- Email: [email protected]
Comparison of murine Epo ELISA and Epo bioassays in detecting serum Epo levels during anemia associated with malaria infection
Journal of Immunological Methods 262 (2002) 129 – 136 www.elsevier.com/locate/jim
Comparison of murine Epo ELISA and Epo bioassays in detecting serum Epo levels during anemia associated with malaria infection Kai-Hsin Chang a,b, Mary M. Stevenson a,b,* a
Institute of Parasitology, Macdonald Campus, McGill University, 21, 111 Lakeshore Road, Ste-Anne-de-Bellevue, Quebec, Canada H9X 1C9 b Centre for the Study of Host Resistance, Montreal General Hospital Research Institute, and Department of Medicine, MUHC, McGill University, 1650 Cedar Avenue, Montreal, Quebec, Canada H3G 1A4 Received 23 August 2001; received in revised form 21 December 2001; accepted 9 January 2002
Abstract A highly sensitive sandwich ELISA specific for murine erythropoietin (mEpo) was developed using commercially available monoclonal anti-mouse Epo antibody and polyclonal anti-human Epo antibody. This newly developed Epo ELISA protocol and the traditional Epo bioassay method were used to analyze Epo production in response to anemia induced during blood-stage Plasmodium chabaudi AS (P. chabaudi AS) malaria infection in C57BL/6 mice. Both methods revealed an inverse correlation between the serum Epo concentration and hematocrit level, but Epo values estimated by the Epo bioassay were between 5- and 20-fold higher than those estimated by the ELISA. Further study demonstrated that the estimated Epo level in bioassay was strongly influenced by other cytokines present in the samples. Therefore, the Epo bioassay detects the net erythropoiesis promoting activities, whereas the ELISA method specifically measures the level of Epo in the samples. Combined with the Epo bioassay, the murine Epo ELISA will be an extremely useful tool in specifically measuring the Epo response and facilitating the understanding of mechanisms involved in the development of anemia-associated diseases using mouse models. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Murine Epo ELISA; Epo bioassay; Malaria; Anemia; Serum factors
1. Introduction Epo is the essential growth factor that promotes the growth, proliferation, and differentiation of erythroid
Abbreviations: Epo, erythropoietin; rmEpo, recombinant murine Epo; hEpo, human Epo. * Corresponding author. Centre for the Study of Host Resistance, Montreal General Hospital Research Institute, 1650 Cedar Avenue, Montreal, Quebec, Canada H3G 1A4. Tel.: +1-514-934-1934x 44508; fax: +1-514-934-8332. E-mail address: [email protected] (M.M. Stevenson).
lineage cells and leads to the increase in circulating red blood cell volume. Epo is mainly produced in adult kidneys and is upregulated in response to tissue hypoxia to correct anemia. Furthermore, Epo is also produced in the brain and can prevent loss of stressed neurons in hypoxia – ischaemia. Epo administration has been proven to be an effective remedy for treating anemia in AIDS (Moore et al., 1998), in renal failure (Tsakiris, 2000), and in chronic inflammatory diseases (Tanaka et al., 1999), as well as for limiting brain damage in experimental autoimmune encephalomyelitis (Brines et al., 2000). Because murine models have been extensively used for studying various anemia-
0022-1759/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 2 2 - 1 7 5 9 ( 0 2 ) 0 0 0 1 2 - 1
130
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
associated disorders as well as for testing drug treatment strategies, it would be extremely helpful to be able to specifically measure murine Epo (mEpo) in serum and in other tissues, such as the kidney and brain. Epo bioassay has been widely used for measuring both human Epo (hEpo) and mEpo in sera and tissue culture media until the development of hEpo ELISA, which is less time-consuming, more convenient, and does not use radioisotopes. Unfortunately, the hEpo ELISA is not sensitive enough to accurately measure basal mEpo levels in mouse serum (Rinaudo and Toniatti, 2000). Recently, ELISAs specific for mEpo were described by two research groups using either self-raised polyclonal antibodies (Noe et al., 1999) or commercially purchased human Epo ELISA kits (Rinaudo and Toniatti, 2000). Here, we describe a mEpo ELISA, which is equally sensitive, more costeffective, and with a larger reliable range, using commercially available anti-Epo antibodies. We further compared the ELISA method and the traditional Epo bioassay in the study of anemia induced during bloodstage Plasmodium chabaudi AS (P. chabaudi AS) malaria infection. The effects of other serum factors on the accuracy of these two methods were also studied.
2. Type of research Quantifying murine erythropoietin in serum.
3. Time required Coating plates with antibodies at 4 C: overnight Wash plates: 3 min/plate (repeated between each step) Blocking plates with blocking buffer at room temperature: 1 h Preparing serum samples and mEpo standard serial dilutions and adding to the plate: 10 min/plate Samples and standards binding at 4 C: overnight Adding detecting antibody and incubating at room temperature: 1 h Adding anti-rabbit IgG-HRPO conjugate and incubating at room temperature: 1 h Adding substrate and color development at room temperature: 30 min.
4. Materials and methods 4.1. Animals and parasites P. chabaudi AS parasites were maintained as previously described (Yap and Stevenson, 1992). Eightweek-old male C57BL/6 mice were purchased from Charles River (St. Constant, QC, Canada). All animal care and handling were carried out in accordance with institutional guidelines of the Montreal General Hospital Research Institute and McGill University. Mice were infected i.p. with 106 parasitized red blood cells (PRBC). Blood was collected by cardiac puncture at different time points following infection, allowed to clot at room temperature, and centrifuged. Serum was collected and stored at 20 C. 4.2. Murine EPO ELISA optimization Monoclonal rat anti-mEpo antibody IgG1 (1 mg/ ml, BD BioSciences, Mississauga, ON, Canada) and polyclonal rabbit anti-hEpo antibody (2 mg/ml, R&D Systems, Minneapolis, MN, USA) were tested as either capturing or detecting antibody. For coating, antibodies were diluted 1:250, 1:500, or 1:1000 in PBS. For detecting, antibodies were diluted 1:200, 1:400, or 1:1000 in blocking buffer, which consisted of PBS and 1% BSA (Gibco Laboratories, Invitrogen, Burlington, ON, Canada). HRPO-conjugated goat anti-rabbit IgG(H + L) (1:3000 dilution, Bio-Rad Laboratories, Hercules, CA, USA) or goat anti-rat IgG (1:1000 dilution, Bio-Rad Laboratories) antibodies diluted in blocking buffer was applied as secondary antibody depending on the primary detecting antibody chosen. All the procedures were performed in the wells of Immulon 2HB U Bottom Microtiter Plates (Dynex Technologies, Chantilly, VA, USA). Following each incubation step, plates were washed three times with 200 ml/well washing buffer, which consisted of 0.1% Tween 20 (Sigma, St. Louis, MO, USA) in PBS. Plates were first coated with 50 ml/well of appropriately diluted capturing antibodies overnight at 4 C and subsequently blocked with 250 ml of blocking buffer for 1 h at room temperature. Recombinant mEpo (rmEpo) standard was serially diluted (0.98 – 500 mU/ml; Boehringer Mannheim-Roche, Laval, QC, Canada) in blocking buffer, and 50 ml/well of
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
each dilution was added to the plates and incubated at either 4 C overnight or at room temperature for 2 h. Furthermore, 50 ml/well of detecting antibodies were added and incubated for 1 h at room temperature, followed by incubating with 50 ml/well of HRPO conjugates for 1 h at room temperature. The bound conjugate was visualized by adding 100 ml/well of ABTS substrate (Boehringer Mannheim-Roche) with 0.036% H2O2. OD values were read at various time points (15 min to 1 h) after addition of substrate in a microplate reader at a wavelength of 405 nm with 492 nm as reference. Once the optimal conditions were determined, changes in final OD values, in the presence of cytokines added to known concentrations of rmEpo standard (0, 1.95, 15.625, and 125 mU/ml) in the step of sample/standard incubation, were determined to evaluate the specificity of this ELISA. Cytokines tested were TNF-a (0.1 –10 ng/ml, R&D systems), IFN-g (0.5 –50 ng/ml, R&D systems), IL-12 (2.8 –100 ng/ml, a gift from Genetics Institute, Cambridge, MA, USA), IL-3 (0.125 – 12.5 ng/ml, Cedarlane Laboratories, Hornby, ON, Canada), and stem cell factor (SCF) (2– 50 ng/ml, Cedarlane Laboratories).
