1. Introduction

The PDGFRB gene encodes the platelet-derived growth factor receptor beta (PDGFRβ) protein, a transmembrane receptor tyrosine kinase (RTK) [1]. Upon platelet-derived growth factor (PDGF) binding, dimerization of the kinase occurs, followed by phosphorylation of the tyrosine residues located in the intracellular domain [2, 3]. The phosphorylated tyrosine residues facilitate the binding of downstream signaling molecules [4, 5]. The cascade of events initiated by phosphorylation of PDGFRβ ultimately affects gene expression and modulates cell proliferation, differentiation, and motility [1, 4]. Dysregulation of the PDGFRβ/PDGF signaling, usually through somatic mutations or fusion with other genes, has been associated with several human disorders including cancers [6–9]. In addition, several pathogenic PDGFRB germline variants have been reported [10].

Gain-of-function germline variants, reviewed by Guérit and colleagues [10], have been linked to several different clinical conditions including infantile myofibromatosis, Kosaki overgrowth syndrome (KOGS), premature aging syndrome, Penttinen type (Penttinen syndrome), and ocular pterygium–digital keloid dysplasia (OPDKD). The PDGFRB variants p.Trp566Arg and p.Pro584Arg located in the juxtamembrane region are linked to KOGS [11–16]. The variants associated with familial cases of infantile myofibromatosis include p.Pro560Leu and p.Arg561Cys [17–19] located in the juxtamembrane region. Variants p.Val6665Ala and p.Asn666Ser, located in the kinase domain, are associated with Penttinen syndrome [20, 21]; p.Asn666Tyr is associated with OPDKD [22]; and p.Asn666His is reported in a single individual, with clinical manifestations described below [23]. These are all constitutively active, ligand-independent variants. One ligand-dependent activating variant has been reported in the transmembrane domain, namely, p.Ser548Tyr, linked to corneal vascular overgrowth in otherwise healthy individuals [24]. In the extracellular domain, the p.Ser493Cys variant has been described in one individual with KOGS [16]. In contrast, loss-of-function variants in PDGFRB have been associated with primary familial brain calcifications. These variants have been observed across all regions of the receptor [25–28].

Intriguingly, the five different known variants of p.Asn666 are associated with strikingly different clinical conditions. A summary of clinical data is presented in Table S1. Briefly, Cheung and coworkers [19] described the p.Asn666Lys variant in a patient with somatic infantile myofibromatosis (the patient also had a germline p.Arg561Cys alteration). This variant was later confirmed in other sporadic infantile myofibromatosis cases, both in combination with the p.Arg561Cys variant and alone [29]. Similarly, a recently identified variant associated with somatic infantile myofibromatosis includes the p.Asn666Thr and p.Ile564_Val572del mutations in PDGFRB [30]. The pathogenicity of this variant remains unconfirmed and was not investigated in this study. The p.Asn666His variant was found in one patient with a complex congenital syndrome with cranial, hand, dental, and growth abnormalities [23]. Furthermore, a recurrent germline p.Asn666Ser variant has been linked to a severe form of Penttinen syndrome [21, 31], a highly debilitating condition characterized by lipodystrophy, acro-osteolysis, chronic ulcers in limbs, and corneal vascularization. More recently, a p.Asn666Tyr variant has been described, associated with OPDKD [22]. This condition manifests as corneal vascular overgrowth in early childhood later followed by keloid formation on digits. While this substitution in patient fibroblasts gives near normal PDGFRβ activation at the core body temperature of 37°C, at the lower physiological temperature of the cornea (32°C), ligand-independent overactivation of PDGFRβ and downstream signaling proteins were found [22]. This can explain why only lower temperature parts of the body (cornea and digits [32, 33]) are affected. Also, in p.Asn666Ser fibroblasts, at 32°C, upregulation of p-STAT1, a downstream signaling protein of PDGFRβ, was observed, indicating partial temperature sensitivity also of this PDGFRB variant [22].

