Homozygous SHBG Variant (rs6258) Linked to Gonadotropin-Independent Precocious Puberty in a Young Girl

Abstract

Sex hormone–binding globulin (SHBG) in the blood is a major determinant of bioactivity for key sex steroids such as testosterone and estradiol. Low serum levels of SHBG have been associated with obesity, polycystic ovaries, and metabolic syndrome, and other states associated with hyperandrogenemia. A 9-year, 6-month-old girl presented with a history of peripheral precocious puberty and aggressive behavior. The patient’s SHBG level was remarkably low for her age, at less than 5 nmol/L (reference range for a girl with a bone age of 10 years, 73 nmol/L [SEM = 10]) [1]. On genetic and protein analysis, the patient was found to have a homozygous missense potentially pathogenic variant in the SHBG gene (c.554C>T, p.P185L); her parents were asymptomatic heterozygote carriers. Laboratory investigations supported the possible involvement of this genetic alteration in the patient’s phenotype. Various analyses of this variant support its pathogenicity, although the exact mechanism remains unclear. In conclusion, we present a genetic SHBG variant in the homozygote state that may have been associated with gonadotropin-independent precocious puberty in a young girl.

Sex hormone–binding globulin (SHBG) functions as a transport glycoprotein for key sex steroids, such as testosterone (T) and estradiol (E₂), affecting their bioavailability and metabolic clearance [2]. Prior studies have shown SHBG preferentially binds 5α-dihydrotestosterone (5α-DHT) and T and E₂ with approximately 5 and approximately 20 times less affinity than DHT, respectively [3]. It has been postulated that women are more sensitive to excess free androgens as a result of low SHBG levels [2]. Sensitivity to the SHBG/free androgen balance likely plays a role in the development and/or maintenance of obesity, polycystic ovary syndrome (PCOS) (independent of weight status), certain aspects of Cushing syndrome, and abnormalities associated with metabolic syndrome [2, 4]. Pediatric female patients diagnosed with precocious puberty have lower levels of SHBG than their counterparts with normal puberty, independent of body mass index (BMI) [5]. Additionally, variants in the SHBG gene (located on chromosome 17p13.1, OMIM 182205) have been associated with premature ovarian insufficiency, PCOS, and other reproductive dysfunction [6].

One study reported a patient who developed severe hyperandrogenism during pregnancy; she was subsequently found to have nearly undetectable levels of SHBG as a result of a compound heterozygote state for 2 defects, a deletion of the SHBG gene and a single amino acid substitution p.P185L (P156L in the mature SHBG polypeptide genetic sequence), also known as rs6258 [6].

In this report, we describe a patient who presented with gonadotropin-independent precocious puberty and had very low SHBG levels; she was homozygous for the p.P185L SHBG gene missense variant, the first such patient in the literature.

Case Presentation

A 9-year, 6-month-old White girl presented for evaluation of drug-resistant precocious puberty. The patient was recruited by studies approved by the institutional review boards of the National Institutes of Health (NIH) at the NIH Clinical Center, under protocol 2000-CH-0180. The patient’s parents signed an informed consent for all clinical and genetic studies and for the publication of this case. At initial presentation at 5 years, 1 month, she had accelerated growth velocity of 10 cm over the past year, breast budding since age 4 years, 5 months, and vaginal discharge and emotional lability since age 4 years, 11 months. All known family pubertal timings were recorded as follows: her mother experienced menarche at age 12 years, father reports completing puberty in high school, maternal grandmother reported menarche at age 13 years, and all 3 paternal sisters experienced menarche between age 12 and 13 years. Pubertal timings of other grandparents were not known at the time of clinical assessment. She was conceived via assisted reproduction because of a maternal history of infertility and paternal history of low sperm count. The patient had no complications at birth and a history of gross motor delay with a poor gross motor quotient of 76 (poor range, 70-79). The midparental height was 160 cm. Her height was 119.1 cm (+2.1 SD), weight 23.6 kg (+1.6 SD), and a normal BMI of 16.6 (+0.9 SD). On physical examination, she had Tanner stage I pubic hair and Tanner stage II breast development with a 5-cm prepubertal, prominent, and slightly pigmented areola and no axillary hair. All gonadotropins and other hormonal laboratory results were prepubertal with luteinizing hormone (LH) of 0.01 mIU/mL, follicle-stimulating hormone (FSH) of 0.165 mIU/mL, and E₂ of less than 1.0 pg/mL. No serial LH or FSH were available to capture fluctuations in pulsatility and amplitude, nor was a standard gonadotropin stimulation test available. Insulin and insulin-like growth factor-1 levels also were not measured. Additional laboratory evaluation was also within normal limits including free thyroxine of 1.18 ng/dL (age-based reference range, 0.85-1.75 ng/dL) and prolactin of 4.9 ng/mL (reference range for females of any age, 3-24 ng/mL). Her imaging at 5 years, 1 month, showed an advanced bone age at 6 years, 10 months and an enlarged uterus (adjusted for age) and ovaries (right 4.7 cc, left 5.3 cc, without other abnormalities [7-9]. Her brain MRI revealed no abnormalities.

