Simultaneous high-performance liquid chromatographic determination of amino acids in a dried blood spot as a neonatal screening test

Random forest classifier improving phenylketonuria screening performance in two Chinese populations

Phenylketonuria (PKU) is a genetic disorder with amino acid metabolic defect, which does great harms to the development of newborns and children. Early diagnosis and treatment can effectively prevent the disease progression. Here we developed a PKU screening model using random forest classifier (RFC) to improve PKU screening performance with excellent sensitivity, false positive rate (FPR) and positive predictive value (PPV) in all the validation dataset and two testing Chinese populations. RFC represented outstanding advantages comparing several different classification models based on machine learning and the traditional logistic regression model. RFC is promising to be applied to neonatal PKU screening.

From bedside to bench-practical considerations to avoid pre-analytical pitfalls and assess sample quality for high-resolution metabolomics and lipidomics analyses of body fluids

The stability of lipids and other metabolites in human body fluids ranges from very stable over several days to very unstable within minutes after sample collection. Since the high-resolution analytics of metabolomics and lipidomics approaches comprise all these compounds, the handling of body fluid samples, and thus the pre-analytical phase, is of utmost importance to obtain valid profiling data. This phase consists of two parts, sample collection in the hospital (“bedside”) and sample processing in the laboratory (“bench”). For sample quality, the apparently simple steps in the hospital are much more critical than the “bench” side handling, where (bio)analytical chemists focus on highly standardized processing for high-resolution analysis under well-controlled conditions. This review discusses the most critical pre-analytical steps for sample quality from patient preparation; collection of body fluids (blood, urine, cerebrospinal fluid) to sample handling, transport, and storage in freezers; and subsequent thawing using current literature, as well as own investigations and practical experiences in the hospital. Furthermore, it provides guidance for (bio)analytical chemists to detect and prevent potential pre-analytical pitfalls at the “bedside,” and how to assess the quality of already collected body fluid samples. A knowledge base is provided allowing one to decide whether or not the sample quality is acceptable for its intended use in distinct profiling approaches and to select the most suitable samples for high-resolution metabolomics and lipidomics investigations.

Improving the Diagnosis of Phenylketonuria by Using a Machine Learning-Based Screening Model of Neonatal MRM Data

Phenylketonuria (PKU) is a common genetic metabolic disorder that affects the infant's nerve development and manifests as abnormal behavior and developmental delay as the child grows. Currently, a triple–quadrupole mass spectrometer (TQ-MS) is a common high-accuracy clinical PKU screening method. However, there is high false-positive rate associated with this modality, and its reduction can provide a diagnostic and economic benefit to both pediatric patients and health providers. Machine learning methods have the advantage of utilizing high-dimensional and complex features, which can be obtained from the patient's metabolic patterns and interrogated for clinically relevant knowledge. In this study, using TQ-MS screening data of more than 600,000 patients collected at the Newborn Screening Center of Shanghai Children's Hospital, we derived a dataset containing 256 PKU-suspected cases. We then developed a machine learning logistic regression analysis model with the aim to minimize false-positive rates in the results of the initial PKU test. The model attained a 95–100% sensitivity, the specificity was improved 53.14%, and positive predictive value increased from 19.14 to 32.16%. Our study shows that machine learning models may be used as a pediatric diagnosis aid tool to reduce the number of suspected cases and to help eliminate patient recall. Our study can serve as a future reference for the selection and evaluation of computational screening methods.

Urea Cycle Related Amino Acids Measured in Dried Bloodspots Enable Long-Term In Vivo Monitoring and Therapeutic Adjustment

BACKGROUND

Dried bloodspots are easy to collect and to transport to assess various metabolites, such as amino acids. Dried bloodspots are routinely used for diagnosis and monitoring of some inherited metabolic diseases.

METHODS

Measurement of amino acids from dried blood spots by liquid chromatography-tandem mass spectrometry.

RESULTS

We describe a novel rapid method to measure underivatised urea cycle related amino acids. Application of this method enabled accurate monitoring of these amino acids to assess the efficacy of therapies in argininosuccinate lyase deficient mice and monitoring of these metabolites in patients with urea cycle defects.

CONCLUSION

Measuring urea cycle related amino acids in urea cycle defects from dried blood spots is a reliable tool in animal research and will be of benefit in the clinic, facilitating optimisation of protein-restricted diet and preventing amino acid deprivation.

Clinical characterization of tremor in patients with phenylketonuria

BACKGROUND

Phenylketonuria (PKU) is due to the deficit of the enzyme phenylalanine hydroxylase, the first step of dopamine synthesis. If not early treated the disease results in severe neurological impairment. Minor neurological signs have been reported in early treated PKU (ETPKU) subjects. Prolactin level is affected by (and reflects) brain dopamine availability. Object of the study was to assess the occurrence, age at onset, distribution, associated neurological signs, and possible pathogenetic biomarkers of tremor in ETPKU.

