Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

Mohindar M. Karunakaran; Carrie R. Willcox; Mahboob Salim; Daniel Paletta; Alina S. Fichtner; Angela Noll; Lisa Starick; Anna Nöhren; Charlotte R. Begley; Katie A. Berwick; Raphaël A.G. Chaleil; Vincent Pitard; Julie Déchanet-Merville; Paul A. Bates; Brigitte Kimmel; Timothy J. Knowles; Volker Kunzmann; Lutz Walter; Mark Jeeves; Fiyaz Mohammed; Benjamin E. Willcox; Thomas Herrmann

doi:10.1016/j.immuni.2020.02.014

Karunakaran, Mohindar M.;Willcox, Carrie R.;Salim, Mahboob;Paletta, Daniel;Fichtner, Alina S.;Noll, Angela;Starick, Lisa;Nöhren, Anna;Begley, Charlotte R.;Berwick, Katie A.;Chaleil, Raphaël A.G.;Pitard, Vincent;Déchanet-Merville, Julie;Bates, Paul A.;Kimmel, Brigitte;Knowles, Timothy J.;Kunzmann, Volker;Walter, Lutz;Jeeves, Mark;Mohammed, Fiyaz;Willcox, Benjamin E.;Herrmann, Thomas;

2020-03-01 00:00:00

University of Birmingham Butyrophilin-2A1 directly binds germline-encoded regions of the Vγ9Vδ2 TCR and is essential for phosphoantigen sensing Karunakaran, Mohindar M.; Willcox, Carrie R.; Salim, Mahboob; Paletta, Daniel; Fichtner, Alina S.; Noll, Angela; Starick, Lisa; Nöhren, Anna; Begley, Charlotte R.; Berwick, Katie A.; Chaleil, Raphaël A.g.; Pitard, Vincent; Déchanet-merville, Julie; Bates, Paul A.; Kimmel, Brigitte; Knowles, Timothy J.; Kunzmann, Volker; Walter, Lutz; Jeeves, Mark; Mohammed, Fiyaz DOI: 10.1016/j.immuni.2020.02.014 License: Creative Commons: Attribution (CC BY) Document Version Publisher's PDF, also known as Version of record Citation for published version (Harvard): Karunakaran, MM, Willcox, CR, Salim, M, Paletta, D, Fichtner, AS, Noll, A, Starick, L, Nöhren, A, Begley, CR, Berwick, KA, Chaleil, RAG, Pitard, V, Déchanet-merville, J, Bates, PA, Kimmel, B, Knowles, TJ, Kunzmann, V, Walter, L, Jeeves, M, Mohammed, F, Willcox, BE & Herrmann, T 2020, 'Butyrophilin-2A1 directly binds germline- encoded regions of the Vγ9Vδ2 TCR and is essential for phosphoantigen sensing', Immunity, vol. 52, no. 3, pp. 487-498.e6. https://doi.org/10.1016/j.immuni.2020.02.014 Link to publication on Research at Birmingham portal General rights Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes permitted by law. •Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. •User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) •Users may not further distribute the material nor use it for the purposes of commercial gain. Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document. When citing, please reference the published version. Take down policy While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been uploaded in error or has been deemed to be commercially or otherwise sensitive. If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access to the work immediately and investigate. Download date: 16. Apr. 2024 Article Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vg9Vd2 TCR and Is Essential for Phosphoantigen Sensing Graphical Abstract Authors Mohindar M. Karunakaran, Carrie R. Willcox, Mahboob Salim, ..., Fiyaz Mohammed, Benjamin E. Willcox, Thomas Herrmann Correspondence b.willcox@bham.ac.uk (B.E.W.), herrmann-t@vim.uni-wuerzburg.de (T.H.) In Brief Karunakaran et al. ﬁnd that butyrophilin 2A1 (BTN2A1) associates with BTN3A1 on the cell surface and binds directly to germline-encoded regions of the Vg9 chain of the Vg9Vd2 TCR. Thus, BTN2A1 collaborates with BTN3A1 to potentiate Vg9Vd2 T cell recognition, playing an essential role in phosphoantigen sensing. Highlights d Radiation hybrids identify BTN2A1 as crucial for Vg9Vd2 phosphoantigen (P-Ag) sensing d BTN2A1 binds directly to the T cell receptor via germline- encoded regions of Vg9 d Cell-surface BTN2A1 associates directly with BTN3A1 independent of P-Ag stimulation d The Vg9-BTN2A1 interaction modality suggests an additional CDR3-dependent TCR ligand Karunakaran et al., 2020, Immunity 52, 487–498 March 17, 2020 ª 2020 The Authors. Published by Elsevier Inc. https://doi.org/10.1016/j.immuni.2020.02.014 Immunity Article Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vg9Vd2TCR andIs Essential for Phosphoantigen Sensing 1,11 2,3,11 2,3 1 1 4 Mohindar M. Karunakaran, Carrie R. Willcox, Mahboob Salim, Daniel Paletta, Alina S. Fichtner, Angela Noll, 1 1 2,3 2,3 5 6,7 Lisa Starick, Anna No¨ hren, Charlotte R. Begley, Katie A. Berwick, Raphael A.G. Chaleil, Vincent Pitard, 6,7 5 8 9 8 4 Julie De´ chanet-Merville, Paul A. Bates, Brigitte Kimmel, Timothy J. Knowles, Volker Kunzmann, Lutz Walter, 10 2,3 2,3,11,12, 1,11, Mark Jeeves, Fiyaz Mohammed, Benjamin E. Willcox, * and Thomas Herrmann * € € Institute for Virology and Immunobiology, University of Wurzburg, Wurzburg, Germany Institute of Immunology and Immunotherapy, University of Birmingham, Birmingham, UK Cancer Immunology and Immunotherapy Centre, University of Birmingham, Birmingham, UK Primate Genetics Laboratory, German Primate Center, Leibniz Institute for Primate Research, Go¨ ttingen, Germany Biomolecular Modelling Laboratory, The Francis Crick Institute, London, UK ImmunoConcEpT Laboratory, Equipe labellise´ e, LIGUE 2017, UMR 5164, Bordeaux University, CNRS, 33076 Bordeaux, France Flow Cytometry Facility, TransBioMed Core, Bordeaux University, CNRS UMS 3427, INSERM US05, 33076 Bordeaux, France € € Medical Clinic and Policlinic II, University of Wurzburg, Wurzburg, Germany School of Biosciences, University of Birmingham, Birmingham, UK Henry Wellcome Building for NMR, Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, UK These authors contributed equally Lead Contact *Correspondence: b.willcox@bham.ac.uk (B.E.W.), herrmann-t@vim.uni-wuerzburg.de (T.H.) https://doi.org/10.1016/j.immuni.2020.02.014 SUMMARY INTRODUCTION Vg9Vd2 T cells respond in a TCR-dependent Human peripheral blood gd T cells are dominated from an early age by Vg9Vd2 lymphocytes (Parker et al., 1990), an innate-like fashion to both microbial and host-derived subset that features a predominant effector status, allowing pyrophosphate compounds (phosphoantigens, or potent cytokine production and cytotoxic capability that is linked P-Ag). Butyrophilin-3A1 (BTN3A1), a protein struc- to a relatively restricted T cell receptor (TCR) repertoire (Davo- turally related to the B7 family of costimulatory deau et al., 1993; Delfau et al., 1992). Vg9Vd2 T cells universally molecules, is necessary but insufﬁcient for this respond in a TCR-dependent fashion to non-peptidic pyrophos- process. We performed radiation hybrid screens phate compounds (phosphoantigens [P-Ag]). These include the to uncover direct TCR ligands and cofactors microbially derived compound (E)-4-hydroxy-3-methyl-but- that potentiate BTN3A1’s P-Ag sensing function. 2-enyl pyrophosphate (HMBPP) (Morita et al., 2007), which is These experiments identiﬁed butyrophilin-2A1 generated by the non-mevalonate isoprenoid synthetic pathway (BTN2A1) as essential to Vg9Vd2 T cell recognition. and is a highly potent activator of Vg9Vd2 T cells. In addition, BTN2A1 synergised with BTN3A1 in sensitizing host-cell-derived isoprenyl pyrophosphate (IPP) can act as a P-Ag and stimulate Vg9Vd2 T cell responses. IPP levels are P-Ag-exposed cells for Vg9Vd2 TCR-mediated re- elevated in some cancer cells and can also be therapeutically sponses. Surface plasmon resonance experiments increased in target cells via aminobisphosphonate drugs that established Vg9Vd2 TCRs used germline-encoded inhibit IPP catabolism, such as Zoledronate (Zol) (Gober et al., Vg9 regions to directly bind the BTN2A1 CFG-IgV 2003; Kunzmann et al., 2000). domain surface. Notably, somatically recombined Vg9Vd2-mediated P-Ag sensing requires cell-cell contact CDR3 loops implicated in P-Ag recognition were (Morita et al., 1995) and depends on both Vg and Vd chains, uninvolved. Immunoprecipitations demonstrated with evidence for involvement of multiple complementarity- close cell-surface BTN2A1-BTN3A1 association in- determining region (CDR) loops (Wang et al., 2010). An essential dependent of P-Ag stimulation. Thus, BTN2A1 is a prerequisite for P-Ag sensing is target-cell expression of butyro- BTN3A1-linked co-factor critical to Vg9Vd2TCR philin (BTN) 3A1 (Harly et al., 2012), a member of a multi-gene recognition. Furthermore, these results suggest a family encoded on chromosome (Chr) 6. BTNs and butyrophi- lin-like (BTNL) molecules are structurally related to the B7 family composite-ligand model of P-Ag sensing wherein of costimulatory molecules, comprising two extracellular immu- the Vg9Vd2 TCR directly interacts with both noglobulin (Ig)-like domains, a transmembrane region, and a BTN2A1 and an additional ligand recognized in a cytoplasmic tail that often contains a B30.2 domain (Rhodes CDR3-dependent manner. Immunity 52, 487–498, March 17, 2020 ª 2020 The Authors. Published by Elsevier Inc. 487 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). et al., 2016). In addition to immunomodulatory effects on anti- (CHO Chr6 cells). Based on this observation, we postulated the gen-presenting cells and conventional ab T cells, several BTN existence of a Factor X encoded on Chr 6, which in addition to and/or BTNL family members are emerging as playing critical BTN3A1 is mandatory for P-Ag-mediated gd T cell stimulation roles in gd T cell development and activation (Boyden et al., (Rian˜ o et al., 2014). 2008; Di Marco Barros et al., 2016; Harly et al., 2012; Melandri To identify Factor X, we used an unbiased genome-based et al., 2018; Vantourout et al., 2018; Willcox et al., 2019). approach involving generation of radiation hybrids between Although the extracellular domain of BTN3A1 was initially re- CHO-Chr 6 cells and BTN3A1-transduced hypoxanthine- ported to present P-Ag and directly bind the Vg9Vd2 TCR (Va- aminopterin-thymidine (HAT)-sensitive rodent fusion partners vassori et al., 2013), other studies have challenged both of these and subsequent analysis of their capacity to stimulate P-Ag ﬁndings and instead support the concept that BTN3A1 senses sensing by Vg9Vd2 T cells (Figure 1A). We postulated that com- P-Ag directly. These data include robust evidence for P-Ag bind- parison of the human gene products transcribed in stimulatory ing to the intracellular B30.2 domain of BTN3A1 and for a P-Ag- radiation hybrids would allow mapping of the gene(s), which induced conformational change (Nguyen et al., 2017; Salim et al., alongside BTN3A1 are mandatory for PAg-mediated stimulation. 2017; Sandstrom et al., 2014). In addition, the importance of We fused CHO-Chr6 cells separately with two HAT-sensitive BTN3A2 and/or BTN3A3 co-expression alongside BTN3A1 for fusion partners—ﬁrst BTN3A1-transduced A23 hamster cells optimal P-Ag sensing has been highlighted, as well as the poten- and second BTN3A1-transduced mouse BW cells (STAR tial of these family members to heterodimerize with BTN3A1 in an Methods)(Sanderson and Shastri, 1994); resulting radiation hy- IgC-dependent manner (Vantourout et al., 2018). brids were assessed for P-Ag-dependent activation of TCR- Vg9Vd2 T cells emerged with the appearance of placental MOP transductants and positive candidates cloned by limiting mammals and have been retained in both primates and species dilution. In some cases, these clones were used as donor cells as diverse as dolphin (Tursiops truncatus) and alpaca (Vicugna for further fusions. A ﬁnal selection of clones (Figure S1A) were pacos)(Fichtner et al., 2018; Karunakaran et al., 2014). Of subjected to RNA sequencing (RNA-seq) analysis (Figure 1B; note, the alpaca is the only non-primate species to date with STAR Methods) alongside CHO-Chr6 cells and rodent fusion proven P-Ag reactivity of Vg9Vd2 T cells and P-Ag binding to partner cells as positive and negative controls, respectively. BTN3 demonstrated (Fichtner et al., 2020). In contrast, rodents A region of 580 kB of Chr 6 permitting P-Ag-mediated stim- lack BTN3, Vd2, and Vg9 homologs (Karunakaran et al., 2014). ulation by the radiation hybrids was identiﬁed (Figures 1C and Consistent with an important role in host immunity, Vg9Vd2 S1B). Analysis of candidate genes within this region revealed T cell expansion and activation is observed in a variety of micro- that the only transmembrane molecules among the expressed bial infections (Morita et al., 2007). Furthermore, attempts to human genes were the major histocompatibility complex therapeutically harness the human gd T cell compartment have (MHC)-class-I-like iron transporter HFE, the BTN3A1 gene hitherto focused predominantly on the Vg9Vd2 subset in the already transduced into rodent fusion partners, BTN3A2, context of both speciﬁc infections (Shen et al., 2019) and cancer BTN3A3, and BTN2A1 and BTN2A2. Since we knew that expres- (Kunzmann et al., 2000; Silva-Santos et al., 2019). From this sion of all three BTN3 genes was insufﬁcient for reconstitution of perspective, the mechanism underpinning Vg9Vd2 T cell activa- the P-Ag response (D.P., A.S.F., M.M.K., and T.H., unpublished tion has been a focus of strong interest. data), BTN2A1 and BTN2A2, which to date have been discussed Our previous studies have established that BTN3A1 is neces- mainly for their immunomodulatory properties (Rhodes et al., sary but not sufﬁcient for P-Ag sensing and indicated the exis- 2016), emerged as the prime candidates for encoding Factor X. tence of an additional putative Chr-6-encoded factor that syner- We then tested the effects on P-Ag-dependent stimulation of gized with BTN3A1 to stimulate P-Ag-mediated responses Vg9Vd2 lymphocytes by human 293T cells after CRISPR-Cas9- (Rian˜ o et al., 2014), which we subsequently coined ‘‘Factor X’’ mediated inactivation of either both BTN2 genes (BTN2 )or / / (Karunakaran and Herrmann, 2014). Here, we set out to identify BTN2A1 (BTN2A1 )or BTN2A2 alone (BTN2A2 ). Inactiva- Factor X using a radiation hybrid approach. We identiﬁed tion of both BTN2 genes completely abolished interferon (IFN)g BTN2A1 as this critical factor and showed it interacts directly production by polyclonal Vg9Vd2 T cell lines in response to with the Vg9Vd2 TCR to potentiate P-Ag-dependent recognition, Zol pulsed cells (Figure 1D). Crucially, both BTN2 and highlighting its role in a ‘‘composite ligand’’ model of Vg9Vd2 BTN2A1 exhibited a complete loss of IL-2 production by T cell recognition. TCR-MOP cells in response to either HMBPP (Figure 1E) or 20.1 mAb (Figure 1F), whereas responses to BTN2A2 were RESULTS similar to wild-type (WT) 293T cells (Figures 1E and 1F). These experiments strongly suggested that alongside BTN3A1, Radiation Hybrids Identify BTN2A1 as Essential for P-Ag BTN2A1 was critical for P-Ag sensing. Sensing We showed previously that a T cell hybridoma expressing BTN2A1 and BTN3A1 Are Sufﬁcient to Potentiate the Vg9Vd2 MOP TCR produced interleukin (IL)-2 in co-culture Vg9Vd2-Mediated P-Ag Sensing with BTN3A1-transduced Chinese hamster ovary (CHO) cells To address whether BTN2A1 was sufﬁcient alongside BTN3A1 incubated with the anti-BTN3A1 monoclonal antibody (mAb) to reconstitute P-Ag sensitization in rodent cells, we transduced + + 20.1 but exhibited a complete lack of response to HMBPP or either one or both genes into both CD80 BW and CD80 CHO Zol (Rian˜ o et al., 2014; Starick et al., 2017). In contrast, HMBPP cells (Figures 2A–2C) and tested their ability to induce IL-2 pro- and Zol sensitivity was restored in co-cultures with human-rodent duction from TCR-MOP cells following incubation with HMBPP. hybrid cells, including CHO cells containing a single human Chr 6 In both cases, whereas transduction of BTN3A1 alone resulted in 488 Immunity 52, 487–498, March 17, 2020 A D ** 100/300 Gy Medium HAT Limiting dilution Zol (25 μM) CHO sensitive Radiation P-Ag or fusion with huChr6 rodent line Hybrid “sensing” P-Ag “sensing” (donor) +BTN3A1 clones RH clone RH donor +CD80 HAT Selection for Fusion selection P-Ag reactivity sequencing BTN3A1 CHO-Chr6 CHO-20-54 BW-2-1-2 CHO-20-20 -/- BTN2 BW-2-10-1 CHO-20-46 -/- BTN2 CHO-100-7 -/- BTN2A1 10 -/- BTN2A1 -/- BTN2A2 -/- BTN2A2 293T 293T 10 0 0.013 0.04 0.12 0.36 1.1 3 10 HMBPP (μM) -/- BTN2 1 -/- BTN2 -/- 80 BTN2A1 -/- BTN2A1 60 -/- BTN2A2 -/- BTN2A2 293T Human Chromosome 6 (distances in Mb) 293T Human Chr 6 BTN3A2 BTN2A2 BTN3A1 BTN2A3P BTN3A3 BTN2A1 BTN1A1P1 BTN1A1 0 0.016 0.008 0.04 0.2 1 26.36 Mb 26.50 Mb 20.1 mAb (μM) Figure 1. Identiﬁcation of BTN2A1 as Factor X (A) Radiation hybrid approach to generate and identify rodent cell-fusion hybrids incorporating portions of human chromosome (Chr) 6 that permit P-Ag sensitization. (B) RNA-seq analysis of prioritized clones generated from fusion with A23 or BW cells. Values for less than three transcripts are merged with the x axis. (C) Arrangement of BTN gene cluster on Chr 6 extracted from genome data viewer GRCh38.p13 (GCF_000001405.39). (D) Production of IFNg from polyclonal Vg9Vd2 T cell lines in response to Zol-treated WT or BTN2 293T cells. Error bars represent standard deviation for three independent experiments. **p < 0.005. / / / (E) Production of IL-2 from TCR-MOP transductants in response to HMBPP-treated WT, BTN2 , BTN2A1 , and BTN2A2 293T cells. / / / (F) Production of IL-2 from TCR-MOP transductants in response to 20.1 mAb-treated WT, BTN2 , BTN2A1 , and BTN2A2 293T cells. In (E) and (F), the different colors indicate results from two independent experiments. See also Figure S1. negligible responses, transduction of both BTN2A1 and BTN3A1 dependent IL-2 response observed in the absence of P-Ag or permitted a robust, HMBPP-dose-dependent IL-2 response, BTN3A1 (Figures 2A and 2B), BTN3A1 co-expression was not conﬁrming their sufﬁciency for P-Ag sensitization (Figures 2B required for BTN2A1-mediated tetramer staining (Figure 2D), and 2C). Interestingly, transduction of BTN2A1 alone resulted nor was exposure to Zol necessary for tetramer staining (Fig- in a weak, HMBPP-dose-independent basal response to both ure 2F). This suggested BTN2A1 might be an independent ligand cell lines (Figures 2B and 2C). for the Vg9Vd2 TCR, the activatory potential of which is critically To assess whether BTN2A1 surface expression was able to augmented in a BTN3A1- and P-Ag-dependent manner. support binding to the Vg9Vd2 TCR, we generated Vg9Vd2 We then investigated why BTN2A2, which shares close 88% TCR tetramers and used them to stain transduced BW and sequence identity with BTN2A1 in its extracellular region, was un- 293T cells (Figures 2D and 2E). BTN2A1 expression on trans- able to potentiate P-Ag sensing alongside BTN3A1. BTN2A2- duced BW and 293T cells was sufﬁcient to enable staining by 293T transductants did not support tetramer staining (Figure S2A), Vg9Vd2 MOP-TCR tetramer (Figures 2D, 2E, and S2A), support- suggesting BTN2A2 might not be able to recognize the Vg9Vd2 ing the idea that BTN2A1 may be a direct TCR ligand; moreover, TCR. However, one major caveat was the considerably lower sur- all Vg9Vd2 TCR tetramers tested stained BTN2A1-transduced face expression of BTN2A2 relative to BTN2A1 in 293T transduc- cells (Figure 2E). Consistent with the minimal basal BTN2A1- tants (Figure S2B), which could also explain this observation. Immunity 52, 487–498, March 17, 2020 489 4,97 11,06 18,15 25,27 27,13 28,73 30,00 31,35 31,15 33,69 36,94 41,78 43,78 52,36 57,97 71,34 78,87 293T 86,02 97,79 105,66 110,59 -/- BTN2 116,56 125,99 132,75 137,99 145,73 151,10 158,00 166,06 170,07 normalized transcripts IL-2 (pg/ml) IL-2 (pg/ml) IFNγ (pg/ml) A Untransduced BTN3A1 BTN2A1 BTN2A1 + BTN3A1 B BW BW BTN2A1 BTN2A1 BW BTN3A1 BTN3A1 60 BW BTN2A1 2° only 40 + BTN3A1 0 1 2 3 4 anti-mouse 2° PE HMBPP (μM) -/- -/- CD E BW 293T BTN2 293T BTN2 + BTN2A1 CHO G115 TCR Untransduced CHO BTN2A1 MOP TCR BTN2A1 CHO BTN3A1 D1C55 TCR BTN3A1 CHO BTN2A1 LES TCR + BTN3A1 BTN2A1 + BTN3A1 HMBPP (μM) TCR tetramer-PE TCR tetramer-PE Untransduced Untransduced BTN2A1 + BTN3A1 BTN2A1 + BTN3A1 - Zol + Zol - Zol + Zol Streptavidin-PE alone MOP TCR Tetramer-PE anti-mouse 2° PE anti-BTN2A1 + anti-mouse 2° PE PE Figure 2. BTN2A1 and BTN3A1 Synergize to Potentiate P-Ag Sensing in Rodent Cells (A) Expression of BTN2A1, BTN3A1, or both genes in transduced BW cells. (B) Production of IL-2 by TCR-MOP transductants in response to HMBPP-treated CD80 BW cells transduced to express BTN2A1, BTN3A1, both, or un- transduced controls. Percentage activation is normalized against the maximum response obtained from CD80 CHO cells expressing both BTN2A1 and BTN3A1 in the presence of 10 mM HMBPP. (C) Production of IL-2 from TCR-MOP transductants in response to HMBPP-treated CD80 CHO cells transduced with either BTN2A1, BTN3A1, both genes, or untransduced controls, with responses normalized as in (B). Error bars in (B) and (C) represent standard deviation for three independent experiments. Differences between untransduced and BTN2-transduced cells were signiﬁcant (p < 0.05), as were those between the BTN2A1-transductant and BTN2A1+BTN3A1-transductant in the presence of HMBPP. (D) MOP-TCR tetramer staining of transduced BW cells. (E) Staining of transduced 293T cells with Vg9Vd2 TCRs. (F) MOP-TCR tetramer staining or anti-BTN2A1 mAb staining of BTN2A1 and BTN3A1-transduced CD80 BW cells versus untransduced controls in the presence and absence of Zol. See also Figure S2. BTN2A1 IgV Domain Directly Binds Germline-Encoded ure 3A); consistent with this, Vg9Vd1 TCR tetramers speciﬁcally Regions of Vg9 TCRs stained BTN2A1-transduced 293T cells (Figure S3A). Taking To establish whether BTN2A1 acted as a direct ligand for the into account the highly similar afﬁnity (K 46.6mM [n = 8]) of the Vg9Vd2 TCR, we expressed the membrane-distal domain of BTN2A1-Vg9Vd1 TCR interaction (Figure 3B) and the radically the BTN2A1 ectodomain and tested direct binding to recombi- divergent CDR3g expressed by this TCR relative to Vg9Vd2 nant Vg9Vd2 TCR using surface plasmon resonance (SPR). In- TCRs (Figure S3A), these results strongly suggested the jection of BTN2A1 IgV produced substantially enhanced signals BTN2A1-Vg9Vd2 interaction focused on germline encoded over surfaces with immobilized Vg9Vd2 TCR relative to Vg4Vd5 regions of the Vg9 IgV domain. This implied that the BTN2A1- and Vg2Vd1 TCRs or control streptavidin surfaces, indicating Vg9Vd2 interaction might be analogous to BTNL3 binding to speciﬁc binding (Figure 3A). Equilibrium afﬁnity measurements human Vg4 TCRs (Melandri et al., 2018; Willcox et al., 2019), of BTN2A1 IgV binding to the G115 and MOP Vg9Vd2 TCRs es- which is similarly focused on germline-encoded regions of the tablished K values of 45.4 mM (n = 9) and 49.9 mM (n = 8), respec- Vg4 chain and allowed us to model BTN2A1-Vg9 interaction tively (Figure 3B). based on the proposed BTNL3-Vg4 interaction mode (Figure 3C). Unexpectedly, experiments also indicated clear binding of An initial homology model suggested strong feasibility of a BTN2A1 IgV