4
$\begingroup$

Assuming it was possible to genetically engineer / modify an adult human, would certain genes be more difficult to edit than other genes and their associated phenotypes? Would this vary by the method of modification (viral vs. pharmacological vs some application of crispr)?

For instance the Y chromosome is smaller than the X chromosome and provides and shields other chromosomes less than an X chromosome (<- I realize that this does not promote planned engineering / mutation, but it serves as an illustrative point)?

$\endgroup$
8
  • 1
    $\begingroup$ Are you asking about the difficulty in the actual modification of genes, or the difficulty in modifying different traits/phenotypes using genetic engineering? $\endgroup$
    – Brad0440
    Feb 27, 2021 at 17:32
  • $\begingroup$ @Brad0440 Good point -- phenotypes included. Edited to match. $\endgroup$ Feb 27, 2021 at 17:33
  • $\begingroup$ The difficulty is accessing, isolating and editing each specific gene depends of course very much on the specific method used for accessing, isolating and editing the genes. To give you a much simpler example: suppose you have an XML file and you want to edit every occurrence of the attribute "朝代" changing the value "南宋" to "元朝". Now, if you try to do it with a simple-minded text editor, such as Notepad, this will be exceedingly difficult and very error-prone; with a better text editor, such as Vim, it will be easier; and if you do it with a tool which understands XML, it will be very easy. $\endgroup$
    – AlexP
    Feb 27, 2021 at 17:43
  • $\begingroup$ Now we use "Cas" protein in bacteria to Photoshop our genes which is much quicker and cheaper than any other conventional methods, currently I heard it is being used to detect novel corona virus and to turn it off ;D $\endgroup$
    – user6760
    Feb 27, 2021 at 17:44
  • $\begingroup$ @user6760 I wouldn't have thought that CRISPR/Cas was even slightly suitable for fighting a virus. Do you have a reference? It could work on retroviruses (which is basically the orginal use-case for Cas) but coronaviruses are certainly not retroviruses. $\endgroup$ Feb 27, 2021 at 17:45

6 Answers 6

1
$\begingroup$

I'll make this a two part answer:

Editing genes

In adult humans, there will be some genes which are physically more difficult to modify than others. Some of the reasons for this are universal to any organism you might want to genetically engineer (including bacteria), but some are unique to higher order organisms like humans.


Most (all?) genetic engineering (GE) techniques require that a specific sequence of DNA is targeted. It is important that this sequence is unique, or at least very rare within the human genome, because otherwise you'd end up targeting other random genes which could have unintended consequences (from nothing to death).

For some genes you'll find it easy to find a relatively unique sequence to target, but for others (such as those with repetitive sequences or which are very similar in sequence to other genes), it will be difficult to target only your GOI.

As you mentioned in your question, different methods of GE do have their own unique requirements which can make editing some genes easier or harder than others. For example, the well known CRISPR/Cas9 system requires a PAM site (for the most commonly used Cas9, this is 'NGG') downstream of the sequence you are targeting. This means that the sequences you can choose to target are limited further.


Another reason some genes will be more difficult to access is because in Humans, there are many proteins which are bound to the DNA. These proteins are varied and have many roles. One example is histones. Histones help pack our DNA up so that it takes up a much smaller space in our cells than if it was just left unwound. If your GOI normally has proteins bound to it, then it will be more difficult to target your own GE machinery towards it.

Editing phenotype

This is where it gets really tricky. Most traits that humans have are controlled by multiple genes, not just one. For example, up to 15 different genes are thought to control eye colour in humans, and those are just the ones we know about! The number of genes involved in each phenotype can vary, so depending on what trait you are trying to change, you would have to modify a different number of genes, which is where the difficulty levels come in.


As well as multiple genes being involved in one phenotype, there are also some genes which influence many phenotypes. Therefore, targeting that gene may also cause other trait changes in ways which are undesirable.


So tl,dr: yes, some genes and phenotypes will be easier to edit than others, and the method of genetic modification you are using will affect which genes these are.

$\endgroup$
7
$\begingroup$

Opening a Can of Worms:

Mammalian genetics is really pretty complex. genetics can be affected by developmental influences (methylation, etc.), position relative to chromosomes and regulatory genes, the regulation of other genes in a biochemical pathway (albinism, for example, can be caused by any of several genes becoming altered/defective), trinucleotide repeats (that can vary the distance from a regulatory site and shift with generations) and so on. There are a hundred (thousand?) ways to alter the DNA and regulation of the DNA to adjust the function of a gene, which is why we can get so many versions of blonde, or curly, or dark-skinned, or web-fingered.

This is a big part of the reason why we haven't tampered much with humans. But to alter an adult human, we need to alter whole tissues or organisms. To make relatively crude alterations (for example, to insert a whole region of DNA coding for a missing enzyme) you can insert a sloppy piece of DNA into semi-random cells and as long as enough cells are able to carry out the final task (like breaking down a metabolic toxin) the change is successful.

