Development of RBC membrane antigen arrays for validating Blood

Subscriber access provided by University of South Dakota

Article

Development of RBC membrane antigen arrays for validating Blood Grouping Reagents Lu Yang, Yang Yu, Chunya Ma, Hongye Wang, Jiayu Dai, Hu Duan, Zhonglin Fu, Ping Wu, Deqing Wang, and Xiaobo Yu J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00370 • Publication Date (Web): 16 Aug 2018 Downloaded from http://pubs.acs.org on August 22, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Development of RBC membrane antigen arrays for validating Blood Grouping Reagents Lu Yang1, Yang Yu1, Chunya Ma1, Hongye Wang2 , Jiayu Dai 2, Hu Duan2, Zhonglin Fu2, Ping Wu2, Deqing Wang1* and Xiaobo Yu2* 1

Department of Blood Transfusion, Chinese PLA General Hospital, Beijing,

100853，China. 2

State Key Laboratory of Proteomics, Beijing Proteome Research Center,

National Center for Protein Sciences (PHOENIX Center, Beijing), Beijing Institute of Lifeomics, Beijing, 102206, China

*Correspondence should be addressed to X.Y. ([email protected]) and D.W. ([email protected])

Abstract Antibody reagents have been remained as a standard approach to characterize blood group (BG) antigens in clinic. The specificity and cross-reactivity of these BG antibodies are routine detected using the gel microcolumn assay (GMA). However, the GMA is neither specific nor sensitive, thus increasing the risk of improperly-matched RBC transfusions. In this work, we describe a bead-based RBC membrane antigen array to detect BG antibody-antigen binding with ~700-fold higher sensitivity and dynamic range than the GMA. RBC membrane antigen arrays were fabricated using 1

ACS Paragon Plus Environment

Journal of Proteome Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

fragmented RBC membranes highly enriched in BG panel antigens. The arrays were then used to screen the interactions of 15 BG reagents to three antigen panels. The majority of the antibody reactions (i.e., 86.7%; 39/45) aligned with those obtained with the GMA. The six cross-reactive, non-specific antibody reactions identified only by our arrays (i.e., 13.3%; 6/45) were confirmed by agglutination inhibition and genotyping assays. These results demonstrate that our RBC membrane antigen array has great potential in screening BG antibodies and improving the safety of RBC transfusions.

Keywords Red blood cell, Antigen array, Antibody, Blood group, Proteomics

2


Page 2 of 34

Page 3 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Red blood cell (RBC) transfusions can cause serious complications if the recipient and donor patients are improperly matched to one another1-6. Therefore, routine screening of the donor’s blood is performed prior to a RBC transfusion to detect the presence of unexpected antibodies, or antibodies that bind to antigens not included in the standard ABO blood group system. These non-ABO blood group (BG) antigens have been routinely characterized over the past decade with commercial BG reagents using the gel microcolumn assay (GMA) or tube assay7. Inaccurate serological results can occur when the non-specific binding of a BG antibody to an off-target antigen is not detected; these spurious results can then increase the risk that a RBC recipient will experience a hemolytic or allergic reaction during or after their transfusion8-12. Over the past decade, donor blood has been screened with commercial BG antibody reagents and the GMA. Reactive BG antibody-RBC antigen complexes are then identified because the gel matrix restricts their migration through the column. However, the GMA is low throughput (2-8 samples within 30 min) and does not provide quantitative information regarding the strength of the antibody-antigen interaction. It is also possible that the GMA has a lower sensitivity compared to fluorescent detection, thereby missing some antibody cross-reactivity events13, 14. To address the low sensitivity issue of the GMA, Barr et al. developed a novel colorimetric immunoassay by printing trisaccharide constructs from an 3



Page 4 of 34

antigen subset of the ABO blood group (A trisaccharide, A Type 1, A Type 2, A Type 3, A Type 4, B trisaccharide, B Type 1, B Type 2, acquired B trisaccharide) and Lewis(a) antigens onto a paper. Antibody binding of 47 monoclonal and 2 polyclonal

ABO

antibodies

were

then

visualized

using

an

alkaline

phosphatase-labeled anti-immunoglobulin antibody and its chromagenic substrate. The results revealed unexpected cross-reactivity of anti-A reagents to the B type 1 antigen

14

. While informative, this colorimetric assay is still low

throughput. A more sensitive assay with multiplexed detection ability is necessary

to

reveal

the

full

breadth

of

antibody

specificity

and

cross-reactivity14. In this work, we developed the first RBC membrane antigen array for BG antibody detection using the bead-based multiplexed immunoassay approach, which is a high throughput and sensitive proteomics technology that has been frequently used over the past decade in translational studies to profile serological proteins and detect tumor cytokines, markers and pathogens

15-20

.

Since the assay is amenable to automation, hundreds of clinical samples can be screened within a short period of time21. Our bead-based RBC array immobilizes fragments of RBC membranes that are highly-enriched with BG panel antigens to beads. Herein, we first optimized the fragmentation of RBC membranes and characterized their antigens. Second, we determined the optimal parameters for assay fabrication and evaluated its performance in detecting antibody binding. Finally, we used our antigen arrays to study the 4


Page 5 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cross-reactivity of 15 BG antibodies and compared those results with the GMA method.