Over the next 6 months, the patient developed behavioral issues, such as aggression and proclivity for obsessive sexual thoughts; her hormonal laboratory findings were, however, unchanged (Table 1). Depot leuprolide 30 mg was initiated at age 5 years, 7 months. Despite leuprolide therapy, she continued to have rapid growth, breast development, and vaginal discharge (see Table 1). At age 5 years, 11 months, she received her second dose of 30 mg of leuprolide. Two months later, she received her third injection at a dose of 15 mg and continued monthly dosing. At age 6 years, 3 months, her bone scan was negative for fibrous dysplasia. At age 6 years, 8 months, she had greater advancement of her bone age to 10 years, 0 months. At this point, the patient was started on letrozole 2.5 mg daily in addition to leuprolide 30 mg every 3 months. Despite changing to a dual-therapy regimen, she further progressed at 7 years, 4 months to Tanner stage III breast development and Tanner stage II pubic hair. Given the likely ineffectiveness of leuprolide in the setting of a clinical diagnosis of gonadotropin-independent precocious puberty, leuprolide was discontinued at age 7 years, 7 months, but she was continued on letrozole. At age 9 years, while on letrozole, she had an episode of menstrual spotting. Her visit 1 month later was concerning for an interval history of hyperphagia. Her measurements were height 149.1 cm (+2.37 SD), weight 56.6 kg (+2.62 SD), and BMI 25.46 (+2.14 SD) with a growth velocity of 6.3 cm/year over the prior year. Given that she had regression of breast tissue, stable pubic hair, and deceleration of bone age progression to 10 years, 8 months, her letrozole was discontinued, and she was referred to the NIH.

Table 1.

Clinical and biochemical evaluations before and after treatment

Age	5 y 1 m	5 y 2 m	5 y 5 m	5 y 6 m	5 y 7 m	5 y 10 m	6 y 1 m	6 y 3 m	6 y 8 m	6 y 9 m	7 y 0 m	7 y 4 m	7 y 7 m	9 y 1 m
Height, cm (SDS)	119.1 (+2.1)	–	121.2 (+2.0)	121.3 (+1.8)	122.1 (+1.8)	124.8 (+2.0)	127.8 (+2.3)	130.3 (+2.5)	132.4 (+2.3)	133.5 (+2.4)	135.3 (+2.3)	137.7 (+2.3)	–	149.1 (+2.37)
BMI, sds	16.6 (+0.9)				17.3 (+1.2)
Tanner stage breast, left/right	3/3	–	2/2	2/2	2/2	2/2	2/2	3/3	2/2	2/2	3/3	3/3	–	2/–
Breasts, left/right, cm	5/5	–	3.5/3.5	4/5	4/4	4/4	6/5	5/5	7/7	5/5	6/7	–	–	–
Tanner stage pubic hair	1	–	1	1	2	2	2	1	1	1	1	2	–	2
Vaginal discharge/menses	–/–	–/–	+/–	+/–	+/–	+/–	+/–	+a/–	+b/–	+/–	+/–	+/–	+/–	+/+
Estradiol, pg/mL	< 1.0	< 1.0	< 1.0	–	1.8	–	2.0	–	2.0	< 1.0	–	–	–	–
LH, mIU/mL	0.010	0.010	0.024	–	0.216	–	0.255	–	–	–	–	–	–	–
FSH, mIU/mL	0.165	0.242	–	–	0.544	–	0.543	–	–	–	–	–	–	–
Bone age	7 y 10m	–	–	–	8 y 5 m	–	–	–	10 y 0 m	–	–	–	–	10 y 8 m
Pelvic ultrasound	Rt ovary 4.7 cc, left ovary 5.3 cc, uterus large for age	Rt ovary 3.5 cc, left ovary 3.5 cc, uterus prepubertal	Rt ovary 6.6 cc, left ovary 4.9 cc, uterus prepubertal		Rt ovary 4.0 cc, left ovary 2.7 cc, uterus 3.8 cc				Rt ovary 2.6 cc, left ovary 2.3 cc, uterus 1.6 cc
Treatment
Leuprolide				Started at 30 mg every 3 moc	30 mg evert 3 mo	30 mg every 3 mo	Increased to 15 mg every 1 moc	15 mg every 1 mo	Decreased to 30 mg every 3 moc	30 mg every 3 mo	30 mg every 3 mo	30 mg every 3 mo	Stopped at visitc
Letrozole										Started at 2.5 mg daily	2.5 mg daily	2.5 mg daily	2.5 mg daily	Stopped at visit

