An invariant VLR-like receptor refines the T-like cell framework in lamprey Authors: Ming Geng, Kang Li, Yating Zhu, Baohua Zou, Jun Li, Haokai Chen, Zhichao Fang, Xinying Jiao, Xue Zhai, Peng Hu, Xiumei Wei, Yue Pang, Jialong Yang Published: 2026 ABSTRACT AbstractJawless vertebrates employ variable lymphocyte receptors (VLRs) to mount adaptive immunity. Although six VLRs (VLRA–F) and T/B-like lymphocytes are identified in lampreys, it remains unclear how VLR diversity is organized at the cellular level and whether individual lymphocytes may integrate multiple VLRs. Using single-cell transcriptomics, we identify a germline-encoded, mature, non-diversified VLR-like receptor (VLR-like) in lamprey Lethenteron reissneri. VLR-like retains structural features of canonical VLRs, and is a GPI-anchored receptor expressed by T-like cells. We generate a VLR-like monoclonal antibody, enabling the identification of VLR-like+ lymphocytes across multiple tissues. Upon PHA stimulation, VLR-like+ lymphocytes exhibit robust ERK and JNK phosphorylation and proliferation, consistent with T-like cell features. In parallel, VLR-like+ cells fail to phagocytose, present antigen, or secrete soluble VLR-like, indicating the absence of B-like characteristics. Notably, VLR-like+ cells are distinct from VLRA+ lymphocytes but comprise both VLR-like+VLRC+ and VLR-like+VLRC- populations. Consistently, VLR-like interacts with VLRC but not VLRA. Together, we demonstrate that VLR-like defines a distinct T-like population with heterogeneous VLRC co-expression, and suggest a mode of receptor organization in which an invariant VLR-like coordinates a diversified VLR within an individual T-like cell, potentially contributing to adaptive immunity in lamprey.Subject terms: Evolutionary genetics, Molecular evolution Introduction Adaptive immunity represents a key evolutionary innovation that enabled vertebrates to survive in a microbially dynamic world1–4. In jawed vertebrates, adaptive immunity relies on somatically diversified antigen receptors like immunoglobulins (Ig) and T-cell receptors (TCR) and major histocompatibility complex molecules to mediate clonal recognition and establish immunological memory1,5–7. In contrast, jawless vertebrates such as lampreys and hagfish evolved a structurally distinct yet functionally analogous system centered on variable lymphocyte receptors (VLRs)8–17. Although these two systems arose independently, both achieve the same evolutionary goal: generating vast repertoires of antigen-specific lymphocytes, highlighting convergent solutions of adaptive immunity in vertebrates. In lampreys, three classical VLR types—VLRA, VLRB, and VLRC—have been extensively characterized and are often considered functionally analogous to T and B cell compartments of jawed vertebrates4,18–23. VLRB is expressed by B-like lymphocytes that secrete soluble multimeric VLR-antibodies, whereas VLRA and VLRC are associated with T-like lymphocyte populations showing transcriptional and functional similarities to αβ and γδ T cells, respectively19,20,24,25. These findings established a foundational framework for functional partitioning within the VLR-based system. While the VLRA–C paradigm has served as a conceptual cornerstone, accumulating genomic and transcriptomic evidence indicates that the system is more complex than originally envisioned. Recent studies have identified additional VLR genes—VLRD, VLRE, and VLRF—in lampreys, suggesting further expansion of the VLR repertoire26,27. Although the immunological functions of these newly identified VLRD–F remain unclear due to the absence of specific monoclonal antibodies (mAbs), available evidence shows that VLRD, VLRE, and VLRF transcripts are predominantly detected in the triple-negative (VLRA−/VLRB−/VLRC−) lymphocyte population, with low-level expression in VLRA+ and VLRC+ cells and no detectable expression in VLRB+ lymphocytes26,27. The identification of these VLRs and their expression patterns not only expands the repertoire of the VLR system but also suggests that its organization may be more complex than previously appreciated. Multiple VLR genes exhibit differential expression within a limited set of T-like cell populations, indicating that VLR expression may not strictly conform to a simple one receptor–one cell type model. In this context, a key question arises as to whether VLR-mediated immunity in lamprey relies solely on an individual receptor, or whether multiple VLRs may be co-expressed within the same cell and collectively contribute to immune responses. If such co-expression occurs, how these molecules are integrated within lymphocyte populations, and whether they participate in distinct modes of receptor organization, remain to be determined. In this study, we performed single-cell RNA sequencing of immune tissues in lamprey (Lethenteron reissneri) and identified a germline-encoded mature VLR-like receptor (VLR-like) that is distinct from canonical VLRs. Using a VLR-like-specific mAb, we examined the broad distribution of VLR-like+ lymphocytes across immune-related tissues. Functionally, VLR-like+ cells exhibit T-like cell features, including ERK and JNK activation and proliferation upon Phytohemagglutinin (PHA) stimulation, while lacking B-like characteristics. Moreover, VLR-like is partially co-expressed with VLRC, and VLR-like+ cells comprise both VLRC+ and VLRC− subsets. Notably, VLR-like can associate with VLRC, but not with VLRA, supporting the possibility that an invariant VLR-like molecule may be present in conjunction with a diversified VLRC within a subset of T-like cells. These observations suggest that VLR-like marks a distinct population within T-like lymphocytes, and that VLR-based immune responses may involve more complex receptor organization than previously appreciated. While the precise functional role of VLR-like in lamprey remains to be determined, our findings provide evidence for an additional layer of receptor organization within the VLR system. Results Single-cell transcriptomic landscape of VLRs in lamprey lymphocytes To characterize the cellular composition and VLR transcription profiles in L. reissneri, we re-analyzed the single-cell transcriptomic datasets (CNGB accession: CNP0005120) derived from blood, gill, and intestine according to the cell-type annotations reported in the previous study28. Our analysis identified several major cell populations across these tissues, including monocytes, granulocytes, red blood cells, epithelial cells, dendritic-like cells (DC-like cells), B-like cells, and T-like cells (Fig. 1a and Supplementary Fig. 1a, b). We then focused on T-like and B-like cell clusters to investigate their transcriptional features. After integrating the lymphoid clusters from these three tissues, we generated differential expression heatmaps (Fig. 1b, c and Supplementary Fig. 1c–e). Based on VLRs and marker genes reported in previous studies19,28–30, largely segregated expression modules and distinct lineage biases were observed between T-like and B-like cells. Specifically, T-like cells were enriched for VLRA, VLRC, VLRD, VLRE, and VLRF, whereas B-like cells preferentially expressed VLRB along with B-lineage-associated molecules such as BLNK and CD22 (Fig. 1b). In addition, several differentially expressed genes further supported transcriptional divergence between T-like and B-like lineages, including CD63, FABP1, VAV2, CREM, CNTFRα, and LRRFIP2 in T-like cells, and TNNC2, TLR2, PLD3, and PTPRJ in B-like cells (Fig. 1b). Notably, a putative novel VLR gene was found to specifically express in T-like population, and was tentatively named VLR-like (Fig. 1b). We next examined the distribution of VLR-like transcript across different annotated cell types according to the previously reported marker gene28. Consistent with other T-like VLRs, VLR-like expression was predominantly restricted to T-like cells (Fig. 1c and Supplementary Fig. 1c–e). Feature plot analysis across blood, gill, and intestine confirmed that VLRB was specifically enriched in B-like cells (Fig. 1d). In contrast, VLRA, VLRC, and VLR-like represented the predominant VLR types in T-like clusters, whereas VLRD, VLRE, and VLRF were restricted to a small subset of T-like cells (Fig. 1d). This transcriptional