Original Paper Acta Haematol 1992;88:175-184
a Second Department of Physiology, Nara Medical University, Nara; b Research Institute of Life Science, Snow Brand Milk Products, Tochigi, Japan
Key Words CFU-E Erythropoietin Fetal mouse liver Monoclonal antibody Mouse plasma Stimulating activity
Neutralization and Immunoaffinity Chromatography of Erythroid Colony-Stimulating Activity in Mouse Plasma by an AntiErythropoietin Monoclonal Antibody
Abstract A relationship between erythropoietin (EPO) and erythroid colony-stimulat ing activity (ECSA) in mouse plasma was examined in fetal mouse liver cell (FMLC) cultures using a monoclonal antibody (MoAb) R2 raised against re combinant human EPO. Most of the ECSA in plasma from normal, anemic, and hypoxic mice was neutralized by MoAb. This neutralization could be re versed by addition of excess of anemic plasma or by preincubation of MoAb with goat anti-mouse IgG antibody. Most of the plasma ECSA was bound to an immunoadsorbent column containing the immobilized MoAb, and the re tained ECSA was completely neutralized by MoAb. The plasma ECSA and standard EPO showed parallel dose-response curves and additive effect on CFU-E stimulation. Based on these findings, we conclude that mouse plasma ECSA detected by CFU-E assay using FMLCs is mainly due to EPO.
Introduction Erythropoietin (EPO) is a glycoprotein hormone that regulates red blood cell production by promoting the pro liferation and differentiation of late erythroid progenitor cells (CFU-E). One of the most useful in vitro bioassays for EPO is CFU-E colony-forming assay, which has the ad vantage of directly measuring EPO activity in stimulating CFU-E growth [1], This assay procedure was used for mea surement of plasma EPO levels in mice [2,3] and also for detection of EPO activity in chromatography of mouse plasma or serum [4,5]. To validate the assay, however, evi-
Received: February 17,1992 Accepted: June 15,1992
dence that erythroid colony-stimulating activity (ECSA) in mouse plasma is ascribable to EPO should be provided, because a wide variety of substances other than EPO, i.e. growth hormones [6], thyroid hormones [7], steroids [8], insulin-like growth factor (IGF) I [9] and II [10], insulin [11], cyclic nucleotides [12],hcmin [13], activin A [14],EPOlike activity (EpLA) [15], erythropoietic stimulating cofac tor (ESCF) [16], and proerythroblast stimulating activity (PSA) [17], have been reported to be potent stimulating agents for CFU-E colony growth. Halvorsen’s group re ported a significant correlation between plasma ECSA levels and EPO levels measured by radioimmunoassay in
Dr. Susumu Sakata Second Department of Physiology Nara Medical University S40 Shijo-cho. Kashihara Nara 634 (Japan)
© 1992 S. Kargcr AG, Basel (XXH-5792/92/0884-0175 $ 2.75/0
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Susumu Sakataa Yasunori Enokia Masatsugu Uedah
Materials and Methods Mice and Plasma Ten-week-old male ICR mice were used throughout the study. Posthemorrhagic anemia was induced by a single or double bleeding of an amount of blood equivalent to 2.0% of the body weight from the tail vein. Twenty-four hours later, heparinized blood was obtained by puncture of the vena cava under ether anesthesia. Hypoxia was pro duced by rearing the animals in a Plexiglas hypobaric chamber con nected with a vacuum pump and keeping them at 0.35 atm for 3 or 4 days (22.5 h/day). Immediately after hypoxia exposure, blood was ob tained as described above. After centrifugation, the plasma samples were sterilized through 0.45-um filters (Millipore, Bedford, Mass., USA), and stored at -9 0 °C until use. CFU-E Assays Assays for CFU-E colony growth were performed in mcthylcellulose microwell cultures. Briefly, liver cells (3-4 x 10J) from the day-12 fetal ICR mice, suspended in 100 pi of a-medium (Flow Labs., McLean, Va., USA) containing 0.8% mcthylccllulose (Shin-ctsu Chemicals, Osaka, Japan), 30% heat-inactivated fetal calf serum (FCS; Flow Labs.), and Ulr'M a-thioglycerol (Nacalai Tesque, Kyoto, Japan) [23], were plated into each well of a 96-well microtiter plate (Corning Glass Works, Corning, N.Y.. USA). Then. 10 ul of standard EPO solution and/or plasma samples were added to each culture. Sheep plasma Step III EPO (3.3 IU/mg; Connaught, Willowdaie, Canada) and/or rHuEPO expressed in baby hamster kidney cells [241 were used as a standard EPO. Cultures were carried out at 37 °C in a fully humidified atmosphere of 5% CO, in air, and crythroid colonies containing eight or more cells were directly scored using an inverted microscope after a 48-hour incubation. In some experiments, the col ony counts were carried out after the culture wells were fixed in situ by flooding with five drops of 0.4% glutaraldchyde in a-medium. All cultures were performed in triplicate or quadruplicate.
176
Sakata/E noki/U cda
Neutralization Studies Neutralization of the biological ECSA in normal, hypoxic, and anemic plasmas was conducted by use of an anti-EPO MoAb R2 which had been shown to neutralize the biological activity of murine EPO [22]. In most experiments, the purified MoAb dissolved in a-medium with 2% FCS and the plasma were added directly to the CFU-E cultures without prior incubation. In some experiments, the plasma was prcincubated with MoAb for 2 h at 37 °C, and subse quently at 4 °C overnight. Then, goat antimousc IgG polyclonal anti body (anti-M-IgG; Tago, Burlingame, Calif., USA) was added to this mixture for further preincubation. The immunocompicxes of anti-MIgG and MoAb were removed from the plasma mixture by refriger ated centrifugation (18,500g, 10 min). The ECSA remaining in the su pernatant was determined in the CFU-E assay. Furthermore, to ex amine whether the neutralization by anti-EPO MoAb is due to spe cific antigen-antibody interactions, MoAb was first prcincubated with an excessive amount of anti-M-IgG as described above. After centrifugation of this mixture, the supernatant was once again prein cubated with mouse plasma. The resulting immune precipitates were removed hy centrifugation, and the supernatant was added to the cul tures for assay of ECSA. In these experiments, MoAb against mouse mammary tumor virus (MMTV) and anti-M-IgG were employed as a control antibody. Irnmunoaffinity Chromatography Mouse plasma sample (1 ml) was loaded at room temperature on an immunoadsorbent column (dimensions 1.2 x 1.8 cm; bed volume 2 ml) containing MoAb R2 fixed on Affi-Gel 10 (Bio-Rad, Richmond. Calif., USA) equilibrated with 10 mM phosphate-buffered saline (PBS), pH 7.4 [22]. Following application of the sample, the column flow was stopped and the column was left overnight at 4°C. The col umn was extensively washed at room temperature with 100 ml of PBS, 20 ml of lOmA/NaPi, pH 7.4, containing0.5 M NaCI, and finally 20 ml of 0.15 M NaCI, in this order, and then eluted by 20 ml of 0.2M acetic acid, pH 2.5, containing 0.15 M NaCI. Fractions of 1 ml were collected. Each fraction thus eluted was immediately neutralized by adding 50 ul of 3.2 M Tris, and was dialyzed against several changes of Dulbccco’s PBS. pH 7.4, and then a-medium. The dialyzed fractions, steril ized through a 0.45-um filter, were tested for ECSA. The ECSA of each fraction was expressed as international units of EPO activity, which was calculated by reference to the dose-response curve for standard Step III EPO. In addition, neutralization of each fraction ECSA was studied using the MoAb R2. Furthermore, neutralization of the ECSA of pooled fractions was attempted by anti-EPO MoAb R6 [22] and rabbit anti-rHuEPO serum. Both the flow-through and the bound fractions were collected separately. These pooled frac tions were dialyzed against dcionized-distilled water, and then lyophilized. The dry fractions were dissolved in 2 ml of a-medium with 2% FCS, and sterilized through a 0.45-um filter. Normal rabbit serum (NRS), heat-inactivated at 56 °C for 30 min. was used as control for irrelevant specificity. At the end of elution, the column was washed and preserved in PBS containing 0.02% sodium azide. Statistics Data arc presented as the mean ±SD . Results were analyzed us ing the Student’s t test.
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
both adult and neonatal mice [18], but main focus of their investigations was to answer a question of whether high plasma ECSA found in neonatal mice is entirely due to EPO [18-21]. Thus, the relationship between the ECSA and EPO has not been fully studied in adult mice yet. Recently, Goto et al. [22] developed a monoclonal anti body (MoAb) R2 directed toward recombinant human EPO (rHuEPO), which was proved to efficiently inhibit the stimulation by anemic mouse plasma of [3H]-thymidine incorporation into DNA in cultured fetal mouse liver cells (FMLCs). To clarify the relationship between the ECSA and EPO in adult mouse plasma, using this highly specific MoAb, we have attempted neutralization and immunoaffinity chromatography of the plasma ECSA of normal, anemic, and hypoxic adult mice. In addition, the mode of action of the ECSA was compared with that of standard EPO in CFU-E colony assay using FMLCs. We present here evidence that the plasma ECSA found in adult mice is predominantly due to EPO.
