinyl pyrrolidone (PVP) and glycerol, respectively, and then freezing in liquid nitrogen. .Animals and inoculation of.E. suis. crossbred pigs were 6- to 13-weeks-old at the time of inoculation. All animals were splenectomized using standard surgical techniques. Ten milliliters of frozen, infected blood was thawed in a 37 °C water bath and inoculated intravenously aJHi/or intramuscularly into the splenectomized pigs. Clinical signs and rectal temperature were assessed daily to monitor the development of clinical disease. When the rectal temperature exceeded 40 °C, blood was collected from the jugular vein in a vacutainer tube (Beckton Dickinson vacutainer Systems, Rutherford, NJ) containing ethylenediamine tetraacetic acid (EDTA) and examined for parasitemia by Wright's stained thin smear. At the height of parasitemia (>75% parasitemia), blood was collected from the anterior vena cava with either sodium heparin (30 U/ml), 0.15% EDTA, 0.38% sodium citrate or 47 glass beads. Blood collected with glass beads was gently agitated until fibrin formation was observed on the beads. At the same time, blood was collected from noninfected pigs which were housed in a separate building. When the febrile phase lasted for more than 3 days or if the animals exhibited severe depression, intramuscular treatment with oxytetracycline (20 mg/kg body weight; Liquamycin LA 200, Pfizer, New York, NY) was initiated to avoid mortality. Cuummr Media. Several media were used and are described as follows: 1) RPMI 1640 (pH 7.4; GIBCO Laboratories, Grand Island, NY) was supplemented with 25 mM N-Z-hydroxy-ethylpiperazine- N'-2-ethanesulfonic acid (HEPES), 0.2% NaHC03 and 1% penicillin-streptomycin solution (Sigma Chemical Co., St. Louis, MO). 2) Mycoplasma medium (MM; pH 7.8) was prepared with PPLO broth (1.5% w/v; Difco), 2.5% yeast extract, 0.1% glucose, 1% penicillin G (1,000 U/ml), 0.2% Thallium acetate and 0.002% phenol red. 3) Eagle's minimum essential medium (EM; pH 7.3) was prepared from powder containing Eagle's medium supplements, non-essential amino acids, Earle's salts and L-glutamine (Sigma Chemical Co.) supplemented with 25 mM HEPES, 0.1% 48 glucose and 0.2% NaHC03, in 1100 ml of distilled water. Reduced EM (rEM) was prepared by further addition of 0.2 mM B-mercaptoethanol, 2.0 mM sodium pyruvate, 0.1 mM hypoxanthine and 0.016 mM thymidine (Brun & Jenni, 1987). Inosine containing rEM (rEMI) was prepared by adding 0.2% inosine (Sigma Chemical Co.) to rEM._ EDTA containing rEMI (rEMI-EDTA) was prepared by adding 0.2% EDTA to rEMI. Incubators. Several incubators were used and are described as follows: 1) Candle jar (Jensen & Trager, 19T7). Briefly, a lighted paraffin candle was placed in a desiccator equipped with a stopcock on the lid. The lid was sealed with silicone grease with the stopcock open to vent the heated air from within the desiccator. When the candle extinguished, the stopcock was closed and the candle jar was placed in a 37 °C incubator. 2) Gas chamber. Carbon dioxide gas was introduced to the gas chamber from one of two stopcocks of the chamber with the other stopcock open. After 5 minutes of introducing the gas, both stopcocks were closed and the gas chamber was placed in a 37 °C incubator. 3) 5% and 10% C02, humidified incubators (Queue C02 incubator, Perkinsburg, WV) and an anaerobic incubator (Forma Scientific, Marietta, OH) were used at 37’°C. 49 Procedures. In the whole blood (WB) culture procedure, approximately 10% parasitemic blood was prepared by mixing infected (>75% parasitemic) and noninfected blood. The mixed WB (1.2 ml) was placed in 35 mm Petri dishes, supplemented with 2% glucose and 0.2% NaHCO3 (150 (:1 of 20% and 2% solutions, respectively), and placed in the incubator. In the red blood cell (RBC) culture procedure, infected (>75% parasitemic) and noninfected blood were centrifuged and the plasma and buffy coat were removed. The infected and noninfected RBC's (IRBC's and URBC's, respectively) were washed twice with the same medium used in each experiment. Approximately 10% parasitized RBC's were prepared by mixing the washed IRBC's and URBC's. The 10% parasitized RBC's were suspended in the appropriate medium to yield a packed cell volume of 5%. The media used for incubation were supplemented with either 10% or 20% heat denatured (56 °C, 30 minutes) pig serum (PS) or fetal calf serum (FCS), as described in each experiment. The mixture (1.5 m1) of the 10% parasitized RBC's and the medium was placed in 35 mm diameter Petri dishes (IRBC culture), and incubated in the appropriate incubator. In EXP II-A, 60 mm diameter Petri dishes were used with 5 ml of the RBC's and medium mixture. In study II, URBC's alone were incubated (URBC culture) under the same conditions to provide a negative control. 50 .Measurements. In study I, sequential samples were taken at various time intervals. In study II, cultures were sampled every 12 hours. Samples were taken at several points of the settled layer of the erythrocytes: RBC thin smears were prepared and stained by Wright's stain to evaluate the percentage of parasitized RBC's by counting the number of parasitized RBC's in over 400 RBC's examined. In study II experiments, glucose concentrations of the culture media were measured (Glucose Procedure No. 510; Sigma Chemical Co.). Cumulative glucose consumption (mg/ml) was calculated by the equation: (initial glucose concentration) - (glucose concentration of the interval sample). In EXP II-A, lactate and pyruvate concentrations of the culture media were measured (Lactate Procedure No. 826-UV, Pyruvate Procedure No. 726-UV; Sigma Chemical Co.). After taking the interval samples, the settled layer of RBC's and the supernatant medium were mixed by gentle swirling before returning to the incubator. Experimental design and statistical analysis. Study 1, consisting of three experiments (EXP's), was a preliminary screening of culture conditions and evaluated the effects of anticoagulants (EXP I-A), gaseous envirOnments (EXP I-B) and media (EXP I-C) on E. suis growth in vitro. Study II was designed to examine the in vitro metabolic activity of E. suis and further refine the in vitro culture 51 conditions. Five EXP's were designed with 3 (EXP II-A, B, C) or 5 (EXP II-D and E) replicates to allow for statistical analysis. EXP II-A examined glucose consumption, and lactate and pyruvate production in IRBC culture. EXP II-B evaluated the effect of inosine addition to the medium and EXP II-C studied the effect of EDTA addition to the medium. EXP II-D assessed serum.supp1ementation (PS vs. FCS) of the medium and the type of incubator (5% C02 incubator vs. candle jar). In EXP II-E, the effect of medium refreshment (every 12 or 24 hours during 72 hours incubation) was assessed. A 10% C02 incubator was used in EXP II-A and 5% C02 incubator was used in EXP's II-B, C, D and E. Statistical analysis was done using the computer program, StatView (version 4.0, Abacus Concepts, Berkeley, CA). Student's t test was used for statistical comparison of 2 groups at each subsampling time. One way analysis of variance and multiple comparisons with Fisher's least significant difference were used for comparing 3 or more groups at each subsampling time. 52 Results STUDY I 0 PRELIMINARY SCREENING OF CULTURE CONDITIONS Combinations of culture condition factors were applied to WB and RBC cultures in several separate experiments. EXP I-A. Anticoagulant. The effects of different anticoagulants on the percent parasitized RBC's in WB culture using blood collected with sodium heparin, EDTA, sodium citrate or glass beads are presented. in Table 2-1. Parasitized erythrocytes were observed for the longest time in heparinized blood incubated in the 10% C02 incubator. Few differences of the effects of anticoagulants were observed in the first 12 hours of culture. However, during the period of 12 to 24 hours of culture, a rapid decrease of the percent parasitized RBC's was observed in the EDTA anticoagulated WB culture. Heparinized and citrate anticoagulated WB cultures did not exhibit a remarkable decrease in the percentage of parasitized RBC's until after 24 hours of culture. Obvious differences were not detected between heparin and glass beads anticoagulated WB cultures. In further experiment, blood was collected with sodium heparin containing solution. EXP I-B. Gaseous environment. The effects of different incubators and gaseous environments on the percentage of parasitized RBC's in WB and IRBC cultures are summerized in Table 2-1. 53 The change in the percentage of parasitized erythrocytes in whole blood culture with different anticoagulants and incubators. Incubation length (hours) Antico_agdant Incubator 0 3 6 9 12 24 36 48 Heparin 1096 C02 10.9 7.8 10.5 14.2 11.5 6.7 1.3 0.9 EDTA 10% C02 13.0 8.4 7.4 9.3 10.3 1.0 O 0 Citrate 10% C02 11.4 14.1 8.4 12.4 11.0 10.8 0.3 0 Incubation length (hours) 0 3 8 24 48 Heparin Candle jar 6.0 7.5 ’ 8.3 4.1 0 Beads Candle jar 4.7 2.7 3.5 0.4 0 Table 2-2. The change in the percentage of parasitized erythrocytes in whole blood (WB) and infected red blood cell (IRBC)‘ cultures. Incubation time (hours) Culture Incubator 0 2 4 6 8 24 W8 5% C02 11.0 11.8 10.2 15.0 17.5 5.4 WB 10% C02 11.0 9.2 13.3 10.9 13.6 7.5 W8 anaerobic 11.0 5.8 16.3 10.5 16.9 2.2 IRBC 5% C02 11.0 4.3 3.5 1.7 1.0 0 IRBC 10% C02 11.0 6.8 1.7 1.2 0.7 0 IRBC anaerobic 1 1.0 4.9 1.2 4.4 0 0 IRBC Candle jar 15.8 N/A'r 1.3 N/A 0 0 IRBC Gas chamber 15.8 N/A 12.2 N/A 0 0 . - IRBC culture was performed with RPMI 1640 contained 20% pig serum. 1 - Not applied 54 Table 2-2. E. suis attachment to erythrocytes was observed at varying intervals during the 24 hour incubation time in WB culture with 5% and 10% C02, and anaerobic incubators. In IRBC culture, longer attachment was observed in either 5% or 10% C02 incubators than in the candle jar or the anaerobic incubator although a hihger percentage of parasitized erythrocytes were observed at 6 hours of culture with the anaerobic incubator compared to the 5 or 10% C02 incubators. In further experiments, 5% or 10% C02 incubators were used. EXP I—C. Medium. The effects of various media on the percentage of parasitized RBC's in IRBC culture are shown in Table 2-3. For comparison, the percentage of parasitized RBC's in WB cultures performed at the same time are also shown. Parasitized RBC's were observed for at least 48 hours in IRBC culture with EM and rEM, and WB culture with glucose and NaHCO3, but not in other media (RPMI 1640 or MM) or in WB culture without glucose or NaHC03 addition. The percentage of parasitized RBC's in IRBC culture with EM and rEM, and WB culture with added glucose and NaHC03 were maintained or increased at 8 hours of incubation but decreased by 24 hours. Since rEM yielded a higher percentage of parasitized RBC's at 8 and 48 hours incubation compared to EM, rEM was selected as the base medium for further experiments, in addition to heparin as the optimal anticoagulant, and 5 and 10% C02 incubators as the optimal gaseous environment. 55 Table 2-3. The change in the percentage of parasitized erythrocytes in infected red blood cell culture with various media‘ and whole blood culture. Incubation lenggmours) Medium 0 2 4 6 8 24 48 RPMI 1640 11.0 6.8 1.7 1.2 0.7 0 MIA? MM? 11.0 6.1 7.9 0.5 0 0 N/A WB—glc, NaHCO3§ 11.0 9.2 13.3 10.9 13.6 7.5 N/A WB-nonll 1 1.0 2.5 3.9 0 0 0 N/A EMS 11.0 N/A N/A N/A 9.1 4.7 2.1 rEM’ 11.0 N/A N/A N/A 15.8 1.4 6.8 we, NaHCO; 11.0 N/A N/A N/A 20.4 3.1 3.9 All media contained 20% pig serum. Not applied. Mycoplasma medium Whole blood culture with 2% glucose and 0.2% NaHCO3 Whole blood culture without glucose and NaHCO3 Eagle's minimum essential medium Reduced Eagle's minimum essential medium with B-mercaptoethanol she-cm-Hu-o-e IIIIIII 56 STUDY II. Msnsmu'r or E. 5015 METABOLIC ACTIVITY AND Funmuak SCREENING or CULTURE CONDITIONS EXP II-A. Glucose consumption and lactate and pyruvate production. Twenty one replicate culture dishes were prepared for each IRBC‘and URBC culture, with rEM containing 10% PS. In each group, three of the replicates were tested at the beginning of incubation and every 12 hours for 72 hours, and measured for the percentage of parasitized RBC's and glucose, lactate and pyruvate concentrations. Figure 2-1 illustrates the percentage of parasitized RBC's (A), and cumulative glucose consumption and lactate production (B) during incubation. During the first 24 hours of incubation, a rapid decrease in the percentage of parasitized RBC's, and a rapid increase in cumulative glucose consumption and lactate production were observed. Positive correlations were observed between the percentage of parasitized RBC's and the rate of glucose consumption and lactate production. Obvious changes in glucose consumption and lactate production in URBC culture were not detected. Pyruvate levels in IRBC and URBC cultures stayed under the detection level of the assay system throughout the experimental period. EXP II-B. Addition of inosine to medium. The effect of providing an energy substrate for erythrocytes by adding inosine in addition to glucose was evaluated by measuring the percentage of parasitized RBC's and cumulative glucose 57 A. Percentage of parasitized RBC‘s 9630 25 2O 15 1O B. Cumulative glucose consumption and lactate production mm 250 t 150 a ‘°° ': + eIc-Iasc -e— GIc-UFIBC 50 .2 + Lac-IRBC —£:— Lac-UFIK) o ‘ , ___ — I '50 I I T l I I I 0 12 24 36 48 60 72 Inanumonfinuwhmus) Figure 2-1. The percentage of parasitized erythrocytes (A) and cumulative glucose (Glc) consumption and lactate (Lac) production (B) in red blood cell (RBC) culture with reduced Eagle's medium. IRBC - Infected RBC culture, URBC a Uninfected RBC culture. Error bars represent standard deviations. 58 consumption in IRBC and URBC cultures with rEM and rEMI, both containing 10% PS (Figure 2-2). The percentage of parasitized RBC's in both media were gradually decreased during the incubation period. Cumulative glucose consumption in both media rapidly increased during the first 36 hours and remained at the same level for the rest of the incubation time. Inosine addition did not significantly change any of the parameters except for increasing the cumulative glucose consumption value at 60 hours. The major advantage of inosine addition was that erythrocyte (both infected and uninfected) integrity was better maintained in rEMI compared to rEM as detemined by the morphology in Wight's stained thin smears. Evaluating the percentage of parasitized RBC's by Wright's stained thin smears was easier. with rEMI. Therefore, rEMI was used as the base media for further experiments. Notably, the coefficients of variation (standard deviation/mean) of the percent parasitized RBC's were much greater than those of the cumulative glucose consumption. The small coeficients of variation in the cumulative glucose consumption were consistently observed in EXP's II-B, C, D and E. EXP II-C. Addition of.EDTA to medium. Under the assumption that E. suis loses its metabolic activity by EDTA treatment, IRBC and URBC cultures were performed with rEMI and rEMI-EDTA (both containing 10% PS) to evaluate if the apparent glucose 59 A. Percentage of parasitized RBC's 9:20 18 -I— rEM 16 ,4 -e— IBM! 1: ‘K 070-500 I I I I I 0 12 24 38 ‘ 48 80 72 B. Cumulative glucose consumption mg/rnl100 . so ‘“ so - 7o ‘ so 50 -s- rEM-Ines -e— rEM-URBC 4o 3 + rEMI-mac -e— rEMI-URBC so 20 10 -,. O T l T I I ' o 12 24 as 48 so 72 lnaflnfionfinmwhdms) Figure 2-2. The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (B) in red blood cell (RBC) culture with reduced Eagle's medium.containing or not containing inosine (rEM and rEMI, respectively). IRBC - Infected RBC culture, URBC = Uninfected RBC culture. *- Significant difference (0.015P<0.05) between rEM‘and rEMI observed in IRBC culture. Error bars represent standard deviations. 60 consumption is altered by the death of E. suis, thus determining that the source of (the observed glucose consumption is the metabolic activity of E. suis itself rather than infected erythrocytes. Changes in the percentage of parasitized RBC's and cumulative glucose consumption, are provided in Figure 2-3. The percent parasitized RBC's in both media were initially increased at 12 hours. However, EDTA treatment caused a rapid decrease in the percentage of parasitized RBC's during 12 to 36 hours of the incubation time. The cumulative glucose consumption in IRBC culture with rEMI-EDTA was significantly lower than observed with rEMI at every sampling time except hour 0. Glucose consumption in IRBC culture with rEMI-EDTA ceased at 24 hours, since the cumulative glucose consumption stayed at the same level from 24 to 72 hours of incubation time. Considering that the number of E. suis attached per infected erythrocyte did not increase during the experimental period, a positive correlation was noted between the number of E. suis attached to RBC's and the rate of glucose consumption. The cumulative glucose consumption values of the rEMI- EDTA in URBC culture gradually decreased over the experimental period. Those values were significantly less than those of the rEMI in URBC culture during 12 to 72 hours of incubation. 61 A. Percentage of parasitized FIBC's 9:30: -IF- dflM '—.-IEMWEDTA 8. Cumulative glucose consumption W140- -|— rEMI-IRBC 12°; + rEMI-EDTA-IRBC i -a— rEMI-URBC 100- ; -e- rEMI-EDTA-UFIBC 80 1 . ,. . ‘l a s 2 t t 60- -l I. 40; -I O. 20 a: 0 d .2 - . :1 z :I '20 I I I I 3? ” O 12 24 36 48 60 72 Incubation time (hours) Figure 2-3. The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (B) in red blood cell (RBC) culture with reduced Eagle's medium containing inosine and EDTA (rEMI-EDTA) and reduced Eagle's medium containing only inosine (rEMI). IRBC - Infected RBC culture, URBC 8 Uninfected RBC culture. to. Significant difference (P<0.01) between rEMI and rEMI-EDTA in IRBC culture. Error bars represent standard deviations. 62 EXP II-D. Selection of serum and incubator. The effect of PS versus FCS addition to the medium, and incubating in 5% C02 incubator versus a candle jar were evaluated. Changes in the percentage of parasitized RBC's in each combination of serum and incubator is shown in Figure 2-4 A. Any indications of the organisms' multiplication or significant differences between the combinations of sera and incubators were not observed, except at the time when the percentage of parasitized RBC's was very low. Significant differences of glycolytic activity were detected with the different sera (Figure 2-4 B). A significantly greater cumulative glucose consumption was observed in rEMI containing FCS compared to PS from 24 to 72 hours. Few differences were observed between the C02 incubator and the candle jar (Figure 2-4 C). The cumulative glucose consumption values in URBC cultures gradually decreased over the experimental period. EXP II-E. Refreshment of medium. Using the optimal medium (rEMI containing 10% FCS) and incubator (5% C02) based on the results of the previous experiments, IRBC cultures composed of 3 groups of 5 replicates were prepared. As a negative control, URBC cultures were also prepared with 2 replicates in each group. In groups 1 and 2, old medium in the Petri dishes was removed and fresh medium was added every 12 and 24 hours, respectively. In group 3, medium was not changed over the experimental period. Changes in the percentage of parasitized RBC's and cumulative glucose consumption are 63 Figure 2-4. The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (B and C) in red blood cell (RBC) culture with reduced Eagle's medium with inosine (rEMI) containing 10% fetal calf serum (FCS) or pig serum (PS), incubated in a 10% C02 incubator or a candle jar. B. Cumulative glucose consumption in IRBC and URBC culture with rEMI containing 10% FCS or PS, incubated in a 10% C02 incubator. C. Cumulative glucose consumption in IRBC and URBC culture with rEMI containing 10% FCS, incubated in a 10% C02 incubator or a candle jar. tt's - Significant differences (P<0.01) between FCS- CO; and PS-CO; (--1), between FCS-IRBC and PS-IRBC (t-Z), and between COz-IRBC and Candle-IRBC (tt3). Error bars represent standard deviations. 64 A. Percentage of parasitized RBC's 9620 15- ; + FCS-CO2 Io-i -e— Ps-ooa ; + FCSCandle 5‘. —e-— PS-Candle O ‘ l O 12 24 36 48 60 72 8. Cumulative glucose consumption - Serum rngllnl 120 100 80 ..2 .‘2 ..2 + FCS-IRBC so “2 + PS-IRBC eez 4° , .9— FCS-URBC 2° . __ c -e— PS-UFIBC ° «fifiw '20 I I I I I I I O 12 24 36 48 6O 72 C. Cumulative glucose consumption - Incubators se3 Inomu12o 100 —I— CO2-IFIK: 80 60 + Candle-IRBC ‘0 43- cos-URBC 20 0 —e— Candle-URBC '20 I I I I I I I o y 12 24 as 48 so 72 Incubation time (hours) Figure 2-4. 65 shown in Figure 2-5. The percentage of parasitized RBC's in groups 1 and 2 remained at the same level, although the standard deviations at each subsampling time were great. In group 3, there was a gradual decrease in the percentage of parasitized RBC's, declining to 0% at 60 hours. Significant differences were observed at several sampling times between groups 1 and 3, and between groups 2 and 3. Refreshment of medium (groups 1 and 2) resulted in higher cumulative glucose consumption at sampling times from 36 to 72 hours. Similar effects on the percentage of parasitized RBC's and cumulative glucose consumption were observed between groups 1 and 2, although group 1 exhibited significantly greater values at 60 hours in the percentage of parasitized RBC's and at 72 hours in the cumulative glucose consumption compared to group 2. 66 A. Percentasge of parasitized RBC's -l- Nochange. + 12hours + 24hours 96 6 ,1 5 t1 ¢t2 .41 4 .32 \\ ee3 3 0‘ ,- ‘ 2 7 kg 1 0 I I T I T o 12 24 as 48 so 72 8. Cumulative glucose consumption ”1 rmghilSO ..2 13:16 160 -I- No change ul 140 -e- 12 hours ./ eel 120 «at- 24hmts “2 Ioo‘ ""1 _ . ‘2 so / so / 40 20 o I I I II I I 0 12 24 38 48 80 72 Incflnmonflme(NNmm Figure 2-5. The percentage of parasitized erythrocytes (A) and cumulative glucose consumption (B) in infected red blood cell (IRBC) culture with medium.refreshment every 12 hours and 24 hours, and without changing medium, t and as . Significant differences ($80.015P<0.05, taaP<0.01) between ”No change" and "12 hours“ (:1, :41), between ”No change” and '24 hours” (#2, «#2) and between '12 hours" and '24 hours“ (#:3). Error bars represent standard deviations. 67 Discussion The successful cultivation of parasitic protozoa requires optimum biophysical growth conditions (temperature, pH, 0; and CO; concentrations) and the nutritional supplementation of amino acids, vitamins, carbohydrates, and sometimes blood or blood derivatives (Enders, 1988). Successful in vitro culture systems have been described for Plasmodium and Babesia (Jensen, 1983; Levy & Ristic,_1983) after a long history of attempts and failures to select the optimum culture conditions. One of the most fascinating findings in the history of Plasmodium in vitro culture is that the parasite was an obligate microaerophile, and favored lower oxygen tensions (5 - 15%) and higher carbon dioxide tensions (3 - 5%) than provided by air (Jensen, 1983; Scheibel et al., 1979; Trager, 1941). With this evidence, Trager and Jensen (1976) introduced the Petri dish-candle jar method for in vitro cultivation of Plasmodiumfalciparum, using RPMI 1640 as a base medium. This technique was applied to the in vitro culture of Anaplasma marginale, and in vitro multiplication and maintenance of infectivity for up to 33 days was achieved (Kessler & Ristic, 1979; Kessler et al., 1979; Mazzola & Kuttler, 1980). In study I, preliminary screening of culture conditions was based on modifications of the culture procedure developed for Plasmodium falciparum (Jensen & Trager, 1977). However in RPMI 1640 systems, parasitized RBC's were no longer 68 visible by 24 hours of incubation, regardless of the types of incubator. Incubating whole blood with glucose and sodium bicarbonate yielded superior results compared to RPMI 1640 systems as parasitized RBC's were observed for up to 48 hours. Since E. suis is attached externally to RBC's or lives free in the plasma, E. suis is exposed to extracellular conditions, unlike malaria parasites which reside inside RBC's. In vitro culture systems for other extracellular blood parasites such as salivarian trypanosomes, had been developed using Eagle's minimum essential medium. For example, ITypanosoma brucei was originally cultivated in feeder layer systems using RPMI 1640 or Eagle's medium. Later, cell free semi-defined medium based on Eagle's medium was established by providing reducing conditions using 6- mercaptoethanol instead of feeder layer cells (Brun & Jenni, 1987). In the experiment of medium selection (EXP I-C), reduced and non-reduced EM were applied to the Petri dish system, and both conditioned medium yielded acceptable results for in vitro maintenance of E. suis attachment on RBC's during cultivation. Glucose consumption, and lactate and pyruvate production by E. suis was expected based on the observation of hypoglycemia in acutely infected animals, and decreased glucose and increased lactate concentration during in vitro incubation of EDTA anticoagulated blood with supplemental glucose (Heinritzi et al., 1990a; Heinritzi et al., 1990b). In the rEM culture system, a rapid decrease in glucose levels 69 and a rapid increase in lactate levels in the culture medium were observed (EXP II-A). These reSults could have been caused by parasite metabolism. alone and/or enhanced erythrocyte metabolism caused by E. suis. It is known that glucose consumption for energy production by pig erythrocytes is extremely low, because pig erythrocytes lack a functional glucose transporter (Harvey, 1989; Rivkin & Simon, 1965). It has been observed that the amount of ATP possessed by pig erythrocytes decreases by 50% at 8 hours of in vitro incubation, regardless of the presence or absence of glucose (Kim & McManus, 1971). Although pig erythrocytes have the capacity for complete glycolysis, glucose uptake is normally limited by the low permeability to glucose (Kim & McManus, 1971), suggesting that pig erythrocytes could consume glucose if glucose enters into pig erythrocytes by some other means. Electronmicrographs have suggested that E. suis parasitism causes membrane deformation of erythrocytes (Zachary & Basgall, 1985). Therefore, it could be hypothesized that E. suis infection may enhance erythrocyte permeability for glucose by disturbing erythrocyte membrane structure, thus the infected erythrocytes may consume more glucose than noninfected erythrocytes. Parasite induced, increased glucose permeability has been observed in Babesia bovis infected erythrocytes (Upston & Gero, 1990). EXP II-C was conducted to determine if E. suis actually consumes glucose during in vitro culture by evaluating if E. suis death affects the apparent glucose consumption in vitro. In this 70 experiment, the assumption was made that E. suis is killed by EDTA treatment, thus losing its glycolytic activity. The effect of EDTA was observed in EXP I-A, in which Eu suis could not maintain its attachment to RBC's in the EDTA anticoagulated blood when compared to heparin or citrate anticoagulated blood. Also, a previous study found that E. suis lost its infectivity with EDTA treatment (Hall et al., 1988). In EXP II-C, EDTA treatment showed profound negative effects on E. suis maintenance of RBC attachment and glycolytic activity, and a positive correlation between the number of attached E. suis and the rate of glucose consumption was observed. This indicates that glucose consumption in the culture system was primarily due to E. suis glycolysis. The possibility that the glycolytic activity of erythrocytes was enhanced by E. suis infection was unsubstantiated by observation that the integrity of infected erythrocytes incubated with IBM (without inosine) was not maintained sufficiently. If the glycolytic activity of infected erythrocytes was enhanced, infected erythrocytes were supposed to sufficiently acquire energy from glycolysis for maintaining their membrane integrity. Therefore, poor maintenance of membrane integrity of infected erythrocytes under the presence of sufficient glucose suggested that the glycolytic activity' of infected. erythrocytes was not sufficiently enhanced to maintain membrane integrity. 