Secondary immune correlate analysis of the moderately protective RV144 vaccine trial showed an association between the reduced risk of HIV acquisition and antibody-dependent cellular cytotoxicity (ADCC) in the absence of high serum IgA . Similarly, mounting evidence suggests that ADCC activity correlates with enhanced HIV control , slower disease progression [3–7] and enhanced neutralizing antibody-mediated sterilizing protection from infection , strongly suggesting that antibody fragment crystalizable (Fc)-effector functions may play a critical role in protective immunity against HIV.
Previous studies in the monoclonal antibody (mAb) therapeutic field have demonstrated that the ADCC potential of an antibody is modulated by specific biochemical properties of the Fc-region of an antibody, including subclass selection and glycosylation . N-linked glycans on antibodies can be classified according to their level of galactosylation: G0, G1, and G2, which, respectively, contain 0, 1 and 2 galactose residues. When a single galactose is added, it can be further identified as either being bound on the α-3 or α-6 mannose arm (α-3-arm G1 or α-6-arm G1, respectively) of the core conserved bi-antennary hepatasaccharide structure. The glycan structure can be further altered by the presence or absence of fucose and sialic acid. In addition, mAbs produced from nonhuman cells, such as Chinese hamster ovary (CHO) cells and mouse myeloma cells, may generate distinct glycan structures due to the expression of enzymes that can add sugars and/or sugar linkages that are not present in humans, including a nonhuman sialic acid, N-glycolylneuraminic acid (Neu5Gc) and galactose in an alpha-1,3 linkage (α-1,3-Gal). Additionally, CHO cells lack GnT III and ST6GalI activity, resulting in a loss of glycan structures that otherwise would be produced in humans.
Previous mAb therapeutic studies have linked the absence of core fucose  with elevated ADCC activity, whereas the role of galactosylation has been more controversial [10–13]. Notably, recent studies in HIV-infected individuals have demonstrated the enrichment of agalactosylated  and afucosylated glycans on both bulk and HIV-specific antibodies, and in particular in controllers who exhibit robust antibody-dependent cellular viral inhibition . In contrast, the specific biophysical features of antibodies that drive more effective antibody-dependent cellular phagocytosis (ADCP), beyond differential binding to Fc-receptors (FcRs) [16,17], have yet to be defined.
With the growing interest in the delivery of HIV-specific mAbs for prophylactic treatment and ‘cure’ strategies [18–20], understanding the specific features of a mAb that promotes the most effective innate immune responses may be of vital importance. However, although several studies have associated strong ADCC activity with antibodies against key epitope specificities [21–23], few studies have taken Fc-glycan modifications into account when comparing multiple mAbs in effector assays. Here we sought to determine whether antibody specificity to key regions of the HIV envelope could predict Fc-effector activity, and/or if the strength of Fc-activity was impacted by changes in glycosylation of the mAb Fc-region.
A selection of mAbs with various specificities including CD4+ binding site: 559/64D, 1331–160E, b6, b12; gp41 cluster I: 50–69D, 240-D; gp41/MPER: 2F5, 4E10; V2: 697-30D, 2158; V3: 2191 and 2219. 559/64D, 1331-160E, 50-69D, 240-D, 697-30D, 2158, 2191 and 2219 – were produced by hybridoma cell lines obtained via fusion of Epstein–Barr virus-transformed peripheral blood mononuclear cells derived from HIV-seropositive individuals, with human × mouse heteromyeloma cells SHM-D33 . Monoclonal antibodies b6 and b12, which were kindly provided by Dr Dennis Burton, and 2F5 and 4E10, provided by the NIH AIDS Reference Reagent Program, produced in CHO cells . All mAbs were of an IgG1 isotype.
Enzyme-linked immunosorbent assay
Antigen avidity was determined by ELISA. Gp140SF162 was coated directly onto ELISA plates at a concentration of 0.5 μg/ml overnight, incubated with titrated mAbs at a dose range (1.0–0.0003 μg/ml) that could capture the three critical phases of mAb binding (saturation, a slope and no reactivity), followed by detection with alkaline phosphatase-conjugated goat antihuman IgG(Fc)-antibody (Southern Biotech, Birmingham, Alabama, USA). ELISAs were repeated twice and in each experiment, mAbs were tested in duplicate. Relative affinity was determined by measuring the concentration of mAb that gave 50% maximal binding by regression analysis (using a curve fit) on Prism 5.