to a Vg9Vd1 TCR, which was derived from the similar interaction mode and highlighted seven amino acids on non-P-Ag reactive Vd1 T cell subset (Halary et al., 2005)(Fig- the face of the BTN2A1 IgV domain incorporating the C, C’, F 490 Immunity 52, 487–498, March 17, 2020 3.3 1.1 0.37 0.12 0.04 0.01 0.1 percent activation percent activation AB 200 200 G115 TCR MOP TCR Vγ9Vδ1 TCR BTN2A1 IgV Vγ9Vδ2 TCR 25 μM Vγ4Vδ5 TCR 150 150 Vγ2Vδ1 TCR Streptavidin 5 6 4 100 50 50 0 0 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 Bound BTN2A1 (RU) Bound BTN2A1 (RU) Bound BTN2A1 (RU) 0 0 0 -30 0 30 60 050 100 0 50 100 0 50 100 Time (s) BTN2A1 IgV (μM) BTN2A1 IgV (μM) BTN2A1 IgV (μM) 300 BTN2A1 IgV C C C G115 TCR 24 μM MOP TCR Vγ9Vδ1 TCR C LES TCR A’ S72 150 F 100 E D R124 E135 90q C’ C’’ R65 -30 0 30 60 K79 Y126 Time (s) Y133 Vγ4 BTNL3-IgV Vγ9 BTN2A1-IgV DF E A’ R20 E70 60 T83 HV4 BC E D V57 T79 40 C’’ I74 C’ T56 CDR2 BTN3A1 + BTN2A1 WT 1 T77 HV4 20 BTN3A1 + BTN2A1 R124E CDR2 BTN3A1 + BTN2A1 R124A E76 CDR1 1 0.1 0 Vγ9 interface surface HMBPP (μM) CDR3 BTN2A1 mutant GH 100 100 MOP WT MOP WT MOP Vγ9 E70A MOP Vδ2 R51A MOP Vγ9 E70R MOP Vδ2 ΔCDR3 80 80 MOP Vγ9 E70K 60 60 40 40 20 20 0 0 1 0.1 0 1 0.1 0 HMBPP (μM) HMBPP (μM) Figure 3. Direct BTN2A1 Binding to Germline-Encoded Regions of Vg9 Is Essential for P-Ag Sensing (A) (Top panel) Injection of BTN2A1 IgV (25 mM) over surfaces with immobilized Vg9Vd2 TCR (2,457 resonance units (RU)) and control surfaces comprising Vg4Vd5 TCR (2,351 RU), Vg2Vd1 TCR (1,800 RU), or streptavidin alone. Notably, signals over streptavidin alone and control TCR surfaces are equivalent. (Bottom panel) Injection of BTN2A1 IgV (24 mM) over surfaces with immobilized G115 (Vg9Vd2; 3,109 RU), MOP (Vg9Vd2; 3,108 RU), and Vg9Vd1 (2,774 RU) TCRs and LES TCR control (Vg4Vd5; 2,885 RU). (B) Equilibrium afﬁnity measurements and Scatchard analysis (inset) of BTN2A1 IgV binding to the G115 (K = 39.5 mM) and MOP (K = 48.4 mM) Vg9Vd2 TCRs and d d Vg9Vd1TCR (K = 47.9 mM). Data in (A) and (B) are representative of eight to nine independent experiments. (C) Model of the BTN2A1-Vg9 interaction mode based on the proposed BTNL3-Vg4 interaction, with expanded panel showing potential contacts at the Vg9- BTN2A1 IgV interface. (D) Effects of seven alanine substitutions in proposed BTN2A1 interface residues on Vg9Vd2 TCR interaction, indicating afﬁnity of mutant BTN2A1 relative to WT BTN2A1 calculated in the same experiment. Data shown are representative of two independent experiments. (E) Effects of BTN2A1 R124A and R124E mutations on IL-2 production by TCR-MOP in response to HMBPP-treated BTN3A1 and BTN2A1 expressing CD80 CHO cells. (F) Predicted involvement of Vg9 HV4 and CDR2 residues in BTN2A1 interaction. (G) Effects of Vg9-E70 mutation (HV4) on IL-2 production by TCR-MOP in response to HMBPP-treated BTN3A1 and BTN2A1 expressing CD80 CHO cells. (legend continued on next page) Immunity 52, 487–498, March 17, 2020 491 WT R65A S72A K79A R124A Y126A Y133A E135A Percent activation WT K /mutant K Response (RU) Response (RU) d d Bound BTN2A1 (RU) Percent activation Percent activation Bound/Free (RU/μM) Bound BTN2A1 (RU) Bound/Free (RU/μM) Bound BTN2A1 (RU) Bound/Free (RU/μM) and G b strands (CFG face), equivalent to the region of BTNL3 IgV IL-2 responses and BTN2A1-dependent P-Ag-independent basal domain involved in binding Vg4, as candidates for alanine muta- responses (Figure 3H). tion (Figure 3C). Individual BTN2A1 alanine mutants were gener- Collectively, these ﬁndings established that Vg9Vd2 TCR ated for these seven residues. Of these, four completely abro- binds BTN2A1 IgV via a binding mode that closely mimics that gated BTN2A1 binding to Vg9Vd2 TCR (R65A, R124A, Y126A, of Vg4 TCR for BTNL3 and that this binding is essential for E135A), a ﬁfth marginally decreased afﬁnity (K79A), Y133A did P-Ag sensing but occurs alongside parallel and essential Vd2 not affect binding, and S72A increased afﬁnity (K 10–15 mM) (Fig- CDR-mediated binding events. ures 3D and S3B). These results allowed generation of an improved, mutationally informed model of BTN2A1/Vg9 interac- BTN2A1 Can Form Disulphide-like Homodimers at the tion using the high-ambiguity driven protein-protein DOCKing Cell Surface (HADDOCK) software, analysis of which outlined a molecular BTN and BTNL molecules have been shown to form either homo- rationale for the effect of each mutation (Figure S3C; STAR or heterodimers (Palakodeti et al., 2012; Vantourout et al., 2018). Methods). Based on comparison of the BTN2A2 IgV sequence To investigate BTN2A1’s propensity for dimer formation, we car- (Figure S3D) and a BTN2A2 homology model (Figure S3E) in the ried out homology modeling of BTN2A1 IgV-C (Figures 4A and context of this BTN2A1 model, we predicted that BTN2A2 IgV 4B) based on superposition of BTN2A1 onto the structure of would also be competent for Vg9 TCR binding, which was subse- the BTN3A1 V-shaped homodimer (Palakodeti et al., 2012). In- quently conﬁrmed using SPR for both Vg9Vd2 TCRs (Figures S3F spection of the model conﬁrmed a viable IgC-IgC homodimer and S3G) and a Vg9Vd1 TCR (Figures S3G and S3H), which indi- interface driven by main-chain-main-chain hydrogen bonding in- cated a similar afﬁnity to BTN2A1 (K 39–50 mM [n = 3]). teractions supplemented by side-chain-dependent hydrophobic To assess the dependence of the functional activity of BTN2A1 contacts; these were predicted to be broadly equivalent to those on TCR binding, we transduced CHO-BTN3A1 cells with the of BTN3A1 IgC-IgC, albeit with increased interchain hydropho- BTN2A1 R124A mutation shown to abrogate Vg9 TCR binding bic contacts in BTN2A1 (M153, F235) versus BTN3A1 (V154, (and also a BTN2A1 R124E charge-reversal mutant) and as- S236), indicating a strong potential for non-covalent dimer for- sessed effects on Vg9Vd2-mediated P-Ag response. Although mation (Figure S4A); in addition, further analyses indicated a permissive for cell-surface BTN2A1 expression (Figure S3I), similar potential for heterodimer formation with other members both mutations completely abrogated both P-Ag-dependent of the BTN family (Figure S4B). IL-2 production and basal P-Ag-independent BTN2A1-mediated Interestingly, the BTN2A1 IgV-C model also highlighted close responses (Figure 3E). Furthermore, BTN2A1 R65A and Y126A proximity of extracellular membrane-proximal cysteine residue mutations that eliminated Vg9 TCR-BTN2A1 interaction also (C247) with its equivalent residue in the opposing monomer (Fig- abrogated P-Ag-dependent and independent responses (Fig- ure 4B), suggesting that the homodimer might be stabilized addi- ure S3J). However, although mCherry reporter signal was de- tionally by an interchain disulphide bond; of note, this cysteine is tected for each construct, it must be noted that these mutant lacking in all other BTN and BTNL molecules (Figure S4C). proteins could not be detected using the anti-BTN2A1 mAb (Fig- Indeed, SDS-PAGE and immunoblot-streptavidin detection of ures S3K and S3L). We therefore could not exclude the possibil- BTN2A1 under reducing and non-reducing conditions conﬁrmed ity that these mutations affected cell-surface expression, that the overwhelming majority of cell surface BTN2A1 was pre- although alternatively, they could be important components of sent as a disulphide-bonded dimer (Figure 4C), even in the pres- the anti-BTN2A1 mAb epitope. ence of BTN3A1 and in the presence and absence of Zol (Fig- The mutationally guided model also indicated involvement of ure S4D), consistent with the BTN2A1 homodimer model multiple TCR residues in the HV4 (including E70, I74, E76, T77, (Figures 4A and 4B). However, transduction of BTN2A1 bearing T79) and CDR2 (G56, T57, V58) loops of the Vg9 IgV domain in a C247W mutation did not affect P-Ag-dependent or P-Ag-inde- BTN2A1 interaction, regions also critical for BTNL3-Vg4interac- pendent basal IL-2 production (Figure 4D); therefore, any disul- tion (Willcox et al., 2019)(Figure 3F). Consistent with this, phide stabilization may be redundant owing to strong existing Vg9Vd2-expressing hybridomas bearing mutations at TCRg HV4 non-covalent homodimer potential. In contrast, BTN2A2, which E70 eliminated BTN2A1-dependent P-Ag-independent IL-2 pro- lacks this cysteine residue, did not form disulphide-linked dimers duction and substantially affected TCR-dependent P-Ag re- (Figure 4C), but analogous structural modeling indicated equiva- sponses, with E70K exhibiting severely reduced activation poten- lent propensity for non-covalent IgC-IgC-mediated homodimer tial (Figure 3G). Although the BTN2A1-Vg9model wassupported formation (Figure S4E). by our BIAcore data (Figure 3A) in indicating no role for Vd in Finally, by combining our HADDOCK-derived model of BTN2A1 recognition, we sought to establish whether BTN2A1- Vg9Vd2/BTN2A1 interaction (Figure S3B) with our BTN3A1- dependent P-Ag sensing was nevertheless affected by Vd2CDR based homology model of the BTN2A1 homodimer (Figure 4A), loops by generating TCR hybridomas bearing either a CDR3 dele- we were able to envisage how BTN2A1 recognition might take tion of TCR-MOP (DCDR3) (Figure S3B) or a R51A substitution in place at the cell surface (Figure 4E). Notably, the Vg9Vd2 TCR- CDR2 (Li, 2010). Each mutation abolished both P-Ag-dependent BTN2A1 interaction mode can in principle allow clustering of (H) Effects of mutations in Vd2 CDR2 (R51A) or a deletion in CDR3 (DCDR3) on IL-2 production by TCR-MOP in response to HMBPP-treated BTN3A1 and BTN2A1 expressing CD80 CHO cells. In (E), (G) and (H), error bars indicate standard deviation for three independent stimulation experiments. Percentage activation is normalized against the maximum response obtained from CHO cells expressing both BTN2A1 and BTN3A1 in the presence of 1 mM HMBPP. Differences between WT and mutants in (E), (G), and (H) were signiﬁcant, except for TCR-MOP E70A at 1 mM. See also Figure S3. 492 Immunity 52, 487–498, March 17, 2020 A BC D NR R MW BTN3A1+ BTN2A1 wt (kDa) BTN3A1+ BTN2A1 C247W 130 60 C247 C247 10 1 0 BTN2A1 IgC-IgC homodimer HMBPP (μM) T cell Vγ9-IgC Vγ9-IgC Vδ2-IgC Vδ2-IgC IgV IgV Vδ2-IgV Vδ2-IgV Vδ2-CDR2 Vγ9-IgV Vδ2-CDR3 Vγ9-IgV Vδ2-CDR2 Vδ2-CDR1 Vγ9-CDR3 Vδ2-CDR3 Vδ2-CDR1 Vγ9-CDR3 IgC IgC Target cell BTN2A1 homodimer Figure 4. BTN2A1 Forms Disulphide-Linked Homodimers at the Cell Surface (A) Homology model of BTN2A1 homodimer. (B) C-terminal region of BTN2A1 homology model indicating close proximity of Cys residues. (C) Non-reducing (NR) or reducing (R) SDS-PAGE analysis of CHO-cell expressed BTN2A1 and BTN2A2 protein. (D) Effects of BTN2A1C247W mutation on IL-2 production by TCR-MOP in response to HMBPP-treated CD80 CHO cells expressing BTN2A1 and BTN3A1. Error bars indicate standard deviation for three independent stimulation experiments. Percentage activation is normalized against the maximum response obtained from CHO cells expressing both BTN2A1 and BTN3A1 in the presence of 1 mM HMBPP. (E) Model of Vg9Vd2-BTN2A1 interaction incorporating BTN2A1 homodimer formation, and bilateral Vg9Vd2 interaction with BTN2A1 IgV domain. See also Figure S4. two TCRs for each BTN2A1 homodimer, each with somatically tagged BTN3A1 (FLAG-BTN3A1). Immunoprecipitation (IP) us- recombined CDR3 loops implicated in P-Ag sensing oriented ing anti-HA beads or 20.1 mAb and subsequent anti-FLAG west- directly toward the target cell surface. ern blot (WB) was used to detect cross-linked BTN2A1-BTN3A1 species (Figure 5). BTN2A1 Is Closely Associated with BTN3A1 at the Cell Following IP of BTN2A1-HA using anti-HA beads and subse- Surface quent WB detection of FLAG-BTN3A1 under reducing condi- To assess whether BTN2A1 and BTN3A1 were associated with tions using an anti-FLAG antibody, two discrete bands were each other at the cell surface either before or after P-Ag expo- detected that considerably exceeded the size of either sure, a membrane-impermeable amine-reactive cross-linker BTN2A1 or BTN3A1 monomers (observed molecular weights incorporating a 16A spacer was used to cross-link proteins on [MWs] 55k Da, FLAG-BTN3A1; 70 kDa, BTN2A1-HA) (Fig- the surface of CHO transductants co-expressing C-terminally ure 5A); each was only detected in the presence of cross-linker. HA-tagged BTN2A1 (BTN2A1-HA) and a N-terminally FLAG- One band was 130 kDa, equivalent to cross-linking of a single Immunity 52, 487–498, March 17, 2020 493 HA-2A1 HA-2A2 HA-2A1 HA-2A2 % activation 15 N (ppm) A B HA IP 20.1 IP CHO cells: 2A1 HA 3A1 FLAG 2A1 HA + 3A1 FLAG 2A1 HA 3A1 FLAG 2A1 HA + 3A1 FLAG Zol -- ++ + + + + -- ++ + + + + FLAG sEGS XL - -- - - - - - ++ + + + + + + BTN2A1 BTN3A1 ~250 kDa MW (kDa) HA BTN2A1 ~125 kDa BTN3A1 HA FLAG BTN3A1 ~55 kDa CD 0.04 119.5 E135 0.03 120.0 N115 120.5 0.02 A109 121.0 0.01 121.5 F55 122.0 0.00 30 35 43 48 54 60 65 70 77 82 87 93 98 104 109 114 120 125 130 135 140 8.70 8.65 8.60 8.55 8.50 8.45 8.40 8.35 H (ppm) BTN3A1 IgV residue number S70 S71 S71 S70 V139 Q74 Q74 H53 V33 V76 V33 L54 V68 V68 Y127 Y127 K136 K136 E135 E135 Y134 Y134 D132 D132 BTN3A1 IgV Figure 5. Cell-Surface Association of BTN2A1 and BTN3A1 Proteins (A) Anti-BTN2A1-HA immunoprecipitation, combined with anti-BTN3A1-FLAG western blot detection, following cell-surface cross-linking of CHO cells expressing BTN2A1-HA, FLAG-BTN3A1, or both. (B) Anti-BTN3A1 IP (20.1 mAb) of the same lysate combined with anti-BTN3A1-FLAG detection. For (A) and (B), likely monomeric or oligomeric species corresponding to appropriate molecular weight bands are indicated on the right-hand side. Data are representative of four independent experiments. 1 15 (C) NMR chemical-shift perturbations (CSPs) in selected residues in H- N-labeled BTN3A1 IgV (100 mM) following addition of BTN2A1 IgV (100 mM). (D) Graph of chemical shift versus residue number in BTN3A1 IgV. Threshold levels for signiﬁcant CSPs are indicated by horizontal lines. (E) Mapping of residues whose amide resonances undergo CSPs on the surface of BTN3A1 IgV domain, showing clustering in the CFG face of the domain. Residues are colored in relation to the size of their CSPs, using the thresholds indicated in (D). See also Figure S5. 494 Immunity 52, 487–498, March 17, 2020 Δδ HN (ppm) ave BTN2A1 monomer and BTN3A1 monomer (125 kDa expected ﬁed BTN2A1 as a critical mediator of P-Ag sensing and a direct MW). A second represented a considerably larger cross-linked ligand for Vg9 TCRs using a TCR-tetramer staining and species (exceeding the weight of the 180 kDa marker), most CRISPR-screen approach (Rigau et al., 2020). likely equivalent to one BTN2A1 homodimer and one BTN3A1 Our results highlight the potential of radiation hybrids as a test homodimer (250 kDa) (Figure 5A). In addition, both bands system for identiﬁcation of genomic regions controlling cellular were also detected when the BTN3A1-speciﬁc mAb 20.1 was phenotypes and function. The relatively simple screening scheme used for the initial immunoprecipitation step (Figure 5B), for identiﬁcation of these regions by comparison of radiation including when polyclonal anti-BTN2A1 antibody was used for hybrid transcriptomes facilitates the generation of custom-made WB detection (Figure S5). Incubation of CHO transductants radiation hybrids, which is an advantage over genetically deﬁned with Zol was also used to assess if the presence of cross-linked radiation hybrid panels (Ross, 2001). Moreover, enabling rodent BTN2A1-BTN3A1 species was dependent upon P-Ag levels (Fig- cells with capacity for P-Ag sensitization will not only help to un- ures 5A, 5B, and S5A). Of note, both higher-MW BTN2A1- derstand Vg9Vd2 T cell function in vitro but also aid in establishing BTN3A1 bands were detected in the presence and absence of much-needed small animal models for the study of P-Ag-reactive Zol (Figures 5A, 5B, and S5A), indicating BTN2A1 association cells. The De Libero group had shown (Kistowska, 2007)that with BTN3A1 occurs constitutively. Vg9Vd2 TCR transgenic mouse cells exhibit a block in thymic To investigate whether BTN2A1-BTN3A1 association involved maturation, which can be overcome by administration of anti- 1 15 IgV-IgV domain interactions, we expressed and puriﬁed H- N- CD3 antibody, suggesting a positive selection signal provided 1 15 labeled BTN3A1 IgV in E. coli and performed H- Nheteronuclear by species-speciﬁc molecules. We hypothesize that BTN2A1 single quantum coherence (HSQC) spectroscopy in the absence and/or BTN3A1 are such molecules and aim to test whether in vivo and presence of an equimolar amount of naturally labeled expression of BTN2A1 and/or BTN3A1 enables Vg9Vd2 T cell E. coli-expressed BTN2A1 IgV (Figures 5C–5E and S5B). Analyses maturation. If established, such a model would allow the determi- were facilitated by our previous assignment of all amide residues nants controlling gd T cell responses and functionality in the of the BTN3A1 IgV domain (Salim et al., 2017) and demonstrated emerging Vg9Vd2 T cell compartment to be studied and impor- small but signiﬁcant chemical-shift perturbations (CSPs) in tantly would allow for development of small animal models for har- numerous BTN3A1 residues in the presence of BTN2A1 IgV (Fig- nessing Vg9Vd2 T cells in pathological conditions such as cancer ures 5C and 5D), indicating direct interaction. Mapping these and infections with Vg9Vd2 T cell activating pathogens. CSPs onto the BTN3A1 IgV domain structure indicated the major- Establishment of direct binding experiments enabled us to ity of these residues clustered around the CFG face of the BTN3A1 probe the interaction mode of BTN2A1 with the Vg9Vd2TCR. domain (Figure 5E). Importantly, this region of BTN3A1 or BTN3A2 Our ﬁnding that TCR binding to BTN2A1 is solely dependent (which share an identical IgV domain) has been highlighted as crit- upon the TCR Vg9 chain is entirely consistent with the ﬁnding ical for P-Ag sensitization (Willcox et al., 2019), including specif- that Vd2 T cells expressing alternative non-Vg9Vg regions are ically Y127, K136, and R73; notably, we show that both Y127 both insensitive to P-Ag and also adopt an adaptive-like biology and K136 displayed detectable chemical shifts upon BTN2A1 fundamentally distinct from the innate-like features of Vg9Vd2 binding, as did S70, S71, and Q74. In contrast, no CSPs were de- Tcells (Davey et al., 2018). However, it was initially surprising, 1 15 tected in similar experiments using H- N-labeled BTN3A1 given that previous studies have highlighted the importance of following addition of a control BTN family IgV domain (unlabeled multiple CDRs of both Vg9and Vd2 TCR chains (including BTNL3; Willcox et al., 2019)(Figures S5C and S5D). These results CDR3g and CDR3d)to Vg9Vd2-mediated P-Ag sensing (Wang indicate that IgV-IgV domain interactions involving the CFG face of et al., 2010); moreover, our mutagenesis studies provided addi- BTN3A1 or BTN3A2 contribute to BTN2A1-BTN3A1 association. tional conﬁrmation of the importance of CDR3d and CDR2d resi- dues for BTN2A1-stimulated P-Ag sensing. Furthermore, the DISCUSSION exclusive focus of BTN2A1 on Vg9 raised the question of whether the Vg9Vd2-BTN2A1 interaction mode was related to that of intes- Here, we identiﬁed BTN2A1 as Factor X via a radiation hybrid tinal Vg4 T cell recognition of BTNL3.8 (Willcox et al., 2019), which involves germline-encoded CDR2 and HV4 regions of the Vg4 approach that highlighted a critical 580 kb region of Chr 6, in TCR chain and the CFG face of the BTNL3 IgV domain. Modeling which we probed candidate genes that were retained in species that bear Vg9Vd2 T cells but were missing or non-conserved in and mutagenesis approaches conﬁrmed a fundamental similarity mouse. BTN2A1-transduced rodent cells were speciﬁcally with this binding mode, indicating both CDR2 and HV4 regions of stained by Vg9Vd2 TCR tetramers, and crucially, BTN2A1 ex- the Vg9 TCR chain, and residues in the CFG face of BTN2A1 were hibited strong functional synergy with BTN3A1, restoring P-Ag critical for recognition. These results highlight clear evolutionary sensing following co-transduction into mouse cells. Moreover, us- conservation of the ‘‘superantigen-like’’ BTN or BTNL-gd TCR ing SPR, we were able to demonstrate speciﬁc binding of BTN2A1 interaction mode across different anatomical sites, which will no IgV domain to Vg9Vd2 TCRs. In addition, target cells transduced doubt be elucidated further by future structural analyses. with BTN2A1 molecules incorporating single amino acid muta- The oligomerisation state and interaction partners of BTN2A1 tions that eliminated Vg9Vd2 TCR binding in SPR experiments on the target cell surface are likely important factors in its mode failed to stimulate P-Ag-speciﬁc effector responses in Vg9Vd2 of action. Speciﬁcally, we show that cell surface BTN2A1 is T cells. These ﬁndings not only establish BTN2A1 as the putative comprised predominantly of homodimers in transduced rodent Factor X co-factor in P-Ag sensing but also highlight that its role as cells and 293T cells. This is consistent with structural work that adirect ligand for theVg9Vd2 TCR is essential to its ability to highlighted the potential of BTN3A1 to form IgC-IgC homo- potentiate P-Ag sensing. Of note, Rigau et al. recently also identi- dimers (Palakodeti et al., 2012), and indeed, our modeling Immunity 52, 487–498, March 17, 2020 495 studies conﬁrmed that BTN2A1 homodimers are likely to form ing site (Sandstrom et al., 2014; Wang et al., 2013). Importantly, highly equivalent IgC-IgC interactions. However, our results our results do not exclude the possibility of direct TCR-BTN3A1 highlight the potential of BTN2A1 homodimer stabilization via interactions, and indeed, recent mutagenesis of BTN3A1 and an interchain disulphide-linkage involving a membrane-proximal BTN3A2 (Willcox et al., 2019) could be interpreted as support- cysteine residue absent in BTN2A2 and BTN3 molecules. Com- ing Vg9-BTN3A1 or BTN3A2 binding using a mode similar to bined with our successful production of BTN2A1 IgV as a soluble both Vg4-BTNL3 and Vg9-BTN2A1. Nor can we discount functional monomeric domain, this suggests that BTN2A1 may the possibility of parallel and/or sequential interaction of Vg9 form a ‘‘Y-shaped’’ dimer analogous to that proposed for TCR with BTN2A1 or BTN3A1 IgV domains. However, the un- BTN3A1 homo- and heterodimers (Palakodeti et al., 2012; Van- equivocal demonstration by our study and by Rigau et al. tourout et al., 2018) and BTNL3.8 and BTNL1.6 heterodimers (2020) of direct Vg9-BTN2A1 interaction, combined with (Melandri et al., 2018; Willcox et al., 2019). Nevertheless, an lack of any compelling evidence for direct TCR-BTN3A1 or important caveat is that although our studies suggest BTN2A1 TCR-BTN3A2 interaction and ﬁnally our detection of direct preferentially homodimerizes even when co-expressed along- BTN2A1-BTN3A1 complexes in this study, point to alternative side BTN3A1, conservation of residues at the IgC interface possibilities. Taken together, these observations strongly sug- means we cannot exclude non-disulphide-stabilized IgC-IgC- gest a composite ligand model of Vg9Vd2 recognition involving mediated heterodimeric interactions with other BTN molecules, coordinate Vg9-germline-mediated interaction with BTN2A1 including potentially BTN2A2 or, alternatively, BTN3A2 or alongside a CDR3-mediated interaction with a separate BTN3A3 (Vantourout et al., 2018). ligand(s). The identity of such a TCR ligand and its potential Importantly, we also establish a close association between association partners at the cell surface is currently a focus of BTN2A1 and BTN3A1 on target cells. While these results are investigation. One possibility that cannot be excluded is that consistent with those of Rigau et al., who determined co-localiza- BTN3A1 is itself recognized in complex with BTN2A1 following tion of BTN2A1 and BTN3A1 to within the 10-nm resolution limit of P-Ag exposure, although our demonstration of constitutive FRET detection (Rigau et al., 2020), our immunoprecipitation BTN2A1-BTN3A1 association might argue against this; given approach employed a membrane-impermeable cross-linker that BTN2A1-BTN3A1 complexes occur in the absence of featuring a 16-A spacer arm, thereby suggestive of a close, P-Ag, this would still require an inside-out mechanism to possibly direct association at the cell surface. While importantly conﬁgure complexes for productive TCR-mediated recogni- these experiments deﬁned discrete, higher-MW species incorpo- tion. Alternatively, it is tempting to speculate that BTN3A1 rating both BTN2A1 and BTN3A1, the requirement for chemical could function (potentially with facilitation by BTN3A2 or cross-linking to immunoprecipitate such complexes suggests BTN3A3) to chaperone a critical additional Vg9Vd2 TCR ligand the association is likely of relatively low afﬁnity. Consistent with to the surface to be coordinately recognized as part of a a direct interaction, our nuclear magnetic resonance (NMR) BTN3A1-ligand complex in a CDR3-mediated fashion along- studies indicate that IgV-IgV interactions likely contribute to this side BTN2A1. One prediction of this model is that such a cis-BTN2A1-BTN3A1 association and corroborate recent muta- CDR3-recognized ligand(s) is likely to be highly conserved be- genesis results (Willcox et al., 2019) that highlighted a critical tween humans and rodents. Of note, instead of invoking direct role for residues on the CFG face of BTN3A1 or BTN3A2 IgV in BTN3A1 IgV interaction with the Vg9Vd2 TCR, this second P-Ag sensing. An interesting precedent for involvement of this model proposes BTN3A1 association with BTN2A1 as a mech- CFG face in IgSF interactions in cis is provided by a recent study anism of recruiting another ligand to the complex following by Chaudhri and colleagues, who showed that PD-L1 may simi- P-Ag exposure; this would allow the Vg9Vd2 T cell compart- larly utilize the CFG face of its IgV domain to mediate cis-interac- ment to continually survey BTN3A1-BTN2A1 complexes for tions with B7.1 (Chaudhri et al., 2018). Notably, similar immuno- the presence of a P-Ag-regulated ligand. In this context, along- precipitation results were obtained in the presence or absence side Vg9 interaction with BTN2A1, such BTN3A1-BTN2A1 of Zol, which stimulates P-Ag accumulation in target cells, indi- interactions most likely serve to spatially orientate BTN3A cating that while likely essential for P-Ag sensing, BTN2A1- homo- or heterodimers and, following P-Ag exposure, an asso- BTN3A1 association per se may not provide the critical molecular ciated ligand, appropriately for TCR CDR3-mediated recogni- signal for Vg9Vd2 activation. Of relevance, the potential for tion. In this second composite ligand model, P-Ag binding to BTN3A1 to heterodimerize with BTN3A2 or BTN3A3 and promote BTN3A1 B30.2 could regulate the strength of BTN3 association optimal P-Ag sensing (Vantourout et al., 2018), the identical IgV with such a ligand, and/or trafﬁcking of such complexes to the domain sequences of BTN3A1 and BTN3A2, and recent mutagen- cell surface. Further studies are required to clarify such mech- esis results on BTN3A1 and BTN3A2 (Willcox et al., 2019)are anistic models, deﬁne structural features of the key interac- important considerations in interpreting these results and suggest tions, address issues such as the role of BTN2A1 B30.2 that BTN2A1 IgV interactions with BTN3 could involve IgV do- domain, and establish relevance to other gd T cell subsets, mains of different BTN3 family members. including the intestinal Vg4 compartment. Collectively, our results and those of Rigau et al. (2020) revise In summary, we show that by acting as a direct ligand for the current models of P-Ag sensing, many of which have previously Vg9Vd2 TCR, BTN2A1 powerfully synergizes with BTN3A1 to focused on BTN3A1 as a ‘‘lone TCR ligand,’’ and proposed potentiate P-Ag sensing. Vg9Vd2 T cells have to date been the either direct presentation of P-Ag by the IgV domain (Vavassori primary focus of therapeutic development for gd T cells. Under- et al., 2013) or ‘‘inside-out’’ models whereby binding of P-Ag to standing their mode of action should facilitate attempts to the intracellular BTN3A1 B30.2 domain is transmitted in some harness them therapeutically for either cell therapy or small way to the extracellular region of BTN3A1, creating a TCR bind- molecule approaches. 496 Immunity 52, 487–498, March 17, 2020 T.J.K., B.K., L.W., M.J., F.M., V.P., J.D.-M., R.A.G.C., and P.A.B.; Writing – STAR+METHODS Original Draft, B.E.W. and T.H.; Writing – Review & Editing, B.E.W., T.H., M.M.K., and C.R.W.; Visualization, M.M.K., C.R.W., T.H., B.E.W., and F.M.; Detailed methods are provided in the online version of this paper Funding Acquisition, T.H. and B.E.W.; Supervision, T.H., B.E.W., C.R.W., and include the following: and M.J. d KEY RESOURCES TABLE DECLARATION OF INTERESTS d LEAD CONTACT AND MATERIALS AVAILABILITY There are no competing interests to declare. d EXPERIMENTAL MODEL AND SUBJECT DETAILS d METHOD DETAILS Received: January 20, 2020 B Generation of radiation hybrids Revised: February 18, 2020 B RNAseq analysis of Radiation Hybrids Accepted: February 24, 2020 B In vitro stimulation with human Vg9Vd2 TCR trans- Published: March 9, 2020 ductants REFERENCES B Expansion of primary human polyclonal Vg9Vd2 T cells B Generation of 293T BTN2 cell lines Boyden, L.M., Lewis, J.M., Barbee, S.D., Bas, A., Girardi, M., Hayday, A.C., B Human IFNg assay Tigelaar, R.E., and Lifton, R.P. (2008). Skint1, the prototype of a newly identi- B Cloning and expression of BTN2A1, BTN2A2 or ﬁed immunoglobulin superfamily gene cluster, positively selects epidermal mutants gammadelta T cells. Nat. Genet. 40, 656–662. B Generation of FLAG/HA tagged BTN3A1 and BTN2A1 Chaudhri, A., Xiao, Y., Klee, A.N., Wang, X., Zhu, B., and Freeman, G.J. (2018). B Generation of Vg9Vd2 TCR (MOP) and mutant TCR PD-L1 Binds to B7-1 Only In Cis on the Same Cell Surface. Cancer Immunol. chains Res. 6, 921–929. B Soluble protein production Davey, M.S., Willcox, C.R., Hunter, S., Kasatskaya, S.A., Remmerswaal, E.B.M., Salim, M., Mohammed, F., Bemelman, F.J., Chudakov, D.M., Oo, B Flow cytometry/TCR tetramer staining Y.H., and Willcox, B.E. (2018). The human Vd2 T-cell compartment comprises B Surface plasmon resonance + - distinct innate-like Vg9 and adaptive Vg9 subsets. Nat. Commun. 9, 1760. B Immunoprecipitation, surface biotinylation, and cross- Davodeau, F., Peyrat, M.A., Hallet, M.M., Gaschet, J., Houde, I., Vivien, R., Vie, linking H., and Bonneville, M. (1993). Close correlation between Daudi and mycobac- B I-TASSER modeling of BTN2A1 and BTN2A2 ecto- terial antigen recognition by human gamma delta T cells and expression of domains V9JPC1 gamma/V2DJC delta-encoded T cell receptors. J. Immunol. 151, B Modeling the BTN2A1-IgV/Vg9 complex 1214–1223. B Analysis of structural modeling and mutagenesis data Delfau, M.H., Hance, A.J., Lecossier, D., Vilmer, E., and Grandchamp, B. B NMR (1992). Restricted diversity of V gamma 9-JP rearrangements in unstimulated B Software human gamma/delta T lymphocytes. Eur. J. Immunol. 22, 2437–2443. d QUANTIFICATION AND STATISTICAL ANALYSIS Di Marco Barros, R., Roberts, N.A., Dart, R.J., Vantourout, P., Jandke, A., Nussbaumer, O., Deban, L., Cipolat, S., Hart, R., Iannitto, M.L., et al. (2016). B Statistical analyses Epithelia Use Butyrophilin-like Molecules to Shape Organ-Speciﬁc gd T Cell d DATA AND CODE AVAILABILITY Compartments. Cell 167, 203–218. Fichtner, A.S., Karunakaran, M.M., Starick, L., Truman, R.W., and Herrmann, SUPPLEMENTAL INFORMATION T. (2018). The Armadillo (Dasypus novemcinctus): A Witness but Not a Functional Example for the Emergence of the Butyrophilin 3/Vg9Vd2 System Supplemental Information can be found online at https://doi.org/10.1016/j. in Placental Mammals. Front. Immunol. 9, 265. immuni.2020.02.014. Fichtner, A.S., Karunakaran, M.M., Gu, S., Boughter, C.T., Borowska, M.T., Starick, L., Noehren, A., Goebel, T.W., Adams, E.J., and Herrmann, T. ACKNOWLEDGMENTS (2020). Alpaca (Vicugna pacos), the ﬁrst non-primate species with a phos- phoantigen-reactive V-gamma-9 V-delta-2 T cell subset. Proc. Nat. Acad. We thank the University of Birmingham Protein Expression Facility for use of their Sci. USA. https://doi.org/10.1073/pnas.1909474117. equipment and Aravindan Viswanathan for help with generation of BTN2A1 and Gober, H.J., Kistowska, M., Angman, L., Jeno¨ , P., Mori, L., and De Libero, G. BTN2A2 CRISPR cell lines. This work was supported by the Wellcome Trust, (2003). Human T cell receptor gammadelta cells recognize endogenous me- United Kingdom (grants 099266/Z/12/Z and 099266/Z/12/A to B.E.W. support- valonate metabolites in tumor cells. J. Exp. Med. 197, 163–168. ing C.R.W., M.S., F.M., and C.B.); Deutsche Krebshilfe, Germany to T.H. (grant 70112079) and V.K. (grant 70112081) supporting D.P. and M.M.K.; Wilhelm– Halary, F., Pitard, V., Dlubek, D., Krzysiek, R., de la Salle, H., Merville, P., Sanderstiftung, Germany, grant 2013.907.2 to T.H. supporting M.M.K.; Deut- Dromer, C., Emilie, D., Moreau, J.F., and De´ chanet-Merville, J. (2005). sche Forschungsgemeinschaft grant HE2346/8-1 to T.H. supporting A.F., Shared reactivity of Vdelta2(neg) gammadelta T cells against cytomegalo- D.P., and M.M.K.; and SIRIC BRIO and the Ligue Nationale contre le Cancer, virus-infected cells and tumor intestinal epithelial cells. J. Exp. Med. 201, France (to J.D.-M. supporting V.P.). NMR studies were supported in part by 1567–1578. the Wellcome Trust, United Kingdom (grant 208400/Z/17/Z to University of Bir- Harly, C., Guillaume, Y., Nedellec, S., Peigne´ , C.M., Mo¨ nkko¨ nen, H., mingham), and we thank HWB-NMR staff at the University of Birmingham for ¨ ¨ Monkkonen, J., Li, J., Kuball, J., Adams, E.J., Netzer, S., et al. (2012). Key providing open access to their Wellcome Trust-funded spectrometers. We implication of CD277/butyrophilin-3 (BTN3A) in cellular stress sensing by a gratefully acknowledge Adrian Hayday, Pierre Vantourout, and Fedor Berditch- major human gd T-cell subset. Blood 120, 2269–2279. evski for useful discussions. Karunakaran, M.M., and Herrmann, T. (2014). The Vg9Vd2 T Cell Antigen Receptor and Butyrophilin-3 A1: Models of Interaction, the Possibility of Co- AUTHOR CONTRIBUTIONS Evolution, and the Case of Dendritic Epidermal T Cells. Front. Immunol. 