Manuscript for ACS journal (Nanoletter) 03/11/03

DNA-directed Synthesis of Generation 7 and 5 PAMAM Dendrimer Nanoclusters

Youngseon Choi1, Almut Mecke2, Timothy Sassanella4, Bradford Orr2,4, Mark Banaszak Holl3,4,

and James R. Baker Jr.1,4*

1Department of Biomedical Engineering, School of Engineering, University of Michigan, Ann Arbor, MI 48109

2Department of Physics, School of Literature, Art and Science, University of Michigan, Ann Arbor, MI 48109

3Chemistry Department, School of Literature, Art and Science, University of Michigan, Ann Arbor, MI 48109

4Center for Biological Nanotechnology, Department of Internal Medicine, University of Michigan Medical School Ann Arbor, MI 48109-0648.

*Corresponding author: James R. Baker Jr., M.D. ; Ruth Dow Doan Professor and Director of Center for Biologic Nanotechnology, The University of Michigan Medical School, 9220 MSRB III, 1150 W.Med.Ctr.Dr., Ann Arbor, MI 48109-0648

ABSTRACT A novel nanostructure was constructed using two different generations of polyamidoamine (PAMAM) dendrimers and complementary 50 base pair oligonucleotides. The oligonucleotides were covalently conjugated to partially acetylated generation 5 and 7 PAMAM dendrimers, and analyzed using UV-vis spectroscopy. These conjugates were then mixed at appropriate temperatures to allow for DNA-directed self-assembly of supramolecular arrays. Overall size and shape of the DNA-linked dendrimer clusters was examined using tapping mode atomic force microscopy (AFM) to resolve the interspatial distance between the linked G7 and G5 dendrimer units. The intra-dendrimer distance was measured to be 21  2 nm, which is in agreement with the theoretical length of the 50-base long oligonucleotide pairs. In contrast, an AFM image of a mixture of non-complexed G7/G5 dendrimers only showed a few dendrimers physically in contact with a inter-dendrimer distance of 8 ~ 10 nm. In addition, a dynamic light scattering (DLS) analysis of the cluster showed a diameter of 37  4 nm, which was the expected diameter of a DNA-linked dendrimer cluster. These results suggest that PAMAM dendrimers can be self-assembled via complementary oligonucleotides to form supramolecular nanoclusters.

KEYWORDS. Oligonucleotide, DNA, PAMAM dendrimer, Self-assembly, Atomic force microscopy.

TITLE RUNNING HEAD. DNA-directed self-assembly of PAMAM dendrimers

Table of Contents Graphic

Figure. A representative three dimensional AFM image of a DNA-linked G7-G5 dendrimers on mica (80 nm  80 nm). The intra-dendrimer distance was measured to be 21  2 nm, which is in good agreement with the theoretical length of the 50-base long oligonucleotide duplex used to connect the polymers. A dynamic light scattering (DLS) measurement of a population of these clusters showed a mean diameter of 37  4 nm.

ABSTRACT.

A novel nanostructure in the size range of 40nm was constructed using two different generations of polyamidoamine (PAMAM) dendrimers and complementary 50 base pair 5’-phosphate modified oligonucleotides oligonucleotides(50-base pairs). The oligonucleotides were Ccovalent; conjugation of the synthetic oligonucleotide to partially acetylated generation 5 and 7 PAMAM dendrimers, approximately 5 nm in diameterand analyzed , were compared with a charge -based dendrimer-oligonucleotide complex using UV-vis spectroscopy. These conjugates were tested for then mixed at appropriate temperatures to allow for DNA-directed self-assembly of supramolecular arrays. Overall size and shape of the DNA-linked dendrimer clusters was examined using tapping mode atomic force microscopy (AFM) to resolve the interspatial distance between the arrayed linked G7 and G5 dendrimer units. TRemarkably, the inter-dendrimer distance was measured to be 212nm, which is in good agreement with the theoretical length of the 50-base long oligonucleotide pairs. In contrast, anComparably,a the non-complexed showed a few dendrimers physically in contact exhibited physically attached dendrimers In addition, a dynamic light scattering (DLS) analysis measurement of the cluster showed a diameter of 37nm4nm, which was the supported our expected diameter of a DNA-linked dendrimer cluster. structure These results suggest that PAMAM dendrimers can be self-assembled via complementary oligonucleotides to form supramolecular nanoclusters.