To understand how p.Asn666Lys, p.Asn666Ser, p.Asn666His, and p.Asn666Tyr variants can lead to so different clinical conditions, we have investigated the phosphorylation of specific tyrosine residues in PDGFRβ and downstream signaling proteins with a particular focus on the effects of physiologically reduced temperature (32°C) and ligand stimulation. We show that all four p.Asn666 PDGFRB variants exhibit distinct phosphorylation patterns both at 32°C and when stimulated with PDGF. All four variants are temperature sensitive, with increased activation at 32°C. Each p.Asn666 PDGFRB variant thus displayed a unique pattern of activation.

2. Materials and Methods

3. Results

3.1. The p.Asn666Tyr, p.Asn666Ser, p.Asn666His, and p.Asn666Lys PDGFRβ Substitutions Cause Ligand-Independent PDGFRβ Activation

In this experiment, immortalized fibroblasts were employed, which had been stably transduced with wild-type PDGFRB (wt), or with vectors expressing p.Asn666Lys, p.Asn666Ser, p.Asn666His, and p.Asn666Tyr substitutions. All four substitutions resulted in increased basal phosphorylation levels of PDGFRβ compared to wild type at 37°C (2.66, 2.51, 2.89, and 2.2 folds, respectively) (Figure 1a). After exposure to 32°C for 6 h, p.Asn666Tyr, p.Asn666Lys, and p.Asn666His showed increased PDGFRβ phosphorylation. The p.Asn666Ser variant showed a trend to higher PDGFRβ phosphorylation at 32°C; however, this was not statistically significant (Figure 1b). In immortalized fibroblasts, similar levels of unphosphorylated PDGFRβ were observed, indicating comparable levels of PDGFRβ expression (Figure 2).

3.3. Specific Downstream Signaling Protein Changes at 37°C and 32°C for the p.Asn666Tyr, p.Asn666Ser, p.Asn666His, and p.Asn666Lys Amino Acid Substitutions

The identification of specific tyrosine residue overactivation prompted an exploration into the temperature effects of the downstream signaling of these variants. To explore the molecular effects of p.Asn666 variants, we systematically examined the selected downstream signaling proteins of PDGFRβ (Figure 3). At 37°C, all p.Asn666 variants exhibited an overactivating effect on one or more signaling mediators, namely, AKT, STAT1, STAT3, STAT6, and PLCγ (Figure 3a,b). Each p.Asn666 substitution demonstrated a distinct pattern. To illustrate, p.Asn666Lys displayed at least a twofold increased phosphorylation level of all examined downstream signaling proteins compared to wild type (Figure 3a,b). At 37°C, comparable levels of phosphorylated STAT1, AKT^Ser473, and PLCγ were seen in p.Asn666Tyr and wild type, while AKT^Thr308, STAT3, and STAT6 showed at least a fourfold increase. Downstream signaling of p.Asn666Ser and p.Asn666His exhibited similar patterns at 37°C, with at least twofold increase of phosphorylated AKT^Thr308, STAT1, and STAT3 while maintaining phosphorylated AKT^Ser473 and PLCγ levels similar to the wild type. Notably, while phosphorylated STAT6 was increased more than 10-fold in p.Asn666Lys, less than a twofold increase was seen in p.Asn666Ser compared to wild type (Figure 3a,b). These distinct phosphorylation patterns in downstream signaling indicate a different constitutively active gain-of-function pattern in all four variants.

We then investigated how downstream signaling proteins respond to a lower physiological temperature (Figure 3a,c). We observed that the activation of the PI3K-AKT pathway was the dominant pattern in p.Asn666Tyr and p.Asn666Lys cells. A particularly prominent 32°C effect was observed for the p.Asn666Tyr substitution (Figure 3a,c). On the other hand, upon exposure to 32°C, p.Asn666Ser, p.Asn666His, and p.Asn666Lys led to increased phosphorylation of various partners in the STAT pathway (Figure 3a,c). At 32°C, the p.Asn666Ser variant exhibited a more than 10-fold increase in p-STAT1 indicating a strong activation of p-STAT1. The p.Asn666Lys substitution had, in addition to p-STAT1, an extensive increase of p-STAT6 at 32°C (more than 50-fold and 4-fold, respectively). For p.Asn666His, all three investigated STAT family members were overactivated at 32°C; p-STAT1 and p-STAT3 increased more than 50-fold, while at least a 10-fold increase was seen for p-STAT6. Notably, all four variants exhibited varying degrees of increased phosphorylation levels in the PLCγ1 pathway at 32°C (Figure 3a,c).

4. Discussion

The identification of five germline variants in the p.Asn666 position associated with severe clinical manifestations suggests that this position may play a crucial role in PDGFRβ signaling and disease development. The p.Asn666Lys substitution in PDGFRβ has been predicted to abolish the interaction between Asn666 and His661, changing potential protein interactions in this region and possibly leading to a structure more similar to the active conformation of KIT (proto-oncogene c-KIT) kinase [19]. The other disease-causing substitutions in neighboring PDGFRB sites further support that notion. The p.Val665Ala variant triggers constitutive activation of PDGFRβ and is linked to Penttinen syndrome [20, 34, 35], while the loss-of-function p.Leu658Pro is associated with primary familial brain calcification [25, 36].

In this study, we confirmed that p.Asn666Tyr, p.Asn666Ser, p.Asn666His, and p.Asn666Lys are gain-of-function PDGFRB variants, causing constitutive ligand-independent activation [19, 21–23] (Figure 1a). The total overactivation of PDGFRβ itself could not account for the distinct and quite different, clinical manifestations seen in these p.Asn666 variants. We observed that p.Asn666Tyr, p.Asn666Lys, and p.Asn666His showed further increased total phosphorylation of PDGFRβ at 32°C measured by ELISA, while p.Asn666Ser did not reach statistical significance (Figure 1b). In addition, all variants exhibited temperature responses at specific PDGFRβ tyrosine residues at 32°C (Figure 2), indicating their temperature-sensitive nature.

The phosphorylation pattern of specific tyrosine residues was observed to be different when cells were stimulated with ligand. The increased phosphorylation was predominantly observed in the higher 180–190 kDa band that represents the fully mature PDGFRβ located in the plasma membrane [37, 38]. The receptor has to be situated in the plasma membrane for activation by extracellular ligands. In contrast, at 32°C, also the lower 160 kDa band showed increased phosphorylation (Figure 2). The 160 kDa band represents an immature form of PDGFRβ [37, 38]. The ligand-dependent activation of the p.Asn666 variants thus revealed variant-specific phosphorylation of investigated tyrosine residues, highlighting distinctive phosphorylation patterns for each p.Asn666 variant (Figure 2).

At 37°C, all variants exhibited overactivation in specific downstream signaling pathways compared to the wild type, confirming their constitutively active nature (Figure 3a,b). At 32°C, an increased activation in various downstream signaling proteins was observed across all four p.Asn666 variants. The p.Asn666Tyr variant exhibited elevated levels of both p-AKT^Thr308 and p-AKT^Ser473 (Figure 3a,c), confirming original findings in patient fibroblasts [22]. Furthermore, the overactivation of STAT1 at 32°C was observed in all four variants, although this ratio might be less accurate in the p.Asn666Tyr variant, as the amount of p-STAT1 at 37°C approached background levels (Figure 3a,c). Such temperature-sensitive overactivation of phosphorylated p-STAT1 in the p.Asn666Ser variant was previously reported in patient fibroblasts [22].

It is thought that specific residues serve as a docking site for distinct downstream signaling proteins [5]. Our results only partly support this notion. For instance, the Y-751 residue serves as one of the docking sites for the PI3K-AKT signaling pathway [39, 40]. In line with this, we find that the p.Asn666Tyr variant exhibited the strongest activation of Y-751, resulting in a strong activation of both p-AKT^Ser473 and p-AKT^Thr308 at 32°C. However, the p.Asn666Lys variant displayed robust AKT signaling at 37°C, but the phosphorylation level of Y-751 was among the lowest observed among the p.Asn666 variants. A similar scenario was observed with the Y-1009 residue, a docking site for PLCγ signaling [41, 42]. The p.Asn666His and p.Asn666Tyr variants had the strongest phosphorylation of the Y-1009 at 32°C, resulting in noticeable activation of PLCγ (Figures 2 and 3a,c). Conversely, the p.Asn666Lys variant exhibited strong phosphorylation of PLCγ with undetectable levels of p-Y-1009 both at 37°C and 32°C (Figures 2 and 3a,c). This may suggest the presence of alternative docking sites for the downstream signaling pathways [43], or it could indicate variations in the timing between tyrosine residue phosphorylation and downstream signaling phosphorylation at different temperatures. These phosphorylation patterns represent a dynamic process captured at a specific time point under defined experimental conditions. This may contribute to the differences in observed phosphorylation peaks at individual PDGFRβ tyrosine residues and in downstream signaling. Variability in the phospho-Y antibody affinity, specificity, and epitope accessibility may also affect the signal intensity and contribute to the consistently strong signals of p-Y-751 and p-Y-1009. Quantitative phospho-proteomics could provide an unbiased assessment of site-specific phosphorylation and clarify potential antibody-dependent variability.

AKT is involved in PDGFRβ-mediated cellular processes, including proliferation, differentiation, survival, growth, migration, and angiogenesis [44, 45]. Furthermore, gene variants causing overactivation of the PI3K/AKT/mTOR pathway are implicated in overgrowth disorders [46]. The clinical manifestations of overgrowth observed in the p.Asn666Tyr variant, affecting the cornea and digits [22], are associated with increased phosphorylation of AKT at 32°C (Figure 3a; Figure S6). A strong AKT involvement at both temperatures was observed in the p.Asn666Lys variant (Figure 3a), which could be associated with tumor overgrowth [47–49].

The STAT1 upregulation in PDGFRβ signaling has been associated with a tissue-wasting phenotype [50]. Although speculative, the severe acro-osteolysis and tissue wasting of limbs described in the p.Asn666Ser variant and the less severe acro-osteolysis of hands in the p.Asn666His variant [21, 23, 31] (Table S1) could be linked to strong STAT1 upregulation at 32°C (Figure 3a). In contrast, the overgrowth phenotype in the p.Asn666Tyr variant exhibited low levels of STAT1 signaling at both temperatures (Figures 3a, 3b, and 3c).

The transcription factors STAT3 and STAT6 regulate cell functions like growth, differentiation, and immune responses, with aberrations in their activation linked to cancer, autoimmunity, and inflammation [51–54]. Similarly, PLCγ, a key enzyme in phosphoinositide signaling, influences cell growth, calcium regulation, and inflammation [55, 56]. A full understanding of the precise roles of PDGFRβ downstream pathways in the development of diseases associated with p.Asn666 variants remains elusive.

Temperature kinetic studies indicate that temperature-induced ligand-independent phosphorylation of all four p.Asn666 PDGFRB variants is a dynamic and reversible process (Figure S6, S7, and S8). This dynamic nature and reversibility of temperature-induced activation adds further complexity to their downstream signaling. This observation prompts the consideration of potential mechanisms underlying the temperature sensitivity of the mutant kinase domain. This may indicate changes in intrinsic stability or efficiency at lower temperatures, or it may be due to indirect effects on phosphorylation regulators like phosphatases or processes such as endocytosis and membrane trafficking. Further research is needed to clarify these interactions and their impact on kinase function at varied temperatures.

5. Conclusions

Understanding the precise contributions of downstream signaling proteins to the clinical manifestations of the four p.Asn666 PDGFRB variants is a complex challenge. However, potential associations have been identified, such as that AKT signaling could influencing overgrowth phenotypes and STAT1 upregulation could be linked to tissue wasting or degenerative clinical features. The distinct pattern of phosphorylation of specific tyrosine residues of PDGFRβ observed in the p.Asn666 variants provides a new possible explanation of how gain-of-function PDGFRB variants result in such strikingly different phenotypes. The four p.Asn666 PDGFRB variants exhibit variant-specific tyrosine residue and downstream signaling overactivation at 32°C, emphasizing the importance of temperature as an environmental factor.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding

This work was supported by grants from the Western Norway Regional Health Authority (912161 and F-12860 to C.B.), the Norwegian Association of the Blind and Partially Sighted (to C.B.), the Dr. Jon S. Larsen Foundation (to C.B.), the Gidske og Peter Jacob Sørensens Forskningsfond (to T.G), and the Olav Raaholt og Gerd Meidel Raagholts stiftelsen (to T.G.).