Original article

M.D. BASSON, B. ZENG, S. WANG

THE C-TERMINAL REGION OF THE FOCAL ADHESION KINASE F1 DOMAIN
BINDS AKT1 AND INHIBITS PRESSURE-INDUCED CELL ADHESION

Departments of Surgery, Pathology, and Biomedical Sciences, School of Medicine and Health Sciences, University of North Dakota,
Grand Forks, ND, USA
Increased extracellular pressure or shear stress activate a complex signal pathway that stimulates integrin binding affinity and potentiates metastatic cell adhesion. Inhibiting either focal adhesion kinase (FAK) and Akt1 can block this pathway, but risks interfering with the diverse other functions of each kinase. However, the mechanotransduced signal pathway involves a novel Akt1-FAK interaction not required for most FAK or Akt1 function, so modeling and blocking this interaction seems a desirable target. Building upon previous work suggesting that FAK-Akt1 binding is mediated by the FAK F1 lobe, we demonstrated that independently expressing the F1 domain in human Caco-2 or murine CT-26 colon cancer cells by transient or stable inducible plasmid expression respectively prevents the stimulation of cancer cell adhesion by increased extracellular pressure. Serial further truncation of the FAK F1 lobe identified shorter regions capable of pulling down Akt1 on a glutathione S-transferase (GST) - conjugated column. Ultimately, we identified a 33 residue segment (residues 94-126) at the C-terminal of the F1 lobe as sufficient to pull down Akt1. These findings raise the possibility of developing a treatment modality around the disruption of the FAK-Akt1 interaction using peptides modeled from FAK.
Key words:
adhesion, Akt, cancer, focal adhesion kinase, metastasis, glutathione s-transferase, pressure-stimulated adhesion

INTRODUCTION

The metastatic spread of cancer from a primary tumor initially develops through an interplay between the dissemination and successful implantation of viable tumor cells. While a primary tumor of 1 cm in size is able to shed one million cells into circulation per day (1), another important contributor is the release of tumor cells by physical manipulation during surgical interventions. Such iatrogenic dissemination not only dislodges tumor cells into the surgical site (2, 3) but also substantially increases the number of circulating tumor cells (4-8). How many of these iatrogenically dispersed tumor cells subsequently develop into metastatic growths is obscured by the population of existing, but undetectable, established metastases. However, up to 1% of curative cancer resections show verifiable signs of wound recurrence and many more develop signs consistent with peritoneal spreading (9-11).

Physical forces such as pressure (12-14) and shear stress (15, 16) have profound effects on cancer cell biology and are ubiquitous in the tumor environment. The increased pressure and shear stress within the circulatory system are well known. In the perioperative setting, tumor manipulation by the surgeon and surgical site irrigation subject tumor cells to pressure and shear stress, while laparoscopic surgery itself generally increases intraperitoneal pressure by 15 mmHg in order to create working space for surgery (17-19). Even relatively brief exposure to such pressures can trigger (20, 21) an intracellular signal cascade that increases the adhesive potential of suspended colon cancer (22), breast cancer (14, 23), squamous cell carcinoma (24, 25) and sarcoma (26). Although most such studies have focused on in vivo adhesion to purified matrix proteins such as collagen I (11, 27), adhesion is also increased to more physiologically relevant substrates like endothelial cells (28, 29), and murine surgical wounds (30-32), and adversely impacts survival in mouse models (32). Parallel mechanosensitive pathways involving either the cytoskeleton and paxillin or Src and phosphatidylinositol 3-kinase (PI3K) detect the pressure stimulus (22). These signaling pathways converge at an interaction between FAK and Akt1 in which Akt1 phosphorylates FAK at serines 517/601/695 to facilitate FAK activation at tyrosine 397 (33). With FAK activation, the consequent increases in b1-integrin affinity and avidity reduce the ligand threshold required for integrin-mediated cell adhesion thus providing a metastatic advantage for pressure-stimulated cells (27).

Integrin profiles are intimately tied to tumor metastasis. While the presence of specific integrins may direct the location of metastatic progression, as is the case with b3 and bone metastases in breast cancer (34), general integrin expression dictates changes in cell adhesion. In particular, decreased expression of b1 integrins is associated with the transition from epithelial to mesenchymal phenotypes that precedes the loss of cell adhesiveness necessary for pathological metastatic dissemination; the reverse process is initiated by the re-expression of these integrins as disseminated tumor cells adhere to distant organ parenchyma to create secondary tumors (35, 36).

Under normal cell physiology, FAK and Akt influence essential cell functions, such as migration (37, 38), survival, and protection against anokis (39) (a mode of cell death to which cancer cells seem somewhat more resistant), which limits the utility of drugs that inhibit their catalytic abilities. However, the dearth of FAK-Akt binding outside the context of mechanical signaling (33, 40) led us to seek to specifically target FAK-Akt interaction in order to prevent FAK activation without compromising alternative activities of either FAK or Akt1. The binding site for FAK on Akt1 seems quite large encompassing the kinase domain but not the Pleckstrin-homology or hinge regions (41). In the current investigation, we sought to determine whether the converse binding site for Akt1 on FAK might be smaller and thus more amenable to future modeling and/or manipulation. Serial truncations isolated a relatively short 33 amino acid subdomain of FAK that is sufficient for Akt1 binding.