KEYWORDS. Oligonucleotide, DNA, PAMAM dendrimer, Self-assembly, Atomic force microscopy.

1

Manuscript for ACS journal (Nanoletter) 03/11/03

Development of two and three-dimensional assemblies of nanoscale molecules with controlled size and shape is a great challenge for current materials chemistry and nanotechnology1. DRecently, due to their precisely controlled mass, surface valency, and surface functionality2,3, PAMAM dendrimers have been employed as molecular components of new mathematically defined, nanoscale dendrimer clusters 4. In these structures, charge interaction neutralization was employed as a basis for forming covalent amide bonds to link the components. However, this approach makes the development assembly of complex molecular arrays different dendrimer modules technically difficult since specific size ratios and harsh chemistry are necessary to coordinate the self assembly core and shell dendrimers. There is also no specificity for the coupling of dendrimers so that it is difficult to assemble specific molecules. Thus, alternative approaches to assembling dendrimers into supramolecular clusters need to be evaluateddeveloped.

Synthetic oligonucleotides have been introduced as a tool to self-assemble molecules and nanoscale objects in precise structural arrangements due to the base specificity of the resulting duplex structure. S s, others 5. . Mirkin, Schultz and their colleaguesMany attempts have reportedbeen made regarding attaced ment of hing ooligonucleotides to gold nanoparticles to direct the formation of periodic nanostructures, which can be reversibly assembled6,7. In contrast, although charge-based PAMAM-oligonucleotide complexes have been employed for the delivery of antisense oligonucleotides and gene transfection in vitro, no one has used covalently bound PAMAM-oligonucleotide conjugates as a means for generating specific supramolecular clusters8,9,10,11

In this study, we attempted to specifically link dendrimers using complementary DNA strands. These clusters were then examined at the nanometer scale by AFM to evaluate inter-dendrimer distance in the a cluster and ______assembly of different components. The resulting DNA-linked dendrimer clusters can be were then employed for the development of biocompatible supramolecular devices that would be biocompatible with different functional ______. and useful in therapeutic applications.

An initial challenge involved linking the negatively charged oligonucleotide to the positively charged amine-terminated PAMAM dendrimers. To minimize the inherent electrostatic interactions between the PAMAM dendrimers and the oligonucleotides, partial acetylation of surface primary amines of dendrimers was performed. approach This would help increase the conjugation efficiency as indicated by Niemeyer et al12. We also expected that limiting the number of available surface primary amines cwould help control the number of oligonucleotides attached on each the dendrimer. However, considering the dense-packed dendrimer surface structure and the potential difficulty of oligonucleotide strand access to dendrimer surface primary amines, acetylation was limited to 90% of the amines, so that on average 12 NH2 groups for per G5 dendrimer would be available for conjugation. For a G7 dendrimer, 89% (48386 NH2 groups out of 434 amines) were acylated, while the . The smaller (G5) dendrimer (G5) was relatively more acetylated (X% or Y of 2 amines) than the larger dendrimer (G7), in orderattach fewer oligonucleotides to assure that G5 dendrimers would have fewer oligonucleotides to limit cross-hybridizationlinking. The molecular characteristics of each of the acetylated dendrimers are summarized in Table 1. To test the working hypothesis of DNA-directed self-assemblyThe , these ppartially acetylated G7 and G5 PAMAM dendrimers were then conjugated to 50 mer complementary oligonucleotides designed with 30 base-pairs overlaps. and 20 base-pairs overhang. After removal of non-conjugated oligonucleotides, Then, the hybridization of the resulting ssDNA-conjugated G7 and G5 dendrimers was performed (Table 2). Hybridization was carried out in excess of G5 to prevent cross-linking and the formation of very large complexes. Ooverall size and shape of the oligonucleotides desired G7-G5 PAMAM dendrimer self-assembly are schematically displayed in Figure 1.

