Rapid derivation of cloning-competent cells from peripheral blood advances conservation biobanking Authors: Laura Knecht, Anthony Mastracci IV, Riley Harbert, Fatima Charles, Laura L Kuperman-Yao, Tyler Miyawaki, Gregory L Gedman, Kathleen M Pirovich, Jason Herrick, Linda Lay, Alba Ledesma, Steve Metzler, Dennis Milutinovitch, James B Papizan, Parker Pennington, Lexie Russell, Kerry Ryan, Kanokwan Srirattana, Kaitlin Steiger, Darya A Tourzani, Kelcey Walker, Rafael V Sampaio, Sven Bocklandt, Ben Lamm, Matt James, Shawn Walker, Beth Shapiro, James Kehler Published: 2026 ABSTRACT AbstractEstablishing viable cell lines from endangered species is essential for conservation, yet traditional fibroblast derivation from skin biopsies faces challenges including contamination risk and extended culture timelines. Here, we demonstrate that endothelial progenitor cells (EPCs) and pericytes isolated from peripheral blood represent robust alternatives to fibroblasts for biobanking. Compared to canid fibroblasts, canid blood-derived cells exhibit 2- to 3-fold faster doubling rates (15 to 20 h vs. ~35 h for fibroblasts) and reduced time to banked cell lines (1.5 to 2 wks vs. 3 to 4 wks for fibroblasts). Proteomic profiling of 32 canonical markers confirmed EPCs and pericytes represent distinct populations with lineage-specific molecular signatures. Optical genome mapping demonstrated equivalent genomic stability across cell types with no detectable structural variants or aneuploidies. Finally, interspecific somatic cell nuclear transfer (iSCNT) experiments confirmed both EPCs and pericytes generate viable canid embryos with efficiency meeting or exceeding fibroblasts. As a proof of concept for conservation cloning, iSCNT embryos made with gray wolf blood-derived cells had a 15% implantation rate following embryo transfer and resulted in six viable fetuses. These findings support integrating blood-derived cell banking into conservation programs, which enables opportunistic genetic preservation during standard management activities and expands options for genetic rescue through assisted reproductive technologies. Introduction Primary cell derivation offers significant potential for species conservation by proactively safeguarding genetic diversity, a crucial resource for species facing population fragmentation, inbreeding, and extinction. Major biobanks, such as the Frozen Zoo at the San Diego Zoo Wildlife Alliance, have amassed substantial collections (Mooney et al. 2023). The Frozen Zoo, for example, holds over 10,000 individual cell lines representing more than 1100 species and subspecies, including ~5% of threatened mammals, birds, amphibians, and reptiles listed on the IUCN Red List (https://sandiegozoowildlifealliance.org/pressroom/news-releases/san-diego-zoo-globals-frozen-zoor-hits-milestone-10000-cell-lines). Cryopreserved cells from collections like the Frozen Zoo can be reprogrammed into stem cells (Hutchinson et al. 2024) and/or used for reproductive cloning by conspecific or interspecific Somatic Cell Nuclear Transfer (iSCNT) (Cowl et al. 2024; Fritts 2022). This conservation cloning approach has been attempted for a wide range of amphibian and mammalian species (Novak et al. 2025a), with recent successful examples including introduction of historic diversity into black footed ferrets, Mustela nigripes (Wisely et al. 2015; Novak et al. 2024) and Przewalski’s horses, Equus przewalskii (Novak et al. 2025b). Tissue sampling techniques, such as taking skin punch biopsies of restrained animals or using dart biopsies of free-ranging wildlife, are two approaches to establishing fibroblast (FB) cultures (Plikus et al. 2021). While FBs are among the most common cell type cultured, obtaining clean FB cultures from skin biopsies is challenging in wild animals. Skin samples are naturally colonized by diverse commensal microbes that readily contaminate cell cultures despite rigorous sterilization efforts in the field, clinic, and laboratory. Over-sterilization through intensive treatment with antimycotics or antibiotics can impair long-term cell proliferation, alter gene expression and regulation, and adversely affect downstream experimental applications (Ryu et al. 2017, Skubis et al. 2017). In an effort to establish a less invasive, sterile sampling technique for bio-banking, genomic, and iSCNT studies, we isolated, expanded, and characterized adherent, stable cell populations from peripheral blood samples from canids under antibiotic free conditions. Endothelial progenitor cells (EPCs) are a rare (~1 in a million) cell type circulating in peripheral blood that have been previously isolated and cultured from blood from mice and humans (Jansen van Vuuren et al. 2023; see review by Yoder 2012). Pericytes are believed to reside on the outside of the endothelial layer of small blood vessels such as capillaries, interacting with EPCs (Attwell et al. 2016; Alarcon-Martinez et al. 2021) and are typically isolated from dissected and digested tissues (McErlain et al. 2025). Here, we show that both EPC and pericyte populations can be isolated and expanded from peripheral blood samples using the same method, and