HIF-1-dependent expression of angiopoietin-like 4 and L1CAM mediates vascular metastasis of hypoxic breast cancer cells to the lungs.

Most cases of breast cancer (BrCa) mortality are due to vascular metastasis. BrCa cells must intravasate through endothelial cells (ECs) to enter a blood vessel in the primary tumor and then adhere to ECs and extravasate at the metastatic site. In this study we demonstrate that inhibition of hypoxia-inducible factor (HIF) activity in BrCa cells by RNA interference or digoxin treatment inhibits primary tumor growth and also inhibits the metastasis of BrCa cells to the lungs by blocking the expression of angiopoietin-like 4 (ANGPTL4) and L1 cell adhesion molecule (L1CAM). ANGPTL4 is a secreted factor that inhibits EC-EC interaction, whereas L1CAM increases the adherence of BrCa cells to ECs. Interference with HIF, ANGPTL4 or L1CAM expression inhibits vascular metastasis of BrCa cells to the lungs.

Hypoxia-induced human endonuclease G expression suppresses tumor growth in a xenograft model.

We have developed a hypoxia-inducible gene therapy approach for the expression of the mature form of human endonuclease G to facilitate cell death in hypoxic regions of the tumor. The chimeric therapeutic gene is placed under the control of a hypoxia response element based promoter and contains a translocation motif linked in frame to an oxygen-dependent degradation domain and the endonuclease G gene. Transient expression of the chimeric therapeutic gene in breast and prostate cancer cell lines resulted in efficient cell death under hypoxia-mimetic conditions. Stable MDA-MB-435 cells expressing the chimeric therapeutic gene under 1% O2 showed an increase in stable HIF-1alpha protein levels and synthesis of the endonuclease G protein in a time-dependent manner. In normoxic conditions, these stable transgenic cells exhibited no change in growth rate, invasion and motility when compared to parental cells. Moreover, xenografts generated using the transgenic cells exhibited highly significant suppression of tumor growth in a preclinical cancer model compared to the parental cell line. Thus, the hypoxia-modulated endonuclease G expression has the potential to be used as a gene-based-therapy system to kill malignant cells within hypoxic regions of tumors.

Oncogenic role of DDX3 in breast cancer biogenesis.

Benzo[a]pyrene diol epoxide (BPDE), the active metabolite of benzo[a]pyrene present in tobacco smoke, is a major cancer-causing compound. To evaluate the effects of BPDE on human breast epithelial cells, we exposed an immortalized human breast cell line, MCF 10A, to BPDE and characterized the gene expression pattern. Of the differential genes expressed, we found consistent activation of DDX3, a member of the DEAD box RNA helicase family. Overexpression of DDX3 in MCF 10A cells induced an epithelial-mesenchymal-like transformation, exhibited increased motility and invasive properties, and formed colonies in soft-agar assays. Besides the altered phenotype, MCF 10A-DDX3 cells repressed E-cadherin expression as demonstrated by both immunoblots and by E-cadherin promoter-reporter assays. In addition, an in vivo association of DDX3 and the E-cadherin promoter was demonstrated by chromatin immunoprecipitation assays. Collectively, these results demonstrate that the activation of DDX3 by BPDE, can promote growth, proliferation and neoplastic transformation of breast epithelial cells.

Collagen gel delivery of Tgf-beta3 non-viral plasmid DNA in rat osteoblast and calvarial culture.

Different forms of collagen as a carrier for naked plasmid DNA have shown potential as vehicles for therapeutic gene delivery and tissue engineering. The objective of this study was to determine the suitability of a dense collagen gel as a vehicle for sustained delivery of plasmid DNA in cell and organ culture. Plasmid DNA encoding Tgf-beta(3) was combined with collagen gel. DNA released into the media was measured by Pico-Green spectrophotometry. Results showed that DNA was released from the collagen gel at a gradual rate for up to 14 days. To evaluate collagen-mediated transfection in tissue, calvariae were exposed to collagen containing plasmid encoding GFP or DsRed. Transfection was visualized by fluorescence localized to tissue adjacent to the vehicle. To evaluate protein production, fetal rat calvarial osteoblasts were cultured with a collagen/Tgf-beta(3) plasmid mixture or in media containing plasmid alone. Media was collected at various time points to measure Tgf-beta(3) protein production. ELISA assays showed that collagen-transfected osteoblasts demonstrated an elevated Tgf-beta(3) protein production for up to 14 days. Therefore, collagen delivery of viable plasmid DNA created a sustained transient transfection of calvarial osteoblasts resulting in prolonged and elevated growth factor production. Together, these results suggest that use of collagen gel as a vehicle may provide a strategy to achieve localized and controlled, non-viral gene delivery in vivo.

Pretargeting of bacterial endocarditis in rats with streptavidin and 111In-labeled biotin.

UNLABELLED: A radioimaging approach for the detection of endocarditis has been investigated using two-step pretargeting with streptavidin and radiolabeled biotin. METHODS: Hemodynamic alterations within the rat heart were induced by placing an in-dwelling catheter into the left ventricle through the aortic valves. The animals were subsequently infected with Staphylococcus aureus through a tail vein. After an incubation period, rats were first injected with streptavidin and, 2 h later, with 111In-labeled ethylene-diaminetetraacetic acid-biotin. Whole-body gamma camera images were taken 4-5 h postinjection of the radiolabeled biotin. Control animals consisted of catheterized but uninfected, infected but uncatheterized and normal untreated rats. As a further control, the labeled biotin was administered to a study animal without the preadministration of streptavidin. RESULTS: Histology showed typical endocarditic changes in the hearts of study animals with massive deposition of gram-positive cocci. Catheterized but uninfected animals showed alterations corresponding to nonbacterial thrombotic endocarditis. Macroautoradiography showed accumulation of radiolabel in the endocarditic vegetations of study animals. Whole-body gamma camera images showed important cardiac uptake in 7 of 8 catheterized and infected animals and in 3 of 6 catheterized but uninfected animals. Normal rats and those infected but not catheterized showed negative results by histology, autoradiography and imaging. The percent uptake of the injected dose in the heart was 0.20 (SD = 0.13) in catheterized and infected animals, 0.12 (SD = 0.10) in catheterized but uninfected animals, 0.10 (SD = 0.04) in infected but uncatheterized animals and 0.04 (SD = 0.01) in normal control animals. CONCLUSION: The two-step pretargeting approach using streptavidin and 111In-labeled biotin was used successfully to detect S. aureus-induced bacterial endocarditis in rats.

Preparation and use of NHS-MAG3 for technetium-99m labeling of DNA.

The chelator mercaptoacetylglycylglycylglycine (MAG3) is on of several amidothiols that have been used successfully to radiolabeled proteins and other molecules with 99mTc. Prior to radiolabeling, the sulfur in these amidothiols is usually protected by a benzoyl group (i.e. S-benzoyl MAG3) which requires extreme alkaline pH or boiling water temperatures for rapid deprotection. As a result, the benzoyl-protected chelator is radiolabeled prior to conjugation (i.e. preconjugation labeling) in the case of carriers such as proteins or polypeptides which cannot withstand harsh conditions. We have employed a simple, two-step, synthesis of the N-hydroxysuccinimide ester of MAG3 in which the sulfur is protected with an acetyl group (i.e. S-acetyl NHS-MAG3). A single-stranded amine-derivitized DNA was coupled with NHS-S-acetyl MAG3. Radiolabeling was accomplished at room temperature and neutral pH by transchelation from 99mTc-tartrate. In comparison to labeled SHNH-DNA, the labeled MAG3-DNA was unstable to cysteine transchelation, however, in contrast to SHNH-DNA, no evidence for serum protein binding of the labeled MAG3-DNA was observed. We conclude that the S-acetyl NHS MAG3 bifunctional chelator may prove to be an attractive alternative method of radiolabeling DNA and other biologically important molecules with 99mTc.

Investigations of directly labeling antibodies with rhenium-188.

