Elevated HMGB1 mediates microbiome-immune axis dysregulation underlying reduced neutralization capacity in obesity-related post-acute sequelae of SARS-CoV-2

doi:10.21203/rs.3.rs-2973496/v1

Download PDF

Article

Elevated HMGB1 mediates microbiome-immune axis dysregulation underlying reduced neutralization capacity in obesity-related post-acute sequelae of SARS-CoV-2

https://doi.org/10.21203/rs.3.rs-2973496/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 03 Jan, 2024

Read the published version in Scientific Reports →

You are reading this latest preprint version

While obesity is a risk factor for post-acute sequelae of SARS-CoV-2 infection (PASC, "long-COVID"), the mechanism(s) underlying this phenomenon is poorly understood. To address this gap in knowledge, we performed a 6-week longitudinal study to examine immune activity and gut microbiome dysbiosis in post-acute stage patients recovering from SARS-CoV-2 infection. Self-reported symptom frequencies and blood samples were collected weekly, with plasma assessed by ELISA and Luminex for multiple biomarkers and immune cell profiling. DNA from stool samples were collected at the early stage of recovery for baseline assessments of gut microbial composition and diversity using 16S-based sequencing. Multiple regression analyses revealed obesity-related PASC linked to a proinflammatory immune profile and reduced adaptive immunity, corresponding with gut microbial dysbiosis. In particular, the high mobility group box 1 (HMGB1) protein was found to be a central mediator of this dysregulation, with its upregulated levels in plasma associated with significantly impaired viral neutralization that was exacerbated with obesity. Plasma levels of HMGB1 also negatively correlated with B-cell activating factor (BAFF), while inducing pro-inflammatory nitric oxide. These findings strongly implicate HMGB1 as a candidate biomarker of PASC, with potential applications for risk assessment and targeted therapies.

Although the number of new cases of SARS-CoV-2 infection globally is subsiding, the prevalence of post-acute sequelae of SARS-CoV-2 (PASC) in individuals recovering from COVID-19 remains a significant public health concern. In particular, individuals with pre-existing conditions associated with immune dysregulation are more susceptible to PASC¹, and collectively, studies have associated systemic inflammation and reduced humoral activity with this condition.^1,2 In addition, given the role of the gut microbiome in regulating immune function, gut microbial dysbiosis also associates with PASC.^3,4 Further, obesity is a common risk factor for PASC in association with a dysregulated microbiome-immune axis.⁵ Thus, the combination of metabolic dysfunction (from insulin resistance and adiposity) and pre-existing disruptions in the microbiome associated with chronic low-grade inflammation in obesity may exacerbate viral persistence that underlies PASC in recovering patients.⁶

PASC is a complex and multifactorial syndrome, and establishing its pathophysiology has proven challenging,⁷ requiring further molecular assessment of implicated processes to distinguish its full biological cascade. PASC's molecular obscurity has prompted several challenges to disease management.⁵ First, without molecular markers for laboratory testing, clinical diagnosis is the only option, likely leading to underdiagnosis. Additionally, without molecular diagnosis, there remains debate when defining and accepting the transition from COVID-19 to PASC.^8,9,10 Although biomarkers linked to COVID-19 severity and PASC have recently been proposed,^11,12 with no clear understanding of their relationship to PASC, these remain to be validated. However, a major challenge is the inability to predict who may develop PASC, which might otherwise facilitate prophylactic measures, or therapeutic intervention, upon symptomatic onset.¹³

Herein, we explored immune- and gut microbiome-associated pathways in individuals recovering from SARS-CoV-2 infection in the post-acute stage to better characterize the development of PASC. We aimed to identify PASC susceptibility biomarkers as candidate PASC predictors and/or therapeutic targets. To molecularly resolve PASC under normal and obesogenic conditions, individuals recovering from COVID-19 were invited to participate in a 6-week longitudinal cohort study via weekly self-reporting their symptoms and providing blood samples. Plasma was subjected to enzyme-linked immunosorbent assays (ELISA) and Luminex technology to measure antibody response and inflammatory/metabolic biomarkers, while peripheral blood mononuclear cells (PBMC) were stained and FACS-sorted for immune cell profiling. Microbial DNA, extracted from a single stool sample obtained at the baseline timepoint, was subjected to 16S-based sequencing using the Ion Torrent platform. Data was integrated and analyzed by Integrated Pathway Analysis (IPA) to identify potential PASC biological pathways, with candidate biomarkers cross-validated using regression analysis to assess their predictive potential.

Using this cohort, we confirmed that obesity associated with increased frequency of PASC. Obesity-associated PASC consisted of a molecular profile of persistently elevated inflammatory responses, corresponding to reduced humoral activity and subsequent loss of viral neutralization. These features also associated with a loss of microbial diversity, suggesting gut dysbiosis. Specifically, IPA identified the high mobility group box 1 (HMGB1) protein as the top pathway implicated by the data. Stepwise pathway analysis of our data highlighted the role of the HMGB1 pathway in regulating the microbiome-immune axis in PASC. Time-series regression was used to determine the mechanism order beginning with obesity-related modulation of the gut microbiome and its downstream effects on HMGB1 and viral neutralization. Significantly, baseline plasma levels of HMGB1 were predictive of neutralization capacity over the course of recovery, with increased levels associated with PASC. Importantly, our data demonstrate HMGB1 as a candidate biomarker for PASC with implications for prophylactic and therapeutic intervention.

Recruitment and Study Cohort

We recruited 48 COVID-19-recovered subjects residing in Oʻahu, Hawaiʻi between June and September 2020, excluding those who missed more than one follow-up. At the time of enrollment, no participants were vaccinated and most participants had recovered from mild symptoms, except for two obese participants who were hospitalized during infection. Participants at entry ranged from 10 to 101 + days since the acute infection phase and, notably, the distribution of time since infection to study recruitment/baseline assessment was similar for all BMI groups (Supplemental Fig. 1). Age range, sex, Body Mass Index (BMI), and ethnicity are shown in Table 1. To further explore the role of obesity in PASC, we stratified our cohort into normal (BMI less than 24.9), overweight (BMI between 25 and 29.9), and obese (BMI greater than 30) groups, and tested for differences between groups. BMI groups were balanced across age and sex but comprised different ethnic compositions as indicated.

Table 1

The table presents demographic information for recovered COVID-19 individuals, further stratified by body mass index (BMI) groups. Categorical variables were tested using Chi-Square analysis, and continuous variables were tested using Kruskal-Wallis one-way analysis of variance.
	Total cohort	BMI groups			P-value
	Total cohort	Normal	Overweight	Obese	P-value
Participants (N; %)	48	14 (29%)	12 (25%)	22 (46%)
Sex (N; %)	0.3
Female	28 (58%)	10 (71%)	5 (42%)	13 (59%)
Male	20 (42%)	4 (29%)	7 (58%)	9 (41%)
Age (mean ± SE)	41.61 ± 3.37	44.93 ± 6.89	54.82 ± 5.47	31.53 ± 4.01
Age categories (N; %)	0.14
0–18	8 (17%)	3 (21%)	0 (0%)	5 (23%)
18–35	11 (23%)	2 (14%)	2 (17%)	7 (32%)
35–50	12 (25%)	4 (29%)	3 (25%)	5 (23%)
50–100	5 (27%)	5 (36%)	6 (50%)	2 (9.1%)
Unknown	4 (8%)	0 (0%)	1 (8.3%)	3 (14%)
BMI (mean ± SE)	30.06 ± 1.13	21.15 ± 0.80	27.71 ± 0.42	37.00 ± 1.05
A1c (mean ± SE)	5.24 ± 0.23	4.74 ± 0.09	5.32 ± 0.66	5.50 ± 0.33
A1c categories (N; %)

Data collection

Figure 1 shows a schematic diagram of the data acquisition and study protocol. Study inclusion required proof of previous SARS-CoV-2 infection by RT-PCR testing, current noninfectious status by a negative RT-PCR result, and a positive immunoglobulin G (IgG) and/or immunoglobulin M (IgM) antibody test at the first study visit. Participants completed a comprehensive health questionnaire regarding their overall health and potential comorbidities. Symptom reports were collected at each visit. Participant samples were collected longitudinally for a duration of 6 weeks; providing one stool sample for microbial sequencing after the first week week, and blood samples every week, for separation into plasma and peripheral blood mononuclear cells (PBMCs).

