Deciphering the Elevated Lipid Via CD36 in Mantle Cell Lymphoma

cancers

Article Deciphering the Elevated Lipid via CD36 in Mantle Cell Lymphoma with Bortezomib Resistance Using Synchrotron-Based Fourier Transform Infrared Spectroscopy of Single Cells

Sudjit Luanpitpong 1,* , Montira Janan 1, Kanjana Thumanu 2, Jirarat Poohadsuan 1, Napachai Rodboon 1, Phatchanat Klaihmon 1 and Surapol Issaragrisil 1,3,4

1 Siriraj Center of Excellence for Stem Cell Research, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand; [email protected] (M.J.); [email protected] (J.P.); [email protected] (N.R.); [email protected] (P.K.); [email protected] (S.I.) 2 Synchrotron Light Research Institute (Public Organization), Nakhon Ratchasima 30000, Thailand; [email protected] 3 Division of Hematology, Department of Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand 4 Bangkok Hematology Center, Wattanosoth Hospital, BDMS Center of Excellence for Cancer, Bangkok 10310, Thailand * Correspondence: [email protected]; Tel.: +66-2-419-2907

Received: 18 February 2019; Accepted: 22 April 2019; Published: 24 April 2019

Abstract: Despite overall progress in improving cancer treatments, the complete response of mantle cell lymphoma (MCL) is still limited due to the inevitable development of drug resistance. More than half of patients did not attain response to bortezomib (BTZ), the approved treatment for relapsed or refractory MCL. Understanding how MCL cells acquire BTZ resistance at the molecular level may be a key to the long-term management of MCL patients and new therapeutic strategies. We established a series of de novo BTZ-resistant human MCL-derived cells with approximately 15- to 60-fold less sensitivity than those of parental cells. Using gene expression profiling, we discovered that putative cancer-related genes involved in drug resistance and cell survival tested were mostly downregulated, likely due to global DNA hypermethylation. Significant information on dysregulated lipid metabolism was obtained from synchrotron-based Fourier transform infrared (FTIR) spectroscopy of single cells. We demonstrated for the first time an upregulation of CD36 in highly BTZ-resistant cells in accordance with an increase in their lipid accumulation. Ectopic expression of CD36 causes an increase in lipid droplets and renders BTZ resistance to various human MCL cells. By contrast, inhibition of CD36 by neutralizing antibody strongly enhances BTZ sensitivity, particularly in CD36-overexpressing cells and de novo BTZ-resistant cells. Together, our findings highlight the potential application of CD36 inhibition for BTZ sensitization and suggest the use of FTIR spectroscopy as a promising technique in cancer research.

Keywords: mantle cell lymphoma; bortezomib resistance; lipid metabolism; CD36; fourier transform infrared spectroscopy; single cells

1. Introduction Mantle cell lymphoma (MCL) is an aggressive non-Hodgkin lymphoma (NHL) that causes signiﬁcant morbidity presumably due to inevitable development of drug resistance [1,2]. With the clinical pattern of poor response to conventional chemotherapy and frequent relapse, MCL remains the

Cancers 2019, 11, 576; doi:10.3390/cancers11040576 www.mdpi.com/journal/cancers Cancers 2019, 11, 576 2 of 21 worst prognosis among all NHL subtypes [3,4]. Bortezomib (BTZ) is a Food and Drug Administration (FDA)-approved first-in-class proteasome inhibitor for the treatment of relapsed/refractory MCL. While a response rate of 30–50% in patients with relapsed disease could be achieved after BTZ treatment irrespective of their sensitivity to prior therapy, more than half of all MCL patients either did not initially respond to BTZ or subsequently acquired resistance [1,5]. Multiple combinations of BTZ and cytotoxic chemotherapy, immunomodulatory drug and/or histone deacetylase inhibitors have been intensively investigated in clinical trials to get higher response and better tolerability [6–8], implying that BTZ would remain one of the core drugs in treatment regimens and that understanding BTZ resistance is necessary for the long-term management of MCL patients and new therapeutic strategies. Drug resistance is a manifestation of cancer, stemming from somatic mutations and genomic plasticity [9]. In this regard, numerous mechanisms associated with molecular changes of drug efflux pumps, drug metabolizing enzymes and apoptosis regulatory pathways have been shown to protect various cancers against chemotherapy [10–12]. Although the information on BTZ resistance in MCL is still very limited, certain characteristics of BTZ-resistant MCL have previously been reported, namely uncontrolled plasmacytic differentiation and activation of B-cell receptor (BCR) signaling [13,14]. Since drug resistance is a multi-factorial phenomenon, we established de novo BTZ-resistant MCL cells and performed large-scale expression analyses at the transcriptional and macromolecular levels using sets of quantitative polymerase chain reaction (PCR) and synchrotron-based Fourier transform infrared (FTIR) spectroscopy. In the present study, we profiled putative cancer-related genes involved in drug resistance and cell survival and observed that most of the tested genes were unexpectedly downregulated in BTZ-resistant MCL. Further single cell analysis with FTIR spectroscopy, however, provided significant information on the dysregulation of lipid in highly resistant cells, leading us to the discovery of an alteration in CD36 receptor that modulates lipid homeostasis. Our findings could be important in understanding BTZ resistance and in developing novel effective adjuvant for MCL.

2. Results

2.1. Establishment of BTZ-Resistant MCL Cells and Their Sensitivity to BTZ BTZ-resistant MCL cells with acquired resistance at diﬀerent degrees were established from BTZ-sensitive human MCL-derived Jeko-1 (J/Parent) cells after sequential treatments with BTZ, i.e., up to 500 nM, and designated as J/BTZ100R, J/BTZ250R and J/BTZ500R (Figure1A). The sensitivity of parental and resistant cells to BTZ at 24 h was ﬁrst evaluated by Hoechst 33342 assay. Hoechst dye stains genomic DNA, allowing the determination of morphologic apoptosis of chromatin condensation and DNA fragmentation. Dose-response curve of BTZ-induced apoptosis was plotted (Figure1B) and lethal concentration 50 (LC50), which causes half of the cells to undergo apoptosis, was calculated from the plot. The LC50 value (mean SE) for parental J/Parent cells was 6.39 1.59 nM versus 87.81 1.14, ± ± ± 236.30 1.22 and 361.10 1.28 nM for resistant J/BTZ100R, J/BTZ250R and J/BTZ500R, respectively. ± ± Resistance index (R) was then calculated according to the relation: R = LC50 resistant cells/LC50 parental cells. Therefore, the degrees of resistance in established resistant cells were approximately 15, 40 and 60-fold more resistant to BTZ than the parental cells. Cancers 2019, 11, 576 3 of 21 Cancers 2019, 11, x 3 of 21

BTZ BTZ BTZ

J/Parent J/100R J/250R J/500R Annexin V+ or PI+

120 J/Parent    J/100R 80   J/250R     J/500R  40 

% Apoptosis % 

    0 01234 Log [BTZ] (nM) Figure 1.1. ResistantResistant phenotypephenotype of of de de novo novo bortezomib bortezomib (BTZ)-resistant (BTZ)-resistant mantle mantle cell lymphomacell lymphoma (MCL) (MCL) cells incells comparison in comparison to parental to parental MCL cells. MCL (A )cells. Schematic (A) Schematic diagram for diagram the establishment for the establishment of BTZ-resistant of MCLBTZ- cellsresistant from MCL parental cells (Jfrom/Parent) parental cells (J/Parent) by stepwise cells selection by stepwise method. selection (B)J/ Parentmethod. cells (B)and J/Parent BTZ-resistant cells and cellsBTZ-resistant at diﬀerent cells degrees at different (J/100R, Jdegrees/250R and (J/100R J/500R), J/250R were treated and J/500R) with various were concentrationstreated with various of BTZ (0–5000concentrations nM) and of apoptosisBTZ (0–5000 was nM) determined and apoptosis by Hoechst was determined 33342 at 24 by h. Hoechst Dose-response 33342 at curves 24 h. Dose- were generatedresponse curves on a logarithmic were generated scale on of druga logarithmic concentration scale of as drug opposed concentration to a linear as scale opposed of percentage to a linear of apoptosis.scale of percentage Lethal concentration of apoptosis. 50Lethal (LC50) concentrat of BTZ inion parental 50 (LC50) and of resistant BTZ in parental cells were and then resistant calculated cells fromwere thethen plot calculated and compared. from the plot and compared.

2.2. Gene Expression Profiling Profiling of BTZ-Resistant MCL In order order to to identify identify genes genes and and pathways pathways import importantant in inthe the development development of BTZ of BTZ resistance, resistance, we weprofiled profiled putative putative cancer-related cancer-related genes genes involving involving inin drug drug resistance resistance and and cell cell survival survival of of MCL using sets of quantitative PCR. These genes werewere grouped into:into: (i) Drug efflux efflux pumps ABCB1, ABCC1 and ABCG2 [[10];10]; (ii) drugdrug metabolizingmetabolizing enzymesenzymes CYP1A2, CYP2D6 and CYP3A4 [[11];11]; (iii)(iii) apoptosisapoptosis regulators BCL2, BCL2L1, MCL1, BAX and TP53 [12];[12]; (iv) cell cycle regulators CCND1,, CDK2,, CDK4,, CDKN18 and CDKN2A [[15];15]; (v) cell survival and growth growth factors NFKB2, MYC,, SOD1,, STK38STK38,, STK38LSTK38L,, and IGF1RIGF1R [16–18];[16–18]; (vi) (vi) Hippo Hippo pathway pathway components components AMOTAMOT, NF2, ,NF2 LATS1, LATS1, LATS2, LATS2, STK3,, STK3STK4, YAP1STK4,, YAP1TAZ and, TAZ KIBRAand KIBRA [19]; (vii)[19 ];epithelial–mesenchymal (vii) epithelial–mesenchymal transition transition (EMT) (EMT)regulators regulators SNAIL2SNAIL2, VIM, ,ZEB1VIM,, ZEB1CTNBB1, CTNBB1, CCND1, CCND1 and CDH1and CDH1[20]; and[20 ];(viii) and pluripotent (viii) pluripotent genes genesOCT4,OCT4 NANOG, NANOG and SOX2and SOX2[21]. Gene[21]. Geneexpression expression levels levels normalized normalized to the to thehousekeeping housekeeping GAPDHGAPDH comparingcomparing parental parental J/Parent J/Parent cells cells and BTZ-resistant JJ/100R,/100R, JJ/250R/250R and JJ/500R/500R cells were averaged, clustered and represented in a heatmap. heatmap. Figure 22AA shows that that most most of of the the tested tested genes genes were were unexpectedly unexpectedly downregulated downregulated in all in resistant all resistant cells. cells.Next, Next,a volcano a volcano plot was plot created was created to better to visualize better visualize the differential the differential gene expression gene expression in J/Parent in J/ Parentversus versusJ/500R cells J/500R using cells more using than more two-fold than two-fold change changeat p < 0.05 at p filter< 0.05 criteria. filter criteria. Figure 2B Figure shows2B showsthat 34% that of 34%these of genes, these genes,including including the well-known the well-known drug metabolizing drug metabolizing enzyme enzyme CYP2D6CYP2D6, proto-oncogenes, proto-oncogenes TAZ TAZand andYAP1YAP1, drug, drug efflux effl uxpump pump ABCC1ABCC1 andand survival survival transcription transcription factor factor NFKB2NFKB2 werewere significantly significantly downregulated in in J/500R J/500R cells cells when when compared compared to toparental parental cells. cells. One Oneimportant important mechanism mechanism for gene for generegulation regulation is epigenetic is epigenetic alterations, alterations, particularly particularly DNA DNA methylation methylation of CpG of CpG dinucleotides dinucleotides [22]. [ 22To]. Toinvestigate investigate the theplausible plausible linkage linkage between between DNA DNA methylation methylation and the and observed the observed gene genesilencing silencing in BTZ- in BTZ-resistantresistant cells, cells,we evaluated we evaluated the global the global DNA DNA methylation methylation status status by specifically by specifically measuring measuring the level the levelof 5-methylcytosine of 5-methylcytosine (5-mC), (5-mC), which which is the is themo mostst common common epigenetic epigenetic mark mark and and an an importantimportant transcriptional repressor [[23].23]. An extensive increa increasese in 5-mC was found in BTZ-resistant JJ/250R/250R and JJ/500R/500R cells when when compared compared to to parental parental cells cells in in a dose-dependent a dose-dependent manner manner (Figure (Figure 2C),2C), indicating indicating the thehypermethylation hypermethylation status status of J/250R of J/250R and and J/500R J/500R ce cells.lls. Treatment Treatment of of these these BTZ-resistant BTZ-resistant cells with hypomethylating agent 5-aza-2′0-deoxycytidine-deoxycytidine (DAC) (DAC) significantly significantly reactivated their expression of top-ranked differentially expressed genes CYP2D6, TAZ and NFKB2 (Figure 2D), suggesting that hypermethylation might be in part responsible for such gene expression perturbation. Cancers 2019, 11, 576 4 of 21 top-ranked differentially expressed genes CYP2D6, TAZ and NFKB2 (Figure2D), suggesting that hypermethylationCancers 2019, 11, x might be in part responsible for such gene expression perturbation. 4 of 21

A C 4 * * 2

CYP2D6 TAZ Level 5-mC Relative NFKB2 0 YAP1 ABCC1 LATS2 STK4 CDH1 NF2 D TP53 CYP2D6 CTNNB1 BAX 2 STK38 KIBRA VIM LATS1 1 # NANOG # OCT4

MCL1 * *

CDK1 Relative Expression CDK2 0 IGF1R SOD1 STK38L CDKN2A BCL2 BCL2L1 TAZ STK3 6 CCND1 # ABCB1 ABCG2 4 CDKN1B MYC # CYP3A4 2 CYP1A2 # SOX2

Relative Expression * * 0 B 10−16 STK4 YAP1 KIBRA NFKB2 TAZ 3 VIM # −8 TP53 10 STK38 Value LATS1 2 P CDH1 # # CYP2D6 ABCC1 NFKB2 1 P = 0.05 BAX * 100 * Relative Expression 0 −7 −2 0 2 7 Log 2 Fold Change FigureFigure 2. 2.Analysis Analysis of of mRNA mRNA expression expression ofof putativeputative cancer-relatedcancer-related genes genes involved involved in in drug drug resistance resistance andand cell cell survival survival of of MCL MCL by by quantitative quantitative real-timereal-time PCR. ( A)) Data Data from from three three independent independent experiments experiments werewere normalized normalized to to house-keeping house-keepingGAPDH GAPDH,, averagedaveraged and represented represented in in heatmap heatmap in in the the three-color three-color scalescale of of red red (upper), (upper), black black (midpoint) (midpoint) and and greengreen (lower).(lower). ( (BB)) Volcano Volcano plot plot of of differentially diﬀerentially expressed expressed genesgenes comparing comparing J/Parent J/Parent and and J/ 500RJ/500R cells cells with with fold fold changechange >>22 and and pp << 0.05;0.05; two-sided two-sided Student’s Student’s t-test.t-test. (C)(C Global) Global DNA DNA methylation methylation was was evaluated evaluated byby measurementmeasurement of 5-methylcytosine (5-mC). (5-mC). Relative Relative absorbance (OD450) of samples to parental cells are shown. Data are mean ± SD (n = 3). * p < 0.05 absorbance (OD450) of samples to parental cells are shown. Data are mean SD (n = 3). * p < 0.05 versus J/Parent cells; two-sided Student’s t-test. (D) J/Parent, J/250R and J/500R cells± were sequentially versus J/Parent cells; two-sided Student’s t-test. (D)J/Parent, J/250R and J/500R cells were sequentially treated with 5-aza-2′-deoxycytidine (DAC) (1 μM) for 72 h and mRNA expression of CYP2D6, TAZ treated with 5-aza-20-deoxycytidine (DAC) (1 µM) for 72 h and mRNA expression of CYP2D6, TAZ and and NFKB2 were evaluated by quantitative real-time PCR. Data are mean ± SD (n = 3). * p < 0.05 versus NFKB2 were evaluated by quantitative real-time PCR. Data are mean SD (n = 3). * p < 0.05 versus J/Parent cells; two-sided Student’s t-test. # p < 0.05 versus non-treated cells;± two-sided Student's t-test. J/Parent cells; two-sided Student’s t-test. # p < 0.05 versus non-treated cells; two-sided Student’s t-test.

Cancers 2019, 11, 576 5 of 21

2.3. FTIR Signatures of BTZ-Resistant MCL FTIR spectroscopy is a sensitive technique capable of providing significant information on the biochemical content of cellular macromolecules, such as proteins, lipids, nucleic acids and carbohydrates, that could be linked to cellular organization and physiological state of the cells [24,25]. Figure3A shows the mean FTIR spectra of the parental J/Parent cells and BTZ-resistant J/250R and J/500R cells 1 recorded from more than 200 single cells in the mid-IR region of 4000 to 800 cm− . Principle component analysis (PCA) is one of the most common multivariate techniques that concentrates the source of variability in the data into few PCs. The PCA of spectroscopic data provide two types of information: (i) visualization of clustering of similar spectra within datasets in scatter plots; and (ii) identification of variables (spectral bands representing various molecular groups within the samples) in loading plots. Figure3B (top) shows the scatter plot of PC1 (PC score; 52%) against PC2 (19%) that could separate the parental cells from BTZ-resistant cells. In brief, J/Parent cells clustered in the positive direction of PC1, while BTZ-resistant cells, particularly J/500R cells, were mostly in the negative direction of PC1 in correlation with the spectral region centered at 2921 (C–H total lipid), 1747 (C=O ester lipid) and 1 1243 (nucleic acid) cm− (Figure3B, bottom). For J /250R cells and J/500R cells, the separation along PC2 can also be explained by the positive loading for PC2 in the spectral region centered at 2917 and 2850 (C–H lipid), 1633 (β-sheet amide I), 1240 (nucleic acid) and 1078 (glycoprotein and carbohydrate 1 1 region) cm− . Hierarchical cluster analysis for the spectral regions from 3000–2800 cm− to 1750–800 1 cm− was performed as a second method to classify and determine the similarity of FTIR spectra of J/Parent, J/250R and J/500R cells. Figure3C shows that the FTIR spectra of J /250 and J/500R cells harbored significant similarities (first group) whereas the J/Parent cells were clustered separately as a second group, indicating that FTIR spectral signatures could clearly distinguish BTZ-resistant cells from the parental cells. To gain more insight into the differences between parental and BTZ-resistant cell lines, various biochemical contents of J/Parent, J/250R and J/500R cells were compared using integral areas of 1 specified FTIR spectral regions corresponding to C–H from lipid (3000–2800 cm− ), C=O ester primarily 1 1 1 from lipid (1750–1735 cm− ), amide I protein (1670–1600 cm− ), amide II protein (1560–1500 cm− ), 1 P=O phosphodiester bond from nucleic acid (1260–1200 cm− ), C–C from nucleic acid and C–O from 1 1 glycoprotein and other carbohydrate (nucleic acid and others; 1180–990 cm− ) and RNA (975–950 cm− ), as illustrated in Figure4A (see also Supplementary Table S1). A remarkable increase of total lipid and ester (lipid) and decrease of amides and nucleic acid and others were observed in highly BTZ-resistant J/500R cells when compared to J/Parent cells (Figure4B). RNA and solely nucleic acid contents were found to be unchanged. These discrepancies in biochemical contents between parental and BTZ-resistant cells might be due to the metabolic reprogramming that is an emerging hallmark of cancer and an attractive target in novel therapeutic strategies for cancer treatment. Cancers 2019, 11, 576 6 of 21 Cancers 2019, 11, x 6 of 21

J/500R Absorbance

J/250R J/Parent

3500 3000 2500 2000 1500 1000

Wavenumber (cm−1)

B C 0.4

J/250R 0.3 J/250R 1 J/500R 0.2 J/500R J/Parent 2 J/Parent 3 0.1 4 5 0.0 J/250R 6 PC2 (19%) PC2 −0.1 7 8 −0.2 9 10 −0.3 1 −0.4 −0.2 0.0 0.2 0.4 0.6 3 PC1 (52%) 10 2 4 J/500R 5 0.10 PC1 6 9 0.00 7 2921 8

1 − 0.10 1243 1747 4 1240 5 0.15 PC2 3 Loadings 2917 2852 7 1078 J/Parent 1633 9 0.00 6 4 − 0.10 8 10 3000 2900 1800 1600 1400 1200 1000 Wavenumber (cm−1) Figure 3. FTIRFTIR spectral spectral signatures signatures of of parent parentalal and BTZ-resistant MCL cells. ( (AA)) Mean Mean FTIR FTIR spectra spectra of of thethe parental JJ/Parent/Parent cellscells andand BTZ-resistantBTZ-resistant J /J/250R250R and and J /J/500R500R cells cells in in the the wavelength wavelength range range of of 4000 4000 to 1−1 to800 800 cm cm− .(. B(B)) (top) (top) Two-dimensional Two-dimensional principal principal componentcomponent analysisanalysis (PCA)(PCA) scorescore plot of all recorded FTIR spectra of J/Parent, J/Parent, J/250R J/250R and JJ/500R/500R cells. Black Black dots dots represent J/Parent J/Parent spectra, while black diamonds representrepresent J/J/250R250R and and grey grey triangles triangles represent represent J/500R J/500R cells. cells. Eclipses Eclipses depicted depicted in the in plot the define plot defineconfidence confidence limit, of limit, which of 95%which of the95% data of the are data allocated. are allocated. (Bottom) (Bottom) Score loadings Score loadings of PC1 (upper) of PC1 and(upper) PC2 and(lower) PC2 to (lower) identify to the identify variables the correspondingvariables corresponding to wavelength to wavelength number. (C )number. Unsupervised (C) Unsupervised hierarchical 1 hierarchicalclassification classification of all recorded of al Jl/ recordedParent, J/ 250RJ/Parent, and J/250R J/500R and spectra J/500R in spectra the range in the of range3000–2800 of 3000–2800 cm− to 1 cm1750–800–1 to 1750–800 cm− . cm–1. Cancers 2019, 11, 576 7 of 21 Cancers 2019, 11, x 7 of 21

A Ester Amide Nucleic Nucleic acid Lipid (Lipid) I & II acid and others

1241 970 2962 Derivative 1515

nd 2854 2 2923 1546

1656

3000 2900 2800 1800 1600 1400 1200 1000

Wavenumber (cm−1)

B Lipid Ester (Lipid) 0.02 * 0.00050 * *

0.01 0.00025 Integral Area Integral Integral Area Integral

0.00 0.00000

Amide I and II Nucleic acid and others 0.050 0.008 * * *

0.025 0.004 Integral Area Integral Area Integral

0.000 0.000

Figure 4.4. SpectralSpectral assignments assignments and and quantiﬁcation quantification of biochemical of biochemical contents contents in parental in parental and BTZ-resistant and BTZ- resistantcells using cells integral using area.integral (A )area. The ( secondA) The derivativesecond derivative spectra obtainedspectra obtained from the from mean the FTIR mean spectra FTIR spectraof the parental of the parental J/Parent J/Parent cells and cells BTZ-resistant and BTZ-resistant J/250R J/250R and Jand/500R J/500R cells cells in the in the wavelength wavelength range range of 1 –1 1–1 of3000–2800 3000–2800 cm cm− to to 1750–800 1750–800 cm cm− .. BandBand assignmentsassignments forfor thethe regionsregions ofof lipid,lipid, ester (lipid), amide amide I, I, amide II,II, nucleicnucleic acidacid (DNA (DNA/RNA)/RNA) and and nucleic nucleic acid, acid, glycoprotein glycoprotein and and other other carbohydrate carbohydrate (nucleic (nucleic acid acidand others)and others) were were illustrated. illustrated. (B) Integral (B) Integral area area of total of total lipid, lipid, ester ester (lipid), (lipid), combined combined amide amide I and I IIand and II andnucleic nucleic acid acid and others,and others, with with observed observed changes, changes, were were plotted. plotted. Data Data are mean are meanSD ± ( nSD= 3).(n = * p3).< *0.05 p < ± versus0.05 versus J/Parent J/Parent cells; cells; two-sided two-sided Student’s Student’st-test. t-test.

2.4. Elevated Elevated Lipid Content and CD36 Upregulation in BTZ-Resistant MCL Lipids, i.e., long chain fatty acids and cholesterols,cholesterols, are essential components of biological membranes and are also used for energy storage, production and cellular signaling [[26,27].26,27]. Enhanced uptake of exogenous lipids and/or and/or overactivated endogenous synthesis provide rapidly proliferating tumor cells with energy and biomassbiomass component,component, conferring both growth andand survivalsurvival advantagesadvantages that are associatedassociated withwith carcinogenesiscarcinogenesis andand tumortumor progressionprogression [[28,29].28,29]. Having demonstrated an increase in cellular lipid composition in BTZ-resistant J/500R cells from FTIR spectral analysis, we next examined the stored lipids using hydrophobic oil red O dye, which binds to lipid droplets due Cancers 2019, 11, 576 8 of 21

increaseCancers 2019 in, cellular 11, x lipid composition in BTZ-resistant J/500R cells from FTIR spectral analysis, we8 next of 21 examined the stored lipids using hydrophobic oil red O dye, which binds to lipid droplets due to theirto their nonpolar nonpolar nature. nature. Figure Figure5A shows5A shows that that J /500R J/500R cells cells had had more more than than three-fold three-fold higher higher in oil in redoil red O stainingO staining when when compared compared to parental to parental cells. cells. Representative Representative micrographs micrographs of oil redof oil O-stained red O-stained cells show cells thatshow an that accumulation an accumulation of lipid dropletsof lipid (reddroplets colored (red dots) colored could dots) be apparently could be visualizedapparently in visualized J/500R cells in (FigureJ/500R5 cellsB). (Figure 5B).

A C CD36 6 6 *

4 * 4

2 2 Relative Oil Red O Relative Expression Relative 0 0

THRSP B 1.5 J/Parent

1.0 *

0.5 Relative Expression 0.0

J/500R

FASN 1.5

1.0 * 0.5 Relative Expression 0.0

FigureFigure 5. 5.Determination Determination of lipidof lipid droplets droplets and expressionand expressi ofon lipid of markerslipid markers in parental in parental and BTZ-resistant and BTZ- cells.resistant (A,B )cells. Parental (A,B J)/ ParentParental cells J/Parent and BTZ-resistant cells and BTZ-resistant J/250 and J/500R J/250 cells and were J/500R ﬁxed cells and were stained fixed with and oilstained red O. with The stainedoil red O. cells The were stained then cells visualized were then under visualized a microscope under or a extracted microscope with or isopropanol extracted with for quantiﬁcation.isopropanol for (A )quantification. Relative absorbance (A) Relative (OD510) absorbance of samples to(OD510) parental of cells samples are shown. to parental Data are cells mean are SD (n = 4). * p < 0.05 versus J/Parent cells; two-sided Student’s t-test. (B) Representative micrographs ±shown. Data are mean ± SD (n = 4). * p < 0.05 versus J/Parent cells; two-sided Student’s t-test. (B) ofRepresentative oil red O-stained micrographs J/Parent and of Joil/500R red cells.O-stai Scalened J/Parent bar = 50 andµm. J/500R (C) mRNA cells. expression Scale bar of= 50 fatty μm. acid (C) synthasemRNA expression (FASN), thyroid of fatty hormone acid synthase responsive (FASN spot), 14thyroid (THRSP hormone) and CD36 responsiveusing quantitative spot 14 (THRSP real-time) and PCR in J/Parent cells in comparison to J/500R cells. Data are mean SD (n = 4). * p < 0.05 versus CD36 using quantitative real-time PCR in J/Parent cells in comparison± to J/500R cells. Data are mean J/±Parent SD (n cells;= 4). * two-sided p < 0.05 versus Student’s J/Parentt-test. cells; two-sided Student’s t-test.

We assessed the gene expression of markers of de novo fatty acid synthesis (FASN) and thyroid We assessed the gene expression of markers of de novo fatty acid synthesis (FASN) and thyroid hormone responsive spot 14 (THRSP) and exogenous fatty acid uptake CD36 [30,31] and observed a hormone responsive spot 14 (THRSP) and exogenous fatty acid uptake CD36 [30,31] and observed a dramatic upregulation of CD36 with subtle changes in FASN and THRSP in J/500R cells when compared to parental cells (Figure 5C). Hence, we postulated that an increase in lipids in BTZ- resistant J/500R cells was likely mediated by an increase in exogenous uptake via CD36 receptor. Flow Cancers 2019, 11, 576 9 of 21 dramatic upregulation of CD36 with subtle changes in FASN and THRSP in J/500R cells when compared to parental cells (Figure5C). Hence, we postulated that an increase in lipids in BTZ-resistant J /500R Cancers 2019, 11, x 9 of 21 cells was likely mediated by an increase in exogenous uptake via CD36 receptor. Flow cytometric analysiscytometric further analysis revealed further an revealed increase inan CD36 increase protein in CD36 expression protein on expression both cell surfaceon both and cell intracellular surface and compartmentintracellular compartment of J/500R cells of when J/500R compared cells when to parentalcompared cells to parental (Figure6 ).cells (Figure 6).

Surface CD36 Intracellular CD36

Unstained J/Parent Unstained J/Parent

0.4% 0.9% 6

SSC SSC

0 2 2 3 4 5 CD36-APC CD36-APC

J/Parent J/Parent

23.4% 51.2%

SSC SSC

2 3 4 5 CD36-APC CD36-APC

J/500R J/500R

75.4% 81.5%

SSC SSC

CD36-APC CD36-APC FigureFigure 6.6. CD36CD36 is is overexpressed overexpressed in in BTZ-resistant BTZ-resistant cells. cells. Flow Flow cytometry cytometry analysis analysis of surface of surface (left) (left) and andintracellular intracellular (right) (right) CD36 CD36 on onparental parental J/Parent J/Parent ce cellslls and and BTZ-resistant BTZ-resistant J/500R J/500R cells. Percentage ofof CD36-positiveCD36-positive cellscells (box)(box) werewere determineddetermined basedbased onon theirtheir internalinternal negativenegative controlcontrol (unstained (unstained cells). cells).

2.5.2.5. CD36 Renders MCL Cells toto BTZBTZ ResistanceResistance andand IsIs anan AttractiveAttractive Co-TargetCo-Target ToTo firstfirst evaluate whether whether CD36 CD36 rendered rendered MCL MCL cells cells to toacquire acquire apoptosis apoptosis resistance resistance to BTZ, to BTZ, we + weisolated isolated parental parental J/Parent J/Parent cells cells into into surface surface CD36-negative CD36-negative (CD36 (CD36−) and−) andCD36-positive CD36-positive (CD36 (CD36+) cells) + cellsusing using fluorescence-activated fluorescence-activated cell sorter cell sorter (FACS) (FACS) (Figure (Figure 7A).7 FigureA). Figure 7B shows7B shows that that CD36 CD36+ cells werecells wereless susceptible less susceptible to apoptosis to apoptosis in response in response to BTZ to BTZ (0–8 (0–8 nM) nM) at 24 at h 24 when h when compared compared to CD36 to CD36− cells.− cells. To Toensure ensure that that these these differences differences in in BTZ BTZ sensitivity sensitivity were were due due to to CD36 CD36 expression, we usedused anti-CD36anti-CD36 + neutralizingneutralizing antibodyantibody (1:500(1:500 oror 22µ μgg/mL)/mL) toto blockblock surface surface CD36 CD36 in in CD36 CD36+ cells.cells. FigureFigure7 7CC shows shows that that + blockingblocking ofof thethe surfacesurface CD36CD36 increasedincreased BTZ-inducedBTZ-induced apoptosis apoptosis in in CD36 CD36+ cellscells toto thethe comparable levellevel toto CD36CD36−− cells. Cancers 2019, 11, 576 10 of 21 Cancers 2019, 11, x 10 of 21

A B J/Parent 120 − CD36− CD36− CD36+ CD36+ 80 * * * SSC 40 * % Apoptosis % * * 0 2 0 2 3 4 5 0369 CD36-APC BTZ (nM)

BTZ 0 nM BTZ 5 nM

120 IgG 1:500 120 IgG 1:500 # 80 80 *

40 40 % Apoptosis % % Apoptosis %

0 0 CD36− CD36+ CD36− CD36+

BTZ 6 nM BTZ 7 nM

120 IgG 1:500 120 IgG 1:500 # # * *

80 80

40 40 % Apoptosis % % Apoptosis

0 0 CD36− CD36+ CD36− CD36+ Figure 7.7. CD36+ cellscells were were less susceptible to BTZ-induced apoptosis. Parental (J/Parent) (J/Parent) cells were − + isolated basedbased onon CD36CD36 expressionexpression intointo CD36CD36− and CD36 cellscells using using fluorescence-activated ﬂuorescence-activated cell sorter (FACS)(FACS) ((AA)) and and apoptosis apoptosis in in response response to to BTZ BTZ (0–8 (0–8 nM) nM) were were determined determined by Hoechstby Hoechst 33342 33342 assay assay at 24 at h (24B). h Data(B). Data are mean are meanSD ± (SDn = (n3). = *3).p *< p0.05 < 0.05 versus versus CD36 CD36cells;− cells; two-sided two-sided Student’s Student’st -test.t-test. ((CC)) CD36CD36++ ± − cells werewere pretreatedpretreated withwith neutralizingneutralizing antibody antibody (1:500) (1:500) for for 1 1 h h and and treated treated with with BTZ BTZ (0–7 (0–7 nM) nM) for for 24 24 h. Afterh. After which, which, apoptosis apoptosis was was determined determined by Hoechstby Hoechst 33342 33342 assay. assay. Data Data are mean are meanSD ± (n SD= 3).(n = * p3).< *0.05 p < + ± # versus0.05 versus BTZ-treated BTZ-treated IgG IgG control control CD36 CD36− or− CD36or CD36cells;+ cells; two-sided two-sided Student’s Student’st-test. t-test. #p p< < 0.050.05 versusversus CD36− control;control; two-sided two-sided Student’s Student’s tt-test.-test.

Next, we manipulated CD36 in parentalparental cells by ectopic gene expressionexpression to ascertain the role of CD36 in BTZ sensitivity. JJ/Parent/Parent cells were transfectedtransfected with CD36 and its level was determined prior to anan evaluationevaluation of of apoptosis. apoptosis. Western Western blot blot and and flow flow cytometric cytometric analyses analyses confirmed confirmed an increase an increase in total in andtotal surface and surface CD36 in CD36 CD36-overexpressing in CD36-overexpressing Jeko-1 cells Jeko-1 (Figure 8cellsA,B). (Figure Importantly, 8A,B). an accumulationImportantly, ofan lipidaccumulation droplets, asof evaluatedlipid drop bylets, oil as red evaluated O staining, by wasoil red apparently O staining, observed was apparently in CD36-overexpressing observed in CD36- cells, whileoverexpressing lipid droplets cells, were while practically lipid droplets minimum were in controlpractically transfected minimum green in fluorescencecontrol transfected protein (GFP)green cellsfluorescence (Figure8 C).protein Figure (GFP)9A shows cells that(Figure CD36-overexpressing 8C). Figure 9A shows cells rendered that CD36-overexpressing the J /Parent cells to BTZcells resistancerendered the when J/Parent compared cells toto GFPBTZ cells.resistance To strengthen when compared this finding, to GFP GFP cells. and To CD36-overexpressing strengthen this finding, cells GFP and CD36-overexpressing cells were treated with various concentrations of BTZ in the presence or absence of anti-CD36 neutralizing antibody (1:1000–1:500). Anti-CD36 antibody exerted a Cancers 2019, 11, 576 11 of 21

wereCancers treated 2019 with, 11, x various concentrations of BTZ in the presence or absence of anti-CD36 neutralizing11 of 21 antibody (1:1000–1:500). Anti-CD36 antibody exerted a pronounced sensitizing effect on BTZ-induced apoptosispronounced in CD36-overexpressing sensitizing effect on cells,BTZ-induced and to a apoptosis lesser extent in CD36-overexpressing in GFP cells, in a dose-dependent cells, and to a lesser manner extent in GFP cells, in a dose-dependent manner (Figure 9B), thereby validating the role of CD36 in (Figure9B), thereby validating the role of CD36 in MCL apoptosis protection. Similar findings on MCL apoptosis protection. Similar findings on the protective role of CD36 were also observed in the protective role of CD36 were also observed in human MCL-derived Granta-519 and SP49 cells human MCL-derived Granta-519 and SP49 cells (Figure 10 and Supplementary Figure S1), suggesting (Figurethe 10 generality and Supplementary of the observed Figure effects S1), in MCL suggesting cells. the generality of the observed effects in MCL cells.

A kDa 93 CD36 72

β-actin 43

B Control GFP Surface CD36

Unstained GFP Unstained GFP

49.6% 1.1%

SSC SSC

FITC CD36-APC

GFP GFP

53.8% 21.0% 6

SSC SSC

0 2 3 FITC CD36-APC

CD36 CD36

1.9% 51.4%

SSC SSC

FITC CD36-APC

C GFP CD36

FigureFigure 8. Overexpression 8. Overexpression of of CD36 CD36 caused caused anan accumulationaccumulation of of lipid lipid content. content. Parental Parental (J/Parent) (J/Parent) cells cells werewere transfected transfected with with CD36 CD36 or or green green ﬂuorescence fluorescence protein (GFP) (GFP) control control plasmid plasmid using using nucleofection. nucleofection. (A) Transfected(A) Transfected cells cells were were analyzed analyzed for for total total CD36CD36 level using using Western Western blotting. blotting. (B) (FlowB) Flow cytometry cytometry analysis of surface CD36 (right) and GFP for transfection control (left). Percentage of CD36- and GFP- analysis of surface CD36 (right) and GFP for transfection control (left). Percentage of CD36- and positive cells (box) were determined based on their internal negative control (unstained cells). (C) GFP-positive cells (box) were determined based on their internal negative control (unstained cells). Analysis of lipid droplets by oil red O staining. Scale bar = 50 μm. (C) Analysis of lipid droplets by oil red O staining. Scale bar = 50 µm. Cancers 2019, 11, 576 12 of 21 Cancers 2019, 11, x 12 of 21

A 120 GFP CD36 80

* * 40 % Apoptosis % * * *

0 0369 BTZ (nM)

BTZ 0 nM BTZ 5 nM

120 IgG 1:1000 1:500 120 IgG 1:1000 1:500 IgG 1:1000 1:500

80 80 # * * * 40

% Apoptosis 40 % Apoptosis %

0 0 GFP CD36 GFP CD36

BTZ 7 nM BTZ 8 nM

120 IgG 1:1000 1:500 120 IgG 1:1000 1:500 # # * * * * * 80 * 80

40 40 % Apoptosis % % Apoptosis%

0 0 GFP CD36 GFP CD36 Figure 9.9. CD36CD36 is is a key mediator ofof BTZ-inducedBTZ-induced apoptosis.apoptosis. ( (AA)) Parental Parental (J/Parent) (J/Parent) cells were transfected with CD36 or GFP control plasmid using nucleofection and apoptosis in response to BTZ (0–8 nM) was was determined determined by by Hoechst Hoechst 33342 33342 assay assay at at24 24h. h.Data Data are aremean mean ± SD (nSD = 4). (n *= p4). < 0.05 * p versus< 0.05 ± versusGFP control; GFP control; two-sided two-sided Student’s Student’s t-test. t(-test.B) CD36-overexpressing (B) CD36-overexpressing Jeko-1 Jeko-1 cells were cells werepretreated pretreated with withneutralizing neutralizing antibody antibody (1:1000–500) (1:1000–500) for 1 for h 1and h and trea treatedted with with BTZ BTZ (0–8 (0–8 nM) nM) for for 24 24 h. h. After After which, apoptosis was determined determined by by Hoechst Hoechst 33342 33342 assay. assay. Data Data are are mean mean ± SDSD (n = ( n3).= *3). p < *0.05p < versus0.05 versus BTZ- ± BTZ-treatedtreated IgG control IgG control GFP GFPor CD36-overex or CD36-overexpressingpressing cells; cells; two-sided two-sided Student’s Student’s t-test.t-test. # p < #0.05p < versus0.05 versus GFP GFPcontrol; control; two-sided two-sided Student’s Student’s t-test.t-test. Cancers 2019, 11, 576 13 of 21 Cancers 2019, 11, x 13 of 21

Granta-519 A GFP CD36

5.6% 35.9%

SSC SSC

1 2 3 4 5 CD36-APC CD36-APC

B 120 GFP CD36 80 * * * 40 * % Apoptosis% * *

0 0369 BTZ (nM)

BTZ 6 nM BTZ 7 nM

120 IgG 1:500 120 IgG 1:500 # #

* 80 * 80

40 40 % Apoptosis % % Apoptosis %

0 0 GFP CD36 GFP CD36 FigureFigure 10.10. GeneralityGenerality of of the the role role of CD36 of CD36 in BTZ-induced in BTZ-induced apoptosis apoptosis in multiple in multiple MCL cell MCL lines. cell Human lines. HumanMCL-derived MCL-derived Granta-519 Granta-519 cells were cells weretransfecte transfectedd with with CD36 CD36 or orGFP GFP control control plasmid usingusing nucleofection.nucleofection. ( (AA)) Flow Flow cytometry cytometry analysis analysis of ofsurface surface CD36. CD36. Percentage Percentage of CD36-positive of CD36-positive cells (box) cells (box)were weredetermined determined based based on their on theirinternal internal negative negative control control (unstained (unstained cells). cells).(B) Apoptosis (B) Apoptosis of CD36- of CD36-overexpressingoverexpressing and GFP and control GFP control cells in cells response in response to BTZ to (0–8 BTZ nM) (0–8 were nM) determined were determined by Hoechst by Hoechst 33342 33342assay assayat 24 ath. 24Data h. Dataare mean are mean ± SD (nSD = 3). (n =* p3). < 0.05 * p < versus0.05 versus GFP GFPcontro control;l; two-sided two-sided Student’s Student’s t-test.t-test. (C) ± (CD36-overexpressingC) CD36-overexpressing Granta-519 Granta-519 cells cells were were pretreated pretreated with with neutralizing neutralizing antibody antibody (1:500) (1:500) for 1 for h and 1 h andtreated treated with with BTZ BTZ (0–7 (0–7 nM) nM) for for 24 24h. h.After After which, which, apoptosis apoptosis was was determined determined by by Hoechst Hoechst 33342 33342 assay. DataData areare meanmean ± SDSD( n(n= = 3)3)..* *p p< < 0.050.05 versusversus BTZ-treatedBTZ-treated IgGIgG controlcontrol GFPGFPor or CD36-overexpressing CD36-overexpressing ± cells;cells; two-sidedtwo-sided Student’sStudent’s t-test.t-test. ## pp << 0.050.05 versusversus GFPGFP control;control; two-sidedtwo-sided Student’sStudent’st -test.t-test. SeeSee alsoalso SupplementarySupplementary FigureFigure S1 S1 for for data data from from human human MCL-derived MCL-derived SP49 SP49 cells. cells.

ToTo provide provide supporting supporting evidence evidence on theon BTZthe BTZ sensitization sensitization by CD36 by CD36 inhibition, inhibition, BTZ-resistant BTZ-resistant J/500R andJ/500R J/250R and cells J/250R were cells treated were with treated BTZ with (0–300 BTZ nM) (0 in–300 the nM) presence in the or presence absence of or anti-CD36 absence of neutralizing anti-CD36 antibodyneutralizing (1:500) antibody and apoptosis (1:500) and was apoptosis determined was determined at 24 h. Figure at 24 11 h.A Figure shows 11A that shows blocking that ofblocking CD36 significantlyof CD36 significantly reduced BTZ reduced resistance BTZ inresistance J/500R cells. in J/500R With thecells. basal With expression the basal ofexpression CD36 in J /of250R CD36 cells, in itJ/250R is not cells, surprising it is not that surprising the inductive that the eff inductivect of CD36e effect inhibition of CD36 oninhibition BTZ-induced on BTZ-induced apoptosis inapoptosis J/250R cellsin J/250R was alsocells significant, was also significant, albeit at a loweralbeit levelat a lo thanwer thoselevel ofthan J/500R those cells of J/500R (Figure cells 11B). (Figure Altogether, 11B). theseAltogether, results these strengthen results the strengthen potential the application potential ofapp CD36lication inhibition of CD36 for inhibition BTZ sensitization. for BTZ sensitization. Cancers 2019, 11, 576 14 of 21 Cancers 2019, 11, x 14 of 21

Figure 11.11. Inhibition of CD36CD36 enhancesenhances BTZBTZ sensitivitysensitivity inin BTZ-resistantBTZ-resistant cells.cells. BTZ-resistantBTZ-resistant JJ/500R/500R cells (A) oror JJ/250R/250R cells (B)) werewere treatedtreated withwith BTZBTZ (0–300(0–300 oror 0–1500–150 nM)nM) inin thethe presencepresence oror absenceabsence ofof CD36 neutralizing antibody (1:500) and apoptosis was determined by Hoechst 33342 33342 assay assay at at 24 h. Data are meanmean ± SD (n = 4).4). * * pp << 0.050.05 versus versus IgG IgG control; control; two-sided two-sided Student’s t-test. ( C) Kaplan Meier ± survival plot of MCL patients from the genegene microarrays (GSE10793) according to the level of CD36 expression. Overall survival of patients with highesthighest expression (cuto(cutoff:ff: Upper quartile, red line) isis compared toto patientspatients withwith lowestlowest expressionexpression (cuto(cutoff:ff: LowerLower quartile,quartile, greengreen line)line) ((nn= = 32).32). (D) KaplanKaplan Meier survivalsurvival plot plot of of MCL MCL patients patients (GSE10793) (GSE10793) according according to the levelto the of levelCD36 ofand CD36ZEB1 andco-expression. ZEB1 co- Overallexpression. survival Overall of patients survival with of patients highest co-expressionwith highest co-expression (cutoff: Upper (cutoff: quartile, Upper red line) quartile, is compared red line) to patientsis compared with to lower patients co-expression with lower (green co-expression line) (n = (green54). line) (n = 54).

It isis well-acceptedwell-accepted that that drug drug resistance resistance is dependent is dependent on multiple on multiple complex complex mechanisms. mechanisms. Since Zeb-1Since hasZeb-1 been has previously been previously reported reported by our by group our group to be anto be important an important therapeutic therapeutic target target that correlates that correlates with patientwith patient overall overall survival survival and and BTZ BTZ resistance resistance in MCLin MCL via via cancer cancer stem stem cell cell regulation regulation [ 32[32],], wewe next performed database analysis based on mRNA expression of CD36 and co-expression of CD36 and ZEB1 inin humanhuman MCLMCL microarraysmicroarrays availableavailable onon GeneGene ExpressionExpression OmnibusOmnibus (GEO;(GEO; accessionaccession numbernumber GSE10793) [[33]33] to examineexamine theirtheir associationassociation toto clinicalclinical outcome.outcome. Our analysis revealed that MCL patients withwith the the highest highest expression expression of CD36of CD36(cuto (cutoff:ff: Upper Upper quartile) quartile) have have shorter shorter overall overall survival survival when comparedwhen compared to MCL to patients MCL withpatients lowest with expression lowest ex (cutopressionff: Lower (cutoff: quartile) Lower (Figure quartile) 11C). (Figure Interestingly, 11C). theInterestingly, hazard ratio the increased hazard ratio approximately increased three-foldapproximately when three-fold compared when to the overallcompared survival to the between overall MCLsurvival patients between with MCL highest patients co-expression with highest of CD36 co-expressionand ZEB1 of(cuto CD36ff: and Upper ZEB1 quartile) (cutoff: to Upper MCL quartile) patients withto MCL lower patients co-expression with lower of CD36 co-expressionand ZEB1 of(Figure CD36 11andD). ZEB1 Notably, (Figure the number11D). Notably, of MCL the patients number with of lowestMCL patients co-expression with lowest (cutoff :co-expression Lower quartile) (cutoff: was notLower suffi quartile)cient to perform was not the sufficient analysis. to These perform clinical the dataanalysis. support These the clinical involvement data support of CD36, the together involvement with Zeb-1,of CD36, in controllingtogether with of aggressiveZeb-1, in controlling MCL. of aggressive MCL.

Cancers 2019, 11, 576 15 of 21

3. Discussion MCL is one of the most aggressive lymphomas attributable to its innate and acquired drug resistance that critically limits the clinical outcome in patients [1–4]. Here, we observed that the development of resistance towards BTZ in MCL was unique, as classical genes involved in drug resistance and lymphoma cell survival were mostly downregulated when compared to their parental cells, opposed of their continued growth (Figure2A,B). We postulated that this phenomenon might occur due to the global DNA hypermethylation (Figure2C,D). The association between drug resistance and global hypermethylation has previously been reported [34,35]. In breast carcinoma, chronic intermittent exposure to adriamycin induced hypermethylation that was associated with multidrug resistant phenotype in correlation to an upregulation of DNMT1, DNMT3A and DNMT3B and an increase in DNA methyltransferase activity [36]. DNA demethylation by antisense targeting or hydralazine was capable of restoring breast carcinoma sensitivity to adriamycin and paclitaxel. Similar findings on the role of hypermethylation in the development of drug resistance were observed in neuroblastoma [37,38]. Several studies have considered an aberrant metabolism as a key feature that leads to the emergence of aggressive tumors [39,40]. Here, we used FTIR spectral analysis to generally examine biochemical changes that differentiate between BTZ-resistant cells and parental cells with an intention to find novel candidate targets (Figure3). We observed the changes in lipid and ester and plausible changes in either nucleic acid, glycoprotein and/or carbohydrate in highly BTZ-resistant J/500R cells that imply the metabolic reprogramming in these cells (Figure4). The critical linkage between metabolism and BTZ sensitivity has recently been reported by our group via a posttranslational medication in the hexosamine biosynthetic pathway called O-GlcNAcylation [41]. An increase of global O-GlcNAcylated protein, a form of glycoprotein that is an ideal sensor for nutritional changes, potentiates MCL response to BTZ and reverses the resistant phenotype, consistent with the findings from FTIR analysis. Dysregulated lipid metabolism has been associated with aggressive forms of numerous solid cancers and served as potential prognostic biomarkers for prostate, breast, ovarian and colorectal cancers [28,29,42]. In the present study, FTIR analysis unveiled for the first time an increase in cellular lipid content in highly BTZ-resistant J/500R cells (Figure4B), which was validated by oil red O staining (Figure5A,B). Excessive lipid has earlier been shown to predict cisplatin resistance in bladder cancer [ 43] and carboplatin resistance in laryngeal carcinoma [44], indicating the involvement of lipid metabolism in drug sensitivity. To further explore the causal relationship between lipid and BTZ resistance at the molecular level, we profiled key players important in the regulation of endogenous and exogenous lipid and identified CD36 as a potential candidate (Figures5C and6). CD36, earlier identified as a platelet integral membrane glycoprotein (GPIV) and a receptor for trombospondin-1 (TSP-1), is a member of class B scavenger receptor family. While its intracellular domains are important in localizing CD36 within caveolae and lipid rafts, which accounts for the diversity of signal transduction, its extracellular domains bind to a vast variety of ligands, including lipid ligands such as native and oxidized lipoproteins, anionic phospholipids and fatty acids [45,46]. Hence, CD36 has subsequent impact on lipid metabolism. We found that overexpression of CD36 caused an increase in cellular lipid droplets (Figure8), indicating that CD36 promotes lipid uptake and /or accumulation in MCL cells. Correlated with invasion of tumors and metastasis, CD36 has been extensively proposed as a prognostic biomarker for various types of cancers, mostly of epithelial origin [45,47]. However, the precise roles of CD36 in hematologic malignancies, particularly the regulation of apoptosis and drug response, are still very limited. In the present study, we provided direct evidence that CD36 protected various human MCL-derived cells against BTZ-induced apoptosis (Figure7, Figure9, Figure 10 and Supplementary Figure S1). Blocking of surface CD36 using neutralizing antibody strongly reversed the protective effect of CD36 in parental cells and importantly reversed the resistant phenotype in BTZ-resistant cells (Figure 11A,B), highlighting the potential application of CD36 inhibition for BTZ sensitization. CD36 antibody therapy has recently been shown to reduce cancer severity in preclinical models of human orthotopic oral cancer and prostate cancer patient-derived xenografts, supporting Cancers 2019, 11, 576 16 of 21 the clinical significance of antibody-based therapeutic targeting of CD36 [47,48]. Our findings are substantiated by survival analyses of MCL patients showing poor clinical outcomes with high CD36 expression that becomes even worse with the co-expression of ZEB1 (Figure 11C,D). It is worth noting that CD36 expression was, however, negatively associated with overall survival of multiple myeloma patients [49]. Zeb-1 (also known as TCF8) is a known zinc finger transcription factor and an EMT activator. In MCL, it has been shown to promote tumor growth in vivo and correlated with poorer clinical prognosis [20]. We earlier found that Zeb-1 prevents various MCL cells against BTZ-induced apoptosis through cancer stem cell expansion and maintenance [32]. Since Zeb-1 has been identified as a driver of adipogenesis in 3T3-L1 fat cell differentiation [50], we evaluated the association between Zeb-1, CD36 and lipid. Overexpression of Zeb-1, although induced lipid accumulation in MCL cells, had minimal effect on CD36, and vice versa (Supplementary Figure S2). These results suggest that Zeb-1 and CD36 independently mediate lipid reprogramming and BTZ response in MCL.

4. Materials and Methods

4.1. Cell Culture and Reagents Human MCL-derived Jeko-1 cells were obtained from American Type Culture Collection (ATCC; Manassas, MA, USA), while Granta-519 and SP49 cells were kind gifts of Dr. Siwanon Jirawatnotai (Systems Pharmacology, Faculty of Medicine Siriraj Hospital, Bangkok, Thailand) [51,52]. Mycoplasma contamination was checked every eight weeks using MycoAlert mycoplasma detection kit (Lonze, Cologne, Germany) and any cell lines found positive were discarded. Cells were cultured in RPMI1640 medium containing 10% fetal bovine serum (FBS), supplemented with 2 mM L-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin, and maintained in a humidiﬁed atmosphere of 5% CO2 environment at 37 ◦C. BTZ was obtained from Janssen-Cilag (Beerse, Belgium) and a stock solution of 1 mg/mL was produced and stored at 20 C until use. Hoechst 33342 was obtained from Molecular − ◦ Probes (Eugene, OR, USA). Antibody for CD36 for ﬂow cytometry was obtained from BioLegend (San Diego, CA, USA), while antibodies for neutralization and Western blotting were from Abcam (Cambridge, UK). All other antibodies were obtained from Cell Signaling Technology (Beverly, MA, USA) and all other reagents, including DAC and oil red O, were obtained from Sigma-Aldrich (St. Louis, MO, USA).

4.2. Generation of De Novo BTZ-Resistant Cells BTZ-resistant cell lines were generated by stepwise selection method as described previously with slight modiﬁcations [13]. Parental MCL-derived Jeko-1 (J/Parent) cells were continuously exposed to increasing concentrations of BTZ to the maximum concentration of 100, 250 and 500 nM. Resistant cells were analyzed by Annexin V-FITC/PI assay, where double negative cells were sorted using BD FACSAria cell sorter (BD Biosciences, San Jose, CA, USA) and designated as J/BTZ100R, J/BTZ250R and J/BTZ500R, respectively. Dead Cell Removal Kit (Miltenyi Biotec, Auburn, CA, USA) was used to remove the dead cells during culture. The dose-response curve of BTZ on resistant cells at diﬀerent degrees was plotted and LC50 was calculated using GraphPad Prism software (La Jolla, CA, USA) as previously described [53].

4.3. Apoptosis Assay Apoptosis was determined by Hoechst 33342 assay. Brieﬂy, cells were incubated with 10 µg/mL Hoechst 33342 for 30 min and analyzed for apoptosis by scoring of cells having condensed (brighter than non-apoptotic) and/or fragmented nuclei by ﬂuorescent microscopy (Eclipse Ti-U with NiS-Elements, Nikon, Tokyo, Japan). The apoptotic index was calculated as the percentage of cells with apoptotic nuclei over total number of cells. Cancers 2019, 11, 576 17 of 21

4.4. RNA Isolation and Quantitative PCR Total RNA was prepared using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Isolated RNA was reverse-transcripted with High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA). Quantitative PCR analysis was carried out on CFX384 Real-Time PCR (Bio Rad, Hercules, CA, USA) using a Power SYBR Green PCR master mix (Applied Biosystems). Initial enzyme activation was performed at 95 ◦C for 10 min, followed by 40 cycles of denaturation at 95 ◦C for 15 s and primer annealing/extension at 60 ◦C for 1 min. Relative expression of each gene was normalized against the housekeeping GAPDH gene product. Heatmap and cluster analysis were performed using MultiExperiment Viewer (MeV) software (http://mev.tm4.org/).

4.5. FTIR Sample Preparation and Microspectroscopy Analysis A drop of 5 µL of 4 105 cells was deposited onto IR-transparent 2-mm-thick barium ﬂuoride × windows, air dried, and stored in a desiccator until spectra were acquired. The spectra were acquired at BL4.1 IR Spectroscopy and Imaging Beamline at the Synchrotron Light Research Institute with a Vertex 70 FTIR Spectrometer (Bruker Optics, Ettlingen, Germany) coupled with an IR microscope (Hyperion 2000, Bruker Optics) and a mercury-cadmium-telluride (MCT) detector cooled with liquid 1 nitrogen over the measurement range from 4000 to 800 cm− . The microscope was connected to a software-controlled microscope stage and placed in a specially designed box that was purged by dry air. The measurements were performed using an aperture size of 10 10 µm with a spectral resolution 1 × of 4 cm− , with 64 scans co-added. Spectral acquisition and instrument control was performed using OPUS 7.2 software (Bruker Optics). Spectra from each sample group were analyzed by using PCA. Data were preprocessed by performing a baseline correction and then normalized using Extended 1 1 Multiplicative Signal Correction using the spectral regions from 3000–2800 cm− and 1800–900 cm− using Unscrambler 10.1 software (CAMO, Oslo, Norway). Next, unsupervised hierarchical cluster 1 1 analysis of FTIR spectra data sets of the regions from 3000–2800 cm− to 1750–800 cm− was performed using Ward’s algorithm, which utilized a matrix deﬁning the inter-spectral distances to identify the two most similar FTIR spectra. The spectral distances between all of the remaining spectra and the new clusters were then recalculated using OPUS 7.2 software.

4.6. DNA Methylation Status Global DNA methylation was determined using the MethylFlashTM Methylated DNA Quantification Kit (Epigentek, Farmingdale, NY, USA), according to the manufacturer’s protocol. In brief, 100 ng genomic DNA was prepared and bound to assay wells that were specifically treated to have high DNA affinity. After which, methylated DNA (5-mC) was captured by specific antibody and signal was enhanced and quantified colorimetrically by measuring an absorbance at 450 nm using a microplate reader (Synergy H1, BioTek, Winooski, VT, USA). Relative 5-mC level was expressed as the ratio of signals from the resistant and control samples.

