A cell-penetrating peptide enhances delivery and efficacy of phosphorodiamidate morpholino oligomers in mdx mice.

Antisense RNA technology is a strategy for the treatment of Duchenne muscular dystrophy (DMD), a progressive and universally fatal X-linked neuromuscular disease caused by frameshift mutations in the gene encoding dystrophin. Phosphorodiamidate morpholino oligomers (PMOs) are an antisense RNA platform that is used clinically in patients with DMD to facilitate exon skipping and production of an internally truncated, yet functional, dystrophin protein. Peptide-conjugated PMOs (PPMOs) are a next-generation platform in which a cell-penetrating peptide is conjugated to the PMO backbone, with the goal of increasing cellular uptake. RC-1001 is a PPMO that contains a proprietary cell-penetrating peptide and targets the Dmd mutation in mdx mice. It was evaluated in mdx mice for exon 23 skipping, dystrophin production, and functional efficacy. Single-dose RC-1001 dose dependently increased exon skipping and dystrophin protein levels in striated muscle and is associated with improvements in muscle function. Dystrophin protein levels were durable for 60 days. Three doses, each given 1 month apart, increased exon skipping to 99% in quadriceps and 43% in heart, with dystrophin protein levels at 39% and 9% of wild type, respectively. These findings support clinical development of PPMO therapies for the treatment of DMD.

Deep Learning Enables Discovery of a Short Nuclear Targeting Peptide for Efficient Delivery of Antisense Oligomers.

Therapeutic macromolecules such as proteins and oligonucleotides can be highly efficacious but are often limited to extracellular targets due to the cell's impermeable membrane. Cell-penetrating peptides (CPPs) are able to deliver such macromolecules into cells, but limited structure-activity relationships and inconsistent literature reports make it difficult to design effective CPPs for a given cargo. For example, polyarginine motifs are common in CPPs, promoting cell uptake at the expense of systemic toxicity. Machine learning may be able to address this challenge by bridging gaps between experimental data in order to discern sequence-activity relationships that evade our intuition. Our earlier data set and deep learning model led to the design of miniproteins (>40 amino acids) for antisense delivery. Here, we leveraged and expanded our model with data augmentation in the short CPP sequence space of the data set to extrapolate and discover short, low-arginine-content CPPs that would be easier to synthesize and amenable to rapid conjugation to desired cargo, and with minimal in vivo toxicity. The lead predicted peptide, termed P6, is as active as a polyarginine CPP for the delivery of an antisense oligomer, while having only one arginine side chain and 18 total residues. We determined the pentalysine motif and the C-terminal cysteine of P6 to be the main drivers of activity. The antisense conjugate was able to enhance corrective splicing in an animal model to produce functional eGFP in heart tissue in vivo while remaining nontoxic up to a dose of 60 mg/kg. In addition, P6 was able to deliver an enzyme to the cytosol of cells. Our findings suggest that, given a data set of long CPPs, we can discover by extrapolation short, active sequences that deliver antisense oligomers.

Pharmacokinetic/Pharmacodynamic Modeling of a Cell-Penetrating Peptide Phosphorodiamidate Morpholino Oligomer in mdx Mice.

PURPOSE: Peptide-conjugated phosphorodiamidate morpholino oligomers (PPMOs) have shown promise in treating Duchenne muscular dystrophy (DMD). We evaluated a semi-mechanistic pharmacokinetic (PK) and pharmacodynamic (PD) model to capture the relationship between plasma and muscle tissue exposure/response in mdx mice treated by mouse surrogate PPMO. METHODS: A single or repeated (every 4 weeks for 20 weeks) intravenous PPMO dose was administered to mdx mice (n = 6/timepoint). A PK/PD model was built to characterize data via sequential modeling. A 2-compartment model was used to describe plasma PK. A simultaneous tissue PK/PD model was subsequently developed: 2-compartment model to describe muscle PK; linked to an indirect response model describing stimulation of synthesis of skipped transcript, which was in turn linked to stimulation of synthesis of dystrophin protein expression. RESULTS: Model performance assessment via goodness-of-fit plots, visual predictive checks, and accurate parameter estimation indicated robust fits of plasma PK and muscle PK/PD data. The model estimated a PPMO tissue half-life of 5 days-a useful parameter in determining the longevity of PPMOs in tissue and their limited accumulation after multiple doses. Additionally, the model successfully described dystrophin expression after single dosing and associated protein accumulation after multiple dosing (increasing ~ twofold accumulation from the first to last dose). CONCLUSIONS: This first PK/PD model of a PPMO in a DMD disease model will help characterize and predict the time course of PK/PD biomarkers in mdx mice. Furthermore, the model framework can be used to develop clinical PK/PD models and can be extended to other exon-skipping therapies and species.