71 In EXP II-C, significant differences were also observed in the cumulative glucose consumption in the URBC culture between rEMI and rEMI-EDTA. EDTA is known to chelate heavy metal ions, such as Ca2+ and Mg”. In the glycolytic pathway, magnesium ions serves as a cofactor for many enzymes. It could be assumed that EDTA. may inhibit glycolysis. Therefore, the observed difference in URBC culture between rEMI and rEMI-EDTA could indicate that pig erythrocytes consumed some glucose, and glycolysis of pig erythrocytes was inhibited by EDTA. On the other hand, the blocking of glycolysis by EDTA may explain in part that E. suis was killed by EDTA treatment. In study I and study II-A, erythrocytes observed in the thin smears, especially in IRBC culture after 24 hours of incubation, exhibited blunt membranes. This observation suggested that the erythrocyte membranes became fragile because of En suis parasitism and/or in vitro incubation. In feline haemobartonellosis, it has been observed that the osmotic fragility of HaemobartonelLa felis infected erythrocytes was increased as a result of decreased erythrocyte lipid concentration (Maede, 1980). The cause of decreased lipids concentration in erythrocyte membrane was not clear, however several possibilities were listed: 1) the erythrocyte membrane may be altered by lipolytic enzymes produced by H; felis; 2) H. felis may utilize the erythrocyte lipids for its own nutrition; and 3) inadequate energy supply to erythrocytes because of hypoglycemia caused by H; felis 72 infection may cause lipid loss (Cooper & Jandle, 1969; Maede, 1980). As discussed above, pig erythrocytes can not acquire glucose sufficiently for its sole energy source. To provide energy to pig erythrocytes, the medium was supplemented with inosine (EXP II-B). Pig erythrocytes may use the nucleoside inosine as an energy source by the following proposed mechanism: inosine + Pi -* ribose-l-P + hypoxanthine ribose-l-P -+ ribose-S-P The ribose-S-P is integrated into the pentose phosphate shunt in the glycolytic pathway. Remarkable effects of inosine addition to the medium on the maintenance of E. suis attachment on RBC's and glycolytic activity were not detected in EXP II-B; however, the in vitro:maintenance of erythrocyte integrity was improved by addition of inosine to the medium and erythrocyte membranes in the thin smears remained sharp for at least 72 hours of incubation. Therefore, inosine addition facilitated easier and possibly more accurate evaluation of the percentage of parasitized RBC's in thin smears. In EXP II-D, the effect of serum and gaseous environment were evaluated using the rEMI system. In the various in vitro culture systems, serum contributes various undefined nutrients including growth factors, vitamins and ions (Brun & Jenni, 1987). In the present experiments, FCS produced a profound positive effect on E. suis glycolytic activity. FCS is known to contain various growth factors which are required '73 for fetal calf growth. It appears that these factors also have beneficial effects on E. suis growth in vitro. The candle jar used for Plasmodium falciparum was re-evaluated in the same experiment using the rEMI system; however, no improvements were observed. In the final experiment (EXP II-E), medium refreshment at either 12 or 24 hour intervals during incubation prolonged attachment of E. suis on erythrocytes and improved glucose consumption. Refreshment of medium provides additional nutrients and removes toxic metabolites. Production of lactate, one of the major metabolites in the culture system, might cause negative effects on E. suis activity by altering the pH of the medium. The negative effect of decreased pH as a result of lactate production was discussed in the in vitro culture system for malaria parasites (Jensen, 1983). In EXP's II-B through II-E, the coefficients of variation from cumulative glucose consumption data were much less than the percentage of parasitized RBC's data. This finding suggested measuring glucose consumption was a more ‘ precise method for evaluating E. suis activity in the developed culture system. The subsamples for thin smears were taken from the several points of the settled erythrocyte layer at the bottom of the culture dishes. Therefore, if E. suis grows as a cluster during cultivation, the percentage of parasitized RBC's was highly dependent on the location from which subsamples were taken. In: addition, Henry (1979) described that artifacts of red cells in blood smears were 74 difficult to distinguish from E. suis bodies in many cases. Especially, small forms of the organism, presumably as a result of budding multiplication, were almost impossible to distinguish from stain deposit or stained dust particles. Measuring glucose consumption has another advantage for evaluating E. suis activity in vitro. Glucose could be consumed by E. suis free in the medium as well as E. suis associated with erythrocytes. Therefore, glucose consumption reflects the activity of E. suis both associated with RBC's and free in the medium, while the percent parasitized RBC's did not account for E. suis free in the medium. The negative values observed in the cumulative glucose consumption in URBC cultures, especially at the longer incubation period, are probably a reflection of the medium evaporation. Since glucose does not evaporate from the medium, the relative glucose concentration in the medium increases as water evaporated. In EXP II-B, water evaporation in each Petri dish between subsampling times was measured by weighing Petri dishes before and after incubation. The weight of Petri dishes decreased by 0.8% of the medium weight in each 12 hour interval with negligible variation between Petri dishes, indicating glucose concentration in each Petri dish was equally concentrated by 0.8%. Since EXP II-B was assumed to be representative for the effect of water evaporation, the weight of the Petri dishes were not measured in other experiments. 75 Throughout the experiments, an increase in the percentage of parasitized RBC's was occasionally observed during the first 12 hours of incubation. It is not certain if this observation was a reflection of active cell to cell transmission or simply a manifestation of an infection initiated in vivo. As a general tendency, the number of parasitized RBC's gradually decreased during the incubation period, except EXP II-E in which a low percentage of parasitized RBC's were maintained by refreshment of the medium, throughout the incubation period. Ichijo et al. (1982) observed a 35% parasitemia in bone marrow of E. ovis infected sheep while the parasitemia in peripheral blood was 4%, and suggested that bone marrow may be a main site for parasite multiplication. According to this finding, bone marrow culture may be a good alternative for the proposed RBC-Petri dish system. Attempts at in vitro culture with bone marrow were reported with Anaplasma marginale with bovine bone marrow (Hruska et al., 1968), and rabbit bone marrow (Marble & Hanks, 1974a; Marble & Hanks, 1974b). However, clear evidence of active replication of A. marginale in the bone marrow systems was not obtained. In an electron microscopic study of bone marrow obtained from an E. suis parasitemic pig, we were unable to detect E. suis parasitism on immature red cells (unpublished data). Therefore, the value of bone marrow culture for E. suis is questionable. 76 In the RBC culture procedure, erythrocytes settled at the bottom of the Petri dishes during incubation period, probably in the first couple of hours. It is suspected that E. suis are immobile organisms because no structures associated with mobility, such as pili, flagella or actin, were observed in electronmicroscopic studies (Pospischil & Hoffmann, 1982; Zachary & Basgall, 1985). Therefore, an opportunity for E. suis to encounter new erythrocytes was limited in the developed culture system. It may be advantageous to apply gentle agitation during cultivation. Since E. suis is an extracellular organism living in blood, gentle agitation may mimic its in vivo environment. The proposed RBC culture is the first description of an in vitro culture system for E. suis using semi-defined medium. Although continuous cultivation was not achieved, the system was successful for short term maintenance of E. suis attachment on erythrocytes and measurement of E. suis glycolytic activity. The 72 hour time period appears to be sufficient for evaluating anti-E. suis compounds in vitro, as demonstrated in EXP II-C using EDTA. Furthermore, the system could be used for protein and DNA analysis by radiolabeling the organism using radiolabeled nutrient, since metabolic activity can be maintained in the system (refer to Chapter III). Further investigation of the in vitro requirements of E. suis growth could lead to the continuous in vitro propagation of E. suis and establishment of in vitro experimental systems for various purposes. 77 References 1. Brun, R. and Jenni, L. (1987): Salivarian trypanosomes: Blood stream forms (trypomastigotes). In In vitro methods for parasite cultivation, Ed. A. E. R. Taylor and J. R. Baker, Academic Press, London, UK, pp. 94-117 2. Cooper, R. A. and Jandle, J. H. (1969): The selective and conjoint loss of red cell lipids. J. Clin. Invest., 48, 906-914 3. Enders, B. (1988): In vitro cultivation of certain parasitic protozoa: Biochemical data and technical requirements for mass production. In Parasitology in focus: Facts and Trends, Ed. H. Mehlhorn, Springer-Verlag, Berlin, Germany, pp. 702-718 4. Geary, T. G., Divo, A. A. and Jensen, J. B. (1983): An in vitro assay system for the identification of potential antimalarial drugs. J. Parasitol., 69, 577-583 5. Hall, S. M., Cipriano, J. A., Schoneweis, D. A., Smith, J. E. and Fenwick, B. W. (1988): Isolation of infective and non-infective Eperythrozoon suis bodies from the whole blood of infected swine. vet. Rec., 123, 651 6. Harvey, J. W. (1989): Erythrocyte Metabolism. In Clinical biochemistry of domestic animals, 4th ed., Ed. J. J. Kaneko, Academic Press, San Diego, CA, pp. 185-234 7. Heinritzi, K., Peteranderl, W. and Plank, G. (1990a): Eperythrozoon-infection in pigs: Influence upon acid-base- balance, and the levels of glucose, lactic, and pyruvic acid in venous blood. Dtsch. tierarztl. wschr., 97, 31-34 8. Heinritzi, K., Plank, G., Peteranderl, W. and Sandner, N. (1990b): Investigation on acid-base balance and carbohydrate metabolism in Eperythrozoon suis infected pigs. J. vet..Med., 37, 412-417 9. Henry, S. (L. (1979): Clinical observation on Eperythrozoonosis. JAVMA, 174, 601-603 10. Hruska, J. C., Kliewer, I. O. and Brock, W. E. (1968): Studies in tissue culture of Anaplasma marginale. In Proc. 5th National Anaplasmosis Conference, Stillwater, OK, pp. 21- 24 . 78 11. Hsu, F. S., Liu, M. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1990): Evaluation of an enzyme-linked immunosorbent assay for the detection of Eperythrozoon suis antibody in swine. In Proc. 11th International Pig Veterinary Society congress, Lausanne, Switzerland pp. 315 12. Ichijo, S., Hosokawa, S., Kim, D.-H. and Konishi, T. (1982): Scanning and transmission electron microscopic observation of Eperythrozoon ovis (E. ovis) . Jpn. J. Vet. Sci., 44, 127-132 13. Jensen, J. B. (1983): Plasmodium. In In vitro cultivation of protozoan parasites, Ed. J. B. Jensen, CRC Press, Boca Raton, FL, pp. 155-192 14. Jensen, J. B. and Trager, W. (1977): Plasmodium falciparum in culture: Use of outdated erythrocytes and description of the candle jar method. J. Parasitol., 63, 883-886 15. Kessler, R. H. and Ristic, M. (1979): In vitro cultivation of Anaplasma marginale: Invasion of and development in noninfected erythrocytes. Am. J. vet. Res., 40, 1774-1776 16. Kessler, R. H., Ristic, M., Sells, D. M. and Carson, C. A. (1979): In vitro cultivation of Anaplasma marginale: Growth pattern and morphologic appearance. Am. J. vet. Res., 40, 1767-1773 17. Kim, H. D. and McManus, T. J. (1971): Studies on the energy metabolism of pig red cells I. The limiting role of membrane permeability in glycolysis. Biochim. Biophys. Acta, 230, 1-11 18. Levy, M. G. and Ristic, M. (1983): Cultivation and in vitro studies of Babesia. In In vitro cultivation of protozoan parasites, Ed. J. B. Jensen, CRC Press, Boca Raton, FL, pp. 221-241 19. Maede, Y. (1980): Studies on feline haemobartonellosis. VI. Changes of erythrocyte lipids concentration and their relation to osmotic fragility. Jpn. J. vet. Sci., 42, 281- 288 20. Marble, D. W. and Hanks, M. A. (1974a): A. marginale grown in stable rabbit bone marrow cells. In Proc. 6th National Anaplasmosis Conference, Las Vegas, NV, pp. 53-54 21. Marble, D. W. and Hanks, M. A. (1974b): Anaplasmosis in primary rabbit bone marrow cells. In Proc. 6th National Anaplasmosis Conference, Las Vegas, NV, pp. 49-50 79 22. Mazzola, V. and Kuttler, K. L. (1980): Anaplasma marginale in bovine erythrocyte cultures. Am. J. vet. Res., 41, 2087-2088 23. Pospischil, A. and Hoffmann, R. (1982): Eperythrozoon suis in naturally infected pigs: A light and electron microscopic study. vet. Pathol., 19, 651-657 24. Rivkin, S. E. and Simon, E. R. (1965): Comparative carbohydrate catabolism and methemoglobin reduction in pig and human erythrocytes. J. cell. Comp. Physiol., 66, 49-56 25. Scheibel, L. W., Ashton, S. H. and Trager, W. (1979): Plasmodium falciparum: Microaerophilic requirements in human red blood cells. Expt. Parasitol., 47, 410-418 26. Sisk, D. B., Cole, J. R. and Pursell, A. R. (1980): Serologic incidence of eperythrozoonosis in Georgia swine. Proc. Annu. Meet. Am. Assoc. vet. Lab. Diag., 23, 91-100 27. Smith, A. R. (1977): Eperythrozoonosis. veterinary (professional topics, university of Illinois: Swine, 5, 2-4 28. Smith, A. R. (1981): Eperythrozoonosis. In Diseases of swine, Ed. B. S. A. D. Leman R. D. Glock, W. L. Mengeling, R. H. C. Penny and E. Scholl, Iowa State University Press, Ames, IA, pp. 683-687 29. Trager, W. (1941): Studies on conditions affecting the survival in vitro of a malaria parasite (Plasmodium lqphurae). J. Exp..Med., 74, 441-461 30. Trager, W. and Jensen, J. B. (1976): Human malaria parasites in continuous culture. Science, 193, 673-675 31. Upston, J. M. and Gero, A. M. (1990): Increased glucose permeability in Babesia Davis-infected erythrocytes. Int. J. Parasitol., 20, 69-76 32. Zachary, J. F. and Basgall, E. J. (1985): Erythrocyte membrane alterations associated with the attachment and replication of Eperythrozoon suis: A light and electron microscopic study. vet. Pathol., 22, 164-170 CHAPTER 3 LACK OF SPECIFICITY OF THE INDIRECT HEMAGGLUTINATION ASSAY FOR SWINE EPERYTHROZOONOSIS AND IDENTIFICATION OF EPERYTHROZOON SUIS SPECIFIC ANTIGENS 80 81 Abstract The indirect hemagglutination assay (IHA) for swine eperythrozoonosis was evaluated for specificity in detecting E. suis specific antibodies by analyzing the effects of incubation temperature and absorption of immune sera with normal pig erythrocytes, and by using IHA antigens prepared from infected and noninfected blood in the IHA, ELISA and western blots. With incubation at 37’°C instead of 4 °C used in the standard IHA procedure, 77% of tested sera showed either an increase or decrease in titer compared to their respective standard IHA titers. More than two fold titer decreases were observed in standard IHA positive (>1:80) and suspicious (1:40) sera, indicating an active role of cold temperature reacting antibodies in IHA. With preincubation of test sera with normal pig erythrocytes at either 4 or 37 °C, the titers of standard IHA positive sera tended to decrease, regardless of preincubation temperature, indicating an active role of autoantibodies in IHA. No differences were observed in either the IHA titers or ELISA optical densities between Es+ (E. suis infected) and Es“ (E. suis noninfected) antigen preparations. E. suis specific antigens were not readily detected in the IHA antigen preparations by western blotting, indicating antibodies measured in IHA were directed mainly at host proteins. Three methods were investigated for identification and partial purification of E. suis antigens: 1) high speed 82 centrifugation of E. suis dissociated from erythrocytes by EDTA; 2) high speed centrifugation and Percoll separation of E. suis free in plasma; and 3) differential centrifugation of E. suis dissociated from hemolysates. Isolated E. suis were solubilized and analyzed by SDS-PAGE and Western blotting. With the first 2 methods, E. suis specific antigens were not readily observed. With the third method, frozen erythrocytes were thawed and centrifuged at 100,0009. The pellet was resuspended in PBS and sequentially centrifuged at 5009, 10,0009 and 100,0009. After each centrifugation, the pellet was removed and analyzed. A number of E. suis specific antigens were readily observed, especially in the 100,000g sediment. Antigen bands at 24, 47, 71, 79, 81, 90, 106 kd were consistently observed in western blots reacted with sera obtained from various pigs and at different post-infection times, and by autoradiography using an antigen prepared from 358 labeled E. suis. Some antigen bands were only recognized by sera collected during the late stage of infection. Sizes of the late appearing antigens and time recognized varied among the pigs. Host protein inclusion was least in the 100,000g sediment. 83 Introduction Swine eperythrozoonosis is caused by the rickettsial organism, Eperythrozoon suis. The disease is characterized by acute ictero-anemia and is associated with impaired weight gains and reproductive performance (Henry, 1979). Swine eperythrozoonosis has been reported world wide including North and South America, Africa, Europe and Asia (Henry, 1979; Hsu et al., 1990; Sisk et al., 1980; Smith, 1977; Smith, 1981). Prior to 1970, the diagnosis of swine eperythrozoonosis was based on herd and individual animal histories describing ictero-anemia, demonstration of Eperythrozoon bodies in blood smears, and the complement fixation test (Splitter, 1958). E. suis was considered primarily a pathogen of feeder pigs, causing acute icteroanemia, with low morbidity and high case mortality. Detection of the organism by blood smears provides a definite diagnosis; however, this method is complicated by difficulties in distinguishing E. suis bodies from artifacts in blood smears (Henry, 1979). The majority of E. suis infections are believed to be subclinical and parasitemia is only detected in acutely affected animals. Smith and Rahn (1975) developed an indirect hemagglutination assay (IHA) for measuring E. suis antibodies. IHA positive, non-parasitemic animals have been observed routinely and are considered to be carriers of E. suis. IHA has become the standard diagnostic test for 84 detecting E. suis antibodies. The IHA antigen is prepared by glutaraldehyde attachment of a partially purified antigen preparation made from the plasma of parasitemic pigs to sheep erythrocytes. Serial dilutions of sera are reacted with the treated sheep erythrocytes to determine the titer. A positive IHA titer is associated with IgM antibodies, correlated with anemia but not with development of detectable parasitemia, and indicates either a recent or persistent/chronic infection (Zachary & Smith, 1985). Hoffmann et al. (1981) classified the anemia in swine eperythrozoonosis as an autoimmune hemolytic anemia and autoantibodies against erythrocytes were thought to play an important role in anemia development. Anti-erythrocytic antibodies have been suggested by several observations such as spontaneous agglutination of infected blood especially at cold temperatures, increased sedimentation rate of infected blood (Doyle, 1932; Robb, 1943; Splitter & Williamson, 1950) and. positive ICoomb's tests (Hoffmann et al., 1981). Recently, Schmidt et-al. (1992) have isolated a cold agglutinin from erythrocytes of infected animals. Biochemical analysis using double immunodiffusion, polyacrylamide gel electrophoresis and western blotting with purified IgM and anti-pig IgM revealed that the isolated cold agglutinins were IgM antibodies. The associations of IgM with IHA reactive antibody, IHA titer with anemia development and anemia development with IgM cold agglutinin production, plus the likelihood that the IHA 85 antigen preparation contains host derived antigens, suggest the possibility that IgM antibodies detected in the IHA test may be directed at host erythrocyte antigens as well as E. suis antigens, thus potentially reducing test specificity. Furthermore, Smith (1981) reported that IHA negative pigs could be carriers because parasitemia was observed in IHA negative pigs after splenectomy, indicating that the test may also lack sensitivity for detecting chronically infected carriers. Recently, enzyme-linked immunosorbent assays (ELISA's) have been developed to detect E. suis specific antibodies and improve test sensitivity for detecting low grade infections (Hsu et al., 1992; Schuller et al., 1990). The coating antigens used in these assays were crude preparations from infected erythrocyte lysates and contained host erythrocyte antigens. Again, considering the autoantibody production induced by E. suis infection, the presence of host proteins in ELISA coating antigens may interfere with test results. The following studies were initiated to evaluate the specificity of the IHA for detecting E. suis specific antibodies and to develop and evaluate improved antigen preparations to facilitate improvements in serodiagnostic tests for swine eperythrozoonosis. 86 Materials and Methods Organism. E. suis was initially isolated from an inappetant, febrile boar housed at the Michigan State University Swine Research Center. A diagnostic CBC revealed that the boar was parasitemic and anemic. This E. suis isolate has been maintained by passage in splenectomized pigs. Infected blood has been cryopreserved by adding 6.7% and 10% (v/v) of 40% polyvinyl pyrrolidone (PVP) and glycerol, respectively, and freezing in liquid nitrogen. Animals and experimental infection. Cross bred pigs were 6- to 13-weeks-old at inoculation. All animals were splenectomized using standard surgical techniques 1 to 2 weeks prior to inoculation and determined to be E. suis noninfected by no evidence of any temperature elevation or development of parasitemia before inoculation. Ten milliliter of frozen infected blood was thawed in a 37 °C water bath and inoculated intravenously and/or intramuscularly into splenectomized pigs. Clinical signs and rectal temperatures were assessed daily for monitoring disease development. When the rectal temperature exceeded 40 °C, blood was collected in a vacutainer tube (Beckton Dickinson Vacutainer Systems, Rutherford, NJ) containing ethylenediamine tetraacetic acid (EDTA) and examined for parasitemia by Wright's stained thin smear. When the febrile phase lasted for more than 3 days or if an animal exhibited 87 severe depression, intramuscular treatment with oxytetracycline (20 mg/kg body weight; Liquamycin LA 200, Pfizer, New York, NY) was initiated to avoid mortality. Serum samples. Serum was harvested from blood obtained from experimentally infected pigs before inoculation and weekly thereafter. Serum samples (field sera) submitted to the Michigan Department of Agriculture for regulatory purposes were kindly provided by Dr. Richard Gatzmeyer. In addition, serum samples submitted to the Michigan Animal Health Diagnostic Laboratory for swine eperythrozoonosis testing were used. The immune reference serum was a pool of sera obtained from blood collected at 2 to 8 weeks post-infection from experimentally infected, splenectomized pigs. The nonimmune reference serum was obtained from blood of a known noninfected, intact pig. Antigen preparation. Antigens were prepared by four different methods: 1) Method IHA antigen: IHA antigen preparation method followed the procedure of Smith and Rahn (1975). Briefly, parasitemic blood was collected using a sodium citrate solution, and plasma was separated by centrifugation. Globulins were removed from the plasma by precipitation with saturated ammonium sulfate (1:1) and the precipitate was discarded. The supernatant was dialyzed against saline 88 borate buffer (pH 8.0) for 12 hours and concentrated to 1/5 of the initial plasma volume by Amicon ultrafiltration using an XM 50 filter (Amicon Division, W. R. Grace & Co., Danvers, MA). Albumin was removed from the concentrated preparation by precipitation with three times its volume of 0.1 M tris buffer (pH 8.0) containing 0.4% ethodin (6,9-diamino-2- ethoxyacridine lactate). The supernatant was again dialyzed and stored at -80 °C. Commercially available IHA antigen (Dr. A. R. Smith, University of Illinois, Urbana-Champaign, IL) was used as a positive control. 2) Method EDTA-Es (Isolation of E. suis dissociated from erythrocytes by EDTA): Blood collected into sodium heparin solution (30 U/ml) was centrifuged at 2759 for 20 minutes, and the plasma and buffy coat were removed. The sediment (erythrocytes) was mixed with phosphate buffered saline (PBS) containing 1% EDTA (EDTA-PBS), incubated for one hour at 37 °C with gentle. rocking, and centrifuged at 5009 for 20 minutes. The supernatant was collected and centrifuged at 20,0009 for 60 minutes at 4 °C. The resulting pellet was solubilized by 3 times freeze-thawing in Tris buffer (pH 8.0) containing 5 mM EDTA, 5 mM iodoacetamide, 1 mM phenylmethylsulfonyl fluoride, 0.1% Nonidet P-40 and 0.1% sodium dodecyl sulfate (SDS), and stored at -80 W2. 