Anti-erythropoietin antibody levels and its association with anaemia in different strains of semi-immune mice infected with Plasmodium berghei ANKA

The procedure for this method has been reported elsewhere[2]. Briefly, four strains of mice BALB/c, C57BL/6 (B6), CBA and New Zealand White (NZW) aged 8 weeks supplied by SLC laboratories, Fukuoka, Japan, were injected intraperitoneally (i.p.) with 10⁴Plasmodium berghei ANKA (PbANKA)–iRBCs. Parasitaemia and reticulocyte levels were monitored every two days by Giemsa-stained thin blood film and are expressed as a percentage of more than 500 RBCs. Haemoglobin (Hb) was measured in a 96-well plate at 570 nm on Bio-Rad Model 3550 Micro plate Reader as previously described[2]. A recent study[20] has shown that estimates of polychromatophilic cells as done after Giemsa staining in this study are typically less than percentages of reticulocytes. However, Giemsa has been successfully used previously for reticulocyte staining[1, 2]. Four microlitres of tail-vein blood was suspended in 1 mL Drabkin reagent (Sigma, St Louis, MO) and absorbance measured, and is expressed as a percentage of baseline levels. To generate semi-immune status, mice were treated at day 6 after infection with chloroquine/ (10 mg/kg intraperitoneally) and pyrimethamine (10 mg/kg intraperitoneally) daily for 6 days. During subsequent rounds of infection, mice were rested for two weeks before being rechallenged with 10⁴ PbANKA, then monitored and drug-cured prior to parasitemia reaching 5%. The mice strains were taken through 7 cycles of infection and treatment to generate the semi immune status, Figure 1. After the 7th cycle of infection and treatment (Figure 1), parasitaemia was checked and when it was negative (zero), which is the recovery stage, blood was collected from the tail, and serum harvested for analysis before the final cycle of infection without treatment. During the final cycle of infection (without treatment), mice were monitored every other day and, on days in which minimum Hb drop (i.e. maximum Hb reduction) was observed, blood was collected and sera were harvested for analysis, as described previously[2]. Sera were stored at −30°C until used. The sera used in this study were raised in previous study[2]. The parasitaemia and haematological profile of all the mice strains were, therefore, not reported here.

The study was conducted strictly according to the recommendation in the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science and Technology, Japan (Notice no. 71). All animal experiments were approved by the Nagasaki University, Board of Animal Research, according to Japanese Guideline for use of experimental animals (Permit Number 0811130716). All efforts were made to humanely minimize animal suffering.

Ultrabind US450 membrane (Gelman Sciences, Japan) was cut according to size and pre-wet with enough volume of tris-buffered saline (TBS). The membrane was placed unto the manifold and 300 μL TBS used to wash wells and membranes. Recombinant mouse EPO (R&D Systems, Japan, Catalogue Number 959-ME) was used as antigen to coat the membrane overnight at 4°C. The membrane in each well was blocked with 100 μL BlockAce (Catalogue No. UK-B25, Dainippon Pharmaceutical Co. Osaka, Japan) for an hour. Membranes in the wells were later washed once with 300 μL tris-buffered saline with tween (TTBS). Primary antibody (from infected sera and uninfected mice sera as negative control) were diluted at 1:10, 1:20 and incubated with the mouse EPO antigen for 3 hours at RT. The membranes in the wells were washed with 300 μL of TTBS twice before they were finally removed from the manifold and washed once in TTBS on a rotating machine. Secondary antibody of goat anti-mouse IgG-conjugated horse-radish peroxidase (HRP) at 1:2,000 dilution was incubated with the membrane for an hour at room temperature (RT) in the dark. Membrane was washed four times with TTBS and once with TBS to remove residual TTBS from membrane. Color development was done by addition of diaminobenzidine (DAB, at 5.0 mg/10 ml PBS + 5 uL H₂0₂) and incubated at RT for 10–15 min.