MATERIALS AND METHODS

Materials

Human Caco-2 and murine CT-26 colon cancer cells were cultured according to American Type Culture Collection (Rockville, MD) recommendations. We obtained Lipofectamine 2000 and other transfection supplies from Invitrogen (Carlsbad, CA), Glutathione Sepharose 4B beads from GE Life Sciences (Pittsburg, PA), and Akt1 and GST antibodies from Cell Signaling Technology (Beverly, MA). All primers were purchased from Integrated DNA Technologies (Coralville, IA). QIAquick Gel Extraction, QIAprep spin Miniprep, QIAquick PCR purification and QIAfilter Plasmid Maxi kits were purchased from Qiagen (Valencia, CA).

Generation of constructs

Mammalian expression vectors pGEX-4T-1 glutathione S-transferase (GST)-FAK-NT1 and its truncations were constructed in vivo PCR of an HA-FAK(WT) plasmid template and introduced into the bacterial expression vector pGEX-4T1 (GE Healthcare, Munich, Germany) through 5'EcoRI and 3'XhoI cut sites. GST-Akt1 was generated by the same manner using a pcDNA3-myr-HA-Akt1 template (Addgene, Cambridge, MA). The HA-FAK(WT) itself, as well as the HA-FRNK (tagged at the COOH terminal), were gifts from Dr. David Schlaepfer. A similar protocol was used to generate the GFP-FAK-NT1 transient expression vector from the pEGFP-C1 vector Clontech (Mountain View, CA). Inducible GFP-FAK-NT1 expression was achieved using the pL6N2-RHS3H/ZF2-PL vector provided by the ARGENT regulated transcription retrovirus kit (now iDimerize inducible heterodimer system from Takara, Mountain View, CA).

Transfections

Caco-2 cells were plated on p100 dishes at 30 – 35% confluence one day prior to transfection. The constructed or empty plasmids were transfected into Caco-2 cells at final concentrations of 2 µg/ml plasmid and 5 µg/ml Lipofectamine 2000. Five hours after transfection, the medium was replaced with 15 ml pre-warmed Caco-2 media without antibiotics. Forty eight hours after DNA transfection, the cells were trypsinized for adhesion or pull-down experiments.

Inducible expression

A stable CT-26 cell line was generated per ARIAD-ARGENT protocols and GFP-FAK-NT1 expression was induced using 50 nM of the provided non-immunosuppressive rapalog, AP21967 with an equal volume of ethanol used as the control vehicle. Forty eight hours after induction, the cells were trypsinized for adhesion experiments.

Glutathione S-transferase pull-down

Glutathione-Sepharose 4B beads (15 µl) were washed twice in ice-cold PBS and resuspended in 400 µl PBS. Bacterial lysate containing GST-Akt1, GST-FAK-NT1 (truncations), or GST proteins (gift of Dr. J. Chen) were then added in excess and incubated with the beads for 1 hour. Glutathione-Sepharose 4B beads coupled to GST-Akt1, GST-FAK (truncations), or GST were then washed twice with PBS by centrifuge for 5 min at 500 g and incubated with nontransfected or transfected Caco-2 cell lysates (600 – 800 µg protein) overnight at 4°C. Transfected Caco-2 cells received plasmids encoding HA-WT-FAK or HA-FRNK. Nontransfected and transfected Caco-2 cell lysates were prepared in cell lysis buffer lysis buffer [50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na3VO4, 50 mM NaF, 10 mM sodium pyrophosphate, 2 mg/ml aprotinin, and 2 mg/ml leupeptin (pH 7.4)]. Following incubation, beads were washed twice with lysis buffer without SDS and protease inhibitors. Proteins were eluted with Laemmli SDS sample dilution buffer, separated by 10% SDS-PAGE, and immunoblotted with GST, Akt1 (Cell Signaling Technology, Danvers, MA), or HA monoclonal antibodies (Covance, Chantilly, VA).

Western blotting

Western blots were performed as previously described (28). Eluate from the pull-downs were resolved by SDS-polyacrylamide gel electrophoresis and transferred to Hybond ECL nitrocellulose membrane (Amersham Pharmacia Biotech, Piscataway, NJ). Membranes were blotted with specific antibodies directed against either their wild-type structures or recombinant tags with the appropriate secondary antibody coupled to horseradish peroxidase. Bands were detected with enhanced chemiluminescence (Amersham) and analyzed with a Kodak Image Station 440CF (Perkin Elmer, Boston, MA). All exposures were within the linear range.

