What's the difference between CREs and DHSs?

What's the difference between CREs and DHSs?

We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

I would like to know what's the difference between cis-regulating elements and DNase I-hypersensitive sites, in order to produce a meaningful segregation of non coding elements affecting gene expression.

Cis-regulatory elements are simply DNA regions upstream or downstream of a gene that can affect its expression (basically they have to be in the same chromosome).

DNAse-I hypersensitive sites (DHS) are regions of chromatin that get digested during the DNAse treatment because they are exposed i.e. not protected by a protein (complex). The protein complex can either be a nucleosome or a transcription factor. It should be noted that a DNA region can be DNAse insensitive also because of its conformation.

Usually the cis-regulatory elements exert their effects by serving as a binding site for a transcription factors (or chromatin remodellers). If the proteins bind at these sites then these regions would be protected from DNAse digestion thereby making them not-DHS. However, it is known that DNA elements can act in trans as well (see papers on 3C- chromosome conformation capture, and its variants). Moreover, whether a site is protected from DNAse or not does not tell anything about its function.

Accessibility Control of V(D)J Recombination: Lessons from Gene Targeting

William M. Hempel , . Pierre Ferrier , in Advances in Immunology , 1998

A A Locus Control Region (LCR) Downstream of the TCRα/δ Locus?

A set of DNase I hypersensitive sites , including one within the well-characterized TCR α gene enhancer (Eα), has been identified downstream of the TCRα constant region. On the basis of transgenic mouse experiments, it was reported that all of the HS sites were required in addition to Eα to confer integration site-independent, copy-number-dependent, and tissue-restricted expression of associated transgenes ( Diaz et al., 1994 Ortiz et al., 1997 ). These properties are characteristic of a type of cis-regulatory element known as a locus control region (LCR), initially described within the β-like globin gene complex ( Martin et al., 1996 , and references within). By analogy with the β-globin LCR, which has been shown to participate in organizing the chromatin configuration of the entire β-like globin gene locus, Winoto and colleagues have proposed a model in which the TCRα downstream sequences, acting as an LCR element, could be shared by the two TCRδ and TCRα genes and play a major role in the control of the differential recombination and expression of the two parts of this locus ( Diaz et al., 1994 ).

This model has been tested by targeted deletion of a 10-kb region immediately 3′ of the Eα element, including most of the HS sites responsible for LCR activity in transgenes ( Hong et al., 1997 also see Fig. 1 ). The neo r gene was left in place at the mutated locus. Possibly because the targeted deletion also affected the third exon of a closely located, inversely oriented gene, DAD1, no animals homozygous for the mutation could be obtained. Instead, the heterozygous mutants were mated with TCRα −/− ( Mombaerts et al., 1992a ) or TCRδ −/− ( Itohara et al., 1993 ) mice, so that only the TCR chains from the mutated allele could be expressed. No difference was found in the development of αβ and γδ cells in the double mutated animals either in the thymus or in the periphery when compared to control wild-type mice. Furthermore, in the whole thymus, the expression of the targeted TCRα allele as measured by RNase protection and flow cytometry was unaffected. The only identified consequence of the mutation seems to be a modest reduction in the number of mature thymocytes expressing high levels of TCRα. One interpretation that the authors provide to explain the minor effect of the LCR deletion is that the transcription of the residual neo r gene may render the chromatin accessible to the recombination machinery, in affect replacing the normal function of the LCR at the targeted TCRα/δ locus. Although this interpretation runs counter to the usual effect of an inserted neo r cassette on gene expression ( Olson et al., 1996 ), β-globin LCR activity ( Kim et al., 1992 ), or V(D)J recombination (see earlier), it is not out of the realm of possibility. In any case, the credibility of the TCRα LCR is at stake unless this hypothesis can be verified by removing the selectable cassette after targeting.

Cis-regulatory elements and human evolution

Modification of gene regulation has long been considered an important force in human evolution, particularly through changes to cis-regulatory elements (CREs) that function in transcriptional regulation. For decades, however, the study of cis-regulatory evolution was severely limited by the available data. New data sets describing the locations of CREs and genetic variation within and between species have now made it possible to study CRE evolution much more directly on a genome-wide scale. Here, we review recent research on the evolution of CREs in humans based on large-scale genomic data sets. We consider inferences based on primate divergence, human polymorphism, and combinations of divergence and polymorphism. We then consider ‘new frontiers’ in this field stemming from recent research on transcriptional regulation.

Current address: Simons Center for Quantitative Biology, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724, USA.


Neural crest was first described in the chick embryo by Wilhelm His Sr. in 1868 as "the cord in between" (Zwischenstrang) because of its origin between the neural plate and non-neural ectoderm. [1] He named the tissue ganglionic crest since its final destination was each lateral side of the neural tube where it differentiated into spinal ganglia. [6] During the first half of the 20th century the majority of research on neural crest was done using amphibian embryos which was reviewed by Hörstadius (1950) in a well known monograph. [7]

Cell labeling techniques advanced the field of neural crest because they allowed researchers to visualize the migration of the tissue throughout the developing embryos. In the 1960s Weston and Chibon utilized radioisotopic labeling of the nucleus with tritiated thymidine in chick and amphibian embryo respectively. However, this method suffers from drawbacks of stability, since every time the labeled cell divides the signal is diluted. Modern cell labeling techniques such as rhodamine-lysinated dextran and the vital dye diI have also been developed to transiently mark neural crest lineages. [6]

The quail-chick marking system, devised by Nicole Le Douarin in 1969, was another instrumental technique used to track neural crest cells. [8] [9] Chimeras, generated through transplantation, enabled researchers to distinguish neural crest cells of one species from the surrounding tissue of another species. With this technique, generations of scientists were able to reliably mark and study the ontogeny of neural crest cells.

A molecular cascade of events is involved in establishing the migratory and multipotent characteristics of neural crest cells. This gene regulatory network can be subdivided into the following four sub-networks described below.

Inductive signals Edit

First, extracellular signaling molecules, secreted from the adjacent epidermis and underlying mesoderm such as Wnts, BMPs and Fgfs separate the non-neural ectoderm (epidermis) from the neural plate during neural induction. [1] [4]

Wnt signaling has been demonstrated in neural crest induction in several species through gain-of-function and loss-of-function experiments. In coherence with this observation, the promoter region of slug (a neural crest specific gene) contains a binding site for transcription factors involved in the activation of Wnt-dependent target genes, suggestive of a direct role of Wnt signaling in neural crest specification. [10]

The current role of BMP in neural crest formation is associated with the induction of the neural plate. BMP antagonists diffusing from the ectoderm generates a gradient of BMP activity. In this manner, the neural crest lineage forms from intermediate levels of BMP signaling required for the development of the neural plate (low BMP) and epidermis (high BMP). [1]

Fgf from the paraxial mesoderm has been suggested as a source of neural crest inductive signal. Researchers have demonstrated that the expression of dominate-negative Fgf receptor in ectoderm explants blocks neural crest induction when recombined with paraxial mesoderm. [11] The understanding of the role of BMP, Wnt, and Fgf pathways on neural crest specifier expression remains incomplete.

Neural plate border specifiers Edit

Signaling events that establish the neural plate border lead to the expression of a set of transcription factors delineated here as neural plate border specifiers. These molecules include Zic factors, Pax3/7, Dlx5, Msx1/2 which may mediate the influence of Wnts, BMPs, and Fgfs. These genes are expressed broadly at the neural plate border region and precede the expression of bona fide neural crest markers. [4]

Experimental evidence places these transcription factors upstream of neural crest specifiers. For example, in Xenopus Msx1 is necessary and sufficient for the expression of Slug, Snail, and FoxD3. [12] Furthermore, Pax3 is essential for FoxD3 expression in mouse embryos. [13]

Neural crest specifiers Edit

Following the expression of neural plate border specifiers is a collection of genes including Slug/Snail, FoxD3, Sox10, Sox9, AP-2 and c-Myc. This suite of genes, designated here as neural crest specifiers, are activated in emergent neural crest cells. At least in Xenopus, every neural crest specifier is necessary and/or sufficient for the expression of all other specifiers, demonstrating the existence of extensive cross-regulation. [4] Moreover, this model organism was instrumental in the elucidation of the role of the Hedgehog signaling pathway in the specification of the neural crest, with the transcription factor Gli2 playing a key role. [14]

Outside of the tightly regulated network of neural crest specifiers are two other transcription factors Twist and Id. Twist, a bHLH transcription factor, is required for mesenchyme differentiation of the pharyngeal arch structures. [15] Id is a direct target of c-Myc and is known to be important for the maintenance of neural crest stem cells. [16]

Neural crest effector genes Edit

Finally, neural crest specifiers turn on the expression of effector genes, which confer certain properties such as migration and multipotency. Two neural crest effectors, Rho GTPases and cadherins, function in delamination by regulating cell morphology and adhesive properties. Sox9 and Sox10 regulate neural crest differentiation by activating many cell-type-specific effectors including Mitf, P0, Cx32, Trp and cKit. [4]

The migration of neural crest cells involves a highly coordinated cascade of events that begins with closure of the dorsal neural tube.

Delamination Edit

After fusion of the neural fold to create the neural tube, cells originally located in the neural plate border become neural crest cells. [17] For migration to begin, neural crest cells must undergo a process called delamination that involves a full or partial epithelial-mesenchymal transition (EMT). [18] Delamination is defined as the separation of tissue into different populations, in this case neural crest cells separating from the surrounding tissue. [19] Conversely, EMT is a series of events coordinating a change from an epithelial to mesenchymal phenotype. [18] For example, delamination in chick embryos is triggered by a BMP/Wnt cascade that induces the expression of EMT promoting transcription factors such as SNAI2 and FoxD3. [19] Although all neural crest cells undergo EMT, the timing of delamination occurs at different stages in different organisms: in Xenopus laevis embryos there is a massive delamination that occurs when the neural plate is not entirely fused, whereas delamination in the chick embryo occurs during fusion of the neural fold. [19]

Prior to delamination, presumptive neural crest cells are initially anchored to neighboring cells by tight junction proteins such as occludin and cell adhesion molecules such as NCAM and N-Cadherin. [20] Dorsally expressed BMPs initiate delamination by inducing the expression of the zinc finger protein transcription factors snail, slug, and twist. [17] These factors play a direct role in inducing the epithelial-mesenchymal transition by reducing expression of occludin and N-Cadherin in addition to promoting modification of NCAMs with polysialic acid residues to decrease adhesiveness. [17] [21] Neural crest cells also begin expressing proteases capable of degrading cadherins such as ADAM10 [22] and secreting matrix metalloproteinases (MMPs) that degrade the overlying basal lamina of the neural tube to allow neural crest cells to escape. [20] Additionally, neural crest cells begin expressing integrins that associate with extracellular matrix proteins, including collagen, fibronectin, and laminin, during migration. [23] Once the basal lamina becomes permeable the neural crest cells can begin migrating throughout the embryo.

Migration Edit

Neural crest cell migration occurs in a rostral to caudal direction without the need of a neuronal scaffold such as along a radial glial cell. For this reason the crest cell migration process is termed “free migration”. Instead of scaffolding on progenitor cells, neural crest migration is the result of repulsive guidance via EphB/EphrinB and semaphorin/neuropilin signaling, interactions with the extracellular matrix, and contact inhibition with one another. [17] While Ephrin and Eph proteins have the capacity to undergo bi-directional signaling, neural crest cell repulsion employs predominantly forward signaling to initiate a response within the receptor bearing neural crest cell. [23] Burgeoning neural crest cells express EphB, a receptor tyrosine kinase, which binds the EphrinB transmembrane ligand expressed in the caudal half of each somite. When these two domains interact it causes receptor tyrosine phosphorylation, activation of rhoGTPases, and eventual cytoskeletal rearrangements within the crest cells inducing them to repel. This phenomenon allows neural crest cells to funnel through the rostral portion of each somite. [17]

Semaphorin-neuropilin repulsive signaling works synergistically with EphB signaling to guide neural crest cells down the rostral half of somites in mice. In chick embryos, semaphorin acts in the cephalic region to guide neural crest cells through the pharyngeal arches. On top of repulsive repulsive signaling, neural crest cells express β1and α4 integrins which allows for binding and guided interaction with collagen, laminin, and fibronectin of the extracellular matrix as they travel. Additionally, crest cells have intrinsic contact inhibition with one another while freely invading tissues of different origin such as mesoderm. [17] Neural crest cells that migrate through the rostral half of somites differentiate into sensory and sympathetic neurons of the peripheral nervous system. The other main route neural crest cells take is dorsolaterally between the epidermis and the dermamyotome. Cells migrating through this path differentiate into pigment cells of the dermis. Further neural crest cell differentiation and specification into their final cell type is biased by their spatiotemporal subjection to morphogenic cues such as BMP, Wnt, FGF, Hox, and Notch. [20]

Neurocristopathies result from the abnormal specification, migration, differentiation or death of neural crest cells throughout embryonic development. [24] [25] This group of diseases comprises a wide spectrum of congenital malformations affecting many newborns. Additionally, they arise because of genetic defects affecting the formation of neural crest and because of the action of Teratogens [26]

Waardenburg's syndrome Edit

Waardenburg's syndrome is a neurocristopathy that results from defective neural crest cell migration. The condition's main characteristics include piebaldism and congenital deafness. In the case of piebaldism, the colorless skin areas are caused by a total absence of neural crest-derived pigment-producing melanocytes. [27] There are four different types of Waardenburg's syndrome, each with distinct genetic and physiological features. Types I and II are distinguished based on whether or not family members of the affected individual have dystopia canthorum. [28] Type III gives rise to upper limb abnormalities. Lastly, type IV is also known as Waardenburg-Shah syndrome, and afflicted individuals display both Waardenburg's syndrome and Hirschsprung's disease. [29] Types I and III are inherited in an autosomal dominant fashion, [27] while II and IV exhibit an autosomal recessive pattern of inheritance. Overall, Waardenburg's syndrome is rare, with an incidence of

2/100,000 people in the United States. All races and sexes are equally affected. [27] There is no current cure or treatment for Waardenburg's syndrome.

Hirschsprung's Disease Edit

Also implicated in defects related to neural crest cell development and migration is Hirschsprung's disease (HD or HSCR), characterized by a lack of innervation in regions of the intestine. This lack of innervation can lead to further physiological abnormalities like an enlarged colon (megacolon), obstruction of the bowels, or even slowed growth. In healthy development, neural crest cells migrate into the gut and form the enteric ganglia. Genes playing a role in the healthy migration of these neural crest cells to the gut include RET, GDNF, GFRα, EDN3, and EDNRB. RET, a receptor tyrosine kinase (RTK), forms a complex with GDNF and GFRα. EDN3 and EDNRB are then implicated in the same signaling network. When this signaling is disrupted in mice, aganglionosis, or the lack of these enteric ganglia occurs. [30]

Fetal Alcohol Spectrum Disorder Edit

Prenatal alcohol exposure (PAE) is among the most common causes of developmental defects. [31] Depending on the extent of the exposure and the severity of the resulting abnormalities, patients are diagnosed within a continuum of disorders broadly labeled Fetal Alcohol Spectrum Disorder (FASD). Severe FASD can impair neural crest migration, as evidenced by characteristic craniofacial abnormalities including short palpebral fissures, an elongated upper lip, and a smoothened philtrum. However, due to the promiscuous nature of ethanol binding, the mechanisms by which these abnormalities arise is still unclear. Cell culture explants of neural crest cells as well as in vivo developing zebrafish embryos exposed to ethanol show a decreased number of migratory cells and decreased distances travelled by migrating neural crest cells. The mechanisms behind these changes are not well understood, but evidence suggests PAE can increase apoptosis due to increased cytosolic calcium levels caused by IP3-mediated release of calcium from intracellular stores. It has also been proposed that the decreased viability of ethanol-exposed neural crest cells is caused by increased oxidative stress. Despite these, and other advances much remains to be discovered about how ethanol affects neural crest development. For example, it appears that ethanol differentially affects certain neural crest cells over others that is, while craniofacial abnormalities are common in PAE, neural crest-derived pigment cells appear to be minimally affected. [32]

DiGeorge syndrome Edit

DiGeorge syndrome is associated with deletions or translocations of a small segment in the human chromosome 22. This deletion may disrupt rostral neural crest cell migration or development. Some defects observed are linked to the pharyngeal pouch system, which receives contribution from rostral migratory crest cells. The symptoms of DiGeorge syndrome include congenital heart defects, facial defects, and some neurological and learning disabilities. Patients with 22q11 deletions have also been reported to have higher incidence of schizophrenia and bipolar disorder. [33]

Treacher Collins Syndrome Edit

Treacher Collins Syndrome (TCS) results from the compromised development of the first and second pharyngeal arches during the early embryonic stage, which ultimately leads to mid and lower face abnormalities. TCS is caused by the missense mutation of the TCOF1 gene, which causes neural crest cells to undergo apoptosis during embryogenesis. Although mutations of the TCOF1 gene are among the best characterized in their role in TCS, mutations in POLR1C and POLR1D genes have also been linked to the pathogenesis of TCS. [34]

Neural crest cells originating from different positions along the anterior-posterior axis develop into various tissues. These regions of neural crest can be divided into four main functional domains, which include the cranial neural crest, trunk neural crest, vagal and sacral neural crest, and cardiac neural crest.

Cranial neural crest Edit

Cranial neural crest migrates dorsolaterally to form the craniofacial mesenchyme that differentiates into various cranial ganglia and craniofacial cartilages and bones. [21] These cells enter the pharyngeal pouches and arches where they contribute to the thymus, bones of the middle ear and jaw and the odontoblasts of the tooth primordia. [35]

Trunk neural crest Edit

Trunk neural crest gives rise two populations of cells. [36] One group of cells fated to become melanocytes migrates dorsolaterally into the ectoderm towards the ventral midline. A second group of cells migrates ventrolaterally through the anterior portion of each sclerotome. The cells that stay in the sclerotome form the dorsal root ganglia, whereas those that continue more ventrally form the sympathetic ganglia, adrenal medulla, and the nerves surrounding the aorta. [35]

Vagal and sacral neural crest Edit

The vagal and sacral neural crest cells develop into the ganglia of the enteric nervous system and the parasympathetic ganglia. [35]

Cardiac neural crest Edit

Cardiac neural crest develops into melanocytes, cartilage, connective tissue and neurons of some pharyngeal arches. Also, this domain gives rise to regions of the heart such as the musculo-connective tissue of the large arteries, and part of the septum, which divides the pulmonary circulation from the aorta. [35] The semilunar valves of the heart are associated with neural crest cells according to new research. [37]

Several structures that distinguish the vertebrates from other chordates are formed from the derivatives of neural crest cells. In their "New head" theory, Gans and Northcut argue that the presence of neural crest was the basis for vertebrate specific features, such as sensory ganglia and cranial skeleton. Furthermore, the appearance of these features was pivotal in vertebrate evolution because it enabled a predatory lifestyle. [38] [39]

However, considering the neural crest a vertebrate innovation does not mean that it arose de novo. Instead, new structures often arise through modification of existing developmental regulatory programs. For example, regulatory programs may be changed by the co-option of new upstream regulators or by the employment of new downstream gene targets, thus placing existing networks in a novel context. [40] [41] This idea is supported by in situ hybridization data that shows the conservation of the neural plate border specifiers in protochordates, which suggest that part of the neural crest precursor network was present in a common ancestor to the chordates. [5] In some non-vertebrate chordates such as tunicates a lineage of cells (melanocytes) has been identified, which are similar to neural crest cells in vertebrates. This implies that a rudimentary neural crest existed in a common ancestor of vertebrates and tunicates. [42]

Ectomesenchyme (also known as mesectoderm): [43] odontoblasts, dental papillae, the chondrocranium (nasal capsule, Meckel's cartilage, scleral ossicles, quadrate, articular, hyoid and columella), tracheal and laryngeal cartilage, the dermatocranium (membranous bones), dorsal fins and the turtle plastron (lower vertebrates), pericytes and smooth muscle of branchial arteries and veins, tendons of ocular and masticatory muscles, connective tissue of head and neck glands (pituitary, salivary, lachrymal, thymus, thyroid) dermis and adipose tissue of calvaria, ventral neck and face

Endocrine cells: chromaffin cells of the adrenal medulla, glomus cells type I/II.