3) Method Plasma-Es; (Isolation of E. suis free in plasma): A) Blood collected into sodium heparin solution (30 U/ml) was incubated at 37 °C for one hour with gentle rocking and then centrifuged at 5009 for 20 minutes. Plasma was 89 collected and centrifuged at 20,0009 for 60 minutes at 4 °C. The resulting pellet was solubilized and stored at -80 °C. B) The pellet obtained by Method Plasma-Es-A, was mixed with isotonic Percoll (70% in final concentration; Pharmacia Fine Chemicals, Uppsala, Sweden) and centrifuged at 60,0009 for 15 minutes. Visible bands were collected and examined for the presence of E. suis by electron microscopy using phosphotungstic acid negative stain. 4) Method D'spin-Es (Isolation of E. suis tw' differential centrifugation of hemolysates): Blood collected into citrate-phosphate-dextrose solution containing 0.2% inosine was centrifuged at 5009 for 20 minutes, and the plasma and buffy coat were removed. Erythrocytes were washed 3 times with Eagle's minimum essential medium containing 0.2% inosine (EMI). After the final wash, erythrocytes were suspended in the same volume of EMI containing 5% fetal calf .serum (FCS). The suspension was mixed with 6.7% and 10% (v/v) of 40% PVP and glycerol, respectively, and stored at -80 °C. The frozen erythrocytes were thawed and centrifuged at 100,0009 for 40 minutes. A small aliquot of the pellet was solubilized and stored at -80 °C (designated as thaw antigen). The remainder of the pellet was resuspended in PBS and sequentially centrifuged at 5009 for 20 minutes and at 10,0009 and 100,0009 for 30 minutes each. After each centrifugation, the pellet was removed, solubilized and stored at -80 °C (designated as 500, 10K and 100K antigens, respectively). 90 Antigens were prepared either from infected (>75% parasitemia, designated as Es+, E. suis positive antigen) or noninfected blood (Es’, E. suis negative antigen) by each preparation Method. In addition, the 100K antigen described in Method D'spin-Es was prepared from the blood of one pig collected: 1) before splenectomy, 2) after splenectomy and before inoculation of E. suis, 3) at the first parasitemic phase, 4) after the first parasitemic phase during which parasitemia was not detectable, 5) at the second parasitemic phase, and 6) after the second parasitemic phase. The protein concentrations of the antigen preparations were measured by the Bio-Rad protein assay (Bio-Rad, Richmond, CA). Preparation of E. suis sensitized sheep erythrocytes (8- RBC) and glutaraldehyde treated sheep erythrocytes (G-RBC) for IHA. Five milliliter of sheep blood was collected with an equal volume of Alsever's solution and the erythrocytes were washed 3 times with PBS (pH 7.2). A half milliliter of packed erythrocytes were suspended in 9.5 ml of PBS in a 50- ml beaker and stirred on a magnetic stirrer. While stirring, 24.7 mg of IHA antigen was added, followed by dropwise addition of 1 ml of 2.5% glutaraldehyde. The mixture was stirred for one hour, then 1 ml of 2% ovalbumin was added and the mixture was stirred for another 30 minutes. S-RBC were washed 3 times and suspended in 9 times its volume of PBS and stored in liquid nitrogen. For use in the IHA, the thawed 91 10% S-RBC solution was diluted 1:20 with PBS. G-RBC were prepared identical to S-RBC, except no antigen was added. G- RBC was used as a 10% solution. Indirect hemagglutination assay (IHA). Several. IHA procedures with modifications of the Smith and Rahn (1975) method were performed: 1) Standard IHA: Test sera were heat inactivated at 56 °C for 30 minutes. To 100 pl of inactivated serum, 0.85 ml of PBS and 50 pl of the 10% G-RBC solution were added and the mixture was incubated for 15 minutes at 25 °C. The mixture was then centrifuged at 500 g for 10 minutes and 50 ul of the supernatant was transferred to a round bottom 96 well microplate. The transferred sera was serially diluted (two fold) with 1% heat inactivated rabbit serum in PBS, with a starting dilution of 1:10. The test antigen, 50 ul of 0.5% S-RBC solution, was added to each well. The plate was covered and incubated at.4 °C over night. The IHA titers on individual samples were interpreted as follows: 1:80 or more, positive; 1:20 or less, negative; and 1:40, suspicious. The IHA titers of duplicate samples were interpreted using geometric means of titer: 1:80 or more, positive; 1:20 or less, negative; and mean titers between 1:20 to 1:80, suspicious. 2) Modified IHA 1 (Assay at 37 °C): The assay was performed with duplicate samples by the same method as the 92 standard IHA procedure except the plates were incubated at 37 °C over night, after addition of S-RBC. 3) Modified IHA 2 (Preincubation of sera with pig erythrocytes): Blood was collected with a 0.15% EDTA solution from six noninfected pigs. The plasma and buffy coat were removed by centrifugation and the erythrocytes were washed 3 times with PBS. The erythrocytes of the 6 pigs were pooled. Inactivated test sera (150 pl) were incubated with the pooled erythrocytes (100 pl) either at 37 °C or 4 W: for 2 hours. After incubation, the mixtures were centrifuged and 100 pl of supernatant (pig erythrocyte absorbed sera) were tested, in duplicate, by the standard IHA procedure. Enzyme-linked immunosorbent assay (ELISA). In 96 well polystyrene ELISA plates (Immulon 2, Dynatech Laboratories Inc., Chantilly, VA), 2 pg of IHA antigen in 100 pl of 50 mM carbonate buffer (pH 8.5) was placed and incubated at 4 °C over night. The plates were washed once with saturation buffer (2% bovine serum albumin (BSA)-tris buffered saline (TBS) containing 0.5% Tween 20 (polyoxyethylene sorbitan monolaurate, pH 7.4) and blocked with 200 pl of saturation buffer for one hour. The plate was washed 3 times with saturation buffer and then incubated with 100 pl of serum, serially diluted by two fold with incubation buffer (0.5% BSA-TBS-Tween, pH 7.2) with a starting dilution of 1:4, for one hour at 25 °C. The plate was again washed 3 times with saturation buffer and incubated with horse radish peroxidase 93 (HRP) conjugated, affinity purified goat anti-pig IgM (100 times diluted in incubation buffer, Kirkegaard & Perry laboratories Inc., Gaithersburg, MD) for one hour at 25 °C. The plate was then washed 3 times with saturation buffer and 100 pl of 0.05 M phosphate-citrate buffer (pH 5.0) containing 0.01% Azino-bis (2,2'-azino-bis(3-ethylbenzthiazoline-6- sulfonic acid) (Sigma Chemical Co.) and 0.01% Hydrogen peroxide was added to each well. After 30 minutes of color development at 25 “C, the optical density of each well was read at 405 nm by an ELISA reader (EIA Reader Model EL 307, Bio-Tek Instruments Inc., Burlington, VT). Electrophoresis and western blot analysis of the antigens. The Es+ and Es' antigens obtained by each preparation method were mixed with sample buffer (100 mM Tris (pH 6.8), 3.3% (w/v) SDS, 16.7% (v/v) glycerol, 4.2% (v/v) 8- mercaptoethanol, and a few crystals of bromophenol blue) in 5 ml glass tubes and put into boiling water for 3 minutes. SDS-polyacrylamide gel electrophoresis (PAGE) was performed with these antigens using a 9% gel (Smith, 1984). The proteins in the gel were visualized using Coomassie brilliant blue stain. Proteins were transferred to nitrocellulose membranes (NC's) electrophoretically by the method of Towbin et al. (1979). The proteins transferred to the NC's were blocked with saturation buffer (same formula as ELISA procedure) at 4 °C over night. The NC's were incubated with sera, diluted 94 1:100 with incubation buffer (same formula as ELISA procedure), at 25 ’C for two hours and washed 3 times with the washing buffer, 0.2% BSA-PBS-0.5% Tween (pH 7.0), 10 minutes each wash. The NC's were incubated with either HRP conjugated goat anti-pig IgM (1:500 in the incubation buffer, Kirkegaard & Perry laboratories Inc.) or rabbit anti-pig IgG (1:2,000 in the incubation buffer; Sigma Chemical Co., St. Louis, M0) for another hour and washed 3 times. The NC's were finally incubated for two hours with triethanolamine buffered saline (pH 7.5) containing 0.05% 4-chloro-1-naphthol (Sigma Chemical Co.) and 0.01% hydrogen peroxide. E. suis specific bands were determined by comparing the bands present in western blots using combinations of Es+ and Es' antigens, and immune and nonimmune sera. Bands detected only in the western blots using-Es+ antigens and immune sera were considered to be E. suis specific antigen bands. Protein standards used for molecular weight (MW) markers (SDS-6H; Sigma Chemical Co.) consisted of bovine erythrocyte carbonic anhydrase (MW 29,000), ovalbumin (MW 45,000), bovine plasma albumin (MW 66,000), rabbit muscle phosphorylase B (MW 97,400), Escherichia coli B-galactosidase (MW 116,000) and rabbit muscle myosin (MW 205,000). The transferred MW standards were visualized by India ink stain (Hancock & Tsang, 1983). 95 35S-methionine radiolabeling of E. suis. Methionine deficient Eagle's minimum essential medium (pH 7.3) was prepared by dissolving powder (for 1 liter of medium) of L- leucine, L-lysine and L-methinonine deficient Eagle's medium (Sigma Chemical Co.) containing Earle's salts and L- glutamine, in 1100 ml of distilled water, supplemented with 0.36 w L-leucine, 0.36 w L-lysine, non-essential amino acids (for one liter medium, Sigma Chemical Co.), 25 mM N-2- hydroxy-ethylpiperazine-N'-2-ethanesu1fonic acid (HEPES), 0.1% (w/v) glucose, 0.2% (w/v) sodium bicarbonate, 0.2 mM 6- mercaptoethanol, 2.0 mM sodium pyruvate, 0.1 mM hypoxanthine, 0.016 mM thymidine and 0.2% (w/v) inosine. Infected (91% parasitemia) and noninfected blood were collected with sodium heparin solution (30 U/ml),.and plasma and buffy coat were removed by centrifugation. The erythrocytes were washed with the deficient medium 3 times. The infected and uninfected erythrocytes were individually mixed with the deficient medium (5% packed cell volume), put into 60 mm Petri dishes, and incubated for 30 minutes in a 5% C02 incubator at 37 °C. After an initial incubation to starve the organisms for methionine, the culture dishes were supplemented with 20 pCi/ml of 35S-methinonine (ICN Biomedicals, Inc., Irvine, CA) and 10% heat denatured FCS, and incubated for another 4 hours. After incubation, the medium was removed by centrifugation. The erythrocytes were washed twice with the deficient medium and suspended in the same volume of the deficient medium containing 10% FCS. The suspension was then 96 supplemented with 6.7% and 10% (v/v) of 40% PVP and glycerol, respectively, and stored at -80 °C. Es"’ and Es’ 100K antigens (antigen preparation Method D'spin-Es) were prepared from the frozen suspensions of infected and uninfected erythrocytes, respectively. Aliquots of the antigen preparations were removed, added to 5 ml of scintillation cocktail (High flash point cocktail, Safety-solve; Research Products International Corp., Mount Prospect, IL), and their radioactivity was measured with a liquid scintillation counter (Beta Trac; Tm Analytic, Elk Grove Village, IL). The protein concentrations of the antigen preparations were measured by Bio-Rad protein assay (Bio-Rad, Richmond, CA). SDS-PAGE was performed with radiolabeled and normal 100K antigens in the same gel to compare the protein bands that appeared by autoradiography and western blotting. After SDS- PAGE was completed, the gel containing MW standards and radiolabeled antigens was cut, stained by Coomassie brilliant blue stain to visualize MW standards, and dried on paper using a gel drier (Bio-Rad). A direct autoradiograph was performed using Kodak X-omat RP film, 10 days exposure at 25 °C. Simultaneously, western blotting was performed with the other portion of the gel containing non-radioactive antigens, using the immune and nonimmune reference sera. Statistical analysis. Statistical analysis was performed using the computer program StatView (version 4.0, Abacus Concepts, Berkeley, CA). The paired sign test was used for 97 comparing the results of the standard IHA and modified IHA's; P values less than 0.05 were considered significant. 98 Results STUDY I. EVALUATION OF IHA The effect of incubation temperature on IHA results. The effects of incubation at 37 "C (modified IHA 1), instead of 4 °C (standard IHA), are presented in Table 3-1. With incubation at 37 °C, the titers of positive and suspicious sera tended to change (either decrease or increase) although the results ‘were not statistically significant. A significant number of negative sera exhibited an increased titer. The magnitude of titer change in most sera was no more than two fold. Notably, 3 positive sera exhibited more than a two-fold reduction from the standard IHA titer when Table 3-1. Alteration of indirect hemagglutination assay (IHA) titer of sera incubated at 37 °C instead of 4 °C used in the standard IHA titer. Change in titer by incubation at 37°C Serum 97009. Total No. Decreased No. Increased No. Unchanged Positive 14 4 (3)1 7 3 Suspicious 10 5 (2) 2 3 Manet 24 4 15 (1) 5 a- - Serum samples were grouped by the interpretation of the standard IHA titer (incubating at 4 °C). 1 - The parentheses show the number of serum samples exhibited more than two-fold change In the titer. In Significant difference (P=0.0192) in the standard titer and modified IHA 1 titer was observed by paired sign test in the negative serum group. 99 incubated at 37 °C. The titers of these 3 samples were 1:80, 1:80 and 1:640 in the standard IHA and 1:20, 1:10 and negative (<1:10) in the modified IHA 1, respectively. The effect of preincubation of serum with pig erythrocytes at different temperatures. Table 3-2 describes the alteration of titers by preincubation with pooled pig erythrocytes at 4 and 37 °C (modified IHA 2). Titers tended to decrease with preincubation at either temperature. Titer reduction with preincubation at 37 °C was greater, especially in the positive serum group than with preincubation at 4 °C (37 °C - 9 of 14 vs 41°C - 5 of 14). A significant number of samples in the positive serum group exhibited a reduction in titer with preincubation at 37 °C when compared to the standard IHA test. IHA using Es+ and Es‘ antigen preparations. Three serum samples were collected from one pig before infection, 15 and 22 days post-infection and tested by the standard IHA using various Es+ and Es‘ IHA antigens (Table 3-3). Pre-infection serum was IHA negative regardless of the antigen preparation. IHA titers of each post infection serum were similar regardless of using Es+ antigen preparations, P2, P3 or P4, or E8“ antigen preparations, N1 or N2, and differences were within a two fold dilution. The titers of each post- infection sera measured using the commercial antigen (P1) were higher than those measured using the prepared antigens. 100 .Ammoodudv Enemies. we; 955 529. 0.6.wa 2: c. u. em as 2382.83 53338.05 5.2. .33 ~ (I. morgue... one .3.» <1. 3%:on 2: c. 353:6 er... I. e .35 9: c. encore So. 025 a 55 So... ooebeo N5 3.9.8 Sam *0 .353: 05 2.05 3355.2. 2.... u .. .Amouxooczubo 9d 5.3 533350... 05 anew canteen 05 5.3 2:332: ma .3: <1. .2393 :2.» 50 cos—3298:. o...» .3 cod-65 82... 3.4.53 823 I . : . E v . a o. . s cm casualzl e . o . E n v w a a o. 8228...... m o 3 a. a a o m 3 2.28.. lumacafica .62 .9888... .02 3.8.80 82 892.8: .62 3829.. 62 83280 62 .98 anew 52am 0. 83:33:82.. .5 t... 5% 0.438238%... .35... c. sunfloll .0. mm new 0. e um 4.3.025»... 9d 5.2. “33338.05 Eon... go .3: 2:: >33 5.3555989. 32.9.. .6 con—23.... .~-m 2me 101 Table 3-3. IHA titers of three serum samples using different antigen preparations. DPI of Antigen preparations serum sample“ P1 I P21 P3* P41 N15 N25 0 negative negative negative negative negative negative 1 5 _ 1 :2560 1 :320 1 :320 1 :640 1 :640 1 :640 22 1 :1280 1:320 1 :160 1 :320 1 :320 1 2160 Number of days post-infection serum Commercial antigen (University of Illinois) - Es'l' antigens prepared from experimentally infected animals - Es" antigens prepared from uninfected animals mid-'4'! ELISA using.Es+ and Es‘ IHA antigen preparations. The IgM ELISA using the IHA antigens were performed to evaluate IgM binding to the antigens. The ELISA optical density (OD) values of pre- and post-infection sera of 2 pigs, using various Es+ and Es' IHA antigen preparations are shown in Figure 3-1. The post-infection sera showed higher OD values than the pre-infection sera, regardless of the antigen preparations, except for the ELISA using P1 antigen with the sera of pig number 2. The Es+ and Es‘ antigens showed a similar pattern of OD value changes by serum dilution, except for the P1 antigen preparation. The back ground values of the P1 antigen preparation (ELISA with pre-infection sera) were higher than those of the other antigen preparations. 102 Figure 3-1. Changes in optical densities by serum dilution in IgM ELISA.using various Eperythrozoon suis positive (Pl-4) and negative (N1-2) antigen preparations for indirect hemagglutination assay (IHA). A and D: The optical density (OD) of ELISA using post-infection sera of pig number 1 and 2. B and E: The ELISA OD using pre-infection serum of pig number 1 and 2. C and F: The difference of OD of pre- and post-infection sera of pig number 1 and 2, rescectively. Pl - E. suis positive antigen obtained from University of Illinois. P2-4 . E. suis positive antigens prepared from experimentally infected pigs. Nl-2 - E. suis negative antigens prepared from uninfected pigs. A. Post-infection serum 0.4 - B. Pre—inlection serum 0.4 - q .0 a: J o to I 1 Optical density .0 d I O C. Difference 0.3 - Optical density 4 16 32 64 128 256 Serumdilution 103 D. E. Pig 2 Post-infection serum 0.4 — q 0.3 d 0.2-1 d 0.1 - O Pro-infection serum 0.4 - 0.3- N '1 0.2 - 4 816 32 64128256 Serumdilution -.- +P3 +P4-B—N1-G—N2 - Figure 3-1. 104 SDS-PAGE and western blot analysis of IHA antigens. SDS- PAGE of various IHA antigens stained by Coomassie brilliant blue are depicted in Figure 3-2. Although there were differences in band intensity, the protein band patterns in SDS-PAGE of Es+ and Es' antigens were nearly identical. Es+ antigen preparation specific bands could not be clearly identified. Intense, wide bands were observed at the 80 to 85 kd area and 65 to 75 kd area in each antigen preparation. Inconsistent, lighter bands were also observed in each antigen preparation. The P1 antigen showed an additional wide band at 60 kd, which was detected as a thin band in other antigen preparations. In IgM western blots, almost identical bands were observed in Es+ (P2-4) and E8" (N1-2) antigens reacted with immune sera; however, they lacked a 60 kd band observed with the P1 antigen preparation (Figure 3-3 A). The 60 kd protein band in the P1 antigen was detected as a dense wide band in blots reacted with both immune and nonimmune sera. The intensity of all bands observed was greater in the blots using the immune reference serum than those using the nonimmune reference serum. In IgG western blots, the band patterns were again similar in all antigen preparations; however, the P1 antigen had denser bands, especially the 60 kd protein band was observed strongly in the western blots with immune and nonimmune reference sera (Figune 3-3 B). Also in blots with no serum applied, the intensity of the 60 kd protein band was similar to the immune and nonimmune sera 105 P1 92 p3 P4 N1 N2 IMVVM kd 205 '— 116 - 97A—-— 66 -— 45— 29 ‘- Figure 3-2. Comparison of the protein composition of various Eperythrozoon suis positive (Es+) and negative (Es‘) antigen preparations for indirect hemagglutination assay (IHA) by SDS-PAGE stained by the Coomassie brilliant blue. Each lane of a 9% gel was loaded with 40 pg of reduced conditioned antigen preparation. P1 = Es+ antigen preparation obtained from the University of Illinois, P2-4 = Es+ antigens prepared from experimentally infected pigs, N1 and 2 = Es' antigens prepared from uninfected pigs. MWM = Molecular weight marker. 106 Figure 3-3. Western blot analysis of indirect hemagglutination assay antigen preparations. A: western blots using 40 pg of Eperythrozoon suis positive (EsT) and negative (Es’) antigens with 1:100 diluted immune and nonimmune reference sera, detected by horse raddish peroxidase (HRP) conjugated anti-pig IgM and 4-chloro-l-naphthol. B: western blots using 40 p9 of Es+ and Es' antigens with 1:100 diluted immune and nonimmune reference sera and without serum, detected by HRP conjugated anti-pig IgG and 4-chloro-1-naphthol. Lane Pl: Es+ antigen obtained from University of Illinois. Lanes P2-4: Es+ antigens prepared from experimentally infected animals. Lanes N1 and 2: Es' antigens prepared from uninfected pigs. 107 A. Immune serum Nonimmune serum 1- 0) V v- N MWM mEmazz £88.12? 1 .. BI Immune serum Nonimmune serum No serum N N V N N N MWM Eammiz EEEEEZ {$831252 kd ' ' 7 205-—- 116 - 97A.- 66 .— 29 _ Figure 3-3. 108 reacted blots. As observed in the IgM blots, the band intensity was greater in the blots using immune reference serum.than those using nonimmune reference serum. In summary, E. suis specific bands were not readily detected in the Es+ IHA antigen preparations either by SDS- PAGE, or IgM or IgG western blotting. STUDY II. EVALUATION OF ANTIENS PREPARED BY THREE DIFFERENT METHODS (Msmaoos EDTA-Es, PLASMA-Es mo D'SPIN-Es) Antigens prepared by Method EDTA-Es. IgG western blots using Es+ and E3" antigens prepared by Method EDTA-Es and immune and nonimmune reference sera are presented in Figure 3-4. Several possible E. suis specific antigens were detected as faint bands at 50 to 55 kd and 25 kd in the Es+ antigen/immune serum lane. However, the predominant protein bands detected were observed both in the Es+ and ES' antigen preparations and were considered to be host derived proteins. Method EDTA-Es was also unacceptable because the procedure resulted in severe hemolysis during the incubation of parasitized RBC's with EDTA-PBS. Also, approximately half of the samples became gelatinous during incubation and no supernatant could be obtained from the preparation following centrifugation. Therefore, clear supernatant containing free E. suis could not be readily obtained. 109 Serum: Immune Nonimmune Antigen: 53+ Es’ Es+Es- MWM _. "‘1' k. f 205 — J 116 _ 97Jl-— ~ I 66 — L1, ”1 l 45 __ ! Figure 3-4. Western blot analysis of Eperythrozoon suis positive (EsI) and negative (Es') antigens prepared by Method EDTA-Es. The blots of the Es+ and Es' antigens (8 p9 in each lane) were incubated with the immune and nonimmune reference sera and detected by horse raddish peroxidase (HRP) conjugated anti-pig IgG and 4-chloro-1- naphthol. MWM = Molecular weight marker. 110 Antigens prepared by Method Plasma-Es. IgG western blotting was performed using Es+ and Es‘ antigens prepared by Method Plasma-Es-A with pre- and 2, 4 and 8 weeks post- infection sera of experimentally infected animals with E. suis (Figure 3-5). A small number of bands including MW 160, 54 and 26 kd antigens were observed in the western blots with the pre-infection sera. In addition, a number of bands were detected by post-infection sera, especially at 4 and 8 weeks. Intense bands were detected at MW 104, 78 and 40 kd. However, most of the bands observed in the blots using the Es+ antigen preparation with post-infection sera were also observed in blots using Es' antigen preparations reacted with post-infection sera. Consequently, most of the bands detected were determined to be host derived proteins. Clearly defined E. suis specific bands were not readily apparent except for a possible E. suis specific band at MW 35 kd. Percoll separation of the antigen pellets was then attempted (Method Plasma-Es-B). Gradient centrifugation resulted in three visible layers with the Es+ antigen pellet and one with the ES' antigen pellet. However, clearly defined E. suis particles were not observed in these layers by electron microscopy (data not shown) and western blotting was not attempted. 111 .uuxuea ucmwoa usasuuaoz u as: .HonunmsciaiouoHcoiv one omH memiwvcs ooueosncou .mmm. unsofixouum nuances mmuoc an convenes whoa seaweeds one .m Henson med scum couuoaaoo uoc ue3 aspen downpoucw upon execs N was .Aucueu on puma Bonn. downpouce upon execs m use v .N as one .cowuuouce museum cuuooaaoo mama nuamesu sauce v use .uwse .N mafia poem cououuce haaeucoaauumxu v scum nose eoameso aspen e one AH. Sauce codename» unease cascade coaua one ..m. sauce on new: essences“ one nuance case use ouo3 ecceususmoum demands Inn one +uu on» no cacao one .uuisausam comes: an cousmuum commends .qu. o>flusmmc one .+uu. u>eveuom menu cooNoueuxuuQfl no uwuaaecs scan assumes .mIm ousmem I. heeemee...E... . e T... 1.. _ ..._ .E.» .. . g . . _ .._. I... r ._ . . _ _ _ 1_. :2; raj... : w... W# . rims . r .. I... . h ._ _r.- _ _ let. 5 H .r _ I r T fi.l ” f I . I .IIo-P .. . _ I . I ..... r .. _ , Ir limom .. . I -31: _ 3...... —.F. ..a—....-... ._u—H_.—. .. —H _.-_ ..aril . I.. —. 2;: so... as... «a... B... _m so... no... «en. B... _m 59.5 Imw ceases +mm 112 Antigens prepared by.Method D'spin-Es. The thaw, 500, 10K and 100K antigens prepared from infected and noninfected blood were compared by IgG western blotting. A number of E. suis specific antigen bands, detected only in the western blots using the Es" antigen preparations with immune reference sera, were observed in each Es+ antigen preparation (Figure 3-6). Strong E. suis specific bands were detected at MW 23, 24, 37, 47, 50, 71, 79, 81, 101 and 106 kd in Es+ thaw, 10K and 100K antigen preparations. The 500 antigen preparation lacked some of these E. suis specific antigens. The 23, 24, 37, 47, 77, 79, and 106 kd antigens appeared most dense in the Es+ 100K antigen. In addition, E. suis specific bands at MW 36 and 77 kd were detected strongly only in the 100K antigen preparation. A band at 90 kd was detected in both Es+ and Es’ preparations of thaw, 500 and 10K antigen preparations. In 100K antigen preparations, however, the band at 90 kd was only detected in the Es+ preparation (shown by arrow). Several host derived antigens, determined by the band appearance in the ES' antigen preparations, were also observed in all preparations. The dense host derived antigen bands were detected at MW 200 (doublet) and 26 kd. However, the 200 kd antigens were found to be less intense in the 100K antigen preparation compared to the 500 and 10K antigen preparations. 113 Immune serum Nonimmune serum Es+ Ag Es' Ag Es+ Ag Es‘ Ag 3 x 3 x a! X M X MWM .5§3§5§.3§_ kd - t. .1 . " ’_ ' SEC L- ."I —— w l \..