High-performance liquid chromatography
N-linked glycans from whole IgG were released from Coomassie blue-stained reducing SDS-PAGE gel bands by enzymatic release using PNGase F (New England Biolabs, Ipswich, Massachusetts, USA). Glycans were fluorescently labeled with 2-aminobenzoic acid (2-AA) [26,27] and the excess dye was removed using a Spe-ed Amide-2 column (Systematic Instruments, Stockport, UK). The 2-AA-labeled glycans were separated using LudgerSep N2amide HPLC column (Ludger, Oxfordshire, UK) and HPLC was carried out in a linear gradient of solvents at 30oC, as previously described [26,27].
Antibody-dependent cellular phagocytosis assay
A phagocytosis assay was performed as previously described . Briefly, rgp140SF162 (Immune Technology, New York, New York, USA) bound to fluorescent beads were incubated with mAb and monocyte THP-1 cells. The cells were analyzed for bead uptake by flow cytometry. Each mAb was tested at 100 μg/ml, twice in replicate experiments.
Rapid fluorescent antibody-dependent cellular cytotoxicity assay
A modified rapid fluorescent ADCC (RFADCC) assay was used as previously described . Briefly, CEM-natural killer resistant cell line (CEM-NKr) T-lymphoblast cells were pulsed with gp140SF162 (60 μg/ml), labeled with intracellular carboxyfluorescein succinimidyl ester (CFSE) and membrane Paul Karl Horan 26 dye (PKH26) dye. mAbs (100 μg/ml) were added to CEM-NKrs and healthy donor natural killer (NK) cells isolated using RosetteSep (Stem Cell Technologies, Vancouver, BC, Canada). Cells that maintained PKH26 but lost intracellular CFSE were analyzed by flow cytometry. mAbs were tested twice in replicate experiments.
Fcγ-receptor surface plasmon resonance analysis
Surface plasmon resonance (SPR) experiments were run on a Biacore 3000 to measure Fcγ receptor (FcγR) binding activity. CM5 chips were coated with FcγRIIa, FcγIIb, FcγIIIa [Cat #1330-CD-050 (Histidine-131), #1875-CD-050, #4325-FC-050 (phenylalanine-158) R&D Systems, Minneapolis, Minnesota, USA] or buffer alone. mAb (100 μg/ml) binding to each FcγR was assessed, tested in replicate experiments and quantified as the relative response units of signal.
Statistical analysis was performed using GraphPad Prism (GraphPad Software, San Diego, California, USA). Spearman-rank correlations were used to examine bivariate associations between continuous outcomes. All P values were two-sided. GENE-E matrix visualization and analysis program was used to create heatmaps .
Fragment antigen binding specificity does not influence monoclonal antibody Fc-effector responses
A panel of mAbs was selected based on their specificity to key immunogenic regions of gp120, including V2, V3, CD4+ binding site (CD4bs) and gp41. The mAb-binding avidity, and ADCP and ADCC activity were all assessed against the same gp140SF162 glycoprotein. Surprisingly, mAbs with similar specificities did not drive similar levels of antibody effector activity for either ADCC (Fig. 1a) or ADCP activity (Fig. 1b), for example, CD4bs-specific mAbs were among the highest and lowest mediators of ADCC and ADCP. Notably, however, gp41-specific mAbs did induce higher ADCP compared to other mAb specificities. Moreover, no association was observed between the capacities of the antibodies to drive both ADCC and ADCP (Fig. 1c). Gp140SF162 avidity was determined by ELISA and relative avidity was determined by measuring the concentration of mAb that gave 50% maximal binding by regression analysis (Fig. 1d). Similarly, gp140SF162 avidity did not predict Fc-effector function (Fig. 1e: ADCC r = 0.2, P = 0.93; Fig. 1f: ADCP r = −0.35, P = 0.25). These data suggest that for the specificities tested, other factors most likely dictated by the Fc-region may predominantly drive Fc-effector functions in in-vitro assays, in this set of mAbs (n = 12).
