A standardized immune phenotyping and automated data analysis platform for multicenter biomarker studies

The analysis and validation of flow cytometry-based biomarkers in clinical studies are limited by the lack of standardized protocols that are reproducible across multiple centers and suitable for use with either unfractionated blood or cryopreserved PBMCs. Here we report the development of a platform that standardizes a set of flow cytometry panels across multiple centers, with high reproducibility in blood or PBMCs from either healthy subjects or patients 100 days after hematopoietic stem cell transplantation. Inter-center comparisons of replicate samples showed low variation, with interindividual variation exceeding inter-center variation for most populations (coefficients of variability <20% and interclass correlation coefficients >0.75). Exceptions included low-abundance populations defined by markers with indistinct expression boundaries (e.g., plasmablasts, monocyte subsets) or populations defined by markers sensitive to cryopreservation, such as CD62L and CD45RA. Automated gating pipelines were developed and validated on an independent data set, revealing high Spearman's correlations (rs >0.9) with manual analyses. This workflow, which includes pre-formatted antibody cocktails, standardized protocols for acquisition, and validated automated analysis pipelines, can be readily implemented in multicenter clinical trials. This approach facilitates the collection of robust immune phenotyping data and comparison of data from independent studies.

[Identification of proteins associated with transcription factors HOXA9 and E2A-PBX1 by tandem affinity purification]

Chimeric transcription factor E2A-PBX1 induces the development of acute lymphoblastic B-cell leukemia in children. Using a transgenic mouse model, we previously demonstrated that homeobox (HOX) gene HOXA9 genetically interact with E2A-PBX1 gene in the development of B-cell leukemia in mice. HOXA9 itself is a potent oncogene resulting in myeloid leukemia when overexpressed, which is strongly accelerated by its collaborator Meis1. HOX, PBX1 and MEIS1 proteins have been shown to form hetero dimeric or trimeric complexes in different combinations. Cooperative interaction between PBX1 and HOX proteins enhances their DNA binding specificity, essential for HOX dependent developmental programs. PBX1 is retained in E2A-PBX1, and thus the strong transcriptional activator properties of E2A-PBX1 may lead to aberrant activation of normally repressed targets of HOX-PBX complexes. However, although there is evidence that E2A-PBX1 could bind to HOX and MEIS1 proteins it is still unclear whether such complexes are actually required for leukemic transformation or whether E2A-PBX1 and HOXA9 are each part of larger protein complexes acting in independent complementing oncogenic pathways. In this study we aim to search for other HOXA9 and E2A-PBX1 interacting proteins. To identify novel proteins interacting with human E2A-PBX1 or HOXA9 we used tandem affinity purification (TAP) of protein complexes from 697 pre-B leukemic and HeLa cell lines transduced to express E2A-PBX1 or HOXA9, respectively, with covalently attached FLAG/HA peptides. The protein composition of each complex was determined using tandem mass-spectrometry. In the E2A-PBX1 containing complex we identified lymphoid transcription factor IKAROS, chromatin remodeling factors of SWI/SNF family while multiple subunits of translation initiation factor eIF3, E3 ubiquitin ligase UBR5 emerged from the HOXA9 complex as potential critical protein partners. This is the first time the protein partners of either E2A-PBX1 or HOXA9 oncoproteins were identified using an unbiased biochemical approach. The identification of translation initiation factors associated with HOXA9 might indicate a novel function for HOX proteins independent of their transcriptional activity.

HOXA4 provides stronger engraftment potential to short-term repopulating cells than HOXB4

Genes of the HOX4 paralog group have been shown to expand hematopoietic stem cells (HSCs). Endogenous expression of HOXA4 is 10-fold higher than HOXB4 in embryonic primitive hematopoietic cells undergoing self-renewal suggesting a more potent capacity of HOXA4 to expand HSC. In this study, we provide evidence by direct competitive bone marrow cultures that HOXA4 and HOXB4 induce self-renewal of primitive hematopoietic cells with identical kinetics. Transplantation assays show that short-term repopulation by HOXA4-overexpressing multilineage progenitors was significantly greater than HOXB4-overexpressing progenitors in vivo, indicating differences in the sensitivity of the cells to external signals. Small array gene expression analysis showed an increase in multiple Notch and Wnt signaling -associated genes, including receptors and ligands, as well as pluripotency genes, for both HOXA4- and HOXB4-overexpressing cells, which was more pronounced for HOXA4, suggesting that both HOX proteins may assert their affects through intrinsic and extrinsic pathways to induce self-renewal of primitive hematopoietic cells. Thus, HOXA4 increases short-term repopulation to higher levels than HOXB4, which may involve Notch signaling.

