Search

The discovery of gene fusion mutations, particularly in leukemia, has consistently identified new cancer pathways and led to molecular diagnostic assays and molecular-targeted chemotherapies for cancer patients. Here, we report our discovery of a novel CREB3L2-PPARγ fusion mutation in thyroid carcinoma with t(3;7)(p25;q34), showing that a family of somatic PPARγ fusion mutations exist in thyroid cancer. The CREB3L2-PPARγ fusion encodes a CREB3L2-PPARγ fusion protein that is composed of the transactivation domain of CREB3L2 and all functional domains of PPARγ1. CREB3L2-PPARγ was detected in

BACKGROUND: Most alloantigens on transfused red blood cells (RBCs) are weakly immunogenic, with only a 2 to 6 percent overall immunization rate even in patients receiving multiple transfusions. Although recipient genetics may contribute to responder and/or nonresponder status, in most cases HLA type does not predict humoral response to RBC antigens. In contrast, rates of alloimmunization do correspond to the underlying disease status of transfusion recipients, suggesting that acquired host factors may play an important role. In this context, it was hypothesized that the inflammatory status of a transfusion recipient would influence immunization to transfused RBCs. STUDY DESIGN AND METHODS: A novel murine model for alloimmunization to RBC antigens was developed with the mHEL mouse, which expresses hen egg lysozyme (HEL) as a model blood group antigen. Leukoreduced mHEL RBCs were transfused into wild-type recipient mice, and anti-HEL responses were monitored. To test the stated hypothesis, some recipient animals were injected with polyinosinic polycytidylic acid (poly(I:C)), a synthetic double-stranded RNA molecule that induces viral-like inflammation. RESULTS: Similar to the immunogenicity of most RBC antigens in humans, transfusion of mHEL RBCs into uninflamed mice was only a weak immunogen. In contrast, poly(I:C)-treated mice had a significant increase in both the frequency and the magnitude of alloimmunization to the mHEL antigen. CONCLUSIONS: These findings demonstrate that recipient inflammation with poly(I:C) significantly enhances humoral immunization to transfused alloantigens in a murine model. Moreover, these data suggest that the inflammatory status of human transfusion recipients may regulate the immunogenicity of transfused RBCs.

Transfusion of red blood cells (RBCs) into patients with anti–donor RBC antibodies (crossmatch-incompatible transfusion) can result in lethal antibody-mediated hemolysis. Less well appreciated is the ability of anti-RBC antibodies to specifically remove their target antigen from donor RBCs without compromising cell survival or adversely affecting the transfusion recipient. In an effort to elucidate the mechanistic details of this process, we describe the first animal model of nonhemolytic antibody-induced RBC antigen loss. RBCs from transgenic mHEL mice express surface hen egg lysozyme (HEL) as a transmembrane protein. Transfusion of mHEL RBCs into mice immunized with HEL results in selective loss of HEL antigen from donor RBCs without affecting other blood group antigens or reducing the circulatory life span of the transfused RBCs. While this process does not require the presence of a spleen, it requires both anti-RBC immunoglobulin G (IgG) antibodies and the FcγIII receptor. These studies provide mechanistic insight into the phenomenon of antigen loss during incompatible transfusion in humans.

Follicular thyroid carcinoma (FTC) frequently harbors the PAX8/PPARgamma fusion gene (PPFP); however, its oncogenic role and mechanism(s) of action remain undefined. We investigated PPFP's effects on cell growth, apoptosis, cell-cell, and cell-matrix interactions in immortalized human thyroid cells (Nthy-ori 3-1) and NIH 3T3 cells. PPFP expression increased the growth of transient and stable Nthy-ori transfectants ( approximately threefold by 72 h). There was an 8.4% increase of cells in the S+G2/M phase, a 7.8% decrease in cells in the G0+G1 phase and a 66% decline in apoptosis at 72 h. Stable Nthy-ori PPFP transfectants grew in soft agar, and PPFP-transfected NIH 3T3 cells exhibited efficient focus formation, suggesting loss of anchorage-dependent growth and contact inhibition, respectively. Overexpression of PPARgamma in Nthy-ori cells did not recapitulate PPFP's growth effects. Treatment of Nthy-ori cells with an irreversible PPARgamma inhibitor mimicked the growth-promoting effects of PPFP and co-expression of PPFP and PPARgamma blocked PPARgamma transactivation activity. Our data provide functional evidence that PPFP acts as an oncoprotein, whose transforming properties depend in part on inhibition of PPARgamma. Our data suggest that PPFP contributes to malignant transformation during FTC oncogenesis by acting on several cellular pathways, at least some of which are normally regulated by PPARgamma.

A series of nonhydrolyzable ubiquitin dimer analogues has been synthesized and evaluated as inhibitors of ubiquitin-dependent processes. Dimer analogues were synthesized by cross-linking ubiquitin containing a terminal cysteine (G76C) to ubiquitin containing cysteine at position 11 ((76-11)Ub(2)), 29 ((76-29)Ub(2)), 48 ((76-48)Ub(2)), or 63 ((76-63)Ub(2)). A head-to-head dimer of cysteine G76C ((76-76)Ub(2)) served as a control. These analogues are mimics of the different chain linkages observed in natural polyubiquitin chains. All analogues showed weak inhibition toward the catalytic domain of UCH-L3 and a UBP pseudogene. In the absence of ubiquitin, isopeptidase T was inhibited only by the dimer linked through residue 29. In the presence of 0.5 microM ubiquitin, isopeptidase T was inhibited by several of the dimer analogues, with the (76-29)Ub(2) dimer exhibiting a K(i) of 1.8 nM. However, USP14, the human homologue of yeast Ubp6, was not inhibited at the concentrations tested. Some analogues of ubiquitin dimer also acted as selective inhibitors of conjugation and deconjugation of ubiquitin catalyzed by reticulocyte fraction II. (76-76)Ub(2) and (76-11)Ub(2) did not inhibit the conjugation of ubiquitin, while (76-29)Ub(2), (76-48)Ub(2), and (76-63)Ub(2) were potent inhibitors of conjugation. This specificity is consistent with the known ability of cells to form K29-, K48-, and K63-linked polyubiquitin chains. While (76-11)Ub(2), (76-29)Ub(2), and (76-63)Ub(2) inhibited release of ubiquitin from a pool of total conjugates, (76-48)Ub(2) and (76-76)Ub(2) showed no significant inhibition. Isopeptidase T was shown to specifically disassemble two conjugates (assumed to be di- and triubiquitin with masses of 26 and 17 kDa) formed in the reticulocyte lysate system. This activity was inhibited differentially by all dimer analogues. The inhibitor selectivity for deconjugation of the 26 and 17 kDa conjugates was similar to that observed for isopeptidase T. The observations suggest that these two conjugated proteins of the reticulocyte lysate are specific substrates for isopeptidase T in lysates.

Increased rates of graft rejection after bone marrow transplantation (BMT) are observed in patients whose illnesses— such as sickle cell disease, thalassemia, and aplastic anemia—necessitate chronic transfusion before BMT. Because BM transplants in these patients are routinely HLA matched, any immunization responsible for increased rejection is likely against minor histocompatibility antigens (mHAs). It has been assumed that contaminating leukocytes in red blood cell (RBC) units are the main sources of immunization to mHAs. However, in this report, we demonstrate that antigens on donor RBCs are presented in the major histocompatibility complex (MHC) class I pathway of recipient antigen-presenting cells, resulting in activation and expansion of recipient CD8+ T cells specific for donor mHAs. Given that human hematopoietic progenitor cells express many of the known mHAs, this observation provides a mechanism by which chronic transfusion of even stringently leukoreduced RBCs may result in sufficient mHA immunization to increase the frequency of BMT rejection.