Results Properties o f ECS A in Mouse Plasma
Neutralization o f Plasma ECSA by Anti-EPO MoAb R2
Anti-EPO MoAb R2 inhibited dose-dependently the formation of CFU-E colonies stimulated by standard EPO or normal or anemic mouse plasma when added simulta neously to the cultures without prior incubation (fig. 2). MoAb dose-response curves for standard EPO and mouse plasmas, expressed as percent inhibition, were similar to one another at antibody doses less than 219 ng/well. Fifty
OOI 0 0 2 0 0 4 0 0 7 014 Q28 0 5 7
Plasma
1.1
2.3
4.5
9.1
16.7
cone. (%)
Fig.I.C omparison of dosc-rcsponse curves for standard EPO (O) and plasma from normal (A ) or hypoxic (A ) mice. Liver cells from the 12th day fetal mice (4 x l()7well) were used. The standard EPO was sheep plasma Step III. Each symbol represents the triplicate mean ± SD.
percent inhibition was obtained at antibody dose of 3.4 ng/ well. At antibody doses more than 219 ng/well, most of the plasma ECSA was neutralized. For mouse plasmas and sheep plasma Step III EPO, the mean percent inhibition even at such high doses of MoAb was not complete, but about 90 and 85%, respectively. For rHuEPO, on the other hand, the complete suppression by MoAb was observed. Similarly, anti-EPO MoAb R2 was capable of neutra lizing the ECSA of fetal mouse plasma [data not shown], suggesting that the direct addition of MoAb to the cul tures neutralizes fetal EPO produced endogenously by macrophages or Kupffer cells in FMLC cultures [25]. Therefore, such direct-additive experiments do not ex clude the possibility that the ECSA found in plasma is not due to EPO but due to an inducer of endogenous EPO production in the cultures. Accordingly, action of MoAb on the neutralization of ECSA in plasma was examined in detail. Mouse plasma was prcincubated with MoAb before addition to the cultures, and then an excess of MoAb was removed from plasma by immunoprécipitation with antiM-IgG. This pretreatment with MoAb almost completely abrogated the ECSA in normal, anemic, and hypoxic plas mas [data not shown]. This result strongly suggests that the ECSA in mouse plasma is EPO but not the inducer of en dogenous EPO. These neutralizing actions of MoAb against ECSA in anemic plasma or rHuEPO were completely overcome by
177
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
At concentrations less than 4.5%, normal mouse plasma stimulated dose-dependently the formation of CFU-E colonies from FMLC without exogenous EPO (fig. 1). Plasma from anemic or hypoxic mice had a similar but stronger stimulatory effect. The dose-response curves for these plasmas were evidently parallel to the curve for standard EPO. At the higher concentrations (9.1-16.7%), however, the number of the colonies decreased in a dosedependent fashion. From this result, mouse plasmas ap pear to contain small amounts of an inhibitor for CFU-E growth. To assess the inhibitory potency at the concentra tions less than 4.5%, an inhibitor assay was performed. Plasma samples were added to the cultures supplemented with an EPO dose (25 mU/wcll) enough to induce the maximum colony formation. Such concentrations of plasma caused no reduction in the colony formation [data not shown]. This fact, as well as parallelism in the dose-response curves, indicates that the inhibitor in mouse plasmas, if present, docs not affect the formation of CFU-E colonies at the additive plasma concentration less than 4.5%. To determine whether the stimulatory effect of mouse plasma on the colony formation is additive or synergistic with standard EPO, the EPO dose-response curves in the absence and presence of plasma were compared. Increas ing doses of EPO were added to the cultures with a fixed amount (0.26%) of normal or anemic plasma. In both the EPO dose-response curves, the plateau levels in the colo ny formation were the same [data not shown]. Thus, the ECSA in mouse plasmas were not synergistic but additive to standard EPO in the present CFU-E assay system. In addition, the ECSA of both standard EPO and mouse plasma remained biologically active even after heating at 56 °C for 30 min or dialysis against distilled water for 3 days (Viskase cellulose tubing; 13,000 MW cut-off). Thus, some biological and biochemical proper ties of mouse plasma ECSA were similar to those of stan dard EPO.
'A
100
B
Ç . . -O -.
.'A' ' Z
'
A
/
> 12.5 25 50 100 2 0 0 4 0 0 0 2 6 052 10 2.1 42 8.3 E P O (mU/well) Plasm a cone. (%)
Fig. 2. Dose-dependent inhibition of CFU-E colony formation by anti-EPO MoAb R2. Different doses of MoAb were added directly to the cultures supplemented with a sufficient amount (25 mU/well) of standard EPO, sheep plasma Step III (O )or rHuEPO ( A ), or 4.2% of plasma from normal (□; mean PCV, 0.45 1/1) or anemic (V ; mean PCV, 0.25 1/1) mice. Liver cells from the 12th day fetal mice (3 x 10V well) were used. The degree of the inhibition was calculated as:
Nmax-N,
% inhibition = _
_
x 100.
“ ^ m in
where Nnm, Nmin and Nxare the number of the colonies formed in the cultures supplemented with 25 mU of standard EPO or 4.2% of mouse plasma, only a-medium, and both the different doses of MoAb and 25 mU of standard EPO or 4.2% of mouse plasma, respectively. Nmax and Nmin arc the mean values. In this experiment, Nmax for the standard EPOs, normal and anemic plasmas were 263, 221, and 254, respectively. Nmi„ was 122. Each symbol represents the triplicate mean.
the addition of higher amounts of the plasma or rHuEPO (fig. 3). They were also abolished by preincubating MoAb with anti-M-IgG (table 1). Anti-MMTV MoAb or anti-MIgG did not neutralize the ECSA at all. These results sug gest that the observed neutralizing action by MoAb is not mediated by its putative cytotoxic action directed to CFU-E but due to its immunospeeific action to EPO. To further confirm the above-mentioned results ob tained in the FMLC cultures, the neutralizing action of MoAb was re-examined in the cultures of bone marrow cells from adult mice (table 2). In the marrow cell cultures, too, MoAb inhibited the formation of erythroid colonies stimulated by mouse plasmas, though incompletely.
178
Sakata/E noki/U eda
Immunoaffinity Chromatography
In control experiments, rHuEPO and sheep plasma EPO (10 units each) were applied on the immunoadsorbent column with the immobilized anti-EPO MoAb R2. The flow-through fractions contained no EPO activity for rHuEPO (fig. 4A) and a trace amount (0.05 U) of the ac tivity for sheep EPO (fig. 4B). Both the standard EPOs re tained on the column were eluted sharply by acetate buffer (pH 2.5). On the other hand, when intact plasma from nor mal, anemic, or hypoxic mice was loaded onto the column, a small quantity of the ECSA, say 0.1 U (table 3), was found in the flow-through fractions, but the majority of the ECSA was recovered in the bound fractions (fig. 5). Al most all the plasma proteins passed freely through the col umn. The total ECSA (expressed as EPO activity) of the bound fractions for anemic and hypoxic plasmas was 2.6—3.1 times as high as that of normal plasma (table 3). In the bound fractions, recovery of the activity was 73-78%. Thus, most of the plasma ECSA could be bound to the immunoadsorbent column, and therefore appeared to be immunochemically identical to EPO. MoAb R2 neutralization experiments on the active fractions were performed to further confirm that the ECSA of the bound fractions, but not of the flow-through fractions, is immunoreactive with MoAb R2. In fraction ation of hypoxic plasma, MoAb did not neutralize the passing ECSA, but did the retained ECSA (fig. 6A). Simi lar results were obtained for normal and anemic plasmas and crude EPO preparation from sheep plasma (fig. 6B).
Mouse Plasma ECSA and A nti-E PO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Anti EPO (ng/well)
Fig. 3. Reversal of anti-EPO MoAb R2-induced CFU-E inhibi tion by increasing doses of rHuEPO (A) or anemic mouse plasma (B). 10 ul of MoAb (7 ng/well; solid symbols) or a-medium (open symbols) was added to 110 pi culture mixture supplemented with different doses of rHuEPO or anemic plasma (mean PCV. 0.15 1/1). The 12th day FMLCs were cultured at cell density of 3 x 103 cclls/wcll. Shaded areas represent the levels of colony formation in the absence of rHuEPO, anemic plasma, and MoAb. Each symbol represents the triplicate mean ±SD.
Table 1. Neutralization of mouse plasma ECSA by anti-EPO
MoAb R2
Table 2. Neutralization of mouse plasma ECSA by anti-EPO MoAb R2 in mouse bone marrow cell cultures
Mean number of colonics ± S D S
Mean number of colonies ± SD* 80 ± 4 80 ± 6
Connaught EPO (25 mU) Connaught EPO (25 mU) +anti-E PO “ Connaught EPO (25 mU) + anti-MMTV1’ Connaught EPO (25 mU) + antimouse IgG° Connaught EPO (25 mU) + (anti-EPO + antimousc IgG)'1
206 ± 8 87±4k 203 ± 7 207 ± 8
rHuEPO (25 mU) rHuEPO (25 mU) +anti-EPO rHuEPO (25 mU) + anti-MMTV rHuEPO (25 mU) +antimouse IgG rHuEPO (25 mU) + (anti-EPO + antimouse IgG)
204 ±7 85 ± 4k 198 ± 6 200 ± 6
Normal plasma1' Normal plasma + anti-EPO Normal plasma + anti-MMTV (Normal plasma + antimouse IgG)f (Normal plasma + (anti-EPO + antimouse IgG))s
134 ±5 88±4k 129 ± 3 131 ±6
Hypoxic plasmab Hypoxic plasma + anti-EPO Hypoxic plasma + anti-MMTV (Hypoxic plasma + antimouse IgG) (Hypoxic plasma + anti-EPO + antimouse IgG))
169 ± 7 89±5k 166 ± 7 166 ± 5
Anemic plasma' Anemic plasma + anti-EPO Anemic plasma + anti-MMTV (Anemic plasma + antimouse IgG) (Anemic plasma + (anti-EPO + antimouse IgG))
208 ±7 90 ± 3k 204 ± 8 204 ± 8
209 ±5
199 ± 7
135 ± 4
171 ±7
Control (0 mU) Control (0 mU) + anti-EPO
1.5 ±0.5 1.5 ±0.5
rHuEPO (25 mU) rHuEPO (25 mU) + anti-EPO“ rHuEPO (25 mU) + antimouse IgGb
60.0 ±3.1 2.5 ± 0.5h 57.0 ±2.5
Normal plasmac Normal plasma + anti-EPO (Normal plasma + antimouse IgG)J
21.0 ±2.2 4.8 ± 0.4' 21.0 ± 1.9
Hypoxic plasma0 Hypoxic plasma + anti-EPO (Hypoxic plasma + antimouse IgG)
49.0 ±1.9 5.8 ±0.4' 47.3 ±1.9
Anemic plasmaf Anemic plasma + anti-EPO (Anemic plasma + antimouse IgG)
54.0 ± 1.9 6.0 ±0.7' 52.8 ±2.2
Bone marrow cells from adult mice (7.5 x 103/well) were used. 11 0.9 ,ug of anti-EPO MoAb R2 was added to cell cultures with rHuEPO or mouse plasma. b 1.2 mg of anti-M-IgG was added to cell cultures with rHuEPO. c Plasma (10 pi, 1/2-fold concentration) from normal mice (mean PCV, 0.47 1/1) was added to cell cultures without EPO. d Mouse plasmas were preincubated with anti-M-IgG before ad dition to cell cultures without EPO. c Plasma (10 pi, 1/2-fold concentration) from mice exposed to 0.35 atm for 3 days was added to cell cultures without EPO. ' Plasma (10 pi, 1/2-fold concentration) from bleeding-induced anemic mice (mean PCV, 0.25 1/1) was added to cell cultures without EPO. p Mean ±SD of quadruplicate wells. h Significantly different from control value: p Mean ±S D of triplicate wells. k Not significantly different from control value: p > 0.05.