Attenuated antibody response and viral neutralization capacity associates with obesity and is characteristic of PASC

To immunologically characterize PASC, we examined changes in the immune response to SARS-CoV-2 infection among individuals in the post-acute stage of recovery stratified by BMI categories. We assessed differences in generalized immunity by measuring an abundant serum antibody, IgG, involved in pathogen protection, using an anti-SARS-CoV-2 IgG ELISA test, and evaluated relative changes by log normalization. We observed lower abundance of anti-SARS-CoV-2 IgG in both the overweight and obese groups when compared to that of normal BMI, although this difference was only significant (p < 0.05) in the obese group (Fig. 2a). Additional analyte immunoassays for Human Coronavirus Ig revealed significantly lower levels of nucleocapsid (p < 0.001), spike trimer (p < 0.05), S1 subunit (p < 0.05), and RBD (p < 0.01) antibodies in obese individuals compared to that of normal BMI (Supplemental Fig. 2a-d). Although participants ranged from 10 to 101 + days since the acute infection phase at study entry and baseline assessment, this distribution was similar by BMI category (Fig. 2b), suggesting that the timing of post-acute recovery is not a contributing factor for the differences in antibody levels we observed to be associated with obesity.

Given the association between reduced humoral response in PASC, we investigated the relationship between antibody abundance and PASC symptomology in our cohort. We first examined the frequency of reported symptoms each week stratified by BMI categories. Initially, lingering COVID-19 symptoms were approximately 1.5-fold (p < 0.05) more frequent in the overweight and obese groups, compared to the normal BMI group. At the final follow-up, all groups still experienced lingering symptoms, but the overweight and obese groups exhibited a delayed recovery phenotype compared to the normal BMI group (Fig. 2c). We then measured relative levels of plasma-bound SARS-CoV-2 antibodies using a surrogate virus neutralization ELISA that mimics blocking of the receptor binding domain (RBD) of the SARS-CoV-2 spike glycoprotein and angiotensin-converting enzyme 2 (ACE2) human cell surface receptor. The obese group exhibited significantly less SARS-CoV-2 neutralization potential (p < 0.001), which was otherwise consistent between the normal and overweight groups (Fig. 2d). These findings indicate that the prolonged PASC symptoms in obesity associate with a lower capacity to maintain robust levels of viral neutralization over the course of recovery.

Next, we examined potential differences in immune cell composition as leukocyte populations have been shown to be perturbed in PASC, including the depletion of B cells.² Upon monoclonal antibody staining and sorting of PBMCs, we observed consistently low B cell populations and high monocyte populations in obese individuals compared to those with normal BMI at baseline (p < 0.05, B cells; p < 0.001, monocytes) and final follow-up (p < 0.001, B cells; p < 0.05, monocytes) (Fig. 2e and f). Cellular populations remained unchanged across all time points for all BMI categories. The diminished B cell population in obesity may explain the reduced levels of anti-SARS-CoV-2 antibodies and the lower degree of viral neutralization observed, which together corresponded to longer symptom duration and delayed recovery among obese individuals.

Obesity-associated inflammation underlies the attenuated immune response in PASC

Due to increased adiposity, increased inflammatory signaling is characteristic of obesity, which may be linked to PASC. Adipose tissue, considered an endocrine organ, secretes inflammatory cytokines and chemokines such as IL-6, TNF-α, and MCP-1, in addition to inflammatory adipokines produced by adipocytes.^14,15 Consequently, we assessed systemic inflammation in our cohort, using a panel of inflammatory biomarkers linked to PASC and COVID-19 severity, including MCP-1, IL-1β, IFN-γ, TNF-α, IL-8, and IL-6.^16,17 Individuals in the normal BMI group exhibited significantly lower levels of IL-6 (p < 0.05), IFN-γ (p < 0.01), and MCP-1 (p < 0.01) compared to the other BMI groups across all timepoints. Overweight individuals exhibited significantly reduced levels of IFN-γ (p < 0.05) and increased levels of IL-1β (p < 0.05) relative to those of normal BMI. Obese individuals exhibited consistently elevated levels of all the inflammatory biomarkers (Fig. 3a-f). We then determined the relationship between the levels of these inflammatory biomarkers and viral neutralization capacity. At baseline, viral neutralization negatively correlated with IL-1β (p < 0.05) and IFN-γ (p < 0.01), but positively correlated with TNF-α (p < 0.0001). These associations were not maintained at the final follow-up, however, most likely due to the dynamics of these inflammatory biomarkers among individuals in the normal and overweight groups (Supplemental Fig. 3a-f). Altogether, these data suggest that chronic inflammation in obesity is detrimental to neutralization capacity.

To examine this possibility further, we measured plasma C-reactive protein (CRP), a biomarker of systemic inflammation,¹⁸ using a CRP Human Instant ELISA™ Kit. Analysis of baseline and final follow-up time points demonstrated a strong inverse relationship between viral neutralization and CRP levels (Supplemental Fig. 4a and b), which were consistently elevated in the obese group (p < 0.01; Supplemental Fig. 4c). Circulating leptin was also measured using a Leptin Human Instant ELISA™ kit as a control for BMI categories due to its proportionality to the amount of adipose tissue,¹⁹ showing significant elevation in the obese BMI group compared to the normal BMI group (Supplemental Fig. 4d).

B-cell activating factor (BAFF) dysregulation in obesity explains attenuated viral neutralization capacity

Persistent inflammation promotes proliferation of innate immune cell lineages, ultimately depleting lymphocytes and leading to lymphopenia (observed in Fig. 2e).^20,21 This could potentially occur via the expansion of adipocytes that express B-cell activating factor (BAFF), a member of the TNF ligand superfamily that facilitates B-cell survival and maturation, both of which are dysregulated in obesity.^22,23 Therefore, we measured plasma levels of BAFF using the BAFF Human Instant ELISA™ Kit and examined its relationship to viral neutralization. We observed that higher levels of BAFF positively correlated with enhanced neutralization of SARS-CoV-2 (p < 0.01; Fig. 4a). We expanded this correlation analysis to our panel of SARS-CoV-2 specific antibodies and observed that concentrations of RDB, nucleocapsid, spike, and S1 antibodies were BAFF-dependent (p < 0.01; Fig. 4b). When stratifying based on BMI categories, we found that BAFF concentrations were significantly decreased in overweight (p < 0.001) and obese (p < 0.001) groups (Fig. 4c) relative to that of normal BMI, which were consistent across all timepoints. Mechanisms that regulate BAFF expression involve nitric oxide (NO), a byproduct of inflammatory monocytes that was elevated in our obese group.²⁴ PASC has also been linked to oxidative stress,⁷ which associates with higher BMI and has been shown to be accompanied by increased serum NO.²⁵ Using a Total Nitric Oxide Assay Kit to measure plasma levels of NO in our cohort, we observed a significant linear relationship between NO and BMI (p < 0.05; Fig. 4d). Altogether, these results suggest NO regulation of BAFF as a mechanism that explains how obesity-mediated inflammation attenuates viral neutralization capacity.

Gut microbial dysbiosis correlates with PASC

Overactivated inflammatory pathways linked to gut microbial dysbiosis have previously been implicated in PASC. Thus, we examined the degree to which microbial dysbiosis might account for the immune dysregulation we observed in our cohort by 16S-based metagenomics profiling. Figure 5 shows the relative abundance of phyla within the cohort and the phylum distributions between the BMI groups. ANOVA indicated altered composition across the BMI groups (p < 0.05), where the obese group exhibited reduced abundance of unclassified species.