Self-enforcing feedback activation between BCL6 and pre-B cell receptor signaling defines a distinct subtype of acute lymphoblastic leukemia

Studying 830 pre-B ALL cases from four clinical trials, we found that human ALL can be divided into two fundamentally distinct subtypes based on pre-BCR function. While absent in the majority of ALL cases, tonic pre-BCR signaling was found in 112 cases (13.5%). In these cases, tonic pre-BCR signaling induced activation of BCL6, which in turn increased pre-BCR signaling output at the transcriptional level. Interestingly, inhibition of pre-BCR-related tyrosine kinases reduced constitutive BCL6 expression and selectively killed patient-derived pre-BCR(+) ALL cells. These findings identify a genetically and phenotypically distinct subset of human ALL that critically depends on tonic pre-BCR signaling. In vivo treatment studies suggested that pre-BCR tyrosine kinase inhibitors are useful for the treatment of patients with pre-BCR(+) ALL.

Inability of HOXB4 to enhance self-renewal of malignant B cells: favorable profile for the expansion of autologous hematopoietic stem cells

Leukemic stem cells share self-renewal properties and slow proliferation with hematopoietic stem cells. Based on expression signatures, it has been suggested that these cells use the same molecular pathways for these processes. However, it is not clear whether leukemic stem cells also respond to factors known to enhance the self-renewal activity of hematopoietic stem cells. The transcription factor homeobox B4 (HOXB4) is known to induce expansion of mouse hematopoietic stem cells. The recombinant TAT-HOXB4 protein also expands human CD34+ cells. In this study we investigated whether overexpression of HOXB4 could increase leukemic initiating cell numbers, an issue that is crucial to its clinical usage. A transgenic mouse model for E2A-PBX1 induced pre-B acute lymphoblastic leukemia was used in combination with HOXB4 transgenic mice to test oncogenic interactions between HOXB4 and E2A-PBX1. The frequency of leukemic initiating cells retrovirally overexpressing HOXB4 was measured by transplantation at limiting dilution and evaluation of leukemia development in recipient mice. Moreover, human B cell lines were evaluated for their colony forming cell potential upon exposure to TAT-HOXB4 protein. Our data with the mouse models show that HOXB4 neither accelerates the generation of E2A-PBX1 B cell leukemia nor expands the number of leukemia initiating cells. Additionally, the growth or colony forming cell proportions of human B cell lines was not changed by HOXB4, suggesting that human B leukemic initiating cells are not affected by HOXB4.

Differentiation and cell-type-restricted expression of HOXC4, HOXC5 and HOXC6 in myeloid leukemias and normal myeloid cells