Monoclonal antibody Fc-effector function is determined by Fcγ-receptor binding
Numerous studies have shown that Fc-effector activity is driven by the antibody affinity to FcγRs . In particular, NK cell-mediated ADCC is governed by FcγRIIIa activation , whereas phagocytes express a wider array of FcγRs, and therefore may be modulated by the synergistic triggering of multiple FcγRs simultaneously . SPR technology was used to evaluate mAb binding to activating FcγRIIa (Fig. 1g), FcγRIIIa (Fig. 1i) and inhibitory FcγRIIb (Fig. 1h). ADCC correlated robustly with FcγRIIIa binding (r = 0.71, P = 0.0032, data not shown), confirming the previous studies . In contrast, no single FcγR binding correlated with ADCP, suggesting that binding to a single FcγR alone may not predict this antibody Fc-activity.
Decreased galactosylation inversely correlates with antibody-dependent cellular phagocytosis, whereas N-glycolylneuraminic acid structures predict stronger antibody-dependent cellular phagocytosis
Monoclonal therapeutics have exploited antibody glycosylation as a mechanism to potentiate the Fc-effector activities of mAbs [32–34]. Therefore, mAb glycosylation was assessed to determine if glycan structures were more highly predictive of potent antibody effector function. Increased galactosylation correlated with stronger ADCP activity (Fig. 2a). Specifically, fully galactosylated glycans (total G2) positively correlated with increased ADCP (r = 0.57, P = 0.02), whereas agalactosylated structures (total G0) inversely correlated with ADCP (r = −0.62, P = 0.01). We further resolved the specific glycoforms that mediated these differential ADCP activities. Both G0 (r = −0.65, P = 0.01) and G0F (r = −0.52, P = 0.04) inversely correlated with ADCP activity, whereas divergent complex relationships were observed for G1 and G2-specific structures. Nearly all Neu5Gc-containing structures, regardless of galactosylation content [with the exception of α-G2F(3)Neu5Gc], significantly correlated (r > 0.54, P < 0.05) or trended towards elevated ADCP activity [α-G2F(6)Neu5Gc: r = 0.52, P = 0.1] (Fig. 2), indicating that agalactosylation reduces ADCP, whereas increased presence of Neu5Gc enhances ADCP activity.
Galactosylation arm linkages predict antibody-dependent cellular cytotoxicity activity
No difference was observed between total galactosylation levels (total G0 or G2) and ADCC activity (Fig. 2a). Interestingly, however, a difference in galactose arm linkages was associated with differential antibody functionality. Specifically when looking at bulk galactosylation, the total α-3-arm G1 levels inversely correlated with ADCC activity (r = −0.65, P = 0.011) and FcγRIIIa binding(r = −0.75, P = 0.002) (Fig. 2b and c), whereas the total α-6-arm G1 structures enhanced ADCC (r = 0.45, P = 0.1) and FcγRIIIa binding (r = 0.58, P = 0.029) (Fig. 2d and e). Upon further analysis, α-3-arm G1F inversely correlated with ADCC (r = −0.54, P = 0.04), whereas α-6-arm G1F positively correlated with ADCC (r = 0.51, P = 0.05). These data suggest that beyond gross galactosylation levels, the arm upon which galactose is added to the core glycan structure may contribute to FcγRIIIa binding and thus Fc-effector functions, and provide a potential explanation of conflicting findings regarding the influence of galactose content on ADCC [10–13].
α-G2F(6)Neu5Gc structures broadly induce Fc-effector activity
As mentioned above, the presence of Neu5Gc correlated with enhanced phagocytosis (Fig. 2a). Of interest, α-G2F(6)Neu5Gc was the only glycan structure that correlated with both ADCC (r = 0.8571, P = 0.01) and ADCP activity (r = 0.52, P = 0.1) (Fig. 2a). Not surprisingly, presence of α-G2F(6)Neu5Gc also correlated highly with FcγRIIa (r = 0.90, P = 0.004), FcγRIIb (r = 0.76, P = 0.03) and trended with FcγRIIIa binding (r = 0.69, P = 0.07). However, the presence of α-1,3-Gal structures has been associated with anaphylaxis in previous mAb therapeutic studies, suggesting that while potently functional, they may be undesirable due to risks of adverse events in vivo. Importantly, α-1,3-Gal structures were only observed on mAbs derived from mouse heteromyeloma cells, but not CHO cells, pointing toward variability in mAb functionality derived from different cell lines.