Covalent bond formation between primary amino groups of dendrimers and phosphate-modified 5’ end of ssDNA was characterized by UV-vis spectroscopy as shown in Figure 2. Compared to the spectra of either oligonucleotide or dendrimer, itself, the oligonucleotide-dendrimer conjugates exhibited a slight shift to longer wavelength of in the DNA absorption peak. A In contrast, a mixture of DNA and dendrimer that was prepared at the same concentration of the ______complex did not show any spectral shift. Even with an increase in ssDNA molar concentration of 5-fold vs. dendrimer, the UV spectra showed an increase in the 260 nm without a corresponding peak shift. (data not shown). This finding was supportedis consistent with the work of by Ottaviani et al.13, who found that the electrostatic association of dendrimer and DNA did not cause a significant change in the UV spectra. Thus, our result suggests that the attachment of ssDNA to dendrimer occurred via a covalent phosphoramidate bond formation. The average number of oligonucleotides attached to the G7 and G5 PAMAM dendrimers was calculated to be 5 and 2 respectively, using the sextinction coefficient of the oligonucleotides at 260 nm.

Initial attempts to hybridize the two ssDNA-dendrimer conjugates used an excess of G7-P1 conjugate (G7-Ac89-P15; 190,299g/mol, 37mol) that was mixed with G5-P2 conjugate (G5-Ac90-P22; 59,867g/mol, 1.2mol) in a 100 mM Tris-EDTA (define TE)TE buffer (pH 8.0, 10mM NaCl). We hypothesized that with for these given conjugates having different numbers of ssDNAs on the dendrimer surface (five linker arms for G7 and two linker arms for G5), limiting the amount of G5-P2 conjugate would statistically prevent network formation between the multivalent dendrimers, yielding instead a dimeric or a trimeric cluster. For hybridization, the reaction mixture was heated above the melting temperature of the oligonucleotides (90°C) for 10 min and then allowed to cool to room temperature over 3 hours, followed by ultrafiltration using a 100k molecular weight cut-off (Centricon YM-100, Milipore®). This purification step removes un-hybridized G5-P2 conjugate while retaining the G7-P1 conjugate and the resulting DNA-linked G7-G5 clusters.

Tapping-mode atomic force microscopy (AFM) images were obtained using a previously reported procedure14,15. Images obtained taken after this ultrafiltration purification showed most of the dendrimers are linked together in consistent distances of 21nm2 nm (Figure 3). All measurements were performed using the previously reported procedure14,15. Figures 3 and 4 show examples of AFM images of dendrimer-oligonucleotide assemblies adsorbed to a mica substrate. Some of the larger objects appeared to be features are most likely aggregates of many individual dendrimers and/or dendrimer clusters. However, there are a number of smaller, dumbbell-shaped clusters whose dimensions are consistent with the ones expected for individual DNA-linked dendrimers clusters.

AFM image analysis of twelve objects (A to L) regarding maximum and minimum bump height, total volume and horizontal distance between the dendrimers is summarized in Table 3. Considering the molecule asymmetry inherent in the DNA-linked dendrimer cluster structure, numerical integration of each feature is performed using the bearing analysis of the Nanoscope III software to yield accurate volume calculations14. Based on a molecular weight and a density of 1.2 g/cm3, the theoretical volume of the smallest possible dendrimer cluster between only one G7 and one G5 PAMAM is 280 nm3. Remarkably, the average volume (286 ± 80 nm3) of the twelve objects seems to suggest a possible assembly of two or three dendrimers, which is close to the theoretical volume for dimeric dendrimer self-assembly. Note that the very low aspect ratio of these topographical features helps to minimize tip convolution effects in the volume measurement.14,15 Many of the particles such as M and N have larger volumes, indicating that more than two dendrimers with several oligonucleotides attached have assembled into a cluster.

