Amino Acid Metabolite Profiling for Predicting and Understanding the Metabolic Effects of Metabolic and Bariatric Surgery

Abstract

Metabolic and bariatric surgery induces metabolic benefits beyond weight loss, including improved insulin sensitivity, type 2 diabetes (T2D) remission, and reduced inflammation. Recent metabolomics research highlights amino acid metabolites—branched-chain amino acids, aromatic amino acids, and tryptophan-derived compounds—as key biomarkers for predicting surgical outcomes. Elevated preoperative levels of isoleucine, phenylalanine, levodopa, and 3-hydroxyanthranilic acid are associated with improved glycemic control and T2D remission. Gut microbiota-derived metabolites, including indole-3-propionic acid and indole-3-pyruvic acid, contribute to insulin sensitivity and lean mass preservation. Additionally, serotonin-related indicators predict postoperative weight loss rates. These metabolites reflect individual metabolic heterogeneity and may outperform conventional clinical models in predicting surgical responsiveness. Integration of metabolomics into preoperative assessment may enhance precision medicine approaches and identify new therapeutic targets.

INTRODUCTION

Metabolic and bariatric surgery (MBS) is known as the most effective method for treating various metabolic diseases, including severe obesity and type 2 diabetes (T2D). Beyond weight loss, systemic metabolic effects such as increased insulin sensitivity, improved glycemic control, reduced hepatic fat, and regulated blood lipid levels have been reported after surgery. However, the molecular mechanisms that induce these clinical effects are not yet clearly understood. Recent advancements in metabolomics technology provide important clues to understanding these mechanisms by quantitatively analyzing changes in metabolites before and after MBS. In particular, among amino acid metabolites, branched-chain amino acids (BCAAs), aromatic amino acids (AAAs), and tryptophan-derived metabolites show a close association with metabolic health and play a key role in predicting and understanding surgical effects.

ASSOCIATION BETWEEN OBESITY AND AMINO ACID METABOLITES

1. BCAA (isoleucine, leucine, valine)

BCAAs consist of isoleucine, leucine, and valine [1], and play important roles in energy metabolism, muscle growth, and the regulation of insulin action [2,3]. In obese patients, blood concentrations of BCAAs are consistently high, which is strongly associated with insulin resistance [4]. This accumulation of BCAAs is related to a decrease in oxidative capacity in the liver and muscles, meaning a reduction in the efficiency of catabolic pathways, and leads to the inhibition of insulin signaling through the mammalian target of rapamycin complex 1 pathway [5,6]. Furthermore, BCAAs contribute to the progression of metabolic diseases through mechanisms such as inducing inflammation in adipocytes, disrupting the tricarboxylic acid cycle, and causing glucose metabolic imbalance [7,8].

Blood BCAA concentrations significantly decrease after MBS, which is interpreted as an indicator reflecting the recovery of metabolic function in the liver and muscles, beyond a simple weight loss effect [9,10]. Especially for isoleucine, studies have shown that higher preoperative concentrations are associated with greater postoperative weight loss, demonstrating its potential as a prognostic biomarker [11]. Additionally, gut microbiota have also been found to influence BCAA metabolism, with reports indicating that changes in microbial communities after surgery affect BCAA concentrations [12].

2. Phenylalanine and tyrosine

Phenylalanine and tyrosine are AAAs that are closely associated with liver function and neurotransmitter biosynthesis [13]. In obese conditions, blood concentrations of these amino acids increase, reflecting impaired liver detoxification capacity and metabolic stress [14,15]. Phenylalanine, in particular, is linked to systemic inflammatory markers, and persistent high concentrations can lead to metabolic diseases such as impaired liver function, increased insulin resistance, and worsening cardiovascular risk factors [16,17]. Tyrosine acts as a precursor for dopamine and norepinephrine, and these neurotransmitters are linked to the regulation of energy intake, reward behavior, and pancreatic beta-cell function [18,19]. Preoperative tyrosine and levodopa (L-DOPA) concentrations showed a correlation with postoperative T2D remission, suggesting that this pathway directly influences insulin secretion and sensitivity regulation [20,21]. Furthermore, dopamine can inhibit insulin secretion through anti-incretin action, so a decrease in dopamine production may improve insulin response [22]. Therefore, these AAAs can be considered causative agents of metabolic diseases and biological indicators that can predict surgical outcomes.

3. Tryptophan and its derivatives

Tryptophan is an essential amino acid that is metabolized through three main pathways: the kynurenine pathway, the serotonin pathway, and the indole pathway [23]. These pathways respectively mediate immune regulation, energy storage and behavioral control, and interaction with gut microbiota [24,25]. In obesity, the kynurenine pathway is abnormally activated, and concentrations of kynurenine and its downstream metabolites (e.g. 3-hydroxyanthranilic acid [3-HAA], kynurenic acid) increase due to the overactivity of the inflammation-inducing enzyme indoleamine 2,3-dioxygenase [26]. These metabolites affect immune cell function and insulin sensitivity, and high concentrations can lead to metabolic disorders [27,28].