METHODS

Fifty-nine ETPKU and 43 control subjects (age range 7-54) underwent individual and familiar tremor history, clinical assessment of tremor by means of the Fahn-Tolosa-Marin Tremor Rating Scale, and IQ evaluation. Historical and concomitant biochemical data (blood levels of Phe) and serum prolactin were included in the analysis.

RESULTS

Thirty-two percent of ETPKU patients were affected by postural and kinetic tremor. We found a significant correlation between severity of tremor and: prolactin level at the day of examination (part A: r_s = 0.320; p = .014; part C: r_s = 0.319; p = .014), Phe fluctuations from 12 years onwards (part B: r_s = 0.300; p = .036). We also found a significant correlation between prolactin (18.2 ± 9.6 ng/ml) and Phe levels (852 ± 472 μmol/l) on the day of assessment (r_s = 0.470; p < .001).

CONCLUSIONS

The main clinical features of tremor in ETPKU evoke those of essential tremor, although with a higher prevalence and an earlier onset than in general population. The severity of tremor was related to concomitant prolactin rather than Phe levels. This pattern suggests that metabolic alterations associated with PKU may result in an anticipation of the tremor onset in subjects who are possibly prone to this disorder.

Preconception Micronutrient Supplementation Reduced Circulating Branched Chain Amino Acids at 12 Weeks Gestation in an Open Trial of Guatemalan Women Who Are Overweight or Obese

Elevated branched chain amino acids (BCAAs: valine, leucine, and isoleucine) are well-established biomarkers of obesity-associated insulin resistance (IR). Mounting evidence suggests that low- and middle-income countries are suffering from a "double burden" of both undernutrition (growth stunting) and overnutrition (obesity) as these countries undergo a "nutrition transition". The purpose of this study was to examine if pre-pregnancy body mass index (BMI, kg/m²) and a daily lipid-based micronutrient supplement (LNS, Nutriset) would lead to cross-sectional differences in circulating levels of branched chain amino acids (BCAAs) in Guatemalan women experiencing short stature during early pregnancy. Using data from an ongoing randomized controlled trial, Women First, we studied women who were normal weight (NW, BMI range for this cohort = 20.1⁻24.1 kg/m²) or overweight/obese (OW/OB, BMI range for this cohort = 25.6⁻31.9 kg/m²), and divided into two groups: those who received daily LNS ≥ 3 months prior to conception through 12 weeks gestation (+LNS), or no LNS (-LNS) (n = 9⁻10/group). BCAAs levels were obtained from dried blood spot card samples (DBS) assessed at 12 weeks gestation. DBS cards provide a stable, efficient, and reliable means of collecting, transporting, and storing blood samples in low resource or field settings. Circulating maternal leptin, adiponectin, and insulin were determined by immunoassays from serum samples collected at 12 weeks gestation. We found maternal pre-pregnancy body mass index (ppBMI) was associated with higher circulating BCAAs (r² = 0.433, p = 0.002) and higher leptin/adiponectin ratio (r = 0.466, p = 0.044) in -LNS mothers at 12 weeks gestation. +LNS mothers demonstrated no correlations between BCAAs or leptin/adiponectin ratio across ppBMI suggesting LNS may be effective at improving metabolic status in OW/OB mothers during early pregnancy.

The spectrum of phenylalanine variations under tetrahydrobiopterin load in subjects affected by phenylalanine hydroxylase deficiency

A fall in blood phenylalanine (Phe) after tetrahydrobiopterin (BH(4)) administration is a common trait in phenylalanine hydroxylase (PAH, EC 1.14.16.1) deficiency (McKusick 261600). To explore the extent and biological correlates of this phenomenon we studied: (a) the spectrum of BH(4) response in patients with PAH deficiency; (b) the variability of BH(4) response according to the severity of the biochemical phenotype; and (c) the variability of the response to BH(4) in subjects with the same genotype. Fifty PAH-deficient subjects (age 1 month-35 years) were enrolled for the study (5 with mild hyperphenylalaninaemia (MHPHE), 15 with mild phenylketonuria (MPKU) and 30 with classic phenylketonuria (CPKU) and underwent an identical schedule of blood samplings 24 h before and after oral BH(4) challenge (6(R)-BH4, 20 mg/kg per day), leaving Phe intake unchanged. The effect of BH(4) on blood Phe concentration was evaluated according to the percent decrease of Phe during the 24 h following the challenge (criterion a), and as variation exceeding the individual variability of blood Phe (criterion b). The number of BH(4)-responders according to criterion b was 31 (including all the 14 detected by criterion a): 17 out of 30 CPKU (57%), 9 out of 15 MPKU (60%), and all the MHPHE subjects (chi(2) = 3.45, df = 2, p = 0.178). The effect of BH(4) showed a large interindividual variability unrelated to diagnostic classification, basal value of blood Phe, maximum percentage of Phe reduction, Phe intake, and genotype. Some inconsistencies were found in patients with identical genotype. The first responsive case homozygous for the severe R408W mutation was found. Two new mutations, Y387X and G352C, were identified (the former was BH(4)-responsive), and the responsiveness of three already reported mutations (R261Q, D338Y, T92I) was substantiated.