2. Materials and methods 2.1. Reagents Sulfo-N-hydroxysuccinimide (NHS) and 1-Ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride (EDC) were purchased from Thermo Fisher Scientific (MA, USA). Na2HPO4 (100mM, pH 6.2) was purchased from Sinopharm (Shang Hai, China). MES (50 mM, pH 5.0) and bovine serum albumin (BSA) was purchased from Sigma-Aldrich (MO, USA). Tween-20 was purchased from Amresco (OH, USA). MagPlex beads were purchased from Luminex Corporation (TX, USA). The 96-well plates were purchased from Corning Inc. (NY, USA). The Alexa Fluor 488 conjugated goat anti-human IgM heavy chain secondary antibody and the Cy3 conjugated donkey anti-human IgG were purchased from Jackson ImmunoResearch (PA, USA). DG Gel® Coombs cards were purchased from Grifols (Barcelona, Spain). Anti-C、E、c、e、Jka、Jkb、M、N、S、s、Fya、K、k、Lea and P1 antibodies were purchased from Sanquin (The Netherlands).

2.2. Preparation of RBC membrane antigens containing antigen panels I, II and III First, 25 mL peripheral blood were collected from healthy human donors 5



expressing either BG antigen panels I, II or III, centrifuged to remove the supernatant, and the resulting RBCs mixed with 25 mL 1x physiological saline (i.e., “saline”). Second, the RBCs were lysed with 3 freeze-thaw cycles (i.e., -80 oC, 30 min and 37 oC, 20 min each). The lysed RBC membranes were centrifuged at 15,000 g for 20 min, and the supernatant was discarded. Third, the precipitated RBCs were washed three times with deionized water (ddH20) to remove free hemoglobin. Fourth, the membrane precipitates were further fragmented into small pieces by a hand homogenizer for 0, 2.5, 5, 10, 20, or 40 min. The resulting fragmented RBC membranes containing the corresponding I, II and III antigen panels were examined by NIKON ECLIPSE Ti microscopy (Tokyo, Japan) with a 60×objective lens while the area of the fragments was measured by NIKON NIS-Element software (Tokyo, Japan). All clinical samples were collected with the written informed consent under the approval of the Institutional Review Boards (IRBs) at the Chinese PLA General Hospital and Beijing Proteome Research Center. All experiments were performed according to the standards of declaration of Helsinki 22.

2.3. Agglutination inhibition test 50 µL of BG antibody diluted 20-fold in saline were mixed with 50 µL of prepared fragmented RBC membranes and incubated for 10 min at room temperature; 50 µL of saline without the BG antibody were used as the control. After centrifugation at 1000 g for 60 s, 25 µL of supernatant and 50 µL RBC 6


Page 6 of 34

Page 7 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


reagent were added into the gel microcolumn cards and centrifuged at 1000 g for 10 min. The agglutination reaction, or the precipitation of RBC-antibody complexes, was assigned a score from 0 to 12 by visual inspection, in which the score number is inversely proportional to the antigenic activity of antigens in fragmented RBC membranes 23.

2.4. Fabrication of RBC membrane antigen arrays The coupling of RBC membrane antigens to the carboxyl-modified magnetic beads (Luminex Inc. Austin, TX, USA) was performed using standard EDC/NHS conjugation chemistry21,

24

. Briefly, 100 µL containing 1.25×106

magnetic beads were washed with 100 µL ddH2O and activated with 80 µL of “activation buffer” (100 mM Monobasic Sodium Phosphate, pH 6.2). Then the activated beads were mixed with 10 µL 50 mg/ml Sulfo-NHS (Thermo Fisher Scientific, IL, USA) and 10 µL of 50 mg/ml EDC (Thermo Fisher Scientific, IL, USA). After incubation for 20 min at room temperature with gentle mixing, the beads were washed twice with 250 µL “coupling buffer” (50 mM MES pH 5.0)(Sigma-Aldrich, St. Louis, MO, USA) and resuspended in 100 µL coupling buffer. Then 45 µg of extracted RBC antigens were added to the bead solution and incubated for 2 h at room temperature. After washing the beads with PBST (PBS with 0.05% Tween-20) again, the antigen-coated beads were stored in 100 µL PBST and counted as previous described 25.

7



2.5. Detection of BG reagents using antigen arrays First, a 50 µL solution containing 2,500 antigen-coated beads was added to each well of a 96-well plate, followed by 50 µL of BG antibody reagents. After incubation for 1 h at 21 ℃ on a shaker, the beads were pulled down with a magnetic separator (Themo Scientific, USA) and washed twice with 100 µL PBST. The detection of BG antibody binding to the antigens on the beads was performed by adding 50 µL fluorophore-conjugated secondary antibody (4 μ g/mL) and incubating the mixture for 30 min at 21 ℃. After washing with PBST three times, the beads were suspended in 200 µL PBST solution and the fluorescent signal was detected using the BD LSRFortessaTM cell analyzer at 355 and 640 nm (San Jose, CA, USA). The generated curves of the bead-based assay across the different concentrations of anti-s antibody were fitted using logistic 5 model in Origin 8.5 software (OriginLab, MA, USA), where the lowest level of detection limit (LOD) was equal to the control signal plus two standard deviations26.

