Neuroproteomics, Protein Folding, Transcription Factors, and Epigenetics for TBI Research

We first examine the characteristics of the eukaryotic chromosome, including anchoring proteins, coiling, and associated methods of compaction. Specific proteins anchor to the plasma membrane and interact with DNA during the supercoiling process. They also function to hold together loop domain centers. More complex arrangements exhibit in eukaryotic chromosome protein anchors, where the histone amino acids bind DNA, changing conformation and effecting the compaction process. Nuclear matrix proteins bond to chromosome 30 nm fibers for compaction through radial loop domains; and during mitosis, chromosomes are anchored to the spindle—a process unique to the eukaryote.

It is instructive to consider the varied roles of the structural maintenance family (SMC)—in particular, the cohesin-protein complex. In addition to its “traditional” roles of controlling sister chromatid separation during metaphase, facilitating chromosome spindle attachment, and assisting in DNA repair, cohesin provides a stabilizing role in insulator effects on the major histocompatibility complex (MHC) transcription, through interactions with the MHC insulators. It is thought that the cohesin ring structure is able to encircle sister chromatids, providing structure and support for uncondensed sister chromatid coherence, condensation, and freeing of chromatids at prophase, and separation of condensed chromatids at anaphase.

Chromosome assembly and segregation in eukaryotes is further assisted by another SMC family member: the large protein complex condensin that travels from the cytoplasm to the nucleus during the start of the M phase. Condensin also binds to chromosomes and assists in compacting radial loops. Two distinct SMC family condensing complexes are also seen to assist with the assembly of condensed chromosomes.

Chromosomes must be coiled for the cell structures to accommodate their size. Coiling may involve circular, closed loop, naked (that is, no histone proteins), double- stranded DNA molecules. These need to be coiled greater than 1,000 fold for sufficient compaction into the cell. Sister chromatids at the conclusion of interphase are double helix DNA, wrapped around histones, forming the nucleosome. This structure is supercoiled into 10- and 30-nm beads. The eukaryotic chromosomes are fairly linear (relative to bacterial chromosomes), and coil into roughly an “X” shape. Seven-fold compacted, “beads on a string” structures are further compacted via histone mediation to 30-nm fibers, getting us to 50-fold compaction.

The next sequence of compaction utilizes nuclear matrix proteins wound around histones, forming coiled beads or nucleosomes. Further twisting into a 700-nm-wide solenoidal structure is accomplished, and radial loop domains provide additional functionality. We call this DNA/histone combination “chromatin.” Further compaction occurs in the heterochromatin. In bacterial chromosomes, DNA binding proteins bind loop domains that then coil in a 10-fold compaction. Then the loop domain structures are “supercoiled” by enzyme catalyzed twisting and compaction of the DNA into itself. After looping and supercoiling, the chromosome then fits into the nucleoid.

Protein domain structure relates to the binding of proteins to DNA, binding of proteins to other proteins, and to domains that activate transcription. We define protein domains as portions of protein sequences with independent function and the ability to fold and reconfigure in an independent manner. One or more motifs comprise the domain, with each motif providing specific structure to the domain. Domains that are part of a macromolecular structure modulate molecular flexibility through conformational changes that serve to alter the accessibility of certain regions, alter reactivity, and positioning—all functioning as a complex lock and key mechanism. We can further analogize domains as Swiss army knives, with varieties of interaction sequences that can be effected in sequencing and activating docking sites, functionalities, and temporal relationships. We suggest as well a thorough examination of the actions of hydrogen bonded water networks that mediate domains and the thermodynamic “Anfinsen” partition function within the cellular confines—where much of the domain activity occurs.

Specific relevance to our three categories of DNA binding, proteinprotein binding, and transcription activation is evidenced in the wealth of evolving knowledge and validation through X-ray diffraction, especially crystallization of metastable domains—and emerging methodologies such as neutron spin echo spectroscopy, and protein tomography, providing visualization of individual domains.

The major groove in a DNA double helix accommodates binding proteins with various domain modules. Dimeric binding proteins provide dimeric receptors and many permutations at a given DNA binding site (in fact there are currently more than 300 complexes and bioinformatic structural predictions on the protein database, along with about 45,000 3D structures confirmed experimentally thus far.) Conserved protein domains provide binding to specific activators controlling signal transduction pathways. A key example of this is the zinc finger domain, exhibiting conformational changes when bound to DNA binding sites.