Peripheral nervous system: Sensory neurons and glia of the dorsal root ganglia, cephalic ganglia (VII and in part, V, IX, and X), Rohon-Beard cells, some Merkel cells in the whisker, [44] [45] Satellite glial cells of all autonomic and sensory ganglia, Schwann cells of all peripheral nerves.

Melanocytes and iris muscle and pigment cells, and even associated with some tumors (such as melanotic neuroectodermal tumor of infancy).


Dissimilar metals and alloys have different electrode potentials, and when two or more come into contact in an electrolyte, one metal (that's more reactive) acts as anode and the other (that's less reactive) as cathode. The electropotential difference between the reactions at the two electrodes is the driving force for an accelerated attack on the anode metal, which dissolves into the electrolyte. This leads to the metal at the anode corroding more quickly than it otherwise would and corrosion at the cathode being inhibited. The presence of an electrolyte and an electrical conducting path between the metals is essential for galvanic corrosion to occur. The electrolyte provides a means for ion migration whereby ions move to prevent charge build-up that would otherwise stop the reaction. If the electrolyte contains only metal ions that are not easily reduced (such as Na + , Ca 2+ , K + , Mg 2+ , or Zn 2+ ), the cathode reaction is the reduction of dissolved H + to H2 or O2 to OH − . [1] [2] [3] [4]

In some cases, this type of reaction is intentionally encouraged. For example, low-cost household batteries typically contain carbon-zinc cells. As part of a closed circuit (the electron pathway), the zinc within the cell will corrode preferentially (the ion pathway) as an essential part of the battery producing electricity. Another example is the cathodic protection of buried or submerged structures as well as hot water storage tanks. In this case, sacrificial anodes work as part of a galvanic couple, promoting corrosion of the anode, while protecting the cathode metal.

In other cases, such as mixed metals in piping (for example, copper, cast iron and other cast metals), galvanic corrosion will contribute to accelerated corrosion of parts of the system. Corrosion inhibitors such as sodium nitrite or sodium molybdate can be injected into these systems to reduce the galvanic potential. However, the application of these corrosion inhibitors must be monitored closely. If the application of corrosion inhibitors increases the conductivity of the water within the system, the galvanic corrosion potential can be greatly increased.

Acidity or alkalinity (pH) is also a major consideration with regard to closed loop bimetallic circulating systems. Should the pH and corrosion inhibition doses be incorrect, galvanic corrosion will be accelerated. In most HVAC systems, the use of sacrificial anodes and cathodes is not an option, as they would need to be applied within the plumbing of the system and, over time, would corrode and release particles that could cause potential mechanical damage to circulating pumps, heat exchangers, etc. [5]

A common example of galvanic corrosion occurs in galvanized iron, a sheet of iron or steel covered with a zinc coating. Even when the protective zinc coating is broken, the underlying steel is not attacked. Instead, the zinc is corroded because it is less "noble" only after it has been consumed can rusting of the base metal occur. By contrast, with a conventional tin can, the opposite of a protective effect occurs: because the tin is more noble than the underlying steel, when the tin coating is broken, the steel beneath is immediately attacked preferentially.

Statue of Liberty Edit

A spectacular example of galvanic corrosion occurred in the Statue of Liberty when regular maintenance checks in the 1980s revealed that corrosion had taken place between the outer copper skin and the wrought iron support structure. Although the problem had been anticipated when the structure was built by Gustave Eiffel to Frédéric Bartholdi's design in the 1880s, the insulation layer of shellac between the two metals had failed over time and resulted in rusting of the iron supports. An extensive renovation was carried out requiring complete disassembly of the statue and replacement of the original insulation with PTFE. The structure was far from unsafe owing to the large number of unaffected connections, but it was regarded as a precautionary measure to preserve a national symbol of the United States. [6]

Royal Navy and HMS Alarm Edit

In the 17th century, [ vague ] Samuel Pepys (then serving as Admiralty Secretary) agreed to the removal of lead sheathing from English Royal Navy vessels to prevent the mysterious disintegration of their rudder-irons and bolt-heads, though he confessed himself baffled as to the reason the lead caused the corrosion. [7]

The problem recurred when vessels were sheathed in copper to reduce marine weed accumulation and protect against shipworm. In an experiment, the Royal Navy in 1761 had tried fitting the hull of the frigate HMS Alarm with 12-ounce copper plating. Upon her return from a voyage to the West Indies, it was found that although the copper remained in fine condition and had indeed deterred shipworm, it had also become detached from the wooden hull in many places because the iron nails used during its installation "were found dissolved into a kind of rusty Paste". [8] To the surprise of the inspection teams, however, some of the iron nails were virtually undamaged. Closer inspection revealed that water-resistant brown paper trapped under the nail head had inadvertently protected some of the nails: "Where this covering was perfect, the Iron was preserved from Injury". The copper sheathing had been delivered to the dockyard wrapped in the paper which was not always removed before the sheets were nailed to the hull. The conclusion therefore reported to the Admiralty in 1763 was that iron should not be allowed direct contact with copper in sea water. [9] [10]

US Navy Littoral Combat Ship Independence Edit

Serious galvanic corrosion has been reported on the latest US Navy attack littoral combat vessel the USS Independence caused by steel water jet propulsion systems attached to an aluminium hull. Without electrical isolation between the steel and aluminium, the aluminium hull acts as an anode to the stainless steel, resulting in aggressive galvanic corrosion. [11]

Corroding lighting fixtures Edit

The unexpected fall in 2011 of a heavy light fixture from the ceiling of the Big Dig vehicular tunnel in Boston revealed that corrosion had weakened its support. Improper use of aluminium in contact with stainless steel had caused rapid corrosion in the presence of salt water. [12] The electrochemical potential difference between stainless steel and aluminium is in the range of 0.5 to 1.0 V, depending on the exact alloys involved, and can cause considerable corrosion within months under unfavorable conditions. Thousands of failing lights would have to be replaced, at an estimated cost of $54 million. [13]

Lasagna cell Edit

A "lasagna cell" is accidentally produced when salty moist food such as lasagna is stored in a steel baking pan and is covered with aluminium foil. After a few hours the foil develops small holes where it touches the lasagna, and the food surface becomes covered with small spots composed of corroded aluminium. [14] In this example, the salty food (lasagna) is the electrolyte, the aluminium foil is the anode, and the steel pan is the cathode. If the aluminium foil touches the electrolyte only in small areas, the galvanic corrosion is concentrated, and corrosion can occur fairly rapidly. If the aluminium foil was not used with a dissimilar metal container, the reaction was probably a chemical one. It is possible for heavy concentrations of salt, vinegar or some other acidic compounds to cause the foil to disintegrate. The product of either of these reactions is an aluminium salt. It does not harm the food, but any deposit may impart an undesired flavor and color. [15]

Electrolytic cleaning Edit

The common technique of cleaning silverware by immersion of the silver or sterling silver (or even just silver plated objects) and a piece of aluminium (foil is preferred because of its much greater surface area than that of ingots, although if the foil has a "non-stick" face, this must be removed with steel wool first) in a hot electrolytic bath (usually composed of water and sodium bicarbonate, i.e., household baking soda) is an example of galvanic corrosion. Silver darkens and corrodes in the presence of airborne sulfur molecules, and the copper in sterling silver corrodes under a variety of conditions. These layers of corrosion can be largely removed through the electrochemical reduction of silver sulfide molecules: the presence of aluminium (which is less noble than either silver or copper) in the bath of sodium bicarbonate strips the sulfur atoms off the silver sulfide and transfers them onto and thereby corrodes the piece of aluminium foil (a much more reactive metal), leaving elemental silver behind. No silver is lost in the process. [16]

There are several ways of reducing and preventing this form of corrosion.

  • Electrically insulate the two metals from each other. If they are not in electrical contact, no galvanic coupling will occur. This can be achieved by using non-conductive materials between metals of different electropotential. Piping can be isolated with a spool of pipe made of plastic materials, or made of metal material internally coated or lined. It is important that the spool be a sufficient length to be effective. For reasons of safety, this should not be attempted where an electrical earthing system uses the pipework for its ground or has equipotential bonding.
  • Metal boats connected to a shore line electrical power feed will normally have to have the hull connected to earth for safety reasons. However the end of that earth connection is likely to be a copper rod buried within the marina, resulting in a steel-copper "battery" of about 0.5 V. For such cases, the use of a galvanic isolator is essential, typically two semiconductor diodes in series, in parallel with two diodes conducting in the opposite direction (antiparallel). This prevents any current while the applied voltage is less than 1.4 V (i.e. 0.7 V per diode), but allows a full current in case of an electrical fault. There will still be a very minor leakage of current through the diodes, which may result in slightly faster corrosion than normal.
  • Ensure there is no contact with an electrolyte. This can be done by using water-repellent compounds such as greases, or by coating the metals with an impermeable protective layer, such as a suitable paint, varnish, or plastic. If it is not possible to coat both, the coating should be applied to the more noble, the material with higher potential. This is advisable because if the coating is applied only on the more active material, in case of damage to the coating there will be a large cathode area and a very small anode area, and for the exposed anodic area the corrosion rate will be correspondingly high.
  • Using antioxidant paste is beneficial for preventing corrosion between copper and aluminium electrical connections. The paste consists of a lower nobility metal than aluminium or copper.
  • Choose metals that have similar electropotentials. The more closely matched the individual potentials, the smaller the potential difference and hence the smaller the galvanic current. Using the same metal for all construction is the easiest way of matching potentials. or other plating can also help. This tends to use more noble metals that resist corrosion better. Chrome, nickel, silver and gold can all be used. Galvanizing with zinc protects the steel base metal by sacrificial anodic action. uses one or more sacrificial anodes made of a metal which is more active than the protected metal. Alloys of metals commonly used for sacrificial anodes include zinc, magnesium, and aluminium. This approach is commonplace in water heaters and many buried or immersed metallic structures.
  • Cathodic protection can also be applied by connecting a direct current (DC) electrical power supply to oppose the corrosive galvanic current. (See Cathodic protection § Impressed current CP.)