2.6. Gel microcolumn assay (GMA) To perform the GMA, 50 µL of RBC and 25 µL of BG reagents were sequentially added to the microcolumn gel cards, and incubated for 15 min at 37 ℃. After centrifugation at 1,030 rpm for 10 min, the score was observed and recorded as previously described 27.

8


Page 8 of 34

Page 9 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.7. Genotyping of BG antigens The genotyping of BG antigens was performed using the polymerase chain reaction-single specific primer (PCR-SSP) method according to the commercial kit from Tianjin Super Biotechnology Developing Co. Ltd (Tianjin, China) and as previously described 28. Briefly, the samples were collected from the same donors as the RBC membrane antigens with informed consent. The genomic DNA was extracted from the serum according to the protocol provided by the commercial kit from Tianjin Super Biotechnology Developing Co. Ltd (Tianjin, China). The DNA was diluted to 30-50 ng/μL. Then, 25μL DNA was mixed with 220μL dNTP-Buffer containing dNTP, gene specific primers and DNA polymerase, in which a 10 μ L mixture was used for PCR amplification at the temperature, time, and cycles that are denoted in the table below. The PCR products were examined with a 2.5% (w/v) DNA agarose gel. Table1. PCR procedure for the genotyping of BG antigens.

Step Temperature/Time

Cycles

1

96℃/2min

1

2

96℃/20s, 68℃/60s

5

3

96℃/20s, 62℃/45s, 72℃/30s 10

4

96℃/20s, 62℃/45s, 72℃/30s 15

5

72℃/3min

1

3. Results 3.1. Preparation and characterization of RBC membrane antigens 9



We developed a method to fragment RBC membranes with highly-enriched BG I, II and III panel antigens, as shown in Figure 1A. Peripheral blood was first collected from healthy donors whose RBCs express the corresponding BG antigen panels. The RBCs were extracted via centrifugation, lysed by 3 freeze-thaw cycles, and washed with deionized water to remove free hemoglobin. Finally, the lysed RBC membranes were further fragmented by a hand homogenizer. The homogenization time (0, 2.5, 5, 10, 20 and 40 min) was optimized by measuring the size of fragmented RBC membranes via microscopy. As shown in Figure 1B, a homogenization time of 0 to 5 min results in large and aggregated pieces. As the homogenization time increases from 0 to 10 min, fragment aggregation and size decreases. The distribution of fragmented membrane sizes was evaluated with the NIKON NIS-Element software (Figure 1C). Without homogenization, the size of the fragmented RBC membranes had a large distribution range from 0.01 μm2 to 46.97 μm2. There was no significant difference in fragment size or fragment size distribution with homogenization longer 10 min (N.S.: no significance) (Figure 1C). The antigenicity of the fragmented RBC membrane antigens was characterized with a competitive agglutination inhibition test with an antibody to the s antigen as a representative

23

. Fragmented RBC membranes

containing the s antigen or saline alone (i.e., negative control) were first mixed with the anti-s antigen antibody in solution (Figure 2A). RBC fragments with 10


Page 10 of 34

Page 11 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


high antigenicity therefore resulted in a low number of free antibodies available for the subsequent agglutination inhibition test containing untreated (i.e., unfragmented) RBCs. The amount of RBC aggregation due to a BG antibody-antigen reaction was then quantified using an agglutination inhibition score from 0 to 12, in which the score number was inversely proportional to the antigenicity of antigens in fragmented RBC membranes. The negative control sample, denoted with an “S” for “saline,” would therefore have the highest agglutination inhibition score (i.e., 12) since the antibodies had not been previously incubated with RBC membrane fragments. Figure 2B shows that agglutination inhibition score of the anti-s antigen antibody mixed with RBC membranes (denoted as “M”) decreased only when the RBCs were homogenized longer than 10 min. Put another way, longer homogenization resulted in the antibodies being able to bind to more antigens on the fragmented RBC membranes. These data are corroborated by Figure 2C, which shows that the agglutination inhibition score decreased from 10 to 8 when the homogenization time was extended to 20 min, but did not decrease further with additional homogenization. The results indicate that 20 min of homogenization is optimal for the preparation of BG antigens.

3.2. Development of RBC membrane antigen arrays Fragmented RBC membranes were first coupled to the carboxyl group of carboxyl group-coated magnetic beads. To optimize the amount of fragmented 11



RBC membranes and antibody dilutions to obtain the maximum signal-to-noise (S/N) ratio, we tested five different amounts of fragmented RBC membranes expressing the s antigen (5.625, 11.25, 22.5, 45 and 90 μg) with 1.25×105 beads as well as three different dilutions of anti-s antibodies spanning two orders of magnitude. The fluorescent signal intensity, which is proportional to the amount of antibody binding, increases with more antigens and lower antibody dilution (i.e., more antibody)(Figure 3B). The S/N ratio was then calculated using the PBS buffer as the background control. The S/N ratio increased when the number of membrane antigens increased and with lower antibody dilution (Figure 3C). The highest S/N ratio occurred with 45 μg of membrane antigens and a 1:10 dilution of the anti-s antibody; these conditions were then used in the following experiments. Two antibody-antigen incubation conditions, 21 ℃ for 1 h and 4 ℃ overnight, were then tested and compared to determine the parameter that would result in the highest S/N ratio. The antigen arrays generated signals following both incubation conditions using the anti-s antigen antibody, but the S/N ratios of antibody to buffer control were all higher at 21 ℃ for 1 h than at 4 ℃ overnight (Figures 3E and 3F). These results may be explained, at least in part, to the significantly lower bead recovery following the overnight incubation at 4 ℃ than with the 1 h incubation at 21 ℃ (Figure 3D). Thus, the shorter incubation at 21 ℃ is ideal. The performance of our RBC fragment antigen arrays was then compared 12


