Germline loss-of-function PAM variants are enriched in subjects with pituitary hypersecretion

Pituitary adenomas (PAs) are common, usually benign tumors of the anterior pituitary gland which, for the most part, have no known genetic cause. PAs are associated with major clinical effects due to hormonal dysregulation and tumoral impingement on vital brain structures. Following the identification of a loss-of-function variant (p.Arg703Gln) in the PAM gene in a family with pituitary gigantism, we investigated 299 individuals with sporadic PAs and 17 familial isolated pituitary adenomas kindreds for PAM variants. PAM encodes a multifunctional protein responsible for the essential C-terminal amidation of secreted peptides. Genetic screening was performed by germline and tumor sequencing and germline copy number variation (CNV) analysis. No germline CNVs or somatic single nucleotide variants (SNVs) were identified. We detected seven likely pathogenic heterozygous missense, truncating, and regulatory SNVs. These SNVs were found in sporadic subjects with GH excess (p.Gly552Arg and p.Phe759Ser), pediatric Cushing disease (c.-133T>C and p.His778fs), or with different types of PAs (c.-361G>A, p.Ser539Trp, and p.Asp563Gly). The SNVs were functionally tested in vitro for protein expression and trafficking by Western blotting, for splicing by minigene assays, and for amidation activity in cell lysates and serum samples. These analyses confirmed a deleterious effect on protein expression and/or function. By interrogating 200,000 exomes from the UK Biobank, we confirmed a significant association of the PAM gene and rare PAM SNVs to diagnoses linked to pituitary gland hyperfunction. Identification of PAM as a candidate gene associated with pituitary hypersecretion opens the possibility of developing novel therapeutics based on altering PAM function.


Introduction 115
The anterior pituitary gland plays a critical role in the dynamic control of major hormonal 116 systems, including growth, fertility, and stress responses. Anterior pituitary adenomas (PAs), 117 also called pituitary neuroendocrine tumors (PitNETs) [1], can be comprised of any of the 118 secretory cell subtypes, such as lactotropes that secrete prolactin, somatotropes (growth 119 hormone (GH)), corticotropes (adrenocorticotropic hormone (ACTH)), gonadotropes (follicle 120 stimulating hormone (FSH) and luteinizing hormone (LH)), and thyrotropes (thyroid 121 stimulating hormone (TSH)) [2]. Although they are usually benign lesions, PAs can have a 122 major impact through hormonal dysregulation and direct mass effects or invasion of brain 123 structures (optic chiasm, cavernous sinus). Epidemiologically, PAs are one of the most 124 frequent intracranial tumor types, and lead to clinically apparent disease with a frequency of 125 approximately 1 per 1,000 in the general population [3,4]. Approximately 95% of PAs occur 126 sporadically. Hereditary PAs are often distinguished by a more severe clinical presentation, 127 such as an earlier age at onset, aggressive growth, larger size, and greater resistance to 128 treatment, and they might coexist with other syndromic components. Therefore, identification 129 7 143 In a large, international cohort of individuals with PAs, we identified germline loss-of-function 144 (LOF) variants of PAM (peptidylglycine α -amidating monooxygenase, MIM: 170270) in 145 subjects with familial and sporadic GH and ACTH hypersecretion. PAM is a highly 146 conserved, multifunctional protein that is increasingly recognized as an important regulator 147 of peptide amidation and secretion, among many other functions, in health and disease [9]. 148 We functionally evaluated 36 SNVs, demonstrating a deleterious effect on PAM 149 function/expression in eight variants. We also report a statistically significant association of This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

164
for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. The third affected member was III-3, a male (son of II-2) who was born with normal length 183 and weight but developed marked overgrowth by the age of 12 months ( Figure 1C). His 184 siblings are dizygotic twins with normal height. III-3 was diagnosed with GH and IGF-1 185 excess (random GH 30 ng/mL and IGF-1 389 ng/mL) before five years of age. An MRI 186 demonstrated a normal pituitary gland ( Figure 1D). To control the excessive growth, the 187 patient was started on the long-acting somatostatin receptor ligand lanreotide autogel at a 188 11 500 ng/mL to 232 ng/mL and growth velocity decreased to 3.6-5.6 cm/year during 193 approximately five years of follow-up. His growth curves for height and weight remain above 194 the 99 th centile but are no longer diverging ( Figure 1E). 195 196 We conducted WES in peripheral blood-derived DNA isolated from the index cases (see the 197 Methods section for details). We applied a variant prioritization strategy ( Figure S1 and 198 Table S3) and selected for further analysis a heterozygous missense SNV in PAM 199 (c.2108G>A, p.Arg703Gln), which was shared by both twins from generation II ( Figure 1A). 200 DNA of their deceased mother was unavailable, so we could not confirm whether they 201 inherited the PAM variant from her or if it arose de novo; the father was WT. The 202 p.Arg703Gln variant was present in the other individual with gigantism, III-3, but also in III-1, 203 his younger brother ( Figure 1A). Endocrine and growth studies showed that III-1 had no 204 evidence of growth excess or hormonal dysregulation at the time of study (age < 10 years). 205 All other family members studied were WT for PAM and had normal growth. Overall, these 206 results suggest incomplete penetrance. No other potentially pathogenic alterations in genes 207 were shared among the three affected members. 208 209 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. gnomAD, and is bioinformatically predicted to be pathogenic (Table 2). 235 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; Arg703 is conserved throughout evolution ( Figure 2B). Along with Tyr651, crystallographic 255 studies identified Arg703 as part of the PALcc catalytic dyad; the catalytic activity of soluble 256 recombinant rat PALcc, in which Arg703 was replaced by Ala or Gln, was greatly reduced 257

