Author Archives: Ruth Mauntz

Wrapping up

Hope you all saw Queen B kill it at the VMAs last night! With class over, it might be a while before any of you get to reading notebook posts, but I wanted to post an update on the sea star ISH since not everyone was involved in that project.

On Friday we grabbed Carolyn and Colleen to examine the sea star slides, but unfortunately it seems like the counter stain wasn’t very effective and we couldn’t really see the tissue cross sections on any of the 12 star ISH slides well at all.

star uw 14 SS 71S tf ep ish (2)

treated (sick) star

We did spend some time examining the H&E slides and got some good pictures.





Have an amazing summer and good luck to those of you starting school this week!



Integrating the integrins

We’re finishing up the sea star ISH this week! This experiment has been going pretty smoothly, other than having to track down some emergency sheep serum today. We’ve added some abalone slides to the mix too, so we have 18 in situ hybridizations in total for counter staining and visualizing under the microscope tomorrow.

The real fun today was dicing up the cell adhesion story hidden in the Pycnopodia helianthoides transcriptome. I had to do a lot of reading and searching to really start understanding focal adhesion complexes and how they involve different proteins and downstream processes, so I started building that database of papers and stuck them in my folder on Google Drive.

I also wanted to better understand which focal adhesion genes change expression, and how the coelomocytes are responding to disease. It makes intuitive sense that focal adhesion might play a role in the clotting response we’re seeing from changes in the coagulation pathway. Let me paint a picture: fibronectin is an abundantly soluble protein in the extracellular matrix (ECM), although not physiologically active until self-assembled into fibrils. It binds to integrin receptors on the cell surface as well as extracellular matrix components such as collagen, fibrin, and heparan sulfate proteoglycans (Pancov 2002). Understanding the backdrop that the extracellular fibronectin matrix scaffolding plays in cell adhesion is important, but the interesting changes in gene expression are also at the cell-surface level: what focal adhesion receptor and substrate genes are changing and what signal pathways might be affected?

Integrins are the transmembrane proteins that make up an important part of the focal adhesion complex on the cell membrane, binding the cell to the ECM and setting off a potential inside-out signaling event between cell and ECM (Widmaier et al 2012). There’s a plethora of proteins that can interact at the focal adhesion junction.

The following table is a list of interacting proteins I pulled from Plow et al, 2000:

Ligand Integrin
Adenovirus penton base protein αvβ3, αvβ5
Bone sialoprotein αvβ3, αvβ5
Borrelia burgdorferi αIIbβ3
Candida albicans αMβ2
Collagens α1β1, α2β1, α11β1, αIbβ3
Denatured collagen α5β1, αvβ3, αIIbβ3
Cytotactin/tenascin-C α8β1, α9β1, αvβ3, αvβ6
Decorsin αIIbβ3
Disintegrins αvβ3, αIIbβ3
E cadherin αEβ7
Echovirus 1 α2β1
Epiligrin α3β1
Factor X αMβ2
Fibronectin α2β1, α3β1, α4β1, α4β7, α5β1, α8β1, αvβ1, αvβ3, αvβ5, αvβ6, αvβ8, αIIbβ3
Fibrinogen α5β1, αMβ2, αvβ3, αxβ2, αIIbβ3
HIV Tat protein αvβ3, αvβ5
iC3b αMβ2, αxβ2
ICAM-1 αLβ2, αMβ2
ICAM-2,3,4,5 αLβ2
Invasin α3β1, α4β1, α5β1, α6β1
Laminin α1β1, α2β1, α6β1, α7β1, α6β4, αvβ3
MAdCAM-1 α4β7
Matrix metalloproteinase-2 αvβ3
Neutrophil inhibitory factor αMβ2
Osteopontin αvβ3
Plasminogen αIIbβ3
Prothrombin αvβ3, αIIbβ3
Sperm fertilin α6β1
Thrombospondin α3β1, αvβ3, αIIbβ3
VCAM-1 α4β1, α4β7
Vitronectin αvβ1, αvβ3, αvβ5, αIIbβ3
von Willebrand factor αvβ3, αIIbβ3


So, that’s cool. You might recognize some proteins in that list. Why are integrins and focal adhesion complexes important?

