A total of 12 diagnostic bone marrow samples were collected from adult patients with B-ALL treated at our hospital. All bone marrows were sampled before the initiation of induction therapy. Of 12 bone marrow samples, 6 samples were from the adult B-ALL patients who achieved complete remission (CR) following the induction therapy with the VDP (vincristine, doxorubicin, prednisone) regimen, and these patients were considered as chemotherapy-sensitive patients (S-ALL) in the study; then, another 6 samples were from the adult B-ALL patients failing to achieve response at the end of the induction therapy with the VDP regimen, who exhibited a minimal residual disease (MRD) >1 × 10^-2 by flow cytometry at the end of the induction therapy, and these patients were considered as chemotherapy-resistant patients (R-ALL) in the study.

The inclusion criteria of patients were as follows: (1) newly diagnosed as B-ALL by ESMO criteria (21); (2) older than 18 years; (3) received VDP regimen induction treatment; (4) had available induction treatment response data; and (5) had accessible bone marrow samples before treatment. The exclusion criteria of patients were as follows: (1) other types of ALL; (2) complicated with or history of other hematological malignancies or solid tumors; and (3) history of chemotherapy. This study was approved by the Ethics Committee of our hospital. The written informed consents were offered by patients.

Total RNA in R-ALL and S-ALL bone marrow samples was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and quantified with Agilent 2100 (Agilent, Santa Clara, CA, USA). The ribosomal RNA was depleted using the Ribo-Zero rRNA Removal Kit (Illumina, San Diego, CA, USA), and the library of mRNA and lncRNA was constructed using the Illumina TruSeq Stranded Total RNA Library Prep Kit (Illumina, USA) according to the kit’s protocol. For the construction of the miRNA library, a miRNA library kit (Qiagen, Hilden, Germany) was applied. HiSeq 2500 (Illumina, USA) was applied to complete the sequencing of lncRNA, mRNA, and miRNA, respectively. After the acquisition of data, the R project (Version 3.6.3) was applied to complete quantile normalization, data processing, bioinformatics analysis, and graph plotting. Briefly, the Factoextra package and Pheatmap package were acquired to perform principal component analysis (PCA) and heatmap analysis, respectively; the DESeq2 and Limma package was adopted to analyze the mean gene expression, fold change (FC), and differentially expressed lncRNA, mRNA, or miRNA (DElncRNA, DEmRNA, or DEmiRNA). The DElncRNAs, DEmRNAs, and DEmiRNAs were shown with a volcano plot. Gene Ontology (GO) and Kyoko Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted via the DAVID web server. Pearson’s correlation coefficient was used to calculate the correlation between the expression of DElncRNAs and DEmRNAs. The potential target of DEmiRNAs was investigated by miRanda. The competing endogenous RNA (ceRNA) network was displayed using the graph in R packages, which contained the following ceRNAs: (1) top 10 deregulated (5 upregulated and 5 downregulated) DElncRNAs according to log₂FC; (2) DEmRNAs, which have a significant correlation with the DElncRNAs mentioned in (1) (Pearson’s correlation coefficient >0.9 or <-0.9); and (3) DEmiRNAs, whose potential targets were DEmRNAs and DElncRNAs mentioned in (1) and (2). The expressions of the selected lncRNA, miRNA, and mRNA were evaluated by reverse transcription-quantitative polymerase chain reaction (RT-qPCR).

RNA-sequencing raw data in the article are available on the public database “National Omics Data Encyclopedia (NODE) database (https://www.biosino.org/node)” with tracking number OEP003016. The link is https://www.biosino.org/node/project/detail/OEP003016.

Human B-ALL cells Nalm-6 and 697 were obtained from DSMZ (Braunschweig, Germany). The doxorubicin (Dox)-resistant Nalm-6 (N6/ADR) cells were obtained from ATCC (Rockville, USA). The Dox-resistant 697 (697/ADR) cells were constructed according to a previous study (22). The Nalm-6 and 697 cells were cultured in RPMI-1640 medium (HyClone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (HyClone, USA). The N6/ADR and 697/ADR cells were cultured in 70 nM Dox (Selleck, Houston, UT, USA) containing RPMI-1640 medium supplemented with 10% FBS. The Dox resistance of cells was confirmed by chemosensitivity to Dox assay. The expressions of lncRNA DUXAP8, miR-29a, PIK3CA in Nalm-6, N6/ADR, 697, and 697/ADR cells were evaluated.

