Two-dimensional gene expression fingerprinting.

In the present study we have developed the two-dimensional Gene Expression Fingerprinting (2-D GEF) procedure suitable for gene expression analysis and new gene discovery. The procedure is based on the two-dimensional gel display of 3'-terminal cDNA restriction fragments produced by one primary (first dimension) and several sequential secondary restriction digestions. Many thousands of individual sequences per cDNA sample can be visualized using this approach, which is also characterized by a high reproducibility, predictable spatial location of cDNA fragments on 2-D gels, and the potential for identifying cDNA fragments solely on the basis of their two-dimensional coordinates. Using this 2-D GEF method, we analyzed and compared the gene expression patterns of two related primitive hematopoietic cell lines, Kg-1 and Kg-1a. A total of 25 candidate differentially expressed sequences were identified, and for 75% of them the presumed expression pattern was confirmed by Northern blotting or reverse transcription-polymerase chain reaction. We also demonstrated that for 70% of bands, correct prediction of their identity could be made on the basis of two-dimensional coordinates, whereas the major part of incorrect predictions was caused by insufficient database quality.

Analysis of gene expression in subpopulations of murine hematopoietic stem and progenitor cells.

OBJECTIVE: The aim of the present work was to study how functional differences between subsets of the murine hematopoietic stem/progenitor cell compartment are manifested on the level of different patterns of gene expression in these subsets. MATERIALS AND METHODS: Amplified 3' terminal total cDNA fragment populations from four stem and progenitor cell fractions sorted using differential staining with Rhodamine 123 were prepared, and gene expression patterns were analyzed by Southern hybridization with a panel of gene markers. RESULTS: For the vast majority of lineage-specific markers, no expression was detected in the long-term repopulating stem cell fraction. Expression of a number of key genes positively regulating entry and progression through the cell cycle was down-regulated in long-term repopulating cells, in accordance with the quiescent state of the latter. In contrast, certain but not all cell division kinase inhibitors were significantly up-regulated in long- and short-term repopulating stem cell fractions. Expression of several genes important for entry into the apoptotic pathway was moderately reduced in long-term repopulating cells. Messenger RNA levels of the transcription factors GATA-1, GATA-2, c-Myb and SCL were down-regulated in long-term repopulating cells, as compared to more mature stem/progenitor cells. Finally, expression of the MDR1a gene encoding the Pgp efflux pump was highest in long-term repopulating cells, and progressively decreased with maturation. CONCLUSION: The patterns of gene expression in the stem/progenitor cell fractions are in good correlation with the known properties of adult hematopoietic stem/progenitor cells and may provide insight into molecular mechanisms underlying stem cell physiology.

Fluorescence-based selection of gene-corrected hematopoietic stem and progenitor cells from acid sphingomyelinase-deficient mice: implications for Niemann-Pick disease gene therapy and the development of improved stem cell gene transfer procedures.

The general utility of a novel, fluorescence-based procedure for assessing gene transfer and expression has been demonstrated using hematopoietic stem and progenitor cells. Lineage-depleted hematopoietic cells were isolated from the bone marrow or fetal livers of acid sphingomyelinase-deficient mice, and retrovirally transduced with amphotropic or ecotropic vectors encoding a normal acid sphingomyelinase (ASM) cDNA. Anti-c-Kit antibodies were then used to label stem- and progenitor-enriched cell populations, and the Bodipy fluorescence was analyzed in each group after incubation with a Bodipy-conjugated sphingomyelin. Only cells expressing the functional ASM (ie, transduced) could degrade the sphingomyelin, thereby reducing their Bodipy fluorescence as compared with nontransduced cells. The usefulness of this procedure for the in vitro assessment of gene transfer into hematopoietic stem cells was evaluated, as well as its ability to provide an enrichment of transduced stem cells in vivo. To show the value of this method for in vitro analysis, the effects of retroviral transduction using ecotropic versus amphotropic vectors, various growth factor combinations, and adult bone marrow versus fetal liver stem cells were assessed. The results of these studies confirmed the fact that ecotropic vectors were much more efficient at transducing murine stem cells than amphotropic vectors, and that among the three most commonly used growth factors (stem cell factor [SCF] and interleukins 3 and 6 [IL-3 and IL-6]), SCF had the most significant effect on the transduction of stem cells, whereas IL-6 had the most significant effect on progenitor cells. In addition, it was determined that fetal liver stem cells were only approximately twofold more "transducible" than stem cells from adult bone marrow. Transplantation of Bodipy-selected bone marrow cells into lethally irradiated mice showed that the number of spleen colony-forming units that were positive for the retroviral vector (as determined by polymerase chain reaction) was 76%, as compared with 32% in animals that were transplanted with cells that were nonselected. The methods described within this manuscript are particularly useful for evaluating hematopoietic stem cell gene transfer in vivo because the marker gene used in the procedure (ASM) encodes a naturally occurring mammalian enzyme that has no known adverse effects, and the fluorescent compound used for selection (Bodipy sphingomyelin) is removed from the cells before transplantation.

