Benefits of Combined All-Trans Retinoic Acid and Arsenic Trioxide Treatment of Acute Promyelocytic Leukemia Cells and Further Enhancement by Inhibition of Atypically Expressed Transglutaminase 2

Randomized trials in acute promyelocytic leukemia patients have shown that treatment with a combination of all-trans retinoic acid (ATRA) and arsenic trioxide (ATO) is superior in efficacy to monotherapy, with significantly decreased mortality. So far, there are little data available to explain the success of the ATRA and ATO combination treatment in molecular terms. We showed that ATRA- and ATO-treated cells had the same capacity for superoxide production, which was reduced by two-thirds in the combined treatment. Secreted inflammatory biomarkers (monocyte chemoattractant protein-1 [MCP-1], interleukin-1 beta [IL-1β] and tumor necrosis factor-α [TNF-α]) were significantly decreased and were further reduced in a transglutaminase 2 (TG2) expression-dependent manner. The amount of secreted TNF-α in the supernatant of NB4 TG2 knockout cells was close to 50 times lower than in ATRA-treated differentiated wild-type NB4 cells. The irreversible inhibitor of TG2 NC9 not only decreased reactive oxygen species production 28-fold, but decreased the concentration of MCP-1, IL-1β and TNF-α 8-, 15- and 61-fold, respectively in the combined ATRA + ATO-treated wild-type NB4 cell culture. We propose that atypical expression of TG2 leads to the generation of inflammation, which thereby serves as a potential target for the prevention of differentiation syndrome.


Introduction
Acute promyelocytic leukemia (APL) is characterized by the presence of a chromosomal translocation between the retinoic acid receptor-alpha (RARα) gene on chromosome 17 and the promyelocytic leukemia protein (PML) gene on chromosome 15, resulting in the PML-RARα fusion gene. The PML-RARα oncoprotein acts as an abnormal retinoid receptor, repressing transcription and resulting in the suppression of retinoic acid-induced myeloid differentiation. All-trans retinoic acid in DS, we further investigated the role of TG2 in limiting toxicities arising from DS. NC9 is a TG2-selective irreversible inhibitor that blocks the enzyme active site and locks TG2 in an open conformation, thereby abrogating GTP binding as well. [11,17]. Our results showed that inhibition of TG2 by NC9 in ATRA + ATO combined treatment significantly reduces not just reactive oxygen species (ROS) but also proinflammatory cytokines and chemokine production that significantly increases the severity of DS.

ATO Alters the Cellular Morphology of Differentiating NB4 Cells
Several clinical studies have reported that the morphology of APL cells changed upon ATRA or ATO treatments. Apoptotic and necrotic cells appeared after ATO-induced cell death, exhibiting a variable size and quality of cytoplasm [26].
To determine the differentiation stages of NB4 cells, a variety of morphological changes were evaluated for the single and the combined treatments. In the case of ATRA treatment, the cells mainly represent differentiated and mitotic cells (blue and green triangles, Figure 1A). ATRA, ATO 0.5 µM and the combination treatment caused less apoptosis in NB4 WT and NB4 TG2-C (virus control) cells, such as TG2-deficient NB4 TG2-KD and NB4 TG2-KO cells (Figure 1(B1,B2), upper panels, green and black bars). Results showed that at day 5, arsenic trioxide induces a time-and dose-dependent cytotoxic effect on morphology, representing damaged and late apoptotic-necrotic phases. Higher concentrations of ATO were associated with an increased number of apoptotic and necrotic cells (Figure 1(B1,B2), lower panels, green and black bars). While the number of differentiated cells was low in NB4 WT and NB4 TG2-C cell lines, a higher apoptotic rate was seen in the ATRA + ATO 2.0 µM combined treatment compared to the ATRA + ATO 0.5 µM treatment of cells (Figure 1(B2), upper panels). There was a higher degree of apoptosis or necrosis in NB4 TG2-KD or NB4 TG2-KO cells (Figure 1(B2), upper panel, black and dark grey bars). ATO treatment alone resulted in minor differentiation, with ATO concentration-dependent apoptosis in the NB4 cell lines (Figure 1(B1,B2), lower panels).

ATRA + ATO Combined Treatment Decreases Differentiated NB4 Cells' Ability to Produce ROS
We previously reported that the atypical expression of TG2 greatly enhances neutrophil granulocytes' production of ROS by enhancing the expression of two important components of the NADPH-oxidase complex, NCF-2/P67PHOX and GP91PHOX. ATO treatment caused significant cellular changes in NB4 cell lines, which may affect the production of ROS. Because the NADPH-oxidase system is responsible for ROS production, we sought to determine the extent of ROS production after ATRA/ATO treatments. Both NCF-2/P67PHOX and GP91PHOX mRNA expression levels were measured at 1 µM ATRA, 0.5 µM, 2.0 µM ATO, respectively, and ATRA + ATO combined treatments at days 0, 3 and 5. While the levels of mRNS expression of both genes showed a similar pattern, especially on the fifth day, displaying a TG2-dependent expression after ATRA treatment, ATO treatments resulted in a magnitude of gene expression almost similar to that of ATRA generated in NB4 WT cells (Figure 2(A1,A2,B1,B2), left side). In combined treatments (ATRA + ATO, 0.5 and 2.0 µM), as a consequence both of the combined treatment and the extent of TG2 quantities, expression values remained low compared to ATRA or ATO treatments alone (Figure 2(A3,A4,B3,B4), right side). These expression values were also reflected in the production of ROS, especially in the ATRA + ATO 2.0 µM treatment, where a 1/3 ROS generating capacity was measured compared to the ROS production with ATRA or ATO treatment alone, depending on the amount of TG2 (Figure 2(C1-C4)).

Combined ATRA + ATO Treatment Markedly Reduces Inflammatory Biomarker Expression
DS, which can be fatal in 2.5-30% of cases in its moderate or severe forms, is characterized by the presence of a large number of inflammatory, differentiated leukemic cells in the bloodstream that synthesize and secrete chemokines and cytokines, triggering a so-called "cytokine storm." We previously showed that MCP-1, IL-1β and TNFα were secreted in a TG2-quantity-dependent manner in differentiated NB4 cell lines. MCP-1, IL-1β and TNFα were measured at the mRNA and protein levels in ATRA, ATRA + ATO 0.5 µM, and ATRA + ATO 2.0 µM differentiated NB4 WT, NB4 TG2-C, NB4 TG2-KD and NB4 TG2 KO-cell culture systems. Overall, at day 5, mRNA levels of MCP-1, IL-1β and TNFα were approximately 50% lower for the combined ATRA + ATO 2.0 µM treatment than for ATRA, but these values were further reduced in a TG2-quantity-dependent manner ( Figure 3(A1,B1,C1)). At day 5, we observed the effective inhibition of MCP-1, IL-1β and TNFα (with 5×, 10× and 20× lower values, respectively, than the controls) in the case of NB4 WT cells in combined therapy, especially ATRA + ATO 2.0 µM, with respect to proinflammatory cytokines and the chemokine content of cell culture supernatants. These values were further reduced in a TG2-quantity-dependent manner (Figure 3(A2,B2,C2), Table S1).  Graphs are the representation of mean ± SD values normalized to 100 µg protein of cell lysate content. Statistical significance was determined via two-way analysis of variance (ANOVA; Bonferroni and Tukey post-hoc test; ATRA vs. ATRA+NC9 * p < 0.05, ** p < 0.01 and *** p < 0.001, **** p < 0.0001).
Normalised production in pg/mL

Discussion
ATRA-based therapy is frequently used in the clinical treatment of APL patients, which leads to the terminal differentiation of leukemic cells towards the neutrophil granulocyte stage. Hyperinflammatory reactions may constitute severe side effects of ATRA treatment, including the infiltration and damaging of soft tissues and organs, such as the lungs and heart, by differentiating APL cells that produce reactive oxygen species. ATRA-induced differentiation of APL cells could generate increased secretion of cytokines, chemokines and interleukins as well as cell adhesion and migration. ATRA-induced differentiation is also associated with the elevated expression of two components of the NADPH-oxidase complex, NCF-2/P67PHOX and GP91PHOX, resulting in the possibility of increased ROS production and, consequently, more severe organ damage.
It has previously been reported that the morphology of APL cells changed upon ATRA or ATO treatment. ATO treatment was associated with apoptotic or necrotic cell death, exhibiting variable size and quality of cytoplasm [26]. Notably, reduced or absent TG2 enhanced the sensitivity of NB4 cells to a combined ATRA + ATO 2.0 µM treatment, with significantly higher apoptotic and necrotic rates.
Oxidative stress caused by reactive oxygen species, a group of oxygen-based reactive molecules produced by ATO activated the NADPH-oxidase system, resulting in the disruption of mitochondrial membrane potential and subsequent apoptosis [27][28][29]. As published previously, the mRNA expression of NCF2/P67PHOX and GPPHOX91 were substantially higher in the presence of TG2 compared to NB4 TG2-KD and TG2-KO cells. Here we find that ATO alone can trigger an increase in the amount of mRNA of NCF2 and GPPHOX91 as well as the production of ROS, similar to ATRA in differentiated NB4 cell lines. So far, no study has shown a 2-fold decrease in ROS production in response to combined ATRA + ATO compared to single ATRA or ATO treatments. While amounts of atypically expressed TG2 in differentiated NB4 cell lines can enhance the function of the NAPDH-oxidase system, leading to high ROS production, TG2 deficiency in TG2-KD cells or TG2-KO cells is associated with significantly lower ROS production, which may be further reduced by the combined ATO treatments.
We have previously shown that the TG2 inhibitor NC9 decreases NF-κB translocation to the nucleus and NF-κB transcriptional activity as well as significantly reducing the production of inflammatory biomarkers, such as MCP-1, IL-1β and TNF-α [20]. Accordingly, our current study demonstrates that combined ATRA + ATO (2.0 µM) with an inhibition of atypically expressed TG2 by NC9 radically limits the expression and secretion concentration of three inflammatory biomarkers and radical oxygen species ( Figure 5(C1-C4)). ATRA + ATO (2µM) + NC9 treatment resulted in an 8-fold decrease in MCP-1, a 15-fold decrease in IL-1β levels and a 61-fold suppression of TNFα; there was also a 28-fold reduction in ROS production, compared to ATRA alone, at day 5.

Cytospin
Samples were taken from a homogeneous suspension of cell cultures. After pre-cleaning the slides (70% alcohol), 10 µL of a homogeneous sample together with 90 µL of 1x phosphate-buffered saline was applied to the Cytospin™ tube (Shandon CYTOSPIN II, 6511 Bunker Lake Blvd. Ramsey, MN 55303 USA), followed by centrifugation at 800 rpm for 3 min. Samples were then fixed at room temperature with methanol and were prepared for further staining.

May-Grünwald Giemsa Staining
May-Grünwald and Giemsa solutions were diluted with distilled water at a 1:10 ratio. Previously fixed samples were stained with a May-Grünwald solution for 10 min and were then rinsed with a diluted Giemsa solution for 5-30 min. Slides were washed and dried at room temperature. Light microscope images and documentation were obtained using a FLoid ® Cell Imaging Station instrument (Life Technologies) at a scale of 200 µm. The ratio of the undifferentiated, differentiated, apoptotic, necrotic and mitotic cells in NB4 cell lines was determined by morphological changes/features during the ATRA, ATO and combined ATRA + ATO treatments. Based on our morphological type evaluations, we classified the morphological types of ATRA-, ATO-and ATRA + ATO-differentiated NB4 cells into six groups: a, undifferentiated (unsegmented nuclear region and thin cytoplasmic region); b, differentiated (a segmented nuclear fraction with the white-grey higher proportion of the cytoplasmic region); c, apoptotic (well-defined membrane changes, with shrinkage) and d: necrotic (lack of nuclear fraction with severe destruction of the membrane structure). Other groups, such as: e, apoptotic-necrotic (strong blue nuclear remnants, disarrayed membrane structure) and f, mitotic (chromatin changes, round shape) were also defined from the ATRA and ATO combined treatment.

Gene Expression Analysis
For Q-PCR measurements, total RNA samples were isolated by TRIzol reagent, following the company's instructions. Total RNA was quantified by a NanoDrop 2000 Spectrophotometer (Thermo Fisher, Waltham, MA USA 02451). Each sample was diluted to 200 ng/µL concentration, followed by reverse transcription using the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher) in a reaction of: 10 µL sample + 10 µL RT-Master mix. The assay and the PCR were performed according to the manufacturer's protocol. For the real-time Q-PCR reaction, the following TaqMan probes (ABI, Applied Biosystems, Waltham, MA USA 02451) were used: NCF2, GPPhox91, IL-1β, MCP-1, TNF-α and GAPDH. The analysis was carried out using the ABI Prism 7900 (ABI, Applied Biosystems). Relative mRNA expression levels were normalized to the level of GAPDH using the ∆∆Ct method.

Superoxide Anion Production
The amount of superoxide radicals was measured by luminol-chemiluminescence assay, using L-012 dye (Wako Pure Chemical Industries, Ltd., 1-2 Doshomachi 3-chome, Chuo-ku, Osaka, Japan) after PMA (50 nM) induction, in a reaction volume of 100 µL medium containing cells and L-012 (50 µM) dye. After 5 min, samples were measured in a Synergy Multimode Microplate Reader (BioTek Instruments, Inc, Winooski, VT, USA). Production of generated light by the reaction was recorded in relative luminescence units (RLUs) and was corrected with the protein concentration levels of the samples.

Conclusions
Together, these results suggest that the atypical expression level of TG2 in ATRA-induced differentiating APL cells is a crucial factor in developing inflammation and in the production of ROS. The lower expression of TG2 may effectively reduce the chance of inflammatory processes and organ damage.
In the combined ATRA + ATO treatment of differentiating APL cells, ATO can restrict ATRA-induced ROS and inflammatory biomarker production capacity in a dose-dependent manner. The irreversible inhibitor of TG2 NC9 not only decreased reactive oxygen species production 28-fold, but decreased the concentration of MCP-1, IL-1β and TNF-α 8-, 15-and 61-fold, respectively in the combined ATRA + ATO-treated wild-type NB4 cell culture.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/12/3/648/s1, Table S1: Protein content of supernatant of NB4 cell lines, Table S2: Protein content of supernatant of NB4 cell lines, Table S3: List of chemicals and reagents were used, Figure S1: Representative images of May-Grünwald-Giemsa staining, Figure S2: Production of ROS in vehicle treated NB4 cells, Figure S3: mRNA expression of MCP-1, IL-1β and TNFα inflammatory biomarkers, Figure S4: The protein content of the supernatants of NB4 cell lines was quantified by ELISA.