Theses Doctoral

The impact of Zfp106 on mouse muscle homeostasis

Quejada, Jose Rafael Navarro

Murine Zfp106 is an 1,888 amino acid protein with two N-terminal and two C-terminal zinc finger domains with a beta-propeller upstream of the C-terminal zinc fingers. The transcription of the protein was found to be controlled through two promoter regions, leading to two isoform families, P1 and P2 Zfp106. The splice variants from each promoter are thought to have distinct starting exons and n-terminal regions. However, the isoforms are not well studied. Since its identification, Zfp106 has been implicated in RNA metabolism, transcription control, immune response, and muscle and testis development. It has been found to be capable of binding C9ORF72 repeats as well as being associated with TDP43 and FUS. However, its function is unknown.

The aim of this study is to understand the role of Zfp106 in vivo through the use and development of various mouse models targeting exons specific to either the P1 or P2 family of isoforms. To begin with, we studied the Zfp106LacZ mouse model whose homozygous mice showed severe muscle atrophy beginning at 4 weeks leading to a premature death by 16 weeks. Research has supported the theory that the muscle atrophy is due to a motor neuron dysfunction potentially stemming from perturbed mitochondrial and spliceosome function. We, along with other researchers, found that this mouse model is not a complete disruption of Zfp106 through qPCR and RNAseq. We then found that this mouse model is an effective depletion of Zfp106 exon 2 and 3 which are exclusive to the P1 Zfp106 isoform family. Additionally, the Zfp106LacZ mouse model has an increased amount of the 1b exon associated with P2-Zfp106 in the skeletal muscle.

Next, we established a CRISPR mouse line (ΔZfp106) targeting an exon common to the full-length splice variants of both the P1 and P2 family of isoforms, exon 5. This was in an attempt to dissect whether or not the muscle atrophy in the Zfp106LacZ mice was due to the interruption of exons 2 and 3 or from the increase in the P2 Zfp106 isoforms. Motor neurons derived from homozygous ΔZfp106 mouse embryonic stem cells, were found to be susceptible to CPA-induced endoplasmic reticulum stress and rotenone induced mitochondrial stress. Interestingly, the in vivo penetration rate of the muscle atrophy phenotype of homozygous ΔZfp106 mice is 60% for male and 12.5% for female mice. This is in stark contrast to the 100% penetration rate of the Zfp106LacZ mice. The reason behind this is currently unclear but may be due to either the incomplete backcrossing of the mouse model, a difference in the splice variants affected by the Zfp106 targeting, or because the muscle atrophy in the Zfp106LacZ mouse model is caused in part by the increase in the expression the P2 Zfp106 family of isoforms. These two mouse models show that affecting the expression of the full-length isoforms of P1-Zfp106 can lead to muscle atrophy.

In an attempt to see if the Zfp106LacZ muscle atrophy was due to a lack of Zfp106 in the skeletal muscle, spinal cord, or necessitated its depletion in both, we derived a mouse line from the Zfp106LacZ that conditionally depletes exon 3 which impairs the expression of full length P1 Zfp106. This was used to target exon 3 removal to the skeletal muscle (Myf5), cholinergic neurons (ChAT), simultaneously (Myf5/ChAT), or a whole body depletion (Ella2). Surprisingly, the whole body depletion of Zfp106 exon 3 did not lead to muscle atrophy even though its removal leads to a frame shift and premature stop codon. The lack of a muscle atrophy phenotype may be because of the expression of a splice variant without exon 3, thereby rescuing the neuromuscular pathology.

Lastly, to better understand the role of the P2 Zfp106 in vivo, we created a mouse line with a CRISPR mediated knockout of exon 1b (ΔP2). Exon 1b is the start exon of the P2 Zfp106 isoform family and the introduction of a destructive INDEL should independently affect the P2 isoform family. Interestingly, this mouse model showed no observed neuromuscular dysfunction or metabolic disorder, responding to a glucose bolus similarly with controls. The lack of a phenotype may be due to compensation by other Zfp106 isoforms or that the P2 isoform family is important in other biological roles.


  • thumnail for Quejada_columbia_0054D_16064.pdf Quejada_columbia_0054D_16064.pdf application/pdf 5.93 MB Download File

More About This Work

Academic Units
Pharmacology and Molecular Signaling
Thesis Advisors
Yazawa, Mazayuki
Ph.D., Columbia University
Published Here
July 30, 2020