Specific Aim II (Chapter 4) 128

The dnMAML transgene used in Aim II inhibits canonical Notch signaling at the level of

transcriptional complex assembly [17]. However, there are other known functions of the Notch

pathway that dnMAML does not affect. NICD binds to the required transcription factor Runx2 to

inhibit osteoblast differentiation [5]. While dnMAML binds to the NICD-RBPjκ complex, it is

unlikely that this impacts the ability of NICD to have other, non-canonical effects. A recent study

also demonstrated non-canonical and cell non-autonomous functions of Notch signaling during

evidence of potential reverse ligand intracellular domain signaling in the ligand-expressing

signaling cell [35, 36]. dnMAML would also not affect this pathway.

Heterozygous dnMAML mice were used in this study. The use of homozygous mice could

have resulted in stronger phenotypes with clearer interpretations into the role(s) of certain cell

populations during repair. However, in general, use of heterozygous mice can be more clinically

relevant since potential therapeutic applications are likely to achieve partial but not complete

ablation of function. Since dnMAML is not an endogenous gene, we were not as concerned with

compensatory effects from a redundant protein that are more likely to occur in heterozygous

mouse models. There are also other models of inhibition of canonical Notch signaling that could

have been used, such as Notch receptor knockout mice [5, 7], or mice with conditional deletion of

RBPjκ [8, 22, 34, 37], but again, or goal was to modulate Notch signaling, not completely ablate

it.

Both males and females were included in this study, but were appropriately separated

into different groups and not compared to each other since they present with different amounts of

bone during development [38]. However, many previous studies have demonstrated similar

responsivity of male and female mice to manipulations of Notch signaling [6, 7, 21, 22] and male

and female mice follow the same spatiotemporal pattern of healing. Therefore, it is scientifically

justifiable to conclude that the phenotype of females during cartilage formation is equivalent to

what would be observed in males, and vice versa during bone formation.

As with all studies, including later time points closer to or after expected complete healing

would allow for better understanding of the final outcome due to Notch inhibition. However, only

three time points were chosen due to resource and time constraints, and the time points chosen

were based on critical stages of fracture healing (5dpf – mesenchymal callus formation; 10dpf –

cartilage formation and early bone formation; 20dpf – bone formation and remodeling) that also

allow for comparison across many studies.

Importantly, our results demonstrated the importance of Notch signaling to resolve the

inflammation, which occurs immediately after injury. Future studies should additionally investigate

the role of Notch activity during peak inflammation.

Because of the complexity of the spatiotemporally changing population of cells and

tissues during healing, we were unable to assess the role of Notch signaling in distinct cell

populations, including osteoblasts and osteoclasts. To address this limitation, future studies could

utilize tissue-specific models of Cre recombinase expression to activate dnMAML in specific

lineages. Utilizing Prx1, Col3.6 or Col2.3 promoters would inhibit Notch signaling in

undifferentiated mesenchymal progenitors, osteoprogenitors, or committed osteoblasts,

respectively. Similarly, TRAP promoters would inhibit Notch signaling in osteoclast lineage cells,

and expressing Cre in lineage-restricted inflammatory cells would be useful for exploring the

contribution of inflammatory cells.

Alternatively, the use of gamma secretase inhibitors (GSI) would allow temporal control of

Notch signaling to isolate or exclude the role of Notch signaling in specific phases of healing. For

example, GSI injections following the conclusion of the acute inflammatory phase could exclude

any secondary effects of altered inflammation on the rest of healing, providing a model to better

understand the direct role of Notch signaling in cartilage formation, callus vascularization, and

bone formation and remodeling. Similarly, GSI injections starting at the cartilage-to-bone

transition would isolate the role of Notch signaling during bone formation and remodeling.

Calvarial defect experiments included in this thesis are at this stage preliminary work

demonstrating the broader application of Notch relevance. More research is needed to fully

understand the role of Notch signaling during calvarial defect healing. However, results from the

tibial fracture model demonstrate that Notch signaling is needed for successful repair. Future

calvarial defect studies should focus on creating a smaller defect since 1.8 mm diameter injuries

result in non-union [39]. Using an intramembranous repair model that normally regenerates would

allow for better understanding into the requirement of Notch signaling for successful

intramembranous fracture healing.

Finally, in Aim I, we identified Jagged1 as the most highly upregulated ligand, suggesting

evaluating the role of Jagged1 during fracture healing using a similar experimental design to Aim

II. However, we have had difficulty in generating Mx1-Cre+;Jagged1f/f mice because of small litter

size and poor animal health.