HIF-1-Induced hsa-miR-429: Understanding Its Direct Targets as the Key to Developing Cancer Diagnostics and Therapies

Simple Summary In this review, we discuss how miRNAs play a critical role in the regulation of mRNA stability and translation, how determining the direct targets of miRNAs in complex networks is extremely difficult, and how translating specific miRNAs to the clinic often results in failure. This has led to concerns that this approach may not be feasible. Using hsa-miR-429 as an example, we discuss the limitations encountered in the development of efficient miRNAs-related therapies and diagnostic approaches and how this can be improved. We also provide a literature analysis of the verified hsa-miR-429 targets in various human research models. A meta-analysis of this work should provide better insights into the role of hsa-miR-429 in cancer diagnosis and any potential therapeutic approaches. Abstract MicroRNAs (miRNAs) play a critical role in the regulation of mRNA stability and translation. In spite of our present knowledge on the mechanisms of mRNA regulation by miRNAs, the utilization and translation of these ncRNAs into clinical applications have been problematic. Using hsa-miR-429 as an example, we discuss the limitations encountered in the development of efficient miRNA-related therapies and diagnostic approaches. The miR-200 family members, which include hsa-miR-429, have been shown to be dysregulated in different types of cancer. Although these miR-200 family members have been shown to function in suppressing epithelial-to-mesenchymal transition, tumor metastasis, and chemoresistance, the experimental results have often been contradictory. These complications involve not only the complex networks involving these noncoding RNAs, but also the problem of identifying false positives. To overcome these limitations, a more comprehensive research strategy is needed to increase our understanding of the mechanisms underlying their biological role in mRNA regulation. Here, we provide a literature analysis of the verified hsa-miR-429 targets in various human research models. A meta-analysis of this work is presented to provide better insights into the role of hsa-miR-429 in cancer diagnosis and any potential therapeutic approach.

This specificity problem could result from a number of reasons including the following: (1) mammalian miRNAs are not required to be perfectly complementary to their target mRNA sequences; (2) each miRNA can affect up to hundreds of targets; (3) miRNAs can promote indirect effects through the modulation of transcription factors (TFs) and the inhibition of their downstream targets, thus generating false positives [32][33][34]. To make matters worse, mRNA stability and translational efficiency can be modulated by different miRNAs simultaneously. Furthermore, although some miRNAs have been shown to be potent post-transcriptional repressors [35][36][37][38][39][40][41][42][43][44][45][46][47][48], the vast majority of these molecules only provide a modest modulation of the subsequent protein levels and thus serve more as 'molecular buffers' than as strong repressors of gene expression [49,50]. Therefore, understanding the complex molecular networks modulated by miRNAs remains a challenging, but mandatory, step towards any successful utilization of these molecules in any therapeutic or diagnostic approach.
In this review, we discuss the molecular networks of miR-429, a member of the miR-200 family that is induced by HIF-1 during hypoxia [51]. miR-429 is commonly associated with the progression and metastatic potential of many human cancers [52][53][54][55][56][57]. The members of the miR-200 family are clustered in two genomic locations: one on chromosome 1 (1p36.33) that comprises miR-200b, miR-200a and miR-429, and the other on chromosome 12 (12p13.31) that includes miR-200c and miR-141 ( Figure 1A) [58]. Despite the fact that these locations encode harpin precursors for both 5′ and 3′ miR-200 family members, the 3' mature miRNA forms are the dominant ones [58], and therefore the -3p postfix is omitted in the majority of reports including in this review.   Although the miR-200 family members have been widely accepted as tumor suppressors, these studies have been controversial with regard to cancer prognoses [52,56,57]. Here, we discuss hsa-miR-429 and the challenges in understanding this miRNA role in cancer development. The first problem is the similarity of seed sequences in all of the miR-200 family members. This makes determining the specificity of each of the different members of this family a challenge. Certainly, not properly addressing this in an experimental model could eventually lead to false conclusions. The seed sequence AAUACUG is common for miRs-200b, miR-200c and miR-429, and there is only a one-nucleotide sequence difference, AACACUG, in miR-200a and miR-141 [23,58]. This identity and similarity make the mRNA target profiles of this family of miRNAs very similar ( Figure 1).
As shown in Figure 1B, the vast majority of the predicted miR-429 high-probability targets (top 5%) are common for all of the miR-429 family members. The only difference is the "unique targets" of miR-429 can also be recognized with a probability higher than 95% [59,60]. A similar picture was obtained for miR-429 targets that were verified experimentally ( Figure 1C) [61]. Furthermore, the expression of specific miR-200 family members such as hsa-miR-429 is modulated by the cellular context, cellular stress conditions, and the stage of tumor progression and metastasis [52][53][54][55][56][57]. How this relates to the other family members is still unclear.
Although the miR-200 family members including hsa-miR-429 were shown in numerous reports to be dysregulated in different types of cancer and assigned functions in suppressing epithelial-mesenchymal transition (EMT), tumor metastasis and chemoresistance [52][53][54][55][56][57], the data are often contradictory and difficult to understand. Therefore, a systematic analysis of the molecular network of their targets and context is needed. Here, we provide a comprehensive literature and meta-analysis of the reports verifying hsa-miR-429 targets in various human research models.

