We obtained the white blood cell count using an automated hematology analyzer (Sysmex KX-21N; Sysmex Corp., Kobe, Japan) to calculate blood volume necessary to obtain 5 × 10
5 white blood cells in each test tube. The red blood cells were lysed in 1% lysis buffer (Nordic Biosite AB, Täby, Sweden) for 10 minutes in the dark at room temperature. We then washed the cells three times by first centrifuging for 5 minutes at 500
g, decanting the supernatant, and then resuspending the cells in an isotonic buffer (BD FACSFlow; BD Biosciences, Franklin Lakes, NJ, USA). We then added marker-specific monoclonal antibodies to the sample tube and added fluorochrome-matched negative isotypes were added to a separately prepared tube (
Supplementary File S1). Tubes were incubated in dark at room temperature for 20 minutes, after which the cells were washed and resuspended in an isotonic buffer (BD Biosciences). Stained cells were analyzed using flow cytometry (BD Biosciences) with a sample size gated for 100,000 singlet leukocytes. We used analytical software (Kaluza version 1.5.20365.16139; Beckman Coulter Inc., Pasadena, CA, USA) for all flow cytometric analyses. Two independent evaluators (YS, MKN) performed all analyses completely blinded to each participant's condition and each other. CD4
+ T cells were identified and gated based on their CXCR3 and CCR6 expression (
Supplementary File S2). Zhang et al.
32 studied Th cell subsets in humans and found that CD4
+CXCR3
+CCR6
− cells had characteristics of Th1, CD4
+CXCR3
−CCR6
− had characteristics of Th2, and that CD4
+CXCR3
−CCR6
+ had characteristics of Th17. We also measured the CD4
+CXCR3
+CCR6
+ cells, which may reflect the Th1/Th17 cell subset (
Supplementary file S2).
27 Treg cells were identified as CD4
+CD127
lowCD25
high cells (
Fig. 1),
18 and Th subsets in Treg cells (Th-like Tregs) were determined (
Supplementary File S3).
27 Nonspecific signaling was eliminated using the corresponding negative isotype control at a threshold of 1%.