Biology, Cell Biology, Biochemistry, and Science: 2011

Wednesday, June 8, 2011

Fall 2010's Biochemistry Notes

Links:

exam 2
https://docs.google.com/document/d/1v-8292bbokdMypiTyumz0QJ9CW0ScJdeyyCSzgNGTOY/edit?hl=en_US&authkey=CL67zqUE

exam 3
https://docs.google.com/leaf?id=0B99-sSwVe231YWRmYzEzOWItNmNiOS00ZWQzLTgxODYtZjhiOWI3ZWYzOGQ1&hl=en_US&authkey=CPf78pwM

exam 4
https://docs.google.com/leaf?id=0B99-sSwVe231YjY1ZWQyMjctZjFmZS00MDBkLTkyNDgtZDFiYzA1YmJlMmQy&hl=en_US&authkey=CJjumagB

Wednesday, May 18, 2011

Japanese Cartoons in Biology

Remember Sonic Hedgehog protein? Well, how about the Raichu probe?

Raichu stands for Ras interacting chimeric unit, where someone has attached a CFP to Raf to make a chimera protein (fusion protein). Then, this Raf-CFP can can interact with Ras engineered with YFP. Normally, Raf only binds to Ras-GTP, not Ras-GDP. Bioengineers invented Raichu to be able to detect and localize Ras-GTP in action, using FRET. When they use fluorescence to zap the cell with CFP's excitation wavelength, CFP is excited and sends out light at the CFP emission wavelength. Conveniently, CFP's emission wavelength is YFP's excitation wavelength, so if Ras-CFP has a GTP and binds Raf-YFP, then your initial light excites CFP which excites YFP so you will see the color of YFP's emission, not CFP's emission. If you see that, that means Ras is bound very closely to Raf.

Raichu actually stands for Ras interacting protein chimeric unit, but it's not surprising they left out the "p" when making the acronym. :P

Apoptosis video

Nice video for apoptosis, covers and LABELS all the proteins we learned about:

http://www.youtube.com/watch?v=hqhxnWty5jc

Tuesday, May 17, 2011

Book Notes: Cancer Part 2

cancer critical genes: small subset of genes whose mutation usually causes cancer

proto-oncogenes: genes who, when mutant, are overactive and drive cells to become cancerous
tumor suppressor genes: genes who prevent cancerous cells (when mutant, they are inactive and allow cells to become cancerous)
DNA maintenance genes: genes who, when mutant, are inactive and allow cells to become genomically unstable, which may or may not lead to cancer

to test possible proto-oncogenes, you insert a copy of the oncogene (mutated version)

if cells become cancerous, then the inserted gene really was an oncogene

to test possible tumor suppressor genes, you have to knockout both copies of the gene in the cell

as long as cells have one copy of a tumor suppressor gene, they can withhold the cell from abnormal growth

tumor viruses: viruses who don't intentionally cause cancer, but by inserting their own genome into the DNA, causes a mutation that leads to cancer

they are usually retroviruses, who carry RNA that is reverse transcribed into DNA and then inserted

DNA decoding of transformed cells helped determine oncogenes

tumor DNA is fragmented and put with fibroblasts that already divide a lot (but are still normal)
if one of the fragment contains an oncogene, some of the cells will start forming masses: transformed!
sequencing the genome of these transformed cells compared to the non transformed cells showed the oncogene
Ras was the first, and is mutated in 1 of every 5 cancers

a point mutation prevents it from hydrolyzing GTP, so it is hyperactive and never shuts off
dominant effect: only one copy of the Ras gene needs to be mutated this way in order to produce an uncontrolled effect

studying retinoblastoma, eye cancer, helped decode tumor suppressor genes

there were 2 types of eye cancer, one hereditary, one just random. The hereditary version impacted both eyes (so it must be a genomic problem in all the eye cells)
they found a deletion on chromosome 13, where one copy of the Rb protein was missing
it's hereditary because families inherit the one good copy and one bad copy, so as soon as the one good copy is mutated, they get retinoblastoma
the non hereditary is random and usually in one eye because there must be 2 separate events that mutate both copies of Rb in one cell
Rb regulates cell cycle, so mutated Rb allows inappropriate and uncontrolled cell proliferation

usually tumor suppressor genes are identified by screening many cancers and finding out what gene is repeatedly missing or defective
tumor suppressor genes can be inactivated in many ways

chromosome deletion, point mutation, segregation/recombination errors, etc, etc
loss of heterozygosity: where one copy of a cancer-critical gene is lost or deactivated, so now the cell has a predisposition to cancer (being cancer-prone)
epigenetics: methylation of a tumor suppressor gene causes it to be stored in heterochromatin and silenced for all the progeny

proto-oncogenes can be overactivated in many ways

point mutation, deletion, gene amplification, chromosomal rearrangement that ends up altering a gene or the regulatory sequence of the gene

3 methods of cancer-critical gene identification

CGH (comparative genomic hybridization): DNA fragments from normal cells and from cancers cells are fluorescently labelled

the labelled fragments hybridize to complementary sequences on a DNA microarray (where each spot corresponds to a known sequence of the genome)
where the tumor complement binds, it means that many copies of the gene exist (gene amplification)
where the normal complement binds, it means there was a deletion
where there is no light, it means that the two copies of the genome exist and are paired up (blocking any probes from hybridizing)

DNA microarray can also be used with mRNA probes
RNAi: siRNA can be engineered to knockout a specific gene by destroying the mRNA of that gene or inhibiting the translation

then you observe whether the knockout cell becomes cancerous or not

best test involves doing knockouts or causing overexpression of a gene in actual mice

mice that develop tumors soon after prove the possibility of that gene being a cancer-critical gene

so what are the pathways in which these cancer-critical genes actually cause cancer?

cell proliferation pathways: just about any and every signaling protein can be affected, usually with proteins that activate or relay signals becoming hyperactive (proto-oncogenes) -- i.e. Ras, EGF receptor

the inhibitors involved in pathways tend to be the tumor suppressor genes -- i.e. Rb

cancer cells might eliminate Rb, or by inhibiting the inhibitor of the inhibitor of Rb (in times of stress, p16 blocks cyclinD/Cdk4 from inhibiting Rb, who inhibits E2F from continuing the cell cycle, SO, if a mutation inhibits p16, cyclinD/Cdk4 is free to inhibit Rb so that E2F happily progresses the cycle)

cell growth pathways: cancer needs to be able to grow uncontrollably, otherwise uncontrolled division will just lead to smaller and smaller cells

PI3/Akt pathway important for growth, cancers activate this pathway in the absence of growth signals so they just grow all the time

overactive Akt helps with protein synthesis, glucose uptake, and lipid synthesis
so it's usually PTEN phosphatase who is messed up, because he's the inhibitor of Akt

apoptosis pathways: apoptosis usually activates in response to DNA damage and mutations, so because cancers need mutations, they must shut off apoptosis so they can keep proliferating

Bcl2 is a protein that normally inhibits apoptosis

in a B-lymphocyte cancer, the Bcl2 gene is translocated to an active promoter, so Bcl2 is overexpressed
this keeps apoptosis from activating

DNA damage pathways: but really the star player is p53

known as the guardian angel gene, p53 is super critical because it happens to be involved in 2 of the above 3 pathways, and in DNA damage systems

to start, p53 is normally kept at very low levels: continuously made and continuously degraded
when a cell is stressed or the genome is damaged, p53 levels shoot up
active p53 (1) stops any cell dividing to buy some time (2) activates DNA repair systems (3) or activates apoptosis when nothing can be done
this is the protection that prevents cells with just a few mutations from blowing over into cancer
p53 is a gene regulatory protein, it activates expression of other proteins that carry out the above-listed functions
mutation of p53 usually happens in its DNA-binding domains, so it can't bind to genes and activate transcription, and the cell can (1) divide uncontrollably (2) avoid apoptosis (3) have genetic instability w/o DNA repair and (4) ignore stresses like radiation and drug treatment

DNA tumor viruses are the lysogenic viruses, who integrate their genome into the host genome and sit quietly

virus genome contains stuff that take control of reproduction machinery to generate more viruses
accidents can occur that activate some components of the virus genome that activate DNA replication and division, leading to tumor formation
e.g. papillomavirus produces a protein E7 that blocks Rb, and an E6 protein that inhibits p53

Metastasis has 2 steps: invasion and re-establishment

EMT (epithelial to mesenchymal transition): normal organism development process where epithelial cells change from adhesive and stable to less adhesive and migratory

this happens when E-cadherin is lost and cells can't form junctions, so they can go out exploring and getting into blood vessels

colonization: cancer cells must acquire the ability to proliferate in a new environment in order to do this

that's why most cancer cells that escape the primary tumor don't actually form new metastases

CASE STUDY: COLORECTAL CANCER
colon epithelium has rapid turnover rate, so it is constantly growing and dividing (at a controlled rate) to replace old cells

renewal depends on special crypts (pockets) of stem cells in the epithelium
benign tumors look like small protruding masses called polyps, the first sign of abnormal growth
the most common mutations happen to 3 proteins: K-Ras, p53, Apc