The serotonin pathway involves the conversion of tryptophan to serotonin via 5-hydroxytryptophan (5-HTP) [29]. Peripheral serotonin induces lipid accumulation in adipocytes and regulates pancreatic insulin secretion [30]. Studies have shown that higher preoperative serotonin concentrations are associated with a slower rate of postoperative weight loss, and the serotonin/5-HTP ratio has been proposed as an indicator for predicting metabolic responsiveness [11].

The indole pathway is formed by the breakdown of tryptophan by gut microbiota and includes indole-3-propionic acid (IPA), indole-3-acetic acid (IAA), and indole-3-lactic acid [31]. In particular, IPA has antioxidant effects, protects the intestinal mucosa, and improves insulin sensitivity, and an increase in IPA concentration after surgery is associated with positive metabolic responses [32,33]. As such, tryptophan-derived metabolites, independently yet interactively, are considered key factors regulating the pathophysiology of obesity and metabolic diseases, as well as surgical responsiveness.

ASSOCIATION BETWEEN MBS AND AMINO ACID METABOLITES

1. Improvement of post-surgical insulin secretion and resistance with amino acid and gut microbiota metabolites

We evaluated the relationship between amino acids and tryptophan-derived gut microbiota metabolites before and after MBS and the improvement of insulin secretion and insulin resistance [34]. Among various amino acid metabolites measured before surgery, phenylalanine, leucine, tryptophan, and tyrosine showed a significant correlation with changes in the insulinogenic index and insulin resistance indices after surgery. In particular, the sum of large neutral amino acids and phenylalanine showed the highest predictive power, with area under the curve (AUC) values of 0.90–0.94 in receiver operating characteristic (ROC) analysis, suggesting their potential as biomarkers to predict improved insulin secretion and sensitivity based solely on preoperative concentrations.

Furthermore, values of 3-HAA, 3-hydroxykynurenine, xanthurenic acid, and anthranilic acid, belonging to the kynurenine pathway of tryptophan metabolism, showed a significant correlation with the improvement of insulin resistance after surgery. These metabolites possess biological properties such as inflammation regulation, oxidative stress alleviation, and pancreatic beta-cell protection [35,36], with 3-HAA in particular demonstrating its potential as a sensitive indicator of surgical responsiveness through its anti-inflammatory and antioxidant functions [37]. Additionally, changes in IAA and IPA among gut microbiota-derived indole metabolites were also associated with improved insulin resistance and secretion, implying that changes in the gut microbiota metabolic environment after MBS can contribute to metabolic improvement [38,39]. This study suggests that various amino acid metabolites and gut microbiota metabolites can act as regulators and predictive indicators of surgical responsiveness, demonstrating the possibility of patient selection based on precision medicine and biomarker-based predictive model development in the future (Table 1).

Table 1. Amino acid metabolites and clinical relevance.

Metabolite	Metabolic role	Clinical insight
Isoleucine	Energy, insulin sensitivity	↑ Pre-op → ↑ post-op weight loss (%EWL)
Phenylalanine	Oxidative stress, insulin resistance	Marker of metabolic stress, predictor of T2D improvement
L-DOPA	β-cell signaling, insulin secretion	↑ Pre-op → predictor of T2D remission
3-HAA	Antioxidant, β-cell protection	Strongest predictor of insulin resistance improvement
Serotonin/5-HTP ratio	Lipogenesis, energy intake	↑ Pre-op → slower weight loss post-op
IPA	Antioxidant, mucosal barrier, insulin sensitizer	↑ Post-op → better insulin function
IPyA	Muscle regeneration, anti-atrophy	↑ Pre-op → predicts FFM preservation post-surgery

%EWL = percentage of excess weight loss, T2D = type 2 diabetes, L-DOPA = levodopa, 3-HAA = 3-hydroxyanthranilic acid, 5-HTP = 5-hydroxytryptophan, IPA = indole-3-propionic acid, IPyA = indole-3-pyruvic acid, FFM = fat-free mass.

2. Prediction of post-surgical diabetes remission based on amino acid metabolites

We evaluated whether preoperative blood amino acid metabolite concentrations can predict remission of T2D and compared their performance with existing clinical predictive models [40]. A total of 16 glucose metabolism-related amino acid metabolites were analyzed, among which L-DOPA (tyrosine pathway) and 3-HAA (kynurenine pathway) were identified as the best metabolites for predicting T2D remission at 12 months post-surgery. The baseline concentrations of these two metabolites were significantly higher in the remission group, and their AUC values were 0.92 and 0.85, respectively, showing superior performance compared to existing clinical predictive models such as ABCD (0.81), DiaRem, and IMS.