Page 12 of 34

Page 13 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


with the GMA, which is routinely used to characterize BG antibody specificity and cross-reactivity. First, we measured the fluorescent signal intensity across 8 different concentrations of anti-s and anti-B antibodies in a 3-fold dilution series spanning ~ 3 orders of magnitude using our antigen array and the GMA (Figures 4A and 4B). The signal intensity increases with increasing concentrations of anti-s and anti-B antibodies in both the antigen arrays (left y-axis, mean fluorescent intensity = MFI) and the GMA (right y-axis, score). The lowest limit of detection (LOD) and dynamic range were then calculated for the bead-based array using the cut-off of control signal plus 2× standard deviations (Figure 4C)(Materials and Methods). The GMA’s LOD for the anti-s and anti-B was 1.1 x 10-2 and 1.23 x 10-3 (i.e., diluted antibody), respectively. The LOD of the antigen array using the same antibodies was 698.81- and 791.39-fold lower than the GMA, with an LOD for anti-s and anti-B of 1.59 x 10-5 and 1.56 x 10-6, respectively. The antigen array also had a wider dynamic range for detecting the anti-s and anti-B antibodies than the GMA. The GMA had a dynamic range of < 1 order of magnitude for the anti-s antibody (1.1 x 10-2 to 1.0 x 10-1) while the dynamic range of the array for the same antibody was > 3 orders of magnitude (1.59 x 10-5 to 1.0 x 10-1). The GMA had a dynamic range of < 1 order of magnitude for the anti-B antibody (1.23 x 10-3 to 1.10 x 10-2). The array, on the other hand, had a dynamic range of > 3 orders of magnitude (1.56 x 10-6 to 1.10 x 10-2) for the anti-B antibody. For the anti-s and anti-B antibodies, the array had a 698.81- and 791.39-fold 13



Page 14 of 34

LOD and wider dynamic range than the GMA, respectively (Figure 4C). The cross-reactivity of our bead-based assay was tested by analyzing the binding of the anti-s and anti-B antibodies to their corresponding target and the negative control, bovine serum albumin (BSA), across the range of antibody concentrations (Supplemental Figure S1). The antibodies bound specifically to their target, but did not bind to the negative control. The reproducibility of our bead-based assay was determined by repeating the anti-B antibody detection within an experiment and across two independent experiments (Supplemental Figure S2). The average intra-CV and inter-CV were 3.32% and 4.34%, respectively. Finally, the S/N ratios of the RBC arrays using an anti-B antibody were similar for experiments performed in 2016 and 2018 using the same membrane

proteins

stored

at

-80

o

C

in

solution

(paired

t-test,

p=0.9487)(Supplementary Table S1). These results demonstrate that our RBC membrane antigen array is reproducible and robust.

3.3. Evaluation of the specificity and cross-reactivity of BG reagents using antigen arrays Using RBC membrane antigen arrays containing three BG antigen panels (I, II and III), we screened 15 BG reagents in triplicate with 18 BG antigens (Rh-hr, Kidd, MNS, Duffy, Kell, Lewis and P), and then compared the results with the GMA (Figure 5A). Antibody-antigen binding (i.e., signal intensity) was determined using a PBS control and t-test analysis with a p-value 98.80% homology to the true antigen targets were present (Figure 6A-C)(Supplementary Figure S3). Since antibodies generally recognize epitopes containing only 4 - 6 amino acids, these results suggest that the BG antibodies likely cross-reacted with non-specific targets having high homology to their true antigens (Figure 6D)

31

. It further underscores the

importance of assessing the cross-reactivity of antibodies. Documented BG 17



Page 18 of 34

mutations are listed in the Blood Group Antigen Gene Mutation Database (https://www.ncbi.nlm.nih.gov/gv/mhc/xslcgi.cgi?cmd=bgmut/home). Like the GMA and agglutination inhibition assay, our antigen arrays also rely on using RBCs with unpurified BG antigens. As such, the BG antibodies may cross-react with non-BG proteins or molecules on the RBC membrane. While these reactions could still identify potentially dangerous cross-reactivity from RBC transfusions, it may be possible to create the arrays using purified BG antigens that are expressed in mammalian cells

32

. Doing so, however, is

not as straight forward as one may believe. Almost all of the BG antigens are membrane proteins with highly complex structures and antigenic epitopes that can include multi-pass membrane proteins and specific antigenic sugar moieties, respectively33, 34. In addition, membrane proteins that are produced in vitro are difficult to display properly. To date, only several BG antigens have been able to be expressed or synthesized successfully in vitro 13, 14, 35.

5. Conclusion We

developed

a

multiplexed

bead-based

immunoassay

using

self-prepared RBC membrane fragments, which has a significantly higher sensitivity, dynamic range, and quantitative capability compared to the GMA assay that is currently being used in the clinic to characterize BG reagents. The results demonstrate that our reproducible and robust RBC membrane antigen array has great potential in screening BG reagents and, consequently, 18


Page 19 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


improving the safety of RBC transfusions.

Abbreviations RBC, Red Blood Cell; BG, Blood group

Supporting Information Supplemental Table S1. Comparing the stability of bead-based arrays using self-prepared RBC antigens.

Supplemental Figure S1. Analysis of the specificity of anti-B and anti-s antibodies to their targets using bead-based arrays. Bovine serum albumin (BSA) is used as a negative control.

Supplemental

Figure

S2.

Analysis

of

assay

reproducibility

on

bead-based arrays. The intra-CV (%) was calculated within an experiment and the inter-CV (%) was calculated between two independent experiments.

Supplemental Figure S3. The alignment of protein sequences in the same blood group.

Acknowledgments This work was supported by the National Program on Key Basic Research 19



Page 20 of 34

Project (2017YFC0906703, 2018ZX09733003 and 2018YFA0507503), the National Natural Science Foundation of China (81673040), the State Key Laboratory

of

Proteomics

(SKLP-O201504,

SKLP-O201703

and

SKLP-K201505) and Capital’s Funds for Health Improvement and Research (2018-2-4034) to X.Y. The authors gratefully acknowledge the Flow Cytometer and the Microscope facilities in the PHOENIX Center as well as the Department of Blood Transfusion at Chinese PLA General Hospital for their support. We also thank Dr. Brianne Petritis for her critical review and editing of this manuscript.

Authorship Contribution L.Y. executed experiments and wrote the paper; Y.Y. and C.M. provided patient samples; H.W., J.D. and H. D. performed statistical analysis; Z.F. and P.W. executed experiments; X. Y. and D. W. designed and supervised all experiments and wrote the paper. All authors approved the final manuscript.

Competing interests The authors declare that no competing interest exists.

References 1.

Elenga, N.; Mialou, V.; Kebaili, K.; Galambrun, C.; Bertrand, Y.; Pondarre, C., Severe neurologic

complication after delayed hemolytic transfusion reaction in 2 children with sickle cell anemia: significant diagnosis and therapeutic challenges. J Pediatr Hematol Oncol 2008, 30, (12), 928-30. 2.

Koelewijn, J. M.; Vrijkotte, T. G.; van der Schoot, C. E.; Bonsel, G. J.; de Haas, M., Effect of

screening for red cell antibodies, other than anti-D, to detect hemolytic disease of the fetus and 20


Page 21 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


newborn: a population study in the Netherlands. Transfusion 2008, 48, (5), 941-52. 3.

Chadebech, P.; Habibi, A.; Nzouakou, R.; Bachir, D.; Meunier-Costes, N.; Bonin, P.; Rodet, M.;

Chami, B.; Galacteros, F.; Bierling, P.; Noizat-Pirenne, F., Delayed hemolytic transfusion reaction in sickle cell disease patients: evidence of an emerging syndrome with suicidal red blood cell death. Transfusion 2009, 49, (9), 1785-92. 4.

Gaines, A. R.; Lee-Stroka, H.; Byrne, K.; Scott, D. E.; Uhl, L.; Lazarus, E.; Stroncek, D. F.,

Investigation of whether the acute hemolysis associated with Rh(o)(D) immune globulin intravenous (human) administration for treatment of immune thrombocytopenic purpura is consistent with the acute hemolytic transfusion reaction model. Transfusion 2009, 49, (6), 1050-8. 5.

Kang, M. G.; Lim, Y. A.; Lee, K. M., [A hemolytic transfusion reaction due to Anti-Ku antibody in a

patient with Knull phenotype: the first case in Korea]. Korean J Lab Med 2009, 29, (3), 238-42. 6.

Koelewijn, J. M.; Vrijkotte, T. G.; de Haas, M.; van der Schoot, C. E.; Bonsel, G. J., Risk factors for

the presence of non-rhesus D red blood cell antibodies in pregnancy. Bjog 2009, 116, (5), 655-64. 7.

Novaretti, M. C.; Jens, E.; Pagliarini, T.; Bonifacio, S. L.; Dorlhiac-Llacer, P. E.; Chamone, D. A.,

Comparison of conventional tube test technique and gel microcolumn assay for direct antiglobulin test: a large study. J Clin Lab Anal 2004, 18, (5), 255-8. 8.

Colwill, K.; Renewable Protein Binder Working, G.; Graslund, S., A roadmap to generate

renewable protein binders to the human proteome. Nat Methods 2011, 8, (7), 551-8. 9.