Several domains can be bound together resulting in further diversity of physicochemical properties and multifunctionality—not just to provide binding sites and regulatory functions—but also as building blocks for biological assemblies and tissue structure. Some excellent examples are binding of DNA helicase to DNA Pol III holoenzyme, regulating DNA strand separation; initiation of eukaryotic DNA replication via the origin recognition protein complex; and finally, the glycolytic enzyme pyruvate kinase, containing a beta regulatory domain, alpha/beta substrate binding domain, and an alpha/beta nucleotide binding domain, all connected with polypeptide linkers.

Protein domains that activate or regulate transcription are exemplified by the leucine zipper domain, mediating DNA binding, dimerization, and transcription initiation. Zinc finger domains (ZnF_GATA) are transcription factors that bind promoter sequences—as a further example of transcription activation domains. In general, we see activators binding enhancers, and repressors binding silencers. Domain dimerization for transcription includes the homodimer and heterodimer formation of identical transcription factors, or different factors, respectively. TFIID and Mediator regulatory transcription factors communicate directly with RNA polymerase, and indirect transcription regulation involves recruiting nucleo- some positioning proteins, as well as histone and DNA modifying proteins in the chromatin-remodeling process.

Transcription factors are influenced by two categories of signals: first are external signals including cytokines, growth factors, hormones, etc., along w'ith stress type responses; and second is a set of intrinsic factors to include transcription, DNA replication, and chromosome modification/segregation. Histone modifications result in repression or alteration of genome function (chromatin deregulation) and, along with extrinsic factor regulation, collectively actualize the epigenetic condition determining modifications and portions of the genetic sequence to be available for transcription. Post-translational modification of histones (HPTMs) includes acetylation, phosphorylation, and methylation, as well as ubiquitylation and sumoylation—the latter two via much larger peptides.

DNA methylation represents a potentially informative form of epigenetic modulation of transcription. Three models describe how these modifications activate or repress transcription: (1) HPTM alters chromatin structure (e.g. by changing electronic charge); (2) HPTM prevents binding of a negative factor to the chromatin template; (3) HPTM creates a binding site for a positive factor. One is typically a cis modifier; and the other two are usually trans modifiers.

Acetylation (adding an acyl to an active hydrogen site [H3N+] on an N tail of lysine residues) activates transcription at H3-K9, K14, K18. K56; H4-K5, K8, K12. K16; H2A, and H2B-K6, K7, К16, and К17. Transcription is also activated by phosphorylation at H3-S10; arginine methylation at H3-K4, K36, and K79; and ubiquitylation at H2B-K123. Transcription is repressed via arginine methylation at H3-K9, K27, and H4-K20; ubiquitylation at H2A-K119; and sumoylation at НЗ, H4-K5, K8, К12, К16, H2A-K126, and H2B-K6, K7, K16, and K17. Acetylation is catalyzed by site-specific histone acetyltransferases and reversed with deacetylase (HDAC). Typically, activators recruit HATs and repressors recruit HDACs. This compacts the nucleosome (model 1) by converting a positive charge at the H,N+ to a partial negative charge due to the =0 bond that repels the DNA and tail-opening up binding sites on the DNA as well as possibly decompacting the nucleosome, making chromatin more easily activated.

Via model 3, bromodomains bind to acetylated lysine residues, and may be part of HAT motifs such as CPB/рЗОО and Gcn5, in remodeling complexes like Swi/Snf, which promote binding to chromatin. Other bromodomains may also be attracted such as Tafl and TFIID complexes, Rsc4 in Rsc, and Brd in various proteins. Transcription repression results from deacetylation via HDAC Sir2 enzymes, acting with cofactors such as Rpd3 in the HDAC Sin3 complex, or Rpd3 in НЗКЗбше. This suppresses RNA Pol II, and regulates the steps of the transcription cycle.

Phosphorylation location is also determined by enzymatic specificity. Proposed mechanisms involve all three models: residue cluster phosphorylation in HI alters DNA binding affinity, locally increasing transcription potential of the chromatin; ala model 2: HP1 binding affinity is lowered throughout mitosis via phosphorylation at H3-S10; and ala model 3: НЗ-SIO phosphorylation is recognized by an adapter protein, inducing transcription. Models 1 and 3 apply to histone lysine methylation, depending on the site. One proposed mechanism is SU V39H1/Clr4 signaling to recruit HP1 or Swi6 and Chp2, effecting replacement of H3 with CENP-A. Demethylation is accomplished via deimination, removing methyls, or via demethylase. Arginine methylation occurs mostly in the nucleus, positively or negatively affecting transcription via methyltransferases that vary in substrate specificity, cellular localization, and targeting. Deimination can activate or repress transcription, depending on tissue and location. Finally, ubiquitylation and sumoylation can modulate transcription by activating methylation or regulating acetylation.

DNA methylation prevents binding of transcription factors to promoters. It is thought that MeCP2 recruits Sin3A HDAC and HKMT to methylated sites (marks), repressing transcription initiation complexes. Gene expression can be activated via mediation of DNA methylation by MDB proteins at CpG islands, which can repress or potentially activate transcription at the CpG islands (an example of RNA activation). In a different example, methylation outside the islands can modulate stem cell differentiation for blood cell type. Imprinting can also result, with parent sex specific positive transcription activation via variable epigenetic cluster modification at combinations of sites, providing transcription activation variants. RISC and RITS processes modulate genetic expression (activate and/or repress) during various phases of the cell cycle.

Comparing repressive effects of chromatin and DNA modifications to those of RNAi, we first provide a basic overview of RISC post-translational silencing and RITS pre-translational silencing. RISC is the RNA silence-inducing complex, directed by siRNAs and miRNAs, plus Sheer; and RITS is the RNA-induced transcriptional silencing complex, which uses siRNA base pairing, stabilized by methylated H3, providing epigenetic chromatin modification. In RISC, siRNA base pairs to mRNAs, recognizing them for degradation and cleavage; while in RITS, the siRNAs target the silencing complex to regions of the chromosome designated for chromatin modification.

RNAi pathways involve enzymatic processing of primary miRNA or dsRNA. Similar to the recruitment via histone modifications of transferases to modified histones, RNAi recruits enzymes to the chromatin that provide modifications. For miRNA, Drosha generates pre-miRNA that is transported to the nucleus by exportin5—while the Dicer cleaves the dsRNA in the cytosol. Dicer removes the stem loop used in miRNA, producing the mature miRNA complex that gets incorporated in the RISC complex after helicase cleavage; while helicase also cleaves the siRNA duplexes for their subsequent incorporation into the RISC. The pathways for miRNA and siRNA to their RISC complexes—along with Argonaute (a common feature of RITS and RISC)—provide a means to target and cleave mRNA in the RISC complex complementary to the RNAi. But in the RITS silencing complex, the target DNA is silenced via the formation of heterochromatin. Argonaute binds the guide strand to the direct silencing of genes.

RISC thereby controls the gene silencing process—which can include the chromatin modifications discussed. RNAi mechanisms that are unique include the immune defense against viruses and transposons—roles performed by RNAi in addition to gene regulation. Demethylation and remethylation during cell differentiation and development temporarily program epigenetic patterns, a more permanent effect than the DNMT catalyzed methylation patterns in non-CpG DNA. RITS controls histone modification (and formation of heterochromatin) prior to transcription. Additional unique roles of RNAi include control of morphogenesis, stem cell maintenance—especially during differentiation and a role as either oncogenes or tumor suppressors. Plural dynamic modifications result in the specificity of position, timing, and systematic control of gene transcription through combinations of the processes discussed.

RNAi can serve to destroy mRNA in addition to modulating translation—another unique process. It can also silence a promoter, in addition to directing chromatin modification. And unlike a simple and direct histone modification that affects more than just the chromosome and branch location of interest—RNAi can be used to switch on and off a specific site of interest—making it quite valuable for manipulation of gene expression. RNAi silencing represents the more efficient of the processes, owing to this specificity and precision of its processes, compared with our menu of histone modifications. Control via RNAi of heterochromatin initiation and assembly requires the process pathways described for RITS, where it is postulated that specific nuclear regions are assembled with specific configurations and timing directed via the RNAi RITS pathways—in which RITS recruits the proper enzymes (such as RDRC) to select RNA templates.

As the human epigenome catalog is established, the direct role of RNAi for epigenetic modulation, gene silencing, chromatin remodeling, cell development, immune responses, and other interrelated roles of the processes discussed will be elucidated. It is anticipated this will lead to an improved understanding of a wide range of interrelated and cascaded pathologies, along with their practicable cognate therapeutic protocols.

