Association between gut microbiota in HIV-infected patients and immune reconstitution following antiretroviral therapy (ART)

Abstract

Background

This study aims to examine the potential link between incomplete immune reconstitution following ART treatment and gut microbiota dysbiosis.

Methods

We collected clinical data and fecal samples from 50 HIV patients undergoing ART and 30 untreated patients. Based on the observed immune function reconstruction, we further categorized the ART(+) group into a responder group (n = 30) and a non-responder group (n = 20). The gut microbiota composition differences were assessed using Alpha diversity and Beta diversity analysis, while differential genera were identified through linear discriminant analysis effect size (LEfSe). Subsequently, functional disparities in the gut microbiota were investigated using PICRUSt2 and metagenomeSeq software.

Results

The results of Alpha diversity and Beta diversity revealed significant differences in the composition of gut microbiota among the three groups. Differential genus analysis identified Morganella as an exclusive genus present only in the Non-responder group, exhibiting a significantly higher relative abundance. Correlation analysis demonstrated a positive association between Morganella and LDL levels. The CAZY analysis revealed that glycosyltransferase 25 (GT25) was significantly expressed in the Non-responder group, whereas it was either undetectable or exhibited extremely low expression levels in both the Responder group and the ART(-) group. Importantly, the correlation analysis indicated a positive association between Morganella and GT25 secretion.

Conclusions

The ecological imbalance of Morganella might be associated with incomplete immune reconstitution following ART, potentially mediated by GT25 secretions. Consequently, Morganella could serve as a promising biomarker for predicting incomplete immune reconstitution in AIDS patients undergoing ART.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-10995-3.

Introduction

Acquired immunodeficiency syndrome (AIDS), caused by HIV infection, is an incurable infectious immunodeficiency disease that poses a significant global threat to human health [1]. Upon HIV infection in the human body, initial targeting of lymphocytes and particularly CD4 + cells occurs within the intestines [2]. Notably, viral replication predominantly takes place in intestinal tissues where approximately 80% of HIV conceals itself, while a mere 2-5% of viral load is detectable in the bloodstream [3]. Numerous studies have consistently demonstrated the pivotal role of gut microbiota in nutrient absorption, food metabolism, intestinal barrier defense against pathogens, and regulation of intestinal immune function [4]. However, HIV infection induces a perturbation in the composition of gut microbiota, leading to pathological immune activation and inflammatory responses that expedite the progression of HIV infection [5, 6]. The progression from HIV infection to AIDS is characterized by the disruption of intestinal barrier function, dysbiosis of the gut microbiota and its metabolites, chronic inflammation, and immune activation, as supported by extensive research [7, 8]. Currently, numerous studies are endeavoring to elucidate the mechanisms underlying immune activation and persistent inflammatory response subsequent to HIV infection by investigating alterations in the gut microbiota [9, 10]. The restoration of a harmonized gut microbial ecosystem is being contemplated as a potential therapeutic target [11, 12].

Antiretroviral therapy is a specialized treatment modality designed to target retroviruses. Its primary mechanism involves inhibiting the activity of viral reverse transcriptase, thereby effectively halting the viral replication process within the human body and controlling viral infection. To date, antiretroviral therapy is universally acknowledged as the most efficacious approach for managing AIDS and has been extensively implemented in clinical settings. ART treatment can effectively suppress HIV virus replication and restore immune function. However, despite effective control of viral load, a subset of patients (approximately 15–30%), particularly those with advanced HIV infection and longer duration of infection, fail to achieve optimal immune restoration, resulting in the phenomenon known as incomplete immune reconstitution [13]. Currently, the incomplete reconstruction of immune function following ART has emerged as a crucial determinant of antiviral therapy efficacy, AIDS incidence, and mortality rates [14, 15]. Previous studies have demonstrated that the diversity, compositional structure, metabolic function, and immune regulatory capacity of the gut microbiota in HIV patients undergo significant alterations before and after antiretroviral therapy (ART). However, the association between incomplete immune reconstitution following ART and the gut microbiota remains underexplored, with limited clear evidence currently available [16–18]. Consequently, this study aims to investigate the effects of ART on the structure and function of the gut microbiota in HIV-infected individuals, while further exploring the potential association between incomplete immune reconstitution after ART and gut microbiota dysbiosis. Furthermore, it seeks to identify specific biomarkers of gut microbiota for immunological non-responder (INR) patients with incomplete function reconstitution.