5, 648. Conceptualization, T.H., M.M.K., C.R.W., and B.E.W.; Investigation, M.M.K., Karunakaran, M.M., Go¨ bel, T.W., Starick, L., Walter, L., and Herrmann, T. C.R.W., M.S., D.P., A.S.F., A. Noll, A. No¨ hren, L.S., V.K., C.R.B., K.A.B., (2014). Vg9 and Vd2 T cell antigen receptor genes and butyrophilin 3 (BTN3) Immunity 52, 487–498, March 17, 2020 497 emerged with placental mammals and are concomitantly preserved in Shen, L., Frencher, J., Huang, D., Wang, W., Yang, E., Chen, C.Y., Zhang, Z., selected species like alpaca (Vicugna pacos). Immunogenetics 66, 243–254. Wang, R., Qaqish, A., Larsen, M.H., et al. (2019). Immunization of Vg2Vd2 T cells programs sustained effector memory responses that control tubercu- Kistowska, M. (2007). Antigen Recognition and Thymic Maturation of Human losis in nonhuman primates. Proc. Natl. Acad. Sci. USA 116, 6371–6378. TCR Vg9-Vd2 Cells (Basel, Switzerland: PhD Basel University), p. 202. Kunzmann, V., Bauer, E., Feurle, J., Weissinger, F., Tony, H.P., and Wilhelm, Silva-Santos, B., Mensurado, S., and Coffelt, S.B. (2019). gd T cells: pleiotropic M. (2000). Stimulation of gammadelta T cells by aminobisphosphonates and immune effectors with therapeutic potential in cancer. Nat. Rev. Cancer 19, induction of antiplasma cell activity in multiple myeloma. Blood 96, 384–392. 392–404. Li, J.-Q. (2010). Modulating the expression of enzymes of isoprenoid synthe- Smedley, D., Haider, S., Durinck, S., Pandini, L., Provero, P., Allen, J., Arnaiz, sis: effects on Vgamma9Vdelta2 T cell activation and tumor cell growth. O., Awedh, M.H., Baldock, R., Barbiera, G., et al. (2015). The BioMart commu- https://nbn-resolving.org/urn:nbn:de:bvb:20-opus-46388. nity portal: an innovative alternative to large, centralized data repositories. Nucleic Acids Res. 43, W589–W598. Melandri, D., Zlatareva, I., Chaleil, R.A.G., Dart, R.J., Chancellor, A., Nussbaumer, O., Polyakova, O., Roberts, N.A., Wesch, D., Kabelitz, D., Soneoka, Y., Cannon, P.M., Ramsdale, E.E., Grifﬁths, J.C., Romano, G., et al. (2018). The gdTCR combines innate immunity with adaptive immunity Kingsman, S.M., and Kingsman, A.J. (1995). A transient three-plasmid expres- by utilizing spatially distinct regions for agonist selection and antigen respon- sion system for the production of high titer retroviral vectors. Nucleic Acids siveness. Nat. Immunol. 19, 1352–1365. Res. 23, 628–633. Morita, C.T., Beckman, E.M., Bukowski, J.F., Tanaka, Y., Band, H., Bloom, Starick, L., Riano, F., Karunakaran, M.M., Kunzmann, V., Li, J., Kreiss, M., B.R., Golan, D.E., and Brenner, M.B. (1995). Direct presentation of nonpeptide Amslinger, S., Scotet, E., Olive, D., De Libero, G., and Herrmann, T. (2017). prenyl pyrophosphate antigens to human gamma delta T cells. Immunity 3, Butyrophilin 3A (BTN3A, CD277)-speciﬁc antibody 20.1 differentially activates 495–507. Vg9Vd2 TCR clonotypes and interferes with phosphoantigen activation. Eur. J. Morita, C.T., Jin, C., Sarikonda, G., and Wang, H. (2007). Nonpeptide antigens, Immunol. 47, 982–992. presentation mechanisms, and immunological memory of human Vantourout, P., Laing, A., Woodward, M.J., Zlatareva, I., Apolonia, L., Jones, Vgamma2Vdelta2 T cells: discriminating friend from foe through the recogni- A.W., Snijders, A.P., Malim, M.H., and Hayday, A.C. (2018). Heteromeric inter- tion of prenyl pyrophosphate antigens. Immunol. Rev. 215, 59–76. actions regulate butyrophilin (BTN) and BTN-like molecules governing gd T cell Nguyen, K., Li, J., Puthenveetil, R., Lin, X., Poe, M.M., Hsiao, C.C., biology. Proc. Natl. Acad. Sci. USA 115, 1039–1044. Vinogradova, O., and Wiemer, A.J. (2017). The butyrophilin 3A1 intracellular van Zundert, G.C.P., Rodrigues, J.P.G.L.M., Trellet, M., Schmitz, C., Kastritis, domain undergoes a conformational change involving the juxtamembrane re- P.L., Karaca, E., Melquiond, A.S.J., van Dijk, M., de Vries, S.J., and Bonvin, gion. FASEB J. 31, 4697–4706. A.M.J.J. (2016). The HADDOCK2.2 Web Server: User-Friendly Integrative Palakodeti, A., Sandstrom, A., Sundaresan, L., Harly, C., Nedellec, S., Olive, Modeling of Biomolecular Complexes. J. Mol. Biol. 428, 720–725. D., Scotet, E., Bonneville, M., and Adams, E.J. (2012). The molecular basis for modulation of human Vg9Vd2 T cell responses by CD277/butyrophilin-3 Vavassori, S., Kumar, A., Wan, G.S., Ramanjaneyulu, G.S., Cavallari, M., El (BTN3A)-speciﬁc antibodies. J. Biol. Chem. 287, 32780–32790. Daker, S., Beddoe, T., Theodossis, A., Williams, N.K., Gostick, E., et al. (2013). Butyrophilin 3A1 binds phosphorylated antigens and stimulates human Parker, C.M., Groh, V., Band, H., Porcelli, S.A., Morita, C., Fabbi, M., Glass, D., gd T cells. Nat. Immunol. 14, 908–916. Strominger, J.L., and Brenner, M.B. (1990). Evidence for extrathymic changes in the T cell receptor gamma/delta repertoire. J. Exp. Med. 171, 1597–1612. Vranken, W.F., Boucher, W., Stevens, T.J., Fogh, R.H., Pajon, A., Llinas, M., Rhodes, D.A., Reith, W., and Trowsdale, J. (2016). Regulation of Immunity by Ulrich, E.L., Markley, J.L., Ionides, J., and Laue, E.D. (2005). The CCPN data Butyrophilins. Annu. Rev. Immunol. 34, 151–172. model for NMR spectroscopy: development of a software pipeline. Proteins 59, 687–696. Rian˜ o, F., Karunakaran, M.M., Starick, L., Li, J., Scholz, C.J., Kunzmann, V., Olive, D., Amslinger, S., and Herrmann, T. (2014). Vg9Vd2 TCR-activation by Wang, H., Fang, Z., and Morita, C.T. (2010). Vgamma2Vdelta2 T Cell Receptor phosphorylated antigens requires butyrophilin 3 A1 (BTN3A1) and additional recognition of prenyl pyrophosphates is dependent on all CDRs. J. Immunol. genes on human chromosome 6. Eur. J. Immunol. 44, 2571–2576. 184, 6209–6222. Rigau, M., Ostrouska, S., Fulford, T.S., Johnson, D.N., Woods, K., Ruan, Z., Wang, H., Henry, O., Distefano, M.D., Wang, Y.C., Raikko¨ nen, J., Mo¨ nkko¨ nen, McWilliam, H.E.G., Hudson, C., Tutuka, C., Wheatley, A.K., et al. (2020). J., Tanaka, Y., and Morita, C.T. (2013). Butyrophilin 3A1 plays an essential role Butyrophilin 2A1 is essential for phosphoantigen reactivity by gd T cells. in prenyl pyrophosphate stimulation of human Vg2Vd2 T cells. J. Immunol. 191, Science 367. 1029–1042. Ross, S.R. (2001). Commentary: phenotypic screening of radiation hybrid Willcox, B.E., Gao, G.F., Wyer, J.R., Ladbury, J.E., Bell, J.I., Jakobsen, B.K., panels. Mamm. Genome 12, 879–881. and van der Merwe, P.A. (1999). TCR binding to peptide-MHC stabilizes a ﬂex- Salim, M., Knowles, T.J., Baker, A.T., Davey, M.S., Jeeves, M., Sridhar, P., ible recognition interface. Immunity 10, 357–365. Wilkie, J., Willcox, C.R., Kadri, H., Taher, T.E., et al. (2017). BTN3A1 Willcox, C.R., Pitard, V., Netzer, S., Couzi, L., Salim, M., Silberzahn, T., Discriminates gd T Cell Phosphoantigens from Nonantigenic Small Moreau, J.F., Hayday, A.C., Willcox, B.E., and De´ chanet-Merville, J. (2012). Molecules via a Conformational Sensor in Its B30.2 Domain. ACS Chem. Cytomegalovirus and tumor stress surveillance by binding of a human gd T Biol. 12, 2631–2643. cell antigen receptor to endothelial protein C receptor. Nat. Immunol. 13, Sambrook, J., and Russell, D.W. (2006). Calcium-phosphate-mediated 872–879. Transfection of Eukaryotic Cells with Plasmid DNAs. CSH Protoc. 2006. Willcox, C.R., Vantourout, P., Salim, M., Zlatareva, I., Melandri, D., Zanardo, L., Sanderson, S., and Shastri, N. (1994). LacZ inducible, antigen/MHC-speciﬁc George, R., Kjaer, S., Jeeves, M., Mohammed, F., et al. (2019). Butyrophilin- T cell hybrids. Int. Immunol. 6, 369–376. like 3 Directly Binds a Human Vg4+ T Cell Receptor Using a Modality ´ ´ Sandstrom, A., Peigne, C.M., Leger, A., Crooks, J.E., Konczak, F., Gesnel, Distinct from Clonally-Restricted Antigen. Immunity 51, 813–825. M.C., Breathnach, R., Bonneville, M., Scotet, E., and Adams, E.J. (2014). The intracellular B30.2 domain of butyrophilin 3A1 binds phosphoantigens to Yang, J., and Zhang, Y. (2015). I-TASSER server: new development for protein mediate activation of human Vg9Vd2 T cells. Immunity 40, 490–500. structure and function predictions. Nucleic Acides Res. 43, W174–W181. 498 Immunity 52, 487–498, March 17, 2020 STAR+METHODS KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies Anti-huBTN3 (CD277) clone 103.2 Gift from Dr. Daniel Olive N/A Anti-huBTN3 (CD277) clone 20.1 Invitrogen Cat# 14-2779-82; RRID: AB_467550 Anti-huBTN2A1 (1C7D) MBL Cat# W005-3 FITC anti-human Vd2 BD Biosciences Cat# 562088; RRID: AB_10892810 F(ab’) Donkey anti mouse IgG (H+L) R-PE Jackson Immunoresearch Cat# 715-116-151; RRID: AB_2340799 mIgG1,k isotype clone p3.6.2.81 eBiosciences Cat#16-4714-85; RRID: AB_470162 mIgG2a,k isotype clone-eBM2a eBiosciences Cat#16-4724-85; RRID: AB_470165 Anti-HA.11 epitope tag afﬁnity matrix (clone 16B12) Biolegend Cat#900801; RRID: AB_2564999 Puriﬁed anti-DYKDDDDK (FLAG) tag antibody (clone L5) Biolegend Cat#637301; RRID: AB_1134266 Anti-HA.11 epitope tag antibody, FITC labeled (Clone 16B12) Biolegend Cat#901507; RRID: AB_2565058 BTN2A1 rabbit polyclonal antibody Sigma Cat#HPA019208; RRID: AB_1845492 Goat anti-rabbit HRP ThermoFisher Cat#G21234; RRID: AB_2536530 Goat anti-rat HRP ThermoFisher Cat#A10549; RRID: AB_2534047 Puriﬁed mouse anti-human TCRg/d, clone 11F2 BD Biosciences Cat# 347900; RRID: AB_400356 Anti-Vg9 antibody, FITC (IMMU360) Beckman Coulter Cat#IM1463; RRID: AB_130871 Bacterial and Virus Strains NEB 5-alpha NEB Cat# C2987H BL21 (DE3) NEB Cat# C2527H Biological Samples BrHPP-expanded Vg9Vd2 T cells This paper N/A Chemicals, Peptides, and Recombinant Proteins HMBPP Sigma Cat#95058 Zoledronate Sigma Cat#SML0223 rhIL-2 AiCuris Ch.B.: ZA4621B/3 Phusion high ﬁdelity DNA polymerase ThermoFisher Scientiﬁc Cat#F530S IN-Fusion HD cloning Kit TAKARA Cat#639649 TOPO TA Cloning kit for sequencing Invitrogen Cat#450071 HAT Media Supplement (50x) Hybrid-Max Sigma Cat#H0262 HT Media Supplement (50x) Hybrid-Max Sigma Cat#H0137 PEG 1500 Roche Cat# 10 783 641 001 Histopaque-1077 Sigma Cat#10711 GeneArt CRISPR Nuclease (OFP reporter) Invitrogen Cat#A21174 GeneArt CRISPR Nuclease (CD4 enrichment) Invitrogen Cat#A21175 EcoRI ThermoFisher Scientiﬁc Cat#ER0271 NdeI Roche Cat# 11 040 227 001 BamHI Roche Cat# 10 567 604 001 BTN2A1 IgV This paper N/A BTN2A2 IgV This paper N/A Soluble T cell receptors (sTCRs) Willcox et al., 2012; this paper N/A Streptavidin-HRP ThermoFisher Scientiﬁc Cat#21130 Streptavidin-PE conjugate ThermoFisher Scientiﬁc Cat#S866 Streptavidin-APC ThermoFisher Scientiﬁc Cat#S868 Sulfo-EGS crosslinker ThermoFisher Scientiﬁc Cat#21566 (Continued on next page) Immunity 52, 487–498.e1–e6, March 17, 2020 e1 Continued REAGENT or RESOURCE SOURCE IDENTIFIER EZ-link Sulfo-NHS-LC biotin ThermoFisher Scientiﬁc Cat#21335 Iodoacetamide Sigma Cat#I6125 Critical Commercial Assays IL-2 mouse uncoated ELISA kit Invitrogen Cat # 88-7024-88 IFN gamma Human uncoated ELISA kit Invitrogen Cat # 88-7316-88 Deposited Data RNA-seq dataset of radiation hybrid clones ﬁltered Mendeley Data https://doi.org/10.17632/ny6bxn4y9s.1 for transcribed human genes Experimental Models: Cell Lines 293T DSMZ Cat#ACC 635; RRID: CVCL_0063 CHO (CHO-K1) ATCC Cat#CCL-61; RRID: CVCL_0214 CHO human Chromosome 6 Coriell Institute for Medical GM11580; RRID: CVCL_V287 research BW36 gal (BW) Dr. Nilabh Shastri Lab; N/A Sanderson and Shastri, 1994 53/4 hybridoma Vg9Vd2 - MOP TCR Starick et al., 2017 N/A A23 Thymidine kinase negative Hamster ﬁbroblast Dr. Carol Stocking Lab N/A (HAT sensitive) BW 58C-CD28+ Harly et al., 2012 N/A Oligonucleotides Primer and CRISPR Sequences in Tables S1 and S2 N/A N/A Recombinant DNA pMIM gift from Dario Vignali Addgene # 52114 pMIG II gift from Dario Vignali Addgene # 52107 pIZ Gift from Dr. Ingolf Berberich N/A pIH Gift from Dr. Ingolf Berberich N/A pIH-FLAG This paper N/A pET23a Merck Millipore Cat# 69745-3 pMT/BiP/V5-HisB Invitrogen Cat# V413020 BTN2A1 IgV in pET23a (wild type and mutants) This paper N/A BTN2A2 IgV in pET23a This paper N/A BTN3A1 IgV in pET23a (Salim et al., 2017) N/A Human and mouse gdTCRs in pMT/BiP/V5-HisB This paper; Willcox et al., 2012 N/A Software and Algorithms FlowJo version 10 FlowJo https://ﬂowjo.co/ PyMOL version 2.0.7 Schrodinger https://pymol.org/2/ GraphPad Prism version 8.0.2 GraphPad Software https://www.graphpad.com BIAevaluation GE Healthcare https://www.gelifesciences.com/en/ gb/shop/protein-analysis/spr-label- free-analysis Origin 2015 OriginLab https://www.originlab.com/ CRISPR design tool Invitrogen N/A ZHANG LAB Vantourout et al., 2018 https://zlab.bio/guide-design-resources CRISPR RGEN tools This paper http://www.rgenome.net/ Other Sensor Chip CM5 GE Healthcare Cat#29149604 Sensor Chip NTA GE Healthcare Cat#BR100407 HBS-P GE Healthcare Cat#BR100368 HBS-EP GE Healthcare Cat#BR100188 streptavidin Sigma Cat#S4622 e2 Immunity 52, 487–498.e1–e6, March 17, 2020 LEAD CONTACT AND MATERIALS AVAILABILITY Further information and requests for resources and reagents should be directed to and will be fulﬁlled by the Lead Contact, Benjamin E. Willcox (b.willcox@bham.ac.uk). Reagents generated in this study are available on request from the Lead Contact with a completed Materials Transfer Agreement. EXPERIMENTAL MODEL AND SUBJECT DETAILS CHO, CHO-chr6, BW, A23, 53/4 hybridoma TCR transductants, and radiation hybrids (CHO Chr6 – rodent fusion hybrids) were cultured with RPMI (GIBCO) supplemented with 10% FCS, 1 mM sodium pyruvate, 2.05 mM glutamine, 0.1 mM nonessential amino acids, 50 mM b-mercaptoethanol, penicillin (100 U/mL) and streptomycin (100 U/mL). Peripheral blood mononuclear cells isolated from healthy volunteers were also maintained as above with or without rhIL-2 (Novartis Pharma). 293T cells were maintained in DMEM (GIBCO) supplemented with 10% FCS. METHOD DETAILS Generation of radiation hybrids 6 7 CHO Chr 6 (10 or 10 cells) were irradiated at Faxitron CP160 (program 160 kV, 6.3 mA, 300 Gy: 60 min, 100 Gy: 20 min). The irra- diated cells and fusion partner (BW or A23) were mixed at 1:1 or 1:3 ratio (irradiated cell:fusion partner) and centrifuged at 461 g for 5 min at RT. The cell pellet was gently tapped and 1 mL PEG1500 was added slowly over a minute with gentle mixing in a prewarmed water-bath. After addition of PEG, cells were resuspended in 50 mL warm serum free RPMI and incubated for 30 min, followed by centrifugation at 461 g for 5 min and careful resuspension in RPMI supplemented with 10% FCS at 10 cells/mL. The cell suspension was seeded in 96 well plate ﬂat bottom (A23 fused) or round bottom (BW fused) plates in 100 ml per well. On the following day, 100 mlof 2X HAT was added and cells were selected for two weeks. The selected clones were supplemented with HT medium and further seeded at limiting dilutions to obtain single cell clones which were tested for P-Ag mediated activation of our Vg9Vd2 TCR (MOP) transductants. P-Ag presentation capable and incapable clones were PCR characterized for human Chr 6 regions with primers listed in Table S1. RNAseq analysis of Radiation Hybrids Knowing the differences in antigen presentation of the various radiation hybrid cell lines, we performed RNA seq to identify those human Chr 6-encoded genes that are expressed in each hybrid line. Cells were stored in TRIzol Reagent (Invitrogen) and total RNA was extracted. Sequencing libraries were produced with an Illumina Truseq RNA preparation kit as described by the supplier’s protocol and were sequenced with an Illumina HiSeq4000. Sequence reads were mapped to the human genome (hg38) with STAR (version STAR_2.50a) and read counts of gene transcripts were determined using gtf ﬁle Homo_sapiens.GRCH38.84.gtf and featur- eCount (v1.5.0-p1). Cell lines were then compared for presence, i.e., expression, of human Chr 6 genes. To ﬁlter out reads descend- ing from mouse or hamster cells, all fastq-ﬁles were initially mapped against the mouse genome (Mus musculus, version GRCm38) using STAR and the corresponding Gene Transfer Format (gtf) ﬁle (version 87). Unmapped reads and those exhibiting more than two mismatches were selected and mapped against the Chinese hamster genome (Cricelulus griseus, version 1). The corresponding gtf- ﬁle was downloaded from the pre-Ensembl ftp site (Cricetulus_griseus.CriGri_1.pre.gtf). Afterward, all unmapped reads and those containing more than two mismatches were again selected to ﬁnally map against the human genome (version hg38; gtf-ﬁle version 84). Only reads showing maximally one mismatch were considered as true. With the help of featureCounts, mapped reads were assigned to genomic features using the above mentioned gtf-ﬁles. The results were summarized within an Excel-ﬁle. Further gene information were extracted from BioMart (Ensembl Genes 84; (Smedley et al., 2015). In vitro stimulation with human Vg9Vd2 TCR transductants For in vitro stimulations, 10 CHO or 293T cells were seeded on day 1 with 50 mL RPMI or DMEM in a 96 well ﬂat bottom cell culture plate and cultured over-night. On day 2, 50 mL of 5x10 - 53/4 hybridoma cells expressing the human MOP Vg9Vd2 TCR and 100 mLof appropriate stimulant such as HMBPP, Zol, or 20.1 mAb were added to the culture and incubated for 22 h. After overnight incubation, the activation of TCR transductants was analyzed by measurement of mouse IL-2 from the supernatants of the co-cultures by ELISA (Invitrogen) as per manufacturer’s protocol. Expansion of primary human polyclonal Vg9Vd2 T cells Fresh peripheral blood mononuclear cells (PBMCs) were isolated from healthy volunteers after obtained written informed consent in accordance with the Declaration of Helsinki and approval by the University of Wurzburg institutional review board. Whole blood was layered over the Histopaque-1077 in a 50 mL falcon tube and centrifuged at 400 g for 30 min at room temperature (RT) with no ac- celeration and brakes. After centrifugation, the opaque interface containing PBMCs were aspirated and washed twice at 461 g for 5 min at RT. Vg9Vd2 T cells were expanded by cultivation of PBMCs with RPMI containing 10% FACS, 1 mM BrHPP and recombinant Immunity 52, 487–498.e1–e6, March 17, 2020 e3 6 human IL-2 100 IU/mL (Novartis Pharma) in 10 cells/mL density in a 96 well U bottom plate for 10 days with 100 mL per well. After 10 days, cells were pooled and washed twice and cultivated to rest without rhIL- 2 at 10 cells/mL density in a 6-well plate. After three days, rested cells were subjected for further experiments. –/– Generation of 293T BTN2 cell lines BTN2A1 and BTN2A2 genes were disrupted in 293T cells using CRISPR. The CRISPR sequencing targeting functional BTN2A genes were designed with the help of online tools mentioned in the table (software section) and sequences were cloned into GeneArt CRISPR Nuclease vector as per manufacturer’s instructions. On day1, 1.5 3 10 293T cells were seeded in a 6 cm cell culture plate with DMEM medium without pyruvate (10% FCS). On day 2, cells were transfected with 5 mg of BTN2A-IgV_CRISPR cloned GeneArt CRISPR Nuclease (CD4 enrichment) Vector or BTN2A1_49FCRISPR cloned GeneArt CRISPR Nuclease (OFP Reporter) Vector or BTN2A2_343CRISPR cloned GeneArt CRISPR Nuclease (CD4 enrichment) Vector in a calcium-phosphate dependent method (Sam- brook and Russell, 2006) (CRISPR sequences are provided in Table S2). 48 h post transfection, the highest reporter expressing (top 3%) cells were sorted and seeded at 1 cell/200 mL medium/well in 96 well plate ﬂat bottom cell culture plate and cultivated till single cell derived clones were visible. Such clones were tested for their capacity to stimulate our 53/4 hybridoma human Vg9Vd2 TCR (MOP) TCR transductants in the presence of 1 mM HMBPP. The clones which exhibited loss of function were subjected to DNA isola- tion, followed by PCR for the ampliﬁcation of genetic loci targeted by CRISPR sequences with appropriate genomic primers (Table S2) complementary to ﬂanking regions of CRISPR target site. The PCR products were cloned into TOPO-TA vector (Invitrogen) and the TOPO-TA clones were analyzed by sequencing for the presence of in/del mutations resulting in loss of gene mutation as shown below. 293T BTN2 cell line harbor BTN2A1 alleles with 10 and 16 nucleotide deletion, BTN2A2 alleles with 1 and 10 nucleotide deletions; 293T BTN2A1 harbors BTN2A1 alleles with 1 nucleotide addition and 10 nucleotide deletion; 293T BTN2A2/ harbors BTN2A2 alleles with 1 nucleotide deletion. CRISPR target sites and allelic phenotypes 1a) BTN2/ allelic phenotype BTN2A1 IgV CRISPR target site 2A1IgV GCAGTGTTTGTGTATAAAGGTGGCAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 55 2A1allele1 GCAGTGTTT——————————-GTGGCAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 45 2A1allele2 GCA————————————————-GTGGCAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 39 *** ************************************ 1b) BTN2A2 IgV CRISPR target site 2A2IgV GCAGTGTTTGTGTATAAGGGTGGGAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 55 2A2allele1 GCAGTGTTTGTGTATA-GGGTGGGAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 54 2A2allele2 GCAGTGTTTG——————————TGGGAGAGAGAGAACAGAGGAGCAGATGGAGGAGT 45 ********** *********************************** 2) BTN2A1/ allelic phenotype 2A1-49F GGACTAGGCTCTAAGCCCCTCATTTCAATGAGGGGCCATGAA-GACGGGGGCATCCGGC 59 Allele1 GGACTAGGCTCTAAGCCCCTCATTTCAATGAGGGGCCATGAAGGACGGGGGCATCCGGC 59 Allele2 GGACTAGGCTCTAAGCCCCTCATTTCAATGAGGGG——————————————GGCATCCGGC 45 *********************************** ********** 3) BTN2A2/ allelic phenotype 2A2 CAAGAAGGCAGGTCCTACGATGAGGCCATCCTACGCC-TCGTGGTGGCA 48 Allele CAAGAAGGCAGGTCCTACGATGAGGCCATCCTACGCCCTCGTGGTGGCA 49 ************************************* *********** Human IFNg assay / 4 293T or 293T BTN2 cells were seeded overnight in triplicates at 23 10 cells/well in 100 mL DMEM containing 10% FCS in a 96 well ﬂat bottom cell culture plate with or without 25 mM Zol. Next day, DMEM with/without Zol was aspirated and cells were washed twice with PBS at RT and 2 3 10 expanded Vg9Vd2 T cells/well in 100mL RPMI was added and cocultured for 4 h. After 4 h, supernatants were collected and frozen at 20 C, until cytokine assay was performed with human IFNg ELISA kit (Invitrogen) was used according to the manufacturer’s instructions. Cloning and expression of BTN2A1, BTN2A2 or mutants The full length human BTN2A1 and BTN2A2 were ampliﬁed from the cDNA obtained from 293T cells with the help of pMIM-BTN2A1/ 2Fwd and pMIM-BTN2A1/2Rev primers (see Table S2) using Phusion DNA polymerase (Thermo Scientiﬁc). The ampliﬁed BTN2A1 or BTN2A2 PCR products were cloned into EcoRI & BamHI digested pMIM or pMIG II vector via In-Fusion HD cloning kit (Takara). BTN2A1 mutants were generated by fusion of two PCR products obtaining from pMIM-BTN2A1Fwd/Mutant-Rev