Disorders caused by genetic defects associated with GH-dependent genes: PAPPA2 defects.

Abstract

Growth hormone (GH) and its mediator, insulin-like growth factor-1 (IGF-1), have long been recognized as central to human growth physiology. IGF-1 is known to complex with IGF binding proteins as well as with the acid labile subunit (ALS) in order to prolong its half-life in circulation. Factors regulating the bioavailability of IGF-1 (i.e. the balance between free and bound IGF-1) were less well understood. Recently, pregnancy-associated plasma protein-A2 (PAPP-A2) was discovered as a protease which specifically cleaves IGF-binding protein (IGFBP)-3 and -5. PAPP-A2 deficient patients present with characteristic findings including growth failure, elevated total IGF-1 and -2, IGFBPs, and ALS, but decreased percentage of free to total IGF-1. Additionally, patients with PAPP-A2 deficiency have impairments in glucose metabolism and bone mineral density (BMD). Treatment with recombinant human IGF-1 (rhIGF-1) improved height SD scores, growth velocity, body composition, and dysglycemia. Mouse models recapitulate many of the human findings of PAPP-A2 deficiency. This review summarizes the function of PAPP-A2 and its contribution to the GH–IGF axis through an examination of PAPP-A2 deficient patients and mouse models, thereby emphasizing the importance of the regulation of IGF-1 bioavailability in human growth.

1. Introduction

The growth hormone (GH) – Insulin-like growth factor (IGF) axis is widely known as one of the critical biological pathways involved in regulating human growth. It has been investigated by many researchers since the late 1940s (Yakar et al., 2018). The GH–IGF axis has traditionally consisted of Somatostatin, GH releasing hormone (GHRH), GH, IGFs, their receptors, and downstream signaling molecules including STAT5B, IGF-binding proteins (IGFBPs), and the acid-labile subunit (ALS). In this manuscript, we will discuss a more recently discovered component to this axis, a regulator of IGF bioavailability, pregnancy-associated plasma protein-A2 (PAPP-A2) (Overgaard et al., 2001), a protease that cleaves IGF binding protein-3 and -5. The secretion of GH from the anterior pituitary is stimulated by GHRH, released by the hypothalamus. Sequentially, GH promotes the production of IGFs, IGFBPs, and ALS from the liver and other tissues. Under normal circumstances, free IGF-1 can circulate with a half-life of up to 10 min in serum. However, the majority of IGF-1 exists in a ternary complex with IGFBP-3 or -5 and the ALS, resulting in an increased half-life of IGF-1 (up to 15 h) (Renes et al., 2019,Yakar et al., 2018). As free IGF-1 is the only form that binds to the IGF receptors, IGF-1 must be liberated from IGFBPs. Therefore, a protease such as PAPP-A2 and/or other proteases are required to cleave IGFBPs, subsequently leading to increased IGF signaling resulting in cell proliferation and growth.

In this review, we discuss PAPP-A2’s contribution to human growth physiology and describe the effects of PAPP-A2 deficiency on serum biochemistry, glucose metabolism, and bone development by exploring the phenotype of human patients bearing PAPPA2 mutations.

