Background
Colorectal carcinoma (CRC) is one of the leading causes of mortality in Western countries. The number of new cases in the United States was predicted to reach 103,170 in the year 2012, with 51,690 deaths expected [
1]. CRC frequently metastasizes into a systemic disease, often invading the lymph nodes and other visceral organs. The occurrence of metastases due to tumour progression is responsible for the vast majority of cancer-related deaths.
The progression of normal colonic epithelium into adenoma and later, into malignant adenocarcinoma, is associated with diverse genetic and epigenetic alterations [
2,
3]. However, the vast majority of CRCs share a well characterized sequence of inactivating mutations in tumor suppressor genes and gain-of-function alterations in oncogenes [
2,
4]. The molecular pathways perturbed by these key genetic changes are relatively well understood. In contrast, there is limited information about the molecular changes that give rise to the subsequent stages of colorectal progression, from carcinoma
in situ, to invasive carcinoma, to lymph node and visceral metastasis [
5]. Understanding how CRC metastases develop is critical for the ultimate control of this cancer.
Metastasis is a complex process that begins with the invasion of cancer cells into the surrounding tissues. The acquisition of enhanced cell motility and invasiveness enable a tumor cell to break away from the primary site, enter and exit the circulation, and successfully establish a metastatic colony [
6]. Cancer cells acquire migratory and invasive properties through disruption of cell-cell junctions, changes in focal adhesion complexes, and extensive reorganization of the actin cytoskeleton [
7,
8]. In mammalian cells, the generation of actin-based dynamic motile structures is regulated by the Rho family of small GTPases, of which the best studied members are Cdc42, Rac1 and RhoA. RhoA is involved in the maintenance of actin stress fibers and focal adhesions, Rac1 in the formation of lamellipodia and membrane ruffles and Cdc42 in the formation of filopodia [
9,
10]. The coordinated activation of these GTPases is thought to be required for efficient migration [
11,
12].
Mutations in Rho GTPases are rarely the cause of cancer; however the increased expression or activation of Rho GTPases is associated with a variety of cancer types [
13], with enhanced invasion and metastatic potential, [
14‐
18] and disease progression [
19,
20]. RhoA is overexpressed in colon carcinoma [
19]. In contrast, Rac1 levels are normal in colorectal tumors, but truncated mutants of adenomatous polyposis coli (APC), which is the cause of sporadic and familial colorectal tumors [
21], stimulate the activity of Asef, a Rac-specific guanine exchange factor [
22]. The finding that Rho GTPase overexpression or hyperactivity, rather than activating mutations, are involved in human cancer suggests their regulatory proteins have a prominent role in deregulated signaling in cancer [
23].
Rho GTPases cycle between inactive GDP-bound and active GTP-bound conformations, which enables them to function as binary switches. GTP-bound Rho proteins interact with multiple downstream effectors to elicit a variety of cellular responses including cytoskeletal reorganization, gene transcription, cell cycle regulation, and vesicular traffic [
24]. The Rho GTPase cycle is tightly regulated by three classes of proteins. Guanine nucleotide exchange factors (GEFs) activate GTPases by promoting the exchange of GTP for GDP, whereas GTPase activating proteins (GAPs) inactivate Rho proteins by enhancing their intrinsic GTPase activity. Guanine nucleotide dissociation inhibitors (GDIs) prevent the dissociation of GDP and maintain the GTPases in an inactive state. GDIs also sequester Rho GTPases as soluble cytosolic complexes in which the C-terminal membrane-targeting lipid moiety of the GTPase is prevented from interacting with membranes [
25,
26]. Since the vast majority of Rho GTPase protein exists in a biologically inactive cytosolic complex with RhoGDI, this is a major point of regulation of Rho GTPase activity and function. A thorough understanding of the mechanisms that regulate Rho GTPases is therefore paramount for understanding deregulated Rho GTPase signaling in cancer.
