- Email: [email protected]
Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria
Accepted Manuscript Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria
Nastassja Himmelreich, Nan Shen, Jürgen G. Okun, Christian Thiel, Georg F. Hoffmann, Nenad Blau PII: DOI: Reference:
S1096-7192(18)30283-X doi:10.1016/j.ymgme.2018.06.011 YMGME 6369
To appear in:
Molecular Genetics and Metabolism
Received date: Revised date: Accepted date:
25 May 2018 21 June 2018 22 June 2018
Please cite this article as: Nastassja Himmelreich, Nan Shen, Jürgen G. Okun, Christian Thiel, Georg F. Hoffmann, Nenad Blau , Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria. Ymgme (2018), doi:10.1016/j.ymgme.2018.06.011
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria
Nastassja Himmelreich1, Nan Shen1,2, Jürgen G Okun1, Christian Thiel1, Georg F Hoffmann1, Nenad Blau1,* 1
PT
Center for Child and Adolescent Medicine and Dietmar-Hopp Metabolic Center,
University of Heidelberg, Heidelberg, Germany; 2Department of Rehabilitation
RI
Medicine, Xin Hua Hospital affiliated to Shanghai Jiao Tong University School of
SC
Medicine, Shanghai, China
NU
*Correspondence:
Dietmar-Hopp Metabolic Center
MA
Prof. Dr. Nenad Blau
University Children’s Hospital Heidelberg
ED
Im Neuenheimerfeld 669 69120 Heidelberg
EP T
Germany
AC C
[email protected]
ACCEPTED MANUSCRIPT ABSTRACT Residual phenylalanine hydroxylase (PAH) activity is the main determinant of the metabolic phenotype in phenylketonuria (PKU). The genotypic heterogeneity of PKU, involving more than 1,000 PAH variants and over 2,500 different genotypes, makes genotype-based phenotype prediction challenging. While a relationship between PAH variants and the metabolic phenotype is well established, we questioned the
PT
importance of PAH expression and residual in vitro activity for the metabolic phenotype.
RI
Thirty-four PAH variants (p.F39L, p.A47V, p.D59Y, p.I65S, p.R68G, p.R68S, p.E76G,
SC
p.A104D, p.D143G, p.R155H, p.R176L, p.V190A, p.G218V, p.R241C, p.R243Q, p.P244L, p.R252W, p.R261Q, p.E280K, p.R297H, p.A300S, p.I306V, p.A309V,
NU
p.L311P, p.A313T, p.L348V, p.V388M, A403V, p.R408Q, p.R408W, p.R413P, p.D415N, p.Y417H, and p.A434D) were transiently transfected into COS-7 cells, and expression of PAH was investigated. Expression patterns were compared with in vitro
MA
PAH activity and allelic phenotype values (APVs).
In vitro PAH activity was significantly higher (p<0.01) in variants associated with mild
ED
hyperphenylalaninemia (PAH activity = 52.1 ± 8.5%; APV = 6.7 - 10.0) than that in classic PKU variants (PAH activity = 21.1 ± 7.0%; APV = 0 - 2.7). Mild PKU variants
EP T
(PAH activity = 40.2 ± 7.6%; APV = 2.8 - 6.6) were not significantly different from mild hyperphenylalaninemia, but there was a difference (p<0.048) compared with classic
AC C
PKU phenotypes.
ACCEPTED MANUSCRIPT INTRODUCTION Phenylketonuria (PKU) is phenotypically a heterogeneous disease, ranging from mild hyperphenylalaninemia (MHP; blood phenylalanine (Phe) 120 - 600 mol/L) to mild PKU (mPKU; blood Phe 600 - 1,200 mol/L) and classic PKU (cPKU; with blood Phe levels >1,200 mol/L) [1]. From a clinical perspective, a finer classification of MHP may have some significance for outcome and treatment of MHP patients. The recent
PT
US guidelines [2] and the NIH PKU scientific review conference [3] splitted MHP into MHP not requiring treatment (blood Phe 120 – 600 mol/L) and MHP-gray zone (360
RI
– 600 mol/L). MHP-grey zone patients may require treatment in some countries,
SC
while MHP patients may not. The metabolic phenotype in PKU is a consequence of loss of phenylalanine hydroxylase (PAH) function, caused by more than 1,000 PAH variants and over 2,500 different genotypes, resulting in hyperphenylalaninemia
NU
(HPA) [4]. Severe cPKU results in almost no or very little residual PAH activity, while in MHP patients substantial residual activity is reported in vivo and in vitro. Data from
MA
the locus-specific PAHvdb database (www.biopku.org/pah) demonstrate that 65% of the variants involved are amino acid substitutions, including changes in nonsense codons, and other common variants, such as deletions (15%), splice variants (12%),
ED
insertions/duplications/indels (4%) and synonymous variants (4%); thus, PKU is a
EP T
highly heterogeneous disease [4,5].
Extensive attempts to correlate genotypes with PKU phenotypes were reported in the past, which achieved results ranging from incomplete or no correlations to substantial
AC C
correlations between PAH variants and corresponding metabolic phenotypes [6,7]. A recent definition of the allelic phenotype value (APV) algorithm assigned PAH variants occurring in functionally hemizygous constellations to one of the three PKU phenotypes. APV-based phenotype prediction was found to be 99% correct for cPKU, 46% for mPKU and 90% for MHP within a cohort of more than 9,000 PKU patients [8]. Correlations between genotypes and in vitro residual PAH activity have been documented for several PAH variants [9-11]. In addition to PAH protein stability and APV, enzyme activity has also been used to predict patients phenotypes [12]. In contrast to APV, it has been shown that calculated enzyme activities based on heterogeneous data from single mutations do not reliably reflect residual activity in the patient [13] and that the two alleles may affect each other, a phenomenon known
ACCEPTED MANUSCRIPT as interallelic complementation [14,15]. In addition, the methods used for the measurement of PAH expression and activity measurement differ greatly, resulting in a broad range of activities for the same variant protein [10,16]. To better understand the relationship between PAH protein expression and enzyme activity, we transiently expressed 34 common PAH variants in COS-7 cells, investigated PAH expression and measured in vitro residual PAH activity under
PT
standardized conditions. Results were then compared with APV data. In addition, we conducted a literature search and compared different expression systems and PAH
SC
RI
activities reported.
MATERIALS AND METHODS
NU
Cloning
The cloning of PAH variants was performed as described previously [15]. In brief, the
MA
expression plasmid pCMV-FLAG-PAH (Promoter-N-Fusion-PAH) was received as a courtesy gift from L. R. Desviat [17]. Mutations in the human PAH-cDNA sequence
ED
were introduced via site-directed mutagenesis using the QuikChange II XL kit from Agilent Technologies (Santa Clara, CA, USA) and confirmed thgrough DNA
EP T
sequencing analysis at SEQLAB Sequence Laboratories (Göttingen, Germany). Transient transfections
For the transient PAH expression COS-7 monkey kidney cells were used. Standard
AC C
culture conditions of 37°C and 5% CO2 in a sterile environment were used. Cells were washed with 1 x PBS every 72 h, and fresh Dulbecco’s modified Eagle’s medium (DMEM), enriched with 10% fetal calf serum (FCS) plus 1% Pen/Strep was added. Then, 106 COS-7 cells plated on 10 cm dishes and were transfected with 10 µg of plasmid DNA and 1 µg of the pSV-β-galactosidase control vector (Promega, Madison, USA) using Fugene HD transfections reagents (Promega, Madison, USA) for 72 h. Finally, the cells were harvested with trypsin and washed 2-times with icecold PBS, then flash frozen in liquid nitrogen for storage at −80°C. Each variant was analyzed in three independent transfection experiments.
ACCEPTED MANUSCRIPT Western blotting Cell lysates were prepared by macerating the cells in 1 M Sucrose, 1 x PBS and 1 x PIM buffer 30 times through a 20G cannula, followed by 3 x 10 sec sonification and centrifugation (30 min/13,000rpm/ 4°C). The high salt content was removed from the supernatant using Zeba Desalting Spin Columns (Pierce Biotechnology, Waltham, MA). For Western blot analyses 20 µg of the sample supernatant proteins was
PT
unstitched through 12% SDS-PAGE and transferred via semi-dry electrophoretic blotting onto a nitrocellulose membrane (GE Healthcare, CT). The membrane was
RI
then blocked for 1 h at room temperature in 5% milk powder in either PBS-T (phosphate-buffered saline with 0.1% Tween) or TBS-T (0.1% Tween in Tris Buffered
SC
Saline). Primary antibodies against PAH (mouse α human, 1:1,000 dilution, Merck Millipore, MA) and ß-actin (mouse α human, 1:10,000, Sigma-Aldrich, MO) were
NU
used, and the samples were incubated overnight at 4°C under constant movement. The secondary species-specific antibody conjugated with a horseradish peroxidase
MA
tag facilitated detection via a chemiluminescent reaction with the ECL reagent (Pierce Biotechnology, Waltham, MA). Quantification of Western blots was performed by ImageJ (https://imagej.nih.gov/ij/) as decribed previously [15]. Data points represent
EP T
PAH activity assay
ED
means of n=3 independent transfection experiments.
In vitro PAH activity measurements were performed for each transfection as described previously using the isotope-dilution liquid chromatography-electrospray
AC C
ionization tandem mass spectrometry technique as triplicates [18]. Allelic phenotype values (APV) The APV value was calculated for PAH variants expressed in COS cells. The APV is based on the frequencies of the metabolic phenotype (i.e., cPKU, mPKU or MHP) for genotypes presenting in a functionally hemizygous state and calculated using the following formula: APV = (%cPKU*0 + %mPKU*5 + %MHP*10)/100 where % indicates the percentage of phenotypes for a given functionally hemizygous genotype [8]. APVs are dynamically calculated from the actual BIOPKU dataset (n=10,340; as of May 2018) and stored in the locus-specific PAHvdb database.
ACCEPTED MANUSCRIPT Searching for a specific genotype in BIOPKU generates a report of the corresponding APV and the predicted phenotype, depending on whether the APVs of both alleles are known (http://www.biopku.org). Prediction of the damaging effects of missense variants The PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2) and SIFT Blink
PT
(http://sift.jcvi.org) algorithms were used to predict the effect of variants on protein. Statistics
RI
All experiments were repeated at least three times. Prism 7 (GraphPad Software Inc.,
NU
SC
CA, USA) was used for descriptive statistics and the unpaired Student’s t-test.
RESULTS
MA
Comparison of in vitro PAH activity in different expression systems We collated the in vitro enzyme activity and protein expression of 87 PAH variants reported in 49 publications (including this one) (Table 1). Most of the variants (51%)
ED
were transiently expressed in monkey kidney COS cells, in E. coli (24%), using the T7-based coupled transcription/translation system (24%) or in human embryonic
EP T
kidney 293 cells (1%). The hepatocytes HepG2 cells, yeast and PRO bacterial expression system were mainly used for PAH protein pattern analyses. We compared the average reported PAH activities determined using the three commonly
AC C
employed expression systems (see above) with the phenotypes (based on APV, when occurring in the functionally hemizygous constellation) of the 73 PAH variants (Figure 1). Based on PAH activity, the COS cells system appeared to be the best system for differentiating between cPKU and mPKU variants (p<0001), while those associated with mPKU and MHP overlapped (p<0.2). E. coli was favored when comparing mPKU with MHP variants (p<0.03) and the cell free TNT-T7 system showed a significant difference (p<0.03) only for cPKU and mPKU. For the most severe cPKU variants with no residual PAH activity, all three systems were comparable (data not shown).
ACCEPTED MANUSCRIPT Phenotype prediction for PAH missense variants Among the 34 variants investigated in this study, 12 were predicted to be associated with cPKU (APV: 0 - 2.7), 8 with mPKU (APV: 2.8 - 6.6) and 10 with the MHP (APV: 6.7 – 10.0) phenotype (Table 2). For four variants (p.I65S, p.D59Y, p.P244L and p.Y417H), no information was provided in the BIOPKU, and APV was not assigned. p.R408W was the most common variant reported in in combination with a severe
PT
null-variant (functional hemizygote) (1,845 patients), followed by p.R261Q (489 patients) and p.R243Q (405 patients) (Table 2). With exception of p.D143G and
RI
p.A313T (2 patients and 1 patient, respectively), all other variants were reported in at
PAH in vitro activity and PKU phenotypes
SC
least four cases.
NU
The PAH activity of 13 variants was investigated for the first time. These included four cPKU variants (p.F39L, p.I65S, p.L311P and p. A434D), four mPKU variants
MA
(p.A104D, p.V190A, p.A313T and p.Y417H) and 5 MHP variants (p.A47V, p.R155H, p.R297H, p.I306V and p.D415N). In vitro PAH activity was only clearly correlated with the predicted phenotype for three variants. The cPKU variants p.L311P and p.
ED
A434D presented very low PAH activity (0% and 9%, respectively), while the MHP variant p.A47V showed 123% activity. For all other variants, PAH activity was
EP T
between 21% and 77% of residual activity not clearly pointing to a phenotype. This is in the line with previous reports (Table 1). The average PAH activity of 10 MHP variants (p.A47V(APV10.0), p.D415N(APV10.0),
AC C
p.R297H(APV9.8), p.R176L(APV9.5), p.I306V(APV9.5), A403V(APV9.2), p.A300S(APV8.9), p.R155H(APV8.8), p.E76G(APV7.1), p.V190A(APV7.1)) was 44 (29)% of the wild-type activity. The highest activity (123%) was determined for the p.A47V variant and the lowest (21%) for the p.R155H variant. The 7 mPKU variants (p.R241C (APV6.0), p.R68S(APV5.5), p.R408Q(APV5.5), p.R68G(APV5.0), p.A313T(APV5.0), p.A309V(APV3.3), p.A104D (APV2.8)) presented with lower average PAH activity 39 (28)%. The p.D143G(APV5.0) variant, with 98% activity and a 40% standard error in three independent measurements, was considered to be unreliable and was excluded from the calculations.
