Veterinary Microbiology 173 (2014) 160–165
Contents lists available at ScienceDirect
Veterinary Microbiology journal homepage: www.elsevier.com/locate/vetmic
Short Communication
Serum protein profiles, circulating immune complexes and proteinuria in dogs naturally infected with Anaplasma phagocytophilum Urska Ravnik a,*, Blanka Premrov Bajuk b, Lara Lusa c, Natasa Tozon a a
Small Animal Clinic, Veterinary Faculty, University of Ljubljana, Cesta v Mestni log 47, SI-1000 Ljubljana, Slovenia Institute of Physiology, Pharmacology and Toxicology, Veterinary Faculty, University of Ljubljana, Gerbiceva 60, SI-1000 Ljubljana, Slovenia c Institute for Biostatistics and Medical Informatics, Faculty of Medicine, University of Ljubljana, Vrazov trg 2, SI-1104 Ljubljana, Slovenia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 March 2014 Received in revised form 7 July 2014 Accepted 10 July 2014
Alterations in serum protein profile, presence of circulating immune complexes (CIC) and proteinuria were investigated in a large group of dogs naturally infected with the Anaplasma phagocytophilum bacterium. Our aim was to evaluate the presence of hypergammaglobulinaemia, CIC and proteinuria as a possible result of an immunemediated disease following infection by or exposure to A. phagocytophilum. Dogs were divided into three groups – IFA positive (188 dogs with confirmed exposure to A. phagocytophilum), PCR positive (31 dogs with confirmed infection), and control (IFA and PCR negative) (19 dogs). Serum and urine protein patterns were determined by electrophoresis and CIC concentrations by absorbance nephelometry. No significant differences in hypergammaglobulinaemia were observed between the different groups, as shown by the presence of acute phase proteins a2 and b1–2 globulins. CIC concentrations in the IFA and PCR positive groups were, on average, higher than in controls by 151.3 mg/ ml, though the differences were not significant. The proportion of dogs with proteinuria did not differ significantly between groups. Our results confirm the assumption that anaplasmosis in dogs is most probably a disease with an acute course, with a good prognosis under the right treatment. ß 2014 Elsevier B.V. All rights reserved.
Keywords: Anaplasma phagocytophilum Circulating immune complex Proteinuria Dogs
1. Introduction Infection of dogs with different tick-borne pathogens of the family Anaplasmataceae, including Anaplasma phagocytophilum, Ehrlichia canis and Anaplasma platys, results in similar clinical and laboratory indications, even though these bacteria attack different target cells – monocytes, granulocytes and platelets. Thus, pathogenesis is suspected to be immune mediated (Harrus et al., 2001; Day, 2008). A few studies have examined cytokine responses to
* Corresponding author. Tel.: +386 1 4779 277; fax: +386 1 4779 349. E-mail address: [email protected] (U. Ravnik). http://dx.doi.org/10.1016/j.vetmic.2014.07.007 0378-1135/ß 2014 Elsevier B.V. All rights reserved.
infection with A. phagocytophilum, in particular in humans, horses, sheep, and dogs (Dumler et al., 2000; Kim et al., 2002; Scorpio et al., 2011). The results show a speciesspecific cytokine response that contributes greatly to controlling infection and preventing subsequent manifestation of the disease. In addition to the host immune response, the pathogenesis in canine tick-borne infections involves secondary immune-mediated sequelae as a result of immune dysregulation (Day, 2011). In monocytic ehrlichiosis, over-activity of B lymphocytes gives rise to hypergammaglobulinaemia and the formation of immune complexes of antigen, antibody and complement, causing immune mediated glomerulonephritis (IMGN)
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
and uveitis (Harrus et al., 2001). Studies on immunemediated disease on infection with A. phagocytophilum in dogs are scarce. The aim of this study was to evaluate the presence of hypergammaglobulinaemia, circulating immune complexes (CIC) and proteinuria as a result of possible IMGN in dogs naturally infected with A. phagocytophilum bacterium. 2. Materials and methods 2.1. Animals Serum samples were collected prospectively from 204 dogs presented to the Small Animal Clinic of Veterinary Faculty in Ljubljana and from 34 military dogs. The only criterion for inclusion was a positive IFA and/or PCR result for A. phagocytophilum, regardless of typical changes in clinical and laboratory findings characteristic of active anaplasmosis (Bjo¨ersdorf, 2005). Dogs were divided into 3 groups–). Dogs were divided into three groups – IFA positive (188 dogs, all PCR negative), PCR positive (31 dogs, regardless of IFA titre), and control (IFA and PCR negative) (19 dogs). 2.2. Determination of A. phagocytophilum antibody titres by IFA Serum samples were analysed for specific antibodies to A. phagocytophilum by IFA (Ristic et al., 1972). A. phagocytophilum, isolated from a Slovenian patient, was grown in a cell culture of HL60 cells.
161
quantitated by scanning the bands at 570 nm with a densitometer. 2.6. Absorbance nephelometry of circulating immune complexes 2.6.1. Simulated immune complexes Heat-aggregated gamma globulins (HAGG) were prepared by heating and centrifugation of dog Cohn Fraction II (NBS Biologicals, UK) as described (Levinson and Goldman, 1983a, 1983b, 1986; Levinson et al., 1984; Feldkamp et al., 1985). The concentration of HAGG was estimated by a biuret method. Calibrators were prepared the day before the assay by serial dilution of the stock HAGG solution with Tris buffer (10 mM Tris, 0.15 M NaCl, 20 g/L bovine serum albumin. 2.6.2. Samples CIC were assayed using polyethylene glycol precipitation (Day, 1987). 2.6.3. Assay of extracts To 770 mL of test (PEG 6000 in phosphate saline buffer with anti-dog IgG (Rockland, USA)) or blank solution (the same but without anti-IgG) 110 mL of extract from patient’s sera was added. The reaction mixtures were incubated at room temperature for 30 min, then transferred to cuvettes and the absorbance measured at 340 nm (Lambda 12, PerkinElmer, USA). The difference in the absorbance between the test and blank reaction mixture is due to turbidity caused by the reaction of anti-IgG with IgG in immune complexes.