179
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Control (0 m il) Control (0 mU) +anti-EPO
Fig. 4. Immunoaffinity chromatography of rHuEPO (A) and sheep plasma Step III EPO (B). • — • = EPO activity (U/fraction) estimated by reference to a standard titration curve using Step III E P O ;....... — absorbance at 280 nm. Two preparations of the stan dard EPO (10 U), dissolved in 1 ml of PBS containing 0.01% bovine scrum albumin, were applied on the immunoadsorbent column in which anti-EPO MoAb R2 fixed on Affi-Gcl 10 was packed. Arrows indicate the positions of change of buffers. The constituent of the buffers used is described in ‘Materials and Methods’. Fig. 5. Immunoaffinity chromatography of plasma from normal (A), anemic (B). and hypoxic (C) mice. • — • = EPO activity (U/
fraction) estimated by reference to a standard titration curve using sheep plasma Step III E P O ;....... = absorbance at 280 nm. Mouse plasmas (1 ml) were applied on the MoAb R2 immunoadsorbent col umn. Arrows indicate the positions of change of buffers. The constit uent of the buffers used is described in ‘Materials and Methods’.
Table 3. Binding of mouse plasma ECSA to immunoadsorbent column containing immobilized anti-EPO MoAb R2
Sample
EPO activity“ mU
557 (100%) Normal plasma0 Anemic plasma'1 1,351 (100%) I lypoxic plasma0 1,650 (100%)
Flow-through fraction
Bound fraction
EPO activity mU
recoveryb %
EPO activity mU
rccovcryb %
95 95 99
17.1 7.0 6.0
404 1.048 1,239
72.5 77.6 75.1
180
Sakata/Enoki/U eda
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
a Before fractionation procedures, the ECSA level of plasma samples was estimated by a multiple-dose parallel-line bioassay, in which sheep plasma Step III EPO was used as a standard. Results are expressed as international units of EPO activity. b Total activity applied per column was defined as 1(10% activity. c Plasma from normal mice (mean PCV, 0.44 l/l). d Plasma from bleeding-induced anemic mice (mean PCV. 0.28 l/l). 0 Plasma from hypoxic mice exposed to 0.35 atm for 4 days.
Fig. 6. Anti-EPO MoAb R2 neutralization of ECSA fractionated by means of immunoaffinity chromatography. Each active fraction of hypoxic mouse plasma (A) and sheep plasma Step III EPO (B) was assayed for ECSA in the presence (solid symbols) and absence (open symbols) of MoAb (0.9 pg/wcll). The 12th day FMI.Cs (3 x lOVwell) were used. Upper and lower shaded areas indicate the levels of colony formation in the presence and absence of Step III EPO (25 mU/well). respectively. Each symbol represents the triplicate mean.
Table 4. Neutralization of ECSA in immunoadsorbent column fractions by anti-F.PO MoAb R2 and/or R6 or rabbit antirHuEPO scrum
Mean number of colonics ± SD' no antibody
R2S
R6h
R2 + R6'
antiserum’
NRSk
Control (0 mU) rHuEOP (25 mU)
130 ±4 263 ± 5
—
—
-
-
-
Normal plasma3 Flow-through fraction1’ Bound fraction1
131 ±3'
128 ±3'
126 ±2'
130 ±4'
264 ±3
180 ±4 144 ± 2 163 ± 4
140 ± 2m 142 ± 2m 127 ± 4'
146 ± 2 m 143 ± 3 m 128 ± 2'
142 ±3™ 143 ±2™ 126 ± 3'
143 ± 4 m 143 ± 3m 130 ±4'
182 ±3 153 ±3 170 ± 3
Hypoxic plasmad Flow-through fraction Bound fraction
243 ±5 143 ±3 219 ±5
143±4m 142 ± 2 m 127 ±3'
149 ± 3 m 142 ±3"’ 129 ±3'
143 ± 3 ” 144 ± 3n 128 ± 41
145 ± 3 m 143 ± 2 m 131 ±2'
240 ±5 152 ±3 222 ±5
Anemic plasma1' Flow-through fraction Bound fraction
267 ±5 144 ± 2 231 ±5
143 ± 3 m 141 ± 3 m 127 ±2'
149 ± 2 m 142 ± 2m 129 ± 2‘
141 ± 3 n 145 ± 2 n 128 ± 4‘
146 ±4"’ 142 ± 2 m 131 ±3'
262 ± 6 154 ±3 233 ±4
181
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
The 12th day FMLCs (3 x lOVwell) were used. a Plasma (10 pi, 1/2-fold concentration) from normal mice (mean PCV, 0.47 1/1) was added to cell cultures without EPO. b Pooled flow-through fractions in immunoaffinity chromatography of 1 ml mouse plasma were adjusted to 2 ml of volume, and 10 ul of the concentrated fraction was added to cell cultures without EPO. c Pooled bound fractions in immunoaffinity chromatography of 1 ml mouse plasma were adjusted to 2 ml of volume, and 10 pi of the con centrated fraction was added to cell cultures without EPO. d Plasma (10 pi, 1/2-fold concentration) from mice exposed to 0.35 atm for 3 days was added to cell cultures without EPO. c Plasma (10 pi, 1/2-fold concentration) from bleeding-induced anemic mice (mean PCV, 0.25 1/1) was added to cell cultures without EPO. f Mean ± SD of triplicate wells. 8 0.9 ug of anti-EPO MoAb R2 was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. h 10 ug of anti-EPO MoAb R6 was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. ‘ 0.9 pg of anti-EPO MoAb R2 and 10 ug of anti-EPO MoAb R6 were simultaneously added to cell cultures with rHuEPO, intact plasma or fractionated plasma. ’ Rabbit anti-rHuEPO scrum (10 pi, 1/2-fold concentration), heat (56 °C, 30 min)-inactivated and dialyzed against Dulbecco’s PBS, was added to cell cultures with rHuEPO. intact plasma or fractionated plasma. k Normal rabbit serum (10 pi, 1/2-fold concentration), heat (56 °C, 30 min)-inactivated and dialyzed against Dulbecco’s PBS, was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. 1 Not significantly different from control values: p > 0.1. m Significantly different from control values: p < 0.05.
Discussion Although several anti-EPO MoAbs with neutralizing ability have been reported from four laboratories so far [22, 26-29], only two of them, R2 [22] and anti-Ep-16 [27], were capable of efficiently neutralizing the biological ac tivity of mouse EPO. MoAb R2 and anti-Ep-16 effectively blocked the capacity of mouse EPO to stimulate the in corporation of [?Hl-thymidine into the DNA of FMLCs and of mouse spleen cells, respectively. Using MoAb R2, we have attempted to demonstrate the identity of mouse plasma ECSA with EPO. In the preliminary experiments, we demonstrated that MoAb R2 inhibits the CFU-E colo ny formation stimulated by standard EPO. This inhibition seems to be due to blockage of EPO binding to CFU-E by MoAb [22], The ECSA of both mouse plasma and crude prepara tion of sheep plasma was not completely bound to the immunoadsorbent column containing the immobilized MoAb R2 (fig. 4B, 5). This seems to be ascribable to the nonimmunoreactive nature of the ECSA present in the flow-through fractions, but not to technical reasons, be cause rHuEPO was completely bound to the column (fig. 4A) and the passing ECSA in the flow-through frac tions was not neutralized by MoAb R2 (fig. 6; table 4). The
182
Saka t a/E noki/U cda
passing ECSA was additive to standard EPO in stimulat ing CFU-E growth and was heat-stable and nondialyzable. Thus, the passing ECSA had some properties similar to EPO. However, the passing ECSA was neutralized neither by anti-rHuEPO scrum nor by MoAb R6 (table 4), sug gesting that the passing ECSA is not an immunochemically different EPO but a non-EPO stimulator. This is fur ther supported by the observation that the contents of the passing ECSA in anemic and hypoxic plasmas were almost equal to that in normal plasma (tables 3, 4). Similarly, it was reported that non-EPO stimulator(s) may contribute to the high ECSA levels of neonatal mouse plasma [18,20, 21]. Normal mouse serum was found to contain two differ ent CFU-E-stimulating factors, ESCF [16] and PSA [17], which both stimulated the growth of murine marrow CFU-E synergistically with EPO. In addition, ESCF was heat-labile [16] and had no effect on fetal mouse liver CFU-E [30]. Based on these properties, ESCF and PSA appear to be different from our passing ECSA. Growth hormones [6], thyroid hormones [7], steroids [8], IGF II [10], cyclic nucleotides [12], and activin A [14] were found to potentiate colony formation of CFU-E in the presence but not the absence of EPO. On the other hand, insulin [11] and hemin [13] were shown to stimulate CFU-E formation independently of EPO. However, both substances are probably dialyzable in view of their smaller molecular sizes than 13,000 (MW cutoff size of dialysis membrane). Accordingly, it seems unlikely that the several candidates above stated represent the passing ECSA. Fagg [15] found a heat-stable and nondialyzable EpLA. which could be dis tinguished from EPO on the basis of affinity for concanavalin A and different molecular weight, in mouse spleen cell conditioned medium. This EpLA [31] and IGF I [9] were found to stimulate CFU-E from FMLCs in the ab sence of and nonsynergistically with EPO. IGF I present in scrum is known to be bound to specific serum proteins and to exist as 50,000- and 150,000-dalton complexes [32], Thus, EpLA and bound IGF I have some biological and bio chemical properties similar to those of the passing ECSA. Therefore, the passing ECSA may be EpLA or IGF I. Fur ther biochemical characterization of the passing ECSA is required for its identification. As shown in figure 1 and tables 1 and 4, the relatively high level of CFU-E colony formation was observed in the control cultures without any exogenous supplementation of EPO, in agreement with the earlier results [33,34], This spontaneous colony formation was not affected by adding anti-EPO MoAb (tabic 1). Therefore, as described by Cole et al. [33], it seems most likely that these colonies were de rived from CFU-Es which had been already triggered by
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
The passing ECSA had some biological and biochem ical properties similar to EPO. This ECSA showed a doseresponse curve parallel to the standard EPO curve, addi tive effect with standard EPO on the colony formation, and heat (56 °C for 30 min)-stable and nondialyzable na tures [data not shown]. To examine whether the passing ECSA is EPO unrcactive to MoAb R2 or non-EPO stim ulator, a neutralization study was carried out using rabbit anti-rHuEPO serum. MoAb R2 and MoAb R6 with neu tralizing activity against mouse EPO [22]. MoAb R2 and R6 were found to recognize different antigenic determi nants [22], R2 and/or R6 or antiserum did not neutralize the ECSA in the pooled flow-through fractions for nor mal, hypoxic and anemic plasmas at all, but did completely neutralize both the ECSA in the pooled bound fractions for these plasmas and rHuEPO (table 4). NRS added as control did not inhibit the colony formation in any cul tures. It should be noted that the colony level in the cul tures with both intact plasmas and MoAbs or antiserum corresponds to that in the cultures with the flow-through fractions. These results suggest that the passing ECSA is the non-EPO stimulator responsible for incomplete neu tralization of intact plasma ECSA by MoAbs or antiserum.
