Literature Review

So, if csy1 is not the capsaicin synthase, what do current researchers think is?

The architects of our metabolic map (Mazourek et al, 2009) have written a nice paper describing how they built it (on SolCyc, before it got wrapped up in the rest of MetaCyc). About the synthase, that they leave a fork connected to so tantalizingly, they say,

- CoA-activated fatty acids form 6x more capsaicin in extracts than free CoA and free fatty acids, suggesting that there is a CoA intermediate between the fatty acid and the capsaicin synthase

-AT3, an acyltransferase which is preferentially expressed in the placenta (where the capsaicin accumulates), when knocked out, causes loss-of-pungency, has been a candidate. However, some loss-of-pungency mutants form capsaicinoids when precursors are added exogenously.

-Another point, off the CS topic but of interest to artisan transgeneticists: most of the capsaicinoid pathway has orthologs within tomato (Solanum lycopersicum), which has a great genome sequence and lots of genetic work done. The fruits actually contain vanillin.

Second paper of interest, referred to earlier, is an RNA-sequencing project which contrasts the expression of the pepper pericarp and placenta (Lei, Liu, et al 2012). Their findings?

-61 genes with specific expression in the placenta, many of which (PAL, 4CL, C3H, COMT) have already found a place in the MetaCyc biosynthetic pathway. Only 27 of those 61 genes had homology with some existing gene, indicating that the CS may be among those 34 unknown genes. Homology means that they were able to match the gene to an already-understood family of genes, generally using BLAST or some other search algorithm that matches gene sequences by the way that they line up. Homology is rarely perfect: just because you have something that really looks like an acyltransferase doesn’t mean it will necessarily transfer acyl groups, and even if it does which acyls it will transfer is hard to know.

-Among the 27 with homology, 10 acyltransferase-like genes were found, including AT3. It could be that AT3 does work as a synthase, but it is operating with other acyltransferases with the same activity. They also refer to an acyltransferase gene which Mazourek, et al, identified back in 2005, AY819027, but it is not clear if they found it.  

We can do the same search that these researchers did to find homologous genes. Pubmed has an easy to use BLAST tool, and using it on AY819027 (specifying that we only want to see Capsicum genes), we find a few genes that are very similar (acyltransferases taken from hot peppers around the Pun1 locus), and also a few that are listed as non-pungent varieties.

So regardless of the fact that this is similar to AT3, these genes are clearly not making capsaicin. This does not disprove that they could, however: they could be missense mutants, not be expressed, or else the intermediates before capsaicin may not be being made.

The proof is in the (poblano corn) pudding! If we want to know for sure what gene in our pepper makes it spicy instead of tasting like vanilla, the most definite method we have is to get these genes out of our pepper, express them in some host, and test them in a solution of of vanillylamine and fatty acid. That’s a lot of process steps to understand, so we’ve got some posts cut out for us:

-First we find the 30-some genes listed in RNA profiling paper that might have a synthase activity

-Then we’ll talk about how to purify the DNA from our peppers (which have recently sprouted, and will probably be slightly dissimilar to the genes in the paper)

-Finally, we’ll talk about how we can express these outside of the pepper, separately, so we can test their activity.

It’s a lot more work then we thought we would have to do at the beginning of this paper, but hey, that’s the burden of proof.

Genetics databases, the burden of proof

Hey everybody!

So last post we said we'd go into accessing the genetics of peppers; our goal is to find the sequence of the capsaicin synthase, so we can create some kind of interference in our peppers to prevent it from expressing.

So we can hop over to pubmed and search "capsaicin synthase" under "Nucleotide" and we get seven or eight hits, all 981 bp.

Unfortunately, we've been reading too much recent literature on the subject and this raises red flags: according to an RNA-sequencing profiling project of Lin, Lei, et al, ( 2012)

1: There is no complete sequence of the pepper genome.

2: Within the more limited sequencing projects that have occurred, no one is quite sure what the capsaicin synthase gene is.

Back to the genes we found on pubmed:

They mostly stem from a paper published in PNAS in 2006, which was retracted in 2008.  The paper is fairly exhaustive: they did a column fractionation of placental enzymes, then tested each fraction for capsaicin synthase activity, by putting the enzyme mixture in a solution of vanilllylamine and 8- methyl-nonenoic acid, and testing the amount of capsaicin that resulted on HPLC. 

They then purified the enzymes on a column containing bound vanillylamine, in the hopes that the affinity CS has for vanillylamine would purify it from other enzymes.

Now that they have a fairly pure product, they sequence it, starting with its amino acids, and go to the point of taking their genomic sequence, putting it into an E.coli vector, and making it heterologously. However, although they give a number for the activity of the heterologous csy1, there are no figures. One would think that you would want to blast that piece of proof right up front where everyone can see it. 

The retraction states that the genetic sequence shows homology to a protein kinase found later, by a M. Rapolu. Since the homology is for a small part of this kinase (981 bp for CS of > 3000 bp for the kinase), this suggests that their gene was wrong, although chances are good that they did purify some kind of synthase, before the sequencing part.

This means that there probably isn't any juice to this protein.  It also suggests that they never inserted the gene in E. coli, or didn't get any activity, or something in the media / E. coli spontaneously creates capsaicin, which is why there are no controls. 

But that is somewhat rude. If biotechnology was as easy as taking one beaker and pouring it into another, we wouldn't have to write this blog. Messing with genes is hard. Why is it so hard?

- Most genes are very hard to see, and when you purify them or do PCR on them, they mutate, sometimes in a way that makes them inactive. 

- Although it is easy enough to align genes with other genes that you may have some idea of their function, actually figuring out the function of any particular gene, in the context of the thousands of other genes, can be a combinatorially hard problem.

With that cautionary note, next time we will discuss the candidates for capsaicin synthase, and how we might go about testing them.

1. 
   

Screening and interference

Earlier we introduced you to some basics of genetic manipulation, then went into Metacyc and the wild-type world of secondary metabolism. We spoke quite casually of "knocking out" some of those genes in our nice graph, as if we could just open it up in MSPaint and use the eraser tool to clear out a section. In reality getting genes to not be expressed is something only slightly less difficult than engineering new ones that work well inside the metabolism.

The original method of messing with genes is mutagenesis and screening. Screening is a big important concept to think about in biotechnology and we're going to use it throughout this project. The essential flavor of it is that biotechnology is very imprecise, but that's okay because we're working with billions to trillions of molecular machines. In the normal engineering shop, if you want to build a house, you source some well-made materials, make careful measurements and cuts using tools that you are trained on, and work according to a plan that you can follow as you build. In the biotech shop, the general idea is that you grow a few billion two by fours, throw in a million nails and hammers, mix well, and have a selection strategy for separating the one well-built house from the hundreds of thousands of piles of debris.

Crop scientists have been doing this since the 1930s: get piles of seeds, irradiate them, and then farmers plant them all and weed out the disgusting mutants and boring regular plants. Most of the grapefruit and peppermint crops in the US were developed using this process. Good magazine suggests that the reason that these foods don't have the bad rap that GMOs do is

1) Because there is no main list of modified crops or way to test for them / people don't know about them.

2) Because radiation mutagenesis is done almost entirely by big public sector institutions that publish their results and work for the public good.

Personally we feel that atomic scientists just have better style than genetic engineers.

 

Googie

Returning to our peppers: if you wanted to knock out and select for vanillin in this method, you would have to grow every plant to the point of producing fruits, then have some chemical or taste test to identify the very few plants that would produce it. You would have to produce a huge amount of plants and take them all the way to fruiting, and it would take you many, many years, which is why we aren't going to do it this way, although we wouldn't have to buy much in the way of equipment.

Engineers these days use what is called an interfering RNA. Remember the central dogma? As you are sitting there doing the hard work of enacting it, transposons and viruses are trying to get your cells to do it for them. Freeloaders! Most multicellular organisms have a protein in charge of kicking RNA out that doesn't belong, it's called RISC, and if it finds double stranded RNA, it tries to transcriptionally silence it through the cell. In plants this kind of silencing can be passed on through generations even without actually pushing it into the genome in some way. In order to get this into our plant, we have a few options.

In the simplest case, we can just make some dsRNA and inject it straight into a plant. We can get a foundry to print us up a 400-600 bp section of the gene, buy some RNA polymerase to produce the dsRNA, and shoot up a healthy plant before the RNA falls apart. This is reasonable, but what we have heard is that this kind of intervention may or may not be systematic or long-lasting. To be more sure that our interfering RNA is being continuously produced;

We can shoot a plant up with a virus containing a piece of the gene we want to knock out: this is called virus-induced gene silencing, and there are published results of someone doing just that in C. annuum.

We can try to integrate that interfering piece into the genome of our Capsicum species (using Agrobacterium or some other transformant) This has also been done.

Unfortunately neither of those papers are immediately accessible from our local university libraries, so there will be a gap until we can talk about exactly what they did.

Next time: we'll talk more about how to access the genetics of Capsicum, both online and in the fleshy pericarp.

Secondary metabolism and Metacyc

Metabolism is a concept we generally think about in terms of diet pills and exercise: the process of burning the complex chemical fuels we need to survive or, on the plant side, turning sunlight and carbon dioxide into those same tasty fuels. Ask a biochem student what metabolism is and (s)he'll have a flashback to a huge org chart of hundreds of small molecule chemical intermediates stemming from just a few enzymatic pathways. Many of which had to be copied out by hand on one term paper or the other, like this.

(Thanks, Steven Hsu!)

 Luckily, we live in the future, and we have the internet. There is a wonderful tool for exploring these pathways which is called MetaCyc. Load it up, and type in a tasty metabolite, like capsaicin, the spicy part of chili peppers.

You get a site with a lot of gibberish and a nice SVG of capsaicin

Under the heading "Reactions known to produce the compound:" you'll see capsaicin biosynthesis.
 
Click on it! 

Now we're in business: those yellow words on the right are the enzymes that the plant uses to transform one chemical intermediate (red) into another. So here is something interesting: capsaicin, a profoundly spicy chemical, is made from the conjugation of a long chain that looks like a fatty acid chain to vanillylamine, which is one step off vanillin. Vanillin is just like it sounds: the main taste component of vanilla beans, a strong taste in a totally different direction. So here is where your artisan transgenic foodie ears should prick up: if one of those genes upstream of the final product were knocked out, would you get a pepper that tasted like vanilla?


In a word, maybe. The reality of enzyme kinetics in a plant are very different from this well organized chart. We know, for example, that many peppers are not pungent at all, while the ones that produce capsaicin produce it only in some tissues (the placenta around the seeds), and at levels that vary hugely: from the relatively tame Jalapeno to the current record holder, the Bhut Jolokia, you have a difference in capsaicin from 200 - 50,000 ppm.

Scrolling down to the bottom of the page, you see the academic papers the curator used to create this graph. If you really want to knock some of these enzymes out and see what happens, you should read those first. Especially the one that talks about "capsaicin synthase", which is the missing enzyme that connects the two main precursors at the bottom.

Metacyc is great to troll around before getting into the nitty gritty of how to modify, but in the end you should read the papers that people have published about actually intervening in those genomes, or what mRNA transcripts come out of pungent / non-pungent peppers. We can wait.

Whither transgenics?

Before we get into a primer about plant metabolism and the things that are and are not easy to do with plants and their genomes, lets get conceptual about some common GM crops.

Most people have heard of Bt corn, there is also Bt cotton, soybeans, potatoes etc. Bt stands for Bacillus thuringiensis, a   microbe whose spores and preparations have been used to control parasite infestations in crops since the 1920s. In its ideal state, Bt-plants have been modified so that every cell of the plant produces one or more cry proteins, which pierce holes in the midgut of many insects, killing them. This protein is fairly directly expressed from a gene in B. thuringiensis, so as long as the plant genome has those cry genes and it doesn't cause toxicity to the plant, those plants will be toxic to pests.

The first commercial Bt-plants were tobacco plants grown back in 1985. This modification is a fairly simple expression of the central dogma of biology: DNA->RNA ->protein. As long as the central dogma holds true, this modification should work.

Toxin expression is fairly boring to the hobbyist, however, and although you could produce other valuable proteins, like insulin or human growth hormone, non-central dogma concerns make this pretty much non-viable. Plant proteins, post-translation, are modified differently than human ones, and we aren't going to go into the difficult and not clinically tested methods for "humanizing" plants.

Introduction.

Backyard Biotech! A build-out blog for transgenic hobbyists. Here we'll discuss how to create your very own genetically modified plants from the dirt up.