Microbial health was also evaluated using the Simpson alpha diversity index and Firmicutes to Bacteroidetes (F:B) ratio. These indices provide insight into the ability of microbial constituents to modulate immunity upon infection, where low diversity and unbalanced F:B ratios link to worse health outcomes.^26–27 Alpha diversity was negatively affected by BMI (p < 0.01) (Fig. 6a). To associate microbial diversity with the PASC phenotype, individuals were regrouped based on the lower and upper 50% diversity cut-offs. Symptom frequencies were recalculated for each diversity group and plotted longitudinally. Normal and overweight individuals represented 83% of the higher diversity group, while 75% of obese individuals corresponded to the low diversity group. Additionally, individuals in the upper 50% cut-off displayed much less frequent PASC symptomology (Fig. 6b), suggesting that higher BMI constrain microbial diversity, as linked to the PASC phenotype.

To examine the influence of microbial diversity on immunity, we longitudinally assessed IgG fold-change for each diversity group. We observed that a poor maintenance of IgG over time corresponded with low microbial diversity (p < 0.05; Fig. 6c). Logarithmic transformation of IgG concentrations correlated with both alpha diversity and F:B ratios, wherein alpha diversity was shown to promote IgG concentrations (p < 0.05), while having a higher F:B ratio was detrimental to IgG concentrations (p < 0.05; Fig. 6e). Concurrently, elevated F:B ratios impeded viral neutralization capacity (p = 0.01; Fig. 6d), but were not affected by BMI (Fig. 6f). These data indicate both microbial diversity and F:B ratio influence immune response. Combined with the results above, this suggests elevated BMI modulates the gut microbiome composition, negatively impacting antibody production and viral neutralization to further exacerbate the PASC phenotype.

High mobility group box 1: a master regulator of diminished humoral response in PASC

Our above data suggests systemic inflammation, compounded by response to viral infection, is deterministic of immune response to SARS-CoV-2 during convalescence. Particularly, the dysregulation of the microbiome-immune axis under obesity-induced constraints largely contributes to reduced immune response. Further, we found inflammatory markers linked to COVID-19 severity and PASC, as highlighted by prior studies, also correlated with decreased viral neutralization underlying PASC. To uncover novel and/or less described biomarkers of PASC, we expanded our investigation to a comprehensive immunological panel used for immune monitoring. We performed a 65plex Luminex assay on plasma samples at baseline and final follow-up timepoints. These longitudinal data were integrated and analyzed using Ingenuity Pathway Analysis (IPA) software to predict potential biomarkers and impacted pathways relevant to PASC. According to IPA, the extracellular high-mobility group box 1 (HMGB1) pathway was the among the top enriched within our cohort. The top four pathways reported by IPA are displayed in Table 2. We chose to further investigate HMGB1 due to previous reports implicating this protein as a prospective biomarker and therapeutic target of COVID-19 severity.¹² HMGB1, being a central mediator of inflammation, generates the secretion of proinflammatory cytokines from dying and/or activated innate immune cells.^11,30 Additionally, HMGB1 signals through the receptor for advanced glycation end products (RAGE, a pattern recognition receptor)²⁸ and toll-like receptors (TLR2 and TLR4), to regulate both sterile and inflammatory responses.²⁹ Taken together, these mechanisms highlight the HMGB1 pathway in explaining PASC.

Table 2

Top canonical pathways enriched in our cohort identified by Ingenuity Pathway Analysis.
Pathway	P-value	Overlap
Airway Pathology in Chronic Obstructive Pulmonary Disease	3.58e-50	24.8% (28/113)
HMGB1 Signaling	4.56e-35	14.6% (23/157)
Role of Cytokines in Mediating Communication in between Immune Cells	7.24e-35	34.6% (18/52)
Role of Pattern Recognition Receptors in Recognition of Bacteria and Viruses	6.37e-34	15.3% (22/14)

A colorimetric ELISA Kit was used to profile HMGB1 protein levels across BMI groups, showing amplified levels in the obese group, as has been previously reported (p < 0.05; Fig. 7a).¹² Despite similar baseline HMGB1 levels across BMI groups, initial levels of HMGB1 negatively correlated with viral neutralization (p < 0.01; Fig. 7b), and we confirmed that HMGB1 associated with the proinflammatory markers TNF-α (p < 0.01), IFN-γ (p < 0.01), and MCP-1 (p < 0.05) (Fig. 7c). We also observed that HMGB1 concentration negatively associated with the levels of nucleocapsid (p < 0.05), spike (p < 0.01), S1 (p < 0.001), and RBD antibodies (p < 0.001; Fig. 7d). Since HMGB1 antagonizes BAFF production in other contexts,³¹ we explored this possibility in PASC, finding that indeed, the levels of HMGB1 negatively associated with that of BAFF (p < 0.05; Fig. 7e), and strongly associated with the inflammatory byproduct NO (p < 0.01; Fig. 7f). In addition to its role in systemic inflammation, HMGB1 also participates in local inflammatory processes of the gut and contributes to the regulation of intestinal homeostasis.^32,33 Indeed, in our cohort we observed that systemic HMGB1 measured in plasma associated with microbial dysbiosis, as shown by lower diversity (p < 0.05, Supplementary Fig. 5a) and higher F:B ratios (p < 0.01, Supplemental Fig. 5b). Together, these data indicate that HMGB1 dysregulation relates to BMI and plays a role in the microbiome-immune axis characterized in PASC.

Pathway analysis shows baseline HMGB1 levels predict viral neutralization capacity

Although previous studies describe independent links between obesity, inflammation, gut microbiome, and HMGB1, none show that these systems together influence PASC. In this study, we attempted to empirically identify a biological pathway to explain the obesity-associated attenuation of viral neutralization capacity we observed using a stepwise data-derived pathway analysis approach as a decision tree (Fig. 8). First, each variable was independently correlated to BMI, revealing associations between baseline viral neutralization capacity (p < 0.05), HMGB1 at final follow-up (p < 0.05), and various microbial abundances at the genus level (p < 0.05, Supplementary Table 1). The strongest association with BMI, microbial abundance, was therefore chosen as the first step in the pathway. While BMI-related genera did not directly influence neutralization, the Bifidobacteriaceae genus most strongly associated with baseline HMGB1 and was therefore chosen as the next step of the pathway. Baseline HMGB1 was used to predict HMGB1 and neutralization at both time points, using time-series regression analysis.

In pathway regression analysis, conducted in the order identified by our data, it was revealed that BMI indirectly affects neutralization capacity (Fig. 9). A one-unit increase in BMI associated with a 0.002% reduction in the relative abundance of the Bifidobacterium genus (p = 0.005), while a 1% reduction in Bifidobacteriaceae was associated with a 3% (p = 0.003) increase in HMGB1 concentration at baseline. Similarly, a one unit increase in baseline HMGB1 significantly associated with (1) 0.48% diminished baseline viral neutralization capacity (p < 0.001), (2) maintainance of elevated HMGB1 levels at the final follow-up by 0.38% (p < 0.001), and (3) a predicted loss of 0.44% of viral neutralization capacity at the final follow-up (p = 0.005). Following the delineation of our pathway, we conducted mediation analysis to investigate whether our step-wise variable relationships were mediated by a third variable. No mediation effects were observed.

The HMGB1 pathway influences the microbiome-immune axis, culminating in PASC

Obesity is known to increase PASC risk in COVID-19 patients.¹⁰ To resolve the molecular basis of PASC in the context of obesity in our cohort, we first confirmed that delayed viral recovery (demonstrated by post-COVID-19 symptom frequency) associated with higher BMI. We then identified a molecular phenotype comprised of persistent systemic inflammation, loss of maintenance of humoral responses, and dysbiosis of the gut microbiome. Time-series regression analysis showed that initial HMGB1 levels predicted neutralization capacity at baseline and after 6 weeks during convalescence.

Analysis of our molecular data revealed numerous biological pathways involved in PASC (Fig. 10). For example, obesity-associated dysbiosis of the gut microbiome likely allows SARS-CoV-2 to exploit inflammation and enhance TNF-α, IL-1β, and IFN-γ cytokine secretion. These inflammatory factors then activate local and peripheral monocytes, upregulating HMGB1 excretion to further activate monocytes and macrophages, producing a proinflammatory state with released NO as a byproduct. Increased NO bioavailability then inhibits BAFF expression (by inducing B cell apoptosis), with loss of B cell propagation decreasing antibody production and thus viral neutralization capacity, prolonging acute COVID-19 symptoms or PASC. This PASC-promoting molecular cascade is particularly susceptible to obesity-mediated dysregulation. Our study highlights the confluence of obesity-associated dysfunction and immune responses in PASC, in which preexisting dysfunction contributes to a more dramatic inflammatory response to viral infection (thus reducing the effectiveness of the immune system). Our model (Fig. 10) suggests that this effect is exacerbated by upregulated HMGB1 due to various possible mechanisms, some of which we identified through pathway analysis.

In our cohort, obese individuals showed greater microbial dysbiosis which has been linked to COVID-19 severity and PASC, irrespective of BMI.^34,36,37 While microbial dysbiosis has been widely observed in obesity,³⁴ we also observed such impairment, although less frequently in non-obese individuals. In vivo mouse studies have shown that SARS-CoV-2 exposure alters gut microbial composition to reduce diversity, in unison with morphological changes that promote mucus and reduce the production of antimicrobial compounds.³⁵ In particular Bifidobacterium abundance has been extensively linked to human health due to its production of health-promoting short-chain fatty acids and protection against infection,³⁶ and we identified this genus to most strongly influence the PASC phenotype, where its lower abundance led to reduced viral neutralization.

This study is not the first to link Bifidobacterium with SARS-CoV-2 infection. For example, loss of Bifidobacterium abundance has been linked to both COVID-19 severity³⁷ and PASC existing 6-months post-infection.³⁸ However, we are the first to identify HMGB1 as a mediator of this association, as characterized in other contexts, e.g., where inoculation with probiotic Bifidobacterium animalis subsp. lactis BB-12 (BB12) decreased HMGB1 levels and partly ameliorated inflammatory responses following infection with the Salmonella Typhimurium strain LT2.³⁹ While our linear regression model suggests that Bifidobacterium modulates HMGB1 reciprocally, HMGB1 may also indirectly modulate Bifidobacterium abundance in PASC pathology, as Bifidobacterium genus members participate in quorum sensing for gut colonization and pathogen protection dependent on iron acquisition.⁴⁰ Iron dysmetabolism has also been observed in SARS-CoV-2 infection, where iron deficiency and anemia of inflammation from excessive ferritin correlate with COVID-19 severity and severe lung pathologies.⁴¹ In extreme responses to infection (sepsis), elevated ferritin and HMGB1 co-occur and contribute to immune activation that results in a cytokine storm.⁴² Moreover, under chronic inflammatory conditions, the development of erythropoietin-refractory anemia (commonly known as anemia of inflammation), being linked to obesity,⁴³ is moderated by HMGB1 inhibition of erythropoiesis due to elimination of erythroid precursors.⁴⁴

Higher serum nitrate levels have been described in individuals who succumbed to COVID-19 illness when compared with COVID-19 survivors.^45,46 However, NO levels in COVID-19 patients showed lower levels of NO when compared to healthy controls,⁴⁶ further suggesting the overamplification of NO is a detrimental phenotype in COVID-19. While NO overproduction in inflammatory diseases occurs systemically, it is also linked to the intestinal lumen. In that regard, in vitro fermentation systems have shown that NO stress modifies microbial composition and functionality. Moreover, NO further induces microbial dysbiosis by depleting butyrate-producing species and selecting for pathogenic species that can withstand high concentrations of NO.⁴⁷ In addition, gut microorganisms can produce NO that diffuses into the blood stream, thus amplifying circulating NO, to contribute to the PASC phenotype through our identified pathway. Our study corroborates a positive relationship between NO bioavailability and HMGB1, where elevated HMGB1 increases NO levels, with several prior studies suggesting that both the gut microbiome and disrupted immune responses contribute to this interaction.^48,49 Considering the regulatory role of HMGB1 in intestinal homeostasis, via a bidirectional relationship with resident microbes,^50,32,33 we expect this balance to be particularly important in diseases linked to chronic inflammation and altered gut microbiomes.^50–51

In our study, COVID-19 recovering individuals with different degrees of microbial dysbiosis exhibited persistent inflammatory signaling of IL-1b, TNF-a, and IFN-γ. Microbial dysbiosis is accompanied by elevated lipopolysaccharides that promote inflammation, in part, by directly activating HMGB1 in resident immune and epithelial cells,³³ while IL-1b inflammatory signaling activates monocytes and stimulates proinflammatory pathways in peripheral tissues.⁵² We observed this phenomenon in the expansion of peripheral blood monocytes in patients with high BMI, a characteristic of obesity.⁵³ Moreover, these monocytes are known to be chronically activated. In addition to proinflammatory signaling, inflammatory monocytes can serve as reservoirs for viral replication and store increased viral titers.⁵⁴ For example, Patterson et al. recently discovered that individuals with severe COVID-19 and PASC, 15 months post-infection, had persistent SARS-CoV-2 S1 protein in non-classical monocytes.⁵⁵

Under obesogenic conditions, elevated HMGB1 has been reported in adipose tissue acting as an alarm that leads to further HMGB1 induction.^30,12,56 As previously mentioned HMGB1 signals through RAGE receptors, shown to be deleterious to the renin-angiotensin system by sustaining inflammation and oxidative stress involved in COVID-19 infection and multiorgan injury⁵⁷ and possibly contributes to multi-organ PASC involvement. Moreover, angiotensin-converting enzyme 2 (ACE2), the SARS-CoV-2 entry receptor, is upregulated by HMGB1, creating a docking station and cache for the SARS-CoV-2 virus. When increased adiposity, HMGB1 overexpression, and SARS-CoV-2 reservoirs coexist, inflammation is persistently amplified. When accompanied by systemic inflammation, additional signaling from SARS-CoV-2 overburdens these processes and delays viral recovery. In that regard, Faizo et al. observed significantly reduced rates and titers of vaccine-induced spike-neutralizing antibodies in obese individuals,⁵⁸ while other recent studies showed that loss of neutralizing antibodies in obese COVID-19 patients with circulating autoimmune antibodies positively correlated with CRP levels and COVID-19 hospitalizations.^59,60 Although we did not expand upon the accumulation of autoantibodies in our cohort, these studies along with our data further indicate that deficiencies in adaptive immunity are inflammation-mediated and impacted by adiposity.

Altogether, our findings show that increased HMGB1 pathway signaling in obesity impairs adaptive immune response to SARS-CoV-2. Such loss of robust immunity in turn propagates PASC in these individuals and can persist chronically. Taken together, our findings suggest HMGB1 could be used as a predictor of PASC risk and a potential candidate for disease mitigation. Although our study was limited by the small sample size due to attrition, we were still powered to examine molecular differences between BMI categories that stratified PASC. While our study cohort may not represent the full spectrum of PASC, our findings provide a valuable insight into the underlying mechanisms of this growing condition and offers a strategy for a quantitative diagnosis.

