Defective IGF-1 prohormone N-glycosylation and reduced IGF-1 receptor signaling activation in congenital disorders of glycosylation

The insulin-like growth factor-1 (IGF-1) signaling pathway is crucial for the regulation of growth and development. The correct processing of the IGF-1Ea prohormone (proIGF-1Ea) and the IGF-1 receptor (IGF-1R) peptide precursor requires proper N-glycosylation. Deficiencies of N-linked glycosylation lead to a clinically heterogeneous group of inherited diseases called Congenital Disorders of Glycosylation (CDG). The impact of N-glycosylation defects on IGF-1/IGF-1R signaling components is largely unknown. In this study, using dermal fibroblasts from patients with different CDG [PMM2-CDG (n = 7); ALG3-CDG (n = 2); ALG8-CDG (n = 1); GMPPB-CDG (n = 1)], we analyzed the glycosylation pattern of the proIGF-1Ea, IGF-1 secretion efficiency and IGF-1R signaling activity. ALG3-CDG, ALG8-CDG, GMPPB-CDG and some PMM2-CDG fibroblasts showed hypoglycosylation of the proIGF-1Ea and lower IGF-1 secretion when compared with control (CTR). Lower IGF-1 serum concentration was observed in ALG3-CDG, ALG8-CDG and in some patients with PMM2-CDG, supporting our in vitro data. Furthermore, reduced IGF-1R expression level was observed in ALG3-CDG, ALG8-CDG and in some PMM2-CDG fibroblasts. IGF-1-induced IGF-1R activation was lower in most PMM2-CDG fibroblasts and was associated with decreased ERK1/2 phosphorylation as compared to CTR. In general, CDG fibroblasts showed a slight upregulation of Endoplasmic Reticulum (ER) stress genes compared with CTR, uncovering mild ER stress in CDG cells. ER-stress-related gene expression negatively correlated with fibroblasts IGF-1 secretion. This study provides new evidence of a direct link between N-glycosylation defects found in CDG and the impairment of IGF-1/IGF-1R signaling components. Further studies are warranted to determine the clinical consequences of reduced systemic IGF-1 availability and local activity in patients with CDG. Supplementary Information The online version contains supplementary material available at 10.1007/s00018-022-04180-x.


Introduction
The insulin-like growth factor-1 (IGF-1) signaling pathway is fundamental for growth regulation, especially in childhood, while also playing a central role in development and metabolic homeostasis. In the bloodstream, IGF-1 circulates primarily as part of a 150 kDa ternary complex with two additional proteins, the IGF binding protein-3 (IGFBP-3) and the acid-labile subunit (ALS). This system extends the half-life of circulating IGF-1 from 10 to 12 min to about 15 h, preserving stable serum levels of IGF-1 [1,2]. Although growth hormone (GH) is a major regulator of hepatic IGF-1 production, IGF-1 is also regulated and secreted by other organs in an autocrine/paracrine manner [3]. Tissue-specific IGF-1 knockout models demonstrated the role of circulating IGF-1 in regulation of body size and tissue growth and revealed autocrine actions in specific tissues such as skeletal muscle [4], bone [5] and nervous system [6]. Thus, the activity of IGF-1 is due to a combination of local expression together with IGF-1 delivered to the tissue from the circulation where high levels are maintained [7]. The function of all the IGF-1 system components is affected by their glycosylation status [8,9]. The mature IGF-1 is a non-glycosylated polypeptide of 7.6 kD derived from post-translational cleavage of the C terminal E peptide from the IGF-1 prohormone (proIGF-1E) [9]. In human, the most ubiquitously expressed prohormone, named proIGF-1Ea, is glycosylated (N-glycosylation site on residue 92 of the Ea petide) and proIGF-1Ea N-glycosylation ensures proper secretion of the mature IGF-1 [9,10]. Both IGFBP-3 and ALS proteins are glycosylated. IGFBP-3 contains three N-glycosylation sites [1] and ALS is a soluble glycoprotein of 85 kDa [11]. Previous studies have demonstrated that N-glycosylation prolongs the half-life of IGFBP-3 [12], increases the affinity of ALS to the IGFBP-3/IGF-1 complex [13] and thus is required for the correct formation and stabilization of the 150 kDa ternary complex [2]. The actions of IGF-1 are regulated by the tyrosine kinase receptor IGF-1R. This receptor is synthesized as a single polypeptide precursor (proreceptor), which undergoes proteolytic cleavage into α (130 KDa) and β (97 kDa) chains that form a tetramer (α-β-β-α). Each α and β subunit contains 11 and 5N-glycosylation sites, respectively [14]; therefore, the α-β-β-α tetramer structure may enclose 32 glycosylation positions [15]. Several studies demonstrated that proper N-glycosylation of IGF-1R proreceptor is required for correct IGF-1R maturation and transport to the cell surface [15][16][17].
Defects in either IGF-1 or its receptor can result in poor pre-and post-natal growth [18,19]. Beside growth failure, patients with IGF-1R haploinsufficiency often exhibit additional clinical features including skeletal malformation, intellectual disability, cardiac defects and facial dysmorphisms [19,20]. In contrast, much less is known about the impact of N-glycosylation genetic defects, such as those found in Congenital Disorders of Glycosylation (CDG), on the IGF-1 system. CDG are a large family of rare inborn errors of metabolism caused by defective glycoprotein and glycolipid glycan synthesis and attachment. They comprise a broad spectrum of clinical manifestations, because glycosylation occurs in every cell and involves all organs and tissues [21]. Major clinical manifestations include nearly always neurological involvement (such as developmental/ intellectual disability, hypotonia, epilepsy, stroke-like episodes, polyneuropathy) besides other organ involvement [21]. Nearly 160 CDG have been described and their number increases exponentially [22]. Glycosylation comprises a large number of biochemical pathways, including N-linked and O-linked pathways, glycosylphosphatidylinositol (GPI) anchor synthesis and glycolipid glycosylation. It occurs in several organelles: the cytoplasm, the endoplasmic reticulum (ER) and ER-Golgi intermediate compartment (ERGIC), the Golgi and vesicular network and the plasma and sarcolemmal membranes [23]. Most CDG are ultrarare disorders, those affecting N-glycosylation are the most common. Several specific therapeutic strategies are under evaluation for these diseases [24,25]. PMM2-CDG is the most common N-glycosylation defect with more than 900 reported patients worldwide [21].
N-linked glycosylation disorders are usually identified by analyzing the serum transferrin isoforms by isoelectric focusing (IEF) or mass spectrometry [23]. On the basis of the transferrin IEF profiles (type 1 or type 2 pattern), N-glycosylation defects are classified as CDG-I or CDG-II. CDG-I results from defects in the glycan assembly in the cytoplasm and ER and is hallmarked by the absence of one or more N-glycans on glycoproteins [23]. CDG type-II is Golgi-related and characterized by defective glycans [23]. Hypoglycosylation leads to protein misfolding and a mild increase of ER-stress markers in patients' fibroblasts [26].
Growing evidence suggests that insufficient levels and/ or activity of growth factors might contribute to some of the clinical manifestations in CDG [8,[27][28][29]. The endocrine data from PMM2-affected children showed that circulating IGF-1 levels were in the low or low normal range in infants and children below the age of 9 years (n = 15 median, 33 µg/L; range, 12-92 µg/L; normal, 77-135 µg/L) and in the low-to-normal IGF-1 levels in adolescents (n = 8; median, 267 µg/L; range, 107-437 µg/L; normal, 350-504 µg/L) [29]. Similarly, Miller et al. [8] reported reduced serum levels of ALS, IGFBP-3, IGF-1 and of ternary complex formation in children with PMM2-CDG (n = 12; age = 8.1 ± 6.6 years) compared with age-matched controls, despite normal or elevated levels of GH. These features were also found in the recently developed mouse model of PMM2-CDG [28].
The molecular basis of the IGF-1 system deficiency found in PMM2-CDG patients is currently unknown.
In this study, we investigated whether glycosylation is directly involved in the function of the IGF-1 system. We hypothesize that primary fibroblasts from patients with CDG show aberrant proIGF-1Ea N-glycosylation and impaired activation of the IGF-1R signaling pathways.
For IGF-1 gene overexpression, fibroblasts were transfected using the Human Dermal Fibroblast Nucleofector™ Kit (Cat #VPD1001; Lonza) using the Nucleofector™ Device (Lonza) following the instruction manual. Briefly, 1 × 10 6 fibroblasts were resuspended in the nucleofector solution, combined with 2.5 µg of plasmid construct containing the class 1 IGF-1Ea isoform [32], transferred to the cuvette supplied and finally transfected with the U-23 program of the Nucleofector™ Device. After electroporation, cells were mixed with 500 µl of DMEM and immediately plated into a 6-well plate. The efficiency of transfection was estimated by GFP-dependent fluorescence and by qPCR for total IGF-1 mRNA levels at 48 h after transfection, as previously described [32]. The fibroblasts' concentration and cell viability were determined by the LUNA-II™ Automated Cell Counter (Logos Biosystems, Twin Helix) with trypan blue staining.

