Green FPs in HEK293T cells

Posted on by kmatreyek

At one point, I was a doe-eyed postdoc reading about new fluorescent proteins (FPs) with improved brightness and thinking it was potentially important to incorporate new FPs into my constructs for cell culture work. I have since come to realize that the intrinsic gains to fluorescence published in those papers do not necessarily translate to brightness in my experiments. The reason are probably multifactorial, including:
1. Commonly accessible equipment is generally optimized for EGFP (or similar FPs), so newer FPs with slightly different excitation and emission spectra may not be captured well with existing microscopes or flow cytometers.
2. The FP brightnesses are typically assess in-vitro, and there may be other factors in eukaryotic cell cytoplasms that may affect the FP brightness (eg. FP half-life).

Well, I’ve still ordered a handful of different green fluorescent proteins anyway, and figured it was worth doing a side-by-side comparison in the transgenic system we use in the lab. This was all done with the HEK293T G542Ac3 (LLP-Tet-Bxb1attP_Int-BFP_IRES-iCasp9-BlastR) cells, which were stably recombined with a single copy of each fluorescent protein at a common genomic locus. The construct organization was: Bxb1attB_[Green FP]_IRES-mCherry-2A-PuroR. Olivia did these recombinations, selected the cells, and ran the cells on the ThermoFisher Attune Flow cytometer, with Sarah’s help. This is what the results look like:

Some interpretations:
1. Rather minimal (~3-fold) difference between mGreenLantern and UnaG. I suppose if we were ever in a situation where we needed every unit of green brightness possible, that we would go with mGreenLantern. That said, UnaG has some benefits; namely, it’s 58% the size of EGFP/mGreenLantern/mNeonGreen, and it lacks the VERY ANNOYING identical sequences at the N- and C-termini (MVSKG … DELYK) which makes molecular cloning a potential pain.
2. Conceptually, I like the idea of fuGFP, but that 20-fold diminished green fluorescence compared to EGFP is potentially problematic. Who knows, maybe I’ll turn this into a target of directed evolution at some point…?

5/22 edit: We recently tested StayGold, and the results were rather underwhelming. In our n=1 experiment, it yielded a 6% increase over mGreenLantern in green MFI within stably expressing landing pad cells. If it were on the EGFP-normalized scale of the chart / experiment above, it would be a value of about 2.15. So ya, not that it’s not potentially better than mGreenLantern; it’s more that, not sure if it’s worth the effort for most applications.

CWRU financial docs

Posted on by kmatreyek

Gotta say; one of the hardest things to deal with in this job are all of the minutiae that come along with the administrative aspects. The topic of today’s post is me keeping notes of my observations with the CWRU financial docs, since 1) I’m just going to forget them otherwise, and 2) it may help other new PIs here.

Salary & Fringe costs terminology: In the summary section for each grant / speedtype / account, personnel costs are summarized. While people’s names are shown in the itemized costs part, the summary section uses vaguer / more confusing language. Here’s the translation:
1. “Faculty Control” <- PI
2. “Academic Support Staff Control” <- Grad students
3. “Research Personnel Control” <- Postdocs
4. “Student Control” <- Not sure yet, since grad students apparently don’t go here?
5. “Non-Academic Professional Control” <- Research Assistant (RA1 and RA2 for me)

Additional personnel costs:
1. Fringe Benefits: Only applies to the faculty (eg. PI) and staff (eg. RAs). As of July 2022, it is 30% for grant accounts, but 34% from startup. Had no clue that difference existed. (2024 update; now the rates are 28% and 34%, respectively)
2. Postdoc insurance: Shows up under “Insurance Control”, and appears to be 12.22% of salary as of July 2022.
3. Apparently there are no additional costs for grad students, as far as I can tell.

Encumbrances: Things that have been charged / ordered, but haven’t been fulfilled yet. My lab has a bunch of backordered items on here.

Core service costs:
We routinely use 1) The CWRU flow cytometry core, and 2) The CWRU genomics core. The charges from them are listed as COR####### numbers billing to Journal numbers, both of which change every month, so there’s no static identifier that can be used to distinguish which is which.

Spent and unspent funds: Probably the clearest place to find these values will be the “contr_summ_by_pi” document. Importantly, the “budget”, “TTD expense”, and “balance” columns have values which are the combination of both direct and indirect cost values. But, as a PI trying to run the lab, I think more in terms of direct costs; both for my personnel salaries and lab purchases, but also for the yearly grant budget. Thus, to convert the “direct+indirect cost” values in the pdf into useful values for lab budgeting, you’ll want to multiply the “direct+indirect” number by 0.625 to get the “direct only” value (at least as of 8/11/2022, when the indirect cost rate here is 61%).

ACE2 dependency paper in PLoS Biology

Posted on by kmatreyek

Our paper, entitled “Expanded ACE2 dependencies of diverse SARS-like coronavirus receptor binding domains” (Roelle et al), is published in PLoS Biology!

Ordering oligos at CWRU

Posted on by kmatreyek

Here’s a price comparison I did back in 2019 (presumably still correct?). But in short, per nt price was cheapest through ThermoFisher.

Thus, we’ve been almost exclusively buying oligos from them, with $7,220 spent (as of June 2022) since our first orders starting December 2019.

Here’s what the histogram of oligo costs have shaped up as.

But, well, don’t order degenerate nucleotides oligos from them as they’ll likely be T biased.

If anyone sees anything better on campus, let me know!

Consistent Plasmidsaurus sequencing miscalls

Posted on by kmatreyek

As I noted in this Twitter exchange, plasmid nanopore sequencing via Plasmidsaurus is great, but not perfect. For example, there seem to be some “achilles heal” sequences, where nanopore reproducibly (like 100% of the time with different plasmid submissions) miscalls certain parts of our plasmids. How do we know they’re miscalls? B/c the Sanger traces of the same exact plasmids show the expected sequence very clearly. Here are two that we commonly see:

A single deleted C nucleotide in the beginning of our IRES sequence:

A phantom T>C base miscall that incorrectly tells us we have a W566R nonsynonymous change in every single one of our human ACE2 constructs.

Both are related to C repeats, but there are plenty of other C repeats in the plasmids we submit and it’s ALWAYS these sequences that give Plasmidsaurus problems. Once I figured this one, it’s really NBD, since I know to ignore these changes, although it did inform our current molecular biology workflow in the lab of 1) Screen colony minipreps via Sanger -> 2) Sequence candidate good constructs with Plasmidsaurus / nanopore -> 3) Sanger to resolve unexpected discrepancies between the expected / intended and Plasmidsaurus sequences.

Command line BLAST

Posted on by kmatreyek

One of the pseudo-projects in the lab requires looking for a particular peptide motif in genomic data. While small scale searches can be done using the web interface, the idea is to do this in a pretty comprehensive / high throughput manner, so shifting to the command line makes sense for this work. I last did this back in 2018 for some preliminary studies, so I’m going to have to re-install the software on my new computer and re-run some of those analyses. I figure I’ll write down my notes as I re-do this, so that I (and others) can use this post as a reference.

Installing BLAST+

The instructions on how to download the program can be found here. I’m on a mac, so I downloaded “ncbi-blast-2.13.0+.dmg” and double clicked and ran the package installer.

Assuming it’s been correctly installed, writing the command …

blastp -task blastp-short -query <(echo -e ">Name\nAAWLIEKGVASAEE") -db nr -remote -outfmt 1

… into the terminal should actually reveal some BLAST-specific output, rather than throw an error.

Running protein motif-specific blast searches

Type in the following into your terminal:

psiblast -phi_pattern PHI-Blast_2A_pattern.txt -db nr -remote -query <(echo -e ">Name\nGATNFSLLKQAGDVEENPGP") -max_hsps 1 -max_target_seqs 10000 -out phi_blast_output.csv -outfmt 10

Note: The above command will require having a text file specifying the pattern constraint (“PHI-Blast_2A_pattern.txt” above), which can be found here. This should yield a 25 KB file csv output, like so.

Extracting just the accession numbers

I don’t remember if there are other BLAST+ outputs that give you the full hit sequence. If so, the method I ended up taking back in 2018 would seem to be unnecessarily roundabout. But, until I figure that out, I’ll follow the old method. As you can see in the aforementioned output format, it doesn’t output the hit protein sequence, and instead just gives the accession number. Thus, the next step is using the accession number to actually figure out the protein sequence. To do this, we’ll use Entrez Direct. To install Entrez Direct, follow the instructions here. Briefly, type in the following into the terminal:

sh -c "$(curl -fsSL ftp://ftp.ncbi.nlm.nih.gov/entrez/entrezdirect/install-edirect.sh

In order to complete the configuration process, execute the following:

echo "source ~/.bash_profile" >> $HOME/.bashrc
echo "export PATH=\${PATH}:/Users/kmatreyek/edirect" >> $HOME/.bash_profile

OK, now that it’s installed, here’s how I’ve used it:

First, the output file above has more info than the accession number. To have it pare down to only the accession number, I used this script, which can be run by entering the following into the terminal, assuming you have the previous output csv file somewhere in the directory with the script (can even be in other folders within that directory):

python3 3_Blast_to_accession.py

This will create a file called “3A_prot_accession_list_complete.txt” (example output file here) which will be the unique-ified list of accession numbers to give to Entrez Direct. (Uniquifying is important if you have multiple .csv outputs you wanted to compile into a single master list).

This can be fed into Entrez Direct using this shell script, which you can run by typing in:

sh 4_Accession_to_fasta.sh

You should now have an output file called “4A_prot_fasta.txt” with the resulting protein sequences in fasta format, like so.

Now you can search for your desired sequence (in its full protein context) within the resulting file.

To be continued…

Are there other steps in this process related to this project? Sure. Like what do you do with all of these full sequences containing the hits? Well, that’s beyond the scope of this post.

ODs on the spec and nanodrop

Posted on by kmatreyek

So there are two ways to measure bacterial culture ODs in the lab. The first is to use the nearby ~ $10,000 Thermofisher Nanodrop One (no cuvette option). The second option is to use a relatively cheaply made cuvette-based spectrophotometer I bought off of Amazon for ~ $100. To make it clear, this comparison is not a statement about the value of a Nanodrop (though I will say that having an instrument like a Nanodrop is essentially a must in a mol biol lab). This is more about if the Nanodrop is already being used by someone and waiting would get in the way of some bacterial speccing timepoints, can I purchase a $100 piece of equipment to relieve such a conflict? Especially for bacterial cultures, where volume isn’t really an issue and the measurement is simply the reading at 600 nm, not even requiring some algebra to make a conversion to more practical units (like ng/uL for DNA).

So to do this comparison, over a number of independent instances, I took the same bacterial culture and put 1mL into a cuvette and ran it on the old spec, and took 2 uL and put it on the Nanodrop pedestal and measured there. I made a table of the results, and graphed it in the plot below.

So the readings on the two instruments certainly correlate (that’s good), although it’s not an exact 1:1 relationship. In fact, the nanodrop gave numbers roughly 1.5 times higher than the spec. But if the two instruments give two different readings, then the question becomes “which is right?”

And to that, I essentially say there is no right answer. Each is a proxy for bacterial cell density (ie. Billions of bacteria / mL), but there’s no “absolute” information encoded in the OD number that tells us that specifically for our bacteria, and we’d still have to come up with a conversion factor either way (ie. my doing limiting dilutions of specc’d cultures and counting colonies), and once we have that, both will be right with that context. Sure, it would be nice if we had a method that was the most in-line with whatever ODs that were being described by various papers in the literature, but who knows what they used (recent papers may be using ODs from the nanodrop [with some perhaps using the cuvette option but many others not], while the older publications certainly didn’t have and instead likely used some old-school form of spec). But even that’s going to be heterogeneous, and will only give limited information anyway.

Well, good record-keeping to the rescue. We’ve transformed the positive control plasmid enough times to sample a range of various ODs just by chance, to see if certain bacterial ODs correlate with transformation efficiency. And boy, there’s been a whole lot of nothing there so far (which is actually quite notable; see below).