Despite multiple studies pointing to a critical role of afucosylation in enhancing ADCC and FcγRIIIa binding, we did not observe this relationship in our relatively small study of mAbs (n = 12). In contrast, as mentioned above, we observed a strong correlation between the incorporation of α-G2F(6)Neu5Gc and ADCC and FcγR binding (Fig. 2a). We therefore speculated that the functional effect of afucosylation on ADCC may have been masked by the presence of α-G2F(6)Neu5Gc. To evaluate this hypothesis, glycan structures containing fucose were further divided into those with and without α-G2F(6)Neu5Gc structures to examine the effect of total fucose without the heavy influence of α-G2F(6)Neu5Gc. Upon separation of the structures, the expected trend was observed for fucose, with increased fucose correlating with decreased ADCC activity (r = −0.48, P = 0.08; Fig. 2f) and FcγRIIIa binding (r = −0.43, P = 0.10; Fig, 2g). These data confirm the importance of afucosylation in enhancing ADCC activity, highlight the importance of other functional glycan modifications, and point toward the complex and context-dependent impacts of these individual hexose structures on the overall activity of a mixed population of glycovariant mAbs.
Whereas multiple studies have associated strong Fc-innate immune recruiting activity to specific mAbs targeting key immunogenic specificities in the setting of HIV infection [21–23], limited attention has been paid to differences in Fc-glycosylation, and its impact on Fc-effector activity. The study presented here suggests that the impact of variant glycan structures may dominate over the variable domain characteristics in driving Fc-effector activity. Hence, future studies examining mAb Fc-mediated control should consider not just antibody epitope-binding properties, but also FcR-binding capacity and glycan structures to fully predict antibody-mediated antiviral control.
Interestingly, these data highlight previous observations that antibody glycosylation can vary between different producer cell lines [9,34,36]. Within this study, immunogenic α-1,3-Gal structures were only present on mAbs derived from mouse heteromyelomas, whereas nonhuman sialic acid Neu5Gc structures were produced by both heteromylomas and CHO cell lines. Although the presence of Neu5Gc structures were associated with increased ADCP activity, mounting evidence suggests that humans may contain high levels of anti-Neu5Gc-specific IgG , bringing into question the potential therapeutic utility of these structures, given both their potential immunogenicity and the impact that pre-existing immunity could have in dampening the efficacy of mAbs containing these structures. Thus, the selection of the most functional glycan structures must be balanced with the potential immunogenic nature of some of the unusual, but most highly functional glycan structures.
Unexpectedly, we observed that decreased galactosylation (G0 and G0F) reduced ADCP activity, whereas, in contrast, fully agalactosylated structures (G0) correlated with increased FcγRIIIa binding. Additionally, specific galactosylation arm linkages were associated with ADCC activity (α-6-arm G1 was positively associated with enhanced ADCC, whereas α-3-arm G1 did not), independent of the presence of fucose, potentially explaining the conflicting observations from numerous groups on the importance of galactosylation on ADCC activity [12,13].
Overall, mAb therapeutics used in HIV therapeutic treatment [18,19], prevention, and potential cure strategies  should take into consideration Fc-glycan modification in order to boost the antiviral ability of the most potent fragment antigen binding domains in their battle against HIV. Over 30 distinct glycan structures have been identified in healthy human serum ; however, only a limited number are being actively exploited by mAb therapeutics. Thus, future efforts should aim to develop understanding of how other novel glycan structures may naturally tune antibody activity in order to define a new set of Fc modifications that may be exploited in the setting of mAb therapeutic functional optimization efforts.
The following reagent was obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: HIV-1 gp41 Monoclonal Antibody (4E10) and HIV-1 gp41 Monoclonal Antibody (2F5) from Dr Hermann Katinger, HIV-1 gp120 CEM.NKR-CCR5 from Dr Alexandra Trkola. Monoclonal Antibody IgG1 b12 and IgG1 b6 were kindly provided by Dr Dennis Burton,
These studies were supported in part by Collaboration for AIDS Vaccine Discovery (CAVD) (OPP1032817: Leveraging Antibody Effector Function) (A.W.C., M.C., L.P., H.R., C.B.K., M.E.A., C.S., G.A.), National Health and Medical Research Council (NHMRC) APP1036470 (A.W.C.), HL59725 (S.Z.P.) and P01 AI100151 (S.Z.P.), and by research funds from the Department of Veterans Affairs.