Moreover, the distance between interthe -linked particles are approximately close to 20 nm in good agreement with the theoretical length of partially hybridized P1-P2 oligonucleotides as schematically shown in Figure 1. A theoretical length of the oligonucleotide was calculated using 4.3A and 3.4A per base for single-stranded DNA and double-stranded DNA, respectively16. The size of an amine-terminated PAMAM dendrimer is approximately 8 nm for G7 and 5 nm for G5 in diameter from according to light scattering and transmission electron microscopy (TEM) measurements17,18. However, the size of partially acetylated dendrimers was found to be used is found out to be 5.0 nm for G7 and 4.7 nm for G5 from SEC analysis (Table 1), which can be explained by a decreasing tendency of G5 dendrimer size according to the increase of acetylation19. Particles with the characteristic topology expected for a dimeric or trimeric G7-G5 dendrimer cluster have been imaged in Figure 4. The three-dimensional image of a dimeric single cluster shows that the total volume is consistent with two oligonucleotide-linked G5 and G7 dendrimers, while the inter-dendrimer distance agrees with length of the P1-P2 oligonucleotides pairs (Figure 4b). Despite some clumping and distortions of the complex morphology due to drying of the dried samples in air, the AFM images clearly demonstrate a variation variety of structures consisting of G7 and G5 dendrimer self-assemblies. Figure 5a exhibits a selection of representative supramolecular array of PAMAM dendrimers, where dimeric or trimeric dendrimer self-assemblies having a consistent inter-dendrimer distance and a total volume ranging from 200 ~ 400 nm3.

65bInterestingly, a few dimer-like G5 or G7 dendrimers appear like DNA-linked structures, however a section analysis indicated the two dendrimers are so close to each other with on average 8 ~ 10 nm (See Supporting Information). This control experiment clearly suggested that the an assembly technique via DNA is necessary for a rational assembly of PAMAM dendrimers.DFurthermore, a supporting information aside from the AFM results was obtained from a dynamic light scattering (DLS) analysis which alloweds for the measurement of the DNA-linked dendrimer clusters size in solution. Compared to the measured average diameter s of the partially acetylated dendrimers (below 10 nm) and the oligonucleotide-conjugated G7 and G5 dendrimers (10 ~ 27 nm), t. T, the DNA-linked cluster showed a diameter of 37 nm4 nm (Table 4). In spite of the limitations of the DLS measurements are limited because they , which assume each s the particle is a sphere. Despite this, these results are in good agreement with AFM image analysis of Figure 3, where the average length of all 56 objects is 34.1 nm. Also, it is not surprising that the dendrimer-DNA conjugates show a little variation in size considering the chain flexibility of the oligonucleotides attached on the dendrimers. However, a predominant population peak of the conjugates falls in between the peak of the dendrimers (5 ~ 10 nm range) and DNA-linked dendrimer clustersa is present . A few clusters forming large aggregates with some uniformity in configuration could indicate problems with purification of the clusters or in stoichiometric control over the number of oligonucleotides on the dendrimers, which should be further investigated.

In conclusion, these results demonstrate that the design of simple and effective systems for DNA-linked dendrimers can be achieved. These unique structures have a number of advantages as supramolecular arrays such as self-assembly at low temperature without additional reactants and the possible assembly of specific dendrimer conjugates that function in fluorescence imaging, cell-receptor targeting20, apoptosis-detection, or as a therapeutic drug21-23. Further studies are underway to use the topographical AFM information as a means on the novel nanostructures based on different conjugation protocols and hybridization reaction conditions in order to obtain higher purity and yield.

ACKNOWLEDGMENTS. We thank the National Cancer Institute for their support on this research under the Contract # NOI-CO-97111. We a

Athank is greatly acknowledged

References:

Majoros, I. K., B.; Woehler, S.; Bull, T.; Baker, J. R. Macromolecules 2003.(1)Storhoff, J. J.; Mirkin, C. A. Chemical Reviews 1999, 99, 1849-1862.

(2)Tomalia, D. A.; Naylor, A. M.; Goddard, W. A. Angewandte Chemie-International Edition in English 1990, 29, 138-175.

(3)Tomalia, D. A. Advanced Materials 1994, 6, 529-539.

(4)Uppuluri, S.; Swanson, D. R.; Piehler, L. T.; Li, J.; Hagnauer, G. L.; Tomalia, D. A. Advanced Materials 2000, 12, 796-800.

(5)Waybright, S. M.; Singleton, C. P.; Wachter, K.; Murphy, C. J.; Bunz, U. H. F. Journal of the American Chemical Society 2001, 123, 1828-1833.

(6)Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607-609.

(7)Alivisatos, A. P.; Johnsson, K. P.; Peng, X. G.; Wilson, T. E.; Loweth, C. J.; Bruchez, M. P.; Schultz, P. G. Nature 1996, 382, 609-611.

(8)Bielinska, A.; KukowskaLatallo, J. F.; Johnson, J.; Tomalia, D. A.; Baker, J. R. Nucleic Acids Research 1996, 24, 2176-2182.

(9)KukowskaLatallo, J. F.; Bielinska, A. U.; Johnson, J.; Spindler, R.; Tomalia, D. A.; Baker, J. R. Proceedings of the National Academy of Sciences of the United States of America 1996, 93, 4897-4902.

(10)DeLong, R.; Stephenson, K.; Loftus, T.; Fisher, M.; Alahari, S.; Nolting, A.; Juliano, R. L. Journal of Pharmaceutical Sciences 1997, 86, 762-764.

(11)Richardson, S. C. W.; Pattrick, N. G.; Man, Y. K. S.; Ferruti, P.; Duncan, R. Biomacromolecules 2001, 2, 1023-1028.

(12)Niemeyer, C. M.; Burger, W.; Hoedemakers, R. M. J. Bioconjugate Chemistry 1998, 9, 168-175.

(13)Ottaviani, M. F.; Furini, F.; Casini, A.; Turro, N. J.; Jockusch, S.; Tomalia, D. A.; Messori, L. Macromolecules 2000, 33, 7842-7851.

(14)Betley, T. A.; Holl, M. M. B.; Orr, B. G.; Swanson, D. R.; Tomalia, D. A.; Baker, J. R. Langmuir 2001, 17, 2768-2773.

(15)Betley, T. A.; Hessler, J. A.; Mecke, A.; Holl, M. M. B.; Orr, B. G.; Uppuluri, S.; Tomalia, D. A.; Baker, J. R. Langmuir 2002, 18, 3127-3133.

(16)Tinland, B.; Pluen, A.; Sturm, J.; Weill, G. Macromolecules 1997, 30, 5763-5765.

(17)Li, J.; Piehler, L. T.; Qin, D.; Baker, J. R.; Tomalia, D. A.; Meier, D. J. Langmuir 2000, 16, 5613-5616.

(18)Jackson, C. L.; Chanzy, H. D.; Booy, F. P.; Drake, B. J.; Tomalia, D. A.; Bauer, B. J.; Amis, E. J. Macromolecules 1998, 31, 6259-6265.

(19)Majoros, I. K., B.; Woehler, S.; Bull, T.; Baker, J. R. Macromolecules 2003.

(20)Esfand, R.; Tomalia, D. A. Drug Discovery Today 2001, 6, 427-436.

(21)Patri, A. K.; Majoros, I. J.; Baker, J. R. Current Opinion in Chemical Biology 2002, 6, 466-471.