ELISA assay
Quantitative determination of IGF-1 concentrations in transfected fibroblasts supernatants was performed by a commercially available ELISA kit following the manufacturer's instructions (Human IGF-I/IGF-1 DuoSet ELISA Cat #DY291-05; R&D Systems). The data were acquired at a wavelength of 450 nm using Model 680 microplate reader (Bio-Rad Laboratories).

Western blot and lectin blot analyses
Fibroblasts were processed for western blot analyses as previously reported [32]. The protein samples (20-40 µg total proteins) were electrophoresed with 10% SDS-PAGE and then transferred to nitrocellulose or PVDF membranes (HRP)-conjugated secondary antibody (against primary antibodies) or streptavidin-HRP (for biotinylated lectins) at room temperature for 1 h and were then washed three times. Blots were developed using Clarity Western ECL Substrate (Bio-Rad Laboratories) and were quantified using the ChemiDoc MP (Bio-Rad Laboratories) equipped with Image Lab software.

RNA extraction, cDNA synthesis and quantitative real-time polymerase chain reaction
The total RNA was extracted and genes of interest analyzed by real-time qPCR as previously described [33]. Briefly, the expression of ER-stress-related genes was monitored by qPCR using TB Green Premix Ex TaqII Mastermix (Takara Bio Europe, France), in a RotorGene 6000 instrument (Corbett life science, Sydney, Australia). As template, 2 ng of total RNA used for cDNA synthesis was used in each PCR tube and a non-template control was included for each primer pair reaction as negative control. All amplification reactions were performed in duplicate. The amplification conditions were: 95 °C for 10 min, 95 °C for 10 s and 60 °C for 50 s for 40 cycles. At the end of each run, a melting curve analysis from 65 °C to 95 °C was performed to exclude the presence of non-specific products or primer dimers. The data were normalized against the reference gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase). The relative expression levels were calculated using the 2 −ΔΔCt method [34].

Statistical analyses
The data are represented as mean ± SEM of at least three independent experiments. Statistical analyses were performed using the t test or one-way ANOVA as appropriate, followed by Tukey's Multiple Comparison post hoc test. Correlation between fibroblasts IGF-1 secretion, densitometric quantification of proIGF-1Ea and IGF-1R immunoreactive bands, lectin binding and ER-stress-related genes was analyzed using Pearson's correlation. The statistical tests were performed using SPSS (IBM SPSS Statistics for Windows, Version 20.0, IBM Corp.) and GraphPad Prism version 5 (GraphPad Software, Inc., La Jolla, CA, USA). A p value ≤ 0.05 was considered statistically significant.

proIGF-1Ea N-glycosylation pattern and mature IGF-1 secretion in CDG fibroblasts
To evaluate the expression level of IGF-1 in CTR and CDG fibroblasts, we conducted preliminary analyses on the IGF-1 mRNA and protein quantity. IGF-1 mRNA quantification by qPCR showed that all fibroblasts had detectable levels of IGF-1 mRNA (mean CT value 30.5); however, the IGF-1 protein level was too low to be detected by western blot or ELISA. We also tried to quantify the IGF-1 secretion in the fibroblast supernatants after concentration using 3 kDa Spin Columns (by about 20-fold; Amicon ® Ultra Merck Millipore), but the IGF-1 concentration still remained below the ELISA sensitivity (93.8 pg/mL). Subsequently, to assess the expression pattern and the N-glycosylation status of the IGF-1 protein, we transiently transfected fibroblasts with a plasmid encoding the IGF-1Ea isoform or an empty vector [32] by electroporation.

