Trying to express a protein in bacteria and it isn’t going well? Maybe you should try expression in an insect cell! Getting the protein recipe in there takes a bit more work but the Baculovirus Expression Vector System (BEVS) is worth it if e. coli make your protein go berserk! Since all organisms, from bacteria to insects to mice to humans can read the same genetic language, you can stick the DNA instructions for making a protein into any type of cell and they can turn it into protein (we call this recombinant protein expression)- but just because they *can* doesn’t mean they *will* and just because they *will* doesn’t mean they’ll do it *well*!
Yesterday I told you about the science of flu shots and how one of the types, FluBlock, uses flu proteins recombinantly expressed in insect cells, and how I use the “same” system to express my proteins (although I am not involved in any way in vaccine development). So I thought today would be a good time to review how this system works. note: this post is a “refreshed” version of a post from March, but I’m busy purifying the proteins I’ve expressed!
Bacteria can read & write but they have different machinery so sometimes they can’t properly fold and/or or modify proteins that are normally expressed in different types of cells. Bacteria are what we call PROKARYOTIC – they don’t have conventional membrane-bound compartments inside their cells. We, and other plants and animals are EUKARYOTIC – we *do* have such cordoned-off rooms in our cells, including a membrane-bound nucleus, where we keep our DNA). And that’s just one of the many ways we differ.
Insect cells are closer to our cells – they have more similar machinery – but they’re easier (and cheaper) to grow than mammalian cells, so they’re a good “next try” if bacterial expression doesn’t work out. We usually try expressing it in bacteria first because it’s way easier, quicker, & cheaper. But that’s only if it works – if doesn’t work well you waste a lot of time, effort, and cash!
Bacterial recombinant protein expression’s a lot easier because you just stick the gene for the protein you want into a circular piece of DNA called a plasmid vector -> stick that vector into bacteria -> tell them to make it and they *will.* More on that here: http://bit.ly/bacoverexpression
For insect cell expression, you still stick the gene into a plasmid, but then you combine that plasmid with another one that has insect-virus-making instructions to make a BACMID -> then you isolate that bacmid and stick it in insect cells -> those insect cells start making & secreting BACULOVIRUS -> then you isolate that virus (V0) and add it to more insect cells -> those cells start making & secreting MORE BACULOVIRUS -> then you isolate that virus (V1) & add it to TONS of insect cells for LARGE—SCALE EXPRESSION -> those cells start making PROTEIN! -> isolate & purify the protein 🙂
“Large-scale” can mean different things, depending on how well the protein you’re interested in expresses, how much of the expressed stuff you can expect to lose during purification, and how much pure protein you need. For most of my preps, I express about 4 L and end up with a few mgs at the end of the multi-step purification process. For industrial purposes, they can do it in giant tank things called bioreactors because they need A LOT! But, except in a few cases when I’ve used a “WAVE bag,” which can hold ~8L, I usually express in flasks. http://bit.ly/2OhWL9t
So let’s talk about how it all (hopefully) works. BaculoViruses (BVs) are a type of virus that infect insect cells. The name “baculovirus” is kinda confusing because it’s not a bacterium & it doesn’t infect bacteria (that’d be a bacteriophage) – the “baculo” in the name just refers to the sticklike shape the viral DNA gets packaged into when it acquires it’s protein coat to become a “nucleocapsid”. This nucleocapsid can infect other cells in different ways depending on what further packaging it gets.
A lot of viruses have linear genomes – sometimes single-stranded and sometimes written in RNA not DNA. This can make things easier for the virus, but harder for scientists to work with in the lab. Thankfully for us, baculoviruses have double-stranded, circular, DNA – similar to the plasmid DNA we can stick into bacterial cells when we want them to make stuff for us.
So scientists have modified baculoviruses to harness their insect-cell-infectiveness (and non-infectiveness for other organisms like us) to get insect cells to make specific proteins for us. Usually the BVs we use are based on naturally-occurring Autographa californica multiple nuclear polyhedrosis virus (AcMNPV) or Bombyx mori (silkworm) nuclear polyhedrosis virus (BmNPV).
Baculoviruses have a biphasic replication cycle where they spread as either a budded virus (BV) or an occlusion-derived virus (ODV). ODV is a “hardier” form they use to spread from insect to insect (horizontal transmission) – it’s coated in a protein called polyhedrin (hence the “polyhedrosis” in the name). Polyhedrin sturdifies it but also makes it harder to get out of the cells, so the virus gets the cell to stock up on bundles of ODVs called occlusion bodies (OBs), then bursts open (lyses) the cell to release them.
The OBs may be sturdy. but even they can’t survive the environment in the insect’s midgut where they end up if an insect eats it. They get dissolved, releasing the ODVs they contain. Then once they’ve infected the insect, they don’t need as much protection, so they switch to the BV form, where the virus buds out of cells through the membrane, taking on a membrane coat as they do so. That coat will help them get into new cells inside the insect (systematic infection) through receptor mediated endocytosis.
