Gut Microbiota-Generated Phenylacetylglutamine and Heart Failure

There are restrictions to the availability of some of the clinical data generated in the present study because we do not have permission in our informed consent from research subjects to share data other than in summary format outside our institution. Where permissible, the datasets generated and/or analyzed during the present studies, or summary data, and all methods used to conduct the research, are available from the corresponding author, upon reasonable request.

All clinical study protocols and informed consent for human subjects complied with the Declaration of Helsinki and received ethical approval by the Cleveland Clinic Institutional Review Board, or by the ethics committee of Charité-Universitätsmedizin Berlin. Written informed consent was obtained from all subjects. All animal model studies were approved by the Institutional Animal Care and Use Committee at the Cleveland Clinic.

Plasma samples and associated clinical data were collected as part of studies at a tertiary care referral center. All subjects gave written informed consent, and the Institutional Review Board of the Cleveland Clinic approved all study protocols. Metabolomics studies were performed on sequential samples from a large and well-characterized longitudinal tissue repository with associated clinical database named GeneBank at the Cleveland Clinic: Molecular Determinants of Coronary Artery Disease. GeneBank is registered under ClinicalTrials.gov Identifier: NCT00590200. It is composed of sequential stable subjects without evidence of acute coronary syndrome (cardiac troponin I <0.03 ng/mL) who underwent elective diagnostic coronary angiography (cardiac catheterization or coronary computed tomography) for evaluation of Coronary Artery Disease (CAD). All subjects had extensive clinical and longitudinal outcome data monitored, including adjudicated outcomes over the ensuing 3 to 5 years after enrollment. History of HF was detected by (a) directly asking patient by research personnel, (b) reviewing medical records for confirmation (all patients were seen by cardiologist at Cleveland Clinic before the left heart catheterization), and (c) International Classification of Diseases codes and adjudication by research personnel.¹² NT-proBNP levels were measured in all GeneBank samples by the Preventive Research Lab (PRL), a CAP and CLIA reference laboratory. Measurements for NT-ProBNP were completed using the Elecsys proBNP II STAT assay on the Roche Cobas e601 analyzer. For the present studies, we leveraged access to the previously reported PAGln levels obtained using stable isotope dilution liquid chromatography tandem mass spectrometry (LC/MS/MS) from sequential consenting subjects enrolled for whom PAGln, left ventricular ejection fraction, and NT-proBNP data were available (n=3256).⁸

Plasma samples (n=833) were also obtained from subjects enrolled in the LipidCardio at the Charité-Universitätsmedizin Berlin: Role of lipoproteins in cardiovascular disease.¹³ The study was approved by the local research ethics committee (approval number: EA1/135/16) under an approved protocol registered under German Clinical Trial Register (drks.de); Identifier: DRKS00020915. All participants provided written informed consent. Patients aged 18 years and older undergoing cardiac catheterization at a single large academic center (Department of Cardiology, Campus Benjamin Franklin, Charité-Universitätsmedizin Berlin), except those with troponin-positive acute coronary syndromes (ACS), were eligible for inclusion. HF was defined as New York Heart Association class ≥2. NT-proBNP was measured in all samples from subjects in the European Cohort using the Elecsys proBNP II STAT assay on the Roche Cobas e601 analyzer.

Stable-isotope-dilution LC/MS/MS was used for quantification of PAGln in human plasma (European Cohort), and PAGly in mouse plasma, as previously described.⁸ Briefly, ice cold methanol containing internal standard (D₅-phenylacetylglutamine) was added to the plasma samples, followed by vortexing and centrifuging (14 000 rpm; 4 °C for 15 min). The clear supernatant was transferred to glass vials with microinserts and submitted to LC/MS/MS analysis. LC/MS/MS analysis was performed on a chromatographic system consisting of 2 Shimadzu LC-30 AD pumps (Nexera X2), a CTO 20AC oven operating at 30 °C, and a SIL-30 AC-MP autosampler in tandem with a triple quadruple mass spectrometer (8050 series, Shimadzu Scientific Instruments, Inc., Columbia, MD). For chromatographic separation, a Kinetex C18 column (50 mm × 2.1 mm; 2.6 μm; Catalog No. 00B-4462-AN, Phenomenex, Torrance, CA) was used. Solvent A (0.1% acetic acid in water) and B (0.1% acetic acid in acetonitrile) were run using the following gradient: 0.0 min (0% B); 0.0 to 2.0 min (0% B); 2.0 to 5.0 min (0%B→20%B); 5.0 to 6.0 min (20%B→60%B); 6.0 to 7.5 min (60%B→70%B); 7.5 to 8.0 min (70%B→100%B); 8.0 to 9.5 min (100%); 9.5 to 10 min (100%B→0%B); 10.0 to 15.0 min (0% B) with a flow rate of 0.4 mL/min and an injection volume of 1 µL. Electrospray ionization in the positive mode was used with multiple reaction monitoring (MRM) for detection of endogenous and stable isotope labeled internal standard. The following transitions were used: m/z 265.2→130.15 for PAGln, m/z 193.8→76.1 for PAGly and m/z 270.1→130.15 for D₅-PAGln. D₅-PAGln was used as internal standard for both PAGln and PAGly. The following ion source parameters were applied: nebulizing gas flow, 3 l/min; heating gas flow, 10 l/min; interface temperature, 300 °C; desolvation line temperature, 250 °C; heat block temperature, 400 °C; and drying gas flow, 10 l/min. Limit of detection and limit of quantification for PAGln were 0.010 and 0.033 µM, respectively; limit of detection and limit of quantification for PAGly were 0.021 and 0.073 µM, respectively. Quality control samples were run with each batch of samples, and interbatch variations expressed as coefficient of variation were <10% for all analytes monitored. Data were collected and analyzed by LabSolution 5.91 software (Shimadzu).

H9c2(2-1) cells (American Type Culture Collection, ATCC; Manassas, VA) were seeded into 10 cm plates at 60% confluency. After seeding (24 hours), cells were then serum starved for 18 hours. Cells were exposed to 100 µM PAGln, 100 µM PAGly, or an equivalent volume of vehicle control for 4 hours. Following exposure, media was removed, cells washed in ice cold PBS, and 0.5 mL RNAlater was added to the plate. Cells were removed by scrapping and stored at −20 ºC for RNA isolation. Prior to isolation, 1 mL of sterile PBS was added to RNAlater samples to allow for the pelleting of cells. Cells were pelleted at 10 000g for 3 minutes. After supernatant was removed, cells were lysed in 250 µL of TRIzol (Ambion, Inc; Austin, TX) by vortexing. Chloroform (50 µL) was added to the TRIzol homogenate and samples incubated at room temperature for 3 minute. Then, tubes were centrifuged at 10 000g for 15 minutes at 4 ºC for phase separation. The aqueous layer was removed, being careful not to disturb the interface, and mixed with 150 µL isopropyl alcohol. Samples were incubated at room temperature for 10 minutes. Tubes were centrifuged at 12 000g for 10 minutes at 4 ºC for pelleting. After the supernatant was removed, RNA pellets were washed twice with 75% ethanol by vortexing and centrifuged at 7500g for 5 minutes at 4 ºC. Supernatant was removed and pellets were allowed to air-dry for 10 minutes before resuspension in DPEC treated water. Nppb (Natriuretic Peptide Precursor B gene) expression was measured using 400 ng of total RNA using the TaqMan RNA-to-Ct 1-step kit (Applied Biosystems, Waltham, MA) on the Applied Biosystems StepOnePlus Real-Time PCR System. Rn00580641_m1 (FAM) was used for detection of Nppb and Rn01775763_g1 (VIC) was used for detection of Gapdh (glyceraldehyde-3-phosphate dehydrogenase).

C57BL6/J male mice (6- to10-weeks-old) were purchased from Laboratory (Bar Harbor, ME) and housed in the Cleveland Clinic Biological Research Unit under strict 14:10 light:dark cycles, with food and water provided ad libitum. Only male mice were used to mitigate the variable of female reproductive hormones on cardiac function.