Rockberg, J.; Lofblom, J.; Hjelm, B.; Uhlen, M.; Stahl, S., Epitope mapping of antibodies using

bacterial surface display. Nat Methods 2008, 5, (12), 1039-45. 10. Taussig, M. J.; Stoevesandt, O.; Borrebaeck, C. A.; Bradbury, A. R.; Cahill, D.; Cambillau, C.; de Daruvar, A.; Dubel, S.; Eichler, J.; Frank, R.; Gibson, T. J.; Gloriam, D.; Gold, L.; Herberg, F. W.; Hermjakob, H.; Hoheisel, J. D.; Joos, T. O.; Kallioniemi, O.; Koegl, M.; Konthur, Z.; Korn, B.; Kremmer, E.; Krobitsch, S.; Landegren, U.; van der Maarel, S.; McCafferty, J.; Muyldermans, S.; Nygren, P. A.; Palcy, S.; Pluckthun, A.; Polic, B.; Przybylski, M.; Saviranta, P.; Sawyer, A.; Sherman, D. J.; Skerra, A.; Templin, M.; Ueffing, M.; Uhlen, M., ProteomeBinders: planning a European resource of affinity reagents for analysis of the human proteome. Nat Methods 2007, 4, (1), 13-7. 11. Uhlen, M.; Bandrowski, A.; Carr, S.; Edwards, A.; Ellenberg, J.; Lundberg, E.; Rimm, D. L.; Rodriguez, H.; Hiltke, T.; Snyder, M.; Yamamoto, T., A proposal for validation of antibodies. Nat Methods 2016, 13, (10), 823-7. 12. Wang, D.; Yang, L.; Zhang, P.; LaBaer, J.; Hermjakob, H.; Li, D.; Yu, X., AAgAtlas 1.0: a human autoantigen database. Nucleic Acids Res 2017, 45, (D1), D769-D776. 13. Seltsam, A.; Blasczyk, R., Recombinant blood group proteins for use in antibody screening and identification tests. Curr Opin Hematol 2009, 16, (6), 473-9. 14. Barr, K.; Korchagina, E.; Ryzhov, I.; Bovin, N.; Henry, S., Mapping the fine specificity of ABO monoclonal reagents with A and B type-specific function-spacer-lipid constructs in kodecytes and inkjet printed on paper. Transfusion 2014, 54, (10), 2477-84. 15. Kricka, L. J.; Master, S. R.; Joos, T. O.; Fortina, P., Current perspectives in protein array technology. Ann Clin Biochem 2006, 43, (Pt 6), 457-67. 16. Wingren, C.; Borrebaeck, C. A., Antibody microarray analysis of directly labelled complex proteomes. Curr Opin Biotechnol 2008, 19, (1), 55-61. 17. Wulfkuhle, J. D.; Speer, R.; Pierobon, M.; Laird, J.; Espina, V.; Deng, J.; Mammano, E.; Yang, S. X.; Swain, S. M.; Nitti, D.; Esserman, L. J.; Belluco, C.; Liotta, L. A.; Petricoin, E. F., 3rd, Multiplexed cell signaling analysis of human breast cancer applications for personalized therapy. J Proteome Res 2008, 21



7, (4), 1508-17. 18. Alhamdani, M. S.; Schroder, C.; Hoheisel, J. D., Oncoproteomic profiling with antibody microarrays. Genome Med 2009, 1, (7), 68. 19. Qiu, J.; Hanash, S., Autoantibody profiling for cancer detection. Clin Lab Med 2009, 29, (1), 31-46. 20. Yu, X.; Schneiderhan-Marra, N.; Joos, T. O., Protein microarrays for personalized medicine. Clin Chem 2010, 56, (3), 376-87. 21. Broger, T.; Basu Roy, R.; Filomena, A.; Greef, C. H.; Rimmele, S.; Havumaki, J.; Danks, D.; Schneiderhan-Marra, N.; Gray, C. M.; Singh, M.; Rosenkrands, I.; Andersen, P.; Husar, G. M.; Joos, T. O.; Gennaro, M. L.; Lochhead, M. J.; Denkinger, C. M.; Perkins, M. D., Diagnostic Performance of Tuberculosis-Specific IgG Antibody Profiles in Patients with Presumptive Tuberculosis from Two Continents. Clin Infect Dis 2017, 64, (7), 947-955. 22. General Assembly of the World Medical, A., World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. The Journal of the American College of Dentists 2014, 81, (3), 14-8. 23. Boettcher, B., Correlation between human ABO blood group antigens in seminal plasma and on seminal spermatozoa. J Reprod Fertil 1968, 16, (1), 49-54. 24. Filomena, A.; Beiter, Y.; Templin, M. F.; Joos, T. O.; Schneiderhan-Marra, N.; Poetz, O., Bead-Based Peptide Arrays for Profiling the Specificity of Modification State-Specific Antibodies. Methods Mol Biol 2015, 1348, 251-65. 25. Yu, X.; Hartmann, M.; Wang, Q.; Poetz, O.; Schneiderhan-Marra, N.; Stoll, D.; Kazmaier, C.; Joos, T. O., microFBI: a microfluidic bead-based immunoassay for multiplexed detection of proteins from a microL sample volume. PLoS One 2010, 5, (10), pii: e13125. 26. Yu, X.; Lv, R.; Ma, Z.; Liu, Z.; Hao, Y.; Li, Q.; Xu, D., An impedance array biosensor for detection of multiple antibody-antigen interactions. Analyst 2006, 131, (6), 745-50. 27. Finck, R.; Lui-Deguzman, C.; Teng, S. M.; Davis, R.; Yuan, S., Comparison of a gel microcolumn assay with the conventional tube test for red blood cell alloantibody titration. Transfusion 2013, 53, (4), 811-5. 28. Rink, G.; Scharberg, E. A.; Bugert, P., PCR with sequence-specific primers for typing of diallelic blood groups. Methods Mol Biol 2015, 1310, 71-81. 29. Weisbach, V.; Kohnhauser, T.; Zimmermann, R.; Ringwald, J.; Strasser, E.; Zingsem, J.; Eckstein, R., Comparison of the performance of microtube column systems and solid-phase systems and the tube low-ionic-strength solution additive indirect antiglobulin test in the detection of red cell alloantibodies. Transfusion medicine 2006, 16, (4), 276-84. 30. Bigalke, B.; Potz, O.; Kremmer, E.; Geisler, T.; Seizer, P.; Puntmann, V. O.; Phinikaridou, A.; Chiribiri, A.; Nagel, E.; Botnar, R. M.; Joos, T.; Gawaz, M., Sandwich immunoassay for soluble glycoprotein VI in patients with symptomatic coronary artery disease. Clin Chem 2011, 57, (6), 898-904. 31. Hebbes, T. R.; Turner, C. H.; Thorne, A. W.; Crane-Robinson, C., A "minimal epitope" anti-protein antibody that recognises a single modified amino acid. Mol Immunol 1989, 26, (9), 865-73. 32. Yu, X.; LaBaer, J., High-throughput identification of proteins with AMPylation using self-assembled human protein (NAPPA) microarrays. Nat Protoc 2015, 10, (5), 756-67. 33. Patnaik, S. K.; Helmberg, W.; Blumenfeld, O. O., BGMUT Database of Allelic Variants of Genes Encoding Human Blood Group Antigens. Transfus Med Hemother 2014, 41, (5), 346-51. 34. Mitra, R.; Mishra, N.; Rath, G. P., Blood groups systems. Indian J Anaesth 2014, 58, (5), 524-8. 35. Seltsam, A.; Wagner, F.; Lambert, M.; Bullock, T.; Thornton, N.; Scharberg, E. A.; Grueger, D.; 22