But if the function is subtle (like triggering foveal development in newborns) there might be a narrow window of life where it can work, or the regulatory systems involved would need to be perfectly inserted in every cell, or a slight variation in regulation (due to insertion in the wrong spot) can cause a problem worse or opposite than the original issue.

It would be very specific to each change to say if there could be an easy or difficult challenge in a given genetic change. And if you want the change to be transmitted to the next generation, then the changes must affect the testes and ovaries (the cells passed on to the next generation). Methods right now aren't really up to the challenges of doing these things (CRISPR comes closest) so I can't say with much accuracy which would be best for what. They may come up with a super-retrovirus next year that flips the current state of science on it's head. THAT is the point where worldbuilding can step in and say whatever you want.

$\endgroup$
3
  • $\begingroup$ Your last paragraph would seem to imply that this is a rapidly evolving field and therefore open science based extrapolation, correct (I know this is an opinion based sub-question)? $\endgroup$ Feb 27, 2021 at 18:37
  • 1
    $\begingroup$ @EnglishmanBob It is a rapidly developing field, but if there's one thing we've learned is that the system is incredibly complex and we have no idea what we're doing. Genes have so many interactions we're mostly doing trial and error at this point. E.g., it turns out mutations in the gene which controls the number of neck vertebrae in humans also have a tendency to cause skin cancer, of all things. $\endgroup$ Feb 27, 2021 at 19:43
  • 1
    $\begingroup$ @Englisman Bob It's pretty wide open going forward, and biology is no where near as fussy or specific as chemistry or physics. It's one of the things I love about it. As long as you don't go into too many specifics, you can justify all sorts of things. $\endgroup$
    – DWKraus
    Feb 27, 2021 at 20:05
5
$\begingroup$

would certain genes be more difficult to edit than other genes and their associated phenotypes?

Modifying a gene isn't that hard. The problem is pinning down all of the effects that the modification will have. Most of the interesting things you'll want to fiddle with in people are controled by the interaction of multiple genes, and a bunch of those genes will control other features and so you may end up in the situation where you can't tweak anything without breaking something else. Ably (if not particularly seriously) illustrated by SMBC.

This is also the reason why humans don't breed true, unlike many of the plant and animal species we've domesticated. Eugenics for peas and dogs and horses: easy. Eugenics for humans, trying to get anything more complex than a particular hair or eye colour? Impractical.

Research into which genes to tweak and how to get the desired effects seems likely to be ethically dubious, to say the least. Obviously, it will happen anyway.

Would this vary by the method of modification (viral vs. pharmacological vs some application of crispr)?

All available methods that looked suitable would likely be tested and trialled if they seemed like they were working out, and the one that did the job would be selected. There's no reason to limit your toolkit, after all.

The same underlying problem of human genetic complexity will affect all of these different processes equally.

As an example of a real-world bit of human genetic engineering, look at two different trial treatments for a kind of sickle-cell disease. Both treatments have the same goal (modify donor stem cells, eradicate host stem cells, implant donor stem cells, wait for healthy blood to be formed) but one uses an engineered virus to modify the haematopoeic stemcells and the other uses CRISPR-Cas9. Both approaches appear to be practical.

$\endgroup$
3
$\begingroup$

Some genes are placed near G-quadruplex sites, structures in our DNA that can be mono, double or quadruple-stranded. Yep, Chromossomes and DNA are much more complex than what we learn in high school.

The G-quadruplex must be unwound before geelne transcription can happen. Sometimes this fails and leads to things like cancer. Some cancers are related to telomeres because that's a point with high concentration of G-quadruplex.

These structures do not show up randomly in the chromossomes, and genes close to them should be harder to edit.

$\endgroup$
3
$\begingroup$

It depends on the defect as well as the gene. For example, dystrophin is 2.3 million base pairs long. If the entire gene were deleted, you would have quite a task to replace it (though there are workarounds) - on the other hand, maybe just one base pair needs to be changed, which makes it as "simple" a problem to fix as any.

Overall, gene therapy remains unreliable with fears of cancer at every turn. However, the ability to create smaller constructs that duplicate needed aspects of a gene's function, or to "knock down" (specifically counteract) transcripts within the cell, should mean that once insertion of new sequence can be done very reliably in many cells, the overall range of difficulty in adult gene therapy should not be as large as one might imagine.

$\endgroup$
1
$\begingroup$

This article: gene therapy discusses the issue in question.

Scientists are working it, there are many complications. The article points to ongoing research. There is work ongoing to gain confidence of doing edits of replacing known bad sequence with a known good sequence.

Biggest difficulty: So many unknowns.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .