A Novel Fluorobenzyl Polyethylene Glycol Conjucated Tetraiodothyroacetic Acid (fb-PMT), Targeting Thyrointegrin αvβ3 in Treatment Acute Myeloid Leukemia

Background: Acute myeloid leukemia (AML) is associated with poor long-term survival, even with newer therapeutic agents. Here we show the results of our pre-clinical study, in which we evaluate the ecacy of a new thyrointegrin αvβ3 antagonist, named fluorobenzyl Polyethylene glycol conjucated tetraiodothyroacetic acid (fb-PMT). Methods and Results: fb-PMT effectively suppresses the malignant growth of human acute myeloid leukemia (AML) after successful engraftment in transgenic NSG-S xenograft mouse models of either established human AML cell line or primary AML cells. Daily treatment with fb-PMT (1-10 mg/kg body weight) subcutaneously (s.c.) for 3-4 weeks was associated with marked regression of leukemogenesis and extended survival in both models. The eciency of the fb-PMT therapy was veried using IVIS imaging, owcytometry and histopathological examination to monitor the engraftment of leukemic cells in the bone marrow and other organs. fb-PMT therapy for 3-4 weeks at 3 and 10 mg/kg daily doses exhibited signicant reduction (P<0.0001) of leukemic cell burden of 74% and >95%, respectively. All fb-PMT-treated mice in the 10 mg/kg treatment arm successfully maintained remission after discontinuing the daily treatment. Comprehensive fb-PMT safety assessments demonstrated excellent safety and tolerability at multiple folds above the anticipated human therapeutic doses. Lastly, our genome-wide microarray screens demonstrated that fb-PMT works through the molecular interference mechanism with multiple signaling pathways contributing to growth and survival of leukemic cells. Conclusion: our preclinical ndings of the potent anticancer activities of fb-PMT and its favorable safety proles warrant its clinical investigation for the effective and safe management of AML. incubated for at once in containing human albumin (HSA), and by ow cytometry. Data acquisition using a FACS Aria III (BD Biosciences) equipped with an argon and was performed using Cell Quest (BD

of compounds were added in the presence or absence of brinogen and incubated for 2 h at room temperature, and then wells were washed three times with buffer A and incubated with a streptavidin − Horseradish peroxidase (HRP) conjugate (1:1000 in buffer A) for 1 h at room temperature. Finally, wells were washed three times with buffer A, and 100 µL peroxidase substrate 3,3′,5,5′tetramethylbenzidine (TMB) was added, and the reaction was terminated after 30 min with 50 µL of 450 nm stop solution for TMB.
Absorbance was determined at 450 nm with a plate reader.

Animals and treatment protocols
Mice were used in accordance with Public Health Service Policy on Humane Care and Use of Laboratory Animals and approved by the Albany VA Medical Center (Albany, NY, USA) institutional animal care and use committee (IACUC) (protocol number 545017). Eighty male NSG-S mice (6-8 weeks of age) were purchased from Jackson Laboratories (Bar Harbor, ME). Preconditioning was done by intraperitoneal injection of busulfan (30 mg/kg, Otsuka America Pharmaceutical Inc., Hayward, CA, USA) 24 hours prior to cell injections. K562-Luc cells and primary AML cells (6373) (5 to10 x 10 6 per mouse) were transplanted via tail vein injection into mice.
For the K562-Luc animal model (40 mice), in vivo imaging system (IVIS) (Perkinelmer, Waltham, MA, USA) scans and the peripheral blood smears examination were performed on animals once a week. The fb-PMT treatment schedule was initiated on day 10 postimplantation when increased counts of blast cells became evident in peripheral blood smears and con rmed by IVIS signals. fb-PMT was administered subcutaneously (s.c.) daily at 3 different doses (1, 3, and 10 mg/kg body weight) or vehicle, PBS (control) for 21 days for both the ON arm (21 days treatment and then 20 mice have been sacri ced) and the ON + OFF arm (21 days treatment followed by 14 days treatment discontinuation and then the remaining 20 mice have been sacri ced). Control animals were administered with vehicle (phosphate buffered saline (PBS), pH 7.4) daily.
In experiments using the primary AML cell animal model (40 mice), the treatment protocol was initiated on animals after con rmation of successful engraftment. Treatment was initiated on day 40 post-implantation with fb-PMT (1, 3, and 10 mg/kg) or control (vehicle, PBS) daily s.c. for 28 days. Twenty animals (ON arm) were humanely sacri ced after 28 days of treatment, and peripheral blood smears and bone marrow aspirates were examined histologically at the end of the 28 days of treatment. To evaluate the relapse after treatment, the remaining 20 mice (ON + OFF arm) were maintained without treatment for an additional 14 days, then they were humanely sacri ced and processed to obtain samples of peripheral blood smears, bone marrow aspirates and organs for histological examination. The maintaining for 14 days (ON + OFF) was established based on the animals' condition in the control groups.