ACCEPTED MANUSCRIPT Finally, for the 12 cPKU variants (p.A434D (APV2.6), p.G218V(APV2.5), p.V388M(APV1.9), p.L348V(APV1.7), p.F39L(APV1.4), p.R261Q(APV1.3), p.R243Q(APV0.4), p.R413P(APV0.1), p.E280K(APV0.04), p.R252W(APV0.04), p.R408W(APV0.03), p.L311P(APV0)), the average PAH activity was 25 (25)%, which was significantly lower than in the MHP and mPKU groups. Four PAH variants without information in the BIOPKU database, thus lacking APV
PT
information, were tentatively classified, based on protein expression and PAH activity, as cPKU (p.Y417H and p.D59Y) and as mPKU or MHP (p.I65S and p.P244L)
RI
(Table 2).
SC
Enzyme activity and PAH expression pattern
The PAH expression pattern and enzyme activity of 34 variants is summarized in
NU
Figure 2. Western blot analysis of expressed proteins revealed three characteristic bands (1, 2 and 3). While band 1 corresponds to the wild-type PAH, bands 2 and 3
further investigated in this study.
MA
represent degradation products of PAH. The nature of the three PAH bands was not
ED
We quantified the total protein expression of PAH and its degradation pattern (bands 1 - 3) and compared the results with in vitro enzyme activity of each of the phenotype groups. The average PAH expression (compared with wild-type PAH) was
EP T
significantly lower (p<0.004) for cPKU variants (35.6%), compared with mPKU variants (93.0%) and MHP variants (82.6%) (Figure 3A). Similarly, the average PAH activity was lower (p<0.048) for cPKU variants (19.2%), compared with mPKU
AC C
variants (40.1%) and MPH variants (46.6%) (Figure 3B). There was no significant difference in protein expression or enzyme activity between mPKU and MHP (Table 2 and Figures 3A and B). Almost no protein expression (<5%) was observed for four cPKU variants: p.L348V, p.R243Q, p.E280K and p.R252W, while for three other cPKU variants: p.R408W and p.L311P, p.R413P, expression was only slightly higher (12%, 13%, and 17% respectively) (Table 2). For the mPKU and MHP variants, PAH expression was always higher than 33%. The degradation pattern was evident for all variants, regardless of the total expression and enzyme activity (Figure 2, Table 2). The ratio of bands 1, 2 and 3 was approximately 8:1:1 for the wild-type PAH (Figure 2) and remarkably different for the p.D59Y (4:2:4), p.I65S (5:4:1), p.R68G (4:3:3), p.R241C (3:5:2), p.I306V (7:3:0),
ACCEPTED MANUSCRIPT p.A403V (6:2:2) and p.Y417H (1:9:0) variants. For the p.P244L, p.P39L and p.A104D variants, the ratio was 7:1:2 (Figure 2, Table 2). Prediction of the damaging effect of variants Two web-based tools for the prediction of variant severity, SIFT and Polyphen-2, were used. Both algorithms predicted two MHP variants (p.D415N, p.E76G) and one mPKU variant (p.A104D) to be benign, while there was no match between the two
SC
RI
possibly damaging, damaging, or deleterious (Table 3).
PT
algorithms for the p.R176L and p.I306V variants. All other variants were predicted as
DISCUSSION
NU
Different approaches can be used to predict the effect of gene variants on the metabolic phenotype. The available web-based tools for the prediction of pathogenic
MA
missense variants include FoldX, PolyPhen-2, SIFT Blink, SNPs3D and a few others. A total of 834 PAH variants and the genotypes of 4,181 PKU patients were recently analyzed using FoldX, PolyPhen-2, SIFT Blink, and SNPs3D, and a quantitative
ED
relationships were found between PAH protein stability and enzyme activity; protein stability and APV; and PAH activity and APV [12]. APV and in vitro PAH activity were
EP T
found to be best for predicting patients’ phenotypes and BH4 responsiveness [12]. As expected, PolyPhen-2 and SIFT Blink predicted most variants to be possibly damaging, damaging or deleterious in the present study. Only two MHP variants
AC C
were predicted to be benign by both algorithms. However, a benign prediction does not necessarily mean that the variant is not disease-causing. The APV expresses the frequency (%) of phenotypes in a homogeneous set of functionally hemizygous PKU genotypes and is defined as a linear scale ranging from 0 (cPKU) to 5 (mPKU) or 10 (MHP). It was developed from a non-linear arbitrary value (AV) system (power of 2) described by Guldberg et al. [19] and used in several studies, with a variable prediction outcome [20-23]. Wettstein at al. [12] correlated AV with several variant damage tools (see above) and in vitro PAH activity. There was a strong positive correlation between in vitro PAH activity and APV (r = 0.799, p<0.01), and mean PAH activity differed significantly between three APV groups p<0.001. PAH activity increased from 3.75 (±4.2%) for cPKU to 40.40 (±15.9%) for mPKU and
ACCEPTED MANUSCRIPT to 51.7 (±25.0%) for MHP. In a more recent study, genotype-phenotype associations were investigated in more than 9,000 PKU patients with various degrees of HPA and known genotypes, and a scale for APV, ranging from 0 for cPKU to 5 for mPKU and 10 for MHP, was introduced [8]. This system enabled the definition of cut-off ranges of for cPKU (0.0 - 2.7), mPKU (2.8 - 6.6) and for MHP (6.7 - 10.0) and was applied to a series of Chinese PKU patients [24]. We showed in this study that APV correlates not only with the metabolic phenotype but also with the low residual in vitro PAH
PT
activity and protein expression in cPKU variants (Figure 3).
RI
Different systems for transient expression of mutant PAH and the measurement of various forms of PAH activity have been used to address the question of how
SC
disease-associated PAH variants cause loss of PAH function [10,11,25-28]. None of the in vitro systems can reflect the hepatic in vivo situation, and reported PAH
NU
activities which differ significantly between the systems used when analyzing the same variant. Critical factors include protein regions (regulatory, catalytic or
MA
oligomerization) and affinity for the substrate Phe and cofactor BH4, all of which affect the sensitivity of the protein to structural changes. Cellular degradation of aberrant PAH protein appears to play a major role in the pathogenicity of numerous mutations
ED
[16]. Thus, in vitro expression analysis may not precisely reflect the level of the hepatic activity of certain variant enzymes, and the obtained activity levels are
EP T
probably slightly overestimated. While COS cells appeared to be favorable for differentiating between cPKU and mPKU variants, the E. coli expression system was superior in distinguishing between mPKU and MHP variants (Figure 1). It should,
AC C
however, be noted that most of the studies conducted in E. coli involved variant protein purification prior to activity measurement. As mentioned previously, not only different cell systems but also different PAH assays have been used to measure enzyme activity. Different methods, such as thinlayer chromatography (TLC) with
14
C-Phe to
14
C-Tyr conversion [27], HPLC with
fluorometric detection [29], or sophisticated mass-spectrometry methods involving the use of stable isotopes [18], have been employed to measure PAH activity at different temperatures, after different reaction times and using different substrates (BH4 or 6methyl-tetrahydropterin; 6MTHP). It is difficult to believe that these strategies produce exactly same results, but there is concordance for variants with a very low activity (see above). Since most of the PKU patients (76%) are compound heterozygotes
ACCEPTED MANUSCRIPT for PAH variants [5], interallelic complementation (positive or negative) may affect the hepatic PAH activity [14,15].Thus, the combination of the genotype-specific PAH activity with phenotype information can be more useful. PAH activity landscape is an alternative method for estimating PAH activity at different substrate and cofactor concentrations [13], and the results are probably closer to the hepatic PAH activities, as previously reported in liver biopsies of PKU patients [30].
PT
Wild-type PAH forms homo-oligomers, with an equilibrium between the tetrameric and dimeric forms. The two homotetramers and heterotetramers can be formed at
RI
different ratios, depending on the effects generated by variants (i.e., folding defects, reduced stability or low expression) [31]. Structural alterations are indicated by the
SC
increased susceptibility of the variant proteins to proteolytic degradation when expressed in E. coli [32]. The structural defects also result in defective oligomeric
NU
assembly. Thus, all of the variant PAH products may exhibit an increased tendency to form aggregates with a specific degradation pattern and reduced PAH activity. The
MA
three degradation products, which we observed in the Western blots of expressed variants, were not identified in this study but point to a specific degradation pattern for each of the variant classes. In particular, the ratio of the three proteins appears to be
ED
characteristic of some variants (Figure 2). The wide range of metabolic (and clinical) phenotypes of PKU has been shown to be
EP T
mainly determined by PAH genotype, although other factors might play a role as well. We investigated PAH and DNAJC12 expression associated with the wild-type enzyme and PAH variants and found that DNAJC12 was upregulated upon PAH
AC C
expression [33]. This observation is interesting, since DNAJC12 was recently reported to be a co-chaperone of PAH, and genetic defects in DNAJC12 result in mild hyperphenylalaninemia [34]. DNAJC12 is, together with the 70 kDa heat shock protein (HSP70) and nucleotide exchange factor (NEF), responsible for the proper folding, intracellular stability and ubiquitin-dependent degradation of PAH and may affect its expression pattern [34,35]. Ongoing investigations will hopefully provide more insights into PAH-DNAJC12 interactions. In summary, this study shows that in vitro PAH activity measurement does not always correlate with the expected metabolic phenotype, when compared in patients harboring the same variants in a hemizygous constellation. However, PAH activity and APV are still the best indicators for the damaging effect of genetic aberrations
ACCEPTED MANUSCRIPT and may be modified by the DNAJC12/HSP70 machinery and several epigenetic factors.
ACKNOWLEDGEMENTS This work is part of the RD-CONNECT initiative and was supported by the FP7-
PT
HEALTH-2012-INNOVATION-1 EU Grant No. 305444 (to NB), by the Dietmar-Hopp Foundation and by a grant from the China Scholarship Council (to NS). The authors
RI
would like to thank Kathrin Schwarz, Jana Hauke, and Peter Monostori for the
AC C
EP T
ED
MA
NU
SC
excellent technical assistance.
ACCEPTED MANUSCRIPT
PRO system
HEK293
APV
TNT-T7
Aberration
COS*
Variant
E. coli
Table 1. Overview of PAH variants and in vitro expression data.
PAH protein (%)***
Reference
PAH activity (%)**
p.D59Y p.T63P
73
c.136G>A
1.5
c.140C>T
10.0 123
c.143T>C
2.4
47
c.175G>T
0
9 92
c.187A>C
5.0 1.4
p.I65T p.I65S p.R68G p.R68S
c.194T>C
c.194T>G
2.5
c.202A>G
5.6
c.204A>T
7.0 6.3
c.227A>G
p.S87R
c.259A>C
10.0
p.T92I
c.275C>T
10.0
p.I95del
c.284_286del
p.D143G p.R155H p.R158Q p.F161S
26 27 33 48 * 32 *
40 100 98 * 25 * 24 47 88
*
22
c.428A>G
5.0
c.464G>A
8.8
52 * 98 * 21
0.2
c.473G>A
2.5
5 9 10 29 7 *
8.8
35
p.E178G
c.533A>G
7.8
31
p.E178K
c.533A>G
-
p.V190A
c.569T>C
32
6.6
100
40
3
; 96 ; 13
4
13
0
100
3
39
43
12
60
29
76
2.3.5
3 ; 100
28
25
27
4
93 ; 12 ; 100
100
100
2.3.5
c.631C>A
9.3
72
p.L212P
c.635T>C
0
17
[8] This study [22]
2.3
1
2
100 100
33
2.1
[8,12,22,23,33] [11,15,32,34,36]
1
[8,38] This study
2.3
[22]
91
[39]
100 ; 20 ; 72
[40] [22,23,39,41] This study [8]
3.5
4
1
100
3
[42] This study This study
4
9 1
100 ; 35 17 21
4
This study [37] This study [22,37] This study
1
42
1
100
3
1
[6,8,12,15,22,43] [44] [22] This study [12]
70
4
63
1
*
p.P211T
5
; 25 ; 22 ; 14
100
26
[22] This study [12,23,33,34]
100
27
2
[22,35]
2.3
27 67
[32] [22,33,34] This study
4
2.3
100
1.7
c.527G>T
p.R176L
0 62 40
46
76
0
c.482T>C
49
82
77
c.365C>A
67
*
6.7
AC C
p.P122Q
c.311C>A
*
EP T
p.E76G
p.A104D
*
PT
1.5
1
RI
p.L48S
c.117C>G
2
SC
p.A47V
2
NU
p.G46S
0
MA
p.F39L
c.1A>G
ED
p.M1V
[45] This study [39] [46]
ACCEPTED MANUSCRIPT 4.5 p.G218V
c.653G>T
p.Q226K
c.676C>A
p.V230I
c.688G>A
10.0
c.721C>T
9.5
c.722G>A
5.0
1
c.728G>A
0
0
10.0
p.V245E
c.734T>A
0
c.734T>C
10.0
p.G247V
c.740G>T
0.9
50 51 4
p.L249F
c.745C>T
5.6
51
p.L249P
c.746T>C
0
7
p.R252Q
c.754C>T
0
c.755G>A
0
c.764T>C
0
p.A259T
c.775G>A
1
p.A259V
c.776C>T
0 * 15 24
0
0
3.9
c.782G>C
2.4
c.809G>A
0
11
p.G272*
c.814G>T
0
0
p.Y277D
c.829T>G
1.1
0
p.T278I
c.833C>T
0
1
c.838G>A
0.1
c.842C>T
0
2 * 11 0 1
p.D282N
c.844G>A
0
p.I283F
c.847A>T
0.4
p.R261Q
p.R261P p.R270K
p.E280K p.P281L
p.R297H
AC C
3.9
c.782G>A
c.890G>A
9.8
[6,49]
1
4
100
1
7
62
0
1
0
8
3
0
3
3
52
48
100
[22]
100
2.3
[12,22]
1
[44] [46] [46]
45
4
[45] [8,49,53-55] This study
1
2
4
3
[29,51]
4
3
[29]
0 ; 100
2 ; 100
11 ; 100 2; 100
3.2
[29,49] [29,39,49] [12] This study
39
28 2
1
1
22 ; 20 ; 30 ; 100
2
[6,8,22,39,49] [11,15,36,43] This study [46]
2
[11,25,46]
0
[42,56] 99
*
[8,52] This study [45]
56
0
2 3 2
[8,45,50,51] This study
4
59
*
23 39 * 23 27 30 43 47 10
p.R261Q
c.781C>T
1
100
EP T
p.L255S
2
10 ; 9 ; 57
ED
p.R252W
0 ; 100
6 9 10 18 * 21 68 70
c.733G>A
-
[39] 1
0
p.V245M
c.749C>T
[22] [48] This study
*
c.731C>T
p.S250F
100
[47] 3
25 * 57 23
p.P244L
p.V245A
63
6
PT
0
52
[8,22,39] This study
3.1
RI
c.727C>T
0.5 p.R243Q
100
SC
p.R243*
63
NU
p.R241H
2
MA
p.R241C
15 * 25 101
0; 10 1
6
1
2
23
10
3
1
2
[8] 1
100 ; 0 ; 2 0
1
1
[57] [6,8,22,49] This study [22,54,58] [22] [22] This study
ACCEPTED MANUSCRIPT
c.896T>G
0.2
c.898G>T
8.9
c.916A>G
10.0
c.926C>T
3.3
c.932T>C
0
12 70 * 0
p.A313T
c.937G>A
5.0
29
p.A322G
c.965C>G
10.0
75
p.L333F
c.997C>T
-
7
p.A342T
c.1024G>A
p.L311P
p.L348V
c.1042C>G
p.S349P
c.1045T>C
0
IVS10-11G>A c.1066-11G>A 0 c.1139C>T 10.0 p.T380M p.F39del
c.116_118del
p.V388M
c.1162G>A
p.E390G
c.1169A>G
p.A395P
c.1183G>C
p.A403V
c.1208C>T
7.6
9.4
p.R413S
AC C
0
p.R408W
*
c.1222C>T
c.1237C>A
5.0
c.1238G>C
0.1
c.1243G>A
10.0
p.Y417H
c.1249T>C
5.0
44
p.Q419R
c.1256A>G
10.0
5.0 p.Y414C p.D415N
c.1241A>G
p.A434D
c.1301C>A
IVS12+1G>A
c.1315+1G>A
41
0
44
0
83
20
23
41
75
15
32 33 * 100 0 1 * 2 3 5 34 2 * 11 28 50 80 * 35
p.R413P
26
28 15 43 * 83 54 62 70
0
0
1
27
[12] This study [22] This study [8] This study [8,62] This study This study [59] [39]
3
[22] [22,24,39,61] This study
2
100 ; 0 100
1
100
2
1
[22,63,64] [8] [46]
1.3
1
100 ; 96 ; 22
85 100
2
2
[22] [11,22,24,36,57] [25,65] This study [12,15,22,39]
16
[22] 100
1
[39,66] This study
1
[8,12,15,22,60,67] This study
1 1
0 ;3
0 38
42
[39] [68] This study
1
2.3
1
100 ; 50 ; 84 72
114
100
2.3
*
70 9
1
100
1
[22] [22] This study This study [69]
*
2.6
[49,61]
1
105
0
EP T
0.9
0
24
1.3 2.5
100
*
25 33 38 0
2
39
*
4.0 3.6
12
100
PT
p.A309V
1
[8,12,22,59,60] This study
1
RI
p.I306V
1
SC
p.A300S
32 * 65 * 25
0 70 ; 91
ED
p.F299C
9
NU
p.R408Q
33 * 41 55 84 2
MA
6.2 c.890G>A
This study 0
1
[70]
ACCEPTED MANUSCRIPT
AC C
EP T
ED
MA
NU
SC
RI
PT
*This study; **Compared with the wild-type activity; **1 = COS; 2 = E.coli; 3 = TNT-T7; 4 = HEK293; 5= Yeast 6 = HepG2; 7 = PRO. APV: Allelic phenotype value (cPKU = 0 – 2.7; mPKU = 2.8 – 6.6; MHP = 6.7 – 10.0)
ACCEPTED MANUSCRIPT Table 2. Summary of PAH variants investigated, number of patients with the corresponding functionally hemizygous genotype in the BIOPKU database, allelic phenotype values, in vitro PAH activity, and PAH expression in COS-7 cells.
*
PAH band 1 ** (%) 83 (9) 33 (19) 68 (17) 78 (16) 74 (8) 80 (4) 79 (16) 70 (18) 80 (3) 77 (18) 38 (17) 30 (29) 58 (17) 68 (14) 57 (14) 66 (16) 66 (9) 25 (11) 68 (9) 79 (3) 59 (9) 78 (13) 3 (1) 63 (8) 34 (19) 2 (1) 11 (4) 1 (1) 0 (0) 10 (6) 6 (1) 2 (1) 43 (6) 7 (5) 56 (11)
PT
100 10.0 100 10.0 96 9.8 91 9.5 90 9.5 83 9.1 79 9.0 75 8.8 41 7.4 43 7.1 20 6.0 9 5.5 9 5.5 0 5.0 0 5.0 0 5.0 0 3.3 0 2.8 0 2.6 0 2.5 0 1.9 0 1.7 0 1.4 0 1.3 0 0.4 0 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RI
0 0 4 9 10 17 21 25 59 57 80 91 91 100 100 100 67 56 52 50 37 34 28 26 7 2 1 1 1 0 0 0 0 0
PAH total protein ** (%) 100 (19) 39 (22) 80 (26) 92 (29) 80 (13) 123 (15) 113 (38) 77 (21) 90 (9) 89 (23) 56 (29) 108 (42) 76 (32) 80 (16) 148 (48) 84 (34) 78 (19) 35 (18) 99 (38) 84 (8) 64 (11) 98 (17) 4 (2) 91 (34) 41 (26) 3 (2) 17 (7) 1 (1) 0 (0) 12 (7) 13 (3) 16 (2) 87 (17) 20 (10) 77 (16)
SC
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 44 48 50 63 66 72 74 93 98 99 99 99 100 0 0 0 0
PAH activity ** (%) 100 (8) 123 (25) 35 (32) 39 (12) 35 (22) 25 (4) 33 (21) 65 (24) 21 (9) 40 (22) 24 (5) 57 (14) 25 (21) 41 (22) 40 (17) 98 (30) 29 (13) 12 (7) 77 (19) 9 (2) 25 (15) 83 (46) 25 (24) 73 (17) 23 (19) 6 (1) 11 (7) 11 (4) 15 (9) 2 (0) 11 (8) 44 (22) 32 (2) 9 (1) 21 (1)
NU
c.140C>T 4 c.1243G>A 66 c.890G>A 23 c.527G>T 11 c.916A>G 78 c.1208C>T 281 c.898G>T 183 c.464G>A 4 c.569T>C 19 c.227A>G 7 c.721C>T 173 c.204A>T 75 c.1223G>A 55 c.202A>G 8 c.428A>G 2 c.937G>A 1 c.926C>T 21 c.311C>A 50 c.1301C>A 27 c.653G>T 10 c.1162G>A 169 c.1042C>G 94 c.117C>G 47 c.782G>A 489 c.728G>A 405 c.1283G>C 109 c.838G>A 136 c.754C>T 142 c.1222C>T 1845 c.932T>C 9 c.1249T>C 0 c.194T>G 0 c.175G>T 0 c.731C>T 0
cPKU mPKU MHP APV (%) (%) (%)
MA
*
ED
n
EP T
PAH_WT p.A47V p.D415N p.R297H p.R176L p.I306V p.A403V p.A300S p.R155H p.V190A p.E76G p.R241C p.R68S p.R408Q p.R68G p.D143G p.A313T p.A309V p.A104D p.A434D p.G218V p.V388M p.L348V p.F39L p.R261Q p.R243Q p.R413P p.E280K p.R252W p.R408W p.L311P p.Y417H p.I65S p.D59Y p.P244L
Nucleotide aberration
AC C
Variant
PAH band 2 ** (%) 9 (6) 3 (2) 6 (6) 10 (9) 3 (3) 38 (10) 16 (10) 5 (2) 5 (2) 9 (4) 9 (5) 61 (6) 10 (9) 11 (1) 43 (11) 6 (6) 6 (5) 5 (4) 11 (10) 5 (5) 4 (2) 19 (3) 1 (1) 11 (10) 5 (5) 1 (1) 5 (2) 0 (0) 0 (0) 1 (0) 5 (1) 14 (1) 37 (10) 3 (2) 6 (3)
PAH band 3 ** (%) 8 (4) 3 (1) 6 (3) 4 (4) 3 (2) 5 (1) 18 (12) 2 (1) 5 (4) 3 (1) 8 (7) 17 (7) 8 (6) 1 (1) 48 (23) 12 (12) 6 (5) 5 (3) 20 (19) 0 (0) 1 (0.4) 1 (1) 0 (0) 17 (16) 2 (2) 0 (0) 1 (1) 0 (0) 0 (0) 1 (1) 2 (1) 0 (0) 7 (1) 10 (3) 15 (2)
number of patients in the BIOPKU database (http://www.biopku.org); **mean (SD); cPKU: classic PKU; mPKU: mild PKU; MHP: mild hyperphenylalaninemia; APV: allelic phenotype value; PAH: phenylalanine hydroxylase; n.d.: not done.
ACCEPTED MANUSCRIPT
AF
SIFT
p.A47V
c.140C>T
0.03
0.03
0.12
Regulatory domain / destabilizing
p.D415N
c.1243G>A
0.44
1.00
0.00
Oligomerization domain / benign
p.R297H
c.890G>A
0.14
0.01
0.84
Catalytic domain / destabilizing
p.R176L
c.527G>T
0.15
0.02
0.00
Catalytic domain / destabilizing
p.I306V
c.916A>G
0.60
0.09
0.90
Catalytic domain/ destabilizing
p.A403V
c.1208C>T
2.40
0.00
1.00
Catalytic domain / destabilizing
p.A300S
c.898G>T
1.50
0.00
1.00
Catalytic domain / destabilizing / T1/2 -variant
p.R155H
c.464G>A
0.03
0.00
1.00
Catalytic domain/ destabilizing
p.V190A
c.569T>C
0.13
0.00
1.00
Catalytic domain/ destabilizing
p.E76G
c.227A>G
0.04
0.51
0.00
p.R241C
c.721C>T
1.10
0.02
1.00
p.R68S
c.204A>T
0.60
0.00
p.R408Q
c.1223G>A
0.63
0.01
p.R68G
c.202A>G
0.05
0.00
1.00
Regulatory domain/ destabilizing
p. D143G
c.428A>G
0.01
0.01
0.15
Catalytic domain/ destabilizing
p.A313T
c.937G>A
0.01
0.00
1.00
Catalytic domain/ destabilizing/ affects splicing
p.A309V
c.926C>T
0.22
0.01
1.00
Catalytic domain/ destabilizing
AC C
p.A104D
Polyphen2 Comment
NU
SC
RI
PT
cDNA
EP T
Protein
ED
Table 3. Alleles frequency (AF), SIFT and Polyphewn-2 damaging effect prediction values and some properties of variants investigated.
Regulatory domain / benign
MA
Catalytic domain/ destabilizing
1.00
Regulatory domain/ destabilizing
1.00
Catalytic domain/ destabilizing
c.311C>A
0.42
0.63
0.00
Regulatory domain / benign-destabilizing
c.1301C>A
0.14
0.01
1.00
Oligomerization domain/ destabilizing
c.653G>T
0.11
0.00
1.00
Catalytic domain/ destabilizing
p.V388M
c.1162G>A
1.80
0.00
1.00
Catalytic domain/ destabilizing / Km-variant
p.L348V
c.1042C>G
0.90
0.00
1.00
Catalytic domain/ destabilizing
p.F39L
c.117C>G
0.42
0.01
1.00
Regulatory domain/ destabilizing
p.R261Q
c.782G>A
5.50
0.00
1.00
Catalytic domain/ destabilizing / CBR1
p.R243Q
c.728G>A
2.60
0.00
1.00
Catalytic domain/ destabilizing
p.R413P
c.1283G>C
0.69
0.00
0.81
Oligomerization domain/ destabilizing
p.A434D p.G218V
ACCEPTED MANUSCRIPT c.838G>A
1.30
0.00
1.00
Catalytic domain/ destabilizing / CBR2
p.R252W
c.754C>T
1.40
0.00
1.00
Catalytic domain/ destabilizing / CBR1
p.R408W
c.1222C>T
19.70 0.00
1.00
Catalytic domain/ destabilizing / Folding variant
p.L311P
c.932T>C
0.17
0.00
1.00
Catalytic domain/ destabilizing
p.Y417H
c.1249T>C
0.03
0.00
1.00
Oligomerization domain / destabilizing / Folding variant
p.I65S
c.194T>G
0.01
0.00
1.00
Regulatory domain/ destabilizing / Folding variant
p.D59Y
c.175G>T
0.01
0.03
0.20
Regulatory domain/ destabilizing
p.P244L
c.731C>T
0.09
0.00
1.00
Catalytic domain/ destabilizing
RI
PT
p.E280K
AC C
EP T
ED
MA
NU
SC
SIFT: <0.06 = deleterious; Polyphen-2: 0-0.15 = benign, 0.15-1.0 = possibly damaging, 0.85-1.0 damaging; A; F: allele frequency (%)CBR1: cofactor (BH4)-binding region 1; CBR2: cofactor (BH4)binding region 2; T1/2: half-time
ACCEPTED MANUSCRIPT REFERENCES [1]
Blau N., Van Spronsen F. J., Levy H.L., Phenylketonuria. Lancet 376 (2010) 1417-1427.
