Human Umbilical Vein Endothelial Cells (HUVEC) were obtained from Cell Applications, Inc. and grown in an endothelial cell growth medium with full growth supplements (Cell Applications, Inc.). Bovine pulmonary artery endothelial cells (BPAEC) and human dermal microvascular endothelial cells (HDMEC) were obtained from Lonza and were grown in an EBM medium plus EGM SingleQuots of growth supplements and EBM-2 medium plus EGM-2 SingleQuots of growth supplements (Lonza), respectively. Glioma cell lines U-87 MG and U251 were kindly provide by Dr. V. Wee Yong (University of Calgary, Calgary, Alberta, Canada). The human epidermoid carcinoma cell line HEp3 was a generous gift from Dr. Andries Zijlstra (Vanderbilt University, USA). Mouse embryonic fibroblast (MEF) cells were isolated from mouse embryos and were used at their early passages (less than passage 4). U-87 MG, U251, HEp3 and MEF cells were cultured in DMEM containing 10% (v/v) fetal bovine serum (FBS, Invitrogen) and 1% penicillin-streptomycin mixtures (Invitrogen). All other cell lines were obtained from American Type Culture Collection (ATCC, Rockville, MD) and were grown in DMEM, MEM or RPMI (Invitrogen) containing 10% (v/v) fetal bovine serum and 1% penicillin-streptomycin mixtures according to ATCC's instructions. All cells were grown in an incubator at 37°C containing 5% CO2. All endothelial cells and primary fibroblast cells were maintained and used before the ninth passage.

The plasmid pET17b-DT385 expressing “receptorless” DT385 was generously provided by Dr. Sundaram Ramakrishnan (Department of Pharmacology, University of Minnesota Medical School, Minneapolis, MN). The plasmid pET17b-p22 was constructed by cloning human plasminogen fragment p22 into pET17b (EMD Biosciences) at the NdeI and BamHI restriction sites. The plasmid pLIC-DT385-p22 was constructed by cloning human plasminogen fragment p22 into the plasmid pLIC-DT385 at the NcoI and HindIII restriction sites, while the plasmid pLIC-DT385 was constructed by inserting DNA encoding DT385 but lacking the stop codon into the pET-30 EK/LIC vector following the manufacturer's protocol (Novagen). The plasmid pSUMO encoding the cDNA for SUMO (small ubiquitin-related modifier, Saccharomyces cerevisiae, Smt3 gene) was kindly provided by Dr. Kaisong Zhou (Dalhousie University). All plasmid constructs were confirmed by DNA sequencing (DalGEN, the Dalhousie University DNA sequencing facility). Recombinant p22, SUMO, DT385, and DT-p22 were expressed in E. coli and purified by Ni-NTA affinity column (Qiagen), pooled and dialyzed overnight against phosphate buffered saline (PBS). All recombinant proteins were purified as a single band as revealed by SDS-PAGE analysis. LPS was removed from purified proteins using polymixin B beads (Detoxi-Gel Endotoxin Removing Gel, Pierce) according to the manufacturer's instructions. Multiple passes through the column were performed until LPS level was less than 30 EU/mg.

Cell viability was assessed by the Cell Titer96 Aqueous One Solution Cell Proliferation Assay (MTS assay-Promega) using the manufacturer's protocol. The IC₅₀ values were calculated by nonlinear least-squares fitting of dose-response curves using the open source computer program QTI PLOT. Data were analyzed with the four-parameter logistic equation f  = (a − d)/[1 + (x/c)b] + d, where a is the asymptotic maximum, b is a slope parameter, c is the value at the inflection point (IC₅₀) and d is the asymptotic minimum. For a control, 0.4 µM or 1.2 µM DT385 was added to the tissue culture media in the absence of cells followed by incubation with the MTS/PMS reagent. Reduction of MTS was not observed under these conditions, confirming that DT385 did not interfere with this assay.