131
4.4. Epo bioassay
4.3. Detecting EPO in murine serum using newly developed murine EPO specific sandwich ELISA
Serum Epo concentration was also measured by the Epo bioassay as previously described (Yap and Stevenson, 1992). Briefly, Iscove’s MDM (IMDM) supplemented with 10% heat-inactivated FCS (HyClone, Logan, UT, USA), 5 10 5 M 2-mercaptoethanol (Sigma), and 48 mg/l Gentamicine (Schering Canada, Pointe-Claire, QC, Canada) was used as culture medium throughout the assay. C57BL/6 mice were rendered anemic by two consecutive daily i.p. injections of phenylhydrazine hydrochloride (60 mg/kg, Fischer Scientific, Fair Lawn, NJ, USA). Spleens were harvested 3 days after the second injection and gently pressed through fine wire meshes and passed through 18-G needles. After washing three times with sterile PBS, the spleen cells were resuspended in culture medium and adjusted to 4 106 cells/ml to be used as responder cells. Aliquots of 100 ml/well of mouse sera diluted four-fold in culture medium as well as serially diluted rmEpo (0.98 – 500 mU/ml) were added into 96-well culture plates (Becton Dickinson Labware, Franklin Lakes, NJ, USA) followed by the addition of 100 ml/well responder cells. After 22 h of incubation at 37 C, 2 mCi 3H-thymidine (specific activity 6.7 Ci/mol, ICN Biomedicals, Irvine, CA, USA) were added to each well and incubated for
Aliquots of 50 ml/well of monoclonal rat anti-mEpo antibody (4 mg/ml in PBS) were added into plates and incubated at 4 C overnight. Following washing and blocking, 50 ml/well aliquots of 0.98 – 500 mU/ml rmEpo standard and 1:3 serial dilutions of serum samples in blocking buffer were added and incubated at 4 C overnight. Aliquots of 50 ml/well of polyclonal rabbit anti-hEpo antibody (10 mg/ml in blocking buffer) were then added and incubated at room temperature for 1 h, followed by incubation with 50 ml/well of HRPO-conjugated goat anti-rabbit IgG(H + L) antibody (1:3000 in blocking buffer) at room temperature for 1 h. After each step, plates were washed three times with washing buffer. The bound conjugate was then detected by adding 0.1 ml/well ABTS substrate with 0.036% H2O2 and developed for 30 min and read at a wavelength of 405 nm with 492 nm as reference. The Epo concentration in serum samples was calculated against the standard curve generated with rmEpo.
Fig. 1. Optimization of the mEpo-specific ELISA. Representative standard curves are shown after color development and OD reading. Plates were coated with monoclonal rat anti-mEpo antibody and the captured mEpo was detected with polyclonal rabbit anti-hEpo antibody at concentrations of 2, 5, or 10 mg/ml.
132
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
2 h. Cells were harvested with an automatic cell harvester and 3H-thymidine incorporation, which reflected the degree of spleen cell proliferation, was determined by scintillation counting. A standard curve was generated by plotting 3H-thymidine incorporation against the concentration of serially diluted rmEpo. Epo concentration in the samples was estimated by calculating the 3H-thymidine incorporation against the standard curve. To determine the specificity of the Epo bioassay, 100 ml/well of responder cells were added to 100 ml/ well of culture medium containing known concentrations of rmEpo standard (0, 1.95, 15.625, or 125 mU/ ml) in the presence or absence of TNF-a (0.1 – 10 ng/ ml), IFN-g (0.5 – 50 ng/ml), IL-12 (2.8 –100 ng/ml), IL-3 (0.125 – 12.5 ng/ml), or stem cell factor (SCF) (2– 50 ng/ml). The amounts of 3H-thymidine incor-
poration were compared between cultures with and without addition of exogenous cytokines. The specificity of the Epo bioassay was also tested by the addition of an antibody cocktail that contained anti-TNF-a (102 Neutralizing U/ml, clone XT22 ascites), anti-IFN-g (1.67 mg/ml, clone XMG1.2), and anti-IL-12 (3.9 mg/ml, clone C17.8) into serum samples to block the effect of these cytokines. The changes in 3H-thymidine incorporation of the responder cells were compared to those incubated with serum samples alone. 4.5. Statistical analyses Values are expressed as mean F standard error of the mean (S.E.M.). Statistical significance of differences was analyzed by Student’s t-test or ANOVA
Fig. 2. Course of infection and changes in hematologic parameters during P. chabaudi AS infection in C57/BL6 mice. (A) Course of parasitemia, (B) degree of anemia as determined by hematocrit level, (C) serum Epo concentration estimated by ELISA, and (D) serum Epo concentration estimated by Epo bioassay. The results are expressed as mean F S.E.M. of three to four mice. Similar results were obtained in replicate experiments.
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
followed by Tukey’s studentized range test. A p < 0.05 was considered to be significant. Serum Epo values were converted to their log values before performing regression analyses. The parallel lines assumption was tested to determine the statistical significance of differences between the slopes of regression lines of the ELISA and bioassay in our model. Statistical analyses were performed using SAS-STAT software (SAS Institute, Cary, NC, USA).
