Gene Drives: so you want to change the world!

By Rich Feldenberg:
Want to change the genetic landscape of whole populations and ecosystems? Tired of having to do it the old fashioned way by genetically engineering one organism at a time? Well now there’s Gene Drive! The fast and efficient way to spread your desired genetic design! Just send $19.99 plus shipping and handling for your Gene Driver Kit today!

If the “Fake Advertisement” in the paragraph above, made it sound as though the Gene Drive concept is some crazy kind of internet scam that is to good to be true, actually nothing could be further from the truth. Well no, you can’t just send in money for a gene drive kit yet, but it turns out that gene drives are real, they’re awesome, they’re controversial, and they can in principle, change the gene pool of an entire population of an organism. In fact, this method of gene editing is so new that very few experiments have even been done, and its founder, Kevin Esvelt, feels that the technology is so powerful that he wants to put a halt on experimentation until society can come together and discuss whether we collectively feel this is an area of science we should pursue, not just one that we can pursue. To understand gene drives we first have to remind ourselves of how the CRISPR-Cas9 system works, which I reviewed in an earlier Darwin’s Kidney post.

Briefly, the CRISPR-Cas9 system is a new and powerful gene editing technique that can be used in living organisms. This system is found naturally in many bacteria, as part of their immune defense mechanism against viral attack. There are two major parts to the system. The first is a guide RNA and the second is the Cas9 enzyme. The guide RNA is a small strand of RNA (somewhere around 20-40 base pairs in length). When the guide RNA finds a perfect base pair match with a DNA strand somewhere in the cell, the Cas9 enzyme cuts that piece of DNA. In the case of bacteria, this allows the them to match one of their guide RNAs to a sequence of DNA from an invading virus, then cut the viral DNA, which disables the virus from taking over the bacterial cell. The guide RNA came from a previous viral attack that the bacteria survived, and when the bacterial enzymes chopped up the invaders DNA into small bits, some was incorporated into CRISPR so that exposure to that same virus the next time would quickly result in recognition by the bacteria – an immune system! In the last few years scientists have discovered how to make guide RNA for any desired gene, and along with the Cas9 enzyme, can then “snip out” the gene or any piece of DNA in question. This can be used to silence genes, or can also be used to replace genes if the cell has access to a DNA sequence that can fill the gap left by the Cas9 enzyme. This may turn out to be a great way to cure genetic diseases through gene therapy.

Gene drive systems, take this concept a step further. Gene drives rely on the gene editing to take place in germ line cells versus somatic cells. Germ-line cells are the cells that will become egg or sperm, and will be used to create new organisms through sexual reproduction. Somatic cells are all the other body cells, such as skin, kidney, brain, pancreas, etc. If a gene is edited in a somatic cell, that change will effect the organism in whom the change was made, but would not be passed down to the next generation.

As an example, lets say you want to be able to provide gene therapy for a genetic disease such as Nephrogenic Diabetes Insipidus (NDI). This disorder is X-linked, meaning that the gene is on the X chromosme. Since males have an X and Y chromosome, with the X coming from their mother and the Y coming from their father, if the mother’s X chromosome has the mutant gene for NDI they will have inherited the disease, which leads to the kidneys inability to regulate water loss. People with this disorder can die of dehydration because even when dehydrated they continue to produce too much urine. A female, having two X chromosomes, one from her mother and the other from her father, might be a carrier for NDI if her mother’s X had the mutant NDI gene, but she still wouldn’t develop the actual disease since her normal NDI gene from her father’s X chromosome will compensate.

In principle you could use the CRISPR system to edit the defect gene, so that the male patient with NDI can now regulate water loss through the kidneys normally. There is still no way to really do this yet. You would need to deliver the CRISPR-Cas9 system, to the appropriate kidney cells of the affected individual. At the present time, a way to target and deliver the system is still not available, but if it could be delivered to the kidney cells it would excise the defective DNA. The cells own repair mechanisms will then look for a replacement to fix the DNA break made by the CRISPR-Cas9. If the normal gene was also delivered to the cell it will be incorporated into the place where CRISPR-Cas9 made the break. This will result in having removed the defective disease causing gene and replaced it with the normal healthy gene, and should therefore cure the disease – Nephrogenic Diabetes Insipidus kidney disease in our example. However, even if this could really work – its never been tried yet for this disease – but was unsuccessfully attempted for Hemophilia, the cured individual would still be able to pass the disease on to their children. The reason is that only the kidney cells were altered, and not the germ-line cells.

Gene drives, on the other hand, effects the germ-line, but they have an even bigger, more ingenious twist to their potential to alter future generations. With gene drives, in addition to supplying the new gene, the genetic code for more CRISPR-Cas9 is also inserted into the target genome. So here is how it might work. Let simplify the example by calling the two alleles of the gene (one allele comes from mom and the other from dad) as Normal and Engineered. It could be any gene in the genome that you’re interested in, such as the gene for making insulin or for making neurotransmitters in the brain, or transcription factors that tell more genes what to do. In this example we want the Engineered gene to take over because it has some trait we have engineered for it that we find desirable. It could be to fix a defective gene or it could be to give the organism some new property. We’ll get to some examples of new properties shortly.

 

The first step might need to take place in the lab when the organism at its earliest stage, the fertilized egg. You place the CRISPR-Cas9 into the fertilized egg with guide RNA that recognizes the Normal gene. You also place the Engineered gene in the cell to replace the Normal gene once Cas9 has cut it. The cell will now have the Engineered gene as part of its entire genome. This will effect both somatic cells and germ-line cells since the fertilized egg will continue its job of dividing into more and more cells, which will eventually become all the cells of the body. Eventually this organism will develop into an adult and find a mate to produce more offspring. The offspring will have an approximately 50% chance that the Engineered gene will be passed on to the the next generation. That is because each offspring will get one copy of Engineered gene from our genetically modified organism and the other gene from its mate, which would carry the Normal gene version since it was never modified. So this is where the special ingenious twist comes in!

Not only does the gene you inserted into the fertilized egg contain the DNA of your engineered gene, but it contains the DNA for making a CRISPR-Cas9 system, as well. This CRISPR-Cas9 is hidden somewhere in the middle of your Engineered gene so that the cells DNA repair enzymes don’t recognize it as being novel to the cell. They only recognize the ends that need to fit in the space that Cas9 cut out. So in this way, the gene we pasted into the genome is Engineered-CRISPR-Cas9. Now when the cell transcribes that gene the CRISPR-Cas9 is also transcribed which leads to a guide RNA and a working Cas9 enzyme. The guide RNA will then match to the Normal gene and Cas9 will cut it. This is important because when the genetically modified organism mates with a wild type organism the offspring will have one Normal gene from the wild type and one Engineered-CRISPR-Cas9 gene from the genetically modified organism. CRISPR-Cas9 then gets transcribed, seeks out the Normal gene, and replaces it with the Engineered-CRISPR-Cas9 gene, so the offspring actually ends up with two copies of the Engineered-CRISPR-Cas9 gene. In this way the rate of transfer of the Engineered gene, to successive generations, goes from 50% to 100%. The Engineered-CRISPR-Cas9 gene effectively edits every other allele that matches its guide RNA, turning it into the Engineered gene. Now the gene can spread rapidly through a population because the odds are always in favor of this gene being passed down to all offspring. See the excellent Mosquito chart in the following article I’ve linked to in order to get a visual on how the inheritance would be effected.

It should be pointed out that this is only effective for organism that reproduce sexually. Asexually reproducing organisms (such as bacteria) won’t be influenced by this mechanism. For organism that have short generation time this is ideal. One proposed problem that gene drives might be able to solve would be in the fight against malaria. Malaria kills millions of people each year (see Darwin’s Kidneys article: Diseases with an Upside). If mosquitos were released into the wild, that were engineered to have a malaria resistant gene and also the CRISPR-Cas9 system, then that gene would spread rapidly throughout the mosquito population. The result would be malaria resistant mosquitos and possibly an end to suffering and death in many parts of the world due to this parasitic infection.

There’s no guarantee, however, that the malaria organism – Plasmodium – would not find a way to evolve around the mosquito’s malaria resistance given enough time. There is also no guarantee that the malaria resistant gene might not somehow decrease the “genetic fitness” of the mosquito making them less likely to survive and reproduce. Mosquitos would be an ideal organism for this type of engineering, however, since they have a rapid generation time, so within several years to decades a gene system of this type could theoretically pass to all members of the population. Humans, on the other hand, reproduce slowly so a gene drive in humans would probably take hundreds of years to spread through the population. Still, you could imagine an attempt to eliminate many genetic diseases completely from existence by using gene drives that over the course of centuries might be effective. One could also imagine the ability to produce a civilization of future generations of humans that are more intelligent, more rational, less violent, more empathetic, and so on, if the genes involved in producing those traits could be identified. It is harder to imagine, however, that society as a whole would ever agree to such a mass alteration of the human genome – creating something beyond human – by directing human evolution in a desired direction. Its too early to know if such changes to the human genome could even be done safely without creating damaging consequence that are impossible to predict. I’m not necessarily advocating for changing the human race for the better, but more just advocating for discussion of the potential positive and negative effects might result from such grandiose dreams.

Because the implications for gene drives are so powerful and large scale, there is currently a call for a hold on research until the ethical considerations can be more fully considered. I think this seems wise at our current state of understanding. Changing an ecosystem could have unforeseen consequences. There may be ways to alter some behavior in organisms with gene drives that would not necessarily eliminate those organisms from the ecosystem – and so may have a mild impact on the ecosystem as a whole. For example, one could engineer a pest to dislike the taste of a crop that it normally damages, and therefore protect the crop without the need for as much pesticide use. The pest is now no longer a pest, but remains in the ecosystem where it can feed on other plants and remain part of the normal food chain for other organisms. Could gene drives be used to engineer plants to more efficiently remove CO2 from the atmosphere, and combat global warming while increasing crop yields?

Gene drives are an exciting new method of changing the genetic makeup of populations of organisms. Whether they will be used to prevent diseases like malaria from killing so many or making crops less prone towards pests and therefore reducing the amount of insecticides released into the environment, is up to society at large to decide if we are ready to pursue such far reaching technology. My hope is that we may find ways to safely use gene drives to improve life on planet earth for ourselves and our fellow species.

References:
1. “Genetically Engineering Almost Anything” by Tim De Chant and Eleanor Nelson, Nova Next. July 17, 2014.
http://www.pbs.org/wgbh/nova/next/evolution/crispr-gene-drives/
2. “Gene Drives and CRISPR could revolutionize ecosystem management”, by Kevin Esvelt, George Church, and Jeantine Lunshof; Scientific American Blog. July 17, 2014.
http://blogs.scientificamerican.com/guest-blog/gene-drives-and-crispr-could-revolutionize-ecosystem-management/
3. Gene Drive Wikipedia: https://en.wikipedia.org/wiki/Gene_drive
4. “Gene editing in Humans”; Neurologica blog by Steven Novella; Nov. 19, 2015
http://theness.com/neurologicablog/index.php/gene-editing-humans/
5. “CRISPR: what’s the big deal?”, Darwin’s Kidney blog by Rich Feldenberg. Nov. 28, 2015.
http://darwinskidneys-science.com/2015/11/28/crispr-whats-the-big-deal/
6. “Can we genetically engineer Rubisco to feed the world?”; Darwin’s Kidney blog by Rich Feldenberg.
July 22, 2015.
http://darwinskidneys-science.com/2015/11/28/crispr-whats-the-big-deal/
7. “Diseases with an upside”; Darwin’s Kidney blog by Rich Feldenberg. July 29, 2015.
http://darwinskidneys-science.com/2015/07/29/diseases-with-an-upside/
8. “Live at the NESS: New Dilemmas in Bioethics”; The Rationally Speaking Podcast. April 24, 2011.
With Massimo Pigliucci and Julia Galef as hosts.
http://rationallyspeakingpodcast.org/show/rs33-live-at-necss-new-dilemmas-in-bioethics.html

9. “Sculpting Evolution”; website of Kevin Esvelt, PhD.  Founder of gene drives.   http://www.sculptingevolution.org/kevin-m-esvelt

 

 

 

Can we genetically engineer Rubisco to feed the world?

Today’s atmosphere is brought to you by Rubisco.

Fine makers of oxygen since 3.5 billion B.C.

By Rich Feldenberg

If you happen to peak outside on a nice sunny summers day to admire the green grass, shady trees, and pleasant bushes, your field of view is, in actuality, filled with Rubisco, busily helping the plants do their special thing of making sugar and churning out oxygen.  Rubisco is by far the most abundant enzyme on the earth and accounts for 30%, or more of the protein found in the green leaves of plants.  Without it there would be no oxygen producing photosynthesis, so if you’re a fan of breathing then you’re probably going to be happy to learn about Rubisco.  And, If you’re thinking to yourself, “if there is that much of it in the world it must be doing something important”, congratulations, you’d be right!

