About

Michelle L. Delco, DVM, PhD, ACVS, is an Assistant Research Professor in the Section of Large Animal Surgery within the Department of Clinical Sciences at Cornell University College of Veterinary Medicine. She holds a DVM from Cornell University and completed a residency in Equine Surgery at the University of California, Davis, where she became a diplomate of the American College of Veterinary Surgeons. Dr. Delco earned her PhD in Comparative Biomedical Sciences from Cornell University in 2016. Her research focuses on understanding how joint injury leads to arthritis in horses and humans, specifically investigating mitochondria as a link between cartilage trauma and osteoarthritis. Her goal is to develop strategies to improve healing and prevent irreversible joint disease and chronic pain in equine and human athletes suffering orthopedic injuries. Professionally, she serves as a Large Animal Staff Surgeon at the Cornell University Hospital for Animals and as an Equine Surgeon at Cornell Ruffian Equine Specialists. Her career includes experience in minimally invasive surgery and the diagnosis and treatment of complex sports injuries in equine athletes, along with recognition through various awards and memberships in professional societies.

Research topics

• Medicine
• Cell biology
• Chemistry
• Pathology
• Biology

Selected publications

• Articular chondrocytes from the knee and ankle have different sensitivities to shear strain

  Journal of Biomechanics · 2026-03-04

  article
  PublisherDOI

• Post‐operative performance in Standardbreds after distal splint bone ostectomy with concurrent suspensory ligament branch desmitis

  Equine Veterinary Education · 2025-06-24 · 1 citations

  article

  Summary Background Distal splint bone fractures are often complicated by suspensory ligament branch (SLB) desmitis. These combined injuries are known to impair postoperative performance, but contemporary data on prognosis is limited. Objective To evaluate the postoperative performance of Standardbred horses following distal splint bone fracture ostectomy with concurrent SLB desmitis. Study design Retrospective cohort study. Methods Clinical, demographic and racing data were retrospectively analysed from Standardbreds racing in Malta. Horses with a splint bone fracture and concurrent SLB desmitis underwent standing ostectomy and were followed for 24 months. Kaplan–Meier survival analysis assessed the time to return to racing. Pre and postoperative performance was compared within the surgery group and against age‐, sex‐ and breed‐matched controls using mixed effects models. The effects of SLB lesion grade and adjunctive platelet‐rich plasma (PRP) therapy were also analysed. Results Eleven horses [median age 10 years (9‐11)] with fractures of the distal 1/3 of the splint bone and concurrent SLB desmitis were included. The mean time to return to racing was 8.5 months (95% CI: 7.5–9.5). Sixty‐four % (7/11) of horses raced at least three times postoperatively. The surgery group raced significantly more than controls over 24 months (mean difference: 2.2 ± 0.4; p < 0.0001). No difference was found between pre and postoperative performance within the surgery group ( p > 0.05). Horses with grade 1 SLB lesions raced more than those with grade 2 or 3 lesions ( p < 0.0001) and controls ( p = 0.0008). Age and PRP treatment had no effect. Main limitations Small sample size ( n = 11) and lack of conservatively treated group for comparison. Conclusion Older Standardbred horses with distal splint bone fractures and mild SLB desmitis can successfully return to racing. Prognosis is favourable for less severe SLB lesions, emphasising the importance of accurate lesion grading in surgical decision‐making.