Placental transfer of vigabatrin (gamma-vinyl GABA) and its effect on concentration of amino acids in the embryo of TO mice

BACKGROUND

The mechanism of the teratogenicity of vigabatrin (VGB) is unknown. The objectives of this study were to determine the placental transfer of VGB and to evaluate the effect of VGB on maternal, placental, and fetal concentrations of amino acids.

METHODS

A single dose of 400 mg/kg VGB in physiological saline was administered intraperitoneally to a group of Theiler outbred (TO) mice on gestational day (GD) 10. The controls received a proportionate volume of saline. Maternal blood samples, embryos, and placentas were collected at 3.5, 6.0, and 9.0 hr after treatment and their total amino acid concentrations determined in an ion-exchange amino acid analyzer.

RESULTS

At 3.5 hr, there was a decrease in concentrations of some amino acids in the blood, placenta, and embryos of VGB-treated mice, but the decrease in methionine was most marked. gamma-aminobutyric acid (GABA) was significantly higher in the VGB group in both the embryos and the placentas at 3.5 hr but at 6.0 and 9.0 hr the differences were not significant. Vigabatrin levels were higher in the placenta than in the embryo at 3.5 hr, but at 6.0 hr there was an overlap of the VGB peak with that of tryptophan with very much lower levels than at 3.5 hr. At 9.0 hr, there was no vigabatrin peak in either the placenta or the embryo.

CONCLUSIONS

Maternal exposure to VGB results in peak levels of the drug after 3.5 hr in the placenta and embryo. Methionine concentration is most severely affected in VGB-treated mothers, placentas, and fetuses. We speculate that this deficiency could be a possible mechanism for the teratogenic effects of vigabatrin.

Trace analysis in chiral separation of selected amino enantiomers

A Crownpack CR(+) column was used for the highly sensitive quantitation of amino acid enantiomers. The use of a mobile phase of exclusively aqueous perchloric acid enabled the use of low-wavelength UV detection (200 nm) before on-line derivatization with o-phthaldialdehyde for fluorimetric detection. The use of simultaneous UV and fluorimetric detection permitted the simultaneous quantification of the impurity and its parent compound. The method was used for the evaluation of some commercially available amino acid standards. Enantiomeric impurities as low as 0.001% (10 ppm) can be determined in some cases. High precision for the determination of trace levels of D-amino acids in the presence of large amounts of corresponding L-enantiomer is demonstrated.

Efficient enantioselective separation and determination of trace impurities in secondary amino acids (i.e., imino acids)

An R-(-)-1-(1-naphthyl)ethyl carbamoylated-beta-cyclodextrin bonded phase in conjunction with a nonaqueous polar mobile phase was used for the highly selective enantioseparation of a number of secondary amino acids after their pre-column derivatization with 9-fluorenylmethyl chloroformate (FMOC). Under the conditions employed, the FMOC reagent served to "lock" the imino acid into their existing conformation thereby preventing the possibility of racemization. Furthermore, it served to increase the sensitivity to the point that trace level enantiomeric impurities were easily detected. Compared with separations that use traditional reversed-phase solvents, this method showed several advantages: higher selectivity towards the imino acid enantiomers investigated, shorter analysis times, faster equilibration of the column, more stable baseline and more sensitive fluorescence detection. The detection limits for FMOC derivatives of proline, trans-4-hydroxyproline, cis-4-hydroxyproline, pyroglutamic acid, 3,4-dehydroproline, thiaproline, penicillamine acetone adduct and pipecolic acid are in the low femtomole range. The method was used for evaluation of enantioselectivity of a number of "optically pure" commercial imino acid standards. Enantiomeric impurities as low as 0.0001% (1 ppm) can be determined in some cases. High precision determination of trace levels of D-imino acids in the presence of large amounts of corresponding (opposite) L enantiomer at 1, 0.1, 0.01% and below are demonstrated.

Applications of automated amino acid analysis using 9-fluorenylmethyl chloroformate

A rapid and sensitive fully automated method for the determination of primary and secondary amino acids in different matrices is described. Amino acids are derivatized with 9-fluorenylmethyl chloroformate using an automated precolumn derivatization technique. Data are presented to show that the technique is both reproducible and highly sensitive. Applications of the technique are presented, including the analysis of peptide and protein hydrolysates and the profiling of free amino acids in physiological fluids.