The majority of reported direct bacteria-virus interactions are associated with viruses infecting the gastrointestinal tract. In this body system, commensal bacteria are considered the first line of defense against invading pathogens by outcompeting their disease-promoting counterparts and limiting tissue accessibility. Undoubtedly, enteric viruses encounter these large numbers of diverse commensal bacteria, but rather than always preventing infection, some viruses evolved to exploit this contact, facilitating the disease process [8].
Under in vitro conditions, viruses may be able to directly bind to their target cell type and undergo replication with ease. However, this strategy may prove problematic in the gastrointestinal tract where a large number of bacteria occupy tissue surfaces, directly competing for receptor binding sites, and reducing the likelihood of pathogenic bacterial proliferation or virus attachment. Other components, like mucus or enzymatic secretions, may also interfere or assist the infection process. To circumvent this, rather than compete for host cell binding sites, some viruses can utilize bacterial ligands to enhance their association with more accessible host cells, initiating infection. This same strategy may be employed by some viruses that may not exclusively target the host’s epithelial cells and use bacteria to assist infection of other cell types in addition to or exclusive of epithelial cells [42,43].
An increasing body of work highlights how certain types of bacteria promote viral disease symptoms, suggesting bacterial populations may aid infection [9]. As research into this area expands, specific bacterial targets are being identified as viral binding sites. Certain enteric viruses illustrate how a bacterial component—attached or independent to the bacterium itself—promotes the virus infection cycle. Poliovirus, which replicates in the intestine prior to systemic dissemination, is a good example [43]. In a comprehensive study [18], mice with and without normal gut microflora were orally challenged with poliovirus. The former group demonstrated mortality twice that of the mice treated with antibiotics. This increase in mortality was associated with increased viral titers in mice with the intact microbiota. These effects were not observed after intraperitoneal challenge in which the virus does not need to interact with the native microbiota prior to infection. When these findings were examined using a cell culture model, exposure of poliovirus to bacteria or bacterial components increased virus titers as much as 500% and doubled the poliovirus adherence to HeLa cells [18]. It was hypothesized that bacterial components—lipopolysaccharides (LPS), peptidoglycan, and other N-acetylglucosamine-containing polysaccharides—increase viral receptor binding and increase viral shedding [18]. Interestingly, evidence has been reported that this also may be the case for murine norovirus, as antibiotic-treated mice show reduced virus titers compared to mice having normal gut microflora [10,39]. A similar mechanism may also be used by human norovirus. Histo-blood group antigens (HBGAs) have been identified as putative host cell receptors or co-receptors for this virus, and HBGA-like moieties have been reported to be present on the surface of some enteric bacteria (i.e., Enterobacter cloacae) [14,44,45]. These motifs were found in the bacterial extracellular polymeric material and bound to representative human norovirus strains. Furthermore, bacterial components have been reported to facilitate viral replication in a BJAB cell culture system. More specifically, the bacterial derived HBGA-like molecules stimulated replication comparable to their synthetic counterparts, in a dose dependent manner for Enterobacter cloacae, a commensal enteric bacteria. Further, the data presented suggests bacterial HBGAs in Enterobacter cloacae enhance norovirus attachment to target cells, increasing viral infectivity [13]. Expanding on these results, a recent study suggests that some commensal Escherichia coli strains expressing HBGA-like moieties may aid norovirus resistance to heat [45], which may have implications for norovirus persistence, however, future study of bacterial effect on viral persistence must be conducted. Interestingly, this is a similar phenomenon to that which was observed with poliovirus, as binding to bacteria also increased the stability of the viral capsid when exposed to heat [19]. Thus, both poliovirus and norovirus provide examples of viruses with enhanced pathogenesis when directly binding commensal enteric bacteria.
Although less prominent in the literature, members of the Reoviridae family of viruses may also exhibit similar gut microbiota interactions. In one study, a cohort of mice was treated with antibiotics prior to challenge, and disease pathology was compared to untreated but challenged animals to test the contribution of the native gut microflora. The untreated mice had classic reovirus strain T3SA+ symptoms with biliary obstructions and enlarged Peyer’s patches. The antibiotic-treated mice appeared normal, and also had significantly lower reovirus titers in the intestine. Similarly, poliovirus infection was enhanced in vivo in a mouse model with wild type mice versus germ-free or antibiotic-treated mice. Additional work also showed that the presence of bacteria enhanced poliovirus replication using a plaque assay [18]. Like norovirus, this may be related to bacteria presenting carbohydrates, since rotavirus and reovirus TS3A+, different genera in the Reoviridae family, exploit HBGAs and sialic acid carbohydrates as cellular receptors, respectively [46,47]. However, the nature of these interactions is poorly characterized, as is the role of HBGA or HBGA-like moiety binding in the rotavirus infection process [20]. Certainly, future work describing the specific mechanisms for the enhancement of reovirus and rotavirus infectivity is forthcoming.
In addition to binding and stabilization of viral capsids, bacterially synthesized enzymes can also stimulate viral infection. In this instance, bacterial presence not only increases influenza virus adhesion [48], but the virus gains a foothold by utilizing bacteria components for infection. To become infectious, the precursor hemagglutinin (HA0) of influenza needs to undergo proteolytic cleavage into HA1 and HA2 fragments. Typically, the host supplies enzymes necessary for activation; however, studies also implicate proteases produced by Staphylococcus aureus and Aerococcus viridans. Such synergism therefore promotes viral pathogenesis [21,22,23]. Unlike the previous examples where enhanced viral pathogenesis likely stemmed from viruses binding the bacteria, this is an example of viruses enhancing pathogenesis by also directly utilizing a bacterial product to aid infection.
Human immunodeficiency virus (HIV) provides a unique example of the interdependence between virus and bacteria. While previous examples dealt with commensal bacteria exploited for viral invasion, HIV recruits another human pathogen. More specifically, some evidence exists that individuals infected with HIV are more prone to Mycobacterium tuberculosis, and infection with the bacteria accelerates the progression of the acquired immune deficiency syndrome (AIDS). Latent infection of individuals with M. tuberculosis has been reported, and in some cases “reactivation” occurs where disease symptoms are observed years after initial infection [49]. Multiple specific mechanisms of HIV promotion of M. tuberculosis reactivation have been reported; for example, depletion of CD4+ T cells and up-regulation of the CD14 (which may aid M. tuberculosis infection) in macrophages, among other proposed mechanisms (reviewed in [25]). During acute M. tuberculosis infection, HIV RNA copy number increases [24,25], possibly due to the interplay between the M. tuberculosis cell wall component lipoarabinomannan and the immune system. M. tuberculosis up-regulates the production of tumor necrosis factor (TNF), an immune system component that controls bacterial infections, which activates HIV replication in macrophages [25]. The immune system also produces interleukin (IL)-6 which, in conjunction with higher TNF, activates transcription of the long terminal repeats in HIV, abetting replication [50]. The immune system is also involved in the bacteria-virus interactions of another retrovirus, mouse mammary tumor virus (MMTV), which interacts with commensal enteric bacteria. In this case, evidence was reported suggesting that MMTV binds enteric bacterial LPS that initiates Toll-like receptor 4 (TLR4, a pattern recognition sensor that targets LPS) that then activates IL-10 and then IL-6 with the effect of allowing the MMTV antigen to evade the immune response and persist in the host. In other words, MMTV binds to and uses bacterial LPS to “cloak” itself from the immune system and persist [26,27]. Both MMTV and HIV are examples of viruses exploiting bacterial effects on the immune system for enhancing infection.
Although the nature of the interaction remains the same, there may be additional benefits to viruses with bacteria interactions other than direct disease progression. Studies have also shown that association with fecal microbiota increased poliovirus environmental fitness and stability, as exposure to bacteria or their polysaccharides decreased the efficacy of virus inactivation by heat and bleach, potentially aiding viral survival in the environment [19]. This observation was further supported by the higher susceptibility to inactivation with heat observed in a poliovirus mutant that did not bind LPS as efficiently. Furthermore, the introduction of these bacterial polysaccharide components has been found to enhance wild type poliovirus binding to its host cells expressing its receptor [19]. In short, gastrointestinal microbiota not only increase poliovirus infectivity, but may also promote virus transfer to the next host. Thus, there are numerous ways that direct viral interaction with bacteria aid viral pathogenesis, and this topic is an emerging area of study in microbiology.