Page 22 of 34

Page 23 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Schneeweiss, C.; Blasczyk, R., Recombinant blood group proteins facilitate the detection of alloantibodies to high-prevalence antigens and reveal underlying antibodies: results of an international study. Transfusion 2014, 54, (7), 1823-30.

Figure legends Figure 1. Preparation of fragmented RBC membranes containing highly-enriched antigens. (A) depicts the workflow for RBC membrane fragmentation. (B) characterizes the RBC membrane fragments with varying homogenization times using microscopy. (C) shows the size distribution of fragmented RBC membranes prepared with different homogenization times. Scale bars represent 20 µm in all images. N.S. means that there is no statistical difference, p-value < 0.01 is *, p-value < 0.001 is * * * . The statistical analysis was performed using the statistical analysis software package SPSS17.0.

Figure 2. Characterization of the antigenicity of RBC membrane antigens using a competitive agglutination inhibition assay. (A) is the workflow of the agglutination inhibition assay used to characterize the antigenicity of the RBC membrane fragments. (B) shows the results of the agglutination inhibition assay using an anti-s antibody with intact RBCs displaying antigen panel I following incubation with RBC membrane fragments homogenized for 0, 2.5, 5, 10, 20, or 40 min. “M” and “S” represent samples that had been previously incubated with fragmented RBC membranes or only saline, respectively. (C) shows the relationship between homogenization time and the agglutination 23



inhibition score from Figure 2B.

Figure 3. Development of RBC membrane antigen arrays. (A) is the workflow of the bead-based RBC membrane antigen array preparation and antibody assay. (B) and (C) show the relationship between the mean fluorescent intensity (MFI) and the signal-to-noise (S/N) ratio of the beads coupled with different amounts of RBC membrane expressing s antigen, and probed with varying dilutions of anti-s antibody. PBS buffer was used as the negative control for the S/N ratio calculations. (D) compares the number of recovered beads following bead-antibody incubation at 4 oC overnight or at room temperature (21 oC) for 1 h. Bead measurement was performed using the Luminex instrument. (E) and (F) show the MFI of the antibody array following antibody-array incubation at 4 oC overnight or at 21 oC for 1 h, respectively. Numbers in red denote the S/N ratio where PBS buffer was used as the control.

Figure 4. Lowest limit of detection (LOD) and dynamic range comparison of the RBC membrane antigen array and gel microcolumn assay. （A） and (B) compares the antibody signal of anti-s and anti-B antibodies binding to their targets, respectively, obtained with the antigen array and GMA. Antigen array data are represented by the black line and “MFI” y-axis, whereas the GMA data are represented by the blue columns and “Score” y-axis. (C) 24


Page 24 of 34

Page 25 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


compares the LOD limit and dynamic range of the antigen array and GMA using anti-s and anti-B antibodies and their corresponding antigens.

Figure 5. Evaluation of the specificity and cross-reactivity of BG reagents. (A) displays the results from the gel microcolumn assay for 15 BG antibodies against antigen panels I, II and III, where “AP” stands for “Antigen Panel.” The 15 antibodies target 18 antigens within the seven major BG families, Rh-hr, Kidd, MNS, Duffy, Kell, Lewis, and P. “+” and “-” represent antibody binding and no binding, respectively. (B) shows antibody reactions that were the same using the antigen array and gel microcolumn assay. (C) shows antibody reactions that were not the same using the antigen array and GMA. “+” and “-” represent antibody binding and no binding, respectively, using the GMA. Black and red labels denote results that were similar and not similar, respectively, using the antigen array and GMA.