-.-- Figure 3-6. Western blot analysis of various antigens prepared by Method D'spin-Es. The blots of 40 pg of each Eperythrozoon suis positive (Es+) and negative (Es‘) thaw, 500, 10K and 100K antigen preparations were incubated with the 1:100 diluted immune and nonimmune reference sera. The antigens were detected by horse radish peroxidase (HRP) conjugated rabbit anti-pig IgG and 4-chloro-l-naphthol. MWM = Molecular weight marker. An arrow shows a 90 kd antigen observed in 100K antigen preparation. 114 Host protein concentration appeared to be the most reduced and the greatest number of E. suis specific bands could be recognized in the 100K antigen preparation. STUDY III. 100K ANTIGEN PREPARATION western blots with various sera. Western blots using Es+ and Es' 100K antigen preparations with pre- and post- infection sera obtained from 5 pigs are shown in Figure 3-7. E. suis specific antigen bands readily detected in Study II were consistently observed in all western blots using the Es+ antigen preparation with post-infection sera. Specifically, antigen bands at molecular weights of 23, 24, 37, 47, 71, 90 and 106 kd were strongest. western blots using antigens prepared at different stages after experimental infection. Figure 3-8 shows the western blots using different 100K antigens prepared from blood collected at 6 different times (prior to infection, during parasitemia and during remission), with the immune and nonimmune reference sera. During the first parasitemia, only the 24, 90 and 106 kd E. suis antigens, were detected in the antigens prepared from parasitemic blood. During the second parasitemia, other E. suis antigens (23, 37, 47, 71, 79 and 81 kd) were observed in addition to the 24, 90 and 106 kd E. suis antigens. 115 .uuxuse cameos usacooao: u as: ..ucmwu cu puma Eouu. cowuomudw uuom execs v one N us one .cOAHOOucM mucuon ouuuwaaoo one: sumo mummy one .meQ vcmnuuuwc mbfiu Esau some when some» one ..2. sumo moccH6uuu mcsEEHcoc .AHV shun Oucmuuuou seesaw .Am. Essen o: and: condensed cue: udoeuuusmmum cumflucs KooH Alan. O>ausmuc one .+uuv ubevenom mesa .w anon uo muoam .mwnu accuoueuummm cues amen covuwude haasvcoawuomxu n Home anon ocean demands zooa on» no weekends uoHn spouses .nin ousmwh , .. _ L. _ . __ _ m z_m men. seamen. we“. 3... 25 comes... 1mm. 59.5 +mm 116 Immune serum Nonimmune serum . N ~ .. Ag. ‘- N v- E N O. P N P O. E z z Ii 2 o. z z 2 (L z E z IWVWM ” ' , kd .... 205 -—- ‘“"’ 116 -- .I‘ 97JI“' "’ -' ‘ .3.»- -— -— ss— «HI-"MTV I . be; .2 ' i 's I" I I —-—— ”CI-O M__—--__ . I)“ . ' . 3'41"!” 3" ' I. Figure 3-8. Western blot analysis of the 100K antigen prepared at the various times before or after infection. The blot of 40 p9 of each antigen preparation was incubated with the 1:100 diluted immune and nonimmune reference sera and detedted with horse raddish peroxidase (HRP) conjugated rabbit anti-pig IgG and 4-chloro-l- naphthol. The antigens used were prepared from blood collected before splenectomy (N1), after splenectomy and before infection (N2), at the first parasitemic phase (P1), after the first parasitemic phase (NPl), at the second parasitemic phase (P2), and after the second parasitemic phase (NP2). MWM = Molecular weight marker. 117 SDS-PAGE and autoradiography using 35S-methionine labeled antigen preparations. The autoradiograph of radiolabeled antigen preparations and western blots of normal antigen preparations are compared in Figure 3-9. Twenty five protein bands were detected in the radiolabeled Es“ antigen preparation (lane 3), while only two light bands were detected in the radiolabeled Es' antigen preparation (lane 4). The protein bands at MW 24, 47, 66, 71, 81, 90, 101 and 106 kd observed in the autoradiograph of the Es" antigen preparation (shown by the left pointing arrows) were also detected in the western blot of the Es+ antigen preparation reacted with the immune reference serum (lane 1), but not in the other western blot lanes (lanes 2, 5 and 6). Therefore, the 8 bands at MW 24, 47, 66, 71, 81, 90, 101 and 106 kd were judged to be E. suis derived antigens. In addition, a band was observed at MW 37 kd in the autoradiograph of the Es+ antigen preparation (shown by the right pointing arrow). A similar band was not detected in the western blots of this experiment. However, a 37 kd E. suis specific antigen was consistently observed in previous western blot experiments. Therefore, the band observed at MW 37 kd in the autoradiograph was also judged to be E. suis derived. western blots with different post-infection sera of the same pigs. For this experiment, eight E. suis specific antigens were selected, according to their appearance in the autoradiograph of radiolabeled antigen preparations and 118 Figure 3-9. Comparison of the autoradiograph of radiolabeled 100K antigen preparations and western blots of the 100K antigen preparations. Lanes 1 and 2 show the western blots of the Eperythrozoon suis positive (Es+) and negative (Es‘) antigen preparations (50 pg each) incubated with the immune reference sera. Lanes 3 and 4 show the autoradiograph of the sulfur 35 labeled Es+ and Es‘ antigen preparations (30 pg each). Lanes 5 and 6 show the western blots of the Es+ and Es' antigen preparations (50 p9 each) incubated with the nonimmune reference sera. MWM = Molecular weight marker. Left arrows show 101, 101, 90, 81, 71, 66, 47 and 24 kd proteins. Right arrow shows 37 kd protein. 119 various western blots previously described. The selected bands were recognized at MW 24, 37, 47, 71, 81, 90, 101 and 106 kd. IgG western blotting was performed with the Es+ and Es‘ 100K antigen preparations using sera obtained weekly from 4 experimentally infected animals up to 21 weeks post- infection (WPI). Table 3-4 summarizes the band detection in western blots using the Es+ and Es' 100K antigen preparations reacted with various pre- and post-infection sera. All E. suis specific antigens (observed only in the Es+ antigen preparation) were detected with all post-infection sera except in those instances reported in Table 3-4. In addition, the pre-infection sera of 2 pigs recognized the 71 kd antigen. In the ES' antigen preparation, bands were observed at MW 71 and 90 kd, the same position as the E. suis specific antigens. The 71 kd antigen in the Es" antigen preparation appeared as a rough broad band. western blots using post-infection sera collected weekly from one pig are shown in Figure 3-10. All 8 E. suis specific bands were detected as intense bands by post- infection sera collected at either 1 or 2 WPI through 15 WPI except for the 14 WPI serum. In addition to the 8 E. suis antigens, several other antigen bands (shown by arrows) were recognized only by sera obtained during the late stage of infection. Late appearing antigen bands were also observed in the other 3 pigs; however, antigen sizes and time recognized varied among the 4 pigs. 120 Table 3-4. Eperythozoon suis specific bands detected in the western blots using 100K E. suis positive (Es+) and negative (Es') antigen preparations with various pre- and post-infection sera from 4 pigs. MW of the band 24 37 47 71 81 so 101 106 Detection in Es+ antigen preparation Detected by all sera from 1-2 up to 21 WPI, except for 14 WPI sera from one pig Detected by all sera from 2-3 up to 21 WPI, except for several post-infection sera from 3 pigs Detected by all sera from 1-2 up to 21 WPI, except for 14 WPI sera from one pig Detected in pre-infection sera of 2 pigs, and all 4 pigs from 1-2 up to 21 WPI, except 14 WPI sera from. one pig Detected by all sera from 2-3 up to 21 WPI, except for 14 WPI sera from one pig Detected by all sera from 1-3 - up to 21 WPI, except for 14 WPI sera from one pig Detected by all sera from 2-3 up to 21 WPI, except for 14 WPI sera from one pig Detected by all sera from 1-2 up to 21 WPI, except for 14 WPI sera from one pig Detection in Es’ antigen preparation Notcbtected Notdetected Notdetected Detected by pre and post- infection sera of 3 pigs‘ Not detected Detected by 2-3 up to 16 WPI sera of 3 pigs Nadetected Notdetected :- = Observed as rough broad bands 121 .wause pcmHms usHsocHoz u :3: .cOHuomucH ousH mcHusc omuouHHoo show an tho owuHcmoomu mcmeucs 30cm w3OHH< .A=OHuooucH upon nx003 Ho Hones: an eBOcm. cOHuowucH umom axow3 mH on H as UODOOHHOO sumo use ...H.m. cOHvowucH cucuon couOOHHoo sumo ..z. aspen occupancy ocsEEHcoc on» ..H. sauce moccuuuuu ucsEEH on» .Am. Epsom Os cuH3 cousnsucH uuw3 msOHHeusmuum cmmHucs .1 mm. 0>Husmoc one .+mH. u>HuHuom swam :OONOhcuANmQN on» HO auoHn use .mHm ouuumucH hHHsucwfiHuomxo cm EOHu thmmk owuowHHoo summ msHms coHuoummwum somHucs MOOH uo mHthscc uoHn cumumOR .oHI m wusmHh . .. . _._....._:.._;:3...nua . . ~ __ . 1W .. t. . ..x . J In l.|..II.|-.II..III.0... ._. II “lulmmmmmumm _- m“? i‘ I ILUHHI/H . 1,1. .. .. f... I wN . / Z W . . - :.--:---2.::_ ,_.. miamaa1lslm an. in-.. I .n I m... _~.m I I I ll .7 a r . 2. . . .mmm mmiim I. 18 I. “Hull IIIHMI II I IIIMMI. I ""'..l E IV.NQ llII .II II "I, I0: uIJILIuflflunflflmmmr l . . , @_.Im8 . I ox P. . . . LII. HmmHNBBLSSVCZIBB NIB mflmmHMSBLSStCZImE NIB 2;: comzcuimm comzcm+mw 122 Discussion The inclusion of host erythrocyte proteins in antigen preparations has been a common cause of false positive (poor specificity) results in immunoassays with various hemoparasitic diseases such as anaplasmosis (Barbet et al., 1983; Budden &'Dimopoullos, 1977; Palmer et al., 1986) and babesiosis (Callow & Dalgliesh, 1982; Goodger et al., 1985). One possible reason for this poor specificity is the reaction of antigen impurities with antibodies in sera (Goodger & Mahoney, 1974). Anti-erythrocytic antibodies in the form of cold agglutinins have been observed in hemoparasitic infections such as eperythrozoonosis, haemobartonellosis, anaplasmosis, malaria and babesiosis and are associated with anemia (Bellamy et al., 1978; Oki & Miura, 1970; Schroeder & Ristic, 1965; Soni & Cox, 1974; Soni & Cox, 1975; Thoongsuwan & box, 1973; Zachary & Smith, l985; Zulty & Kociba, 1990). A majority of these cold agglutinins are IgM antibodies directed at glycoproteins (mainly I antigen in humans) located on the erythrocyte membrane (Lau & Rosse, 1975; Pruzanski & Shumak, 1977a; Pruzanski & Shumak, 1977b). In addition to cold agglutinins, isoerythrocyte antibodies have been frequently observed in experimental infection or vaccination systems usinginfected blood as an inoculum (Barry et al., 1986; Dimmock, 1973; Duzgun et al., 1988). The role of isoerythrocyte antibodies transferred via colostrum from dams vaccinated with lyophilized A. marginale 123 infected erythrocytes has been proposed in neonatal isoerythrolysis of calves (Ristic, 1976). Study I examined the role of cold agglutinins in IHA by comparing the standard IHA to several modified procedures. With the modified IHA 1 (incubation at 37 °C), 77% of the sera (37 of 48 tested) showed either an increase or decrease in titer compared to their respective standard IHA titers (Table 3-1). A significant number of samples in the negative serum group showed an increased titer with the modified IHA 1. However, the magnitude of the titer change of most samples was no more than two-fold. More than two-fold titer changes were observed mainly in the positive and suspicious sera, which exhibited marked decreases in titers with the modified IHA 1. The decrease in titers observed with positive sera indicates an active role of cold temperature reacting antibodies in IHA and demonstrates that the incubation temperature used in the standard IHA (4 °C) needs to be maintained to provide consistent results. The modified IHA 2 (preincubation of sera with normal pig erythrocytes) was designed to evaluate the role of autoantibodies to erythrocyte antigens in IHA. Two different temperatures were used for preincubation of sera in order to evaluate the temperature dependency of the autoantibodies. A significant number of positive sera (9 of 14) exhibited a decreased titer with the modified IHA 2, preincubation at 37 °C, compared to their respective standard IHA titers. In addition, 5 of 14 positive sera showed a decreased titer with 124 the modified IHA 2 with preincubation of sera at 4 °C (P value in the paired sign test was 0.0625). These results indicated that the IHA titers of positive sera decreased with preincubation of sera with pig erythrocytes regardless of the preincubation temperature and suggested that temperature independent autoantibodies also play an active role in IHA. The role of autoantibodies in the IHA was further evaluated in subsequent experiments using Es+ and Es‘ antigen preparations. The observation that few differences were found in IHA and in IgM ELISA when using Es+ (P2-4) or 83' (N1 and N2) antigen preparations indicated that the main antibodies detected in IHA and ELISA were autoantibodies to erythrocyte antigens. This finding was further confirmed by failure to detect E. suis specific antigens by western blot analysis of the Es+ and E3' IHA antigens. In the western blot procedure, antigen preparations were boiled in 308 containing buffer. Consequently, the proteins were presumably, denatured and may have lost important conformational epitopes. Previously, Shankarappa and Dutta (1989) could not demonstrate protein specificity with a monoclonal antibody to Ehrlichia risticii by western blotting and they concluded that 'heat d'enaturation sufficiently altered the specific epitope, making it nonreactive with the monoclonal antibody. The presence of conformational antigens in the antigen preparations could not be evaluated by western blot analysis. However, antigens with altered conformation could be potentially detected as differences in protein 125 composition in SDS-PAGE analysis between the Es+ and Es" antigen preparations. In this study, no bands were detected specifically in the Es+ antigen preparations compared to Es- antigen preparation by SDS-PAGE analysis. This finding indicates that the Es" antigen preparations contain insufficient E. suis specific antigens to be detected by SDS- PAGE, and that the majority of proteins and antigens are host derived. Interestingly, the P1 antigen (obtained from the University of Illinois) contained an exceptionally high level of 60 kd protein. The 60 kd band in the P1 antigen was observed with equal intensity in IgG western blots reacted with immune and nonimmune sera, and also in western blots reacted with no serum. The strong band appearance in western blots with no serum indicated that the 60 kd protein reacted with anti-pig 196, thus suggesting that the 60 Rd protein was a part of IgG, possibly the heavy chain portion based on the protein molecular weight. This observation alternatively suggests a possibility of the interference in test results by the IgG (either .E. suis specific antibodies or erythrocyte autoantibodies induced by E. suis infection) inclusion in IHA antigen preparation. For example, hemagglutination could occur due to bridging E. suis specific IgG present in the IHA antigen preparation by E. suis bodies present free in serum obtained from parasitemic pigs, which could increase titer levels and test sensitivity, complicating interpretation of true E. suis IgM levels/seroconversion. Alternatively, 126 hemagglutination could result from bridging of erythrocyte autoantibodies present in IHA antigen preparation by erythrocyte antigens present in serum, possibly reducing test specificity. The 60 Rd band was also observed in IgM western blot using the P1 antigen with immune and nonimmune sera, although the band intensity was reduced. This observation could be interpreted as a cross reaction between anti-pig IgM and the 60 kd protein. This cross reaction could also explain the high background values observed in the IgM ELISA using the P1 antigen. 1n: the IHA experiment using various antigen preparations, the P1 antigen yielded 2-3 fold higher titers compared to other antigens.' The differences observed in SDS— PAGE between the Pl antigen and other antigens may explain the titer differences. If one assumes that the 60 kd protein is 196 heavy chain, this protein is not ”a functional protein” for the agglutination reaction. Sheep erythrocytes were coated with the same amount of each IHA antigen preparation based on the total weight of proteins. Since the P1 antigen preparation contains an exceptional amount of the 60 kd protein, the relative amount of "functional proteins" in the P1 antigen used for coating sheep erythrocytes would be lower than the relative amount of "functional proteins" in other antigen preparations (P2, P3, P4, N1 and N2). If it is assumed that the amount of "functional proteins" in the P1 antigen was optimal for the agglutination reaction, and that 127 the amounts of "functional proteins" in the other preparations were greater but sub-optimal for the agglutination reaction, IHA using P1 antigen preparation may have yielded higher titers than IHA using other antigen preparations. Isolation of Eperythrozoon bodies has been attempted with E. ovis and E. suis. Lang et a1. (1986) described successful separation of E. ovis bodies from hemolysates. Following this method, Hall et al. (1988) failed to completely separate E. suis bodies from erythrocyte antigens by lysing erythrocytes. Alternatively, they used EDTA to release the organisms from erythrocytes, and then separated the free organisms from erythrocytes by low speed centrifugation. The organisms were then concentrated by high speed centrifugation, yielding a high concentration of E. suis bodies. In the first experiment of Study II (Method EDTA-Es), the method of Hall et al. (1988) was applied to separate E. suis and found to be inadequate for purifying the organisms. In addition, severe hemolysis and the formation of a gelatinous precipitate of parasites and erythrocytes occurred during the incubation of the parasitized RBC's with EDTA-PBS, making this method unacceptable for separating E. suis from blood components. The cause of the severe hemolysis may be explained by an increased fragility of the erythrocytes due to E. suis infection. Increased fragility of infected erythrocytes has also been observed in babesiosis (Wright, 1973) and haemobartonellosis (Maede, 1980). In 128 Haemobartonella felis infected erythrocytes, increased osmotic fragility was found to be related to decreased erythrocyte lipids concentration (Maede, 1980). The cause of the gelatinous precipitation of infected erythrocytes and the EDTA-PBS mdxture during the incubation was not determined although it may be related to activated platelets that remained after the washing step. The activation of platelets during parasitemia has been observed in E. suis infected pigs (Plank & Heinritzi, 1990; Zachary & Smith, 1985) and E. ovis infected sheep (Overas, 1969). In the second experiment of Study II (Method Plasma-Es), an attempt was made to isolate E. suis free in plasma. However, the antigen preparation contained excessive host blood components, mostly platelets as determined by light microscopy, and clear separation of E. suis was not achieved. The source of the various bands observed in the western blots was not clear, but some of them may have been derived from platelet receptor proteins and also 196 bound to platelets. In the subsequent Percoll separation procedure, E. suis bodies could not be detected by electron microscopy in the visible layers formed, indicating Method Plasma-Es was not sufficient for separating E. suis free in plasma from plasma components. In the final experiment of Study II (Method D'spin-ES), E. suis isolation was attempted from infected erythrocyte lysates. Each pellet of the differential centrifugation procedure contained various E. suis specific antigens. The 129 observation that the 100K antigen preparation contained the least amount of host antigens and the most variety of E. suis specific antigens indicated that differential centrifugation concentrated E. suis antigens effectively. Method D'spin-Es provided another advantage that frozen infected erythrocytes could be used. In Methods EDTA-Es and Plasma-Es, parasitemic pigs were required as fresh infected blood was used. With Method D'spin-Es, infected erythrocytes could be stored frozen and the antigen could be prepared or the antigen preparation method could be modified at any time. In Study III, the 100K antigen preparation was further evaluated for E. suis specificity by various methods. western blots using sera from 5 pigs revealed several E. suis specific antigens at the same molecular weights as observed in Study II, further confirming the E. suis specificity of the 100K antigen preparation. However, there was a possibility that some of the observed E. suis specific antigens may have been erythrocyte antigens altered by E. suis infection. These altered erythrocyte antigens could be retained in the antigen preparation prepared from the post parasitemic blood in the second experiment of Study III, because it has been observed that erythrocyte membrane alteration was retained after E. suis was eliminated from the blood by tetracycline treatment (Zachary & Basgall, 1985). In the present experiment, E. suis specific antigens were detected only in antigen preparations from parasitemic blood and not from blood collected during remission, indicating 130 that the observed antigens were E. suis derived proteins and not proteins derived from erythrocytes that may have been altered by E. suis infection. In the same experiment, the numbers of E. suis specific antigens observed in the 100K antigen preparations from the lst and the 2nd parasitemic phase blood were different. This was probably due to the difference in concentrations of E. suis bodies in the 2 antigen preparations. The degree of parasitemia at the 1st and 2nd parasitemic phase were 72 and 96%, respectively. The first parasitemic phase antigen was the only antigen prepared from less than 90% parasitemic blood in all of the experiments. Considering the differences in the number of E. suis specific antigens detected in the 1st and 2nd parasitemic phase antigens, it was assumed that preparing the antigen from blood with a lower parasitemia resulted in lower concentration of E. suis specific antigens. Antigens prepared from infected erythrocyte lysates have been developed in various hemoparasites infection. In swine eperythrozoonosis, whole lysate of infected erythrocytes have been used in the complement fixation test (Splitter, 1958), and ELISA (Schuller et al., 1990). Hsu et al. (1992) developed an E. suis antigen prepared from infected erythrocyte lysate with 2 purification steps. In their procedure, lysed erythrocytes were centrifuged at 1,0009 and the supernatant was discarded. The pellet was suspended in PBS and frozen. The frozen suspension was then thawed and centrifuged at 11,0009. The resultant supernatant was used 131 for the antigen in ELISA. In ovine and bovine eperythrozoonosis, semi-purified organisms from infected erythrocyte lysates have been used for complement fixation, ELISA and passive hemagglutination tests (Daddow, 1977 ; Finerty et al., 1969; Kawazu et al., 1990; Lang et al., 1987). In anaplasmosis, purification of A. marginale from erythrocyte lysates has been attempted using enzymatic digestion of erythrocyte components (Hart et al., 1981) and density gradient centrifugation (McCorkle-Shirley et al., 1985). Palmer and McGuire (1984) purified A. marginale by erythrocyte disruption with minimal sonication and centrifugation at a force not sufficient to sediment erythrocyte stroma. However, none of these antigen preparation methods could guarantee complete absence of host proteins or complete integrity of the parasite. Therefore, analysis of such preparations would not reliably characterize the protein structure of the organism. In the erythrocyte culture system used for A. marginale, it has been demonstrated that infected erythrocytes incorporated radiolabeled amino acids but uninfected erythrocytes did not (Davis et al., 1978). A. marginale protein structure has been analyzed under the complete absence of host erythrocyte protein contamination, using organisms radiolabeled by metabolic incorporation of 358- methionine (Barbet et al., 1983). A methionine requirement was also demonstrated in malaria parasites (Jensen, 1983). To identify E. suis protein structures, radiolabeling of E. 132 suis by 358-methionine was accomplished using the erythrocyte culture system developed in a previous study (Chapter 2). The 100K antigen preparation, using 358-methionine labeled E. suis, demonstrated 25 distinct protein bands. Only 2 faint bands were observed in the antigen preparation from uninfected erythrocyte culture. Several of the radiolabeled E. suis protein bands showed the same molecular weight as the E. suis specific bands observed in the western blot experiments, further confirming the E. suis specific bands observed in the western blots were E. suis derived. However, some of the E. suis specific bands observed in the western blot experiments were not detected by autoradiography. Since radiolabeling was performed for only 4 hours, it could be assumed that only proteins with sufficiently rapid turnover were radiolabeled effectively and proteins with a relative slow turnover were not. Alterations in erythrocyte membrane proteins due to infection may be another potential reason for the inconsistencies of E3 suis specific antigens between the autoradiographs and western blots. In bovine anaplasmosis, it has been observed that a high molecular weight glycoprotein observed in normal bovine red cells was not present in infected red cells. Instead, four new lower molecular weight glycoproteins were found in infected erythrocytes (Nordelo & Ysern-Caldentey, 1982). These new glycoproteins were suggested to be the result of structural modifications of erythrocyte proteins due to Anaplasma 133 infection. Goff et al. (1986) found that reactions to lectins by bovine erythrocytes were altered by A. marginale and E. wenyoni infection. They suggested that additional lectin binding sites were exposed by alteration of erythrocyte membrane proteins by rickettsial infection. In addition, alterations of erythrocyte proteins have been postulated in Plasmodium infection as a result of changes in the relative intensity of the protein bands in SDS-PAGE, phosphorylation patterns, insertion of a new polypeptide, and recognition of the erythrocyte proteins by monoclonal antibodies (Chaimanee & Yuthavong, 1979; Crandall & Sherman, 1991; Yuthavong et al., 1979). Surface proteins and glycoproteins have been observed to be altered during infection with Babesia bovis (Howard et al., 1980). In swine eperythrozoonosis, alteration of erythrocyte membranes has been observed by