Introduction
Phenylketonuria (PKU; OMIM 261600) is a genetic metabolic disorder, caused by insufficient activity of the hepatic enzyme phenylalanine hydroxylase (PAH). PAH converts the essential amino acid phenylalanine (Phe) to tyrosine (Tyr). Untreated patients with classical PKU demonstrate several cerebral complications including intellectual disability, seizures, and psychiatric symptoms. Neonatal screening and a Phe-restricted diet initiated directly afterwards, prevent most, but not all, of these complications (Blau et al. Reference Blau, van Spronsen and Levy2010).The neurocognitive outcome of early- and continuously treated PKU patients is generally within the normal range, although variations in outcome exist. Some patients may still suffer from a wide range of symptoms, including difficulties with social skills, lower intelligence quotient (IQ), impairments in executive functioning, and symptoms of attention-deficit/hyperactivity disorder (Huijbregts et al. Reference Huijbregts, de Sonneville, van Spronsen, Licht and Sergeant2002b , Reference Huijbregts, Gassio and Campistol2013; Waisbren et al. Reference Waisbren, Noel, Fahrbach, Cella, Frame, Dorenbaum and Levy2007; Antshel, Reference Antshel2010; Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010).
Shortages of brain dopamine and other monoamines like serotonin are thought to play a key role in the origin of neurocognitive sequelae in early-treated PKU (Surtees & Blau, Reference Surtees and Blau2000; van Spronsen et al. Reference van Spronsen, de Groot, Hoeksma, Reijngoud and van Rijn2010). Two mechanisms have been proposed for this theory (online Supplementary Fig. S1). First, high blood Phe levels competitively inhibit the transport of Tyr and tryptophan (Trp), amino acid precursors of catecholamines and serotonin, respectively, across the blood–brain barrier. Second, high brain Phe levels inhibit the activity of Tyr and Trp hydroxylases, the enzymes being responsible for the rate-limiting steps in catecholamine and serotonin synthesis (Pascucci et al. Reference Pascucci, Ventura, Puglisi-Allegra and Cabib2002; van Spronsen et al. Reference van Spronsen, Hoeksma and Reijngoud2009).
However, most evidence for monoamine deficiencies in PKU is indirect, relying mainly on: (1) measurements of monoamine (metabolite) levels in body fluids like urine or cerebrospinal fluid (e.g. Douglas et al. Reference Douglas, Jinnah, Bernhard and Singh2013), (2) postmortem data (McKean, Reference McKean1972), and (3) studies in rodents (e.g. Puglisi-Allegra et al. Reference Puglisi-Allegra, Cabib, Pascucci, Ventura, Cali and Romano2000; Joseph & Dyer, Reference Joseph and Dyer2003; van Vliet et al. Reference van Vliet, Bruinenberg, Mazzola, van Faassen, de Blaauw, Pascucci, Puglisi-Allegra, Kema, Heiner-Fokkema, van der Zee and van Spronsen2016; Winn et al. Reference Winn, Scherer, Thony and Harding2016). A limitation of these studies is that the degree and locus of monoaminergic abnormalities in the living human PKU brain remain unclear. Molecular imaging studies, however, allow the in vivo quantification of many aspects of neurotransmitter functions in the brain. Only two such studies have been performed in PKU. The findings of one positron emission tomography (PET) study in three PKU adults suggested an increased number of available striatal dopamine D2 receptors (Paans et al. Reference Paans, Pruim, Smit, Visser, Willemsen and Ullrich1996). Another PET study reported lower presynaptic striatal 6-[18F]fluoro-L-dopamine uptake in seven adults with PKU compared with seven healthy controls (Landvogt et al. Reference Landvogt, Mengel, Bartenstein, Buchholz, Schreckenberger, Siessmeier, Scheurich, Feldmann, Weglage, Cumming, Zepp and Ullrich2008). To date, no one has demonstrated, to our knowledge, a link between monoaminergic markers and the neurocognitive impairments in patients with PKU.
In the present study, we tested the hypothesis that adults with PKU have higher striatal dopamine D2/3 receptor (D2/3R) availability with [123I]iodobenzamide (IBZM) single-photon emission computed tomography (SPECT), as a proxy marker of a reduced brain synaptic dopamine concentration, and lower peripheral monoamine metabolite levels, relative to healthy controls. We also hypothesized that adults with PKU demonstrate relationships between D2/3R availability with brain and blood Phe levels, and with self-reported levels of impulsivity and neurocognitive performances.
