Tissues analyzed in this study were harvested from Brangus (Angus and Brahman cross) steers (Bos taurus) 1817 and 1851, housed and immunized at Texas A&M University Veterinary Medical Park under TAMU IACUC 2022−0059. The animals were immunized with two TIM-3 (T cell immunoglobulin and mucin domain 3) were given intradermally to prime and boost an immune response in steer 1817, and against PD-1 (programmed cell death 1 antigen) in steer 1851. Both immunogens were recombinantly produced extracellular domains of the human protein. While the two immunization strategies suggest different experiments, the data were drawn with distinct metadata contexts, allowing the evaluation of biological patterns that represent the natural range of responsiveness within this sample set. See Supplemental Fig. 1 for timeline, components of injections, and antibody titers from isolated peripheral blood confirming immunization through antigen-specific IgG enzyme-linked immunosorbent assays (ELISA). Briefly, 96-well plates (BD Biosciences, San Jose, CA) were coated with 100 µL of 1 ng/µL antigen in filtered, autoclaved coating buffer at pH 9.6, containing 2.93 g NaHCO3 and 1.5 g Na2CO3 in 1000mL dH2O. The plates were then sealed and incubated overnight at 4℃. The next day, the plates were decanted and blotted on a paper towel three times before washing the wells five times with 200 µL of room-temperature 1X Tris-Buffered Saline (TBS) containing 0.1% Tween 20 (TBST). Next, 200 µL of blocking buffer consisting of 10% filtered equine whole sera in 1X TBST. The plates were then sealed and incubated at 37℃ for 60 min. Plates were washed five times as previously described before incubating each well with 150 µl of the respective diluted bovine serum or blocking buffer (background well). The plates were then sealed and incubated at 37℃ for 120 min. Thereafter, the wells were washed five times as described earlier before placing 150 µl of 1:4000 Peroxidase-conjugated AffiniPure goat anti-bovine IgG (H + L) (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) or 150 µl of the blocking buffer. The plates were then sealed and incubated at 37℃ for 60 min. Afterwards, the wells were washed four times with 200µL of room temperature 1X TBST, then washed three times with 200µL of room temperature 1X TBS. Subsequently, 150µL of ready-to-use 3,3′,5,5′-Tetramethylbenzidine substrate was added to each well and incubated at room temperature for 10 min or until a blue color appears. If no color develops after 10 min, further incubation was performed at 37℃ and reassessed periodically. Once developed, the reaction was stopped with 150µL of 1 M H2SO4 and then read on a OD-450 plate reader.

Figure 1 depicts the 24 tissues collected from both steers in order from anterior to posterior based on four major tissue categories: systemic, secondary lymphoid, gastrointestinal tract, hematopoietic, and circulatory immune tissues. Tissues are listed as follows alphabetically with an unique acronym: abomasum (ABOM), bone marrow (BOMA), brain (BRAI), caudal lobe of the lung (CAUD), cecum (CECU), colon (COLO), cranial lobe of the lung (CRAN), duodenum (DUOD), epigastric lymph node (EPLN), gallbladder (GALL), ileum (ILEU), ileal Peyer’s patch (ILPP), jejunum (JEJU), mandibular lymph node (MALN), mesenteric lymph node (MELN), medial retropharyngeal lymph node (MRLN), muscle (MUSC), omasum (OMAS), peripheral blood mononuclear cells (PBMC), reticulum (RETI), rumen (RUME), superficial cervical lymph node (SCLN), spleen (SPLE), and subiliac lymph node (SULN). Once collected, all tissues were preserved for 24 h in RNAlater (ThermoFisher Scientific, Waltham, MA, USA) at 4 °C and then stored at −20 °C until processing.

Tissues collected from steers. Tissues are in order from anterior to posterior based on four categories (systemic, secondary lymphoid tissues, gastrointestinal tract, and hematopoietic and circulatory immune tissues). Each tissue has been assigned a four-letter acronym. Created in BioRender. Hissen, K. (2026) https://BioRender.com/cn2orzb

Total RNA was purified from each tissue using Ambion TRIzol reagent (ThermoFisher Scientific) following the manufacturer’s instructions. Briefly, about 100 mg of tissue was homogenized in TRIzol reagent using a Tissuelyser II (Qiagen, Hilden, Germany). Additionally, 9.0 × 106 (steer 1817) and 3.3 × 106 (steer 1851) PBMCs were suspended in TRIzol and homogenized by passing the suspension through a 20-gauge 1.5” needle attached to a 1 mL syringe (BD, Franklin Lakes, NJ, USA). RNA quantity (ng/µl) and quality (260/280) were measured using a Nanodrop® ND-1000 spectrophotometer (ThermoFisher Scientific).

Using the mRNA, cDNA was synthesized using SuperScript III First-Stand System (Invitrogen, Waltham, MA, USA) reverse transcriptase following the manufacturer’s instructions with 5 µg of RNA. Both quantity and quality of the synthesized cDNA were measured using a Nanodrop®. The cDNA concentrations were standardized from all tissues to 400 ng/µl for PCR reactions.

Using Geneious Prime 2022.0.1 (https://www.geneious.com), a single forward primer was designed to a conserved region of both IGHV1-7 (ultralong HC V) and IGHV1-10 (canonical HC V), for ultralong CDR H3 arises from multiple VH families. Reverse primers were designed to the CH1 domain exon of IgM, IgD, IgG, IgE, and IgA constant regions (Walther et al. 2013. Deiss et al. 2019). Primers were tested both individually and as a multiplex of all five reverse primers using pooled calf cDNA (as above), choosing to use a multiplex of the reverse primers in subsequent reactions. DreamTaq PCR master mix (Invitrogen) was used to amplify target sequences and PCR products were visualized using a 1.5% agarose gel to confirm band sizes (Supplemental Fig. 2). Each primer was then appended with partial Illumina adaptors for sequencing. Unique barcodes of eight base pair (bp) in length were created and inserted between the P5 Illumina adaptor and the forward primer sequence (P5 – barcode – primer) to create primers designed to distinguish sequenced amplicons by tissue (Supplemental Table 1). This second set of primers was retested using the same process as above, in addition to Phusion High-Fidelity PCR Master Mix polymerase (ThermoFisher Scientific).

To obtain IgH VH-CH1 amplicon sequences from each tissue, a final 50 µl PCR reaction was ran using 400 ng cDNA, primers (5 µM forward and a 5 µM reverse primer mix) containing adaptors and barcodes, and Phusion High-Fidelity PCR Master Mix with the following cycling parameters: 30 s at 98 °C (initial denaturation), 30 cycles of 30 s at 98 °C (denaturation), 30 s at 58 °C (annealing), and 60 s at 72 °C (extension), with a final extension of 10 min at 72 °C. After PCR amplification, the presence of resulting bands was confirmed and amplicon lengths were validated using a 1% agarose gel (run for 75 min at 83 V using 5 µl GelGreen, 1 µl GeneRuler 1 kb Plus DNA Ladder (ThermoFisher Scientific), and 5 µl of the sample mixed with 1 µl of 6x loading buffer) as shown in Supplemental Figs. 3 and 4. ExoSAP-IT Express reagent (ThermoFisher Scientific) was used to purify the remaining PCR product (45 µl) enzymatically, then the amplicon was measured for DNA concentration using a Qubit 4 fluorometer (Invitrogen). These targets were confirmed through sequencing.

We submitted a pooled sample containing 2 µl of each tissue PCR product for amplicon sequencing (Amplicon-EZ Illumina-based sequencing services; Azenta Life Sciences, South Plainfield, NJ, USA). A total of 566,628 paired reads were obtained via Amplicon-EZ sequencing. Using default options of the BBMerge paired read merger in Geneious Prime (Biomatters, Ltd, Auckland, New Zealand, v2023.1.1), the reads were merged into 180,787 sequences and sorted by barcode, allowing a single mismatch in the barcode sequence. The tissue indicated by the barcode was appended to sequence names. VDJ-C germline reference sequences were created for each isotype by concatenating germline V (canonical IGHV1-10; ultralong IGHV1-7), D (canonical IGHD6-2*02; ultralong IGHD8-2*02), and J (IGHJ2-04) germline gene segments with constant regions for IgM, IgD, IgG, IgE, and IgA. Simultaneously, the 180,787 sequences were mapped to Ig reference sequences for both the canonical HC and ultralong HC using default mapper settings in Geneious Prime. Geneious mapper assembled 165,374 reads to two reference sequences to create two contigs.

We confirmed that each contig contained either the canonical HC or ultralong HC sequences using NCBI BLASTn (Zhang et al. 2000) and validated all amplicons within a contig to ensure they contained either the canonical HC or ultralong HC sequences by searching for the encoded YYC motif typical of canonical (IGHV1-10) or the 8-bp duplication encoding the TTVHQ motif found in ultralong CDR H3 sequences (IGHV1-7). For this study both the very short CDR H3s peaking at 8 aa and the dominant population peaking at 25 aa up until 39 aa are all considered canonical (Fig. 2). We extracted the nucleotide sequence from each contig between the encoded aspartic acid residue (D) at position 98 (DXATYYC) of the V segment and residue 24 of the C region. We then reassembled sequences to specific germline reference sequences for all five isotypes, creating five contigs each for canonical HC and ultralong HC amplicons. We extracted all sequences greater than 150 bp in length and validated isotypes within each contig using the constant region sequences, moving sequences to appropriate isotype files as necessary. Finally, we aligned canonical HC or ultralong HC sequences by isotype, and appended to the sequence name for analysis. We have provided our sequences in Excel files for each steer within the supplementary data. Each file contains two tabs of the raw, unfiltered output that was exported. The functional tab includes sequences that have not been aligned to the canonical HC or ultralong HC but meet the criteria indicated above. The unique tab contains the sequences that have been aligned to canonical HC or ultralong HC and was used for data analysis.

Steers share similar patterns of CDR H3 amino acid (aa) length. Within the canonical (CN, blue) CDR H3, there was a shoulder peak consisting of ≤ 11 aa, in addition to the main peak 12–39 aa in length. Ultralong (UL, orange) CDR H3s were defined as anything ≥ 40 aa in length

All statistical analyses were performed using SAS Analytics software (version 9.4) and JMP Pro 17 software (SAS Institute Cary, NC). A paired t-test was used to compare the proportions of unswitched and switched isotypes in each tissue across different CDR H3 lengths. Factorial analyses of variance (ANOVA) were used to evaluate the effects of CDR H3 length, immunoglobulin isotype, and tissue on response variables. Three-way, two-way, and one-way ANOVA models were fitted as appropriate to the experimental design with a significance of α = 0.05. When the overall ANOVA was significant (α = 0.05), pairwise comparisons among group means were performed using Tukey’s honestly significant difference (HSD) post hoc test.