5. Results 5.1. Murine Epo ELISA optimization In preliminary studies, to establish the optimal conditions for the ELISA, we found that the use of polyclonal rabbit anti-hEpo antibody as capturing antibody yielded high background and low sensitivity regardless of the dilution chosen (data not shown). When monoclonal rat anti-mEpo was used as the capturing antibody, the reliable range of the standard curve increased when lower dilutions of both capturing and detecting antibodies were used. We also found that incubation of rmEpo and serum sample at 4 C overnight yielded a lower background and higher sensitivity than short time incubation at room temperature. The optimal conditions of this Epo ELISA were determined to be the use of 4 mg/ml of monoclonal rat anti-mEpo antibody as the capturing antibody and 10 mg/ml of polyclonal rabbit anti-hEpo antibody as the detecting antibody with overnight incubation of test samples and standards at 4 C. Under these conditions, the ELISA can detect mEpo as low as 1 mU/ml and the reliable range extends to a high of 300 mU/ml (Fig. 1). When this ELISA was tested for its specificity, addition of a wide concentration range of TNF-a, IFN-g, IL-12, IL-3, and SCF to known concentrations of rmEpo standard at the standard/sample incubation step did not alter the OD values compared to incubation with rmEpo standard alone (data not shown).
133
parasitemia then gradually declined and became almost undetectable by day 11 post infection (p.i.). Anemia, as measured by hematocrit level, developed progressively as the parasite load increased (Fig. 2B) and became most severe at day 8 to day 9 p.i. We measured the serum Epo concentrations by both the sandwich ELISA described here (Fig. 2C) and by the bioassay (Fig. 2D). Both assays showed that Epo production reached its highest level between day 8 and day 9 p.i. when the mice were most anemic. Regression analysis of serum Epo levels against hematocrit levels revealed an inverse correlation in both the ELISA and bioassay ( p < 0.05) with high R2 values (Fig. 3). The parallel lines assumption was tested and the slopes were not significantly different from one another but the estimated Epo levels were between 5- and 20-fold higher in bioassay than in ELISA ( p < 0.05). 5.3. Effects of other serum factors in the accuracy of bioassay The specificity of the Epo bioassay was tested by incubating known concentrations of rmEpo in the presence or absence of additional serum factors including TNF-a, IFN-g, IL-12, IL-3, and SCF. For all
5.2. Epo production during blood-stage P. chabaudi AS infection Following P. chabaudi AS infection, C57BL/6 mice developed increased parasitemia which reached peak level at day 7 after infection (Fig. 2A). The
Fig. 3. Inverse correlation between hematocrit level and serum Epo concentration measured by murine-specific Epo ELISA and by bioassay. Twenty-five samples were measured in each group. p < 0.05 for both regression lines.
134
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
concentrations tested, TNF-a (0.1 – 10 ng/ml) and IFN-g (0.5 – 50 ng/ml) significantly suppressed the proliferation of responder cells in the absence of rmEpo or with 1.95 – 125 mU/ml rmEpo (Fig. 4 and data not shown). However, IL-12 (2.8 – 100 ng/ml) and SCF (2 –50 ng/ml) stimulated the proliferation of responder cells in the absence of rmEpo or with 1.953 mU/ml rmEpo but inhibited the proliferation induced by higher amounts of rmEpo (15.625 – 125 mU/ml). Furthermore, IL-3 at higher concentrations (1.25 – 12.5 ng/ml) showed significant stimulatory effects on cultures in the absence of rmEpo or with 1.953 – 15.625 mU/ml rmEpo. However, at a lower concentration (0.125 ng/ml), IL-3 stimulated the proliferation in the absence of rmEpo or with 1.953 mU/ml rmEpo but inhibited the proliferation induced by a higher dose of rmEpo (125 mU/ml). Based on the previously published serum cytokine data in malaria-infected mice or patients, we summarized the effects of 1 ng/ ml TNF-a, 50 ng/ml IFN-g (Su and Stevenson, 2000), 2.8 ng/ml IL-12 (Sam and Stevenson, 1999; Su and Stevenson, 2000), 0.125 ng/ml IL-3, and 2 ng/ml SCF (Burgmann et al, 1997) on the Epo bioassay in Fig. 4. Furthermore, when a mixture of anti-IL-12, anti-TNFa, and anti-IFN-g antibodies was added to sera from
Fig. 4. Effects of cytokines on the proliferation of bioassay responder cells. Spleen cells from phenylhydrazine-treated mice were incubated with four different concentrations of rmEpo in the presence or absence of exogenous cytokines. Each bar represents the amount of 3 H-thymidine incorporated and is expressed as mean F S.E.M. of four samples. Similar results were obtained in replicate experiments. * p < 0.05 compared with no addition of exogenous cytokine.
infected mice to neutralize the activity of these cytokines, the 3H-thymidine incorporation in the Epo bioassay increased significantly in comparison with responder cells cultured with sera alone (data not shown).
6. Discussion Epo has been shown to be important in correcting anemia as well as protecting neuron cells in hypoxia– ischaemia. To facilitate research in both hematology as well as neurology, a specific and accurate method of measuring Epo concentration in serum and tissues is required. For human Epo, ELISA kits, which provide a quick and convenient way of measuring Epo, are commercially available. Unfortunately, although human and murine Epo molecules share 80% homology, hEpo ELISA is not sensitive enough for detecting murine Epo (Rinaudo and Toniatti, 2000). Recently, an ELISA specific for mEpo with a reliable range of 0.6 –30 mU/ml was described by Noe et al. (1999), who combined self-raised polyclonal anti-hEpo and anti-mEpo antibodies. This was followed by a mEpo ELISA developed by Rinaudo and Toniatti (2000), who used two monoclonal anti-hEpo antibodies from commercially purchased hEpo ELISA kits and achieved a sensitivity of 1 mU/ml. Here, we report an mEpo-specific ELISA, which is equally sensitive (1 mU/ml) and yet provides a wider reliable range (up to 300 mU/ml) using commercially available antibodies. As serum Epo levels can be elevated up to 1000 mU/ml (Fig. 2C and Noe et al., 1999), an ELISA with the ability to detect a wider reliable range eliminates the need of further diluting samples and thus avoids the error introduced by high sample dilutions. Prior to the development of Epo ELISAs, Krystal (1983) established the Epo bioassay, which was considered to be more specific for Epo than in vitro 59Fe bioassay methods. This bioassay has been widely used for detecting Epo levels in serum and culture medium. When we compared the Epo levels in sera of normal and P. chabaudi AS-infected mice detected by the newly established ELISA and by the bioassay, an inverse correlation between the serum Epo concentrations and hematocrit levels during P. chabaudi AS infection was observed regardless of the assay used.