Apc is an inhibitor of the Wnt pathway but causing degradation of B-catenin (recall B-catenin can go in to the nucleus to activate transcription)
but tumors tend to form faster with a hyperactive B-catenin rather than an inactive APC
the high frequency of APC mutations comes from APC being involved in other pathways as well and getting mutated affects the mitotic spindle formation, so chromosome segregation is messed up

one form of colon cancer is hereditary, due to a mutation in DNA mismatch repair

the inherited predisposition involves having one messed up copy of the repair gene and one good copy
once a mutation inactivates the one good copy, cells will start building mutations, increasing chances for cancer

tumors often have specific order of mutations

in most cases, Apc is one of the first to go

mutations w/o Apc's guardianship will build up

then Ras falls apart

cells now proliferate w/o substrate or neighbor inhibition

then p53 is messed up

now cells can mutate faster and avoid apoptosis

each tumor will mutate its own variation of similar genes (like messing up the Ras regulator as opposed to Ras itself), but the control mechanisms need to be disabled in a proper order
even different patients with the same type of cancer will have different mutations, which makes cancer treatment very difficult

cancer prevention or early diagnosis of a primary tumor is the best method of avoiding cancer

treatment is difficult because you must remove ALL the bad cells, or else even just one cancer stem cell will regenerate a tumor
surgery makes it difficult to get every tiny metastasis, and drug/chemical treatments also hurt good cells

traditional cancer treatment exploits genetic instability

radiation and drugs kill cancer cells because they can't repair the DNA damage, while normal cells will pause cycling (so you lose hair) but will fix the damage
unfortunately, cancer cells can evolve resistance to treatment (like bacteria!)

new treatments exploits CAUSE of genetic instability

first, consider that there are multiple ways to repair DNA damage

a single strand break is fixed by one system
if that system is down, the stalled replication fork that results when DNA Pol encounters the ssbreak can be fixed by homologous recombination

second, consider that cancer cells tend to lose some of the repair systems in order to become cancer

for example, let's say a cancer cell can't repair a stalled fork, but it can fix ssbreaks

third, consider that we can devise drugs to impair repair systems

in the example with the cancer cell that can't repair stalled forks, what if we give it a drug that inhibits the repair of ssbreaks?
now the tumor can't repair ssbreaks, and it already couldn't repair stalled forks, so it will die....
BUT the normal cells live, because unlike the cancer cell, they CAN repair stalled forks, even if the drug also affects them and inhibits the repair of ssbreaks

these treatments take advantage of the fact that normal cells have multiple options, while cancer cells have very few repair/survival options because of their mutations

cancer cells have still another advantage: genomic heterogeneity

even though they are all descended from the same ancestor, they have such a high rate of mutation that much of the population have different variations in their genome
this means you can't just use one type of treatment to eradicate a whole tumor
multidrug resistance: phenomenon where a cell, after surviving just one drug treatment, somehow is resistant to a bunch of other drugs that it's never met before

this happens when a gene called Mdr1 is amplified (has multiple copies now)
Mdr1 makes a transporter that pumps out drugs, overexpression of the protein makes sure no drugs hang around and damage the cell: they just get pumped out

new therapies are being developed due to our increasing knowledge

targeting the one hyperactive protein that a cancer may be dependent upon (oncogene addiction: the cell is addicted to/dependent on the oncogene)
send in antibodies that either carry toxins or enzymes that cleave pro-drugs into toxins at the cancer site (because antibodies bind so specifically, you use anti-cancer antibodies)

we can also engineer small molecules that target the one hyperactive protein of oncogene addiction

e.g. CML (chronic myelogenous leukemia) is associated with Philadelphia chromosome
this chromosome snaps the genes Abl and Bcr and rejoins them so they fuse into Bcr-Abl
the fusion protein combines the Tyr kinase of Abl and the active expression of Bcr to make a hyperactive kinase
this prevents apoptosis of unnecessary lymphocytes, so they build up in the blood
a special molecule called Gleevec blocks Bcr-Abl, and effectively eliminates CML in early diagnosis patients
less effective with later patients b/c cancers have usually evolved resistance to Gleevec binding

probably going to have to do a cocktail to fully eradicate cancerous cells

another approach is to target the tumor blood vessels

there are drugs that inhibit the VEGF receptor
also, endothelial cells that are about to make new blood vessels have distinguishable characteristics that we can engineer molecules to attack them specifically

the final method is to enhance the existing immune system to target tumors

we can tag cancer cells with antibodies to attract macrophages to gobble them up

the future of treatment is probably going to involve a simultaneous combinations of drugs and molecules, because no cell will have resistance to all of them without some time to evolve (i.e. if we used the drugs one by one, every time, there would always be some cells resistant the single drug used)

pg. 1230 - 1240

Book Notes: Cancer Part 1

normal cells work cooperatively, and obey signals and controls on division and growth so the organism can survive

cancer cells work "selfishly", growing and dividing out of control, nabbing up nutrients and creating blood vessels so they can survive
know the difference between BENIGN and MALIGNANT (is it invasive)

several types of cancer by tissue of origin

carcinoma: from epithelium, most common because epithelium is usually exposed to stresses and dangers (serves as protective barrier, making it vulnerable to damage and mutation)
sarcoma: from muscle/connective tissue
leukemia/lymphoma: from white blood cells, lymphocytes, and nerve cells

cancer cells tend to be all clones of the primary tumor

patients with CML, a leukemia with the mutation Philadelphia chromosome, show that the chromosome break happens at exactly the same place for all cells in the tumor, but it happens a few hundred basepairs different between patients
this implies cells of one tumor all derive from one ancestral cancerous cell and are all clones

cancer results from a genetic mutation, which may be from chemical carcinogens or radiation/UV light

however, you need more than one mutation in ONE CELL LINE to have cancer
organisms have a high overall rate of mutation, but that rate is including all the cells of a human (there are several trillion cells in the body, a human has about several trillion mutations over the lifetime, so it's like maybe one mutation per cell per day, and remember that repair machinery usually fixes the mutation)
cancer usually happens in old age because that's when you've likely built up a lot of mutations and the tumor has had some time to grow into a visible mass

e.g. leukemia didn't show up until 5 years after the atomic bombs razed Hiroshima, and lung cancer takes 20 years of heavy smoking before surprise!

rule of thumb: minimum 5 mutations to become cancerous
creating a mutant cancerous cell takes microevolution

mutation rate: how fast does this kind of cell get mutations?
number of reproducing individuals: how many cells in this tissue divide at a time?
rate of reproduction: how fast can a cell divide and how many times can it divide?
selective advantage: does the mutation kill the cell or give it a bonus or do nothing?
you need several rounds of division (epithelial cells divide constantly) with advantageous mutations carrying through each step until you shake loose the body's regulatory systems

cancer also involves epigenetic mutations

epigenetics involving histone modification is a way of controlling which genes are active and which are silent
mutations happen with the enzymes that modify histones, the proteins that interpret the histone code, etc et
these mutations are also passed on to daughter cells

because cancer cells proliferate uncontrollably, they are also genetically unstable

it's usually because the cancers have a mutation in DNA repair/maintenance genes
the increase in mutation rate plus their rapid division rate makes their evolution rate really fast--cancers with genetic instability speed rapidly towards malignancy
cancer cells can also accumulate bad mutations that end up killing it, so it is the cancer cell with the right mutations that will persist into tumors and metastases

the core mutations of cancer cells tend to be in control of cell death, control of cell differentiation, or both (for optimal cancer growth :D )

any mutation or external condition that promotes cell growth can and will help cancerous cells
normal tissue growth controlled by apoptosis, especially when apoptosis serves as a control to destroy mutated cells

cells that don't do apoptosis when they're supposed to keep growing and passing on mutations, potentially forming cancer

stem cells will produce daughter cells that proliferate for a bit before committing to differentiation

if the differentiation step is blocked, daughter cells just keep dividing and dividing

other core mutations are DNA damage response and other stress responses

the DNA damage response goes hand in hand with the cell cycle checkpoints: mutations in either system won't be able to stop a damaged cell from continuing to divide and pass on the mutations
being able to survive through these mutations makes them more undefeatable

the final barrier to cancers is replicative cell senescence, when the telomeres run out, the cell would normally do apoptosis

cancers can avoid the chkpt when telomeres run out and just keep dividing without telomeres (meaning every round shrinks the chromosome by a bit)
cancers can also have a mutation that reactivates telomerase or a mutation that creates something similar to telomerase

new theory suggests tumors are organized with cancer stem cells at the top and limited dividing lower cancerous daughter cells

if the regular cancer daughter cell is implanted in a mouse, it can't generate a new tumor because it has limited dividing capacity
only a small percent of cells in a tumor (<1) can propagate indefinitely

where do cancer stem cells come from?