EPO in the fetal circulation. Another possible explanation might be that IGF I [35], which is present in FCS, stim ulates the colony formation. However, IGF I in FCS ap pears to play a minor role in the spontaneous colony for mation, because just a slight reduction in the spontaneous colony formation was caused by decreasing the concentra tion of FCS from 30 to 10% in our culture system [data not shown]. This study provides evidence that mouse plasma ECSA detectable in FMLC cultures is mainly due to EPO. First, most of the plasma ECSA was neutralized by MoAbs against EPO or anti-rHuEPO serum (fig. 2; tables 1, 4). This neutralization could be reversed by excess of EPO or anemic plasma (fig. 3) or by preincubation of MoAb with anti-M-IgG (table 1). These results suggest that the inhibi tion by MoAb is caused by immunospccific binding of MoAb to EPO. In fact, MoAb • EPO complex appears to have little or no affinity to EPO receptor [16], Second, most of the plasma ECSA was bound to the immunoadsorbent column (fig. 5), and the retained ECSA was totally blocked by MoAbs or anti-rHuEPO scrum (fig. 6A; table 4). Moreover, plasmas from mice under erythropoietic stimulation contained larger amounts of the retained ECSA compared with normal mouse plasma (fig. 5; tables
3, 4). In contrast, plasma from transfusion-induced poly cythemic mice contained no or decreased detectable ECSA [data not shown]. Such quantitative changes of the ECSA are comparable to a well-known physiologic re sponse of EPO. The circulating EPO levels arc normally increased by tissue hypoxia and decreased by tissue hyperoxia. Third, mode of action of the ECSA in nonfractionated plasma was similar to that of EPO. The plasma ECSA and standard EPO showed parallelism in the doseresponse curves (fig. 1) and additivity with each other in their function. From these results, CFU-E assay method using FMLCs is considered to be valid for mouse plasma EPO assay. However, it should be noted that a small por tion of the detected ECSA may be due to non-EPO stim ulator.
Acknowledgements We thank Dr. J. Morimoto for his generous gifts of anti-MMTV MoAb. This work was supported in part by grants-in-aid for Scientific Research (grant No. 01570048) and for Encouragement of Young Sci entists (grant No. 01770062) from the Ministry of Education, Science and Culture of Japan.
References 8 Singer JW. Samuels AI. Adamson JW: Steroids and hematopoiesis. I. The effect of steroids on in vitro erythroid colony growth: Structure ac tivity relationships. J Cell Physiol 1976:88:127134. 9 Kurtz A. Jelkmann W. Bauer C: A new candi date for the regulation of erythropoiesis. Insu lin-like growth factor I. FF.BS Lett 1982:149: 105-108. 10 Dainiak N. Kreczko S: Interactions of insulin, insulinlike growth factor II. and platelet-de rived growth factor in erythropoietic culture. J Clin Invest 1985:76:1237-1242. 11 Kurtz A. Jelkmann W. Bauer C: Insulin stim ulates erythroid colony formation independ ently of erythropoietin. Br J Haematol 1983:53: 311-316. 12 Brown JE, Adamson JW: Modulation of in vitro erythropoiesis: Enhancement of erythroid colony growth by cyclic nucleotides. Cell Tissue Kinet’1977:10:289-298. 13 Porter PN. Meints RH. Mesner K: Enhance ment of erythroid colony growth in culture by hernin. Exp Hematol 1979:7:11-16.
14 Yu J. Shao L. Lemas V. Yu AL. Vaughan J. Rivier J. Vale W: Importance of FSH-releasing protein and inhibin in erythrodifferentiation. Nature 1987;330:765-767. 15 Fagg B: Is erythropoietin the only factor which regulates late erythroid differentiation? Na ture 1981:289:184-186. 16 Blanchet JP. Arnaud S, Samarut J. Bouabdelli M: A factor present in normal mouse serum stimulates late erythroid precursor prolifera tion. Fxp Hematol 1984;12:595-602. 17 Udupa KB. Lipschitz DA: Proerythroblast stimulating activity: Its purification from mouse serum and its effect on mouse erythroid cell proliferation in vitro. Br J Haematol 1988:69:157-162. 18 Widness JA, Sanengen T, I laga P. Clemons GK, Myhre K. Halvorsen S: Correlation of plasma erythropoiesis stimulating factor(s) and immunoreactive erythropoietin levels during rapid growth in the mouse. Acta Physiol Scand 1989:136:527-533.
183
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
1 Hâgâ I’. Falkanger B: In vitro assay for erythro poietin: Erythroid colony formation in methyl cellulose used for the measurement of ery thro poietin in plasma. Blood 1979:53:1172-1181. 2 Mcberg A. Hâgâ 1’, Johansen M: Plasma eryth ropoietin levels in mice during the growth pe riod. Br J Haematol 1980:45:569-574 3 Hâgâ P, Kristiansen S: Role of the kidney in foetal erythropoiesis: Erythropoiesis and erythropoietin levels in newborn mice with re nal agenesis. J Embryol Exp Morph 1981; 61:165-173. 4 Cutler RL. Johnson GR. Nicola NA: Charac terization of murine erythropoietin. Exp He matol 1985:13:899-905. 5 Choppin J. Egrié J. Casadevall N. Lacombe C. Muller O. Varet B: Biochemical analyses of murine erythropoietin from plasma and from cloned erythroleukemia cells. Exp Hematol 1987;15:171-176. 6 Golde DW, Bersch N. l.i CH: Growth hor mone: Species-specific stimulation of erythro poiesis in vitro. Science 1977:196:1112-1113. 7 Golde DW, Bersch N. Chopra IJ. Cline MJ: Thyroid hormones stimulate erythropoiesis in vitro. Br J Haematol 1977:37:173-177.
25 Gruber DF. Zucali JR, Mirand EA: Identifica tion of erythropoietin producing cells in fetal mouse liver cultures. Exp Hematol 1977:5:392398. 26 Sytkowski AJ. Donahue KA: Immunochemical studies of human erythropoietin using site-spe cific anti-peptide antibodies. J Biol Client 1987; 262:1161-1165. 27 Wognum AW, Lansdorp PM. Krystal G: Immu nochemical analysis of monoclonal antibodies to human erythropoietin. Exp Hematol 1990: 18:228-233. 28 D’Andrea AD, Szklut PJ. l.odish HE. Alder man EM: Inhibition of receptor binding and neutralization of bioactivity by anti-erythro poietin monoclonal antibodies. Blood 1990:75: 874-880. 29 Feldman L. Heinzcrling R. Hillam RP. Chern YE, Frazier JG, Davis Kl„ Sytkowski AJ: Four unique monoclonal antibodies to the putative receptor binding domain of erythropoietin in hibit the biological function of the hormone. Exp Hematol 1992:20:64-68. 30 Arnaud S. Blanche! JP: The erythropoietic stimulating activity of normal mouse serum: Effect on erythroid precursors from different haematopoietic tissues and suggested role of accessory cells. Scand J Haematol 1986:37:180188. 31 Fagg B. Roitsch CA: Characterization of eryth ropoietin-like activity from mouse spleen cells. J Cell Physiol 1986;126:1-9.
32 Daughaday WH. Ward AP. Goldberg AC. Trivedi B. Kapadia M: Characterization of soma tomedin binding in human scrum by ultracen trifugation and gel filtration. J Clin Endocrinol Metab 1982;55:916-921. 33 Cole RJ, Regan T. White SL. Check EM: The relationship between erythropoietin-depend ent cellular differentiation and colony-forming ability in prenatal haemopoietic tissues. J Embryol Exp Morph 1975:34:575-588. 34 Kanamaru A. Okamoto T. Hara H. Nagai K: Developmental changes in erythropoietin re sponsiveness of late erythroid precursors in mouse hemopoietic organs. Dev Biol 1982:92: 221-226. 35 Kurtz A. Hartl W, Jelkmann W, Zapf J. Bauer C: Activity in fetal bovine serum that stimulates erythroid colony formation in fetal mouse liv ers in insulin-like growth factor I. J Clin Invest 1985;76:1643-1648.
184
Sakata/Enoki/U cda
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
19 Sanengen T. Halvorscn S: Plasma erythropoiesis stimulating factor(s) in neonatal mice: In vitro dose response and chromatography stud ies. Pediatr Res 1987;21:148-151. 20 Sanengen T. Myhre K. Halvorscn S: Erythro poietic factors in plasma from neonatal mice. In vivo studies by the exhypoxic polycythacmic mice assay for erythropoietin. Acta Physiol Seand 1987;129:381-386. 21 Sanengen T, Clemons GK, Halvorscn S, Widncss JA: Immunoreactive erythropoietin and erythropoiesis stimulating factor(s) in plasma from hypertransfused neonatal and adult mice. Studies with a radioimmunoassay and a cell culture assay for erythropoietin. Acta Physiol Scand 1989:135:11-16. 22 Goto M. Murakami A. Akai K. Kawanishi G. Ueda M. Chiba H. Sasaki R: Characterization and use of monoclonal antibodies directed against human erythropoietin that recognize different antigenic determinants. Blood 1989: 74:1415-1423. 23 Iscovc NN. Siebcr F: Erythroid progenitors in mouse bone marrow detected by macroscopic colony formation in culture. F.xp Hematol 1975:3:32-43. 24 Goto M. Akai K. Murakami A. Hashimoto C, Tsuda F.. Ueda M. Kawanishi G. Takahashi N. Ishimoto A. Chiba H. Sasaki R: Production of recombinant human erythropoietin in mam malian cells: Host-cell dependency of the bio logical activity of the cloned glycoprotein. Bio/ Technology 1988;6:67-71.
a Second Department of Physiology, Nara Medical University, Nara; b Research Institute of Life Science, Snow Brand Milk Products, Tochigi, Japan
Key Words CFU-E Erythropoietin Fetal mouse liver Monoclonal antibody Mouse plasma Stimulating activity
Neutralization and Immunoaffinity Chromatography of Erythroid Colony-Stimulating Activity in Mouse Plasma by an AntiErythropoietin Monoclonal Antibody
Abstract A relationship between erythropoietin (EPO) and erythroid colony-stimulat ing activity (ECSA) in mouse plasma was examined in fetal mouse liver cell (FMLC) cultures using a monoclonal antibody (MoAb) R2 raised against re combinant human EPO. Most of the ECSA in plasma from normal, anemic, and hypoxic mice was neutralized by MoAb. This neutralization could be re versed by addition of excess of anemic plasma or by preincubation of MoAb with goat anti-mouse IgG antibody. Most of the plasma ECSA was bound to an immunoadsorbent column containing the immobilized MoAb, and the re tained ECSA was completely neutralized by MoAb. The plasma ECSA and standard EPO showed parallel dose-response curves and additive effect on CFU-E stimulation. Based on these findings, we conclude that mouse plasma ECSA detected by CFU-E assay using FMLCs is mainly due to EPO.
Introduction Erythropoietin (EPO) is a glycoprotein hormone that regulates red blood cell production by promoting the pro liferation and differentiation of late erythroid progenitor cells (CFU-E). One of the most useful in vitro bioassays for EPO is CFU-E colony-forming assay, which has the ad vantage of directly measuring EPO activity in stimulating CFU-E growth [1], This assay procedure was used for mea surement of plasma EPO levels in mice [2,3] and also for detection of EPO activity in chromatography of mouse plasma or serum [4,5]. To validate the assay, however, evi-
Received: February 17,1992 Accepted: June 15,1992
dence that erythroid colony-stimulating activity (ECSA) in mouse plasma is ascribable to EPO should be provided, because a wide variety of substances other than EPO, i.e. growth hormones [6], thyroid hormones [7], steroids [8], insulin-like growth factor (IGF) I [9] and II [10], insulin [11], cyclic nucleotides [12],hcmin [13], activin A [14],EPOlike activity (EpLA) [15], erythropoietic stimulating cofac tor (ESCF) [16], and proerythroblast stimulating activity (PSA) [17], have been reported to be potent stimulating agents for CFU-E colony growth. Halvorsen’s group re ported a significant correlation between plasma ECSA levels and EPO levels measured by radioimmunoassay in
Dr. Susumu Sakata Second Department of Physiology Nara Medical University S40 Shijo-cho. Kashihara Nara 634 (Japan)
© 1992 S. Kargcr AG, Basel (XXH-5792/92/0884-0175 $ 2.75/0
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Susumu Sakataa Yasunori Enokia Masatsugu Uedah
Materials and Methods Mice and Plasma Ten-week-old male ICR mice were used throughout the study. Posthemorrhagic anemia was induced by a single or double bleeding of an amount of blood equivalent to 2.0% of the body weight from the tail vein. Twenty-four hours later, heparinized blood was obtained by puncture of the vena cava under ether anesthesia. Hypoxia was pro duced by rearing the animals in a Plexiglas hypobaric chamber con nected with a vacuum pump and keeping them at 0.35 atm for 3 or 4 days (22.5 h/day). Immediately after hypoxia exposure, blood was ob tained as described above. After centrifugation, the plasma samples were sterilized through 0.45-um filters (Millipore, Bedford, Mass., USA), and stored at -9 0 °C until use. CFU-E Assays Assays for CFU-E colony growth were performed in mcthylcellulose microwell cultures. Briefly, liver cells (3-4 x 10J) from the day-12 fetal ICR mice, suspended in 100 pi of a-medium (Flow Labs., McLean, Va., USA) containing 0.8% mcthylccllulose (Shin-ctsu Chemicals, Osaka, Japan), 30% heat-inactivated fetal calf serum (FCS; Flow Labs.), and Ulr'M a-thioglycerol (Nacalai Tesque, Kyoto, Japan) [23], were plated into each well of a 96-well microtiter plate (Corning Glass Works, Corning, N.Y.. USA). Then. 10 ul of standard EPO solution and/or plasma samples were added to each culture. Sheep plasma Step III EPO (3.3 IU/mg; Connaught, Willowdaie, Canada) and/or rHuEPO expressed in baby hamster kidney cells [241 were used as a standard EPO. Cultures were carried out at 37 °C in a fully humidified atmosphere of 5% CO, in air, and crythroid colonies containing eight or more cells were directly scored using an inverted microscope after a 48-hour incubation. In some experiments, the col ony counts were carried out after the culture wells were fixed in situ by flooding with five drops of 0.4% glutaraldchyde in a-medium. All cultures were performed in triplicate or quadruplicate.
176
Sakata/E noki/U cda
Neutralization Studies Neutralization of the biological ECSA in normal, hypoxic, and anemic plasmas was conducted by use of an anti-EPO MoAb R2 which had been shown to neutralize the biological activity of murine EPO [22]. In most experiments, the purified MoAb dissolved in a-medium with 2% FCS and the plasma were added directly to the CFU-E cultures without prior incubation. In some experiments, the plasma was prcincubated with MoAb for 2 h at 37 °C, and subse quently at 4 °C overnight. Then, goat antimousc IgG polyclonal anti body (anti-M-IgG; Tago, Burlingame, Calif., USA) was added to this mixture for further preincubation. The immunocompicxes of anti-MIgG and MoAb were removed from the plasma mixture by refriger ated centrifugation (18,500g, 10 min). The ECSA remaining in the su pernatant was determined in the CFU-E assay. Furthermore, to ex amine whether the neutralization by anti-EPO MoAb is due to spe cific antigen-antibody interactions, MoAb was first prcincubated with an excessive amount of anti-M-IgG as described above. After centrifugation of this mixture, the supernatant was once again prein cubated with mouse plasma. The resulting immune precipitates were removed hy centrifugation, and the supernatant was added to the cul tures for assay of ECSA. In these experiments, MoAb against mouse mammary tumor virus (MMTV) and anti-M-IgG were employed as a control antibody. Irnmunoaffinity Chromatography Mouse plasma sample (1 ml) was loaded at room temperature on an immunoadsorbent column (dimensions 1.2 x 1.8 cm; bed volume 2 ml) containing MoAb R2 fixed on Affi-Gel 10 (Bio-Rad, Richmond. Calif., USA) equilibrated with 10 mM phosphate-buffered saline (PBS), pH 7.4 [22]. Following application of the sample, the column flow was stopped and the column was left overnight at 4°C. The col umn was extensively washed at room temperature with 100 ml of PBS, 20 ml of lOmA/NaPi, pH 7.4, containing0.5 M NaCI, and finally 20 ml of 0.15 M NaCI, in this order, and then eluted by 20 ml of 0.2M acetic acid, pH 2.5, containing 0.15 M NaCI. Fractions of 1 ml were collected. Each fraction thus eluted was immediately neutralized by adding 50 ul of 3.2 M Tris, and was dialyzed against several changes of Dulbccco’s PBS. pH 7.4, and then a-medium. The dialyzed fractions, steril ized through a 0.45-um filter, were tested for ECSA. The ECSA of each fraction was expressed as international units of EPO activity, which was calculated by reference to the dose-response curve for standard Step III EPO. In addition, neutralization of each fraction ECSA was studied using the MoAb R2. Furthermore, neutralization of the ECSA of pooled fractions was attempted by anti-EPO MoAb R6 [22] and rabbit anti-rHuEPO serum. Both the flow-through and the bound fractions were collected separately. These pooled frac tions were dialyzed against dcionized-distilled water, and then lyophilized. The dry fractions were dissolved in 2 ml of a-medium with 2% FCS, and sterilized through a 0.45-um filter. Normal rabbit serum (NRS), heat-inactivated at 56 °C for 30 min. was used as control for irrelevant specificity. At the end of elution, the column was washed and preserved in PBS containing 0.02% sodium azide. Statistics Data arc presented as the mean ±SD . Results were analyzed us ing the Student’s t test.
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
both adult and neonatal mice [18], but main focus of their investigations was to answer a question of whether high plasma ECSA found in neonatal mice is entirely due to EPO [18-21]. Thus, the relationship between the ECSA and EPO has not been fully studied in adult mice yet. Recently, Goto et al. [22] developed a monoclonal anti body (MoAb) R2 directed toward recombinant human EPO (rHuEPO), which was proved to efficiently inhibit the stimulation by anemic mouse plasma of [3H]-thymidine incorporation into DNA in cultured fetal mouse liver cells (FMLCs). To clarify the relationship between the ECSA and EPO in adult mouse plasma, using this highly specific MoAb, we have attempted neutralization and immunoaffinity chromatography of the plasma ECSA of normal, anemic, and hypoxic adult mice. In addition, the mode of action of the ECSA was compared with that of standard EPO in CFU-E colony assay using FMLCs. We present here evidence that the plasma ECSA found in adult mice is predominantly due to EPO.