pattern was further supported by qPCR analysis (Supplementary Fig. 2). Meanwhile, we found that trace levels of VLRA and VLRC were detected in a minority of B-like cells from blood and intestine (Fig. 1d). Since further analysis within B-like cells revealed low co-expression between VLRB and each T-like VLR (Supplementary Fig. 3), these VLRA and VLRC signals within B-like cells most likely arise from background noise or ambient RNA contamination rather than genuine biological co-expression. Overall, these findings delineate the landscape and transcriptional features of VLRs in L. reissneri and identify a putative novel VLR-like gene specifically expressing in T-like cells. Single-cell transcriptomic landscape of lamprey VLRs. a Uniform manifold approximation and projection (UMAP) visualization of the integrated single-cell atlas, shown separately for blood, gill, and intestine, showing the major annotated cell populations across tissues. T-like cells and B-like cells form transcriptionally distinct clusters in all three tissues. T-like and B-like cell clusters are outlined with black borders. Numbers indicate cluster IDs. b Heatmap showing curated lineage-associated marker genes (upper block) together with the top-ranked differentially expressed genes (DEGs; lower block) from the T-like versus B-like comparison across blood, gill, and intestine. DEGs were identified using Seurat FindMarkers (min.pct = 0.25, logfc.threshold = 0.25) and ranked by average log2 fold change (avg_log2FC). Expression values were derived from log-normalized RNA assay data and are displayed as row-wise Z-score-scaled values. Cells were ordered first by cell type and then by tissue. c Heatmap showing the expression specificity of representative marker genes across the annotated cell populations in the integrated single-cell atlas. Expression is shown as row-wise Z-score-scaled average expression. d Tissue-resolved feature plots showing the distribution of germline plus mature transcripts of the indicated VLR genes (VLRA, VLRC, VLRD, VLRE, VLRF, VLR-like, and VLRB) across blood, gill, and intestine. Grey denotes low (or undetectable) expression levels, whereas purple denotes relatively high expression levels. Lamprey VLR-like is a germline-encoded mature VLR-like receptor Through single-cell transcriptomic profiling, we identified a putative novel VLR-like gene distinct from VLRA–F (NCBI GenBank accession no. WOJ44305.1). VLR-like encodes a mature peptide comprising a 37-residue signal peptide (SP), a 37-residue N-terminal LRR (LRRNT), a 19-residue LRR1, two 24-residue LRRVs, a 24-residue LRRVe, a 12-residue connecting peptide (CP), a 60-residue C-terminal LRR (LRRCT), and a 74-residue C-terminal stalk region (Supplementary Fig. 4). Multiple sequence alignment showed that the LRRNT and LRRCT regions—key domains for distinguishing VLR isotypes—exhibit low similarity to those of other known mature VLRs in L. reissneri (Fig. 2a), indicating that VLR-like does not belong to previously defined VLR types. Nevertheless, VLR-like retains the characteristic modular architecture of VLRs, including LRRNT, LRR1, multiple LRRV modules, CP, LRRCT, and a C-terminal stalk region (Fig. 2b). Structural modeling further predicted that VLR-like adopts a conserved solenoid fold typical of lamprey VLRs (Fig. 2c). Together, these results indicate that VLR-like represents a structurally conserved yet distinct VLR-like molecule. Genomic analysis revealed that the VLR-like gene consists of a single exon, in contrast to canonical VLR loci, suggesting that it is a germline-encoded, non-diversified receptor (Supplementary Fig. 5). To validate this, we performed PCR amplification of VLR-like using both genomic DNA and cDNA as templates, each yielding a single consistent band (Fig. 2d). Sequencing of multiple VLR-like cDNA clones further showed that LRR module sequences were highly uniform, with no detectable polymorphism (Supplementary Fig. 6a), indicating the absence of somatic diversification. In addition, transmembrane prediction for VLR-like indicated no canonical transmembrane domain (Supplementary Fig. 6b), whereas a glycosylphosphatidylinositol (GPI) anchoring site was predicted within the C-terminal stalk region (Supplementary Fig. 6c). To experimentally assess this, NIH/3T3 cells expressing VLR-like were treated with phosphatidylinositol-specific phospholipase C (PI-PLC). Similar to the GPI-anchored protein Thy-1 membrane glycoprotein (CD90)31,32, cell-surface VLR-like was released into the supernatant upon PI-PLC treatment (Fig. 2e), supporting that VLR-like is expressed as a GPI-anchored membrane protein. Notably, VLR-like is highly conserved across multiple lamprey species (Fig. 2f). Phylogenetic analysis further showed that VLR-like clusters with VLRD and VLRE, forming a branch that groups with VLRA, VLRC, and VLRF, whereas VLRB forms a separate outgroup (Fig. 2g). These relationships indicate that VLR-like is evolutionarily more closely related to VLRs associated with T-like lymphocyte populations than to VLRB. Collectively, these findings identify VLR-like as a germline-encoded, mature, and non-diversified VLR-like receptor with a highly conserved sequence. Sequence, structural, and phylogenetic analyses of lamprey VLR-like. a Multiple sequence alignment of mature VLR proteins from lamprey showing different regions. Conserved cysteine residues are highlighted in yellow. b Schematic domain organization of lamprey VLR proteins, including signal peptide (SP), LRRNT, LRR1, LRRV, connecting peptide (CP), LRRCT, and C-terminal stalk region. c Predicted 3D structure of VLR-like. The LRRNT is shown in purple, LRR1 in blue, LRRVs in green, CP in orange, and LRRCT in red. d, e The experiments were independently repeated three times with similar results. d PCR amplification of VLR-like from genomic DNA and cDNA templates. Arrows indicate the amplified products. e Phosphatidylinositol-specific phospholipase C (PI-PLC) cleavage assay of NIH/3T3 cells expressing VLR-like. Supernatant and cell lysate fractions were analyzed by immunoblotting before and after PI-PLC treatment. Mouse leukocytes and Thy-1 membrane glycoprotein (CD90) were included as controls. Molecular weight (kDa) markers adjacent to the target bands are indicated. f Multiple sequence alignment of VLR-like LRR regions across lamprey species. Conserved residues are shaded. g Phylogenetic tree of VLRs from different lamprey species. Bootstrap support values are indicated at internal branches. Generation of VLR-like mAb and identification of VLR-like+ lymphocytes To enable specific detection of VLR-like+ lymphocytes, we generated a mouse mAb against lamprey VLR-like. The full-length VLR-like coding sequence was cloned into the MIGR1 vector, and retroviral particles were produced using BOSC23 packaging cells (Fig. 3a). NIH/3T3 cells were subsequently infected with VLR-like retrovirus to establish stable VLR-like-expressing target cells for mouse immunization (Fig. 3b). Following hybridoma fusion, the supernatant from clone 1B2E8 specifically identified a distinct positive population among lymphocytes (Fig. 3c, d). This mAb recognized VLR-like-expressing NIH/3T3 cells but did not cross-react with cells expressing VLRA, VLRB, or VLRC (Fig. 3e). To further assess specificity in primary cells, we sorted VLR-like+ and VLR-like− lymphocytes and analyzed their gene expression profiles. VLR-like transcripts were exclusively detected in VLR-like+ cells, whereas VLRA was confined to the VLR-like− population (Fig. 3f). Notably, VLR-like+ lymphocytes showed expression of VLRC, whereas VLR-like− cells did not (Fig. 3f), indicating a potential cross-reactivity of this mAb with VLRC or a co-expression between VLR-like and VLRC. Then, we sought to confirm whether this VLR-like mAb cross-recognizes VLRC. A specific protein band detected in gill leukocyte lysates by this mAb (Fig. 3g) was confirmed to correspond to VLR-like, with no detectable peptides derived from VLRC or other canonical VLRs (Fig. 3h and Supplementary Fig. 7a). These results demonstrate that the mAb specifically recognizes VLR-like without detectable cross-reactivity. This antibody belongs to the IgG2a subclass (Supplementary Fig. 7b). IF analysis suggested that VLR-like