2.3. PCR detection of A. phagocytophilum DNA DNA was extracted from leucocytes with QiaAmp DNA mini kit according to the manufacturer’s instructions (Qiagen GmbH, Germany). The primer pair Ehr521 and Ehr790 was used for screening all samples (Kolbert, 1996). All positive samples were additionally tested for the groESL operon. All the amplicons were verified by sequencing. 2.4. Haematology A complete blood count was performed using an automated laser haematology analyser H*1 (Siemens/ Bayer, Germany) with species specific software (H*1 MultiSpecies V30 Software, USA). 2.5. Serum protein electrophoresis Serum for protein analysis was separated by centrifugation 1 h after blood collection and stored at 20 8C until analysed. Serum protein electrophoresis (SPE) was performed using the HYDRASYS system, a semi-automated multi-parameter instrument. The automated steps include processing of HYDRAGEL agarose gels in the following sequence: sample application, electrophoretic migration, drying, staining, destaining and final drying. We used 10 mL of thawed serum. The electrophoresis was performed at 20 8C using 10 W for approximately 7 min. Products were stained with amidoblack stain. Serum proteins were
2.6.4. Calculation of CIC concentration in sera and calibrators For each run calibrators ranging from 12.5 to 400 mg/L were prepared by serial dilution. The calibrators were added directly to test or blank solutions. One set of calibrators was assayed along with duplicate sets of the extracted sera. The difference in the absorbance values of calibrators in test or blank solution was plotted versus calibrator concentration to provide a calibration curve. CIC concentration was estimated from the calibration curve from the average absorbance difference of each serum sample. 2.7. Urinalysis Urine samples were analysed by urine analyser CLINITEK 50 (Siemens/Bayer, Germany), using a standard multitest urine dipstick, and microscopic examination of urine sediment within 2 h of sample collection. In samples in which proteinuria was detected, proteins were measured by automated chemistry analyser Daytona RX (Randox, UK) and the urine protein:creatinine ratio (UPC) was calculated. 2.8. Urine protein electrophoresis The pattern and molecular masses of proteins in urine samples were monitored by SDS-PAGE (Laemmli, 1970). A 10.4% resolving gel was used on a mini-Protean II Slab cell apparatus (Bio-Rad, USA). Prior to loading the samples,
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
162
urine samples were mixed (1:3) with 4 SDS Laemmli buffer. The samples were visualized with Coomassie brilliant blue R-250 (Bio-Rad, USA). Molecular masses were determined using molecular mass standards of 10–170 kDa (Fermentas, Thermo scientific, USA). 2.9. Statistics For purposes of statistical evaluation dogs were divided into four groups. DefNeg contained dogs without active anaplasmosis according to currently accepted criteria (Bjo¨ersdorff, 2005) (IFA titre 1:1024). DefPos contained dogs with active anaplasmosis (IFA titre 1:2048 + thrombocytopenia and/or clinical signs). The other two groups comprised PCR positive and PCR and IFA negative (control group) dogs. Groups were compared to each other using the Kruskal–Wallis and chi-squared tests. Statistical significance was defined as P < 0.05. Restricted cubic splines were used to flexibly model the relationship between CIC or PLT and serum or urine proteins; the estimated association was displayed graphically. 3. Results 3.1. Serum protein electrophoresis results One hundred and fifty-nine samples – 127 IFA positive, 18 PCR positive and 14 control serum – were subjected to SPE and their serum protein fraction concentrations determined (Table 1). Albumin/globulin (A/G) ratios less than 0.8, considered as pathological, were observed in 32 samples, 27 of which were IFA positive, 4 PCR positive and 1 control (Table 2). A/G values in a further 127 samples were within the reference range and, although there were minor changes in electrophoretogram in 48 of these samples, they were not considered pathological. Changes were limited to increased concentrations of both albumin and globulins. A/G ratios of all samples did not differ significantly from those in controls (Table 2). Further, changes in different protein fractions were compared only in samples with pathological A/G ratios. Thrombocytopenia is a hallmark of acute anaplasmosis. Comparison of relevant serum protein fractions showed statistically significant associations between thrombocytopenia and hypoalbuminaemia, hypo-a1-globulinaemia, and hypergammaglobulinaemia (Fig. 1). Table 1 Serum protein electrophoresis results (g/dl) (mean standard deviation) in IFA positive, PCR positive and control group.
n TP A G A/G a1 a2 b1–2 b3
g
IFA positive
PCR positive
Control
127 7.19 1.25 3.56 0.81 3.66 0.84 1.14 0.86 0.30 0.26 1.43 0.41 0.64 0.22 0.77 0.25 0.56 0.24
18 6.43 1.03 3.13 0.64 3.31 0.72 0.98 0.28 0.25 0.07 1.26 0.36 0.61 0.20 0.65 0.16 0.55 0.18
14 6.26 0.46 3.38 0.53 2.91 0.42 1.20 0.28 0.35 0.11 0.98 0.13 0.38 0.09 0.62 0.09 0.55 0.16
n, number of samples; TP, total proteins; A, albumins; G, globulins; A/G, albumin to globulin ratio; a1–g, different protein fractions.
Table 2 Number of samples with A/G < 0.8 and their changes on electrophoretogram in different groups.
n A"/A# G"/G# a1"/a1# a2"/a2# b1–2"/b1–2# b3"/b3# g"/g#
DefNeg
DefPos
PCR
Control
24 0/23* 23*/0 11/6* 17*/2 9*/0 16*/2 6/0
3 0/3* 3*/0 0/1* 3*/0 2*/0 2/0 2/1
4 0/4* 3*/0 2/1* 2*/0 1*/0 2/0 2/0
1 0/1 0/0 1/0 0/0 0/0 1/0 1/0
n, number of dogs; DefNeg, dogs with IFA titres 1:1024; DefPos, dogs with IFA titres 1:2048; PCR, PCR positive dogs; ", increased; #, decreased; A, albumin; G, globulins; a1–g, serum protein fractions; *, significantly different than control group.