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
While these two methods provided comparable slopes ( 0.0461 and 0.0621, respectively) and R2 values (0.8556 and 0.9063, respectively), serum Epo levels estimated by the Epo bioassay were significantly higher, which was consistent with the estimated intercepts of both regression lines (4.4342 vs. 3.9078, respectively). Since the bioassay estimates Epo levels in samples based on the amount of 3H-thymidine incorporated by responder cells, the effects of other blood elements besides Epo on the proliferation of responder cells might be partially responsible for the difference observed. To test this hypothesis, the responder cells from spleens of phenylhydrazine-treated mice were incubated with known concentrations of rmEpo supplemented with several cytokines known to be present in the sera of malaria-infected patients and experimental animals (Burgmann et al., 1997; Fried et al., 1998; Othoro et al., 1999; Sam and Stevenson, 1999; Su and Stevenson, 2000). Consistent with their inhibitory role in erythropoiesis (Allen et al., 1999), TNF-a and IFNg suppressed the Epo-stimulated spleen cell proliferation at all concentrations tested. Surprisingly, IL-12, which was shown to enhance murine erythropoiesis in vitro in combination with IL-4 and SCF (Dybedal et al., 1995), exerted stimulatory effects when cultured with lower doses of rmEpo but exerted negative effects when cultured with higher doses of rmEpo. This phenomenon was also observed at all concentrations of SCF and the lowest dose of IL-3 tested. Therefore, when measuring lower doses of Epo in the presence of IL-12, SCF, or IL-3 with the bioassay, the stimulatory effects of these cytokines increase 3Hthymidine incorporation and lead to an overestimation of Epo concentrations. In contrast, when measuring higher doses of Epo in the presence of IL-12, SCF, TNF-a, and IFN-g, the inhibitory effects of these cytokines resulted in an underestimation of Epo concentrations, which is supported by the significantly increased serum bioactivity when TNF-a, IFN-g, and IL-12 were neutralized. The 5 –20-fold higher levels of Epo detected by the bioassay in comparison to those detected by the ELISA at all Epo concentrations were possibly due to the presence of other factors not included in this study, such as IL-10, IL-4, GM-CSF, TGF-b, and other unknown serum factors. These serum factors may synergistically promote erythropoiesis with high levels of Epo and cause an over-
135
estimation of overall Epo concentrations in the bioassay. Taken together, our data suggest that the cell proliferation in the bioassay, which estimates the Epo concentrations in samples, is greatly influenced by several factors including the presence of other hematopoietic cytokines in the samples, the concentrations of these cytokines, and the interaction of Epo and these cytokines. To estimate Epo concentrations accurately using the bioassay, it is essential to either remove all these cytokines from the samples, or extensively characterize and exclude the impact of all serum factors in each sample on the proliferation of responder cells. Without these modifications, data obtained from the Epo bioassay represent the net erythropoiesis promoting activity of the sample, which is the balance of the inhibitory and stimulatory elements. In contrast, the Epo ELISA, which is not influenced by the presence of other serum factors in the samples, provides a more reliable estimation of Epo concentrations. In conclusion, our newly developed murine Epo ELISA provides a convenient, sensitive, cost-effective, and highly reliable way to measure murine Epo levels in serum and other organs. In contrast, data generated by the Epo bioassay should be interpreted as net erythropoiesis promoting/inhibitory activities rather than the level of Epo specifically. A combination of these two assays will facilitate the research in malaria and other diseases using murine models.
7. Essential references Krystal, 1983. Noe et al., 1999. Tsakiris, 2000.
Acknowledgements The authors wish to thank Fabrice Rouach for statistical consultation. This research was supported by the Medical Research Council of Canada (Canadian Institutes of Health Research) grant number MT14663. K-H Chang was a recipient of the Lloyd Carr-Harris McGill Major Fellowship from McGill University, Montreal, Quebec, Canada.
136
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
References Allen, D.A., Breen, C., Yaqoob, M.M., Macdougall, I.C., 1999. Inhibition of CFU-E colony formation in uremic patients with inflammatory disease: role of IFN-gamma and TNF-alpha. J. Investig. Med. 47, 204. Brines, M.L., Ghezzi, P., Keenan, S., Agnello, D., de Lanerolle, N.C., Cerami, C., Itri, L.M., Cerami, A., 2000. Erythropoietin crosses the blood – brain barrier to protect experimental brain injury. PNAS 97, 10526. Burgmann, H., Looareesuwan, S., Wiesinger, E.C., Winter, W., Graninger, W., 1997. Levels of stem cell factor and interleukin-3 in serum in acute Plasmodium falciparum malaria. Clin. Diagn. Lab. Immunol. 4, 226. Dybedal, I., Larsen, S., Jacobsen, S.E., 1995. IL-12 directly enhances in vitro murine erythropoiesis in combination with IL-4 and stem cell factor. J. Immunol. 154, 4950. Fried, M., Muga, R.O., Misore, A.O., Duffy, P.E., 1998. Malaria elicits type 1 cytokines in the human placenta: IFN-gamma and TNF-alpha associated with pregnancy outcomes. J. Immunol. 160, 2523. Krystal, G., 1983. A simple microassay for erythropoietin based on 3 H-thymidine incorporation into spleen cells from phenylhydrazine treated mice. Exp. Hematol. 11, 649. Moore, R.D., Keruly, J.C., Chaisson, R.E., 1998. Anemia and survival in HIV infection. J. Acquired Immune Defic. Syndr. 19, 29.
Noe, G., Riedel, W., Kubanek, B., Rich, I.N., 1999. An ELISA specific for murine erythropoietin. Br. J. Haematol. 104, 838. Othoro, C., Lal, A.A., Nahlen, B., Koech, D., Orago, A.S., Udhayakumar, V., 1999. A low interleukin-10 tumor necrosis factoralpha ratio is associated with malaria anemia in children residing in a holoendemic malaria region in western Kenya. J. Infect. Dis. 179, 279. Rinaudo, D., Toniatti, C., 2000. Sensitive ELISA for mouse erythropoietin. BioTechniques 29, 218. Sam, H., Stevenson, M.M., 1999. In vivo IL-12 production and IL12 receptors beta1 and beta2 mRNA expression in the spleen are differentially up-regulated in resistant B6 and susceptible A/J mice during early blood-stage Plasmodium chabaudi AS malaria. J. Immunol. 162, 1582. Su, Z., Stevenson, M.M., 2000. Central role of endogenous gamma interferon in protective immunity against blood-stage Plasmodium chabaudi AS infection. Infect. Immun. 38, 4399. Tanaka, N., Ito, K., Ishii, S., Yamazaki, I., 1999. Autologous blood transfusion with recombinant erythropoietin treatment in anaemic patients with rheumatoid arthritis. Clin. Rheumatol. 18, 293. Tsakiris, D., 2000. Morbidity and mortality reduction associated with the use of erythropoietin [Review]. Nephron 85 (Suppl. 1), 2. Yap, G.S., Stevenson, M.M., 1992. Plasmodium chabaudi AS: erythropoietic responses during infection in resistant and susceptible mice. Exp. Parasitol. 75, 340.