Literature Search Strategy
The review was prepared according to recommendations for systematic reviews and meta-analyses [62,63]. This comprehensive literature search included two online databases, PubMed and Scopus, up to February 2023. The terms "miR-429, hsa-miR-429, microRNA-429, miR-429-3p, hsa-miR-429-3p" were used as search identifiers in the relevant literature search.

Inclusion and Exclusion Criteria
The studies identified in these searches were included for further evaluation when the following criteria were fulfilled: (1) miR-429 was investigated in any type of human cell type; (2) the experimental approach aimed to determine miR-429 direct mRNA targets; (3) the study included experimental confirmation of miR-429 direct binding to the target sequence using a target mask (protector), luciferase reporters or RNA immunoprecipitation; (4) the target validation included miR-429 overexpression by synthetic analogs, such as mimics, pre-miRs, or vectors and inhibition by specific miR-429 inhibitors. Furthermore, studies were excluded according to following criteria: (1) non-English articles; (2) retracted articles; (3) errata and other types of articles such as conference records, abstracts, reviews or metanalyses; (4) insufficient data to confirm direct miR-429 interactions; (5) the prediction data were obtained from TCGA and GEO datasets. The schematic workflow of the data selection is presented in Figure 2.

Data Extraction
All included studies were independently identified by two investigators (S.B. and R.B.), and uncertain data were verified by the third author (J.C.F). The following information was collected: (1) authors names, (2) DOI, (3) publication year, (4) sample types, (5) cell types, (6) type of miRNA, (7) type of biochemical strategy used to verify hsa-miR-429 targets, (8) assays used to evaluate functional role of identified interactions and (9) translational models.

Data Extraction
All included studies were independently identified by two investigators (S.B. and R.B.), and uncertain data were verified by the third author (J.C.F). The following information was collected: (1) authors names, (2) DOI, (3) publication year, (4) sample types, (5) cell types, (6) type of miRNA, (7) type of biochemical strategy used to verify hsa-miR-429 targets, (8) assays used to evaluate functional role of identified interactions and (9) translational models.

Prediction of mRNA Targets and Gene Ontology Analysis
The miRNA-mRNA target interactions were predicted and analyzed with the use of mirDIP (microRNA Data Integration Portal and miRTarBase). The mirDIP database provides nearly 152 million human miRNA-target predictions, which were collected across 30 different resources and does not accumulate prediction bias toward biological processes or pathways [60,61]. miRTarBase contains miRNA-target interactions which were collected by manually surveying pertinent literature [61]. The Enrichr (https://amp.pharm.mssm.edu/Enrichr/, accessed on 2 February 2023) [64] and WebGestalt (https://www.webgestalt.org/, accessed on 2 February 2023) [65] webservers were applied to the analysis results into the "Gene Ontology categories" with the selection based on a Q value < 0.05. Furthermore, the analyses were limited to experimentally verified interactions, and no extended gene enrichment set analyses were performed.