Conflicts of interest
There are no conflicts of interest.
1. Haynes BF, Gilbert PB, McElrath MJ, Zolla-Pazner S, Tomaras GD, Alam SM, et al. Immune-correlates analysis of an HIV-1 vaccine efficacy trial
. N Engl J Med
2. Ahmad R, Sindhu ST, Toma E, Morisset R, Vincelette J, Menezes J, et al. Evidence for a correlation between antibody-dependent cellular cytotoxicity-mediating anti-HIV-1 antibodies and prognostic predictors of HIV infection
. J Clin Immunol
3. Hessell AJ, Hangartner L, Hunter M, Havenith CE, Beurskens FJ, Bakker JM, et al. Fc receptor but not complement binding is important in antibody protection against HIV
4. Forthal DN, Gilbert PB, Landucci G, Phan T. Recombinant gp120 vaccine-induced antibodies inhibit clinical strains of HIV-1 in the presence of Fc receptor-bearing effector cells and correlate inversely with HIV infection rate
. J Immunol
5. Forthal DN, Landucci G, Haubrich R, Keenan B, Kuppermann BD, Tilles JG, et al. Antibody-dependent cellular cytotoxicity independently predicts survival in severely immunocompromised human immunodeficiency virus-infected patients
. J Infect Dis
6. Baum LL, Cassutt KJ, Knigge K, Khattri R, Margolick J, Rinaldo C, et al. HIV-1 gp120-specific antibody-dependent cell-mediated cytotoxicity correlates with rate of disease progression
. J Immunol
7. Chung AW, Navis M, Isitman G, Wren L, Silvers J, Amin J, et al. Activation of NK cells by ADCC antibodies and HIV disease progression
. J Acquir Immune Defic Syndr
8. Jefferis R. Isotype and glycoform selection for antibody therapeutics
. Arch Biochem Biophys
9. Shields RL, Lai J, Keck R, O’Connell LY, Hong K, Meng YG, et al. Lack of fucose on human IgG1 N-linked oligosaccharide improves binding to human Fcgamma RIII and antibody-dependent cellular toxicity
. J Biol Chem
10. Kumpel BM, Rademacher TW, Rook GA, Williams PJ, Wilson IB. Galactosylation of human IgG monoclonal anti-D produced by EBV-transformed B-lymphoblastoid cell lines is dependent on culture method and affects Fc receptor-mediated functional activity
. Hum Antibodies Hybridomas
11. Kumpel BM, Wang Y, Griffiths HL, Hadley AG, Rook GA. The biological activity of human monoclonal IgG anti-D is reduced by beta-galactosidase treatment
. Hum Antibodies Hybridomas
12. Boyd PN, Lines AC, Patel AK. The effect of the removal of sialic acid, galactose and total carbohydrate on the functional activity of Campath-1H
. Molec Immunol
13. Shinkawa T, Nakamura K, Yamane N, Shoji-Hosaka E, Kanda Y, Sakurada M, et al. The absence of fucose but not the presence of galactose or bisecting N-acetylglucosamine of human IgG1 complex-type oligosaccharides shows the critical role of enhancing antibody-dependent cellular cytotoxicity
. J Biol Chem
14. Moore JS, Wu X, Kulhavy R, Tomana M, Novak J, Moldoveanu Z, et al. Increased levels of galactose-deficient IgG in sera of HIV-1-infected individuals
15. Ackerman ME, Crispin M, Yu X, Baruah K, Boesch AW, Harvey DJ, et al. Natural variation in Fc glycosylation of HIV-specific antibodies impacts antiviral activity
. J Clin Investig
16. Dugast AS, Tonelli A, Berger CT, Ackerman ME, Sciaranghella G, Liu Q, et al. Decreased Fc receptor expression on innate immune cells is associated with impaired antibody-mediated cellular phagocytic activity in chronically HIV-1 infected individuals
17. Ackerman ME, Dugast AS, McAndrew EG, Tsoukas S, Licht AF, Irvine DJ, et al. Enhanced phagocytic activity of HIV-specific antibodies correlates with natural production of immunoglobulins with skewed affinity for FcgammaR2a and FcgammaR2b
. J Virol
18. Barouch DH, Whitney JB, Moldt B, Klein F, Oliveira TY, Liu J, et al. Therapeutic efficacy of potent neutralizing HIV-1-specific monoclonal antibodies in SHIV-infected rhesus monkeys
19. Shingai M, Nishimura Y, Klein F, Mouquet H, Donau OK, Plishka R, et al. Antibody-mediated immunotherapy of macaques chronically infected with SHIV suppresses viraemia
20. Klein F, Mouquet H, Dosenovic P, Scheid JF, Scharf L, Nussenzweig MC. Antibodies in HIV-1 vaccine development and therapy
21. Ferrari G, Pollara J, Kozink D, Harms T, Drinker M, Freel S, et al. An HIV-1 gp120 envelope human monoclonal antibody that recognizes a C1 conformational epitope mediates potent antibody-dependent cellular cytotoxicity (ADCC) activity and defines a common ADCC epitope in human HIV-1 serum
. J Virol
22. Tudor D, Bomsel M. The broadly neutralizing HIV-1 IgG 2F5 elicits gp41-specific antibody-dependent cell cytotoxicity in a FcgammaRI-dependent manner
23. Tyler DS, Stanley SD, Zolla-Pazner S, Gorny MK, Shadduck PP, Langlois AJ, et al. Identification of sites within gp41 that serve as targets for antibody-dependent cellular cytotoxicity by using human monoclonal antibodies
. J Immunol
24. Gorny MK, Xu JY, Gianakakos V, Karwowska S, Williams C, Sheppard HW, et al. Production of site-selected neutralizing human monoclonal antibodies against the third variable domain of the human immunodeficiency virus type 1 envelope glycoprotein
. Proc Natl Acad Sci U S A
25. Burton DR, Pyati J, Koduri R, Sharp SJ, Thornton GB, Parren PW, et al. Efficient neutralization of primary isolates of HIV-1 by a recombinant human monoclonal antibody
26. Yu X, Vasiljevic S, Mitchell DA, Crispin M, Scanlan CN. Dissecting the molecular mechanism of IVIg therapy: the interaction between serum IgG and DC-SIGN is independent of antibody glycoform or Fc domain
. J Mol Biol
27. Neville DC, Coquard V, Priestman DA, te Vruchte DJ, Sillence DJ, Dwek RA, et al. Analysis of fluorescently labeled glycosphingolipid-derived oligosaccharides following ceramide glycanase digestion and anthranilic acid labeling
. Anal Biochem
28. Ackerman ME, Moldt B, Wyatt RT, Dugast AS, McAndrew E, Tsoukas S, et al. A robust, high-throughput assay to determine the phagocytic activity of clinical antibody samples
. J Immunol Methods
29. Gomez-Roman VR, Florese RH, Patterson LJ, Peng B, Venzon D, Aldrich K, et al. A simplified method for the rapid fluorometric assessment of antibody-dependent cell-mediated cytotoxicity
. J Immunol Methods
31. Nimmerjahn F, Ravetch JV. Fcgamma receptors as regulators of immune responses
. Nat Rev Immunol
32. Arnold JN, Wormald MR, Sim RB, Rudd PM, Dwek RA. The impact of glycosylation on the biological function and structure of human immunoglobulins
. Annu Rev Immunol
33. Raju TS. Terminal sugars of Fc glycans influence antibody effector functions of IgGs
. Curr Opin Immunol
34. Hristodorov D, Fischer R, Linden L. With or without sugar? (A)glycosylation of therapeutic antibodies
. Molec Biotechnol
35. Chung CH, Mirakhur B, Chan E, Le QT, Berlin J, Morse M, et al. Cetuximab-induced anaphylaxis and IgE specific for galactose-alpha-1,3-galactose
. N Engl J Med
36. Frenzel A, Hust M, Schirrmann T. Expression of recombinant antibodies
. Front Immunol
37. Ghaderi D, Taylor RE, Padler-Karavani V, Diaz S, Varki A. Implications of the presence of N-glycolylneuraminic acid in recombinant therapeutic glycoproteins
. Nature Biotechnol
38. Routier FH, Hounsell EF, Rudd PM, Takahashi N, Bond A, Hay FC, et al. Quantitation of the oligosaccharides of human serum IgG from patients with rheumatoid arthritis: a critical evaluation of different methods
. J Immunol Methods