Chaetocin antileukemia activity against chronic myelogenous leukemia cells is potentiated by bone marrow stromal factors and overcomes innate imatinib resistance.

Chronic myelogenous leukemia (CML) is maintained by a minor population of leukemic stem cells (LSCs) that exhibit innate resistance to tyrosine kinase inhibitors (TKIs) targeting BCR-ABL. Innate resistance can be induced by secreted bone marrow stromal cytokines and growth factors (BMSFs) that protect CML-LSCs from TKIs, resulting in minimal residual disease. Developing strategies to eradicate innate TKI resistance in LSCs is critical for preventing disease relapse. Cancer cells balance reactive oxygen species (ROS) at higher than normal levels, promoting their proliferation and survival, but also making them susceptible to damage by ROS-generating agents. Bcr-Abl increases cellular ROS levels, which can be reduced with TKI inhibitors, whereas, BMSFs increase ROS levels. We hypothesized that BMSF-mediated increases in ROS would trigger ROS damage in TKI-treated CML-LSCs when exposed to chaetocin, a mycotoxin that imposes oxidative stress by inhibiting thioredoxin reductase-1. Here, we showed that chaetocin suppressed viability and colony formation, and induced apoptosis of the murine hematopoietic cell line TonB210 with and without Bcr-Abl expression, and these effects were potentiated by BMSFs. In contrast, imatinib activities in Bcr-Abl-positive TonB210 cells were inhibited by BMSFs. Further, BMSFs did not inhibit imatinib activities when TonB210 cells expressing Bcr-Abl were cotreated with chaetocin. Chaetocin showed similar activities against LSC-enriched CML cell populations isolated from a murine transplant model of CML blast crisis that were phenotypically negative for lineage markers and positive for Sca-1 and c-Kit (CML-LSK). BMSFs and chaetocin increased ROS in CML-LSK cells and addition of BMSFs and chaetocin resulted in higher levels compared with chaetocin or BMSF treatment alone. Pretreatment of CML-LSKs with the antioxidant N-acetylcysteine blocked chaetocin cytotoxicity, even in the presence of BMSFs, demonstrating the importance ROS for chaetocin activities. Chaetocin effects on self-renewal of CML-LSKs were assessed by transplanting CML-LSKs into secondary recipients following ex vivo exposure to chaetocin, in the presence or absence of BMSFs. Disease latency in mice transplanted with CML-LSKs following chaetocin treatment more than doubled compared with untreated CML-LSKs or BMSFs-treated CML-LSKs. Mice transplanted with CML-LSKs following chaetocin treatment in the presence of BMSFs had significantly extended survival time compared with mice transplanted with CML-LSKs treated with chaetocin alone. Our findings indicate that chaetocin activity against CML-LSKs is significantly enhanced in the presence of BMSFs and suggest that chaetocin may be effective as a codrug to complement TKIs in CML treatment by disrupting the innate resistance of CML-LSKs through an ROS dependent mechanism.

Yeast two-hybrid screening of cyclic peptide libraries using a combination of random and PI-deconvolution pooling strategies.

We developed a high throughput yeast two-hybrid (Y2H) assay for screening pools of combinatorial cyclic peptide preys against pools of bait proteins. The assay used the PI (pooling with imaginary tags) deconvolution pooling strategy to generate pools of baits and a random pooling strategy to generate pools of cyclic peptide preys. Haploid yeast, expressing pools of baits or preys, were arrayed and mated to each other to generate diploid arrays, where the yeast express both baits and preys. Diploid arrays were scored for activation of the Y2H reporter genes. We used this Y2H pooling strategy, referred to as 'PI-pool-on-random pool', to screen a cyclic peptide library for interactions against Bcr-Abl domains. Seven Bcr-Abl domain baits and LexA control were pooled using the PI deconvolution pooling strategy. The cyclic peptide library was randomly arrayed into pools of ~1000 members. Cyclic peptides were isolated for six of seven Bcr-Abl domain baits. The PI-pool-on-random pooling Y2H assay using high stringency Y2H reporter genes produced no false positives, while missing 20% of real interactions. The high specificity of the PI-pool-on-random pooling Y2H assay eliminates the need to validate interactions. Pooling of baits and preys allows large prey libraries to be screened against multiple baits and takes advantage of PI-deconvolution to determine protein interactions with high sensitivity and specificity. The scalability of this assay allows the peptide preys to be isolated in a high throughput manner against a large number of baits and provides an avenue for generating affinity agents against entire proteomes in the future.

