Tese de Doutoramento Medicina, Investigação Clínica 2022 Faculdade de Ciências Médicas, Universidade NOVA de Lisboa

Introduction:Sulfotransferase enzymes (SULT) catalyzesulfoconjugation of drugs, as well as endogenousmediators, gut microbiota metabolites and environmental xenobiotics. To address the limitedevidence on sulfonation activity from clinical research, we conceiveda clinical metabolic phenotypingmethod using paracetamol,a drug with significant metabolism by SULT1A1 and SULT1A3, as a probe substrate. Our aim was to estimate sulfonation capabilityof phenolic compounds and study its intra and interindividual variabilityin vivoin man. Population and Methods: This thesis included two separate clinical trials:Clinical trial “Variability of Sulfotransferase 1A1 Activity in Humans: An Approach to Improve Predictive Drug Response –Part I: Analysis of Intraindividual Variation in Healthy Adults”(EudraCT nº 2016-001395-29)was approved by the Portuguese National Ethics Committee. We recruited 36 healthy adults (12 men and 24 women, 12 on oral contraceptives). Baseline samples were collected. Subjects received 1 g of oral paracetamol on three different occasions and had blood and urine samples collected after 2 hours. Paracetamol (P) and its metabolites (sulfate, glucuronate, cysteine-S-conjugate and mercapturate) were measured in plasma and spot urine samples using liquid chromatography-high resolution mass spectrometry(LC-HRMS). Paracetamol sulfonation index (PSI) was used to express SULT activity. PSI was calculated for urine (uPSI) and plasma (pPSI) samples. Twenty-nine subjects were genotyped for SULT1A1and genotypes were compared to phenotype measured with PSI.We measured expression of SULT1A1 and SULT1A3 in whole blood samples. Non-targeted metabolomics was performed in 27 pre-dose urine samples and 33 pre-dose plasma samples using LC-MS. LC-MS raw data was preprocessed using XCMS (R package xcms 3.6.2). Data was normalized by total area, centered and unit variance-scaled. Principal Component Analysis (PCA) was performed to assess the main sources of the variance of the data. Partial Least Square (PLS) analysis was performed to evaluate the correlation between the metabolites and the PSI using SIMCA (MKS Umetrics, Umeå, Sweden Umetrics, version 16.0.1). Metabolite annotation was performed by matching MSMS data against metabolomics databases (Metlin,HMDB).Clinical Trial “Variability of Sulfotransferase 1A1 Activity in Humans: An Approach to Improve Predictive Drug Response –Part II: Analysis of Interindividual Variation in Hypertensive Patients”(EudraCT 2019-002266-12)was approved by the Portuguese National Ethics Committee. We recruited 37patients with arterial hypertension. Baseline samples were collected. Subjects were phenotyped with the metabolic ratio PSI using paracetamol as probe drug, as previously described. PSI was calculated for urine (uPSI) samples.Both cohorts were combined to explore interindividual variability.Means, standard deviations, frequencies and percentages were used to describe the population under study.Multiple linear regression models wereperformed to study the role of several measurementsin the association withPSI. The linear regression model assumptions, namely the residuals homocedasticity and normality were checked and confirmed for all models obtained. 95% confidence intervals for the model’s coefficient estimates were obtained and presented. A confidence level of =0.05 was used throughout. Results:PSI was defined as the ratio between paracetamol sulfate (PS) and the sum of PS, paracetamol glucuronate (PG) and P: PSI=PS/(PS+PG+P). PSI showed low intraindividual variability, with a good correlation between values in plasma and spot urine samples. SULT1A1 exhibitedgenetic variation. We found a prevalence of 0.24 for the SULT1A1*2 allele, the most common SULT1A1 variant allele.Only two individuals possessed the allele SULT1A1*3. We found no association of genotype with expression or phenotype measuredwith uPSI.SULT1A3 was found to be correlated with uPSI.In untargeted analysis, PCA showed an influence of sex and contraceptive use on the baselinemetabolic profile. A PLS model was built between the pre-dose metabolic profile and the uPSI (R2(X)=0.249, R2(Y)=0.898, Q2=0.535, p=0.016). A total of 22 features in urine were selected as the most relevant ions of the PLS model (variable importance of the projection (VIP) value >2.0). Although no conclusive metabolite annotation was achievedfor any of the ions, MSMS information led as to elucidate that one of the metabolites was a glucuronidated metabolite. This metabolite increased with the PSI.In hypertensive patients, uPSI retained the capacity to segregate high and low sulfonators. Logistic regression revealed the influence of eGFR and of the presence of autoimmune disease on uPSI.Combining both cohorts, we found a coefficient of variation of uPSI between individuals of 21.9%. Logistics regression revealed an influence of age and of plasma creatinine on uPSI.Using the antimodewe defined 0.44 as the cut-offvalue between low and high sulfonate conjugators phenotypes. We did not detect differences between both phenotypes.Discussion: Our SULT phenotyping method is a simple non-invasive procedure requiring urine spot samples, using the safe and convenient drug paracetamol as a probe substrate, and with low intraindividual coefficient of variation. Although it will not give us mechanistic information, it will provide us an empirical measureof an individual’ssulfonator status.For the first time we have unveiled the presence of SULT1A1genetic variants in a Portuguese population and we did not find a genotype-phenotype relation. In the population studied, we have found an association between an individual’s baseline urinary metabolite profile and PSI, suggesting it could be used to estimate and compare sulfonation capacity between individuals.We discovered a modest interindividual variability of phenol sulfonation phenotype measured with uPSI. We detected an influence of plasma creatinine and age on uPSI that is yet to be understood. Conclusion: To the best of our knowledge, PSIprovides the first standardized in vivo empirical measure of an individual’s phenol sulfonation capability and of its intraindividual and interindividual variability.Furthermore, we have established the basis for the identification of a biomarker of sulfonation that could be useful for predicting sulfonator status from baseline metabolic profile. Both these initiatives constitute relevant tools in precision medicine.