The early phase of engraftment after murine blood cell transplantation is mediated by hematopoietic stem cells.

Blood cells transplantation is largely replacing bone marrow transplantation because engraftment is more rapid. This accelerated engraftment is thought to be mediated by relatively mature committed hematopoietic progenitor cells. Herein, we have used a modified rhodamine (Rho) staining procedure to identify and purify Rho+/++ (dull/bright) and Rho- (negative) subpopulations of hematopoietic progenitor cells in murine cytokine-mobilized blood. The Rho+/++ cell population contained > 99% of committed progenitor cells with in vitro colony-forming ability. The Rho- cell population contained the majority of hematopoietic stem cells with in vivo marrow repopulating ability. The rate of hematopoietic reconstitution was identical in recipients of grafts containing only purified Rho- stem cells or purified Rho- stem cells in combination with large numbers of Rho+/++ committed progenitor cells. In contrast, transplantation of 3-fold more hematopoietic stem cells resulted in accelerated reconstitution, indicating that the reconstitution rate was determined by the absolute numbers of Rho- stem cells in the graft. In addition, we observed a 5- to 8-fold reduced frequency of the subset of hematopoietic stem cells with long-term repopulating ability in cytokine-mobilized blood in comparison to steady-state bone marrow. Our results indicate that hematopoietic stem cells and not committed progenitor cells mediate early hematopoietic reconstitution after blood cell transplantation and that relative to bone marrow, the frequency of stem cells with long-term repopulating ability is reduced in mobilized blood.

Evaluation of the silicon phthalocyanine pc 4 for photodynamic bone marrow purging.

The silicon phthalocyanine Pc 4 was tested as a photosensitizer for the selective photoinactivation of malignant cells in bone marrow transplantation samples. Using a murine model system, incubation of 1.5×107 cells/mL with 15 nM Pc 4 followed by exposure to red light (λ>600 nm, fluence of 18 J/cm2) was shown to result in a greater than 6 log10 reduction of the clonogenic growth for the murine cell lines ABE-8.1/2, BC3A and L1210. The clonogenic growth of WEHI-3 and P815 cells was reduced by more than 5 log10 and more than 3 log10, respectively. Late murine hematopoietic progenitor cells were less sensitive than cancer cells; the surviving fractions were 0.084 for the colony forming unit, megakaryocyte (CFU-Mk); 0.038 for the colony forming unit, granulocyte macrophage (CFU-GM); 0.0018 for the colony forming unit, mix (CFU-mix) and <0.003 for burst forming units, erythroid (BFU-E). Early hematopoietic progenitor cells, assayed by the in vitro cobble stone area forming cell assay, were not affected by the photodynamic treatment. Likewise, in vivo assays of early hematopoietic progenitor cells showed no reduction of their ability to repopulate the bone marrow. Irradiation of the samples following incubation of 1.5×106 cells/mL with Pc 4 resulted in increased photosensitivity of all cell types, including the early and late hematopoietic progenitor cells. Flow cytometric analysis of Pc 4 uptake by the cells revealed that the increased photosensitivity could be traced to increased Pc 4 uptake; however, Pc 4 uptake among cell types did not correlate with photosensitivity. When mixed with bone marrow (BM) cells, Pc 4 uptake in the cell lines increased as the fraction of BM increased from 0.5 to 0.95. These observations suggest that Pc 4 may be a suitable photosensitizer for bone marrow purging. © 1998 Society of Photo-Optical Instrumentation Engineers.

Bone marrow transplantation in acid sphingomyelinase-deficient mice: engraftment and cell migration into the brain as a function of radiation, age, and phenotype.