Reexpression of epigenetically silenced AML tumor suppressor genes by SUV39H1 inhibition.

Reexpression of hypermethylated tumor suppressor genes using DNA methyltransferase (DNMT) and histone deacetylase inhibitors occurs by a mechanism whereby promoter demethylation is the dominant event. In support of this model, we found in acute myeloid leukemia cells with hypermethylated p15INK4B and E-cadherin promoters that the DNMT inhibitor, 5-aza-2'-deoxycytidine, induced p15INK4B and E-cadherin expression, and decreased levels of DNA methylation, histone H3 lysine 9 (H3K9) methylation and SUV39H1 associated with p15INK4B and E-cadherin promoters. On the basis of these observations, we examined whether promoter demethylation was dominant to H3K9 demethylation in p15INK4B and E-cadherin reexpression. We observed that SUV39H1 short hairpin RNA and chaetocin, a SUV39H1 inhibitor, induced p15INK4B and E-cadherin expression and H3K9 demethylation without promoter demethylation. Reexpression of hypermethylated p15INK4B and E-cadherin required histone H3K9 demethylation that was achieved directly by inhibiting SUV39H1 expression or activity, or indirectly by decreasing the amount of SUV39H1 associated with the p15INK4B and E-cadherin promoters using 5-aza-2'-deoxycytidine. The results from this study highlight the potential of H3K9 methyltransferases as therapeutic targets for reactivating expression of hypermethylated genes.

Bcr-Abl induces autocrine IGF-1 signaling.

Bcr-Abl oncogene is responsible for the initial phase of chronic myelogenous leukemia (CML), which is effectively treated by the Bcr-Abl inhibitor imatinib. Over time patients become resistant to treatment and progress to blast crisis, an event that is driven by additional genetic and epigenetic aberrations. Recently, we showed that Riz1 expression decreases in blast crisis and that re-expression of Riz1 inhibits IGF-1 expression. IGF-1 signaling is required in many stages of hematopoiesis and inappropriate activation of autocrine IGF-1 signaling may facilitate transformation to blast crisis. We observed that in 8 out of 11 matched CML patient biopsies the IGF-1 expression is elevated in blast crisis. We examined mechanisms used by CML blast crisis cell lines to activate IGF-1 expression. We found that Bcr-Abl activates autocrine IGF-1 signaling using Hck and Stat5b. Inhibition of these signaling components using small molecule drugs or shRNA decreases proliferation and enhances apoptosis. Together, our study suggests that aberrant IGF-1 signaling is an important event in blast crisis transformation and it provides a mechanism to explain the activity of IGF-1R and Hck inhibitors in blocking CML blast crisis phenotypes.

RIZ1 repression is associated with insulin-like growth factor-1 signaling activation in chronic myeloid leukemia cell lines.

RIZ1 is a histone methyltransferase whose expression and activity are reduced in many cancers. In chronic myelogenous leukemia (CML), blastic transformation is associated with loss of heterozygosity in the region where RIZ1 is located and with decreased RIZ1 expression. Forced RIZ1 expression in model CML blast crisis (BC) cell lines decreases proliferation, increases apoptosis and enhances differentiation. We characterized molecular mechanisms that may contribute to potential CML tumor suppressor properties of RIZ1. Several RIZ1-regulated genes involved in insulin-like growth factor-1 (IGF-1) signaling were identified using cDNA microarrays. RIZ1 was shown to associate with promoter regions of IGF-1 and to increase histone H3 lysine 9 methylation using chromatin immunoprecipitation assays. IGF-1-blocking antibody was used to demonstrate the importance of autocrine IGF-1 signaling in CML-BC cell line viability. Forced RIZ1 expression in CML-BC cell lines decreases IGF-1 receptor activation and activation of downstream signaling components extracellular signal-regulated kinase 1/2 and AKT. These results highlight the therapeutic potential of inhibiting IGF-1 pathway in the acute phase of CML.

Peptide aptamers: dominant "genetic" agents for forward and reverse analysis of cellular processes.