Results Properties o f ECS A in Mouse Plasma
Neutralization o f Plasma ECSA by Anti-EPO MoAb R2
Anti-EPO MoAb R2 inhibited dose-dependently the formation of CFU-E colonies stimulated by standard EPO or normal or anemic mouse plasma when added simulta neously to the cultures without prior incubation (fig. 2). MoAb dose-response curves for standard EPO and mouse plasmas, expressed as percent inhibition, were similar to one another at antibody doses less than 219 ng/well. Fifty
OOI 0 0 2 0 0 4 0 0 7 014 Q28 0 5 7
Plasma
1.1
2.3
4.5
9.1
16.7
cone. (%)
Fig.I.C omparison of dosc-rcsponse curves for standard EPO (O) and plasma from normal (A ) or hypoxic (A ) mice. Liver cells from the 12th day fetal mice (4 x l()7well) were used. The standard EPO was sheep plasma Step III. Each symbol represents the triplicate mean ± SD.
percent inhibition was obtained at antibody dose of 3.4 ng/ well. At antibody doses more than 219 ng/well, most of the plasma ECSA was neutralized. For mouse plasmas and sheep plasma Step III EPO, the mean percent inhibition even at such high doses of MoAb was not complete, but about 90 and 85%, respectively. For rHuEPO, on the other hand, the complete suppression by MoAb was observed. Similarly, anti-EPO MoAb R2 was capable of neutra lizing the ECSA of fetal mouse plasma [data not shown], suggesting that the direct addition of MoAb to the cul tures neutralizes fetal EPO produced endogenously by macrophages or Kupffer cells in FMLC cultures [25]. Therefore, such direct-additive experiments do not ex clude the possibility that the ECSA found in plasma is not due to EPO but due to an inducer of endogenous EPO production in the cultures. Accordingly, action of MoAb on the neutralization of ECSA in plasma was examined in detail. Mouse plasma was prcincubated with MoAb before addition to the cultures, and then an excess of MoAb was removed from plasma by immunoprécipitation with antiM-IgG. This pretreatment with MoAb almost completely abrogated the ECSA in normal, anemic, and hypoxic plas mas [data not shown]. This result strongly suggests that the ECSA in mouse plasma is EPO but not the inducer of en dogenous EPO. These neutralizing actions of MoAb against ECSA in anemic plasma or rHuEPO were completely overcome by
177
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
At concentrations less than 4.5%, normal mouse plasma stimulated dose-dependently the formation of CFU-E colonies from FMLC without exogenous EPO (fig. 1). Plasma from anemic or hypoxic mice had a similar but stronger stimulatory effect. The dose-response curves for these plasmas were evidently parallel to the curve for standard EPO. At the higher concentrations (9.1-16.7%), however, the number of the colonies decreased in a dosedependent fashion. From this result, mouse plasmas ap pear to contain small amounts of an inhibitor for CFU-E growth. To assess the inhibitory potency at the concentra tions less than 4.5%, an inhibitor assay was performed. Plasma samples were added to the cultures supplemented with an EPO dose (25 mU/wcll) enough to induce the maximum colony formation. Such concentrations of plasma caused no reduction in the colony formation [data not shown]. This fact, as well as parallelism in the dose-response curves, indicates that the inhibitor in mouse plasmas, if present, docs not affect the formation of CFU-E colonies at the additive plasma concentration less than 4.5%. To determine whether the stimulatory effect of mouse plasma on the colony formation is additive or synergistic with standard EPO, the EPO dose-response curves in the absence and presence of plasma were compared. Increas ing doses of EPO were added to the cultures with a fixed amount (0.26%) of normal or anemic plasma. In both the EPO dose-response curves, the plateau levels in the colo ny formation were the same [data not shown]. Thus, the ECSA in mouse plasmas were not synergistic but additive to standard EPO in the present CFU-E assay system. In addition, the ECSA of both standard EPO and mouse plasma remained biologically active even after heating at 56 °C for 30 min or dialysis against distilled water for 3 days (Viskase cellulose tubing; 13,000 MW cut-off). Thus, some biological and biochemical proper ties of mouse plasma ECSA were similar to those of stan dard EPO.
'A
100
B
Ç . . -O -.
.'A' ' Z
'
A
/
> 12.5 25 50 100 2 0 0 4 0 0 0 2 6 052 10 2.1 42 8.3 E P O (mU/well) Plasm a cone. (%)
Fig. 2. Dose-dependent inhibition of CFU-E colony formation by anti-EPO MoAb R2. Different doses of MoAb were added directly to the cultures supplemented with a sufficient amount (25 mU/well) of standard EPO, sheep plasma Step III (O )or rHuEPO ( A ), or 4.2% of plasma from normal (□; mean PCV, 0.45 1/1) or anemic (V ; mean PCV, 0.25 1/1) mice. Liver cells from the 12th day fetal mice (3 x 10V well) were used. The degree of the inhibition was calculated as:
Nmax-N,
% inhibition = _
_
x 100.
“ ^ m in
where Nnm, Nmin and Nxare the number of the colonies formed in the cultures supplemented with 25 mU of standard EPO or 4.2% of mouse plasma, only a-medium, and both the different doses of MoAb and 25 mU of standard EPO or 4.2% of mouse plasma, respectively. Nmax and Nmin arc the mean values. In this experiment, Nmax for the standard EPOs, normal and anemic plasmas were 263, 221, and 254, respectively. Nmi„ was 122. Each symbol represents the triplicate mean.
the addition of higher amounts of the plasma or rHuEPO (fig. 3). They were also abolished by preincubating MoAb with anti-M-IgG (table 1). Anti-MMTV MoAb or anti-MIgG did not neutralize the ECSA at all. These results sug gest that the observed neutralizing action by MoAb is not mediated by its putative cytotoxic action directed to CFU-E but due to its immunospeeific action to EPO. To further confirm the above-mentioned results ob tained in the FMLC cultures, the neutralizing action of MoAb was re-examined in the cultures of bone marrow cells from adult mice (table 2). In the marrow cell cultures, too, MoAb inhibited the formation of erythroid colonies stimulated by mouse plasmas, though incompletely.
178
Sakata/E noki/U eda
Immunoaffinity Chromatography
In control experiments, rHuEPO and sheep plasma EPO (10 units each) were applied on the immunoadsorbent column with the immobilized anti-EPO MoAb R2. The flow-through fractions contained no EPO activity for rHuEPO (fig. 4A) and a trace amount (0.05 U) of the ac tivity for sheep EPO (fig. 4B). Both the standard EPOs re tained on the column were eluted sharply by acetate buffer (pH 2.5). On the other hand, when intact plasma from nor mal, anemic, or hypoxic mice was loaded onto the column, a small quantity of the ECSA, say 0.1 U (table 3), was found in the flow-through fractions, but the majority of the ECSA was recovered in the bound fractions (fig. 5). Al most all the plasma proteins passed freely through the col umn. The total ECSA (expressed as EPO activity) of the bound fractions for anemic and hypoxic plasmas was 2.6—3.1 times as high as that of normal plasma (table 3). In the bound fractions, recovery of the activity was 73-78%. Thus, most of the plasma ECSA could be bound to the immunoadsorbent column, and therefore appeared to be immunochemically identical to EPO. MoAb R2 neutralization experiments on the active fractions were performed to further confirm that the ECSA of the bound fractions, but not of the flow-through fractions, is immunoreactive with MoAb R2. In fraction ation of hypoxic plasma, MoAb did not neutralize the passing ECSA, but did the retained ECSA (fig. 6A). Simi lar results were obtained for normal and anemic plasmas and crude EPO preparation from sheep plasma (fig. 6B).
Mouse Plasma ECSA and A nti-E PO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Anti EPO (ng/well)
Fig. 3. Reversal of anti-EPO MoAb R2-induced CFU-E inhibi tion by increasing doses of rHuEPO (A) or anemic mouse plasma (B). 10 ul of MoAb (7 ng/well; solid symbols) or a-medium (open symbols) was added to 110 pi culture mixture supplemented with different doses of rHuEPO or anemic plasma (mean PCV. 0.15 1/1). The 12th day FMLCs were cultured at cell density of 3 x 103 cclls/wcll. Shaded areas represent the levels of colony formation in the absence of rHuEPO, anemic plasma, and MoAb. Each symbol represents the triplicate mean ±SD.
Table 1. Neutralization of mouse plasma ECSA by anti-EPO
MoAb R2
Table 2. Neutralization of mouse plasma ECSA by anti-EPO MoAb R2 in mouse bone marrow cell cultures
Mean number of colonics ± S D S
Mean number of colonies ± SD* 80 ± 4 80 ± 6
Connaught EPO (25 mU) Connaught EPO (25 mU) +anti-E PO “ Connaught EPO (25 mU) + anti-MMTV1’ Connaught EPO (25 mU) + antimouse IgG° Connaught EPO (25 mU) + (anti-EPO + antimousc IgG)'1
206 ± 8 87±4k 203 ± 7 207 ± 8
rHuEPO (25 mU) rHuEPO (25 mU) +anti-EPO rHuEPO (25 mU) + anti-MMTV rHuEPO (25 mU) +antimouse IgG rHuEPO (25 mU) + (anti-EPO + antimouse IgG)
204 ±7 85 ± 4k 198 ± 6 200 ± 6
Normal plasma1' Normal plasma + anti-EPO Normal plasma + anti-MMTV (Normal plasma + antimouse IgG)f (Normal plasma + (anti-EPO + antimouse IgG))s
134 ±5 88±4k 129 ± 3 131 ±6
Hypoxic plasmab Hypoxic plasma + anti-EPO Hypoxic plasma + anti-MMTV (Hypoxic plasma + antimouse IgG) (Hypoxic plasma + anti-EPO + antimouse IgG))
169 ± 7 89±5k 166 ± 7 166 ± 5
Anemic plasma' Anemic plasma + anti-EPO Anemic plasma + anti-MMTV (Anemic plasma + antimouse IgG) (Anemic plasma + (anti-EPO + antimouse IgG))
208 ±7 90 ± 3k 204 ± 8 204 ± 8
209 ±5
199 ± 7
135 ± 4
171 ±7
Control (0 mU) Control (0 mU) + anti-EPO
1.5 ±0.5 1.5 ±0.5
rHuEPO (25 mU) rHuEPO (25 mU) + anti-EPO“ rHuEPO (25 mU) + antimouse IgGb
60.0 ±3.1 2.5 ± 0.5h 57.0 ±2.5
Normal plasmac Normal plasma + anti-EPO (Normal plasma + antimouse IgG)J
21.0 ±2.2 4.8 ± 0.4' 21.0 ± 1.9
Hypoxic plasma0 Hypoxic plasma + anti-EPO (Hypoxic plasma + antimouse IgG)
49.0 ±1.9 5.8 ±0.4' 47.3 ±1.9
Anemic plasmaf Anemic plasma + anti-EPO (Anemic plasma + antimouse IgG)
54.0 ± 1.9 6.0 ±0.7' 52.8 ±2.2
Bone marrow cells from adult mice (7.5 x 103/well) were used. 11 0.9 ,ug of anti-EPO MoAb R2 was added to cell cultures with rHuEPO or mouse plasma. b 1.2 mg of anti-M-IgG was added to cell cultures with rHuEPO. c Plasma (10 pi, 1/2-fold concentration) from normal mice (mean PCV, 0.47 1/1) was added to cell cultures without EPO. d Mouse plasmas were preincubated with anti-M-IgG before ad dition to cell cultures without EPO. c Plasma (10 pi, 1/2-fold concentration) from mice exposed to 0.35 atm for 3 days was added to cell cultures without EPO. ' Plasma (10 pi, 1/2-fold concentration) from bleeding-induced anemic mice (mean PCV, 0.25 1/1) was added to cell cultures without EPO. p Mean ±SD of quadruplicate wells. h Significantly different from control value: p Mean ±S D of triplicate wells. k Not significantly different from control value: p > 0.05.
179
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
Control (0 m il) Control (0 mU) +anti-EPO
Fig. 4. Immunoaffinity chromatography of rHuEPO (A) and sheep plasma Step III EPO (B). • — • = EPO activity (U/fraction) estimated by reference to a standard titration curve using Step III E P O ;....... — absorbance at 280 nm. Two preparations of the stan dard EPO (10 U), dissolved in 1 ml of PBS containing 0.01% bovine scrum albumin, were applied on the immunoadsorbent column in which anti-EPO MoAb R2 fixed on Affi-Gcl 10 was packed. Arrows indicate the positions of change of buffers. The constituent of the buffers used is described in ‘Materials and Methods’. Fig. 5. Immunoaffinity chromatography of plasma from normal (A), anemic (B). and hypoxic (C) mice. • — • = EPO activity (U/
fraction) estimated by reference to a standard titration curve using sheep plasma Step III E P O ;....... = absorbance at 280 nm. Mouse plasmas (1 ml) were applied on the MoAb R2 immunoadsorbent col umn. Arrows indicate the positions of change of buffers. The constit uent of the buffers used is described in ‘Materials and Methods’.
Table 3. Binding of mouse plasma ECSA to immunoadsorbent column containing immobilized anti-EPO MoAb R2
Sample
EPO activity“ mU
557 (100%) Normal plasma0 Anemic plasma'1 1,351 (100%) I lypoxic plasma0 1,650 (100%)
Flow-through fraction
Bound fraction
EPO activity mU
recoveryb %
EPO activity mU
rccovcryb %
95 95 99
17.1 7.0 6.0
404 1.048 1,239
72.5 77.6 75.1
180
Sakata/Enoki/U eda
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
a Before fractionation procedures, the ECSA level of plasma samples was estimated by a multiple-dose parallel-line bioassay, in which sheep plasma Step III EPO was used as a standard. Results are expressed as international units of EPO activity. b Total activity applied per column was defined as 1(10% activity. c Plasma from normal mice (mean PCV, 0.44 l/l). d Plasma from bleeding-induced anemic mice (mean PCV. 0.28 l/l). 0 Plasma from hypoxic mice exposed to 0.35 atm for 4 days.
Fig. 6. Anti-EPO MoAb R2 neutralization of ECSA fractionated by means of immunoaffinity chromatography. Each active fraction of hypoxic mouse plasma (A) and sheep plasma Step III EPO (B) was assayed for ECSA in the presence (solid symbols) and absence (open symbols) of MoAb (0.9 pg/wcll). The 12th day FMI.Cs (3 x lOVwell) were used. Upper and lower shaded areas indicate the levels of colony formation in the presence and absence of Step III EPO (25 mU/well). respectively. Each symbol represents the triplicate mean.
Table 4. Neutralization of ECSA in immunoadsorbent column fractions by anti-F.PO MoAb R2 and/or R6 or rabbit antirHuEPO scrum
Mean number of colonics ± SD' no antibody
R2S
R6h
R2 + R6'
antiserum’
NRSk
Control (0 mU) rHuEOP (25 mU)
130 ±4 263 ± 5
—
—
-
-
-
Normal plasma3 Flow-through fraction1’ Bound fraction1
131 ±3'
128 ±3'
126 ±2'
130 ±4'
264 ±3
180 ±4 144 ± 2 163 ± 4
140 ± 2m 142 ± 2m 127 ± 4'
146 ± 2 m 143 ± 3 m 128 ± 2'
142 ±3™ 143 ±2™ 126 ± 3'
143 ± 4 m 143 ± 3m 130 ±4'
182 ±3 153 ±3 170 ± 3
Hypoxic plasmad Flow-through fraction Bound fraction
243 ±5 143 ±3 219 ±5
143±4m 142 ± 2 m 127 ±3'
149 ± 3 m 142 ±3"’ 129 ±3'
143 ± 3 ” 144 ± 3n 128 ± 41
145 ± 3 m 143 ± 2 m 131 ±2'
240 ±5 152 ±3 222 ±5
Anemic plasma1' Flow-through fraction Bound fraction
267 ±5 144 ± 2 231 ±5
143 ± 3 m 141 ± 3 m 127 ±2'
149 ± 2 m 142 ± 2m 129 ± 2‘
141 ± 3 n 145 ± 2 n 128 ± 4‘
146 ±4"’ 142 ± 2 m 131 ±3'
262 ± 6 154 ±3 233 ±4
181
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
The 12th day FMLCs (3 x lOVwell) were used. a Plasma (10 pi, 1/2-fold concentration) from normal mice (mean PCV, 0.47 1/1) was added to cell cultures without EPO. b Pooled flow-through fractions in immunoaffinity chromatography of 1 ml mouse plasma were adjusted to 2 ml of volume, and 10 ul of the concentrated fraction was added to cell cultures without EPO. c Pooled bound fractions in immunoaffinity chromatography of 1 ml mouse plasma were adjusted to 2 ml of volume, and 10 pi of the con centrated fraction was added to cell cultures without EPO. d Plasma (10 pi, 1/2-fold concentration) from mice exposed to 0.35 atm for 3 days was added to cell cultures without EPO. c Plasma (10 pi, 1/2-fold concentration) from bleeding-induced anemic mice (mean PCV, 0.25 1/1) was added to cell cultures without EPO. f Mean ± SD of triplicate wells. 8 0.9 ug of anti-EPO MoAb R2 was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. h 10 ug of anti-EPO MoAb R6 was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. ‘ 0.9 pg of anti-EPO MoAb R2 and 10 ug of anti-EPO MoAb R6 were simultaneously added to cell cultures with rHuEPO, intact plasma or fractionated plasma. ’ Rabbit anti-rHuEPO scrum (10 pi, 1/2-fold concentration), heat (56 °C, 30 min)-inactivated and dialyzed against Dulbecco’s PBS, was added to cell cultures with rHuEPO. intact plasma or fractionated plasma. k Normal rabbit serum (10 pi, 1/2-fold concentration), heat (56 °C, 30 min)-inactivated and dialyzed against Dulbecco’s PBS, was added to cell cultures with rHuEPO, intact plasma or fractionated plasma. 1 Not significantly different from control values: p > 0.1. m Significantly different from control values: p < 0.05.
Discussion Although several anti-EPO MoAbs with neutralizing ability have been reported from four laboratories so far [22, 26-29], only two of them, R2 [22] and anti-Ep-16 [27], were capable of efficiently neutralizing the biological ac tivity of mouse EPO. MoAb R2 and anti-Ep-16 effectively blocked the capacity of mouse EPO to stimulate the in corporation of [?Hl-thymidine into the DNA of FMLCs and of mouse spleen cells, respectively. Using MoAb R2, we have attempted to demonstrate the identity of mouse plasma ECSA with EPO. In the preliminary experiments, we demonstrated that MoAb R2 inhibits the CFU-E colo ny formation stimulated by standard EPO. This inhibition seems to be due to blockage of EPO binding to CFU-E by MoAb [22], The ECSA of both mouse plasma and crude prepara tion of sheep plasma was not completely bound to the immunoadsorbent column containing the immobilized MoAb R2 (fig. 4B, 5). This seems to be ascribable to the nonimmunoreactive nature of the ECSA present in the flow-through fractions, but not to technical reasons, be cause rHuEPO was completely bound to the column (fig. 4A) and the passing ECSA in the flow-through frac tions was not neutralized by MoAb R2 (fig. 6; table 4). The
182
Saka t a/E noki/U cda
passing ECSA was additive to standard EPO in stimulat ing CFU-E growth and was heat-stable and nondialyzable. Thus, the passing ECSA had some properties similar to EPO. However, the passing ECSA was neutralized neither by anti-rHuEPO scrum nor by MoAb R6 (table 4), sug gesting that the passing ECSA is not an immunochemically different EPO but a non-EPO stimulator. This is fur ther supported by the observation that the contents of the passing ECSA in anemic and hypoxic plasmas were almost equal to that in normal plasma (tables 3, 4). Similarly, it was reported that non-EPO stimulator(s) may contribute to the high ECSA levels of neonatal mouse plasma [18,20, 21]. Normal mouse serum was found to contain two differ ent CFU-E-stimulating factors, ESCF [16] and PSA [17], which both stimulated the growth of murine marrow CFU-E synergistically with EPO. In addition, ESCF was heat-labile [16] and had no effect on fetal mouse liver CFU-E [30]. Based on these properties, ESCF and PSA appear to be different from our passing ECSA. Growth hormones [6], thyroid hormones [7], steroids [8], IGF II [10], cyclic nucleotides [12], and activin A [14] were found to potentiate colony formation of CFU-E in the presence but not the absence of EPO. On the other hand, insulin [11] and hemin [13] were shown to stimulate CFU-E formation independently of EPO. However, both substances are probably dialyzable in view of their smaller molecular sizes than 13,000 (MW cutoff size of dialysis membrane). Accordingly, it seems unlikely that the several candidates above stated represent the passing ECSA. Fagg [15] found a heat-stable and nondialyzable EpLA. which could be dis tinguished from EPO on the basis of affinity for concanavalin A and different molecular weight, in mouse spleen cell conditioned medium. This EpLA [31] and IGF I [9] were found to stimulate CFU-E from FMLCs in the ab sence of and nonsynergistically with EPO. IGF I present in scrum is known to be bound to specific serum proteins and to exist as 50,000- and 150,000-dalton complexes [32], Thus, EpLA and bound IGF I have some biological and bio chemical properties similar to those of the passing ECSA. Therefore, the passing ECSA may be EpLA or IGF I. Fur ther biochemical characterization of the passing ECSA is required for its identification. As shown in figure 1 and tables 1 and 4, the relatively high level of CFU-E colony formation was observed in the control cultures without any exogenous supplementation of EPO, in agreement with the earlier results [33,34], This spontaneous colony formation was not affected by adding anti-EPO MoAb (tabic 1). Therefore, as described by Cole et al. [33], it seems most likely that these colonies were de rived from CFU-Es which had been already triggered by
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
The passing ECSA had some biological and biochem ical properties similar to EPO. This ECSA showed a doseresponse curve parallel to the standard EPO curve, addi tive effect with standard EPO on the colony formation, and heat (56 °C for 30 min)-stable and nondialyzable na tures [data not shown]. To examine whether the passing ECSA is EPO unrcactive to MoAb R2 or non-EPO stim ulator, a neutralization study was carried out using rabbit anti-rHuEPO serum. MoAb R2 and MoAb R6 with neu tralizing activity against mouse EPO [22]. MoAb R2 and R6 were found to recognize different antigenic determi nants [22], R2 and/or R6 or antiserum did not neutralize the ECSA in the pooled flow-through fractions for nor mal, hypoxic and anemic plasmas at all, but did completely neutralize both the ECSA in the pooled bound fractions for these plasmas and rHuEPO (table 4). NRS added as control did not inhibit the colony formation in any cul tures. It should be noted that the colony level in the cul tures with both intact plasmas and MoAbs or antiserum corresponds to that in the cultures with the flow-through fractions. These results suggest that the passing ECSA is the non-EPO stimulator responsible for incomplete neu tralization of intact plasma ECSA by MoAbs or antiserum.
EPO in the fetal circulation. Another possible explanation might be that IGF I [35], which is present in FCS, stim ulates the colony formation. However, IGF I in FCS ap pears to play a minor role in the spontaneous colony for mation, because just a slight reduction in the spontaneous colony formation was caused by decreasing the concentra tion of FCS from 30 to 10% in our culture system [data not shown]. This study provides evidence that mouse plasma ECSA detectable in FMLC cultures is mainly due to EPO. First, most of the plasma ECSA was neutralized by MoAbs against EPO or anti-rHuEPO serum (fig. 2; tables 1, 4). This neutralization could be reversed by excess of EPO or anemic plasma (fig. 3) or by preincubation of MoAb with anti-M-IgG (table 1). These results suggest that the inhibi tion by MoAb is caused by immunospccific binding of MoAb to EPO. In fact, MoAb • EPO complex appears to have little or no affinity to EPO receptor [16], Second, most of the plasma ECSA was bound to the immunoadsorbent column (fig. 5), and the retained ECSA was totally blocked by MoAbs or anti-rHuEPO scrum (fig. 6A; table 4). Moreover, plasmas from mice under erythropoietic stimulation contained larger amounts of the retained ECSA compared with normal mouse plasma (fig. 5; tables
3, 4). In contrast, plasma from transfusion-induced poly cythemic mice contained no or decreased detectable ECSA [data not shown]. Such quantitative changes of the ECSA are comparable to a well-known physiologic re sponse of EPO. The circulating EPO levels arc normally increased by tissue hypoxia and decreased by tissue hyperoxia. Third, mode of action of the ECSA in nonfractionated plasma was similar to that of EPO. The plasma ECSA and standard EPO showed parallelism in the doseresponse curves (fig. 1) and additivity with each other in their function. From these results, CFU-E assay method using FMLCs is considered to be valid for mouse plasma EPO assay. However, it should be noted that a small por tion of the detected ECSA may be due to non-EPO stim ulator.
Acknowledgements We thank Dr. J. Morimoto for his generous gifts of anti-MMTV MoAb. This work was supported in part by grants-in-aid for Scientific Research (grant No. 01570048) and for Encouragement of Young Sci entists (grant No. 01770062) from the Ministry of Education, Science and Culture of Japan.
References 8 Singer JW. Samuels AI. Adamson JW: Steroids and hematopoiesis. I. The effect of steroids on in vitro erythroid colony growth: Structure ac tivity relationships. J Cell Physiol 1976:88:127134. 9 Kurtz A. Jelkmann W. Bauer C: A new candi date for the regulation of erythropoiesis. Insu lin-like growth factor I. FF.BS Lett 1982:149: 105-108. 10 Dainiak N. Kreczko S: Interactions of insulin, insulinlike growth factor II. and platelet-de rived growth factor in erythropoietic culture. J Clin Invest 1985:76:1237-1242. 11 Kurtz A. Jelkmann W. Bauer C: Insulin stim ulates erythroid colony formation independ ently of erythropoietin. Br J Haematol 1983:53: 311-316. 12 Brown JE, Adamson JW: Modulation of in vitro erythropoiesis: Enhancement of erythroid colony growth by cyclic nucleotides. Cell Tissue Kinet’1977:10:289-298. 13 Porter PN. Meints RH. Mesner K: Enhance ment of erythroid colony growth in culture by hernin. Exp Hematol 1979:7:11-16.
14 Yu J. Shao L. Lemas V. Yu AL. Vaughan J. Rivier J. Vale W: Importance of FSH-releasing protein and inhibin in erythrodifferentiation. Nature 1987;330:765-767. 15 Fagg B: Is erythropoietin the only factor which regulates late erythroid differentiation? Na ture 1981:289:184-186. 16 Blanchet JP. Arnaud S, Samarut J. Bouabdelli M: A factor present in normal mouse serum stimulates late erythroid precursor prolifera tion. Fxp Hematol 1984;12:595-602. 17 Udupa KB. Lipschitz DA: Proerythroblast stimulating activity: Its purification from mouse serum and its effect on mouse erythroid cell proliferation in vitro. Br J Haematol 1988:69:157-162. 18 Widness JA, Sanengen T, I laga P. Clemons GK, Myhre K. Halvorsen S: Correlation of plasma erythropoiesis stimulating factor(s) and immunoreactive erythropoietin levels during rapid growth in the mouse. Acta Physiol Scand 1989:136:527-533.
183
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
1 Hâgâ I’. Falkanger B: In vitro assay for erythro poietin: Erythroid colony formation in methyl cellulose used for the measurement of ery thro poietin in plasma. Blood 1979:53:1172-1181. 2 Mcberg A. Hâgâ 1’, Johansen M: Plasma eryth ropoietin levels in mice during the growth pe riod. Br J Haematol 1980:45:569-574 3 Hâgâ P, Kristiansen S: Role of the kidney in foetal erythropoiesis: Erythropoiesis and erythropoietin levels in newborn mice with re nal agenesis. J Embryol Exp Morph 1981; 61:165-173. 4 Cutler RL. Johnson GR. Nicola NA: Charac terization of murine erythropoietin. Exp He matol 1985:13:899-905. 5 Choppin J. Egrié J. Casadevall N. Lacombe C. Muller O. Varet B: Biochemical analyses of murine erythropoietin from plasma and from cloned erythroleukemia cells. Exp Hematol 1987;15:171-176. 6 Golde DW, Bersch N. l.i CH: Growth hor mone: Species-specific stimulation of erythro poiesis in vitro. Science 1977:196:1112-1113. 7 Golde DW, Bersch N. Chopra IJ. Cline MJ: Thyroid hormones stimulate erythropoiesis in vitro. Br J Haematol 1977:37:173-177.
25 Gruber DF. Zucali JR, Mirand EA: Identifica tion of erythropoietin producing cells in fetal mouse liver cultures. Exp Hematol 1977:5:392398. 26 Sytkowski AJ. Donahue KA: Immunochemical studies of human erythropoietin using site-spe cific anti-peptide antibodies. J Biol Client 1987; 262:1161-1165. 27 Wognum AW, Lansdorp PM. Krystal G: Immu nochemical analysis of monoclonal antibodies to human erythropoietin. Exp Hematol 1990: 18:228-233. 28 D’Andrea AD, Szklut PJ. l.odish HE. Alder man EM: Inhibition of receptor binding and neutralization of bioactivity by anti-erythro poietin monoclonal antibodies. Blood 1990:75: 874-880. 29 Feldman L. Heinzcrling R. Hillam RP. Chern YE, Frazier JG, Davis Kl„ Sytkowski AJ: Four unique monoclonal antibodies to the putative receptor binding domain of erythropoietin in hibit the biological function of the hormone. Exp Hematol 1992:20:64-68. 30 Arnaud S. Blanche! JP: The erythropoietic stimulating activity of normal mouse serum: Effect on erythroid precursors from different haematopoietic tissues and suggested role of accessory cells. Scand J Haematol 1986:37:180188. 31 Fagg B. Roitsch CA: Characterization of eryth ropoietin-like activity from mouse spleen cells. J Cell Physiol 1986;126:1-9.
32 Daughaday WH. Ward AP. Goldberg AC. Trivedi B. Kapadia M: Characterization of soma tomedin binding in human scrum by ultracen trifugation and gel filtration. J Clin Endocrinol Metab 1982;55:916-921. 33 Cole RJ, Regan T. White SL. Check EM: The relationship between erythropoietin-depend ent cellular differentiation and colony-forming ability in prenatal haemopoietic tissues. J Embryol Exp Morph 1975:34:575-588. 34 Kanamaru A. Okamoto T. Hara H. Nagai K: Developmental changes in erythropoietin re sponsiveness of late erythroid precursors in mouse hemopoietic organs. Dev Biol 1982:92: 221-226. 35 Kurtz A. Hartl W, Jelkmann W, Zapf J. Bauer C: Activity in fetal bovine serum that stimulates erythroid colony formation in fetal mouse liv ers in insulin-like growth factor I. J Clin Invest 1985;76:1643-1648.
184
Sakata/Enoki/U cda
Mouse Plasma ECSA and A nti-EPO MoAb
Downloaded by: Leiden University Medisch Centrum 132.229.13.63 - 11/3/2018 1:56:56 PM
19 Sanengen T. Halvorscn S: Plasma erythropoiesis stimulating factor(s) in neonatal mice: In vitro dose response and chromatography stud ies. Pediatr Res 1987;21:148-151. 20 Sanengen T. Myhre K. Halvorscn S: Erythro poietic factors in plasma from neonatal mice. In vivo studies by the exhypoxic polycythacmic mice assay for erythropoietin. Acta Physiol Seand 1987;129:381-386. 21 Sanengen T, Clemons GK, Halvorscn S, Widncss JA: Immunoreactive erythropoietin and erythropoiesis stimulating factor(s) in plasma from hypertransfused neonatal and adult mice. Studies with a radioimmunoassay and a cell culture assay for erythropoietin. Acta Physiol Scand 1989:135:11-16. 22 Goto M. Murakami A. Akai K. Kawanishi G. Ueda M. Chiba H. Sasaki R: Characterization and use of monoclonal antibodies directed against human erythropoietin that recognize different antigenic determinants. Blood 1989: 74:1415-1423. 23 Iscovc NN. Siebcr F: Erythroid progenitors in mouse bone marrow detected by macroscopic colony formation in culture. F.xp Hematol 1975:3:32-43. 24 Goto M. Akai K. Murakami A. Hashimoto C, Tsuda F.. Ueda M. Kawanishi G. Takahashi N. Ishimoto A. Chiba H. Sasaki R: Production of recombinant human erythropoietin in mam malian cells: Host-cell dependency of the bio logical activity of the cloned glycoprotein. Bio/ Technology 1988;6:67-71.
