The authors of the present papers have previously characterized human plasmacytoid dendritic cells (pDCs) and human type 2 conventional dendritic cells (cDC2s) for their ability to attract and activate T cells in vitro and for the mobilization or de novo induction of anti-tumoral immune responses upon adoptive autologous cell transfer in melanoma patients. Human type 1 conventional dendritic cells (cDC1s) have been reported to be more efficient than other human DC types for the cross-presentation of cell-associated antigens and have been therefore proposed to be critical for anti-tumor immunity, in consistency with the experimental demonstration that this is the case for their mouse counterparts in animal preclinical models of cancer. This hypothesis is supported by recent analyses demonstrated that high expression of cDC1-specific genes in tumors is generally associated with a better clinical outcome. However, due to their rarity and frailness, studies that examined he functional properties of human cDC1s are scarce. In particular, it is not clearly established whether human cDC1s could exert tolerogenic functions under certain circumstances, and, if this is the case, whether this property could be hijacked by tumors for their own benefit to escape immune control. This hypothesis is supported by the fact that human cDC1s selectively express high levels of the gene encoding an enzyme involved in tryptophan catabolism and shown to contribute to immune tolerance / immunosuppression, namely the indoleamine 2,3-dioxygenase 1 (IDO1). However, gene and protein expression are not always correlated, and the enzymatic activity of IDO1 is controlled by post-translational mechanisms. Hence, further studies are required to evaluate the ability of cDC1s to exert immune tolerogenic / immunosuppressive functions, and the dependency of this process on IDO1 activity, as compared to pDCs and cDC2s. This is the aim of the current study.

This study confirms that human cDC1s express higher levels of the IDO1 gene than pDCs and cDC2s, both ate steady state and upon stimulation. Moreover, for the first time to the best of my knowledge, this study shows that human cDC1s indeed express higher levels of the IDO1 enzyme than pDCs and cDC2s, both ate steady state and upon stimulation, and that it is associated with a higher ability to catabolize tryptophan into its immunosuppressive product Kynurenine. In addition, this study shows induction of IDO1 expression and enzymatic function in human cDC2s but not in pDCs upon activation, although to levels remaining significantly lower than in cDC1s. The study further shows that supernatants from activated human cDC1s and cDC2s but not pDCs partly inhibit the activation of peripheral blood leukocytes upon cross-linking of CD3 and CD28, which is reversed upon pharmacological inhibit of IDO1 activity in the DCs by incubating them with the drug epacadostat. In contrast, DC activation did not decrease but on the contrary enhanced the activation of allogeneic peripheral blood leukocytes in co-cultures, and this was not significantly affected by IDO1 pharmacological inhibition. Finally, all three DC types, cDC1s, cDC2s and pDCs were able to cross-present a soluble long peptide to transgenic T cells ectopically expressing the corresponding cognate T cell receptor, without any significant effect of IDO1 pharmacological inhibition on this function.

The amount and quality of the work performed is adequate, involving several complementary readouts and techniques. The data are of adequate quality. The figures are properly designed, and sufficiently self-explanatory. The text is well written, in a sufficient precise yet concise manner. To the best of my knowledge, this paper is the first to measure IDO1 expression and activity in primary cDC1s and cDC2s and to demonstrate that it might exert immunosuppressive functions under certain conditions. Hence, the present work is original, definitely deepen our understanding of the functions of cDC1s and cDC2s, and bears important implications on whether and how harnessing these cells for cancer immunotherapy. However, a few major and minor points need to be addressed.

Major points.

1/ Necessity to include a reference positive control for IDO1 immunosuppressive activity.
To interpret the data from the current paper, it is important to include a reference positive control for IDO1 immunosuppressive activity. This is especially important under experimental conditions where no effects of epacadostat are observed, i.e. in Figures 4B and Figure 5. The best positive control would be comparison of the DC types already included in the manuscript with IDO1-expressing antigen-presenting cells that have been previously reported to inhibit T cell activation in an IDO1-dependent manner in similar experimental settings, such as mature monocyte-derived DCs (MoDCs) (Hwu P, Du MX, Lapointe R, Do M, Taylor MW, Young HA. Indoleamine 2,3-dioxygenase production by human dendritic cells results in the inhibition of T cell proliferation. J Immunol. 2000. PMID: 10725715; Munn DH, Sharma MD, Mellor AL. Ligation of B7-1/B7-2 by human CD4+ T cells triggers indoleamine 2,3-dioxygenase activity in dendritic cells. J Immunol. 2004. PMID: 15034022). Otherwise, it is difficult to assess the significance of the negative results shown. This is all the more the case for the assay of soluble antigen cross-presentation, since it is performed by measuring a different readout and on a much shorter time frame than for the assay performed with CD3+CD28 cross-linking or for the mixed lymphocyte reaction (MLR). For cross-presentation, the authors measure CD69 up-regulation 24 hours after co-culture with antigen-loaded DCs, whereas for the anti-CD3/CD28 or MLR assays they measure T cell proliferation 3 days after their stimulation. In the published experiments that I have found examining the role of IDO1 activity in the regulation of T cell activation by MoDCs, the readout generally consisted in measuring T cell proliferation after 4 to 5 days of co-culture. It is not clear whether IDO1 expression by MoDCs would affect CD69 up-regulation in the soluble antigen cross-presentation assay used in the current manuscript.