Materials and methods
Participants
Eighteen patients with PKU (nine females; online Supplementary Table S1) were recruited through treating physicians. The mean (±s.d.) age was 31.2 ± 7.4 (range 18–42) years. Full Scale IQ (FSIQ) ranged from 88 to 113 (n = 17, one patient declined testing), as assessed with the Wechsler Adult Intelligence Scale, 3rd ed. Seventeen patients had been diagnosed by neonatal screening. One patient (patient 3, online Supplementary Table S1) was missed by neonatal screening and was identified at 3 years of age after her brother (patient 16) was diagnosed. Sixteen patients were continuously treated with a Phe-restricted diet since diagnosis with Tyr-enriched protein substitutes. Four patients (patients 3, 6, 7, and 16, online Supplementary Table S1) were also proven responsive to, and treated with, oral tetrahydrobiopterin (BH4), a pharmacological chaperone that promotes correct folding and stability of the PAH enzyme (Strisciuglio & Concolino, Reference Strisciuglio and Concolino2014), which has been found to lower blood Phe levels in a subset of PKU patients (Hegge et al. Reference Hegge, Horning, Peitz and Hegge2009). No patient received extra free Tyr supplementation. The mean levels on the study day for the 18 subjects were 706.8 ± 347.1 µmol/l for blood Phe and 42.2 ± 17.1 µmol/l for blood Tyr.
To compare the data obtained in PKU patients, we used historical data available for 12 healthy controls (seven females, aged 20–39 years with a mean of 27.0 ± 6.1 years) (Boot et al. Reference Boot, Booij, Zinkstok, Abeling, de Haan, Baas, Linszen and van Amelsvoort2008, Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). There were no significant between-group differences in gender (p = 0.72) or age (p = 0.12).
All assessments were conducted in the Academic Medical Center in Amsterdam, The Netherlands. Exclusion criteria for all participants included (i) a current or past psychiatric history, (ii) current or previous exposure to anti-psychotic or psychostimulant medication; (iii) a lifetime history of alcohol or substance abuse or dependence; (iv) a concomitant or past severe medical condition; (v) pregnancy; and (vi) iodine allergy. Additional exclusion criteria for PKU patients included (vii) lack of fluency in Dutch. The study protocol was approved by the Institutional Review Board, i.e. the Medical Ethics Committee of the Academic Medical Center of Amsterdam. Each participant gave written informed consent.
Study procedures
A schematic representation of the study procedures is shown in Fig. 1. All PKU patients were asked to complete a dietary record on the 2 days prior (D−2 and D−1) to the study day (D0) to evaluate daily consumption of natural protein and amino acid supplements, in particular with regard to the intake of large neutral amino acids (LNAAs). The PKU patients presented at 6:30 or 7:30 AM on the study day. Venous blood samples were drawn approximately 15–30 min after arrival (D0) for determination of plasma monoamine metabolite and blood amino acid levels. All blood samples were obtained after overnight fasting to eliminate prandial effects on amino acid concentrations. Blood sample collections were immediately followed by a proton magnetic resonance spectroscopy (1H MRS) scan to assess brain Phe levels. All participants then completed a selection of executive function tasks from the Amsterdam neuropsychological tasks (ANT) program (de Sonneville, Reference de Sonneville, den Brinker, Beek, Brand, Maarse and Mulder2009), completed the Barratt Impulsiveness Scale (BIS), version 11 (Patton et al. Reference Patton, Stanford and Barratt1995), and underwent an assessment of postsynaptic striatal D2/3R availability with [123I]IBZM SPECT.
Schematic representation of study procedures. In the patients with PKU, all the assessments, except for two of three capillary bloodspots (D−2, D−1), were performed on the same day. Data from healthy controls were available from historical records (indicated in upper three boxes). Text in italics represents assessments that provide information on the study population (data are provided in the online Supplementary material). *Indicates assessments that were conducted to test the study hypotheses. D−2, 2 days prior to the study day; D−1, 1 day prior to the study day; 1H MRS, proton magnetic resonance spectroscopy; ANT, the Amsterdam neuropsychological tasks program; IBZM SPECT, [123I]iodobenzamide single-photon emission computed tomography.
Previously published data from healthy controls (SPECT data, plasma monoamine metabolite levels, and self-reported levels of impulsivity) were available for comparison records (Boot et al. Reference Boot, Booij, Zinkstok, Abeling, de Haan, Baas, Linszen and van Amelsvoort2008, Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010).
Amino acid levels in blood and capillary bloodspots
To provide information on the study population, amino acid levels were assessed in all PKU patients. Blood levels of amino acids were measured with automated ion-exchange liquid chromatography followed by postcolumn derivatization with ninhydrin and photometric detection (JEOL JLC-500W Aminotac Amino Acid Analyzer, JEOL Ltd, Tokyo, Japan) (Moore et al. Reference Moore, Spackman and Stein1958). Fasting capillary bloodspot samples were obtained on 3 consecutive days (D−2, D−1, D0). Phe and Tyr levels were determined using tandem mass spectrometry (standard neutral loss method) (Rashed et al. Reference Rashed, Ozand, Bucknall and Little1995). Six bloodspot samples from four patients did not meet the quality criteria on the Quattro Premier XE (Waters, Milford, Massachusetts, USA) and were rejected by the laboratory.
