The entire DCD gene is approximately 3.9 Kb; starts at 54644589 and ends at 54648493 bp; orientation: reverse strand.

The DCD gene encodes 3 transcript variants (splice variants). There are two transcript variants deposited in the NCBI database (https://www.ncbi.nlm.nih.gov/gene) and one additional transcript variant reported in Ensembl (http://www.ensembl.org/). The transcript variant 1 is formed by 5 exons and 5 coding exons, resulting in a transcript length of 519 bps and a translation length of 110 amino acids (aa). The transcript variant 2 is the longest variant, with 6 exons and 6 coding exons (transcript length of 670 bps) and a translation length of 121 aa. The transcript variant 3 is formed by 6 exons and 3 coding exons (transcript length of 773 bps), which results in the shorter translation variant, with 77 aa.

The full-length dermcidin protein has 110 amino acids (aa) that consists of a 19-amino-acid signal peptide, a 43-amino-acid prodomain and a 48-amino-acid antimicrobial peptide domain (Schittek, 2012). This precursor is proteolytically processed into several derivatives. DCD-1 (47 aa), DCD-1L (48 aa), SSL-25 (25 aa) and SSL-23 (23 aa), the derivatives with the best described function, are secreted by eccrine sweat glands as antimicrobial peptides (Schittek et al., 2001; Steffen et al., 2006). The Y-P30 (30 aa) derivative has been described as a neuron survival-promoting peptide (Cunningham et al., 2000), while the PIF (20 aa) - Proteolysis-Inducing Factor, a proteolytic glycoprotein, is associated to a severe loss of skeletal muscles in cachectic cancer patients (Todorov et al., 1996; Lorite et al., 1997). The full primary structure of DCD is illustrated in Figure 1.

Dermcidin is constitutively expressed by the dark mucous cells of the secretory coil of eccrine sweat glands, unlike the other antimicrobial peptides that are induced by injury and inflammation (Schittek et al., 2001; Rieg et al., 2004; Rieg et al., 2006). In these glands, the DCD mRNA is translated into a 110-amino acid peptide that can be processed and secreted into sweat or cleaved by several proteases in this fluid (Schittek et al., 2001; Schittek, 2012; Rieg et al., 2004; Flad et al., 2002). There are 14 DCD-derived peptides that occur in sweat. The C-terminal region of the protein contains the majority of the peptides with antimicrobial activity: DCD-1, DCD-1L, SSL-23, SSL-25, SSL-45 and SSL-46 (Schittek et al., 2001; Steffen et al., 2006; Mühlhäuser et al., 2017). Furthermore, there is one peptide derived from the prodomain YDP-42 (Schittek, 2012). Even DCD-1 or DCD-1L can be processed in sweat into smaller peptides like LEK-24 and SSL-25 by enzymes such as cathepsin D, 1,10-phenanthroline-sensitive carboxypeptidase and endoproteases (Baechle et al., 2006).
Dermcidin is also expressed by sebaceous glands and neutrophils in immunity-related circumstances (Dahlhoff et al., 2016; Lominadze et al., 2005) and, moreover, has been shown to be overexpressed in the context of HIV infection of monocytoid cells (Pathak et al., 2009). Additionally, oxidatively stressed neural cell lines express dermcidin in the form of the signal peptide Y-P30, described firstly as DSEP, the diffusible survival/evasion peptide, which can, indeed, be detected in culture media (Cunningham et al., 2002).
The human term placental tissue may express DCD as two splice variants under a restricted spatiotemporal pattern, which may be related to the involvement of dermcidin in the molecular mechanisms of pregnancy. Placental dermcidin can be processed into DCD-1 and defend the fetus against invading microorganisms (Motoyama et al., 2007), or into Y-P30 to exert neuritogenic activity during fetus brain development (Mikhaylova et al., 2014).
In plasma, dermcidin may be found in different conditions: 1) in hepatocellular carcinoma patients (Qiu et al., 2018); 2) overexpressed in patients with arterial hypertension, as a player in modulation of NO levels and induction of platelet aggregation (Ghosh et al., 2011; Ghosh et al., 2012a); 3) in type I diabetes mellitus, inhibiting the synthesis of insulin in hepatocytes and pancreatic cells (Ghosh et al., 2011; Ghosh et al., 2012b). Dermcidin is also secreted by ischemic skeletal muscles, enhancing cardiomyocytes apoptosis under hypoxic conditions and infarct size after permanent coronary artery ligation (Esposito et al., 2015). Moreover, dermcidin is abundantly present in exhaled breath condensate of asthmatic and lung cancer patients, and it can be regarded as a biomarker in these diseases (Bloemen et al., 2001; Chang et al., 2010).
Also, dermcidin has revealed a relevant role in cancer; it offers a survival advantage in some tumors, as reported for breast cancers (Porter et al., 2003; Brauer et al., 2014); prostate cancer (Wang et al., 2003; Stewart et al., 2007); lung cancer, as a biomarker in exhaled breath condensate (Chang et al., 2010); hepatic cancer cells (Lowrie et al., 2006; Shen et al., 2011; Qiu et al., 2018; Lowrie et al., 2011); gastro-oesophageal cancer (Deans et al., 2006); pancreatic cancer cells (Stewart et al., 2008a); a myelogenous leukemia cell line (Stocki et al., 2011); and melanoma (Smith et al., 2005; Rieg et al., 2004; Trzoss et al., 2014). Additionally, tumor cells express PIF, which is resistant to proteolytic digestion by trypsin and the human cancer expression has been related to inhibition of muscle cell differentiation and a high weight loss (Majczak et al., 2007; Todorov et al., 1997; Wigmore et al., 2000; Jiang and Clemens, 2006; Wieland et al., 2007). In fact, PIF was discovered in 1996 in mice and later described as a potent catabolic factor that acts as a cachectic cancer factor (Tisdale et al., 2004; Schittek et al., 2012; Todorov et al., 1996). Although the role of murine PIF is well established, its biology remains speculative in humans (Stewart et al., 2008b).

Dermcidin is mainly found in the cytoplasm or extracellularly.
In the dark mucous cells of the secretory coil of eccrine sweat glands, dermcidin is detected in the Golgi complex and in the secretory granules, released in the cell surface and transported to epidermal surface by the sweat duct (Schittek et al., 2001; Rieg et al., 2006, Sakurada et al., 2010). There are differences in the levels of DCD-derived peptides in eccrine sweat of distinct anatomic sites; highest concentrations are found in palms, arms and forehead, where the density of eccrine sweat glands is higher and the body is more exposed to pathogens and minor trauma (Rieg et al., 2004).
Although less profusely, dermcidin is also found in basal tears, in cervicovaginal fluid and in breast milk, in which it also assumes a function associated to innate immunity (You et al., 2010; Shaw et al., 2007; Azkargorta et al., 2015; Chow et al., 2016). Furthermore, dermcidin-derived peptides can be found in lipid droplets secreted by the holocrine sebaceous glands and in the granules of neutrophils, suggesting that this protein has, likewise, an immunity related purpose in these cells (Dahlhoff et al., 2016; Lominadze et al., 2005).

DCD-1, DCD-1L, SSL-25, SSL-23, SSL-45 and SSL-46 peptides have a broad and overlapping spectrum regarding antimicrobial activity against, for example, Staphylococcus aureus, Enterococcus faecalis, Staphylococcus epidermidis, Escherichia coli, Pseudomonas aeruginosa and Candida albicans (Schittek et al., 2001; Steffen et al., 2006; Mühlhäuser et al., 2017). Anionic DCD-1L monomers can bind bacterial membrane and fold into an amphiphilic alpha-helix, interacting parallelly to the surface. Further studies indicate the proteins can self-assemble into high-order oligomers and form ion channels, which can lead to bacterial cell death by abrogation of the transmembrane potential (Paulmann et al., 2012; Song et al., 2013). Such conformation is indeed favored by transmembrane potential as well as by the presence of Zn²⁺ due to neutralization of protein charge. On the other hand, SSL-29, LEK-44, LEK-45 and YDP-42 peptides showed no antimicrobial activity against the same bacterial species: S. aureus, E. faecalis, S. epidermidis, E. coli and P. aeruginosa (Rieg et al., 2005; Steffen et al., 2006).
DCD-1L has been studied as a stimulant for production of pro-inflammatory cytokines and chemokines by human keratinocytes cells controlled by G-protein and mitogen-activated protein kinase (MAPK) signaling pathways, thus contributing to the cutaneous immunity (Niyonsaba et al., 2009).
High blood levels of dermcidin variant 2 was found in patients following acute ischemic heart disease. There have been further investigations on rather such variant may inhibit synthesis and release of insulin by the pancreas, an effect that was then shown to be reversed by aspirin or insulin administration (Bank et al., 2014; Ghosh et al., 2014). As a consequence of insulin resistance, there may be increase in blood pressure and atherosclerosis through inhibition of NO production and promotion of prothrombotic effects (Ghosh et al., 2011). The same role for dermcidin was observed in ischemic and hemorrhagic stroke patients (Bank et al., 2015).
Dermcidin has been shown to interact with few proteins within the cell. In this context, phosphorylation at Tyr20 enabled dermcidin to interact with the SH2 domain of NCK1, which, through effector proteins like the PAK1 /ARHGEF6 (also called PIX) complex, stimulates Rho GTPases RAC1 and CDC42 and reorganizes actin cytoskeleton, ultimately promoting cell migration in a hepatocellular carcinoma cell line (SK-HEP1) (Shen et al., 2011). Furthermore, in a gastric cancer cell line (BGC-823), dermcidin protein was one of the proteins bound to a long non-coding RNA (STCAT3), while overexpression of both dermcidin and STCAT3 were related to poor prognosis for patients (Zhang et al., 2018). Considering small molecules - natural products, in this case - dermcidin has been shown to bind paclitaxel and seriniquinone, a marine bacterium derived compound. In the first setting, dermcidin precursor was recovered, along with HSP90AA1 and actinin, from a paclitaxel-biotin probe fed to drug-sensitive breast cancer cell line (MCF7), but not from their paclitaxel-resistant analogue, suggesting a role of dermcidin in the cell resistance profile to this compound and cross-resistance to other anticancer drugs (Zuo et al., 2010). Regarding the later natural product , dermcidin protein was revealed to be a direct target of a seriniquinone probe, which was found to be further bound to other proteins, such as GAPDH and Hsp70, in a colorectal carcinoma cell line (HCT 116) model (Trzoss et al., 2014).
Y-P30 was first seen overexpressed as a neuron survival-promoting peptide in response to oxidative stress (Cunningham et al., 1998), promoting survival of cortical neurons after cerebral lesions by retaining calreticulin in the cytosol and reducing calcium signaling to a degree in which it is protective, regardless of any immune cell inhibition (Cunningham et al., 2000). During hypoxia and hypoglycemia, Y-P30 expression is associated with reduced neural cell death in vitro (Schneeberg et al., 2009). This protein is also related to induction of neurite outgrowth in cortical neurons during early brain development, a role that is suggested by the effect of binding of Y-P30 to PTN (pleiotrophin) in the extracellular axonal membrane, thus increasing syndecan local concentrations and promoting neurite growth (Landgraf et al., 2008). Furthermore, this mechanism can be related to calcium/calmodulin-dependent serine kinase ( CASK) migration to the nucleus, where Y-P30 interacts differently with syndecan-2 ( SDC2) or -3 ( SDC3) in immature or mature neurons, contributing with migration and axonal outgrowth (Landgraf et al., 2014; Neumann et al., 2019).
PIF has been investigated as a catalytic protein, which can induce proteasome expression through NF-kB activation (Wyke and Tisdale, 2005). Consequently, this process results in decreased protein synthesis and severe degradation of skeletal muscle, thus correlating with the weight loss in cachectic cancer patients, an effect typically attributed to PIF (Todorov et al., 1996; Lorite et al., 1997; Todorov et al., 2007). PIF was also associated with the release of Ca²⁺ from intracellular stocks, suggesting PIF receptor (a zinc-sensing receptor) is coupled to a G-protein to cause calcium release from endoplasmic reticulum (Mirza and Tisdale, 2012). Reinforcing these data, direct and indirect inhibitors of NF-kB were able to attenuate the development of muscle loss (Wyke et al., 2004; Russell et al., 2007).

The DCD gene is highly homologous among different non-human primate species, as shown in Table 1, which compares DCD isoform 1 among species.
Table 1. Comparative identity of human DCD with other species.

% Identity for: Homo sapiens CDC	Symbol	Protein	DNA
vs. P. paniscus	DCD	100	100
vs. P. abelii	DCD	96.9	98.2
vs. N. leucogenys	DCD	93.8	98.4
vs. H. moloch	DCD	92.8	91.7
vs. T. francoisi	DCD	91.7	95.3
vs. R. roxellana	DCD	91.7	94.2
vs. C. angolensis palliatus	DCD	91.7	94
vs. C. jacchus	DCD	91	96
vs. P. tephrosceles	DCD	90.7	92.7
vs. M. mulatta	DCD	89.7	95.2
vs. T. gelada	DCD	89.7	91.1
vs. M. leucophaeus	DCD	89.7	95.9
vs. M. nemestrina	DCD	89.7	89.9
vs. C. sabaeus	DCD	84.8	93.7
vs. A. nancymaae	DCD	82.3	96.8
vs. S. boliviensis boliviensis	DCD	80.4	95.0
vs. S. apella	DCD	77.3	94.3
vs. C. capucinus imitator	DCD	76.5	95.2

(Source: http://www.ncbi.nlm.nih.gov/homologene)

From 37420 unique samples reported in COSMIC (Catalogue of Somatic Mutations in Cancer; http://cancer.sanger.ac.uk/cancergenome/projects/cosmic), a total of 145 unique samples presented DCD mutations, of which 67 different mutations are distributed among 47 missense substitutions, 16 synonymous substitutions, 2 frameshift insertion and 2 frameshift deletions. Somatic mutation frequency of 0.3% is similar to that reported in cBioPortal (http://www.cbioportal.org). From 10967 samples, 31 mutations are classified as 25 missense substitutions and 6 truncated mutations.