Transferrin Isoforms, Old but New Biomarkers in Hereditary Fructose Intolerance

Journal of Clinical Medicine Article Transferrin Isoforms, Old but New Biomarkers in Hereditary Fructose Intolerance Ainara Cano 1, Carlos Alcalde 2, Amaya Belanger-Quintana 3, Elvira Cañedo-Villarroya 4, Leticia Ceberio 5, Silvia Chumillas-Calzada 6, Patricia Correcher 7, María Luz Couce 8, Dolores García-Arenas 9, Igor Gómez 10, Tomás Hernández 11, Elsa Izquierdo-García12, Dámaris Martínez Chicano 9, Montserrat Morales 6, Consuelo Pedrón-Giner 13, Estrella Petrina Jáuregui 14, Luis Peña-Quintana 15 , Paula Sánchez-Pintos 8, Juliana Serrano-Nieto 16, María Unceta Suarez 17, Isidro Vitoria Miñana 7and Javier de las Heras 1,18,19,*   Citation: Cano, A.; Alcalde, C.; Belanger-Quintana, A.; Cañedo-Villarroya, E.; Ceberio, L.; Chumillas-Calzada, S.; Correcher, P.; Couce, M.L.; García-Arenas, D.; Gómez, I.; et al. Transferrin Isoforms, Old but New Biomarkers in Hereditary Fructose Intolerance. J. Clin. Med. 2021,10, 2932. https://doi.org/ 10.3390/jcm10132932 Academic Editor: John Griffith Jones Received: 21 May 2021 Accepted: 28 June 2021 Published: 30 June 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1Biocruces Bizkaia Health Research Institute, 48093 Barakaldo, Spain; [email protected] 2Paediatrics Unit, Río Hortega University Hospital, 47012 Valladolid, Spain; [email protected] 3Metabolic Diseases Unit, Department of Paediatrics, Ramon y Cajal Hospital, 28034 Madrid, Spain; [email protected] 4Department of Metabolism Diseases and Nutrition, Niño Jesús University Children’s Hospital, 28009 Madrid, Spain; [email protected] 5Internal Medicine Service, Cruces University Hospital, 48903 Barakaldo, Spain; [email protected] 6 12 de Octubre University Hospital, CIBERER, 28041 Madrid, Spain; [email protected] (S.C.-C.); [email protected] (M.M.) 7Nutrition and Metabolic diseases Unit, La Fe University Hospital, 46026 Valencia, Spain; [email protected] (P.C.); [email protected] (I.V.M.) 8Unit of Diagnosis and Treatment of Congenital Metabolic Diseases, Department of Paediatrics, IDIS-Health Research Institute of Santiago de Compostela, CIBERER, MetabERN, Santiago de Compostela University Clinical Hospital, 15704 Santiago de Compostela, Spain; [email protected] (M.L.C.); [email protected] (P.S.-P.) 9Department of Paediatric Gastroenterology, Hepatology and Nutrition, Sant Joan de Déu Hospital, 08950 Barcelona, Spain; dgar[email protected]g (D.G.-A.); [email protected]g (D.M.C.) 10 Araba University Hospital, 01009 Gasteiz, Spain; igor[email protected] 11 Paediatric Service, Albacete University Hospital, 02006 Castilla-La Mancha, Spain; [email protected] 12 Pharmacy Department, Infanta Leonor University Hospital, 28031 Madrid, Spain; [email protected]g 13 Gastroenterology and Nutrition Section, Niño Jesús University Children’s Hospital, 28009 Madrid, Spain; [email protected]g 14 Clinical Nutrition Section, Navarra University Hospital, 31008 Pamplona, Spain; [email protected] 15 Pediatric Gastroenterology, Hepatology and Nutrition Unit, Mother and Child Insular University Hospital Complex, Asociación Canaria para la Investigación Pediátrica (ACIP), CIBEROBN, University Institute for Research in Biomedical and Health Sciences, University of Las Palmas de Gran Canaria, 35016 Las Palmas de Gran Canaria, Spain; [email protected] 16 Paediatric Service, Málaga Regional University Hospital (HRU), 29010 Málaga, Spain; [email protected] 17 Biochemistry Laboratory, Metabolism Area, Cruces University Hospital, 48903 Barakaldo, Spain; [email protected] 18 Division of Paediatric Metabolism, CIBERER, Cruces University Hospital, 48093 Barakaldo, Spain 19 Department of Paediatrics, University of the Basque Country (UPV/EHU), 48940 Leioa, Spain *Correspondence: javier[email protected] Abstract: Hereditary Fructose Intolerance (HFI) is an autosomal recessive inborn error of metabolism characterised by the deficiency