BTK inhibitors differentially induce apoptosis but similarly suppress chemotaxis and lipid 1 accumulation in mantle cell lymphoma 2 3


 Background: The more selective second-generation BTK inhibitors (BTKis) Acalabrutinib and Zanubrutinib and the first-generation BTK inhibitor (BTKi) Ibrutinib are highlighted by their clinical effectiveness in mantle cell lymphoma (MCL), however, similarities and differences of their biological and molecular effects on anti-survival of MCL cells induced by these BTKis with distinct binding selectivity against BTK remain largely unknown. Methods: AlamarBlue assays were performed to define cytotoxicity of BTKis against MCL cells, Jeko-1 and Mino. Cleaved PARP and caspase-3 levels were examined by immunoblot analysis to study BTKi-induced apoptotic effects. Biological effects of BTKis on MCL-cell chemotaxis and lipid droplet (LD) accumulation were examined in Jeko-1, Mino and primary MCL cells via Transwell and Stimulated Raman scattering imaging analysis respectively. Enzyme-linked immunoassays were used to determine CCL3 and CCL4 levels in MCL-cell culture supernatants. RNA-seq analyses identified BTKi targets which were validated by quantitative RT-PCR (qRT-PCR) and immunoblot analysis. Results: Acalabrutinib and Zanubrutinib induced moderate apoptosis in Ibrutinib high-sensitive JeKo-1 cells and Ibrutinib low-sensitive Mino cells, which was accompanied by cleaved PARP and caspase-3. Such effects might be caused by the stronger ability of Ibrutinib to upregulate the expression of pro-apoptotic genes, such as HRK, GADD45A , and ATM , in JeKo-1 cells than in Mino cells, and the expression of such apoptotic genes was slightly changed by Acalabrutinib and Zanubrutinib in both JeKo-1 and Mino cells. Further, Acalabrutinib, Zanubrutinib and Ibrutinib reduced MCL-cell chemotaxis with similar efficiency, due to their similar abilities to downmodulate chemokines, such as CCL3 and CCL4. Also, these three BTKis similarly suppressed MCL-cell LD accumulation via downregulating lipogenic factors, DGAT2, SCD, ENPP2 and ACACA without significant differences. Conclusion: BTKis demonstrated differential capacities to induce MCL-cell apoptosis due to their distinct capabilities to regulate the expression of apoptosis-related genes, and similar biological and molecular inhibitory effects on MCL-cell chemotaxis and LD accumulation.

6 function was then used to quantify the area fractions of LDs in the whole image area, then normalized to 145 the cell number counted from the same image. 146 147

Cell-viability assay 148
Cell viability was evaluated using AlamarBlue assay according to manufacturer's protocol 149 Serotec). JeKo-1 and Mino cells were seeded in 96-well cell culture plates at 5×10 4 cells/100 μl/well (n 150 = 5). AlamarBlue solution (10 μl) was added to each well after cells were treated with BTKis. 151 Fluorescence values were determined with a 560 nm excitation and 590 nm emission wavelengths after 152 3 h incubation in 37℃. Cell-viability was calculated based on manufacturer's instruction. 153 154

Quantitative RT-PCR (qRT-PCR) 155
Total RNA was extracted with TRIzol (#15596018; Thermo Fisher Scientific, Beijing) and RNA samples 156 that meet following requirements were used in subsequent experiments: RNA integrity number (RIN) > 157 7.0 and a 28S:18S ratio > 1.8. To avoid genomic DNA contamination, total RNA samples were treated 158 with a RNase-Free DNase Kit (Invitrogen) following manufacturer's instructions and cDNA was 159 synthesized from 2 μg of total RNA using Superscript III First-strand Synthesis System (#18080-051; 160 Invitrogen, Beijing) according to manufacturer's instructions. qRT-PCR was performed using PowerUp 161 SYBR Green Mix (#00710493; Applied Biosystems, Beijing). The primers used for qRT-PCR were 162 shown in Additional file 2: supplementary Table S1. 163 164

Statistical analysis 178
Statistical analysis was performed with GraphPad Prism 8.0 (GraphPad Software Inc.). Data were shown 179 as mean ± SEM. Differences between two groups or among multiple groups were determined by unpaired 180 2-tailed Student's t-test or by one-way ANOVA with Tukey's multiple comparisons test respectively. P-181 values less than 0.05 were considered significant. 182 183 Results 184

BTKis induced differential cytotoxicity against MCL cell lines 185
We firstly aimed to compare the cytotoxic effectiveness of Ibrutinib, Acalabrutinib and Zanubrutinib 186 observed in MCL. The peak plasma concentration of Ibrutinib, Acalabrutinib and Zanubrutinib in 187 patients treated with these drugs is about 0.5 μM [21,22] were treated with Ibrutinib, Acalabrutinib or Zanubrutinib individually at dose of 0, 0.25, 0.5, 1, 2 and 5 190 μM for 24, 48 and 72 h and measured cell-viability using AlamarBlue assay (Fig. 1a), which showed that 191 Ibrutinib presented a stronger cytotoxic activity in JeKo-1 cells than in Mino cells, and the cytotoxicity 192 of Acalabrutinib and Zanubrutinib were weak in both JeKo-1 and Mino cells over time. Such cytotoxic 193 effects were accompanied by slight cleavage of caspase-3 and PARP (Fig. 1b), which were consistent 194 with cytotoxic capacities of these three BTK inhibitors. 195

RNA-seq identified apoptosis-related targets of BTKis 197
To understand the comprehensive mechanisms involved in MCL-cell apoptosis, we analyzed

RNA-seq identified chemotaxis-related target genes of BTKis 225
Chemokines-activated BCR signaling facilitates cell migration (pseudoemperipolesis) beneath stromal 226 cells [23]. In the current study, conditioned medium from HS-5 human marrow stromal cells induced 227 migration of JeKo-1, Mino and primary MCL cells, which was inhibited by BTKis without significant 228 differences between drug treatments at 2 μM (Fig. 2a). To understand the potentially distinct mechanisms 229 underlying BTKi-controlled biological processes including MCL-cell migration, KEGG enrichment 230 analysis was used to perform chemokine signaling pathway assignment of DEGs regulated by Ibrutinib, 231 Acalabrutinib or Zanubrutinib individually in JeKo-1 and Mino cells, which was included in the immune 232 system subcategory (Additional file 1: Figure S2). We then analyzed the unions of chemotaxis-related 233 DEGs regulated by each BTKi in JeKo-1 or Mino respectively, followed by generating intersections of 234 9 the DEGs in JeKo-1 and Mino, by which downregulated chemotaxis-related genes were screened, 235 including CCL3L1, CCL3, CCL4, CCL4L2, CXCL16 and CXCR5 (Fig. 2b). Accordingly, we examined 236 the levels of CCL3 and CCL4 in the supernatants of JeKo-1, Mino and primary MCL cells treated with 237 or without BTKis individually for 48 h in the presence of anti-human IgM F(ab)2, and the results indicated 238 that these three BTKis significantly suppressed the production of CCL3 and CCL4 with similar degree 239 in MCL cells (Fig. 2c). These data suggested that the three BTKis demonstrated similar abilities to 240 modulate chemotaxis-related genes, which may result in their similar MCL-cell chemotaxis inhibition. 241 242

BTKis inhibited LD accumulation in MCL 243
Considering the critical role of LDs for MCL-cell survival, we next aimed to investigate the capability 244 of BTKis to block LD accumulation. Firstly, SRS imaging analysis was performed to test the LD 245 accumulation in primary MCL cells treated with each BTKi for 24 h versus cells treated with vehicle 246 control, by which revealed that these three BTKis could significantly suppress LD accumulation (Fig.  247 4a, b). Next, JeKo-1 and Mino cells were treated based on the procedure shown in Additional file 1: 248 Figure S3, and analyzed by SRS imaging. The results showed that MCL cells cultured in 10% FBS media 249 had abundant LDs, which diminished after serum-deprivation for 24 h unless provided with 10% FBS 250 again, however, treatment with BTKis inhibited the capacity of FBS to stimulate the production of LDs 251 significantly (Fig. 4c, d). 252

RNA-seq identified LD accumulation related target genes of BTKis 254
To reveal the molecular mechanism underlying BTKi-inhibited MCL-cell LD accumulation, KEGG 255 enrichment analysis showed that BTKis could modulate the expression of some critical lipid metabolism-256 related DEGs (Additional file 1: Figure S2), whose number was shown in Venn diagram (Fig. 5a). We 257 then screened the unions of lipid metabolism-related DEGs regulated by each BTKi in JeKo-1 or Mino 258 respectively, followed by generating intersections of the DEGs in JeKo-1 and Mino (Fig. 5b), and four 259 downregulated LD accumulation-related genes, DGAT2, SCD, ACACA (ACC1) and ENPP2, were 260 selected and validated by qRT-PCR analysis (Fig. 5c) and immunoblot analysis (Fig. 5e). Since the 261 inhibition of these four lipogenic genes could initiate tumor cell apoptosis [24][25][26][27], such findings may 262 demonstrate novel mechanism underlying the anti-survival effects of these three BTKis in MCL via 263 impairing lipid biosynthesis at least partially. 264

Discussion 265
BTK is a major kinase in BCR signaling pathway, which is highlighted by the clinical effectiveness of 266 irreversible small-molecule BTKis, Ibrutinib, Acalabrutinib and Zanubrutinib. In this study, we made a 267 parallel analysis of Ibrutinib, Acalabrutinib and Zanubrutinib to uncover their potential similarities and 268 differences in anti-survival effects in MCL, since they have differential binding selectivity against their 269 common target, BTK, but have similar biologic effects and comparable clinical responses. Our MCL-270 cell viability data showed that Ibrutinib demonstrated a high cytotoxic activity in JeKo-1 cells, but not 271 in Mino cells, which was consistent with previous findings that JeKo-1was classified as an Ibrutinib-272 sensitive cell line, while Mino was classified as an Ibrutinib-resistant cell line [28]. Also, the sensitivities 273 of both JeKo-1 and Mino cells to Acalabrutinib and Zanubrutinib were low. In order to explain the 274 mechanism underlying such effects, RNA-seq analysis followed by KEGG analysis were performed to 275 identify the critical upregulated pro-apoptotic genes and downregulated anti-apoptotic genes controlled 276 by each BTKi, and the data showed that Ibrutinib was more powerful to upregulate pro-apoptotic genes, 277

HRK, GADD45A and ATM in JeKo-1 cells than in Mino cells. In addition, Acalabrutinib and 278
Zanubrutinib had low capacities to modulate the expression of such three apoptotic genes in both JeKo-279 1 and Mino cells. Of note, these three apoptosis-related genes identified in this study were well-known 280 in apoptosis signaling. Detailly, GADD45A is associated with DNA damage and is proapoptotic [29], 281 and ATM, as a tumor suppressor gene, plays a role in the initiation and/or progression of MCL [30]. As 282 a member of the pro-apoptotic subgroup of BCL-2 family, HRK is an essential initiators of apoptosis 283 that can function as tumor suppressors [31]. All of these findings could support the idea that these genes 284 play roles in mediating MCL-cell apoptosis induced by BTK inhibitors, even though overall apoptotic 285 effects induced by BTKis were moderate in MCL. 286 Both Ibrutinib and Acalabrutinib have been shown to decrease levels of CCL3 and CCL4, two 287 critical chemokines inducing migration or homing of leukemia cells, in CLL-cell cultures and their 288 separate clinical trials [14][15][16]32]. Zanubrutinib could inhibit homing of CLL cells through 289 downregulating CXCR5, a homing receptor mediating migration or homing and BCR signaling 290 activation [18,33]. However, regulatory impact of these three BTKis on chemotaxis and chemotaxis-291 related genes in MCL still need to be analyzed. Our study showed that conditioned medium from HS-5 292 human marrow stromal cells induced migration of MCL cells, which was inhibited by Ibrutinib, 293 Acalabrutinib and Zanubrutinib without significant differences between drug treatments. Consistent with 294 11 these functional data, our results indicated that Ibrutinib, Acalabrutinib and Zanubrutinib similarly 295 reduced the expression of CCL3, CCL4 and CXCR5 via RNA-seq followed by KEGG analysis, and the 296 production of CCL3 and CCL4 was validated by ELISA quantification, which did not show significant 297 differences between drug treatments as well. Except these known chemotaxis-related target genes of 298 BTKis, we also found that Ibrutinib, Acalabrutinib and Zanubrutinib could similarly reduce the 299 expression of CCL3L1, CCL4L2 and CXCL16. 300 Since elevated LDs could enhance MCL-cell survival [34], we detected whether these three BTKis 301 could inhibit pro-survival LD accumulation in MCL-cell via SRS imaging analysis, a label-free live-cell 302 imaging technique for testing intracellular components accumulation, including LDs [35]. As expected, 303 quantitative analysis of lipogenesis at single-cell level via SRS imaging revealed that treatment with 304 BTKis significantly reduced the accumulation of LDs in MCL. Based on the KEGG classification of 305 RNA-seq data, we found that BTKis treatment dramatically reduced the expression of several pivotal 306 lipogenic genes, DGAT2, ENPP2, SCD and ACACA (ACC1). DGAT2 catalyzes the final step in synthesis 307 triglyceride, which is a major component of LDs [36], and genetic deletion of DGAT2 was lethal with 308 knockout mice presenting severe and systemic reductions in triglyceride [37]. SCD is a principal enzyme 309 responsible for fatty acid desaturation, which is critical for growth, survival and tumorigenesis [25,38]. and LPA is responsible for cancer cells growth and anti-cancer therapy resistance of many cancer cells 313 [27]. Combined with our findings showing BTKi-associated ENPP2 downregulation, the downregulation 314 of SCD might be caused by the reduction of LPA levels at least partially, since LPA could stimulate SCD 315 expression and therefore accelerate the formation of lipid droplets [41]. Besides, ACC1, also known as 316 ACACA, controls de novo lipogenesis, whose chemical inhibition suppresses lipogenesis and induces 317 apoptosis in cancer cells [26]. Previous study showed that Zanubrutinib could downregulate the 318 expression of ACACA in MCL [19], and our study demonstrated that both Ibrutinib and Acalabrutinib 319 downmodulated ACACA expression in MCL as well. Importantly, inhibition of fatty acid synthesis, a 320 crucial step of LDs accumulation, triggers significant apoptosis in MCL [17]. Accordingly, our data 321 suggested that BTKi-induced downregulation of DGAT2, ENPP2, SCD and ACACA might result in LD 322 accumulation inhibition, by which trigger modest MCL-cell death at least partially, and such findings 323 provide a new evidence that targeting the lipid metabolism pathway might be a strategy to treat MCL, or 324 Acalabrutinib or Zanubrutinib in a dose-dependent manner for 24, 48 and 72h (n=5). The cell-viability 516 was determined by AlamarBlue assay. Data are shown as mean ± SEM; ****P < 0.0001, as calculated 517 using the Student's t-test. b Cleaved caspase-3 and cleaved PARP were increased in MCL cells, JeKo-1 518 and Mino, treated with Ibrutinib, Acalabrutinib or Zanubrutinib in a dose-dependent manner for 24 h, 519 which were examined by immunoblot analysis. GAPDH was used as loading control. shown as mean ± SEM; *P<0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001, as calculated using one-529 way ANOVA with Tukey's multiple comparisons test.

Fig. 4 BTKis inhibited LD accumulation in MCL cells. a SRS imaging analysis of cellular LDs in 543
primary MCL cells (n = 3) treated as the indicated conditions for 24 h. One representative sample is 544 shown. Scale bars, 10 μm. b Lipid area in the fields of view (n = 3) obtained from each sample was 545 quantified based on the SRS images from (a) by ImageJ software. Data are shown as mean ± SEM; 546 *P<0.05; **P < 0.01, as calculated using the Student's t-test. c SRS imaging analysis of cellular LDs in 547 JeKo-1 or Mino cells treated as the indicated conditions on the top of the figures. One representative 548 MCL sample is shown. Scale bars, 10 μm. d Lipid area was quantified based on the SRS images from (c) 549 (n = 5) by ImageJ software. Data are shown as mean ± SEM; *P<0.05; **P < 0.01; ***P < 0.001; ****P 550 < 0.0001, as calculated using the Student's t-test. Zanubrutinib (2 μM). Data are shown as mean ± SEM; **P < 0.01; ***P < 0.001; ****P < 0.0001, as 560 calculated using the Student's t-test. d Analysis of protein expression of DGAT2, ENPP2, SCD and 561 ACACA in MCL cells via immunoblot analysis, which were treated with Ibrutinib, Acalabrutinib or 562 Zanubrutinib (2 μM). 563