Hacker News new | past | comments | ask | show | jobs | submit login
Attempting to modify e-coli with CRISPR in my bathroom (benjamin.computer)
56 points by Schiphol on July 26, 2018 | hide | past | favorite | 44 comments



The post mentions CRISPR in the title and at several points but I can’t actually discern at what step the CRISPR gene editing would have been performed. As described, this mostly resembles a classical transgenic experiment where a plasmid with an antibiotic resistance gene is introduced (transfected) into the bacteria. This has nothing to do with CRISPR, which allows more targeted modifications of a host genome than simple plasmid transfection.

At any rate it’s a fun experiment that is performed, in similar form, in biology in French high schools. Contrary to the tone in some of the comments it’s not particularly dangerous (the bacteria can’t survive outside their agar plates), though local regulations in some countries still might forbid it from being performed at home.

Unfortunately the description is too imprecise to be sure what exactly the author did.


Near as I can tell the only real reference to CRISPR is that he purchased a plasmid that had been modified with CRISPR. Note: This is not what-so-ever like using the CRISPR method to modify DNA. Modifying bacteria with a plasmid as he is demonstrating here is no where near its equivalent. Every introduction to biology course that I've ever seen performs the identical procedure (and generally with much higher rates of success).


I think I got it: the plasmids (plural) contain the Cas9 and the gRNA targeting rpsL in the host organisms [1] (plus the DNA template for the homology directed repair).

[1] http://www.the-odin.com/crispr-bacterial-guide/


It's kind of insane to modify e coli with crispr anyways, that's kind of like using kubernetes to get a list of files on your hard drive.


It entirely depends why you’re modifying it. Introducing a transgene into E. coli can be done much easier than via CRISPR, true. But that’s not the only (or even main) use of CRISPR. I know of several relevant CRISPR experiments performed in E. coli that can’t really be performed any other way.


Those usually use a variant cas9 protein which is not what is included in Odin.


No, I'm talking about CRISPR/Cas9. For instance, there are experiments knocking out lots of loci simultaneously. I'm not aware of any other gene editing techniques that could be used by here, at least if you want to preserve some selectivity.


Don't know if they are more or less of a pain but lambda red I-SceI works quite well, and I think there is a kunkel based whole genome modifications option.


Some issues and general observations:

* Making plates is hard. You really need an autoclave, and a laminar flow hood is also very helpful. For a DIY approach, you need a pressure cooker and a bunsen burner (in a relatively still room, the bunsen burner creates a radius of sterile space around it. Yes, this is legitimate lab technique, if a bit old-school). Those plates were clearly contaminated, and without proper sterile technique, nothing whatsoever is going to work. The kit's instructions, in my view, are unlikely to produce sterile plates. They provide sterilized agar, but need to be poured in a sterile area, but make no provision for this.

* The instructions for the kit are just barely sufficient. There are a million different ways you can do things incorrectly, so very clear and thorough steps are needed for a novice to have a chance at success. A lot of troubleshooting material should be included as well. This is a setup for failure in my view. For instance, an un-inoculated control plate is a bare minimum control. For a HN-friendly analogy, it would be like debugging software without any error reporting, print statements, etc.

* Lab equipment is mostly fridges, water baths, plastic disposables, and moderately specialized stuff like tabletop centrifuges and thermocyclers. It seems relatively accessible, which gets people excited about DIY microbiology, but it's deceptive. Temperatures, and the stability of temperatures, can matter a lot. Reliable equipment is very important. Someone seriously looking to do this (it's certainly possible!) needs to be able to devote a fair amount of space and resources to their home lab.

* The #1 difference between a 'real' lab and a home lab is access to troubleshooting resources. Shit goes wrong all the time, and experienced post-docs, lab techs, colleagues, etc. are what make research possible. You also very often need access to quality materials for positive and negative controls. It's not uncommon for entire labs to grind to a halt for weeks or months until some tricky bit of troubleshooting is resolved.

Some people are commenting that this isn't 'CRISPR'. It is. The transformation provides a plasmid that will not convey resistance on it's own (unless they're being dishonest about what's in the kit). Only when a stable transformation of the genome occurs will resistance be seen.


I spent a year in a lab learning how to clone genes into E.Coli. it's amazing how easy it is to screw up. Once, we got a bottle of broth back from the autoclave (it should be completely sterile) and within a week or so of sitting on the shelf, unopened, there was already a cloudy membrane growing inside.

What I learned from the process is that it's all about control experiments and troubleshooting. To make a real discovery today, or even just convince yourself you succeeded in a process, you need to be outright paranoid about every component in your system. I wasted a lot of time because I started with only negative controls, and learned the hard way that you need positive controls.

That doesn't even begin to describe all the problems I had extracting reliable data from gels. I'm convinced that all the dark secrets in biology are in the gels.

Even when I worked for a major biotech company, when i talked to the guys who ran the continuous fermentors said they always saw strange fluctuations in things like bulk metabolic rate, drug production, etc, that they couldn't explain, and they were the experts in reliably growing liters of e.coli.


What worried me most about those pictures is that there is a can in that fridge that says 'breakfast pizza'. Is that a real thing or is this very subtle trolling?


It's a can of olives.