of the hepatic enzyme aldolase B. Its treatment consists in adopting a fructose-, sucrose-, and sorbitol (FSS)-restrictive diet for life. Untreated HFI patients present an abnormal transferrin (Tf) glycosylation pattern due to the inhibition of mannose-6-phosphate isomerase by fructose-1-phosphate. Hence, elevated serum carbohydrate-deficient Tf (CDT) may allow the prompt detection of HFI. The CDT values improve when an FSS-restrictive diet is followed; however, previous data on CDT and fructose intake correlation are inconsistent. Therefore, we examined the complete serum sialoTf profile and correlated it with FSS dietary intake and with J. Clin. Med. 2021,10, 2932. https://doi.org/10.3390/jcm10132932 https://www.mdpi.com/journal/jcm J. Clin. Med. 2021,10, 2932 2 of 14 hepatic parameters in a cohort of paediatric and adult fructosemic patients. To do so, the profiles of serum sialoTf from genetically diagnosed HFI patients on an FSS-restricted diet (n= 37) and their age-, sexand body mass index-paired controls (n= 32) were analysed by capillary zone electrophoresis. We found that in HFI patients, asialoTf correlated with dietary intake of sucrose (R = 0.575, p< 0.001 ) and FSS (R = 0.475, p= 0.008), and that pentasialoTf+hexasialoTf negatively correlated with dietary intake of fructose (R = − 0.386, p= 0.024) and FSS (R = − 0.400, p= 0.019). In addition, the tetrasialoTf/disialoTf ratio truthfully differentiated treated HFI patients from healthy controls, with an area under the ROC curve (AUROC) of 0.97, 92% sensitivity, 94% specificity and 93% accuracy. Keywords: hereditary fructose intolerance; fructose; sucrose; sorbitol; diet; sialotransferrin profile; biomarker; aldolase B 1. Introduction Hereditary fructose intolerance (HFI; OMIM 229600) is an autosomal recessive inborn error of metabolism [ 1 ]. HFI was first reported in 1956 by Chambers and Pratt [ 2 ]. It is characterized by deficiency of the enzyme fructose-1,6-bisphosphate aldolase (aldolase B; E.C. 4.1.2.13), which is predominantly expressed in the liver, kidney and small intestine [ 3 ]. Aldolase B catalyses different reactions including cleavage of fructose-1-phosphate (F1P) and reversible cleavage of fructose-1,6-bisphosphate (FBP) into glyceraldehyde phosphate and dihydroxyacetone phosphate (DHAP) [ 1 ]. Therefore, this enzyme plays a key role in the control of fructose and glucose metabolism, regulating both glycolysis and gluconeogenesis. HFI is caused by homozygous or compound heterozygous mutations in the aldolase B gene (ALDOB; 612724) on chromosome 9q31 [ 4 ]. Based on the carrier frequency of the most common mutations in neonates, it has been estimated that the prevalence of HFI is around 1 in 26,000 live births in Europe [ 5 ] and 1 in 20,000 births in the US [ 6 ], yet many authors agree that it may be significantly underdiagnosed. Symptoms in HFI patients are usually initiated five or six months after birth due to the introduction of complementary feeding in the infant, who reacts with a variety of clinical signs such as failure to thrive, accompanied by vomiting, abdominal pain and acute liver failure [ 7 ]. HFI is also characterised by a set of metabolic alterations that include hypoglycaemia, metabolic acidosis, hypophosphatemia, hyperuricemia, hypermagnesemia and hyperalaninaemia after fructose loading [ 1 ]. Early diagnosis of HFI prevents the continued intake of fructose, which would otherwise lead to renal and hepatic seizures, coma and even death [ 1 ]. It allows dietary treatment, consisting mainly of a fructose-, sucrose-, and sorbitol (FSS)-restricted diet for life [ 1 ]. Recurrent symptoms make the diagnosis possible in childhood; however, many patients remain undiagnosed until adulthood [8]. Even though fructosemic patients are believed to remain quite healthy when kept on a fructose-restricted diet, compliance with this regime is often difficult, especially when taking into consideration the overload of fructose in our alimentary routines observed in recent years. There is a growing use of fructose, as it is widely used as a food additive, and small amounts of this sugar are hidden in many foods. Thus, it is not surprising that HFI patients develop previously unreported complications, such as liver steatosis [ 7 , 9 ] or signs of proximal tubular dysfunction [10,11]. Whereas easy detection of galactosemia, another error of carbohydrate metabolism, can be performed by means of metabolic analysis using dried blood spot (DBS) testing [ 12 ], no simple metabolic test is available for the rapid identification of HFI. The diagnosis of HFI used to be based on the clinical features of fructose intolerance, liver biopsy to assess aldolase B activity, and/or a fructose challenge test [ 13 ]. Nowadays, the diagnosis of HFI consists in identifying either biallelic pathogenic variants of ALDOB by molecular genetic testing or deficient hepatic aldolase B activity from a liver biopsy [ 13 ]. Despite liver biopsy being very specific, it is invasive and costly. Therefore, the high sensitivity and J. Clin. Med. 2021,10, 2932 3 of 14 the non-invasive nature of ALDOB genetic testing make it the preferred method for HFI diagnosis [13]. In the mid-1970s, a temporary change in the transferrin (Tf) glycoform profile in the serum and cerebrospinal fluid was demonstrated to be associated with sustained heavy alcohol consumption [ 14 ]. The abnormal serum Tf pattern, initially presenting as an increased amount of Tf bands with an isoelectric point at or above 5.7 (corresponding to asialo-, monosialo-, and disialo-Tf) in isoelectric focusing (IEF), improved on abstinence, with a half-life of ~10 days [ 15 ]. This protein fraction, later named serum carbohydrate-deficient Tf (CDT), was suggested as a specific biomarker, and its measurement was proposed to identify sustained heavy alcohol consumption and monitor abstinence during treatment [ 16 ]. In vitro studies using cultured rat liver hepatocytes suggest that acetaldehyde resulting from alcohol dehydrogenase oxidation of ethanol causes the disruption of endoplasmic reticulum function and thus interferes with glycosylation [ 17 ]. At present, loss of sialylation on serum Tf is used as a screening test both for chronic alcohol consumption and for congenital disorders of glycosylation (CDG) [15,16]. Serum Tf is a glycoprotein that transports iron (Fe 3+ ), which is mainly synthesised and metabolised in liver hepatocytes. Two complex chains of oligosaccharides, which vary in their degree of ramification, form the glycolic portion of the glycoprotein; each of them can have two or three external chains or anthems, with a sialic acid residue in the terminal position (Figure 1). Then, Tf has different isoforms depending on the number of sialic acid residues present on its oligosaccharide chain: asialo-, monosialo-, disialo-, trisialo-, pentasialoand hexasialo-Tf. As stated by Helander et al. [ 15 ], Tf shows natural microheterogeneity, owing to variations in iron load, amino acid sequence (i.e., genetic variants) and the structure of the two N-linked oligosaccharides (N-glycans). In human serum, tetrasialoTf is usually the most abundant glycoform (~80%), followed by pentasialo- (~14%), trisialo- (~4%), disialo- (~1%) and hexasialoTf (~1%) [18,19]. Previous studies have shown that untreated fructosemic patients have an abnormal Tf glycosylation pattern as a consequence of F1P-mediated competitive inhibition of mannose6-phosphate isomerase (MPI) [ 20 ]. In particular, patients with HFI present a Tf isoelectric focussing (IEF) type Ib pattern, in which defects in the synthesis and assembly of the glycans occur, indicating that HFI is a secondary CDG syndrome [ 21 ]. Therefore, as glycosylation of Tf is a measure of intrahepatic F1P concentrations, elevated serum CDT values allow the quick detection of HFI [ 22 ]. This biomarker is valuable for detecting the persistence of some abnormalities caused by trace amounts of fructose ingestion and/or non-adherence to the prescribed diet in HFI patients. Moreover, CDT values improve after 2–4 weeks when a fructose-restricted diet is taken, and then, CDT measurements are recommended for HFI diagnosis and treatment monitoring [22]. However, there is no agreement among authors regarding the correlation of the sialoTf profile with dietary fructose consumption, nor about which isoform of Tf is more valuable for monitoring fructosemic patients [ 21 – 25 ]. Thus, the objective of the present study was to determine the usefulness of the entire sialoTf profile for monitoring an FSS-restricted diet in HFI patients by assessing the correlation between the different sialoTf isoforms and dietary fructose, sucrose, and sorbitol intake. J. Clin. Med. 2021,10, 2932 4 of 14 J. Clin. Med. 2021, 10, x FOR PEER REVIEW 3 of 13 assess aldolase B activity, and/or a fructose challenge test [13]. Nowadays, the diagnosis of HFI consists in identifying either biallelic pathogenic variants of ALDOB by molecular genetic testing or deficient hepatic aldolase B activity from a liver biopsy [13]. Despite liver biopsy being very specific, it is invasive and costly. Therefore, the high sensitivity and the non-invasive nature of ALDOB genetic testing make it the preferred method for HFI diagnosis [13]. Figure 1. The link between fructose metabolism and the profile of sialotransferrins in hereditary fructose intolerance. Dietary sorbitol and sucrose, i.e., a disaccharide formed by glucose and fructose, are precursors of fructose. The catabolic pathway of dietary fructose, sucrose and sorbitol is altered in patients with hereditary fructose intolerance (HFI). In aldolase B deficiency, the catabolism of fructose-1-P (F1P) is impaired (red bar), and this molecule is accumulated largely in the liver of HFI patients. Consequently, HFI patients have an abnormal transferrin (Tf) glycosylation pattern because of F1P-mediated competitive inhibition of mannose-6-phosphate isomerase (MPI). Tf exhibits different isoforms depending on the number of sialic acid residues present on its oligosaccharide chain; asialo-, monosialo-, disialo-, trisialo-, pentasialo- , and hexasialo-Tf. The sum of asialo-, monosialo-, and disialo-Tf is called carbohydrate-deficient Tf (CDT). In the mid-1970s, a temporary change in the transferrin (Tf) glycoform profile in the serum and cerebrospinal fluid was demonstrated to be associated with sustained heavy alcohol consumption [14]. The abnormal serum Tf pattern, initially presenting as an increased amount of Tf bands with an isoelectric point at or above 5.7 (corresponding to asialo-, monosialo-, and disialo-Tf) in isoelectric focusing (IEF), improved on abstinence, with a half-life of ~10 days [15]. This protein fraction, later named serum carbohydratedeficient Tf (CDT), was suggested as a specific biomarker, and its measurement was proposed to identify sustained heavy alcohol consumption and monitor abstinence during treatment [16]. In vitro studies using cultured rat liver hepatocytes suggest that acetaldehyde resulting from alcohol dehydrogenase oxidation of ethanol causes the disruption of endoplasmic reticulum function and thus interferes with glycosylation [17]. At present, Figure 1. The link between fructose metabolism and the profile of sialotransferrins in hereditary fructose intolerance. Dietary sorbitol and sucrose, i.e., a disaccharide formed by glucose and fructose, are precursors of fructose. The catabolic pathway of dietary fructose, sucrose and sorbitol is altered in patients with hereditary fructose intolerance (HFI). In aldolase B deficiency, the catabolism of fructose1-P (F1P) is impaired (red bar), and this molecule is accumulated largely in the liver of HFI patients. Consequently, HFI patients have an abnormal transferrin (Tf) glycosylation pattern because of F1Pmediated competitive inhibition of mannose-6-phosphate isomerase (MPI). Tf exhibits different isoforms depending on the number of sialic acid residues present on its oligosaccharide chain; asialo-, monosialo-, disialo-, trisialo-, pentasialo-, and hexasialo-Tf. The sum of asialo-, monosialo-, and disialo-Tf is called carbohydrate-deficient Tf (CDT). 2. Experimental Section 2.1. Participants A cross-sectional study was conducted from October 2019 to November 2020. The study population comprised 37 genetically diagnosed HFI patients from 31 unrelated families. The recruited fructosemic patients had been on dietary treatment with fructose, sucrose and sorbitol exclusion for at least two years. Their age-, sex-, and BMI-matched healthy controls (n= 32) were also studied. Eleven Spanish hospitals participated in the study: Cruces University Hospital, Basque Country [host]; Araba University Hospital, Basque Country; Navarra University Hospital; Navarra, 12 de Octubre University Hospital, Madrid; Niño Jesús University Children’s Hospital, Madrid; Ramón y Cajal University Hospital, Madrid; La Fe University Hospital, Valencian Community; Málaga Regional University Hospital, Andalusia; Santiago de Compostela University Clinical Hospital, Galicia; Río Hortega University Hospital, Castile and Leon; Mother and Child Insular University Hospital complex, Canary Islands. The study protocol was performed according to the ethical guidelines of the revised 1975 Declaration of Helsinki [ 26 ] and approved by the Research Ethics Committee of the J. Clin. Med. 2021,10, 2932 5 of 14 Basque Country (CEIm-E), ethic approval code: PI2019072. Written informed consent was obtained from parents or legal guardians of children (below 18 years of age) and adult study participants. 2.2. Study Visits Study visits were performed at the Cruces University Hospital, Spain. Blood samples were collected at the same time, at 8:30 a.m., after at least 8 h of fasting. Patients were weighed, and their height and abdominal perimeters were measured. Standing height was measured with a wall-mounted stadiometer, and patients were weighed barefoot, to the nearest 100 g, with digital scales. 2.3. Genotype Analysis HFI patients were genetically diagnosed in their hospitals of origin by Sanger sequencing of the ALDOB gene [27] or by NGS gene panel screening. 2.4. Biochemical Analyses Concentrations of iron, Tf and ferritin in the plasma were determined by routine clinical techniques. 2.5. Transferrin Quantification The Tf profile was determined in serum by capillary zone electrophoresis (CZE) using a commercially available system (Sebia, Capillarys 2TM, France) in 37 HFI patients and in 32 healthy volunteers as described before [ 18 ]. Briefly, the Tf isoforms were separated into five fractions according to their degree of sialylation, i.e., asialoTf (non-sialylated form), disialo-, trisialo-, tetrasialoand the sum of pentasialoand hexasialo-Tf, and expressed as a percentage of the total sialoTf. Data analysis was performed with the software package Phoresis 8.6.3 (Sebia, France). The sum of the asialoTf and disialoTf fractions was expressed as a CDT percentage. Additionally, the tetrasialoTf-to-disialoTf ratio was calculated. 2.6. Assessment of Diet Dietary information was collected in a self-administered nutritional record of dietary intake on three days (two during the week and one on the weekend) in 34 of the 37 HFI patients, with a special emphasis on determining fructose, sucrose and sorbitol dietary intake. Once completed, the different carbohydrate composition, expressed in mg per day, was calculated using the Nutritional Calculation DIAL Program (Version 3.10.5.0), of ALCEGENI INERIA [ 28 ]. In addition, to complete the information of fructose and sorbitol dietary intake, the following databases were consulted: Australian Food Composition Database—Release 1.0 [ 29 ], Danish Frida Food Database [ 30 ], and German Food Composition and Nutrition Tables [ 31 ]. Data obtained on fermentable mono-, di-, oligosaccharides and polyols (FODMAP) were also used [32]. 2.7. Statistical Analysis Continuous variables were represented as mean ± standard deviation and range. Fructosemic patients on an FSS-restrictive diet (n= 37) were compared with their age-, sex-, and BMI-marched controls (n= 32) by unpaired Student’s t-test. The Mann–Whitney U test was used to compare continuous variables between HFI patients with and without asialoTf expression. the Wallis test was used to compare continuous variables among the FFS-intake tertiles. p-values were based on two-tailed comparisons as appropriate, and those less than 0.05 were considered to indicate a statistically significant difference. Pearson correlation was used to assess bivariate relationships between variables. The diagnostic accuracy of the model that differentiates between HFI patients and nonHFI patients was assessed by the area under the ROC curve (AUC) by exporting the obtained data to the MetaboAnalyst software package (version 5.0). The optimal score cut-off value for the estimation group was selected based on sensitivity and specificity, J. Clin. Med. 2021,10, 2932 6 of 14 and positive and negative likelihood ratios as that at which the sum of sensitivity and specificity was maximum. Statistical analyses were performed using SPSS software, version 23 for Windows (IBM, Chicago, IL), R Software (R version 3.2.0; R Development Core Team, 2010; http://cran.r-project.org) and Microsoft Office Excel (v19.0). 3. Results 3.1. Study Visits The study population comprised 37 genetically diagnosed HFI patients from 31 unrelated families and their 32 healthy controls (Table 1). The HFI patients were 22 females and 15 males; 23 patients were below 18 years of age, and 14 were adults. The controls were 19 females and 13 males, and of these volunteers, 20 were below 18 years of age, and 12 were adults. Table 1. Principal features in HFI patients and in their healthy controls. Continuous variables are represented as mean ±standard deviation and range. Differences in continuous variables between HFI patients and controls were calculated by using the Student’s t-test. Transferrin (Tf). Controls HFI Patients p-Value n32 37 Women/men, n/n19/13 22/15 0.594 Age, year 21.5 ±15.1 (2.1–61.3) 19.7 ±14.8 (3.7–63.4) 0.629 Weight, kg 50.5 ±15.3 (11.9–71.0) 45.3 ±16.4 (17–73) 0.178 Height, cm 155.5 ±19.5 (85.5–181.0) 151.0 ±21.1(104.2–190.7) 0.368 BMI, kg/m220.2 ±3.2 (14.6–27.1) 19.0 ±3.1 (14.0–27.4) 0.107 Waist circumference, cm 68.3 ±9.4 (48–88) 67.5 ±9.7 (51–87) 0.736 CDT, % (asialoTf + disialoTf) 0.7 ±0.1 (0.4–1.0) 2.4 ±1.8 (0.6–8.4) 9<0.001 AsialoTf, % 0.0 ±0.0 (0.0–0.0) 0.10 ±0.3 (0.0–0.1) 0.027 DisialoTf, % 0.7 ±0.1 (0.4–1.0) 2.3 ±1.6 (0.6–7.8) <0.001 TrisialoTf, % 4.4 ±1.7 (2.5–6.7) 5.3 ±1.2 (3.2–7.7) <0.001 TetrasialoTf, % 84.4 ±1.6 (81–87) 79.0 ±2.7 (74–84) <0.001 Pentaand hexa-sialoTf, % 10.8 ±1.1 (9.0–14.1) 13.3 ±1.7 (10.3–17.1) <0.001 Tetra-/di-sialoTf 133.4 ±27.7 (81.3–204.8) 51.8 ±32.9 (9.4–137.5) <0.001 Serum iron (mg/dL) 98.7 ±39.7 (34–205) 85.3 ±30.6 (53–216) 0.133 Tf (mg/dL) 270.9 ±37.8 (210–397) 288.1 ±36.5 (222–378) 0.064 Ferritin (mg/dL) 50.4 ±37.9 (10–176) 73.2 ±73.4 (4–284) 0.124 Significant p-values are marked in bold. The majority of the HFI patients enrolled were diagnosed during their childhood, i.e., at 3.0 ± 2.6 years, except four patients, who were diagnosed at 34, 43, 48,and 63 years of age. There were no differences in weight, height, BMI or waist circumference between the HFI patients and their respective controls (Table 1). 3.2. Biochemical Analyses The relative percentages of all sialoTf isoforms analysed were altered in HFI patients when compared with their controls, as shown in Table 1and in Figure 2. The concentration of CDT, disialo-, trisialoand the sum of pentasialoand hexasialo-Tf in the HFI patients was higher by 3.7-fold, 3.5-fold, 28%,and 23%, respectively, when compared with the controls. In contrast, tetrasialoTf concentration was lower by 6% in the HFI patients in comparison with their healthy controls. AsialoTf was present in 6 out of the 37 HFI patients and was not detected in any control participant. J. Clin. Med. 2021,10, 2932 7 of 14 J. Clin. Med. 2021, 10, x FOR PEER REVIEW 7 of 13 Figure 2. Sialotransferrin (sialoTf) isoform percentages in HFI patients (n = 37) and in their respective healthy controls (n = 32). Differences in continuous variables between HFI and their controls were calculated by using the Student’s t-test. 3.3. Serum Transferrin Isoforms, Hepatic Parameters and Fructose Consumption in Patients with Fructosemia Table 2 shows Tf isoforms, hepatic parameters and FSS consumption in HFI patients, all together and divided by FSS-intake tertiles, and shows that there were no significant differences in sialoTf or the studied hepatic parameters among the three tertiles of FSSintake in HFI participants. Table 2. Dietary intake of fructose, sucrose, sorbitol and their sum (FSS), sialoTf percentages and liver parameters in HFI patients, all together (n = 34) and divided into FSS-intake tertiles: 1st (n = 12), 2nd (n = 11) and 3rd (n = 11). Continuous variables are represented as mean ± standard deviation and range. Differences in continuous variables among the three tertiles were calculated using the Kruskal–Wallis test. HFI patients all together FSS-intake 1st tertile FSS-intake 2nd tertile FSS-intake 3rd tertile p Figure 2. Sialotransferrin (sialoTf) isoform percentages in HFI patients (n= 37) and in their respective healthy controls (n= 32). Differences in continuous variables between HFI and their controls were calculated by using the Student’s t-test. 3.3. Serum Transferrin Isoforms, Hepatic Parameters and Fructose Consumption in Patients with Fructosemia Table 2shows Tf isoforms, hepatic parameters and FSS consumption in HFI patients, all together and divided by FSS-intake tertiles, and shows that there were no significant differences in sialoTf or the studied hepatic parameters among the three tertiles of FSSintake in HFI participants. J. Clin. Med. 2021,10, 2932 8 of 14 Table 2. Dietary intake of fructose, sucrose, sorbitol and their sum (FSS), sialoTf percentages and liver parameters in HFI patients, all together (n= 34) and divided into FSS-intake tertiles: 1st (n= 12), 2nd (n= 11) and 3rd (n= 11). Continuous variables are represented as mean ± standard deviation and range. Differences in continuous variables among the three tertiles were calculated using the Kruskal–Wallis test. HFI Patients all Together FSS-Intake 1st Tertile FSS-Intake 2nd Tertile FSS-Intake 3rd Tertile p n34 12 11 11 Fructose intake (mg/day) 322 ±304 (0–1323) 70 ±78 (0–220) 296 ±161 (98–604) 621 ±311 (238–1323) <0.001 Sucrose intake (mg/day) 503 ±651 (15–3747) 131 ±99 (15–308) 378 ±70 (255–530) 1032 ±944 (344–3747) <0.001 Sorbitol intake (mg/day) 50 ±93 (0–443) 11 ±20 (0–57) 32 ±40 (0–90) 11 ±142 (0–443) 0.049 FSS intake (mg/day) 1764 ±938 (15–4366) 212 ±157 (15–446) 707 ±193 (457–936) 1764 ±937 (1020–4366) <0.001 AsialoTf, % 0.08 ±0.25 (0–0.8) 0.08 ±0.21 (0.0–0.6) 0.09 ±0.22 (0.0–0.8) 0.14 ±0.32 (0.00–1.00) 0.934 DisialoTf, % 2.3 ±1.6 (0.6–7.8) 2.2 ±2 (0.6–7.8) 2.4 ±1.5 (0.9–5.4) 2.3 ±1.3 (0.9–4.6) 0.612 TrisialoTf, % 5.3 ±1.2 (3.2–7.7) 5.1 ±1.1 (3.4–6.8) 5.2 ±1.2 (3.2–7.1) 5.7 ±1.1 (4.0–7.7) 0.395 TetrasialoTf, % 79.0 ±2.7 (73.7–84.3) 78.6 ±2.9 (73.7–82.5) 79.6 ±2.9 (74.4–84.3) 78.7 ±2.1 (76.7–82.7) 0.674 Pentaand hexa-sialoTf, % 13.3 ±1.7 (10.3–17.1) 14.1 ±1.8 (10.8–17.1) 12.7 ±1.6 (10.3–15.9) 13.1 ±1.5 (10.37–14.6) 0.262 Tetra-/di-sialoTf 51.8 ±32.9 (9.4–137.5) 62.2 ±40.8 (9.4–137.5) 47.3 ±29.1 (13.8–93.7) 46.1 ±27.9 (16.7–91.9) 0.758 GGT (U/L) 15.3 ±5.9 (9–36) 17.4 ±8.2 (9–36) 14.4 ±4.9 (9–27) 14.2 ±3.5 (10–21) 0.221 GOT, AST (U/L) 26.5 ±9.5 (16–66) 45.8 ±30.4 (19–45) 28.5 ±4.8 (19–33) 27.9 ±4.7 (16–66) 0.322 GPT, ALT (U/L) 30.4 ±31.4 (9–198) 44.3 ±50.7 (16–198) 22.8 ±9.5 (10–47) 24.7 ±15.7 (9–68) 0.234 INR 1.00 ±0.08 (0.9–1.1) 0.96 ±0.08 (0.9–1.1) 1.00 ±0.67 (0.9–1.1) 1.00 ±0.77 (0.9–1.1) 0.151 Significant p-values are marked in bold. There were six fructosemic participants with asialoTf expression. There were no significant differences in the studied hepatic parameters or in the FSS-intake between HFI patients who expressed asialoTf and those who did not express it (Table 3). Table 3. Dietary intake of fructose, sucrose, sorbitol, their sum (FSS), liver parameters and sialoTf isoform percentages in HFI patients with (n= 6) or without (n= 31) the asialoTf isoform. Continuous variables are represented as mean ± standard deviation and range. Differences in continuous variables between HFI patients with or without asialoTf isoform were calculated using the Mann–Whitney U test. Transferrin (Tf). No AsialoTf Expression AsialoTf Expression p-Value Fructose intake (mg/day) 354 ±321 (0–1323) 434 ±327 (220–620) 0.254 Sucrose intake (mg/day) 434 ±327 (15–1334) 1096 ± 1496 (117–3747) 0.232 Sorbitol intake (mg/day) 64 ±104 (0–443) 19 ±41 (0–93) 0.192 FSS intake (mg/day) 868 ±596 (15–2119) 1535 ± 1618 (337–4366) 0.487 J. Clin. Med. 2021,10, 2932 9 of 14 Table 3. Cont. No AsialoTf Expression AsialoTf Expression p-Value GGT (U/L) 15.5 ±6.1 (9–36) 14.7 ±5.2 (9–21) 1 GOT, AST (U/L) 27.2 ±10.3 (13–66) 23 ±2.8 (19–27) 0.606 GPT, ALT (U/L) 31.5 ±34.2 (9–198) 24.7 ±6.2 (19–35) 0.223 INR 1.0 ±0.1 (0.9–1.1) 1.0 ±0.1 (1.0–1.1) 0.295 CDT, % (asialoTf+disialoTf) 1.8 ±1.0 (0.6–4.5) 5.5 ±1.7 (3.9–8.4) <0.001 AsialoTf, % 0 0.6 ±0.3 (0.2–1.0) <0.001 DisialoTf, % 1.8 ±1.0 (0.6–4.5) 4.9 ±1.5 (3.7–7.8) <0.001 TrisialoTf, % 5.3 ±1.2 (3.2–7.7) 5.6 ±1.2 (4.0–7.1) 0.575 TetrasialoTf, % 79.5 ±2.5 (76.5–84.3) 76.5 ±2.2 (73.5–79.8) 0.005 Pentaand hexa-sialoTf, % 58.6 ±31.7 (10.3–17.1) 16.6 ±4.3 (10.3–14.4) 0.427 Tetra-/di-sialoTf 13.4 ±1.7 (17.0–137.5) 12.5 ±1.6 (9.4–20.9) <0.001 Significant p-values are marked in bold. 3.4. Correlations between Transferrin Isoforms, Hepatic Parameters and Fructose Consumption in Patients with Fructosemia The correlations between the dietary intake of fructose, sucrose, sorbitol and their sum (FSS) with the entire sialoTf profile were assessed in HFI patients (Table 4). The highest correlation values were found for the asialoTf percentage, which correlated with dietary intake of sucrose (R = 0.575, p< 0.001) and FSS (R = 0.475, p= 0.008). The pentasialoTf+hexasialoTf fraction correlated negatively with fructose (R = − 0.386, p= 0.024) and with FSS intake ( R = −0.400 ,p= 0.019), and tetrasialoTf correlated with dietary intake of sorbitol (R = 0.361, p= 0.036). Table 4. Correlation analysis of sialotransferrin (Tf) fraction levels with dietary intake of fructose, sucrose, sorbitol and their sum (FSS) and with liver parameters in HFI patients. Pearson correlation analysis was used to assess bivariate relationships between variables. Significant p-values are marked in bold. Dietary Intake CDT, % A-Tf, % Di-Tf, % Tri-Tf, % Tetra-Tf, % Pentaand Hexa-Tf, % Tetra-/di-Tf Ratio Fructose R0.063 0.107 0.046 0.244 0.092 −0.386 −0.085 p0.725 0.573 0.794 0.169 0.606 0.024 0.631 Sucrose R0.242 0.575 0.18 0.238 −0.073 −0.303 −0.187 p0.168 <0.001 0.309 0.175 0.68 0.08 0.29 Sorbitol R−0.197 −0.175 −0.197 −0.181 0.361 −0.253 0.209 p0.264 0.354 0.264 0.305 0.036 0.15 0.236 FSS R0.187 0.475 0.133 0.25 0.016 −0.4 −0.151 p0.29 0.008 0.453 0.154 0.929 0.019 0.392 Liver Parameters GGT R0.061 0.006 0.077 0.049 −0.12 0.091 −0.107 p0.719 0.974 0.653 0.774 0.48 0.592 0.528 GOT (AST) R−0.213 −0.135 −0.212 −0.004 0.184 −0.059 0.184 p0.212 0.453 0.214 0.982 0.282 0.733 0.281 GPT (ALT) R0.044 −0.01 0.061 0.076 −0.185 0.193 −0.092 p0.795 0.955 0.72 0.656 0.272 0.253 0.587 INR R−0.026 0.082 −0.033 −0.064 −0.055 0.162 0.103 p0.882 0.662 0.849 0.717 0.752 0.351 0.555