Plasma monoamine metabolite levels
Fasting venous blood was used to obtain plasma levels of four monoamine metabolites. Homovanillic acid (pHVA; the predominant dopamine metabolite), vanillylmandelic acid (pVMA) and 3-methoxy-4-hydroxy-phenylglycol (pMHPG) levels were measured using reverse-phase high-performance liquid chromatography and coulometric electrochemical detection, with a modified method as previously described (Boot et al. Reference Boot, Booij, Zinkstok, Abeling, de Haan, Baas, Linszen and van Amelsvoort2008; Hartleb et al. 2013). 5-Hydroxyindoleacetic acid (p5HIAA; the predominant serotonin metabolite) levels were measured in 17 PKU patients. Patient 10 was excluded due to bromocriptine (dopamine D2 agonist) use that has been found to produce a significant effect on pHVA (Kendler et al. Reference Kendler, Heninger and Roth1982). Theoretically, flunarizine, a calcium channel blocker that may affect striatal D2/3R availability (Brücke et al. Reference Brücke, Wöber, Podreka, Wöber-Bingöl, Asenbaum, Aull, Wenger, Ilieva, Harasko-van der Meer, Wessely and Deecke1995), could potentially also influence plasma monoamine metabolite levels. However, to the best of our knowledge, this has not been reported. Therefore, we repeated the analyses of plasma monoamine metabolites after excluding the single patient taking flunarizine (patient 6).
Proton magnetic resonance spectroscopy
Fasting 1H MRS scans were obtained for 18 PKU patients to measure brain Phe levels. They were examined on a Philips Ingenia 3.0 Tesla MR system (Philips Medical Systems, Best, The Netherlands) equipped with a 16-channel head coil (Philips) using a point-resolved spectroscopy sequence to select a voxel of interest in the parietal white matter. Details are given in the online Supplementary material.
[123I]IBZM SPECT
We assessed striatal D2/3R availability with [123I]IBZM, using the validated equilibrium/constant infusion technique (Laruelle et al. Reference Laruelle, Abi-Dargham, van Dyck, Rosenblatt, Zea-Ponce, Zoghbi, Baldwin, Charney, Hoffer, Kung and Innis1995), and a brain-dedicated SPECT system (Neurofocus). The SPECT protocol was performed as described in our previous report (Boot et al. Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). To optimize image quality, PKU patients received a bolus of approximately 80 MBq instead of 56 MBq, followed by a bolus to hourly infusion ratio of approximately 4.0, identical to the abovementioned study. Fifteen PKU patients completed the SPECT protocol. Data of the 12 age-matched controls were published before (Boot et al. Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). All SPECT data were reconstructed and analyzed blind to clinical data by the same investigator (E.B.) as described in our previous report (Boot et al. Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). Two PKU patients (patients 6 and 10, online Supplementary Table S1) who received medication (flunarizine and the dopamine agonist bromocriptine, respectively) that influences D2/3R availability (Brücke et al. Reference Brücke, Wöber, Podreka, Wöber-Bingöl, Asenbaum, Aull, Wenger, Ilieva, Harasko-van der Meer, Wessely and Deecke1995; Lam, Reference Lam2012), and one other (patient 15) because of a technical failure of the infusion pump during imaging, were excluded from [123I]IBZM SPECT.
Impulsivity
The BIS-11, Dutch version 11 (Patton et al. Reference Patton, Stanford and Barratt1995), a validated 30-item self-report questionnaire, widely used as a measure of impulsivity, was administered to all participants on the study day.
Executive function performance
The ANT program (de Sonneville, Reference de Sonneville, den Brinker, Beek, Brand, Maarse and Mulder2009) was used to evaluate executive function performance in the PKU patients. The ANT is a computer-aided assessment battery that allows for the systematic evaluation of information processing capacities. It has been used successfully to determine cognitive deficit profiles for various clinical conditions, including PKU (Huijbregts et al. Reference Huijbregts, de Sonneville, Licht, van Spronsen and Sergeant2002a ). For this study, we selected three subtasks: (i) the Memory Search Two-Dimensions (MS2D) task, to assess working memory; and (ii) the Shifting Attentional Set – visual (SSV) task, to assess inhibitory control (subtask, part 2) and cognitive flexibility (subtask, part 3). More details on the subtasks can be found in the online supplementary material. One PKU participant (patient 9, online Supplementary Table S1) did not complete the MS2D task.
