It was in 1999 that I wrote a research grant at Stanford University, in which I postulated that some of the chemokine receptors (especially CXCR4 and CCR7) could be involved in organ-specific tumor metastasis of for example breast cancer - a topic I was already interested for many years due to former work on CD44 and tumor progression. Unfortunately, I didn't get the grant and my visa ended, but others in an independent company on Stanford's campus - really not using my approach, but a totally different one - published in 2001 in Nature a correlation of those receptors and organ-specific metastasis - although the mechanism remained unclear. I later (in 2001, unpubl.) found that already normal breast tissue showed some expression of those receptors, which was in contrast to that Nature paper (the main authors later confirmed to me, I was right).
I may not have followrf that research in every detail, but would like to know, if there is now more clarity in that field, i.e. did I miss any break-throughs. I also invented later a very different approach to better understand the possible mechanism of integrin activation by chemokines; therefore I developed a new method: to the best of my knowledge to date, I became in 2001 the first to measure on a single-molecule level with AFM (atomic force microscopy) the specific (!) interaction between a single integrin and its ligand (VLA-4 / VCAM-1) between two living (!) cells (inhibeted by anti-VLA-4 or VCAM-1 MAbs, respectively, but not control antibodies). Shortly later in 2003, i.e. after two years on my original idea, I successfully developed the method to measure what I originally wanted to measure: the (physiologic) activation of a single integrin receptor on a living cell, mediated by a chemokine (VLA-4 / VCAM-1 / SDF-1; control: pertussis toxin). Unfortunately, others in those labs were supported by Leopoldina members and continued on my exclusive projects, collaborations and published even without even acknowledging me ... Since only leukocytes appeared to have integrins with different activation states related to chemokine-induced activation, I used mouse lymphoma cell lines (also for safety reasons, since I did not want to have any physicists work on human cells) - although I already had organized the human cells, human chemokine (SDF-1) and a temporary inhibitor for CXCR4.
Is there any knowledge I may have missed:
1) are there in the meantime any other cells than leukocytes and derived tumor or stem cells known, which show such an activation of integrins by chemokines (my impression was, all cancer cells I checked had a constitutively active form of integrins, but in some cases totally incactive, but highly expressed other cell adhesion receptors)
2) is the concept of chemokines directing organ-specific metastases further confirmed? any more specific chemokine receptors related to metastases?
Thanks for your input.
Aren't integrins always on the cell surface, and ready for action. I work on Adam family members, which contain Disintegrin domains. They are membrane proteinases that cleave things like CXCR4. So Adams, which are known to promote cancer and metastasis when overexpressed, or just expressed in tumor cells, can bind to integrins, and inhibition of this interaction is known to prevent things like migration of the tumor cells through matrices in vitro. In Vivo, and in the clinic, there are antibodies that bind to integrins.
So, yes, I think that integrins, via interactions with Adams, may cause release of chemo mine receptors, which can regulate the way in which tumor cells migrate. Adams can process probably all type I and II membrane proteins, so they affect many things, ie even soluble icam, vcam, etc. And the balance between soluble receptors and ligands is what controls how TUMORS grow and spread.
Thank you, Marcia, for your qualified answer and bringing other proteins, like ADAMs to my attention !
And yes, integrins usually should be ready for action, i.e. binding to their ligands, BUT only on certain types of cells -usually members of the leukocyte family- the VLA4 and LFA1 also can exist in at least two, perhaps three, different, but physiologic, activation states: low- and high-affinity, perhaps one or more intermediate states as well. They also can interfere with the cytoskeleon. In addition, under experimental conditions, probably all integrins can be inactivated. The activation of VLA4 or LFA1 on leukocytes involves the binding of a chemokine to a chemokine receptor, which immediately activates the integrin (to better binding).
So, Adams appear to be not very specific, but worth to look more carefully how they could relate to organ-specific metastasis. My ideas and hypotheses are more in the direction of "metastasis stem cells", a hypothesis, which I was trying to develop for the last 2 decades - and which does not seem to contradict the famous "tumor stem cell" theory from my former sponsor about a decade ago.
To refine my question:
Are there any tumor stem cells known, in addition to the lymphoma and leukemia lineages, that might have - at the tumor stem cell level - a similar integrin-chemokine-crosstalk as the rolling leukocytes? If yes, the mechanism might also be similar; if not, there must be a different mechanism - a new potential target for future cancer therapies.