(1) from real stem cells and (2) from normal proliferating cells that developed a mutation that makes them propagate indefinitely like a stem cell
cancer stem cells divide slowly, so treatments that target rapidly dividing cells won't harm them, and then the tumor mass will regrow

metastasizing cancer is the most dangerous, but it requires cancer cells to overcome some barriers first

1st, the tumor cells must invade neighboring normal tissue and keep spreading
2nd, the tumor cells must find a blood or lymphatic vessel and get in
3rd, the tumor cells must grab onto a new site while floating through the vessel and form a small clump (micrometastases)
4th, the clump must develop into a large tumor for stability
many cells are usually able to get into the vessel, but very few can attach and colonize in a new location

even forming micrometastases is no guarantee of continued survival
metastasis usually requires a LOT of mutations in all the right places

a large tumor also needs a supply of nutrients

a tumor over 1-2 mm will need to induce angiogenesis: formation of new blood vessel
normal and tumor cells secrete angiogenic signals in response to hypoxia (lack of nutrients and oxygen)
these signals activate transcription and secretion of VEGF
VEGF attract endothelial cells and stimulate growth of a blood vessel
induced vessels are usually disorganized and go random places with dead ends

this leaves a lot of area still in hypoxia, which causes natural selection for cells that survive in tough conditions, which makes cancer even tougher to eradicate

stroma: the surrounding connective tissue, even tumor cells talk with them like normal cells

stroma contains fibroblasts, white blood cells, etc, for support
cancer cells send signals to the stroma
the stroma responds with signals that stimulate growth and secrete proteases to loosen the ECM for invasion

tumor and stroma evolve together (experiment where a tumor is plucked out and put next to normal fibroblasts showed the tumor can't survive)
possible source of treatment: inhibit deranged stroma to kill tumors

sum up of characteristic cancer behaviors:

survive and proliferate in weird conditions (not attached to substrate, etc, etc)
insensitive to anti-proliferation signals
avoid apoptosis
avoid stress and damage responses
induce help from stroma
induce angiogenesis
invade and proliferate far far away
genetically unstable
have stable telomeres

pg. 1205 - 1223

Monday, May 16, 2011

Book Notes: Signaling Part 5

pathways involving gene regulatory proteins (who activate gene expression) depend on regulated proteolysis to control them

Notch, Wnt, and Hedgehog are all these pathways, and they are especially involved in animal development

Notch is famous for affecting neural cell development

lateral inhibition: if one cell develops into a neuron, it signals to its neighbors to NOT being nerve cells and instead become epidermal
this occurs because the developing neuron displays the Delta signal protein on its surface
neighbors all have Notch receptors that bind Delta and receive the signal to develop into epidermis
Notch signaling can also be used for promoting neighbors to become the same cell type in other cases

when Notch binds Delta, a membrane-bound protease cleaves Notch's tail

the tail zooms off into the nucleus and activates a set of genes by binding and converting a transcriptional repressor into an activator
Notch is actually cleaved 3 times in its lifetime: (1) during synthesis, it is cleaved by Golgi to form heterodimer, (2) during ligand-binding, when the extracellular head is cleave off, and (3) after ligand-binding, when the cytoplasmic tail is cleaved off
cleavage is irreversible, meaning Notch is a one-time use receptor

Notch and Delta are glycoproteins, and their specific match-up depends on how their sugar sidechains are modified (glycosylation is important!)
Wnt is famous for its versatility: it plays a role in catenins (i.e. jxns), cell polarity, etc, etc

Wnt has 3 well-known pathways:

Wnt/beta-catenin pathway: Wnt affects the gene regulatory protein beta-catenin
planar polarity pathway: coordinate polarization in plane of epithelium
Wnt/Ca2+ pathway: stimulates increase of calcium level

Wnt is a secreted molecule that binds to Frizzled receptors

Wnt-bound Frizzled recruits Dishevelled, which then relays the signal down one of the 3 pathways

for the beta-catenin pathway, Wnt binds to both Frizzled and an LRP (LDL-receptor-related protein)

any beta-catenins not involved in jxn-forming is bound by degradation complex, which keeps it away from the nucleus and promotes its degradation
2 kinases (GSK3 and CKIgamma) phosphorylate C tail of LRP, which inactivates axin (part of the degradation complex), so free beta-catenin accumulates and migrates to the nucleus

without Wnt, the set of genes controlled by Wnt are silent

LEF1/TCF proteins bound to Groucho (a co-repressor)
when Wnt activates, beta-catenin comes in and displaces Groucho as the co-partner
Beta-catenin serves as a co-ACTIVATOR for the genes
c-Myc is one of the proteins expressed by Wnt signaling, and it stimulates growth and proliferation (potential cancer!!!)

Hedgehog is similar to Wnt: is secreted, modified by lipids, trigger switch from repression to activation, and leads to growth/proliferation response

3 transmembrane proteins mediate responses to the Hedgehog proteins:

Patched: multipass with some of it in intracellular vesicles, and some on the cell surface to bind Hedgehog
Smoothened: 7pass transmembrane
iHog: 4-5 immunoglobulin domains and 2-3 fibronectin domains on cell surface, serve as co-receptors with Patched
Patched normally sequesters Smoothened
Hedgehog binding to iHog partnered with Patched inhibits Patched by causing endocytosis and degradation of Patched

result is that Smoothened is free and phosphorylated and continues the signaling pathway

Cubitus interruptus (Ci): gene regulatory protein that is a target of Hedgehog

in normal conditions, Ci is proteolysed into smaller proteins that hang in the nucleus as repressors
this proteolysis is dependent on phosphorylation by PKA and GSK3 and CK1 (all Ser/Thr kinases)
when Hedgehog binds, Smoothened runs free and recruits Costal2, a scaffold protein, and Fused, another Ser/Thr kinase

Costal2 normally holds the proteolytic kinases together to process Ci, but binding to Smoothened blocks this ability

uncleaved Ci also goes to the nucleus, but acts as an activator
uncleaved Ci can also express the Patched gene = more Patched inhibits further Hedgehog signaling

pg. 946 - 952

Book Notes: Signaling Part 4

G proteins do more than play with cAMP and Ca2+

the a subunit of G12 activates GEF for the Rho GTPase
others bind and open/close ion channels to change membrane potential
others phosphorylate channels or change levels of cNMP (e.g. cAMP) that affect channels

olfactory receptors utilize cAMP-gated ion channels

odor particle binds to olfactory receptor
the receptor activates the G protein Golf (G subscript olf for olfactory)
Golf activates adenylyl cyclase = lots of cAMP
cAMP binds to cation channels that allow Na+ to flood in
neuron containing the olfactory receptor is depolarized and sends a potential down the axon to the brain
each neuron makes only 1 type of receptor that senses only a small set of odors

visual receptors utilize cGMP-gated ion channels

rod photoreceptors are cells with a rodlike outer and inner segment, followed by the body, followed by the axon and synapse
the outer segment looks like a pancake stack with lots of rhodopsin molecules in them
the plasma membrane coating the stacks have lots of cGMP-gated cation channels
in the dark, cGMP remains bound to the channels, keeping them open
in the light, the cis chromophore in the rhodopsin absorbs a photon and isomerizes into trans
the changed chromophore causes the protein opsin to change conformation and activates the G protein transducin (Gt)
transducin a subunit activates cGMP PDE, which removes cGMP

visual system resets itself VERY QUICKLY so that we can see the change in light in the next instant

rhodopsin kinase phosphorylates the tail of activated rhodopsin, to prevent further activation of Gt
arrestin binds to Pi-rhodopsin to fully inhibit rhodopsin activity
RGS acts as GAP to make Gt hydrolyze the GTP
light also causes calcium channels to close, so calcium levels fall, so guanylyl cyclase hurries to make more cGMP, returning to normal "dark" mode
negative feedback allows system to adapt to continuous light, and be sensitive to CHANGES in light (seeing camera flash in broad daylight)

all the steps of intracellular signaling pathways are potential SIGNAL AMPLIFICATION steps

but for rapid sensing, cells must be able to both amplify a signal quickly and destroy the response just as quickly
therefore, any and every protein/molecule in the pathway can be a target for regulation

GPCRs are usually desensitized in 3 ways

receptor inactivation: receptor interaction with G protein is blocked

e.g. GRK (GPCR kinase) like RK that phosphorylates GPCR after the receptor has been activated by ligand-binding
then (as with rhodopsin) an inhibitor like arrestin binds and blocks the G protein interaction

receptor sequestration: move receptor inside the cell to block access to ligand signal

arrestin can also couple the receptor to endocytosis machinery and clathrin
later dephosphorylation returns the receptor to the surface

receptor down-regulation: receptor is destroyed post-activation

endocytosis of receptor in some cases lead to the receptor ubiquitylation and lysosomal degradation

EZCR (enzyme-coupled receptor): transmembrane receptor with a cytosolic domain that is an enzyme or associates with another enzyme (note, the abbreviation is my own invention)

exist 6 classes of EZCRs

receptor Tyr kinase: phosphorylate Tyr residues on itself and some signaling proteins
Tyr kinase associated receptor: recruit cytoplasmic Tyr kinase
receptor Ser/Thr kinase: phosphorylate Ser/Thr on itself and some gene regulatory proteins
His kinase associated receptor: phosphorylate itself on His and transfers the Pi to a signaling protein
receptor guanylyl cyclase: catalyzes production of cGMP
receptorlike Tyr phosphatase: remove Pi from Tyr of signaling proteins (ligands are unknown)

RTK (receptor Tyr kinase): bind hormones like insulin, growth factors, and ephrins (cell-surface ligands as opposed to free floating signals)

ephrins and EPHrin receptors both make responses in the receptor cell and the ligand cell: "bidirectional signaling"
ligand binding causes RTKs to dimerize and cross-phosphorylate each other
the phosphorylated Tyr residues serve as docking sites for specific intracellular signaling proteins
these signaling proteins are either activated by docking or by receiving phosphorylation from the RTK
proteins dock by modular interaction domains (remember SH2, SH3, etc etc), sites of protein protein binding that don't affect function

examples of docked proteins are PI3-kinase and PLCgamma (PLCbeta associates with GPCRs)
adaptor proteins with lots of modular domains connect docked proteins to other proteins that DON'T have these awesome domains

example being RAS

Ras and Rho families are the only monomeric GTPase families to relay signals from surface receptors

as a family, they can coordinate one signal to many pathways
Ras has lipid groups to keep it anchored to the cytoplasmic leaflet of the membrane

RTKs activate Ras through adaptors and Ras-GEF

case study: fly eye is composed of ommatidia, which are sets of 8 photoreceptor cells and 12 accessory cells
the Sevenless (Sev) mutant gene causes flys to be blind in UV light, because mutant Sev prevents normal development of the R7 photoreceptor
BUT it turns out for some blind flies, the Sev gene is kept normal: it's the protein Bride of Sevenless (BOSS) who's mutant
Boss is a 7pass transmembrane ligand on the R8 cell, and binds to the R7 Sev RTK (see, Sevenless is an RTK, a receptor. BOSS is the ligand. That's why it's "bride of Sevenless," because it binds to the Sevenless receptor)
when they bind, the R7 precursor cell is induced to develop into R7
even if other cells express Sev (which they do), only R7 touches R8, so only the Sev on R7 receives the ligand
Steps to signal transduction

Boss binds Sev-RTK
Sev-RTK is phosphorylated and an adaptor DRK with an SH2 domain docks
a special Ras-GEF called Son of Sevenless (SOS) is also bound by DRK's 2 SH3 domains
SOS makes membrane-bound Ras get a new GTP and the activated Ras continues the signal with cytoplasmic free proteins

see, SOS is Son of Sevenless because it's activated when Bride of Sevenless binds to Sevenless receptor = love child (haha, biologists are both bored and lewd)

Now that Ras is active, where do we go next?

Ras-MAP-kinase signaling pathway: Ras activates MAPKKK (Raf), who activates MAPKK (Mek), who activates MAPK (Erk), who activates gene regulatory proteins in the nucleus
these gene regulatory proteins quickly turn on "early genes," who are part of the primary response
early genes produce more gene regulatory proteins that go to activate later genes for the secondary response
in this way, there is a rapid response and a more delayed response
this pathway is how the signal goes from surface (where RTKs and Ras await) to the nucleus (where DNA and genes hide)
lots of positive and negative feedback: MAP kinases activate both downstream targets and its own activators, BUT they also increases transcription of phosphatases and deactivate its own activators (Erk blocks Raf)

different sets of 3 MAP kinases form MAP modules in different pathways

however, many modules share one or more of the same kinases
scaffolds prevent cross-talk by associating with the sensor, so that activated kinases that bind to the scaffold are automatically next to its downstream targets also docked to the scaffold protein

this makes sure the activated kinase doesn't wander off and find other targets of other pathways

Rho GTPases regulate the cytoskeleton (actin and microtubules) in response to special guidance receptors

inactive Rho is bound to GDI, which prevents association with the GEF
Eph RTKs are the receptors that activate surface-bound GEFs that activate Rho
ephrins are the guidance ligands that bind Eph RTK

e.g. most commonly used for directing axonal growth
target cells express the appropriate ephrin on its surface
the axon's growth cone has Eph RTK that binds ephrin
Rho changes the cytoskeleton to grow in the direction of the bound ephrin to shift the axon thataway

PI3-kinase is another protein that docks to RTK

PI3 phosphorylates PIs, usually to make PIP3
PIP3 floats around until the PTEN phosphatase turns it into PIP2
PIP2 is nabbed by PLCbeta or PLCgamma and cleaved into IP3 and DAG for signaling
RTKs activate PI3-kinase to make PIP3 available to make PIP2 available for signaling
PIP3 have special PH modular domains (pleckstrin homology) that are super specific (because PIP3 and PI3 tend to be involved in growth and related pathways that are vulnerable to become cancerous)

PI3-kinase-Akt pathway receives signals from IGF, the "SURVIVE AND GROW" signal

RTK activates PI3 kinase, which makes PIP3, which recruits Akt and PDK1 (phosphoinositide-dependent protein kinase 1) to the plasma membrane
Akt is activated there and phosphorylates many targets, producing a growth response

e.g. one target is called Bad, which causes apoptosis, but phosphorylation of Bad makes it dock onto a scaffold protein, preventing its function

a special TOR protein promotes ribosome production, protein synthesis, and nutrient uptake/metabolism, so it is activated by the Akt pathway for growth purposes

pg. 916 - 935

Sunday, May 15, 2011

Book Notes: Signaling Part 3

GPCRs all are from the same family and share characteristics

sevenpass transmembrane receptors that associate with G proteins
examples include rhodopsin and olfactory receptors

signal binds GPCR, GPCR goes through conformational change, activates a bound trimeric G protein, who relays the signal to the rest of the cell

each GPCR has its own set of G proteins
the G protein is either normally bound to the receptor or binds after ligand binds to receptor

G protein has alpha subunit, beta subunit, and gamma subunit (beta and gamma usually go together as the beta-gamma complex)
in normal state, a is bound to a GDP

the activated receptor acts like GEF and makes a drop the GDP and get a GTP
the G protein changes conformation and the subunits can now act on enzymes or ion channels in the plasma membrane to continue the signaling cascade
when a hydrolyzes the GTP (usually with help from a regulator of g-protein signaling-RGS protein), it becomes inactive

adenylyl cyclase is one possible target of the G protein

it makes cAMP from ATP and is membrane-bound
change of [cAMP] is resisted by cAMP phosphodiesterases (PDEs)
GPCRs can activate either Gs or Gi to stimulate or inhibit the cyclase
2 kinds of toxins target this pathway:

cholera: bacteria makes enzyme that adds an ADP ribose to the a of a Gs protein, so it can't hydrolyze GTP, so it is continuously on: continuous adenylyl cyclase = continuous high [cAMP] = diarrhea in intestinal cells
pertussis: bacteria makes enzyme that adds and ADP ribose to the a of a Gi protein, so it can't interact with GPCR to receive a new GTP, so it is continuously off: continuous adenylyl cyclase = continuous high [cAMP] = fluid flooding the lungs

so with cAMP now present in high concentration, where does the signal go next?

cAMP activates PKAs (cAMP-dependent protein kinase), a Ser/Thr kinase
PKA normally exists as 2 regulatory subunits bound to 2 catalytic subunits
4 cAMP binding simultaneously to PKA causes the r subunits to release the c subunits, which go to do the phosphorylation
one of PKA's targets is PDE, which lowers the cAMP, quickly shutting down the signal to a brief local pulse rather than a weak extended signal (like a quick mosquito bite as opposed to the annoying buzz buzz buzzing in your ear)

A-kinase: another name for the regulatory subunits, help localize PKAs to the right place

AKAP (A-kinase anchoring protein): an adaptor that tethers the A kinase (and bound c subunits if not activated) to cytoskeleton or organelle membrane or other signaling proteins to make a complex

so PKA is active, what happens next?

one option is activation of gene expression

target gene has a special sequence in the regulatory region of the gene: CRE (cAMP response element)
the CREB protein recognizes and binds to this sequence
PKA phosphorylates CREB
active CREB recruits CBP (CREB binding protein), who stimulates transcription where the CREB is sitting (in that way, the Cre gene is specified by the CREB)

G proteins, in addition to adenylyl cyclase, also act on PLC (phospholipase C)

PLC is activated by the Gq protein, and then cleaves PIP2 into IP3 and DAG
IP3 is a water-soluble molecule and goes into the cytoplasm

binds to IP3 receptors at the ER, which are calcium channels
the channels open and calcium ions flood the cytoplasm
calcium flood is later reverted by the opening of other calcium channels in the ER and by calcium pumps that expel the ions into the ECM

DAG is not water-soluble and remains in the membrane, where it is further cleaved into arachidonic acid and later becomes eicosanoids-small lipid signaling molecules (e.g. prostaglandin involved in pain/inflammation)