(22)Patri, A. K.; Thomas, T.; Baker, J. R.; Bander, N. H. Abstracts of Papers of the American Chemical Society 2002, 223, D62-D62.

(23)Quintana, A.; Raczka, E.; Piehler, L.; Lee, I.; Myc, A.; Majoros, I.; Patri, A. K.; Thomas, T.; Mule, J.; Baker, J. R. Pharmaceutical Research 2002, 19, 1310-1316.

(24)Dvornic, P. R.; Tomalia, D. A. Macromolecular Symposia 1995, 98, 403-428.

(25)Chu, B. C. F.; Wahl, G. M.; Orgel, L. E. Nucleic Acids Research 1983, 11, 6513-6529.

(1)Storhoff, J. J.; Mirkin, C. A. Chemical Reviews 1999, 99, 1849-1862.

(2)Tomalia, D. A.; Naylor, A. M.; Goddard, W. A. Angewandte Chemie-International Edition in English 1990, 29, 138-175.

(3)Tomalia, D. A. Advanced Materials 1994, 6, 529-539.

(4)Uppuluri, S.; Swanson, D. R.; Piehler, L. T.; Li, J.; Hagnauer, G. L.; Tomalia, D. A. Advanced Materials 2000, 12, 796-800.

(5)Waybright, S. M.; Singleton, C. P.; Wachter, K.; Murphy, C. J.; Bunz, U. H. F. Journal of the American Chemical Society 2001, 123, 1828-1833.

(6)Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607-609.

(7)Alivisatos, A. P.; Johnsson, K. P.; Peng, X. G.; Wilson, T. E.; Loweth, C. J.; Bruchez, M. P.; Schultz, P. G. Nature 1996, 382, 609-611.

(8)Bielinska, A.; Kukowska-Latallo, J. F.; Johnson, J.; Tomalia, D. A.; Baker, J. R. Nucleic Acids Research 1996, 24, 2176-2182.

(9)Kukowska-Latallo, J. F.; Bielinska, A. U.; Johnson, J.; Spindler, R.; Tomalia, D. A.; Baker, J. R. P Natl Acad Sci USA 1996, 93, 4897-4902.

(10)DeLong, R.; Stephenson, K.; Loftus, T.; Fisher, M.; Alahari, S.; Nolting, A.; Juliano, R. L. Journal of Pharmaceutical Sciences 1997, 86, 762-764.

(11)Richardson, S. C. W.; Pattrick, N. G.; Man, Y. K. S.; Ferruti, P.; Duncan, R. Biomacromolecules 2001, 2, 1023-1028.

(12)Niemeyer, C. M.; Burger, W.; Hoedemakers, R. M. J. Bioconjugate Chemistry 1998, 9, 168-175.

(13)Ottaviani, M. F.; Furini, F.; Casini, A.; Turro, N. J.; Jockusch, S.; Tomalia, D. A.; Messori, L. Macromolecules 2000, 33, 7842-7851.

(14)Betley, T. A.; Holl, M. M. B.; Orr, B. G.; Swanson, D. R.; Tomalia, D. A.; Baker, J. R. Langmuir 2001, 17, 2768-2773.

(15)Betley, T. A.; Hessler, J. A.; Mecke, A.; Holl, M. M. B.; Orr, B. G.; Uppuluri, S.; Tomalia, D. A.; Baker, J. R. Langmuir 2002, 18, 3127-3133.

(16)Tinland, B.; Pluen, A.; Sturm, J.; Weill, G. Macromolecules 1997, 30, 5763-5765.

(17)Li, J.; Piehler, L. T.; Qin, D.; Baker, J. R.; Tomalia, D. A.; Meier, D. J. Langmuir 2000, 16, 5613-5616.

(18)Jackson, C. L.; Chanzy, H. D.; Booy, F. P.; Drake, B. J.; Tomalia, D. A.; Bauer, B. J.; Amis, E. J. Macromolecules 1998, 31, 6259-6265.