may localize on the cell surface (Fig. 3i). Consistently, flow cytometry analysis across multiple tissues, including gill, intestine, liver, and peripheral blood, confirmed surface expression of VLR-like on lymphocytes (Fig. 3j). VLR-like+ cells accounted for approximately 10–20% of leukocytes in these immune-associated tissues (Fig. 3j, k), indicating their broad distribution within the lamprey immune system. Collectively, these results establish a high-specificity mAb as a reliable tool for phenotypic and functional characterization of VLR-like+ lymphocytes. Generation and validation of VLR-like mAb. a, b Flow-cytometric analysis of VLR-like-GFP expression in BOSC23 cells (a) and NIH/3T3 cells (b). c, d Detection of VLR-like on NIH/3T3 cells using isotype control or the anti-VLR-like mAb 1B2E8, shown as dot plots (c) and overlaid histograms (d). e Specificity analysis of 1B2E8 using NIH/3T3-GFP cells expressing lamprey VLR-like, VLRA, VLRB, or VLRC, respectively. f, g The experiments were independently repeated three times with similar results. f Semi-quantitative analysis of VLRs expression in sorted VLR-like⁺ and VLR-like− cell populations. DNA Ladder (in bp) adjacent to the amplification products is indicated. g, h Western blot assay using gill leukocytes showed the specificity of VLR-like mAb (g) and representative LC-MS/MS spectra of VLR-like-derived peptides (h). i Immunofluorescence detection of VLR-like expression. The graph is representative of at least three independent experiments. j Detection of VLR-like⁺ lymphocytes in lamprey gill, intestine, liver, and peripheral blood leukocytes (PBL) by flow cytometry. k Quantification of VLR-like⁺ lymphocytes in the indicated tissues, data are presented as mean ± SEM for n = 3 biologically independent lampreys per experiment. Individual data points are shown as dot plots. VLR-like+ cells are broadly distributed across immune-associated tissues To investigate the anatomical distribution of VLR-like+ lymphocytes, we examined major immune-associated organs—including the gill, retropharyngeal spine, kidney, spinal cord, intestine, liver, and skin—to obtain a systematic overview of VLR-like+ lymphocytes in lamprey. Structural features of L. reissneri tissues were first characterized by hematoxylin and eosin (H&E) staining (Fig. 4a–f), followed by immunofluorescence (IF) and immunohistochemistry (IHC) to localize VLR-like-expressing cells. VLR-like+ lymphocytes were widely distributed across multiple examined immune tissues, with tissue-specific distribution patterns, suggesting potential regional differences in immune context. The VLR-like signals were consistently detected in mucosal tissues located at environmental interfaces. The skin, serving as the first line of defense, displayed notable distribution of VLR-like+ lymphocytes (Fig. 4g). Similarly, abundant VLR-like+ cells were observed in mucosa-associated lymphoid structures—particularly adjacent to gill filaments (Fig. 4h) and within the intestinal typhlosole (Fig. 4i), which was further confirmed by IHC (Fig. 4j, k). Beyond these mucosal sites, VLR-like+ cells were also detected in internal organs associated with immune regulation and hematopoiesis. The liver harbored VLR-like+ cells scattered within the parenchyma and perivascular regions (Fig. 4l); while the kidney, a known hematopoietic site in lampreys, contained dense clusters of VLR-like+ lymphocytes (Fig. 4m). Together, these findings provide a systematic anatomical and histological map of VLR-like+ lymphocytes in lamprey, highlighting their broad tissue distribution and potential immunological function. Localization of VLR-like+ cells in lamprey. Cross-sections of lamprey tissues were processed for H&E staining, immunohistochemistry (IHC), or immunofluorescence (IF). a Anterior transverse section showing immune-related organs, including the epipharyngeal ridge (ep), spinal cord (sc), notochord (n), and gill. b Section at the level of the anterior gut showing the kidney (k), intestine (i), and typhlosole (ty). c–f Higher-magnification H&E-stained sections of the gill (c), kidney (d), intestine and typhlosole (e), and liver (f). g–i, l, m IF analysis showing the distribution of VLR-like⁺ cells in the skin (g), gill (h), intestinal typhlosole (i), liver (l), and kidney (m). Arrows indicate VLR-like⁺ cells. j, k IHC staining showing VLR-like⁺ cells in the gill ( j ) and intestinal typhlosole (k). Arrows indicate VLR-like⁺ cells. Scale bars for (a, b), 1 mm; (c–e, j), 100 μm; (f, h, k), 50 μm; (g, i, l, m), 10 μm. The graphs are representative of at least three independent experiments. VLR-like+ lymphocytes respond to PHA stimulation, consistent with T-like activation features PHA is a well-established T-cell mitogen that activates T cells through receptor cross-linking and downstream signaling activation33. PHA can also activate VLRA+ T-like cells in lamprey19. To assess whether VLR-like+ lymphocytes exhibit similar activation properties to T cells in jawed vertebrates and VLRA+ T-like cells in jawless vertebrates, lamprey leukocytes were stimulated with PHA in vitro. This treatment elicited a robust, time-dependent activation of canonical T-cell signaling pathways34–36 in VLR-like+ lymphocytes (Fig. 5a). Flow cytometric analysis revealed a marked increase in the phosphorylation of ERK1/2 and JNK, with p-ERK1/2⁺ cells rising from 39.1 to 74.9% at 9 h, and p-JNK+ cells increasing from 45.5 to 68.0% over the same period (Fig. 5a). This coordinated phosphorylation pattern reflects the dynamic intracellular signaling cascade characteristic of T-cell activation. Consistent with these in vitro findings, in vivo PHA injection induced a pronounced expansion of the VLR-like+ lymphocyte population—from approximately 15% in controls to 39% at day 10 post-injection (Fig. 5b). 5-Ethynyl-2’-deoxyuridine (EdU) incorporation assays revealed a striking accumulation of proliferating VLR-like+EdU+ cells following PHA treatment, whereas such cells were rare in controls (Fig. 5c). Together, these results demonstrate that VLR-like+ lymphocytes undergo activation and proliferation in response to PHA stimulation, similar to those of VLRA+ T-like cells. These findings support the classification of VLR-like+ lymphocytes as a T-like population with proliferative and signaling responses to mitogenic stimulation. Activation of VLR-like⁺ lymphocytes in lamprey following PHA stimulation. a Flow cytometric analysis of phosphorylated ERK1/2 (p-ERK1/2) and JNK (p-JNK) in gill VLR-like⁺ lymphocytes after phytohemagglutinin (PHA) stimulation for 3, 6, and 9 h. b VLR-like⁺ cells in intestinal lymphocytes 10 days after PHA injection. c Immunofluorescence detection of proliferating 5-Ethynyl-2’-deoxyuridine (EdU)⁺VLR-like⁺ cells in typhlosole 10 days after PHA injection. Scale bars: 100 μm. All experiments were performed in three independent biological replicates. VLR-like+ lymphocytes lack canonical B-cell functions B cells are defined by several hallmark effector functions, including phagocytosis, antigen presentation, and soluble antibody secretion37–41. In lamprey, VLRB+ lymphocytes can also secrete VLR antibody19. To determine whether VLR-like+ lymphocytes share any of these B-like characteristics, we performed several analyses. Phagocytosis assays using fluorescent microspheres revealed that more than 20% of VLR-like− cells (likely macrophages or VLRB⁺ lymphocytes) could phagocytose microspheres, whereas almost none of VLR-like+ lymphocytes exhibited phagocytic activity (Fig. 6a, b). In addition, following in vitro incubation of leukocytes with inactivated Aeromonas hydrophila, proteolytic activity associated with antigen processing was detected in VLR-like− lymphocytes, but not in VLR-like+ lymphocytes (Fig. 6c, d). These results do not support detectable phagocytic or antigen-processing activity in VLR-like+ lymphocytes in lamprey. Then, we examined whether VLR-like+ lymphocytes secrete soluble antibody-like proteins like VLRB+ lymphocytes19,20. Immunoblot analysis detected the VLR-like in leukocyte lysates (Fig. 6e), but not in serum obtained from lampreys with or without inactivated A. hydrophila immunization (Fig. 6e). Together, these results suggest that VLR-like+ lymphocytes do not exhibit