Kaja
Zuwala‡
ab,
Anton A. A.
Smith‡
a,
Martin
Tolstrup
*b and
Alexander N.
Zelikin
*ac
aDepartment of Chemistry, Aarhus University, Aarhus, Denmark. E-mail: zelikin@chem.au.dk
bDepartment of Infectious Diseases, Aarhus University Hospital, Denmark. E-mail: marttols@rm.au
ciNano Interdisciplinary Nanoscience Centre, Aarhus University, Aarhus, Denmark
First published on 5th January 2016
Histone deacetylase inhibitors (HDACi) and panobinostat in particular are currently in the focus of intensive investigation as latency reversing agents against the human immunodeficiency virus (HIV). Regretfully, HDACi have dose limiting side-effects making controlled, optimized methods for delivery of panobinostat highly warranted. This has proven to be highly challenging, predominantly because panobinostat has no readily available classic sites for bioconjugation. In this work, we address this challenge and present the first macromolecular prodrugs of panobinostat engineered using self immolative linkers (SIL) and a disulfide trigger for drug release upon cell entry. Synthetic methodology involved the development of a novel monomer with functionalities of SIL and activated ester for one-step polymer-analogous conjugation to drugs. In agreement with the design set forward, copolymers were stable in buffered solutions and released panobinostat at reducing conditions. Synthesized polymers were highly efficacious as latency reversing agents as monitored in three cell lines harboring latent HIV, at no expense to the cytotoxicity of treatment. The data presented herein provide broad pre-in vivo characterization of a promising prodrug system developed to address a global healthcare challenge, safe and efficient reversal of HIV latency.
One possible approach to minimize systemic side effects is to extend the residence of PANO in the blood (passive targeting) or engineer chemical methods to bring PANO to the target cells (active targeting). This can be accomplished using macromolecular prodrugs (MP)5–7 and/or antibody–drug conjugates (ADC),8,9 that is, macromolecular agents. Either case relies on the development of facile bioconjugation and engineered mechanisms of drug release.10 However, the chemistry of PANO does not present typical handles for bioconjugation and creation of biodegradable linkages (such as an amine or a thiol functionality). To our knowledge, there are no literature reports on ADC or MP of PANO. In this work, we specifically address this challenge and develop MP of PANO equi-efficacious to the parent drug in activating latent HIV as studied in three different cell lines, providing a sound in vitro evaluation of these prodrugs.
Therapeutic benefit delivered by MP is well celebrated in the anti-cancer treatment whereby high molar mass of the polymer–drug conjugates leads to an increased hydrodynamic radius of the molecule, enhanced blood residence time as compared to the parent drug, and accumulation of the conjugate in the tumour via the enhanced permeation–retention effect (EPR).5,6 Prolonged circulation times are also highly beneficial for delivery of anti-HIV medication. Thus, short half-life of the drugs can result in too low exposure and subsequently lead to development of drug resistance.11 In contrast, extended blood residence time provides a sufficient concentration of the circulating drug even in case of non-adherence and extended time between drug dosing.12 For PANO, the main obstacle to this is that chemistry of this molecule does not present the “classic” handles for bioconjugation and creation of biodegradable linkages (Fig. 1A). To solve this chemical incompatibility, we propose to use self-immolative linkers (SIL), a recently developed cunning tool of organic chemistry.13 With the view to achieve a triggered intracellular drug release, we used a thiol-containing SIL and conjugation to polymers via a disulfide linkage.14 The latter is among the most popular tools of bioconjugate chemistry and releases the drug once inside the cell via the glutathione-triggered thiol-disulfide exchange.15
MP were engineered based on a polymer with a documented characterization in polymer therapeutics and clinical evaluation, poly(N-2-hydroxypropyl methacrylamide), PHPMA.16 For bioconjugation, we designed a monomer with a functionality of activated ester connected to the methacrylate via a spacer comprised of a disulfide linkage and an SIL, Fig. 1B. Attractive aspects of this monomer are the ease of its synthesis and its capability of forming SIL to drugs with diverse structure, prior to or after the polymerization. In contrast to the existing “activated esters” or similar monomers designed towards conjugation of drugs to carrier (bio)polymers,17 this monomer has a built-in disulfide linkage – for a specific intracellular drug release; and an SIL – which tremendously accelerates drug release.18,19 We anticipate that this monomer will find diverse use in polymer therapeutics and broader drug delivery applications.
The activated monomer and HPMA were co-polymerized via the reversible addition–fragmentation chain transfer (RAFT) mechanism, Fig. 1C.20,21 We observed that conversions in these polymerizations were rather low, a likely explanation being a small release of 4-nitrophenol inhibiting the polymerization. Nevertheless, polymers were successfully obtained in a range of molar masses and content of the “activated” monomer. Conjugation of PANO to a polymer via a polymer-analogous reaction was first investigated via an in situ NMR monitoring which revealed a progressive conjugation reaction which was complete (with resolution of this technique) within 7 h (Fig. 2). This synthetic route afforded polymers with narrow dispersity (Ð < 1.3), molar mass from 6 to 13 kDa, and PANO content from 6 to 11 mol%.