and Mutant- Fwd/pMIM- BTN2A1/2Rev and cloned as above. Such cloned BTN2 genes and their corresponding mutants were expressed in target cells through retroviral transduction (Soneoka et al., 1995). e4 Immunity 52, 487–498.e1–e6, March 17, 2020 Generation of FLAG/HA tagged BTN3A1 and BTN2A1 For the N terminus FLAG or HA tagged proteins, FLAG or HA tag was inserted into the BamHI+EcoRI digested pIH or pIZ vector back bone with digested FLAG/HA Fwd and Rev oligonucleotide sequences with appropriate restriction sites and linker sequences ﬂanking the FLAG/HA tag sequences. BTN3A1 and BTN2A1 full length amplicons without leader peptide were ampliﬁed either cDNA or above mentioned pMIM-BTN2A1 as template. BTN2A1-HA tagged constructed ampliﬁed with pMIM-BTN2A1 as template and pIZ-BTN2A1Fwd and BTN2A1- HA-Rev primers was cloned into pIZ vector backbone via In-Fusion-HD cloning as per manufac- turer’s protocol. Generation of Vg9Vd2 TCR (MOP) and mutant TCR chains Vg9Vd2 TCR (MOP), Vd2-R51A and Vd2-CDR3 deletion mutant (DCDR3: CDR3d sequence CACD——–YTDKLIF) TCR chains were generated as reported earlier (Li, 2010; Starick et al., 2017). Vg9-E70A, -E70R and -E70K mutants were generated by fusion of two PCR products ampliﬁed by MOP- Vg9 Fwd/Mut-Rev and Mut-Fwd/ MOP-Vg9-Rev primers with pEGN-MOP-Vg9 as template using Phusion polymerase. Such generated wild type TCR chains and mutant TCR chains were expressed in 53/4 hybridoma cells by retro- viral transduction (Soneoka et al., 1995). Soluble protein production cDNA encoding wild type BTN2A1 (S27 to V142) or BTN2A2 IgV domains (S31 to V146), or BTN2A1 IgV incorporating the described mutations, were generated as gblocks (Integrated DNA Technologies) including the sequence for a C-terminal 6x Histidine tag and cloned into the pET23a expression vector (Novagen). Proteins were overexpressed, puriﬁed and refolded as described (Willcox et al., 2019). BTN2A1 and BTN2A2 IgV domains were refolded by dilution in 100 mM Tris, 400 mM L-Arginine- HCl, 2 mM EDTA, 6.8 mM cystamine, 2.7 mM cysteamine, 0.1 mM PMSF, pH 8, overnight at 4 C. The refolding mixture was concentrated and puriﬁed by size exclusion chromatography on a Superdex-200 column (GE Healthcare) pre-equilibrated with 20 mM Tris, 150 mM NaCl, pH 8, or 20 mM Na3PO4 pH 7.4 buffer, or PBS. BTN3A1 IgV was expressed, refolded, and puriﬁed as described (Salim et al., 2017). Soluble gd TCRs were generated in Drosophila S2 cells and puriﬁed by nickel chromatography as pre- viously described (Willcox et al., 2012). TCRs were then biotinylated via a C-terminal BirA tag. Flow cytometry/TCR tetramer staining Flow cytometry staining of the samples were performed with the below mentioned antibodies and samples were measured on FACSCalibur or LSRII ﬂow cytometer (BD). The expression of BTN3A1 and BTN2A1 were detected with anti-huBTN3 (CD277) clone 103.2 (gift from David Olive) and anti-huBTN2A1 clone 1C7D (MBL), followed by secondary antibody F(ab’) Donkey anti mouse IgG (H+L) R-PE (Jackson Immunoresearch). mIgG1,k isotype clone p3.6.2.81 (eBiosciences) and mIgG2a,k isotype clone-eBM2a (eBiosciences) were used a isotype controls and were detected by above mentioned secondary antibody. N-terminal HA-tagged BTN2A1 or BTN2A2 were detected using anti-HA-FITC (Biolegend). PBMC expanded human Vg9Vd2 T cells were detected with FITC- conjugated anti-human Vd2 (BD Biosciences). Biotinylated soluble Vg9Vd2 TCRs were tetramerized by the addition of Strep- tavidin-PE conjugate (ThermoFisher Scientiﬁc) at room temperature, and 1-2mg of tetramer used to stain 10 cells at 4 C. Surface plasmon resonance SPR was performed as previously described (Willcox et al., 1999) on a BIAcore3000 using streptavidin-coated CM5 chips and HBS- EP buffer (GE Healthcare). Biotinylated Vg9 TCRs, and control Vg2, or Vg4 TCRs (2000-3000 RU), were captured on the Streptavidin chip. Analyte concentrations ranged from 1-200 mM. Immunoprecipitation, surface biotinylation, and crosslinking / + 293T cells in which the BTN2A1 and BTN2A2 loci have been functionally inactivated (293T BTN2 cells), or CHO CD80 cells, were transduced to overexpress HA-tagged BTN2A1 or BTN2A2, or BTN3A1, as indicated. Cells were surface biotinylated using EZ-Link Sulfo NHS-LC-biotin (ThermoFisher, 0.8mg/mL in PBS) for 30 min on ice, quenched with 20mM Tris pH 7.5 for 5 min, washed in TBS, and lysed in lysis buffer containing 1% NP40 in 20mM Tris pH 7.5, 150mM NaCl ± 10mM iodoacetamide (Sigma). HA-tagged BTN2A1 or BTN2A2 was immunoprecipitated using anti-HA resin (BioLegend). Immunoprecipitations were washed in lysis buffer and eluted in nonreducing (NR) or reducing (R) SDS sample buffer and boiled, or incubated at 37C for 5 min before separation on 4%–20% SDS- PAGE gels (BioRad). Proteins were transferred to PVDF using the BioRad TransBlot Turbo system, blocked in 3% BSA, then incu- bated with streptavidin-HRP (Thermo). To investigate potential association of BTN2A1 and BTN3A1 at the cell surface, CHO cells overexpressing BTN2A1-HA, FLAG-BTN3A1, or both, were treated with the soluble, membrane-impermeable crosslinker sulfo-EGS (ThermoFisher) at 0.5mM in PBS, at 4 C for 2 h. Following this, the reaction was quenched by addition of Tris pH 7.5 to 20mM. Cells were washed in TBS and lysed in 1% NP40 lysis buffer. After centrifugation to remove insoluble material, immuno- precipitation was carried out using anti-HA resin or 20.1 antibody bound to protein A Sepharose (GE Healthcare). Immunoprecipitates were run on duplicate 4%–20% gels (BioRad) and blotted with anti-BTN2A1 or anti-FLAG antibodies. I-TASSER modeling of BTN2A1 and BTN2A2 ectodomains The ectodomain structures of BTN2A1 (residues Q29-A248) and BTN2A2 (residues Q33-M265), were generated using the I-TASSER (Iterative Threading ASSEmbly Reﬁnement) server (Yang and Zhang, 2015). Brieﬂy, the target sequences were initially threaded Immunity 52, 487–498.e1–e6, March 17, 2020 e5 through the Protein Data Bank (PDB) library by LOMETS2, an online meta- threading server system for template-based protein pre- diction. Continuous fragments were excised from LOMETS2 alignments and structurally reassembled by replica-exchange Monte Carlo simulations. The simulation trajectories were then grouped and used as the initial state for second round I-TASSER assembly simulations. Finally, lowest energy structural models were identiﬁed and reﬁned by fragment-guided molecular dynamic simulations to improve hydrogen-bonding contacts and omit steric clashes. Models were ranked based on their I-TASSER conﬁdence (C) score (range 5 to +2 with a higher score correlating with a higher conﬁdence model). Modeling the BTN2A1-IgV/Vg9 complex The BTN2A1-IgV/Vg9 complex was modeled with HADDOCK (van Zundert et al., 2016). BTN2A1 residues (R65, K79, R124, Y126 and E135) were classiﬁed as active in Vg9 binding based upon the results of SPR binding experiments. ‘Passively involved’ residues were selected automatically. Vg9 residues (R20, D72, E70 and E76) selected for use as ambiguous interaction restraints to drive the dock- ing process with BTN2A1 were predicted from an initial homology model (generated by superimposing BTN2A1-IgV and Vg9 onto the previously published BTNL3-IgV/Vg4 complex model (Melandri et al., 2018). Analysis of structural modeling and mutagenesis data R65 (located in the IgV domain of BTN2A1) forms a salt bridge interaction with E76 (in the Vg9 TCR chain). This interaction is likely to be abolished by introducing Ala at this position in BTN2A1 IgV domain (R65A), consistent with abrogation of binding by the R65A mutant. The hydroxyl group of S72 (BTN2A1) is in close proximity to V58 (TCR). Juxtaposition of this polar residue (S72) with a hydrophobic residue (V58) is likely to be energetically unfavorable for binding in this region. By substituting an Ala (ie a non-polar res- idue) at this position, the S72A mutation is likely to introduce hydrophobic interactions with V58, consistent with enhanced binding compared to wild-type (11-15mM (S72A) versus 50mM (Wild-type)). K79A leads to reduced binding to TCR (100mM). K79 forms a salt bridge interaction with E76 (TCR). Change to Ala will result in loss of this interaction consistent with a reduction in binding afﬁnity. The fact that binding is not totally abolished suggests that this inter- action is a not a major contributor to the binding energy. Note however that E76 also contacts R65 (see above). R124A mutation in BTN2A1 abolishes binding to TCR. The HADDOCK model suggests that R124 forms a salt bridge interaction with E70 (Vg9-IgV TCR) and a hydrogen bonding interaction with the hydroxyl group of T83. These interactions will be lost when intro- ducing an Ala at this position. Y126A abolishes binding to TCR. Y126 forms multiple hydrophobic stacking interactions with I74 (HV4 region of Vg9-IgV). In addi- tion, the hydroxyl group of Y126 forms a hydrogen bonding interaction with T77. These will be lost upon alanine substitution. Although Y133A is located at the interface with Vg9, it does not mediate interactions with Vg9 TCR residues and hence it is unsur- prising that substitution to alanine does not affect binding afﬁnity. E135 forms a salt bridge interaction with R20 (TCR). Substitution to Ala will result in a loss of this interaction, and consistent with this, E135A mutation abolishes binding to the TCR. NMR HSQC experiments were performed at 298K on 600MHz Bruker Avance III spectrometer equipped with a 5 mm TCI cryogenically 1 15 cooled triple resonance probe. Spectra were acquired using 100 mM H- N-labeled BTN3A1. Experiments were processed using Topspin 3.2 (Bruker). All analysis was performed using CCPN Analysis (Vranken et al., 2005). For analysis of BTN3A1/BTN2A1 inter- action, the ﬁnal concentration of each protein was 100mM, and a threshold of 0.015 weighted average ppm difference was used as a cut-off to identify chemical shift perturbations in BTN3A1 residues upon BTN2A1 addition. Software Structural ﬁgures were generated in PyMOL (version 2.0.7; Schrodinger, LLC). SPR data was analyzed in BIAevaluation (GE Health- care) and Origin 2015 (OriginLab). QUANTIFICATION AND STATISTICAL ANALYSIS Statistical analyses Stimulation data and transcript visualization in Figures 1 and S1A were calculated and depicted using GraphPad Prism. Differences between transduced and untransduced cells were tested using 2-way ANOVA and unpaired multiple t test using the Holm-Sidak method. DATA AND CODE AVAILABILITY The RNaseq data of radiation hybrid clones ﬁltered for transcribed human genes are available at Mendeley data https://doi.org/10. 17632/ny6bxn4y9s.1 e6 Immunity 52, 487–498.e1–e6, March 17, 2020

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Immunity Unpaywall

http://www.deepdyve.com/lp/unpaywall/butyrophilin-2a1-directly-binds-germline-encoded-regions-of-the-v-9v-2-ruEzZPeCGm

Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

Loading next page...

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher: Unpaywall
ISSN: 1074-7613
DOI: 10.1016/j.immuni.2020.02.014
Publisher site: See Article on Publisher Site

Abstract

Journal

Immunity – Unpaywall

Published: Mar 1, 2020

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

Butyrophilin-2A1 Directly Binds Germline-Encoded Regions of the Vγ9Vδ2 TCR and Is Essential for Phosphoantigen Sensing

References

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies