☣ HTGAA 2015 ☣

Welcome

Class-specific

General biology notes/links

• Fun
  
  • Jonathan Coulton song about DNA
  • Protein synthesis dance
  • PCR song
• DIY
  
  • CRISPR
  • DIY BIO
  • Gaudi labs open lab equipment
  • Open Wetware
  • Kits
    
    • Amino DIY bio kit
    • Synbiota

Hardware inventory @ AS220 Labs

Glossary (TODO: merge with Romain's):

Final project ideas

• Somewhat feasible
  
  • Biological version control?
    
    • "Clocked" enzyme– an enzyme which does nothing unless exposed to a certain frequency of light
      
    • nucleotide-level XOR gate
      
    • A nucleotide-level hash function
      
    • in-vivo sequencing (prerequisite)
      

      People are approaching this by increasing mRNA expression or Ed Boyden style, with lasers.

    • Ron Weiss's lab does something like this with logic gates triggered by Micro RNA
      
  • timed drug delivery
    
    • synthetic OMV with a tunable repressilator
      
  • Hormones from yams
    
    • With e. coli?
    • Estrogen? Estradiol?
  • Giant mushrooms
    
    • For composites! I think Anna might be doing this.
    • That express psilocybin?
  • Ergot that directly produces LSD
    
    • Or maybe just LSA.
  • UVC evolution accelerator
    
    • There is one of these at NASA ames
• Way out there/completely nuts
  
  • Biological synthesizer/voicebox
    
  • Countermeasures to biological surveillance
    
    • other people's dust?
  • Make one of George Church's organs
    
  • Sequel to Biosphere 2
    
  • XY->XX transition virus (but how to trigger regeneration of every cell?)
    

Week 1, overview [2015-08-25]

• Big picture
  
  • "universities are obsolete and we're working on the other half" –N.G.
  • "We're applying Shannon's theorem's to everything– bio is next" –N.G.
  • data<->bio
  • MIT is a mainframe, MOOCs are timesharing/bitnet, fabacad is the internet
  • HTGAA 2015 goals: local guru in each lab for 2016, working lab infrastructure.
  • design->build->test->analyze-design loop
• Biology
  
  • "the gut is the rainforest of the body" –David Kong
  • DNA nanostructures are about using the Franklin basepairings– A-T and C-G– to move tiny 3D objects
  • Gene drives
  • Phenyl rings? Benzene?
  • DIY inkjet DNA printer
  • ticker tape ion flux eavesdropping
  • There's a BSL-line between recombinant and non-recombinant
  • There is a gene therapy, Vivera, in clinical trials in Europe
  • "It's a good business model, but it's not a solution" re: some kind of bio-DRM
  • Gene drives: not a virus
• Legal/ethics
  
  • Links/local regulations
    
    • Rhode Island College biosafety group
    • FDA biologic products regulations
  • Notes/questions
    
    • who's responsible for the products of my cells?
    • what is the risk of synthetic biology as a surveillance tool? (human dust contains 10 micrograms of DNA/gram)
    • what about the potential flood of cheap drugs?
    • How do we keep our ability to analyze systems in pace with our ability to manipulate them?
    • What sort of "governors" (like steam governors) can we grow?
    • import/export controls on information
    • "bolster detection and response capabilities"
    • iGEM has an ethics requirement
  • What should we do about malicious state actors?
    
    • bioweapons convention
    • protective norms

Week 2, DNA nanostructures [2015-09-02]

• Safety
  
  • How to throw things away (again, BSL 1: 1:10 bleach mixture, down the sink.)
  • Brown and RIC have bio labs, the people there will know the rules.
  • Curved cleanable floors required in some places
• CADNAno
  
  • Installation
    
    • CADNAno 2.2
      
      • Download cadnano2 from github
      • run python main.py

      This worked flawlessly on all my GNU/Linux computers.

