- Email: [email protected]
Monoclonal antibodies raised to human cells — specificity for pig leukocytes
Veterinary Immunology and Immunopathology 80 (2001) 175–186
Monoclonal antibodies raised to human cells — specificity for pig leukocytes K. Haversona,*, M. Baileya, C.R. Stokesa, A. Simonb, L. LeFlufya, G. Banfielda, Z. Chenc, E. Hollemweguerc, J.A. Ledbetterd a
Division of Molecular and Cellular Biology, University of Bristol, Langford BS40 5DU, UK b LEBAO, Hannover, Germany c Pharmingen, San Diego, CA, USA d Bristol-Myers Squibb, Seattle, WA, USA
Abstract A total of 27 monoclonal antibodies raised to human targets were included in the present Pig CD workshop. 14 of these had been tested in previous workshops and had been reported as crossreactive, a further 13 had been reported as cross-reactive during the Human Leukocyte Differentiation Antigens Workshop VI (HLDA VI) and/or by the donor (a commercial company submitting these mAb for validation by the workshop community). Of the 27 antibodies, three antibodies with previously reported reactivity for pig cells were eliminated from the workshop following preliminary tests due to lack of reactivity. Nine antibodies, although initially positive, gave inconsistent results during the course of the workshop. We found consistent reactivity for 15 antibodies. However, the cellular distribution of the target molecules on pig and human cells was shown to be different for three of these antibodies. These findings have important implications for the usefulness of these antibodies as research tools in the pig. # 2001 Elsevier Science B.V. All rights reserved. Keywords: Swine; Cluster of differentiation (CD); Antigens; Monoclonal antibodies (mAb); Cross-reactivity; CD11b; CD11c; CD50; CD14; CD40; CD47; CD49; CD92; CD93
1. Introduction In spite of the success of the previous and present International Swine CD workshops in the characterization of cluster determinants on pig leucocytes, there is still a shortage of reagents to cell surface molecules relative to those available to our colleagues working *
Corresponding author. Tel.: þ44-117-928-9289; fax: þ44-44-928-9505. E-mail address: [email protected] (K. Haverson). 0165-2427/01/$ – see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 0 1 ) 0 0 2 8 5 - 9
176
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
with human or rodent cells. The preparation of such antibodies is time-consuming and there is therefore a constant search for well-characterization antibodies raised to human cells with cross-reactivity for other species. Although there have been frequent reports of such antibodies, few systematic studies have been conducted into the quality of such reagents for use in the pig. Two main questions need to be answered: The first question concerns the consistent reactivity of such reagents, reduced affinity to other than the original target species may cause problems in this respect. If consistent cross-reactivity is observed, then we need to know if the ligands recognized are homologous to those in humans and if their cellular distribution is the same.
2. Methods Table 1 is a list of all putatively cross-reactive antibodies donated/submitted to this workshop. All antibodies were tested and titrated for optimal use in preliminary work Table 1 List of anti-human mAb submitted to this workshop, with name of donors Number
mAb
Specificity
Donor
14 15 16 23 24 25 33 34 35 36 37 38 39 40 41 42 43 44 45 143 144 145 146 147 148 149 199
TMG6-5 S-Hcl3 BiG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 HH2 FN4 AHN16 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
CD11b CD11c CD14 CD32 CD35 CD40 CD47 CD47 CD49b CD49d CD49f CDw50 CwD75a CDw78a CD63a CD79b CD80 CDw92 CD93 CD11b CD39 CD2 CD21 HLA-DP, DQ, DR HLA-DQ HLA-DR CD49e
Ando Becton Dickenson Schutt Pulford Stein Ledbetter Horejsi Pharmingen Pharmingen Becton Dickenson Pharmingen Sanchez–Madrid Funderud Funderud Skubitz Pharmingen Pharmingen Majdic Majdic Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Van Kooyk
a
Withdrawn after negative preliminary tests.
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
177
before the beginning of the first round by the author, leading to the elimination of three antibodies from the workshop. All remaining antibodies were tested by all workshop participants on many target cells. However, for the purpose of this comparison, only nine selected datasets for pig cells and four for human cells from three different laboratories, with comparable targets and gating of the three major leukocyte populations of peripheral blood, were used for this comparative investigation. Target cells A–D were from slaughter-weight Landracecrossbred pigs, target cells E and F from two normal adult humans, the remaining targets were data from the first round (Appendix). Antibodies were tested in single colour flow cytometry, gating for both species was based on their FCS/SSC characteristics and cells were assigned to the lymphocyte, monocyte and granulocyte gate on this basis. However, it was apparent that the FCS/SSC characteristics of pig monocytes and lymphocytes were frequently overlapping (using the expression of SWC3 as definitive monocyte marker, not shown), therefore, the ‘lymphocyte’ gate would contain small monocytes and the ‘monocyte’ gate would contain large lymphocytes. For this reason, the co-expression of the test antibody with the specific antibodies to CD3, CD21 and SWC3 was also investigated for three selected antibodies, using Ficoll-separated PBMC as targets.
3. Results Tables 2 and 3 show the percentage of positive cells for each antibody in the lymphocyte, monocyte and granulocyte gates from pig (Table 2) and human (Table 3) PBMC, based on FCS/SSC characteristics, for all 24 antibodies. Fig. 1 summarises the mean reactivity of antibodies observed for pig (Fig. 1a) and human (Fig. 1b) cells. See also Appendix for further data. A detailed examination of the reactivities of each antibody revealed that antibodies could be grouped into three main categories. 1. Antibodies with reactivity consistent with similar cellular distribution of the ligand for both species. Antibodies falling into category 1 were: #16 biG10 (CD14), #25 G28.5 (CD40), #33 MEM-122 and #34 CIKM (CD47), #36 L25 (CD49d), #44 VIM15 (CD92), #45 VIMD2 (CD93), #145 RPA-2.10 (CD2), #146 B-ly4 (CD21) as well as the anti-MHC class II antibodies #147 TU39, #148 TU169 and #149 TU36. The reactivity of some of these antibodies is also confirmed in the relevant section reports. 2. Antibodies with consistent reactivity, but identifying cellular subsets which appeared different for each species. Antibodies falling into category 2 were: #14 TMG6-5 (CD11b), #15 S-Hcl3 (CD11c) and #38 HP2/19 (CD50). 3. Antibodies which gave negative, inconsistent or very weak reactivity on pig cells (identified by standard deviations @ mean of positive cells) in comparison to those found on human cells. The remaining 10 antibodies fell into this category. One antibody appeared to have deteriorated even for use on human cells. The results for antibodies in categories 1 and 3 will not be shown in more detail. However, antibodies in category 2 were investigated in more detail in one experiment
178
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Table 2 Percentage of positive cells in gates (pig leukocytes)a Number
Specificity
mAb
A
B
C
D
T8
T9
T10 T11 T12 Mean S.D.
Pig lymphocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
12 33 1 0 3 22 ND 82 1 15 3 44 2 3 0 ND 0 3 36 9 16 8 15 1
7 32 ND ND 3 12 ND 69 1 0 1 ND 3 1 0 ND 1 2 31 6 21 6 10 1
18 19 5 0 7 20 ND 89 2 1 ND 42 5 1 1 ND 0 2 24 9 13 4 18 0
27 32 2 0 2 21 ND 79 4 42 1 ND 5 2 1 ND 5 14 54 19 59 28 46 2
27 20 20 72 73 37 94 97 80 76 13 21 6 4 18 22 6 5 34 11 12 30 22 2
11 5 4 4 4 43 100 89 4 75 29 52 11 4 5 18 4 13 89 7 79 40 80 71
6 5 1 0 0 7 33 44 8 11 8 26 50 2 23 7 1 3 44 7 8 2 4 2
2 2 0 0 0 7 78 67 1 47 4 54 31 0 15 8 0 0 65 9 26 5 14 1
8 2 0 17 0 5 90 60 49 68 69 59 23 0 34 13 0 0 64 7 33 2 4 0
13 17 4 12 10 19 79 75 17 37 16 42 15 2 11 13 2 5 49 9 30 14 24 9
9 14 7 25 24 14 27 17 28 31 23 14 16 2 12 6 2 5 21 4 24 14 25 23
Pig monocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36
20 34 25 1 4 33 ND 90 1 17 4 50 4 4 1 ND 0 6 32 11 26 13 22
15 41 ND 0 9 43 ND 81 2 9 3 ND 5 4 1 ND 3 6 26 8 40 20 31
26 30 20 0 15 43 ND 94 6 1 2 49 11 4 3 ND 0 3 13 12 34 8 36
37 35 15 1 4 44 ND 84 6 29 2 ND 12 4 3 ND 6 27 29 27 73 50 66
6 11 49 0 0 8 31 14 4 27 16 63 63 1 45 24 0 1 12 2 33 0 4
10 25 21 4 4 21 100 93 6 58 29 55 11 5 4 18 88 12 12 16 80 73 88
15 12 52 0 0 14 90 55 0 57 13 80 50 0 80 55 0 0 17 1 30 1 3
6 7 41 38 0 4 95 49 47 67 78 70 33 0 86 43 0 0 2 0 45 1 0
8 66 76 88 87 58 92 98 90 90 32 52 9 5 18 35 4 7 4 4 12 36 25
16 29 37 15 14 30 82 73 18 40 20 60 22 3 27 35 11 7 16 9 41 23 30
10 18 21 30 28 19 29 28 31 30 25 12 22 2 35 15 29 9 11 9 22 26 30
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
179
Table 2 (Continued ) Number
Specificity
mAb
A
B
C
D
T8
199
CD49e
SAM-1
2
3
2
4
7
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
46 4 ND 0 1 6 ND 81 1 0 3 ND 2 1 1 ND 13 1 1 1 1 1 1 1
51 4 57 0 1 10 ND 88 2 2 3 10 4 1 1 ND 1 1 3 1 2 2 3 1
53 7 77 1 2 7 ND 71 4 7 8 ND 4 3 6 ND 18 4 3 4 10 3 8 1
45 8 94 1 1 6 ND 71 1 6 3 35 2 3 1 ND 0 1 2 1 5 1 2 8
Pig granulocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e a
T9
T10 T11 T12 Mean S.D.