Experimental Approaches to Define the Molecular Network of hsa-miR-429
The identification and solid experimental verification of both the mRNA targets as well as the transcriptional and posttranscriptional regulators of a specific miRNAs remain a major obstacle to understand their role in cellular signaling pathways. These types of analyses are mandatory for the further efficient translation of miRNAs into therapeutic approaches. Hence, although many reports regarding the mRNA targets and the

Prediction of mRNA Targets and Gene Ontology Analysis
The miRNA-mRNA target interactions were predicted and analyzed with the use of mirDIP (microRNA Data Integration Portal and miRTarBase). The mirDIP database provides nearly 152 million human miRNA-target predictions, which were collected across 30 different resources and does not accumulate prediction bias toward biological processes or pathways [60,61]. miRTarBase contains miRNA-target interactions which were collected by manually surveying pertinent literature [61]. The Enrichr (https://amp.pharm.mssm. edu/Enrichr/, accessed on 2 February 2023) [64] and WebGestalt (https://www.webgestalt. org/, accessed on 2 February 2023) [65] webservers were applied to the analysis results into the "Gene Ontology categories" with the selection based on a Q value < 0.05. Furthermore, the analyses were limited to experimentally verified interactions, and no extended gene enrichment set analyses were performed.

Experimental Approaches to Define the Molecular Network of hsa-miR-429
The identification and solid experimental verification of both the mRNA targets as well as the transcriptional and posttranscriptional regulators of a specific miRNAs remain a major obstacle to understand their role in cellular signaling pathways. These types of analyses are mandatory for the further efficient translation of miRNAs into therapeutic approaches. Hence, although many reports regarding the mRNA targets and the biological function of hsa-miR-429 accumulated over the last decade, here we focused only on the rigorous ones that were based on experimental verification.
There is no single experimental method, unfortunately, that provides researchers with total confidence that any miRNA binding to mRNA is physiologically relevant. Therefore, multiple independent experimental approaches are required. For example, miRNA overexpression with the use of synthetic analogs (miRNA mimics containing the sequences that are identical to the guide strands of the mature miRNAs) results in the dramatic and nonphysiological overexpression (from hundred-to thousandfold) of these molecules in cells. This level of overexpression oversaturates the RNA-induced silencing complexes (RISC complexes) and results in binding to mRNA targets that may not be physiologically rele-vant [66,67]. Alternative strategies to limit the miRNA overexpression effects on target gene expression in a more physiologically appropriate way include designing longer synthetic miRNA precursors such as pre-miRNAs that undergo Dicer processing in the cytoplasm or the use of pri-miRNAs that require delivery to the nucleus for processing [68][69][70][71].
A similar false positive bias can be generated by luciferase reporter approaches in which easily transfected human cell lines such as HEK293 cells serve as an environment for vectors expressing target 3'UTRs attached to the luciferase reporter as well as the miRNA analogs being tested [67,[72][73][74][75][76]. In this type of system, both the target sequences and the miRNAs are elevated to nonphysiological levels, and the cell serves as the source of the RNAi machinery. Thus, even if mutagenesis, which often relies on removing the entire target sequence, confirms a potential miRNA-mRNA binding, it still needs to be considered as an overexpression approach.
Another approach for monitoring miRNA-mRNA interactions can be achieved by the utilization of specific miRNA inhibitors that impair the endogenous miRNA levels. AntagomiRs are synthetic chemically modified ssRNAs that are fully complementary miRNAs and thereby effectively sequester mature miRNAs [77,78]. The mechanism of miRNA inhibition by these molecules depends on the type of chemical modifications used. For example, high-affinity inhibitors form heteroduplexes with the targeted miRNAs, whereas low-affinity ones promote miRNA degradation [79][80][81][82]. There are concerns even about the use of antagomiRs. For example, this approach affects the entire miRNA target network, and this could allow other genes to modulate the expression of the candidate target and therefore lead to false positives based on indirect effects [83].
A more specific and appropriate approach to verify mRNA targets utilizes target protectors (morpholinos) that prevent endogenous miRNA binding to an mRNA region containing a specific miR seed sequence [84]. These molecules are ssRNAs (of~25 bases), chemically modified to prevent them from triggering the RNAi pathway, and are complementary to an mRNA target sequence (of at least 14-15 contiguous bases). This approach blocks the relevant mRNA in question and provides the needed specificity that the other approaches lack [8,[84][85][86][87]. That being said, although they provide a better physiological insight into miRNA-mRNA interactions, it remains possible that the binding in the 3 mRNA regions may contain other miRNA target sites that could still be affected. If the target protector blocks the significant decreases in the target mRNA expression, then this concern is minimized. This target protector approach, unfortunately, would not differentiate between the targets of miR-429 and other miR-200 family members.
Another viable approach to verify miRNA-mRNA interactions relies on analyzing the composition of RISC complexes (RNA Immunoprecipitation Chip Assay, RIP assay) [88,89]. In these assays, immunoprecipitation of Argonaute RISC Catalytic Component 2 (Ago2), containing the other RISC components including all of the complexed mRNAs and miRNAs, is carried out. The verification of the miRNA and mRNA of interest is verified by qPCR or Next-Generation Sequencing (NGS) [50]. Although these assays are a solid prognostic of functional miRNA-mRNA interactions, the RISC complexes reflect all of the bound miRNAs and mRNAs in the cells. Therefore, the limitation of this approach is that, though it identifies the mRNA and miRNA contained in the RISC, it does not demonstrate that the two components actually exist in a functional hybrid. Furthermore, numerous research applications ignore the fact that even if the miRNA-related mRNA expression levels do change, it does not mean that this will lead to a change in the protein levels or the phenotype [90].
Taken together, the verification of miRNA-mRNA interactions is a complex challenge that should be based on multiple independent experimental strategies ( Figure 3). One should also consider process-related changes in the endogenous miRNA levels as well as the complexity of signaling pathways that the miRNA's target mRNA plays. Despite the large number of reports regarding hsa-miR-429 function, our inclusion criteria resulted in a final selection of 47 genes (Figure 2) that were identified either as a direct target of hsa-miR-429 or as transcriptional or posttranscriptional modulators, e.g., transcription factors or lncRNAs (sponges), respectively (Table 1).
One should also consider process-related changes in the endogenous miRNA levels well as the complexity of signaling pathways that the miRNA's target mRNA plays. D spite the large number of reports regarding hsa-miR-429 function, our inclusion crite resulted in a final selection of 47 genes (Figure 2) that were identified either as a dir target of hsa-miR-429 or as transcriptional or posttranscriptional modulators, e.g., tra scription factors or lncRNAs (sponges), respectively (Table 1). 10.1016/j.biocel.2019.05.021 10.1016/j.lfs.2020.118480 10.1096/fj.14-267054  To obtain insights into the biological role of the hsa-miR-429 network, we performed gene enrichment analyses using two well established integrative web tools, Enrichr and Webgestalt [64,65], and applied a strict False Discovery Rate-based (FDR) criteria analysis [133,134]. This approach assigned the hsa-miR-429-related target genes and indicated that this miRNA is an important regulator of processes that are either activated during development or dysregulated in human cancer and cardiovascular diseases.
In many cancers models, the repression of the endogenous levels of hsa-miR-429 leads to increased levels of ZEB1 and ZEB2 and consequently to reduced E-cadherin synthesis and increased cellular motility [138]. This mechanism is even more interesting in light of reports showing that all miR-200 family members are directly repressed transcriptional targets during ZEB1 and ZEB2 activation [139,140]. Thus, hsa-miR-429, along with the others miR-200 family members, and ZEB1 or ZEB2 function within a double-negative feedback loop, and consequently, changes in the levels of the miRNA or the ZEB factors have dramatic effects on each other. Thus, depending on the cellular signaling network, a cell can switch to either the epithelial or the mesenchymal phenotype [138].
The cellular signals that initiate increased ZEB factor expression and the related feedback loop activation include the TGF-β and TNF-α signaling pathways and the cellular responses from steroid hormones and hypoxia [138]. Interestingly, hsa-miR-429 participates in a negative feedback loop as an important direct regulator of HIF-1 signaling in normal human endothelial cells [101,103,141]. Thus, hsa-miR-429 could work as an additional safety mechanism by preventing extensive HIF-1-driven expression of ZEB factors during hypoxia. However, the consequences of such a mechanism for both the development processes and the hypoxic cancer microenvironment will require further validation.
Establishing the specific impact of hsa-miR-429 on ZEB factor expression is very challenging, since all miR-200 members are part of this negative feedback loop [138]. Furthermore, cancer-specific epigenetic changes such as silencing of Achaete-Scute Homolog (ACLS2) [96] or increasing the levels of lncRNA sponges such as MAPKAPK5 Antisense RNA 1 (MAPKAPK5-AS1) [94], circular RNA protein tyrosine kinase 2 (circPTK2) [115], cir-cLIFR [116] or metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) [99,117] that limit the miR-200 family expression, were also shown to correlate with an increased metastatic potential. The role of hsa-miR-429, however, in modulating EMT is not limited to the ZEB factors, since other direct targets include other important positive transcriptional regulators of this pathway such as EZH2 [142], RBBP4 [143] and NOTCH1 [144]. Thus, depending on the cancer type, the reduction of hsa-miR-429 levels could also accelerate the EMT completely independently of the ZEB pathway.
EZH2 has been shown to increase the expression of circRNA_0082835, an hsa-miR-429 sponge in cancer cells, creating another negative feedback loop-based regulation between EZH2 and hsa-miR-429 [123]. Interestingly, however, AKT1, that is a negative regulator of EMT [145], has been shown to be a direct target of hsa-miR-429, along with tight junction protein 1 (TJP1) and occludin (OCLN) that are repressed during the EMT process [146,147]. Thus, reduced hsa-miR-429 levels can lead to increased AKT1, TJP1 and OCLN levels and consequently limit the EMT, counteracting this miRNA effect on ZEB factors and other EMT-favoring targets.
Both hypoxia and the activation of HIF-1 signaling are common features of tumor microenvironments [162][163][164]. They lead to both hsa-miR-429 induction and the accumulation of lncRNAs that are capable of reducing this miRNA expression levels. Given the inhibitory effect of hsa-miR-429 on EMT, we speculate that HIF-1 tumor-specific impact on lncRNAs levels, e.g., MALAT1 or XIST, could potentially help cancer cells to overcome the hsa-miR-429-related blockaid and shift the cells towards EMT. Furthermore, since oncogenic alterations in miRNA levels are often associated with drug resistance [165,166] and chemoresponses [167,168], the cancer-specific deregulation of HIF-1-related hsa-miR-429 induction may have a significant impact on therapeutic efficiency. This hypothesis, however, will require in-depth mechanistic tumor-specific studies in order to understand the impact of HIF-1-related tuning of hsa-miR-429 expression on the metastatic potential and the possibility of any viable drug responses.
This complex contradictory mechanism of hsa-miR-429 involvement in EMT signaling ( Figure 4) requires further study to determine the usefulness of miR-200 family members to modulate EMT-related responses. Nevertheless, considering that miRNA acts as a buffer in gene expression profiles, the complex interactions of hsa-miR-429 may be critical for the proper tuning of developmental processes that include neurogenesis. Indeed, ASCL2, which is an important regulator of cellular differentiation [169], has been shown to induce the expression of hsa-miR-429 [96]. Obviously, further research is necessary to determine if feedback loop-based changes in hsa-miR-429 levels can work as an EMT switch during development.

Responses to Hypoxia
In both physiological and pathological conditions, the process of cellular adaptation to insufficient oxygen levels is based on the wide transcriptional reprogramming of gene expression. Reprograming is needed in order to alter metabolism toward glycolysis, prevent cell death and restore oxygen homeostasis through the activation of angiogenesis and erythropoiesis [170]. The master regulators of these cellular activities are hypoxiainducible factors (HIFs) that during hypoxia form active heterodimers of alpha subunits with their oxygen-independent beta subunits [171][172][173]. The HIF-1 complex mediates the initial and very wide reprograming of gene expression during acute hypoxia [174]. Extensive cellular adaptation to low oxygen levels, along with the intensive effort to restore the normal oxygen pressure through the upregulation of angiogenesis and erythropoiesis [50,51,174], creates a potential risk of oxidative damage (reoxygenation injury) upon the rapid restoration of the cellular oxygen levels [175].

Responses to Hypoxia
In both physiological and pathological conditions, the process of cellular adaptation to insufficient oxygen levels is based on the wide transcriptional reprogramming of gene expression. Reprograming is needed in order to alter metabolism toward glycolysis, prevent cell death and restore oxygen homeostasis through the activation of angiogenesis and erythropoiesis [170]. The master regulators of these cellular activities are hypoxia-inducible factors (HIFs) that during hypoxia form active heterodimers of alpha subunits with their oxygen-independent beta subunits [171][172][173]. The HIF-1 complex mediates the initial and very wide reprograming of gene expression during acute hypoxia [174]. Extensive cellular adaptation to low oxygen levels, along with the intensive effort to restore the normal oxygen pressure through the upregulation of angiogenesis and erythropoiesis [50,51,174], creates a potential risk of oxidative damage (reoxygenation injury) upon the rapid restoration of the cellular oxygen levels [175].
During chronic hypoxia in endothelial cells in vascularized organs such as the lung, heart, placenta and kidney [176][177][178], the other HIF isoform, HIF-2α (EPAS1), eventually replaces HIF-1 and modulates its activity to a more balanced adaptative response [173,[179][180][181]. Importantly, passing the transcriptional signal from HIF-1 to HIF-2 is necessary for proper vascularization during development [182,183]. However, the switch from HIF-1 and HIF-2 is also utilized by many cancer cells as an adaptive mechanism to the chronic oxygen deficiency in the tumor microenvironment [179,184]. Furthermore, another tissue-specific HIF alpha isoform, HIF-3α-2, accumulates during prolonged hypoxia in order to either support erythropoietin production in hepatoma cells [185] or induce the mTOR pathway in endothelial cells and eventually to direct the cells toward apoptosis [186,187].
This relationship between the HIFs and hsa-miR-429 has two important functional implications in endothelial cells. First, HIF-1 creates a negative regulatory loop with hsa-miR-429 that is responsible for decreasing the HIF1A message during hypoxia as the EPAS1 levels increase. Secondly, it prevents an extensive HIF3A accumulation during acute hypoxia that could potentially limit HIF-1 activity and induce cell death. In contrast, during prolonged hypoxia, it attenuates the continuous HIF-1 transcriptional activity through the dominant negative properties of HIF-3α [189] as well as promotes the initiation of apoptosis [186].