the canonical features of B cells—phagocytosis, antigen presentation, and antibody secretion—and are therefore functionally distinct from B-like cells. Together with their proliferative and signaling responses to PHA stimulation, these observations are consistent with the classification of VLR-like+ lymphocytes as a T-like population. Lamprey VLR-like⁺ lymphocytes lack B-like cell functional characteristics. a Flow cytometric assessment of fluorescent microsphere (FM) uptake by intestinal VLR-like⁺ and VLR-like− lymphocytes following incubation for 4 h. b Statistic figures showing the phagocytosis percentage in VLR-like+ and VLR-like− cells, data are presented as mean ± SEM for n = 3 biologically independent lampreys per experiment, p = 0.0033. c Flow cytometric analysis of DyeQuenched casein (DQ-casein) fluorescence in VLR-like⁺ and VLR-like− lymphocytes following stimulation of intestinal leukocytes with inactivated Aeromonas hydrophila for 8 h. d Statistic figures showing the DQ-casein+ cell percentage in VLR-like+ and VLR-like− populations, data are presented as mean ± SEM for n = 3 biologically independent lampreys per experiment, p = 0.0018. VLR-like− cells from the same animals served as an internal negative control, eliminating the need for an external control group (a–d). e Western blot (immunoblot, IB) analysis of VLR-like expression in serum from lamprey after 14 days of immunization with inactivated A. hydrophila, with gill leukocytes serving as a positive control. Arrows indicate the VLR-like protein. Coomassie blue staining demonstrated the presence of proteins in the lamprey serum, as shown by SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis). All experiments were performed in three independent biological replicates. The p-value is determined by a two-tailed Student’s t-test. **p < 0.01. The corresponding data points are shown as dot plots. VLR-like+ represent a T-like population distinct from VLRA+ cells but include both VLRC+ and VLRC- lymphocytes Given that VLR-like+ lymphocytes exhibit clear T-like features, we next sought to determine how VLR-like+ population is positioned relative to the previously characterized VLRA+ and VLRC+ T-like populations. Using the same retroviral expression strategy applied to VLR-like, we generated VLRA- and VLRC-expressing retroviruses (Supplementary Fig. 8a, b) and established NIH/3T3 cells stably expressing each receptor (Supplementary Fig. 8c, d). These cells were used to immunize mice, followed by hybridoma fusion, yielding mAbs that identified distinct populations among lamprey lymphocytes (Supplementary Fig. 8e–h). To confirm the specificity of these mAbs, we sorted VLRA+ or VLRC+ lymphocytes and their corresponding negative cells for gene expression analysis. In sorted VLRA+ and VLRA− lymphocytes, VLRA was exclusively expressed in VLRA+ cells, whereas VLRC and VLR-like expression was detected in VLRA− but not VLRA+ population (Fig. 7a), indicating the high specificity of VLRA mAb. Similarly, VLRC transcript was restricted to VLRC+ lymphocytes, whereas VLRA was detected in VLRC− cells (Fig. 7b), supporting the specificity of VLRC mAb. However, we found that both VLRC+ and VLRC− lymphocytes expressed VLR-like (Fig. 7b). Together with our earlier observation that VLRC transcripts are enriched in VLR-like+ cells (Fig. 3f), these results suggest that VLR-like and VLRC may be co-expressed in a subset of lymphocytes. Co-expression of VLR-like with VLRC in lamprey. a, b Semi-quantitative analysis of VLRs expression in sorted VLRA⁺ and VLRA− (a) and VLRC⁺ and VLRC− (b) cell populations. DNA Ladder (in bp) adjacent to the amplification products is indicated. c–f FACS plots showing the double staining of VLRC and VLRA (c), VLR-like and VLRA (d), VLR-like and VLRC (e, f) in gill leukocytes of lamprey. g Immunofluorescence assay showing the double staining of VLR-like and VLRC in lamprey typhlosole. Scale bars: 50 μm. h, i Overlay histograms showing p-ERK1/2 (h) and p-JNK (i) in VLR-like⁺VLRC⁺ and VLR-like⁺VLRC− cell populations under control conditions or after 6 h PHA stimulation. ( j, k) Co‑immunoprecipitation (Co‑IP) analysis was performed to examine the interaction between VLR‑like‑HA and VLRC-Flag ( j ) or VLRA-Flag (k) in cotransfected cells. Lysates were immunoprecipitated (IP) with anti-Flag Ab-conjugated agarose beads and then immunoblotted (IB) with anti‑HA or anti‑Flag antibodies; Input controls are shown below. Molecular weight (kDa) markers adjacent to the target bands are indicated. l Schematic showing the subsets of T-like cells in lamprey. m Schematic showing the comparison of receptor organization strategy between lamprey T-like cells and mammalian iNKT cells and MAIT cells. TCR, T cell receptor; iNKT, invariant natural killer T cell; MAIT, mucosal-associated invariant T cell. All experiments were performed in three independent biological replicates. Next, we further examined this possibility using flow cytometry. As previously observed in sea lampreys, VLRA+ and VLRC+ lymphocytes in L. reissneri form two distinct, non-overlapping populations (Fig. 7c). Notably, VLR-like+ cells also form a population largely non-overlapping with VLRA+ lymphocytes (Fig. 7d and Supplementary Fig. 9a). More importantly, we found that VLR-like is co-expressed with VLRC in a fraction of lymphocytes (Fig. 7e). VLR-like+ cells encompass a broader population than VLRC+ cells, as most of VLRC+ lymphocytes are contained within the VLR-like+ population, while the VLR-like+ compartment also includes VLR-like+VLRC− cells (Fig. 7e, f and Supplementary Fig. 9b). The co-expression of VLRC within a subset of VLR-like+ cells was further validated by IF (Fig. 7g). Considering VLR-like+ compartment also includes VLRC+ lymphocytes, we sought to know whether the T-like cell properties upon PHA stimulation are just contributed by the VLRC+ cells. The phosphorylation of ERK1/2 and JNK was obviously enhanced upon PHA stimulation in both VLR-like+VLRC+ and VLR-like+VLRC− cells (Fig. 7h, i), indicating that VLR-like+ lymphocytes can respond to stimulation irrespective of VLRC co-expression. Finally, we investigated whether VLR-like can interact with VLRC. The Co-IP assay showed that VLR-like interacts with VLRC (Fig. 7j), but not with VLRA (Fig. 7k), suggesting a potential association between these receptors. Collectively, these findings indicate that VLR-like+ lymphocytes represent a T-like population distinct from VLRA+ cells and comprise both VLRC+ and VLRC− subpopulations. The observed co-expression and association between VLR-like and VLRC further suggest a potential mode of receptor organization that warrants further investigation. Discussion In this study, we identified and characterized a germline-encoded, mature, non-diversified VLR-like receptor (VLR-like) in the lamprey L. reissneri, revealing an additional component of the VLR-based immune system that differs from canonical diversified VLRs. Like the six previously characterized VLRs (VLRA–F)19–21,26,27,42, VLR-like adopts the canonical modular architecture of lamprey VLRs; however, it exhibits pronounced sequence divergence from VLRA–F in key regions, including LRRNT, LRR1, and LRRCT. Phylogenetic analysis further supports that VLR-like does not group within the known diversified VLR clades, but instead forms a separate branch, consistent with its distinct molecular features. Single-cell RNA-seq further reinforces the distinctiveness of VLR-like. VLR-like-expressing cells display transcriptional features consistent with T-like population, while lacking characteristics of VLRB+ B-like cells. VLR-like shows expression comparable to VLRA and VLRC, supporting its role as a major T-like population. Notably, unlike canonical VLRs that undergo somatic diversification, VLR-like is encoded in the germline and exhibits an invariant sequence, indicating that it does not participate in the classical diversification process that underlies antigen receptor diversity in jawless vertebrates. This distinction raises the possibility that VLR-based immunity may not rely exclusively on diversified receptors, but may also incorporate invariant receptor components. Defining the tissue distribution patterns of VLR-expressing cells is fundamental to understanding