Statistical analysis
All statistical analyses were conducted using IBM SPSS Statistics 22 for Windows (SPSS Inc., Chicago, Illinois, USA). A probability value of <0.05 was selected as the level of significance for all tests. Kolmogorov–Smirnov tests were used to examine normality. Between-group differences for gender distribution were tested using a χ2 test. Between-group differences were tested using an independent-samples t test or Mann–Whitney U test as appropriate. Pearson correlation coefficients or Spearman's rank correlation coefficients were calculated with two-tailed tests of significance to investigate the relationships between biological markers, and between biological markers and neurocognitive performances and impulsivity levels, as appropriate. All data are presented as mean ± 1 s.d. unless indicated otherwise.
Results
Striatal D2/3 receptor availability
Consistent with our hypothesis, the striatal D2/3 receptor binding potential (D2/3R BPND), as a measure of striatal D2/3R availability, was 13% higher in adults with PKU (1.34 ± 0.19 v. 1.18 ± 0.17 for controls; p = 0.032; Fig. 2). This finding remained significant after excluding the PKU patients who received BH4 (PKU, n = 12, 1.33 ± 0.19; p = 0.048). There were no significant correlations between (brain or blood) Phe levels and striatal D2/3R BPND (data not shown). Excluding the patients who received BH4 did not change these results.
Increased mean striatal D2/3 receptor availability in patients with phenylketonuria (PKU). Horizontal lines indicate mean striatal D2/3 receptor binding potential (D2/3R BPND). Triangles mark the PKU patients who were treated with oral tetrahydrobiopterin. D2/3R BPND, as a proxy marker of brain synaptic dopamine concentration, was significantly higher in patients with PKU (n = 15) compared with healthy controls (HCs; n = 12), indicating more available dopamine receptors for the radiopharmaceutical [123I]iodobenzamide (IBZM), which binds selectively to dopamine D2/3 receptors.
Plasma monoamine metabolite levels
As expected, plasma HVA and pMHPG levels were significantly lower in PKU patients than in controls (41.4 ± 13.6 nmol/l v. 76.1 ± 24.8 and 10.7 ± 3.0 nmol/l v. 26.4 ± 9.3, respectively; Fig. 3a , b ). After excluding patient 6, who received flunarizine, a compound that may possibly influence peripheral monoamine levels, the differences remained significant for pHVA (p < 0.001) and pMHPG (p = 0.0001). After excluding the four patients who received oral BH4, the significance levels for pHVA and pMHPG remained unchanged (p < 0.001 and p < 0.0001, respectively). In contrast, plasma VMA levels were not significantly different between the two groups (Fig. 3c ). This result remained unchanged after excluding patient 6 and/or the patients who received BH4. The mean plasma 5HIAA level in the PKU group was 22.2 ± 7.7 nmol/l. The range (7.4–37.3 nmol/l) indicated that for all participants, levels were lower than the laboratory reference values (44.0–79.0 nmol/l; Fig. 3d ).
Plasma monoamine metabolite levels in 17 adults with phenylketonuria (PKU) and 12 healthy controls (HCs). (a) Mean plasma homovanillic acid levels (pHVA, nmol/l). (b) Mean plasma 3-methoxy-4-hydroxy-phenylglycol levels (pMHPG, nmol/l). (c) Mean plasma vanillylmandelic acid levels (pVMA, nmol/l). (d) Mean plasma 5-hydroxyindoleacetic acid levels (p5HIAA); p5HIAA was not assessed in the HCs. Laboratory reference values for p5HIAA are indicated with diagonal stripes. Error bars indicate ± 1 s.d.
Impulsivity
Self-reported impulsivity levels were significantly higher in PKU patients compared with healthy controls (60.2 ± 7.7 v. n = 12, 54.5 ± 5.1, p = 0.033; Fig. 4a ). D2/3R BPND correlated positively with impulsivity levels in the PKU patients (p = 0.003; Fig. 4b ) but not in the controls (p = 0.219). Given that later onset of dietary management could potentially influence impulsivity levels, we reran analyses after excluding patient 3. The between-group differences (p = 0.042) and correlation with D2/3R BPND (p = 0.002) remained significant. No significant correlations were found between (blood or brain) Phe levels and scores of impulsivity on the BIS-11 (data not shown).
Relationship between striatal D2/3 receptor availability [D2/3 receptor binding potential (D2/3R BPND)] and impulsivity. (a) Patients with phenylketonuria (PKU; n = 18) reported higher levels of impulsivity than healthy controls (HCs; n = 12). (b) In patients with PKU, D2/3R BPND correlated positively with impulsivity levels.
