Unfortunately, the identification of genetic correlates of high-risk/poor prognosis neuroblastoma (e.g. BDNF/TrkB expression, MYCN amplification) has not consistently translated into equal gains in treatment efficacy, with long-term overall survival for children with advanced disease remaining less than 40% in spite of increasingly intensive chemotherapy regimens.26, 27
Therefore, continued examination for additional developmental and molecular alterations contributing to the etiology of neuroblastoma initiation or progression is of paramount clinical importance.
Previous reports have postulated a putative tumor suppressor role for ACY1.9, 10
Despite this proposed role, ACY1 expression was paradoxically greater in MYCN-amplified SMS-KCNR cells than non-MYCN-amplified SK-N-SH cells. Yet, ACY1 was the only aminoacylase whose expression was markedly up-regulated upon differentiation in media containing 10% serum, supporting a role for ACY1 in differentiation. Conversely, ACY1 expression was decreased in SMS-KCNR cells upon differentiation. Differential subcellular localization is one possible explanation for these incongruent observations. In SK-N-SH cells, ACY1 expression was primarily centriolar, but in SMS-KCNR cells, expression was more diffusely cytoplasmic. ACY1 regulates the subcellular localization of sphingosine kinase type 1 and, thereby, its product sphingosine-1-phosphate, promotes cell growth and inhibits apoptosis of tumor cells.28
Therefore, the distinct subcellular localization of ACY1 might similarly be involved in targeting specific protein substrates in neuroblastoma cells. Curiously, induction of neural differentiation in SK-N-SH cells altered expression of the 65 kDa, but not the predicted 50 kDa, ACY1 isoform. All three aminoacylases exist in multimeric forms that are enzymatically active deacetylases29–31
and are quite atypical “cytoplasmic” enzymes in that they may undergo glycosylation and phosphorylation.12
Glycosylation represents the most plausible explanation for the 15 kDa increase in molecular weight, but future studies will need to be undertaken to confirm this notion. The disparate result of increased ACY1 expression in MYCN-amplified cells, yet increased expression upon differentiation is mimicked in vivo
. Kaplan-Meier survival data indicate that high ACY1 expression is correlated with better prognosis, as one would predict for a tumor suppressor. However, high expression is also associated with increased mortality initially (Oncogenomics database: http://pob.abcc.ncifcrf.gov/cgi-bin/JK?rm=get_kms;db=nbProg;exprs=log2ctr;threshold=2;limit=100;geneid=4117
); thus, the relevance of ACY1 in neuroblastoma is, at present, unclear.
Although ASPA expression was more highly expressed in the non-MYCN-amplified SK-N-SH line than the MYCN-amplified SMS-KCNR line, this is likely due more to cell type than MYCN status. ASPA is primarily expressed in CNS glial cells and may contribute to a less malignant neuroblastoma phenotype by supporting glial cell differentiation. The enrichment of ASPA in SK-N-SH and SH-EP S-type cells is consistent with Kaplan-Meier survival data that low ASPA expression correlates with an unfavorable prognosis (Oncogenomics database: http://pob.abcc.ncifcrf.gov/cgi-bin/JK?rm=get_kms;db=nbProg;exprs=log2ctr;threshold=2;limit=100;geneid=9710
) in that patients with tumors possessing low levels of Schwannian stroma have a less favorable prognosis.25
Although ASPA is best characterized for its role in myelin lipid synthesis, by use of the acetate liberated upon NAA deacetylation11
, recent evidence suggests a link between ASPA and cell cycle regulation.32
ASPA knockout mice exhibited increased G1→S cell cycle progression and increased histone acetylation, consistent with previous findings of nuclear ASPA localization.22, 30
ASPA may also contribute to neuroblastoma tumorigenesis via its suppressive effect on BDNF expression. ASPA-null rats exhibit abnormally elevated BDNF levels that are diminished upon transfection with constitutive ASPA, placing ASPA regulatory hierarchically above BDNF.33
Based on its punctate membrane-associated expression, it has been suggested that ASPA may subserve a signaling function.33
Given that BDNF promotes tumor survival, angiogenesis, and drug resistance in TrkB-positive neuroblastoma tumors4, 5
and the mechanisms of BDNF/TrkB regulation are poorly understood, a possible link between ASPA and BDNF warrants further investigation.
ACY3 is the least understood aminoacylase primarily due to its limited substrate presence in eukaryotic cells (e.g., N-acetylated aromatic amino acids and mercapturic acids).13, 14
Nonetheless, data are consistent with a role for ACY3 in neuroblastoma promotion: 1) the 80 kDa ACY3 isoform was more robustly expressed in MYCN-amplified cells, 2) ACY3 was expressed in the juvenile and adult adrenal medulla and abundantly expressed in neuroblastoma surgical specimens, and 3) Kaplan-Meier survival data indicates that high ACY3 expression correlates with poor prognosis (Oncogenomics database: http://pob.abcc.ncifcrf.gov/cgi-bin/JK?rm=get_kms;db=nbProgSg;exprs=log2ctr;threshold=2;limit=100;geneid=30370
). However, the one inconsistent finding in this study was ACY3 up-regulation upon cAMP-induced neural differentiation of SK-N-SH cells. Yet, unlike ACY1, this increase only occurred in media containing 1% serum. Another perplexing finding is ACY3’s nuclear localization. While nuclear ASPA expression has been reported30
, to our knowledge ACY3 has not been previously detected within the nucleus. The ACY3 peptide used as an immunogen shows no homology to ASPA, ACY3 sera does not cross react with recombinant ASPA, and immunolabeling is blocked by peptide pre-incubation; hence, we believe this immunolabeling to be authentic. However, based on our limited understanding of ACY3’s normal physiological role, it is difficult to infer the functional importance of ACY3 expression in neuroblastoma.
The primary role of aminoacylases is to deacetylate amino acids for subsequent use in protein synthesis, or, in the case of ASPA, to provide acetate for myelin synthesis. Although the conservation of the essential zinc-binding functional groups and zinc-dependence for enzymatic activity demonstrate that aminoacylases are zinc–dependent metalloproteinases34
, thus far no proteolytic substrates have been identified. Therefore, future studies will need to be undertaken to determine whether the aminoacylases function via their traditional, amino acid deacetylation activity or some, as yet unidentified, novel mechanism and whether their role is restricted to neuroblastoma or is of relevance to other tumors.