ORIGINAL ARTICLE
Hypoxic regulation of the PERK/ATF4/LAMP3-arm of the unfolded protein response in head and neck squamous cell carcinoma Anika Nagelkerke, PhD,1,2 Fred C. G. J. Sweep, PhD,2 Hanneke Stegeman, PhD,1 Reidar Grenman, MD, PhD,3 Johannes H. A. M. Kaanders, MD, PhD,1 Johan Bussink, MD, PhD,1 Paul N. Span, PhD1* 1
Department of Radiation Oncology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 2Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 3Department of Otorhinolaryngology–Head and Neck Surgery and Department of Medical Biochemistry, Turku University Hospital and University of Turku, Turku, Finland.
Accepted 11 March 2014 Published online 27 June 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.23693
ABSTRACT: Background. The purpose of this study was to examine the hypoxic regulation of the PKR-like endoplasmic reticulum kinase (PERK)/ activating transcription factor-4 (ATF4)/lysosome-associated membrane protein 3 (LAMP3)-arm of the unfolded protein response (UPR) in head and neck squamous cell carcinoma (HNSCC). Methods. LAMP3 expression was determined in patient biopsies by immunohistochemistry and correlated to clinicopathological parameters. mRNA and protein expression for PERK, ATF4, and LAMP3 was evaluated after hypoxic exposure of HNSCC cell lines. Results. In patients with HNSCC, high LAMP3 expression correlated with N classification (p 5 .019) and the occurrence of distant metastases during follow-up (p 5 .039). Patients with high LAMP3 levels had a worse metastasis-free survival (p 5 .008). Intriguingly, LAMP3 expres-
sion was localized exclusively in normoxic areas of tumors and xenografts. Expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 was not universally induced in hypoxic HNSCC cell lines. Exposure to endoplasmic reticulum-stress stimulated PERK, ATF4, and LAMP3 expression. Conclusion. LAMP3 is relevant for prognosis in HNSCC. However, the PERK/ATF4/LAMP3-arm of the UPR responds differently to hypoxia in C 2014 Wiley Periodicals, Inc. HNSCC compared to other tumor types. V Head Neck 37: 896–905, 2015
INTRODUCTION
apoptotic cell death.15 Apparently, hypoxic regions represent a niche in which tumor cells can survive and enhance their aggressiveness. Cell survival under hypoxic conditions can be mediated by the unfolded protein response (UPR).16–19 Under conditions of stress, such as hypoxia, misfolded proteins will accumulate in the endoplasmic reticulum. The UPR is aimed at restoring homeostasis by alleviating this endoplasmic reticulum stress. In other words, the UPR provides cells with ways to survive cellular stress. Three fundamental endoplasmic reticulum-resident transmembrane proteins convey UPR signaling: PKR-like endoplasmic reticulum kinase (PERK), inositol-requiring protein 1 (IRE1), and activating transcription factor-6 (ATF6).16,17,19 Of these 3 arms, the PERK-arm has been examined most extensively. PERK is a kinase that can phosphorylate eukaryotic initiation factor 2a (eIF2a), thereby preventing general mRNA translation.20 Nevertheless, translation of a small number of mRNAs will be preferentially induced under stress conditions. Activation transcription factor 4 (ATF4) is one of the factors induced by endoplasmic reticulum stress in a PERK-dependent manner.21 ATF4 itself is an indispensable transcription factor, overexpressed in cancer,22 inducing the expression of several genes essential for cell survival.23
The presence of hypoxic regions within solid tumors is of unequivocal importance for patient prognosis in a variety of cancer types. Hypoxia arises as the consequence of an imbalance between the supply and consumption of oxygen within a tumor. Inadequate vascularization is one of the underlying mechanisms impairing oxygenation. As a consequence, tumor cells are either too remote from blood vessels to receive sufficient amounts of oxygen (diffusion-limited hypoxia) or vessels are structurally and functionally not correctly formed restricting the oxygen supply (perfusion-limited hypoxia).1–3 The consequences of hypoxia in tumors are severe. Patients with hypoxic tumors have a worse prognosis compared to patients with well-oxygenated tumors.4–11 Hypoxia can cause resistance to various treatment modalities,12 of which radiotherapy is a well-known example.13 In addition, hypoxia can influence gene expression, can cause genetic instability, and promote invasiveness and metastasis.14 Hypoxic cells are also less susceptible to
*Corresponding author: P. N. Span, Department of Radiation Oncology 874, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: [email protected]
896
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
KEY WORDS: lysosome-associated membrane protein 3 (LAMP3), unfolded protein response (UPR), hypoxia, endoplasmic reticulum stress, head and neck squamous cell carcinoma (HNSCC)
LAMP3
Recently, a novel target gene of the UPR was identified: lysosome-associated membrane protein 3 (LAMP3).21,24 This protein is induced by hypoxia as part of the PERK/ATF4-arm. Indeed, LAMP3 was found upregulated by hypoxia in a panel of different tumor cell lines.24 In addition, LAMP3 has prognostic value in patients with breast cancer and patients with cervical cancer.25,26 Increasing evidence from our laboratory emphasizes the importance of LAMP3 in tumor biology with functions in hypoxia-induced metastasis and therapy resistance.27–30 This prompted us to explore whether LAMP3 expression is also relevant for prognosis in a cohort of patients with head and neck squamous cell carcinoma (HNSCC), because hypoxia is a strong adverse factor in this tumor type.
MATERIALS AND METHODS Patients In this study, 29 patients with HNSCC located in the oral cavity, oropharynx, hypopharynx, or larynx were included. All patients were treated at the Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, and all gave written informed consent. Pretreatment tumor biopsies were obtained and immediately stored in liquid nitrogen until further processing. Before taking biopsies, the hypoxic cell marker pimonidazole was administered to patients, as previously described.31,32
IN
HNSCC
TABLE 1. Patient and tumor characteristics and LAMP3 expression levels.
Age, y, mean (range)
59 (36–84)
Total no. (%) 29 (100) Sex Male 26 (90) Female 3 (10) T classification T1/2 10 (34) T3 13 (45) T4 6 (21) N classification N0 7 (24) N1 22 (76) Tumor site Larynx 11 (38) Hypopharynx 9 (31) Oropharynx 7 (24) Oral cavity 2 (7) Distant metastasis occurrence No 21 (72) Yes 8 (28)
LAMP3 expression, median (min–max)
p value
135 (65–487) .474* 137 (65–487) 108 (84–207) .736† 148 (84–240) 131 (65–487) 136 (119–164) .019* 103 (65–146) 152 (84–487) .225† 131 (65–487) 187 (91–240) 110 (84–225) 153 (146–160) .039* 127 (65–487) 178 (135–240)
Abbreviation: LAMP3, lysosome-associated membrane protein 3. * Mann–Whitney U test. † Kruskal–Wallis test.
Growth of head and neck squamous cell carcinoma xenografts
Hypoxic conditions were induced with a H35 Hypoxystation (Don Whitly Scientific, Shipley, UK). For treatment with tunicamycin, cells were incubated with 1 lg/mL tunicamycin from Streptomyces sp., T7765 (Sigma Aldrich, St. Louis, MO) for 24 hours.
HNSCC xenografts of 10 other patient biopsies and 4 UT-SCC cell lines were established, as previously described.33,34
RNA isolation, cDNA synthesis, and quantitative polymerase chain reaction
Immunohistochemistry, image acquisition, and analysis Immunohistochemical staining of xenografts was performed, as described earlier.26 Patient material was stained similarly, with the exception that vessels were stained using the mouse antihuman endothelium antibody PAL-E (Euro Diagnostica, Arnhem, The Netherlands) diluted 1:10. Alexa-Fluor-647-conjugated chicken-antimouse immunoglobulin G (A21463; 2 mg/mL, Invitrogen) diluted 1:100 was used as the secondary antibody. All images were acquired using a Zeiss Axioskop fluorescence microscope (Zeiss, G€ottingen, Germany) in combination with IP-lab imaging software (Scanalytics, Fairfax, VA). LAMP3 expression was established by using an adapted semiquantitative H-score. Average staining intensity was determined by IP-lab and multiplied by an estimate of the percentage of positive cells.
Cell culture, hypoxic incubations, and tunicamycin treatment Twelve UT-SCC cell lines were maintained in Dulbecco’s modified Eagles medium supplemented with 10% (vol/vol) fetal bovine serum, 20 mM Hepes, 1 3 nonessential amino acids, 2 mM L-glutamine, 10 U/mL penicillin, and 10 lg/mL streptomycin (all from PAA Laboratories, C€ olbe, Germany) at 37 C with 5% CO2.