(FYI: I don’t remember which instrument I used to measure the OD A600 readings. Probably mostly the old spec, tho).

So yea, I’ve generally used cultures with ODs at the time of collection between 0.1 and 0.45, and they’ve collectively given me transformation rates of ~ 20,000 using our standard “positive control” plasmid. So there seems to be a pretty wide window of workable ODs. But generally speaking, I see no issue with having a culture of 0.1 to 0.4 OD as measured with either machine for use with chemical transformation.

Setting up a hybrid lab meeting

Posted on by kmatreyek

Both due to child-care and pandemic reasons, our originally 100% in-person lab meeting for a time was 100% remote and for the last few months have been 100% hybrid. For overall accessibility reasons, I’ll likely have hybrid remain the default option, and only not bother to set up the Zoom when it’s clear absolutely everybody is going to be in attendance in-person. Over time, I think I’ve better learned how I should be setting up the hybrid lab meeting, and I figure I’d write down the steps here so I can remember (and anybody else can do so if they’re setting things up).

Standard flexible format (ie. Nobody needs presenter mode)
Here, the laptop plugged into the projector is providing the sights and sounds of the conference room, but is simply serving as a “viewer” of the slides in Zoom. Here, I’ll assume this is being done with the common lab laptop (Kenny’s old laptop from 2019), although anyone’s laptop should work.

  1. Plug in the 360 degree camera into the common lab laptop. If you also want to use a different external microphone (ie. if you don’t want to use the microphone associated with the 360 degree camera), then plug that in now too.
  2. Log into the lab meeting on Zoom. Confirm that the right camera and microphone are selected. Make sure the sound is up to the maximum, and that this computer remains unmuted.
  3. Using the USB-C adapter, hook up the laptop to either the projector or the wheeled TV. The adapter will allow for connecting to the projector with the existing VGA cord, or the TV with an HDMI connection.
  4. Make sure the Zoom screen is showing on the projector / TV screen.
  5. To actually use this setup, the idea will be that 1) anybody with their own computer can log into the lab meeting Zoom and share their desktop or window with the presentation (powerpoint file or google slides, for example), or 2) if it’s someone without their own computer, they can use Kenny’s or Anna’s computer to screen share (assuming the presentation file is somewhere easily accessible, like the “Lab_meetings” directory of the lab Google Drive).
    Note: While these computers are being used to share the slides, all sound (input and output) should be happening from the common lab computer connected to the projection device.

One speaker / Longer-form talk where someone needs presenter mode
The main difference here is that the laptop presenting the slides has to be plugged into the projector / monitor, and is thus not simply a “viewer” in the Zoom call. Here, I’ll assume this is being done with the common lab laptop, a

  1. Plug in the 360 degree camera into the common lab laptop. If you also want to use a different external microphone (ie. if you don’t want to use the microphone associated with the 360 degree camera), then plug that in now too.
  2. Log into the lab meeting on Zoom. Confirm that the right camera and microphone are selected. Make sure the sound is up to the maximum, and that this computer remains unmuted.
  3. Open the presentation file. Windows may get much harder to navigate once connected to the second screen, so you may as well get everything set up beforehand.
  4. Using the USB-C adapter, hook up the laptop to either the projector or the wheeled TV. The adapter will allow for connecting to the projector with the existing VGA cord, or the TV with an HDMI connection.
  5. Now, go to Zoom and hit “Share screen”. Probably makes sense to choose the screen with the presentation on it, though it doesn’t really matter at this point since you can adjust it later.
  6. Once the screen is sharing, go to the slide / presentation software you’re using. Assuming it’s Powerpoint, then hit “presenter view”.
  7. If the wrong screen is showing on the projector / TV monitor, then hit swap displays in the Zoom panel until it does.
  8. Now, if the wrong screen is being shared on Zoom (ie. people in the Zoom call are saying they’re seeing your presenter view), then hit the “Share screen” button again and choose the correct screen to cast.

That should do it!