There are three things I’d like to point out about Rubisco. One is that Rubisco is freaking amazing.  It is an awesome protein with interesting molecular properties, catalyzing fascinating chemistry, and dating back to some of the earliest life on earth.  The second thing is that as amazing as Rubisco is, it is incredibly poorly mesigned (mesigned is my word meaning designed by natural selection).  Rubisco is horribly inefficient and slow, and it’s amazing it hasn’t been fired from it’s post and replaced with a new, younger, more hip version.  And finally, Rubisco could, in principle, be engineered to be much better, possibly increasing crop yields to feed an increasing global population, and removing CO₂ from the air to combat global warming.  Let’s tackle each of these points.

Point One:, Rubisco is amazing.  Without it life on earth would likely still be living as simple single celled mats of slime on the ocean floor in a oxygen free world.  Rubisco is the abbreviated form for the formal name of the enzyme Rubulose-1,5-bisphosphate carboxylase/oxygenase.  Yeah, that’s why Rubisco (rhymes with San Francisco) rolls off the tongue so much nicer, and is way easier to say three times fast.  This enzyme goes way back to the good old days when singled cell organisms ruled the world, and appears to have a common origin in all three of the major kingdoms of living things -bacteria, archaea, and eukaryotes- indicating a very early origin sometime around the time of the Last Universal Common Ancestor (LUCA).  It appears to have arisen even before the evolution of oxygen producing photosynthesis.  There are, by the way, other types of photosynthesis that do not produce oxygen as a byproduct – so called anaerobic photosynthesis- which are less efficient than the oxygen producing types.  Based on the study of related proteins, known as Rubisco-like proteins (RLPs), Rubisco may have evolved from RLPs that performed other enzymatic functions before it was eventually modified to it’s modern role in photosynthesis.

Rubisco is a big protein, that is itself composed of two main types of subunits – the large subunit (L) and the small subunit (S).  There are three major forms of Rubisco, but in most plants and algae, the Rubisco is composed of a combination of eight L-subunits and eight S-subunits.  Rubisco catalyzes the first step in the photosynthetic process taking CO₂ and making it react with the compound ribulose 1,5-bisphosphate (RuBP).  That is way the Rubisco enzyme is named Rubulose-1,5-bisphosphate carboxylase/oxygenase.  In one of it’s reactions, it is carboxylating the substrate RuBP.  This leads to the formation of two molecules of phosphoglyceric acid (PGA), that then go through a metabolic pathway called the Calvin cycle.  The PGA products eventually go on to other metabolic pathways, and the result is sweet sweet sugar!

Space filling model of the Rubisco protein structure 

The L subunit has the active site with a critical lysine residue for binding CO₂.  It actually takes two CO₂ molecules to get things going.  The first CO₂ molecule is used just as an activator for the enzyme’s active site, but isn’t used in the carboxylation reaction.  The second CO₂ molecule is what is used to react with RuBP, and it is this carbon that is added onto the molecule.  The O₂ that is eventually released at the end of photosynthesis does not come from the CO₂ but is taken from water, which is also necessary in the reaction.   The genes for the L subunit of Rubisco are found in the chloroplasts, tiny organelles within the cells, where Rubisco is conducting its important job.  The S subunit is more of a stabilizing part of the protein and its gene is located in the nucleus, and once the protein is made, needs to be shuttled into the chloroplasts.

 

 

Also necessary for the enzyme to function is the presence of an ion of Mg⁺² ,which acts to stabilize the activation site.  This process allows one CO₂ molecule, along with a molecule of H₂O to become incorporated in RuBP.   There is a whole lot of Rubisco in the green leaves of plant to carry out this important chemical reaction.   As we said above, about 30% or more of the protein in the leaves of plants is in the form of Rubisco, so it therefore accounts for a huge amount of nitrogen stored in the biosphere, since proteins contain nitrogen as part of their structure.  

Point Two: Rubisco is so mind numbingly inefficient I am almost embarrassed for our plant cousins.  It turns out that the carboxylase function of Rubisco (you know the really important thing it does by taking CO₂ from the air and attaching it to RuBP to begin the process of making carbohydrate) is not the only reaction it performs.  In fact, it’s very name -the unabbreviated one that is- tells you right off that it it also is an oxygenase.  That is the carboxylase/oxygenase last portion of the name.  This means that Rubisco is not terribly selective for CO₂, but can also react at the activation site with a molecule of molecular oxygen (O₂), which has some chemically similar properties.  This leads to a horribly counterproductive metabolic pathway called photorespiration.  In other word, it is not very selective, and is so nearsighted that it may grab onto an O₂ as easily as a CO₂.  Normally about 25% of the reactions that Rubisco is catalyzing are with oxygen going down the photorespiration pathway.  Also, keep in mind that in the atmosphere today, and for at least the last billion years, the concentration of O₂ has been way in excess of that of CO₂.  The atmosphere these days is 21% O₂ and only 0.04% CO₂, so that makes it even more difficult for poor little Rubisco to discriminate effectively.

Ribulose 1,5-bisphosphate (RuBP)

Photorespiration leads to RuBP being converted into one molecule of PGA and one molecule of 2-phosphoglycolate.  This doesn’t lead to carbohydrate production.  Even worse this uses energy in the form of ATP and released CO₂ into the air.  Totally wrong if you want to store energy from the sun in the form of yummy sugar molecules.  So we can clearly say that Rubisco has a poor affinity for CO₂.  An enzyme’s affinity for its substrate is measured by a characteristic called K_m, and Rubisco’s K_m is kind of wimpy.   This relative non-selectivity may be a reflection of the world in which Rubisco first evolved.  At that time the concentration of CO₂ in the atmosphere would have been much higher and the concentration of O₂ would have been extremely low since photosynthesis was just getting started.  Rubisco probably didn’t need to be too selective since O₂ was just a trace gas back then.

The selectivity of Rubisco for CO₂ over O₂ is affected by temperature.  Warmer temperatures decrease the selectivity making Rubisco even more inefficient.  That can be a problem for plants in a hot dry climate.  Also a change in the amounts of CO₂ to O₂ with respect to each other will influence the enzyme efficiency.  These two factors may become a significant concern in a world of global climate change where the both the temperature and concentration of CO₂ are on the increase.  How this could affect the world’s already insufficient food supplies will have to be seen.

Besides it’s affinity for reacting with a substrate, another characteristic of an enzyme is it’s rate of reaction called the Vmax.  Guess what, Rubisco’s Vmax also really sucks.  Probably not what you would expect for an enzyme that is the most abundant in the world.  Where as most enzymes catalyze thousands of reactions per second, Rubisco is only able to catalyze about 10 reactions per second.  Now, I hate to sound so judgmental, but that is really pathetic!  It is certainly possible that Rubisco was never able to evolve to be more efficient due to constraints on its structure once it became vital to the plant way of life.  Any alteration in the critical active site may have affected too many other protein-protein interactions necessary for normal function, and so never took place.  Alternatively, there may be some advantages to photorespiration, after all, so that completely shutting down that pathway would, likewise, be detrimental to growth.  There seems to be a trade off between having organisms who’s Rubisco has good affinity for CO₂ (K_m) and those who’s Rubisco has a fast reaction rate (Vmax).  It’s a case of, you can’t have your cake and eat it too.  If you favor one quality then you suffer in the other.

Plants have come up with a few smart ways of helping to boost the efficiency of Rubisco.  One way is to attempt to concentrate the amount of CO₂ around the enzyme.  C4 plants do this by adding the carbon from a CO₂ molecule to phosphenolpyruvate (PEP), then through a series of chemical reactions, the organic compound malate is produced.  The malate is shuttled to the plant cells that contain Rubisco and the CO₂ is removed.  This concentrates the CO₂ in the vicinity of Rubisco so it can act more efficiently.  The waste in energy to produce the malate is more than made up for by the better efficiency of the Rubisco in C4 plants due to this CO₂ concentrating ability.  C4 plants are a more recent evolutionary development, but only represents about 3% of land plants.  They are well suited for living in desert conditions where C3 plants would not be able to photosynthesize effectively and would rapidly lose too much water.  C3 plants to well in moderate climates with only moderate sun light.  The lower temperatures helps to improve Rubisco efficiency at utilizing CO₂ over O₂.  

 

C4 plants are therefore more efficient, especially in warm dry climates.  CAM (Crassulacean acid metabolism) plants close their stomata in the day to prevent fluid loss and open them at night to allow diffusion of CO2 into the leaves, where it is stored in malate.  During the day the CO₂ is again removed from the malate so it can be used by Rubisco to make carbohydrate.  CAM plants can be either land or aquatic.  

 

The C4 plant, Maize, busily concentrating CO2 to boost Rubisco efficiency

Point Three: Perhaps we can make a better Rubisco, one that can select CO₂  over O₂ more effectively, and react more quickly.   Nature has had billions of years to figure this out, so maybe its our turn now to design a Rubisco that can be improved in a variety of different ways.  This might be accomplished by artificial selection or genetic engineering – to produce a Genetically modified organism (GMO) with the desirable traits we choose.  In fact, there is a great deal of research looking into possible ways to improve Rubisco, but so far progress seems to have been rather modest.  A super Rubisco could in theory produce more carbohydrate under warmer, drier, and lower light conditions, decrease the amount of nutrient nitrogen necessary for plant growth, and remove more CO₂ from the atmosphere, and release more O₂.  This could be vital to consider for a growing global population that is outstripping its food resources and heading towards potential disaster due to global warming.  How could it be done?

It is known that red algae has a Rubisco with the highest value yet found for CO₂   affinity.  It is nearly 3 times better at discriminating CO₂ from O₂ than is the Rubisco from crops, like corn.  It may be theoretically possible to engineer crop plants to express the red algae Rubisco.   Other studies have looked to genetically engineering Rubisco by substituting key amino acid residues in critical areas of the enzymes protein structure and observing the effect.  This has resulted in some mild success.  In one study, by switching a particular alanine residue in the L subunit with a asparagine, the affinity was increased by 9%.  Not a huge increase, but potentially a good starting point.  

Other research has focused on speeding up Rubisco’s slow rate of reaction.  One way to accomplish this could be to create a CO₂ concentrating mechanism in C3 plants like corn and rice, that is similar to the natural CO₂ concentrating mechanisms found in C4 plants.  The ways to make this happen are less clear, but could involve manipulations that would put certain types of CO₂ transporter in the membranes of chloroplasts to help concentrate the gas where it needs to be.  

It should also be noted, that while the intended effects for changes to Rubisco protein would be for the common good of the planet, if we get to the point where such genetically modified plants are possible, it would need to be studied, not only to determine that there are no unintended consequences on the environment, but also that these changes actually result in greater plant growth and yield.  There may be some reasons why photorespiration is allowed to occur at the high rates it does.  One theory is that this is a protective mechanism for the plant so that in intense light conditions energy overload does not occur that could result in oxidative damage to the plant.  There may be a certain limiting factor where carboxylation can be maximized to a certain degree, but once you cross some threshold it actually becomes detrimental to the organism.  

There is no doubt that Rubisco is a curious and fascinating protein, and one on which our lives, and continued survival, are completely dependent.  It is certainly worthy of our admiration for its important and ancient role in maintaining earths biosphere.  There seems to be much more we need to understand about its biochemistry before we can tell if it will be a tool we can utilize to improve and protect our planet.  It could also potentially be altered in algae or cyanobacteria to terraform other planets like Mars, which although it has an extremely thin atmosphere, does have an abundance of CO₂ over O₂.  If we eventually discover life on other world that have evolved some form of photosynthesis, it will be interesting to learn what proteins or other methods they came up with to catalyze the carboxylation reaction that Rubisco serves for us here on earth.

References:

1. M. A. J. Parry, et al., “Manipulation of Rubisco: the amount, activity, function, and regulation”. Journal of Experimental Botany, Vol 54, No. 386,  pp. 1321-1333, May 2003.

2. Spencer M. Whitney, et. al., “Advancing our understanding and capacity to engineer natures CO2-sequestering enzyme, Rubisco”, Plant  Physiology, Vol. 155, pp. 27-35, Jan. 2011.

3. Wikipedia article on Rubisco:  https://en.wikipedia.org/wiki/RuBisCO

4. Wikipedia article on Photorespiration:  https://en.wikipedia.org/wiki/Photorespiration

5. Article on Mesign:     http://darwinskidneys.blogspot.com/2015/07/another-clever-mesign-brought-to-you-by.html