Laura Tom

Date of Award

January 2011

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)



First Advisor

Deepak Narayan

Subject Area(s)



Introduction: Hemangiomas are the most common benign tumor of infancy. They are unique in that their behavioral course is well-described with a microvascular proliferative phase, quiescent phase and regressive fatty involuting phase. First, it is posited in this work that the cell of origin will have angiogenic and adipogenic properties, like the pericyte. Second, the predilection of the lesion to areas of embryologic fusions in the head and neck indicate that the neural crest may be involved. Previous research focused on the angiogenic characteristics of proliferating hemangiomas, this work is novel in that it investigates markers involved in adipogenesis. This study explores the possibility that hemangiomas are pericyte-like stem cell tumors, derived from the neural crest, capable of adipocyte differentiation.

Methods: Markers for pericytes, stem cells and neural crest cells as well as adipocyte differentiation were analyzed using using immunohistochemistry, immunofluorescence and quantitative RT-PCR. These markers included: NG2, SMA, calponin, CD90(Thy1), GLUT1, dlk (pericytes); Pauf5 (Oct4), Nanog, Sox2, Sox9, Sox10, c-myc (stem cells); Nestin, Sox10, NGFR (p75), NG2 (neural crest cells) and C/EBPá, C/EBPâ, C/EBPä, PPARã, dlk, RXR, SREBF, Sox9, and Krûppel-like factors (KLF2, KLF4, KLF5, KLF6, KLF15) (adipocyte differentiation markers). Immunohistochemistry and immunofluorescence was done on formalin-fixed, paraffin-embedded archival tissue. qRT-PCR was performed on 15 hemangioma specimens (5 proliferating, 3 early quiescent, 4 late quiescent, and 5 involuting) and compared to Human Dermal Microvascular Endothlial Cells (HDMEC) utilizing the delta-delta Ct method of analysis. Hemangioma phase was determined by clinical notes. To compare hemangioma phases, a one-way ANOVA was done with a Tukey HSD post hoc analysis.

Results: Immunohistochemistry identified these markers in the periluminal structures within the hemangiomas: GLUT1, dlk, SMA, and Nestin. Immunofluorescence identified these markers in a confirmatory pattern: GLUT1, dlk, SMA, Nestin, Sox2, Calponin and NG2. The qRTPCR analysis revealed that the relative expression of all (pericyte, stem cell, neural crest and adipocyte differentiation) analyzed markers was greater than that of the HDMEC control. Significant differences were found between the hemangioma phases for Sox9, C/EBPâ and KLF2.

Conclusion: The work presented in this thesis is novel and informative. It demonstrated that hemangiomas indeed express pericyte, stem cell and neural crest cell markers. This is the first study to identify factors involved in adipocyte differentiation including dlk, PPARã and KLFs. The complex adipogenesis transcription cascade complicates the ability to draw conclusion in regard to adipocyte differentiation. However, the data presented provides support for the theory that hemangiomas are pericyte-like stem cell tumors derived from the neural crest capable of adipocyte differentiation.


This is an Open Access Thesis.

Open Access

This Article is Open Access