Discussion
In this study, we provide new evidence for IGF-1 system impairment in CDG fibroblasts due to defective N-glycosylation of proIGF-1Ea and IGF-1R, which partially disrupts IGF-1 signaling pathway activation. We previously demonstrated that, under physiological conditions, intracellular IGF-1 is mainly expressed as a heavily N-glycosylated prohormone of 17-22 kDa [10]. Here, using CTR and CDG fibroblasts transiently transfected with the IGF-1Ea isoform, we monitored the N-glycosylation pattern of proIGF-1Ea and IGF-1 secretion. Two distinct bands, likely representing glycosylated and unglycosylated proIGF-1Ea, were found in CTR, PGAP2-CDG and most PMM2-CDG fibroblasts by western blotting. Conversely, a unique band was present in fibroblasts derived from ALG3-CDG, ALG8-CDG and GMPPB-CDG. In both ALG3-CDG, we found an abnormal band of about 14 kDa, while in ALG8-CDG and in GMPPB-CDG, we found a marked accumulation of a 12 KDa isoform. These bands probably represent hypoglycosylated isoforms of the proIGF-1Ea. To evaluate if aberrant proIGF-1Ea glycosylation impaired IGF-1 secretion, we quantified the IGF-1 protein level in cell culture supernatants and found that most CDG fibroblasts released less IGF-1 in the culture media compared to CTR. These data corroborate our previous findings that N-glycosylation of the proIGF-1Ea is essential for correct IGF-1 secretion [10]. Normal proIGF-1Ea N-glycosylation and IGF-1 secretion in PGAP2-CDG fibroblasts was expected, since PGAP2 is not involved in N-glycosylation but in GPI-anchor synthesis of GPI-anchored proteins. The IGF-1 serum level was evaluated in seven patients, whose fibroblasts have been included in this work (ALG3-CDG, ALG8-CDG, PGAP2-CDG and PMM2_p1,_p2,_p4,_p7-CDG). ALG3-CDG and ALG8-CDG patients showed low IGF-1 serum levels, while lowto-normal IGF-1 serum levels were found in studied patients with PMM2-CDG. These data remain provisional, due to the low number of patients tested, but they are in line with the published data on reduced circulating IGF-1 levels in CDG patients [8,29].
Because the effects of IGF-1 are mediated principally through the IGF-1R, we subsequently analyzed the IGF-1R levels in primary fibroblasts from different CDG. Previous studies have shown that N-glycosylation contributes to ligand binding, kinase activity and turnover of the IGF-1R [35]. Accordingly, we found that IGF-1R expression and activity were reduced in primary fibroblasts from different CDG. Levels of IGF-1R were particularly low in ALG3-CDG, ALG-8-CDG and in some of the PMM2-CDG fibroblast cell lines tested. We did not find an accumulation of the IGF-1 proreceptor in these cells, suggesting that defective N-glycosylation causes IGF-1R proreceptor instability and degradation [27]. In PMM2-CDG, we also analyzed the IGF-1-induced IGF-1R pathway activation. Most PMM2-CDG fibroblasts showed reduced IGF-1R and ERK1/2 phosphorylation compared to CTR, while the Akt activation did not differ significantly probably due to the high variability of the Akt response among PMM2-CDG fibroblasts. 1R, Akt and ERK1/2 phosphorylation level in CTR (mean of three technical replicates for each of five biological replicates) and PMM2-CDG (mean of three technical replicates for each of seven biological replicates) fibroblasts (E). *significantly different from CTR fibroblasts; #significantly different from recombinant IGF-1 untreated cells *** and ### p < 0.0001, ## **p < 0.001, *p < 0.01; # p < 0.05 Fig. 4 Lectin-binding analysis. Concanavalin A (ConA) and Phaseolus vulgaris leucoagglutinin (PHA-L) binding to CTR (mean of three technical replicates for each of five biological replicates) and CDG (mean of three technical replicates for each of twelve biological replicates) fibroblasts. Con A and PHA-L recognize high mannose and complex type N-glycans, respectively. *significantly different from CTR fibroblasts, *p < 0.0001 Altogether, these data corroborate and expand previous observations demonstrating an IGF-1 system deficiency in CDG [8,[27][28][29]. In particular, our results suggest that the glycosylation defect found in CDG might impair the IGF-1 system not only indirectly, by destabilizing the ternary complex [8], but also directly due to proIGF-1Ea hypoglycosylation and IGF-1 secretory defect. Furthermore, local growth-promoting effects of IGF-1 might be compromised by the reduced IGF-1R proreceptor glycosylation and related impairment of IGF-1 signal transduction. We believe that these findings are relevant in the context of CDG since both systemic IGF-1 production and local bioactivity are essential to support normal growth during development [5,7]. Patients with genetic IGF-1R defects, show symptoms (i.e., pre and postnatal growth retardation, microcephaly, cardiac defects and dysmorphic features) that significantly overlap with those presented in various CDG types [19,20]. For example, the majority of CDG patients analyzed in the present study showed growth failure and microcephaly (Supplementary Table 1), which are also clinical hallmarks of IGF-1/IGF-1R defects [36][37][38]. Further studies are needed, also taking advantage of the recently developed CDG animal models [28], to evaluate the impact of glycosylation defects on each component of the IGF-1 system. This, in turn, will help to clarify the association between clinical CDG features and IGF-1/ IGF-1R signaling abnormalities. Interestingly, a case study showed that treatment with recombinant IGF-1 prompted linear growth in a child with PMM2-CDG [39], suggesting the potential of IGF-1 therapy in the improvement of clinical outcome of CDG patients. However, taking into account our data together with previous findings [27], the residual IGF-1R activity should be measured before starting a recombinant IGF-1 therapy.
Although typical clinical features of CDG were present in patients whose fibroblasts we studied, clinical manifestations and severity have a wide spectrum, which is probably related to the specific gene variants and/or enzyme residual activity [30 and Supplementary Table 1]. Global quantitative protein glycosylation performed with lectins that have specific affinity for different types of N-glycans (ConA and PHA-L) failed to highlight difference between different CDG fibroblast subtypes. We found a modest, and not significant, decrease of PHA-L reactivity, indicating reduced levels of β1-6 branching structures and complex N-glycans. A positive association between IGF-1 secretion and PHA-L binding was found. In addition, the analysis of ER-stressrelated gene expression showed a slight increase of CHOP/ DDIT3, uXBP1 and ATF4 mRNA level in CDG compared to CTR fibroblasts, which negatively correlate with fibroblasts IGF-1 secretion. A correlation between ER stress and IGF-1 system alterations has been proposed in several neurodegenerative diseases that share as common feature the accumulation of misfolded proteins [40]. Although further studies are needed, the broad modification of protein N-glycosylation patterns and the mild ER-stress in CDG fibroblasts might also have contributed to the IGF-1/IGF-1R impaired signaling observed in CDG fibroblasts [40,41].  level of selected  ER-stress-related genes: CHOP/  DDIT3, sXBP1, uXBP1, HSPA5,  CEBPB, CHAC1, MAP1LC3B and ATF4 in CTR (mean of three technical replicates for each of five biological replicates) and CDG (mean of three technical replicates for each of 12 biological replicates) fibroblasts. *significantly different from CTR fibroblasts, **p < 0.01, *p < 0.05

Conclusions
In this study, we have shown that primary fibroblasts from CDG patients have reduced levels of proIGF-1Ea and IGF-1R proreceptor glycosylation. The decreased glycosylation of IGF-1/IGF-1R signaling pathway components reduced the IGF-1 bioactivity in CDG fibroblasts. Thus, our molecular data provide new insight into CDG pathogenesis and suggest that both IGF-1 production (circulating/local) and bioactivity might be compromised in CDG due to reduced N-glycosylation.
These findings pave the way to future studies focused on the impact of CDG on the different IGF-1 system components, which include IGF-1, IGFBPs and the IGF-1R. More efforts will be needed to clarify the correlation between reduced IGF-1/IGF-1R signaling in CDG patients and their clinical manifestations.
Acknowledgements "The Cell Line and DNA Biobank from Patients Affected by Genetic Diseases", member of the Telethon Network of Genetic Biobanks (project no. GTB18001), funded by Telethon Italy, provided us with specimens". Thanks to the Professor Timothy Bloom for a critical reading of this work. We gratefully acknowledge AMMeC (Associazione Malattie Metaboliche Congenite, Italy), for technical assistance and collaboration to the Molecular and Cell Biology Laboratory of Neurometabolic Diseases at the Meyer Children's Hospital of Florence, Italy.
Author contributions LDP, GA and EB conceived and designed the study. LDP, GA, RS and MB performed cell culture experiments, western blot and lectin blot analyses and ELISA assay. LG and AD performed the quantification of ER-stress-related genes. GA, AM, LF and EB analyzed the data and drafted the manuscript. MAD, RB, RG, JJ, VS and EB provided overall direction to the project and revised the manuscript. All the authors contributed to manuscript revision, read and approved the submitted version.
Funding The laboratory part of this research did not receive any specific Grant from funding agencies in the public, commercial or notfor-profit sectors. Patients were studied in the context of the Tuscany Region Call for Health 2018 (Grant DECODE-EE).

Availability of data and materials
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflict of interest
The authors declare no competing interests.

Code availability Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.