From G-Quartet to G-Quadruplex and its Nanoarchitectures

Chapter Editor: Lea Spindler

In 1910 Ivar Bang reported that concentrated solutions of guanylic acid formed a gel. It took half a century before Gellert et al. in 1962 discovered the structural motif, a guanine-quartet, to be the basis for guanylic acid gelation. This chapter starts with a historical overview and gives credit to Ivar Bang and several other pioneers in the field. Then the basic principles of guanine self-assembly are explored, either for individual molecules like guanosine and its derivatives, or for guanine-rich oligonucleotides and G-rich DNA-sequences. Finally, we show how these guanine-based nanoarchitectures are visualised by surface techniques like scanning tunnelling microscopy or atomic force microscopy.

GUANYLIC ACID SELF-ASSEMBLY: 100 YEARS LATER

Gang Wu

Department of Chemistry, Queen's University, Kingston, Ontario, Canada K7L 3N6

1 INTRODUCTION

It is commonly accepted that the work of Gellert, Lipsett, and Davies published in 1962 marks the beginning of modern research on G-quartet related molecular systems. Interestingly, the opening sentence of this seminal paper stated, "In 1910, Bang reported that concentrated solutions of guanylic acid formed a gel." This referred to a paper published half of a century earlier. They then described two structural models used to explain the experimental X-ray diffraction patterns obtained for guanylic acid fibers. The fibers were drawn from two different types of guanylic acid gels: one from concentrated aqueous solutions of guanosine 3'-monophosphate (3'-GMP) under an acidic condition (pH 5) and the other from the 5' isomer, 5'-GMP, under a similar condition. The central structural building block proposed for the 3'-GMP gel consists of four hydrogen bonded guanine bases in a planar fashion, which is now known as a G-quartet or G-tetrad. The structural model proposed for the 5'-GMP gel was initially ambiguous, due to the limited quality of the X-ray data. Later, Davies and co-workers showed that the correct structural model for 5'-GMP gels obtained at pH 5 consists of hydrogen bonded guanine bases that form a continuous helix similar to the shape of a lock-washer. These two models are illustrated in Figure 1.

In the past 20 years, research activities in the field of G-quartet related molecular systems have grown exponentially. Recent discoveries of the existence of the G-quartet motif in many biologically important systems such as telomeres, promoters of many genes, and sequences related to various human diseases have triggered tremendous research interest in this unusual type of nucleic acid structures. Now G-quartet structures can be found in such diverse areas as molecular biology, medicinal chemistry, supramolecular chemistry, and nanotechnology. It is interesting to note that, as the field of G-quartet research expands in recent years, one can also find increasingly more references to the aforementioned 1910 article by Bang. Unfortunately, many citations are either inaccurate in terms of what exactly Bang reported in the 1910 paper or completely wrong in terms of who Bang was. For example, he was sometimes misidentified as being a German biochemist (obviously his 1910 paper was published in Biochemische Zeitschrift and indeed in German). One may wonder how many modern researchers in the field have actually read Bang's 1910 article. It seems, however, that all we can learn from the modern literature is that Bang somehow noticed gelation of guanylic acid under certain conditions, but we cannot find any clue as to the context he made this important observation. So who was Bang and why did he study guanylic acid in the first place? In this article, we set out to examine briefly the history centered on the discovery of guanylic acid (1894-1912), then discuss some early results on its self-assembly properties/structures (1962-1990), and finally describe some recent progress related to guanylic acid (since 2000). The main objective of this article is to provide the reader with a sense of history and a big picture about what we have learned about this fascinating molecule and its self-assembly properties over a period of 100 years.