HOX genes have shown a lineage-specific expression in hematopoiesis and are suggested as being involved in the expression of certain adhesion molecules. Recently, we have demonstrated that HOXC4 and HOXC6, but not HOXC5, are expressed during lymphoid differentiation. Reports on the expression of these genes in myeloid leukemias and normal myeloid cells are still scarce. Therefore, we have investigated the expression of HOXC4, HOXC5 and HOXC6 in purified subpopulations of bone marrow in addition to 36 specimens of acute myeloid leukemias (AMLs), eight chronic myeloid leukemias (CMLs), several myeloid cell lines and cutaneous localizations of three myelomonocytic leukemias and one granulocytic sarcoma by RT-PCR and partly by RNA in situ hybridization (RISH). HOXC4 and HOXC6 transcripts were both detected by RT-PCR in 22/36 and 24/36 AMLs, respectively. The distribution of HOXC4 and HOXC6 gene expression over the different types of AML was largely similar and covered all types of AML. In contrast, HOXC5 gene expression was found in only 6/32 AMLs. Expression of HOXC5 was restricted to AMLs of the granulocytic (FAB M1-M3), early monocytic (FAB M4) and early erythroid (FAB M6) lineage. In general, except in one FAB M5b case, no expression of HOXC5 was found in AMLs derived from late stages of monocytic (FAB M5) and megakaryocytic (FAB M7) lineages. As for HOXC4 and HOXC6, expression of HOXC5 was absent in CMLs. Using RISH significant HOXC4, HOXC5 and HOXC6 expression was found in a number of additionally studied AML samples of different FAB classification (M2, M4, M5b and M5b), (M2 and M5b) (M2, M4, M5b), respectively. In tissue localizations of leukemias a different expression pattern of HOXC4, HOXC5 and HOXC6 was found. In contrast to mature leukemic stages of myeloid differentiation, these skin localizations of leukemias expressed HOXC5 and HOXC6. HOXC4 expression was found both in leukemic cells derived from peripheral blood and from cutaneous localizations. Besides HOXC4 expression in monocytes no expression of HOXC4, HOXC5 and HOXC6 was found in granulocytes and monocytes, colonies of growth factor-induced CD34+ bone marrow cells. In earliest CD34+/CD38low and high cell fractions of bone marrow only HOXC4 and in megakaryocytic cells both HOXC4 and HOXC6 were found. Thus, the expression patterns of these HOXC genes found in the limited number of cell fractions of normal bone marrow suggest that the expression patterns found in AMLs and CMLs might reflect the normal situation. Furthermore, the presence of HOXC5 and HOXC6 expression specifically in skin infiltrates of late differentiation stages of myeloid leukemias, suggests an additional role for these genes in the positioning of these myeloid cells in skin tissue.

HOXC4, HOXC5, and HOXC6 expression in non-Hodgkin's lymphoma: preferential expression of the HOXC5 gene in primary cutaneous anaplastic T-cell and oro-gastrointestinal tract mucosa-associated B-cell lymphomas

Most of the 39 members of the homeobox (HOX) gene family are believed to control blood cell development. HOXC4 and HOXC6 gene expression levels increase with differentiation of lymphoid cells. In contrast, HOXC5 is not expressed in the lymphoid lineage, but was found in lymphoid cell lines, representing the neoplastic equivalents of various differentiation stages of T and B lymphocytes. In the present study, we investigated the HOXC4, HOXC5, and HOXC6 gene expression pattern in 89 non-Hodgkin's lymphomas (NHLs) of different histologic subtypes and originating from different sites. Using RNA in situ hybridization and semiquantitative reverse transcription-polymerase chain reaction, we found expression of HOXC4 in 83 of 88 and HOXC6 in 77 of 88 NHLs and leukemias investigated. In contrast, HOXC5 expression was found in only 26 of 87 NHLs and appeared to be preferentially expressed by two specific subsets of lymphomas, ie, primary cutaneous anaplastic T-cell lymphomas (9 of 9) and extranodal marginal zone B-cell lymphomas (maltomas; 7 of 9). These results indicate that, in contrast to HOXC4 and HOXC6, HOXC5 shows a type- and site-restricted expression pattern in both T- and B-cell NHLs.

HOXC4, HOXC5, and HOXC6 expression in primary cutaneous lymphoid lesions. High expression of HOXC5 in anaplastic large-cell lymphomas