Peptide aptamers are the newest in the class of "genetic" agents that aid in the analysis of cellular processes. Peptide aptamers interact with and can inactivate gene products, but don't mutate the DNA that encodes them. Reverse analysis with peptide aptamers involves isolating aptamers that interact with a specific protein and monitoring the resulting aptamer-induced phenotype. Conversely, forward analysis with peptide aptamers involves expressing combinatorial libraries of aptamers within an organism and screening for aptamer-induced variations in their phenotypes. The two-hybrid system is used in both processes to help identify the protein interactions, and variations from the basic procedure as described in UNIT are presented here.

Use of intrinsic binding energy for catalysis by a cofactor-independent DNA enzyme.

The concept of the Circe effect, according to which an enzyme's substrate-binding energy is utilized to destabilize the substrate towards the reaction transition state, has been shown to be a relevant catalytic strategy for naturally occurring protein enzymes and for two ribozymes that use nucleotide-based substrates and metal ion cofactors. We wished to investigate whether such a catalytic strategy extends even to divergent and unevolved catalysts constructed from biopolymers. We examined the properties of a small, in vitro selected, and cofactor-independent DNA enzyme, PS5.M, which catalyzes porphyrin metallation. The metallation reaction is unique, in that the energies for binding and for metallation of both the substrate and of a transition-state analogue (TSA) can be measured. We report that PS5.M, originally selected for binding to the TSA, displays the Circe effect in channeling a significant component of entropy-rich "intrinsic" binding energy to distort and to alter the basicity of the bound substrate. The study demonstrates that nucleic acids are, by themselves, capable of creating active sites for the catalysis of chemical reactions involving non-nucleotide substrates. Furthermore, the study of the metallation of the TSA provides a quantitative estimate of the effectiveness of such a compound in mimicking the true transition state for porphyrin metallation.

"Mutagenesis" by peptide aptamers identifies genetic network members and pathway connections.

We selected peptide aptamers from combinatorial libraries that disrupted cell-cycle arrest caused by mating pheromone in yeast. We used these aptamers as baits in two-hybrid hunts to identify genes involved in cell-cycle arrest. These experiments identified genes known to function in the pathway, as well as a protein kinase, the CBK1 product, whose function was not known. We used a modified two-hybrid system to identify specific interactions disrupted by these aptamers. These experiments demonstrate a means to perform "genetics" on the protein complement of a cell without altering its genetic material. Peptide aptamers can be identified that disrupt a process. These aptamers can then be used as affinity reagents to identify individual proteins and protein interactions needed for the process. Forward genetic analysis with peptide aptamer "mutagens" should be particularly useful in elucidating genetic networks in organisms and processes for which classical genetics is not feasible.

Lanthanide probes for a phosphodiester-cleaving, lead-dependent, DNAzyme.

Lanthanides are useful probes for studying metal ion interactions in biological systems. The trivalent cations of the lanthanide metals are unique in that their ionic radii and the first pKa values of bound water molecules vary monotonically along the period. In addition, the europium and terbium cations have the useful property that their luminescence is enhanced when bound to nucleic acids. We have found that lanthanide ions can function as effective co-factors for a lead-dependent, phosphodiester-cleaving catalytic DNA (DNAzyme). We used the unique properties of the lanthanide co-factors to study the metal binding site as well as the catalytic mechanism of the DNAzyme. The catalyzed lanthanide-mediated cleavage occurred at neutral pH and at room temperature, with multiple turnovers of substrate. A range of lanthanide ions could act as co-factors, but differentially, with the smaller lanthanides (Tb, Tm, Lu) being the most effective. The rate of cleavage of the phosphodiester did not vary linearly with either the ionic radius or the first pKa of lanthanide-coordinated water molecules. The DNAzyme appeared to use only a single bound lanthanide ion as co-factor. Luminescence spectroscopy with terbium revealed the importance of the 2' hydroxyl group at the cleavage site in lanthanide ion binding, and the substrate molecule alone appeared to generate substantially the catalytically relevant metal-binding site. This model system demonstrated further the utility of complexing lanthanide ions directly to DNA molecules for catalytic purposes. The use of lanthanide ions also provides a means for investigating the metal ion binding sites of nucleic acid enzymes in general.

DNA enzymes.

The past year has seen a coming-of-age in DNA enzyme research. Far from being laboratory curiosities, the activities of new DNA enzymes have broadened the known catalytic repertoire of nucleic acid enzymes, provided valuable insights into different mechanistic possibilities open to nucleic acid catalysts, and explored the importance for catalysis of native functionalities within DNA and RNA, as well as of a diversity of extrinsic cofactors. Thus, the first amino acid cofactor-utilizing DNA enzyme has been described, as well as DNA enzymes that cleave RNA without the assistance of any external cofactor. On the practical side, the most efficient RNA-cleaving nucleic acid enzyme described to date is a DNA enzyme.

Evidence for the metal-cofactor independence of an RNA phosphodiester-cleaving DNA enzyme.

BACKGROUND: RNA and DNA are polymers that lack the diversity of chemical functionalities that make proteins so suited to biological catalysis. All naturally occurring ribozymes (RNA catalysts) that catalyze the formation, transfer and hydrolysis of phosphodiesters require metal-ion cofactors for their catalytic activity. We wished to investigate whether, and to what extent, DNA molecules could catalyze the cleavage (by either hydrolysis or transesterification) of a ribonucleotide phosphodiester in the absence of divalent or higher-valent metal ions or, indeed, any other cofactors. RESULTS: We performed in vitro selection and amplification experiments on a library of random-sequence DNA that incorporated a single ribonucleotide, a suitable site for cleavage. Following 12 cycles of selection and amplification, a 'first generation' of DNA enzymes (DNAzymes) cleaved their internal ribonucleotide phosphodiesters at rates approximately 10(7)-fold faster than the spontaneous rate of cleavage of the dinucleotide ApA in the absence of divalent cations. Re-selection from a partially randomized DNA pool yielded 'second generation' DNAzymes that self-cleaved at rates of approximately 0.01 min-1 (a 10(8)-fold rate enhancement over the cleavage rate of ApA). The properties of these selected catalysts were different in key respects from those of metal-utilizing ribozymes and DNAzymes. The catalyzed cleavage took place in the presence of different chelators and ribonuclease inhibitors. Trace-metal analysis of the reaction buffer (containing very high purity reagents) by inductively coupled plasma-optical emission spectrophotometry indicated that divalent or higher-valent metal ions do not mediate catalysis by the DNAzymes. CONCLUSIONS: Our results indicate that, although ribozymes are sometimes regarded generically to be metalloenzymes, the nucleic acid components of ribozymes may play a substantial role in the overall catalysis. Given that metal cofactors increase the rate of catalysis by ribozymes only approximately 10(2)-10(3)-fold above that of the DNAzyme described in this paper, it is conceivable that substrate positioning, transition-state stabilization or general acid/base catalysis by the nucleic acid components of ribozymes and DNAzymes may contribute significantly to their overall catalytic performance.

Recognition of anionic porphyrins by DNA aptamers.

DNA sequences were isolated by in vitro selection for binding to N-methylmesoporphyrin IX (NMM), a molecule that behaves as a stable transition-state analogue for porphyrin chelatases. Clones approximately 280 and approximately 120 nucleotides long were obtained, which bound to NMM with sub-micromolar affinity but bound mesoporphyrin IX (MPIX), as well as various metalloderivatives of MPIX, with lower affinity. Footprinting experiments with dimethyl sulfate, DNase I, and bound hemin molecules activated by superoxide identified a series of short guanine-rich motifs to be the binding sites for the various porphyrins. One clone, PS2, examined in depth, gave a methylation footprint with bound NMM but not with bound MPIX nor with a number of metalloporphyrins. The binding domain PS2, synthesized as a short oligonucleotide, itself showed high-affinity binding to NMM. The binding sequences from different clones were loosely homologous, and the footprinting data were consistent with their folding to form one or more guanine quartets in the presence of NMM. Ultraviolet--visible absorption and circular dichroism spectroscopy of the DNA--NMM complexes indicates, however, that the interaction is not primarily intercalative in nature. The preferential binding of NMM by these aptamers raises the possibility of their being able to catalyze the chelation of metal ions by the porphyrin MPIX.

Promoting the quality of life in nursing homes: can regulation succeed?

It has been assumed that the best policy for promoting quality of life in nursing homes is direct regulation. In this paper it is argued that if our experience in regulating quality of care is any indication, we may not possess the political will to successfully regulate quality of life. Moreover, from a legal perspective, the less concrete nature of the concept of quality of life may make it more difficult to regulate than quality of care. Finally, although regulation would probably be necessary if potential nursing home residents (and their agents) lacked the information or rationality to make choices that promoted their interests, this has never been shown to be the case empirically. Therefore, we may not be forced to choose regulation to achieve an adequate quality of life. Alternative--and perhaps better--policies may be available.