ALBERT

Description: Abstract 3414 Mesenchymal stromal cells (MSC) are of promising therapeutic use to suppress immunogenic responses following transplantation, and to support expansion of hematopoietic stem- and progenitors cells (HSPC) from small transplants derived for instance from cord blood. Culture-expanded MSC produce a wide variety and quantity of Wnt-proteins and the crucial role of Wnt-signaling in the hematopoietic stem cell niche is well established. However, studies addressing Wnt-signaling in MSC have (i) only focused on culture-expanded MSC and (ii) did not discriminate between phenotypically distinct subpopulations which are present in bulk cultures of expanded MSC. Recently we identified three new subpopulations of MSC in human bone marrow (BM) based on expression of CD271 and CD146: CD271brightCD146−, CD271brightCD146+, CD271−CD146+. These fractions co-express the “classical” MSC markers CD90 and CD105 and lack expression of CD45 and CD34 (Maijenburg et al, Blood 2010, 116, 2590). We and others demonstrated that the adult BM-derived CD271brightCD146− and CD271brightCD146+ cells contain all colony forming units-fibroblasts (Maijenburg et al, Blood 2010, 116, 2590; Tormin et al, Blood 2010, 116, 2594). To investigate how these primary subsets functionally compare to conventional, culture-expanded MSC, we investigated their Wnt-signature and hematopoietic support capacity. To this end, we sorted CD271brightCD146− and CD271brightCD146+ cells from human adult BM (n=3) and compared their Wnt-signatures obtained by Wnt-PCR array to the profiles from cultured MSC from the same donors. Fifteen genes were consistently differentially expressed in the two sorted uncultured subsets compared to their conventionally cultured counterparts. Expression of CCND1, WISP1 and WNT5B was strongly increased, and WNT5A was only detected in the conventionally cultured MSC. In contrast, WNT3A was exclusively expressed by sorted primary CD271brightCD146− and CD271brightCD146+ cells, that also expressed higher levels of JUN, LEF1 and WIF1. The differences in Wnt (target)-gene expression between CD271brightCD146− and CD271brightCD146+ cells were more subtle. The Wnt-receptors LRP6 and FZD7 were significantly higher expressed in CD271brightCD146+ cells, and a trend towards increased expression in the same subset was observed for CTNNB1, WNT11 and MYC. When the sorted subsets were cultured for 14 days (one passage), the differences in Wnt-related gene expression between the subsets was lost and the expanded sorted cells acquired an almost similar Wnt-signature as the MSC cultured from BM mononuclear cells from the same donors. The cultured subsets lost the expression of Wnt3a and gained the expression of Wnt5a, similar to the unsorted MSC cultured from the same donors in parallel. Despite the loss of a distinct Wnt-signature, co-culture experiments combining the sorted MSC subsets with human HSPC revealed that CD271brightCD146+ cells have a significantly increased capacity to support HSPC in short-term co-cultures (2 weeks) compared to CD271brightCD146− cells (p

Description: A single recombinant immunoglobulin G1 (IgG1) anti-RhD antibody (MonoRho) was compared with a currently used polyclonal anti-RhD product (Rhophylac) in a phase 1 study for safety, efficacy of Rhesus D (RhD)–positive red blood cell (RBC) clearance, and prevention of RhD immunization in RhD-negative men challenged with 15 mL RhD-positive RBCs. Both the polyclonal product and recombinant anti-RhD effectively cleared RhD-positive RBCs after intravenous and intramuscular injection. The recombinant anti-RhD demonstrated a slower clearance rate compared with the polyclonal anti-RhD. There was no dose response, and there was considerable variation among subjects who received the same dose of recombinant anti-RhD. Interestingly, RhD-positive RBC clearance rates were strongly associated with Fcγ receptor IIA (FcγRIIA) and FcγIIIA but not with FcγIIIB polymorphisms. Subjects homozygous for FcγRIIA-131H or FcγRIIIA-158V allotypes showed a faster clearance rate compared with both the heterozygote and the corresponding alternative homozygote allotypes. A similar but less marked trend was seen for the polyclonal anti-RhD. Despite the variation in clearance rates there was no evidence of anti-RhD alloantibodies in any of the subjects at +6 months after the RBC challenge.

Description: Abstract 3854 Mesenchymal stromal cells (MSC) have a promising potential for cellular therapies, based on their immunomodulatory functions and their differentiation capacity. However, it remains unclear how MSC exert their functions and especially homing towards effector sites is poorly understood. An imperative for future therapies is to understand the process of MSC migration. We have previously shown that only 10–20% of culture-expanded MSC is capable to migrate in vitro, but these migratory cells can not be discriminated by markers. To identify genes specific for migratory MSC, we performed a gene expression study. Using a Transwell system (12mm pore size), migrating and non-migrating fetal bone marrow MSC (FBMSC) were physically separated and RNA was extracted. MSC that were exposed to a SDF-1 gradient (0.4mm pore size) and cultured FBMSC were included as controls. Only 9 genes were differentially expressed between migratory and non-migratory MSC. This set did not include cell surface markers that enable selection for a migratory subset. The nuclear orphan receptor family members Nurr1 and Nur77 were 2-fold upregulated in migratory MSC. These results were confirmed by RQ-PCR. Nurr1 and Nur77 are members of the NR4A nuclear orphan receptor family and were first described as early response transcription factors upon growth factor stimulation. We observed that the chemokine SDF-1 and the growth factor PDGF-BB increased the expression of Nur77 and Nurr1 in MSC. To investigate the role of Nur77 and Nurr1 in migration, FBMSC were transduced with lentivirus to achieve overexpression. Both overexpression of Nur77 and Nurr1 increased specific migration towards SDF-1 compared to Mock transduced FBMSC (n=4) (Nur77 mean 297% ± 186 relative to Mock, p≤ 0.014; Nurr1 mean 227% ± 79, p≤ 0.014). We previously described that cell cycle influences MSC migration, with S- and G2/M-phase of the cell cycle negatively influencing migration. In MSC cultures overexpressing Nur77 or Nurr1, the percentage of cells in S-phase of the cell cyle was significantly reduced (Nur77 15.2±7.8%, p≤0.007; Nurr1 16.4±9.1%, p≤0.039) compared to Mock-transduced cells (24.0±8.9%). Possibly, part of the Nur77 and Nurr1 effect is regulated by an effect on cell cycle. Together these data suggest that Nur77 and Nurr1 are involved in migration of MSC, which offers perspectives to modulate migration of MSC. Because these nuclear orphan receptors were not implied in MSC biology to date, we study their immunomodulatory effect. NR4A genes modulate immune responses in a variety of cells by influencing cytokine production. The immunomodulatory effect of MSC also seems to be based on secretion of a wide variety of cytokines, chemokines and growth factors. We therefore investigated whether Nur77 and Nurr1 play a role in cytokine and growth factor production in MSC. Compared to Mock transduced MSC, overexpression of Nur77 and Nurr1 in MSC resulted in increased basal levels of IL-6 protein (Nur77 3.4±2.4 fold, p≤0.007; Nurr1 3.8±3.1 ns) and IL-8 protein (Nur77 5.9±3.3 fold, p≤0.046; Nurr1 4.7±2.2, p≤0.031) and HGF mRNA expression (Nur77 5.4±0.8 fold, p≤0.021; Nurr1 3.4±1.18, p≤0.026). This was further enhanced upon stimulation with TNFα for 24hrs relative to the unstimulated Mock (IL-6 Mock 22.9±19.5 fold; Nur77 35.1±21.9 fold; Nurr1 35.8±24.1 fold; IL-8 Mock 58±37 fold; Nur77 160±141 fold; Nurr1 262±181 fold). HGF mRNA levels were slightly elevated by stimulation with IFNγ for 24hrs (Mock 1.8±1.6 fold; Nur77 7.0±3.8 fold; Nurr1 7.1±1.7 fold; ns). Expression levels of IDO1 and TGF-ß, also involved in immune modulation by MSC, were not influenced by Nur77 or Nurr1. In conclusion, we identified two novel genes involved in MSC migration. Intriguingly these genes are also involved in cytokine production, suggesting that the fraction with the best migratory capacity is able to modulate the immune response at its effector site distinct from the bulk of culture expanded MSC that lack migratory capacity. Modulation of Nur77 and Nurr1 expression in MSC could therefore offer perspectives to ameliorate cellular therapy. Disclosures: No relevant conflicts of interest to declare.

Description: Autoimmune neutropenia (AIN) in children can be divided into 2 forms. In primary AIN, neutropenia is the sole abnormality, and although neutrophil counts are generally below 500 μL−1, mild bacterial infections occur. Primary AIN is mostly seen in young children and shows a self-limited course. AIN occurring in association with autoimmune diseases (secondary AIN) often shows more severe infectious complications. We analyzed clinical and serological data from 28 pediatric patients with AIN to evaluate whether there is a possible relationship between specificity of the neutrophil autoantibodies and the clinical course of the disease. Specificity of the circulating antibodies was determined with the indirect granulocyte immunofluorescence test (GIFT) and a panel of phenotyped donor neutrophils. The samples were further analyzed in the monoclonal antibody immobilization of granulocyte antigens assay (MAIGA) for neutrophil antigen (NA)1, NA2, CD11a, and CD11b specificity. With the indirect GIFT, an antibody specificity was deduced in 26 of the 28 analyzed samples. In all but 3 sera from patients with primary AIN, NA1-(76%) or NA2-(10%) specific antibodies were detected with the indirect GIFT. In 2 samples, the reactivity in the indirect GIFT was too weak to draw conclusions, but the MAIGA showed NA1 and/or NA2 specificity of the antibodies. One serum, from a patient with primary AIN with a persistent neutropenia for more than 6 years, contained NA1, possibly pan-FcγRIIIb, and CD11a antibodies. In 4 sera from patients with primary AIN, weak antibodies with CD11a or CD11b specificity were detected with the MAIGA. Sera from 7 patients with secondary AIN contained in all cases antibodies with pan-FcγRIIIb specificity, as deduced from the indirect GIFT results and absorbance/elution experiments performed with 2 sera. The MAIGA confirmed this for only 1 of the 5 tested sera. Furthermore, CD11a antibodies were detected in 1 of the 5 tested sera. In conclusion, our results indicate that primary AIN is usually associated with NA-specific antibodies, whereas secondary AIN seems to be associated with pan-FcγRIIIb antibodies. Thus, characterization of the antibodies in sera from children with AIN discriminates patients with primary AIN from those with secondary AIN.

Description: Human adult bone marrow-derived mesenchymal stromal cells MSC (ABMSC) are increasingly applied in the clinic to decrease graft versus host disease and to enhance hematopoietic recovery. Fetal bone marrow-derived MSC (FBMSC) display similar immune suppressive and regenerative capacities as adult MSC, and have been transplanted into patients. The proliferation capacity of FBMSC, however, is much larger than that of ABMSC. The aim of our studies is to understand the molecular mechanism of proliferation and hematopoietic support by MSC to optimize the expansion of functional MSC for clinical use. Comparison of gene expression between ABMSCs and FBMSCs identified 687 differentially expressed genes. Of these, 16 were Wnt-related, mainly Wnt-inhibitors and Frizzled receptors. Expression of SFRP4, WISP1, WISP2, WISP3, FZD1, FZD5, FZD8 and MYCBP2 was upregulated in ABMSC, whereas DKK1, DKK2, CCND2, WNT5a, MYC, FZD2, FZD6 and FZD7 are expressed at a higher level in FBMSC. Although the expression of few genes (e.g. DKK1) was culture density dependent, other genes such as Wnt5a, DKK2 and SFRP4 were consistently differentially expressed independent of culture conditions. Therefore we investigated the role of Wnt signaling in adult and fetal bone marrow-derived MSC. Wnt3a induced a concentration dependent increase of the canonical Wnt-target genes TCF and LEF in both ABMSC and FBMSC. However, ABMSC responded faster, and at a lower concentration of Wnt3a compared to FBMSC. In addition, Wnt3a increased the proliferation of ABMSC, but not of FBMSC. Interestingly, a complete medium change was already sufficient to increase TCF/LEF expression in ABMSC, but not in FBMSC, suggesting that ABMSC produced a soluble Wnt-inhibitor. Moreover, switching MSC conditioned medium between FBMSC and ABMSC indicated that FBMSC conditioned medium significantly stimulated the expansion of ABMSC while the reverse experiments did not show an inhibiting effect of ABMSC conditioned medium on the expansion of FBMSC. Thus, ABMSC produce a factor that only affects ABMSC, but not (the factors produced by) FBMSC. To block autocrine Wnt production, MSC were exposed for 48 h to the Inhibitor of Wnt Production 2 (IWP2). Abrogation of Wnt-production in FBMSC modestly decreased beta-catenin expression, and strongly decreased TCF/LEF expression, but did not affect ABMSC. Addition of IWP-2 to long-term cultures strongly inhibited proliferation of FBMSCs compared to ABMSCs. To unravel the role of MSC-produced Wnt factors in hematopoiesis, we co-cultured adult or fetal MSCs together with cord blood derived CD34+ cells in the presence or absence of IWP2 inhibitor. Addition of IWP2 to ABMSC decreased the short term support of hematopoietic stem and progenitors (HSPC), while IWP2 did not affect the support of HSPCs by FBMSC. Overall, ABMSCs provided a significant better short term hematopoietic support than FBMSCs. In conclusion, our data demonstrate that ABMSC produce both Wnt factors and inhibitors. FBMSC, in contrast, produce Wnt-related factors that seem to contribute more to the expansion capacity of FBMSC than to their hematopoietic support. To identify factors we current use mass spectroscopy of supernatant to determine the secretome. Modulation of (parts) of the Wnt pathway may improve clinical expansion protocols of ABMSC. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2590 Mesenchymal stromal cells (MSC) in bone marrow (BM) consist of a heterogeneous population of cells. MSC orchestrate the BM microenvironment, and therefore have a pivotal role in hematopoietic support. Since MSC produce a large quantity and variety of Wnt proteins, it is likely that Wnts are involved in their hematopoietic support function Recent studies show enrichment for BM-derived MSC by sorting for CD271+ or CD146+ cells. However, it is unclear whether these markers are co-expressed with classical markers MSC markers CD73, CD105 and CD90. In addition mainly adult samples were studied, while it is known that MSC frequency declines with age. Most in-vitro studies with MSC as well as all therapeutic applications have been performed with a heterogeneous population of cultured cells. In this study we first characterized the different MSC subpopulations in BM from 81 donors (adult n=55 mean age 62 (19-82), pediatric n=22 mean age 4 (0-16) and fetal n=6, second trimester) by 6-color FACS analysis. Next we analyzed the subsets with respect to their distribution in different age groups, colony forming ability, and Wnt (target)-gene expression compared to cultured MSC. Three putative MSC subsets were identified in CD45−/dimCD34− cells from adult BM, based on expression of CD271 and/or CD146, CD105 and CD90. These populations were CD271brightCD146−CD105+CD90+, CD271brightCD146+CD105+CD90+ and CD271−CD146+CD105+CD90+, hereafter referred to CD271 single positive (sp) (0.011±0.001% from the mononuclear cell fraction (MNC)), double positive (0.003±0.001% from MNC) and CD146 sp (0.007±0.001%). The largest population of putative MSC in adult BM was CD271 sp (n=55, 55.9±24.3%). In contrast to adult BM, in pediatric BM the dominant population was double positive (n=22, 78.2±20.9%, 0.0115±0.003% MNC count), while the CD271 sp (0.0005±0.001% from MNC) and CD146 sp (0.0005±0.0001% from MNC) populations contained less than 25% of the CD271 and/or CD146+ cells. Remarkably, in fetal BM CD146 sp cells accounted for 35.5±8.8% (0.107±0.004% from MNC) of all presumptive MSC. Double positive cells represented 60.8±8.2% (0.1739±0.025% from MNC). CD271 sp cells, dominant in adults, were nearly absent (0.0115% ±0.004% from MNC)in fetal BM. The proportion of CD271 sp cells in BM was positively correlated with donor age (n=83, R2=0.41, p

Description: Autoimmune neutropenia (AIN) in children can be divided into 2 forms. In primary AIN, neutropenia is the sole abnormality, and although neutrophil counts are generally below 500 μL−1, mild bacterial infections occur. Primary AIN is mostly seen in young children and shows a self-limited course. AIN occurring in association with autoimmune diseases (secondary AIN) often shows more severe infectious complications. We analyzed clinical and serological data from 28 pediatric patients with AIN to evaluate whether there is a possible relationship between specificity of the neutrophil autoantibodies and the clinical course of the disease. Specificity of the circulating antibodies was determined with the indirect granulocyte immunofluorescence test (GIFT) and a panel of phenotyped donor neutrophils. The samples were further analyzed in the monoclonal antibody immobilization of granulocyte antigens assay (MAIGA) for neutrophil antigen (NA)1, NA2, CD11a, and CD11b specificity. With the indirect GIFT, an antibody specificity was deduced in 26 of the 28 analyzed samples. In all but 3 sera from patients with primary AIN, NA1-(76%) or NA2-(10%) specific antibodies were detected with the indirect GIFT. In 2 samples, the reactivity in the indirect GIFT was too weak to draw conclusions, but the MAIGA showed NA1 and/or NA2 specificity of the antibodies. One serum, from a patient with primary AIN with a persistent neutropenia for more than 6 years, contained NA1, possibly pan-FcγRIIIb, and CD11a antibodies. In 4 sera from patients with primary AIN, weak antibodies with CD11a or CD11b specificity were detected with the MAIGA. Sera from 7 patients with secondary AIN contained in all cases antibodies with pan-FcγRIIIb specificity, as deduced from the indirect GIFT results and absorbance/elution experiments performed with 2 sera. The MAIGA confirmed this for only 1 of the 5 tested sera. Furthermore, CD11a antibodies were detected in 1 of the 5 tested sera. In conclusion, our results indicate that primary AIN is usually associated with NA-specific antibodies, whereas secondary AIN seems to be associated with pan-FcγRIIIb antibodies. Thus, characterization of the antibodies in sera from children with AIN discriminates patients with primary AIN from those with secondary AIN.

Description: We analyzed a genetic polymorphism of Fcγ receptor IIIa (CD16) that is present on position 158 (Phe or Val) in the membrane-proximal, IgG-binding domain. With a polymerase chain reaction–based allele-specific restriction analysis assay we genotyped 87 donors and found gene frequencies of 0.57 and 0.43 for FcγRIIIA-158F and −158V, respectively. A clear linkage was observed between the FcγRIIIA-158F and −48L genotypes on the one hand and the FcγRIIIA-158V and −48H or −48R genotypes on the other hand (χ2 test; P 〈 .001). To determine the functional consequences of this FcγRIIIa-158V/F polymorphism, we performed IgG binding experiments with natural killer (NK) cells from genotyped donors. All donors were also typed for the recently described triallelic FcγRIIIa-48L/R/H polymorphism. NK cells were treated with lactic acid to remove cell-associated IgG. FcγRIIIaNK158F bound significantly less IgG1, IgG3, and IgG4 than did FcγRIIIaNK-158V, irrespective of the FcγRIIIa-48 phenotype. Moreover, freshly isolated NK cells from FcγRIIIa-158VV individuals carried significantly more cytophilic IgG than did NK cells from FcγRIIIa-158FF individuals. In addition, CD16 monoclonal antibody (MoAb) MEM154 bound more strongly to FcγRIIIa-158V, compared with -158F, again independently of the FcγRIIIa-48 phenotype. The binding of MoAb B73.1 was not influenced by the FcγRIIIa-158V/F polymorphism, but proved to depend solely on the amino acid present at position 48 of FcγRIIIa. In conclusion, the previously reported differences in IgG binding among the three FcγRIIIa-48L/R/H isoforms are a consequence of the linked, biallelic FcγRIIIa-158V/F polymorphism at amino-acid position 158.

Description: Recently, a new alloantigen on IgG Fc receptor type IIIb (FcγRIIIb), SH, was described (Bux et al, Blood 89:1027, 1997). We identified three healthy individuals whose neutrophils reacted positively with the SH antiserum. The neutrophil antigen (NA) phenotype of all three donors was NA(1+,2+). Analysis of genomic DNA showed that the three donors were positive for the described SH-encoding mutation in the NA2-FcγRIIIB gene, 266C→A. However, NA(1,2) genotyping and nucleotide sequencing of an NA2-specific fragment amplified from the genomic DNA fragment showed that these individuals carried three FcγRIIIBgenes, namely, NA1-FcγRIIIB, NA2-FcγRIIIB, andSH-FcγRIIIB, encoding NA1-FcγRIIIb, NA2-FcγRIIIb, and SH-FcγRIIIb, respectively. Southern blot analysis confirmed these findings. Furthermore, all three transcripts were isolated from neutrophil mRNA. To investigate whether the presence of threeFcγRIIIB genes resulted in a higher membrane expression of FcγRIIIb, we measured the reactivity of neutrophils from NA(1+,2+)SH(+) individuals with a panel of CD16 monoclonal antibodies (MoAbs) in comparison with neutrophils from NA(1+,2+)SH(−) controls. Reactivity of four different anti–pan-FcγRIII MoAbs and NA2-specific MoAb GRM1 was higher with SH(+) neutrophils compared with controls, whereas that of NA1-specific MoAbs was similar, which is in concordance with the results from the genomic analysis. We observed that reactivity with NA2-specific CD16 MoAb PEN1 was sixfold higher in SH(+) individuals compared with controls. Apparently, the 60Ala→Asp substitution in SH-FcγRIIIb influences the epitope recognized by PEN1. In conclusion, we identified three NA(1+,2+)SH(+) individuals carrying three FcγRIIIB genes and observed a clear gene-dosage effect on the level of expression of neutrophil FcγRIIIb.