can serve as useful cell types for biobanking and cloning in conservation. Materials and methods Blood derivation and culture We collected blood samples from gray wolf (Canis lupus) and domestic dog (C. l. familiaris). All samples were collected in ACD blood tubes and transported overnight at room temperature (RT), under the supervision of veterinary staff. After recording blood volume, we diluted each sample with 1xDPBS (Cat#: 14190-144; Fisher Scientific) and 3% Heat inactivated FBS (HiFBS, Cat#: 16140071; Gibco) at a 1:3 ratio (1 volume blood:3 volumes of DPBS plus 3% HiFBS). We then gently and equally layered the diluted blood on top of 15 mL of Ficoll Paque Plus (Cat#: 95021-205; VWR) in 50 mL conical tubes. For example, 60 mL of diluted blood would be split into two 50 mL conicals with 15 mL Ficoll Paque Plus per conical. Next, we centrifuged the diluted blood with the Ficoll Paque Plus layer at 800 g for 25 min at RT with the break set to 0, and then aspirated the plasma layer. We then transferred the PBMC layer to another 50 mL conical tube and brought the total volume up to 50 mL with DPBS. We centrifuged the conical tube at 800 g at RT for 7 min with the brake set to 9. We aspirated the supernatant and then resuspended it in 30 mL of DPBS, and centrifuged at 400 g at RT for 5 min with a brake of 9. We then aspirated the supernatant and resuspended it in 1 mL of supplemented EGM (EGM with a total of 20% HiFBS) (Cat#: CC-3124; Lonza minus the gentamicin supplement). All EGM from this point forward is a total of 20% HiFBS EGM without gentamicin or other antibiotics. The cells were then counted using AOPI Cell Staining Solution (Cat#: CS2-0106-25 mL; Fisher Scientific), and resuspended in 25 mL of EGM and transferred to a Biocoat Collage T175 Flask (Cat#: 356487; Corning) and incubated at 37 °C, 5% CO2. The next day, we changed the media, and then continued media changes every other day until colonies had formed. Once colonies appeared, we performed daily half media changes until the colonies were 5 mm in diameter or were greater than 20 total colonies per flask. All media was fully antibiotic free for the entire duration of derivation and propagation. We deattached the cells after two washes of DPBS, using 50:50 ratio of 0.05% Trypsin (Cat#: 25300-054; Fisher Scientific) and Accutase (Cat#: SF006; Millipore) for 3 to 5 min at 37 °C, 5% CO2. Once cells were lifted, we quenched the solution with EGM (2× volume of trypsin:accutase) and then centrifuged it at 400 g for 5 min. The media was aspirated and then resuspended in 1 mL of EGM. Finally, we counted cells using AOPI and plated them onto Biocoat Collagen T175 Flasks and seeded at 1 to 2 × 106 per flask. When the cells were confluent, we repeated the same cell deattachment procedure and, instead of the cells being replated, they were centrifuged at 400 g for 5 min at RT and then resuspended at a concentration of 1 × 106/mL in EGM freezing media (30% EGM, 60% HiFBS, and 10% DMSO (Cat#: D2650-5 × 5 mL)). Blood derived cultures were antibiotic free from the time of derivation and remained that way for the life of the culture. FB derivation and culture We obtained dermal tissue for FB cell line derivation from wolves under managed care undergoing castration, as they were not needed for breeding. All biopsies were obtained by veterinary staff under anesthesia and analgesia using aseptic technique. Domestic dog tissue was not obtained due to the historical success and commercial availability of dog FBs. Ear notches were under 2 cm × 2 cm in size, and biopsies were 5 mm in diameter. To account for potential contamination from the field, we divided the wolf skin samples in half and generated an explant derived cell line and a digested derived FB cell line. We generated explant derived FBs by mincing the tissue into ~1 to 2 mm × 1 to 2 mm pieces. Pieces were then distributed across multiple scratched wells of a 6-well plate, with at least 6 minced pieces per well. These plates were scratched with an 18-25G syringe in a cross hatch manner in vertical, horizontal, and diagonal directions with at least 5 scratches per direction. After the tissue was placed onto the scratch marks with the epidermis oriented away from the bottom of the plate, the plate was placed into the 37 °C, 5% CO2 incubator for 5 to 10 min. After the incubation, the explants were attached to the bottom of the plate and 2 to 3 mL of cDMEM [DMEM high glucose, + sodium pyruvate, − glutamine (Cat#: 10313-021; Fisher Scientific), 10% HiFBS, 1× Glutamax (Cat#: AI2860-01; Fisher Scientific), 1× MEM-NEAA (Cat#: 11140-050; Fisher Scientific), 1× 2-mercaptoethanol (Cat#: 21985023; Gibco), 1× Penicillin–Streptomycin (P/S, Cat #: 15140122, Fisher Scientific)] was gently added down the walls to each well. Cells migrated out of the explants on average 5 days postplating. We generated digested derived FBs by mincing in 5 mg/mL collagenase IV (Cat#: C4-28-100MG; Sigma) and cDMEM and then incubating at 37 °C, 5% CO2 for at least 1 h. Every 15 min the mixture was minced for at least 3 to 5 min to aid the digestion process, in addition to adding 300 to 500 μL of the collagenase IV solution