RNA was isolated using the Total RNA purification kit (Norgen Biotek, Thorold, Canada), in accord with the manufacturer’s instructions. Next, 1 lg of RNA was reverse transcribed using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Richmond, CA). Quantitative polymerase chain reaction was performed with SYBR Green (Applied Biosystems, Foster City, CA) on a CFX96 real-time PCR detection system (Bio-Rad) and the following primers: PERK FW: 50 -CTGATTTTGAGCC AATTC-30 and RV: 50 -CCGGTACTCGCGTCGCTG-30 ; ATF4 FW: 50 -CCTTCACCTTCTTACAACCT-30 and RV: 50 -GTAGTCTGGCTTCCTATCTC-30 ; LAMP3 FW: 50 TGAAAACAACCGATGTCCAA-30 and RV: 50 -TCAGA CGAGCACTCATCCAC-30 ; 18s FW: 50 -GTAACCCGTT GAACCCCATT-30 and RV: 50 -CCATCCAATCGGTAGT AGCG-30 . A pre-developed assay (Applied Biosystems) was used for the reference gene hypoxanthine-guanine phosphoribosyl transferase.35 For analysis of gene expression, 18s was used as the reference gene during hypoxic conditions.36
Western blot analysis Western blot analysis was performed as previously described.26 Primary antibodies used were: rabbit antiPERK (#5683), rabbit anti-phospho-PERK (#3179), rabbit anti-phospho-eIF2a (#3597), and rabbit anti-ATF4 HEAD & NECK—DOI 10.1002/HED
JUNE 2015
897
NAGELKERKE ET AL.
FIGURE 1. Association of lysosome-associated membrane protein 3 (LAMP3) expression levels with clinicopathological parameters. Association between LAMP3 and N classification (A) or development of metastases (B). Horizontal lines represent median values and dashed horizontal lines represent the median value of the entire population. Open symbols are outliers above the range of the chosen y-axis. Kaplan–Meier estimates of overall (C), disease-free (D), and metastasis-free (E) survival.
(#11815) (all from Cell Signaling Technology, Beverly, MA), rabbit anti-LAMP3 (LS-B7698; Lifespan Biosciences, Seattle, WA) and mouse anti-a-tubulin (Calbiochem, San Diego, CA). All antibodies were diluted 1:1000. Western blots were quantified using ImageJ (National Institutes of Health, Bethesda, MD).
Statistical analysis Statistical analyses were performed using GraphPad Prism 4.0 software (GraphPad Software, La Jolla, CA). Associations between LAMP3 expression and clinicopathological parameters were assessed using Mann–Whitney U or Kruskal–Wallis tests. Patient survival was evaluated using the Kaplan–Meier method with log-rank test. Quantitative polymerase chain reaction data were analyzed using the t test and correlations were assessed using Spearman’s rank correlation test. Any p value < .05 was indicative of statistical significance. Asterisks indicate statistical signifi898
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
cance relative to the corresponding control, where * is p < .05, ** is p < .01, and *** is p < .001.
RESULTS High LAMP3 expression is associated with N classification, metastasis formation, and survival Twenty-nine patients were included in this study. The clinical characteristics are presented in Table 1. The median duration of follow-up for all patients was 28.2 months and for surviving patients it was 83.9 months. Patient biopsies were stained for LAMP3, the hypoxic cell marker pimonidazole and vessels. Expression of LAMP3 was scored semiquantitatively and correlated with clinicopathological parameters (Table 1). A significant association was found between LAMP3 expression and N classification, where patients with positive lymph nodes had higher levels of LAMP3 (p 5 .019; Figure 1A). In addition, patients who developed distant
LAMP3
IN
HNSCC
FIGURE 2. Merged pseudo-colored fluorescent images of lysosome-associated membrane protein 3 (LAMP3; red), pimonidazole (green), and vessels (blue) in a biopsy from a squamous cell carcinoma of the larynx (A). A larger magnification of the white rectangle is shown in (B). Original magnification is 3100. Bars equal 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary. com.]
metastases within the first 24 months of follow-up had higher LAMP3 expression compared with patients without distant metastases (p 5 .039; Figure 1B). Next, patients were dichotomized based on the median LAMP3 expression score. Patients with high LAMP3 levels in their primary tumor displayed a reduced overall, disease-free, and metastasis-free survival, but differences with the low LAMP3 group were only statistically significant for
metastasis-free survival (p 5 .008; Figure 1C, 1D, and 1E).
LAMP3 expression in tumor biopsies and xenografts is localized exclusively in normoxic areas As previous data showed that LAMP3 is associated with hypoxia in breast cancer,26 we next examined HEAD & NECK—DOI 10.1002/HED
JUNE 2015
899
NAGELKERKE ET AL.
FIGURE 3. Merged pseudocolored fluorescent images of lysosome-associated membrane protein 3 (LAMP3; red), pimonidazole (green), and vessels (blue) in a SCCNij202 xenograft (A), and a UT-SCC45 xenograft (B). Original magnification is 3100. Bars equal 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
LAMP3 expression in relation to the exogenous hypoxia marker pimonidazole. Unexpectedly, co-staining of LAMP3 with pimonidazole in the patient biopsies revealed that LAMP3 expression was localized exclusively in normoxic areas. A representative example of a staining is given in Figure 2. To further substantiate these findings, LAMP3 expression relative to pimonidazole and vessels was evaluated in a panel of different HNSCC xenografts. Ten SCCNij and 4 UT-SCC xenografts were stained. Six of the 10 SCCNij xenografts and 1 of the 4 UT-SCC xenografts showed expression of LAMP3. Two representative images are shown in Figure 3. Similar to the patient biopsy material, in these xenografts, LAMP3 900
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
expression was again limited to the non-hypoxic areas close to the vasculature (Figure 3A and 3B).
PERK, ATF4, and LAMP3 expression is not universally induced after hypoxic exposure of UT-squamous cell carcinoma cells To further examine how hypoxia affects the PERK/ ATF4/LAMP3-arm of the UPR in HNSCC, a panel of UT-SCC cells was cultured under hypoxic conditions. First, basal expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 was evaluated in the UT-SCC cells (Figure 4A–4C, and 4D). Levels were found to vary
LAMP3
IN
HNSCC
FIGURE 4. Basal mRNA expression levels of (A) PKR-like endoplasmic reticulum kinase (PERK), (B) activating transcription factor-4 (ATF4), and (C) lysosome-associated membrane protein 3 (LAMP3) in a panel of 12 UT-SCC cell lines. Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each. (D) Basal protein expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 in the UT-SCC cell panel. a-tubulin represents the loading control. Numbers below the bands indicate densitometric analysis corrected for the corresponding a-tubulin. (E) Correlation between mRNA and protein expression of PERK (left), ATF4 (middle), and LAMP3 (right) in 12 UT-SCC cell lines.
considerably between different cell lines. In addition, mRNA and protein expression did not show similar levels of expression (Figure 4E). Direct comparison of LAMP3 mRNA expression with that previously found in breast cancer cells26 revealed that mRNA expression in UT-SCC cells was not substantially lower or higher. Exposure of UT-SCC cells to 0.1% O2 for up to 72 hours led to diverse effects on the mRNA expression of PERK, ATF4, and LAMP3 (Figure 5A–5C). The mRNA expression of both PERK and ATF4 was not universally induced after hypoxic exposure (Figure 5A and 5B). A significant induction of expression by hypoxia was found in 5 and 2 cell lines, respectively. The other cell lines showed a significant inhibition of expression (5 and 2 cell lines, respectively) or effects were not significant (2 and 8 cell lines, respectively). Figure 5C shows that LAMP3 mRNA expression is significantly downregulated in 7 of the 12 UT-SCC cell lines tested. In contrast, 2 cell lines
showed a significant induction after hypoxic exposure, and, in 3 cell lines, effects were not significant. Next, 3 UT-SCC cell lines were selected for analysis of protein expression after hypoxia. UT-SCC5 (PERK, ATF4, and LAMP3 mRNA induction after hypoxia), UTSCC24A (induction of PERK mRNA, no effect on ATF4 mRNA, and inhibition of LAMP3 mRNA after hypoxia), and UT-SCC45 (PERK, ATF4, and LAMP3 mRNA inhibition after hypoxia) were chosen to obtain a diverse panel with regard to the effect of hypoxia on the UPR. Results on the protein level were not always identical to the effects found on the mRNA level (Figure 5D). Intriguingly, expression of PERK, p-PERK, ATF4, and LAMP3 was reduced in both UT-SCC5 and UT-SCC24A cells. UT-SCC45 cells showed a slight induction at 24 hours of hypoxic exposure for PERK, p-PERK, and LAMP3, whereas ATF4 expression was inhibited at all time points. Data on expression of phospho-eIF2a are in HEAD & NECK—DOI 10.1002/HED
JUNE 2015
901
NAGELKERKE ET AL.
