Therapeutic Potential of HLA-I Polyreactive mAbs Mimicking the HLA-I Polyreactivity and Immunoregulatory Functions of IVIg

HLA class-I (HLA-I) polyreactive monoclonal antibodies (mAbs) reacting to all HLA-I alleles were developed by immunizing mice with HLA-E monomeric, α-heavy chain (αHC) open conformers (OCs). Two mAbs (TFL-006 and TFL-007) were bound to the αHC’s coated on a solid matrix. The binding was inhibited by the peptide 117AYDGKDY123, present in all alleles of the six HLA-I isoforms but masked by β2-microglobulin (β2-m) in intact HLA-I trimers (closed conformers, CCs). IVIg preparations administered to lower anti-HLA Abs in pre-and post-transplant patients have also shown HLA-I polyreactivity. We hypothesized that the mAbs that mimic IVIg HLA-I polyreactivity might also possess the immunomodulatory capabilities of IVIg. We tested the relative binding affinities of the mAbs and IVIg for both OCs and CCs and compared their effects on (a) the phytohemagglutinin (PHA)-activation T-cells; (b) the production of anti-HLA-II antibody (Ab) by B-memory cells and anti-HLA-I Ab by immortalized B-cells; and (c) the upregulation of CD4+, CD25+, and Fox P3+ T-regs. The mAbs bound only to OC, whereas IVIg bound to both CC and OC. The mAbs suppressed blastogenesis and proliferation of PHA-activated T-cells and anti-HLA Ab production by B-cells and expanded T-regs better than IVIg. We conclude that a humanized version of the TFL-mAbs could be an ideal, therapeutic IVIg-mimetic.


Introduction
Personalized, passive immunotherapy encompasses the transfer of monoclonal antibodies (mAbs), polyclonal antibodies (Abs), and purified natural Abs for the specific purpose of either upregulating or suppressing immune functions, depending on the nature and status of the underlying disease. A variety of mAbs, capable of performing specific functions, are available. These include antibodies capable of depleting various cell types such as B-cells (rituximab), CD52-bearing leukocytes (alemtuzumab), and CD33-bearing cells (gemtuzumab). Additionally, mAbs can inhibit various cytokine pathways such as the TNF-TNFR axis (etanercept, adalimumab, infliximab) and the IL-6-IL-6R axis (tocilizumab, actemra). In contrast, therapeutic intravenous immunoglobulin (IVIg) contains purified natural Abs, prepared by pooling and purifying IgG from the plasma of 10 to 60 thousand normal and healthy donors. The United States Food and Drug Administration (FDA) has approved immunoprophylaxis with IVIg therapy for chronic inflammatory demyelinating polyneuropathy (CIDP), immune thrombocytopenic purpura (ITP), primary immunodeficiency (PI), secondary immunodeficiency in chronic lymphocytic leukemia, pediatric human immunodeficiency virus (HIV) infection, Kawasaki disease, prevention of graft vs. host disease in adult bone marrow transplant recipients, and organ transplantation [1].
The presence of pre-existing Abs in a potential organ transplant recipient is referred to as sensitization. It remains one of the significant immunological barriers to receiving an organ. As well as restricting the availability of a given organ for a particular patient, these antibodies cause long-term graft attrition when present post-transplantation [2,3]. IVIg is administered to end-stage organ disease patients pre-transplantation and allograft recipients post-transplantation for reduction (desensitization) of these HLA Abs. The exact mechanism of desensitization by IVIg has not been elucidated. However, IVIg per se contains polyreactive, natural IgGs against multiple targets, including endogenous and exogenous Abs, immunomodulating peptides, blood group antigens, and various cytokines. Whereas some of the immunoregulatory mechanisms of the action of IVIg have been proven in vitro and animal models, many proposed mechanisms remain enigmatic due to IVIg polyantigen reactivity, polyclonality, and diversity in the preparations of IVIg. Several mechanisms of action of IVIg have been proposed [4]. These include (a) Fc-receptor blockade, (b) neutralization of Abs by idiotypic and anti-idiotypic Abs, (c) blockade of the Fas apoptotic pathway by anti-Fas auto-Abs, (d) regulation of complement components, (e) modulation of cytokine secretion, (f) hindrance of natural-killer cell activity, (g) inhibition of matrix metalloproteinase-9, (h) suppression of NF-kB activation and IkB degradation, (i) G1 cell cycle arrest, (j) prevention of tumor growth, (k) decrease in leukocyte recruitment, (l) attenuation of T-cell stimulation, (m) effects on Ab kinetics, and (n) effects on dendritic cells. These mechanisms of IVIg are not mutually exclusive and operate synergistically.
Additionally, more observations have accrued to document that normal, healthy human sera contain a diverse variety of IgG Abs against allo-HLA molecules [5][6][7][8][9]. Therefore, the IVIg purified from the pooled plasma of thousands of healthy individuals clearly carries allo-HLA Abs.
Our research commenced with studying the structure of a non-classical HLA class-I (HLA-I) antigen, namely HLA-E and anti-HLA-E mAbs [10,11]. Anti-HLA antibodies do not react with an entire antigen; like all antibodies, the reaction is against a portion of the antigen termed an epitope. At the center of the epitope is an eplet consisting of one to several amino acids recognized by the complementarity-determining regions of the antibody. Epitopes may be specific to a single HLA locus or HLA allele (private epitopes, vide infra). Other times, the epitope may be shared by other loci or alleles (public epitopes, vide infra). The unique and shared amino acid sequences of the diverse HLA-I antigens were identified to elucidate whether the mAbs generated by immunizing an HLA-I antigen recognized public or private epitopes. In addition, the relative immunoregulatory capability of antibodies against either public or private epitopes deserves consideration. The objective of this review is to compare the nature and functional characteristics of the polyreactive anti-HLA-E mAbs (TFL-006 and TFL-007) with those of the commercial preparations of therapeutic IVIg as follows:  Table 1 shows both HLA-E amino acid sequences that are specific and those that are shared with alleles of other loci of HLA-I. Some amino acid sequences or epitopes are specific to the HLA-E locus because they are not found in other HLA-Ia or Ib loci. The most striking private epitopes or sequences are 65 RSARDTA 70 and 143 SEGKSNDASE 152 . Table 2 summarizes the results of comparing HLA-E sequences with hundreds of alleles from the HLA-Ia and HLA-Ib loci. Some sequences of HLA-E are found in only one allele. The HLA-E sequences, or PRAPWMEQE, and EPPKTHVT are found in HLA-A*3306 and HLA-B*8201, respectively, whereas HLA-E sequences 117 AYDGKDY 123 and 126 LNEDLRSWTA 135 are shared with all other HLA loci (HLA-A, HLA-B, HLA-C, HLA-F, and HLA-G) that we have examined. These latter sequences perfectly befit the definition of "shared" or "public epitopes." Table 1. Amino acid sequences of HLA-E compared with HLA-F and HLA-G. There are several sequences restricted to HLA-E, as illustrated by a short sequence, 66 SARDT 70 , and a long sequence, 143 SEGKSNDASE 152 . HLA-E shares several sequences with HLA-F and HLA-G. In addition, HLA-E may selectively share several sequences with HLA-G and a few sequences ( 175 GKETL 179 and 193 PISDH 197 ) with HLA-F. Above all, HLA-E shares 117 AYDGKDY 123 and 126 LNEDLRSWTA 135 with all other HLA isoforms (HLA-A, HLA-B, HLA-C, HLA-F, and HLA-G).
Vaccines 2021, 9, x 4 of 33 Table 1. Amino acid sequences of HLA-E compared with HLA-F and HLA-G. There are several sequences restricted to HLA-E, as illustrated by a short sequence, 66 SARDT 70 , and a long sequence, 143

HLA-E G S H S L K Y F H T S V S R P G R G E P R F I S V G Y V D D T Q F V R F D N D A A S P R M V P R A P HLA-G G S H S M R Y F S A A V S R P G R G E P R F I A M G Y V D D T Q F V R F D S D S A C P R M E P R A P HLA-F G S H S L R Y F S T A V S R P G R G E P R Y I A V E Y V D D T Q F L R F D S D A A I P R M E P R E P
a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 a1 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

HLA-E W M E Q E G S E Y W D R E T R S A R D T A Q L F R V N L R T L R G Y Y N Q S E A G S H T L Q W M H G HLA-G W V E Q E G P E Y W E E E T R N T K A H A Q T D R M N L Q T L R G Y Y N Q S E A S S H T L Q W M I G HLA-F W V E Q E G P Q Y W E W T T G Y A K A N A Q T D R V A L R N L L R R Y N Q S E A G S H T L Q G M N G
a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 HLA  a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2 a2  151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200     Thus, the HLA-I molecules occur as heterodimers complexed with β2m on the cell surface and as β2m-free α-HC OCs. Schnabl et al. [13] reported that T lymphocytes activated in vitro or in vivo, but not resting, expressed many OCs. Using the mAb W6/32 that specifically recognized HLA-I CCs and the mAbs L45 and HC10 that bound explicitly to OCs but not to CCs, they confirmed the existence of HLA-OCs. Immunoprecipitation and co-capping experiments showed that LA45 was bound to HLA-I OCs at the cell surface. The mAb L45 bound to PHA-activated T-cells from a panel of 12 people with different HLA types, suggesting that LA45 might bind to epitopes shared by all HLA-I α-HCs. The HLA-I OCs expressed on the cell surface of activated T-cells-or EBV-transformed Bcells-are referred to as "peptide-binding empty HLA" [14,15]. The presence of HLA-I OCs was confirmed on activated T-cells in vitro and in vivo and on B cell lines (RAJI, NALM6), EBV-transformed B-cells, and the myeloid cell line KG-1A [16,17]. Interestingly, the expression of OCs on the cell surface in normal human T-cells upon activation and cell division correlated with the level of proliferation [18]. Indeed, the OCs on the cell surface were glycosylated. The inhibition of tyrosine phosphorylation with the Src-family tyrosine kinase inhibitor PP2 resulted in the enhanced release of HLA-I α-HC from the cell surface of activated T-cells. Further studies indicated that the OCs were regulators of ligand-receptor interactions and had potential implications for immune activation [18][19][20] and immune regulation [21]. The inflamed tissues in spondyloarthropathy showed increased levels of OCs on CD14b+ monocytes compared to other leukocyte subsets. The level of OCs also increased on activated dendritic cells of the extravillous trophoblast [22]. Interestingly, the OCs maintained the peptide-binding groove in vitro [23]. Khare et al. [24] reported that the β2m-free HCs of HLA-B27 might induce arthritis in transgenic mice, Figure 1. HLA-E structure shows specific amino acid sequences, private epitopes, shared amino acid sequences, or public epitopes. (A). Private epitopes recognized by the anti-HLA-E monospecific mAb, TFL-033 (B). Public epitopes are recognized by the anti-HLA-E mAbs, TFL-006, and TFL-007. Both private and public epitopes are shown in yellow.

The Most Widely Shared Public Epitopes Are Cryptic in HLA-I
The HLA-I molecule is a heterodimer because the α-HC is closely associated with β2m ( Figure 1). As a consequence, not all sequences of HLA α-HC are exposed. Figure 1A shows that the specific sequences, or the private epitopes, of HLA-E (shown in yellow), 65 RSARDTA 70 and 143 SEGKSNDASE 152 , are exposed for direct binding of Abs. Figure 1B shows that the most commonly shared sequences or the public epitopes (shown in yellow) 117 AYDGKDY 123 and 126 LNEDLRSWTA 135 are cryptic, as they lie in close proximity with β2m. Figure 2 illustrates the crypticity of the shared epitope, 117 AYDGKDY 123 . The exposure of these shared epitopes depends on the nature of the HLA molecule. Most often, the HLA molecules occur as heterodimers, also designated as "closed conformers" or CCs. In CCs, the most prevalent shared epitopes, 117 AYDGKDY 123 and 126 LNEDLRSWTA 135 , are masked by β2m. However, in β2m-free α-HCs, designated as "open conformers" (OCs), they are exposed. and β2m-free, HC-specific Abs decreased the disease incidence in this model. OCs exist as dimers or in clusters at the cell surface in vivo [23,[25][26][27], profoundly impacting receptor engagement. Recognition of OCs as ligands by members of the killer Ig receptor family (KIR) and Ig-like transcript (ILT)/LIR/LILR family could influence their immunomodulatory function in inflammatory disease [28]. The shared sequences presented in HLA-E OCs contribute to the generation of HLA-I polyreactivity Abs. Though the OCs are found on activated immune cells, the soluble forms of HLA-I present in humans also occur as OCs. Demaria et al. [29,30] found that the levels of OCs in circulation were regulated by proteolytic cleavage. The release was mediated by a Zn (2+)dependent, membrane-bound metalloprotease (MMP). Specific separation by the metalloprotease occurs at a site close to the papain cleavage site in the α3 domain of HLA-I HCs. This site is not accessible to MMP in CCs. During the shedding of HLA-I trimers under different pathological conditions, the exogenous peptide and β2m dissociate from these CCs, and subsequent cleavage of OCs ( Figure 3) may be partially responsible for controlling the levels of OC on the surface of activated cells. Since the shed OCs expose previously cryptic epitopes, they may be recognized by B-cells for antibody production. Thus, the HLA-I molecules occur as heterodimers complexed with β2m on the cell surface and as β2m-free α-HC OCs. Schnabl et al. [13] reported that T lymphocytes activated in vitro or in vivo, but not resting, expressed many OCs. Using the mAb W6/32 that specifically recognized HLA-I CCs and the mAbs L45 and HC10 that bound explicitly to OCs but not to CCs, they confirmed the existence of HLA-OCs. Immunoprecipitation and co-capping experiments showed that LA45 was bound to HLA-I OCs at the cell surface. The mAb L45 bound to PHA-activated T-cells from a panel of 12 people with different HLA types, suggesting that LA45 might bind to epitopes shared by all HLA-I α-HCs. The HLA-I OCs expressed on the cell surface of activated T-cells-or EBV-transformed B-cells-are referred to as "peptide-binding empty HLA" [14,15]. The presence of HLA-I OCs was confirmed on activated T-cells in vitro and in vivo and on B cell lines (RAJI, NALM6), EBV-transformed B-cells, and the myeloid cell line KG-1A [16,17]. Interestingly, the expression of OCs on the cell surface in normal human T-cells upon activation and cell division correlated with the level of proliferation [18]. Indeed, the OCs on the cell surface were glycosylated. The inhibition of tyrosine phosphorylation with the Src-family tyrosine kinase inhibitor PP2 resulted in the enhanced release of HLA-I α-HC from the cell surface of activated T-cells. Further studies indicated that the OCs were regulators of ligand-receptor interactions and had potential implications for immune activation [18][19][20] and immune regulation [21]. The inflamed tissues in spondyloarthropathy showed increased levels of OCs on CD14b+ monocytes compared to other leukocyte subsets. The level of OCs also increased on activated dendritic cells of the extravillous trophoblast [22]. Interestingly, the OCs maintained the peptide-binding groove in vitro [23]. Khare et al. [24] reported that the β2m-free HCs of HLA-B27 might induce arthritis in transgenic mice, and β2m-free, HC-specific Abs decreased the disease incidence in this model. OCs exist as dimers or in clusters at the cell surface in vivo [23,25,26], profoundly impacting receptor engagement. Recognition of OCs as ligands by members of the killer Ig receptor family (KIR) and Ig-like transcript (ILT)/LIR/LILR family could influence their immunomodulatory function in inflammatory disease [27]. The shared sequences presented in HLA-E OCs contribute to the generation of HLA-I polyreactivity Abs.
Though the OCs are found on activated immune cells, the soluble forms of HLA-I present in humans also occur as OCs. Demaria et al. [28] found that the levels of OCs in circulation were regulated by proteolytic cleavage. The release was mediated by a Zn (2+) -dependent, membrane-bound metalloprotease (MMP). Specific separation by the metalloprotease occurs at a site close to the papain cleavage site in the α3 domain of HLA-I HCs. This site is not accessible to MMP in CCs. During the shedding of HLA-I trimers under different pathological conditions, the exogenous peptide and β2m dissociate from these CCs, and subsequent cleavage of OCs ( Figure 3) may be partially responsible for controlling the levels of OC on the surface of activated cells. Since the shed OCs expose previously cryptic epitopes, they may be recognized by B-cells for antibody production.

Antigenicity Rank of the Most Widely Shared Public Epitopes of HLA-E
An Ab directed against a protein antigen can recognize either a linear or a discontinuous sequence (still closely aligned in space due to protein folding) in its native conformation. This so-called eplet is present at the center of an epitope. The immunogenicity of an epitope is defined as its ability to actively induce an immune response. In contrast, the antigenicity of an epitope is defined as its passive ability to be recognized by the immune system. The antigenicity of a sequence, or epitope, in a polypeptide can be predicted using parameters such as hydrophilicity, flexibility, accessibility, beta turns, cell surface exposure, and polarity. Several methods are known for predicting the antigenicity of a continuous sequence. Chou and Fasman [31] developed a methodology, based on α-turns in

Antigenicity Rank of the Most Widely Shared Public Epitopes of HLA-E
An Ab directed against a protein antigen can recognize either a linear or a discontinuous sequence (still closely aligned in space due to protein folding) in its native conformation. This so-called eplet is present at the center of an epitope. The immunogenicity of an epitope is defined as its ability to actively induce an immune response. In contrast, the antigenicity of an epitope is defined as its passive ability to be recognized by the immune system. The antigenicity of a sequence, or epitope, in a polypeptide can be predicted using parameters such as hydrophilicity, flexibility, accessibility, beta turns, cell surface exposure, and polarity. Several methods are known for predicting the antigenicity of a continuous sequence. Chou and Fasman [31] developed a methodology, based on α-turns in protein secondary structures, for predicting antigenicity. Kolaskar and Tongaokar [32] developed a semi-empirical method using the physicochemical properties of amino acid residues and their frequency of occurrence in segmental epitopes. Karplus and Scholtz [33] predicted antigenicity based on three flexibility scales. Parker et al. [34] predicted antigenicity with a novel hydrophilicity scale. In these methods, when computing the score for given residue i, the amino acids in an interval of the chosen length, centered around residue i, are considered.
Further details are provided on the antibody epitope prediction web (http://tools. immuneepitope.org/tools/bcell/iedb_input (Last updated on 30 May 2021, see also iedb.org for details). The methods were collectively employed to assess the antigenicity ranking of the different sequences of HLA-E listed in Table 2. Among the shared peptide sequences listed in the table, the most prevalent shared epitope, 117 AYDGKDY 123 , ranked first in antigenicity. We have used these peptide sequences and the second-ranking epitope for peptide inhibition studies of anti-HLA-E mAbs and established that mAbs TFL-006 and TFL-007 are the most polyreactive [10,11,35,36].

Anti-HLA-E mAb Production
The mAbs were produced following the recommendations of the National Research Council's Committee on Methods of Producing mAbs [37]. The recombinant α-HCs HLA-E R107 and HLA-E G107 (source: Immune Monitoring Laboratory, Fred Hutchinson Cancer Research Center, concentration: 10 mg/mL in MES buffer) were used. Two mice were immunized with 50 mg of each antigen in PBS (pH 7.4, 100 mL) admixed with TiterMax (100 mL) (Sigma-Aldrich, St Louis, MO, USA) adjuvant before injection into the footpad and intraperitoneally. Three immunizations were given at about 12-day intervals, with an additional immunization after 12 days for mice receiving HLA-E G107 .
The clones were cultured in a medium (RPMI 1640 w/L-glutamine and sodium bicarbonate, with 15% FBS, penicillin, streptomycin, L-glutamine, and sodium pyruvate. Hybridomas were cryopreserved in RPMI containing 10% DMSO and 20% FBS. All culture supernatants were screened for IgG reacting to HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, and HLA-G using single-antigen microbeads on a Luminex platform. The mean fluorescent intensity (MFI) was determined for supernatants after diluting to 1/2. The MFI values were corrected against those obtained with negative control beads. Further details are provided elsewhere [35,38].

HLA-1 Antigen-Coated Single Antigen Beads (SABs) and the Luminex Platform for
Monitoring the Reactivity and the Density of These mAbs The affinity of Abs to different HLA-I antigens was monitored on a Luminex Platform, using HLA-I molecules, coated as single antigens, on the multiplex, fluorescinated microbeads. Three different kinds of microbeads were available for investigation, as illustrated in Figure 4: (i) beads with an admixture of CCs and OCs (Regular LABScreen SAB, source: One Lambda, Inc., Canoga Park, CA, USA), (ii) beads restricted to CCs only (iBeads developed by trypsinization of regular LABScreen beads and Immucore LIFECODES SAB), and (iii) beads restricted to OCs (alkali or acid-denatured beads). The regular LABScreen microbeads were coated with HLA-A (31 different antigens), HLA-B (50 different antigens), and HLA-C (16 different antigens). In addition, the HLA-Ia microbeads had built-in control beads: positive beads coated with human IgG and negative beads coated with serum albumin (human or bovine). For HLA-Ib, the control beads (both positive and negative) were added separately.

The HLA Reactivity Groups of mAbs Generated by Recombinant HLA-E OCs
More than 200 hybridomas were generated using the OCs of HLA-E. All mice exp iments were carried out under the guidance of Professor Paul I. Terasaki, who owned O Lambda, Inc., in Canoga Park, CA, U.S.A., with animal subject committee approval. T mAbs secreted by these hybridomas were both HLA-E monospecific and HLA-Ia a HLA-Ib polyreactive. These mAbs could be categorized into eight different groups, shown in Table 3. Group 1 corresponds to monospecific mAbs reacting restrictively w HLA-E. Group 4 refers to HLA-Ib-specific mAbs. Group 10 represents mAbs, recognizi both HLA-Ia and HLA-Ib molecules (Table 4). 24 The different kinds of microbeads were characterized using three different mAbs ( Figure 4). The mAb W6/32 (IgG2a) bound to CCs, but not to OCs [39], and bound to both peptide-associated and peptide-free CCs [40,41]. The mAb HC-10 (IgG2a) recognized an epitope in HLA-I between amino acid positions 57 and 62, with arginine at position 62 (R62) being crucial for HC-10 binding [42]. HC-10 recognized cell surface CCs devoid of a peptide, whereas the presence of a peptide reduced HC-10 reactivity [43]. The mAb TFL-006 (IgG2a) bound to OCs of all HLA-I loci and was inhibited by peptides from the amino acid sequences shared by all HLA-I loci [35,36].

The HLA Reactivity Groups of mAbs Generated by Recombinant HLA-E OCs
More than 200 hybridomas were generated using the OCs of HLA-E. All mice experiments were carried out under the guidance of Professor Paul I. Terasaki, who owned One Lambda, Inc., in Canoga Park, CA, USA, with animal subject committee approval. The mAbs secreted by these hybridomas were both HLA-E monospecific and HLA-Ia and HLA-Ib polyreactive. These mAbs could be categorized into eight different groups, as shown in Table 3. Group 1 corresponds to monospecific mAbs reacting restrictively with HLA-E. Group 4 refers to HLA-Ib-specific mAbs. Group 10 represents mAbs, recognizing both HLA-Ia and HLA-Ib molecules (Table 4).    Table 4 compares the HLA-I reactivities of an HLA-E monospecific mAb, TFL-033, with two HLA-I polyreactive mAbs, namely TFL-006 and TFL-007. TFL-006 showed reactivity with 32 alleles of HLA-A, 48 alleles of HLA-B, and 16 alleles of HLA-C, whereas TFL-007 reacted with 24 HLA-A, 44 HLA-B, and 16 HLA-C alleles. TFL-006 did not react with B*4901 and B*5701, although shared sequences (public epitopes) were present. TFL-006 and TFL-007 reacted with all three HLA-Ib isoforms as follows: HLA-E (+++) > HLA-G (++) > HLA-F (+). Among HLA-Ia isoforms, the reactivities in general were as follows: HLA-C (+++) > HLA-B (++) > HLA-A (+). The MFIs of the mAbs against A*1101 were the highest among HLA-A isoforms, with B*4006 the highest among HLA-B isoforms and C*1802 the highest among HLA-C isoforms. Of the seven HLA polyreactive mAbs belonging to group 10, TFL-006 ranked first, and TFL-007 ranked second. The reactivities of other mAbs of group 10 are presented elsewhere [44].
To determine the affinity of the mAbs TFL-006 and TFL-007 for shared sequences in the OCs, synthetic peptides purified by reverse-phase HPLC from GenScript Corporation (Piscataway, NJ, USA) were obtained [10]. Table 2 compares peptide inhibition using three commonly shared peptides, namely DTAAQI, AYDGKDY, and LNEDLRSWTA. These peptides were used, separately, to block the binding of the mAbs to the regular LABSCreen SABs, in which the OCs (β2m-free α-HC) were admixed with intact HLA trimers (CCs). Of the three peptides tested, AYDGKDY blocked more than 52% of the binding to the mAb TFL-006 compared to the other two peptides. Similar results were obtained with mAb TFL-007.
HLA-I reactivities of the different dilutions (1/2 to 1/128) of IVIg preparations were tested on regular LABScreen SABs, acid-denatured LABScreen SABs (OCs only), and on enzymatically treated regular LABScreen SABs, called iBeads (CCs only). The details of this methodology are presented elsewhere [34]. Figure 5 compares the MFIs, signifying the density of HLA-Ia and HLA-Ib IgG Abs on the three different SABs for IVIg preparations from GamaSTAN ( Figure 5A) and octagam ( Figure 5B). Table 6 summarizes the HLA-I polyreactivity of commercial preparations of IVIg from different sources. The data presented in Figure 5A,B reveal that the strength of the Abs is much higher in denatured SABs, which is predominant with OCs, than in the CCs restricted to the iBeads. These findings confirmed that the anti-HLA IgG Abs in the IVIg preparations recognized both the CCs and OCs of HLA molecules, but with a higher prevalence of Abs recognizing OCs. To ascertain whether the immunoreactivity of IVIg to HLA-Ia alleles was due to the OCs of HLA-I, the anti-HLA-1a Abs were adsorbed out with Sephadex gel conjugated with OCs of HLA-E and then tested for HLA-E and HLA-Ia reactivity [35]. IVIg immunoreactivity to HLA-Ia was minimized after the adsorbing-out process. Most importantly, HLA-E OCs alone could absorb out a major fraction of anti-HLA-1a antibodies, indicating the prevalence of polyreactive antibodies against HLA-I OCs in IVIg preparations. This unique finding formed the basis for comparing the functional characteristics of IVIg with HLA-I polyreactive mAbs (TFL-006 and TFL-007).  (B) Figure 5. HLA-I polyreactivity of IVIg. IVIg was tested on three kinds of SABs: (i) regular LABScreen bead set coated with an admixture of CCs and OCs, (ii) regular LABScreen bead set treated with a mild acid to convert OCs to CCs. The bead set was loaded with OCs. (iii) "iBeads" generated from regular LABScreen bead sets to enzymatically eliminate OCs, and therefore containing only CCs. The Luminex immunoassay with IVIg from GammaSTAN (A) and Octagam (B) confirmed the HLA-I polyreactivity of IVIg. Importantly, IVIG reacted with the monomeric variants on the acid-denatured SABs better than on iBeads. The density of IgG binding is illustrated by mean fluorescent intensity (MFI), as shown in different colors. The binding affinity of IVIg to monomeric variants was much higher than those recognizing the CC on 'iBeads.'. HLA-I polyreactivity of IVIg. IVIg was tested on three kinds of SABs: (i) regular LABScreen bead set coated with an admixture of CCs and OCs, (ii) regular LABScreen bead set treated with a mild acid to convert OCs to CCs. The bead set was loaded with OCs. (iii) "iBeads" generated from regular LABScreen bead sets to enzymatically eliminate OCs, and therefore containing only CCs. The Luminex immunoassay with IVIg from GammaSTAN (A) and Octagam (B) confirmed the HLA-I polyreactivity of IVIg. Importantly, IVIG reacted with the monomeric variants on the acid-denatured SABs better than on iBeads. The density of IgG binding is illustrated by mean fluorescent intensity (MFI), as shown in different colors. The binding affinity of IVIg to monomeric variants was much higher than those recognizing the CC on 'iBeads.'.
Kaveri et al. [59] demonstrated that Abs to a conserved region of HLA-Ia, present in pooled therapeutic IVIg, were capable of modulating CD8+ T-cell-mediated function. Klaessn et al. [60] showed that IgG and F(ab)2 fractions in IVIg were responsible for the inhibitory function. In contrast, Miyagi et al. [61] attributed the inhibition of blastogenesis and proliferation of activated T-cells by IVIg to the Fc receptors for IgG, FcγRI (CD23), FcγRII (CD32), FcγRIII (CD16), and FcγRIV (CD64), expressed on the immune cells upon activation. MacMillan et al. [70] documented further the IVIg-mediated suppression of T-cell proliferation, with or without CD28 co-stimulation. Sali et al. [62] demonstrated that 2% intact or Fab 2 fragments of IVIg could penetrate immune cells to modulate these activities. The penetrating fraction of IVIg inhibited the upregulation of activation marker CD25 on CD4+ splenocytes. In a placebo-controlled trial, the administration of IVIg to patients with inflammatory myopathies was associated with a significant reduction of the number of T-lymphocytes in vivo [74].
Based on these observations, it was hypothesized that HLA-I polyreactive mAbs that mimicked HLA-I reactivity of IVIg preparations might suppress activated T-cells, similarly to IVIg. The hypothesis was tested by comparing dose-dependent effects on the suppression of PHA-activated T-cells of three entities: IVIg, HLA-I polyreactive mAbs that mimicked IVIg (TFL-006 and TFL-007), and mAbs that did not mimic IVIg (TFL-033 and TFL-037).

Hypothesis Testing: Measurement of T-Lymphocyte Proliferation
More details are presented elsewhere, but all experimental analyses were carried out at Terasaki Foundation Laboratory with Review Board approval [58]. Briefly, we labeled purified human T-lymphocytes (freshly collected from a healthy, young adult male) with the intracellular fluorescent dye Carboxyfluorescein N-succinimidyl ester (CFSE), a cellpermeable dye that remained in the cell for several mitotic divisions. The technology used is illustrated in Figure 6A. Most importantly, the cessation of the progress of mitotic activity could be monitored as successive two-fold reductions in the fluorescent intensity after 72 h of treatment, including the addition of PHA and PHA plus IVIg or mAbs [58].

The Suppression of Activated T-Cells by IVIg vs. mAbs
A summary of the findings is presented in Figures 6B, 7 and 8. The results established the differential effects on the suppression of PHA-activated T-cells, both by IVIg and by HLA-I polyreactive mAbs (TFL-006 and TFL-007). The mAbs (TFL-033 and TFL-037) that did not mimic the HLA-I reactivity of IVIg did not affect the activation or proliferation of T-cells. The mAbs TFL-006 and TFL-007 appeared to be more potent suppressors of the blastogenesis and proliferation of activated CD4+ T lymphocytes than IVIg. The concentrations of the mAbs required for the suppression of T-cell proliferation were 50fold lower than the required concentration of IVIg. The suppression of blastogenesis and proliferation of T-cells by both IVIg and the anti-HLA-E mAbs was dose-dependent, and the dose required with mAbs was 50-150-fold lower than with IVIg. The mAb binding to OCs might have signaled T cell deactivation because the OCs have an elongated cytoplasmic tail with phosphorylation sites (tryosine320/serine335).

The Suppression of Activated T-Cells by IVIg vs. mAbs
A summary of the findings is presented in Figures 6B, 7 and 8. The results established the differential effects on the suppression of PHA-activated T-cells, both by IVIg and by HLA-I polyreactive mAbs (TFL-006 and TFL-007). The mAbs (TFL-033 and TFL-037) tha did not mimic the HLA-I reactivity of IVIg did not affect the activation or proliferation o T-cells. The mAbs TFL-006 and TFL-007 appeared to be more potent suppressors of th blastogenesis and proliferation of activated CD4+ T lymphocytes than IVIg. The concen trations of the mAbs required for the suppression of T-cell proliferation were 50-fold lower than the required concentration of IVIg. The suppression of blastogenesis and pro liferation of T-cells by both IVIg and the anti-HLA-E mAbs was dose-dependent, and th dose required with mAbs was 50-150-fold lower than with IVIg. The mAb binding to OC might have signaled T cell deactivation because the OCs have an elongated cytoplasmi tail with phosphorylation sites (tryosine320/serine335).  A tentative model of PHA-mediated activation and HLA polyreactive, mAb-mediated deactivation is proposed (Figure 9). It is known that PHA activation initiates phosphorylation of the cytoplasmic domain of CD3 and the activation of transcription factors. TCR cross-linking leads to the expression and phosphorylation of cell surface molecules such as IL-Rα [74] and the OCs of HLA-I [13][14][15][16][17][18][19][20][21][22]. The binding of mAbs to the shared amino acid sequences or epitopes exposed on the OCs may dephosphorylate the tyrosyl and seryl residues on the elongated cytoplasmic tails of the HLA-I OCs [75,76]. This may simultaneously lead to dephosphorylation of CD3 and revert the PHA-activation of CD3 on T-cells (for further details, see the legend for Figure 9). Upon dephosphorylation, Tcells were deactivated, resulting in the suppression of blastogenesis and proliferation. Based on the observations that there was a suppression of the blastogenesis and proliferation of PHA-activated CD4+ T-cells by HLA-I polyreactive, anti-HLA-E mAbs (TFL-006 and TFL-007) but not by HLA-I non-reactive, anti-HLA-E mAbs (TFL-033 and TFL-007), it was inferred that the IVIg-mediated suppression of the blastogenesis and proliferation of PHA-activated CD4+ T-cells could also be due to binding of the HLA-I OC-reactive IgG fraction in IVIg. Furthermore, when equal concentrations of the HLA-I polyreactive mAb TFL-007 and IVIg were compared, the suppression by the mAbs was greater than that of the IVIg. Evidently, the admixture of other IgGs with HLA-polyreactive IgGs in IVIg might have hindered the T-cell suppressive efficacy of the IVIg. A tentative model of PHA-mediated activation and HLA polyreactive, mAb-mediated deactivation is proposed (Figure 9). It is known that PHA activation initiates phosphorylation of the cytoplasmic domain of CD3 and the activation of transcription factors. TCR cross-linking leads to the expression and phosphorylation of cell surface molecules such as IL-Rα [74] and the OCs of HLA-I [13][14][15][16][17][18][19][20][21][22]. The binding of mAbs to the shared amino acid sequences or epitopes exposed on the OCs may dephosphorylate the tyrosyl and seryl residues on the elongated cytoplasmic tails of the HLA-I OCs [75,76]. This may simultaneously lead to dephosphorylation of CD3 and revert the PHA-activation of CD3 on T-cells (for further details, see the legend for Figure 9). Upon dephosphorylation, T-cells were deactivated, resulting in the suppression of blastogenesis and proliferation. Based on the observations that there was a suppression of the blastogenesis and proliferation of PHA-activated CD4+ T-cells by HLA-I polyreactive, anti-HLA-E mAbs (TFL-006 and TFL-007) but not by HLA-I non-reactive, anti-HLA-E mAbs (TFL-033 and TFL-007), it was inferred that the IVIg-mediated suppression of the blastogenesis and proliferation of PHA-activated CD4+ T-cells could also be due to binding of the HLA-I OC-reactive IgG fraction in IVIg. Furthermore, when equal concentrations of the HLA-I polyreactive mAb TFL-007 and IVIg were compared, the suppression by the mAbs was greater than that of the IVIg. Evidently, the admixture of other IgGs with HLA-polyreactive IgGs in IVIg might have hindered the T-cell suppressive efficacy of the IVIg. That elongation resulted in the exposure of cryptic tyrosine (Tyr320) and serine (Ser355) residues in the cytoplasmic tail. It might have been the binding of the mAbs to the shared peptide sequences that initiated the phosphorylation, leading to signal transduction. A final step involved initiation of dephosphorylation of the cytoplasmic domain of CD3, resulting in arrest of activation or suppression. That seemed plausible, as the phosphorylation was known to be reversible.

Background and Hypothesis
Previous literature documented that commercial IVIg not only inhibited PHA-and cytokine-induced T-cell activation and proliferation and significantly reduced the number of T-lymphocytes in vivo in a placebo-controlled trial in patients with inflammatory myopathies, but it was also capable of suppressing antibody production in patients under different disease conditions. Therefore, IVIg has become not only a substitution therapy for patients with immunodeficiencies [78], but also a therapeutic agent in autoimmune and systemic inflammatory diseases [79], as well as in organ and bone marrow transplantation [3,80].
IVIg is extensively used in patients with end-stage organ disease as well as allograft recipients. High levels of HLA antibodies, caused by various sensitizing events such as previous transplantation, pregnancy, or blood transfusion, are observed in patients with end-stage organ disease. High levels of such allo-HLA Abs can more likely produce positive crossmatch results with potential organ donors and preclude transplantation. Consequently, antibody-positive patients may experience prolonged waiting time. Among renal transplant recipients alone, such sensitized patients constitute more than one-third of those on the waiting list. For highly sensitized patients (with a panel-reactive HLA antibody (PRA) greater than 80%), the prospects of transplantation are grim. Previous literature documented that commercial IVIg not only inhibited PHA-and cytokine-induced T-cell activation and proliferation and significantly reduced the number of T-lymphocytes in vivo in a placebo-controlled trial in patients with inflammatory myopathies, but it was also capable of suppressing antibody production in patients under different disease conditions. Therefore, IVIg has become not only a substitution therapy for patients with immunodeficiencies [78], but also a therapeutic agent in autoimmune and systemic inflammatory diseases [79], as well as in organ and bone marrow transplantation [3,80].
IVIg is extensively used in patients with end-stage organ disease as well as allograft recipients. High levels of HLA antibodies, caused by various sensitizing events such as previous transplantation, pregnancy, or blood transfusion, are observed in patients with end-stage organ disease. High levels of such allo-HLA Abs can more likely produce positive crossmatch results with potential organ donors and preclude transplantation. Consequently, antibody-positive patients may experience prolonged waiting time. Among renal transplant recipients alone, such sensitized patients constitute more than one-third of those on the waiting list. For highly sensitized patients (with a panel-reactive HLA antibody (PRA) greater than 80%), the prospects of transplantation are grim.
The formation of Abs against allo-antigens depends on both T and B-cells. Therefore, aggressive suppressive strategies have been developed to deplete both T and B-cells in order to reduce the generation of allo-HLA Abs formed before and after transplantation. One such strategy is induction therapy with rabbit or horse anti-human thymocyte globulin, a polyreactive, polyclonal mixture of non-specific cytotoxic Abs capable of killing immune cells [81]. However, many clinical transplant centers worldwide have formulated alternate protocols to suppress the formation of anti-HLA Abs. These protocols may include plasmapheresis (PP), high-dose IVIg, or a combination of PP with low-dose IVIg [2,[49][50][51][52][53][54][82][83][84][85][86][87] or Rituximab, a monoclonal Ab (mAb) that depletes CD20+ B-cells [88][89][90][91]. Although several immunotherapeutic potentials are attributed to IVIg, its mechanism of action is far from certain. Possibly, it is due to polyclonality and the mixture of several kinds of IgG Abs present, including Abs against all HLA class I loci and alleles, as illustrated in Table 6, and Figure 5A,B.
A most interesting finding [35] was that the HLA-Ia reactivity of IVIg was significantly abolished when anti-HLA-E Abs were depleted, specifically by passing IVIg through HLA-E heavy chain-conjugated Affi-gel affinity columns. This suggested that IVIg's HLA-Ia reactivity could possibly be due to the presence of HLA-E-specific IgG Abs, and that there were anti-HLA-E mAbs that might simulate the HLA reactivity of IVIg.
These observations led us to hypothesize that the anti-HLA-E mAbs that simulated the HLA-reactivity of IVIg could mimic IVIg by suppressing B-cells from producing Abs. We compared the efficacy of IVIg versus that of mAb TFL-007 (HLA-I polyreactive, anti-HLA-E mAb) (Table 4) in suppressing the B-cell blastogenesis, proliferation, and production of Abs.

Methodology to Test the Hypothesis
More details are presented elsewhere [92]. Two commercial IVIg preparations were used. They were (1) IVIg-GamaSTAN™ (Lot 26NJ651; Talecris Biotherapeutics, Inc., Research Triangle Park, NC, USA) formulated as a 15-18% protein solution at a pH of 6.4-7.2 in 0.21-0.32 M glycine and, (2) IVIgGamunex ® -C (Lot 26NKLG1, Talecris), a solution at a pH of 6.4-7.2 in 0.16-0.24 M glycine, albumin < 20 µg/mL. The HLA-I polyreactive HLA-E mAbs used were TFL-007s (culture supernatant) and TFL-007a (ascites). For suppressing antibody production by B-cells, two different experiments were carried out: Experiment # 1 was on freshly purified B-cells from the peripheral blood lymphocytes of a woman alloimmunized with her husband's HLA DRB1*0101 antigens during her first pregnancy.
On day 9, the cells were pooled, aliquoted, and maintained without the cytokine combo or anti-CD40 mAb. These wells were exposed to medium or IVIg (1/100, 1.5 mg/mL) or mAb TFL-007s (1/100, 5 µg/mL) for 72 h. To study the effect of IVIg on the secretion of allo-HLA IgG Abs, we used IVIg at a protein concentration 300-fold higher than that of the purified culture supernatant of TFL-007s (5 µg/mL) used in the treatment of B-cells in culture. The supernatants recovered from the respective wells were screened for the HLA-alloAbs.
Experiment # 2 was on the human hybridoma cell line HML16, generated from the resting B-cells of a multiparous woman.
The The expression of CD19 was monitored using the fluorescein isothiocyanate (FITC)labeled, anti-human CD19 (mAb HIB19). On days 0 and 7, both resting and activated B-cells were stained with phycoerythrin (PE) anti-human CD20 (mAb 2H7), peridinin chlorophyll (PerCP) antihuman CD27 (mAb 0323), and FITC anti-human CD38 (mAb HIT2) to examine the differential activation of B-cells. The source of the mAbs was from BioLegend. Prior to staining with Abs, Human TruStain FcX™ (BioLegend, San Diego, CA, USA) was used to block FcR-involved, unwanted staining.

TFL-007 Suppressed HLA-Antibody Production Better Than That of IVIg
In the first experiment, both the sera of the woman and of those secreted by activated memory B-cells (culture supernatants) showed the presence of several alloAbs directed against DRB1*0102, DRB1*0404, DRB1*0405, DRB1*1402, and DRB1*0401, in addition to the primary alloantibody, anti-DRB1*0101 IgG. The secondary alloAbs were not directed against the husband's alleles. They might have represented cross-reactive alloAbs in that they occurred in cultures containing the primary alloantibody, anti-DRB1*0101. These alleles shared amino acids or amino acid sequences with the primary allele, DRB1*0101. The details were presented elsewhere [92]. Figure 10 documents that the GamaSTAN IVIg suppressed the secretion of the primary alloantibody against DRB1*0101 at different time points (p 2 < 0.01). Similarly, the HLA polyreactive mAb TFL-007s also significantly reduced the secretion of the primary alloantibody (p 2 < 0.0005). However, the suppression of the secretion of the primary alloantibody by the mAb TFL-007s was highly significant compared to the IVIg-induced suppression.
they occurred in cultures containing the primary alloantibody, anti-DRB1*0101. These alleles shared amino acids or amino acid sequences with the primary allele, DRB1*0101. The details were presented elsewhere [92]. Figure 10 documents that the GamaSTAN IVIg suppressed the secretion of the primary alloantibody against DRB1*0101 at different time points (p 2 < 0·01). Similarly, the HLA polyreactive mAb TFL-007s also significantly reduced the secretion of the primary alloantibody (p 2 < 0·0005). However, the suppression of the secretion of the primary alloantibody by the mAb TFL-007s was highly significant compared to the IVIg-induced suppression. Figure 10. The impact of IVIg (GamaSTAN) and HLA-I polyreactive mAb (TFL-007) on the primary alloantibody DRB1*0101 (IgG), secreted by activated B-lymphocytes obtained from an alloimmunized woman. GamaSTAN S/D IVIg was used at 1:100 dilution, 1·5 mg. protein/mL. At the time when the IVIg was added, the cytokine combo and anti-CD40 antibody were not added. In all panels, the mean fluorescent intensity (MFI) of alloantibody secretion is compared between the medium control and treatment with IVIg and HLA-I polyreactive IgG2a mAb TFL-007. IVIg inhibited the secretion of the primary alloantibody at a significant level (p 2 = 0·01). HLA-I polyreactive mAb inhibited the secretion of the primary alloantibody at a higher significant level (p 2 = 0.0005). This is an original figure .
In the second experiment, the treatment effects of two preparations of IVIg (GamaS-TAN and Gamunex) were tested on HM16 at different dilutions and protein concentrations and compared with the media control [92]. The IVIg concentration used for the hybridoma cell line HML16 was comparable to the high-dose IVIg used for desensitization and post-transplant therapy. Neither of the IVIg preparations showed any significant suppression of the secretion of allo-HLA-B IgG by the hybridoma cells. In contrast, the mAb TFL-007a had a strikingly significant suppressive effect on the secretion of both HLA-B*0702 and B*8101. More importantly, the mAb TFL-007a showed dosimetric suppression of allo-HLA Abs. When HML16 was treated with the highest concentration of TFL-007a (100 μg/mL), suppression was 33% for B*0702 and 34% for B*8101, compared with the Figure 10. The impact of IVIg (GamaSTAN) and HLA-I polyreactive mAb (TFL-007) on the primary alloantibody DRB1*0101 (IgG), secreted by activated B-lymphocytes obtained from an alloimmunized woman. GamaSTAN S/D IVIg was used at 1:100 dilution, 1.5 mg. protein/mL. At the time when the IVIg was added, the cytokine combo and anti-CD40 antibody were not added. In all panels, the mean fluorescent intensity (MFI) of alloantibody secretion is compared between the medium control and treatment with IVIg and HLA-I polyreactive IgG2a mAb TFL-007. IVIg inhibited the secretion of the primary alloantibody at a significant level (p 2 = 0.01). HLA-I polyreactive mAb inhibited the secretion of the primary alloantibody at a higher significant level (p 2 = 0.0005). This is an original figure.
In the second experiment, the treatment effects of two preparations of IVIg (GamaS-TAN and Gamunex) were tested on HM16 at different dilutions and protein concentrations and compared with the media control [92]. The IVIg concentration used for the hybridoma cell line HML16 was comparable to the high-dose IVIg used for desensitization and posttransplant therapy. Neither of the IVIg preparations showed any significant suppression of the secretion of allo-HLA-B IgG by the hybridoma cells. In contrast, the mAb TFL-007a had a strikingly significant suppressive effect on the secretion of both HLA-B*0702 and B*8101. More importantly, the mAb TFL-007a showed dosimetric suppression of allo-HLA Abs. When HML16 was treated with the highest concentration of TFL-007a (100 µg/mL), suppression was 33% for B*0702 and 34% for B*8101, compared with the medium control group. When the dosage of TFL007a was decreased, the suppression effect declined. In short, in marked contrast to the IVIg preparations, the HLA-I polyreactive mAb TFL-007a significantly suppressed the secretion of both allo-HLA-B Abs.
The therapeutic mAb, tocilizumab, given to patients with rheumatoid arthritis, increases T-regs and correlates with clinical response [108]. Upregulation of T-regs by IVIg is considered a critical factor in controlling experimental autoimmune encephalomyelitis; IVIg is known to upregulate T-regs [109].
Because IVIg can upregulate T-regs, it is hypothesized that the HLA-I polyreactive mAbs TFL-006 and TFL-007 may also induce proliferation of T-regs. For testing this hypothesis, CD4+CD25+foxp 3 + Tregs were obtained from the peripheral blood of normal and healthy donors after obtaining necessary consent and institutional IRB approval. A variety of cell surface markers, including CD4, CD25 (IL-2Rα), CD45RA, and FoxP3, were monitored, using their respective monoclonal Abs.
We compared the impact of different commercial preparations of IVIg and the HLA-I polyreactive mAb TFL-007 in triplicate on untreated T-regulatory cells (CD4+/CD25+/FoxP 3+ ) obtained from the blood of a healthy volunteer from TFL. The mAb purified from ascites was used throughout. Figure 11 illustrates that the different commercial preparations of IVIg at two different dilutions (1/10 and 1/80) failed to upregulate the Tregs, while the mAb TFL-007a showed a significant increase in the number of cells as compared to controls. mAb TFL-007a significantly suppressed the secretion of both allo-HLA-B Abs.
The therapeutic mAb, tocilizumab, given to patients with rheumatoid arthritis, increases T-regs and correlates with clinical response [108]. Upregulation of T-regs by IVIg is considered a critical factor in controlling experimental autoimmune encephalomyelitis; IVIg is known to upregulate T-regs [109].
Because IVIg can upregulate T-regs, it is hypothesized that the HLA-I polyreactive mAbs TFL-006 and TFL-007 may also induce proliferation of T-regs. For testing this hypothesis, CD4+CD25+foxp 3 + Tregs were obtained from the peripheral blood of normal and healthy donors after obtaining necessary consent and institutional IRB approval. A variety of cell surface markers, including CD4, CD25 (IL-2Rα), CD45RA, and FoxP3, were monitored, using their respective monoclonal Abs.

Discussion
IVIg is a mixture of polyclonal IgG Abs, pooled and purified from thousands of normal individuals. The binding affinity of IVIg is multivarious because the IgG Abs in IVIg can bind to blood groups, MHC complex antigens, cytokines, chemokines and their receptors, and several other antigens including human albumin. Most importantly, in the last three decades [2,3,[49][50][51][52][53][54][55][56][57][82][83][84][85][86][87][88][89][90][91], IVIg has been administered to HLA-sensitized patients who possess a higher level of HLA allo-Abs, with the intention to lower these Abs. Due to the high level of these allo-HLA Abs, such patients are more likely to have a positive crossmatch result with potential organ donors pre-transplantation, and they often languish for years on the waiting list. For highly sensitized patients, the prospects of transplantation get grim. Realizing the imminent need to lower the impact of HLA sensitization, the US FDA approved IVIg for therapeutic administration in HLA-sensitized patients. However, due to its lack of efficacy as a monotherapy [55][56][57], IVIg is often combined with other therapeutic agents [89][90][91] such as plasmapheresis and rituximab (a mAb that depletes CD20+ B-cells). Commercial therapeutic preparations of IVIg, formulated in different countries, possess allo-HLA IgG Abs against almost all alleles of all six loci of HLA-I [35,57]. In this regard, HLA polyreactive mAbs developed from HLA-E αHCs mimic the allo-HLA antibody profile of IVIg [the U.S. Patent 10,800,847; 13 October 2020].
The structure of a single HLA-I antigen has two different profiles, as illustrated in Figure 2. The most well-known profile is the cell surface heteromer, consisting of α-HC associated with β2m ( Figure 1). In the heteromeric profile, many of the most commonly shared amino acid sequences (Tables 1 and 2), such as 117 AYDGKDY 123 and 127 LNEDLRSWTA 135 are cryptic, hidden by β2m. The second less-known profile of a single HLA-I antigen is a β2m-free a-HC, i.e., an OC ( Figure 2). Although this structure is naturally occurring, it is often misconstrued to be identical to denatured HLA antigens, as found on alkali or acid-treated SABs [46,110]. In fact, the misconception that the monomeric versions of naturally occurring HLA molecules are "denatured" HLA is well-documented in several reports [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]. These reports demonstrate several novel aspects of the monomeric versions. The shared amino acid sequences common to all six loci of HLA are exposed for immune recognition and serve as receptors for signal transduction. Notably, the long cytoplasmic tails of the monomeric versions on cells with tyrosyl and seryl residues ( Figure 2) are shown to be involved in signal transduction (Figure 9). Arosa et al. [19,111] elucidated further the difference between the intact HLA antigens, as CCs, and the monomeric variants, as OCs, when expressed in vivo on the cell surface ( Figure 4). One difference was that HLA molecules shed from the cell surface [112] occurred as monomeric variants, or OCs, in the circulation ( Figure 3). Since they exposed cryptic epitopes, they served as immunogens for allo-HLA Ab production. Indeed, the immunogenicity and antigenicity of the shared cryptic epitope 117 AYDGKDY 123 were much greater than other epitopes ( Table 2).
Visualizing the antibody recognition sites (epitopes) of CCs and OCs would clarify and elucidate the HLA profile instrumental for generation of HLA-I polyreactive IgG mAbs and naturally occurring IgG Abs. Theoretically, one can expect the presence of polyreactive IgG Abs recognizing the OCs of HLA in IVIg. Indeed Figure 5 confirms the prevalence of two commercial preparations of IVIg reacting to monomeric variants on the acid denatured SABs. The density of IgG binding to monomeric variants was much higher than those recognizing the CCs on 'iBeads' (Figure 5). Kaveri et al. [59] pointed out "Abs to a conserved region (cryptic domain) of HLA class I molecules, capable of modulating CD8 T cell-mediated function, are present in pooled normal immunoglobulin for therapeutic use" (p. 865). Strikingly, HLA-polyreactive IgG2a mAbs (TFL-006 and TFL-007) do not recognize CCs, but recognize only the OCs on the SABs (Table 5). Importantly, as noted above, these naturally occurring OCs existing on the cell surface in vivo should not be considered synonymous with chemical treatment (acid or alkali) on synthetic beads disrupting intact CCs to produce denatured α-HCs in vitro.
The binding of the HLA-I polyreactive mAbs to OCs is responsible for the suppression of blastogenesis and proliferation. It may involve the reversal of phases of activation of T-lymphocytes, mediated by signal transduction. The elongation of the cytoplasmic tail of the HLA-I OCs exposes otherwise cryptic tyrosine-320 [76] and serine-335 [113] residues, with a provision for phosphorylation (Figure 9). Although serine-335 is generally considered the primary site of phosphorylation in this tail, the phosphorylation of tyrosine-320 has been indicated by others [76]. Thus, the HLA-I OCs in activated T-cells are associated with tyrosine phosphorylation and are capable of enabling cis interactions with cell surface receptors or other signaling molecules [76,77,[113][114][115][116]. The binding by TFL mAbs may result in dephosphorylation of the cytoplasmic tails of CD3 molecules by activating phosphatases. The result is to arrest transcription factors and inhibit the synthesis of the proteins involved in blastogenesis and mitosis (Figure 9). These events suggest that the suppression of the activation of T-cells could be due to the binding of HLA-I polyreactive mAbs, mimicking the HLA-I polyreactivity of IVIG, to the shared amino acid sequences exposed on the naturally occurring OCs of HLA-I molecules (Figure 2). This activation involves not one HLA-I locus but all alleles of the six loci (HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, and HLA-G) expressed on the surface of activated T-cells. The efficacy of the monoclonality of the HLA-I polyreactive mAbs and their F(ab )2 binding is restricted to OCs. Furthermore, poor efficacy of IVIg is due to admixture with CC-binding Abs, other non-HLA Abs, anti-idiotypic Abs, immune complexes, and other chemicals such as sugars added to render stability to IVIg. Possibly, the same or similar phosphorylation and dephosphorylation mechanisms may be involved in suppressing antibody production by B-cells. The HLA-I polyreactive mAb, but not IVIg, suppressed allo-HLA-I and allo-HLA-II Abs production by B-memory cells and immortalized B-cells. While monitoring the efficacy of the HLA-I polyreactive mAbs, the cytokine combination and anti-CD40 antibody were both removed to precisely evaluate the impact of the antibody. Possibly, HLA-I polyreactive mAbs may also help to suppress other Abs produced by B-cells and alloAbs as well.
In addition, the HLA-I polyreactive mAbs upregulated T-regs better than IVIg ( Figure 11). T-regs are well known for their immunoregulatory properties. They suppress Ab production by downregulating B-memory and plasma cells [102] and by depleting CD4+ [103] and CD8+ [104,105] T-cells. We believe that both IVIg and the HLA-I polyreactive TFL mAbs perform the immunosuppressive functions stated above, including the upregulation of T-regs, by binding to the OCs of HLA.

Conclusions
The observations point out conclusively that IVIg (a) suppressed blastogenesis and proliferation of T-cells, (b) minimized the HLA-II allo-antibody production by B-memory cells of parous women and HLA-I antibody production of an immortalized cell line, and (c) expanded T-regs. These effects could be due to the HLA OC-polyreactive Abs present in IVIg. These findings are further strengthened by the several similarities observed in the nature and functional characteristics of IVIg and HLA-I polyreactive, OC-specific TFL mAbs (TFL-006 and TFL-007) ( Table 7). The mAbs, which recognize shared epitopes on naturally occurring OCs, with no ability to bind to the CCs, perform the same suppressive functions better than the different preparations of IVIg. Possibly, anti-CC Abs present in IVIg may impact their above-mentioned immunomodulatory functions. Hence, HLA-I polyreactive TFL mAbs are capable of serving as IVIg-mimetics, perhaps more efficiently than IVIg itself. Clinical trials are clearly warranted. Possibly, humanized versions of these TFL-mAbs, either combined or alone, can be better therapeutic tools than IVIg to suppress HLA sensitization and minimize antibody production post-transplantation. Table 7. Comparison of natural and functional characteristics of IVIg and HLA-I polyreactive mAbs.