90
12
8
6
15
28
61 7 3 3 4 10 93 57 4 9 13 37 8 4 3 13 46 14 5 3 11 6 52 33
40 8 7 0 0 1 18 0 0 2 25 23 12 0 28 50 0 1 0 0 0 0 0 2
55 6 22 0 0 6 82 24 0 32 44 54 7 0 75 84 0 0 1 0 1 0 0 2
77 0 4 19 0 2 71 2 0 3 18 55 1 0 86 61 0 0 0 0 0 0 0 4
53 6 38 3 1 6 66 49 1 8 15 36 5 1 25 52 10 3 2 1 4 2 8 6
12 3 38 7 1 3 33 36 2 10 14 17 4 1 35 30 16 5 2 1 4 2 18 11
Gating based on FSC/SSC characteristics Leukocytes.
using gating based on FCS/SSC characteristics as well as the co-expression of three characteristic pig cell surface molecules, CD3, CD21 and SWC3. 3.1. #14, TMG6-5 (CD11b) CD11b (MAC-1) is found on human granulocytes, monocytes and NK cells. Fig. 2 shows that the distribution on pig cells is different. Whereas both human monocytes and granulocytes are strongly positive for this antibody (Fig. 2a), pig cells in all gates including granulocytes have both positive and negative subsets (Tables 2 and 3 and Fig. 2b). Additionally, the expression levels on the positive subsets are lower than those on human cells. A more detailed investigation, based on the co-expression of specific cell surface molecules on PBMC (Fig. 2c), shows that CD3þ T-cells and CD21þ B-cells were negative for the TMG6-5 ligand. Additionally and surprisingly, the SWC3þ monocyte fraction expressed only minimal levels of the TMG6-5 antigen. However, other TMG6-5-positive cells were clearly present in the PBMC preparation. This pattern of expression indicates that TMG6-5 only binds to approximately 50% of granulocytes and
180
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Table 3 Percentage of positive cells in gates (human leukocytes)a Number
Specificity
mAb
E
F
T34
T35
Mean
Human lymphocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
38 8 2 1 11 12 ND 96 1 2 0 46 ND 8 1 ND 23 2 84 8 7 6 10 1
Human monocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36
98 78 86 38 95 76 ND 93 30 58 13 99 ND 81 5 ND 96 36 57 38 61 52 76
S.D.
46 4 3 1 8 7 ND 96 1 3 1 51 4 5 1 ND 23 2 79 5 5 4 7 2
34 17 0 11 16 17 95 70 8 87 31 97 17 14 29 12 30 13 81 14 29 14 20 51
19 8 3 4 11 24 94 100 13 83 50 97 15 8 63 8 16 11 91 9 13 10 12 44
34 9 2 4 12 15 94 90 6 44 21 73 12 9 23 10 23 7 84 9 13 8 12 25
12 5 1 5 3 7 0 14 6 47 24 28 7 4 29 3 6 6 5 4 11 5 6 27
98 42 84 8 90 38 ND 93 10 29 7 98 11 34 2 ND 96 11 12 11 4 21 39
95 89 96 11 61 75 81 84 77 77 82 100 19 8 89 84 87 22 14 11 86 29 81
84 74 66 7 26 73 66 97 71 88 84 78 12 11 93 74 75 27 14 15 49 36 39
93 71 83 16 68 65 74 92 47 63 46 94 14 33 47 79 88 24 24 19 50 34 59
6 20 13 15 32 18 10 5 32 26 42 10 4 34 50 7 10 10 22 13 34 13 23
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
181
Table 3 (Continued ) Number
Specificity
mAb
E
F
T34
T35
Mean
199
CD49e
SAM-1
59
53
91
86
72
19
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
100 48 91 3 100 45 ND 91 12 16 2 99 ND 81 4 ND 100 8 23 29 5 4 54 1
100 24 87 1 93 12 ND 95 5 11 2 100 7 40 ND ND 100 5 5 5 2 1 21 2
99 85 78 0 58 19 65 85 45 12 54 98 3 0 97 60 98 2 0 0 2 1 10 59
100 97 92 9 80 54 79 95 74 15 90 100 9 17 100 85 100 17 9 11 9 9 33 84
100 63 87 3 83 33 72 91 34 14 37 99 6 35 67 73 99 8 9 12 5 4 30 37
1 34 6 4 18 20 10 5 32 2 43 1 4 35 55 18 1 6 10 13 3 4 19 42
Human granulocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e a
S.D.
Gating based on FSC/SSC characteristics human leukocytes.
binds only minimally to pig monocytes, but its expression is consistent with the presence of the TMG6-5 ligand on pig NK cells. 3.2. #15, S-cHl3 (CD11c) CD11c is found on human granulocytes, monocytes, NK cells and some T- and B-cell subsets. Tables 2 and 3 and Fig. 3 shows that the distribution on pig cells is different. Pig granulocytes, in contrast to human granulocytes, were not recognised by S-cHl3 (Fig. 3b). However, the majority of pig monocytes, defined by SWC3 co-expression on PBMC, were positive (Fig. 3c). Additionally, similar to humans, a subset of CD3þ cells showed variable expression of the S-cHl3 ligand (Fig. 3c) and CD21þ B-cells appeared to express minimal levels of the antigen. 3.3. #38, HP2/19 (CD50) The vast majority of human leukocytes express CD50 (ICAM-3) at a relatively uniform level (Fig. 4a). However, most pig granulocytes were negative (Fig. 4b) and the
182
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Fig. 1. (a) Percentage of positive porcine cells, (b) percentage of positive human cells, gated by FSC/SSC characteristics. Figures based on data in Tables 2 and 3.
expression of the antigen appeared variable on some other pig cells: whereas all monocytes expressing SWC3 were positive, only small subsets of CD3þ T-cells and CD21þ B-cells expressed the HP2/19 antigen at variable levels (Fig. 4c).
4. Discussion A preliminary glance at Fig. 1a and b illustrates that in general, the mean percentage of cells which were positively identified by these reagents was comparatively low for pig cells in comparison to human cells. The reasons for this may be several: (a) reduced affinity of the antibody for the corresponding ligand on pig cells, leading to similar
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
183
Fig. 2. Flow cytometric staining with TMG6-5. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the TMG6-5 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
cellular distribution, but a lowered threshold of detection of relatively sparse ligands. (b) Many antibodies appeared to have deteriorated during storage, thus leading to variable results during the course of this workshop. Some of this apparent deterioration may be due to a reduced affinity, with relatively large effects of lowered effective amounts of antibody. (c) A different cellular distribution of the ligand on human and pig cells, with fewer pig cell types expressing the ligand for the antibody. (d) A different composition of pig and human blood. For example, the number of NK cells in pig blood are relatively high and may constitute as many as 60% of lymphocytes (Raymond et al., 1998). Additionally, we found that setting gates based on FCS/SSC characteristics for pig blood was relatively difficult. There appears to be large amount of overlap between lymphocytes, NK cells and monocytes, so that the so-called ‘monocyte’ gate almost certainly contains many lymphocytes and NK cells and vice versa. We attempted to overcome this difficulty by the use of characteristic cell surface molecule expression in
184
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Fig. 3. Flow cytometric staining with S-cHl3. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the S-cHl3 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
two-colour flow cytometry. (e) The antibody may bind to allotypic determinants, thus leading to erratic results. Finally, for some antibodies, a combination of these may apply. It is not within the scope of this work to identify all of these effects. Therefore, this work is to a large extent preliminary. However, it suffices to illustrate the general problems associated with the use of such reagents. For antibodies classified in category 1, whose cellular reactivity was consistent with, but not proven to be identical to the one found on human cells, no detailed characterisation was carried out. In order to use these reagents for pig cells, it has to be recommended that further characterisation is carried out to establish this. However, the specificity of several of these antibodies has been confirmed in the relevant workshop section reports.
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
185
Fig. 4. Flow cytometric staining with HP2/19. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the HP2/19 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
For antibodies in category 2, this workshop has established that the cellular distribution of the ligands are clearly different for the different species. Of particular interest is the distribution of the ligand for the anti-human CD11b antibody #14 TMG6-5. This antibody clustered very closely with an anti pig antibody, #122 MIL4 (see First Round Report as well as the myeloid section report). In turn, #122 also cross-reacts with human cells and has been submitted to the next human CD workshop. A comparison of the reactivity of both antibodies has previously implied that both antibodies recognise the same ligand on pig cells. The reactivity of MIL4 had been described previously (Haverson et al., 1994). MIL4 identifies pig NK cells, approximately 50% of PMN and eosinophils. Like TMG6-5, it does not recognise pig monocytes. This is in contrast to CD11b expression on human monocytes, in fact, CD11b (MAC-1) is frequently used as a characteristic cell surface molecule of human monocytes. A further complication is added by the fact that an anti pig CD11b antibody has been reported previously (Bullido et al.,
186
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
1996), with similar cellular distribution, functional characteristics and molecular weight as that reported for CD11b on human cells, but clearly different from MIL4 and TMG6-5. A detailed comparison of all three antibodies, as well other CD11-specific antibodies, was carried out in a special investigation and is reported by Ezquerra et al. in this edition. The ligand for the anti-human CD11c antibody #15, S-cHl3, was also expressed differently on pig cells. CD11c is expressed by all human myeloid cells, but not by pig granulocytes. However, like in humans, the antigen was expressed by most pig monocytes as well as by a T-lymphocyte subset. There also appears to be very low expression on pig B-cells. Similarly, the expression of CD50 (ICAM-3), although ubiquitous for the majority of human leukocytes, was more restricted on pig cells. Most pig granulocytes were negative, but monocytes were positive and subsets of T- and B-cells showed variable expression. This work illustrates the importance of carrying out a thorough characterisation of reagents raised and characterised for cells of one species for use in another species. Establishing that cross-reactivity occurs is clearly not enough if meaningful conclusions are to be drawn. In addition, the functional and evolutionary implications of these findings are of interest. Possibly, different species are using similar molecules for different functions and conversely, different molecules for similar functions. An alternative explanation is that the observed cross-reactivity is mainly due to chance similarity of epitopes on different molecules.
References Bullido, R., Alonso, F., delMoral, M.G., et al., 1996. Monoclonal antibody 2F4/11 recognizes the alpha chain of a porcine beta 2 integrin involved in adhesion and complement mediated phagocytosis. J. Immunol. Methods 195 (1/2), 125–134. Haverson, K., Bailey, M., Higgins, V.R., Bland, P.W., Stokes, C.R., 1994. Characterization of monoclonalantibodies specific for monocytes, macrophages and granulocytes from porcine peripheral-blood and mucosal tissues. J. Immunol. Methods 170 (2), 233–245. Raymond, C., Wilkie, B., Mallard, B., Quinton, M., 1998. Natural killer cell frequency and function in Yorkshire pigs selectively bred for high or low antibody and cell-mediated immune response. Natural Immunity 16 (4), 127–136.