Executive function performance
D2/3R BPND also correlated positively with error rate during a cognitive flexibility task (SSV-subtask, part 3; Fig. 5a ). Results were not significant for error rate during a working memory task (MS2D-task; p = 0.075; Fig. 5b ) or performance on an inhibitory control task (Fig. 5c ). Excluding patient 3 did not materially change results (data not shown). There were no significant correlations between blood or brain Phe levels and executive function performance (data not shown).
Relationship between striatal D2/3 receptor availability and executive function performance in patients with phenylketonuria (PKU). (a) Relationship between striatal D2/3 receptor availability [D2/3 receptor binding potential (D2/3R BPND)] and error rate during the Shifting Attentional Set – visual task (SSV), part 3; a subtask of the Amsterdam neuropsychological tasks (ANT) program that requires cognitive flexibility (n = 15). (b) Relationship between D2/3R BPND and error rate during the Memory Search Two-Dimensions task (MS2D); a working memory task (n = 14). (c) Relationship between D2/3R BPND and error rate during the ANT-SSV, part 2, that requires inhibitory control (n = 15).
Discussion
A longstanding pathophysiological theory postulates that high blood Phe concentrations in PKU patients may lead to neurocognitive deficits by impairing brain dopamine and other monoamine synthesis (Pascucci et al. Reference Pascucci, Ventura, Puglisi-Allegra and Cabib2002; Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010). In support of this theory, the present study shows, for the first time, significantly higher striatal D2/3R availability in adult PKU patients in comparison with control participants, suggesting that there may be reduced concentrations of dopamine in the synapse. These findings can be explained by the dopamine receptor competition model that predicts that a lower dopamine concentration in the synapse will lead to a lower occupancy of dopamine D2/3R (and possibly also a compensatory upregulation of the presence of dopamine D2/3R) and consequently higher binding of the radiotracer [123I]IBZM to dopamine D2/3R (see discussion in Boot et al. Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). The present study complements the two previous molecular imaging studies in PKU that suggested reduced dopamine synthesis (Paans et al. Reference Paans, Pruim, Smit, Visser, Willemsen and Ullrich1996; Landvogt et al. Reference Landvogt, Mengel, Bartenstein, Buchholz, Schreckenberger, Siessmeier, Scheurich, Feldmann, Weglage, Cumming, Zepp and Ullrich2008), by providing further support for the assumption of dopamine deficiencies at the level of the synapse of these treated patients.
Results of our study found no significant correlations between blood or brain Phe levels and striatal D2/3R BPND. Several possible mechanisms may have contributed. In addition to power issues, these include between-subject variability in D2/3R availability (Kegeles et al. Reference Kegeles, Zea-Ponce, Abi-Dargham, Rodenhiser, Wang, Weiss and van Heertum1999), challenging aspects of measuring in vivo brain Phe levels (Kreis et al. Reference Kreis, Zwygart, Boesch and Nuoffer2009), and blood concentrations of other LNAAs competing with Tyr and Phe for transport into brain (van Vliet et al. Reference van Vliet, Bruinenberg, Mazzola, van Faassen, de Blaauw, Kema, Heiner-Fokkema, van Anholt, van der Zee and van Spronsen2015b ). Reduced availability of monoaminergic precursors in the brain may be an even greater limiting factor in monoamine biosynthesis than increased brain Phe levels (van Vliet et al. Reference van Vliet, Bruinenberg, Mazzola, van Faassen, de Blaauw, Kema, Heiner-Fokkema, van Anholt, van der Zee and van Spronsen2015b ; van Vliet et al. Reference van Vliet, Bruinenberg, Mazzola, van Faassen, de Blaauw, Pascucci, Puglisi-Allegra, Kema, Heiner-Fokkema, van der Zee and van Spronsen2016) (online Supplementary Fig. S1).
Our results add to previous human studies of PKU that assessed monoamine (metabolite) levels in body fluids such as urine, and found lower levels compared with controls (e.g. Nadler & Hsia, Reference Nadler and Hsia1961; Douglas et al. Reference Douglas, Jinnah, Bernhard and Singh2013). In the present study, PKU patients showed lower levels of pHVA and pMHPG, and p5HIAA levels below the lower limit of the reference range, suggesting reduced synthesis of dopamine, norepinephrine, and serotonin, respectively. It however remains unclear to what extent these abnormalities may reflect brain monoaminergic alterations (Elsworth et al. Reference Elsworth, Leahy, Roth and Redmond1987; Pickar et al. Reference Pickar, Breier and Kelsoe1988).