4.7. Oil Red O Staining and Quantification Cell suspension at 2 105 cells were collected onto a 24-well plate, washed twice with phosphate × buffered saline (PBS) and fixed with formalin vapor for 15 min. After fixation, the cells were washed twice with distilled water, air dried and stained with filtered oil red O working solution (60% oil red O stock solution in distilled water; stock solution, 5 mg/mL oil red O powder in isopropanol) for 20 min at room temperature. The cells were then washed twice with distilled water to remove an unbound dye and visualized under a microscope. After microscopic examination, the amount of oil red O was quantified in each well by extracting the dye using isopropanol and gently pipetting and its absorbance was read at 510 nm using a microplate reader (Synergy H1, BioTek). Cancers 2019, 11, 576 18 of 21

4.8. CD36 Staining and Flow Cytometric Analysis For surface staining, cells at the density of 2 105 cells/100 µL in PBS were labeled with × 2 µL of APC-conjugated antibody against CD36 (BioLegend, San Diego, CA, USA) for 15 min at room temperature. The cells were then washed, fixed in 2% paraformaldehyde for 30 min and resuspended in PBS for analysis by flow cytometry. For intracellular staining, cells were first fixed in 2% paraformaldehyde and permeabilized with 0.5% Tween-20 in PBS for 15 min, followed by antibody incubation.

4.9. Plasmids and Transfection mCherry-CD36-C10 (CD36) plasmid was a gift from Prof. Michael Davidson (Addgene # 55011). Cells were transfected with CD36 or GFP control plasmid by nucleofection using 4D-NucleofectorTM (Lonza, Morristown, NJ, USA) with EW113 (Jeko-1 cells) or with DN100 (Granta-519 and SP49 cells) device program. The transfected cells were analyzed for CD36 prior to use by ﬂow cytometry and/or Western blotting.

4.10. Western Blotting After specific treatments, cells were incubated in a commercial lysis buffer (Cell Signaling Technology) and a protease inhibitor mixture (Roche Molecular Biochemicals, Indianapolis, IN, USA) at 4 ◦C for 30 min. Protein content was analyzed using bicinchoninic acid (BCA) protein assay (Pierce Biotechnology, Rockford, IL, USA) and 50 µg of proteins were resolved under denaturing conditions by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes. Membranes were blocked with 5% nonfat dry milk, incubated with appropriate primary antibodies at 4 ◦C overnight, and subsequently incubated with peroxidase-conjugated secondary antibodies for 1 h at room temperature. The immune complexes were analyzed by enhanced chemiluminescence detection system on a digital imager (ImageQuant LAS, GE Healthcare, Pittsburgh, PA, USA).

4.11. Gene Microarray Dataset and Survival Analysis The published dataset GEO ID GSE10793, which were obtained from tumor biopsies from 71 untreated MCL patients, was used to calculate overall survival associated to Zeb-1 and CD36. Genechip measurements were performed using NCI/Staudt human 15K v13 arrays [33]. Kaplan Meir survival plot of MCL patients was generated using GraphPad Prism software according to the level of gene expression with the cutoﬀ value for the highest expression at upper quartile.

4.12. Statistical Analysis The data represent means SD from three or more independent experiments as indicated. ± Statistical analysis was performed by two-sided Student’s t-test at a signiﬁcance level of p < 0.05.

5. Conclusions In conclusion, we demonstrate here the involvement of lipid metabolism in the regulation of apoptosis in MCL cells. Increased lipid content by means of lipid droplets, in part via the upregulation of CD36, reduced BTZ sensitivity of MCL cells, causing BTZ resistance. Therefore, lipid accumulation and CD36 overexpression could be important in predicting therapeutic responses of MCL. We further advise the potential application of CD36 inhibition in combination therapy to battle against drug-resistant MCL. Additionally, we also suggest the use of FTIR spectroscopy as a promising technique in cancer research, particularly when exploring the molecular mechanisms involved in metabolic reprogramming and drug resistance. Cancers 2019, 11, 576 19 of 21

Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/11/4/576/s1. Figure S1: CD36 is a key mediator of BTZ-induced apoptosis, Figure S2: Overexpression of Zeb-1 and CD36 in MCL Jeko-1 cells, Table S1: FTIR Band Assignment for Biological Samples. Author Contributions: Conceptualization, S.L. and S.I.; formal analysis, S.L., M.J. and K.T.; funding acquisition, S.L. and S.I.; investigation, S.L., M.J., J.P., N.R. and P.K.; methodology, S.L. and K.T.; supervision, S.I.; visualization, S.L.; writing—original draft, S.L.; writing—review and editing, S.I. Funding: S.I. is a Senior Research Scholar of Thailand Research Fund. This work was supported by grants from Thailand Research Fund (TRG5980013, to S.L.; RTA 488-0007, to S.I.) and the Commission on Higher Education (CHE-RES-RG-49, to S.I.). Acknowledgments: We would like to acknowledge Siwanon Jirawatnotai from Systems Pharmacology, Department of Pharmacology, Faculty of Medicine Siriraj Hospital, Mahidol University for his technical assistance and Davor Solter and Barbara Knowles for their comments on the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

1. Perez-Galan, P.; Dreyling, M.; Wiestner, A. Mantle cell lymphoma: Biology, pathogenesis, and the molecular basis of treatment in the genomic era. Blood 2011, 117, 26–38. [CrossRef][PubMed] 2. Shah, B.D.; Martin, P.; Sotomayor, E.M. Mantle cell lymphoma: A clinically heterogeneous disease in need of tailored approaches. Cancer Control 2012, 19, 227–235. [CrossRef] 3. Prichard, M.; Harris, T.; Williams, M.E.; Densmore, J.J. Treatment strategies for relapsed and refractory aggressive non-Hodgkin’s lymphoma. Expert. Opin. Pharmacother. 2009, 10, 983–995. [CrossRef] [PubMed] 4. Zaja, F.; Federico, M.; Vitolo, U.; Zinzani, P.L. Management of relapsed/refractory mantle cell lymphoma: A review of current therapeutic strategies. Leuk. Lymphoma 2013, 55, 988–998. [CrossRef] 5. Diefenbach, C.S.; O’Connor, O.A. Mantle cell lymphoma in relapse: The role of emerging new drugs. Curr. Opin. Oncol. 2010, 22, 419–423. [CrossRef][PubMed] 6. Hambley, B.; Caimi, P.F.; William, B.M. Bortezomib for the treatment of mantle cell lymphoma: An update. Ther. Adv. Hematol. 2016, 7, 196–208. [CrossRef][PubMed] 7. Arkwright, R.; Pham, T.M.; Zonder, J.A.; Dou, Q.P. The preclinical discovery and development of bortezomib for the treatment of mantle cell lymphoma. Expert Opin. Drug Discov. 2017, 12, 225–235. [CrossRef] 8. Yazbeck, V.; Shafer, D.; Perkins, E.B.; Coppola, D.; Sokol, L.; Richards, K.L.; Shea, T.; Ruan, J.; Parekh, S.; Strair, R.; et al. A phase II trial of bortezomib and vorinostat in mantle cell lymphoma and diffuse large b-cell lymphoma. Clin. Lymphoma Myeloma Leuk. 2018, 18, 569–575. [CrossRef] 9. Kruh, G.D. Introduction to resistance to anticancer agents. Oncogene 2003, 22, 7262–7264. [CrossRef] [PubMed] 10. Gottesman, M.M.; Fojo, T.; Bates, S.E. Multidrug resistance in cancer: Role of ATP-dependent transporters. Nat. Rev. Cancer 2002, 2, 48–58. [CrossRef][PubMed] 11. McFadyen, M.C.E.; Melvin, W.T.; Murray, G.I. Cytochrome P450 enzymes: Novel options for cancer therapeutics. Mol. Cancer Ther. 2004, 3, 363–371. 12. Pommier, Y.; Sorder, O.; Antony, S.; Hayward, R.L.; Kohn, K.W. Apoptosis defects and chemotherapy resistance: Molecular interaction maps and networks. Oncogene 2004, 23, 2934–2949. [CrossRef][PubMed] 13. Pérez-Galán, P.; Mora-Jensen, H.; Weniger, M.A.; Shaffer III, A.L.; Rizzatti, E.G.; Chapman, C.M.; Mo, C.C.; Stennett, L.S.; Rader, C.; Liu, P.; et al. Bortezomib resistance in mantle cell lymphoma is associated with plasmacytic differentiation. Blood 2011, 117, 542–552. [CrossRef] 14. Kim, A.; Seong, K.M.; Kang, H.J.; Park, S.; Lee, S.S. Inhibition of Lyn is a promising treatment for mantle cell lymphoma with bortezomib resistance. Oncotarget 2015, 6, 38225–38238. [CrossRef] 15. Shah, M.A.; Schwartz, G.K. Cell cycle-mediated drug resistance: An emerging concept in cancer therapy. Clin. Cancer Res. 2001, 7, 2168–2181. [PubMed] 16. Dai, B.; Grau, M.; Juilland, M.; Klener, P.; Höring, E.; Molinsky, J.; Schimmack, G.; Aukema, S.M.; Hoster, E.; Vogt, N.; et al. B-cell receptor-driven MALT1 activity regulates MYC signaling in mantle cell lymphoma. Blood 2017, 129, 333–346. [CrossRef][PubMed] Cancers 2019, 11, 576 20 of 21

