Article Text

Inflammation-associated gut microbiome in postacute sequelae of SARS-CoV-2 points towards new therapeutic targets
  1. Valdirene Leao Carneiro1,
  2. Katherine M Littlefield2,
  3. Renee Watson2,
  4. Brent E Palmer2,
  5. Catherine Lozupone3
  1. 1 Department of Life Sciences, University of Bahia State, Salvador, Brazil
  2. 2 Department of Medicine, University of Colorado - Anschutz Medical Campus, Aurora, Colorado, USA
  3. 3 Department of Biomedical Informatics, University of Colorado - Anschutz Medical Campus, Aurora, Colorado, USA
  1. Correspondence to Dr Catherine Lozupone, Department of Biomedical Informatics, University of Colorado - Anschutz Medical Campus, Aurora, Colorado, USA; catherine.lozupone{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

We read with interest the recent report by Liu et al 1 describing faecal microbiome differences with postacute sequelae of SARS-CoV-2 (PASC), commonly referred to as ‘Long-COVID’. We have previously reported elevated levels of SARS-CoV-2-specific T cells with PASC compared with resolved COVID-19 (RC; no lingering symptoms at the time of sample collection) that correlated with increased levels of the inflammatory marker IL-6, suggesting that elevated inflammation in PASC may be related to immune response to residual virus.2 Although several studies have reported gut microbiome differences during acute COVID-19,3 PASC has received less attention. We, thus, sought to characterise gut microbiome differences in PASC versus RC using faecal samples from our study2 and to relate these differences to inflammation.

The faecal microbiome was evaluated using 16S rRNA gene sequencing. Plasma levels of inflammatory markers IL-6 and C reactive protein (CRP) were measured with ELISA (see online supplemental methods). Cohort information is in table 1. IL-6 and CRP were elevated with PASC (figure 1A). Gut microbiome composition did not significantly differ between the PASC and RC cohorts (PERMANOVA; p=0.087; figure 1B), but did correlate with IL-6 and CRP levels (Adonis; IL-6 p=0.03; CRP p=0.01). IL-6 and CRP also correlated with PC1 from a principal coordinates analysis (figure 1C,D), suggesting a relationship between microbiome composition and inflammation in PASC. Using SELBAL,4 which identifies ratios or ‘Balances’ of microbes that can differentiate between groups, we found that the faecal microbiomes of individuals with PASC had a lower ratio of an amplicon sequence variant (ASV) highly related to Faecalibacterium prausnitzii over ASVs related to species in the genus Bacteroides (B. dorei, B. massiliensis and B. thetaiotaomicron) (figure 1E), which provided an area under the curve (AUC) of 0.863 for differentiating individuals with PASC from RC. Balance values also negatively correlated with IL-6 (r=−0.44, p=0.01). These microbiome differences are consistent with Liu et al,1 who also reported higher levels of Bacteroides (B. vulgatus specifically) and lower F. prausnitzii with PASC. Liu et al also reported higher Ruminococcus gnavus with PASC, and lower Collinsella aerofaciens, and Blautia obeum. Interestingly, an ASV highly related to R. gnavus (100% identity over V4 read) correlated positively with IL-6 and ASVs related to F. prausnitzii (98.7% ID), C. aerofaciens (100% ID) and B. obeum (100% ID) all negatively correlated with IL-6 and/or CRP levels in our study (online supplemental table 1). Thus, our results are consistent with those of Liu et al and extend their findings by showing associations between the microbiome and markers of systemic inflammation.

Supplemental material

Figure 1

(A) Plasma levels of IL-6 and CRP measured by ELISA and separated by cohort. The difference between cohorts was determined by Mann-Whitney tests. (B) PCoA decomposition of unweighted UniFrac distances for PASC and RC participants. Relationships between PC1 from (B) and circulating levels of IL-6 (C) and CRP (D) for PASC and RC. Statistical significance was measured using linear regression (IL-6/CRP~PC1). Each point represents data from one participant where teal represents PASC and orange represents RC. (E) The groups of taxa that form the global balance are specified at the top of the plot. The box plot represents the distribution of the balance scores for PASC and RC individuals. (F) ROC curve with its AUC value (0.863) for differentiation PASC from RC based on the balance in E. AUC-ROC, area under the reciever operating curve; CRP, C reactive protein; FPR, false postivie rate; PASC, postacute sequelae of SARS-CoV-2; PC, principal coordinate; PCoA, principal coordinates analysis; RC, resolved COVID-19; TPR, true positive rate.

Table 1

Cohort demographics

A mechanistic link between microbiome differences and high inflammation in PASC is supported by studies showing anti-inflammatory effects of F. prausnitzii 5 and proinflammatory effects of R. gnavus.6 Higher levels of particular Bacteroides and R. gnavus in PASC may also be interesting because R. gnavus, B. thetaiotaomicron and B. vulgatus all produce sialidases that can liberate sialic acids from host mucin.7–9 B. thetaiotaomicron can also increase the ratio of sialylated to sulfated mucins in mono-associated rats, and this effect was diminished by co-colonisation with F. prausnitzii.10 Sialic acids may be important in SARS-CoV-2 pathogenesis because the receptor-binding domain of the spike protein recognises sialic acid-containing oligosaccharides.11 Increased levels of microbially liberated sialic acids could support immune evasion by SARS-CoV-2 when free sialic acids bind to the spike protein; sialic acids are used for host immune evasion by multiple bacterial pathogens.12 Sialidase and sialyltransferase inhibitors have been effective in influenza prophylaxis and symptom relief,13 and also prevented pathogen outgrowth due to liberation of sialic acids from gut mucins by B. vulgatus in mice8, suggesting that, if experimentally validated, sialidase inhibition has therapeutic potential for PASC.

Ethics statements

Patient consent for publication

Ethics approval

This protocol was approved by Colorado Multiple Institutional Review Board (CoMIRB #20-1219). Participants gave informed consent to participate in the study before taking part.


We would like to thank our study participants for their time and contribution of samples. We thank Casey Martin for data analysis guidance, Sarah Jolley for help with recruitment, and Min Zhang and Eiko Yamada for support in the lab.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Twitter @CathyLozupone

  • BEP and CL contributed equally.

  • Contributors BEP conceptualised the study and supervised recruitment, sample collection and immune data generation. VLC performed all microbiome data analyses and contributed to results interpretation and writing of the letter. KL recruited individuals, generated and analyzed data, and contributed to results interpretation and writing of the letter. RW compiled metadata and contributed to literature searches. CL supervised microbiome data generation and analysis and results interpretation and writing of the letter.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.