2/ Necessity to rephrase some of the conclusions.
The following sentences are misleading, since the T cell proliferation was induced by TCR cross-linking and not upon co-culture with DCs. Hence, these sentences must be rephrased. Page 6, lines 18-19, “Taken together, epacadostat enhances T cell proliferation in presence of TLR-stimulated cDCs, while we did not find evidence that it interferes with DC maturation or their ability to cross-present soluble antigen”, should be edited as “Taken together, epacadostat enhances the T cell proliferation induced by TCR cross-linking in the presence of the supernant of TLR-stimulated cDCs […]”. Page 7, lines 19-20, “In our assays, IDO had a functional impact on T cell proliferation induced by TLR-stimulated cDC1s as well as cDC2s” should be rephrased as “In our assays, the supernatant from TLR-stimulated cDC1s as well as cDC2s had a functional impact on the T cell proliferation induced by TCR crosslinking”. Then, the next sentence does specify that this inhibition is dependent on IDO1 activity in DCs and speculates about its mechanisms of action.

Page 8, lines 2-3, “By using the IDO1 selective inhibitor epacadostat, we were able to increase T cell proliferation in conditioned medium of cDC1s or cDC2s, and in a classical MLR with cDC2s”. The end of the sentence (“, and in a classical MLR with cDC2s”) is incorrect and must be removed, because the data on MLRs, Figure 4B, does not show any significant effect of epacadostat with any of the DC types tested including cDC2s.

Minor points.

a) The authors should at least mention and discuss the selective high expression of IDO2 in human cDC1s, even if they do not assess it themselves which would the best.
IDO1 and IDO2 are paralog genes coding for enzymes with identical enzymatic activities but differential expression patterns. Yet, amongst human DC and other immune cell types, human cDC1s specifically express high levels of both IDO1 and IDO2. This was shown at steady state in a paper referenced in the current manuscript (30. Crozat, K., Guiton, R., Guilliams, M., Henri, S., Baranek, T., Schwartz-Cornil, I., Malissen, B., et al., Comparative genomics as a tool to reveal functional equivalences between human and mouse dendritic cell subsets. Immunol. Rev. 2010. 234: 177–198). This was confirmed both at steady state and after stimulation in another study (Balan S, Ollion V, Colletti N, Chelbi R, Montanana-Sanchis F, Liu H, Vu Manh TP, Sanchez C, Savoret J, Perrot I, Doffin AC, Fossum E, Bechlian D, Chabannon C, Bogen B, Asselin-Paturel C, Shaw M, Soos T, Caux C, Valladeau-Guilemond J, Dalod M. Human XCR1+ dendritic cells derived in vitro from CD34+ progenitors closely resemble blood dendritic cells, including their adjuvant responsiveness, contrary to monocyte-derived dendritic cells. J Immunol. 2014. PMID: 25009205). Thus, the effect of epacadostat observed could result from the inhibition of both IDO1 and IDO2. This needs to be stated in the manuscript.

b) Literature citation.
Whereas the literature is covered in a rather exhaustive and fair manner globally, some editions are required in a few instances.

Page 3, lines 49-50, “Interestingly, RNA profiling revealed high IDO expression of human steady-state cDC1s [30]–[32].”, please edit as ““Interestingly, RNA profiling revealed high IDO expression of human steady-state and activated cDC1s [30]–[32]” and add the following reference (Balan S, Ollion V, Colletti N, Chelbi R, Montanana-Sanchis F, Liu H, Vu Manh TP, Sanchez C, Savoret J, Perrot I, Doffin AC, Fossum E, Bechlian D, Chabannon C, Bogen B, Asselin-Paturel C, Shaw M, Soos T, Caux C, Valladeau-Guilemond J, Dalod M. Human XCR1+ dendritic cells derived in vitro from CD34+ progenitors closely resemble blood dendritic cells, including their adjuvant responsiveness, contrary to monocyte-derived dendritic cells. J Immunol. 2014. PMID: 25009205).

Page 5, lines 53-54, “In humans, all blood DC subsets are capable of cross-presenting soluble antigens [40], although cDC1s are superior at cross-presenting cellular antigens [41]–[43]”. Please add the following reference, since this paper was published back-to-back with reference 41-42, and together with them was the first study reporting a higher ability of human cDC1s to cross-present a cell-associated antigen as compared to cDC2s and pDCs (Crozat K, Guiton R, Contreras V, Feuillet V, Dutertre CA, Ventre E, Vu Manh TP, Baranek T, Storset AK, Marvel J, Boudinot P, Hosmalin A, Schwartz-Cornil I, Dalod M. The XC chemokine receptor 1 is a conserved selective marker of mammalian cells homologous to mouse CD8alpha+ dendritic cells. J Exp Med. 2010. PMID: 20479118).

Page 6, line 44, “Human cDC1s are considered to be the equivalent of murine CD8a+ DCs [44]”. To the best of my knowledge, the first papers to have provided strong evidence that human cDC1s are the equivalent of murine CD8a+ DCs are the following, first “Robbins SH, Walzer T, Dembélé D, Thibault C, Defays A, Bessou G, Xu H, Vivier E, Sellars M, Pierre P, Sharp FR, Chan S, Kastner P, Dalod M. Novel insights into the relationships between dendritic cell subsets in human and mouse revealed by genome-wide expression profiling. Genome Biol. 2008. PMID: 18218067) followed by another paper from the same group cited in the current manuscript (30. Crozat, K., Guiton, R., Guilliams, M., Henri, S., Baranek, T., Schwartz-Cornil, I., Malissen, B., et al., Comparative genomics as a tool to reveal functional equivalences between human and mouse dendritic cell subsets. Immunol. Rev. 2010. 234: 177–198). Hence, these two references must be cited in this sentence. Moreover, the 3 other papers on the same topics that were published back-to-back to reference 44 must also be cited, namely two references already included in the current manuscript (41. Jongbloed, S.L., Kassianos, A.J., McDonald, K.J., Clark, G.J., Ju, X., Angel, C.E., Chen, C.-J.J., et al., Human CD141+ (BDCA-3)+ dendritic cells (DCs) represent a unique myeloid DC subset that cross-presents necrotic cell antigens. J. Exp. Med. 2010. 207: 1247–1260; 42. Bachem, A., Güttler, S., Hartung, E., Ebstein, F., Schaefer, M., Tannert, A., Salama, A., et al., Superior antigen cross-presentation and XCR1 expression define human CD11c+CD141+ cells as homologues of mouse CD8+ dendritic cells. J. Exp. Med. 2010. 207: 1273–1281) as well as the fourth following reference (Crozat K, Guiton R, Contreras V, Feuillet V, Dutertre CA, Ventre E, Vu Manh TP, Baranek T, Storset AK, Marvel J, Boudinot P, Hosmalin A, Schwartz-Cornil I, Dalod M. The XC chemokine receptor 1 is a conserved selective marker of mammalian cells homologous to mouse CD8alpha+ dendritic cells. J Exp Med. 2010. PMID: 20479118).

Page 7, lines 8-9, “In addition, upregulation of IDO by stimulated cDC1s might serve as a negative feedback mechanism to prevent excessive T cell stimulation. This might explain why in some studies cDC1s were associated with a tolerogenic function [48],[49]”. The identity of the CD141+ cells studied in these papers should be considered with caution as they may not be cDC1s but rather cells of the monocyte/macrophage. Indeed, in reference [49] the human dermal CD141(high) cells studied also expressed CD14, a combination of marker not found on bona fide cDC1s but on inflammatory cells derived from monocytes. The study cited as reference [48] was mostly performed in mice where CD141 expression is not a marker of cDC1s. Moreover, the functional analyses in this paper were performed from mouse or human DCs derived in vitro from monocytes, and the CD141+ DCs characterized in mice expressed CD11b contrary to bone fide mouse cDC1s. Hence, to the best of my knowledge, no published study has yet reported a tolerogenic function of bona fide human cDC1s, whether IL-10 expression or the induction of Tregs.

Page 7, lines 36-39, “A subset of pDCs present in tumor-draining lymph nodes of mice has been reported to induce IDO-mediated T cell suppression, Treg differentiation and activation of resting Tregs [1],[2],[51]”. The identity of the IDO+ cells studied in these papers should be considered with caution. Indeed, the same authors later published that the putative CD19+ IDO+ immunosuppressive cells that they had identified in these studies and considered as a subset of pDCs were actually “B lymphoid cells” (Johnson BA 3rd, Kahler DJ, Baban B, Chandler PR, Kang B, Shimoda M, Koni PA, Pihkala J, Vilagos B, Busslinger M, Munn DH, Mellor AL. B-lymphoid cells with attributes of dendritic cells regulate T cells via indoleamine 2,3-dioxygenase. Proc Natl Acad Sci U S A. 2010. PMID: 20498068).

To the best of my appreciation, the authors have satisfactorily answered the comments from both reviewers.