Association of the M1V PRKAR1A Mutation with Primary Pigmented Nodular Adrenocortical Disease in Two Large Families

Abstract

Background: Carney complex (CNC) is a familial multiple neoplasia syndrome frequently associated with primary pigmented nodular adrenocortical disease (PPNAD), a bilateral form of micronodular adrenal hyperplasia that leads to Cushing’s syndrome (CS). Germline PRKAR1A mutations cause CNC and only rarely isolated PPNAD.

Patients and Methods: PRKAR1A mutation analysis in two large families with CS and no other CNC manifestations demonstrated a M1V germline mutation; a total of 21 asymptomatic individuals were screened, and mutation carriers were evaluated for CNC. The mutation was expressed in vitro and functionally tested for its effects on protein kinase A function.

Results: Presymptomatic testing identified five first-degree relatives who were M1V carriers and who were all diagnosed with subclinical, mild CS at ages ranging from 20–56 yr. There were no other signs of CNC. In a cell-free system, we detected a shorter compared with the wild-type type 1α regulatory subunit of protein kinase A (PRKAR1A) protein (43 kDa). This was not identified in cell lines from the patients or in transfection experiments in HEK293 cells that showed no detectable PRKAR1A protein from the M1V-bearing constructs. In these cells, the mutant mRNA was expressed in a 1:1 ratio.

Conclusion: In two large families, the M1V PRKAR1A mutation resulted in a PPNAD-only phenotype with significant variability both in terms of age of onset and clinical severity. Expression studies showed a unique effect of this sequence change. This study has implications for genetic counseling of carriers of this PRKAR1A mutation and patients with CNC and PPNAD and for the study of PRKAR1A-related tumorigenesis.

Mutation in the initiation codon of PRKAR1A leads to a mild Carney complex phenotype.

Carney complex (CNC; MIM no. 160980) is an autosomal dominant multiple neoplasia syndrome, characterized by cardiac and extracardiac myxomas and other tumors along with spotty skin pigmentation and endocrine overactivity (1,2). The most frequent endocrine manifestation of CNC is ACTH-independent Cushing’s syndrome (CS) caused by primary pigmented nodular adrenocortical disease (PPNAD), a form of micronodular bilateral adrenocortical hyperplasia (3). Mutations of the gene coding for the protein kinase A (PKA) regulatory subunit PRKAR1A [the type 1α regulatory subunit (RIα) protein], cause the disease in most CNC patients; few apparently sporadic PPNAD patients also carry PRKAR1A mutations (4,5).

Scant information is available for PRKAR1A genotype and PPNAD/CNC phenotype correlation. We recently reported the first PRKAR1A mutation c.709(−7−2)del6 that was almost exclusively associated with mild CS due to PPNAD; c.709(−7−2)del6 was particularly prevalent among the patients with this phenotype, being present in approximately 20% of them (6). We now report the second PRKAR1A mutation (c.1A→G/p.M1V) that appears to be associated almost exclusively with PPNAD. This observation has implications for genetic counseling and for the investigations related to RIα’s role in adrenal function and tumor formation.

Subjects and Methods

Informed consent was obtained as part of a protocol approved by the institutional review boards of the participating institutions. The index patients described in this study were referred for the investigation of CS; their relatives were recruited after the identification of the mutation (see supplemental Subjects and Methods, published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org).

DNA was extracted from peripheral blood leukocytes using the Wizard genomic DNA purification kit (Promega, Madison, WI), and the 12 exons and the flanking intronic sequences of the PRKAR1A gene were amplified using the primers and the conditions described previously (6,7).

Lymphocyte cell lines from patients or control subjects were established and treated as previously described (6,7) (see supplemental Subjects and Methods).

To study the functional effect of M1V on PRKAR1A expression, a PCR-based cloning method was used to generate both the wild- type and mutant expression constructs, according to previously published methods (8). The effect of mutated PRKAR1A on PKA activity was determined as previously described (8).

To assess the expression of the M1V-bearing PRKAR1A gene in the absence of cellular factors, we used a transcription and translation cell-free system (TNT T7 coupled reticulocyte lysate system; Promega) with l-[³⁵S]methionine (EasyTag; PerkinElmer Life and Analytical Sciences, Norwalk, CT), according to the manufacturer’s instructions.

To study the lack of the M1V protein in the cell-based system, a proteasome proteolytic activity inhibitor, clasto-lactacystin β-lactone, was applied to transfected cells as described (9). This treatment functions to inhibit the degradation of the proteins by the proteasome.