prior to placing it back in the incubator. The cells were then washed with cDMEM and plated onto a tissue culture treated T75 flask (Cat#: 229340; CELLTREAT) in cDMEM. For both methods, the cells were then cultured, passaged, and frozen at p2 to p3 in FB Freezing Media (30% cDMEM, 60% HiFBS, and 10% DMSO). At passage 1 onward, antibiotics were no longer included in the media. Doubling rate We used the following format to calculate the doubling rate across all species and cell types: doubling rate = hours in culture × LN(2)/LN(final cell number/initial cell number). LN is the equation for natural log. The cells were counted using a 1:1 dilution of cell suspension to AOPI on the Denovix CellDrop. Proteomics Mass spectrometry analysis was performed to quantitatively compare protein expression across different cell populations. We maintained one line each of blood-derived cells and FBs, under standard culture conditions until passage five and cultured to 80% confluency in two six-well plates for biological replicates. On the day of lysis, we gently washed cells twice with 1 mL of 1× DPBS, and discarded the supernatant. We lysed the cells by adding 75 μL of RIPA buffer (Cat#: J63306.AP) pre-mixed with HALT protease inhibitor (Cat#: 87785). Subsequently, we incorporated 1 μL of nuclease (Cat#: 88700) and mixed post-lysis. We quantified total protein concentration using ThermoFisher’s BCA Gold assay (Pub. No. MAN0029413 Rev. B.0). Lysates were stored under -80C in 1.5 mL low bind Eppendorf tubes (ThermoFisher Cat#: 90410) until all samples were collected for downstream mass spectrometry library preparation. We generated peptides via trypsin digestion and labeled using the TMT10plex isobaric labeling kit and manual (Cat#: 90406; ThermoFisher), (MAN0015866; Thermo Fi. We pooled labeled samples by species and subjected each pool to high-pH fractionation to yield eight fractions using methods adapted from Ruprecht et al. (2017). We analyzed each fraction using an Orbitrap Eclipse mass spectrometer processed with Proteome Discoverer 3.0 software. Peptides were annotated against species-specific UniProt databases. Following data acquisition, we filtered proteins based on false discovery rate (FDR), retaining only those with an FDR < 1% and a normalized average intensity >2000 units across grouped samples for downstream analysis. Reference genome assemblies We downloaded the gray wolf assembly GCA_905319855.2, which includes autosomes, sex chromosomes, and a mitochondrial genome for downstream Optical Genome Mapping and WGS analyses. FACS staining and analysis FACS staining and analysis of wolf blood-derived cells and FBs was performed to better characterize blood derived cell subpopulations. FACS was performed on the same lines as proteomics with one line per cell type. Staining was limited to two antibodies that had been raised and validated against dog CD34+ hematopoietic stem cells and CD90+ FBs. We prepared single cell suspensions of wolf blood derived cells and FBs at 3 × 105 cells per condition per line and then incubated cells for 30 min on ice in blocking buffer [90% DPBS (Cat#: 14190-144;Fisher Scientific) and 10% Goat serum (Cat#: 16210-072; VWR)]. The cells were then pelleted and washed twice with DPBS at 400 g for 5 min at RT. We incubated cell aliquots for 1-h at RT in the dark in 200 μL of Flow Cytometry Antibody Dilution Buffer (Cat#: 13616; Cell Signaling Technology) in parallel with the corresponding antibodies. We diluted the Rat IgG2b Isotype Control (APC) (Cat#: 17-4031-82; eBioscience), Mouse IgG1 Isotype Control (PE) (Cat#: 12-4714-82; eBioscience) 1:200, anti-dog CD90 (APC) (Cat#: 17-5900-42; eBioScience) at 1:20, and the anti-dog CD34 (PE) (Cat#: 12-0340-42; eBioscience) at a final concentration of 3 μg/mL. The dual stain condition was the same with CD90 at a 1:20 dilution and CD34 at 3 μg/mL dilution. All of these antibodies were diluted in the Flow Cytometry Antibody Dilution Buffer. After incubation, we pelleted and washed the cells three times with DPBS at 400 g for 5 min at RT. We resuspended cells into 200uL of cold FACS buffer (DPBS, 2% FBS (Cat#: 16140071; Gibco), 2.5 mM EDTA pH 8.0 (Cat#:15575020; Thermofisher), 25 mM HEPES (Cat#: 15630-080; Fisher Scientific), and 1% P/S and filtered through a 5 mL 40 micron capped FACS tube (Cat#: 196-19881-083; Spectrum Laboratory Products). We then analyzed cells on BD FACSDiscover S8 Cell Sorter, with a VBYR laser configuration, 100 μm nozzle size, and standard configuration and settings while using the YG1 (575 nm) and R1 (655nm) lasers. Angiogenesis A key feature of EPCs is their ability to spontaneously form microtubules in a dish, when plated in a supportive medium and matrix. We thawed Matrigel (Cat#: 354230; Corning) on ice and then plated 75 μL per well of a 96-well plate. The plate was then incubated at 4 °C for at least 1 h. We gently aspirated the Matrigel prior to plating (for blood derived cells: Cat#: 354409; Corning; For FBs: Cat#: 229196; CELLTREAT), and then incubated at 37 °C, 5% CO2 for 10 min. One of each of the three cell types was lifted by normal culture conditions, and plated at 1 × 104 for EPCs and Pericytes, and 