Deep learning to design nuclear-targeting abiotic miniproteins.

There are more amino acid permutations within a 40-residue sequence than atoms on Earth. This vast chemical search space hinders the use of human learning to design functional polymers. Here we show how machine learning enables the de novo design of abiotic nuclear-targeting miniproteins to traffic antisense oligomers to the nucleus of cells. We combined high-throughput experimentation with a directed evolution-inspired deep-learning approach in which the molecular structures of natural and unnatural residues are represented as topological fingerprints. The model is able to predict activities beyond the training dataset, and simultaneously deciphers and visualizes sequence-activity predictions. The predicted miniproteins, termed 'Mach', reach an average mass of 10 kDa, are more effective than any previously known variant in cells and can also deliver proteins into the cytosol. The Mach miniproteins are non-toxic and efficiently deliver antisense cargo in mice. These results demonstrate that deep learning can decipher design principles to generate highly active biomolecules that are unlikely to be discovered by empirical approaches.

p21-activated kinase 2 regulates HSPC cytoskeleton, migration, and homing via CDC42 activation and interaction with ß-Pix.

Cytoskeletal remodeling of hematopoietic stem and progenitor cells (HSPCs) is essential for homing to the bone marrow (BM). The Ras-related C3 botulinum toxin substrate (Rac)/cell division control protein 42 homolog (CDC42) effector p21-activated kinase (Pak2) has been implicated in HSPC homing and engraftment. However, the molecular pathways mediating Pak2 functions in HSPCs are unknown. Here, we demonstrate that both Pak2 kinase activity and its interaction with the PAK-interacting exchange factor-ß (ß-Pix) are required to reconstitute defective ITALIC! Pak2 (ITALIC! Δ/Δ)HSPC homing to the BM. Pak2 serine/threonine kinase activity is required for stromal-derived factor-1 (SDF1α) chemokine-induced HSPC directional migration, whereas Pak2 interaction with ß-Pix is required to regulate the velocity of HSPC migration and precise F-actin assembly. Lack of SDF1α-induced filopodia and associated abnormal cell protrusions seen in ITALIC! Pak2 (ITALIC! Δ/Δ)HSPCs were rescued by wild-type (WT) Pak2 but not by a Pak2-kinase dead mutant (KD). Expression of a ß-Pix interaction-defective mutant of Pak2 rescued filopodia formation but led to abnormal F-actin bundles. Although CDC42 has previously been considered an upstream regulator of Pak2, we found a paradoxical decrease in baseline activation of CDC42 in ITALIC! Pak2 (ITALIC! Δ/Δ)HSPCs, which was rescued by expression of Pak2-WT but not by Pak2-KD; defective homing of ITALIC! Pak2-deleted HSPCs was rescued by constitutive active CDC42. These data demonstrate that both Pak2 kinase activity and its interaction with ß-Pix are essential for HSPC filopodia formation, cytoskeletal integrity, and homing via activation of CDC42. Taken together, we provide mechanistic insights into the role of Pak2 in HSPC migration and homing.

Loss of function of TET2 cooperates with constitutively active KIT in murine and human models of mastocytosis.

Systemic Mastocytosis (SM) is a clonal disease characterized by abnormal accumulation of mast cells in multiple organs. Clinical presentations of the disease vary widely from indolent to aggressive forms, and to the exceedingly rare mast cell leukemia. Current treatment of aggressive SM and mast cell leukemia is unsatisfactory. An imatinib-resistant activating mutation of the receptor tyrosine kinase KIT (KIT D816V) is most frequently present in transformed mast cells and is associated with all clinical forms of the disease. Thus the etiology of the variable clinical aggressiveness of abnormal mast cells in SM is unclear. TET2 appears to be mutated in primary human samples in aggressive types of SM, suggesting a possible role in disease modification. In this report, we demonstrate the cooperation between KIT D816V and loss of function of TET2 in mast cell transformation and demonstrate a more aggressive phenotype in a murine model of SM when both mutations are present in progenitor cells. We exploit these findings to validate a combination treatment strategy targeting the epigenetic deregulation caused by loss of TET2 and the constitutively active KIT receptor for the treatment of patients with aggressive SM.

A unique carboxyl-terminal insert domain in the hematopoietic-specific, GTPase-deficient Rho GTPase RhoH regulates post-translational processing.

