Integrated Molecular Profiling of Clinical and Preclinical Cardiac Xenotransplants

F. Becker (München)¹, N. Hadi Zadeh (München)², J. Ji (München)², B. Reichart (München)³, M. Bender (München)⁴, H. Kartmann (München)², A. Vincent (Martinsried)⁵, J. Stöckl (München)⁶, J. Adelssen (München)², A. Ali (München)⁷, E. Kemter (München)⁷, B. P. Griffith (Baltimore)⁸, A. K. Singh (Baltimore)⁸, B. Kessler (Oberschleißheim)⁹, M. Kurome (Oberschleißheim)⁹, D. DeLaughter (Boston)¹⁰, G. Puga Yung (Geneve)¹¹, J. Denner (Berlin)¹², A. Godehardt (Langen)¹³, C. Walz (München)¹⁴, S. Massberg (München)¹, T. Fröhlich (München)⁶, J. Galindo (Baltimore)¹⁵, A. Tully (Baltimore)¹⁶, A. Grazioli (Baltimore)¹⁷, P. Brenner (München)¹⁸, J. G. Seidman (Boston)¹⁰, C. Seidman (Boston)¹⁰, T. Nordmann (Martinsried)¹⁹, D. Ayares (Blacksburg)²⁰, J. Abicht (München)⁴, M. Mann (Martinsried)²¹, E. Wolf (Oberschleissheim)²², M. Längin (München)⁴, M. M. Mohiuddin (Baltimore)²³, E. L. Lindberg (München)², D. Reichart (München)^1
1LMU Klinikum der Universität München Medizinische Klinik und Poliklinik I München, Deutschland; ²LMU Klinikum Medizinische Klinik und Poliklinik I München, Deutschland; ³LMU Klinikum der Universität München Herzchirurgische Klinik und Poliklinik München, Deutschland; ⁴LMU Klinikum Klinik für Anaesthesiologie München, Deutschland; ⁵Max-Planck-Institut Institute of Biochemistry Martinsried, Deutschland; ⁶Ludwig-Maximilians-Universität Gene Center - Laboratory for Functional Genome Analysis München, Deutschland; ⁷Ludwig-Maximilians-Universität Gene Center and Department of Biochemistry München, Deutschland; ⁸University of Maryland Cardiothoracic Surgery Baltimore, USA; ⁹Ludwig-Maximilians-Universität Center for Innovative Medical Models Oberschleißheim, Deutschland; ¹⁰Harvard Medical School Department of Genetics Boston, USA; ¹¹Universite de Geneve Department of Medicine Geneve, Schweiz; ¹²Freie Universität Berlin Fachbereich Veterinärmedizin - Institut für Virologie Berlin, Deutschland; ¹³Paul-Ehrlich-Institut Bundesinstitut für Impfstoffe und biomedizinische Arzneimittel Langen, Deutschland; ¹⁴LMU Klinikum Pathologisches Institut München, Deutschland; ¹⁵University of Maryland Baltimore, Deutschland; ¹⁶University of Maryland General Surgery Baltimore, USA; ¹⁷University of Maryland Cardiac Surgery Baltimore, USA; ¹⁸LMU Klinikum Herzchirurgische Klinik und Poliklinik München, Deutschland; ¹⁹Max-Planck-Institut Department of Biochemistry Martinsried, Deutschland; ²⁰Revivicor Blacksburg, Deutschland; ²¹Max-Planck-Institut für Biochemie Martinsried, Deutschland; ²²LMU München Lehrstuhl für molekulare Tierzucht und Biotechnologie Oberschleissheim, Deutschland; ²³University of Maryland School of Medicine Baltimore, USA

Introduction: Heart failure is a leading cause of death worldwide, with donor shortage limiting allogeneic transplantation. Advances in genetic engineering of pigs and improved immunosuppressive regimens have enabled cardiac xenotransplantation to reach clinical stage. However, both first-in-human cardiac xenotransplantation ended in graft failure, highlighting the need for mechanistic insights to advance clinical application.

Material and Methods: We combined single-nucleus RNA sequencing and serial blood protein profiling to map porcine hearts adaptation or failure after implantation into non-human primates and compared findings to myocardial biopsies from two human xenografts. Two end-stage heart failure patients ineligible for allogeneic transplantation received genetically modified porcine hearts. The first patient survived 60 days, with porcine cytomegalovirus/roseolovirus (PCMV/PRV) detection thereafter. The second patient experienced progressive xenograft failure after 30 days and died of acute humoral rejection on day 40. The preclinical cohort comprising 19 pig-to-baboon xenotransplants was established for comparative analysis. Donors were, crossbred German Landrace and Large White pigs with triple genetic modifications. Recipients were Papio anubis and hamadryas baboons under immunosuppression.

Results: Both models showed near-complete replacement of porcine by recipient immune cells. The clinical rejection case exhibited marked expansion of human myeloids with upregulation of hypoxia-related genes (HIF1A, HK2, PFKFB3), ECM-remodeling factors (SPP1, VCAN) and innate immune mediators (C5AR1, CTSB).

Similarly, rejection xenografts (n=4) in the preclinical cohort exhibited profound immune dysregulation with expanded myeloid/lymphoid infiltration, marked by pro-inflammatory macrophage phenotypes (ISG induction, hypoxia-associated signatures) and loss of homeostatic programs. Vascular endothelial cells (ECs) in rejecting xenografts showed impaired JAG1-NOTCH signaling and elevation of remodeling pathways (VEGF, TGFβ, HSPG). Cardiomyocytes displayed a metabolic shift with depleted fatty acid oxidation, downregulated Ca²⁺-handling genes (CAMK2D, RYR2, CAMTA2) and induction of endoplasmic reticulum stress pathways, mirroring clinical rejection phenotypes. In the PCMV/PRV-positive human xenograft, myeloids displayed interferon-stimulated gene (ISG) enrichment (IFIT1/2/3, MX1/2, OAS1/2/3) and elevated cytokine/chemokine transcripts (CCL2, IL6), indicating a conserved interferon-dominated immune response. Consistent with this clinical observation, in three preclinical xenografts, PCMV/PRV RNA was detected in porcine vascular ECs, with capillary EC decrease and expansion of interferon-stimulated EC subsets. PCMV/PRV-positive ECs upregulated antiviral ISGs (MX1, IFIT1, OAS1/2/3) and pro-inflammatory mediators (CXCL10), with reduced junctional stability (TSPAN18 and VWF reduction) and pro-coagulant phenotypes (SERPINE1 elevation).

Conclusion: Human xenografts display striking molecular similarities with the preclinical model, including immune cell replacement and conserved cardiomyocyte stress signatures. These shared signatures provide a mechanistic framework for xenograft behavior and support translation towards clinical application.