their physiological and immunological roles. Using specific mAbs, VLRA+, VLRB+, and VLRC+ lymphocytes have been shown to distribute across major immune-relevant organs in sea lamprey, with VLRC+ cells exhibiting preferential enrichment at epithelial barrier sites such as the skin19,20. In contrast, the tissue distribution of more recently identified VLRD–F-expressing cells remains less well defined, largely due to the lack of specific mAbs, although evidence from RNA-based approaches suggests several expression patterns26,27. In this study, using a VLR-like-specific mAb, we found that VLR-like+ cells are broadly distributed across multiple immune-associated tissues of L. reissneri, including the gill, intestine, liver, peripheral blood, skin, and kidney. This widespread distribution—spanning both mucosal and systemic sites—resembles the general tissue pattern of T cells in jawed vertebrates43–46 and is consistent with their proposed T-like immune role. Integrating our findings with prior knowledge of VLRA–F localization, it becomes clear that the T-like compartment of jawless vertebrates exhibits an intricate tissue organization. Distinct VLRs occupy characteristic tissue niches, likely reflecting a degree of functional specialization within this ancient immune system. These spatial patterns provide a perspective for understanding how cellular distribution may support functional diversification within the VLR-based immune system. Clarifying the immunological functions of different VLRs is essential for deciphering the architecture of adaptive immunity in jawless vertebrates. Although six distinct VLRs have been identified, functional dissection has been limited by the scarcity of specific mAbs; thus far, only VLRA, VLRB, and VLRC have been characterized in depth19,20. In sea lamprey, Bacillus anthracis spore antigen can induce proliferation of VLRA+, VLRB+, and VLRC+ lymphocytes19,20. VLRA+ and VLRC+ lymphocytes behave as canonical T-like populations and mount strong proliferative responses to the T-cell mitogen PHA19. Although their global transcriptional programs are broadly similar, VLRA+ and VLRC+ lymphocytes exhibit distinct functional biases: VLRA+ lymphocytes preferentially express TCF1 and CTLA4, whereas VLRC+ lymphocytes are enriched for SOX13 and LFA1, and can be induced to proliferate and upregulate IL-16 following poly I:C stimulation20. By contrast, VLRB binds native antigens and differentiates into antibody-secreting cells, thereby mediating humoral immunity in jawless vertebrates19. Against this functional backdrop, our study extends the functional characterization of VLR-based lymphocytes. We show that VLR-like+ lymphocytes, similar to VLRA+ and VLRC+ populations, undergo strong proliferation following PHA stimulation and rapidly activate ERK and JNK signaling pathways, consistent with a T-like activation program. At the same time, VLR-like+ cells lack the antibody-secreting capacity of VLRB+ B-like cells and do not display detectable phagocytic or antigen-presentation activities, thereby excluding a B-like identity. Together, these functional features support the classification of VLR-like+ lymphocytes as a T-like population and provide additional insight into the functional diversification of VLR-based immunity in jawless vertebrates. In jawless vertebrates, the VLR-based immune system has often been viewed in a manner consistent with a one VLR–one cell type pattern, in which distinct VLRs demarcate discrete lymphocyte populations, as exemplified by VLRA+ and VLRC+ T-like cells20,22. However, the expanding repertoire of VLR genes and the limited functional characterization of many of these receptors suggest that this simplified model may not fully capture the organizational complexity of the VLR system. In the present study, multiple lines of evidence consistently point to a more intricate organization of VLRs within T-like lymphocytes. At the transcriptional level, analysis of sorted populations revealed that VLR-like transcripts are detectable in both VLRC+ and VLRC− lymphocytes, whereas VLRC is also expressed in VLR-like+ compartment. At the protein level, flow cytometry and IF analyses further demonstrate that a substantial fraction of lymphocytes co-express VLR-like and VLRC. Importantly, biochemical assays show that VLR-like can associate with VLRC, but not with VLRA. Together, these findings support coordinated expression or interaction between distinct VLR molecules within the same cell, rather than strictly segregated receptor usage. In parallel, these findings also refine the cellular organization of the T-like compartment, in which VLR-like+ cells define a population distinct from VLRA+ lymphocytes while encompassing both VLRC+ and VLRC− subsets (Fig. 7l). Notably, VLR-like differs from previously characterized VLRs, and it is germline-encoded and non-diversified. This feature suggests that VLR-like may not function as a conventional antigen receptor in the same manner as diversified VLRs. Instead, its association with VLRC raises the possibility that VLR-like may cooperate with diversified VLRs within individual cells, potentially contributing to receptor organization or functional modulation of T-like lymphocytes. Based on these observations, we propose a model in which an invariant VLR-like molecule is coordinately present with a diversified VLR (such as VLRC) within an individual T-like lymphocyte (Fig. 7m). This mode of receptor organization introduces an additional layer of complexity to the VLR system, extending beyond receptor diversification alone. Conceptually, such an arrangement bears resemblance to invariant or semi-invariant receptor usage observed in certain T-cell populations of jawed vertebrates, including invariant natural killer T (iNKT) cells and mucosal-associated invariant T (MAIT) cells, in which an invariant TCRα chain is paired with a restricted or semi-invariant TCRβ chain to mediate antigen recognition (Fig. 7m)47–49. Importantly, this comparison is intended as a conceptual parallel rather than a direct functional equivalence. Further studies will be required to determine whether VLR-like directly participates in antigen recognition, contributes to receptor signaling, or primarily functions in receptor organization. Nevertheless, our findings provide initial evidence that VLR-based immunity in jawless vertebrates may involve coordinated receptor systems within individual cells, thereby offering a new perspective on how immune recognition can be organized in the absence of canonical TCR and co-receptor architectures. Methods Experimental animals Lampreys (Lethenteron reissneri) were collected from the Yalu River in Liaoning Province, China, with body lengths of 10–15 cm and weights of 6–8 g. The larval stage lasts 2.5 to 3 years before metamorphosis. The lampreys were maintained in oxygenated circulating water at 16 °C. BALB/c mice were raised in Minhang Laboratory Animal Center of East China Normal University. All experimental procedures were conducted according to the Guide for the Care and Use of Laboratory Animals of the Ministry of Science and Technology of China. This study was approved by the East China Normal University Experimental Animal Ethics Committee and Liaoning Normal University Experimental Animal Ethics Committee with approval numbers of f + m 20240703 and LL2023088. scRNA-seq library preparation and sequencing Single-cell RNA-seq libraries from gill, intestine, and blood samples of lamprey were prepared using the DNBelab C Series Single-Cell Library Prep Set (MGI, 1000021082) according to the protocol described by ref. 28. In brief, single-cell suspensions were encapsulated into droplets, followed by emulsion disruption, bead collection, reverse transcription, and cDNA amplification to generate sequencing libraries. The libraries were sequenced on the DIPSEQ T1 platform at BGI-Qingdao (Qingdao, China). Sequencing was performed using the following configuration: Read 1, 30 bp (comprising 10-bp cell barcode 1, 10-bp cell barcode 2, and 10-bp unique molecular identifier [UMI]); Read 2, 100 bp for transcript sequences, with an additional 10-bp sample index. Single-cell sequencing data processing and cell clustering Previously published lamprey single-cell RNA-seq