Monoclonal antibodies raised to human cells — specificity for pig leukocytes K. Haversona,*, M. Baileya, C.R. Stokesa, A. Simonb, L. LeFlufya, G. Banfielda, Z. Chenc, E. Hollemweguerc, J.A. Ledbetterd a
Division of Molecular and Cellular Biology, University of Bristol, Langford BS40 5DU, UK b LEBAO, Hannover, Germany c Pharmingen, San Diego, CA, USA d Bristol-Myers Squibb, Seattle, WA, USA
Abstract A total of 27 monoclonal antibodies raised to human targets were included in the present Pig CD workshop. 14 of these had been tested in previous workshops and had been reported as crossreactive, a further 13 had been reported as cross-reactive during the Human Leukocyte Differentiation Antigens Workshop VI (HLDA VI) and/or by the donor (a commercial company submitting these mAb for validation by the workshop community). Of the 27 antibodies, three antibodies with previously reported reactivity for pig cells were eliminated from the workshop following preliminary tests due to lack of reactivity. Nine antibodies, although initially positive, gave inconsistent results during the course of the workshop. We found consistent reactivity for 15 antibodies. However, the cellular distribution of the target molecules on pig and human cells was shown to be different for three of these antibodies. These findings have important implications for the usefulness of these antibodies as research tools in the pig. # 2001 Elsevier Science B.V. All rights reserved. Keywords: Swine; Cluster of differentiation (CD); Antigens; Monoclonal antibodies (mAb); Cross-reactivity; CD11b; CD11c; CD50; CD14; CD40; CD47; CD49; CD92; CD93
1. Introduction In spite of the success of the previous and present International Swine CD workshops in the characterization of cluster determinants on pig leucocytes, there is still a shortage of reagents to cell surface molecules relative to those available to our colleagues working *
Corresponding author. Tel.: þ44-117-928-9289; fax: þ44-44-928-9505. E-mail address: [email protected] (K. Haverson). 0165-2427/01/$ – see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 0 1 ) 0 0 2 8 5 - 9
176
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
with human or rodent cells. The preparation of such antibodies is time-consuming and there is therefore a constant search for well-characterization antibodies raised to human cells with cross-reactivity for other species. Although there have been frequent reports of such antibodies, few systematic studies have been conducted into the quality of such reagents for use in the pig. Two main questions need to be answered: The first question concerns the consistent reactivity of such reagents, reduced affinity to other than the original target species may cause problems in this respect. If consistent cross-reactivity is observed, then we need to know if the ligands recognized are homologous to those in humans and if their cellular distribution is the same.
2. Methods Table 1 is a list of all putatively cross-reactive antibodies donated/submitted to this workshop. All antibodies were tested and titrated for optimal use in preliminary work Table 1 List of anti-human mAb submitted to this workshop, with name of donors Number
mAb
Specificity
Donor
14 15 16 23 24 25 33 34 35 36 37 38 39 40 41 42 43 44 45 143 144 145 146 147 148 149 199
TMG6-5 S-Hcl3 BiG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 HH2 FN4 AHN16 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
CD11b CD11c CD14 CD32 CD35 CD40 CD47 CD47 CD49b CD49d CD49f CDw50 CwD75a CDw78a CD63a CD79b CD80 CDw92 CD93 CD11b CD39 CD2 CD21 HLA-DP, DQ, DR HLA-DQ HLA-DR CD49e
Ando Becton Dickenson Schutt Pulford Stein Ledbetter Horejsi Pharmingen Pharmingen Becton Dickenson Pharmingen Sanchez–Madrid Funderud Funderud Skubitz Pharmingen Pharmingen Majdic Majdic Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Pharmingen Van Kooyk
a
Withdrawn after negative preliminary tests.
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
177
before the beginning of the first round by the author, leading to the elimination of three antibodies from the workshop. All remaining antibodies were tested by all workshop participants on many target cells. However, for the purpose of this comparison, only nine selected datasets for pig cells and four for human cells from three different laboratories, with comparable targets and gating of the three major leukocyte populations of peripheral blood, were used for this comparative investigation. Target cells A–D were from slaughter-weight Landracecrossbred pigs, target cells E and F from two normal adult humans, the remaining targets were data from the first round (Appendix). Antibodies were tested in single colour flow cytometry, gating for both species was based on their FCS/SSC characteristics and cells were assigned to the lymphocyte, monocyte and granulocyte gate on this basis. However, it was apparent that the FCS/SSC characteristics of pig monocytes and lymphocytes were frequently overlapping (using the expression of SWC3 as definitive monocyte marker, not shown), therefore, the ‘lymphocyte’ gate would contain small monocytes and the ‘monocyte’ gate would contain large lymphocytes. For this reason, the co-expression of the test antibody with the specific antibodies to CD3, CD21 and SWC3 was also investigated for three selected antibodies, using Ficoll-separated PBMC as targets.
3. Results Tables 2 and 3 show the percentage of positive cells for each antibody in the lymphocyte, monocyte and granulocyte gates from pig (Table 2) and human (Table 3) PBMC, based on FCS/SSC characteristics, for all 24 antibodies. Fig. 1 summarises the mean reactivity of antibodies observed for pig (Fig. 1a) and human (Fig. 1b) cells. See also Appendix for further data. A detailed examination of the reactivities of each antibody revealed that antibodies could be grouped into three main categories. 1. Antibodies with reactivity consistent with similar cellular distribution of the ligand for both species. Antibodies falling into category 1 were: #16 biG10 (CD14), #25 G28.5 (CD40), #33 MEM-122 and #34 CIKM (CD47), #36 L25 (CD49d), #44 VIM15 (CD92), #45 VIMD2 (CD93), #145 RPA-2.10 (CD2), #146 B-ly4 (CD21) as well as the anti-MHC class II antibodies #147 TU39, #148 TU169 and #149 TU36. The reactivity of some of these antibodies is also confirmed in the relevant section reports. 2. Antibodies with consistent reactivity, but identifying cellular subsets which appeared different for each species. Antibodies falling into category 2 were: #14 TMG6-5 (CD11b), #15 S-Hcl3 (CD11c) and #38 HP2/19 (CD50). 3. Antibodies which gave negative, inconsistent or very weak reactivity on pig cells (identified by standard deviations @ mean of positive cells) in comparison to those found on human cells. The remaining 10 antibodies fell into this category. One antibody appeared to have deteriorated even for use on human cells. The results for antibodies in categories 1 and 3 will not be shown in more detail. However, antibodies in category 2 were investigated in more detail in one experiment
178
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Table 2 Percentage of positive cells in gates (pig leukocytes)a Number
Specificity
mAb
A
B
C
D
T8
T9
T10 T11 T12 Mean S.D.