Pressure regulation

Experimental pressure conditions were controlled using an airtight box with inlet and outlet valves, thumb screws, a pressure gauge and an O-ring for an airtight seal as previously described in detail (11, 30). The box was prewarmed to 37°C for 1 hour to minimize temperature fluctuations experienced by the cells. The gas used for pressurization was a filtered 5% CO2/95% air mixture consistent with the atmosphere in which the cells were routinely cultured. The temperature was maintained within ± 2°C and the pressure within ± 1.5 mmHg. Partial pressures of O2 and CO2 and pH do not change appreciably during pressurization (11). Cells were plated as for adhesion assays described below, and the six well plates were then placed into the prewarmed pressure box which was rapidly repressurized and monitored throughout the experiment with adjustment of the pressure if required. The pressure box was maintained in an incubator at 37°C, and control cells, incubated at ambient pressure, were maintained in the same incubator. Pressure was maintained throughout the adhesion experiment, which lasted 30 minutes.

Cell adhesion studies

100,000 Caco-2 or CT-26 cells were seeded to collagen I coated 6-well plates under ambient or increased pressure conditions for 30 minutes. Non-adherent cells were washed away, and the remainder fixed with 0.01 M NaIO4, 0.75 M lysine, 0.0375 M sodium phosphate buffer, pH 7.4, and 2% paraformaldehyde on ice for 1 hour. The adherent cells were counted microscopically in at least 20 random high power fields per well under a fluorescent microscope.

FAK1 structural analysis

Structures were obtained from the Protein Data Bank (http://www.rcsb.org): 2AL6 (42). Structures were analyzed using Pymol from DeLano Scientific (San Carlos, CA).

Statistical analysis

Results were compared by Student's unpaired t-test and considered statistically significant when P < 0.05. All experiments were done independently at least three times unless indicated otherwise. All data are expressed as mean ± S.E.M.

RESULTS

FAK-related non-kinase, 67 kDa (FRNK) is not the primary binding site by which FAK binds to Akt1 in suspended Caco-2 cells

FRNK is a segment from the COOH-terminal region of the FAK molecule, the C-terminal focal adhesion targeting domain, which functions as an endogenous FAK inhibitor by competitively binding to focal contacts while lacking catalytic capability (43). It might have been predicted that the FRNK sequence is also important in Akt-binding. We transfected either HA-tagged wild type FAK or HA-tagged FRNK plasmids into Caco-2 cells, and incubated the resulting cell lysate with Sepharose beads conjugated to GST-Akt1 before Western blotting for HA. Passage over the GST-Akt1 column enriched the resulting eluate for HA-conjugated wild type FAK while the amount of HA-conjugated FRNK was markedly reduced by this procedure (Fig. 1, 1 of 2 similar).

Figure 1 Fig. 1. FRNK alone is not sufficient to bind Akt1. Pull-down assays used GST or GST-Akt1 (prey, bottom blot) to bind HA-FRNK or HA-WT-FAK (bait, top blot) found in the cell lysate of transiently transfected Caco-2 cells. Caco-2 cells expressing HA-FRNK control showed low levels of GST-Akt1 pull-down relative to those expressing HA-WT-FAK (one of two similar). Lysate from cells transfected with HA-WT-FAK or HA-FRNK were used as references.

Transient expression of FAK-NT1 inhibits pressure-induced adhesion

In contrast, previous preliminary observations suggested that the F1 lobe of FAK (herein referred to as the NT1 region) of wild-type FAK was sufficient to pull down Akt1 (41). To test whether this interaction of NT1 with Akt1 might have biological effects, we evaluated the effect of overexpressing NT1 in Caco-2 cells on the adhesive response to increased extracellular pressure, which requires FAK-Akt1 interaction (27, 33). Caco2 cells transiently expressing the GFP-FAK-NT1 construct were therefore exposed to ambient or 15 mmHg increased pressure for 30 minutes. Pressure-induced adhesion was inhibited in the cells expressing GFP-FAK-NT1 but not in those expressing the control GFP alone (Fig. 2). Interestingly, the basal levels of adhesion were also reduced in the GFP-FAK-NT1 population.

Figure 2 Fig. 2. Transient expression of FAK-NT1 inhibits pressure-induced adhesion. Caco-2 cells transiently expressing the GFP control demonstrated increased cell adhesion after exposure to 15 mmHg pressure. Transient expression of GFP-FAK-NT1 blocked pressure-induced cell adhesion. Decreases in basal levels of cell adhesion were also seen in the cells expressing GFP-FAK-NT1 (N = 6; *P < 0.05 versus the ambient GFP control).

Inducible expression of FAK-NT1 inhibits pressure-induced adhesion

To establish the generalizability of this phenomenon, we constructed a stable murine CT26 colon cancer line which expressed GFP-FAK-NT1 on induction with rapalog, along with a control cell line which only expressed GFP on induction. Expression of the GFP-FAK-NT1 construct in CT26 cells was induced 48 hours prior to exposure to ambient or 15 mmHg increased pressure for 30 minutes. Inducing expression of the GFP-FAK-NT1 construct prevented the stimulation of cell adhesion by increased pressure. In contrast, cells in which only GFP was inducibly expressed did display increased adhesion in response to increased extracellular pressure (Fig. 3). Basal levels of adhesion were reduced in the GFP-FAK-NT1 population.

Figure 3 Fig. 3. Inducible expression of FAK-NT1 inhibits pressure-induced adhesion. As with the transient expression model, CT-26 cells inducibly expressing the GFP control showed increased cell adhesion after exposure to 15 mmHg pressure. Inducible expression of GFP-FAK-NT1 blocked pressure-induced cell adhesion, however it did cause a larger of a decrease in basal cell adhesion (N = 6; *P < 0.05 versus the ambient GFP control).

The FAK-NT1-2-2 region is sufficient to bind Akt1

Because the NT1 region of FAK is still quite large, we further truncated FAK-NT1 in an attempt to specify the region responsible for Akt1 binding. Five truncations were generated: NT1 (residues 1-126), NT1-1 (residues 1-60), NT1-2 (residues 61-126), NT1-1-2-1 (residues 61-93), and NT1-2-2 (residues 94-126) (Fig. 4A). The truncations that contained the 33 amino acids found in the NT1-2-2 truncation (NT1, NT1-2, and NT1-2-2) were able to pull down more significantly more Akt1 than the constructs that did not (GST, NT1-1, and NT1-2-1) (Fig. 4B and 4C; P < 0.05).

Figure 4 Fig. 4. Akt1/FAK binding requires the Akt1 kinase domain. (A) Recombinant GST-FAK truncations NT1 (residues 1- 126), NT1-1 (residues 1-60), NT1-2 (residues 61-126), NT1-1-2-1 (residues 61-93), and NT1-2-2 (residues 94-126) were generated to test their ability to pull down Akt1. (B, C) All truncations containing the NT1-2-2 region (NT1, NT1-2, and NT1-2-2) pulled down Akt1. The constructs that did not (GST, NT1-1, and NT1-2-1) pulled down a significantly smaller amount (N = 4; *P < 0.05 versus GST control; #P < 0.05 versus GST-NT1-1; ^P < 0.05 versus GST-NT1-2-1).

DISCUSSION

Extracellular forces like pressure and shear stress increase the binding affinity of surface integrins effectively decreasing the ligand threshold required for cell adhesion (16, 22, 40). Pressure stimulation provides dislodged tumor cells with such a metastatic advantage through a mechanism that involves elements such as FAK, Akt1, and Src. Common signaling components, these kinases are well-defined pharmacologic targets but also essential to normal cell physiology. However, the relationship between FAK and Akt1 presents an opportunity to interfere with this protein-protein interaction without compromising other catalytic potentials of these kinases. A previous preliminary study suggested that the FAK F1 lobe can by itself associate with Akt1 (41). Here we demonstrate that overexpression of this FAK domain can itself inhibit pressure-induced adhesion. While the F1 lobe itself is still quite large, serial truncations demonstrated that the interaction between FAK and Akt1 depends upon a 33 amino acid region in the C-terminal of the FAK F1 lobe. These findings support the possibility of using therapeutics modeled after this much smaller region of FAK to disrupt the pressure signaling pathway.

In our model, 15 mmHg increased extracellular pressure is achieved through the isothermic injection of a filtered 5% CO2/95% air mixture into airtight boxes; pressurization occurs at 37°C for 30 min. While CO2 was our gas of choice for its wide use in laparoscopic surgery, it should be noted that the exact effect of the type of gas is still debated: some groups show CO2 to produce less cell adhesion when compared to nitrogen and helium (44) and others show no difference (45). Ma and colleagues reported that pure CO2 pressurization, particularly at high pressures, actually decreases both the expression of adhesion molecules and the adhesive and invasive potential of colon cancer cells (46). Our results here and in vivo (31), as well as those of others (47) differ from the report by Ma and colleagues. High pressure pure carbon dioxide gassing of cell suspensions with limited buffering capacity tends to be associated with substantial acidosis (48), which is itself very toxic and may thus have effects beyond the signaling events studied here. The addition of the 95% room air has the added benefit of matching the atmosphere within most cell incubators which helps us focus on the impact of just the mechanical forces at work. Previous work from our laboratory has demonstrated that pressure stimulates adhesion even when the experiment is done in a pure nitrogen atmosphere (11).

Pressure-stimulated adhesion can be blocked and indeed survival improved in murine tumor models using either colchicine (31) or siRNA to alpha-actinin-1 (32) to interrupt cytoskeletal mechanotransduction, but the concentration of colchicine required to achieve these effects is substantially higher than that acceptable in humans, while molecular modification by siRNA techniques would be challenging in the clinical setting. Inhibitors that target FAK and Akt1 directly can also prevent pressure-stimulated adhesion in vivo (27, 33, 40). However, these important kinases have diverse cell functions and blocking either can produce substantial side effects. For instance, FAK inhibitors such as Y15 may cause peritonitis with fatal complications (49, 50). Direct Akt1 blockade reduces cell proliferation in a way that diminishes the effect of chemotherapeutic agents reliant on DNA replication (50). Furthermore, agents that act upstream of these kinases may manifest varied, bordering unpredictable, consequences. In the case of Akt, the triterpene Celastrol decreases osteosarcoma invasion through inhibition of the PI3K/Akt pathway (50, 51) yet it promotes myofibril hypertrophy by activating the Akt1/ERK1/2 pathway (52). Such examples emphasize the benefit of targeting specific protein interactions over general function.

Mechanical forces recruit FAK and Akt1 to one another to stimulate cancer cell adhesion, an interaction not common among previously described signal pathways. This force-activated FAK-Akt1 interaction is therefore an enticing target because it may have less side effects. While the apparent novelty of this FAK-AKT1 interaction may reflect the limits of our knowledge about these kinases, preventing this specific interaction seems highly likely to have less off-target effects than the global consequences of blocking all catalytic functions of either kinase. We now show that such an interaction can be blocked by the expression of FAK fragments modeled after key components involved in the binding of these two kinases.

The FAK molecule is functionally divided into the N-terminal erythrocyte band four.1-ezrin-radixin-moesin (FERM) domain (residues 35-362), the central kinase domain (residues 416-676), and the C-terminal focal adhesion targeting (FAT) domain (residues 677-1025) (53). The FERM domain both connects components of the cell membrane to the cytoskeleton and regulates FAK activity (53). The FERM domain is further divided into three lobes: F1 (residues 35-130), F2 (residues 131-255), and F3 (residues 256-362) (Fig. 3). The F2 lobe regulates FAK catalytic activity by binding the kinase domain and folding the FERM over it to physically occlude the active site (54). The F3 lobe exhibits homology with regions of other FERM domains that bind the cytoplasmic tails of b integrins and ICAM-2 when activated (42). Compared to the rest of the FERM domain, less is known about the function of the F1 lobe. However, we have consistently shown, before with the F1 lobe in its entirety and here with the FAK NT1-2-2 region (residues 94-126), that the FAK F1 lobe is involved in Akt1 binding (41). Furthermore, while the activation site of FAK (tyrosine 397) is outside this region, changes to the residues within the F1 lobe, which do not physically contact tyrosine 397, are yet capable of triggering FAK activation (42). Such an allosteric activation may reconcile the importance of the F1 lobe with the Akt1-dependent FAK activation seen with pressure stimulation. Additionally, the specificity of this relationship is demonstrated by the inability of HA-FRNK (FAK-related non-kinase) to binds Akt1. This is noteworthy as FRNK is an established FAK truncation consisting of only the FAK C-terminal FAT domain and is often used for it's ability to bind, but not phosphorylate, FAK targets (43). Specifically, FRNK binds to focal adhesion complexes. This further coincides with our findings that pressure induces FAK-Akt1 association in suspended cells which have yet to form focal adhesion complexes (28). Under this lens, we better understand the involvement of the F1 lobe in the FAK-Akt1 interaction.

The expression of the FAK-NT1 region successfully inhibited the stimulatory effect of pressure on cell adhesion in two different cell lines, using transient or stable inducible overexpression. This is consistent with previous observations showing binding between the NT1 region of FAK and Akt1 (41) as well as with the postulated necessity of FAK-Akt1 interactions for pressure-induced adhesion to occur (33). Taken together, these observations in two different model systems strongly support the conclusion that FAK interacts with Akt1 in response to pressure stimulation in vivo the FAK-NT1 region. Transient or induced overexpression of the GFP-FAK-NT1 plasmid in Caco-2 cells also decreased basal cell adhesion suggesting some tonic activity of this force-activated pathway even in the absence of increased pressure stimulation. However, FAK-NT1 is 126 amino acids in length. Such a large protein would be challenging either to dose pharmacologically or to mimic with small molecule analogs. A parallel set of studies examining Akt1 truncations was not able to narrow the Akt1 binding site for FAK down to a single small domain. The FAK binding site on Akt1 seems to span across the entirety of the Akt1 kinase domain as both N- and C-terminal based truncations of the region were equally capable of pulling down FAK (41). By contrast, our serial truncation studies of FAK were able to narrow down the region on FAK required for Akt1 binding to the 33 amino acid sequence in the FAK-NT1-2-2 truncation. Indeed, the FAK-NT1-2-2 region contains a segment that is surface accessible and may be responsible for orchestrating the Akt1 binding we observe through these pull-down assays (Fig. 5A and 5B). About one quarter the size of the larger FAK-NT1, this smaller subdomain may prove much easier to model or manipulate and may be an important target for future study.

Figure 5
Fig. 5. The FAK molecule. Cartoon (A) and surface (B) depictions of the crystal structure of the FAK FERM domain containing the F1, F2, and F3 lobes (PDB code: 2AL6). The NT1-2-2 segment and its surface accessible region are contained within the F1 lobe.

Ninety percent of cancer deaths are attributed not to the original tumor but to metastatic growths (55). Such metastasis requires many steps but one key step is the adhesion of disseminated or circulating tumor cells to a remote substrate. Exposure to increased extracellular pressure initiates changes that decrease cell motility and thus migratory potential (56). At least some of the effect of pressure on cell motility may be its influence on b1-integrin subunit phosphorylation and thus cell adhesiveness (57). When this occurs in cancer cells after dissemination, the net effect may be to promote new tumor formation through the establishment of cell-cell and cell-matrix adhesions (35).

It is noteworthy in this regard that physical forces activate a very different pathway in cancer cells that are already adherent (14, 58, 59), so that the same stimuli that promote the adhesiveness of circulating cells would not be expected to prevent the motility and invasion of tumor cells within a primary tumor. The force-activated pro-adhesive pathway targeted here has been observed in malignant cell types including colon cancers (11), squamous head and neck cancers (25), breast cancers (60), and even sarcomas (26). Thus, inhibiting this pro-metastatic signal cascade might have substantial benefits for reducing perioperative tumor dissemination and even longer term metastasis from unresectable tumors. Our results here raise the possibility that this pathway might be inhibited by interfering with FAK-Akt1 binding, in a fashion that may bypass the off-target effects common to currently available therapeutics.

Acknowledgements: This work was supported by the National Institutes of Health (RO1DK060771 to M.D.B.).

Conflict of interests: None declared.

REFERENCES

  1. Fidler IJ. Cancer biology is the foundation for therapy. Cancer Biol Ther 2005; 4: 1036-1039.
  2. Sugarbaker PH. Successful management of microscopic residual disease in large bowel cancer. Cancer Chemother Pharmacol 1999; 43 (Suppl): S15-S25.
  3. Allardyce R, Morreau P, Bagshaw P. Tumor cell distribution following laparoscopic colectomy in a porcine model. Dis Colon Rectum 1996; 39 (10 Suppl.): S47-S52.
  4. Weitz J, Koch M, Kienle P, et al. Detection of hematogenic tumor cell dissemination in patients undergoing resection of liver metastases of colorectal cancer. Ann Surg 2000; 232: 66-72.
  5. Choy A, McCulloch P. Induction of tumour cell shedding into effluent venous blood breast cancer surgery. Br J Cancer 1996; 73: 79-82.
  6. Uchikura K, Takao S, Nakajo A, et al. Intraoperative molecular detection of circulating tumor cells by reverse transcription-polymerase chain reaction in patients with biliary-pancreatic cancer is associated with hematogenous metastasis. Ann Surg Oncol 2002; 9: 364-370.
  7. Yamaguchi K, Takagi Y, Aoki S, Futamura M, Saji S. Significant detection of circulating cancer cells in the blood by reverse transcriptase-polymerase chain reaction during colorectal cancer resection. Ann Surg 2000; 232: 58-65.
  8. Hayashi N, Egami H, Kai M, Kurusu Y, Takano S, Ogawa M. No-touch isolation technique reduces intraoperative shedding of tumor cells into the portal vein during resection of colorectal cancer. Surgery 1999; 125: 369-374.
  9. Nelson H, Sargent DJ, Wieand HS, et al. Clinical outcomes of surgical therapy study. A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 2004; 350: 2050-2059.
  10. Turnbull RB, Kyle K, Watson FR, Spratt J. Cancer of the colon: the influence of the no-touch isolation technic on survival rates. Ann Surg 1967; 166: 420-427.
  11. Basson MD, Yu CF, Herden-Kirchoff O, et al. Effects of increased ambient pressure on colon cancer cell adhesion. J Cell Biochem 2000; 78: 47-61.
  12. Shiratsuchi H, Basson MD. Extracellular pressure stimulates macrophage phagocytosis by inhibiting a pathway involving FAK and ERK. Am J Physiol Cell Physiol 2004; 286: C1358-C1366.
  13. Thubrikar MJ, Robicsek F. Pressure-induced arterial wall stress and atherosclerosis. Ann Thorac Surg 1995; 59: 1594-1603.
  14. Basson MD, Zeng B, Downey C, Sirivelu MP, Tepe JJ. Increased extracellular pressure stimulates tumor proliferation by a mechanosensitive calcium channel and PKC-beta. Mol Oncol 2015; 9: 513-526.
  15. Huynh KC, Gyenes M, Nguyen TH, Vo TV, Stoldt VR. Impact of shear stress on Src and focal adhesion kinase phosphorylation in fibrinogen-adherent platelets. Blood Coagul Fibrinolysis 2017; 28: 279-285.
  16. Thamilselvan V, Patel A, van der Voort van Zyp J, Basson MD. Colon cancer cell adhesion in response to Src kinase activation and actin-cytoskeleton by non-laminar shear stress. J Cell Biochem 2004; 92: 361-371.
  17. O'Rourke N, Kodali BS. Laparoscopic surgery during pregnancy. Curr Opin Anaesthesiol 2006; 19: 254-259.
  18. Daskalakis M, Scheffel O, Weiner RA. High flow insufflation for the maintenance of the pneumoperitoneum during bariatric surgery. Obes Facts 2009; 2 (Suppl. 1): 37-40.
  19. Neudecker J, Sauerland S, Neugebauer E, et al. The European Association for Endoscopic Surgery clinical practice guideline on the pneumoperitoneum for laparoscopic surgery. Surg Endosc 2002; 16: 1121-1143.
  20. Thamilselvan V, Basson MD. The role of the cytoskeleton in differentially regulating pressure-mediated effects on malignant colonocyte focal adhesion signaling and cell adhesion. Carcinogenesis 2005; 26: 1687-1697.
  21. Craig DH, Haimovich B, Basson MD. Alpha-actinin-1 phosphorylation modulates pressure-induced colon cancer cell adhesion through regulation of focal adhesion kinase-Src interaction. Am J Physiol Cell Physiol 2007; 293: C1862-C1874.
  22. Basson MD. An intracellular signal pathway that regulates cancer cell adhesion in response to extracellular forces. Cancer Res 2008; 68: 2-4.
  23. Less JR, Posner MC, Boucher Y, Borochovitz D, Wolmark N, Jain RK. Interstitial hypertension in human breast and colorectal tumors. Cancer Res 1992; 52: 6371-6374.
  24. Gutmann R, Leunig M, Feyh J, et al. Interstitial hypertension in head and neck tumors in patients: correlation with tumor size. Cancer Res 1992; 52: 1993-1995.
  25. Conway WC, Van der Voort van Zyp J, Thamilselvan V, Walsh MF, Crowe DL, Basson MD. Paxillin modulates squamous cancer cell adhesion and is important in pressure-augmented adhesion. J Cell Biochem 2006; 98: 1507-1516.
  26. Perry BC, Wang S, Basson MD. Extracellular pressure stimulates adhesion of sarcoma cells via activation of focal adhesion kinase and Akt. Am J Surg 2010; 200: 610-614.
  27. Thamilselvan V, Craig DH, Basson MD. FAK association with multiple signal proteins mediates pressure-induced colon cancer cell adhesion via a Src-dependent PI3K/Akt pathway. FASEB J 2007; 21: 1730-1741.
  28. Thamilselvan V, Basson MD. Pressure activates colon cancer cell adhesion by inside-out focal adhesion complex and actin cytoskeletal signaling. Gastroenterology 2004; 126: 8-18.
  29. van Zyp J, Conway WC, Craig DH, van Zyp N, Thamilselvan V, Basson MD. Extracellular pressure stimulates tumor cell adhesion in vitro by paxillin activation. Cancer Biol Ther 2006; 5: 1169-1178.
  30. van der Voort van Zyp J, Conway WC, Thamilselvan V, Polin L, Basson MD. Divalent cations influence colon cancer cell adhesion in a murine transplantable tumor model. Am J Surg 2005; 190: 701-707.
  31. Craig DH, Owen CR, Conway WC, Walsh MF, Downey C, Basson MD. Colchicine inhibits pressure-induced tumor cell implantation within surgical wounds and enhances tumor-free survival in mice. J Clin Invest 2008; 118: 3170-3180.
  32. Craig DH, Downey C, Basson MD. SiRNA-mediated reduction of alpha-actinin-1 inhibits pressure-induced murine tumor cell wound implantation and enhances tumor-free survival. Neoplasia 2008; 10: 217-222.
  33. Wang S, Basson MD. Akt directly regulates focal adhesion kinase through association and serine phosphorylation: implication for pressure-induced colon cancer metastasis. Am J Physiol Cell Physiol 2011; 300: C657-C670.
  34. Pecheur I, Peyruchaud O, Serre CM, et al. Integrin alpha(v)beta3 expression confers on tumor cells a greater propensity to metastasize to bone. FASEB J 2002; 16: 1266-1268.
  35. Neal CL, McKeithen D, Odero-Marah VA. Snail negatively regulates cell adhesion to extracellular matrix and integrin expression via the MAPK pathway in prostate cancer cells. Cell Adh Migr 2011; 5: 249-257.
  36. Wang JH, Manning BJ, Wu QD, Blankson S, Bouchier-Hayes D, Redmond HP. Endotoxin/lipopolysaccharide activates NF-kappa B and enhances tumor cell adhesion and invasion through a beta 1 integrin-dependent mechanism. J Immunol 2003; 170: 795-804.
  37. Walsh MF, Ampasala DR, Hatfield J, et al. Transforming growth factor-beta stimulates intestinal epithelial focal adhesion kinase synthesis via Smad- and p38-dependent mechanisms. Am J Pathol 2008; 173: 385-399.
  38. Gayer CP, Chaturvedi LS, Wang S, Alston B, Flanigan TL, Basson MD. Delineating the signals by which repetitive deformation stimulates intestinal epithelial migration across fibronectin. Am J Physiol Gastrointest Liver Physiol 2009; 296: G876-G885.
  39. Walsh MF, Thamilselvan V, Grotelueschen R, Farhana L, Basson M. Absence of adhesion triggers differential FAK and SAPKp38 signals in SW620 human colon cancer cells that may inhibit adhesiveness and lead to cell death. Cell Physiol Biochem 2003; 13: 135-146.
  40. Craig DH, Gayer CP, Schaubert KL, et al. Increased extracellular pressure enhances cancer cell integrin-binding affinity through phosphorylation of beta1-integrin at threonine 788/789. Am J Physiol Cell Physiol 2009; 296: C193-C204.
  41. Basson MD, Zeng B, Wang S. Akt1 binds focal adhesion kinase via the Akt1 kinase domain independently of the pleckstrin homology domain. J Physiol Pharmacol 2015; 66: 701-709.
  42. Ceccarelli DF, Song HK, Poy F, Schaller MD, Eck MJ. Crystal structure of the FERM domain of focal adhesion kinase. J Biol Chem 2006; 281: 252-259.
  43. Sieg DJ, Hauck CR, Schlaepfer DD. Required role of focal adhesion kinase (FAK) for integrin-stimulated cell migration. J Cell Sci 1999; 112: 2677-2691.
  44. Tan BJ, Dy JS, Chiu PY, et al. Effects of pneumoperitoneal gases and pressures on transitional cell carcinoma adhesion, growth, apoptosis and necrosis: an in vitro study. J Urol 2005; 174: 1463-1467.
  45. Ludemann, Watson, Smith, Ellis, Jamieson. Tumor implantation during laparoscopy using different insufflation gases - an experimental study using cultured cancer cells. Minim Invasive Ther Allied Technol 2003; 12: 310-314.
  46. Ma JJ, Feng B, Zhang Y, et al. Higher CO2-insufflation pressure inhibits the expression of adhesion molecules and the invasion potential of colon cancer cells. World J Gastroenterol 2009; 15: 2714-2722.
  47. Zhang X, Guo X, Zhang A, Wang Y, Zhao J. Influence of carbon dioxide pneumoperitoneum environment on adhesion and metastasis of a human ovarian cancer cell line. Surg Endosc 2009; 23: 108-112.
  48. Wildbrett P, Oh A, Naundorf D, Volk T, Jacobi CA. Impact of laparoscopic gases on peritoneal microenvironment and essential parameters of cell function. Surg Endosc 2003; 17: 78-82.
  49. Golubovskaya V, Curtin L, Groman A, Sexton S, Cance WG. in vivo toxicity, metabolism and pharmacokinetic properties of FAK inhibitor 14 or Y15 (1, 2, 4, 5-benzenetetramine tetrahydrochloride). Arch Toxicol 2015; 89: 1095-1101.
  50. Galvez-Peralta M, Flatten KS, Loegering DA, et al. Context-dependent antagonism between Akt inhibitors and topoisomerase poisons. Mol Pharmacol 2014; 85: 723-734.
  51. Yu X, Wang Q, Zhou X, et al. Celastrol negatively regulates cell invasion and migration ability of human osteosarcoma via downregulation of the PI3K/Akt/NF-kappaB signaling pathway in vitro. Oncol Lett 2016; 12: 3423-3428.
  52. Gwag T, Park K, Park J, Lee JH, Nikawa T, Choi I. Celastrol overcomes HSP72 gene silencing-mediated muscle atrophy and induces myofiber preservation. J Physiol Pharmacol 2015; 66: 273-283.
  53. Parsons JT. Focal adhesion kinase: the first ten years. J Cell Sci 2003; 116: 1409-1416.
  54. Lietha D, Cai X, Ceccarelli DF, Li Y, Schaller MD, Eck MJ. Structural basis for the autoinhibition of focal adhesion kinase. Cell 2007; 129: 1177-1187.
  55. Sporn MB. The war on cancer. Lancet 1996; 347: 1377-1381.
  56. Kovalenko PL, Flanigan TL, Chaturvedi L, Basson MD. Influence of defunctionalization and mechanical forces on intestinal epithelial wound healing. Am J Physiol Gastrointest Liver Physiol 2012; 303: G1134-G1143.
  57. Flanigan TL, Craig DH, Gayer CP, Basson MD. The effects of increased extracellular deformation, pressure, and integrin phosphorylation on fibroblast migration. J Surg Res 2009; 156: 103-109.
  58. Walsh MF, Woo RK, Gomez R, Basson MD. Extracellular pressure stimulates colon cancer cell proliferation via a mechanism requiring PKC and tyrosine kinase signals. Cell Prolif 2004; 37: 427-441.
  59. Downey C, Craig DH, Basson MD. Isoform-specific modulation of pressure-stimulated cancer cell proliferation and adhesion by alpha-actinin. Am J Surg 2011; 202: 520-523.
  60. Downey C, Alwan K, Thamilselvan V, et al. Pressure stimulates breast cancer cell adhesion independently of cell cycle and apoptosis regulatory protein (CARP)-1 regulation of focal adhesion kinase. Am J Surg 2006; 192: 631-635.
R e c e i v e d : April 30, 2017
A c c e p t e d : May 29, 2017
Author’s address: Prof. Marc D. Basson, Senior Associate Dean for Medicine and Research, School of Medicine and Health Sciences, University of North Dakota, 1301 North Columbia Road, Grand Forks, ND 58202, U.S.A.; e-mail: marc.basson@med.und.edu