Cell proliferation was also evaluated with crystal violet staining. At the end of treatment with DT385, cells were fixed with 100% methanol, and stained with 0.5% crystal violet in 20% methanol for 15 minutes at room temperature. Stained cells were photographed at 25× magnification on a Zeiss Axiover 200 inverted microscope (Carl Zeiss, Germany).

Apoptotic and necrotic cells were visualized with the apoptosis and necrosis assay kit (Biotium, Inc) following the manufacturer protocol. Stained cells were photographed on a Zeiss Axioplan II fluorescence microscope (Carl Zeiss, Germany), with appropriate filter settings for fluorescein isothiocyanate (FITC) and Texas Red fluorescence. Digital images were processed in Adobe Photoshop (Adobe Inc.).

U-87 MG or Hela cells were seeded at a ratio of 1∶10 in 1 ml of DMEM plus 10% FBS per well in a 12-well plate overnight. Cells were treated with 1 µM DT385 for the indicated time, incubated for 15 minutes at 37°C with 0.5 ml labeling medium (methionine-free DMEM, 10% dialyzed FBS and 50–100 µCi Pro-Mix L-[³⁵-S]-(Amersham)) and cell lysates analyzed by SDS-PAGE. Radioactive bands were visualized by radiography using a phosphorImager after overnight exposure on a phosphor screen. For quantitation, [³⁵S] incorporation was measured by liquid scintillation counting.

The anti-angiogenic activity of proteins was tested on the CAM as described [24]. When analyzing tumor-induced vascularization, 50,000 HEp3 cells replaced bFGF/ VEGF as the angiogenic stimulus [25]. Four implants were placed on each CAM and ten embryos were used for each experimental compound.

HEp3-GFP tumour cells (100,000 cells/10 µl DMEM media) were applied directly to a filter-disc abraded area of the CAM of 9-day old chick embryos as detailed by [26]. Chicks were incubated under standard conditions until Day 15 when tumors were imaged and sorted for random distribution to control and treatment groups. Daily systemic intravenous injections of DT385 (2 µg in 50 µL PBS) or PBS (50 µL) were performed on day 15, 16 and 17. On day 18, tumors were excised, cleaned of CAM and then weighed. Alternatively, fluorescent (GFP) Hep3 tumors were imaged in vivo using a Zeiss Lumar fluorescence stereomicroscope. The outline of the tumor was determined using the fluorescence signal of the tumor cells (GFP channel (ex470/em525)). Specifically, Volocity software was calibrated to select areas with fluorescence intensity corresponding to 1 or more standard deviations above background. The area of the tumour was defined and calculated using Improvision Volocity software (Perkin Elmer, Inc.)(Version 5.3.0 Build 0). A unit of measure (1 mm) was calibrated using the grid of a hemocytometer. The determination of tumor volume assumes the shape to be hemiellipsoidal and the equation for the tumor volume is: V  =  π /6 • (length) • (width) • (height) where height  = 1.63√(Length •Width). Data were analyzed using Students t-test.

All animal work was carried out at the animal facility of Dalhousie University in accordance with the guidelines set forth in the Care and Use of Laboratory Animals by Dalhousie University. All investigations were approved by the Dalhousie Animal Research Ethics Board. Female 6-to 8-week old C57BL6/J mice (Jackson Laboratories) were used. Tumors were induced by subcutaneous injection of LLC cells (10⁶ cells) in 100 µl of sterile PBS. Palpable tumors were established three to four days after injection, at which point the mice were randomly assigned to two experimental groups, those that received recombinant SUMO (control) and those that received recombinant DT385 treatment. SUMO and DT385 were administered to mice, peritumorally at day 5 (25 µg in 100 µL PBS), and at days 9, 12 and 15 (10 µg in 100 µL PBS each injection).

DT385 and BSA were FITC-labeled using the EZ-label FITC protein labeling Kit following the manufacturer's protocol (Pierce).