This was done based on modified method according to Tzioufas et al.[9]. The recombinant mouse EPO (R&D Systems, Japan, Catologue Number 959-ME) was used as antigen. It was dissolved in phosphate buffered saline (PBS, pH 7.2), and coated in the polystyrene micro titer plate at 0.09 μg per well. After incubation overnight at 4°C, plates were washed with 0.1% Tween 20/PBS and blocking was done with Block Ace (Catalogue No. UK-B25, Dainippon Pharmaceutical Co. Osaka, Japan) for 1 hour. Diluted serum samples from the infected mice and uninfected mice (serving as negative control) all at 1:1,000 dilutions was added to the wells in duplicate and incubated for three hours at 37°C. After washing for five times with 0.1% Tween 20/PBS, 100 μL horse radish peroxidase (HRP)-conjugated goat anti-mouse IgG (Catalogue number 55563, MP Biomedicals, LLC 29525 Fountain Pkwy Solon, OH 44139, USA) at 1:2,500 was added to the wells in the 96-well ELISA plates and incubated for an hour at RT. Washing was done five times with 0.1% Tween 20/PBS and reaction visualized by the addition of 100 μL stabilized chromogen (Part/Lot number SS01/302008, Biosource, Carmarillo, CA, USA). Reaction was stopped 20 minutes later by the addition of 50 μL 1 N H₂SO₄ and absorbance was measured in the ELISA reader machine at 450 nm.

Measurement of mouse EPO was done according to manufacturer’s instructions of the Quantikine ELISA kit for mouse/rat EPO Immunoassay (R and D Systems, Japan, Catalogue number MEP00). Briefly 50 μL of standards, control and samples were pipette into the wells of a monoclonal antibody specific for mouse/rat EPO pre-coated micro- plate, which already has 50 μL of the assay diluents added. This was incubated at RT for 2 h on a shaker. The wells were washed five times and 100 μL of the conjugate added, then incubated at RT for 2 h on the shaker. Again washing was done five times with 100 μL substrate solution added to each well and incubated for 30 min on the bench top in the dark. After 100 μL stop solution was added to each and OD read at 450 nm.

Cytokines measurement was done using Procarta Mouse Cytokine Assay Kit plex according to manufacturer’s instructions (Luminex, Affymetrix). Briefly, filter plate was wetted with reading buffer and incubated at room temperature for 5 min, then filtered. Antibody beads (50 μL) were added to each well, filtered and washed once with 150 μL of washing buffer. Twenty-five microlitres of serum standard buffer was added to all the sample wells, and later equal volume (25 μL) of serum was added to each well, incubated at room temperature for 60 min. Later, washing was done three times, after which 25 μL premixed detection antibody was added, incubated for 30 min on the shaker at room temperature. After, washing and filtration was done three times. During each wash 150 μL of washing buffer was used. Streptavidin-PE (50 μL) was later added and incubated for 30 min, later washed and filtered three times. After the third wash the plate was prepared for analysis. Reading buffer (120 μL) was added to each instrument.

Serum transferrin was measured according to manufacturer’s instruction (ELISA kit from Alpha diagnostic International, USA, Catologue number 6390). Briefly, 100 uL of calibrators, control and samples were pipette into pre-designated wells already coated with anti-transferin antibodies, all in duplicates. This was incubated at room temperature (RT) for thirty minutes. The plate was covered during incubation. The contents of the wells were aspirated following incubation. Thereafter, each well was completely filled with diluting wash solution and aspirated. This was repeated three times, for a total of four washes. Thereafter, 100 uL of appropriately diluted enzyme-antibody conjugate was added to each well and incubated at room temperature for 30 min in the dark. The plate was covered and kept level during incubation. Washing was done and plates blotted four times as done previously. Thereafter, 100 μL of 3,3′,5,5′ tetramethylbenzidine (TMB) was added to each well and incubated for 10 min in the dark; after which 100 uL of stop solution was added to each well. The absorbance of each well was read at 450 nm.

Data analysis was done using the GraphPad Prism Version 5.00 for Windows, GraphPad Software, San Diego California, USA,[21]. Data are expressed as the mean with standard deviation (SD) unless otherwise stated. Data were log transformed to ensure normal distribution before one-way analysis of variance (ANOVA, with Tukey’s post-test, two tailed), were performed. Pearson correlation analysis was performed on the transformed data of variables to compare the relationship between them. Values were considered significant when p < 0.05.