A trial of depot leuprolide was initiated at age 5 years 7 months, and 12 hours after injection, LH was 0.537 mIU/mL, FSH was 1.5 mIU/mL, and estradiol was 11 pg/mL. The patient had continuing rapid linear growth, breast development, and vaginal discharge on leuprolide despite increasing from 30 mg every 3 months to 15 mg every 1 month, confirming that her precocious puberty was not likely central in origin. She was treated with letrozole 2.5 mg daily from age 6 years 9 months to age 9 years and 1 month, with a decrease in the amount of glandular breast tissue and slowing of bone age advancement.

Abbreviations: BMI, body mass index; FSH, follicle-stimulating hormone; LH, luteinizing hormone; Rt, right.

^a Decreased amount of vaginal discharge.

^b Vaginal discharge resolved initially then recurred.

^c Date of drug status or change at that visit.

On arrival at the NIH, she was seen by the pediatric endocrinology team as well as psychiatry. Her mean height was 151.96 cm (+2.4 SD), weight 60 kg (+2.6 SD), and BMI 26 (+2.1 SD) (Fig. 1). No café au lait macules were noted on the patient at presentation. Pubertal examination yielded early Tanner stage III for breast development (measured as right 4 cm, left 5 cm) and Tanner III for pubic hair development, normal clitoral size, and an estrogenized vagina with discharge present. Biochemical evaluation showed pubertal levels of gonadotropins and undetectable T and E₂ levels (Table 2) but she also had extraordinarily low SHBG levels at 4 to 5 nmol/L (reference range for bone age 10 years, 73 [SEM = 10]) [1]. Imaging of the patient’s pituitary and brain did not reveal any abnormalities. She was evaluated by psychiatry and reported having persistent visualizations of murder, episodes of sexual thoughts and behaviors including vaginal penetration, and obsessive-compulsive behaviors since age 8 years.

Figure 1.

Growth chart of stature and weight. Measurements obtained prior to her National Institutes of Health (NIH) evaluation are labeled as a black circle, and measurements from the NIH are an X. The patient’s height and weight were greater than the 95% percentile at age 5 years, 6 months. At age 5 years, 6 months, she was started on leuprolide. Letrozole was added at age 6 years, 9 months. Leuprolide was discontinued at age 7 years, 7 months.

Table 2.

National Institutes of Health clinical and biochemical evaluation

Measurements/laboratory results, units	Value	Reference range
Height, cm	151.96, 99.3%ile, +2.4 SD
Weight, kg	60, 99.6%ile, +2.6 SD
BMI	26, 98.3%ile, +2.1 SD	Elevated > 85%ile
Blood pressure, mm Hg
Systolic	101 (< 50%ile)	Elevated > 90%ile
Diastolic	70 (50-90%ile)	Elevated > 90%ile
TS breast	Early III
TS pubic hair	III
Follicle-stimulating hormone, mIU/mL	1.8	Early puberty: 1.0-12.8
		Late puberty: 1.0-11.7
Luteinizing hormone, mIU/mL	1.7	Early puberty: 0.02-12.0
		Late puberty: 0.40-11.7
Dehydroepiandrosterone serum, ng/dL	271	0-599
Sex hormone–binding globulin, nmol/L	4	For bone agea
		< 10 y: 77 ± 7 (44-127)
		10 y: 73 ± 5 (39-131)
		11 y: 65 ± 7 (39-118)
Testosterone, ng/dL	< 20.0	Premenopausal < 81
Cortisol, mcg/dL	7.9	5.0-25.0
Dihydrotestosterone, pg/mL	< 50	TS II+: ≤ 300
Adrenocorticotropin, pg/mL	43	5.0-46.0
17-OH progesterone, ng/dL	14	< 13-6216
Estradiol serum, pg/mL	< 5	TS I: undetectable-20
		TS II: undetectable-24
		TS III: undetectable-60
Thyrotropin, mcIU/mL	5.95	Female 6 y-10 y: 0.48-4.20
		Female 11 y-15 y: 0.76-4.20
Growth hormone, ng/mL	0.39	0.00-8.00
Prolactin, mcg/L	6.7	2.0-25.0

Physical and biochemical evaluation was completed at the NIH Clinical Center in Bethesda, Maryland, at patient age 9 years and 6 months. She was noted to be above the 98th percentile for height, weight, and BMI. Blood pressure percentile based on height and weight: normal less than 90th percentile. She was found to be early TS III for breast development (right 4 cm, left 5 cm) and TS III for pubic hair development. Additionally, she was found to have above normal thyrotropin and a below normal level of sex hormone–binding globulin. All other laboratory results were within normal limits. Laboratory values and reference ranges are from the NIH Clinical Center unless otherwise specified.

Abbreviations: %ile, percentile; BMI, body mass index, NIH, National Institutes of Health; TS I, Tanner stage I, TS II, Tanner stage II, TS III, Tanner stage III.

^a Reference range as mean ± SEM with minimum and maximum values [1].

Genetic Studies

Whole-exome sequencing was performed (obtained at GeneDx), and the patient was found to have a rare (Genome Aggregation Database frequency = 0.4%) homozygous variant that was predicted to be a pathogenic variant in SHBG (NM_001040.5, rs6258, c.554C>T, p.P185L). Both parents were heterozygous for the same sequence alteration, and her unaffected brother was not a carrier.

There were no other pathogenic variants in the sequence. We also looked carefully at the KISS1, KISS1R, PROKR2, MKRN3, and DLK1 genes and no such variations were found within the patient’s whole-exome sequence.

While looking for a possible impairment of the protein function, the variant was found and evaluated by MutationTaster (http://www.mutationtaster.org), Polymorphism Phenotyping v2 algorithm tool (PolyPhen-2) (http://genetics.bwh.harvard.edu/pph2), and the Sorting Tolerant From Intolerant (SIFT) algorithm (http://sift.jcvi.org). The variant was predicted to be likely pathogenic by all in silico analyses. The SHBG promoter region (1400 bp) was amplified by end point polymerase chain reaction (GoTaq Green Master Mix, Promega) and subjected to direct bidirectional sequencing (BigDye Terminator 3.1 Cycle Sequencing Kit, Applied Biosystems) using a 3500xL Genetic Analyzer (Applied Biosystems). Primers for polymerase chain reaction and Sanger sequence are available on request.

Analyses of Sex Hormone–Binding Globulin

An established ligand saturation analysis [10] was used to determine the steroid-binding capacity of SHBG heparinized plasma from the patient and her relatives. Briefly, plasma samples were diluted (1:100) in and preincubated (30 minutes at room temperature) with a dextran-coated charcoal (DCC) slurry in phosphate-buffered saline assay to remove endogenous steroids. After centrifugation to remove DCC, supernatants were incubated with [³H] DHT in the presence or absence of 100-fold molar excess of unlabeled DHT to monitor nonspecific binding. Unbound steroids were removed by exposure to DCC for 10 minutes at 0 °C followed by centrifugation to allow SHBG-bound [³H]DHT to be measured. The relative binding affinities (RBAs) of steroids were determined using [³H]DHT as the labeled ligand and increasing amounts of DHT, T, or 17β-E₂ as competitors [10] in culture medium from CHO cells expressing wild-type (WT) SHBG or the SHBG p.P185L variant. In addition, plasma SHBG levels were also determined using a Delfia time-resolved immunofluorescence assay [11].

