Logotyp: till Uppsala universitets webbplats

The generation and coordination of locomotion comprise of inhibitory and excitatory interneurons that govern motor neuron output to muscles. This neuronal network is known as the central pattern generator (CPG). Spinal inhibitory interneurons that originate from the dI6 population play a crucial role in the coordination of locomotor output. This thesis provides new insights into the role of two subsets of interneurons within the dI6 population, marked by the expression of Doublesex and mab-3 related transcription factor 3 (dmrt3a) or Wilms tumor 1 (wt1a). 

We showed that dmrt3a neurons play an important role during zebrafish locomotion by using a transgenic line lacking dmrt3a expression. The absence of Dmrt3a generated acceleration and coordination problems in larvae and reduced the maximum speed in juveniles. Analysis of the transcriptome of the dmrt3a subpopulation in both mice and zebrafish established their characteristics as well as revealed unique markers for novel dI6 subpopulations. The knock-down of Wt1a protein resulted in impaired left-right alternation and a reduction of fast swims. On a cellular level, we found alterations in dmrt3a-, wt1a- and evx2- interneuron composition, indicating a link between the two dI6 populations and changes in their fate assignments affecting cell type composition within the locomotor network. Moreover, we linked the absence of specific subtypes of dI6 interneurons with the observed locomotor phenotype in knock-down animals.

While searching for unique markers genes within dmrt3a-expressing interneurons, we observed calb2b (encoding for calretinin) expression in one specific subpopulation. Following the generation of a calb2b transgenic line, we found it expressed in interneurons and motor neurons. Knock-down of calretinin generated a development disorder of dmrt3a interneurons and motor neurons, leading to a disruption of the left-right alternation and escape responses performance. 

Together, the work presented here provides new cellular, molecular, and behavioral information related to the dI6 population, helping us to better understand their role within the CPG of zebrafish specifically and vertebrates as a whole.

Neural networks are the functional building blocks of the central nervous system. To better understand how these networks develop and operate, we turned to the zebrafish locomotor network, with a focus on subtypes of interneurons expressing dmrt3a and wt1a. These neurons first gained interest when a mutation in the Dmrt3 gene was found to be responsible for Icelandic horses’ ability to perform additional gaits, indicating a flexibility within their locomotor central pattern generator.

In zebrafish, the Dmrt3 population is known to be commissural, inhibitory and involved in escape behaviors and left-right alternation during locomotion. We characterized the locomotor behavior at embryonic, larval and juvenile stages in dmrt3a mutants. A strong phenotype was observed in larval escape behavior, showing reduced top speed while the animals spent more time accelerating. While the phenotype subdued as the animals developed, juveniles still maintained a lower maximum locomotor speed.

To get a more detailed understanding of the observed phenotypes, an experimental setup was established combining dual ventral root recordings with calcium imaging and various sensory stimuli to induce diverse locomotor outputs in fictively behaving larva. Implementing this method, we investigated the function of Dmrt3 and Wt1 expressing interneurons in escape behaviors and found that knock-down of Dmrt3a disturbed the fast phase of tail evoked escapes, while knock-down of Wt1a lead to aberrant looming evoked escapes, indicating sub-functionalization. 

Finally, calcium imaging was employed to reveal the activity of Dmrt3 neurons at a population level. The fraction of active cells steadily increased during development and small clusters of correlated Dmrt3 interneuron ensembles were observed within a segment. This work provides insights into how parallel motor networks are orchestrated to generate a flexible behavioral output, revealing fundamental principles extending to the workings of our own brain.