Bioresponsive small molecule polyamines as noncytotoxic alternative to polyethylenimine

articles

Bioresponsive Small Molecule Polyamines as

Noncytotoxic Alternative to Polyethylenimine

Christopher R. Drake,^†,‡,§Abderrahim Aissaoui,^†Orestis Argyros,^†

James M. Serginson,^†,‡Bryn D. Monnery,^†,‡Maya Thanou,^†,|

Joachim H. G. Steinke,*^,‡and Andrew D. Miller*^,†,⊥

Imperial College Genetic Therapies Centre, Department of Chemistry, Imperial College

London, Flowers Building, Armstrong Road, London SW7 2AZ, U.K., Department of

Chemistry, Imperial College London, South Kensington Campus, London, SW7 2AZ, U.K.,

and ImuThes Ltd, Flowers Building, Armstrong Road, London, SW7 2AZ, U.K.

Received September 4, 2009; Revised Manuscript Received August 18, 2010; Accepted

August 30, 2010

Abstract: Nonviral gene therapy continues to require novel synthetic vectors to deliver

therapeutic nucleic acids effectively and safely. The majority of synthetic nonviral vectors

employed in clinical trials to date have been cationic liposomes; however, cationic polymers

are attracting increasing attention. One of the few cationic polymers to enter clinical trials has

been polyethylenimine (PEI); however, doubts remain over its cytotoxicity, and in addition it

displays lower levels of transfection than viral systems. Herein, we report on the development

of a series of small molecule analogues of PEI that are bioresponsive to the presence of pDNA,

forming poly(disulﬁde)s that are capable of efﬁcacious transfection with no associated toxicity.

The most effective small molecule developed, a cyclic disulﬁde based upon a spermine backbone,

is shown to form very well-deﬁned polyplexes (100-200 nm in diameter) that mediate murine

lung transfection in vivo to within an order of magnitude of in vivo jetPEI, and at the same time

display a much improved cytotoxicity proﬁle.

Keywords: Nonviral; gene therapy; reducible; cyclic disulﬁde; polyamine synthesis

clinical trials^1,2with viral systems. Polycations,^3-5including

Introduction

branched and linear poly(ethylenimine) (PEI), poly(L-lysine)

(PL) and poly(amido amine) dendrimers, have attracted

widespread attention within the ﬁeld due to the stable

complexes (termed polyplexes) they form with plasmid DNA

(pDNA) which have proved able to transfect cells effectively

both in Vitro and in ViVo. Linear PEI, currently widely

accepted as the commercial “gold standard”, has progressed

Synthetic gene therapy vectors aim to match or surpass

the ability of viruses to deliver therapeutic DNA without the

associated health hazards that have been highlighted during

* Corresponding author. Mailing address: Department of Chem-

istry, Imperial College London, South Kensington Campus,

London, SW7 2AZ, U.K. Tel: +44 (0)2075945852. Fax: +44

(0)2075945804. E-mail: j.steinke@imperial.ac.uk.

^†Imperial College Genetic Therapies Centre, Department of

Chemistry, Imperial College London.

(1) Check, E. A Tragic Setback. Nature 2002, 420 (6912), 116–118.

(2) Couzin, J.; Kaiser, J. As Gelsinger Case Ends, Gene Therapy

Suffers Another Blow. Science 2005, 307 (5712), 1028.

(3) Jeong, J. H.; Kim, S. W.; Park, T. G. Molecular Design of

Functional Polymers for Gene Therapy. Prog. Polym. Sci. 2007,

32, 1239–1274.

(4) Lungwitz, U.; Breunig, M.; Blunk, T.; Go¨pferich, A. Polyethyl-

enimine-based Non-viral Gene Delivery Systems. Eur. J. Pharm.

Biopharm. 2005, 60, 247–266.

^‡Department of Chemistry, Imperial College London, South

Kensington Campus.

^§Present address: Department of Radiology, University of Cali-

fornia, San Francisco, 185 Berry St., Suite 350, San Francisco,

CA 94107.

^|Present address: King’s College London, Pharmacy Department,

Pharmaceutical Science Division, Franklin-Wilkins Building,

150 Stamford St., London, SE1 9NH, U.K.

(5) Park, T. G.; Jeong, J. H.; Kim, S. W. Current Status of Polymeric

Gene Delivery Systems. AdV. Drug DeliVery ReV. 2006, 58, 467–

486.

^⊥ImuThes Ltd.

2040 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6, 2040–2055

10.1021/mp9002249  2010 American Chemical Society

Published on Web 10/07/2010

Noncytotoxic AlternatiVe to Polyethylenimine

articles

into phase I clinical trials;⁶however, doubts over its toxicity

combined with its lower efﬁcacy compared to viruses⁷may

limit its future use.

polycations due to their subsequent breakdown into smaller,

less toxic, units. In addition the destabilization of polyplexes

that results from the degradation of the polymeric vector

should also encourage pDNA release which, if it occurs in

close enough proximity to a cell’s nucleus, might also raise

levels of transfection.

A number of different biodegradable bonds have been

incorporated into polycationic vectors. Broadly speaking they

can be grouped into three categories: hydrolyzable,¹⁷

acid-triggerable^18-21and reducible.^22-48Reducible bonds

are particularly advantageous as they will be stable within

the body until they reach the cytoplasm of a cell, where the

The root of the toxicity of PEI is the interactions this

polycation establishes with cellular membranes, disrupting

them and leading to severe damage to the cell.⁸There is

considerable evidence that the extent of this damage is related

to the molecular weight of PEI, with higher molecular weight

species causing higher levels of toxicity, presumably due to

the stronger nature of the interactions they form.^9-12

However, in order to affect efﬁcacious pDNA delivery, PEI

must also be of at least an intermediate molecular weight (2

kDa is generally accepted as the lower limit for this

parameter), otherwise they are unable to form polyplexes of

sufﬁcient stability to protect the pDNA from premature

degradation by nucleases.^9,13-16This conﬂict of interests with

respect to molecular weight has led to a plethora of

biodegradable polycations, able to condense pDNA ef-

fectively while also improving upon the toxicity of stable

(17) Luten, J.; van Nostrum, C. F.; De Smedt, S. C.; Henninck, W. E.

Biodegradable Polymers as Non-viral Carriers for Plasmid DNA

Delivery. J. Controlled Release 2008, 126, 97–110.

(18) Kim, Y. H.; Park, J. H.; Lee, M.; Kim, Y.-H.; Park, T. G.; Kim,

S. W. Polyethylenimine with Acid-Labile Linkages as a Biode-

gradable Gene Carrier. J. Controlled Release 2005, 103, 209–

219.

(19) Shim, M. S.; Kwon, Y. J. Controlled Delivery of Plasmid DNA

and siRNA to Intracellular Targets Using Ketalized Polyethyl-

enimine. Biomacromolecules 2008, 9 (2), 444–455.

(20) Knorr, V.; Russ, V.; Allmendinger, L.; Ogris, M.; Wagner, E.

Acetal Linked Oligoethylenimines for Use as pH-Sensitive Gene

Carriers. Bioconjugate Chem. 2008, 19 (8), 1625–1634.

(21) Ko, I. K.; Ziady, A.; Lu, S.; Kwon, J. K. Acid-degradable Cationic

Methacrylamide Polymerized in the Presence of Plasmid DNA

as Tunable Non-viral Gene Carrier. Biomaterials 2008, 28 (28),

3872–3881.

(22) Breunig, M.; Lungwitz, U.; Liebl, R.; Goepferich, A. Breaking

up the Correlation Between Efﬁcacy and Toxicity for Nonviral

Gene Delivery. Proc. Natl. Acad. Sci. U.S.A. 2007, 104 (36),

14454–14459.

(23) Carlisle, R. C.; Etrych, T.; Briggs, S. S.; Preece, J. A.; Ulbrich,

K.; Seymour, L. W. Polymer-Coated Polyethylenimine/DNA

Complexes Designed for Triggered Activation by Intracelular

Reduction. J.Gene Med. 2004, 6, 337–344.

(24) Gosselin, M. A.; Guo, W.; Lee, R. J. Efﬁcient Gene Transfer Using

Reversibly Cross-Linked Low Molecular Weight Polyethylen-

imine. Bioconjugate Chem. 2001, 12 (6), 989–994.

(25) Gosselin, M. A.; Guo, W.; Lee, R. J. Incorporation of Reversibly

Cross-Linked Polyplexes into LPDII Vectors for Gene Delivery.

Bioconjugate Chem. 2002, 13 (5), 1044–1053.

(26) Kwok, K. Y.; Park, Y.; Yang, Y.; McKenzie, D. L.; Liu, Y.; Rice,

K. G. In ViVo Gene Transfer Using Sulfhydyl Cross-Linked PEG-

Peptide/Glycopeptide DNA Co-Condensates. J. Pharm. Sci. 2003,

92 (6), 1174–1185.

(27) Manickam, D. S.; Oupicky, D. Multiblock Reducible Copolypep-

tides Containing Histidine-Rich and Nuclear Localization Se-

quences for Gene Delivery. Bioconjugate Chem. 2006, 17 (6),

1395–1403.

(28) McKenzie, D. L.; Kwok, K. Y.; Rice, K. G. A Potent New Class

of Reductively Active Peptide Gene Delivery Agents. J. Biol.

Chem. 2000, 275 (14), 9970–9977.

(29) McKenzie, D. L.; Smiley, E.; Kwok, K. Y.; Rice, K. G. Low

Molecular Weight Disulﬁde Cross-Linking Peptides as Nonviral

Gene Delivery Carriers. Bioconjugate Chem. 2000, 11 (6), 901–

909.

(6) www.polyplus-transfection.com (accessed 25/08/08).

(7) Varga, C. M.; Teford, N. C.; Thomas, M.; Klibanov, A. M.;

Grifﬁth, L. G.; Lauffenburger, D. A. Quantitative Comparison of

Polyethylenimine Formulations and Adenoviral Vectors in Terms

of Intracellular Gene Delivery Processes. Gene Ther. 2005, 12,

1023–1032.

(8) Moghimi, S. M.; Symonds, P.; Murray, J. C.; Hunter, A. C.;

Debska, G.; Szewczyk, A. A Two Stage Poly(ethylenimine)-

Mediated Cytotoxicity: Implications for Gene Transfer/Therapy.

Mol. Ther. 2005, 11 (6), 990–995.

(9) Breunig, M.; Lungwitz, U.; Liebl, R.; Fontanari, C.; Klar, J.; Kurtz,

A.; Blunk, T.; Goepferich, A. Gene Delivery with Low Molecular

Weight Polyethylenimines. J. Gene Med. 2005, 7, 1287–1298.

(10) Fischer, D.; Bieber, T.; Li, Y.; Elsasser, H.-P.; Kissel, T. A Novel

Non-Viral Vector for DNA Delivery Based on Low Molecular

Weight, Branched Polyethylenimine: Effect of Molecular Weight

on Transfection Efﬁciency and Cytotoxicity. Pharm. Res. 1999,

16 (8), 1273–1279.

(11) Jeong, J. H.; Song, S. H.; Lim, D. W.; Lee, H.; Park, T. G. DNA

Transfection Using Linear Poly(ethylenimine) Prepared by Con-

trolled Acid Hydrolysis of Poly(2-ethyl-2-oxazoline). J. Controlled

Release 2001, 73, 391–399.

(12) Kunath, K.; von Harpe, A.; Fischer, D.; Petersen, H.; Bickel, U.;

Voigt, K.; Kissel, T. Low-Molecular-Weight Polyethylenimine

as a Non-Viral Vector for DNA Delivery: Comparison of

Physiochemical Properties, Transfection Efﬁciency and In ViVo

Distribution with High-Molecular-Weight Polyethylenimine. J.

Controlled Release 2003, 89, 113–125.

(13) Breunig, M.; Lungwitz, U.; Liebl, R.; Klar, J.; Obermayer, B.;

Blunk, T.; Goepferich, A. Mechanistic Insights into Linear

Polyethylenimine-Mediated Gene Transfer. Biochim. Biophys.

Acta 2007, 1770, 196–205.

(14) Choosakoonkriang, S.; Lobo, B. A.; Koe, G. S.; Koe, J. G.;

Middaugh, C. R. Biophysical Characterisation of PEI/DNA

Complexes. J. Pharm. Sci. 2003, 92 (8), 1710–1722.

(15) Forrest, M. L.; Koerber, J. T.; Pack, D. W. A Degradable

Polyethylenimine Derivative with Low Toxicity for Highly

Efﬁcient Gene Delivery. Bioconjugate Chem. 2003, 14 (5), 934–

940.

(16) Godbey, W. T.; Wu, K. K.; Mikos, A. G. Size Matters: Molecular

Weight Affects the Efﬁciency of Poly(ethylenimine) as a Gene

Delivery Vector. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5177–

5181.

(30) Miyata, K.; Kakizawa, Y.; Nishiyama, N.; Harada, A.; Yamasaki,

Y.; Koyama, H.; Kataoka, K. Block Catiomer Polyplexes with

Regulated Densities of Charge and Disulﬁde Cross-Linking

Directed to Enhance Gene Expression. J. Am. Chem. Soc. 2004,

126 (8), 2355–2361.