L-DOPA, as a dopamine precursor, is involved in the auto-inhibitory signals of pancreatic beta-cells, and higher preoperative concentrations were associated with better recovery of blood glucose control and insulin secretion after surgery. On the other hand, 3-HAA, an anti-inflammatory metabolite of the tryptophan-kynurenine pathway, is interpreted to have a clearer effect in reducing metabolic stress and protecting beta-cells at higher concentrations. The levels of these metabolites maintained their predictive power up to 3 months post-surgery, suggesting that the initial metabolic state can influence long-term glycemic control outcomes after surgery. In particular, this study demonstrated that metabolites can provide 'qualitative' information about metabolic status that existing clinical indicators do not capture, highlighting the potential of metabolites as unique indicators reflecting physiological heterogeneity at the molecular level, independent of quantitative values (age, body mass index, diabetes duration). In conclusion, this study provides strong evidence that L-DOPA and 3-HAA can be used as preoperative predictive biomarkers, supporting the need for metabolite-based precision predictive model development.

3. Prediction of early weight loss rate using metabolites

The rate of weight loss among patients undergoing sleeve gastrectomy varies significantly, and preoperative metabolic status can influence the outcome. This study showed that some of the 20 obesity-related amino acid metabolites measured preoperatively were significantly associated with the percentage of excess weight loss (%EWL) at 3 and 6 months post-surgery [11]. Specifically, isoleucine among BCAAs showed a significant positive correlation with postoperative %EWL, while metabolites of the serotonin pathway, namely serotonin, 5-hydroxyindoleacetic acid (5-HIAA), and the serotonin/5-HTP ratio, showed strong performance in predicting slow weight loss after surgery. Higher serotonin concentrations and higher serotonin/5-HTP ratios were associated with slower weight loss, while conversely, higher 5-HIAA and 5-HIAA/serotonin ratios tended to be associated with faster weight loss. This result is interpreted as being related to the energy storage promotion, increased insulin secretion, and fat accumulation promotion functions of peripheral serotonin. In ROC analysis, serotonin and the serotonin/5-HTP ratio showed high predictive power with AUC values of 0.79–0.81, suggesting that preoperative serotonin metabolic profiles can be valid biomarkers for predicting early weight loss rates.

This study also analyzed ratio indicators reflecting enzyme activity of the serotonin pathway (tryptophan hydroxylase, aromatic L-amino acid decarboxylase, monoamine oxidase A), emphasizing the importance of sophisticated metabolite analysis reflecting even the enzyme activity status in metabolic pathways for predicting surgical responsiveness. This approach, by utilizing information on pathway-based functional metabolic flow in addition to single metabolite concentrations, provides important evidence for the establishment of future metabolite-based precision medicine strategies.

4. Preservation of lean body mass after surgery with tryptophan-derived metabolites

Rapid weight loss after MBS inevitably accompanies the loss of fat-free mass (FFM), which can lead to a decrease in basal metabolic rate and musculoskeletal function. Consequently, recent studies have highlighted the importance of FFM preservation in addition to the amount of weight loss as a clinically significant indicator. This study analyzed the effect of indole-3-pyruvic acid (IPyA) concentrations, measured in preoperative serum from 42 patients who underwent sleeve gastrectomy, on changes in FFM after surgery [41]. The patient group with relatively preserved FFM at 3–6 months showed significantly higher preoperative IPyA concentrations, and in ROC analysis, IPyA showed a predictive power of AUC 0.763 (P=0.006) for predicting FFM gain.

IPyA is a tryptophan-derived metabolite produced by gut microbiota and muscle stem cells, and it activates the aryl hydrocarbon receptor (AhR) to regulate inflammation and contribute to muscle regeneration. In particular, this study suggested the possibility that IPyA can not only preserve FFM but also reduce the risk of postoperative weight regain. This is based on a physiological mechanism where inflammation suppression and atrophy inhibition functions act together through AhR signaling in muscles. Furthermore, the FFM preservation group also had a higher rate of weight loss, showing significant differences at both 6 and 12 months post-surgery. As such, IPyA can function as a precise biomarker reflecting muscle regeneration and energy metabolism regulation, beyond simple body composition changes, and suggests the possibility of therapeutic intervention through tryptophan-IPA metabolic pathway modulation in the future. This study is one of the first cases to emphasize the role of metabolites in the novel clinical goal of FFM preservation, expanding the potential of biological indicators reflecting preoperative metabolic status.

CONCLUSIONS

MBS induces various metabolic effects beyond simple weight loss, including remission of T2D, improvement of insulin resistance, and inflammation suppression. Metabolomics is gaining attention as a tool to understand the molecular basis of these effects and to predict patient-specific surgical responsiveness. In particular, amino acid metabolites such as BCAAs, AAAs, and tryptophan derivatives are closely associated with obesity and metabolic diseases, and their preoperative concentrations can be useful in predicting the extent of metabolic improvement after surgery. In the future, it is expected that personalized surgical strategies based on metabolomics will become possible through multi-omics integration analysis and machine learning-based predictive model development. Ultimately, such research can extend beyond predicting postoperative prognosis to the discovery of non-surgical alternative therapeutic targets.