In this study, we have shown that inhibitors of PI 3-kinase negatively regulate branching morphogenesis in SMG by acting directly on the epithelium. Consistent with a role in branching, PI 3-kinase is localized primarily within the epithelium. We have rescued the effect of the PI 3-kinase inhibitors by addition of an immediate target of PI 3-kinase, PIP3
. We also report that phosphorylation of Akt, another downstream effector of PI 3-kinase, decreases (at Ser473
) with increasing incubation time with PI 3-kinase inhibitors and correlates with inhibition of branching morphogenesis. Together, these data indicate that PI 3-kinase is involved in branching morphogenesis and that its effects are mediated through signaling via PIP3
and possibly Akt. Since the addition of PIP3
can stimulate cleft formation, we propose a model in which PIP3
-dependent signaling initiates cleft formation. Alternatively, PIP3
may not initiate cleft formation but may provide a cleft-stabilizing signal. PIP3
is generated by phosphorylation of PIP2
, which also has signal transduction properties (Czech, 2000
). These data suggest that the formation of PIP3
, and not the reduction in PIP2
levels, is the critical event downstream of PI 3-kinase signaling during branching morphogenesis. Since little is known regarding the mechanism of cleft formation, the role of PIP3
in SMG branching morphogenesis will be the subject of future studies.
Because the specificity of the chemical inhibitors is a concern in inhibitor studies (Davies et al., 2000
), we have used two chemically independent PI 3-kinase inhibitors, LY294002 (LY) and wortmannin (wmn), and observed the same effects on branching morphogenesis with both. Wmn has been reported to affect PI4Kβ
(Meyers and Cantley, 1997
), although not at the levels used in this study. LY has inhibitory effects on the mammalian target of rapamycin (mTOR) (Brunn et al., 1996
), yet mTOR does not appear to be a target in SMG branching morphogenesis, since rapamycin itself had no effect on branching (). LY has also been reported to inhibit DNA-dependent protein kinase (Hartley et al., 1995
) and casein kinase-2 (CK2) (Davies et al., 2000
). While we found that apigenin, a CK2 inhibitor, did inhibit branching morphogenesis (), apigenin is not specific for CK2 and affects a dozen different kinases (Pinna, 2002
). With the lack of specific CK2 inhibitors, it is currently difficult to test the role of CK2 in branching morphogenesis. Nevertheless, as a result of these studies, we can conclude that PI 3-kinase is involved in branching morphogenesis; however, we cannot exclude the possibility of a contributions from other molecules.
PI 3-kinase is a mediator of many signal transduction pathways in many systems, which include both EGF- and FGF-stimulated signaling (Carballada et al., 2001
; Okano et al., 2000
; Toker, 2000
). In agreement with previous data in which inhibitors of EGFR moderately inhibit branching morphogenesis (Kashimata and Gresik, 1997
; Kashimata et al., 2000
), we have shown that the highly specific EGFR inhibitor (reported IC50
of 25 pM) inhibits branching morphogenesis in our SMG organ cultures. Likewise, we found that PD98059 moderately inhibited branching morphogenesis () similar to previous reports (Kashimata et al., 2000
). In our culture system, not dependent on externally added growth factors, the effects of the PI 3-kinase inhibitor LY were greater than the effects of the MAP kinase inhibitor PD98059 (). When LY was combined with either PD98059 or PD153035, there was no additive effect in this system. However, when combined with the FGFR1 inhibitor, SU5402, there was a combinatorial effect on branching morphogenesis, suggesting that LY and SU5402 affect separate pathways. Additional evidence in support of the hypothesis that PI 3-kinase is not a component of FGF signaling comes from the results of the mesenchyme-free cultures. It was previously reported that, in mesenchyme-free cultures, EGF stimulates bud formation, while FGF7 stimulates duct elongation (Morita and Nogawa, 1999
). In our mesenchyme-free cultures, LY () and wmn () appear to allow ductal extension but inhibit branching. Together, these data indicate that PI 3-kinase is a downstream effector of EGF, but not FGF, action during SMG branching morphogenesis.
While the EGFR inhibitor PD153035 is more specific (Bridges et al., 1996
) than the tryophostin RG50864 (Nowak et al., 1997
), which was previously used to demonstrate the significance of EGF signaling during SMG development (Kashimata and Gresik, 1997
), it inhibits not only EGFR1 (ErbB1) but also the EGFR family members ErbB2, -3, and -4, indicating that one or more of these family members may be involved in SMG branching morphogenesis. In fact, EGFR1 is not a strong activator of PI 3-kinase (Yarden and Sliwkowski, 2001
). Further, it has been reported that the EGF family member most effective in stimulating PI 3-kinase activation is ErbB3 (Soltoff et al., 1994
). During SMG branching morphogenesis, it is not known which ErbB family member simulates PI 3-kinase.
The fact that the inhibitory effects of PI 3-kinase inhibitors appear to be greater than that of EGF pathway inhibitors indicates that PI 3-kinase may participate in an additional pathway involved in branching morphogenesis. Another stimulator of branching morphogenesis is integrin α
6, as was shown by the inhibition of branching by integrin α
6 inhibitory antibodies in organ culture (Kadoya et al., 1995
), yet the mechanism remains unknown. We added a function-inhibiting integrin α
6 antibody (GoH3) to SMG cultures and assayed for phosphorylation of Akt at Ser473
as an indicator of PI 3-kinase activity. However, we found that there was no effect on Akt phosphorylation by GoH3 (data not shown), suggesting that PI 3-kinase does not mediate integrin α
6 signaling. The possibility remains that PI 3-kinase contributes to an additional, possibly unknown, pathway involved in SMG branching morphogenesis.
What additional downstream mediators of PI 3-kinase might be involved in branching morphogenesis is not known. Since loss of Akt phosphorylation correlates with addition of LY and inhibition of branching, Akt may also be involved. Consistent with this possibility is the fact that ML-9 inhibits branching morphogenesis at 35 μ
M (data not shown). Although ML-9 inhibits Akt phosphorylation (Hernandez et al., 2001
), it is also an inhibitor of myosin-light chain kinase and protein kinase A at these levels. Rapamycin, besides inhibiting mTOR, also inhibits p70S6 kinase (p70S6K), which is a downstream mediator of some PI 3-kinase activities. However, since rapamycin did not inhibit branching morphogenesis (), p70S6K does not appear to play a role in branching morphogenesis. Additional molecules activated by PI 3-kinase signaling, including isoforms of protein kinase C (Nakanishi et al., 1993
; Toker et al., 1994
) and Rac (Hawkins et al., 1995
), are potential targets involved in branching morphogenesis. Future studies will seek to identify specific molecules down-stream of PI 3-kinase/PIP3
in branching morphogenesis.
Proliferation, migration, cell-shape change, matrix synthesis, matrix degradation, and other changes in cell physiology are components of the process of branching morphogenesis. The processes are different in each branching organ. In contrast to the kidney, in which branching morphogenesis initiates with ureteric bud outgrowth, branching initiates with cleft formation in the SMG. It is interesting that PI 3-kinase plays a role in the initiation of branching in the kidney and SMG, despite the different mechanisms involved. In the SMG, it is clear that PI 3-kinase/PIP3
plays a role in initiation of cleft formation, although it is not clear whether other physiological mechanisms are involved, except that the effects of PI 3-kinase do not appear to depend on changes in proliferation. In the presence of PI 3-kinase inhibitors, proliferation within the epithelium is maintained, as determined by BrdU incorporation assays (data not shown). It is not known whether PI 3-kinase can affect some cell migratory processes during SMG cleft formation as it does during the active cell migration during ureteric bud outgrowth (Tang et al., 2002
). In the future, it will be interesting to identify the downstream mediators of PI 3-kinase/PIP3
signaling in order to understand the detailed mechanisms of this pathway during mouse SMG branching morphogenesis.