17. Bisikirska, B.C.; Adam, S.J.; Alvarez, M.J.; Rajbhandari, P.; Cox, R.; Lefebvre, C.; Wang, K.; Rieckhof, G.E.; Felsher, D.W.; Califano, A. STK38 is a critical upstream regulator of MYC’s oncogenic activity in human B-cell lymphoma. Oncogene 2013, 32, 5283–5291. [CrossRef] 18. Vishwamitra, D.; Shi, P.; Wilson, D.; Manshouri, R.; Vega, F.; Schlette, E.J.; Amin, H.M. Expression and effects of inhibition of type I insulin-like growth factor receptor tyrosine kinasein mantle cell lymphoma. Haematologica 2011, 96, 871–880. [CrossRef][PubMed] 19. Auer, R. Discovery of Hippo in MCL. Blood 2010, 116, 861–862. [CrossRef][PubMed] 20. Sánchez-Tilló, E.; Fanlo, L.; Siles, L.; Montes-Moreno, S.; Moros, A.; Chiva-Blanch, G.; Estruch, R.; Martinez, A.; Colomer, D.; Gy˝orffy, B.; et al. The EMT activator ZEB1 promotes tumor growth and determines differential response to chemotherapy in mantle cell lymphoma. Cell Death Differ. 2014, 21, 247–257. [CrossRef] 21. Samaniego, F.; Sehgal, L.; Braun, F.K.; Berkova, Z.; Romaguera, J.E.; Wang, M.; Rodriguez, M.A.; Neelapu, S.S.; Mathur, R. Molecular signatures of tumor-initiating cells unveil Wnt pathway as a therapeutic target in mantle cell lymphoma. Blood 2014, 124, 2148. 22. Curradi, M.; Izzo, A.; Badaracco, G.; Landsberger, N. Molecular mechanisms of gene silencing mediated by DNA methylation. Mol. Cell. Biol. 2002, 22, 3157–3173. [CrossRef] 23. Breiling, A.; Lyko, F. Epigenetic regulatory functions of DNA modifications: 5-methylcytosine and beyond. Epigenetics Chromatin. 2015, 8, 24. [CrossRef] 24. Güler, G.; Acikgoz, E.; Karabay Yavasoglu, N.Ü.; Bakan, B.; Goormaghtigh, E.; Aktug, H. Deciphering the biochemical similarities and differences among mouse embryonic stem cells, somatic and cancer cells using ATR-FTIR spectroscopy. Analyst 2018, 143, 1624–1634. [CrossRef] 25. Mourant, J.R.; Yamada, Y.R.; Carpenter, S.; Dominique, L.R.; Freyer, J.P. FTIR spectroscopy demonstrates biochemical differences in mammalian cell cultures at different growth stages. Biophys. J. 2003, 85, 1938–1947. [CrossRef] 26. Boroughs, L.K.; DeBerardinis, R.J. Metabolic pathways promoting cancer cell survival and growth. Nat. Cell Biol. 2015, 17, 351–359. [CrossRef] 27. Sunshine, H.; Iruela-Arispe, M.L. Membrane lipids and cell signaling. Curr. Opin. Lipidol. 2017, 28, 408–413. [CrossRef][PubMed] 28. Beloribi-Djefaflia, S.; Vasseur, S.; Guillaumond, F. Lipid metabolic reprogramming in cancer cells. Oncogenesis 2016, 5, e189. [CrossRef][PubMed] 29. Bozza, P.T.; Viola, J.P.B. Lipid droplets in inflammation and cancer. Prostaglandins Leukot. Essent. Fatty Acids 2010, 4–6, 243–250. [CrossRef] 30. Danilova, O.V.; Dumont, L.J.; Levy, N.B.; Lansigan, F.; Kinlaw, W.B.; Danilov, A.V.; Kaur, P. FASN and CD36 predict survival in rituximab-treated diffuse large B-cell lymphoma. J. Hematop. 2013, 6, 11–18. [CrossRef] 31. Kuemmerle, N.B.; Rysman, E.; Lombardo, P.S.; Flanagan, A.J.; Lipe, B.C.; Wells, W.A.; Pettus, J.R.; Froehlich, H.M.; Memoli, V.A.; Morganelli, P.M.; et al. Lipoprotein lipase links dietary fat to solid tumor cell proliferation. Mol. Cancer Ther. 2011, 10, 427–436. [CrossRef][PubMed] 32. Luanpitpong, S.; Poohadsuan, J.; Samart, P.; Kiratipaiboon, C.; Rojanasakul, Y.; Issaragrisil, S. Reactive oxygen species mediate cancer stem-like cells and determine bortezomib sensitivity via Mcl-1 and Zeb-1 in mantle cell lymphoma. Biochim. Biophys. Acta Mol. Basis Dis. 2018, 1864, 3739–3753. [CrossRef] 33. Blenk, S.; Engelmann, J.C.; Pinkert, S.; Weniger, M.; Schultz, J.; Rosenwald, A.; Müller-Hermelink, H.K.; Müller, T.; Dandekar, T. Explorative data analysis of MCL reveals gene expression networks implicated in survival and prognosis supported by explorative CGH analysis. BMC Cancer 2008, 8, 106. [CrossRef] 34. Nyce, J. Drug-induced DNA hypermethylation and drug resistance in human tumors. Cancer Res. 1989, 49, 5826–5836. 35. Nyce, J.W. Drug-induced DNA hypermethylation: A potential mediator of acquired drug resistance during cancer chemotherapy. Mutat. Res. 1997, 386, 153–161. [CrossRef] 36. Segura-Pacheco, B.; Perez-Cardenas, E.; Taja-Chayeb, L.; Chavez-Blanco, A.; Revilla-Vazquez, A.; Benitez-Bribiesca, L.; Duenas-González, A. Global DNA hypermethylation-associated cancer chemotherapy resistance and its reversion with the demethylating agent hydralazine. J. Transl. Med. 2006, 4, 32. [CrossRef] [PubMed] 37. Qiu, Y.Y.; Mirkin, B.L.; Dwivedi, R.S. Differential expression of DNA-methyltransferases in drug resistant murine neuroblastoma cells. Cancer Detect. Prev. 2002, 26, 444–453. [CrossRef] Cancers 2019, 11, 576 21 of 21

38. Qiu, Y.Y.; Mirkin, B.L.; Dwivedi, R.S. Inhibition of DNA methyltransferase reverses cisplatin induced drug resistance in murine neuroblastoma cells. Cancer Detect. Prev. 2005, 29, 456–463. [CrossRef][PubMed] 39. Hanahan, D.; Weinberg, R.A. Hallmarks of cancer: The next generation. Cell 2011, 144, 645–674. [CrossRef] [PubMed] 40. Cheong, H.; Lu, C.; Lindsten, T.; Thompson, C.B. Therapeutic targets in cancer cell metabolism and autophagy. Nat. Biotechnol. 2012, 30, 671–678. [CrossRef] 41. Luanpitpong, S.; Chanthra, N.; Janan, M.; Poohadsuan, J.; Samart, P.; U-Pratya, Y.; Rojanasakul, Y.; Issaragrisil, S. Inhibition of O-GlcNAcase sensitizes apoptosis and reverses bortezomib resistance in mantle cell lymphoma through modification of truncated Bid. Mol. Cancer Ther. 2018, 17, 484–496. [CrossRef] 42. Zhao, Y.Y.; Miao, H.; Cheng, X.L.; Wei, F. Lipidomics: Novel insight into the biochemical mechanism of lipid metabolism and dysregulation-associated disease. Chem. Biol. Interact. 2015, 240, 220–238. [CrossRef] [PubMed] 43. Lee, M.Y.; Yeon, A.; Shahid, M.; Cho, E.; Sairam, V.; Figlin, R.; Kim, K.H.; Kim, J. Reprogrammed lipid metabolism in bladder cancer with cisplatin resistance. Oncotarget 2018, 9, 13231–13243. [CrossRef] 44. Rak, S.; De Zan, T.; Stefulj, J.; Kosović,M.; Gamulin, O.; Osmak, M. FTIR spectroscopy reveals lipid droplets in drug resistant laryngeal carcinoma cells through detection of increased ester vibrational bands intensity. Analyst 2014, 139, 3407–3415. [CrossRef] 45. Enciu, A.M.; Radu, E.; Popescu, I.D.; Hinescu, M.E.; Ceafalan, L.C. Targeting CD36 as biomarker for metastasis prognostic: How far from translation into clinical practice? Biomed Res. Int. 2018, 2018, 7801202. [CrossRef][PubMed] 46. Febbraio, M.; Hajjar, D.P.; Silverstein, R.L. CD36: A class B scavenger receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J. Clin. Investig. 2001, 108, 785–791. [CrossRef] [PubMed] 47. Pascual, G.; Avgustinova, A.; Mejetta, S.; Martín, M.; Castellanos, A.; Attolini, C.S.; Berenguer, A.; Prats, N.; Toll, A.; Hueto, J.A.; et al. Targeting metastasis-initiating cells through the fatty acid receptor CD36. Nature 2017, 541, 41–45. [CrossRef] 48. Watt, M.J.; Clark, A.K.; Selth, L.A.; Haynes, V.R.; Lister, N.; Rebello, R.; Porter, L.H.; Niranjan, B.; Whitby, S.T.; Lo, J.; et al. Suppressing fatty acid uptake has therapeutic effects in preclinical models of prostate cancer. Sci. Transl. Med. 2019, 11. [CrossRef] 49. Alaterre, E.; Raimbault, S.; Goldschmidt, H.; Bouhya, S.; Requirand, G.; Robert, N.; Boireau, S.; Seckinger, A.; Hose, D.; Klein, B.; et al. CD24, CD27, CD36 and CD302 gene expression for outcome prediction in patients with multiple myeloma. Oncotarget 2017, 8, 98931–98944. [CrossRef] 50. Gubelmann, C.; Schwalie, P.C.; Raghav, S.K.; Röder, E.; Delessa, T.; Kiehlmann, E.; Waszak, S.M.; Corsinotti, A.; Udin, G.; Holcombe, W.; et al. Identification of the transcription factor ZEB1 as a central component of the adipogenic generegulatory network. Elife 2014, 3, e03346. [CrossRef] 51. Amin, H.M.; McDonnell, T.J.; Medeiros, L.J.; Rassidakis, G.Z.; Leventaki, V.; O’Connor, S.L.; Keating, M.J.; Lai, R. Characterization of 4 mantle cell lymphoma cell lines. Arch. Pathol. Lab. Med. 2003, 127, 424–431. [PubMed] 52. Kridel, R.; Meissner, B.; Rogic, S.; Boyle, M.; Telenius, A.; Woolcock, B.; Gunawardana, J.; Jenkins, C.; Cochrane, C.; Ben-Neriah, S.; et al. Whole transcriptome sequencing reveals recurrent NOTCH1 mutations in mantle cell lymphoma. Blood 2012, 119, 1963–1971. [CrossRef][PubMed] 53. Woods, J.R.; Mo, H.; Bieberich, A.A.; Alavanja, T.; Colby, D.A. Fluorinated amino-derivatives of the sesquiterpene lactone, parthenolide, as (19)f NMR probes in deuterium-free environments. J. Med. Chem. 2011, 54, 7934–7941. [CrossRef][PubMed]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).