2. Physiology and function of PAPP-A2

In 2001, Overgaard et al. (2001) identified and cloned cDNA encoding the PAPP-A2 protein, a protein with 46% homology to the functionally related protein PAPP-A (Overgaard et al., 2001). The PAPPA2 gene, located at chromosome 1q25.2, contains the elongated zinc-binding motif, HEXXHXXGXXH, which works as the catalytic zinc-binding site, similar to PAPP-A (Boldt et al., 2001). PAPP-A2 and PAPP-A were classified as “pappalysins,” new members of the metzincin family of metalloproteinases (Boldt et al., 2001). Given that PAPP-A was known to target select IGF binding proteins for proteolysis, the experiments incubating PAPP-A2 with each of the six IGFBPs revealed that IGFBP-3 and -5 were specific substrates for PAPP-A2 (Overgaard et al., 2001). PAPP-A2 is expressed abundantly in the human placenta, and it’s also detected in other tissues throughout the body, including the kidney, mammary tissue, pituitary, brain, and pancreas. Unlike PAPP-A which cleaves its target IGFBP-4 while anchored to the cell surface, PAPP-A2 does not bind the cell surface but is rather found in circulation (Laursen et al., 2002,Oxvig, 2015). PAPP-A2 protein has been detected at extremely high levels in the circulation of women during pregnancy (Kloverpris et al., 2013,Wang et al., 2009). Therefore, it is thought that PAPP-A works as a local modulator for IGF-1 bioavailability by cleavage of IGFBP-2, -4, and -5 (Argente et al., 2017,Oxvig, 2015). However, PAPP-A2 may play a larger role in increasing serum free IGF-1 concentration systemically via liberating IGF-1 from its ternary complex with IGFBP-3 or -5 and ALS. It is also possible that PAPP-A2 has local tissue effects as well but this requires further exploration.

Stanniocalcin-1 (STC1) and -2 (STC2) are secreted glycoproteins that are widely expressed in human and mouse tissues and are highly conserved throughout evolution from fish to higher vertebrates (Chang et al., 2008,Chang et al., 1998). Kløverpris et al. (2015) and Jepsen et al. (2015) reported that STCs work as protease inhibitors for PAPP-A and PAPP-A2, suggesting that they too may play an important role in regulating the GH–IGF axis (Jepsen et al., 2015,Kloverpris et al., 2015). Interestingly, while STC1 appears to play a role in calcium regulation, transgenic overexpression of human STC2 in mice causes growth failure (Gagliardi et al., 2005). To date, there have been no human cases with a pathogenic mutation in either STC1 or 2.

Recent genomic studies have highlighted the importance of PAPPA, PAPPA2, and STC2 in height biology. In 2010, the Genetic Investigation of Anthropometric Traits Consortium performed a large genome-wide association study (GWAS) examining the role of common genetic variation in determining adult stature (Lango Allen et al., 2010). They identified 180 different loci associated with height and then examined the biological connections between the genes within these loci. This analysis highlighted a number of biological pathways and included the PAPPA, PAPPA2, and STC2 genes. In a follow-up GWAS (Marouli et al., 2017), the same group of investigators identified 83 rare and low frequency coding variants to be associated with height. Of these 83 variants, a missense variant in STC2 was found to have the largest effect on height. The heights of carriers with this rare STC2 missense variant were approximately 2.1 cm taller than non-carriers. These studies highlight that genetic variation around the PAPPA, PAPPA2, and STC2 genes play a role in determining an individual’s height in the general population.

3. PAPP-A2 deficiency in humans

3.1. Descriptions of patients with PAPP-A2 deficiency

In 2016, the first two families bearing rare mutations in PAPPA2 were identified using whole-exome sequencing (Dauber et al., 2016). The two families were of Spanish (Family 1) and Palestinian (Family 2) ancestry and the affected children presented with short stature and markedly elevated IGF-1 levels. In the Spanish family, two of four siblings were found to carry a novel homozygous frameshift mutation in PAPPA2 (c.1927_1928insAT, p.D643fs25*). The affected children presented with heights in the low-normal range (−1.1 and −0.96 SDS) but significantly below their mid-parental target height. Their biochemical profiles at presentation were notable for extreme elevations in IGF-1 and milder elevations of IGFBP-3. PAPP-A2 levels were undetectable. Subtle dysmorphic features consisted of small chins, microcephaly, and long fingers and toes (Dauber et al., 2016). In the second family, three of five siblings carried a novel homozygous missense mutation in PAPPA2 (c.3098C > T, p.A1033V). They presented with more severe postnatal growth retardation and short stature with heights ranging from −2.8 to −3.8 SDS. They had a very similar biochemical profile with extreme elevations in IGF-1 and milder elevations of IGFBP-3 but these patients had detectable levels of PAPP-A2. In vitro studies demonstrated that the missense mutation led to the inability of the mutated PAPP-A2 to cleave IGFBP-3 and IGFBP-5. Similar to the children in the Spanish family, the affected siblings, had small chins, microcephaly, long thin fingers, and delayed dental eruption. In addition, two of the five affected patients were born small for gestational age (SGA). The parents of both families bearing heterozygous mutations in PAPPA2 were of normal height. Follow up studies of the growth profile from the two older siblings in the second family showed the absence of an apparent pubertal growth spurt (Cabrera-Salcedo et al., 2017,Dauber et al., 2016).