Human study participants

The protocol for this study was approved by the University of Hawaii Office of Research Compliance 2019 − 00376 and performed in accordance to said guidelines. Informed consent was obtained from all participants and/or their legal guardians. Recruitment was advertised through flyers posted at the University of Hawaiʻi at Manoa and media outlets. Data were collected from Oʻahu residents who tested positive for COVID-19 at any time during the recruitment period, but not longer than three months prior to the first data collection. Participants were eligible for the study if they provided proof of a previous positive nasopharyngeal RT-PCR test but were currently non-infectious as deemed by a negative RT-PCR result. Initially, 71 participants were enrolled between June 1, 2020 and September 25, 2020. Following enrollment, confirmation of a previous SARS-CoV-2 infection was determined by specific IgG/IgM rapid antibody testing, at the first visit, by a finger blood prick. A positive result for SARS-CoV-2 IgG or IgM antibodies was necessary to advance the into the study; five participants were excluded due to negative antibodies. During the first visit, participants completed a health questionnaire, provided demographic data, and received a stool collection kit. To ensure privacy, all participants, and their respective survey responses were anonymized using a unique numerical ID. Participants were asked to return weekly for five additional follow-up visits. During the second follow-up, subjects submitted a stool sample and recorded their height and weight, for BMI calculation by dividing weight in kilograms by the square of height in meters. At each of the five follow-up visits, participants provided a blood sample and reported symptoms such as fever, cough, difficulty breathing, fatigue, muscle or body aches, headache, new loss of taste or smell, sore throat, congestion or runny nose, nausea, and diarrhea. Symptom percentages for each follow-up visit were calculated by summing the number of participants experiencing any number of symptoms, in that week, divided by the total number of participants having no symptoms during that week. Symptoms were deemed resolved if no longer reported during any of the follow-up visits. Individuals who missed more than two appointments were excluded from the study. A total of 48 participants were included in the final analysis. All participants received monetary incentives after each weekly appointment.

IgG/IgM Antibody detection test

Rapid patient screening for SARS-CoV-2 specific antibody tests was performed using RayBio® Coronavirus (SARS-CoV-2) IgM/IgG test kits (Colloidal Gold Method) (RayBiotech, Peachtree Corners, GA, USA, cat #CG-CoV-IgM, CG-CoV-IgG) to confirm COVID-19 detection in whole blood from a finger prick.

Blood Processing

Blood was processed within 24 h of collection to separate plasma and PBMCs using PBMC isolation tubes (STEMCELL Technologies, Vancouver, BC, Canada). Plasma samples were kept at -80 C, and PBMCs at -150°C, until ready to assay.

Stool sample DNA/RNA extraction

DNA and RNA were extracted from stool samples using AllPrep® PowerFecal® DNA/RNA Kits (Qiagen. Inc., Valencia, CA, USA). From each sample, 1–2µl of purified DNA and RNA were analyzed using a NanoDrop™ Microvolume Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) to assess sample quality, in addition to a Qubit™ Fluorometer (Thermo Fisher).

16S rRNA Sequencing and Analysis

An Ion Torrent 16S™ Metagenomics Kit (Thermo Fisher) was used to amplify 16S rRNA genes of all stool samples. Following the amplification conditions indicated in the manufacturer’s protocol, PCR was performed in two pools, each containing specific primer sets. After PCR amplification, equal volumes (20 µL) of PCR products from pools 1 and 2 of each sample were combined into a single PCR tube. Then, 40 µL of the combined PCR product was purified using the Agencourt Ampure XP kit, according to the manufacturer’s instructions (Beckman Coulter, Brea, CA, USA). DNA was quantified using the Qubit Fluorometer with the Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific, Warrington, England), and the DNA library was prepared using the Ion Plus Fragment Library Kit (Thermo Fisher Scientific, Austin, TX, USA) and Ion Xpress Barcodes Adapters 1–80 Kit (Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol. Each library was diluted to a concentration of 80 pM and equal volumes of each library were pooled. The Ion 510™ & Ion 520™ & Ion 530™ Kit (Thermo Fisher Scientific, Austin, TX, USA) was used to clonally amplify the pooled library of nanosized ionosphere particles by emulsion PCR. Bead enrichment and chip loading were conducted using an Ion Chef Instrument. The parallel sequencing reaction was performed using the Ion S5 Next-Generation Sequencing system with Ion 530 chips.

After obtaining sequencing reads (Supplemental Table 1), we performed rarefaction at n = 10,000, on microbial absolute abundance, to eliminate sample bias from uneven counts, further excluding six samples, to leave 42 for further analysis. The results were analyzed using the Ion Reporter Software on the Ion 16S Metagenomics Kit analysis module. In the scope of analysis, the reads were mapped to two reference 16S rRNA databases, Greengenes v13.5 and MicroSEQ ID v3.0, following the manufacturer’s instructions. Chimeric sequences were automatically identified and removed using Ion Reporter software.^61,62 To cluster the sequences, operational taxonomic unit (OTU) analysis was performed using Ion Reporter Software, which runs Quantitative Insights into Microbial Ecology (QIIME). Alpha and beta diversity graphics created by QIIME software were exported from Ion Reporter Software. For alpha diversity analysis, the observed species and Shannon, Fisher, and Simpson indices were generated to analyze species diversity within samples.

Antibody detection ELISA

The SeroFlash™ SARS-CoV-2 IgG/IgM ELISA Fast Kit (Epigentek Inc., Farmingdale, NY, USA) was used to quantify SARS-CoV-2 spike protein IgM and IgG antibodies in plasma samples. For further confirmation and characterization, the Human Coronavirus Ig Total 11-Plex ProcartaPlex™ Panel (Thermo Fisher) was used for screening of four anti-SARS-CoV-2 antibodies (spike trimer, S1 subunit, receptor binding domain (RBD), and nucleocapsid), and six anti-coronavirus strains (SARS-CoV-1, MERS, CoV-NL63, CoV-KHU1, CoV-229E, and CoV-OC43). Fluorescent signals were analyzed using a Luminex 200 instrument (R&D Systems, Inc., Minneapolis, MN, USA), and data processed using Bio-Plex Manager™ software (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Inflammation factors concentration measurements

Inflammatory and metabolic biomarkers: Exploration of immunological biomarkers was performed for plasma cytokine concentrations of G-CSF (CSF-3), GM-CSF, IFN alpha, IFN gamma, IL-1 alpha, IL-1 beta, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8 (CXCL8), IL-9, IL-10, IL-12p70, IL-13, IL-15, IL-16, IL-17A (CTLA-8), IL-18, IL-20, IL-21, IL-22, IL-23, IL-27, IL-31, LIF, M-CSF, MIF, TNF alpha, TNF beta, TSLP; chemokines BLC (CXCL13), ENA-78 (CXCL5), Eotaxin (CCL11), Eotaxin-2 (CCL24), Eotaxin-3 (CCL26), Fractalkine (CX3CL1), Gro-alpha (CXCL1), IP-10 (CXCL10), I-TAC (CXCL11), MCP-1 (CCL2), MCP-2 (CCL8), MCP-3 (CCL7), MDC (CCL22), MIG (CXCL9), MIP-1 alpha (CCL3), MIP-1 beta (CCL4), MIP-3 alpha (CCL20), SDF-1 alpha (CXCL12; growth factors FGF-2, HGF, MMP-1, NGF beta, SCF, VEGF-A and soluble receptors APRIL, BAFF, CD30, CD40L (CD154), IL-2R (CD25), TNF-RII, TRAIL (CD253) and TWEAK, using a human Immune Monitoring 65-Plex Human Procarta™ Plex Panel (ThermoFisher part no. EPX650-10065-901), according to the manufacturer’s instructions. The samples were then centrifuged at 13,000 × g for 2 min to pellet the aggregates, with a standard curve generated using the manufacturer's antigen standards. Bead counts below 35 were insufficient for the analysis and were excluded from the analysis. Fluorescent signals were analyzed using a Luminex 200 instrument (R&D Systems, Inc., Minneapolis, MN, USA). The Bio-Plex Manager™ software (Bio-Rad Laboratories, Inc., Hercules, CA, USA) was used for data processing.

Other Biomarkers: To evaluate the general indications of inflammation, CRP plasma levels were measured using a Human CRP Instant ELISA Kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer’s instructions. The samples were diluted 1:500 with assay buffer before use. Samples that were above the standard curve were further diluted and assayed a second time. The concentrations of NO were assessed using the Total Nitric Oxide Assay Kit (catalog no. EMSNOTOT, Thermo Fisher) according to the manufacturer’s instructions. BAFF measurements were performed using the BAFF Human Instant™ ELISA Kit (catalog no. BMS2007INST Thermo Fisher), according to the manufacturer’s instructions. HMGB1 plasma levels were assessed using a colorimetric Elisa Kit (Part No. NBP2-62766, Novus Biologicals, Littleton, CO, USA), according to the manufacturer’s instructions, without sample dilution. All ELISA assays were performed using a SpectraMax ABS/ABS Plus Microplate Reader (Molecular Devices, San Jose, CA, USA). Leptin plasma concentrations were measured using a Leptin Human Instant™ ELISA kit (Thermo Fisher), according to the manufacturer’s instructions. Samples were diluted 1:25 with assay buffer before use.