2 × 104 for FBs in duplicates into both Matrigel coated wells and uncoated 96 wells with 100 μL media. The plate was then incubated at 37 °C, 5% CO2 for a maximum of 24 h, and imaged at least 4 times over the time period. In general, the assay showed tubule formation by 12 h, which demonstrates a population of endothelial origin rather than a population from a mesenchymal or other origin. Low density lipoprotein and acylated low density lipoprotein uptake EPCs are also known to preferentially uptake acylated Low Density Lipoproteins (acLDL), providing a second functional assay to interrogate cell identity in vitro. We grew one type of each blood derived cell and FBs, as a negative biological control, under standard cell culture conditions until the cell lines had reached 65% to 80% confluency. On the day of passaging, we collected 1.0 × 105 cells from each cell line, pelleted at 400 g for 5 min at RT and washed cells with 1× DPBS, and then resuspended them in serum-free DMEM (DMEM, FBS, 0.3% MACS BSA (Cat#: 130-091-376, Miltenyi Biotec)) for FBs and serum free EGM (EGM,-FBS, 0.3% MACS BSA) for blood-derived cells. We then transferred cells onto Corning 96-well plates for blood-derived cells and CELLTREAT 96-well plates for FBs with 1.0 × 104 cells per condition. The plates were then incubated at 37 °C, 5% CO2 for 16 h. We washed the cells with LDL Assay buffer (1× DPBS, +calcium, +magnesium (Cat#: 14040133; Gibco), 0.3% MACS BSA) and resuspended in serum-free DMEM for FBs and serum free EGM for blood derived cell lines. Low Density Lipoprotein from Human Plasma, BODIPY FL complex (Cat#: L3483, Invitrogen) was diluted 1:100 for a final concentration of 10 μL/mL and added to the appropriate wells in replicates. Low Density Lipoprotein from Human Plasma, Acetylated, Alexa Fluor 594 Conjugate (Cat#: L35353; Invitrogen) was also diluted 1:100 for a final concentration of 10 μL/mL and added to the appropriate wells in replicates. Cells were returned to the incubator at 37 °C, 5% CO2 for 2.5 h. After incubation, we washed the cells three times with the LDL Assay Buffer and then imaged and analyzed on the Incucyte SX5 utilizing the Orange/Green optical modules. Karyotyping To determine whether canid EPCs and Pericytes remained karyotypically stable in culture, we shipped live canid (gray wolf) cell cultures to an independent cytogenetics laboratory (KaryoLogic Inc.) in T25 flasks at 25% and 50% confluency so that they would arrive subconfluent, thus minimizing contact inhibition of the cells. Mitotic spreads were prepared and G-banded karyotypes assembled from 20 cells per line to determine the average karyotype and detect the presence of any aneuploidies. Optical genome mapping We performed Optical Genome Mapping (OGM) analyses of all derived cell lines as a higher resolution genomic integrity assay, using the Bionano Saphyr platform according to manufacturer specifications. Briefly, we isolated ultra-high-molecular weight (UHMW) genomic DNA from 500 k to 750 k cells per line, as described in the Bionano Prep SP-G2 Fresh Cell Pellet DNA Isolation Protocol (Bionano, Doc no. CG003). We then fluorescently labeled UHMW gDNA by the Direct Label Enzyme (DLE-1) staining and backbone visualization (Bionano Prep Direct Label and Stain (DLS) Protocol, Doc no. 30206). We loaded the labeled and stained UHMW gDNA sample onto a Saphyr Chip and into the Saphyr System Instrument (Bionano Saphyr System User Guide, Doc no. 30247). We identified structural variations and low variant allele fractions using Bionano software and protocols (Solve Theory of Operation: Structural Variant Calling, Doc no. CG-30110). The resolution for detecting variants with this method was >500 bp up to the chromosome level, when compared to short-read WGS assemblies. Interspecies somatic cell nuclear transfer For gray wolf iSCNT, we used either EPCs or pericytes from C. lupus as nuclear donors. Synchronization of EPCs and pericyte donor cells for iSCNT was achieved by allowing cultures to grow to complete confluence to induce contact inhibition and enter cell cycle (G0) arrest. We isolated oocytes from donor dogs in spontaneous estrus and performed enucleation, reconstruction, electroporation, subsequent activation, and overnight in vitro culture (IVC), as described previously (Kim et al. 2012). As prolonged canid IVC has not been established, embryo transfer was performed to assess the developmental competence of these iSCNT gray wolf embryos. A single recipient dog in natural estrus underwent surgical oviductal embryo transfer and was supported during surgical recovery under IACUC approved veterinary care and analgesia, following AVMA and USDA required best practices. As the two unrelated donor male wolves under managed care were being castrated, and their cloned offspring were not needed for captive breeding programs, the single pregnancy was terminated at mid-gestation by surgical collection. The domestic dog recipient underwent a timed ovario-hysterectomy at 33 days of pregnancy. The donor dog was supported during surgical recovery, under IACUC-approved veterinary care and analgesia, following AVMA and USDA best practices, and