All metals can be classified into a galvanic series representing the electrical potential they develop in a given electrolyte against a standard reference electrode. The relative position of two metals on such a series gives a good indication of which metal is more likely to corrode more quickly. However, other factors such as water aeration and flow rate can influence the rate of the process markedly.

The compatibility of two different metals may be predicted by consideration of their anodic index. This parameter is a measure of the electrochemical voltage that will be developed between the metal and gold. To find the relative voltage of a pair of metals it is only required to subtract their anodic indices. [17]

To reduce galvanic corrosion for metals stored in normal environments such as storage in warehouses or non-temperature and humidity controlled environments, there should not be more than 0.25 V difference in the anodic index of the two metals in contact. For controlled environments in which temperature and humidity are controlled, 0.50 V can be tolerated. For harsh environments such as outdoors, high humidity, and salty environments, there should be not more than 0.15 V difference in the anodic index. For example: gold and silver have a difference of 0.15 V, therefore the two metals will not experience significant corrosion even in a harsh environment. [18] [ page needed ]

When design considerations require that dissimilar metals come in contact, the difference in anodic index is often managed by finishes and plating. The finishing and plating selected allow the dissimilar materials to be in contact, while protecting the more base materials from corrosion by the more noble. [18] [ page needed ] It will always be the metal with the most negative anodic index which will ultimately suffer from corrosion when galvanic incompatibility is in play. This is why sterling silver and stainless steel tableware should never be placed together in a dishwasher at the same time, as the steel items will likely experience corrosion by the end of the cycle (soap and water having served as the chemical electrolyte, and heat having accelerated the process).


There has been much interest in determining the mechanisms that govern the increase of thickness of the oxide layer over time. Some of the important factors are the volume of oxide relative to the volume of the parent metal, the mechanism of oxygen diffusion through the metal oxide to the parent metal, and the relative chemical potential of the oxide. Boundaries between micro grains, if the oxide layer is crystalline, form an important pathway for oxygen to reach the unoxidized metal below. For this reason, vitreous oxide coatings – which lack grain boundaries – can retard oxidation. [9] The conditions necessary, but not sufficient for passivation are recorded in Pourbaix diagrams. Some corrosion inhibitors help the formation of a passivation layer on the surface of the metals to which they are applied. Some compounds, dissolved in solutions (chromates, molybdates) form non-reactive and low solubility films on metal surfaces.

Discovery Edit

In the mid 1800s, Christian Friedrich Schönbein discovered that when a piece of iron is placed in dilute nitric acid, it will dissolve and produce hydrogen, but if the iron is placed in concentrated nitric acid and then returned to the dilute nitric acid, little or no reaction will take place. Schönbein named the first state the active condition and the second the passive condition. If passive iron is touched by active iron, it becomes active again. In 1920, Ralph S. Lillie measured the effect of an active piece of iron touching a passive iron wire and found that "a wave of activation sweeps rapidly (at some hundred centimeters a second) over its whole length". [10] [11]

Surface passivation Edit

The surface passivation process, also known as the Atalla passivation technique, [12] was developed by Mohamed M. Atalla at Bell Telephone Laboratories (BTL) in the late 1950s. [6] [13] In 1955, Carl Frosch and Lincoln Derick at Bell Telephone Laboratories (BTL) accidentally discovered that silicon dioxide (SiO2) could be grown on silicon. They showed that oxide layer prevented certain dopants into the silicon wafer, while allowing for others, thus discovering the passivating effect of oxidation on the semiconductor surface. [14] In the late 1950s, Atalla further discovered that the formation of a thermally grown SiO2 layer greatly reduced the concentration of electronic states at the silicon surface, [13] and discovered the important quality of SiO2 films to preserve the electrical characteristics of p–n junctions and prevent these electrical characteristics from deteriorating by the gaseous ambient environment. [15] He found that silicon oxide layers could be used to electrically stabilize silicon surfaces. [16] J.R. Ligenza and W.G. Spitzer, who studied the mechanism of thermally grown oxides, managed to fabricate a high quality Si/SiO2 stack, with Atalla and Kahng making use of their findings. [17] [18] [19] Atalla developed the surface passivation process, a new method of semiconductor device fabrication that involves coating a silicon wafer with an insulating layer of silicon oxide so that electricity could reliably penetrate to the conducting silicon below. By growing a layer of silicon dioxide on top of a silicon wafer, Atalla was able to overcome the surface states that prevented electricity from reaching the semiconducting layer. [6] [7] For the surface passivation process, he developed the method of thermal oxidation, which was a breakthrough in silicon semiconductor technology. [20]

Before the development of integrated circuit chips, discrete diodes and transistors exhibited relatively high reverse-bias junction leakages and low breakdown voltage, caused by the large density of traps at the surface of single crystal silicon. Atalla's surface passivation process became the solution to this problem. He discovered that when a thin layer of silicon dioxide was grown on the surface of silicon where a p–n junction intercepts the surface, the leakage current of the junction was reduced by a factor from 10 to 100. This showed that the oxide reduces and stabilizes many of the interface and oxide traps. Oxide-passivation of silicon surfaces allowed diodes and transistors to be fabricated with significantly improved device characteristics, while the leakage path along the surface of the silicon was also effectively shut off. This became one of the fundamental isolation capabilities necessary for planar technology and integrated circuit chips. [21]

Atalla first published his findings in BTL memos during 1957, before presenting his work at an Electrochemical Society meeting in 1958. [22] [23] The same year, he made further refinements to the process with his colleagues E. Tannenbaum and E.J. Scheibner, before they published their results in May 1959. [24] [25] According to Fairchild Semiconductor engineer Chih-Tang Sah, the surface passivation process developed by Atalla's team "blazed the trail" that led to the development of the silicon integrated circuit. [21] [24] Atalla's surface passivation method was the basis for several important inventions in 1959: the MOSFET (MOS transistor) by Atalla and Dawon Kahng at Bell Labs, the planar process by Jean Hoerni at Fairchild Semiconductor, and the monolithic integrated circuit chip by Robert Noyce at Fairchild in 1959. [22] [23] [21] [24] By the mid-1960s, Atalla's process for oxidized silicon surfaces was used to fabricate virtually all integrated circuits and silicon devices. [26]

In solar cell technology, surface passivation is critical to solar cell efficiency. [27] In carbon quantum dot (CQD) technology, CQDs are small carbon nanoparticles (less than 10 nm in size) with some form of surface passivation. [28] [29] [30]

Aluminium Edit

Aluminium naturally forms a thin surface layer of aluminium oxide on contact with oxygen in the atmosphere through a process called oxidation, which creates a physical barrier to corrosion or further oxidation in many environments. Some aluminium alloys, however, do not form the oxide layer well, and thus are not protected against corrosion. There are methods to enhance the formation of the oxide layer for certain alloys. For example, prior to storing hydrogen peroxide in an aluminium container, the container can be passivated by rinsing it with a dilute solution of nitric acid and peroxide alternating with deionized water. The nitric acid and peroxide mixture oxidizes and dissolves any impurities on the inner surface of the container, and the deionized water rinses away the acid and oxidized impurities. [31]

Generally, there are two main ways to passivate aluminium alloys (not counting plating, painting, and other barrier coatings): chromate conversion coating and anodizing. Alclading, which metallurgically bonds thin layers of pure aluminium or alloy to different base aluminium alloy, is not strictly passivation of the base alloy. However, the aluminium layer clad on is designed to spontaneously develop the oxide layer and thus protect the base alloy.

Chromate conversion coating converts the surface aluminium to an aluminium chromate coating in the range of 0.00001–0.00004 inches (250–1,000 nm) in thickness. Aluminium chromate conversion coatings are amorphous in structure with a gel-like composition hydrated with water. [32] Chromate conversion is a common way of passivating not only aluminium, but also zinc, cadmium, copper, silver, magnesium, and tin alloys.

Anodizing is an electrolytic process that forms a thicker oxide layer. The anodic coating consists of hydrated aluminium oxide and is considered resistant to corrosion and abrasion. [33] This finish is more robust than the other processes and also provides electrical insulation, which the other two processes may not.

Ferrous materials Edit

Ferrous materials, including steel, may be somewhat protected by promoting oxidation ("rust") and then converting the oxidation to a metalophosphate by using phosphoric acid and further protected by surface coating. As the uncoated surface is water-soluble, a preferred method is to form manganese or zinc compounds by a process commonly known as parkerizing or phosphate conversion. Older, less-effective but chemically-similar electrochemical conversion coatings included black oxidizing, historically known as bluing or browning. Ordinary steel forms a passivating layer in alkali environments, as reinforcing bar does in concrete.

Stainless steel Edit

Stainless steels are corrosion-resistant, but they are not completely impervious to rusting. One common mode of corrosion in corrosion-resistant steels is when small spots on the surface begin to rust because grain boundaries or embedded bits of foreign matter (such as grinding swarf) allow water molecules to oxidize some of the iron in those spots despite the alloying chromium. This is called rouging. Some grades of stainless steel are especially resistant to rouging parts made from them may therefore forgo any passivation step, depending on engineering decisions. [34]

Common among all of the different specifications and types are the following steps: Prior to passivation, the object must be cleaned of any contaminants and generally must undergo a validating test to prove that the surface is 'clean.' The object is then placed in an acidic passivating bath that meets the temperature and chemical requirements of the method and type specified between customer and vendor. While nitric acid is commonly used as a passivating acid for stainless steel, citric acid is gaining in popularity as it is far less dangerous to handle, less toxic, and biodegradable, making disposal less of a challenge. [35] Passivating temperatures can range from ambient to 60 degrees C, or 140 degrees F, while minimum passivation times are usually 20 to 30 minutes. After passivation, the parts are neutralized using a bath of aqueous sodium hydroxide, then rinsed with clean water and dried. The passive surface is validated using humidity, elevated temperature, a rusting agent (salt spray), or some combination of the three.. [36] The passivation process removes exogenous iron, [37] creates/restores a passive oxide layer that prevents further oxidation (rust), and cleans the parts of dirt, scale, or other welding-generated compounds (e.g. oxides). [37] [38]

Passivation processes are generally controlled by industry standards, the most prevalent among them today being ASTM A 967 and AMS 2700. These industry standards generally list several passivation processes that can be used, with the choice of specific method left to the customer and vendor. The "method" is either a nitric acid-based passivating bath, or a citric acid-based bath, these acids remove surface iron and rust, while sparing the chromium. The various 'types' listed under each method refer to differences in acid bath temperature and concentration. Sodium dichromate is often required as an additive to oxidise the chromium in certain 'types' of nitric-based acid baths, however this chemical is highly toxic. With citric acid, simply rinsing and drying the part and allowing the air to oxidise it, or in some cases the application of other chemicals, is used to perform the passivation of the surface.

It is not uncommon for some aerospace manufacturers to have additional guidelines and regulations when passivating their products that exceed the national standard. Often, these requirements will be cascaded down using Nadcap or some other accreditation system. Various testing methods are available to determine the passivation (or passive state) of stainless steel. The most common methods for validating the passivity of a part is some combination of high humidity and heat for a period of time, intended to induce rusting. Electro-chemical testers can also be utilized to commercially verify passivation.

Nickel Edit

Nickel can be used for handling elemental fluorine, owing to the formation of a passivation layer of nickel fluoride. This fact is useful in water treatment and sewage treatment applications.

Silicon Edit

In the area of microelectronics and photovoltaics surface passivation is usually implemented by oxidation to a coating of silicon dioxide. The effect of passivation on the efficiency of solar cells ranges from 3-7%. Passivation is effected by thermal oxidation at 1000 °C. The surface resistivity is high, >100 Ωcm. [39]


Cell culture

K562, HeLa, and HEK293 cells were used for transient transfection to examine the enhancer activities in the DLR assay. K562 cells were cultured in RPMI1640 Medium (Gibco) with 10% fetal bovine serum (Hyclone) and penicillin (100 U/ml)-streptomycin (0.1 mg/ml) (Invitrogen), and HeLa and HEK293 cells were cultured in Dulbecco’s Modified Eagle Medium (Gibco) with 10% fetal bovine serum (Hyclone) and penicillin (100 U/ml)-streptomycin (0.1 mg/ml) (Invitrogen). All cells were maintained at 37 °C with 5% CO2 in a humidified incubator.

Transcriptome sequencing and gene expression analysis

mRNA-seq was originally designed to explore the dynamic transcriptomes during human erythroid differentiation and development (Yang Y, Wang H, Chang KH, Qu H, Zhang Z, Xiong Q, Qi H, Cui P, Lin Q, Ruan X, et al: Transcriptome dynamics during human erythroid differentiation and development, submitted). In Brief, we extracted total RNA from HESC, ESER, FLER, and PBER, and depleted 18S and 28S ribosomal RNAs before constructing cDNA libraries. Next, we used the ABI SOLiD System to perform massively parallel ligation sequencing and mapped the sequence reads to human reference sequence [release Mar. 2006 (NCBI36/hg18)]. Gene expression intensity was calculated by normalizing the read counts to RPKM according to the gene length and total mapped reads, and genes with RPKM < 0.01 were removed.

Quantitative real-time PCR

Total RNA was extracted from HESC, ESER, FLER, and PBER cells using TRIZOL® Reagent (Invitrogen, 15596–018) and DNA contamination was removed using the TURBO DNA-free™ Kit (Ambion, AM1907). DNA-free RNA was reverse transcribed using the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, K1622) according to manufacturer’s instruction. Primers were designed using Primer 5 (Additional file 2: Table S4). PCR were performed in triplicate using Maxima® SYBR Green/ROX qPCR Master Mixes (2×) (Fermentas, K0223) and CFX96™ Real-Time PCR Detection System (Bio-rad), and data were analyzed using the CFX Manager™ Software. The KLF transcript levels were first calculated by referring to those of 18S ribosomal RNA, and the expression levels of KLFs in ESER, FLER, and PBER were normalized to those in HESC. The statistical significance of differences between individual KLFs’ expressions in erythroid cell types and those in HESC were calculated using the independent-samples t-test.

Digital DNase I sequencing and erythroid-specific or putative erythroid-specific DHS selection

In this study, the DNase-seq data used were obtained from the University of Washington [41, 60], and are available through the UCSC Genome Browser ( and the NCBI Gene Expression Omnibus (GEO) data repository under accessions GSE29692 and GSE32970. DHSs were identified using an algorithm developed by the University of Washington [65]. In this study, FDR threshold of 0.5% was used to define DHS for each cell type. The domains of KLF loci were defined as extensions from 70 kb upstream of TSSs to 20 kb downstream of the poly (A) sites. ESER, FLER, and PBER cells represent primary erythroid cells at different developmental stages, and K562 cells represent erythroleukemia cells. In DHS screening, all other cell types were employed as non-erythroid control cell types. DHSs in these domains were considered to be erythroid specific if they were only present in erythroid cells and were identified as putative erythroid specific if they were present in erythroid cells and exhibited much lower peaks in one or two non-erythroid cell types.

DNA manipulation

To generate firefly luciferase reporter constructs with minP, the identified 23 erythroid-specific or putative erythroid-specific DHSs were amplified from human blood genomic DNA with Pfu DNA Polymerase (Promega, M7741) and inserted upstream of minP in the pGL4.23 expression vector (Promega, E8411). To further generate firefly luciferase reporter constructs with KLF-Ps, TSSs of individual KLFs were predicted from UCSC Genome Browser, and fragments of approximately 1 kb in length upstream of TSSs (Additional file 2: Table S2) were amplified and cloned into pGL4.10 vector, followed by inserting DHSs upstream of the corresponding KLF-Ps. The activities of KLF-Ps were examined and used as baselines. HS2, a classical enhancer in β-globin LCR, was cloned into the corresponding vectors upstream of minP or KLF-Ps and used as positive controls in the DLR assay [16]. The primers used in this study are listed in Additional file 2: Tables S5 and S6. The integrity of the reporter constructs was confirmed using restriction digestions and sequencing.

Transient transfection and DLR assay

Cells were seeded into 48-well plates. K562 cells (1.5 × 10 5 /well) were transiently transfected with 500 ng of firefly luciferase vector and 0.75 ng of a Renilla luciferase vector, pRL-TK (Promega, E2441), using Lipofectamine LTX and Plus Reagent (Invitrogen, 15338–100). HeLa (7 × 10 4 /well) and HEK293 (7 × 10 4 /well) cells were similarly transfected using the Lipofectamine 2000 Transfection Reagent (Invitrogen, 11668–019) as per the manufacturer’s instructions. Forty-eight hours after transfection, cells were harvested to prepare for the cell lysates, and luciferase activities were immediately measured with the Dual-Luciferase Reporter Assay System (Promega, E1910) as per the manufacturer’s instructions. Transient transfections were repeated at least twice, and every construct was transfected in triplicates. Standard deviations were shown as error bars above respective columns. For data processing, firefly luciferase activity was normalized to that of Renilla luciferase in all the groups, and the relative activity of each promoter was normalized as 1. The statistical significance of differences between promoters and DHSs were analyzed using one-way ANOVA function in R language.

Bioinformatic analyses

Data of conserved elements in placental mammals [51], layered H3K4me1 and layered H3K27ac [52], and Txn Factor ChIP data [53] in the regions of 23 DHSs were obtained from UCSC Genome Browser and summarized in Table 1.

TFs binding to the 23 DHSs were collected from Txn Factor ChIP track on UCSC Genome Browser. A heat map was drawn after K-means clustering using R language.

Conserved motifs embedded in erythroid-specific KLF enhancers were analyzed using the MEME software (, and these de novo discovered motifs were searched against the ENCODE-motifs database (

Bio-11-dUTP (Ambion, AM8450) and TdT (New England Biolabs, M0315s) were used to label the 3′-OH of single-stranded oligos (5′-AGC ATG AAG TAG GAG AGT GAT GAT GAC AGT GCT GCT TTG CAC AGA TAA GCC TGG CGG A-3′, 5′-TCC GCC AGG CTT ATC TGT GCA AAG CAG CAC TGT CAT CAT CAC TCT CCT ACT TCA TGC T-3′) and complementary oligos were annealed as per the manufacturer’s instructions. Nuclear proteins of K562, HeLa, and HEK293 cells were extracted using the rapid micro-preparation method (lysis buffer: 10 mM Hepes [pH7.9], 10 mM KCl, 1.5 mM MgCl2, 0.5 mM PMSF, 0.5 mM DTT high-salt extraction buffer: 20 mM Hepes [pH7.9], 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM PMSF, 0.5 mM DTT) [66]. Protein concentrations were measured using the BCA Protein Assay Kit (Pierce, 23225). EMSA was performed in 20 μl reaction mixture, containing 30 fmol biotin-labeled oligos and 7 μg nuclear extract with or without 7.2 pmol unlabeled oligos using the LightShift Chemiluminescent EMSA Kit (Pierce, 20148) according to the manufacturer’s instruction (Figure 7B).

The Evolution of the Superbugs

In August 2016, a woman in her late 20s arrived in a hospital in Reno, Nevada, carrying a particularly deadly strain of the bacteria Klebsiella pneumoniae. The woman, who recently had traveled to India, and while there was admitted to several hospitals over complications caused by a thighbone fracture, proved to be resistant to all 26 types of antibiotics by the time she was hospitalized in the United States—including carbapenem, often considered the “last resort” antibiotic, because it is used when all other antibiotics fail.

What is a Superbug?

The strain responsible for the woman’s infection,Klebsiella pneumoniae carbapenemase, or KPC, is one of several types of bacteria that have become known as “superbugs”. At least 18 of these diseases and infections have been labeled by the CDC, and three in particular have been classified as “urgent” drug-resistant threats: Neisseria gonorrhoeae, which causes the STD gonorrhea, Clostridium difficil (CDIFF), which causes life-threatening diarrhea, and a class of bacteria known as Carbapenem-resistant Enterobacteriaceae (CREs), which includes KPC.

Of these, CREs are the most worrisome for scientists and doctors. Though no less deadly than other types of drug-resistant diseases, CRE drug resistance is found within genes that lie in plasmids, pieces of DNA that can easily move between bacterial species of the same family. Enterobacteriaceae (The “E” in CREs), a particularly large group of related bacteria that include species such as Salmonella, E. coli, and Shigella, are particularly susceptible to the spread of plasmid-based genes between bacterium. Although all of these diseases are easily treatable with antibiotics, they can quickly prove to be deadly without proper medication.