PBMC Phenotyping

Viably cryopreserved PBMCs of approximately 1.0 × 10⁷ from individuals, at weeks 2 and 6 of our protocol, were first thawed in AIM-V serum-free media (Thermo Fisher) supplemented with 1:50 DNase (Sigma-Aldrich, St. Louis, MO, USA), washed, and resuspended in wash buffer (PBS, 3% BSA, and 1 mM EDTA). Aliquots of 1.25 × 10⁵ PBMCs were taken from each sample for flow cytometry-based cellular phenotyping assays, to determine cell type compositions. Cells were counted using a Countess Automated Cell Counter (Life Technologies, Carlsbad, CA, USA). Aliquots were stained with yellow amine fluorescent reactive dye (YARD; Thermo Fisher Scientific) and then with anti-CD16 Brilliant Violet 421 (Clone 3G8), anti-CD3 V500 (Clone UCHT1), anti-CD14 Qdot®605 (Clone TüK4), anti-CD56 Pe-Cy7 (Clone B159), anti-CD19 PE-Cy7 (Clone SJ25C1), anti-CD20 Pe-Cy7 (Clone 2H7), and anti-HLA-DR APC-H7 (Clone G46-6) to identify leukocyte subpopulation frequencies. Anti-CD16 was purchased from BioLegend, Inc. (San Diego, CA, USA). Anti-CD3, anti-CD56, anti-CD20, anti-CD19, and anti-HLA-DR antibodies were obtained from BD Biosciences (San Jose, CA). Anti-mouse Ig/Negative Control (FBS) Compensation Particle Set (BD Biosciences) was used for compensation analysis of fluorescent signals emitted by each fluorochrome from the multicolored cellular phenotyping panel employed. Anti-mouse Ig compensation beads were stained with fluorochrome-conjugated antibodies in separate wells. ArC™ Amine Reactive Compensation Bead Kit (ThermoFisher) reactive bead/negative beads were used to compensate for YARD (Live/Dead stain) fluorescent signals. Stained cells from PBMCs and compensation particles were analyzed using a 4-laser BD LSRFortessa flow cytometer (BD Biosciences). Data were analyzed using FlowJo software (Tree Star, Inc., Ashland, OR, USA). The frequency (%) of each cell type was determined using the event count (specific events/total events) with debris exclusion.

Statistical analysis

Comparative analyses of clinical, immunological, cell flow cytometry, and qPCR data were performed using ANOVA Mixed-effects analysis with Sidak’s multiple comparisons test correction. Non-parametric tests for clinical, immunological, and cell flow cytometry data were used primarily because the majority of data were non-normally distributed, as measured using the D’Agostino and Pearson tests for normality. To predict associative outcomes, several stepwise multiple logistic regression models were tested for clinical features, immunological features, and qPCR results that were identified by Wilcoxon Rank-Sum tests followed by multiple testing corrections (FDR-adjusted p < 0.01). Graphing and statistical analyses were performed using Prism 9, version 9.0c (GraphPad Software, La Jolla, CA, USA). Statistical significance was set at p < 0.05.

A t-test was performed to assess the similarity between the overall (n = 48) and rarefied (n = 42) gut microbiome datasets. Correlation analysis was used to assess the association between the relative abundance of the gut microbiome and BMI. Moreover, path analysis was used to shape the research question of how differences in BMI impact PASC.

Ingenuity Pathway Analysis (IPA)

The differential expression of cytokines, chemokines, and growth factors was used as an input for the Ingenuity Pathway Analysis (IPA) (Qiagen, United States). The core analysis module in IPA was used to deduce differentially regulated canonical pathways, upstream regulators, diseases and biological functions, and novel gene networks based on Fisher’s exact test (p-value cutoff at 0.05).

SARS-CoV‐2 surrogate virus neutralization test (sVNT)

The surrogate virus neutralization test (GenScript, NJ, USA) is a blocking ELISA that mimics the virus neutralization process, detecting circulating neutralizing SARS-CoV‐2 antibodies that block the interaction between the RBD and ACE2 on the cell surface receptor of the host. The test was isotype- and species-independent. Plasma samples were diluted 10x with sample dilution buffer and assayed according to the GenScript protocol. The absorbance of the sample was inversely dependent on the titer of the anti-SARS-CoV-2 neutralizing antibodies.

Data Availability Statement

Data is available upon reasonable request to AKM.

Acknowledgements

We would like to thank all participants who made this study possible. We greatly appreciate the support for this study from Hawaiʻi Integrated Analytics (HIʻA). This manuscript and its contents are solely the responsibility of the authors and do not represent the official view of HIʻA or sponsors. Illustrations were created using BioRender.com.

Author Contributions

RJ and AKM concieved of the study and obtained funding. NR, NA, AKM wrote the manuscript. NR, RP, BK, TM, RL, RJ and AKM obtained the data. NR, RW, LU, AC performed experimental assays. NR, RP, KP, and AKM analyzed the data. All authors contributed to the article and approved the submitted version.

Statement of Ethics

This study and its protocol were approved by the University of Hawaiʻi Office of Research Compliance 2019-00376.

Competing Interests Statement

The authors of this research paper declare the following potential conflicts of interest regarding patentable product associated with this study. Ruben Juarez and Alika Maunakea from HIʻA may file a patent application related to technology described in this paper. Furthermore, Ruben Juarez and Alika Maunakea have a financial interest in HIʻA, a company that holds the exclusive license for commercializing of this technology. The involvement of the authors in the patent and the company does not influence the objectivity, integrity, or validity of the research findings presented in this paper. All data and results have been analyzed impartially, and appropriate steps have been taken to mitigate any potential bias. This conflict of interest disclosure is made to ensure transparency and maintain the highest ethical standards in scientific research. The other authors declare no competing interests exist.

Funding Statement

This project was funded by Hawaii Community Foundation grant number 20HCF-101573, Army grant number W911NF21P0025 and National Science Foundation award number 2036240.