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2041

articles

Drake et al.

high concentrations of glutathione maintained there will

cleave them.^49,50This triggered degradation mechanism

should ensure that the pDNA is not released until it has

reached the vicinity of its site of action. Strategies involving

disulﬁde bonds realized to date include the reversible cross-

linking of low molecular weight PEI,^22,24,35the reversible

lateral stabilization of polyplexes,^23,25,30,32linking amine

containing pendant side chains to the polymer backbone via

disulﬁde bonds,^36,39bioreducible polypeptides,^{26-29,33,34,37,38}

bioreducible rotaxanes³¹and the synthesis of polyamines with

disulﬁde linkages within their backbone,^40-48including the

recently published work of Park et al. on the synthesis of

linear poly(ethylenimine sulﬁde)s.⁴²

(31) Ooya, T.; Choi, H. S.; Yamashita, A.; Yui, N.; Sugaya, Y.; Kano,

A.; Maruyama, A.; Akita, H.; Ito, R.; Kogure, K.; Harashima, H.

Biocleavable Polyrotaxane-Plasmid DNA Polyplex for Enhanced

Gene Delivery. J. Am. Chem. Soc. 2006, 128 (12), 3852–3853.

(32) Oupicky, D.; Carlisle, R. C.; Seymour, L. W. Triggered Intra-

cellular Activation of Disulﬁde Crosslinked Polyelectrolyte Gene

Delivery Complexes with Extended Systemic Circulation In ViVo.

Gene Ther. 2001, 8, 713–724.

(33) Oupicky, D.; Parker, A. L.; Seymour, L. W. Laterally Stabilised

Complexes of DNA with Linear Reducible Polycations: Strategy

for Triggered Intracellular Activation of DNA Delivery Vectors.

J. Am. Chem. Soc. 2002, 124 (1), 8–9.

(34) Parker, A. L.; Eckley, L.; Singh, S.; Preece, J. A.; Collins, L.;

Fabre, J. W. (LYS)₁₆-Based Reducible Polycations Provide Stable

Polyplexes with Anionic Fusogenic Peptides and Efﬁcient Gene

Delivery to Post-Mitotic Cells. Biochim. Biophys. Acta 2007, 1770,

1331–1337.

(35) Peng, Q.; Zhong, Z.; Zhuo, R. Disulﬁde Cross-Linked Polyeth-

ylenimines (PEI) Prepared via Thiolation of Low Molecular

Weight PEI as Highly Efﬁcient Gene Vectors. Bioconjugate Chem.

2008, 19 (2), 499–506.

We were interested in designing and developing a biore-

ducible analogue of PEI that would retain or improve on its

efﬁcacy while remaining nontoxic. In order to achieve this

we sought to incorporate regular disulﬁde linkages within a

linear polyamine backbone. As described above this part of

our design strategy was with signiﬁcant precedent, therefore

we decided to seek further innovation by constructing small

molecule polyamines that were also either dithiols or cyclic

disulﬁdes in character. The molecular weight of any thera-

peutic polymer affects its biological and pharmacological

proﬁle, and even relatively small changes can have dramatic

effects in a clinical setting.^51,52Indeed, the molecular weight

of PEI has been shown to affect both efﬁcacy of pDNA

delivery and cytotoxicity.¹⁶However, in spite of its clear

importance, and the development of living polymerizations

that allow for a degree of control to be exerted over it, this

crucial variable often remains under weak synthetic control.⁵¹

In contrast small molecules such as dithiols and cyclic

disulﬁdes have a clearly deﬁned molecular weight which is

under strict synthetic control, allowing for consistency of

production across any number of different batches. How

would such small molecules be able to condense pDNA into

the stable polyplexes required for gene delivery? In both

(36) Pichon, C.; LeCam, E.; Gue´rin, B.; Coulaud, D.; Delain, E.;

Midoux, P. Poly[Lys-(AEDTP)]: A Cationic Polymer that Allows

Dissociation of pDNA/Cationic Polymer Complexes in a Reduc-

tive Medium and Enhances Polyfection. Bioconjugate Chem. 2002,

13 (1), 76–82.

(37) Read, M. L.; Singh, S.; Ahmed, Z.; Stevenson, M.; Briggs, S. S.;

Oupicky, D.; Barrett, L. B.; Spice, R.; Kendall, M.; Berry, M.;

Preece, J. A.; Logan, A.; Seymour, L. W. A Versitile Reducible

Polycation-Based System for Efﬁcient Delivery of a Broad Range

of Nucleic Acids. Nucleic Acids Res. 2005, 33 (9), e86.

(38) Trentin, D.; Hall, H.; Wechsler, S.; Hubbell, J. A. Peptide-Matrix-

Mediated Gene Transfer of an Oxygen Insensitive Hypoxia-

Inducible Factor-1R Variant for Local Induction of Angiogenesis.

Proc. Natl. Acad. Sci. U.S.A. 2006, 103 (8), 2506–2511.

(39) Zugates, G. T.; Anderson, D. G.; Little, S. R.; Lawhorn, I. E. B.;

Langer, R. Synthesis of Poly(ꢀ-amino ester)s with Thiol-Reactive

Side Chains for DNA Delivery. J. Am. Chem. Soc. 2006, 128

(39), 12726–12734.

(40) Christensen, L. V.; Chang, C.-W.; Kim, W. J.; Kim, S. W.; Zhong,

Z.; Lin, C.; Engbersen, J. F. J.; Feijen, J. Reducible Poly(amido

ethylenimine)s Designed for Triggered Intracellular Gene Deliv-

ery. Bioconjugate Chem. 2006, 17 (5), 1233–1240.

(41) Kloeckner, J.; Wagner, E.; Ogris, M. Degradable Gene Carriers

Based on Oligomerised Polyamines. J. Pharm. Sci. 2006, 29, 414–

425.

(45) Lin, C.; Zhong, Z.; Lok, M. C.; Jiang, X.; Henninck, W. E.; Feijen,

J.; Engbersen, J. F. J. Random and Block Copolymers of

Poly(amido amine)s with High- and Low-Basicity Amino Groups:

Study of DNA Condensation and Buffer Capacity on Gene

Transfection. J. Controlled Release 2007, 123 (67), 75.

(46) Lin, C.; Blaauboer, C.-J.; Timoneda, M. M.; Lok, M. C.; van

Steenbergen, M.; Hennink, W. E.; Zhong, Z.; Feijen, J.; Engber-

sen, J. F. J. Bioreducible Poly(amido amine)s with Oligoamine

Side Chains: Synthesis, Characterisation, and Structural Effects

on Gene Delivery. J. Controlled Release 2008, 126, 166–174.

(47) Ou, M.; Wang, X.-L.; Xu, R.; Chang, C.-W.; Bull, D. A.; Kim,

S. W. Novel Biodegradable Poly(disulﬁde amine)s for Gene

Delivery with High Efﬁciency and Low Cytotoxicity. Bioconju-

gate Chem. 2008, 19 (3), 626–633.

(48) Wang, X.-L.; Jensen, R.; Lu, Z.-R. A Novel Environment Sensitive

Biodegradable Polydisulﬁde with Protonatable Pendants for

Nucleic Acid Delivery. J. Controlled Release 2007, 120, 250–

258.

(49) Meister, A.; Anderson, M. E. Glutathione. Annu. ReV. Biochem.

1983, 52, 711–760.

(42) Lee, Y.; Mo, H.; Koo, H.; Park, J.-Y.; Cho, M. Y.; Jin, G.-W.;

Park, J.-S. Visualisation of the Degradation of a Disulﬁde Polymer,

Linear Poly(ethyleneimine sulﬁde), for Gene Delivery. Biocon-

jugate Chem. 2007, 18 (1), 13–18.

(43) Lin, C.; Zhong, Z.; Lok, M. C.; Jiang, X.; Henninck, W. E.; Feijen,

J.; Engbersen, J. F. J. Linear Poly(amido amine)s with Secondary

and Tertiary Amino Groups and Varying Amounts of Disulﬁde

Linkages: Synthesis and In Vitro Gene Transfer Properties. J.

Controlled Release 2006, 116 (130), 137-.

(44) Lin, C.; Zhong, Z.; Lok, M. C.; Jiang, X.; Henninck, W. E.; Feijen,

J.; Engbersen, J. F. J. Novel Reducible Poly(amido amine)s for

Highly Efﬁcient Gene Delivery. Bioconjugate Chem. 2007, 18

(1), 138–145.

(50) Saito, G.; Swanson, J. A.; Lee, K.-D. Drug Delivery Strategy

Utilising Conjugation via Reversible Disulﬁde Linkages: Role and

Site of Cellular Reducing Activities. AdV. Drug DeliVery ReV.

2003, 55 (2), 199–215.

(51) Ali, M.; Brocchini, S. Synthetic Approaches to Uniform Polymers.

AdV. Drug DeliVery ReV. 2006, 58, 1671–1687.

(52) Hunter, A. C.; Moghimi, S. M. Therapeutic Synthetic Polymers:

a Game of Russian Roulette. Drug DiscoVery Today 2002, 7 (19),

998–1001.

2042 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

Materials and Methods

General Considerations. Dried CH₂Cl₂, MeOH and THF

were obtained from a pure solv-MD standard design solvent

puriﬁcation system (Innovative Technologies, U.K.) using

columns ﬁlled with 3 mm aluminum oxide balls (Sigma-

Aldrich, U.K.). Dry TFA was obtained by distillation over

CaH₂at ambient pressure. All organic solvents used were

dry, and reactions were performed under nitrogen unless

stated. TLC was performed on Merck Kieselgel 60 F254

aluminum backed plates and visualized under UV light or

stained with KMnO₄solution. Puriﬁcation over silica was

achieved using Merck Kieselgel 60 (230-400 mesh). ¹H and

¹³C NMR were recorded on a Bruker DRX400 spectrometer

and processed using MestRe-C software (version 4.7.0.0,

available from www.mestrec.com) using residual isotopic

solvent as an internal reference for organic deuterated

solvents. Samples run in D₂O were referenced using an

external solution of dioxane in D₂O. The following abbrevia-

tions were used during assignment: s ) singlet, d ) doublet,

t ) triplet, q ) quartet, quint ) quintet, m ) multiplet, br

) broad. Chemical ionization (CI), fast atom bombardment

(FAB) and electronic ionization (EI) mass spectra were

recorded by Mr. John Barton using VG-70-250E and VG

autospec Q instruments. Elemental analysis was performed

by Stephen Boyer at London Metropolitan University.

Melting points were measured on a Sanyo Gallenkamp hot

stage. All chemicals were purchased from Sigma Aldrich,

Lancaster or VWR unless otherwise stated and were used

as received. 2-[(4-Methoxybenzyl)sulfanyl]ethanol⁵⁵(2),

2-[(4-methoxybenzyl)sulfanyl]acetaldehyde⁵⁵(3) and di-tert-

butyl ethane-1,2-diylbis[(2-hydroxyethyl)carbamate]⁵⁶(11)

were synthesized according to published procedures. Deacety-

lated linear PEI was synthesized according to the procedure

of Thomas et al.⁵⁷

Figure 1. Schematic representation of the mechanisms

of condensation for small molecule polyamine cyclic

disulﬁdes.

cases we envisaged that they would respond to the introduc-

tion of pDNA by forming high molecular weight poly(dis-

ulﬁde)s. In the case of dithiols, we anticipated that they might

associate with pDNA through electrostatic interactions with

phosphodiester backbone charges and then undergoing a

sequence of controlled oxidations, utilizing the pDNA as a

kinetic template (kinetically templated polymerization, KCP).

This use of pDNA as a kinetic template has been described

previously within the ﬁeld of gene therapy, successfully

resulting in the formation of polyplexes after polyconden-

sations and vinylic polymerizations.^21,53In the case of the

cyclic disulﬁdes (Figure 1), we anticipated that they should

also interact electrostatically with pDNA and then undergo

a sequence or series of thiol-disulﬁde exchange reactions to

form a poly(disulﬁde) in a process that we hypothesized

would be driven by entropic considerations (thermodynami-

cally controlled templated polymerization, TCTP). This

dynamic exchange should be catalyzed by the presence of

small concentrations of thiols and has been widely used to

great effect in dynamic combinatorial chemistry under

biologically relevant conditions (i.e., pH 7.4).⁵⁴From the

outset, we appreciated that our combined approach would

not solve the polymer molecular weight problem directly per

se, in that the ﬁnal condensing species were still likely to be

polydisperse, but would remove any doubts about the

inﬂuence of variations in the synthetic conditions employed

on the biological properties of the vector in question.

pEGFP-Luc-DNA (1 mg/mL) was supplied by BD Bio-

sciences, U.K. pUMV-C1-nt-beta gal was obtained from the

University of Michigan Vector Core and was ampliﬁed by

Plasmid Factory, Germany. Luciferase gene expression was

measured with the Luciferase Assay System purchased from

Promega, USA. Light output readings were done on a

Berthold Lumal LB 9507 luminometer. Transfection ef-

ﬁciency, measured in relative light units (RLU), was normal-

ized to the total mass of cellular protein using the BCA

protein assay kit (Thermo Scientiﬁc, USA). LDH assay was

Here we report the synthesis of small molecule polyamine

dithiols and cyclic disulﬁdes based upon a variety of di-,

tri- and tetraamine starting materials. These novel compounds

were then tested for their ability to complex and cage pDNA

into condensed nanoparticles in a manner suitable for gene

delivery, in Vitro and in ViVo. Studies involving efﬁcacy of

pDNA delivery and cytotoxicity were measured and studied

throughout.

(55) Gordon, E. M.; Godfrey, J. D.; Delaney, N. G.; Asaad, M. M.;

Von Langen, D.; Cushman, D. W. Design of Novel Inhibitors of

Aminopeptidases. Synthesis of Peptide Derived Diamino Thiols

and SUlfur Replacement Analogues of Bestatin. J. Med. Chem.

1988, 31 (11), 2199–2211.

(56) Solodin, I.; Brown, C. S.; Bruno, M. S.; Chow, C.-Y.; Jang, E.-

H.; Debs, R. J.; Heath, T. D. A Novel Series of Amphiphilic

Imidazolinium Compounds for In Vitro and In ViVo Gene

Delivery. Biochemistry 1995, 34 (41), 13537–13544.

(57) Thomas, M.; Lu, J. J.; Ge, Q.; Zhang, C.; Chen, J.; Klibanov,

A. M. Full Deacetlyation of Polyethylenimine Dramatically Boosts

its Gene Delivery Efﬁcacy and Speciﬁcity to Mouse Lung. Proc.

Natl. Acad. Sci. U.S.A. 2005, 102 (16), 5679–5684.

(53) Trubetskoy, V. S.; Budker, V. G.; Hanson, L. J.; Slattum, P. M.;

Wolff, J. A.; Hagstrom, J. E. Self-Assembly of DNA-Polymer

Complexes Using Template Polymerisation. Nucleic Acids Res.

1998, 26 (18), 4178–4185.

(54) Otto, S.; Furlan, R. L. E.; Sanders, J. K. M. Selection and

Ampliﬁcation of Hosts from Dynamic Combinatorial Libraries

of Macrocyclic Disulﬁdes. Science 2002, 297, 590–593.

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2043

articles

Drake et al.

carried out using the CytoTox 96 Non-Radioactive Cyto-

toxicity Assay (Promega, USA). PCS size measurements

were carried out on a Coulter N4 dynamic light scattering

instrument. SEM images were taken on a Leo-40 spectrom-

eter, using samples gold plated in Vacuo using standard

techniques. All media and PBS were purchased from

Invitrogen, U.K. Lysis buffer (5×) was obtained from Roche

Diagnostics, U.K. Agarose, FCS, penicillin-streptomycin,

HEPES free acid, glutaraldehyde, benzyl benzoate and benzyl

alcohol were obtained from Sigma, U.K. OCT embedding

matrix was purchased from Cellpath Ltd., and eosin was

purchased from VWR. CHO-K1 and A549 cells were from

the American Type Culture Collection and were incubated

at 37 °C in 10% CO₂. F12 was used as the base medium for

CHO-K1 cells and DMEM was used for A549 cells. 10%

fetal calf serum and 1% penicillin-streptomycin were added

to the base media to give serum containing media. Female

BALB/c mice were purchased from Charles River Ltd., U.K.,

and housed under identical conditions. Murine experiments

were performed in accordance with the UK Home Ofﬁce

regulations for the use of animals in research. Lung samples

were homogenized using a Ultra-Turrax T25 homogenizer

(Fisher Bioblock Scientiﬁc, U.K.), and ꢀ-gal expression was

measured using a ꢀ-gal ELISA assay (Roche Diagnostics,

U.K.). Histological images were taken using a Nikon Eclipse

E600 microscope. Statistical analysis on the in Vitro data

was performed using Welch’s t test (2-tailed) at a 0.05

signiﬁcance level. In ViVo data was analyzed by Imperial

College Statistical Advisory Service using a Bonferroni

corrected ANOVA test at a 0.05 signiﬁcance level.

Synthetic Chemistry. General Procedure A: Bisalkyla-

tion of Small Molecule Polyamines with {[(4-Methoxyben-

zyl)sulfanyl]ethyl} Moieties. The small molecule polyamine

(1 equiv) and [(4-methoxybenzyl)sulfanyl]acetaldehyde (3)

(2 equiv) were dissolved in dry DCM (100 mL). MgSO₄

(∼1 g) was added, and the resulting suspension was stirred

for 6 h. The suspension was ﬁltered, and dry MeOH (100

mL) was added, followed by NaBH₄(10 equiv), and the

resulting suspension was stirred for 15 h. Concentrated HCl

(20 mL) was added dropwise (effervescence observed), and

the resulting suspension was reﬂuxed for 6 h. The suspension

was concentrated under reduced pressure, and the resulting

slurry was ﬁltered and then washed with EtOAc (100 mL)

to give a white solid, which was recrystallized from H₂O to

give the desired product as a white solid.

(C₂₇H₄₃N₃O₂S₂· H⁺) 506.2875; found 506.2897. Elemental

anal.: calculated (C₂₇H₄₃N₃O₂S₂·3PF₆) C 52.72% H 7.54%

-

N 6.83%; found C 52.77% H 7.54% N 6.78%.

General Procedure B: RemoVal of 4-Methoxybenzyl

Protecting Groups To ReVeal Dithiol. Dry anisole (10 equiv)

was dissolved in dry TFA (20 mL), and the resulting solution

was degassed using the freeze-pump-thaw method (3

cycles). The bis[(4-methoxybenzyl)sulfanyl] (1 equiv) was

added, and the resulting solution was heated at reﬂux for 3

days. The solution was then concentrated under reduced

pressure, and the residue was washed with dry degassed Et₂O

(50 mL). Degassed MeOH (20 mL) was added to the solid,

and the resulting suspension was ﬁltered. The solvent was

then removed under reduced pressure and the solid redis-

solved into a minimum amount of degassed MeOH. De-

gassed Et₂O was added until the cloud point was reached,

allowing the product to be puriﬁed via recrystallization to

give the desired dithiol as a white powder.

3,7,12-Triazatetradecane-1,14-dithiol (Tris-triﬂuoroacetate

Salt) (TriN-3,4-DT). 1,14-Bis[(4-methoxybenzyl)sulfanyl]-

3,7,12-triazatetradecane (5) (400 mg, 0.650 mmol) was

reacted according to general procedure B to give TriN-3,4-

DT as a white powder (174 mg, 0.287 mmol, 44%). Mp >

1

220 °C. H NMR (400 MHz, d₄-MeOH): δ 1.62 (br, 4H,

NHCH₂CH₂CH₂CH₂NH), 1.94 (quint, 2H, NHCH₂CH₂-

CH₂NH), 2.80 (m, 4H, 3.10, CH₂SH) 3.00-3.25 (m, 12H,

NHCH₂). ¹³C NMR (400 MHz d₄-MeOH): δ 20.9, 20.9, 24.0,

24.1, 24.2, 45.6, 45.8, 47.9, 48.2, 51.4, 51.5. HR-MS (m/z,

+ve FAB): calculated (C₁₁H₂₇N₃S₂· H⁺) 266.1725; found

266.1730. Elemental anal.: calculated (C₁₁H₂₇N₃S₂·

3C₂HF₃O₂) C 33.61% H 4.98% N 6.92%; found C 33.74%

H 4.90% N 6.87%.

General Procedure C: Oxidation of Dithiols to Cyclic

Disulﬁdes. The dithiol was dissolved in concentrated HCl

(10 mL). DMSO (0.1 mL) was added, and the resulting

solution was stirred for 6 h, during which time a white

precipitate developed. The solution was then concentrated

under reduced pressure, and the product was precipitated with

an excess of acetone (25 mL). The solid was ﬁltered, washed

with further portions of acetone (2 × 25 mL), dried under

reduced pressure at 70 °C for 12 h and then redissolved in

H₂O and lyopholized to give the desired cyclic disulﬁde as

a white powder.

1,2-Dithia-5,9,14-triazacyclohexadecane (Tris-hydrochlo-

ride Salt) (TriN-3,4-CDS). 3,7,12-Triazatetradecane-1,14-

dithiol (TriN-3,4-DT) (98 mg, 0.16 mmol) was reacted

according to general procedure C to give TriN-3,4-CDS as

a white powder (51 mg, 0.057 mmol, 87%). Mp > 220 °C.

¹H NMR (D₂O, 400 MHz): δ 1.83 (br, 4H, NHCH₂CH₂-

CH₂CH₂NH), 2.19 (br quint, 2H, NHCH₂CH₂CH₂NH),

3.10-3.30 (m, 12H, NHCH₂CH₂CH₂CH₂NH + NHCH₂-

CH₂CH₂NH + CH₂S), 3.50 (m, 4H, NHCH₂CH₂S). ¹³C NMR

(D₂O, 100 MHz): δ 22.5, 22.5, 22.7, 31.7, 32.5, 44.3, 44.4,

45.6, 45.8, 46.8, 46.9. ES-MS (m/z): 263.6 (100%), 526.9

(24%). HR-MS (m/z, ES): calculated (C₁₁H₂₅N₃S₂· H⁺)

1,14-Bis[(4-methoxybenzyl)sulfanyl]-3,7,12-triazatetrade-

cane (Tris-hydrochloride Salt) (5). Spermidine (0.82 mL, 5.2

mmol) was reacted according to general procedure A to give

5 as a white solid (1.63 g, 2.65 mmol, 53%). Mp > 220 °C.

¹H NMR (D₂O, 400 MHz): δ 1.76 (br, 4H, NHCH₂CH₂-

CH₂CH₂NH), 2.12 (br quint, 2H, NHCH₂CH₂CH₂NH), 2.82

(t, J 7.0 Hz, 4H, NHCH₂CH₂S), 3.00-3.30 (m, 12H,

NHCH₂), 3.82 (s, 6H, CH₃O), 3.88 (s, 4H, SCH₂Ar), 7.05

(d, 4H, J 8.5 Hz, ArH), 7.40 (d, 4H, J 8.5 Hz, ArH). ¹³C

NMR (D₂O, 100 MHz): δ ) 22.4, 22.6, 22.8, 26.1, 26.2,

34.1, 43.9, 44.0, 45.8, 45.9, 46.0, 46.2, 55.0, 113.9, 129.6,

130.0, 158.3. HR-MS (m/z, +ve FAB): calculated

2044 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

264.1568; found 264.1577. Elemental anal.: calculated

(C₁₁H₂₅N₃S₂· 3HCl) C 35.43% H 7.57% N 11.27%; found

C 35.56% H 7.54% N 11.15%.

130.8, 131.0, 159.0. HR-MS (m/z, ES⁺): calculated

(C₂₇H₄₄N₄O₂S₂· H⁺) 521.2984; found 521.2981. Elemental

anal.: calculated (C₂₇H₄₄N₄O₂S₂· 4HCl) C 48.65% H 7.26%

N 8.40%; found C 48.51% H 7.16% N 8.49%.

3,6,10,13-Tetraazapentadecane-1,15-dithiol (Tetra-triﬂuo-

roacetate Salt) (TetraN-2,3,2-DT). 1,15-Bis[(4-methoxyben-

zyl)sulfanyl]-3,6,10,13-tetraazapentadecane (13) (300 mg,

0.450 mmol) was reacted according to general procedure B

to give TetraN-2,3,2-DT as a white powder (77 mg, 0.10

mmol, 23%). Mp > 220 °C. ¹H NMR (400 MHz, d₄-MeOH):

δ 2.21 (br quint, 2H, NHCH₂CH₂CH₂), 2.84 (t, J 7.0 Hz,

1,12-Bis[(4-methoxybenzyl)sulfanyl]-3,6,10-triazadode-

cane (Tris-hydrochloride Salt) (12). 1,6-Diamino-3-azahex-

ane (0.32 mL, 2.6 mmol) was reacted according to general

procedure A to give 12 as a white solid (703 mg, 1.28 mmol,

49%). Mp > 220 °C. ¹H NMR (D₂O, 400 MHz): δ 2.13 (br

quint, 2H, NHCH₂CH₂CH₂NH), 2.81 (br t, 4H,

NHCH₂CH₂S), 3.00-3.50 (m, 8H, NHCH₂CH₂CH₂NH +

NHCH₂CH₂S), 3.46 (s, 4H, NHCH₂CH₂NH), 3.70-3.85 (m,

10H, SCH₂Ar + CH₃O), 6.98 (br, 4H, ArH), 7.33 (br, 4H,

ArH³). ¹³C NMR (D₂O, 100 MHz): δ 23.0, 27.1, 27.2, 35.4,

43.6, 43.8, 44.9, 45.7, 47.3, 47.8, 56.4, 115.2, 130.8, 130.8,

159.0. HR-MS (m/z, +ve FAB): calculated (C₂₅H₃₉N₃O₂S₂·

H⁺) 478.2562; found 478.2576. Elemental anal.: calculated

(C₂₅H₃₉N₃O₂S₂· 3HCl) C 49.41% H 6.85% N 7.52%; found

C 49.47% H 6.94% N 7.43%.

CH₂SH), 3.10-3.40 (m, 8H, NHCH₂CH₂CH₂NH

+

NHCH₂CH₂SH), 3.48 (s, 8H, NHCH₂CH₂NH). ¹³C NMR

(100 MHz, d₄-MeOH): δ 21.0, 24.4, 44.7, 44.9, 46.2, 51.9.

HR-MS (m/z, ES): calculated (C₁₁H₂₈N₄S₂· H⁺) 281.1834;

found 281.1686. Elemental anal.: calculated (C₁₁H₂₈N₄S₂·

4CF₃CO₂H) C 30.98% H 4.38% N 7.61%; found C 30.90%

H 4.33% N 7.56%.

3,6,10-Triazadodecane-1,12-dithiol (Tris-triﬂuoroacetate

Salt) (TriN-2,3-DT). 1,12-Bis[(4-methoxybenzyl)sulfanyl]-

3,6,10-triazadodecane (12) (550 mg, 0.984 mmol) was

reacted according to general procedure B to give TriN-2,3-

DT as a white powder (312 mg, 0.538 mmol, 55%). Mp >

220 °C. ¹H NMR (400 MHz, d₄-MeOH): δ 2.14 (quint, 2H,

NHCH₂CH₂CH₂NH), 2.83 (m, 4H, CH₂SH) 3.10-3.35 (m,

8H, NHCH₂CH₂CH₂NH, NHCH₂CH₂SH), 3.47 (s, 4H,

NHCH₂CH₂NH). ¹³C NMR (400 MHz d₄-MeOH): δ 20.4,

20.4, 23.1, 43.5, 43.7, 44.7, 45.6, 50.5, 51.0. HR-MS (m/z,

+ve FAB): calculated (C₉H₂₃N₃S₂· H⁺) 238.1412; found

238.1418. Elemental anal.: calculated (C₁₁H₂₇N₃S₂·

3C₂HF₃O₂) C 31.09% H 4.52% N 7.25%; found C 30.97%

H 4.60% N 7.42%.

1,2-Dithia-5,8,12-triazacyclobutadecane (Tris-hydrochlo-

ride Salt) (TriN-2,3-CDS). 3,6,10-Triazadodecane-1,12-

dithiol (TriN-2,3-DT) (150 mg, 0.259 mmol) was reacted

according to general procedure C to give TriN-2,3-CDS as

a white powder (81 mg, 0.24 mmol, 92%). Mp > 220 °C.

¹H NMR (D₂O, 400 MHz): δ 2.17 (br quint, 2H,

NHCH₂CH₂CH₂NH), 3.06 (br t, 4H, CH₂S), 3.25 (m, 4H

NHCH₂CH₂CH₂NH), 3.45-3.55 (m, 8H, NHCH₂CH₂S +

NHCH₂CH₂NH). ¹³C NMR (D₂O, 100 MHz): δ 23.3, 32.4,

43.7, 43.8, 45.1, 45.7, 46.6, 47.0. ES-MS (m/z): 236 (100%),

471 (58%). HR-MS (m/z, ES): calculated (C₉H₂₁N₃S₂· H⁺)

236.1255; found 236.1261. Elemental anal.: calculated

(C₁₁H₂₅N₃S₂· 3HCl) C 31.35% H 7.02% N 12.19%; found

C 31.29% H 6.93% N 12.08%.

1,2-Dithia-5,8,12,15-tetraazacycloheptadecane (Tetra-

hydrochloride Salt) (TetraN-2,3,2-CDS). 3,6,10,13-Tetraaza-

pentadecane-1,15-dithiol (TetraN-2,3,2-DT) (80 mg, 0.11

mmol) was reacted according to general procedure C to give

TetraN-2,3,2-CDS as a white powder (38 mg, 0.090 mmol,

81%). Mp > 220 °C. ¹H NMR (D₂O, 400 MHz): δ 2.16 (br

quint, 2H, NHCH₂CH₂CH₂NH), 3.06 (br t, 4H, CH₂S), 3.24

(m, 4H, NHCH₂CH₂CH₂NH), 3.40-3.60 (m, 8H,

NHCH₂CH₂NH + NHCH₂CH₂S). ¹³C NMR (D₂O, 100

MHz): δ 23.6, 32.5, 43.90 (broad signal), 45.6, 47.0. HR-

MS (m/z, ES): calculated (C₁₁H₂₆N₄S₂·H⁺) 279.1677; found

279.1683. Elemental anal: (C₁₁H₂₆N₄S₂· 4HCl) C 31.14% H

7.13% N 13.20%; found 31.09% H 5.45% N 9.02%.

1,18-Bis[(4-methoxybenzyl)sulfanyl]-3,7,12,16-tetraazaoc-

tadecane (Tetra-hydrochloride Salt) (14). Spermine (516 mg,

2.55 mmol) was reacted according to general procedure A

to give 14 as a white solid (944 mg, 1.27 mmol, 50%). Mp

1

> 220 °C. H NMR (D₂O, 400 MHz): δ 1.79 (br, 4H,

NHCH₂CH₂CH₂CH₂NH), 2.10 (br quint, 4H, NHCH₂-

CH₂CH₂NH), 2.79 (t, 4H, J 7.0 Hz, NHCH₂CH₂S), 3.00-3.20

(m, 16H, NHCH₂), 3.76 (s, 4H, SCH₂Ar), 3.79 (s, 6H,

CH₃O), 6.95 (d, 4H, J 8.5 Hz, ArH), 7.30 (d, 4H, J 8.5 Hz,

ArH). ¹³C NMR (D₂O, 100 MHz): δ 23.0, 23.2, 27.2, 35.3,

45.0, 45.1, 47.3, 47.6, 56.4, 115.2, 130.7, 130.8, 159.0. HR-

MS (m/z, ES): calculated (C₃₀H₅₀N₄O₂S₂· H⁺) 563.3453;

found 563.3438. Elemental anal.: calculated (C₃₀H₅₀N₄O₂S₂·

4HCl) C 48.35% H 7.44% N 7.52%; found C 48.55% H

7.61% N 7.35%.

1,15-Bis[(4-methoxybenzyl)sulfanyl]-3,6,10,13-tetraazap-

entadecane (Tetra-hydrochloride Salt) (13). 3,7-Diazanonane-

1,9-diamine (0.43 mL, 2.6 mmol) was reacted according to

general procedure A to give 13 as a white solid (475 mg,

0.714 mmol, 23%). Mp > 220 °C. ¹H NMR (D₂O, 400 MHz):

δ 2.18 (br quint, 2H, NHCH₂CH₂CH₂NH), 2.78 (t, 4H, J

6.5 Hz, NHCH₂CH₂S), 3.20-3.30 (m, 8H, NHCH₂CH₂-

CH₂NH + NHCH₂CH₂S), 3.45 (s, 8H, NHCH₂CH₂NH), 3.79

(s, 4H, SCH₂Ar), 3.82 (s, 6H, OCH₃), 6.99 (d, 4H, J 8.5 Hz,

ArH), 7.35 (d, 4H, J 8.5 Hz, ArH). ¹³C NMR (D₂O, 100

MHz): δ 23.2, 27.2, 35.3, 43.6, 43.8, 45.7, 47.8, 56.4, 115.2,

Tetraazaoctadecane-1,18-dithiol (Tetra-triﬂuoroacetate Salt)

(TetraN-3,4,3-DT). 1,18-Bis[(4-methoxybenzyl)sulfanyl]-

3,7,12,16-tetraazaoctadecane (14) (750 mg, 1.01 mmol) was

reacted according to general procedure B to give TetraN-

3,4,3-DT as a white powder (507 mg, 0.651 mmol, 64%).

1

Mp > 220 °C. H NMR (400 MHz, D₂O): δ 1.74 (br, 4H,

NHCH₂CH₂CH₂CH₂NH), 2.11 (br quint, 4H, NHCH₂CH₂-

CH₂NH) 2.83 (t, J 6.5 Hz, 4H, CH₂SH), 3.05-3.20 (m, 12H,

NHCH₂CH₂CH₂CH₂NH + NHCH₂CH₂CH₂NH), 3.26 (t, J

6.5 Hz, 4H, NHCH₂CH₂SH). ¹³C NMR (100 MHz, D₂O): δ

20.4, 23.1, 23.3, 44.9, 45.0, 47.6, 50.5. HR-MS (m/z, +ve

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2045

articles

Drake et al.

FAB): calculated (C₁₄H₃₄N₄S₂·H⁺) 323.2303; found 323.2311.

Elemental anal.: calculated (C₁₄H₃₄N₄S₂·3C₂HF₃O₂) C 33.93%

H 4.92% N 7.20%; found C 34.06% H 5.02% N 7.16%.

ES): calculated (C₆H₁₆N₂S₂·H⁺) 181.0833; found 181.0842.

Elemental anal.: calculated (C₆H₁₆N₂S₂· 2HCl) C 28.46% H

7.16% N 11.06%; found C 28.55% H 7.05% N 10.95%.

1,2-Dithia-5,8-diazacyclodecane (Bis-hydrochloride Salt)

(DiN-2-CDS). 1,8-Dithio-3,6-diazaoctane (DiN-2-DT) (150

mg, 0.592 mmol) was reacted according to general procedure

C to give DiN-2-CDS as a white powder (136 mg, 0.541

1,2-Dithia-5,9,14,18-tetraazacycloeicosane (Tetra-hydro-

chloride Salt)] (TetraN-3,4,3-CDS). 3,7,12,16-Tetraazaoc-

tadecane-1,18-dithiol (TetraN-3,4,3-DT) (150 mg, 0.193

mmol) was reacted according to general procedure C to give

TetraN-3,4,3-CDS as a white powder (80 mg, 0.17 mmol,

89%). Mp > 220 °C. ¹H NMR (D₂O, 400 MHz): δ 1.77 (br,

4H, NHCH₂CH₂CH₂CH₂NH), 2.13 (br quint, 4H,

NHCH₂CH₂CH₂NH), 3.00-3.20 (m, 16H, CH₂S + NHCH₂-

CH₂CH₂CH₂NH + NHCH₂CH₂CH₂NH), 3.47 (m, 4H,

NHCH₂CH₂S).¹³C NMR (D₂O, 100 MHz): δ 23.3, 23.4, 32.5,

45.1, 45.2, 46.5, 47.7. Each peak assigned in the ¹³C NMR

is 2 signals with almost identical chemical shifts. MS (m/z,

ES): 321 (100%), 641 (14%). HR-MS (m/z, ES): calculated

(C₁₄H₃₂N₄S₂·H⁺) 321.2147; found 321.2151. Elemental anal.:

calculated (C₁₁H₂₅N₃S₂·3HCl) C 36.05% H 7.78% N 12.01%;

found C 35.89% H 7.73% N 11.89%.

1

mmol, 91%). Mp > 220 °C. H NMR (D₂O, 400 MHz): δ

3.06 (t, J 6.5 Hz, 4H, CH₂S), 3.53 (m, 8H, NHCH₂). ¹³C

NMR (D₂O, 100 MHz): δ 32.4, 43.8, 47.0. HR-MS (m/z,

ES): calculated (C₆H₁₆N₂S₂·H⁺) 179.0677; found 179.0673.

Elemental anal.: calculated (C₆H₁₄N₂S₂· 2HCl) C 28.68% H

6.42% N 11.15%; found C 28.80% H 6.36% N 11.00%.

Biological Assays. Formation of Polyplexes. A solution

of pDNA (10 mg mL^-1) in the appropriate buffer was further

diluted with the buffer before the appropriate amount of the

cyclic disulﬁde solution (either 1 mg mL^-1or 0.1 mg mL^-1

in the same buffer) was added. The resulting solution was

shaken, centrifuged at 13,200 rpm for 5 s and then allowed

to stand for 30 min before use.

N³,N⁶-Di(tert-butoxycarbonyl)-1,8-dithioacetate-3,6-diaza-

octane (11). PPh₃(6.02 g, 22.1 mmol) was dissolved in dry

THF (200 mL), and the resulting solution was cooled to 0

°C. Diisopropyl azodicarboxylate (4.52 mL, 22.1 mmol) was

added to it, and the reaction mixture was stirred at 0 °C for

a further 45 min, during which time a precipitate developed.

N³,N⁶-Di(tert-butoxycarbonyl)-3,6-diazaoctane-1,8-diol (10)

(1.75 g, 5.02 mmol) and thioacetic acid (1.64 mL, 22.1

mmol) were dissolved in dry THF (50 mL), and this solution

was added to the previous suspension dropwise over 5 min.

The resulting suspension was stirred at 0 °C for 1 h and

then at ambient temperature for 15 h, during which time the

suspension cleared to leave a yellow solution. This solution

was concentrated under reduced pressure, and the crude

product was then puriﬁed over silica (4:1, v:v, hexane:

EtOAc) to give 11 as a white solid (1.53 g, 3.28 mmol, 65%).

Agarose Gel Electrophoresis. Standard Conditions. Poly-

plex solutions of various N/P ratios containing 500 ng of

pDNA in 8 µL of 4 mM HEPES (pH ) 7.4) were made up

using the procedure described previously. These solutions

were diluted with 8 µL of HEPES and then 4 µL of loading

buffer and run for 1 h at 110 V through a 0.8% agarose gel

plate containing 20 ng µL^-1of ethidium bromide. The

resulting plate was visualized under UV light.

ReductiVe Conditions. Polyplex solutions of various N/P

ratios containing 500 ng of pDNA in 8 µL of 4 mM HEPES

(pH ) 7.4) were made up using the procedure described

previously. These solutions were diluted with 8 µL of a 200

mM DTT solution (in HEPES) and then 4 µL of loading

buffer and run for 1 h at 110 V through a 0.8% agarose gel

plate containing 20 ng µL^-1of ethidium bromide. The

resulting plate was visualized under UV light.

PCS Measurements. Polyplex solutions of various N/P

ratios containing 10 µg of DNA in 200 µL of 4 mM HEPES

(pH ) 7.4) were made up using the procedure described

previously. Three readings were taken for each sample, and

the results were expressed as the mean and standard deviation

of these three readings.

Scanning Electron Microscopy. A polyplex solution con-

taining 10 µg of pDNA was formed with TetraN-3,4,3 at an

N/P ratio of 10 in 30 µL of 4 mM HEPES (pH ) 7.4). This

solution was applied to a standard SEM stub, and the solvent

was evaporated off under ambient conditions. The stub was

then gold plated using standard techniques and analyzed via

SEM.

1

R_f) 0.29 (4:1, v:v, hexane:EtOAc). Mp ) 76-78 °C. H

NMR (CDCl₃, 400 MHz): δ 1.42 (s, 18H, OCCH₃), 2.28 (s,

6H, SC(O)CH₃), 2.90 - 3.10 (m, 4H, CH₂S), 3.20-3.35 (m,

8H, NCH₂). ¹³C NMR (CDCl₃, 100 MHz): complex due to

4 possible rotamers present. HR-MS (m/z, CI⁺): found

465.2097; calculated (C₂₀H₃₆N₂O₆S₂·H⁺) 465.2093. Elemen-

tal anal.: found C 51.80% H 7.93% N 6.15%; calculated

(C₂₀H₃₆N₂O₆S₂) C 51.70% H 7.81% N 6.03%.

3,6-Diazaoctane-1,8-dithiol (Bis-hydrochloride Salt) (DiN-

2-DT). N³,N⁶-Di(tert-butoxycarbonyl)-1,8-dithioacetate-3,6-

diazaoctane (11) (1.14 g, 2.54 mmol) was dissolved in

concentrated HCl (50 mL), a process which caused ef-

fervescence to be observed. The resulting solution was heated

at reﬂux for 1 h. The solvent was then removed under

reduced pressure, and the resulting white powder was dried

under reduced pressure to give DiN-2-DT as a white powder

In Vitro Transfections. 50,000 cells were seeded in a 24-

well plate and allowed to grow in 500 µL of serum containing

medium until they were 90% conﬂuent (usually 48 h). The

medium was then removed, and each well was washed with

500 µL of serum free medium. 200 µL of Opti-MEM was

then added to each well, followed by a 150 µL polyplex

solution, each containing 2 µg of plasmid DNA at various

N/P ratios, in 4 mM HEPES (pH ) 7.4). The cells were

1

(543 mg, 2.14 mmol, 84%). Mp > 220 °C. H NMR (D₂O,

400 MHz): δ 2.87 (t, J 6.5 Hz, 4H, CH₂SH), 3.34 (t, J 6.5

Hz, 4H, NHCH₂CH₂SH), 3.51 (s, 4H, NHCH₂CH₂NH). ¹³

C

NMR (D₂O, 100 MHz): δ 20.5, 43.5, 51.0. HR-MS (m/z,

2046 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

of PBS before being incubated in serum-containing medium

for a further 19 h. 20 µL of a 5 mg/mL solution of 3-(4,5-

dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)

was then added to each of the samples, which were then

incubated at 37 °C for 5 h. The medium was removed, and

200 µL of DMSO was added to each well. The absorbance

of each well at 550 nm was then measured and the % cell

viability calculated as a percentage of the positive control.

jetPEI was used in ratios in accordance with the manufac-

turer’s instructions in identical procedures to those described

above. Each sample was analyzed in triplicate, allowing for

the calculation of a mean and standard deviation for each

one.

incubated at 37 °C for 6 h, after which time the polyplex

solutions were removed and replaced with 500 µL of serum

containing medium. The cells were incubated for a further

40 h. The medium was then removed from the cells, which

were washed with 500 µL of PBS before 300 µL of 1× lysate

buffer was added. After being shaken at ambient temperature

for 30 min the lysate was removed, using vigorous means,

from the wells and centrifuged at 13,200 rpm for 2.5 min.

50 µL of the lysate supernatant was analyzed for luciferase

content, and 25 µL was analyzed for protein content using

standard assays. Results were presented as relative light units

(RLU) per mg of protein. jetPEI was used in ratios in

accordance with the manufacturer’s instructions in identical

procedures to those described above using HEPES as the

buffer. Each sample was analyzed in triplicate, allowing for

the calculation of a mean and standard deviation for each

one.

Cytotoxicity (LDH Assay). 20,000 cells were seeded in

a 96-well plate and allowed to grow in 200 µL of serum

containing medium until they were 90% conﬂuent (usually

48 h). The medium was then removed, and each well was

washed with 200 µL of serum free medium. 75 µL of Opti-

MEM was then added to each well, followed by a 75 µL

polyplex solution, each containing 1 µg of plasmid DNA at

various N/P ratios, in 4 mM HEPES (pH ) 7.4). In addition

a blank sample, consisting of 75 µL of Opti-MEM and 75

µL of 4 mM HEPES, was also added to one set of cells to

allow the background level of LDH released to be measured.

The cells were incubated at 37 °C for 24 h, after which time

the medium was removed, centrifuged at 13,200 rpm for 2.5

min and kept at 4 °C. The cells were then washed with 200

µL of PBS before 75 µL of 1× lysate buffer was added.

After being shaken at ambient temperature for 30 min the

lysate was removed, using vigorous means, from the wells,

diluted with 75 µL of Opti-MEM and centrifuged at 13,200

rpm for 2.5 min. 50 µL from each of the medium samples

and lysates was analyzed for LDH content according to the

manufacturer’s instructions. Prior to performing the colori-

metric measurement for LDH content the spectrometer was

zeroed using a mixture of ddH₂O and Optim-MEM (1:1, v:v).

The % cell death was calculated as the amount of LDH

released divided by total LDH content from the well. jetPEI

was used in ratios in accordance with the manufacturer’s

instructions in identical procedures to those described above.

Each sample was analyzed in triplicate, allowing for the

calculation of a mean and standard deviation for each one.

Cytotoxicity (MTT Assay). 20,000 cells were seeded in

a 96-well plate and allowed to grow in 200 µL of serum

containing medium until they were 90% conﬂuent (usually

48 h). The medium was then removed and each well was

washed with 100 µL of PBS. 50 µL of DMEM was then

added to each well, followed by a 50 µL polyplex solution,

each containing 1 µg of plasmid DNA at various N/P ratios,

in 4 mM HEPES (pH ) 7.4). In addition cells treated with

just DMEM (100 µL) were used as the positive control. The

cells were incubated at 37 °C for 4 h before the polyplex

solutions were removed and the cells washed with 100 µL

Gene Delivery to Mouse Airways and ꢀ Gal Expres-

sion in ViWo. Transfection Protocol. The mice were anes-

thetized, using isoﬂurane, and polyplex solutions, containing

25 µg of pDNA in 100 µL of 5% glucose, were applied

intranasally in one dose. After 24 h the mice were humanely

sacriﬁced with an intraperitoneal injection of pentobarbital

and their lungs removed. Those mice that were pretreated

with N-acetylcysteine (NAC) had 100 µL of a 50 mM NAC

solution in 5% glucose applied intranasally 30 min prior to

the instillation of the polyplexes.

ELISA Assay. The lung tissue from above was placed in

lysis buffer (1×) and homogenized on ice for 30 s before

being lysed using three freeze-thaw cycles. The lysate was

then centrifuged and the supernatant was analyzed for both

ꢀ-gal and total protein content using standard assays.

Histology. The lung tissue originating from the left lobe

was ﬁxed by placing it into a 0.5% glutaraldehyde solution

for 10 min. It was then washed repeatedly with PBS before

being incubated for 24 h at 37 °C in an X-gal solution in

order to stain it for ꢀ-gal. The tissue was then washed with

PBS and dehydrated by sequential immersion in 70%, 95%

and ﬁnally absolute ethanol. It was then cleared by soaking

it in a mixture of benzyl benzoate and benzyl alcohol (2:1,

v:v) and examined under a light microscope. Two separate

histochemical analyses were then performed: OCT embed-

ding after cryosectioning or parafﬁn embedding. For the

former the samples were decleared via sequential immersion

in 100% MeOH, MeOH in PBS and ﬁnally PBS. They were

then inﬁltrated with 20% sucrose in PBS before being

embedded in a mixture of OCT and 20% sucrose in PBS

and frozen via immersion in isopentane, previously precooled

with liquid N₂. Cryostat sections 9 µm thick were counter-

stained with eosin to provide the ﬁnal sample to be analyzed.

Other samples were embedded in parafﬁn and then cut into

4 µm sections, which were then counterstained with eosin.

Both groups of samples were then analyzed for ꢀ-gal staining

by direct light microscopy.

Results

Synthesis of Dithiols and Cyclic Disulﬁdes. The synthetic

methodology developed to selectively difunctionalize ter-

minal primary amines of tri- and tetraamines with 2-mer-

captoethyl groups and their subsequent conversion to dis-

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2047

articles

Drake et al.

Scheme 1^a

^a(i) 4-Methoxybenzyl chloride, K₂CO₃, MeOH; 2 h; 96%. (ii) SO₃-pyridine, DMSO, DIPEA, DCM; 30 min; 52%. (iii) (a) MgSO₄, DCM, 8 h; (b) NaBH₄,

DCM:EtOH (1:1, v:v); 15 h; (c) HCl, DCM:EtOH (1:1, v:v); 60 °C; 6 h; 49-52%. (iv) TFA; reﬂux; 3 days; 44-64%. (v) Concentrated HCl, DMSO; 6 h;

38-92%. Abbreviations: MeBn ) 4-methoxybenzyl.

Scheme 2^a

performed according to a previously published procedure,⁵⁸

followed by universal deprotection in reﬂuxing concentrated

HCl giving DiN-2-DT as the dihydrochloride salt. Oxidation

with DMSO resulted in the expected cyclic disulﬁde DiN-

2-CDS. With all these compounds in hand, a series of pDNA

condensation experiments were performed to assess whether

either the dithiols or cyclic disulﬁdes formed polyplexes

suitable for gene delivery.

Polyplex Formation and Characterization. Initially, the

pDNA condensing capacities of two spermidine derivatives,

TriN-3,4-DT and TriN-3,4-CDS, were investigated by aga-

rose gel electrophoresis and photon correlation spectroscopy

(PCS). Polyplexes of varying N/P ratios were assembled by

the addition of the cationic species to pDNA in 4 mM

HEPES buffer (pH 7.4). Both compounds were found to fully

condense pDNA at an N/P ratio of 2.5 (Figure 2). Addition

of dithiothreitol (DTT) (100 mM, 1 h) was found to enable

clean release of pDNA as illustrated by the reappearance of

pDNA bands on the agarose gel for all N/P ratios where

pDNA condensation had previously been observed.

^a(i) (Boc)₂O, CHCl₃; 5 h; 48%. (ii) PPh₃, DIAD, AcSH, THF; 0 °C-rt;

15 h; 65%. (iii) Concentrated HCl; reﬂux; 1 h; 84%. (iv) Concentrated HCl,

DMSO; 6 h; 91%. Abbreviations: Boc ) tert-butoxy carbonyl, Ac ) acetyl.

ulﬁdes is outlined in Scheme 1. After an extensive search

of suitable reactions, reductive amination of spermidine with

aldehyde 3 (synthesized according to previous literature

example),⁵⁵followed by extended reﬂux in triﬂuoroacetic

acid, gave the desired dithiol, TriN-3,4-DT, in good yield.

Facile oxidation of this species under high dilution, using

dimethyl sulfoxide (DMSO) as the oxidant, furnished the

corresponding cyclic disulﬁde, which was labeled TriN-3,4-

CDS. In this nomenclature the preﬁx refers to the number

of secondary amine functionalities between each disulﬁde

bond and the numbers refer to the methylene spacings

between them. The code ﬁnishing the label identiﬁes the

compound as either the dithiol (DT) or cyclic disulﬁde

(CDS). Three further amines (6-8) were also successfully

treated via the same methodology to give the corresponding

dithiols and cyclic disulﬁdes. ES-MS analysis of all four

cyclic disulﬁdes showed peaks corresponding to mono- and

bicyclic species. No evidence for any higher order molecular

weight cyclic oligomers or polymers was obtained. No further

puriﬁcation of these cyclic compounds was found to be

necessary other than heating under reduced pressure to

remove excess DMSO. This was undertaken to ensure that

a small concentration of free thiols remained to assist the

anticipated thiol-disulﬁde exchange reaction during pDNA

condensation.

Particle sizing of the polyplexes was determined by means

of PCS (Figure 3a). The condensation of pDNA with TriN-

3,4-DT resulted in the formation of large aggregates of well

over 1000 nm in diameter that were not considered suitable

for gene delivery. Pleasingly, however, polyplexes of be-

tween 150 and 250 nm were obtained when pDNA was

condensed with TriN-3,4-CDS at an N/P ratio >2.5. Scanning

electron microscopy (SEM) images of polyplexes formed

with TetraN-3,4,3-CDS at an N/P ratio of 10 conﬁrm the

presence of such nanometric structures, circular in shape with

diameters of around 200 nm (Figure 3b). Therefore, given

these data, we came to the conclusion that the cyclic disulﬁde

approach should be a more effective strategy for pDNA

condensation than dithiols. Accordingly DiN-2-CDS, TriN-

2,3-CDS, TriN-3,4-CDS, TetraN-2,3,2-CDS and TetraN-

3,4,3-CDS were taken forward for further in Vitro evaluation.

For the functionalization of diamines, an alternative route

(Scheme 2) was sought due to our inability to successfully

functionalize 1,2-ethylenediamine with aldehyde 3. After

Boc-protection of amine functional groups on diol 9,⁵⁶

Mitsunobu displacement of the alcohols with thioacetate was

(58) Oiry, J.; Pue, J. Y.; Laval, J. D.; Fatome, M.; Imbach, J. L.

Synthesis and Radioprotective Activity of WR-1065 Derivatives:

N-(2-acetylthioethyl)-1,3-propanediamine and N, N′-bis(2-acetylth-

ioethyl)-1,3-propanediamine. Eur. J. Med. Chem. 1995, 30 (1),

47–52.

2048 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

Figure 2. Agarose gels showing condensation of pDNA

and its release under reductive conditions (100 mM

DTT) for (a) TriN-3,4-DT and (b) TriN-3,4-CDS.

Polyplex solutions containing 500 ng of pDNA were

made up in 10 µL of 4 mM HEPES at various N/P

ratios. After 30 min one set of samples was further

diluted with 10 µL of 4 mM HEPES while the other set

was diluted with 10 µL of 200 mM DTT. Following a

further 30 min incubation the samples were run on a

0.8% agarose plate a 110 V for 1 h. Key: 1, naked

pDNA; 2, N/P ) 0.5; 3, N/P ) 1; 4, N/P ) 2.5; 5, N/P )

5; 6, N/P ) 12.5; 7, deacetylated PEI N/P ) 5; 8,

spermine N/P ) 25; 9, N/P ) 0.5 (100 mM DTT); 10, N/

P ) 1 (100 mM DTT); 11, N/P ) 2.5 (100 mM DTT);

12, N/P ) 5 (100 mM DTT); 13, N/P ) 12.5 (100 mM

DTT); 14, deacetylated PEI N/P ) 5 (100 mM DTT).

Gratifyingly, all four analogues based on tri- and tetraamine

structures were able to condense pDNA into polyplexes with

diameters of 150-250 nm (Table 1). Agarose gel electro-

phoresis showed that DiN-2-CDS was unable to condense

pDNA even at an N/P ratio of 12.5 (data not shown),

indicating that it was unlikely to be an effective vector;

however, it was still tested in Vitro to conﬁrm this assumption.

In Vitro Transfection. The ability of the four cyclic

disulﬁdes to transfect both a CHO-K1 and an A549 cell line

with a luciferase expressing pDNA was assessed in com-

parison with jetPEI, which was used in ratios in accordance

with the manufacturer’s guidelines⁶for each speciﬁc cell line.

Polyplexes containing 2 µg of pEGFP-Luc-DNA were

prepared at various N/P ratios in 4 mM HEPES and applied

to cells for a period of 6 h at 37 °C. Serum-free conditions

were used during this incubation as it has been reported

previously that serum can lower the transfection efﬁcacies

of polyamine vectors such as PEI. This is probably due to

its destabilizing effect on polyplexes,^13,59although its effect

on the cyclic disulﬁde’s polyplexes was not tested and hence

it is not known whether they would be similarly affected.

The polyplex solutions were then removed and replaced with

media, and the levels of luciferase expression were assessed

after a further 40 h of cell incubation using standard assays

(Figure 4).

Figure 3. (a) PCS measurements of polyplexes formed

from TriN-3,4-DT and TriN-3,4-CDS. Polyplexes con-

taining 10 µg of pDNA in 200 µL of 4 mM HEPES buffer

(pH

) 7.4) were made up with TriN-3,4-DT and

TriN-3,4-CDS at various N/P ratios and measured by

PCS. (b) SEM images of polyplexes formed in 4 mM

HEPES from pDNA and TetraN-3,4,3-CDS at N/P ratio

of 10.

Table 1. Particle Sizes of Polyplexes Formed with

TriN-2,3-CDS, TriN-3,4-CDS, TetraN-2,3,2-CDS and

TetraN-3,4,3-CDS as Measured by PCS

0.5

312 ( 126 282 ( 119 195 ( 77

193 ( 86 351 ( 160 3000 ( 1350 221 ( 92 101 ( 38

1

2.5

5

12.5

TriN-2,3

TriN-3,4

201 ( 78 128 ( 51

TetraN-2,3,2 291 ( 129 260 ( 112 2163 ( 966 204 ( 76 197 ( 79

TetraN-3,4,3 388 ( 163 2191 ( 984 118 ( 46 200 ( 80 149 ( 60

When tested on CHO-K1 cells (Figure 4a), at N/P ratios

of 5 and above, the four analogues based upon tri- and

tetraamine backbones were able to mediate levels of trans-

fection to within an order of magnitude of jetPEI. In contrast,

DiN-2-CDS was unable to mediate transfection above

background levels. For each of the four efﬁcient analogues

the levels of transfection rose to a maximum before falling

again as the N/P ratio was increased further, presumably due

to the increasing cytotoxicity of the vector. There is very

(59) Lim, Y.-B.; Kim, S.-M.; Suh, H.; Park, J.-S. Biodegradable,

Endosome Disruptive, and Cationic Network-Type Polymer as a

Highly Efﬁcient and Nontoxic Gene Delivery Carrier. Bioconju-

gate Chem. 2002, 13 (5), 952–957.

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2049

articles

Drake et al.

Cytotoxicity. The cytotoxicity of the ﬁve cyclic disulﬁdes,

along with jetPEI, was measured using a standard lactate

dehydrogenase (LDH) release assay. LDH is a cytoplasmic

enzyme that is released into medium due to cell death, hence

the LDH assay provides a sensitive method to assess

cytotoxicity. The cytotoxicity of one analogue (TetraN-3,4,3-

CDS) was also measured using the MTT assay, however this

assay was judged to be less sensitive than the LDH one and

hence was not used for all analogues (see Supporting

Information). Polyplexes containing 1 µg of pEGFP-Luc-

DNA were prepared at various N/P ratios in 4 mM HEPES

and applied to cells, previously grown to 90% conﬂuence,

for a period of 24 h at 37 °C. It is conceivable that the high

conﬂuence of the cells might reduce the cytotoxic effect of

the polyplexes, however such an effect should apply equally

to all experiments and hence any differences observed should

be the direct result of the characteristics of the polyplexes

themselves. Covering media were then removed from the

cells in each well, which were then independently lysed.

Analysis of the LDH content of both the media and lysate

was performed using a standard kit, and the level of cell

death was calculated by taking the LDH content of the media

as a percentage of the total LDH content measured in a

corresponding untreated population of cells in a well (Figure

5).

Figure 4. Transfection of (a) CHO-K1 and (b) A549 cells

with luciferase pDNA using cyclic disulﬁdes. Naked

pDNA and jetPEI were used as negative and positive

controls respectively. Polyplex solutions, made up in a

mixture of 4 mM pH 7.4 HEPES and serum-free

medium, containing 2 µg of luciferase pDNA were

incubated with the cells for 6 h before being replaced

with serum-containing medium for a further 40 h. The

cells were then lysed, and the luciferase and protein

content of the lysates was measured (CHO-K1: for

cyclic disulﬁdes n ) 3 ( standard deviation; jetPEI n )

12 ( standard deviation; DNA n ) 12 ( standard

deviation. A549: for cyclic disulﬁdes n ) 4 ( standard

deviation; jetPEI n ) 12 ( standard deviation; DNA n )

20 ( standard deviation).The optimal N/P ratio for each

cyclic disulﬁde was compared to jetPEI using Welch’s t

test (p ) 0.05), and those cyclic disulﬁdes able to

transfect at levels statistically indistinguishable from

jetPEI are marked with *.

When tested on CHO-K1 cells both analogues based upon

a propylene/butylene inter-nitrogen spacing (i.e., TriN-3,4-

CDS and TetraN-3,4,3-CDS) showed no cytotoxicity levels

above background at N/P ratios of 5 and 7.5, at which ratios

optimal transfection was also mediated. This compares

favorably with jetPEI, which caused the release of 53 ( 14%

of the LDH, signiﬁcantly above background levels (21 (

7%) thereby demonstrating that the bioreducible vectors

employed here are able to alleviate some of the cytotoxicity

associated with PEI. When higher N/P ratios were tested, it

was noted that TetraN-3,4,3-CDS proved to be marginally

more toxic than TriN-3,4-CDS, perhaps indicating a negative

consequence of raising the number of intradisulﬁde amines.

In comparison, two of the cyclic disulﬁdes based upon

ethylene/propylene inter-nitrogen spacings (DiN-2-CDS and

TriN-2,3-CDS) exhibited high cytotoxicity even at the

relatively low N/P ratios (5 and above). This cytotoxicity

could be obviated by preparing the polyplexes in a stronger

buffer solution (40 mM HEPES, data shown in Supporting

Information) to buffer the greater acidity of these polymers

compared with TriN-3,4-CDS and TetraN-3,4,3-CDS. In-

terestingly TetraN-2,3,2-CDS, despite its ethylene/propylene

spacings, exhibited no such extreme toxicity for reasons

which are currently unclear. Similar overall trends were

observed with A549 cells although these were far less

susceptible to polyamine induced damage. Indeed even jetPEI

did not provoke any LDH release signiﬁcantly above

background levels.

little to choose between the optimal performance of any of

the tri- and tetraamine analogues in this cell line, in contrast

to the results obtained when A549 cells were used (Figure

4b). Here the two tetraamine analogues TetraN-2,3,2-CDS

and TetraN-3,4,3-CDS mediated optimal levels of transfec-

tion at least half an order of magnitude higher than the tri-

and diamines. Noticeably this cell line proved harder to

transfect with the cyclic disulﬁdes, as their performance

remained orders of magnitude beneath that of jetPEI. These

results demonstrate that small molecule cyclic disulﬁdes with

tri- and tetraamine backbones can successfully transfect both

A549 and CHO-K1 cells signiﬁcantly above background

levels and, in the case of CHO-K1 cells, to within half an

order of magnitude of jetPEI, currently considered the most

effective commercial cationic polymeric vector.

The cytotoxicity of TetraN-3,4,3-CDS, one of our most

successful compounds in terms of mediating high levels of

transfection at low N/P ratios with no associated cytotoxicity,

was also assessed as the free material in direct comparison

2050 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

Figure 6. Cytotoxicity of jetPEITM and TetraN-3,4,3-

CDS to CHO-K1 cells as free materials measured by

LDH release assay. Free material was diluted to the

appropriate ﬁnal concentrations with a mixture of 4 mM

HEPES and Opti-MEM (150 µL, 1:1, v:v) and then

Figure 5. Cytotoxicity caused by polyplexes containing

cyclic disulﬁdes to (a) CHO-K1 and (b) A549 cells

assessed via LDH release assay. Blank solutions of

buffer and medium (1:1, v:v) and jetPEI were used as

negative and positive controls respectively. Polyplex

solutions, made up in a mixture of 4 mM pH 7.4 HEPES

and serum-free medium (1:1, v:v), containing 1 µg of

luciferase pDNA were incubated with the cells for 24 h

prior to measurement of LDH content of both media and

cell lysate (n ) 3, ( standard deviation). (CHO-K1: for

cyclic disulﬁdes n ) 3 ( standard deviation; jetPEI n )

24 ( standard deviation; blank n ) 24 ( standard

deviation. A549: for cyclic disulﬁdes n ) 4 ( standard

deviation; jetPEI n ) 12 ( standard deviation; DNA n )

24 ( standard deviation).

incubated with the cells for 24

h prior to the

measurement of LDH content of both media and cell

lysate (n ) 3, ( standard deviation). (a) Results

presented as a function of concentration of free material

(mg/mL). (b) Results presented as a function of the

concentration of amine groups (mmol).

jetPEI in 5% glucose, the recommended solvent for in ViVo

jetPEI, at various N/P ratios with a ꢀ-galactosidase (ꢀ-gal)

expressing pDNA and delivered intranasally to the lungs of

female BALB/c mice. At 24 h postinstillation, the murine

lungs were removed and analyzed for ꢀ-gal expression using

standard assays. The amount of ꢀ-gal expressed (ng of ꢀ-gal

protein/100 mg of protein) was then calculated, and the

results obtained are shown in Figure 7 (see Supporting

Information for the individual results for each mouse). While

no signiﬁcant ꢀ-gal expression levels were observed in

nontreated mice, a signiﬁcant level was clearly measurable

in the lung homogenates of mice treated with polyplexes

formed with TetraN-3,4,3-CDS at both N/P ratios of 5 and

8 (respectively 1.663 ( 0.3976, n ) 10, and 1.608 ( 0.3180,

n ) 5), with the highest activity being observed at N/P of 5.