Alloantibodies that inhibit the function of wildtype (wt) FVIII molecules constitute the most serious complication of replacement therapy for hemophilia (H)A patients. Although it is not possible to accurately predict which patients will develop inhibitors, ethnicity is a known risk factor as alloimmunization occurs ~2−fold more often in African Americans (AAs) than in Caucasians (Cs). Analysis of data from a resequencing study revealed that the FVIII gene (F8) contains four nonsynonymous−single nucleotide polymorphisms (ns−SNPs) that encode six distinct wt FVIII proteins designated haplotype (H)1−H6. While AAs express all but H6, only H1 and H2, the two recombinant (r)−FVIII proteins used clinically, are expressed by Cs. About 25% of AAs express H3, H4, or H5, proteins that differ structurally from r−FVIII at three ns−SNPs (R484H, D1241E, M2238V), two of which are located in dominant inhibitor epitopes and have AA−restricted minor−alleles. We designed the Pharmacogenetics of Inhibitor Risk (PIR) study to test the hypothesis that these ns−SNPs, when allelically mismatched in replacement therapy, are pharmacogenetic risk factors that contribute to the greater incidence of inhibitor development in patients of African−descent. We enrolled 94 of 223 total AA HA patients from collaborating treatment centers. Table 1 presents characteristics of the first 47 subjects (whose functional F8 regions have been resequenced entirely) including: FVIII haplotype (exposure) − H3/H4/H5 vs H1/H2; inhibitor status (outcome) − yes/no; F8 mutation − high−risk vs low−risk types; other covariates and potential confounders. A preliminary analysis demonstrated an odds ratio of 3.6 for inhibitors developing in the exposed subjects (P=0.10; 95% CI=0.7–17.6). Our finding suggests that pharmacogenetic factors may indeed contribute to the increased incidence of this alloimmune complication in AAs. To increase the power to detect a true association and allow adjustment for confounders, we are currently completing F8 resequencing in the next 47 patients. Table 1. Relevant characteristics of the first 47 African-American hemophilia-A PIR study patients*

Background: Transfusion of red blood cells (RBC) into patients with anti-donor RBC antibodies (crossmatch incompatible transfusion) can result in antibody mediated hemolysis. Less well appreciated is the ability of anti-RBC antibodies to specifically remove their target antigen from donor RBCs without compromising cell survival. This phenomenon has now been reported for the major clinically significant blood group antigens, including Rh, Kell, Kidd and Duffy. Although this has been described multiple times in humans, no mechanistic elucidation has been accomplished. In an effort to investigate the mechanism of this process, we describe the first animal model of non-hemolytic antibody induced RBC antigen loss. Methods: mHEL mice express the model antigen Hen Egg Lysozyme (HEL) as a cell surface protein on RBC. Since mHEL mice are on a C57BL/6 background, the mHEL antigen represents a single antigenic difference between donor RBC and recipient mice. Immunizing C57BL/6 mice with HEL/CFA results in the generation of high titer IgG anti-HEL responses rendering the mice crossmatch incompatible with mHEL RBC. This system was utilized to study the effects of transfusing mHEL RBC into crossmatch incompatible recipients. Results: Similar to the antibody induced antigen loss observed in humans, transfusion of donor mHEL RBC into crossmatch incompatible mice results in selective loss of HEL antigen from donor RBC without affecting other blood group antigens or reducing the circulatory lifespan of the donor RBC. In addition, recovered RBC that have lost their antigen have normal morphology. This process is antigen specific and occurs in mice that have received passive injections of anti-HEL antisera. A spleen is not required for antigen loss to occur. However, antigen loss does not occur in animals with a targeted deletion of the FcγIII receptor. Although polyclonal anti-HEL antisera consistently causes antigen loss, and IgG1 and IgG2b are the predominant subclasses of anti-HEL IgG in the antisera, no antigen loss is observed in response to purified monoclonal anti-HEL antibodies of the IgG1 and IgG2b subclass. Conclusion: These studies demonstrate that antibody induced antigen loss is a process that involves interaction of RBC, anti-RBC IgG and FcγIII receptors, thus providing mechanistic insight into the phenomenon of antigen loss during incompatible transfusion. The lack of antigen loss in response to monoclonal anti-HEL IgG1 or IgG2b suggests that antigen loss occurs in response to a minor IgG subtype in antisera, depends upon biological properties of the antibody (such as affinity), or that additional serum cofactors are involved.

