The effects of stress on gut virome: Implications on infectious disease and systemic disorders

2.1 Gut virome

The virome, which refers to the collection of viruses that make up the intestinal microbiota, represents 1% (or less) of the total metagenomic reads, according to metagenomic studies on human fecal samples. It is estimated that every gram of feces contains 10⁸–10⁹ virus-like particles (VLP) (Iliev & Cadwell, 2021). The human intestinal virome includes both eukaryotic viruses, which replicate by infecting human cells, and bacteriophages, which replicate by infecting bacteria (Lecuit & Eloit, 2017). The eukaryotic virome primarily consists of the families Adenoviridae, Anelloviridae, Astroviridae, Parvoviridae, Picornaviridae, and Picobirnaviridae (Matijašić et al., 2020). Eukaryotic viruses are acquired progressively with age, unlike bacteriophages, which are highly present in the early stages of life and then decrease in their relative abundance over time (Lim et al., 2015). Bacteriophages constitute the most substantial component of the intestinal virome, and their presence is strongly influenced by the presence of their bacterial hosts (Lecuit & Eloit, 2017). Despite this, the most abundant are the DNA viruses, particularly Caudovirales (double-strand DNA viruses) and Microviridae (single-strand DNA viruses) (Cao et al., 2022). Bacteriophages are divided into lytic and lysogenic based on the type of interaction established with the bacterial host, although some can perform both lytic and lysogenic cycles (e.g., Escherichia coli lambda phage). During the lytic cycle, a bacteriophage infects a live bacterial target cell, replicates inside it, kills the bacterium through lysis, and releases hundreds or thousands of new bacteriophages, potentially playing a role in the vehiculation of genic traits (Żbikowska et al., 2020). For example, an investigation conducted on the growth cycle of the lytic bacteriophage k3w7, a ds-DNA virus belonging to the myoviridae family, potentially useful in therapies against Klebsiella pneumoniae infections, has been shown to produce carbapenemase (Baqer et al., 2023). Also, results showed that this virus has a short latency period (approximately 20 min) and a burst size of approximately 220 plaque-forming units per infected cell (Baqer et al., 2023). This strategy enables a swift modulation of the overall microbiota composition, by acting either directly or indirectly on the most abundant microbial fraction (i.e., bacteria), besides conditioning the biology of both the microbiota and the host (Hsu et al., 2019). Quantitative modulation of the microbiota composition has been attributed to bacteriophages at the expense of the low- and high-abundant bacterial specimens in a strictly specific manner (Reyes et al., 2013). In turn, the altered quantitative composition is mirrored in third, nontargeted, microbial specimens as the consequence of the changing biotic and abiotic milieu (Hibbing et al., 2009; Hsu et al., 2019; Schmidt et al., 2018).

In contrast, the lysogenic cycle leads to the integration of the viral genome, in a prophage form, into the bacterial chromosome (Żbikowska et al., 2020). It has been observed that E. coli prophages encode genes for resistance to antibiotics and other environmental stresses (Minot et al., 2011), underlying the interdependencies among the microorganisms within the ecological niche. Additionally, bacteriophage members can be classified as monovalent, infecting only one bacterial species, and polyvalent phages, capable of infecting two or more bacterial species (Żbikowska et al., 2020). Nevertheless, this classification is somewhat arbitrary since it is based on in vitro tests relying on a limited number of bacterial isolates, thus limiting the definition of the bacteriophage's tropism itself, which is likely to change in the context of the human gut (Chung et al., 2023; Kim et al., 2021).

3.1 Factors of virome dysbiosis

A recent study by Nishijima et al. identified as many as 97 factors that influence the virome structure. These factors are intrinsic (age, sex) or extrinsic (lifestyle, diet), most of which concern both bacteriophages and their bacterial hosts. These represent a basis for understanding the symbiotic relationships occurring among the bacterial communities and their viruses, supporting the medical and industrial applications of indigenous viruses (Nishijima et al., 2022). Mounting evidence suggests that the human gut virome displays distinct, age-dependent patterns of diversity, in response to an array of factors including immune status fluctuations throughout life stages (infant, child, adult, elderly). The highest overall viral richness is observed in infants and adults: there are significant decreases between infants and children or adults and elderly individuals, but there are significant increases between children and adults. These overall trends, however, did not apply evenly across virus types. For example, eukaryotic virus richness (mostly human Anelloviruses) is high in infancy, presumably driven by an underdeveloped immune system, and then decreases into childhood and remains constant and low through the rest of life. The bacteriophage family Siphoviridae (the most abundant in the human microbiota) mirrors the overall trend, but on the contrary, Microviridae reach a modest peak in infancy, decrease in childhood, and then slowly increase across the rest of the lifespan (Gregory et al., 2020).

