ncborcherding / scRepertoire

A toolkit for single-cell immune profiling
https://www.borch.dev/uploads/screpertoire/
MIT License
311 stars 54 forks source link

Problems in analyzing the VDJ of the antibody of a quasi-monoclonal B cell mouse #426

Closed JavierMolinaV closed 1 month ago

JavierMolinaV commented 1 month ago

Hello!

We have developed a monoclonal mouse model using the genetic strategy published in Jacobsen JT et al One-step generation of monoclonal B cell receptor mice capable of isotype switching and somatic hypermutation. J Exp Med. 2018 Oct 1;215(10):2686-2695. doi: 10.1084/jem.20172064. Epub 2018 Sep 4. PMID: 30181412; PMCID: PMC6170169.

The genetic strategy consists in inserting a pre-rearanged complete antibody in the Igh locus with the light chain (VJ) and the heavy chain (VDJ) all transcribed by a unique mRNA. This mRNA is transcribed as a unique transcript and later on the heavy chain and the light chain are cleaved at protein level thanks to a P2A sequence in the mRNA. Another property of the model is that it has a human constant light chain instead of a murine constant light chain in order to identify by flow citometry the cells carrying the antibody with an anti-huIgk. I have modified the IGMT reference in order to include the human constant light chain sequence adding a line where we include the sequence of the huIgk found in IGMT.

We know by FACS that a high percentage of B cells express our construct, but we have performed a SC experiment and I am not able to detect the sequence of the antibody by TRUST4 (I do detect it in Seurat RNA features because we have included the whole sequence in the fasta reference when running cellranger).

I don´t know if I should modify the code of TRUST4 in order to improve the detection of this VDJ. It is true that the mRNA is very large (as it contains the heavy and light chain together) and also the human kappa constant light chain does not contain a stop codon. Maybe some of these features make dificoult the detection of this antibody by TRUST4.

If you could help me I would be very gratefull.

Thank you very much!!

Javi

ncborcherding commented 1 month ago

Hey Javi,

Sounds like an interesting model/experiment. The fact you are picking up RNA counts is a good sign that the sequences are there.

This is probably a better question to ask to the TRUST4 group: https://github.com/liulab-dfci/TRUST4. @mourisl is extremely good and responsive.

Nick