Background: HLA matched related bone marrow transplantation (BMT) following myeloablative treatment with busulfan, cyclophosphamide and ATG is an effective therapy for children with severe sickle cell disease (SCD). However, a five percent risk for transplant related mortality precludes its use in less severely affected children, and a much higher risk in adults prevents its use in older patients. Low dose total body irradiation based conditioning regimens may provide a safer approach to BMT for patients with SCD. However, although these non-myeloablative regimens reliably allow for sustained engraftment in patients with malignant diseases, there is a high incidence of graft rejection in patients with SCD. One distinct variable in the sickle cell population is chronic transfusion with multiple units of red blood cells (RBC), which may lead to immunization to antigens present on donor bone marrow. Because BMTs in this setting are HLA matched, any immunization responsible for increased rejection is likely against minor histocompatibility antigens (mHAs). It has been assumed that contaminating leukocytes in RBC units are the main source of immunization to mHAs. However, a recent analysis of patients undergoing HLA matched BMT for aplastic anemia performed by the International bone marrow transplant registry, indicates that rejection of multiply transfused patients remains high despite the use of leukoreduced blood products. We hypothesized that RBC are alone sufficient to immunize to mHAs through consumption of donor RBC by recipient antigen presenting cells (APC) and crosspresentation of donor peptides in recipient MHC I molecules. Methods: Chicken Ovalbumin (OVA) is a well studied antigen. We converted OVA into a model RBC associated antigen using a maleimide-sulfhydryl based crosslinking strategy to couple OVA to RBC (OVA-RBC). To monitor activation of CD8+ T cells specific for OVA, we utilized mice transgenic for a T cell receptor that recognizes an OVA peptide 257–264 (SIINFEKL) presented by MHC class I H-2Kb (OT-I mice). Leukoreduced (OVA-RBC) from B10.BR (H-2k) donor mice were transfused into C57BL/6 (H-2b) mice that had been adoptively transferred with OT-I T cells. Activation, and expansion of OT-I cells was assessed by flow cytometry using specific tetramer reagents. Results: Although their lifespan is somewhat decreased, OVA-RBC continue to circulate for 10–14 days with stable persistence of OVA on the RBC. In response to transfusion of OVA-RBC, OT-I T cells expanded 49 fold compared to untransfused mice (p value = .0002). This was antigen specific, as no expansion was seen with RBC crosslinked to a third party antigen (hen egg lysozyme). Moreover, the observed T cell expansion required that the antigen be associated with the RBC, as no expansion was observed after injection with soluble OVA. This experimental design precludes direct presentation of OVA by donor cells, as donor cells do not express the MHC (H-2Kb) to which OT-I T cells are restricted. Discussion: Based upon the above data, we conclude that mHAs on transfused RBC crossprime into the MHC class I pathway of recipient APC. Since human hematopoietic progenitor cells express many of the known mHAs, this observation provides a mechanism by which chronic transfusion of even stringently leukoreduced RBC may result in sufficient mHA immunization to increase the frequency of BMT rejection.

A complementary DNA (cDNA) for ubiquitin carboxyl-terminal hydrolase isozyme L3 was cloned from human B cells. The cDNA encodes a protein of 230 amino acids with a molecular mass of 26.182 daltons. The human protein is very similar to the bovine homolog, with only three amino acids differing in over 100 residues compared. The amino acid sequence deduced from the cDNA was 54% identical to that of the neuron-specific protein PGP 9.5. Purification of bovine PGP 9.5 confirmed that it is also a ubiquitin carboxyl-terminal hydrolase. These results suggest that a family of such related proteins exists and that their expression is tissue-specific.