Processes 2018, 6,42
Based on these results, we believe that the generation of novel miRNA regulatory networks is a
critical function of ADAR editing, and, notably, that dysregulated editing may create susceptibilities
that allow tumorigenesis and tumor progression to occur. Corroborating this idea, several studies have
already established a clear precedent for ADAR activity being implicated in cancer biology. Recently,
Chen et al. [15] described direct involvement of ADAR editing in human hepatocellular carcinoma
(HCC), showing how the transcripts of an oncoprotein degrader and confirmed contributor to HCC
pathology, antizyme inhibitor 1 (AZIN1), are modified at specific sites by ADAR1, and that ADAR1
is commonly upregulated in HCC patient tumors resulting in even higher AZIN1 editing frequency
and poorer prognosis. In addition, the authors were able to successfully demonstrate that higher
levels of edited AZIN1 promoted an increased incidence of tumor formation and invasive ability.
Over-editing of AZIN1 has also been implicated in other cancers, such as esophageal squamous cell
carcinoma [13]. Other recent studies suggest that ADAR1 might also play a pathogenic role in chronic
myeloid leukemia (CML). Jiang et al. [14] have recently shown that overexpression of ADAR1 in
cultured blood progenitor cells can promote reprogramming of myeloid progenitor cells resulting
in heightened hematopoietic differentiation toward the myeloid lineage. Increased ADAR1 levels
were repeatedly found in CML patient samples leading the authors to speculate that ADAR played a
causal role. In fact, a related study recently found CML could not be induced in mice following a bone
marrow transplant of marrow cells carrying an ADAR deletion suggesting ADAR1 may be essential
for leukemia cell survival [14].
In contrast to the previous examples linking hyper-editing to malignancy, the opposite scenario,
hypo-editing, has also been implicated as contributing to various cancers, specifically in relation to
miRNAs. For instance, it has been shown by Choudhury et al. [17] that reduced editing of miR-376a
promotes glioblastoma cell invasion in orthotopic glioma. Normally-edited miR-376a targets and
suppresses the receptor for the autocrine motility factor (AMF) that stimulates tumor motility via base
pair complementarity with the 3 ′ UTR of the AMF receptor mRNA; however, when unedited, the
miRNA loses this ability. It was also demonstrated that unedited miR-376a binds to the 3 ′ UTR of the
RAP2A mRNA transcript (coding for a protein known to suppress glioblastoma cell invasion), causing
the RAP2A protein’s function to be inhibited. This report does an excellent job of demonstrating how
ADAR-induced single base pair changes in miRNAs can alter their target specificity and ultimately
lead to pathologically significant ramifications. Further, while it is clear that RNA editing can be
fundamentally linked to cancer via sequence alteration and the expression/repression of oncogenes,
there is also evidence of involvement in other tumorigenic pathways. For instance, a correlation has
been shown between reduced editing of Alu elements and multiple tumors, including brain, prostate,
lung, and kidneys [14,18]. Additionally, chronic inflammation related to viral infection has been
previously implicated in tumorigenesis and this may be due, in part, to overexpression of ADAR1
mediated by inflammation [45]. Of note, in this work we identify 19,462 unique edit sites in MCF-7
cells versus 35,090 unique sites in MDA-MB-231s suggesting generally higher ADAR1 activity in this
more aggressive breast cancer cell line.
Importantly, the work presented here represents the most comprehensive of only a handful
of analyses of the effects of mRNA A-to-I editing on miRNA targeting published to date [30–32],
and represents only the second ever experimental evidence indicating that the modulation of
miRNA targeting through ADAR editing may directly contribute to breast cancer pathology [33].
When taken together, this report along with recently published studies suggesting mRNA editing
can alter microRNA regulations [30–33] (all published within the last few months) strongly
suggest that the participation of A-to-I editing in directing microRNA targeting is currently
significantly underappreciated.
That said, our analysis of the RNA editing data from two breast cancer cell lines demonstrate that
miR-140-3p is able to regulate the apoptosis inducing gene DFFA in MCF-7 but not in MDA-MB-231.
DFFA is the larger of two protein subunits that comprise caspase-activated DNase (CAD) and, when
bound to CAD, DFFA inhibits its ability to degrade DNA and condense chromatin, but during
139
Based on these results, we believe that the generation of novel miRNA regulatory networks is a
critical function of ADAR editing, and, notably, that dysregulated editing may create susceptibilities
that allow tumorigenesis and tumor progression to occur. Corroborating this idea, several studies have
already established a clear precedent for ADAR activity being implicated in cancer biology. Recently,
Chen et al. [15] described direct involvement of ADAR editing in human hepatocellular carcinoma
(HCC), showing how the transcripts of an oncoprotein degrader and confirmed contributor to HCC
pathology, antizyme inhibitor 1 (AZIN1), are modified at specific sites by ADAR1, and that ADAR1
is commonly upregulated in HCC patient tumors resulting in even higher AZIN1 editing frequency
and poorer prognosis. In addition, the authors were able to successfully demonstrate that higher
levels of edited AZIN1 promoted an increased incidence of tumor formation and invasive ability.
Over-editing of AZIN1 has also been implicated in other cancers, such as esophageal squamous cell
carcinoma [13]. Other recent studies suggest that ADAR1 might also play a pathogenic role in chronic
myeloid leukemia (CML). Jiang et al. [14] have recently shown that overexpression of ADAR1 in
cultured blood progenitor cells can promote reprogramming of myeloid progenitor cells resulting
in heightened hematopoietic differentiation toward the myeloid lineage. Increased ADAR1 levels
were repeatedly found in CML patient samples leading the authors to speculate that ADAR played a
causal role. In fact, a related study recently found CML could not be induced in mice following a bone
marrow transplant of marrow cells carrying an ADAR deletion suggesting ADAR1 may be essential
for leukemia cell survival [14].
In contrast to the previous examples linking hyper-editing to malignancy, the opposite scenario,
hypo-editing, has also been implicated as contributing to various cancers, specifically in relation to
miRNAs. For instance, it has been shown by Choudhury et al. [17] that reduced editing of miR-376a
promotes glioblastoma cell invasion in orthotopic glioma. Normally-edited miR-376a targets and
suppresses the receptor for the autocrine motility factor (AMF) that stimulates tumor motility via base
pair complementarity with the 3 ′ UTR of the AMF receptor mRNA; however, when unedited, the
miRNA loses this ability. It was also demonstrated that unedited miR-376a binds to the 3 ′ UTR of the
RAP2A mRNA transcript (coding for a protein known to suppress glioblastoma cell invasion), causing
the RAP2A protein’s function to be inhibited. This report does an excellent job of demonstrating how
ADAR-induced single base pair changes in miRNAs can alter their target specificity and ultimately
lead to pathologically significant ramifications. Further, while it is clear that RNA editing can be
fundamentally linked to cancer via sequence alteration and the expression/repression of oncogenes,
there is also evidence of involvement in other tumorigenic pathways. For instance, a correlation has
been shown between reduced editing of Alu elements and multiple tumors, including brain, prostate,
lung, and kidneys [14,18]. Additionally, chronic inflammation related to viral infection has been
previously implicated in tumorigenesis and this may be due, in part, to overexpression of ADAR1
mediated by inflammation [45]. Of note, in this work we identify 19,462 unique edit sites in MCF-7
cells versus 35,090 unique sites in MDA-MB-231s suggesting generally higher ADAR1 activity in this
more aggressive breast cancer cell line.
Importantly, the work presented here represents the most comprehensive of only a handful
of analyses of the effects of mRNA A-to-I editing on miRNA targeting published to date [30–32],
and represents only the second ever experimental evidence indicating that the modulation of
miRNA targeting through ADAR editing may directly contribute to breast cancer pathology [33].
When taken together, this report along with recently published studies suggesting mRNA editing
can alter microRNA regulations [30–33] (all published within the last few months) strongly
suggest that the participation of A-to-I editing in directing microRNA targeting is currently
significantly underappreciated.
That said, our analysis of the RNA editing data from two breast cancer cell lines demonstrate that
miR-140-3p is able to regulate the apoptosis inducing gene DFFA in MCF-7 but not in MDA-MB-231.
DFFA is the larger of two protein subunits that comprise caspase-activated DNase (CAD) and, when
bound to CAD, DFFA inhibits its ability to degrade DNA and condense chromatin, but during
139