FIGURE 5. Effect of hypoxic exposure (0.1% O2) on (A) PKR-like endoplasmic reticulum kinase (PERK), (B) activating transcription factor-4 (ATF4), and (C) lysosome-associated membrane protein 3 (LAMP3) mRNA expression in 12 UT-SCC cell lines. Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each. (D) Protein expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 after hypoxic exposure (0.1% O2) of 3 UT-SCC cell lines. a-tubulin represents the loading control. Numbers below the bands indicate densitometric analysis corrected for the corresponding a-tubulin.
conflict with PERK, ATF4, and LAMP3 as a substantial induction in all 3 cell lines after hypoxic incubation was found (Figure 5D).
Endoplasmic reticulum stress induces PERK, ATF4, and LAMP3 mRNA expression in UT-SCC5, -24A, and -45 cells
Carbonic anhydrase IX expression is induced in UTSCC5, -24A, and -45 cells upon hypoxic exposure
Next, the effect of endoplasmic reticulum stress on mRNA expression of PERK, ATF4, and LAMP3 was examined in 3 UT-SCC cell lines. Therefore, cells were exposed to tunicamycin, an inhibitor of N-linked glycosylation, which induces endoplasmic reticulum stress and consequentially the UPR. A 24-hour tunicamycin treatment was sufficient to induce a vast increase in PERK, ATF4, and LAMP3 mRNA expression in all 3 cell lines (Figure 6B–6D). This indicates that the PERK/ATF4/ LAMP3-arm is not defective in these cell lines.
Next, we wanted to verify that hypoxic culture actually induces a transcriptional response in UT-SCC cells. Therefore, the mRNA expression of the endogenous hypoxia marker carbonic anhydrase IX (CA-IX) was evaluated in UT-SCC5, -24A, and -45 cells. CA-IX is a known downstream target of hypoxia-inducible factor 1 the master regulator of the hypoxic response. All cell lines displayed a massive induction of CA-IX expression, ranging from a 50-fold induction in UT-SCC45 at 24 hours to a 25,000-fold induction in UT-SCC5 after 72 hours (Figure 6A). This indicates that these cells have a functional HIF-pathway and can respond to hypoxic conditions. 902
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
DISCUSSION In this study, expression of LAMP3 was assessed in a cohort of HNSCCs. Patients with positive lymph nodes and patients developing metastases showed higher
LAMP3
IN
HNSCC
FIGURE 6. (A) Effect of hypoxic exposure (0.1% O2) on carbonic anhydrase IX (CA-IX) mRNA expression in 3 UT-SCC cell lines. Effect of treatment with 1 lg/mL tunicamycin for 24 hours in 3 UT-SCC cell lines on mRNA expression of (B) PKR-like endoplasmic reticulum kinase (PERK), (C) activating transcription factor-4 (ATF4), and (D) lysosome-associated membrane protein 3 (LAMP3). Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each.
expression levels of LAMP3 in their primary tumor. In addition, patients with high LAMP3 expression had a significantly worse metastasis-free survival than patients with low LAMP3 expression in their tumors. These data indicate that also in HNSCCs LAMP3 is of importance for tumor progression. Further examination of the association of LAMP3 with hypoxia revealed that, in contrast to other tumor types,24,26 LAMP3 was not linked with hypoxia in HNSCC. Instead, expression was located exclusively in normoxic areas of both primary tumors and xenografts. Culturing HNSCC cells under hypoxic conditions revealed that, in most cell lines, hypoxia did not induce expression of PERK, ATF4, or LAMP3. Instead, in most cell lines, expression levels of these factors remained unaltered or were even inhibited after exposure to hypoxic conditions. This is distinctly different from tumor cells of other origins, in which the UPR was vastly induced by hypoxia.24,26 The observation that phosphorylated eIF2a accumulates during hypoxia, whereas its downstream targets ATF4 and LAMP3 do not, underlines the unusual manner in which the UPR seems to be regulated in these cells. However, phosphorylation of eIF2a does not rely exclusively on PERK. eIF2a can be phosphorylated by other factors, but this is in response to stresses other than hypoxia.37,38 Whether eIF2a is phosphorylated by a PERK-independent mechanism in HNSCC cells and what effect this has on ATF4 activity remains to be elucidated.
The data presented in the current study show that expression of PERK and ATF4 is still present in HNSCC cells, but is not always induced by hypoxia. Several studies have shown that PERK and ATF4 are induced during conditions of (severe) hypoxia,20–23,39 and that these factors are essential for tumor survival.18,39,40 In xenografts of transformed mouse embryonic fibroblasts ATF4 colocalized with hypoxia, as indicated by the marker EF-5.39 In addition, tumors with defective PERK signaling had a higher number of apoptotic cells in hypoxic areas compared to tumors with intact PERK signaling. Hypoxia induces LAMP3 expression via PERK and ATF4 in tumor cell lines of multiple origins (ie, colon, breast, prostate, lung, and cervical cancer cells).24 However, none of the studies cited above examined HNSCC. We found that endoplasmic reticulum stress provoked by tunicamycin in HNSCC cells did induce a response of the UPR. This demonstrates that the PERK-arm of the UPR is not defective in HNSCC. Previously, PERK-/- tumors were found to exhibit less and smaller hypoxic areas compared to tumors where PERK is still intact.39 Extensive hypoxic regions were still present in HNSCC xenografts, which is different from the effects of defective PERK signaling. The expression of other hypoxia-regulated genes was conventionally induced under hypoxic conditions. Together, these results clearly show that HNSCC represent a group of tumors that respond atypically to hypoxia with regard to the PERK/ATF4/LAMP3-arm of the UPR. We HEAD & NECK—DOI 10.1002/HED
JUNE 2015
903
NAGELKERKE ET AL.
hypothesize that there are 2 possibilities as to what is happening in these cells. Either 0.1% O2 for up to 72 hours is not sufficient to induce endoplasmic reticulum stress in these cells, or pathways other than the PERK/ ATF4-arm are active to deal with hypoxia-induced stress. Our current in vitro data are limited to examination of effects after hypoxia at 0.1% O2. The UPR is maximally induced under conditions of anoxia.41 However, 0.1% O2 is also sufficient for its activation, albeit with different, slower kinetics.26 Nevertheless, it would be interesting to validate these results in a broader range of oxygen concentrations, as this may provide more information on the oxygen dependency of the observed effects. However, if hypoxia does not induce stress in HNSCC cells, then apparently these cells are very well adapted to hypoxic conditions. Previously, survival of UT-SCC cells was found to be relatively insensitive to 0.5% O2,34 indicating that hypoxia is well-tolerated in these cell lines. The absence of a hypoxia-induced response on the PERK/ ATF4-arm in HNSCCs may have important implications for how these tumors deal with conditions of hypoxia. Perhaps HNSCC cells behave differently under hypoxic conditions in terms of their metabolism. As the UPR is mainly activated to deal with an overload of misfolded proteins in the endoplasmic reticulum, absence of a response during hypoxic conditions suggests that these cells have a low requirement for protein folding. In addition, it would be interesting to see how hypoxic exposure affects the other 2 pathways of the UPR besides the PERK-arm. So far, it is unclear whether there is crosstalk between the different arms of the UPR, although redundancy has been suggested.17,42 In conclusion, this study examined the prognostic value of LAMP3 in HNSCC. Patients with high LAMP3 levels had a worse survival, which, despite the small size of the cohort, was highly statistically significant for metastasisfree survival. Intriguingly, LAMP3 expression was localized exclusively in normoxic areas of tumors and xenografts. Analysis of mRNA and protein expression of PERK, ATF4, and LAMP3 in HNSCC cell lines revealed that, in a number of cell lines, expression was not induced, but unaltered or even inhibited after exposure to 0.1% O2. These data suggest that HNSCC represent a subset of tumors that deal with cellular stress induced by hypoxia in an atypical manner.
Acknowledgment We are grateful to Jasper Lok for technical assistance.
REFERENCES 1. Brown JM. Evidence for acutely hypoxic cells in mouse tumours, and a possible mechanism of reoxygenation. Br J Radiol 1979;52:650–656. 2. Thomlinson RH, Gray LH. The histological structure of some human lung cancers and the possible implications for radiotherapy. Br J Cancer 1955;9: 539–549. 3. Rademakers SE, Span PN, Kaanders JH, Sweep FC, van der Kogel AJ, Bussink J. Molecular aspects of tumour hypoxia. Mol Oncol 2008;2:41–53. 4. Brizel DM, Scully SP, Harrelson JM, et al. Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res 1996;56:941–943. 5. Brizel DM, Sibley GS, Prosnitz LR, Scher RL, Dewhirst MW. Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 1997;38:285–289. 6. Fyles AW, Milosevic M, Wong R, et al. Oxygenation predicts radiation response and survival in patients with cervix cancer. Radiother Oncol 1998;48:149–156.
904
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
7. Hockel M, Schlenger K, Aral B, Mitze M, Schaffer U, Vaupel P. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 1996;56:4509–4515. 8. Nordsmark M, Overgaard M, Overgaard J. Pretreatment oxygenation predicts radiation response in advanced squamous cell carcinoma of the head and neck. Radiother Oncol 1996;41:31–39. 9. Sundfor K, Lyng H, Rofstad EK. Tumour hypoxia and vascular density as predictors of metastasis in squamous cell carcinoma of the uterine cervix. Br J Cancer 1998;78:822–827. 10. Janssens GO, Rademakers SE, Terhaard CH, et al. Accelerated radiotherapy with carbogen and nicotinamide for laryngeal cancer: results of a phase III randomized trial. J Clin Oncol 2012;30:1777–1783. 11. Kaanders JH, Wijffels KI, Marres HA, et al. Pimonidazole binding and tumor vascularity predict for treatment outcome in head and neck cancer. Cancer Res 2002;62:7066–7074. 12. Vaupel P, Thews O, Hoeckel M. Treatment resistance of solid tumors: role of hypoxia and anemia. Med Oncol 2001;18:243–259. 13. Hockel M, Vaupel P. Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst 2001;93:266–276. 14. Sutherland RM. Tumor hypoxia and gene expression–implications for malignant progression and therapy. Acta Oncol 1998;37:567–574. 15. Harris AL. Hypoxia–a key regulatory factor in tumour growth. Nat Rev Cancer 2002;2:38–47. 16. Feldman DE, Chauhan V, Koong AC. The unfolded protein response: a novel component of the hypoxic stress response in tumors. Mol Cancer Res 2005;3:597–605. 17. Koumenis C, Wouters BG. “Translating” tumor hypoxia: unfolded protein response (UPR)-dependent and UPR-independent pathways. Mol Cancer Res 2006;4:423–436. 18. Rouschop KM, van den Beucken T, Dubois L, et al. The unfolded protein response protects human tumor cells during hypoxia through regulation of the autophagy genes MAP1LC3B and ATG5. J Clin Invest 2010;120:127–141. 19. Wouters BG, Koritzinsky M. Hypoxia signalling through mTOR and the unfolded protein response in cancer. Nat Rev Cancer 2008;8:851–864. 20. Koumenis C, Naczki C, Koritzinsky M, et al. Regulation of protein synthesis by hypoxia via activation of the endoplasmic reticulum kinase PERK and phosphorylation of the translation initiation factor eIF2alpha. Mol Cell Biol 2002;22:7405–7416. 21. Blais JD, Filipenko V, Bi M, et al. Activating transcription factor 4 is translationally regulated by hypoxic stress. Mol Cell Biol 2004;24:7469–7482. 22. Ameri K, Lewis CE, Raida M, Sowter H, Hai T, Harris AL. Anoxic induction of ATF-4 through HIF-1-independent pathways of protein stabilization in human cancer cells. Blood 2004;103:1876–1882. 23. Rzymski T, Milani M, Pike L, et al. Regulation of autophagy by ATF4 in response to severe hypoxia. Oncogene 2010;29:4424–4435. 24. Mujcic H, Rzymski T, Rouschop KM, et al. Hypoxic activation of the unfolded protein response (UPR) induces expression of the metastasisassociated gene LAMP3. Radiother Oncol 2009;92:450–459. 25. Kanao H, Enomoto T, Kimura T, et al. Overexpression of LAMP3/ TSC403/DC-LAMP promotes metastasis in uterine cervical cancer. Cancer Res 2005;65:8640–8645. 26. Nagelkerke A, Mujcic H, Bussink J, et al. Hypoxic regulation and prognostic value of LAMP3 expression in breast cancer. Cancer 2011;117:3670–3681. 27. Mujcic H, Nagelkerke A, Rouschop KM, et al. Hypoxic activation of the PERK/eIF2alpha arm of the unfolded protein response promotes metastasis through induction of LAMP3. Clin Cancer Res 2013;19:6126–6137. 28. Nagelkerke A, Bussink J, Mujcic H, et al. Hypoxia stimulates migration of breast cancer cells via the PERK/ATF4/LAMP3-arm of the unfolded protein response. Breast Cancer Res 2013;15:R2. 29. Nagelkerke A, Bussink J, van der Kogel AJ, Sweep FC, Span PN. The PERK/ATF4/LAMP3-arm of the unfolded protein response affects radioresistance by interfering with the DNA damage response. Radiother Oncol 2013;108:415–421. 30. Nagelkerke A, Sieuwerts AM, Bussink J, et al. LAMP3 is involved in tamoxifen resistance in breast cancer through modulation of autophagy. Endocr Relat Cancer 2014;21:101–112. 31. Hoogsteen IJ, Lok J, Marres HA, et al. Hypoxia in larynx carcinomas assessed by pimonidazole binding and the value of CA-IX and vascularity as surrogate markers of hypoxia. Eur J Cancer 2009;45:2906–2914. 32. Nijkamp MM, Hoogsteen IJ, Span PN, et al. Spatial relationship of phosphorylated epidermal growth factor receptor and activated AKT in head and neck squamous cell carcinoma. Radiother Oncol 2011;101:165–170. 33. Troost EG, Laverman P, Philippens ME, et al. Correlation of [18F]FMISO autoradiography and pimonidazole [corrected] immunohistochemistry in human head and neck carcinoma xenografts. Eur J Nucl Med Mol Imaging 2008;35:1803–1811. 34. Stegeman H, Kaanders JH, Wheeler DL, et al. Activation of AKT by hypoxia: a potential target for hypoxic tumors of the head and neck. BMC Cancer 2012;12:463. 35. de Kok JB, Roelofs RW, Giesendorf BA, et al. Normalization of gene expression measurements in tumor tissues: comparison of 13 endogenous control genes. Lab Invest 2005;85:154–159. 36. Nagelkerke A, Mujcic H, Wouters B, Span PN. 18S is an appropriate housekeeping gene for in vitro hypoxia experiments. Br J Cancer 2010; 103:590; author reply 591–592.
LAMP3 37. de Haro C, Mendez R, Santoyo J. The eIF-2alpha kinases and the control of protein synthesis. FASEB J 1996;10:1378–1387. 38. Kaufman RJ. Stress signaling from the lumen of the endoplasmic reticulum: coordination of gene transcriptional and translational controls. Genes Dev 1999;13:1211–1233. 39. Bi M, Naczki C, Koritzinsky M, et al. ER stress-regulated translation increases tolerance to extreme hypoxia and promotes tumor growth. EMBO J 2005;24:3470–3481.
IN
HNSCC
40. Blais JD, Addison CL, Edge R, et al. Perk-dependent translational regulation promotes tumor cell adaptation and angiogenesis in response to hypoxic stress. Mol Cell Biol 2006;26:9517–9532. 41. Mazure NM, Pouyssegur J. Hypoxia-induced autophagy: cell death or cell survival? Curr Opin Cell Biol 2010;22:177–180. 42. Lee AH, Iwakoshi NN, Glimcher LH. XBP-1 regulates a subset of endoplasmic reticulum resident chaperone genes in the unfolded protein response. Mol Cell Biol 2003;23:7448–7459.