3.2. Biochemical consequences of PAPP-A2 deficiency

As noted above, PAPP-A2 protein cleaves IGFBP-3 and -5 thereby freeing IGF-1 from its ternary complex allowing it to bind to its receptor. Therefore, PAPP-A2 deficiency should result in a decrease in free IGF-1 levels leading to a lack of negative feedback at the level of the pituitary ultimately resulting in increased growth hormone levels. This subsequently should cause an increase in all of the growth hormone dependent proteins including total IGF-1 itself as well as IGF-2, IGFBP-3, IGFBP-5, and ALS levels (Figure 1). Further investigations into the affected patients supported this hypothesis. All affected patients presented with elevated serum total IGF-1 and IGFBP-3 levels (Dauber et al., 2016). Bioactive IGF-1 levels measured by a cell-based kinase receptor activation (KIRA) assay (Chen et al., 2003,Sorensen et al., 2015) were reduced in the three prepubertal patients, while those levels in pubertal patients were within normal range. Additionally, free IGF-1 levels measured by enzyme-linked immunosorbent assay (ELISA) showed the same trend as bioactive IGF-1 levels. However, when analyzed as a percentage of bioactive IGF-1 or free IGF-1 compared to the total IGF-1 (i.e. bioactive/total IGF-1 or free/total IGF-1), these percentages were significantly decreased in all patients (Fujimoto et al., 2020,Sorensen et al., 2015). This suggests that the absolute circulating levels of free IGF-1 may be a less important biomarker of PAPP-A2 function, and perhaps of IGF-1 action, than the percentage of free IGF-1. Additionally, extreme elevations in total IGF-I values are a useful clinical clue to suggest the diagnosis of PAPP-A2 deficiency. While mild elevations may be present in heterozygous IGF1R mutations, the degree of elevation seen in the PAPP-A2 deficient patients would only be seen in individuals with homozygous mutations of IGF1R which are extremely rare and lead to severe short stature and microcephaly. As predicted by the hypothesis stated earlier, the patients’ spontaneous GH secretion or peak GH values after stimulation tests were also elevated. Furthermore, other growth hormone dependent factors including serum IGF-2, IGFBP-5, and ALS levels were elevated in the majority of patients. All of the heterozygous unaffected siblings had normal biochemical profiles.

Figure 1. Schematic of the GH-IGF axis highlighting the effects of PAPP-A2 deficiency.

PAPP-A2 deficiency results in insufficient cleavage of the ternary complex containing IGFs, IGFBP-3 or -5, and ALS, resulting in decreased levels of free IGF-1. Reduced free IGF-1 leads to a lack of negative feedback at the level of the pituitary, resulting in increased GH secretion. Ultimately, this leads to increased IGFBPs, ALS, and total IGF-1 without an increase in free IGF-1.

