Per Borgström


Borgström Lab

Molecular Targeting of Cells Program


A cancerous tumor is not only composed of tumor cells. There are numerous other types of cells such as fibroblasts, endothelial cells, and immune cells. These cells make up the stroma of the tumor. An important criterion in the treatment of solid tumors is the understanding of the molecular and microenvironmental events that influence tumor formation, growth, metastasis and vascularization. Currently used rodent tumor models, including transgenic tumor models, or subcutaneously-growing human tumors in immunodeficient mice, do not sufficiently represent clinical cancer, especially with regard to metastasis and drug sensitivity. Preclinical tumor model systems employed to evaluate potential new treatment strategies should aim to represent the process and patterns of metastasis of their clinical counterparts as closely as possible. The lack of adequate mouse models that recapitulate the tumor microenvironment has significantly delayed advances in the understanding of the role of the tumor microenvironment.
We have developed a tumor model that more accurately depicts human tumors in many ways beyond the standard subcutaneous models used predominately in the past. We use a chamber technique, called the dorsal skinfold chamber. Our model in awake mice, using fluorescence video-microscopy allows us to evaluate not only tumor growth and angiogenic activity, but also mitotic and apoptotic indices as well as tumor cell migration of implanted tumor spheroids. Under high magnification, the underlying mechanisms behind tumor regression such as cell cycle arrest, mitotic catastrophe, or apoptosis can be distinguished. Temporal and spatial relation of vessel density to cell death allows detailed quantitative evaluation of regimen effectiveness. This model thus provides a unique means to acquire real time data of micro tumor biological events in the context of the host.
In our chamber, we have successfully grafted a variety of organ tissues, including prostate, liver, lung, mammary fat pad and ovary tissue, and shown that it is possible to grow tumor spheroids in explants from these different tissues.
Our system allows continuous measurements of both growth/regression and angiogenesis of the small tumor spheroids. We are using this system to evaluate various therapeutic interventions on prostate (TRAMP-C2), mammary carcinoma (N202), as well as ovarian epithelial cancer (MOVCAR). We also use our model to clarify the role of tumor stroma and the genetic changes in stromal cells in the initiation and progression of tumors.

Research Interests

Angiogenesis. Tumors can not grow beyond a certain size without angiogenesis. Every successful increase in the tumor cell population must be preceded by an increase in new capillaries that converge upon the tumor. Without angiogenesis, potentially malignant tumors can stay in a "dormant" state for a number of years without invading surrounding tissues. The critical event converting such a dormant tumor into a rapidly growing malignancy, is the switch to the angiogenic phenotype demarcating two stages in the development of the tumor - the prevascular phase and the vascular phase.

Angiogenesis, a fundamental process by which new blood vessels are formed, is rare in adult mammals under normal physiological conditions. In adult physiological angiogenesis, i.e., wound healing, negative inhibitory influences ultimately take over and the process of neovascularization ceases at the completion of the wound healing. However, in pathophysiological conditions like cancer and chronic inflammation, these inhibitory controls fail and angiogenesis persists.

The idea of anti-angiogenic drugs as a strategy for cancer therapy was first proposed nearly 25 years ago. Since then, many compounds have been shown to inhibit angiogenesis in various in vitro and in vivo systems. To date, many angiostatic drugs are being tested in clinical trials.

The formation of new blood vessels during angiogenesis can be broken down into a number of distinct yet overlapping processes. Angiogenesis begins with the early inflammatory phase, characterized by dilated and permeable vessels. This early inflammatory response is followed by a proteolytic step where the basement membrane of the endothelium is degraded through the action of a variety of proteases. All metastatic tumor cells synthesize the required protease for remodeling endothelial basement membranes. Several metalloproteinase inhibitors are currently being tested in clinical trials. Proteolysis is followed by nonmitogenic migration of capillary endothelial cells towards the angiogenic stimuli such that endothelial cells migrate from the vascular wall, through perivascular connective tissue and parenchyma. The next step in the process of angiogenesis is the proliferation of endothelial cells behind the leading front of migrating endothelial cells. A variety of soluble polypeptides and other factors have been described that promote such endothelial proliferation. One such factor is the vascular endothelial growth factor (VEGF), an endothelial cell-specific mitogen and a growth factor which is secreted by a variety of human tumors. Various antagonists of the VEGF signaling pathway are currently in Phase II/III trials.

