|Home | About | Journals | Submit | Contact Us | Français|
Several investigators have shown that the SDF1α/CXCR4 signaling pathway has a role in promoting prostate cancer metastasis. It has been shown previously that a neutralizing antibody directed toward CXCR4 inhibits prostate cancer cell adhesion and invasion in vitro, inhibits localization to the bone when the prostate cancer cells are injected into the heart, and inhibits tumor growth when the prostate cancer cells are injected directly into the tibial bone. An interesting paper in the current issue of this journal by Xing and colleagues 1 entitled “Tumor cell specific blockade of CXCR4/SDF1α interaction in prostate cancer cells by hTERT promoter induced CXCR4-knockdown: a possible metastasis preventing and minimizing approach” describes the effects of stable down-regulation of CXCR4 on prostate cancer cell adhesion, invasion, and growth in the bone. These data provide additional insights into the growing list of potential mechanism(s) by which CXCR4 and SDF1 may contribute to the metastatic process (see Figure 1).
The chemokine receptor CXCR4 is one member of a large family (≈ 18) of G-protein coupled receptors and recognizes peptides (i.e., chemokines) as ligands 2,3. SDF1α (also known as CXCL12) is the predominant ligand for CXCR4 4. The effects of the SDF1α/CXCR4 signaling pathway on prostate cancer are considered to be multifactorial, and it has been implicated in the homing of tumor cells to specific organs (metastasis), as well as the establishment and growth of tumor cells at specific sites, which are most likely mediated by the effects of CXCR4 on cancer cell adhesion, invasion, and proliferation. The current study by Xing et al 1 builds on the work of other investigators that have implicated the CXCR4/SDF1α pathway in promoting prostate cancer cell invasion and metastasis. Examination of CXCR4 expression in patient tumors has shown that it is frequently expressed on metastatic cancer cells 5. A number of organs express SDF1α, including the lung, bone, and lymph nodes. Other investigators have used blocking of CXCR4 function with a neutralizing antibody to show that CXCR4 is involved in prostate cancer cell adhesion and invasion in vitro, and the localization of prostate cancer cells to bone when injected into the heart in vivo. Blocking of CXCR4 function with a neutralizing antibody also had been shown to inhibit tumor growth when prostate cancer cells were injected directly into the tibial bone 6–9. The Xing et al article both confirms and extends these findings by showing that the stable down-regulation of CXCR4 inhibits SDF1α-promoted prostate cancer cell adhesion to osteosarcoma and endothelial cells, inhibits invasion through Matrigel, and reduces the incidence of prostate cancer establishment and growth when the tumor cells are injected into the tibial bone of mice. Thus, these studies represent an important advance in our knowledge of the role of CXCR4/SDF1α signaling in prostate cancer metastasis. They also raise several important new questions regarding CXCR4/SDF1α signaling and most especially the effects of this process in the context of other signaling events.
SDF1α can bind CXCR7, as well as CXCR4, and it has been reported very recently that the CXCR7/SD1α signaling pathway plays a role in prostate cancer biology 10. It is not clear whether CXCR7 is expressed on the PC3 prostate cancer cells utilized by Xing et al. 1. It is possible, however, that the expression of CXCR7 on these cells accounts for the absence of complete inhibition of tumor establishment and growth in the tibia and the absence of complete inhibition of cell adhesion and invasion in vitro that these investigators found utilizing PC3 cells with down-regulated CXCR4. Thus, it would be of interest to determine the relative effect of SDF1α signaling through CXCR4 as compared to CXCR7 on prostate cancer cell adhesion, invasion and metastasis perhaps through down-regulation of these individual receptors in a cell line that is known to express both CXCR4 and CXCR7. It would also be interesting to assess whether cells in which CXCR4 is down-regulated compensate for this loss by upregulating CXCR7 or other receptors.
The finding of Xing et al., 1 that SDF1α stimulation increases prostate cancer cell invasion through Matrigel and that this is inhibited with the down-regulation of CXCR4, suggests one possible mechanism whereby the SDF1α/CXCR4 signaling pathway could promote metastasis. The effects of SDF1α on cell adhesion have been attributed for the most part to changes in the integrin family of cell adhesion receptors. The integrin family of receptors largely recognizes extracellular matrix proteins as ligands and initiates reorganization of the actin cytoskeleton that can lead to cell movement. Specifically, either increased expression of a particular integrin or an increased activation state of an integrin has been reported on SDF1α stimulation. Other investigators analyzing prostate cancer invasion through Matrigel have shown an increase in invasion with SDF1α stimulation together with increased expression or activation of integrin αvβ3 or α5β1 8,6. Thus, SDF1α could be promoting invasion in the experiments described by Xing et al. by increasing the expression or activity of integrin αvβ3 or α5β1. Increased expression or activity of these two integrins could also promote pro-survival or pro-proliferation signals 11. The latter signals attributed to engagement of integrin αvβ3 or α5β1 are consistent with the reported increased activation of ERK and Akt with SDF1α stimulation of prostate cancer cells reported by other investigators 12,9. In addition, SDF1α stimulation has been reported to increase expression of several proteases: MMP-2, MMP-9, stromelysin-1, stromelysin-2, and the membrane type 1 metalloprotease (MMP-14) 13,12. The increased expression of proteases on SDF1α stimulation most likely creates a pathway for cell movement/invasion.
An increase in adhesion of the PC3 prostate cancer cells to endothelial cells and to osteosarcoma cells was also observed with SDF1α stimulation and an inhibition of this adhesion with the down-regulation of CXCR4 in the Xing et al paper 1. Cell adhesion to endothelial cells likely involves the participation of the selectin or the ICAM/VCAM family of cell adhesion receptors. Engl et al 6 reported, as data not shown that the DUI45 and LNCaP prostate cancer cells attach to E-selectin, but minimally to P-selectin, ICAM-1 or VCAM-1. Bone sinusoidal endothelium is distinct from endothelium in other vessels 14. Moreover, although not addressed in the Xing et al paper, bone endothelium differs from endothelium in other organs for several properties, depending upon where in the bone the endothelium resides 15.
As always, care must be taken in extrapolating data generated using animal models to the metastasis in humans. Xing and colleagues used an in vivo model of metastasis in which a post-homing event is measured, i.e., the establishment and growth of the prostate cancer cells in the bone environment after direct injection into the tibia. Investigators commonly use experimental models of prostate cancer establishment and growth in the bone that are based on injection of the tumor cells directly into bones other than the vertebrae, most likely due to the difficulty in injecting tumor cells into the vertebra of mice. In patients, however, prostate cancer metastases are frequently localized to the vertebrae (spine) 16,17. As pointed out by Xing and colleagues, this is likely influenced by the promotion of homing of the cells by the extensive vascularization of the spine. In terms of the post-homing events, the preferential localization of prostate cancer metastases to the vertebrae in humans raises the question as to whether the levels or availability of SDF1α differ in the vertebrae as compared to other bones. Interestingly, Sun et al. 7 have reported that although SDF1α protein levels (as measured by ELISA) are similar in the spine and lung, their expression in these tissues is significantly lower than in the tibia, femur and pelvis. This would suggest that factors other than the levels of SDF1α can contribute to the preferential metastases of prostate cancer to the vertebrae (spine) in patients, although it is possible that the levels of SDF1α in the bones varies among individuals. For example, the availability of SD1α may be enhanced when bone or lung tissue is compromised by age-related changes, trauma, or behavioral factors, such as smoking. Thus, studies in nude mice that utilize the tibia for assessment of prostate cancer establishment and growth may not be completely representative of human prostate cancer metastasis.
In summary, the paper by Xing et al. solidifies a role for CXCR4 in promoting prostate cancer cell adhesion to bone-derived cells, prostate cancer cell invasion through Matrigel, and prostate cancer cell establishment and growth in the tibial bone. However, the molecular mechanisms promoting these effects need further elucidation. For example, is the effect of SDF1α on invasion due solely to a change in the expression or activity of integrins αvβ3 or α5β1, and is the effect of SDF1α on cell-cell adhesion due to changes in the selectin family, ICAM/VACM family, or another family of cell-cell adhesion 7 receptors? What are the negative checks on the SDF1α/CXCR4 pathway such that the metastatic prostate cancer cell can remain stationary and proliferate in appropriate parts of the metastatic tumor; i.e., does local degradation of SDF1α by the CD26/dipeptidyl-peptidase IV occur in prostate cancer metastasis in vivo 18. Collectively, these issues impact decisions concerning the best strategy for inhibiting the SDF1α/CXCR4 pathway in prostate cancer. Currently, a small molecule inhibitor of CXCR4 (CTCE-9908, Chemokine Therapeutics Corp.) is being tested in animal models of cancer.