U87 cells were incubated in the absence or presence of DT385 (2 µM) alone or in combination with 10 mM ammonium chloride. The medium was removed at various time points and replaced with fresh media. Thirty six hours after the addition of test compounds, cell viability was measured by the MTS assay.

The significance of the data was determined using the Student's t-test (one-tailed). P values of <0.05 were regarded as significant.

Recent studies from our laboratory identified and characterized a novel antiangiogenic fragment of plasminogen called p22 [27]. Since p22 was a potent and specific inhibitor of capillary endothelial cells, it was envisioned that this protein could be used to target DT to newly forming vasculature. We expressed a fusion protein in E. coli in which p22 was substituted for the receptor-binding domain of DT (DT385-p22). The activity of this fusion protein was compared with recombinant DT385 and recombinant p22. To determine the impact of these compounds on cell viability, MTS assays were performed on bovine pulmonary artery endothelial cells (BPAEC) after a three-day treatment. Interestingly, we observed that unlike the p22 prepared by proteolytic digestion of human plasminogen [27], the recombinant p22 failed to inhibit the viability of the BPAEC (Figure 1A). Unfortunately we have been unable to produce active recombinant p22 in E. coli or yeast (data not shown). Of particular interest was our observation that both the recombinant DT385-p22 construct and recombinant DT385 dramatically decreased the number of viable BPAEC. We determined an IC₅₀ value of 0.21 µM and 0.14 µM for DT385-p22 and DT385 respectively (Table 1).

Since the recombinant p22 was inactive whereas DT385-p22 retained activity, it was unclear from these results if p22 might have retained activity when expressed as the fusion protein, DT385-p22. Considering the specificity of plasminogen-derived p22 for endothelial cells, we expected that if the p22 component of the DT385-p22 fusion protein, then DT385-p22 should only target endothelial cells and not cancer cells, as has been demonstrated for plasminogen-derived p22 [27]. We therefore tested the cytotoxic activity of DT385-p22 and DT385 with cancer cell lines. For these experiments we initially tested two commonly used human cancer cell lines, the HT1080 fibrosarcoma cell and the HeLa cervical carcinoma cell. Unexpectedly, we found that DT385 caused a dramatic loss in cell viability of the cancer cell lines in a dose-dependent manner (Figure 1B). At the highest dose tested (2.4 µM), 10% of viable Hela cells remained, while about 50% of viable HT1080 cells remained. DT385-p22 did also cause a loss in cell viability but was less potent than DT385. To determine whether the viability of human endothelial cells was affected by DT385-p22, we incubated HUVEC and HDMEC with these proteins (Figure 1C). Surprisingly, the viability of both cell lines was not affected by DT385, D385-p22 or p22 at the concentration up to 2.5 µM. We extended the incubation time to 7 days with 2.4 µM of these recombinant proteins, and no loss of cell viability was observed (Figure S1,A). However, with very high concentrations of DT385, loss of viability of the HUVEC and HDMEC was observed (Figure S1,B). We determined an IC₅₀ of about 5.77 and 7.54 µM DT385 for the HUVEC and HDMEC, respectively. These results suggested that the cytotoxic activity of DT385-p22 was due to the activity of the DT385 domain of the fusion protein and not the p22 domain. We also concluded that DT385 can kill human cancer cells at a concentration not affecting primary endothelial cell lines. The observation that DT385 had cytotoxic activity was novel and unexpected. We therefore investigated the possibility that DT385 might kill other cancer cells.

We extended the cytotoxicity assay to 18 human cancer cell lines (Table 1). Fifteen cancer cell lines were affected by DT385 and efficacy of DT385 varied among the cancer cells. Cancer cell lines with highest sensitivity to DT385 (IC₅₀ less than 0.5 µM DT385), included U-87 MG, U251, 293T, HEK293, Hela and Calu-3 cells. The group with intermediate sensitivity to DT385 (IC₅₀ between 0.5-1.5 µM) included Colo201, Colo205, LNCap, PC-3, HT1080, and MDA-MB-231 cells. Weakly sensitive cancer cell lines with IC₅₀ greater than 1.5 µM included MCF7, HCT116, BT-20, NB4, HL-60, and HEp3 cells. In contrast, recombinant p22 had no effect on cell viability at concentrations as high as 2.4 µM (data not shown). The three human cancer cell lines not affected by DT385 at the highest dose tested (2.4 µM) were the promyelocytic leukemia cell lines (HL-60 and NB4) and the epidemoid carcinoma cell line (HEp3).