Phetsouphanh C, Darley DR, Wilson DB, et al. Immunological dysfunction persists for 8 months following initial mild-to-moderate SARS-CoV-2 infection. Nat Immunol. 2022;23(2):210-216. doi:10.1038/s41590-021-01113-x
Davitt E, Davitt C, Mazer MB, Areti SS, Hotchkiss RS, Remy KE. COVID-19 disease and immune dysregulation. Best Pract Res Clin Haematol. 2022;35(3):101401. doi:10.1016/j.beha.2022.101401
Su Q, Lau RI, Liu Q, Chan FKL, Ng SC. Post-acute COVID-19 syndrome and gut dysbiosis linger beyond 1 year after SARS-CoV-2 clearance. Gut. Published online August 8, 2022. doi:10.1136/gutjnl-2022-328319
Liu Q, Mak JWY, Su Q, et al. Gut microbiota dynamics in a prospective cohort of patients with post-acute COVID-19 syndrome. Gut. 2022;71(3):544-552. doi:10.1136/gutjnl-2021-325989
Aminian A, Bena J, Pantalone KM, Burguera B. Association of obesity with postacute sequelae of COVID-19. Diabetes Obes Metab. 2021;23(9):2183-2188. doi:10.1111/dom.14454
Scherer PE, Kirwan JP, Rosen CJ. Post-acute sequelae of COVID-19: A metabolic perspective. Zaidi M, ed. eLife. 2022;11:e78200. doi:10.7554/eLife.78200
Pierce JD, Shen Q, Cintron SA, Hiebert JB. Post-COVID-19 Syndrome. Nurs Res. 2022;71(2):164-174. doi:10.1097/NNR.0000000000000565
CDC. Post-COVID Conditions. Centers for Disease Control and Prevention. Published September 1, 2022. Accessed December 12, 2022. https://www.cdc.gov/coronavirus/2019-ncov/long-term-effects/index.html
Post-acute sequelae of COVID-19 (PASC): a meta-narrative review of pathophysiology, prevalence, and management - PMC. Accessed February 22, 2023. https://www-ncbi-nlm-nih-gov.eres.library.manoa.hawaii.edu/pmc/articles/PMC8977184/
Vimercati L, De Maria L, Quarato M, et al. Association between Long COVID and Overweight/Obesity. J Clin Med. 2021;10(18):4143. doi:10.3390/jcm10184143
Extracellular HMGB1: a therapeutic target in severe pulmonary inflammation including COVID-19? | Molecular Medicine | Full Text. Accessed February 22, 2023. https://molmed.biomedcentral.com/articles/10.1186/s10020-020-00172-4
Chen R, Huang Y, Quan J, et al. HMGB1 as a potential biomarker and therapeutic target for severe COVID-19. Heliyon. 2020;6(12):e05672. doi:10.1016/j.heliyon.2020.e05672
Tjendra Y, Al Mana AF, Espejo AP, et al. Predicting Disease Severity and Outcome in COVID-19 Patients: A Review of Multiple Biomarkers. Archives of Pathology & Laboratory Medicine. 2020;144(12):1465-1474. doi:10.5858/arpa.2020-0471-SA
Mohamed-Ali V, Pinkney JH, Coppack SW. Adipose tissue as an endocrine and paracrine organ. Int J Obes. 1998;22(12):1145-1158. doi:10.1038/sj.ijo.0800770
Fantuzzi G. Adipose tissue, adipokines, and inflammation. Journal of Allergy and Clinical Immunology. 2005;115(5):911-919. doi:10.1016/j.jaci.2005.02.023
Schultheiß C, Willscher E, Paschold L, et al. The IL-1β, IL-6, and TNF cytokine triad is associated with post-acute sequelae of COVID-19. Cell Reports Medicine. 2022;3(6):100663. doi:10.1016/j.xcrm.2022.100663
Chen Y, Wang J, Liu C, et al. IP-10 and MCP-1 as biomarkers associated with disease severity of COVID-19. Molecular Medicine. 2020;26(1):97. doi:10.1186/s10020-020-00230-x
Karadag F, Kirdar S, Karul AB, Ceylan E. The value of C-reactive protein as a marker of systemic inflammation in stable chronic obstructive pulmonary disease. Eur J Intern Med. 2008;19(2):104-108. doi:10.1016/j.ejim.2007.04.026
Martins M do C, Lima Faleiro L, Fonseca A. Relationship between leptin and body mass and metabolic syndrome in an adult population. Revista Portuguesa de Cardiologia (English Edition). 2012;31(11):711-719. doi:10.1016/j.repce.2012.09.008
Mishra KP, Singh AK, Singh SB. Hyperinflammation and Immune Response Generation in COVID-19. Neuroimmunomodulation. 2020;27(2):80-86. doi:10.1159/000513198
Wong RSY. Inflammation in COVID-19: from pathogenesis to treatment. Int J Clin Exp Pathol. 2021;14(7):831-844.
Kim YH, Choi BH, Cheon HG, Do MS. B cell activation factor (BAFF) is a novel adipokine that links obesity and inflammation. Exp Mol Med. 2009;41(3):208-216. doi:10.3858/emm.2009.41.3.024
Do MS, Jeong HS, Choi BH, et al. Inflammatory Gene Expression Patterns Revealed by DNA Microarray Analysis in TNF-α-treated SGBS Human Adipocytes. Yonsei Medical Journal. 2006;47(5):729-736. doi:10.3349/ymj.2006.47.5.729
Giordano D, Draves KE, Li C, Hohl TM, Clark EA. Nitric oxide regulates B cell activating factor (BAFF) expression and T-cell independent antibody responses. J Immunol. 2014;193(3):1110-1120. doi:10.4049/jimmunol.1303158
Foroumandi E, Alizadeh M, Kheirouri S, Asghari Jafarabadi M. Exploring the role of body mass index in relationship of serum nitric oxide and advanced glycation end products in apparently healthy subjects. PLoS One. 2019;14(3):e0213307. doi:10.1371/journal.pone.0213307
Hagerty SL, Hutchison KE, Lowry CA, Bryan AD. An empirically derived method for measuring human gut microbiome alpha diversity: Demonstrated utility in predicting health-related outcomes among a human clinical sample. PLOS ONE. 2020;15(3):e0229204. doi:10.1371/journal.pone.0229204
Wagner BD, Grunwald GK, Zerbe GO, et al. On the Use of Diversity Measures in Longitudinal Sequencing Studies of Microbial Communities. Front Microbiol. 2018;9. doi:10.3389/fmicb.2018.01037
Zhang J, Zhang L, Zhang S, et al. HMGB1, an innate alarmin, plays a critical role in chronic inflammation of adipose tissue in obesity. Molecular and Cellular Endocrinology. 2017;454:103-111. doi:10.1016/j.mce.2017.06.012
Ibrahim ZA, Armour CL, Phipps S, Sukkar MB. RAGE and TLRs: Relatives, friends or neighbours? Molecular Immunology. 2013;56(4):739-744. doi:10.1016/j.molimm.2013.07.008
Holms RD. Long COVID (PASC) Is Maintained by a Self-Sustaining Pro-Inflammatory TLR4/RAGE-Loop of S100A8/A9 > TLR4/RAGE Signalling, Inducing Chronic Expression of IL-1b, IL-6 and TNFa: Anti-Inflammatory Ezrin Peptides as Potential Therapy. Immuno. 2022;2(3):512-533. doi:10.3390/immuno2030033
Kalled SL. The role of BAFF in immune function and implications for autoimmunity. Immunological Reviews. 2005;204(1):43-54. doi:10.1111/j.0105-2896.2005.00219.x
Vasconcelos RM, Sanfilippo N, Paster BJ, et al. Host-Microbiome Cross-talk in Oral Mucositis. J Dent Res. 2016;95(7):725-733. doi:10.1177/0022034516641890
Wang FC, Pei JX, Zhu J, et al. Overexpression of HMGB1 A-box reduced lipopolysaccharide-induced intestinal inflammation via HMGB1/TLR4 signaling in vitro. World Journal of Gastroenterology. 2015;21(25):7764-7776. doi:10.3748/wjg.v21.i25.7764
Geng J, Ni Q, Sun W, Li L, Feng X. The links between gut microbiota and obesity and obesity related diseases. Biomedicine & Pharmacotherapy. 2022;147:112678. doi:10.1016/j.biopha.2022.112678
Gut microbiome dysbiosis in antibiotic-treated COVID-19 patients is associated with microbial translocation and bacteremia - PMC. Accessed March 6, 2023. https://www-ncbi-nlm-nih-gov.eres.library.manoa.hawaii.edu/pmc/articles/PMC9626559/
Arboleya S, Watkins C, Stanton C, Ross RP. Gut Bifidobacteria Populations in Human Health and Aging. Frontiers in Microbiology. 2016;7. Accessed April 2, 2023. https://www.frontiersin.org/articles/10.3389/fmicb.2016.01204
Hazan S, Stollman N, Bozkurt HS, et al. Lost microbes of COVID-19: Bifidobacterium, Faecalibacterium depletion and decreased microbiome diversity associated with SARS-CoV-2 infection severity. BMJ Open Gastroenterol. 2022;9(1):e000871. doi:10.1136/bmjgast-2022-000871
Liu Q, Mak JWY, Su Q, et al. Gut microbiota dynamics in a prospective cohort of patients with post-acute COVID-19 syndrome. Gut. 2022;71(3):544-552. doi:10.1136/gutjnl-2021-325989
Splichal I, Donovan SM, Kindlova Z, et al. Release of HMGB1 and Toll-like Receptors 2, 4, and 9 Signaling Are Modulated by Bifidobacterium animalis subsp. lactis BB-12 and Salmonella Typhimurium in a Gnotobiotic Piglet Model of Preterm Infants. Int J Mol Sci. 2023;24(3):2329. doi:10.3390/ijms24032329
Christiaen SEA, O’Connell Motherway M, Bottacini F, et al. Autoinducer-2 plays a crucial role in gut colonization and probiotic functionality of Bifidobacterium breve UCC2003. PLoS One. 2014;9(5):e98111. doi:10.1371/journal.pone.0098111
Sonnweber T, Boehm A, Sahanic S, et al. Persisting alterations of iron homeostasis in COVID-19 are associated with non-resolving lung pathologies and poor patients’ performance: a prospective observational cohort study. Respir Res. 2020;21(1):276. doi:10.1186/s12931-020-01546-2
Karakike E, Giamarellos-Bourboulis EJ. Macrophage Activation-Like Syndrome: A Distinct Entity Leading to Early Death in Sepsis. Frontiers in Immunology. 2019;10. Accessed April 28, 2023. https://www.frontiersin.org/articles/10.3389/fimmu.2019.00055
Weiss G, Ganz T, Goodnough LT. Anemia of inflammation. Blood. 2019;133(1):40-50. doi:10.1182/blood-2018-06-856500
Dulmovits BM, Papoin J, Li JH, et al. HMGB1 Causes Anemia of Inflammation By Modulating Erythropoietin Signal Transduction. Blood. 2018;132:628. doi:10.1182/blood-2018-99-119553
Lorente L, Gómez-Bernal F, Martín MM, et al. High serum nitrates levels in non-survivor COVID-19 patients. Med Intensiva. 2022;46(3):132-139. doi:10.1016/j.medin.2020.10.003
Dominic P, Ahmad J, Bhandari R, et al. Decreased availability of nitric oxide and hydrogen sulfide is a hallmark of COVID-19. Redox Biology. 2021;43:101982. doi:10.1016/j.redox.2021.101982
Leclerc M, Bedu-Ferrari C, Etienne-Mesmin L, et al. Nitric Oxide Impacts Human Gut Microbiota Diversity and Functionalities. mSystems. 2021;6(5):e0055821. doi:10.1128/mSystems.00558-21
Jiang W, Pisetsky DS. The Role of IFN-α and Nitric Oxide in the Release of HMGB1 by RAW 264.7 Cells Stimulated with Polyinosinic-Polycytidylic Acid or Lipopolysaccharide. The Journal of Immunology. 2006;177(5):3337-3343. doi:10.4049/jimmunol.177.5.3337
Chakraborty R, Bhatt KH, Sodhi A. High mobility group box 1 protein synergizes with lipopolysaccharide and peptidoglycan for nitric oxide production in mouse peritoneal macrophages in vitro. Mol Immunol. 2013;54(1):48-57. doi:10.1016/j.molimm.2012.10.042
Biscetti F, Flex A, Alivernini S, Tolusso B, Gremese E, Ferraccioli G. The Role of High-Mobility Group Box-1 and Its Crosstalk with Microbiome in Rheumatoid Arthritis. Mediators Inflamm. 2017;2017:5230374. doi:10.1155/2017/5230374
Chen Y, Wu D, Sun L. Clinical Significance of High-Mobility Group Box 1 Protein (HMGB1) and Nod-Like Receptor Protein 3 (NLRP3) in Patients with Ulcerative Colitis. Med Sci Monit. 2020;26:e919530-1-e919530-8. doi:10.12659/MSM.919530
Snodgrass RG, Huang S, Choi IW, Rutledge JC, Hwang DH. Inflammasome-mediated secretion of IL-1β in human monocytes through TLR2 activation; Modulation by dietary fatty acids. J Immunol. 2013;191(8):10.4049/jimmunol.1300298. doi:10.4049/jimmunol.1300298
Friedrich K, Sommer M, Strobel S, et al. Perturbation of the Monocyte Compartment in Human Obesity. Front Immunol. 2019;10:1874. doi:10.3389/fimmu.2019.01874
Sala E, Kuka M. The Suppressive Attitude of Inflammatory Monocytes in Antiviral Antibody Responses. Viral Immunol. 2020;33(4):327-333. doi:10.1089/vim.2019.0132
Patterson BK, Francisco EB, Yogendra R, et al. Persistence of SARS CoV-2 S1 Protein in CD16+ Monocytes in Post-Acute Sequelae of COVID-19 (PASC) up to 15 Months Post-Infection. Frontiers in Immunology. 2022;12. Accessed April 2, 2023. https://www.frontiersin.org/articles/10.3389/fimmu.2021.746021
Al-Benna S. Association of high level gene expression of ACE2 in adipose tissue with mortality of COVID-19 infection in obese patients. Obes Med. 2020;19:100283. doi:10.1016/j.obmed.2020.100283
Chiappalupi S, Salvadori L, Donato R, Riuzzi F, Sorci G. Hyperactivated RAGE in Comorbidities as a Risk Factor for Severe COVID-19-The Role of RAGE-RAS Crosstalk. Biomolecules. 2021;11(6):876. doi:10.3390/biom11060876
Faizo AA, Qashqari FS, El‐Kafrawy SA, et al. A potential association between obesity and reduced effectiveness of COVID‐19 vaccine‐induced neutralizing humoral immunity. J Med Virol. Published online September 14, 2022:10.1002/jmv.28130. doi:10.1002/jmv.28130
Frasca D, Reidy L, Romero M, et al. The majority of SARS-CoV-2-specific antibodies in COVID-19 patients with obesity are autoimmune and not neutralizing. Int J Obes. 2022;46(2):427-432. doi:10.1038/s41366-021-01016-9
Chang SE, Feng A, Meng W, et al. New-onset IgG autoantibodies in hospitalized patients with COVID-19. Nat Commun. 2021;12(1):5417. doi:10.1038/s41467-021-25509-3
Relative reduction of biological and phylogenetic diversity of the oral microbiota of diabetes and pre-diabetes patients - PubMed. Accessed December 16, 2021. https://pubmed-ncbi-nlm-nih-gov.eres.library.manoa.hawaii.edu/30625362/
Karabudak S, Ari O, Durmaz B, et al. Analysis of the effect of smoking on the buccal microbiome using next-generation sequencing technology. J Med Microbiol. 2019;68(8):1148-1158. doi:10.1099/jmm.0.001003