Here we should credit the title of this article to several previous workers in the field. In particular, Guschlbrauer, Chantot, and Thiele published the first review article on the subject in 1990 which was entitled "Four-stranded nucleic acid structures 25 years later: From guanosine gels to telomer DNA". In 2004, Davis wrote the most popular review in the field with a similar title: "G-quartet 40 years later: From 5'-GMP to molecular biology and supramolecular chemistry". Thus it seems appropriate for us to follow this good tradition, except that, as the reader may have already noticed, we extend the beginning of our time line to Bang's 1910 discovery.

2 THE BEGINNING

In 1894, Olof Hammarsten (1841-1932) isolated from ox pancreas a substance that he called β-nucleoprotein. The non-protein portion of β-nucleoprotein was given the name of guanylic acid, because on hydrolysis it releases an excessive amount of guanine but no trace of other bases such as thymine and cytosine. This appears to be very different from other nucleic acids known at the time. From 1897-99, Ivar Christian Bang (1869-1918, a Norwegian physician and clinical chemist), who had obtained his medical degree two years earlier from Oslo, joined Hammarsten's Physiological Chemistry Laboratory in Uppsala, Sweden, as a trainee (perhaps modern equivalent of a postdoctoral fellow). In the next decade, with a two-year interruption (1900-1902) when he practiced medicine in Oslo, Bang attempted to elucidate the chemical structure of guanylic acid. He spent a great deal of effort trying to obtain a reasonably large quantity of pure guanylic acid for chemical analysis, which turned out to be particularly difficult. In the meantime, several other groups were also studying guanylic acid." Because of the difficulties involved in producing pure, crystalline guanylic acid for proper chemical analysis, different researchers obtained different results, triggering a lively debate that lasted more than 10 years.

Bang's 1910 paper was the last and also most comprehensive one of the series of papers that he published on the subject of guanylic acid (afterwards he turned his attention to some other problems in clinical chemistry and died suddenly from coronary occlusion in 1918 at age of 49). In this lengthy paper (18 pages), he correctly identified for the first time the chemical compositions of guanylic acid as being composed of guanine, pentose, and phosphoric acid in equal molar amounts. However, because his elemental analysis was inaccurate, he was unable to account for an unknown residue of C4H10O2. In this work, he also described an observation that, after a concentrated guanylic acid solution was neutralized with KOH, re-acidifying the solution with acetic acid led to gel formation. This kind of gel formation was commonly known for thymus nucleic acids at that time. For this reason, Bang argued forcefully (sometimes filled with personal anger) but incorrectly against the view of other contemporary researchers who considered guanylic acid to be a simple nucleotide just like inosinic acid, a mononucleotide discovered by Liebig in 1848 from beef booth. This latter opinion was clearly expressed by Albrecht Kossel (1853-1927, a German biochemist) in his Nobel lecture presented in December of 1910, in which, after describing the complex nature of nucleic acids, he stated, "The composition of inosinic and guanylic acids is still simpler."

The controversy around the chemical nature of guanylic acid was finally settled in 1912 by an American biochemist, Phebus Aaron Levene (1869-1940) who had worked on guanylic acid since 1901. In 1909, Levene successfully obtained crystalline guanosine, a guanine nucleoside, through hydrolysis of guanylic acid. So he was well aware of the fact that, upon cooling, a hot guanosine aqueous solution will form a gel in the presence of a small amount of K+ ions; see Figure 1(e). In a paper published in 1912, he argued, "However, guanosine — a simple guanine-pentoside — shares with guanylic acid the property of gelatinizing when it contains only slight proportion of mineral impurity." Therefore, gel formation alone was insufficient to disprove the simple nature of guanylic acid, as Bang had attempted to do. To address the chemical composition problem, Levene performed a more accurate elemental analysis that showed unambiguously the correct chemical formula for guanylic acid to be C10 H14N5PO8 with a structure shown in Figure 2. Now we know that this particular structure corresponds to a slightly different isomer, 3'-GMP, rather than the most common one, 5'-GMP. However, this work was a truly remarkable achievement, considering the following three facts: (1) X-ray was discovered by Röntgen in 1895; (2) Its diffraction effect on crystals was discovered by Friedrich, Knipping, and Laue in 1912, the same year as the publication of Levene's work; and (3) Of course, nuclear magnetic resonance (NMR) was not discovered until 1945. It is perhaps very hard for those students in my 2nd year organic spectroscopy course to grasp the concept of chemical analysis or structural elucidation without modern spectroscopic techniques. It is also amazing to see how science progresses over time. Thus, as for the chemical nature of guanylic acid, Bang was unfortunately incorrect. However, very few people in the field of G-quartet research are aware of the fact that Bang is actually best known for his pioneering contributions to clinical chemistry; he was considered to be the founder of modern clinical chemistry. For example, he invented the technique for analyzing blood sugar levels on a microscale. To return to the subject on hand, now what is the structural basis for guanylic acid gelation? The answer to this question would have to wait for another 50 years.

3 EARLY STUDIES

Since 1955, David Davies at the National Institute of Health had been studying RNA structures using X-ray fiber diffraction. During an effort to prepare polyguanylic acid, poly G, he and his co-workers accidentally discovered the G-quartet motif in 1962. Davies described this serendipitous discovery in his autobiography published in 2005: "Marie [Lipsett] originally thought that she had been able to make poly G but was then disappointed to discover that what she had was unpolymerized GMP that was forming a viscous solution that looked just like DNA." Recall that this was exactly what Bang had seen 50 years earlier! Then Davies continued, "As soon as she told me this I rushed over and pulled some fibers that gave diffraction patterns that could be explained by the formation of G-quartets." Soon after, Iball, Morgan, and Wilson reported that deoxyguanosine nucleosides and nucleotides also form similar helices. Many guanine nucleosides and nucleotides were subsequently found to be able to form gels containing similar helices.

In 1972, based on infra-red (1R) spectroscopic evidence, Todd Miles at the NIH reported a new helical structure formed by 5'-GMP in neutral or slightly basic solution (pH 7~8) in the presence of Na+ ions. As 5'-GMP behaves as a regular liquid under such conditions, it is suitable for solution NMR studies. Thus he approached his NMR colleague at the NIH, Ted Becker. At that time, Becker' lab was equipped with one of the early high-field 220 MHz NMR spectrometers. Together, they established that the planar G-quartet motif is responsible for the 5'-GMP helix formation in neutral solution. Years later, Becker described this discovery in Encyclopedia of Nuclear Magnetic Resonance: "Tom Pinnavaia came from Michigan State University to spend his sabbatical leave in my lab in 1974-75. He discovered that guanosine-5'-monophosphate (GMP) in neutral or basic solution self-associates by hydrogen bonding to give a stable structure in which the H-bonds exchange only slowly on the NMR timescale. We were able to interpret the NMR, along with ancillary infrared spectra data, in terms of an unexpected tetrameric structure." Remarkably, the 5'-GMP self-assembly process in neutral solution exhibits great sensitivity towards the type of alkali metal ions present in solution. As shown in Figure 3, while Li and Cs ions are inactive, Na+, K+ and Rb appear to promote different self-assembled structures. The very clean four-peak pattern (α, β, γ, and δ signals for H8) observed for Na2(5'-GMP) strongly suggests the presence of well-defined aggregates, yet the proper interpretation of this spectral pattern had remained illusive for more than 30 years until recently (vide infra). In 1976, Zimmerman confirmed the helix formation of Na2(5'-GMP) under neutral conditions with X-ray fiber diffraction data. In the several years that followed, 1H, 13C and 31P NMR techniques were used to gain insights into the various aspects of this new type of 5'-GMP self-assembly. In the meantime, Laszlo and co-workers used 23Na NMR to probe ion binding in the 5'-GMP self-assembly. In the 1990s, Gottarelli, Spada, Mariani and their co-workers reported extensive studies on the various liquid crystal phases formed from guanylic acid and related derivatives.