Comparison of murine Epo ELISA and Epo bioassays in detecting serum Epo levels during anemia associated with malaria infection Kai-Hsin Chang a,b, Mary M. Stevenson a,b,* a
Institute of Parasitology, Macdonald Campus, McGill University, 21, 111 Lakeshore Road, Ste-Anne-de-Bellevue, Quebec, Canada H9X 1C9 b Centre for the Study of Host Resistance, Montreal General Hospital Research Institute, and Department of Medicine, MUHC, McGill University, 1650 Cedar Avenue, Montreal, Quebec, Canada H3G 1A4 Received 23 August 2001; received in revised form 21 December 2001; accepted 9 January 2002
Abstract A highly sensitive sandwich ELISA specific for murine erythropoietin (mEpo) was developed using commercially available monoclonal anti-mouse Epo antibody and polyclonal anti-human Epo antibody. This newly developed Epo ELISA protocol and the traditional Epo bioassay method were used to analyze Epo production in response to anemia induced during blood-stage Plasmodium chabaudi AS (P. chabaudi AS) malaria infection in C57BL/6 mice. Both methods revealed an inverse correlation between the serum Epo concentration and hematocrit level, but Epo values estimated by the Epo bioassay were between 5- and 20-fold higher than those estimated by the ELISA. Further study demonstrated that the estimated Epo level in bioassay was strongly influenced by other cytokines present in the samples. Therefore, the Epo bioassay detects the net erythropoiesis promoting activities, whereas the ELISA method specifically measures the level of Epo in the samples. Combined with the Epo bioassay, the murine Epo ELISA will be an extremely useful tool in specifically measuring the Epo response and facilitating the understanding of mechanisms involved in the development of anemia-associated diseases using mouse models. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Murine Epo ELISA; Epo bioassay; Malaria; Anemia; Serum factors
1. Introduction Epo is the essential growth factor that promotes the growth, proliferation, and differentiation of erythroid
Abbreviations: Epo, erythropoietin; rmEpo, recombinant murine Epo; hEpo, human Epo. * Corresponding author. Centre for the Study of Host Resistance, Montreal General Hospital Research Institute, 1650 Cedar Avenue, Montreal, Quebec, Canada H3G 1A4. Tel.: +1-514-934-1934x 44508; fax: +1-514-934-8332. E-mail address: [email protected] (M.M. Stevenson).
lineage cells and leads to the increase in circulating red blood cell volume. Epo is mainly produced in adult kidneys and is upregulated in response to tissue hypoxia to correct anemia. Furthermore, Epo is also produced in the brain and can prevent loss of stressed neurons in hypoxia – ischaemia. Epo administration has been proven to be an effective remedy for treating anemia in AIDS (Moore et al., 1998), in renal failure (Tsakiris, 2000), and in chronic inflammatory diseases (Tanaka et al., 1999), as well as for limiting brain damage in experimental autoimmune encephalomyelitis (Brines et al., 2000). Because murine models have been extensively used for studying various anemia-
0022-1759/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 2 2 - 1 7 5 9 ( 0 2 ) 0 0 0 1 2 - 1
130
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
associated disorders as well as for testing drug treatment strategies, it would be extremely helpful to be able to specifically measure murine Epo (mEpo) in serum and in other tissues, such as the kidney and brain. Epo bioassay has been widely used for measuring both human Epo (hEpo) and mEpo in sera and tissue culture media until the development of hEpo ELISA, which is less time-consuming, more convenient, and does not use radioisotopes. Unfortunately, the hEpo ELISA is not sensitive enough to accurately measure basal mEpo levels in mouse serum (Rinaudo and Toniatti, 2000). Recently, ELISAs specific for mEpo were described by two research groups using either self-raised polyclonal antibodies (Noe et al., 1999) or commercially purchased human Epo ELISA kits (Rinaudo and Toniatti, 2000). Here, we describe a mEpo ELISA, which is equally sensitive, more costeffective, and with a larger reliable range, using commercially available anti-Epo antibodies. We further compared the ELISA method and the traditional Epo bioassay in the study of anemia induced during bloodstage Plasmodium chabaudi AS (P. chabaudi AS) malaria infection. The effects of other serum factors on the accuracy of these two methods were also studied.
2. Type of research Quantifying murine erythropoietin in serum.
3. Time required Coating plates with antibodies at 4 C: overnight Wash plates: 3 min/plate (repeated between each step) Blocking plates with blocking buffer at room temperature: 1 h Preparing serum samples and mEpo standard serial dilutions and adding to the plate: 10 min/plate Samples and standards binding at 4 C: overnight Adding detecting antibody and incubating at room temperature: 1 h Adding anti-rabbit IgG-HRPO conjugate and incubating at room temperature: 1 h Adding substrate and color development at room temperature: 30 min.
4. Materials and methods 4.1. Animals and parasites P. chabaudi AS parasites were maintained as previously described (Yap and Stevenson, 1992). Eightweek-old male C57BL/6 mice were purchased from Charles River (St. Constant, QC, Canada). All animal care and handling were carried out in accordance with institutional guidelines of the Montreal General Hospital Research Institute and McGill University. Mice were infected i.p. with 106 parasitized red blood cells (PRBC). Blood was collected by cardiac puncture at different time points following infection, allowed to clot at room temperature, and centrifuged. Serum was collected and stored at 20 C. 4.2. Murine EPO ELISA optimization Monoclonal rat anti-mEpo antibody IgG1 (1 mg/ ml, BD BioSciences, Mississauga, ON, Canada) and polyclonal rabbit anti-hEpo antibody (2 mg/ml, R&D Systems, Minneapolis, MN, USA) were tested as either capturing or detecting antibody. For coating, antibodies were diluted 1:250, 1:500, or 1:1000 in PBS. For detecting, antibodies were diluted 1:200, 1:400, or 1:1000 in blocking buffer, which consisted of PBS and 1% BSA (Gibco Laboratories, Invitrogen, Burlington, ON, Canada). HRPO-conjugated goat anti-rabbit IgG(H + L) (1:3000 dilution, Bio-Rad Laboratories, Hercules, CA, USA) or goat anti-rat IgG (1:1000 dilution, Bio-Rad Laboratories) antibodies diluted in blocking buffer was applied as secondary antibody depending on the primary detecting antibody chosen. All the procedures were performed in the wells of Immulon 2HB U Bottom Microtiter Plates (Dynex Technologies, Chantilly, VA, USA). Following each incubation step, plates were washed three times with 200 ml/well washing buffer, which consisted of 0.1% Tween 20 (Sigma, St. Louis, MO, USA) in PBS. Plates were first coated with 50 ml/well of appropriately diluted capturing antibodies overnight at 4 C and subsequently blocked with 250 ml of blocking buffer for 1 h at room temperature. Recombinant mEpo (rmEpo) standard was serially diluted (0.98 – 500 mU/ml; Boehringer Mannheim-Roche, Laval, QC, Canada) in blocking buffer, and 50 ml/well of
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
each dilution was added to the plates and incubated at either 4 C overnight or at room temperature for 2 h. Furthermore, 50 ml/well of detecting antibodies were added and incubated for 1 h at room temperature, followed by incubating with 50 ml/well of HRPO conjugates for 1 h at room temperature. The bound conjugate was visualized by adding 100 ml/well of ABTS substrate (Boehringer Mannheim-Roche) with 0.036% H2O2. OD values were read at various time points (15 min to 1 h) after addition of substrate in a microplate reader at a wavelength of 405 nm with 492 nm as reference. Once the optimal conditions were determined, changes in final OD values, in the presence of cytokines added to known concentrations of rmEpo standard (0, 1.95, 15.625, and 125 mU/ml) in the step of sample/standard incubation, were determined to evaluate the specificity of this ELISA. Cytokines tested were TNF-a (0.1 –10 ng/ml, R&D systems), IFN-g (0.5 –50 ng/ml, R&D systems), IL-12 (2.8 –100 ng/ml, a gift from Genetics Institute, Cambridge, MA, USA), IL-3 (0.125 – 12.5 ng/ml, Cedarlane Laboratories, Hornby, ON, Canada), and stem cell factor (SCF) (2– 50 ng/ml, Cedarlane Laboratories).