This study is the first to demonstrate a relationship between a neurochemical monoaminergic brain marker and measures of executive functioning in adult PKU patients, most of whom were treated early and continuously. Previous studies have found that in early treated adult PKU patient, executive functioning is particularly affected (Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010). The prefrontal cortex and striatum play a crucial role in executive functioning, and cognitive processes are believed to be largely modulated by monoamines, including dopamine (Cropley et al. Reference Cropley, Fujita, Innis and Nathan2006; Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010). Consistent with the potential importance of this in PKU is the finding from the present study that there was a significant positive correlation between striatal D2/3R availability and error rate on a cognitive flexibility task in the adults with PKU. The correlation between D2/3R BPND and error rate during a working memory task did not however reach statistical significance (p = 0.075). Performance on the inhibitory control task was also not correlated with D2/3R BPND, although it is possible that ceiling effects associated with this relatively easy task could have hampered the possibility of finding a correlation. Notably, it is possible that associations between neuropsychological test performance and D2/3R BPND could be accounted for by differences in general attention performance. In this study, different from other studies (e.g. Jahja et al. Reference Jahja, Huijbregts, de Sonneville, van der Meere and van Spronsen2014), a correlation with executive functioning was not observed with (blood or brain) Phe levels, nor with other biochemical markers assessed. We note that Ullrich et al. did not find an effect of levodopa treatment on visual-evoked potentials or neuropsychological test performance in adults with PKU (Ullrich et al. Reference Ullrich, Weglage, Oberwittler, Pietsch, Funders, van Eckhardstein and Colombo1994, Reference Ullrich, Weglage, Oberwittler, Pietsch, Funders, van Eckhardstein and Colombo1996).
Higher levels of impulsivity in patients with PKU than in healthy controls, consistent with previous reports (Hendrikx et al. Reference Hendrikx, van der Schot, Slijper, Huisman and Kalverboer1994; Stemerdink et al. Reference Stemerdink, Kalverboer, van der Meere, van der Molen, Huisman, de Jong, Slijper, Verkerk and van Spronsen2000), and the strong positive correlation of self-reported levels of impulsivity with striatal D2/3R BPND within the PKU group, are also in line with a possible ‘hypodopaminergic state’ in PKU. Here, it would be surmised that patients with a higher striatal dopamine D2/3 receptor (D2/3R) availability may have a more pronounced shortage of synaptic dopamine that could then lead to increased impulsivity. Impulsivity is believed to be mediated by dopaminergic and serotonergic activity (Dalley & Roiser, Reference Dalley and Roiser2012) and to be directly related to dysfunctional inhibitory processes (Bari & Robbins, Reference Bari and Robbins2013).
While previous studies in PKU patients have also shown reduced peripheral monoamine (metabolite) levels and negative correlations between blood Phe levels and peripheral monoamine (metabolite) levels (Nadler & Hsia, Reference Nadler and Hsia1961; Douglas et al. Reference Douglas, Jinnah, Bernhard and Singh2013), their correlation with availability of monoamines in the brain, and their relationship with neurocognitive performance, remain unclear. Hence, it is uncertain to what extent these peripheral markers could be of value for monitoring brain pathophysiology. This is also the case for plasma prolactin, a neuroendocrine marker for cerebral dopamine (van Vliet et al. Reference van Vliet, Anjema, Jahja, de Groot, Liemburg, Heiner-Fokkema, van der Zee, Derks, Kema and van Spronsen2015a ; Juhász et al. Reference Juhász, Kiss, Simonova, Patocs and Reismann2016). In this regard, it is of interest that in the present study, no significant correlations were found between D2/3R BPND and plasma monoamine levels.