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
905
Hypoxic regulation of the PERK/ATF4/LAMP3-arm of the unfolded protein response in head and neck squamous cell carcinoma Anika Nagelkerke, PhD,1,2 Fred C. G. J. Sweep, PhD,2 Hanneke Stegeman, PhD,1 Reidar Grenman, MD, PhD,3 Johannes H. A. M. Kaanders, MD, PhD,1 Johan Bussink, MD, PhD,1 Paul N. Span, PhD1* 1
Department of Radiation Oncology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 2Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 3Department of Otorhinolaryngology–Head and Neck Surgery and Department of Medical Biochemistry, Turku University Hospital and University of Turku, Turku, Finland.
Accepted 11 March 2014 Published online 27 June 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.23693
ABSTRACT: Background. The purpose of this study was to examine the hypoxic regulation of the PKR-like endoplasmic reticulum kinase (PERK)/ activating transcription factor-4 (ATF4)/lysosome-associated membrane protein 3 (LAMP3)-arm of the unfolded protein response (UPR) in head and neck squamous cell carcinoma (HNSCC). Methods. LAMP3 expression was determined in patient biopsies by immunohistochemistry and correlated to clinicopathological parameters. mRNA and protein expression for PERK, ATF4, and LAMP3 was evaluated after hypoxic exposure of HNSCC cell lines. Results. In patients with HNSCC, high LAMP3 expression correlated with N classification (p 5 .019) and the occurrence of distant metastases during follow-up (p 5 .039). Patients with high LAMP3 levels had a worse metastasis-free survival (p 5 .008). Intriguingly, LAMP3 expres-
sion was localized exclusively in normoxic areas of tumors and xenografts. Expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 was not universally induced in hypoxic HNSCC cell lines. Exposure to endoplasmic reticulum-stress stimulated PERK, ATF4, and LAMP3 expression. Conclusion. LAMP3 is relevant for prognosis in HNSCC. However, the PERK/ATF4/LAMP3-arm of the UPR responds differently to hypoxia in C 2014 Wiley Periodicals, Inc. HNSCC compared to other tumor types. V Head Neck 37: 896–905, 2015
INTRODUCTION
apoptotic cell death.15 Apparently, hypoxic regions represent a niche in which tumor cells can survive and enhance their aggressiveness. Cell survival under hypoxic conditions can be mediated by the unfolded protein response (UPR).16–19 Under conditions of stress, such as hypoxia, misfolded proteins will accumulate in the endoplasmic reticulum. The UPR is aimed at restoring homeostasis by alleviating this endoplasmic reticulum stress. In other words, the UPR provides cells with ways to survive cellular stress. Three fundamental endoplasmic reticulum-resident transmembrane proteins convey UPR signaling: PKR-like endoplasmic reticulum kinase (PERK), inositol-requiring protein 1 (IRE1), and activating transcription factor-6 (ATF6).16,17,19 Of these 3 arms, the PERK-arm has been examined most extensively. PERK is a kinase that can phosphorylate eukaryotic initiation factor 2a (eIF2a), thereby preventing general mRNA translation.20 Nevertheless, translation of a small number of mRNAs will be preferentially induced under stress conditions. Activation transcription factor 4 (ATF4) is one of the factors induced by endoplasmic reticulum stress in a PERK-dependent manner.21 ATF4 itself is an indispensable transcription factor, overexpressed in cancer,22 inducing the expression of several genes essential for cell survival.23
The presence of hypoxic regions within solid tumors is of unequivocal importance for patient prognosis in a variety of cancer types. Hypoxia arises as the consequence of an imbalance between the supply and consumption of oxygen within a tumor. Inadequate vascularization is one of the underlying mechanisms impairing oxygenation. As a consequence, tumor cells are either too remote from blood vessels to receive sufficient amounts of oxygen (diffusion-limited hypoxia) or vessels are structurally and functionally not correctly formed restricting the oxygen supply (perfusion-limited hypoxia).1–3 The consequences of hypoxia in tumors are severe. Patients with hypoxic tumors have a worse prognosis compared to patients with well-oxygenated tumors.4–11 Hypoxia can cause resistance to various treatment modalities,12 of which radiotherapy is a well-known example.13 In addition, hypoxia can influence gene expression, can cause genetic instability, and promote invasiveness and metastasis.14 Hypoxic cells are also less susceptible to
*Corresponding author: P. N. Span, Department of Radiation Oncology 874, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: [email protected]
896
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
KEY WORDS: lysosome-associated membrane protein 3 (LAMP3), unfolded protein response (UPR), hypoxia, endoplasmic reticulum stress, head and neck squamous cell carcinoma (HNSCC)
LAMP3
Recently, a novel target gene of the UPR was identified: lysosome-associated membrane protein 3 (LAMP3).21,24 This protein is induced by hypoxia as part of the PERK/ATF4-arm. Indeed, LAMP3 was found upregulated by hypoxia in a panel of different tumor cell lines.24 In addition, LAMP3 has prognostic value in patients with breast cancer and patients with cervical cancer.25,26 Increasing evidence from our laboratory emphasizes the importance of LAMP3 in tumor biology with functions in hypoxia-induced metastasis and therapy resistance.27–30 This prompted us to explore whether LAMP3 expression is also relevant for prognosis in a cohort of patients with head and neck squamous cell carcinoma (HNSCC), because hypoxia is a strong adverse factor in this tumor type.
MATERIALS AND METHODS Patients In this study, 29 patients with HNSCC located in the oral cavity, oropharynx, hypopharynx, or larynx were included. All patients were treated at the Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, and all gave written informed consent. Pretreatment tumor biopsies were obtained and immediately stored in liquid nitrogen until further processing. Before taking biopsies, the hypoxic cell marker pimonidazole was administered to patients, as previously described.31,32
IN
HNSCC
TABLE 1. Patient and tumor characteristics and LAMP3 expression levels.
Age, y, mean (range)
59 (36–84)
Total no. (%) 29 (100) Sex Male 26 (90) Female 3 (10) T classification T1/2 10 (34) T3 13 (45) T4 6 (21) N classification N0 7 (24) N1 22 (76) Tumor site Larynx 11 (38) Hypopharynx 9 (31) Oropharynx 7 (24) Oral cavity 2 (7) Distant metastasis occurrence No 21 (72) Yes 8 (28)
LAMP3 expression, median (min–max)
p value
135 (65–487) .474* 137 (65–487) 108 (84–207) .736† 148 (84–240) 131 (65–487) 136 (119–164) .019* 103 (65–146) 152 (84–487) .225† 131 (65–487) 187 (91–240) 110 (84–225) 153 (146–160) .039* 127 (65–487) 178 (135–240)
Abbreviation: LAMP3, lysosome-associated membrane protein 3. * Mann–Whitney U test. † Kruskal–Wallis test.
Growth of head and neck squamous cell carcinoma xenografts
Hypoxic conditions were induced with a H35 Hypoxystation (Don Whitly Scientific, Shipley, UK). For treatment with tunicamycin, cells were incubated with 1 lg/mL tunicamycin from Streptomyces sp., T7765 (Sigma Aldrich, St. Louis, MO) for 24 hours.
HNSCC xenografts of 10 other patient biopsies and 4 UT-SCC cell lines were established, as previously described.33,34
RNA isolation, cDNA synthesis, and quantitative polymerase chain reaction
Immunohistochemistry, image acquisition, and analysis Immunohistochemical staining of xenografts was performed, as described earlier.26 Patient material was stained similarly, with the exception that vessels were stained using the mouse antihuman endothelium antibody PAL-E (Euro Diagnostica, Arnhem, The Netherlands) diluted 1:10. Alexa-Fluor-647-conjugated chicken-antimouse immunoglobulin G (A21463; 2 mg/mL, Invitrogen) diluted 1:100 was used as the secondary antibody. All images were acquired using a Zeiss Axioskop fluorescence microscope (Zeiss, G€ottingen, Germany) in combination with IP-lab imaging software (Scanalytics, Fairfax, VA). LAMP3 expression was established by using an adapted semiquantitative H-score. Average staining intensity was determined by IP-lab and multiplied by an estimate of the percentage of positive cells.
Cell culture, hypoxic incubations, and tunicamycin treatment Twelve UT-SCC cell lines were maintained in Dulbecco’s modified Eagles medium supplemented with 10% (vol/vol) fetal bovine serum, 20 mM Hepes, 1 3 nonessential amino acids, 2 mM L-glutamine, 10 U/mL penicillin, and 10 lg/mL streptomycin (all from PAA Laboratories, C€ olbe, Germany) at 37 C with 5% CO2.