electron microscopy and shown to persist even after the parasites were eliminated by tetracycline treatment (Zachary & Basgall, 1985). Zachary and Smith (1985) suggested that recognition of altered erythrocyte antigens as foreign antigen by the host's immune system may cause autoantibody production in swine eperythrozoonosis. Therefore, it could be assumed that some of the apparent E. suis specific antigens may be altered erythrocyte proteins caused by E. suis infection. The source of the two faint bands observed in the autoradiograph of the antigen preparation from the uninfected erythrocyte culture material was not clear. Since mature 134 erythrocytes do not synthesize any proteins, the proteins of mature erythrocytes could not be radiolabeled by' incorporating 355-methionine. Possible sources of the bands may be a small number of metabolically active leukocytes left over after buffy coat removal, or the presence of immature erythrocytes in the culture system. Based on IgG western blotting and autoradiography results, eight protein bands with MW 24, 37, 47, 71, 81, 90, 101 and 106 kd, were selected as the most reliable E. suis specific bands. The final experiment was designed to evaluate the time and duration that the sera of experimentally infected pigs recognized these eight antigens after E. suis inoculation. Nearly all of the eight antigen bands were consistently recognized in western blots using sera collected from 2 to 21 weeks post infection from all four pigs evaluated, indicating that the selected antigens were highly antigenic. One week post-infection sera of all four pigs failed to recognize the antigens of MW 37, 81 and 101 kd and some sera obtained at one week post infection failed to recognize other antigens as well. Therefore, the IgG immune response occurred within 1 to 2 weeks and lasted up to 21 weeks post infection. However, sera obtained 2 to 21 weeks post infection could not always recognize all of the eight selected antigens. Since E. suis is an extracellular parasite and exposed to immunoglobulins in the plasma, it could be expected that the quantity of E. suis specific IgG in the 135 plasma, collected at the parasitemic phase, may be significantly reduced by binding to organisms present in the blood. The inability of all post-infection sera to recognize all antigens may be related to the fact that some of the sera were collected during parasitemia. This might partly explain why some of the 1 week post-infection sera failed to recognize the antigens, besides the time lag between infection and the initiation of a measurable IgG immune response. Three of the four 1 week post infection sera were collected during a parasitemic (>85%) period. On the other hand, some of the sera failed to recognize the antigens even when the sera were collected during remission. Among the 8 selected proteins, the 90 Rd protein was of specific interest. This protein was detected as an intense band in the western blots using the Es" 100K antigen preparation with. the immune reference serum" and. by autoradiography. The band appearance in the autoradiograph indicated that the 90 kd protein was derived from E. suis. However, in Study II, a similar band was observed at the same location (90 kd) in the Es' preparations of thaw, 500 and 10K antigen. This band, presumed to be host derived 90 kd protein, was not detected in the Es‘ 100K antigen indicating that the host derived 90 kd protein was eliminated in the 100K antigen preparation. However, in the western blots performed in Study III, the host derived 90 kd band was occasionally observed in the Es‘ 100K antigen, indicating the band could not always be eliminated by differential 136 centrifugation. These observations suggested that the Es+ antigen preparation could contain two proteins, both having similar molecular weight of about 90 kd, one derived from E. suis and one from the host. Therefore, the 90 kd antigen band was not considered to be an appropriate band for evaluating E. suis specificity. The 71 kd protein was recognized by the pre-infection sera of 2 pigs. This may indicate that the 2 pigs were already infected with E. suis prior to inoculation or had persisting colostral antibodies. Faint broad bands were observed in the Es' antigen preparation around the 70 kd region. These bands could be confused with the 71 kd protein; although they were not sharp, and might be a reflection of nonspecific binding of negatively charged IgG to positively charged proteins located around the 70 kd region. Interestingly, various antigens were recognized only by late infection stage sera. The cause of this observation was not elucidated in this study but indicates that the recognition of some antigens by the host's humoral immune system was delayed. This may be related to parasite evasion mechanisms from the host's immune system. Parasites can develop escape mechanisms to avoid the destructive effects of the immunologically hostile host. Cohen (1982) summarized various immune evasion mechanisms of parasites including anatomical seclusion of parasites, modification of parasite antigens, and modification of host immune responsiveness. 137 The immune evasion mechanisms of Eperythrozoon in general are poorly understood. In experimentally infected, parasitemic pigs, lymphocytes blastogenesis induced by lipopolysaccharide was increased (Zachary & Smith, 1985). Hyperglobulinemia was also observed during parasitemia. These findings indicate that polyclonal B cell activation took place during parasitemia. Polyclonal B cell activation is an immune evasion mechanism developed by many parasites, leading to misdirected responses by the host's immune system, resulting in generalized immune suppression (Cohen, 1982). It has been commonly observed that polyclonal B cell activation induces autoantibody production, consisted with one of the characteristics of swine eperythrozoonosis. As an alternative consideration, antigen mimicry of the parasite antigens to host antigens has been discussed as a cause of autoantibody induction in rickettsial and bacterial infections (Ristic, 1976; Steere, 1989). This mechanism should be considered as a potential immune evasion mechanism developed by E. suis, resulting in the host's immune system failure to recognize certain E. suis antigens as foreign. Surface antigens continually exposed to the host might be considered as "evolutionary antigens" by which the parasite has responded in various ways to the host response and developed various immune evasion mechanisms (Lloyd & Soulsby, 1988). For parasite survival in the host, the functional antigens of the parasite, which induce protective immune responses, need to escape from host immune system 138 attack. (hi the other hand, antigens non-functional for development of host resistance could remain susceptible to host attack. Therefore, the observed antigens recognized only by sera.collected during late infection could be considered as functional antigens of E. suis. This study analyzed E. suis antigens by western blotting. The study elucidated a lack of specificity of IHA, for detecting E. suis specific antibodies. Therefore, the development of improved serodiagnostic methods is a critical requirement to gain a better understanding of the disease. The 100K antigen preparation developed in the present study could provide a vital component for developing more accurate serodiagnostic methods. 139 References 1. Barbet, A. F., Anderson, L. W., Palmer, G. H. and Mcguire, T. C. (1983): Comparison of proteins synthesized by two different isolates of Anaplasma marginale. Inf. Immun., 40, 1068-1074 2. Barry, D. N., Parker, R. J., De Vos, A- J., Dunster, P. and Rodwell, B. J. (1986): A microplate enzyme-linked immunosorbent assay for measuring antibody to Anaplasma marginale in cattle serum. Aust. vet. J}, 63, 76-79 3. Bellamy, J. E. C., MacWilliams, P. s. and Searcy, G. P. (1978): Cold-agglutinin hemolytic anemia and Haemobartonella canis infection in a dog. JAVMA, 173, 397-401 4. Callow, L. L. and Dalgliesh, R. J. (1982): Immunity and immunopathology in babesiosis. In Immunology of parasitic infections, 2nd ed., Ed. S. Cohen and K. S. warren, Blackwell Scientific Publications, Oxford, UK, pp. 475-526 5. Chaimanee, P. and Yuthavong, Y. (1979): Phosphorylation of membrane proteins from Plasmodium berghei-infected red cells. Biochem. Biophys. Res. cammun., 87, 953-959 6. Cohen, S. (1982): Survival of parasites in the immunocompetent host. In Immunology of parasitic infections, 2nd ed., Ed. S. Cohen and K. S. Warren, Blackwell Scientific Publications, Oxford, UK, pp. 138-161 7. Crandall, I. and Sherman, I. W. (1991): Plasmodium falciparum (human malaria)-induced modifications in human erythrocyte band 3 protein. Parasitology, 102, 335-340 8. Daddow, K. N. (1977): A complement fixation test for the detection of Eperythrozoon infection in sheep. Aust. Vet. Jo, 53' 139-143 9. Davis, W. C., Talmadge, J. E., Parish, S. M., Johnson, M. I. and Vibber, S. D. (1978): Synthesis of DNA and protein by Anaplasma marginale in bovine erythrocytes during short- term culture. Inf. Immun., 22, 597-602 10. Dimmock, C. K. (1973): Blood group antibody production in cattle by a vaccine against Babesia argentina. Res. Vet. SCio' 15' 305-309 11. Doyle, L. P. (1932): A rickettsia-like or anaplasmosis- like disease in swine. JAVMA, 81, 668-671 140 12. Duzgun, A., Schunter, C. A., Wright, I. G., Leatch, G. and Waltisbuhl, D. J. (1988): A sensitive ELISA technique for the diagnosis of Anaplasma marginale infections. Vet. Parasitol., 29, 1-7 13. Finerty, J. F., Hidalgo, R. J. and Dimopoullos, G. T. (1969): A. passive ihemagglutination procedure for the detection of Eperythrozoon infection in calves. Am. J. Vet. Res., 30, 43-45 14. Goff, W. L., Johnson, W. and Kuttler, K. L. (1986): Anaplasma marginale, Eperythrozoon wenyoni: Lectin reactions with bovine erythrocytes. Expt. Parasitol., 61, 103-113 15. Goodger, B. V. and Mahoney, D. F. (1974): Evaluation of the passive haemagglutination test for the diagnosis of Babesia argentina infection in cattle. Aust. vet. J., 50, 246-249 16. Goodger, B. V., Wright, I. G. and Waltisbuhl, D. J. (1985): Babesia bovis: the effect of acute inflammation and isoantibody production in detection of babesial antigens. EXperimentia, 41, 1577-1579 17. Hall, 8. M., Cipriano, J. A., Schoneweis, D. A., Smith, J. E. and Fenwick, B. W. (1988): Isolation of infective and non-infective Eperythrozoon suis bodies from the whole blood of infected swine. vet. Rec., 123, 651 18. Hancock, K. and Tsang, V. C. W. (1983): India Ink Staining of proteins on nitrocellulose paper. Analytical Biochem., 133, 157-162 19. Hart, L. T., Larson, A. D., Decker, J. L., weeks, J. P. and Clancy, P. L. (1981): Preparation of intact Anaplasma marginale devoid of host cell antigens. Current Microbiol., 5, 95-100 20. Henry, S. C. (1979): Clinical observation on Eperythrozoonosis. JAVMA, 174, 601-603 21. Hoffmann, R., Schmid, D. O. and Hoffmann-Fezer, G. (1981): Erythrocyte antibodies in porcine eperythrozoonosis. vet. Immunol. Immunopathol., 2, 111-119 22. Howard, R. J., Rodwell, B. J., Smith, P. M., Callow, L. L. and Mitchell, G. F. (1980): Comparison of the surface proteins and glycoproteins on erythrocytes of calves before and during infection with Babesia Davis. J. Protozool., 27, 241-247 141 23. Hsu, F. S., Liu, M. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1990): Evaluation of an enzyme-linked immunosorbent assay for the detection of Eperythrozoon suis antibody in swine. In Proc. 11th International Pig Veterinary Society congress, Lausanne, Switzerland pp. 315 24. Hsu, F. S., Liu, M. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1992): Evaluation of an enzyme-linked immunosorbent assay for detection of Eperythrozoon suis antibodies in swine. Am. J. vet. Res., 53, 352-354 25. Jensen, J. B. (1983): Plasmodium. In In vitro cultivation of protozoan parasites, Ed. J. B. Jensen, CRC Press, Boca Raton, FL, pp. 155-192 26. Kawazu, S., Nakamura, Y., Kamio, T., Fujisaki, K. and Minami, T. (1990): Enzyme-linked immunosorbent assay for detection of antibodies to Eperythrozoon wenyoni in cattle. Jpn. J. vet. Sci., 52, 1297-1300 27. Lang, F. M., Ferrier, G. R. and Nicholls, T. J. (1986): Separation of Eperythrozoon ovis from erythrocytes. vet. Rec., 119, 359 28. Lang, F. M., Ferrier, G. R. and Nicholls, T. J. (1987): Detection of antibodies to Eperythrozoon ovis by the use of an enzyme-linked immunosorbent assay. Res. vet. Sci., 43, 249-252 29. Lau, F. O. and Rosse, W. F. (1975): The reactivity of red blood cell membrane glycophorin with "cold-reacting" antibodies. Clinic. Immunol. Immunopathol., 4, 1-8 30. Lloyd, S. and Soulsby, E. J. L. (1988): Immunological responses of the host. In Parasitology in focus: Facts and Trends, Ed. H. Mehlhorn, Springer-Verlag, Berlin, Germany, pp. 619-650 31. Maede, Y. (1980): Studies on feline haemobartonellosis. VI. Changes of erythrocyte lipids concentration and their relation to osmotic fragility. Jpn. J. vet. Sci., 42, 281- 288 32. McCorkle-Shirley, 8., Hart, L. T., Larson, A. D., Todd, W. J. and Myhand, J. D. (1985): High-yield preparation of purified Anaplasma marginale from infected bovine red blood cells. Am. J. vet. Res., 46, 1745-1747 33. Nordelo, M. A. and Ysern-Caldentey, M. (1982): Abnormal bovine erythrocyte membrane proteins and glycoproteins during and after infection with Anaplasma marginale. Biochem. Biophys. Res. Commun., 104, 664-672 142 34. Oki, Y. and Miura, K. (1970): Characteristics and roles of red cell autoantibodies in equine infectious anemia. Jpn. J. vet Res., 32, 217-227 35. Overas, J. (1969): Studies on Eperythrozoon ovis infection in sheep. Acta vet. Scand., 28 (Sup.), 1-147 36. Palmer, G. H. and McGuire, T. C. (1984): Immune serum against Anaplasma marginale initial bodies neutralizes infectivity for cattle. J. Immunol., 133, 1010-1015 37. Plank, G. and Heinritzi, K. (1990): Disseminated intravascular coagulation in porcine eperythrozoonosis. Berl..Mfinch. Tierarztl. WSchr., 103, 13-18 38. Pruzanski, W. and Shumak, K. H. (1977a): Biological activity of cold-reacting autoantibodies (first of two parts). New Eng. J. Med., 297, 538-542 39. Pruzanski, W. and Shumak, K. H. (1977b): Biological activity of cold-reacting autoantibodies (second of two parts). New'Eng. J. Med., 297, 583-589 40. Ristic, M. (1976): Immunologic systems and protection in infections caused by intracellular blood protista. vet. Parasitol., 2, 31-47 41. Robb, A. D. (1943): Ictero—anemia in growing swine. vet. Med. Sm. Anim. Clinic., 38, 271-274 42. Schmidt, P., Kaspers, B., Jfingling, A., Heinritzi, K. and Lbsch, U. (1992): Isolation of cold agglutinins in Eperythrozoon suis-infected pigs. Vet. Immunol. Immunopathol., 31, 195-201 43. Schroeder, W. F. and Ristic, M. (1965): Anaplasmosis XVII. The relation of autoimmune processes to anemia. Am. J. vet. Res., 26, 230-245 44. Schuller, W., Heinritzi, K., Al-Nutha, S., Kolbl, S. and Schuh, M. (1990): Serological studies by means of CF- tests and ELISA for evaluation of antibodies against Eperythrozoon suis. Berl..Mfinch. Tierarztl. WSchr., 103, 9- 12 45. Shankarappa, B. and Dutta, S. K. (1989): Production and characterization of monoclonal antibodies to Ehrlichia 46. Sisk, D. H., Cole, J. R. and Pursell, A. R. (1980): Serologic incidence of eperythrozoonosis in Georgia swine. Proc. ARIN]. Meet. Am. ASSOC. Veto Lab. Diag., 23' 91-100 143 47. Smith, A. R. (1977): Eperythrozoonosis. Veterinary (professional topics, university of Illinois: Swine, 5, 2-4 48. Smith, A. R. (1981): Eperythrozoonosis. In Diseases of swine, Ed. B. S. A. D. Leman R. D. Glock, W. L. Mengeling, R. H. C. Penny and E. Scholl, Iowa State University Press, Ames, IA, pp. 683-687 49. Smith, A. R. and. Rahn, T. (1975): An indirect hemagglutination test for the diagnosis of Eperythrozoon suis infection in swine. Am. J. vet. Res., 36, 1319-1321 50. Smith, B. J. (1984): SDS polyacrylamide gel electrophoresis of proteins. In Proteins: Methods in molecular biology, vol.1, Ed. J. W. Walker, Humana Press, Clifton, NJ, pp. 41-55 51. Soni, J. L. and Cox, H. W. (1974): Pathogenesis of acute avian malaria: I: Immunologic reactions associated with anemia, splenomegaly, and nephritis of acute Plasmodium gallinaceum infections of chickens. Am. J. Trop. Med. Hyg., 23, 577-585 52. Soni, J. L. and Cox, H. W. (1975): Pathogenesis of acute avian malaria: II: Anemia mediated by a cold-active autohemagglutinin from the blood of chickens with acute Plasmodium gallinaceum infection. Am. J. Trop. Med. Hyg., 24, 206-213 53. Splitter, E. J. (1958): The complement-fixation test in diagnosis of eperythrozoonosis in swine. JAVMA, 132, 47-49 54. Splitter, E. J. and Williamson, R. L. (1950): Eperythrozoonosis in swine. A preliminary report. JAVMA, 116, 360-364 55. Steere, A. C. (1989): Lyme disease. New Eng. J. Med., 321, 586-596 56. Thoongsuwan, S. and Cox, H. W. (1973): Antigenic variants of the haemosporidian parasite, Babesia rodohaini, selected by in vitro treatment with immune globulin. Ann. Trop. Med. Parasitol., 67, 373-385 57. Towbin, H., Staehelin, T. and Gordon, J. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci., USA, 76, 4350-4354 58. Wright, I. G. (1973): Osmotic fragility of erythrocytes in acute Babesia argentina infections in splenectomized Bos taurus calves. Res. vet. Sci., 15, 299-305 144 59. Yuthavong, Y., Wilairat, P., Panijan, B., Potiwan, C. and Beale, G. H. (1979): Alterations in membrane proteins of mouse erythrocytes infected with different species and strains of malaria parasites. cemp. Biochem. Physiol., 63, 83-85 60. Zachary, J. F. and Basgall, E. J. (1985): Erythrocyte membrane alterations associated with the attachment and replication of Eperythrozoon suis: A light and electron microscopic study. vet. Pathol., 22, 164-170 61. Zachary, J. F. and Smith, A. R. (1985): Experimental porcine eperythrozoonosis: T-lymphocyte suppression and misdirected immune responses. Am. J. Vet. Res., 46, 821-830 62. Zulty, J. C. and Kociba, G. J. (1990): Cold agglutinins in cats with haemobartonellosis. JAVMA, 196, 907-910 CHAPTER 4 DEVELOPMENT OF IMPROVED SERODIAGNOSTIC TESTS FOR SWINE EPERYTHROZOONOSIS 14S 146 Abstract Enzyme-linked immunosorbent assays (ELISA) using horse radish peroxidase (HRP) conjugated rabbit anti-pig IgG (IgG- ELISA) and HRP conjugated protein A (ProA-ELISA), or a dot blot (DB) using 12-"I-protein A were performed on sera obtained from experimentally infected pigs and naturally exposed pigs (field sera) using Es+ (E. suis positive) and E8“ (E. suis negative) 100K antigen preparations described in Chapter 3. Serum antibody levels were determined by two methods: 1) actual optical density (OD) in ELISA or percent binding of protein A in DB, using Es+ antigen only (CrudeEs-Test); or 2) adjusted OD or percent binding calculated by subtracting the value obtained with E3“ antigen from Es“ antigen value (NetEs-Test). The ELISA and DB results were compared to the indirect hemagglutination assay (IHA). The specificity of the rabbit anti-pig IgG and protein A reagents for detecting immunoglobulins bound to antigens were compared by western blotting and no differences in the specificities of the two reagents were found as the band patterns were identical. By 2 weeks post-infection (WPI), most sera were positive by ELISA and DB as determined by CrudeEs-Test or NetEs-Test, and by IHA. However, IHA failed to detect antibodies in all of the 1 WPI sera, while the ProA- and IgG-ELISA's and the DB detected antibodies in 71, 63 and 88% of the 1 WPI sera by CrudeEs-Test, and 57, 75 and 88% by NetEs-Test, respectively. The time course of IHA titer development following 147 experimental infection did not correlate with the development of E. suis specific antibodies detected by the other assays, using CrudeEs-Test or NetEs-Test. However, IHA titers did correlate with results obtained with the ELISA's and DB using the Es' antigen, indicating that the IHA titer was not specific for E. suis antibodies. The IHA was detecting erythrocytic authntibodies induced by E. suis infection. The results obtained with the field sera varied considerably among the tests. For example, 91% of sera were positive by IgG-ELISA CrudeEs-Test, while only 9% of sera were positive by DB NetEs-Test. However, the positive or negative results obtained by CrudeEs-Test or NetEs-Test were significantly associated between the two ELISA's and the DB, but not between the IHA and the other tests. The detection of 37, 47, 71, 81 and 101 kd antigens by field sera in IgG western blots were closely associated with the results of the two ELISA's and the DB and the NetEs-Test results with the two ELISA's were significantly associated with the detection of all 5 of the antigens. However, there was no association between IHA results and the appearance of E. suis specific antigen bands in the western blots. It was concluded that NetEs-Test using either ELISA or DB could provide better serodiagnostic assays for swine eperythrozoonosis compared to the standard IHA. 148 Introduction Swine eperythrozoonosis is caused by the rickettsial organism, Eperythrozoon suis. The disease is characterized by acute ictero-anemia and is associated with poor weight gain and reproductive failure (Henry, 1979). Swine eperythrozoonosis has been reported world wide including North and South America, Africa, Europe and Asia (Henry, 1979; Hsu et al., 1990; Sisk et al., 1980; Smith, 1977; Smith, 1981). ' The diagnosis of swine eperythrozoonosis has been based on herd and individual animal histories describing icteroanemia and on the demonstration of Eperythrozoon bodies in blood smears. Detection of the organism provides definite diagnosis; however, artifacts in 'blood smears have been difficult to distinguish from E. suis bodies in many cases (Henry, 1979). Moreover, most E. suis infections are subclinical without detectable parasitemia. Serodiagnostic assays such as the complement fixation test, indirect hemagglutination assay (IHA) and enzyme-linked immunosorbent assay (ELISA) have been developed for detecting low level infections not detected by traditional blood smears (Hsu et al., 1992; Schuller et al., 1990; Smith & Rahn, 1975; Splitter, 1958). The antigens used in these assays were crude preparations from parasitemic plasma or the lysate of infected erythrocytes and thus contained host plasma and erythrocyte antigens. In swine eperythrozoonosis, 149 autoantibodies against erythrocytes have been postulated to play an important role in anemia development (Hoffmann et al., 1981), in the occurrence of agglutination of infected blood especially at cold temperature, and increased sedimentation rate of infected blood (Doyle, 1932; Robb, 1943; Splitter & Williamson, 1950). The occurrence of anti- erythrocyte antibodies in E. suis infected pigs has been shown by positive Coomb's tests (Hoffmann et al., 1981). Considering the presence of host proteins in the antigen preparations used in the serodiagnostic assays for swine eperythrozoonosis, test results may be greatly influenced by immune complex formation between host proteins and the autoantibodies induced by E. suis infection. As reported in Chapter 3, immunoblot analysis of the antigen preparation used in the indirect hemagglutination assay (IHA), a widely used serodiagnostic technique for E. suis infection, indicated that antibodies detected by IHA are mainly directed at host derived proteins. The following studies were conducted to develop improved serodiagnostic assays for swine eperythrozoonosis. 150 Materials and Methods Organism. E. suis was initially isolated from an inappetant boar housed at the Michigan State University Swine Research Center. A diagnostic CBC revealed that the boar was parasitemic and anemic. The E. suis isolate has been .maintained by passage in splenectomized pigs. Infected blood also has been cryopreserved by adding 6.7% and 10% (v/v) of 40% polyvinyl pyrrolidone (PVP) and glycerol, respectively, and freezing in liquid nitrogen. Animals and experimental infection. Cross bred pigs were 6- to 13-weeks-old at the time of inoculation with E. suis. All animals were splenectomized using standard surgical techniques 1 to 2 weeks prior to inoculation and determined to be E. suis noninfected by' no evidence of any temperature elevation or development of parasitemia before inoculation. Ten milliliter of frozen infected blood was thawed in a 37 °C water bath and inoculated intravenously aJui/or intramuscularly into the splenectomized pigs. Clinical signs and rectal temperatures were assessed daily for monitoring disease development. When the rectal temperature exceeded 40 °C, blood was collected in a vacutainer tube (Beckton Dickinson Vacutainer Systems, Rutherford, NJ) containing ethylenediamine tetraacetic acid (EDTA), and examined for parasitemia by Wright's stained thin smear. When the febrile phase lasted for more than 3 days or if an animal exhibited 151 severe depression or anemia (less than 20% of packed cell volume), intramuscular treatment with oxytetracycline (20 mg/kg body weight; Liquamycin LA 200, Pfizer, New York, NY) was initiated to avoid mortality. Serum samples. Serum samples were obtained from 8 experimentally infected animals prior to inoculation with E. suis and weekly for the first 4 weeks post-infection (WPI) and every other week from 4 to 12 WPI. A total of 140 serum samples obtained from naturally exposed pigs (field sera), consisting of 7 sera each from 20 herds, submitted to the Michigan Department of Agriculture for regulatory purposes were kindly provided by Dr. Richard Gatzmeyer. The immune reference serum was a pool of sera obtained from blood collected at 2 to 8 WPI from 3 experimentally infected animals. The nonimmune reference serum was obtained from blood of a specific-pathogen-free (SPF) pig, known to be E. suis free. Indirect hemagglutination assay (IHA). The IHA procedure of Smith and Rahn (1975) was followed. 1) Preparation of sensitized sheep erythrocytes (S-RBC) with IHA antigen: Five milliliters of sheep blood was collected with equal volume of Alsever's solution and erythrocytes were washed 3 times with phosphate buffered saline (PBS; pH 7.2). A half milliliter of packed erythrocytes were suspended in 9.5 ml of PBS in a 50-ml 152 beaker and Stirred on a magnetic stirrer. While stirring, 24.7 mg of IHA antigen (provided by Dr. A. R. Smith, University of Illinois, Champaign-Urbana, IL) was added followed by dropwise addition of 1 ml of 2.5% glutaraldehyde. The mixture was stirred for one hour, then 1 ml of 2% ovalbumin was added and the mixture was stirred for another 30 minutes. S-RBC were washed 3 times and resuspended in 9 times its volume with PBS and stored in liquid nitrogen. For use in the IHA, 10% S-RBC solution was diluted 20 times with PBS. The glutaraldehyde treated sheep erythrocytes (G- RBC) were prepared identical to S-RBC except no antigen was added. G-RBC was used as a 10% solution. 0 2) Assay: Test sera were heat inactivated at 56 °C for 30 minutes. To 100 pl of inactivated serum, 0.85 ml of PBS and 50 ul of the 10% G-RBC solution were added and the mixture was incubated for 15 minutes at 25 °C. 'The mixture was centrifuged at 500 g for 10 minutes and 50 ul of the supernatant was transferred to a round bottom 96 well microplate. The transferred sera were diluted serially (two fold) using 1% heat inactivated rabbit serum in PBS with a starting dilution of either 1:10 or 1:20. The test antigen, 50 ul of 0.5% S-RBC solution, was added to each well. The plate was covered and incubated at 4 °C over night. The IHA titer (geometric mean for duplicate or triplicate samples) was interpreted as follows: 1:80 or more was positive; 1:20 or less was negative; and a titer between 1:20 to 1:80 was suspicious. 153 100K antigen preparation. Blood collected with citrate phosphate dextrose solution (14 ml/100 ml blood) containing 0.2% inosine was centrifuged at 5009 for 20 minutes and the plasma and buffy coat were removed. Erythrocytes were washed 3 times with Eagle's minimum essential medium containing 0.2% inosine (EMI). After the final wash, erythrocytes were resuspended in the same volume of EMI containing 5% fetal calf serum. The suspension was mixed with 6.6% and 10% (v/v) of 40% PVP and glycerol, respectively, and stored at -80 °C. The frozen erythrocytes were thawed and centrifuged at 100,000g for 40 minutes. The pellet was resuspended in PBS (pH 7.4) and sequentially centrifuged at 500g for 20 minutes and at 10,0009 and 100,000g for 30 minutes each. The pellet from the 100,000g centrifugation was solubilized by 3 times freeze-thawing in Tris buffer (pH 8.0) containing 5 mM EDTA, 5 mM iodoacetamide, 1 mM phenylmethylsulfonyl fluoride, 0.1% Nonidet P-40 and 0.1% sodium dodecyl sulfate (SDS), and stored at -80 °C (100K antigen). 100K antigens were prepared from noninfected and infected blood (>75% parasitemia) collected before and after infection from the same animal (Es‘ and Es+ 100K antigens, respectively). The protein concentrations of the antigen preparations were measured by the Bio-Rad protein assay (Bio- Rad, Richmond, CA). 154 Enzyme-linked immunosorbent assay (ELISA). Two ELISA's were performed using either horse radish peroxidase (HRP) conjugated protein A (ProA-ELISA) or HRP conjugated rabbit anti-pig IgG (IgG-ELISA). The wells of polystyrene ELISA plates (Immulon 2, Dynatech Laboratories Inc., Chantilly, VA) were coated with 0.5 ug (ProA-ELISA) or 1 pg (IgG-ELISA) of the Es+ or E3” 100K antigens in 100 pl of 50 mM carbonate buffer (pH 8.5), sealed, and incubated at 37 °C over night. After coating the wells, the plates were washed 3 times with ELISA washing buffer, composed of PBS (pH 7.4) containing 0.05% Tween 20 (polyoxyethylene sorbitan monolaurate), and blocked for one hour with 200 pl of 2% bovine serum albumin (BSA) in 50 mM carbonate buffer (pH 8.0). The plate was then incubated with 100 pl of the 1:100 dilution of the test sera, diluted with ELISA incubation buffer (0.5% BSA-PBS-0.05% Tween, pH 7.2), for 2 hours at 251°C. The plate was washed 4' times with ELISA washing buffer, followed by the addition of 100 ul of l:10,000 dilution of HRP conjugated protein A (Kirkegaard & Perry Laboratories Inc., Gaithersburg, MD) diluted with ELISA incubation buffer, or 1:5,000 dilution of HRP conjugated rabbit anti-pig IgG (Sigma Chemical Co.) diluted with ELISA incubation buffer. The plates were then incubated for another hour at 25 °C. Following incubation, the plates were washed 4 times with the ELISA washing buffer. For the ProA-ELISA, plates were incubated with 100 (:1 of 0.03% ABTS (2,2'-azino-bis(3-ethy1benzthiazoline-6-sulfonic) acid); Kirkegaard & Perry) in 50 mM phosphate citrate buffer 155 containing 0.01% hydrogen peroxide for 15 minutes. For the IgG-ELISA, plates were incubated with 100 pl of 0.01% ABTS (Sigma Chemical Co.) in 50 mM phosphate citrate buffer containing 0.01% hydrogen peroxide for 30 minutes. After incubation, the optical density (OD) of each well was measured at 405 nm by an ELISA reader (EIA Reader Model EL 307, Bio-Tek Instruments Inc., Burlington, VT). The results were reported as two values: 1. CrudeEs-Test the OD obtained with the Es+ antigen; the difference between the OD's obtained 2. NetEs-Test with the Es“ antigen compared to the E3“ antigen, calculated by subtracting OD value of the ES' antigen from the Es+ antigen OD value. Individual cut off values (mean + 2 standard deviations) were determined for each of the two ELISA's (CrudeEs-Test or NetEs-Test) using the data from the pre-infection sera obtained from the 8 experimentally infected pigs. The means. and standard deviations of OD values of pre-infection sera and cut off values for each test are shown in Table 4-1. The positive/negative results of each CrudeEs-Test and NetEs-Test of the two ELISA's were determined as follows: in each test, sera showing OD's equal to or greater than the respective cut off value were classified as positive; and sera showing OD's less than the respective cut off value were classified as negative. Table 4-1. Determination of cut off values used for classification of test sera to positive 156 or negative by ProA (protein A)- and lgG-EUSA's and dot blot. A59)! Test Mean‘ Standard deviation1 Cut off valuei ProA-ELISA CrudeEs-Test 0.202 0.053 0.308 NetEs-Test 0.1 17 0.045 0.207 IgG-ELISA CrudeEs-Test 0.079 0.023 0.1 25 NetEs-Test 0.045 0.018 0.081 Dot blot 15 CrudeEs-Test 1.460 0.518 2.496 NetEs-Test 0.605 0.209 1.023 Dot blot 2" CrudeEs-Test 1.291 0.155 1.601 NetEs-Test 0.281 0.268 0.817 : - Mean values of ELISA 00's or dot blot percent bindings of pre-infection sera obtained from blood of 8 experimentally infected pigs, determined by CrudeEs-Test or NetEs-Test. T - Standard deviations of EUSA 00's or dot blot percent bindings of pre-infection sera determined by CrudeEs-Test or NetEs-Test. - determined by mean + 2 standard deviations. Using ‘ZSI-protein A batch 1. - Using 125l-protein A batch 2. =co'a-H- I 157 Dot blot (DB). For this assay, protein A was labeled with 1251 by the method of Fraker and Speck (1978). Three micrograms of 100K antigen (either Es+ or Es‘) were placed in the respective wells of a blotter (Bio-Blot; Bio-Rad). Antigen was applied to the nitrocellulose membrane (NC) using vacuum and the wells were washed once with PBS (pH 7.4). The blotted NC was blocked overnight in blot blocking buffer, 2% BSA in tris buffered saline (TBS, pH 7.4) with 0.5% Tween 20. After blocking, the NC-was cut into pairs of Es+ and Es- antigen dots. Each pair of dots were placed into the wells of 24 well culture plates (Corning, Corning, NY), and incubated with 0.5 ml of a 1:200 dilution of serum diluted with blot incubation buffer (0.5% BSA-TBS-0.5% Tween, pH 7.2) for 2 hours. The.dots were then washed 3 times with blot washing buffer (0.2% BSA-TBS-0.5% Tween, pH 7.0), followed by incubation with 0.5 ml of 125I-protein A (20 ng/ml in 0.5% BSA-TBS-0.5% Tween) for one hour. The dots were washed 3 times with blot washing buffer, the pairs were separated, and the radioactivity of each dot was measured in a gamma counter (Gamma Counting System, RIA 300; Tm Analytic, Elk Grove Village, IL). To adjust for the effect of radioactive decay over time, the percent binding was calculated by the following equation: Count per minute (CPM) of a dot Percent binding = x 100 CPM of total amount added to a well 158 As used in the ELISA, test results were determined by two methods: 1. CrudeEs-Test = the percent binding obtained with the Es" antigen dot; 2. NetEs-Test = the difference between the Es“ and Es" antigen dots, calculated by subtracting the percent binding with the Es' antigen dot from the percent binding with the Es+ antigen dot. Individual cut off values (mean + 2 standard deviations) were determined for each CrudeEs-Test or NetEs-Test using pre-infection sera from 8 experimentally infected pigs. The means and standard deviations for percent binding of pre- infection sera and cut off values for each test are shown in Table 4-1. For performing DB, two batches of protein A were 1251-iodinated separately. Because the 125I-iodination efficiency of the two batches of protein A were different, cut off values for each batch were determined individually using the same sera (shown as dot blot 1 and 2 in Table 4-1). The cut off value of dot blot 1 was used for experiments evaluating sera from experimentally infected pigs. The cut off value of dot blot 2 was used for experiments evaluating field serum samples. The positive/negative results of each CrudeEs-Test and NetEs-Test were determined as follows: in each test, sera showing percent binding equal to or greater than the respective cut off values were classified as positive; and sera showing percent binding less than the respective cut off values were classified as negative. 159 Comparison of specificities of the anti-pig IgG and protein A reagents by western blot. To evaluate whether there were differences in the specificities of the rabbit anti-pig IgG and protein A reagents, western blotting using the 100K antigen was performed in parallel using HRP conjugated anti- pig IgG and 1251 labeled protein A. The Es+ and E8' 100K antigens were mixed with sample buffer (100 mM Tris (pH 6.8), 3.3% (w/v) SDS, 16.7% (v/v) glycerol, 4.2% (v/v) fi-mercaptoethanol, and a few crystals of bromophenol blue) and placed in boiling water for 3 minutes. SDS-polyacrylamide gel electrophoresis (PAGE) was performed with these antigens using a 9% gel (Smith, 1984). The proteins were transferred to NC's electrophoretically by the method of Towbin et al. (1979). The proteins transferred to the NC's were blocked with blot blocking buffer at 4 °C over night. The NC's were incubated with either immune or nonimmune reference sera diluted 1:100 with blot incubation buffer at 25 °C for two hours, followed by washing 3 times with blot washing buffer, 10 minutes each wash. For 196 western blotting, the NC's were incubated with 1:2,000 dilution of HRP conjugated rabbit anti-pig IgG (Sigma Chemical Co.) diluted with blot incubation buffer for another one hour and washed 3 times with blot washing buffer. The NC's were finally incubated for two hours with triethanolamine buffered saline (pH 7.5) containing 0.05% 4- chloro-l-naphthol (Sigma Chemical Co.) and 0.01% hydrogen peroxide. 160 For protein A western blotting, the NC's were incubated 'with 125I-protein A (105 CPM/ml in blot incubation buffer) at 25 °C for one hour. Following incubation, the NC's were washed 3 times with blot washing buffer and air dried. Direct autoradiography was performed with the dried NC's using Kodak x-omat RP film, over night exposure at -80 «3. Protein standards used for molecular weight (MW) markers (SDS-6H; Sigma Chemical Co.) consisted of bovine erythrocyte carbonic anhydrase (MW 29,000), ovalbumin (MW 45,000), bovine plasma albumin (MW 66,000), rabbit muscle phosphorylase B (MW 97,400), Escherichia coli fi-galactosidase (MW 116,000) and rabbit muscle myosin (MW 205,000). The transferred MW standards were visualized by India ink stain (Hancock & Tsang, 1983). western blot analysis of field samples. Of the 140 field sera tested by the two ELISA's, DB and IHA, 63 samples were selected for western blot analysis so that representative numbers of sera yielding different combinations of results in the three new assays (ProA- and IgG-ELISA's and DB) were included. The selection procedure was as follows: The 140 sera were stratified into 6 groups by combinations of the positive/negative results of ProA-ELISA CrudeEs-Test, IgG- ELISA CrudeEs-Test and DB CrudeEs-Test (all ProA-ELISA positive sera were also positive in IgG-ELISA). The cut off values used to stratify the sera were the mean OD or percent binding of pre-infection sera from the 8 experimentally 161 infected pigs plus 3 standard deviations, instead of 2 standard deviations, to increase test specificity so that only highly positive samples in each assay could be classified as positive in the selection process. A similar number of sera were selected from each of the 6 combinations. The procedures of SDS-PAGE and western blotting were done as described above with the following modifications. SDS-PAGE was performed using a 9% gel consisting of two wide lanes, one with Es+ and the other with E3“ 100K antigen, and a lane for MW standards. After electrophoresis and electrotransfer of the proteins to NC, the NC was blocked with blot blocking buffer at 4 “C over night. The NC was then cut into narrow strips of Es+ and Es' 100K antigens. The strips were incubated individually with the serum sample, diluted 1:100 with blot incubation buffer, at 25 °C for two hours (one Es+ and one E3“ antigen strip for each sample), and washed 3 times with blot washing buffer, 10 minutes each wash. The NC strips were incubated with HRP conjugated rabbit anti-pig IgG for one hour, then washed 3 times with blot washing buffer, and were then incubated for two hours with triethanolamine buffered saline (pH 7.5) containing 0.05% 4-chloro-1-naphthol and 0.01% hydrogen peroxide. The blots were evaluated for the presence of E. suis specific antigen bands at MW's 24, 37, 47, 71, 81, 101 and 106 kd, as presented in Chapter 3. 162 Statistical analysis. Statistical analysis was done using the computer program StatView (version 4.0, Abacus Concepts, Berkeley, CA). Chi-square contingency analysis was performed to evaluate the associations between each positive/negative result of the two ELISA's, the DB and the IHA, and the associations of the two ELISA's, the DB and the IHA positive/negative results with the detection of each E. suis specific antigens in the IgG western blots. For 2x2 contingency analysis with expected values less than 5, a continuity correction was made (Gill, 1978). A P value less .than 0.05 was considered significant. 163 Results Comparison of western blots using anti-pig IgG and Protein .A. The results of western blots using Es+ and Es' 100K antigens, immune and nonimmune reference sera, and HRP conjugated rabbit anti-pig IgG and 1251 labeled protein A are summarized in Figure 4-1. No differences were detected in the bands present in the western blots developed with either rabbit anti-pig IgG or Protein A. Serum samples from experimentally infected pigs. Triplicate samples of pre- and post-infection sera obtained from 8 experimentally infected pigs were tested by ProA- and IgG-ELISA, DB and IHA. The positive/negative results of each assay are summarized in Table 4-2. The two ELISA's and DB exhibited a high degree of sensitivity for detecting antibody responses in post-infection sera whether evaluated by CrudeEs-Test (94% to 96%) or NetEs-Test (91% to 96%). The pre-infection sera were negative in all tests, except for one sample that was positive in the IgG-ELISA NetEs-Test. Negative results with post-infection sera in the two ELISA's and DB tended to occur with sera collected during the early stage of infection. However, more than half of the 1 WPI sera were positive in the ProA- and IgG-ELISA's and DB (71, 63 and 88% sera were positive by CrudeEs-Test, and 57, 75 and 88% sera were positive by NetEs-Test, respectively). Eighty- 164 Figure 4-1. Comparison of antibody detection by protein A and rabbit anti-pig 196 by western blot. Western blotting were performed using Eperythrozoon suis positive (33*) and negative (Es') 100K antigens and immune (IS) and nonimmune (NS) reference sera. Antibodies bound to antigens were detected by 1‘zf’I-protein A (105 CPM/ma of the buffer) or horse radish peroxidase (HRP) conjugated rabbit anti-pig IgG (1:2000 diluted in the buffer). u”‘I-protein A treated nitrocellulose membranes (NC's) were exposed to Kodak x-omat RP film over night at -80 'C. HRP conjugated anti-pig IgG treated NC's were incubated with hydrogen peroxide containing 4-chloro-1-naphthol for two hours. MWM - Molecular weight marker. 165 125l-protein A HFlP conjugated anti-pig IgG Es+anfigen ES'anfigen Es+anfigen Es’anfigen MWM IS NS IS NS IS NS IS NS '.| Figure 4-1. 166 2:0 9.0m corona—5.88 .6 .33 on... n a .cozooecmumoo axon; go .3832 u H .3388 .6 Coo-:3: 03:82.83 ozgmoo eo Coo—:3: of. u H Scam ego So: misfits»? at 3 one tonne: 9.95 3383 3:83 .ootoo 53085.33 mcoEm pot—2, moi—tom eo .352. .33 och 63332.. 203 money acetate an 33 N Eot “538:8 Sow 53085.05 icon—Boo N .3: can 63 you {flow—6m. :. u .. 3.}... 33m cmhm 3: a «Q 3. 38.. 8:; 2.33 v). v). .3. v? .E. . V). «3 N. So So So So So me Sc 8 m3 m; {m m; m3. m3 3m m «.3 as . as a; m3 m3 Sm c as as as . a: as at. ‘ 8» e a; as we as m8 m8 «a m to to to 2.: ts to to N as a: a: So 8m Ce 2m E; F o to o to o to o P: o to 8o 1. So 8:03... 883 umohImmuoz amp-.3330 30.7%“.qu umoHImmoo-zo «mohImmuoz 30.7338 833 <1. 83 58 2:: ‘6me 5.3556980... «omens ecu ~03 you .m.83 3.3558059. 32.02. .3 05 J03 Doc 0:... m.88 50.855.88.05 .0050... 0:0 .03 000 6.4.9490... 0cm -2 503.3 (0.5 .3 0055.300 mm 0.0m 0.0.. 050050.. .o 050.005 NI? 030... 179 80.08000 8:08.800 n . . .Amod v a. 000.00.000.00 0:00...:m.m u .. .8 ... .00 ... 3.0.. 8.0.. . 408.. 3.0.. .00... .000 5:080:00 0404. 844. 008. . 8.0. .80. 88. 800. 00.0) n. <:. .80.. .80.. .08.. .08.. . .8. ~08. 300.. .000 8:88:00 88.. 8. .. 0. . .. 0000. .008. 08.. 080.. 00.0, 0 30.00.02 .08.. 3.0.8 .08.. .4. .. 3.0.. 3.4.. .00... .000 0:88:00 008. .88. 800. 448. .0000. . .08.v 0:... 00.0) .. 00.000020 0.0 .00 A .0... .08.. .040. .08.. 2.04.3 84.. .40... .000 6:88:00 .. 4004. 1.000. .408. .08. .408. .008. 080.. 00..) n. 00.3.02 .400. 640.. A .0... .40... 3.83 3.4. 340.. .000 5:080:00 88.. 00... .044. 88. 8.0. . .08. 080.. 00.2. n. 8.000020 830.09 8... 840.. .48.. 80... 8.0.. 080.. 34.0.. 0000 6:003:00 88. .0400. . .80. . .0 .0. . .000.v .480. 88.. 00.0., .. 0088.02 3.0.. .08.. .8... 3.0.. .84.. .80.. 3.0 .. .0000 8:080:00 88. .020. 88. 04.0. ..000.v . .08.v :8. 00..., 0 00.000020 830-8... 00. .0. .0 z. .4 .0 40 0.0.005 .8. .300... 8:3 0000:. 00.00:. 0.3. .0 0:. .0 30.18.02 3.30.0: .30... 800003 on. 5 0:00 000.80 0.20000 0.50 .m 3000 .0 5.000000 05 0:0 .21.. >080 5005028080.. .0205 0:0 00.0 .00 029.300. 0:0 -2 5000.0. <0... .3 0058.300 00 00.0800 0.0... .0 00.000. 0200005025000 05 c0050.. 0.3.000 8:09.800 0.8.00.5 0+ 0.005 180 Discussion REAGENTS In the developed assays, anti-pig IgG or protein A were used to detect immunoglobulins bound to antigens. Protein A has been shown to bind to 196 of most mammalian species and also to IgM, IgA and 193 with lower reactivity (Lindmark et al., 1983). It has been postulated that protein A binding to IgG occurs via the Fc portion of IgG and binding to other classes of immunoglobulins is mainly via the Fab portion. In pigs, protein A binds effectively to the main IgG subclasses, 1961 and Ing (more than 90% reactivity to both subclasses), and less effectively to IgM (35 to 40% reactivity) and 19A (23% reactivity) (Lindmark et al., 1983). The anti-pig IgG used in the study was affinity purified from the sera of rabbits that was immunized with the whole molecule of pig IgG. Since the light chain (kappa or lambda chain) is shared by all classes of immunoglobulins, antibodies against the whole molecule of 196 could potentially react to all classes of immunoglobulins. Therefore, both reagents, protein A and anti-pig 196, could recognize IgM to some degree but the degree of specificities of the two reagents for detecting IgM are probably different. In the first experiment, the band appearances in the western blots using HRP conjugated anti-pig IgG and 125I- protein A were compared to evaluate the specificity and the 181 suitability of two reagents for use in the immunoassays. No differences were found in the specificity of anti-pig IgG and protein A ability to detect antigen bands in western blots, for the band pattern and band intensity in the western blots using either anti-pig IgG or protein A were identical. EXPERIMENTAL INFECTION In the experimental infection, the proposed ELISA's and DB detected an antibody response sooner in post-infection sera from experimentally infected animals than the IHA. Two test methods, CrudeEs-Test and NetEs-Test, were evaluated in each ELISA and DB. The two tests showed similar sensitivities for detecting an antibody response in post- infection sera of experimentally infected pigs. CrudeEs-Test represented the IgG binding to the Es+ antigen, while NetEs- Test represented the difference of IgG binding to Es+ and Es“ antigens, thus more specifically representing 196 binding to E. suis specific antigens. Therefore, NetEs-Test should eliminate the effects of autoantibodies induced by E. suis infection and also erythrocyte antigens in the inoculum. It is assumed that the increase in CD or percent binding values in NetEs-Test of the two ELISA's or DB could represent a specific IgG immune response in the experimentally infected pigs with E. suis. 182 Indeed, the autoantibody production in pigs experimentally infected with E. suis was demonstrated in Figure 4-2. The increase in on values and percent binding of the two ELISA's and DB using the Es‘ antigen preparation after infection are potential evidence of an immune response to host derived proteins. Similar observations were reported in experimental infections with Anaplasma and Babesia when infected blood was used as an inoculum (Barbet et al., 1983; Budden & Dimopoullos, 1977; Callow & Dalgliesh, 1982; Goodger et al., 1985). It was suggested that isoerythrocyte antibodies were produced as a result of host immune reaction to isoerythrocyte antigens in the inoculum (Barry et al., 1986; Dimmock, 1973; Duzgun et al., 1988). In studies directed at the development of immunoassays for hemoparasitic infection using crude antigen preparations, it is expected that an apparent high sensitivity for detecting post infection sera may be interpreted by a reaction of isoerythrocytic antibodies to impurities of antigen preparation. Budden and Dimopoullos (1977) suggested that a capillary-tube agglutination test developed for anaplasmosis (Ristic, 1962) had detected only antibodies directed at erythrocytes, since purified Anaplasma bodies were not agglutinated by the infected serum, while erythrocyte stromata separated from Anaplasma bodies was agglutinated by the infected serum. In addition to isoerythrocyte antibodies directed at non-self erythrocytes, the development of autoantibodies to host self erythrocytes has been observed as 183 a form of cold agglutinin in hemoparasite infections such as eperythrozoonosis, haemobartonellosis, anaplasmosis, malaria and babesiosis, which are also associated with anemia (Bellamy et al., 1978; Oki & Miura, 1970; Schroeder & Ristic, 1965; Soni & Cox, 1974; Soni & Cox, 1975; Thoongsuwan & Cox, 1973; Zachary & Smith, 1985; Zulty & Kociba, 1990). To understand the effect of antibodies directed at erythrocytes in diagnosis for bovine anaplasmosis, Duzgun et al. (1988) developed a similar ELISA system to the NetEs-Test in the present study, using 2 different antigens prepared before and after infection from the same animals (named positive and negative antigens, respectively). lThe ”net reading", calculated by subtracting the negative antigen result from the positive antigen result, was evaluated for A. marginale infection. This test showed excellent specificity for it could differentiate between the sera of animals experimentally infected with B. bovis or B. bigemina. If the negative antigen test results were not subtracted from the positive antigen test results, 37% of the sera would have been falsely classified as positive for A. marginale infection. In swine eperythrozoonosis, a similar test system using positive and negative antigens in ELISA has been developed using lysates of the whole noninfected and infected erythrocytes as ELISA coating antigens (Schuller et al., 1990). The N-Ag values plotted in Figure 4-2 illustrate autoantibody development during the course of infection. The 184 N-Ag values demonstrated an initial peak at 4 WPI in each ELISA and DB. Thus an observed rapid increase in the CD or percent binding value of CrudeEs-Test in either ELISA or DB during the first 4 WPI was largely reflected by the autoantibody development. This demonstrates the possible bias, caused by autoantibodies, in the interpretation of specific humoral immune response against hemoparasites if antibodies are measured by an assay using a crude antigen prepared only from infected blood (Es+ antigen). In the two ELISA's and DB, more than 90% of post- infection sera of experimentally infected animals were positive by both CrudeEs-Test and NetEs-Test. Most of the negative results were observed in the sera of early infection stage, presumably because of the time required for the host immune system to respond to E. suis infection. With the IHA, 1 WPI sera were all negative. It has been reported that IHA measures IgM titer. Generally, with humoral immune responses in a primary infection, IgM levels rise first followed by increasing 196 levels. The failure of IHA to detect 1 WPI sera may be due to its inability to detect low concentrations of IgM. However, considering the observation in Chapter 3 that the IHA antigen consisted of mainly host proteins and that antibodies reacting to the IRA antigen were mainly autoantibodies, the failure of IHA to detect 1 WPI sera could be due to a delay of autoantibody production after B. suis infection. This is consistent with the hypothesis that autoantibody production may be induced by alterations of 185 erythrocyte membrane proteins or the exposure of hidden erythrocyte antigens due to E. suis infection (Zachary & Smith, 1985). Evidence supporting the hypothesis that IHA may detect only autoantibodies is found in Figure 4-2; the observation that IHA titer development was identical to N-Ag value development in both ELISA's suggests that IHA and ELISA's using Es‘ antigen detected the same antibody, presumably autoantibodies to host derived antigens. Hsu et al. (1992) observed a high correlation (r=0.8601) between log IHA titers and ELISA OD values using an antigen prepared from an infected erythrocyte lysate. In their procedure, lysed infected erythrocytes were centrifuged at 1000g and the supernatant discarded. The pellet was suspended in PBS and frozen. The frozen suspension was thawed and centrifuged at 110009. The resultant supernatant was used for the coating antigen in ELISA. Their results are inconsistent with the results obtained here, in which the log IHA titers and OD values of CrudeEs-Test and NetEs-Test in the two ELISA's and DB did not take similar courses during the experimental period. The correlation between the log IHA titers and the OD values of their ELISA may indicated that their ELISA was primarily detecting autoantibodies. 186 FIELD 833A The results of testing sera from naturally exposed pigs showed considerable variation among the 4 assays. The numbers of positive samples in CrudeEs-Test and NetEs-Test in each ELISA's or DB were quite different. This indicates that a considerable amount of antibodies reactive to the Es' antigen (autoantibodies or antibodies crossreactive to host antigens) had been produced in pigs raised under field condition, regardless of E. suis infection status. A possible explanation of this observation is that animals in the field are exposed to various etiologic agents which could induce development of agent specific antibodies, antibodies crossreactive to host antigens due to the similarity of agents and host antigens, and sometimes non-specific antibodies via polyclonal B cell activation. Induction of autoantibodies as a result of polyclonal B cell activation has been commonly observed in parasitic diseases including trypanosomiasis, leishmaniasis, malaria, babesiosis and schistosomiasis (Cohen, 1982). Therefore, it is likely that antibodies reactive or crossreactive to host antigens may develop in pigs under field conditions. This hypothesis was supported by the result of the subsequent western blot experiments with the field serum samples. In several western blots, strong bands were observed at different molecular weight positions from E. suis specific antigens (data not shown). These antigens were 187 determined to be derived from.the host because the bands were observed in both Es+ and ES' antigen preparations. Under field conditions, the presence of anti-erythrocyte antibodies (IgM and IgG) not associated with E. suis infection has been observed in sows with nonhemolytic anemia (van Leengoed et al., 1992). Disagreements between experimentally infected animals versus animals infected in field conditions have been reported previously with A. marginale. Serotests for bovine anaplasmosis such as complement fixation, capillary tube agglutination, card agglutination and indirect fluorescent antibody tests were more accurate for diagnosing experimental infections than for diagnosing field infections (Gonzalez et al., 1978; Rogers, 1971). The IgG-ELISA of this study classified a greater number of field sera as positive compared to the ProA-ELISA and the DB, regardless of whether the results were evaluated by CrudeEs-Test or NetEs-Test (Table 4-3). The percentage of the positive sera in IgG-ELISA CrudeEs-Test and NetEs-Test were 91 and 56%. On the other hand, the percentage of the positive sera in ProA-ELISA CrudeEs-Test and NetEs-Test were 50 and 17%. Since true status of E. suis infection of field sera is difficult to assess, the ability of those tests to detect true infection status (sensitivity and specificity) could not be evaluated in this study and thus neither could the true percentage of the E. suis positive sera be determined (true prevalence). Unfortunately, definite diagnosis of E. suis infection is possible only by detection 188 of parasitemia in acutely affected, febrile animals (Smith, 1981). The duration of the acute parasitemic phase is short and the organism can not be detected in animals that have recovered from the febrile stage (Splitter, 1951). Even though there was a disagreement of the number of positive sera detected by the two ELISA's and DB, the three assays showed a similar ability to distinguish positive and negative sera, based on the contingency analysis of the positive/negative results (Table 4-4). In another words, positive sera determined by one of the two ELISA's or DB tended to be positive in other tests. WESTERN BLOT ANALYSIS When the recognitions of each E. suis specific antigen by the field serum samples in the western blots were compared, all the field samples recognized the 24 and the 106 kd antigen bands with a few exceptions (Table 4-7). On the other hand, the 71 and the 81 kd antigens were recognized only by 15 and 11 of the 63 sera, respectively. This indicates that there were considerable variations in the E. suis antigen recognition by the field samples. As discussed above, antibodies produced in response to other infectious agents may cross-react to some extent with E. suis antigens, and these cross-reactions may contribute to this variation. 7 189 There ‘were significant associations between the detections of the 37, 47, 71, 81 and 101 Rd E. suis antigen bands in the western blots and the positive/negative results ‘with the two ELISA's and DB (Table 4-8). The NetEs-Test's of the two ELISA's were positively associated with all 5 antigens, indicating that positive results of the NetEs- Test's with both ELISA's were strongly associated with the presence of E. suis specific antibodies in the sera. This was further confirmed by the observations reported in Table 4-6, that NetEs-Test of either ELISA exhibited a higher specificity for detecting sera that reacted with 6 or more of the selected 7 E. suis specific antigen bands, compared to CrudeEs-Test. Even though there were significant associations between the detections of the 37, 47, 71, 81 and 101 kd E. suis antigen bands in the western blots and the positive/negative results with the two ELISA's and DB, the levels of association (contingency coefficients) were relatively low. The criteria used for selection of E. suis specific antigen bands was the band appearance in the autoradiograph of radiolabeled 100K antigen preparation (refer to Chapter 3). Since radiolabeling of E. suis was performed with in vitro culture for only 4 hours, it could be assumed that only proteins with sufficiently rapid turnover were radiolabeled effectively and proteins with a relative slow turnover were not. The 100K antigen used in the two ELISA's and DB contained the selected 7 E. suis antigens and possibly slow 190 turnover antigens. The results of the two ELISA's and DB reflected antibody reactions to both the rapid (the 7 E. suis antigens) and slow turnover antigens. Therefore, relatively low association levels of the two ELISA's and DB results and the reaction to the selected 7 E. suis antigen bands in western blots may be explained by the assumption that some of the two ELISA's and DB results are due to antibody reactions with slow turnover antigens. CONCLUSION Recently, E. suis DNA detection techniques have been developed using whole organism DNA hybridization (Oberst et al., 1990b) and recombinant DNA hybridization (Oberst et al., 1990a). These techniques detected E. suis specific DNA from the high salt lysate of whole blood samples, and apparently offer high sensitivity and specificity for detecting E. suis infection. Unfortunately at this time, a positive result with DNA hybridization indicates the presence of a current infection and does not detect past infections. These techniques could potentially lead to the development of a more powerful diagnostic technique using the DNA polymerase chain reaction (PCR), in which very low levels of infection could be detected by amplification of E. suis specific DNA. Even though PCR provides a more powerful diagnostic tool, serological diagnostic methods are still important for 191 conducting disease prevalence surveys. Unlike PCR, which requires sophisticated laboratory environment and relatively expensive equipment, ELISA can be performed at mmch less expensive and con be adapted for field use, allowing for quick assessment of antibody status. 192 References 1. Barbet, A. F., Anderson, L. W., Palmer, G. B. and Mcguire, T. C. (1983): Comparison of proteins synthesized by two different isolates of Anaplasma marginale.' Inf. Immun., 40, 1068-1074 2. Barry, D. N., Parker, R. J., De Vos, A. J., Dunster, P. and Rodwell, B. J. (1986): A microplate enzyme-linked immunosorbent assay for measuring antibody to Anaplasma .marginale in cattle serum. Aust. vet. J., 63, 76-79 3. Bellamy, J. E. C., MacWilliams, P. S. and Searcy, G. P. (1978): Cold-agglutinin hemolytic anemia and Haemobartonella canis infection in a dog. JAVMA, 173, 397-401 4. Budden, J. R. and Dimopoullos, G. T. (1977): Finite purification of Anaplasma marginale: Serologic inactivity of the Anaplasma body. Am. J. vet. Res., 38, 633-636 5. Callow, L. L. and Dalgliesh, R. J. (1982): Immunity and immunopathology in babesiosis. In Immunology of parasitic infections, 2nd ed., Ed. S. Cohen and K. S. warren, Blackwell Scientific Publications, Oxford, UK, pp. 475-526 6. Cohen, S. (1982): Survival of parasites in the immunocompetent host. In Immunology of parasitic infections, 2nd ed., Ed. 8. Cohen and K. S. Warren, Blackwell Scientific Publications, Oxford, UK, pp. 138-161 7. Dimmock, C. K. (1973): Blood group antibody production in cattle by a vaccine against Babesia argentina. Res. vet. Sci., 15, 305-309 8. Doyle, L. P. (1932): A rickettsia-like or anaplasmosis- like disease in swine. JAVMA, 81, 668-671 9. Duzgun, A., Schunter, C. A., Wright, I. G., Leatch, G. and Waltisbuhl, D. J. (1988): A sensitive ELISA technique for the diagnosis of Anaplasma marginale infections. Vet. Parasitol., 29, 1-7 10. Fraker, P. J. and Speck, J. C., Jr. (1978): Protein and cell membrane iodinations with a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a-diphenylglycoluril. Biochem. Biophys. Res. Commun., 80, 849-857 11. Gill, J. L. (1978): Design and analysis of experiments in the animal and medical sciences. Iowa State University Press, Ames, IA 193 12. Gonzalez, F. F., Long, R. F. and Todorovic, R. A. (1978): Comparisons of the complement-fixation, indirect fluorescent antibody, and card agglutination tests for the diagnosis of bovine anaplasmosis. Am. J. Vet. Res., 39, 1538-1541 13. Goodger, B. V., Wright, I. G. and Waltisbuhl, D. J. (1985): Babesia Davis: the effect of acute inflammation and isoantibody production in detection of babesial antigens. Experimentia, 41, 1577-1579 14. Hancock, K. and Tsang, V. C. W. (1983): India Ink Staining of proteins on nitrocellulose paper. Analytical 15. Henry, S. C. (1979): Clinical observation on Eperythrozoonosis. JAVMA, 174, 601-603 16. Hoffmann, R., Schmid, D. O. and Hoffmann-Fezer, G. (1981): Erythrocyte antibodies in porcine eperythrozoonosis. vet. Immunol. Immunopathol., 2, 111-119 17. Hsu, F. S., Liu, H. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1990): Evaluation of an enzyme-linked immunosorbent assay for the detection of Eperythrozoon suis antibody in swine. In Proc. 11th International Pig veterinary Society congress, Lausanne, Switzerland pp. 315 18. Hsu, F. S., Liu, H. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1992): Evaluation of an enzyme-linked immunosorbent assay for detection of Eperythrozoon suis antibodies in swine. Am. J. Vet. Res., 53, 352-354 19. Lindmark, R., Thorén-Tolling, K. and Sj6quist, J. (1983): Binding of immunoglobulins to protein A and immunoglobulin levels in mammalian sera. J. Immunol. Methods, 62, 1-13 20. Oberst, R. D., Hall, S. M., Jasso, R. A., Arndt, T. and Wen, L. (1990a): Recombinant DNA probe detecting Eperythrozoon suis in swine blood. Am. J. Vet. Res., 51, 1760-1764 21. Oberst, R. D., Hall, S. M. and Schoneweis, D. A. (1990b) : Detection of Eperythrozoon suis DNA from swine blood by whole organism DNA hybridizations. Vet. Microbiol., 24, 127-134 22. Oki, Y. and Miura, K. (1970): Characteristics and roles of red cell autoantibodies in equine infectious anemia. Jpn. 194 23. Ristic, M. (1962): A capillary tube-agglutination test for anaplasmosis - A preliminary report. JAVMA, 141, 588-594 24. Robb, A. D. (1943): Ictero-anemia in growing swine. vet. Med. Sm. Anim. Clinic., 38, 271-274 25. Rogers, R. J. (1971): Bovine anaplasmosis: an evaluation of the complement-fixation and capillary tube- agglutination tests and the incidence of antibodies in northern Queensland cattle herds. Aust. vet. J;, 47, 364-369 26. Schroeder, W. F. and Ristic, M. (1965): Anaplasmosis XVII. The relation of autoimmune processes to anemia. Am. J. Vat. Res., 26, 230-245 27. Schuller, W., Heinritzi, K., Al-Nutha, S., Kolbl, S. and Schuh, M. (1990): Serological studies by means of CF- tests and ELISA for evaluation of antibodies against Eperythrozoon suis. Berl..Mfinch. Tierarztl. WSchr., 103, 9- 12 28. Sisk, D. B., Cole, J. R. and Pursell, A. R. (1980): Serologic incidence of eperythrozoonosis in Georgia swine. Proc. Annu. fleet. Am. Assoc. vet. Lab. Diag., 23, 91-100 29. Smith, A. R. (1977): Eperythrozoonosis. Veterinary professional topics, university of Illinois: Swine, 5, 2-4 30. Smith, A. R. (1981): Eperythrozoonosis. In Diseases of swine, Ed. B. S. A. D. Leman R. D. Glock, W. L. Mengeling, R. H. C. Penny and E. Scholl, Iowa State University Press, Ames, IA, pp. 683-687, 31. Smith, A. R. and Rahn, T. (1975): An indirect hemagglutination test for the diagnosis of Eperythrozoon suis infection in swine. Am. J. Vet..Res., 36, 1319-1321 32. Smith, B. J. (1984): SDS polyacrylamide gel electrophoresis of proteins. In Proteins: Methods in molecular biology, vol.1, Ed. J. W. Walker, Humana Press, Clifton, NJ, pp. 41-55 33. Soni, J. L. and Cox, H. W. (1974): Pathogenesis of acute avian malaria: I: Immunologic reactions associated with anemia, splenomegaly, and nephritis of acute Plasmodium gallinaceum infections of chickens. Am. J. Trop. Med. Hyg., 23, 577-585 195 34. Soni, J. L. and Cox, H. W. (1975): Pathogenesis of acute avian malaria: II: Anemia mediated by a cold-active autohemagglutinin from the blood of chickens with acute Plasmodium gallinaceum infection. Am. J. Trop. Med. Hyg., 24, 206-213 35. Splitter, E. J. (1951): Eperythrozoonosis in swine. Iowa St. Cbllege vet., 13, 77-81 36. Splitter, E. J. (1958): The complement-fixation test in diagnosis of eperythrozoonosis in swine. JAVMA, 132, 47-49 37. Splitter, E. J. and Williamson, R. L. (1950): Eperythrozoonosis in swine. A preliminary report. JAVMA, 116, 360-364 38. Thoongsuwan, S. and Cox, H. W. (1973): Antigenic variants of the haemosporidian parasite, Babesia rodohaini, selected by in vitro treatment with immune globulin. Ann. Trop..Med. Parasitol., 67, 373-385 39. Towbin, H., Staehelin, T. and Gordon, J. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci., USA, 76, 4350-4354 40. van Leengoed, L., van Dijk, J., Ham, M. and verheijden, J. (1992): Nonhemolytic anemia (NHA) in sows. In Proc. 12th International Pig Veterinary Society Congress, Hague, Netherlands pp. 394 41. Zachary, J. F. and Smith, A. R. (1985): Experimental porcine eperythrozoonosis: T-lymphocyte suppression and misdirected immune responses. Am. J. Vet. Res., 46, 821-830 42. Zulty, J. c. and Kociba, c. 'J. (1990): Cold agglutinins in cats with haemobartonellosis. JAVMA, 196, 907-910 OVERALL CONCLUSION In swine eperythrozoonosis, similar to other hemoparasitic diseases, the induction of autoantibodies against erythrocyte proteins is a major cause of confusion in the analysis of the organism proteins and results in poor specificity with serodiagnostic assays which use crude antigen preparations. The present study revealed a lack of specificity for detecting E. suis specific antibodies in the widely used IHA because of host derived impurities in the crude antigen preparation used in the test. The in vitro culture system developed in this study facilitated the incorporation of a radiolabeled methionine by E. suis, thus enabling the analysis of organism specific proteins. However, to maximize the advantages of in vitro culture, a continuous culture system, as established in Plasmodium falciparum, needs to be developed for E. suis. Further screening of culture conditions, including optimum incubation temperature and pH, the supplementation of fetal pig serum, and additional understanding of the basic biology of EL suis will be indispensable for establishing a continuous in vitro culture system for E. suis. The glycolytic activity of E. suis observed in vitro suggests a 196 197 potentially significant role of hypoglycemia in the in vivo pathophysiology of swine eperythrozoonosis. Among the several antigen preparations developed in this study, the 100K antigen preparation exhibited the greatest E. suis specificity. The specificity of the 100K antigen preparation, obtained by sequential differential centrifugation, was confirmed by the appearance of the E. suis antigen bands in western blots using only post infection sera and antigens prepared from parasitemic, blood. Specificity was further confirmed by the band identity in the western blots using the 100K antigen and the autoradiograph of the radiolabeled E. suis. The ProA- and IgG-ELISA, and DB developed in this study demonstrated high sensitivity for detecting antibodies in post-infection sera of experimentally infected pigs, regardless of using either CrudeEs-Test or NetEs-Test. However, impurities in the semi-purified 100K antigen still resulted in the detection of autoantibodies and isoerythrocytic antibodies. The NetEs-Test was developed to compensate for the detection of these antibodies and improved the detection of E. suis specific antibody responses. Indeed, subsequent western blot analysis that compared the results of the two ELISA's revealed that the NetEs-Test in either ProA- or IgG-ELISA was most significantly associated with the recognition of E. suis specific antigen bands in western blots. 198 Further purification of the 100K antigen preparation is recommended to eliminate interactions of host protein impurities in the antigen preparation with autoantibodies induced by E. suis infection. This would eliminate the necessity of using Es+ and Es’ antigen preparations in the ELISA's or DB. The development of a rapid, sensitive and specific serodiagnostic assay could lead to a better understanding of the nature of disease including the true prevalence of E. suis infection and help to resolve the questionable associations of E. suis infection with impaired growth and reproductive performance. The characterization of E. suis specific antigens also could lead to the development of an effective vaccine. OVERALL BIBLIOGRAPHY OVERALL BIBLIOGRAPHY 1. Adams, E. W., Lyles, D. I. and Cockrell, K. O. (1959): Eperythrozoonosis in a herd of purebred ~1andrace pigs. 2. Anthony, H. D., Kelley, D. C., Nelson, D. L. and Twiehaus, M. J. (1962): Suppressing Eperythrozoon infection in swine. vet..ued., 57, 702-703 3. Augsten, K. (1982): Identification of Eperythrozoon coccoides by scanning electron microscope. z . versuchstierk., 24, 169-170 4. Avakyyan, A. A., Dyakonov, L. P. and Nadtochey, G. A. (1973): Ultrastructure and nature of representatives of the genus Eperythrozoon (Schilling, 1928). Isv. Akad. Mauk SSSR, 1, 135-138 5. Baker, H. J., Cassell, G. H. and Lindsey, J. R. (1971): Research complications due to Haemobartonella and Eperythrozoon infections in experimental animals. Am. J. Pathol., 64, 625-656 6. Baljer, G., Heinritzi, K. and Wieler, L. (1989): Demonstration of Eperythrozoon suis antibodies in experimentally and naturally infected pigs with indirect hemagglutination. J. vet. Med., B36, 417-423 7. Barbet, A. F., Anderson, L. W., Palmer, G. H. and Mcguire, T. C. (1983): Comparison of proteins synthesized by two different isolates of Anaplasma marginale. Inf. Immun., 40, 1068-1074 8. Barry, D. N., Parker, R. J., De Vos, A. J., Dunster, P. and Rodwell, B. J. (1986): A microplate enzyme-linked immunosorbent assay for measuring antibody to Anaplasma marginale in cattle serum. Aust. vet. J., 63, 76-79 9. Bellamy, J. E. C., MacWilliams, P. S. and Searcy, G. P. (1978): Cold-agglutinin hemolytic anemia and Haemobartonella canis infection in a dog. JAVMA, 173, 397-401 199 200 10. Berkenkamp, s. D. and wescott, R. B. (1988): Arthropod transmission of Eperythrozoon coccoides in mice. Lab. Anim. Sci., 38, 398-401 11. Berrier, H. H. and Gouge, R. E. (1954): Eperythrozoonosis transmitted in utero from carrier sows to their pigs. JAVMA, 124, 98-100 12. Biberstein, E. L., Barr, L. M., Larrow, L. L. and Roberts, 3. J. (1956): Eperythrozoonosis of swine in New York State. cornell vet., 46, 288-297 . 13. Blood, D. C., Radostits, O. M. and Henderson, J. A. (1983): Veterinary Medicine. 6th ed., Bailliere Tindall, London, UK 14. Brownback, A. (1981): Eperythrozoonosis as a cause of infertility in swine. Vet. Med. Sm. Anim. Clinic., 76, 375 and 378 15. Brun, R. and Jenni, L. (1987): Salivarian trypanosomes: Blood stream forms (trypomastigotes). In In vitro methods for parasite cultivation, Ed. A. E. R. Taylor and J. R. Baker, Academic Press, London, UK, pp. 94-117 16. Budden, J. R. and Dimopoullos, G. T. (1977): Finite purification of Anaplasma marginale: Serologic inactivity of the Anaplasma body. Am. J. vet. Res., 38, 633-636 17. Callow, L. L. and Dalgliesh, R. J. (1982): Immunity and immunopathology in babesiosis. In Immunology of parasitic infections, 2nd ed., Ed. 8. Cohen and K. S. Warren, Blackwell Scientific Publications, Oxford, UK, pp. 475-526 18. Campbell, R. L. (1945): Icteric condition found in hogs in northern Missouri. N. Am. vet., 26, 347-348 19. Chaimanee, P. and Yuthavong, Y. (1979): Phosphorylation of membrane proteins from Plasmodium berghei-infected red cells. Biochem. Biophys. Res. cemmun., 87, 953-959 20. Clark, R. (1942): Eperythrozoon felis (sp. nov.) in a cat. J. S. Afr. vet. Med. Assoc., 13, 15-16 21. Claxton, M. and Kunesh, J. P. (1975): Eperythrozoonosis in swine. Iowa St. Uhiv. vet., 3, 822-83 22. Cohen, S. (1982): Survival of parasites in the immunocompetent host. In Immunology of parasitic infections, 2nd ed., Ed. S. Cohen and K. S. Warren, Blackwell Scientific Publications, Oxford, UK, pp. 138-161 201 23. Cooper, R. A. and Jandle, J. H. (1969): The selective and conjoint loss of red cell lipids. J. Clin. Invest., 48, 906-914 24. Cox, F. E. G. (1966): Acquired immunity to Plasmodium vinckei in mice. Parasitology, 56, 719-732 25. Cox, H. W. and Calaf-Iturri, G. (1976): Autoimmune factors associated with anaemia in acute Haemobartonella and Eperythrozoon infections of rodents. Ann. Trop. Med. Parasitol., 70, 73-79 26. Crandall, I. and Sherman, I. W. (1991): Plasmodium falciparum (human malaria)-induced modifications in human erythrocyte band 3 protein. Parasitology, 102, 335-340 27. Daddow, K. N. (1977): A complement fixation test for the detection of Eperythrozoon infection in sheep. Aust. veto J., 53’ 139-143 28. Daddow, K. N. (1979): The natural occurence in a goat of an organism resembling Eperythrozoon ovis. Aust. vet. J., 55, 605-606 29. Daddow, K. N. (1980): Culex annulirostris as a vector of Eperythrozoon ovis infection in sheep. vet. Parasitol., 7, 313-317 30. Daddow, K. N. (1982): The protection of lambs from Eperythrozoon infection while suckling Eperythrozoon ovis carrier ewes. vet. Parasitol., 10, 41-45 31. Davis, W. C., Talmadge, J. E., Parish, S. M., Johnson, M. I. and Vibber, S. D. (1978): Synthesis of DNA and protein by Anaplasma marginale in bovine erythrocytes during short- term culture. Inf. Immun., 22, 597-602 32. Dicke, W. E. (1934): Anaplasmosis-like disease in swine. vet. Med. Sm. Anim. Clinic., 29, 288 33. Dimmock, C. K. (1973): Blood group antibody production in cattle by a vaccine against Babesia argentina. Res. vet. Sci., 15, 305-309 34. Doyle, L. P. (1932): A rickettsia-like or anaplasmosis- like disease in swine. JAVMA, 81, 668-671 35. Doyle, L. P. (1945): Anaplasmosis-like disease in hogs. N. Am. vet., 26, 677 202 36. Duzgun, A., Schunter, C. A., Wright, I. G., Leatch, G. and Waltisbuhl, D. J. (1988): A sensitive ELISA technique for the diagnosis of Anaplasma marginale infections. Vet. Parasitol., 29, 1-7 37. Eliot, C. P. (1936): The insect vector for the natural tansmission of Eperythrozoon coccoides in mice. Science, 84, 397 38. Elko, E. E. and Cantrell, W. (1968): Phagocytosis and anemia in rats infected with Haemobartonella muris. J. Inf. Dis., 118, 324- 332 39. Enders, B. (1988): In vitro cultivation of certain parasitic protozoa: Biochemical data and technical requirements for mass production. In Parasitology in focus: Facts and Trends, Ed. H. Mehlhorn, Springer-Verlag, Berlin, Germany, pp. 702-718 40. English, P. R. and Smith, W. J. (1975): Some causes of death in neonatal piglets. vet. Ann., 15, 95-104 41. Ewers, W. H. (1971): Eperythrozoon mariba sp. nov. (Protophyta: Order Rickettsiales) a parasite of red blood cells of the flying fox (Peteropus macroti epularius) in New Guinea. Parasitology, 63, 261-269 42. Finch, S. C., Kawasaki, S., Prosnitz, L., Clemett, A. R., Jonas, A. M. and Perillie, P. E. (1968): Ethyl palmitate induced medical splenectony in rats. In Baker, H. J., Cassell, G. H. and Lindsey, J. R. (1971): Research complications due to Haemobartonella and Eperythrozoon infections in experimental animals. Am. J. Pathol., 64, 625- 656 43. Finerty, J. F., Hidalgo, R. J. and Dimopoullos, G. T. (1969): A. passive ihemagglutination. procedure for ‘the detection of Eperythrozoon infection in calves. Am. J. vet. Res., 30, 43-45 44. Foote, L. E., Brock, W. E. and Gallaher, B. (1951): Ictero-anemia, eperythrozoonosis, or anaplasmosis-like disease of swine proved to be caused by a filtrable virus. .N. Am. vet., 32, 17-23 45. Fraker, P. J. and Speck, J. C., Jr. (1978): Protein and cell membrane iodinations with a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a-diphenylglycoluril. Biochem. Biophys. Res. Commun., 80, 849-857 203 46. Geary, T. G., Divo, A. A. and Jensen, J. B. (1983): An in vitro assay system for the identification of potential antimalarial drugs. J. Parasitol., 69, 577-583 47. Gill, J. L. (1978): Design and analysis of experiments in the animal and medical sciences. Iowa State University Press, Ames, IA 48. Glasgow, L. A., Murrer, A. T. and Lombardi, P. S. (1974): Eperythrozoon coccoides. II. Effect on interferon production and role of humoral antibody in host resistance. Inf. Immun., 9, 266-272 49. Glasgow, L. A., Odugbemi, T., Dwyer, P. and Ritterson, A. L. (1971): Eperythrozoon. coccoides. I. Effect on the interferon response in mice. Inf. Immun., 4, 425-430 50. Gledhill, A. W., Bilbey, D. L. J. and Niven, J. S. F. (1965): Effect of certain murine pathogens on phagocytic activity. Br. J. Exp. Pathol., 46, 433-442 51. Gledhill, A. W. and Dick, G. W. A. (1955): The nature of mouse hepatitis virus infection in weaning mice. J. Path. Bac., 69, 311-320 52. Goff, W. L., Johnson, W. and Kuttler, K. L. (1986): .Anaplasma marginale, Eperythrozoon wenyoni: Lectin reactions ‘with bovine erythrocytes. Expt. Parasitol., 61, 103-113 53. Gonzalez, F. F., Long, R. F. and Todorovic, R. A. (1978): Comparisons of the complement-fixation, indirect fluorescent antibody, and card agglutination tests for