Introduction

While the prevalence of HIV-associated morbidity, mortality and the occurrence of opportunistic infections has diminished considerably with the advent of antiretroviral therapy (ART), there has been a corresponding increase in the risk for other chronic diseases.¹ People living with HIV (PLHIV) demonstrate higher risk of fatal cardiovascular disease than the general population,² which is, like many other complications among PLHIV, associated with chronic inflammation or immunodeficiency.^1-3 Another factor adversely affecting the quality of life (QoL) of HIV+ patients undergoing ART is a non-infectious, often chronic, diarrhoea^{4, 5} which has limited possibilities for management.⁶ The primary cause of such non-infectious diarrhoea in HIV+ patients receiving ART is the treatment itself, of which diarrhoea is a fairly common side effect.⁷ However, non-infectious diarrhoea can also derive from pathological effects of HIV on cells and on the immune system in the gastrointestinal tract.^{7, 8} A model trying to explain the emerging risk of non-AIDS-related morbidity emphasizes HIV-mediated mucosal disruption leading to chronic systemic inflammation.³

In recent years, the implications of the gut microbiome and its metabolites have gained increasing attention. One such group of metabolites are short-chain fatty acids (SCFA), which are produced by multiple bacterial species in the intestinal tract through fermentation of dietary fibre and other resistant starches.^{9, 10} They are mainly used as source of energy for colonic epithelial cells or transported via blood stream to other tissues from where they intervene in the host’s cellular processes.^{9, 11} The most studied compounds of SCFA are acetic, propionic and butyric acids, with two, three and four carbons in their chemical structure respectively.¹¹ They constitute 90-95% of the SCFA present in the colon.¹⁰ SCFA have been described as for their anti-inflammatory properties^{12, 13} and in the amelioration of the course of several diseases through a positive influence on autoimmune reactions and metabolic functions. In particular, its positive effects in multiple sclerosis,^{14, 15} Parkinson’s disease,¹⁶ cardiovascular diseases,¹⁷ inflammatory bowel disease,^{18, 19} respiratory inflammatory diseases¹³ and type 2 diabetes^{10, 20} have been demonstrated. The modulation occurs through activation of G protein-coupled receptors (FFAR2, FFAR3, GPR109a, Olfr78) and through acetylation and methylation of the host’s gene expression.^{11, 19, 21, 22} By acting as inhibitors of histone deacetylase (HDAC) they play an important role in epigenetically regulating the colonic regulatory T-cell homeostasis.^{9, 23} Orally administered SCFA have been shown to increase the amount of intestinal Tregs by enabling the differentiation of CD4+ T cells into regulatory IL-10 producing T cells.^{14, 23-25} SCFA have also been described to suppress the conversion into pathogenic TH1 and TH17 cells.^{26, 27} However, it has been indicated that SCFA can not only promote T-cell differentiation into regulatory T cells but also into less inflammatory TH1 and TH17 effector cells, depending on the cytokine milieu.^{28, 29} In addition, it has been found that SCFA facilitate primarily the extra-thymic generation of Tregs. Such a de novo Treg cell generation in the periphery was described to be especially potentiated by butyrate and propionate.³⁰ While suppressing the pro-inflammatory cytokines including TNF-α, IL-6, IL-4, IL-5, IL-17a, IL-2 and TGFβ^{9, 10, 31, 32} SCFA have been documented to increase the release of the anti-inflammatory cytokine IL-10.^{9, 23, 28, 31, 33} A similar beneficial effect has also been reported during sodium propionate (SP) supplementation.^{14, 15, 34}

Taken together, these reports draw our attention to the potential modulatory role of dietary habits in the control of infectious diseases. Likewise, the role of SCFA in the control of insulin sensitivity and hyperlipidaemia^{10, 20, 35-37} could also be potentially providing crucial benefits in compensating the negative effects of ART. Therefore, in the current study we evaluated the effects of SCFA-supplementation on metabolic and immunological parameters, as well as on the microbiome of PLHIV and patient-reported QoL. For better comparison of the effects we used SP as SCFA, since SP in particular has been used for dietary supplementation in recent studies.^{14, 15, 34}

Methods

Results

Sociodemographic data and clinical characteristics

As shown in Table 1 the mean age of study participants (N = 32) was 50 years (24–73 years). The largest included age group was between 40 and 60 years old (n = 20, 63%). Only 15% were over 60 years old. 7 (22%) of the participants were female, the majority with 25 subjects (78%) were male. The HIV transmission of all female participants occurred through heterosexual contact. In the male collective, 21 out of the 25 participants were classified as men who have sex with men (MSM) transmissions (84%), one subject reported a heterosexual transmission. 3 participants did not provide any detailed information (12%).

Table 1. Sociodemographic data and clinical characteristics of the HIV+ study cohort at baseline

For the largest group of participants, time since initial HIV diagnosis was 10–20 years prior to the study (47%). For almost a third of the participants (31%), < 10 years had lapsed since the first diagnosis, while 22% stated that the first diagnosis was made over 20 years ago. All subjects had an undetectable HIV viral load throughout the study. The mean value of CD4 nadir of the participants were 322 cells/µl (with a range from 6 to 912 cells/µl).