As shown in Fig. 2, SHBG mass concentration measured by Delfia vs steroid-binding capacity in the proband and family members behaved normally within the experimental error of each assay and correlated well, indicating no abnormality in the steroid-binding affinity of SHBG of the variant protein [12]. The SHBG mass concentration for the WT brother was within the expected normal reference ranges [13]. The SHBG levels for the heterozygous mother and father were about half the mean of the normal ranges (ie, ~ 50 nmol/L in women and ~ 30 nmol/L in men), and for the proband they were far below the normal range for female children of a similar age using these assays [1].

Figure 2.

Sex hormone–binding globulin (SHBG) binding assay vs SHBG levels. The graph contrasts the SHBG mass concentration measured by the Delfia assay (x-axis) vs the steroid-binding capacity of each patient’s SHBG (y-axis). The line on the graph is the expected perfect correlation between the Delfia and SHBG binding capacity assay values. The SHBG mass concentration for the heterozygous mother and father were about half the mean of the normal ranges, the wild-type brother’s SHBG mass concentration was within the expected normal reference range (ie, ~ 50 nmol/L in women and ~ 30 nmol/L in men), and the proband’s SHBG mass concentration and steroid-binding capacity were well aligned with each other and very much lower than the normal range for female children of a similar age (minimum being 44 nmol/L).

When the steroid-binding affinities of the WT SHBG and p.P185L variant SHBG gene studied in CHO cells were determined, their RBAs for T and E₂ were similar, although the mutant protein exhibited a slightly higher RBA for E₂ vs WT SHBG (Fig. 3). The reason for this is unclear because the P156 residue in the mature protein structure is on the surface of the protein, far from the steroid-binding site and not located in the vicinity of proposed entrances of E₂ to the steroid binding site [12]. It is, however, close to residues that form a calcium-binding pocket that plays a key role in determining steroid-binding affinity [14]. Additionally, a minigene assay was used to determine whether this genetic variant affects protein splicing, and no differences in splicing or expression between the WT and mutant SHBG genes were identified (data not shown).

Figure 3.

Steroid-binding properties of wild-type (WT) sex hormone–binding globulin (SHBG) and SHBG p.P185L variant. A competitive steroid binding assay of recombinant WT SHBG gene vs p.P185L SHBG variant produced in CHO cells using [H³]5-dihydrotestosterone (DHT) as labeled ligand demonstrates that the SHBG genetic variant has a normal affinity for DHT and testosterone but a slightly higher affinity for estradiol (E₂) when compared to the WT SHBG gene.

Discussion

We describe a young girl whose rapid growth, premature thelarche, and aggressive behavior led to a diagnosis of gonadotropin-independent precocious puberty and the discovery of very low SHBG levels. Whole-exome sequencing revealed that the patient was homozygous for a previously described (in the heterozygote state), likely to be pathogenic, missense variant of the SHBG gene.

Human SHBG is a homodimeric glycoprotein [15] that binds sex steroids, such as T, E₂, and DHT [13]. Plasma SHBG plays a key role in the transportation, bioavailability, and metabolic clearance of these steroids to their target tissues [6]. Low plasma levels of SHBG and hyperandrogenism have been found in patients with PCOS, type 2 diabetes mellitus, and obesity [4], similar to our patient. The pathophysiology between this association is unclear but given our patient’s hyperandrogenism and SHBG deficiency, her presentation, including obesity, is likely multifactorial. Thyroid hormones and weight reduction have opposing effects with high SHBG and subsequently lower free androgen levels [16]. While the cause of low SHBG in hyperandrogenic states is unknown, prior studies have alluded to an increased metabolic clearance of SHBG, or a reduction in SHBG secretion by the hepatocytes [17].

We speculate that an increase in free sex hormone levels due to an exceptionally low plasma SHBG level was responsible for our patient’s gonadotropin-independent precocious puberty and symptoms associated with elevated sex hormones. Her ovaries were somewhat large for a prepubertal girl, which was an unexpected finding in the setting of gonadotropin-independent precocious puberty, and no cysts or masses were noted, ruling out ovarian cysts or masses as a cause for her clinical findings. We speculate that the increase in ovarian size may be an early indicator of impending PCOS, given her subsequent development of obesity and hyperandrogenism, although this was not further evaluated in the context of her other clinical findings. Of note, her behavioral symptoms were likely caused by hyperandrogenism, which has been described in patients with familial male precocious puberty [18] and congenital adrenal hyperplasia [19, 20]. Girls with congenital adrenal hyperplasia have been noted to have various behavioral issues [19], and the literature suggests causation between hyperandrogenism and emotional, conduct, and hyperactivity-inattention symptoms [20]. For our patient, her gonadotropin-independent precocious puberty appropriately did not respond to treatment with a gonadotropin-releasing hormone agonist, although the exact mechanism for her clinical signs remains unclear. The germline variant p.P185L in SHBG that was identified as homozygous in the proband, heterozygous among her parents, and not present in her brother, is predicted to be pathogenic by all in silico analyses.