Types A and B Niemann-Pick disease (NPD) result from the deficient activity of the lysosomal hydrolase, acid sphingomyelinase (ASM). A long-term goal of our research is to evaluate the effects of bone marrow transplantation (BMT) and hematopoietic stem cell gene therapy (HSCGT) on the NPD phenotype. As an initial step toward this goal, we have undertaken a study aimed at optimizing hematopoietic cell engraftment in acid sphingomyelinase "knock-out" (ASMKO) mice. Several parameters were analyzed, including the effects of radiation and donor cell number on survival and engraftment of newborn and adult animals, the number of donor cells detected in the brain posttransplantation, and the levels of ASM activity achieved in the brain. A total of 202 ASMKO and normal animals were transplanted and studied, and the overall conclusions were: (1) newborn ASMKO animals were more susceptible to radiation-induced mortality than normal animals, (2) at low radiation doses, increasing the donor cell number improved engraftment, while this was less evident at the higher radiation doses, (3) engraftment was easier to achieve in normal as compared with ASMKO animals, (4) among newborn transplants, the number of donor cells detected in the brain was directly correlated with engraftment in the blood, (5) more donor cells were detected in the brains of newborn ASMKO animals as opposed to newborn normal animals, and (6) no donor cells were found in the brains of animals transplanted as adults, including those that were highly engrafted in the blood. These results provide important information regarding the design of future BMT and HSCGT studies in ASMKO mice and other mouse models and demonstrate the potential of altering the NPD phenotype by these therapeutic strategies.

Bone marrow transplantation in newborn rats with mucopolysaccharidosis type VI: biochemical, pathological, and clinical findings.

BACKGROUND: Mucopolysaccharidosis type VI (MPS VI) is the lysosomal storage disorder caused by the deficient activity of arylsulfatase B (ASB). In this study, we evaluated bone marrow transplantation (BMT) for the treatment of MPS VI and the effects of irradiation on the survival and engraftment of bone marrow-transplanted neonatal rats. METHODS: One- to 2-day-old MPS VI rats were injected with normal bone marrow after irradiation with 200, 400, or 800 cGy. Ninety percent of the animals receiving a single dose of 200 cGy (n=30) survived the procedure, whereas irradiation with 400 cGy (n=23) or 800 cGy (n=12) resulted in significant mortality (78% and 100%, respectively). Engraftment was monitored by determining ASB activities in peripheral white blood cells and by Y chromosome in situ hybridization analysis. Fifty-two percent of the animals from the 200-cGy group engrafted for up to 8 months after BMT; among the five animals that survived the 400-cGy dose, all engrafted. In comparison, only 20% of nonirradiated animals engrafted at low levels. Of the 24 engrafted animals that were monitored for 8 months after BMT, clinical and/or radiographic improvements were noted in only one (BMT animal 3). Enzymatic analysis revealed that the ASB activities in the reticuloendothelial organs of this animal, as well as two other engrafted but clinically unimproved animals (BMT animals 1 and 2), were normal or near normal; correspondingly, the glycosaminoglycan levels in these organs were significantly reduced. Consistent with the clinical and biochemical observations, light and electron microscopic findings were more improved in BMT animal 3 as compared with BMT animals 1 and 2, although a reduction of storage was evident in each of these transplant recipients, particularly in the trachea and aorta, two tissues that are characteristic sites of pathology in human patients. CONCLUSIONS: These results indicate that BMT in newborn MPS VI patients may prevent many of the pathological and clinical findings in this disorder, but is likely to have very limited and unpredictable effects on the skeletal abnormalities.

Purification of repopulating hemopoietic cells based on binding of biotinylated Kit ligand.