DAG also activates PKC: PKC 1st senses the calcium flood and moves to the plasma membrane, where it binds DAG
triple binding of PKC to calcium, DAG, and phosphatidylserine on the membrane fully activates PKC

calcium is a universal signal because it CAN cause a sudden flood

this happens because intracellular [Ca] is kept very very very low, while [Ca] in ER and ECM is very very very high
as soon as channels open, calcium ions zoom in
in normal conditions, ATP pumps and antiporters in the plasma membrane and ER membrane expel or suck up calcium out of the cytoplasm

calcium signals tend to occur in multiple spikes rather than a continuous period of high concentration, because there are so many pumps and other calcium binding proteins

also, calcium causes its own positive feedback, but too much calcium causes its own negative feedback, producing the oscillation of calcium spikes
a strong signal produces rapid oscillations, while a weak signal = low frequency of oscillations

therefore, some calcium sensitive proteins are actually calcium-frequency sensitive: low frequency activates one set of genes, while high frequency activates another

calmodulin: calcium binding protein that relays the calcium signal

2 calcium ions bind calmodulin, causing a conformational change
the Ca2+/calmodulin complex goes and binds other target molecules, undergoing a second conformational change

target molecules include the pumps that get rid of calcium

one important target is CaMk (Ca2+/calmodulin-dependent kinase): CaMk phosphorylates itself and other target molecules

the autophosphorylation enables CaMk to remain active even when calcium goes away, until phosphatases come by

CaMk is also the protein that is sensitive to calcium frequency!

low frequency calcium allows time for the CaMk to lose activity and go back to normal before the next pulse
high frequency prevents CaMk levels from falling back to zero before the next pulse, so the total CaMk activity goes up, falls a little bit, then goes up even more, falls a little bit, and goes up even even more
also, CaMk is a multisubunit protein, so a cell can make CaMk's with different subunits to do different things at different frequencies

pg. 903 - 916 (from Chapter 15)

Book Notes: Signaling Part 2

even prior to multicellular organisms, it was important for unicellular things to sense signals from each other

quorum sensing: receive chemical signals from neighbor bacteria to coordinate motility, antibiotic production, and conjugation (bacteria sex)
yeast cells secrete mating factor to get neighbors to stop cloning and make a haploid cell for sexual fusion (communication is important in a relationship!)

general multicellular signaling process

signaling molecule released into ECM (amino acid, peptide, protein, steroid, fatty acid, dissolved gas, etc, etc)
target cell binds the signal with a receptor protein on its surface
binding causes cytoplasmic domain of receptor to activate or act on other proteins and molecules within the cell
intracellular signaling pathway involves many molecules activating/binding to each other
effector proteins are expressed/activated to cause the effect that the signal was going for

2 types of short distance signaling:

contact-dependent: cells must be touching each other (i.e. signal molecule is attached to the source cell, and the receptor is attached to the target cell, so signal-receptor binding can only happen if the two cells are right next to each other)
paracrine signaling: secreted signal only hangs around in the ECM near the source cell, affecting only locals

how do you keep the signal from flowing too far away? The neighbor target cells take up the signal or ECM enzymes destroy the signal, or the ECM itself traps the signal like superglue
autocrine signaling: similar to paracrine, except secreted signal only affects itself and other neighbors of the same cell type

2 types of long distance signaling:

synaptic signaling: neurons emit chemical signals at a synapse right next to the target cel (however, the axon can extend literally from your head to your toes!)
endocrine signaling: source cells spit out hormones who travel by bloodstream to find target cells all over the body (usually one kind of hormone causes different effects to different cell types for a coordinated response between multiple tissues and organ systems)
compare & contrast: synaptic is targeted precisely and super fast (electric zap!), while endocrine is targeted generally and pretty slow (diffusion and blood flow)

in general, response time also depends on effect

causing an effect with existing proteins is fast
causing an effect that requires synthesis/gene expression is slow

remember gap junctions? They help with "the most intimate of all forms of cell communication"

they equalize cytoplasmic conditions between connected cells
they allow sharing of second messengers, especially important where some cells of a tissue cannot be reached by nerve impulses, so the signal is passed on through sharing second messengers (cAMP, calcium ions, etc, etc)

cells, based on cell type, respond to specific combos of signals (if every single signal molecule meant something different, cells would get so confused if they happened to receive both a DEATH signal and a KEEP LIVING signal)

combinations prevent confusion and accidental wrong signals (maybe you sometimes get half the molecules necessary to spell DEATH but you always get all the letters of KEEP LIVING)

different cells types produce different response to the same signal molecule

response based on receptor type, intracellular signaling molecules (they serve as the signal interpreter), and which effectors are activated

what happens when the signal is removed?

in some cases, the changes caused by the signal are permanent, so removing the signal does nothing
in other cases, the changes caused by the signal are temporary, so removing the signal removes the effect it caused

how does this work? usually by having second messengers/proteins that are quickly destroyed (very short half lives)
having a continuous signal causes continuous production to maintain concentration of the second messenger/protein
no signals means no production means the existing messengers get degraded quickly and the cell stops producing the response

some signals skip the outside receptor and go directly into the cell to cause response

gases like nitric oxide pass through the plasma membrane and bind to proteins to cause effects (is short-lived and local, through, because it can only diffuse)

nerve signals endothelial cells to make NO, NO diffuses out and into neighbor cells, binds guanylyl cyclase to make cGMP that relaxes muscle to increase blood flow

steroid hormones also pass through the plasma membrane: they bind to intracellular nuclear receptors

the nuclear receptors are either free-floating in the cytoplasms or already bound to DNA
normally the receptors are bound by inhibitors, and binding of the steroid causes release of the inhibitors
the receptors then go ahead and activate DNA transcription (all receptors bind as dimers)
some receptors are not inhibited normally, but get inhibited by the steroid
these receptors cause the primary response: transcription of a few genes
the products of those genes go and activate transcription of a lot of effector genes to make the delayed secondary response

many cell types use the same receptors: it's the combination of other molecules who join forces with the receptor to cause different effects (for example, activation of gene transcription requires a whole slew of factors, activators, removal of repressors, etc, etc)

most signals meet up with surface receptors, who act as signal transducers (because they convert outside signal to internal signaling pathway without actually going in)
3 main types of surface receptors:

ion-channel-coupled receptors (ICCR) : rapid signaling, usually with synapses

signal attaching to receptor causes ion channels to open, which alter the concentration of ions on either side of the membrane, depolarizing the membrane and passing on the electrical signal

G-protein-coupled receptors (GPCR) : acts by regulating a neighbor membrane-bound trimeric GTP-binding protein (G protein)

the G protein then activates a free target protein that initiates the signaling pathway
all GPCRs are multipass transmembrane proteins of one superfamily

enzyme-coupled receptors (EZCR) : are enzymes of target molecules and/or activators of other enzymes

usually single pass transmembrane, kinases or kinase activators

signals are usually relayed from cell surface to internal organelles/proteins/genes by both small secondary messengers (mediators) and large intracellular signaling proteins

they are small molecules, normally not present or only in low concentrations, that get mass-produced in response to signal
their high concentration allows rapid diffusion throughout cell (or membrane, in the case of diacylglycerol)
they bind to specific signaling and effector proteins to pass on the message by altering function and/or conformation in the molecules they bind to

the large signaling proteins have multiple functions

receive signal and pass it on
serve as scaffold for multiple signaling guys to dock and efficiently trade messages
transduce signal (i.e. from kinase cascade to secondary messenger diffusion)
amplify signal (produce lots of messengers or activate lots of downstream proteins)
integrate/receive 2 signals and pass only 1 signal
spread to other pathways (activate proteins in this pathway and other pathways)
anchor proteins to a specific location to attract molecules or create a localized response
regulate other signaling proteins (regulates signal strength within cell)

usually these large signaling proteins are like ON-OFF switches: activated/deactivated with phosphorylation and/or GTP hydrolysis

with phosphorylation, a protein is active or inactive depending on balance of kinase and phosphatase activities (kinase and phosphotases are also coordinated to either be all kinase active at one time or all phosphatase active at one time)

phosphorylation cascade: kinase activates a kinase that activates a kinase that activates a kinase and so on.
kinases phosphorylate either serine/threonine or tyrosine residues

with GTP binding domains, they turn on with GEF making them get GTP and turn off when GAP forces them to hydrolyze it into GDP

because so many proteins are also part of other pathways, how can a cell avoid cross talk (activation of another pathway through proteins from a signaled pathway)?