The DT385-mediated loss in cell viability was also confirmed using crystal violet staining (Figure 2). For example, less than 10% of glioma U-87 MG cells remained after DT385 treatment. To confirm cross-species sensitivity to DT385, we tested mouse skin melanoma (B16-F10), Lewis lung carcinoma (LLC) and mouse embryonic fibroblast (MEF) cells. While the B16-F10 and MEF were highly sensitive to DT385, the LLC were only weakly sensitive (Table 1). Together the data from MTS assays and Crystal Violet staining indicate that the “receptorless” DT, DT385, can indeed be cytotoxic to many tumor cell lines. The molecular mechanism that accounts for the broad range in sensitivity to DT385 is currently under investigation.

To evaluate the cytotoxicity of DT385 on other primary cell lines, primary cultures of three human fibroblast cell lines (CCD-1064Sk, BJ, and IMR-90) and one monocyte (SC) cell line were incubated with increasing concentrations of DT385 and the cell viability was determined by the MTS assay. CCD-1064Sk and BJ fibroblasts had an IC₅₀ of 2.22 µM and 2.37 µM DT385, respectively. The lung fibroblasts (IMR-90) had an IC₅₀ of 1.44 µM and the viability of the monocytes (SC) was unaffected by DT385 at concentrations up to 2.4 µM (Table 1). Again, recombinant p22 had no effect on cell viability at concentrations up to 2.4 µM (data not shown). In contrast, when confluent CCD-1064Sk, BJ and IMR-90 fibroblasts were treated for 3 days with 2 µM DT385, we observed no significant loss of cell viability (Figure S2). These data indicate that even when DT385 is effective in decreasing the viability of proliferating primary cells, it fails to kill these cells when they are quiescent/confluent. Taken together, these data support our previous conclusion that DT385 alone has the potential to kill cancer cells with minimum effects on primary cells.

To determine if the resistance to DT385 could be overcome by continuous incubation, cells with weak or no detectable sensitivity received a second dose, 48 hours after the first treatment and cell viability was measured 2 days later. The human epidermoid carcinoma cell line, HEp3 and breast cancer cell line MDA-MB-231 both exhibited enhanced sensitivity with continuous incubation with DT385. Compared to a single application of DT385, which did not affect HEp3 cells significantly, two applications of DT385 decreased the cell viability (IC₅₀ of 2.07 µM (Figure S3). Similarly, the IC₅₀ for the breast cancer cell line, MDA-MB-231 decreased from 1.02 µM to 0.66 µM upon a second incubation with DT385 (Table 1). In contrast, a second incubation with DT385 was without effect for the primary endothelial cells, HUVEC or HDMEC (Figure S3). Taken together, the results show that receptorless DT in the form of DT385 is cytotoxic to a large number of tumor cells. While the sensitivity to DT385 varies, continuous exposure diminishes the viability of even the most resistant cancer cell lines.

DT kills cells by a mechanism involving cellular apoptosis. We also observed increased apoptosis in DT385 treated cells (Figure 3). The incubation of cultured U-87 with DT385 resulted in a dramatic increase in cellular apoptosis as measured by an increase in annexin V staining (85%) and the uptake of ethidium homodimer (87%).

