Clinical immune diagnostics are currently confined to the detection of specific antibodies in serum and other bodily fluids. Diagnostic decisions on whether a person has been infected by, and subsequently developed immunity, e.g., to a particular virus or vaccination, and the magnitude of the ensuing immune response are presently deduced from measurements of serum antibody titers. In this chapter we will describe why it is necessary to include memory B cell (Bmem) measurements into such considerations, and how this can be reliably accomplished with minimal cell material and labor by performing B cell ELISPOT/FluoroSpot, collectively “ImmunoSpot®” assays (Note 1).
There are conceptual and practical reasons why, thus far, mainstream efforts towards monitoring antigen-specific Bmem have not been sufficiently undertaken. Namely, it has largely been accepted that antibody titer measurements directly reflect Bmem frequencies. This notion has been based on the antiquated assumption that during the B cell response to antigen, long-lived Bmem and long-lived plasma cells (PCs) arise in a fixed ratio to each other. Thus, the feeling had been that cumbersome assays to detect Bmem utilizing fragile live cell material are unnecessary because simple measurements of stable proteins in serum reliably provide the sought-after information. However, recent advances in the understanding of B cell lineage differentiation pathways implies the requirement for a different approach. This topic has been the subject of several excellent reviews recently [1–3], and is only outlined briefly in the following section.
1.1 Differential fate decisions for B cell differentiation into the memory vs. plasma cell lineages
During the antigen-driven B cell response, within a germinal center (GC) present in the draining lymph node(s), Bmem and PCs arise along differential, affinity-driven maturation pathways (Fig. 1). Briefly, when germinal center B cells (GCB) undergo proliferation and acquire somatic hypermutations (SHM), (Note 2) daughter cells arise with mutated antigen-binding sites of their B cell antigen receptors (BCR) (Note 3). Because SHM occur somewhat randomly, a subset of these mutated BCR can acquire an increased affinity for the antigen, although the affinity of most BCR is either unaltered or attenuated as a result of these mutations. In subsequent steps, the progeny of GCB that gained an increased affinity for the antigen undergo additional cycles of proliferation, SHM, and affinity-based positive selection. Multiple repetitions of this process eventually lead to the differentiation of PC which constitute the cellular basis of the affinity-matured antibody responses. Progeny of GCB that do not meet the increasingly stringent antigen-driven affinity selection criterion (Note 4) exit the GC and become long-lived Bmem.
Bmem and PCs therefore emerge along alternative routes driven by different selection criteria, and subsequently their frequencies are not necessarily linked. Consequently, one cannot expect antigen-specific serum antibody levels and Bmem frequencies to be proportional to each other. From the above understanding of antigen-driven B cell differentiation it also follows that, while both PC and Bmem are antigen-specific (that is, their BCR exceeds a minimal binding constant for antigen), PC (and hence secreted antibodies) constitute primarily the high affinity end of this affinity spectrum, while Bmem also encompass a lower affinity fraction. Recent interpretation of these findings even implies that PC and Bmem play fundamentally different roles in humoral immune defense [1].
1.2 Serum antibodies reflect the first wall of adaptive humoral defense, Bmem the second
While PC elicited during the primary immune response can secrete large amounts of antibodies, their lifespans are heterogenous and, contrary to the previous assumption that PC are long-lived, likely fall on a continuum with possibly only a fraction of them surviving long term [4, 5]. The antibody molecules they secrete are also relatively short-lived in the body and possess half-lives of several weeks, at best (Note 5). Thus, serum antibodies that are detected at any one time have been recently produced, and therefore the maintenance of serum antibody levels depends on constant active replenishment by PC. Such antibodies in bodily fluids and their cell-bound variants (Notes 5–7) constitute the first wall of the adaptive humoral immune defense since they are already present prior to antigen reencounter. Such antibodies (IgA and IgM) are transported across the mucosa and can confer instant host protection by preventing entry of the pathogen/antigen into the body. If the offending pathogen/antigen succeeds in crossing this interface and gains access into the body, pre-formed antibodies (IgG and IgM) already present in serum can still confer protection through direct neutralization. Serum antibodies also possess precipitating, opsonizing, and complement activating activity to further combat dissemination of the pathogen/antigen. Pre-formed antibodies can additionally initiate inflammatory reactions through their interactions with Ig-binding receptors expressed on various immune cell populations (IgE, IgG or IgM).