3.2. Circulating immune complexes concentration measurement results Presence of CIC was assessed in 223 serum samples – 180 were IFA positive, 24 PCR positive and 19 control. CIC were confirmed in 79.4% of IFA positive samples, 83.3% of PCR positive and 89.5% of control samples. Mean CIC concentrations and standard deviations were determined (Table 3). CIC concentrations of the IFA and PCR positive groups did not differ significantly from those of the control group, although their average CIC concentrations were higher by 1513 mg/ml. In contrast, statistically significant associations existed between CIC concentration and thrombocytopenia, hypergammaglobulinaemia and low A/G ratio. Dogs with higher CIC concentrations had lower platelets, higher gammaglobulins, and lower A/G ratios (Fig. 2). 3.3. Urinalysis results Urinalysis was conducted on 58 samples – 36 were IFA positive, 3 PCR positive and 19 control samples. Urine multitest dipstick revealed few changes in specific gravity (SG). In 15 samples SG was above, and in 5 samples below, the reference range of values. Because of only one time measurement of SG in a random urine sample and, mostly, because of no correlation with pathologic urine sediment or proteinuria (with the exception of two samples with low SG), these SG changes could not be considered to be pathological. In a microscopic evaluation of urine sediment we focused on casts and renal cells. Hyaline and fine granular casts were found in 12 samples, all in the IFA positive group. In seven of those samples, UPC was high and UPE revealed the presence of MMW and HMW proteins, with the exception of one sample, which contained only LMW proteins. 3.4. Urine protein electrophoresis results Urine protein electrophoresis (UPE) was conducted in 62 samples – 42 IFA positive, 3 PCR positive and 17 control. Proteins differing in molecular weight groups were found in 58.1% of the samples. Urine samples were divided into three groups according to the presence of low (LMW)
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
163
Fig. 1. (a–c) Significant correlation between thrombocytopenia and hypoalbuminaemia, hypo-a1-globulinaemia and hypergammaglobulinaemia. PLT, platelets; ALB, albumin; Alpha1, a1-globulins; gamma, g-globulins; P, probability value.
4. Discussion Table 3 Circulating immune complexes concentration (mg/ml) (mean standard deviation) in IFA positive, PCR positive and control group.
n CIC SD
IFA
PCR
Control
143 271.5 298.4
20 290.9 230.2
17 127.3 84.5
n, number of CIC positive samples; CIC, circulating immune complexes; SD, standard deviation.
(76 kDa) proteins. UPE results for 10 dogs in the control group showed the presence of LMW proteins – bands were observed at 10, 15, 25, and 55 kDa. All these dogs were clinically healthy, had inactive urine sediment and UPC within the reference range. All surveyed samples with LMW proteins had bands with the same molecular weight as the control group, inactive urine sediment and UPC in the reference range, therefore we considered those proteins to be physiological. Proteins with higher molecular weight were detected in 11 urine samples (30.6%), all of which were IFA positive and PCR negative. UPC in all those samples was above the reference range. These results confirmed glomerular proteinuria. CIC concentration was also found to correlate significantly with the presence of MMW and HMW proteins in urine – an increase in CIC concentration increases the occurrence of proteinuria (Fig. 3).
Fig. 2. (a and b) Significant correlation between concentration of CIC and gammaglobulins and albumin: globulin ratio. CIC, circulating immune complexes; gamma, g-globulins; AG, albumin:globulin ratio; P, probability value.
This study presents the serum protein profiles, presence of CIC and proteinuria in a large group of dogs naturally infected with A. phagocytophilum. SPE is indicated in infectious diseases, which are known to cause dysproteinaemia. Hypoalbuminaemia, hyperglobulinaemia (mainly hypergammaglobulinaemia) and hypo-a1-globulinaemia are characteristic for monocytic ehrlichiosis and leishmaniosis, which can both proceed to a chronic phase. In our study, dysproteinaemia was not significant, not even in the acute phase of anaplasmosis. Lower A/G ratios were found in 32 samples, the majority (84.4%) in groups with IFA titre 1:1024 that did not meet the criteria for active anaplasmosis (Bjo¨ersdorff, 2005). Alterations in various protein fractions correlated with those in monocytic ehrlichiosis and leishmaniosis, with the exception of gammaglobulins, which did not differ significantly from those in the control group. The major contributors to hyperglobulinaemia were the acute phase proteins a2 and b1–2 globulins. Twenty (62%) dogs in the group with lower A/G ratio had thrombocytopenia and 23
Fig. 3. Significant correlation between CIC concentration and the presence of MMW proteins in urine. CIC, circulating immune complexes; MMW, middle molecular weight proteins; P, probability value.
164
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
(71.8%) dogs had elevated body temperature, depression, and reluctance to move, which could be associated with an acute infectious process. This could mean that dogs with low IFA titres, with subclinical or persistent infection, can develop an acute form of anaplasmosis. However, dysproteinaemia could possibly be the result of some other disease. Based on our results, SPE cannot be recommended as a diagnostic tool that would help in understanding the course of anaplasmosis. However, the significance of thrombocytopenia can be emphasized, based on its statistically significant association to acute phase proteins. In the case of clinical signs accompanied by thrombocytopenia, treatment should be considered, regardless of the level of the IFA titre. This complements our previous findings (Ravnik et al., 2011). The presence of CIC has too been described in many infectious diseases and, if the disease proceeds to the chronic phase, can cause glomerulonephritis, polyarthritis or uveitis (Pedersen, 1999). In monocytic ehrlichiosis, CIC were found in naturally and experimentally infected dogs in acute and subclinical phases of the disease (Harrus et al., 2001). The latter’s results show that the severity of the disease correlated with the presence of CIC. In our study, CIC concentration was high both in the control and study groups, which is due to the method used. Nephelometry recognizes CIC as a molecular aggregate, therefore does not distinguish between aggregates of immunoglobulin molecules alone and immune complexes that contain antigens. And since self-reactive antibodies are always present in healthy adults, those can be detected by nephelometry. But importantly, there was no significant difference in CIC concentration in any group comparing to the control group. CIC concentration was however correlated to thrombocytopenia, hypergammaglobulinaemia and lower A/G ratio. CIC concentration was significantly higher in dogs with thrombocytopenia, regardless of the IFA titre or a positive PCR result, which suggests an ongoing immunological response and again