[2]
Vockley J., Andersson H. C., Antshel K. M., Braverman N. E., Burton B. K., Frazier D. M., Mitchell J., Smith W. E., Thompson B. H., Berry S. A., Phenylalanine hydroxylase deficiency: diagnosis and management guideline.
[3]
PT
Genet Med 16 (2014) 188-200.
Camp K. M., Parisi M. A., Acosta P. B., Berry G. T., Bilder D. A., Blau N.,
RI
Bodamer O. A., Brosco J. P., Brown C. S., Burlina A. B., Burton B. K., Chang C. S., Coates P. M., Cunningham A. C., Dobrowolski S. F., Ferguson J. H.,
SC
Franklin T. D., Frazier D. M., Grange D. K., Greene C. L., Groft S. C., Harding C. O., Howell R. R., Huntington K. L., Hyatt-Knorr H. D., Jevaji I. P., Levy H.
NU
L., Lichter-Konecki U., Lindegren M. L., Lloyd-Puryear M. A., Matalon K., Macdonald A., McPheeters M. L., Mitchell J. J., Mofidi S., Moseley K. D.,
MA
Mueller C. M., Mulberg A. E., Nerurkar L. S., Ogata B. N., Pariser A. R., Prasad S., Pridjian G., Rasmussen S. A., Reddy U. M., Rohr F. J., Singh R. H., Sirrs S. M., Stremer S. E., Tagle D. A., Thompson S. M., Urv T. K., Utz J.
ED
R., van Spronsen F., Vockley J., Waisbren S. E., Weglicki L. S., White D. A., Whitley C. B., Wilfond B. S., Yannicelli S., Young J. M., Phenylketonuria
EP T
scientific review conference: State of the science and future research needs. Mol Genet Metab 112 (2014) 87-122. [4]
Blau N., Shen N., Carducci C., Molecular genetics and diagnosis of
[5]
AC C
phenylketonuria: state of the art. Expert Rev Mol Diagn 14 (2014) 655-671. Blau N., Genetics of Phenylketonuria: Then and Now. Hum Mutat 37 (2016) 508-15. [6]
Desviat L. R., Perez B., Gamez A., Sanchez A., Garcia M. J., Martinez-Pardo M., Marchante C., Boveda D., Baldellou A., Arena J., Sanjurjo P., Fernandez A., Cabello M. L., Ugarte M., Genetic and phenotypic aspects of phenylalanine hydroxylase deficiency in Spain: molecular survey by regions. Eur J Hum Genet 7 (1999) 386-92.
[7]
Daniele A., Scala I., Cardillo G., Pennino C., Ungaro C., Sibilio M., Parenti G., Esposito L., Zagari A., Andria G., Salvatore F., Functional and structural characterization of novel mutations and genotype-phenotype correlation in 51
ACCEPTED MANUSCRIPT phenylalanine hydroxylase deficient families from Southern Italy. FEBS J 276 (2009) 2048-59. [8]
Garbade S., Shen N., Himmelreich N., Haas D., Trefz F.K., Hoffmann G.F., Burgard P., Blau N., Allelic phenotype values: a model for genotype-based phenotype prediction in phenylketonuria. Genet Med (2018) DOI: 10.1038/s41436-018-0081-x.
[9]
Okano Y., Eisensmith R. C., Güttler F., Lichter-Konecki U., Konecki D. S.,
PT
Trefz F. K., Dasovich M., Wang T., Henriksen K., Lou H., et al., Molecular basis of phenotypic heterogeneity in phenylketonuria. N Engl J Med 324
Waters P. J., Parniak M. A., Nowacki P., Scriver C. R., In vitro expression
SC
[10]
RI
(1991) 1232-8.
analysis of mutations in phenylalanine hydroxylase: linking genotype to
[11]
NU
phenotype and structure to function. Hum Mutat 11 (1998) 4-17. Pey A. L., Desviat L. R., Gamez A., Ugarte M., Perez B., Phenylketonuria: genotype-phenotype correlations based on expression analysis of structural
[12]
MA
and functional mutations in PAH. Hum Mutat 21 (2003) 370-8. Wettstein S., Underhaug J., Perez B., Marsden B. D., Yue W. W., Martinez A.,
ED
Blau N., Linking genotypes database with locus-specific database and genotype-phenotype correlation in phenylketonuria. Eur J Hum Genet 23 (2015) 302–309.
Danecka M. K., Woidy M., Zschocke J., Feillet F., Muntau A. C., Gersting S.
EP T
[13]
W., Mapping the functional landscape of frequent phenylalanine hydroxylase (PAH) genotypes promotes personalised medicine in phenylketonuria. J Med
[14]
AC C
Genet 52 (2015) 175-85.
Leandro J., Nascimento C., de Almeida I. T., Leandro P., Co-expression of different subunits of human phenylalanine hydroxylase: evidence of negative interallelic complementation. Biochim Biophys Acta 1762 (2006) 544-50.
[15]
Shen N., Heintz C., Thiel C., Okun J. G., Hoffmann G. F., Blau N., Coexpression of phenylalanine hydroxylase variants and effects of interallelic complementation on in vitro enzyme activity and genotype-phenotype correlation. Mol Genet Metab 117 (2015) 328–335.
[16]
Waters P. J., How PAH gene mutations cause hyper-phenylalaninemia and why mechanism matters: insights from in vitro expression. Hum Mutat 21 (2003) 357-69.
ACCEPTED MANUSCRIPT [17]
Aguado C., Perez B., Ugarte M., Desviat L. R., Analysis of the effect of tetrahydrobiopterin on PAH gene expression in hepatoma cells. FEBS Lett 580 (2006) 1697-701.
[18]
Heintz C., Troxler H., Martinez A., Thony B., Blau N., Quantification of phenylalanine hydroxylase activity by isotope-dilution liquid chromatographyelectrospray ionization tandem mass spectrometry. Mol Genet Metab 105 (2012) 559-65. Guldberg P., Rey F., Zschocke J., Romano V., Francois B., Michiels L., Ullrich
PT
[19]
K., Hoffmann G. F., Burgard P., Schmidt H., Meli C., Riva E., Dianzani I.,
RI
Ponzone A., Rey J., Güttler F., A European multicenter study of phenylalanine
SC
hydroxylase deficiency: classification of 105 mutations and a general system for genotype-based prediction of metabolic phenotype. Am J Hum Genet 63
[20]
NU
(1998) 71-9.
Kasnauskiene J., Cimbalistiene L., Kucinskas V., Validation of PAH genotypebased predictions of metabolic phenylalanine hydroxylase deficiency
MA
phenotype: investigation of PKU/MHP patients from Lithuania. Med Sci Monit 9 (2003) CR142-6.
Santos L. L., Fonseca C. G., Starling A. L., Januario J. N., Aguiar M. J.,
ED
[21]
Peixoto M. G., Carvalho M. R., Variations in genotype-phenotype correlations in phenylketonuria patients. Genet Mol Res 9 (2010) 1-8. Zhu T., Ye J., Han L., Qiu W., Zhang H., Liang L., Gu X., Variations in
EP T
[22]
genotype-phenotype correlations in Phenylalanine hydroxylase deficiency in Chinese Han Population. Gene 529 (2013) 80-87. Zhu T., Ye J., Han L., Qiu W., Zhang H., Liang L., Gu X., The Predictive Value
AC C
[23]
of Genetic Analyses in the Diagnosis of Tetrahydrobiopterin (BH4)Responsiveness in Chinese Phenylalanine Hydroxylase Deficiency Patients. Sci Rep 7 (2017) 6762. [24]
Wang R., Shen N., Ye J., Han L., Qiu W., Zhang H., Liang L., Sun Y., Fan Y., Wang L., Wang Y., Gong Z., Liu H., Wang J., Yan H., Blau N., Gu X., Yu Y., Mutation spectrum of hyperphenylalaninemia candidate genes and the genotype-phenotype correlation in the Chinese population. Clin Chim Acta 481 (2018) 132-138.
[25]
Gjetting T., Petersen M., Guldberg P., Guttler F., In vitro expression of 34 naturally occurring mutant variants of phenylalanine hydroxylase: correlation
ACCEPTED MANUSCRIPT with metabolic phenotypes and susceptibility toward protein aggregation. Mol Genet Metab 72 (2001) 132-43. [26]
Waters P. J., Scriver C. R., Parniak M. A., Homomeric and heteromeric interactions between wild-type and mutant Phenylalanine hydroxylase subunits: Evaluation of two-hybrid approaches for functional analysis of mutations causing hyperphenylalaninemia. Mol Genet Metab 73 (2001) 230-8.
[27]
Gamez A., Perez B., Ugarte M., Desviat L. R., Expression analysis of
PT
phenylketonuria mutations. Effect on folding and stability of the phenylalanine hydroxylase protein. J Biol Chem 275 (2000) 29737-42.
Leandro P., Lechner M. C., Tavares de Almeida I., Konecki D., Glycerol
RI
[28]
SC
increases the yield and activity of human phenylalanine hydroxylase mutant enzymes produced in a prokaryotic expression system. Mol Genet Metab 73
[29]
NU
(2001) 173-8.
Doskeland A. P., Doskeland S. O., Ogreid D., Flatmark T., The effect of ligands of phenylalanine 4-monooxygenase on the cAMP-dependent
MA
phosphorylation of the enzyme. The Journal of biological chemistry 259 (1984) 11242-8.
Bartholomé K., Lutz P., Bickel H., Determination of phenylalanine hydroxylase
ED
[30]
activity in patients with phenylketonuria and hyperphenylalaninemia. Pediatr Res 9 (1975) 899.
Erlandsen H., Stevens R. C., A structural hypothesis for BH4 responsiveness
EP T
[31]
in patients with mild forms of hyperphenylalaninaemia and phenylketonuria. J Inherit Metab Dis 24 (2001) 213-30. Bjorgo E., Knappskog P. M., Martinez A., Stevens R. C., Flatmark T., Partial
AC C
[32]
characterization and three-dimensional-structural localization of eight mutations in exon 7 of the human phenylalanine hydroxylase gene associated with phenylketonuria. Eur J Biochem 257 (1998) 1-10. [33]
Himmelreich N., Hoffmann G.F., Blau N., Expression pattern of phenylalanine hydroxylase variants is regulated by co-chaperone DNAJC12. Mol Genet Metab 123 (2018) 210.
[34]
Anikster Y., Haack T. B., Vilboux T., Pode-Shakked B., Thony B., Shen N., Guarani V., Meissner T., Mayatepek E., Trefz F. K., Marek-Yagel D., Martinez A., Huttlin E. L., Paulo J. A., Berutti R., Benoist J. F., Imbard A., Dorboz I., Heimer G., Landau Y., Ziv-Strasser L., Malicdan M. C., Gemperle-Britschgi C.,
ACCEPTED MANUSCRIPT Cremer K., Engels H., Meili D., Keller I., Bruggmann R., Strom T. M., Meitinger T., Mullikin J. C., Schwartz G., Ben-Zeev B., Gahl W. A., Harper J. W., Blau N., Hoffmann G. F., Prokisch H., Opladen T., Schiff M., Biallelic Mutations in DNAJC12 Cause Hyperphenylalaninemia, Dystonia, and Intellectual Disability. Am J Hum Genet 100 (2017) 257-266. Blau N., Martinez A., Hoffmann G. F., Thony B., DNAJC12 deficiency: A new strategy in the diagnosis of hyperphenylalaninemias. Mol Genet Metab 123
EP T
ED
MA
NU
SC
RI
PT
(2018) 1-5.
AC C
[35]
ACCEPTED MANUSCRIPT LEGEND TO FIGURES Figure 1. Comparison of the PAH activity (mean, 25th - 75th percentile, min, max) of variants associated with three metabolic phenotypes (cPKU, mPKU and MHP) transiently expressed in (A) COS cells, (B) E. coli and (C) in TNT-T7 cells. The assignment of variants to metabolic phenotypes was carried out using APV (see Materials and Methods). Average PAH activity was calculated when there was more
PT
than one report for the same variant. The unpaired t-test was considered statistically significant when p<0.02. cPKU: classic phenylketonuria; mPKU: mild
RI
phenylketonuria; MHP: mild hyperphenylalaninemia.
SC
Figure 2. PAH expression patterns and PAH activities of 34 variants transiently expressed variants in COS-7 cells. Gray-shaded bares represent the amount (%) of
NU
PAH protein (bands 1, 2 and 3), and open bars represent PAH activity (average of three measurements in both cases). For details, see the text and Table 2.
MA
Figure 3. Comparison of (A) PAH protein expression and (B) PAH activity (mean, 25th - 75th percentile, min, max) of 34 variants in COS-7 cells. The assignment of variants to metabolic phenotypes was carried out using APV (see Materials and
ED
Methods). The unpaired t-test was considered statistically significant when p<0.02. cPKU: classic phenylketonuria; mPKU: mild phenylketonuria; MHP: mild
AC C
EP T
hyperphenylalaninemia.
