Inactivation or loss of TTP promotes invasion in head and neck cancer via transcript stabilization and secretion of MMP9, MMP2, and IL-6

Abstract

Purpose: Invasion is the critical step in progression of a precancerous lesion to squamous cell carcinoma of the head and neck (HNSCC). Invasion is regulated by multiple proinflammatory mediators. Tristetraprolin (TTP) is an mRNA-degrading protein that regulates multiple proinflammatory mediators. TTP may serve as an excellent treatment target. Rap1 is a ras-like oncoprotein that induces critical signaling pathways. In this study, the role of rap1 in TTP-mediated invasion was investigated.

Experimental design: Using complementary approaches, we modulated TTP and altered expression of interleukin (IL)-6 and matrix metalloproteinase (MMP) 2/9, which were quantified by ELISA and zymogram. Invasion was evaluated in vitro using the oral-cancer-equivalent (OCE) three-dimensional model and in vivo in the chick chorioallantoic membrane (CAM). The role of rap1 and p38 were established using knockdown strategies.

Results: Downregulation of TTP significantly increased invasion via secretion of MMP9/2 and IL-6. In the novel OCE and CAM invasion models of HNSCC, cells with downregulated TTP destroyed the basement membrane to invade the underlying connective tissue. Rap1 induces p38 mitogen-activated protein kinase (p38)-mediated inactivation of TTP. Inactive TTP enhances transcript stability via binding to the 3'-untranslated region (UTR). High IL-6 and MMP9 are prognostic for poor clinical outcomes in patients with HNSCC.

Conclusions: Targeting the rap1-p38-TTP cascade is an attractive novel treatment strategy in HNSCC to concurrently suppress multiple mediators of invasion.

Figure 1. IL-6 and MMP9 are predictive of poor outcome in SCCHN

Immunohistochemistry was performed on tissue sections of a human SCCHN tissue microarray. Interactions with IL-6 and MMP9 were determined with a COX interaction model. Patient groups are indicated by color lines. Low MMP9 and low IL-6 staining intensity (Red, n=14); Low MMP9 and high IL-6 (Yellow, n=8); High MMP9 and low IL- (Blue, n=6); High MMP9 and high IL-6 (Green, n=7). Subjects who did not experience the events were censored (p=0.020).

Figure 2. Downregulation of TTP promotes invasion

(A) UM-SCC-1 cells transfected with NT or siTTP were analyzed for invasion and migration. (B) Whole cell lysates were immunobloted with TTP and actin. Signal intensity was quantified as arbitrary densitometric unit (DU) by Image J software, normalized to actin and presented as percent of NT. (C) Whole cell lysates from UM-SCC-1 cells stably transduced with shControl (shC) and shTTP were immunoblotted with TTP and actin, as a loading control. Signal intensity (DU) was quantified, normalized to actin and presented as percent of shC. (D) Left Panel: Hematoxylin and eosin staining of UM-SCC-1-shC and -shTTP cells grown as an OCE. Black arrows depict invasive islands, which are more visible at the higher magnification (right). Middle panel: Quantification of total number of invasive islands. Right panel: Quantification of total number of invaded cells, normalized to total cell number. Data are representative of 2 independent experiments with two replicates in each experiment. (E) Left panel: UM-SCC-1shC and -shTTP CAM sections were stained with hematoxylin and eosin. Images were taken under bright-field and fluorescence (GFP) at 20X and merged to visualize cells invading past the basement membrane (BM - white arrows). Right panel: Quantification of invasive islands. Results are representative of 3 independent experiments. (F) CAM sections were immunostained for collagen IV to identify the BM (yellow arrows). Images were taken at 20X under fluorescence and merged (green-tumor cells, red-Col IV, blue-DAPI). White arrows show invading tumor cells (also Fig S3). (*p<0.01)

Figure 3. Downregulation of TTP promotes invasion via IL-6, MMP9 and MMP2

(A) UM-SCC-1 cells were transiently transfected with NT or siTTP. Gelatin zymography (left panel) was used to quantify MMP9/2 in conditioned medium and expressed as percent of corresponding control (middle panel). TTP knockdown was verified (right panel). Data are representative of 3 independent experiments. (B) Whole-cell lysates from UM-SCC-1-shTTP cells transiently transfected with siIL-6, siMMP9, siMMP2 or NT were immunobloted with IL-6, MMP9, or MMP2 and actin antibodies. Signal intensity (DU) was quantified, normalized to actin, and expressed as percent of control. (C) Invasion of UM-SCC-1-shTTP cells transiently transfected with siIL-6, siMMP9, siMMP2 or NT was quantified and expressed as percent of NT. Data are representative of two independent experiments with three replicates in each experiment. (*p<0.01)

Figure 4. TTP regulates the stability of transcripts of IL-6, MMP9 and MMP2 via the 3′-UTR

(A) Left three panels: IL-6, MMP9 and MMP2 mRNA levels were quantified by Q-RT-PCR from UM-SCC-1 cells transfected with NT and siTTP. Data are representative of two independent experiments with three replicates in each experiment. Right Panel: Whole-cell lysates were immunoblotted with TTP and actin. Signal intensity (DU) was quantified, normalized to actin, and expressed as percent of NT. (B) Left three panels: UM-SCC-1shC or shTTP cells were treated with actinomycin D followed by RNA isolation and quantification via Q-RT-PCR. Right panel: Whole cell lysates were immunoblotted with TTP and actin. Signal intensity (DU) was quantified, normalized to actin and expressed as percent of shC. (C) Left three panels: UM-SCC-1 cells transfected with NT or siTTP were co-transfected with a luciferase reporter construct containing the 3′-UTR of IL-6, MMP9 or MMP2 and a Renilla construct for normalization. Data are representative of three independent experiments with three replicates in each experiment. Right panel: Whole cell lysates were immunoblotted with TTP and actin antibodies. Signal intensity (DU) was quantified, normalized to actin, and expressed as percent of NT. (*p<0.01)

Figure 5. Rap1B induces p38-mediated phosphorylation of TTP, which promotes secretion of pro-inflammatory mediators

(A) Whole cell lysates from UM-SCC-1 cells treated with IL-1β were immunoblotted with phospho-p38 (pp38), p38, rap1B and actin. Active GTP bound rap1B (rap1B_GTP) was retrieved by a ralGDS pull-down assay. p38 was used for normalization of pp38 and rap1B was used for normalization of rap1B_GTP. Signal intensity (DU) was quantified, normalized to its corresponding loading control and expressed as percent of time zero. Data are representative of three independent experiments. (B) Cells were transfected with two siRNAs targeting rap1B (si6 and si7) or NT and stimulated with IL-1β or PBS (control) for 10 minutes. Whole cell lysates were immunoblotted with pp38, p38 and rap1B. Signal intensity (DU) was quantified, normalized to p38 and expressed as percent of stimulated NT. Experiments were performed in triplicate. (C) Whole cell lysates from SCCHN and normal keratinocytes (HOK) were immunoblotted for pp38, p38 and actin. Signal intensity (DU) was quantified, normalized to p38 then actin and expressed as percent of HOK. (D) TTP was immunoprecipitated from UM-SCC-1, -11A and -81B and were immunoblotted with phosphoserine or TTP. Signal intensity (DU) was quantified, normalized to input and expressed as percent of UM-SCC-1. (E) UM-SCC-1 cells were serum starved and treated with 10μM of SB203580. Clarified cell lysates were immunoprecipitated with TTP antibody, and blotted with anti-TTP and anti-phosphoserine antibodies. (F) Cells transfected with sip38 or NT were incubated with PBS (control) or IL-1β for 10 minutes. Top two panels: Immunoprecipitation was performed as above. Signal intensity (DU) was quantified, normalized to input and expressed as percent of unstimulated NT. Lower three panels: Whole cell lysates from NT and sip38 were also immunoblotted with pp38, p38 and actin. Signal intensity (DU) was quantified, normalized to p38 then actin and expressed as percent of unstimulated NT. Three independent experiments were performed. (G) Cells were transfected with sip38 or NT and cell lysates were immunoblotted with p38 and actin antibodies. Signal intensity (DU) was quantified, normalized and expressed as percent of NT. IL-6 secretion was determined with two replicates in each of three independent experiments. (*p<0.01)

Figure 6. Proposed model for role of TTP in invasion

TTP, downregulated or functionally inactivated by rap1B and p38, promotes tumor invasion and progression via mRNA stabilization and increased secretion of IL-6, MMP9 and MMP2.