datasets generated by Pang et al. were used for further analysis in this study28. Raw reads generated from the DIPSEQ T1 platform were processed using the DNBelab C Series high-throughput scRNA-seq software (MGI) to decode cell barcodes and align transcripts to the lamprey reference genome50. To enhance annotation accuracy, existing gene models, particularly exon boundaries, were manually curated and updated. The resulting expression matrix was subjected to doublet detection using DoubletFinder, and approximately 3% of cells with the highest similarity scores were removed. Cells with fewer than 200 detected genes per library were also excluded. The filtered data were subsequently analyzed in Seurat v5.2.1. Normalization, scaling, and identification of highly variable features were performed using the SCTransform function. For datasets derived from the same tissue, integration was achieved using the Harmony algorithm following principal component analysis (RunPCA), and the integrated data were visualized in uniform manifold approximation and projection (UMAP) space. Cell clustering was conducted using the FindClusters function at a resolution of 0.5, based on the nearest-neighbor graph derived from Harmony reduction. Differentially expressed genes (DEGs) across clusters were identified using the FindAllMarkers function (min.pct = 0.25, logfc.threshold = 0.25). Cluster annotations were then assigned by comparing cluster-specific DEGs with established marker gene expression profiles. Heatmaps were generated using pheatmap v1.0.13 and ComplexHeatmap v2.18.0, and scatter plots were generated using ggplot2 v4.0.0. Co-expression analysis of VLR genes in B-like cells. B-like cells were extracted from the blood, intestine, and gill single-cell transcriptomic datasets. Co-expression between VLRB and VLRA/C/D/E/F/VLR-like was determined from raw UMI counts, with double-positive cells defined by a count threshold of ≥2 for both genes. Double-positive fractions were summarized by heatmaps, and single-cell co-expression patterns were visualized by scatter plots. The accession numbers of the marker genes used in the single-cell sequencing analysis are provided in Supplementary Table 1. Identification of VLRs sequences in lampreys Previously reported VLR sequences (VLRA, VLRB, and VLRC) were used as queries for TBLASTN searches51 against the lamprey genome50. Sequences corresponding to VLRD, VLRE, and VLRF in the lamprey were identified based on homology to previously reported sea lamprey VLR sequences. A similar homology-based strategy was employed to annotate VLRD, VLRE, VLRF, and VLR-like in other lamprey species. Detailed sequence information is provided in Supplementary Table 2. Sequence and phylogenetic analysis Leucine-rich repeat (LRR) domains were identified using SMART and manual inspection. Signal peptides were predicted using SignalP (https://services.healthtech.dtu.dk/services/SignalP-6.0/). Transmembrane domains were predicted using TMHMM. Multiple sequence alignments were performed with Clustal X v2.1. Three-dimensional structure predictions were generated using AlphaFold252 and visualized with PyMOL v2.0. VLR-like locus visualization was performed using gggenes v0.6.0. GPI anchor sites were predicted using the PredGPI algorithm (https://mendel.imp.ac.at/gpi/cgi-bin/gpi_pred.cgi). Phylogenetic trees were reconstructed using the neighbor-joining method in MEGA v1153 software. The tree topology was assessed by bootstrap analysis with 1000 replicates, and evolutionary distances were computed using the JTT matrix-based method. Leukocyte isolation Leukocytes were isolated from lamprey tissues and peripheral blood by Percoll (GE Healthcare, 10246712) density gradient centrifugation. Peripheral blood was collected from the heart into pre-cooled anticoagulant solution (8.17 g/L NaCl, 19.8 g/L glucose, 5.46 g/L citric acid, 8.82 g/L sodium citrate, 3.72 g/L EDTA; pH 6.6). Tissue samples were gently homogenized in Leibovitz’s L-15 medium (Gibco, 41300039) and filtered through a 200-mesh sieve to obtain single-cell suspensions. A discontinuous Percoll gradient was prepared (4 mL each of 52 and 34% Percoll), and then 3 mL of cell suspension was layered on top. Samples were centrifuged at 500 × g for 30 min at 16 °C with minimal acceleration and deceleration. Leukocytes collected from the interface between 34 and 52% Percoll were harvested for subsequent experiments. Leukocyte stimulation Gill leukocytes (1 × 106 cells) were cultured in L-15 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. Cells were either stimulated with 2 µg/mL PHA (Sigma, L2769) or left unstimulated as controls. Following incubation at 16 °C for 3, 6, and 9 h, leukocytes were harvested for assays. Bacterial infection Aeromonas hydrophila was cultured overnight in Luria-Bertani (LB) medium at 37 °C. Bacteria in the exponential growth phase were collected, washed three times with sterile phosphate-buffered saline (PBS) at 2000 × g for 5 min, and then resuspended in 4% paraformaldehyde (PFA) for inactivation. The suspension was fixed overnight at 4 °C and washed three times with sterile PBS. Lampreys were intraperitoneally (i.p.) injected with 100 µL of the inactivated bacterial suspension at a concentration of 1 × 108 CFU/mL, while control animals received an equal volume of sterile PBS. Gill leukocytes and serum were collected at designated time points for subsequent assays. Retrovirus packaging and cell infection The full-length coding sequences of lamprey VLRA, VLRC, and VLR-like were synthesized by Sangon Biotech (China) and cloned into the pEGFP-C1-MIGR1 vector. For retrovirus production, 5 μg of pCL-Eco and 10 μg of pEGFP-MIGR1-VLRA, -VLRC, or -VLR-like were dissolved in HEPES-buffered saline containing 0.1 mM CaCl2 and co-transfected into 2 × 106 BOSC23 cells. After 8 h of transfection, the medium was replaced with fresh DMEM (Gibco, 11995065) supplemented with 10% FBS and 1% penicillin/streptomycin. Retrovirus-containing supernatants were harvested 48 h post-transfection. To infect target cells, 600 μL of retrovirus supplemented with 5 μg/mL polybrene (Sigma, H9268) was added to 2 × 105 NIH/3T3 cells. After incubation for 6 h, the virus was diluted by adding 1.5 mL fresh DMEM, and the medium was fully replaced with 6 mL of fresh DMEM the following day. Infected NIH/3T3 cells were harvested 48 h post-infection for subsequent mouse immunization experiments. Development of mouse anti-lamprey VLRA, VLRC, and VLR-like mAbs To generate hybridomas producing mouse mAbs against lamprey VLRA, VLRC, and VLR-like, mice were intraperitoneally immunized with 2 × 107 VLRA-, VLRC-, or VLR-like-expressing NIH/3T3 cells twice at a two-week interval, followed by two weekly booster immunizations. Three days after the final booster, spleen cells were harvested and fused with SP2/0 myeloma cells using 50% polyethylene glycol 4000 (Sigma, 95904). The fusion cells were cultured in RPMI 1640 medium (Gibco, 11875093) supplemented with 10% FBS and 1% HAT (Gibco, 21060017) for 7–10 days. Hybridoma supernatants were screened for anti-VLRA, anti-VLRC, and anti-VLR-like activity using flow cytometry, and positive clones were subsequently subcloned via limiting dilution to establish monoclonal hybridoma lines. PI-PLC cell digestion assay 3 × 106 mouse spleen leukocytes or VLR-like-overexpressing NIH3T3 cells were incubated in PBS with or without 0.5 U of PI-PLC (Sigma, P5542) at 37 °C with gentle shaking for 5 h. After incubation, the supernatant and cells were collected separately. The cells were washed twice with PBS and then lysed with NP-40 lysis buffer for 30 min. After centrifugation at 12,000 × g for 10 min, the supernatant (cell lysate) was collected. Both the digestion supernatant and the cell lysate were mixed with SDS-PAGE loading buffer, boiled for 10 min, and analyzed by western blot. Flow cytometry and cell sorting To screen for hybridomas secreting anti-lamprey VLRA, VLRC, and VLR-like mAbs and to quantify VLRA-, VLRC-, and VLR-like-expressing cells in lamprey tissues and peripheral blood, hybridoma supernatants were incubated with 1 × 106 leukocytes, followed by staining with Alexa Fluor 647-conjugated goat anti-mouse IgG H&L (Abcam, ab150115, 1:2000). To assess the specificity