RNA was isolated using the Total RNA purification kit (Norgen Biotek, Thorold, Canada), in accord with the manufacturer’s instructions. Next, 1 lg of RNA was reverse transcribed using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Richmond, CA). Quantitative polymerase chain reaction was performed with SYBR Green (Applied Biosystems, Foster City, CA) on a CFX96 real-time PCR detection system (Bio-Rad) and the following primers: PERK FW: 50 -CTGATTTTGAGCC AATTC-30 and RV: 50 -CCGGTACTCGCGTCGCTG-30 ; ATF4 FW: 50 -CCTTCACCTTCTTACAACCT-30 and RV: 50 -GTAGTCTGGCTTCCTATCTC-30 ; LAMP3 FW: 50 TGAAAACAACCGATGTCCAA-30 and RV: 50 -TCAGA CGAGCACTCATCCAC-30 ; 18s FW: 50 -GTAACCCGTT GAACCCCATT-30 and RV: 50 -CCATCCAATCGGTAGT AGCG-30 . A pre-developed assay (Applied Biosystems) was used for the reference gene hypoxanthine-guanine phosphoribosyl transferase.35 For analysis of gene expression, 18s was used as the reference gene during hypoxic conditions.36
Western blot analysis Western blot analysis was performed as previously described.26 Primary antibodies used were: rabbit antiPERK (#5683), rabbit anti-phospho-PERK (#3179), rabbit anti-phospho-eIF2a (#3597), and rabbit anti-ATF4 HEAD & NECK—DOI 10.1002/HED
JUNE 2015
897
NAGELKERKE ET AL.
FIGURE 1. Association of lysosome-associated membrane protein 3 (LAMP3) expression levels with clinicopathological parameters. Association between LAMP3 and N classification (A) or development of metastases (B). Horizontal lines represent median values and dashed horizontal lines represent the median value of the entire population. Open symbols are outliers above the range of the chosen y-axis. Kaplan–Meier estimates of overall (C), disease-free (D), and metastasis-free (E) survival.
(#11815) (all from Cell Signaling Technology, Beverly, MA), rabbit anti-LAMP3 (LS-B7698; Lifespan Biosciences, Seattle, WA) and mouse anti-a-tubulin (Calbiochem, San Diego, CA). All antibodies were diluted 1:1000. Western blots were quantified using ImageJ (National Institutes of Health, Bethesda, MD).
Statistical analysis Statistical analyses were performed using GraphPad Prism 4.0 software (GraphPad Software, La Jolla, CA). Associations between LAMP3 expression and clinicopathological parameters were assessed using Mann–Whitney U or Kruskal–Wallis tests. Patient survival was evaluated using the Kaplan–Meier method with log-rank test. Quantitative polymerase chain reaction data were analyzed using the t test and correlations were assessed using Spearman’s rank correlation test. Any p value < .05 was indicative of statistical significance. Asterisks indicate statistical signifi898
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
cance relative to the corresponding control, where * is p < .05, ** is p < .01, and *** is p < .001.
RESULTS High LAMP3 expression is associated with N classification, metastasis formation, and survival Twenty-nine patients were included in this study. The clinical characteristics are presented in Table 1. The median duration of follow-up for all patients was 28.2 months and for surviving patients it was 83.9 months. Patient biopsies were stained for LAMP3, the hypoxic cell marker pimonidazole and vessels. Expression of LAMP3 was scored semiquantitatively and correlated with clinicopathological parameters (Table 1). A significant association was found between LAMP3 expression and N classification, where patients with positive lymph nodes had higher levels of LAMP3 (p 5 .019; Figure 1A). In addition, patients who developed distant
LAMP3
IN
HNSCC
FIGURE 2. Merged pseudo-colored fluorescent images of lysosome-associated membrane protein 3 (LAMP3; red), pimonidazole (green), and vessels (blue) in a biopsy from a squamous cell carcinoma of the larynx (A). A larger magnification of the white rectangle is shown in (B). Original magnification is 3100. Bars equal 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary. com.]
metastases within the first 24 months of follow-up had higher LAMP3 expression compared with patients without distant metastases (p 5 .039; Figure 1B). Next, patients were dichotomized based on the median LAMP3 expression score. Patients with high LAMP3 levels in their primary tumor displayed a reduced overall, disease-free, and metastasis-free survival, but differences with the low LAMP3 group were only statistically significant for
metastasis-free survival (p 5 .008; Figure 1C, 1D, and 1E).
LAMP3 expression in tumor biopsies and xenografts is localized exclusively in normoxic areas As previous data showed that LAMP3 is associated with hypoxia in breast cancer,26 we next examined HEAD & NECK—DOI 10.1002/HED
JUNE 2015
899
NAGELKERKE ET AL.
FIGURE 3. Merged pseudocolored fluorescent images of lysosome-associated membrane protein 3 (LAMP3; red), pimonidazole (green), and vessels (blue) in a SCCNij202 xenograft (A), and a UT-SCC45 xenograft (B). Original magnification is 3100. Bars equal 100 lm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
LAMP3 expression in relation to the exogenous hypoxia marker pimonidazole. Unexpectedly, co-staining of LAMP3 with pimonidazole in the patient biopsies revealed that LAMP3 expression was localized exclusively in normoxic areas. A representative example of a staining is given in Figure 2. To further substantiate these findings, LAMP3 expression relative to pimonidazole and vessels was evaluated in a panel of different HNSCC xenografts. Ten SCCNij and 4 UT-SCC xenografts were stained. Six of the 10 SCCNij xenografts and 1 of the 4 UT-SCC xenografts showed expression of LAMP3. Two representative images are shown in Figure 3. Similar to the patient biopsy material, in these xenografts, LAMP3 900
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
expression was again limited to the non-hypoxic areas close to the vasculature (Figure 3A and 3B).
PERK, ATF4, and LAMP3 expression is not universally induced after hypoxic exposure of UT-squamous cell carcinoma cells To further examine how hypoxia affects the PERK/ ATF4/LAMP3-arm of the UPR in HNSCC, a panel of UT-SCC cells was cultured under hypoxic conditions. First, basal expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 was evaluated in the UT-SCC cells (Figure 4A–4C, and 4D). Levels were found to vary
LAMP3
IN
HNSCC
FIGURE 4. Basal mRNA expression levels of (A) PKR-like endoplasmic reticulum kinase (PERK), (B) activating transcription factor-4 (ATF4), and (C) lysosome-associated membrane protein 3 (LAMP3) in a panel of 12 UT-SCC cell lines. Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each. (D) Basal protein expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 in the UT-SCC cell panel. a-tubulin represents the loading control. Numbers below the bands indicate densitometric analysis corrected for the corresponding a-tubulin. (E) Correlation between mRNA and protein expression of PERK (left), ATF4 (middle), and LAMP3 (right) in 12 UT-SCC cell lines.
considerably between different cell lines. In addition, mRNA and protein expression did not show similar levels of expression (Figure 4E). Direct comparison of LAMP3 mRNA expression with that previously found in breast cancer cells26 revealed that mRNA expression in UT-SCC cells was not substantially lower or higher. Exposure of UT-SCC cells to 0.1% O2 for up to 72 hours led to diverse effects on the mRNA expression of PERK, ATF4, and LAMP3 (Figure 5A–5C). The mRNA expression of both PERK and ATF4 was not universally induced after hypoxic exposure (Figure 5A and 5B). A significant induction of expression by hypoxia was found in 5 and 2 cell lines, respectively. The other cell lines showed a significant inhibition of expression (5 and 2 cell lines, respectively) or effects were not significant (2 and 8 cell lines, respectively). Figure 5C shows that LAMP3 mRNA expression is significantly downregulated in 7 of the 12 UT-SCC cell lines tested. In contrast, 2 cell lines
showed a significant induction after hypoxic exposure, and, in 3 cell lines, effects were not significant. Next, 3 UT-SCC cell lines were selected for analysis of protein expression after hypoxia. UT-SCC5 (PERK, ATF4, and LAMP3 mRNA induction after hypoxia), UTSCC24A (induction of PERK mRNA, no effect on ATF4 mRNA, and inhibition of LAMP3 mRNA after hypoxia), and UT-SCC45 (PERK, ATF4, and LAMP3 mRNA inhibition after hypoxia) were chosen to obtain a diverse panel with regard to the effect of hypoxia on the UPR. Results on the protein level were not always identical to the effects found on the mRNA level (Figure 5D). Intriguingly, expression of PERK, p-PERK, ATF4, and LAMP3 was reduced in both UT-SCC5 and UT-SCC24A cells. UT-SCC45 cells showed a slight induction at 24 hours of hypoxic exposure for PERK, p-PERK, and LAMP3, whereas ATF4 expression was inhibited at all time points. Data on expression of phospho-eIF2a are in HEAD & NECK—DOI 10.1002/HED
JUNE 2015
901
NAGELKERKE ET AL.
FIGURE 5. Effect of hypoxic exposure (0.1% O2) on (A) PKR-like endoplasmic reticulum kinase (PERK), (B) activating transcription factor-4 (ATF4), and (C) lysosome-associated membrane protein 3 (LAMP3) mRNA expression in 12 UT-SCC cell lines. Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each. (D) Protein expression of PERK, p-PERK, p-eIF2a, ATF4, and LAMP3 after hypoxic exposure (0.1% O2) of 3 UT-SCC cell lines. a-tubulin represents the loading control. Numbers below the bands indicate densitometric analysis corrected for the corresponding a-tubulin.
conflict with PERK, ATF4, and LAMP3 as a substantial induction in all 3 cell lines after hypoxic incubation was found (Figure 5D).
Endoplasmic reticulum stress induces PERK, ATF4, and LAMP3 mRNA expression in UT-SCC5, -24A, and -45 cells
Carbonic anhydrase IX expression is induced in UTSCC5, -24A, and -45 cells upon hypoxic exposure
Next, the effect of endoplasmic reticulum stress on mRNA expression of PERK, ATF4, and LAMP3 was examined in 3 UT-SCC cell lines. Therefore, cells were exposed to tunicamycin, an inhibitor of N-linked glycosylation, which induces endoplasmic reticulum stress and consequentially the UPR. A 24-hour tunicamycin treatment was sufficient to induce a vast increase in PERK, ATF4, and LAMP3 mRNA expression in all 3 cell lines (Figure 6B–6D). This indicates that the PERK/ATF4/ LAMP3-arm is not defective in these cell lines.
Next, we wanted to verify that hypoxic culture actually induces a transcriptional response in UT-SCC cells. Therefore, the mRNA expression of the endogenous hypoxia marker carbonic anhydrase IX (CA-IX) was evaluated in UT-SCC5, -24A, and -45 cells. CA-IX is a known downstream target of hypoxia-inducible factor 1 the master regulator of the hypoxic response. All cell lines displayed a massive induction of CA-IX expression, ranging from a 50-fold induction in UT-SCC45 at 24 hours to a 25,000-fold induction in UT-SCC5 after 72 hours (Figure 6A). This indicates that these cells have a functional HIF-pathway and can respond to hypoxic conditions. 902
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
DISCUSSION In this study, expression of LAMP3 was assessed in a cohort of HNSCCs. Patients with positive lymph nodes and patients developing metastases showed higher
LAMP3
IN
HNSCC
FIGURE 6. (A) Effect of hypoxic exposure (0.1% O2) on carbonic anhydrase IX (CA-IX) mRNA expression in 3 UT-SCC cell lines. Effect of treatment with 1 lg/mL tunicamycin for 24 hours in 3 UT-SCC cell lines on mRNA expression of (B) PKR-like endoplasmic reticulum kinase (PERK), (C) activating transcription factor-4 (ATF4), and (D) lysosome-associated membrane protein 3 (LAMP3). Data are presented as mean 1 SD. Results are from 2 representative experiments with 3 replicates each.
expression levels of LAMP3 in their primary tumor. In addition, patients with high LAMP3 expression had a significantly worse metastasis-free survival than patients with low LAMP3 expression in their tumors. These data indicate that also in HNSCCs LAMP3 is of importance for tumor progression. Further examination of the association of LAMP3 with hypoxia revealed that, in contrast to other tumor types,24,26 LAMP3 was not linked with hypoxia in HNSCC. Instead, expression was located exclusively in normoxic areas of both primary tumors and xenografts. Culturing HNSCC cells under hypoxic conditions revealed that, in most cell lines, hypoxia did not induce expression of PERK, ATF4, or LAMP3. Instead, in most cell lines, expression levels of these factors remained unaltered or were even inhibited after exposure to hypoxic conditions. This is distinctly different from tumor cells of other origins, in which the UPR was vastly induced by hypoxia.24,26 The observation that phosphorylated eIF2a accumulates during hypoxia, whereas its downstream targets ATF4 and LAMP3 do not, underlines the unusual manner in which the UPR seems to be regulated in these cells. However, phosphorylation of eIF2a does not rely exclusively on PERK. eIF2a can be phosphorylated by other factors, but this is in response to stresses other than hypoxia.37,38 Whether eIF2a is phosphorylated by a PERK-independent mechanism in HNSCC cells and what effect this has on ATF4 activity remains to be elucidated.
The data presented in the current study show that expression of PERK and ATF4 is still present in HNSCC cells, but is not always induced by hypoxia. Several studies have shown that PERK and ATF4 are induced during conditions of (severe) hypoxia,20–23,39 and that these factors are essential for tumor survival.18,39,40 In xenografts of transformed mouse embryonic fibroblasts ATF4 colocalized with hypoxia, as indicated by the marker EF-5.39 In addition, tumors with defective PERK signaling had a higher number of apoptotic cells in hypoxic areas compared to tumors with intact PERK signaling. Hypoxia induces LAMP3 expression via PERK and ATF4 in tumor cell lines of multiple origins (ie, colon, breast, prostate, lung, and cervical cancer cells).24 However, none of the studies cited above examined HNSCC. We found that endoplasmic reticulum stress provoked by tunicamycin in HNSCC cells did induce a response of the UPR. This demonstrates that the PERK-arm of the UPR is not defective in HNSCC. Previously, PERK-/- tumors were found to exhibit less and smaller hypoxic areas compared to tumors where PERK is still intact.39 Extensive hypoxic regions were still present in HNSCC xenografts, which is different from the effects of defective PERK signaling. The expression of other hypoxia-regulated genes was conventionally induced under hypoxic conditions. Together, these results clearly show that HNSCC represent a group of tumors that respond atypically to hypoxia with regard to the PERK/ATF4/LAMP3-arm of the UPR. We HEAD & NECK—DOI 10.1002/HED
JUNE 2015
903
NAGELKERKE ET AL.
hypothesize that there are 2 possibilities as to what is happening in these cells. Either 0.1% O2 for up to 72 hours is not sufficient to induce endoplasmic reticulum stress in these cells, or pathways other than the PERK/ ATF4-arm are active to deal with hypoxia-induced stress. Our current in vitro data are limited to examination of effects after hypoxia at 0.1% O2. The UPR is maximally induced under conditions of anoxia.41 However, 0.1% O2 is also sufficient for its activation, albeit with different, slower kinetics.26 Nevertheless, it would be interesting to validate these results in a broader range of oxygen concentrations, as this may provide more information on the oxygen dependency of the observed effects. However, if hypoxia does not induce stress in HNSCC cells, then apparently these cells are very well adapted to hypoxic conditions. Previously, survival of UT-SCC cells was found to be relatively insensitive to 0.5% O2,34 indicating that hypoxia is well-tolerated in these cell lines. The absence of a hypoxia-induced response on the PERK/ ATF4-arm in HNSCCs may have important implications for how these tumors deal with conditions of hypoxia. Perhaps HNSCC cells behave differently under hypoxic conditions in terms of their metabolism. As the UPR is mainly activated to deal with an overload of misfolded proteins in the endoplasmic reticulum, absence of a response during hypoxic conditions suggests that these cells have a low requirement for protein folding. In addition, it would be interesting to see how hypoxic exposure affects the other 2 pathways of the UPR besides the PERK-arm. So far, it is unclear whether there is crosstalk between the different arms of the UPR, although redundancy has been suggested.17,42 In conclusion, this study examined the prognostic value of LAMP3 in HNSCC. Patients with high LAMP3 levels had a worse survival, which, despite the small size of the cohort, was highly statistically significant for metastasisfree survival. Intriguingly, LAMP3 expression was localized exclusively in normoxic areas of tumors and xenografts. Analysis of mRNA and protein expression of PERK, ATF4, and LAMP3 in HNSCC cell lines revealed that, in a number of cell lines, expression was not induced, but unaltered or even inhibited after exposure to 0.1% O2. These data suggest that HNSCC represent a subset of tumors that deal with cellular stress induced by hypoxia in an atypical manner.
Acknowledgment We are grateful to Jasper Lok for technical assistance.
REFERENCES 1. Brown JM. Evidence for acutely hypoxic cells in mouse tumours, and a possible mechanism of reoxygenation. Br J Radiol 1979;52:650–656. 2. Thomlinson RH, Gray LH. The histological structure of some human lung cancers and the possible implications for radiotherapy. Br J Cancer 1955;9: 539–549. 3. Rademakers SE, Span PN, Kaanders JH, Sweep FC, van der Kogel AJ, Bussink J. Molecular aspects of tumour hypoxia. Mol Oncol 2008;2:41–53. 4. Brizel DM, Scully SP, Harrelson JM, et al. Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma. Cancer Res 1996;56:941–943. 5. Brizel DM, Sibley GS, Prosnitz LR, Scher RL, Dewhirst MW. Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 1997;38:285–289. 6. Fyles AW, Milosevic M, Wong R, et al. Oxygenation predicts radiation response and survival in patients with cervix cancer. Radiother Oncol 1998;48:149–156.
904
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
7. Hockel M, Schlenger K, Aral B, Mitze M, Schaffer U, Vaupel P. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 1996;56:4509–4515. 8. Nordsmark M, Overgaard M, Overgaard J. Pretreatment oxygenation predicts radiation response in advanced squamous cell carcinoma of the head and neck. Radiother Oncol 1996;41:31–39. 9. Sundfor K, Lyng H, Rofstad EK. Tumour hypoxia and vascular density as predictors of metastasis in squamous cell carcinoma of the uterine cervix. Br J Cancer 1998;78:822–827. 10. Janssens GO, Rademakers SE, Terhaard CH, et al. Accelerated radiotherapy with carbogen and nicotinamide for laryngeal cancer: results of a phase III randomized trial. J Clin Oncol 2012;30:1777–1783. 11. Kaanders JH, Wijffels KI, Marres HA, et al. Pimonidazole binding and tumor vascularity predict for treatment outcome in head and neck cancer. Cancer Res 2002;62:7066–7074. 12. Vaupel P, Thews O, Hoeckel M. Treatment resistance of solid tumors: role of hypoxia and anemia. Med Oncol 2001;18:243–259. 13. Hockel M, Vaupel P. Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst 2001;93:266–276. 14. Sutherland RM. Tumor hypoxia and gene expression–implications for malignant progression and therapy. Acta Oncol 1998;37:567–574. 15. Harris AL. Hypoxia–a key regulatory factor in tumour growth. Nat Rev Cancer 2002;2:38–47. 16. Feldman DE, Chauhan V, Koong AC. The unfolded protein response: a novel component of the hypoxic stress response in tumors. Mol Cancer Res 2005;3:597–605. 17. Koumenis C, Wouters BG. “Translating” tumor hypoxia: unfolded protein response (UPR)-dependent and UPR-independent pathways. Mol Cancer Res 2006;4:423–436. 18. Rouschop KM, van den Beucken T, Dubois L, et al. The unfolded protein response protects human tumor cells during hypoxia through regulation of the autophagy genes MAP1LC3B and ATG5. J Clin Invest 2010;120:127–141. 19. Wouters BG, Koritzinsky M. Hypoxia signalling through mTOR and the unfolded protein response in cancer. Nat Rev Cancer 2008;8:851–864. 20. Koumenis C, Naczki C, Koritzinsky M, et al. Regulation of protein synthesis by hypoxia via activation of the endoplasmic reticulum kinase PERK and phosphorylation of the translation initiation factor eIF2alpha. Mol Cell Biol 2002;22:7405–7416. 21. Blais JD, Filipenko V, Bi M, et al. Activating transcription factor 4 is translationally regulated by hypoxic stress. Mol Cell Biol 2004;24:7469–7482. 22. Ameri K, Lewis CE, Raida M, Sowter H, Hai T, Harris AL. Anoxic induction of ATF-4 through HIF-1-independent pathways of protein stabilization in human cancer cells. Blood 2004;103:1876–1882. 23. Rzymski T, Milani M, Pike L, et al. Regulation of autophagy by ATF4 in response to severe hypoxia. Oncogene 2010;29:4424–4435. 24. Mujcic H, Rzymski T, Rouschop KM, et al. Hypoxic activation of the unfolded protein response (UPR) induces expression of the metastasisassociated gene LAMP3. Radiother Oncol 2009;92:450–459. 25. Kanao H, Enomoto T, Kimura T, et al. Overexpression of LAMP3/ TSC403/DC-LAMP promotes metastasis in uterine cervical cancer. Cancer Res 2005;65:8640–8645. 26. Nagelkerke A, Mujcic H, Bussink J, et al. Hypoxic regulation and prognostic value of LAMP3 expression in breast cancer. Cancer 2011;117:3670–3681. 27. Mujcic H, Nagelkerke A, Rouschop KM, et al. Hypoxic activation of the PERK/eIF2alpha arm of the unfolded protein response promotes metastasis through induction of LAMP3. Clin Cancer Res 2013;19:6126–6137. 28. Nagelkerke A, Bussink J, Mujcic H, et al. Hypoxia stimulates migration of breast cancer cells via the PERK/ATF4/LAMP3-arm of the unfolded protein response. Breast Cancer Res 2013;15:R2. 29. Nagelkerke A, Bussink J, van der Kogel AJ, Sweep FC, Span PN. The PERK/ATF4/LAMP3-arm of the unfolded protein response affects radioresistance by interfering with the DNA damage response. Radiother Oncol 2013;108:415–421. 30. Nagelkerke A, Sieuwerts AM, Bussink J, et al. LAMP3 is involved in tamoxifen resistance in breast cancer through modulation of autophagy. Endocr Relat Cancer 2014;21:101–112. 31. Hoogsteen IJ, Lok J, Marres HA, et al. Hypoxia in larynx carcinomas assessed by pimonidazole binding and the value of CA-IX and vascularity as surrogate markers of hypoxia. Eur J Cancer 2009;45:2906–2914. 32. Nijkamp MM, Hoogsteen IJ, Span PN, et al. Spatial relationship of phosphorylated epidermal growth factor receptor and activated AKT in head and neck squamous cell carcinoma. Radiother Oncol 2011;101:165–170. 33. Troost EG, Laverman P, Philippens ME, et al. Correlation of [18F]FMISO autoradiography and pimonidazole [corrected] immunohistochemistry in human head and neck carcinoma xenografts. Eur J Nucl Med Mol Imaging 2008;35:1803–1811. 34. Stegeman H, Kaanders JH, Wheeler DL, et al. Activation of AKT by hypoxia: a potential target for hypoxic tumors of the head and neck. BMC Cancer 2012;12:463. 35. de Kok JB, Roelofs RW, Giesendorf BA, et al. Normalization of gene expression measurements in tumor tissues: comparison of 13 endogenous control genes. Lab Invest 2005;85:154–159. 36. Nagelkerke A, Mujcic H, Wouters B, Span PN. 18S is an appropriate housekeeping gene for in vitro hypoxia experiments. Br J Cancer 2010; 103:590; author reply 591–592.
LAMP3 37. de Haro C, Mendez R, Santoyo J. The eIF-2alpha kinases and the control of protein synthesis. FASEB J 1996;10:1378–1387. 38. Kaufman RJ. Stress signaling from the lumen of the endoplasmic reticulum: coordination of gene transcriptional and translational controls. Genes Dev 1999;13:1211–1233. 39. Bi M, Naczki C, Koritzinsky M, et al. ER stress-regulated translation increases tolerance to extreme hypoxia and promotes tumor growth. EMBO J 2005;24:3470–3481.
IN
HNSCC
40. Blais JD, Addison CL, Edge R, et al. Perk-dependent translational regulation promotes tumor cell adaptation and angiogenesis in response to hypoxic stress. Mol Cell Biol 2006;26:9517–9532. 41. Mazure NM, Pouyssegur J. Hypoxia-induced autophagy: cell death or cell survival? Curr Opin Cell Biol 2010;22:177–180. 42. Lee AH, Iwakoshi NN, Glimcher LH. XBP-1 regulates a subset of endoplasmic reticulum resident chaperone genes in the unfolded protein response. Mol Cell Biol 2003;23:7448–7459.
HEAD & NECK—DOI 10.1002/HED
JUNE 2015
905