EDIT: specifically, this: https://www.instacart.com/stop-shop/products/3191201-sb-medi...


No, that probably isn't a can of breakfast pizza, just looks like a recipe that's usually printed on cans of vegetables. Further, if you were worried that breakfast pizza was something unusual, it's just normal pizza, not specifically for breakfast, but using breakfast ingredients (eggs, bacon, potatoes.) Often times cans of vegetables include a recipe to... fill space? Provide value?

The can itself looks like a can of olives (warning at the bottom says to watch out for pits, the ingredient line for olives is very specific as to what type of olives to use, olives are fruits in a botanical sense.)


Oh right, you can zoom those pictures, now I see the details you mention. Yeah googling 'breakfast pizza' does show a lot of recipes, I'm gonna surprise my kids this weekend, what got me extra worried though is that this was somehow pizza from a can - but it's not, I can now sleep soundly tonight, thanks :)


Is there a large part of the world I'm ignorant of which considers olives and mushrooms to be breakfast ingredients?


they are not uncommon in omlettes. Spanish Omlette- olives & red pepper. Also, I would imagine that the Japanese eat miso soup with mushrooms for breakfast.


It's a real thing, if Bill Bryson is to be believed, although he never elaborates on its precise nature: https://www.independent.co.uk/arts-entertainment/brysons-ame...


Intro contains two of the greatest lines ever blogged:

>Grow some bacteria

>Take some bacteria and make it competent


I wonder if anyone has genetically modified microorganisms to synthesize illegal drugs. Seem like if this happens, the DEA is going to have a really rough time.


You can synthesize a useful precursor to enantiomerically pure methamphetamine from baker's yeast and benzaldehyde. It's been found in an actual clandestine drug lab at least once:

https://www.ncbi.nlm.nih.gov/pubmed/19443151

It's rare for organized crime to resort to sophisticated precursor-manufacturing because they can buy/smuggle such chemicals more easily than doing multi-stage manufacturing starting from commodity chemicals.

It's rare for solo operators[1] because most "meth cooks" are not capable bench chemists. Following simple recipes assembled by others is about all they can do.

[1] Ones that get caught, anyway.



Wow! Very interesting, thanks for those links


is he risking his family, and possibly neighbors, with this setup in his bathroom?

asking with genuine concern - i have no idea how contagious e-coli would be and what its medium of spreading is.


Your intestines are filled with e-coli. It spreads via contaminated water (or food, usually water).


They usually don't sell deadly bacteria in kits aimed at high schoolers...


"The e-coli experiment you can buy is an attempt to create bacteria that can survive on an anti-bacterial plate of agar."

That's exactly what we need!


This is standard procedure in microbiology. Resistance is to one known antibiotic rather than an attempt to create a MSR monster. This allows one to quickly verify if a trait intended to be conveyed with the antibiotic resistance was transferred, since cells without the package will just die.


This is so hacky compared to working in a well-equipped lab. Not having analysis instruments also limits how much can be accomplished.


You say this like it's a bad thing? I'm more amazed that anything like this is even possible outside of a lab environment.


I've engineered a 20kbp yac in my kitchen, but also I have a PhD in chemistry and real molecular biology experience.


Pretty much anybody with enough training in the right techniques and access to the right catalogs could do this in a garage. It's harder (labs have a lot of conveniences), but not impossible.


>This is so hacky compared to working in a well-equipped lab

>Hacker News

Seems about right


>Attempting to modify e-coli with CRISPR in my bathroom

This either sounds like the third sentence uttered in a 911 call, The FBI justification for a lifetime supermax incarceration, or the cause of death and mortality summary on a medical report.


I know the leaders in the DIY bio movement. They've all talked to FBI agents, and the consistent message is: we know you're allowed to do this, thanks for following the law, and by the way, if you see anything, say something (to us).


It made me think, "I wonder if you could modify whatever bacteria is growing in my bathroom to emit a fresh lemon scent, so it would always smell freshly cleaned."



Hmm, am I the only one concerned here? This technology should be strongly regulated. I don't want some "biohacker" accidentally creating AB resistent bacteria in their back garden...


> I don't want some "biohacker" accidentally creating AB resistent bacteria in their back garden...

There are much easier ways of creating AB resistant bacteria than CRISPR. In some countries, they even give out the necessary chemical reagents for free to anyone with a bit of patience!


Would somebody explain if that actually makes (sort of) sense or it's pure trolling?


Benjamin seems to be a bona fide geek, and the rest of his website is quite something. The [electronic wedding invitation](https://benjamin.computer/posts/2016-11-29-WeddingInvite.htm...) is another favorite of mine. I don't think he is trolling, no.


The accusations of trolling can do a lot of damage.

Some people find hacking ecoli is cool - I am one of them, even if this experiment failed it is great to see what is possible right now.


Where is the accusation? Mine was a question. And, according to other comments, the title is actually a click-bait (no CRISPR involved). Probably I used the word "trolling" with a bit of a lateral meaning (something like "make fun of other people"), improperly.


> no CRISPR involved

Please see my follow-up comment: I was able to find the protocol of the kit that the author likely used and it looks like a CRISPR experiment after all. Its description just didn’t sound like it (the actual steps are somewhat similar between the two experiments).


Thanks for the clarification. Mine was still an honest question.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: