Development of transplantable B-cell lymphomas in the MHC-defined miniature swine model

The MGH MHC-defined miniature swine model used in this study has been previously described [5]. A subline of the SLA^dd (DD) swine herd has undergone 11 generations of sequential brother-sister mating and have achieved a coefficient of inbreeding that exceeds 94%. This inbreeding has allowed for animals of the DD line to accept heart and reciprocal skin grafts without immunosuppression [9].

Heparinized whole blood samples were processed for peripheral blood mononuclear cell (PBMC) isolation by diluting blood with Hank’s balanced salt solution (HBSS, Thermo Fisher Scientific, Waltham, MA). Diluted blood was overlayed onto lymphocyte separation medium (Histopaque, Sigma-Aldrich, St Louis, MO) and underwent gradient centrifugation for cell isolation. The separated cell fraction was removed and washed twice with HBSS before resuspension in tissue culture medium. Tissue harvest from tumor recipients was performed in sterile manner. Cells were dissociated from connective tissue by mashing the piece of tissue with the flat end of a syringe plunger and passed through a 40 µm cell strainer placed over a 50 ml conical tube. Cells were washed twice with HBSS before being resuspended in tissue culture medium.

Tumor cell lines of porcine origin were cultured in RPMI 1640 media with 12% fetal bovine serum (FBS), 2 mM Glutamine, 0.1 mM non-essential amino acid mixture, 1 mM sodium pyruvate, 10 mM HEPES buffer (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), and 25 µM 2-Mercaptoethanol. All cultures were maintained at 37 °C with 5.0% CO₂. Tumor culture viability was determined by stable expansion for at least 1 month, with passage occurring weekly. Subclones were established by limiting dilution. Each sublcone plate was examined for colonies every 10–14 days, and wells of interest were further expanded. Cultures were frozen in liquid N2 for storage at − 180 °C with a concentration of 1 × 10⁷cells/ml in cryoprotective media (Lonza, Walkersville, MD) diluted to a final concentration of 50% FBS.

Cell turnover was characterized by establishing growth curves for each clone. Parent cultures were plated on 6-well plates in triplicate at an initial concentration of 5 × 10³ cells/well in 2 ml of culture media. Cell counts were performed daily using trypan blue exclusion dye (Corning, Manassas, VA) until a plateau was noticed. Culture media was replenished weekly.

PBMCs, tissues, and cell cultures of both porcine and murine origin were analyzed by flow cytometry using either porcine specific or reactive antibodies directly conjugated to either Fluorescein or Biotin. Biotin labeled antibodies were detected using streptavidin phycoerythrin

(PESA; PharMingen, San Diego, California, USA). The list of antibodies tested include, CD1 (76-7-4 IgG2a), CD2 (MSA-4 IgG2a), CD3e (898H2-6-15 IgG2a), CD4 (74-12-4 IgG2b), CD5 (9G12 IgG1), CD8 (76-2-11 IgG2a), CD9 (1038H-4-6 IgM), CD11b (M1/70 IgG2b, BioLegend), CD16 (G7 IgG1), CD21 (B-ly4 IgG1), CD25 (231.3B2 IgG1), CD45RA (Fg-2F9 IgG1), CD172a (74-22-15 IgG1), CD80 (16-10A1 IgG, BioLegend), Class I D (2.12.3a IgM), Class II DR(40D IgG2b), Class II DQD (1052 h-5-21 IgG1), anti-mu heavy chain (5C9 IgG1), anti-lambda light chain (K139.3E1 IgG2a), CD79a (HM-57 IgG1, Bio-Rad). Cell labeling was performed as previously described [10]. Flow cytometry was tested with FACSCalibur (Becton–Dickinson, San Jose, CA). Data was analyzed using FlowJo software version 10.

Tissue samples were harvested from all animals at end of study necropsy and preserved in formalin for hematoxylin and eosin staining and immunohistochemistry. The list of antibodies used for immunohistochemistry include, CD45, CD79a (HM-57 IgG1, Bio-Rad), anti-lambda light chain (K139.3E1 IgG2a).

PLHV-1 and PLHV-3 status was determined from stored PBMCs and tumor cells after recovery of DNA using the Gentra Puregene Kit (Qiagen Sciences, Germantown, MD), followed by Real Time TaqMan polymerase chain reaction (PCR) techniques described in detail [11, 12]. The PCR primers and probes used are listed in Table 1.

Figure 1 is a schematic representation of the experimental plan. In vitro tumor cultures were expanded over 1–2 months, pooled, and washed twice with phosphate buffered saline (PBS, Corning, Manassas, VA) before propagation in the pig model.

Porcine model: A mild conditioning regimen was initiated in the pig recipient 4 days prior to cell transplantation. The regimen included 4 days of intravenous (IV) anti-porcine CD3 immunotoxin [13] (0.05 mg/kg/dose administered twice daily), 100 cGy of total body irradiation, and 12 days of IV cyclosporine (target plasma range of 400-800 ng/ml). On day 0, the animal was injected with LCL-21353.2F8 at a dose of 1 × 10⁸ cells subcutaneously (SC) and 2.85 × 10⁸ cells/kg (4 × 10⁹ total cell dose) IV through an indwelling catheter placed in the external jugular vein.

Mouse model: Tumor cultures were removed from storage, propagated through two passages, and washed twice with PBS before transfer into NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ (NSG) mice (Jackson Laboratory, Bar Harbor, ME). Mice were given IV and or intraperitoneal (IP) injections as follows: Group 1 (n = 6) received 1 × 10⁷–2 × 10⁷ IP and 1 × 10⁷ cells IV from LCL-21353.2F8 cell line. Group 2 (n = 4) received either no or 0.5 × 10⁷ cells IP and 1 × 10⁷ cells IV from LCL-21353.2F8-23482 cell line.

