June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Machine learning used to identify diverse cellular roles in the corneal wound healing response
Author Affiliations & Notes
  • Kristen Segars
    Pharmacology, Boston University School of Medicine, Boston, Massachusetts, United States
  • Nicholas Azzari
    Biochemistry, Boston University School of Medicine, Boston, Massachusetts, United States
  • Celeste B Rich
    Biochemistry, Boston University School of Medicine, Boston, Massachusetts, United States
  • Vickery E Trinkaus-Randall
    Biochemistry, Boston University School of Medicine, Boston, Massachusetts, United States
  • Footnotes
    Commercial Relationships   Kristen Segars None; Nicholas Azzari None; Celeste Rich None; Vickery Trinkaus-Randall None
  • Footnotes
    Support  1F30EY033647-01, R21EY029097-01, Massachusetts Lions Eye Research Foundation, New England Corneal Transplant Fund
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 3110. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Kristen Segars, Nicholas Azzari, Celeste B Rich, Vickery E Trinkaus-Randall; Machine learning used to identify diverse cellular roles in the corneal wound healing response. Invest. Ophthalmol. Vis. Sci. 2023;64(8):3110.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : When the corneal epithelium is injured, ATP released into the local environment initiates calcium signaling events in cells near the wound. Propagation of these events from cell to cell is necessary for cellular motility and wound healing. We have developed computational analysis programs to characterize these signaling events and have identified high-signaling cells in the wound response that recruit their neighbors into “clusters” of signaling events. We hypothesize that individual cells play different roles in the wound healing response based on their signaling behavior, with high-signaling cells driving the wound healing response.

Methods : Live cell imaging was performed on both ex vivo globes from male C57Bl6 mice and cultured Human Corneal Limbal Epithelial (HCLE) cells using the Zeiss LSM 880 confocal microscope. Laser ablation was performed using the Bleaching feature. Globes were pre-incubated with CellMask DeepRed and Fluo-4AM, and stabilized using a 3D printed holder. Images were collected after injury on the corneal epithelium and corneal-limbal regions. SiRNA knockdown or pharmacological inhibition was performed on HCLE cells prior to staining with SiR actin and Fluo-4AM, wounding and imaging. Machine learning analysis was performed using Zen, Matlab and ImageJ.

Results : Recruitment of cells into signaling clusters is observed near the wound in both cell culture and organ culture models. After a wound to the central cornea, recruitment is also observed in the corneal-limbal interface. Hierarchical clustering analysis revealed a subpopulation of high-signaling cells that initiate recruitment in their neighbors. KnKnockdown of P2X7 attenuates the conductor cell phenotype by reducing the high-signaling cell population. Knockdown of Pannexin-1 and Connexin-43 also attenuate a distinct conductor cell phenotype, but do so by promoting an intermediate-signaling phenotype. Ablating a conductor cell causes neighboring cells to take on a higher-signaling phenotype than previously observed.

Conclusions : A diverse range of cell signaling behavior is present in a successful wound healing response, with high-signaling “conductor cells” recruiting their lower-signaling neighbors into a signaling cluster. Future studies will focus on the hypothesized loss of this functionally distinct cellular phenotype with age and disease, conditions where wound healing in the cornea becomes inefficient.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×