Materials and methods

Study population

This is a single-center prospective cohort study conducted at First Affiliated Hospital of USTC, which is a designated hospital for antiretroviral therapy for HIV/AIDS in Anhui Province (China). From January to December 2022, a total of 59 AIDS patients receiving ART were enrolled in the ART(+) group. These patients underwent consistent ART treatment for one year. To minimize intergroup disparities, we selected 32 AIDS patients who had not initiated ART treatment in the control group [ART(-) group], ensuring their age and gender distribution matched that of the ART(+) group. Inclusion criteria for AIDS patients: (1) meeting the diagnostic criteria for AIDS, laboratory tests revealed positive HIV antibody results, which were corroborated by a clear epidemiological history; (2) age ≥ 18 years. Exclusion criteria: (1) non-compliance with medical advice; (2) pregnant or lactating women; (3) severe organic diseases such as cirrhosis, renal failure, heart failure, pneumonia; (4) patients who have undergone surgery within the past 3 months.

According to the 2021 edition of the “Chinese Guidelines for Diagnosis and Treatment of AIDS”, immunological responder is defined as a patient whose CD4 + T lymphocyte count has increased by at least 30% or by at least 100 cells/µL relative to the baseline CD4 + T cell count at the time of initial diagnosis, one year following the initiation of antiretroviral therapy. Patients who fail to meet these criteria are categorized as immunological non-responders (INR). The ART(+) group was stratified into two subgroups based on the reconstruction of immune function following ART treatment for one year, namely the responder group and non-responder group. This study obtained approval from the Ethics Committee of First Affiliated Hospital of USTC (No. 2024RE284). Prior to the initiation of the study, informed consent forms were duly signed by all enrolled patients.

Clinical data

Demographic and clinical data of patients are collected and recorded. Demographic information includes age, gender, body mass index (BMI), diabetes status, hypertension status, alcohol consumption habits, and smoking habits. Clinical data includes laboratory test results, follow-up records, and medication history. After the initial diagnosis and admission, all patients underwent a comprehensive medical history inquiry and recording, thorough physical examination, meticulous medication record review, as well as extensive laboratory tests. Following discharge, the experimental group conducted regular follow-up through telephone calls to individuals undergoing ART treatment, consistently inquiring about their adherence to medication.

The diagnostic criteria for HIV/AIDS, as outlined in the 2021 edition of the Chinese Guidelines for Diagnosis and Treatment of HIV/AIDS, are as follows: (1) Positive HIV antibody screening, confirmed by Western blot test, which is considered the gold standard for AIDS diagnosis; (2) Epidemiological history encompassing unsafe sexual behavior, intravenous drug use, children born to HIV-positive individuals, and occupational exposure history; (3) Clinical manifestations including fever, diarrhea, lymphadenopathy, and weight loss. The antiretroviral therapy drugs primarily comprise non-nucleoside reverse transcriptase inhibitors (NNRTIs), protease inhibitors (PIs), and integrase strand transfer inhibitors (INSTIs).

The ART(+) group collected blood samples at the time of initial diagnosis and one year after ART treatment, while the ART(-) group collected blood samples only at the time of initial diagnosis. A total volume of 5 mL venous blood was obtained using vacuum blood collection tubes, which were subsequently stored at room temperature, centrifuged at 3000 g for 10 min, and immediately frozen at -80℃. Subsequently, Alanine transaminase (ALT), Aspartate transaminase (AST), Total Protein (TP), Albumin (ALB), Prealbumin (PA), Creatinine (Cr), Blood urea nitrogen (BUN), C-reaction protein (CRP), White blood cell count (WBC), Interleukin-6 (IL-6), Procalcitonin (PCT), N-terminal pro-brain natriuretic peptide (NT-ProBNP), CD4 cell count, CD8 cell count, and CD3 cell count were examined.

DNA extraction and purification

Patients in the ART (+) group, who had received antiretroviral therapy for one year, and patients in the ART (-) group, who were newly diagnosed and had not yet initiated ART treatment, were required to provide fecal samples within 2 h of admission and deliver them to the hospital laboratory. Trained laboratory physicians used disposable sterile cotton swabs to gently collect at least 1 gram of feces by swiping over the surface, which was then placed into a sterile storage tube and immediately frozen at -80 °C. Genomic DNA was extracted using the PowerMax extraction kit (MoBio Laboratories, Carlsbad, CA, USA) and stored at -20 °C. The DNA concentration was determined using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and sample integrity was assessed by 2% agarose gel electrophoresis. The V4 region of the 16 S rRNA gene was amplified using PCR with forward primer F (5’-GTGCCAGCAGCCGCGGTAA-3’) and reverse primer R (5’-GGACTACCAGGGTTTCTAAT-3’). The PCR products were purified and quantified using the EasyPure Quick Gel Extraction Kit (TransGen, Beijing, PR China) and PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, USA), respectively.