Assessment of Leukemic cells engraftment by ow cytometry and immunohistostaining
Human AML engraftment was assessed by owcytometry and de ned as the percentage of human CD45+/CD33 + cells in total live mononuclear cells. Fresh bone marrow cells from NSG-S mice engrafted with K562-Luc and primary AML cells (6373) were collected at day 10 and 40 post-engraftment, respectively, once the blast cells detected in peripheral blood smear. samples were stained with antibodies for cell surface markers: anti-human CD45-PE, anti-human CD33-FITC (BD Biosciences, San Jose, CA, USA). Cells were incubated with monoclonal antibodies for 15 min at room temperature, washed once in PBS containing 0.1% human serum albumin (HSA), and analyzed by ow cytometry. Data acquisition was performed using a FACS Aria III (BD Biosciences) equipped with an argon and red diode laser, and analysis was performed using Cell Quest software (BD Biosciences).
As the K562-Luc cells express dim CD34, immunohistochemistry was performed for Formalin-xed decalci ed femurs from primary AML cells (6373) transplanted mice were para n-embedded and sectioned at 5 µm sections. Slides were stained using human anti-CD34 primary antibody (R&D system, Minneapolis, MN USA) then with HRP conjugated secondary antibody (Cell Signaling Technology Inc. Danvers, MA, USA). HRP activity was detected by diaminobenzidine tetrahydrochloride (DAB) and the slide were counterstain by Methyl green.

Safety study in animal models
The objective of this study was to evaluate the safety of fb-PMT after a period of daily s.c injections for 28 days in multiple species.
Clinical chemistry, biomarkers and histopathological examination were carried out (data not shown).

RNA isolation from AML cells and microarrays
K562-Luc and KG1a cells were cultured in 50 cm² cell culture asks with 10 mL phenol red free RPMI media containing 10% FCS to 75% con uence. K562-Luc and KG1a cells were treated (at 50% con uence) with 30 µM fb-PMT for 48 hours. Total RNA was immediately isolated from harvested cells using Trizol and checked for quality using an Agilent Bioanalyzer (Agilent Technologies, Santa Clara, CA) before being used for microarray analysis. The quality and the concentrations of the extracted RNA were analyzed using a NanoDrop (Thermo Fisher Scienti c, Waltham, MA) and the Agilent Bioanalyzer. RNA samples (100 ng) deemed to be of su cient quality (RIN greater than 8) were processed according to the standard Affymetrix RNA labeling protocol. At least two independent biological replicates of control and treated samples were concurrently interrogated in microarray analyses. In preliminary experiments, the treatment dose and duration were carefully selected so as to not signi cantly affect growth and survival of target cells for the duration of experiments.
Microarray Analysis. Labeled RNA samples were processed for hybridization employing the Clariom™ S human array platform (Affymetrix, Santa Clara, CA) at the Center for Functional Genomics, University at Albany, Rensselaer, NY. Brie y, 100 ng of total RNA was processed using the WT Plus Reagent kit (Affymetrix). Sense target Complementary DNA (cDNAs) were generated using the standard Affymetrix WT protocol and hybridized to Affymetrix Human Clariom S arrays. Arrays were washed, stained, and scanned on a GeneChip 3000 7G scanner using Affymetrix GeneChip Command Console Software (AGCC). Transcriptome Analysis Console Software (TAC v3.0.1.5) was used to identify differentially expressed genes (DEGs). Brie y, the CEL les were summarized using the SST-RMA algorithm in TAC and the normalized data were subjected to one-way ANOVA with a Benjamin Hochberg False Discovery Rate correction included (P < 0.05). A 1.5-fold expression change cut-off was used to select entities that were statistically differentially expressed between the conditions being compared (treated and untreated groups). In the standard work ow protocol, the fragmented biotinlabeled cDNAs were hybridized for 16 h to Affymetrix Arrays, scanned on an Affymetrix Scanner 3000 7G using AGCC software, and processed as described above. Alternatively, CEL les after QC screening using Affymetrix Expression Console software were imported into GeneSpring GX11.5 (Agilent Technologies). The data was then quantile normalized using PLIER and baseline transformed to the median of the control samples. The probe sets were further ltered to exclude the bottom 20th percentile across all samples. The resulting entity lists were subjected to an unpaired T-test with the Benjamini-Hochberg False Discovery rate correction and a 1.5-fold expression changes lter to identify differentially expressed transcripts between the control and test conditions at a p-value < 0.05. All analyzed and reported data are MIAME compliant and the raw data have been deposited in Gene Expression Omnibus (GEO; GSE95790) as detailed on the Microarray Gene Expression Data Society (MGED) website (http://www.mged.org/Workgroups/MIAME/miame.html).
Overall, the work ow of the microarray analyses was modeled based on previously published contributions [21].
Gene set enrichment analyses of DEGs were done using the Enrichr bioinformatics platform, which enables the interrogation of nearly 200,000 gene sets from more than 100 gene set libraries. The Enrichr API (January 2018 through October 2020 releases) [22,23] was used to test genes of interest for signi cant enrichment in numerous functional categories. When technically and analytically feasible, different sets of DEGs de ned at multiple signi cance levels of statistical metrics and comprising from dozens to several thousand individual genetic loci were analyzed using differential Gene set enrichment analysis (GSEA) to gain insights into biological effects of DEGs and infer potential mechanisms of anticancer activities. This approach was successfully implemented for identi cation and characterization of human-speci c regulatory networks governed by human-speci c transcription factor-binding sites [24] and functional enhancer element [25], 13,824 genes associated with 59,732 human-speci c regulatory sequences [26], 8,405 genes associated with 35,074 human-speci c neuroregulatory single-nucleotide changes [27]. Initial GSEA entail interrogations of each speci c set of DEGs using 29 distinct genomic databases, including comprehensive pathway enrichment Gene Ontology (GO) analyses followed by in-depth analyses of the selected genomic databases deemed most statistically informative. In all tables and plots (unless stated otherwise), in addition to the nominal p values and adjusted p values, the "combined score" calculated by Enrichr software is reported, which is a product of the signi cance estimate and the magnitude of enrichment (combined score c = log(p) * z, where p is the Fisher's exact test p-value and z is the z-score deviation from the expected rank).

Statistical analysis
An overall comparison of the means for all groups was carried out using a one-way ANOVA. Tukey con dence intervals were used to test for differences in means for each experimental group versus the control group. Results are presented as means ± S.D. A value of P < 0.05 indicated a statistically signi cant difference.

Results
In Vitro Binding A nity of fb-PMT with the thyrointegrin αvβ3.
In binding a nity experiments of fb-PMT, we con rmed that fb-PMT has a high a nity for the thyrointegrin αvβ3 receptors with a lower IC 50 (50% inhibitory concentration) of 0.23 nM (Fig. 1).

Preclinical in vivo therapy experiments revealed potent fb-PMT anticancer activities.
Effects of fb-PMT therapy on K562-Luc human leukemic cell line FLT3-ITD primary human AML cells engrafted in transgenic mice.
In the K562-Luc engrafted in transgenic mice, blast cells appeared in the blood smears of NSG-S mice 10 days after engraftment, with an average value of 40%. After 21 days of fb-PMT daily s.c. injection, blast cell counts continually and consistently decreased in a dosedependent manner in treated versus control groups, while animals in the control group showed increased blast cells in the peripheral blood. No blast cells could be detected in fb-PMT-treated animals (10 mg/kg) at the end of the treatment. Furthermore, there was no rebound increase in peripheral blast cells at 1 and 3 mg/kg with full sustained remission at fb-PMT dose of 10 mg/kg at 1-2 weeks post-discontinuation of treatment (Fig. 2, Supplementary Figure S1A).
On the other hand, primary AML cells (6373-FLT3-ITD) cells appeared in the blood smears of NSG-S mice 40 days after engraftment, with an average value of 26%. After 28 days of daily s.c. treatment, peripheral smears of treated animals were entirely normal at fb-PMT dose of 10 mg/kg. Daily s.c. injections of fb-PMT at 1, 3, and 10 mg/kg doses prevented blast cell expression/reproduction compared to controls by 54%, 75%, and 90.5%, respectively ( Fig. 2 and Supplemental Figure S1B).
After termination, the bone marrow K562-Luc engrafted mice with a daily treatment of fb-PMT at 3 mg/kg manifested 30-40% in ltration with blast cells, while 70% maturation could be detected. fb-PMT-treated animals at 10 mg/kg dose presented bone marrows with blast cell counts < 5% and > 95%, normal maturation have been documented (segmented neutrophils). The remission was maintained in all treated mice at least 2 weeks after fb-PMT therapy discontinuation (Fig. 3A, 3B and Supplemental Figure S2A). IVIS scans and histopathological results at sacri ce showed a dose-dependent decrease of brain, lung, liver, and spleen in ltration with the leukemic cells in the group of fb-PMT-treated mice (10 mg/kg) in comparison to control group Supplemental Figure S3. The fb-PMT therapy at 10 mg/kg dose in the ON + OFF treatment group resulted in successfully maintained remission in all animals 2 weeks after withdrawal of the daily treatment. The sustained remission was con rmed using blood smear analyses, IVIS scans, owcytometry and histopathological examinations.
Regarding the bone marrow samples from mice engrafted with primary AML cells (FLT3-ITD), the fb-PMT-treated group (10 mg/kg) restored the normal bone marrow maturation with abundant megakaryocytes in comparison to control animals ( Fig. 3C and 3D). The results were con rmed with owcytometry and immunohistochemistry analysis (Supplemental Figure S2) Furthermore, we evaluated the splenic in ltration in our animal models. Histopathological results showed a marked decrease of splenic metastases of the leukemic cells in the group treated with fb-PMT (10 mg/kg) compared to control group (Fig. 4). Similarly, to the K562-Luc AML experiments, the primary AML model (ON + OFF) group (10 mg/kg) manifested the successful maintenance of remission 2 weeks after withdrawal of daily therapy. The splenic weight showed marked decrease (80%) even with the low dose (1 mg/kg). The ON + OFF groups maintained normal splenic weight in comparison to control, which may re ect successful prevention of engraftment (   Legend: GES were identi ed based on the analyses of 12 down-regulated genes in fb-PMT-treated K562 cells; *, Statistical metrics were de ned by the Enrichr bioinformatics platform (Methods).   Table 1 identi ed a total of 25 genes, differential expression of which appears to de ne molecular signals of either activation of or interference with transcriptional pathways in fb-PMT-treated human AML cells (Fig. 5). GSEA of genes comprising the 25-gene and 12-gene expression signatures validated their signi cance in de ning observations of the molecular mimicry of transcriptional pathways' activation and interference induced by fb-PMT treatment in human AML cells (Figs. 5-7, Table 2). GSEA of all signi cant DEGs con rmed and extended these ndings.
Interestingly, GSEA identi ed the SNAI transcriptional pathway as the most signi cantly enriched pathway of the molecular interference observed in K562 cells treated with fb-PMT among either down-regulated or up-regulated DEGs (Supplemental Fig. 4). Additional examples of the speci c genes and pathways of potential functional signi cance revealed by the GSEA of 233 genes down-regulated in KG1a cells after fb-PMT treatment are shown in Table 3. Of note, GSEA of the LINCS L1000 Ligand Perturbations database of upregulated genes revealed evidence of molecular interference with functions of multiple growth factors in human cancer cell lines (Fig.  7).  [9]. Integrin αVβ3 has been reported to be more expressed in AML cells especially CD34-positive cells, monocytic leukemias, patient with NPM, FLT3-ITD [12].
Consistent with the previous reports on mechanisms of anticancer actions of thyrointegrin αvβ3 antagonists [29,[33][34][35][36][37], genome-wide microarray screens reported here demonstrated that fb-PMT appears to exert its potent anticancer actions on human AML cells through the molecular interference mechanism with multiple signaling pathways supporting growth and survival of leukemic cells. We detected signi cant molecular signals of transcriptional interference with gene expression induced in human cancer cells in response to multiple growth factors such as EGF, IGF-1, TGFA, and many others. Consistently, examples of the fb-PMT-induced GES of transcriptional pathway's activation include RB1, IRF9, MAML1, RAP1A, and GATA4 pathways, known biological functions of which appear highly consistent with the hypothesis that activation of these pathways may contribute to fb-PMT anticancer activity [53][54][55][56].
Finally, consistent with our previous reports on the crosstalk between integrin αvβ3 and estrogen receptor α (ERα), which contributes to the induced proliferation of cancer cells [57-60], we found that fb-PMT interfered with estrogen signaling in human AML cells. The αvβ3 agonist (thyroid hormone) was associated with increased phosphorylation and nuclear enrichment of ERα. Confocal microscopy indicated that both T4 and estradiol (E2) caused nuclear translocation of integrin αv and phosphorylation of ERα. The speci c ERα antagonist (ICI 182,780; fulvestrant) blocked T4-induced ERK1/2 activation, ERα phosphorylation, PCNA expression, and proliferation [57].
Furthermore, GLP toxicity study showed that fb-PMT is safe and well tolerated at > 60 fold higher than the effective anticancer doses after daily s.c injection for 4 weeks in rat and monkey models (data not shown).
fb-PMT is an effective anticancer agent against solid tumors and hematological malignancies, with broad spectrum, potent antiangiogenic activity against all known growth factors and other pro-angiogenesis stimuli [28, 61, 62]. Collectively, preclinical ndings of fb-PMT warrant its clinical investigation for the effective and safe management of AML.

Conclusion
Our novel thyrointegrin αvβ3 antagonist, fb-PMT, is preferable for potential clinical use because of its e cacy against human xenograft models of AML as well as its safety, even at high doses. Our genomic data demonstrated the potent anticancer actions on human AML through the molecular interference mechanism with multiple signaling pathways supporting growth and survival of leukemic cells. fb-PMT could have a broader application because it could be utilized, either alone or in combination with chemotherapeutic agents, to treat AML cancer or other cancers.