The rolling leukocytes are mediated by adam17 cleavage of l selectin and tumor cells also do this. The tumor cells typically can be derived from any tissue. Forexample the breast cancer cell line, BT747 expresses l selectin. For tumor stem cells, I don't know. But I bet your hypothesis is right. I know that other Adam members may be involved in tumor stem cell survival. But I would suggest you contact Peter Dempsey, a colleague of mine for a definitive answer.
Robert, Paul Saftig also knows about this topic. Please look at his publication list. And maybe reach out to him. If you are interested in this topic, there is a Gorden Research Conference to be help in March on Regulated proteolysis of cell surface proteins.
In 1997 I found (unpubl.) only very, very few tumor cells capable of rolling on endothelial monolayers; that included mostly the known homing receptors (as far as I could identify and confirm them) and their known corresponding counter-receptors; it also included a selectin (in my findings not L-selectin, but P-selectin). I couldn't really find CD44 as rolling mediator - as I had suggested years before (the main reason why I had chosen in 1995 to go to Stanford or Harvard - at a time none of the established CD44 experts ever could really think of CD44 as a rolling mediator..., was published a year later in Science, after I asked at Stanford, if there was any possibility to check it there).
So, now I definitely will read more about the adam17 and the physiologic shedding of L-selectin. Perhaps this could lead to a project with AFM (atomic force microscopy).
CD44 cleavage is mediated by ADAM10 for sure. But other Adam family members could be involved. I think the atomic force method would be viable. There are now specific inhibitors of Adam 10, 17 and 8. It would be interesting
It would be interesting to look at rolling, minus and plus inhibitor. Joerg Bartsch in Marburg has the adam8 inhibitor. Gill Murphy has the adam17 inhibitor, and Incyt
Is there any more news on the chemokine receptors potentially involved in organ-specific metastasis?
Robert, there are lots of papers on the subject. But I don't know much more about it.
Thank you, Alasdair, for your answer. When I found in mid-2001 that CXCR4 and CCR7 were indeed expressed in all types and stages of mammary carcinoma (Eibl, unpubl.), I also found them expressed in some normal epithelial cells within the breast tissue, which was in contrast to the Nature paper, which somehow excluded the expression in normal tissue, but Albert Zlotnik (the senior author) and others shortly later confirmed to me that I was right. It was around that time that the cancer stem cell theory came up - not sure if that could bring together those findings. I mean the metastasis stem cell (which I wanted to find in the 1990s) just is an organ stem cell with all migration tools including the chemokine receptor needed, it only needs to grow and already can migrate.
Just found this paper on MMP8 cleavage of adam10 and spc cells. ADAM10 also cleaves e cadherin. And apparently in this paper, mmp8 can cleave from the membrane Adam10. Cleavage of adam10 is known to regulate its ability to cleave membrane proteins.
We found that SPCs in atheromas expressed MMP8 and that MMP8 knockout significantly reduced SPC numbers in atherosclerotic lesions in apolipoprotein E (ApoE)-deficient mice fed a Western diet. Further in vivo experiments showed that ApoE(-/-)/MMP8(-/-) mice injected with stem cells isolated from bone marrows of ApoE(-/-)/MMP8(-/-) mice had fewer SPCs in atheromas and smaller lesions than ApoE(-/-)/MMP8(-/-) mice injected with stem cells isolated from bone marrows of ApoE(-/-)/MMP8(+/+) mice. Ex vivo experiments showed that MMP8 deficiency inhibited the ability of SPCs to migrate from the arterial lumen and the adventitia into atherosclerotic lesions. In vitro assays indicated that MMP8 facilitated SPC migration across endothelial cells and through Matrigel or collagen I. We also found that MMP8 cleaved a-disintegrin-and-metalloproteinase-domain-10 and that MMP8 deficiency reduced mature a-disintegrin-and-metalloproteinase-domain-10 on SPCs. Knockdown of MMP8 or incubation with the a-disintegrin-and-metalloproteinase-domain-10 inhibitor GI254023X decreased E-cadherin shedding on SPCs. The decrease in migratory ability of SPCs with MMP8 knockdown was reduced by incubation of such cells with culture supernatant from SPCs without MMP8 knockdown, and this compensatory effect was abolished by an antibody against soluble E-cadherin.
- Badges
- Science topic
- Similar topics
- Biological Science
- Immunology