131
4.4. Epo bioassay
4.3. Detecting EPO in murine serum using newly developed murine EPO specific sandwich ELISA
Serum Epo concentration was also measured by the Epo bioassay as previously described (Yap and Stevenson, 1992). Briefly, Iscove’s MDM (IMDM) supplemented with 10% heat-inactivated FCS (HyClone, Logan, UT, USA), 5 10 5 M 2-mercaptoethanol (Sigma), and 48 mg/l Gentamicine (Schering Canada, Pointe-Claire, QC, Canada) was used as culture medium throughout the assay. C57BL/6 mice were rendered anemic by two consecutive daily i.p. injections of phenylhydrazine hydrochloride (60 mg/kg, Fischer Scientific, Fair Lawn, NJ, USA). Spleens were harvested 3 days after the second injection and gently pressed through fine wire meshes and passed through 18-G needles. After washing three times with sterile PBS, the spleen cells were resuspended in culture medium and adjusted to 4 106 cells/ml to be used as responder cells. Aliquots of 100 ml/well of mouse sera diluted four-fold in culture medium as well as serially diluted rmEpo (0.98 – 500 mU/ml) were added into 96-well culture plates (Becton Dickinson Labware, Franklin Lakes, NJ, USA) followed by the addition of 100 ml/well responder cells. After 22 h of incubation at 37 C, 2 mCi 3H-thymidine (specific activity 6.7 Ci/mol, ICN Biomedicals, Irvine, CA, USA) were added to each well and incubated for
Aliquots of 50 ml/well of monoclonal rat anti-mEpo antibody (4 mg/ml in PBS) were added into plates and incubated at 4 C overnight. Following washing and blocking, 50 ml/well aliquots of 0.98 – 500 mU/ml rmEpo standard and 1:3 serial dilutions of serum samples in blocking buffer were added and incubated at 4 C overnight. Aliquots of 50 ml/well of polyclonal rabbit anti-hEpo antibody (10 mg/ml in blocking buffer) were then added and incubated at room temperature for 1 h, followed by incubation with 50 ml/well of HRPO-conjugated goat anti-rabbit IgG(H + L) antibody (1:3000 in blocking buffer) at room temperature for 1 h. After each step, plates were washed three times with washing buffer. The bound conjugate was then detected by adding 0.1 ml/well ABTS substrate with 0.036% H2O2 and developed for 30 min and read at a wavelength of 405 nm with 492 nm as reference. The Epo concentration in serum samples was calculated against the standard curve generated with rmEpo.
Fig. 1. Optimization of the mEpo-specific ELISA. Representative standard curves are shown after color development and OD reading. Plates were coated with monoclonal rat anti-mEpo antibody and the captured mEpo was detected with polyclonal rabbit anti-hEpo antibody at concentrations of 2, 5, or 10 mg/ml.
132
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
2 h. Cells were harvested with an automatic cell harvester and 3H-thymidine incorporation, which reflected the degree of spleen cell proliferation, was determined by scintillation counting. A standard curve was generated by plotting 3H-thymidine incorporation against the concentration of serially diluted rmEpo. Epo concentration in the samples was estimated by calculating the 3H-thymidine incorporation against the standard curve. To determine the specificity of the Epo bioassay, 100 ml/well of responder cells were added to 100 ml/ well of culture medium containing known concentrations of rmEpo standard (0, 1.95, 15.625, or 125 mU/ ml) in the presence or absence of TNF-a (0.1 – 10 ng/ ml), IFN-g (0.5 – 50 ng/ml), IL-12 (2.8 –100 ng/ml), IL-3 (0.125 – 12.5 ng/ml), or stem cell factor (SCF) (2– 50 ng/ml). The amounts of 3H-thymidine incor-
poration were compared between cultures with and without addition of exogenous cytokines. The specificity of the Epo bioassay was also tested by the addition of an antibody cocktail that contained anti-TNF-a (102 Neutralizing U/ml, clone XT22 ascites), anti-IFN-g (1.67 mg/ml, clone XMG1.2), and anti-IL-12 (3.9 mg/ml, clone C17.8) into serum samples to block the effect of these cytokines. The changes in 3H-thymidine incorporation of the responder cells were compared to those incubated with serum samples alone. 4.5. Statistical analyses Values are expressed as mean F standard error of the mean (S.E.M.). Statistical significance of differences was analyzed by Student’s t-test or ANOVA
Fig. 2. Course of infection and changes in hematologic parameters during P. chabaudi AS infection in C57/BL6 mice. (A) Course of parasitemia, (B) degree of anemia as determined by hematocrit level, (C) serum Epo concentration estimated by ELISA, and (D) serum Epo concentration estimated by Epo bioassay. The results are expressed as mean F S.E.M. of three to four mice. Similar results were obtained in replicate experiments.
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
followed by Tukey’s studentized range test. A p < 0.05 was considered to be significant. Serum Epo values were converted to their log values before performing regression analyses. The parallel lines assumption was tested to determine the statistical significance of differences between the slopes of regression lines of the ELISA and bioassay in our model. Statistical analyses were performed using SAS-STAT software (SAS Institute, Cary, NC, USA).