Basically, embedded in the coat is a glycoprotein (protein with a stuck-on sugar) called gp64 that lets it stick onto receptors on the membranes of other cells which then engulf it sink-hole-style (“receptor mediated endocytosis” is where the target cell basically pinches in the part of the membrane including the receptor – and the bound virus in this case). But in the late stages of infection the insect will get really sick so the virus needs to find a new host, so it switches back to the ODV form. And this requires getting the cell to make a lot of polyhedrin to form that strong coat.
Viruses are able to make different proteins at different stages in their viral lifecycle by controlling transcription. For a cell to make a protein, it first has to make a messenger RNA (mRNA) copy of the genetic recipe (transcription) and then use that copy of the recipe to make the corresponding protein (translation). DNA sequences called promoters located in front of genes help act as transcription “signal lights” – repressors can bind to give you a “red light” and promoters bind to give the “green light” to start making mRNA and protein. It’s kinda like you have a bunch of TV channels that only broadcast at certain times.
Different promoters are green-lighted and red-lighted at different times, so different proteins are made at different times depending on what the cell needs, so different combinations of channels can be airing at different times. But they’re not totally uncoordinated. If multiple proteins are under the control of the same promoter (as is the case with bacterial operons), or multiple promoters are regulated together so that they’re activated at the same time (like if the same activators bind them), they can come on like multiple TV shows all airing in the same time slot.
Inducible promoters, like the lac promoter we often use when we express proteins in bacterial cells, are like “On Demand” programming – when we add a chemical called IPTG it “green lights” it by binding to a lac promoter-bound repressor, getting the repressor to change shape and fall off so transcription can begin.
But with insect cell expression, it’s more like you take over a specific channel – time slot (like airing PBS when people think they’re getting the Disney Channel). But you still have to wait for when the show you replaced normally airs. In BEVS, we usually replace the polyhedrin gene with our gene, but we have to wait for virus to get the cell to turn it on, which happens in the “very late phase” of the baculovirus infection cycle.
During the very late phase of infection, when the virus is preparing the OBs, it has to make a lot of polyhedrin, so there’s a strong promoter in front of the polyhedrin (polH) gene that “turns green” very late in the viral lifecycle.
But the polyhedrin is only needed for the occlusion body type of infection, where it has to survive harsh conditions it might face before finding a host cell. We don’t need this because the cells are right next door!
So the key modification for turning the baculovirus into an expression system is swapping out the gene for polyhedrin with a gene for the protein you want to make, but leaving the “traffic light” in front. This way, when the cell thinks it’s turning the light green to prepare to lyse it’s actually making your protein!
But we can’t see that our protein’s being made, so we use a reporter protein that provides evidence that we’re at least in the same “time slot” – p10 is another very late protein that gets hyper-expressed around the same time as polyhedrin & helps with occlusion body formation & lysis – neither of which we need since we want the budded form of the virus.
So, just like we replaced the polyhedrin gene with the gene for the protein we really want, we can replace the p10 gene with the gene for a reporter protein – something whose “channel” we can watch. YFP is great for this because if you shine blue light on it, it it will emit yellow/green light.
So, during the cell culturing, when the cells start glowing under the microscope when I shine that light on them, I know that the p10 promoter has been activated – so I know that the very late phase time slot has started and that my protein should be getting made – but since I can’t actually watch my protein’s “channel” there could be technical difficulties or my show could have been canceled – only purification will tell!
When I pass it through an affinity chromatography column with little beads (resin) that bind specifically to a tag that I have on the end of my protein – the yellow flows right on through, while my protein stays stuck. And then I can compete my protein of the column with a tag mimic, and see what’s there – and then purify it even further.
So how does it work in practice? There are a few types of insect cells that are commonly used. The one our lab uses the most is Sf9 (which comes from the fall armyworm Spodoptera frugiperda, as does Sf21). Sometime we use Hi5 (from the cabbage looper Trichoplusia ni) which work better for some proteins, but they’re a lot “goopier” to work with. But before we can infect any of them we need to make the virus. There are several steps to BEVs.
SUBCLONING) First you have to subclone your gene into a suitable transfer vector (aka donor vector). This is just like when you clone a gene into a plasmid for bacteria -> you take the gene (the cDNA version, which is edited to remove regulatory “introns”) from one place & stick it into a donor vector (you can do this by “cutting & pasting” using restriction enzymes or “copy & stapling” using PCR-based methods like SLIC). more here: http://bit.ly/2Tkg5B1
But this vector isn’t designed for expression, instead it’s designed to help get your gene into a bacmid, which is a modified version of baculovirus DNA (no coat or anything, just the circular DNA)
BACMID PRODUCTION: Next you have to produce a bacmid containing your gene. This just requires DNA-combining & copying, not actual virus production, so bacteria are up to the task. We take the transfer vector which has our gene and combine it with the bacmid vector which has all the baculovirus-making genes using site-specific Tn7-based transposition. Our lab uses DH10MultiBac vectors which are a modified form of AcMNPV. The “multi” is because they also have a Cre-lox recombination site so you can co-express things, but I’m usually just using Tn7 transposition.