RhoH is a hematopoietic-specific, GTPase-deficient member of the Rho GTPase family that was first identified as a hypermutable gene in human B lineage lymphomas. RhoH remains in a constitutively active state and thus its effects are regulated by expression levels or post-translational modifications. Similar to other small GTPases, intracellular localization of RhoH is dependent upon the conserved "CAAX" box and surrounding sequences within the carboxyl (C) terminus. However, RhoH also contains a unique C-terminal "insert" domain of yet undetermined function. RhoH serves as adaptor molecule in T cell receptor signaling and RhoH expression correlates with the unfavorable prognostic marker ZAP70 in human chronic lymphocytic leukemia. Disease progression is attenuated in a Rhoh(-/-) mouse model of chronic lymphocytic leukemia and treatment of primary human chronic lymphocytic leukemia cells with Lenalidomide results in reduced RhoH protein levels. Thus, RhoH is a potential therapeutic target in B cell malignancies. In the current studies, we demonstrate that deletion of the insert domain (LFSINE) results in significant cytoplasmic protein accumulation. Using inhibitors of degradation pathways, we show that LFSINE regulates lysosomal RhoH uptake and degradation via chaperone-mediated autophagy. Whereas the C-terminal prenylation site is critical for ZAP70 interaction, subcellular localization and rescue of the Rhoh(-/-) T cell defect in vivo, the insert domain appears dispensable for these functions. Taken together, our findings suggest that the insert domain regulates protein stability and activity without otherwise affecting RhoH function.

RhoH is critical for cell-microenvironment interactions in chronic lymphocytic leukemia in mice and humans.

Trafficking of B-cell chronic lymphocytic leukemia (CLL) cells to the bone marrow and interaction with supporting stromal cells mediates important survival and proliferation signals. Previous studies have demonstrated that deletion of Rhoh led to a delayed disease onset in a murine model of CLL. Here we assessed the impact of RhoH on homing, migration, and cell-contact dependent interactions of CLL cells. Rhoh(-/-) CLL cells exhibited reduced marrow homing and subsequent engraftment. In vitro migration toward the chemokines CXCL12 and CXCL13 and cell-cell interactions between Rhoh(-/-) CLL cells and the supporting microenvironment was reduced. In the absence of RhoH the distribution of phosphorylated focal adhesion kinase, a protein known to coordinate activation of the Rho GTPases RhoA and Rac, appeared less polarized in chemokine-stimulated Rhoh(-/-) CLL cells, and activation and localization of RhoA and Rac was dysregulated leading to defective integrin function. These findings in the Rhoh(-/-) CLL cells were subsequently demonstrated to closely resemble changes in GTPase activation observed in human CLL samples after in vitro and in vivo treatment with lenalidomide, an agent with known influence on microenvironment protection, and suggest that RhoH plays a critical role in prosurvival CLL cell-cell and cell-microenvironment interactions with this agent.

GLP-1-derived nonapeptide GLP-1(28-36)amide inhibits weight gain and attenuates diabetes and hepatic steatosis in diet-induced obese mice.

BACKGROUND: The metabolic syndrome is an obesity-associated disease manifested as severe insulin resistance, hyperlipidemia, hepatic steatosis, and diabetes. Previously we proposed that a nonapeptide, FIAWLVKGRamide, GLP-1(28-36)amide, derived from the gluco-incretin hormone, glucagon-like peptide-1 (GLP-1), might have insulin-like actions. Recently, we reported that the nonapeptide appears to enter hepatocytes, target to mitochondria, and suppress glucose production and reactive oxygen species. Therefore, the effects of GLP-1(28-36)amide were examined in diet-induced obese, insulin-resistant mice as a model for the development of human metabolic syndrome. METHODS AND RESULTS: Three- to 11-week infusions of GLP-1(28-36)amide were administered via osmopumps to mice fed a very high fat diet (VHFD) and to control mice on a normal low fat diet (LFD). Body weight, DXA, energy intake, plasma insulin and glucose, and liver triglyceride levels were assessed. GLP-1(28-36)amide inhibited weight gain, accumulation of liver triglycerides, and improved insulin sensitivity by attenuating the development of fasting hyperglycemia and hyperinsulinemia in mice fed VHFD. GLP-1(28-36)amide had no observable effects in control LFD mice. Surprisingly, the energy intake of peptide-infused obese mice is 25-70% greater than in obese mice receiving vehicle alone, yet did not gain excess weight. CONCLUSIONS: GLP-1(28-36)amide exerts insulin-like actions selectively in conditions of obesity and insulin resistance. The peptide curtails weight gain in diet-induced obese mice in the face of an increase in energy intake suggesting increased energy expenditure. These findings suggest utility of GLP-1(28-36)amide, or a peptide mimetic derived there from, for the treatment of insulin resistance and the metabolic syndrome.