Tissue microarray analyses provide links between global histone modifications (such as acetylation and methylation of lysine and arginine residues) and susceptibility to oncogene promotion and recurrence of prostate cancer. Mechanisms for this may include chromosomal translocation enhancement via modifications of exchange kinetics (by modulating the accessibility, reactivity, and positioning of protein domains that activate transcription) between nonhomologous chromosomes, producing constitutively active chimeric proteins or oncogenic receptor activation, and inappropriate expression of genes regulating growth and amplification of protooncogenic segments—along with angiogenesis increase, avoidance of apoptosis, diminution of DNA repair capability, and increased tissue invasion capacity. DNA microarray analysis of cancer phenotypes indicates signaling pathway aberrations effected via the disruption of intracellular antagonists, receptor blockers, reconfigured enhanceo- somes, and spatiotemporal signal chains, some of which result in oncogenic progression. Changes in gene expression patterns resulting from chromatin modifications manifest, for example, as various lymphomas that are otherwise phenotypically similar. Additional examples are loss of function mutation in Swi/Snf, increasing proliferation of E2F and leading to cancer progression, and activation of prometastatic genes such as urokinase plasminogen activator and heparanase.

We’ve explored the conserved role of intragenic differential methylation across the gene body, at the 5' promoter, in the introns (intragenic), at the 3' end of the DNA transcript, and in the gene-sparse intergenic region, and where the C-phosphate-G (CG) nucleotides occur in higher concentrations or islands (CpGs). Tissue-specific methylation via DNMT enzymes indeed regulates promoter activity in the intragenic and 3' transcript end, and in intergenic regions; regulating alternative (distant) promoters (also termed alternative CGI promoters); underscoring the importance of histone methylation and acetylation (along with phosphorylation, ubiquitination, ADP ribosylation, etc.) in genetic and alternate promoter regulation, as potential oncogene-inducing factors. Aberrant DNA methylation (hyper and hypo) of non- CpG island promoters, and genome wide hypo, can lead directly to cancer. Finally, interactions of various disease factors create large number of permutations, providing a diversity of genetic diseases. For instance, DNA methylation may cause histone deacetylation—and vice versa; and methyl binding proteins such as MeCP and MBD can deliver HDACs and HMTs that target, for instance, repressive lysines, deacety- late histones, or remodel chromatin.

CTCF-cohesin interaction with the DNA substrate could involve DNA loop(s) (influenced by insulators); or maybe spatially proximal chromatinized helixes (influenced by sister chromatid pairing). CTCF dimerization may provide the basis for MHC-II insulators “holding together” the stabilization of gene expression via levels of functional cohesin in the cell, and MHC.-II insulators serving as nucleation points for the transcription functional complex—as well as cohesin regulation of transcription via stabilization of bound distal insulators with proximal promoter sequences. It is suggested that the cohesin ring structure is able to encircle the sister chromatids, providing structure and support. Additionally, cellular responses to RNAi silencing, and to the enzymatic processing of primary miRNA or dsRNA, represent additional mechanisms for dysregulation/disease of genes and cells when subject to improper “interference” of gene regulation and integrity. All such factors illustrate specific mechanisms for the dysregulation of genes and cells, and causations for disease.

Examples of genetic dysfunction related diseases involve single causative factors or various combinations of the following: uniparental disomy, deletions, imprint defects, point mutations, translocation mutations, alternate splicing and duplications. Some selected examples of specific diseases and their mechanisms include the following. Fragile X syndrome symptoms include retardation, large forehead and ears, with some neurodevelopmental phenotypes. Etiology is mapped to the “fragile” Xq27.3 chromosome; and FMRl (encodes FMRP protein) “malexpression” causes unstable CGG repeat, mediated in cis (aberrant methylation, acetylation). Fragile X involves a primary mutation plus secondary epigenetic mutation; FMRP malexpression is also related to synaptic developmental issues. In Rett syndrome, a neurological disorder exhibiting multiple symptoms, the q27 arm of the X chromosome in the gene encoding MECP2 methyl binding protein is bound with methylated CpG dinucleotides, acting to repress transcription through recruitment of the Sin3A, SWI-SNF, and HDAC repressors by maintaining localized heterochromatin. The same epigenetic mechanisms discussed herein may also lead to a better understanding of complex diseases, in addition to those directly related to epigenetic abnormalities—and to therapies based on both histone modifications and DNA methylation and RNAi pathway mediation.

Epigenetics, as the totality of chromatin template alterations effecting transcription and silencing of genes from a common genome, extends our understanding of genomic bioinformatics and the biotechnology arena—along with our perspectives on causality, previously alleged “junk” DNA, and the fundamental considerations for use of cancer and stem cell lines. Epigenetics modifies the means by which the internal code of the genotype manifests its external product, the phenotype—resulting in diversity of form and function (morphology). [Ala the Central Dogma, DNA to RNA (transcription) to protein (translation) to traits]. Epigenetics describes the interactions of the genotype with the environment to produce the phenotype—extending the traditional Mendelian genetics of heredity with a much broader paradigm. The new paradigm includes a diversity of pathways linking phenotype to genotype. These include DNA-protein interactions, transcription-translation relationships based on temporal sequences of events and environments within and without the cell/nucleus, and molecular process stochastics. Epigenetic dialectic derives from the complexity of protein effectors modulated by histone modifications, and enzyme substrate specificity directing repression or activation of transcription; all bestowed with additional specificity through chromatin remodeling, HATs, DNMTs, HDACs, HDMs, DDMs, and DNA methylytransferases and demethylases—as well as even more specific transcription factors representing the genotype contribution to the epigenetic manifestation of the phenotype.

Significant changes in epigenetic coding occur during the development life cycle. Embryonic stem cells can reprogram the genome’s epigenetic marks, providing diversity of gene regulation and expression (pluripotency), and potentially extending into successive generations; while adult stem cells can be used for more specialized purposes and medical therapies. Epigenetic programming involves intervention in the epigenetics of pluripotent stem cells. This can be accompanied by germ cell line production in culture for research or therapies—perhaps bringing us very close to the controversial, albeit illegal, practice of human cloning—where distinction might be needed for cells, tissues, embryos, and humans—and a legal delineation established for the question of where life begins? I propose it begins with consciousness. For example, Henrietta Lacks’ cancer cells—while “immortal” are not conscious.

The idea of personalized embryonic stem cells addressing individual needs is certainly an attractive prospect. The key benefit to be derived from such research is obtained by identifying an abnormality and studying the associated epigenetics— rather than justifying the effort by the lives it may save; since the genetic diseases may present in only a very small percentage of the population, thereby obviating a prioritization based on saving lives.

But if by researching the effects, for instance, of retroviral insertion of genetic information into the nucleus, we discover a means to control or reverse a certain cancer—or to “heal” damaged brain tissue—then the focused research would indeed benefit many more lives than the expectation based on the specific research agenda at initiation. So when the biotechnologist indicates he or she is “working on a cure for cancer”—we cannot say with certainty this is not the case, despite the initially limited experimental domain. Biochemistry affords the merger of bioinformatics, genomics, epigenomics and proteomics—providing rich potential for rapid advancements—thereby further justifying the expansion of research in these areas. Undesired avoidance of apoptosis and growth prevention factors leading to cancer may provide positive benefit to proliferation and immortality of, for instance, a HeLa culture for epigenetic mapping experiments. Therefore, the benefits of directed research are not totally predictable.

Since mammals cannot reproduce via parthenogenesis, parent-specific epigenetic processes are needed for the genome during gametogenesis; i.e., differential expression of imprinting by the two parental genetic alleles, unless—and we know our exceptions are plenteous—the female gamete epigenotype can be manipulated, for example, to produce males that are exclusively of maternal origin. Perhaps that’s “not the android we’re looking for”—as natural processes of evolution and genetic conservation could lead to unanticipated results such as disease vulnerability or extinction. A totipotent mammal could indeed prove quite lacking in the raison d’Stre, or the claimed reason for existence, that provides the moral fabric upon which we base our ethical value system. Not until we’ve achieved a closed-form solution to the systematic relationships between genotype, phenotype, disease, and the environment should we propose optimization or global bioreprogramming routines. To this end— none of the DNA is garbage—as indirect and combinatorial processes may utilize any portions not currently translating to protein. We should not perturb the chromatin structure until we’ve solved and resolved the complete set of algorithms represented in the physical solution space. The epigenetic solution space involves putting garbage DNA to work in a productive manner, supervised in part by microRNA, environmental and epigenetic fabrics, and temporal sequencing of logical biosystem space, tempered by advantageous transfection of desirable information and knowledge.