Release of PANO from MP was studied using HPLC, Fig. 3. We have previously shown that the SIL linkage employed in this work remains stable at physiological conditions and undergoes a fast degradation upon a treatment with dithiothreitol or a natural intracellular thiol containing tripeptide, glutathione.18,19 In full agreement with this, PANO containing polymers showed only a negligible spontaneous drug release in physiological buffers but readily released PANO under reducing conditions (Fig. 3).
The utility of the MP containing PANO as LRA against HIV was first investigated in ACH2 and U1 cell models (T cell line and promonocyte cell line respectively) latently infected with HIV-1 by incubating the cells with the polymers at concentration 1 mg L−1. Pristine PHPMA revealed no activity as a latency reversing agent, Fig. 4. In contrast, MPs proved to be highly efficacious and for both cell lines, optimal compositions of MP matched the efficacy of 100 nM PANO (being therapeutically relevant concentration) at no expense to the cell toxicity of treatment. With regard to the optimal molar mass of the MP, no clear trend emerged from this dataset, most likely due to narrow range of studied molar mass. With regard to the MP drug loading, 11 kDa polymer sample with double the amount of conjugated drug was markedly more effective in activating the latent virus (see data for U1 cell line). EC50 value for MP (13 kDa, 6% drug loading) was established at 1 mg L−1 for both cell lines (for titration curves, see ESI Fig. SI5†). Expressed in mole equivalents of PANO, the polymers were ∼10-fold less potent than the pristine drug. This pharmacodynamic loss is not unexpected for polymer therapeutics22,23 and is due to e.g. decreased cell entry of MP compared to the free drug as well as incomplete release of the drug from the carrier, whereas the overall benefit of MP is typically due to optimization of pharmacokinetics of the drug on systemic and sub-cellular levels.5,6
Further characterization of the anti-latency treatment using MP was performed using a J-Lat full length clone 6.3 cell line which is a Jurkat derived cell line infected with a retroviral construct consisting of full length HIV-1 genome without env and nef, but containing eGFP (green fluorescent protein).24 eGFP is being co-expressed together with the HIV genome upon reactivation from latency. The green fluorescence of the cells can be quantitatively tracked with flow cytometry in parallel to the live/dead assay, Fig. 5. The two ways to activate the virus – using MP or pristine PANO – were near identical in terms of fraction of cells becoming positive for eGFP (in each data set – top left quadrangle) as well as the associated toxicity effect (bottom right quadrangle) – illustrating the success in the design of MP. Curiously, flow cytometry analysis revealed that toxicity of treatment with both MP and PANO agents was not due to activation of HIV and dead cells were devoid of the HIV-associated fluorescence. Upon a treatment with PANO or MP, a population of cells with activated HIV was not stained with a reagent for cell death, whereas the cells staining positive for cell death were devoid of activated HIV. This observation may present significant novelty for the rapidly developing field of HIV anti-latency treatment; we are now investigating this in detail.
Earlier we have explored the broader impact of PANO on immunomodulation.25,26 We identified a small but significant increase in T cell activation by PANO, an effect that may be significant for the drugs' ability to turn on transcription of integrated proviruses due to a more favorable cellular milieu. Therefore, we sought to determine the activation of T cells by MP through quantifying the expression of the early T cell activation marker CD69. The activation of human CD4+ T cells by MP of PANO was comparable to or higher than activation by pristine PANO used at 50 nM concentration, Fig. 6. Importantly, no cell death above background was detected in any of the samples as determined by flow cytometry. The highest expression of CD69 was achieved with MPs of PANO was at 11.1 mg L−1 at which concentration activation of T cells was comparable or exceeding that achieved by the pristine drug. Pristine polymer PHPMA did not induce any activation of T cells clearly indicating that activity of MPs lies primarily in the release of drug itself.
Taken together, the results presented above establish a novel bioconjugation approach to create MP of PANO, an HDACi agent with virtually no arms for conjugation. Resulting polymers were equi-efficacious to the parent drug in terms of net virus activation and CD4+ T cell activation. We expect that these results would lead to the advent of MP and ADC for this and other HDACi with similar, highly challenging for conjugation structure.
J-Lat cells are Jurkat derived cells virally infected with retroviral construct: HIV-R7/E-/GFP; full length HIV-1 without env and nef, which means that GFP (green fluorescent protein) is expressed together with HIV genome upon reactivation from the state of latency. The green fluorescence of the cells can be tracked by flow cytometry. J-Lat cells were seeded at initial density of 1 × 106 cells per well in a 24-well format in 500 μL of complete RPMI media. Compounds were added at desired concentrations in 500 μL media. Identical dimethyl sulfoxide (DMSO) levels were used for untreated control. 24 h later cells were harvested and stained with Life/Dead near-IR (Invitrogen, Denmark) and then fixed in 1% paraformaldehyde solution. Expression of GFP and cell viability were analyzed by flow cytometry on BD FACSVerse. Data was analyzed using FlowJo Version 10.0.7 (TreeStar, USA).
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5sc03257a |
‡ Authors contributed equally. |
This journal is © The Royal Society of Chemistry 2016 |