form complexes of all the correct signaling molecules on pre-formed scaffolds or temporarily on the tail of a receptor with multiple docking sites
having all the right ones in close proximity makes signaling fast, efficient, and specific

the complex can be induced to form by signals and by interaction domains

the domains are small binding pockets and complementary motifs on the receiving protein that don't interfere with shape or function
examples are SH2, SH3, PTB, and PH domains
complexing can also be structurally important: the way the complex is built can form a protein "trail" that leads to a specific location
adaptor proteins are important in complexing (protein whose job is to connect protein A and protein B)

cell response can either be ON/OFF (all-or-none) or gradual increase/decrease based on signal concentration (gradient)

a lot of gradient response can look ON/OFF if the concentration is high enough

e.g. 4 cAMP binds to PKA in order to activate it: a lot of cAMP will suddenly turn on all the PKA in the cell, a slowly growing concentration of cAMP will gradually turn on the PKA one by one as they grab up 4 (the 4 must bind simultaneously)
the slope of the graded response looks steeper (more on-off) as the number of required cooperative molecules increase

another way a graded response appears ON-OFF is if enzyme A that activates enzyme B also deactivates the inhibitor of enzyme B
a true all-or-none response needs a positive feedback loop

enzyme A activates enzyme B, who comes back around to further activate enzyme A (as well as doing its other functions)
because 1 enzyme A can activate, say, 15 enzyme B, all 15 enzyme B can come back and activate 15 enzyme A's each (totaling up to 225+1 enzyme A's) who can now activate 15 more enzyme B's each (totaling up to 3390 + 15 enzyme B's).
essentially, a small response very quickly EXPLODES into a huge response
even if the original signal fades, the response can keep going until another signal comes by to stop it (through another pathway)
this form of signal memory can be passed on to daughter cells (epigenetics)

negative feedback can provide 2 different effects:

a quick acting negative feedback shuts down a response as soon as the signal is gone and the cell goes back to normal (adaptive)
a slow acting negative feedback doesn't shut down a response until later, when the signal might reappear again, causing the cell to oscillate (go back and forth) between response and no response (some oscillating pathways continue without new signal)

desensitization: even if signal sticks around for a while, cell returns back to normal by not being able to continue sensing the signal

prevents uncontrolled response (i.e. EGF causing growth continuously)
cells sense change in signal concentration, not presence of signal
five ways of desensitizing:

ligand binding causes response and causes endocytosis of the receptor, so cell no longer has receptors to sense signal until receptors get returned by exocytosis
endocytosis of receptor leads to destruction of receptor (even longer delay until cell can re-synthesize receptors)
activated receptors get quickly modified (i.e. phosphorylated) and become inactive
downstream signaling protein is inhibited: so active receptor doesn't lead to response because the signal is interrupted somewhere along the line
part of the response is to synthesize an inhibitor of signaling protein

pg. 879 - 903 (from Chapter 15)

Book Notes: Signaling Part 1

allosteric enzymes have multiple sites

active: binds substrate and performs reaction
regulatory: binds regulatory molecule (activator or inhibitor) and alters active site to modify function
active sites and regulatory sites are often very far away, how can one affect the other?

binding of regulatory molecule causes conformational change of entire protein, which may "open" or "close" the active site

coupled binding: essentially binding of substrate at one site affects binding at another site

affinity coupling: binding at one site increases binding affinity at the other site (I bind, you bind, we all bind!)
reciprocal coupling: binding at one site decreases affinity at the other site (I bind, you don't bind, *insert evil laughter*)

cooperative allosteric transition: swapping a collection of proteins from active to inactive (or vice versa) is made faster and more efficient when you group the proteins into SYMMETRIC gangs

binding of a few inhibitor upsets the symmetry, and the whole system wants to go back to symmetry (a symmetrical arrangement is energetically favorable), so all the other proteins in the group quickly grab inhibitors
they go back to being symmetric, but are now all turned off

proteins are also regulated by phosphorylation

attaching a phosphate gives the protein 2 more negative charges, which can change the protein shape by attracting a positive cluster somewhere along the polypeptide
the extra phosphate can also be (1) blocking an active site (2) serving as part of the active site to bind the substrate or (3) serving as the complement of another protein's binding site--i.e. the substrate protein is phosphorylated, enabling the enzyme to bind to the phosphorylated substrate
the reversibility of phosphorylation makes it a very versatile form of protein control
kinases: attach phosphate
phosphatases: remove an attached phosphate

GTPases are regulated by their own activity

its own hydrolytic degradation of GTP to GDP turns off the GTPase, but the exchange of a used GDP for a new GTP turns on the GTPase
GAPs help the hydrolytic process, while GEF helps the GDP release process so the GTPase can grab a new GTP

a small chemical change can generate large protein movements (to create the new conformation)

e.g. EF-TU protein shifts about a tenth of a nm at the GTP binding site when dephosphorylated

a switch helix nearby can no longer bind and swings loose
the domains of the protein are connected on a hinge, and the switch helix is like the deadbolt: when the lock is released, the "door" swings open, so the two domains open up about 4 nm, releasing the substrate (in EF-TU's case, it's the tRNA with an A.A. bound)

pg. 171 - 181 (from Chapter 3)

Sunday, April 17, 2011

Cell Bio Exam 3 Study Guide

complete: https://docs.google.com/leaf?id=0B99-sSwVe231NTE5MjExMmMtNDY2YS00NjUwLWJkNzItM2UxYWM1ZDBhZmYz&hl=en

http://dl.dropbox.com/u/22080433/cell%20bio%20exam%203%20study%20guide.pdf

Apoptosis Pathway

Saturday, April 16, 2011

Book Notes: ECM Part 3

integrins: transmembrane receptors that connect cells to the ECM

external conditions and internal signals are connected and transduced by the integrin
integrin binds tightly in presence of tension and loosely without
general shape is 2 subunits alpha and beta that span the membrane (short cytosolic C' and long extracellular N')

cytosolic portions attach to actin filmanets through talin (binds to beta) and vinculin (similar to adherens jxns)
extracellular portions attach to specific sequences of laminin and fibronectin or things hanging off of other cells

cell-cell jxns connect cytoskeleton and signals between cells, while cell-matrix jxns connect cytoskeleton to matrix skeleton and signals
focal adhesions: sites of cell-matrix adhesion

integrins swap between actin and inactive conformation (migrating macrophages have to alternate between binding and releasing in order to crawl form binding site to binding site)

"outside-in activation"

integrin is tightly folded in absence of ligand: when ligand partially associates, the integrin unfolds and ligand binding affinity increases
when integrin is folded, the intracellular domain is also tightly closed (alpha and beta bound together)
when integrin unfolds, the alpha and beta chain separate, exposing the talin binding site
sum up: grabbing ligand causes connection to internal cytoskeleton to form so that force can be applied for movement (it's like if your arm was separated at your elbow, but if your hands grabbed a dumbbell, your arm connects again at the elbow, and then you can use your biceps to lift the dumbbell, eh?)

"inside-out activation"

other signals activate signal cascade, producing PIP2
PIP2 activates talin to bind to beta integrin
extracellular unfolds, ready to find a ligand
sum up: connection at the internal cytoskeleton readies the cell to grab a ligand for next action (it's like if you spot a dumbbell, you hook up your arm to your elbow in preparation, then grab the dumbbell and start exercising :D)

integrins tend to cluster to form strong adhesions

their "high-affinity binding" conformation is relatively low affinity anyways
clustering many integrins is like combining weak noncovalent interactions: strength in numbers

ECM controls cell proliferation and survival through the integrins

anchorage dependence: cells depend on ECM connection to decide to grow, divide, or even survive

muscle, epithelial, and endothelial cells commit suicide if they fall off the ECM
some cells will not grow even if food is plentiful, unless they are attached
this dependence ensures cells only grow or divide when the situation is right (as determined by the ECM)
falling off the ECM allows cells to decide their own fate (dangerous rebels! just look at those naughty cancer cells, growing and wandering about on their own like that), which is no good for the overall health of the organism
binding to ECM also spreads cells apart, so they have room to grow and divide (see, ECM provides for all)--if cells are lost, it'll be faster to refill that space if all the remaining cells get spread out and grow than if only the neighbors are able to divide (they would reach senescence much faster that way *insert frown* )

focal adhesion kinase (FAK): a cytoplasmic Tyr kinase

talin and paxillin (which binds to alpha) recruit FAK
FAK molecules at the focal adhesion point cross-phosphorylate each other

the phospho-tyrosine is a docking site for Src family of Tyr kinases, who then phosphorylate other proteins and more Tyrs on the FAKS, creating more sites for Src binding

FAK also helps with removing focal adhesions (observed overexpression of FAK in motile cancer cells, and overproduction of focal adhesions in cells with no FAK)
FAK and similar pathways produce global (entire) cell response, including affecting gene expression

integrins can also produce small, local effects (i.e. only in the cytoplasm/membrane close to focal adhesion)

activate local population of Rho GTPases to cause effect (for example, changing direction of a growing axon)

basal lamina is only one type of ECM, the connective tissue is also a form of ECM

variations in composition give it different properties (from jellyfish jelly to lobster carapace to teeth and bone)
formation and orientation of this ECM is controlled by the cells living in it (mostly fibroblasts!)

fibroblast: matrix cells of connective tissue

secrete matrix macromolecules (chondroblasts make cartilage, osteoblasts make bone, etc)
matrix similar to lamina in that it has (1) proteoglycans and (2) FIBROUS proteins
proteoglycans from jelly substance to resist compression and hold the fibroblasts within, while collagen fibers give tensile and resilient strength
other matrix proteins direct, stabilize, and signal the cells within