14
[13], a result consistent with its location at the active site ( Figure 2C). Altogether, these data 258 suggested that a LOF PAM variant could be associated with pituitary gigantism, prompting 259 us to pursue this lead. 260 261 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; 262 263 536 539 542   This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; https://doi.org/10.1101/2023.01.20.23284646 doi: medRxiv preprint  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; Germline CNVs at the PAM locus were analyzed by ddPCR in a cohort of 137 individuals 294 with PAs (16 with gigantism and 121 with different types of PAs). Germline DNA sequencing 295 detected 51 SNVs (Table 2 and Table S4). There were no germline CNVs, or 296 rare/pathogenic somatic DNA SNVs (data not shown). A subset of the germline SNVs (15 of 297 51) was prioritized for further screening together with the p.Arg703Gln variant (Table 2). This 298 selection was based on the following non-mutually exclusive criteria: MAF < 1%, in silico 299 analysis with multiple algorithms predicting a deleterious effect on PAM function or splicing, 300 and published functional data indicating pathogenicity [15]. To complement these data, we 301 used Clustal and structure-based analyses to identify variants that could impact protein 302 structure/function ( Figure 2B and 2C, Figure S3). The remaining, non-characterized germline 303 SNVs are listed in Table S4. PHMcc and six of the seven in PALcc were predicted to be pathogenic ( Table 2). The 309 prioritized variants were not significantly spatially clustered when we considered the 310 geometric mean distance between all pairs of variants normalized to cDNA length (p = 0.35). 311 However, when we classified the variants based on their location within the N-terminus, 312 PHMcc, linker, PALcc or C-terminus (Figure 2A), the N-terminus harbored more variants 313 than expected by random chance (p = 0.02, Table S5). This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; particulate fraction ( Figure 3A). PAM expression levels and protein integrity were evaluated 322 by Western blot analysis ( Figure 3B) and PAM trafficking was assessed by analysis of N-323 and O-linked oligosaccharide maturation ( Figure 3C and 3D). 324 To establish the efficacy of our expression system, we compared the properties of WT PAM 325 to those of missense PAM variants designed to inactivate PHMcc (p.Thr189Ile) or PALcc 326 (p.His529Arg and p.Gly796Glu, Figure S4). PHM and PAL activity measurements for WT 327 PAM and for each engineered control were normalized to the expression of that PAM protein 328 [(PHM or PAL activity)/PAM protein], providing a measure of its specific activity; the PHM 329 and PAL specific activities of each engineered control were then compared to the PHM and 330 PAL specific activities of WT PAM. As expected, the PHM activity of the p.Thr189Ile mutant 331 and the PAL activity of the p.His529Arg and p.Gly796Glu mutants were reduced more than 332 ten-fold compared to WT PAM ( Figure S4 and Table 3). 333 334 335 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; The structure of bifunctional PAM has not yet been experimentally determined, but for all 344 three engineered variants, inactivating mutations placed into one domain resulted in a 345 modest decrease in the activity of the other domain. 346 Activity and Western blot data for all studied SNVs are shown in Figure 3A and 3B, 347 respectively. No significant decrease in PHM activity was seen for SNVs p. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; engineered control lacking PHM activity (Thr189Ile) and that each of these six variants had 372 been N-glycosylated. As expected, neuraminidase treatment brought about a slight 373 decrease in the apparent molecular mass of WT PAM. Strikingly, neuraminidase treatment 374 failed to reduce the mass of any of these six SNVs, indicating that their glycans had not 375 undergone normal sialylation. Despite this, p.His240Tyr, one of the variants exhibiting 376 altered sialylation, had normal levels of both PHM and PAL activity. Altered trafficking in the 377 secretory and/or endocytic pathway could limit sialylation. 378 379 A summary of the effect of each prioritized and engineered SNV on PAM expression, 380 enzymatic activity and oligosaccharide maturation is presented in Table 3. 381 382 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ;  In N-Gly Con In N-Gly Con In N-Gly Con In N-Gly Con In N-Gly Con In N-Gly Con This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023.  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; are significantly more prevalent in our cohort vs. controls, and are without homozygous 416 variant entries in gnomAD. The p.Arg80Gln SNV is located within PHMcc, while the 417 p.Pro412Leu and p.Glu491Asp SNVs are within the linker region. The p.Arg80Gln did not 418 segregate with the phenotype in the FIPA kindred and functional studies showed no 419 significant effect of p.Arg80Gln on PHM enzymatic activity. In contrast, the p.Pro412Leu 420 variant segregated with the FIPA phenotype and was seen in both the proband (acromegaly) 421 and the daughter (prolactinoma). Neither of the p.Pro412Leu or p.Glu491Asp variants in the 422 non-enzymatic linker region had a functional effect on PHM or PAL activity ( Figure 3A, Table  423 3  Figure S6). Skipping of exon 21 was confirmed by testing both the affected subject and 436 parent blood-extracted RNA ( Figure 4A and 4B) and by using a minigene splicing assay 437 ( Figure 4C). Transient expression of p.His778fs in PEAKrapid cells demonstrated, as 438 expected, that the truncated protein lacked detectable PAL activity; in addition, significantly 439 reduced levels of PHM activity were also observed ( Figure 3A, Table 3). Western blotting 440 indicated that the mutant protein is produced but at lower levels than those observed for WT. 441 Moreover, the observed band was smaller than predicted (75 vs. 90 kDa, respectively), 442 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; suggesting that the mutant might lack some post-translational modifications and/or is 443 cleaved at a pair of basic residues introduced by the frameshift (Figure 4D This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; predicted for this protein -which includes only the first 777 residues of WT PAM-1 but extends 45 466 residues beyond residue 777 before reaching a stop codon -is 90.31 kDa. NT, not transfected. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; As a result, we could only assess PHM and PAL activity in the small number of serum 509 samples collected from our cohort. As shown in Figure 6, we measured PHM and PAL 510 activity and calculated the ratio of PAL to PHM activity for 13 subjects (11 affected 511 individuals/carriers and two WT controls). To control for the fact that the serum samples 512 assayed had been stored for varying periods of time, we evaluated the statistical 513 significance of changes in activity by normalizing PAL activity to PHM activity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; https://doi.org/10.1101/2023.01.20.23284646 doi: medRxiv preprint variants. Also consistent with the data shown in Table 3 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. . For this analysis, we considered 20 genes that harbor germline or somatic 600 variants known to predispose to PAs in order to estimate a GDI cutoff above which a gene is 601 unlikely to cause a pituitary disease, i.e., it is considered a false positive. PAM was reported 602 to have a Phred-scaled GDI score of 3.98, below the calculated cutoff of 4.84 (Table S7) This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. The finding of two non-coding SNVs affecting PAM splicing (albeit with different outcomes) 621 and recent studies underscoring the importance of evaluating splicing defects as a disease 622 mechanism for both conditions of hyper and hypopituitarism [31, 32], prompted us to extend 623 the corresponding functional studies to selected synonymous and intronic variants found in 624 our cohort ( Figure S12A and Table S8). The selection criteria were analogous to those 625 applied for other prioritized variants. This analysis did not reveal any splicing impairment. 626 We observed only a possible reduction in mRNA expression by the c.195G>T allele (Figure  627 S12B), but this was not subsequently confirmed by qPCR studies ( Figure S12C). 628 Taken together, these findings indicate that impairment of PAM function or expression via 629 abnormal splicing is not frequently associated with PAs. 630 631 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; Our in vitro functional tests indicated an impairment in enzymatic activity for five missense 686 SNVs and the splicing variant (in PAL). PAM expression was decreased for two 5'UTR 687 variants and the COOH-terminal region splicing variant. In addition, glycosylation was 688 abnormal for four of the five missense SNVs. Experimental truncations used to define the C-689 terminus of rat PALcc are consistent with the lack of PAL activity observed for p.His778fs, 690 with the same inability of the truncated rat protein to fold correctly, thereby preventing its This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; explored. Taken together, these data underline the multiple roles that PAM can play in 806 secretion of hormones. Studies of emerging PAM actions will be needed to address the 807 influence of genetic variants in this study on PAM-mediated pituitary hormone release. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. to the limited amount of DNA available, Sanger sequencing was targeted to specific regions 890 of PAM where the majority of likely pathogenic variants were previously observed (Figure 2), 891

Materials and Methods
i.e., the proximal promoter and exons 9, 16, 18, 19, 20, and 21. The primers included in 892 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; Single cell RNA sequencing (scRNA-seq) analysis: Figure S2A  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; • weight of 100 for transcript ablation. 1037 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; https://doi.org/10.1101/2023.01.20.23284646 doi: medRxiv preprint Statistical significance for gene-based tests was set at a Bonferroni-corrected threshold of p for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. ; for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted January 20, 2023. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.