Animals were IP injected with pH-neutralized PAGln (50 mg/kg; n=16), PAGly (50 mg/kg; n=16) or isotonic vehicle control (n=14). PAGln and PAGly were dissolved in normal saline and neutralized with NaOH; the vehicle control solution consisted of an equivalent amount of NaOH/HCl added to the normal saline. Fifteen minutes after injection, animals were euthanized by Ketamine/Xylazine overdose (300 mg/kg + 30 mg/kg); blood was collected by cardiac puncture of the right ventricle. Right after blood collection, hearts were excised from the animal and stored in RNAlater. At a later time, the entire left atria was isolated from each heart. Atria were lysed in 250 µL of TRIzol by bead beating with a single 5.0-mm zirconium oxide bead. Chloroform (50 µL) was added to the TRIzol homogenate and samples were incubated at room temperature for 3 minutes. Tubes were centrifuged at 10 000g for 15 minutes at 4 ºC for phase separation. The aqueous layer was removed, being careful not to disturb the interface, and mixed with 150 µL isopropyl alcohol. Samples were incubated at room temperature for 10 minutes. Tubes were centrifuged at 12 000g for 10 minutes at 4 ºC for pelleting. After the supernatant was removed, RNA pellets were washed twice with 75% ethanol by vortexing and centrifuged at 10 000g for 8 minutes at 4 ºC. Supernatant was removed, and pellets were allowed to air-dry for 10 minutes before resuspension in DPEC treated water. Nppb expression was measured using 1.5-µL total RNA using the TaqMan RNA-to-Ct 1-step kit (Applied Biosystems; Waltham, MA) on the Applied Biosystems StepOnePlus Real-Time PCR System. Mm00435304_g1 (FAM) was used for detection of Nppb and Mm99999915_g1 (VIC) was used for detection of Gapdh. The investigators were not blinded for these studies.

Adult wild-type mouse cardiomyocytes were isolated, and contractility studies were performed utilizing the IonOptix System (Myopace, Milton, MA) as previously described.^14,15 Briefly, myocytes were isolated and plated on glass chamber slides, which were placed on a Leica microscopic stage connected to a field stimulator specifically designed for driving isolated myocytes (MyoPace, IonOptix). Cardiomyocytes were stimulated at 3 Hz and imaged with a variable field-rate camera (MyoCam, IonOptix) using edge-detection and sarcomere length technology. Myocytes were treated with 10 µM of epinephrine and cardiomyocytes contractility, transients were recorded in the presence or absence of 100 µM PAGln or PAGly, and pre-incubated for 15 minutes, before addition of epinephrine. Single sarcomere contraction cycles were recorded with time as contractility transients. Sarcomere length (µm) and sarcomere shortening (µm) were measured considering 5 different scanning windows.

We previously reported that plasma levels of PAGln are associated with CVD and incident major adverse cardiovascular events (myocardial infarction, stroke, or death).⁸ In this same study, we showed, through a variety of mechanistic investigations, that PAGln signaled via adrenergic receptors. Because altered sympathetic tone is a key contributor to the development of HF,^18,19 we sought to assess if systemic PAGln levels show a clinical association with HF risk and severity. For initial analyses, we leveraged access to previously reported PAGln data (US Cohort; see Methods section for details), in order to examine the association between PAGln levels, HF, LVEF, and NT-proBNP. Subject (n=3256) clinical characteristics, demographics, and laboratory values are provided in Table 1, which shows a middle-aged cohort with high prevalence of CVD and related risk factors. Subjects with the adjudicated diagnosis of HF (characterized in Table S1) had higher systemic levels of PAGln compared with nonHF subjects (P<0.0001; Figure 1A). Further, individuals with HF without CAD had higher levels of PAGln than control subjects (no-HF and no-CAD; Figure S1A), and individuals with CAD and HF had higher levels of PAGln than subjects with CAD alone (Figure S1A). Moreover, an increased risk for HF was observed among those with higher circulating levels of PAGln (eg, fourth versus first quartile), even following adjustments for traditional CVD risk factors and indices of kidney function (Figure 1B).

We next sought to validate the observed association between circulating PAGln levels and HF risk using an independent cohort. For these analyses, we measured serum PAGln levels from subjects enrolled in a similar registry of sequential cardiology patients undergoing coronary angiography at a tertiary referral center in Europe (European Cohort; for details, see Methods above). Subject clinical characteristics, demographics, and laboratory values are provided in Table 2, which shows an elderly cohort with high CVD risk factor prevalence. As observed in the US Cohort, subjects with HF in the European Cohort (characterized in Table S1) had significantly higher serum levels of PAGln compared with those without HF (Figure 1C). Further, individuals with HF without coronary artery disease (CAD) had higher levels of PAGln than control subjects (no-HF and no-CAD; Figure S1B), and individuals with CAD and HF had higher levels of PAGln than subjects with CAD alone (Figure S1B). Moreover, circulating PAGln levels were again associated with increased risk of HF (fourth versus first quartile), even after multivariable logistic regression adjustments for both traditional CVD risk factors and measures of kidney function (Figure 1D). The association between elevated levels of PAGln and HF risk observed within both the US and European Cohorts also appears to hold true when individuals (combined cohort analysis) are categorized by LVEF status (ie, HFpEF, LVEF ≥50; HF with mildly reduced ejection fraction [HFmEF], 40<LVEF<50; or HF with reduced ejection fraction [HFrEF], LVEF ≤40; Figure 2).

In both the US and European Cohorts, increasing plasma levels of PAGln were dose-dependently associated with declining LVEF (Figure 3A and 3B; ρ=−0.144 [P<0.0001] and ρ=−0.185 [P<0.0001]). Similarly, increasing levels of PAGln in both the US and European Cohorts showed a dose-dependent association with increasing serum levels of N-terminal prohormone of BNP (NT-proBNP), a cardiac hormone generated in response to subclinical myocardial strain²⁰ (Figure 3C and 3D; ρ=0.293 [P<0.0001] and ρ=0.324 [P<0.0001]). Of interest, we found that the association of PAGln with NT-proBNP remained significant in both the US and European Cohorts, when looking at the subsets of subjects with preserved left ventricular systolic function (eg, LVEF ≥50; P<0.0001 [Kruskal-Wallis test] for both cohorts; Figure S2A and S2B), as well as in the subsets with normal renal function (eg, estimated glomerular filtration rate ≥90 mL/min per 1.73 m²; P<0.0001 and P=0.001 [Kruskal-Wallis test] for US and EU Cohorts, respectively; Figure S3). Moreover, an increased risk for HF was observed among those with higher circulating levels of PAGln (eg, fourth versus first quartile), even following adjustments for traditional CVD risk factors, indices of kidney function and LVEF and NT-proBNP (Figure S4).

Given the striking association observed between PAGln and NT-proBNP levels in the 2 clinical cohorts, we sought to determine if PAGln, or its rodent counterpart phenylacetylglycine (PAGly),⁸ could directly induce BNP (Nppb) gene expression. To initially test this hypothesis in vitro, we used the immortalized rat cardiomyoblast cell line, H9c2. After 4 hours of exposure to pathophysiologically relevant levels of either PAGln or PAGly (Methods), Nppb (BNP) expression (normalized to Gapdh) was observed to be significantly increased (3.71-fold for PAGln [P=0.03], 3.66-fold for PAGly [P=0.03]) compared with vehicle-treated cells (Figure 4A). Advancing next to in vivo studies, we observed that when systemic levels of either PAGln or PAGly were acutely elevated after intraperitoneal injection (versus vehicle control), Nppb expression (normalized to Gapdh) in the left atria was significantly increased compared with vehicle-treated animals (1.66- and 1.46-fold increase, P=0.01 and P=0.02, 15 min after injection of PAGln or PAGly, respectively; Figure 4B). Intraperitoneal injection of PAGly also lead to increased Nppb expression within left ventricle (1.5-fold, P=0.001; data not shown).

Having observed a clinical association between PAGln and reduced LVEF, we next examined the effect of PAGln or PAGly on cardiomyocyte function. When examining freshly isolated primary murine cardiomyocytes, PAGln or PAGly treatment alone did not alter the baseline shortening of cardiomyocyte sarcomere length in response to repetitive field stimulation compared with vehicle treatment (Figure 5A through 5D, black vehicle versus blue line, PAGln or PAGly). Epinephrine (Epi) treatment alone resulted in a significant increase in inotropic response of the primary adult murine cardiomyocytes (Figure 5A through 5D, green line, Epi). However, when cardiomyocytes were incubated with both Epi and either PAGln or PAGly, the positive inotropic effect of Epi was significantly blunted, as indicated by the reduction in Epi-induced increase in contraction (sarcomere shortening; Figure 5A through 5D, red line, Epi+PAGln or Epi+PAGly). These results are consistent with PAGln (and PAGly) fostering an overall myocardial depressant effect during sympathetic stimulation.

This work is supported by grants from the National Institutes of Health (NIH) and the Office of Dietary Supplements (P01 HL147823, R01HL103866, R01HL126827) and the Foundation Leducq (17CVD01). Dr Romano was supported in part by training grant T32HL134622 from the National Heart, Lung, and Blood Institute of the NIH. Dr Witkowski was supported by an award from the Deutsche Forschungsgemeinschaft (WI 5229/1-1). Mass spectrometry studies were performed on instrumentation housed in a facility supported in part through a Shimadzu Center of Excellence award.

Table S1

Figures S1–S4