DEPARTMENT OF PATHOLOGY, UNIVERSITY MEDICAL CENTER GRONINGEN
The role of IL-13 in the pathogenesis of classical Hodgkin lymphoma
Iris Baars S2369885
Master Research Project 1 Biomedical Sciences Under supervision of:
Prof. dr. Anke van den Berg Rianne Veenstra
2
Abstract
The cytokine IL-13 has been suggested to stimulate proliferation of Hodgkin Reed–
Sternberg cells, the malignant cells in classical Hodgkin lymphoma (cHL), through phosphorylation of STAT6. A recent GWAS study found an association between cHL and the rs20541 IL-13 variant, which has been considered to have increased activity. However, the effects of wild-type and variant IL-13 on the pathogenesis of cHL are still unclear.
Therefore, the effects of wild-type and variant IL-13 on cell growth and phosphorylation of STAT6 in cHL were investigated in the current study. We show here that certain cHL cell lines express IL-13 and that phosphorylation of STAT6 is dependent on IL-13. Moreover, antibody-mediated inhibition of IL-13 decreases cell growth in L428, a nodular sclerosis (NS) cell line, but not in L1236, a mixed cellularity cell line. No differences between wild- type and variant IL-13 on cell growth were observed. These results suggest that IL-13 is an important growth factor in certain cHL subtypes, such as NS and that the rs20541 IL-13 variant does not have increased activity compared to wild-type IL-13.
3
Content
Introduction……….4-7 Methods………7-10 Results………10-20 Discussion………20-24 References……….24-27 Supplementary………28-76
4
1. Introduction
1.1 Hodgkin Lymphoma
Hodgkin lymphoma (HL) is a lymph node cancer of germinal center B cell origin and is subdivided into classic Hodgkin lymphoma (cHL) and a nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL). cHL, consisting of the subtypes nodular sclerosis, mixed cellularity, lymphocyte-rich, and lymphocyte-depleted HL, accounts for approximately 95%
of cases, whereas NLPHL represents only 5% of all HL cases (Küppers, 2012). NLPHL is characterized by the presence of lymphocyte-predominant (LP) cells (Küppers, 2012), while cHL is characterized by the presence of rare malignant Hodgkin Reed–Sternberg (HRS) cells (Küppers et al., 2012), which are giant multinucleated cells derived from germinal center (GC) B cells (Schmitz et al., 2009; Küppers, 2012; Agostinelli and Pileri, 2014). HRS cells lack expression of many B lineage genes, including many B cell antigen receptor (BCR) components (Schwering et al., 2003). In contrast, HRS cells do express proteins involved in antigen presentation, such as major histocompatibility complex (MHC) class II antigens, CD40, CD80, and CD86 (Poppema, 1996). The markers CD30 and usually CD15 are also frequently expressed on HRS cells (Liu et al., 2014). However, these HRS cells constitute only 1% of the tumor (Küppers, 2012). The majority of the malignancy is composed of inflammatory cells including lymphocytes, eosinophils, fibroblasts, macrophages, and plasma cells (Skinnider and Mak, 2002; Greaves et al., 2013). Most of the reactive cells are CD4 positive memory T cells (Poppema, 1996), which contribute to the survival of HRS cells (Liu et al., 2014). These CD4 positive T cells show both a transcription factor profile and a cytokine profile that match anergic Th2 and Treg cells (Schreck et al., 2009; Liu et al., 2014).
1.2 Classical Hodgkin Lymphoma
The most common form of cHL, nodular sclerosis (NS) accounts for 75–80% of cHL cases and is characterized by fibrosis originating from thickened lymph node capsule (Pileri et al., 2002; Gobbi et al., 2013). NS is further defined by lacunar cells, a variant of HRS cells, (Pileri et al., 2002; Anagnostou et al., 1977) and is associated with nodular pattern (Pileri et al., 2002; Agostinelli and Pileri, 2014). The second cHL subtype, called mixed cellularity (MC), constitutes about 15–25% of cHL cases (Pileri et al., 2002) and is characterized by a relatively large number of HRS cells and by the presence of plasma cells, epithelioid histiocytes, eosinophils, and T cells (Pileri et al., 2002). The third cHL subtype, called lymphocyte-depleted cHL, is very rare, accounting for approximately 1% of HL cases (Pileri et al., 2002). It shows the worst clinical behavior and prognosis and is associated with a large number of HRS cells and a small number of reactive lymphocytes (Pileri et al., 2002; Agostinelli and Pileri, 2014). The last type, accounting for approximately 6% of cHL cases (Gobbi et al., 2013), is lymphocyte-rich cHL and is characterized by the presence of small lymphocytes, neutrophils, eosinophils and plasma cells and a small number of HRS cells with a classical immunophenotype consistent with cHL (Nam-Cha et al., 2009; Pileri et al., 2002; Anagnostopoulos et al., 2000; Gobbi et al., 2013). The tumor cells in cHL are infected by Epstein–Barr virus (EBV) in approximately 40% of all cHL patients (Küppers, 2012) depending on sex, age, ethnicity and geographic area (Glaser et al., 1997; Elgui et al., 2002; Araujo et al., 2006; Armstrong et al., 1998). Specifically, EBV-positive cHL is more frequently observed in males, children and older adults, non-whites, in less economically developed regions of the world and in patients with the mixed cellularity subtype (Glaser et al., 1997).
5
1.3 Cytokines in Classical Hodgkin Lymphoma
HL is currently treated by chemotherapy (ABVD or BEACOPP) and/or radiation therapy, which is effective in about 70–80% of the HL patients (Pileri et al., 2002).
However, these treatments have rather detrimental side effects (Pileri et al., 2002). That is, current treatment is associated with toxicity, which can lead to infertility, secondary acute myeloid leukaemia and myelodysplastic syndromes (Bauer et al., 2011). Moreover, HL patients are more likely to die from late treatment-related toxicities than from HL itself (Re et al., 2005). New treatment strategies are therefore desirable. Targeting specific cancer survival pathways would be a promising approach to improve HL therapy. A treatment strategy aimed to restore cytokine production could be successful, since cHL is characterized by an abnormal expression of cytokines, produced primarily by the HRS cells (Skinnider et al., 2002). The abnormal expression of cytokines associated with cHL might also explain a number of symptoms related to cHL, such as enlarged lymph nodes, fever, weight loss, and night sweats, the latter three representing the so called B-symptoms (Skinnider et al., 2002), which are observed in approximately 25% of all cHL patients (Pileri et al., 2002).
Typically, a cytokine either acts in a paracrine manner to modulate the activity of surrounding cells, or in an autocrine manner to affect the cell that produced it. In the context of cHL, cytokines produced by HRS cells are thought to contribute to the pathogenesis of this disease both by acting as autocrine growth factors and by regulating the reactive milieu in a paracrine manner (Skinnider and Mak, 2002).
1.4 Interleukin-13
One of the cytokines associated with cHL is interleukin-13 (IL-13), which has multiple immunomodulatory and anti-inflammatory biologic activities (Skinnider et al., 2002). IL-13 has similar functions to IL-4 (Zurawski and de Vries, 1994). That is, when IL- 4Rα associates with the IL-13Rα1 chain to form the type II IL-4R complex, it can bind both IL-4 and IL-13 (Skinnider et al., 2002). This then results in binding of the IL4Ra chain by JAK1 kinase, which phosphorylates a specific tyrosine residue on the IL-4Ra chain. This allows for recruitment of STAT6, a transcription factor that exists in a latent unphosphorylated form in the cytoplasm. JAK1 then phosphorylates tyrosine residue 641 of STAT6 which promotes homodimerization of STAT6 and its translocation to the nucleus. In the nucleus, phosphorylated STAT6 activates transcription of target genes (Skinnider et al., 2002; Leonard and O'Shea, 1998; Wurster et al., 2000; Mikita et al., 1996). Thus, IL-13 functioning is IL-4Rα/STAT6 mediated but IL-4 independent. IL-13 is a cytokine usually produced by Th2 cells, but can also be produced by Th0 cells, Th1 cells, CD8 positive T cells, basophils and mast cells, unlike IL-4, which can only be produced by Th2 cells (de Waal Malefyt et al., 1995; Burd et al., 1995; Ochensberger et al., 1996). IL-13 is especially involved in the humoral immune response andis associated with B cell proliferation and immunoglobulin (Ig) class switching (Zurawski and de Vries, 1994). IL-13 stimulates B cell proliferation upon binding of either anti-CD40 or anti-Ig antibodies and inhibits B cell apoptosis upon CD40 binding (McKenzie et al., 1993; Lømo et al., 1997). Resting normal B cells do not express IL-13, but can be stimulated to express low levels of this cytokine following engagement of CD40, which is highly expressed on HRS cells (Murray et al., 1995;
Poppema, 1996) or EBV infection, which is frequently observed in cHL patients (de Waal Malefyt et al., 1995; Punnonen et al., 1993; Kindler et al., 1995). Furthermore, IL-13 production continues for longer periods than IL-4 production under the same conditions (de Waal Malefyt et al., 1995). Although human T cells do not express the IL-13Rα1 chain (Graber et al., 1998; Zurawski and de Vries, 1994), IL-13 is able to amplify an existing Th2 response by stimulating other cell populations to produce chemokines which attract Th2 cells. IL-13 stimulates monocytes to release macrophage-derived chemokine (MDC)
6
(Bonecchi et al., 1998; Andrew et al., 1998) and fibroblasts to release eotaxin in a STAT6- dependent manner (Matsukura et al., 2001; Hoeck and Woisetschläger, 2001). Both these chemokines recruit Th2 cells through interaction with the chemokine receptors CCR3, which binds eotaxin, and CCR4, which binds MDC, expressed on activated Th2 cells (Sallusto et al., 1998). Eotaxin, otherwise known as CCL11, also contributes to the recruitment of eosinophils, basophils and mast cells (Daugherty et al., 1996; Sallusto et al., 1997; Yamada et al., 1997) which are cells contributing to IL-13 production (de Waal Malefyt et al., 1995;
Burd et al., 1995; Ochensberger et al., 1996) (Figure 1).
Figure 1. IL-13 functioning in cHL. IL-13, produced by Th2 cells, Th1 cells, CD8+ T cells, basophils and mast cells, binds the type II IL-4R complex, consisting of an IL- 4Rα and an IL-13Rα1 chain, which results in JAK1 kinase activation. JAK1 kinase phosphorylates a specific tyrosine residue on the IL-4Ra chain resulting in STAT6 recruitment and STAT6 translocation into the nucleus where it activates transcription. IL- 13, also produced by HRS cells, stimulates fibroblasts to release eotaxin, which can result in collagen production and thereby can cause fibrosis. Eotaxin also contributes to the recruitment of eosinophils, basophils and mast cells, which are cells involved in IL-13 production.
1.5 Interleukin-13 in Classical Hodgkin Lymphoma
Previous studies have demonstrated that IL-13 and IL-13Rα are frequently expressed in HRS cells and that IL-13 may play a role in tumor growth (Skinnider et al., 2002;
Skinnider et al., 2001; Kapp et al., 1999). A study by Skinnider and colleagues showed a decrease in cell proliferation and phosphorylated STAT6 levels upon antibody mediated IL- 13 inhibition in HL cell lines (Skinnider and Elia et al., 2002). Furthermore, an increase in cell proliferation was observed upon exogenous IL-13 exposure (Skinnider and Elia et al., 2002). Their study also observed a correlation between phosphorylated STAT6 and cHL (Skinnider and Elia et al., 2002). Moreover, STAT6 phosphorylation was shown to be IL-13 dependent and inhibition of STAT6 has been shown to induce apoptosis in cHL cell lines (Skinnider and Elia et al., 2002). Furthermore, certain STAT6 target genes in cHL cell lines
7
have been associated with the regulation of proliferation and apoptosis (Baus et al., 2009).
A study by Jundt and colleagues showed that fibroblasts in HL produce eotaxin (Jundt et al., 1999), which may also result from IL-13 stimulation. The effect of IL-13 in cHL might differ between different single-nucleotide polymorphisms (SNPs) in the IL-13 gene as demonstrated before in asthma patients (Vladich et al., 2005). This assumption is in line with the previous finding that the rs20541 SNP at 5q31 located in regions related to the IL- 13 and IL-4 genes is associated with cHL as determined using meta-analysis (Urayama et al., 2012; Cozen et al., 2014). The rs20541 SNP causes the replacement of the positively charged arginine 130 (R130) with a neutral glutamine, which results in increased IL-13 activity and in increased STAT6 phosphorylation (Vladich et al., 2005). These findings together indicate that specific IL-13 inhibition might be a successful target for cHL treatment. The lack in IL-13 upon IL-13 inhibition could even be compensated by IL-4, since IL-4 has similar functions to IL-13, as suggested in a previous study (Skinnider et al., 2002). It is hypothesized that cell growth of HRS cells is stimulated by IL-13 via phosphorylation of STAT6 and that the rs20541 IL-13 variant is more successful in stimulating cell growth of HRS cells than wild-type IL-13. However, the effects of wild-type IL-13 and the variant IL-13 containing the rs20541 SNP on the pathogenesis of cHL are still largely unknown. Therefore, the effect of both wild-type and variant IL-13 on various cHL cell lines will be investigated in the current study.
2. Methods 2.1 Cell cultures
Seven different HL cell lines were used in this study; L428 (nodular sclerosis; DSMZ No ACC 197), L1236 (mixed cellularity; DSMZ No ACC 530), KM-H2 (mixed cellularity;
DSMZ No ACC 8), L591 (nodular sclerosis; DSMZ No ACC 602), L540 (nodular sclerosis;
DSMZ No ACC 72), SUP-HD1 (nodular sclerosis; DSMZ No ACC 574) (all from Leibniz- Institut DSMZ, Braunschweig, Germany) and DEV (NLPHL; cell line developed in-house).
Cells were subcultured three times a week according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). The cell lines L1236, KM-H2 and L591 were cultured in RPMI 1640 media (Cambrex Biosciences, Manassas, VA) supplemented with 10%
fetal calf serum (FCS) (Cambrex Biosciences). L428 was cultured in RPMI 1640 media containing 5% FCS (Cambrex Biosciences) and L540 and DEV were cultured in RPMI 1640 media containing 20% FCS (Cambrex Biosciences). SUP-HD1 was cultured in McCoys5A media (Cambrex Biosciences) containing 20% FCS (Cambrex Biosciences). All media were supplemented with 2 mMUltraglutamine 1, 100 U/ml Penicillin, 100 µg/ml Streptomycin (all Cambrex Biosciences). Cells were counted using a hemacytometer according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). All steps were performed in a safety cabinet and cell cultures were stored at 37°C in a humidified chamber in an atmosphere containing 5% CO2.
2.2 IL-13 inhibition with anti-IL-13
To investigate the effect of IL-13 inhibition on STAT6 and pSTAT6 expression, IL-13 was inhibited with anti-IL-13 (R&D systems, Minneapolis, MN, USA). Cells were diluted to obtain 5∙105 cells/ml in a total of 4 ml RPMI 0% FCS. Cells were then left untreated or were treated with 4µg/ml anti-IL-13 (AF-213-NA, R&D systems). Cells were incubated for 16 hours 37°C in a humidified chamber in an atmosphere containing 5% CO2 after which cell lysates were prepared.
8
2.3 Gene knockdown
To generate lentiviruses for IL-13 knockdown, HEK293T cells were transduced with plasmids harboring mZIP-sh-IL-13#4, mZIP-sh-IL-13#5, mZIP-sh-IL-13#6, mZIP-sh-IL- 13#7 (Table 1) or the controls mZIP-sh-NT1, mZIP-sh-NT2. A mixture of 2,5M CaCl2 and 4 plasmids (2 µg lentivector, 1 µg pMSCV-VSV-G, 1 µg pRSV.REV and 1 µg pMDL-gPRRE) was used to transfect HEK293T cells to produce lentiviral particles. After 48 hours, the supernatant was filtered using a 0,45 µM millex HV PVDF filter (Millipore, Amsterdam, The Netherlands) to harvest the virus. L428 and L1236 were transduced with lentiviral particles and 4 µg/ml polybrene (Sigma, St. Louis MO) to increase the efficiency of transduction.
L428 cells (5∙105 cells/ml) were transduced with 5µl/ml, 50µl/ml or 800µl/ml virus and L1236 cells (5∙105 cells/ml) were transduced 5µl/ml, 50µl/ml or 100µl/ml virus. IL-13 knockdown efficiency was determined by GFP competition assay and Real-Time Quantitative Polymerase Chain Reaction (qRT-PCR). RNA samples and cell lysates of L428 (800µl/ml virus) and L1236 (100µl/ml virus) were prepared on day 8 after transfection.
Table 1. Sequences of short hairpin RNAs to IL-13
Construct Sequence 5’ 3’
IL-13#4 Sense GATCCCGCGAGGGACAGTTCAACTGAATTCAAGAGATTCAGTTGAACTGTCCCTCGCGTTTTTG Antisense AATTCAAAAACGCGAGGGACAGTTCAACTGAATCTCTTGAATTCAGTTGAACTGTCCCTCGCGG IL-13#5 Sense GATCCGGTCAACATCACCCAGAACCTTCAAGAGAGGTTCTGGGTGATGTTGACCTTTTTG Antisense AATTCAAAAAGGTCAACATCACCCAGAACCTCTCTTGAAGGTTCTGGGTGATGTTGACCG IL-13#6 Sense GATCCGACCTGACTATTGAAGTTTTCAAGAGAAACTTCAATAGTCAGGTCTTTTTG Antisense AATTCAAAAAGACCTGACTATTGAAGTTTCTCTTGAAAACTTCAATAGTCAGGTCG
IL-13#7 Sense GATCCGACTATTGAAGTTGCAGATTCATTCAAGAGATGAATCTGCAACTTCAATAGTCTTTTTG Antisense AATTCAAAAAGACTATTGAAGTTGCAGATTCATCTCTTGAATGAATCTGCAACTTCAATAGTCG
2.4 GFP competition assay
GFP expression was measured on a FACS Calibur flow cytometer (BD PharMingen, San Diego, USA) at day 4 post-transfection and monitored for three weeks tri-weekly. The percentage of GFP positive cells was analyzed using FlowJo software (version 10, Treestar, Ashland, OR). The GFP positive percentage was normalized to day 4 post-transfection.
2.5 RNA isolation
RNA was isolated from cHL cell lines L1236, L428, KM-H2, L591, L540, SUPHD1 and DEV and from the IL-13 knockdown samples of both L428 and L1236 using the miRNeasy mini kit (Qiagen GmbH, Hilden, Germany). Samples were prepared using 600ul QIAzol Lysis Reagent (Invitrogen™). Cells were spun down in eppendorf tubes containing Phase Lock Gel Heavy for 30 seconds at 14000 g at RT in a centrifuge with swinging buckets. Tubes containing homogenate were then put at RT for 5 min. Thereafter, the total homogenate was transferred to the Phase Lock Gel containing tubes and 140ul chloroform was added.
Samples were mixed and put at RT for 2-3 minutes. Next, samples were centrifuged for 15 minutes at 12000 g at 4°C. The upper aqueous phase was transferred to a new tube and 1.5 volumes of 100% ethanol were added. Samples were then put in RNeasy Mini columns and centrifuged at ≥8000 g for 15 seconds at RT after which the flow-through was discarded. Next, 700µl Buffer RWT was added to the RNeasy Mini column and columns were centrifuged for 15 seconds at ≥8000 g. Thereafter, 500μl Buffer RPE was added onto RNeasy Mini column and columns were centrifuged for 15 seconds at ≥8000 g. Another 500µl Buffer RPE was added onto RNeasy Mini column and columns were centrifuged for 2 minutes at ≥8000 g after which the columns were placed into new 2ml collection tubes, which were then centrifuged for 1 minute at full speed. The RNeasy Mini column was
9
transferred to a new 1.5ml collection tube and 30–50µl RNase-free water was added directly onto RNeasy Mini column membrane. The columns were centrifuged for 1 minute at ≥8000 g to elute. The RNA concentration was measured using a NanoDrop ND-1000 Spectophotometer (Thermo Fisher Scientific Inc., Waltham, USA) and samples were stored at -80°C
2.6 Real-Time Quantitative PCR
A qRT-PCR was performed to measure IL-13, IL-13Rα1 and IL-4R expression and to measure IL-13 knockdown efficiency on day 8 post-transfection. The experiments were performed in triplicate. cDNA was synthesized with Superscript II reverse transcriptase and random primers according to the company instruction starting from 500ng total RNA (Invitrogen, Carlsbad, USA). PCR was performed using 10ng cDNA as input in a final volume of 30μl containing 5x First-Strand Buffer and 2ul0.1M DTT (final concentration 10mM), 1μl RNaseOUT Recombinant Ribonuclease Inhibitor (40 units/μl) (Invitrogen, Carlsbad, USA) and 1ul Superscript II (200 units). Amplification consisted of 35 cycles using a thermocycler (Bio-Rad, Hercules, USA). PCR products were analyzed on a 3% agarose gel, purified using Zymoclean™ Gel DNA Recovery Kit (Zymo research, Irvine, USA) and the RT-qPCR was run using LightCycler® 480 (Roche, Basel, Switzerland). Agarose gel pictures were captured using Gel Doc XR+ System (Bio-Rad, Hercules, USA).
2.7 Western blot analysis
Western blot analysis was performed to measure STAT6 and pSTAT6 protein expression in the cHL cell lines L1236, L428, KM-H2 and L591. STAT6 and pSTAT6 protein expression was also measured after transduction with mZIP-sh-IL-13 constructs and after antibody-mediated inhibition of IL-13. Whole proteins were extracted using Cell Lysis Buffer containing 0.1% PMSF (Cell Signalling) and stored at -20˚C. A volume of 1 ml Cell Lysis Buffer containing 0.1% PMSF per 20∙106 cells was used. Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Scientific, Massachusetts, USA) and measured on a Varioskan Flash Multimode Reader (Thermo Scientific) at a wavelength of 562 nm with reference to several dilutions of bovine serum albumin (BSA) of known concentration. Samples were prepared and separated on 8% sodium dodecyl sulfate- polyacrylamide gels (0,75mm) by electrophoresis at 130 V for ±90 min for which a protein concentration of 22.5µg was used. Trichloroethanol (TCE) activation was done using Gel DocTM EZ imager (Bio-Rad, Hercules, USA) to confirm presence of proteins. Samples were subsequently transferred onto a 0,45µm Hybond ECL nitrocellulose membrane (Bio-Rad, Hercules, USA) at 100 V for 75 min. Transfer was confirmed by visualization using Gel DocTM EZ imager (Bio-Rad, Hercules, USA). The membranes were blocked for 1 hour in Tris- buffered saline plus Tween20 (TBST) containing 5% skimmed milk (Campina, The Netherlands) at RT to reduce non-specific protein binding. Thereafter, blots were incubated with either rabbit polyclonal anti-STAT6 antibody (clone S-20, Santa Cruz, CA, USA) diluted 1:500 in 5% skimmed milk in TBST or rabbit polyclonal anti-pSTAT6 antibody (clone Y641, Cell Signaling Technology, Danvers, MA) diluted 1:1000 in 5% skimmed milk in TBST overnight at 4ºC. The membranes were subsequently incubated with secondary antibody (Dako, A/S, Glostrup, DK) diluted 1:1000 in 5% skimmed milk in TBST for 1 hour at RT.
Following this, blots incubated with tertiary antibody (Dako, A/S, Glostrup, DK) diluted 1:1000 in 5% skimmed milk in TBST for 1 hour at RT. Prior to incubation with secondary and tertiary antibody, blots were washed three times with TBST. Immunoblots were developed using SuperSignal® West Pico Chemiluminescent Substrate Kit (Thermo Scientific) for which they were incubated with SuperSignal® West Pico Chemiluminescent Substrate (ThermoScientific) for 5 min at RT. Mouse monoclonal anti-GAPDH antibody (clone 0411, Santa Cruz, CA, USA) diluted 1:20,000 in 5% skimmed milk in TBST was used
10
as a control. ChemiDocTM XRS+ System with Image LabTM Software (Bio-Rad, Hercules, USA) was used for visualization and quantification of protein bands, respectively. The expression of STAT6 and pSTAT6 were quantified relative to the GAPDH expression.
2.8 AlamarBlue® assay
AlamarBlue® assay (Invitrogen™, Carlsbad, CA, USA) was used to measure the effect of wild-type and variant IL-13 on cell growth before and after IL-13 knockdown and after treatment with anti-IL-13 neutralizing antibody. AlamarBlue® uses the natural reducing power of living cells to convert the blue resazurin to the red fluorescent molecule, resorufin, resulting from cell growth. Cell growth before IL-13 knockdown was measured in L1236, L428 and KM-H2. Cell growth after IL-13 knockdown was measured in L428 and L1236. L1236 and L428 cells were also used to measure cell growth upon anti-IL-13 treatment. Cells were washed 3 times with RPMI 1640 media containing 0% FCS for 5 minutes at 1200 rpm at RT and were serially diluted to obtain 1∙104 cells in 200 µl RPMI 0%
FCS. L1236, L428 and KM-H2 cells were then left unstimulated or were stimulated with 10ng/ml wild type IL-13 (213-ILB, R&D systems) or 10 ng/ml variant IL-13 (Z02711, GenScript, Piscataway, NJ, USA). The same protocol was used for L428 and L1236 IL-13 knockdown samples. For the treatment with anti-IL-13, L428 and L1236 were left untreated or were treated with 0.2µg/ml, 1µg/ml, 2µg/ml or 4µl/ml anti-IL-13 (AF-213-NA, R&D systems). Hereafter, cells were plated in quadruplicates on a 96 well plate. Next, 10 µl/well AlamarBlue® (BUF012B, Bio-Rad) was added. Cells were incubated at 37°C in a humidified chamber in an atmosphere containing 5% CO2.After 4, 24, 48 and 72 hours, absorption was measured at a wavelength of 570 nm.
2.9 ELISA
IL-13 secretion in the cHL cell lines L1236, L428, KM-H2, L591, L540, SUP-HD1 and DEV was measured by ELISA. IL-13 secretion was also measured in the L428 and L1236 IL- 13 knockdown samples on day 47 after transfection. Cells (5∙105 cells/ml) were washed with RPMI containing 0% FCS and were incubated at 37°C in a humidified chamber in an atmosphere containing 5% CO2. Supernatants of cell cultures were collected after 24 hours and IL-13 secretion was determined using the DuoSet ELISA Development kit and was measured using a Varioskan Flash Multimode Reader (ThermoScientific) at a wavelength of 492 nm. The sensitivity of IL-13 detection was 15.625pg/ml.
2.10 Statistical analysis
Statistical analysis for the AlamarBlue® assay experiments was performed using two-way ANOVA analysis using PRISM. In the present study, a value of p<0.05 was considered significant.
3. Results
3.1 IL-13 expression in HL cell lines
First, IL-13 expression was determined by qRT-PCR in 7 different HL cell lines (Figure 2A). IL-13 expression was observed in SUP-HD1, L428, L1236 and L540. However, IL-13 expression in SUP-HD1 was more than ten-fold higher than in L428, L1236 and L540. No IL- 13 expression was detected in the other HL cell lines. Second, IL-13 secretion in the 7 HL cell lines was measured by ELISA (Figure 2B). IL-13 secretion was observed in SUP-HD1, L1236 and L428. However, IL-13 secretion in SUP-HD1 was more than fifteen-fold higher than in L428 and L1236. No IL-13 secretion was detected in the other HL cell lines. Next, IL-
11
13-Rα1 expression was measured using qRT-PCR (Figure 2C). IL-13-Rα1 expression was detected in all HL cell lines, apart from L540. However, IL-13-Rα1 expression in L591 and DEV was minimal. Lastly, IL-4R expression was determined using qRT-PCR (Figure 2D). IL- 4R expression was observed in all HL cell lines. However, expression in L1236 and KM-H2 was more potent than the expression in the other 5 HL cell lines. These results show that not all HL cell lines produce IL-13. However, IL-13 expression and secretion was observed in SUP-HD1, L428 and L1236. Furthermore, expression of both receptors needed for binding of IL-13 was detected in these cell lines, indicating IL-13 might be able to function as an autocrine growth factor in these cell lines.
IL-13 expression
L591 KM-H
2 DEV
SUPHD1 L540
L1236 L428 0
2 4 6 8
A
Relative expression
IL-13 secretion
L591 KM-H
2 DEV
SUPHD1 L540
L1236 L428 0
100 200 300 400
B
Concentration (pg/ml)
IL13-R1 expression
L591 KM-H
2 DEV
SUPH D1
L540 L1236
L428 0
1 2 3 4
C
Relative expression
IL-4R expression
L591 KM-H
2 DEV
SUPH D1
L540 L1236
L428 0
1 2 3 4
D
Relative expression
Figure 2. IL-13 expression in HL cell lines. (A) IL-13 expression in HL cell lines determined by qRT-PCR. TBP was used as a housekeeping gene. One single experiment is shown and the error bars represent the SD of qRT-PCR measurements in triplicate (mean ± error). (B) IL-13 secretion in HL cell lines determined by ELISA. One single experiment is shown and the error bars represent the SD of ELISA measurements in triplicate (mean ± error). (C) IL-13-Rα1 expression in HL cell lines determined by qRT-PCR. TBP was used as a housekeeping gene. One single experiment is shown and the error bars represent the SD of qRT-PCR measurements in triplicate (mean ± error).
(D) IL-4R expression in HL cell lines determined by qRT-PCR. TBP was used as a housekeeping gene. One single experiment is shown and the error bars represent the SD of qRT-PCR measurements in triplicate (mean ± error).
3.2 STAT6 and pSTAT6 protein expression in cHL cell lines
STAT6 and phospho-STAT6 (pSTAT6) protein expression in 4 cHL cell lines was measured using western blot analysis (Figure 3). The experiment was performed in duplicate (Supplemental Figure 1 and 2). STAT6 protein expression was detected in all cHL cell lines (Figure 3B). pSTAT6 expression was only observed in the cHL cell lines L1236 and L428 (Figure 3C). Thus, pSTAT6 expression was only observed in cHL cell lines that express and secrete IL-13, suggesting IL-13 is needed for the phosphorylation of STAT6.
12
Figure 3. STAT6 and pSTAT6 protein expression in different cHL cell lines determined by western blot analysis. (A) Immunoblots for STAT6 and pSTAT6. A representative western blot is shown. (B) Relative STAT6 expression in the different cHL cell lines. (C) Relative pSTAT6 expression in the different cHL cell lines. GAPDH was used as a housekeeping gene. The average of an experiment performed in duplicate of one single batch is shown (mean and SD).
3.3 Cell growth in cHL cell lines upon IL-13 stimulation
Cell growth upon IL-13 stimulation was measured in L428, L1236 and KM-H2 using AlamarBlue® assay. The displayed results represent the average of three independently performed experiments (Supplemental Figure 3, 4 and 5). No significant differences in cell growth were observed upon IL-13 stimulation in L428 and L1236 (Figure 4A and B). A significant (P<0.001) effect on cell growth was observed upon variant IL-13 stimulation in KM-H2 (Figure 4C). Thus, increasing IL-13 levels in IL-13-producing cells does not affect cell growth. However, stimulation with variant IL-13 does seem to increase cell growth in the cHL cell line KM-H2, which does not express IL-13, but does express the receptors needed for IL-13 binding. These results together suggest that IL-13 is able to stimulate cell growth in certain cHL cell lines and that variant IL-13 might be more successful in stimulating cell growth in cHL than wild-type IL-13.
13
Cell growth L428
4 24 48 72
0 500 1000 1500 2000
WT IL-13 Variant IL-13
A
Unstimulated
Time afte r stimulation (h)
Fluorescence (normalized to 4 hrs) Cell growth L1236
4 24 48 72
0 500 1000 1500
WT IL-13 Variant IL-13
B
Unstimulated
Time afte r stimulation (h)
Fluorescence (normalized to 4 hrs)
Cell growth KM-H2
4 24 48 72
0 500 1000 1500 2000
WT IL-13 Variant IL-13 Unstimulated
***
C
Time after stimulation (h)
Fluorescence (normalized to 4 hrs)
Figure 4. Cell growth in cHL cell lines upon IL-13 stimulation determined by AlamarBlue assay. Cell growth in unstimulated, wild-type (WT) IL-13-stimulated and variant IL-13-stimulated (A) L428 (B) L1236 and (C) KM-H2 cells. The graphs show the average of three independent experiments (mean and SD). Significance was measured using PRISM statistical software and unstimulated cells as control (**p<0.01, ***p<0.001).
3.4 Cell growth in cHL cell lines upon anti-IL-13 treatment
Cell growth in L428 and L1236 was measured upon anti-IL-13 treatment using AlamarBlue® assay. The results shown represent the average of three independently performed experiments (Supplemental Figure 6, 7 and 8). A decrease in cell growth was observed in L428 upon anti-IL-13 treatment. The difference in cell growth is significant in cells treated with 2µg/ml neutralizing anti-IL-13 antibody (P<0.001) compared to cell growth in untreated cells (Figure 5A). Interestingly, the decrease in cell growth was diminished when L428 cells were treated with 4µg/ml anti-IL-13. No differences in cell growth upon anti-IL-13 treatment were observed in L1236 (Figure 5B). Thus, IL-13 seems to play a role in cell growth in L428, but not in L1236.
14
L428 Cell growth
4 24 48 72
0 500 1000 1500
Untreated 0.2 ug/ml anti-IL-13 1 ug/ml anti-IL-13 2 ug/ml anti-IL-13
A
***
4 ug/ml anti-IL-13
Time after treatment (h)
Fluorescence (normalized to 4 hrs)
L1236 Cell growth
4 24 48 72
0 500 1000 1500 2000
Untreated 0.2 ug/ml anti-IL-13 1 ug/ml anti-IL-13 2 ug/ml anti-IL-13
B
4 ug/ml anti-IL-13
Time afte r tre atme nt (h)
Fluorescence (normalized to 4 hrs)
Figure 5. Cell growth in L428 and L1236 upon anti-IL-13 treatment determined by AlamarBlue assay. Cell growth upon anti-IL-13 treatment in (A) L428 and (B) L1236. The graphs show the average of three independent experiments (mean and SD). Significance was measured using PRISM statistical software and untreated cells as control (**p<0.01, ***p<0.001).
3.5 STAT6 and pSTAT6 expression in cHL cell lines upon anti-IL-13 treatment
pSTAT6 and STAT6 protein expression was measured upon anti-IL-13 treatment in L428 and L1236 using western blot analysis (Figure 6). The experiment was performed in duplicate (Supplemental Figure 9-12). No clear differences were observed in STAT6 expression upon anti-IL-13 treatment (Figure 6B and D). A decrease in pSTAT6 protein expression was observed in both L428 (73.0%) and L1236 (80.5%) upon anti-IL-13 treatment (Figure 6E and F). Thus, inhibition of IL-13 leads to a decrease in pSTAT6 protein expression, suggesting phosphorylation of STAT6 is dependent on stimulation by IL-13.
15
Figure 6. STAT6 and pSTAT6 expression in L428 and L1236 upon anti-IL-13 treatment determined by western blot analysis. Immunoblots for STAT6 and pSTAT6 for (A) L428 and (B) L1236. A representative western blot is shown.
Relative STAT6 expression in untreated or anti-IL-13 treated (C) L428 and (D) L1236. Relative pSTAT6 expression in untreated or anti-IL-13 treated (E) L428 and (F) L1236. GAPDH was used as a housekeeping gene. The average of an experiment performed in duplicate of one single batch is shown (mean and SD).
3.6 IL-13 expression in cHL cell lines after IL-13 knockdown
Knockdown efficiency at mRNA level was determined by qRT-PCR in L428 and L1236.
A decrease in IL-13 expression in mZIP-sh-IL-13#5 (62.2%), mZIP-sh-IL-13#6 (61.3%) and mZIP-sh-IL-13#7 (82.4%) samples compared to the average relative IL-13 expression of the non-targeting (NT) controls was observed in L428 (Figure 7A). Results for mZIP-sh- IL-13#4 and the corresponding TBP varied, so data are not shown. No effect of IL-13 knockdown at mRNA level was observed for L1236 (Figure 7B). These results indicate that IL-13 knockdown was successful in L428, but not in L1236.
16
IL-13 expression L428
NT1 NT2
Aver
age NT IL-13#
4 IL-13#
5 IL-13#
6 IL-13#
7 0.0
0.5 1.0 1.5 2.0
62.2% 61.3%
82.4%
Relative expression X
IL-13 expression L1236
NT1 NT2
Aver
age NT IL-13#
4 IL-13#
5 IL-13#
6 IL-13#
7 0.00
0.05 0.10 0.15
20.0%
40.0%
Relative expression
A B
Figure 7. mRNA IL-13 expression in L428 and L1236 IL-13 knockdown samples determined by qRT-PCR. (A) mRNA IL-13 expression in L428 cells transduced with mZIP-sh-NT1, NT2, IL-13#5, IL-13#6 and IL-13#7. Results for mZIP-sh-IL-13#4 and the corresponding TBP varied, so data are not shown. (B) mRNA IL-13 expression in L1236 cells transduced with mZIP-sh-NT1, NT2, IL-13#4, IL-13#5, IL-13#6 and IL-13#7. TBP was used as a housekeeping gene. Non-targeting (NT) shRNAs were used as control. One single experiment is shown and the error bars represent the SD of qRT-PCR measurements in triplicate (mean ± error).
3.7 Cell proliferation in cHL cell lines after IL-13 knockdown
GFP Competition Assay L428
4 6 8 11 13 15 18 20 22
0 50 100
NT1 NT2 IL-13#4 IL-13#5 IL-13#6 IL-13#7
Days afte r infe ction
% GFP+ cells (normalized to day 4)
GFP Competition Assay L1236
4 6 8 11 13 15 18 20 22
0 50
100 NT1
NT2 IL-13#4 IL-13#5 IL-13#6 IL-13#7
Days afte r infe ction
% GFP+ cells (normalized to day 4)
A
B
Figure 8. Cell growth in L428 and L1236 IL-13 knockdown samples determined by GFP competition assay. GFP positive cells in (A) L428 and (B) L1236 cells transduced with mZIP-sh-NT1, NT2, IL-13#4, IL-13#5, IL-13#6 and IL-13#7. Non-targeting (NT) shRNAs were used as control. GFP positive cells are IL-13 negative.
17
Cell proliferation in cHL cell lines after IL-13 knockdown was determined using GFP competition assay (Figure 8). Cell proliferation was measured for three weeks tri-weekly (Supplemental Figure 25 and 26). Less GFP positive cells were observed in mZIP-sh-IL- 13#4, mZIP-sh-IL-13#5 and mZIP-sh-IL-13#7 compared to the NT controls in both L428 (Figure 8A) and L1236 (Figure 8B) indicating a role for IL-13 in cell proliferation. However, the decrease in GFP positive cells in mZIP-sh-IL-13#4, mZIP-sh-IL-13#5 and mZIP-sh-IL- 13#7 transduced cells compared to NT controls was limited.
To investigate the effect of IL-13 knockdown on cell proliferation under stress conditions, cells were cultured without serum or with decreasing levels of serum and monitored for 2 weeks tri-weekly. However, cells started dying immediately when cultured without serum (Supplemental Figure 27 and 28), making pursuing the experiment impossible. Therefore, no results were obtained from this experiment.
3.8 IL-13 secretion in cHL cell lines after IL-13 knockdown
IL-13 secretion L428
WT NT1 NT2
Aver
age NT IL-13#4 IL-13#5
IL-13#6 IL-13#7 0
50 100 150 200
36.5%
20.2%
38.3%
42.1%
Concentration (pg/ml)
IL-13 secretion L1236
WT NT1 NT2
Aver
age NT IL-13#4 IL-13#5
IL-13#6 IL-13#7 0
20 40 60
15.5%
13.3%
21.5%
11.8%
Concentration (pg/ml)
A B
Figure 9. IL-13 secretion in L428 and L1236 IL-13 knockdown samples determined by ELISA. IL-13 secretion in (A) L428 and (B) L1236 cells transduced with mZIP-sh-NT1, NT2, IL-13#4, IL-13#5, IL-13#6 and IL-13#7. Wild- type (WT) and non-targeting (NT) shRNAs were used as controls. One single experiment is shown and the error bars represent the SD of ELISA measurements in quadruplicate (mean ± error).
IL-13 secretion was determined in mZIP-sh-IL-13 transduced cHL cell lines using ELISA. The experiment was performed in quadruplicate. A decrease in IL-13 secretion was observed upon IL-13 knockdown. However, differences in IL-13 secretion observed in mZIP- sh-IL-13 transduced cHL cell lines compared with the average of the NT controls were rather small for both L428 and L1236 (Figure 9). It is important to note that a large difference in IL-13 secretion between L428 NT1 and NT2 was observed. IL-13 secretion in NT1 seems more accurate, since it is more comparable to the IL-13 secretion in L428 wild-type (WT).
When mZIP-sh-IL-13 transduced L428 cells are compared to NT1, no difference in IL-13 secretion can be observed. Thus, no effect on protein level was observed after IL-13 knockdown, suggesting that knockdown of IL-13 was not efficient in L428 and L1236 cells.
3.9 STAT6 and pSTAT6 protein expression in cHL cell lines after IL-13 knockdown
STAT6 and pSTAT6 protein expression in cHL cell lines was measured after IL-13 knockdown using western blot analysis (Figure 10 and 11). The experiment was performed in duplicate (Supplemental Figure 13-16). The relative STAT6 expression was only
18
determined once for mZIP-sh-IL-13#5 and mZIP-sh-IL-13#6, since problems occurred with the housekeeping gene (Supplemental Figure 13). The average of the two NT controls was determined and was used as a reference. No significant differences in STAT6 (Figure 10B) and pSTAT6 (Figure 10C) expression were observed upon IL-13 knockdown in L428 (Figure 10).
A small decrease in STAT6 (Figure 11B) and pSTAT6 (Figure 11C) expression was observed upon IL-13 knockdown in L1236 compared to the average of the NT controls.
However, the decrease in STAT6 and pSTAT6 protein expression was minimal (Figure 11).
Thus, STAT6 and pSTAT6 expression was not affected in L428 knockdown samples. A small decrease in both STAT6 and pSTAT6 expression was observed in L1236 knockdown samples. However, the decrease in pSTAT6 protein expression possibly resulted from the decrease in STAT6 protein expression and not from the IL-13 knockdown. These results seem to support the previous assumption that IL-13 knockdown was not successful.
Figure 10. STAT6 and pSTAT6 protein expression in L428 IL-13 knockdown samples determined by western blot analysis. (A) Immunoblots for STAT6 and pSTAT6 for L428 cells transduced with mZIP-sh-NT1, NT2, IL-13#4, IL- 13#5, IL-13#6 and IL-13#7. A representative western blot is shown. (B) Relative STAT6 expression in the different samples. (C) Relative pSTAT6 expression in the different samples. GAPDH was used as a housekeeping gene. L428 wild-type (WT) and non-targeting (NT) shRNAs were used as controls. The average of an experiment performed in duplicate of one single batch is shown (mean and SD).
19
Figure 11. STAT6 and pSTAT6 protein expression in L1236 IL-13 knockdown samples determined by western blot analysis. (A) Immunoblots for STAT6 and pSTAT6 for L1236 cells transduced with mZIP-sh-NT1, NT2, IL-13#4, IL- 13#5, IL-13#6 and IL-13#7. A representative western blot is shown. (B) Relative STAT6 expression in the different samples. (C) Relative pSTAT6 expression in the different samples. GAPDH was used as a housekeeping gene.
L1236 wild-type (WT) and non-targeting (NT) shRNAs were used as controls. The average of an experiment performed in duplicate of one single batch is shown (mean and SD).
3.10 Cell growth in cHL cell lines after IL-13 knockdown
Cell growth upon IL-13 stimulation was measured in IL-13 knockdown samples on day 40, 47 and day 54 post-transfection using AlamarBlue® assay (Supplemental Figure 17-24). A significant increase in cell growth was observed upon WT IL-13 stimulation in L428 NT1 (P<0.01) and mZIP-sh-IL-13#4 (P<0.01). No significant differences in cell growth were observed upon variant IL-13 stimulation in L428. An overall significant increase in cell growth was observed for WT IL-13 (P<0.05), but not for variant IL-13 in L428 (Figure 12A).
A significant increase in cell growth was observed upon WT IL-13 stimulation in L1236 mZIP-sh-IL-13#4 (P<0.01), mZIP-sh-IL-13#5 (P<0.01) and mZIP-sh-IL-13#7 (P<0.05). A significant increase in cell growth was observed upon variant IL-13 stimulation in L1236 NT1 (P<0.05), mZIP-sh-IL-13#4 (P<0.05) and mZIP-sh-IL-13#5 (P<0.01). An overall significant increase in cell growth was observed for both WT (P<0.05) and variant (P<0.05) IL-13 in L1236 (Figure 12B). No obvious differences in cell growth between controls and IL-13 knockdown samples were observed.
These results suggest that WT IL-13 affects cell growth in L428 more than variant IL- 13, but that both WT IL-13 and variant IL-13 affect cell growth in L1236. However, WT IL- 13 seems to increase cell growth to a greater extent than variant IL-13 in L428 and L1236, suggesting WT IL-13 is more successful in stimulating cell growth in cHL than variant IL-13.
20
Cell growth L428
WT NT1 NT2
IL-13#4 IL-13#5
IL-13#6 IL-13#7 0
500 1000 1500
Unstimulated WT IL-13 Variant IL-13
Fluorescence (normalized to 4 hrs)
Cell growth L1236
WT NT1 NT2
IL-13#4 IL-13#5
IL-13#6 IL-13#7 0
500 1000
1500 Unstimulated
WT IL-13 Variant IL-13
Fluorescence (normalized to 4 hrs)
A
B
Figure 12.Cell growth in L428 and L1236 IL-13 knockdown samples upon 72 hrs of IL-13 stimulation determined by AlamarBlue assay. Cell growth upon wild-type (WT) and variant IL-13 stimulation in (A) L428 and (B) L1236 cells transduced with mZIP-sh-IL-13#4, #5, #6 and #7. The graphs show the average of three independent experiments (mean and SD). Significance was measured using PRISM statistical software and unstimulated cells as control.
4. Discussion
We show that certain cHL cell lines express and secrete IL-13 and that STAT6 expression is frequent in cHL. Furthermore, phosphorylation of STAT6 was exclusively observed in cell lines in which IL-13 production was detected. Stimulation with wild-type and variant IL-13 in IL-13-producing cell lines did not affect cell growth, whereas stimulation with variant IL-13 did affect cell growth in a cell line in which no IL-13 expression was found. Moreover, cell growth and phosphorylation of STAT6 were shown to be inhibited upon treatment with a neutralizing IL-13 antibody. These results indicate that
21
cell growth in certain cHL cell lines is affected by IL-13 and that IL-13 is responsible for phosphorylation of STAT6 in these cell lines.
Our finding that certain cHL cell lines express and secrete IL-13 and express IL- 13Rα1 and IL-14R is supported by other studies, which have demonstrated IL-13 and IL- 13Rα1 expression in cHL cases (Skinnider and Kapp and Mak, 2001) and IL-13Rα1 expression in most cHL cell lines (Skinnider et al., 2001). Furthermore, we detected phosphorylated STAT6 in L428 and L1236, which are cells that showed IL-13 expression and secretion in our study. We did not observe phosphorylated STAT6 in cells that did not show IL-13 production. In contrast to findings in our study, a previous study observed pSTAT6 expression in KM-H2 (Skinnider and Elia et al., 2002), but also detected IL-13 secretion in KM-H2 in a previous study (Kapp et al., 1999). However, other studies supported our results regarding IL-13 production in KM-H2 (Natoli et al., 2013; Kis et al., 2000). The contradicting findings regarding KM-H2 might be explained by heterogeneity in the cell line.
These results taken together indicate that IL-13 stimulates STAT6 phosphorylation, which is supported by previous findings (Skinnider and Elia et al., 2002; Takeda et al., 1996).
Variant IL-13 was shown to be associated with cHL in a recent GWAS (Urayama et al., 2012). However, other studies do not support this finding (Cozen et al., 2012; Enciso- Mora et al., 2010; Frampton et al., 2013). A study by Vladich and colleagues demonstrated that variant IL-13 was more effective in inducing phosphorylation of STAT6 (Vladich et al., 2005). However, we did not observe an obvious difference between wild-type and variant IL-13 functioning. Stimulation with variant IL-13 increased cell growth in KM-H2, whereas stimulation with wild-type IL-13 did not. However, cell growth in L428 and L1236 was not affected upon stimulation with wild-type or variant IL-13, which suggests that IL-13- producing cHL cell lines cannot be further stimulated with either wild-type or variant IL-13.
In contrast to our findings, cell growth in L1236 was shown to be enhanced upon exogenous IL-13 exposure in a study by Skinnider and colleagues (Skinnider and Elia et al., 2002).
However, they used a 10 times higher IL-13 concentration than was used in the present study. Cell growth after IL-13 knockdown in L428 and L1236 was also measured, since it was suggested that these cells already produced IL-13 levels sufficient to stimulate maximal cell growth. In disagreement with the findings by Vladich and colleagues and by our finding that cell growth KM-H2 was more affected by variant-IL-13, wild-type IL-13 seemed more effective in increasing cell growth in knockdown samples than variant IL-13.
The contradicting findings between our study and that of Vladich and colleagues could result from differences in IL-13 secretion between studies. A more than 40 times higher IL-13 secretion had been observed in L428 by Kapp and colleagues (Kapp et al., 1999) compared to the IL-13 secretion measured in our study. Since we observed a more than six-fold difference in IL-13 secretion for L428 and a more than two-fold difference in IL-13 secretion for L1236 in two independent measurements, it seems IL-13 secretion may vary tremendously in time. This suggests that the IL-13 concentration used in our study to stimulate cHL cells might not have been sufficient to elicit an effect on cell growth. However, the IL-13 concentration used in our study (10ng/ml) seemed sufficient for the IL-13 secretion measured in our study in L428 and L1236 (17-108pg/ml and 20-43pg/ml respectively). It would be of interest to measure IL-13 secretion in cHL cell lines on different time points and to repeat IL-13 stimulation experiments with higher wild-type and variant IL-13 concentrations in future experiments.
Another possible explanation for the contradicting findings is the correction factor used in the study by Vladich and colleagues (Vladich et al., 2005). They adjusted the variant IL-13 concentrations for differences in binding affinity between wild-type and variant IL-13 by their ELISA antibodies. It is however questionable if this is a proper way to analyze the results. Moreover, the study by Vladich and colleagues used monocytes obtained from normal human PBMCs for their experiments, while we used cHL cell lines. Another explanation for our results regarding wild-type and variant IL-13 could be related to the receptors involved in IL-13 binding. It has been shown that the rs20541 SNP occurs in α-
22
helix D, the region of IL-13 that is thought to interact with IL-4Rα/IL-13Rα1 heterodimers (Madhankumar et al., 2002). Vladich and colleagues examined the difference in affinity for IL-13Rα2 between wild-type and variant IL-13 (Vladich et al., 2005). It is however more interesting to investigate differences in affinity for the IL-13Rα1, since it has recently been demonstrated that HRS cell lines do not express IL-13Rα2 mRNA (Oshima and Puri, 2001).
Differences in affinity for the IL-13Rα1 receptor might be able to explain the results obtained in our study.
The results obtained in the present study do not show a clear difference between wild-type and variant IL-13 functionality. However, since only small effects on cell growth were found for both wild-type and variant IL-13, it seems of interest to analyze the effect of variant IL-13 versus wild-type IL-13 on cHL pathogenesis in more detail. It would therefore be interesting to test the difference between stimulation with wild-type and variant IL-13 on phosphorylation of STAT6 in cHL cell lines, since variant IL-13 has previously been associated with increased STAT6 phosphorylation (Vladich et al., 2005). Furthermore, cell growth upon IL-13 stimulation using various concentrations of wild-type and variant IL-13 should be measured in multiple cHL cell lines. Another interesting approach would be to investigate the difference in affinity for IL-13Rα1.
The present study demonstrated a decrease in cell growth in L428, but not in L1236 upon inhibition of IL-13 by a neutralizing antibody. In contrast to these findings, a study by Skinnider and colleagues showed that antibody-mediated neutralization of IL-13 in L1236 led to a dose-dependent inhibition of proliferation and to a decrease in STAT6 phosphorylation (Skinnider and Elia et al., 2002). The difference in results found between our studies could result from differences in anti-IL-13 antibody concentrations. The decrease in cell proliferation upon IL-13 inhibition was only observed when a concentration of 20µl/ml was used, which is 5 times higher than the concentration used in our study. The results obtained in the present study suggest that L1236 might be responsive to IL-13, but to a lesser extent than other cHL cell lines. That is, a higher IL-13 secretion was detected in L428, but a more potent decrease in cell growth was detected using a lower anti-IL-13 concentration than the concentration that affected cell growth in L1236. This indicates that L1236 might rely on other growth factors. This is supported by the knowledge that L428 is a NS cell line, whereas L1236 is a MC cell line. NS has previously been shown to be related to fibrosis on which IL-13 was considered to have an effect via the stimulation of fibroblasts (Ohshima et al., 2001; Oriente et al., 2000). Furthermore, a previous study showed that higher numbers of IL-13 positive cells were found in NS cases than in MC cases (Ohshima et al., 2001). Moreover, IL-4, which is another cytokine able to activate STAT6, has recently been shown to be expressed in L1236 cells, whereas no IL-4 expression was observed in L428 cells (Malec et al., 2004). IL-4 might act as a growth factor in L1236, which could explain the difference in the effect of IL-13 inhibition in the two cell lines. However, a decrease in STAT6 phosphorylation upon IL-13 inhibition was observed in both L428 and L1236, which indicates that IL-13 inhibition affects STAT6 phosphorylation in both cell lines.
Nevertheless, the decrease in STAT6 phosphorylation seems to affect cell growth in L428 cells more than in L1236 cells. This indicates that a pathway unrelated to STAT6 may be involved in the cell growth of L1236.
In opposition to the hypothesis regarding the role of IL-13 in NS, a study by Kapp and colleagues treated L428 cells with a 75 times higher anti-IL-13 antibody concentration than the concentration used in the present study and did not find an effect on cell growth.
The finding by Kapp and colleagues was suggested to be due to the vigorous IL-13 secretion measured in L428 (Kapp et al., 1999). However, the amount of anti-IL-13 used in their study was more than 30 times higher than the IL-13 secretion measured in the same study and should not explain the obtained results. The knowledge that IL-13 secretion seems to vary tremendously between experiments might however support their explanation. The difference in results found between our studies could also result from differences in anti-IL- 13 antibody concentrations. That is, the present study showed that cell growth in L428 cells