emphasises the significance of thrombocytopenia in treatment decision making in an A. phagocytophilum seropositive dog. A statistically significant association is suggested between high CIC concentration and the presence of proteinuria, but the number of samples was low (only four samples). Larger numbers of samples, together with histological and immunohistochemical analysis of the kidney tissue, need to be assessed. If the above mentioned correlation were then confirmed, CIC concentration could be used as a risk factor for the development of glomerulonephritis in A. phagocytophilum infected dogs. Proteinuria was detected in 36 (58.1%) samples. LMW proteins, present in 15 (41.7%) samples, were considered physiological since the bands were at the same positions as in the control group. MMW and HMW proteins were present in 11 (30.5%) samples, which suggest the presence of glomerulonephritis. Proteinuria can be present transiently in a febrile phase of any disease although none of these dogs were febrile at the time of urine sampling. The occurrence of proteinuria in the different groups and the control group did not differ significantly, but the fact that MMW and HMW proteins were detected exclusively in A. phagocytophilum seropositive dogs warrants regular
follow-up of A. phagocytophilum infected dogs for any later occurrence of proteinuria. Microscopic evaluation of all urine samples revealed the presence of different, mainly fine granular, casts in all proteinuric samples. We therefore recommend urine electrophoresis for all samples with microscopically confirmed casts and/or high urinary protein to creatinine ratios. Proteinuria was detected in combination with high CIC levels (100%) and thrombocytopenia (75%) in four samples with various IFA titres. Our results indicate that high CIC levels and thrombocytopenia occur in the acute phase of anaplasmosis and proteinuria probably only after a chronic antigenic stimulation (Pedersen, 1999). This suggests that persistent infection by A. phagocytophilum can result in the development of IMGN and reoccurrence of clinical signs at a particular moment. 5. Conclusions The results of this study do not support the reported development of a chronic phase of anaplasmosis in dogs (Egenvall et al., 2000; Alleman et al., 2006). Given that anaplasmosis in dogs is typically manifested seasonally, with acute clinical signs, and that the percentage of asymptomatic seropositive dogs in endemic areas is high, this is probably a disease with an acute onset and a good prognosis following the right treatment. Conflict of interest None of the authors has a financial or personal relationship with other persons or organizations that could bias the content of the article. Acknowledgments The authors thank Katarina Babnik for excellent laboratory work with CIC measurements and urine protein electrophoresis and Dr. Alberto Franceschi for serum protein electrophoresis. We also thank Prof. Dr. Tatjana Avsic Zupanc and Dr. Katja Strasek Smrdel for performing PCR and IFA for A. phagocytophilum and Prof. Roger Pain for review of English. The authors acknowledge the financial support of the Slovenian Research Agency (Project No. P4-0053). References Alleman, A.R., Chandrashekar, R., Beall, M., Cyr, K., Barbet, A., Lundgren, A., Sorenson, H., Wamsley, H., Wong, S., 2006. Experimental inoculation of dogs with a human isolate (Ny18) of Anaplasma phagocytophilum and demonstration of persistent infection following doxycycline therapy (abst). J. Vet. Intern. Med. 20, 763. Bjo¨ersdorff, A., 2005. Ehrlichiosis and anaplasmosis. Part 2: Granulocytic ehrlichiosis: anaplasma phagocytophilum comb. nov. (E. phagocytophila genogroup) infection. In: Shaw, S.E., Day, M.J. (Eds.), ArthropodBorne Infectious Diseases of the Dog and Cat. Manson Publishing, London, pp. 127–132. Day, M.J., 1987. A study of the immune response in canine disseminated aspergillosis. (Ph.D. thesis)Murdoch University. Day, M.J., 2008. The Clinical Immunology of the Dog and Cat. Manson Publishing, London, pp. 94–115. Day, M.J., 2011. The immunopathology of canine vector-borne diseases. Parasites Vectors 4, 48.
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165 Dumler, J.S., Trigiani, E.R., Bakken, J.S., Aguero-Rosenfeld, M.E., Wormser, G.P., 2000. Serum cytokine responses during acute human granulocytic ehrlichiosis. Clin. Diagn. Lab. Immunol. 7, 6–8. Egenvall, A., Lilliho¨o¨k, I., Bjo¨ersdorff, A., Olsson Engvall, E., Karlstam, E., Artursson, K., Heldtander, M., Gunnarsson, A., 2000. Detection of granulocytic Ehrlichia species DNA by PCR in persistently infected dogs. Vet. Rec. 146, 186–190. Feldkamp, C.S., Levinson, S.S., Perry, M., Amin, V., 1985. Anti-IgG combined with rate nephelometry for measuring polyethylene glycol-precipitated circulating immune complexes. Clin. Chem. 31, 2024–2027. Harrus, S., Day, M.J., Waner, T., Bark, H., 2001. Presence of immunecomplexes, and absence of antinuclear antibodies, in sera of dogs naturally and experimentally infected with Ehrlichia canis. Vet. Microbiol. 83, 343–349. Kim, H.Y., Mott, J., Zhi, N., Tajima, T., Rikihisa, Y., 2002. Cytokine gene expression by peripheral blood leukocytes in horses experimentally infected with Anaplasma phagocytophila. Clin. Diagn. Lab. Immunol. 9, 1079–1084. Kolbert, C.P., 1996. Detection of the agent of human granulocytic ehrlichiosis by PCR. In: Persing, D.H. (Ed.), PCR Protocols for Emerging Infectious Diseases. ASM Press, Washington, pp. 106–111. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. Levinson, S.S., Goldman, J., 1983a. Absorbance nephelometry of immune complexes by reaction with anti-IgG after treatment with polyethylene glycol. Clin. Chem. 29, 2035–2039.
165
Levinson, S.S., Goldman, J., 1983b. Measurement of circulating Clq precipitable immune complexes by a nephelometric type assay after polyethylene glycol extraction in model system. Am. Soc. Clin. Pathol. 79, 451–457. Levinson, S.S., Goldman, J., 1986. Fluorescent anti-immunoglobulin G assay of circulating immune complexes extracted with polyethylene glycol. J. Clin. Microbiol. 23, 29–32. Levinson, S.S., Goldman, J.O., Feldkamp, C.S., 1984. Anti-IgG binding test to assay circulating IgG-containing immune complexes from polyethylene glycol precipitates. Clin. Chem. 30, 1502–1506. Pedersen, N.C., 1999. A review of immunologic diseases of the dog. Vet. Immunol. Immunopathol. 69, 251–342. Ravnik, U., Tozon, N., Strasek Smrdel, K., Avsic Zupanc, T., 2011. Anaplasmosis in dogs: the relation of haematological, biochemical and clinical alterations to antibody titre and PCR confirmed infection. Vet. Microbiol. 149, 172–176. Ristic, M., Huxsoll, D.L., Weisiger, R.M., Hildebrant, P.K., Nyindo, M.B.A., 1972. Serological diagnosis of tropical canine pancytopenia by indirect immunofluorescence. Infect. Immun. 6, 226–231. Scorpio, D.G., Dumler, J.S., Barat, N.C., Cook, J.A., Barat, C.E., Stillman, B.A., DeBisceglie, K.C., Beall, M.J., Chandrashekar, R., 2011. Comparative strain analysis of Anaplasma phagocytophilum infection and clinical outcomes in a canine model of granulocytic anaplasmosis. Vector Borne Zoonotic Dis. 11, 223–229.