Diacylglycerol kinases (DGKs) phosphorylate the lipid second messenger diacylglycerol (DAG) to yield PA. There are ten mammalian DGK isoforms (α, β, γ, etc.), each with specific patterns of expression, localization within cells and distinct structural domains [
27]. Our previous studies demonstrated that DGKζ regulates both Rac1 and RhoA activation [
28,
29]. DGKζ forms two independent multiprotein signaling complexes with Rac1 and RhoA that function as selective RhoGDI dissociation factors. In DGKζ-deficient fibroblasts, Rac1 and RhoA activation are decreased and cell migration is significantly reduced [
28,
29]. In light of these findings, we hypothesized that increased DGKζ expression in cells would lead to enhanced Rho GTPase activity and increased migratory potential. DGKζ mRNA is highly expressed in colon cancer tissue relative to normal colonic epithelium [
30]. To investigate potential roles of DGKζ in colon cancer metastasis, we used a cellular, isogenic model of human CRC metastatic transition. The SW480 and SW620 cell lines were established from biopsies taken at different intervals from the same 50-year-old CRC male patient [
31]. SW480 cells derive from the primary tumor, a poorly differentiated (grade 4) CRC invading the muscularis propria. SW620 cells derive from a lymph node metastasis taken from the same individual six months later, when recurrent cancer with liver and mesenteric lymph node metastases was discovered.
Gene expression profile data available in the National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO) repository show the DGKζ transcript is increased in the metastatic SW620 cell line relative to the SW480 primary tumor line [
5]. Here, we demonstrate that increased DGKζ protein levels in SW620 cells are associated with increased Rho GTPase activity and downstream signaling. Silencing of DGKζ expression in SW620 cells decreased Rac1 and RhoA activity and attenuated cell invasion. DGKζ silencing also attenuated the invasiveness of PC-3 prostate cancer and MDA-MB-231 breast cancer cells. Collectively, our findings suggest interfering with DGKζ function or expression may be a potential route to block the invasiveness of metastatic cancer cells.
Methods
Cell lines and culture conditions
Human colorectal tumor cell lines SW480 (ATCC CCL-227) and SW620 (ATCC CCL-228), prostate cancer cell line PC-3 (ATCC CRL-1435), and breast cancer cell line MDA-MB-231 (ATCC HTB-26) were obtained from the American Type Culture Collection (ATCC, Manassas, VA). The cells were verified according to ATCC Technical Bulletin No. 8 (2008) and grown at 37°C, 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin.
Antibodies
Affinity-purified antibody to the N-terminus of DGKζ was prepared from DGKζ antisera and has been thoroughly characterized [
28,
32‐
34]. Monoclonal Rac1 antibody (Catalogue number: 610650) was purchased from BD Biosciences (San Jose, CA). A polyclonal antibody to PAK1 (Catalogue number: 2602) was purchased from Cell Signaling Technologies (Danvers, MA). Anti-pPAK1 was a gift from Dr. Jonathan Chernoff (Fox Chase Cancer Center, Philadelphia, PA) [
28,
35]. Monoclonal anti α-tubulin antibody (Catalogue number: T5168) was purchased from Sigma-Aldrich (St. Louis, MO). Horseradish peroxidase-conjugated anti-rabbit (Catalogue number: 711-035-152) and anti-mouse (Catalogue number: 715-035-150) secondary antibodies were from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).