All animal experiments were approved by and in compliance with Massachusetts General Hospital (MGH) Institutional Animal Care and Use Committee (IACUC). MGH is an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) recognized research institution.

Malignancy was suspected in a 2.5 year old female generation (G)11 inbred DD animal 21353, which presented with weight loss, mild head tilt, lethargy, and an elevated white blood cell count. Peripheral blood samples were drawn, processed, placed in culture and subcloned by limiting dilution in attempt to establish a tumor cell line in vitro. Subclone LCL-21353.2F8 was successfully propagated and selected for further study. Examination of cultures by light microscopy revealed large clusters of cells aggregated in suspension with rapid doubling time and stable growth characteristics, depicted in Fig. 2A–C. Immunophenotyping by flow cytometry identified MHC class I expression as being comparable to normal whole blood lymphocyte population, while all cells stained positive for MHC class II, representing a homogenous population (Fig. 3a, b). Lack of CD16 and CD172a surface markers typically expressed in myeloid lineage, as well as high expression of mu heavy chain and lambda light chain, suggests the clone is derived from a B-cell malignancy (Fig. 3a, b). Elevated expression of lymphocyte activation markers CD80 and CD45RA were also observed. Another subclone derived from animal 21353, LCL-21353.1B2, demonstrated less robust growth characteristics (Fig. 2D), as well as the absence of lambda light chain surface expression (Fig. 3a, b). Intracellular staining revealed LCL-21353.1B2 was positive for cytoplasmic lambda light chain (Fig. 3c).

Porcine lymphotrophic herpesvirus (PLHV) status was examined for 21353 PBMC banked shortly after disease onset, and was found to be positive for both PLHV-1 and PLHV-3 (primers and probes in Table 1). Following subcloning and long-term culture, subclones LCL-21353.2F8/1B2 derived from 21353 PBMC remained PCR positive for PLHV-3 but were negative for PLHV-1 (Additional file 1: Figure S1).

After a subcloned and characterized tumor line was established, we sought to propagate growth in vivo using a histocompatible miniature swine, G10 inbred DD animal 23482 (see Additional file 2: Table S1 for general characteristics of the animals). The recipient received mild conditioning prior to IV and SC transplantation of cell line LCL-21353.2F8 (see methods for detailed conditioning regimen). Cyclosporine levels remained within therapeutic range throughout the protocol. Unfortunately, the animal succumbed to acute respiratory complications 11 days after receiving tumor cell injections. The SC injection site was biopsied during necropsy, and a solid mass was removed.

The SC mass was harvested and cultured; after multiple passages, the remaining cells proliferated with similar physical and phenotypic characteristics to those of the primary cell line LCL-23153.2F8 (Fig. 2C, D). Activation markers CD80 and CD45RA remained consistent (Fig. 3a), while mu heavy chain intensity decreased and anti-lambda light chain increased after transplantation (Fig. 3a–c). Of note, LCL-23153.2F8-23482 displayed more homogeneity in both MHC class I and lambda light chain surface expression (Fig. 3b). In addition to the cell line derived from the SC site, tumor cultures were established from various sources including the spleen and peripheral blood, which were harvested at the time of necropsy. (Data not shown)

Tumor formation can be visualized macroscopically in the mass removed from the SC injection site (Fig. 4A). When analyzed histologically, tumor infiltrates became apparent in this space (Fig. 4B–E). Figure 4C, clearly shows an abundance of hematopoietic cells in this region, as characterized by CD45 staining, while CD79a cytoplasmic staining was evident throughout (Fig. 4D). Lambda light chain immunostaining was less convincing, as only a limited number of cells within the sample were positive (Fig. 4E).

To confirm the in vivo viability of each swine derived cell line, NSG mice were selected to receive both primary and secondary swine tumor cells at varying doses and routes of administration (described in Tables 2 and 3). Recipients of LCL-21353.2F8 manifested clinical signs of malignancy within 1 month of injection. The symptoms ranged from mild hind limb paralysis to hunched posture and decreased activity. Solid masses were found throughout the abdomen of each recipient upon necropsy (Fig. 5A, B). Mass size corresponded with the end of study date, at which time ~ 2 cm inguinal tumors were identified in all animals that continued until day 70 (Fig. 5A, B). In each of these cases, the solid tumors were vascularized and developed in strikingly similar fashion, surrounding adjacent organs without the formation of ascites. Tumor cells were demarcated within the mass by histologic analysis (Fig. 5C–E). The cells were large in size, with scant cytoplasm and the characteristics of a lymphocytic cell (Fig. 5C). The tumor cells were positive for CD45 confirming their hematopoietic origin (Fig. 5D) and positive for CD79a (Fig. 5E), validating their derivation from the original tumor cell line, which demonstrated CD79a staining in vitro. Recipients with earlier end of study dates had less significant gross tumor formation, with < 1 cm masses observed sporadically throughout the abdomen. The abdominal mass of an early terminated mouse was examined by flow cytometry using porcine specific anti-lambda light chain. A considerable shift was noted in the tumor mass, which stained brightly positive (Fig. 6a).

Unlike the primary tumor line, secondary transfer of LCL-21353.2F8-23482 into NSG mice did not lead to solid mass formation by end of study. Despite the absence of solid tumor masses in this cohort, 3 of 4 mice that received LCL-21353.2F8-23482 developed moderate hind limb paralysis within 2 weeks of cell injection. The presence of porcine tumor cells was confirmed at time of sacrifice by flow cytometry of isolated splenocytes. Figure 6b demonstrates a brightly stained population of surface porcine Ig lambda light chain and porcine MHC Class II among splenocytes from an NSG recipient of LCL-21353.2F8-23482.