Homeobox (HOX) genes are involved in the lineage-specific differentiation of bone marrow stem cells. Recently, we reported a largely similar expression pattern of HOXC4 and HOXC6 in normal and neoplastic cells of the lymphoid lineage. In contrast, HOXC5 was specifically expressed in a subset of B-cell non-Hodgkin's lymphomas (B-NHL) but not in normal lymphocytes or lymphoid leukemias. This might suggest a role for HOXC5 in the pathogenesis of these lymphomas. In the present study the expression of HOXC4, HOXC5, and HOXC6 in primary cutaneous lymphomas was investigated. Using RNA in situ hybridization (RISH) and semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR), we found strong expression of HOXC4 and HOXC6 in all, except one, primary cutaneous lymphomas and all reactive cutaneous lymphoid infiltrates. Interestingly, a strong expression of HOXC5 in primary anaplastic CD30+ large T-cell lymphomas was found. RISH was consistently negative for HOXC5 in all other types of primary cutaneous B- and T-cell lymphomas. However, by semiquantitative RT-PCR these lymphomas showed a weak expression of HOXC5 mRNA. Therefore, we concluded that these lymphomas express low constitutive levels of HOXC5 mRNA. Furthermore, HOXC5 expression was consistently absent in reactive cutaneous lymphoid infiltrates, hyperplastic tonsils and lymph nodes, and peripheral blood lymphocytes either unstimulated or stimulated by a cocktail of CD3 and CD28 antibodies. As a strong expression of HOXC5 in primary cutaneous lymphomas was observed only in anaplastic large T-cell lymphomas and reactive control tissues lacked HOXC5 expression, these data strongly support a role for HOXC5 in the genesis of anaplastic large-T-cell lymphomas.

Expression of the homeobox gene HOXC4 in keratinocytes of normal skin and epithelial skin tumors is correlated with differentiation

Homeobox (HOX) genes share a highly conserved 183-bp sequence. The encoded proteins are capable of binding to specific DNA sequences and functioning as transcription factors. HOX genes play a critical role in the temporal and spatial differentiation of cells during embryogenesis. In several adult tissues, HOX genes are expressed in a constant, tissue-specific pattern, whereas in malignant tumors of these tissues an altered expression pattern was found. We investigated the expression of HOXC4 in adult normal skin by reverse-transcription polymerase chain reaction and non-radioactive RNA in situ hybridization. Moreover, HOXC4 expression was studied in various epidermal neoplasms (solar keratosis, six specimens; Bowen's disease, four; squamous cell carcinoma, nine; basal cell carcinoma, three) by RNA in situ hybridization. HOXC4 was found to be expressed in the suprabasal layers of the epidermis in normal skin specimens and the adjacent non-lesional epidermis of all other specimens. Atypical keratinocytes of solar keratoses and Bowen's disease as well as basaloid cells of basal cell carcinomas were negative. In squamous cell carcinoma, well differentiated areas with keratinization showed HOXC4 expression, whereas poorly differentiated areas were negative. Immunostaining with an antibody against cytokeratin 10, a marker of epidermal differentiation, was performed. A good correlation between the distribution pattern of HOXC4 and cytokeratin 10 in the lesions examined was found. These results suggest that HOXC4 is expressed mainly in differentiated keratinocytes. Lack of differentiation (as in neoplastic cells) is accompanied by downregulation of HOXC4 expression.

Quantification of biotinylated RNA probes for in situ hybridization using chemiluminescence

For reliable detection of mRNA by non-radioactive in situ hybridization, calibration and standardization of the individual steps involved are essential. We describe a method that allows determination of the size and integrity as well as quantification of biotinylated RNA probes in a single experiment. Serial dilutions of biotinylated RNA probes generated by promoter-mediated in vitro transcription were size-separated by gel electrophoresis in the presence of known amounts of 5'-biotinylated oligomers which served as internal standard. Following immobilization onto nylon membranes and visualization by chemiluminescence, optical densities of probes and internal standards were measured by densitometry and analysed by linear regression. RNA probes complementary to the human homeobox genes HOX-C6, -C8 and -C9 were quantified. Four different 5'-biotinylated oligomers (20, 35, 50 and 75 bases) were tested as internal standards. Concerning the separation of probe and oligomer in the gel, transfer properties and efficiency of binding to the membrane, the oligomer of 35 bases was found to be the best internal standard with highest reproducibility. Comparison of probe concentration by spectrophotometry and the described method showed a good correlation, indicating that our method is a reliable assay for quantitative and qualitative control of biotin-labelled probes.

Two computer programs using logit transformation for the analysis of S-shaped curves

Two BASIC computer programs using logit transformation for the analysis of S-shaped curves are presented. The first program estimates the median effective dose (ED50) of drugs. A maximum likelihood method similar to Finney's algorithm of probit analysis is applied. The second fits S-shaped curves to empirical data which cannot be transformed easily to proportions. Logit transformation proves to be a simple and accurate alternative to probit transformation when computing memory is limited or calculation speed is critical.