5. Results 5.1. Murine Epo ELISA optimization In preliminary studies, to establish the optimal conditions for the ELISA, we found that the use of polyclonal rabbit anti-hEpo antibody as capturing antibody yielded high background and low sensitivity regardless of the dilution chosen (data not shown). When monoclonal rat anti-mEpo was used as the capturing antibody, the reliable range of the standard curve increased when lower dilutions of both capturing and detecting antibodies were used. We also found that incubation of rmEpo and serum sample at 4 C overnight yielded a lower background and higher sensitivity than short time incubation at room temperature. The optimal conditions of this Epo ELISA were determined to be the use of 4 mg/ml of monoclonal rat anti-mEpo antibody as the capturing antibody and 10 mg/ml of polyclonal rabbit anti-hEpo antibody as the detecting antibody with overnight incubation of test samples and standards at 4 C. Under these conditions, the ELISA can detect mEpo as low as 1 mU/ml and the reliable range extends to a high of 300 mU/ml (Fig. 1). When this ELISA was tested for its specificity, addition of a wide concentration range of TNF-a, IFN-g, IL-12, IL-3, and SCF to known concentrations of rmEpo standard at the standard/sample incubation step did not alter the OD values compared to incubation with rmEpo standard alone (data not shown).
133
parasitemia then gradually declined and became almost undetectable by day 11 post infection (p.i.). Anemia, as measured by hematocrit level, developed progressively as the parasite load increased (Fig. 2B) and became most severe at day 8 to day 9 p.i. We measured the serum Epo concentrations by both the sandwich ELISA described here (Fig. 2C) and by the bioassay (Fig. 2D). Both assays showed that Epo production reached its highest level between day 8 and day 9 p.i. when the mice were most anemic. Regression analysis of serum Epo levels against hematocrit levels revealed an inverse correlation in both the ELISA and bioassay ( p < 0.05) with high R2 values (Fig. 3). The parallel lines assumption was tested and the slopes were not significantly different from one another but the estimated Epo levels were between 5- and 20-fold higher in bioassay than in ELISA ( p < 0.05). 5.3. Effects of other serum factors in the accuracy of bioassay The specificity of the Epo bioassay was tested by incubating known concentrations of rmEpo in the presence or absence of additional serum factors including TNF-a, IFN-g, IL-12, IL-3, and SCF. For all
5.2. Epo production during blood-stage P. chabaudi AS infection Following P. chabaudi AS infection, C57BL/6 mice developed increased parasitemia which reached peak level at day 7 after infection (Fig. 2A). The
Fig. 3. Inverse correlation between hematocrit level and serum Epo concentration measured by murine-specific Epo ELISA and by bioassay. Twenty-five samples were measured in each group. p < 0.05 for both regression lines.
134
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
concentrations tested, TNF-a (0.1 – 10 ng/ml) and IFN-g (0.5 – 50 ng/ml) significantly suppressed the proliferation of responder cells in the absence of rmEpo or with 1.95 – 125 mU/ml rmEpo (Fig. 4 and data not shown). However, IL-12 (2.8 – 100 ng/ml) and SCF (2 –50 ng/ml) stimulated the proliferation of responder cells in the absence of rmEpo or with 1.953 mU/ml rmEpo but inhibited the proliferation induced by higher amounts of rmEpo (15.625 – 125 mU/ml). Furthermore, IL-3 at higher concentrations (1.25 – 12.5 ng/ml) showed significant stimulatory effects on cultures in the absence of rmEpo or with 1.953 – 15.625 mU/ml rmEpo. However, at a lower concentration (0.125 ng/ml), IL-3 stimulated the proliferation in the absence of rmEpo or with 1.953 mU/ml rmEpo but inhibited the proliferation induced by a higher dose of rmEpo (125 mU/ml). Based on the previously published serum cytokine data in malaria-infected mice or patients, we summarized the effects of 1 ng/ ml TNF-a, 50 ng/ml IFN-g (Su and Stevenson, 2000), 2.8 ng/ml IL-12 (Sam and Stevenson, 1999; Su and Stevenson, 2000), 0.125 ng/ml IL-3, and 2 ng/ml SCF (Burgmann et al, 1997) on the Epo bioassay in Fig. 4. Furthermore, when a mixture of anti-IL-12, anti-TNFa, and anti-IFN-g antibodies was added to sera from
Fig. 4. Effects of cytokines on the proliferation of bioassay responder cells. Spleen cells from phenylhydrazine-treated mice were incubated with four different concentrations of rmEpo in the presence or absence of exogenous cytokines. Each bar represents the amount of 3 H-thymidine incorporated and is expressed as mean F S.E.M. of four samples. Similar results were obtained in replicate experiments. * p < 0.05 compared with no addition of exogenous cytokine.
infected mice to neutralize the activity of these cytokines, the 3H-thymidine incorporation in the Epo bioassay increased significantly in comparison with responder cells cultured with sera alone (data not shown).
6. Discussion Epo has been shown to be important in correcting anemia as well as protecting neuron cells in hypoxia– ischaemia. To facilitate research in both hematology as well as neurology, a specific and accurate method of measuring Epo concentration in serum and tissues is required. For human Epo, ELISA kits, which provide a quick and convenient way of measuring Epo, are commercially available. Unfortunately, although human and murine Epo molecules share 80% homology, hEpo ELISA is not sensitive enough for detecting murine Epo (Rinaudo and Toniatti, 2000). Recently, an ELISA specific for mEpo with a reliable range of 0.6 –30 mU/ml was described by Noe et al. (1999), who combined self-raised polyclonal anti-hEpo and anti-mEpo antibodies. This was followed by a mEpo ELISA developed by Rinaudo and Toniatti (2000), who used two monoclonal anti-hEpo antibodies from commercially purchased hEpo ELISA kits and achieved a sensitivity of 1 mU/ml. Here, we report an mEpo-specific ELISA, which is equally sensitive (1 mU/ml) and yet provides a wider reliable range (up to 300 mU/ml) using commercially available antibodies. As serum Epo levels can be elevated up to 1000 mU/ml (Fig. 2C and Noe et al., 1999), an ELISA with the ability to detect a wider reliable range eliminates the need of further diluting samples and thus avoids the error introduced by high sample dilutions. Prior to the development of Epo ELISAs, Krystal (1983) established the Epo bioassay, which was considered to be more specific for Epo than in vitro 59Fe bioassay methods. This bioassay has been widely used for detecting Epo levels in serum and culture medium. When we compared the Epo levels in sera of normal and P. chabaudi AS-infected mice detected by the newly established ELISA and by the bioassay, an inverse correlation between the serum Epo concentrations and hematocrit levels during P. chabaudi AS infection was observed regardless of the assay used.
