Proteasomal cysteine deubiquitinase inhibitor b-AP15 suppresses migration and induces apoptosis in diffuse large B cell lymphoma
The first line therapy for patients with diffuse large B cell (DLBCL) is R-CHOP. About half of DLBCL patients are either refractory to, or will relapse, after the treatment. Therefore, identifying novel drug targets and effective therapeutic agents is urgently needed for improving DLBCL patient survival. b-AP15, a selective small molecule inhibitor of proteasomal USP14 and UCHL5 deubiquitinases (DUBs), has shown selectivity and efficacy in several other types of cancer cells. This is the first study to report the effect of b-AP15 in DLBCL.
Cell lines of two DLBCL subtypes, Germinal Center B Cell/ GCB (SU-DHL-4, OCI-LY-1, OCI-LY-19) and Activated B Cell/ABC (SU-DHL-2), were used in the current study. Cell viability was measured by MTS assay, proliferation by trypan blue exclusion staining assay, cellular apoptosis by Annexin V-FITC/PI staining and mitochondrial outer membrane permeability assays, the activities of 20S proteasome peptidases by cleavage of specific fluorogenic substrates, and cell migration was detected by transwell assay in these GCB- and ABC-DLBCL cell lines. Mouse xenograft models of SU-DHL-4 and SU-DHL-2 cells were used to determine in vivo effects of b-AP15 in DLBCL tumors.
b-AP15 inhibited proteasome DUB activities and activated cell death pathway, as evident by caspase activation and mitochondria apoptosis in GCB- and ABC- DLBCL cell lines. b-AP15 treatment suppressed migration of GCB- and ABC-DLBCL cells via inhibiting Wnt/β-catenin and TGFβ/Smad pathways. Additionally, b-AP15 significantly inhibited the growth of GCB- and ABC DLBCL in xenograft models.
These results indicate that b-AP15 inhibits cell migration and induces apoptosis in GCB- and ABC-DLBCL cells, and suggest that inhibition of 19S proteasomal DUB should be a novel strategy for DLBCL treatment.
KeywordsB-AP15 Diffuse large B cell lymphoma Apoptosis Migration
Activated B cell-like
Cytoxan, hydroxyrubicin, oncovin, and prednisone
Diffuse large B cell lymphoma
Germinal center B cell-like
Poly adenosine diphosphate ribose polymerase
Primary mediastinal B cell lymphoma
Diffuse large B cell lymphoma (DLBCL) is the most common non-Hodgkin’s lymphoma which is highly heterogeneous . Gene expressional profiling classifies DLBCL into at least three distinct molecular subtypes: an activated B cell-like (ABC), a germinal center B cell-like (GCB), and a primary mediastinal B cell lymphoma (PMBCL) [2, 3, 4]. Most of DLBCLs belong to GCB and ABC subtypes, representing up to 41 and 35%, respectively . GCB subtype is characterized by the activation of Bcl-2 and c-Myc [5, 6], while ABC subtype is featured by constitutively activation of NF-κB pathway . Interestingly, in response to standard CHOP (Cytoxan, Hydroxyrubicin, Oncovin, and Prednisone) chemotherapy, GCB-DLBCL patients have a significantly better outcome with relatively favorable 5-year overall survival rates compared to ABC-DLBCL patients [8, 9, 10]. However, the molecular basis for these differential responses of these two DLBCL subtypes remains unknown. While researchers have been looking for subtype-specific therapies for ABC or GCB, until now, there is no success .
Our current research is related to the involvement of proteasome ubiquitin system in DLBCL development and therapy-resistance. 20S proteasome inhibitor bortezomib, which was approved as a single agent in patients with multiple myeloma (MM), was evaluated in clinical phase III studies in DLBCL [1, 12], but the toxicity and limitation of bortezomib have been observed . Compared to traditional 20S proteasome inhibitors, targeting the particular deubiquitinase in the ubiquitin proteasome system is a more selective and less toxic therapy strategy.
Deubiquitinases (DUBs) are important regulators in protein degradation and have been suggested to play an important role in cancer development and therapy resistance [14, 15]. In mammalian cells, there are three DUBs present in the 19S proteasome: USP14, UCHL5 and Rnp11. USP14 and UCHL5 are not constitutive proteasome subunits but are reversibly associated with the Rpn1 and Rpn13 subunits of the 19S RP base, respectively, whereas Rnp11 is an important part of 19S proteasome structure and activity. Following the recruitment of poly-ubiquitin chain-tagged substrate protein locates to 19S, USP14 and UCHL5 trim ubiquitin chains from the distal end while Rnp11 performs cleaving entire chains from substrates, which would then obtain entry into the proteolytic chamber of 20S core region for substrate protein degradation [16, 17]. It has been reported that USP14 and UCHL5 are highly expressed in various tumors and play an important role in regulating oncogenic signaling [18, 19, 20, 21]. A recent study, for instance, showed that USP14 and UCHL5 were detected in tumor cell cytoplasm in 77 and 74% of the DLBCL cases, respectively . UCHL5 and USP14 should thus be considered as new targets in DLBCL therapy. It has been reported that b-AP15, a small molecule inhibitor of USP14 and UCHL5 , is able to induce apoptosis and overcome bortezomib resistance in multiple myeloma and Waldenstroms macroglobulinemia [24, 25]. The effect of b-AP15 on DLBCL, however, has not been evaluated.
In the current report, we investigated the anti-tumor activity of b-AP15 in DLBCL. We found that cells of both ABC- and GCB-subtypes were sensitive to b-AP15 treatment. Our results from both in vitro and in vivo studies suggested that b-AP15, by inhibiting the activities of USP14 and UCHL5 deubiquitinases, can suppress migration and induce apoptosis in GCB- and ABC-DLBCL cells. This study illustrates the potential of b-AP15 to be a candidate therapy for DLBCL, providing a basis for clinical evaluation.
Materials and methods
Chemicals and reagents
b-AP15 was purchased from Merk Millipore (Darmstadt, Germany). The proteasome inhibitor, bortezomib (PS341), was purchased from BD Biosciences (San Jose, CA).SKL2001, IWR-1-endo, TP0427735 HCl, and SIS3 HCl were from SelleckChemicals (Huston, TX). TGFβ1 was purchased from Peprotech. Suc-LLVY-AMC, Z-LLE-AMC, Boc-LRR-AMC were obtained from BostonBiochem (Cambridge, MA). These reagents were dissolved in dimethyl sulfoxide (DMSO) as a stock solution, and stored at − 20 °C. In all experiments, final concentration of DMSO did not exceed 0.3%. Antibodies to the following proteins were purchased from Cell Signaling Technology (Danvers, MA) and used at a dilution of 1:1000: poly adenosine diphosphate ribose polymerase (PARP) (clone 4C10–5, #9532), phospho-Erk1/2 (T202/Y204, #4370), Erk1/2 (#4348), phospho-Akt (#2965), Akt (#4685), p27 (#3688), XIAP (#2045), caspase-8 (#9746), caspase-9 (#9504), Cleaved Caspase-3(9661S), apoptosis-inducing factor (AIF) (#5318), Bax (#5023), phospho-STAT5A/B (Y694/Y699; clone 8–5-2, #9314) and STAT5 (#9358), Bcl-2 (15071S), Smad2/3 (8685S), p-Smad2/3 (8828S), Dvl2 (3224S), LRP6 (3395S), p-LRP6 (Ser1490; 2568S), β-Catenin (8480S), Snail (3879S), Slug (9585S), E-Cadherin (14472S), and N-Cadherin (14215S). Antibodies against ALK-5 (mab5871) was purchased from (Minneapolis, MN). Antibodies against ubiquitin (P4D1) (sc-8017), USP14 (SC-515812) and Ki-67 (sc-23,900) were from Santa Cruz Biotechnology (Dallas, Texas).. Antibodies against cleaved-caspase-3 (AV00021), cytochrome c (C5118) and survivin (S8191) were from Sigma-Aldrich (St. Louis, MO). Anti-UCH37/UCHL5 antibody (ab124931) was from abcam (Cambridge, MA). Anti-GAPDH (#60630) and anti-Actin (#0768) antibodies were from Bioworld Technology (Minnesota, USA). HRP-conjugated goat anti-rabbit (AP132P) and anti-mouse (12–349) antibodies were from Merk Millipore.
The DLBCL cell lines SU-DHL-4, OCI-LY-1, OCI-LY-19 (GCB-DLBCL) and SU-DHL-2 (ABC-DLBCL) were purchased from ATCC (Manassas, VA) and incubated in RPMI 1640 medium (LifeTechnologies, Waltham, MA) supplemented with 10% fetal calf serum (Hyclone, Waltham, MA), 100 unit/ml penicillin, and 0.1 mg/ml streptomycin. Cells were incubated at 37 °C and in water vapor–saturated air with 5% CO2 at one atmospheric pressure.
Cell viability assay
MTS assay (CellTiter 96 Aqueous One Solution reagent, Promega, Madison, WI) was used to measure cell viability. Briefly, 2 × 104 cells in 100 μl were treated with b-AP15 for 48 h. Control cells received DMSO for a final concentration the same as the highest concentration of b-AP15 but less than 0.3% (v/v). Four hours before culture termination, 20 μl MTS was added to the wells. The absorbance density was read on a 96-well plate reader at wavelength 490 nm.
Cell counting assay
SU-DHL-4 and SU-DHL-2 cells were seeded into 24-well plates (2 × 105 cells/ml, 1 ml/well) and treated with various concentrations of b-AP15 for indicated duration. Then 0.4% trypan blue (Sigma-Aldrich) was added to count the number of live and dead cells under a light microscope.
Cell death assay
DLBCL cells were treated with various concentrations of b-AP15 for 24 h. Apoptosis was determined by flow cytometry using Annexin V-fluoroisothiocyanate (FITC)/PI double staining (Sungene Biotech, TianJin, China). DLBCL cells were collected, washed with PBS and resuspended with binding buffer (Sungene Biotech). The cell preparation was then stained with Annexin V and PI following manufacturer’s protocol. Samples were analyzed using FACSCalibur flow cytometer and CellQuestPro software. The Annexin V/PI positive cells in the culture dish were also imaged with an inverted fluorescence microscope equipped with a digital camera (AxioObsever Z1, Zeiss, Germany).
Western blot analysis
Whole cell lysates were prepared in RIPA buffer (1 × PBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 10 mM b-glycerophosphate, 1 mM sodium orthovanadate, 10 mM NaF, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 × Roche Protease Inhibitor Cocktail (Roche, Indianapolis, IN). To detect the level of cytochrome C and AIF, the cytosolic fraction was prepared with a digitonin extraction buffer (10 mM PIPES, 0.015% digitonin, 300 mM sucrose, 100 mM NaCl, 3 mM MgCl2, 5 mM EDTA, and 1 mM PMSF). Western blotting was performed as we previously described , using specific primary antibodies as indicated and appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies as indicated.
Measurement of mitochondrial membrane permeability
The mitochondrial membrane potential of cells treated with b-AP15 or untreated were assayed by mitochondrial membrane potential kit (Sigma-Aldrich, St. Louis, MO), following manufacturer’s instruction. DLBCL cells were treated with various doses of b-AP15, and after 24 h the cells were harvested, prepared in 1 ml warm medium, and then 5 μl cationic hydrophobic mitochondrial potential dye was added. The cells were incubated for 30 min in a 5% CO2, 37 °C incubator. After centrifugation the cells were resuspended with 500 μl assay buffer, followed by monitoring the cells using a flow cytometry with ƛex = 635 nm, ƛem = 660 nm at APC channel.
Proteasomal activity assay
The 20S proteasomal peptidase activities were measured using synthetic fluorogenic substrates. To evaluate in vitro proteasome inhibition, cells were lysed in ice-cold lysis buffer (25 mM Tris-HCl, pH 7.4) for 10 min. Equal amounts of protein from each sample were then treated with various concentration of b-AP15 for 30 min, and then incubated at 37 °C with specific fluorogenic substrates (25 μM) for 2 h in dark. The substrates used were Suc-LLVY-AMC for chymotrypsin-like activity, Z-LLE-AMC for caspase-like activity and Boc-LRR-AMC for trypsin-like activity. Fluorescence intensity was measured using a spectrophotometer at excitation of 350 nm and emission of 438 nm (Varioskan Flash 3001, Thermo,Waltham, MA).
Real-time quantitative polymerase chain reaction (PCR)
Bcl-2 forward, 5′-AACATCGCCCTGTGGATGAC-3′;
Bcl-2 reverse, 5′-AGAGTCTTCAGAGACAGCCAGGAG-3′;
c-Myc forward, 5′-GGAGGCTATTCTGCCCATTTG-3′;
c-Myc reverse, 5′-CGAGGTCATAGTTCCTGTTGGTG-3′;
P65 forward, 5′-ACCTCGACGCATTGCTGTG-3′;
P65 reverse, 5′-CTGGCTGATCTGCCCAGAAG-3′.
SU-DHL-4 and SU-DHL-2 cells were harvested after treatment with or without b-AP15 over 3 h. The cells lysed using DUB buffer (25 mM Tris-HCl, 20 mM NaCl, 5 mM MgCl2, 200 μM ATP), then added HA-Ub-VS (1 μM) and incubated in 37 °C 30 min. The samples were boiled with SDS-PAGE sample loading buffer and subjected to Western blot analysis.
Cell migration assays
SU-DHL-4 and SU-DHL-2 cells were treated with indicated concentration of b-AP15, SKL2001, IWR-1-endo, TP0427735 HCl, SIS HCl, and TGFβ1for 24 h. Thereafter, 2 × 106 cells/ml of two cell types were starved in serum-free RPMI 1640 medium for 1 h at 37 °C in 5% CO2. Cell suspensions (2 × 105 in 100 μl) were added to the upper chambers with a pore size of 8 μm (Corning) and 600 μl of complete medium to the lower chambers. After the plate incubated 2–3 h at 37 °C in 5% CO2, the cells in the lower chamber were counted.
Nude Balb/c mice were bred at the animal facility of Guangzhou Medical University. The mice were housed in barrier facilities with 12 h light dark cycle, with food and water available ad libitum. Totally 3 × 107 cells of SUDHL-4 and SU-DHL-2 cells were inoculated subcutaneously on the flanks of 5-week-old mice, each subtype include 12 mice. After inoculation for 5–6 days, 12 mice evenly separated to vehicle and b-AP15 group randomly, then treated with either vehicle (Cremophor EL: PEG400: saline = 2: 2: 4) or b-AP15 (5 mg/kg/day) for a total of 11 days. The tumor sizes were measured and tumor volumes were calculated by the following formula: a2 × b × 0.4, where “a” is the smallest diameter and “b” is the diameter perpendicular to “a”. Tumor xenografts were removed, weighed, stored and fixed at day 11 after treatment. All experiments were performed in conformity to relevant guidelines and regulations. All animal studies were conducted with the approval of the Institutional Animal Care and Use Committee of Guangzhou Medical University.
Immunohistochemical staining (IHC)
Formalin-fixed xenografts were embedded in paraffin and sectioned using standard techniques. Tumor xenograft sections were immunostained for Ubs, Ki67 and p-Smad2/3.MaxVisionTM reagent (MaixinBiol, Fuzhou, FuJian) was applied to each slide following the manufacturer’s instructions. Color was developed with 0.05% diaminobenzidine and 0.03% H2O2 in 50 mmol/L Tris-HCl, pH 7.6, and the slides were counterstained with hematoxylin. A negative control for every antibody was also included for each xenograft specimen by substituting the primary antibody with preimmunize serum.
All experiments were performed at least thrice, and the results were expressed as mean ± SD where applicable. GraphPad Prism 5.0 software (GraphPad Software) was used for statistical analysis. Comparison of multiple groups was made with one-way ANOVA followed by Tukey’s test or Newman-Kueuls test. Value of p < 0.05 was considered statistically significant.
b-AP15 inhibits cell viability and proliferation in cell lines of GCB- and ABC-DLBCL
We then performed a trypan blue exclusion assay to confirm the capacity of b-AP15 to inhibit proliferation in the two subtypes of DLBCL cell lines. As shown in Fig. 1c, b-AP15 decreased cell growth in a dose- and time-dependent manner.
b-AP15 induces cell death in both GCB- and ABC-DLBCL cell lines
We next assessed the cell death-inducing ability of b-AP15 in GCB- and ABC-DLBCL cells by using Annexin V/PI staining assay. After SU-DHL-4, OCI-LY-1, OCI-LY-19 and SU-DHL-2 cell lines were treated with different concentrations of b-AP15 for 24 h, a significant increase of Annexin V+/PI+ cell populations was detected by inverted fluorescence microscopy, as shown in left panel of Fig. 1d. Similar results were obtained by flow cytometry analysis (Fig. 1d, right panel), confirming that b-AP15 triggered cell death in the two subtypes of DLBCL in a dose dependent manner.
b-AP15-induced apoptosis was associated with activation of caspase and inhibition of anti-apoptotic protein expression
b-AP15 inhibits proteasome function in GCB- and ABC-DLBCL cells
b-AP15 suppresses DLBCL cell migration
b-AP15 down-regulates molecular players involved in DLBCL progression
b-AP15 restrains the growth of xenografted GCB- and ABC-DLBCL tumors in nude mice
The ubiquitin proteasome pathway has been validated as a novel therapeutic target in cancer. The first proteasome inhibitor, bortezomib, has been approved by the US FDA as a single agent or in combination in multiple myeloma. Recent preclinical and clinical studies demonstrated that targeting the canonical NF-κB pathway through inhibition of the 20S proteasome with bortezomib could kill DLBCL cells [32, 33]. Unfortunately, not all DLBCL are bortezomib-sensitive, and patients may eventually develop bortezomib-resistant disease . It has been reported that USP14 and UCHL5 are involved in the development of tumor and are potential new targets for proteasome inhibition in DLBCL . In the current study we planned to figure out whether b-AP15 could inhibit the progression of DLBCL, and we report that b-AP15 can do so through inhibiting the deubiquitinases activities of USP14 and UCHL5.
We found that b-AP15, a novel molecule inhibitor of USP14 and UCHL5 , significantly inhibited the viability and induced apoptosis of GCB- and ABC-DLBCL cells. In addition, we also found that treatment with b-AP15 suppressed the migration of GCB- and ABC-DLBCL cells. Results from nude mouse xenograft models of two types DLBCLs also showed that b-AP15 inhibited tumor growth in vivo.
Our study revealed that b-AP15-induced apoptosis was associated with caspase activation and mitochondria apoptosis (Figs. 1 and 2). b-AP15 downregulates the protein level of XIAP, Bcl-1, Bcl-xl and Survivin. The altered ratio of anti-apoptosis and pro-apoptosis proteins triggered potential reduction in mitochondria, resulting in cytochrome C and AIF release and caspase activation and cell death.
We next investigated the mechanism underlying pro-apoptotic activity of b-AP15. We showed that b-AP15 induced a rapid and significant accumulation of ubiquitin-proteins and substrate protein p27 and b-AP15 has no marked influence on peptidases of 20S proteasome (Fig. 3). In a short period of time, b-AP15 inhibited the function of proteasome, following by the cleavage of PARP. Recent reports have identified that b-AP15 treatment led to the accumulation of misfolded proteins to trigger ER stress . It is a widely accepted concept that ER stress can activate caspase pathway and induce cell apoptosis . We speculated that b-AP15 targeted the DUB function of USP14 and UCHL5, a large amount of unfolded proteins triggered ER stress to induce the cell apoptosis. On the other hand, our study showed that b-AP15 distinctly downregulated those proteins associated with cancer progression in ABC- and GCB-DLBCLs (Figs. 4 and 5). We detected the mRNA and protein levels of p65, Bcl-2 and c-Myc, and the results showed that both the mRNA and protein levels were all decreased except the protein level of Bcl-2. Together, these data may explain the growth and migration inhibition as well as apoptosis induction effects of b-AP15 on both ABC- and GCB-DLBCL.
It is well established that metastasis is an important cause for highly lethality. Recent studies showed that USP14 is overexpressed in colorectal cancer and esophageal squamous cell carcinoma (ESCC) [18, 37]. Downregulation of USP14 resulted in accumulation of poly-ubiquitinated forms of Dvl, which significantly impairs downstream Wnt signaling . The HA-Ub-VS assay demonstrates that b-AP15 inhibits the deubiquitinase activity of both USP14 and UCHL5. b-AP15 treatment induces the decreases of Dvl, β-catenin and c-Myc resulting in inhibition of Wnt signaling and the cell migration of ABC- and GCB-DLBCL cells (Fig. 4a and b). Our data shows the cell migration was activated by SKL2001 (Wnt/β-catenin signaling activator) and decreased by IWR-1-endo (β-catenin pathway inhibitor). Meanwhile, the inhibition of b-AP15 in cell migration was antagonized by SKL2001 (Fig. 4d and e), showing that the Wnt/β-catenin signaling plays an important role in regulating DLBCL cells migration. Like USP14, UCHL5 is also involved in tumorigenesis and progression . It has been reported that UCHL5 combined with transcription factor Smad2/3, can regulate TGFβ signaling [38, 39]. Our result illustrates that b-AP15 decreases the protein level of Smad2/3, and the phosphorylated Smad2/3 (Fig. 4c). Furthermore, both SIS3 HCl and TP0427735 HCl (TGFβ/Smad signaling inhibitors) exhibit a suppression function on both p-Smad 2/3 protein level (Fig. 4f) and cell migration (Fig. 4g) in DLBCL cells, indicating that inhibiting TGFβ/Smad pathway could inhibit cell migration of DLBCL cells, which is similar to b-AP15 function. Moreover, we observe that TGFβ1 induces p-Smad 2/3 and partially rescues the inhibition of p-Smad 2/3 by b-AP15 (Fig. 4f), while its effect on the proportion of cell migration is not significant. Taken together, these findings suggest that b-AP15-regulated cell migration in DLBCL cells is associated with Wnt/β-catenin and TGFβ/Smad signaling pathways, whereas the Wnt/β-catenin signaling pathway may play a more important role in b-AP15 regulated cell migration. Besides, the suppressing function of b-AP15 in cell migration has been confirmed in vitro but should be further investigated in vivo.
In conclusion, our research confirmed that b-AP15 inhibits the activity of two proteasomal DUBs, USP14 and UCHL5, leading to induce ABC- and GCB-DLBCL cell apoptosis. b-AP15 also inhibits Wnt and TGFβ signaling pathways and suppresses ABC- and GCB-DLBCL cells migration. Our studies on the basic research of b-AP15 suggest the feasibility of the clinical application of b-AP15 in DLBCLs.
XPS and JBL designed the experiments and analyzed the data. LLJ, QYH, XMC and YNS performed most of the experiments. JXW, XYL and JHC provided administrative, technical or material support. XPS, JBL and QPD wrote the manuscript. All authors reviewed the manuscript. All authors read and approved the final manuscript.
This work was supported by NSFC (81670154/H0812,81470355/H1616 and 81100378/H0812), Projects (201707010352, 1201410214, 2014A030313492, 201528042) from the Foundation of Guangzhou Science and Technology Innovation Committee, Bureau of Education of Guangzhou Municipality, GD-NSF and Guangdong special support scheme (to XS); The National Funds for developing local colleges and universities (B16056001), GD-NSF for research team (2018B030312001), NSFC (81272451/H1609, 81472762/H1609) and MOE (20134423110002; to JL).
Ethics approval and consent to participate
The animal study has been examined by the Ethics Committee of the Guangzhou Medical University.
Consent for publication
The authors declare no competing interests.
- 11.Abramson JS. Hitting back at lymphoma: how do modern diagnostics identify high-risk diffuse large B-cell lymphoma subsets and alter treatment? Cancer-Am Cancer Soc. 2019;125(18):3111–20.Google Scholar
- 17.D'Arcy P, Wang X, Linder S. Deubiquitinase inhibition as a cancer therapeutic strategy. PharmacolTher. 2015;147:32–54.Google Scholar
- 24.Tian Z, D'Arcy P, Wang X, Ray A, Tai YT, Hu Y, Carrasco RD, Richardson P, Linder S, Chauhan D, et al. A novel small molecule inhibitor of deubiquitylating enzyme USP14 and UCHL5 induces apoptosis in multiple myeloma and overcomes bortezomib resistance. Blood. 2014;123(5):706–16.CrossRefGoogle Scholar
- 25.Chitta K, Paulus A, Akhtar S, Blake MK, Caulfield TR, Novak AJ, Ansell SM, Advani P, Ailawadhi S, Sher T, et al. Targeted inhibition of the deubiquitinating enzymes, USP14 and UCHL5, induces proteotoxic stress and apoptosis in Waldenstrommacroglobulinaemiatumour cells. Br J Haematol. 2015;169(3):377–90.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.