  PublisherDOI

• Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli

  Zenodo (CERN European Organization for Nuclear Research) · 2024-01-25

  datasetOpen access

  This is the dataset for our paper titled " Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli". Corresponding author information: Email: ht452@cornell.edu (Han Kheng Teoh); ic64@cornell.edu (Itai Cohen) This dataset is shared under a Creative Commons Attribution 4.0 International license (CC BY 4.0); the data will be openly available to share and adapt, but appropriate credit to the original data creators is required upon reuse. When using this dataset, please cite: The dataset: Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Data from: Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli [dataset]. Zenodo. https://doi.org/10.5281/zenodo.10565588 AND the paper: Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli. PLOS One https://doi.org/10.1371/journal.pone.0297947 The work was supported by the NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases, Contract: K08AR068470, R03AR075929, and The Harry M. Zweig Fund for Equine Research. This work was also supported by the NIH National Institute of Neurological Disorders and Stroke. Contract: R01NS116595. Additionally, this work was supported by the National Science Foundation grants DMR-1807602, CMMI 1927197, and BMMB-1536463. Lastly, this work made use of the Cornell Center for Materials Research Shared Facilities, which are supported through the NSF MRSEC program (DMR-1719875). DATA & FILE OVERVIEW ------------------------------------------------- The dataset contains two folders : Data and Code. In the Data folder, the experimental data is organized into three subfolders, specifying the date when the experiment was performed. Each subfolder contains the following files: all_locs.mat - contains the cells (x,y) position. The data is organized as a Nx2 array, where N is the number of cells in the sample. blue_all.mat - contains the post-impact NMP (cell death) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the NMP (cell death) intensity was measured. green_all.mat - contains the post impact Ca^{2+} intensity for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured. red_all.mat - contains the post impact TMRM (mitochrondrial polarity) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the TMRM (mitochondrial polarity) intensity was measured. impact_intensity.mat - contains the Ca^{2+} intensity during impact for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured. impact_locs.mat - contains the cells' (x,y) position within the impact site. The data is organized as a Nx2 array, where N is the number of cells in the sample. D_skl_dd_mm_yy.p - contains the symmetrized KL divergence between cells' latent representation. The data is organized as a N by N array, where N is the number of cells. In addition, the Data folder also contains : model_weights.p file - contains the weights and biases for the trained VAE network used in this study. The Code folder contains: decoders.py - contains a class function for the VAE decoder. encoders.py - contains a class function for the VAE encoder. loaders.py - contains a function that partitions the cell data into a training set and a test set. wrapper.py - contains a class function that trains a VAE. Cartilage VAE - Part I.ipynb - contains the code necessary to generate Figures 1 to 5 in the manuscript. Cartilage VAE - Part II.ipynb - contains the code necessary to generate Figures 5 to 9 in the manuscript.

  PublisherDOI

• Degradation of lubricating molecules in synovial fluid alters chondrocyte sensitivity to shear strain

  Journal of Orthopaedic Research® · 2024-08-25 · 8 citations

  articleOpen access

  Articular joints facilitate motion and transfer loads to underlying bone through a combination of cartilage tissue and synovial fluid, which together generate a low-friction contact surface. Traumatic injury delivered to cartilage and the surrounding joint capsule causes secretion of proinflammatory cytokines by chondrocytes and the synovium, triggering cartilage matrix breakdown and impairing the ability of synovial fluid to lubricate the joint. Once these inflammatory processes become chronic, posttraumatic osteoarthritis (PTOA) development begins. However, the exact mechanism by which negative alterations to synovial fluid leads to PTOA pathogenesis is not fully understood. We hypothesize that removing the lubricating macromolecules from synovial fluid alters the relationship between mechanical loads and subsequent chondrocyte behavior in injured cartilage. To test this hypothesis, we utilized an ex vivo model of PTOA that involves subjecting cartilage explants to a single rapid impact followed by continuous articulation within a lubricating bath of either healthy synovial fluid, phosphate-buffered saline (PBS), synovial fluid treated with hyaluronidase, or synovial fluid treated with trypsin. These treatments degrade the main macromolecules attributed with providing synovial fluid with its lubricating properties; hyaluronic acid and lubricin. Explants were then bisected and fluorescently stained to assess global and depth-dependent cell death, caspase activity, and mitochondrial depolarization. Explants were tested via confocal elastography to determine the local shear strain profile generated in each lubricant. These results show that degrading hyaluronic acid or lubricin in synovial fluid significantly increases middle zone chondrocyte damage and shear strain loading magnitudes, while also altering chondrocyte sensitivity to loading.

  PublisherOA PDFDOI

• Preliminary in vivo investigation of the mesenchymal stromal cell secretome as a novel treatment for methicillin‐resistant <i>Staphylococcus aureus</i> in equine skin wounds

  Veterinary Surgery · 2024-10-05 · 3 citations

  articleOpen access

  OBJECTIVE: We aimed to study the antimicrobial and pro-healing potential of equine mesenchymal stromal cell secreted products (i.e. secretome), collected as conditioned media (mesenchymal stromal cell-conditioned media, MSC CM), in a novel in vivo model of methicillin-resistant Staphylococcus aureus (MRSA)-inoculated equine thorax wounds. STUDY DESIGN: Prospective in vivo study. ANIMALS: Two Thoroughbred geldings. METHODS: Six full-thickness cutaneous wounds were created bilaterally on the dorsal thorax of two horses (n = 12 wounds/horse). Wounds on the left thoraces were inoculated with MRSA on day 0. All wounds were then treated with either mupirocin ointment, MSC CM, or vehicle control (n = 4 wounds per group) once daily for 3 days. Photographs were taken to quantify wound scores and sizes, as well as samples to determine bacterial colony forming units (CFUs), at days 0, 1, 2, 3, 7, 14, 21, and 28. The wound edge was biopsied on days 0, 7, and 28, and scored histologically. RESULTS: Inoculated wounds had more bacterial CFUs at day 1 (p < .0001) and were larger in size at day 28 (p = .0009) than noninoculated wounds. Mupirocin-treated wounds were smaller than MSC CM and vehicle control-treated wounds at day 28 (p = .003). Mesenchymal stromal cell-conditioned media did not affect CFU numbers in inoculated and noninoculated wounds. Moreover, MSC CM did not affect the parameters of wound size or gross or microscopic wound scores over time. CONCLUSION: Mesenchymal stromal cell-conditioned media did not exhibit antimicrobial or pro-healing properties in the current study; however, the in vivo model of inoculated equine thorax wounds requires further optimization. CLINICAL SIGNIFICANCE: This pilot study contributes to a growing understanding of the equine MSC secretome as an antimicrobial and pro-healing therapeutic for equine wounds.

  PublisherOA PDFDOI

• Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli

  Zenodo (CERN European Organization for Nuclear Research) · 2024-01-25

  datasetOpen access

  This is the dataset for our paper titled " Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli". Corresponding author information: Email: ht452@cornell.edu (Han Kheng Teoh); ic64@cornell.edu (Itai Cohen) This dataset is shared under a Creative Commons Attribution 4.0 International license (CC BY 4.0); the data will be openly available to share and adapt, but appropriate credit to the original data creators is required upon reuse. When using this dataset, please cite: The dataset: Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Data from: Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli [dataset]. Zenodo. https://doi.org/10.5281/zenodo.10565588 AND the paper: Jingyang Zheng, Han Kheng Teoh, Michelle L. Delco, Lawrence J. Bonassar , and Itai Cohen. (2024) Application of a Variational Autoencoder for Clustering and Analyzing in situ Articular Cartilage Cellular Response to Mechanical Stimuli. PLOS One https://doi.org/10.1371/journal.pone.0297947 The work was supported by the NIH National Institute of Arthritis and Musculoskeletal and Skin Diseases, Contract: K08AR068470, R03AR075929, and The Harry M. Zweig Fund for Equine Research. This work was also supported by the NIH National Institute of Neurological Disorders and Stroke. Contract: R01NS116595. Additionally, this work was supported by the National Science Foundation grants DMR-1807602, CMMI 1927197, and BMMB-1536463. Lastly, this work made use of the Cornell Center for Materials Research Shared Facilities, which are supported through the NSF MRSEC program (DMR-1719875). DATA & FILE OVERVIEW ------------------------------------------------- The dataset contains two folders : Data and Code. In the Data folder, the experimental data is organized into three subfolders, specifying the date when the experiment was performed. Each subfolder contains the following files: all_locs.mat - contains the cells (x,y) position. The data is organized as a Nx2 array, where N is the number of cells in the sample. blue_all.mat - contains the post-impact NMP (cell death) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the NMP (cell death) intensity was measured. green_all.mat - contains the post impact Ca^{2+} intensity for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured. red_all.mat - contains the post impact TMRM (mitochrondrial polarity) intensity for each cell. The data is organized as a TxN array, where T is the number of time points the TMRM (mitochondrial polarity) intensity was measured. impact_intensity.mat - contains the Ca^{2+} intensity during impact for each cell. The data is organized as a TxN array, where T is the number of time points the Ca^{2+} intensity was measured. impact_locs.mat - contains the cells' (x,y) position within the impact site. The data is organized as a Nx2 array, where N is the number of cells in the sample. D_skl_dd_mm_yy.p - contains the symmetrized KL divergence between cells' latent representation. The data is organized as a N by N array, where N is the number of cells. In addition, the Data folder also contains : model_weights.p file - contains the weights and biases for the trained VAE network used in this study. The Code folder contains: decoders.py - contains a class function for the VAE decoder. encoders.py - contains a class function for the VAE encoder. loaders.py - contains a function that partitions the cell data into a training set and a test set. wrapper.py - contains a class function that trains a VAE. Cartilage VAE - Part I.ipynb - contains the code necessary to generate Figures 1 to 5 in the manuscript. Cartilage VAE - Part II.ipynb - contains the code necessary to generate Figures 5 to 9 in the manuscript.

  PublisherDOI

• Connexin 43 regulates intercellular mitochondrial transfer from human mesenchymal stromal cells to chondrocytes

  Stem Cell Research & Therapy · 2024-10-10 · 34 citations

  articleOpen accessSenior author

  BACKGROUND: The phenomenon of intercellular mitochondrial transfer from mesenchymal stromal cells (MSCs) has shown promise for improving tissue healing after injury and has potential for treating degenerative diseases like osteoarthritis (OA). Recently MSC to chondrocyte mitochondrial transfer has been documented, but the mechanism of transfer is unknown. Full-length connexin 43 (Cx43, encoded by GJA1) and the truncated, internally translated isoform GJA1-20k have been implicated in mitochondrial transfer between highly oxidative cells, but have not been explored in orthopaedic tissues. Here, our goal was to investigate the role of Cx43 in MSC to chondrocyte mitochondrial transfer. In this study, we tested the hypotheses that (a) mitochondrial transfer from MSCs to chondrocytes is increased when chondrocytes are under oxidative stress and (b) MSC Cx43 expression mediates mitochondrial transfer to chondrocytes. METHODS: Oxidative stress was induced in immortalized human chondrocytes using tert-Butyl hydroperoxide (t-BHP) and cells were evaluated for mitochondrial membrane depolarization and reactive oxygen species (ROS) production. Human bone-marrow derived MSCs were transduced for mitochondrial fluorescence using lentiviral vectors. MSC Cx43 expression was knocked down using siRNA or overexpressed (GJA1 + and GJA1-20k+) using lentiviral transduction. Chondrocytes and MSCs were co-cultured for 24 h in direct contact or separated using transwells. Mitochondrial transfer was quantified using flow cytometry. Co-cultures were fixed and stained for actin and Cx43 to visualize cell-cell interactions during transfer. RESULTS: Mitochondrial transfer was significantly higher in t-BHP-stressed chondrocytes. Contact co-cultures had significantly higher mitochondrial transfer compared to transwell co-cultures. Confocal images showed direct cell contacts between MSCs and chondrocytes where Cx43 staining was enriched at the terminal ends of actin cellular extensions containing mitochondria in MSCs. MSC Cx43 expression was associated with the magnitude of mitochondrial transfer to chondrocytes; knocking down Cx43 significantly decreased transfer while Cx43 overexpression significantly increased transfer. Interestingly, GJA1-20k expression was highly correlated with incidence of mitochondrial transfer from MSCs to chondrocytes. CONCLUSIONS: Overexpression of GJA1-20k in MSCs increases mitochondrial transfer to chondrocytes, highlighting GJA1-20k as a potential target for promoting mitochondrial transfer from MSCs as a regenerative therapy for cartilage tissue repair in OA.

  PublisherOA PDFDOI

• Connexin 43 Regulates Intercellular Mitochondrial Transfer from Human Mesenchymal Stromal Cells to Chondrocytes

  bioRxiv (Cold Spring Harbor Laboratory) · 2024-03-20 · 6 citations

  preprintOpen accessSenior authorCorresponding

  Background: ) and the truncated internally translated isoform GJA1-20k have been implicated in mitochondrial transfer between highly oxidative cells, but have not been explored in orthopaedic tissues. Here, our goal was to investigate the role of Cx43 in MSC to chondrocyte mitochondrial transfer. In this study, we tested the hypotheses that (a) mitochondrial transfer from MSCs to chondrocytes is increased when chondrocytes are under oxidative stress and (b) MSC Cx43 expression mediates mitochondrial transfer to chondrocytes. Methods: Oxidative stress was induced in immortalized human chondrocytes using tert-Butyl hydroperoxide (t-BHP) and cells were evaluated for mitochondrial membrane depolarization and reactive oxygen species (ROS) production. Human bone-marrow derived MSCs were transduced for mitochondrial fluorescence using lentiviral vectors. MSC Cx43 expression was knocked down using siRNA or overexpressed (GJA1+ and GJA1-20k+) using lentiviral transduction. Chondrocytes and MSCs were co-cultured for 24 hrs in direct contact or separated using transwells. Mitochondrial transfer was quantified using flow cytometry. Co-cultures were fixed and stained for actin and Cx43 to visualize cell-cell interactions during transfer. Results: Mitochondrial transfer was significantly higher in t-BHP-stressed chondrocytes. Contact co-cultures had significantly higher mitochondrial transfer compared to transwell co-cultures. Confocal images showed direct cell contacts between MSCs and chondrocytes where Cx43 staining was enriched at the terminal ends of actin cellular extensions containing mitochondria in MSCs. MSC Cx43 expression was associated with the magnitude of mitochondrial transfer to chondrocytes; knocking down Cx43 significantly decreased transfer while Cx43 overexpression significantly increased transfer. Interestingly, GJA1-20k expression was highly correlated with incidence of mitochondrial transfer from MSCs to chondrocytes. Conclusions: Overexpression of GJA1-20k in MSCs increases mitochondrial transfer to chondrocytes, highlighting GJA1-20k as a potential target for promoting mitochondrial transfer from MSCs as a regenerative therapy for cartilage tissue repair in OA.

  PublisherOA PDFDOI

• Return to racing in Standardbreds after distal splint bone ostectomy with concurrent suspensory ligament branch desmitis

  2024-10-30

  preprintOpen access

  Background : Distal splint bone fractures, common in racing Standardbreds and Thoroughbreds, are often complicated by secondary suspensory ligament branch (SLB) desmitis. These combined injuries are known to impair post-operative performance, but contemporary data on prognosis is limited. Objective : To evaluate the post-operative performance of horses following distal splint bone fracture ostectomy with concurrent SLB desmitis. Study Design: Retrospective cohort study Methods: Clinical, demographic, and racing data were retrospectively analyzed. Horses with a splint bone fracture and concurrent SLB desmitis underwent standing ostectomy and were followed for 24 months. Kaplan-Meier survival analysis assessed the time to return to racing. Pre- and post-operative performance was compared within the surgery group and against age-, sex-, and breed-matched controls using mixed-effects models. The effects of SLB lesion grade and adjunctive platelet-rich plasma (PRP) therapy were also analyzed. Results: Eleven horses (mean age 9 ± 1.6 years) with fractures of the distal 1/3 of the splint bone and concurrent SLB desmitis were included. The mean time to return to racing was 8.5 months (95% CI: 7.5–9.5). 64% (7/11) of horses raced at least three times post-operatively. Surgery group horses raced significantly more than controls over 24 months (mean difference: 2.2 ± 0.4; p &lt; 0.0001). No difference was found between pre- and post-operative performance within the surgery group (p &gt; 0.05). Horses with grade 1 SLB lesions raced more than those with grade 2 or 3 lesions (p &lt; 0.0001) and controls (p = 0.0008). Age and PRP treatment had no significant effect. Main limitations : Relatively small surgery population size (n = 11) and lack of conservatively treated group for comparison Conclusion: Distal splint bone ostectomy with concurrent SLB desmitis had a favorable prognosis in cases of less severe SLB lesions.

  PublisherOA PDFDOI

• Application of a variational autoencoder for clustering and analyzing in situ articular cartilage cellular response to mechanical stimuli

  PLoS ONE · 2024-05-20 · 2 citations

  articleOpen accessCorresponding

  In various biological systems, analyzing how cell behaviors are coordinated over time would enable a deeper understanding of tissue-scale response to physiologic or superphysiologic stimuli. Such data is necessary for establishing both normal tissue function and the sequence of events after injury that lead to chronic disease. However, collecting and analyzing these large datasets presents a challenge-such systems are time-consuming to process, and the overwhelming scale of data makes it difficult to parse overall behaviors. This problem calls for an analysis technique that can quickly provide an overview of the groups present in the entire system and also produce meaningful categorization of cell behaviors. Here, we demonstrate the application of an unsupervised method-the Variational Autoencoder (VAE)-to learn the features of cells in cartilage tissue after impact-induced injury and identify meaningful clusters of chondrocyte behavior. This technique quickly generated new insights into the spatial distribution of specific cell behavior phenotypes and connected specific peracute calcium signaling timeseries with long term cellular outcomes, demonstrating the value of the VAE technique.

  PublisherOA PDFDOI

Recent grants

• Mitochondrial dysfunction as a link between cartilage injury and osteoarthritis

  NIH · $724k · 2021–2022

• The role of mitochondrial Damage Associated Molecular Patterns (mDAMPs) in posttraumatic osteoarthritis.

  NIH · $157k · 2019–2022

Frequent coauthors

• Lisa A. Fortier

  New York State College of Veterinary Medicine

  56 shared

• Lawrence J. Bonassar

  Cornell University

  26 shared

• Itai Cohen

  17 shared

• Edward D. Bonnevie

  University of Pennsylvania

  12 shared

• Megan J. Fahey

  Cornell University

  10 shared

• Brenna Pugliese

  Cornell University

  9 shared

• Matthew A. Thomas

  Cornell University

  8 shared

• L.A. Seewald

  Cornell University

  8 shared

Labs

• Michelle L. Delco LaboratoryPI

Awards & honors

• Young Investigator Award, Arnold and Madalene Penner 5 th An…
• Coleman Graduate Leadership Program, Cornell University (201…
• Dante and Sharon Ferrini Award for Veterinary Thoroughbred H…
• Diplomate, American College of Veterinary Surgeons (Large An…
• Center for Equine Health Research Award, University of Calif…