Control lncRNA overexpression plasmid, lncRNA DUXAP8 overexpression plasmid, control lncRNA short hairpin RNA (shRNA) plasmid, and lncRNA DUXAP8 shRNA plasmid were constructed by Shanghai GenePharma Co., Ltd (Shanghai, China), with pEX-2 vector (GenePharma, Shanghai, China) and pGPU6 vector (GenePharma, China), respectively. The plasmids were transfected into N6/ADR or 697/ADR cells in the presence of Lipofectamine™ 2000 (Invitrogen, USA). After transfection, the cells were divided into oeCTL, oeLNC, shCTL, and shLNC groups. The N6/ADR and 697/ADR cells that were not transfected served as blank control.

Control miR shRNA and miR-29a shRNA plasmids were constructed with the pGCMV/miR/inhibit vector (GenePharma) by Shanghai GenePharma Co., Ltd (Shanghai, China). After the construction, the control miR shRNA plasmid, miR-29a shRNA plasmid, control lncRNA shRNA plasmid, and lncRNA DUXAP8 shRNA plasmid were co-transfected into N6/ADR and 697/ADR cells by applying Lipofectamine™ 2000, accordingly. The cells were then divided into 5 groups: (a) blank group, no transfection cells; (b) shCTL group, cells were transfected with control lncRNA shRNA plasmid and control miR shRNA plasmid; (c) shLNC group, cells were transfected with lncRNA DUXAP8 shRNA plasmid and control miR shRNA plasmid; (d) shMIR group, cells were transfected with control lncRNA shRNA plasmid and miR-29a shRNA plasmid; and (e) shLNC + shMIR group, cells were transfected with lncRNA DUXAP8 shRNA plasmid and miR-29a shRNA plasmid.

The pEX-2 vector and pGCMV/miR vector (GenePharma, China) were applied to construct the control mRNA overexpression, PIK3CA overexpression, control miR overexpression, and miR-29a overexpression plasmids by Shanghai GenePharma Co., Ltd. (Shanghai, China). To perform the co-transfection of control mRNA overexpression plasmid, PIK3CA overexpression plasmid, control miR overexpression plasmid, and miR-29a overexpression plasmid, Lipofectamine™ 2000 was adopted with the manufacturer’s instruction followed. The following groups were generated after transfection: (a) blank group, cells without transfection; (b) oeCTL group, cells were transfected with control mRNA overexpression plasmid and control miR overexpression plasmid; (c) oeMIR group, cells were transfected with miR-29a overexpression plasmid and control mRNA overexpression plasmid; (d) oePIK3CA group, cells were transfected with control miR overexpression plasmid and PIK3CA overexpression plasmid; and (e) oeMIR + oePIK3CA group, cells were transfected with miR-29a overexpression plasmid and PIK3CA overexpression plasmid.

The PIK3CA overexpression plasmid or mTOR activator (MHY1485) (2 μM, MCE, China) was cocultured with lncRNA DUXAP8 knockdown plasmid in N6/ADR and 697/ADR cells, followed by the detection of cell viability via CCK-8 assay, to explore the role of the PI3K-AKT-mTOR pathway in the restored chemosensitivity by lncRNA DUXAP8 knockdown.

The TRIzol Reagent was adopted to extract total RNA. The concentration of total RNA was evaluated by NanoDrop 2000 (Thermo, USA). 1 μg total RNA was transcribed into cDNA using the RT Reagent Kit (Takara, Shiga, Japan) and subjected to the following thermal cycling: 37°C for 15 minutes (min) and 85°C for 5 seconds (sec). The qPCR mix (Takara, Japan) was used to complete qPCR. β-Actin was used as an internal reference for lncRNA and mRNA, and U6 served as an internal reference for miRNA. Primers were listed as follows: lncRNA DUXAP8, forward, 5′ ACCAGCCTCACTAGCACTCTC 3′, reverse, 5′ CTTCCAGCCTCAGCCTCCTAA 3′; PIK3CA, forward, 5′ TTCTCAACTGCCAATGGACTGT 3′, reverse, 5′ AGCACGAGGAAGATCAGGAATG 3′; β-actin, forward 5′ GGCACCACACCTTCTACAATGA 3′, reverse, 5′ TCTCCTTAATGTCACGCACGATT 3′; miR-29a, forward, 5′ ACACTCCAGCTGGGTAGCACCATCTGAAAT 3′, reverse, 5′ TGTCGTGGAGTCGGCAATTC 3′; U6, forward, 5′ GCTCGCTTCGGCAGCACATA3′, reverse, 5′ AATATGGAACGCTTCACGAATTTGC 3′.

