Shortly after the beginning of the COVID-19 pandemic, the study conducted by Zuo and colleagues already showed that opportunistic pathogens were overrepresented whilst beneficial commensals were underrepresented in the gut of COVID-19 patients (
Table 1)
[20]. Moreover, they found several associations between changes in the abundance of certain bacterial taxa and the severity of COVID-19
[20]. Curiously, this gut dysbiosis was already present at the time of hospitalization, was detected at all timepoints during hospitalization, and persisted even after the clearance of SARS-CoV-2 and the resolution of the respiratory symptoms (
Table 1)
[20]. Tian and colleagues used 16S rRNA gene sequencing to compare the gut microbiota of seven uninfected controls and seven COVID-19-recovered patients at three months after discharge (
Table 1)
[24]. Coherently, they observed that alpha and beta diversity values were significantly different when comparing both groups (
Table 1)
[24]. Another work performed by Upadhyay and colleagues analysed the gut microbiota at later timepoints after SARS-CoV-2 initial infection (up to 154 days) (
Table 1)
[25]. Notably, this study could give us some clues about why no association has been found between long COVID onset or severity and the severity of the acute COVID-19 disease, thus finding many severely ill long COVID patients with a history of mild or even asymptomatic COVID-19 (
Table 1)
[26][27]. On the one hand, these authors demonstrated that SARS-CoV-2 altered the gut microbiota not only in the K18-humanized angiotensin-converting enzyme 2 mouse model, which is susceptible to SARS-CoV-2 infection, but also in wild-type C57BL/6J mice that are resistant to severe lung pathology from SARS-CoV-2 infection
[25]. On the other hand, they found that mild SARS-CoV-2 infection resulted in long-lasting disruption and instability of the gut microbial ecology, with the Firmicutes and Actinobacteriota phyla and the
Rothia genus as the most variable bacterial taxa following SARS-CoV-2 infection (
Table 1)
[25]. For this, they used 16S rRNA gene sequencing and shotgun metagenomic sequencing (MGS) to evaluate the gut microbiota in stool samples of SARS-CoV-2-positive human patients with mild symptoms at different times after initial infection (
Table 1)
[25]. Taken together, these results suggest that the long-lasting gut dysbiosis resulting from SARS-CoV-2 infection, which could be later playing a pivotal role on the pathogenesis of long COVID, is independent on the severity of the acute phase of the COVID-19 disease. It is worth highlighting that the vast majority of studies on long COVID enrol COVID-19 patients with a wide spectrum of disease severities, without stratifying them by severity in most cases (
Table 1). This is to avoid any possible biases that could arise if a given COVID-19 disease severity is over- or under-represented. However, the work performed by Upadhyay and colleagues is particularly useful as they exclusively focused on mild COVID-19 patients
[25]. Hence, they obtained valuable information regarding the possible associations between the severity of COVID-19 and long-lasting changes in the gut microbiota, which could lead to the onset of long COVID
[25]. Unfortunately, although interesting, these three already-mentioned studies presented very low sample sizes as their main limitation, which could compromise the soundness of their conclusions (
Table 1)
[20][24][25]. Moreover, none of the three followed the evolution of the patients to know which ones developed long COVID to compare them with those others who did not develop the disease (
Table 1)
[20][24][25].