16S rRNA gene high-throughput sequencing

The sequencing platform used was Illlumina HiSeq4000 pair-end 2 × 150 bp platform. After obtaining the raw data, the Barcode sequences and primer sequences were removed, and Vsearch v2.4.4 was used to assemble the reads of the samples. QIIME v1.8.0 (http://qiime.org/) was employed to filter the sequences, resulting in the final Effective Tags. Using a similarity threshold of 97%, tags with a similarity above 97% are clustered into OTUs. The Vsearch 2.4.4 software and SILVA128 database are used to annotate the species, and the microbial community composition between samples and groups is analyzed. The abundance and classification of all OTUs in each sample are recorded, and any OTUs with an abundance below 0.001% of the total sequences in all samples will be removed.

Bioinformatics analysis and statistical analysis

The focus of our study was primarily on analyzing intergroup differences at the phylum and genus levels. Therefore, we did not include any discussion of bioinformatics analysis between samples in this article. For bioinformatics analysis, we mainly utilized QIIME software and R software (v3.2.0). We used the Venn diagram function in the R package to generate visualizations that depict shared and unique operational taxonomic units (OTUs) among different groups. Alpha diversity primarily assesses species diversity in samples based on the total number of OTUs (Chao1, ACE) and microbial community diversity (Shannon, Simpson). The QIIME software is used for calculating alpha diversity indices and goods coverage, generating ranked abundance curves at the OTU level, as well as plotting the curves.

The beta diversity was based on Bray-Curtis distance and calculated using Qimme software to generate a principal coordinate analysis (PCoA) plot. The analysis of similarities (ANOSIM) test, implemented in the “Function betadisper” function of the R package, was used to assess the significance of differences between groups. Linear discriminant analysis effect size (LEfSe) analysis, performed with the ldamarker function from R package v3.2.0, was employed to detect biomarkers showing significant differences between groups. OTUs with LDA scores > 4 were considered as indicating significant intergroup differences. Calculate the Spearman correlation between differential species and clinical indicators in terms of their belonging to a certain level, and plot them as heatmaps using pheatmap (R package v3.2.0). Analyze the functional differences of microorganisms between groups based on the GO and KEGG databases using PICRUSt2 software. Use metagenomeSeq software to analyze the differential metabolic functions of bacterial genera.

SPSS 24.0 software was utilized to conduct statistical analysis on the data. Descriptive statistics such as mean ± standard deviation were employed to describe metric variables that followed a normal distribution. T-tests were conducted to compare two groups, while F-tests were applied to compare three or more groups. Count data were presented as [n (%)] and compared using chi-square tests across different groups. A significance level of p < 0.05 indicated a statistically significant difference.

Results

Baseline characteristics

The study enrolled a total of 91 AIDS patients, with 59 receiving ART and 32 not receiving ART. We excluded a total of 11 patients, including 5 whose fecal samples were improperly collected, 4 who did not adhere to medication regimen, 1 who had an inflammatory lung infection, and 1 who underwent appendicitis surgery prior to fecal collection. Consequently, the final analysis included a cohort of 50 AIDS patients receiving ART and 30 patients without ART treatment. In the ART(+) group, there were 46 males and 4 females with an average age of (45.80 ± 10.64) years old. Among them, successful immune function reconstruction was achieved after one year of treatment in 30 cases, while reconstruction failed in 20 cases. Consequently, they were categorized into a responder group (n = 30) and a non-responder group (n = 20). The study included thirty untreated AIDS patients, comprising 28 males and 2 females, with an average age of (39.80 ± 11.58) years old. The ART(+) group underwent a one-year follow-up, successfully completing the follow-up for 59 patients. Fortunately, no fatalities were recorded throughout the entire follow-up period.

The demographic and clinical characteristics of the non-responder group, the responder group and the ART(-) group are detailed in Table 1. There were statistically significant differences in the levels of CRP, IL-6, CD4 cell count, CD8 cell count, CD3 cell count, and Cr in three groups. The remaining variables did not show any differences. There were statistically significant differences between the non-responder group and responder group in ALB, CRP, IL-6, CD4 cell count, CD8 cell count, CD3 cell count, and Cr levels.

Table 1.

Baseline characteristics of study populations

Variables	ART(+) group			ART(-) group (n = 30)	p-value#
	Non-responder group (n = 20)	Responder group (n = 30)	p-value*
Demographic
Age, years	44.10 ± 10.37	46.73 ± 10.95	0.173	39.80 ± 11.58	0.171
Male	20 (100.00)	26(86.67)	0.089	27 (90.00)	0.251
Smoking	4 (20.00)	7 (23.33)	0.533	5 (16.67)	0.812
Drinking	5 (25.00)	6 (20.00)	0.676	8 (26.67)	0.822
Hypertension	4 (20.20)	8 (26.67)	0.589	3 (10.00)	0.250
Diabetes	1 (5.00)	2 (6.67)	0.808	0 (0.00)	0.375
BMI, kg/m2	22.72 ± 3.48	23.45 ± 2.38	0.563	23.26 ± 2.04	0.273
Clinical variables
ALT, IU/L	35.20 ± 7.67	25.53 ± 4.17	0.263	36.67 ± 4.61	0.814
AST, IU/L	27.30 ± 8.94	20.47 ± 8.54	0.123	36.60 ± 5.04	0.414
ALB, g/L	35.91 ± 7.33	43.27 ± 9.15	0.022	38.05 ± 6.92	0.079
PA, g/L	158.30 ± 30.28	180.09 ± 76.56	0.055	226.14 ± 23.35	0.083
TP, g/L	66.03 ± 8.77	67.87 ± 8.88	0.772	68.89 ± 7.48	0.719
CRP, mg/L	41.21 ± 8.94	9.73 ± 2.08	0.035	19.80 ± 9.59	0.016
PCT, ng/mL	0.96 ± 0.67	0.16 ± 0.13	0.287	0.84 ± 0.65	0.155
WBC, 109/L	6.42 ± 2.19	5.18 ± 1.73	0.275	6.08 ± 1.76	0.616
IL-6, ng/L	106.91 ± 27.81	35.47 ± 9.10	0.001	31.47 ± 9.53	0.046
BUN, mmol/L	23.60 ± 5.35	5.54 ± 2.06	0.289	5.26 ± 1.83	0.149
CD4 count, cell/µL	87.60 ± 12.54	321.67 ± 95.01	0.006	241.40 ± 38.88	0.026
CD8 count, cell/µL	328.89 ± 87.86	943.47 ± 117.16	0.049	693.47 ± 40.96	0.004
CD3 count, cell/µL	455.78 ± 58.59	1308.13 ± 97.23	0.021	988.60 ± 102.19	0.004
Cr, µmoI/L	125.90 ± 36.43	76.60 ± 7.78	0.025	63.87 ± 10.97	0.001
ART regimen
NNRTI	15 (75.00)	23 (76.67)	0.892	/	/
PI	7 (35.00)	9 (30.00)	0.710	/	/
INSTI	18 (90.00)	26 (86.67)	0.722	/	/

Data presented as n (%), mean ± standard deviation. * Comparisons between risk group and non-risk group, ^# comparisons between three group. BMI = body mass index, ALB = albumin, PA = prealbumin, NT-proBNP = N-terminal Pro-B-type natriuretic peptide, INSTI = integrase strand transfer inhibitor, NNRTI = nonnucleoside reverse transcriptase inhibitor, PI = protease inhibitor, ALT = alanine aminotransferase, AST = aspartate transaminase, TP = total protein, CRP = C-reaction protein, PCT = procalcitonin, WBC = white blood cell, IL-6 = interleukin-6, BUN = blood urea nitrogen, Cr = creatinine

Effect of ART treatment efficacy on the gut microbiota diversity

The samples from the three groups yielded a total of 5724 amplicon sequence variants (ASVs), and the observed OTUs in each sample exhibited good’s coverage indices above 99%, indicating a good efficacy of sequencing for all samples. Alpha rarefaction analysis of OTUs demonstrated sufficient sequencing depth across the three groups’ samples, enabling successful detection of most microbial populations and mitigating biases arising from intergroup differences in sample size (Supplementary Fig. 1). The Venn diagram shows a total of 13,663 OTUs, with 9,412 in the Non-responder group, 11,911 in the Responder group, and 10,313 in the ART(-) group. Notably, there are 7,089 OTUs shared among all three groups (Fig. 1A).

Fig. 1.

Venn diagram and diversity analysis. (A) A Venn diagram among the non-responder group, responder group, and ART(-) group. (B) Alpha diversity indices for Chao1index. (C) Alpha diversity indices for Shannon index. (D) Alpha diversity indices for Simpson index. (E) Principal coordinate analysis (PCoA) based on Bray-Curtis distance among the non-responder group, responder group, and ART(-) group

Significant differences in inter-group Alpha diversity were observed, with p values of 0.017, 0.012, and 0.004 for Chao1, Shannon, and Simpson indices respectively. Notably, Responder group results for Chao1 and Shannon diversity indices exhibited significantly higher values compared to the Non-responder group. These findings suggest that the efficacy of ART treatment is associated with Alpha diversity (Fig. 1B-D). The Beta diversity analysis, employing the Bray-Curtis distance metric, revealed a significant variation in the structure of gut microbiota prior to ART treatment. However, following ART treatment, there was an observed tendency for stabilization in the microbial community structure (Fig. 1E). To assess differences in community composition among the three groups, an ANOSIM test was conducted and yielded results indicating a statistically significant difference between them (R = 0.041, p = 0.045).

Differences in gut microbiota composition of HIV patient after ART treatment

To elucidate the impact of ART and its efficacy on the composition of the gut microbiota, we analyzed the dominant components of the gut microbiota at both the phylum and genus levels. To refine our research focus, we concentrated on genera that exhibited significant differences at the genus level, thereby elucidating the precise impact of immune function reconstitution following ART treatment on specific intestinal bacteria. The taxonomic profiles at the phylum level for the three groups are shown in Fig. 2A. The top five dominant phyla for all groups were Firmicutes, Bacteroidota, Proteobacteria, Verrucomicrobiota, and Actinobacteriota. The p-values for the differences between these phyla among the three groups were 0.161, 0.513, 0.013, 0.272 and 0.263 respectively. It is worth noting that the ART(-) group and Non-responder group have similar relative abundance of Proteobacteria, while the Responder group shows a significant decrease in Proteobacteria relative abundance (Fig. 2B). At the genus level, the top 10 dominant genera were Bacteroides, Prevotella, Escherichia-Shigella, Enterobacter, Faecalibacterium, Parabacteroides, Veillonella, Blautia, Megamonas, and Phascolarctobacterium (Fig. 2C).

Fig. 2.

Taxonomic features of gut microbiota in HIV patients with non-responder group, responder group, and ART(-) group. (A) Relative abundances of bacteria at the phylum level. (B) Relative abundance of Proteobacteria. (C) Relative abundances of bacteria at the genus level

To differentiate the microbial communities among various groups, we employed the linear discriminant analysis (LDA) effect size (LEfSe) algorithm. We identified a total of 7 microbial taxa with LDA scores greater than 3, including 2 significantly expressed taxa in the Non-responder group, 2 in the Responder group, and 3 in the ART(-) group (Fig. 3A). At the genus level, significant differences were observed in the relative abundances of Enterobacter, Ruminococcus, and Anaerotruncus.The cladogram illustrated distributional differences of microbial populations among the three groups (Fig. 3B). For accurate differentiation at genus level among these groups, a heatmap of relative abundance was generated for differential analysis of genera (Fig. 3C).

Fig. 3.

Taxonomic differences between gut microbiota among the non-responder group, responder group, and ART(-) group and heatmap of correlation. (A) LDA analysis results. (B) LEfSe analysis clustering tree. ASV, amplicon sequence variants; p, c, o, f, and g represent phylum, class, order, family and genus, respectively. (C) Heatmap displaying the relative abundance at the genus level. (D) Relative abundance of the most significant genus. (E) Heatmap of correlation between gut microorganisms and clinical indicators. *, p < 0.05; **, p < 0.01; NS, not significant; NA, not applicable

The genera with significant differences are depicted in Fig. 3D, and we provided a comprehensive analysis of these distinct genera from various perspectives. Following ART treatment, Mitsuokella, Acidaminococcus, Tyzzerella, Clostridia_UCG-014, and Ruminococcus exhibited enrichment in the ART(+) group, while Enterobacter showed a significant decrease. It is noteworthy that Mitsuokella, Clostridia_UCG-014, and Ruminococcus were further enriched in the Responder group. However, Acidaminococcus, Tyzzerella, and Enterobacter displayed a further decline. After ART treatment, the immune recovery is directly correlated with the abundance of Faecalibacterium and Alloprevotella. In the Responder group, their relative abundance is higheras compared to ART(-) group, whereas in the Non-responder group it is relatively lower. The presence of Morganella and Cloacibacillus was undetected in both the Responder group and the ART(-) group, whereas these two microorganisms were exclusively detected in the Non-responder group. Moreover, the relative abundances of Morganella and Cloacibacillus in the Non-responder group exhibited a significant increase. Furthermore, following ART treatment, there are no alterations in the relative abundance of Catenibacterium and Lachnospiraceae_NK4A136_group within the Non-responder group. However, these genera experience a substantial increase within the Responder group.

Correlation between gut Microbiome and clinical indicators

In order to investigate the association between specific bacterial genera and lipid metabolism, inflammation, immunity, and nutrition, we assessed blood lipid indicators, nutritional markers, as well as levels of inflammatory factors and immune factors in patientsat baseline (pre-ART). The correlation between clinical indicators of HIV patients and their gut microbiota is depicted in Fig. 3E. It is noteworthy that the Lachnospiraceae_NK4A136_group exhibits a negative correlation with IL-6, whereas Morganella, Acidaminococcus, and LDL show positive correlations. Catenibacterium demonstrates a positive correlation with ALB, while Faecalibacterium displays positive correlations with TG, TC, and CD4 levels. Additionally, Enterobacter shows positive correlations with LDL, PA, TP, and ALB.

Analysis of metabolic pathway among groups and specific gut bacteria

We analyzed the metabolic pathways of gut microbiota based on PICRUSt2 (Fig. 4A), and found that lactate consumption I (MF0079), maltose degradation (MF0008), lysine degradation I (MF0058), trehalose degradation (MF0012), phenylalanine degradation (MF0024), putrescine degradation (MF0082), 4-aminobutyrate degradation (MF0076), ferredoxin oxidoreductase (MF0069) and arginine degradation II (MF0052) were significantly decreased in the Responder group. In addition, KEGG analysis showed that the metabolic levels of aminobenzoate degradation (ko00627), tetracycline biosynthesis (ko00253), glutathione metabolism (ko00480), caprolactam degradation (ko00930), and Biosynthesis of siderophore group nonribosomal peptides (ko01053) were significantly reduced in the Responder group (Fig. 4B).

Fig. 4.

Gut microbiological function prediction. (A) Metabolic pathways based on PICRUSt2. (B) KEGG analysis based on PICRUSt2. (C) CAZY analysis based on PICRUSt2. (D) Correlation between KEGG metabolic modules and differential microbiota. *, p < 0.05; **, p < 0.01; NS, not significant

We conducted CAZY analysis using PICRUSt2, which revealed a significant reduction in metabolic levels of GT8 and GH37 in the Responder group. Interestingly, GT25 was significantly expressed in the non-responder group but was undetectable in the responder group. Additionally, it exhibited only an extremely low level of expression in the ART(-) group (Fig. 4C). Notably, our CAZY and genus correlation analysis demonstrated a positive association between Morganella and GT25 secretion (Fig. 4D).

Discussion

Recent research has demonstrated that ART significantly alters the composition of gut microbiota in HIV patients [19–20]. However, the relationship between immune function reconstitution following ART and alterations in the gut microbiota remains poorly understood. Thus, this study aims to investigate the potential link between incomplete immune reconstitution after ART and gut microbiota dysbiosis, as well as to identify specific marker microbiota that could serve as predictors of incomplete immune reconstitution following ART. Therefore, we performed 16 S rRNA gene sequencing analysis on fecal samples obtained from ART(-) patients, IR patients, and INR patients.