We see transcriptome-wide changes in at least 24 integrins or integrin-binding genes:

Protein names log2Fold Change pvalue Entry Length (aa)
Disintegrin and metalloproteinase domain-containing protein 10 (ADAM 10) 2.77 1.42E-07 Q8JIY1 749
A disintegrin and metalloproteinase with thrombospondin motifs 3 (ADAM-TS 3) 3.77 6.09E-13 O15072 1205
Disintegrin and metalloproteinase domain-containing protein 12 (ADAM 12) 4.17 3.68E-14 Q61824 903
A disintegrin and metalloproteinase with thrombospondin motifs 6 (ADAM-TS 6) -2.33 0.002229 Q9UKP5 1117
A disintegrin and metalloproteinase with thrombospondin motifs 13 (ADAM-TS 13) (vWF-cleaving protease) 7.96 1.97E-16 Q76LX8 1427
Sushi/vWF, EGF and pentraxin domain-containing protein 1 (Polydom) 2.20 0.00026 A2AVA0 3567
Sushi/vWF, EGF and pentraxin domain-containing protein 1 (Polydom) 4.20 1.75E-06 A2AVA0 3567
Sushi/vWF, EGF and pentraxin domain-containing protein 1 (Polydom) 1.61 1.26E-05 A2AVA0 3567
Sushi/vWF, EGF and pentraxin domain-containing protein 1 (Polydom) -1.91 0.005433 A2AVA0 3567
Integrin beta-PS 1.74 0.007869 P11584 846
Integrin alpha-4 (CD antigen CD49d) 3.07 6.60E-09 Q00651 1039
Integrin alpha-4 (CD antigen CD49d) 3.63 6.34E-06 Q00651 1039
Integrin beta-1 (Fibronectin receptor subunit beta) (CD antigen CD29) -1.40 0.001249 B0FYY4 798
Matrix metalloproteinase-19 (MMP-19) 4.22 3.85E-06 Q9JHI0 527
Matrix metalloproteinase-24 (MMP-24) 2.24 0.000174 Q9R0S2 618
Collagenase 3 (MMP-13) 3.39 2.10E-05 O77656 471
Collagenase 3 (MMP-13) 10.84 3.86E-43 O77656 471
Stromelysin-1  (MMP-3) 3.97 2.12E-06 P28863 478
Collagenase 3 (MMP-13) 7.82 1.48E-18 O77656 471
Matrix metalloproteinase-17 (MMP-17) 4.08 8.08E-09 Q9ULZ9 603
Matrix metalloproteinase-16 (MMP-16) 3.43 7.61E-07 Q9WTR0 607

A lot of these proteins pop up as important clotting or anticlotting responses, especially sushi, ADAM proteins, and MMPs. It may be reasonable to think more in depth about the role that changes in focal adhesion have on signaling events. How are the cells, specifically the coelomocytes, responding to this infection?

Side note! In my search today I came across a really cool protein called integrin-linked kinase (ILK). In addition to being an awesome part of the focal adhesion complex, it’s a biochemically inactive kinase…. it doesn’t phosphorylate proteins! Here’s a phylogenetic tree I generated in String with ILK (first, left-most column of green and orange) and its known interacting proteins.



Fishing for complements (no, not that kind!)

This just in! Based off a preliminary look at the genes in the sick stars, the complement cascade and coagulation pathways are both undergoing changes in expression levels. The complement cascade, an integral part of the innate immune response with downstream effects including opsonization and phagocytosis, also has changes in gene expression of key complement factors.

vWF is an important clotting protein, both there are also several ADAMS (vWF-cleaving) proteins that are changing expression levels. Interestingly, I’m having a hard time finding any serine proteases (clotting factor activators) with differential expression levels.

What a weekend! The RCN talks were extremely interesting and I could go on about the exciting strides these scientists are making in marine disease diagnostics, outbreak forecasting, and marine disease communication.

complement and coagulation cascades


BLAST it all!!

I was kind of all over the place after lunch running the DIG-PCR and finally working out the script for the sea urchin BLAST, but am happily ending the day searching spIDs for under-expressed adhesion proteins for the sea star transcriptomics project. While I felt a little scatterbrained with the RCN approaching and with the culmination of these projects looming, I’m happy to report two successes:

1) Monica and I successfully synthesized the DIG-probes for the candidate microbial pathogen today. After we ran the probe out on the gel we got a bright band right at 250bp. We have probe, and although one of our controls boiled off during the PCR, I ran another PCR synthesis after dinner and have that metaphorical bun in the oven. The plan is to move forward with the sea star ISH this Sunday as the RCN winds down, so we have time to pull data together for our speed talks.

2) I also finally was able to BLAST our transcriptome against the sea urchin nucleotide base from Spbase, but am only getting a fraction of homologous genes that should be there. Collin is also curious at the general characterization of our transcriptomic data, although he’s approaching it a little differently. According to Collin, the annotations we’re working with include a total of 3947 urchin genes (not necessarily significantly differentially expressed). I’m getting 782 genes (I think maybe my BLAST parameters are too stringent). Take a look at my notebook and let me know what you think (hint: the most interesting part is at the bottom).

