Deep Lessons From the Uncultured.

(

In the current issue of The Journal of Infectious Diseases, Hage et al [1] report on the use of modern DNA sequencing methods in a study of the genetic diversity of human cytomegalovirus (HCMV). Information about HCMV population dynamics provided by this and other studies of HCMV genetic diversity reveal a richer and more subtle biology of the virus than previously documented and suggest that more forms of population diversity remain to be identified. The results provide important guidance to researchers and point to future clinical applications during acute and long-term care of patients suffering from or at risk of HCMV infection.

GOING DEEP INTO THE WILD

Recent advances in nucleotide sequencing have enabled novel insights related to genetic variation in populations of organisms and microbes throughout nature. On slowly changing genetic backdrops that help to define species, a plethora of biological mechanisms have evolved for generating genetic and epigenetic adjustments that enhance biological fitness by enabling rapid responses to changes in the biologic milieu. Methods for laboratory culture of bacteria and fungi were developed in the 1800s, and then for replication of viruses in cultured cells in the middle of the 20th century. Much of what we know about the genetics of microbes is based on study of material propagated in laboratories. Information gained in recent years makes it clear that much can be learned from uncultured, wild microorganisms.

Typical contemporary DNA sequencing methods rely on sequencing large numbers of random fragments generated from target sequences [2]. Accurate sequencing is dependent on determining every position in a sequence tens to hundreds if not thousands of times. An important byproduct of such depth of coverage is reliable detection of sequence heterogeneities that are present even at low frequencies. Use of these methods to sequence virus genomes present in uncultured clinical specimens has led to discovery of novel free and endogenous viruses [3] and provides important information about the extent and nature of inter- and intrahost virus genetic diversity [4-10].

HUMAN CYTOMEGALOVIRUS: A SHAPE-SHIFTING CHAMELEON

Human cytomegalovirus is a pathogen of high societal importance [11]. It is a major cause of morbidity and mortality in immunocompromised patients [12], but its major societal impact is due to transplacental congenital infections that cause a spectrum of effects, including severe, often life-threatening disease in neonates, hearing loss (sometimes delayed) in young children [13,14], and associations with other developmental disorders [15].

Human cytomegalovirus has complex biology that includes a protracted replication cycle that is dependent on multiple virally encoded tools that regulate intrinsic, innate, and adaptive defenses (reviewed in [12,16]). After primary infection, the virus persists for the lifetime of the host, making use of a repository of latently infected cells from which sporadic reactivations of lytic replication produce infectious virions that reseed the latent repository, stimulate the immune system, and are vehicles for cell-to-cell and host-to-host transmission.

The HCMV genome is the largest (approximately 236 kbp) of any known human virus. Of its approximately 170 protein-coding genes, only 45 are required for production of infectious virions in cultured cells. Most of the other genes, referred to as accessory genes, play roles in modulating interactions with host immune functions and enabling efficient entry and replication in various cell types. The genome, thus, represents a robustly complex toolbox for evolutionary survival in a host that is well equipped to defend itself.

Much of what is known of HCMV molecular and cell biology is the product of studies that used strains of the virus that were cultured from clinical specimens obtained in the 1950s (reviewed in [17]). In 1996, Cha et al reported that genomes of the 2 most widely used laboratory strains of the virus (AD169 and Towne) lack similar 13–15-kB segments that are present in viruses with less extensive culture passage history [18]. Subsequent studies made it clear that the HCMV genome changes during replication in cultured cells, with some changes occurring in a specific sequence during adaptation of virus present in fresh clinical specimens to replication in cultured cells [19]. Genomic changes that occur during culture include point mutations and small insertions or deletions (indels) [6] and deletion of large gene blocks, such as described above. Many of these changes affect genes that regulate cell tropism and evasion of intrinsic, innate, and adaptive defenses.

Studies of genes modified or lost upon passage in culture have indicated that HCMV is a shape-shifting chameleon for which replication in 1 cell type results in production of virions more suited to infecting cells of other types [20,21]. This enables infection of a wide range of cell types (eg, fibroblasts, endothelial cells, epithelial cells, monocytes) that populate a wide variety of tissues (eg, lung, intestinal, renal, retinal, nervous system, placental, fetal) that are important sites of HCMV’s clinical manifestations.