Many times, however, the first wall of adaptive humoral immune defense fails to prevent reinfection. Such is the case when antibody levels either decline to sub-protective levels, or when antigenic viral variants are encountered that evade neutralization, as evidenced, e.g., in the recent COVID pandemic. Boosting the levels of specific antibodies that were already established through infection or vaccination against the previously circulating virus strain, the “homotype”, will not confer sterilizing immunity against a newly emerging antigenic variant strain, the “heterotype”. In this case, Bmem provide the second wall of adaptive humoral immune defense. Within the Bmem pool established during a preceding immune response, BCR specificities possessing a reduced affinity for the homotypic virus/antigen will be present that, by chance, have an increased affinity for the variant, heterotypic virus/antigen. As such heterotypic antigen-reactive Bmem occur at higher frequencies than would occur within a B cell repertoire that is naïve to the homotypic virus/antigen; moreover, as such Bmem have already undergone immunoglobulin (Ig) class switch recombination (CSR), they can readily engage into faster and more efficient recall responses, including potential re-entry into a GC for acquisition of further affinity-enhancing SHM. While this Bmem-based second wall of adaptive humoral immunity is not able to prevent infection with a heterotypic virus, it still mediates immune protection against it by enabling the host to rapidly mount a secondary-type immune response at the first encounter of a heterotypic virus.
Studying antigen-specific serum antibodies vs. Bmem, therefore, provides fundamentally different information on adaptive humoral immunity. The former reflects only on the past remnants of previously established, still deployable, but passive, and fading immunological memory. In contrast, studies of Bmem provide insights into the ability of an individual to actively mount secondary immune responses against homo- and/or heterotypic antigens. Thus, studying Bmem permits one to take a glance into the future.
1.3 Emerging evidence for serum antibody levels and memory cell frequencies providing divergent information on the magnitude of B cell-mediated immunity
Recently, we concluded a systematic study (to the best of our knowledge the first on this subject) in which we measured circulating antibody levels to SARS-CoV-2, EBV, and different seasonal influenza virus strains and compared them with the frequencies of Bmem reactive with the same antigens [6]. These ImmunoSpot®-based Bmem detections were enabled by our newly-acquired ability to achieve high density antigen coating, thus reliable detection of Bmem specific for essentially any antigen ([11] and Note 8). Representative results for the SARS-CoV-2 Spike protein are shown in Fig. 2. Notably, elevated frequencies of Spike-specific, Bmem-derived antibody-secreting cells (ASCs) were detected in all of these convalescent COVID-19 donors despite variable levels of IgG antibody reactivity in the plasma at the time of sample collection. In this same study [6], discordance between antibody levels and Bmem-derived ASC frequencies was also noted for the SARS-CoV-2 NCAP antigen and multiple antigenically unrelated seasonal influenza strains, all of which represent respiratory viruses that cause acute, short-term infections. Furthermore, discordance between antibody levels and Bmem-derived ASC frequencies was also evidenced against the EBNA1 antigen of Epstein-Barr virus (EBV), a latent herpesvirus that occasionally reactivates [7].
1.4 Emerging evidence for Bmem detection being more reliable for revealing past antigen exposure than serum antibody measurements
It is not uncommon to find individuals who exhibit little if any seropositivity for an antigen, yet possess significant, often high, frequencies of antigen-specific, Bmem-derived ASC (Fig. 2, [6], and Kirchenbaum et al., manuscript in preparation). Among the antigens tested, SARS-CoV-2 proteins are the most informative in this regard because SARS-CoV-2 Spike and NCAP antigen-specific Bmem-derived ASC are not detected in samples from individuals cryopreserved prior to the onset of the COVID-19 pandemic (which therefore are verifiably immunologically naïve to this virus owing to the time period in which these samples were collected). In contrast, each of the seronegative, yet Bmem-positive, donors in our study cohort had undergone PCR-verified SARS-CoV-2 infections (Note 9). In these instances, serum antibody measurements provided clearly false negative results on the infection history of these individuals, while the presence of Bmem-derived ASC reliably revealed it.
We have also previously noted clearly false negative serological assessments for human cytomegalovirus (HCMV) exposure [8]. Obtaining PBMC (peripheral blood mononuclear cells) from different FDA-approved blood banks, all of which screen their donors for HCMV seropositivity, we found that PBMC from many such HCMV-seronegative donors possessed HCMV antigen-reactive, Bmem-derived ASC in ImmunoSpot® assays following in vitro polyclonal stimulation. Such subjects also exhibited CD4+ and CD8+ T cell memory to HCMV. Thus, the presence of Bmem-derived ASC reactivity (and T cell memory) more reliably revealed the HCMV exposed status of these donors than did serum antibody levels.
In yet another independent line of investigation, studying patients with multiple sclerosis (MS), an autoimmune disease of the central nervous system (CNS), we detected neuroantigen-specific Bmem, in the absence of detectable serum antibody levels in most patients, whereas such Bmem were absent in healthy controls [9]. Collectively these findings suggest that detecting Bmem-derived ASC might be a more reliable way to diagnose past infections and possibly also autoimmune diseases than afforded presently by serum antibody measurements.
1.5 Assessing Bmem might be important, but how, and why?
Presently there are two major techniques that permit the detection of antigen-specific Bmem in blood, lymphoid tissues, and other primary cell material. One approach is based on labeling B cells with fluorescently-tagged antigen(s) followed by their detection using flow cytometry; the other is by ImmunoSpot®. While the strength of the former approach is that it allows for segregation of antigen probe-binding B cells into phenotypically distinct subsets based on their surface marker expression, it has several disadvantages compared to ImmunoSpot®. One crucial difference is the lower limit of detection of the former. While ImmunoSpot® enables measurement of a single ASC within a bulk population of cells with essentially no intrinsic lower detection limit (Note 10), flow cytometry falls short for detecting antigen-specific B cells when they are present as low frequency events. Bmem, however, frequently occur in very low frequencies in PBMC (Fig. 3). Additionally, the number of PBMC required for flow cytometric detection of antigen-specific B cells is considerably higher compared to ImmunoSpot® (Note 11). Notably, a surface staining-only approach for identification of antigen-specific B cells does not reliably reveal the Ig class/subclass usage of ASC, while 4-color ImmunoSpot® assays provide this key information with ease as part of routine Bmem frequency measurements (Note 12). Lastly, unlike for flow cytometry, the actual wet lab implementation of B cell ImmunoSpot® assays is scalable for high-throughput analysis (Note 13) and multi-color ImmunoSpot® analysis can be fully automated (see chapter by Karulin et al. in this volume, [10]).
1.6 Affinity coating: enabling B cell ImmunoSpot® assays “to see”
With so many potential advantages in favor of ImmunoSpot® assays for the detection of Bmem, the question arises why this technique has not become more widely used for immune monitoring purposes. The answer is simple and practical: the original protocol (that involves direct coating of the antigen to the membrane) only works well for a rather limited set of antigens (Note 14). Consequently, many investigators likely gave up on this assay after not being successful in their initial attempts to establish such assays for the antigen(s) of interest. Our recent introduction of affinity capture coating [11] represents a breakthrough to this end, as it provides a universal strategy for successful assay development for essentially any antigen: the membrane is first coated with an anti- (His- or other) affinity tag-specific antibody followed by the addition of recombinant (His- or other) affinity-tagged recombinant antigen. In this way, low-affinity absorption of the antigen to the membrane via weak, non-specific binding forces (primarily hydrophobicity) is replaced by specific, high-affinity binding. The assay principle is depicted in Fig. 4B, and the protocol is described in detail below. This version of the B cell ImmunoSpot® assay is particularly well-suited for detecting and characterizing rare antigen-specific, Bmem-derived ASC in a test sample, whereas an alternative variant of the B cell ImmunoSpot® assay approach, which enables assessment of ASC functional affinity and is described in another chapter of this volume [12], is better suited for samples in which antigen-specific ASC are present at an elevated frequency among all ASC (Note 15).
1.7 High content information provided by assessing individual ASC via ImmunoSpot® vs. serum antibody measurements
Owing to the single-cell resolution afforded by ImmunoSpot® assays, they are ideally-suited to study individual ASC that comprise an antigen-specific B cell response. While traditional serum antibody measurement techniques readily perceive large increases in antigen-specific antibody titers, they fail to appreciate more subtle changes in antibody levels, especially when the abundance of antibody is very small, e.g. in the context of allergen-specific IgE [13], or when an elevated level of preexisting antibody reactivity is already present, e.g. in the context of studying seasonal influenza vaccine responses [14], [15]. In contrast, B cell ImmunoSpot® assays circumvent ambiguity in both cases through quantifying the precise number of cells that are actively secreting antigen-specific antibody and thus offers increased sensitivity and resolution.
When optimal antigen coating conditions exist in the direct B cell ImmunoSpot® assay, each secretory footprint will be composed of antibody originating from a single ASC and thus permits assessment of antibody affinity similar to that achieved through studying an individual monoclonal antibody (mAb). In this setup, the morphology of an ASC-derived secretory footprint is primarily defined by the affinity of the secreted antibody for the membrane-bound antigen. In agreement with prior computational modeling [16], we routinely observe variable spot morphologies in such antigen-specific B cell ImmunoSpot® assays [6, 11]. Importantly, beyond enumeration of such antigen-specific secretory footprints, a multitude of morphological features including metrics of intensity and size are also captured for each spot-forming unit (SFU) and are readily exported as flow cytometry standard (FCS) files by the ImmunoSpot® software. Such FCS files can be leveraged to visualize variable spot morphology, as was done using scatter plots in our recent publications [6, 11]. Here we direct the reader to the chapter by Karulin et al. in this volume dedicated to high content analysis of spot morphologies in this issue [10]. Additionally, in the context of an inverted B cell ImmunoSpot® assay (Note 15) in which ASC-derived secretory footprints are efficiently captured irrespective of antigen specificity, secretory footprints originating from individual antigen-specific ASC are revealed by their ability to retain an antigen probe. While also briefly introduced below, a chapter in this volume by Becza et al. [12] describes in detail how the inverted ImmunoSpot® assay variant and high content data analysis enable assessment of the functional affinity distribution present in a polyclonal population of ASC, such as the ASC response elicited following COVID-19 vaccination.
1.8 Frequency of antigen-specific, Bmem-derived ASC reflect memory potential
Because ImmunoSpot® assays detect the secretory footprints of individual ASC, frequency information is revealed by counting the numbers of antigen-specific SFU, either per cells plated per well, or, better, as the frequency of antigen-specific ASC secreting a given Ig class/subclass among all ASC producing that Ig class/subclass (Note 16). In either case, however, the crowding of secretory footprints along with the ELISA effect (Note 17) can lead to undercounting of secretory footprints [6]. Systematically studying this phenomenon previously, we found that there exists a range in which cell numbers plated per well are directly proportional to the number of SFU detected. Depending on the nature of the assay and morphology of the resulting SFU, however, the corresponding SFU counts measured at higher cell inputs may break down at approximately 100–200 SFU per well (Note 18). This is a major concern for establishing accurate frequencies of antigen-specific, Bmem-derived ASC, especially in light of our previous data demonstrating that frequencies may span orders of magnitude for the same antigen in different individuals (Fig. 2, and [6]). Also, within any single individual, the frequency of Bmem-derived ASC specific for different antigens can span a similarly wide distribution [6]. Furthermore, even the number of ASC producing a given Ig class/subclass, irrespective of antigen specificity, exhibits considerable inter-individual variation. ELISA effects in B cell ImmunoSpot® assays can also jeopardize high content analysis of spot morphologies. A simple and efficient solution to this problem is to serially dilute the cell input to establish the linear range of SFU counts from which the accurate frequency of ASC can be extrapolated by linear regression ([6], and Note 19). Karulin et al. in this volume [10] introduce software for automatically establishing frequencies from such serial dilution experiments making this approach suitable for high-throughput workflows.
1.9 Defining the Ig class/subclass usage of antigen-specific Bmem
The different Ig classes and subclasses are endowed with distinct effector functions and each contributes non-redundant roles towards maintaining host defense (see above, and reviewed in greater detail in [17]). During the primary immune response, B cells can transition from IgM-expressing naïve B cells into effector cells (PC) and Bmem that have undergone class switch recombination (CSR) [1]. CSR is an irreversible process involving the excision of DNA encompassing the exons of the Igµ heavy chain required for expression of IgM and juxtaposition of the upstream variable region genes (VDJ; which jointly define the antigen specificity of the BCR or secreted antibody) with downstream exons encoding alternative Ig classes or IgG subclasses [18]. Class switching of the BCR to downstream Ig classes or IgG subclasses is an instructed process and can be influenced by the cytokine milieu and co-stimulation provided by CD4+ T helper cells. For the latter, the differentiation of naïve CD4+ T cells into different T helper cell classes (Th1, Th2, Th17, etc.) is defined by the circumstances of antigen encounter, in particular by Toll-like receptor (TLR)-derived signals [19, 20]. Triggering the “appropriate" type of T helper cells capable of stimulating the optimal Ig class usage during an infection or following vaccination is vital to successful host defense and the avoidance of collateral immune-mediated pathology (reviewed in [21]). The same applies for reinfection with the same pathogen.
Upon antigen reencounter, and subsequent reactivation, Bmem rapidly differentiate into PC that secrete the same Ig class/subclass expressed by the parental Bmem. Therefore, by detecting the Ig class/subclass that Bmem-derived ASC produce in B cell ImmunoSpot® assays permits the prediction of the specific types of antibodies that will be produced following the next antigen encounter (Notes 20–22). Learning about the full spectrum of Ig classes/subclasses that the antigen-specific Bmem repertoire will produce when the antigen is reencountered is thus essential for predicting the protective efficacy of future antibody responses to an antigen. Moreover, such B cell ImmunoSpot® assays may also predict the likelihood of antibody-mediated complications occurring upon antigen re-exposure (as illustrated by the example of antibody-dependent enhancement leading to exacerbation of clinical disease following secondary dengue infection [22]). Importantly, the frequency of antigen-specific Bmem capable of secreting different Ig classes or IgG subclasses can be orders of magnitude apart [6] (and see also Yao at al. in this volume [23]). Thus, it is not only recommended to assess all antigen-specific Ig classes/subclasses, but also to do so over a wide frequency range for each. The latter can be readily accomplished by combining the serial dilution strategy described above with 4-color ImmunoSpot® analysis (for details, see the chapter of Yao et al. in this volume, [23]).
1.10 Assessing the affinity distribution of the antigen-specific Bmem repertoire
As described above in this chapter, and briefly recapitulated here, SHM leads to diversification of Bmem cells that previously participated in a GC reaction [24]. Upon reencounter with the same (homo-) or a modified (heterotypic) antigen, previously generated Bmem endowed with high-affinity BCR for the re-challenging antigen will readily differentiate into PC that are capable of rapidly increasing antibody titers [3]. Moreover, Bmem can also be re-recruited into a GC following antigen reencounter, where they will undergo further rounds of proliferation and acquire additional SHM necessary to refine and improve their BCR affinity for the offending antigen. Again, the GCB cell progeny endowed with the highest affinity BCR for the eliciting antigen will be selected for PC differentiation and ideally will enter the potentially long-lived fraction of the PC compartment. Thus, a Bmem repertoire that entails a higher frequency of high-affinity B cells specific for the (homo- or heterotypic) antigen, can generate a faster and more robust anamnestic antibody response upon antigen (re)encounter.
Thus far, the study of BCR/antibody affinity for a particular antigen has largely been confined to the generation of B cell hybridomas and/or paired IgH/IgL sequencing followed by expression and purification of individual mAb in order to establish their specificities and affinity using surface plasmon resonance or biolayer interferometry. While these approaches are the gold standard for studying individual mAbs (i.e., single B cells), it would require generation and subsequent characterization of hundreds of such mAbs to appreciate the underlying affinity distribution of antigen-specific Bmem repertoire in just one donor, at a single time point, and against only a single antigen of interest. While such an exercise is, in theory at least, conceptually possible for a very limited number of individuals, it would be inconceivable both in magnitude and cost to attempt such an objective for a larger donor cohort, e.g., as part of a clinical trial. Consequently, this traditional method for characterizing the affinity distribution of the antigen-specific B cell repertoire is not realizable for high-throughput immune monitoring efforts.
The chapter in this volume by Becza et al. [12] demonstrates that the affinity distribution information of the antigen-specific B cell repertoire can be established with ease, and in a readily scalable manner, using two B cell ImmunoSpot® assay variants. For both, to characterize the affinity distribution of, e.g., vaccine-elicited IgG+ ASC, a first experiment would need to be performed that establishes the frequency of such antigen-specific IgG+ ASC using the serial dilution approach (Note 23). This information is essential for being able to seed in a follow-up test the PBMC at a “Goldilocks” cell density in which the individual secretory footprints can be studied without interfering with each other. In this second experiment, using an independent aliquot of cryopreserved cell material, the donor PBMC are seeded at the so-called “Goldilocks number”, that is assay dependent and between 50–100 SFU per well. As each SFU per well represents the secretory footprint of an individual ASC, at 50 SFU per well, the secretory footprints of 50 distinct ASC would be assessed. Seeding additional replicate wells at this “Goldilocks number” would increase the number of antigen-specific IgG+ ASC being characterized. We recommend to study a fixed number of ≥ 300 ASC for each antigen concentration as this number gives a considerable (and representative) sample size for assessment of the antigen-specific Bmem repertoire elicited in a particular person, against an individual antigen, at the time of sample collection.
Leveraging the inverted ImmunoSpot® approach for B cell affinity distribution measurements, the soluble antigen probe is added in decreasing concentrations to the same number of replicate wells for each antigen concentration, whereby each well was previously seeded with a “Goldilocks number” of cells. At the highest antigen probe concentration, the secretory footprints of all antigen-specific ASC, high- and low-affinity alike, will be detected as SFU when the membrane-retained antigen probe is visualized. In wells containing decreasing concentrations of the antigen probe, however, only the secretory footprints originating from ASC that produced an antibody with high affinity for the antigen probe will retain an adequate amount for their eventual detection as SFU; B cells with affinity lower than dictated by the actual antigen concentration go undetected, and the SFU counts will decrease by their number. Therefore, establishing the SFU counts across a range of antigen probe concentrations provides insights into the affinity distribution of the antigen-specific B cell repertoire.
The second B cell ImmunoSpot® assay variant suited for assessment of the affinity distribution present in a polyclonal, antigen-specific ASC compartment relies on a direct assay in which the antigen coating density is graded. Such assays (Fig. 4) inherently provide information on the affinity of the “monoclonal antibody” that each antigen-specific ASC secretes. Following the basic rules of antibody-antigen binding, ASC that produce high-affinity antibodies generate dense, bright and tight (sharp and small) secretory footprints (“spots”), while ASC that produce antibodies with lower affinity yield secretory footprints that are fainter and more diffuse [16]. Coating of the membrane with decreasing densities of antigen and monitoring the resulting changes in SFU number and morphology therefore provides information regarding the affinity distribution of an antigen-specific ASC repertoire (Note 24). The chapter by Becza et al. in this volume [12] lays out in detail both assay variants for B cell affinity measurements.
1.11 Measuring cross-reactivities of individual Bmem-derived ASC
As described above, Bmem also provide a “second wall of adaptive humoral immunity” against mutated antigens/viruses (heterotypic antigens) that evolved to evade neutralizing antibodies induced by the original (homotypic) virus/antigen. While the primary response leads to fine tailoring of the B cell repertoire directed against the homotypic antigen, including affinity maturation, among the variants created by random SHM, Bmem endowed with BCR possessing affinity for the heterotypic antigen will be present. Even if such Bmem would be rare, and even if their initial affinity would be modest, they still would occur in much higher frequencies compared to the naïve B cell repertoire. Moreover, most of these Bmem have already undergone Ig class switching. Therefore, when such clonally expanded, semi-affinity matured, and class-switched, cross-reactive Bmem engage in a primary immune response against the heterotypic antigen, they can generate a more rapid and efficient antibody response compared to individuals who have not been previously immunized/infected (Note 25). Measuring existing serum antibodies does not detect such cross-reactive Bmem, and thus does not provide predictive information about cross-reactive protection, whereas ImmunoSpot®-based assessments at the level of individual Bmem-derived ASC does.
There are two ways to measure Bmem-derived ASC cross-reactivity by ImmunoSpot®. For both, the appropriate cell input per well first needs to be defined which will provide the “Goldilocks” SFU count for the homotypic antigen (between 50 and 100 SFU per well, see above). In the simple version of B cell cross-reactivity studies, a second experiment is performed, an inverted B cell assay, in which this Goldilocks input of PBMC for the particular donor is seeded into all wells. Setting up replicate wells permits to assess ≥ 300 individual ASC-derived secretory footprints in independent single-color ImmunoSpot® assays, comparing the numbers of SFU detected in all replicates for the homotypic antigen, vs. the cumulative SFU number in the same number of replicate wells detected using the heterotypic antigen probe. The ratio of these two numbers reveals the percentage of cross-reactive ASC at equimolar concentrations of the respective antigen probes (Note 26, and Fig. 5). Raising/lowering the concentration of heterotypic antigen probe enables assessment of the affinity spectrum of Bmem-derived ASC that cross-react with the heterotypic antigen.
In a second, somewhat more complex, but perhaps more elegant approach, Bmem-derived ASC cross-reactivity is evaluated using alternatively-labelled homotypic and heterotypic antigens. As before, the PBMC are plated at the Goldilocks input number into replicate wells of an inverted B cell ImmunoSpot® assay. The “tag 1” labelled homotypic antigen probe (e.g. His tag) and the “tag 2” labelled heterotypic antigen probe (e.g., FLAG tag) are added simultaneously, at equimolar concentrations, followed by dual-color detection of the SFU-bound antigen probes. The ratio of “tag 1” single-color positive SFU and of “tag 1 + tag 2” double-color positive SFU is established, revealing the frequency of cross-reactive ASC under equimolar conditions. Raising and lowering the concentration of the heterotypic antigen probe provides insights into the affinity of the individual homotype-primed Bmem for the heterotypic antigen.
1.12 Concluding remarks
The aim of this chapter was to draw attention to the tremendous, thus far largely unrealized, potential of B cell ImmunoSpot® assays, and their potential to revolutionize immune diagnostics. Although the ELISPOT assay was originally introduced 40 years ago for detecting antigen-specific ASC [25, 26], it was T cell ELISPOT assays (introduced 5 years later [27]) that took the limelight. For the latter to occur, however, we needed to introduce fundamental modifications to the originally described protocol so that the assay lived up to its potential and was capable of reliably revealing secretory footprints of individual T cells responding to antigen ([28], and Note 27). Our introduction of automated, objective, and scientifically validated software-assisted machine reading of T cell-derived secretory footprints [29] also largely contributed to the success of T cell ImmunoSpot® assays (Note 28).
Despite being introduced earlier into the literature, B cell ImmunoSpot® assays have, thus far, not become a mainstay in B cell immune monitoring efforts. One likely reason is that for most antigens the classic protocol of directly absorbing the antigen to the assay membrane simply does not work (see above); only our recent introduction of the affinity coating approach [11] enables one to rapidly develop ImmunoSpot® assays for detecting ASC with specificity for essentially any affinity-tagged antigen of interest. But perhaps the primary reason why B cell ImmunoSpot® assays have not been sufficiently pursued until now is the widely held, and flawed, assumption that serum antibody measurements yield sufficient insight on underlying B cell-mediated adaptive humoral immunity. As outlined above, from the basic science perspective, this standpoint is no longer tenable. Importantly, a major component of the critical information pertaining to Bmem-mediated immunity is now easily attainable through implementation of B cell ImmunoSpot® assays, as outlined above, and is neglected by studying antigen-specific serum reactivity/titers alone.
Ease of implementation, low cost, and moderate labor investment are major requirements for the success of any assay. The efficiency of B cell ImmunoSpot® in PBMC utilization has been repeatedly highlighted in this chapter (see also Note 23) and is critical in clinical trial settings. Cryopreserved PBMC can be used without impairing B cell function [30], permitting batch testing of dozens of samples by a single investigator in a single experiment (and hundreds of samples with a well-trained team). Fully automated image analysis of the assay results is now available, along with retention of audit trails [10]. The reproducibility of B cell ImmunoSpot® is quite remarkable for a cellular assay [6], and the serial dilution strategy detailed above permits to extend the upper and lower detection limits of the assay over orders of magnitude [6]. Moreover, B cell ImmunoSpot® assays have been shown to be suitable for regulated testing [31]. For all of the above reasons, we believe B cell ImmunoSpot® testing will soon become an indispensable component of the immune monitoring repertoire.
Detailed methodology for the affinity capture-based B cell ImmunoSpot® assay is described in another chapter contributed by Yao et al. in this volume [23]. This is the method of choice for establishing frequencies of antigen-specific ASC via the serial dilution strategy; including for all Ig classes, or IgG subclasses, simultaneously. Detailed methodology for the single-color, antigen-specific inverted B cell ImmunoSpot® assay described in Fig. 5 is detailed below. Both the direct and inverted ImmunoSpot® assay formats are also well-suited for assessment of ASC cross-reactivity. However, while direct ImmunoSpot® assays provide insights into the affinity distributions through studying spot morphologies, the inverted assay described in detail by Becza et al. [12] is the method of choice for such high-content analysis of antigen-specific ASC repertoires. Software solutions for B cell ImmunoSpot® analysis are detailed in another chapter of this volume by Karulin et al. [10].