Scientists Crack the Code for Mapping Gene Activity Across Whole Transplanted Limbs

Hand and face transplants can restore form and function to people who have lost limbs or suffered devastating injuries, but they come with a unique immunological puzzle. Unlike a heart or kidney, a transplanted limb is not a single organ. It is a composite of skin, muscle, nerve, blood vessel, fat, tendon and bone, each with its own biology and each capable of provoking a different immune response. A team of researchers at the University of Freiburg has now built the first reproducible workflow that lets scientists read the gene activity of an entire transplanted rat limb while preserving both its RNA and its anatomy, opening a new window onto how rejection unfolds across these extraordinarily complex grafts.

The technique at the heart of the study is spatial transcriptomics, a technology that maps gene expression directly onto intact tissue sections rather than requiring cells to be dissolved into a suspension first. The approach has been applied successfully to tumors, lung, ureter and skin, but until now it had never been established for a whole-limb vascularized composite allograft. The obstacle was not the analysis platform itself but the pre-analytical handling of large, heterogeneous tissue. A rat hindlimb contains compartments with sharply different mechanical properties and densities, and mishandling at any stage of harvest, freezing, embedding, sectioning or permeabilization can destroy RNA integrity, distort morphology or blur spatial resolution.

Writing in Frontiers in Molecular Biosciences, Marie Dreck, Elena Grabis and colleagues describe a refined workflow that spans every step from surgery to sequencing. The team used Lewis rats as limb donors and transgenic Wistar Kyoto rats carrying a human CD68 promoter driving green fluorescent protein as recipients, which allowed them to visualize host-derived macrophages as they infiltrated the graft. Tissue was harvested between 18 and 90 days after transplantation, with ischemia time kept as short as possible to protect the fragile RNA molecules inside the cells.

Several critical design decisions emerged from the team’s trial and error. First, the tibia and fibula were removed by careful periosteal elevation before embedding. Early cryosectioning trials showed that retained bone mechanically sheared and fragmented the adjacent soft tissue, producing uneven sections unsuitable for either histology or spatial analysis. Second, the freezing method mattered enormously. Rapid immersion in liquid nitrogen degraded RNA and introduced severe freezing artifacts, so the researchers instead used controlled freezing in an isopropanol bath on dry ice, which cooled the tissue more gradually at around minus 78 degrees Celsius and yielded uniformly preserved, intact sections with excellent adhesion to the slide.

Formalin fixation and paraffin embedding, the standard preservation method in most pathology laboratories, proved fatal for the workflow. When the team compared FFPE specimens with cryopreserved ones at multiple quality control checkpoints, the FFPE samples showed collapsed ribosomal RNA peaks and an accumulation of short fragments, indicating extensive degradation. Their cDNA profiles were fragmented and weak, and their sequencing libraries contained little detectable material in the 200 to 600 base pair range needed for the Visium platform. The cryopreserved samples, by contrast, showed strong 18S and 28S ribosomal peaks, broad cDNA distributions between 200 and 9,000 base pairs, and a clean library peak near 482 base pairs. Cryopreservation, the authors conclude, is essential because it maintains the intact poly(A) tails that the spatial barcoded oligonucleotides on the capture slide require to hook released messenger RNA.

RNA quality also depended on the extraction chemistry. RNA isolated from sections using the conventional phenol-guanidine thiocyanate-chloroform method yielded RNA Integrity Number values between 3.7 and 4.2, while a column-based Qiagen RNeasy approach pushed values up to between 6.0 and 6.9. That is still below the threshold of 7 recommended by 10x Genomics, the manufacturer of the Visium platform, yet every sample passed the downstream quality controls, with median log10 unique molecular identifiers per spot between 2.95 and 3.85, acceptable transcript complexity and low mitochondrial read percentages. The finding suggests that conventional RIN cutoffs may be overly restrictive for structurally complex composite tissue, and that RNA integrity measured after extraction does not always reflect the true biological state of the tissue.

Permeabilization, the step in which cell membranes are dissolved so that messenger RNA can diffuse to the capture spots, required its own careful calibration. Because permeabilization efficiency is tissue-dependent, the team tested incubation times of 30, 33, 36, 42 and 48 minutes on hindlimb cross-sections placed on tissue optimization slides. After reverse transcription with fluorescently labeled nucleotides, the fluorescence-positive area was quantified in ImageJ using an identical intensity threshold across conditions. The 33-minute sample showed the highest fluorescence-positive area at 32.4 percent, compared with 9.2 percent at 36 minutes and 17.4 percent at 42 minutes, indicating that this timing released RNA efficiently while preventing overpermeabilization and signal diffusion.

The immunofluorescence side of the workflow required separate ingenuity. The endogenous GFP signal in the hCD68 reporter rats survived only a brief fixation protocol: 15 minutes in pre-warmed 4 percent paraformaldehyde, followed by washes and blocking in goat serum with Tween-20. Even then the native fluorescence was faint, so the researchers added an anti-GFP antibody to amplify the signal, enabling reliable visualization of recipient-derived monocytes and macrophages infiltrating skin, muscle and connective tissue. Consecutive 10 micrometer cryosections were stained in parallel with the spatial transcriptomics runs, so the protein-level maps serve as complementary validation of the transcriptional data.

The payoff is visible in the downstream analysis of the representative sample. Unsupervised clustering identified seven spatially organized transcriptional compartments that corresponded cleanly to known anatomy: activated stroma, inflammatory myeloid regions, adipose tissue, B-cell and plasma cell zones, extracellular-matrix-remodeling stroma, keratinizing epithelium and peripheral glial structures. Histologically annotated vessels matched spatial enrichment of endothelial markers, and a vessel-distance map computed in the SPATA2 software confirmed the spatial correspondence. Cell-type composition was further refined by integrating the spatial data with a published Rattus norvegicus single-cell reference atlas using the cell2location and CytoSPACE algorithms.

The authors are candid about trade-offs. Removing the bone improved sectioning but excluded the osseous compartment from transcriptomic analysis, and immunofluorescence and spatial transcriptomics were performed on adjacent rather than identical sections, so precise transcript-to-protein colocalization remains out of reach. Yet the framework’s generality is its most exciting feature. The same principles of controlled freezing, bone removal, optimized permeabilization and GFP-preserving fixation can be transferred to other large, complex tissues, from musculoskeletal specimens to peripheral nerve and organ transplants. For a field in which rejection biology has long been read through macroscopic graft changes and conventional histology, the ability to watch immune cells arrive, localize and change their transcriptional programs across an entire living limb composite promises to reshape how acute and chronic rejection are understood, and ultimately how they are treated.

Subject of Research: Optimized spatial transcriptomics and immunofluorescence protocol for whole rat vascularized composite allograft hindlimb sections

Article Title: Preserving RNA integrity and morphology in whole rat allograft hindlimb sections: an optimized protocol for spatial transcriptomics and immunofluorescence

Article References: Dreck, M., Grabis, E., Schneider, L., Lind, M., Heinrichs, P., Christmann, T., Neidert, N., Brose, T. Z., Kreuzaler, S., Pankratz, F., Heiland, D. H., Eisenhardt, S. U., & Thomé, J. (2026). Preserving RNA integrity and morphology in whole rat allograft hindlimb sections: an optimized protocol for spatial transcriptomics and immunofluorescence. Frontiers in Molecular Biosciences, 13, Article 1910122. https://doi.org/10.3389/fmolb.2026.1910122

Image Credits: AI Generated

DOI: 10.3389/fmolb.2026.1910122

Keywords: spatial transcriptomics, vascularized composite allotransplantation, rat hindlimb model, RNA integrity, immunofluorescence, cryopreservation, Visium, graft rejection, macrophages, GFP reporter rats, cryosectioning, reconstructive surgery