Generally, viruses are packages for minimal genomes Viruses need to carry in their genomes only a minimum of information: to construct particles that can enter a cell, replicate their nucleic acid and essential proteins, and then exit the cell. In many cases some information is taken from the host genome; for example, many viruses use host polymerases to replicate their genomes.
This was what we knew for years. But with results from new methodologies appearing, our views are subject to major changes. First, we now recognize that phages are the most abundant and diverse biological entities on the planet, they outnumber bacteria by about tenfold in certain ecosystems. They are found everywhere, from the ocean and its basement to the human gastrointestinal tract. According to the nucleotide database of the National Center for Biotechnology Information (NCBI), we, as of September 2019, have the complete genome sequences of about ~8500 phages. Phages have a wide range of genome sizes, the smallest being ~2500 bp. The so-called jumbo phages contain >200 kb, carrying genes involved in replication and nucleotide metabolism. Finally, a new group of phages, called megaphages, have genomes >540 kb; they were uncovered from human and animal guts.
What is at the center of the new methodologies? First and foremost, viral metage nomics. The method assesses the total viral nucleic acids, mostly dsDNA, isolated from a specific environment, including the human gut. The methodology, as useful as it is, has a major drawback: it lacks the high resolution needed to reconstruct genomes of closely related species cohabiting the same environment. The drawback is overcome by single-virus genomics, which sorts individual phage particles prior to sequencing their genomes. High levels of microdiversity were revealed within phage populations. Finally, these two methods are complemented by the use of transmission electron microscopy, revealing an astounding number of different morphologies, with capsid (see below) structures of various shapes (e.g., polyhedral, filamentous, pleomorphic) or connected to a tail.
Viral genomes come in a remarkable variety of forms. The polynucleotide may be DNA or RNA, and it may be single-stranded or double-stranded. The protein shell that surrounds the viral DNA or RNA, known as the capsid, is just sufficient to contain the genome, to permit the virus to get into a host cell, and to allow the newly replicated viruses within to escape to infect other cells. These requirements are met in a host of ways in different viruses, each using minimal machinery. As shown in the example of bacteriophage φ29 (Figure 1), some viruses are constructed in a way that allows them to inject their DNA into a host cell; in these viruses, the DNA is tightly coiled within the capsid, and energy for injection is stored in the bending of the DNA and in electrostatic repulsion between DNA phosphates. This means that energy must be supplied to force the DNA into the capsid, for example, by ATP hydrolysis.
Fig1. φ29 bacteriophage-like viruses. (A) Non-enveloped, head–tail structure. The tail is noncontractile and has a collar plus 12 tail fibers. The genome is double-stranded linear DNA of ~16–20 kb, encoding between 20 and 30 genes. (B) Typical cryo-EM images of mature φ29 particles at ~50,000-fold magnification. (C) Three-dimensional reconstruction of the full-length φ29 packaged genome. The degree of order increases from the center toward the capsid, with six concentric DNA shells visible in longitudinal and transverse sections of the volume. (D) Monte Carlo simulation of DNA packaging in φ29: consecutive images show stages at which 30%, 50%, 70%, and 100% of the genome has been packaged. The chain is colored blue at the end entering the confinement and red at the free end within the capsid at full packaging. The capsid is not shown. [A, adapted from ViralZone, Swiss Institute of Bioinformatics, Lausanne, Switzerland. B–D, from Comolli LR, Spakowitz AJ, Siegerist CE et al. (2008) Virology 371: 267–277. With permission from Elsevier.]
Other viruses, especially those that attack eukaryotic hosts, use a completely different method: they have an external membrane (envelope) that can fuse with the host’s cell membrane to ensure internalization of the viral genome. The proteins or other molecules needed to accomplish this are synthesized by the host cell, using genetic instructions from the viral genome. The compaction of the nucleic acid to virus-compatible dimensions is in this case accomplished by the viral coat proteins, together with the presence of positive ions or small molecules bound to the DNA that help to neutralize the polynucleotide charge. The structural organization of the RNA genome of the human immunodeficiency virus, HIV, as a representative example of such enveloped viruses is presented in Figure 2. Many RNA viruses, including those that cause influenza and the common cold, are constructed in this fashion.