Establishment of DGKζ-knockdown SW620 and PC-3 cell lines
A lentiviral vector containing a small hairpin RNA (shRNA) construct targeted to human DGKζ gene (Catalog no. RHS3979-9569052) and a pLKO.1 empty lentiviral vector were purchased from Open Biosystems. The empty pLKO.1 vector (Catalog #RHS4080) contains a 18 bp stuffer sequence between the AgeI and EcoRI restriction sites. The shRNA oligonucleotides (oligo ID: TRCN0000000668, Open Biosystems) corresponding to the sequence on human DGKζ gene are: sense, 5′ TCG CAC AGG ATG AGA TTT ATA 3′; antisense, 5′ TAT AAA TCT CAT CCT GTG CGA 3′. Ultra-pure plasmid DNAs were prepared according to the manufacturer’s protocol. To generate stable knockdown cell lines, SW620 cells were transfected with the shRNA vector using FuGENE 6 Transfection Reagent (Roche-applied-science). After transfection, cells were incubated for 24 h. Transfectants were then selected with 7 μg/ml puromycin (Cellgro Catalog no. 61-385-RA). After two weeks, the stable clones were transferred to 96-well plates using sterile cloning discs (Bel-Art Products), grown until confluent, and then transferred to 60 mm cell plates. DGKζ levels in various clones were analyzed by immunoblotting. Clones with DGKζ protein levels that were substantially reduced compared to the controls were selected and maintained in medium containing 7 μg/ml puromycin. SW620 cells stably transfected with pLKO.1 empty vector were used as a control. DGKζ knockdown PC3 cell lines were generated in the same manner using Attractene transfection reagent to transfect the cells and 2.2ug/ml puromycin for selection.
Lentiviral knockdown of DGKζ expression in MDA-MB-231 cells
A set of 3 lentiviral vectors containing shRNA targeted to the DGKζ gene (Thermoscientific; Catalogue no. RHS4531-EG8525) were used to generate lentivirus using the second generation packaging plasmids pMD2.G and psPAX2 from Addgene. MBA-MD-231 cells were infected with mixture of the 3 lentivirus and incubated for 30 hours at 37°C with 5% CO2 before being used in the invasion assays or extracted for western analysis.
Rac1 activity assay
Levels of active Rac1 and RhoA were measured using a GST-PAK1 PBD and GST-Rhotekin RBD pull-down assay, respectively [
36]. Cells were immediately harvested in chilled lysis buffer (50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 20 mM MgCl
2, and protease inhibitors). Lysates were centrifuged at 12,000 ×
g for 5 min. Equivalent amounts of protein were incubated with GST-PBD or -RBD beads for 30 min at 4°C. The beads were washed with lysis buffer, boiled in reducing sample buffer, and eluted proteins assayed for bound Rac1 or RhoA by immunoblotting.
Western blot
Cells were lysed in an ice-cold lysis buffer (50 mM Tris–HCl, pH7.5, 150 mM NaCl, 50 mM MgCl2, 1% Triton X-100, 1 μg/ml antipain, 1 μg/ml pepstatin, 1 μg/ml leupeptin, 0.5 mM AEBSF, and 1 mM benzamidine hydrochloride). Cellular debris was removed by centrifugation (14,000 × g for 10 min at 4°C). Total protein concentration of the supernatants was determined using a colorimetric assay method (Bio-Rad). 100 μg of total protein from each sample was resolved by SDS-PAGE, transferred onto PVDF membrane (Millipore), immunoblotted with the affinity-purified polyclonal antibodies to DGKζ (1:100) and horseradish peroxidase-conjugated goat anti-rabbit secondary antibodies (1:800), and detected using enhanced chemiluminescence (Pierce Biotechnology). Differences in protein loading were monitored by probing membranes with monoclonal anti-α-Tubulin antibody.
Invasion assay
Cellular invasive ability was evaluated using Corning 6.5 mm Transwell inserts (8 um pore size, 24 well plate, Fisher Scientific). For the SW620 and SW480 cell lines, the upper surface of the inserts was coated with 100 ul of 500 ug/ml Matrigel and the underside was treated with either 15 ug/ml (SW480 versus SW620) or 100 ug/ml (vector control versus shRNA) collagen type I. The cells were serum starved for 24 hours in DMEM containing 0.25% FBS, then re-suspended in DMEM/0.25% FBS/20 mM HEPES [pH 7.5] and seeded at 50,000 cells per insert. The medium in the lower chamber consisted of 600 ul DMEM/20% FBS/20 mM HEPES [pH 7.5], and 10 ug/ml collagen type I as chemoattractants. The cells were incubated at 37°C in a humidified atmosphere containing 5% CO2 for approximately 70 hours. The PC-3 cell lines were starved in serum-free DMEM, then resuspended in DMEM/0.1% FBS/20 mM HEPES [pH 7.5], and 25,000 cells each were seeded on inserts coated with 50 ul of 2 mg/ml Matrigel on the upper surface and 15 ug/ml collagen type I on the lower surface. The media in the lower chamber was the same as for the SW620 and SW480 lines. MDA-MB-231 cells were seeded at 25,000 cells per insert in serum-free DMEM/20 mM HEPES [pH 7.5] on inserts coated with 50 ul of 1 mg/ml Matrigel on the upper surface and 15ug/ml collagen type I on the lower surface. The lower chamber contained DMEM/10% FBS/10 ug/ml collagen type I. Both the PC3 and the MDA-MB-231assays were incubated for 24 hrs. Following incubation, the Matrigel was removed from the upper chambers using a cotton swab and the cells were fixed with 4% paraformaldehyde in PBS for 10 minutes, permeabilized with 0.5% Triton X-100 in PBS for 15 minutes, and stained with 30 ug/ml propidium iodide in PBS with 0.03% Triton X-100 for 6 hours. To compare the invasiveness of the cell lines, the inserts were placed on 24 × 50 mm coverslips and imaged on a Zeiss Observer D1 microscope fitted with a 10× objective. For each insert, five fields of view were imaged in a cross pattern and the number of invading cells counted. The counts were then averaged to obtain an invasive index. For each invasion assay plate, the invasive index of the inserts was normalized to the vector control cell lines (or to the SW620 cell line when compared to the SW480 line). Two control SW620/vector and two SW620/shRNA knock down lines were compared as well as three PC-3/vector and three PC-3/shRNA lines. Lentiviral shRNA infected verses vector infected MDA-MB-231 cells were also compared and the average reduction in DGKζ expression following infection was determined by western analysis.
Discussion
The development of metastatic tumors is a major cause of death in many human cancers. CRC progresses from adenoma to malignant adenocarcinoma to invasive carcinoma, and finally, to metastatic cancer. The origin of the SW480 and SW620 cell lines from the spontaneous progression of a human CRC in a single patient makes this a useful system for the analysis of gene expression changes during the transition from invasive carcinoma to metastasis. This isogenic, cellular model of CRC has been extensively validated and several studies showed SW620 cells are more invasive than SW480 cells
in vitro[
38,
39]. In this study, we documented a ~3-fold increase in the level of DGKζ protein in SW620 cells, as compared to SW480 cells. Silencing of DGKζ expression by ~50% reduced the invasiveness of SW620 cells, suggesting DGKζ contributes significantly to the increased motility of this cancer cell line. Moreover, silencing DGKζ expression in PC-3 and MDA-MB-231 cells also lead to reductions in their invasiveness. Taken together, these findings strongly suggest DGKζ contributes to the overall invasive potential of SW620, PC-3 and MDA-MB-231 cells.
The significance of these findings relates to our previous work, which established DGKζ as a critical regulator of both Rac1 and RhoA activity [
28,
29]. In the former case, we showed DGKζ-derived PA activates PAK1, which phosphorylates RhoGDI, allowing for the release and subsequent Rac1 activation. Mouse embryonic fibroblasts deficient in DGKζ have reduced Rac1 activity and reduced Rac1-related structures such as lamellipodia and membrane ruffles [
28]. Consistent with these findings, knockdown of DGKζ expression in SW620 cells significantly reduced Rac1 activity. Rac1 protein levels remained constant however, suggesting DGKζ acts primarily at the level of Rac1 activation. In DGKζ-null fibroblasts, even the complete absence of DGKζ only decreased Rac1 activity by ~ 50% suggesting other mechanisms contribute to Rac1 activation. Indeed, at least two additional DGK isoforms, DGKζ and DGKζ reportedly contribute to the regulation of Rac1 activation. DGKζ-dependent activation of atypical PKCζ/ζ mediates the release of Rac from RhoGDI in epithelial cells in response to hepatocyte growth factor [
41,
42], while DGKζ acts as an upstream suppressor of Rac1 activity in fibroblasts [
43]. However, the polysomal mRNA expression of DGKζ or DGKζ was not substantially different in SW480 and SW620 cells and therefore the increased migration of SW620 cells is not likely due to changes in the expression of these isoforms.
RhoA activity was increased approximately 3-fold in SW620 cells compared to SW480 cells. Furthermore, there was a comparable increase in both RhoA and DGKζ expression. Since DGKζ is required for efficient RhoA activation [
29], the combination of increased DGKζ and RhoA expression likely accounts for the increased RhoA activity in SW620 cells. However, since the level of RhoA activity but not protein was decreased by DGKζ silencing, it appears unlikely that DGKζ directly regulates RhoA expression. Thus, our findings in SW620 cells are consistent with our previous studies in mouse embryonic fibroblasts, which indicated that DGKζ regulates RhoA activity [
29].
Activating mutations in Rho GTPases are rarely detected in human cancers. More frequently however, overexpression and/or hyperactivation of Rho proteins contribute to tumor progression and metastasis [
19,
20,
44‐
48]. One study found that Rac1 plays a key role in the progression of CRC
in vivo: decreased Rac1 expression blocked tumor formation in an orthotopic model of colorectal adenocarcinoma, whereas its overexpression in SW620 cells accelerated colorectal adenocarcinoma progression when the cells were injected into athymic nude mice [
49]. In another study, RhoA activity correlated with lymph node metastasis in human colorectal cancer. More active RhoA in tumors with lymph node involvement than in those that did not metastasize suggests increased RhoA function is associated with enhanced tumor cell motility [
46]. Together, these findings suggest decreasing Rac1 or RhoA expression, or alternatively, interfering with their ability to achieve or maintain the active GTP-bound state [
50], is a viable strategy to reduce CRC progression and metastasis. The results presented herein suggest decreasing DGKζ expression or function is a potential route to reducing Rac1 and RhoA activity and the migratory ability of colon cancer cells. This strategy may be beneficial not only in cancers where DGKζ is overexpressed, but possibly also in cases where DGKζ is expressed at normal levels but Rac1 or RhoA are overexpressed or hyperactive.
Analysis of data deposited in the Oncomine database reveals DGKζ mRNA is highly expressed in several different colon cancer cell lines and in colon cancer tissue relative to normal colonic epithelium [
30,
51]. Moreover, DGKζ expression in CRC is high in comparison with other cancer types [
52]. Thus, DGKζ and its downstream signaling pathways may be important factors influencing colon cancer progression. However, a limitation of our studies is the lack of correlative clinical data showing the DGKζ protein level is elevated in metastatic cancer. Thus, it will be important to validate our findings by comparing DGKζ protein levels in primary and metastatic tumor specimens. Moreover, the effect of silencing DGKζ expression on the
in vivo metastatic potential of tumor cells with elevated DGKζ levels or high Rho GTPase activity remains to be investigated.
Finding that DGKζ is upregulated in other metastatic cancers would suggest interfering with its function might allow for a more general role in inhibiting tumor cell motility and invasion. DGKζ is also overexpressed several-fold in a variety of breast carcinoma and breast adenocarcinoma cell lines [
53]. We found that silencing DGKζ expression in highly metastatic MDA-MB-231 cells decreased their invasiveness, suggesting DGKζ signaling also plays a role in the overall invasive potential of these cells. Similar results were obtained with PC-3 prostate cancer cells. Therefore, targeting DGKζ function in these cancers as well may provide new avenues for therapeutic strategies.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
KC, KM, RA and TN performed the experiments and analyzed the data. SG prepared the manuscript. All authors approved the final version of the manuscript.