Sea Urchin BLAST, e<-05

Thank you Maya for the fun lecture and lab time on ecological modeling today!


PCR, probes, paths….. and a really cool video!!

Well, we got bands again after running the PCR with our sea star primers. This time we ran the full gamut of tissues with primer 1 (top lanes, 218bp), getting bands in the diseased cecum. Primer 2 (bottom lanes, 250bp) again shows bands in diseased cecum, as well as  healthy cecum and a healthy tube foot.  Apparently both our recipes for cPCR and DIG-PCR use the same concentration of MgCl, so I’m not sure exactly we’re having problems getting a probe product. We might need another DIG-PCR run before we put this ISH to rest.


Also spent a lot of time working in IPython trying to blast the star transcriptome against the urchins. The script hasn’t really changed much from yesterday, I think I just have to get my blast path. Almost there, promising amazing results soon!

……….On a more fun note, check out this video I found of sea stars feasting on mussels!! They’re so cool!!


Sea urchin BLAST

Yesterday I set a goal for myself, BLAST the transcriptome against the sea urchin rather than the universal NCBI nt database. has an databases on sea urchin (echinodermata) RNA and protein, which would be informative to compare 1) contigs or 2) annotations to our sea stars spIDs. Suffice it to say, it was a struggle getting my script together and figuring out what monster I had decided to tackle. Here’s my code from the day, I am trying to repeat the BLAST today but this time against the genome (not a blastx) and with an Eval<-05.

August 12 Sea Urchin BLAST notebook:

I spent some time during the blast playing around with our enriched processes from DAVID. Thoughts on GO terms- DAVID outputs a multitude of GO terms with our spIDs but PANTHER, another gene ontology software, seems to parse down the number of GO terms. Not that this is changing the annotations, but it might be useful when looking at the overall changes in metabolic process.


Sea star PCR

We got bands! And it looks very exciting as we’re getting hits in both of our diseased sea stars only in the coelomic fluid (important for wound healing and immunity, Holm et al, 2010). Lanes 2-5 are the first set of primers we redesigned, lanes 6-19 are the second set we ordered (we’re getting a product in the 250bp range), and the bottom half of the gel is the primer set we used in week 1. Again, we are not getting DNA amplification with that set but we have at least one primer set that works, so we can move forward with our ISH experiment and start making probes tomorrow.

Also very excited to start team meetings tomorrow and set some RNAseq goals for the next few days.

EDIT: We used tissue from the cecum, not the coelomic fluid. Lanes 2-5 are, respectively, H1_T, H1_D, D1_T, D1_D.



Peeking into the transcriptome

On Friday [I missed my log on Friday so this Sunday's post is going to be a two-fer] we spent the morning in R using Lauren’s WGCNA script to look at different ways to model the sea star transcriptomic. It’s possibly a very useful way to look at differential gene correlations between the 3 sites our stars were collected from, but apparently might need some code tweaking to make a better fit model.

Later we converged to make a game plan, a wise decision tackling a project this immense. We decided the best way to start would be to divide and conquer, splitting into 3 teams that look at the transcriptome, the DEGs (differentially expressed genes), and the enriched data set. After hours spent sifting through spIDs (, I found some interesting proteins that have the most highly differentiated expression from our data mining session. More strategy sessions, goal-setting, and assignment breakdowns are going to be the name of the game as we continue breaking down what’s happening to our infected sea stars!

Gene Name Subcellular location Notes
Interleukin-6 receptor subunit cell membrane/secreted Two isoforms that localize either to cell membrane or secretion. Ciliary neurotrophic factor
Beta-ketoacyl-acyl-carrier-protein synthase I Lipid metabolism; fatty acid biosynthesis
Collagenase 3 secreted Plays a role in the degradation of extracellular matrix proteins including fibrillar collagen, fibronectin, TNC and ACAN. Also important for wound healing and tissue remodeling.
Polyketide synthase 1 Secondary metabolite biosynthesis; flavonoid biosynthesis
Putative D-arabinono-1,4-lactone oxidase mitochondrial membrane Cofactor biosynthesis; D-erythroascorbate biosynthesis; dehydro-D-arabinono-1,4-lactone from D-arabinose
Quinone oxidoreductase cytoplasm putative superoxide scavenger. Xenobiotic catabolic process
Skeletor, isoforms B/C cytoskeleton/nuclear mitotic spindle Provides structural support and reorganization during meiosis/mitosis
Probable polyketide synthase 1 Secondary metabolite biosynthesis; flavonoid biosynthesis
vWF-cleaving protease secreted Has a disintegrin domain, which is a family of proteins that inhibit integrin and cell-cell and cell-ECM adhesion. Inhibits platelet adhesion and blood clotting.
Fibropellin-3 secreted extracellular matrix glycoprotein that’s present in echinoids