of VLR-like mAb and evaluate potential cross-reactivity with VLRA/B/C isoforms, VLR-like hybridoma supernatants were separately incubated with 1 × 106 NIH/3T3 cells expressing VLR-like, VLRA, VLRB, or VLRC, followed by staining with Alexa Fluor 647-conjugated goat anti-mouse IgG H&L. To assess the co-expression of lamprey VLRA/VLRC, VLRA/VLR-like, VLRC/VLR-like by flow cytometry, cells were stained with distinct antibody combinations as follows: for VLRA/VLRC, cells were sequentially stained with VLRC mAb, followed by Alexa Fluor 488-conjugated goat anti-mouse IgG H&L (Abcam, ab150113, 1:2000), then with biotin-conjugated VLRA mAb (1 mg/mL, 1:400), and finally with APC-streptavidin (BioLegend, 405207, 1:400). For VLRA/VLR-like, cells were stained with VLRA mAb, followed by Alexa Fluor 647-conjugated goat anti-mouse IgG H&L, and finally with FITC-conjugated VLR-like mAb (1 mg/mL, 1:400). For VLRC/VLR-like, cells were stained with VLRC mAb, followed by Alexa Fluor 647-conjugated goat anti-mouse IgG H&L, and finally with FITC-conjugated VLR-like mAb. All incubations were performed on ice for 30 min. After each staining, cells were centrifuged at 800 × g for 3 min at 4 °C and washed twice with FACS buffer (PBS containing 2% FBS). For intracellular staining of signaling pathways, gill leukocytes were first stained for the target cell population according to the method described above, sequentially. Cells were then fixed and permeabilized on ice for 2 h using fixation/permeabilization solution (eBioscience, 00-5123-43). Subsequently, cells were incubated with APC-conjugated antibodies against phosphorylated Erk1/2 Thr202/Tyr204 (BioLegend, 935703, 1:200) or phosphorylated JNK Thr183/Tyr185 (Cell Signaling Technology, 9257S, 1:200). After incubation, cells were washed and resuspended in BD Perm/Wash buffer (eBioscience, 00-8333-56) prior to analysis. All stained cells were analyzed using a BD FACS Canto II flow cytometer. For cell sorting, lamprey leukocytes were stained as described above, and both positive and negative cells for VLRA, VLRC, and VLR-like were sorted with a BD FACS Aria II flow cytometer. Quantitative real-time RT-PCR (qPCR) and PCR Total RNA was extracted from the gills, intestines, or peripheral blood of lamprey using TRIzol reagent (Invitrogen, 15596018CN) following the manufacturer’s instructions. RNA was reverse transcribed into cDNA using the Plus All-in-one 1st Strand cDNA Synthesis SuperMix (Novoprotein, E047-01A). qPCR reactions were carried out using 2 × NovoStart SYBR qPCR SuperMix Plus (Novoprotein, E096-01A) on a Bio-Rad CFX Duet RT-PCR instrument, with diluted cDNA serving as the template. Relative mRNA expression levels of target genes were calculated using the 2−ΔΔCT method, with GAPDH as an internal control. PCR amplification of the VLR-like gene was performed using cDNA and genomic DNA templates. Amplification products were analyzed by agarose gel electrophoresis and confirmed by size comparison with DNA markers. Uncropped scans of the key blots are provided in the Source Data file. The primers used for qPCR and PCR were synthesized by Sangon Biotech (China), and the sequences are provided in Supplementary Table 3. Antibody sub-typing The isotypes and subclasses of lamprey VLR-like mAb were determined using a double-antibody sandwich ELISA according to the manufacturer’s instructions for the Mouse Monoclonal Antibody Isotyping ELISA Kit (Biodragon, KLB4011216). Briefly, microplates were first coated with a capture antibody recognizing the common Fc region of mouse Ig. Supernatants from VLR-like hybridoma cultures were then added to the wells. Subsequently, horseradish peroxidase-conjugated detection antibodies specific for individual mouse isotypes and subclasses were applied in separate reactions. Each incubation step was performed at 37 °C for 30 min, with five washes using PBST between steps to remove unbound components. Colorimetric detection was carried out using a TMB substrate system, and the reaction was stopped with dilute sulfuric acid at 37 °C for 20 min. Absorbance at 450 nm was measured with a microplate reader to determine the antibody class or subclass. Western Blot assay Isolated tissues or cells were lysed in NP-40 lysis buffer on ice for 30 min, followed by centrifugation at 12,000 × g for 10 min. The supernatant was collected, mixed with 5× protein loading buffer, and then heated in a boiling water bath for 10 min. Proteins were separated by 12% SDS-PAGE and transferred onto nitrocellulose (NC) membranes. Membranes were blocked with 4% non-fat milk in PBST at room temperature for 1 h, followed by incubation with VLR-like mAb or CD90 (Beyotime, AF1636) overnight at 4 °C. Subsequently, membranes were incubated with Alexa Fluor 680-conjugated goat anti-mouse IgG H&L (Abcam, ab175775, 1:10,000) or Alexa Fluor 800-conjugated goat anti-rabbit IgG H&L (Abcam, ab150115, 1:30,000) at room temperature for 1 h in the dark. After each incubation step, membranes were washed three times with PBST and imaged using the Odyssey CLx Imaging System. Uncropped scans of the key blots are provided in the Source Data file. Immunofluorescence To examine VLR-like localization, gill leukocytes were collected by centrifugation onto glass slides using a Cytospin device (Thermo). Cells were fixed in methanol for 5 min, followed by blocking with 1% bovine serum albumin (BSA) at 37 °C for 1 h. Samples were then incubated with VLR-like mAb as the primary antibody, followed by incubation with Alexa Fluor 488-conjugated goat anti-mouse IgG H&L (Abcam, ab150113, 1:800) in the dark. For tissue-level analysis, lamprey tissues were fixed in 4% PFA at room temperature for 24 h, subjected to gradient dehydration and clearing, and embedded in paraffin. Sections were cut at an appropriate thickness and mounted onto glass slides. After baking at 60 °C for 2 h, sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Heat-induced epitope retrieval (HIER) was performed using citric acid buffer (pH 6.0). Sections were then incubated with primary and secondary antibodies as described above. To detect the co-expression of VLRC and VLR-like, sections were subjected to HIER and blocking. Subsequently, the sections were incubated with VLRC mAb, followed by an Alexa Fluor 594-conjugated goat anti-mouse IgG (H&L) (Abcam, ab150116; 1:800), and finally with a FITC-conjugated VLR-like mAb (1:400). Each antibody incubation was performed at room temperature for 1 h. After each incubation, samples were washed three times with PBS and stained with Hoechst 33342 (Beyotime, P0133). Fluorescent images were captured using a fluorescence microscope. Immunohistochemistry Tissue sections were prepared as described above. Following antigen retrieval, endogenous peroxidase activity was quenched. Sections were blocked with 1% BSA at 37 °C for 1 h and then incubated overnight with VLR-like mAb as the primary antibody at 4 °C. After primary antibody incubation, sections were sequentially treated with a signal amplification solution and an enzyme-conjugated goat anti-mouse IgG polymer (ZSGB-BIO, PV-9000) at 37 °C for 20 min. Each step was followed by three washes with PBS. Color development was performed using freshly prepared DAB substrate solution (ZSGB-BIO, ZLI-9018) at room temperature for 5–8 min. Sections were rinsed under running water, counterstained with hematoxylin (Servicebio, G1077) for 20 s, and subjected to differentiation and bluing. Finally, the sections were dehydrated, cleared, mounted, and examined under a light microscope for imaging and analysis. EdU staining Following a pre-injection with 25 μg PHA, lampreys received an intraperitoneal injection of EdU (RiboBio, C00052; 5 mg/kg) in 100 μL PBS 10 h prior to tissue collection to assess cell proliferation. Tissues were collected, and briefly rinsed in PBS to remove excess blood, then fixed and embedded in paraffin as described previously. Tissue sections were prepared as described above. To reduce autofluorescence, sections were washed with 2 mg/mL glycine in PBS for 10 min. Sections were then permeabilized with 0.5% Triton X-100 (Beyotime, ST1723) in PBS for 10 min and blocked with 1% BSA. IF staining for VLR-like was performed using the primary antibody and Alexa Fluor 594-conjugated