3.3. Glucose metabolism in PAPP-A2 deficiency

PAPP-A2 most likely plays a role in glucose metabolism. Fasting glucose levels in all PAPPA2 deficient patients were normal to high-normal with hyperinsulinemia suggesting a degree of insulin resistance (Cabrera-Salcedo et al., 2017,Dauber et al., 2016,Munoz-Calvo et al., 2016). Oral glucose tolerance tests performed in the second family revealed that two of the patients had elevated blood glucose levels (141 and 152 mg/dL, respectively) at the two-hour time point. The exact mechanism of dysglycemia in our patients is not completely known but may be due to the effects of high growth hormone levels and low free IGF-1 levels caused by PAPP-A2 deficiency. Elevated baseline GH levels induce insulin resistance via effects on hepatic gluconeogenesis, glyconenolysis, and lipolysis (Pal et al., 1992,Vijayakumar et al., 2010). However, the administration of a very low dose of recombinant GH in healthy young adults (19–29yrs) improved insulin sensitivity presumably through the effects of increases in free IGF-1 (Yuen et al., 2004). The administration of recombinant IGF-1 for patients with type 1 and type 2 diabetes mellitus, without PAPP-A2 deficiency, increased serum IGF-1 levels resulting in improved insulin sensitivity (Simpson et al., 2004). This data suggests that free IGF-1 levels may play an important role in modulating insulin sensitivity (Yuen et al., 2007).

3.4. Bone mineral density (BMD) in patients with PAPP-A2 deficiency

All patients had varying degrees of decreased bone mineral density (BMD) with the majority of patients having BMD in the low-normal range (Argente et al., 2019,Cabrera-Salcedo et al., 2017,Hawkins-Carranza et al., 2018). In Family 1, bone microarchitecture strength was measured using the Trabecular Bone Score (TBS). TBS was mildly low in one of the siblings and increased in both siblings after two years of treatment with recombinant human IGF-1 (rhIGF-1). In Family 2, all of the patients had bone density below the mean for the population with one of the brothers having a whole-body BMD z-score below −2 SD. The youngest sibling showed significant gains in BMD after 1 year of treatment with rhIGF-1.

3.5. Treatment of patients with PAPP-A2 deficiency

Currently, there is no approved treatment for PAPP-A2 deficiency. However, two therapeutic approaches have been attempted; rhIGF-1 and plasma transfusion as a means of PAPP-A2 replacement (Andrew et al., 2018,Cabrera-Salcedo et al., 2017,Munoz-Calvo et al., 2016). Four of the five affected patients underwent treatment with subcutaneous rhIGF-1 injection (40–120 μg/kg/dose, twice a day, progressively increased). In Family 2, a detailed pharmacokinetic study was performed. After 120 μg/kg/day of subcutaneous rhIGF-1 injection, the peak free IGF-1 levels were measured and the pharmacokinetic parameters were similar between the patients, their heterozygous relatives and unaffected controls (Cabrera-Salcedo et al., 2017). Three of the patients were successfully treated with rhIGF-1 therapy resulting in a first-year change in height SDS of +0.4 SDS in all three patients and accelerated growth velocity from −1.97 SD to +1.29 SD. Additionally, treatment resulted in resolution of their hyperinsulinemia and improvement in fasting blood glucose levels. Treatment outcomes of the patients are summarized in Table 1. Basal GH levels decreased with rhIGF-1 treatment which may be the cause of the improvement seen in glucose metabolism.

Table 1.