she was adopted following IACUC and American Humane guidelines. Results Derivation success and timeline We successfully derived adherent cell populations from peripheral canid blood samples requiring a minimum of 15 mL of blood. (Table 1, Fig. 1). Blood-derived cell cultures showed an overall derivation success rate of 80% (Table 1) with initial colony growing onset of 7 ± 1 days post-processing. Due to rapid proliferation, all blood-derived cultures generated 20 million low-passage (P1–2) cells within 10 to 14 days of initial plating. Table 1Cell line derivation success rates by species and source.Blood-derived CultureEar Notch Fibroblast Culture Genus Species Success Failure % Success Success Failure % Success Canis lupus 20100%30100% Canis lupus familiaris *2167%00 Total Success Rate 480% *Excluded fetal skin derivation Cell line derivation success rates by species and source. *Excluded fetal skin derivation Cell lines can be generated from both fibroblasts and blood-derived cells (pericytes and endothelial progenitor cells, or EPCs) isolated from blood draws. Cells isolated from blood show an expedited timeline to handoff for downstream biobanking applications compared to fibroblasts. Growth characteristics, morphology, and surface markers Doubling rates differed consistently between cell types (Fig. 2). EPCs exhibited the fastest average doubling rate of 15 h, while pericytes averaged 20 h, and FBs showed more variable rates exceeding 35 h. Morphologically, FBs presented as elongated spindle-shaped cells with central nuclei across species (Fig. 3). EPCs displayed small, cobblestone-like morphology with high cell aggregation. Pericytes showed FB-like morphology, but with less cellular elongation (Fig. 3). Doubling times by passage for canid cell types (a–c). Data were collected via Incucyte cell imaging with the cell-by-cell AI analysis software trained to analyze each cell type. Cells were imaged daily to capture live cell count, which was used to calculate the average doubling rate across each passage. Lonza EGM is Lonza’s endothelial growth media and cDMEM is complete DMEM (10% FBS, 1× Glutamax, 1× MEM-NEAA, and 1×2-mercaptoethanol). Morphology of canid (a–c) cells. All photos were taken under the phase imaging of the Incucyte. Images were captured with the 10× objective with a scale bar length of 400um. Flow cytometric analysis of canid cells using anti-dog CD34 and CD90 antibodies showed distinct surface marker profiles (Fig. S1). EPCs contained a discrete 25% population of CD34+ cells. Pericytes contained a minor ~12% sub-population of CD90+ cells with negligible (<1%) detection of CD34. FBs were uniformly 100% CD90+ with a minor 3.5% subpopulation % CD34+ and CD90+ cells, possibly reflecting heterogeneity in early passage cultures. Functional characterization In vitro angiogenesis assays demonstrated that EPCs formed tubule-like networks when plated on extracellular matrix (ECM) substrate, while FBs showed no morphological changes (Fig. 4a). Pericytes adhered to the outer surface of EPC-derived tubule structures in co-culture experiments. Neither pericytes nor FBs formed tubules when plated alone. Functional assays for cell characterization. (a) Results of the angiogenesis assay. Canid EPC, pericyte, and EPC + pericyte images were taken using the Zeiss Primovert with a 10× objective, and have scale bars at 100 μm. The remaining image was taken using the Incucyte at a 10× objective; scale bars at 800 μm. EPCs formed tubule networks, while pericytes adhered to the outer layer of the EPC-tubule network. Pericytes and fibroblasts plated alone did not form tubules. (b) Results of the low Density Lipoprotein & Acylated low density lipoprotein uptake assay. A “−” below a bar indicates negative controls, while E, P, and F indicate EPCs, pericytes, and fibroblasts, respectively. Cell fluorescence was measured using Incucyte Cell-by-Cell AI Analysis software trained on these cell types. (c) Fluorescent microscopy images of cells taken by the Incucyte at a 10× objective with the Green and Orange fluorescence channels to generate positive cell fluorescence data. Fibroblasts have a lower level of BodipyFL-tagged (green) LDL uptake than those of blood derived cell lines, while pericytes have lower levels of AlexaFluor594-tagged (red) AcLDL uptake than EPCs and fibroblasts from the same species. These scale bars are all at 800 μm. LDL and AcLDL uptake assays revealed differential internalization patterns (Fig. 4b and c). Both EPCs and pericytes showed elevated LDL uptake compared to FBs. For AcLDL uptake, EPCs and FBs demonstrated the highest uptake, with pericytes showing the lowest uptake levels. Uptake is defined here as the percent of cells which have fluorescence rather than the absolute intensity of the cells themselves. This differential acLDL uptake can be used to distinguish EPCs from pericytes in cultures originating from blood samples. Genomic stability assessment OGM with ~80-fold genomic coverage confirmed expected chromosome copy numbers and sex chromosome compositions for all cell lines (Fig. 5). OGM data from male canid samples were concordant with independent G-banded karyotype analysis (Fig. 5). Both male gray wolves from which cell lines were derived had 38 pairs of autosomes and one X and one Y chromosome. No significant structural variants or aneuploidies were detected, confirming karyotypically normal primary cells suitable for downstream applications. (a) Optical genome mapping (OGM) plots summarizing chromosome copy numbers for canid cell lines. (b) G-banded karyotype analysis of canid pericytes showing normal karyotype. Proteomic characterization of cell populations Analyses of gene expression data from 33 canonical cell markers revealed distinct clustering of the three primary cell populations: EPCs, FBs, and pericytes (Fig. 6). Results of proteomic analyses for replicate samples of each cell type. (a) Heatmap showing relative expression of canonical protein markers for cell type determination in canid cells. Protein expression was normalized, and color intensity reflects z-scores from −2.58 (blue, lowest) to 2.58 (red, highest); gray indicates undetected protein from the mass spectrometer. Log2(fold-change) and −Log10(P-value) for each gene are provided in Table S1; (b) principal component analysis (PCA) of all samples using 18 canonical markers present in all cell lines: VIM, COL1A1, ANXA5, COL3A1, CSPG4, ITGA5, ITGB1, CD44, ENG, ICAM1, PDGFRA, MMP14, THY1, COL4A2, PDGFRB, DCN, TGFBI, and ITGB5. The first two principal components account for 97% of the total variance. Fibroblasts. Canid FBs exhibited elevated expression of ECM proteins and markers associated with activated FBs, including Fibroblast Activation Protein Alpha (FAP), which is an established marker for activated FBs in tissue remodeling (Tillmanns et al. 2015), and Insulin-like Growth Factor-binding Protein 3 (IGFBP3), which is involved in modulating IGF signaling and ECM interactions. We observed a high abundance of major structural ECM components Collagen Type I Alpha 1 Chain (COL1A1), Collagen Type III Alpha 1 Chain (COL3A1), which aligns with the known matrix-synthesizing capacity of FBs. Vimentin (VIM), which is crucial for cell structure, and Platelet-Derived Growth Factor Receptor Beta (PDGFRB), a key receptor in FB proliferation, as well as Tenascin C, an ECM glycoprotein induced during tissue injury and remodeling (Midwood et al. 2011). When comparing EPCs to FBs, Log₂(Fold-Change) values for these markers (Table S1) were consistently negative and statistically significant (Log₁₀(P-value) > 2.8). Pericytes. Pericytes demonstrated a high abundance of proteins involved in cell adhesion and ECM interaction, including Integrin Beta 5 (ITGB5) and Intercellular Adhesion Molecule 1 (ICAM1), both of which are critical for cell–cell and cell–matrix interactions (Pang et al. 2018). High expression of matrix metalloproteinase-14 (MMP14) suggested potential for ECM degradation and turnover, key functions in vascular remodeling. Other highly expressed proteins include Transforming Growth Factor Beta Induced Protein (TGFBI) and Decorin (DCN), suggesting active ECM engagement, and Melanoma Cell Adhesion Molecule (MCAM/CD146), an important surface molecule on pericytes that downregulates TGFB signaling to impact endothelial cells junction integrity (Chen et al. 2017). In addition to FBs, PDGFRB is a canonical marker of human and mouse pericytes involved in their recruitment and activation (reviewed by Yao 2022). PDGFRB was abundant in canid pericytes and negligible in EPCs. EPCs. EPCs exhibited markers consistent with endothelial-like cells, including high expression of Endoglin (ENG), a TGF-β co-receptor, and Cadherin 5 (CDH5/VE-Cadherin), which is essential for endothelial cell–cell junctions (Siavashi et al. 2016), detected specifically in EPCs. Platelet-Derived Growth Factor Receptor Alpha (PDGFRA) is expressed during fetal cardiac development and in rare post-natal cardiac endothelial cells (Chong et al. 2013) Interestingly, EPCs had abundant expression of PDGFRA when compared to pericytes and FBs. Canid EPCs expressed canonical human and mouse EPC markers (reviewed by Hassanpour et al. 2023) including CD34, PECAM1, SELE, NOS3, ITGA9, ESAM, and HGF in abundance that were not detected in canid FBs or pericytes (Caiado and Dias 2012; Liu et al. 2019). A PCA used data from 18 markers quantified across all cell lines (VIM, COL1A1, ANXA5, COL3A1, CSPG4, ITGA5, ITGB1, CD44, ENG, ICAM1, PDGFRA, MMP14, THY1, COL4A2, PDGFRB, DCN, TGFBI, ITGB5; the other 14 were excluded due to missing data; Fig. 6b). Some marker expression patterns deviated from canonical assignments (e.g. low VWF in EPCs, high VWF in canid FBs; variable FN1 and CD44 expression across cell types), reflecting context-dependent expression profiles. Interspecies somatic cell nuclear transfer Both pericytes and EPCs successfully generated viable iSCNT gray wolf embryos. iSCNT was performed as existing canid IVC systems do not support development beyond early cleavage, such that transfer of iSCNT embryos into a surrogate dog is required to assess their developmental potential. We transferred 21 embryos reconstructed from pericytes and 18 from EPCs into a single recipient domestic dog in estrus (Fig. 7a and b). Transabdominal ultrasound at Days 21 and 23 post-embryo transfer, the earliest time when ultrasound can be definitively used to diagnose pregnancy in the dog, confirmed implantation of six viable embryos with active heartbeats (Fig. 7c). Timed ovariohysterectomy at Day 33, the half-way point in canine gestation, yielded six morphologically normal fetuses. These time points were selected to provide insight into the potential of cloned embryos to implant, establish pregnancy, and continue fetal development to mid- and full-term gestation. Day 33 fetuses were evaluated for gross morphology prior to tissue collection, and all were comparable to standards published for non-cloned, age-matched canine fetuses (“Canine Embryonic Atlas”, 2017). To establish cellular origin for each fetus, we developed a custom SNP panel for a specific locus on chromosome 2. A 272 bp PCR allowed us to characterize four SNPs differentiating donor lines: wolf 1 (pericytes) and wolf 4 (EPCs). This genotyping panel identified four fetuses derived from pericytes and two from EPCs, demonstrating viable implantation rates of 19% (pericytes) and 11% (EPCs). Brightfield images of preimplantation embryos and postimplantation ultrasound images. (a) Early cleavage stage embryos are visible in overnight cultures of iSCNT gray wolf embryos reconstructed with EPCs and (b) pericytes. (c) A transabdominal ultrasound image shows a sagittal image of one of six viable gray wolf fetuses (orange arrow) at 23 days after ET. Discussion Blood-derived cells offer practical advantages for conservation biobanking This study demonstrates that EPCs and pericytes isolated from peripheral blood are a robust alternative to dermal FBs for conservation biobanking and iSCNT. Blood-derived cells exhibit 2 to 3-fold faster doubling rates (15 to 20 h vs. >35 h for FBs) and reduce the timeline for generating banked cell lines from 3 to 4 wks to 1.5 to 2 wks (Figs 1 and 2). Combined with genomic stability equivalent to tissue-derived cells (Fig. 5) and demonstrated viability for generating cloned embryos across three mammalian genera (Fig. 7), blood-derived cells offer practical advantages for accelerating genetic rescue efforts. The primary advantages of peripheral blood over skin biopsies are the initial cell propagation length and ease of acquisition during existing veterinary workflows. While FB derivation from skin is well-established, the small tissue fragments typical of dart biopsies may prolong the culture time due to limited starting material. Peripheral blood offers a more abundant starting material, which may mitigate the risks associated with sample degradation or contamination, both factors relevant in field-based collection where laboratory proximity varies. The sterile, closed-system collection of blood in anticoagulant tubes contrasts with the microbial burden inherent to skin surfaces. The consistency of blood-derived cell doubling rates may facilitate standardized biobanking workflows across diverse conservation programs. Whole blood samples are already routinely collected with anticoagulants for genomic studies and non-viably frozen in the field without cryoprotectants for DNA extraction. With advanced planning, blood collection protocols could preserve options for both immediate genomic analysis and future cell line establishment, maximizing the value of each sampling event. This dual-purpose approach is particularly relevant for cryptic, nocturnal, or otherwise difficult-to-sample species where opportunities for biological sample collections are rare. Molecular and functional characterization confirms distinct cell populations Comprehensive proteomic profiling established that EPCs, pericytes, and FBs are biologically distinct populations with lineage-specific molecular signatures (Fig. 6). FBs exhibited elevated ECM production markers (COL1A1, COL3A1, FAP), pericytes showed high abundance of adhesion and vascular remodeling proteins (ITGB5, ICAM1, MMP14, MCAM), and EPCs expressed endothelial-associated markers (ENG, CDH5, PDGFRA) along with canonical EPC markers (CD34, PECAM1, SELE, NOS3) detected exclusively in EPCs (Fig. 6, Table S1). The statistical significance of expression differences (Log₁₀(P-value) > 2.8) and consistency across biological replicates confirm that blood-derived cultures yield distinct cell types rather than heterogeneous mixtures. For practical biobanking characterization without mass spectrometry, morphological assessment combined with doubling rate measurements provides reliable first-pass identification (Figs 2 and 3). EPCs consistently display cobblestone morphology and ~15-h doubling times, while FBs show spindle morphology with >35-h doubling times. Pericytes exhibit intermediate characteristics with FB-like morphology but faster doubling (~20 h). Functional assays further distinguish cell types within 24 h: tubule formation on ECM occurs exclusively in EPC cultures, while differential uptake of fluorescently labeled LDL versus acetylated-LDL distinguishes EPCs from pericytes (Fig. 4). These species-agnostic methods should enable biobank curators to characterize cell lines from taxonomically diverse samples as collections expand to under-represented vertebrate lineages. Both EPCs and pericytes support viable iSCNT outcomes Our iSCNT experiments demonstrate that blood-derived cells generate viable embryos with efficiency comparable to or exceeding traditional FB donors. In gray wolves, both pericytes (19% implantation) and EPCs (11% implantation) contributed to viable fetuses, with the combined rate (15%) exceeding prior published wolf iSCNT rates using FBs (0.8%) (Kim et al. 2007). Collectively, these results indicate that cell source (blood versus tissue) and specific blood-derived cell type do not compromise cloning outcomes. Current limitations and opportunities for optimization The stochastic emergence of either EPC-dominant or pericyte-dominant cultures from blood samples presents both a challenge and an opportunity. In our study, replating the total buffy-coat fraction resulted in primary colonies of both cell types adhering to traditional cell culture substrates and media optimized for EPC derivation and expansion. Both cell types formed primary colonies in a similar time window and could be dissociated, replated, and grown rapidly. While admixture of both cell types in cultures cannot be ruled out without single-cell analyses, the formation of cultures appeared stochastic with either EPCs or pericytes forming the predominant cell type in individual cultures. Selective enrichment may be achievable through substrate modification, growth factor supplementation, or differential adhesion protocols, enabling targeted derivation based on intended downstream applications. Both EPCs and pericytes participate in angiogenic signaling and repair and provide cell models for studying pathophysiological processes like atherosclerosis, making such optimization valuable for future studies beyond conservation applications. Future work will be necessary to optimize derivation success for different species. Species-specific body size, hematocrit, and circulating EPC/pericyte frequency are likely to influence blood volume requirements, and establishing taxon-specific guidelines would improve collection success rates while minimizing physiological stress to sampled individuals. In addition, robust field-to-lab workflows that maintain both cell viability and genomic quality would maximize the utility of opportunistic sampling, particularly for endangered species where repeated sampling is not always feasible. Integration with conservation strategies and existing biobanking infrastructure The choice between blood and tissue sampling should ultimately be guided by species-specific factors, including restraint safety, veterinary access, and protocol success rates. Blood sampling and viable cell propagation provides a complementary approach to biobanking to intervene in species with compressed timelines facing precipitous population declines. Recent successes introducing cloned historically underrepresented founders into black-footed ferret and Przewalski’s horse populations (Novak et al. 2024; Novak et al. 2025b) illustrate how biobanked cells restore genetic variation unavailable through conventional breeding. Here we demonstrate how blood-derived cells can also be used to expand the pool of individuals to be biobanked before unique genotypes are lost. Integration of blood-derived cell lines with existing repositories such as the Frozen Zoo requires no additional steps. Blood-derived cells can be processed using standard cryopreservation protocols and stored under identical conditions as tissue-derived lines. Unlike the technical demands of generating and biobanking stem cells (Hutchinson et al. 2024), blood-derived cells can be processed using standard cryopreservation protocols and stored under identical conditions as tissue-derived lines. Ultimately, however, the success of any biobanking-to-cloning effort depends on co-development of Assisted Reproductive Technologies (ARTs). The iSCNT results reported here were only possible because of established ART protocols in domestic dogs, emphasizing that cloning advances depend on robust foundational reproductive biology research (Cowl et al. 2024; Hildebrandt and Holtze 2024). Finally, the selection of genetically valuable individuals for cloning should be performed in coordination with local and international community members invested in the propagation of endangered species and populations (Gray et al. 2023; Novak et al. 2025a). Support for international and Indigenous groups advocating for in situ and ex situ conservation efforts is a critical parallel need alongside biobanking and cloning efforts (Bolton et al. 2022; Hayah et al. 2025). By integrating blood-derived cell collection into routine management activities, conservation programs can opportunistically capture genetic diversity during standard veterinary procedures, expanding options for genetic rescue when traditional reproductive technologies cannot achieve sufficient genetic diversity. The minimal infrastructure requirements make this approach accessible to field programs with limited resources, broadening the pool of individuals and species represented in global biobanking efforts. Supplementary Material