Pig lymphocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
12 33 1 0 3 22 ND 82 1 15 3 44 2 3 0 ND 0 3 36 9 16 8 15 1
7 32 ND ND 3 12 ND 69 1 0 1 ND 3 1 0 ND 1 2 31 6 21 6 10 1
18 19 5 0 7 20 ND 89 2 1 ND 42 5 1 1 ND 0 2 24 9 13 4 18 0
27 32 2 0 2 21 ND 79 4 42 1 ND 5 2 1 ND 5 14 54 19 59 28 46 2
27 20 20 72 73 37 94 97 80 76 13 21 6 4 18 22 6 5 34 11 12 30 22 2
11 5 4 4 4 43 100 89 4 75 29 52 11 4 5 18 4 13 89 7 79 40 80 71
6 5 1 0 0 7 33 44 8 11 8 26 50 2 23 7 1 3 44 7 8 2 4 2
2 2 0 0 0 7 78 67 1 47 4 54 31 0 15 8 0 0 65 9 26 5 14 1
8 2 0 17 0 5 90 60 49 68 69 59 23 0 34 13 0 0 64 7 33 2 4 0
13 17 4 12 10 19 79 75 17 37 16 42 15 2 11 13 2 5 49 9 30 14 24 9
9 14 7 25 24 14 27 17 28 31 23 14 16 2 12 6 2 5 21 4 24 14 25 23
Pig monocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36
20 34 25 1 4 33 ND 90 1 17 4 50 4 4 1 ND 0 6 32 11 26 13 22
15 41 ND 0 9 43 ND 81 2 9 3 ND 5 4 1 ND 3 6 26 8 40 20 31
26 30 20 0 15 43 ND 94 6 1 2 49 11 4 3 ND 0 3 13 12 34 8 36
37 35 15 1 4 44 ND 84 6 29 2 ND 12 4 3 ND 6 27 29 27 73 50 66
6 11 49 0 0 8 31 14 4 27 16 63 63 1 45 24 0 1 12 2 33 0 4
10 25 21 4 4 21 100 93 6 58 29 55 11 5 4 18 88 12 12 16 80 73 88
15 12 52 0 0 14 90 55 0 57 13 80 50 0 80 55 0 0 17 1 30 1 3
6 7 41 38 0 4 95 49 47 67 78 70 33 0 86 43 0 0 2 0 45 1 0
8 66 76 88 87 58 92 98 90 90 32 52 9 5 18 35 4 7 4 4 12 36 25
16 29 37 15 14 30 82 73 18 40 20 60 22 3 27 35 11 7 16 9 41 23 30
10 18 21 30 28 19 29 28 31 30 25 12 22 2 35 15 29 9 11 9 22 26 30
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
179
Table 2 (Continued ) Number
Specificity
mAb
A
B
C
D
T8
199
CD49e
SAM-1
2
3
2
4
7
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
46 4 ND 0 1 6 ND 81 1 0 3 ND 2 1 1 ND 13 1 1 1 1 1 1 1
51 4 57 0 1 10 ND 88 2 2 3 10 4 1 1 ND 1 1 3 1 2 2 3 1
53 7 77 1 2 7 ND 71 4 7 8 ND 4 3 6 ND 18 4 3 4 10 3 8 1
45 8 94 1 1 6 ND 71 1 6 3 35 2 3 1 ND 0 1 2 1 5 1 2 8
Pig granulocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e a
T9
T10 T11 T12 Mean S.D.
90
12
8
6
15
28
61 7 3 3 4 10 93 57 4 9 13 37 8 4 3 13 46 14 5 3 11 6 52 33
40 8 7 0 0 1 18 0 0 2 25 23 12 0 28 50 0 1 0 0 0 0 0 2
55 6 22 0 0 6 82 24 0 32 44 54 7 0 75 84 0 0 1 0 1 0 0 2
77 0 4 19 0 2 71 2 0 3 18 55 1 0 86 61 0 0 0 0 0 0 0 4
53 6 38 3 1 6 66 49 1 8 15 36 5 1 25 52 10 3 2 1 4 2 8 6
12 3 38 7 1 3 33 36 2 10 14 17 4 1 35 30 16 5 2 1 4 2 18 11
Gating based on FSC/SSC characteristics Leukocytes.