    • CADNAno 2.5
      

      The following should work for most people– install QT5, PyQT5svg, some dev tools (in case you don't have them), download SIP and PyQT5 and build those, and "radnano" should run fine as well. I found it's less stable than 2.2, but just switched back and forth as each crashed.

      sudo apt-get install gcc make g++ 
      sudo apt-get install libqt5svg5 libqt5svg5-dev python-dev qt5-qmake qt5-default
      wget http://sourceforge.net/projects/pyqt/files/sip/sip-4.16.9/sip-4.16.9.tar.gz
      tar xvzf sip-*.tar.gz
      cd sip-*/
      python configure.py
      make
      sudo make install
      cd ..
      wget http://sourceforge.net/projects/pyqt/files/PyQt5/PyQt-5.5/PyQt-gpl-5.5.tar.gz
      tar xvzf PyQt-gpl-5.5.tar.gz
      cd PyQt-gpl-5.5
      python configure.py --sip-incdir ../sip*/siplib --confirm-license
      make -j2 # one plus the number of cores in your computer
      sudo make install
      cd ..
      git clone https://github.com/cadnano/cadnano2.5.git
      cd cadnano2.5
      python bin/main.py
      

      I recommend running each step on its own, in case something screws up, and not just blindly cutting and pasting. ;p This page was very helpful for building PyQT5.

• Related software/links
  
  • Tools from a biomod team, including one to check your staples
  • a .json example of the Rothemund rectangle on the CanDo DNA origami repository
  • Plotting microtiter platemaps in R
  • Lecture notes
    
    • "DNA origami"
    • short "staple" strands link the scaffold together
    • More like DNA knitting than DNA origami
    • incubate with a thermocycler– or just a styrofoam container and water
    • incubate at around 50 degrees (C, of course) for at least 18 hours
    • We can make tubes
    • We can make spiky tubes with arbitrary attachments
    • We can attach a lipid bilayer to protect the DNA nanostructure the same way cells are. This prevents them from being digested.
    • We can make tubes that open when attached to a specific cell, which can tell them to undergo apoptosis
    • a structure-specific aptomer is capable of recognizing certain unique binding sites on a cell surface
    • aptomers are "kind of like a molecular AND gate"
    • "cascades" of aptomers that check for various cell-surface markers
    • we can make tubes– we can make tiny tiny syringes that punch through cell walls
    • they are literally making a tiny pill
    • DNA nanostructures need submodules and software to keep track of them.
    • strands come in sets of 192
    • the button on the far right lets you select how many basepairs
  • homework
    
    • [X] read Paul Rothemond's 2006 nature paper
    • [X] Build flattened Rothemund rectangle in caDNAno (output: json file)
    • [X] Generate staple strand sequences in caDNAno (output: list of sequences in tab delimited text file)
    • [X] Color staple strands to create a black-and-white pattern in caDNAno (output: json file) I made a simple alternating black-and-white pattern, which comes out to a line of dumbells every row.
    • [X] Generate ASCII file black-and-white image (output: text file)
    • [X] Generate list of wells to pipette to generate black-and-white pattern (output: zeros ones)
    • [X] Python program to insert dumbell strand into the middle of anything (I know, not quite what we want)
    • [ ] order strands
    • Notes on platemap format
      
      • Each plate is 12x8, 96 wells
      • ASCII table is in the form [plate,column,row]
      • The columns go alphabetically from A to H (8 columns)
      • The rows go hexadecimally from 1 to C (12 rows)
      • The plates count from 0 (unlike the rows)

Week 3 Synthetic minimal cells [2015-09-09]

• cadnano review
  
  • There are staples in plates 0 and 1, dumbells in plates2 and 3.
  • We are just ordering staple strands, dumbells, and the virus.
  • the platemaps are just for local use.
  • There's a Python script for generating the json files for cadnano in the extra credit folder
• synthetic minimal cells
  
  • People making synthetic minimal cells have three main goals: reproduction, metabolism, and information storage
  • This is bottom-up, Craig Venter's efforts are top down
  • Synthetic cell walls are semi-permeable phospholipid bilayers, just like natural cells, but they use different lipids
  • "semipermeable" means permeable to small, non-charged and non-polar molecules.
  • You can choose either phospholipids + cholesterol or fatty acids
  • One of the most commonly used phospholipids is 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)
  • As an alternative to cell walls, you can put your enzymes/reagents in water droplets in oil
  • Picking a cell wall is easier than deciding which enzymes and reagents to put in
  • Cholesterol makes the cell walls more fluid and less viscous
  • Tx/TL stands for "transcription/translation"
  • The advantage of synthetic cells is that you have much more control over the system
  • Synthetic minimal cells "aren't alive" in that things toxic to naturally occuring cells don't hurt them
  • "cells don't like anyone messing with them"
  • Cell (e. coli) extract vs. PURE (protein expression using recombinant lements) system
  • This presentation seems like an add for PURE
  • here's an actual ad for PURE, somewhat informative
  • Cell extract is higher yield, PURE has (appropriately) fewer impurities
  • Alternatives to e. coli extract: wheat germ, insect, rabbit, human (HeLa)
  • There's no yeast-based one available yet
  • Purifying membrane proteins is very tough, so PURE is much better for extracting membrane proteins
  • Synthetic cells can be used to image mRNA expression (almost in-vivo sequencing!)
  • You can "evolve" synthetic minimal cells, but they don't divide
  • "Thus far the phi29 genomic replication system is the simplest and best characterized isothermal system available"
  • homework
    
    • read papers
      
      • [X] Integration of biological parts toward the synthesis of a minimal cell
      • [X] Build life to understand it
      • Integrating artificial with natural cells to translate chemical messages that direct E. coli behaviour
      • [X] Synthetic biology Minimal cell mimicry
      • [X] Semi-synthetic minimal cells as a tool for biochemical ICT
      • [X] Semi-synthetic minimal cells origin and recent developments
      • [X] Piecing Together Cell-like Systems
    • main assignment
      
      • pick a function (preferably one normal bacteria can't do)
        
        • Draft design of a timed drug-delivery cell
        • [X] design components (pick a membrane, enzymes, other internal molecules), consider outputs
        • [X] list all lipids, enzymes, genes, and source cells. Confirm reasonable output (GFP? an enzyme? or something other cells respond to? PH change?)
        • [ ] Express GFP in a cell free environment :( we tried, it didn't work.

Week 4, Next Generation Synthesis [2015-09-16]

• synthetic minimal cells review
  
  • Someone in Baltimore designed a system of synthetic cells to produce testosterone+estrogen!
  • "snare" proteins let you fuse liposomes
  • sterols "flip-flop" through cell walls
• next gen synthesis
  
  • downloaded the FSA file of PUC19
  • Got an SVG of unmodified pUC19 from NUPAC
  • wrote a fun translation Python script: sequence-> amino acids, here's the result
  • homework
    
    • read papers
      
      • [X] Read Photoelectrochemical synthesis of DNA microarrays
      • [X] Read Programming cells by multiplex genome engineering and accelerated evolution
    • main assignment
      
      • [ ] select 18 bp priming sites that amplify a ~2.25 kb region of pUC19 immediately upstream of the plac promoter and downstream of the start of lacZalpha.
      • [ ] Design one pair that primes optimally: I looked at pUC19 in LabGenius and in AddGene LacZ starts at 146, AddGene says the lac_\promoter is at 142, LabGenius shows LacZ starting at 148, so I'll just assume AddGene is right.
      • [ ] and another that primes poorly.
      • [ ] Describe the PCR thermocycling program

Week 5, Bio Production [2015-09-23]

• next gen synthesis review
  
  • If your primers bind to one another, you get primer dimers and can't amplify
  • Storing enzymes requires a fridge that can go to -20 degrees
• bio production
  
  • feedstock -> fermentation -> ingredients
  • ginkgo bioworks modifies yeast to optimise the fermentation step
  • Metabolic pathways as a subway map
  • When copying a pathway into a new organism, it's better to find enzyme analogs from similar organisms instead of copying genes over
  • Overexpressing an enzyme can overwhelm a cell
  • When you modify a pathway, you can get around organism specific regulations by introducing pathways from other organisms
  • Start with the end product and try to work backward until you find a compound that already exists
  • software
    
    • COBRA:
      

      sudo apt-get install libglpk-dev libsbml && sudo pip install cobra
      

  • homework
    
    • [ ] design a biosynthetic strategy for a compound of your choice

Week 6, Darwin on Steroids [2015-09-30]

• review
  
  • Methanophages eat methane and produce biopolymers
  • There are bacteria that produce magnets
  • Related genes are often next to one another (and also often grouped into operons
  • $500 for 1M oligonucleotides, 1/20,000 bad after error correcting
  • Biolistics is just shooting DNA into something to mutate it
  • A collection of DNA/RNA sequences related to some biological task
  • Syntrophy is like artificially introduced mutualism
  • Novel amino acids can drastically increase bioproduction of certain molecules
• Homework
  
  • A draft/overview of my ideas on ethics+risks+implementation

Week 7, Genome Engineering [2015-10-06]

• Review
  
  • There are patents being applied for in the US
  • You don't know which genetic disorders you have until a kid is born expressing them, or you sequence
  • Mary presented on making people tiny
• Genome engineering
  
  • Throw a lot of transposons at a collection of reproducing cells
  • If the cells survive, the genes the transposons knocked out aren't essential.
  • cells include a lot of DNA with unknown function, test one thing at a time
  • DNA sections go into "cassettes"
  • "recombinant mediated cassette exchange"
  • this synthetic genome is tetracycline resistant
  • the wild-type ones aren't
  • one round of transposon bombardment+bio knowledge isn't enough
  • they didn't increase the mutation rate
  • Defragment the genome!
  • Synthetic genomes not yet medically important
  • hack genes in yeast, transplant them back to microbes
  • some bacterial genes are toxic in yeast
  • they developed an e. coli that donates genes to diatoms
  • novartis asked them to modify e. coli's cell wall so it doesn't look like e. coli
  • then they lost funding for that project
  • they're developing artificial mammalian chromosomes
  • bacteria "conjugate", exchanging DNA!
  • yeast -> e. coli is hard, but yeast -> mammals and e. coli -> mammals is easy
• Homework
  
  • [ ] Come up with an algorithm for minimal genome creation

Week 8, Florescent In Situ Sequencing [2015-10-14]

• FISSEQ
  
  • We want to do in situ sequencing for all the normal engineering reasons
  • Designing plasmids without FISSEQ is like designing circuits without an oscilloscope
  • Modifying mRNA and cDNA to floresce based on sequence
  • FISSEQ still requires killing the cell
  • gene ontology database
  • Not every expressed gene is visible in FISSEQ
  • With FISSEQ, we kill the cells instantly, so they have no time to change
  • Classic next-gen sequencing sees much more mRNA– FISSEQ is like zooming in.
• Homework:
  
  • Computational assignment: Analyze a FISSEQ dataset and find some in situ sequences.
    
  • Lab assignment: Create an in situ sequencing library inside a polyacrylamide hydrogel, and detect the sequencing amplicons using fluorescent sequencing by hybridization.
    
  • Design assignment: Answer these questions based on past experiences in HTGAA:
    
    • Were there any experiments where in situ data of RNA, DNA, protein, or other cellular features would be helpful in understanding the engineering process?
      
      • The only experiment we did in lab was expression of GFP in a cell-free system. I don't think FISSEQ would have helped.
    • What are some reasons in situ data could be better than bulk data for this experiment?
      
      • There aren't any.
    • What kinds of molecules would you like to detect?
      
    • What factors would limit your ability to detect the things you are interested in?
      

Week 9, Synthetic developmental biology [2015-10-21]

• FISSEQ review
  
  • Someone wanted to use FISSEQ on biofoam, which makes me wonder:
• Synthetic developmental biology
  
  • A bioreactor is a device for inducing cell differentiation in different areas, a kind of 3D printer for organs
  • Image analysis with ImageJ
  • You need image analysis to determine things like rate of change of tissue or how fast bacteria swim
  • Electrodes tend to break down, this can be mitigated by salt reservoirs that store the components
  • wetpong.net
  • how does kombucha respond to electricity? What about pressure?
• Demo
  
  • Cartridges with different organisms that slot in!
• Homework (paramecium!)
  
  • one of the labs just found paramecia in a nearby lake and cultured them that way and found different paramecia go toward different poles
  • [X] Build an Arduino-paramecium interface
  • How fast are the paramecia? From our video data, about 1mm/second when stimulated. This is a very rough figure.
  • potential games
    
    • Hexagon
    • 2048
• Design homework (…okay…)
  
  • Problem 1
    
    • (a) It becomes rounded due to capacitance and lower voltage due to the resistive load of the bioreactor.
    • (b) The bioreactor has some nonzero capacitance and resistance, distorting the waveform, drawing current, and bringing the voltage down.
    • (c) The load will be greater, so the voltage will be lower.
    • (d) Match the impedance by adding a capacitor.
  • Problem 2
    
    • No idea how to use imageJ
  • Problem 3
    

Week 10, Biofabrication and additive manufacturing [2015-10-28]

• Homework review
  
  • You can feed paramecium with the stuff that comes out of a fish tank filter when you squeeze it
  • When you have too many cells, either freeze them or toss them (my policy with kombucha)
• Silk
  
  • We don't yet know how to make high quality sink in vitro, so we're extracting the silk protein from cocoons
  • Silk is a 400kilodalton protein
  • Silk is edible
  • Silk can be mixed with chitosan to make a clear transparent plastic substance
  • You can make sensors with blood (spooooky)
  • You can copy micro-optics, like holograms and DVDs, with silk
• Homework
  
  • [X] Acquire Lithium Bromide
  • [ ] Extract fibroin
  • [ ] Reconstitute silk

Week 11, Gene drives and synthetic ecosystems [2015-11-04]

• Homework review
  
  • Even the labs that don't yet have all the material have some cool results, breaking down and reconstituting bioplastics
  • It matters a lot what you start with
  • Lithium bromide has to be disposed of the way photoprinter chemicals would (luckily we have a darkroom here!)
  • An alternate solvent, formic acid, was discussed
  • Gaudilabs dissolved PVA while they were waiting.
  • We tried to build a centrifuge, but had trouble making it fast enough.
  • Play with blending stuff with the silk
  • Water+ethanol+calcium chloride can be used as a LiBr replacement
  • San Diego used Sigma-Aldrich's "plastic atmos bags" instead of a fume hood
• Gene drives
  
  • "Biology works through selective stickiness"
  • "If you think there's something bacteria can't do because we're special, you're probably wrong"
  • Cas9 pretty much always works
  • And generally ignores mutations
  • Cas9 can also be used as a DNA-binding protein for transcription regulation
  • Why is biology so much harder than other technologies? Evolution! Too many parts.
  • Biology is full of spaghetti code, too many interactions
  • Engineered organisms rarely survive in the wild
  • Gene drives cheat inheritance
  • Cas9 only works on sexually reproducing organisms
  • Around half of all bacteria have some sort of CRISPR system
  • You can "crash a population" through a number of methods
  • A Y chromosome + a gene drive would likely destroy an entire population
  • But these "suppression drives" are hard to make
  • You can use gene drives to make a species more susceptible to gene drives
  • By next year we should have a gene drive on nematodes as a demo
  • Yeast have gene drives because they occasionally sexually reproduce, but they tend to be selected away in organisms that reproduce often sexually
  • What can go right?
    
    • We can target specific pests
    • We can end dengue, malaria, etc.
    • We can save species that are going to go extinct
    • But how do we get public acceptance?
  • Design requirements
    
    • The organism needs to reproduce sexually
    • You need the sequence of the organism
    • You need to know that you can make the organism transgenic
    • You need to use a highly specific, very strong promoter that stays in the nucleus (U6 promoters are the suggested candidates)
  • Homework
    
    • [ ] Design a custom discussion forum for gene drives (…okay…)
    • [ ] Describe a problem that could be fixed with a CRISPR gene drive
    • [ ] Select target organism
    • [ ] Explain why a gene drive is the ideal solution vs. other options
    • [ ] What could go wrong?
    • [ ] Who should be involved in the discussion?
    • [ ] Design a gene drive
    • [ ] Identify target gene
    • [ ] Find CRISPR target sites
    • [ ] Explain safety measures

Week 12, Gut Microbiome [2015-11-11]

• Review of gene drives
  
  • Cloud services LITERALLY ON A CLOUD lolol
• Class
  
• homework
  

Week 13, Protein Folding [2015-11-18]

• Homework
  

Week 15, Open Hardware [2015-12-02]

• Review of protein folding
  
• Homework
  
  • [ ] Design open hardware tool for biology

Week 16, IP [2015-12-09]

• Review of Open Hardware
  
  • A device that measures the capacitance of cells as they grow
  • A device that makes bio-friendly tubes
  • 3D prints of pictures of… something
  • A bioprinter with a syringe
• Setting up biohacking labs (Urs Gaudenz)
  
  • "growing this aeroplane as we've been flying it"
  • "a lab is not a set of tools, it's a group of people"
  • "it's important to work with others so you have someone else to blame"
  • "a kitchen is a perfect first biolab"
  • You can find a lot of parts of dumpsters
  • Start with projects– projects inform parts, not the other way around
  • bentolab
  • http://openhardware.science
• Patents and IP (George Church)
  
  • Myriad genetics had patents on genes for breast cancer until challenged by ACLU
  • Because use of PFGE was restricted, better alternatives were developed
  • all of you are part of the revolution
  • Mutations could create to germ-free organisms that can't feel pain
  • comment: documentary about people staying behind in chernobyl
• BK bioreactor: hacking the genetic heritage of a superfund site (Dr. Elizabeth Henaff)
  
  • FLOATING BIOLAB
• Homework
  
  • [ ] complete feedback form

Week 17, final presentation [2016-01-13]