CDC stages: 28% of participants were classified as stage A, one fourth of the collective were of stage B. The highest number of participants (47%) could be subsumed in stage C, with CDC stage C3 being represented the most (Table 1).

As shown in Table 1, most subjects (56%) were in the normal weight range (BMI 18.5–24.9), 25% were overweight, 15% were obese (BMI >29.9). Only 1 participant was underweight with a BMI <18.5. Furthermore, 44% of the participants were currently smoking, 25% self-reported to be former smokers. 31% have never consumed any tobacco. The most represented comorbidities were depression and vitamin D deficiency. Hypercholesterolemia was found in 25% of participants.

Assessment of metabolic profile

As seen in Fig. 1 HbA1c and HbA1c% concentration decreased noticeably (R² > 0.9) at both end point measurements (V6 and V9). Cholesterol (R² > 0.7) and the LDL/HDL ratio also showed a decrease throughout the whole study (R² > 0.9). No changes in triglycerides and serum glucose level were detected.

Evaluation of Th1, Th2, Th17, Tregs, lymphocyte subpopulations and cytokine profile

The results of the T-cell differentiation analysis are shown in Fig. 2. Among T-cell populations, the percentage of CD4+ T helper cells and Type 2 helper cells (Th2) increased (R² > 0.9). Regulatory T cells (Tregs), Type 1 helper cells (Th1), T helper 17 cells (Th17) all decreased in their percentage (R² > 0.9). By the end of the intervention, we observed an increase in CD4, CD8/38 (R² > 0.9) and CD19 (R² > 0.7). No changes in CD8, CD3, NK and HLADR act. cells were observed.

As seen in Fig. 3, determination of the inflammatory parameters showed a noticeable decrease in IL-2 (R² > 0.9) and IL-10 (R² > 0.7). Interestingly, we observed an increase in IL17-a during SP supplementation, which decreased again during follow-up. No changes were found in IFN- γ, IL-4 and TGF- β levels. IL-23 could not be measured as it was under the limit of quantification (LOQ) uniformly across all probes.

Microbiome and bowel regulation

The analysis of the microbiome between baseline and V9 showed a noticeable change in the taxonomic composition as illustrated in Fig. 4. At the taxonomic level of family, a dominance of the prevotella enterotype was observed in the study collective compared baseline. By V9, we observed a tendency towards a reduction in Prevotellaceae, Lachnospiraceae and Veillonellaceae. We have also found an increase in Ruminococcaceae, Porphyromonadaceae, Bacteroidaceae, Coriobacteriaceae and Erysipelotrichaceae.

The supplementation of SP was well tolerated. As represented in Data S2, during SP supplementation about half of the subjects, especially those who have suffered from diarrhoea or loose stools, started reporting a subjective improvement of stool consistency and bowel regulation.

Assessment of health-related Quality of Life with the SF-36 questionnaire

The results of the questionnaire evaluation, which are shown in Table 2, can be compared to the standardized results of the German normative population from 1998.³⁸ Overall, the scores of the HIV+ study cohort are at any point much lower as compared to the general population, without any exceptions.

Table 2. Comparison of Short Form (36) Health Survey parameters at baseline (B), visit 6 (V6) and visit 9 (V9) – self-reported by PLHIV (sodium propionate study cohort) (n = 23) vs. German normative population 1998 (n = 6967)

Scales	B	V6	V9	Data '98
Physical Function	68 ± 28	61 ± 30	68 ± 28	85 ± 21
Role Physical	40 ± 40	50 ± 49	50 ± 38	82 ± 33
Bodily Pain	45 ± 29	45 ± 24	55 ± 25	67 ± 26
General Health	48 ± 22	48 ± 24	50 ± 20	66 ± 18
Vitality	36 ± 21	35 ± 25	38 ± 18	60 ± 18
Social Function	52 ± 27	53 ± 32	52 ± 30	86 ± 20
Role Emotional	36 ± 43	42 ± 47	35 ± 42	89 ± 27
Mental Health	50 ± 28	52 ± 22	53 ± 23	72 ± 17
Physical Component Score	43 ± 11	42 ± 11	45 ± 12	48 ± 9
Mental Component Score	30 ± 17	32 ± 17	29 ± 16	51 ± 9

Corresponding to Fig. 5, the subscales in the PCS group are Physical Function (PF), Role Physical (RP), Bodily Pain (BP) and General Health (GH). The two parameters RP and BP increased slightly (R² > 0.7) indicating the reduction of impairment through physical problems in everyday life and less physical pain. The subscales in the MCS group are Vitality (VT), Social Function (SF), Role Emotional (RE) and Mental Health (MH). The MH scale can be described analogue to the GH scale as the general perception of mental health in the individual. It showed a consistent rise (R² > 0.8).

Discussion

To the best of our knowledge, dietary supplementation of SP in PLHIV has to date remained uninvestigated; therefore, we have for the first time evaluated the effects of SP supplementation, focusing on immunological/metabolic effects, changes in the microbiome and patient-reported QoL.

We observed a rise in CD4+ cell count, a decrease in pro-inflammatory cytokines and a decrease in Tregs. Also noticeable were improvements of metabolic parameters, individual items in patient-reported QoL, taxonomic changes in microbiome and improvement of bowel regulation.

Consistent with previous findings,^{9, 10, 15, 23, 33, 34} we observed a decrease in IL-2, depletion of TH1/TH17 and increase in CD4+ cells. Interestingly, while this is in line with previously reported effects, the decrease in Tregs and the decrease in IL-10 appears to be different. And while in the context of HIV infection Tregs regulate the immune system to suppress inflammation and spread of virus, the suppression of the immune activation also promotes latent HIV-1 reservoir formation.^{39, 40} It has been hypothesized that targeting Tregs for depletion might be a potential target for HIV research with HDAC inhibitors as potential cure strategies.^{39, 41}

Interestingly, while in the absence of infection short-chain fatty acids do increase IL10+ Tregs²⁹ the opposite effects have also been described, depending on environment, organ and disease.³¹ Two examples are the SCFA-induced inhibition of lipopolysaccharide stimulated IL-10 production¹² and the increase in inflammatory cytokines in the liver during increased C2 levels.⁴² It should be noted that in the latter study the intervention has been performed with acetate as SCFA. It has been indicated that acetate has low to none HDAC inhibitor activity in contrast to propionate.^{9, 31} It remains unclear in which pathological contexts the SCFA functions may be altered.⁴² In addition, it has also been shown that higher than physiological levels of SCFA can cause dysregulated T-cell responses and cause an inflammatory response.⁴³ While there are many studies describing a general dysbiotic state of the gut microbiome in PLHIV regardless of age,^44-49 a representative study demonstrating the physiological levels of SCFA in the gut of PLHIV has not yet been executed. There are only limited, inconsistent studies describing SCFA levels. In an elderly HIV+ Mexican population, unbalanced faecal SCFA levels were demonstrated at significantly higher levels of propionic acid than in the corresponding HIV- control group.⁵⁰ In contrast, a study performed in a Chinese HIV+ population demonstrated lower levels of SCFA.⁵¹ It is unclear what the actual SCFA basal levels are in German PLHIV, and how they contextually vary with age and sexual preferences. It is also unclear if higher or lower physiological levels may affect the outcome of SP supplementation.

In our HIV+ study cohort, the analysis of taxonomic composition of microbiome revealed a dominance of the Prevotella enterotype, thereby providing a rationale for study participants with looser stool.⁵² Consistent with our observation, such an enrichment of Prevotella has also been independently reported by other studies in PLHIV^{45-47, 53-55} and MSM.^{44, 56-58} It is important to note within this context that SCFA are fermented among others through Prevotellaceae. In our cohort, Prevotellaceae decreased by the end of the study, with about half of the cohort reporting a subjective improvement of stool consistency during the intervention. Furthermore, our data demonstrated lower basal Ruminococcaceae in PLHIV, that increased upon SP supplementation. Other studies describing common markers of PLHIV-microbiome found a relative depletion of Ruminococcaceae. Ruminococcaceae are not only associated with anti-inflammatory properties and gut homeostasis,^{44, 59, 60} but its depletion has also been described in young adults with depression.⁶¹ Interestingly, depression was also the most common comorbidity in our study group.

Our data demonstrate metabolic changes post-SP supplementation, such as a decrease in total cholesterol, LDL concentrations and in HbA1c. SCFA have been reported to participate in the glucose homeostasis by reducing serum glucose levels and insulin resistance.^{10, 20, 35, 36} This may be due to the stimulation of the release of the anorectic gut hormones peptide YY (PYY) and glucagon-like peptide-1 (GLP-1).^{62, 63} Furthermore, a recent study has suggested that SCFA could reduce lipogenesis and enhance lipolysis through regulating related hormones and genes,⁶⁴ which supports previous findings of blood lipid profile improvements.^{37, 65} This indicates that the mechanisms underlying beneficial metabolic actions of SCFA could also work in an HIV+ cohort. These and improvements concerning bowel regulation were associated with a better QoL. Despite our increased understanding of the impact of SP supplementation on PLHIV, the relationship between HIV infection and SCFA remains undeciphered at several levels and hence warrants further in-depth study.

Limitations of our study are that it was designed as a monocentric study with only a small number of subjects, a short duration of oral supplementation without documentation of nutrition and the absence of a standardized scale for stool consistency. Therefore, randomized controlled trials with a larger cohort are needed to further investigate the clinical effects of SP supplementation in PLHIV.

Patient consent

Obtained. The patients in this manuscript have given written informed consent to publish their case details.