Submission Details

AGL was first reported in relation to autosomal recessive glycogen storage disease III (GSD III; also known as Cori or Forbes disease) in 1992 (Yang et al., PMID: 1374391). AGL encodes the glycogen debrancher enzyme (GDE), which uses both glucosidase and transferase activities to break down glycogen into glucose. GSD III is diagnosed by finding a lack of GDE activity in patient cells or through identifying biallelic pathogenic variants in AGL.

Clinical manifestations of GSD III include excess glycogen in the liver, hepatomegaly, hepatic fibrosis, and hypoglycemia. Skeletal muscle and cardiac muscle involvement is also common in patients diagnosed with subtype IIIa, while patients with IIIb tend to only demonstrate liver involvement (GeneReviews). Furthermore, the c.16C>T (p.Q6X) and c.18_19delAG (p.Gln6HisfsX20) variants in exon 3 are exclusively associated with the GSD IIIb subtype (PMID:20648714). Although the genotypic and phenotypic differences behind the IIIa and IIIb subtypes is not fully understood, one explanation is that AGL isoforms are differentially expressed in liver and muscle tissues, and exon 3 could be bypassed in the muscle-specific isoforms. There are also cases of subtypes IIIc and IIId reported in the literature, which are suspected to be caused by the isolated loss of either glucosidase (IIIc) or transferase (IIId) activity in the GDE (PMID:20648714). Per criteria outlined by the ClinGen Lumping and Splitting Working Group, we found no difference in inheritance pattern and phenotypic variability (apart from the variable phenotypes caused by isoforms). Therefore, the following disease entities have been lumped into one disease entity (GSD III): GSD IIIa, GSD IIIb, GSD IIIc, GSD IIId.

Over 100 individual AGL variants have been identified in GSD III patients (reviewed in Goldstein et al. 2010, PMID: 20648714). 15 variants (1 missense, 6 nonsense, 5 frameshift, 3 splicing) that have been reported in 11 probands in 6 publications (PMIDs: 26913919, 23430490, 19834502, 11757581, 25602008, 8755644) are included in this curation. More evidence is available in the literature, but the maximum score for genetic evidence (12 pts.) has been reached. The mechanism of pathogenicity is known to be LOF.

This gene-disease relationship is also supported by biochemical assays, expression studies, and animal models (PMIDs: 24613482, 1856189, 2961257). First, biochemical experiments have demonstrated that AGL breaks down glycogen into glucose, which is consistent with the hallmark phenotype of excess glycogen observed in patient cells (PMID:2961257). Secondly, an experiment by Hermans et al. established that a similar protein, GAA, also acts as a glucosidase (PMID:1856189). GAA was classified as definitively associated with GSD II by the ClinGen General Gene Curation Expert Panel in 2019. As a result, a similar gene being implicated in a similar disease is evidence that supports the relationship between AGL and GSD III. Next, an immunoblot analysis showed the presence of GDE in the muscle and liver tissues of controls, but was absent in patients diagnosed with GSD III. Finally, an Agl knockout mouse model demonstrated similar phenotypes to human patients including hepatomegaly, excess glycogen levels, liver fibrosis, muscle weakness, and elevated CK levels when fasted (PMID: 24613482).

In summary, AGL is definitively associated with autosomal recessive GSD III. This has been repeatedly demonstrated in both the research and clinical diagnostic settings, and has been upheld over time. This classification was approved by the ClinGen General IEM GCEP on the meeting date 2/24/23 (SOP Version 9).