To study the effects of anti-angiogenic substances, a variety of in vitro assays, which model the different steps of angiogenesis have been developed over the years. The traditional models include the cornea pocket and the chorioallantoic membrane of the chick embryo (CAM). A number of experimental models are also available for the study of angiogenesis in normal and tumor tissues. However, no model allows detailed description of early phases of tumor induced angiogenesis with simultaneous visualization of tumor growth in vivo. To fill this gap, we developed a system for non-invasive, in vivo and in situ study of tumor angiogenesis in conscious mice, using video-microscopy of tumor spheroids implanted in dorsal skin fold chambers in nude mice.


 Click to view an MPEG illustrating the usefulness of histone H2B-GFP labeling of tumor cells.

Comparative Evaluation of Systemic Inhibitors. Currently we are using this system to study prostate cancer angiogenesis. Our in vivo model permits the evaluation of a molecule’s angiostatic potential as well as its consequences on cell cycle and apoptosis. Several metastatic variants of prostate carcinoma are being evaluated in this micrometastasis model environment using intravital microscopy in the dorsal skinfold chamber of athymic mice. Three variants of PC3 and LNCAP, which vary in their metastatic potential, have been transduced with a retroviral vector expressing Histone-H2B as a fusion protein with enhanced green fluorescent protein. This fusion protein, implemented in our laboratory, through collaboration with Dr. Geoff Wahl’s group at the Salk Institute in La Jolla, permits the evaluation of mitotic and apoptotic indices by way of its association with chromatin in a non-invasive manner. Valuable mechanistic information regarding cell cycle and apoptosis may therefore be obtained from sample fields of the prostate cancer micrometastasis model, growing in the dorsal skinfold chamber of a nude mouse.

This model permits the evaluation of angiostatic agents within a 14 - 21 day period. In other orthotopic or experimental metastasis models, this process requires several months per experiment, is laborious to perform, and may lead to ambiguous results given the inherent latency of prostate carcinogenesis. Moreover, this model is being developed with the goal of comparative analysis of different angiostatic agents currently in preclinical development.

Prostate Angiogenesis Model. Variants of different metastatic potential from LnCAP and PC3 are transduced with VSV pseudotyped histone H2B-GFP retrovirus. Tumor spheroids prepared using the liquid overlay technique ensures reproducibility of tumor spheroid size. Spheroids are implanted in dorsal skinfold chambers in nude mice, and animals are observed at certain intervals during a two week observation period, and evaluated for tumor area, vascular density as well as mitotic and apoptotic indices. The system allows evaluation of systemic as well as local treatment regimens.


By comparing the efficacy of a candidate therapeutic to other treatment regimens within the same model, a given therapeutic may be evaluated on the competitive basis of 1) the extent of angiostasis 2) the effects of such angiostasis on cell cycle progression within the core and periphery of a micrometastasis and 3) the relative cell death as a result of angiostasis. The ability to weigh all three parameters gives a more balanced perspective to the potential of a candidate therapeutic in this relatively new field. The use of the LNCAP and PC3 variant models will hopefully permit the evaluation of the potential synergy of angiostatic therapy combined with androgen ablation and/or chemotherapy regimens. In earlier experiments we implanted tumor spheroids from the human prostate carcinoma cell line DU 145. Results from these experiments demonstrated that anti-VEGF treatment results in complete inhibition of tumor angiogenesis and no growth beyond an initial prevascular phase.

Photomicrographs illustrating the complete inhibition of tumor angiogenesis after anti VEGF treatment of nude mice implanted with the human prostate cell carcinoma cell line DU 145. In tumors of control animals, vascular networks with high vdensity were induced, wheras in tumors treated with an anti-VEGF moAb (A4.6.1, 200 ug twice weekly IP, Genentech Inc.,) angiogenic activity was reduced to at most budding.


Tetracycline Regulated Expression Vectors: With the rapid advancement of novel gene therapy based strategies for angiostasis, and given the obvious potential for potent "bystander effects", we have constructed a new expression system to generate tetracycline regulated cDNA constructs of angiostatic agents. This vector eliminates much of the labor intensive screening of multiple clones in the presence and absence of tetracycline. pTRE contains a minimal CMV promoter which is fused to an upstream tetracycline response element, TRE. For the rapid generation TET-regulated expression of candidate genes in TET-OFF cell lines, we have included an internal ribosomal entry site (IRES) with green fluorescent protein (TRE-IRES-eGFP). This vector allows the generation of TET regulated cell lines which express essentially any gene of interest. As the candidate gene is located between the promoter and the eGFP, transcription of the candidate gene is required to allow the transcription of the downstream reporter gene eGFP using the IRES. Other vectors that use reporter genes under separate promoters frequently result in varied levels of expression of the reporter and the gene of interest. By performing first positive and then negative FACs sort on the drug resistant population of cells, in the absence and presence of tetracycline respectively, a GFP inducible population of cells can be isolated that is regulated by the presence of tetracycline or doxycycline. Moreover, as the eGFP and the gene of interest are transcribed under the same mRNA, levels of GFP florescence correlate with levels of expression of the candidate gene of interest, allowing a sort of cells with a specific level of maximal gene induction. RT-PCR / Western Blot analysis of the gene of interest from the sorted population can then be used to assure the inducible expression of the gene of interest. Initial generation of Tet-Off cell lines is achieved using the TET-OFF vector (Clontech, Palo Alto, CA). We have used this TETOFF IRES2GFP model to evaluate the angiostatic potential of several candidate angiostatic genes in the highly angiogenic HT1080 model.

Candidate Angiostatic Genes. Illustration of the TET-OFF system using the highly angiogenic cell line HT1080. When implanted in dorsal skinfold chambers in nude mice, these cells induce a very high density vascular network (Panel A). Panel B demonstrates that the gene is effectively turned OFF, i.e., no GFP fluorescence. Panel C illustrates the angiostatic activity when the gene is turned on as illustrated by the GFP fluorescence (Panel D).


In addition to this plasmid-based model, we have now successfully generated and tested a 2-vector tetracycline regulated retroviral system using the TRE-IRES2GFP model and a modified TET-ON system. In this system, the TETON retroviral vector contains the VP16 TET transactivator followed by an IRES sequence and the TTS sequence which encodes a novel repressor protein that prevents gene expression until the levels of doxycycline reach 10ng/ml, after which the TETON transactivator drives gene expression. This model results in a more tightly regulated expression system than the older "leaky" TETON vectors. The retroviral system is packaged in an HEK-293 system pseudotyped with the VSV envelope, allowing concentration of virions to very high titers (>109).

Evaluation of the angiogenic potential of tumor tissues derived from human prostate cancer thin needle biopsies: In ongoing investigations, we are also using our in vivo system to evaluate the angiogenic activity of tumor tissues derived from human prostate cancer thin needle biopsies and are comparing these with the established prostate cancer cell lines mentioned above. We also examine the capacity of a neutralizing anti-VEGF moAb, and a soluble VEGF receptor fusion protein (fIt-IgG) to inhibit the angiogenesis and growth of the thin needle biopsies. While these experiments are designed to examine the dependence of heterogeneous prostate tumor xenograff biopsies on VEGF mediated angiogenesis, we also intend to examine whether neutralization of VEGF is synergistic or antagonistic to some of the chemotherapy regimens used for prostate cancer. We will combine anti-VEGF treatment with the following two agents:
a) Estramustine phosphate (EMP), a unique antitumour agent, which is selectively taken up by prostate cells and exerts antineoplastic effects by interfering with microtubule dynamics and by reducing plasma levels of testosterone. It has been demonstrated that the antitumour response obtained with oral EMP as a single agent is comparable with that obtained with the best intravenous chemoregimens but EMP induces no myelotoxicity, and other adverse effects of treatment are usually mild.
b) Taxol (paclitaxel) is a chemotherapeutic agent derived from the yew tree that show important activity in metastatic prostate cancer. Paclitaxel, the first of a new class of compounds, antimicrotubule agents called taxanes, has a unique mechanism of action in blocking cells at multiple phases of the cell cycle, inhibiting cell division and rendering cells non-functional. Interestingly, combining paclitaxel with EMP, results in enhancement of the effect of paclitaxel.

© 2009 Vaccine Research Institute of San Diego. All rights reserved.