Figure 1
Figure 2
Figure 3
Nastassja Himmelreich, Nan Shen, Jürgen G. Okun, Christian Thiel, Georg F. Hoffmann, Nenad Blau PII: DOI: Reference:
S1096-7192(18)30283-X doi:10.1016/j.ymgme.2018.06.011 YMGME 6369
To appear in:
Molecular Genetics and Metabolism
Received date: Revised date: Accepted date:
25 May 2018 21 June 2018 22 June 2018
Please cite this article as: Nastassja Himmelreich, Nan Shen, Jürgen G. Okun, Christian Thiel, Georg F. Hoffmann, Nenad Blau , Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria. Ymgme (2018), doi:10.1016/j.ymgme.2018.06.011
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Relationship between genotype, phenylalanine hydroxylase expression and in vitro activity and metabolic phenotype in phenylketonuria
Nastassja Himmelreich1, Nan Shen1,2, Jürgen G Okun1, Christian Thiel1, Georg F Hoffmann1, Nenad Blau1,* 1
PT
Center for Child and Adolescent Medicine and Dietmar-Hopp Metabolic Center,
University of Heidelberg, Heidelberg, Germany; 2Department of Rehabilitation
RI
Medicine, Xin Hua Hospital affiliated to Shanghai Jiao Tong University School of
SC
Medicine, Shanghai, China
NU
*Correspondence:
Dietmar-Hopp Metabolic Center
MA
Prof. Dr. Nenad Blau
University Children’s Hospital Heidelberg
ED
Im Neuenheimerfeld 669 69120 Heidelberg
EP T
Germany
AC C
[email protected]
ACCEPTED MANUSCRIPT ABSTRACT Residual phenylalanine hydroxylase (PAH) activity is the main determinant of the metabolic phenotype in phenylketonuria (PKU). The genotypic heterogeneity of PKU, involving more than 1,000 PAH variants and over 2,500 different genotypes, makes genotype-based phenotype prediction challenging. While a relationship between PAH variants and the metabolic phenotype is well established, we questioned the
PT
importance of PAH expression and residual in vitro activity for the metabolic phenotype.
RI
Thirty-four PAH variants (p.F39L, p.A47V, p.D59Y, p.I65S, p.R68G, p.R68S, p.E76G,
SC
p.A104D, p.D143G, p.R155H, p.R176L, p.V190A, p.G218V, p.R241C, p.R243Q, p.P244L, p.R252W, p.R261Q, p.E280K, p.R297H, p.A300S, p.I306V, p.A309V,
NU
p.L311P, p.A313T, p.L348V, p.V388M, A403V, p.R408Q, p.R408W, p.R413P, p.D415N, p.Y417H, and p.A434D) were transiently transfected into COS-7 cells, and expression of PAH was investigated. Expression patterns were compared with in vitro
MA
PAH activity and allelic phenotype values (APVs).
In vitro PAH activity was significantly higher (p<0.01) in variants associated with mild
ED
hyperphenylalaninemia (PAH activity = 52.1 ± 8.5%; APV = 6.7 - 10.0) than that in classic PKU variants (PAH activity = 21.1 ± 7.0%; APV = 0 - 2.7). Mild PKU variants
EP T
(PAH activity = 40.2 ± 7.6%; APV = 2.8 - 6.6) were not significantly different from mild hyperphenylalaninemia, but there was a difference (p<0.048) compared with classic
AC C
PKU phenotypes.
ACCEPTED MANUSCRIPT INTRODUCTION Phenylketonuria (PKU) is phenotypically a heterogeneous disease, ranging from mild hyperphenylalaninemia (MHP; blood phenylalanine (Phe) 120 - 600 mol/L) to mild PKU (mPKU; blood Phe 600 - 1,200 mol/L) and classic PKU (cPKU; with blood Phe levels >1,200 mol/L) [1]. From a clinical perspective, a finer classification of MHP may have some significance for outcome and treatment of MHP patients. The recent
PT
US guidelines [2] and the NIH PKU scientific review conference [3] splitted MHP into MHP not requiring treatment (blood Phe 120 – 600 mol/L) and MHP-gray zone (360
RI
– 600 mol/L). MHP-grey zone patients may require treatment in some countries,
SC
while MHP patients may not. The metabolic phenotype in PKU is a consequence of loss of phenylalanine hydroxylase (PAH) function, caused by more than 1,000 PAH variants and over 2,500 different genotypes, resulting in hyperphenylalaninemia
NU
(HPA) [4]. Severe cPKU results in almost no or very little residual PAH activity, while in MHP patients substantial residual activity is reported in vivo and in vitro. Data from
MA
the locus-specific PAHvdb database (www.biopku.org/pah) demonstrate that 65% of the variants involved are amino acid substitutions, including changes in nonsense codons, and other common variants, such as deletions (15%), splice variants (12%),
ED
insertions/duplications/indels (4%) and synonymous variants (4%); thus, PKU is a
EP T
highly heterogeneous disease [4,5].
Extensive attempts to correlate genotypes with PKU phenotypes were reported in the past, which achieved results ranging from incomplete or no correlations to substantial
AC C
correlations between PAH variants and corresponding metabolic phenotypes [6,7]. A recent definition of the allelic phenotype value (APV) algorithm assigned PAH variants occurring in functionally hemizygous constellations to one of the three PKU phenotypes. APV-based phenotype prediction was found to be 99% correct for cPKU, 46% for mPKU and 90% for MHP within a cohort of more than 9,000 PKU patients [8]. Correlations between genotypes and in vitro residual PAH activity have been documented for several PAH variants [9-11]. In addition to PAH protein stability and APV, enzyme activity has also been used to predict patients phenotypes [12]. In contrast to APV, it has been shown that calculated enzyme activities based on heterogeneous data from single mutations do not reliably reflect residual activity in the patient [13] and that the two alleles may affect each other, a phenomenon known
ACCEPTED MANUSCRIPT as interallelic complementation [14,15]. In addition, the methods used for the measurement of PAH expression and activity measurement differ greatly, resulting in a broad range of activities for the same variant protein [10,16]. To better understand the relationship between PAH protein expression and enzyme activity, we transiently expressed 34 common PAH variants in COS-7 cells, investigated PAH expression and measured in vitro residual PAH activity under
PT
standardized conditions. Results were then compared with APV data. In addition, we conducted a literature search and compared different expression systems and PAH
SC
RI
activities reported.
MATERIALS AND METHODS
NU
Cloning
The cloning of PAH variants was performed as described previously [15]. In brief, the
MA
expression plasmid pCMV-FLAG-PAH (Promoter-N-Fusion-PAH) was received as a courtesy gift from L. R. Desviat [17]. Mutations in the human PAH-cDNA sequence
ED
were introduced via site-directed mutagenesis using the QuikChange II XL kit from Agilent Technologies (Santa Clara, CA, USA) and confirmed thgrough DNA
EP T
sequencing analysis at SEQLAB Sequence Laboratories (Göttingen, Germany). Transient transfections
For the transient PAH expression COS-7 monkey kidney cells were used. Standard
AC C
culture conditions of 37°C and 5% CO2 in a sterile environment were used. Cells were washed with 1 x PBS every 72 h, and fresh Dulbecco’s modified Eagle’s medium (DMEM), enriched with 10% fetal calf serum (FCS) plus 1% Pen/Strep was added. Then, 106 COS-7 cells plated on 10 cm dishes and were transfected with 10 µg of plasmid DNA and 1 µg of the pSV-β-galactosidase control vector (Promega, Madison, USA) using Fugene HD transfections reagents (Promega, Madison, USA) for 72 h. Finally, the cells were harvested with trypsin and washed 2-times with icecold PBS, then flash frozen in liquid nitrogen for storage at −80°C. Each variant was analyzed in three independent transfection experiments.
ACCEPTED MANUSCRIPT Western blotting Cell lysates were prepared by macerating the cells in 1 M Sucrose, 1 x PBS and 1 x PIM buffer 30 times through a 20G cannula, followed by 3 x 10 sec sonification and centrifugation (30 min/13,000rpm/ 4°C). The high salt content was removed from the supernatant using Zeba Desalting Spin Columns (Pierce Biotechnology, Waltham, MA). For Western blot analyses 20 µg of the sample supernatant proteins was
PT
unstitched through 12% SDS-PAGE and transferred via semi-dry electrophoretic blotting onto a nitrocellulose membrane (GE Healthcare, CT). The membrane was
RI
then blocked for 1 h at room temperature in 5% milk powder in either PBS-T (phosphate-buffered saline with 0.1% Tween) or TBS-T (0.1% Tween in Tris Buffered
SC
Saline). Primary antibodies against PAH (mouse α human, 1:1,000 dilution, Merck Millipore, MA) and ß-actin (mouse α human, 1:10,000, Sigma-Aldrich, MO) were
NU
used, and the samples were incubated overnight at 4°C under constant movement. The secondary species-specific antibody conjugated with a horseradish peroxidase
MA
tag facilitated detection via a chemiluminescent reaction with the ECL reagent (Pierce Biotechnology, Waltham, MA). Quantification of Western blots was performed by ImageJ (https://imagej.nih.gov/ij/) as decribed previously [15]. Data points represent
EP T
PAH activity assay
ED
means of n=3 independent transfection experiments.
In vitro PAH activity measurements were performed for each transfection as described previously using the isotope-dilution liquid chromatography-electrospray
AC C
ionization tandem mass spectrometry technique as triplicates [18]. Allelic phenotype values (APV) The APV value was calculated for PAH variants expressed in COS cells. The APV is based on the frequencies of the metabolic phenotype (i.e., cPKU, mPKU or MHP) for genotypes presenting in a functionally hemizygous state and calculated using the following formula: APV = (%cPKU*0 + %mPKU*5 + %MHP*10)/100 where % indicates the percentage of phenotypes for a given functionally hemizygous genotype [8]. APVs are dynamically calculated from the actual BIOPKU dataset (n=10,340; as of May 2018) and stored in the locus-specific PAHvdb database.
ACCEPTED MANUSCRIPT Searching for a specific genotype in BIOPKU generates a report of the corresponding APV and the predicted phenotype, depending on whether the APVs of both alleles are known (http://www.biopku.org). Prediction of the damaging effects of missense variants The PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2) and SIFT Blink
PT
(http://sift.jcvi.org) algorithms were used to predict the effect of variants on protein. Statistics
RI
All experiments were repeated at least three times. Prism 7 (GraphPad Software Inc.,
NU
SC
CA, USA) was used for descriptive statistics and the unpaired Student’s t-test.
RESULTS
MA
Comparison of in vitro PAH activity in different expression systems We collated the in vitro enzyme activity and protein expression of 87 PAH variants reported in 49 publications (including this one) (Table 1). Most of the variants (51%)
ED
were transiently expressed in monkey kidney COS cells, in E. coli (24%), using the T7-based coupled transcription/translation system (24%) or in human embryonic
EP T
kidney 293 cells (1%). The hepatocytes HepG2 cells, yeast and PRO bacterial expression system were mainly used for PAH protein pattern analyses. We compared the average reported PAH activities determined using the three commonly
AC C
employed expression systems (see above) with the phenotypes (based on APV, when occurring in the functionally hemizygous constellation) of the 73 PAH variants (Figure 1). Based on PAH activity, the COS cells system appeared to be the best system for differentiating between cPKU and mPKU variants (p<0001), while those associated with mPKU and MHP overlapped (p<0.2). E. coli was favored when comparing mPKU with MHP variants (p<0.03) and the cell free TNT-T7 system showed a significant difference (p<0.03) only for cPKU and mPKU. For the most severe cPKU variants with no residual PAH activity, all three systems were comparable (data not shown).
ACCEPTED MANUSCRIPT Phenotype prediction for PAH missense variants Among the 34 variants investigated in this study, 12 were predicted to be associated with cPKU (APV: 0 - 2.7), 8 with mPKU (APV: 2.8 - 6.6) and 10 with the MHP (APV: 6.7 – 10.0) phenotype (Table 2). For four variants (p.I65S, p.D59Y, p.P244L and p.Y417H), no information was provided in the BIOPKU, and APV was not assigned. p.R408W was the most common variant reported in in combination with a severe
PT
null-variant (functional hemizygote) (1,845 patients), followed by p.R261Q (489 patients) and p.R243Q (405 patients) (Table 2). With exception of p.D143G and
RI
p.A313T (2 patients and 1 patient, respectively), all other variants were reported in at
PAH in vitro activity and PKU phenotypes
SC
least four cases.
NU
The PAH activity of 13 variants was investigated for the first time. These included four cPKU variants (p.F39L, p.I65S, p.L311P and p. A434D), four mPKU variants
MA
(p.A104D, p.V190A, p.A313T and p.Y417H) and 5 MHP variants (p.A47V, p.R155H, p.R297H, p.I306V and p.D415N). In vitro PAH activity was only clearly correlated with the predicted phenotype for three variants. The cPKU variants p.L311P and p.
ED
A434D presented very low PAH activity (0% and 9%, respectively), while the MHP variant p.A47V showed 123% activity. For all other variants, PAH activity was
EP T
between 21% and 77% of residual activity not clearly pointing to a phenotype. This is in the line with previous reports (Table 1). The average PAH activity of 10 MHP variants (p.A47V(APV10.0), p.D415N(APV10.0),
AC C
p.R297H(APV9.8), p.R176L(APV9.5), p.I306V(APV9.5), A403V(APV9.2), p.A300S(APV8.9), p.R155H(APV8.8), p.E76G(APV7.1), p.V190A(APV7.1)) was 44 (29)% of the wild-type activity. The highest activity (123%) was determined for the p.A47V variant and the lowest (21%) for the p.R155H variant. The 7 mPKU variants (p.R241C (APV6.0), p.R68S(APV5.5), p.R408Q(APV5.5), p.R68G(APV5.0), p.A313T(APV5.0), p.A309V(APV3.3), p.A104D (APV2.8)) presented with lower average PAH activity 39 (28)%. The p.D143G(APV5.0) variant, with 98% activity and a 40% standard error in three independent measurements, was considered to be unreliable and was excluded from the calculations.