GAGs: glycosaminoglycans, chain of repeats of one disaccharide units, unbranched

the disaccharide is 1 of either N-acetylglucosamine or N-acetylgalactosamine (simply put: modified glucose or modified galactose) plus a uronic acid
GAGs are negatively charged due to sulfate/carboxyl groups on sugars
4 types:

hyaluronan
chondroitin sulfate/dermatan sulfate
heparan sulfate
keratan sulfate

GAGs are hydrophilic and stiff (can't fold like polypeptides), so they take up lots of space and volume for low mass, making porous gels even at low concentrations
because of this jelly feature, GAGs usually just fill space in the ECM
their negativity attracts positive ions, who attract water (water likes to follow high ion concentrations, remember hypo and hypertonic?), creating swelling and turgor pressure so that the ECM can resist squishing and compression

hyaluronan: simplest of GAG, 25000 disaccharide units, no sulfated sugars, no generally linked to any core protein, synthesized on cell surface straight into the ECM (other GAGS are the exact opposite: sulfated, protein-linked, synthesized and exported by exocytosis)

serves as space filler and shaping (small quantity swells with water to occupy large volume to move aside parts or expand shape)
synthesized from basal side, it can deform the epithelium to create a cell-free space below for other cells to migrate in

after cell migration, hyaluronidase degrades the excess hyaluronan

the other GAGs are attached to protein as proteoglycans

the core protein goes through the ER processing and receives the sugar chain in the Golgi

linkage tetrasaccharide: first attachment on protein at a serine, serves as primer for additional sugar attachment
glycosyl transferases adds sugars one at a time to the sugar primer
the sugars are also modified in the Golgi: epimerization alters configuration, while sulfations increase negativity
general proteoglycans can have any number of sugar side chains, but one of them must be a GAG [glycoprotein is less than 60% carb by weight, while proteoglycan is 95% carb by weight]

variation comes from type of GAG, length of GAG, modifications of GAG, and combos of GAGs on one core protein

proteoglycans have various functions

GAG chains form gels of varying charge density and pore size (act as sieves to regulate traffic of molecules through the ECM based on charge and size, just like gel electrophoresis :D )
control activity of secreted proteins by binding to them:

binding limits their diffusion, their activity (depending on where they are bound), and their lifetime
example: heparan sulfate binds FGF (fibroblast growth factor), causing them to oligomerize (aggregate) to enable crosslinking and activation of receptors

in this way, proteoglycan binding enhances signal transduction

example: heparan sulfate bind and trap chemokines on inflamed site, chemokines are continuously signaling white blood cells to get here and fix the probelm

in this way, proteoglycan binding localizes a signal at the needed site

can act as co-receptors

some proteoglycans stay attached to the plasma membrane by having their core protein be transmembrane or bound to a GPI anchor (lipid buddy)
these bound GAGs collaborate with cell-surface receptors or the receptors themselves can be proteoglycans by having a GAG attached
syndecan: transmembrane proteoglycan, cytosolic domain interacts with actin and signaling molecules in cortex

bind FGF and pass it on to FGF receptors

collagen: fibrous proteins

3 polypeptide chains (all alpha chains) are wound in superhelix, is long and STIFF
rich in Pro (which stabilizes the helical conformation), and Gly (at every 3rd residue to help pack tightly since Gly is super small)
triplet Gly-X-Y sequence
exist tons of different alpha chains, so different cell types produce different combos if 3 alphas to make different collagen fibers, although the number of possibilities is limited by the individual chains' ability to combine into a superhelix (some A.A. sequences may not fit well with certain other sequences)
fibrillar collagens: forms fibrils (long rope with no interruptions) 10-300 nm diameter and 100s um long, consists of Type I,II,III,V,XI collagen (bone, cartilage, skin, V similar to I, XI similar to II)

collagen fibrils aggregate into bundled cables called collagen fibers (several um in diameter)--this method of organization is similar to muscle fibers and fibrils

Type IX and XII are "fibril-associated collagens" because they decorate surface of fibrils ad link them to one another and other ECM components
Type IV is part of basal lamina (see previous notes)
Type VII form dimers that form anchoring fibrils that attach basal lamina to connective tissue
Type XVII are "collagen-like" proteins that have transmembrane domain and exist in hemidesmosomes
Type XVIII form core protein of a proteoglycan in basal lamina

collagen gets lots of post-translational modifications

pro-alpha chains: large precursor that is synthesized by ribosomes and put into ER, flanked on either end by "propeptides" that will be cleaved
in ER, Pro and Lys are hydroxylated (hydroxy-Pro and hydroxy-Lys)
certain hydroxy-Lys get glycosylated
pro-alpha chains combine into triple helix held by H-bonding = procollagen
hydroxy-Lys and hydroxy-Pro provide the the H for H-bonding

propeptides are cleaved post-secretion by enzymes outside the cell

allows cell to delay collagen fibril formation until outside (otherwise it'd be too big and bulky to get outside)
propeptides also guide the triple helix formation of procollagen but prevent aggregation into fibrils (characteristic of self-assembly)

collagen is cross-linked to form fibrils, and there are striations every 67 nm of fibril (produced by staggered packing of collagen)

cross-linking gives tensile strength

GAGs resist compression while collagen resists stretching

tendons have parallel collagen bundles to prevent muscle pulling, while skin has woven network of fibers to resist in all directions (example is toughness of leather)
arrangement of fibrils is determined by the fibril-associated collagens (i.e. type IX and XII)

their helix chain is interrupted by nonhelical domains to make it more flexible than regular collagen
they are not cleaved and keep their propeptides to block aggregation
they bind periodically to fibrils and mediate interactions btwn fibrils and other matrix stuff

cells also assist fibril organization by using BRUTE FORCE

since ECM is connected to cell cytoskeleton, tugging on cytoskeleton impacts outside networks
fibroblasts can draw in collagen and contract gel to smaller volume, or a cluster of fibroblasts and draw collagen around them into a dense capsule where the collagen goes in circles around them

elastic fibers: fibers that enable stretching and recovery post-stretching

inelastic collagen fibrils can be woven in to limit stretching ability and protect from tearing
elastin: hydrophobic protein that is Pro/Gly rich but not glycosylated and has only hydroxy-Pro

after secretion, elastin crosslinks to each other and forms network of fibers and sheets
the chain alternates between hydrophobic segments (gives elasticity by adopting loose random coil conformation) and Ala-Lys alpha helical segments (forms crosslinks)

elastin core is covered by sheath of microfibrils (10 nm diameter), who are synthesized before elastin and provides scaffold for elastin deposits

microfibrils are also somewhat elastic
made of glycoproteins, like fibrillin

fibronectin: matrix glycoprotein that attach cells to ECM

dimer with two large subunits joined by disulfide bonds, each subunit has multiple domains, each domain has multiple modules, each module is encoded in one exon
the fibronectin gene is one large sequence that gets alternatively spliced to make different fibronectins
one type of module is made of type III fibronectin repeat, which binds to cell integrins

fibronectin fibrils: insoluble form in which the dimers crosslink by more disulfide bonds and combine with ECM

not self-assembly, forms where cell integrins stretch fibronectins to expose an extra disulfide binding site, enabling the extra cross-linking
makes sure fibronectin only forms fibrils where cells need them, not randomly when they're floating through the blood vessel

RGD sequence (Arg-Gly-Asp): sequence on fibronectin that binds to cells

anything with an RGD sequence will bind to a cell
RGD sequence is not the only cell-binding motif, and integrins need to recognize other matrix molecules to form a completely tight bind

cells can degrade matrix too

necessary for repair and restructuring whole organs/tissues
necessary for cell movement and division (making space)

example: white blood cell localization to injury site

pg. 1169 - 1193

Book Notes: ECM Part 2

cell-cell junctions can send signals into the interior

disintegrating the adherens jxns frees the adaptor catenins to wander around cytosol to spread a signal
nonclassical cadherins may serve as co-receptors for signal receptors (i.e. VEGF is a survival signal that a receptor tyrosine kinase binds, but the receptor cannot bind the signal without VE-cadherin's help)

selectins: cell-surface proteins that bind carbs (they are otherwise known as lectins)

they also bind junctions, but specialize in temporary cell-cell adhesion

example are white blood cells who patrol around and use this temporary binding to follow tracks so they can migrate between the bloodstream and tissue (i.e. mustering of white blood cells to location of inflammation)
"rolling" motion along vessel endothelium

transmembrane protein with lectin domain that binds specifically to oligosaccharides on another cell

L-selectin: on white blood cells
P-selectin: on blood platelets and inflamed endothelial cells
E-selectin: on endothelial cells

integrins: cell-surface proteins that bind other surface proteins

also specialize in temporary cell-cell adhesion but is stronger than selectins
work with selectins on white blood cells, but the stronger adhesion is necessary to overcome the selectins' affinity for blood vessel endothelium so that the white blood cell can escape from the vessel into tissue to save the day :D
both selectins and integrins are heterophilic, they don't bind to selectins and integrins of the same type

selectins bind glycoproteins and glycolipids
integrins bind transmembrane immunoglobulins!

immunoglobulin (Ig) superfamily: cell surface proteins with extracellular domains that are similar to antibodies

intracellular cell adhesion molecule (ICAM): type of Ig bound by white blood cell integrin, heterophilic
vascular cell adhesion molecule (VCAM): type of Ig bound by white blood cell integrin, heterophilic
neural cell adhesion molecule (NCAM): type of Ig on neurons, homophilic, contribute to cell-type specific organization but not as necessary as cadherins

synapse formation depends on axon finding the correct target tissue to form synapse with

Ig adhesion molecules help by pairing up homotypically with target cells that express the same type of Ig

example is Fasciclin3, who leads neuronal growth cones to connect with motor neurons in the muscle
another example is Sidekicks, who matches neurons in different layers of the retina to growing axons

synapse formation also requires complex assembly of receptors, ion channels, synaptic vesicles, docking proteins, etc, etc.

scaffold protein: chains of proteins with various binding domains

PDZ domains: about 70 A.A. long domains that bind cytosolic C-term tails of specific transmembrane proteins
the various domains connect to the cadherins, channels, exo/endocytosis regulation protein, cytoskeleton, etc, etc
multiple scaffold proteins can also unite to form network connecting all the components of a sypnase

Discs large (Dlg): protein that assists in synapse formation, occluding jxn formation, control of cell polarity, and control of cell proliferation

all epithelial cells (1) are polarized because one side faces basal lamina while the other faces the lumen and (2) form selectively permeable barrier by utilizing occluding jxns to prevent unfiltered leakage between cells

tight jxn: the occluding jxn of vertebrates, prevent molecules from seeping past the cells in any direction and prevents membrane pumps from going to the wrong side (i.e. in gut lining, nutrients must go unidirectionally, from lumen to bloodstream, occluding jxn prevents them from leaking backwards and prevents nutrient pumps from going to the wrong side and pumping them backwards)
experimentally show that tight jxns seal cell sheets by putting in a traceable molecule of a certain size

visualization of molecule will show it never passes through the cell sheet

sealing not COMPLETELY impermeable; small molecules can diffuse through

but size limit of small molecules varies between tissues: bladder is almost totally impermeable to tiny sodium ions, while small intestine allows their passage

sometimes cells loosen their junctions to allow nutrients to diffuse through easily, especially after meal, when the concentration gradient is high enough to have nutrients go in the correct direction by themselves: "paracellular transport"

sealing strands: chains of adhesion proteins along the length of the cell close to the apex side that bind to partner chains on neighboring cells (not like one pocket or jxn like in cadherins, but chains of jxns all around the cell)

claudins: the transmembrane adhesion proteins forming the sealing strands
assisted by occulin and tricellulin, who seal the corners where 3 cells meet
claudin family has members of varying structure to form different sized pockets to change the permeability

paracellular pores: the selective channels formed by various combos of claudins

occluding jxns have to occur above adherens and desmosomes, so all three are held together in junctional complexes

Tight junction protein (TJP): a.k.a. ZO protein is a cytosolic scaffold protein that organize and anchor the adhesion proteins of the 3 types of jxns

septate jxns: occluding jxns of invertebrates, very similar to vertebrate version
cell-cell junctions and basal lamina govern the polarity

experiment: grow MDCK cells suspended in gel all by itself, shows no polarity

let them divide into small colony, the colony will organize itself into a vesicle shape, secrete basal lamina outside the vesicle, and all the cells now show polarity between the outside and inside of the vesicle
polarization is spontaneous but depends on contact with neighboring cells

experiment: screen for mutations in C. elegans that affect polarization of embryo

PAR genes: family of genes that control polarization
LKB1: vertebrate homolog of Par4, loss of LKB1 causes loss of polarization, overexpression of LKB1 causes self-polarization (not neighbor dependent), so cells can't orient TOGETHER into a sheet

therefore, proteins assisting with polarity come in two categories: producing polarization in general, and orienting groups of cells correctly with landmarks like the basal lamina

category 1: produce polarization in individual cell

Par3: scaffold protein with PDZ domain, binds to the other two
Par6: similar to Par3
atypical protein kinase C (aPKC): binds to the two Pars
the complex has binding sites for small GTPase Rac and Cdc42 and others
Rac and Cdc42 control actin assembly (lack of Rac causes cells to polarize in the wrong direction)
assembly of the complex is one specific part of the cell cortex recruits Rac and Cdc42 to polarize the cell towards that one part
the complex also controls the Crumbs complex (with scaffolds Discs-lost and Stardust, assembles on apical side) and Scribble complex (with scaffolds Discs-large and Scribble, assembles on basal side)

category 2: orient polarized cells to each other and to environment

the proteins get coordinated to the different junctions by binding to tails of the transmembrane adhesion proteins
from apical to basal:

Crumbs complex
tight jxn, Par3-Par6-aPKC complex
Scribble complex

Rac and friends induce cell to secrete basal lamina ECM components to opposite end of cell (since most of the proteins were associated near the apical side with the tight jxns, the "opposite" side is basal lamina)

planar cell polarity: cells in a tissue with apico-basal polarity show additional polarity towards another direction perpendicular to the apico-basal axis.

example: in trachea, all cells are lined from apical to basal side horizontally, BUT their cilia must sweep in a vertical direction to shift mucus and dust upwards and out
studying mutations revealed important proteins whose function is not exactly known yet (they are also oddly named)

Frizzled and Dishevelled (these two part of a signal pathway called "Wnt"), Flamingo and Dachsous (these two code for cadherins)

gap junction: channel between cells that allow them to share cytoplasm without having to exocytose stuff and the other cell endocytose it (plant counterpart is plasmodesmata)

channel is about 2-4 nm long formed by connexins and innexins.
ions and small molecules flow easily through them (dyes and electrical currents transfer quickly throughout a sheet of cells)
channel is about 1.5 nm wide (ions, sugars, A.A, N.T., vitamins, cyclic AMP and other second messengers) but can't share macromolecules (proteins, DNA, etc)

connexon: a hemichannel formed by six connexins (4 pass transmembrane protein)

the two connexons of two cells align and stick together to form whole channel
gap junction consists of a lot of connexons, so it looks more like a sieve of pores rather than one giant channel
connexons can be assembled out of combos of different kinds of connexins to produce different permeabilities
even two hemichannels can be made of radically different connexins and still connect together
connexin have half life of a few hours, connexons are always added around existing gap jxns and removed from center of existing gap jxns
connexins are inserted into membranes and delivered to the plasma membrane by exocytosis

they diffuse around the plasma membrane until they hit an existing gap jxn and get trapped there
unpaired hemichannels are either closed or can be opened to serve as channels for release or entrance of small molecules

gap jxns enable instantaneous transfer of action potential between joined cells

useful in nerve cells where speed and synchrony is important in the response (i.e. escaping from surprise attack)
useful also in maintaining homeostasis of a tissue (small molecule concentration can fluctuate from cell to cell, but having that channel allows them to diffuse out of high concentration cells to boost low concentration cells)

even whole channels can flip between open and closed, and permeability is reduced by low pH or high calcium concentration

ECM is full of calcium, so if one cell is punctured, calcium rushes in, the gap junction senses this and closes, protecting neighbors from the damaged cell (remaining connected means whatever gets into the puncture cell can also get into the healthy cells)

plasmodesmata: the only jxn for plants, is a pore in which the plasma membranes of different cells are CONNECTED, not just connection by a protein channel

channel about 20-40 nm diameter
desmotubule: narrow cylindrical tubule that connects to the ER in the center of the channel

cytosol is shared around the desmotubule

basement membrane (basal lamina): thin, tough, but flexible sheet of ECM that is essential to all epithelia

40-120 nm thick, surrounds muscle and fat and epithelial cells
function in forming structure, filtering between its two sides, influence metabolism, promote cell survival, proliferation/differentiation, act as pathway for migratory cells
skin depends on connection the basal lamina which holds it fast to the underlying dermis

basal lamina made of FIBROUS PROTEINS (glycoproteins with short oligosaccharides) and GLYCOSAMINOGLYCANS (GAGs, linked to proteins to make proteoglycans)

lamina composition varies, but there are key components

glycoproteins: laminin, type IV collagen, nidogen
proteoglycan: perlecan
associate with other molecules: collagen XVIII, fibronectin

laminin-1: classical laminin, has 3 polypeptide chains arraned like bunch of flowers in a vase

exist various isoforms of the each chain type, so laminins come in various combos like connexons
interact with each other at heads to form a sheet (PRIMARY ORGANIZER)

type IV collagen: 3 long protein chains that form superhelix, interrupted in several places to allow bending

interact with each other at terminal domains to form flexible but strong network (TENSILE STRENGTH)

laminin binds to perlecan and nidogen and cell laminin receptors, while collagen binds to perlecan and nidogen
so perlecan and nidogen connect the networks of laminin and collagens (LINKERS)
cell receptors like dystroglycan grab laminin by their feet, laminin heads network together to form a sheet, nidogen and perlecan link the laminin to collagen network

examples of diverse functions of basal lamina

kidney: superthick lamina blocks maromolecules from leaving blood when filtering for urine
epithelium movement: blocks fibroblasts of connective tissue from touching epithelial cells, but allows passage by macrophages and lymphocytes (they have proteases to cut holes through the lamina for them)
regeneration: studied at neuromuscular jxn where nerve meets muscle, but they are separated by basal lamina

lamina has special isoforms of all of its components
when the nerve cell happens to be destroyed, new axons locate the old site by recognizing the special markers of the basal lamina that still marks the junction
works similarly for muscle cell regeneration
nerve cell deposits agrin proteoglycan at the lamina site, who induces muscle cells to form acetylcholine receptors on their surface, while muscle cells deposit special laminins that bind to voltage gated ion channels of nerve cells
when either or both nerve and muscle gets destroyed, muscle regenerates first by natural wound response, but gets positioned correctly by the existing agrin, then growing axons come down to find the old site by binding to the existing laminins

pg. 1145 - 1169
A.A. = amino acids
N.T. = nucleotides

Blog Archive