The cells were lysed with RIPA buffer (Beyotime, Shanghai, China). The BCA protein assay kit (Beyotime, China) was used to determine the protein concentration. After thermal denaturation, 20 μg protein was separated with 4%–20% precast gel (Beyotime, China). The protein was then transferred to a nitrocellulose membrane (PALL, Port Washington, NY, USA). After being blocked, the membrane was incubated with diluted primary antibodies and secondary antibodies, successively. Finally, the protein bands were visualized with the ECL Kit (Beyotime, China). The antibodies’ information was as follows: PIK3CA monoclonal antibody (1: 2,000) (Invitrogen, USA); Bcl-2 monoclonal antibody (1: 500) (Invitrogen, USA); cleaved caspase 3 monoclonal antibody (1: 1,000) (CST, UK); AKT monoclonal antibody (1:1,000) (CST; UK); phospho-AKT (pAKT) monoclonal antibody (1:1,000) (CST, UK); mTOR monoclonal antibody (1:3,000) (Invitrogen, USA); phospho-mTOR (pmTOR) monoclonal antibody (1:1,000) (CST, UK); S6K monoclonal antibody (1:3,000) (CST, UK); phospho-S6K (pS6k) monoclonal antibody (1:1,000) (CST, UK); and anti-rabbit IgG, HRP-linked antibody (1: 10,000) (Invitrogen, USA).

The Cell Counting Kit-8 (Beyotime, China) was applied to assess cell proliferation. 3 × 10³ cells were loaded into a 96-well plate. 10 μl reagent was added and incubated with cells for 2 h after transfection. A microplate reader was used to read optical density (OD) values. The TUNEL Apoptosis Assay Kit (Beyotime, China) was used to assess the cell apoptosis rate according to the instructions, with 2.5 × 10⁴ cells seeded into a 24-well plate.

For the confirmation of Dox resistance, un-transfected Nalm-6, N6/ADR, 697, and 697/ADR were seeded in a 96-well plate with the number of 1 × 10⁴ and incubated with a range concentration (0, 5, 10, 20, 40, and 80 nM) of Dox for 24 h referring to a previous paper (23). For chemosensitivity detection of transfected cells, the cells were cultured with a range concentration (0, 5, 20, 80 nM) of Dox for 24 h. After the incubation, the cell viability was detected with Cell Counting Kit-8 as described in Cell Proliferation.

To detect the binding of lncRNA DUXAP8 and miR-29a, the wild type and mutant type of lncRNA DUXAP8 were cloned into the pGL6 vector (Beyotime, China). The wild-type and mutant-type lncRNA DUXAP8 plasmids, miR-29a mimic, and control mimic were co-transfected into 293T cells (ATCC, USA) by Lipofectamine™ 2000. After 48 h of incubation, the cells were lysed and the fluorescence value was detected using Luciferase Reporter Gene Assay Kit (Beyotime, China). To detect the binding of PIK3CA and miR-29a, the PIK3CA wild-type and mutant-type plasmids were constructed with a pGL6 vector. With the application of Lipofectamine™ 2000, the PIK3CA wild-type plasmid, PIK3CA mutant-type plasmid, miR-29a mimic, and control mimic were co-transfected into 293T cells. At 48 h, the fluorescence value was detected as mentioned above.

The control lncRNA overexpression plasmid, lncRNA DUXAP8 overexpression plasmid, control lncRNA shRNA plasmid, and lncRNA DUXAP8 shRNA plasmid were transfected into N6/ADR and 697/ADR cells according to the methods described in the LncRNA DUXAP8 Plasmid Transfection. The N6/ADR and 697/ADR cells that were not transfected served as blank control. Then, CMC-544 (Pfizer Inc., New York, NY, USA) was added into cells with various concentrations of 0, 0.5, 1, 2, 4, and 8 ng/ml (24). After incubation for 48 h, cell viability was evaluated with Cell Counting Kit-8, and the half-maximal inhibitory concentration (IC₅₀) was calculated using Probit regression analysis by SPSS 23.0 (IBM, Armonk, NY, USA). At last, the transfected and blank control cells were incubated with CMC-544 at an amount of 2.670 or 2.794 ng/ml for 48 h, with the apoptosis rate determined by TUNEL Apoptosis Assay Kit. The cell viability and apoptosis detection were carried out as described in Cell Proliferation and Cell Apoptosis.