We now examine transport and the nuclear pore complex (NPC). “Random” diffusion of the import complex is unidirectional, from the relatively high cytoplasmic concentration to the lower concentration in the nucleoplasm; and similarly the random diffusion across a concentration gradient occurs for importin transport out of the nucleoplasm to the cytoplasm, again from high to low concentration. Ran-GEF and Ran-GAP promote cargo dissociation in the nucleoplasm, and importin dissociation in the cytoplasm, respectively; serving to maintain the concentration gradients of cargo complex and Ran~GTP across the NPC—the macromolecular cylinder spanning the nuclear membranes.

Diffusion is thought to be enabled by hydrophobic FG-repeat structures called nucleoporins (FG-nucleoporins) lining the entire surface of the NPC basket.

The hydrophobic importin cargo complexes are able to diffuse through the basket, while hydrophilic proteins are slowed or precluded. I would suggest this may be due to the effect of elimination or minimization of hydrogen bonded networks transiently forming and breaking at the cargo-NPC interface, due to the bonding exclusion resulting from the hydrophilic chaperone complex; while at the same time the FG repeats transiently bond to NLSs. Unaccompanied hydrophilic proteins could form hydrogen bonds with surface areas not entirely “covered” by FGs, thereby slowing or preventing passive diffusion through the basket. This could be assisted via the dynamic reconfiguration of the Phe-Gly repeats in the Nups—and consistent with the consensus three-step mechanisms.

Or perhaps the Nups lining the basket exhibit some periodicity corresponding to the hydrophobic/hydrophilic sequence distributions along the cargo complex. This would be an interesting problem to model/simulate with computational molecular dynamics—and to then validate with an artificial nanotube Nup construct. Perhaps the clathrin and coat protein folds similar to those of the Nups would reveal the nature of the hydrophobic interactions occurring as the FG’s bind to the cognate- shuttling proteins. These periodic interactions might serve as a transport mechanism that “moves the cargo complex along” as it diffuses through the NPC. The EM imaging that demonstrated the dynamic and flexible nature of the NPC itself is exciting— and the “nanoaquarium” developed at the University of Pennsylvania (Mauk, M., private conversation 2012) may provide a means to conduct in vivo dynamic images to further elucidate the NPC diffusion processes—as well as mRNA filtering effects.

Similar to the mechanism described for import, export from the nucleus to the cytoplasm involves shuttle proteins containing nuclear export and localization signal sequences (NES and NLS, respectively). First, a trimolecular complex is formed in the nucleus, consisting of Ran~GTP, exportin 1 (nuclear-export receptor), and the cargo to be exported. Perhaps in a manner similar to the hydrophobic effect described for import above, the trimolecular complex diffuses (again from high to low concentration gradient of cargo complex from high to low) through the NPC to the cytoplasm. There, the cargo dissociates from the complex via Ran-GAP hydro- lization of the Ran~GTP, and release of the complex constituents. One notable difference between the import and export processes is the Ran~GTP presence in the cargo complex for export, but not for import.

The hydrophobic karyopherin family of importins and exportins is highly conserved; and some of the cognate NES/NLS proteins are seen to function as both importins and exportins. The shuttling mechanisms for alternate cargos resemble the processes described—and most are Ran-dependent. Thus the NPC serves the role of transporting cargo back and forth across the nuclear envelope. As we see in the next question, the NPC also delivers genetic programming to the cytoplasm for protein synthesis.

In addition, epigenetic controls associated with the NPC include chromatin boundary delineation, modulation of transition regions of chromatin density distributions associated with varied transcription and/or mRNA genesis and repair “stations.” It is hypothesized that the NPC serves as a staging area for congregation of active genes and nucleation sites for biogenesis. Again, this staging area is replete with chromatin modeling complexes—also recruited to the “loading dock,” to be activated, modulated, regulated, and “shipped out.” Add to this, chromatin maintenance and repair, including telomere functions, and finally chromatin segregation during mitosis— and we see the vast array of functionality at the NPC, and the specifics of the gene expression portion of this activity.


Lodish, H., A. Berk, C.A. Kaiser, et al. 2008. Molecular cell biology, 6th edition. New York: W.H. Freeman and Company.

< Prev   CONTENTS   Source   Next >