Figure 6. Validation of cross-reactive antigens. (A) is the validation of cross-reactive antigens using the competitive agglutination inhibition test in which a higher score denotes higher antibody reactivity. “M” and “S” represent samples that had been previously incubated with fragmented RBC membranes or only saline, respectively. (B) shows the representative genotyping results that antigen panels I, II and III have varying levels of JKa and JKb antigens as determined with PCR-based genotyping and DNA agarose gels. (C) 25



summarizes the genotyping results for antigens C, c, E, e, M, N, S, s, JKa, and JKb. Labels in “black” and “red” represent similar and dissimilar results, respectively, from the antigen array and gel microcolumn assay. (D) compares the protein sequences of antigens in the same BG. The amino acids highlighted in yellow are those that are different between the similar antigens.

26


Page 26 of 34

Page 27 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For TOC Only

27



Figure 1. Preparation of fragmented RBC membranes containing highly-enriched antigens. (A) depicts the workflow for RBC membrane fragmentation. (B) characterizes the RBC membrane fragments with varying homogenization times using microscopy. (C) shows the size distribution of fragmented RBC membranes prepared with different homogenization times. Scale bars represent 20 µm in all images. N.S. means that there is no statistical difference, p-value < 0.01 is *, p-value < 0.001 is * * * . The statistical analysis was performed using the statistical analysis software package SPSS17.0. 170x143mm (300 x 300 DPI)


Page 28 of 34

Page 29 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Characterization of the antigenicity of RBC membrane antigens using a competitive agglutination inhibition assay. (A) is the workflow of the agglutination inhibition assay used to characterize the antigenicity of the RBC membrane fragments. (B) shows the results of the agglutination inhibition assay using an anti-s antibody with intact RBCs displaying antigen panel I following incubation with RBC membrane fragments homogenized for 0, 2.5, 5, 10, 20, or 40 min. “M” and “S” represent samples that had been previously incubated with fragmented RBC membranes or only saline, respectively. (C) shows the relationship between homogenization time and the agglutination inhibition score from Figure 2B. 189x123mm (300 x 300 DPI)



Figure 3. Development of RBC membrane antigen arrays. (A) is the workflow of the bead-based RBC membrane antigen array preparation and antibody assay. (B) and (C) show the relationship between the mean fluorescent intensity (MFI) and the signal-to-noise (S/N) ratio of the beads coupled with different amounts of RBC membrane expressing s antigen, and probed with varying dilutions of anti-s antibody. PBS buffer was used as the negative control for the S/N ratio calculations. (D) compares the number of recovered beads following bead-antibody incubation at 4 oC overnight or at room temperature (21 oC) for 1 h. Bead measurement was performed using the Luminex instrument. (E) and (F) show the MFI of the antibody array following antibody-array incubation at 4 oC overnight or at 21 oC for 1 h, respectively. Numbers in red denote the S/N ratio where PBS buffer was used as the control. 175x131mm (300 x 300 DPI)


Page 30 of 34

Page 31 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Lowest limit of detection (LOD) and dynamic range comparison of the RBC membrane antigen array and gel microcolumn assay. （A） and (B) compares the antibody signal of anti-s and anti-B antibodies binding to their targets, respectively, obtained with the antigen array and GMA. Antigen array data are represented by the black line and “MFI” y-axis, whereas the GMA data are represented by the blue columns and “Score” y-axis. (C) compares the LOD limit and dynamic range of the antigen array and GMA using antis and anti-B antibodies and their corresponding antigens. 143x102mm (300 x 300 DPI)



Figure 5. Evaluation of the specificity and cross-reactivity of BG reagents. (A) displays the results from the gel microcolumn assay for 15 BG antibodies against antigen panels I, II and III, where “AP” stands for “Antigen Panel.” The 15 antibodies target 18 antigens within the seven major BG families, Rh-hr, Kidd, MNS, Duffy, Kell, Lewis, and P. “+” and “-” represent antibody binding and no binding, respectively. (B) shows antibody reactions that were the same using the antigen array and gel microcolumn assay. (C) shows antibody reactions that were not the same using the antigen array and GMA. “+” and “-” represent antibody binding and no binding, respectively, using the GMA. Black and red labels denote results that were similar and not similar, respectively, using the antigen array and GMA. 215x135mm (300 x 300 DPI)


Page 32 of 34

Page 33 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. Validation of cross-reactive antigens. (A) is the validation of cross-reactive antigens using the competitive agglutination inhibition test in which a higher score denotes higher antibody reactivity. “M” and “S” represent samples that had been previously incubated with fragmented RBC membranes or only saline, respectively. (B) shows the representative genotyping results that antigen panels I, II and III have varying levels of JKa and JKb antigens as determined with PCR-based genotyping and DNA agarose gels. (C) summarizes the genotyping results for antigens C, c, E, e, M, N, S, s, JKa, and JKb. Labels in “black” and “red” represent similar and dissimilar results, respectively, from the antigen array and gel microcolumn assay. (D) compares the protein sequences of antigens in the same BG. The amino acids highlighted in yellow are those that are different between the similar antigens. 164x128mm (300 x 300 DPI)



79x36mm (300 x 300 DPI)


Page 34 of 34

Development of RBC membrane antigen arrays for validating Blood

Recommend Documents