All data in this study were expressed as mean ± standard deviation. The difference among groups and between two groups was assessed by one-way ANOVA followed by Dunnett’s test or t-test. GraphPad Prism 7.02 (GraphPad Software Inc., La Jolla, CA, USA) was applied to analyze data and make graphs. The statistical significance was defined as p < 0.05. No significance: NS (p>0.05); *p<0.05, **p<0.01, ***p<0.001.

The current study consisted of two main parts: the clinical part and the experimental part ( Supplementary Figure 1 ). In the clinical part, 6 R-ALL patients and S-ALL patients were enrolled to detect the lncRNA, miRNA, and mRNA profiles via RNA-seq and RT-qPCR validation to sort key networks related to chemoresistance. Then, experiments were performed via modifying lncRNA DUXAP8, miR-29a, and PIK3CA to explore their effect (alone or in combination) on Dox-resistant B-ALL cell proliferation, apoptosis, Dox chemosensitivity, etc.

The enrolled B-ALL patients were aged 40 ± 9 years, with 50% males and 50% females. Meanwhile, the mean WBC was 59 ± 45 × 10⁹/l, and the mean bone marrow blast percentage was (83 ± 10)%. The detailed information of their characteristics is shown in Table 1 . It was notable that by comparison, S-ALL patients showed a trend of lower WBC (p = 0.064) and CD19 percentage (p = 0.065) compared to R-ALL patients but lacked statistical significance. Furthermore, other characteristics were of no difference between them ( Table 1 ).

The lncRNA expression profile could clearly distinguish R-ALL patients from S-ALL patients via PCA plot analysis ( Figure 1A ) and showed a good inter-consistent trend in R-ALL patients and S-ALL patients, respectively, by heatmap analysis ( Figure 1B ). Then, a total of 759 upregulated and 579 downregulated lncRNAs were identified in R-ALL patients compared with S-ALL patients ( Figure 1C ).

Besides, the miRNA expression profile also distinguished R-ALL patients from S-ALL patients via PCA plot analysis ( Figure 1D ) and exhibited an acceptable inter-consistent trend by heatmap analysis ( Figure 1E ), with 46 upregulated/29 downregulated miRNAs discovered in R-ALL patients compared to S-ALL patients ( Figure 1F ). Furthermore, the mRNA expression profile was also able to differentiate R-ALL patients from S-ALL patients ( Figure 1G ), and disclosed an inter-consistent trend by heatmap analysis ( Figure 1H ), with 676 upregulated/944 downregulated mRNAs uncovered in R-ALL patients compared to S-ALL patients ( Figure 1I ). In addition, the detailed information of each dysregulated lncRNA, miRNA, and mRNA is exhibited in Supplementary Tables 1–3 , respectively.

KEGG enrichment analyses of dysregulated lncRNA ( Figure 2A ), miRNA ( Figure 2B ), and mRNA ( Figure 2C ) expression profiles were carried out, which observed that they were all closely related to the PI3K-AKT pathway. Inspiringly, a comprehensive ceRNA network was established based on dysregulated lncRNA, miRNA, and mRNA expression profiles, which discovered that the lncRNA DUXAP8/miR-29a/PIK3CA network was a key component ( Figure 2D ). What is more, PIK3CA is the key component involved in the PI3K-AKT pathway, which is closely engaged in ALL pathogenesis and treatment sensitivity (25–27). Subsequently, RT-qPCR was performed to verify their expressions, which also demonstrated that lncRNA DUXAP8 was upregulated ( Figure 2E ), miR-29a was downregulated ( Figure 2F ), and PIK3CA was overexpressed ( Figure 2G ) in R-ALL patients compared to S-ALL patients (all p < 0.05). Based on the abovementioned information, the lncRNA DUXAP8/miR-29a/PIK3CA network and downstream PI3K-AKT pathway were essential in chemotherapy resistance, then subsequent experiments were performed to explore and validate this issue.

Firstly, Dox resistance was confirmed in N6/ADR and 697/ADR cells ( Supplementary Figures 2A, B ), then upregulated lncRNA DUXAP8 and PIK3CA and downregulated miR-29a were observed in N6/ADR and 697/ADR cells (all p < 0.05, Supplementary Figures 2C, D ).

After confirmation of transfection efficiency (all p < 0.01, Figures 3A, B ), lncRNA DUXAP8 overexpression was observed to promote cell proliferation in N6/ADR ( Figure 3C ) and 697/ADR ( Figure 3D ) cells (both p < 0.05). Then, lncRNA DUXAP8 overexpression increased Bcl-2 expression but repressed cleaved caspase-3 expression in N6/ADR ( Figures 3E, G ) and 697/ADR ( Figures 3F, H ) cells (all p < 0.05). In addition, lncRNA DUXAP8 overexpression reduced the cell apoptosis rate reflected by TUNEL-positive cell percentage in N6/ADR ( Figures 3I, K ) and 697/ADR ( Figures 3J, L ) cells (both p < 0.05). Besides, lncRNA DUXAP8 overexpression decreased Dox chemosensitivity in N6/ADR cells to some extent ( Figure 3M ) (p < 0.05) but did not affect Dox chemosensitivity in 697/ADR cells ( Figure 3N ) (p > 0.05). Inversely, lncRNA DUXAP8 knockdown exhibited the opposite effect to its overexpression in regulating the abovementioned cell functions; in particular, lncRNA DUXAP8 knockdown restored Dox chemosensitivity to some extent in N6/ADR ( Figure 3M ) and 697/ADR ( Figure 3N ) cells (all p < 0.05).

It was then observed that lncRNA DUXAP8 negatively regulated miR-29a expression in both N6/ADR ( Figure 4A ) and 697/ADR ( Figure 4B ) cells (all p < 0.05). By the Luciferase reporter gene assay with the designed binding sequence shown in Figure 4C , lncRNA DUXAP8 was found to be directly bound to miR-29a to regulate its expression (p < 0.01, Figure 4D ).

MiR-29a knockdown did not affect lncRNA DUXAP8 expression in both N6/ADR ( Figure 5A ) and 697/ADR ( Figure 5B ) cells (both p > 0.05), while lncRNA DUXAP8 knockdown increased miR-29a expression in both N6/ADR ( Figure 5C ) and 697/ADR ( Figure 5D ) cells (both p < 0.01). Subsequent experiments revealed that miR-29a knockdown enhanced cell proliferation in N6/ADR ( Figure 5E ) and 697/ADR ( Figure 5F ) cells (both p < 0.05); it also suppressed the cell apoptosis rate reflected by the TUNEL-positive cell percentage in N6/ADR ( Figures 5G, I ) and 697/ADR ( Figures 5H, J ) cells (both p < 0.05); then, it promoted Bcl-2 expression but reduced cleaved caspase-3 expression in N6/ADR ( Figures 5K, M ) and 697/ADR ( Figures 5L, N ) cells (all p < 0.05). Besides, miR-29a knockdown decreased Dox chemosensitivity in N6/ADR ( Figure 5O ) and 697/ADR ( Figure 5P ) cells (both p < 0.05).

Most notably, miR-29a knockdown compensated for the effects of lncRNA DUXAP8 knockdown on cell proliferation ( Figure 5E ), cell apoptosis rate reflected by TUNEL-positive cell percentage ( Figures 5G, I ), apoptotic markers ( Figures 5K, M ), and Dox chemosensitivity ( Figure 5O ) in N6/ADR cells (all p < 0.05). Similarly, miR-29a knockdown also compensated for the effects of lncRNA DUXAP8 knockdown on cell proliferation ( Figure 5F ), cell apoptosis rate reflected by TUNEL-positive cell percentage ( Figures 5H, J ), apoptotic markers ( Figures 5L, N ), and Dox chemosensitivity ( Figure 5P ) in 697/ADR cells (all p < 0.05).

LncRNA DUXAP8 knockdown decreased the PIK3CA gene expression in N6/ADR ( Figure 6A ) and 697/ADR ( Figure 6B ) cells (both p < 0.05); it also repressed PIK3CA protein expression in N6/ADR ( Figures 6C, E ) and 697/ADR ( Figures 6D, F ) cells (both p < 0.05), whereas miR-29a knockdown increased PIK3CA gene expression in N6/ADR ( Figure 6A ) and 697/ADR ( Figure 6B ) cells (both p < 0.001), as well as elevated PIK3CA protein expression in N6/ADR ( Figures 6C, E ) and 697/ADR ( Figures 6D, F ) cells (both p < 0.01). Subsequent Luciferase reporter gene assay observed that miR-29a was directly bound to PIK3CA to regulate its expression (p < 0.01, Figure 6G ), with the designed binding sequence shown in Figure 6H .

PIK3CA overexpression did not affect miR-29a expression in N6/ADR ( Figure 7A ) and 697/ADR ( Figure 7B ) cells (both p > 0.05), while miR-29a overexpression inhibited PIK3CA expression in N6/ADR ( Figure 7C ) and 697/ADR ( Figure 7D ) cells (both p < 0.05). It was then noted that miR-29a overexpression inactivated the PI3K-AKT-mTOR pathway in N6/ADR ( Figures 7E, G ) and 697/ADR ( Figures 7F, H ) cells (all p < 0.05), while PIK3CA overexpression activated the PI3K-AKT-mTOR pathway in N6/ADR ( Figures 7E, G ) and 697/ADR ( Figures 7F, H ) cells (all p < 0.05).

MiR-29a overexpression decreased cell proliferation in N6/ADR ( Figure 8A ) and 697/ADR ( Figure 8B ) cells (both p < 0.05), while it increased the cell apoptosis rate reflected by the TUNEL-positive cell percentage in N6/ADR ( Figures 8C, E ) and 697/ADR ( Figures 8D, F ) cells (both p < 0.01). MiR-29a overexpression also inhibited Bcl-2 expression but promoted cleaved caspase-3 expression in N6/ADR ( Figures 8G, I ) and 697/ADR ( Figures 8H, J ) cells (all p < 0.05). Besides, miR-29a overexpression increased Dox chemosensitivity in N6/ADR ( Figure 8K ) and 697/ADR ( Figure 8L ) cells (both p < 0.05). However, PIK3CA overexpression showed the opposite effect as that of miR-29a overexpression in N6/ADR and 697/ADR cells (all p < 0.05).

It was inspiring that PIK3CA overexpression compensated for the effects of miR-29a overexpression on cell proliferation ( Figure 8A ), cell apoptosis rate reflected by the TUNEL-positive cell percentage ( Figures 8C, E ), apoptotic markers ( Figures 8G, I ), and Dox chemosensitivity ( Figure 8K ) in N6/ADR cells (all p < 0.05). Similarly, PIK3CA overexpression also compensated for the effects of miR-29a overexpression on cell proliferation ( Figure 8B ), cell apoptosis rate reflected by the TUNEL-positive cell percentage ( Figures 8D, F ), apoptotic markers ( Figures 8H, J ), and Dox chemosensitivity ( Figure 8L ) in 697/ADR cells (all p < 0.05).

In addition, PIK3CA overexpression plasmid or mTOR activator (MHY1485) was cocultured with lncRNA DUXAP8 knockdown plasmid to explore the role of the PI3K-AKT-mTOR pathway in the restored chemosensitivity by lncRNA DUXAP8 knockdown. Then, it was observed that PI3K-AKT-mTOR pathway activation decreased Dox chemosensitivity in N6/ADR ( Supplementary Figure 3A ) and 697/ADR cells ( Supplementary Figure 3B ) and also attenuated the effect of lncRNA DUXAP8 knockdown.

Inotuzumab ozogamicin is a novel drug recommended for the treatment of refractory B-ALL, so we further explored whether modification of lncRNA DUXAP8 affected inotuzumab ozogamicin treatment in Dox-resistant B-ALL cells. It was observed that lncRNA DUXAP8 overexpression increased the IC₅₀ of inotuzumab ozogamicin in N6/ADR cells ( Figure 9A ) (p < 0.05), but not in 697/ADR cells ( Figure 9B ) (p > 0.05); it also decreased the cell apoptosis rate reflected by the TUNEL-positive cell percentage in inotuzumab ozogamicin-treated N6/ADR cells ( Figures 9C, E ) (p < 0.05), but not in 697/ADR cells ( Figures 9D, F ) (p > 0.05). Besides, lncRNA DUXAP8 knockdown decreased the IC₅₀ of inotuzumab ozogamicin in N6/ADR cells ( Figure 9A ) and 697/ADR cells ( Figure 9B ) (both p < 0.05), which also enhanced the cell apoptosis rate reflected by the TUNEL-positive cell percentage in inotuzumab ozogamicin-treated N6/ADR cells ( Figures 9C, E ) and 697/ADR cells ( Figures 9D, F ) (both p < 0.05).