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
While these two methods provided comparable slopes ( 0.0461 and 0.0621, respectively) and R2 values (0.8556 and 0.9063, respectively), serum Epo levels estimated by the Epo bioassay were significantly higher, which was consistent with the estimated intercepts of both regression lines (4.4342 vs. 3.9078, respectively). Since the bioassay estimates Epo levels in samples based on the amount of 3H-thymidine incorporated by responder cells, the effects of other blood elements besides Epo on the proliferation of responder cells might be partially responsible for the difference observed. To test this hypothesis, the responder cells from spleens of phenylhydrazine-treated mice were incubated with known concentrations of rmEpo supplemented with several cytokines known to be present in the sera of malaria-infected patients and experimental animals (Burgmann et al., 1997; Fried et al., 1998; Othoro et al., 1999; Sam and Stevenson, 1999; Su and Stevenson, 2000). Consistent with their inhibitory role in erythropoiesis (Allen et al., 1999), TNF-a and IFNg suppressed the Epo-stimulated spleen cell proliferation at all concentrations tested. Surprisingly, IL-12, which was shown to enhance murine erythropoiesis in vitro in combination with IL-4 and SCF (Dybedal et al., 1995), exerted stimulatory effects when cultured with lower doses of rmEpo but exerted negative effects when cultured with higher doses of rmEpo. This phenomenon was also observed at all concentrations of SCF and the lowest dose of IL-3 tested. Therefore, when measuring lower doses of Epo in the presence of IL-12, SCF, or IL-3 with the bioassay, the stimulatory effects of these cytokines increase 3Hthymidine incorporation and lead to an overestimation of Epo concentrations. In contrast, when measuring higher doses of Epo in the presence of IL-12, SCF, TNF-a, and IFN-g, the inhibitory effects of these cytokines resulted in an underestimation of Epo concentrations, which is supported by the significantly increased serum bioactivity when TNF-a, IFN-g, and IL-12 were neutralized. The 5 –20-fold higher levels of Epo detected by the bioassay in comparison to those detected by the ELISA at all Epo concentrations were possibly due to the presence of other factors not included in this study, such as IL-10, IL-4, GM-CSF, TGF-b, and other unknown serum factors. These serum factors may synergistically promote erythropoiesis with high levels of Epo and cause an over-
135
estimation of overall Epo concentrations in the bioassay. Taken together, our data suggest that the cell proliferation in the bioassay, which estimates the Epo concentrations in samples, is greatly influenced by several factors including the presence of other hematopoietic cytokines in the samples, the concentrations of these cytokines, and the interaction of Epo and these cytokines. To estimate Epo concentrations accurately using the bioassay, it is essential to either remove all these cytokines from the samples, or extensively characterize and exclude the impact of all serum factors in each sample on the proliferation of responder cells. Without these modifications, data obtained from the Epo bioassay represent the net erythropoiesis promoting activity of the sample, which is the balance of the inhibitory and stimulatory elements. In contrast, the Epo ELISA, which is not influenced by the presence of other serum factors in the samples, provides a more reliable estimation of Epo concentrations. In conclusion, our newly developed murine Epo ELISA provides a convenient, sensitive, cost-effective, and highly reliable way to measure murine Epo levels in serum and other organs. In contrast, data generated by the Epo bioassay should be interpreted as net erythropoiesis promoting/inhibitory activities rather than the level of Epo specifically. A combination of these two assays will facilitate the research in malaria and other diseases using murine models.
7. Essential references Krystal, 1983. Noe et al., 1999. Tsakiris, 2000.
Acknowledgements The authors wish to thank Fabrice Rouach for statistical consultation. This research was supported by the Medical Research Council of Canada (Canadian Institutes of Health Research) grant number MT14663. K-H Chang was a recipient of the Lloyd Carr-Harris McGill Major Fellowship from McGill University, Montreal, Quebec, Canada.
136
K.-H. Chang, M.M. Stevenson / Journal of Immunological Methods 262 (2002) 129–136
References Allen, D.A., Breen, C., Yaqoob, M.M., Macdougall, I.C., 1999. Inhibition of CFU-E colony formation in uremic patients with inflammatory disease: role of IFN-gamma and TNF-alpha. J. Investig. Med. 47, 204. Brines, M.L., Ghezzi, P., Keenan, S., Agnello, D., de Lanerolle, N.C., Cerami, C., Itri, L.M., Cerami, A., 2000. Erythropoietin crosses the blood – brain barrier to protect experimental brain injury. PNAS 97, 10526. Burgmann, H., Looareesuwan, S., Wiesinger, E.C., Winter, W., Graninger, W., 1997. Levels of stem cell factor and interleukin-3 in serum in acute Plasmodium falciparum malaria. Clin. Diagn. Lab. Immunol. 4, 226. Dybedal, I., Larsen, S., Jacobsen, S.E., 1995. IL-12 directly enhances in vitro murine erythropoiesis in combination with IL-4 and stem cell factor. J. Immunol. 154, 4950. Fried, M., Muga, R.O., Misore, A.O., Duffy, P.E., 1998. Malaria elicits type 1 cytokines in the human placenta: IFN-gamma and TNF-alpha associated with pregnancy outcomes. J. Immunol. 160, 2523. Krystal, G., 1983. A simple microassay for erythropoietin based on 3 H-thymidine incorporation into spleen cells from phenylhydrazine treated mice. Exp. Hematol. 11, 649. Moore, R.D., Keruly, J.C., Chaisson, R.E., 1998. Anemia and survival in HIV infection. J. Acquired Immune Defic. Syndr. 19, 29.
Noe, G., Riedel, W., Kubanek, B., Rich, I.N., 1999. An ELISA specific for murine erythropoietin. Br. J. Haematol. 104, 838. Othoro, C., Lal, A.A., Nahlen, B., Koech, D., Orago, A.S., Udhayakumar, V., 1999. A low interleukin-10 tumor necrosis factoralpha ratio is associated with malaria anemia in children residing in a holoendemic malaria region in western Kenya. J. Infect. Dis. 179, 279. Rinaudo, D., Toniatti, C., 2000. Sensitive ELISA for mouse erythropoietin. BioTechniques 29, 218. Sam, H., Stevenson, M.M., 1999. In vivo IL-12 production and IL12 receptors beta1 and beta2 mRNA expression in the spleen are differentially up-regulated in resistant B6 and susceptible A/J mice during early blood-stage Plasmodium chabaudi AS malaria. J. Immunol. 162, 1582. Su, Z., Stevenson, M.M., 2000. Central role of endogenous gamma interferon in protective immunity against blood-stage Plasmodium chabaudi AS infection. Infect. Immun. 38, 4399. Tanaka, N., Ito, K., Ishii, S., Yamazaki, I., 1999. Autologous blood transfusion with recombinant erythropoietin treatment in anaemic patients with rheumatoid arthritis. Clin. Rheumatol. 18, 293. Tsakiris, D., 2000. Morbidity and mortality reduction associated with the use of erythropoietin [Review]. Nephron 85 (Suppl. 1), 2. Yap, G.S., Stevenson, M.M., 1992. Plasmodium chabaudi AS: erythropoietic responses during infection in resistant and susceptible mice. Exp. Parasitol. 75, 340.