Treatment outcomes

Case	Sex	Treatment period	rhIGF-1 therapy	Age (y.o.)	Height, cm (SDS)	ΔHtSDS	GV, cm/y (SDS)		Bone age (y.o.)	BMD (SDS)	Body composition change
F1-P1	F	0	40–80 μg/kg, twice daily	10.5	132 (−1.25)	-		3.7 (−1.5)	10.8	−0.1*	-
		12M	120 μg/kg, twice daily	11.5	139.6 (−0.86)	+0.4		7.6 (+1.6)	11.3	−0.4*	Increases in the percentage of LBM (+4.8 %) and decreases in PBFM (-4.5 %)
										After 24M, −0.3*
F1-P2	M	0	40–80 μg/kg, twice daily	6.1	111.5 (−0.74)	-		5.8 (-1.6)	4.5	−0.1*	-
		12M	120 μg/kg, twice daily	7.1	118.5 (−0.34)	+0.4		7.0 (+1.1)	5.5	−0.7*	Increases in the percentage of LBM (+5.0 %) and decreases in PBFM (-4.7 %)
										After 24M, −0.2*
F2-P1	F	Untreated	-	18.6	138.4 (−3.81)	-		NA	NA	−0.5†	NA
F2-P2	M	0–1	60 μg/kg, twice daily	14.5	142.4 (−2.9)	-		4.3 (−1.6)	13.5	−2.5†	-
		12M		15.5	147.2 (−3.0)	−0.1		4.8 (+1.5)	14.5	−2.2†	Increases in LBM and decreases in PBFM
F2-P3	M	0	60	10.4	122.0 (−2.9)	-		3.0 (−2.8)	9.0	−0.7†	-
		12M	120 μg/kg, twice daily	11.4	128.2 (−2.5)	+0.4		6.2 (+1.2)	10.0	−0.1†	Increases in LBM and decreases in PBFM
F1-P1	F	0	957	1.4	ND‡	5912	3745 (U/L)	4.8	15.1
		12M	No significant change	No significant change in bioactive IGF-1	NA	No significant change	NA	0.8	11
F1-P2	M	0	882	0.3	ND‡	4850	3625 (U/L)	5.3	27
		12M	No significant change	No significant change in bioactive IGF-1	NA	No significant change	NA	0.7	8
F2-P1	F	Untreated	1060	2.0	0.2	4557	2445 (U/L)	NA	Hyperinsulinemia
F2-P2	M	0	573.2	5.6	0.1	8997	23	0.1	17.1
		12M	756.2	6.1	0.1	8983	17	0.1	Hyperinsulinemia
F2-P3	M	0	546.8	0.8	0.1	>10000	18	0.2	7
		12M	709.1	1.2	0.2	6362	14	0.1	Improved

^*The height-adjusted Z-score was for the lumber spine.

^†The height-adjusted Z-score was for the whole body less head.

^‡ELISA kit could not detect affected PAPP-A2 caused by the frameshift mutation.

Abbreviations: BA, bone age; BMI, body mass index; HC, head circumference; GV, growth velocity F1, Spanish family; F2, Palestinian family; P1, P2, P3 – Proband 1, 2, 3; LBM, lean body mass; PBFM, percentage of total body fat mass; ND, not detectable; NA, not available

In one of the four treated patients, intracranial hypertension and headache occurred early during rhIGF-1 treatment resulting in discontinuation of therapy (Dauber et al., 2016). No patient experienced any episodes of hypoglycemia.

Prior to initiation of rhIGF-1 treatment in Family 2, an attempt was made to explore the possibility of PAPP-A2 replacement therapy via plasma transfusion. A single transfusion of 20 mL/kg of fresh frozen plasma was given to the eldest affected sibling with PAPP-A2 deficiency (Andrew et al., 2018). The plasma transfusion resulted in a 2.5-fold increase of free IGF-I levels on day 1 post-transfusion but there were no significant changes in total IGF-1, total IGFBP-3 or intact IGFBP-3 levels. Additionally, no detectable changes in PAPP-A2 levels were found suggesting that this transfusion did not substantially increase circulating PAPP-A2 levels. However, the increase in free IGF-1 raises the possibility that recombinant PAPP-A2 could be an effective medication for PAPP-A2 deficiency and growth failure. Additional studies are needed to understand the pharmacokinetics and pharmacodynamics of PAPP-A2.