Contents lists available at ScienceDirect
Veterinary Microbiology journal homepage: www.elsevier.com/locate/vetmic
Short Communication
Serum protein profiles, circulating immune complexes and proteinuria in dogs naturally infected with Anaplasma phagocytophilum Urska Ravnik a,*, Blanka Premrov Bajuk b, Lara Lusa c, Natasa Tozon a a
Small Animal Clinic, Veterinary Faculty, University of Ljubljana, Cesta v Mestni log 47, SI-1000 Ljubljana, Slovenia Institute of Physiology, Pharmacology and Toxicology, Veterinary Faculty, University of Ljubljana, Gerbiceva 60, SI-1000 Ljubljana, Slovenia c Institute for Biostatistics and Medical Informatics, Faculty of Medicine, University of Ljubljana, Vrazov trg 2, SI-1104 Ljubljana, Slovenia b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 March 2014 Received in revised form 7 July 2014 Accepted 10 July 2014
Alterations in serum protein profile, presence of circulating immune complexes (CIC) and proteinuria were investigated in a large group of dogs naturally infected with the Anaplasma phagocytophilum bacterium. Our aim was to evaluate the presence of hypergammaglobulinaemia, CIC and proteinuria as a possible result of an immunemediated disease following infection by or exposure to A. phagocytophilum. Dogs were divided into three groups – IFA positive (188 dogs with confirmed exposure to A. phagocytophilum), PCR positive (31 dogs with confirmed infection), and control (IFA and PCR negative) (19 dogs). Serum and urine protein patterns were determined by electrophoresis and CIC concentrations by absorbance nephelometry. No significant differences in hypergammaglobulinaemia were observed between the different groups, as shown by the presence of acute phase proteins a2 and b1–2 globulins. CIC concentrations in the IFA and PCR positive groups were, on average, higher than in controls by 151.3 mg/ ml, though the differences were not significant. The proportion of dogs with proteinuria did not differ significantly between groups. Our results confirm the assumption that anaplasmosis in dogs is most probably a disease with an acute course, with a good prognosis under the right treatment. ß 2014 Elsevier B.V. All rights reserved.
Keywords: Anaplasma phagocytophilum Circulating immune complex Proteinuria Dogs
1. Introduction Infection of dogs with different tick-borne pathogens of the family Anaplasmataceae, including Anaplasma phagocytophilum, Ehrlichia canis and Anaplasma platys, results in similar clinical and laboratory indications, even though these bacteria attack different target cells – monocytes, granulocytes and platelets. Thus, pathogenesis is suspected to be immune mediated (Harrus et al., 2001; Day, 2008). A few studies have examined cytokine responses to
* Corresponding author. Tel.: +386 1 4779 277; fax: +386 1 4779 349. E-mail address: [email protected] (U. Ravnik). http://dx.doi.org/10.1016/j.vetmic.2014.07.007 0378-1135/ß 2014 Elsevier B.V. All rights reserved.
infection with A. phagocytophilum, in particular in humans, horses, sheep, and dogs (Dumler et al., 2000; Kim et al., 2002; Scorpio et al., 2011). The results show a speciesspecific cytokine response that contributes greatly to controlling infection and preventing subsequent manifestation of the disease. In addition to the host immune response, the pathogenesis in canine tick-borne infections involves secondary immune-mediated sequelae as a result of immune dysregulation (Day, 2011). In monocytic ehrlichiosis, over-activity of B lymphocytes gives rise to hypergammaglobulinaemia and the formation of immune complexes of antigen, antibody and complement, causing immune mediated glomerulonephritis (IMGN)
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
and uveitis (Harrus et al., 2001). Studies on immunemediated disease on infection with A. phagocytophilum in dogs are scarce. The aim of this study was to evaluate the presence of hypergammaglobulinaemia, circulating immune complexes (CIC) and proteinuria as a result of possible IMGN in dogs naturally infected with A. phagocytophilum bacterium. 2. Materials and methods 2.1. Animals Serum samples were collected prospectively from 204 dogs presented to the Small Animal Clinic of Veterinary Faculty in Ljubljana and from 34 military dogs. The only criterion for inclusion was a positive IFA and/or PCR result for A. phagocytophilum, regardless of typical changes in clinical and laboratory findings characteristic of active anaplasmosis (Bjo¨ersdorf, 2005). Dogs were divided into 3 groups–). Dogs were divided into three groups – IFA positive (188 dogs, all PCR negative), PCR positive (31 dogs, regardless of IFA titre), and control (IFA and PCR negative) (19 dogs). 2.2. Determination of A. phagocytophilum antibody titres by IFA Serum samples were analysed for specific antibodies to A. phagocytophilum by IFA (Ristic et al., 1972). A. phagocytophilum, isolated from a Slovenian patient, was grown in a cell culture of HL60 cells.
161
quantitated by scanning the bands at 570 nm with a densitometer. 2.6. Absorbance nephelometry of circulating immune complexes 2.6.1. Simulated immune complexes Heat-aggregated gamma globulins (HAGG) were prepared by heating and centrifugation of dog Cohn Fraction II (NBS Biologicals, UK) as described (Levinson and Goldman, 1983a, 1983b, 1986; Levinson et al., 1984; Feldkamp et al., 1985). The concentration of HAGG was estimated by a biuret method. Calibrators were prepared the day before the assay by serial dilution of the stock HAGG solution with Tris buffer (10 mM Tris, 0.15 M NaCl, 20 g/L bovine serum albumin. 2.6.2. Samples CIC were assayed using polyethylene glycol precipitation (Day, 1987). 2.6.3. Assay of extracts To 770 mL of test (PEG 6000 in phosphate saline buffer with anti-dog IgG (Rockland, USA)) or blank solution (the same but without anti-IgG) 110 mL of extract from patient’s sera was added. The reaction mixtures were incubated at room temperature for 30 min, then transferred to cuvettes and the absorbance measured at 340 nm (Lambda 12, PerkinElmer, USA). The difference in the absorbance between the test and blank reaction mixture is due to turbidity caused by the reaction of anti-IgG with IgG in immune complexes.