The results of the present study may have implications for the monitoring of PKU patients that has thus far focused primarily on reducing blood Phe levels (Weglage et al. Reference Weglage, Wiedermann, Denecke, Feldmann, Koch, Ullrich, Harms and Möller2001). While blood Phe levels can relatively easily be monitored, and blood Phe is the major marker for dietary adherence, its utility as a biomarker for brain pathophysiology and clinical monitoring appears suboptimal. A significant correlation between blood and brain Phe levels, as found in this and other studies (Pietz et al. Reference Pietz, Kreis, Boesch, Penzien, Rating and Herschkowitz1995; Leuzzi et al. Reference Leuzzi, Tosetti, Montanaro, Carducci, Artiola, Carducci, Antonozzi, Burroni, Carnevale, Chiarotti, Popolizio, Giannatempo, D’Alesio and Scarabino2007; Kreis et al. Reference Kreis, Zwygart, Boesch and Nuoffer2009), is not supported by other studies (Weglage et al. Reference Weglage, Wiedermann, Denecke, Feldmann, Koch, Ullrich, Harms and Möller2001; Moats et al. Reference Moats, Moseley, Koch and Nelson2003; Sijens et al. Reference Sijens, Oudkerk, Reijngoud, Leenders, de Valk and van Spronsen2004). A linear relationship between blood and brain Phe levels has also been deemed unlikely under the assumption of a saturable transport mechanism (Weglage et al. Reference Weglage, Wiedermann, Denecke, Feldmann, Koch, Ullrich, Harms and Möller2001) and there have been anecdotal reports of PKU patients achieving normal intellect with low brain Phe levels, despite elevated blood Phe levels (Weglage et al. Reference Weglage, Moller, Wiedermann, Cipcic-Schmidt, Zschocke and Ullrich1998, Reference Weglage, Wiedermann, Denecke, Feldmann, Koch, Ullrich, Harms and Möller2001; Koch et al. Reference Koch, Moats, Guttler, Guldberg and Nelson2000). In addition, findings concerning relationships between blood Phe and neurocognitive performance are mixed (for an overview see Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010). In the present study, neither blood nor brain Phe levels correlated significantly with any of the clinical outcome measures. Given the finding that striatal dopamine D2/3R availability, however, was significantly correlated with measures of executive functioning in this study, we propose that brain monoaminergic markers may be potential biomarkers of long-term cognitive outcome in PKU (van Vliet et al. Reference van Vliet, Bruinenberg, Mazzola, van Faassen, de Blaauw, Kema, Heiner-Fokkema, van Anholt, van der Zee and van Spronsen2015b ). In this context, a parabolic (‘inverted U’) relationship between cerebral dopamine concentrations and neurocognitive performance (Cools & D'Esposito, Reference Cools and D’Esposito2011), and between striatal dopamine D2/3R availability and impulsivity (Gjedde et al. Reference Gjedde, Kumakura, Cumming, Linnet and Møller2010), should be considered.
Methodological considerations
Neurochemical studies provide the most direct method to assess monoamine systems in the living human brain currently available. In the present study, we performed [123I]IBZM SPECT, a well-validated reproducible modality to assess striatal D2/3R availability that has been used successfully for several disorders by our group (e.g. Boot et al. Reference Boot, Booij, Zinkstok, de Haan, Linszen, Baas and van Amelsvoort2010). We used a bolus/infusion technique; this approach is not sensitive to changes in cerebral blood flow and measures D2/3R availability in a state of equilibrium (Booij et al. Reference Booij, Korn, Linszen and van Royen1997). Nevertheless, differences in neuroreceptor density in PKU patients and healthy controls cannot be ruled out with this modality and may have influenced the results observed. In addition, [123I]IBZM SPECT is not suited to examining the dopamine system in extrastriatal regions. It would be of interest to examine other brain areas in PKU, in particular the prefrontal cortex, given its crucial role in executive functioning and its dependence on dopamine activity, with other PET or SPECT radioligands (Cropley et al. Reference Cropley, Fujita, Innis and Nathan2006; Barnes et al. Reference Barnes, Dean, Nandam, O’Connell and Bellgrove2011). The caudate nucleus has widespread connections with the prefrontal cortex. Future study design could thus be improved by using magnetic resonance imaging to co-register the SPECT images, which would permit delineation of [123I]IBZM binding in the caudate nucleus from that in the putamen (Cropley et al. Reference Cropley, Fujita, Innis and Nathan2006). Evaluating the left and right caudate nucleus separately would also allow assessment of lateralization of D2/3R availability in relation to neurocognitive deficits. Future studies investigating absolute quantification of brain Phe concentrations using 1H MRS (Kreis et al. Reference Kreis, Zwygart, Boesch and Nuoffer2009), and those studying other neurotransmitters in human PKU brain in vivo, in particular serotonin (Pascucci et al. Reference Pascucci, Ventura, Puglisi-Allegra and Cabib2002; van Spronsen et al. Reference van Spronsen, Hoeksma and Reijngoud2009; Christ et al. Reference Christ, Huijbregts, de Sonneville and White2010), and neurotransmitter system interactions, are also needed.
As in another Dutch study (Schmaal et al. Reference Schmaal, Veltman, Nederveen, van den Brink and Goudriaan2012), in the present study, the levels of self-reported impulsivity in healthy controls were relatively low (and the levels in patients with PKU were similar) to those previously reported for two US community samples (Patton et al. Reference Patton, Stanford and Barratt1995; Reise et al. Reference Reise, Moore, Sabb, Brown and London2013). This raises the possibility of cultural differences and/or differences due to translation of the questionnaire. Considerations for future studies with respect to assessment of impulsivity include task-based tests instead of self-reported impulsivity scores, and accounting for the clinical psychiatric state of the patient since impulsivity ratings may be state dependent (e.g. Corruble et al. Reference Corruble, Benyamina, Bayle, Falissard and Hardy2003).
