39 ms·

Ask HN: Can you program B cells to make antibodies for viral spike proteins?

1 points by respinal 6y ago

JPLeRouzic 6y ago
It is possible to change the DNA of a cell to make changes in protein expression. There are several methods, some low cost like using a virus vector [0], some much more complex and costly. See for example the Astrazeneca candidate vaccine [1] which creates spike-specific antibodies. [0] https://en.wikipedia.org/wiki/Vectors_in_gene_therapy https://en.wikipedia.org/wiki/Vectors_in_gene_therapy [1] https://en.wikipedia.org/wiki/COVID-19_vaccine#Vaccine_candidates https://en.wikipedia.org/wiki/COVID-19_vaccine#Vaccine_candi...
- respinal 6y ago
  Thank you for your reply! I was asking more specifically, could not write it due to characters limitation, about programming B cells (or any antibody generating system) to make antibodies to target viral spike proteins (peplomers), in an interactive or dynamic way (changing as virus mutates) , until a strong bond is reached/found (e.g., YES/NO decision). I was reading that most viruses have viral spike proteins, so if someone could do that it would be great for the treatment and detection of viral infections. Also, it could save a lot of time and money.
  - JPLeRouzic 6y ago
    As far I understand, you want a better biological mechanism than the current B-cell? There are three problems: 1. Making a nano machine like a B cell, possibly by re-engineering a B cell with genetic therapy. It is a challenge in the realm of synthetic biology. 2. Making a smart detector for viruses, something that is done already in the innate immune system, but obviously it should be better and be equipped with a "stop" system. 3. Producing an antigen/antibody. A difficulty is that not all humans react in the same manner to antibodies. This is why designing vaccines is so hard. You should look at MHC I and II complexes. There are hundreds or more MHC subtypes and each subtype roughly means a different antigen. There are prediction servers that help to design antigens [0]. [0] http://www.cbs.dtu.dk/services/NetMHCpan/ http://www.cbs.dtu.dk/services/NetMHCpan/
    - respinal 6y ago
      Yes, but not better per se. I would say more adaptive or controllable (which is probably very hard to do). Nature/biology has been doing that by trial and error for years, but it takes time and it is not very efficient. I am more interested in #2 as this could significantly improve and make robust detection/sensing systems, since viruses constantly mutate changing the binding affinities of detection/sensing systems. My background is in bioanalytical chemistry, so as soon as I read about the commonality of viral spike proteins among viruses that idea crossed my mind. Definitely, the human body is a complex system. So if you block the viral spike protein of a virus (e.g., SARS-CoV-2), that could have some adverse effects in some people? Interesting. Thank you for your comments! Very helpful!
      - JPLeRouzic 6y ago
        Thank you also for the discussion. Controlling a cell population in the body is yet another great challenge! Maybe it could be more feasible if there were two sub-systems: 1. A new virus detector, it doesn't have to be based on B cells, an implantable device could perhaps fill that role at least as a first step. 2. A platform can produce a personalized vaccine. I made a toy version of this a few years ago for peptide vaccines [0]. 2.1. On the software side there are minor challenges. 2.2. On the legal aspect of a personalized vaccine, its transport and administration, it is much more complicated, those are real challenges. [0] https://padiracinnovation.org/en/peptide_PoC/ https://padiracinnovation.org/en/peptide_PoC/
        respinal 6y ago
        You are welcome! For #1, I totally agree, used B cells as an example, and also implantable devices/sensors could bring more challenges (foreign body response, encapsulation, a big issue for implantable sensors). I am thinking more of a new virus detector that can be used in vitro or outside that can automatically adapt to viral changes (mutations). Let's call it "molecular morphing" virus detector, if you will (https://www.researchgate.net/figure/Principles-of-the-molecular-morphing-approach-Molecular-morphing-generates-a-path-in_fig3_261032438 https://www.researchgate.net/figure/Principles-of-the-molecu...). For #2, sweet! Thanks for sharing! The new virus detection platform could be used to create personalized vaccines, yes. For #2.1, oh. Software can definitively do things faster like try million iterations until they find the "perfect match" or strong bond (from my previous comments) for viral spike proteins. For #2.2, also agree, I am more interested in the detection/screening side of things for diseases in general.