2.3. PCR detection of A. phagocytophilum DNA DNA was extracted from leucocytes with QiaAmp DNA mini kit according to the manufacturer’s instructions (Qiagen GmbH, Germany). The primer pair Ehr521 and Ehr790 was used for screening all samples (Kolbert, 1996). All positive samples were additionally tested for the groESL operon. All the amplicons were verified by sequencing. 2.4. Haematology A complete blood count was performed using an automated laser haematology analyser H*1 (Siemens/ Bayer, Germany) with species specific software (H*1 MultiSpecies V30 Software, USA). 2.5. Serum protein electrophoresis Serum for protein analysis was separated by centrifugation 1 h after blood collection and stored at 20 8C until analysed. Serum protein electrophoresis (SPE) was performed using the HYDRASYS system, a semi-automated multi-parameter instrument. The automated steps include processing of HYDRAGEL agarose gels in the following sequence: sample application, electrophoretic migration, drying, staining, destaining and final drying. We used 10 mL of thawed serum. The electrophoresis was performed at 20 8C using 10 W for approximately 7 min. Products were stained with amidoblack stain. Serum proteins were
2.6.4. Calculation of CIC concentration in sera and calibrators For each run calibrators ranging from 12.5 to 400 mg/L were prepared by serial dilution. The calibrators were added directly to test or blank solutions. One set of calibrators was assayed along with duplicate sets of the extracted sera. The difference in the absorbance values of calibrators in test or blank solution was plotted versus calibrator concentration to provide a calibration curve. CIC concentration was estimated from the calibration curve from the average absorbance difference of each serum sample. 2.7. Urinalysis Urine samples were analysed by urine analyser CLINITEK 50 (Siemens/Bayer, Germany), using a standard multitest urine dipstick, and microscopic examination of urine sediment within 2 h of sample collection. In samples in which proteinuria was detected, proteins were measured by automated chemistry analyser Daytona RX (Randox, UK) and the urine protein:creatinine ratio (UPC) was calculated. 2.8. Urine protein electrophoresis The pattern and molecular masses of proteins in urine samples were monitored by SDS-PAGE (Laemmli, 1970). A 10.4% resolving gel was used on a mini-Protean II Slab cell apparatus (Bio-Rad, USA). Prior to loading the samples,
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
162
urine samples were mixed (1:3) with 4 SDS Laemmli buffer. The samples were visualized with Coomassie brilliant blue R-250 (Bio-Rad, USA). Molecular masses were determined using molecular mass standards of 10–170 kDa (Fermentas, Thermo scientific, USA). 2.9. Statistics For purposes of statistical evaluation dogs were divided into four groups. DefNeg contained dogs without active anaplasmosis according to currently accepted criteria (Bjo¨ersdorff, 2005) (IFA titre 1:1024). DefPos contained dogs with active anaplasmosis (IFA titre 1:2048 + thrombocytopenia and/or clinical signs). The other two groups comprised PCR positive and PCR and IFA negative (control group) dogs. Groups were compared to each other using the Kruskal–Wallis and chi-squared tests. Statistical significance was defined as P < 0.05. Restricted cubic splines were used to flexibly model the relationship between CIC or PLT and serum or urine proteins; the estimated association was displayed graphically. 3. Results 3.1. Serum protein electrophoresis results One hundred and fifty-nine samples – 127 IFA positive, 18 PCR positive and 14 control serum – were subjected to SPE and their serum protein fraction concentrations determined (Table 1). Albumin/globulin (A/G) ratios less than 0.8, considered as pathological, were observed in 32 samples, 27 of which were IFA positive, 4 PCR positive and 1 control (Table 2). A/G values in a further 127 samples were within the reference range and, although there were minor changes in electrophoretogram in 48 of these samples, they were not considered pathological. Changes were limited to increased concentrations of both albumin and globulins. A/G ratios of all samples did not differ significantly from those in controls (Table 2). Further, changes in different protein fractions were compared only in samples with pathological A/G ratios. Thrombocytopenia is a hallmark of acute anaplasmosis. Comparison of relevant serum protein fractions showed statistically significant associations between thrombocytopenia and hypoalbuminaemia, hypo-a1-globulinaemia, and hypergammaglobulinaemia (Fig. 1). Table 1 Serum protein electrophoresis results (g/dl) (mean standard deviation) in IFA positive, PCR positive and control group.
n TP A G A/G a1 a2 b1–2 b3
g
IFA positive
PCR positive
Control
127 7.19 1.25 3.56 0.81 3.66 0.84 1.14 0.86 0.30 0.26 1.43 0.41 0.64 0.22 0.77 0.25 0.56 0.24
18 6.43 1.03 3.13 0.64 3.31 0.72 0.98 0.28 0.25 0.07 1.26 0.36 0.61 0.20 0.65 0.16 0.55 0.18
14 6.26 0.46 3.38 0.53 2.91 0.42 1.20 0.28 0.35 0.11 0.98 0.13 0.38 0.09 0.62 0.09 0.55 0.16
n, number of samples; TP, total proteins; A, albumins; G, globulins; A/G, albumin to globulin ratio; a1–g, different protein fractions.
Table 2 Number of samples with A/G < 0.8 and their changes on electrophoretogram in different groups.
n A"/A# G"/G# a1"/a1# a2"/a2# b1–2"/b1–2# b3"/b3# g"/g#
DefNeg
DefPos
PCR
Control
24 0/23* 23*/0 11/6* 17*/2 9*/0 16*/2 6/0
3 0/3* 3*/0 0/1* 3*/0 2*/0 2/0 2/1
4 0/4* 3*/0 2/1* 2*/0 1*/0 2/0 2/0
1 0/1 0/0 1/0 0/0 0/0 1/0 1/0
n, number of dogs; DefNeg, dogs with IFA titres 1:1024; DefPos, dogs with IFA titres 1:2048; PCR, PCR positive dogs; ", increased; #, decreased; A, albumin; G, globulins; a1–g, serum protein fractions; *, significantly different than control group.
3.2. Circulating immune complexes concentration measurement results Presence of CIC was assessed in 223 serum samples – 180 were IFA positive, 24 PCR positive and 19 control. CIC were confirmed in 79.4% of IFA positive samples, 83.3% of PCR positive and 89.5% of control samples. Mean CIC concentrations and standard deviations were determined (Table 3). CIC concentrations of the IFA and PCR positive groups did not differ significantly from those of the control group, although their average CIC concentrations were higher by 1513 mg/ml. In contrast, statistically significant associations existed between CIC concentration and thrombocytopenia, hypergammaglobulinaemia and low A/G ratio. Dogs with higher CIC concentrations had lower platelets, higher gammaglobulins, and lower A/G ratios (Fig. 2). 3.3. Urinalysis results Urinalysis was conducted on 58 samples – 36 were IFA positive, 3 PCR positive and 19 control samples. Urine multitest dipstick revealed few changes in specific gravity (SG). In 15 samples SG was above, and in 5 samples below, the reference range of values. Because of only one time measurement of SG in a random urine sample and, mostly, because of no correlation with pathologic urine sediment or proteinuria (with the exception of two samples with low SG), these SG changes could not be considered to be pathological. In a microscopic evaluation of urine sediment we focused on casts and renal cells. Hyaline and fine granular casts were found in 12 samples, all in the IFA positive group. In seven of those samples, UPC was high and UPE revealed the presence of MMW and HMW proteins, with the exception of one sample, which contained only LMW proteins. 3.4. Urine protein electrophoresis results Urine protein electrophoresis (UPE) was conducted in 62 samples – 42 IFA positive, 3 PCR positive and 17 control. Proteins differing in molecular weight groups were found in 58.1% of the samples. Urine samples were divided into three groups according to the presence of low (LMW)
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
163
Fig. 1. (a–c) Significant correlation between thrombocytopenia and hypoalbuminaemia, hypo-a1-globulinaemia and hypergammaglobulinaemia. PLT, platelets; ALB, albumin; Alpha1, a1-globulins; gamma, g-globulins; P, probability value.
4. Discussion Table 3 Circulating immune complexes concentration (mg/ml) (mean standard deviation) in IFA positive, PCR positive and control group.
n CIC SD
IFA
PCR
Control
143 271.5 298.4
20 290.9 230.2
17 127.3 84.5
n, number of CIC positive samples; CIC, circulating immune complexes; SD, standard deviation.
(76 kDa) proteins. UPE results for 10 dogs in the control group showed the presence of LMW proteins – bands were observed at 10, 15, 25, and 55 kDa. All these dogs were clinically healthy, had inactive urine sediment and UPC within the reference range. All surveyed samples with LMW proteins had bands with the same molecular weight as the control group, inactive urine sediment and UPC in the reference range, therefore we considered those proteins to be physiological. Proteins with higher molecular weight were detected in 11 urine samples (30.6%), all of which were IFA positive and PCR negative. UPC in all those samples was above the reference range. These results confirmed glomerular proteinuria. CIC concentration was also found to correlate significantly with the presence of MMW and HMW proteins in urine – an increase in CIC concentration increases the occurrence of proteinuria (Fig. 3).
Fig. 2. (a and b) Significant correlation between concentration of CIC and gammaglobulins and albumin: globulin ratio. CIC, circulating immune complexes; gamma, g-globulins; AG, albumin:globulin ratio; P, probability value.
This study presents the serum protein profiles, presence of CIC and proteinuria in a large group of dogs naturally infected with A. phagocytophilum. SPE is indicated in infectious diseases, which are known to cause dysproteinaemia. Hypoalbuminaemia, hyperglobulinaemia (mainly hypergammaglobulinaemia) and hypo-a1-globulinaemia are characteristic for monocytic ehrlichiosis and leishmaniosis, which can both proceed to a chronic phase. In our study, dysproteinaemia was not significant, not even in the acute phase of anaplasmosis. Lower A/G ratios were found in 32 samples, the majority (84.4%) in groups with IFA titre 1:1024 that did not meet the criteria for active anaplasmosis (Bjo¨ersdorff, 2005). Alterations in various protein fractions correlated with those in monocytic ehrlichiosis and leishmaniosis, with the exception of gammaglobulins, which did not differ significantly from those in the control group. The major contributors to hyperglobulinaemia were the acute phase proteins a2 and b1–2 globulins. Twenty (62%) dogs in the group with lower A/G ratio had thrombocytopenia and 23
Fig. 3. Significant correlation between CIC concentration and the presence of MMW proteins in urine. CIC, circulating immune complexes; MMW, middle molecular weight proteins; P, probability value.
164
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165
(71.8%) dogs had elevated body temperature, depression, and reluctance to move, which could be associated with an acute infectious process. This could mean that dogs with low IFA titres, with subclinical or persistent infection, can develop an acute form of anaplasmosis. However, dysproteinaemia could possibly be the result of some other disease. Based on our results, SPE cannot be recommended as a diagnostic tool that would help in understanding the course of anaplasmosis. However, the significance of thrombocytopenia can be emphasized, based on its statistically significant association to acute phase proteins. In the case of clinical signs accompanied by thrombocytopenia, treatment should be considered, regardless of the level of the IFA titre. This complements our previous findings (Ravnik et al., 2011). The presence of CIC has too been described in many infectious diseases and, if the disease proceeds to the chronic phase, can cause glomerulonephritis, polyarthritis or uveitis (Pedersen, 1999). In monocytic ehrlichiosis, CIC were found in naturally and experimentally infected dogs in acute and subclinical phases of the disease (Harrus et al., 2001). The latter’s results show that the severity of the disease correlated with the presence of CIC. In our study, CIC concentration was high both in the control and study groups, which is due to the method used. Nephelometry recognizes CIC as a molecular aggregate, therefore does not distinguish between aggregates of immunoglobulin molecules alone and immune complexes that contain antigens. And since self-reactive antibodies are always present in healthy adults, those can be detected by nephelometry. But importantly, there was no significant difference in CIC concentration in any group comparing to the control group. CIC concentration was however correlated to thrombocytopenia, hypergammaglobulinaemia and lower A/G ratio. CIC concentration was significantly higher in dogs with thrombocytopenia, regardless of the IFA titre or a positive PCR result, which suggests an ongoing immunological response and again