4. Regulation of GH–IGF axis related proteins by PAPP-A2

4.1. Insights from mouse studies

To date, there have been three mouse models described with PAPP-A2 deficiency and each shows reduced body weight and size with significant decreases in free IGF-1 compared to the wild-type control. Conover et al. (2011) described the first global knock-out (KO) mouse model in which PAPP-A2 protein was absent from all tissue in which it is normally expressed (Conover et al., 2011). Serum from these KO mice was subsequently assayed for free and total IGF-1, and similar to the human patients, PAPP-A2 deficiency led to markedly decreased levels of free IGF-1 despite elevation in total IGF-1 (Dauber et al., 2016). Next, Christians et al. (2015) created a second Pappa2 KO mouse model (Christians et al., 2015). Like the KO created by Conover, Christians saw a significant increase in total IGF-1 in the KO mouse (945 ± 24 ng/mL) compared to littermate wild-type (WT) controls (582 ±28 ng/mL). Finally, a knock-in (KI) mouse model (p.V1034A) was created via CRISPR-Cas9 which recapitulates the human PAPPA2 missense mutation (p.V1033A) (Fujimoto et al., 2019). Similar to the Christians model, homozygous KI mice both show extremely reduced free IGF-1 levels compared to WT littermate controls and markedly higher circulating levels of total IGF-1 in the homozygous KI compared to WT with heterozygotes having an intermediate phenotype.

As seen in the humans, the pattern of reduced body size is noted in every mouse model paper to date. The first KO mouse model showed postnatal growth retardation in male and female mice of 10% and up to 30% compared to WT controls, respectively (Conover et al., 2011). The reduction of body weight and length were also noted in the KI mouse model. Again, males showed ~10% reduction in body length and weight, while females show a 20% reduction in body weight compared to littermate WT controls, both with reduced relative lean body mass (Fujimoto et al., 2019). However, in a reanalysis of the data from Fujimoto et al. (2019), there was no significant sex by genotype interaction either for body length or body weight at 8 and 16 weeks. Similar to the biochemical data, the heterozygous KI and KO mice display as expected, an intermediate phenotype compared to the homozygous KI and WT littermates (Christians et al., 2015,Fujimoto et al., 2019). In addition, bone appears to be affected in all three described mouse models. Embryonic KO mice at E16.5 have smaller skeletons compared to wild-type controls (Conover et al., 2011). In the KI mice, adults have inhibited periosteal bone expansion which results in slender bones (Fujimoto et al., 2019). However, bone mineral density does not appear to be affected in either model. Also reported, Pappa2 KO mice at all ages have shorter bone growth (Christians et al., 2018) and without PAPP-A2 produced locally in bone osteoblasts, normal growth will not occur (Amiri et al., 2015). Finally, in a thorough examination of bone regulation in a Pappa2 deletion mouse model, Christians et al. (2019) showed that PAPP-A2 plays a role in bone structure and mass through a possible regulation of IGFBP-5 in a sex- and age- specific manner (Christians et al., 2019).

In further biochemical studies, Fujimoto et al. (2019) monitored circulating levels of intact IGFBP-3 as a measure of PAPP-A2 activity. As such, serum collected from the KI homozygous mice showed significant increases in intact IGFBP-3 recapitulating the human increases in total IGFBP-3 seen in those with PAPP-A2 deficiency. However, Christians et al. (2015) did report elevations of IGFBP-5 but IGFBP-3 levels showed an approximate 15-fold decrease in KO mice compared to WT controls. As of today, there is no sufficient explanation for these differences in IGFBP-3 levels between models.

To follow up on the human data suggesting impaired glucose tolerance, Fujimoto et al. (2019) performed glucose and insulin tolerance tests (GTT and ITT) in the KI mouse model. In GTT, homozygous KI mice presented significantly elevated glucose levels after glucose injection (glucose values; 422 mg/dL (KI-M), 307 mg/dL (KI-F); 323 mg/dL (WT-M), 234 mg/dL (WT-F)). In ITT, homozygous KI mice did not decrease blood glucose significantly compared to WT after insulin injection (the lowest glucose values; 221 mg/dL (KI-M), 175 mg/dL (KI-F); 125 mg/dL (WT-M), 119 mg/dL (WT-F)). Therefore, both insulin resistance and glucose intolerance were noted in homozygous KI mice. KI male and female mice both show significant decreases in % lean mass and increased levels of % fat mass. These changes in lean muscle and fat mass correspond with insulin resistance in humans and mirror the changes seen in the PAPP-A2 deficient humans (Cabrera-Salcedo et al., 2017). In sum, the totality of the mouse data supports a causal role of PAPP-A2 deficiency in short stature and associated metabolic derangements seen in the human patients.