using gating based on FCS/SSC characteristics as well as the co-expression of three characteristic pig cell surface molecules, CD3, CD21 and SWC3. 3.1. #14, TMG6-5 (CD11b) CD11b (MAC-1) is found on human granulocytes, monocytes and NK cells. Fig. 2 shows that the distribution on pig cells is different. Whereas both human monocytes and granulocytes are strongly positive for this antibody (Fig. 2a), pig cells in all gates including granulocytes have both positive and negative subsets (Tables 2 and 3 and Fig. 2b). Additionally, the expression levels on the positive subsets are lower than those on human cells. A more detailed investigation, based on the co-expression of specific cell surface molecules on PBMC (Fig. 2c), shows that CD3þ T-cells and CD21þ B-cells were negative for the TMG6-5 ligand. Additionally and surprisingly, the SWC3þ monocyte fraction expressed only minimal levels of the TMG6-5 antigen. However, other TMG6-5-positive cells were clearly present in the PBMC preparation. This pattern of expression indicates that TMG6-5 only binds to approximately 50% of granulocytes and
180
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Table 3 Percentage of positive cells in gates (human leukocytes)a Number
Specificity
mAb
E
F
T34
T35
Mean
Human lymphocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
38 8 2 1 11 12 ND 96 1 2 0 46 ND 8 1 ND 23 2 84 8 7 6 10 1
Human monocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36
98 78 86 38 95 76 ND 93 30 58 13 99 ND 81 5 ND 96 36 57 38 61 52 76
S.D.
46 4 3 1 8 7 ND 96 1 3 1 51 4 5 1 ND 23 2 79 5 5 4 7 2
34 17 0 11 16 17 95 70 8 87 31 97 17 14 29 12 30 13 81 14 29 14 20 51
19 8 3 4 11 24 94 100 13 83 50 97 15 8 63 8 16 11 91 9 13 10 12 44
34 9 2 4 12 15 94 90 6 44 21 73 12 9 23 10 23 7 84 9 13 8 12 25
12 5 1 5 3 7 0 14 6 47 24 28 7 4 29 3 6 6 5 4 11 5 6 27
98 42 84 8 90 38 ND 93 10 29 7 98 11 34 2 ND 96 11 12 11 4 21 39
95 89 96 11 61 75 81 84 77 77 82 100 19 8 89 84 87 22 14 11 86 29 81
84 74 66 7 26 73 66 97 71 88 84 78 12 11 93 74 75 27 14 15 49 36 39
93 71 83 16 68 65 74 92 47 63 46 94 14 33 47 79 88 24 24 19 50 34 59
6 20 13 15 32 18 10 5 32 26 42 10 4 34 50 7 10 10 22 13 34 13 23
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
181
Table 3 (Continued ) Number
Specificity
mAb
E
F
T34
T35
Mean
199
CD49e
SAM-1
59
53
91
86
72
19
TMG6-5 S-Hcl3 biG10 2ZC115 Ber-MAC-DRC G28.5 MEM-122 CIKM AK7 L25 GOH3 HP2/19 CB3-1 Bb-1 VIM15 VIMD2 ICRF44 TU66 RPA-2.10 B-ly4 TU39 TU169 TU36 SAM-1
100 48 91 3 100 45 ND 91 12 16 2 99 ND 81 4 ND 100 8 23 29 5 4 54 1
100 24 87 1 93 12 ND 95 5 11 2 100 7 40 ND ND 100 5 5 5 2 1 21 2
99 85 78 0 58 19 65 85 45 12 54 98 3 0 97 60 98 2 0 0 2 1 10 59
100 97 92 9 80 54 79 95 74 15 90 100 9 17 100 85 100 17 9 11 9 9 33 84
100 63 87 3 83 33 72 91 34 14 37 99 6 35 67 73 99 8 9 12 5 4 30 37
1 34 6 4 18 20 10 5 32 2 43 1 4 35 55 18 1 6 10 13 3 4 19 42
Human granulocyte gate 14 CD11b 15 CD11c 16 CD14 23 CD32 24 CD35 25 CD40 33 CD47 34 CD47 35 CD49b 36 CD49d 37 CD49f 38 CDw50 42 CD79b 43 CD80 44 CDw92 45 CD93 143 CD11b 144 CD39 145 CD2 146 CD21 147 HLA-DP 148 HLA-DQ 149 HLA-DR 199 CD49e a
S.D.
Gating based on FSC/SSC characteristics human leukocytes.
binds only minimally to pig monocytes, but its expression is consistent with the presence of the TMG6-5 ligand on pig NK cells. 3.2. #15, S-cHl3 (CD11c) CD11c is found on human granulocytes, monocytes, NK cells and some T- and B-cell subsets. Tables 2 and 3 and Fig. 3 shows that the distribution on pig cells is different. Pig granulocytes, in contrast to human granulocytes, were not recognised by S-cHl3 (Fig. 3b). However, the majority of pig monocytes, defined by SWC3 co-expression on PBMC, were positive (Fig. 3c). Additionally, similar to humans, a subset of CD3þ cells showed variable expression of the S-cHl3 ligand (Fig. 3c) and CD21þ B-cells appeared to express minimal levels of the antigen. 3.3. #38, HP2/19 (CD50) The vast majority of human leukocytes express CD50 (ICAM-3) at a relatively uniform level (Fig. 4a). However, most pig granulocytes were negative (Fig. 4b) and the
182
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Fig. 1. (a) Percentage of positive porcine cells, (b) percentage of positive human cells, gated by FSC/SSC characteristics. Figures based on data in Tables 2 and 3.
expression of the antigen appeared variable on some other pig cells: whereas all monocytes expressing SWC3 were positive, only small subsets of CD3þ T-cells and CD21þ B-cells expressed the HP2/19 antigen at variable levels (Fig. 4c).
4. Discussion A preliminary glance at Fig. 1a and b illustrates that in general, the mean percentage of cells which were positively identified by these reagents was comparatively low for pig cells in comparison to human cells. The reasons for this may be several: (a) reduced affinity of the antibody for the corresponding ligand on pig cells, leading to similar
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
183
Fig. 2. Flow cytometric staining with TMG6-5. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the TMG6-5 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
cellular distribution, but a lowered threshold of detection of relatively sparse ligands. (b) Many antibodies appeared to have deteriorated during storage, thus leading to variable results during the course of this workshop. Some of this apparent deterioration may be due to a reduced affinity, with relatively large effects of lowered effective amounts of antibody. (c) A different cellular distribution of the ligand on human and pig cells, with fewer pig cell types expressing the ligand for the antibody. (d) A different composition of pig and human blood. For example, the number of NK cells in pig blood are relatively high and may constitute as many as 60% of lymphocytes (Raymond et al., 1998). Additionally, we found that setting gates based on FCS/SSC characteristics for pig blood was relatively difficult. There appears to be large amount of overlap between lymphocytes, NK cells and monocytes, so that the so-called ‘monocyte’ gate almost certainly contains many lymphocytes and NK cells and vice versa. We attempted to overcome this difficulty by the use of characteristic cell surface molecule expression in
184
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
Fig. 3. Flow cytometric staining with S-cHl3. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the S-cHl3 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
two-colour flow cytometry. (e) The antibody may bind to allotypic determinants, thus leading to erratic results. Finally, for some antibodies, a combination of these may apply. It is not within the scope of this work to identify all of these effects. Therefore, this work is to a large extent preliminary. However, it suffices to illustrate the general problems associated with the use of such reagents. For antibodies classified in category 1, whose cellular reactivity was consistent with, but not proven to be identical to the one found on human cells, no detailed characterisation was carried out. In order to use these reagents for pig cells, it has to be recommended that further characterisation is carried out to establish this. However, the specificity of several of these antibodies has been confirmed in the relevant workshop section reports.
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
185
Fig. 4. Flow cytometric staining with HP2/19. (a) Results for human leukocytes; (b) results for pig leukocytes, with gating based on FCS/SSC characteristics; (c) co-expression of the HP2/19 ligand with CD3, CD21 and the monocyte/granulocyte marker SWC3 on pig peripheral blood mononuclear cells.
For antibodies in category 2, this workshop has established that the cellular distribution of the ligands are clearly different for the different species. Of particular interest is the distribution of the ligand for the anti-human CD11b antibody #14 TMG6-5. This antibody clustered very closely with an anti pig antibody, #122 MIL4 (see First Round Report as well as the myeloid section report). In turn, #122 also cross-reacts with human cells and has been submitted to the next human CD workshop. A comparison of the reactivity of both antibodies has previously implied that both antibodies recognise the same ligand on pig cells. The reactivity of MIL4 had been described previously (Haverson et al., 1994). MIL4 identifies pig NK cells, approximately 50% of PMN and eosinophils. Like TMG6-5, it does not recognise pig monocytes. This is in contrast to CD11b expression on human monocytes, in fact, CD11b (MAC-1) is frequently used as a characteristic cell surface molecule of human monocytes. A further complication is added by the fact that an anti pig CD11b antibody has been reported previously (Bullido et al.,
186
K. Haverson et al. / Veterinary Immunology and Immunopathology 80 (2001) 175–186
1996), with similar cellular distribution, functional characteristics and molecular weight as that reported for CD11b on human cells, but clearly different from MIL4 and TMG6-5. A detailed comparison of all three antibodies, as well other CD11-specific antibodies, was carried out in a special investigation and is reported by Ezquerra et al. in this edition. The ligand for the anti-human CD11c antibody #15, S-cHl3, was also expressed differently on pig cells. CD11c is expressed by all human myeloid cells, but not by pig granulocytes. However, like in humans, the antigen was expressed by most pig monocytes as well as by a T-lymphocyte subset. There also appears to be very low expression on pig B-cells. Similarly, the expression of CD50 (ICAM-3), although ubiquitous for the majority of human leukocytes, was more restricted on pig cells. Most pig granulocytes were negative, but monocytes were positive and subsets of T- and B-cells showed variable expression. This work illustrates the importance of carrying out a thorough characterisation of reagents raised and characterised for cells of one species for use in another species. Establishing that cross-reactivity occurs is clearly not enough if meaningful conclusions are to be drawn. In addition, the functional and evolutionary implications of these findings are of interest. Possibly, different species are using similar molecules for different functions and conversely, different molecules for similar functions. An alternative explanation is that the observed cross-reactivity is mainly due to chance similarity of epitopes on different molecules.
References Bullido, R., Alonso, F., delMoral, M.G., et al., 1996. Monoclonal antibody 2F4/11 recognizes the alpha chain of a porcine beta 2 integrin involved in adhesion and complement mediated phagocytosis. J. Immunol. Methods 195 (1/2), 125–134. Haverson, K., Bailey, M., Higgins, V.R., Bland, P.W., Stokes, C.R., 1994. Characterization of monoclonalantibodies specific for monocytes, macrophages and granulocytes from porcine peripheral-blood and mucosal tissues. J. Immunol. Methods 170 (2), 233–245. Raymond, C., Wilkie, B., Mallard, B., Quinton, M., 1998. Natural killer cell frequency and function in Yorkshire pigs selectively bred for high or low antibody and cell-mediated immune response. Natural Immunity 16 (4), 127–136.