Major changes in biosynthesis and extracellular matrix organization

Major changes in biosynthesis and extracellular matrix organization






Gene Name Sub cellular location Notes
Cytochrome P450 2J6 ER, cell membrane arachidonic acid metabolic pathway. Known to interact with myocilin
Heparanase secreted Enzyme that cleaves heparan sulfate proteoglycans for ECM degradation and remodeling. Interestingly, in mammalian studies increased activity of this enzyme results in increased blood borne tumor metastasis, probably because of its potential to degrade endothelial linings.
Protein arginine N-methyltransferase 2 cytoplasm to nucleus represses transcriptional activity through methylation of histone H4, may have role in hormone signaling pathway as well
Dipeptidase 1 apical cell membrane hydrolyzes dipeptides in renal metabolism, microvilli projections
Lutropin-choriogonadotropic hormone receptor (GPCR) cell membrane GPCR hormone receptor
Zinc finger and SCAN domain-containing protein 2 nucleus transcriptional regulation
Sodium-dependent noradrenaline transporter cell membrane multi-pass membrane protein for amine transporter. Terminates the action of norepinephrine by sodium-dependent reuptake into presynaptic terminals. Target of psychomotor stimulants such as amphetamines or cocaine
Myocilin ER, extracellular exosome, extracellular space glycolytic signalling role and seems to have a strong role in the cytoskeletal organization of neural cells, with processes controlling cell-cell adhesion and cell-matrix adhesion. Mutations in the human eye MYOC can lead to glaucomas.
Patatin-like phospholipase domain-containing protein 4 lipid hydrolase
Poly(U)-specific endoribonuclease secreted Has 2 SMB (growth hormone) binding domains, cleaves ssRNA
Cytochrome P450 1A5 ER membrane, endomembrane arachidonic acid metabolic pathway. Known to interact with myocilin


Changes in transcriptional regulation, lowered expression of neural and hormonal receptor pathways,  as well as decrease in cell-matrix adhesion and cytoskeletal reorganization

Changes in transcriptional regulation, lowered expression of neural and hormonal receptor pathways, as well as decrease in cell-matrix adhesion and cytoskeletal reorganization


I also sifted through a asterias proteomics paper Morgan found but only came up with 2 proteins (RAB and GTP-binding proteins) that matched spIDs from the coelomocyte protein they extracted. Not sure why that would be so low. Maybe annotation has come a long ways since 2011. Here is the paper:



I got DAVID working!

After hours spent working out erroneous scripts of course. I had a bright idea to rerun DESeq2 in an effort to compare transcriptomes from 2 out of the 3 sites where we collected our sea stars to see if immune response changes in stars with different exposure histories, but kept running into problems with scripts, R codes, and file jumbling. The good news is my scripting tool belt is definitely expanding.

After cleaning up my database files and retracing my database libraries, I decided to take another shot at joining our differentiated transcriptome with spIDs and loading onto DAVID for bioinformatics analysis. I had a much easier time getting to DAVID and running my DESeq contigs against the Phel database and got some fun pathways up on the screen. Next up, ReviGO!

I’ll post my IPython notebook as soon as they’re loaded onto lilmax.

Update: Here I am cleaning up house:

….and here I am attempting to compare site effects:

Had a great time at the speed talks tonight! EIMD covered a lot of ground…. paleobiology, algal phycology, fly genetics and behavior, and coral disease.


Transcriptome enrichment

Lots of time in computer lab today and it’s getting slightly less stressful every time. We went back to SQLShare this morning to join our full annotated star transcriptome with our SwissProt protein BLAST. Afterward we ran a statistical program in R  to decide which transcripts are significantly differentiated (p<0.05) between the sick and healthy sea stars. This  DeSeq output file we then joined with our SwissProt database in SQLShare to save spIDs for each annotation and ran through DAVID (some of us more successfully than others) to essentially normalize our differentially expressed transcripts to the entire transcriptome and look at which processes are affected by this infection.

Here’s a recap of most of my scripts today, I hope you enjoy my frenzied note-taking and repeated command attempts!

red is significant, p< 0.05

red is significant, p< 0.05

If learning to script is like learning a language I’d say I might be able to stutter my way through a sentence.