The human immune system consists of various regiments of specialised cells, that float around in our bodies and patrol the perimeter, to detect the presence of both infectious agents from outside as well as aberrant and damaged cells on the inside. The earliest response to any attack comes from a group of general-purpose defence cells calles macrophages or phagocytes. These aren’t particularly sophisticated, they just go “oh this isn’t right, I better eat it, nom nom nom”, and they are also able to alert more specialised cells, but this takes a bit of time. A schematic of the two prongs of the human immune response looks like this :
When a cell in our body mutates, and that happens hundreds of times every day in each of us, those first-line defences kick in and kill off the aberrant cell (I’m simplifying things, there are also repair mechanisms etc). But since this is evolution we are talking here, cells can figure out a way to trick the immune system into leaving them alone, and that’s the way that a neoplasm may get established and spread. So naturally, research has looked into these immune processes for a long time, but to my knowledge until now there wasn’t a real breakthrough made.
Now research just published online in PNAS seems to indicate such a breakthrough. Essentially, it turns out that one way tumour cells trick the immune system into leaving them alone is that they express on their cell surface a protein called CD47, which signals to the macrophages “nothing to see here, move along, and don’t eat me”.It also turns out that all solid tumours in humans express CD47, and in higher concentrations than normal cells. Higher CD47 concentration in tumour cells is associated with shorter survival. Again, here’s a schematic :
By binding CD47, the signaling protein SIRPα on the macrophage initiates a cascade that eventually results in the inhibition of phagocytosis (The “don’t eat me signal”). And it is possible to block this cascade with an antibody, thereby enabling the immune system to attack those cells.
So when scientists implanted human breast, ovarian, and colon cancer cells into mice, waited for the cancer to get established and then treated the mice with an antibody that blocks CD47, the mice’s immune system kicked in and either significantly shrunk or even completely killed the tumours, and inhibited metastatic spread. This seems to be related to tumour size, if it’s big to start with, lesser mice were cured. The authors conclude :
Here we demonstrate that expression of CD47 is a general mechanism used by human patient solid tumor cells to evade phagocytosis. Blocking mAbs that disrupt the interaction be- tween CD47 and SIRPα enabled the phagocytosis of solid tumor cells in vitro and inhibited tumor growth in several orthotopic xenotransplantation models. Moreover, anti-CD47 antibodies prevented the formation of tumor metastases. These results es-
tablish CD47 as a therapeutic target on solid tumor cells.
First up, once again, so far it’s mice, not people. But it’s a good sign that this works in immunocompetent as well as immunocompromised mice, because one of the questions I had was whether the response was so dramatic partly because human tissue was being implanted into a different species. It’s also a good sign that there seems to be no showstopping side effects. You will remember that I said CD47 is expressed in normal human cells, but it doesn’t seem that the immune system, after CD47 blockade, starts eating up human cells indiscriminately. The main side effect seems to be a transient anaemia or neutropenia.
But any treatment on humans with this antibody is years away, although the authors seem rather keen to go and start phase 1 and 2 trials as soon as possible. So don’t get your hopes up too soon, this may in the end just not work. But it certainly looks very promising to me.
—————————————————————————-
The CD47-signal regulatory protein alpha (SIRPa) interaction is a therapeutic target for human solid tumors,
Proceedings of the National Academy of Sciences of the United States of America
Stephen B. Willingham,
Jens-Peter Volkmer et al
Published online before print March 26, 2012, doi: 10.1073/pnas.1121623109