To characterize Kit expressing mouse bone marrow (BM) cells, and to determine their contribution to short- and long-term repopulation of the hemopoietic system of irradiated recipients, we have purified Kit+ BM cells by flow cytometry. A high level of Kit expression was detectable on 1-2% of BM cells after staining with biologically active biotinylated Kit ligand (KL) or with anti-Kit antibodies (ACK-2). Compared to unfractionated BM, the Kit+ fractions were enriched for immature hemopoietic cells, as shown by morphological differentiation, in vitro culture, and spleen colony formation. Enrichment of colony-forming cells was higher in biotin-KL+ than ACK-2+ fractions. Colony-forming cells were not found in the Kit- subsets. To study the hemopoietic repopulation capacity of the Kit+ and Kit- cells, serial dilutions of the sorted fractions were transplanted into irradiated mice, and peripheral blood of these recipients was monitored regularly for the presence of donor-derived cells during a 1 year period. Nucleated blood cell repopulation by male donor cells in female recipients was assessed using a Y-chromosome specific DNA probe; erythroid repopulation by normal donor cells in W/Wv recipients was examined flow cytometrically by measuring the forward light scatter of donor- and host-type erythrocytes. A 25- to 100-fold enrichment of long-term repopulating ability in the sorted Kit+ fractions showed that Kit+ cells are capable of reconstitution of circulating erythrocytes and nucleated blood cells after BM transplantation. Transient repopulation of the red blood cell lineage was observed after transplantation of Kit- cells. Detection of donor-derived nucleated cells 1 year after transplantation showed that Kit+ cells contributed to donor-type repopulation of bone marrow, spleen and thymus. Our data demonstrate that isolation of BM cells on the basis of Kit expression is a useful addition to the methods that are commonly applied in stem cell enrichment protocols.

A novel murine cathelin-like protein expressed in bone marrow.

A novel cDNA encoding a putative secreted protein was isolated from murine bone marrow. The encoded protein named MCLP (murine cathelin-like protein) was found to be highly homologous to the pig cathelin, and to four neutrophil antimicrobial polypeptides: CAP 18, indolicidin, Bac 5 and FALL-39. Secondary structure prediction studies identified a highly cationic region in the C-terminal part of prepro-MCLP with a tendency to adopt an amphipathic alpha-helical conformation, as observed in many antimicrobial peptides. However, no antibacterial activity was observed with the synthetic peptide corresponding to this region of MCLP.

Homing of fluorescently labeled murine hematopoietic stem cells.

PKH-26 was used as a viable fluorescent membrane stain for murine hematopoietic stem cells. The presence of the dye on the cells was shown not to interfere with their ability to form day-8 and -12 spleen colonies in lethally irradiated mice. To study their in vivo homing behavior in detail, 10(4) labeled cells from a population enriched for CFU-S were injected intravenously into nonirradiated mice and into mice irradiated 3 hours previously. At 17, 41, and 65 hours after injection, the numbers of labeled cells per organ were quantified using the specialty developed flow cytometric fluorescence hypercompensation procedure for the detection of rare events, which allows a detection sensitivity of 1 per 10(6). Spleen homing in irradiated and nonirradiated mice was virtually identical, whereas homing to nonirradiated bone marrow was 2.5 times higher than to irradiated bone marrow. This indicates a different homing mechanism for spleen and bone marrow. The results of this direct homing assay were placed in perspective with results of indirect homing studies from the literature, introducing a new "h-factor." From the CFU-S data, putative specific enrichment factors for spleen-specific and bone marrow-specific homing were derived. Examination of the fluorescence intensity distribution among the labeled cell population indicated that virtually all cells started to proliferate rapidly after injection into both irradiated and nonirradiated animals. This indicates that specific signals from stromal elements in the stem cell niches are needed to keep the cells quiescent and that the majority of the transplanted stem cells do not home to such niches. The potential use of PKH-26 for in vivo characterization of stem cell niches is discussed.

Genomic structure and alternative splicing of the murine bhk/ctk/ntk gene.

Recently, we and others have cloned cDNAs encoding a second member of the Csk family of inhibitory protein kinases, which we termed Bhk [M.A. Ershler et al. (1994) Dokl. Akad. Nauk. 339, 679-683]. In the present study, two new distinct types of bhk mRNA were found in addition to the third form described previously. Analysis of the bhk genomic structure established that three exons participate in the alternative splicing of bhk mRNA.

Monitoring of leukemia growth in a rat model using a highly sensitive assay for the detection of LacZ marked leukemic cells.

A very sensitive assay for the detection of LacZ marked cells of an in vitro growing subline of the brown Norway rat myelocytic leukemia (BNML) model was developed. By combining cytochemical X-gal staining with D-galactose mediated suppression of endogenous background beta-galactose activity, a detection sensitivity of one leukemic cell per 10(8) normal bone marrow cells could be achieved. A detailed analysis of the in vivo growth pattern and kinetics of this cell line is presented. Also, it is shown that after cyclophosphamide treatment of leukemic rats no leukemic colonies are formed in an agar-colony assay, whereas the leukemic cells remain detectable in the bone marrow for a considerable time period. Eventually, however, all leukemic cells disappear from the marrow. These findings are discussed in the light of prolonged detection of rare leukemic cells in patients in continuing remission.