4.2. Relationship between PAPP-A2 and GH–IGF axis related proteins in the general population

Recently, Fujimoto et al. (2020) and DiPrisco et al. (2019) assayed PAPP-A2 levels throughout childhood or at birth and examined the relationship between PAPP-A2 levels and GH-related proteins (DiPrisco et al., 2019,Fujimoto et al., 2020). PAPP-A2 levels consistently decreased throughout childhood (3–18 years) (Age, 50%ile value [male, female] (ng/mL); 3.5 year-old, 0.345, 0.334; 17.5 year-old, 0.104, 0.132). Mean cord blood PAPP-A2 levels in infants were 2.2 ng/mL. However, in adulthood (20–70 years), PAPP-A2 levels were positively correlated with age (Steinbrecher et al., 2017). In short, these findings suggest that PAPP-A2 levels show a U-shaped curve throughout the lifespan. As expected, PAPP-A2 had a strong inverse correlation with serum intact IGFBP-3 levels, while positively correlated with the percentage of free to total IGF-1 in childhood (Fujimoto et al., 2020). This supports PAPP-A2’s role in regulating IGF-1 levels on a larger population wide scale.

DiPrisco et al. (2019) reported that PAPP-A2 levels in cord blood inversely associated with birth weight and length (DiPrisco et al., 2019). PAPP-A2, even in adjusted linear regression models for gestational age, race, sex, delivery mode, maternal age, and Apgar scores, was inversely associated with birth length Z-score (β = −0.279, P = 0.017). Cord blood total IGF-1 levels in small for gestational age (SGA) infants was significantly lower than in appropriate for gestational age (AGA) infants as previously reported (Randhawa, 2008,Tzschoppe et al., 2015). In contrast, PAPP-A2 levels in the SGA group were significantly higher than in the AGA group. This unexpected finding may be a compensatory response to increase the percentage of free IGF-1 in these children with low IGF-1 and growth impairment.

5. Conclusions

PAPP-A2 has recently been recognized as an important component to the GH–IGF axis. It appears to play a key role in regulating the percentage of free IGF-1 through its actions as a protease of IGFBP-3 and IGFBP-5. PAPP-A2 deficiency results in growth impairment in humans with expected derangements in biochemical parameters including increased levels of total IGF-1, IGFBP-3, ALS, and growth hormone. Mouse models now exist which recapitulate the human phenotype of PAPP-A2 deficiency. Early studies exploring the role of PAPP-A2 in the general population provide supportive evidence for it being an important player in growth physiology.

Highlights:

• PAPP-A2, an IGFBP-3 and -5 protease, is a key regulator of free IGF-1.

• PAPP-A2 deficiency causes growth failure with high total IGF-1 and IGFBP-3 levels.

• PAPP-A2 levels positively correlate with the percentage of free IGF-1 in childhood.

• rhIGF-1 treatment may improve PAPP-A2 deficient patients’ growth and metabolism.

Abbreviations

AGA

appropriate for gestational age

GH

growth hormone

GTT

glucose tolerance test

IGFBP

insulin-like growth factor-binding protein

ITT

insulin tolerance test

KI

knock-in

KO

knock-out

PAPP-A

Pregnancy-associated plasma protein-A

PAPP-A2

Pregnancy-associated plasma protein-A2

SGA

small-for-gestational age

STC

Stanniocalcin

WT

wild-type