In the transfer vector, your gene’s flanked by Tn7 “donor sites” (Tn7L & Tn7R) that match up to an “attachment site” attTn7 on the bacmid, so with the help of the Tn7 transposase made from a “helper plasmid” the bacteria also host, your gene “jumps in” to the bacmid at a precise location – and this location is in the middle of the lacZ gene, so cells that successfully recombined them don’t make functional lacZ so they can’t convert X-gal into a blue product so you look for white colonies in blue-white screening to see what bacteria inserted your gene into their bacmid. more on that here: http://bit.ly/2MxNPs2
You purify the bacmid out of those bacteria similarly to how you’d purify any other plasmid (think miniprep, more here: http://bit.ly/2Ty5HVS ), but you take extra sterility measures.
V0 PRODUCTION: So the bacterial cells made more bacmid, but without the viral capsule stuff, that bacmid can’t get into the insect cells on its own. So you have to help it out through TRANSFECTION. Transfection is just a fancy term for when we stick stuff into cells – when we do it to bacterial cells we often call it transformation, but it’s just semantics.
A lot of the time, when we do bacterial transformation we use heat shock where we bathe chemically-weakened bacteria in the DNA we want to put in, then briefly heat them up to open up bacterial pores to let the DNA in. more here: http://bit.ly/2Jj7L47
But insect cells are a lot more sensitive than bacteria, so we need a gentler method. So we do our transfections with the help of a transfection agent – it’s a propriety blend, but it likely contains cationic (+-charged) molecules that coat the DNA’s negative charge so it doesn’t clash with the insect cell membrane’s negative charge & instead it sticks & gets taken in. more here: http://bit.ly/2HsWlsV
At later steps, we won’t need this “artificial” help because the infected insect cells will do the coating for us! And it’s a lot better at it so we just have to dump some virus on the cells and they’ll get infected. But at this point we only have bacmid not baculovirus, so it’s much less efficient. To help things out, we do this step on plated cells, not cells floating in suspension. The plated format maximizes the contact between the transfection agent-aided bacmid & the cells & prevents the cells from “running away” from it.
Stick some cells in a plate and they’ll settle down & stick to the base of the well. Then you add bacmid & transfection agent and wait. Cells will start taking in the bacmid but you won’t be able to tell yet. The way you can tell is when the cells start making YFP. YFP stands for Yellow Fluorescent Protein & it’s just a modified version of GFP (the green version). As a fluorescent protein, it absorbs light of one color & emits light of another color (yellow in our case, though it looks green through the scope) (more here): http://bit.ly/2HRnRQH You monitor the YFP production, keeping an eye out for when most of the cells are glowing green but they still look healthy.
During the later phases of infection, the insect cells will start releasing virus. Unlike the bacmid we took out of the bacteria, this is ready-to-ship (coated with the “landing module” and insertion machinery). This virus is capable of infecting other insect cells, but at this point, there’s not really any cells for them to infect – time to change that!
We “harvest” this original virus, which we call the V0, by using a syringe to suck up the liquid media the cells have secreted the virus into & then pushing it through a filter to get rid of any cell bits, etc. Now we have a little tube of virus. Time to give it some cells to play with!
V1 PRODUCTION: When you want to do large-scale expression, this little tube of virus isn’t enough – you’ll need a lot more – so the next step is a virus scale-up step where the goal is still virus-making, not protein-making. We take that V0 virus and add it to more cells – this time to cells growing in suspension (floating in liquid) which allow you to make a lot more. Depending on how much expression we want to use it for we normally infect somewhere between 100mL & 1L (we need ~20mL per L of expression)
When it comes to harvesting the V1, you can’t just suck it up like you did for the V0 because in the V0 the cells were already separated from the media because they were physically stuck to the plate – but here you have to do the separating yourself first.
We do this separating by centrifugation – we pour the cell & virus-containing liquid into bottles & spin them to pellet out the cells (works cuz they’re heavier than the liquid). Then we collect that liquid containing the virus and filter it. Since we have a lot more liquid now we use a bottle-top vacuum filter instead of a syringe filter. You screw the filter onto a bottle top, pour on the liquid & let a vacuum line help suck it through the filter. Our old filters took forever but our new ones seem to work a lot better. At this point you have a lot of baculovirus, so it’s time to start making a lot of protein (there’s some protein in the pellet from the V1 too so you can keep that too).
LARGE-SCALE PROTEIN EXPRESSION: To get the major protein-making going we need to add that V1 to more cells – usually we add ~20-30mL depending on cell density to 1L volumes in 2L flasks (they need room to breathe). We let them grow for a couple days, monitoring their YFP production & overall cell health to make sure all’s going well and then we “harvest” the CELLS this time – we want to KEEP THE PELLET (assuming the protein you’re expressing isn’t excreted)
We pour off the liquid into some bleach, resuspend the pellet in some clean buffer (pH-stabilized salt water) with protease inhibitor (to prevent protein-chewing), then flash-freeze in liquid nitrogen & store in the -80°C freezer until we’re ready to break them open (lyse them) & take out & clean up the protein! more on how we do that here: http://bit.ly/2JY6cWi