Results

We studied two families with the M1V mutation with PPNAD only (Fig. 1A). The proband of the first (CAR 19.01) was first reported in 1957 (10) and is now deceased from, reportedly, complications of pneumonia. Neither his father nor his mother was affected by history. Three of the proband’s daughters (CAR 19.03, 19.05, and 19.06), were M1V carriers, and all presented with CS due to PPNAD. To date, they have been followed each for more than 15 yr after bilateral adrenalectomies and have not had any other CNC manifestations.

Figure 1.

Pedigrees of families CAR 19 (A) and CAR 51 (B). The genotype of the individuals available for analysis is shown above each symbol; the stars indicate heterozygoous presence of M1V substitution; the arrows indicate the probands.

The index patient (CAR 51.01) of the second family was diagnosed with CS at age 12 yr. Clinical signs were present since the age of 7, and his growth declined after the age of 10. After the identification of the M1V mutation in PRKAR1A, family screening identified three more carriers of the mutation: the patient’s father and two siblings (Fig. 1B). Her father (CAR 51.04) had been diagnosed with severe CS at the age of 36, after being admitted for a traumatic femur fracture and multiple abnormalities of the spinal vertebrae. Family pictures revealed a plethoric face since 1945, his asthma had disappeared since 1944, and hypertension was diagnosed in 1948. He has no facial or other pigmented lesions. His second daughter (CAR 51.10) to this date remains completely asymptomatic but with biochemically confirmed hypercortisolemia. Finally, a son (brother of the proband, CAR 51.13), had mild, subclinical ACTH-independent CS, but other CNC manifestations, again, were absent. We also studied 10 relatives of the father of the proband and did not find the mutation there, suggesting that it most likely occurred de novo in the father.

Western blot of extracts from patient cell lines carrying the M1V mutation showed only the normal PRKAR1A protein, as we have demonstrated elsewhere (4) (data not shown). To study the expression of the M1V PRKAR1A isoform, we used first a cell-free transcription and translation TNT T7 coupled reticulocyte lysate system in experiments controlled with the wild-type protein (Fig. 2A). We detected expression of both wild-type and mutant constructs; the wild type corresponded to the expected size (49 kDa), and the mutant was shorter (approximately 43 kDa).

Figure 2.

A, Autoradiography after cell-free transcription and translation in TNT T7 coupled reticulocyte lysate system. A shorter (43 kDa) compared with the wild-type PRKAR1A (49 kDa) was detected after transfection with vectors bearing the M1V substitution. B, Western blot analysis of the expression of PRKAR1A in HEK293 cells after transfection with the wild-type and the M1V PRKAR1A constructs; parallel experiments were done with (+) and without (−) clasto-lactacystin β-lactone, a proteasome proteolytic activity inhibitor.

To assess the expression of the M1V-harboring PRKAR1A isoform in a cell-based system, we performed parallel transfection of HEK293 cells with the wild-type and the M1V-bearing construct. Similarly to other nonsynonymous substitutions in PRKAR1A, the M1V cDNA was expressed, as confirmed by sequencing (8). Western blot analysis, however, demonstrated presence of only the wild-type size PRKAR1A (Fig. 2B), after transfection with both the wild-type and the mutant constructs. The protein expression levels from the M1V-transfected cells were lower compared with the wild type and similar to the PRKAR1A levels after transfection with an empty vector (mock). This observation suggested that the detected protein most probably corresponds to the endogenous PRKAR1A in the HEK293 cells.

We then analyzed the effects of the transfection with M1V bearing constructs on the cellular PKA activity. The results of our transfection experiments are presented in the Supplementary Data and Supplementary Figure 1. Briefly, higher PKA activity and reduced binding to cAMP was measured after the transfection with M1V bearing construct–a result consistent with an overall decrease in PRKAR1A protein levels.

Discussion

This study of two large CNC families demonstrated that a mutation in the initiation codon of PRKAR1A, M1V, resulted in a specific phenotype characterized by PPNAD only and often mild atypical CS. The identified mutation carriers, who underwent extensive and detailed clinical screening, all developed CS but no other classical CNC manifestations, like myxomas, schwannomas, or spotty skin pigmentation. The M1V PRKAR1A mutation has never been seen in normal controls; in addition, more than 2000 chromosomes have been tested in our laboratory over the last 10 yr.

Although the mutant mRNA is found to be expressed equal to the wild-type levels in carriers of the mutation, the mutant protein does not seem to be expressed in cells. Notably, in a cell-free system, we detected a shorter PRKAR1A of approximately 43 kDa. This protein corresponds to an alternative translation product starting from a surrogate initiation site-in-frame ATG in the context of the Kozak sequence located 141 bp downstream from the original initiation codon. However, when we transfected the same vectors into HEK293 cells, we were not able to detect shorter PRKAR1A, with or without proteasome proteolytic activity inhibition, suggesting that there is a mechanism different from proteolytic degradation that prevents the expression of M1V in the cell. This observation raises the question of the specific phenotype observed among the carriers of M1V as opposed to the majority of CNC patients, in whom no genotype-phenotype correlation has been established.

The majority of the PRKAR1A mutations lead to nonsense RNA; the mutant RNA species are then degraded by nonsense-mediated decay (NMD) (7). Some aspects of the CNC phenotype were reproduced in the Prkar1a^+/− and transgenic, prkar1a-antisense-expressing mice (11,12). The latter developed a late-onset, mild form of adrenal hyperplasia and adenomas associated with lack of suppression of corticosterone secretion in response to dexamethasone (12,13). The variability of CS and corticosterone secretion among patients with PPNAD and CNC and transgenic mice, respectively, has been attributed to possibly incomplete NMD or mutations expressed at the protein level (14), loss of heterozygosity (LOH) in PPNAD nodules in contrast to preservation of hemizygosity for the normal allele in surrounding adrenal cortex (13,15), and possibly other factors such as the presence of modifying genetic defects and a female gender (16).

Recently, a small number of mutations were described that were not subject to NMD; they all led to expressed RIα variants that were associated with increased PKA activity in vitro, just like PRKAR1A mutations that undergo NMD (8,17,21). Interestingly, when LOH could be tested in adrenal tissue from patients with expressed RIα variants, there were no allelic losses (14). Thus, deficient control of the PKA catalytic subunits by either NMD of the mutant allele and LOH of the normal allele (complete RIα loss) or expressed RIα variants that have lost the ability to regulate PKA activity (and thus without need for LOH of the normal allele) is what is responsible for increased PKA activity in PPNAD and other CNC-affected tissues (8,17). These laboratory data indicated that alteration of RIα function alone (and not only its complete loss) is sufficient for increasing PKA activity leading to tumorigenesis and CNC; these data also predicted that quantitative differences of PKA activity in adrenal tissue would be associated with variable forms and/or severity of PPNAD and CS in affected patients.

So far, there has been no significant genotype-phenotype correlation, except in the case of the c.709(−7−2) del6 PRKAR1A mutation, which, similar to M1V, was associated with PPNAD only (6). However, this splice site deletion did not lead to an expressed RIα variant and was subject to NMD. In addition, in contrast to M1V, it appeared to be a low-penetrance variant with several of its carriers manifesting no signs of CS or CNC even after complete hormonal and radiological investigations.

Among the M1V mutation carriers, variability was present in age of onset and severity of CS. Whereas the index patients developed symptoms related to CS before the age of 10 yr, the father of the second proband developed severe CS at the age of 36 yr. Her two siblings were presymptomatically tested (at ages 56 and 45), and both appeared to have only a form of CS. In addition to this intra-familial variability, this family clearly demonstrated differences when compared with previously described CNC patients. One of the hallmarks of CNC, spotty skin pigmentation, which is typically found in more than three fourths of the patients, was absent from both M1V-carrying families. Because PPNAD is the only manifestation of CNC in these two families, we may conclude that the adrenal cortex may be particularly sensitive to minor changes of PKA activity and/or PRKAR1A levels.

Mutation p.M1V results from a change of the first codon in exon 1 from ATG to GTG. This change might cause reduced or absent translation of the wild-type protein or expression of an altered PRKAR1A; however, the actual consequences in vivo are not completely clear. In Smith-Lemli-Opitz syndrome, an autosomal recessive multiple congenital malformation and mental retardation syndrome, M1V has also been suggested to be a mild mutation (18). The reason may be that GTG in this position also may function as a start codon (19). Indeed, although never detected in vivo, the second ATG codon in PRKAR1A, located 141 bp downstream, might be used as an alternative translation initiator codon acting synergistically or independently of the normal start site. Such a shorter protein would lack the dimerization/docking domain, which is critical for the binding to kinase-anchoring proteins and, respectively, for the proper cellular localization of PKA (22).

In conclusion, the M1V mutation resulted in a specific phenotype in two large CNC families in which the affected patients were predominantly characterized by PPNAD and mostly mild atypical CS. There is significant, however, intra-familial variability. This observation has implications for genetic counseling and for the investigations related to the role of PRKAR1A in adrenal function and tumor formation.