ACCEPTED MANUSCRIPT Finally, for the 12 cPKU variants (p.A434D (APV2.6), p.G218V(APV2.5), p.V388M(APV1.9), p.L348V(APV1.7), p.F39L(APV1.4), p.R261Q(APV1.3), p.R243Q(APV0.4), p.R413P(APV0.1), p.E280K(APV0.04), p.R252W(APV0.04), p.R408W(APV0.03), p.L311P(APV0)), the average PAH activity was 25 (25)%, which was significantly lower than in the MHP and mPKU groups. Four PAH variants without information in the BIOPKU database, thus lacking APV
PT
information, were tentatively classified, based on protein expression and PAH activity, as cPKU (p.Y417H and p.D59Y) and as mPKU or MHP (p.I65S and p.P244L)
RI
(Table 2).
SC
Enzyme activity and PAH expression pattern
The PAH expression pattern and enzyme activity of 34 variants is summarized in
NU
Figure 2. Western blot analysis of expressed proteins revealed three characteristic bands (1, 2 and 3). While band 1 corresponds to the wild-type PAH, bands 2 and 3
further investigated in this study.
MA
represent degradation products of PAH. The nature of the three PAH bands was not
ED
We quantified the total protein expression of PAH and its degradation pattern (bands 1 - 3) and compared the results with in vitro enzyme activity of each of the phenotype groups. The average PAH expression (compared with wild-type PAH) was
EP T
significantly lower (p<0.004) for cPKU variants (35.6%), compared with mPKU variants (93.0%) and MHP variants (82.6%) (Figure 3A). Similarly, the average PAH activity was lower (p<0.048) for cPKU variants (19.2%), compared with mPKU
AC C
variants (40.1%) and MPH variants (46.6%) (Figure 3B). There was no significant difference in protein expression or enzyme activity between mPKU and MHP (Table 2 and Figures 3A and B). Almost no protein expression (<5%) was observed for four cPKU variants: p.L348V, p.R243Q, p.E280K and p.R252W, while for three other cPKU variants: p.R408W and p.L311P, p.R413P, expression was only slightly higher (12%, 13%, and 17% respectively) (Table 2). For the mPKU and MHP variants, PAH expression was always higher than 33%. The degradation pattern was evident for all variants, regardless of the total expression and enzyme activity (Figure 2, Table 2). The ratio of bands 1, 2 and 3 was approximately 8:1:1 for the wild-type PAH (Figure 2) and remarkably different for the p.D59Y (4:2:4), p.I65S (5:4:1), p.R68G (4:3:3), p.R241C (3:5:2), p.I306V (7:3:0),
ACCEPTED MANUSCRIPT p.A403V (6:2:2) and p.Y417H (1:9:0) variants. For the p.P244L, p.P39L and p.A104D variants, the ratio was 7:1:2 (Figure 2, Table 2). Prediction of the damaging effect of variants Two web-based tools for the prediction of variant severity, SIFT and Polyphen-2, were used. Both algorithms predicted two MHP variants (p.D415N, p.E76G) and one mPKU variant (p.A104D) to be benign, while there was no match between the two
SC
RI
possibly damaging, damaging, or deleterious (Table 3).
PT
algorithms for the p.R176L and p.I306V variants. All other variants were predicted as
DISCUSSION
NU
Different approaches can be used to predict the effect of gene variants on the metabolic phenotype. The available web-based tools for the prediction of pathogenic
MA
missense variants include FoldX, PolyPhen-2, SIFT Blink, SNPs3D and a few others. A total of 834 PAH variants and the genotypes of 4,181 PKU patients were recently analyzed using FoldX, PolyPhen-2, SIFT Blink, and SNPs3D, and a quantitative
ED
relationships were found between PAH protein stability and enzyme activity; protein stability and APV; and PAH activity and APV [12]. APV and in vitro PAH activity were
EP T
found to be best for predicting patients’ phenotypes and BH4 responsiveness [12]. As expected, PolyPhen-2 and SIFT Blink predicted most variants to be possibly damaging, damaging or deleterious in the present study. Only two MHP variants
AC C
were predicted to be benign by both algorithms. However, a benign prediction does not necessarily mean that the variant is not disease-causing. The APV expresses the frequency (%) of phenotypes in a homogeneous set of functionally hemizygous PKU genotypes and is defined as a linear scale ranging from 0 (cPKU) to 5 (mPKU) or 10 (MHP). It was developed from a non-linear arbitrary value (AV) system (power of 2) described by Guldberg et al. [19] and used in several studies, with a variable prediction outcome [20-23]. Wettstein at al. [12] correlated AV with several variant damage tools (see above) and in vitro PAH activity. There was a strong positive correlation between in vitro PAH activity and APV (r = 0.799, p<0.01), and mean PAH activity differed significantly between three APV groups p<0.001. PAH activity increased from 3.75 (±4.2%) for cPKU to 40.40 (±15.9%) for mPKU and
ACCEPTED MANUSCRIPT to 51.7 (±25.0%) for MHP. In a more recent study, genotype-phenotype associations were investigated in more than 9,000 PKU patients with various degrees of HPA and known genotypes, and a scale for APV, ranging from 0 for cPKU to 5 for mPKU and 10 for MHP, was introduced [8]. This system enabled the definition of cut-off ranges of for cPKU (0.0 - 2.7), mPKU (2.8 - 6.6) and for MHP (6.7 - 10.0) and was applied to a series of Chinese PKU patients [24]. We showed in this study that APV correlates not only with the metabolic phenotype but also with the low residual in vitro PAH
PT
activity and protein expression in cPKU variants (Figure 3).
RI
Different systems for transient expression of mutant PAH and the measurement of various forms of PAH activity have been used to address the question of how
SC
disease-associated PAH variants cause loss of PAH function [10,11,25-28]. None of the in vitro systems can reflect the hepatic in vivo situation, and reported PAH
NU
activities which differ significantly between the systems used when analyzing the same variant. Critical factors include protein regions (regulatory, catalytic or
MA
oligomerization) and affinity for the substrate Phe and cofactor BH4, all of which affect the sensitivity of the protein to structural changes. Cellular degradation of aberrant PAH protein appears to play a major role in the pathogenicity of numerous mutations
ED
[16]. Thus, in vitro expression analysis may not precisely reflect the level of the hepatic activity of certain variant enzymes, and the obtained activity levels are
EP T
probably slightly overestimated. While COS cells appeared to be favorable for differentiating between cPKU and mPKU variants, the E. coli expression system was superior in distinguishing between mPKU and MHP variants (Figure 1). It should,
AC C
however, be noted that most of the studies conducted in E. coli involved variant protein purification prior to activity measurement. As mentioned previously, not only different cell systems but also different PAH assays have been used to measure enzyme activity. Different methods, such as thinlayer chromatography (TLC) with
14
C-Phe to
14
C-Tyr conversion [27], HPLC with
fluorometric detection [29], or sophisticated mass-spectrometry methods involving the use of stable isotopes [18], have been employed to measure PAH activity at different temperatures, after different reaction times and using different substrates (BH4 or 6methyl-tetrahydropterin; 6MTHP). It is difficult to believe that these strategies produce exactly same results, but there is concordance for variants with a very low activity (see above). Since most of the PKU patients (76%) are compound heterozygotes
ACCEPTED MANUSCRIPT for PAH variants [5], interallelic complementation (positive or negative) may affect the hepatic PAH activity [14,15].Thus, the combination of the genotype-specific PAH activity with phenotype information can be more useful. PAH activity landscape is an alternative method for estimating PAH activity at different substrate and cofactor concentrations [13], and the results are probably closer to the hepatic PAH activities, as previously reported in liver biopsies of PKU patients [30].
PT
Wild-type PAH forms homo-oligomers, with an equilibrium between the tetrameric and dimeric forms. The two homotetramers and heterotetramers can be formed at
RI
different ratios, depending on the effects generated by variants (i.e., folding defects, reduced stability or low expression) [31]. Structural alterations are indicated by the
SC
increased susceptibility of the variant proteins to proteolytic degradation when expressed in E. coli [32]. The structural defects also result in defective oligomeric
NU
assembly. Thus, all of the variant PAH products may exhibit an increased tendency to form aggregates with a specific degradation pattern and reduced PAH activity. The
MA
three degradation products, which we observed in the Western blots of expressed variants, were not identified in this study but point to a specific degradation pattern for each of the variant classes. In particular, the ratio of the three proteins appears to be
ED
characteristic of some variants (Figure 2). The wide range of metabolic (and clinical) phenotypes of PKU has been shown to be
EP T
mainly determined by PAH genotype, although other factors might play a role as well. We investigated PAH and DNAJC12 expression associated with the wild-type enzyme and PAH variants and found that DNAJC12 was upregulated upon PAH
AC C
expression [33]. This observation is interesting, since DNAJC12 was recently reported to be a co-chaperone of PAH, and genetic defects in DNAJC12 result in mild hyperphenylalaninemia [34]. DNAJC12 is, together with the 70 kDa heat shock protein (HSP70) and nucleotide exchange factor (NEF), responsible for the proper folding, intracellular stability and ubiquitin-dependent degradation of PAH and may affect its expression pattern [34,35]. Ongoing investigations will hopefully provide more insights into PAH-DNAJC12 interactions. In summary, this study shows that in vitro PAH activity measurement does not always correlate with the expected metabolic phenotype, when compared in patients harboring the same variants in a hemizygous constellation. However, PAH activity and APV are still the best indicators for the damaging effect of genetic aberrations
ACCEPTED MANUSCRIPT and may be modified by the DNAJC12/HSP70 machinery and several epigenetic factors.
ACKNOWLEDGEMENTS This work is part of the RD-CONNECT initiative and was supported by the FP7-
PT
HEALTH-2012-INNOVATION-1 EU Grant No. 305444 (to NB), by the Dietmar-Hopp Foundation and by a grant from the China Scholarship Council (to NS). The authors
RI
would like to thank Kathrin Schwarz, Jana Hauke, and Peter Monostori for the
AC C
EP T
ED
MA
NU
SC
excellent technical assistance.
ACCEPTED MANUSCRIPT
PRO system
HEK293
APV
TNT-T7
Aberration
COS*
Variant
E. coli
Table 1. Overview of PAH variants and in vitro expression data.
PAH protein (%)***
Reference
PAH activity (%)**
p.D59Y p.T63P
73
c.136G>A
1.5
c.140C>T
10.0 123
c.143T>C
2.4
47
c.175G>T
0
9 92
c.187A>C
5.0 1.4
p.I65T p.I65S p.R68G p.R68S
c.194T>C
c.194T>G
2.5
c.202A>G
5.6
c.204A>T
7.0 6.3
c.227A>G
p.S87R
c.259A>C
10.0
p.T92I
c.275C>T
10.0
p.I95del
c.284_286del
p.D143G p.R155H p.R158Q p.F161S
26 27 33 48 * 32 *
40 100 98 * 25 * 24 47 88
*
22
c.428A>G
5.0
c.464G>A
8.8
52 * 98 * 21
0.2
c.473G>A
2.5
5 9 10 29 7 *
8.8
35
p.E178G
c.533A>G
7.8
31
p.E178K
c.533A>G
-
p.V190A
c.569T>C
32
6.6
100
40
3
; 96 ; 13
4
13
0
100
3
39
43
12
60
29
76
2.3.5
3 ; 100
28
25
27
4
93 ; 12 ; 100
100
100
2.3.5
c.631C>A
9.3
72
p.L212P
c.635T>C
0
17
[8] This study [22]
2.3
1
2
100 100
33
2.1
[8,12,22,23,33] [11,15,32,34,36]
1
[8,38] This study
2.3
[22]
91
[39]
100 ; 20 ; 72
[40] [22,23,39,41] This study [8]
3.5
4
1
100
3
[42] This study This study
4
9 1
100 ; 35 17 21
4
This study [37] This study [22,37] This study
1
42
1
100
3
1
[6,8,12,15,22,43] [44] [22] This study [12]
70
4
63
1
*
p.P211T
5
; 25 ; 22 ; 14
100
26
[22] This study [12,23,33,34]
100
27
2
[22,35]
2.3
27 67
[32] [22,33,34] This study
4
2.3
100
1.7
c.527G>T
p.R176L
0 62 40
46
76
0
c.482T>C
49
82
77
c.365C>A
67
*
6.7
AC C
p.P122Q
c.311C>A
*
EP T
p.E76G
p.A104D
*
PT
1.5
1
RI
p.L48S
c.117C>G
2
SC
p.A47V
2
NU
p.G46S
0
MA
p.F39L
c.1A>G
ED
p.M1V
[45] This study [39] [46]
ACCEPTED MANUSCRIPT 4.5 p.G218V
c.653G>T
p.Q226K
c.676C>A
p.V230I
c.688G>A
10.0
c.721C>T
9.5
c.722G>A
5.0
1
c.728G>A
0
0
10.0
p.V245E
c.734T>A
0
c.734T>C
10.0
p.G247V
c.740G>T
0.9
50 51 4
p.L249F
c.745C>T
5.6
51
p.L249P
c.746T>C
0
7
p.R252Q
c.754C>T
0
c.755G>A
0
c.764T>C
0
p.A259T
c.775G>A
1
p.A259V
c.776C>T
0 * 15 24
0
0
3.9
c.782G>C
2.4
c.809G>A
0
11
p.G272*
c.814G>T
0
0
p.Y277D
c.829T>G
1.1
0
p.T278I
c.833C>T
0
1
c.838G>A
0.1
c.842C>T
0
2 * 11 0 1
p.D282N
c.844G>A
0
p.I283F
c.847A>T
0.4
p.R261Q
p.R261P p.R270K
p.E280K p.P281L
p.R297H
AC C
3.9
c.782G>A
c.890G>A
9.8
[6,49]
1
4
100
1
7
62
0
1
0
8
3
0
3
3
52
48
100
[22]
100
2.3
[12,22]
1
[44] [46] [46]
45
4
[45] [8,49,53-55] This study
1
2
4
3
[29,51]
4
3
[29]
0 ; 100
2 ; 100
11 ; 100 2; 100
3.2
[29,49] [29,39,49] [12] This study
39
28 2
1
1
22 ; 20 ; 30 ; 100
2
[6,8,22,39,49] [11,15,36,43] This study [46]
2
[11,25,46]
0
[42,56] 99
*
[8,52] This study [45]
56
0
2 3 2
[8,45,50,51] This study
4
59
*
23 39 * 23 27 30 43 47 10
p.R261Q
c.781C>T
1
100
EP T
p.L255S
2
10 ; 9 ; 57
ED
p.R252W
0 ; 100
6 9 10 18 * 21 68 70
c.733G>A
-
[39] 1
0
p.V245M
c.749C>T
[22] [48] This study
*
c.731C>T
p.S250F
100
[47] 3
25 * 57 23
p.P244L
p.V245A
63
6
PT
0
52
[8,22,39] This study
3.1
RI
c.727C>T
0.5 p.R243Q
100
SC
p.R243*
63
NU
p.R241H
2
MA
p.R241C
15 * 25 101
0; 10 1
6
1
2
23
10
3
1
2
[8] 1
100 ; 0 ; 2 0
1
1
[57] [6,8,22,49] This study [22,54,58] [22] [22] This study
ACCEPTED MANUSCRIPT
c.896T>G
0.2
c.898G>T
8.9
c.916A>G
10.0
c.926C>T
3.3
c.932T>C
0
12 70 * 0
p.A313T
c.937G>A
5.0
29
p.A322G
c.965C>G
10.0
75
p.L333F
c.997C>T
-
7
p.A342T
c.1024G>A
p.L311P
p.L348V
c.1042C>G
p.S349P
c.1045T>C
0
IVS10-11G>A c.1066-11G>A 0 c.1139C>T 10.0 p.T380M p.F39del
c.116_118del
p.V388M
c.1162G>A
p.E390G
c.1169A>G
p.A395P
c.1183G>C
p.A403V
c.1208C>T
7.6
9.4
p.R413S
AC C
0
p.R408W
*
c.1222C>T
c.1237C>A
5.0
c.1238G>C
0.1
c.1243G>A
10.0
p.Y417H
c.1249T>C
5.0
44
p.Q419R
c.1256A>G
10.0
5.0 p.Y414C p.D415N
c.1241A>G
p.A434D
c.1301C>A
IVS12+1G>A
c.1315+1G>A
41
0
44
0
83
20
23
41
75
15
32 33 * 100 0 1 * 2 3 5 34 2 * 11 28 50 80 * 35
p.R413P
26
28 15 43 * 83 54 62 70
0
0
1
27
[12] This study [22] This study [8] This study [8,62] This study This study [59] [39]
3
[22] [22,24,39,61] This study
2
100 ; 0 100
1
100
2
1
[22,63,64] [8] [46]
1.3
1
100 ; 96 ; 22
85 100
2
2
[22] [11,22,24,36,57] [25,65] This study [12,15,22,39]
16
[22] 100
1
[39,66] This study
1
[8,12,15,22,60,67] This study
1 1
0 ;3
0 38
42
[39] [68] This study
1
2.3
1
100 ; 50 ; 84 72
114
100
2.3
*
70 9
1
100
1
[22] [22] This study This study [69]
*
2.6
[49,61]
1
105
0
EP T
0.9
0
24
1.3 2.5
100
*
25 33 38 0
2
39
*
4.0 3.6
12
100
PT
p.A309V
1
[8,12,22,59,60] This study
1
RI
p.I306V
1
SC
p.A300S
32 * 65 * 25
0 70 ; 91
ED
p.F299C
9
NU
p.R408Q
33 * 41 55 84 2
MA
6.2 c.890G>A
This study 0
1
[70]
ACCEPTED MANUSCRIPT
AC C
EP T
ED
MA
NU
SC
RI
PT
*This study; **Compared with the wild-type activity; **1 = COS; 2 = E.coli; 3 = TNT-T7; 4 = HEK293; 5= Yeast 6 = HepG2; 7 = PRO. APV: Allelic phenotype value (cPKU = 0 – 2.7; mPKU = 2.8 – 6.6; MHP = 6.7 – 10.0)
ACCEPTED MANUSCRIPT Table 2. Summary of PAH variants investigated, number of patients with the corresponding functionally hemizygous genotype in the BIOPKU database, allelic phenotype values, in vitro PAH activity, and PAH expression in COS-7 cells.
*
PAH band 1 ** (%) 83 (9) 33 (19) 68 (17) 78 (16) 74 (8) 80 (4) 79 (16) 70 (18) 80 (3) 77 (18) 38 (17) 30 (29) 58 (17) 68 (14) 57 (14) 66 (16) 66 (9) 25 (11) 68 (9) 79 (3) 59 (9) 78 (13) 3 (1) 63 (8) 34 (19) 2 (1) 11 (4) 1 (1) 0 (0) 10 (6) 6 (1) 2 (1) 43 (6) 7 (5) 56 (11)
PT
100 10.0 100 10.0 96 9.8 91 9.5 90 9.5 83 9.1 79 9.0 75 8.8 41 7.4 43 7.1 20 6.0 9 5.5 9 5.5 0 5.0 0 5.0 0 5.0 0 3.3 0 2.8 0 2.6 0 2.5 0 1.9 0 1.7 0 1.4 0 1.3 0 0.4 0 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RI
0 0 4 9 10 17 21 25 59 57 80 91 91 100 100 100 67 56 52 50 37 34 28 26 7 2 1 1 1 0 0 0 0 0
PAH total protein ** (%) 100 (19) 39 (22) 80 (26) 92 (29) 80 (13) 123 (15) 113 (38) 77 (21) 90 (9) 89 (23) 56 (29) 108 (42) 76 (32) 80 (16) 148 (48) 84 (34) 78 (19) 35 (18) 99 (38) 84 (8) 64 (11) 98 (17) 4 (2) 91 (34) 41 (26) 3 (2) 17 (7) 1 (1) 0 (0) 12 (7) 13 (3) 16 (2) 87 (17) 20 (10) 77 (16)
SC
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 33 44 48 50 63 66 72 74 93 98 99 99 99 100 0 0 0 0
PAH activity ** (%) 100 (8) 123 (25) 35 (32) 39 (12) 35 (22) 25 (4) 33 (21) 65 (24) 21 (9) 40 (22) 24 (5) 57 (14) 25 (21) 41 (22) 40 (17) 98 (30) 29 (13) 12 (7) 77 (19) 9 (2) 25 (15) 83 (46) 25 (24) 73 (17) 23 (19) 6 (1) 11 (7) 11 (4) 15 (9) 2 (0) 11 (8) 44 (22) 32 (2) 9 (1) 21 (1)
NU
c.140C>T 4 c.1243G>A 66 c.890G>A 23 c.527G>T 11 c.916A>G 78 c.1208C>T 281 c.898G>T 183 c.464G>A 4 c.569T>C 19 c.227A>G 7 c.721C>T 173 c.204A>T 75 c.1223G>A 55 c.202A>G 8 c.428A>G 2 c.937G>A 1 c.926C>T 21 c.311C>A 50 c.1301C>A 27 c.653G>T 10 c.1162G>A 169 c.1042C>G 94 c.117C>G 47 c.782G>A 489 c.728G>A 405 c.1283G>C 109 c.838G>A 136 c.754C>T 142 c.1222C>T 1845 c.932T>C 9 c.1249T>C 0 c.194T>G 0 c.175G>T 0 c.731C>T 0
cPKU mPKU MHP APV (%) (%) (%)
MA
*
ED
n
EP T
PAH_WT p.A47V p.D415N p.R297H p.R176L p.I306V p.A403V p.A300S p.R155H p.V190A p.E76G p.R241C p.R68S p.R408Q p.R68G p.D143G p.A313T p.A309V p.A104D p.A434D p.G218V p.V388M p.L348V p.F39L p.R261Q p.R243Q p.R413P p.E280K p.R252W p.R408W p.L311P p.Y417H p.I65S p.D59Y p.P244L
Nucleotide aberration
AC C
Variant
PAH band 2 ** (%) 9 (6) 3 (2) 6 (6) 10 (9) 3 (3) 38 (10) 16 (10) 5 (2) 5 (2) 9 (4) 9 (5) 61 (6) 10 (9) 11 (1) 43 (11) 6 (6) 6 (5) 5 (4) 11 (10) 5 (5) 4 (2) 19 (3) 1 (1) 11 (10) 5 (5) 1 (1) 5 (2) 0 (0) 0 (0) 1 (0) 5 (1) 14 (1) 37 (10) 3 (2) 6 (3)
PAH band 3 ** (%) 8 (4) 3 (1) 6 (3) 4 (4) 3 (2) 5 (1) 18 (12) 2 (1) 5 (4) 3 (1) 8 (7) 17 (7) 8 (6) 1 (1) 48 (23) 12 (12) 6 (5) 5 (3) 20 (19) 0 (0) 1 (0.4) 1 (1) 0 (0) 17 (16) 2 (2) 0 (0) 1 (1) 0 (0) 0 (0) 1 (1) 2 (1) 0 (0) 7 (1) 10 (3) 15 (2)
number of patients in the BIOPKU database (http://www.biopku.org); **mean (SD); cPKU: classic PKU; mPKU: mild PKU; MHP: mild hyperphenylalaninemia; APV: allelic phenotype value; PAH: phenylalanine hydroxylase; n.d.: not done.
ACCEPTED MANUSCRIPT
AF
SIFT
p.A47V
c.140C>T
0.03
0.03
0.12
Regulatory domain / destabilizing
p.D415N
c.1243G>A
0.44
1.00
0.00
Oligomerization domain / benign
p.R297H
c.890G>A
0.14
0.01
0.84
Catalytic domain / destabilizing
p.R176L
c.527G>T
0.15
0.02
0.00
Catalytic domain / destabilizing
p.I306V
c.916A>G
0.60
0.09
0.90
Catalytic domain/ destabilizing
p.A403V
c.1208C>T
2.40
0.00
1.00
Catalytic domain / destabilizing
p.A300S
c.898G>T
1.50
0.00
1.00
Catalytic domain / destabilizing / T1/2 -variant
p.R155H
c.464G>A
0.03
0.00
1.00
Catalytic domain/ destabilizing
p.V190A
c.569T>C
0.13
0.00
1.00
Catalytic domain/ destabilizing
p.E76G
c.227A>G
0.04
0.51
0.00
p.R241C
c.721C>T
1.10
0.02
1.00
p.R68S
c.204A>T
0.60
0.00
p.R408Q
c.1223G>A
0.63
0.01
p.R68G
c.202A>G
0.05
0.00
1.00
Regulatory domain/ destabilizing
p. D143G
c.428A>G
0.01
0.01
0.15
Catalytic domain/ destabilizing
p.A313T
c.937G>A
0.01
0.00
1.00
Catalytic domain/ destabilizing/ affects splicing
p.A309V
c.926C>T
0.22
0.01
1.00
Catalytic domain/ destabilizing
AC C
p.A104D
Polyphen2 Comment
NU
SC
RI
PT
cDNA
EP T
Protein
ED
Table 3. Alleles frequency (AF), SIFT and Polyphewn-2 damaging effect prediction values and some properties of variants investigated.
Regulatory domain / benign
MA
Catalytic domain/ destabilizing
1.00
Regulatory domain/ destabilizing
1.00
Catalytic domain/ destabilizing
c.311C>A
0.42
0.63
0.00
Regulatory domain / benign-destabilizing
c.1301C>A
0.14
0.01
1.00
Oligomerization domain/ destabilizing
c.653G>T
0.11
0.00
1.00
Catalytic domain/ destabilizing
p.V388M
c.1162G>A
1.80
0.00
1.00
Catalytic domain/ destabilizing / Km-variant
p.L348V
c.1042C>G
0.90
0.00
1.00
Catalytic domain/ destabilizing
p.F39L
c.117C>G
0.42
0.01
1.00
Regulatory domain/ destabilizing
p.R261Q
c.782G>A
5.50
0.00
1.00
Catalytic domain/ destabilizing / CBR1
p.R243Q
c.728G>A
2.60
0.00
1.00
Catalytic domain/ destabilizing
p.R413P
c.1283G>C
0.69
0.00
0.81
Oligomerization domain/ destabilizing
p.A434D p.G218V
ACCEPTED MANUSCRIPT c.838G>A
1.30
0.00
1.00
Catalytic domain/ destabilizing / CBR2
p.R252W
c.754C>T
1.40
0.00
1.00
Catalytic domain/ destabilizing / CBR1
p.R408W
c.1222C>T
19.70 0.00
1.00
Catalytic domain/ destabilizing / Folding variant
p.L311P
c.932T>C
0.17
0.00
1.00
Catalytic domain/ destabilizing
p.Y417H
c.1249T>C
0.03
0.00
1.00
Oligomerization domain / destabilizing / Folding variant
p.I65S
c.194T>G
0.01
0.00
1.00
Regulatory domain/ destabilizing / Folding variant
p.D59Y
c.175G>T
0.01
0.03
0.20
Regulatory domain/ destabilizing
p.P244L
c.731C>T
0.09
0.00
1.00
Catalytic domain/ destabilizing
RI
PT
p.E280K
AC C
EP T
ED
MA
NU
SC
SIFT: <0.06 = deleterious; Polyphen-2: 0-0.15 = benign, 0.15-1.0 = possibly damaging, 0.85-1.0 damaging; A; F: allele frequency (%)CBR1: cofactor (BH4)-binding region 1; CBR2: cofactor (BH4)binding region 2; T1/2: half-time
ACCEPTED MANUSCRIPT REFERENCES [1]
Blau N., Van Spronsen F. J., Levy H.L., Phenylketonuria. Lancet 376 (2010) 1417-1427.
[2]
Vockley J., Andersson H. C., Antshel K. M., Braverman N. E., Burton B. K., Frazier D. M., Mitchell J., Smith W. E., Thompson B. H., Berry S. A., Phenylalanine hydroxylase deficiency: diagnosis and management guideline.
[3]
PT
Genet Med 16 (2014) 188-200.
Camp K. M., Parisi M. A., Acosta P. B., Berry G. T., Bilder D. A., Blau N.,
RI
Bodamer O. A., Brosco J. P., Brown C. S., Burlina A. B., Burton B. K., Chang C. S., Coates P. M., Cunningham A. C., Dobrowolski S. F., Ferguson J. H.,
SC
Franklin T. D., Frazier D. M., Grange D. K., Greene C. L., Groft S. C., Harding C. O., Howell R. R., Huntington K. L., Hyatt-Knorr H. D., Jevaji I. P., Levy H.
NU
L., Lichter-Konecki U., Lindegren M. L., Lloyd-Puryear M. A., Matalon K., Macdonald A., McPheeters M. L., Mitchell J. J., Mofidi S., Moseley K. D.,
MA
Mueller C. M., Mulberg A. E., Nerurkar L. S., Ogata B. N., Pariser A. R., Prasad S., Pridjian G., Rasmussen S. A., Reddy U. M., Rohr F. J., Singh R. H., Sirrs S. M., Stremer S. E., Tagle D. A., Thompson S. M., Urv T. K., Utz J.
ED
R., van Spronsen F., Vockley J., Waisbren S. E., Weglicki L. S., White D. A., Whitley C. B., Wilfond B. S., Yannicelli S., Young J. M., Phenylketonuria
EP T
scientific review conference: State of the science and future research needs. Mol Genet Metab 112 (2014) 87-122. [4]
Blau N., Shen N., Carducci C., Molecular genetics and diagnosis of
[5]
AC C
phenylketonuria: state of the art. Expert Rev Mol Diagn 14 (2014) 655-671. Blau N., Genetics of Phenylketonuria: Then and Now. Hum Mutat 37 (2016) 508-15. [6]
Desviat L. R., Perez B., Gamez A., Sanchez A., Garcia M. J., Martinez-Pardo M., Marchante C., Boveda D., Baldellou A., Arena J., Sanjurjo P., Fernandez A., Cabello M. L., Ugarte M., Genetic and phenotypic aspects of phenylalanine hydroxylase deficiency in Spain: molecular survey by regions. Eur J Hum Genet 7 (1999) 386-92.
[7]
Daniele A., Scala I., Cardillo G., Pennino C., Ungaro C., Sibilio M., Parenti G., Esposito L., Zagari A., Andria G., Salvatore F., Functional and structural characterization of novel mutations and genotype-phenotype correlation in 51
ACCEPTED MANUSCRIPT phenylalanine hydroxylase deficient families from Southern Italy. FEBS J 276 (2009) 2048-59. [8]
Garbade S., Shen N., Himmelreich N., Haas D., Trefz F.K., Hoffmann G.F., Burgard P., Blau N., Allelic phenotype values: a model for genotype-based phenotype prediction in phenylketonuria. Genet Med (2018) DOI: 10.1038/s41436-018-0081-x.
[9]
Okano Y., Eisensmith R. C., Güttler F., Lichter-Konecki U., Konecki D. S.,
PT
Trefz F. K., Dasovich M., Wang T., Henriksen K., Lou H., et al., Molecular basis of phenotypic heterogeneity in phenylketonuria. N Engl J Med 324
Waters P. J., Parniak M. A., Nowacki P., Scriver C. R., In vitro expression
SC
[10]
RI
(1991) 1232-8.
analysis of mutations in phenylalanine hydroxylase: linking genotype to
[11]
NU
phenotype and structure to function. Hum Mutat 11 (1998) 4-17. Pey A. L., Desviat L. R., Gamez A., Ugarte M., Perez B., Phenylketonuria: genotype-phenotype correlations based on expression analysis of structural
[12]
MA
and functional mutations in PAH. Hum Mutat 21 (2003) 370-8. Wettstein S., Underhaug J., Perez B., Marsden B. D., Yue W. W., Martinez A.,
ED
Blau N., Linking genotypes database with locus-specific database and genotype-phenotype correlation in phenylketonuria. Eur J Hum Genet 23 (2015) 302–309.
Danecka M. K., Woidy M., Zschocke J., Feillet F., Muntau A. C., Gersting S.
EP T
[13]
W., Mapping the functional landscape of frequent phenylalanine hydroxylase (PAH) genotypes promotes personalised medicine in phenylketonuria. J Med
[14]
AC C
Genet 52 (2015) 175-85.
Leandro J., Nascimento C., de Almeida I. T., Leandro P., Co-expression of different subunits of human phenylalanine hydroxylase: evidence of negative interallelic complementation. Biochim Biophys Acta 1762 (2006) 544-50.
[15]
Shen N., Heintz C., Thiel C., Okun J. G., Hoffmann G. F., Blau N., Coexpression of phenylalanine hydroxylase variants and effects of interallelic complementation on in vitro enzyme activity and genotype-phenotype correlation. Mol Genet Metab 117 (2015) 328–335.
[16]
Waters P. J., How PAH gene mutations cause hyper-phenylalaninemia and why mechanism matters: insights from in vitro expression. Hum Mutat 21 (2003) 357-69.
ACCEPTED MANUSCRIPT [17]
Aguado C., Perez B., Ugarte M., Desviat L. R., Analysis of the effect of tetrahydrobiopterin on PAH gene expression in hepatoma cells. FEBS Lett 580 (2006) 1697-701.
[18]
Heintz C., Troxler H., Martinez A., Thony B., Blau N., Quantification of phenylalanine hydroxylase activity by isotope-dilution liquid chromatographyelectrospray ionization tandem mass spectrometry. Mol Genet Metab 105 (2012) 559-65. Guldberg P., Rey F., Zschocke J., Romano V., Francois B., Michiels L., Ullrich
PT
[19]
K., Hoffmann G. F., Burgard P., Schmidt H., Meli C., Riva E., Dianzani I.,
RI
Ponzone A., Rey J., Güttler F., A European multicenter study of phenylalanine
SC
hydroxylase deficiency: classification of 105 mutations and a general system for genotype-based prediction of metabolic phenotype. Am J Hum Genet 63
[20]
NU
(1998) 71-9.
Kasnauskiene J., Cimbalistiene L., Kucinskas V., Validation of PAH genotypebased predictions of metabolic phenylalanine hydroxylase deficiency
MA
phenotype: investigation of PKU/MHP patients from Lithuania. Med Sci Monit 9 (2003) CR142-6.
Santos L. L., Fonseca C. G., Starling A. L., Januario J. N., Aguiar M. J.,
ED
[21]
Peixoto M. G., Carvalho M. R., Variations in genotype-phenotype correlations in phenylketonuria patients. Genet Mol Res 9 (2010) 1-8. Zhu T., Ye J., Han L., Qiu W., Zhang H., Liang L., Gu X., Variations in
EP T
[22]
genotype-phenotype correlations in Phenylalanine hydroxylase deficiency in Chinese Han Population. Gene 529 (2013) 80-87. Zhu T., Ye J., Han L., Qiu W., Zhang H., Liang L., Gu X., The Predictive Value
AC C
[23]
of Genetic Analyses in the Diagnosis of Tetrahydrobiopterin (BH4)Responsiveness in Chinese Phenylalanine Hydroxylase Deficiency Patients. Sci Rep 7 (2017) 6762. [24]
Wang R., Shen N., Ye J., Han L., Qiu W., Zhang H., Liang L., Sun Y., Fan Y., Wang L., Wang Y., Gong Z., Liu H., Wang J., Yan H., Blau N., Gu X., Yu Y., Mutation spectrum of hyperphenylalaninemia candidate genes and the genotype-phenotype correlation in the Chinese population. Clin Chim Acta 481 (2018) 132-138.
[25]
Gjetting T., Petersen M., Guldberg P., Guttler F., In vitro expression of 34 naturally occurring mutant variants of phenylalanine hydroxylase: correlation
ACCEPTED MANUSCRIPT with metabolic phenotypes and susceptibility toward protein aggregation. Mol Genet Metab 72 (2001) 132-43. [26]
Waters P. J., Scriver C. R., Parniak M. A., Homomeric and heteromeric interactions between wild-type and mutant Phenylalanine hydroxylase subunits: Evaluation of two-hybrid approaches for functional analysis of mutations causing hyperphenylalaninemia. Mol Genet Metab 73 (2001) 230-8.
[27]
Gamez A., Perez B., Ugarte M., Desviat L. R., Expression analysis of
PT
phenylketonuria mutations. Effect on folding and stability of the phenylalanine hydroxylase protein. J Biol Chem 275 (2000) 29737-42.
Leandro P., Lechner M. C., Tavares de Almeida I., Konecki D., Glycerol
RI
[28]
SC
increases the yield and activity of human phenylalanine hydroxylase mutant enzymes produced in a prokaryotic expression system. Mol Genet Metab 73
[29]
NU
(2001) 173-8.
Doskeland A. P., Doskeland S. O., Ogreid D., Flatmark T., The effect of ligands of phenylalanine 4-monooxygenase on the cAMP-dependent
MA
phosphorylation of the enzyme. The Journal of biological chemistry 259 (1984) 11242-8.
Bartholomé K., Lutz P., Bickel H., Determination of phenylalanine hydroxylase
ED
[30]
activity in patients with phenylketonuria and hyperphenylalaninemia. Pediatr Res 9 (1975) 899.
Erlandsen H., Stevens R. C., A structural hypothesis for BH4 responsiveness
EP T
[31]
in patients with mild forms of hyperphenylalaninaemia and phenylketonuria. J Inherit Metab Dis 24 (2001) 213-30. Bjorgo E., Knappskog P. M., Martinez A., Stevens R. C., Flatmark T., Partial
AC C
[32]
characterization and three-dimensional-structural localization of eight mutations in exon 7 of the human phenylalanine hydroxylase gene associated with phenylketonuria. Eur J Biochem 257 (1998) 1-10. [33]
Himmelreich N., Hoffmann G.F., Blau N., Expression pattern of phenylalanine hydroxylase variants is regulated by co-chaperone DNAJC12. Mol Genet Metab 123 (2018) 210.
[34]
Anikster Y., Haack T. B., Vilboux T., Pode-Shakked B., Thony B., Shen N., Guarani V., Meissner T., Mayatepek E., Trefz F. K., Marek-Yagel D., Martinez A., Huttlin E. L., Paulo J. A., Berutti R., Benoist J. F., Imbard A., Dorboz I., Heimer G., Landau Y., Ziv-Strasser L., Malicdan M. C., Gemperle-Britschgi C.,
ACCEPTED MANUSCRIPT Cremer K., Engels H., Meili D., Keller I., Bruggmann R., Strom T. M., Meitinger T., Mullikin J. C., Schwartz G., Ben-Zeev B., Gahl W. A., Harper J. W., Blau N., Hoffmann G. F., Prokisch H., Opladen T., Schiff M., Biallelic Mutations in DNAJC12 Cause Hyperphenylalaninemia, Dystonia, and Intellectual Disability. Am J Hum Genet 100 (2017) 257-266. Blau N., Martinez A., Hoffmann G. F., Thony B., DNAJC12 deficiency: A new strategy in the diagnosis of hyperphenylalaninemias. Mol Genet Metab 123
EP T
ED
MA
NU
SC
RI
PT
(2018) 1-5.
AC C
[35]
ACCEPTED MANUSCRIPT LEGEND TO FIGURES Figure 1. Comparison of the PAH activity (mean, 25th - 75th percentile, min, max) of variants associated with three metabolic phenotypes (cPKU, mPKU and MHP) transiently expressed in (A) COS cells, (B) E. coli and (C) in TNT-T7 cells. The assignment of variants to metabolic phenotypes was carried out using APV (see Materials and Methods). Average PAH activity was calculated when there was more
PT
than one report for the same variant. The unpaired t-test was considered statistically significant when p<0.02. cPKU: classic phenylketonuria; mPKU: mild
RI
phenylketonuria; MHP: mild hyperphenylalaninemia.
SC
Figure 2. PAH expression patterns and PAH activities of 34 variants transiently expressed variants in COS-7 cells. Gray-shaded bares represent the amount (%) of
NU
PAH protein (bands 1, 2 and 3), and open bars represent PAH activity (average of three measurements in both cases). For details, see the text and Table 2.
MA
Figure 3. Comparison of (A) PAH protein expression and (B) PAH activity (mean, 25th - 75th percentile, min, max) of 34 variants in COS-7 cells. The assignment of variants to metabolic phenotypes was carried out using APV (see Materials and
ED
Methods). The unpaired t-test was considered statistically significant when p<0.02. cPKU: classic phenylketonuria; mPKU: mild phenylketonuria; MHP: mild
AC C
EP T
hyperphenylalaninemia.
Figure 1
Figure 2
Figure 3