Sera of the mice strains were screened by immunodot blot for the possible identification of anti-EPO antibody levels. It can be seen from Figure 2 that sera from all the mice strains showed the presence of anti-EPO antibody, but negative for uninfected control sera, suggesting that parasitaemia might contribute to the production of the anti-EPO antibody. The anti-EPO antibody levels were measured, at infection and at recovery. Anti-EPO antibody was significantly different in the mice strains (mean OD_450nm values: Balb/c (2.1); B6 (1.3); CBA (1.4); NZW (1.7)) at infection (p = 0.045). While Balb/c anti-EPO Ab levels at infection differed from B6 (p < 0.05), the others were not significantly different when paired. The anti-EPO antibody is significantly higher in all four mice strains at infection when compared with the uninfected control (mean OD_450nm value = 0.1) (p < 0.0001). Anti-EPO Ab levels at recovery in the mice strains (Balb/c, 1.8; B6, 1.1; CBA, 1.5; NZW, 1.0) were significantly different, p = 0.0004. Again anti-EPO Ab levels of Balb/c at recovery were significantly different from that of B6 and NZW (p < 0.05), while the others did not differ (p > 0.05). The means of anti-EPO Ab levels at infection and recovery differed (p = 0.0008) (Figure 3).

To explore the association of anti-EPO antibody with anaemia, the level of anti-EPO antibody was plotted against Hb loss in all the mice strains giving interesting results as seen in Figure 4 (combined results for all the mice strains) and Figure 5 (individually). Anti-EPO antibody correlated significantly with Hb loss when all the mice strains was combined, Figure 4 (r = 0.5681; p = 0.003). Plotting anti-EPO antibody against Hb loss per mice strains (Figure 5), it was observed that the correlation was significant (r = 0.7651; p = 0.028) in only Balb/c but not in other strains B6 (r = 0.304; p = 0.5991), CBA (r = 0.2592; p = 0.619) and NZW (r = 0.8223; p = 0.1777).

To further understand the erythropoietic response in all the semi-immune mice strains the EPO levels were evaluated at both infection and at recovery. It was consistently observed for each mice strain that, EPO at recovery was significantly lower than that at infection (Figure 6), confirming earlier reports[5‐7]. However, when EPO levels at infection was compared in all the mice strains, EPO levels in NZW was significantly higher than Balb/c (p < 0.05).

Since EPO level is a measure of how active the individual is able to respond to Hb drop, the association of EPO level with the anti-EPO antibody level in all the semi-immune mice strains was analyzed but found no significant correlation. However, when this association was analysed per mice strain, a significant negative association was found only in Balb/c mice strain at infection stage, Figure 7 (r = −0.83; p = 0.01). No significant association was found in the other mice strains both at infection and recovery, Figure 7.

While IL-6 and INF-γ are known to suppress erythropoiesis, TNF and IL-10 are thought to contribute to the degree of anaemia in children with falciparum malaria[22, 23]. These cytokines were measured to evaluate the extent to which they may be implicated in anaemia and anti-EPO antibody production indirectly. IL-17 is thought to be associated with autoimmunity. Significantly high levels of IL6, IL-17 and INF-γ were observed in Balb/c, p < 0.05 (Table 1). Interleukin 10 (IL-10) and TNF were similar in all the mice strains, p = 0.08, 0.07 respectively; TNF/IL10 ratio on the other hand were significantly different in all the mice strains, p = 0.02. To further understand the relationship between the cytokines and anti-EPO antibody, anti-EPO antibody was plotted against the cytokine levels. It was observed that there was no correlation (p > 0.05) between all the cytokines and anti-EPO antibody in all the mice strains.

Transferrin is known to regulate iron metabolism and also an indicator of inflammatory response during infection. Thus, to evaluate the extent to which iron is metabolized at minimum Hb (maximum Hb reduction, see Figure 1 in[2]) and possible implication with immune response transferrin levels were measured at infection and found out that the levels was significantly low in the Balb/c when compared with the other mice strains. An elevated level of transferrin was also observed at infection for all the semi-immune mice strains when compared with the uninfected mice (Figure 8).