DT has been shown to enter toxin-sensitive mammalian cells by receptor-mediated endocytosis which involves the interaction of the receptor-binding domain of the protein with its extracellular receptor. Since DT385 does not possess a receptor-binding domain, it should be incapable of binding to cells and becoming internalized. To investigate if DT385 was internalized, we tracked fluorescently labeled DT385 with confocal microscopy. Tumor cells sensitive to DT385 were incubated with FITC-DT385 and imaged with confocal microscopy. As shown in Figure 4A , incubation of cells with DT385 resulted in the internalization of the toxin which was detectable in perinuclear vesicles. In contrast, incubation of cells with similar concentrations of FITC-bovine serum albumin did not result in internalization of the protein (Figure S4). This experiment suggested that the uptake of DT385 by cells was not due to non specific cellular uptake of the protein.

Ammonium chloride is a weak base that diffuses into the endosome and serves as a proton reservoir, thus inhibiting the acidification of the endosome. We utilized ammonium chloride treatment of cells to investigate whether DT385 entered the cell by endocytosis. We observed that incubation of U87 cells with DT385 for as little as 2 hours resulted in significant cell death (Figure 4B). However, the simultaneous addition of ammonium chloride and DT385 abrogated the cytotoxic activity of DT385. The observation that ammonium chloride blocked the cytotoxic activity of DT385 suggests that DT-385 enters cells through an acidic endocytic pathway.

Diphtheria toxin kills cells by catalysing the ADP-ribosylation of EF-2, leading to inhibition of protein synthesis [5], [28]. To investigate if the DT385 decreased cell viability by inhibiting protein synthesis, glioma U-87 MG cells were treated with 1 µM DT385 for 36 hours followed by labeling with [³⁵S-] methionine for 15 minutes. As shown in Figure 5A, SDS-PAGE analysis of cell lysate showed that protein synthesis was severely reduced in DT385 treated cells. This inhibition occurred within 24 hrs of treatment and persisted during the duration of the assay (48 hr) (Figure 5B). The 36-h treatment gave the greatest decrease in protein labeling. These data suggest that, mechanistically, DT385 decreases cell viability by blocking protein synthesis. Similar results were obtained for Hela cells (data not shown).

The progression and metastasis of neoplastic disease requires extensive vascularization of the tumor to sustain the nutritional and oxygen demands of the proliferating cancer cells. This is generally achieved through a process called angiogenesis [29]. The chick chorioallantoic membrane assay (CAM) is a useful in vivo model system to investigate the effects of toxins on angiogenesis. As shown in Figure 6A, DT385 at a concentration of 40 nM (1.2 pmol in 30 µL) caused complete inhibition of HEp3-induced (angiogenic) vascular sprouting. This concentration is far below the IC₅₀ of the tumor cells and therefore a direct effect of DT385 on HEp3 viability is unlikely.

To investigate whether DT385 could reduce tumor growth, HEp3 tumors were grown on the CAM and the 6-day old tumors were then treated with 2 µg of DT385 injected intravenously daily for three days. Assuming that the total blood volume of a 15–17 day chick embryo is approximately 2 mL [30], the initial systemic concentration of DT385 used in these experiments would be about 20 nM. Tumors were removed and weighed 24 hours after the last injection. We observed that DT385 decreased tumor mass by 41% (Figure 6B). We also observed a significant loss of 66% in tumor volume after treatment with DT385 (Figure 6C). These results indicate that DT385 can inhibit tumor angiogenesis and subsequently reduce tumor growth.

To study the effect of DT385 in another in vivo model, LLC were implanted in mice and tumors were allowed to grow for a 19-day period. DT385 was administered subcutaneously on days 5, 9, 12 and 15 at a dosage of 25 µg for first injection and 10 µg for the remaining injections. We observed that tumor growth was significantly inhibited in mice receiving DT385 by day 19 (23%) when compared to control mice treated with an identical concentration of the recombinant control protein, SUMO (Figure 7). We did not observe any toxicity to the mice at these dosages. However, we observed an LD₅₀ of 120 µg of DT385 (n = 10, data not shown) suggesting that higher doses of DT385 were lethal to mice.