Methods for labeling antibodies with 99mTc cannot be used without modification for radiorhenium despite the similar chemistries, in part because of a lower redox potential of rhenium and therefore a greater tendency to reoxidize. We have investigated conditions for directly labeling B72.3 IgG with 188Re via both mercaptoethanol and stannous ion antibody reduction. The reduced 188Re was stabilized for transchelation as the glucoheptonate complex and transchelated in the presence of excess stannous ion. End points were low "non-specific" binding (i.e. labeling in the absence of antibody reduction) and increased stability to cysteine challenge. By both methods, labeling efficiencies after about 15 minutes averaged 58.77% with as little as 4% non-specific binding. Specific activities of 15 muCi/microgram was achieved after 1.5 hours. By investigating labeling condition, it was possible to improve the stability of the label on stannous ion reduced antibody such that the in vitro and in vivo properties of 188Re were largely independent of labeling method. For example, losses of 188Re due to oxidation (16%) and to cysteine (7%) during 37 degrees C serum incubations for 24 hours were identical for both methods. Furthermore, after the administration to normal mice, whole body clearance and the accumulations of 188Re at 2.5 and 24 hours in blood and in most organs were also independent of labeling method. In conclusion, two different direct labeling methods provided a 188Re-labeled antibody with identical stabilities and with in vivo properties not greatly different from that seen for the same antibody radiolabeled directly with 99mTc.

Comparative properties of a technetium-99m-labeled single-stranded natural DNA and a phosphorothioate derivative in vitro and in mice.

Oligonucleotides, particularly single stranded, may ultimately be of considerable use as radiopharmaceuticals. We have compared a synthetic 22-base single-stranded phosphodiester DNA with its phosphorothioate analog after both were radiolabeled with 99mTc via the hydrazino nicotinamide chelator. Whole body clearance of the label in mice was much slower when introduced on the phosphorothioate (30% vs. 75% clearance at 6 hr) because of immediate and persistent accumulation in liver (47% vs. 2% injected dose/g at 4 hr). The label in both cases was present in urine primarily on low molecular weight catabolites. High-performance liquid chromatography analysis of 37 degrees C serum incubates showed serum protein binding of 99mTc in both cases (about 100% bound at 24 hr) but to different proteins. Different behavior with respect to protein binding was also observed in the analysis of liver and kidney homogenates: the phosphodiester label was almost quantitatively converted to lower molecular weight catabolites after only 15 min, whereas the phosphorothioate label was primarily on proteins. The rapid digestion of the phosphodiester by nucleases was not observed, probably because protein binding of the labeled oligonucleotides stabilized against degradation. Thus the phosphodiester DNA may be the preferred 99mTc-labeled oligonucleotide in certain circumstances to avoid the high and persistent liver uptake observed with the phosphorothioate DNA.

Protein labelling via deoxyribonucleic acid hybridization.

A novel method of radiolabelling antibodies and other proteins is described. A small single-stranded DNA was covalently conjugated to an antibody and labelled by hybridization following the addition of the complementary single-stranded DNA labelled with technetium-99m (99Tc(m)) or indium-111 (111In). Antibody labelling efficiencies were 100% in about 1 h at room temperature with specific activities of up to 30 microCi micrograms-1 of IgG for 99Tc(m). Both diester and thioate DNAs were used. Both the diester- and thioate-labelled antibodies showed complete label stability in 37 degrees C saline. After 24 h in 37 degrees C serum, however, about 40% of the label in the case of the diester antibody was on low molecular weight species--probably labelled catabolites from nuclease degradation of the phosphodiester DNA. In contrast, the 99Tc(m) label on the thioate antibody was immediately and quantitatively bound to serum proteins--probably due to non-specific binding through the sulphur groups. Biodistribution studies in normal mice reflect these in vitro observations: 99Tc(m) on the diester antibody was rapidly cleared through the kidneys, probably as low molecular weight catabolite, while on the thioate antibody, the 99Tc(m) label was predominately deposited in the liver. In conclusion, by modifying with a single-stranded DNA, proteins may be readily labelled with a variety of radionuclides by DNA hybridization. The properties of the radiolabel are strongly influenced by the nature of the DNA.

Technetium-99m labeling of DNA oligonucleotides.

UNLABELLED: Single-stranded RNA and DNA oligonucleotides may be useful as radiopharmaceuticals for antisense and other in vivo applications if convenient methods for stably attaching radionuclides such as 99mTc can be developed. METHODS: To radiolabel DNA with 99mTc, we have used the hydrazino nicotinamide (SHNH) moiety developed elsewhere. The diethylenetriaminepentacetic acid (DTPA) chelate was used to label DNA with 111In for comparison. Complementary 22-base, single-stranded oligonucleotides were obtained, each with a primary amine attached to either 3' or 5' end with a biotin moiety on the opposite end. The DNA was conjugated with SHNH by a N-hydroxysuccinimide derivative with DTPA by the cyclic anhydride. RESULTS: Reversed-phase HPLC analysis showed that essentially complete conjugation was achieved in both cases. The purified SHNH-DNA was radiolabeled with 99mTc by transchelation from glucoheptonate at labeling efficiencies of up to 60% and DTPA-DNA with 111In acetate at up to 100% efficiency. After labeling, the ability of the DNAs to bind to streptavidin through the biotin moieties and to hybridize with their complementary DNA in saline was retained for both radiolabels as determined by size-exclusion HPLC analysis. HPLC radiochromatograms of serum incubates showed a shift to 99mTc, but not 111In, to a high molecular weight, strongly suggesting serum protein binding in the former case only. Low-molecular weight degradation products were seen with 111In, but not with 99mTc and may be related to the use of phosphodiester-linked oligonucleotides. As a further measure of label stability, the DNAS were bound to streptavidin-conjugated magnetic beads and incubated in fresh 37 degrees C human serum. Less than 4% of 99mTc and 14% of 111In was lost in 24 hr. CONCLUSION: Amino-modified, single-stranded DNA can be stably radiolabeled with 99mTc by the SHNH moiety without loss of function.

Recombinant metallothionein-conjugated streptavidin labeled with 188Re and 99mTc.

Consideration is now being given to the use of avidin (or streptavidin) and biotin for radiotherapy of tumor. Accordingly, the goal of this study was to radiolabel a mouse metallothionein-streptavidin fusion protein with 188Re and to compare its properties to those of the same fusion protein radiolabeled with 99mTc. A recombinant metallothionein-streptavidin fusion protein was radiolabeled by transchelation with 99mTc- and 188Re-glucoheptonate. Labeling efficiency, which was not optimized for either radionuclide, was approximately 60% for 99mTc and 20% for 188Re. Radiochemical purity was demonstrated by size exclusion HPLC both by nearly quantitative shifts of the 188Re label to higher molecular weight upon the addition of biotinylated antibody and by the absence of a shift with biotinsaturated 188Re-metallothionein-streptavidin. Stability of the labels in 37 degrees C serum was evaluated by comparing the HPLC radiochromatograms of serum samples both before and after the addition of biotinylated antibody. The 188Re label behaved like 99mTc in that the same peaks were evident, including one prominent peak due to labeled cysteine. Recoveries during HPLC analysis of serum samples showed that oxidation rates to perrhenate and pertechnetate were identical. However, instability to cysteine challenge was greater for 188Re; for example, the loss of label to cysteine after 24 h under one set of conditions was 41% for 188Re and 22% with 99mTc. Analysis by HPLC of liver and kidney homogenates from mice administered the labeled antibodies were qualitatively and, in large measure, quantitatively independent of label. Biodistributions at 5 h in normal mice were statistically identical between the two labels in blood and in most tissues. In conclusion, streptavidin may be radiolabeled with radiorhenium using recombinant mouse metallothionein as a bifunctional chelator, and under one set of labeling conditions at least, 188Re showed similar in vitro and in vivo behavior to that of 99mTc labeled to the same fusion protein.

Can a cysteine challenge assay predict the in vivo behavior of 99mTc-labeled antibodies?

Recent investigations have shown that transchelation to cysteine in a principal mode of in vivo instability of 99mTc-labeled antibodies. In this investigation, a cysteine challenge assay was used to measure the in vitro instability of 99mTc directly labeled to two IgG antibodies (B72.3 and C110) via two established direct labeling methods employing mercaptoethanol and stannous ion for antibody reduction and by a novel method using glutathione for this purpose. For both antibodies, the greatest instability to cysteine occurred with stannous ion reduction. The stability of glutathione-reduced B72.3 was indistinguishable from mercaptoethanol-reduced B72.3 whereas glutathione-reduced C110 showed stability roughly intermediate between that of the other reducing agents for this antibody. Results obtained in normal mice were in the direction predicted by the assay: for both antibodies, urinary clearance of 99mTc was fastest in mice receiving antibodies labeled via stannous ion reduction, presumably because of the increased transchelation of label to cysteine in vivo. Urinary clearance was slower and identical in mice receiving B72.3 labeled via glutathione or mercaptoethanol whereas clearance in the case of glutathione-reduced C110 was intermediate between that of the other two reducing agents. At both time points, higher radioactivity levels were observed in kidneys and lower levels in blood and most other tissues for both antibodies in the case of stannous ion reduction as expected for the label of greatest instability. In the B72.3 case, with only one exception, tissue and blood levels following administration of glutathione-reduced antibody were indistinguishable from that following administration of mercaptoethanol-reduced antibody. In the C110 case, significant differences in activity levels were observed in several tissues between glutathione- and mercaptoethanol-reduced antibodies. In conclusion, the relative in vivo behaviour of 99mTc when administered to mice while labeled to two IgG antibodies were successfully predicted based on the results of an in vitro cysteine challenge assay.