Nevertheless, these ꢀ-gal levels were still lower than those

obtained with in ViVo jetPEI (3.059 ( 0.2644, n ) 15), hence

we sought an adjuvant that might improve the performance

of the cyclic disulﬁde. N-Acetylcysteine (NAC) is a major

component of Nacystelyn, an agent widely used in the clinic

to alter mucus properties.⁶¹It is also known to be a precursor

to jetPEI by means of the same LDH assay (Figure 6).

Happily TetraN-3,4,3-CDS exhibited a markedly better

cytotoxicity proﬁle than jetPEI, causing minimum cell death

without toxicity when equivalent levels of jetPEI resulted

in near 90% cell death (0.02 mg mL^-1). In order to ensure

that this effect was not the result of TetraN-3,4,3-CDS’s

lower density of amines, the results were also converted to

a function of amine concentration (Figure 6b) and, although

less pronounced, the less cytotoxic nature of TetraN-3,4,3-

CDS was still evident.

Gene Transfection into the Mouse Airways in ViWo. The

two most promising candidates identiﬁed during the in Vitro

tests, TetraN-3,4,3-CDS and TetraN-2,3,2-CDS, were taken

forward for in ViVo transfection analysis. PEI has proved to

be a successful vector for transfection of the lung when

instillated intranasally,⁶⁰marking it out as having great

potential for the treatment of lung disorders such as cystic

ﬁbrosis. Hence we decided to compare our two lead

compounds to in ViVo jetPEI in this situation. Polyplexes

were formulated initially with TetraN-3,4,3-CDS and in ViVo

(60) Wiseman, J. W.; Goddard, C. A.; McLelland, D.; Colledge, W. H.

A Comparison of Linear and Branched Polyethylenimine (PEI)

with DCChol/DOPE Liposomes for Gene Delivery to Epithelial

Cells In Vitro and In ViVo. Gene Ther. 2003, 10, 1654–1662.

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2051

articles

Drake et al.

Figure 7. Summary of ꢀ-Gal expression from murine

lung following intranasal instillation of polyplexes con-

structed with TetraN-3,4,3-CDS and jetPEI. Polyplexes

containing 25 µg of pDNA in 100 µL of 5% glucose

were introduced intranasally into female BALB/c mice.

Twenty-four hours later the mice were sacriﬁced and

the ꢀ-Gal levels in their lungs measured via a ꢀ-Gal

ELISA assay. The groups of mice which were

pretreated with N-acetylcysteine (marked as NAC) were

perfused intranasally with 100 µL of a 50 mM NAC

solution 30 min prior to the instillation of the polyplex

solutions. Those groups marked with + achieved levels

of transfection signiﬁcantly higher than the nontreated

mice (ANOVA, Bonferroni corrected, p < 0.05) and

those marked with * were statistically indistinguishable

from jetPEI (ANOVA, Bonferroni corrected, p > 0.05).

of glutathione,⁴⁹hence we postulated that it might be able

to aid a reductively active vector such as TetraN-3,4,3-CDS.

Mice were perfused with NAC (100 µL, 50 mM in 5%

glucose) intranasally 30 min prior treatment with the poly-

plexes, signiﬁcantly increasing ꢀ-gal expression with TetraN-

3,4,3-CDS (3.662 ( 0.3228, n ) 5) to the levels exhibited

by in ViVo jetPEI, which itself was unaffected by pretreatment

with NAC. TetraN-2,3,2-CDS was subsequently tested, using

the NAC pretreatment, however it was unable to match

TetraN-3,4,3-CDS, mediating levels of transfection that were

signiﬁcantly lower (0.733 ( 0.279, n ) 5).

Histological analysis of lung tissues transfected with ꢀ-gal

using either TetraN-3,4,3-CDS or jetPEI as a vector was then

performed to precisely localize the transfected areas. Figure

8a shows an in toto x-Gal staining performed on the whole

left lobes extracted from untreated mice and those treated

with either TetraN-3,4,3-CDS or jetPEI. A strong blue

staining is observed in both TetraN-3,4,3-CDS- and jetPEI-

treated tissues, while only a light staining is seen in untreated

mice. Subsequently histological analysis of the murine lung

tissues was performed using two separate staining protocols,

namely OCT embedding after cryosectioning and parafﬁn

embedding. In both cases eosin counterstaining was used to

aid the identiﬁcation of the blue ꢀ-gal expressing areas

Figure 8. Histological analysis of lung tissue transfected

with ꢀ-gal using either TetraN-3,4,3-CDS or jetPEI as a

vector. (a) ꢀ-Gal expression in the mouse lungs after an

in toto x-Gal staining of the whole left lobe: A-left,

nontreated;A-right,TetraN-3,4,3-CDS;B-left,TetraN-3,4,3-

CDS

+ NAC; B-right, jetPEI. (b) Localization of

blue-stained cells using lung tissues produced either by

cryosectioning followed by OCT-embedding procedure

(C to F) or via parafﬁn treatment (G to H). Scale bars: 5

µm.

(Figure 8b). Both methods highlighted the successful trans-

fection achieved with TetraN-3,4,3-CDS, with close exami-

nation revealing that the major sites of transfection, marked

with arrowheads in Figure 8b, were airway epithelia from

the bronchioles and bronchioli (C, D). Evidence was also

observed for the transfection of macrophages and possibly

pneumocytes in the parenchymal area (E, F), however the

signiﬁcant levels of background expression found in these

cells do not allow for ﬁrm conclusions to be made on this

point.

(61) App, E. M.; Baran, D.; Dab, I.; Malfroot, A.; Cofﬁner, M.;

Vanderbist, F.; King, M. Dose-Finding and 24-h Monitoring for

Efﬁcacy and Safety of Aerosolised Nacystelyn in Cystic Fibrosis.

Eur. Respir. J. 2002, 19, 294–302.

2052 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

effective pDNA condensation. Taken together the perfor-

mance of these vectors provides excellent evidence for the

proposed TCTP, although to date no direct evidence of its

occurrence has been obtained. Dithiols were also assessed

as potential gene delivery vectors, however despite proving

able to condense pDNA at low N/P ratios via a kinetically

controlled polymerization (KCP) the resultant polyplexes

were large aggregates of the order of thousands of nanome-

ters in diameter, in contrast to the discrete polyplexes of

between 100 and 250 nanometers formed with the cyclic

disulﬁdes, indicating that they were not suitable vectors.

Discussion

The success of gene therapy is dependent on the develop-

ment of safe, efﬁcient vectors. Cationic polyamines, espe-

cially PEI, have long been considered as promising, however

to date their progress toward the clinic has been hampered

by concerns over their toxicity and a lack of efﬁcacy when

compared to viruses. This paper has described the synthesis

and biological characterization of several small molecule

cyclic disulﬁde vectors that were designed to respond to the

presence of pDNA by forming bioreducible poly(disulﬁde)s

via a TCTP. The most effective analogues were able to effect

high levels of transfection, statistically indistinguishable from

jetPEI, currently one of the leading commercially available

synthetic vectors, under certain conditions both in Vitro and

in ViVo. Crucially they also exhibited a markedly improved

cytotoxicity proﬁle compared to jetPEI, potentially making

them more viable for future use in the clinic.

The cyclic disulﬁdes were synthesized from the analogous

dithiols under high dilution in order to ensure that they

remained small molecules, with ES-MS showing the presence

of only mono- and bicyclics. Such species should be

ineffectual vectors due to their low molecular weight,

however we postulated that the introduction of pDNA would

initiate a TCTP resulting in the formation of suitable

poly(disulﬁde)s which would subsequently be able to effect

pDNA condensation. It was considered that the addition of

pDNA to a solution of a cyclic disulﬁde, buffered at a pH

of 7.4, would act in the same way as the introduction of a

template molecule to a dynamic combinatorial library in that

it creates a new energy minimum for the system to access.

It has been established that the process of pDNA condensa-

tion by polycationic species is entropically driven, with the

entropic term being a combination of both the favorable

release of small molecule counterions and the putative effect

of bimolecular combination.^62,63Considering a range of

molecular weights from small molecule cyclic disulﬁdes to

high molecular weight polymers, it can be seen that the

favorable release of counterions is constant, however the

bimolecular combination penalty paid is greater the lower

the molecular weight of the polycation. In response to this

entropic driving force the system will strive to reduce this

penalty via the rearrangement of the small molecule cyclics

into high molecular weight polymers. All the biological data

obtained for the cyclic disulﬁdes points to the presence of

high molecular weight poly(disulﬁde)s upon the addition of

pDNA as both pDNA condensation and successful transfec-

tion were mediated at low N/P ratios (2.5 and 5 respectively

for TriN-3,4-CDS). When exposed to strongly reductive

conditions (100 mM DTT) complete pDNA release was

observed for all the cyclic disulﬁdes, demonstrating clearly

that the analogous monomeric species were unable to effect

Molecular weight is known to be a key determinant of a

polymer’s characteristics and has been demonstrated to

impact both PEI’s efﬁcacy and its cytotoxicity. But, despite

its prominence in determining the success of a polymeric

vector, it is a factor that remains under poor synthetic control

with difﬁculties remaining in achieving consistency between

separate batches of the same compound. These issues are

resolved by the use of a small molecule such as the cyclic

disulﬁdes, which can be consistently and reproducibly

produced due the high dilutions employed during their

synthesis. In addition structure-activity relationships can be

proposed without concerns that they might be inﬂuenced by

differing molecular weights. This is not to say that the ﬁnal

molecular weight distributions of the condensing species are

identical as it is likely that each cyclic disulﬁde will respond

to the introduction of pDNA uniquely, however such

differences are inherent properties of the cyclic disulﬁdes

themselves rather than the consequence of synthetic variations.

The range of cyclic disulﬁdes accessible using the synthetic

methodology described allowed for various structure-activity

relationships to be probed. Two variables were of particular

interest: ﬁrst the number of amines between each disulﬁde

and second the buffering capacity of the amines. This second

factor was inﬂuenced by varying the inter-nitrogen spacing,

which has a direct effect of the acidity of the protonated

amines⁶⁴with shorter spacings resulting in more acidic

protons. The effect of employing more acidic protons should

be an increased buffering capacity at physiological pH, a

factor that has been suggested to be crucial to PEI’s ability

to affect endosomal escape via the so-called “proton sponge

hypothesis”^65,66and hence mediate high levels of transfection.

The transfection data from the ﬁve analogues synthesized

allows us to draw several conclusions. The most obvious is

that at least three secondary amines between each disulﬁde

(64) Geall, A. J.; Taylor, R. J.; Earll, M. E.; Eaton, M. A. W.;

Blagborough, I. S. Synthesis of Cholesteryl Polyamine Carbam-

ates: pKa Studie and Condensation of Calf Thymus DNA.

Bioconjugate Chem. 2000, 11 (3), 314–326.

(65) Akinc, A.; Thomas, M.; Klibanov, A. M.; Langer, R. Exploring

Polyethylenimine-Mediated Transfection and the Proton Sponge

Hypothesis. J. Gene Med. 2005, 7, 657–663.

(66) Boussif, O.; Lezoualc’h, F.; Zanta, M. A.; Mergny, M. D.;

Scherman, D.; Demeneix, B. A.; Behr, J.-P. A Versatile Vector

for Gene and Oligonucleotide Transfer into Cells in Culture and

In ViVo: Polyethylenimine. Proc. Natl. Acad. Sci. U.S.A. 1995,

92 (16), 7297–7301.

(62) Bronich, T.; Kabanov, A. V.; MArky, L. A. A Thermodynamic

Characterisation of the Interaction of a Cationic Copolymer with

DNA. J. Phys. Chem. B 2001, 105 (25), 6042–6050.

(63) Ou, Z.; Muthukamar, M. Entropy and Enthalpy of Polyelectrolyte

Condensation: Langevin Dynamics Simulations. J. Chem. Phys.

2006, 124 (15), 154902-.

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2053

articles

Drake et al.

bond are required to mediate any transfection above that

shown by naked pDNA. DiN-2-CDS, based upon an ethyl-

enediamine structure, proved unable to transfect either cell

line investigated, a result that was expected due to its inability

to condense pDNA even at an N/P ratio of 12.5. However

those structures based upon a tri- and tetraamine structure

successfully mediated transfection at N/P ratios of 5 and

above. There was little to choose between the performances

of these vectors in CHO-K1 cells, with the optimal levels

of transfection being similar across all four analogues. At

the highest N/P ratio tested, 25, falls in efﬁcacy were noted

for each vector, which was ascribed to their toxicity

beginning to overwhelm the cells. Pleasingly the optimal

levels of transfection achieved were within an order of

magnitude of jetPEI for each of these four analogues,

demonstrating their efﬁcacy for this cell line. When examined

using Welch’s t test (2-tailed) no signiﬁcant difference was

found between jetPEI and either TriN-3,4-CDS, TetraN-

2,3,2-CDS or TetraN-3,4,3-CDS (p > 0.05), showing that

we had succeeded in matching the efﬁcacy of the current

industrial gold standard. In contrast when tested on A549

cells the cyclic disulﬁdes mediated levels of transfection that

were orders of magnitude beneath those of jetPEI. However

this cell line, possibly as it proved to be harder to transfect,

did highlight some interesting differences between the

vectors. Both tetraamine derivatives proved to be more

efﬁcacious vectors than their triamine analogues, mediating

optimal levels of transfection that were signiﬁcantly higher.

Lengthening the number of amine functionalities between

each disulﬁde should allow for the formation of more stable

polyplexes, one possible explanation for the improved

transfections observed. In addition it is also conceivable that

lowering the density of disulﬁdes renders the polyplex less

susceptible to reductive degradation, allowing it to persist

in the cytoplasm for a longer period of time and hence

perhaps allowing it to deliver its cargo closer to the cell’s

nucleus. While this may be beneﬁcial for the transfection,

the LDH assays on the CHO-K1 reveal that this might be

hazardous in terms of cytotoxicity with TetraN-3,4,3-CDS

provoking higher levels of LDH release than TriN-3,4-CDS

at N/P ratios of 7.5 and 12.5. Again such a relationship might

be expected as a higher number of amines would result in

increased interactions with cellular membranes, potentially

leading to greater damage and hence the observed rise in

cytotoxicity.

3,4,3-CDS mediates its highest levels of transfection at lower

N/P ratios compared to TetraN-2,3,2-CDS, an observation

that again can be explained by its greater charge at physi-

ological pH. The lack of any evidence for an advantage to

shorter amine spacings leads to the conclusion that buffering

capacity is not a dominant factor in the performance of these

vectors. However when considering the cytotoxicity of the

polyplexes, it is striking that both DiN-2-CDS and TriN-

2,3-CDS are far more cytotoxic than the other analogues.

The fact that this relatively extreme cytotoxicity can be

improved by raising the strength of the buffer used to form

the polyplexes to 40 mM HEPES shows that this difference

is simply due to incomplete buffering of the more acidic

amines from these compounds rather than any inherent

feature of the polyamine structure itself, however as we

would prefer to employ lower strength buffers in ViVo this

would be a considerable disadvantage for these analogues.

Precisely why such an effect was not observed for TetraN-

2,3,2-CDS is not clear at present, although the clear

implication is that the pK_avalues for its amine are higher

than those of DiN-2-CDS and TriN-2,3-CDS.

Many of the structure-activity trends described above

correlate well with those already reported in the literature.

Of particular interest is the work of Reineke and co-workers

on glycopolyamines, wherein the number of secondary

amines between the sugar groups and their basicity is

varied.^67,68They noted a distinct increase in transfection upon

raising the number of amines between the glycol units, an

observation in line with our data. Lee et al. noted a similar

trend when they tested various poly(ethylene sulﬁde)s, with

those polymers with six or eight inter-disulﬁde nitrogens

showed higher transfection than those with either two or

four.⁴²Other groups have also investigated this parameter,

however their results are clouded somewhat by the varying

degrees of branching produced by their synthetic routes,

which introduces another factor which must be taken into

account.^40,41Reineke et al. also found no correlation between

buffering capacity and the ability to transfect cells,⁶⁸indeed

their analogues based upon spermine demonstrated far higher

transfection than similar analogues incorporating ethylene

spacings despite their lower buffering capacities. Similar

results were also reported by Zhang et al. during their work

on poly(phosphoramidate)s.⁶⁹Forrest et al. varied the degree

of acetylation of PEI in order to control its buffering capacity,

observing an increase in transfection upon lowering the

In asking whether the buffering capacity can inﬂuence the

efﬁcacy of the cyclic disulﬁde for gene delivery it is

instructive to examine any differences between TriN-2,3-

CDS and TriN-3,4-CDS and between TetraN-2,3,2-CDS and

TetraN-3,4,3-CDS. When considering their optimal levels

of transfection across the two cell lines the only discernible

difference is TriN-3,4-CDS’s greater efﬁcacy in A549 cells.

It seems reasonable to assign the relatively poor performance

of TriN-2,3-CDS in this situation to the fact that it will be

less charged than TriN-3,4-CDS, potentially rendering it

unable to form polyplexes of sufﬁcient stability to transfect

this cell line. The only other trend of note is that TetraN-

(67) Liu, Y.; Reineke, T. M. Hydroxyl Stereochemistry and Amine

Number Within Poly(glycoamidoamine)s Affect Intracellular DNA

Delivery. J. Am. Chem. Soc. 2005, 127 (9), 3004–3015.

(68) Liu, Y.; Reineke, T. M. Poly(glycoamidoamine)s for Gene

Delivery. Structural Effects on Cellular Internalisation, Buffering

Capacity and Gene Expression. Bioconjugate Chem. 2007, 18 (1),

19–30.

(69) Zhang, P.-C.; Wang, J.; Leong, K. W.; Mao, H.-Q. Ternary

Complexes Comprimising Polyphosphoramidate Gene Carriers

with Different Types of Charge Groups Improve Transfection

Efﬁcacy. Biomacromolecules 2005, 6 (1), 54–60.

2054 MOLECULAR PHARMACEUTICS VOL. 7, NO. 6

Noncytotoxic AlternatiVe to Polyethylenimine

articles

ability of PEI to absorb protons at physiological pH.⁷⁰

However, while these results demonstrate that buffering

capacity is not always a crucial factor, it should be noted

that when the buffering capacity of PEI is completed ablated,

as shown by Thomas et al. when they per-alkylated PEI,

transfection does indeed suffer signiﬁcantly.⁷¹It is also

instructive to note that PL’s transfection efﬁciency can be

raised via the incorporation of histidine residues able to buffer

at physiological pH.⁷²Collectively these results, along with

our own, strongly suggest that while it is crucial for a vector

to possess some buffering capacity, it is frequently not the

dominant factor that determines efﬁcacy. Indeed recent

research has cast considerable doubt upon the validity of the

proton sponge hypothesis by suggesting that those polyplexes

that become transfectionally active are taken up into the cell

via neutral caveosomes,^73,74rendering the endolytic ability

of PEI irrelevant to its efﬁcacy.

The improvement in cytotoxicity for the cyclic disulﬁdes

compared to jetPEI was highlighted for both the polyplexes

and the free materials. When tested in CHO-K1 cells jetPEI

provoked the release of around 50% of their LDH, signiﬁ-

cantly higher than the background levels. In contrast, at N/P

ratios at which they were able to achieve statistically

indistinguishable levels of transfection (Welch’s t test, p )

0.05), neither TetraN-2,3,2-CDS nor TetraN-3,4,3-CDS

showed any evidence of cytotoxicity despite the long

incubation times used (24 h). This improvement was also

evident when the jetPEI and TetraN-3,4,3-CDS were com-

pared as free materials, with far higher concentrations of the

cyclic disulﬁde being tolerated before any LDH release above

background was observed. This effect could not be ascribed

solely to the lower density of amine functionalities found in

TetraN-3,4,3-CDS as when the results were presented as a

function of this variable its cytotoxicity remained markedly

lower implying that its bioreducible nature was central to

its improved performance.

tion. Pleasingly TetraN-3,4,3-CDS, used at an N/P ratio of

5, proved able to effect levels of transfection that were

signiﬁcantly above nontreated mice (ANOVA analysis,

Bonferroni corrected, p ) 0.12), although still below those

exhibited by in ViVo jetPEI. However when the mice were

pretreated with NAC, the levels of ꢀ-gal expression observed

with TetraN-3,4,3-CDS rose signiﬁcantly to levels that were

statistically indistinguishable from our positive control

(ANOVA analysis, Bonferroni corrected, p ) 1.0). Interest-

ingly no such response was seen with in ViVo jetPEI,

indicating that the effect may have been the result of the

temporary raising of glutathione levels in the lung rather than

the mucolytic properties of NAC. TetraN-2,3,2-CDS was also

tested, however even after pretreatment with NAC it did not

perform as well as TetraN-3,4,3-CDS or jetPEI. Histological

analysis of the transfected tissue showed that the major sites

of transfection were airway epithelial cells in the bronchioles

and bronchioli. These cells are the target for the treatment

of cystic ﬁbrosis, highlighting the potential of TetraN-3,4,3-

CDS to be used for this purpose.

In conclusion we were able to develop synthetic method-

ology to synthesize small molecule cyclic disulﬁdes, able to

form stable, biologically active polyplexes via a TCTP on

pDNA, from a range of di-, tri- and tetraamines. Those

analogues based upon tri- and tetraamine structures were able

mediate successful transfection, at the same level as jetPEI

in CHO-K1 cells, while remaining essentially nontoxic.

While they were not as effective when tested in A549 cells,

our lead compound, TetraN-3,4,3-CDS, was able to match

the ability of jetPEI to transfect murine lungs if used after

pretreatment with NAC. Initial structure-activity investiga-

tions suggested that the buffering capacity of these species

is unlikely to have a profound effect of their ability to

transfect cells, however the number of amines between the

disulﬁdes may prove to be crucial. Further investigations are

needed though before any precise relationships can be settled

upon. In addition the ability of the cyclic disulﬁdes to

transfect cells effectively in the presence of serum, which

has been reported to reduce the efﬁcacy of polyamine vectors

such as PEI, might be tested to ascertain whether they are

able to retain their performance under such conditions. The

efﬁcacy shown by TetraN-3,4,3-CDS in transfecting the

airway epithelia of the murine lungs paves the way for it to

be examined in more detail as a potential vector for the

treatment of cystic ﬁbrosis, one of the major targets for the

gene therapy community.

Having established that cyclic disulﬁdes were able to

mediate transfection in Vitro we progressed onto testing their

efﬁcacy in ViVo, speciﬁcally for transfection of murine lungs

with a ꢀ-gal containing plasmid after intranasal administra-

(70) Forrest, M. L.; Meister, G. E.; Koerber, J. T.; Pack, D. W. Partial

Acetylation of Polyethylenimine Enhances In Vitro Gene Delivery.

Pharm. Res. 2004, 21 (2), 365–371.

(71) Thomas, M.; Klibanov, A. M. Enhancing Polyethylenimine’s

Delivery of Plasmid DNA into Mammalian Cells. Proc. Natl.

Acad. Sci. U.S.A. 2002, 99 (23), 14640–14645.

(72) Midoux, P.; Monsigny, M. Efﬁcient Gene Transfer by Histidylated

Polylysine/pDNA Complexes. Bioconjugate Chem. 1999, 10 (3),

406–411.

(73) Rejman, J.; Bragonzi, A.; Conese, M. Role of Clathrin- and

Caveolae-Mediated Endocytosis in Gene Transfer Mediated by

Lipo- and Polyplexes. Mol. Ther. 2005, 12 (3), 468–474.

(74) van der Aa, M. A. E. M.; Huth, U. S.; Ha¨fele, S. Y.; Schubert,

R.; Oosting, R. S.; Mastrobattista, E.; Henninck, W. E.; Peschka-

Su¨ss, R.; Koning, G. A.; Crommelin, D. J. A. Cellular Uptake of

Cationic Polymer-DNA Complexes via Caveolae Plays a Pivotal

Role in Gene Transfection in COS-7 Cells. Pharm. Res. 2007,

24 (8), 1590–1598.

Acknowledgment. We would like to acknowledge

ﬁnancial support from the EPSRC (DTA studentship for

C.R.D.). Thanks also to Peter Haycock and Dr. Richard

Sheppard for their help with the more complex NMR

experiments.

Supporting Information Available: Additional data on

LDH, MTT and ꢀ-Gal expression assays. This material is

available free of charge via the Internet at http://pubs.acs.org.

MP9002249

VOL. 7, NO. 6 MOLECULAR PHARMACEUTICS 2055

Article Doi

Bioresponsive small molecule polyamines as noncytotoxic alternative to polyethylenimine

DOI: 10.1021/mp9002249

Source and publish data:

Authors:

Article abstract of DOI:10.1021/mp9002249

Full text of DOI:10.1021/mp9002249

Products guided by the article

R&D Labs maybe for 35378-93-9

Relevant to this article

Hot Product