HUMAN CYTOMEGALOVIRUS GENETIC DIVERSITY AND VIRUS BIOLOGY

In the work by Hage et al [1], modern sequencing methods were used to obtain 57 complete HCMV genome sequences from uncultured clinical specimens obtained from various bodily compartments in renal and stem cell transplant recipients, as well as from congenitally infected children and their mothers. In addition, the authors capitalized on the power of longitudinal specimen collection [22] by collecting specimens at multiple time points. Although only minor population shifts were observed over time in some patients, in several individuals, the major HCMV population in blood switched from 1 genotype to another over the course of the study. Whole-genome genotype switching is consistent with long-standing evidence that individuals can be infected by, and then harbor, multiple strains of virus. Clinically, organ transplant donor– recipient mismatches with respect to HCMV glycoprotein H genotypes are more likely to have adverse outcomes [23].

The authors also detected minor virus subpopulations that harbored markers for antiviral resistance, providing the virus the ability to rapidly respond to therapeutic intervention. A patient with HCMV retinitis had distinct virus populations in the vitreous humor compared with blood. Other recent studies of HCMV diversity have also shown that major populations of the virus can differ between blood and urine in congenitally infected children and that population diversity can shrink dramatically (a genetic bottleneck) as infection proceeds from 1 compartment to another [9,10]. In addition, the diversity of HCMV populations undergoes a substantial genetic bottleneck at the time of human-to-human transmission [10], providing evidence that HCMV transmission can be initiated by a very small number of infectious virions.

The level of diversity seen for HCMV in uncultured specimens is in the range of diversities seen for RNA viruses [24], which have historically been thought to have greater genetic diversity than DNA viruses. In contrast to RNA viruses, wild populations of HCMV exhibit extensive interstrain recombination that appears to maintain core replicative functions while generating diversity in genes that facilitate immune escape [4,6,9]. These recombinational genotypic shifts are loosely analogous to segment-shuffling of influenza viruses.

Somewhat surprisingly, accessory genes with long evolutionary histories are inactivated in some wild lineages [25]. This can occur through small changes, such as point mutations or indels that can inactivate genes by mechanisms such as translational frameshifting, without discarding all of the information gained during their evolution. Under subsequent selective pressure, relatively small changes would be sufficient for reactivation of these genes at much lower biological cost than waiting for de novo evolutionary emergence of a tool that serves the same purpose. Interestingly, no evidence for virus gene inactivation was seen in a study of uncultured HCMV genomes present in malignant glioma specimens, suggesting that these genomes might represent replication competent virus [26].

DEEP LESSONS FROM THE UNCULTURED

It is important to conduct research with viruses that are genetically similar to uncultured wild virus and to understand the genotypic limits of whatever strain is used. For animal models that use viruses such as murine cytomegalovirus, guinea pig cytomegalovirus, and rhesus cytomegalovirus, it is important to understand how the strain studied compares with uncultured virus from the same host. In studies of humans, assays for endpoints such as stimulation of protective immunity need to be evaluated against collections of strains that reflect the diversity of the target population.

Nomenclature matters. As suggested by Wilkinson et al [17], it seems reasonable to apply the term clinical strain to virus present in clinical specimens that has not been passaged in cultured cells. Descriptions of strains termed low- passage should be accompanied by details of their provenance, passage number, nature of passage (through infected cells or cell-free virus), and the cell types used for passage.

The form of virus diversity in the wild changes in response to environmental change, including changes due to human behavior. For HCMV, such changes include immune suppression in the context of organ and stem cell transplantation and antiviral therapy. Herpesviruses of farmed animals provide excellent examples of the speed with which related viruses evolve in response to environmental factors such as crowding, life-span changes, and vaccination [27-29]. We anticipate changes in HCMV populations in response to increased use of daycare and look forward to the problem of monitoring HCMV in vaccinated populations.