Fig2. HIV and organization of its genome. (A) Anatomy of HIV. The outer coat of HIV is a layer of cell membrane stolen from the infected cell upon viral particle release. The coat is studded with protein spikes composed of two glycoproteins, gp120 and gp41, which serve to dock the virus onto receptor proteins on the surface of the host cell. This membrane layer envelops the matrix and capsid struc tures that encase the genetic material. The matrix proteins form a coat on the inner surface of the viral membrane and play a central role when new viruses bud from the surface of infected cells. The capsid proteins form a cone-shaped coat around the viral RNA, delivering it into the cell during infection. The proteins that form these structures are colored according to the color scheme representing the primary structure of the genome in part B. Three kinds of proteins that are essential for viral replication—reverse transcriptase (RT), integrase, and protease—are encoded by viral RNA and are present in the virions or mature viral particles. The HIV replication cycle is covered in more detail in Chapter 20. (B) Primary organization of the genetic material along the RNA molecule. (C) The secondary structure of the viral RNA is highly complex and is further coiled in three dimensions. The coiling is so dense that it was practically impossible to see any detail by electron microscopy. [C, adapted from Watts JM, Dang KK, Gorelick RJ et al. (2009) Nature 460: 711–716. With permission from Macmillan Publishers, Ltd.]
Finally, and in view of the current pandemic of coronavirus infections, we present the present knowledge of the viral structure, its genome and proteins, and its replication cycle (Box 8.1).
Bacterial chromosomes are organized structures in the cytoplasm
Most but not all bacteria possess a single, larger circular chromosome, often accompanied by one or more small circular plasmids. The large chromosome is free in the cell cytoplasm, as bacteria have no morphologically defined nucleus. The cytoplasmic complex of DNA and proteins is called the nucleoid (Figure 3). At a global scale, the bacterial chromosome is spirally folded to fit within the bacterial cell. For Escherichia coli, the unfolded circumference of the duplex DNA circle would be ~1.6 mm, whereas the dimensions of the bacterium are around 2 μm. This corresponds to a packing ratio of about 1000. Regions of the chromosome sequentially close to each other interact in three-dimensional space. At a scale of ten to hundreds of kilo bases, the bacterial chromosome is further partitioned into chromosome interaction domains (CIDs). CIDs exhibit a high degree of self-interaction and are insulated from flanking regions.
Fig3. The bacterial nucleoid. (A) E. coli under the electron microscope. The cell has just undergone DNA replication and consequently has two nucleoids. (B) Transmission electron microscopy, TEM: osmium fixation of E. coli shows a confined nucleoid containing a network of randomly oriented DNA fibers. (C) Schematic representing one of the four DNA topological macrodomains in the nucleoid of E. coli and an adjacent less-structured region. Large portions of the circular chromosome occupy the same cellular space, forming macro domains that are separated by less-structured regions; these regions prevent the domains from colliding. The DNA in the macrodomains is highly packed through interactions with specific protein factors, shown as green squares. [A, from Menge B & Wurtz M. Biozentrum, University of Basel/Science Photo Library. B, from Eltsov M & Zuber B (2006) J Struct Biol 156: 246–254. With permission from Elsevier. C, adapted from Boccard F, Esnault E & Valens M (2005) Mol Microbiol 57: 9–16. With permission from John Wiley & Sons, Inc.]
A major factor contributing to compaction is supercoiling of the bacterial DNA. The bacterial DNA is highly negatively supercoiled. The equilibrium superhelical density is about –0.06; given a genome size of 4.6 × 106 bp, or 4.6 × 105 turns of DNA, this indicates about 3 × 104 superhelical twists. This can be seen in vitro as a highly branched plectonemic writhing of the DNA (see Figure3). Whether this is the exact mode of supercoiling in vivo is unknown. The supercoiling is created and maintained by several factors. These factors include the binding of structural proteins that are a part of the bacterial nucleoid, processes such as transcription and replication, and the actions of gyrases, which are topoisomerases that introduce negative supercoiling into topologically con strained DNA molecules. Other topoisomerases regulate the level of superhelical stress.
DNA-bending proteins and DNA-bridging proteins help to pack bacterial DNA
A partial list of bacterial nucleoid-associated proteins is given in Table 1. The most extensively studied of these are the DNA-bending proteins heat unstable (HU) protein, and integration host factor (IHF), both of which exist in slightly postsynthetically modified forms and function as heterodimers. HU and IHF are structurally very similar, and as minor-groove binders, they are capable of inducing very sharp bends and kinks in DNA. Recent evidence shows that both HU and IHF, in high concentrations, can induce helical rodlike structures in DNA. Examples from X-ray and atomic force microscopy (AFM) imaging studies are shown in Figure 4.
Table1. Major structural proteins of the E. coli nucleoid. [Adapted from Johnson RC, Johnson LM, Schmidt JW et al. (2005) In The Bacterial Chromosome, Higgins PN ed., pp. 65–132.]
Fig4. Interactions of DNA bending proteins IHF and HU with DNA. (A) Protein–DNA co-crystal structures. IHF is bound sequence-specifically to the H’ site from phage λ. (B) AFM images of protein–DNA complexes. (C) At low concentrations, both proteins, represented by blue spheres, bend the DNA. At high concentrations, the proteins induce the formation of rigid filaments, as illustrated in the right-hand image for HU; a similar structure has also been proposed for IHF–DNA filaments at high concen trations of the protein. (A, from Swinger KK, Lemberg KM, Zhang Y et al. (2003) EMBO J 22: 3749–3760. With permis sion from John Wiley & Sons, Inc. B, from Luijsterburg MS, Noom MC, Wuite GJL et al. (2006) J Struct Biol 156: 262 272. With permission from Elsevier.]
Both HU and IHF bind nonspecifically but preferentially to highly supercoiled regions. IHF possesses some tendency to bind to specific nucleotide sequences as well. It is probably appropriate to think of these proteins as primarily having a structural function, although it is clear from Table 1 that they do not constrain, even together, a large fraction of the genome. Alone, they could not yield the extensive, stable higher-order structure that we find in the eukaryotic genome. The level of stability that they provide is appropriate for the bacterial genome, which is continually involved in processes such as transcription, replication, recombination, and trans position. This high activity level would be inhibited by highly stable structures.
Another group of proteins, including histone-like nucleoid structuring protein (H-NS) and structural maintenance of chromosome (SMC) proteins, are known to bridge DNA. The action of these proteins in compacting DNA is illustrated in Figure5. As we shall see in Chapter 11, H-NS also plays a significant role in transcriptional regulation. Other nucleoid proteins play predominantly functional roles. Factor for inversion stimulation (Fis), for example, acts rather like a transcriptional activator in higher organisms: it can bind upstream from a promoter site and interact with the C-terminal tail of the RNA polymerase, so as to activate transcription from that promoter. In addition, it can bend DNA and help to compact it (Figure 6). DNA protection during starvation protein (Dps) is a peculiar, specialized protein that accumulates only when bacteria are starved and enter a stationary phase. Under these circumstances, it forms very extensive and regular liquid-crystal co-structures with the DNA, protecting the genome from disintegration. In vitro, the protein inter acts with DNA and through self-association (aggregation compacts the DNA; see Figure 6). Overall, the bacterial nucleoid appears to be a very fluid, dynamic structure, well suited to the hectic metabolism of these organisms.
Fig5. Interactions of DNA-bridging proteins H-NS and SMC with DNA. (A) H-NS. Top, structure of the protein dimer; middle, AFM image of DNA loops formed by duplex bridging by H-NS; bottom, two possible models of binding. (B) SMC. Top, AFM image of rosettelike SMC structures in solution. A typical field shows rosettelike structures. Bottom, speculative model of rosette formation by SMC dimers, in which SMC head domains form the central part of the structures and arms extend outward; DNA loops could be trapped by the arms. [A, top, adapted from Luijsterburg MS, Noom MC, Wuite GJL et al. (2006) J Struct Biol 156: 262–272. With permission from Elsevier. B, adapted from Mascarenhas J, Volkov AV, Rinn C et al. (2005) BMC Cell Biol 6: 28, doi: 10.1186/1471-2121-6-28. With permission from BioMed Central.]
Fig6. Interactions of nucleoid associated proteins Fis and Dps with DNA. (A) Crystal structures of Fis-DNA complex and pure Dps protein, a huge dodecamer. (B) AFM images of protein–DNA complexes created in vitro. (C) Perceived modes of binding to create the structures in the AFM images. [A, B, top, from Luijsterburg MS, Noom MC, Wuite GJL et al. (2006) J Struct Biol 156: 262–272. With permission from Elsevier. A, bottom, courtesy of Andrea Ilari, University of Rome, Rome, Italy. B, bottom, from Ceci P, Cellai S, Falvo E et al. (2004) Nucleic Acids Res 32: 5935–5944. With permission from Oxford University Press.]