goat anti-mouse IgG H&L (Abcam, ab150116, 1:800) as described in the “Immunofluorescence” section. After completing the VLR-like staining, EdU detection was carried out according to the manufacturer’s instructions for the Cell-Light EdU Apollo Stain Kit (RiboBio, C10310). Finally, nuclei were stained with Hoechst 33342, and sections were imaged using a fluorescence microscope. Proteolytic enzyme activity assay The proteolytic enzyme activity of lamprey VLR-like⁺ lymphocytes was measured to investigate their capacity for antigen processing and presentation. Intestinal leukocytes were isolated from three biologically independent lampreys. Then, 1 × 106 intestine leukocytes were incubated with inactivated A. hydrophila at 16 °C for 8 h. After incubation, cells were collected, washed, and incubated with BODIPY-labeled casein working solution (from the EnzChek Protease Assay Kit; Invitrogen, E6638) at room temperature for 1 h. Cells were then washed and stained with VLR-like mAb. Finally, samples were analyzed by flow cytometry. Phagocytosis assay To assess phagocytic activity, fluorescent microspheres (Thermo, F13081) were first pre-incubated in L-15 medium supplemented with 10% FBS and 1% penicillin/streptomycin at 16 °C for 30 min. Intestinal leukocytes were isolated from three biologically independent lampreys. Subsequently, 1 × 106 intestine leukocytes were incubated with fluorescent microspheres at a cell/bead ratio of 1:10 at 16 °C for 4 h. After incubation, the cells were collected, washed with cold PBS, and stained with VLR-like mAb. To quench the fluorescence of any beads attached to the outside of the cells, trypan blue (0.2% w/v) was added for 1 min. Finally, samples were analyzed by flow cytometry. Antibody secretion detection Gill leukocytes were lysed in NP-40 lysis buffer (Beyotime, P0013F) on ice for 30 min, followed by centrifugation at 12,000 × g at 4 °C for 10 min to collect the protein supernatant. Lamprey whole blood was allowed to stand at room temperature for 1 h and then centrifuged at 2000 × g at 4 °C for 10 min to obtain serum. Both protein supernatant and serum were mixed with 5× protein loading buffer and boiled for 10 min, and then subjected to western blot assay. For immunoprecipitation, proteins were resolved under reducing SDS-PAGE, stained with Coomassie brilliant blue, and gel bands corresponding to the target were excised and sent for liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. Co-immunoprecipitation (Co-IP) assay The HA- or Flag-tagged full-length coding regions of VLRA, VLRC, and VLR-like were double-digested and cloned into the pEGFP-C1 vector to generate the corresponding transfection plasmids. The primers were synthesized by Sangon Biotech (China), and their sequences are provided in Supplementary Table 3. The plasmids were purified with the E.Z.N.A. Endo-Free Plasmid DNA Midi Kit (Omega Bio-Tek, D6915). Subsequently, the plasmids were added to Opti-MEM containing Lipofectamine 2000 and incubated at room temperature for 20 min. The mixture was gently added to HEK 293 T cells. At 6 h post-transfection, the medium was replaced with DMEM supplemented with 10% FBS and 1% penicillin-streptomycin. At 48 h post-transfection, cells were harvested and lysed on ice for 30 min in NP-40 lysis buffer containing PMSF (MCE, HY-B0496) and Protease Inhibitor (MCE, HY-K0010). Then, 10 μL of anti-Flag Ab-conjugated agarose beads (Beyotime, P2271) were added to the lysate and incubated overnight at 4 °C with gentle shaking. After four washes with NP-40 buffer, the beads were boiled in SDS-PAGE loading buffer for 10 min and analyzed by western blot. Uncropped scans of the key blots are provided in the Source Data file. Statistical analysis GraphPad Prism v10.0 software was used for statistical analysis. All data were presented as mean ± SEM, and statistical significance was determined using the Student’s t-test. The p-values were as follows: *p < 0.05, **p < 0.01, ***p < 0.001. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary information Transparent Peer Review file Reporting Summary Source data Source Data Acknowledgements We thank the Instrument-sharing Platform of the School of Life Sciences of East China Normal University (ECNU) for instrument sharing, the Flow Cytometry Core Facility of the School of Life Sciences of ECNU for the FACS analysis. Author contributions M.G. designed and performed experiments, analyzed data, and drafted the manuscript. K.L., Y.Z., B.Z., J.L., H.C., Z.F., X.J., X.Z., and P.H. performed experiments. X.W., Y.P., and J.Y. conceived the project, designed experiments, analyzed data, and drafted the manuscript. Peer review Peer review information Nature Communications thanks Max Cooper and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Funding This study was supported by grants from National Natural Science Foundation of China (No. 42576117), Natural Science Foundation of Shanghai (24ZR1419700), and Shanghai Sailing Program (24YF2710300). Data availability The single-cell RNA-seq datasets used in this study were generated and published previously by Pang et al. (10.1038/s41467-025-56153-w). The raw sequencing data are available in the CNGB Nucleotide Sequence Archive under accession code CNP0005120 and in the NCBI BioProject database under accession number PRJNA1194219. The processed single-cell RNA-seq data and annotation metadata from Pang et al. are available in the Zenodo repository under 10.5281/zenodo.14338297. The integrated single-cell RNA-seq dataset generated and analyzed in this study has been deposited in the Zenodo repository under 10.5281/zenodo.20263264. All other data are in the main paper or the supplementary materials. The mass spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository under dataset identifier PXD072176. The proteomics data are freely available for reuse through the ProteomeXchange/iProX platform, and the processed proteomics data can be interactively explored and downloaded from the same repository. Source data are provided with this paper. Code availability Custom code used for data analysis is available in GitHub (https://github.com/Yating-Zhu/lamprey) and Zenodo (10.5281/zenodo.20263264). The quality control and cell clustering steps in this work follow the methods provided in the original GitHub repository (https://github.com/lskfs/lamprey-single-cell-atlas) and Zenodo (10.5281/zenodo.14338297). Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Ming Geng, Kang Li, Yating Zhu, Baohua Zou. Contributor Information Xiumei Wei, Email: xmwei@bio.ecnu.edu.cn. Yue Pang, Email: pangyue@lnnu.edu.cn. Jialong Yang, Email: jlyang@bio.ecnu.edu.cn. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-74420-2. References Associated Data Supplementary Materials Supplementary information Transparent Peer Review file Reporting Summary Source Data Data Availability Statement The single-cell RNA-seq datasets used in this study were generated and published previously by Pang et al. (10.1038/s41467-025-56153-w). The raw sequencing data are available in the CNGB Nucleotide Sequence Archive under accession code CNP0005120 and in the NCBI BioProject database under accession number PRJNA1194219. The processed single-cell RNA-seq data and annotation metadata from Pang et al. are available in the Zenodo repository under 10.5281/zenodo.14338297. The integrated single-cell RNA-seq dataset generated and analyzed in this study has been deposited in the Zenodo repository under 10.5281/zenodo.20263264. All other data are in the main paper or the supplementary materials. The mass spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository under dataset identifier PXD072176. The proteomics data are freely available for reuse through the ProteomeXchange/iProX platform, and the processed proteomics data can be interactively explored and downloaded from the same repository. Source data are provided with this paper. Custom code used for data analysis is available in GitHub (https://github.com/Yating-Zhu/lamprey) and Zenodo (10.5281/zenodo.20263264). The quality control and cell clustering steps in this work follow the methods provided in the original GitHub repository (https://github.com/lskfs/lamprey-single-cell-atlas) and Zenodo (10.5281/zenodo.14338297).