ZNF337

As found through the MyHits program (found on ExPasy), there are about 6 different motifs (or pfams) present in ZNF337.[11]

Table 1. Six different motifs within the ZNF337 protein. The KRAB box, PHD finger, Rpr2, Zinc finger (C2H2 type), Zinc finger (C3HC4 type - RING finger), and Zinc finger (MYM-type zinc finger with FCS sequence motif) all play different functions and roles.

The secondary structure of ZNF337 is predicted to have many helices, sheets, turns and coils (especially random coils) as shown below.[12][13]

Both the H. Sapiens and P. troglodyte secondary structures are extremely similar; however, it is interesting to compare to S. dumerili where there is a stronger presence of sheets and coils between both 200-300 bp and 400-500 bp positions instead of sheets and helices. Additionally, comparing the beginning of the secondary structure (0-14 bp) of all species/orthologs shows that coils and turns make up the majority of the beginning, but not as much in some species such as S. dumerili (more helices and sheets instead).

P. troglodytes (95.6% identical to human protein) secondary structure

S. dumerili (30.9% identical to human protein) secondary structure

C. japonica (1.9% identical to human protein) secondary structure

H. Sapiens ZNF337 secondary structure

Several tertiary structure modeling programs were unable to construct a model for ZNF337. When using the SWISS-model program, some models were constructed, however, to ZNF568. The ZNF568 protein sequence is 45.20% identical to that of ZNF337, has a sequence similarity of 0.44, and coverage of 0.37 with a range between the 345-623 bp amino acids in the ZNF337 protein sequence.[14] The predicted tertiary structure is shown in Figure 1. In this figure, there are several zinc ion ligands.

Figure 1. Predicted tertiary structure of ZNF568, which is 45.20% identical to the ZNF337 protein sequence. Created using SWISS-MODEL.

ZNF568 is a protein coding gene, associated with diseases such as transient neonatal diabetes mellitus. It has transcriptional repression activity, partially through the recruitment of the co-repressor TRIM28, but also has repression activity independently of this interaction. It is specifically important during embryonic development, where it acts as a direct repressor of a placental-specific transcript of IGF2 in early development and regulates convergent extension movements required for axis elongation and tissue morphogenesis in all germ layers. It is also crucial for normal morphogenesis of extraembryonic tissues including the yolk sac, extraembryonic mesoderm and placenta. Interestingly, it may enhance proliferation or maintenance of neural stem cells [15]

The promoter region was chosen using ElDorado at Genomatrix, which assessed the ZNF337 gene locus for possible promoter regions. Out of the six possible promoter regions and sets, promoter set 6 (GXP_8991829) was chosen as it is the one best supported by transcripts (has six transcript ID’s). Its start position is 25696627, its end position is 25697904 and its length is 1278 base pairs. Within GXP_8991829 (-), coding transcript GXT_26235925 was chosen as it has 5 exons, 37,403 CAGE tags, and corresponds with accession number XM_006723558 in NCBI (see Figure 2).

The promoter sequence contains a CpG island with a CpG count of 138. There is also a DNAse cluster (score =1000) present within the promoter sequence.

Possible transcription factors for the ZNF337 promoter region were determined using ElDorado at Genomatrix. These are listed below in Table 2.

Table 2. Possible transcription factors for ZNF337 promoter region.

There are 340 factors from 129 cell types of Transcription Factor ChIP-seq Clusters (from Encode3).[16] With that said, only the strong ones (indicated as black or dark grey) that also contain peaks within the promoter or enhancer regions are shown in Table 3.

Table 3. Transcription Factor-ChIP Clusters associated with specific cell types.

According to ORegAnno (literature curated TFBSs), there is no TF-ChIP signal overlap within the promoter/enhancer regions. Most of the ORegAnno citations correlate with a “NANP” gene, while transcription factors CTCF and CEBPA are confirmed in the enhancer region for the ZNF337 gene.[16]

Both RNA sequence data from the Gene database records at NCBI and the Human Protein Atlas [17] using immunohistochemical staining to determine protein in various tissues show that the ZNF337 protein is expressed in many tissues. While ZNF337 mRNA tissue specificity is expressed in low tissue specificity levels, the mRNA is notably expressed in the cerebellum (brain) but is also more highly expressed in all tissues (distribution in all) compared to protein expression, especially higher in female tissues.

An antibody was developed against a recombinant protein corresponding to amino acids: ESSQGQRENPTEIDKVLKGIENSRWGAFKCAERGQDFSRKMMVIIHKKAHSRQKLFTCRECHQGFRDESALLLHQN. The specificity of human ZNF337 antibody was verified on a Protein Array containing target protein plus 383 other non-specific proteins. This isotype is IgG, its clonality is polyclonal, its host is rabbit, and its purity is immunogen affinity purified. This staining of human cerebellum shows cytoplasmic positivity in Purkinje cells (which regulate and coordinate motor movements through inhibitory functions and neurotransmitters).[18]

While there is little-some expressivity in a wide range of tissues, together, these results indicate a trend that expressivity is highest and most present in the brain, particularly the cerebellum. A few experiments and results also indicate expressivity in female (and some male) reproductive tissues.

The prediction for localization of ZNF337 is highest in the nucleus (nuclear) at 95.7% followed by 4.3% in the mitochondria (mitochondrial).[19]

ZNF337 contains many predicted post-translational domains such as phosphorylation (serine and tyrosine kinases),[20] PEST motifs,[21] O-GlcNAc sites,[22] SUMOylation,[23] and glycation [24] as seen below:

Table 4. Different post-translation modifications. Modifications can alter protein structure, thus affecting overall protein function and viability.

• PEST motifs are high in proline (P), glutamic acid (E), serine (S), and threonine (T). PEST motifs are linked to a reduced half-life of proteins and are potential targets for proteolytic degradation/cleavage sites.
• O-GlcNAc modifications may compete with phosphorylation for control of a protein’s activation state.
• SUMOylation can interfere with the interaction between the target and its partner, can provide a binding site for an interacting partner, result in a conformational change of the modified target, or facilitate or antagonize ubiquitination.
• Glycation sites indicate the potential for a nonenzymatic process to occur, in which proteins react with reducing sugar molecules and therefore impair the function and change the characteristics of the protein. Because there are quite a few of these glycation sites, there could be a somewhat high probability that it is easy to change the function or characteristic of the protein.

No predicted transmembrane domains were identified from tests run through SOSUI.[25] A prediction for a new signal peptide is very low and negative at -3.83. The GvH is also very negative at -8.69 (with a possible cleavage site between amino acids 56 and 57), indicating a low possibility that it has a cleavable signal sequence. Thus, ZNF337 is predicted to have no N-terminal signal peptide. Also, Reinhardt’s method for cytoplasmic/nuclear discrimination has a cytoplasmic prediction for ZNF337 with a reliability score of 94.1.[19]

The nuclear localization signal is somewhat low at 0.75. Orthologs (P. troglodytes, S. dumerili, and C. asiatica ) were used to confirm the significance of these predictions. Likewise, there were no predictions of no N-terminal peptide signals and transmembrane domains. All these ZNF337 orthologous proteins confirmed the prediction of nuclear location at 95.7%.