Interestingly, Chen and colleagues went one step further in terms of the time elapsed from the initial infection to the time of analysis (
Table 1)
[28]. Thus, they studied the gut microbiota at three different timepoints after the SARS-CoV-2 infection, the longest being 6 months after discharge (
Table 1)
[28]. Specifically, they utilised 16S rRNA gene sequencing to monitor alterations in the faecal microbiota of COVID-19 patients with diverse disease severity at three timepoints: (1) acute phase, from illness onset to viral clearance; (2) convalescence, from viral clearance to 2 weeks after hospital discharge; and (3) postconvalescence, 6 months after hospital discharge (
Table 1)
[28]. Importantly, their data also supported the idea that SARS-CoV-2 infection results in a long-lasting disruption of the gut microbiota
[25], since they found that bacterial alpha diversity was significantly reduced in acute phase, convalescence, and postconvalescence patients to the same levels compared to uninfected controls (
Table 1)
[28]. However, their sample sizes were also low, they did not perform any relative abundance analyses, and the potential correlations between changes in alpha diversity and the onset or severity of long COVID were not explored (
Table 1)
[28]. As in the previous work, Liu and colleagues also extended their gut microbiota analyses up to 6 months after initial SARS-CoV-2 infection (
Table 1)
[29]. However, unlike the four studies already mentioned, in this case they established a group of long COVID patients (
Table 1)
[29]. These authors used MGS to analyse changes in faecal samples of 106 COVID-19 patients with different disease severity from admission until 6 months later, and then correlated the results with persistent symptoms at 6 months (
Table 1)
[29]. They observed that the gut microbiota of COVID-19 patients who did not develop long COVID was totally recovered and indistinguishable from that of uninfected controls at 6 months (
Table 1)
[29]. Nevertheless, the 76% of patients who developed long COVID presented significant compositional alterations of gut microbiota at the same time after infection (
Table 1)
[29]. Importantly, their analyses revealed significant correlations between the abundance of certain bacterial taxa and the presence of some of the most characteristic symptoms of long COVID (
Table 1)
[29]. Hence, the abundance of nosocomial gut pathogens, including
Clostridium innocuum, was high in patients with fatigue and neuropsychiatric symptoms, whereas the abundance of opportunistic gut pathogens was high in patients with persistent respiratory symptoms (
Table 1)
[29]. This study avoided the main limitations of the other works mentioned before, as sample sizes were reasonably high, and samples were homogeneously collected at 6 months after initial infection (
Table 1)
[29]. However, the COVID-19 disease severity of their enrolled patients was not taken into account, losing the opportunity to perform additional interesting comparisons, especially after having found that the gut microbiota composition at admission was associated with the occurrence of long COVID (
Table 1)
[29]. At this point, an apparent contradiction arises between the finding that the gut microbiota was restored in COVID-19 patients who did not develop long COVID observed by Liu and colleagues
[29], and the long-lasting gut dysbiosis in all COVID-19 patients described in the previously discussed works by Tian and colleagues
[24], Upadhyay and colleagues
[25], and Chen and colleagues (
Table 1)
[28]. Nevertheless, these differences are probably because, while Liu and colleagues separated COVID-19 patients who developed long COVID and those who did not into two different groups (
Table 1)
[29], all COVID-19 patients were analysed together in the other three studies (
Table 1)
[24][25][28]. Therefore, in these three studies the long-lasting gut dysbiosis observed in patients who developed long COVID could be concealing the recovery of gut microbiota in patients who did not develop long COVID (
Table 1)
[24][25][28]. To elucidate this issue, Zhang and colleagues published what probably is the most clarifying study on this topic to date, mainly because their sample sizes were very high, and they extended the time at which the gut microbiota was analysed up to 1 year after discharge (
Table 1)
[30]. They performed 16S gene sequencing of stool samples from three different groups of subjects at 1 year after discharge: (1) healthy controls; (2) COVID-19-recovered patients without persistent symptoms; and (3) COVID-19 patients who presented long COVID symptoms (
Table 1)
[30]. Then, they analysed the correlations between the gut microbiota and long COVID, finding that, consistent with the results from Chen and colleagues
[28], alpha diversity was significantly reduced in long COVID patients compared to both control groups (
Table 1)
[30]. However, while Chen and colleagues observed that alpha diversity decreased in all their groups of subjects up to 6 months after discharge compared to uninfected controls
[28], Zhang and colleagues found that alpha diversity was normal in both uninfected controls and COVID-19 patients who did not develop long COVID at 1 year after discharge (
Table 1)
[30]. Thus, this sheds light on the previous discussion and reinforces the idea that the gut microbiota of COVID-19 patients who did not develop long COVID has probably been restored at 6 months after initial infection as shown by Liu and colleagues
[29], and at 1 year after discharge as shown by Zhang and colleagues
[30]. In any case, besides these changes in alpha diversity, Zhang and colleagues also analysed the compositional structure of the gut microbiota (
Table 1)
[30]. They observed that the relative abundance of the bacterial genera
Eubacterium,
Agathobacter,
Subdoligranulum, and
Ruminococcus was significantly different in COVID-19 patients who developed long COVID compared to both control groups, whereas, on the contrary, the genus
Veillonella was overrepresented in long COVID patients (
Table 1)
[30].
In conclusion, taken together, all these results strongly suggest that SARS-CoV-2 infection induces long-lasting gut dysbiosis in COVID-19-recovered patients who develop long COVID. However, the gut microbiota of COVID-19 patients who do not develop long COVID is already restored at 6 months after the initial SARS-CoV-2 infection, or probably even earlier, although further research would be necessary to determine this. Therefore, new studies analysing multiple timepoints after the SARS-CoV-2 initial infection will be necessary to characterize the exact dynamics of changes in the gut microbiota of patients in recovery, as well as to elucidate the involvement of gut dysbiosis in the pathogenesis of long COVID.
3. The Upper Respiratory Tract Microbiota in Long COVID
The upper respiratory tract, and especially the nasopharynx, has been shown to be key for SARS-CoV-2 infection and proliferation
[16][21]. The nasopharynx presents a common meeting place for the ear, nose, and oral cavities, which are the main portals of entry for the virus
[16][21]. Thus, the nose and oral cavities are particularly important for the entry of respiratory viruses, including SARS-CoV-2, being the first places where their replication starts
[16][21]. In the human body, only the gut microbiota is larger than that of the mouth, whose microbial community comprises over 1000 species of commensal bacteria, viruses, fungi, and protozoa
[31]. Microorganisms from the oral cavity can be aspirated or ingested, altering the microbiota of the new anatomical areas to which they reach and potentially producing diseases such as pneumonia
[32][33]. Importantly, oral dysbiosis has been linked to the pathogenesis of multiple systemic diseases
[34][35]. Undoubtedly, the most important contribution to date to the knowledge of the relationship between the oral microbiota and long COVID was the work published by Haran and colleagues
[36]. They collected tongue swabs from SARS-CoV-2-infected patients presenting COVID-19 symptoms and followed their evolution, finding that 37% of them developed long COVID
[36]. Importantly, they found that the oral microbiota of long COVID patients presented higher abundance of the bacterial genera
Prevotella and
Veillonella [36]. Interestingly, it has been described that these bacterial genera can reach and infect the lungs through the oral–lung aspiration axis
[37][38], and have been linked to systemic diseases
[39].
Prevotella and
Veillonella induce inflammation by TLR-2 activation and induction of the cytokines IL-23 and IL-1 in the case of
Prevotella [40][41], and mainly by a strong induction of IL-6 in the case of
Veillonella [42]. Notably, the low-grade inflammation induced by members of the
Prevotella genus is known to be systemic
[39]. Moreover, authors found increased abundance of lipopolysaccharide-producing bacterial species in samples from long COVID patients compared to controls, including
Veillonella dispar,
Veillonella infantium,
Veillonella atypica,
Leptotrichia wadei, and
Megasphaera micronuciformis [36][43][44][45]. Another interesting observation is that metabolic pathways known to have anti-inflammatory properties were reduced
[36]. Therefore, all these findings suggest that the inflammation induced by some of the bacterial genera that were overrepresented in the oral cavity of long COVID patients could be playing a pivotal role in the origin of their long-lasting symptoms. This is consistent with the widespread hypothesis that systemic chronic inflammation may be involved in the pathogenesis of long COVID
[46][47]. One of the main strengths of this study was that authors collected samples at initial COVID-19 stages when patients had not received any treatment that could alter their microbiota
[36]. In addition, they obtained reliable control samples from patients who did not develop long COVID but that were collected exactly in the same conditions and at the same timepoints as the long COVID samples
[36]. However, sample sizes were low (with only 10 long COVID patients), and the study was devoid of another control group composed of uninfected subjects which would have allowed for additional interesting comparisons
[36].
In summary, although the involvement of the upper respiratory tract microbiota in the onset and/or severity of long COVID has barely been studied yet, there are already important findings describing the existence of oral dysbiosis in long COVID patients
[36]. As it has been amply demonstrated that the microbiota of other parts of the upper respiratory tract, especially that of the nasopharynx, is significantly altered in COVID-19 patients
[16], further research will be necessary to determine whether it plays a role in the pathogenesis of long COVID.