Experiments on subjects who followed a different diet have shown that the intestinal virome changes significantly based on the type of diet. The pre-existing proportions in the virome were altered, while the subjects who followed the same diet had a more similar composition of the virome (but not identical because of the interpersonal variation). Specifically, dietary intervention suggests a trend toward enrichment in Siphoviridae and depletion of Myoviridae (Minot et al., 2011), although tailored changes are expected to happen depending on the change and duration of the dietary intervention. Interestingly, a study analyzing the fecal virome of 930 healthy adult subjects from two regions in China (Hong Kong and Yunnan), spanning six ethnicities (Han, Zang, Miao, Bai, Dai, and Hani), and including urban and rural residents for each ethnicity observed that gut-DNA virome differed significantly with geography- and ethnicity-specific compositional differences. Similarly, bacterial community composition was significantly affected, although overall human gut virome was more heterogeneous than the gut bacteriome at the population level (Zuo et al., 2020).

Gut virome differences have been linked to the presence of other pathological conditions. For example, in stroke patients, an alteration in the structures and composition of the microbiome was observed, which is probably due to changes in the virome. In fact, in these subjects, there was an increase in the relative abundance of Bacteroides phage B40_8 and Cronobacter phage CS01. Furthermore, in patients' virome, the relative abundance of the Streptococcus phages was positively correlated with that of their hosts, while in the healthy volunteers' virome, the relative abundance of the Faecalibacterium phage, Bilophila phage, and Roseburia phage was positively correlated with that of their hosts. These underline the connection existing between the composition of the bacteriophage and the bacterial fraction (Wang et al., 2021; Zhang et al., 2024).

5.1 Stress-induced susceptibility to infectious diseases through virome alteration

Both stress and the virome are considered significant factors in hindering the functions and efficiency of the immune system. In this scenario, two independent factors have converging outcomes, that is, host immune suppression, enabling the growth of pathogens and/or pathobionts, and flare-ups of infections.

The impact of stress on the gut virome is evaluated by considering the duration of the stress period. There is a notable difference in the effects of acute and chronic stress. Acute stressors (e.g., lasting minutes) can cause downregulation of specific immunity and upregulation of natural immunity. For example, mental stress due to exams tends to suppress cellular immunity and preserve humoral immunity. In contrast, chronic stressors are associated with decreased functionality of both cellular and humoral immunity. Depending on the type of stressful situation, the sequence of events affecting the immune system (trauma or loss) may vary. Nevertheless, diverse responses to the same stressor (and stressor entity) have been observed depending on subjective variables (e.g., age). This, in turn, is mirrored as variable immunological impairment (Segerstrom & Miller, 2004). A study performed on rats observed that a brief acute restraint stress led to an increased number of memory T-cells before immunization and helper T effector cells after immunization, suggesting short-term stress as an important immune booster and an effective strategy for the control of infective agents. This also appears to influence a more robust immune response following repeated stimulation with the same antigen even after months (Vitlic et al., 2014). Chronic stress, on the other hand, is generally recognized as immunosuppressive (Sorrells & Sapolsky, 2007), negatively influencing pathological outcomes of both infectious and neoplastic diseases. Also, allergic and autoimmune disease outcomes are affected by prolonged stress. Although apparently in contrast, one possible explanation suggests that chronic stress alters cytokine expression patterns, causing the suppression of Th1 cytokines (such as IFN-γ), involved in defense against intracellular infections and some neoplastic diseases, and the activation of Th2 cytokines (such as IL-10), involved in the onset and progression of allergies and various autoimmune diseases (Vitlic et al., 2014). Additionally, it has been demonstrated that acute stress increases the lymphocyte count in the blood, while chronic stress causes an overall decrease in the number of circulating lymphocytes, but there is an increase in the number of immature T lymphocytes. This suggests one of the potential mechanisms through which the immunosuppression associated with chronic stress can influence and exacerbate autoimmune diseases (Moroda et al., 1997).

In summary, acute stress factors, except for natural killer (NK) cells and superoxide production by neutrophils, enhance the immune system, particularly its innate component that can act rapidly. In any case, the greatest number of chronic stress factors are associated with general immunosuppression which affects both components of the immune system.

Evaluating antibody titers against latent viruses is an effective method for analyzing the consequences of psychological stress on the immune system (Vitlic et al., 2014). Stress resulting from marriage discordances negatively affects endocrine responses and immunological activation. An increase in antibody titers against the Epstein–Barr virus (EBV) and the blastogenic response to mitogens of T cells was observed in both men and women (Kiecolt-Glaser et al., 1997). Another study analyzed a group of caregivers and higher antibody titers against cytomegalovirus (CMV) were observed, indicating poorer control of the latent virus due to psychological stress events (Pariante et al., 1997).

