ANGIOSTATIN AND CANCER
Angiostatin was isolated in the 1990s by Judah Folkman and his colleagues at the Boston Children’s Hospital. They inoculated mice with cells from a tumour that metastasises to the lungs: over a few weeks the cells grew to form primary skin tumours about 1 cm in diameter at which point they were surgically removed (Fig. 1).
1. Mouse metastasis model.
At this stage, micro-metastases could be detected throughout the lungs but none had developed beyond the critical size of about 1 to 2 mm or created their own blood supply – they were dormant. However, once a primary tumour was removed, the dormant tumours developed rapidly so that, after a further three weeks, the lungs doubled in weight due to the metastases. This suggested that a primary tumour itself had been producing an anti-angiogenic factor that prevented blood vessel development in the dormant tumours. If a protein spreads throughout the body it’s pretty likely that some will be excreted. Accordingly, Folkman and colleagues collected urine from their mice and analysed the proteins, eventually isolating angiostatin. Purified angiostatin given daily to mice after removal of a primary tumour completely prevented the development of micro-metastases. Angiostatin is active against primary tumours established in mice from inoculated human tumour cells and it also inhibits the proliferation of endothelial cells in culture (Fig. 2).
2. Endothelial cells in culture: effect of reduced oxygen.
As so often happens in cancer, clinical trials of angiostatin and endostatin were disappointing and they have been superseded by a recently developed angiogenesis inhibitor, bevacizumab (Avastin®), a monoclonal antibody against VEGFA, which has been used with some success in treating metastatic colon cancer and glioblastoma.
In an astonishing illustration of their adaptability, some types of tumour cells are able to differentiate into endothelial-like cells, thereby forming a neovasculature that is not host-derived. This behaviour has been detected in glioblastoma, a particularly aggressive type of a brain tumour, and confirmed both by immunohistochemical analysis of tissue sections and by showing that the tumour-derived endothelial cells share their mutational profile with a parent tumour. Thus their neovasculature is derived from tumour cells rather than, for example, from a fusion with host endothelial cells. This differential flexibility extends to the generation of smooth muscle cells and when these tumours are transplanted into mice their growth is supported by the expansion of the human, tumour-derived vascular tissue.
Tumours are abnormal growths and it is not surprising that the blood vessels they create are also pretty weird. Usually they don’t have any discernible pattern: sometimes they just stop like a kind of cul-de-sac, sometimes blood flows into them from both ends generating a form of traffic chaos and, in general, they are leaky and tortuous and have unstable blood flow patterns compared with normal tissues.
This picture of chaotic structure and flow suggests that some parts of a tumour may get less oxygen than others and measurements show that not only is this true but that cells within tumours often survive on much less oxygen (lower partial pressure, pO2) than normal cells in an adjacent tissue. In fact, the centre of a growing tumour often becomes so hypoxic that the cells die, forming a necrotic core, while the outer regions of a tumour continue to grow. The instability of tumour vasculature suggests that the degree of hypoxia varies not only between different regions of a tumour but also with time. There is considerable evidence that what has been termed ‘cycling hypoxia’ can occur both relatively rapidly, due to fluctuation in blood flow through tumour microvessels, and more slowly, probably as a result of vascular remodelling. In the face of this behaviour cancer cells have shown great versatility in the adaptations they have evolved to help their survival. These centre around the metabolic pathways that convert glucose into the principal energy currency of the cell, ATP.
There are two main stages: glycolysis and oxidative phosphorylation. Glucose to pyruvate: this pathway is an anaerobic reaction (it doesn’t need oxygen and it parallels what happens when yeast makes alcohol, except that yeast, converts pyruvate to ethanol rather than lactic acid). The most important thing, however, is that glycolysis generates two molecules of ATP for every glucose that is turned into pyruvate, the final step in the pathway being catalysed by pyruvate kinase. The pathways of the tricarboxylic acid cycle (TCA cycle) and oxidative phosphorylation take the pyruvate generated by glycolysis and turn it into water and carbon dioxide. This process requires oxygen (it is aerobic), it takes place inside a specialised organelle – the mitochondrion – and the energy released is converted into about 30 ATP molecules per glucose molecule consumed.