Investigations of ascorbate for direct labeling of antibodies with technetium-99m.

UNLABELLED: Recently, a method for the direct labeling of antibodies with 99mTc was described in which sulfhydryls were reportedly generated by reduction of antibody disulfides with ascorbic acid. Thereafter, these proteins may be labeled at high efficiency with 99mTc following reduction of pertechnetate with dithionite. This investigation was initially conducted to evaluate the mechanism of the increased stability towards cysteine challenge reported for the label and subsequently to determine the role of ascorbate in the labeling process. METHODS: It was possible to reproduce the reported high labeling efficiencies by increasing the dithionite concentration fivefold, presumably because of variabilities among lots of commercial sodium dithionite. RESULTS: Despite success in labeling, it was not possible to confirm that antibody reduction followed the treatment with ascorbate. Using both Ellman's reagent and 2,2' dithiodipyridine as indicators, we were unable to detect sulfhydryls on one IgG antibody treated at ten times the suggested ascorbate-to-antibody molar ratio. It was estimated that the number of sulfhydryls generated could not have been more than 1% (dithiodipyridine) to 2% (Ellman's). Furthermore, radiolabeling efficiencies for two IgG antibodies and stabilities of the label to cysteine challenge were unchanged when the ascorbate was eliminated. The number of sulfhydryls generated by treatment of the antibody with dithionite at 1-2 times the concentration required for adequate labeling was about 1% (dithiodipyridine) to 5% (Ellman's). CONCLUSION: For the conditions of this investigation and for the antibodies employed, ascorbate apparently played no more than a minor role at best in the labeling process. If antibody reduction occurred, this most likely was a result of residual dithionite presented to the protein along with the reduced 99mTc.

Directly and indirectly technetium-99m-labeled antibodies--a comparison of in vitro and animal in vivo properties.

To investigate the in vivo and in vitro properties of 99mTc when labeled to antibodies via one direct and one indirect method, the B72.3 and C110 IgG antibodies were radiolabeled directly via stannous ion reduction and indirectly via the hydrazino nicotinamide chelator and compared in vitro and in vivo. Antibody avidity (but not immunoreactive fraction) appeared to be independent of labeling methods for both antibodies. Following stannous ion reduction, antibodies were fragmented by denaturing SDS PAGE although only slight evidence of fragmentation was found in vivo. The direct label was instable to transchelation to cysteine and glutathione in vitro and in vivo. Following intravenous administration, urinary excretion of activity was threefold greater for the direct label and was almost exclusively labeled cysteine and glutathione. Significant differences in the biodistribution of 99mTc were also observed: liver levels were lower, kidney levels were higher and clearance of label from blood and tissues was faster for the direct label. At Day 1, tumor accumulation was threefold lower for the direct label although most normal tissues were also lower. In conclusion, when labeled to two antibodies by one direct method, 99mTc is unstable towards transchelation relative to one indirect method. These relative instabilities greatly influenced the biodistributions in mice and may influence the quality of images obtained in patients.

The molecular weight and oligomerization of rabbit thrombomodulin as assessed by sedimentation equilibrium.

Lubrol-solubilized rabbit thrombomodulin has been examined by equilibrium sedimentation in buffers that include sufficient D2O to make the detergent neutrally buoyant. Data were acquired at rotor speeds from 12,000 to 28,000 rpm from two thrombomodulin preparations, at protein concentrations from 0.01 to 0.07%, and in buffer containing 0.01 to 0.23% Lubrol. Examination of the data from different rotor speeds shows that the thrombomodulin exists as a heterogeneous mixture containing monomer (Mr 65,000), trimer, and higher oligomers. The oligomers do not equilibrate over the time scale of the experiment. The weight fraction as monomer varies from preparation to preparation, and appears to be independent of detergent concentration. Thus, experimenters should be cautious when interpreting binding or kinetic results obtained under similar buffer conditions.