No competing interests reported.

PASCsupplementalfigures.pdf

Download PDF

Journal Publication

published 03 Jan, 2024

Read the published version in Scientific Reports →

Editorial decision: Major revision
09 Aug, 2023
Reviews received at journal
20 Jul, 2023
Reviewers agreed at journal
17 Jul, 2023
Reviewers invited by journal
07 Jul, 2023
Editor assigned by journal
07 Jul, 2023
Editor invited by journal
02 Jun, 2023
Submission checks completed at journal
02 Jun, 2023
First submitted to journal
23 May, 2023

You are reading this latest preprint version

Elevated HMGB1 mediates microbiome-immune axis dysregulation underlying reduced neutralization capacity in obesity-related post-acute sequelae of SARS-CoV-2

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Recruitment and Study Cohort

Data collection

Attenuated antibody response and viral neutralization capacity associates with obesity and is characteristic of PASC

Obesity-associated inflammation underlies the attenuated immune response in PASC

B-cell activating factor (BAFF) dysregulation in obesity explains attenuated viral neutralization capacity

Gut microbial dysbiosis correlates with PASC

High mobility group box 1: a master regulator of diminished humoral response in PASC

Pathway analysis shows baseline HMGB1 levels predict viral neutralization capacity

Discussion

The HMGB1 pathway influences the microbiome-immune axis, culminating in PASC

Materials & Methods

Human study participants

IgG/IgM Antibody detection test

Blood Processing

Stool sample DNA/RNA extraction

16S rRNA Sequencing and Analysis

Antibody detection ELISA

Inflammation factors concentration measurements

PBMC Phenotyping

Statistical analysis

Ingenuity Pathway Analysis (IPA)

SARS-CoV‐2 surrogate virus neutralization test (sVNT)

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1