emphasises the significance of thrombocytopenia in treatment decision making in an A. phagocytophilum seropositive dog. A statistically significant association is suggested between high CIC concentration and the presence of proteinuria, but the number of samples was low (only four samples). Larger numbers of samples, together with histological and immunohistochemical analysis of the kidney tissue, need to be assessed. If the above mentioned correlation were then confirmed, CIC concentration could be used as a risk factor for the development of glomerulonephritis in A. phagocytophilum infected dogs. Proteinuria was detected in 36 (58.1%) samples. LMW proteins, present in 15 (41.7%) samples, were considered physiological since the bands were at the same positions as in the control group. MMW and HMW proteins were present in 11 (30.5%) samples, which suggest the presence of glomerulonephritis. Proteinuria can be present transiently in a febrile phase of any disease although none of these dogs were febrile at the time of urine sampling. The occurrence of proteinuria in the different groups and the control group did not differ significantly, but the fact that MMW and HMW proteins were detected exclusively in A. phagocytophilum seropositive dogs warrants regular
follow-up of A. phagocytophilum infected dogs for any later occurrence of proteinuria. Microscopic evaluation of all urine samples revealed the presence of different, mainly fine granular, casts in all proteinuric samples. We therefore recommend urine electrophoresis for all samples with microscopically confirmed casts and/or high urinary protein to creatinine ratios. Proteinuria was detected in combination with high CIC levels (100%) and thrombocytopenia (75%) in four samples with various IFA titres. Our results indicate that high CIC levels and thrombocytopenia occur in the acute phase of anaplasmosis and proteinuria probably only after a chronic antigenic stimulation (Pedersen, 1999). This suggests that persistent infection by A. phagocytophilum can result in the development of IMGN and reoccurrence of clinical signs at a particular moment. 5. Conclusions The results of this study do not support the reported development of a chronic phase of anaplasmosis in dogs (Egenvall et al., 2000; Alleman et al., 2006). Given that anaplasmosis in dogs is typically manifested seasonally, with acute clinical signs, and that the percentage of asymptomatic seropositive dogs in endemic areas is high, this is probably a disease with an acute onset and a good prognosis following the right treatment. Conflict of interest None of the authors has a financial or personal relationship with other persons or organizations that could bias the content of the article. Acknowledgments The authors thank Katarina Babnik for excellent laboratory work with CIC measurements and urine protein electrophoresis and Dr. Alberto Franceschi for serum protein electrophoresis. We also thank Prof. Dr. Tatjana Avsic Zupanc and Dr. Katja Strasek Smrdel for performing PCR and IFA for A. phagocytophilum and Prof. Roger Pain for review of English. The authors acknowledge the financial support of the Slovenian Research Agency (Project No. P4-0053). References Alleman, A.R., Chandrashekar, R., Beall, M., Cyr, K., Barbet, A., Lundgren, A., Sorenson, H., Wamsley, H., Wong, S., 2006. Experimental inoculation of dogs with a human isolate (Ny18) of Anaplasma phagocytophilum and demonstration of persistent infection following doxycycline therapy (abst). J. Vet. Intern. Med. 20, 763. Bjo¨ersdorff, A., 2005. Ehrlichiosis and anaplasmosis. Part 2: Granulocytic ehrlichiosis: anaplasma phagocytophilum comb. nov. (E. phagocytophila genogroup) infection. In: Shaw, S.E., Day, M.J. (Eds.), ArthropodBorne Infectious Diseases of the Dog and Cat. Manson Publishing, London, pp. 127–132. Day, M.J., 1987. A study of the immune response in canine disseminated aspergillosis. (Ph.D. thesis)Murdoch University. Day, M.J., 2008. The Clinical Immunology of the Dog and Cat. Manson Publishing, London, pp. 94–115. Day, M.J., 2011. The immunopathology of canine vector-borne diseases. Parasites Vectors 4, 48.
U. Ravnik et al. / Veterinary Microbiology 173 (2014) 160–165 Dumler, J.S., Trigiani, E.R., Bakken, J.S., Aguero-Rosenfeld, M.E., Wormser, G.P., 2000. Serum cytokine responses during acute human granulocytic ehrlichiosis. Clin. Diagn. Lab. Immunol. 7, 6–8. Egenvall, A., Lilliho¨o¨k, I., Bjo¨ersdorff, A., Olsson Engvall, E., Karlstam, E., Artursson, K., Heldtander, M., Gunnarsson, A., 2000. Detection of granulocytic Ehrlichia species DNA by PCR in persistently infected dogs. Vet. Rec. 146, 186–190. Feldkamp, C.S., Levinson, S.S., Perry, M., Amin, V., 1985. Anti-IgG combined with rate nephelometry for measuring polyethylene glycol-precipitated circulating immune complexes. Clin. Chem. 31, 2024–2027. Harrus, S., Day, M.J., Waner, T., Bark, H., 2001. Presence of immunecomplexes, and absence of antinuclear antibodies, in sera of dogs naturally and experimentally infected with Ehrlichia canis. Vet. Microbiol. 83, 343–349. Kim, H.Y., Mott, J., Zhi, N., Tajima, T., Rikihisa, Y., 2002. Cytokine gene expression by peripheral blood leukocytes in horses experimentally infected with Anaplasma phagocytophila. Clin. Diagn. Lab. Immunol. 9, 1079–1084. Kolbert, C.P., 1996. Detection of the agent of human granulocytic ehrlichiosis by PCR. In: Persing, D.H. (Ed.), PCR Protocols for Emerging Infectious Diseases. ASM Press, Washington, pp. 106–111. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. Levinson, S.S., Goldman, J., 1983a. Absorbance nephelometry of immune complexes by reaction with anti-IgG after treatment with polyethylene glycol. Clin. Chem. 29, 2035–2039.
165
Levinson, S.S., Goldman, J., 1983b. Measurement of circulating Clq precipitable immune complexes by a nephelometric type assay after polyethylene glycol extraction in model system. Am. Soc. Clin. Pathol. 79, 451–457. Levinson, S.S., Goldman, J., 1986. Fluorescent anti-immunoglobulin G assay of circulating immune complexes extracted with polyethylene glycol. J. Clin. Microbiol. 23, 29–32. Levinson, S.S., Goldman, J.O., Feldkamp, C.S., 1984. Anti-IgG binding test to assay circulating IgG-containing immune complexes from polyethylene glycol precipitates. Clin. Chem. 30, 1502–1506. Pedersen, N.C., 1999. A review of immunologic diseases of the dog. Vet. Immunol. Immunopathol. 69, 251–342. Ravnik, U., Tozon, N., Strasek Smrdel, K., Avsic Zupanc, T., 2011. Anaplasmosis in dogs: the relation of haematological, biochemical and clinical alterations to antibody titre and PCR confirmed infection. Vet. Microbiol. 149, 172–176. Ristic, M., Huxsoll, D.L., Weisiger, R.M., Hildebrant, P.K., Nyindo, M.B.A., 1972. Serological diagnosis of tropical canine pancytopenia by indirect immunofluorescence. Infect. Immun. 6, 226–231. Scorpio, D.G., Dumler, J.S., Barat, N.C., Cook, J.A., Barat, C.E., Stillman, B.A., DeBisceglie, K.C., Beall, M.J., Chandrashekar, R., 2011. Comparative strain analysis of Anaplasma phagocytophilum infection and clinical outcomes in a canine model of granulocytic anaplasmosis. Vector Borne Zoonotic Dis. 11, 223–229.