Modification of rhodamine staining allows identification of hematopoietic stem cells with preferential short-term or long-term bone marrow-repopulating ability.

We have developed a modified rhodamine (Rho) staining procedure to study uptake and efflux in murine hematopoietic stem cells. Distinct populations of Rho++ (bright), Rho+ (dull), and Rho- (negative) cells could be discriminated. Sorted Rho- cells were subjected to a second Rho staining procedure with the P-glycoprotein blocking agent verapamil (VP). Most cells became Rho positive [Rho-/Rho(VP)+ cells] and some remained Rho negative [Rho-/Rho(VP)- cells]. These cell fractions were characterized by their marrow-repopulating ability in a syngeneic, sex-mismatch transplantation model. Short-term repopulating ability was determined by recipient survival for at least 6 weeks after lethal irradiation and transplantation--i.e., radioprotection. Long-term repopulating ability at 6 months after transplantation was measured by fluorescence in situ hybridization with a Y-chromosome-specific probe, by graft function and recipient survival. Marrow-repopulating cells were mainly present in the small Rho- cell fraction. Transplantation of 30 Rho- cells resulted in 50% radioprotection and > 80% donor repopulation in marrow, spleen, and thymus 6 months after transplantation. Cotransplantation of cells from both fractions in individual mice directly showed that within this Rho- cell fraction, the Rho-/Rho(VP)+ cells exhibited mainly short-term and the Rho-/Rho(VP)- cells exhibited mainly long-term repopulating ability. Our results indicate that hematopoietic stem cells have relatively high P-glycoprotein expression and that the cells responsible for long-term repopulating ability can be separated from cells exhibiting short-term repopulating ability, probably by a reduced mitochondrial Rho-binding capacity.

Biotinylation of interleukin-2 (IL-2) for flow cytometric analysis of IL-2 receptor expression. Comparison of different methods.

The main prerequisites for the use of biotinylated ligands to study the expression of growth factor receptors on heterogeneous cell populations, such as peripheral blood or bone marrow, by flow cytometric methods, are that the biotinylated ligand retains its binding ability and that binding of the biotinylated ligand to the receptor does not inhibit the subsequent interaction of biotin with fluorescently tagged avidin or streptavidin. Using interleukin-2 (IL-2), we compared the usefulness of various biotinylation reagents, NHS-biotin, S-NHS-biotin, S-NHS-LC-biotin, DBB and photobiotin, and developed optimal biotinylation conditions for the preparation of biologically active biotin-labeled IL-2 and the detection of IL-2 receptor expressing cells by flow cytometry. As determined by spot blot analysis, biotinylation of IL-2 was most efficient at the highest biotin-to-protein (B:P) ratio used. At a B:P ratio of 100, most of the biological activity of IL-2 was retained when S-NHS-LC-biotin was used. In contrast, most of the biological activity of IL-2 samples that were labeled with NHS-biotin or photobiotin was lost under these conditions. Biotin-labeled IL-2 preparations were tested in order to detect IL-2 receptors on IL-2 dependent CTLL-2 cells by flow cytometry after sequential staining with the biotinylated IL-2 and fluorescence tagged streptavidin. A high B:P ratio generally resulted in a high specific fluorescence intensity of the cells, particularly when S-NHS-LC-biotin was used as the biotinylation reagent. Biotin-IL-2 could also be used to detect IL-2 receptors expressed by lymphocytes in peripheral blood and bone marrow. Comparison of staining of lymphocytes with biotinylated IL-2 and an antibody against the IL-2 receptor alpha chain demonstrated that only a subset of the cells that showed a strong fluorescence signal after staining with biotinylated IL-2 expressed high numbers of the IL-2 receptor alpha chain. This is in agreement with the expression of functional IL-2 receptors on resting T cells and NK cells which do not express the alpha chain. After stimulation with PHA, virtually all lymphocytes expressed the alpha chain, whereas only part of these cells showed a strong fluorescence signal after staining with biotin-IL-2, while the rest of the cells had very low numbers of IL-2 binding sites. Our results demonstrate that, in addition to staining individual receptor subunits with antibodies, staining with biotinylated IL-2 is a useful indicator of functional IL-2 receptor expression.

High-speed photodamage cell selection using a frequency-doubled argon ion laser.

A flow cytometer was developed for the high-speed "sorting" of desired cells by selectively irradiating (zapping) the undesired cells from a population. After previous efforts to photoinactivate cells with photosensitizers had failed, it was decided to exploit the photosensitivity of the cell's DNA at 257 nm. It was shown that a 257 nm laser output power of 20-100 mW was sufficient to induce a 4.5 log cell kill after the cells were processed through a focused 257 nm laser beam. Experiments proved that the photodamage flow cytometer (ZAPPER) could selectively photoinactivate cells at rates over 22,000 events/s, and selection purities ranged from 81% to 100%. The yields of the desired cells depended on the selection mode. In the Enrichment mode, the zap laser was not aimed at the jet, and only undesired cells were exposed to a brief ultraviolet (UV) pulse after modulation of the UV laser beam. The yields of desired cells ranged from 95% to 105%. In the Purge mode, the zap laser beam was aimed onto the jet, and only desired cells were allowed to pass after deflection of the UV laser beam; the yields of desired cells ranged from 12% to 52%. The cause of the reduced yields in the PURGE mode was traced to the fact that the Electro-Optic Modulator was used to modulate the zap laser proved too slow for the intended application. The lifetime of the frequency-doubling crystal used for the generation of the 257 nm beam was found to be limited to several days.(ABSTRACT TRUNCATED AT 250 WORDS)

Classmode: a new data format for real-time multiparameter data analysis and data compression.

A new data acquisition and analysis format (classmode) was developed that allows real-time data classification in a flow cytometer. In our cytometer, detected events were classified in real time by their presence or absence in a set of look-up tables (LUT). A modification of the cytometer hardware allows the exclusive transfer of the LUT data to the acquisition/storage computer. Using a combination of 8 LUTs, the analyzed events can be classified into 256 subpopulations. Real-time data classification results in an increased data transfer rate and a significant compression of the data.

Differential suppression of background mammalian lysosomal beta-galactosidase increases the detection sensitivity of LacZ-marked leukemic cells.

A method is described for the detection of Escherichia coli beta-galactosidase-expressing leukemic cells in ex vivo bone marrow samples. 4-Methylumbelliferyl-beta-D-galactopyranoside is used as a substrate in a kinetic assay. D-Galactose is used to suppress endogenous lysosomal beta-galactosidase activity, yielding a sixfold increase in sensitivity. With this assay, the detection limit is one leukemic cell per 10(4) normal bone marrow cells.

The expression of cytokine receptors by purified hemopoietic stem cells.

Sorted fractions from mouse bone marrow containing highly purified hemopoietic stem and progenitor cells were studied for the expression of growth factor receptors. With the use of rhodamine 123 WGA+, 15-1.1-, low density cells were separated into quiescent pluripotent stem cells and committed progenitor cells. RNA was extracted and cDNA was prepared by reverse transcription. Using primers specific for growth factor receptors, the cDNA of each sorted fraction was amplified by polymerase chain reaction (PCR). The quiescent rhodamine 123 dull stem cell fraction was found to express the interleukin 3 (IL-3) receptor beta unit and c-kit, but not the granulocyte-macrophage colony stimulating factor (GM-CSF) receptor beta unit nor flk-2. The rhodamine 123 bright fraction with activated stem cells and mostly committed progenitor cells similarly expressed the IL-3-R beta, and c-kit. However, this fraction also expressed flk-2 and GM-CSF-R beta. Since the expression of c-kit in the stem cell fraction does not correspond with the poor response to the kit-ligand stem cell factor (SCF) by these cells, we further analyzed the fractions with respect to their binding of biotinylated SCF. The SCF-binding cells were found to be all rhodamine 123 bright. This indicates that the expression of c-kit is not sufficient to yield a functional receptor for SCF; c-kit probably needs a partner molecule to form a functional high-affinity binding site for SCF. Similar to the beta unit of the GM-CSF receptor, this partner is then not expressed in the stem cell fraction.(ABSTRACT TRUNCATED AT 250 WORDS)