CREs are particularly prevalent in China and South Asia, but are becoming increasingly more common in the United States as well: the CDC estimates that this year alone, at least 175 cases have been reported in hospitals. Luckily, most CREs can still be cured with a few antibiotics that are used only in extreme cases, but even these may be effective for only so long.

Colistin and the MCR1 Gene

In China, a new variant of a gene, called MCR-1, has been found in cases of Klebsiella pneumoniae and E. coli. Like CREs, Mcr-1 easily moves from one bacterium to another via plasmids, but unlike CREs, this variant of MCR-1 allows bacteria to be resistant to the absolute last resort antibiotic, colistin. The reason is not due to an overuse of the drug in treating illnesses, but because because China has historically used colistin as an animal growth hormone in livestock feed. As humans consume colistin-fed beef, pork, or chicken, the bacteria in our bodies in turn become used to the presence of the antibiotic, and over time adapt to be able to resist it.

While China has since made it illegal to use colistin in animal feed as of April of this year, it has also since approved the use of the drug in hospitals in order to tackle a growing number of CRE cases, themselves likely caused by the overuse of carbapenems in agriculture. And as colistin is used more and more, so too will it become more likely for bacteria to evolve a resistance for it in humans, paving the way for a string of superbugs incurable by modern science.

The Threat of SuperBug Evolution

It’s a textbook case of organisms adapting to respond to their environments that has been ignored for decades, with people concerned only with the short-term benefits of overusing antibiotics without looking at the long-term consequences. It’s a problem that Alexander Fleming, the creator of penicillin, warned of it his Nobel Prize acceptance speech, saying, “There is the danger that the ignorant man may easily underdose himself, and by exposing his microbes to non-lethal quantities of the drug make them resistant.”

Without the creation of new forms of antibiotics that can be used to fight future CREs and colistin-resistant superbugs, scientists predict that antibiotic-resistant diseases could kill up to 10 million people a year, mostly from common infections that today are easily treated. The unfortunate reality is that scientists had warned of this scenario since the creation of the first antibiotics, and by not heeding their advice, we have set ourselves up for a future of superbugs that may not be able to be cured by modern science.

For More Information:

  • Klebsiella pneumoniae: By NIAID (Klebsiella pneumoniae Bacterium) [CC BY 2.0 (], via Wikimedia Commons
  • CRE and antibiotic resistance graphics – courtesy of the CDC.

Article by Devin Windelspecht. Devin is a junior at Northeastern University in Boston MA where he majors in international relations. Devin is responsible for background work on many of the articles on the site, as well as some science writing.

Laboratory results are reported in LIMS and reports are available to print as they are completed. Expected turnaround time is:

Grp A Beta Strep/Throat one business day
Respiratory (Nasopharyngeal, Sputum) two business days
Wound & Miscellaneous two business days
Fecal two - three business days
Urine two business days
Genital Tract two business days

Unveiling the gene regulatory landscape in diseases through the identification of DNase I‑hypersensitive sites (Review)

Copyright: © Chen et al. This is an open access article distributed under the terms of Creative Commons Attribution License.

This article is mentioned in:


1. Introduction

Definition of DNase I-hypersensitive sites (DHSs)

DNase I is an endonuclease with little DNA-sequence specificity (1). In the early 1960s, DNase I was used to probe how the nucleosome was organized (2). Weintraub and Groudine (3) found that active chromatin was prioritized for decomposition by this enzyme. These specific regions are termed DHSs, which are distinct markers of active chromatin that co-position with various cis -regulatory elements (CREs), including the expression regulation sequences of enhancers and promoters, negative regulation sequence of insulators, silencers, and certain locus control regions. DHSs usually disperse around transcriptionally active genes and are accessible to regulatory proteins. Therefore, by contrast, certain regions within DHSs are resistant to degradation as they are protected by these gene regulatory proteins, including transcription factors.

Generation and erasure of DHSs

The underlying molecular mechanisms of gene regulation remain to be fully elucidated. The first step of gene regulation is that the cells adopt an ‘open’ structure in response to external stimulation at a group of specific sequences located in certain chromatin regions. These exposed sequences are bound by site-specific transcriptional regulatory elements, leading to chromatin structure remodeling marked by significant accessibility to the nucleases (4). DHSs in the open chromatin serve important roles in chromatin rearrangement (5,6), and are stimulated by the different state of histone acetylation (7,8) and the binding capacity of the chromatin remodeling multiprotein complex (9). DHSs act as binding anchors of activator proteins and mediating cofactors, and interact with the preinitiation complex at promoters (2,10-14), regulating gene expression.

DHSs can be eliminated by site-specific factors. It was reported that DHS at a CCCTC-binding factor (CTCF)-dependent silencer can be eliminated due to the eviction of CTCF and remodeling of a nucleosome caused by inducible non-coding RNA transcription in the chicken lysozyme (15). DHSs are dynamic and fine-tuned by different classes of remodeling enzymes. For example, TFE3 combines with ACF to stimulate the occurrence of an active DHS site in the IgH intronic enhancer, whereas PU.1 has been demonstrated to recruit Mi2β and subsequently erase this DHS (16).

Characteristic features of DHSs

In mammalian cells, >3% of the genome is found to be DNase I-hypersensitive (17). To date, 290,000,000 DHSs have been recognized. Each tissue/cell type is represented by multiple distinguished DHS profiling derived from different individuals. DHSs are considered to be the one of the most useful discriminative features between cell types (18) and have several distinctive characters.

First, DHSs are typically characterized by high sensitivity to DNase I, particularly when the related gene is actively transcribed. Regions with high transcriptional activity are reported to be even more sensitive than those with no transcriptional activity. That is, there exists two states of DHSs, open and closed, in which the accessibility features of chromatin is increased or decreased, respectively, associated with gene expression (19).

Second, DHSs are short sequences of

200 base pairs with low methylation, and the majority are no more than several hundred base pairs long. These low-methylation regions are co-positioned at or close to the transcription starting sites (20,21), which affect gene transcription according to the degree of methylation.

Third, DHSs are representative markers of regulatory DNA and overlap with multiple CREs, including promoters, enhancers and active transcription sites (22). In addition, DHSs have underpinned the identification of other CREs, including insulators, silencers and locus control regions (10).

Fourth, although each tissue/cell exhibits distinguished DHS signatures, there exists specific core regions in DHSs that can be identified by sequence-specific DNA-binding proteins. The core regions are conserved in different cell types across species and are enriched with binding sites of HMG14 and HMG17 proteins (23).

2. High-throughput sequencing methods for DHS identification

The techniques used for DHS identification do not vary substantially, all of which are novel techniques based on high-throughput sequencing. The differences are compared in Table I.

Table I

Comparison of different techniques to identify DHSs.

Table I

Comparison of different techniques to identify DHSs.

[i] Accurate DHSs, DNase I-hypersensitive sites DNase-seq, DNase-sequencing scDNase-seq, single-cell DNase-seq liDNase-seq, low-input DNase-seq.

DNase-sequencing (DNase-seq)

Decades ago, Southern blot hybridization was the major method used to identify DHSs by characterizing digested DNAs following the titration of DNase I (24). However, the low-throughput nature of this strategy profoundly restricted its further application. Improvements and the wide application of the massively parallel sequencing technique has allowed high-resolution genome-scale mapping of various DHSs, which lays a foundation for assembling comprehensive catalogs of regulatory sequences (25,26). The first method used for identifying thousands of DHSs simultaneously was introduced by Crawford et al (27). Firstly, the nuclei are cleaved by DNase I, and the two ends are then digested blunt using T4 DNA polymerase. The genome is then cleaved by adding Bam HI and Bgl II. The digested blunt or sticky fragments are ligated into the pBluescript SK(+) plasmid and sequenced. This method enriches the sequence within the genome that relates to active chromatin and identifies DHSs on a genome-wide scale. Two years later, these high-throughput strategies were renewed by attaching a biotinylated linker to the DNase-digested ends (Fig. 1). The linker tags are used to extract short joint DNA sequences, which can be identified by DNase-based high-throughput next-generation sequencing (DNase-seq) (28) or DNase-based microarrays (DNase-chip) (26). Similar strategies providing accurate mapping of DHSs further assist in revealing a large category of CREs in all types of mammalian tissues and cells (29,30).

Figure 1.

Experimental procedures of the high-throughput DHSs identification protocol. Intact nuclei are digested with DNase I and then made blunt, followed by ligation of a biotinylated linker and sonication for shearing. The products are incubated on a streptavidin column for pooling of the DNase I-cleaved ends. The extracted short adjacent DNA fragments are either hybridized to tiled arrays (DNase-chip) or subjected to library construction and next generation sequencing (DNase-sequencing). DHSs, DNase I-hypersensitive sites.

Morin et al (31) simulated the whole exon sequencing paradigm, and developed a customized capture panel for known DHSs (‘immune sequences’), specific for DHS detection in immune cells and genetic variation in immune-related diseases.

Single-cell DNase-seq (scDNase-seq)

Despite the robustness of DNase-seq technology, millions of cells are required, which limits its application in rare cases with limited cells, such as in certain cells from patients. In addition, traditional DNase-seq suffers from low sensitivity as a result of DNA loss during the multiple purification steps. Therefore, scDNase-seq, also known as Pico-seq, was developed to minimize DNA loss, and has been applied in the analysis of chromatin accessibility using single cells (32,33). To prevent loss of the small quantity of DNase I-hypersensitive DNA released by DNase I digestion of single cells, a large amount of circular plasmid DNA is added as carrier DNA in the subsequent steps of library preparation (Fig. 2). The previous application of scDNase-seq to tumor cells, NIH3T3 cells and pools of normal cells has shown that DHS patterns at the single-cell level are highly reproducible among individual cells (33). This method enables the generation of a genome-wide DHS map for rare samples, which is more valuable for clinical application.

Figure 2.

Schematic of single-cell DNase sequencing. Intact nuclei are digested by DNase I and followed by library construction including end-repair, ligation of the adaptors, PCR amplification with circular carrier DNA and high-throughput sequencing. DHSs, DNase I-hypersensitive sites.

Low-input DNase-seq (liDNase-seq)

Although scDNase-seq can identify DHSs from a small quantity of starting material, the annotation of sequencing results requires a pre-known DHS database, which is a challenge in the identification of de novo DHSs sites. Lu et al (34) introduced liDNase-seq by modifying the scDNase-seq method to achieve de novo genome-wide DHS identification using no more than 30 cells. The major technical improvements, including reducing the complexity of the purification process prior to the adaptor ligation reaction and the first amplification reaction, and modifying the size selection step by using SPRI affinity beads in place of gel purification. The process for generating DHS maps is similar to that of the ENCODE project ( and and allows the identification of de novo DHSs at much higher resolution than in previous methods.

ImmunoSEQ technique

Morin et al (31) developed an ImmunoSEQ technique for the detection of known DHSs. Using this technique, whole-exome sequencing or customized DHS region sequencing can be efficiently performed. The analysis focuses on the variation of non-coding regions of immune-related diseases.

3. Functions of DHSs

DHSs are involved in gene expression regulation

DHSs are essential features of all defined types of active CREs and are often co-positioned with them (35-37). DHSs are directly involved in chromatin modeling and structural reestablishment, the recognition of regulatory proteins and regulating the initiation of transcription. DHSs are associated with nearby gene expression changes through the binding of certain regulatory proteins to their specific sequence at promoters or other CREs regions, and are thus involved in cell fate decisions, individual variation and development (22). Frank et al (19) detected thousands of CREs at which the accessibility of chromatin increased or decreased. These changes coincided with the transcription level of adjacent genes, which is important in the regulation of global gene expression, and most likely infers activation or deactivation of enhancer elements. Huang and Liew (38) identified DHSs in the 4-kb upstream locus of the cardiac myosin heavy chain-α (MHC-α) gene in the hamster and revealed a conserved GATA-motif site that interacts specifically with GATA-binding factors at different stages of cardiomyocyte development, which provided evidence for the role of GATA factors in the gene expression of cardiac MHC-α.

Open DHSs often mark increases in local transcription levels, which supports the observation that open DHSs are enhancers. Similarly, closed DHSs may represent reduced enhancer activity (19). This influence is more apparent when genes are associated with two or more directional matched DHS changes. The findings identified in genome-wide association studies (GWASs) show that genetic variations frequently lie in non-coding regions of the genome that contain CREs, which suggests that gene expression change underlies the development of several complex traits.

DHSs exhibit distinguished profiles and contribute to define CREs

Cells in a specific stage and stature possess a fixed set of CREs that are accessible to trans-acting factors, and thus underlie a complex controlling network of chromatin (35,39). Each cell type has a specific set of regulatory sequences and the cumulative span of those sequence consists of >80% of the non-coding region of the genome. Studies on DHSs help to disclose delicate gene regulation mechanisms and enable extensive annotation of the genome. The genome-wide mapping of DHSs provides a novel platform for the promising investigation of a specific molecular biological problems affected by the regulation of a given gene or a group of genes (17).

In addition, DHSs form a complicated, spatially- and temporally-specific network. Certain DHSs identified in one cell type by DNase-seq may not occur in the other cell types. Pan et al (40) reported that 12 DHSs in chromatin related to the Msx2 gene varied in different cell types in the chicken, when they examined anterior and posterior limb mesenchymal cells, calvarial osteoblasts and fibroblasts in embryos. Most of the DHSs were not detected as active in any of the four typed of cells, and only the DHS in the basal promoter region was present in all four tissues. One DHS was active and unique in the cells with Msx2 transcripts, and a secondary DHS was unique in non-expressing cells. The anterior and posterior limb mesenchyme cells had a distinct group of DHSs, which were more complex than those detected in calvarial osteoblast cells, which suggested that a complicated DHS pattern may be involved in the different regulation models of the Msx2 gene in these two tissues, and is involved in cell fate decision by interaction with cell-specific transcription factors to guide the transcription program of cell fate decision and development. DHSs of a certain gene may also change in response to different transcription activity. Grünweller et al (41) examined the 5'-end of the vigilin gene in chickens using the DNase-seq technique and reporter gene analysis method. They identified two candidate DHSs. One DHS was active and unique under high transcriptional activity of the vigilin gene promoter in the chicken cells, which was termed DHS1, and a secondary DHS was only found under low transcriptional activity, which was termed DHS2. The activity of the promoter of the vigilin gene was enhanced over 10 times by upstream sequences of the transcription start site (TSS). Identifying DHSs and comparing their features differs among various cell types or within a similar cell type, but culture in different circumstances is essential for revealing gene expression patterns under different conditions. This can effectively complement current understanding and may have potential clinical applications for disease treatment. The exploitation of variable and plastic patterns of active DHSs offers potential for the identification of certain cell or tissue states, which may have potential to be applied to clinical diagnoses and predictions or the evaluation of therapeutic effects.

Studies investigating DHSs facilitate the identification of novel CREs, as DHSs are more promising indicators for the identification of chromatin accessibility, which have been widely used to map functional regulation elements. DHSs overlie CREs with parallel degrees of nuclease sensitivity and cover the main sequence of regulatory factor (42). DHSs usually contain CREs related to transcriptional activation on the reporter locus, such as enhancers, but can also contain transcription inhibition, such as silencers (17). A DHS map reveals the state and pattern of the presence of CREs, in addition to the variable and plastic states of chromatin in various cell types (25,29,33,43). Liu et al (44) identified 17,472 specific DHSs and transcription factor binding sites in two cell lines, the hESC H1 cell line and trophoblast (TB)-treated cell line, and constructed a transcription factor network for placental development. The specific DHSs in the TB-treated cells were found in the ‘blood vessel’ and ‘trophectoderm’, including members of the transcription factor motif family: Leucine zipper, helix-loop-helix, GATA and ETS. The model of a TB system induced by bone morphogenetic protein 4 (BMP4) was demonstrated to be important in investigating the mechanism of trophoblast development and revealed novel candidate genes involved in the regulation of human placental development. These findings indicate that DHSs enable the precise delineation of genomic CREs. Further investigations on DHSs are expected to reveal more novel regulatory elements.

Identifying sequence variations of the DHSs of phylogenetic trees instead of the coding region of genes may assist in disclosing the changes and evolution of certain phenotypes. Dong et al (45) analyzed and evaluated the accelerated evolution of orthologous sequences at DHSs from the human genome and primate genomes using systematic biology methods, and constructed a comparison map between the DHSs and ancient repeat elements (AREs). Their analysis identified the local AREs of all DHSs and demonstrated that they were neutrally evolving. Therefore, it is noteworthy that

0.44% of DHSs in the human genome are undergoing accelerated evolution (termed ace-DHSs). Further analysis of ace-DHSs is warranted for investigating the evolution of human-specific phenotypes. These DHS analyses are important in basic studies and may be of potential value in translational medicine and personalized medicine.

4. Available data of DHSs

The ENCODE project ( and aims to evolve comprehensive schemes to list all human DHSs in order to map and catalog genome-wide CREs. DHSs mark transcriptionally active sites of chromatin, which may be the origin of cell selectivity.

The ENCODE research institutes have performed genome-wide mapping of DHSs in >100 human cell and tissue types, and almost 3,000,000 DHSs have been identified, including 71 normal differentiated primary cells, 16 immortalized primary cells, 30 malignancy-derived cell lines and eight multipotent and pluripotent progenitor cells. The 20-50-bp reads from the DNase-seq experiments enabled unique mapping to 86.9% of the genomic sequence, allowing the interrogation of a large fraction of transposon sequences. The DHS profiles of 125 different human cell types were obtained, and of these, 34% were specific to individual cell types and only a minority were detected in all cell types (3,692). The open state of DHSs varied >100 times, but the constitution pattern was consistent in distinct cell types. It was demonstrated that

5% of DHSs were detected in the TSS region, while the remaining 95% represented distal DHSs dispersed uniformly in intronic and intergenic regions. These data provide additional information for disclosing the mechanism of transcription.

5. DHSs and diseases

DHSs are associated with multiple diseases and have been suggested to serve distinct roles in the etiology of cancer, immune-related diseases, inflammatory bowel disease, Alzheimer's disease, bone marrow density problems, coronary artery disease, autism, and certain common diseases and complex traits (Fig. 3) (46). Recent evidence demonstrates the potential value of cell-specific and disease-related DHSs in personalized medicine. Evidence showing high overlap between human diseases and CREs has been well-documented, which confirm that ‘critical’ cell types may function as causal factors for certain diseases or help to maintain certain phenotypic traits (46).

Figure 3.

Process of identification of disease-related DHSs. DHSs, DNase I-hypersensitive sites CRE, cis-regulatory element TSS, transcription start site.

The accessibility or inaccessibility of the state of DHSs is reported to be associated with diseases. An increasing number of novel DHSs have been found to be associated with diseases. Specific cell and tissue types have been identified as being associated with different diseases. For example, specific immune cell types are involved in immune-related diseases (inflammatory bowel disease), and specific tissue types are involved in diseases affecting specific organs (coronary artery disease), with other associations including adrenal glands in coronary artery disease, immune systems in Alzheimer's disease and kidneys with bone marrow density (46).

Thousands of tumor-specific DHSs located at promoter and enhancer regions have been detected, which have been shown to be involved in the occurrence and development of cancer.

Function-related mutations of the DHS region are closely correlated with transcription initiation activity and thus result in the occurrence of certain diseases (Fig. 4). There are >100 studies on DHSs that assessed various cell or tissue types by ENCODE Alliance (124 different cell types) and NIH roadmap epigenomics group (342 different adult fetal tissue samples), which demonstrate that an overlap exists between mutations at non-coding DNA regulatory sequences and diseases and traits.

Figure 4.

Pathological mechanism of variation at DHS sites of disease-related genes. (A) DHSs act as binding anchors of activator proteins and cofactors. A multiprotein complex is assembled and bound to the ‘open’ sequences of DHSs, including promoter or enhancer regions, and after chromatin rearrangement this leads to gene transcription. (B) When pathological mutations or variations occur at crucial regions of DHSs, it influences the identification and interaction of regulatory binding proteins and the transcription function is interrupted or altered, which may subsequently cause phenotypes. DHSs, DNase I-hypersensitive sites.

GWASs have identified numerous single nucleotide polymorphisms (SNPs) at DHSs, which are associated with various types of quantitative traits and complex disorders. Local mutation density is variable throughout the genome (47). A study on 1,161 human cancer genomes revealed that the density of point mutations at the center of the DHS in the gene promoter region of somatic cells was increased (48). Numerous tissue types, including brain, pancreatic and liver tissue, have also demonstrated the enrichment of SNPs associated with DHSs in major depressive disorder (49). A 14-kb Down syndrome cell adhesion molecule deletion sequence, containing 12 CNS DHSs, was found in an autistic family, with regulatory potential affecting the biology of the central nervous system (50). De novo mutations, rich in DHSs and proximal genes, have been significantly predicted to result in the loss of transcription binding factors. For example, deletion of lysine-specific demethylase 5B binding was found at the promoter of the candidate autism risk gene, EFR3A (51).

A mutation at the SNP (chr18:52417839 G>C) site was reported to be correlated with follicular thyroid cancer, which had an influence on the binding of tumor suppressor protein p53 and subsequently resulted in the decreased gene expression of thioredoxin-like 1 ( TXNL1) (33). The highest prevalence of mutations was found in the hypothetical driving factor DHS chr5:1325957-1328153, located in an intron of the Cleft lip and palate transmembrane 1-like (CLPTM1L) gene and 30 kb upstream of telomerase reverse transcriptase (TERT), results in the overexpression of six adjacent genes and four of these genes [TERT, CLPTM1L, thyroid hormone receptor interactor 13 (TRIP13), lysophosphatidylcholine acyltransferase 1 (LPCAT1)] are known to be associated with cancer (52-54).

Recently, a statistical method has been developed to identify distal regulatory elements with hypothetical driving mutations in breast cancer, to identify DHSs in non-coding genomic sequences associated with significant mutations in breast cancer and abnormal expression of adjacent genes, which may be important in the development of cancer (55). The mutation of chr5:1325957-1328153 at the DHS region in breast cancer was reported to be associated with the overexpression of oncogene tripartite motif containing 27 (TRIM27). In addition, abnormal activity was found with the mutation of chr6:28948439-28951450 at the DHS region. Using data from The Cancer Genome Atlas (TCGA) and breast cancer International Alliance (metabonomics) molecular taxonomy, Guo et al (56) found two hypothetical functional variants, rs62331150 and rs73838678, located at the DHS site and transcription factor binding region. Among them, rs62331150 was associated with the expression of tet methylcytosine dioxygenase 2 (TET2) in normal breast tissue and tumor tissue. Two new SNP (rs12309362 and rs9970827) were found to be significantly associated with reducing the risk of hepatocellular carcinoma (HCC) by measuring the mutations at the peak of DHSs in 1,538 patients with HCC and 1,465 normal controls (57).

A study on endometriosis, by re-sequencing 1.29 mb of the 9p21 region, revealed that the mutation of rs17761446 at the DHS was associated with endometriosis, the protective G allele at this site had a strong interaction with the ANRIL promoter. Further chromatin immunoprecipitation analysis confirmed that the protective G allele also had a preferential binding capacity with transcription factor 7-like 2, EP300 and may be involved in the development of endometriosis (58).

Studies on prostate cancer and breast cancer cells have revealed that different DHS patterns of androgen receptor (AR) and estrogen receptor 1 (ESR1) were of high predictive value for hormone receptor binding and may be involved in the development of these types of cancer. The quantitative measurement of DHS changes can predict the binding sites of perturbation-inducible transcription factors (59).

Following activation of the transcription of AR in LNCaP cells, 244 upregulated and 486 downregulated accessible DHS regions were detected to be the candidate sites for further investigation, which may be associated with prostate cancer. CTCF and the ELK1-ETS transcription factor are potential upstream regulator elements, which are rich in open promoter regions of downregulated genes. The inhibitor of DNA-binding 1 HLH protein (ID1) is the only transcription factor that is significantly upregulated, exhibiting basal sequence enrichment in the promoter region of the upregulated gene. Therefore, CTCF, ELK1 and ID1 may be potential targets for the treatment of prostate cancer (60). Increasing evidence shows that changes in the expression of BMP4 are involved in the pathogenesis of cancer, which is associated with cancer metastasis and progression, including rectal, hepatocellular and ovarian cancer (61). In order to determine the characteristics of BMP4 transcription mediators in breast cancer, RNA-Seq and DNase-seq were analyzed in T-47D and MDA-MB-231 breast cancer cells treated with BMP4. It was confirmed that MBD2, core-binding factor-β and hypoxia-inducible factor 1α were downstream regulators of the BMP4 signal, which enhanced cell migration and decreased cell growth (62).

In tumor therapy, particularly in acute myeloid leukemia (AML), intratumoral heterogeneity caused by clonal evolution has been found, which may have an influence on the effect of treatment. In order to solve this problem, the chromatin accessibility of subclones of AML was compared directly using unsupervised clustering analysis. Marked differences in the chromatin landscape and transcriptional regulation among the subclones were detected and confirmed. The data indicated that the common DHSs of individual AML subclones dominated in the clustering analysis over the subclone-specific DHSs, most likely due to the impact of shared founder mutations at the DHSs in each AML subclone. Clone-specific DHSs, runt-related transcription factor and ETS motifs are expressed in abundance in the two clones of DHSs, although GATA motifs are particularly abundant in FLT3-WT clones (63). It may be a potential strategy to use DHSs analysis to improve the treatment effect, particularly for those types of cancer with features of intraturmoral heterogeneity.

Genetic variation at DHSs has been reported to be correlated with carcinogenesis (64). By analyzing 1,161 human cancer samples from 14 types of cancer, DHS profiles and SNP distributions were mapped to link to promoter activity, some of which were involved in differential nucleotide excision repair (NER) and resulted in carcinogenesis (48). Consistent with this finding, genome-wide maps of NER regions show that the repair ability of nucleotide excision was decreased with mutation at the DHS of gene promoter regions.

Jin et al (33) reported that thousands of tumor-specific DHSs were identified on cells dissected from follicular thyroid carcinoma samples fixed on formalin-fixed paraffin-embedded slides. Numerous DHSs have been reported to be correlated with the development of thyroid cancer (33). A de novo mutation (chr18:52417839 G>C) at a DHS located downstream of the TXNL1 gene is associated with the formation of thyroid carcinoma. It was reported that rs62331150 located at promoter region and rs73838678 located at the enhancer region of the gene, increased the risk of breast cancer. These two SNP sites were found to be in linkage disequilibrium with rs9790517 of the adjacent TET2 gene (55). It was also found that, in samples with mutation at the rs12309362 and rs9970827 sites, the risk of being affected with HCC decreased significantly (57).

Ten DHSs were identified as being mutated with abnormal expression of target genes in breast cancer (55). Mutation at the DHS chr5:1325957-1328153, was present at a high frequency in the cancer cells, resulting in the high level of transcription of certain genes close to it, including TERT, CLPTM1L, TRIP13 and LPCAT1, which has been confirmed to be associated with cancer (52,53,65). Mutations at DHSs chr5:1325957-1328153 were found to result in the high expression of TRIM27, and certain mutations in this region caused abnormal accessibility of DHS chr6:28948439-28951450, which are associated with cancer.

The pattern of DHSs can be stimulated by hormones through regulation of the binding capacity of AR and ESR1 in prostate cancer cells and breast cancer cells. Following binding with AR or ESR1, the DNase I-hypersensitivity of certain sequences was found to be altered, and the regional nucleosome occupancy changed for AR binding but not for ESR1b binding, which indicated different interaction modes in AR and ESR1 regulation (59).

In Gene Ontology analysis, genes associated with tumor-specific DHSs are abundant in biological processes, including the regulation of GTPase activity and response to hypoxia, and cancer-related pathways. Understanding the accessibility dynamics of chromatin in the process of disease occurrence and development can provide insights into how cell fate is regulated, and how transcriptional systems are organized and regulated in different tissues and how they are destroyed in disease states. In addition, the application of DHSs in biomedical research can expand the field of cell-selective gene regulation analysis, enabling the identification of long-range regulatory patterns of the system and previously undescribed phenomena, such as DHS activation patterns and mutation rates in abnormal and immortal cells.

6. Discussion

Although DHSs occupy a small portion of human genome,

2% of the genome, a relatively large proportion of CREs may be involved in the establishment of well-organized expression networks in each cell type and thus contribute to the etiology of a certain disease. The comprehensive delineation of distribution, constituents and biological activities of DHSs help to map and classify functional CREs. The identification of CREs is critical for elucidating the mechanism of gene expression regulation underlying biological events, and the development and progression of certain diseases.

To date, the DNase-seq technique remains one of the most efficient techniques for disclosing the known and unknown regulatory elements of diverse target cells. Cooper et al (32) released a detailed protocol in Nature that may facilitate the spread of DNase-seq analysis. However, the number of researchers able to utilize the high-throughput DHS capture technique well is limited. In addition, careful manipulation is required due to high background noise. Difficulties in performing experiments are not the greatest challenge for its wider application bioinformatics analysis is difficult for the majority of laboratories. An insufficient number of bioinformatics technicians, particularly those with the required programming skills and familiarity in this field, is the main concern. The exploitation of software or online tools is required to simplify bioinformatics analysis and enable easier understanding by researchers.

Although evidence has identified an association between DHSs and diseases, using DHSs as a marker for prediction, prevention and pre-clinical diagnosis remains a challenge due to the complexity of regulation of the DHS profile. The uncertainty of genome-wide prediction of CREs and specific DHS related to gene function may increase the challenge of its application in clinical practice. Future work should focus on the investigation of more delicate methods to locate DHSs precisely, help disclose the mechanisms underlying gene expression differences, determine how to modify chromatin accessibility, reveal how changes in transcription factor binding are driven by genetic variations, and guide how to integrate DHSs in clinical practice.



This study was supported by the National Natural Science Foundation of China (grant no. 81671473), the Key Talents of Jiangsu Province (grant nos. WSW-108 and FRC201754) and the Innovation Team Project of Wuxi (grant no. CXTDJS003).