The results of the good’s coverage and alpha rarefaction analysis for all samples indicate that the sequencing depth, species richness, and evenness between groups fulfill the prerequisites for subsequent analyses. Furthermore, to facilitate a more precise comparison of the correlation between existing literature reports and the findings of this study, several pertinent studies were selected and summarized. The detailed summaries are presented in Supplementary Table 2 [21–24]. After reviewing the studies on the impact of ART treatment on the gut microbiota in HIV patients, it was observed that immunodiscordant responders exhibited an augmentation in diversity, whereas immunological non-responders displayed a reduction in diversity, aligning with previous research findings [16–17]. In this study, Alpha diversity analysis revealed that the Responder group exhibited the highest number of observed genera, indicating a greater bacterial richness in comparison to the Non-responder group, which demonstrated the lowest number of observed genera. These findings suggest that post-ART immune function reconstruction status has a significant impact on gut microbiota diversity. Beta diversity analysis revealed higher variability in the distribution of intestinal microbiota in the ART(-) group, while the Non-responder and Responder groups showed further separation after ART treatment. This finding suggests that the gut microbiota of the ART(-) group exhibits significant inter-individual variations, and following long-term ART treatment, these changes in gut microbiota gradually become associated with immune reconstitution within the host. Recently, a study revealed that the gut microbiota composition of ART(-) individuals exhibits the highest similarity to those with INR, while distinctly separating from both INR and IR. Furthermore, INR and IR groups formed separate clusters [18]. Compared with this report, we observed distinct separation phenomena in the composition of the gut microbiota in the INR and IR groups, which aligns with previous findings; however, regarding the distribution of the gut microbiota in the ART(-) group, research results remain inconsistent, and no consensus has yet been reached. The differences between our findings and this reports may be ascribed to various factors, such as geographical location or dietary patterns. Previous investigations were predominantly conducted in European countries and western China, whereas our study was carried out in eastern China, implying that regional and environmental factors might exert an influence on the composition of gut microbiota [25–26]. In conclusion, despite the lack of consistency in the gut microbiota composition among ART(-) individuals, there is consensus regarding the separate clustering of gut microbiota between INR and IR patients.

To identify the differential genera between groups, we performed LEfSe analysis and generated a genus-level heatmap depicting relative abundance. The results revealed that the Responder group exhibited enrichment of Mitsuokella, Clostridia_UCG-014, Ruminococcus, Faecalibacterium, Alloprevotella, Catenibacterium, and Lachnospiraceae_NK4A136_group. Whereas Acidaminococcus, Tyzzerella, Morganella, and Cloacibacillus were enriched in the Non-responder group. These findings suggest their potential involvement in modulating immune recovery mechanisms following ART treatment. Reviewing previous research reports, we observe that differences at the genus level lack high consistency, and the findings reported in each study vary considerably [21–24]. This variation is likely to be closely associated with factors such as sample size, geographical distribution, and ethnic diversity. Notably, our study identifies Morganella and Cloacibacillus as specific genera for the Non-responder group, with Morganella exhibiting relatively high abundance. To the best of our knowledge, this finding has not been documented in existing literature. Morganella, a gram-negative bacterium, is considered a conditional pathogen associated with various nosocomial infections, particularly urinary tract and wound infections [27–28]. Multiple studies have consistently demonstrated a positive correlation between Morganella and the expression of inflammatory factors (IL-6, IL-8, and TNFα), while showing a negative association with levels of short-chain fatty acids (SCFAs). Furthermore, Morganella exhibits the ability to secrete diverse small molecule metabolites that impact intestinal lipid metabolism, auxiliary factors, and amino acid metabolism [29–30].

HIV infection can induce diverse metabolic alterations, including impaired glucose metabolism, hypertriglyceridemia, and downregulation of phospholipid metabolism [31–32]. Additionally, ART treatment may further disrupt metabolic function by affecting abnormal amino acid breakdown metabolism, imbalances in phospholipid and sphingolipid metabolism, and upregulation of mitochondrial toxicity [33]. HIV infection and ART often lead to abnormal emaciation, which is attributed to impaired gastrointestinal barrier function, opportunistic infections, and accelerated gastrointestinal transit time. These factors increase the risk of malnutrition for individuals with INR compared to those with IR [34]. We investigated the relationship between different bacterial genera and lipid metabolism, inflammation, immunity, and nutrition. The findings revealed that Lachnospiraceae_NK4A136_group exhibited a significant association with inflammatory factors, while Morganella, Acidaminococcus, Enterobacter, and Faecalibacterium demonstrated a notable correlation with lipid metabolism. Catenibacterium and Enterobacter were found to be linked to nutritional levels, whereas Faecalibacterium also displayed an association with immune factors.

In order to further investigate the relationship between microbial metabolic pathways and immune recovery in HIV patients, we employed PICRUSt for predictive functional analysis. Our findings indicate that the Responder group exhibited significant reductions in glycolysis, lactate consumption, amino acid degradation, and antimicrobial substance synthesis compared to both the ART (-) and Non-responder groups. These results suggest that patients in the Responder group have experienced notable improvements in nutritional status and inflammation conditions, but metabolic abnormalities still persist. A report from China indicates that after ART treatment, there is some improvement in patients’ immune function and metabolic abnormalities. However, problems related to impaired glucose metabolism, abnormal lipid metabolism, and amino acid catabolism caused by HIV infection still persist [35].

Furthermore, we performed CAZY analysis and observed a positive correlation between Morganella and the secretion of GT25. Notably, the Non-responder group exhibited a significant augmentation in GT25 secretion. Consequently, our findings suggest that Morganella potentially modulates GT25 secretion, thereby contributing to immune non-response in HIV patients undergoing ART treatment. The GT25 enzyme belongs to the GT-A type glycosyltransferase family, and GT-A type glycosyltransferases can modulate intestinal inflammation response and immune function through diverse mechanisms. Firstly, glycosyltransferases can penetrate intestinal epithelial cells and interact with their proteins, resulting in aberrant protein glycosylation. Consequently, the original functions of these proteins are compromised, thereby impacting signal transduction, gene transcription, and immune response of intestinal epithelial cells [36–37]. Moreover, the intestinal epithelium abundantly expresses N-glycans and O-glycans, which play a pivotal role in regulating intestinal barrier function and defending against foreign pathogens and toxins. Intestinal epithelial glycosylation is complex and associated with cytokines, glycosyltransferases, glycosidases, and ion channels. Dysregulation of intestinal epithelial glycosylation can exacerbate intestinal inflammation [38–39].

There are several limitations to consider in our study. Firstly, the observed variations in gut microbiota among the study population can be attributed to differences in geographical locations and dietary habits. The HIV patients included in this study primarily originated from Anhui Province, which may somewhat restrict the generalizability and applicability of our findings. Additionally, our sample size is relatively limited, which may constrain the generalizability of our research findings. Furthermore, this study did not conduct stratified analyses by gender, nor did it account for certain specific scenarios of male-male sexual behavior. Lastly, a healthy control group was not included in this research, preventing us from drawing comparative conclusions between HIV-infected and uninfected individuals. Subsequently, our research will primarily focus on nursing interventions in the ART treatment process for HIV patients, attempting to regulate gut microbiota dysbiosis through probiotics with the objective of enhancing patients’ nutritional status and mitigating the risk of immunological non-response to treatment.

Conclusion

Our research demonstrates that the reconstitution of immune function in HIV patients following ART is closely linked to alterations in the gut microbiota. Specifically, the ecological dysbiosis of Morganella may be associated with incomplete immune reconstitution post-ART, potentially mediated by GT25 secretions. Consequently, Morganella holds promise as a potential biomarker for predicting incomplete immune reconstitution in HIV patients undergoing ART.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

HIV

Human immunodeficiency virus

ART

Antiretroviral therapy

INR

Immunological non-responder

LEfSe

Linear discriminant analysis effect size

GT-25

Glycosyltransferase 25

AIDS

Acquired immunodeficiency syndrome

BMI

Body mass index

NNRTI

Non-nucleoside reverse transcriptase inhibitor

PI

Protease inhibitor

INSTI

Integrase strand transfer inhibitor

ALT

Alanine transaminase

AST

Aspartate transaminase

TP

Total Protein

ALB

Albumin

NT-proBNP

N-terminal Pro-B-type natriuretic peptide

PA

Prealbumin

Cr

Crea

BUN

Blood urea nitrogen

CRP

C-reaction protein

WBC

White blood cell count

IL-6

Interleukin-6

PCT

Procalcitonin

QIIME

Quantitative insights into microbial ecology

OTU

Operational taxonomic units

PCoA

Principal coordinate analysis

ACV

Amplicon sequence variant

ANOSIM

Analysis of similarities

LDA

Linear discriminant analysis

KEGG

Kyoto encyclopedia of genes and genomes

SCFA

Short-chain fatty acids

Author contributions

Y.S assisted with the statistical analysis and drafted the manuscript. M.H assisted with the statistical analysis, and revised the manuscript. J.W, and T. L collected the data and performed the statistical analysis. Y.Q and A.L assisted with designing the study and revised the manuscript. All authors read and approved the manuscript.

Funding

This study was supported by Seventh Cycle Clinical Key (Cultivation) Specialty in Hefei Department Construction Project (Clinical Laboratory) (No. Z033-1).

Data availability

The raw microbiome data have been archived in the NCBI Sequence Read Archive (SRA) repository under the accession number PRJNA1219356.

Declarations

Ethics approval and consent to participate

The present study was performed according to the Declaration of Helsinki or relevant guidelines and regulations, and was approved by the Ethics committee of First Affiliated Hospital of USTC (No. 2024RE284). This study has obtained the written informed consent forms provided by all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.