Extraneous factors may have influenced the outcome of the assessments. For example, one patient took his Phe-free amino acid mixture 2 h prior to the blood sample (patient 11). Theoretically, this could have led to increased monoamine synthesis, and consequently higher monoamine metabolite levels in this patient. Still, pHVA, pMHPG, and p5HIAA levels in this patient were all lower than the mean levels in the PKU group. One patient had also stopped smoking 17 days prior to the study (patient 7), and one patient smoked approximately five cigarettes per day (patient 8). Although cerebral dopamine is released when smoking, these two patients had D2/3R BPND values close to the mean (1.33 and 1.36 v. 1.34) in the PKU group. This is consistent with one study reporting that there is no effect of smoking on striatal D2/3R availability over the long term (Yang et al. Reference Yang, Yao, McEvoy, Chu, Lee, Chen, Yeh and Chiu2006).
Some data collected in the PKU patients were not available for the control participants, e.g. p5HIAA levels. However, the fact that none of the PKU patients reached a p5HIAA level above the lower limit of the reference range makes the null hypothesis (no between-group differences in p5HIAA levels) unlikely. Because formal executive functioning was not assessed in controls, and in the absence of normative data for the ANT, it is unclear to what extent the executive deficits present in our PKU cohort would be clinically relevant. It should be noted that all patients had an FSIQ score within the normal range. Also, blood samples for plasma monoamine metabolite levels were not necessarily drawn in the fasted state in the controls. Although this could have influenced the results for between-group differences in plasma monoamine metabolite levels (Doran et al. Reference Doran, Labarca, Wolkowitz, Roy, Douillet and Pickar1990), the other findings would not be affected.
There was no between-group difference in sex distribution; however, we cannot rule out an effect of sex on our findings. Males and females are equally likely to be affected with PKU. However, one could speculate that sexually dimorphic characteristics of the monoaminergic system of the brain (e.g. Harrison and Tunbridge, Reference Harrison and Tunbridge2008) could produce differing effects on executive functioning in men and women with PKU.
The present study did not evaluate the effect of any dietary intervention. A challenge study with equal assessments in a diet v. placebo condition would allow testing the hypothesis that dietary factors may affect cerebral monoamine synthesis and availability in PKU, and consequently neurocognitive functioning (e.g. Mehta et al. Reference Mehta, Gumaste, Montgomery, McTavish and Grasby2005).
Although the present study is to date the largest involving neurochemical imaging techniques in PKU, the sample sizes were small and the number of outcome variables relatively large. On the one hand, the results should therefore be interpreted with caution and this work needs to be replicated in an independent larger cohort of PKU patients. On the other hand, the range of blood Phe levels, striatal D2/3R binding, and other variables enabled the performance of correlational analyses within the PKU group. Future studies, including a prospective design, are required to elucidate the effects of aging on monoamine systems in PKU.
Conclusions
This study provides further support for the hypothesis of cerebral dopamine deficiency as a possible pathophysiological explanation of executive functioning deficits in early and continuously treated adult PKU.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0033291717001398.
Acknowledgements
This work was financially supported by the Dutch Brain Foundation (E.B., grant number KS2012(1)-31); the Brain & Behavior Research Foundation (T.A.M.J.v.A., formerly NARSAD); the Dutch PKU Research Foundation (F.J.v.S.); NutsOhra Fund (F.J.v.S., grant number 1003–061); and the Division of Metabolic Diseases of the Beatrix Children's Hospital of the University Medical Centre Groningen, The Netherlands. The funding agencies had no role in study design, in the collection, analysis and interpretation of data, in the writing of the report, or in the decision to submit the paper for publication. The authors are grateful to all study participants. They thank the staff of the Department of Nuclear Medicine, including Bastiaan Kee for excellent technical assistance and Corry Stoelman-Andreas for administrative support, Raschel van Luijk-Snoeks and Sandra van den Berg for conducting the 1H MRS scans, Lydia Veerhuis for administrative support, and Corrie Timmer and Cora Jonkers for evaluating the patient's dietary records. Parts of this study have been presented as an oral presentation at the 19th International Research Symposium of the Society for the Study of Behavioural Phenotypes (SSBP) in Siena, Italy in September 2016 and as posters at the annual symposium of the Society for the Study of Inborn Errors of Metabolism (SSIEM) in Rome, Italy in September 2016, the 23rd Annual Meeting & Exhibition of the International Society for Magnetic Resonance in Medicine (ISMRM), Toronto, Ontario, Canada in June 2015, and the Society of Biological Psychiatry's (SOBP) 70th annual meeting in Toronto in May 2015.
Declaration of Interest
None.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