Taken together, in endothelial cells at least, hsa-miR-429 serves as a crucial modulator of HIF signaling. This miRNA maintains low levels of HIF1A and HIF3A expression during normal oxygen conditions, while during the early stages of hypoxia, when HIF-1α is dramatically elevated, hsa-miR-429 elevated expression maintains low levels of HIF3A mRNA and protein. During prolonged hypoxia, the HIF-1α levels decline via hsa-miR-429 mediated effects on HIF1A mRNA levels. This then leads to a reduction in has-miR-429 levels and a subsequent elevation of the HIF3A message and HIF-3α protein accumulation over time. Thus, hsa-miR-429 regulates the transitional switch between HIF-1-, HIF-2-and HIF-3-mediated transcriptional reprogramming in endothelial cells during hypoxia ( Figure 5). Although the described mechanisms are well characterized in normal endothelial cells, it should be emphasized that they may also be active in the vessels of hypoxic solid tumors [190,191]. Therefore, the tumor microenvironment may be an important determinant of hsa-miR-429 pathological function. Importantly, additional hsa-miR-429 mRNA targets were reported in hypoxia-exposed cancer cells, stressing this miRNA role in controlling both HIF signaling and hypoxia-induced angiogenesis. Similar to what was found for EMT, AKT1 as an hsa-miR-429 target, was shown to limit HIF-1α expression [192]. Thus, the induction of hsa-miR-429 by HIF-1 during acute hypoxia can serve a similar role as that of HIF-3 and prevent AKT1-related HIF-1 inactivation early during hypoxia, while this miRNA's decline during prolonged hypoxia could facilitate the HIF switch. hsa-miR-429 can also modulate the extent of hypoxia-induced angiogenesis indirectly by controlling the extent of HIF signaling and thus VEGF expression [197]. hsa-miR-429 can also modulate angiogenesis through direct interactions with other regulators including v-Crk avian sarcoma virus CT10 oncogene homolog-like (CRKL) [198], fibroblast growth factor receptor substrate 2 (FRS2) [199], tissue inhibitor of metalloproteinases 2 (TIMP2) [200] and Janus kinase 1 (JAK1) [201]. Since some of these hsa-miR-429 targets either have antiangiogenic functions (e.g., TIMP2) [200] or are cancer-type specific, further studies are necessary in order to understand how this miRNA modulates angiogenesis in different types of cancer.
Additional levels of hsa-miR-429 impact on the HIF pathway can also be provided by the direct modulation of PTEN levels [129]. PTEN is a repressor of HIF-1 transcriptional activity [193], and interestingly, the loss of expression of this gene is observed in many types of cancers [194]. Hence, hypoxic changes in hsa-miR-429 can also modulate the extent of HIF-1 transcriptional activity. Furthermore, the hypoxic elevation of hsa-miR-429 can also contribute to HIF-1-mediated reduction in peroxisome proliferator-activated receptor alpha (PPAR-α) expression during hypoxia [195]. This mechanism complements the role of the HIF switch in the adaptative response to hypoxia, since a decrease in PPARα will increase the efficiency of cellular oxygen use through the modulation of mitochondrial metabolism by decreasing the beta-oxidation of fatty acids [196].
hsa-miR-429 can also modulate the extent of hypoxia-induced angiogenesis indirectly by controlling the extent of HIF signaling and thus VEGF expression [197]. hsa-miR-429 can also modulate angiogenesis through direct interactions with other regulators including v-Crk avian sarcoma virus CT10 oncogene homolog-like (CRKL) [198], fibroblast growth factor receptor substrate 2 (FRS2) [199], tissue inhibitor of metalloproteinases 2 (TIMP2) [200] and Janus kinase 1 (JAK1) [201]. Since some of these hsa-miR-429 targets either have antiangiogenic functions (e.g., TIMP2) [200] or are cancer-type specific, further studies are necessary in order to understand how this miRNA modulates angiogenesis in different types of cancer.
As already mentioned, numerous lncRNAs that sequester (sponges) hsa-miR-429, such as MALAT1, LINC01234, XIST, MAPKAPK5-AS1 and PVT1, are dysregulated by hypoxia (usually, induced) [154][155][156][157][158][159][160][161] or promote oncological transformation [202]. However, many of them bind many other miRNAs under the very same conditions, including all miR-200 family members [203]. Our understanding of their dynamic changes during hypoxia and their preferences and efficiency in binding specific miRNAs in more physiologically relevant models remains limited. Thus, more studies are needed to establish how effective these lncRNAs are in terms of reducing the cellular hsa-miR-429 expression levels when other mRNA targets are available.

Discussion
Taken together, hsa-miR-429 appears to be critical for the regulation of cellular adaptation to hypoxia and EMT. Both of these processes are extremely important for development and for maintaining cellular homeostasis. Since these aspects of cell biology are dysregulated in human pathologies such as cardiovascular disease and cancer, understanding the role of hsa-miR-429 could be a starting point for novel therapeutic and diagnostic strategies. Nevertheless, understanding the complex and overlapping networks of hsa-miR-429 mRNA target genes that are master transcriptional factors for EMT and HIF pathways remains a difficult task. Performing mechanistic studies that would involve all of them simultaneously is currently beyond the current research possibilities.
Importantly, this problem can be partially solved by the proper identification of hsa-miR-429 direct targets and by analyzing the kinetics and priority of specific miRNA-mRNA interactions within signaling pathways, as well as the related changes in hsa-miR-429 levels and other miR-200 family members that share the same or very similar seed sequences. There are only a few reports, however, that utilize such an approach. Unfortunately, a plethora of proposed hsa-miR-429 targets were not rigorously verified and are often downstream targets of transcription factors that are themselves direct targets of hsa-miR-429. Furthermore, in terms of cancer cell lines, both miR-200 family members as well as the related signaling pathways can be affected by oncogenic transformation in a cancer/patientspecific manner, and generalizing the observation regarding hsa-miR-429 role requires direct verification in various cancer cell lines.
Taken together, despite hsa-miR-429 role (along with that of the other miR-200 family members) as an important regulator of both physiological and pathological processes, more rigorous and extensive studies are needed to decipher the network of has-miR-429 molecular targets and to estimate this miRNA family's role on cellular homeostasis. Furthermore, since both EMT and the cellular response to hypoxia are important aspects of development, hsa-miR-429 may also be involved in the modulation of these processes. Nevertheless, only a few studies have considered the role of this miRNA in development [152,204,205], and therefore more research is needed to verify this regulatory role.

Conclusions
Although the role of miRNAs in the regulation of both physiological and pathological aspects of cell biology is undisputable, translation of these ncRNAs into the clinic remains difficult. As discussed here with the hsa-miR-429 example, the main limitations of the development of efficient miRNAs-related therapies and diagnostic approaches are not only the complex networks involving these noncoding RNAs, but also the difficulties in identifying false positives. Thus, to provide miRNA-related clinical benefits, more rigorous and comprehensive research is needed to increase our understanding of the mechanisms underlying their biological role in models that are as physiologically relevant.
Author Contributions: S.B., J.S., J.F.C. and R.B. wrote, read and revised the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.