the diagnosis of bovine anaplasmosis. Am. J. Vet. Res., 39, 1538-1541 54. Goodger, B. V. and Mahoney, D. F. (1974): Evaluation of the passive haemagglutination test for the diagnosis of Babesia argentina infection in cattle. Aust. vet. J., 50, 246-249 55. Goodger, B. V., Wright, I. G. and Waltisbuhl, D. J. (1985): Babesia bovis: the effect of acute inflammation and isoantibody production in detection of babesial antigens. Experimentia, 41, 1577-1579 56. Gothe, R. and Kreier, J. P. (1977): Aegyptianella, Eperythrozoon, and Haemobartonella. In Parasitic protozoa, vol.4, Ed. J. P. Kreier, Academic Press, New York, NY, pp. 251-294 204 57. Gulland, F. M., Doxey, D. L. and Scott, G. R. (1987): Changing morphology of Eperythrozoon ovis. Res. vet. Sci., 43, 88-91 58. Hall, S. M., Cipriano, J. A., Schoneweis, D. A., Smith, J. E. and Fenwick, B. W. (1988): Isolation of infective and non-infective Eperythrozoon suis bodies from the whole blood of infected swine. vet. Rec., 123, 651 59. Hancock, K. and Tsang, v. C. W. (1983): India Ink Staining of proteins on nitrocellulose paper. Analytical Biocnem., 133, 157-162 60. Hannon, P. G., Lochmiller, R. L., Mellen, J. W., Craig, T. M. and Grant, W. E. (1985): Eperythrozoon in captive juvenile collared peccaries in Texas. Wildlife Dis., 21, 439-440 61. Hart, L. T., Larson, A. D., Decker, J. L., weeks, J. P. and Clancy, P. L. (1981): Preparation of intact Anaplasma marginale devoid of host cell antigens. Current.Microbiol., 5, 95-100 62. Harvey, J. W. (1989): Erythrocyte Metabolism. In Clinical biochemistry of domestic animals, 4th ed., Ed. J. J. Kaneko, Academic Press, San Diego, CA, pp. 185-234 63. Heinritzi, K. (1989): Eperythrozoon suis infection as a multifactorial disease. Berl..Mfinch. Tierarztl. WSchr., 102, 337-342 64. Heinritzi, K., Jfingling, A., Kfihlmann, R. and Lbsch, U. (1990a): .Eperythrozoon suis-infection: Occurrence and importance of cold agglutinin. In Proc. 11th International Pig Veterinary Society congress, Lausanne, Switzerland pp. 329 65. Heinritzi, K., Peteranderl, W. and Plank, G. (1990b): Eperythrozoon-infection in pigs: Influence upon acid-base- balance, and the levels of glucose, lactic, and pyruvic acid in venous blood. Dtscb. tierarztl. WSchr., 97, 31-34 66. Heinritzi, K., Plank, G., Peteranderl, W. and Sandner, N. (1990c): Investigation on acid-base balance and carbohydrate metabolism in Eperythrozoon suis infected pigs. J. vet. Med., 37, 412-417 67. Henry, S. C. (1979): Clinical observation on Eperythrozoonosis. JAVMA, 174, 601-603 205 68. Hoffmann, R., Schmid, D. O. and Hoffmann-Fezer, G. (1981): Erythrocyte antibodies in porcine eperythrozoonosis. vet. Immunol. Immunopathol., 2, 111-119 69. Howard, G. W. (1975): The experimental transmission of Eperythrozoon ovis by masquitoes. In Proc. 33th Conference, British Society for Parasitology, pp. 71 70. Howard, R. J., Rodwell, B. J., Smith, P. M., Callow, L. L. and Mitchell, G. F. (1980): Comparison of the surface proteins and glycoproteins on erythrocytes of calves before and during infection with Babesia bovis. J. Protozool., 27, 241-247 71. Hoyte, H. M. D. (1962): Eperythrozoon teganodes sp. nov. (Rickettsiales), parasitic in cattle. Parasitology, 52, 527-532 72. Hruska, J. C., Kliewer, I. O. and Brock, W. E. (1968): Studies in tissue culture of Anaplasma marginale. In Proc. 5th National Anaplasmosis Conference, Stillwater, OK, pp. 21-24 73. Hsu, F. S., Liu, M. C., Chou, s. M., Zachary, J. F. and Smith, A. R. (1990): Evaluation of an enzyme-linked immunosorbent assay for the detection of Eperythrozoon suis antibody in swine. In Proc. 11th International Pig veterinary Society congress, Lausanne, Switzerland pp. 315 74. Hsu, F. S., Liu, M. C., Chou, S. M., Zachary, J. F. and Smith, A. R. (1992): Evaluation of an enzyme-linked immunosorbent assay for detection of Eperythrozoon suis antibodies in swine. Am. J. Vet. Res., 53, 352-354 75. Hung, A. L. and Lloyd, 8. (1985): Humoral immune response of sheep to infection with Eperythrozoon ovis. Res. vat. SCio, 39' 275-278 76. Hung, A. L. and Lloyd, S. (1989): Role of the spleen and resette-formation response in experimental Eperythrozoon ovis infection. vet. Parasitol., 32, 119-126 77. Hyde, C. L., Finerty, J. F. and Evans, C. B. (1972): Antibody and immunoglobulin synthesis in germfree and conventional mice infected with Eperythrozoon coccoides. Am. J. Trop. Med. Hyg., 21, 506-511 78. Ichijo, S., Hosokawa, A., Kin, T., Tojo, M. and Konishi, T. (1982a): Studies on Eperythrozoonosis in Sheep. I. Clinical and hematological observations on spontaneous and experimental cases. J. Jpn. vet. Med. Assoc., 35, 76-85 206 79. Ichijo, S., Hosokawa, 8., Kim, D.-H. and Konishi, T. (1982b): Scanning and transmission electron microscopic observation of Eperythrozoon ovis (E. ovis). Jpn. J. vet. Sci., 44, 127-132 80. Ilemobade, A. A. and Blotkamp, C. (1978a): Eperythrozoon ovis. I. Serological diagnosis of infection by the indirect immunofluorescent antibody test. Tropenmed. Parasitol., 29, 307-310 81. Ilemobade, A. A. and Blotkamp, C. (1978b): Eperythrozoon ovis. III. The effect of infection on blood pH, concentrations of pyruvate, lactate and glucose in blood, haematology, and on concurrent and superimposed Trypanosoma vivax infection in sheep. Tropenmed. Parasitol., 29, 443-450 82. Inada, I., Ichijo, S., Sarashina, T. and Osame, s. (1985): Studies on Eperythrozoonosis in sheep -Clinical, hematological and electronmicroscopic finding in cases experimentally infected with E. ovis-. 14, 195-209 83. Iralu, V. and Ganong, K. D. (1983): Agglutination of mouse erythrocytes by Eperythrozoon coccoides. Inf. Immun., 39, 963-965 84. Jansen, B. C. (1952): The occurrence of Eperythrozoon parvum Splitter, 1950 in South African swine. Understepoort J. vet Res., 25, 5-6 85. Jennings, A. R. and Seamer, J. (1956): A new blood parasite in British pigs. Nature, 178, 153-154 86. Jensen, J. B. (1983): Plasmodium. In In vitro cultivation of protozoan parasites, Ed. J. B. Jensen, CRC Press, Boca Raton, FL, pp. 155-192 87. Jensen, J. B. and Trager, W. (1977): Plasmodium falciparum in culture: Use of outdated erythrocytes and description of the candle jar method. J. Parasitol., 63, 883-886 88. Kawazu, 8., Nakamura, Y., Kamio, T., Fujisaki, K. and Minami, T. (1990): Enzyme-linked immunosorbent assay for detection of antibodies to Eperythrozoon wenyoni in cattle. Jpn. J. vet. Sci., 52, 1297-1300 89. Keeton, K. S. and Jain, N. C. (1973): Eperythrozoon wenyoni: A scanning electron microscope study. J. Parasitol., 59, 867-873 207 90. Kessler, R. H. and Ristic, M. (1979): In vitro cultivation of Anaplasma marginale: Invasion of and development in noninfected erythrocytes. Am. J. vet. Res., 40, 1774-1776 91. Kessler, R. H., Ristic, M., Sells, D. M. and Carson, C. A. (1979): In vitro cultivation of Anaplasma marginale: Growth pattern and morphologic appearance. Am. J. vet. Res., 40, 1767-1773 92. Kim, H. D. and McManus, T. J. (1971): Studies on the energy metabolism of pig red cells I. The limiting role of membrane permeability in glycolysis. Biochim. Biophys. Acta, 230, 1-11 93. Kingsley, K. and Hibbs, C. M. (1968): Eperythrozoonosis in swine: a review. vet. Med. Sm. Anim. Clinic., 63, 971-972 94. Kinsley, A. T. (1932): Protozoan like body in blood of swine. vet. Med., 27, 196 95. Kinsley, A. T. (1939): Anaplasmosis-like disease in swine. vet. Med., 34, 104 96. Kinsley, A. T. and Ray, J. D. (1934): Anaplasmosis-like disease in swine. JAVMA, 84, 391-392 97. Kocan, K. M., Stiller, D., Goff, W. L., Claypool, P. L., Edwards, W., Ewing, S. A., Mcguire, T. C., Hair, J. A. and Barron, S. J. (1992): Development of Anaplasma marginale in male Dermacentor andersoni transferred from parasitemic to susceptible cattle. Am. J. vet. Res., 53, 499-507 98. Kreier, J. P. and Ristic, IL. (1963): Morphologic, antigenic, and pathogenic characteristics of Eperythrozoon ovis and Eperythrozoon wenyoni. Am. J. vet. Res., 24, 488- 500 99. Kreier, J. P. and Ristic, M. (1968): Haemobartonellosis, eperythrozoonosis, grahamellosis and ehrlichiosis. In Infectious blood diseases of man and animals, Ed. D. weinman and M. Ristic, Academic Press, New York, NY, pp. 387-472 100. Kreier, J. P. and Ristic, M. (1984): Genus IV. Eperythrozoon Schilling 1928, 1854. In Bergey's manual of systematic bacteriology, vol.1, Ed. E. N. R. Krieg volume 1; J. G. Holt, editor-in-chief, Williams & Wilkins, Baltimore, MD, pp. 726-729 208 101. Lang, F. M., Ferr-ier, G. R. and Nicholls, T. J. (1986): Separation of Eperythrozoon ovis from erythrocytes. vet. Rec., 119, 359 102. Lang, F. M., Ferrier, G. R. and Nicholls, T. J. (1987): Detection of antibodies to Eperythrozoon ovis by the use of an enzyme-linked immunosorbent assay. Res. vet. Sci., 43, 249-252 103. Laskowski, L., Stanton, M. F. and Pinkerton, H. (1954): Chemotherapeutic studies of alloxan, dehydroascorbic acid, and related conpounds in murine haemobartonellosis. J. Inf. Dis., 95, 182-190 104. Lau, F. O. and Rosse, W. F. (1975): The reactivity of red blood cell membrane glycophorin with "cold-reacting" antibodies. Clinic. Immunol. Immunopathol., 4, 1-8 105. Levy, M. G. and Ristic, M. (1983): Cultivation and in vitro studies of Babesia. In In vitro cultivation of protozoan parasites, Ed. J. B. Jensen, CRC Press, Boca Raton, FL, pp. 221-241 106. Lindmark, R., Thorén-Tolling, K. and Sjoquist, J. (1983): Binding of immunoglobulins to protein A and immunoglobulin levels in mammalian sera. J. Immunol. 107. Lloyd, 8. and Soulsby, E. J. L. (1988): Immunological responses of the host. In Parasitology in focus: Facts and Trends, Ed. H. Mehlhorn, Springer-Verlag, Berlin, Germany, pp. 619-650 108. Maede, Y. (1978): Studies on feline haemobartonellosis. v. role of the spleen in cats infected with Haemobartonella felis. Jpn. J. vet. Sci., 40, 141-146 109. Maede, Y. (1979): Sequestration and phagocytosis of Haemobartonella felis in the spleen. Am. J. vet. Res., 40, 691-695 110. Maede, Y. (1980): Studies on feline haemobartonellos is . VI . Changes of erythrocyte lipids concentration and their relation to osmotic fragility. Jpn. J. Vat. Sci., 42, 281-288 111. Marble, D. W. and Hanks, M. A. (1974a): A. marginale grown in stable rabbit bone marrow cells. In Proc. 6th National Anaplasmosis Conference, Las Vegas, NV, pp. 53-54 209 112. Marble, D. W. and Hanks, M. A. (1974b): Anaplasmosis in primary rabbit bone marrow cells. In Proc. 6th National Anaplasmosis Conference, Las Vegas, NV, pp. 49-50 113. Marmoston, J. (1935): Effect of splenectomy on a latent infection, Eperythrozoon coccoides, in white mice. J. 114. Martin, B. J., Chrisp, C. E., Averill, D. R. and Ringler, D. H. (1988): The identification of Eperythrozoon ovis in anemic sheep. Lab. Anim. Sci., 38, 173-177 115. Mason, R. W., Manuel, M., Daddow, K. N. and Witt, D. J. (1981): Eperythrozoon ovis recorded in Tasmania in association with Heinz body development. Aust. vet. J., 57, 46 ° 116. Mason, R. W. and Statham, P. (1991): The determination of the level of E. ovis parasitemia in chronically infected sheep and its significance to the spread of infection. Aust. vet. J., 68, 115-116 117. Matamoros, I. A., Cox, N. M. and Moore, A. B. (1991): Effects of exogenous insulin and body condition on metabolic hormones and gonadotropin-induced follicular development in prepuberal gilts. J. Anim. Sci., 69, 2081-2091 118. Mazzola, V. and Kuttler, K. L. (1980): Anaplasma marginale in bovine erythrocyte cultures. Am. J. vet. Res., 41, 2087-2088 119. McCorkle-Shirley, 8., Hart, L. T., Larson, A. D., Todd, W. J. and Myhand, J. D. (1985): High-yield preparation of purified Anaplasma marginale from infected bovine red blood cells. Am. J. vet. Res., 46, 1745-1747 120. McKee, A. E., Ziegler, R. F. and Giles, R. C. (1973): Scanning and transmission electronmicroscopy of Haemobartonella canis and Eperythrozoon ovis. .Am. J. vet. Res., 34, 1196-1201 121. McLaughlin, B. G., Evans, C. N., McLaughlin, P. 8., Johnson, L. W., Smith, A. R. and Zachary, J. F. (1990): An Eperythrozoon-like parasite in llamas. JAVMA, 197, 1170-1175 122. Nakamura, Y. Personal communication, 1991 123. Neitz, W. O. (1937): Eperythrozoonosis in sheep. Onderstepoort J. vet Res., 9, 9-30 210 124. Nicholls, T. J., Veal, P. I. and Overend, D. (1989): The effect of artificial Eperythrozoon ovis infection on the growth rate of stressed and non-stressed sheep. Aust. vet. J., 66, 184-186 125. Nicholls, T. J. and Veale, P. I. (1986): A modified indirect immunofluorescent assay for the detection of antibody to Eperythrozoon ovis in sheep. Aust. vet. J., 63, 157-159 126. Nikolskii, 8. N. and Slipchenko, 8. N. (1969): Experiments in the transmission of Eperythozoon ovis by ticks H. plumbeum and Rh. bursa. Veterinariia, 5, 46 127. Nordelo, M. A. and Ysern-Caldentey, M. (1982): Abnormal bovine erythrocyte membrane proteins and glycoproteins during and after infection with Anaplasma marginale. Biochem. Biophys. Res. Commun., 104, 664-672 128. Oberst, R. D., Hall, 8. M., Jasso, R. A., Arndt, T. and Wen, L. (1990a): Recombinant DNA probe detecting Eperythrozoon suis in swine blood. Am. J. vet. Res., 51, 1760-1764 129. Oberst, R. D., Hall, 8. M. and Schoneweis, D. A” (1990b): Detection of Eperythrozoon suis DNA from swine blood by whole organism DNA hybridizations. vet. Microbiol., 24, 127-134 130. Oki, Y. and Miura, K. (1970): Characteristics and roles of red cell autoantibodies in equine infectious anemia. Jpn. J. vet Res., 32, 217-227 131. Ott, K. J., Astin, J. K. and Stauber, L. A. (1967): Eperythrozoon coccoides and rodent malaria: Plasmodium chabaudi and Plasmodium Berghei. Expt. Parasitol., 21, 68-77 132. Overés, J. (1969): Studies on Eperythrozoon ovis infection in sheep. Acta vet. Scand., 28 (Sup.), 1-147 133. Palmer, G. H. and McGuire, T. C. (1984): Immune serum against Anaplasma marginale initial bodies neutralizes infectivity for cattle. J. Immunol., 133, 1010-1015 134. Peters, D. and Wigand, R. (1955): Bartonellaceae. Bacterial. Rev., 19, 150-159 135. Peters, W. (1965): Comparative relationship between Eperythrozoon coccoides and Plasmodium berghei in the mouse. Expt. Parasitol., 16, 158-166 211 136. Plank, G. and Heinritzi, K. (1990): IDisseminated intravascular coagulation in porcine eperythrozoonosis. Berl..Mfinch. Tierarztl. Nechr., 103, 13-18 137. Pomerat, C. M., Frieden, E. H. and Yeager, E. (1947): Reticulo-endothelial immune serum (REIS). V. An experimental anemia in Bartonella infected rats produced by anti-blood immune serum. J. Inf. Dis., 80, 154-163 138. Pospischil, A. and Hoffmann, R. (1982): Eperythrozoon suis in naturally infected pigs: A light and electron microscopic study. vet. Pathol., 19, 651-657 139. Preston, K. 8. and Greve, J. H. (1965): Eperythrozoonosis in 4-week-old pigs. Iowa St. Uhiv. vet., 27, 119, 127 140. Prullage, J. B. (1989): Mechanical transmission of Eperythrozoon suis Splitter by arthropod vectors. Doctoral Dissertation, Purdue University 141. Pruzanski, W. and Shumak, K. H. (1977a): Biological activity of cold-reacting autoantibodies (first of two parts). New'Eng. J. Med., 297, 538-542 142. Pruzanski, W. and Shumak, K. H. (1977b): Biological activity of cold-reacting autoantibodies (second of two parts). New'Eng. J. Med., 297, 583-589 143. Quin, A. H. (1938): A herd condition of swine characterized by icterus and anemia. JAVMA, 93, 327-328 144. Riley, V. (1964): Synergism between a lactate dehydrogenase elevating virus and Eperythrozoon coccoides. Science, 146, 921-922 145. Riley, V., Loveless, J. D. and Fitzmaurice, M. A” (1964): Comparison of a lactate dehydrogenase elevating virus-like agent and Eperythrozoon coccoides. Proc. Soc. Exp. Biol. Med., 116, 486-490 146. Ristic, M. (1962): A capillary tube-agglutination test for anaplasmosis - A preliminary report. JAVMA, 141, 588-594 147. Ristic, M. (1976): Immunologic systems and protection in infections caused by intracellular blood protista. vet. Parasitol., 2, 31-47 148. Rivkin, S. E. and Simon, E. R. (1965): Comparative carbohydrate catabolism and methemoglobin reduction in pig and human erythrocytes. J. Cell. Comp. Physiol., 66, 49-56 212 149. Robb, A. D. (1943): Ictero-anemia in growing swine. vet. Med. Sm. Anim. Clinic., 38, 271-274 150. Rogers, R. J. (1971): Bovine anaplasmosis: an evaluation of the complement-fixation and capillary tube- agglutination tests and the incidence of antibodies in northern Queensland cattle herds. Aust. vet. J., 47, 364-369 151. Rogers, R. J. (1974): Serum opsonins and the passive transfer of protection in Babesia rodhaini infections of rats. Int. J. Parasitol., 4, 197-201 152. Rosenkrans, C. F., Jr., Zinn, G. M. and Jesse, G. W. (1984): Effects of Chlortetracycline as a treatment against eperythrozoonosis. J. Anim. Sci., 59 (Suppl.1), 388 153. Rouse, B. and Johnson, R. (1966): Eperythrozoon ovis. 154. Scheff, G. J., Scheff, I. M. and Eiseman, G. (1956): Concerning the mechanism of Bartonella anemia in the splenectomized rat. J. Inf. Dis., 98, 113-120 155. Scheibel, L. W., Ashton, 8. H. and Trager, W. (1979): Plasmodium falciparum: Microaerophilic requirements in human red blood cells. Expt. Parasitol., 47, 410-418 156. Schilling, V. (1928): Eperythrozoon coccoides, eine neue durch splenektomie aktivierbare dauerinfektion der weissen maus. Klin. wochenschr., 7, 1853-1855 157. Schmidt, P., Kaspers, B., Jfingling, A., Heinritzi, K. and Lbsch, U. (1992): Isolation of cold agglutinins in Eperythrozoon suis-infected pigs . Vet . Immunol . Immunopathol., 31, 195-201 158. Schroeder, W. F. and Ristic, M. (1965): Anaplasmosis XVII. The relation of autoimmune processes to anemia. Am. J. vet. Res., 26, 230-245 159. Schuller, W., Heinritzi, K., Al-Nutha, 8., Kolbl, S. and Schuh, M. (1990): Serological studies by means of CF- tests and ELISA for evaluation of antibodies against Eperythrozoon suis. Berl. Mfinch. Tierarztl. wechr., 103, 9- 12 160. Scott, J. K. and Stannard, J. N. (1954): Relationship between Bartonella muris infection and acute radiation effects in the rat. J. Inf. Dis., 95, 302-308 161. Seamer, J. (1960): Studies with Eperythrozoon parvum Splitter, 1950. Parasitology, 50, 67-80 213 162. Seamer, J., Gledhill, A. W., Barlow, J. L. and Hotohin, J. (1961): Effect of Eperythrozoon coccoides upon lymphocytic choriomeningitis in mice. J. Immunol., 86, 512- 515 163. Shankarappa, B. and Dutta, 8. K. (1989): Production and characterization of monoclonal antibodies to Ehrlichia 164. Sheriff, D. (1978): The pathology of E. ovis. .N. z. vet. J., 26, 315 165. Sisk, D. B., Cole, J. R. and Pursell, A. R. (1980): Serologic incidence of eperythrozoonosis in Georgia swine. PIOC. Anna. Meet. Me ASSOC. veto Labo Diago' 23' 91-100 166. Small, E. and Ristic, M. (1967): Morphologic features of Haemobartonella felis. Am. J. vet. Res., 28, 845-851 167. Smith, A. R. (1977): Eperythrozoonosis. veterinary professional topics, university of Illinois: Swine, 5, 2-4 168. Smith, A. R. (1981): Eperythrozoonosis. In Diseases of swine, Ed. B. 8. A. D. Leman R. D. Glock, W. L. Mengeling, R. H. C. Penny and E. Scholl, Iowa State University Press, Ames, IA, pp. 683-687 169. Smith, A.‘R. and Rahn, T. (1975): An indirect hemagglutination test for the diagnosis of Eperythrozoon suis infection in swine. Am. J. vet. Res., 36, 1319-1321 170. Smith, B. J. (1984): SDS polyacrylamide gel electrophoresis of proteins. In Proteins: Methods in molecular biology, vol.1, Ed. J. W. Walker, Humana Press, Clifton, NJ, pp. 41-55 171. Soni, J. L. and Cox, H. W. (1974): Pathogenesis of acute avian malaria: I: Immunologic reactions associated with anemia, splenomegaly, and nephritis of acute Plasmodium gallinaceum infections of chickens. Am. J. Trop. Med. Hyg., 23, 577-585 172. Soni, J. L. and Cox, H. W. (1975): Pathogenesis of acute avian malaria: II: Anemia mediated by a cold-active autohemagglutinin from the blood of chickens with acute Plasmodium gallinaceum infection. Am. J. Trop. Med. Hyg., 24, 206-213 173. Sonoda, M., Takahashi, K., Tamura, T. and Koiwa, M. (1977): Clinical and hematological observation on spontaneous and experimental cases of eperythrozoonosis in sheep. J. Jpn. vet. Med. Assoc., 30, 374-380 214 174. Spencer, R. (1940): Anaplasmosis-like disease of swine. vet. Med. Sm. Anim. Clinic., 35, 294-295 175. Splitter, E. J. (1950a): Eperythrozoon suis n. sp. and Eperythrozoon parvum n. sp., two new blood parasites of swine. Science, 111, 513-514 176. Splitter, E. J. (1950b): Eperythrozoon suis, the etiologic agent of ictero-anemia or an anaplasmosis-like disease in swine. Am. J. vet. Res., 11, 324-330 177. Splitter, E. J. (1950c): Neoarsphenamine in acute eperythrozoonosis of swine. JAVMA, 117, 371-373 178. Splitter, E. J. (1951): Eperythrozoonosis in swine. Iowa St. college vet., 13, 77-81 179. Splitter, E. J. (1952): Eperythrozoonosis in swine - filtration studies. Am. J. vet. Res., 13, 290-297 180. Splitter, E. J. (1953): Eperythrozoon parvum, a filtrable blood parasite of swine. Nature, 172, 40 181. Splitter, E. J. (1958): The complement-fixation test in diagnosis of eperythrozoonosis in swine. JAVMA, 132, 47- 49 182. Splitter, E. J. and Castro, E. R. (1957): Antibiotic therapy in acute eperythrozoonosis of swine. JAVMA, 131, 293-294 183. Splitter, E. J. and Williamson, R. L. (1950): ,Eperythrozoonosis in swine. A preliminary report. JAVMA, 116, 360-364 184. Stansly, P. G. a. N., C. F. (1965): Relationship between spleen weight increase factor SWIF of mice and Eperythrozoon coccoides. Proc. Soc. Exp. Biol. Med., 119, 1059-1063 185. Stauch, D. W. (1951): Eperythrozoonosis. Iowa vet., 22, 9-11 186. Steere, A. C. (1989): Lyme disease. New Eng. J. Med., 321, 586-596 187. Suntharasamai, P. and Rytel, M. W. (1973): Interferon response and interferon action in Eperythrozoon coccoides infection of mice. Proc. Soc. Exp. Biol. Med., 142, 811-816 215 188. Sutton, R. H. (1976): Eperythrozoon ovis: The difference in carbohydrate metabolism between infected and uninfected sheep erythrocytes. AJEBAK, 54, 449-458 189. Sutton, R. H. (1977): The effect of Eperythrozoon ovis infection on the glucose level and some acid-base factors in the venous blood of sheep. Aust. vet. J., 53, 478-481 190. Sutton, R. H. (1979a): Effct of Eperythrozoon ovis infection on the reductive potential of sheep erythrocytes. vet. Parasitol., 5, 11-15 191. Sutton, R. H. (1979b): The pathology of E. ovis. .N. 192. Tanaka, H., Hall, W. T., Sheffield, J. B. and Moore, D. H. (1965): Fine structure of Haemobartonella muris as compared with Eperythrozoon coccoides and Mycoplasma (pulmonis. J. Bacteriol., 90, 1735-1749 193. Thomas, T. B., Pomerat, C. M. and Frieden, E. H5 (1949): Cellular reactions in Haemobartonella infected rats with anemia produced by anti-blood immune serum. J. Inf. Dis., 84, 169-186 194. Thoongsuwan, S. and Cox, H. VL. (1973): Antigenic variants of the haemosporidian parasite, Babesia rodohaini, selected by in vitro treatment with immune globulin. Ann. Trop..Med. Parasitol., 67, 373-385 195. Thurston, J. P. (1955): Observation on the course of Eperythrozoon coccoides infections in mice, and the sensitivity of the parasite to external agents. Parasitology, 45, 141-151 196. Towbin, H., Staehelin, T. and Gordon, J. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci., USA, 76, 4350-4354 197. Trager, W. (1941): Studies on conditions affecting the survival in vitro of a malaria parasite (Plasmodium lophurae). J. Exp..Med., 74, 441-461 198. Trager, W. and Jensen, J. B. (1976): Human malaria parasites in continuous culture. Science, 193, 673-675 199. Tuomi, J. (1966): A microorganism affecting bovine platelets. Experimentia, 22, 458-459 216 200. Tuomi, J. and von Bonsdorff, C.-H. (1967): Ultrastructure of a microorganism associated with bovine platelets. Experimentia, 23, 111-112 201. Tyzzer, E. (1942): A comparative study of grahamellae, haemobartonellae, and eperythrozoa of small mammals. Proc. 202. Uilenberg, G. (1967): Eperythrozoon tuomii, n. sp. (Rickettsiales), the third species of Eperythrozoon of bovines in Madagascar. Rev..Elev., 20, 563-569 203. Upston, J. M. and Gero, A. M. (1990): Increased glucose permeability in Babesia Davis-infected erythrocytes. Int. J. Parasitol., 20, 69-76 204. Valli, v. E. O. (1985): The hematopoietic system. In Pathology of domestic animals, 3rd ed., vol.3, Ed. K. V. F. Jubb, P. C. Kennedy and N. Palmer, Academic Press., Orlando, FL, pp. 83-236 205. van Leengoed, L., van Dijk, J., Ham, M. and Verheijden, J. (1992): Nonhemolytic anemia (NBA) in sows. In Proc. 12th International Pig Veterinary Society Congress, Hague, Netherlands pp. 394 206. Voller, A. and Bidwell, D. E. (1968): The effect of Eperythrozoon coccoides infection in mice on superimposed Plasmodium chabaudi and semliki Forest virus infections. Ann. Trop. Med. Parasitol., 62, 342-348 207. Weinman, D. (1944): Infectious anemias due to Bartonella and related red cell parasites. Trans. Am. Philos. Soc., 33, 243-345 208. Weiss, L., Geduldig, U. and Weidanz, W. (1986): Mechanisms of splenic contol of murine malaria: Reticular cell activation and the development of a blood-spleen barrier. Am. J. Anat., 176, 251-285 209. wensing, T., Nouwens, G., Schotman, A. J. H., Vernooy, J. and Zwart, D. (1974): The effect of Eperythrozoon wenyoni on the glucose level and acid base balance of bovine blood in vivo and in vitro. Tijdschr. Diergeneesk., 99, 136-142 210. Wigand, R. (1956): A complement-fixation reaction with Haemobartonella muris and Eperythrozoon coccoides. Nature, 178, 1288-1289 211. Williams, R. E. Personal communication, 1990 217 212. Williams, R. E. and McCain, T. L. (1982): Insects and EPE. Pig America, Watt Publishing, Mt. Morris, IL, PP. 37, (June) 213. Wright, I. G. (1973): Osmotic fragility of erythrocytes in acute Babesia argentina infections in splenectomized Bos taurus calves. Res. Vet. Sci., 15, 299- 305 214. Yuthavong, Y., Wilairat, P., Panijan, B., Potiwan, C. and Beale, G. H. (1979): Alterations in membrane proteins of mouse erythrocytes infected with different species and strains of malaria parasites. Comp. Biochem. Physiol., 63, 83-85 215. Zachary, J. F. and Basgall, E. J. (1985): Erythrocyte membrane alterations associated with the attachment and replication of Eperythrozoon suis: A light and electron microscopic study. vet. Pathol., 22, 164-170 216. Zachary, J. F. and Smith, A. R. (1985): Experimental porcine eperythrozoonosis: T-lymphocyte suppression and misdirected immune responses. Am. J. Vet. Res., 46, 821-830 217. Zinn, G. M., Jesse, G. W. and Dobson, A. W. (1983): Effect of eperythrozoonosis on sow productivity. JAVMA, 182, 369-371 218. Zulty, J. C. and Kociba, G. J. (1990): Cold agglutinins in cats with haemobartonellosis. JAVMA, 196, 907-910 MICHIGAN STATE UNIV. LIBRnRIES W ”WI N ill‘ ”I W M WI W W W I" "I“ WI 31293009140025