1 points by respinal 6y ago

It is possible to change the DNA of a cell to make changes in protein expression. There are several methods, some low cost like using a virus vector [0], some much more complex and costly. See for example the Astrazeneca candidate vaccine [1] which creates spike-specific antibodies. [0] https://en.wikipedia.org/wiki/Vectors_in_gene_therapy https://en.wikipedia.org/wiki/Vectors_in_gene_therapy [1] https://en.wikipedia.org/wiki/COVID-19_vaccine#Vaccine_candidates https://en.wikipedia.org/wiki/COVID-19_vaccine#Vaccine_candi...

respinal 6y ago
Thank you for your reply! I was asking more specifically, could not write it due to characters limitation, about programming B cells (or any antibody generating system) to make antibodies to target viral spike proteins (peplomers), in an interactive or dynamic way (changing as virus mutates) , until a strong bond is reached/found (e.g., YES/NO decision). I was reading that most viruses have viral spike proteins, so if someone could do that it would be great for the treatment and detection of viral infections. Also, it could save a lot of time and money.
- JPLeRouzic 6y ago
  As far I understand, you want a better biological mechanism than the current B-cell? There are three problems: 1. Making a nano machine like a B cell, possibly by re-engineering a B cell with genetic therapy. It is a challenge in the realm of synthetic biology. 2. Making a smart detector for viruses, something that is done already in the innate immune system, but obviously it should be better and be equipped with a "stop" system. 3. Producing an antigen/antibody. A difficulty is that not all humans react in the same manner to antibodies. This is why designing vaccines is so hard. You should look at MHC I and II complexes. There are hundreds or more MHC subtypes and each subtype roughly means a different antigen. There are prediction servers that help to design antigens [0]. [0] http://www.cbs.dtu.dk/services/NetMHCpan/ http://www.cbs.dtu.dk/services/NetMHCpan/
  - respinal 6y ago
    Yes, but not better per se. I would say more adaptive or controllable (which is probably very hard to do). Nature/biology has been doing that by trial and error for years, but it takes time and it is not very efficient. I am more interested in #2 as this could significantly improve and make robust detection/sensing systems, since viruses constantly mutate changing the binding affinities of detection/sensing systems. My background is in bioanalytical chemistry, so as soon as I read about the commonality of viral spike proteins among viruses that idea crossed my mind. Definitely, the human body is a complex system. So if you block the viral spike protein of a virus (e.g., SARS-CoV-2), that could have some adverse effects in some people? Interesting. Thank you for your comments! Very helpful!
    - JPLeRouzic 6y ago
      Thank you also for the discussion. Controlling a cell population in the body is yet another great challenge! Maybe it could be more feasible if there were two sub-systems: 1. A new virus detector, it doesn't have to be based on B cells, an implantable device could perhaps fill that role at least as a first step. 2. A platform can produce a personalized vaccine. I made a toy version of this a few years ago for peptide vaccines [0]. 2.1. On the software side there are minor challenges. 2.2. On the legal aspect of a personalized vaccine, its transport and administration, it is much more complicated, those are real challenges. [0] https://padiracinnovation.org/en/peptide_PoC/ https://padiracinnovation.org/en/peptide_PoC/
      - respinal 6y ago
        You are welcome! For #1, I totally agree, used B cells as an example, and also implantable devices/sensors could bring more challenges (foreign body response, encapsulation, a big issue for implantable sensors). I am thinking more of a new virus detector that can be used in vitro or outside that can automatically adapt to viral changes (mutations). Let's call it "molecular morphing" virus detector, if you will (https://www.researchgate.net/figure/Principles-of-the-molecular-morphing-approach-Molecular-morphing-generates-a-path-in_fig3_261032438 https://www.researchgate.net/figure/Principles-of-the-molecu...). For #2, sweet! Thanks for sharing! The new virus detection platform could be used to create personalized vaccines, yes. For #2.1, oh. Software can definitively do things faster like try million iterations until they find the "perfect match" or strong bond (from my previous comments) for viral spike proteins. For #2.2, also agree, I am more interested in the detection/screening side of things for diseases in general.