It is often thought that eukaryotic viruses are responsible exclusively for asymptomatic infections, but active viral replication can influence the immune system, with positive or negative outcomes, depending on the circumstances (Iliev & Cadwell, 2021). Moreover, bacteriophages can transfer, through their nucleic acid, genes for antibiotic resistance or genes that code for the immune system escaping, thus increasing enteric bacterial pathogenicity (Colomer-Lluch et al., 2011). For example, phages sakϕC and sak42D (Staphylococcus aureus) encode staphylokinase (sak) that neutralizes antimicrobial peptides thus eluding the immune system and causing host tissue death (Nguyen & Vogel, 2016; Van Wamel et al., 2006). Phages can also encode inhibitors of complement (SCIN), proteins (CHIP), superantigens, and other substances that are responsible for the lethal outcome of S. aureus infection (De Jong et al., 2018). Also, emerging S. aureus strains can become resistant to ampicillin thanks to β-lactamase genes, blaTEM and blaCTX-M9, that can be transferred by S. aureus phage. Salmonella typhimurium strains appear to be multiantibiotic resistant acknowledging the transfer of resistance genes by temperate phages (Colomer-Lluch et al., 2011).

A study on the role of virome in bacterial adaptation, after antibiotic-induced stress in animal models revealed that phage genes undergo robust enrichment following antibiotic treatment. Such gene enrichment could promote the production of resistance genes against the administered drug in phages and their subsequent transfer to the bacterial community, thereby acting as a reservoir for resistant strains (Spencer et al., 2022). For example, research about patients undergoing antibiotics therapy for Helicobacter pylori demonstrates that following Helicobacter pylori eradication, gut virome diversity decreased (especially after multiple treatment cycles), community diversity increased, and there was a greater proportion of core viruses. The duration of these effects is 6 months, while phage–bacteria interactions are enhanced for up to 6 weeks (Wang et al., 2023) (Table 1).

Nevertheless, information regarding virome and the immune system is still insufficient. A possible explanation relies on the fact that the immune system normally suppresses these viruses to an abundance below the limit of detection of the most sensitive technologies, such as metagenomics (Ouabbou et al., 2020). The effects of intestinal viral infections are often mediated by interferons (IFNs), particularly type I, such as IFNA and IFNB, and type III, such as IFNLs (Sun et al., 2015). On the other hand, in case of infection by a virus that establishes symbiotic relationships, the IFN response may be necessary.

5.2 Implications of stress-altered virome on systemic disorders

Both mental and psychological stress are acknowledged to influence the host immune system and thus, the balance of the host-virome. It is straightforward to assume a cascade effect on the emergence and progression of systemic disorders, in addition to those of infectious origin.

Studies in subjects with type 1 diabetes provide the strongest evidence that intestinal viruses can cause chronic immune disease in humans (Notkins et al., 1979). Autoantibodies against insulin have been associated with chronic enterovirus B infection and, in a smaller number of cases, mastadenovirus C infection. Small single-stranded DNA viruses belonging to the Circoviridae family have also been associated with the development of diabetes-associated serum antibodies in the absence of early infection. Infection with group B Coxsackievirus reduces disease in younger nonobese diabetic (NOD) mice but accelerates it in older mice at a time when insulin values are already altered. Rotaviruses accelerate the autoimmune reaction by stimulating the production of type I INF by the pancreas, where it arrives through the lymph nodes; IFNs, in turn, activate self-reactive lymphocytes. Murine norovirus (MNV) infection, on the other hand, protects against the development of diabetes in NOD mice (Iliev & Cadwell, 2021).

Altered bacteriophage composition is associated with pathological inflammatory conditions affecting the intestine (Gogokhia et al., 2019). Specifically, clonal expansion of Caudovirales bacteriophages has been observed in patients with Crohn's disease (Norman et al., 2015). It has also been noted that, in murine models, viral infections induce characteristics associated with Crohn's disease and celiac disease. Particularly, they induce defects in Paneth cells and the synthesis of gluten-reactive lymphocytes (Iliev & Cadwell, 2021). The increase in bacteriophage levels (mainly those with tropism toward E. coli, Lactobacillus, and Bacteroides) in germ-free mice can alter the mucosal immune status and exacerbate colitis by inducing the production of IFN-γ through TLR9 (Gogokhia et al., 2019). Furthermore, the Spounaviridae phages and those targeting Streptococcus and Alistipes have been proposed as informative markers for murine colitis analogous to IBD (Duerkop et al., 2018). Interestingly, increased levels of bacterial hosts were not observed, as opposed to previous studies on this topic, suggesting that phage abundance is not always correlated with bacterial abundance and that inflammation might trigger phage replication. Furthermore, a comparison of the metagenomes of experimental and human models of intestinal bacteriophages showed good correspondence, suggesting how the murine colitis model could be employed in translational medicine studies for the human model (Matijašić et al., 2020).

Inflammatory bowel diseases (IBD), encompassing two major forms: ulcerative colitis (UC) and Crohn's disease (CD), are closely associated with alterations in the gut–brain axis and stress. Previous studies, using patient-reported outcomes (PROs) (Liuzza et al., 2021; Spagnuolo, Basile, et al., 2022), have highlighted how anxiety, depression, fatigue, and satisfaction in social roles may be compromised in these diseases both during flare-ups and during clinical remission phases (Iaquinta et al., 2022; Mancina et al., 2020; Spagnuolo, Iaquinta, et al., 2022). Also, these factors worsen the natural history of the disease in terms of surgical recourse and hospitalizations (Fairbrass et al., 2022). In addition, the evolution of IBD is strictly linked to the effects that stress has on the gastrointestinal microbiota (Konturek et al., 2011). Increased microbial load in the colon tissue, excessive cytokine release, and partially attenuated immune reactivity due to stress are among the leading factors with a negative impact on IBD. Depression is associated with elevated levels of C-reactive protein and tumor necrosis factor-alpha (TNF-α) in patients with IBD (Brzozowski et al., 2016). An increasing amount of data confirms that disturbances in the eukaryotic virome are linked to IBD pathogenesis: a study analyzing colon samples from IBD patients and healthy controls has shown that IBD is characterized by increased levels of the Herpesviridae family, as well as increased expression of relevant endogenous viral sequences (Wang et al., 2015). Studies on mice carrying mutations in IL-10 or Atg16L1 demonstrate that eukaryotic viruses can encourage the emergence of IBD by interacting with risk genes (Basic et al., 2014). The enteric virome is also abnormal in Crohn's disease and ulcerative colitis patients (Norman et al., 2015). Patients with UC exhibit a higher abundance of Pneumoviridae compared to the control group, while the opposite is observed for the Anelloviridae family (Zuo et al., 2019). There is elevated abundance of Hepadnaviridae and Hepeviridae in ulcerative colitis patients compared to controls, while diet-related viral families Polydnaviridae and Tymoviridae are less enriched. A similar trend has been observed for Virgaviridae in CD patients (Ungaro et al., 2019). Intestinal inflammation caused by eukaryotic viruses has also been studied in animal models, with a primary focus on Norovirus (Matijašić et al., 2020).

Based on current achievements, we can hypothesize that specific variations in the viral component of the microbiota could lead to IBD and that stress, in turn, could have a negative impact, causing disease flare-ups or exacerbation of symptoms. On the other hand, stressful conditions may induce inflammatory states that could lead to IBD, potentially affecting the viral component of the microbiota. The table below summarizes the main alterations of the virome associated with major system and gastrointestinal diseases (Table 2).

Taken together, the above observation suggests a potential implication of modulating the gut virome to address or at least ameliorate, these system disorders. Interestingly, a recent study demonstrated fecal virome transplantation (FVT) as a suitable method for restoring the gut microbiota of mice exposed to social chronic stress, with a consequent improvement of the clinical manifestation (Ritz et al., 2024). Similarly, FVT has proven successful in the remission of recurrent Clostridium difficile infections (Kao et al., 2019). Also, virome transfer from lean-to-obese mice restored the weight gain effect of the obese phenotype (Rasmussen et al., 2020). Similarly, targeted approaches such as phage therapy resulted in the selective reduction of Fusobacterium nucleatum, a bacterial specimen commonly associated with colorectal cancer conditions (Zheng & Dong, 2019); besides scoring success in reducing rates of antimicrobial resistance specimens, whose diffusion is currently posing serious threats to a broad spectrum of pathological conditions of both infectious and noninfectious origin (Duan et al., 2019; Heuler et al., 2021).

However, viral infections alone are not sufficient to induce a confirmed multifactorial disease. Viruses can, in combination with additional environmental factors and commensal bacteria, determine the phenotype of hosts carrying inflammatory diseases (Cadwell et al., 2010). The need for additional factors could explain the reduced frequency of association of common viruses with confirmed diseases (1% or less of the population). Moreover, studies in murine models of Crohn's and Celiac disease describe that the effects of viruses are strain-specific, complicating the linkage of viral profiles with various pathological conditions (Ottman et al., 2012).