Discussion
This study documents the IgG subclass cross-reactivity pattern of
several IgG subclass-specific mAbs that are used for the detection of
DSA subclasses in prospective and actual transplant recipients. This
study is the first to document an extensive titrimetric analysis of IgG
subclass cross-reactivity pattern that is dependent on the concentration
of both IgG subclass-specific mAbs and IgG1-4 protein targets. While the
results of this study have a minimal impact on the general clinical
findings about IgG subclass-specific DSAs, the awareness of such
cross-reactivity may help the refinement and understanding of the
complex distribution of IgG subclass-specific DSAs found in transplant
patients (10,15,19,20). Surprisingly, only the anti-human-IgG3 mAb
(clone: HP6050) was truly mono-specific at all concentrations tested,
where it recognized only IgG3 proteins and not IgG1, IgG2 or IgG4
proteins. All other mAbs tested were binding to IgG subclass proteins in
addition to their respective immunogen, thereby being cross-reactive.
This extensive cross-reactivity pattern of the different IgG
subclass-specific mAbs was not realized before, although minimal
cross-reactivity was detected for anti-IgG1 and -IgG2 mAbs particularly
(15,17). Only a few studies have validated the modified single antigen
beads (SAB) assay procedure for the screening of IgG subclass DSA
(14–17,21). The validation of the modified SAB assay relies on several
parameters, such as custom IgG subclass-specific coated-beads, IgG
subclass-specific mAbs concentrations, that vary across HLA
laboratories. Nonetheless, this study supports the use of Luminex®
multiplex platform as an effective approach to assess IgG subclass
cross-reactivity of currently available and newly generated mAbs.
In allograft recipients, several studies have demonstrated a clinical
impact of IgG subclass DSAs on allograft survival, outcomes, and adverse
events, where IgG3 DSAs have been shown to have the most deleterious
impact. Indeed, the presence of IgG3 DSAs was associated with acute
ABMR, shorter time to rejection, allograft failure (10), poor graft
survival (21), and biopsies positive for C4d (17) in adult kidney
recipients, and with a higher risk of graft dysfunction in pediatric
kidney recipients (22). In adult liver recipients, IgG3 DSAs were
associated with an increased risk of graft loss (16) and with death
(23). These findings are not impacted by the use of the anti-human-IgG3
mAb (clone: HP6050) since it is mono-specific to IgG3 proteins, and thus
considerably reduces the rate of false-positive IgG3 recipients.
However, IgG3 DSA-positive recipients will most likely be IgG1
DSA-false-positive when cross-reactive anti-human-IgG1 mAbs (clone:
HP6001 and 4E3) are used at high concentration or when serum
concentration of IgG3 DSA is high.
Regarding IgG4 DSAs, their presence pre-transplant, in kidney
recipients, was an independent predictor of early rejection, graft
failure, and graft survival (21), while their presence post-transplant
was associated with subclinical ABMR (10). In pediatric liver
recipients, IgG4 DSAs are associated with a histopathological and
transcriptional features indicative of active controlled alloimmune
response (T cell mediated rejection) (14). The IgG4 DSA false-positive
rate in these findings should be low since the anti-human-IgG4 mAb
(clone: HP6025) cross-reacts with IgG2 proteins at high concentrations
only, and possibly because IgG2 do not constitute a large proportion of
IgG DSAs (8). In contrast, IgG1 and IgG2 false-positive rates may be
more impacted by our findings, where IgG1 and IgG2 results should be
viewed in the context of cross-reactivity, as previously suggested (17).
Although recipients with IgG1 DSAs only can clearly be identified by the
anti-IgG subclass mAbs pattern of reactivity described in this study,
and IgG1 DSAs unequivocally represents the predominant HLA subclass
(10,14–17,19–22,24), recipients with IgG2 and/or IgG3 DSAs are likely
to exhibit IgG1 DSA false positivity as a result of anti-IgG1 mAbs
cross-reactivity (Table 2). Similarly, anti-IgG2 mAbs exhibit the
greatest cross-reactivity, particularly clone 31-7-4 (14), and patients
with IgG1, IgG3, and/or IgG4 DSAs are likely to be IgG2
DSA-false-positive. In support, IgG2 DSA-positive only recipients are
absent (10,17,21,24) or extremely rare (11,15,20,22). Therefore, the
distribution of IgG subclass DSAs in transplant recipients should be
reevaluated in light of our findings, in order to confirm, for example,
whether the presence of specific IgG subclass DSAs are associated with
the type of immunization (transplantation, transfusion, pregnancy) in
HLA-sensitized patients (15,19), or whether the presence of multiple IgG
subclass was associated with chronic rejection in adult liver recipients
(16). Overall, our study confirms that IgG subclass reagents,
particularly IgG subclass-specific mAbs, must be optimized before the
implementation of IgG subclass DSA analyses in clinical practice (10).
Differences in the generation of IgG subclass-specific beads, anti-IgG
subclass mAb concentration, and anti-IgG subclass mAb clone selection
could explain the differences in MFI reported in different
cross-reactivity assessments (14–17,21). Nonetheless, several
approaches may contribute to the validation and standardization of IgG
subclass DSA screening in anticipation of newly developed anti-IgG
subclass mono-specific mAbs. First, our study provides the detailed
description for the generation of beads covalently coated with specific
IgG subclass proteins. This process facilitates the control of the
amount of IgG subclass protein coated onto the beads, which is an
important factor in the observed cross-reactivity of the different
anti-IgG subclass mAbs in our study. Covalent and non-covalent
IgG-protein coating are two methods commonly used to generate IgG
subclass-specific coated-beads but parameters, such as the amount of IgG
subclass protein coated onto the beads, are usually not reported.
Secondly, there is no consensus for the working concentration of each
IgG subclass mAbs with maximum of an 80-fold difference between reports
(Table 1); however, our study demonstrates that the concentration of the
IgG subclass mAbs is a critical factor that determines the relationship
between maximum MFI detection capacity (sensitivity) and IgG subclass
protein cross-reactivity (specificity). We demonstrated, for
cross-reactive IgG subclass mAbs, that a high concentration increases
the sensitivity while reducing the specificity, conversely a low
concentration reduces the sensitivity while increasing the specificity.
Moreover, our results confirm that the affinity and avidity varies
between anti-IgG subclass mAbs, as the direct MFI comparison of each mAb
at different concentration does not reflect relative amount of IgG
subclass protein, and thus supports the use of a unique working
concentration for each anti-IgG subclass mAbs. Thirdly, our study
suggests that the use of clones HP6001, HP6014, HP6050 and HP6025 to
detect IgG1, IgG2, IgG3, and IgG4 DSAs, respectively, is the optimal
anti-IgG subclass mAbs combination that maximizes the sensitivity and
specificity of subclass DSA screenings using Luminex® SAB assays. In
terms of cross-reactivity, only the anti-IgG4 mAbs, HP6025 and HP6023,
did not exhibit notable differences; however, HP6025 seems most suitable
to use with HP6001, HP6014, and HP6050. The affinity of HP6025 for its
target is closer to that of other IgG subclass mAbs, whereas HP6023
appears to have a higher affinity since the concentration of HP6023 was
lowest to reach saturation of the beads compared to other IgG subclass
mAbs; moreover, HP6025 was used more frequently to detect IgG4 DSAs,
thus providing consistency for the comparison of MFIs with the reports
of previous studies. Last, the disparity in the methods defining the
threshold for IgG subclass DSA positivity warrants further
investigations. Yet, the appropriate threshold for positivity should be
determined for each IgG subclass mAb separately. In spite of the
well-known limitations of the SAB assays (25,26), variations in the
modified protocols for the IgG subclass DSA screening impair the direct
comparison of different studies’ results, particularly for MFI
strengths.
The Luminex® SAB assay has not been approved as a quantitative assay by
the US Food and Drug Administration to measure the concentration of DSAs
(27). However, it is evident that allo-sensitization, defined by a
marked increase of IgG DSA production, is detected by the SAB assay.
Extensive serum titration analyses revealed that some DSA still show
high strength (MFI>15,000) at a serum dilution of 1:1024
(28). Similarly, at a serum dilution of 1:16, Navas et al. demonstrated
that the IgG subclass DSA strength correlated with the ability to bind
complement particularly when IgG1 DSAs were accompanied with IgG2 DSAs
(20). In addition, C1q-binding DSAs usually exhibit high titers in the
regular SAB assay (28), and anti-IgG2 mAb (clone: 31-7-4, at 20µg/mL)
cross-reacted with IgG1 proteins at high concentrations (1µg and 2µg per
1x106beads); therefore, it suggests that serum DSA
concentrations (or density of bead-bound DSAs) clearly impact the
cross-reactivity level of IgG subclass mAbs and prevent the clear
distinction of patients’ IgG subclass DSA distribution. Furthermore,
weak DSAs (MFI<2,000) detected with the regular SAB assay, can
be negative in the modified IgG subclass specific SAB assay (8),
particularly if the concentration of the anti-IgG subclass mAb is low.
Similarly, weak HLA antibodies (MFI<5,000) detected with the
conventional PE-conjugated polyclonal anti-human-IgG had a significantly
lower MFI when detected with a PE-conjugated anti-human-IgG mAb, while
the opposite was true for de novo DSA particularly (18). In spite of the
extreme variability of serum IgG subclass DSA concentrations and the
impact on DSA detection, the goal in validating the use of PE-conjugated
mAbs in IgG subclass DSA screening is directly related to the serum
dilutions used for testing, where optimal conditions increases the
sensitivity and decreases the cross-reactivity of the assay. The
refinement of patients’ IgG subclass distribution assessment can benefit
from serum dilution experiments, despite the cost-prohibitive nature of
serum dilution screening using the SAB assay.
This empirical analysis of IgG subclass mAbs cross-reactivity has a few
limitations. First, interfering factors present in the serum may affect
the cross-reactivity of IgG subclass mAbs. A recent study purified serum
IgG to remove factors that can interfere with the SAB assay and thus may
enhance MFI signal of DSA (14); however, the purification of serum IgG
reveals an extensive HLA reactivity that was masked in the native sera
(29), which may affect IgG subclass DSA distribution of purified serum
IgG samples and further studies are needed to address this problem.
Second, besides the <90% homology between IgG subclasses,
which makes the generation of mAbs delicate for producing mono-specific
ones, the presence of different IgG subclass allotypes further
complicates this problem. Since different IgG subclass allotypes may be
implicated in their immunogenicity by inducing allotype-specific
antibodies (30), it is important to verify that each IgG
subclass-specific mAbs recognize each allotypes equally, and that there
is no impact on IgG subclass cross-reactivity. Last, we have tested only
one lot for each IgG subclass mAbs, but Cicciarelli et al. have
suggested that different lots of mAbs may have different pattern of
cross-reactivity (17). Indeed, the anti-IgG1 (clone: HP6001) was shown
to be cross-reactive to IgG4, and, interestingly, we observed
cross-reactivity to IgG2 and IgG3, but not to IgG4, for the same clone
from a different source.
In conclusion, we have characterized and validated the IgG1-4 subclass
cross-reactivity pattern of 8 anti-IgG subclass-specific mAbs. To our
surprise, only one anti-IgG3 specific mAb (clone: HP6050) was
mono-specific, whereas all other anti-IgG subclass mAbs were
cross-reactive. Importantly, IgG subclass mAb cross-reactivity is
dependent on both their concentration and the concentration of the
IgG1-4 subclass protein targets. Despite the observed cross-reactivity
of certain mAbs, unique patterns of reactivity, which correspond to a
unique IgG subclass composition, can be identified. Although the impact
of IgG subclass DSAs on allograft outcomes that was previously
demonstrated is not critically influenced by our data, the dilution of
serum samples and IgG subclass-specific mAbs must be further optimized
before their implementation of IgG subclass DSA screening in allograft
recipients.