Skip to main content
NCBI home page
iScience logoLink to iScience
. 2020 Jan 18;23(2):100850. doi: 10.1016/j.isci.2020.100850

PLOD2 Is Essential to Functional Activation of Integrin β1 for Invasion/Metastasis in Head and Neck Squamous Cell Carcinomas

Yushi Ueki 1,2, Ken Saito 1,∗, Hidekazu Iioka 1, Izumi Sakamoto 3, Yasuhiro Kanda 4, Masakiyo Sakaguchi 5, Arata Horii 2, Eisaku Kondo 1,6,∗∗
PMCID: PMC6997870  PMID: 32058962

Summary

Identifying the specific functional regulator of integrin family molecules in cancer cells is critical because they are directly involved in tumor invasion and metastasis. Here we report high expression of PLOD2 in oropharyngeal squamous cell carcinomas (SCCs) and its critical role as a stabilizer of integrin β1, enabling integrin β1 to initiate tumor invasion/metastasis. Integrin β1 stabilized by PLOD2-mediated hydroxylation was recruited to the plasma membrane, its functional site, and accelerated tumor cell motility, leading to tumor metastasis in vivo, whereas loss of PLOD2 expression abrogated it. In accordance with molecular analysis, examination of oropharyngeal SCC tissues from patients corroborated PLOD2 expression associated with integrin β1 at the invasive front of tumor nests. PLOD2 is thus implicated as the key regulator of integrin β1 that prominently regulates tumor invasion and metastasis, and it provides important clues engendering novel therapeutics for these intractable cancers.

Subject Areas: Molecular Biology, Cancer

Graphical Abstract

graphic file with name fx1.jpg

Open in a new tab

Highlights

  • •

    PLOD2 specifically regulates intracellular localization of integrin β1

  • •

    The hydroxylation on integrin β1 by PLOD2 is critical for its stability

  • •

    PLOD2 and integrin β1 expression colocalized at the invasive front of SCC tissues

  • •

    PLOD2 is necessary for tumor invasion/metastasis through the integrin β1 maturation

Molecular Biology; Cancer

Introduction

Head and neck cancers are the sixth most common malignant tumors worldwide. They are mainly treated with extended resection or radiation/chemotherapy, but functional deteriorations in swallowing, vocalization, and respiration lead to clinical decline in quality of life (QOL). Even in cases of combined surgical treatment with radiation therapy and chemotherapy, local recurrence including primary lesion and surrounding lymph nodes, is usually observed in approximately 30% of the patients. As distant metastases, as in the lung, can be seen in 25% of cases, recurrence and therapeutic resistance are very conspicuous. The 5-year survival rate in advanced cases of stage III and above is about 40%; thus the patients still present a poor prognosis (Laramore et al., 1992). Recently, cetuximab, a molecular targeting agent, has been administered in combination with cisplatin to patients with head and neck malignancies, showing some benefit to outcomes. However, it remains that subjects suffer from serious side effects such as dermatitis and mucositis, and novel avenues for cancer therapeutics are needed for improved antitumor properties and patient QOL (Bonner et al., 2006).

Previous studies revealed that 2-oxoglutarate and the iron-dependent dioxygenases superfamily function as a hydroxylase/demethylase and that they hydroxylate or demethylate molecules such as transcription factor, histones, and DNA as substrates. Indeed, it has been reported that these enzymes play various roles in cell cycle and gene expression and control of invasion/metastasis of cancer cells in multiple cell lines via modified molecules (Markolovic et al., 2015).

There are some reports that procollagen lysyl hydroxylase (PLOD), which belongs to this superfamily, is involved in the hydroxylation of collagen molecules as the only target molecule identified to date, and it is considered to be an essential enzyme for cross-linking reaction between collagen molecules outside the cell (Wu et al., 2006). PLOD is composed of an N-terminal endoplasmic reticulum-body translocation signal, a hydroxyl-catalyzed domain at the C-terminus, and hydroxylates lysine residues located in the terminal region of the collagen precursor in the ER.

Currently, three members of the PLOD family molecular group are found to share high homology: PLOD1, PLOD2, and PLOD3 (Valtavaara et al., 1997, Passoja et al., 1998a). Tissue expression analysis by Northern blot showed that PLOD2 and PLOD3 had tissue-specific expression unlike the broad expression of PLOD1. In particular, PLOD2 has been reported to be induced in response to differentiation of bone marrow stromal cells (Valtavaara et al., 1998, Uzawa et al., 1999). Gene mutations in PLOD1 and PLOD3 have been identified in the case of Ehlers-Danlos syndrome (fragile skin, hematomas, and joint hypermobility), whereas PLOD2 mutations are clinically reported as involved in the Bruck syndrome characterized by imperfect bone formation (Walker et al., 2004, van der Slot et al., 2003).

Furthermore, based on the previous studies, PLOD2 has been reported to be specifically induced by activation of transcription factor HIF-1α in response to hypoxia and TGF-β1 stimulation on tumor stroma, cancer cells (breast cancer, hepatocellular carcinoma), and sarcoma (mouse sarcoma model) (Chen et al., 2015, Gilkes et al., 2013a, Gilkes et al., 2013b, Eisinger-Mathason et al., 2013).

It is known that collagen secretion and remodeling of extracellular matrix (ECM) are accelerated especially in cancer stromal cells, leading to invasion and metastases of cancer cells (Gilkes et al., 2013a). The association of HIF-1α-dependent increase in expression of PLOD2 and epithelial mesenchymal transition (EMT) has been established in human glioma cells, and glioma patients with high expression of PLOD2 showed a poor prognosis (Xu et al., 2017). However, the intracellular substrate and tumorigenic activity of PLOD2 expressed in cancer cells remains unclear, especially in refractory cancers, such as oral, head, and neck cancers.

In this study, apart from tumor microenvironmental factors, we focused on biological function of PLOD2 on cancer cell itself using oral squamous cell carcinoma (SCC) lines as a model because they are high in PLOD2 expression, especially in tumor invasion/metastasis highlighting specific interaction between PLOD2 and integrin β1. We have found that PLOD2 induces the hydroxylation of integrin β1, which plays an essential role in the stability and cellular localization of integrin β1 for its functional activation. The integrin β1 requires specific interaction with PLOD2, which will lead to greater understanding of molecular mechanisms for cancer invasion/metastasis and foster innovation in therapies against refractory malignancies.

Results

Specific Upregulation of PLOD2 and Its Effect on Cellular Motility of Oral SCC Cells

Since the 2-oxoglutarate and iron-dependent dioxygenases have been classified into several groups according to the target amino acid residue, such as proline hydroxylase, asparagine/aspartate hydroxylase, etc., we first examined their expression in each category to identify specific hydroxylases highly expressed only in SCC lines from the oral cavity, in contrast to the non-neoplastic cellular counterpart (non-tumorigenic immortalized keratinocyte), HaCaT. Three representative human oral SCC lines (HSC-2, HSC-3, Ca9-22) derived from different patients were examined in comparison with HaCaT, and gene expression of typical hydroxylases for each category was examined by a semiquantitative RT-PCR method (Figure S1A). The expression of JMJD2A, JMJD6, EGLN2, MINA53, NO66, HIF1AN, FTO, and TET1 did not show a significant difference between the tumor cells and HaCaT. On the other hand, the expression of PLOD2, EGLN1, EGLN3, OGFOD1, ASPH, and ALKBH3 revealed certain differences in expression among these cell lines, especially in that the mRNA level of PLOD2 in the SCC lines showed the most prominent and consistent increase at four to eight times compared with HaCaT (Figures S1A and 1A). This upregulation of PLOD2 mRNA in tumor cells was also corroborated by protein levels in immunoblot analysis even among the lysyl hydroxylase family including PLOD1 and PLOD3 (Figure 1B). Immunofluorescence analysis with anti-PLOD2 antibody revealed that endogenous PLOD2 was predominantly localized to the ER which was confirmed by GFP-labeled ER marker (Figure 1C). Thus, tumor-specific upregulation of the expression was observed only in PLOD2 among the examined hydroxylase and demethylase family, and the mRNA and protein levels of PLOD1 and PLOD3 were not specifically elevated in the tumor cells although they are categorized to the same family.

Figure 1.

Figure 1

Open in a new tab

Expression of the Various Hydroxylases in Oral SCC Cells

(A) The expression level of mRNAs in oral SCC cells was determined by quantitative PCR compared with that of HaCaT. Data are means ± s.d. from three biological replicates (*p < 0.05, Student's t-test).

(B) Protein expression of PLOD family in SCC lines and HaCaT by immunoblotting.

(C) Immunofluorescence of PLOD2 in oral SCC lines (HSC-2, HSC-3, and Ca9-22) and non-neoplastic keratinocyte (HaCaT). Colocalization of PLOD2 with ER marker (ER-GFP) was indicated by arrowhead. Nuclei were stained with Hoechst 33258. Scale bar = 20 μm.

(D) RNA interference (siRNA)-mediated knockdown of PLOD in oral SCCs demonstrated the attenuated protein expression by immunoblotting.

(E) GFP-expressing SCC cells were transfected with control siRNA (siCtrl) or with PLOD-siRNA (siPLOD1, siPLOD2, and siPLOD3). Cell migration was evaluated by wound healing assay. Images were taken at 0 and 24 h after wound formation (scale bar = 400 μm). The wound width was estimated using fluorescence microscopic images. Each symbol represents siCtrl (circle, black), siPLOD1 (square, blue), siPLOD2 (triangle, red), and siPLOD3 (cross, green). Asterisk indicates p < 0.05 as compared with siCtrl. Data are means ± s.d. from three technical replicates for one biological replicate.

Based on these results, we next studied the cellular motility of the SCC cells, HSC-2, HSC-3, and Ca9-22, treated with specific siRNA against PLOD2 (siPLOD2) in comparison with the cells treated with siPLOD1 or siPLOD3. Specificity of the siRNA for each PLOD isoform (siPLOD1, siPLOD2, and siPLOD3) was confirmed by immunoblotting, which demonstrated that knockdown by these siRNAs did not affect other isotypes (Figures S1B and 1D). In the wound-healing assay, each stable GFP-expressing clone established from parental HSC-2, HSC-3, and Ca9-22 was employed for evaluation. Mobilization of each clone, visualized by fluorescence microscope 24 h after treatment with the siRNA, revealed that only attenuated expression of PLOD2 significantly affected cellular migration among these three SCC cells, whereas neither PLOD1-knockdown nor PLOD3-knockdown critically disturbed migration of the tumor cells (Figure 1E, upper images and Figure S1B). Inhibition of wound closure was over 50% in siPLOD2-treated cells at 24 h following siRNA treatment (Figure 1E, lower graphs); however, the MTT assay indicated that tumor cell proliferation was not affected by siRNA-treatment in all PLOD isoforms (Figure S1C). These data implied that PLOD2 might be deeply involved in regulating tumor cell motility.

Crosstalk between PLOD 2 and Integrin β1 in Cellular Motility

On the basis of these findings, we focused on the specific role of PLOD2 in tumor cell motility. Generally, acceleration of cell mobility is closely related to invasive properties of tumor cells, and we examined whether expression of E-cadherin (CDH1) as a marker of epithelial-mesenchymal transition (EMT) was altered with or without siPLOD2-treatment of the SCC cells. We found that PLOD2-knockdown did not affect the expression and membrane localization of CDH1 for all SCC cells (Figures 2A and S2A). Because all of these three SCC lines highly express integrin β1 as one of the critical motor molecules for invasion, we next examined the effect of PLOD2 on integrin β1 therein. In siPLOD2-treated tumor cells, lack of filopodia development was observed as a morphological change as confirmed by phalloidin staining, and a marked decrease in integrin β1 was revealed, in contrast to the cells treated with control siRNA (Figures 2B, S2B, S2C, and S3). This phenomenon was not seen in siPLOD1- or siPLOD3-treated cells (Figure S1D). The siPLOD2-treated tumor cells with decreased integrin β1 showed no altered expression of CDH1 or SNAIL, suggesting the PLOD2 seemed not to be involved in EMT at least in these SCC cells (Figures 2C and S4A). Involvement of integrin β1 in migration of these SCC cells was demonstrated by knockdown assay using siIntegrin β1 (Figures S4B and S4C). Taken together, our data indicate that integrin β1 appears directly regulated by PLOD2 for these tumor cells in an EMT-independent manner.

Figure 2.

Figure 2

Open in a new tab

PLOD2 Is Essential for Stabilization of Integrin β1

(A) Immunofluorescence revealed expression, and localization of CDH1 was not affected by siPLOD2-treatment in SCC cells.

(B) Expression and intracellular localization of integrin β1 of the SCC cells was examined at 48 h after treatment with siPLOD2. Cytoskeleton and nuclei were stained with phalloidin and Hoechst, respectively. Scale bar = 20 μm.

(C) Expression of integrin β1, CDH1, and SNAIL in the siPLOD2-transfected cells by immunoblotting using anti-PLOD2, anti-integrin β1, anti-CDH1, and anti-SNAIL Ab, respectively.

(D) Semiquantitative expression of integrin β1 mRNA by RT-PCR with or without siPLOD2-treatement.

(E) Comparative ratio of integrin β1 mRNA in siPLOD2-treated cells based on the quantitative PCR results. Quantitative results are mean ± s.d. from three biological replicates (n.s. = not significant, Student's t-test).

(F) Restoration of integrin β1 by treatment with MG132 and chloroquine (CHQ). HSC-2 cells pretreated with siPLOD2 were examined for integrin β1 expression 18 h after treatment with MG132 (1 nM) or CHQ (50 μM), respectively. Expression of integrin β1 protein by immunoblotting (upper panel), intracellular localization of integrin β1 by immunofluorescence using anti-integrin β1 Ab (lower panel). Integrin β1 (red) was merged with lysosome marker (Lyso-GFP). Scale bar = 20 μm.

(G) Effect of PLOD2 mutant lacking the catalytic domain (ΔPKHD) to integrin β1. Integrin β1 of the HSC-2 transfected with myc-tagged PLOD2 lacking the hydroxylase domain (ΔPKHD) compared with that of the cells transfected with the WT. Reduction of integrin β1 detected by immunoblotting (upper panel) and the loss of plasma membrane localization indicated by arrowhead with immunofluorescence (lower panel). Scale bar = 20 μm.

(H) Wound healing assay revealed cell migration was affected in the ΔPKHD-transfected cells as shown in the graph (upper panel) and migratory images (lower panel). Each symbol in the graph represents empty vector (circle, black), PLOD2 WT (square, blue), and PLOD2 ΔPKHD mutant (triangle, red). Data are means ± s.d. from three technical replicates for one biological replicate (*p < 0.05, Student's t-test as compared with empty vector).

Next, to clarify whether PLOD2 affects induction of integrin β1 mRNA, or directly modifies the integrin β1 protein, RT-PCR was first performed to examine fluctuations in mRNA levels. Ultimately, no significant alteration in integrin β1 mRNA expression with or without siPLOD2 introduction was detected in SCC cells, which was further confirmed by qPCR (Figures 2D and 2E). Therefore, siPLOD2 did not affect induction of integrin β1 mRNA, i.e. the result suggested that integrin β1 protein might be persistently produced in tumor cells but may require a certain modification by PLOD2 for stabilization.

Following the recent study reporting degradation of integrin α at lysosome post-ubiquitination (Lobert et al., 2010), we examined the effects of proteasome inhibitor, MG132, and chloroquine (CHQ)—which is known to interfere with endocytosis processes and with ligand delivery to the lysosome (Erbacher et al., 1996)—as to whether they might restore the integrin β1 protein in siPLOD2-treated tumor cells. Immunoblotting demonstrated that both MG132 and CHQ successfully restore integrin β1 to the level of the control siRNA-treated cell as shown in Figure 2F (upper panel) and Figure S5A and that intracellular localization of integrin β1 was observed in the siPLOD2-introduced cells treated with MG132 and with CHQ, whereas membranous integrin β1 was only restored in MG132-treated cells (Figure S5B). The incidence of lysosome localization of integrin β1 remains at the same level in all cases (Figure 2F lower panel and Figure S5C). Based on this finding, the role of PLOD2 as a hydroxylase was examined with respect to stability of integrin β1 in the wild-type PLOD2-transfected HSC-2 vs. the ΔPKHD-PLOD2-transfected cell (ΔPKHD; ΔLysyl hydroxylase/prolyl 4-hydroxylase domain, inactive form lacking the catalytic domain of PLOD2) (Pirskanen et al., 1996, Passoja et al., 1998b, Heikkinen et al., 2000, Ruotsalainen et al., 2006, Kati et al., 2007). Significant enhancement of integrin β1 was not observed in the wild-type PLOD2-transfected HSC-2 but its membrane expression (filopodial localization) was prominently enhanced (Figure 2G, lower panel images) due to the stabilization of integrin β1 protein by sufficient PLOD2. On the other hand, immunoblotting revealed reduction of integrin β1 in the ΔPKHD-PLOD2-transfected cells compared with empty vector or wild-type PLOD2-transfected HSC2 (WT), and loss of the membranous integrin β1 was observed in contrast to the WT and the vector cells (Figure 2G). In this regard, decrease of integrin β1 in ΔPKHD-PLOD2-transfected cells was supposed to be due to destabilization of integrin β1 lacking active PLOD2, which was explained by MG132 assay shown in Figure S5D. Cellular motility of the ΔPKHD-PLOD2-transfected cells was significantly attenuated as shown by wound-healing assay (Figure 2H). These data imply that hydroxylase activity of PLOD2 was required for protein stability and functional localization of integrin β1.

PLOD2 Regulates Intracellular Localization of Integrin β1 Through Their Coupling

To examine whether PLOD2 specifically regulates intracellular localization of integrin β1, "Fluorescent based technology detecting Protein-Protein Interaction in living cells" (the Fluoppi system) was employed to visualize the movement of integrin β1 in the presence of PLOD2 (Figure 3A). When the two proteins of interest specifically bind to each other, it yields aggregated fluorescence foci (red colored dots) inside cells under the Fluoppi system. As shown in Figure 3B, large fluorescence foci in HeLa cotransfected with PLOD2-Red and integrin β1-Ash demonstrated specific binding of PLOD2 to integrin β1 within cells. Further we did semiquantification of fluorescent dot intensity on Fluoppi specimens between PLOD2-alone-expressing cells and integrin β1–PLOD2-coexpressing cells (Figures S6A and S6B). We also repeated Fluoppi assay to obtain clear image of fluorescent dot formation only on the integrin β1–PLOD2 cotransfection into the PLOD2-knockout cells (Figure S6C). Tracking the PLOD2-integrin β1 complex by Fluoppi system using integrin β1-Red expressor in combination with PLOD2-Ash, the fluorescent foci were first colocalized with ER-GFP in the early phase of transfection, suggesting they were at the ER (Figures 3C and 3D) and then Monti-red-labeled integrin β1 protein transited to filopodia and the plasma membrane in the cotransfected cells (Figure 3D). Immunoprecipitation assay using anti-Flag Ab on the cotransfected HEK293T cell lysates indicated binding of integrin β1 both to full-length PLOD2 and ΔPKHD-PLOD2, which demonstrated direct binding of these two proteins, and integrin β1 bound to the site outside of the PKHD domain of PLOD2 (Figures 3E and 3F). Further immunoprecipitation experiments of endogenous PLOD2 in HSC-2 cells revealed an interaction with the immature integrin β1 (Figure S6D).

Figure 3.

Figure 3

Open in a new tab

Specific Coupling of Integrin β1 with PLOD2

(A) The construction of PLOD2 expressor fused Monti-red tag (PLOD2-Red) and integrin β1 expressor fused Ash tag (Integrin β1-Ash) (left panel). Schematic representation of mechanism for intracellular fluorescent dot formation in response to specific protein-protein interaction (right panel).

(B) Marked aggregation of fluorescent dots was detected only in the HeLa cells cotransfected with PLOD2-Red and Integrin β1-Ash. Scale bar = 20 μm.

(C) The fluorescent dots were merged with GFP-labeled ER marker (ER-GFP). Arrowhead indicated each PLOD2, ER-GFP, and colocalized proteins in the transfected HeLa cells, respectively. Scale bar = 20 μm.

(D) The fluorescent dots were observed at lamellipodia of the PLOD2-Red-Integrin β1-Ash cotransfected cells or the Integrin β1-Red-PLOD2-Ash cotransfected cells. Hatched box indicated hyperview field in the right. Scale bar = 20 μm.

(E) Domain organization of the FLAG epitope tagged-Integrin β1 expressor, Myc epitope-tagged PLOD2 mutant expressor lacking PKHD domain (ΔPKHD) used in the immunoprecipitation assay.

(F) Integrin β1 was coprecipitated both with WT-PLOD2 and ΔPKHD-PLOD2 in HEK293T cells. Expression of each PLOD2 expressor and EGFP expressor as a control in whole cell lysates (left), immunoprecipitated PLOD2, and its variant (probed with anti-Myc Ab) using anti-Flag Ab for precipitation (right). IgG heavy chain (H.C.) and light chain (L.C.) are shown.

Direct Hydroxylation of Integrin β1 by PLOD2 for Its Activation

Because the result of the specific binding between PLOD2 and integrin β1 would indicate that integrin β1 might be a substrate for PLOD2, hydroxylation of the integrin β1 protein purified from the lysate of the 293T transfected with wild-type (WT) PLOD2 was analyzed by mass spectrometry (LC/MS) using the integrin β1 derived from ΔPKHD-PLOD2 transfected cells. The main spike, indicated with the hatched box in the overviewed LC/MS chromatogram of integrin β1 from the cotransfected cells with integrin β1 and WT-PLOD2 (Figure 4A), contained the unique spike (indicated by arrow) representing the molecular weight of triple lysine hydroxylation (encoded amino acids between 651 a.a. and 658 a.a., “AFNKGEKK”), which was not observed in the integrin β1 protein from the integrin β1-ΔPKHD-PLOD2 cotransfected sample (Figure 4B; hyperview of 4A). The shift of the integrin β1 peak indicated by the arrow in the WT-PLOD2 cotransfected sample is two-fold higher than that of the integrin β1 in the ΔPKHD-PLOD2 cotransfected sample (Figure 4B). Additionally, we performed an experiment monitoring the activity of PLOD2 employing substrate peptides according to the previous reports (Takaluoma et al., 2007, Guo et al., 2017). The result showed specific increase of succinate derived from α-ketoglutarate and substrate (integrin β1 substrate peptide.; AFNKGEKK) incubated with purified PLOD2 as well as positive control sample using collagen peptide (IKGIKGIKG) sample (Figures 4C and 4D). The stability of integrin β1 harboring the single Lys-replacement with Ala (K564A, K657A, and K658A) did not seem significantly impaired inside cells transfected with each integrin β1 mutant; however, triple Lys-mutated integrin β1 (K564A + K657A + K658A) seemed markedly unstable (Figure 4E). Moreover, the triple Lys-mutative integrin β1 was not recruited to the plasma membrane in the transfected HSC-2 cells, in contrast to the cells with the single Lys-mutative integrin β1, which still retained its membranous localization as well as in those with WT-integrin β1 (Figure 4F). In this regard, flow cytometry analysis revealed that approximately 20% of the WT-transfected BHK cells showed plasma membrane expression of integrin β1, whereas no significant number of cells showed membrane expression of integrin β1 in the triple K-A mutant-transfected cells, which indicated loss of membrane localization without PLOD2 activity (Figures S7A and S7B). Thus, the magnitude of hydroxylation on integrin β1 by PLOD2 might be critical for its intracellular stability.

Figure 4.

Figure 4

Open in a new tab

Hydroxylation of Integrin β1 in Presence of PLOD2 and Its Significance in Intracellular Localization of Integrin β1

(A) LC/MS of integrin β1 purified from lysate of the 293T cells coexpressing integrin β1-Flag and WT-PLOD2 in comparison with that from the cells expressing integrin β1-Flag and ΔPKHD-PLOD2. The main peaks of integrin β1 were highlighted with the hatched box.

(B) Hyperview of the highlighted peaks. Arrowhead represented the fragment of integrin β1 (position of #651-658 a.a. containing three lysines; AFNKGEKK) from the WT-PLOD2-transfected or form the ΔPKHD-PLOD2-transfected lysate, which showed shift of the peak between these two integrin β1 (spectra; 484.5 to 486 m/z).

(C) Recombinant PLOD2-6xHis protein from HEK293T was purified in 150 mM imidazole using cobalt resin. Purification of PLOD2 was confirmed by Coomassie staining and immunoblotting.

(D) Hydroxylation reaction of PLOD2 was carried out in vitro as described in Methods. Collagen peptides or no peptide substrates were used as controls for reaction. Data are means ± s.d. from three technical replicates for one biological replicate (*p < 0.05, Student's t-test).

(E) Expression of integrin β1 mutants replacing Lys#654 to Ala (K654A), Lys#657 to Ala (K657A), Lys#658 to Ala (K658A), or the triple Ala-substituted mutant replacing the lysine (3KA).

(F) Intracellular localization of the integrin β1 of these three mutants on HSC-2. Arrowhead indicated integrin β1 at the plasma membrane. Only in the transfectant with triple Ala mutant did integrin β1 lack its membranous localization. Scale bar = 20 μm.

PLOD2 Facilitates Metastasis of SCC Cells In Vivo

Based on these findings, we examined whether PLOD2-affected invasion/metastasis of SCC cells in vivo using the intrathoracic metastatic mouse model xenografted with HSC-2 cells. Stably GFP-expressing HSC-2 cells from the PLOD2-knockout clone (PLOD2-KO; KO#31, Figure S8A) were xenografted into the thoracic cavity of athymic nude (nu/nu) mice and evaluated for development of metastatic foci 40 days post-implant in comparison with mice xenografted with parental HSC-2 cells (PLOD2-WT). As to cellular behaviors, PLOD2-KO (KO#31) carrying the heterozygous knockout of PLOD2 did not show significant differences in cellular growth compared with parent HSC-2 (Figure S8B), but endogenous expression of integrin β1 was seriously attenuated (Figure S8C). In the KO#31 clone, plasma membrane localization of integrin β1 was also disturbed, which suggested impaired function of integrin β1 (Figure S8D). Reflecting these results, cellular motility of KO#31 was also seriously decreased in comparison with the parental HSC-2 cells (Figure S8E, image and graphs). Whereas, using PLOD2-KO clone (#31) of HSC-2, we performed rescue experiment by reintroducing wild-type PLOD2. The result showed the revertant (revPLOD2) restored cellular motility, which was similar to the original HSC-2 by regaining expression of maturated integrin β1 (Figures S9A and S9B). This corroborated requirement of integrin β1 expression for invasive property of HSC-2 cells, and it was regulated by PLOD2. Further we performed the time-dependent chase experiment for integrin β1 under presence of PLOD2 on PLOD2-KO clone with or without WT-PLOD2 and ΔPKHD-PLOD2 transfection. Result showed significant induction of mature integrin β1 was detected in response to time-dependent increase of WT-PLOD2 (Figure S10). Whereas overexpression of ΔPKHD-PLOD2, deficient form of hydroxylation activity (Figure 4B), induced the instability of integrin β1, which suggested the necessity of hydroxylation for integrin β1 maturation. Consequently, in the tumor metastasis mouse model, implant of the PLOD2-KO (KO#31) HSC-2 clone showed drastic elimination of metastatic lesions inside the thoracic cavity in vivo except for local growth of lung tumors, which was clearly different from the PLOD2-WT cell-implanted mice with multiple metastatic foci on pleura (Figures 5A and S11). Immunohistochemistry of the tumors derived from the PLOD2-WT cell-xenograft and KO#31 cell-xenograft was corroborative with tumor cells with or without coexpression of PLOD2 and functional integrin β1 (Figure 5B).

Figure 5.

Figure 5

Open in a new tab

Deficiency of PLOD2 Inhibited Metastasis of SCC Cells In Vivo

(A) In vivo development of metastatic foci inside thoracic cavity in mice xenografted with GFP-expressing HSC-2 bearing PLOD2-WT (left panel) and the PLOD2-deficient HSC-2 (PLOD2-KO; right panel). Fluorescence imaging was performed 40 days post-implant.

(B) Histological examination of the metastatic tumor with wild type of PLOD2 (PLOD2-WT) and of that without PLOD2 (PLOD2-KO) in mice tumor models. Immunohistochemistry using anti-PLOD2 Ab and anti-integrin β1 Ab were performed in the same tumor. Box at the HE images (upper panel) indicated the field shown as hyperview in the middle and lower panel. Scale bar = 50 μm.

Coexpression of PLOD2 and Integrin β1 in Oral, Pharyngeal, and Laryngeal SCC Tissues from the Patients

Finally, endogenous expression of PLOD2 and integrin β1 in SCC tissues from patients was examined. As shown in Figure 6, enhanced expression of PLOD2 was detected in SCC cells from different origins such as oral cavity (case 1, case 2: well-differentiated type; case 3: moderately differentiated type), pharynx (case 4: well-differentiated type; case 5: moderately differentiated type), and larynx (case 6: moderately differentiated type), which was consistent and coincident with integrin β1 expression, especially at the marginal region and invasive front of the tumor nests (Figure S12A). Moreover, we demonstrated coexpression of these two proteins at invasive cancer nest using double immunofluorescence (Figure S12B).

Figure 6.

Figure 6

Open in a new tab

Expression of PLOD2 and Integrin β1 in Head and Neck SCCs from the Patients

Immunohistochemistry using anti-PLOD2 Ab and anti-integrin β1 Ab were performed on the SCC tissues derived from oral cavity, pharynx, and larynx, respectively. Box at the HE images (left column) indicated the field shown as hyperview in the middle and right column. Scale bar = 50 μm.

Discussion

Intracellular molecules undergoing hydroxylation by various hydroxylases have been reported to significantly alter cellular functions by regulation of protein degradation, resulting in translational activation or repression of particular proteins in various cancer cells. These intracellular events consequently affect cell-cycle progression, gene expression and activation, and the triggering of cancer invasion/metastases (Ploumakis and Coleman, 2015, Ivan et al., 2001, Zhang et al., 2009, Singleton et al., 2014). Previous studies have shown that PLOD2, one of the lysine hydroxylases, which is highly expressed in cancer cells, plays a major role to process collagens as substrate to form a suitable tumor ECM (Chen et al., 2015, Gilkes et al., 2013a, Gilkes et al., 2013b, Eisinger-Mathason et al., 2013). Accordingly, previous reports on the biological significance of PLOD2 in tumors focused on ECM regulation through its interaction with collagen fibers comprising the tumor matrix.

In this study, we found an additional functional aspect of PLOD2 as a direct regulator of cancer invasion/metastasis through its specific interaction with a major integrin family molecule, integrin β1. Namely, PLOD2 is a key regulator for activation of integrin β1 expressed in head and neck SCCs, which stabilizes and activates integrin β1 by hydroxylation via specific binding. This event leads to accelerated cellular motility needed to form metastatic sites for the SCC cells in vivo. Based on results of our analysis, tumor invasion via the PLOD2-mediated activation of integrin β1 seems not to be mediated by the EMT (epithelial-mesenchymal transition) process in cancer cells but by the direct effect of interactions of these two molecules. Marked increase in cellular motility of SCC cells was observed by integrin β1-PLOD2 protein-protein interaction without altering CDH1 or SNAIL expression (Friedl, 2004, Friedl et al., 2004, Canel et al., 2013) as shown in Figure 2C, and the increase appeared strongly affected by catalytic activity of PLOD2 (Figures 2G, 2H, and 4D). Importantly, instability of integrin β1 protein from knockdown of PLOD2 (downregulation of PLOD2) was observed in cancer cells of diverse origins such as esophageal SCC cells (KYSE30), lung adenocarcinoma cells (A549), uterine cervical SCC cells (HeLa), hepatocellular carcinoma cells (KYN-2), and breast invasive ductal adenocarcinoma cells (MCF7) as well (Figure S13A). In addition, we examined change of integrin β6 and integrin α5 expression by treating with PLOD2-siRNA in HSC-2, HSC-3, and Ca9-22 cells, because integrin β6 and α5 were reported to be expressed, in addition to β1 on head/neck SCCs (Koivistro et al., 2000). The result was shown as Figure S13B, which demonstrated significant attenuation of integrin β6 and α5 by PLOD2-siRNA treatment in these cells. As an exception, in HSC-3 cells, unlike other two SCC lines, decrease of α5 was not clearly observed, whereas integrin β6 was decreased as well as β1. This might be because of the HSC-3's different cellular property from HSC-2 and Ca9-22 (both were cloned from primary SCC lesion), which was cloned from the lymph nodal metastasis and with highly expressed Vimentin (Figure 2C). These observations imply that PLOD2 might play a pivotal role for regulation of integrin β1 involved in invasion/metastasis of tumors with diverse lineages. Additionally, integrin β1 knockdown had less of an effect on migration in Figure S4B, than siPLOD2 throughout. PLOD2 may in fact provide an approach to target integrin β1 along with some other invasion-promoting targets. A large screening for PLOD2 targets may be required in the future.

We demonstrated that PLOD2 hydroxylase modified integrin β1 protein as a direct substrate. In the intracellular molecular binding assay (Fluoppi assay), PLOD2 and integrin β1 formed complex at the ER (ER) as shown in Figures 3B, 3C, and S6, then the integrin β1 moved to filopodia and the plasma membrane (Figure 3D). ER localization of PLOD2 has been reported in the previous studies (Chen et al., 2017, Gjaltema et al., 2016), which is consistent with our present result, indicating that PLOD2 first hydroxylizes integrin β1 at the ER, following which the integrin β1 is recruited to the cell surface functional site after the modification.

In previous studies on modification of collagen by PLOD2, hydroxylation of lysine within the X-K-G motif was reported to be essential for polymerization of collagen fibers (Kivirikko and Pihlajaniemi, 1998, Myllyharju and Kivirikko, 2004, Takaluoma et al., 2007, van der Slot et al., 2003). Based on earlier findings, the putative lysine-rich motif for hydroxylation was found at position of #651-658 a.a., “AFNKGEKK,” in integrin β1 (Figure S13C), which may point to a target motif for PLOD2. In fact, the mass spectrometry result revealed “AFNKGEKK” as the supershifted peak among integrin β1 fragments, which appeared only in the active PLOD2-treated sample (Figures 4A and 4B), and coincidently, the triple Lys substitution with Ala (K654A/K657A/K658A) abolished membrane localization of integrin β1 (Figures 4F and S7). Taking these together, the hydroxylation of these lysine within AFNKGEKK sequence by PLOD2 is critically implicated in gain of integrin β1 function in vivo.

We generated the PLOD2-knockout HSC-2 clone, PLOD2-KO (KO#31; heterozygous genomic knockout of PLOD2, PLOD2−/+), and tracked its dynamics in vivo in comparison with those of the parental HSC-2 (wild type; PLOD2-WT) using the mouse xenografted model. Intrathoracic implant of PLOD2-KO cells showed loss of pleural dissemination and pulmonary metastasis; it only showed local growth of tumor, whereas PLOD2-WT cells developed multiple metastatic foci covering the entire thoracic cavity. These data suggested that cancer cell motility was critically regulated by the presence of PLOD2. Immunohistochemistry demonstrated coexpression of PLOD2 and integrin β1 at invasive cancer nests in the metastatic tumor tissue derived from the mice implanted with PLOD2-WT cells, whereas integrin β1 expression in cancer nests was markedly attenuated by defect in PLOD2. This histological finding corroborated the significant relationship between integrin β1 and PLOD2 in vivo as was proven, further, by in vitro molecular assay. Finally, we confirmed that the expression pattern of integrin β1 and PLOD2 in SCC tissues surgically obtained from patients was similar to that of tumors in the mouse model.

Therefore, we conclude that integrin β1, a critical motor molecule for SCC invasion/metastasis, requires PLOD2 for its stabilization and functional cellular localization via their specific interaction through hydroxylation (Figure 7). Our results indicate that specific suppression of PLOD2 activity may provide one of the chief clues for inhibiting cancer invasion/metastasis in future cancer therapeutics.

Figure 7.

Figure 7

Open in a new tab

Intracellular Dynamics of Integrin β1 Mediated by PLOD2 in SCC Cells

Head and neck squamous cell carcinomas retain high-level expression of PLOD2, which induces hydroxylation of integrin β1. Integrin β1 protein is stabilized by the hydroxylation and recruited to cell membrane as its functional site, which contributes to invasion/metastasis of SCCs. Loss of PLOD2 by contrast causes instability of integrin β1, which results in loss of tumor metastasis. PM (plasma membrane), ER (endoplasmic reticulum), integrin α (yellow), and β1 (red).

Limitations of the Study

This study provides a crosstalk between PLOD 2 and integrin β1 in cellular motility, and the hydroxylation on integrin β1 by PLOD2 is critical for its intracellular stability. However, we could not optimize the in vitro reaction of integrin β1 hydroxylation, because of an insoluble integrin β1 mutant substrate, and we were unable to directly detect the function of PLOD2 and integrin β1 in patient-derived primary SCC samples. Further enzymatically and clinically studies remain to be determined.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Acknowledgments

We thank Dr. K. Honma and Dr. T. Kawasaki (Niigata Cancer Center Hospital) for pathological informations. The collaboration was approved both by Niigata University ethical committee and Niigata Cancer Center ethical committee. We thank Dr. K. Myrick for critical reading of the manuscript. We are also grateful to Ms. A. Ageishi and Ms. N. Sumi for kind assistance for office procedures. This work was supported by Grant-in-Aid for Challenging Exploratory Research (Grant Number JP 24659170; K.S.), JAPAN SOCIETY FOR THE PROMOTION OF SCIENCE (JSPS).

Author Contributions

Conceptualization, K.S., E.K.; Methodology, Y.U., K.S.; Investigation, Y.U., K.S., H.I., I.S., Y.K., M.S.; Writing—Original Draft, K.S., E.K.; Writing—Review & Editing, E.K.; Funding Acquisition, Y.U., K.S., E.K.; Resources, Y.U., K.S, A.H.; Supervision, E.K.

Declaration of Interests

The authors declare no competing interests.

Published: February 21, 2020

Footnotes

Supplemental Information can be found online at https://doi.org/10.1016/j.isci.2020.100850.

Contributor Information

Ken Saito, Email: kens@med.niigata-u.ac.jp.

Eisaku Kondo, Email: ekondo@med.niigata-u.ac.jp.

Supplemental Information

Document S1. Transparent Methods, Figures S1–S13, and Table S1
mmc1.pdf (4.1MB, pdf)

References

  1. Bonner J.A., Harari P.M., Giralt J., Azarnia N., Shin D.M., Cohen R.B., Jones C.U., Sur R., Raben D., Jassem J. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N. Engl. J. Med. 2006;354:567–578. doi: 10.1056/NEJMoa053422. [DOI] [PubMed] [Google Scholar]
  2. Canel M., Serrels A., Frame M.C., Brunton V.G. E-cadherin-integrin crosstalk in cancer invasion and metastasis. J. Cell Sci. 2013;126:393–401. doi: 10.1242/jcs.100115. [DOI] [PubMed] [Google Scholar]
  3. Chen Y., Terajima M., Yang Y., Sun L., Ahn Y.H., Pankova D., Puperi D.S., Watanabe T., Kim M.P., Blackmon S.H. Lysyl hydroxylase 2 induces a collagen cross-link switch in tumor stroma. J. Clin. Invest. 2015;125:1147–1162. doi: 10.1172/JCI74725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen Y., Terajima M., Banerjee P., Guo H., Liu X., Yu J., Yamauchi M., Kurie J.M. FKBP65-dependent peptidyl-prolyl isomerase activity potentiates the lysyl hydroxylase 2-driven collagen cross-link switch. Sci. Rep. 2017;7:46021. doi: 10.1038/srep46021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Eisinger-Mathason T.S., Zhang M., Qiu Q., Skuli N., Nakazawa M.S., Karakasheva T., Mucaj V., Shay J.E., Stangenberg L., Sadri N. Hypoxia-dependent modification of collagen networks promotes sarcoma metastasis. Cancer Discov. 2013;3:1190–1205. doi: 10.1158/2159-8290.CD-13-0118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Erbacher P., Roche A.C., Monsiqny M., Midoux P. Putative role of chloroquine in gene transfer into a human hepatoma cell line by DNA/lactosylated polylysine complexes. Exp. Cell Res. 1996;225:186–194. doi: 10.1006/excr.1996.0169. [DOI] [PubMed] [Google Scholar]
  7. Friedl P. Prespecification and plasticity: shifting mechanisms of cell migration. Curr. Opin. Cell Biol. 2004;16:14–23. doi: 10.1016/j.ceb.2003.11.001. [DOI] [PubMed] [Google Scholar]
  8. Friedl P., Hegerfeldt Y., Tusch M. Collective cell migration in morphogenesis and cancer. Int. J. Dev. Biol. 2004;48:441–449. doi: 10.1387/ijdb.041821pf. [DOI] [PubMed] [Google Scholar]
  9. Gilkes D.M., Bajpai S., Chaturvedi P., Wirtz D., Semenza G.L. Hypoxia-inducible factor 1 (HIF-1) promotes extracellular matrix remodeling under hypoxic conditions by inducing P4HA1, P4HA2, and PLOD2 expression in fibroblasts. J. Biol. Chem. 2013;288:10819–10829. doi: 10.1074/jbc.M112.442939. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  10. Gilkes D.M., Bajpai S., Wong C.C., Chaturvedi P., Hubbi M.E., Wirtz D., Semenza G.L. Procollagen lysyl hydroxylase 2 is essential for hypoxia-induced breast cancer metastasis. Mol. Cancer Res. 2013;11:456–466. doi: 10.1158/1541-7786.MCR-12-0629. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  11. Gjaltema R.A., van der Stoel M.M., Boersema M., Bank R.A. Disentangling mechanisms involved in collagen pyridinoline cross-linking: the immunophilin FKBP65 is critical for dimerization of lysyl hydroxylase 2. Proc. Natl. Acad. Sci. U S A. 2016;113:7142–7147. doi: 10.1073/pnas.1600074113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Guo H.F., Cho E.J., Devkota A.K., Chen Y., Russell W., Phillips G.N., Jr., Yamauchi M., Dalby K.N., Hurie J.M. A Scalable lysyl hydroxylase 2 expression system and luciferase-based enzymatic activity assay. Arch. Biochem. Biophys. 2017;618:45–51. doi: 10.1016/j.abb.2017.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Heikkinen J., Risteli M., Wang C., Latvala J., Rossi M., Valtavaara M., Myllylä R. Lysyl hydroxylase 3 is a multifunctional protein possessing collagen glucosyltransferase activity. J. Biol. Chem. 2000;275:36158–36163. doi: 10.1074/jbc.M006203200. [DOI] [PubMed] [Google Scholar]
  14. Ivan M., Kondo K., Yang H., Kim W., Valiando J., Ohh M., Salic A., Asara J.M., Lane W.S., Kaelin W.G., Jr. HIFα targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science. 2001;292:464–468. doi: 10.1126/science.1059817. [DOI] [PubMed] [Google Scholar]
  15. Kati T., Juha L., Johanna M. Lysyl hydroxylase 2 is a specific telopeptide hydroxylase, while all three isoenzymes hydroxylate collagenous sequences. Matrix Biol. 2007;26:396–403. doi: 10.1016/j.matbio.2007.01.002. [DOI] [PubMed] [Google Scholar]
  16. Kivirikko K.I., Pihlajaniemi T. Collagen hydroxylases and the protein disulfide isomerase subunit of prolyl 4-hydroxylases. Adv. Enzymol. Relat. Areas Mol. Biol. 1998;72:325–398. doi: 10.1002/9780470123188.ch9. [DOI] [PubMed] [Google Scholar]
  17. Koivistro L., Grenman R., Heino J., Larjava H. Integrin a5b1, avb1, and avb6 collaborate in squamous cell spreading and migration on fibronectin. Exp. Cell Res. 2000;255:10–17. doi: 10.1006/excr.1999.4769. [DOI] [PubMed] [Google Scholar]
  18. Laramore G.E., Scott C.B., al-Sarraf M., Haselow R.E., Ervin T.J., Wheeler R., Jacobs J.R., Schuller D.E., Gahbauer R.A., Schwade J.G. Adjuvant chemotherapy for resectable squamous cell carcinomas of the head and neck: report on Intergroup Study 0034. Int. J. Radiat. Oncol. Biol. Phys. 1992;23:705–713. doi: 10.1016/0360-3016(92)90642-u. [DOI] [PubMed] [Google Scholar]
  19. Lobert V.H., Brech A., Pedersen N.M., Wesche J., Oppelt A., Malerød L., Stenmark H. Ubiquitination of α5β1 integrin controls fibroblast migration through lysosomal degradation of fibronectin-integrin complexes. Dev. Cell. 2010;19:148–159. doi: 10.1016/j.devcel.2010.06.010. [DOI] [PubMed] [Google Scholar]
  20. Markolovic S., Wilkins S.E., Schofield C.J. Protein hydroxylation catalyzed by 2-oxoglutarate-dependent oxygenases. J. Biol. Chem. 2015;290:20712–20722. doi: 10.1074/jbc.R115.662627. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Myllyharju J., Kivirikko K.I. Collagens, modifying enzymes and their mutations in humans, flies and worms. Trends Genet. 2004;20:34–45. doi: 10.1016/j.tig.2003.11.004. [DOI] [PubMed] [Google Scholar]
  22. Passoja K., Rautavuoma K., Ala-Kokko L., Kosonen T., Kivirikko K.I. Cloning and characterization of a third human lysyl hydroxylase isoform. Proc. Natl. Acad. Sci. U S A. 1998;95:10482–10486. doi: 10.1073/pnas.95.18.10482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Passoja K., Myllyharju J., Pirskanen A., Kivirikko K.I. Identification of arginine-700 as the residue that binds the C-5 carboxyl group of 2-oxoglutarate in human lysyl hydroxylase 1. FEBS Lett. 1998;434:145–148. doi: 10.1016/s0014-5793(98)00966-1. [DOI] [PubMed] [Google Scholar]
  24. Pirskanen A., Kaimio A.M., Myllylä R., Kivirikko K.I. Site-directed mutagenesis of human lysyl hydroxylases expressed in insect cells. Identification of histidine residues and aspartic acid residue critical for catalytic activity. J. Biol. Chem. 1996;271:9398–9402. doi: 10.1074/jbc.271.16.9398. [DOI] [PubMed] [Google Scholar]
  25. Ploumakis A., Coleman M.L. OH, the places you'll go! Hydroxylation, gene expression, and cancer. Mol. Cell. 2015;58:729–741. doi: 10.1016/j.molcel.2015.05.026. [DOI] [PubMed] [Google Scholar]
  26. Ruotsalainen H., Sipilä L., Vapola M., Sormunen R., Salo A.M., Uitto L., Mercer D.K., Robins S.P., Risteli M., Attila A. Glycosylation catalyzed by lysyl hydroxylase 3 (LH3) is essential for basement membranes. J. Cell Sci. 2006;119:625–635. doi: 10.1242/jcs.02780. [DOI] [PubMed] [Google Scholar]
  27. Singleton R.S., Liu-Yi P., Formenti F., Ge W., Sekirnik R., Fischer R., Adam J., Pollard P.J., Wolf A., Thalhammer A. OGFOD1 catalyzes prolyl hydroxylation of RPS23 and is involved in translation control and stress granule formation. Proc. Natl. Acad. Sci. U S A. 2014;111:4031–4036. doi: 10.1073/pnas.1314482111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Takaluoma K., Lantto J., Myllyharju J. Lysyl hydroxylase 2 is a specific telopeptide hydroxylase, while all three isoenzymes hydroxylate collagenous sequences. Matrix Biol. 2007;26:396–403. doi: 10.1016/j.matbio.2007.01.002. [DOI] [PubMed] [Google Scholar]
  29. Uzawa K., Grzesik W.J., Nishiura T., Kuznetsov S.A., Robey P.G., Brenner D.A., Yamauchi M. Differential expression of human lysyl hydroxylase genes, lysine hydroxylation, and cross-linking of type I collagen during osteoblastic differentiation in vitro. J. Bone Miner. Res. 1999;14:1272–1280. doi: 10.1359/jbmr.1999.14.8.1272. [DOI] [PubMed] [Google Scholar]
  30. van der Slot A.J., Zuurmond A.M., Bardoel A.F., Wijmenga C., Pruijs H.E., Sillence D.O., Brinckmann J., Abrahan D.J., Black C.M., Verzijl N. Identification of PLOD2 as telopeptide lysyl hydroxylase, an important enzyme in fibrosis. J. Biol. Chem. 2003;278:40967–40972. doi: 10.1074/jbc.M307380200. [DOI] [PubMed] [Google Scholar]
  31. Valtavaara M., Papponen H., Pirttila A.-M., Hiltunen K., Helander H., Myllyla R. Cloning and characterization of a novel human lysyl hydroxylase isoform highly expressed in pancreas and muscle. J. Biol. Chem. 1997;272:6831–6834. doi: 10.1074/jbc.272.11.6831. [DOI] [PubMed] [Google Scholar]
  32. Valtavaara M., Szpirer C., Szpirer J., Myllylä R. Primary structure, tissue distribution, and chromosomal localization of a novel isoform of lysyl hydroxylase (lysyl hydroxylase 3) J. Biol. Chem. 1998;273:12881–12886. doi: 10.1074/jbc.273.21.12881. [DOI] [PubMed] [Google Scholar]
  33. Walker L.C., Overstreet M.A., Willing M.C., Marini J.C., Cabral W.A., Pals G., Bristow J., Atsawasuwan P., Yamauchi M., Yeowell H.N. Heterogeneous basis of the type VIB form of Ehlers-Danlos syndrome (EDS VIB) that is unrelated to decreased collagen lysyl hydroxylation. Am. J. Med. Genet. A. 2004;131:155–162. doi: 10.1002/ajmg.a.30302. [DOI] [PubMed] [Google Scholar]
  34. Wu J., Reinhardt D.P., Batmunkh C., Lindenmaier W., Far R.K., Notbohm H., Hunzelmann N., Brinckmann J. Functional diversity of lysyl hydroxylase 2 in collagen synthesis of human dermal fibroblasts. Exp. Cell Res. 2006;312:3485–3494. doi: 10.1016/j.yexcr.2006.07.013. [DOI] [PubMed] [Google Scholar]
  35. Xu Y., Zhang L., Wei Y., Zhang X., Xu R., Han M., Huang B., Chen A., Li W., Zhang Q. Procollagen-lysine 2-oxoglutarate 5-dioxygenase 2 promotes hypoxia-induced glioma migration and invasion. Oncotarget. 2017;8:23401–23413. doi: 10.18632/oncotarget.15581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Zhang Q., Gu J., Li L., Liu J., Luo B., Cheung H.W., Boehm J.S., Ni M., Geisen C., Root D.E. Control of cyclin D1 and breast tumorigenesis by the EglN2 prolyl hydroxylase. Cancer Cell. 2009;16:413–424. doi: 10.1016/j.ccr.2009.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Transparent Methods, Figures S1–S13, and Table S1
mmc1.pdf (4.1MB, pdf)

Articles from iScience are provided here courtesy of Elsevier

RESOURCES