Ohlsson et al analyzed the SHBG affinity for testosterone using Scatchard plots (using [3H]T) of SHBG in serum of men for the variant p.P185L, and revealed a higher mean dissociation constant indicative of a lower affinity in the homozygous mutant than the WT [21]. However, in the Ohlsson study the serum proband samples were diluted less than the WT, which is problematic because of its influence on SHBG-binding affinity. In that study the relative affinity of this variant for E₂ was not determined, and this has also not been reported previously. The reason why this p.P185L variant might have altered affinity for E₂ is unknown but the substitution of a proline in this location by leucine may influence the positioning of neighboring residues that participate in binding calcium, which influences the steroid-binding properties of SHBG [14].

It is important to note that the SHBG levels found in the proband ranged between 4 and 5 nmol/L, which is far lower than the minimum normal value of 44 nmol/L in girls of similar age [1]. On biochemical evaluation of her parents, they were found to have lower SHBG concentrations, around half the mean of the normal ranges. We also checked the KISS1, KISS1R, PROKR2, MKRN3, and DLK1 genes and no variations were found within the patient’s whole-exome sequence.

To determine if the very low levels of SHBG in the proband were associated with mutation in cis-regulatory sequences within the SHBG promoter that drives expression in the liver, we sequenced 1400 bp upstream of the SHBG transcription start site. These sequences in the proband perfectly matched the consensus sequences. This region also contained the TAAAA polymorphic region (repeats length varying between 5 and 10) that influence SHBG transcription levels [22], with repeat numbers over 6 being associated with modest reductions in plasma SHBG levels. The proband was heterozygous for 7 and 8 TAAAA repeats while her mother and father were homozygous for 7 and 8 TAAAA repeats, respectively. Since the Genome Aggregation Database frequency for 7 repeats is 7.42%, a TAAAA repeat polymorphism does not account for the very low plasma levels of SHBG in the proband, and we were left with 3 explanations for this. There is either a long-range effect of the genetic mutation within the mutant SHBG gene that reduces its transcription; there is an abnormal abundance of key transcription factors that control SHBG transcription in the liver, such as HNF-4α or COUP-TF [23], or the plasma clearance rate of the mutant protein is abnormally fast. A deeper analysis of these possibilities in this patient and her family might provide new information on the regulation of plasma SHBG levels.

We conclude that a female patient homozygous for the p.P185L genetic variant of SHBG presented with gonadotropin-independent precocious puberty. The patient did not have any other genetic variants in her whole-exome sequencing, and it appears possible that the alteration in the SHBG gene may be at least contributory to the observed phenotype.

Glossary

Abbreviations

5α-DHT

5α-dihydrotestosterone

BMI

body mass index

DCC

dextran-coated charcoal

E₂

estradiol

FSH

follicle-stimulating hormone

LH

luteinizing hormone

NIH

National Institutes of Health

PCOS

polycystic ovary syndrome

RBA

relative binding affinity

SHBG

sex hormone–binding globulin

T

testosterone

WT

wild-type

Additional Information

Disclosures: C.A.S. holds patents on the function of the PRKAR1A, PDE11A, and GPR101 genes and related issues; his laboratory has also received research funding on the GPR101 gene, abnormal growth hormone secretion, and its treatment by Pfizer, Inc; F.R.F. holds a patent on the GPR101 gene and/or its function. Since the submission of this manuscript C.A.S was named chief scientific officer of ELPEN, Inc and now serves as a consultant to Lundbeck, Pharma, Inc, and Sync, Inc; he is also the director of genetics and precision medicine at FORTH (www.forth.gr); none of these activities are in conflict with the contents of this manuscript. The other authors have nothing to disclose.

Data Availability

Some or all data sets generated during and/or analyzed during the present study are not publicly available but are available from the corresponding author on reasonable request.