Experimental and clinical studies indicate that low molecular weight heparins (LMWH) may inhibit cancer and/or metastasis. The purpose of this study was to investigate whether it is possible to design non-anti-coagulant, anti-metastatic compounds based on heparin. The LMWH Tinzaparin and a series of non-anti-coagulant (NAC) heparin derivatives, varying in size from 2,500 to 10,000 Da, were tested for their anti-metastatic activity in an experimental B16F10 metastasis model. The most promising NAC heparin drug candidate and Tinzaparin were further evaluated in B16F10 model with spontaneous metastasis from a primary subcutaneous tumor. In the experimental model, Tinzaparin, NAC2500, and NAC6000 were inactive whereas both NAC8000 and NAC10000 significantly inhibited the number of induced experimental metastases by 69 and 73%, respectively. NAC8000 was chosen over NAC10000 for further studies because of its lower molecular weight with an expected better bioavailability. In the spontaneous model, Tinzaparin had no inhibitory effect on metastatic activity. In contrast, NAC8000 significantly inhibited the number of metastases by 58%. Neither Tinzaparin nor NAC8000 inhibited primary subcutaneous tumor growth. Together, these results indicate that the anti-metastatic effect of heparin derivatives is not a result of anti-coagulant activity. The non-anti-coagulant NAC8000 specifically inhibits early establishment of tumor cells, but not primary tumor growth. Therefore, NAC8000 is a promising non-anti -coagulant compound for preventing tumor metastasis.

We recently described an in vivo angiogenesis assay for rats--an optimized Matrigel plug assay based on subcutaneously implanted chambers with fixed volume and shape. Here we examine the possibility of switching the host animal from rat to mouse, thereby reducing the requirement for test compound an order of magnitude. The chambers consist of a plexiglas ring with a 0.2 ml volume and two nylon net filters. Chambers containing growth factor-reduced Matrigel supplemented with basic fibroblast growth factor (bFGF) were subcutaneously implanted in the right flank of three different strains of mice; BALB/c, C57BL/6J, and NMRI-nu. On day 10 post-implantation: i) each chamber was taken out, ii) a picture of the induced angiogenic response was taken, and iii) the redness of the chamber content was quantified by computer image analysis. The level of bFGF-induced angiogenesis in the mouse assay was lower than in the previously published rat assay. Importantly, the background angiogenesis in mice in chambers containing Matrigel alone was correspondingly decreased. Therefore, a more sensitive threshold for the computer image analysis was used. In all three strains of mice, bFGF-induced angiogenesis was significantly increased compared to Matrigel alone. Furthermore, the positive anti-angiogenic control compound TNP-470 (10 mg/kg/d s.c.) completely inhibited the bFGF-induced angiogenesis. The in vivo chamber angiogenesis assay allows quantitative analysis of angiogenic and anti-angiogenic activity in mice. The model is very robust and only little influenced by the choice of mouse strain.

Efficient in vitro and in vivo angiogenesis assays, to assess and compare anti-angiogenic activity are a prerequisite for the discovery and characterization of anti-angiogenic targets. Here we describe an optimized Matrigel plug assay based on subcutaneously implanted chambers and two fast and reproducible measuring techniques. Plexiglas ring/nylon net filter-chambers (0.2 ml) containing growth factor-reduced Matrigel and 300 ng basic fibroblast growth factor (bFGF) were subcutaneously implanted into the right flank of rats. Chamber angiogenesis was scored on day 5 and day 10 post-implantation by computer image analysis of the chamber, and by optical density reading at 415 nm of a PBS solution of the chamber content. bFGF significantly induced chamber angiogenesis and histological examination confirmed that numerous blood vessels were present in the bFGF-induced chambers. The anti-angiogenic control compound TNP-470 (10 mg/kg/d s.c.) completely inhibited the bFGF-induced angiogenesis. In contrast, the anti-inflammatory or immuno-suppressive compounds cyclosporin A (15 mg/kg/d p.o.), indomethacin (1 mg/kg/d p.o.), and prednisolone (5 mg/kg/d p.o.) showed no anti-angiogenic activity, indicating that the bFGF-induced angiogenesis was not driven by an inflammatory response or by a foreign body reaction. Finally, two candidate anti-angiogenic compounds were tested in the assay. Continuous low-dose therapy with cyclophosphamide (25 mg/kg/d p.o.) significantly inhibited bFGF-induced angiogenesis, whereas 1alpha,25-dihydroxyvitamin D3 (0.5 micro g/kg/d p.o.) showed no significant anti-angiogenic activity. In conclusion, this in vivo chamber angiogenesis assay is a useful new tool for drug evaluation as well as research into anti-angiogenesis.

Little is known about the effects of antiangiogenic therapy on perfusion of human tumors and the mechanisms by which tumors can adapt to these treatments and recur. Here, we examined the effects of serial passaging of LN-229 human glioma xenografts overexpressing thrombospondin (TSP)-1 on tumor growth, vascularity, and perfusion. Persistence of TSP-1 overexpression was confirmed after three serial s.c. passages of small xenografted tumor blocks of cells stably transfected with TSP-1 cDNA (clones C9 and E7) or vector controls (pooled clones A7-A9) in immunodeficient nu/nu mice. The tumor vascularity was estimated by noninvasive near infrared spectroscopy measuring blood volume at 800 +/- 10 nm and by histological vessel scores in CD31-immunostained cryosections. The tumor perfusion was assessed by noninvasive laser Doppler flowmetry. Overexpression of TSP-1 significantly inhibited tumor growth. In size-matched tumors (approximately 300 mm(3)), the blood volume and the histological vessel scores were lower in the TSP-1-transfected tumors than in controls, and this effect was more pronounced in tumors derived from the clone with the highest TSP-1 expression (clone E9). Despite this clear reduction in tumor vascularity, the tumor perfusion was the same in TSP-1-transfected tumors and controls. This study shows that TSP-1 overexpression slows glioma growth in vivo but does not prevent it from reaching a large size (300 mm(3)). Whereas a clear reduction in blood volume during tumor growth and a reduced vascular index at sacrifice are observed in TSP-1-transfected tumors, this did not affect perfusion when size-matched comparisons were performed. Given the increased time needed to reach equal size, it indicates that a fixed rate of perfusion must be maintained in the tumor to allow for growth. Elucidation of the mechanisms that allow this to happen has important consequences for the understanding of tumor recurrence after antiangiogenic therapy.

The potential of noninvasive laser Doppler flowmetry (LDF) and near infrared spectroscopy (NIRS) to detect acute effects of different vascular-modifying agents on perfusion and blood volume in tumors was evaluated. C3H mouse mammary carcinomas (∼200 mm3) in the rear foot of CDF1 mice were treated with flavone acetic acid (FAA, 150 mg/kg), 5,6-dimethylxanthenone-4acetic acid (DMXAA, 20 mg/kg), combretastatin A-4 disodium phosphate (CAMP, 250 mg/kg), hydralazine (HDZ, 5 mg/kg), or nicotinamide (NTA, 500 mg/kg). Tumor perfusion before and after treatment was evaluated by noninvasive LDF, using a 41°C heated custombuilt LDF probe with four integrated laser/receiver units, and tumor blood volume was estimated by MRS, using light guide coupled reflectance measurements at 800±10 nm. FAA, DMXAA, CAMP, and HDZ significantly decreased tumor perfusion by 50%, 47%, 73%, and 78%, respectively. In addition, FAA, DMXAA, and HDZ significantly reduced the blood volume within the tumor, indicating that these compounds to some degree shunted blood from the tumor to adjacent tissue, HDZ being most potent. CAMP caused no change in the tumor blood volume, indicating that the mechanism of action of CAMP was vascular shut down with the blood pool trapped in the tumor. NTA caused no change in either tumor perfusion or tumor blood volume. We conclude that noninvasive LDF and MRS can determine acute effects of vascular modifying agents on tumor perfusion and blood volume.

The use of laser Doppler flowmetry (LDF) and near-infra-red spectroscopy (NIRS) for non-invasive in vivo measurements of angiogenic and anti-angiogenic activity in nude mice was evaluated. Angiogenic foci were induced in the skin by implantation of slow release pellets containing 200 ng basic fibroblast growth factor (bFGF). LDF and NIRS recordings from induced foci were significantly higher than placebo implants (P<0.05) and controls (P<0.001), proving that LDF and NIRS provide measures of angiogenic activity. Correspondingly, by these methods, an anti-angiogenic activity was significantly demonstrated in bFGF-stimulated animals treated with either the specific anti-angiogenic compound TNP-470 (P<0.05) or the anti-inflammatory agent dexamethasone (P<0.001). We conclude that LDF and NIRS, alone or in combination, are useful non-invasive tools for early evaluation of angiogenic and anti-angiogenic activity in vivo.

We examined the relationship between non-invasive estimates of the tumor hemoglobin concentration by near-infrared spectroscopy (NIBS) and histological scores of tumor vascularity by Chalkley counts in seven tumor lines in nude mice [malignant gliomas: U87, U118, U373; small cell lung cancers (SCLC): 54A, 5413, DMS79; prostate cancer: MatLyLu (MILL) ]. We also evaluated the effect of continuous anti-angiogenic treatment with TNP-470 on tumor hemoglobin concentration and tumor vascularity in U87 and MILL tumors. Non-invasive LAIRS recordings were performed with a custom-built flash near-infrared spectrometer using light guidecoupled reflectance measurements at 800±10 nm. Chalkley counts were obtained from CD31-immunostained cryosections. The LAIRS recordings in arbitrary absorbance units increased with tumor size in the individual tumors until a plateau was reached at approximately 150 mm3. This plateau was relatively tumor line-specific. LAIRS recordings at the plateau phase were strongly correlated (P<.001, n=71) to the histological vessel score (Chalkley count) of the same individual tumors excised immediately after the LAIRS was performed. Non-invasive LAIRS recordings of the highly vascularized gliomas (U87, U118, and U373) plus the MatLyLu tumor line were significantly higher than the three less vascularized SCLC tumor lines (P<.001). Continuous treatment with the anti-angiogenic compound TNP-470, an endothelial cell inhibitor, significantly retarded tumor growth in both U87 and MILL tumors, but all tumors eventually grew. When comparing treated and untreated tumors of similar size, both LAIRS recordings and Chalkley counts were significantly lower in TNP-470-treated tumors (P<.05). In conclusion, the LAIRS technique provides a non-invasive measure of the degree of vascularization in untreated tumors and the LAIRS technique can measure modifications in tumor vascularization by anti-angiogenic therapy.

We examined the effect of the anti-angiogenic compound TNP-470 on early tumor growth characteristics following subcutaneous implantation of 1 mm3 tissue blocks of human glioblastoma U87, in nude mice. The mice received daily injections with TNP-470, 7 mg/kg, from one day before until either 3, 7, 11, or 15 days after inoculation. The time from inoculation until initiation of exponential tumor growth was determined along with the post-therapeutic growth delay and the initial tumor doubling time (TD) for each individual tumor (n=103) on the basis of tumor volume growth curves. We found that: i) the onset of growth of U87 xenografts was effectively inhibited by concurrent treatment with TNP-470 beyond the first three days after inoculation, ii) this effect was fully reversible upon termination of therapy, and iii) the post-therapeutic growth delay was independent of the accumulated dose. These findings demonstrate that the in vivo effect of TNP-470 on tumor growth is tumor inhibitory rather than cytotoxic. The lack of effect of the anti-angiogenic compound, TNP-470, in the early 3-day schedule is consistent with the existence of an early avascular phase which precede the angiogenesis-dependent tumor growth.

The urea-extractable proteins from calcified regions of intermoult. cuticle of the lobster, Homarus americanus, have been separated by two-dimensional electrophoresis, showing that the extracts contain a large number of proteins. The major proteins have isoelectric points between 4 and 9, and their apparent molecular weights are between 5 and 30 kDa.Two of the proteins have been purified by a combination of ion-exchange chromatography, gel-filtration and RP-HPLC, and their complete amino acid sequences were determined by a combination of mass spectrometry and automated Edman degradation.Although they were purified from a single animal, both proteins were obtained as two isoforms. The isoforms of the smaller protein (HaCP4.6) differed only in a single position (phenylalanine/isoleucine), and the isoforms of the larger protein (HaCP11.6) differed in two positions (valine/isoleucine and glutamine/lysine). HaCP11.6 is N-terminally blocked by a pyroglutamate residue.Variants of an 18-residue motif are a characteristic feature of both sequences: it occurs twice in HaCP4.6 and four times in HaCP11.6. Comparison of the sequences to sequences published for cuticular proteins from other arthropods shows that the repeated motif is also present in proteins from the exoskeleton of the Bermuda land crab, Gecarcinus lateralis, but not in the single shrimp protein (Pandalus borealis) sequenced so far. The amino acid compositions of the lobster proteins are similar to that of flexible cuticles in locusts, but no convincing sequence similarities were found between the lobster proteins and cuticular proteins from locusts or other insects. (C) 1997 Elsevier Science Inc.