Abnormalities in iNKT cells are associated with impaired ability of monocytes to produce IL‐10 and suppress T‐cell proliferation in sarcoidosis

Sarcoidosis is a multisystem granulomatous disorder characterized by marked T‐cell expansion of T helper 1 (Th1) cells. The cause of T‐cell overactivity is unknown. We hypothesized that interleukin‐10 (IL‐10) production by a yet undefined cell type might be defective, resulting in loss of regulation of T‐cell activity. Focusing on IL‐10‐producing monocytes, we first showed that monocytes isolated from the peripheral blood of corticosteroid‐naïve sarcoidosis patients (n = 51) produced less IL‐10 compared to controls, and were less able to suppress T‐cell proliferation. In addition, monocytic IL‐10 production correlated negatively with disease activity score. As invariant natural killer T (iNKT) cells are known to both interact with monocytes and be reduced in sarcoidosis patients, we then asked whether iNKT‐specific defects might be responsible for this reduced IL‐10 production. We found that greater numbers of circulating iNKT cells was associated with higher IL‐10 production. Moreover, iNKT cells enhanced monocytic IL‐10 production in vitro. Defective IL‐10 production and T‐cell suppression by sarcoidosis monocytes could be restored following their coculture with iNKT cells, in a CD1d‐ and cell contact‐dependent process. We suggest that reduced iNKT‐cell numbers in sarcoidosis may lead to impaired monocytic IL‐10 production and unchecked T‐cell expansion in sarcoidosis. These findings provide fresh insight into the mechanism of sarcoidosis disease, and interaction between iNKT cells and monocytes.


Regarding the Abstract section:
a) From this text we don't know, where iNKT and IL-10 concentration were tested, in a blood or BAL? 2. Regarding the Introduction section: a) Authors should describe a role of iNK-cells in the context of immunity, especially in sarcoidosis. b) Authors should write about immune response in blood and in lungs because there is an opposite immunological response in peripheral blood and in the lung. c) IL-10 level in a peripheral blood from SA patients is increased in many studies, why in the current study it was decreased? In some studies, a low production of IL-10 was present also in TB.We know that increased and chronic TNF-α and IL-6 production induces a secretion of anty-proinflammatory IL-10. Comments to the Author Major comments: 1. Based on the methods section, there appears to be an important typo in the second paragraph of the results section where there is reference to "negative selection" of monocytes. Please clarify as whether the study involved negative or positive selection seems critical to this study……the text in question is as follows: "We next questioned if IL-10-producing monocytes suppressed T cell proliferation. PBMCs were isolated from healthy individuals (n=12) and CD14+ cells were first removed using CD14 MACS bead positive selection system to provide a monocyte-free cell system for subsequent addition of fixed numbers of monocytes. These peripheral blood lymphocytes (PBL) were then CFSE-stained, and allogeneic monocyte-derived DCs were added [1:4, DC:PBL]. Autologous CD14+ monocytes (generated using CD14 MACS beads negative selection) or CD19+ B cells as control [1:1, monocytes (or B cells):PBL] were then added." 2. This reviewer has concerns regarding the authors' composite score. CXR manifestations in sarcoidosis connote severity of disease, not necessarily activity of disease. BAL lymphocytosis is an accepted measure of disease "activity". Also, the authors' composite measurement has not been validated which they state in the text: "no validated universal score currently exists" so why do the authors think that their score/measurement is the "best"? Further, how did the authors arrive at the score cut offs between low and high disease activity? Why not use the median value instead?
3. The text in the results section for Fig 4A is misleading in current form and should state what is stated in the legend. Specifically, that there is NO correlation between iNKT cell frequencies vs. IL10 concentration in sarcoidosis subjects. 4. In the discussion, the authors should acknowledge the use of an NKT cell clone and its limitations in generalizing the findings without replication with ex vivo iNKT cells.
Minor comments: 1. Consider changing to TNF-α and IFN-γ rather than what appears in the text which is TNF -a and IFNg when referring to TNF and IFN 7. Is reference #35 correctly cited in the text related to NKT cell clones?

Figure legend 3B. This is not technically a correlation
First Revision -authors' response -6 December 2013

Point-by-point response
Reviewer: 1 Comments to the Author The manuscript is very interesting but it needs some improvements: Important question is: if the Authors repeat the same study using CD14+CD16+ monocytes, they will obtain a low level of Il-10? Probably not. It is well known that in contrast to CD14+CD16+ monocytes, CD14+CD16-produces a low level of IL-10.
The study was performed by MACS bead CD14 negative selection kit. This kit isolates CD14+ cells without touching CD14+ cells. We optimised this kit with Miltenyi in order to include CD16+CD14+ monocytes (the original kit had CD16 mAbs which meant that CD16+ cells including CD14+CD16+ monocytes would have been depleted). Thus our monocytes contained both CD16+ and CD16populations. Hardly any of the CD16+ monocytes produced IL-10 after LPS stimulation (we showed this in Figure 1c). These are all freshly isolated monocytes used immediately (ie not thawed from frozen). We have made this clearer in view of this reviewer's comments.
1. Regarding the Abstract section: a) From this text we don't know, where iNKT and IL-10 concentration were tested, in a blood or BAL?
Apologies, these are from bloodthis is now in the abstract.

Peer review correspondence
Yes, agree -we didn't put it in originally to keep the introduction succinct and targeted but we agree it would be informative. Now added in (page 3). c) IL-10 level in a peripheral blood from SA patients is increased in many studies, why in the current study it was decreased? In some studies, a low production of IL-10 was present also in TB. We know that increased and chronic TNF-α and IL-6 production induces a secretion of anty-proinflammatory IL-10.
Most studies refer to IL-10 levels in bronchoalveolar lavage and even then the levels were low, and almost no studies compared sarcoidosis patients with healthy controls (mainly between acute and chronic). Data on IL-10 in peripheral blood in sarcoidosis is conflicting with some reports of it being reduced and others increased when measured directly in whole blood or serum in which case, the source of the IL-10 is unclear/ multiple. We are not commenting on serum IL-10 levels, rather the IL-10 production by a subset cells (monocytes) in the peripheral blood. Thank you. We were particularly interested in the monocytes in the circulation and not in the BAL because we were focusing on the upstream behaviour of monocytes before these cells migrate to the lungs or other organs. This is to provide us with a means of potentially addressing altered behaviour in these monocytes before they differentiate into macrophage or contribute to granuloma formation. The phenotype of monocytes in the peripheral blood was extensively studied, with particular focus on identifying surface markers/ other phenotypic characteristics for the IL-10 producing subset. None of the following expression: CD11b, CD11c CD14, CD15, CD16, CD32, CD62l, CD115, CD163, CCR2, CX3CR1 and HLADR were increased or reduced in the IL-10 producing subset (Figure 1c). c) Why have the Authors tested subset of monocytes CD14+CD16-CD206-CD115+CD15-, not nonclassical CD14+CD16+, which increased occurrence is characteristic for sarcoidosis and produces increased level of IL-10? Maybe it is a reason of obtained low level of IL-10.
We have tested both CD14+CD16+ and CD14+CD16-, ofcourse! They are in the same panel of mAbs we used to examine whether the there are any markers that would help us identify the IL-10 producing monocytes (see figure 1c). Sorry this is not clearwe have made it clearer by explaining that we have used these mAbs to determine it the IL-10 expressing monocytes expressed any of these markers. As per response to comment (1), hardly any of the CD16+ monocytes produced IL-10 after LPS stimulation.
d) The Discussion section is too long and it is not concrete; there are a lot of hypotheses Agreehave cut down.

Reviewer: 2
Comments to the Author Major comments: 1. Based on the methods section, there appears to be an important typo in the second paragraph of the results section where there is reference to "negative selection" of monocytes. Please clarify as whether the study involved negative or positive selection seems critical to this study……the text in question is as follows: "We next questioned if IL-10-producing monocytes suppressed T cell proliferation. PBMCs were isolated from healthy individuals (n=12) and CD14+ cells were first removed using CD14 MACS bead positive selection system to provide a monocyte-free cell system for subsequent addition of fixed numbers of monocytes. These peripheral blood lymphocytes (PBL) were then CFSE-stained, and allogeneic monocyte-derived DCs were added [1:4, DC:PBL]. Autologous CD14+ monocytes (generated using CD14 MACS beads negative selection) or CD19+ B cells as control [1:1, monocytes (or B cells):PBL] were then added." For this sentence '..We next questioned if IL-10-producing monocytes suppressed T cell proliferation.
PBMCs were isolated from healthy individuals (n=12) and CD14+ cells were first removed using CD14 MACS bead positive selection system to provide a monocyte-free cell system for subsequent addition of fixed numbers of monocytes…', the CD14+ cells have to be removed from the PBMC in order that the there are no monocytes in these cells (the 'monocyte-free cell system'). This necessarily **has** to be done by CD14 positive selection. These monocytes are then discarded. Then, to add in a controlled and standard number of monocytes, we derived allogeneic monocytes using CD14 negative selection and added this into the monocyte-free cell system, as in this part of the statement "…Autologous CD14+ monocytes (generated using CD14 MACS beads negative selection) or CD19+ B cells as control [1:1, monocytes (or B cells):PBL] were then added." 2. This reviewer has concerns regarding the authors' composite score. CXR manifestations in sarcoidosis connote severity of disease, not necessarily activity of disease. BAL lymphocytosis is an accepted measure of disease "activity". Also, the authors' composite measurement has not been validated which they state in the text: "no validated universal score currently exists" so why do the authors think that their score/measurement is the "best"? Further, how did the authors arrive at the score cut offs between low and high disease activity? Why not use the median value instead?
Yes, apologies -in retrospect we agree that we should not have used the word 'best' because indeed validation of this has not been published. We have however, since validated this against BAL lymphocytosis, sol-IL2r levels, and abnormalities on high resolution CT scan; but this has not been published yet. So, we have changed the word 'best' to ' a defined and standardized' quantification of disease activity. The high and low score cut offs were arbitrary. We have now put in the whole range of disease activity score in order to remove the uncertainties over the definition of high and low activity. This showed a highly significant correlation between proportion of IL-10+ monocytes (and IL-10 levels in culture supernatant) and the disease activity score.  the table or get rid of the table and put them in the text, or   the figure legends?) Apologies, yes agree. Fig 2A's reference in Table 2 was a typothis should state healthy control. We have now kept the table simple and shown demographics for figures where cells were derived from sarcoidosis patient and compared to healthy controls only. For the rest, we have placed the sample sizes in the legend. 5. On pg 10 of results, the authors state that "only CD14+ cells (monocytes) in the co-culture expressed IL-10, implicating these cells as the sole source of IL-10 in the supernatant". But didn't the experimental design consist of sorting out CD14+ monocytes from 36 controls and culturing them with NKT clones? So are you saying that the NKT clone was not the source of IL-10? If this is not correct, please considering clarifying the sentence.
Yeswe sorted out CD14+ cells and cultured with iNKT clones, and when we examined IL-10 expression by FACS in these cells, only CD14+ cells expressed IL-10. This means that iNKT clones is not the source of IL-10. 8. Figure legend 3B. This is not technically a correlation Thank you, agree. In fact figure 3B and 3C are now modified to remove the arbitrary division to low and high activity. We have amended legend accordingly.